Óïðàâëåíèå áèçíåñîì
Ïðîåêòèðîâàíèå (GEP)

GMP ïî ðàçäåëàì:


Êîíòðîëü êà÷åñòâà
Àóòñîðñèíãîâàÿ äåÿòåëüíîñòü
Òåðìèíîëîãèÿ GMP
Ýêîëîãè÷åñêèé ìåíåäæìåíò
Î ïðîåêòå

Äëÿ ñîäåðæèìîãî ýòîé ñòðàíèöû òðåáóåòñÿ áîëåå íîâàÿ âåðñèÿ Adobe Flash Player.

Ïîëó÷èòü ïðîèãðûâàòåëü Adobe Flash Player


1. Introduction 1
2. People: Leadership, Visionaries, Acknowledgments, and Awards 5
3. Organization 49
4. Process Safety 65
5. The Environment 87
6. Regulatory Affairs: Meeting the U.S. Food and Drug
Administration (FDA) Requirements 109
7. Patents 141
8. Chemical Engineering 165
9. Excursions in the ?-Lactam and Steroid Fields 203
10. Case Studies 267
11. The Future 317
Author Index 389
Subject Index 401

Several very useful books on the subject of chemical process development have been
published.1 These have been written largely from the point of view of the bench
chemist or chemical engineer. Emphasis in this collection of books is on the work
needed to ensure that practical chemical reactions are created for scale-up, that the
chemistry is understood, that the theory and mechanics needed to engineer scaleup
are addressed, and that Safety, Environment and Food and Drug Administration
requirements are met.
This book is about the management of the people, organization, and the main
disciplines which have to be to be integrated to create and develop a chemical process
to meet all the needs.
Management recognizes that people are the most important assets in their organization
and that inspiring leadership provides the best driving force for success.
The major requirements for such leadership are reviewed. In today’s pharmaceutical
industry, leaders need to be visionaries with the ability to motivate their scientist
and engineer co-workers to express themselves, to take risks, and to harness sound
judgment in fusing together the many components that form a chemical process.
Personal examples are used throughout the book to illustrate this. A few of the
1(a) Lee, S., and Robinson, G. Process Development, Oxford University Press, Oxford, 1996. (b) Repic, O.
Principles of Process Research and Chemical Development in the Pharmaceutical Industry, John Wiley
& Sons, New York, 1998. (c) Process Chemistry in the Pharmaceutical Industry, Ed. Gadamasetti, K.
G. Marcel Dekker, New York, 1999. (d) Anderson, N. G. Practical Process Research and Development,
Academic Press, New York, 2000. (e) Griskey, R. G. Chemical Engineering for Chemists, American
Chemical Society, Washington, D.C., 1997. (f) McConville, F. X. The Pilot Plant Real Book, FXM
Engineering and Design, Worcester MA, 2002.

Copyright C 2008 John Wiley & Sons, Inc.

frameworks through which people are recognized and rewarded for their achievements
are described. People recognition and rewards are undertaken in partnership
with the company Human Resources function.
Organization of the work of scientists and engineers and how this is integrated
with other disciplines to provide the foundations for success in achieving defined
missions is outlined. It is recognized that organizations need to be flexible and be
prepared to change to meet the unexpected and also the different needs of different
The main “outside” disciplines influencing the progress of chemical process development
in the pharmaceutical industry are process safety, environmental considerations,
and FDA regulatory affairs. The basic principles governing these disciplines
and the major activities needed to meet the requirements in these areas are summarized.
Beyond the regulatory disciplines, the vital importance of patenting and
defending intellectual property is also emphasized. An outline of the chemical engineer’s
role in chemical process development is given with particular emphasis on
chemical plant equipment requirements for the major unit operations.
Two case studies are provided to illustrate how the work of chemical process
development is carried out and how this work is changing with time. Two essays
describing technical excursions in two of the major fields I worked in, ?-lactams
and steroids, place chemistry in a historical perspective and provide a picture of the
excitement and variety of challenges that come with a career in chemical process
The final chapter, on the future, provides a personal summary of a few of the
major endeavors I believe should be pursued in order to address today’s realities,
including the consequences of past neglect. These endeavors require that we raise
education—in our case, chemistry education and in particular its integration with
the analytical, biological, and engineering sciences—to a much higher level of importance.
They include finding ways to overcome the rising monster of intrusive
regulation; to address the consequences of outsourcing; to increase the use of biological
systems in synthesis; to simplify and contain chemical processes; to promote
evaluation of newer technologies and reexamine some old ones; and to prevent and
reduce waste. Preparing for the future also requires that all thinking people need to
fantasize, in our case to stimulate debate on how the major chemistry challenges
in the world should be tackled. Such debates must lead to the creation and funding
of feasible programs—I offer one “starter,” tongue-in-cheek fantasy of my own. By
promoting new chemistry-based thinking, we might breathe new life into the old
DuPont slogan “Better things for better living through chemistry,” with the twist that
“chemistry” be defined in the broader interdisciplinary context referred to above.
This book draws on my own experience and observations from over 10 years of
working at the bench and over 30 years growing through the management ranks in
chemical process research and development, the last 14 at the vice-presidential level.
The book is thus a summary of the work of many co-workers, to whom I am forever
indebted, and is written in the hope of stimulating others to create new futures.
Chemists and engineers joining chemical process development organizations
quickly recognize that although we grow from our roots in chemistry or engineering,
we need to adapt quickly by embracing and incorporating all manner of inputs, sometimes
unforeseen, into our work. We have to adapt to the turbulence that goes with
practicing chemistry in the real world of tackling often urgent problems in R&D, in
manufacturing and in pertinent business areas. Thus we have to accommodate the
needs of government, secure intellectual property, and aid marketing, sales, finance,
law, and so on, at the same time as providing supplies and information in order to
bring new drugs to the market place as quickly as possible. The practical combination
of these activities creates the life of a company more or less under the rule of
imperfect and changing laws.
The chapters in this book started out as handouts for a series of talks prepared for
students of chemistry interested in the possibilities of a career in the chemical process
development field. Some were also presented to my manufacturing colleagues
at Schering–Plough. The chapters are based on the work carried out during my employment
at several pharmaceutical companies (Arapahoe Chemicals/Syntex, Glaxo,
Bristol–Myers, and Schering–Plough) in both the R&D and manufacturing areas.
This diversity of experience enabled me to appreciate the need to accommodate the
different objectives and philosophies that drive each company, and frequently divisions
within companies. Add to this the iterative nature of the drug development
field and one soon understands the need for flexibility in progressing the work of
any organisation. Above all, it is worth repeating that success in any organization is
dependent on well-equipped people working together in a creative and disciplined
environment to address the common need. People are the key. Creative individuals,
working collaboratively in a team, which accommodates a little heresy, are more
important than buildings, machinery, budgets, balance sheets and bureaucracies, and
all the other components of any endeavor.
Although the core professional discipline in chemical process development is
chemistry, success in finding the best chemistry to develop to a pilot plant and manufacturing
scale is dependent on many factors and disciplines. In a chemical process
development department that is part of a pharmaceutical research organization, the
mission to produce the active pharmaceutical ingredients (APIs) and intermediates
needed by one’s research colleagues for their work to identify new drug candidates
is the first priority. The early API supply mission usually comprises using research
chemistry, often in a raw state (I refer to this as the Recipe stage), to produce needed
supplies. To meet further urgent (usually larger) API needs, the Recipe stage evolves,
for safe scale-up, into the Method stage. As the likelihood increases that a potentially
marketable API is emerging, the chemical process development department works to
cultivate a deeper understanding of what is needed to create chemical transformations
that are practical and broadly acceptable, in safety, environmental, regulatory, and
economic terms. This begins the real Process Development phase of a project. In this
phase, one needs to give thinking people in the immediate organization—especially
chemists, analytical chemists, and chemical engineers—increasing “space” to express
themselves in building the research transformations, or new ones they can predict
will be better, into the beginnings of a process.
As the momentum in this direction increases, the disciplines of chemical engineering,
of patents, and of the regulations which guide process development work
(safety, the environment and FDA regulatory affairs) become increasingly important.
In addition, one needs to seek the input of the manufacturing people in creating the
manufacturing process and, as the project develops, to assist in process design and
the implementation of a system of operations suited to the ultimate manufacturing
process and manufacturing site. Integrating the sometimes seemingly conflicting activities
of API supply with chemical process research and development inevitably
creates a chaotic environment. However, chaos can be dealt with through proper
staffing and with agreed prioritizations. In my mind the process that develops from
integration of these activities is better than one that develops by separatingAPI supply
from process research and development. The simple reason for this is that gaining
experience in the overall system enormously enhances the ability of scientists and
engineers to see what is really needed in generating a manufacturing process.
This book is intentionally broad in scope. I recognize that some chapters may
lack in depth, but I hope the collection will provide readers with human perspective
on what is involved in chemical process development. I am aware that there are
omissions, such as to the broad uses of computers and applications of statistics,
which may intensify concealment of their value in developing chemical processes.
I therefore urge practitioners to consult with their leaders for guidance on questions
regarding other disciplines to accommodate in progressing their work.
The final reality is that every one of us working in chemical process development
could write a different book drawing on their personal experiences. It would move
the field along to a greater state of appreciation and understanding if more of us did.

The right people are the most important assets in any organization.

The major factors I wish to address in recognizing the vital importance of people are
leadership, the influence of visionaries, outstanding scientists and engineers, the value
of consultants, and the recognition of the achievements of people through awards and
a scientific/engineering ladder of promotion.
Organizations strive for success in their chosen businesses. To achieve success,
nothing is more important and complex than finding, organizing, and keeping the
right people to work in it and creating the environment for them to express their
talents. The right people share the goals of a good organization and believe they are
in a good place to meet their own needs. The leaders in the organization are, for
their part, in general agreement with this assessment, especially in recognizing that
both parties need to work to sustain their relationship and to accommodate changing
The right people come from all walks of society, embracing everyone from the
most gifted professionals to the cleaners. Understandably, it is visionaries and leaders

Copyright C 2008 John Wiley & Sons, Inc.

and those who generate the successes who receive the most attention and publicity.
However, it is vital that everyone understand that achievements also owe much to
those working in the lower ranks of the organization, not forgetting those outside the
organization who provide support, including families at home. All have an influence
and need to feel that their contributions are appreciated.
Although this presentation is concerned with people in chemical process development
organizations in the pharmaceutical industry, there is much that is applicable
to people in almost all industries. First, it is worth placing people in the context of
the most important element in an organization, leadership, recognizing that infinite
variations are needed to suit infinite circumstances. Leadership sets the tone, evolving
as objectives change.
Textbooks and educational courses may provide the principles of leadership, but
it is human application and successes that identify the leader. Leaders are people
who need to take responsibility for running an organization, at the same time as
accommodating factors beyond their control.
In the scientificworld it seems obvious that leaders in a given area should be highly
qualified (or, rarely, just very, very experienced) in the major discipline they are leading
and that they should understand the importance of related disciplines. In chemical
process development a highly trained chemist leader needs to have experience in areas
such as chemical engineering, biological sciences, and analytical sciences. Leaders
of chemical process development may also come from these other sciences, provided
they have the talent and supporting people to uphold their leadership.
Leaders need many abilities:
 The ability to identify the people needs of the organization and also to find,
attract, develop, and keep real talent. It is not enough to find someone for one
immediate kind of work. One may need a specialist, but such a person in today’s
fast-moving risk-taking technical world must be able to adapt to changes and
challenges that stretch his/her specialization and imagination. The final judges
in the selection process need experience, and sometimes even an instinctive
feel, in choosing their co-workers. It is necessary to ensure good mentoring and
training to develop one’s people resource over time. In the course of such a
process, future leaders are identified.
 The ability to delegate and trust. These are important requirements in pursuing
any endeavor. At the same time, especially early in a relationship, one generally
needs to remain “unobtrusively interested” (e.g., through project review
meetings) until progress reveals that the trust is well-placed.
 The ability to be flexible and to act to correct one’s failures on the one hand as
well as to selflessly represent outstanding people on the other. Leaders who fail
to deal with poor performance do not inspire their subordinates. Leaders who
neglect superior talent or hog their credit do a disservice to the organization,
and ultimately to themselves. Leaders need to recognize and reward outstanding
ability. Salary is only one way. Organizational ladders of professional growth
equal to managerial ladders is another. An awards system (see later) is yet
 The ability to listen, communicate, promote action and collaborate, clearly, on
the issues in a wide variety of situations. Each issue may require its own minimission
statement, worked out by the principals to define a needed objective,
within the constraints of other commitments, and to marshal the resources to
meet it. Given such definition, motivating the players needs enthusiasm and
resolve and as good a grasp of the problems as can be mustered. This can
be extraordinarily difficult if there is great uncertainty regarding the facts, or
competing demands. Nevertheless, shrewd risk-taking needs to be encouraged,
and, if unsuccessful, responsibility needs to be accepted. Keeping a wise focus
on the essentials, including thorough project reviews, is often vital to success.
 The ability to promote the scientific/engineering dialogue and project vision at as
high a professional level as is feasible, or appropriate. Scientists and engineers
are usually very good at responding to technical challenges in an adventurous
way, but wise counsel may occasionally be needed to avoid projects drifting
far from addressing the core problem—still allowing that there is a chance
for a maverick solution! The scientific/engineering dialogue extends beyond
chemical development to require interactions with other disciplines, including
pharmaceutical sciences and regulatory affairs, and it is in accommodating these
interactions that listening ability, wisdom, and vision are most needed.
 The ability to succinctly and modestly keep one’s own superiors abreast of issues,
progress, setbacks, and individual contributions. In this arena, one needs to
accommodate (although not necessarily always accept) the thoughts and advice
of those with greater perspective.
 The ability and courage to deal with project failure, usually without entirely
abandoning the fight to salvage something useful. Few events are more difficult
to handle, especially if one has been personally committed. Mourning is brief
for leaders since they need to take stock of the realities, reassess the facts,
dissolve project teams, and redeploy resources on new initiatives. Leaders give
credit for achievements in failed projects and encourage appropriate use or
publication of worthwhile findings. Another positive is that failures give leaders
the opportunity to show they care for individual workers.
 The ability to continually adapt to an increasingly problematic regulatory world
and persevere in efforts to improve operations and to deal with the bureaucracy.
Governments have, quite properly, reacted to the overly self-serving activities
of some companies and individual entrepreneurs by creating strict rules of governance.
Since breaching the rules leads to regulatory problems and causes
business delays, industry has reacted by creating internal compliance groups
to avoid such problems. Compliance groups, striving to help their company be
“whiter than white,” have set up internal controls and bureaucracies that, unfortunately,
further stifle creativity and change. As a result, in the pharmaceutical
industry, process development chemists and engineers are obliged to define an
industrial process for producing an active pharmaceutical ingredient (API) at
the earliest possible development stage. Freezing or minimizing change, at say
the IND filing stage, until the NDA has been approved by the FDA has greatly
inhibited the creative drive for better processes, if not for new products. Given
that rules impact on all phases of development and that the development phase
of bringing an API to the market is the most costly phase, it is inevitable that
if creative drive continues to be inhibited, the cost of drugs to the consumer
will continue to be high. Thus, rules, lawyers, relentless media attention, the
remorseless and often short-term demands of the financial markets and their
analysts, and the increased politicization of the alleged obscene profitability of
the pharmaceutical industry, at least in the United States, make for a difficult
 The ability to work for the love of it, as if the company is your own. This
is generally an inspiration to all around you. Such a commitment requires
a complex combination of qualities, notably a personal passion for the job,
wisdom, aggression, humility, creativity, a sense of humor, obsession, relentless
drive, occasional ruthlessness, and the ability to stay hungry, inter alia. People
working for the love of it generally have a passion to promote excellence.
A continuous search for leaders is a vital part of every company’s mission. The
following statement1 by Charles D. Miller, Chairman and CEO of Avery Dennison,
is illustrative:
My personal specifications for successful leaders are very simple. I look for people
who possess the character to succeed in a highly competitive environment; who have
the courage to take risks; who speak candidly and with confidence; who exercise good
judgement, often with little information; who think creatively and inventively; and who
have a community spirit to work collaboratively in a team-supported environment. One
of our most important challenges today is to nurture and develop our next generation
of leaders who will be successful in diverse global environments and who will, in turn,
develop other leaders to capitalize on the Company’s many strengths.
In conclusion, leadership has never been more needed, in every area, to overcome
situations and inertias that take an inevitable toll on the competitiveness of the
advanced nations (see Chapter 11).
It is worthwhile for all of us to look back and reflect on the individuals who really
made a difference to our professional careers. It usually begins with supportive parents
and inspiring teachers, enabling one to emerge from university with the knowledge
and certificates that are the tickets allowing you to travel. Once “on the road,” it is
up to you and to all the professionals around you. In most respects you find these
professionals yourself in joining companies of people whom you feel are of like mind
and whom you can convince would benefit from employing you.
Although most of the legion of people who made a difference to my own career are
little known, except through their scientific papers and local recognition, it seemed
1Avery Dennison Annual Report to Stockholders, March 1, 1995.
to me worthwhile to introduce the most influential ones to you. These are the people
who illustrate particular abilities needed to succeed in chemical process development
projects. Perhaps these “sketches” will encourage readers to reflect on corresponding
people in their own careers.
Of the many people to whom I reported, I found only a few to be exceptionally
visionary and brilliant leaders. Five were the sort of leaders anyone would be privileged
to work with; the sixth was more of a maverick superbly suited to particular
situations and circumstances. While the visionaries were indispensable to all our
successes, it was the hundreds of scientists and engineers who I had the good fortune
to work with, and whose sustained technical achievements over many years created
the chemistry and engineered the processes, who provided the company with benefits
and breakthroughs. In completing this section of the presentation, I pay tribute to
several of our consultants and particularly to three professors who consulted for us
over long periods and who proved particularly inspiring.
These are the people who generally see, as part of their professional brief, that
there must be opportunity for revolutionary as well as evolutionary approaches to
“business” creation, development, and improvement. They have ideas of their own
but are open to outside stimulation and willing to run with the ideas of others.
Visionaries recognize the importance of giving talented people their head. In our field
they encourage and support such people in their scientific enterprise and quest for
scientific understanding. They are willing to give talented people time and resources
and willing to beat back naysayers and senior managers who all too often call for
short-term solutions or strict adherence to organizational boundaries. Visionaries
believe in their people, they tolerate a little heresy, they possess personal courage and
have the good judgement to know how far “vision” can be taken. Visionaries by their
enterprise often acquire more than their normal fair share of luck and, as a result, are
often responsible for many of the great advances in anything.
In my experience, process technology is advanced significantly under such leadership.
This leads me to the people who, in the periods indicated, contributed so much
to my own career.
Drs. Tom and Richard Waugh (1960–1966). These exceptionally adventurous and
courageous brothers, together with an engineer, Oscar Jacobsen, raised the capital
to found Arapahoe Chemicals in Boulder, Colorado, simply because they wanted
to work there (rather than continue working with Standard Oil in Gary, Indiana).
They perceived Boulder as a better place to raise their families, and they needed a
workplace environment in which they could better express their technical abilities.
In founding the company, much thought went into tapping the most singular
quality of the Colorado climate, its dry air. This led them to the production and sale
of Grignard reagents and later other metallo-organics. They were willing to tackle
all manner of hazardous chemical reactions, some of which led to fires and the loss
of physical plant. The insurance money enabled them to learn from mistakes and
rebuild. In my time I recall the rupturing of a bursting disc following a runaway
Grignard reaction—a large quantity of ethyl chloride had been added to slowly
activating magnesium. A spurting jet of ethyl magnesium chloride blew onto an
aggressively sited MacDonald’s hamburger stand. Tom and Dick took the whole
affair very seriously, paying for the cleaning and repair of damage to customer cars.
But they couldn’t gag the jokers who suggested that the hamburgers never tasted so
Arapahoe won the respect of major customers around the United States, not only
for the custom work done for them by Arapahoe, but by reacting to quality issues in a
fundamental way. Thus, by becoming aware of the instability of the N-bromoamides
they made for others, particularly in the steroid industry, they continually improved
and patented2 their processes thereby producing stable N-bromoamides which became
another foundation of Arapahoe’s business. The culmination of this work was a
process2c wherein a solution of the amide in a cold (5–15?C) freshly prepared solution
of HBrO3 was treated with bromine to give the N-bromoamide. The key step was to
form HBrO3 by passing a concentrated solution of NaBrO3 through a column of a
strong acid resin (Dowex 50W-X8). Bromide ion produced in the bromination was
reoxidized to bromine. The process was particularly useful for the preparation of the
relatively unstable N-bromoacetamide.
Product purity became a passion at Arapahoe Chemicals, as well as a formidable
marketing tool. It became an unwritten trademark in all of Arapahoe’smarketed products,
including DDQ, organic scintillators, numerous pharmaceutical intermediates,
and metallocenes.
The scientific environment at Arapahoe Chemicalswas stimulating and successful.
Tom and Dick supported scientists in their efforts to further their education through
course work at Colorado University and by encouraging dialogue and consulting
sessions with several of the chemistry departments professors. Their leadership and
family orientation as employers owed much to their commitment to the company,
their love of their jobs, their sense of purpose, their energy and enthusiasm, and their
willingness to accept difficult projects and to listen to everybody’s ideas for solutions.
Not surprisingly, they attracted entrepreneurial people to the company. They
also established a strong business/science culture. This was always evident at our
frequent open-ended project reviews in which the responsible scientists presented
their project work, fielded questions, ideas, and suggestions, and made appropriate
accommodations in presenting an ongoing course of action. In the scientific arena
we accomplished a great deal, even if it seemed small in the greater scheme of science.
Tom and I made a useful contribution to the organic scintillator field with the
invention of dimethyl-POPOP, a commercially successful more soluble successor
to the original organic scintillator, POPOP.3 We created practical chemistry, with
2(a) Waugh, R. C., and Waugh, T. D. U.S. Patent 2,971,959, 1961. (b) Waugh, R. C., and Waugh, T. D.
U.S. Patent 2,971,960, 1961. (c) Robertson, D. N. U.S. Patent 3,187,044, 1965 (to Arapahoe Chemicals,
3Walker, D., andWaugh, T. D. J. Heterocyclic Chem., 1964, 1, 72. Dimethyl-POPOP is still on the market,
40 years after its invention.
Dr. Bill Coleman, for several chemical steps in Syntex’s synthesis of the oral contraceptive
chlormadinone. With Haldor Christensen, sodium dispersion chemistry led
to a superior process for the manufacture of the Eli Lilly herbicide, diphenamid. We
devised novel patented chemistry, with the inspiration of Dr. Martin Hultquist, for the
manufacture of DDQ. The list could go on and on, but the essence is that in Arapahoe
we became chemical process development chemists. We learned that there were no
such chemists as steroid chemists, organometallic chemists, heterocyclic chemists,
and so on. There are only process development chemists, capable of synthesizing
anything. Being scientists in a small company we also learned to accommodate other
disciplines and business requirements in creating our chemical processes.
As a result of its successes, Arapahoe Chemicals became a takeover target for
Syntex. Once taken over, the ensuing changes disturbed the magic of the original
company. It was not the same and many of us moved on. But all of us owed a debt to
the genius and vision of Drs. Tom and DickWaugh. I built on this unique experience
for the rest of my career.
Dr. Arthur Best (1966–1975). Moving to the penicillin and fledgling cephalosporin
production facility of Glaxo Laboratories in Ulverston, Lancashire, introduced
me to the more structured rigors of the pharmaceutical industry. The Ulverston
factory synthesized chemical intermediates and APIs as well as many dosage
forms for the marketplace. The move from working in a small, fast-moving, freewheeling,
all-encompassing, practical chemistry organization to heading the chemistry
component of a large process investigation department came as an immense
shock. The chemical process challenges were enormous, but the whole thrust of
the department—troubleshooting and improving existing processes with limited
resources—severely restricted the opportunities for real process understanding, redefinition,
development, and improvement. Itwas clear that we needed process revolution
as well as evolution.
It was fortunate for Glaxo, as well as myself, that Dr. Arthur Best was the technical
director of the Ulverston factory at the time and, moreover, that he subscribed to the
view that only people on the ground in Ulverston could do the process development
and process troubleshooting work he thought was needed. He saw that the process
research and development people in Glaxo, Greenford, were much too involved in
serving research needs for clinical supplies of the company’s new APIs to have the
time and effort to provide the dedicated technical power needed for all the process
evolution/revolution opportunities in Ulverston. They were also far away and did not
have the laboratory space to enable them to increase staff to meet the needs. He also
perceived a conservatism in the Greenford process development department. Thus in
selecting and developing a second process4 for the manufacture of cephalexin,5 the
Greenford development group opted to develop Eli Lilly’s chemistry in the belief that
4The first process, which was already in production in Ulverston (and, in part, in Montrose, Scotland),
utilized the 2,2,2-trichloroethyl (TCE) group for the protection of the carboxyl group in the starting
penicillin G sulfoxide acid; for more detail of the need to change, see Chapter 7.
5Eli Lilly was the discoverer of cephalexin. They used p-nitrobenzyl (PNB) protection of the penicillin
carboxyl group in their manufacturing process. Glaxo had rights to this process, as well as to market
thiswould speed change to a newprocess inUlverston and Montrose.We inUlverston
argued that the Eli Lilly chemistry was undesirable for safety and environmental
reasons.6 To the chagrin of some in the Greenford process development group, Dr.
Best encouraged and supported (by approving the conversion of existing and available
space in Ulverston to laboratories and adding scientific manpower and equipment)
my proposal to explore and develop the DPM alternative to the PNB group despite
enormous risks to himself.
Dr. Best’s initiative set in place an unprecedented and competitive collaboration
between the Greenford and Ulverston process development groups. This was administered
through frequent technical review meetings in Greenford. Greenford concentrated
on developing the chemistry to use the PNB group while we in Ulverston set
about proving that use of the diphenylmethyl (DPM) group would give cephalexin
yields equal to those obtainable via use of PNB and also generating the information
to prove that the DPM group offered a safer, more environmentally friendly option.
Making the choice between the two protecting groups was accelerated by a letter
received from Ciba pointing out that Glaxo’s use of the TCE group was covered by
a Woodward patent to Ciba. The final selection between PNB and DPM was made
at a meeting in Greenford. Dr. Best’s position, based on the equivalence of yields,
cephalexin product quality, and the equal state of advancement of the two processes,
was that it was unacceptable to introduce the Lilly-patented PNB process (despite our
NRDCrights allowing us to use it) versus the Ulverston, Glaxo-patentedDPMprocess
when the Lilly process introduced so much more in the way of hazard and waste.
We argued that the use of p-nitrobenzyl bromide, a proven vesicant, in introducing
the PNB group and the hazardous waste produced in removing it were undesirable
burdens in amanufacturing situation. In addition, cost calculations showed amarginal
advantage in favor of using DPM protection. The decision to adopt the DPM process
was made by Glaxo senior management after the technical meeting.
During the nine years I worked with Dr. Best he regularly demonstrated that an
eloquently argued, well-supported case would generally overcome a weaker case,
however passionately argued.
Dr. Robert A. Fildes (1975–1980). Bob Fildes was one of the most dynamic and
controversial people I ever had the pleasure to work with, as a colleague in Glaxo
(1968–1974) and in Bristol–Myers. As a biochemist in Glaxo, he saw the immense
opportunities to be gained through “neutralizing” the amino group in the
?-aminoadipoyl side chain of cephalosporinCusing a d-amino acid oxidase (DAAO).
He was years ahead of his time, but unfortunately his staff in Sefton Park and ourselves
in Ulverston were not able to generate an economic process for the recovery of
the product.Dr. Fildes no doubt feels somewhat vindicated today by the later adoption
of his process by Farmitalia (now Antibioticos) as part of their successful technology.
cephalexin, through the blanket license agreements with the National Research and Development Council
(NRDC), which owned all the patent rights to cephalosporins and derivatives thereof.
6We opted to develop diphenylmethyl (DPM) protection as an alternative to PNB. More detail of the
chemistry is provided in Chapter 9.
They coupled Bob Fildes’ DAAO-enzyme first step with an acylase-cleavage step to
generate a commercially successful process for producing 7-aminocephalosporanic
When the senior management in Glaxo Laboratories changed (1974), a harsh compartmentalization
of responsibilities occurred, wherein factories such as Ulverston
were restricted to process investigation and troubleshooting and responsibility for
process research and development was returned, fully, to Greenford. It seemed to me
a form of organizational terrorism. Dr. Fildes left Glaxo to become Vice President of
all development (primarily fermentation, chemistry, and chemical engineering) in the
Industrial Division of Bristol–Myers in East Syracuse, New York. He persuaded me
to join him. At the time, control of the Industrial Division was in the hands of a very
tough Italian, Dr. Abramo Virgilio, whose mission for development was that they
create process cost reduction and quality improvement as rapidly as possible, and
whose mission for his marketing arm was that they pursue sales of existing products,
notably 6-APA, ampicillin, amoxicillin, 7-ACA, kanamycin, and amikacin to meet
agreed, but aggressive, targets. In defining “as rapidly as possible” for development,
he required that any money spent on process cost reduction had to produce full
payback in no more than 18 months! Bob Fildes provided the vital buffer between
ourselves and the short-term thinkers in senior management and encouraged the science
that led to the many successes of our chemical process development group. Our
group was also funded to develop processes and to produce supplies of APIs for
the Research Division’s drug discovery and development programs. Our successes
led to a close and harmonious relationship with the Research Division. However,
neither the Research nor the Industrial Division would countenance delay of their
programs by any perception that we were favoring one Division’s requirements over
the other’s. Although we were well-staffed to meet the needs of both, we had to be
careful and realistic in making promises to either. In reality, the careful balance of
resource utilization was only seen to be acceptable if we exceeded expectations for
both divisions! Bob Fildes proved to be masterful in handling the balance despite
his many other roles which required that he travel extensively worldwide. He proved
quite adept at managing all his responsibilities at 40,000 feet!
Our workload became more realistic for a while when Dr. Virgilio was posted to
manage Bristol–Myers’ Far Eastern Division, and Dr. Filippo LaMonica took over.
This continued for a couple of years when numerous changes occurred. Dr. Irwin
Pachter, Vice President of Research, retired and Dr. Julio Vita took over. Dr. Virgilio
returned to take over the Industrial Division and Dr. LaMonica left. Bob Fildes moved
on to become President of Biogen and later Cetus. Dr. David Johnson replaced Bob
and I moved to take Dr. Johnson’s place as director of development chemistry and
engineering. Dr. Vita decided that Research should control its own API supply and
began building his own facility—there was no Bob Fildes to argue against this.
Dr. Fildes’ courageous, persevering British bulldog approach to problems and
issues was admired and needed. He was never afraid of controversial combat, including
with the FDA. Unfortunately, the bulldog image was seen by many as
7See Chapter 9 for an account of this work.
metamorphosing into that of a Rottweiler. Nevertheless, his career flourished in a
different way beyond Cetus.
Dr. David Johnson (1975–1982). Of all my senior managers, Dave Johnson was the
one who knew most about organic chemistry and synthesis. He was a hard-driving
chemist with a “nose” for practical solutions to process development problems. Being
a student of Professor John Sheehan, his knowledge of ?-lactam chemistry was
extensive. Indeed he was called on to represent Bristol–-Myers in its many patent
battles with Beecham in which Bristol–Myers staked out its own patent position
covering ampicillin and amoxicillin trihydrates.8
Dave Johnson generated many outstanding synthesis proposals during our frequent
technical meetings—he always tried to stay involved—and stimulated the thinking
of all around him. He had a synthesis vision that he promoted through in-depth
discussion of specific chemical reactions and brainstorming with our chemists and
me in intense sessions. No problem ever seemed insoluble to him, and as a result we
all rose to the occasion. I particularly remember Dave’s exhortations on the problem
of overcoming Beecham’s patent on amoxicillin synthesis, a patent that, if it could
not be overcome, would shut down Bristol–Myers’ efforts to gain a share of the
lucrative Japanese amoxicillin market. Dave was relentless in goading us to search
for a newer/better way of acylating 6-APA (preferably solubilized in an organic
solvent) with p-hydroxyphenylglycyl chloride hydrochloride. There is no doubt that
his efforts to stretch our minds to the limit, search our imaginations, and rummage in
the most abstruse literature, for this newer/better synthesis were chiefly responsible
for the practical success we finally achieved—which evolved from a finding in an
obscure Russian journal.9 I have no doubt that this success would not have arisen
without Dave Johnson’s perseverance.
Above his chemical vision, Dave Johnson was both a friend10 and a mentor for me
and many of my staff during the period of organizational upheaval at Bristol–Myers
described above. Dr. Vita’s initiatives broke up the chemical development organization
and resulted in Bristol–Myers losing many fine scientists and engineers. I
was fortunate to be identified by a headhunter and recruited by the Schering–Plough
Research Institute to become theirVice President of chemical development. This coincided
with the time when Schering–Plough was seeking revolutionary changes under
the exceptional and inspiring leadership of their CEO, Robert Luciano. I joined them
reporting to Dr. Hal Wolkoff, Senior V.P. of all development operations, including
pharmaceutical sciences, analytical chemistry, organic chemistry and biotechnology.
Dr. Hal Wolkoff (1982–1992). My years reporting to Dr. Wolkoff were the most
exciting, productive, and satisfying of my entire career. Hal Wolkoff was, to me,
8Once, while on a fishing trip by flying boat into northern Canadian Lakes, he was desperately needed
to aid a patent action. Dr. Roy Abraham, at headquarters in New York, was able to call out the Canadian
Mounties to find him—true to the legend they again got their man!
9See Chapter 7 for detail of this work.
10Inter alia he introduced my boys and me to the bone-chilling “sport” of ice-fishing on lakes Cazenovia
and Oneida.
the most level-headed yet courageous visionary of all the people I worked for. He
saw the big picture and agreed that chemical process development was not about
chemistry alone. However, he needed a good case justifying our vision of what a
modern chemical process development organization should look like. We had to
convince him that the additional functions we wanted to adopt would fit with all
the components of his larger development organization and also with the relevant
groups in other parts of the company. He needed to know how we thought all the
new functions we proposed adding would actually work, both together and in the
larger organization. Although he might have needed to make a few leaps of faith, Dr.
Wolkoff accepted the overall logic of our proposals and gave his unstinting support.
He backed and often represented our case to senior management. Slowly a new
comprehensive chemical process development function emerged.
As a result of Dr. Wolkoff’s efforts, the following initiatives were supported by
the company:
 Headcount was increased by recruiting many high-quality people into Chemical
Process Development.
 Funds were secured for modern laboratory and pilot plant, equipment.
 In-house support groups were funded (Analytical, Safety, Environmental, and
Regulatory Affairs).
 A chemical biotransformation group was introduced.
These initiatives are described in more detail in Chapter 3.
These enhancements took several years, in all, to introduce but provided the
backbone of technical power that had so much impact on company operations, in
both manufacturing and research.
Dr. Wolkoff deftly handled his position of power within the Schering–Plough
Research Institute. His grasp of what was needed to achieve desired goals and his
ability to distill the essentials from complex information and then to make concise
and focused decisions that went to the heart of a problem were rare and admirable
qualities. In keeping with my other visionaries, he recognized that people were the
most important assets in any organization, and his efforts to acclaim what his people
had achieved were widely appreciated. Also, he did not shrink from constructive
criticism. I always knew where I stood.
These were the people who provided sustained scientific/engineering leadership in
the pharmaceutical company settings I worked in.
To quote Stephen Mulholland,11 “Scientific leadership is a useful and necessary
drive in those industrial scientists who have it in them to make an impact on
their organization through their own achievement. Scientific leadership requires the
11South African Times, January 17, 1999.
assumption of risk, the acceptance of failure, and the determination to overcome it
when it strikes.”
“What is useful to bear in mind is that very few people are willing to assume
leadership in the sense of being prepared to assume risks and assume responsibility.
Many of course desire the fruits of leadership but only a tiny proportion of people are
willing to expose themselves to the risk of failure. An even smaller proportion truly
wish to have responsibility. The hard truth is that the vast majority, notwithstanding
their almost universal desire for recognition and the fruits of success, are not chosen,
or they hang back, because they are not well-equipped for leadership.”
Scientific/engineering specialists in the field of chemical process development
need to acquire a complex blend of skills. Scientists and engineers may be wellendowed
intellectually and by training to imagine synthetic schemes for the preparation
of an API, and go into the laboratory to test them. They may have the right
gifts of curiosity and imagination. They may have the energy, tenacity, and skills
to implement imagination, but that is seldom enough. Some of the most overlooked
additional requirements for becoming a successful chemical process development
chemist are gaining experience, recognizing and cultivating practical solutions to
problems, satisfying the regulatory disciplines, and accommodating the bottom line.
To prepare for leadership in chemical process development, one needs to draw on an
apprenticeship integrating chemistry with pertinent disciplines in a practical fashion.
The following pays tribute to a few of the people who made the most memorable
contributions to the shaping of my own chemical process development career.
Dr. Martin Hultquist (1960–1966). Martin Hultquist was one of the most gifted,
practical, ingenious, and generous process development chemists I ever met. He
worked for American Cyanamid in Bound Brook, New Jersey, for many years, but
his dream (like the dreams of Tom and DickWaugh)was to return to Colorado (hewas
born in the tiny hamlet of Laird close to the Nebraska border). To that end, he pursued
Arapahoe Chemicals for years, ultimately persuading them to give him a job. My own
“training” was immeasurably enhanced by Martin’s amiably intense and imaginative
approach to chemical process development and scale-up. His vast experience was a
technical resource for all of Arapahoe’s laboratory scientists. Chemistry thoughts and
advice were given unstintingly and always with a view to enhancing the Arapahoe
mission. His work bench may have been a mind-boggling jumble of glassware, as
though an earthquake had passed through, but, diving through it for a thermometer
or a dropping funnel or anything else, he demonstrated he knew where everything
was! He was a master of speed, convenience, and multi-tasking, often to be found
smoking a pipe and watching a reaction going on a hotplate while exploring ideas for
new reactions with his trademark test-tube experiments—generally a prelude to his
next flask-sized experiment. It all seemed like wizardry—a power of transforming
something common into something special.
Martin Hultquist had a rare instinct for organic chemistry and a “green thumb”
that provided an education for us raw young chemists. Many simple solutions came
from his work. He found ways to work in water as a solvent whenever he could. He
would often acidify basic solutions of acid-sensitive compounds with methyl formate.
He encouraged the use of isopropyl acetate (b.p. 89?C) instead of ethyl acetate (b.p.
77?C) because of its reduced water solubility and greater stability to hydrolysis. He
preceded the phase transfer catalysis era using detergents to speed reaction rates and
increase yields. His bag of tricks, as he would whimsically refer to his armory of
techniques, was an eye opener for his more conventional disciples.
Whenever he had a spare moment, he could be found thumbing through the latest
chemistry journals. Martin Hultquist had an infectious passion for chemistry and was
an inspiration to the entire laboratory staff.Most of all, when your experiments failed,
he was always there with an encouraging word, a story of his own tribulations, and a
few good thoughts and suggestions.
Glaxo Co-workers (1966–1975). There were many co-workers in Glaxo who contributed
significantly to the successes of our laboratory, pilot plant, and plant programs.
The following were kindred spirits in our efforts to break out of the conventional
mold and do something new and better:
Brian Clegg led our chemical engineering department and later the entire development
department. His chemical engineering training, his exploratory spirit, and
his judgment and leadership were vital assets during our pilot plant and plant work
to prove that the diphenylmethyl (DPM) group for carboxyl protection was a safe
and practical option. Brian Clegg, convinced by our laboratory data, enthusiastically
endorsed scale-up of our initial process which involved handling hundreds of kilos
of peracetic acid and the separate preparation of hundreds of kilos of diphenyldiazomethane
(DDM). Many were nervous about the risk of a runaway reaction, or an
explosive decomposition.12 Subsequent to this work, Brian Clegg made many enormous
contributions to process engineering and process safety in Glaxo over many
years, most noteworthy being his work with Hans Weibel of Rosenmund AG which
led to the development of better filters. Later, Brian Clegg played a vital role in
Glaxo’s plant engineering projects both in the United Kingdom and in Singapore.
Dr. TedWilson added considerable technical strengths to our Ulverston chemical
process development group when he, along with Glaxo, Greenford, colleagues, Drs.
Brian Laundon, and George Taylor, decided to leave Glaxo Research and join us
in Ulverston. Ted Wilson demonstrated his practical creativity in his work to generate
a phase transfer catalyst approach to the preparation of DDM. He defined the
structural requirements in the phase transfer catalyst for the best yields of DDM. He
made other notable contributions, particularly in discovering penicillin G 1(S)–oxide
acetone solvate, a compound that could be produced in a very pure state. Ted Wilson’s
scientific leadership was recognized as an important asset in his further career
development—he later went on to head the Greenford process development group,
and a few years after that he moved to Bristol-Myers to take over the post I vacated!
Dr. Roy Bywood no doubt made many contributions to Glaxo’s Evans Medical
Division before this unit’s research effortwas shut down.Wewere fortunate to engage
Roy Bywood. His persnickety, quantitative approach to organic synthesis contributed
12Fortunately, thanks to the work of our Gerard Gallagher and Drs. Ted Wilson and Roy Bywood, in
particular, we were later able to create a process using DDM generated and consumed in situ.
much to many of our Ulverston projects, but he will be most remembered for his
unraveling of the role of iodine in the oxidation of benzophenone hydrazone to DDM,
a discovery that enabled us to explain previous yield vagaries and that set the DPM
process on a firm foundation.
Others. There were many others in our Ulverston laboratories to whom both I and
Glaxo owe debts of gratitude for their valuable contributions to laboratory and pilot
plant programs. Several moved on to production roles, notably Drs. George Taylor,
Brian Laundon, Jim Patterson, David Eastlick, Colin Robinson, Phil Chapman, and
Mr. Chris Dealtry. One of our most effective laboratory chemists, especially on our
DPM ester project, was Gerard Gallagher. I can also pay tribute to two other bachelor’s
degree chemists, Ray Holligan and Eric Thompson, and two with no formal
chemistry qualifications, Harry Stables and Gordon Bottomley. Their practical creativity
progressed many Glaxo projects. Lastly, I would be remiss in not mentioning
Dr. EricMartlew, an unsung scientist with formidable analytical skills whose passion
for chromatography proved invaluable in our projects and whose willingness to test
out new ideas gave us some insight into the potential for polymer-supported synthesis
(see Chapter 11).
Dr. Gordon Gregory. Apart from Dr. Arthur Best, Dr. Gordon Gregory (“Greg” as
he was affectionately known) was my other mentor in Glaxo—he worked in Glaxo
Research in Greenford. I had previously reported to him when we both worked in
Britain’s Atomic Weapons Research Establishment in Aldermaston (1955–1957). In
addition to our many scientific discussions, mostly about cephalosporin chemistry,
Greg provided wise counsel on ways of working with the Glaxo Development group
in Greenford. His insights into the personalities in Greenford was extraordinarily
helpful; and his rapport with his supervisors—Dr. Joe Elks and, to a lesser extent, Dr.
TomWalker and the director of all research, Dr. B. A. Hems, FRS—undoubtedly contributed
to my being a better-known quantity than might otherwise have been the case.
I was a fairly frequent visitor to Greenford, which helped to create the understandings
that developed, especially during the competitive phase of our PNB//DPM ester
interactions. Through Greg, I was also introduced to several of Glaxo’s consultants,
notably the formidable Professor Derek Barton (Imperial College) and Professors
E. R. H. Jones (Oxford), Maurice Stacey (Birmingham), and Malcolm Clark (Warwick).
Occasionally, I was invited to selected consulting sessions. All these consultants
visited us inUlverston, lending to the credibility of science on theUlverston site.
Bristol-Myers Co-workers (1975–1982). Scientific life in Bristol’s East Syracuse
Industrial Division was driven by hard-nosed practical considerations and financial
realities. Chemists and engineers adapted well to being perennially on the front line
in fielding process yield and product quality problems. There was, however, thanks to
Bob Fildes and Dave Johnson, time to spend on ideas for process improvement under
the 18-month payback rule set by Dr. Abramo Virgilio, and, as in most major organizations,
there were several chemists and engineers who rose to the challenge in both
Syracuse and our major manufacturing facility in Sermoneta, Italy. The enthusiastic
leadership of Drs. Bob Fildes and David Johnson created the environment enabling a
few people to emerge as successful doers and leaders of important scientific/business
Dr. Chester Sapino applied NMR instrumentation to the solution of intricate
problemswith a verve, tenacity, and brilliance that even doubters of his strategy agreed
wasworth pursuing, for a while. Eventually, as a result of his outstanding achievement
inworking out and optimizing the chemical transformation of l(+)-glutamic acid into
l(?)-4-benzyloxycarbonylamino-2-hydroxybutyric acid (BHBA, the N-blocked side
chain for Amikacin) inD2Oin anNMRtube, he gained the credibility needed to apply
dynamic NMR, as we called it, to other major projects. Probably the most important
of these was his application of NMR to the identification and characterization of the
trimethylsilylcarbamate obtained by gassing bistrimethylsilyl 6-APA with CO2 (see
Chapter 7). This finding was vital in enabling Bristol to market amoxicillin in Japan.
Dr. Ettore Visibelli, as head of the process investigation and development group
in Sermoneta, Italy, was the “spiritual leader” of our chemical process improvement
efforts in our Italian plant. His scientific ability and leadership role seemed at times
under siege in the intense rough and tumble promoted by the hard-headed leaders
of this prime manufacturing location. Ettore was a major player in cost reduction
efforts and played a vital role in implementing the technology transfers needed for
the Sermoneta factory to meet production targets. Dr. Visibelli became the beacon for
science in Sermoneta; indeed his scientific skills, coupled with his talent for diplomacy
became crucial in the area of implementing the systems essential for meeting
environmental regulations and liaising with government officials on environmental
Glenn Johnson became the chemical engineering process automation guru for
Bristol–Myers during my time there. He introduced me to the power of computerdriven
process control with his pioneering work in the East Syracuse plant. His
principal achievement was in creating the computer program for automating the
PCl5-mediated cleavage of penicillin V to 6-APA and the corresponding cleavage of
theN-isobutylcarbamate of cephalosporinCto 7-ACA. This programwas particularly
demanding in requiring precise operation at lowtemperatures (?30?C) and in needing
that all process steps be adapted to eliminate physical handling; thus solid PCl5 was
prepared in situ by adding chlorine to PCl3. The same process plant was used for
producing both 6-APA and 7-ACA. Because this usage raised regulatory concerns
associated with the possibility of contaminating one product with another, the cleaning
of the plant between campaignswas regarded as an essential part of the manufacturing
process. Glenn was able to build an efficient automated process for clean-out between
campaigns by simply running the entire cleavage process through the plant without
using any penicillin or cephalosporin.
Others. In any appreciation of the work of a department, one can always identify
many dedicated, hard-working chemists and engineers who played important roles
in the department’s technical achievements. Among the people who made my seven
years at Bristol-Myers so successful were chemical engineers Walt Williams, Bruce
Shutts, StephenYu,DaveWarner, and Dave Angel and chemists,Drs. Chester Sapino,
Chou Tann, Marty Cron, and Messrs. Glenn Hardcastle, Herb Silvestri, Mario Ruggeri,
Nikki Rousche, Steve Brundidge, Jack Ruby, Kenny Shih, and J. S. Lin. I was
later flattered to have four of these join me when I moved to Schering-Plough (see
In addition, there was always a good collaborative spirit between ourselves in
chemical process development and fermentation process development, thanks to
excellent rapport with Drs. Richard (Dick) Elander, David Lowe, and Leonardo
Schering–Plough Co-workers (1982–1996). It was clear, even before I joined
Schering-Plough, that the company was on a mission to revolutionize the way it
did business, largely seen in the appointment of the dynamic Robert Luciano to the
post of CEO.Major changes in senior management, decisions to increase funding for
Research, inter alia, and decisions to lure in a new cadre of leaders augured well for
the future. Mr. Luciano created an adventurous climate and urged on the subsequent
progress by encouraging and inspiring employees to rise to the new challenges which
inevitably developed. Many great people from the outside saw the opportunities and
joined the company. Change was easier to introduce in chemical process development
when Bruce Shutts, Dr. Chou Tann, Steven Yu, and Mario Ruggeri joined us from
Bristol–-Myers and Dr. George Love joined us fromMerck. These people, along with
like-minded people already in the organization (notably Drs. Marty Steinman and
Doris Schumacher and Messrs. RayWerner and Bob Jaret), were instrumental over a
relatively short time in changing the culture of our organization to one more focused
on science and the fundamentals of process engineering. The latter was key. Prior to
the arrival of Bruce Shutts and Steven Yu, no chemical engineers had been hired for
more than 15 years—chemists (who had lower salary requirements) were believed to
be perfectly satisfactory substitutes!
Bruce Shutts, like his supervisor at Bristol, Walt Williams, was born and raised
in Pittsfield, Massachusetts, and was schooled in chemical engineering at Cornell
University, New York. The Cornell chemical engineering program provides a comprehensive
chemistry training as well as an excellent training in the core chemical
engineering discipline. As a result, Bruce proved quite conversant in both chemistry
and analytical chemistry. He quickly picked up the skills needed to run analytical
instruments, notably NMR instruments, and, in the days before his managerial talents
were recognized, he was frequently to be found in the laboratory carrying out the
experiments needed to define a pilot plant process. This hands-on approach served
him well in his dialogue with chemists and enabled him to appreciate and help them in
creating processes. He used his training effectively, and often brilliantly, in the chemical
engineering aspects of process development. He pioneered, within Schering, the
technology of process containment and became as familiar with the nuances of operating
a controlled environment room as in identifying, and spearheading, Schering’s
investment in process equipment wherein the plant itself served as the controlled
environment room (introduction of the Kraus–Maffei Titus system to Schering—see
the case study on Dilevalol Hydrochloride–Development of a Commercial Process—
was entirely Bruce’s brainchild). Bruce played major roles in both (a) running process
development projects for preparing APIs and (b) our programs with manufacturing
(identifying process equipment needs for particular chemical reactions and aiding
Puerto Rico in its programs to raise steroid process yields and reduce costs). Over
time, Bruce worked hard to familiarize himself with the main Regulatory disciplines,
safety, environmental and FDA regulatory affairs. Bruce Shutts became a
well-rounded and adventurous engineer/scientist/manager asset and played a major
role in our successes.
My almost two decades of working with Dr. Chou-Hong (“Joe”) Tann was
undoubtedly the most scientifically productive and successful period of my career.
Chou Tann served with the military after graduation from university in Taiwan. He
gained his doctorate from Catholic University in Washington, D.C. with Professor
John Eberhardt and went on to “post-doc” with Professor Steven Gould. I hired Chou
to work in our development groups in Bristol–Myers to augment our efforts to use
NMR to understand the chemical transformations going on in process development
work. Initially, Chou worked with Dr. Chester Sapino, his mentor and first supervisor,
and raised the science of using NMR (both in process research for leads and
in the development and optimization of processes) to a level well beyond anything
previously achieved. Also, it was not just Chou’s NMR skills in analyzing chemical
reactions that set him apart. He joined my Schering-Plough chemical process development
team in 1983 and quickly demonstrated a creative ability much needed both
in rapidly searching for new approaches to the synthesis of Schering’s new APIs and,
equally important, in the revolution of long-standing manufacturing processes. Chou
also proved he had a gifted approach to people selection and attracted many fine
young scientists into our organization (Drs. T. K. Thiruvengadam, Xiaoyong Fu, and
Junning Lee all introduced major advances in several projects). The group worked as
more than just a team; in fact, it worked as a family striving to rise in the world.
Many examples of the successes of Chou Tann and his team are detailed in the
following pages. His impact on the manufacturing operations of Schering-Plough,
especially in Puerto Rico and Mexico, was truly immense. I can mention one contribution
to manufacturing which demonstrated the value of his attention to detail and
his zeal to fully understand what was going on in a chemical reaction.
Chou had brominated steroid I with 1,3-dibromo-5,5-dimethylhydantoin (DBDMH)
to give the bromohydrin, II, which in turn was formylated (Vilsmeier reagent)
and treated with base to give epoxide III:
Vilsmeier Base
This reaction scheme had been successfully carried out in the laboratory, giving III
of high purity (ca. 99.5%). Before the process was introduced into the plant in Puerto
Rico, Chou and his team undertook a number of large-scale runs in our Union, New
Jersey, pilot plant using Puerto Rico intermediate I and their new batch of DBDMH
(a batch not yet used by Puerto Rico) received from our normal supplier. Chou
observed, in all of the pilot plant runs, that the yield of epoxide was as expected but
was puzzled by the purity number (99%), which was consistently 0.5% lower than
typically found. Chou Tann and his team undertook many laboratory reactions with
different lots of intermediate I, different lots of DBDMH, and different solvents in
an attempt to resolve their quality finding. This led them to undertake a mass spectral
analysis of the new DBDMH which revealed the presence of the fire-retardant,
octabromobiphenyl (IV), as a trace contaminant.
This very insoluble compound accumulated in product III at a low level but proved
to be undetectable in the final betamethasone product. Despite this, Schering decided
that no betamethasone should be made using DBDMH contaminated with IV on the
grounds that polybrominated biphenyls are known to concentrate in body fat and
that hexabromobiphenyl was implicated in a large-scale poisoning of dairy cattle in
Michigan in 1973.13,14 Other steroid manufacturers used this DBDMH, unaware of
the contamination, and later were embarrassed intomultimillion dollar recalls of their
products from the marketplace. In short, Chou Tann’s vigilance and high standards
saved Schering from a similar fate.
Chou Tann was mostly responsible for numerous other innovations in other
projects. Picking up on the trimethylsilylation approach to solubilizing aminoglycosides
(see Chapter 7), Chou and his team created new processes for the selective
acylation of polytrimethylsilylsisomicin and polytrimethylsilylgentamicin B which
led to the current manufacturing processes for the preparation of netilmicin and isepamicin.
During this work he created a valuable new formylating agent, formylmercaptobenzthiazole,
a reagent that deserves wider attention. The very significant contribution
he and his team made to improving Schering-Plough’s steroid manufacturing
operations are summarized in Chapter 9.
Chou Tann’s selfless ability in encouraging his co-workers to express themselves
provided the environment leading to Dr. T. K. Thiruvengadam’s invention of the
13I had earlier encountered this probably worthy philosophy at Bristol–Myers when Joe Bomstein, our
QC Director, dismissed efforts to completely segregate Kanamycin production from penicillin production
with the words “If you cannot detect penicillin in Kanamycin, your test in no good!”
14Sax’s Dangerous Properties of Industrial Materials, 8th edition, R. J. Lewis, Sr., Ed., Van Nostrand
Reinhold, New York, 1982, p. 2830.
process for the manufacture of Schering-Plough’s highly successful cholesterol absorption
inhibitor, ezetimibe (see Chapter 9).
Looking back over my 43 years working in the pharmaceutical industry, I can unequivocally
say that Chou Tann was the best chemical process development scientist
I ever had the privilege of working with.
Ray Werner obtained his degree in chemical engineering at the New Jersey
Institute of Technology and was already established when I arrived. Ray was one
of our greatest assets in advising us on the way the organization worked at the time
and thus became an invaluable resource in enabling us to climb out of the era of
chemist domination of pilot plant operations. To his credit, Ray quickly recruited
chemist and analyst help to supplement his engineering skills in creating pilot plant
procedures. Our takeover of the manufacturing operations of the Union site and
adaptation of the large-scale equipment would not have happened in the desired time
frame without Ray’s evaluations and advice. Ray continued to be a major asset and
chemical engineering resource with respect to our programs in the Manufacturing
Steven Yu obtained his chemical engineering training at the Massachusetts Institute
of Technology and honed with it an incredible work ethic, a can-do attitude,
and an ability to see how his engineering skills needed to be applied in any project.
His affable and outgoing personality brought people together, even under the most
harried of circumstances, qualities that promoted him into significant management
roles within the chemical development organization. Steven welcomed dialogue with
the many chemists who sought his advice before writing their pilot plant procedures.
He was also much in demand as an evaluator of plant equipment needs for the Union,
Puerto Rico, and Singapore sites. His initiatives, in seeking further education in
the regulatory requirements associated with chemical and API processes, led to his
becoming responsible for the Union-site manufacturing operations. Steven became
an important asset in the organization as well as being recognized as a chemical
engineer’s engineer.
Dr. Ernst Vogel came to lead our Swiss Chemical Development Operation in
Schachen, near Lucerne, with both impressive credentials (Ph.D. from ETH, Zurich,
and postdoctoral experience with Professor David Evans at Caltech in California)
and industrial experience working in the Vitamins Division of Hofmann LaRoche.
Some would say his genes were also right. His father was a co-founder of the
chemical supply house Fluka. Ernst led his organizationwith gentlemanly courage and
enterprise and made many scientific contributions to numerous projects, especially
in the areas of preparing and/or outsourcing intermediates for such as our penem,
ACE inhibitor, and antifungal projects. He also played a major role in setting up the
Schering Biotechnology program in Switzerland.
Ernst could always be relied on, greatly relishing adventurous projects. He was
personally involved in transferring the chemistry for producing the sulfur-containing
fragment of our Spirapril (ACE inhibitor) project to Schering’s Mexican plant
(and climbed Popocatapetl (?19,000 ft) while waiting for plant engineering modifications!).
He took on new technology, in setting up plant to run a process at
?80?C, when my Union colleagues got “cold feet.” This equipment was then very
successfully used in carrying out a chiral hydroxylation of an olefine using a chiral
dichlorocamphorylsulforyloxaziridine (discovered by Franklin Davis at Drexel
and made “practical” primarily by our Dr. Dinesh Gala—see Chapter 4). Once installed,
this equipment became very useful in several other projects that required
low-temperature chemistry.
Ernst Vogel built on the support of several outstanding direct reports, notably by
Ruedi Bolzern, his plant engineer (highly regarded, and always on top of every imaginable
kind of engineering project), Dr. Ingrid Mergelsberg (an experienced chemistry
“all-rounder,” especially talented in techniques for producing chiral molecules), and
Kurt Jost (who managed the pilot plant with impeccable thoroughness andwas “ahead
of the curve” in waste disposal and environmental matters).
Dr. Doris Schumacher graduated from Gettysburg College, Pennsylvania, gained
her master’s at Johns Hopkins, Baltimore, Maryland, and continued her further education
in part-time study while working for Schering. It took her eight years, working
with Professor Stan Hall at Rutgers University, New Jersey, to complete her Ph.D.
Doris’ career owed much to her incredible sense of purpose, towering determination,
and hard work. These qualities, infused with humility, a common touch, and a willingness
to pick up on the ideas of others, served her extraordinarily well during her
long career, which was rewarded by scientific recognition (Presidents Award) and
promotions. Doris was a wonderful role model for other aspiring people. She and
her co-workers made a number of very important contributions to Schering-Plough
programs. The key steps of the Schering manufacturing processes for Loratadine and
Florfenicol were invented by her and her team. She showed enormous tenacity in
pursuing chemical transformations she believed should work, her ultimate achievement
being to demonstrate that a previously unsuccessful attempt to use Ishikawa’s
reagent, for the step of converting CH2OH to CH2F in Florfenicol manufacture, could
indeed be made to work—in nearly quantitative yield (see Chapter 7).
Finally, to underline Doris’ restless quest for further education, she completed a
law degree at Seton Hall University, New Jersey, in 2004!
Dr. George Love brought a vital discipline, physical organic chemistry, to our
organization. He studied with Professor Harold Hart, Michigan State University, for
his Ph.D. and did postdoctoralwork with ProfessorRobertMoss atRutgersUniversity,
New Jersey. He went on to Merck and gained valuable experience in chemical process
development work before joining Schering. George was one of the key figures in
changing the Schering way of thinking in two key areas. One was to persuade
Schering’s manufacturing people in Puerto Rico and Mexico to provide theoretical
yield data in addition to the weight/weight yield data they used in their accounting.
This was achieved by their acquisition of purity data, especially on intermediates,
enabling us to make better sense of every step in each process. George’s effort,
supported by the Manufacturing V.P., Jim Confroy, was no mean feat considering the
expense of adding people and modern analytical instrumentation to the manufacturing
site. The effortwas absolutely vital in enabling us to provide a scientific basis for yield
improvement, especially in the steroid manufacturing processes. The other change in
the way of thinking was in the Regulatory Affairs area. George was seconded to the
Regulatory Affairs Department for several months, where he acquired the insights
needed to enable Chemical Development to gain a real voice in decisions on what
technical information should be included in our INDs and NDAs. On his return from
this “sabbatical,” his efforts enabled us to preserve some flexibility in our written
submissions to the FDA, especially in submitting information on the early steps of
a process. We were able later to accommodate crucial, if sometimes seemingly only
minor, process changes in our operating procedures through mechanisms agreed with
our regulators.
By approaching his chemical process development work from a quantitative analytical
point of view, George was one of the key people, along with Chou Tann and a
few others, who demonstrated that fundamental understanding of the process chemistry
and identification of the impurities in every process step was essential to yield
improvement. The process improvements made through these efforts, especially over
the years in the steroid processes, were worth millions of dollars to the company both
from yield increases and in avoiding the need for capital investment in additional
processing equipment to meet the requirements of our growing steroid markets.
Dr. Junning Lee was one of several outstanding people in Dr. Chou Tann’s
organization, in addition to Drs. T. K. Thiruvengadam and Xiaoyong Fu. I had the
opportunity to work closely with Junning Lee for about 4 years in the area of finding
better chemistry for themanufacture of Ceftibuten, licensed by Schering-Plough from
Shionogi (see Chapter 9). He was seconded to work directly with me and with the
several other parties also involved in the project, namely, Colorado State University in
Fort Collins, Antibioticos in Milan, and the Electrosynthesis Company near Buffalo,
New York. Dr. Lee proved to be not only a gifted laboratory experimentalist but also
superb in liaison initiatives with the other three laboratories. His scientific insights,
business acumen, and ability to get the right work done at the bench level were major
factors in the technical success of the project.
Although Dr. Ashit Ganguly, Vice President of Schering’s Drug Discovery operations
on the Kenilworth site, was in the research arm of the Schering-Plough Research
Institute, he was an extremely important collaborator. His genius has been wellrecognized
in numerous awards for his many avant-garde scientific achievements. He
was an organizational peer of mine but, with respect to meeting his research needs
for API supplies and for chemical intermediates, my role was a subordinate one. In
short we did everything possible to help him move his research programs along as
rapidly as possible. We also worked closely on the chemistry aspects of a few of
the projects assigned directly to development, where we played the lead role in the
efforts to find a lower cost process for the manufacture of Ceftibuten. The liaison and
rapport that we built with his research group was enhanced during the period when
we occupied laboratories alongside those in his organisation. We benefited greatly
from interactions with his people, notably Drs. Girijavallabhan (Giri), Stuart Mc-
Combie, Mike Green, Elliot Shapiro, Paul McNamara, Adrian Afonso, Vince Gullo,
John Piwinski, and more. A particularly strong and invaluable rapport was also established
with Dr. Ganguly’s structural chemistry colleagues, specifically Dr. Birendra
(Ben) Pramanik (see Case Studies—Temozolomide). Research also benefited from
Chemical Development’s discoveries that we freely passed on through ongoing scientific
dialogue—for example, our Dr. T. K. Thiruvengadam’s brilliant chiral ?-lactam
synthesis (see Chapter 9). Dr. Thiruvengadam’s synthesis became the vehicle through
which research synthesized many new cholesterol absorption inhibitors. The team
spirit was also enhanced by the several consulting professors we shared, notably
Professors Sir Derek Barton, Ronald Breslow, and Paul A. Bartlett.
The close interactions between our two groups led to the acquisition of several
of our best contributors from the Research organization. Before my time, these were
Drs. Marty Steinman, Dick Draper, and John Jenkins, and later Drs. Shen-chun Kuo
and David Andrews. One of our Development team, Dr. Nick Carruthers, even went
the other way, with considerable success.
Others. Our chemical development organization was driven, in every sense of this
word, by the enormous enthusiasm, commitment, and professionalism of all of our
personnel. I owe a great deal to Dr.Marty Steinman, who, especially in the early days,
selflessly advised me through the intricacies of the changes I needed to make. He
served as a sounding board, restrained some of my excesses, and went on to demonstrate
steady leadership in managing a large section of our laboratory operations.
Marty later played an important role in our outsourcing mission.
Drs. Don Hou and Nick Carruthers joined us from Professor Paul A. Bartlett’s
Group in the University of California, Berkeley. Don proved diligent and creative in
learning the “development trade” and made outstanding contribution tomany projects.
His ingenuity in identifying an avant-garde synthesis of our D2 antagonist CNS drug
(Sch 39166) and hiswork on enantioselective alkylation (Farnesyl Protein Transferase
Inhibitor Project) provided outstanding examples of “out-of-the-box” thinking. Nick
Carruthers had earlier worked for Roussel–UCLAF in the United Kingdom on penem
syntheses.More than most, he demonstrated that chemistry training enables one to be
comfortable undertaking chemical process discovery and development in any field of
chemistry. His synthesis contributions to the transformation of 9?-hydroxyandrost-
4-ene-3,17-dione into intermediates useful for Schering’s manufacturing processes
were particularly creative (see Chapter 9). Several of our Ph.D. chemists had a hand
in our steroid process discovery and improvement programs. Notably, Dr. Richard
Draper made many visits to Mexico City and provided valuable insight and inputs
into their operations. The two who later did the most work in Mexico City were
Drs. Donal Maloney and David Tsai. Donal was seconded from Schering’s process
R&D operation in Rathdrum, Ireland, and spent a couple of years working in our
Mexican production plant before joining our chemical development organization
in Union, New Jersey. Donal’s chemistry and analytical inputs into the processes
being run in Mexico City demonstrated the inestimable value of seconding a highpowered
scientist, and especially one with production experience, to work on the
ground at the plant site. David Tsai traveled numerous times to Mexico City and
became a respected visitor who, like Dr. Maloney, did much to bring new chemistry,
new analytical techniques, and better process understanding to the site. These efforts
enabled us to make rational changes to the plant processes. As a result of this work
and the efforts of all the support people on the Mexico City site, process yields
improved and product costs declined substantially over the years.
Therewere otherswho contributed greatly to our programs to improve plant steroid
processes. Dr. Xiaoyong Fu, in collaboration with Drs. Chou Tann, T. K. Thiruvengadam
(T.K. for short) and Junning Lee, was one of the principal architects in our
successful introduction of our new process for “dehydrating” 11?-hydroxysteroids
to 9,11-steroids (see Chapter 9). T.K. proved to be very special and one of our
most gifted scientists from the very beginning when Chou Tann recruited him into his
group. Although T.K’s lovely exploitation of the Passerini reaction, to create albuterol,
never did take off his brilliantly successful ezetimibe synthesis did (see Chapter 9).
T.K. made many other contributions—for example, to Schering’s aminoglycoside
processes. Anantha Sudhakar, who is not just another Ph.D., demonstrated extraordinary
creativity in utilizing allene chemistry in two of our projects, one to
establish 9?-hydroxyandrost-4-ene-3,17-dione as a starting material for Schering’s
anti-inflammatory steroids (see Chapter 9), and the other in our highly successful
program to create a manufacturing process for the chiral left hand fragment of Schering’s
superior new antifungal, Posaconazole (see Scheme 1 in Chapter 8). When I
graduated (retired), it was clear that Anantha’s accomplishments and talents would
lead him on to greater things. Also in this category was Dr.GeorgeWu, whose highly
creative chemistry and irrepressible enthusiasm bore fruit in several synthesis challenges,
particularly in Schering’s florfenicol and farnesyl protein transfer inhibitor
projects. In the latter, his creative use of a variant of the Heck reaction (converting
a 2-bromopyridine to a carboxyanilide with CO and aniline in the presence of a Pd
catalyst) led to a highly efficient commercial process. Dr. Dinesh Gala broke new
ground for us on many projects, with the chiral hydroxylation of olefins at very low
temperature being one of the most memorable. Dinesh was one of the few who made
time to write papers and publish his work. (The problem is partly, if not mostly,
of management’s making, resulting from pressing people to move on quickly from
one “completed” project to a new one.) Bill Leong should be mentioned along with
Junning Lee, for their efforts within the American Chemical Society, New Jersey
local section, and the Sino-American chemistry society, respectively, to promote the
profession of chemistry on the larger stage outside the internal activities of their
We were fortunate in employing many very talented, hard-working bachelor’s
and master’s degree chemists without whom we could not have succeeded. Bob
Jaret, despite being labeled early on as “outspoken,” was recognized rather late in
his career as a person with a considerable grasp of the broad requirements needed to
synthesise anAPI.He came into his ownwhen we promoted him to lead our pilot plant
operation. Bob had a practical “bottom line” vision as well as a great appreciation
of the people needs in organizing the work of engineering and implementing a
chemical process on a pilot plant scale. He became a valuable asset, and the flow of
APIs from his pilot plant was testimony to his leadership. Lou Herczeg blossomed
as a chemist working in George Love’s group. He quickly picked up on George’s
fervor for process understanding: One outstanding achievement was his isolation,
identification, and quantification of all the impurities produced in manufacturing the
final steroid intermediate produced in our Mexico City plant.Hewas a frequent visitor
to Mexico, greatly aiding their process improvement efforts—he survived the 1986
Mexico City earthquake with vivid memories of the walls of his hotel cracking open!
Lou later used his acquired knowledge and skills to take on the task of writing our
Development Reports (essential for our interactions with the FDA). Mario Ruggeri,
with his Sicilian flair, perfectly mirrored the picture of Mt. Etna on his office wall. He
was seconded to our manufacturing plant in Puerto Rico, where he worked long hours
to introduce them to the routine use of HPLC to gather the fundamental information
needed for process control and improvement. I personally appreciated theworkMario
did to lay the groundwork for later successes. I also remember him for his incredible
tomato plants, which grew over the roof of his Puerto Rico house but set no fruit!We
lost an enthusiastic chemist and a great character when he was headhunted away to
manage the plant of a generic penicillins manufacturer in Columbia, Maryland.
There were many, many more bachelor’s/master’s chemists deserving of thanks.
Richard Rausser (el barrelito, as he was referred to in Mexico City), Pete Tahbaz
(who, it seemed, could do anything), Tim McAllister, John Chiu, John Clark,
Michael Green (all quiet, reliable, technically accomplished, hard-working doers),
Cesar Colon, Kim Belsky, Jan Mas, Bruce Murphy, Gene Vater, and on and on.
One person deserving special mention is Alan Miller, who worked with passion and
energy in pilot plant scale-up. His motto is “If you enjoy what you do you never need
to work!” In regard to environmental matters, our operations were fortunate to be in
the hands of our most experienced chemical engineer, Bob Emery. Environmental
Compliance became more difficult with time, and we came to be dependent on the
competent, conscientious, and exacting Liz Dirnfeld to keep us “clean.”
Our process safety people, notably Dr. Rick Kwasny and Messrs. Joe Buckley,
Bob Giusto, Howard Camp, and Jay Marino, proved wise and dedicated professionals
who thoroughly educated us in calorimetry, the tests to run, and the practices
to adopt to ensure we met the requirements for safe operation.
Our successes owed much to the rigor of the analysts in our chemical development
analytical team who worked vigorously and tirelessly to ensure we met the set quality
standards and who worked collaboratively to resolve issues. Their responsiveness at
times seemed superhuman. I particularly recall Paul Sandor, Robert Strack, and
Paul Johnston, who in turn relied on the dedication of co-workers including Fred
Roberts, Alicia Duran-Capece, Jian Ning, and others. In the larger analytical context,
our colleagues in the separate, core analytical department were true colleagues
in their enormous efforts to help progress our projects—Gene McGonigle, Nick
DeAngelis, Van Rief, Don Chambers, and Caesar Snodgrass Pilla, to name only
a few. Their commitment and involvement were essential to our progress.
Our biotransformation group (Drs. DavidDodds, Alex Zaks, and BrianMorgan)
contributed to most of our chiral synthesis projects, although in most cases enzymebased
routes were not selected over chiral induction or classical resolution processes
for the short-term needs in API synthesis. This area, however, remains one of huge
promise with the prospect of working in water being one of its most appealing
The quality and professionalism of our large-scale work improved significantly
through the hiring of several gifted engineers, Bruce Shutts, Steve Yu, Al DiSalvio,
Noel Dinan, “Perry” Lagonikos, Joe Cerami, Vince Djuhadi, Andy Ye, and, later,
Guy Gloor and Anthony Toto, to add to the able hard-pressed people already in
the organization, Bob Emery, Ray Werner, Don Beiner, Lydia Peer, and Ron
DeVelde, conscientiously assisted by a chemist-turned-engineer, Stan Rosenhouse.
One of our big plusses was our employment of an electrical engineer, Tom Brennan,
who proved to be an invaluable asset in many projects. Successful operation of our
pilot plants and large-scale plant depended on our forepersons (notably John Junio,
Ed Coleman, Al Regenye, Dan Simonet, and Al Winkelman) and operators.
Good operators are well-trained, experienced, proactive and reliable. They show a
shrewd understanding of plant equipment and often ran a procedure on the knife
edge of operability with the critical eye needed to improve it. Good operators never
allow stressed equipment to become a problem. They behave as if they were owners,
developing an instinct for what looks, sounds, feels, and smells like normal.
They continually involve others in getting things right and, as needs change, which
in a development situation is all the time, they are the people who adapt, learn,
and do. They briefly mourn the loss of failed projects and generate the enthusiasm
and drive to move on to new challenges. There were dozens of process operators and
support people on whom successful operation depended. I talked to many of them
fairly regularly in the course of “rounds” of our facilities and in reviewing projects
on the “shop floor.” All appreciated being appreciated! A few I can recall, many
years later, are Al White, Khalif Rashid, Elvie Cooper, Bill Hood, Bill Fee, Dan
Coakley, Lewis Balcom, Al Fiers, George Dietrich, Henry Hill, Steve Zimenoff,
John Czerwinski, and our diligent maintenance leader Tony Meyer and his assistant
Pete Ruffo.
The entire operation of a plant is dependent on the supply and warehousing of
chemicals. Here the dedication of talented professionals (Jeff Samuel and Jenny
Dong) provided a vital service in ensuring the timely delivery of quality materials.
For the warehousing and stringent documentation covering receipt, storage, and
distribution, we were fortunate to be in the hands of Dennis Von Linden and his
No people acknowledgment would be complete without paying tribute to the enormously
talented and well-organized administrative assistants I relied on, especially
in my Schering years, to ensure that the organization ran smoothly. They were called
secretaries, but they took on a much more proactive guidance role, beyond the routine
definition of secretary. Those who had the greatest impact, over many years,
were Elaine Piete, Janet LaMorte, Gina Alcaide, Lavonne Wheeler, and Kathy
On the larger stage, our interactions with the Schering manufacturing organization
were strongly supported by John Nine, President of Worldwide Manufacturing,
and his vice presidents, Jim Confroy and Michael Monroe. They enthusiastically
encouraged our collegial rapport with the technical movers and shakers in all their
major manufacturing plants in Rathdrum, Ireland, in Mexico City, in Manati, Puerto
Rico, and, later, in Singapore.
Of all the technical people in manufacturing, the greatest concentration of talent
was in our Rathdrum, Ireland, facility. Drs. Brian Brady, Henry Doran, and
Maurice Fitzgerald provided an enthusiastic and extraordinarily creative technical
resource. Their practical genius enabled them to design manufacturing processes that
were simple, efficient, productive, and economical. It was essentially their chlorpheniramine
process which convinced Schering that purchase of their originally tiny
company was a good investment—and it was. During our 15 years of close association
with them—including the frequent visits of people, both ways, to promote
practical chemistry and technology transfer—we made tremendous progress in all
the projects we handled together. Their “chemistry” (between people as well as at the
bench and in the pilot plant) had a practical elegance that had a major impact both
on their own processes and on manufacturing scale operations all over the Schering
organization, notably in Singapore. Brian Brady was the consummate leader—he
had grown up, as I had, exceeding the offerings of his home chemistry set, carrying
out experiments such as the spectacular Thermit reaction in his own back garden.
Because he was given responsibility for the Analytical/QC function, as well as the
chemistry R&D function, he harnessed the combination to the benefit of Rathdrum
synthesis programs as well as in the exquisite resolution of many impurity problems.
Henry Doran possessed a nearly incandescent practical creativity and needed Brian
to temper the ardor of his fertile mind—he had wonderful and invaluable insights
in process chemistry and was an engaging companion in discussing chemistry anywhere.
15 Maurice Fitzgerald was one who just got on with the business of chemistry.
He was quite the reverse of Henry in demeanor but no less a powerful practical
chemist whose incredible persistence wrung chemical processes out of the most unyielding
situations. In broad terms the Irish group was one of exuberant creativity
which employed an abundance of great characters. Tony Smith was the affable general
manager for many years and magically overcame his English heritage in being
embraced as a virtual Irishman. Stephen Barrett, whose other passion was sporting
dogs, took over on Tony’s retirement. Conor O’Brien was their marvelously crusty
and colorful purchasing manager, as well as a collector of Irish silver.
My only regret with the Irish was that I did not get them involved sooner in
polishing Chou Tann’s Albuterol process. If the Irish sodium borohydride process for
the final triple reduction step (see Chapter 5) had been proved earlier, Albuterol would
be being produced today using it.We wasted too much time expecting a third party to
come through employing the original reduction using borane-dimethylsulfide, such
that both process justification and momentum were lost. It was my failure. I also wish
that more of the work of the Irish had been published. For one, Professor Lawesson
would have been delighted that his quirky reagent (for converting –CO to –CS) had
actually been adopted by Rathdrum on a commercial scale!
Puerto Rico was, culturally, quite different and, although the production support
scientists and engineers did not have the entrepreneurial spirit of the Irish, given
our technical support and the enthusiastic encouragement of their Polish-American
leader, Rich Murawski, they played a large part in helping us to introduce better
technology. In particular, Puerto Rico was Chou Tann’s “field of dreams,” where
he and his staff, working with Puerto Ricans Dr. Yvonne Lassalle, Ms. Iliana
15I recall our last uproarious dinner at my house before I “graduated” when Henry consumed more than
anyone else of five Grand Cru Bordeaux’s. At the end he was found drinking the last of the bottle, heavy
tannins and all, of a memorable 1989 Chateau Figeac, or was it the 1990 Lynch Bages, or . . .?
Quinones, and Messrs. Luis Rios, Luis Gil, and Kenny Llaurador, broke new
ground in both aminoglycoside and steroid projects. Our interaction with our plant
in Mexico City was probably the most intense. Ing. Miguel Escobar, our Mexico
City general manager, persuaded his own senior management and ourselves that
their process issues needed urgent attention. This began a long and fruitful period of
collaboration, some of which is outlined under “Excursions in Steroid Chemistry.”
The Mexico City plant did, indeed, have much to contend with. Quite apart from
chemistry/engineering issues, our nervous corporate security people worked with
Miguel Escobar to avoid his being kidnapped, and they made draconian changes to
the security system after payroll robbers, armed with a machine gun, broke into the
factory at a time he was not there. Despite his heavy administrative duties, especially
in finding, hiring, and keeping staff to support his core of long-serving people and
dealing with the trade unions,Miguel Escobar made a special point of being involved
in our technical discussions whenever he could. Of his technical staff, two stood out:
Dr. GildaMorales, for her outstanding technical abilities, and Mr. Sergio Sanchez,
who proved to be a reliable, always-interested contributor over many years. But it
was Miguel Escobar’s grasp of the ever-changing global nature of the steroid raw
material supply situation and his constant effort and pressure to create safe processes,
and to secure cost reduction, year after year, which became his greatest legacy to
Consultants. The value of consultants can sometimes be difficult to quantify. To
me they were an indispensable component of our organization. It was not just their
analyses of our process strategies, their contribution of ideas or critiques, or their
ability to refer us to literature which we might not have seen, but it was also the
stimulus that outside minds, with none of the inside “baggage,” provided which lifted
us to contribute at a higher level.
Over the years we benefited from visits by many consultants, but it was the
Professors I referred to as the three great B’s of our chemistry who stimulated us the
most. These were Professors Paul A. Bartlett (University of California, Berkeley),
Professor Sir Derek H. R. Barton (Imperial College, London, and Texas A&M,
State College), and Professor Ronald Breslow (Columbia University, New York).
They each brought different qualities to aid us in our work.
We interacted with Professor Bartlett both at the time of his visits and whenever
we needed to follow up on any problem with any of the projects under discussion
with him. He worked with us through a preset agenda and written progress reports
that we sent to him prior to his visits. He wrote detailed reports on our projects and his
ideas following his visits. His modus operandi created a disciplined structure for our
engagements with him and made all his visits extraordinarily successful. Anecdotally,
the success of many of his avant-garde ideas reflected a risk-taking creative style that
was evident both in his own research and in the adventurous sky-diving/hang-gliding
activities in his leisure life! Professor Sir Derek Barton was of altogether a different
stripe. Our scientists were in awe16 of his enormous intellect—he invariably could
16When I worked in Glaxo, awe was more like dread. Especially after he won his Nobel prize, many saw
him as tyrannical, especially if they inadvertently erred in presentation and explanations of their chemistry.
offer several solutions to synthesis problems to our one, and he would refer you to
papers, authors, dates, and once even a page number from his stupendous memory
bank. In his later years he was a delight to work with, but the chemists coming to
our consulting sessions had usually “burned the midnight oil” to ensure that all their
data were “bulletproof” and that they were prepared to answer deep and searching
questions. There was no doubt that Sir Derek “raised the level of the game” of
everyone who worked with him.
Professor Ronald Breslow’s vast knowledge of steroid chemistry, and natural products
in general, coupled with extensive experience in consulting with totally different
industries (such as General Motors), brought a perspective to our consultations with
him which was without equal. His flair for the practical aspects of synthesis and his
appreciation of the accommodations needed to meet the requirements of impacting
disciplines (pharmaceutical sciences, safety, engineering, etc.) enhanced his comment
and suggestions. The many years he worked with us were testimony to the enormous
value of his consulting visits.
Over my many years in the chemical process development “business,” I encountered
several invaluable consultants. Notable were those in Glaxo (in addition to Sir
Derek)—Professors ERH Jones (Oxford University), Maurice Stacey (Birmingham
University), and Dr. (later Professor) Richard Stoodley (Newcastle University and
University of Manchester Institute of Science and Technology). The latter visited us
more in a lecturing capacity butwas always scrupulously careful in what he said about
penicillin chemistry since he also consulted for Glaxo’s then arch rival, Beecham.
We first benefited from consultations with Professor Paul A. Bartlett at Bristol-
Myers and were fortunate to engage him to consult with us in Schering. Another
consultant of note at Schering, in addition to the three great B’s, was Professor Jerry
Meinwald (Cornell University), who brought an enthusiasm and a sense of joy to
his and our chemistry. Although not everyone I encountered believed in the merit of
engaging consultants, I have no doubt that our own track record of process discovery
and the rate of progress of our projects was enormously enhanced by their presence.
Awards. The recognition, appreciation, and advancement of people is one of the
most important activities in a successful organization. Intelligent people know their
strengths, weaknesses, and desires and recognize that everyone cannot reach the
very top. However, they need to see that those who do reach the top are worthy
of the position. They also need to see that there are many “tops,” other than the
very top, which they can aspire to reach. Periodic (usually annual) salary increases
provide only one way of recognition and appreciation. They often follow difficult
(and occasionally debasing) performance appraisals. Top companies usually attract
top people who have never suffered the indignities that go with comparison with
similar peers. For this reason, top companies understand that performance appraisal
and salary increases must be only one of many ways of recognizing talent. Nor
Later, when he came to consult with us at Schering, post his time at Gif-sur-Yvette, he had mellowed
considerably. Over a luncheon one time, he even shared with Dr. Ganguly and me that “Glaxo was the
only company that ever fired me!”
is a performance appraisal system, usually an annual event, adequate to account
for a whole year’s work. Great leaders make performance appraisal a continuum
and build other systems to recognize and appreciate what all of their people do. A
caring approach by leaders for the welfare of their co-workers needs to be created.
This may be through first-name greetings, periodic walk-abouts, thanks for jobs
well done at project reviews or in meetings, and so on. Knowing the people and
people knowing their leaders is an essential part of good management. If your people
understand the reasons for company difficulties, as well as successes, they better
accommodate the realities of company life for themselves. Open communications
(perhaps including through a company newspaper) make people aware of reasons
for company restraint (thereby avoiding the appearance of stinginess, or, in the
unfortunate case of layoffs, giving the company a draconian hard-nosed image).
Whenever possible, in good times and bad, companies should be continually striving
(and to be seen as doing so) to promote other dimensions and definitions of “top”
in order to recognize and reward the people whose work will take the company out
of a bad position or raise the prospects for enhanced company performance—at all
In the scientific/engineering arena, encouraging publication and the presentation of
papers at professional meetings promotes a good image of the company and enhances
the prestige of individuals. Such an exercisemay make an employee as valuable to the
outside as he/she is within the company, thereby making it even more important that
the company recognize individual talent. Awards and a technical ladder of promotion
help to meet the needs and create a good image of the company.
Dr. Ganguly established a President’s Award to recognize outstanding achievements
in his drug discovery organization, and we defined an equivalent award for
all of Development, including analytical development, biotechnology development,
chemical development, and pharmaceutical development.
In brief, the President’s Award for Development was created to recognize outstanding
achievements (by individuals or a team) in the discovery and development
of innovations that measurably contributed to the advancement of company projects
and business.
A committee of three expert reviewers, drawn from the above development areas,
supplemented by one reviewer each from drug discovery and research administration,
was set up to judge the submissions. Most of the reviewers were vice presidents.
Guidelines were created and published for the use of potential candidates who
were made aware that successful submissions would be for exceptionally creative
and meritorious, as well as complete, pieces of work that were beyond normal performance
expectations. The criteria used for judgment were based on creativity, novelty,
value, the difficulties faced and overcome, and the level of cooperation evident in
advancing the achievement.
Candidates were required to provide perspective with a review of the problem
(including an analysis of competing situations and literature references). The
novel solution of the problem had to be shown to be truly innovative (and probably
patentable—see Chapter 7—or publishable in a major journal). The weighting given
to innovation was generally 35–40% of the total score.
The innovation also had to be shown to be of significant value to the company.
The weighting given to value was set at 30–35%, with novelty and value together
comprising 70% of the total score.
Themagnitude of the technical challenge overcome had to be convincingly demonstrated.
This element was given a weighting of 20%.
The remaining 10% of the score was given for participative openness in the
advancement of the project.
All four categories above were scored by the “expert reviewers” on a 1–5 basis:
Acceptable Fair Adequate Good Very Good Excellent Outstanding
1.0–2.0 2.0–2.5 2.5–3.0 3.0–3.5 3.5–4.0 4.0–4.5 4.5–5.0
Addition of the scores, corrected for weightings, gave a final score in the 1–5 range.
Generally, our evaluations only awarded those with scores exceeding 4.0. This
was because a separate award, called the Impact Award, could be given if a more
senior person or champion of the project could make a case that the submission was
worthy of a separate, albeit lesser, award. No one could be given both awards.
It was recognized that some employees would regard the President’s Award as
somewhat elitist, open to relatively few. On these grounds the Impact Award gained
wider significance being open to all personnel. In practice, nominations were generally
submitted through lower-level supervisors on behalf of the submitters, though
self-nomination was also permitted. Every effort was made to avoid trivialization of
Impact Awards.
Awards may also be given to employees who obtain patents from which the
company benefits financially—such recognition is often long after the President’s or
Impact Awards have been made.
Promotions. It seems obvious that some scientists/engineers are not suited to, or
not interested in, taking on a management role. In order to recognize the importance
of significantly creative individuals, many companies have developed a ladder of
promotion parallel to that leading to the vice presidential rank. This may be along
the lines of the following:
Presidential Fellow Vice President
Senior Development Fellow Senior Principal Scientist/
Department Head
Development Fellow Principal Scientist/
Group Leader
Senior Scientist
As with all organizations, rising to more exalted positions is not necessarily
permanent. One’s standing in all positions has to be earned. Thus a Development
Fellowneeds to demonstrate, year after year, a sustained level of scientific/engineering
performance in terms of innovation and the implementation of innovation. The higher
the position, the higher the bar. A presidential Fellow, for example, would have to
be well-recognized outside the company, both in industry and academia, as well as
within it, for his/her technical achievements, thereby earning national/international
Put simply, people are everything in any organization. They provide the leadership,
the character, the vision, the technical innovations, and the day-to-day effort necessary
for the discovery, development, and progression of an organization’s mission. In
the chemical process development field people create the collaborative mechanisms
needed to bring together the various disciplines required to effectively advance science
and technology into plant operation. People express the social concerns needed
to meet and exceed the standards set by regulatory authorities (in safety, environment
and FDA regulatory affairs). People, through their achievements, gain personal satisfaction
and recognition. People also recognize for themselves the vital importance
of continuing education in order to stay “at the top of their game.”
My abiding memories spring from the richness of the people component of the
companies I have worked in, as well as from recollection of the many chemistry/
engineering successes we created over the years. It is a source of reassurance for
the future to realize how technically oriented people from all over the world can
come together to produce solutions to problems and implement them in all manner
of settings.
Drs. Tom and Richard Waugh with engineer Oscar Jacobsen
Dr. Arthur Best
Dr. Robert A. Fildes
Dr. David Johnson
Dr. Hal Wolkoff
Dr. Martin Hultquist
Dr. Chou-Hong Tann
Engineer Brian Clegg
Dr. Ettore Visibelli Mario Ruggeri
Dr. Ted Wilson
Engineer Bruce Shutts
Engineer Ray Werner Dr. Martin Steinman
Engineer Steven Yu
Dr. Ernst Vogel
Dr. Doris Schumacher Dr. Junning Lee
Dr. George Love
Dr. T. K. Thiruvengadam
Dr. Brian Brady Ingeniero Miguel Escobar
Dr. Birendra Pramanik
Dr. Ashit Ganguly with Professor Sir Derek Barton
Autographed British Postage stamp honouring Nobel
Laureate, Professor Derek Barton
A formal people structure is needed to effectively create, implement and continually
update collective strategies and tactics in pursuit of a mission.
Organizations are assembled from diverse human resources to achieve defined missions
through orderly action plans. In industry, organizations and their people are
generally in perpetual competition with others in their effort to “create profit.” They
strive to find pathways to distinguish themselves from competitors in the expectation
of generating opportunities for company and personal growth within the social
system. Robert Frost eloquently illuminated opportunities for distinction in his poem
“The Road Not Taken,” which ends:
Two roads diverged in a wood, and I —
I took the one less traveled by,
And that has made all the difference.
Although Frost’s words crystallize the spirit of adventure, in encouraging a journey
by a less-traveled route, they do not speak to the possibility of encountering setbacks.
Thus the adventurer may, later, need to deal with the unexpected by products of
his/her adventurous spirit. President John F. Kennedy, in the 1960s, recognized the
issue with words to the effect that when you scientists invent something new, I have to
invent a way of dealing with it. In short, organizations built to create something new

Copyright C 2008 John Wiley & Sons, Inc.
seldom deal with all the consequences of their actions. As a result, and perennially
lagging behind, they have to address the need to continually retool their organizations
to try to meet the overall needs. In this way, responsible organizations react to their
failings and work to accommodate a broader (including social) approach in their
organizational thinking. Thus organizations, like people, continually evolve in order
to stay alive and prosper. Unfortunately, in recent years, pharmaceutical organizations
in general seem to have lost their way in addressing social (public) setbacks.
Organizations undertaking drug discovery, development, and marketing are often
taken for granted as sources of cures for illnesses. As a result, the general public
mostly forms its impressions of pharmaceutical organizations not through the valuable
contributions the discovery and development componentsmake to treating and curing
disease, but through the more visible marketing component (the major player in
setting drug prices) and associated media and stock market analyst attention. The
United States media, for instance, pays disproportionate attention to United States
drug price differentials versus other countries, to market withdrawals due to adverse
medical revelations, to aggressive drug company advertising, to alleged bribes and
shady financial practices, to large payouts to failed executives, and so on. Most of the
good works in terms of discovering life-saving drugs, in extending and improving the
quality of life through drugs, in reducing costs by reducing patient time in hospitals,
in donating drugs at times of world crises, in making orphan drugs available, in doing
everything reasonable to minimize animal testing, and in sponsoring educational,
arts, and social programs, and so on, are suppressed by the cacophony of adverse
Although pharmaceutical organizations are not blameless, the media and stock
market analysts share in the responsibility for the poor image of pharmaceutical
organizations. For instance, all three seem to conspire in creating all too frequent
public statements on research findings that can skew the merits and gloss over the
uncertainties in drug research and development. The quarterly reporting of progress
for stock market analysts seems much too aggressive for a complex industry that
advances only slowly. Some calming in the frenzy for information (affecting stock
prices, inter alia), as well as curbing marketing excesses, would seem in order to
allow pharmaceutical organizations time to rethink, retool, and reconfigure the many
components that go into creating a new drug and pricing it for the marketplace.
Without such time to build a coherent position and to harmonize ideas on how drug
research and development should be funded in the world as a whole, pharmaceutical
organizations will remain on the knife-edge of credibility risking government price
controls. Price controls will only curtail the expensive wide-ranging research spirit
of inquiry, which is the cornerstone of future success, and is needed to enable the
knowledge societies to advance at a harmonious rate, at the same time as adapting
their systems to accommodate a changingworld and the emergence of newknowledge
In general, pharmaceutical organizations and the Pharmaceutical Research and
Manufacturers Association (PhRMA) do an unconvincing public relations job in providing
believable explanations of their activities—in particular, why it is so difficult
(expensive) to discover new drugs, why it takes so long to bring new drugs to the
marketplace, why better drugs reduce overall treatment costs, and what goes into
making drugs so relatively costly. To compound their problems, pharmaceutical organizations
often seem flat-footed in responding to criticism, such as in countering
suggestions that they may be covering up adverse information on their drugs. Unfortunately,
the public perception of the pharmaceutical industry is not unlike that of
some of the disciplines, certainly the chemistry discipline, that are needed to produce
new drugs. Chemistry, for instance, also has a poor image in the eyes of the public,
often being seen as spawning an industry that is dirty, odorous, dangerous, and environmentally
harmful. To digress further, chemistry is also seen as a difficult subject
in which to gain a university degree. Small wonder that the number of young people
in Western countries wishing to pursue a career in chemistry has proportionately
declined over the decades and that funds are presently not there to support university
chemistry departments, leading to a few chemistry departments being closed in the
United Kingdom.
Recent signs are more encouraging as Chemical Societies, individuals, and industrial
companies work to publicize the vital importance of the sciences to civilization
and particularly how important science is to the economic life of the advanced (developed)
nations. Much of this appears in learned journals and scientific society
magazines. Efforts to reach out to the public through the mainstream media seem to
be increasing,1 but scientific organizations and industrial companies still have much
to do to help those at the interface with the public rebuild the valued image they once
had. Fortunately, the organizational crisis at the marketing interface with the public
is virtually absent at the Discovery/Development level. There is therefore plenty of
opportunity for pharmaceutical organizations to improve their image.
Therapeutic Teams and the Chemical Development Role
In contrast with the turbulence at the top of pharmaceutical organizations, the organizational
situations at the Research and Development level are challenging in a
different way. In Research and Development the task is to build a team of organizations
that can work together to meet the requirements for successful drug discovery
and development. The chemical development organization is usually considered to be
a part of the Research drug discovery organization, although it can also work as part
of a forward-thinking Manufacturing organization provided that Manufacturing can
ensure dedicated, enterprising, visionary, collaborative, and supportive leadership.
The discovery and early development of APIs is handled in a variety of ways
by pharmaceutical companies. This presentation is limited to describing a Research
organizational structure I have worked in and which continues to work well. Themost
important element in making drug discovery and development effective is to have
visionary open-minded people in the important leadership roles, especially from the
discovery point of view, and in the conceptual areas (also involving marketing, and
a business development organization assiduously screening third-party prospects).
1For example, the New York Times Tuesday Science Section and BBC and American TV programs of the
Discovery type.
Therapeutic Team chairpersons report (dotted line) to President
Other groups reporting to President – Human Resources, Planning, Ancillary Research Groups
FIGURE 1. Matrix organization for the drug discovery and development process.
Chemical process development’s early role is usually to provide quality API as
rapidly as possible, often using the research department chemical methodology (the
Recipe) for the earliest supplies.
In most major pharmaceutical organizations, Therapeutic Teams are created to
shepherd the discovery and development process. Each of these Teams is responsible
for a given therapeutic area—for example, anti-infectives, cardiovascular, oncology,
and so on. Each Team creates a mission statement to formalize its objectives. To meet
its objectives, each Team draws its pertinent human and physical resources from
the Research line organizations, Discovery, Development, and Medical. The people
drawn from the line organizations are those with full authority to speak for their
particular discipline. Therapeutic Team leaders are generally experts in the field they
lead, and they are of high standing (e.g., Vice Presidents) in the overall organization.
A representation of such an organizational structure is provided in Figure 1.
It will be clear that conflicts of resource availability and utilization will occur.
Conflict resolution and overall guidance of the various Therapeutic Team programs
is handled by a committee of the most senior executives in the research organisation,
if conflicts cannot be resolved at a lower level.
The Drug Discovery, Medical, and Development organizations incorporate the
usual interacting disciplines needed to meet their objectives. Drug Discovery incorporates
chemists, biochemists, microbiologists, structural analysts, drug metabolism
scientists, and toxicologists. Medical incorporates physicians, FDA regulatory affairs
(mixed disciplines), drug safety specialists, statisticians, and medical writers. Drug
Development incorporates biotechnologists, microbiologists, biochemical engineers,
chemists, chemical engineers, pharmacists, pharmaceutical engineers, and analytical
chemists (QC).Numerous other disciplines are intended to be included—for example,
virologists, geneticists, bioethicists, computer scientists, QA, and crystallographers,
inter alia. Legal, human resource, finance, business, and other administration disciplines
are also incorporated, as needed, in setting up and running Team programs.
To ensure the broadest reach, Therapeutic Team members are also generally coopted
from the Marketing and Business Development organizations. They also draw
members from company International divisions to provide world perspective. The
Team organizations can only be effective with the very best people in leadership roles
and the very best people from the line organizations to do the work, all collaborating
and coordinating with each other to move programs along in an agreed time frame.
As indicated, the Chemical Development organization provides the APIs for Therapeutic
Team programs. Once some credibility regarding the value of a drug lead
is established (this can often take many years—more than 10 years in the case of
Schering–Plough’s Loratadine), programs are projected by the Team to provide some
kind of order enabling contributing parties to properly plan their inputs. It is recognized
that no program can be definitely spelled out; all depends on the issues
encountered as each information gathering phase progresses. An indefinite drug development
program, outlining the major activities, is sketched out in Figure 2. It will
be appreciated that since so many activities are going on at once, a setback in any
activity will slow (or even terminate) any program.
The timing of Chemical Development’s involvement is generally determined by
Drug Discovery’s conclusion that a given API is a drug development candidate.
Once a candidate has been accepted, estimates of kilo requirements and the timing
of deliveries can be made. Often these projections are quite aggressive such that
Chemical Development is best served by working with Drug Discovery at an early
stage in order to get a head start and, discretely, provide information on potential scaleup
issues foreseen in the preparation of large quantities of emerging API candidates.
It is usually more difficult to decide that a developing API is really not worth
pursuing and to terminate the program. This is often a major factor in terms of
efficient utilization of the organization’s resources.
The Chemical Development Mission and Structure
In my time in Schering–Plough the Chemical Process Development organization was
given a dual role in being staffed to support both the Research drug discovery/
development organization and the Schering–Plough Manufacturing organization.
This arrangement owed much to the fact that Chemical Development was ceded
the entire Union site manufacturing operation, and responsibility to continue manufacture
of a few residual small volume APIs, when manufacturing moved offshore.
Our mission statement reflected this unique state of affairs.
Chemical Development Objectives
 Chemical Development’s primary objective is to provide quality active pharmaceutical
ingredients (APIs) for company Therapeutic Team programs, on time
Ye ar 1 2 3 4 5 6
Synthesis and
API Supply and
creation of synthesis
Synthesis Development
and generation of a
“PROCESS” for scale-up
Inspection (PAI)
Drug Product scale-up
and specifications
Toxicology 2 week 3 mos. > > > 6 mos. 2 year oncogenicity study Extended Safety Studies
ADME* Studies Pharmacokinetic
Understanding biological
Studies with thousands of
volunteer patients to
determine efficacy, compare
with competition and assess
adverse reactions
effects and
Dose range
Phase Phase I Safety and Phase II Phase III
Clinical Preliminary
First in man
Dose range
Refine dose ranging
FDA Filing _ _ IND†_ _ _ __ _ _ NDA?_ _ _ _ NDA APPROVAL
10 – 12 years?
Provide API
to market and
work to submit
data for FDA
Post marketing
testing and
(Phase IV)
*ADME–Absorption, Distribution, Metabolism, Excretion
†IND–Investigational New Drug application
? NDA–New Drug Application
? From Inception –break through API’s may be fast-tracked
FIGURE 2. Indefinite drug development program.
and in a cost-effective manner, at the same time as meeting all Safety, cGMP,
and Environmental Regulations.
 The following objectives are integrated with the primary objective as the API
develops to a marketed product:
 Provide support for Therapeutic Team activities and Manufacturing.
 Create and optimize safe, well-engineered commercial chemical processes
by Phase III meeting all cGMP, Environmental, and cost-of-goods
requirements—usually involving Manufacturing.
 Engineer a total technology package suitable for designing a production
facility elsewhere—with Manufacturing.
 Where justified, undertake manufacture of early launch bulk actives for Marketing,
allowing the company to delay capital investment until the market
needs are fully known and the best process technology is worked out.
 Transfer Technology to company and/or third-party production plants.
 Hire, and work to keep, the critical force of capable people needed to meet the
above objectives, and provide support and training to keep them up-to-date.
In the course of meeting its objectives, Chemical Development provides the bridge
between new drug discovery and Manufacturing.
The above objectives deserve qualification and further explanation:
Remarkably, we were mostly able to meet the primary objective, with very few
supply glitches, largely because of our early involvement with Research and our
emphasis on quality and rigorous attention to detail (see Chapter 6). It took us some
time, starting essentially at ground zero, to assemble and tune the systems needed to
meet all the Regulatory requirements (see later).
Support for the Therapeutic Teams was wide-ranging, covering such items as the
identification and preparation of API impurities for analytical and toxicity work,
analyses of supply quantities and timelines, and, later, projections on possible API
manufacturing strategies and the cost of goods, and so on. Support for Manufacturing
is illustrated in Chapter 9.
The creation of a safe, well-engineered commercial process by phase III was
seldom a task that could be completed to our satisfaction, even with the early involvement
of our manufacturing colleagues. We could always create a practical method
(interim process) for manufacture, usually with careful attention to outsourcing raw
materials and intermediates (see Chapters 4–6 and 8).
Our chemical engineering staff, in collaboration with their manufacturing counterparts,
rose wonderfully to the challenges of creating manufacturing technology
packages, a task made more difficult by all the ongoing, sometimes excruciating
requirements to show that the proposed package would meet all FDA requirements
(see Chapter 6 and 8).
Undertaking manufacture of early launch bulk actives for Marketing was always
controversial. It occurred in one case (see presentation on dilevalol hydrochloride)
and, as it happened, the early launch manufacturing strategy saved enormously, by
avoiding capital investment in a manufacturing plant, when the drug had to be withdrawn
shortly after its market launch. Today manufacture is often undertaken, at least
for intermediates, using third parties. Third-party involvement brings its own problems:
in the diversion of resources to meet the needs for confidentiality and intellectual
property agreements, technology transfer, process support, and administration.
In today’s climate, where everyone seeks to accelerate the supply of API [and all
other activities leading to the filing of a New Drug Application (NDA)], technology
transfer starts very early in the API supply program. It usually commences with
efforts to outsource early intermediates and ensure the production of a quality product.
Technology transfer requires that attention be paid to all the problems cited in the
previous paragraph (see also Chapter 6).
Hiring and keeping the best people for the work needed is the most important
mission in any organization (see Chapter 2).
Chemical Development Organization
It will be readily apparent that given the foregoing mission, a Chemical Process
Development organization needs core skills in chemistry, analysis, and chemical
engineering, and it also needs people with the ability to interact effectively with those
in many other areas and disciplines (Figure 3).
From the outset our mission, enhanced by manufacturing responsibilities, required
that our organizational structure should not be solely that of a service department for
the Therapeutic Teams, and Discovery and Development departments, even though
this was our primary mission.
Biotech Engineering
Research Safety
Manufacturing Pharm Sci.
Teams FDA
Health Services
Public Relations
Domestic &
Animal Health
Capital Investment
Materials Management
Waste Treatment/
Emissions Control
Domestic &
Risk Management
Third Party
FIGURE 3. Chemical development interactions.
In meeting our primary mission, the need for close interaction with our Discovery
colleagues was paramount. In this regard the positioning of some of our laboratories
adjacent to those in Discovery was especially helpful in securing strong dialogue and
liaisons on a scientist to scientist level. Interactions with the overall Development
organization’s Analytical Research and Development (QC) organization was also
vigorously promoted (they created the first analytical methods and specifications
and carried the responsibility for quality control on all APIs). Interactions with the
Pharmaceutical Development organization were also strong. Our collaborations with
them ensured that the API being produced was suitable, particularly in its physical
form, for their dosage form preparations. Chemists were the primary people involved
in the Discovery and Analytical liaisons, with chemical engineers being heavily
involved, aided by the chemists, in interactions with Pharmaceutical Development
and Manufacturing.
Our interactions with Biotechnology Development and the company Safety and
Environmental departments depended on the project. The Safety and Environmental
Departments were particularly involved when new chemicals were being handled.
Interactions with Patents department were always strong, if intermittent, since the
creation of patentable intellectual property was a frequent outcome of our work.
At the time I joined Schering–Plough (1982), the Chemical Process Development
department was minimally staffed and equipped and unable to carry out all the duties
it shouldered. Fortunately, our visionary leader of all Development, Dr. HalWolkoff,
shared our view of the need to introduce small Analytical, Safety, Environmental,
Regulatory Affairs and also Biotransformation functions into the Chemical Process
Development organization and was instrumental in successfully making the case for
funds for the following:
 Headcount, primarily chemists, chemical engineers and analysts, was increased
to meet the needs of the added functions.
 Buildings were modified and upgraded to accommodate modern laboratories,
analytical instruments and some pilot plant equipment.
 An in-house process safety group, with its own laboratory and calorimetry
equipment, was introduced, in agreement with the existing company Industrial
Safety and Hygiene group. Our in-house group became a resource for the
broader company organization, especially Manufacturing, and its efforts led
to wider recognition of the need for some of the same calorimetry capabilities
on Manufacturing sites.
 An in-house process and intermediate chemicals analysis group was created, for
liaison with the central Research analytical chemistry and quality control organization
(housed in the Development component of the Research organization).
 An environmental scientist was recruited to liaise with the company’s existing
environmental affairs organization—the role grew to enabling chemical process
development personnel to better “translate” environmental regulations into
working practice in the chemical development organisation.
 One of our senior scientists was seconded to the Regulatory Affairs department
in the Research organization to become conversant in the process for writing
and submitting the CMC (Chemistry, Manufacturing and Controls) sections to
the FDA (as part of their IND and NDA submissions). On his return to Chemical
ProcessDevelopment, he became a vital asset enabling us tomakemore effective
contributions to Regulatory Affairs and in enabling Regulatory Affairs to better
represent the Chemical Process Development position in creating their CMC
 A chemical biotransformation group was created to harness opportunities for the
enzyme-mediated synthesis of chiral intermediates needed in our API programs.
Later this group also went on to aid Research’s evaluations of metabolism and
chiral chemistry issues in progressing API candidates.
Over the course of time, our organization became fully fledged, as outlined in
Figure 4.
One of the most difficult tasks was to create the operating structure to enable the
organization to work as seamlessly as possible. In my time leading the organization,
the Process Research and the Biotransformations functions in Figure 4 were in
laboratories adjacent to those of the Discovery Research scientists’ laboratories.
Research’s laboratories were nearly three miles away from the pilot plant/production
site where all other functions, except the Swiss operation, were situated. The Swiss
operation based near Luzern overcame their distance away from the core site by
virtue of frequent interactions through their key personnel, all outstanding people
(see Chapter 2).
Themost important vehicles for ensuring strong interactions were monthly project
reports, manufacturing plant visits (for our manufacturing role), appropriately timed
technical meetings, and internal, in-depth symposia. The internal symposia were international
in scope and usually lasted for about three days. They were attended by
selected senior technical personnel from the major manufacturing sites, principally
Ireland, Mexico, and Puerto Rico (and later Singapore), and key technical people
from these sites who were involved in particular programs. The major players from
our Swiss operation were always present. Program reviews usually included contributions
providing invaluable perspective from the Coordinators of the pertinent
Therapeutic Teams. The main contributors to the symposia programs, in addition to
our people from Chemical Process Development, were the people from Research
(drug discovery), Pharmaceutical Sciences, Analytical Research and Development,
and often Patents department. The most important document (needed to promote
preparation for the symposium and to provide the framework for the technical meetings)
was a carefully crafted agenda based, principally, on important current development
projects involving all relevant disciplines, including analytical, pharmaceutical
sciences, and manufacturing (technology transfer). Research often provided an introductory
overview of a major project, which could include a review of how a drug
candidate was faring in a toxicological study or in the clinic. Technology transfer
and manufacturing issues always had a significant place in the agendas. The impact
of safety, environmental, and FDA regulatory affairs was included where pertinent.
Although the symposia were usually held in the United States, they were occasionally
Large Scale
Production &
• Liaises strongly
with Research
• Supply of
initial Clinical
bulk actives.
• Process R&D
• Undertakes
programs, e.g.
metabolites, etc.
• Evaluates
chemical sources
• Integrates
into Chem.
Small-Scale Chem.
(Pilot Plant)
• Produces bulk
actives for Clinical
• Carries out
process R&D to
define processes
through to NDA
• Transfers
wherever needed
• Undertakes technical
support of
offshore operations
Large-Scale Chem.
• Produces large volume
intermediates & API’s
for clinical programs
and the market
• Evaluates process
• Undertakes support
• Materials
Management &
• Maintenance
• Technology Transfer
• Process Development
• Produces certain bulk
actives & intermediates
for early market launch
• Technology
• Plant design
• Automation
• Process
• Permit
• Liaison
• Testing
• Support
* Group liaises
strongly with
central groups
?Guide interactions with all disciplines identified in Fig. 3.
• Liaises strongly
with central
Analytical group
• Analytical
• Analytical
• Process
• Validates
• Liaises strongly
with central
Chem. Dev.
• Produces bulk
actives for
• Carries out
process R&D
• Transfers
• Sources
chemicals in
Sr. Director?
Laboratory &
Pilot Plant
Sr. Director?
Dev. and
Support Services
Regulatory Affairs
FIGURE 4. Chemical development organization and principal functions.
held off-shore at a manufacturing facility. In pursuing the manufacturing aspects of
our mission, we also often held symposia on particular manufacturing issues on the
manufacturing site in need.
The symposium agendas assigned individuals to prepare and speak on given
subjects. The symposium agendas were issued weeks before the meeting such that
speakers always came prepared to speak knowledgably about their projects. Minutes
(mostly agreed action plans and assignments of responsibility) were published to
ensure that all knew where they stood. International Development Symposia were
usually held twice a year. Regular technical meetings were set up on a more frequent
(as needed) basis to enable us to power projects along as efficiently as possible.
The composition of the teams that were forming and dissolving along with
the projects, as well as the leadership of these teams, usually depended on the stage of
the project. Chemists involved from the inception of a project frequently stayed with
the project throughout its “life” in Chemical Process Development, though leadership
could change depending on the aptitude and interest of the chemist.
Some experienced chemists could lead a project from the beginning to its implementation
in a production plant. Frequently, chemical engineers took over at some
stage in the development. Management generally had a strong input into such decisions.
As in any organization within a larger organization, governing day-to-day activities
was an important function. To this end, having set up the organization’s structure
provided the human and physical resources, and having defined the organization’s
mission, constant attention was paid to its operation. This was done by developing
a portfolio of standard operating procedures (SOPs) and codes of practice. The
corporate SOP portfolio provided an umbrella of guidance documents that were developed,
as one goes down the organization, into a cascade of more specific umbrellas
governing, eventually, the operation of each subordinate organization or department.
Each subordinate organization, such as Chemical Process Development, essentially
put together its own portfolio of SOPs under the guidance and tutelage of an overall
administrative (Regulatory) mogul within the organization. These SOPs, which were
common-sense summaries of a logical system for running our technical operation,
were important documents for everyone. In particular, they served as training documents
for employees and were invaluable in demonstrating to outside regulators and
auditors just how the organization ran.
In broad general outline the SOP manuals governing the Chemical Process Development
organization, assigning duties, and ensuring the correct discharge of responsibilities
were collected under the main categories of Administration, Documentation,
and Operations. Each department within the Chemical Process Development organization
was responsible for creating and working to its own specific SOP portfolio.
The following is a partial list of SOP’s governing pilot plant operations:
Administrative Procedures
Including an SOP Numbering System/SOP
Distribution and Review/Master Batch Records/
Implementation of New Batches/and so on.
Documentation Procedures ??
Including Personnel Signature and Initialing lists/
Equipment Status Labeling/Sample Labeling/and
so on.
Operational Procedures
Including Dress Codes for Various Rooms/
Instrument Calibration/Weighing Procedures/
Operating Instructions for Equipment/and so on.
CLEANING PROCEDURES – Including Equipment
Cleaning, Use and Maintenance/Specific
Equipment Cleaning Procedures/Storage of Clean
Equipment/and so on.
We tried to design the SOP’s structures and train all staff in their implementation so
that they would become second nature.
Organization Development
Organisations are concerned with the development of people as well as the development
of products and processes. Human Resources (HR) is the company organization
responsible for undertaking the work needed to meet the diverse people requirements
of the company. In this endeavor, HR works with the leaders of all the disciplines
needed for the company to succeed. HR aids in the identification and hiring of the
best people they can attract to the company often with the help of their contacts,
specialist personnel recruiters, and through recruiting visits to universities. HR also
aids in setting up the mechanisms to follow the progress of people in the organization.
Performance reviews provide one such mechanism. These evaluate the work
of individuals and are often linked with salary awards. They also identify strengths
and weaknesses and provide pointers for further development and improvement. In
my experience, most scientists, analysts, and engineers prefer to grow within their
professional discipline—though many like to be challenged by exposure to other
areas. Thus another function of HR is to follow the development of a person by
identifying his/her abilities, character, aptitudes for types of work, and so on. Over
the longer term, performance reviews provide the basis for career progression, for
promotions, further training and succession planning, inter alia. Occasionally, performance
reviews identify people who are incompatible with the organization and its
mission. Such revelations need to be shared with the person concerned, leading to
programs to help the person become compatible or to find a suitable role elsewhere
in the organization. It is very important for reviewers to listen carefully to possible
misfits since they may be saying something the organization really needs to know.
Should no progress be made, over say a several-month period, termination may be
the only way to help an incompatible person move into more suitable employment.
Such cases are relatively rare but, when identified, need to be documented every step
of the way.
Organizations are also governed by budgets based on goals, manpower deployment,
and time frames—with reports required at reasonable intervals, reviewing
progress. In these areas, technical assistance, especially in finance, was provided by
the overall Research Administration’s Planning, Finance, and Accounting departments.
All the uncertainties in R&D projects led to lively interactions on money,
manpower, and time. Changes and new agreements were facts of life, and often dif-
ficult to deal with. However, like all organizations, ours developed and changed with
the continuum that is the world of Research and Development.
Organizations are fluid “living” mechanisms for implementing visions and achieving
advantage. The more responsible ones change continuously to deal with those
consequences of the activities which adversely affect the social system as a whole.
In pharmaceutical research and development the success of matrix organizations in
progressing drug discovery and development demonstrates the importance of integrating
all the disciplines needed to discover and develop new drugs and illustrates
the importance of conflict resolution in moving forward. The constant foment in discovery/
development organizations requires that individuals and teams of individuals
become aware of what is going on around them, enabling them to ensure that their
position in the overall scheme of things is compatible with the overall objectives. The
operation of an R&D matrix organization is quite different from the operation of the
more complex company organization (especially marketing) at the public interface.
The creative component in drug discovery is the R&D wild card. However, unlike
the wild cards at the public interface, creativity is relatively sheltered from public
scrutiny (excepting where it involves animal testing!). R&D organizations work in a
more science-based, fact-driven framework than public interface organizations. Scientists
live more by their brains than by their wits. In such as marketing, wits aremore
important. Of course both are vital to success. Public interface components create
a marketing program based on drug facts and marketing “wild cards”—open-ended
imaginations on what is needed to maximize public demand for the drug at a price
they think they can get! Therein lies the major basis for the public perception of
pharmaceutical organizations.
In the more down-to-earth chemical process development organization, embedded
in a Research organization, the role that the chemists, chemical engineers, and
analysts play is dependent on the openness and vision of the Research leadership.
In the Schering–Plough Research organization there was agreement, before I joined,
that the Chemical Process Development organization would take on a small manufacturing
role, despite the possible downsides of a dual function. There was concern
that a manufacturing involvement would divert resources from our API supply and
new process development mission. For its part, Manufacturing also needed assurance
that their budget contribution to the added staff, facilities, and functions (outlined
earlier) was being used to meet their needs. The key was not to overpromise to either
party. I regarded staffing for both roles as a boon—in all my previous appointments
(with Arapahoe Chemicals/Syntex, Glaxo, and Bristol–Myers), manufacturing had
been a considerable component of our activities. Involvement with the day-to-day
operations of Manufacturing sharpened attention to production detail, to all the Regulatory
disciplines (Safety, Environment, and FDA Regulatory Affairs), to the importance
of all our people (including people at the working level—for example, process
operators), to diligent record-keeping, to disciplined warehousing, to consistent plant
maintenance, and so on. These activities and collaboration with the Manufacturing
sites generated invaluable feedback that contributed to the creation of more rounded
processes, better meeting both FDA and Manufacturing needs. Both our chemist
and chemical engineering staff were particular beneficiaries, especially in terms of
training and exposure to the real world of manufacturing during their plant visits. As
a result of this experience, all our personnel became more powerful contributors to
the organization. Our Research colleagues, in seeing us (as many of them did) take
their bench chemistry, with some modifications (especially for safety), into a pilot
plant scale-up appreciated the attention to detail and the practical decision-making involved
in selecting process parameters—solvents, reaction conditions, crystallization
techniques, and so on. The success of this rigorous approach helped to foster the vital
mutual respect that existed between the Research and Development organizations as
well as between Development and Manufacturing.
Communicationwith all interacting parties is a core activity in running a successful
organization. In addition towritten reports, along with technical meetings on a regular
basis, one of the most important activities in promoting the enthusiasm, sense of
involvement, and passion for the work was to organize technical symposia to review
broad progress on projects with a wide range of project participants. The personal
interactions that grew from these meetings did much to move projects along.
Operating the Chemical Process Development organization was formally documented
via manuals of Standard Operating Procedures (SOPs) governing all critical
operations. We worked to train people to work to these SOPs, often finding that
feedback from the training process caused SOPs to be changed. We tried to make
them “applied common sense” such that time would not be wasted by continually
having to refer to them.
Engaging and organizing the right people was the most important activity in the
organization. In this we needed the close involvement of company Human Resources
people, not only in the hiring process but, subsequently, to ensure that those engaged
were the right people, that they were satisfied with their jobs, and that we set up the
mechanisms to recognize their achievements, to deal with their (and the organization’s)
failings, to provide development opportunities, and to identify and groom the
next generation of leaders.
As previously stated, there is no more important function than engaging and
organizing the right people and creating the professional environment in which they
can grow. The success of all organizations depends on it.
Also as stated, organizations are living entities created to deal with an infinite
variety of unique missions. Looking back at the organizations I had a hand in creating,
even my last one at Schering–Plough, they were not perfect.We could have done some
things better. Thus this review of my last organization may contain much of interest
to organizations elsewhere, but it is already pass?e and not intended as a blueprint for
other organizations.
The safety of the republic is the supreme law.
—— Ancient Latin Watchword
Throughout history, the chemical and pharmaceutical industries have gained mindboggling
unexpected experience in the hazards of working with chemicals. The
safety literature provides a sobering and dark commentary with regard to explosions,
runaway reactions, fires, toxic emissions, asphyxiations, spills, and so on, and their
consequences. Consequences are seen in the injuries and deaths of people and in
physical, social, and environmental damage around the world.
Industry has learned greatly from these experiences but has also had to accept
outside analysis and governance in the safety field. As a result, many government
agencies have grown to provide guidance, oversight, and regulation for allwho handle
chemicals. Notably, in the United States, the main agency is the Occupational Safety
and Health Administration (OSHA).1 The National Institute of Safety and Health
(NIOSH),2 part of the Center for Disease Control and Prevention (CDC), conducts
research and makes recommendations for the prevention of work-related injuries and
illnesses. OSHA’s mission is to protect workers from hazards, which in part includes
setting limits for exposure to chemicals. Other safety organizations have developed,

Copyright C 2008 John Wiley & Sons, Inc.
such as the Chemical Safety and Hazard Investigation Board (CSB),3 which may be
described as a hybrid of OSHA and the Environmental Protection Agency (EPA),
various Manufacturers Associations, and the Laboratory Safety Institute.4 Fittingly,
the American Chemical Society’s Division of Chemical Health and Safety (CHAS)5
does much to help educate people at the student level regarding the nature of chemical
hazards. The American Institute of Chemical Engineers’ Center for Chemical Process
Safety also provides invaluable Safety information.6 Similar organizations have been
established around the world.
Numerous books on the hazards and safety aspects of handling chemicals have
been published. Two of the now standard works to be found in most technical libraries
are under the original authorship of Sax7 and of Bretherick.8 Those engaged in
assessing health risks can gain further insight into the variability of human response
to chemical exposure from the book by Neumann and Kimmel.9
The EPA is also building a RiskManagement Program (RMP) database from fiveyear
reports required from those manufacturing facilities covered by the RMP rule
(see the Compendial Federal Register 40CFR 68). It is expected that the collected
five-year histories, disclosing broad accident information, will allow a more proactive
approach to predicting future safety performance, avoiding accidents and improving
safety management.
However, despite the initiatives of governments, outside agencies, and internal
company safety organizations, more needs to be done by the individual. Specific
problems require that every scientist and engineer has to be his/her own safety officer
in working with chemicals in laboratory and pilot plant situations. Every scientist and
engineer also has to be aware of health issues that can arise in handling chemicals. The
chemist’s earliest introduction is often through reading Material Safety Data Sheets
(MSDSs) that he/she puts into the perspectives of his/her experience in handling
This chapter provides an introduction, which, hopefully, will enable chemists
and engineers to appreciate the major Safety/Health issues faced by people working
with chemicals. The most immediately devastating are obviously explosion and fire.
Adverse health effects resulting from the exposure of people to certain chemicals
can also be immediate (e.g., exposure to methyl isocyanate in Bhopal), but health
effects can also take time to manifest themselves. In view of the unknowns, it is best
to maintain caution in all situations even though, putting matters into perspective,
exposure to many chemicals (solvent vapors are the most common) can be tolerated
3www.chemsafety.gov/circ—this site provides often detailed reports on incidents and accidents.
7Sax’s Dangerous Properties of IndustrialMaterials, Lewis, J. R., Ed., Van Nostrand Reinhold, New York,
8Bretherick’s Handbook of Reactive Chemical Hazards, Urben P., Ed., Academic Press, Oxford, 2006.
9Neumann, D. A. and Kimmel, C. A. Human Variability in Response to Chemical Exposures: Measures,
Modeling and Risk Assessment, CRC Press, Boca Raton, FL, 1998.
provided that they are below published time-weighted averages.10 Notwithstanding
all of the above, safety infractions still occur for many reasons, the most common
often being at the extremes of a lack of knowledge and comprehension on the one
hand and too much familiarity with a given safety risk (especially if it is reinforced
by a long record of no incidents or near misses) on the other. Such conditions can
lead to complacency and erosion of vigilance.
Inadequate knowledge and comprehension are best overcome by gathering experimental
information, including calorimetric data, on the chemical reactions being
undertaken. Gathering calorimetric information is useful not only in identifying immediate
dangers, but also in gaining a wealth of invaluable data on such as heats
of reaction and crystallization, polymorphic form evaluation, and thermal stability,
inter alia. Examples of the use of the differential scanning calorimeter (DSC) and
the accelerated rate calorimeter (ARC) in evaluating potentially explosive situations
are provided in this chapter. The use of the reactive system screening tool (RSST)
and the Radex safety calorimeter in evaluating reactions for runaway potential is also
described. In addition, a short account of the use of a reaction calorimeter (RC1) in
gaining thermal information on a chemical process, as an aid to process development,
is provided.
An outline of the main functions of the above five instruments follows:
Differential Scanning Calorimeter (DSC). A DSC is a versatile instrument allowing
the chemist and engineer to screen for thermal hazards and also to determine the
heat capacity and purity of a chemical. Another area of significant use is in the
determination of polymorphism in crystals of a given chemical. One of the main
advantages of aDSC is that it enables the user to gain a great deal of information from a
small sample size (1–5mg). The instrument is robust and easy to use, enabling the user
to rapidly obtain and quantify results. The DSC provides information concerning the
onset of thermal events including exothermic/endothermic decompositions without
jeopardizing the instrument or user. It helps by providing data for a “go” or “nogo”
process decision and indicates whether additional testing is needed with larger
samples in other calorimeters.
Accelerating Rate Calorimeter (ARC). TheARC is sturdily constructed for the main
purpose of simulating runaway reaction conditions on a small scale, typically using
a 2 to 5 g sample. The sample is heated to a predetermined starting temperature in a
spherical metal bomb. The sample is allowed to incubate at this temperature while the
instrument control system scans for initiation of an exotherm. If no exothermic activity
is found, the sample temperature is raised, and the “wait-exotherm search” routine is
10(a) NIOSH Pocket Guide to Chemical Hazards, U.S. Department of Health and Human Services,
Public Health Services, Centers for Disease Control and Prevention, National Institute for Occupational
Safety and Health, June 1997, published by the U.S. Government Printing Office, Superintendent of
Documents, Washington, D.C. 20402. (b) Handbook of Chemistry and Physics, 78th edition, Lide, D. R.,
Ed., 1997–1998, Section 16–20. The time-weighted average on a given chemical is the limit a worker
can be exposed to in an 8-hour working day and a 40-hour working week. It is recognized there may be
individual exceptions.
continued until either an exotherm is found or a temperature limit of the system is
reached. When an exotherm is detected, the controller maintains the temperature of
the calorimeter wall at the same value as the spherical bomb. Time, temperature, and
pressure data are recorded while the sample self-heats under adiabatic conditions. The
heat evolved in the exothermic process is moderated by the bomb and its contents.
This moderating effect is taken into account when the data are analyzed by using a
thermal inertia factor to adjust both the self-heating rate and the observed adiabatic
temperature rise.
The ARC provides accurate temperature, pressure, and time data to enable a
protocol to be devised which avoids operating problems when reactions are scaled
Reactive System Screening Tool (RSST). The RSST is a thermal screening instrument.
The sample to be tested is placed in a 10ml insulated spherical glass test cell
that in turn is placed in a 0.5 liter pressure vessel equipped with a simple heating system.
A thermocouple is placed in the sample to be tested; the thermocouple provides
feedback control to the pressure vessel heater, enabling heat losses to be overcome,
thereby ensuring a linear sample temperature ramp from 0.25?C/min to 2?C/min. The
whole system can be pressured up to 800 psi in part to suppress the boiling of light
solvents that could mask an exothermic onset. The addition of reagents to the test
sample can also be arranged through a valve and syringe directly to the sample (to
estimate heat of reaction). The system carries provision for accumulating temperature,
time, and pressure data via a dedicated microprocessor. The accumulated data
can be plotted for the interpretation of test results.
Radex Safety Calorimeter. The Radex calorimeter is a modular instrument that can
simultaneously evaluate six different samples (size range 0.5 to 5 ml), or one substance
under a variety of conditions. Each module is a separate entity with its own calibrated
oven capable of being operated under an open, closed, or pressurized condition,
with all temperature differences between the sample and the oven being stored in a
microprocessor for further analysis. The Radex calorimeter is very versatile; samples
can be tested in either an isothermal or ramp mode. In the isothermal mode, each
oven is heated to a preset temperature and held at that temperature throughout the
experiment. In the ramp mode of operation, the oven is heated linearly to a preset
temperature, or can be maintained at a given temperature for a predetermined time.
The flexibility of oven function in the Radex calorimeter enables the user to determine
the intrinsic stability of a chemical and to also compare the impact of such parameters
as temperature, atmosphere, and impurities on the stability of a given substance.
Reaction Calorimeter (RC 1). This calorimeter has grown in popularity as a practical
process development tool. Its value is based on the precise measurement of thermal
events occurring at each step of carrying out a chemical transformation. The reaction
calorimeter enables the chemist to gain a realistic insight into heating and cooling a
reaction on a large scale. In running an exothermic reaction in a small reaction flask
in the laboratory, the chemist generally relies on a large cooling bath with a very
large temperature differential, versus the reaction mixture, to maintain a reaction at a
desired temperature. In this situation, the chemist essentially conceals the importance
of the ratio of cooling/heating surface area to reaction volume. In using a reaction
calorimeter, wherein the reaction vessel is cooled and heated via an external jacket,
much like a pilot plant vessel, the chemist acquires a better appreciation of the
limitations of a jacketed vessel—in short, the effects of a smaller surface-area-tovolume
ratio. A reaction run in a reaction calorimeter allows the chemist to determine
the importance of heat transfer, and the need to address it, in developing a chemical
process. In another sense, the process development chemist can often save many
hours of operation by obtaining heat of reaction data from a reaction calorimeter to
determine when a reaction is complete, thereby often avoiding the convenient practice
of “stirring a reaction mixture overnight to complete the reaction.”
Application of the above instruments in identifying the potential hazards associated
with any process or chemical enables the chemist and the engineer to make
recommendations for the safest possible operation of that process and the best way
of handling a particular chemical, or, indeed, whether to do so at all.
Chemical Explosion. Efforts to avoid explosion and fire have the highest priority
in creating a safe operation. In the laboratory, the trained chemist and engineer
usually recognize the dangers in working with particular chemical structures. The
most important structures with intrinsic explosion potential are listed in Table 1.
Most of the groups in Table 1 may be regarded as metastable intermediates which
are on their way to carbon monoxide, carbon dioxide, nitrogen, or another more
thermodynamically stable structure.
A simplistic way of classifying these structures is through comparison with the
dyestuff industry, where groups are assigned chromophore or auxochrome status. A
chromophore is a chemical group that gives rise to color when allied in a suitable
manner and in sufficient number with hydrocarbon moieties—for example,
C C C O, C N , N N , N(O) N N O , ,
By analogy, NO2, CN2, C.OO.C , N3, ClO4, and C C groups, which
can produce explosion, may be named plosophores. Plosophores are defined as
chemical structures which are predisposed to cause molecules containing them to
decompose violently when they absorb energy (e.g., shock or heat). This intrinsic
explosivity characteristic is reduced when the oxygen balance with carbon declines
(e.g., trinitrobenzene versus nitrobenzene) or when the plosophore content is diluted
(e.g., diazomethane versus diphenyldiazomethane) or by dissolving (diluting) the
compound in a compatible solvent.
TABLE 1. Chemical Structures with Intrinsic Explosion Potential
Group Structures
Nitrates and nitro compounds C NO2 ; C O NO2 ; C N NO2 ; NO3
? salts
Diazo compounds and diazoalkanes C N2
+X- ; CN2
Peroxides and ozonides C OOH ; C O O C C
; C
Peracids and peresters C CO3H ; C CO3R
Tetrazoles and triazoles
N ;
Hydrazoic acids and azides HN3 , Mx(N3)x e.g. Pb(N3)2
Acetylenes, especially polyacetylenes C C C C n
Chlorates and perchlorates, especially in
the presence of organic matter
HClO3, Mx(ClO3)x, HClO4, Mx(ClO4)x
Continuing with this theme, an auxochrome is a group that can deepen or intensify
color—for example, NH2 or OH. By analogy, these same groups can be regarded
as auxoploses, since combination with a plosophore will enhance the potential for
explosion (see Table 1 for auxoplose–plosophore combinations).
Application of Differential Scanning Calorimetry (DSC) in Dealing with a
Potentially Explosive Situation.11 Everyone working with chemicals containing
plosophoric groups needs to address the hazards long before scale-up is contemplated.
Information gained from calorimetry studies enables those responsible for
scale-up to specify conditions that will prevent explosion. However, as an illustration
of how potential problems can creep up on the unwary, it is pertinent to describe a
situation occurring in Schering–Plough’s process for manufacturing Nitro-dur skin
patches (prescribed for the treatment of angina). The patches comprise nitroglycerine
incorporated into a polymer matrix via an aqueous emulsion process. During the
drying operation, water is removed which passes through a charcoal scrubbing unit.
During an intersite meeting involving chemical development, the manager of our
Chemical Development Safety Group, Joe Buckley, asked for samples of charcoal
from the Nitro-dur scrubbing unit to ascertain whether any nitroglycerine vapor had
condensed and accumulated on the charcoal. Differential scanning calorimetry (DSC)
analyses were carried out on the charcoal versus fresh unused carbon. Each sample
11Buckley, J. T., Marino, J. P., Emery, R. L. Process Hazard Identification in the Pharmaceutical Industry,
ACS 1991 Spring National Meeting, April 19, 1991.
FIGURE 1. DSC of unused charcoal heated at 10?C/min from 50?C to 300?C in a hermetically
sealed aluminum pan.
FIGURE 2. DSC of scrubber charcoal heated at 10?C/min from 50?C to 300?C in a hermetically
sealed aluminum pan.
was heated at 10?C/min from 50?C to 300?C in a hermetically sealed pan. The unused
charcoal exhibited a rapid endotherm at 154?C as water desorbed (Figure 1). DSC
scans of all scrubber charcoal samples (Figure 2 is representative) showed a small
water endotherm at 140?C superimposed on a highly energetic exotherm that began
at approximately 100?C, peaking at 197?C and generating 2.3 kJ/g of energy per
sample. This information, indicating that substantial nitroglycerine was absorbed on
the charcoal, led us to undertake an impact sensitivity test to gain insight into whether
or not the scrubber charcoal might be detonated by an impact or friction shock, or
by a spark. A JANAF drop weight test was carried out by specialists in explosives.12
12Hazards Research Corporation, 200 Valley Rd., Mt. Arlington, NJ 07856.
TABLE 2. Comparison of Explosivity of Nitro-dur Scrubber
Charcoal with Known Explosives Using the JANAF Drop Weight Test
Height of Drop Weight Causing
Initiation of Explosion (Inches)
Nitroglycerine (neat) 0.39
Benzoyl peroxide 5.20
RDXa 8.00
Scrubber charcoal sample 12.50
Picric acid crystals 13.00
TNT 15.00
Ammonium perchlorate 21.60
This test utilizes a known weight—2 kg in our case—that is dropped from a series
of increasing heights up to 36 in. onto a sample in a specially designed cup. A
loud report, flame or other indication of combustion indicates a positive test. The
test result is regarded as the height which indicates a 50% probability of initiation
of an explosion. The JANAF test was carried out 20 times on scrubber charcoal
samples and indicated a 50% probability of initiation at a 12.5-in. drop height. This
sensitivity result was consistent with results on other known explosives. The results
are summarized in Table 2.
Given the results in Table 2, indicating that scrubber charcoal had an impact
sensitivity between picric acid and the commercial explosive, RDX, new procedures
were adopted for the use and storage of scrubber charcoal:
 The charcoal in the scrubber unit was replaced more frequently to minimize
nitroglycerine accumulation.
 Contaminated charcoal was stored in a segregated area and mixed with an inert
material (vermiculite) prior to being sent out for disposal.
 The company undertaking the disposal (incineration) was advised of the new
composition of scrubber material.
Application of Accelerating Rate Calorimetry (ARC) in Evaluating a Reaction with
a Potentially Explosive Nitro Compound. One of our process development projects
required the preparation of 2-hydroxy-1-nitro-2-phenylethane via the addition of
sodium methoxide to a mixture of one mole of benzaldehyde and one mole of
nitromethane in methanol (Scheme 1).
The nitronate salt precipitated as a thick slurry, posing heat transfer and stirring
problems. Before taking the process into the pilot plant, we needed information on the
thermal stability of the nitronate in order to provide operating guidelines to eliminate
the risk of explosive decomposition. Our accelerating rate calorimeter was selected
for the test since it afforded great sensitivity and the best user protection for the
technician running the test in the event of a detonation.
Nitronate salt
SCHEME 1. Preparation of 2-hydroxy-1-nitro-2-phenylethane.
FIGURE 3. Temperature versus time.Nitronate heated in theARCunder adiabatic conditions.
A sample of nitronate (0.8 g) was heated in a titanium “bomb” in 10?C increments
from ambient temperature. Under adiabatic conditions the nitronate began to selfheat
at 70?C and violently deflagrated at 106?C (Figure 3). The deflagration produced
temperature rates of 700?C/min, pressure rates of 6000 psi/min, and nitrogen oxide
off gases.
The ARC test enabled us to set up guidelines for the preparation and use of the
nitronate salt on a pilot plant scale:
 The reaction temperature was held at no more than 30?C during formation of
the nitronate.
 The solvent quantity used in the reaction was increased in order to efficiently
stir and cool the nitronate.
 The nitronate salt was not isolated or dried. Instead, acidification of the salt was
carried out to produce the more stable nitro alcohol.
(i) CF =CFCF 2 3 + HN(Et)2
(ii) 2OH + I
CH Cl 2 2
+ HF
Ishikawa reagent
SCHEME 2. Conversion of CH2OH to CH2F with Ishikawa reagent.
Runaway Reactions. Such reactions may not have the catastrophic severity of an
explosion, but thermal runaway reactions can also be extremely destructive. The
thermal runaway may not be in a reaction solution itself, but may occur as a result
of say a condenser failing, leading to loss of solvent from the reactor such that the
residue of reagents alone is now being subjected to heat.
Application of the Reactive System Screening Tool (RSST) in Examining the Potential
for a Runaway Reaction Associated with Plant Failure. A “what-if” scenario
was examined by Schering–Plough in the in situ manufacture of Ishikawa reagent for
a subsequent reaction with a primary alcohol group to produce an intermediate for
manufacture of the antibiotic Florfenicol (Scheme 2). We asked the question, What
if a leak developed in the pressure reactor used for the preparation and use of the
Ishikawa reagent such that all the methylene chloride solvent distilled out and the
residual reagents were now being heated on their own?
The stability of the Ishikawa reagent itself [(I) in Scheme 2] was examined in our
reactive system screening tool (RSST); this equipment was used, since, by design, it
allowed us to test the effects of heat in a pressure vessel. The reactor and contents
were heated at a rate of 1?C/min under a pressure of 200 psi.
As seen in Figure 4, a small exotherm occurred over the temperature range
65–100?C, possibly coinciding with Ishikawa reagent (I) converting to the enamine
(II) and HF.
A maximum self-heating rate of 2.2?C/minute was reached at 91?C in 16 min.
The system stabilized and no further exotherm activity occurred until the temperature
reached 136?C. At this temperature a violent exotherm ensued reaching 495?C, with
a maximum self-heating rate of 2000?C/min being reached at 340?C within 23 min.
In regard to pressure, other measurements showed that the violent exotherm produced
a pressure of 387 psi/min at 405?C and a maximum pressure of 1600 psi at 307?C.
Venting of the RSST at room temperature released large amounts of HF. The contents
of the RSST cell were a polymerized mass. It was reasoned that HF was causing
an acid-catalyzed polymerization of the enamine, but we have no absolute proof of
this. Nevertheless, the RSST test enabled us to impose the following safeguards for
reagent preparation (i) and reaction with the alcohol (ii):
FIGURE 4. Temperature rate versus temperature. Ishikawa reagent heated in the RSST at
1?C/min with a 200 psia nitrogen atmosphere.
 During the reagent preparation step, process temperature limits were set at
ambient temperature to reduce enamine formation.
 A Hazard and Operability Study (HAZOP) of the proposed operating procedure
in the plant was undertaken to ensure that the “fluorination reaction” heat
exchange system could not exceed 100?C.
 The batch sheet for operation of the process was written to ensure that the
correct amount of the alcohol intermediate (RCH2OH) was added to consume
the Ishikawa reagent produced.
Guarding against runaway reactions has become one of the more important activities
resulting from HAZOP studies now undertaken more or less routinely before
running a process in the pilot plant. In more cavalier times past, most readers can
probably recall bursting discs failing and the contents of their reactors spewing over
the rooftops of their pilot plants. I can personally recall a runaway decomposition
of a hot solid on the M6 motorway in England in 1970. A pilot plant peracetic acid
oxidation of potassium penicillin G had been carried out in water at our Ulverston
plant and the penicillin G sulfoxide (25% water content) had been taken by car to the
Midlands of England for drying tests using a fluid bed dryer (previously a vacuum
tray drier had been used at 40–45?C without incident). After the fluid bed drying was
complete (to approximately 0.5% moisture level), the hot solid (later estimated at
>50?C) was packed into a plastic lined fiber drum and loaded into the back of the car
N2/THF, -90 oC
93% yield
92% ee
SCHEME 3. Chiral hydroxylation of a sodium enolate.
(a station wagon estate car). Everyone wanted to get back north as rapidly as possible
and could not wait for the solid to cool. Some 30 miles north, the fiber drum lid
blew off with a gentle pop and the car occupants witnessed the decomposing product
fizzing slowly to the front of the car, causing them to abandon the vehicle. This traumatic
experience led us to investigate the decomposition (probably acid-catalyzed by
traces of sulfuric acid—the peracetic acid contained small amounts of sulfuric acid).
Rather than relying only on improving the washing step before drying, we created a
process for producing a sugar-like crystal of the sulfoxide as an acetone solvate that
was free of acid contaminants and dried without incident.13
Application of the Radex Calorimeter in Examining the Effects of Iron Contaminants
on the Stability of a Potentially Labile Reagent. The synthesis of chiral
compounds is now a de rigueur activity in the search for active pharmaceutical ingredients
(APIs). Frequently chiral chemical reagents are employed for the introduction
of chemical groups in a chirally specific manner. One such reagent, employed for the
introduction of a chiral hydroxy group onto a prochiral carbon atom in a Z-enolate,
is (?)-(2S,8aR)-[(8,8-dichlorocamphoryl)sulfonyl] oxaziridine14,15 (Scheme 3).
The chiral oxaziridine is very reactive, even at lowtemperature, leading us to query
its stability. Since we were proposing to carry out the preparation of the oxaziridine
in a steel vessel, we investigated the thermal stability of the chiral oxaziridine on its
own, in the presence of stainless steel and in the presence of ferric ion. The stability
test was carried out in a RADEX safety calorimeter. The results are summarized in
Figure 5.
As can be seen, the neat sample (Trace C) began to decompose at approximately
165?C. Decomposition occurredmuch earlier, at approximately 84?C, in the presence
of ferric ions (Trace A) whereas stainless steel, in the form of filings, caused decomposition
to commence at 148?C (Trace B). In all cases, once decomposition started,
it became self-propagating within 15 min—a 1-g sample underwent a 40–50?C temperature
rise within seconds and violently frothed out of the sample tube.
The effect of ferric ion content on the temperature at which decomposition started
was studied, with the results summarized in Table 3.
13Wilson, E. M., and Taylor, A. B. U.S. Patent 3,853,850, 1974 (to Glaxo).
14Mergelsberg, I., Gala, D., Scherer, D., DiBenedetto, D., and Tanner, M. Tetrahedron Lett., 1992, 33,
15Gala, D., DiBenedetto, D., Mergelsberg, I., and Kugelman, M. Tetrahedron Lett., 1996, 37, 8117.
TABLE 3. Effect of Fe3+ on the Onset of Decomposition of
Sample Onset of Decomposition (?C)
Neat 165
0.04% Fe3+ 120
0.09% Fe3+ 112
0.60% Fe3+ 84
?T (Del C)
85.0 100. 115. 130.
Temperature (°C)
145. 160. 175. 190. 205.
FIGURE 5. Thermal stability of (?)-(2S,8aR)-[(8,8-dichlorocamphoryl)sulfonyl]
As a result of this work, the following recommendations were adopted to avoid
possible runaway decomposition:
 The temperature limit in the preparation of the oxaziridine was set at 50?C.
 All solvents were tested to ensure that they were free of ionic transition metals
before use.
 Only deionized water was used in processing the oxaziridine.
 Glass-lined vessels, passivated with dilute nitric acid or alcoholic EDTA, were
used in reactions with the oxaziridine.
 No metallic tools (scoops, etc.) were used in working with the oxaziridine.
NH2.HCl NaBH4 in
(S) (S)
SCHEME 4. Sodium borohydride reduction of methyl (S)-phenylglycinate hydrochloride.
Application of the Mettler RC 1 Reaction Calorimeter in Optimizing a Process for
the Reduction ofMethyl (S)-Phenylglycinate. The sodium borohydride reduction of
methyl (S)-phenylglycinate (hydrochloride salt) to the corresponding primary amino
alcohol (Scheme 4) had been worked out in the laboratory and taken to a pilot plant
scale where an unexpected exotherm was observed.
In outline the procedure was as follows:
Methyl (S)-phenylglycinate hydrochloride (POX-C, 10 kg, 49 moles) in 50%
aqueous ethanol (20 liters) was added over approximately 2 hr to a solution of sodium
borohydride (7.5 kg, 198 moles) in 50% aqueous ethanol (30 liters—previously
adjusted to pH 9.5 with aqueous 2 N sodium hydroxide and cooled to 0–5?C). The
operating procedure for the reduction step called for the reaction solution to be held
at 5–10?C during POX-C addition. In the actual pilot plant run, the temperature
was held in the range –2.5?C to 9?C during the addition (in fact, the temperature
was difficult to control because the cooling circuit to the pilot plant reactor was
at ?40?C). The addition of the ester (POX-C) was also prolonged and intermittent
because each addition caused a rapid temperature increase requiring time to drop to
approximately 0?C before the next aliquot was added. At the end of the ester addition
(batch temperature 1.2?C), the batch was warmed slowly. At approximately 10?C a
rapid temperature increase to 56?C occurred, coinciding with a vigorous evolution of
hydrogen, in spite of maximizing flow of the cooling circuit.
This event, which caused no personnel injuries nor equipment or batch loss, was
unexpected—there had been no indications of unexpected exotherms in the laboratory
process development work.
In continuing the pilot plant reaction, once the temperature had subsided to 25?C,
the solution was stirred for 2 hr to ensure that the reduction was complete (HPLC).
The excess sodium borohydride was consumed by the addition of acetone (20 liters)
over approximately 90 min. The exothermic reaction was readily controlled in the
temperature range 30–35?C. To recover the product, salts were filtered and the acetone
distilled out. The desired product was extracted with n-butanol and the extract was
washed with water. The n-butanol layer was stripped and the residue was dissolved
in toluene. Further salts were filtered, the toluene was stripped out, and the (S)-2-
amino-2-phenylethanol product was distilled under vacuum; the yield was 81%.
In order to understand, and avoid, the exotherm observed in the reduction reaction,
the following process steps were examined using the Mettler RC-1 reaction
Step 1: Cooling the NaBH4/ethanol/water solution from 2?C to 7?C.
Step 2: Warming the mixture obtained after adding the methyl (S)-phenylglycinate
to the NaBH4/ethanol/water from 7?C to 25?C.
FIGURE 6. POX-C/EtOH/water/NaBH4. Temperature ramp from 20?C to 7?C of NaBH4/
EtOH/water; Qrxn = ?0.7 kcal/mole of NaBH4; adiabatic temperature rise is 4.33?C.
As can be seen from Figure 6, a small exotherm (adiabatic temperature rise of
4.33?C) is apparent, corresponding with the heat of crystallization of some of the
sodium borohydride.
Figure 7 shows the result of warming the reaction mixture. At approximately 10?C
a rapid heat evolution occurs (adiabatic temperature rise of 44.33?C). It was reasoned
that this exotherm was a composite of an endotherm resulting from dissolution of the
crystallized sodium borohydride and a larger exotherm corresponding to the reduction
of the ester, POX-C.
As a result of these observations, the following action steps were taken and
reinvestigated using the RC-1 reaction calorimeter:
1. The volumes of 50% aqueous ethanol used for dissolving the sodium borohydride
and POX-C were increased by 233% and 75%, respectively.
2. The sodium borohydride solution was cooled to approximately 10?C (instead
of the original 0–5?C) prior to adding the POX-C.
3. The addition of POX-C was carried out at 10–15?C instead of 5–10?C.
FIGURE 7. POX-C/EtOH/water/NaBH4. Temperature ramp from 7?C to 25?C in 20 min;
Qrxn = ?41.8 kcal/mole of phenylglycine Me ester HCl; adiabatic temperature rise is 44.33?C.
The RC-1 results indicated that no unexpected exotherm would occur. This was
confirmed on the pilot plant scale.
The RC-1 work was extended to a study of the exotherm resulting from the acetone
addition step (to consume the excess sodium borohydride). The data obtained from
RC-1 work was invaluable in calculating the cooling requirements needed for largerscale
work and for further optimization of the process.
As an important aside, the reader will have gathered from the foregoing examples
that uncontrolled foaming and violent gas release are fairly common outcomes of
runaway reactions. Such events have resulted in the need for emergency venting to
prevent or minimize catastrophic damage. This is usually done via a properly sized
vent line fitted with a bursting disc rated at less than the pressure rating of the vessel.
Dust Explosion. Dusts, like solvent vapors, when in the presence of oxygen concentrations
that will support combustion, represent potentially dangerous situations.
Explosion and/or fire can be triggered by a spark, derived from an impact or friction
or an electrostatic discharge or from hot spots (e.g., overheated motor bearings)
that expose flammable chemicals to temperatures above the flash point. The thermal
degradation of materials may produce flammable gases amplifying the destructive
Dusts are most often encountered in milling operations and drying operations,
with dust explosions being mostly associated with organic compounds. The need for
caution is often the greatest for final APIs since a fine particle size, particularly a form
suitable for preparation of the marketed dosage form, is frequently a requirement for
bioavailability. The problem of potential explosivity of dusts can be an even greater
factor in the preparation of drug products since formulations frequently contain sugars
as a major ingredient; sugar dusts can be particularly hazardous in situations where
there is a risk of such as an electrostatic discharge.
The potential risk of initiating a dust explosion is best determined by contracting
specialists to carry out explosivity tests on dusts.16,17 Such testing provides better
understanding of the explosion, fire, and thermal properties of the powders being
processed. In this way, measures to deal with the potential hazards can be selected
to greatly reduce or eliminate the risk. General measures include a HAZOP study to
identify and exclude potential ignition sources, backed up by inert gas blanketing.
Grounding of equipment is a first line of defense against electrostatic ignition of an
explosion. Large vents are also usually provided for explosion relief.
Vapor Explosion. Vapor explosions occur when an ignition source causes a
flammable organic vapor and oxygen to burn at an increasingly very fast rate to
create a high pressure shock wave. This becomes a detonation when the front velocity
of the shock wave exceeds the speed of sound. Chemical plants have the potential
to bring all three elements (ignition source, fuel, and oxygen) together. Elimination
of any one of these three elements creates a safe condition.
Elimination of possible sources of ignition is addressed by building plants with
explosion-proof equipment (motors, light switches, etc.), by equipping operators
with clothing that will not build up an electrostatic charge, by providing them with
conducting safety shoes, and by using only nonsparking tools. Plants also use inert gas
blanketing in operations where there is a potential to generate sparks. Maintenance
programs need to be such that there is no chance of hot spots developing in any of
the equipment, and they also need to ensure that equipment is properly grounded.
One operation requiring especially high vigilance is the use of a centrifuge. The
bowl of a centrifuge spins at a high speed, and any mechanical failure of a moving
part (such as a bearing) might generate frictional sparks or a hot spot. The centrifuge’s
own switches and drive motor are also potential sources of electrical sparks.
The possibility of electrostatic discharges from ungrounded metal parts, operators’
clothing, or old safety shoes offer additional concerns. Footwear and flooring need
to be inspected regularly to ensure that their anti-static properties are maintained:
16The Health & Safety Laboratory, Broad Lane, Sheffield S37HQ, Yorkshire, England. e-mail:
17Factory Mutual Research Corporation, 1151 Boston-Providence Turnpike, P.O. Box 9102, Norwood,
MA 02062. Tel: 617-762-4300.
TABLE 4. Flammability Characteristics of Heptane18
Item Result Notes
Flash point ?4?C Temperatures above -4?C produce a vapor
concentration above the liquid which can be
Flammability limit 1–7% Concentrations between 1% and 7% heptane in
air are flammable. Above and below these
concentrations, combustion cannot be
204?C Heptane spontaneously combusts above 204?C.
Minimum ignition
0.24 mJ@
At this concentration it takes a spark of only
0.24 mJ to ignite the vapor. Higher-energy
sparks are needed when the vapor
concentration becomes more or less than 3.4%.
Resistance between a person and the ground should not exceed 108 ohms. Operations
employing insulating liquids of low flash point need special attention. The use of
inert gas blanketing is standard practice in operating a centrifuge; this is usually
supplemented by installing oxygen meters that shut off the equipment when oxygen
levels rise into the explosibility limits zone for the solvent being used; this zone varies
from solvent to solvent.18
The use of flammable solvents always touches off a “spark” of concern! I recall
one serious incident occurring in the drying of a penicillin wet with heptane. After
filtration the heptane wet product was loaded into plastic lined fiber board drums
and moved to the drying area. During the process of digging out the wet product
and spreading it on trays for drying, an explosion occurred that severely burned the
two operators involved. The subsequent investigation led to the conclusion that an
electrostatic charge generated by the use of plastic scoops had created a spark which
set off the explosion.
The potential dangers of working with heptane can be seen by reference to its
flammability characteristics (Table 4).
A short account of the major considerations for the assurance of a safe working
environment follows. This account provides perspective on the culture of safety
needed in a chemical process development organization.
18Handbook of Chemistry and Physics, 78th edition, Lide, D. R., Ed., Chemical Rubber Company, Boca
Raton, FL, 1997–1998, Section 15, pp. 14–18.
19 Plant/Operation Progress, 1992, 11, No. 2.
TABLE 5. Major plant service Functions of In-House safety organisation
Item Service
Safety training Documented in-house induction and safety reinforcement. OSHA
mandated requirements including evacuation, emergency response,
fork-lift, fire extinguisher, respiratory, eye and hearing protection,
HAZCOM, vessel entry, hotwork, lock-out/tag-out, laboratory
safety and documentation
Safety inspections  Plant inspection with team of rotating membership
 Local housekeeping with team of rotating membership
 Follow-up meetings
 Incident review (including outside incidents for information)
 Audits with company Safety/Health organisation
 Action plans and compliance reviews
Supplies & MSDSs  Ensuring the availability of protective equipment [head (eyes &
ears), hands and feet]
 Ensuring on-site limits on dangerous chemicals are being
observed (quantity, warehousing and conditions)
 Update MSDS logs
The Chemical Safety Organization. Every company has its own umbrella
Safety/Health organization to promote its particular safety culture, to provide the
leadership for the creation of safe operating conditions, and to ensure that such
conditions are adopted. Those departments, or divisions, wherein safety requires
far more emphasis than in such as office areas, create their own in-house “mini-
Safety/Health” organizations with dotted line relationships to the company-wide
(umbrella) Safety/Health organization. In short, the chemical process development
in-house organization, which faces many unknowns in its day-to-day work with new
chemical syntheses, has to have all the capabilities needed to ensure that both laboratory
and pilot plant operations are intrinsically safe. The in-house organization is
also accorded the power to enforce safe operations. To ensure that priority is given
to chemical development needs, the in-house Safety/Health organization has the laboratories,
the instrumentation (notably calorimetric), and the staff to undertake the
evaluations needed in the timeframe needed to meet the urgent requirements of clinical
supply programs. This in-house organization also works closely with the umbrella
Safety/Health organization in implementing a variety of functions needed to ensure
safe operation and worker health, and to provide the data needed to assure safety
compliance. These areas are listed in Table 5.
The in-house Safety/Health organization works with the umbrella organization in
liaison activities with outside groups, including manufacturing, maintenance, environmental,
and others, as needed.
Beyond the above framework (focusing on training, follow-up, attention to detail,
and vigilance in the pursuit of safety), companies endeavor to continuously improve
their safety performance. One popular approach has developed from a critical evaluation
of behavior and an in-depth examination of the “why’s and wherefore’s” of the
choices people make in adjusting their actions to meet both relatively routine and also
sometimes unpredictable situations on their work site. Adoption of behavior-based
safety programs20 appears to be a promising addition to the armory of approaches
for improving safety.
The mechanisms of protecting people from exposure to unsafe situations are too
numerous to summarize here, but one preemptive action that stood out for me was
practiced by the physician in the Glaxo, Ulverston penicillin/cephalosporin factory
in the 1960s. He looked askance at ginger-haired, blond, and generally fair-skinned
people when they interviewed for employment in areas where they may be exposed
to the potentially allergenic compounds being produced, purely on the grounds that
the record showed that they were much more likely to suffer allergenic reactions than
dark-haired or darker-skinned people!!
Operating Procedures. After having gained a great deal of information regarding the
potential hazards of a chemical reaction, the chemist intent on scaling the chemistry
up to a pilot plant scale needs to work with chemical engineers, process operators, and
Safety/Health, environmental, and regulatory people in the creation of an operating
procedure, often referred to as a Standard Operating Procedure (SOP). All involved
in creating SOPs work under the guidelines set by the various government agencies
charged with oversight in these areas: OSHA, EPA, FDA.
The Safety Section of the SOP is the most vital component, since it deals with the
education and protection of those who will run the process. However, it covers more
than the dangers addressed in the hazard evaluation phase. It requires the creation of
Material Safety Data Sheets (MSDSs) for all the chemicals needed in operating the
process. It provides overall guidance on the safe operation of the process and identifies
what protective gear to wear and what precautions are needed in the operation. These
efforts are also inextricably linked with what needs to be done regarding process
emissions and wastes.
Generally, in the early phases of an API synthesis project,many new chemicals are
prepared and used for which there is no or very little safety information or industrial
hygiene data. However, rather than delay the project by waiting to gather all the
safety data (calorimetric data are always gathered to determine whether there are
any major risks), a judgment call is usually made to go ahead using the best safety
protection available, and with the most appropriate plant systems, to deal with the
eventualities that can be perceived. This judgment call usually follows a meeting
to evaluate line-by-line detail of the new process, and also the detail of how the
plant equipment will be used—a form of the hazard operability study. Emissions and
wastes are captured or contained using the best available technology and are usually
sent away to approved disposal experts. In general, a conservative stance is adopted
in all first-time activities.
Eventually, as a project matures to become a candidate for development to a
manufacturing scale, it becomes necessary to gather a great deal of additional data,
20Behavioral Science Technology, Inc., 417 Bryant Circle, Ojai, CA 93023. E-mail:bstojai@bstsolutions.
both to create comprehensive MSDSs on all the chemicals involved and to deal with
the testing requirements for storage of the chemical and for its shipment between
locations. Particular concerns are the stability of materials and dealing with spills.
Happily, the approach to dealing with spills has improved dramatically over the last
four decades. I recall as a young chemist in the 1960s being the sole technical person
in the office area of Arapahoe Chemicals in Boulder, Colorado, when a frantic call
came in from a man on the dock in New Orleans. One of his men had punctured
a drum of 3 M ethylmagnesium chloride in diethyl ether with his forklift truck and
was urgently seeking information on how to deal with it. I ran across the office for a
disposal procedure manual, but in the 30 seconds taken to get back to the telephone,
a calmer voice at the other end said “it’s OK buddy, somebody kicked the drum into
the harbor and it’s buzzing around putting on quite a show.” Today the procedures
for dealing with such eventualities are an essential part of the package of documents
sent out with the chemicals being shipped.
Material Safety Data Sheets. MSDSs should be available for all chemicals sold in
the chemical marketplace. No transportation of any chemical should be undertaken
without an MSDS. In building an MSDS for a new chemical to be shipped, the
following information needs to be provided:
1. Chemical product and company identification
2. Composition/Information on ingredients (formula and CAS number)
3. Hazards identification
4. First aid measures
5. Fire-fighting measures
6. Accidental release measures
7. Handling and storage
8. Exposure controls/personal protection
9. Physical and chemical properties
10. Stability and reactivity
11. Toxicological information
12. Ecological information
13. Disposal considerations
14. Transport information
15. Regulatory information
16. Other information
Provision of all of the above information is primarily for the shipment of larger
quantities of materials. The reality is that gram quantities of laboratory samples,
especially research samples, are shipped with whatever data can be quickly mustered.
The shipper of laboratory samples is responsible for ensuring that proper precautions
are taken in the shipment of potentially hazardous small samples.
Safety Award Programs. Although organizations do much to make people more
conscious of safety—for example, through doormats displaying safety messages at
the building entries, through safety posters on the wall, or through case studies
in safety meetings—my experience with safety awards suggests that, if they are
adopted at all, they are devised to avoid fudging and cover-ups. I have known injured
employees (notably with back problems resulting from poor lifting practices or bad
lacerations) to take vacation days rather than spoil their department’s safety record!
Although this presentation has provided little more than an introduction to the Safety
field, I have tried to promote an awareness of the need to work respectfully with
chemicals for the good of all those involved in a project.
Safety is the most important of the prominent regulated activities (Safety, Environmental,
and FDA Regulatory Affairs) encountered in progressing chemical process
development projects. Death, serious injury, and property damage still result far more
often from events occurring in the manufacture of a drug than from environmental
excursions or during the process of developing the drug efficacy and adverse reaction
information for a new drug application (NDA).
Safety/Health and Environmental Affairs are often interwoven in practice, especially
where process emissions, chemical exposure, and waste disposal can impact
on public health. Many, if not most, companies with chemical synthesis plants in
populated areas work with local communities to foster good relations sometimes via
open house days or in the creation of action plans to deal with adverse events which
may occur in plant operation. The presentation on Environment (Chapter 5) addresses
the canon that has developed to deal with exposure to chemicals, with the impact of
spills and emissions on all life forms and with waste recycle, treatment, and disposal.
As will be seen in the presentation on FDA Regulatory Affairs (Chapter 6), human
safety issues are an integral part of the development of the API through the FDA.
In closing, it is worth pointing out that the controls, discipline, and documentation
built into chemical processes for the manufacture of APIs, to satisfy FDA requirements,
has also greatly improved approaches and attitudes to safety and environmental
Where there are threats of serious or irreversible damage, scientific uncertainty shall not
be used to postpone cost-effective measures to prevent environmental degradation.
——1992 Rio Declaration on Environment and Development
Environmental issues continue to occupy a prominent, real and emotive place inworld
thinking as populations increase and the earth’s declining resources are developed to
meet a variety of “needs,” from growth and employment to survival and creation of
a steady state, and much in between. As a result, the world’s thinkers have become
polarized, with developers and conservationists trying to agree on the best ways of
moving “civilization” into the future.
The effect on industries, which generally try but are also obliged to be “wise” in
the development of anything, has been equivocal, with one good result being that they
have joined, to one degree or another, themovement to build aworld public conscience
on the environment. Governments, aware that conscience was not enough, have
worked to overcome egregious environmental exploitation, while still encouraging
development. As a result, as in other fields subject to government oversight (including
Safety and FDA control over pharmaceutical development), environmental guidance
has been established with the Environmental Protection Agency (EPA) taking a
leading role in providing regulatory oversight. Apart from government activities,
Chemical Societies everywhere are promoting environmentally friendly chemistry.
The American Chemical Society’s Green Chemistry Institute, for example, provides

Copyright C 2008 John Wiley & Sons, Inc.
information and encouragement worldwide in a drive for everyone to create processes
that minimize the release of chemicals into the environment.
Although pharmaceutical companies are small players with respect to the volume
of chemicals prepared and used, they subscribe to the high standards of the chemical
industry, generally with special operations for dealing with the new, exotic and
hazardous chemicals frequently encountered in synthesizing APIs. Chemical Process
development chemists and engineers, because of their bridging role in developing the
processes to be used in the eventual manufacture of an API, need to be fully aware
of the hazards associated with exposure to chemicals and wherever possible to avoid
the use of dangerous chemicals. In this area, safety and environmental protection are
closely integrated, and process chemists and engineers share a common interest in
knowing what to avoid and what laws apply to the use of new chemicals as well
as toxic chemicals and what limits need to be observed with respect to personnel
An outline of the major environmental laws in the United States follows. These
laws provide a framework for the basic standards used in the governance of environmental
matters. The most important laws are the Comprehensive Environmental
Response, Compensation and Liability Act (CERCLA—now more often referred to
as Superfund, a name derived from the passage of a later supplement to CERCLA),
the Resource Conservation and Recovery Act (RCRA), the CleanWater Act (CWA),
the Clean Air Act (CAA), the Safe Drinking Water Act (SDWA), the Emergency
Planning and Community Right-to-Know Act (EPCRA), and the Toxic Substances
Control Act (TSCA). A brief description of these acts follows:
Comprehensive Environmental Response, Compensation and Liability Act
(CERCLA). The original Act (1980) was amended in 1986 by passage of the Superfund
Amendment and Reauthorization Act (SARA). The law governs sites that have
been contaminated by hazardous substances or could become contaminated (e.g.,
by leaks developing in corroded drums of hazardous materials held on a site). The
law imposes liabilities on all those with any connection to the site at the time the
hazardous substance was left there. This includes the generator of the hazard as well
as the transporter; it also includes the owner or operator of the site, as well as any
future acquirer (say, through a merger). Liabilities include investigation of the site
and cleanup charges. The only defenses against liability are (a) Act of God, (b) Act
of War, and (c) Act of Omission (such as an innocent purchaser of the site, proving
that tests of the site had been undertaken and no environmental concerns found).
Another requirement of Superfund is that notification of a hazardous substance
release (in reportable quantities) must be made at the time of the release – failure to
report is treated as a felony.
The Oil Pollution Act is roughly similar to Superfund but is specifically applied
to releases of oil and petroleum products.
Resource Conservation and Recovery Act (RCRA). This law governs the “cradle
to grave” tracking of hazardous wastes all the way from generation through treatment,
recycling, storage, shipping, and disposal. A permit system governs the entire
tracking process through to approved destruction in a licensed, permitted facility.
As in other regulated disciplines, documentation tracking every phase of hazardous
waste movement is required.
The law allows a generator of hazardous waste to undertake treatment, storage,
and disposal, provided that these steps are properly documented; in the early days of
governing RCRA, the steps of treatment, storage, and disposal led to much confusion
(see later).
Amendments to RCRA were introduced in 1984, notably governing the use of underground
storage tanks,many ofwhich were found to be leaking. This amendment essentially
led to underground storage becoming a less favorable and expensive option –
today, all storage tanks need to be in bunded containment areas, with above ground
storage being preferred.
Clean Water Act (CWA). The discharge of wastewater from industrial facilities in
the United States is controlled by a permit system (National Pollution Discharge
Elimination System). The limits of pollutants allowed by the permit depends on both
the nature of the pollutants present and also the situation prevailing in the receiving
body of water, whether this be a river, estuary, lake, publicly owned treatment works
(POTW), wetland, or any other. Effluent limitations are set by the EPA, but individual
states may require stricter limits to ensure that the receiving body of water can absorb
the discharge volumes and pollutant levels proposed. The company discharging the
waste is required to monitor discharge composition and periodically provide a report,
available to the public, logging the levels of pollutants discharged.
The discharge of wastewater to a POTW is usually strictly controlled since the
POTW itself is a permitted facility.Wastewater such as process wastewater, received
from industrial facilities, must meet pretreatment standards so as not to compromise
the treatment undertaken by the POTW. Usually, industrial wastewater is subjected
to pH adjustment and has to meet carbon oxygen demand (COD), biological oxygen
demand (BOD), and particulate content standards set in collaborationwith the POTW.
The discharge of all wastewater from a given industrial site is controlled, including
storm water collected from company land, roads, rooftops, and car parks.
Discharges in excess of permits can exact considerable financial penalties, and
even prison sentences.
Clean Air Act (CAA). In the United States the CAA enjoins the EPA to set ambient
air quality standards and emission limitations that have been adopted by the states
under federally approved plans. The CAA standards and limitations apply to all
sources that might pollute the air including power stations, automobiles, and industrial
sources. Operating permits require emissions monitoring to show compliance with
the standards. As with the CWA, periodic reports on emissions are required which,
again, are available to the public. The public can use reports, showing noncompliance,
in citizen suits against an infringer.
There have been several amendments to the original act including the highly publicized
permit requirements for volatile organic compounds (VOCs) that cause ozone
depletion; some VOCs such as the chlorofluorocarbons are in the process of being
completely replaced by compounds found to be less damaging to the environment.
Safe Drinking Water Act (SDWA). This act regulates suppliers of drinking water to
the public and also those companies that supply their own water (e.g., on-site well
water) to their employees. Maximum contaminant levels and maximum contaminant
level goals have been developed to try to ensure that the risks in deaths from diseases
such as cancer are controlled to a low range [e.g., for cancer, one excess death (over
average) in 106]. Standards are continually improving; for example, the maximum
allowed level of lead in drinking water has been reduced from 50 parts per billion to
5 parts per billion.
The standards developed under the SDWA are often used in setting the standards
for clean up under RCRA and CERCLA.
Emergency Planning and Community Right-to-Know (EPCRA). This 1986 actwas
a component of the (SARA), and it mandates that the use of hazardous materials in a
manufacturing operation is preceded by the provision of information on the “intentto-
use” to regulatory authorities and the local community. This information is usually
supplemented with a response plan addressing the actions that would be taken in the
event of a spill or other release.
EPCRA also requires that companies compile a hazardous chemical inventory
record and report releases. The EPA has used this information to create a nationwide
record on the usage and release of hazardous materials.1 The dialogue resulting from
the availability of such information to the public led most companies to subscribe to
voluntary hazardous emission reduction programs.
The Pollution Prevention Act of 1990 essentially formalized a movement that
now requires that industrial companies report their progress in toxic chemical source
reduction and recycling for each toxic chemical during the prior calendar year. The
information collected is available to the public.
The Toxic Substances Control Act (TSCA). One of the most important aspects of
TSCA lies in the power given to the EPA to regulate the use, storage, disposal,
and clean-up of hazardous substances. The most publicized example of this power
is the case of the enforced clean-up of polychlorinated biphenyls (PCBs) from the
Hudson River in New York. After years of legal wrangling, General Electric was
ordered (in 2001—24 years after production stopped) to dredge the river to remove
PCBs. Another high-profile PCB case, concerning dumping in landfills and local
waterways near Anniston, Alabama, is being defended by Solutia, Inc., a spin-off
from Monsanto.
TSCA has given the EPA broad control over the production and importation of
new chemical compounds. Before production of a new chemical can commence, the
EPAcan require that the substance be tested. In essence, this situation is little different
1For example, Consolidated List of Chemicals Subject to the EPCRA and Section 112(r) of the CAA.
Title III Lists of Lists EPA 550-B-01-003, October 2001. www.epa.gov/ceppo
from the long-standing practice in Europe, embodied in their efforts to register all
chemical substances, both old and new. The drive for this came in part from releases
of chemicals (some accidental, some unauthorized) and the exposure of untrained
people during the transportation of chemicals. The European Dangerous Substances
Directive of June 27, 1967 is the European equivalent of TSCA. This directive led to
the compilation of a list of all known chemical substances made and used in commerce
between January 1, 1971 and September 18, 1981, and it was formally published on
June 15, 1990 under the title “European Inventory of Existing Chemical Substances”
(EINECS). All known chemical substances appearing in this list were essentially
“grandfathered” on the assumption that they had been shipped and used previously
such that the hazards associated with them were presumably known and methods of
dealing with spills and emissions were already available. However, in cases where
the quantities marketed have grown substantially, further testing may be required.
APIs (active pharmaceutical ingredients) are themselves exempt from TSCA since
they are the subject of comprehensive toxicity testing prior to FDA approval.
New chemicals produced after the EINECS list was closed were required to be the
subject of a battery of tests, depending on the quantity, before they could be registered
for use and shipment. The ever-growing list of new substances was first published
under the title “European List of Notified Chemicals Substances” (ELINCS) on May
29, 1991 and is updated periodically. A brief outline of the major tests required before
a new substance can be registered is provided later.
The framework of basic environmental standards is governed by a quality system
that companies set up to oversee all activities. One such quality system, set up in
Europe by the International Organization for Standardization (ISO), requires that a
quality management and registration operation is built into a process plant and that
manufacturers govern themselves through policy and procedure manuals, training
operations, and audit programs to ensure that their activities are properly validated,
documented, and continually comply. The European System covering environmental
standards is often referred to as ISO 14000. ISO qualification generally results from
passing a one-time inspection with the applicants being left with the responsibility
of ensuring that their approved validated control systems are a starting point for
long-term improvement and that renewal audits show they continually comply.
In practice, responsible companies processing chemicals operate under an ISO 14000
framework or an equivalent. Their policy and procedure manuals detail the control
of activities all the way from receiving and storing chemicals to moving and processing
them and, finally, dealing with wastes. They cover worker protection (often
with Safety), equipment requirements and setup, chemical and solvent dispensation,
processing operations, emissions capture, intermediate and product isolations, and
wastewater/waste solvent treatment and disposal.
It is pertinent to provide an overview of the environmental concerns associated
with the major aspects of chemical processing in large-scale operations, particularly
pilot plants. These are process emissions to the air, chemicals’ handling, organic
process wastes, and wastewater.
Process Emissions to the Air
The final regulation covering hazardous air pollutants was published in the Federal
Register on September 21, 1998.2 At the time the laws were set up in the State of
New Jersey, I failed in efforts to gain an exemption for R&D pilot plant operations,
up to the final stage of defining a process that would be taken into large-scale manufacture.
My grounds were that in the early stages of development we needed to
delay work on environmental monitoring, data collection, and control issues until
we had determined the best chemistry to use and that, anyway, pilot plants work
with only small quantities and create little pollution. To capture this little pollution,
I proposed that we should set up a low-temperature system for trapping all emissions
(i.e., an overkill chiller system backed by scrubbers and/or carbon absorption),
protect operators in the most practical and aggressive way (in collaboration with
Safety), and drum waste solvent streams for incineration – I agreed that process
wastewater should meet the standards needed for discharge to the local Publicly
Owned Treatment Works (POTW). In this way I hoped to (a) delay the considerable
data collection and paperwork associated with reportable operations until the final
process selection had been made and (b) use the time and scientist/engineer/operator
expertise we would free up to responsibly and diligently speed process selection,
including factoring in related environmental issues into the equation. Unfortunately,
this did not happen because no one saw an easy way of separating out R&D from
our small-scale manufacturing activity. In short, we fell afoul of a strict interpretation
of the regulation (reference 2, p. 50294). Therefore, before every pilot plant run we
made, we were obliged to carry out “pre-” and “post-emissions” calculations. This
involved calculating the vapor emissions from the chemical reactions being undertaken
over the time of an operation, at all temperatures and accounting for all gas
blanketing. The “post-emission” calculations were done to account for any deviations
in the operating procedure during the actual running of the process. The objective of
carrying out emissions calculations in the first place was to ensure that emissions did
not exceed the capabilities of our scrubbing systems. I eventually only succeeded in
eliminating the “post-emission” calculations when no significant process deviation
The control of emissions to the air both protects process operators and preserves the
quality of the air for local communities. In addition, controlminimizes any impact that
volatile organic compounds (VOCs) may have on the larger environmental picture
(including ozone layer depletion and global warming). The New Jersey State and
Federal environmental control programs also require control over so-called fugitive
emissions (leaks from piping flanges and valves for example).
240 CFR Parts 9 and 63, pp. 50279–50386.
The protection of workers has been well dealt with by OSHA. Limits on worker
exposures allowed have been published3 in the form of threshold limit values (TLVs)
and time-weighted averages (TWAs), the latter being the amount of substance a person
can be exposed to in a normal 8-hr day in a 40-hrwork week. In both production plants
and pilot plants, emissions’ capture is most usually achieved by coupling reaction
vessels to a scrubbing system, a carbon bed absorption system, a low-temperature
condensing system, or a combination of these. Incineration and catalytic oxidation
are also used to destroy emissions. In opening reactors, companies frequently employ
“elephants’ trunking” exhaust for emissions’ capture at the open manhole, but worker
protection can extend all the way to the use of “breathing-air” suits when particularly
noxious materials are being used. In manufacturing practice, programs are set up to
obtain emissions data for all chemical process operations to ensure that the plant is
operatingwithin its permit. This is often a one-time operationwith occasional audits of
routine production to assure compliance. Frequently, purpose-built emissions control
units are installed to capture process emissions.4
Manufacturing processes generally meet environmental requirements, but a restless
search is always going on to improve them (e.g., reducing solvent usages and
waste volumes). I recall one such effort in the manufacture of albuterol (salbutamol),
wherein work to introduce a simpler, environmentally beneficial and lower-cost process
led to air emissions problems which spurred further beneficial changes.
The old process for the manufacture of albuterol5 requires the use of a number
of chemicals, particularly formaldehyde and bromine, which are classified as extremely
hazardous substances (and subject to reporting requirements under EPCRA
and CAA—see later). Clearly, we preferred not to handle these. At the time, albuterol
manufacture, because of its small volume, was being carried out in Chemical Development
plants. Because we were enjoined to reduce wastes as part of our RCRA and
EPCRA commitments, we initiated a search for a better and simpler process. Such a
process evolved from our work on themanufacturing process for dilevalol hydrochloride
(q.v.). Scheme 1 summarizes our first new process starting with low-cost methyl
High-quality albuterol was obtained in good yield from this process. However,
several environmental disadvantages were identified. The preparation of the keto
aldehyde hydrate (KAH) generated dimethyl sulfide, methyl bromide, and trimethylsulfonium
bromide (this compound sublimed in the condenser). In addition, reduction
of the Schiff base with dimethylsulfide borane, although very attractive in simplifying
3Handbook of Chemistry and Physics, 78th edition, Lide, D. R., Ed., Chemical Rubber Company, Boca
Raton, FL,1997–1998, Section 16, pp. 32–28.
4The Robinson Brothers Ltd. Manufacturing Plant inWest Bromwich, England, has a ring main emissions
control system around its plant for capturing odorous sulfur compounds from its processes, for central
treatment and/or incineration. Robinson works on the principle that a good environmental standing in the
local community is easily lost and hard to regain.
5Kleeman, A., Engel, J., Kutscher, B., and Reichert, D. Pharmaceutical Substances, Syntheses, Patents,
Applications, 4th edition, Georg Thieme Verlag, Stuttgart, 2001, p. 1849.
6Tann, C. H., Thiruvengadam, T. K., Chiu, J., Green, M., McAllister, T. L., Colon, C., and Lee, J. U.S.
Patent 5,283,359, 1994 (to Schering Corp.).
in CH2Cl2
1. DMSO/c.HBr
PriOH @ 88–90 C
2. aq. H2SO4, 70–75
under vacuum
H2NBut (CH3)2S BH3
SCHEME 1. Process for the preparation of albuterol from methyl salicylate.
the process by achieving three reduction steps in the one pot, gave odorous dimethyl
sulfide as a byproduct. Our proposal was to source KAH from a third-party supplier
(instead of the 5-bromoacetyl methyl salicylate produced for the old process). However,
scale-up of the Scheme 1 process to KAH revealed considerable difficulties
in dealing with the sublimed trimethylsulfonium bromide and in accommodating
the high costs for the pollution control equipment required to remove the unreacted
dimethyl sulfide and methyl bromide. In addition, traces of 3-bromo KAH were
found in the KAH produced by the Scheme 1 process. The third-party’s price idea for
KAH, from initially appearing attractive, escalated considerably because of the added
charges calculated as needed to depreciate the cost of emissions control equipment.
In spite of this, the Scheme 1 process provided a foundation for Sepracor, Inc., to
build a chiral process route to synthesize the allegedly more active (S)-albuterol.7
As far as the conversion of KAH to albuterol was concerned, our production
colleagues in Ireland (Drs. Brian Brady and Maurice Fitzgerald) were able to find
practical process conditions under which sodium borohydride replaced dimethylsul-
fide borane, with the same quality and yield result.
Nevertheless, the setback on the cost of KAH coupled with the relatively low kilo
requirements for albuterol and plans to move production off shore quenched interest
in the Scheme 1 process despite my rearguard laboratory efforts (with our Dr. C. H.
Tann and Dr. Beat Zehnder, Fachhochschule Nordwest Schweiz) to identify the basis
of an alternative much cleaner route into KAH (Scheme 2).
7Gao, Y., Hong, Y., and Zepp, C. M. U.S. Patent 5,442,118, 1995 (to Sepracor Inc.).
in CH2Cl2
1. O3/CH3OH/H+
2. Aq. H2SO4
SCHEME 2. Proposed new synthesis of ketoaldehyde hydrate for albuterol.
The principal raw materials were low in cost: Methyl salicylate was $5.50/kg and
crotonic acid was $3.85/kg in 1996.
Although there was useful literature precedent for crotonoyl chloride acylation8
of bromo methyl salicylate, there was none for the ozonolysis reaction. There was
speculation that ozone would open the phenol ring, but for my part I argued that
the carbonyl groups would inhibit this. Based on my oft-stated homily that one
should “never allow theory to abort an experiment,” our Dr. Tann showed that the
crotonoylation reaction in Scheme 2 could be carried out directly in apparently high
yield with aluminum chloride in methylene chloride and that ozone did not open the
phenol ring.
The ozonolysis step was studied in Switzerland using methanol-methylene chloride
as the solvent at ?15?C to ?70?C. The starting methyl 5-crotonylsalicylate
was fully consumed in less than 30 min. The presence of KAH was determined by
thin-layer chromatography (Figure 1).
The two ozone-resistant impurities in the starting methyl 5-crotonoylsalicylate
were speculated9 to be
The principal low-level impurity was later identified to be the keto acid produced
by oxidation of the ketoaldehyde. A very small amount of the benzoic acid resulting
from hydrolysis of the methyl ester could also be detected.
It appeared to us that given a successful development effort, the methyl 5-
crotonoylsalicylate route to KAH would eliminate all the air emission problems
associated with the original manufacturing process for albuterol intermediates as
8(a) Kawano, S., Komaki, T., and Watanabe, H. Japanese Patent 44077571, 1969 (to Eisai Co. Ltd.). (b)
Kono, S., Komaki, T., and Watanabe, H. Japanese Patent 43013619, 1968 (to Eisai Co. Ltd.).
9Suggestion of Dr. J. Gosteli, Cerecon AG, CH-4416 Bubendorf, Switzerland. Dr. Gosteli also reasoned
that the stability of methyl salicylate in the Friedel–Crafts reaction may be understood by regarding it as
a vinylogous carbonate, noting that anthranalate esters are also slow to hydrolyze.
Thin Layer Chromatographic Analysis
By-product 1 (contaminant)
Methyl 5-crotonyl salicylate
Methyl 5-dihydroxyacetylsalicylate (KAH)
By-product 2 (contaminant)
Unidentified, isolated by-product A.
E : Methyl 5-crotonylsalicylate (Marker)
6-24’ : Ozonolysis time
Stationary Phase
Mobile Phase
Amount charged
: HPTLC-Plate
: Dichloromethane/Toluene/Ethylacetate 5:2:1
: 1µ , streak
: UV 254
: Marker solution : 5 mg. Methyl 5-crotonylsalicylate in 2 ml.
: Reaction Solution : 0.5 ml Reaction solution in 2 ml methanol
E 6’ 12’ 18’ 24’
FIGURE 1. Thin-layer chromatogram of the solution obtained by the ozonolysis of methyl
well as those described above for the Scheme 1 process. However, because of other
priorities/plans and the relatively low return on investment, support for exploring the
new process lead withered away.
In the broader scheme of things, air emissions issues and especially setting limits
on the amounts released, continue to spur debate. For the chemist and engineer
developing a process in a responsible company, emissions need to be taken into
account. As seen in the foregoing albuterol process example, air emissions can
sometimes govern selection of the process to be developed for larger-scale use.
Alternatively, one can work hand in hand with a third party who has the demonstrated
capability to deal with specific emissions. In transferring technology to third parties,
emissions calculations are always very helpful, even early in the life of a project, in
establishing a good rapport with the third party.
Chemicals Handling and Organic Process Wastes
Chemicals Handling. The chemicals of most concern to all workers (including
those who might become exposed—e.g., transporters via a spill) and to the EPA are
categorized in published lists. Thirteen chemicalswith proven carcinogenic properties
are regulated by OSHA. These are:
Methyl chloromethyl ether
Bischloromethyl ether
3,3’-Dichlorobenzidine and its salts
Further additions to this list can be envisaged by structural implication—for example,
tolidines, propyleneimine, nitrosoethylmethylamine, and so on. In practice, every
effort is generally made by companies to avoid using these compounds.
Beyond the list of 13 substances, a larger list of chemicals has been consolidated
and is subject to reporting requirements under EPCRA and CAA (see footnote
1). Most are classified as extremely hazardous substances (EHSs) and are subject
to limits on the quantities allowed on site (threshold planning quantities—TPQs).
The use of EHSs requires that documentation, training, surveillance, and emergency
planning protocols are created for dealing with everyday use and inadvertent
Hazardous wastes containing listed toxic substances and wastes that are reactive,
ignitable, corrosive, or toxic are covered under RCRA.
New chemical substances have to be qualified for use and transport by building
a database that becomes more comprehensive as the quantities handled grow. In the
United States, the shipment of small quantities (e.g., analytical samples) can be done
with a very limited MSDS (especially if the sample is judged to pose little risk).
Substances thought to be very toxic or carcinogenic need more testing, including an
Ames test and an acute toxicity test (e.g., LD50—lethal dose at which 50% of the
rats/mice used in the test die). TSCArequirements for the shipment of larger quantities
are similar to those used in Europe. An outline of European testing requirements for
the Notification of New Substances (NONS) up to the level of one tonne/annum (or
five tonnes, cumulative) is given in Table 1.
Increased production levels require additional toxicology and ecotoxicology data
to ensure that prolonged exposure effects are understood.10 Today, the ELINCS
system is being progressed under European proposals labeledREACH—Registration,
Evaluation and Authorization of Chemicals.
The above indicates that as a result of the increased knowledge base on chemical
substances, chemicals handling has become more formalized. Gone are the times, as
in my school days, when one could go down to a local chemicals supplier to buy
small volumes of chemicals such as mineral acids and chemicals beyond the usual
commercial chemistry set—for example, ammonium dichromate for the “Green Tea”
10It is appropriate to add that there is a growing effort to minimize the use of animals in all chemical
testing programs.
TABLE 1. Outline of Testing Requirements to Qualify a New Substance for Inclusion in the European List of Notified Chemical Substances
Kilo Threshold Physicochemical Toxicology Ecotoxicology
Up to 10 kg  Chemical identity (Spectra/HPLC)  Acute toxicity (1 route) —
 Physical state
10–100 kg As above, plus:
 Flash point  Additional acute toxicity  Acute toxicity to Daphnia
 Flammability  Ease of biodegradability
100–1000 kg As above, plus: As above, plus: As above
 Melting point  Skin irritation
 Boiling point  Skin sensitization
 Water solubility  Eye irritation
 Partition coefficient (octanol–water)  Bacterial cell mutagenicity
 Vapor pressure
1000 kg to cumulative level
of 5000 kg/manufacturer
As above, plus: As above, plus: As above, plus:
 Relative density  Acute oral toxicity  Acute toxicity to fish
 Surface tension  Acute dermal toxicity  Algal growth inhibition test
 Explosivity  Acute inhalation toxicity  Bacterial respiration inhibition assay
 Auto-flammability  Subacute (28-day) toxicity  Hydrolysis screening
 Oxidizing properties  Bacterial cell mutagenicity  Soil absorption/desorption screening
experiment, poisonous mercuric thiocyanate for creating “Pharaoh’s Serpent,” or
magnesium ribbon, aluminum powder, and ferric oxide to demonstrate the “Thermit
Reaction” in your back garden.11 Sadly, even the spectacular Lassaigne nitrogen
test, once used in college chemistry courses as a test for determining whether your
unknown organic substance contained nitrogen, has also been consigned to chemical
history; there was always an excitement to fusing your organic unknown in a Bunsen
burner flame with a small “pea” of sodium, quenching the red hot fusion tube in
water(!), and then adding a ruby red sodium nitroprusside solution and watching it
turn Prussian blue if cyanide had been produced from nitrogen in your unknown!
As an aside, one wonders whether loss of some of this old magic of chemistry
has unwittingly dulled curiosity and enthusiasm for chemistry; however, given the
insurance costs to cover legal liability, and probably a declining market for Bunsen
burners, it had to be.
Although the professional chemist can still go to the company or university store
or the Catalogue Supply houses for laboratory chemicals, control over the purchasing
and storage of chemicals for pilot plant and plant use has become a highly organized
and paper (or electronic)-intensive operation.12 In regular commerce, warehouses
are built for chemicals storage and divided into separate areas [receiving,
in-process intermediates, APIs, quarantined chemicals, hazardous chemicals, and
chemicals (wastes) for disposal]. The warehouse is governed by operating procedure
manuals with particular attention being paid to the analytical status of every
The removal of chemicals from the warehouse and their processing in pilot plants
and plants is controlled through Standard Operating Procedures (SOPs) protecting
both the operators and the environment. SOPs are also developed for dealing with
The major initial concerns of process development chemists lie in the selection
of the process chemistry to be used, with safety, product quality, cost-of-goods, raw
material/intermediate sourcing, process equipment requirements, and speed of implementation
being the primary driving forces early in process selection. Environmental
issues are an early consideration in process selection only when the use of environmentally
noxious chemicals is proposed. Such chemicals are those that are acutely
toxic (e.g., methyl isocyanate and phosgene), those affecting operator/community
health (carcinogens, vesicants and lachrymators), those that may seriously compromise
air quality (odorous chemicals, NOx, SOx , and ozone depleters), and those
posing wastewater disposal problems (bactericides, heavy metals, ammonia, and
phosphates). Some can be readily scrubbed (NOx, SOx). The use of noxious chemicals
especially on a larger scale is often best left to third parties with facilities for
handling them (e.g., phosgene).
11A large crystal of (NH4)2Cr2O7, when ignited, burns spectacularly to produce a green volcano of
Cr2O3. The ignition of a trail of Hg(SCN)2 causes decomposition with astonishing swelling. Iron objects
are welded together by the intense heat generated when Mg ribbon in a mixture of Al and Fe2O3 is ignited.
12Unfortunately, the systems are not yet good enough to stop the acquisition of raw materials for making
explosives by terrorists and chemicals for illegal drug manufacture.
Many hazardous and environmentally undesirable chemicals are in use in commerce
and are appropriately controlled until another generation of chemical processes
evolves (often in cost reduction exercises, or in securing a desirable patent
position) to displace the original. The manufacture of 7(R)-amino-3-methylceph-
3-em-4-carboxylic acid (often referred to as 7-ADCA) is a case in point where the
original carboxyl protecting group, p-nitrobenzyl(using a vesicant, p-nitrobenzyl bromide,
for esterification), was superseded by diphenylmethyl (using in situ diphenyldiazomethane
for esterification), which in turn was superseded by the trimethylsilyl
group (see Chapters 7 and 9).
Organic ProcessWastes. Pollution from all industries, including that from chemical
and pharmaceutical manufacturing plants, has over the years raised an enormous
public outcry. Long ago, even in the so-called “developed world,” industry was
primarily concerned with the chemicals they could sell and paid relatively little attention
to the wastes they produced. In my time, I can recall the Grand Canal on
fire in Mexico City, resulting from dumped, organic-solvent-contaminated waste being
set alight, and I remember the local “fallout” when irresponsible waste haulers
in England illegally dumped spent fermentation waste in country woods. All of us
are aware of acid rain, toxic sites, and their clean-up and have read of rapacious
manufacturers around the world with little concern for anything other than maximizing
their “bottom lines,” dumping their wastes wantonly or illegally in landfills, in
rivers and in the sea, or wherever they could. Rightly, such environmental atrocities,
often compromising public health, led to government controls, massive fines, and
even jail time for perpetrators in those countries that have enforceable environmental
Today, most manufacturers are responsible and have been “ahead of the curve” for
many years in waste recycling, treatment, and disposal. A waste-avoidance culture
is also emerging in the selection of processes to be developed for use on a manufacturing
scale. The aforementioned albuterol process work at least testifies that
environmental issues are being raised and, in the 7-ADCA case, that dirty processes
(the use of p-nitrobenzyl protection) are being replaced by cleaner ones (trimethylsilyl
The reality in developing and operating chemical processes on a pilot plant scale
is that waste treatment goes on continuously as part of the process, particularly to
render wastes safe for disposal. Thus activated carbon cakes (say, from a decolorizing
step), or a spent hydrogenation catalyst, often need to bewashed to remove flammable
solvent and treated withwater to render them safe for disposal.Organic processwastes
are frequently disposed of by incineration, often after a simple solvent stripping
operation is used to recover a volatile solvent for recycle. Solvent recycle, which
helps to meet environmental goals to reduce chemical usage, needs to be done with
an eye to quality. This is particularly important in the last steps of a process—the
recovered solvent needs to be subjected to careful analytical screening to ensure that
the quality of the API, and even late intermediates, is the same as when fresh solvent
is used.
There are other ways of reducing solvent (chemical) usage such as:
 Increasing reaction concentration (and with it plant productivity). An ultimate,
if rarely achievable, goal would be to run the reaction without solvents.
 Harmonizing solvent usage within a plant by switching the solvent used in a
process to one already established (and recovered) in the manufacturing plant
receiving the technology.
 Redesigning the process to reduce the number of steps and solvent usage. An
ultimate achievement in redesign would be to use water as the solvent and
enzymes to carry out desired transformations.13
In a process development/improvement setting in the pharmaceutical industry, the
process changes identified above are all likely to require regulatory approval before
adoption, especially for late stages in the synthesis of an API (see Chapter 6).
There has been considerable growth of interest in so-called “Green Chemistry”
or “Sustainable Chemistry” over the last quarter century. The terms “Green” and
“Sustainable” have given new prominence to fermentation and enzyme-mediated
processes and to systems that operate in water. Such processes build on the already
major contribution that fermentation processes make to the pharmaceutical industry.
As an aside, several important classes of API owe their commercial success to the
fermentation of microorganisms:
Antibiotics: Penicillins and cephalosporins
Anticancer agents: Nucleosides (e.g., Bleomycins)
Cholesterol biosynthesis inhibitors: Statins (e.g., Lovastatin)
Some vitamins B12, Riboflavin (some synthesized)
Fermentation is also the basis for the manufacture of biomass foodstuffs (primarily
protein for animal and human consumption), amino acids (especially monosodium
glutamate and l-lysine), and the major industrial feedstock and gasoline additive,
More specific and growing uses of microorganisms are in the areas of finding
enzymes for specific tasks, especially if one can integrate with established biological
transformations. Thus, the Antibioticos success in harnessing an amino acid oxidase
to convert the aminoadipic acid side chain of cephalosporin C into a glutaroyl side
13More speculatively, “cascade chemistry” has been proposed as an environmentally friendly approach to
chemical synthesis. Process design mirrors nature in that the process of producing an intermediate leads
to an active product that progresses to a further activated intermediate and so on down the cascade to a
needed product; see Hall, N. Science, 1994, 266, 32.
chain enabled them to tap into established amidase process systems to create an
all-aqueous process for the manufacture of 7-ACA14 (see Chapter 9).
There is little doubt that microbiological approaches to chemical transformations,
and not only chiral transformations, are likely to grow substantially as the wasteavoidance
culture becomes more established, and the laws are fully addressed.
In the early days of companies coming to terms with environmental laws, especially
laws concerning process wastes, there was considerable overreaction by all
parties involved—that is, internal company lawyers, waste haulers, and government
inspectors charged with ensuring compliance. A pair of cautionary tales illustrate the
disconnects that can occur in setting up new systems.
Case 1: Incident in the Early Days of Dealing with Process “Wastes”. We in Chemical
Development always felt that because of our knowledge base we were the right
people to determine the best way of treating organic process wastes before disposal,
regarding such work as part of any program to develop a process for use in a manufacturing
plant. We felt even more strongly about this when it was decided that five
drums of six-year-old lithium hexamethyldisilazane (prepared by us and erroneously
labeled “Hazardous Waste”) should be disposed of (a waste hauler quoted $100,000
for taking it away!). We relabeled the drums and neutralized the contents by adding
them to cooled dilute sulfuric acid, disposing of the aqueous salts to the sewer and
drumming the resultant relatively innocuous hexamethyl disiloxane for regular disposal.
The Legal Department took us to task for relabeling the drums and for treating
the waste without a permit. The Legal Department conducted an investigation; and
based on their interpretation of the guidelines being adopted by the State of New
Jersey and the self-reporting philosophy they had embraced, they sent a full report to
the New Jersey State’s Department of Environmental Protection confessing to a perceived
illegality. The DEP made no issue about the case and it was dropped. Clearly,
the Legal Department’s interpretation of the embryo regulations was in error.
Case 2: Another Over-Strict Interpretation of the Wording of the Laws by a Government
Inspector. I heard of this case on the radio in England during “Gardeners
Question Time” one Sunday afternoon in June 1988. In creating the waste disposal
laws, Parliament allowed that agricultural waste such as cow manure from farms
could be spread on fields as a fertilizer, recognizing that this was, anyway, a longstanding
practice. However, possibly useful industrial waste such as water waste from
scrubbing ammonia could not. A large amateur gardener’s allotment operation for
growing vegetables and flowers had for years been taking deliveries of horse manure
from a local racing stable. After auditing the racing stables, an officious environmental
inspector came to the allotment and issued a court summons on the grounds that
they were illegally disposing of industrial waste. It transpired that the stables had
registered themselves as being in the horse racing industry and that race horse manure
had been reclassified as an industrial waste! After the initial shock, the dilemma was
resolved by allowing that the stable manure could be disposed of to the allotment if
14Cambiaghi, S., Tomaselli, S., and Verga, R. U.S. Patent, 5,424,196, 1995 (to Antibioticos, S.p.A.).
the allotment operators took charge of horse manure pick-up rather than continuing
the previous practice wherein deliveries were initiated by the stables!15
As the waste-avoidance culture develops, it is expected that other environmentally
friendly technologies will emerge to, directly or indirectly, aid in finding better chemistry
or avoid or reduce process wastes. Several of these technologies are outlined in
the presentation on the future, including the use of polymer supports, electrochemistry,
chemistry in greenhouse gases, and process hydration. Since waste reduction is
a contributor to cost reduction, it is already in the mainstream of chemical process
development activities for large-scale operation. It can thus be seen that, as well as
being a socially responsible activity,waste reduction is a contributor to cost reduction,
bringing it into the mainstream of chemical process development considerations.
Wastewater. Most pilot plants have a permit to treat (commonly by adjusting pH
and removing volatile solvents) and discharge their waste water directly to a publicly
owned treatment works (POTW) or, if the pilot plant is on a chemical manufacturing
site, to an internal waste water treatment facility prior to discharge, under strict permit
control, to a POTW or a natural body of water. As chemical process development
work progresses to the stage where a process becomes a candidate for transfer to a
manufacturing site, wastewater issues need to be addressed at the pilot plant level
and with the manufacturing site taking over the project. Few of the large number
of organic chemicals in use for API manufacture are the subject of guidelines on
permissible levels that can be discharged in wastewater in the United States. Those
that have been listed16 are summarized in Table 2.
Frequently the manufacturing site will evaluate the waste water stream for compatibility
with their existing waste water treatment/disposal systems and particularly
to determine that the microbes used in COD/BOD reduction can accommodate the
new wastewater when diluted in their existing wastewater feed. The most important
parameters to control for wastewater disposal are pH, the volatile organic compound
content, certain soluble salts and suspended solids, carbon oxygen demand (COD),
and biological oxygen demand (BOD). Of these, pH is usually the easiest to maintain
in the generally required range of 6–9 as it leaves the plant.
When the solvent content of the wastewater is high, incineration may be the
lowest cost option for disposal. However, VOCs are often stripped for reuse or
separate incineration. Soluble salts such as of widely used metals (iron, aluminum,
and chromium) and of commonly used anions (cyanide and fluoride) can pose waste
disposal problems. Excessive levels of such algal bloom promoters as ammonia and
phosphate introduce effluent problems on a large scale.
15The story on the disposal of manure took another turn when the EU decreed restrictions in the amount
of manure from pig, dairy, and poultry farms which could be spread over fields in “nitrate-vulnerable
zones”—ostensibly to protect drinking water, rivers, streams, and coastal estuaries. Greater restrictions
have been applied to farm waste disposal over sandy and shallow soils in Denmark and Holland in that
disposal is banned over the period August to November. The smell accompanying the resumption of
disposal in Holland can be detected on the east coast of England! Uhlig, R. The Daily Telegraph, March
12, 2002.
16Federal Register, 63, No. 182, Sept. 21, 1998, 50434.
TABLE 2. Guidelines on Permissible Levels of Common Organic Contaminants in Wastewater.
Pretreatment Standardsa Pretreatment Standardsa
Max. Daily
Avg. Monthly Discharge
Must Not Exceed Compound
Max. Daily
Avg. Monthly Discharge
Must Not Exceed
Ammonia (as Nb) 84.1 29.4 Benzene 3.0 0.7
Acetone 20.7 8.2 Toluene 0.3 0.1
Methyl Isobutylketone 20.7 8.2 Xylenes 3.0 0.7
Isobutyraldehyde 20.7 8.2 n-Hexane 3.0 0.7
n-Amyl acetate 20.7 8.2 n-Heptane 3.0 0.7
n-Butyl acetate 20.7 8.2 Methylene chloride 3.0 0.7
Ethyl acetate 20.7 8.2 Chloroform 0.1 0.03
Isopropyl acetate 20.7 8.2 1,2-Dichloroethane 20.7 8.2
Methyl formate 20.7 8.2 Chlorobenzene 3.0 0.7
Methyl Cellosolve R 275.0 54.7 o-Dichlorobenzene 20.7 8.2
Isopropyl ether 20.7 8.2 Diethylamine 255.0 100.0
Tetrahydrofuran 9.2 3.4 Triethylamine 255.0 100.0
a Mg/liter (ppm).
b Not applicable to sources that discharge to a POTW with nitrification capability.
The pharmaceutical industry, because of the enormous diversity of chemistry used
in the synthesis of APIs and their intermediates, probably carries more wastewater
treatment/disposal problems than any other industry. Fortunately, the relatively small
scale of production and the economic well-being of the industry has allowed all
kinds of accommodations that would be major problems on a very large scale. Thus,
gelatinous hydroxides of chromium and aluminum are frequently precipitated in
small lagoons or basins and dug out for landfill disposal. Where this is not possible,
imaginative alternatives are devised (e.g., Shionogi’s disposal of waste from
its aluminum chloride–anisole cleavage of cephalosporin esters is done via malic
acid chelation and the solution shipped to a licensed processor). Cyanide in aqueous
waste is usually oxidized to relatively harmless cyanate (by hydrogen peroxide,
alkaline chlorination, or ozone), and fluoride ion is generally precipitated as calcium
fluoride to reduce the fluoride concentration to a desired level (frequently
<5–6 ppm). Where water waste streams contain large-molecular-weight organic
compounds, membrane filtration technology (particularly ultrafiltration and reverse
osmosis) offers a useful technology for concentrating the waste stream to recover
water and an aqueous organic waste for incineration. This approach has been successfully
applied to the disposal of a polyol waste stream in a petrochemical plant.17
The primary means of reducing COD/BOD in industrial wastewater is via settling
basins and aeration in lagoons, biotowers, or large tanks. Tanks are frequently lidded
and equipped with a scrubber if noxious off gases are present. A cascade of lagoons is
often necessary to bring COD/BOD levels down to compliance levels for discharge to
a POTW or a natural body of water. At Bristol–Myers’ plant in Latina, Italy, the last
lagoon is covered in water hyacinth, which absorbs “nutrients” and, through harbored
aerobic and anaerobic microorganisms, also metabolizes polluting chemicals and
absorbs metals (particularly lead and zinc). In addition, the visual effect of a field of
water hyacinth at work, often with fish swimming about, is quite pleasing.18 Related
to this, treatment of industrial wastewater via manmade reed beds is now being
practiced on a very large scale using macroscopic plant life (including bulrushes and
reeds), often to valuable effect (e.g., the dechlorination of chlorophenols).19
Another illustration of the potential in harnessing plant life for soil remediation is
the finding20 that the fern, Pteris vittata,when grown in soil containing 6 ppm arsenic,
hyper-accumulated 755 ppm of this metalloid in its fronds in only two weeks. When
Pteris vittata was grown in artificially contaminated soil (1500 ppm As), the fronds
took in 15,861 ppm As in the same two-week time frame. Similarly, research in both
the United States and the United Kingdom has demonstrated the potential of using
plants from the family Brassicacae in the remediation of soils heavily contaminated
with zinc, cadmium, nickel, lead, and selenium.21
17Pearson, D. Chemical Processing, January 2002, 24. Website: www.chemicalprocessing.com
18However, it should be noted that the above-described Latina wastewater treatment system has now been
replaced by an activated sludge treatment system.
19Cobban, R., Gregson, D., Phillips, P. Chemistry in Britain, 1998, 40.
20Ma, L.Q., Komar, K. M., Tu, C., Zhang, W., Cai, Y., and Kennelley, E. D. Nature, 2001, 409, 579.
21Rouhi, A. M. Chemical and Engineering News, 1997, Jan. 13, 21.
Destructive Methods Separation Methods
Recycling Ag & Cr
Salt Splitting
Anodic oxidation
Cathodic reduction
Indirect Membrane
Processes Direct
SCHEME 3. Outline of major electrochemical wastewater treatment options.
More recently, growing awareness of the presence of APIs in wastewater has
drawn the attention of Environmental Protection Agencies in both the United States
and Europe, and it has also drawn the attention of the U.S. Geological Survey (USGS).
This has resulted from increased recognition that the following have come together
to reveal new issues: (a) wastewater from feeding large quantities of antibiotics to
livestock, (b) the common practice of flushing unused, outdated (and excreted) medications
down the toilet, and (c) the development of exquisite analytical methodology
to detect extremely low levels of APIs in water. For instance, the USGS has detected
(at the part per billion level) almost all of 95 selected APIs (mostly antibiotics, antidepressants,
anti-inflammatories, analgesics, antacids, and cardiovascular drugs) in
streams across the United States.22 In theUnitedKingdom, the Environmental Agency
(EA) has gone even further in one case. Thus the EA has called for water companies
in England and Wales to investigate sewage treatment technologies to effectively
remove estrogenic steroids from rivers. Work at Brunel and Exeter Universities has
indicated that 17?-ethinyl-estradiol (component of contraceptive pills) is having an
adverse effect on the reproductive ability of male fish, even at concentrations lower
than 1 ng/liter.23
A few other waste treatment technologies are outlined below.
Low-cost sources of electrical power have stimulated the widespread application
of a number of electrochemically based technologies in wastewater treatment
(Scheme 3).
The electrical generation of ozone is used in municipal water treatment. Ozone
is a very powerful oxidizing agent (its oxidation potential being exceeded only by
fluorine) and has the advantage of being about 12 times more soluble in water than
oxygen. Ozone also has the advantage over chlorine (still the most favored oxidant
22Hileman, B. Chemical and Engineering News, 2001, Dec. 3, 31. See alsoHTTP://PUBS.ACS.ORG/CEN
23Chemistry in Britain, May 13, 2002; and Burke, M. Chemistry in Britain, January 30, 2003. This work
is extending to wastewater from agricultural operations and aquaculture where steroid hormones are in
use; see Nicholls, H. Chemistry World, October 2004, 21.
for municipal water treatment, and itself generated by electrical means) in that its use
avoids the formation of chlorinated hydrocarbons from any organic materials which
may be present. As an aside, ozone is also used in destroying odorous gas emissions
and in chemical ozonolysis.
The electrochemical oxidation of metal ions (e.g., Ag+ >Ag2+) for the catalytic
oxidation of organic compounds has been practiced on a small scale.24 The feasibility
of recycling Cr3+ produced from Cr6+ in theMarker degradation of diosgenin acetate
is outlined in Chapter 11.
The direct electrochemical oxidation (no cell divider membrane) of wastewater
has been employed in the textile industry. Typically, this industry produces an
organic-contaminated wastewater that also contains sodium chloride; sodium chloride
is desirable in promoting anodic oxidation. The presence of sodium chloride is
fortuitous for textile manufacturers since the hypochlorite byproduct produced in the
electrochemical oxidation process is used for textile bleaching operations.24
Of the separationmethods, electrodialysis is themost widely employed, especially
in the removal of nitrates from water. The electrochemical splitting of sodium sulfate
in industrial wastewater streams has been employed to regenerate sodium hydroxide
(and hydrogen) at the cathode and sulfuric acid (and oxygen) at the anode, for use in
other processes, thereby greatly reducing the burden of disposing of sodium sulfate
waste. In a nonelectrochemical sense, membrane technology (particularly ultrafiltration
and reverse osmosis) is now used on an enormous scale for the purification of
brackish water. Membrane technologies will no doubt grow in importance as a means
of producing high-quality water from wastewater streams and, as already described
(see footnote 17) a low-cost way of concentrating wastewater in some situations.
It is worth mentioning the use of supercritical water oxidation as a means of destroying
organic compounds in complex effluents produced by the pharmaceutical
industry, since this technology has been evaluated for treating a variety of biotechnology
and chemical process wastes.25 When water is heated under pressure above
its critical point (374?C and 218 atmospheres), its character changes significantly.
Its dielectric constant and viscosity are greatly reduced, and it becomes an excellent
solvent for organic substances and oxygen. The SmithKline and Johns Hopkins
University workers25demonstrated that the technology, operated at 650?C and 252 atmospheres,
essentially destroyed all organic compounds, including microorganisms
and protein in recombinant fermentation broth.
As far as the environment is concerned, the pharmaceutical industry occupies a unique
niche since process development chemists, chemical engineers, and manufacturing
people generally deal with relatively small quantities of complex waste which, not
24Dr. Guillermo Zappi, private communication.
25Johnston, J. B., Hannah, R. E., Cunningham, V. L., Daggy, B. P., Sturm, F. J., and Kelly, R. M.
Bio/Technology, 1988, 6, 1423.
infrequently, contain quite hazardous chemicals. All involved in developing chemical
processes for scale-up to a manufacturing operation need to increasingly embrace the
air, chemicals’ handling, organic waste, and water waste issues as an integral part of
their thinking during the period of developing an API to the marketing stage. The
search for the highest goal, namely to find the simplest, safest, most environmentally
friendly, and lowest-cost process to produce a quality API, is frequently rendered extremely
difficult given the prevailing climate calling for the fastest possible delivery
of the needed material. Although the effort to meet all the regulatory requirements
inhibits the will to seek potentially better process options, chemical process developmentworkers
need the towering strength of purpose, along with management support,
to rise above the deadening weight of bureaucracy, as well as delivery and quality
goals, to seek the highest goal—inadequate compromises are better than nothing at
all, at least allowing a start to creating the next generation of processes. Outside
efforts with universities and research contractors often provide a good start.
However, compliance with the legal requirements outlined in this presentation
has to also be achieved. All changes in process chemistry, operating procedures,
and equipment need to be supported by environmental calculations showing that
comparison of emissions versus the original process are within the boundaries of
the operating permit and that consistency has been achieved. The onus is on the
manager to ensure that the calculations are without error and available on site for
environmental auditors. The most important factor as far as regulators are concerned
is that a written record has been kept documenting compliance with the laws.
Environmental matters need the same enthusiastic personal involvement as one
gives to safety and regulatory affairs matters in order to reduce risks, protect people,
and meet Food and Drug Administration requirements. When a major investment in
a new process is being proposed, it is fairly common for the manufacturing plant
management to initiate a dialogue with the local community by way of advice and
to gain feedback. Such considerations are vital in an industry that needs to build and
maintain a high standing in the local community.
Our objective is to have you build quality into your drugs, not test it in.
—Henry Avallone, FDA
Regulations, whether in the process safety, environmental, or food and drug field,
were originally introduced in response to significant events such as factory explosions
or catastrophic releases of environmentally damaging chemicals, or to stop
the promotion and sale of useless or potentially dangerous pseudo medicines. Once
established, regulations and regulatory agencies gained a life of their own, providing
a framework to accommodate refinements such as creating legislation in response to
safety risks or for eliminating potential causes of environmental damage, or in the
case of this presentation, for building quality into processes for producing APIs.
Henry Avallone’s compelling statement, made during a pre-approval inspection
(PAI) of a Schering–Plough Chemical Development manufacturing facility in the
late 1980s, left us in no doubt that the FDA wanted us to create a comprehensive
system to ensure that the manufacture of active pharmaceutical ingredients (APIs),
as well as dosage forms, was being undertaken to guarantee the quality of our

Copyright C 2008 John Wiley & Sons, Inc.
products. Although we thought we had been doing a good job in ensuring that our
APIs were of the highest quality, the first FDA inspection of one of our process
development/production operations1 made us aware of our inadequacies in light of
the FDA’s pursuit of quality2 and how they were working in their role as the public’s
champion to promote drug safety, efficacy, and quality.
The Food, Drug and Cosmetic Act of 1938, requiring that new drugs be tested
for safety, was quickly approved in response to a 1937 tragedy caused when the SE
Massengill Company used diethylene glycol, without testing it for safety, in its syrup
formulation of sulfanilamide – 108 people died, mostly children, from ingestion of
the glycol. In 1962 the Act was overhauled. Safety testing was made more rigorous,
and proof of a drug’s efficacy was added. The reach of the FDA’s mission was slowly
extended to cover the quality of APIs, with their key precursors and the synthesis
sequences perceived as having an effect on product quality (by introducing a liability
to cause API contamination). In the 1960s the FDA began to promote the concept
of Good Manufacturing Practice (GMP) as a foundation for ensuring API quality.
GMP has become cGMP (current Good Manufacturing Practice) to reflect continuing
evolution in quality assurance. The FDA also continues to refine its guidance role
and is addressing the future through an ambitiously titled document3 “Pharmaceutical
cGMP’s for the 21st Century: A Risk-Based Approach.” One can sympathize with the
process development chemist in the pharmaceutical industry who is still struggling to
fully understand and implement the broader principles and interpretation of system
needs for cGMP documentation, validation, and compliance, if he/she sometimes
feels like the Red Queen in “Alice in Wonderland.” Her observation “it takes all the
running I can do to keep in the same place . . .” mirrors the seemingly endless effort
needed to meet the high standards set by the FDA to protect public health. Few can
even deal with her next sentence “If you want to get somewhere else, you must run
twice as fast as that!”
Some, and especially those manufacturing APIs, argue that there have been relatively
few adverse public health effects, and that the in-depth focus on quality and
the excessive validation and documentation associated with conforming to the requirements
of cGMP is not justified. However, there have long been concerns about
contamination of APIs and drug products, whether these have been due, inter alia, to
impurities generated in the synthesis of the API, or via cross-contamination from the
air, or from the use of equipment also employed in producing other products. One of
the earliest examples concerned potential allergenic reactions to penicillin contamination.
This led to the complete segregation of penicillin (and later cephalosporin
and other related ?-lactams) production from other operations. It can also be seen
from FDA Inspection Reports on Pharmaceutical Companies that sometimes serious
“quality” excursions still occur, creating chronic problems and, rarely, serious events.
1For more detail see the case study on Dilevalol Hydrochloride: Development of a Commercial Process.
2The equivalent European, Japanese, and other agencies similarly strive, with leaner resources, to promote
drug safety, efficacy, and quality. These agencies, although not as far-reaching or aggressive in their
oversight as the FDA, contribute invaluable perspective—for example, through joint efforts to harmonize
regulations. Such efforts ensure that industry and its regulators are all working with the same script.
Examples from the drug product area and from the API production area illustrate the
continuing need for vigilance.
The variable potency of one manufacturer’s sodium levothyroxine tablets, prescribed
for hyperthyroidism, led to numerous adverse reaction reports, starting in
the 1980s. Some patients were getting too little drug and others too much. Underdosed
patients exhibited a greater incidence of fatigue, depression, fuzzy-headedness
and itching—some gained weight, others reported brittle hair! Overdosed patients
reported more muscle tremors, insomnia, heart palpitations, and heart rhythm abnormalities.
It took considerable time for the findings to play out to the point of
linking patient symptoms with variability of tablet potency. Ultimately, after much
dialogue with the mostly disbelieving manufacturer, efforts were initiated to qualify
other suppliers and phase out the now wayward manufacturer.4
Another case, much more serious and also more pertinent to the chemical process
development area, occurred in 1989 when over 1600 people became ill with
eosinophilia–myalgia syndrome (EMS) and 38 died, worldwide, after taking ltryptophan
(Trp) manufactured by one producer in Japan. Prior to the outbreak,
this producer whose Trp met the >98.5% purity specification had decided to employ
a new genetically modified strain of the established Bacillus amyloliquefaciens and
also to halve the amount of activated charcoal used in the purification step. These
changes cause the Trp product to become contaminated with several new impurities,
principally I to III, all associated to some extent (using a crude animal model) with
The total impurity content of the Japanese Trp was also greater than in the product
made by other producers. In order to be allowed back into production, draconian
changes had to be made to the manufacturing process. In addition, a reverse-phase
4During this phase, the wayward manufacturer suppressed contradicting information and argued that its
product was better than that of others. This led to a class action lawsuit against the wayward manufacturer
which was settled in August 2000 with a $100 million payout to affected consumers. The wayward
company was taken over by a large pharmaceutical company that resolved the problems and continues to
market the drug.
5(a) Simat, T. J., vanWickern, B., Eulitz, K. D., and Steinhart, H. J. Chromatography, B: Biomedical
Applications, 1996, 685, 41. (b) Simat, T. J., Eulitz, K. D., and Steinhart, H. GIT Fachzeitschrift fuer das
Laboratorium, 1996, 40(4), 339.
HPLC analytical specification for I (<8 ppm) was introduced. The sum of detectable
contaminant peaks eluting prior to Trp was reduced to <100 ppm and the sum of
those eluting after Trp was reduced to <300 ppm. This case, more than any other,
has served to reinforce the FDA’s position on the vital importance of building quality
into process operations.
It seemed for the longest time that the chemist’s search for high-yielding processes
giving the highest-quality intermediates and APIs was dependent on no more than
scientific common sense. The term scientific common sense is, however, too vague,
open-ended, and subjective to be embraced as the sole foundation of quality. Even
given that the instruments used by the scientist to determine API quality are rigorously
maintained and calibrated and that the API analytical standard is impeccable, how
many chemists have found (to their consternation) that an HPLC trace on their API,
say when inadvertently run out over the lunch period, has revealed unsuspected peaks
at unacceptable levels? How many have found that they not infrequently have needed
to recrystallize an API more than once to achieve the desired purity and yet failed
to seek the reasons why the level of causative impurity had risen to require multiple
recrystallizations? How many have introduced a raw material from a new supplier, or
tweaked a process step, or changed to a lower-cost source of solvent, or reduced the
level of carbon used for color removal, or made some seemingly innocent change,
often in the name of convenience or cost reduction, only to create some unanticipated
problem? Again, how many chemical engineers have transferred a process from one
vessel to another with, say, a different stirrer configuration, or a different temperature
control capability, only to find an adverse effect on product quality? Such occurrences
may not happen every day, but they happen frequently enough to make the concept
of building quality into an API something of an imperative.
This presentation provides an outline of many facets of the work undertaken
by chemists and engineers to produce an API of acceptable quality, in a way that
satisfies regulatory requirements. It is not intended to be a guidance document outlining
all the activities that must be undertaken to satisfy every detail of the requirements.
In this sense it is incomplete. The purpose is to give the process development
chemist/engineer an overview of the combination of science, technology, and quality,
which is the basis of assuring that API production will meet FDA requirements.
The science/technology/quality combination essentially provides the information for
the Investigational New Drug (IND) and New Drug Application (NDA). The work
needed to create these documents is summarized, leading to a Chemistry, Manufacturing,
and Controls (CMC) document which is the foundation of the chemical process
development contribution to the NDA. Before the NDA is approved, the company is
subjected to a Pre-Approval Inspection (PAI). In order to initiate this, the chemical
process development staff concerned with the project works with others as needed
(particularly the company Regulatory Affairs, Manufacturing, and Quality Control
groups) to provide additional information for the FDA to commence the PAI inspection.
The principal document is the Development Report. The other major interests
of the FDA are Technology Transfer and Validation.
The undertaking, from the beginning, needs to adapt to the continuum of change
which is the reality in developing a process to manufacture an API. Initially, most
effort is devoted tomodifying the usually rawresearch API synthesis scheme (Recipe)
to eliminate the obvious safety hazards and thereby make it acceptable for scale-up to
supply API for Toxicology and Pharmacology programs. These two most important
disciplines essentially decide whether the subject API is sufficiently safe and effective
to be considered a potential drug development candidate and to determine whether
metabolites are factors and what program should be followed in further development.
The initial quantities of API needed for the analytical and pharmaceutical dosage
form disciplines are also produced by chemical development in this early phase of
scale-up and process development, though scale-up difficulties often limit API supply.
The major governing factors in the process development program to produce
quality APIs within cGMPs (and provide for regulatory submissions) are summarized
under the following headings:
 Building a Quality System
 The Toxicology Batch
 Establishing the API Quality Specification and the Last Process Steps
 The R&D Work Needed to Define the Synthesis Methodology Before the IND
Filing (including solvent, raw material and intermediate quality considerations)
 Creation of the CMC Section for the FDA
 The Pre-Approval Inspection (PAI)—The Development Report, Technology
Transfer and Validation
Building a Quality System
This starts with assembling an organization of people not only capable of meeting
the technical needs in producing an API, but also educated and trained to provide the
documented record that they have done so, with full attention paid to assuring the
quality of the API and the quality of the data. This needs investments in analysts and
analytical instrumentation, in laboratories, in a well-equipped pilot plant containing
a controlled environment room, and in a comprehensive warehouse facility. It is also
desirable, particularly in the preparation of parenteral or inhaled drugs, to invest in a
water purification system if the API is to be prepared in water or in a water-containing
medium [alternatively, capital investment can be avoided by the relatively expensive
purchase of water-for-injection (WFI)].
As a general rule, the analysis of APIs is undertaken by an independent analytical
research/quality control (QC) unit. Nevertheless, in my experience, the chemical
development organization greatly benefits from having its own QC organization to
provide analytical support on a fast-track basis. This organization, which maintains
a documented profile on all of its instruments (including operating procedures and
calibrations as well as files on all standards), undertakes the rapid well-documented
analysis and labeling of raw materials and intermediates and plays a vital role in
the management of the chemical development warehouse. The warehouse should
have areas for receiving and weighing chemicals, segregating them according to their
status, and for the storage of final APIs—refrigerated if stability tests indicate a need to
do so. Thewarehousing operation is governed by amanual of operating procedures. In
a corresponding way, the chemical development organization benefits from assigning
and training one of its employees to become a regulatory affairs “expert” working
to represent the chemical development point of view with the central regulatory
affairs organization (which interacts with the FDA). Both of these satellite QC and
regulatory affairs operations have dotted line relationships with their respective central
organizations. Although somewhat controversial, these organizational arrangements
continue to work well in Schering–Plough (see Chapter 3).
All personnel, including pilot plant management and operating personnel, need
to prepare and continually update curriculum vitae documenting their qualifications,
experience, present job descriptions, and their training in cGMPs. This record shows
that those producing APIs are qualified to do so. It also recognizes that pilot plant
personnel contribute invaluable process observations and detail, and essential input
in maintaining the documentation (batch records, standard operating and cleaning
procedures, etc.), as well as equipment upkeep and equipment logs (history, operating
instructions, maintenance, and calibration records).
The Toxicology Batch
The chemical development organization usually produces the proposed API for the
animal toxicity tests to ensure that the API is both safe and can be safely administered
to humans in the Phase I human safety study. Since toxicology tests are among the
first activities undertaken in the drug development process, it follows that a great
deal of attention is paid to the quality of the toxicology batch, including the impurity
profile. Ideally, one would like to use the same quality API in the toxicology tests
as would eventually be produced in the final manufacturing plant. However, since
the final process and the API quality are usually unknown, all that can be done is to
produce the API by the most practical process available at the time. The early focus
on quality generally leads to a major effort being disposed to establishing the last
process steps (and particularly the final step) first (see next section: “Establishing the
API Quality Specification and the Last Process Step(s)”).
In producing API by the most practical process available, one generally strives to
achieve an API quality of ?98% and to identify all the impurities present in >0.1%
amount. In practice, especially if the anticipated dose is likely to be high (as say
with an antibiotic) or an API is to be delivered to a sensitive organ (e.g., the lung
by inhalation), we have generally identified all impurities >0.05% and in one case
I recall, >0.02%. There is nothing sacrosanct about ?98%. I recall that the purity
specification on one of our antibioticswas set at?97%. In practice, the impurity levels
in toxicology batches are usually higher than those in the final marketed product. This
provides chemical process development with some flexibility since the FDA readily
accepts changes to lower levels of impurities. It is very difficult to gain approval for
higher levels without further toxicology tests.
Usually, in progressing process development work, one eventually finds a better
process (e.g., lower in cost and/or safer ormore environmentally friendly) that creates
a different profile of impurities. Since no company is likely to sanction delays in its
drug development program by undertaking the added cost of further toxicology work,
a better process is generally only acceptable if the new impurities in the API can be
held to <0.1%. There is some flexibility. A few years ago, because of the urgent need
to progress a chiral antifungal candidate, preliminary toxicologyworkwas undertaken
on a several-hundred-gram sample of the desired enantiomer (purity 100%) prepared
by separation using chiral chromatography. Limited acute toxicology work was also
carried out on the pure unwanted enantiomer. The agreement with Toxicology was
that additional toxicologyworkwould be undertaken onmaterial produced later using
a classical resolution process to qualify the API containing the unwanted enantiomer.
In this case, the unwanted enantiomer content of the classically prepared product
appeared likely to be at a level of about 1–2%. This dialogue with Toxicology
indicated that impurities that are close in structure to the API are more likely to be
similar to the API in toxicology profile, and therefore more acceptable as impurities,
than those impurities that have large structural differences versus the API. However,
the <0.1% new impurity limit is still generally preferred for progressing a new
synthesis option to IND filing. In the above case we did not have the opportunity to
establish the resolution process because the particular antifungal was dropped.
In practice, it is fair to say that chemical development organizations have coped
fairly well with the inhibitions to change, which are the fallout of adopting the ad
hoc “specification” set in using the toxicology batch. Nevertheless, because of such
inhibitions, companies have undoubtedly restricted opportunities to find the best
(and lowest cost) commercial process for manufacturing the API in the interest of the
fastest possible rate of development of a drug to the marketplace. In my experience,
the inhibition to change has adversely affected process research. There may be ways
of changing this situation, which I will address in Chapter 11.
Establishing the API Quality Specification and the Last Process Step(s)
These are often very difficult tasks because chemical development is usually drawn
into its API supply mission at a very early juncture. At the start of a program there
are many uncertainties to resolve before identifying the most desirable API structure
and scoping out the market opportunities. Is the desirable API one of several chiral
options? Is the desired activity associated with the API as synthesized, or one of its
metabolites? The corollary of this is, Will the API structure need to be modified to
prevent an unwanted metabolic conversion—often by substitution of the metabolic
site (e.g., an H atom may be replaced by F)? Will the API need to be delivered in an
oral, topical, parenteral, or inhalation form, or more than one of these? Once selected,
will the desired API be a salt or a pro-drug? Even then there will be questions as
to which salt or what structure will be selected for the pro-drug moiety. Inevitably,
adding to the uncertainty, the question of establishing the polymorphic form will also
It will be appreciated that some of the changes in direction that result from
addressing these questions are often momentous enough that the toxicology program
is extended or restarted, thereby giving chemical development more time to carry
out experiments to help determine the best synthesis option to pursue. Frequently,
however, chemical development effort has to be diverted into the synthesis of large
quantities of one or more key intermediates to enable research to accelerate their
programs to identify the desired API.
The chemical process development work to define the final process step and aid in
setting the API specification is undertaken, as far as possible, outside the API supply
program and is the initial component of the exploratory effort needed to determine
the eventual industrial process.
API Quality Specification
Setting the API specification is one of the prime tasks undertaken by the central
independent analytical research and quality control unit using the data gathered over
the course of early research studies and in preparing the toxicology batch. The process
of setting the specification occurs over a period of time, evolving to accommodate the
findings made as knowledge is gained, uncertainties are resolved, and the synthesis
of the API develops.
Many factors are tracked in order to create the API specification, which is part of
the Investigational New Drug (IND) and New Drug Application (NDA) filed with the
FDA, or other regulatory agencies. The major factors are:
 The API structure itself [including identification of the active enantiomer,
metabolite, salt, solvate (hydrate) or pro-drug as needed]
 The crystal form (in particular the polymorph) and particle size
 The API assay, the assay of impurities, and product stability
The API Structure. In searching for the most active API structure to develop, it
is routine today to separate and test the enantiomers of a racemic molecule since
desired biological activity generally resides mostly in one enantiomer. This point was
appreciated long ago in the marketing of the oral ?-lactam antibiotics ampicillin,
cephalexin, amoxicillin, and cefadroxil, all carrying either the (R) phenylglycyl or
(R) p-hydroxyphenylglycyl side chain. The potential for enhanced biological activity
with single enantiomers has been realized in other therapeutic areas, though follow-up
has not been universal (e.g., ?-andrenergic blockers; see footnote 1).
More recently, partially as a result of increased research sophistication and observations
made in ADME (Absorption, Distribution, Metabolism, and Excretion)
studies, increased attention is being given to evaluating metabolites of APIs. An old
example (not being pursued because of the lack of patent protection) is the metabolite
NC C(CH3)2
Flutamide (Eulexin) Not marketed
SCHEME 1. Metabolism of Flutamide.
Loratadine (Claritin) "Desloratadine"(Clarinex)
SCHEME 2. Metabolism of Loratadine.
of the antiandrogen, Flutamide (Eulexin). The metabolite was later shown to be the
true API (see Scheme 1) .
A recent switch to a metabolite of a compound already in the marketplace is the
move from the nonsedating antihistamine, loratadine (Claritin), to “desloratadine”
(Clarinex) (Scheme 2).
Desloratadine is the most abundant of the several compounds produced when
loratadine is metabolized.
There are other considerations in searching for the most active API structure,
creating the need for close collaborationwith Pharmaceutical Development scientists.
Many APIs are marketed in a salt form. Salt formation can confer a variety
of physical, chemical, and biological properties on the API without changing its
basic chemical structure. A few of the important properties are water solubilization,
modified dissolution rates, improved stability, and beneficial pharmacological effects.
Preferred cations in salt form with API acids are sodium, distantly followed
by potassium and calcium. Organic cations [e.g., diethanolamine and Nmethylglucamine
(meglumine)] are used to a lesser extent. Preferred salts of basic
APIs are the hydrochloride distantly followed by the sulfate, bromide, and phosphate.
A large number of organic acids are also used (again to a lesser extent), notably tartaric,
citric, maleic, methanesulfonic, and acetic acids. The reader is referred to
a review article by Monkhouse and co-workers6 for a comprehensive, if old, list of
acids and bases employed in the pharmaceutical industry. This article also reviews the
effects of salt formation on bioavailability and on physiochemical, pharmacological,
and toxicological properties.
6Berge, S. M., Bighley, L. D., and Monkhouse, D. C. J. Pharm. Sci., 1977, 66, 1.
CO2 . CH
Sodium Cefuroxime Cefuroxime axetil
Ampicillin Pivampicillin
SCHEME 3. Preparation of penicillin and cephalosporin pro-drugs.
Pro-drugs are precursors to the API itself, being metabolized to the API in the
body. Essentially, both loratadine and flutamide above are pro-drugs. Pro-drugs are
often created to improve the oral absorption of the API, thereby creating patentable
advantage. Such initiatives have extended the original patent holders rights or enabled
competitors to gain a market niche. Two examples are sodium cefuroxime, which
became the pro-drug cefuroxime axetil,7 and ampicillin, which became the pro-drug
pivampicillin8 (Scheme 3).
Various other acyloxyalkyl esters of penicillins and cephalosporins were patented,
which gave the inventors positions in the penicillin/cephalosporin market. In short,
the pro-drug concept affords many opportunities to impart desirable properties to
Although chemical development organizations are not directly involved in identifying
the API structure to be developed, they are involved in providing research with
the building blocks, in the form of large quantities of advanced key intermediates,
to help speed their search. Chemical Development, especially during its emphasis
on defining the last synthesis step and the purification process, can also contribute,
peripherally, if its chemists or engineers identify stable salts, or solvates, or polymorphs
that have desirable properties (especially if these properties create a patentable
The reader will appreciate that the greater the complexity involved in identifying
tomorrow’s APIs, the more difficult will be the challenge of creating the ultimate
manufacturing process in a timely manner.
7(a) Cefuroxime: Cook, M. C., Gregory, G. I., and Bradshaw, J. U.S. Patent 3,974,153,1976 (to Glaxo);
(b) Cefuroxime axetil: Gregson, M., and Sykes, R.B. U.S. Patent 4,267,320,1981 (to Glaxo).
8(a) Ampicillin: Doyle, F. P., Nayler, J. H. C., and Smith, H. U.S. Patent 2,985,648,1961 (to Beecham).
(b) Pivampicillin: Frederiksen, E. K., and Godtfredsen, W. O. U.S. Patent 3,660,575, 1972 (to Lovens
Kemiske Fabrik).
The Crystal Form and Particle Size. Many APIs exist in more than one crystal
or polymorphic form. Since different crystal forms can possess quite different
properties—for example, melting point, solubility, and rate of dissolution—it is essential
at the start of the API development program, to establish which crystal form
(or reproducible mixture of crystal forms) of the API will be developed. As an illustration,
riboflavin (vitamin B2) can exist in three different crystal forms varying in
solubility in water at 25?C from 60 mg/liter to 1200 mg/liter.9 Generally speaking,
the high-solubility form of an API is the metastable form, usually with the lowest
melting point and also the fastest dissolution rate. Such a situation raises the concern
that a fast dissolution rate will lead to a faster absorption rate such that the therapeutic
efficacy of an API may vary depending on the polymorph administered. A similar
consideration exists if the API is isolated in an amorphous form (as happens with
many aminoglycosides). Amorphous materials are always more soluble than their
crystalline counterparts simply because more energy is required for a molecule of
a crystalline API to leave the crystal lattice than is the case with the amorphous
form. The form of the API can have a major impact on therapeutic properties. The
antibiotic novobiocin provides a dramatic example.10 The crystalline form is very
slow to dissolve and produces no detectable blood levels after oral administration.
In contrast, administration of the amorphous form leads to measurable blood levels
and significant biological activity. Amorphous forms of an API can be produced by
freeze-drying or the spray-drying of aqueous solutions.
Utilizing an amorphous form of an API is not, however, universally desirable.
Amorphous compounds are often metastable. As a result, there is a real risk that
they will transform to crystalline materials in the final dosage form. Novobiocin
again provides a case in point. The amorphous form, in aqueous suspension, will
transform on standing into the inactive crystalline form.10 Similarly, the highly soluble
metastable crystals of riboflavin revert to less soluble forms if they are washed with
water above 10?C.9
The amorphous and crystalline forms of chloramphenicol stearate provide a further
example,11 underlining the importance of establishing the form of the API at the start
of the development program. Finding the precise conditions for routinely reproducing
the needed form of an API (also meeting other analytical criteria—for example,
purity and particle size range) frequently requires considerable work12 and thorough
documentation, especially through the critical scale-up process.
9Dale, J. K. U.S. Patent 2,603,633, 1952 (to Commercial Solvents Corporation).
10Mullins, J. D., and Macek, T. J. J. Am. Pharm. Assoc. (Sci. Ed.), 1960, 49, 245.
11Alimarante, L., DeCarneri, I., and Coppi, G. Farmaco (Pavia) Ed. Prat. 1960, 15, 471; Chem. Abstr.
1961, 905.
12There is always the risk that work done to find the needed form of an API will not succeed in a given
time frame and that only after some time in production will the thermodynamically most stable form
emerge. This happened to Abbott Laboratories in 1998. They were obliged to withdraw the capsule form
of their HIV drug Ritonavir, because of the appearance of a new crystal form that possessed different
dissolution and absorption characteristics (see Pharm. J. 1998, 261, 150). The crystalline form in capsules
was later replaced by a gel capsule formulation that could not crystallize. We in Glaxo also encountered a
disappearing polymorph in manufacturing an early intermediate for Cephalexin (Bywood, R., Gallagher,
G., Sharma, G. K., andWalker, D. J. Chem. Soc., Perkin I, 1975, 2030). In this case, our first preparations
The particle size of a crystalline API is often an important factor in achieving
desired physical properties, such as reasonable drying times (see Chapter 8) and
desired pharmaceutical properties, such as dissolution rates and blood levels. The
FDA, in addition to requiring data demonstrating that the particle size reduction
process is consistently under control, is also likely to require proof that no thermal
degradation has occurred outside acceptable limits (i.e., no new impurity exceeding
the 0.1% level is produced during milling or micronization).
The API Assay, the Assay of Impurities and Product Stability. The central independent
analytical research and quality control unit is responsible for the analytical
release of both the API and the formulated drug product for the drug development
programs. The central independent QC unit provides all the analytical data needed
to build the analytical specification for the IND. It is recognized by those involved
that the IND is a relatively raw document compared with the later NDA, which is
built on data from a more developed process situation, using more refined analytical
The major concerns of the analyst in finding and developing API analytical procedures
are to provide the methodology for quantifying API purity, to work with
process chemists in identifying, preparing, and quantifying impurities, to work with
pharmacists and chemists in establishing the polymorph requirement, to provide
methodology for determining solvent and water content, and to ensure appropriate
limits are set for heavy metal, particulate, and residue-on-ignition (ROI) content. A
starting point for an API assay may look something like the following:
Assay Usually >97% pure (dry basis). A range is often given—for
example, 97.0–103.0%.
Chiral purity >95% e.e.
Polymorph Stable and reproducible
Impurities Total, ideally, ?2%? with no single impurity >0.5%
Solvents Levels depend on solvent (see later)
Heavy metals Generally <20 ppm
ROI Usually <0.5%; later this parameter may be <0.1%
LOD Mirrors solvent and water content (excepting specified solvate)
? Initially, the assumption (for HPLC assay) is that all substances have the same UV
extinction coefficient as the API.
Microbial contamination counts are often determined. Sterility and pyrogen tests
are needed for parenteral APIs.
The research analyst also undertakes a major program to determine the stability
of the API, including determining degradation pathways occurring under various
of the crystalline diphenylmethyl ester of penicillin G sulfoxide possessed an m.p. of 127?C to 128?C.
Later a new form emerged, m.p. 146?C. The new crystalline form posed no manufacturing problems. We
never saw the low-melting form again! Further examples of disappearing polymorphs have been cited
(Dunitz, J. D., and Bernstein, J. Acc. Chem. Res. 1995, 28, 193).
storage conditions and when the API is blended with the excipients to be used in the
dosage form.
A few comments on impurities and on stability are worth making.
The major impurities produced in the process to be scaled up are usually identified
at the research stage or in the early phase of developing the process for scale-up. To aid
the analyst, impurities are frequently recovered from mother liquors obtained from
the final crystallization step—for example, by preparative HPLC. The major ones are
synthesized and purified to provide the analytical “standards” needed to quantify the
amounts produced in the API synthesis. Over the course of time, the obvious impurity
collection is supplemented by those substances that might be produced in the process,
including other enantiomers. These “theoretical” impurities help to provide answers
to almost every query on the impurity profile of the API.
Levels of impurities found in the toxicology batch are usually accepted as the
allowable upper limits for the IND/NDA. Levels can be as much as a percent or two,
depending on the product and vagaries of the individual synthesis. High levels of impurities
can be quite acceptable provided that the toxicology work to prove the safety
of the API was carried out with API containing the same high level of impurities.
The startingmaterials in a synthesis are obvious potential impurities. More difficult
to deal with are the impurities deriving from the impurities in the starting materials;
situations developing from this often provide a good reason to set high quality
standards for all the key starting raw materials.
The stability of the API (and also of chemical intermediates) is often a key
factor in determining process requirements and API or chemical intermediate storage
conditions. This is especially true for relatively unstable compounds such as the
penicillin antibiotics. Stability testing can be important in evaluating variations in a
process, especially in evaluating minor changes in a manufacturing process, or the
impact of adverse shipping or storage conditions, or testing the compatibility of an API
with various packaging materials. Differential Scanning Calorimetry (see Chapter 4)
is often used to assess quality and gain information on the tendency of an API to
degrade. It is also helpful in assessing the effect of impurities on stability. Frequently,
simple heating of an API in an oven in an accelerated test at a given temperature
for an appropriate time can give useful stability information, such as on potency
loss or color generation, and can provide information on the impact of air versus
nitrogen blanketing on the rate of degradation. However, the reader should be aware
that accelerated tests (high temperatures for a short time) can exaggerate the actual
results obtained by storing at room temperature for times up to the projected expiry
date—usually a few years. Nevertheless, impurities produced in such as accelerated
stability tests are often isolated and identified for use in assessing the effects of aging
on the quality of APIs held to their expiry dates.
The most important objective of the above work on stability is to provide the data
to qualify the sought-after expiration date of the drug product, and with it the API.
The independent research analysts and the chemical development QC analysts
are vital players in the process development program. A strong interactive dialogue
between them and the process chemist/engineer invariably pays enormous dividends.
The research analyst generally has considerable sophistication in his/her armory of
Convergent Synthesis
Linear Synthesis
Many steps
Diosgenin Betamethasone
SCHEME 4. Two simple last process steps.
instruments (e.g., multinuclear and 2D-NMR, X-ray diffraction, FAB-MS, ICP-MS,
GC-MS, and HPLC-MS) to aid in learning about chemical purity, the chemical
transformations going on, the structure of impurities being produced, the polymorph
profile, and the stability of materials at any stage of the process.
Establishing the Last Process Step
The work needed to define and establish the last process step is generally difficult
inasmuch as it starts with the decisions to be made on what should be the final API
structure. Initially, therefore, it embraces all the uncertainties associated with defining
the API structure and establishing the crystal form and particle size.
The construction of the API molecule depends on the synthesis selected for
assembling the molecule. Many APIs are, or can be, simple to put together—for
example, in convergent syntheses such as when acylation of an amine is a logical
last step, as in the synthesis of most ?-lactam antibiotics and peptides, or in linear
syntheses such as the manipulation of a steroid molecule produced from an earlier
long sequence (Scheme 4). In these cases, simple synthesis strategies allow the major
quality concerns to be focused on controlling the quality of the intermediates IV, V,
and VIII, as well as the quality of reagents and solvents used to effect the conversions
to VI and IX, including the subsequent recrystallization solvent(s), if needed.
The process development chemist faces amarkedlymore complex problem in quality
control if the convergent synthesis is more open-ended as in Scheme 5. In addition
to ensuring that process conditions do not cause racemization of the chiral centers, the
number of key intermediates (X to XIII) being produced in such a convergent synthesis
usually slows the selection of the best intermediate structure to use and may lead to
variations in the impurity profile both of these intermediates and the final API (XV).
As an illustration, a change from the p-chlorobenzenesulfonyl leaving group in X
could conceivably lead to XI being alkylated by the X fragment at sites other than the
phenolic oxygen. Furthermore, the commercial availability of low-cost intermediates
Sch 56592
SCHEME 5. More complex assembly in last process steps.
(building blocks) from other chemical intermediate manufacturers already producing
building blocks for other companies may change the synthesis strategy. The availability
of such building blocks may not be known at the start of a synthesis program but,
when discovered, can lead to an outsourcing program to qualify the new supplier and
often a new specification of impurities. The further back in the synthesis one goes to
make a process change generally lessens the problems of qualifying new impurities
(usually because the number of subsequent steps provides more opportunity to purge
impurities—for example, in crystallization steps). Thus, changes in the impurity pro-
file of XI in Scheme 5 might be expected to be easier to accommodate than changes
in the impurity profile of XIII. For instance, if an impurity in XIII became combined
with the large molecular fragment XII, the insolubility conferred by this fragment may
make it difficult to separate two quite similarmolecules carrying onlyminor structural
Despite such considerations, economics generally favor adoption of a convergent
synthesis over a linear synthesis. This is because outsourcing of building blocks
minimizes capital investment in chemical production plants and, given detailed attention
to the quality of outsourced building blocks, can minimize the steps subjected
to regulatory scrutiny. The other side of the outsourcing strategy is, however,
that companies investing in the plants to undertake a long synthesis themselves
develop know-how and a position of controlling their destiny after patent expiration.
Thus, they have a better chance of holding onto their market position long
after their patents have expired. A good illustration of this is the steroid field—
many of the same major manufacturers of steroids who were producers in the 1950s
are still in production today. However, major steroid manufacturers are also leveraging
their manufacturing capabilities for additional profit by offering late-stage
building blocks, such as compound VIII, in the commercial marketplace. This is
leading to consolidation in steroid manufacturing as less efficient manufacturers
of steroids contract the most efficient to manufacture their late intermediates and
even their APIs. The downside of such an outsourcing strategy is that the sourcing
company does not always know whether a takeover of the efficient manufacturer
by another company may change the operating plan to the exclusion of third-party
Nevertheless, third party manufacture of fine chemical intermediates has grown
substantially in the last decade, primarily to reduce pharmaceutical companies’ financial
exposure and speed API development. Pharmaceutical companies are farming out
their late-stage intermediates technology to third parties as the number of their API
candidates increases, thereby avoiding investment in capital and human resources.
Third-party manufacturers are building on their strengths—in particular, chemistry
and technology (e.g., phosgene, explosives, handling odorous sulfur compounds,
etc.)—and on their high-quality specialty operations to enable pharmaceutical companies
to move more quickly in the development of APIs.
Returning to the main theme, one of the most important components in defining
the last process step is to determine the purification scheme to be used for meeting
quality and crystal form/particle size criteria for the API.
On the assumption that the synthesis steps for constructing the precise structural
features of the API have been carried out from well-defined, good-quality building
blocks, the main task of the process development chemist is to find a process
for purifying the API to give the required physical form meeting the purity requirements.
This is usually a crystallization process utilizing an acceptable solvent,
or mixture of solvents, having low toxicity. Preferred solvents are listed in Table
1. Water was not included in the ICH list of preferred solvents, but where it can
be used, most process chemists and engineers would regard it as the solvent of
Less preferred solvents (Class 2) are listed in Table 2. As stated in the ICH document,
these solvents need to be limited because they are suspected to be nongenotoxic
animal carcinogens or possible causative agents of some irreversible toxicity such
as neurotoxicity or teratogenicity. They may also cause other significant, frequently
TABLE 1. Preferred Solvents for Use in the Purification of
APIs (Class 3)13
Acetic acid Formic acid
Acetone n-Heptane
Amyl alcohol Isoamyl alcohol
Anisole Isobutyl acetate
1-Butanol Isobutyl alcohol
2-Butanol Isopropyl acetate
Butyl acetate Isopropyl alcohol
t-Butylmethyl ether Methyl acetate
Cumene Methyl ethyl ketone
Dimethyl sulfoxide Methyl isobutyl ketone
Diethyl ether n-Pentane
Ethanol 1-Propanol
Ethyl acetate Propyl acetate
Ethyl formate Tetrahydrofuran
TABLE 2. Less Preferred Solvents for Use in the
Purification of APIs (Class 2)13
Acetonitrile Methanol
Chlorobenzene 2-Methoxyethanol
Chloroform Methyl n-butyl ketone
Cyclohexane Methylcyclohexane
1,2-Dichloroethylene N-Methylpyrrolidone
Dimethylacetamide Methylene dichloride
Dimethyl formamide Nitromethane
Dioxane Pyridine
2-Ethoxyethanol (Cellosolve) Sulfolane
Ethylene glycol Tetralin
Ethylene glycol dimethyl ether Toluene
Formamide 1,1,2-Trichloroethylene
Hexane Xylene
reversible, toxicity. The ICH guideline13 provides PDE data (mg/day) and concentration
limits (ppm) for Class 2 solvents in drug products.
Although Class 3 solvents have no known human health hazards at levels normally
found in drug products, generally corresponding with a level of <0.5% in the API,
the solvents in Class 2 need to be controlled at levels which are less, sometimes
substantially less, than those allowed for Class 3 solvents. The level of Class 2
13List taken from International Conference on Harmonization (ICH), harmonized tripartite (Europe,
Japan, United States) guideline entitled Impurities Guideline for Residual Solvents. The above solvents
are categorized as Class 3 solvents, with low toxic potential to man. Class 3 solvents have permitted daily
exposures (PDEs) of 50 mg or more per day.
TABLE 3. Solvents to be Avoided in the
Purification of API (Class I)13
Benzene 1,1-Dichloroethylene
Carbon tetrachloride 1,1,1-Trichloroethane
Source: ICH Impurities Guideline for Residual Solvents.
solvents allowed in the API is generally determined by reference to the amount
of the solvent calculated to be present in the drug product (i.e., after dilution with
the excipients) and PDE data. Calculation methods are given in the ICH Impurities
Guideline for Residual Solvents.13
Process development chemists make every effort to avoid using Class 1 solvents
(Table 3) for the crystallization of APIs. The solvents in Table 3 pose carcinogenic,
environmental, or other toxicity risks.
Solvents other than the above Class 3 or Class 2 solvents may be used, but doing
so requires that the user provide information to assure regulatory agencies that there
will be no untoward health consequences from the level of their presence in an API.
Clearly, PDE information showing that the drug product containing the expected level
of residual solvent was safe would be best.
In directing substantial effort into establishing the last process step, the development
chemist must pay close attention to creating a written record that will not
only provide others with precise detail on how to reproduce the process but what has
been done to build quality into the process. The written record embraces the company
cGMP policy and procedure manuals, chemists’ notebooks, analytical control
and product release data files, pilot plant SOPs, batch sheets and cleaning records,
cGMP training manuals, and staff curriculum vitae. This documentation, initially in
a relatively raw state, leads to the filing of the Investigational New Drug Application
(IND) and eventually, in a refined state, to the New Drug Application (NDA). The
company also prepares a dossier describing in detail the process for the manufacture
of an API—referred to in the United States as the Drug Master File (DMF). Technology
transfer, based on most of the above documentation, usually begins before
the NDA is filed. The Development Report, summarizing the company’s journey to
the NDA process, is being built at the same time. Validation of the process and the
manufacturing system follows. The validation record provides the written proof that
you are doing what you say you are doing.
Establishing the last process step marks only the beginning of the effort to create a
quality culture. The effort reaches both forward into ensuring that the pharmaceutical
development regulatory needs are met and backwards into ensuring that the process
steps leading to themolecular assembly of theAPI structurewill guarantee the quality
of the final API.
The iterative nature of the work to define the last process step also applies to the
R&D work needed to define the synthesis methodology for the key steps leading to
the last process step.
The R&D Work Needed to Define the Synthesis Methodology
Themore uncertainty there is in identifying the synthesis to be developed for the early
steps in the preparation of an API, the more difficult it is to define the last API step
in a needed time frame. Experience in process chemistry and open-minded dialogue
among proponents of the various synthesis strategies are essential components in
the early selection of the best synthesis strategy. In my experience, exploration of
more than one strategy is the norm with intense effort focused on trying to overcome
the perceived weaknesses in each strategy—in short order! There is also the odd
“bootlegger” who needs to get an idea out of his/her system—and who, maybe
infrequently, wins the day.
Today, the real problem is time and setting a decision-making timetable. This has
led to a sharper focus such that the research recipe for producing the API as quickly
as possible for toxicology/pharmacology work is often driven into a larger-scale
operation with no more than evolutionary changes to make the chemistry safer. The
urgency generally causes effort that might otherwise be devoted to finding a better
synthesis strategy to be diverted into improving the research recipe. One answer
to this that has proved helpful is to establish a dialogue with research chemists to
enable them to address development issues and specific concerns as soon as there is
an indication that a particular API might become a development candidate. Another
is to generate secrecy agreements with third parties, preferably those who might
be expert in the technologies and the structural chemistry related to the API or its
intermediates. However, identifying the right third party is not easy. One has the most
control, including over costs, if you own (i.e., have patented) the intermediates being
sought. This is usually a later consideration being helpful at the commercialization
stage, enabling one to “shop around” to find the best supplier.
As far as research collaboration is concerned, the Schering–Plough manufacturing
process for its cholesterol absorption inhibitor, Ezetimibe (Zetia), exemplifies the
collaboration case. The close interaction between Research and Development, aided
by the delay caused by the realization that the first API structure (XVI) had to be
modified to overcome metabolism issues, provided the intellectual resource and the
time for a fuller understanding of the chemistry needed to create the chiral ?-lactam
Sch 48461
This led our Dr. T. K. Thiruvengadam to invent an exquisitely elegant synthesis of
Ezetimibe—Scheme 13 in Chapter 9.
H2/Lindlar catalyst Peracid
OH NHCOCHCl2 Reduction
(Ph3P)3PdCl2/Cu I
(Yarovenko reagent)
SCHEME 6. Florfenicol synthesis explored by a third party.
In regard to third-party collaboration in the manufacture of Schering–Plough’s
antibiotic, Florfenicol (XVII), the initial thrust by our Animal Health Division,
developers of the molecule, was to involve a third party with an excellent track record
in another project with them. The third party partially explored the synthesis outlined
in Scheme 6.
Several problems with Scheme 6 were identified. The Yarovenko reagent led to
several percent of an impurity containing ?C?C·CH2Cl, which proved difficult to
remove, and the azide ring opening of the epoxide led to both possible hydroxyl
azides. These problems and the need for a resolution step led to abandonment of
Scheme 6.
A concurrent initiative with Zambon S.p.A. in Italy was much more successful
since it was based on working with Zambon to utilize their commercially available
thiamphenicol intermediate, XVIII, a molecule already possessing the structural
requirements of Florfenicol, including chirality.
Zambon succeeded in converting the terminal CH2OH, in an N-protected derivative
of XVIII, intoCH2 F confirming the viability of the option [see Chapter 7 for both the
Zambon synthesis of Florfenicol (Scheme 9) and further ramifications of the project].
Every chemical process development organization will have its own experiences
in the value of collaboration both with its own research organization and third parties.
Because of the structural novelty of newAPIs, one rarely finds commercial sources
of needed late intermediates such as in the Florfenicol case above. The problem then
becomes either to produce the target intermediates yourself or to find and qualify one
or more third parties, preferably experienced in the chemical technologies needed in
the synthesis, to do so.
Qualifying a third party requires that detailed synthesis, analytical, and GMP
information are provided to the third party. This is usually passed to the third party in
a technology transfer package (see later). Eventually the third party is registered in
the NDA as a supplier who has committed to meet an agreed analytical specification.
For materials well back in the synthesis and for raw materials, including solvents, one
may need no more than basic analytical release information—appearance, identity,
purity (sometimes it may be necessary to set limits on specific impurities).
For late-stage intermediates, a very stringent set of additional criteria will usually
be required. These will depend on many factors, but some of the major ones are:
 Impurities. Do some impurities persist through to the API or become converted
to other undesirable impurities which are difficult to remove? If so, what levels
should be set in the specification for the intermediate? Do transport or storage
conditions affect impurity levels?
 Water/Solvents. Are the levels of water or solvent contaminants important,
requiring a tight specification? If so, what levels should be set and what process
(e.g., drying conditions) is needed to meet the specification?
 Optical Purity. Is optical purity crucial? Is optical purity changed by transport
or storage conditions?
As the raw material/intermediate sourcing program develops, it is usual to qualify
more than one supplier for “insurance” purposes, and to eventually encourage
competitive pricing.
Establishing and reproducing quality is one of the most important objectives
in developing a chemical process. Generally, the quality focus starts with the raw
materials and, as indicated above, becomes a crucial factor in sourcing late stage
intermediates. Today, in recognition of the importance of quality, most chemical
raw material manufacturers in Europe and the United States have subscribed to a
quality assurance standard initiated in the United Kingdom but developed, from
1987, by the International Organization for Standardization (ISO) in Geneva. Five
quality assurance standards, applying to manufacturing quality in any business, are
embraced under the logo ISO 9000.
ISO 9000. This standard, titled “Quality Management and Quality Assurance
Standards: Guidelines for Selection and Use,” is advisory describing the use of
the standards in establishing supplier contracts as well as providing guidance
on the use of the other four standards.
ISO 9001. This is a comprehensive standard that includes the requirements of
both ISO 9002 and ISO 9003. It describes what is needed for quality assurance
in services and products all the way from design, development, and installation
to production, servicing, and supply.
ISO 9002. This limited standard applies to quality assurance in installation and
ISO 9003. This standard only provides a guideline for the requirements of quality
assurance in final inspections and testing.
ISO 9004. This standard, like ISO 9000, is advisory providing guidelines for the
development and application of an internal quality management system.
ISO qualification is essentially a generic qualification specifying all the elements
of a system that need to be in place in order to effectively control quality. It is flexible
in that it is up to the user to determine how the elements are implemented. It does
not tell users how to do their job. ISO qualification generally results from passing an
audit which shows that you have the system in place to meet the requirements of the
ISO standard. In regard to chemical manufacture, ISO qualification has been much
sought after and recognized by those sourcing chemicals as a desirable requirement
in a supplier of raw materials and intermediates.
It needs to be stressed that although raw materials, solvents, and intermediates
made by ISO qualified producers are usually well regarded as inputs for the manufacture
of APIs, the ISO system is not an alternative to cGMPs. One of my European
colleagues has likened ISO qualification to passing the tests needed to obtain a driving
license. Possessing a driving license does not say anything about how well you
control your driving.
Returning to the theme of searching for the best synthesis methodology, the main
driving force is usually provided by the imagination of chemists and collaboration
with those expert in sourcing chemicals, with analysts, with chemical engineers, and
frequently with their counterparts in the manufacturing division. Having identified
a synthesis, it is relatively straightforward to make changes in the chemistry or
introduce new chemistry before the IND is filed, although every change needs to be
justified and proven using analytical findings to show that both process intermediates
and API are within the quality parameters set in the toxicology batch and within
those being developed for the IND. The pilot plant used in scale-up is generally
qualified with records kept on all the equipment needed for all processes. This record
shows that instruments are regularly checked and calibrated and that the process
equipment can deliver what it is required to deliver—in terms of temperature control,
rates of cooling and heating, stirring requirements, filtration, washing and drying
characteristics, and so on. Company audits as well as calibration and maintenance
records on equipment, especially equipment used for carrying out API production,
are important in establishing a quality system. Quality risks, such as those evident
when new suppliers of raw materials, intermediates, and solvents are being evaluated
or when multiple recrystallizations are needed to meet quality criteria, have to be
identified, the causes of potential problems understood, and steps taken to avoid
compromising API quality. Activities to identify and resolve quality risks go on
continuously. It is generally recognized that all quality risks and causes of potential
quality problems cannot be resolved by the time the IND is filed. Such work goes
on into the pre-NDA phase. However, during the period between filing the IND
and submitting the NDA, it is much more time-consuming to make changes. Small
changes—for example, a switch to using sulfuric acid instead of hydrochloric acid
(corrosion issue)—may be introduced on a well-documented basis, proving that the
change has had no adverse effect on quality. Large changes can be made after the
IND submission if rigorous analytical and process operating scrutiny is undertaken
to ensure that the quality of key intermediates and particularly the API is maintained.
However, the opposite of this is not uncommon wherein processes are “frozen” post
the INDfiling. After theNDAfiling, it is considered unwise tomake changes (see case
study on “Dilevalol Hydrochloride: Development of a Commercial Process,”—FDA
Review and Compliance Activities, page 288) since they may delay FDA review of
the NDA.
In cases where changes are being made between the IND and NDA, the API
produced is generally put into a Restricted-Use category. Such materials may be used
for work that does not find its way into any Regulatory submission. It may be used,
for instance, in preliminary analytical, stability, and regulatory qualification work
if it is agreed that there is overwhelming justification for the changed process, say
on the grounds of significantly lower costs or considerable superiority from an API
manufacturing standpoint. Restricted-Use API may also be valuable in evaluating
API formulation ideas or for evaluating process conditions or alternative formulation
machinery for producing the dosage form. Once this latter kind of work has been
completed, the product from the work is usually destroyed.
Creation of the Chemistry, Manufacturing, and Controls (CMC)
Document for the FDA
The CMC section of the NDA is a small component of the company’s overall submission
of data to the FDA seeking approval to market a new drug. The NDA submission
is made to the Review Branch of the FDA in Washington. Final submission of the
CMC section is usually preceded by a presentation of the proposed content to the
Review Branch. Feedback from the ensuing discussion is incorporated into the final
submission. Once a drug is approved, responsibility for FDA oversight in the implementation
of manufacturing and marketing shifts to the Compliance Branch of the
FDA, specifically to field offices close to the manufacturing sites.
As part of the preparation for creating the CMC document, the chemical development
organization (or the manufacturing organization if technology transfer has
already occurred) usually produces a minimum of three large-scale batches of the
subject API using the procedure to be filed in the CMC documents. The results of
this three-batch exercise demonstrate that the process operation and API quality are
consistent with the criteria established for the CMC document.
The CMC section of the NDA is written by the company Regulatory Affairs
organization in close collaborationwith theChemical, Pharmaceutical, and Analytical
Development organizations. The CMC section contains the following information:
 A formula outline of the synthesis chemistry.
 Acatalogue of solvents, rawmaterials, and sourced intermediates, vendor identi-
fication, and analytical release specifications for all thesematerials. (It is usually
desirable to identify two or more vendors.)
 A journal-style description of the process used for API manufacture.
 A review of the critical parameters requiring specific control to secure quality.
 A summary of the analytical methodologies used to establish the quality of all
chemicals used in the API manufacturing operation, as well as to ensure quality
control throughout the process.
 A succinct description of the work done to establish the structure of the API and
the crystal form.
 The API specification.
 Impurity identification and a specification for the level of impurities.
 API stability information.
 Corresponding information to the above on the formulation of the dosage form
(drug product).
The pre-NDA meeting with the Review Branch of the FDA usually takes the form
of a presentation much along the lines of a scientist presenting a paper at a scientific
meeting. A pre-NDA meeting is invaluable in identifying and dealing with FDA
concerns before the final NDA is submitted. The pre-NDA meeting also enables the
company to gain approval, or other guidance, on its synthesis strategy in regard to
the starting point of its API synthesis. Normally the FDA requires that the synthesis
starting point should be a commercially available chemical. However, pharmaceutical
companies, recognizing the uncertain nature of drug development and their financial
exposure if they had to invest in all the manufacturing plant needed to produce the
API, increasingly work with third parties enabling them to become suppliers of latestage
intermediates. Novel late-stage intermediates produced by third parties can
sometimes qualify as being commercially available even if they are made exclusively
for the contracting pharmaceutical company and are not offered for sale to others.14
These circumstances require that the third party essentially create a DrugMaster File,
and thereby be subject to FDA inspection, to qualify themselves as a supplier of a
specific late-stage intermediate in the NDA.
14There have been a few cases wherein the FDA has accepted that a late-stage new intermediate made
exclusively under contract for one party can be classed as commercially available, not requiring FDA audit.
These cases are notable in being more science-based and in exhibiting an extraordinarily high level of
quality control throughout production. Tight quality control over the starting chemicals (including impurity
levels), tight in-process controls in producing the late-stage intermediate, rigorous impurity mapping, and
an exacting quality specification on the compound have been major factors enabling the FDA to approve
the third-party supplier as a source of the late-stage new intermediate, without requiring DMF status.
These few cases provide an indication that the FDA is beginning to apply the principles outlined in its
document “Pharmaceutical cGMPs for the 21st Century: A Risk-Based Approach.” However, it is clear
that “risks” allowed by one Division of the FDA Review Branch may not be acceptable to another.
Analogous presentations to the Review Branch are made for the drug product.
Both the chemical and pharmaceutical development presentations are supported by
analytical development presentations providing the data to satisfy the FDA that all
components of the API synthesis and dosage form preparation are well-controlled.
Before and following the NDA filing, many other activities are going on which
are pertinent to regulatory matters and preparatory to the FDA Compliance Branch’s
Pre-Approval Inspection (PAI). The most important activities are the writing of the
Development Report, Technology Transfer, and Validation of the operation.
The Pre-Approval Inspection (PAI)—The Development Report,
Technology Transfer, and Validation
The filing of the NDA with the Review Branch of the FDA leads to the Compliance
Branch being notified and given the CMC section. In turn, the Compliance Branch
contacts the company, indicating its readiness to undertake a Pre-Approval Inspection
of the facilities to be used for manufacturing the PAI. The company decides when
it is ready for the PAI. Readiness is determined, inter alia, by the availability of the
documentation needed to demonstrate that the systems for manufacturing the API are
in place. This often starts with the Development Report, summarizing the genesis of
the synthesis scheme and how this was developed to the NDA process. The process of
Technology Transfer provides further insight. This embodies the mechanics of how
themanufacturing technologywasmoved and controlled in scale-up to the production
plant and demonstrates the discipline employed by the manufacturer in dealing with
every component of the process, including plant equipment, operator training, and
analytical oversight. Validation, which is not always completed by the time of the
PAI, is an essential final component of the NDA approval process. The bottom line
is, however, that APIs cannot be marketed without completed validations.
The Development Report
The Development Report is prepared via a succession of interim reports, usually
over several years. When interim reports detail findings critical to the synthesis,
they appear in an addendum form in the final report. The interim reports provide a
journal-style write-up describing how the chemistry used for synthesizing the API
was carried out. The identification of impurities and the determination of their level
is a priority. Ways of maximizing the yield of the desired product, especially by
overcoming side reactions (impurity formation), are described. Recrystallization and
other purification methods (including chromatography) are detailed. The recovery
and recycling of reagents and solvents is introduced as quality information on the
acceptability of recycling becomes available. The Development Report also includes
raw material specifications, analytical test methods, in-process control methods, and
intermediate and API stability information.
The following gives an outline of the structure of a typical Development Report
with some notes on the kind of information that can appear:
 Introduction. Provides a brief outline of the discovery of the API and the chemical
process for its preparation.
 Process Flow Chart. Gives the sequence of reactions in chemical structure terms.
 Nomenclature. Provides an index linking the chemical structure of the intermediates
and API with the shorthand designations used by the company and the
Chemical Abstracts Name.
 Description of the Process. The typical large-scale batch operation (same size
as in the NDA) is described in journal-style detail for each step.
 Critical parameters are identified— that is, those which have a defined range.
Range excursions may affect API quality. The consequences of range excursion
are discussed.
 Process control conditions are summarized. These may affect safety, yield,
process consistency, or intermediate quality. Some of the common concerns
are solvent hazards, moisture content of solvents or reagents, reaction temperatures
and times, reagent addition rates, crystallization conditions, and drying
 In-process and final-test instruments are described along with analytical methods
and reference charts (e.g., HPLC chromatograms) for assaying intermediates
and final API. Typical analytical results profiling impurities, monitoring
for reaction completion, and determining data such as the water and solvent
content of products are provided.
 Raw material specifications—lists of vendors and specifications for each material
used in the step with particular attention to the impurity content of
intermediates that contribute to the structure of the final API.
 Stability of intermediates—provides data on the storage conditions (for solids,
this is frequently polythene bags in fiber drums) that are desired for each
intermediate. Such data are usually gathered to show that the intermediate is
stable for X months (or years) under suggested (desired) storage conditions,
alleviating concerns that out-of-specification materials might be used.
 Yield data—the range of yield that is expected from the process is identified.
Yields outside the range require investigations as to cause and what will be
done to overcome the issues created.
 Impurities—a list of impurities obtainable for each step is provided separately
in the description of the process used for each step. The origin of
each impurity, how each impurity is controlled, and each impurity’s fate are
described—typically most potential impurities are eliminated or reduced to
acceptable (defined) levels during intermediate purification steps.
 Process development—this section ends each step, providing a summary of
factors involved in the selection of reagents, solvents, their amounts, process
conditions, detail of critical parameters, efforts to streamline the process (e.g.,
increase concentrations and combine steps), and so on. When events occur
that give special concern, these are detailed. One unexpected example in one
project was the impact that light had on samples of an intermediate being
analyzed by HPLC. Light caused degradation of the molecule, leading to
erroneous results. In broad terms, the reaction was
Preventing exposure to light avoided the problem.
 Ancillary formation. This can include, inter alia:
 An outline of the NDA process
 Detail of in-process and final test methods for raw materials, intermediates,
and products
 Description of special analytical method(s) for chiral assays
 Preparation of chiral auxiliaries
 HAZOP reports
In summary, the Development Report provides a comprehensive overviewof the many
factors addressed in creating and implementing the process for the manufacture of
the API.
Technology Transfer
Technology transfer from a chemical development operation commences as soon as
the decision is made to move any part of a process, say for preparation of an early
intermediate, to a third party, or the company’s own manufacturing division. Where
third parties are involved, the transfer is usually undertaken after a confidentiality
or secrecy agreement has been signed to ensure that no unwanted disclosures can
occur—for example, to competitors. Early dialogue with potential manufacturers of
intermediates is often informal and may comprise no more than the transfer of a
laboratory procedure, analytical methodology, a sample of typical product, and the
analytical “standard” existing at the time. Ideally, the third party will reproduce the
chemistry exactly as provided, or agreed, and the analytical instruments, analytical
methodology, and results will be compared to establish that the third party can
reproduce the methods satisfactorily. Any changes in the procedure that may be
necessary for scale-up have to be approved before implementation. As always, every
step of the process is recorded to demonstrate that the operation is being properly
Transfer of the last process step for producing the API, generally to one’s own
manufacturing division, is a major endeavor. This often occurs between the IND and
NDA filings. One of the desired objectives in doing this is to enable the company
to carry out as much of the Phase III clinical and toxicology work as possible using
API produced by the commercial manufacturer. This timing, which helps to simplify
the dialogue with the FDA, is not always achievable if the putative commercial site
requires an investment in specialized equipment which takes a long time to deliver,
install, and commission. It can be seen from this that investment in the manufacturing
plant needs to be recognized and addressed early in the development of the final
process step. Situations such as this demonstrate why great importance needs to be
attached to developing the last process step as soon as possible.
In conducting technology transfer, most companies work to a “standard” operating
procedure (SOP). The basic content of such an SOP may include some of the
Purpose: • Establish the requirements for the orderly transfer of technology
from one site to another in order to ensure equivalence of
Responsibility: • Outline those with the accountability.
Nomenclature: • Define terms—for example, API, Analytical Comparison Report,
ELINCS, Experimental Protocol, and so on.
Procedural Steps: • Identify team members and their function on both sites.
• Create the technology package (process/analytical batch
information, safety requirements, training/operating protocols,
projected timelines, etc.).
• Establish plan to deal with deviations/variances.
• Coordinate efforts during preparation and analysis of
demonstration batches.
• Review the results.
• Write and issue a Technology Transfer Summary Report.
• Create a Validation Readiness Checklist to document completion
of the items that must be addressed prior to manufacture of the
validation batches.
In technology transfer, it is necessary to establish that the plant equipment, instrumentation,
and operating systems in the receiving plant will provide the conditions
the process needs at every step and that the plant maintenance program is such as to
assure continuity of these conditions. Plant cleaning is also a major consideration,
especially if the plant is being used to produce several products. Procedures for the
removal of solvents and process residues from previous batches are essential to the
concept of “building in quality.” This is particularly important in dryers, which are
also used for several products. However, even dedicated dryers need to be cleaned
using a validated procedure to prevent build-up of degradation residues. If the API
is for parenteral or inhaler use—that is, produced in a clean environment room (see
Chapter 8)—cleaning procedures become vitally important, not only for minimizing
impurity contamination but for eliminating or greatly limiting pyrogen or bacterial
contamination. In short, cleaning is a vital component of API manufacture.
Inevitably, as a process develops, improvements are made and changes are needed
to accommodate them. Chemistry and operating changes, even seemingly very small
ones, can adversely affect API quality, raising the need for all process changes to be
governed by a change control system. The system should be flexible. Process changes
in producing raw materials and early intermediates generally need less attention than
changes made in the last API step or in steps to produce key intermediates where
quality changes can impact on the quality of the final API. Whereas a laboratory “use
test” on a “new” raw material or early intermediate (i.e., using the “new” material
to make the API and showing that API quality is unaffected)15 might suffice for
chemicals well back in the synthesis, a change in a process step late in the synthesis
requires increasingly stringent scrutiny the closer to the API one gets. Change in
the late steps of an API synthesis, or in the synthesis of the API itself, require a
change control procedure monitoring every step of the improved process to ensure
reproducibility and consistencywith previous quality criteria andcGMPs—especially
to ensure that the quality of the API is either improved or not adversely affected. The
same diligence applies to the manufacturing plant operation.
Technology transfer is usually uneventful if it is carried out in a structured and disciplined
way, with dedicated supervision, meticulous attention to detail, and constant
This exacting discipline has come to the fore as API and drug product producers have
been obliged to provide the proof that their process conditions, their instruments, and
their plant equipment are verified, calibrated, and demonstrated to work as they say
they work to consistently deliver API meeting the NDA specification.
The successful operation of a production chemical process to give a desired API
depends on exactly reproducing the chosen chemistry, on the analytical instrumentation
used to control the chemistry and determine the quality of the API, and on the
equipment and services used to carry out the process steps as well as the cleaning
procedures needed before equipment reuse. To ensure that these main components of
the operating system are working optimally to deliver high-quality API, they all need
to be validated. It is essential to start validation with a protocol detailing the work to
be undertaken to prove that all components of the system set up for manufacture of
the API consistently provide quality product. Implementation of the protocol work
leads to the documentation establishing the proof.
The process chemistry chosen is defined in detail by determining the limits of
process conditions (i.e., the operating ranges for temperature, time, pH, etc.) which
consistently give APIwithin the specification filed in theNDA. Really critical process
parameters are flagged for particular vigilance—ideally the chemistry chosen will be
robust enough to minimize the need for very tight control. Validation of analytical
instrumentation through regular calibration using unimpeachable chemical standards
provides the cornerstone of quality assurance. Plant equipment and instrumentation
(i.e., measuring such as temperature ramps, pressure, pumping rates, pH, etc.) are
correspondingly demonstrated to provide the expected services and correct read-outs
15It is always a good idea to monitor the quality of the “new” material for a time to provide assurance that
its quality is being maintained.
in the ranges needed for controlling the chemical process. Protocols for cleaning validation
are a vital component of the validation package. Cleaning validation comprises
proving that the mechanics of cleaning, followed by the sampling of cleaned surfaces
and analysis, together create assurance that the plant will not cause contamination of
the API.
All validation activities associated with producing a minimum of three validation
batches of the API are drawn together in the form of a Validation Report. This report
essentially documents company compliance with the NDA process, making it a vital
document in the effort to gain FDA approval of the NDA.
Although very considerablework is undertaken in validating anAPI manufacturing
process, one should not overlook some of the benefits. The main ones are as follows:
 Afuller understanding of the process is gained during the exercise of determining
the process conditions to be used in running the validation batches.
 Fuller understanding enables the plant to reach optimal process efficiency faster.
 Plant failures are minimized.
 In-process assays provide data for the statistical analysis of trends for
yield/quality improvement or warnings of potential adverse events.
Validation is one of the most important exercises in completing the information
requirements for NDA approval. As a result, considerable importance is attached to
this discipline in pharmaceutical companies such that validation professionals have
a place in all domestic and international Regulatory compliance organizations across
the company.
The PAI carried out by theCompliance Branch of the FDAuses theNDAsubmitted
by the company as the basis for the inspection. The inspection comprises an in-depth
scrutiny of records and operations as well as a wide-ranging dialogue with the
company, usually over several days. The inspectors, in looking for out-of-compliance
situations, may focus on failed batches and their reprocessing or rework, or on the
retesting of expired materials, or on the SOP for analytical instrument calibration, or
on the conditions of storage for standard samples, or on process deviations and how
they were addressed, or on cleaning procedures and the volumes of solvent used, or on
warehousing and the label status of materials stored there, or . . . , the list could go on
and on. In short, an inspection may take the FDA into any aspect of the manufacture
of the API.
At the end of the inspection, a meeting is held at which the FDA summarizes
its findings and outlines the deficiencies they found. Deficiencies are referred to as
FDA 483’s. Their findings are summarized in a written follow-up. The company
then responds in writing, providing a program for overcoming the deficiencies. The
dialogue continues until the FDA is satisfied that the company’s manufacturing
operation is in compliance and is properly documented and validated.
In order to prepare for a PAI, the company usually sets up an “independent” audit
of the facilities well before the formal PAI. This audit, usually carried out by the
company’s own Quality Assurance/Regulatory Compliance groups, often assisted by
outside specialists, enables the company to address most of the items likely to be
spotted by the FDAinspectors. Such audits serve as a form of training of the personnel
in operations and enhance awareness of the Regulatory needs which should be being
addressed as the development of information for the NDA is going on.
Irregular internal PAI-type audits of operations, like safety inspections, enable the
company to maintain its vigilance and ensure that its operations are sustained at a
high level of compliance with agency regulations.
The creation of a process to meet FDA requirements is a multidisciplinary activity.
Although the Chemical Process Development organization generates the core process,
its shaping and implementation to meet the needs of all other parties involved
(particularly Regulatory, Manufacturing, Pharmaceutical Sciences, and Quality Assurance)
requires an extraordinary level of collaboration. The principal objective is
to produce, and to demonstrate that you have indeed produced, a high-quality API in
a well-controlled system of operations.
Selection of the process is an iterative activity adapting to the changes occurring
during the selection of the specific API and the definition of the last process step(s)
to produce it. Most attention is devoted to producing a quality API for research and
creating an efficient practical synthesis sequence for its preparation, in collaboration
with analytical groups, chemical engineers, pharmaceutical scientists, and manufacturing
operations.Outside vendors of rawmaterials and intermediates are “cultivated”
and approved.
The relatively free-wheeling synthesis selection phase through the IND gives way
to a much more controlled development phase, wherein the quality of the Toxicology
Batch (and especially impurity levels) dictates the quality of the API batches to be
produced, slowing process change. Analytical methodologies and specifications for
theAPI, intermediates, and rawmaterials becomemore refined. Impurity and stability
profiles are established. Process control mechanisms are developed and plant SOPs
incorporate the better controls. The NDA process slowly takes shape.
The documentation package covering the operation also begins to emerge as
the program moves toward NDA submission. The components of the Development
Report and information leading to the CMC section are pulled together. Technology
Transfer is formalized, and the time-consuming process of validation is begun.
Preparation for the PAI is completed, and eventually NDA approval is gained. These
activities, demonstrating that quality is being built in, also create the foundation for
the system of governance needed for maintaining quality in ongoing operations.
The focus of the FDA continues to be guaranteeing that the drug products in the
public domain have the reproducible purity and stability profile which they originally
approved. Detail of the regulations continues to be debated and the FDA remains
open to ideas for reducing the Regulatory burden carried by the pharmaceutical
companies.16 The spirit of this dialogue is seen in the Pharmaceutical Research and
Manufacturers Association (PhRMA) initiative to ensure quality while streamlining
both regulatory and development operations.17 A greater need, to enable process
innovation to continue well into the NDA phase, is addressed later (see Chapter 11).
Whatever changes are proposed, they all need to be based on scientific evidence that
they are justified and that they do not jeopardize the FDA’s main thrust, which is to
ensure that the pharmaceutical companies produce the highest-quality APIs.
In this presentation, I have provided only a personal overview of how chemical
process development scientists and engineers approach the task of creating a
chemical process for the manufacture of an API and assembling the information
and documentation needed for submission and approval of an NDA. Much detail is
lacking. It is therefore important that all readers embarking on a chemical process
development program work with their own regulatory specialists to ensure that they
are accommodating the requirements of their own company’s culture.
16Current FDA guidance documents can be accessed from the internet at http://www.fda.gov/cder/
17Cupps, T., Fritschel, B., Mavroudakis, W., Mitchell, M., Ridge, D., and Wyvratt, J. Pharm. Technol.,
February 2003, 34.
Inventions, protected for a single generation, promote the creation of new businesses,
social prosperity, and better inventions.
There are many stories connected with every process discovery and development
project. One of the least publicized is the patent story that usually unfolds in a wellorchestrated
fashion through the dialogue of individual and company inventors with
patent attorneys. Patent protection covering a given discovery lasts 20 years from the
date of filing. In practice, this time is reduced by the several years taken to test and gain
regulatory approval to market a drug product. Because of such delays, it is possible
to gain additional years of patent protection in several countries to compensate for
the time spent in the regulatory review process.1 In the United States, the extension
provided is for up to half the IND time and all the NDA time. However, the total
extension time cannot exceed 5 years and the extended patent term cannot exceed
14 years from FDA approval.2
Patent exclusivity enables a company to recoup its research and development
investment and make an appropriate profit. Once patents expire, generic competition
greatly erodes sales, although generic competition is frequently delayed by the
portfolio of patents usually created by a company. It will be apparent that because
of the enormous costs of drug development and the need to recoup these costs, all
drugs taken to the marketplace should have some form of patent protection, even if
1Health News Daily, October 1997.
2Associated Press, October 1997.

Copyright C 2008 John Wiley & Sons, Inc.
coverage only pertains to the use of an old (unpatentable) chemical structure for a
specific hitherto unrecognized and therefore new medical indication.
One of the greatest satisfactions obtained by the chemist seeking a patent to
protect his or her invention lies in meeting the intellectual challenge associated with
analyzing the existing literature (art) and defining the invention. In competitive fields,
the challenge is often to stake out a strategic position free of the impediment of thirdparty
patent rights, while building a patent portfolio to protect one’s own position.
This may entail licensing third-party rights, or, where validity is in doubt, attacking
the patent before the courts in proceedings, which can be complex and very costly.
In this chapter, I will provide background and a few personal experiences to illustrate
the great importance of protecting intellectual property through the patent system.
The worldwide patent system harmonized considerably during the 1980s and
1990s. It continues to harmonize as more countries sign the General Agreement on
Trades and Tariffs (GATT) Treaty, requiring enhanced patent protection for pharmaceutical
inventions over what many countries had previously provided. The subject
matter of a patent filing not onlymay be the drug itself, as a novel chemical compound,
but also may be later-found advantageous forms of a drug; these include polymorphs,
salts, dosage formulations, and combinations of drugs. Most important, from the
standpoint of Chemical Development, process features may be patented such as
new synthetic methods and intermediates used for drug manufacture, unprecedented
separation, and purification technologies and novel equipment applications.
A brief review of the following aspects of patent activity provides perspective:
1. Patent content
2. The driving forces leading individuals/companies to seek patents
3. Factors influencing the worth of a patent
4. Timeliness in seeking a patent
5. The defense of patent property
6. Designing around process patents
7. Patenting versus trade secrets
8. Patent aspects of the development of processes for Florfenicol manufacture
9. A lighter side of the patent literature
Several examples, most of which are based on the author’s personal experience,
are used to illustrate these areas.
To obtain grant of a patent, it is necessary to prepare and file a patent application
describing the alleged invention and then legally defining what the inventor believes
his invention to be in terms of a series of patent claims.
In writing a patent application, the inventors generally give a broad overview
of the field (with references) and describe their invention in relation to the public
disclosures. To support patentability, the inventors will often provide “surprising”
information showing why the invention would not be obvious to those “skilled in
the art.” It is incumbent on the inventor(s) to disclose in the patent application the
best method known to them for carrying out the process claimed in the application.
The claimed invention will identify the features responsible for the patentability
of the process (e.g., temperature, pressure, pH, and other ranges) by a cascade of
patent claims defining the invention from the broadest which can be justified down
to the invention in its narrowest terms. The latter is essentially the very specific
and preferred core process that the inventor actually carries out in practice. Where
possible, compound claims will be included covering key useful intermediates.
The filed patent application will be examined by the receiving patent office to
ensure that the claimed invention meets the statutory requirements for patentability.
In order to be patentable, the claimed invention must meet three key requirements:
 It must be novel (i.e., be a novel compound/process/medical use not previously
described or used anywhere).
 It must be nonobvious (i.e., it would not occur to one skilled in the art to arrive
at the claimed invention).
 It must be useful (i.e., it must concern a composition, process, or novel device
that is capable of commercial application).
Patent examiners, in examining a patent application, will search the literature to
determine whether, in their view, the claimed invention is already described in the
literature or “prior art” (i.e., the invention is not novel), or whether it can be said to
be obvious in view of what is described in the literature. The findings of the examiner
will be reported to the inventor in an Official Action or rejection notice, identifying
what the examiner considers to be the closest “relevant art.” For the purposes of
illustration, a U.S. rejection on the grounds of obviousness will usually start:
The following is a quotation of 35 USC 103(a) which forms the basis for all
obviousness rejections set forth in this Office action:
(a) A patent may not be obtained though the invention is not identically disclosed or
described as set forth in section 102 of this title, if the differences between the subject
matter sought to be patented and the prior art are such that the subject matter as a whole
would have been obvious at the time the invention was made to a person having ordinary
skill in the art to which said subject matter pertains. Patentability shall not be negatived
by the manner in which the invention was made.
Invariably, the examiner will find what he or she considers to be “relevant art” even
though in practice it may not be pertinent. It is the inventor’s task to explain why the
relevant art is not pertinent.
In process chemistry terms, change is a continuum. Discoveries and their development
lead to processes that give higher yields, create cost reduction, introduce
environmentally cleaner and safer chemistry, provide a more stable product, and so
on, all highly desirable and advantageous results. Where patentable inventions arise,
individuals (in companies individuals generally assign their inventions to the company
they work for) file patent applications directed to those inventions. In this way,
companies create a portfolio of patents protecting the inventions and covering their
drug substance or product, and processes thereto. This portfolio extends patent coverage,
and with it company sales, often for years beyond expiration of the original patent
granted by the patent office. It is important to reiterate that invention is generally not
just an incremental improvement or evolution of a process. It must be a significant
or even revolutionary change, a new and unprecedented step, which is commercially
advantageous over the existing “art.” New patents that cover a drug substance, such
as a more stable form (e.g., a salt or a solvate) or novel and useful intermediates,
represent the strongest patent protection; all of these have to show some unexpected
advantage over the previous “art” to be patentable. “Use patents,” covering a new
therapeutic use of an old compound, may also provide effective protection. Patents
that cover processes only are generally more easily circumvented (see section entitled
“Designing Around Process Patents”).
The cost of obtaining and maintaining a patent has to be justified by the potential
value of the patent obtained. Clearly, a new drug substance patent provides a very
broad scope of patent protection against copiers for the life of the patent. Then if later
on a market switch from the original drug to a more pharmacologically advantageous
form is found (say a salt, solvate, or pro-drug), this new form may itself be patentable.
This situation would arise if the new form demonstrably enhances the drug product’s
performance (e.g., giving higher blood levels that increase therapeutic value). Such
findings often extend the effective patent protection for a product and hence the
associated market.
Sometimes patent protection may extend beyond the literal wording of what was
initially protected. An example may illustrate this. In the 1970s, Eli Lilly’s very
successful antibiotic Cephalexin was the subject of a patent. Squibb then came onto
the market with Cephradine. The Courts found that Squibb was infringing Lilly’s
patent on the basis that Squibb’s Cephradine contained a few percent of Cephalexin.
Reportedly, Squibb settled by agreeing to a royalty.
Cephradine Cephalexin
Patents, besides protecting a marketing position, may also act as a source of licensing
income or as negotiating tools.
Diligence of pursuit is often critical in winning a patent race in a competitive market
area. Where several parties are working in the same field and independently arrive
at essentially the same invention, it is generally the first party to file an adequately
supported patent application that achieves grant of a valid patent and hence is in a
position to dominate (or roof) the other party (it should be noted that a patent gives a
right to exclude others, rather than a positive right of use). This system of according
priority to competing inventions is termed the “first to file” system. Critics say that it
leads to a counterproductive “race to the patent office.” It is, however, a simple system
for according priority to inventions and is now universally adopted throughout the
world, with the notable exception of the United States. In the United States the patent
goes to the “first to invent” rather than the “first to file.” I do not need to comment
any further on the U.S. “first to invent” system because the examples on timeliness
given below relate to activities judged primarily according to who was the first to file.
One example illustrating the importance of diligently pursuing an invention occurred
in Glaxo in the late 1960s during a period when the PCl5-cleavage process was
being exploited for the conversion of the amide side chain of fermented or semisynthetic
penicillins and cephalosporins to the corresponding 6-aminopenicillanates and
7-aminocephalosporanates (Scheme 1).
Temporary protecting groups [R’ = residue of a carboxylic anhydride,
trimethylsilyl (Beecham and Gist-brocades), chlorodimethylsilyl (Bristol–Myers)]
were usually used for the production of 6-aminopenicillanic acid (6-APA) and 7-
aminocephalosporanic acid (7-ACA). Labile carbon esters were often preferred where
it proved preferential to leave the carboxylate protecting group intact while additional
manipulations were carried out on the bicyclic ?-lactam [e.g., p-nitrobenzyl (Lilly),
diphenylmethyl (Glaxo), p-methoxybenzyl (Otsuka), 2,2,2-trichloroethyl (Ciba)]. Labile
refers to the ready removal of the carbon ester—p-nitrobenzyl by hydrogenolysis,
diphenylmethyl and p-methoxybenzyl by acids, and 2,2,2-trichloroethyl by zinc reduction.
R" = R' or H
R' = Carboxyl protecting group
1. PCl5/Me2NPh
2. Alcohol
3. H2O/H+
SCHEME 1. The temporary protection of carboxyl using phosphorus trichloride.
In order to provide a protecting group of our own for 7-ACA manufacture, we
in Glaxo evaluated many options for the temporary blockade of the carboxylic acid
group of cephalosporinCand its derivatives. These included compounds such as BCl3,
POCl3, SOCl2, and PCl3. Unfortunately the evaluation of too many compounds all at
once delayed our realization that PCl3 at low temperature (approximately ?20?C to
?30?C) was a viable candidate. Although Glaxo’s patent department filed our patent
application within six weeks of realization of its superiority, we learned several
months later that Professor Ishimaru of Osaka University had filed a patent some four
months before us.3(a) Professor Ishimaru visited Glaxo. Negotiations took place, but
no agreement could be reached. Although Glaxo allowed its patent to go through to
publication,3(b,c) it dropped the development of PCl3 as a temporary protecting group.
A happier result occurred a few years ago in our Chemical Development group
in the Schering–Plough Research Institute. We sought a process for producing 2-
phenyl-1,3-propanediol (PPD), an intermediate for the antiepileptic drug Felbamate,
which Schering licensed from Carter–Wallace for marketing in Europe.
The then-commercial process required several steps, used relatively hazardous
reagents, and employed a difficult hydrogenation step. We projected the cost-ofgoods
(COG) for PPD produced by the commercial process4(a) would lead to a COG
for Felbamate that was more than double the figure desired by our marketing group,
even allowing for the economies of large-scale operation. A low COG was considered
essential because of the indicated high dosage (2 g/patient/day, with some groups of
epileptics receiving a dosage of more than double this figure in early treatment).
We set ourselves the urgent objective of finding a shorter, simpler, safer process
based on methyl phenyl acetate (cost for large tonnages: ?$5.50/kg in 1992). Based
on our ideas for the conversion of methyl phenyl acetate to PPD, we calculated that
the COG for PPD might be reduced to about a third of the COG projection for
the commercial process. The marketing group agreed that in order to control our
own destiny, we needed to be in charge of the technology and supported an urgent
evaluation program. We proposed to prove and, if possible, patent our technology,
and then take it to several potential partners with a view to identifying the one or two
most suited to adopting the technology and meeting our COG target.
3(a) Ishimaru, T., and Kodama, Y. U.S. Patent 3,896,118,1975 (priority date November 17, 1970, to
Toyama). (b) Chapman, P. H., and Holligan, J. R. U.S. Patent 3,882,108,1975 (to Glaxo). (c) British Patent
1,391,437,1975 (priority date April 7, 1971, to Glaxo).
4(a) Choi, Y. M. U.S. Patent 4,982,016,1991 (to Carter Wallace).
Ethyl tropate
SCHEME 2. Possible ethyl tropate route to PPD.
Our task was complicated by the fact that we were aware, from our licenser,
that others were working on the same problem, might be ahead of us, and may be
developing patent positions that could potentially negate our efforts.
Work published in 19894(b,c) showed that Choi (Carter–Wallace) had prepared
PPD by the reduction of diethyl phenylmalonate using lithium aluminum hydride or
lithium borohydride. This approach did not, however, appear to us likely to meet the
requirement for a low COG.
In June 1992, we reasoned that the preparation of methyl tropate via the reaction
of formaldehyde with methyl phenyl acetate may provide a low-cost intermediate
capable of being reduced to PPD. The preparation of ethyl tropate via this route
was already published.5 However, the German workers found that formation of ethyl
2-phenylacrylate was substantial. In addition, no ongoing reduction of methyl tropate
to PPD was described (Scheme 2).
Our Chemical Development team, under Dr. Chou-Hong Tann, energetically and
creatively undertook a vigorous laboratory effort to try to improve the above process
to enhance the yield of methyl tropate and also to investigate the reduction of this
compound to PPD. Study of the formaldehyde reaction did not quickly yield much
improvement, so within weeks and not without overcoming some reluctance to abandon
the methyl tropate route, we phased in an exploration of the reaction of methyl
formate with methyl phenylacetate, followed by reduction with sodium borohydride.
This route almost immediately showed great promise (Scheme 3).
Within a matter of 4–6 weeks after initiation of the methyl formate program,
we had defined an outline of a process and gathered sufficient results to initiate
(b) Choi, Y. M., Emblidge, R. W., Kucharczyk, N., and Sofia, R. D. J. Org. Chem., 1989, 54, 11194. (c)
Choi, Y. M., and Emblidge, R. W. J. Org. Chem., 1989, 54, 1198. A somewhat lower cost process via
2-nitro-2-phenyl-1,3-propanediol [Stiefel, F. J., U.S. Patent 4,868,327, 1989 (to Carter–Wallace)] was also
practiced commercially.
5(a) Schwenker, G., Prenntzell, W., Gassner, U., and Gerber, R. Chem Ber., 1966, 99, 2407. (b) USSR
Patent Application 322,988,1976; this describes a method for preparing tropic acid by the reduction of
alkyl esters of formyl phenylacetic acid using borohydrides. No reference is made to the reduction of both
CHO and CO2C2H5 groups in this patent.
Not Isolated
SCHEME 3. Methyl formylphenylacetate route to PPD.
preparation of patent applications6 supporting our earlier calculations that the cost
of PPD might well be reduced to one-third the cost of PPD made by the commercial
process. We also perceived environmental and quality advantages. A priority patent
application covering this route was filed in the U.S. Patent office on September 18,
1992.6(a) In addition to claiming the overall process outlined above, and the best range
of reaction conditions, we also sought claims for boron intermediates:
wherein X, Y, and Z independently represent H, OH, O , OR or OCOR 2
wherein R is alkyl or aralkyl
wherein R is C to C alkyl (a further definition of Y in the nonionic 2 1 6
formula above is OCH CH(Ph)CH OH) 2 2
wherein M is a metal of groups I, II, or III of the periodic table
The Schering U.S. priority application dated September 18, 1992 formed the basis
for an international Patent Convention Treaty (PCT) application, covering Europe,
which was filed eight months later.6(b)
As should be common practice when working in a competitive field, we maintained
a watch worldwide for relevant third-party patents and patent applications as
they were published. In Europe and most other countries, patent applications are
published 18 months from the earliest priority date. In the United States, at that time,
patents were not published until grant (the law has recently changed7), but grant can
often occur earlier in the United States than in other major countries. That was the
6(a) The original U.S. Patent Application, filed September 18, 1992, by Walker, D., Babad, E., and Tann,
C.-H., was not pursued to the publication stage when Schering–Plough abandoned the Felbamate project.
(b)Walker, D., Babad, E., Tann, C.-H., Tsai, D. J., Kwok, D-I., Belsky, K. A., and Herczeg, L. International
Patent Applications WO 94/27941, Priority date May 25, 1993 and WO 94/06737 (March 31, 1994) to
Avondale Chemical Company Division of Schering–Plough.
7In keeping with the worldwide patent law harmonization effort, the U.S. House of Representatives and
the Senate Judiciary Committee approved a Bill in 1997 requiring the publication of Patent Applications
in 18 months; the 1997 House version did, however, carry an amendment exempting individual inventors
(not companies) and Universities.
case here when the U.S. patent8(a) to Johnson et al. was granted and published on
August 24, 1993.With that publication, we realized that unbeknownst to us, Johnson
et al. had been working along similar lines to ourselves. The Johnson et al. patent was
similarly concerned with 2-aryl-1,3 propanediols, more specifically PPD. It essentially
disclosed the preparation, isolation, and reduction of methyl tropate (prepared
from methyl phenylacetate) and claimed a priority date of September 16, 1992 (versus
our September 18, 1992 for the reduction of methyl 2-formylphenylacetate). Of
most concern, however, was that the U.S. Johnson et al. patent, in addition to having
claims to the process of its invention, also had product claims to boron intermediates
of the structure:
wherein M+ is a metal of Groups I to III of the periodic table, or quaternary ammonium.
There was no spectral or other evidence to support the structure claimed.
We realized that should Johnson et al. obtain grant in Europe of valid claims to those
boron intermediates, then those claims could be an impediment to our proposed manufacture
of PPD in Europe; Johnson et al. had a priority date two days earlier than
We anticipated that Johnson et al. would file in the PCT countries within the
permitted one year period expiring September 16, 1993. We needed to see what
Johnson et al. would claim in Europe. In Europe, in contrast to the United States,
the patent prosecution file is open to public inspection. It was thus possible to find
out that Johnson et al. had indeed filed a corresponding patent application8(b) in
Europe and to see what patent claims they were pursuing. We found that the claims
being pursued were essentially the same as those in the United States. Thus, only the
claims to boron intermediates were a potential impediment to our planned European
manufacture. Could Johnson et al. obtain valid claims to such intermediates?
In considering this question, we noted the following. It seemed to us that Choi,
in his earlier 1991 disclosure of the lithium borohydride reduction of diethyl 2-
phenylmalonate, may well have produced boron intermediates falling within the
Johnson et al. intermediate patent claims. If this were so, then the Johnson et al.
patent claims to boron intermediates would be invalid if they were the same as the
boron intermediates previously produced by Choi. In short, the Johnson et al. claims
would be open to attack as lacking novelty. We investigated. We undertook a boron
NMR analysis of the following fully reduced solutions:
 Reduction of diethyl 2-phenylmalonate with lithium borohydride precisely following
the Choi procedure.4
8(a) Johnson, F., and Miller, R. U.S. Patent 5,239,121,1993 (filed September 16, 1992 to Ganes Chemicals).
(b) Equivalent European Patent Application 588652 A1 (filed September 16, 1993).
 Reduction of ethyl tropate with sodium borohydride precisely following the
Johnson et al. procedure.8(a)
 Reduction of methyl 2-formyl 2-phenylacetate with sodium borohydride using
the Schering procedure.6
All three solutions showed the presence (NMR) of the same boron intermediates.
In light of this evidence, it was clear that Johnson et al. could not obtain a grant in
Europe of valid patent claims to boron intermediates that would be an impediment
to us. (Incidentally, for the same reason, Schering could not obtain such intermediate
claims.) On the other hand, the Johnson et al. methyl tropate reduction process was
patentable over the disclosed Choi diethyl phenylmalonate reduction process and
Schering’s formyl phenylacetate process was patentable over both the disclosed Choi
and Johnson et al. processes. Both were novel and both had a basis for showing
Johnson et al. likely became aware that they could not obtain valid intermediate
claims in Europe that would impede Schering because eventually they withdrew their
European patent application.
Johnson et al. had attempted to cover the Schering reduction process in a later
patent filing9 having a priority date of November 17, 1992. But they were too late.
Schering had beaten them with a priority date of September 18, 1992. Thus the speed
with whichDr Tann and his group phased in themethyl formyl phenylacetate approach
provided Schering with the lowest-cost technology and a vital patent position.
As indicated, pharmaceutical companies protect the fruits of their research—valuable
newdrugs—by patents. Where countries provide product protection for newdrugs, the
protection is strong. However, in days gone by, and to a lesser extent even now, some
countries only provided, or still only provide, protection for the patentee’s disclosed
process for making the drug. Such “process protection” is not strong protection
because generic companies can frequently devise other different processes, outside
the process patent, for producing the drug—for instance, by reversing the order of
process steps specified in the process patent claim.
Alternatively, where the basic patent to the new drug has expired, and hence others
are free to use the old process disclosed in the patent, the generic company can assert
that it is using that old process to prepare the drug even though in the meantime the
patentee will have devised newer, more efficient improved processes, the subject of
later patents, and the generic company is really using one of those newer patented
processes. In such circumstances, the patentee is faced with the question, Which
process is the generic company using, the less efficient off-patent process or the
improved patented process?
9Johnson, F., and Miller, R. U.S. Patent 5,250,744,1993 (filed November 17, 1992, to Ganes Chemicals).

Polytrimethylsilyl Kanamycin A
SCHEME 4. Bristol–Myers process for the production of Amikacin.
The most widely used method of challenging the legitimacy of the process used
to manufacture the generic product is to undertake an “impurity profile” study of the
drug substance extracted from the competing product, in short to check its impurity
“fingerprint” vs. your own product. Most companies can describe cases. One in my
experience occurred when an Italian company decided to market Amikacin, at that
time a patented Bristol–Myers’ product, in Korea.
After years ofwork, we in Bristol–Myers created an elegant and very practical process
for the manufacture of Amikacin which we commercialized. In brief, the process
involves solubilization of Kanamycin A in organic solvents by trimethylsilylation
(hexamethyldisilazane-HMDS) followed by acylation with S-4-benzyloxycarbonyl
amino-2-hydroxybutyric acid activated by formation of an active ester with Nhydroxynorbornene-
2,3-dicarboximide (BHBA active ester) (Scheme 4).
The impurities were primarily compounds which carried the 4-amino-2-
hydroxybutyric acid side chain on other than the C-1 amino group of Kanamycin
A, including diacyl products. One extraordinary and unexpected result of using
the trimethylsilylation approach to solubilizing Kanamycin A in organic solvents
preparatory to acylation was the total absence of any acylation at the C-3 amino
group. Thus the impurity profile of the Amikacin prepared via polytrimethylsilyl
Kanamycin A was quite unique—no other of the several processes described for the
acylation of Kanamycin A gave Amikacin containing no C-3-amino product. Our
work was patented10 and published.11 Later, at a meeting on another subject with the
10Cron, M. J., Keil, J. G., Lin, J. S., Ruggeri, M. V., and Walker, D. U.S. Patent 4,424,343, 1984 (to
11Cron, M. J., Keil, J. G., Lin, J. S., Ruggeri, M. V., and Walker, D. Chem. Commun., 1979, 266.
Italian company which was about to market generic Amikacin in Korea, they congratulated
me, on the side, for our work on the Amikacin process. As soon as the Italian
company’s Amikacin appeared on the Korean market, we obtained samples and analyzed
them. When the result came back showing that the Italian Amikacin contained
the same profile of impurities as the Bristol–Myers product, and specifically that it
contained no C-3-amino product, we were able to force the Italian product off the
Korean market on the grounds of their infringement of the Bristol–Myers patent.
The competitive nature of the pharmaceutical industry often results in lead compounds
being identified which are picked up by many companies working to grow
their franchises in core businesses. Thus there are many companies with substantial
business based on the same molecules. Three of the major molecular foundations of
my experience are as follows:
Steroids Pfizer (Upjohn), Aventis (Roussel), Glaxo–Smith–Kline,
Schering–Plough, AKZO (Organon), Hofmann LaRoche
(Syntex), Merck, American Home Products (Wyeth), etc.
Penicillins/cephalosporins Eli Lilly, Glaxo–Smith–Kline (Beecham),
Bristol–Myers–Squibb, Pfizer, Shionogi, Fujisawa, Takeda,
Aventis (Roussel/Hoechst), Merck, Novartis (Ciba), etc.
Aminoglycosides Meiji, Eli Lilly, Schering–Plough, Pfizer,
There are many companies working on many other molecular foundations—for example,
alkaloids, peptides, proteins, benzodiazepines, statins, and so on.
Companies working on the same core molecule build patent franchises to protect
their positions, often designing around the patent positions of competitors. One such
situation arose when Beecham saw Bristol–Myers enter the Japanese Amoxicillin
market. Beecham’s belief was that Bristol–Myers was using its patented process for
the manufacture of Amoxicillin (Scheme 5).
Beecham commenced court action in Japan to get Bristol–Myers off the market,
accusing them of infringing its patented process. As the reports of that litigation show,

SCHEME 6. Bristol-Myers process for the production of Amoxicillin.
Bristol–Myers denied the Beecham accusation, stating that it had a process of its own
which did not involve the acylation of the bis-trimethylsilyl derivative of 6-APA.
However, Bristol–Myers refused to disclose its process to Beecham on the grounds
that the process was still under patent prosecution with the U.S. Patent Office. This
dilemma was resolved when Bristol–Myers and Beecham agreed that the Japanese
judge presiding over the court action could visit Bristol–Myers’ plant in Sermoneta,
Italy, where the process was being run, to judge whether or not patent infringement
was occurring.
In reaching its decision to launch Amoxicillin on the Japanese market,
Bristol–Myers had recognized the need to design around the Beecham patent and
identify and develop a process which would be competitive.
Our Chemical Development group in Bristol–Myers had earlier held many brainstorming
sessions to find a solution to the problem. The solution came from a reading
of Russian literature12 describing the conversion of TMSNH· to TMSO2CNH·
We found that the bis-trimethylsilyl derivative of 6-APA reacted quantitatively with
dry CO2 to give a trimethylsilylcarbamate intermediate which in turn reacted almost
quantitatively with p-hydroxyphenylglycyl chloride to give Amoxicillin (Scheme 6).
The formation of the carbamate could be readily followed by NMR, and sufficient
spectral evidence was gathered to allow Bristol–Myers to unequivocally claim the
trimethylsilylcarbamate derivative of monotrimethylsilyl 6-APA.
The mechanism of the acylation reaction was never studied. We, perhaps naively,
hypothesized the mechanism involved:

However, since a process patent was granted to Bristol–Myers13 no further work was
Assertion of property rights of a different kind occurred in the late 1960s in the then
commercially important era of penicillin sulfoxide ring expansion to cephalosporins.
An early focus of commercial attention was finding a commercially viable blocking
group for the 3-carboxyl group of the penicillin.Ciba, through itsWoodwardResearch
Institute in Basel, found that the trichloroethyl ester of penicillin sulfoxides could be
produced in high efficiency, the product ring could be expanded to the cephalosporin,
and the trichloroethyl ester protection could be removed by treatment with zinc.14
The process is outlined in Scheme 7.
The use of the 2,2,2-trichloroethyl (TCE) group for carboxyl protection became
of interest following publications byWoodward et al.14(c,d,e) As a result, TCE protection
was used by Glaxo in the early stages of developing a process for Cephalexin
manufacture. At the same time, Glaxo initiated work to find an alternative to TCE,
recognizing that Ciba may patent its TCE-based process, and also recognizing that
TCE protection was unsuitable from an industrial standpoint. Glaxo’s plant in Montrose,
Scotland, where the step of removing the TCE group was being carried out, was
encountering zinc/zinc chloride waste disposal problems owing to the spontaneous
combustion properties of this waste.
Glaxo initiated evaluation of two other carboxyl protecting groups, namely
diphenylmethyl (DPM) and p-nitrobenzyl (PNB). Of these, development of the PNB
group was already well-advanced by Eli Lilly, with whom Glaxo had a working relationship
through the U.K. National Research Development Council (NRDC). [The
NRDC held the patents on Cephalosporins and licensed rights to several companies,
including Glaxo and Eli Lilly.] Evaluation of the two blocking groups was resolved
13Walker, D., Silvestri, H. H., Sapino, C., and Johnson, D. A. U.S. Patents 4,240,960,1980; 4,278,600,1981;
4,310,458,1982 and 4,351,796,1982 (all to Bristol–Myers).
14(a)Woodward, R. B. British Patent 1,155,016,1969. (b)Woodward R. B. U.S. Patent 3,828,026,1974 (to
Ciba–Geigy). (c) Woodward, R. B. Science, 1966, 153, 487. (d) Woodward, R. B. Angew. Chemie, 1966,
78, 557 (e)Woodward, R. B., Heusler, K., Gosteli, J., Naegeli, P., Oppolzer,W., Ramage, R., Ranganathan,
S., and Vorbr?uggen, H. J. Am. Chem. Soc., 1966, 88, 852.
in favor of the DPM ester since we in the Chemical Development Group (in Glaxo’s
Ulverston manufacturing plant in the United Kingdom) had created and developed a
novel low-cost process for introducing DPM ester protection. We also demonstrated
its value in the manufacture of Cephalexin DPM ester and showed that the DPM
group was readily removed to give high-quality Cephalexin. The new process was
patented.15 Glaxo paid a royalty to Ciba during the short period of its use of the
Woodward patent.
Most improvements to commercial processes are not patented. These generally result
from efforts by the process improvement and development groups on the production
site to increase productivity and reduce costs. These efforts include increasing yield
by better understanding of the chemical transformations, increasing reaction concentration,
minimizing isolations (e.g., by combining process steps), recycling waste
more efficiently, changing solvents to harmonize with other plant processes, improving
environmental compliance, improving product quality, and finding lower-cost,
quality sources of raw materials. Not infrequently, better chemistry may be devised,
steps reversed advantageously, and other innovations made which could create a
patentable situation. Often the judgment call is made to keep such innovations as
trade secrets. This can be dangerous and expensive, since if others obtained a patent
covering the technology, those who practice the innovation as a trade secret could
be considered infringers who may be forced to either stop using the technology or
pay royalties to the patent holder! Happily, this way of thinking is being outmoded
to allow the original trade secret user to keep using the technology without license.
Europe now recognizes prior use as a defense to an infringement action, and the
United States is moving toward recognizing the right of prior use.
Several patent situations emerged during the course of the development of a process
for the manufacture of Florfenicol. Those described in this section illustrate the
importance of the following:
1. When third parties are involved in process research and development work,
agreements are needed to cover both the objectives of the joint program of work
and definition of the ownership of intellectual property (patents) discovered by
the third party. Such agreements benefit all parties.
15(a) Bywood R., Gallagher, G., Sharma, G. K., andWalker, D. German Patent 2,201,018,1972 (to Glaxo).
(b) Bywood, R., Gallagher, G., and Walker, D. German Patent 2,311,597,1973 (to Glaxo).
2. The company that owns the patents is in charge of the technology that develops;
this is vital for controlling Active Pharmaceutical Ingredient (API) or Intermediate
supply and provides flexibility in selecting manufacturing partners—for
example, to obtain the lowest cost of goods (COG).
3. Quality issues can raise questions that galvanize the search for improved processes,
often leading to new and patentable technology.
4. Failure to demonstrate a chemical synthesis idea in one laboratory need not
discourage others who believe they can make that idea work.
5. Designing around the patent positions of others frequently requires the kind of
thinking that identifies new and sometimes revolutionary approaches to a synthesis;
competitive instincts and seemingly insurmountable obstacles stimulate
new insights and speed innovation.
Initial Process Exploration with Third Parties
Florfenicol (IX) was patented by Schering–Plough as a broad spectrum antibiotic
with Gram-positive and Gram-negative activity comparable to chloramphenicol (X).
Chloramphenicol had become severely restricted in use, owing to its propensity to
cause blood dyscrasia (aplastic anemia) in some patients.
(R) (S)
(R) (R)
(R) (R)
Florfenicol was believed to be more like thiamphenicol (XI), which appeared to be
free of the blood toxicity problems associated with chloramphenicol. The toxicity of
chloramphenicol was loosely linked to the presence of the nitro group. Nevertheless,
Schering regarded Florfenicol as too risky to develop for human use, leaving the
compound to be picked up by Schering’s Animal Health Division for use as an
animal antibiotic.
In order to progress the development of Florfenicol, Schering produced its earliest
needs (for safety and early clinical studies) using a process based on thiamphenicol
(Scheme 8):16
Schering–Plough undertook a broad-based effort with third parties, both university
and industrial, to search for a more practical process than that outlined in Scheme 8.
The most important collaboration was that with Zambon S.p.A in Italy.
Zambon already marketed Thiamphenicol, which it produced in Italy via the chiral
intermediate XII.17 In 1980 Zambon undertook the search for a synthesis scheme
16Nagabhushan, T. L. U.S. Patent 4,235,892, 1980 (to Schering–Plough).
17(a)Kleeman, A., Engel, J.,Kutscher, B., and Reichert, D. Pharmaceutical Substances; Syntheses, Patents
and Applications, 4th edition, Georg Thieme Verlag, Stuttgart, 2001, p. 2016. (b) Jacquez, J., Collet, A.,
and Wilen, S. Enantiomers, Racemates and Resolutions, John Wiley & Sons, New York, 1981, p. 223.
OH NPhth
1. NH2OH
2. Cl2CHCO2CH3
SCHEME 8. Conversion of Thiamphenicol to Florfenicol.
1. CH3SO2Cl/base
2. KF/polyethylene glycol
SCHEME 9. Zambon process for producing Florfenicol.
which would give a less expensive product than that afforded by Scheme 8. The
major problems with Scheme 8 lay in the cost and potential hazard associated with
the production and use of the DAST reagent, and also with the expense of removing
the difluoro impurity resulting from “fluorination” at both the primary and secondary
alcohol positions of XIII.
Zambon devised the “fluorination” process outlined in Scheme 9, and used the new
intermediate XIV to produce useful quantities of Florfenicol for Schering–Plough’s
ongoing development programs.
The process steps up to compound XV were the subject of disclosure in a Zambon
patent.18 The Zambon patent was part of a family of patents that Zambon was to build
up over time, in both the United States and Europe, covering its process R&D work.
Claim 1 in Zambon’s patent, covering the process to compound XIV is provided
to illustrate the breadth and complexity of their coverage. It can be adduced that
compound XVI in Claim 1 embraces the key intermediate XIV. Claim 1 reads as
18Jommi, G., Chiarino, D., and Pagliarin, R. U.S. Patent 4,743,700,1988 (to Zambon S.p.A.).
A process for preparing a compound of formula XVI (formula III in the U.S.
X1 X2
R is a methylthio, methylsulfoxy, methylsulfonyl, or nitro group, and
X1 is hydrogen, 1–6 carbon alkyl, 1–6 carbon haloalkyl, 3–6 carbon cycloalkyl,
phenyl or 1–6 carbonphenylalkyl, whose phenyl ring may be substituted by one
or two halogen atoms, 1–3 carbon alkyl, 1–3 carbon alkoxy or nitrogroups; or
X1 together with X2 is an oxygen atom or an alkylene having from 2–5 carbon
atoms; and
X2 is hydrogen, 1–6 carbon alkyl, 1–6 carbon haloalkyl, 1–6 carbon cycloalkyl
or phenyl which may be substituted by one or two halogen atoms, 1–3 carbon
alkyl, 1–3 carbon alkoxy or nitro groups; or
X2 together with X1, is an oxygen atom or an alkylene having from 2–5 carbon
atoms; or
X2 is covalently linked to X3; and
X3 is hydrogen or –COR4 wherein
R4 is hydrogen, 1–6 carbon alkyl, 1–6 carbon haloalkyl, 3–6 carbon cycloalkyl,
phenyl or 1–6 carbon phenylalkyl, which phenyl ring may be substituted by one
or two halogen atoms, 1–3 carbon alkyl, 1–3 carbon alkoxy or nitro groups; or
R4 together with X2 is
(CH2)m CH=CH
(CH2)m CH2 CH2
n is 1 or 2; m is 0 or 1; X is hydrogen, a halogen atom, 1–3 carbon alkyl, 1–3
carbon alkoxy or nitro groups; or
R4 together with X2 and X1 is a chain of formula
(CH2)p CH (CH2)q
p is 3 or 4 and q is 1 or 2; or
X3 is covalently linked with X2; a process that comprises reacting one mole of a
compound of formula
X1 X2
R, X1, X2, and X3 are as is defined above, and where
X4 is –OSO2R6 where R6 is methyl, trifluoromethyl, phenyl, p-methyl phenyl,
2-, 4-, 6-trimethylphenyl, 2- naphthyl, or 2-pyridyl with 1–15 moles of a nongaseous
inorganic fluoride in a polyglycol having at least four alkylidene oxide
units at a temperature of 40?C to 150?C and retaining said compound of formula
III formed in the reaction mixture until completion of the reaction.
This broad claim covers processes to compound XIV and XVII the compounds
most germane to Schering–Plough’s subsequent Florfenicol process interests (see
(Ishikawa reagent)
(not identified) CH3SO2 CH CHCH2F
SCHEME 10. Use of new fluorinating reagents in the synthesis of Florfenicol intermediates.
The Zambon U.S. Patent18 covered processes only. Product claims to intermediates
were divided out from that parent patent, in both the United States and Europe, and
became the subject of a later patent.19
One interesting feature of Zambon’s U.S. Patents, which might have become
fortuitous as far as Schering–Plough was concerned, was that no examples or claims
were made to the use of intermediates, produced according to their process, in the
production of Florfenicol. In short, the dotted line process steps in Scheme 9, covering
the production of Florfenicol using the process and intermediates of its U.S. Patent,
were not claimed by Zambon. Thus, had Schering decided to import Florfenicol
produced in a country that did not recognize patents (e.g., China or India at that
time), Zambon could not have prevented this. In short, Schering would not have
been infringing Zambon’s process claims in the United States by importation since
Zambon in this patent did not claim Florfenicol made by its process!
In regard to other process exploration work carried out in Universities under
Schering–Plough auspices, it is pertinent to mention the work of Professor Szarek and
his postdoctoral student at Queens University in Kingston Ontario. One particularly
interesting reaction carried out by them20 is outlined in Scheme 10. This work was
carried out in 1982. Although the hydroxymethyl methyloxazoline did not appear to
give the desired fluoromethyl compound with Ishikawa reagent, we later found that
other similar oxazolines did.
The Search for a Better Process for Florfenicol Manufacture
The Schering–Plough chemical process development group was drawn into the Florfenicol
clinical supply program in 1984 with a view to determining whether a lowercost
process might be found which would overcome the need to routinely recrystallize
the Florfenicol produced by Zambon. Recrystallization had proved necessary to
19Jommi, G., Chiarino, D., and Pagliarin, R. U.S. Patent 5,153,328,1992 (to Zambon S.p.A.)
20Szarek, W. A., and Matsuura, D. Unpublished results, November 1982.
(Ishikawa reagent)
1. Hydrolysis
2. Cl2CHCO2CH3
where R = C6H5
where R = Cl2CH
1. H2O
XIX a R=C6H5
b R=Cl2CH
SCHEME 11. Schering–Plough process for the manufacture of Florfenicol.
remove an impurity we identified to be XVIII.
resulting from step 2 of the Zambon process outlined in Scheme 9.
Work to improve the cost of goods and the quality of Florfenicol led to the process
outlined in Scheme 11. The Schering process exploration program, led by Dr. Doris
Schumacher, is noteworthy for the striking success of the Ishikawa reaction identified
in Scheme 11 versus the results obtained in Scheme 10. Dr. Schumacher’s initiative
nicely illustrates a homily I often use: Never let theory abort an experiment.
This process was the subject of two process patents,21 one for the selective oxazoline
formation step and the other for the “fluorination” step. The key factor in
obtaining a patent for the “fluorination” step was the use of pressure. Application
of process conditions previously described by Ishikawa and co-workers22 gave poor
yields. When the reaction was carried out in dry methylene chloride at 100?C in
an autoclave, the yield of high-quality Florfenicol intermediates (XIV or XVII) was
almost quantitative.
A study of the hydrolysis of XIX(b) revealed that the “free base” form was very
stable to the action of water, whereas the Ishikawa reaction mixture (containing HF)
21(a) Clark, J. E., Schumacher, D. P., and Wu, G.-Z. U.S. Patent 5,382,673,1995 (to Schering–Plough).
(b) Schumacher, D. P., Clark, J. E., and Murphy, B. L. U.S. Patent 4,876,352,1989 (to Schering–Plough).
22Takaoka, A., Iwagiri, H., and Ishikawa, N. Bull. Chem. Soc. Japan 1979, 52, 3377.
XIX (b)
"free base"
in CH2Cl2
XIX (b)
reaction solution in
2 moles H2O
3 days at 20°C
2 moles H2O
< 1 hr at 20%
100% Florfenicol
No reaction
SCHEME 12. Hydrolysis of XIX b.
was rapidly hydrolyzed to Florfenicol (Scheme 12). Thiswork led to the establishment
of a “one-pot” process for the manufacture of Florfenicol.23
Schering–Plough’s Improved Process in Relation to
Zambon’s Patent Rights
During the period leading to Schering–Plough’s new process for the manufacture of
Florfenicol, Zambon had continued its efforts to complete its process and compound
claims to the intermediates disclosed in its U.S. Patents. In order to ensure the
best possible patent protection for Florfenicol processes and intermediates thereto,
Schering–Plough acquired rights to the Zambon patents, thus closing a chapter of
fruitful collaboration.
As the patent system harmonizes and more countries recognize the importance of
patenting intellectual property, the value of patents grows. This leads to more people
and companies seeking patents for even the most abstruse and maybe bizarre inventions.
Patents issued by the U.S. Patent Office can be viewed in some large libraries
or on the Web at www.uspto.gov. The following patents indicate there are no limits
to human imagination:
U.S. Patent 5,911,805. The inventor claims a “unique die-cut confetti that has
unusual aerodynamic features that create visually pleasing flight patterns that
have not been previously observed with confetti.” To achieve the pleasing flight
patterns, the inventor merely stamps a hole into the middle of each piece of
paper, which itself can be any desired shape (e.g., a bell, dove, etc.).
U. S. Patent 5,876,995. The inventor claims a way of making glow-in-the-dark
party drinks such as champagne. The glow is created by the interaction of
luciferin and luciferase as in the bioluminescence process, which lights the
firefly’s tail. One way of achieving this is to clone the luciferase gene into the
23(a) Example 7 in Reference 21(b) describes the “one-pot” process. (b) Wu, G., Schumacher, D. P.,
Tormos, W., Clark, J. E., and Murphy, B. L. J. Org. Chem., 1997, 62, 2996.
FIGURE 1. Illustration from French Patent 2,327,599.
yeast that makes champagne and arrange for luciferin to be introduced when
the cork is popped. One imagines that the product would require FDA approval!
My own all-time favorite bizarre patent I owe to Dr. Ken Kerridge, librarian at
Bristol–Myers. He and his colleagues recognized, in sending out a weekly summary
of relevant new patents, that the Bristol–Myers scientists needed occasional light
relief from scanning the almost mind-numbing torrent of patent information. The
following patent abstract served the need:
French Patent 2,327,599 (1977) to E. Kimmerle. This patent describes a new
bank security system:
The bank counter conceals an automatic system for seizing robbers, encircling them
with cables, releasing tear gas, etc. When a pedal behind a counter is pressed, alternate
sections of the counter screen tip over, bringing tear gas tubes to the robbers’ head
level—tear gas is released automatically as this happens. At the same time, bulletproof
screens are raised in front of the staff. In tipping, the counter screens/sections
release reels of cable stored behind them, the reels roll down ramps and then circle,
some to the left and some to the right, encircling the robbers legs. Some reels roll
straight across the banking hall; their cables catch in the legs of furniture fixed to the
floor and serve as trip wires. Small balls are distributed over the floor, and stupefying
gas is released. These measures are added to those described in the parent patent
(Figure 1).
Unfortunately, today, few chemists appear to have the time to read the patent
literature. The proliferation of journals and the pressure to accelerate projects appears
to create neglect of patents as a source of technical information. Process development
chemists and engineers do get involved in cases of patent interference, extending
patent life, and in situations where newprocesses for important compounds have some
priority. However, these are relatively minor activities. Neglect of the patent literature
is seen in the paucity of reference to patents in publications in the major journals.
Even in journals such as Organic Process Research and Development (OPRD),
one finds relatively few references to patents. This situation is recognized by the
editor of OPRD,24 and initiatives25 to bring patents to the attention of readers of
this journal have been taken. In view of the growing importance of intellectual
property and the increasing strength of a harmonizing patent system, the reading
and understanding of patents needs to be given more attention. Certainly, chemists
involved in creative/inventive endeavors need to see their work in the perspective of
patent applications. Patents are more than just a nuisance slowing the publication of
scientific findings. They are the pillars on which companies around the world build
their businesses.
24Laird, T. J. Organic Proc. Res. Dev., 2000, 4, 61.
25(a) Turner, K. J. Organic Proc. Res. Dev., 2000, 4, 68; (b) idem ibid., 2000, 4, 246.
The philosopher may be delighted with the extent of his views, the artificer with the
readiness of his hands, but let one remember without mechanical performance refined
speculation is but an empty dream and let the other remember that without theoretical
reasoning, dexterity is little more than brute instinct.
——Samuel Johnson (1709–1784)
Though the chemist may not be the philosopher in Samuel Johnson’s sense, nor
the chemical engineer merely an artificer, the quotation aptly reinforces the need for
the closest possible integration of the chemist’s and chemical engineer’s skills in the
creation of a chemical process.
The descriptive approach to illustrating the chemical engineer’s contribution to
chemical process development, which is presented in the following pages, provides
no more than an introduction to the chemical engineering discipline. There
is no substitute for a dialogue with a professional chemical engineer when developing
a process for scale-up. Published texts—for example, Griskey’s Chemical
Engineering for Chemists1 and Perry et al.’s Perry’s Chemical Engineers Handbook2—
provide more fundamental information on the field. I have personally found
1Griskey, R. G. Chemical Engineering for Chemists,American Chemical Society,Washington, D.C., 1997.
2Perry, R. H., Green, D. W., and Maloney, J. O. Perry’s Chemical Engineers Handbook, 7th edition,
McGraw-Hill, New York, 1997.

Copyright C 2008 John Wiley & Sons, Inc.
Perry’s Handbook to be particularly useful. It gives masses of physical and chemical
data on chemicals and learned information on topics such as heat and mass transfer,
pumping, flow measurement, liquid–liquid extraction, distillation, evaporation, fractionation,
reaction kinetics, membrane separation processes, chromatography, crystallization,
filtration, drying, process control, materials of construction, corrosion,
waste water management, biochemical engineering, pollution, accounting, and more.
More recently, a splendid publication by McConville3 provides an excellent source
of chemical engineering information and, more particularly, a chemical engineer’s
insights and perspective on chemical process scale-up. For a more formal chemical
engineering education, The American Institute of Chemical Engineers offers a wide
range of courses (see www.aiche.org/education).
The process of developing the chemistry needed to produce a target molecule
usually starts with the research chemist’s raw chemical transformation (a notebook
“Recipe”). Building on this, development chemists, analysts, and chemical engineers
work to better understand the chemistry and modify or change it to make it safer,
more efficient, and more suitable for larger-scale operation—during this exercise the
Recipe generally becomes a Method. Transformation of the Method into a Process is
a major endeavor often taking many years of close collaboration between chemists,
analysts, chemical engineers, and chemical manufacturing people. Nor does the
work end when the developers of a production process transfer the technology into
the manufacturing plant. The manufacturing group responsible for producing the
chemical generally achieves substantial cost reduction over the years by increasing
productivity, in particular by reducing labor costs, minimizing equipment usage, and
generally reducing plant overhead. Thus the development of a plant process is a
continuum requiring dedicated teamwork over a long period of time.
Chemists start by manipulating the chemistry, in concert with analysts. Chemists,
however, are not usually trained in chemical engineering such that integration of
the chemistry and chemical engineering disciplines is often unwittingly delayed to
the detriment of process development. As an aside, in my experience, the practical
aspects of creating and developing a chemical process are hardly studied at all in
Universities except in a small number of colleges that offer courses in industrial
chemistry or chemical technology.
The chemist does not think much about how he or she manipulates the mechanics
of a process, how chemicals are transported to the site, how they are moved from
store to bench, how these often noxious chemicals are moved from their containers
and weighed, how they are added to flasks, how the flasks are stirred and heated
and cooled, how reflux and distillation are carried out, how and even why solvent
recovery is done, how crystals are formed for the best filtration, how filtrations are
done and products washed and taken from the filter to a drier and how the drier is
operated, how dry powders are handled or offloaded, and how they are milled or
micronized. In reality, the differences between the chemist’s view of the preparation
of a chemical and the chemical engineer’s view are profound. In scale-up, the chemist
3McConville, F. X. The Pilot Plant Real Book, FXM Engineering and Design, Worcester, MA, 2002.
not unnaturally looks on the chemical plant as a scaled-up laboratory. In contrast, the
chemical engineer looks at laboratory equipment in terms of a scaled-down plant.
This is not to denigrate the chemist and his glassware. The chemist grew up
drawing chemicals from a store, manually pouring or scooping them from their
containers into graduated receptacles or onto a balance, manually adding them to
flasks set up in heating mantles or cooling/heating baths in a fume hood, manually
setting up condensers, soxhlets, Dean-Stark traps, distillation receivers, and so on,
carrying out chemical additions, heating and cooling, crystallizing and manually
lifting the flask and pouring the contents into a Buchner filter and manually washing
the filter cake, sometimes eliminating cracks in the cake with a spatula, digging out
the wet cake and loading it thinly onto a dish or paper for drying, and so on. Having
grown up with glassware and having been trained in laboratory manipulations and
chemical reactions observed through glass in a fume hood, the manual approach to
running chemical reactions is almost second nature to the chemist. He doesn’t have
to think much about solvent recovery, the handling of chemicals, or the exposure of
plant operators. The trained graduate chemist has a fair knowledge of hazards and
a rudimentary experience of safety, and he knows from Material Safety Data Sheets
(MSDSs) the dangers of handling chemicals. The chemist thinks mostly about the
molecular transformations going on and how to manage and improve them to get the
best reaction yield and product quality.
The process development chemist, by dint of exposure to scale-up, rapidly becomes
aware of the challenges posed by chemical operations on a larger scale and recognizes
the need for thinking about howto carry out routine laboratorymanipulations in larger
equipment. In this domain the chemical engineer provides an educational resource
in respect of identifying the large-scale equipment needed to meet the requirements
of the process. The engineer also provides considerations for changing the process to
meet the limitations or harness the advantages of operating in large-scale equipment.
In the pharmaceutical industry, scale-up of a chemical recipe, method, or process
invariably comes before chemists or chemical engineers are fully ready for it; this
is generally because of the urgency of the clinical supply program and the relentless
need for speed in defining and implementing a process suitable for commercial
operation. Clearly, raw chemical recipes pose the greatest problem especially from
the safety, environmental, regulatory, and logistics points of view. So little is generally
known about the chemicals, the chemical sources, outsourcing possibilities,
the chemistry involved in manipulating the transformations, the effects of changing
reaction conditions in larger-scale equipment, the limitations of available equipment,
what equipment modifications are needed to reproduce defined reaction conditions,
and how long all the necessary steps needed to commence scale-up will take. A written
procedure for every step in order of their proposed implementation is provided,
along with the chemicals, to enable a hazard analysis and hazard operability study
to be undertaken before a procedure can be approved for scale-up—the safety of
operators, chemists, and chemical engineers is paramount. The best available procedures
for analyzing the reactions and products are vital for control of the process.
It bears repeating that the thorough preparation of the batch sheet (manufacturing
directions) is a team effort that reflects the input of all those who will be involved in
running the process.
The project team in charge of scale-up usually comprises chemists, chemical engineers,
and analysts. These people are responsible for asking the sourcing specialists
to order needed chemicals, for modifying the large-scale equipment, for working out
analytical methods, and for prioritizing and scheduling. They are also simultaneously
called on to provide a likely timeline for operations to commence and for delivery of
the final active pharmaceutical ingredient (API), meeting the desired quality criteria.
At the same time, they are exhorted to do everything possible to stay within budget!
The timeline, kilos expected, and product quality are the main interests. This information
enables others, and especially those involved in toxicology, pharmaceutical
dosage form development, pharmacology, and drug metabolism, to plan their own
programs for the prioritized development of the API.
The formal program outlined above, leading to delivery of the API, meeting the
desired quality criteria, varies with the length and difficulty of the API synthesis.
Chemical engineers may not be involved to any extent in the provision of the first kilo
if the API synthesis is relatively short and straightforward. At the other extreme, the
synthesis of the first kilo may immediately require the input of chemical engineers.
One such example was provided by the synthesis of the Schering–Plough antifungal
I, wherein 1.2 tonnes of chemicals, including solvents, were required for the synthesis
of the first kilogram (Scheme 1).
The chemical engineer’s “laboratory” is a physicalworld apart. Chemicals arrive in
drums, containers, and cylinders requiring closely regulated storage; largewarehouses
have separate areas for new chemicals, for released (analyzed) chemicals, for inprocess
chemicals, for quarantined chemicals, and for final APIs (often requiring
refrigerated storage). Dangerous and controlled chemicals all need special (lock-up)
storage. Chemicals are moved by forklift trucks (often with explosion proof motors)
or “dollies,” to weighing areas (properly ventilated) for weighing or measuring the
amounts required by the operating procedure. Frequently the weighing scales are set
by the reaction vessels, allowing weighed chemicals to be directly delivered to the
vessel (sometimes the last API step chemicals have to be delivered through a glove
box or protected system to minimize microbiological contamination). Operating
staff are also trained and are issued with special protective equipment as needed, to
ensure their own protection from exposure to chemicals at all appropriate steps of the
Suitable reaction vessels are selected by the chemical engineers. These may be
glass-lined steel, steel (for good heat transfer), or vessels made of corrosion-resistant
materials (e.g., Hastelloy, teflon-coated steel, etc.). Reaction vessels in pharmaceutical
pilot plants may range from a few liters to thousands of liters and are generally
equipped for multipurpose use. Although reaction vessels come in many different
configurations, they preferably have a variable-speed agitator, attachment to an emission
control device, a sight-glass, a thermometer pocket, a pressure gauge, a nitrogen
1.0 kg 56592
1.4 kg
0.5 kg
2 steps
"Left-Hand Piece"
• 7 steps
• 40 kg of intermediates
• 900 kg of reagents, solvents, etc.
"Center Piece"
• 3 steps
• 10-kg of intermediates
• 120 kg of reagents, solvents
"Right-Hand Piece"
• 9 steps
•15 kg of intermediates
• 250 kg of reagents, solvents
2.2 kg
"Total Synthesis"
• 21 steps
• 65 kg of intermediates
• 1.27 Tonnes of reagents,
SCHEME 1. Initial chemical usage in the preparation of Sch 56592.
(or argon) blanketing port, a pressure relief line (with bursting disc), condenser access,
inlet(s) for the addition of chemicals, and capabilities for measuring liquid
levels. They are usually also connected to a similarly equipped receiver vessel for
distillation and, subsequently, process work-up. Heating and cooling are provided by
an external jacket (often split level in larger vessels), sometimes supplemented with
capability for pumping the reactor contents through an external heat exchanger. A
baffle is often a valuable addition to aid mixing, especially in large vessels. A bottom
valve (most frequently a ball valve, a rising stem valve, or a butterfly valve) enables
removal of vessel contents. A typical multipurpose pilot plant reactor package is
outlined in Figure 1.
Since most chemical processes use flammable organic solvents, the electrics in all
pilot plant and plant facilities are invariably explosion-proof.
There are many differences between pilot plant and laboratory operation. These
can be illustrated by reference to the following operations:
Heat Exchange
The scale of pilot plant operations results in slower heating and cooling rates since the
ratio of cooling or heating surface to reaction volume ismuch smaller in a plant versus
in small-scale laboratory equipment. This can pose serious problems in controlling
strongly exothermic reactions, especially if the chemist has found that the rate of
vent to catch
Vent to
scrubber for
off gas
disk Pressure
Sight glass
Split heating/
Weigh cells
Sight glass for
Vessel made of various materials, most common are stainless steel,
Hastelloy, glass-lined steel, teflon-coated steel
FIGURE 1. Typical multipurpose pilot plant reactor package.
addition of one reactant to another has to be done quickly for the best result. The heat
exchange problem is exacerbated if the reaction mixture is corrosive, necessitating
the use of glass or teflon-lined vessels. Such a situation can be accommodated to
some extent by the use of a corrosion-resistant external heat exchanger or by greater
dilution, though the latter, using a large volume of solvent as the heat sink, is not
favored by those who eventually want to create high productivity in a process. The
chemical engineer’s approach may be to pump solutions of the reactants continuously
into a cooled mixing chamber and immediately through an appropriately sized heat
exchanger into the reaction vessel.
In most pilot plants, the reactor jackets are equipped for heating with hot water
or steam under varying degrees of pressure. Similarly, in cooling, ice water (0?C)
or ice/methanol (?20?C) are frequently employed, with lower temperatures being
obtained through the use of dry ice/acetone. These systems are familiar to the chemist,
but water-based heating and cooling systems in particular bring a sense of unease,
especially to the chemical engineer asking the “what if ” question, What if water
leaked from a failing reflux condenser into say a Grignard reaction?
For the safety reason, newpilot plants are often equipped with inert cooling/heating
fluids in reactor jackets and condensers. Aromatic compounds and mixtures thereof
and silicones are widely used as heat exchange fluids. The range of operating temperatures
desired generally dictates the choice of heat exchange fluid. Aromatic heat
exchange fluids are generally made up from biphenyl–terphenyl diphenyloxide combinations
or from di- or polyalkylated benzenes. Two drawbacks of these fluids lie
in their odor and their propensity to burn in a fire. On the other hand, the more expensive
silicone fluids are somewhat safer and less odorous. The use of single fluids
for temperature control also avoids the discontinuity inherent in two media systems
(e.g., water and steam) where heat exchange is lost in switching from one medium to
In other situations, the heat of reaction in, say, a water quenching step may be
simply accommodated by feeding a reaction mixture into a vessel containing crushed
The mixing of chemicals in a pilot plant vessel, especially when a reaction is heterogeneous,
can be critical such that the design of the agitator in the pilot plant reactor
is important—for example, to avoid local concentrations of a reactant or to quickly
disperse reaction heat. The most commonly used agitators are of a propeller or anchor
type, but the chemical engineer is well-versed in setting up continuous external
reactant mixing and heat exchange devices or fabricating items such as a submerged
perforated pipe to enhance the distribution of a reactant in as even a way as possible.
Reaction quenching (e.g., with water) or solvent extraction of wanted or unwanted
reaction products from an aqueous solution generally leads to the need for the separation
of aqueous and organic solvent layers. The laboratory approach, manually
shaking a conical separating funnel and separating layers, often repeating a solvent
extraction several times for the best efficiency, is not an option in the plant. Separating
aqueous and organic layers in a plant setting poses no difficulty, using the vessel’s
bottom sight glass, when the separation of layers is clean, but poses great difficulty
when a “rag” or emulsion confuses layer separation. The development chemist works
to overcome or minimize the problem using techniques to enhance layer separation
such as the use of solvents with much higher or lower density than the water layer,
or adding salts to promote separation by increasing the ionic character of the water
layer, or by using nonpolar solvents. The chemical engineer can help by using devices
to detect major differences in the conductivity of organic and aqueous layers.
If multiple solvent extraction operations are a necessity, the chemical engineer
usually recommends a less labor-consuming process. A frequently used approach,
which can generally be automated, is to build a simple low-cost solvent extraction
column (Figure 2), wherein efficient mixing of organic solvent and aqueous layers in
a countercurrent flow system elegantly serves the purpose.
Improvements on the simple device have been patented, such as the use of a gentle
vertical agitation of the passing layers with perforated plates to improve contact of
aqueous and organic phases, thereby increasing the theoretical plate capability of
solution of
Product in
FIGURE 2. Simplified sketch of a liquid–liquid extraction column.
the column; the agitated device is known as a Karr column, after its inventor. Other,
even more sophisticated extraction equipment has been invented to meet the needs of
the fermentation industry. This is especially impressive in the penicillin fermentation
industry wherein the relatively stable penicillin salts and mycelium in the broth are
mixed with an organic solvent and acidified to give a very unstable penicillin acid in
the water medium. The penicillin acid is then immediately extracted into the organic
solvent in a mixer chamber and centrifugally separated to give a more stable organic
solution of the free penicillin acid, which is quickly precipitated as a relatively stable
salt, filtered, and dried (Figure 3). The wholemixing/acidifying/extracting/separating
process is engineered to be carried out in a fewseconds, thereby minimizing decomposition
of the penicillin molecule. Themajor process equipment firms in the centrifugal
mixer separator field are Westfalia and Alfa-Laval.
From a commercial point of view, there is an environmental downside to the
commonly used solvent extraction of aqueous solutions, namely that the extracted
aqueous layer needs to be processed (often by nomore than stripping volatile solvents)
before being sent on to standard wastewater treatment systems.
There are many ramifications to solvent extraction, especially from aqueous
processing streams. Sometimes a polar, water-soluble molecule can be advantageously
rendered lipophilic by carrying out a chemical reaction to nullify solubilizing
factors; for example, zwitterionic amino acids can be converted to extractable
acids by acylation of the NH2 group, rendering the molecule suitable for conventional
solvent extraction. One commercial example of this is the extraction of
Cephalosporin C (II) from a filtered fermentation broth by first acylating the amino
group with isobutyl chloroformate and extracting the resultant isobutyloxycarbonyl
Centrifuge bowl Acid
Whole or
penicillin broth
Mycelium (knifed off
Penicillin salt to filter
(centrifuge and drier)
FIGURE 3. Simplified sketch of centrifugal separation system used for the extraction of
penicillin from broth.
1. ClCO2Bui
of III
SCHEME 2. Process of the extraction of a Cephalosporin C derivative from fermentation
derivative (III) with methyl isobutyl ketone.4 A pure salt of II is produced by precipitation
with dicyclohexylamine5 (Scheme 2).
Another major application of extraction techniques lies in the use of resins to
recover polar and nonpolar materials from both aqueous and nonaqueous process
streams, commonly using solid ion exchange resin beads or, increasingly, macroreticular
resins, often packed in columns. The writer is most familiar with the recovery of
cephalosporin C itself from filtered broth using IRC-50/IRA-68 Resins. In addition
to removing desired, and also undesired cephalosporins (Figure 4), by absorbing and
concentrating them on the resin beads, the elution of the cephalosporins from the
resin affords some opportunity for chromatographic separation, although in practice
only lactone and desacetoxycephalosporin C can be significantly reduced by such
In the case of cephalosporin C extraction, chemical engineers played the major
role in designing the large plant needed to meet the projected tonnage off-takes. Plant
design was based on the results of resin evaluation and selection, hydraulic studies
on the resin bead size and column dimensions, Cephalosporin C loading capacities,
4Johnson, D. A., Richardson, E. J., Rombie, J. M., and Silvestri, H. H. U.S. Patent 3,573,296,1971 (to
5Brooks, T. J. U.S. Patent 3,830,809, 1974 (to Bristol–Myers).
Desacetylcephalosporin C Lactone Desacetoxycephalosporin C
FIGURE 4. Desired and undesired Cephalosporins removed by absorption on ion exchange
elution conditions, resin regeneration, and sizing of the columns and associated
equipment, including the selection of equipment for the evaporation, crystallization,
filtration, and drying steps needed to produce the isolated cephalosporin C potassium
salt. Activities of this kind, in any scale-up or plant design project, are often regarded
as the core activities of chemical engineers in the chemical process development field
and would make an interesting case study. Such a study is beyond the introduction
provided in this presentation.
Although widely practiced both in the laboratory and in the plant, it is always necessary
to consider the consequences of subjecting a reaction mixture being distilled
to heat. There are safety concerns, product stability issues, environmental considerations,
solvent recycling or disposal factors, time limitations, and cost constraints.
Safety and product stability issues associated with the distillation/evaporation
process are discussed in Chapter 4.
Distillation is carried out for many reasons: concentration of a substrate in a solvent,
removal of one solvent from a mixture and often its replacement by another,
azeotropic removal of unwanted solvents (very frequently azeotropic drying), fractionation
to separate a pure solvent for reuse, and removal of a low-boiling product
of a reaction frequently to prevent its further reaction with the initial substrates or
product (e.g., removal of water by passing a wet distillate through molecular sieves).
In distilling mixed solvents it is generally necessary to consult azeotrope data
books6 for mixture compositions. Chemists generally prefer to carry out reactions in
relatively low-boiling low-cost solvents. Most of the common solvents are listed in
Table 1, with a few notes on issues.
Vacuum distillation of solvents is frequently undertaken to reduce the temperature
of the solution, thereby reducing the risk of decomposition. Vacuum distillation
also speeds the removal of solvents. In a pilot plant, setting vacuum is usually
generated through a great variety of sealed mechanical pumps, or, simply, using steam
6Horsley, L. H. Azeotropic Data—III, Advances in Chemistry Series 116, American Chemical Society,
Washington, D.C., 1973.
TABLE 1. Common Reaction Solvents, Costs,a Boiling Points, Flash Points and Comment
Solvent Cost ($/unit)a B.Pt. (?C) Flash Point (?C) Density Comment
Methylene chloride 1.65/kg 40 — 1.325 Suspected carcinogen-declining use
Acetone 1.10/kg 56 ?17 0.79
Methanol 0.37/kg 65 11 0.791
Isopropanol 1.50/kg 83 12 0.785
Tetrahydrofuran 4.03/kg 67 ?17 0.887
Ethyl acetate 1.54/kg 77 ?2 0.90
Isopropyl acetate 2.66/kg 89 4 0.87 Less water-soluble than ethyl acetate
Dimethyl carbonate 2.78/kg 90 18 1.069 Underutilized, but note m.p. = 4?C
Diethyl carbonate 2.95/kg 128 31 0.975
Propylene carbonate 2.05/kg 239 123 1.205
Tert-butyl methyl ether 2.39/kg 56 ?25 0.742 Increased cost due to phase out in gasoline
Heptane 1.75/kg 99 ?4 0.684 Concerns re static charges in working with plastic
Toluene 0.61/liter 111 4 0.867 A perennial favorite
Chlorobenzene 1.54/kg 133 29 1.107
Xylenes 0.60/liter 137–144 29 0.860 Boiling Points: o-145?C, m-139?C, p-138?C
Water Virtually free 100 — 1.0 The preferred solvent
Acetonitrile 2.31/kg 81 5 0.786
Methyl isobutyl ketone 2.75/kg 118 13 0.80
Methyl ethyl ketone 2.15/kg 103 ?6 0.805
DMF 0.93/kg 153 57 0.946
DMA 2.04/liter 166 70 0.938
DMSO 2.85/liter 189 95 1.102
All are water-soluble.
Recovery from aqueous
solutions is fairly expensive
aCommercial costs for bulk quantities. Data from Dr. P. Savle, Schering–Plough, 1Q ’07, and Chemical Market Reporter.
Dilute solution
delivered to wall
Jacket heated
with steam or
hot oil
Agitator spreads solution
against wall of evaporator
FIGURE 5. Thin-film evaporators.
ejectors. The use of vacuum distillation always raises environmental concerns: solvent
contamination of the oil in a vacuum pump or of water in a steam ejector system.
Generally, in small-scale pilot plant operations, contaminated oil, distilled waste
solvent, and aqueous solvent wastes are sent to outside specialists for incineration:
Non-halogen-containing solventwastes are segregated from halogenatedwastes,with
the latter being incinerated in furnaces constructed to withstand hydrogen halide
Another vacuum distillation technique, including for water removal in large-scale
operations, is evaporation. Thin-film evaporators are designed to expose large surface
areas of liquid to heat and/or vacuum in order to speed the evaporation process. The
design principle is illustrated in Figure 5.
When a chemical process becomes a manufacturing proposition, chemical engineering
data on solvent recovery may well be needed in order to make the capital
investments necessary to achieve cost reduction and meet environmental requirements.
Frequently, however, one arrives at the manufacturing stage with minimal
solvent recovery data, necessitating the continued incineration of solvents using environmentally
approved waste disposal protocols. In those cases where relatively
small quantities of waste solvent are being generated, incineration often remains the
most economic option. In other situations, efforts are made to reuse solvents stripped
from processes; however, the main justification for cost saving and waste reduction
is negated if contaminants in the distilled solvent have an adverse effect on the yield
or quality of the desired product.
In creating a manufacturing process for a given site, solvents are often changed to
harmonize with the solvent recovery capabilities of that site. This process is not always
straightforward when contaminants from various process streams compromise
the analytical specification of the solvent required for a given NDA process. Under
such circumstances, contaminated solvents often have to be qualified to meet FDA
requirements, with the extent of qualification usually depending on whether the solvent
is used for production of an early intermediate or for an API. In a manufacturing
setting, where large volumes of waste solvent are generated, solvent recovery to meet
the desired specification is often contracted out to third parties.
Distillation is one of those unit operations that assumes its importance based on
the scale of operation. Clearly, distillation is one of the most important operations in
a petroleum manufacturing plant where separations of closely boiling liquids require
sophisticated distillation column design, built upon detailed theoretical plate determinations,
material of construction issues, and secondary distillation requirements.
In the pharmaceutical industry, the recovery and recycle of a solvent (versus incineration)
depends on process scale. Further insight into the problems that might be faced
in recovering solvents (and reagents) is provided in the section entitled “Dilevalol
Hydrochloride: Development of a Commercial Process.”
In conclusion, it is worth digressing to describe another widely used technology
that can be used as an alternative to some distillations, namely the concentration of a
chemical, particularly in aqueous solution, by reverse osmosis (RO). This technology
depends on reversing the normal osmosis process, wherein water, on one side of
a semipermeable membrane, will flow through the membrane to dilute an aqueous
solution of such as a salt on the other side of the semipermeable membrane. Reversal,
in commerce (e.g., obtaining potable water from brackish water), is achievable by
applying pressure generally of many hundreds of psi to the salt solution, so that water
will flow back through the membrane, concentrating the salt solution. RO is widely
practiced in the chemical, pharmaceutical, and wastewater treatment industries. The
technology of producing strong semipermeable membranes and their supports is a
critical requirement, especially in cases where membrane materials may be compromised
by the presence of organic solvents. In some endeavors (e.g., RO of sea water
to produce potable water), energy costs are a major operating consideration.
Crystallization is one of the most important unit operations in the manufacture of
APIs and to a lesser extent in the manufacture of intermediates for APIs (since later
process steps often provide opportunities for purging impurities).
An API is closely controlled in terms of crystal form, polymorph identity, particle
size, impurity profile and content, solvent, and water levels. All of these “quality
parameters” are defined in creating a drug product that has the desired pharmacological
properties (e.g., tablet dissolution rate to give needed blood levels) and desired
physical properties (e.g., stability and compatibility with drug delivery systems).
Solid API intermediates are crystallized, and thereby purified, to meet specified
quality standards. The specificationmay cover nomore than appearance, identity, and
purity parameters; these are often set by finding the right solvent that consistently
provides the needed crystal qualities or, in some instances, an amorphous solid
product. In contrast, the crystallization of APIs themselves is usually the subject
of intensive research efforts in collaboration with the pharmaceutical scientists to
provide a desired crystalline (rarely amorphous) API that will consistently allow
production of a dosage form meeting the criteria needed for the marketplace. Every
API has its own specific requirements, all of which need chemical engineering input
to one degree or another. Sometimes the attention to detail can be very demanding.
One case in the author’s experience concerned a process for the crystallization of an
API·hydrochloride (API·HCl) that required major reengineering after production had
Crystallization of an API·HCl. Chemical engineers in process research and development
operations gain specialist expertise in a great variety of process problems such
that they are often called upon to solve production problems. Areas of involvement
include increasing the productivity of a plant to delay the need for further equipment
investment, reducing manufacturing costs by reducing labor requirements and
helping resolve regulatory issues.
When production problems persist for a long time, especially if an FDA inspection
references the chronic aspects of a problem, a sense of urgency is created which
encourages the involvement of chemical process R&D engineers. One such event occurred
when the hydrochloride salt of an active pharmaceutical ingredient (API·HCl),
being produced by a third party manufacturer, proved difficult to dry to the solvent
level needed (<0.1% ethyl acetate) to avoid the development of an ethyl acetate odor
in tablets of the API·HCl.
During the period of developing the process for the manufacture of the API·HCl,
the ethyl acetate level was set at <0.2%. This level was filed in the New Drug Application
(NDA). However, over the course of time, it became apparent that the tablets
made from API·HCl containing <0.2% ethyl acetate developed an odor of ethyl
acetate—an issue manifested in customer complaints (as an aside, the FDA generally
targets customer complaints as one of its focus areas in initiating a production plant
As a result of these findings, the ethyl acetate specification was reduced to the
<0.1% figure based on an investigation of ethyl acetate levels in the API·HCl versus
odor development in the product tablets; in practice the API·HCl was dried to a target
specification of <0.08% ethyl acetate to provide a comfort level. However, since the
API·HCl tenaciously held onto ethyl acetate, the drying times to meet the new target
varied from 100 hr to as much, on rare occasions, as 300 hr even under extreme
drying conditions—110?C under vacuum! The FDA regarded the process as out of
Fortunately, in our particular case, the API·HCl proved sufficiently stable, even
under the extreme temperature condition, to permit bulk drug production within the
desired specification. Nevertheless, the extension of the drying time from ?24 hr,
to get down to an ethyl acetate level of <0.2%, to the 100–300 hr needed to reach
the 0.08% figure proved to be quite unacceptable. A rough comparison of API·HCl
particle size versus drying time to meet the <0.08% ethyl acetate specification revealed
that the routine API·HCl production material contained ?80% of crystals >
200 µ.Micronizing the API·HCl to reduce the particle size greatly reduced the drying
times, but this approach introduced another step, adding unwanted GMP qualification
requirements and, even more important, creating safety and industrial hygiene
concerns associated with dusts and worker exposure.
As a result of the above, we defined the objective as improving the process while
staying within the process filed in the NDA. This essentially defined our objective as
one of finding crystallization conditions that would give a smaller crystal of the same
polymorph as that described in the NDA.
The NDA process for preparing the API·HCl comprised addition of 100 mol.%
of 35% hydrochloric acid to a stirred solution of the acetate salt of the API in ethyl
acetate (65%), water (35%), and methanol (5%) at a temperature of 24–28?C. This
process produced crystals of variable size, with the said >80% being over 200 µ.
A solution to the problem was created as a result of the initiatives of two chemical
engineers, RayWerner and Lydia Peer. Their crystallization process studies led them
to the following variant of the NDA process:
Addition of 50 mol.% of the 35% hydrochloric acid at 24–28?C to the stirred
solution of the acetate salt of the API in the same composition of ethyl acetate, water,
and methanol as in the NDA process gave a solution of the mixed salts. Interestingly,
most of the heat evolution in formation of the API·HCl occurred during this first half
addition. More important, the resulting solution could be cooled virtually to ?10?C
without crystallization of API·HCl. In short, the mixture of salts was far more soluble
than API·HCl on its own. In practice the solution was cooled to ?5?C to ensure that
ice crystals did not crash out before the final addition of the remaining 50 mol.% of
35% hydrochloric acid. In the laboratory the final 50 mol.% of 35% hydrochloric
acid was added over about 1 min, resulting in fine crystals of API·HCl which dried to
<0.08% ethyl acetate in approximately 12 hr or less. In the first plant trial, addition
of the hydrochloric acid was carried out over 45 min and still produced much smaller
crystals of API·HCl than those produced by the third-party manufacturer, despite the
addition time being far longer than the 1min used in the laboratory. The plant trialwas
based on holding the temperature at ?5?C rather than adding the 35% hydrochloric
acid as rapidly as possible. Although the plant trial produced beautiful, perfectly
formed fine crystals, they, disappointingly, dried to the <0.08% ethyl acetate level at
the lower end of the time range (i.e., approximately 100 hr) of the API·HCl prepared
by the third party manufacturer. Microscopic comparison of the two lots of crystals
revealed that those formed over 45 min were uniformly dense, whereas those formed
in 1 min were much more fractured. Further physical comparison showed that both
lots of crystals were of the same polymorphic form and identical in most other ways,
principally in melting point, stability, and overall particle size range. When the plant
trial was repeated using a 1-min addition, the fine fractured crystals produced were
identical to those formed in the laboratory. The filtration and washing of both types
of crystals were similar and close to times observed with API·HCl prepared by the
NDA Process
100x Magnification
NDA Modified Process
100x Magnification
FIGURE 6. Comparison of photomicrographs of API·HCl prepared using the NDA process
and the modified NDA process.
0 20 40 60
Time (minutes)
Amount Dissolved (%)
NDA Process
Modified NDA
FIGURE 7. Comparison of rate of dissolution of API·HCl prepared by the NDA process and
the modified NDA process.
third-party manufacturer. The photomicrographs in Figure 6 illustrate the enormous
differences in the appearance of the crystals.
The range of drying times for various plant lots of fractured crystals, to meet the
<0.08% ethyl acetate specification, was 8–24 hr. The heat generated in the process
using a 1-min addition of the last half of the 35% hydrochloric acid did not prove
problematic, being held in the range ?5 to +15?C under full jacket cooling using
coolant at a temperature of approximately ?20?C. The particle size range of the
desired fractured crystals was generally 80% <150 µ. The process for preparing
API·HCl by the modified NDA process was not without its detractors since one
difference found versus the API·HCl prepared by the NDA process was in the relative
rates of their dissolution (Figure 7).
Roughly 85% of the API·HCl prepared using the modified NDA process was
dissolved in 20 min versus 45 min for the API·HCl from the NDA process. Although
the drug is administered on a twice-a-day regimen, the concern was raised that
there may be an earlier spike in the blood level of the API·HCl from the modified
NDA process. A medical review of the data led to the conclusion that in the 12-hr
time frame between API·HCl administrations the 25-min difference in reaching the
specified 85% dissolved material was of little consequence. It was also observed
that by the use of a higher pressure in tablet production, a tablet dissolution rate
conforming to tablets of marketed product could be created.
The modified NDA process was adopted.
Although the above case centers on the creation of a desired crystalline product
that dries well, similar permutations of solvent, solvent mixtures, salt selection, and
crystallization conditions (temperature, concentration, pH, seeding, stirring, rates of
addition, etc.) are generally applicable in creating crystals that filter and wash well.
The chemical engineer is a vital ally of the chemist in such work.
Filtration, Washing, and Drying
These unit operations are closely integrated in plant practice. The chemist’s Buchi
filter, with choice of the best filter paper, manipulation of the wet cake on the filter
with a spatula (to improve washing), and smoothing to eliminate cracking of the
cake, is a far cry from the situation faced by the chemical engineer dealing with
large volumes of crystal slurry, or a hot solution of a product being carbon treated,
say before crystallization. In the latter case, the carbon is removed in a closed filter
system—for example, a Sparkler filter at a temperature above the crystallization
temperature of the product. Not infrequently, when carbon peptization has occurred,
it is necessary to add a filter aid to prevent clogging of the closed filter and to ensure
that fine-particle carbon is held by the filter aid – many APIs are subjected to a
filter test in concentrated solution, a gray filter paper usually signaling that a batch is
carbon contaminated and must be reworked.
In engineering practice at the pilot plant stage, the chemical engineer employs
a great variety of Buchi-type filters. Whereas open large Buchi filters are still used
in pilot plant practice (with “elephant trunking” exhaust to remove fumes and with
worker protection as needed, all the way up to “breathing air” suits), the equivalents
favored by chemical engineers are closed filters of various types.
The simplest are closed filters of the Sparkler or Aurora type. The most versatile is
theRosenmund type, which can undertake increasingly sophisticated functions:These
range from stirring the filter cake in a slurry mode, to building in a drying operation
via hot nitrogen passed upwards through the filter plate, and on to offloading through
side ports into containers, with the whole system being virtually closed to minimize
bacterial contamination and worker exposure. Sketches indicating the design features
of the major Buchi-type closed filters are provided in Figure 8.
Most of these filters, especially the closed ones, are relatively easily cleaned,
usually by flooding the filter with an appropriate hot solvent.
Centrifugation is one of the most widely employed types of filtration, notable
for operational speed (assuming the solid packs well on the filter cloth, or sintered
metal bowl, and the porosity of the solid cake is maintained to allow thorough washing)
and for producing a relatively dry cake—it is not uncommon for a centrifuged
filter cake to be “dried out” to a solvent content of approximately 20%. Sometimes
Feed N2
Sight Glass
Agitator which
rotates and
up and down
Solids discharge with agititator
Solids discharge manually
after filtration
Sight Glass
GlovePorts Discharge
Solids discharge manually
using glove ports
Sight Glass
FIGURE 8. Buchi-type closed filters.
centrifugation causes dense packing of a solid product, making washing slow and
inefficient; occasionally this can be overcome by loading the original crystal slurry
on to the centrifuge at relatively low revolutions/minute, or if the solid is an unwanted
product of a process (e.g., a fermentation broth mycelium) by mixing the slurry with a
filter aid. In some cases, unwanted products that do not filter (e.g., muddy materials)
can be removed by centrifugation in a solid bowl centrifuge, wherein the mud is
retained in the centrifuge bowl and the needed liquid flows out over the edges.
One of the major considerations in operating a centrifuge for the filtration of crystal
slurries in flammable organic solvents is to eliminate the possibility of explosion by
using safeguards such as a nitrogen blanket to provide oxygen levels belowthe critical
level. Oxygen sensors are usually installed that switch off the centrifuge if prescribed
oxygen levels are exceeded.
The unloading of a centrifuge once the spin-dry cycle has been completed is best
accomplished without exposing process operators to harmful fumes, while at the same
time minimizing the biological contamination of the product. In this regard, bottomoffloading
centrifuges minimize product handling by process operators, but do not
Top Unloading
Cake is manually dug out
through top hatch
Top Hatch
into cake
Bottom Drop
Cake is discharged through slots in
the bottom of the centrifuge bowl
Bowl Slotted
at bottom to allow for
product discharge
Liquor Mother
Knife cuts into cake and “peels”
solid out through discharge chute
Peeler Moves
into cake
Solid Bowl
Top Hatch
FIGURE 9. Types of conventional centrifuge.
overcome the handling needed in transferring the product to a conventional dryer.
Sketches indicating the design features of conventional centrifuges are provided in
Figure 9.
A valuable variant of the conventional centrifuge is the horizontal spindle centrifuge,
especially the Heinkel type. This centrifuge has a unique offloading feature
in that the filter cloth loaded with washed filter cake is pulled out to allow a reversed
centrifugal removal of the solid as illustrated in Figure 10.
The most advantageous features of the Heinkel-type centrifuge are the ability to
handle difficult-to-filter materials by building up and washing thin cakes and offloading
them rapidly. Thus, the Heinkel-type centrifuge is generally more versatile and
can be faster in operation than the conventional type. As is the case with conventional
centrifuges, the windage release of solvent vapors requires that solvent capture
systems (e.g., carbon beds or air scrubbing) be in place.
A few other kinds of filters are used in manufacturing plants (Figure 11), particularly
for intermediates. One is the plate and frame filter press, which is relatively
low in cost but labor-intensive. Another is the vacuum rotary filter wherein the filter
cake rotates through a washing segment, as well as through a “pulling-dry” segment
During Loading/ Washing
Solids are stopped by the cloth
Mother liquors pass through cloth
Solids are represented by “S”
Filter Cloth
Filter Cloth
During Unloading
The Filter inverts causing the
solid to chute off the filter onto
the housing and out the
discharge chute
Discharge Chute
FIGURE 10. Operation of a Heinkel-type centrifuge.
Knife for
Rotating Vacuum Drum Filter
Filtrate outlet
Cake Discharged during
plate disassembly
Filter Press
Filter Plates
& Wash
Filter Plate
Product Slurry
& Wash Inlet
FIGURE 11. Other filters used in commerce.
prior to being knifed-off or blown-off. These two filters are somewhat tedious to
operate and environmentally problematic, though enclosed versions of rotary filters
are marketed.
A very large assortment of ovens and dryers is in use in chemical and pharmaceutical
plants. Simple forced-air-heated tray dryers are increasingly required to have
solvent capture attachments; old permits allowed specific levels of solvent in the emitted
air and a finite emission rate and time for each individual drying operation. Such
environmental strictures often persuade the chemical engineer to use a centrifuge for
filtration in order to minimize the amount of solvent needing removal in the drying
Sight Glass
Agitator which
rotates and
up and down
Buchi-type Drier (e.g., Rosenmund)
Rotating Agitator
Horizontal Paddle Drier
Conical Drier Spherical Drier
M Charge
Rotating and Revolving
Rotating Agitator
FIGURE 12. Popular vacuum dryers with a mechanical “paddle.”
step. However, as mentioned, solvent capture from windage emissions is also needed
Vacuum drying is the most popular drying option, and dryers of many different
types equipped with appropriate emissions capture devices have been designed to
suit all manner of requirements. One of the most versatile is the vacuum tray drier,
which can accommodate solids that dry at variable rates. Greater sophistication, not
infrequently leading to a loss of versatility, has been introduced with the addition of
paddles to mechanically move the solid during the drying period, accelerating the
rate of drying. In many cases the paddles also function to break up lumps; however,
the reverse can also occur, leading to balling and/or lining the dryer wall with a
veneer of solid—obviously undesirable mechanical and insulating situations. Dryers
with various types of mechanical stirring devices are available, notably in Buchitype,
horizontal, or conical paddle configurations, the latter with couplings allowing
the screw flights to turn and sweep round the wall of the drier at the same time
(Figure 12).
Spherical dryers have recently become more popular in the drying of APIs, one
reason being that hold-up of API in the discharge process is less than that with other
types of drier.
Fluid bed dryers (wherein hot air or a hot gas is driven upwards through a bed of
wet crystals to fluidize the mass, with fine material being caught by filter socks) are
particularly useful for drying water wet filter cakes. Spray dryers are also normally
employed for the removal of water, though in this case the rapid drying process
(from a spray of an aqueous solution falling in a countercurrent or co-current flow
of hot air or gas) often leads to an amorphous material. Similarly, lyophilization
of an aqueous solution under high vacuum at low temperatures (freeze drying) is
often employed for the production of drug substances, again often in an amorphous
There are environmental (and sometimes safety) concerns with evaporative techniques
of the above types whenever the solid to be dried, or the solution to be spray
dried or lyophilized, contains volatile organic solvents. Of the above three drying
techniques, spray drying is probably the most widely used, especially in recovering
water-soluble drug substances from water solution (e.g., the aminoglycosides).
A variety of dryers is often available in model pilot plants since not every product
will drywithout causing problems of the balling or veneering type, or in the case of the
conical dryer without clogging at the bottom outlet. Test drying in a variety of dryers
is often needed to determine the optimum dryer and optimum drying conditions.
Drying tests often lead to the reinvolvement of the chemist in efforts to change the
crystallization to improve the handling properties of the crystal.
Frequently, filtration, washing, and drying operations are integrated especially
where noxious substances are being handled or when a crystal slurry of an API is
being processed to a dry solid in a controlled environment room (CER). In the latter
case, all types of combinations of filters and dryers are used (Figure 13).
The operation of CERs is governed by validated SOPs covering topics such as
cleaning and microbiological testing, ingress and egress protocols (for both people
and materials), air quality and air pressure differentials and equipment calibration,
and maintenance, inter alia. The entire operation is covered by a comprehensive
documentation program.
Pilot plants generally produce APIs using controlled environment rooms of the
above type. Manufacturing units also often operate using this basic type of controlled
environment room. However, when specific large-scale operations are designed, it
is often desirable to minimize operator involvement by integrating filtration, washing,
drying, and product offloading steps into a single closed unit. In short, the
enclosed plant is the clean room. This strategy of containment has many advantages,
1. Elimination/minimization of exposure of both product and operators to contamination
2. Enhancement of process safety
3. Minimization of solvent and product releases into the environment
Filter Unit* Drier Unit#
HEPA Filtered Air
Air Pressure vs. Outside is ++
Product &
Air Pressure
is +
Air Pressure
is +
* Favored Units are of the Rosenmund or Heinkel type
# Favored Vacuum Drier units are of the closed type such as
Rosenmund, the Conical drier, the vacuum tray drier, or the
vacuum paddle drier type
API Crystal
FIGURE 13. Illustration of major features of a controlled environment room. Controlled
environmental rooms (CERs) are operated by gowned, specially trained operators working
to standards/guidelines identified as appropriate for producing APIs meeting specific quality
criteria including microbiological and particulate requirements for each specific drug form or
use—parenteral, inhaler, oral, or topical (in practice the standards applied routinely may be
the more exacting requirements).
4. Closed system recovery of solvents
5. Process automation
6. Enhanced GMP compliance
There are some downsides too:
1. Containment systems are often inflexible, thereby making them unattractive
for multiproduct use.
2. When two ormore products are produced in a contained system, the turnaround
time from one product to another is generally very long, owing to the need to
thoroughly clean the entire equipment train and to validate that it is clean.
3. Long, large-scale product cycles are usually needed to justify the investment.
4. Purpose-built containment units are generally best for single-product operations,
though industry has overcome the “one-product” downside in many
The use of a Kraus–Maffei type of containment unit is described in the case study
entitled “DilevalolHydrochloride: Development of aCommercial Process” (q.v.). The
Containment Technology Containment Technology
Containment Technology Containment Technology
Dryer Dry Product
Glove Box
Glove Ports
Dryer Dry Product
Glove Box
Dryer Dry Product
Glove Box
Clamp Ring
Swing Port
FIGURE 14. Sequence of steps in the contained off-loading of a drier.
operation of the Krauss–Maffei Titus System, diagrammatically outlined in Figure 3
of this case study (Chapter 10), is basically as follows:
The process for producing the crystals of product to be filtered is carried out in a
closed system in such a way as to virtually eliminate any microbiological or particulate
contamination of the crystals. The crystal slurry is filtered and washed using the closed
centrifuge. The filter cake is knifed off and picked up by circulating hot nitrogen gas
and carried to the conical drier for the final drying operation. The hot nitrogen passes
through the filter socks to a condenser where solvent is condensed for recovery and
recycle. The nitrogen is then blown through the heater for repetition of the cycle. For
complete containment of the product, wherein the plant equipment itself becomes the
controlled environment room, the offloading of the drier is carried out through a glove
box into a plastic container (bags stored inside the glove box) within a drum as depicted
in Figure 14.
Other ingenious efforts have been made by chemical engineers to integrate crystallization
with filtration, washing, and drying operations. One such unit built by
Rosenmund AG in Switzerland is their “Nutrex” unit. This unit is designed to carry
out a reaction—namely a crystallization—and, by inverting the unit on gimbals and
Dryer Dry Product
Glove Box
Containment Technology
Dryer Dry Product
Glove Box
Containment Technology
Dryer Dry Product
Glove Box
Containment Technology
Dryer Dry Product
Glove Box
Containment Technology
Dryer Dry Product
Glove Box
Containment Technology
Dryer Dry Product
Glove Box
Containment Technology
FIGURE 14. (Continued)
lowering the stirrer, to also undertake filtration, washing, drying, and offloading steps.
A sketch illustrating the cycles is provided in Figure 15.
Although the Nutrex unit is a testimonial to the creativity of chemical engineers, it
is obvious that inflexibility increases as more potentially problematic unit operations
are integrated. In addition, the mechanical and operating requirements (e.g., piping)
for large-scale production do not lend themselves well to the operating principles of
the Nutrex unit.
Reaction & Crystallisation Filtration & Drying Offloading
Filter Plate
FIGURE 15. Operating cycles in the use of a Rosenmund “Nutrex” unit.
Producing a Product of Small Particle Size (Milling, Micronization,
and Precipitation)
Reducing the particle size of a material for heterogeneous reactions or for preparing
a drug product from a very insoluble API is an important area of endeavor.
Delumpers and hammer mills are frequently used to reduce the particle size
of products, thereby creating a manageable physical form for ongoing processing,
usually improving rates of dissolution or an even distribution of a product when
blended in a mixture.
On occasion, it is necessary to produce a very fine particle size in order to effect
a reaction—for example, reactions where a reagent is insoluble in the medium used
for a reaction. In this case, a very small particle size (e.g., 5–10 µ) can enhance
chemical reactivity by exposing a very large surface area of an insoluble chemical to
an appropriate reactant. One example of this is the preparation of a so-called sodium
dispersion, which has been used in a variety of commercial processes (Scheme 3).
Sodium dispersion is prepared by melting sodium in toluene at just over 100?C
in a reactor equipped with a high-speed stirrer engineered to maximize shear of the
liquid sodium. On cooling, stable particle sizes of 5–10 µ can be achieved. The stirrer
design may be any of the types illustrated in Figure 16.
Micronization in the pharmaceutical industry is most often associated with producing
an API, generally a water-insoluble compound, of very small particle size to
enhance its absorption into the bloodstream. The design principles of the commonly
used micronizers are illustrated by reference to Figure 17.
Chemical engineers are the principal managers of micronization studies, as well
as of other technologies used in producing a fine particle size product. The following
case study describes the work involved in producing an insoluble API in a fine
particle form suitable for dosage form preparation. The case specifically illustrates
the importance of observation and also looking beyond traditional micronization for
methods to produce fine particle size products.
A Micronization Study. The particle size of an API is often a key parameter to
control to ensure consistency in bioavailability of the API in the drug product. For
this reason, micronization of APIs has become a standard operating procedure, at
Mg stearate
Dispersator Sodium
SCHEME 3. Uses of sodium dispersion in commerce.
FIGURE 16. Stirrer designs for the production of sodium dispersion.
the research stage, in a number of pharmaceutical companies, especially for waterinsoluble
molecules. The particle size of an inhaled steroid, for example, is carefully
controlled to meet the exacting consistency standards set by a company and approved
by the Pulmonary Division of the FDA.
Cyclone or Filter
to Collect Product
Top View
FIGURE 17. Design feature of commonly used micronizers.
Setting a particle size specification for an API is seldom straightforward. Many
factors have to be considered, such as the flow properties needed for tableting,
capsule filling, and filling inhalation devices, the tendency of the API to pick up
a static charge, the achievement of desired rates of dissolution to meet blood level
requirements in the body, the availability of equipment on the manufacturing sites,
and even the stability of the API and any propensity to change polymorphic form
during micronization. Although many APIs can be micronized without difficulty, it
cannot be predicted that all of them will. For example, one needs to know whether a
commercially viable API feed rate to a micronizer can be achieved, and whether the
crystal of the API will clog the feed device or, after micronization, the micronizer
In overcoming problems in the crystallization and micronization of a waterinsoluble
antifungal compound, we first tried to meet desired particle size speci-
fications by conventional means such as adding a solution of the compound in an
acceptable GRAS (Generally Regarded as Safe) water-soluble solvent into water under
conditions that might lead to a fine particle size, in an acceptable crystal form
and with acceptable filtration and drying characteristics. Unfortunately, this work did
not initially lead to desired particle size requirements, although the product filtered
well. Drying of the product in an INOX-GLATT dryer with wet milling capabilities
led to caking. Drying in a regular vacuum tray dryer followed by delumping gave a
uniform product suitable for use as the feed to a micronizer, but not fine enough for
use in a dosage form.
Micronization of our antifungal using the favored Jet pulverizer micronizer
(Figure 17) led to the desired particle size product; however, the micronizer Venturi
feed mechanism rapidly (1–3 min) became clogged by a ceramic-like coating on the
0 1 2 3 4 5 6 7
Venturi Clogging Time (Min)
Water & Solvent Content (LOD) in %
FIGURE 18. Relationship between Venturi clogging time and Water/Solvent content of an
antifungal API.
wall of the Venturi nozzle, just beyond the Venturi outlet. This necessitated frequent
dismantling of the micronizer for cleaning, severely hampering the drug delivery
program and necessitating rework of the recovered material; this also created an
analytical nightmare. A different feed device (obtained from a Hosokawa mill) gave
no better results, nor did we find that efforts to embrittle the particles by pre-cooling
them in a dry-ice box were any more successful. These failures led us to initiate an
urgent program of work to find a way of producing product of an acceptable particle
size. The chemical engineers in charge of the project (Noel Dinan and Steven Yu)
examined the pilot plant data generated in every phase of the project. Noel Dinan
observed that the caking occurring in the INOX-GLATT dryer did not happen until
quite late in the drying cycle. He went on to show that the transition from a free-
flowing crystal to the caking stage occurred when the water level in the solid dropped
below approximately 5%. This key observation led to the proposal that product of
water content >5–6% be used as micronizer feed. The engineers’ suggestion was not
accepted at first for various reasons. Some pointed out that since the crystal used as
input for the micronization step came from a process in which a methanol solution
was added to water, it seemed likely that methanol levels in the final micronized API
would be unacceptable. Others raised questions on the crystal form of the product:
Would it be a hydrate or be in a different, perhaps even metastable, crystal form?
Analytical studies using differential scanning calorimetry and infrared showed that
no new polymorph or hydrate form was produced. Noel Dinan undertook a study
feeding material of various water contents into the micronizer (Figure 18). No clogging
of the Venturi inlet occurred when the water of the API was >7%. Moreover the
water and methanol levels were greatly reduced during the course of micronization
owing to the sweep of dry nitrogen through the micronizer. The study results are
summarized in Table 2.
TABLE 2. Micronization Studiesa with “Wet” API Using a Four-Inch Jet Pulverizer
(LODb), %
Particle Size Distribution
<2 µ
<7.5 µ
<30 µ
LOD (%)
Content %
H2O Content
2.0 39.0 91.0 99.9 0.2 0.01 0.3
3.75 48.8 96.9 100.0 0.21 0.01 0.25
9.8 58.2 99.1 100.0 0.22 <0.01 0.77!
12.8 65.2 98.9 100.0 0.22 0.01 0.17
aFeed rate 50 g/min. Injector pressure 120 psi. Milling pressure 100 psi. Venturi 316 SS.
bLOD, loss on drying.
Use of a smallermicronizer (3/4 in.) gave anomalous results.As seen from Table 2,
the water methanol and particle size targets set for the product were routinely met:
Target particle size: 98% ? 30 µ, 75% ? 7.5 µ, and 30% ? 2 µ
Target methanol content: 0.01%
Most important, the micronizer clogging problem was overcome. The success of the
project owed everything to the simple observation that in the drying step a marked
adverse change in the rheological properties of the API occurred when the water level
dropped below approximately 5%. In further engineering work to try to simplify the
process and to generate an optimal production process, Noel Dinan reinvestigated the
original (failed) precipitation concept by producing an acceptable particle size range
using a continuous crystallization scheme. A 10% solution of the API in methanol
(1 part) was fed simultaneously and equivalently with a stream containing 25%
methanol and 75% water (40 parts) into a mixing chamber and on into a stirred
crystallizer containing 2.5% of the batch weight in seeds to produce a very uniform
crystal: 100% <40 µ, 80% <7 µ.
The methanol level in the precipitation vessel (25%) was optimized in studies
of process conditions for the precipitation: When methanol levels below 25% were
used, gumming of the product was observed. The product of the above precipitation
process filtered and washed well, did not agglomerate during drying, and, physically,
appeared and behaved the same as micronized material. Such a continuous process
was operationally attractive despite the dilution, since it shortened the time cycle (a
large crystallizer was available) and avoided the labor intensive and dusty micronization
process. However, the micronization process using “wet” API was adopted since
it was deemed useable on all production sites.
In concluding this section on producing small particle size powders, I would like
once again to pay tribute to the chemical engineers’ ability to grasp the practical
significance of observations made regarding the properties of materials. Earlier I
passed over the idea of embrittling a solid as a technique for making it suitable for
milling. It is pertinent to recount one case where embrittling is used commercially on
a tens of thousands of tonnes/annum scale. TheMayekamaManufacturing Company,
Ltd., Tokyo, Japan, uses embrittling via a secondary cooling system to cool used tires
to the point where they can be milled easily to a powder for recycle. A two-stage
refrigeration system using ammonia on the high side and ethane on the low side
chills a hydrofluoroether, C4F9OCH3, which in turn cools air to ?87?C. This cold
air is passed over incoming tires, cooling them to the point that they shatter easily.
Selection of C4F9OCH3 was made because of its low-temperature properties and
because it is nonflammable, non-ozone-depleting, and low in toxicity. This cooling
system was found to be cheaper and actually friendlier to the environment than liquid
nitrogen systems used earlier.
It is worth describing other operations where differences between the chemist’s
laboratory and the chemical engineer’s pilot plant and plant create the need for
different approaches. Pumping, flow measurement, and reactor volume measurement
are a few of the more common operations deserving the chemist’s attention.
The chemist’s manual pouring techniques are necessarily replaced in large-scale
operation by more practical techniques. The preferred method of moving liquids
(solutions) during operation of a chemical process is by gravity through a multistorey
train of reactors. However, very frequently the lifting of liquids to elevated levels to
permit gravity movement is done using pumps. Pumps are of many types and are
designed to handle all types of flammable, noxious, viscous, and corrosive liquids,
as well as heavy slurries and gases. Delivery of the latter is more often handled by
weight from a cylinder of the compressed gas.
The major kinds of pumps used in the chemical and pharmaceutical industries are
centrifugal, positive displacement, and turbine pumps.
In a centrifugal pump, rotating flared blades in the pump housing essentially suck
in a liquid from a delivery pipe at the eye of the driveshaft, ejecting it outwards from
the blades into the shell of the housing carrying the outlet pipe. Centrifugal pumps
are probably the most widely used pump types and are available for capacity needs
as small as 2–3 gals/min and as large as 100,000 gals/min!
A positive displacement pump is equipped with both valve inlet and outlet pieces:
The inlet opens with the suction of liquid in the priming stroke of the piston, and
it closes as the outlet valve opens in the discharge stroke. The drive for movement
of the liquid may be a piston, a plunger, or a diaphragm, the latter being a flexible
material fabricated of rubber, plastic, or metal. Rotary pumps can be considered as
positive displacement pumps. Such pumps depend upon themechanical displacement
of liquid by rotation of an “impeller” within a stationary housing. In a gear pump the
impeller is a pair of rotating meshed gears that impound liquid from the inlet pipe in
the outer tooth gaps and carry it a round the periphery of the housing to the outlet pipe.
Although the capacity of positive displacement pumps is not as great as centrifugal
pumps, they are more efficient because internal losses are minimized. Far greater
pressures can be exerted on a liquid with a positive displacement pump, leading to
such pumps being widely used for pumping to high heads. Diaphragm pumps have
great appeal for handling hazardous and toxic liquids since their construction eliminates
exposure of seals and packing to hazardous and noxious liquids. Diaphragm
pumps are also popular for pumping slurries, especially where gentle handling of
suspended solids is required to minimize crystal degradation. The major downside
to the use of a diaphragm pump is the inevitability of failure of the membrane; such
pumps therefore require that a rigorous inspection and maintenance program is in
place. Peristaltic pumps, despite their weak point of potential failure of the flexible
hose, have found some favor where nonaggressive liquids, or slurries in such liquids,
are being pumped. Pumping rates of 350 gals/min at over 2000 psi have been routinely
achieved in a 24-hr-a-day operation. Again a rigorous inspection/maintenance
policy needs to be in place to ensure that flexible hoses are changed before
they fail.
Turbine pumps mix features of a simple propeller (axial flow) pump with a centrifugal
pump and are often referred to as units with mixed flow. A simple turbine
pump carries curved vanes on a central rotating spindle. Such pumps are often immersed
in the liquid and find use in closed-loop circulation systems, in condenser
circulating water, and in sumps and wells. Turbine pumps have noteworthy pumping
capacity, and like positive displacements pumps are often used for heads up to about
100 ft/stage with capacities of up to several hundred gallons/minute.
Flow Measurement
The chemist’s “eyeballing” a graduated dropping funnel to deliver reactants at desired
feed rates into a glass reactor, or setting up precise delivery using a validated peristaltic
pump, is generally replaced by quite different flow measurement systems in the pilot
plant and plant worlds.
Pumps, such as piston pumps, can meter liquids into a reactor fairly precisely,
but the chemical engineer uses a flow measurement device for greater precision.
The most commonly used flowmeters are rotameters that are calibrated to translate
the lifting of a float in a vertical slightly tapered tube (small diameter at the inlet of the
flowmeter) into a measure of the amount of liquid delivered in a given timeframe. For
greatest precision the rotameter is calibrated with the specific fluid being metered.
Most modern rotameters are provided with a calibration plot that corresponds to
There are many other flow measurement devices including Orifice/Venturi meters,
turbine meters, and more sophisticated instruments using ultrasonic, magnetic, and
Coriolis effect techniques. Orifice/Venturi type meters have a restriction causing a
pressure drop related to the flow rate of liquid. Such meters are popular because of
their low cost; however, their accuracy can be compromised by upstream elbows and
valves. Turbine meters are designed so that rotation speed varies linearly with the
Fluid Force
Fluid Force
FIGURE 19. Mass flowmeters.
flow rate. Magnetic pick-up of turbine blade rotation is translated into voltage pulses,
which in turn are converted into a measure of flow rate. Again, such meters operate
best in situations where flow is not restricted.
Ultrasonic flowmeters depend upon measuring time delays in received sound
waves from a pair of opposing transducers, one set downstream from the other in
a pipe carrying liquid. The transducers measure the difference between the velocity
at which sound travels with the flow of liquid and against the flow of liquid, the
signals being translated into a measure of liquid flow. Another ultrasound flowmeter
for liquids containing bubbles or particulates utilizes the Doppler effect—that is, the
change in frequency of a returned sound wave bounced from a particle or bubble
moving away. Ultrasonic flowmeters, like magnetic flowmeters and Coriolis mass
flowmeters, have no moving parts and can be used for measuring flow within a
broad range of viscosities and temperatures. Magnetic flowmeters sense the voltage
induced when a conductive fluid flows through a magnetic field. The induced voltage,
proportional to the flowvelocity, is fed to ameasuring amplifier by a pair of electrodes.
Coriolis mass flowmeters (Figure 19) depend on creating and measuring Coriolis
acceleration in a flow loop.7 This is done by passing flowing liquid through a flow
loop that is being vibrated, usually with an amplitude of approximately 2 mm at a
frequency of approximately 80 cycles per second.
The inflowing liquid exerts a small opposite directional force on the upward and
downward vibrations of the tube. Going in at an upward vibration, the liquid assumes
the force of the up motion halfway around the loop. This upward flow force exerts
the opposite effect on the down cycle of the vibration the net effect, causing the tube
to twist. Electromagnetic velocity detectors located on each side of the flow tube
measure the velocity of the tube. The twist results in a time difference between the
tube velocity signals; this time difference is directly proportional to mass flow. From
once being regarded as esoteric and expensive, Coriolis flowmeters are now in the
mainstream, largely because of their accuracy of mass flow measurement and the
7The discovery of the Coriolis effect resulted from work done by Gaspar Gustav de Coriolis (1792–1843)
at the behest of Napoleon, who wanted to know why his cannon balls never went straight. There is no
historical record to indicate that Coriolis cast any light on Napoleon’s problem but, as often happens in
research, unexpected findings can stimulate curiosity and lead to other useful outcomes—as Pasteur once
said, “Chance only visits the prepared mind.”
availability of lower priced meters which deliver almost the same performance level
as higher priced units.
In modern plants it is important to add that reliance on pumping liquids through
flow measurement devices to deliver precise weights or volumes is often replaced by
load cells that weigh the process vessel and its contents and accurately record weights
of incoming reactants and solvents.
In a point of validation it is worth saying that all weight or volume measurement
devices need to be well maintained, calibrated, proven, and regularly audited to ensure
that the device is delivering what the operations people running a process batch sheet
say it is delivering.
Reactor Volume Measurement
The chemist operating a laboratory experiment in glassware relies on the use of
balances for weighing reagents and graduated cylinders or dropping funnels for
measuring volumes. Although graduated glass vessels are used to control the addition
of liquid reagents and solutions to a reactor, the chemical engineer, operating a pilot
plant reactor in which volumes cannot be checked by eye, works to ensure that a pilot
plant standard operating procedure (SOP) is being carried out as well as possible by
double-checking reaction volumes.
In past times the dipstick, often made of wood, was widely used to make a volume
measurement check even in a calibrated vessel. In reference, perhaps reverence, to
the wooden dipstick it is pertinent to mention that the Ford Motor Company, in
celebrating its centennial in 2003, built six all-new Model T’s, each being equipped
(as was the original) with a ruler for checking the fuel level in the tank! (New York
Times, May 18, 2001, page F1). Today, opening a reactor for liquid level checks, or
for sampling the contents, is considered to be undesirable. (In regard to sampling,
contained devices can be fitted to draw reaction samples for the chemist’s process
monitoring work.) In measuring reaction volumes, reactor calibration is often used.
There are also a great number of both simple and sophisticated level measurement
A simple device is based on the pressure required to slowly bubble a gas (nitrogen
or air) through a dip tube to the bottom of a vented reactor the pressure differential
being translated into a level measurement. At the other extreme, the application of
ultrasonic and radar technologies has led to much more sophisticated devices.
Ultrasonic level measurement depends on a transducer sending an ultrasonic pulse
to the liquid surface which is reflected back to the transducer. Electronics convert
the ultrasonic lag time into a distance (D) corresponding to depth. Ultrasonic level
measurement is based on the equation
D = V ? t/2
where D is a function of the round-trip time (t) required for an electronic pulse to
travel at the speed of sound (V) from the face of the transducer to the reflecting
Although ultrasonic level indicators require little maintenance and are unaffected
by the nature of the liquid (acidity/basicity, dielectric constant, or specific gravity),
they cannot be used in circumstances of excessive foaming or turbulence or in hightemperature
situations where stratified vapor layers may be present.
On the other hand, radar detectors can be used in vessels exhibiting a wide variety
of conditions: Radar energy passes through air or vapor space with imperceptible
changes. The greatest limitation to the use of radar detectors is price, because of
the complexity of the microwave and timing circuitry. In an advanced measurement
unit, the radar level gauge sends out a continually swept microwave signal with
varying frequency. The transmitted signal is compared with the signal sent back in a
microwave mixer device, the difference in frequency being directly proportional to
the distance. It will be obvious that high-quality signal processing is necessary since
the liquid surface in a working reactor is never calm and unwanted echoes within the
reactor need to be separated out.
The chemical engineer is the best process technology team member to oversee the
general operation of a plant. In addition to managing the chemistry and engineering,
this oversight also covers GMP, Safety & Environmental requirements to meet regulatory
needs (dealt with separately), and such physical responsibilities as are involved
in waste management, plant maintenance, process automation, process containment,
plant equipment evaluation (e.g., for capital projects), technology transfer, and all
aspects of the management of the human resources. In all of these activities, the
chemical engineer collaborates with the needed specialists and experts and is often
responsible for the documentation required in validating such operations. A brief outline
of waste management, computer applications, and pilot plant/plant maintenance
is appropriate here.
Waste Management
As stated earlier, in the process development phase of a project, wastes are generally
sent out for disposal by licensed practitioners in the field. Nevertheless, the chemical
engineer is required to meet the constraints required by the operating permits for the
pilot plant or plant. Thus, it is necessary to ensure that incoming processmaterials and
outgoing products are managed (see earlier) and that in plant operations, all emissions
are properly controlled (often by condensation or by validated entrapment, absorption
or scrubbing systems) and that wastewater going to the general plant sewer meets the
standards set for disposal to a public water treatment facility. For larger operations
this may require dialogue with the public treatment facility. In these endeavors some
in-house treatment (e.g., pH adjustment, solvent stripping, and filtration) is usually
required to meet biological oxygen demand (BOD), carbon oxygen demand (COD),
and particulates content standards set for disposal to the public treatment facility.
Waste treatment and management work often provides information of immense value
should the process be subsequently taken on to a manufacturing scale.Waste disposal
issues are not infrequently a factor in process selection, or in generating a program of
work to assist the manufacturing site taking on a process in preparing its own waste
treatment operation; this may be no more than the manufacturing plant determining
whether the wastewater can be accommodated by the organisms in its own waste
treatment facility. It can become much more if waste solvent recovery and recycle
become major capital and GMP considerations. In such situations, outside specialists
are often engaged to help deal with waste issues—as an aside I remember that in
the late 1960s a lorry drove around Ulverston, England, where Glaxo has a large
manufacturing facility, which proudly displayed the slogan “We specialize in talking
Computer Applications
The application of computers in the chemical process development industry is a
field in itself, beyond the scope of this presentation. However, the following limited
overview provides some idea of the chemical engineer’s contribution in this area of
The chemical engineer is generally the person most involved in the application of
computers to process engineering problems, as well as in the application of computers
in such areas as process control, process modeling, robotics, cost calculations, and so
on. At the pilot plant level of process monitoring and control, the usual practice is to
build automation systems into process operations at late stages in the development of
a process, when the chemistry is better understood, and provide only limited process
control in the early stages of development. The sophisticated modeling of processes
is particularly valuable when a process is selected for development to a large-scale
production operation.
In the early stages of process operation at the pilot plant level, the parameters
that lend themselves well to automation are pH adjustment, pressure control, and
temperature control, the latter by automating rates of heating and cooling and rates
of addition of reagents. Simple process monitoring by automating sampling and
metering to an analytical instrument for feedback to process operators, or for utilizing
the information obtained to initiate subsequent process control steps, is also often
employed at the pilot plant level. The application of computers in the pilot plant
phase of developing a process provides an invaluable opportunity for data collection
and analysis, speeding the identification of optimal process conditions.
The automation of laboratory operations to enhance the process development
chemist’s armory has also grown as an activity. Equipment is available (e.g., Buchi
Syncore8) to aid in the evaluation and optimization of such as time-temperature cycles
in a given reaction, or in combinatorial and parallel synthesis endeavors.
Computer applications also spread to the preparation of batch sheets and other
written requirements for Regulatory purposes as well as to training courses for
8Buchi Labortechnik AG, CH-9230 Flawil, Switzerland.
operations people wherein instructional programs can aid in the education and testing
of workers, such as in Safety, Environmental, and GMP areas.
Pilot Plant and Plant Maintenance
Maintenance is a vital component of all pilot plant and plant manufacturing operations.
Any organization that strives to meet deadlines for the preparation of clinical
supplies and commercial APIs requires a plant that ideally does not break down.
To achieve this, all responsible managements require a strong, well-staffed maintenance
department that has the trained manpower, budget, spare parts, mission, and
organization to keep the plant operating.
Maintenance is not merely called upon to repair failed equipment (breakdown
maintenance). Over the course of time, maintenance workers have teamed with plant
operations personnel to develop preventive maintenance (PM), predictive maintenance
(PdM), and condition-based maintenance (CBM) programs, all in an effort to
get ahead of problems occurring in plant operations, and to do so as efficiently as
Preventive maintenance is probably the most common type of maintenance practiced
in chemical manufacturing plants. This usually means carrying out work to
ensure that a piece of equipment will operate continuously and efficiently. PM tasks
include regularly performed lubrication, parts replacement, and so on. PM is generally
carried out routinely at set intervals, usually at a scheduled shutdown. PM has
the downside of requiring work that may not be necessary and replacing parts that
may still have useful life.
Predictive maintenance (PdM) tries to avoid some of the wasted effort called for in
a PMprogram. Equipment monitoring programs are set up to provide data on a given
piece of equipment, with maintenance decisions being based on the data collected.
Condition-based maintenance (CBM) operates more like PdM with an overview
philosophy, enabling plant operators and maintenance departments to do the right
work at the right time. CBM enables those trends leading to problems to be identified
and proactive steps taken to head off breakdowns. An operating history, especially
of the critical pieces of equipment and their integration in a processing system,
is assembled. This log provides the baseline for plant “health” checks carried out
by operations and shop floor maintenance people on a continuous basis; the plant
essentially signals when it is getting sick. The charting of sensor outputs from a
given piece of equipment can enable operating and maintenance personnel to spot,
early on, changes occurring, such as vibration problems, equipment overheating,
and deterioration in the analysis of lubricating oil. These changes trigger actions
to avoid occurrence of a problem. The documented record of repairs, calibrations,
validations, and so on [frequently held in a computerized maintenance management
system (CMMS)], is additionally important in providing information to show that the
chemical process is being run according to the requirements of the StandardOperating
Procedure (SOP); the equipment log is an important component of the documentation
needed to establish GMP (Good Manufacturing Practice) requirements for the FDA.
The importance of the chemical engineer’s input in a chemical process development
program cannot be overemphasized. The chemical engineer is, in every way,
a team member and contributor whose input should be sought by every process development
chemist. As Samuel Johnson said “. . . without mechanical performance,
refined speculation is but an empty dream . . . .”
“Nothing is Forever.”
——Allen Read
Some of the most satisfying experiences in chemical process development come from
those times when one ventures into a chemical manufacturing process investigation
program in a major field. In major fields such as ?-lactam or steroid chemistry,
the subjects of this chapter, one quickly becomes absorbed by the contributions the
founders made to science, medicine, and the understanding of disease. Many of the
chemical and biological schemes they devised for themanufacture of their APIs arose
from the science of the times. Nevertheless, it is instructive to read the historical
beginnings, if for no other reason than to follow the evolution of chemical and
microbiological thinking over the decades.More than sharing in the thought processes
of the times, chemical process development chemists can contribute from a modern
chemistry perspective, even if in only a small way, to the further advancement of the
field, especially by creating new, lower-cost, safer or environmentally advantageous
molecular transformations. These can be magical experiences, properly tempered by
the humbling thought that one is frequently climbing on the shoulders of giants to
add to the vast knowledge base which is the historical foundation of the field.

Copyright C 2008 John Wiley & Sons, Inc.
I have been fortunate over a 43-year career in chemical process development in
the pharmaceutical industry to have listened to the thoughts of a few of the greats and
to work with many like-minded collaborators, both in R&D and Manufacturing, in
proposing and advancing synthesis schemes and achieving practical progress. Useful
steps forward bring considerable satisfaction, even if they are not as momentous as the
enlightenment which came to those who discovered biological activity or, painstakingly
over many years, unraveled the chemical structure of an active compound using
degradative methods.
In this spirit, it is pertinent to recount some of the work done under Arapahoe
Chemicals/Syntex, Glaxo, Bristol–Myers, and Schering–Plough auspices, much of
which illustrates the influence of competition, patent portfolios, business, newdiscoveries,
safety, the environment, and marketing, some of which shows that compelling
opportunities can resurrect chemical transformations previously cast off as unworkable
or commercially uninteresting.
The Fermentation of Penicillin G and Cephalosporin C (Glaxo)
In commencing work in the penicillin/cephalosporin field, one is easily enchanted by
the excitement attending the original discovery of penicillin by Alexander Fleming
and its later applications during World War II. In regard to the latter, one can marvel
at the vision of Florey and others who realized the contribution penicillin wouldmake
in overcoming infection and, at the basic chemistry level, appreciated the competitive
aspects of chemistry in the grudging acceptance by Sir Robert Robinson that Robert
Woodward’s proposed ?-lactam structure for penicillin was indeed correct.
In a much smaller historic way, and from my vantage point of working in Glaxo’s
Ulverston Penicillin/Cephalosporin factory under the guidance and steadying leadership
of Dr. Arthur Best, and with wonderfully committed co-workers, it is worth
recounting some of the scientific, manufacturing, and “chemosocial” issues of the
times from the mid-1960s.
The intellectual property (the patent portfolio) covering the cephalosporin field
was held by the National Research and Development Council (NRDC) in the United
Kingdom, largely because early recognition of the value of cephalosporins had resulted
from the pioneering work at Oxford University on strains of a microorganism
first harvested by Professor Brotzu from a sewage outfall in Sardinia.
The biosynthesis of penicillins and especially cephalosporin C was a hot academic
and industrial field during my time in Glaxo (1966–1975). The work of Arnstein
and co-workers1 had more or less established a broad outline of the transformations
occurring in the biosynthesis of penicillins, via the oxidative cyclization of
the tripeptide, l-?-aminoadipoyl-l-cysteinyl-d-valine (ACV-tripeptide). I had the
privilege of sitting in with the fermentation people in the Ulverston factory (notably,
Drs. Don Hockenhull, Bob Fildes, and Steve Goulden) during discussions with
1Arnstein, H. R.V., Morris, D. Biochem. J. 1960, 76, 357.
Professor Edward Abraham, Oxford University, on his many ideas to determine the
precise pathway for the biosynthesis of cephalosporin C and on his several unmet
requests that Glaxo provide him with its high-yielding cephalosporin C producing
strain for his work on the biosynthesis. Professor Abraham’s gentlemanly requests
contrasted sharply with the anecdotal stories of earlier rapacious times when opportunistic
entrepreneurs tried to obtain fermentation strains by laying agar plates
outside Glaxo’s Barnard Castle Fermentation Plant in the hope of collecting biological
fallout from factory air exhaust systems.2,3 None of us in Ulverston contributed
to the unraveling of the biosynthesis of Cephalosporin C, but work continued in Professor
Abraham’s laboratories. A few decades later, Professor Jack Baldwin crowned
the Oxford work in brilliantly defining the entire Cephalosporin C biosynthesis
As with most manufacturing operations, the main focus of the technical support
work in the Ulverston factory was to:
1. Reduce the cost of fermentation products, primarily penicillin G and
Cephalosporin C, by improving process performance—for example, increase
the fermentation titer and extraction efficiency at the same time meeting or
improving product quality; process improvement frequently involved much
troubleshooting work to resolve production problems.
2. Reduce the cost of chemical intermediates by improving existing chemistry.
Troubleshooting frequently provided leads for process evolution and also ideas
leading to better chemistry for the conversion of fermentation products to chemical
intermediates—for example, cephalosporin C to 7-ACA. Again, product
quality is the most important guiding parameter.
3. Reduce the cost of manufacturing and improve the quality of bulk APIs made
from the chemical intermediates.
Although the chemical development organization in Ulverston had little to do
with improving the fermentation processes, it was always necessary to stay vigilant to
ensure that when a change was made to the fermentation process, there was no impact
on the subsequent chemical transformation steps. Changes were sometimes imposed
by natural events outside anyone’s control. I recall that the change in fermenter
feed from one season’s molasses to another could play havoc with the fermentation
titer. On another occasion during a drought, the well water level fell to levels that
led to high salts content in the water used for fermentation, adversely affecting the
titer. Such events only added to the difficulties of making progress in fermentation
2Dr. Brian Boothroyd, personal communication.
3Fermentation factories are always sensitive to the possibility of industrial espionage negating commercial
advantages they had gained from R&D work to improve their fermentation strains. For this reason, before
Glaxo would send out waste proteinaceous spent mycelium for animal feed, they always pre-baked it to
destroy viable microorganisms.
4Baldwin, J. E., and Schofield, C. The Chemistry of ?-Lactams, Page, M. I., Ed., Blackie Academic and
Professional, Glasgow, 1992, p. 1.
development. In the early days, there was no good method for gas analysis to study
oxygen absorption and carbon dioxide expiration during the growth and producing
stages of the microorganism. More productive strains were discovered by treating
the regular strains with chemical mutants or UV light and plating out the surviving
organisms. In the manufacturing plant, stirring rates and the delivery of air were
optimized by trial and error. Statistical analysis of plant data played a large part in
process improvement.
When problems occurred in fermentation operations, it could take weeks to identify
the cause of poor performance and take corrective action. As a result, there was a
considerable degree of empiricism and “feel” associated with fermentation development
and controlling a “living process.” It was therefore not surprising that the filing
cabinets of the fermentation development and production staffs contained enormous
numbers of reports on problems that came back time and again. Unfortunately, the
empiricism associated with managing fermentation development all too often spilled
over into the chemical development side of operations, such that one of the greatest
managerial challenges was to ensure that a quantitative, rational analysis and testing
program was used to resolve the usually more quantifiable chemical synthesis
Despite the turbulence encountered in all process improvement and process development
work, there were the occasional triumphs of applied common sense. One
of considerable social as well as economic consequence, occurring in the manufacture
of penicillin G, resulted from the Ulverston process improvement attending the
switch from buying solid phenylacetic acid to purchasing an aqueous solution containing
50% sodium phenyl acetate. In using the solid, our factory workers weighed
the required amount of phenylacetic acid and then dissolved it in aqueous sodium
hydroxide to produce the aqueous solution used for feeding the penicillin fermenters.
The handling of solid phenylacetic acid had created problems for many years, both
for the Ulverston factory and for our vendor, Albright and Wilson (A&W). Solid
phenylacetic acid introduced an obnoxious, pervasive, sweaty aroma to the penicillin
G buildings and the workers’ clothes and homes in Ulverston. The odors in Ulverston
and surrounding communities were, however, almost trivial besides those encountered
by theA&Wworkers.Keith Partridge, Sales and MarketingManager forA&W
during my time at Ulverston (1966–1975), recounted the privations of process operators
manufacturing phenylacetic acid in A&W’s Ann Street works in Widnes. They
were paid a “social bonus” for working in the plant. Keith said you could walk down
the street where they lived and identify their houses by the smell!! No one would sit in
their seats at the local public house—even their beer glasses were segregated!! The ultimate
indignity occurredwhenVernons, the football pool company, asked oneworker
not to send in hisweekly pool coupon because of odor complaints from the clerks who
processed it!!!
The common sense use of the 50% aqueous solution of phenylacetic acid virtually
eliminated the handling of solid material, though it took some time and a
few plant trials to convince the production managers that there was nothing else in
the A&W aqueous solution which might have an adverse effect on the penicillin
G titer.
Filter mycelium
D-amino acid oxidase*/H2O2
Chromatographic purification
SCHEME 1. ?D-AAO derived from Aspergillus and Trigonopsis strains.
† Added H2O2 improves the conversion of the intermediate keto acid to the glutaroyl side
chain. Indeed H2O2 and a ketone also effect the conversion of I to II.7
The scientists and engineers in the Ulverston factory interacted with the Glaxo
Research and Development people in Greenford (Chemistry) and Sefton Park
(Microbiology/Biochemistry and Fermentation/Extraction)—an interaction that
added vitality, good ideas, some rivalry, and great exchanges to the benefit of all
who participated. Ultimately, the Ulverston factory benefited the most. A number of
notable advances in improving process efficiency came out of the Research Groups,
some of which were implemented. Some were before their time, but should not have
One such advance was the work carried out in Sefton Park by Dr. Robert Fildes and
co-workers5,6 on the enzyme-mediated conversion of the zwitterionic aminoadipic
acid side chain in Cephalosporin C into a glutaroyl side chain (a process that rendered
the molecule more amenable to solvent extraction from filtered fermentation broths;
see Scheme 1.
At the time (early 1970s) the enzyme reaction worked well, and the
3-acetoxymethyl-7(R)-glutaroylaminoceph-3-em-4-carboxylic acid (II) produced
could be extracted into n-butanol fairly readily. However, isolation of the free acid
(II) or salt, added to the losses in the extraction step and created only a marginal
economic advantage versus the existing process [chromatographic enrichment of
cephalosporin C in the filtered broth followed by concentration (Luwa evaporator)
and precipitation as the potassium salt]. Glaxo Research, and to a lesser extent the
Ulverston staff involved in the recovery of potassium Cephalosporin C, worked on the
process for some time but did not succeed in making it practical. The idea was a good
one that, finally (almost 30 years later), came to fruition in the hands of Antibioticos
S.p.A. in Settimo-Torinese, Italy. Antibioticos succeeded in building on Glaxo’s
original ideas by finding and immobilizing both an amino acid oxidase and, separately,
a glutaroylamidase on polymer supports.8 These innovations, coupled with the
adroit use of macroreticular resins for purifying their cephalosporin intermediates,
enabled them to create an all aqueous process for the manufacture of the important
cephalosporin intermediate, 3-acetoxymethyl-7-aminoceph-3-em-4-carboxylic acid
(7-ACA-III) (Scheme 2).
5Arnold, B. H., Fildes, R. A., Gilbert, D. A. U.S. Patent 3,658,649, 1972 (to Glaxo).
6Fildes, R. A., Potts, J. R., and Farthing, J. E. U.S. Patent 3,801,458, 1974 (to Glaxo).
7Suzuki, N., Sowa, T., and Murakami, M. U.S. Patent 4,079,180, 1978 (to Asahi).
8Cambiaghi, S., Tomaselli, S., and Verga, R. U.S. Patent 5,424,196, 1995 (to Antibioticos S.p.A.).
Immobilized glutaroyl
7-ACA acylase
H to pH 3.5
Filter mycelium
Chromatographic purification
(macroreticular resins)
Immobilized D-AAO/H2O2
SCHEME 2. Antibioticos’ industrial process for the conversion of Cephalosporin C in fermentation
broths to 7-ACA.
By combining immobilized enzyme and chromatography technologies, Antibioticos
realized the long-standing objective of avoiding intermediate isolations, thereby
making 7-ACA the first product isolated from the fermenter. Asahi’s position in the
same field7 contributed to the Antibioticos achievement. Biochemie in Austria has
also had notable success in a similar, independent program.
Environmentally clean technologies such as those outlined in Scheme 2 have, in
addition to reducing the cost of 7-ACA, introduced new opportunities for the preparation
of other commercially useful cephalosporin intermediates (see later section
entitled “Back to Classical Cephalosporins”).
Fifty years ago, the simple molecular manipulation of penicillins isolated from
the fermenter (Penicillins G and V) was limited to producing 6-aminopenicillanic
acid and acylating the amino group with different acylating agents, thereby providing
novel products with an improved biological spectrum. Looked at in the same way,
7-ACA offered two sites for molecular manipulation: the 7-amino group and the
3-acetoxymethyl group. Many cephalosporins with new and improved biological
spectra resulted from replacing these groups; the interested reader is referred to
books and symposia publications on this subject.9
The Ring Expansion of Penicillin Sulfoxides to Cephalosporins
In the early days, much was done by the major pharmaceutical companies to
build on the vast chemistry opened up when Morin and Jackson (Eli Lilly) discovered
their pioneering process for ring expanding penicillin sulfoxides (IV) to
3-methylcephalosporins (V).10 This breakthrough subsequently led to the discovery,
development, and marketing of many new antibiotics.
9(a) Cephalosporins and Penicillins: Chemistry and Biology, Flynn, E. H., Ed., Academic Press, New
York, 1972. (b) Chemistry and Biology of ?-Lactam Antibiotics, Morin, R. B., and Gorman, M., Eds.,
Academic Press, New York, 1982. (c) Recent Advances in the Chemistry of ?-Lactam Antibiotics, Special
Publication No. 28, J. Elks, Ed., The Chemical Society London, 1977. (d) idem, Special Publication No.
38, Gregory, G. I., Ed., 1981. (e) idem, Special Publication No. 51, Brown, A. G., and Roberts, S. M.,
Eds., 1984.
10(a) Morin, R. B., Jackson, B. G., Mueller, R. A., Lavagnino, E. R., Scanlon, W. B., Andrews, S. L., J.
Am. Chem. Soc. 1963, 85, 1896. (b) Idem, ibid. 1969, 91, 1401. (c) Morin, R. B., Jackson, B. G. U.S.
Patent 3,275,626, 1966 (to Eli Lilly).
Strong acid (e.g., p.TSA)
Heat @ 125–150°C in
inert solvent
R = C6H5OCH2CO. or C6H5CH2CO.
R' = Readily removed carboxyl protecting group
SCHEME 3. Outline of the Morin – Jackson ring expansion.10 c
The Morin–Jackson ring expansion discovery led first to the orally absorbed
antibiotic, Cephalexin, marketed by Eli Lilly and Glaxo under National Research and
Development Council (NRDC) licenses.
The Morin–Jackson ring expansion as first described required that the 3-carboxyl
group be blocked by a readily removed protecting group (Scheme 3).
Several companies were involved in the search for the best carboxyl protecting
group, notably Ciba–Geigy, Eli Lilly, Glaxo, and later Gist–Brocades. Ciba–Geigy,
through the work of Professor R. B. Woodward,11 discovered the value of the 2,2,2-
trichloroethyl (TCE) group. This group, like the ?-lactam ring itself, proved to be
stable during the conditions used in the ring expansion process, and also the conditions
used in the subsequent PCl5 cleavage reaction to give the 7-aminocephalosporin TCE
ester. The TCE group was usually removed by treating the TCE ester with zinc.12
At about the same time (early 1960s), Eli Lilly found that the p-nitrobenzyl (PNB)
groupwould provide protection equivalent to that provided by the TCE group with the
further merit of introducing good crystallinity, and therefore ease of purification, to
the products of the ring expansion process. Early ways of introducing the pNB group
involved reacting p-nitrobenzyl bromide (a vesicant) with the penicillin carboxylate.
Reductive removal of the PNB group was also found, by Eli Lilly workers, to give a
hazardous waste (allegedly containing carcinogenic tolidines13). In its early work on
the ring expansion process, Glaxo followed Ciba–Geigy’s TCE lead, recognizing that
Ciba–Geigy would subsequently patent the process. This enabled Glaxo to get off
to a fast start to find a commercially useful ring expansion process and, separately,
to undertake work to develop a process based on an alternative protecting group.
Glaxo Research chose to follow Eli Lilly’s lead to develop a process based on PNB
protection since Glaxo had rights, through the umbrella license which the NRDC had
11(a) Woodward, R. B. Science, 1966, 153, 487. (b) Woodward, R. B. Angew. Chemie, 1966, 78, 557. (c)
Woodward, R. B., Heusler, K., Gosteli, J., Naegeli, P., Oppolzer, W., Ramage, R., Ranganathan, S., and
Vorbr?uggen, H., J. Am. Chem. Soc. 1966, 88, 852.
12(a) Woodward, R. B. Brit. Pat 1,155,016, 1969. (b) Woodward, R. B. U.S. Patent 3,828,026, 1974 (to
Ciba Geigy).
13In commercializing the use of pNB protection, Eli Lilly built a special containment facility in its Clinton,
Illinois, Cephalexin plant in order to segregate the waste from removal of the pNB group for separate
created with several companies, including Eli Lilly, to sublicense the findings made
by all members of the “NRDC club.”
A few of us in the Glaxo factory organization in Ulverston thought that the hazard
to process operators that would be introduced by using p-nitrobenzyl bromide and by
implementing the reductive removal of the PNB group were undesirable in a factory
setting. On these grounds, Dr. Arthur Best supported my proposal to evaluate the
diphenylmethyl (DPM) group, with the conceptual objective being to introduce DPM
via reaction of penicillin G sulfoxide acid with diphenyldiazomethane, produced in
situ, and to subsequently remove the DPM group by acid-catalyzed solvolysis.14
The competition that ensued with the Glaxo, Greenford, chemical development
group (developing PNB protection) and the Ulverston team was a healthy, if tortured
one. Many in Greenford believed that Ulverston staff should be doing process troubleshooting
work on the existing chemical processes and working to increase process
yields and reduce costs. But Dr. Best was having none of it, at the same time recognizing
that the Ulverston activities were considered heretical by a number of people
in Greenford. Looking back, the risks to Ulverston’s management were undoubtedly
considerable. Nevertheless, joint meetings were organized with Greenford R&D scientists
and I personally was given tacit support by Dr. Joseph Elks (at the time the
Greenford Director of Chemistry). Dr. B. A. Hems (later Dr. B. A. Hems, FRS), who
was the managing director of all R&D in Greenford and himself a visionary,15 chose
to let the Ulverston work go on.
As detailed in our later publications and patents,16 we found that diphenyldiazomethane
(DDM) could be prepared in high yield by the peracetic acid oxidation
of benzophenone hydrazone in methylene chloride in the presence of a base. We
used peracetic acid in the presence of a nonoxidizable base, tetramethylguanidine
being the base of choice. Since, at that time, the separate preparation of DDM was
perceived as adding an unsafe step to the process of producing our target molecule,
penicillin G sulfoxide diphenylmethyl ester, we quickly moved on to finding a practical
process that would enable us to prepare DDM in the presence of penicillin G
sulfoxide acid under process conditions wherein the DDM would be immediately
scavenged by the sulfoxide acid. We found that this could simply be done by adding
peracetic acid to a solution of benzophenone hydrazone and penicillin G sulfoxide
acid in methylene chloride. However, we found that the yield to the target DPM ester
was quite variable. Fortunately, our Dr. Roy Bywood made the connection between
yield of the target ester and the source of penicillin G sulfoxide acid. He observed
that old samples of sulfoxide acid gave better yields of ester than new ones and
traced the difference to their methods of preparation. Old samples had been prepared
14This proposal was based on speculation that the work of Horner, L., and Fernekess, H. Chem. Ber., 1961,
94, 712 (in which they showed that moderate yields of esters and ethers could be obtained by adding
peracetic acid to benzophenone hydrazone in the presence of excesses of carboxylic acids and phenols)
may provide a basis for a commercial process to DPM esters.
15Jack, Sir D., and Walker, T. Roy. Soc. Biograph. Mem. 1997, 217.
16(a) Adamson, J. R., Bywood, R., Eastlick, D. T., Gallagher, G.,Walker, D., andWilson, E. M., J. Chem.
Soc., Perkin Trans. I, 1975, 2030. (b) Gallagher, G., andWalker, D. U.S. Patent 4,083,837, 1978 (to Glaxo).
(c) Bywood, R., Gallagher, G., Sharma, G. K., andWalker, D., J. Chem. Soc., Perkin Trans. I, 1975, 2019.
1. PhC=NNH2, I2, CH3CO3H
in CH2Cl2
2. Ring expansion* (Scheme 3)
1. PCl5 cleavage of the
2. Activated N-blocked
3. Deblocking (HCO2H or
dil. aq. HCl)
*Glaxo employed a bidentate catalyst, pyridinium dichloromethylphosphonate, in
refluxing dioxane to carry out the ring expansion.
SCHEME 4. Glaxo process for the manufacture of Cephalexin.
by the oxidation of penicillin G with sodium metaperiodate whereas new samples
had been prepared using the much cheaper peracetic acid. He proved that the yield
discrepancy was related to contamination of the penicillin G sulfoxide acid by iodide
ion. Higher ester yields were obtained in the presence of traces of iodine or iodide
ion; typically, when peracetic acid (1.4 mole) was added to a solution of penicillin G
sulfoxide acid (1 mole), and benzophenone hydrazone (1.3 mole) in chloroform in
the presence of iodine (0.002 mole), an almost quantitative yield of the target DPM
ester was obtained. Levels of iodine significantly lower or higher than 0.0015 mole
equivalents with respect to the hydrazone gave poorer yields.
This work led on to a Glaxo process patent for the preparation of cephalexin (VI)
using DPM protection17 (Scheme 4). The Ulverston process was developed through
the pilot plant phase and proven to give high-quality Cephalexin.Moreover, the DPM
ester wastes were shown to be readily oxidizable to benzophenone using aqueous
nitric acid.18 Successful plant trials followed to compare both the DPM and PNB
At a final process selection meeting in Greenford, Dr. Best presented the case
in favor of adopting the DPM process. The process economics vs. PNB were only
marginally in favor of DPM. The decision to adopt DPM was based largely on our
own patent position and operator safety and environmental considerations. The DPM
process was scaled up in Ulverston (through the ring expansion step) and Glaxo’s
factory in Montrose, Scotland, and served for many years to produce Cephalexin for
Glaxo’s marketing needs. Later on, Dr. Hems, during a visit to the Ulverston factory,
congratulated us for “sticking to our guns” in developing the DPM process.
17Bywood, R., Gallagher, G., and Walker, D., German Patent 2,311,597, 1973 (to Glaxo).
18The oxidation is carried out by refluxing the “benzhydryl” waste in aqueous 30% nitric acid (see footnote
17, example 37). The benzophenone produced is easily converted to benzophenone hydrazone for recycle
to the DDM process. See also Rivkin, S. M. J. Appl. Chem. (USSR), 1938, 11, 83 (Chem. Abstr., 1938,
32, 4566). The nitric acid oxidation process was adopted by Albright and Wilson, our benzophenone
hydrazone supplier.
1. Acylate amine
2. Overlayer with
DDM inCH2Cl2
3. H+
(in filtered broth)
Yield >90%
SCHEME 5. Extractive esterification of Cephalosporin C derivatives.
We went on, in a lower-key way, to explore the possibilities of using diphenyldiazomethane
(DDM) for the extractive esterification of cephalosporins from aqueous
solution (Scheme 5).19 This worked very well. It was not followed up for commercial
use at the time, but did find use in a later process development project carried out
with Schering–Plough (see later). Two key factors in the decision to further evaluate
DDM as a potentially useful reagent for extractive esterification were its better-thanexpected
stability20 and the creation of an improved process for producing DDM.
Our Ted Wilson16(a) discovered that the previously used, and expensive, tetramethylguanidine
base could be replaced by aqueous sodium hydroxide and a phase transfer
catalyst (Aliquat 336) in a two-phase system with methylene chloride. This process
also gave excellent yields of DDM (>90%).
We extended our process work to prepare new polystyrene resins carrying the
diazoalkane group (see Chapter 11).
All of this work introduced us to the world of patents and led to a valuable,
independent patent portfolio for Glaxo.
It is pertinent to point out that, after the capital investment program in equipment
for the DPM-based process for Cephalexin manufacture was well advanced,
we became aware of the pioneering work of Gist–Brocades (now a part of DSM) in
using the trimethylsilyl (TMS) group as a temporary protecting group for carboxyl
in the ring expansion reaction.21 Although the TMS group was removed by hydroxylic
materials during the work-up of every process step, its advantages of low cost,
a very simple procedure for preparing TMS esters (carboxylate and TMS Cl), and
its ease of removal (water regenerates the carboxylic acid and hexamethyldisiloxane
when a little acid is present) recommended TMS protection to many companies
(Bristol–Myers went further in utilizing the even cheaper dichlorodimethylsilane as
a protecting group; in practice, the waste polysiloxane was generally disposed of via
waste haulers). Gist–Brocades developed their discovery into a very large market for
7(R)-amino-3-methylceph-3-em-4-carboxylic acid (7-ADCA). Glaxo looked at the
19Robinson, C. H., and Walker, D. U.S. Patent 4,059,573, 1977 (to Glaxo).
20Sakamoto, D., Hirayama, Y., Kohno, Y., Sakai, T., Shiraishi, Y., and Saijo, S. European Patent 177,248,
1987 (to Taoka Chemical Co., Ltd).. These workers report that crystalline DDM undergoes no decomposition
when held at 5?C for 100 hr, but that solutions in CH2Cl2 deteriorate faster: A 50% solution
decomposes 4.9% in 20 days and 30.5% in 40 days at 0?C.
21(a) de Koning, J. J., Kooreman, H. J., Tan, H-S., and Verweij, J. J. Org. Chem. 1975, 40, 1346. (b)
Verweij, J., and DeVroom, E. Recl. Trav. Chim. Pays-Bas, 1993, 112, 66–81.
possibility of producing Cephalexin via the Gist 7-ADCA process (or to purchase
Gist 7-ADCA), but decided that the economics did not justify the change. [Although
the chemical ring expansion of penicillins to 3-methylcephalosporins continues to
be used for the manufacture of the orally active 3-methylcephalosporins, there remains
that never-say-die belief among biotransformation afficionados, particularly
Professor Arnold Demain, that one day the biological ring expansion of Penicillin
G to deacetoxycephalosporin G (DAOG) using deacetoxycephalosporin C synthase
(DAOCS) will succeed to the point of becoming an industrial process:
Penicillin G
Streptomyces clavuligerus
Source of DAOCS
Work published relatively recently22 indicates that the yield in this process has been
improved to 10%.
Given a much higher yield, the enzymatic process could save operating costs
and reduce waste disposal and environmental problems vs. the established penicillin
sulfoxide ring expansion process. However, the major issue is whether an
old industry with declining profitability, would be able to justify the cost of investment
in better technology (cf. the case of ceftibuten discussed later). In related
work, 6-adipoylaminopenicillanic acid has been shown to be convertible to
7-adipoylaminocephalosporanic acid in very low yield by Streptomyces clavuligerus
expandase gene.]23
It was during this period that Glaxo’s Research management, and also the Manufacturing
management, changed with the retirement of Dr. Hems and with other
organizational maneuvers. The effect of these changes on the small Ulverston process
development operation was traumatic. It was decreed that all process development
work would return to the R&D group in Greenford and that Ulverston would concentrate
on troubleshooting and cost reduction. Dr. Robert Fildes, by then head of the
Ulverston fermentation operation, left to join the Bristol–Myers Industrial Division.
He persuaded me to follow him. Other co-workers departed to join other parts of
the Glaxo organization or other companies. The rough organizational trampling by
Glaxo’s senior management brought to mind Machiavelli’s 16th-century observation:
There is nothing more difficult to carry out,
nor more doubtful of success, nor more dangerous to handle,
than to initiate a new order of things.
22Adrio, J. L., Demain, A. L., J. Org. Proc. Res. Dev., 2002, 6, 427.
23Crawford, L., Stepan, A. M., McAda, P. C., Rambosek, J. A., Conder, M. J., Vinci, V. A., and Reeves,
C. D., Bio/Technology, 1995, 13, 58.
The changes could have been made with more consideration of the consequences
—how to keep the talent and experience lost!
Semisynthetic Penicillins and Cephalosporins: Adventures in Product
Quality (Bristol–Myers)
Bristol–Myers built on Eli Lilly’s discovery of Cephalexin by funding the research
that led to the development of Cefadroxil (VIII), just as Beecham had done in
building on its own discovery of ampicillin with the invention of amoxicillin. Both
amoxicillin and cefadroxil carried the d(-)p-hydroxyphenylglycyl side chain which
served to extend biological activity and increase blood levels versus ampicillin and
All of these semisynthetics had their own quirks in the sense that routinely achieving
high quality on a production scale proved very difficult. The major source of most
quality problems with all four of these compounds lay in the quality of the 6-APA or
7-ADCA used for their manufacture. Traces of metals such as iron, as well as traces
of unknown impurities, affected the color and even the taste of the products.
I recall Bristol–Myers’ Managing Director in Brazil, Paulo Mendez, loudly complaining
of ampicillin and amoxicillin quality problems that were traced back to the
quality of some lots of 6-APA made in Syracuse, New York. “You were supposed to
ship 6-APA,” he forcefully reminded us, “and I emphasize the 6. Your last shipment
was only 51
Similarly, Banyu in Japan bought Cephalexin from Bristol–Myers’ facility in
Latina, Italy. In the early days Banyu complained that the taste and odor of Latina’s
Cephalexin batches varied considerably, with many batches failing to meet Banyu’s
organoleptic criteria, particularly for taste. Since taste and odor are not quantifiable
parameters, the dilemma persisted for some time. During a visit to Latina, Banyu’s
President, Dr. Iwadare, politely explained that Japanese children actually retched
when obliged to take Banyu’s Cephalexin oral suspension. The seriousness of the
situation became clear to Latina management when Banyu refused to accept free
goods that failed their organoleptic tests. The rest of the story was reported to me by
my Latina colleague, Dr. Ettore Visibelli. In an effort to resolve the problem, Banyu
sent their President’s son and their Director of R&D to conduct taste tests on batches
of Cephalexin and to “train” Latina people to carry out the taste tests to identify
Cephalexin batches suitable for shipment to Japan. Dr. Visibelli said that there was
a considerable uproar at the start of the exercise since the Banyu team proved to
be quite persnickety in selecting a room where the taste tests would be performed.
They wanted a room with no distracting odors, good air conditioning and excellent
lighting—the appearance of the powder was important to them. They sniffed the air in
many rooms in several buildings before deciding that the Engineering Department’s
draftsmen’s room met their standards (the story quickly spread around the Latina
factory that the draftsmen didn’t sweat as much as everyone else!). But that was not
all. The Banyu team spent some time in selecting the right Italian water for mouth
rinsing after each test.
The taste test itself was a ceremony to behold (as many did, taking turns to look
through the window). It seemed to take on the aura of a religious occasion as each
taster sniffed his freshly opened bottle and then carefully spooned a very small amount
of powder onto a clean plate. After a debate about the powder’s appearance, they lifted
their plates to theirmouths in unison and licked the sample off. A pause followed with
exquisite facial expressions registering their reactions to their palates. The mouth was
rinsed and the results of the tasting were written down. Only six licks per day could
be undertaken such that it took the Japanese several days to lick their way through
the assembled batches. The ruling on acceptable batches and unacceptable batches
was handed down and arrangements made to ship the acceptable batches to Japan.
The use of Latina staff to routinely taste test the Cephalexin batches never happened.
Quite apart from the concern that the unions might react to people being used as
“tasting guinea pigs,” it was agreed that a scientific solution was needed. Indeed, Dr.
Visibelli and his staff, working with Latina analysts and production people, found
that the presence of acetone in the last process step led to the formation of traces of an
unstable Schiff base with Cephalexin which degraded to give unidentified products
causing the odor and taste problems.
Beyond quality exercises, there were many patent issues to address, a few of
which are detailed in the patents presentation (q.v.). Work in the cephalosporin field
continued at Bristol–Myers, but the compounds produced and marketed, cephapirin
and ceforanide, were overtaken by products of other companies, notably cefuroxime
axetil (Glaxo), ceftriaxone (Roche), and cefotaxime (Aventis—at one time Hoechst
Roussel). Later the penems and carbopenems further diminished interest in the earlier
In 1981, management changes in the Bristol–Myers organization formalized a
split in the duties of our chemical process development group. The Research arm of
the organization built its own development group to supply the small quantities of its
API’s for initial screening tests. As the scientific challenges narrowed, several of us
left—I being attracted to Schering–Plough (1982).
Penems and Cephalosporins—Flourishing Science (Schering–Plough)
Although Schering–Plough was active in the ?-lactam field through its discovery of
several interesting penems, none of them made it to the marketplace. However, the
period 1982–1996 (when I “graduated” rather than retired!!) was one of the most
fruitful periods of my career, notably for the opportunity to work with outstanding
people doing great science and to build a chemical development organization of
considerable power, with the indispensable encouragement and support of one of
only a few visionaries encountered in my career, Dr. Hal Wolkoff.
The first few turbulent years were distinguished by our hiring of many chemical
engineers and chemists, by the vital acquisition of modern instrumentation (NMR,
HPLC, GC, and IR), and by the creation of new laboratories. An adventitious instant
increase in plant capacity, enabling chemical development to meet growing demands
for bulk APIs, resulted from the transfer of Schering’smanufacturing assets in Union,
NJ [including personnel and responsibility for continuing production of a few low
volume products (albuterol, aurothioglucose, and others)] to chemical development.24
Several of my Bristol – Myers colleagues (Bruce Shutts, Steven Yu, Mario Ruggeri,
and Dr. Chou Tann) followed me to Schering–Plough. The agreement with
Schering was that in addition to meeting our Research commitments and supervising
the manufacture of the low volume products, we would also support the manufacturing
people in their efforts to improve the performance of steroid manufacturing
operations in Mexico City and Manati, Puerto Rico. The manufacturing division
funded the chemical development support team in a collaborative effort with Mexico
City and Manati scientists. It was agreed that this effort would be ancillary to the
main objective, producing quality APIs for Schering Research clinical, toxicology,
and pharmaceutical development programs.
Our work to develop the synthesis methods that would deliver kilos of penems to
Research owed much to the brilliant scientific achievements of Schering–Plough
Research’s Drs. Ashit Ganguly, Stuart McCombie, Viyyor Girijavallabhan (Giri),
and Adrian Afonso, who defined the synthesis sequence. However, in reducing their
work to practice in the chemical development laboratories and pilot plants, much
was owed to the skills and dedication of our own chemists and chemical engineers,
notably Ray Werner, Bruce Shutts, Lydia Peer, Stan Rosenhouse, Pete Tahbaz, Bob
Jaret, and Drs. Marty Steinman and Dick Draper. None of our own achievements in
delivering kilos of penems would have been possible without the starting material,
initially methyl 6,6-dibromo penicillanate (IX) and later methyl (5R,6S,8R)-6-(1-
trichloroethoxycarbonyloxyethyl) penicillanate (X) produced with enthusiastic and
practical flair in Schering’s Union pilot plant by Stan Rosenhouse et al., and subsequently
in Schering’s manufacturing plant in Ireland by Drs. Brian Brady, Henry
Doran, and Maurice Fitzgerald, very ably supported by Michael Miley, who ran the
pilot plant. Our Swiss chemical development pilot plant group, enterprisingly led by
Phil Ottiger (and later by Dr. Ernst Vogel) and Kurt Jost, were responsible for sourcing
the side-chain and blocking groups as well as implementing several of the steps
beyond compound X. The Irish group became responsible for the implementation of
all of Scheme 6.
24Such a transfer was made possible by Schering–Plough’s investment in new production facilities in
Manati, Puerto Rico.
Zn pH 7–8
Purification N
SCHEME 6. Schering–Plough method for the preparation of methyl (6S, 7R, 8R) 6-(1-
trichloroethoxycarbonyloxyethyl) penicillanate (X).
Skilled adaptation of the laboratory methodology outlined in Scheme 6 created a
practical method for the preparation of high-quality X in batches of several hundred
kilograms. Ireland went on to become the main supplier of X, thus providing the
Union, NJ and Swiss pilot plants with all the raw materials needed for the subsequent
steps. This strategy also eliminated penicillin contamination concerns from the Union
site—at least until the last steps. The Unionmethodology for convertingXto Schering
penem Sch 34343 is outlined in Scheme 7.
It can be quickly appreciated that the methodology illustrated in Schemes 6 and 7,
although radically improved to render it practical, was far from providing the basis of
an industrial process. Themajor burdens were the use and disposal of mercuric acetate,
the use of other reagents and solvents such as methyl iodide (suspected carcinogen),
allyl iodoacetate (lachrymator), and chloroform (carcinogen)25 which posed health
hazards, and the large number of synthesis steps, some giving relatively poor yields,
which created a significant cost of goods problem. Moreover, the synthesis lacked
in elegance since only the core ?-lactam ring of the starting 6-APA is incorporated
into the final penem. In addition, FDA requirements that we use dedicated process
equipment in segregated areas to avoid “penicillin” cross-contamination possibilities
raised considerable logistics problems.
In progressing the API supply program in the period 1982–1985, we managed to
meet rising expectations, mostly for increased kilo requirements, at the same time
25Although chloroform is classified as a “confirmed carcinogen . . .” in Sax’s Dangerous Properties of
Industrial Materials, 8th edition, Volume II, Lewis, R. J., Ed., van Nostrand Reinhold New York, 1992,
p. 815, its status is questioned by the former director of the Bioassay Segment of the National Cancer
Institute (see Weisburger, J. H., Chemical and Engineering News, December 17, 2001, p. 8). In addition
to giving equivocal results in rat and mice tests, Dr. Weisburger points out that chloroform is neither
genotoxic nor mutagenic and does not combine with DNA.
89 % N
SCH 34343
Deblock with
(Ph3P)4 Pd
TCE = —CH2CCl3 ; Tr = (C6H5)3C? ; Naphth = ; LiHMDS = LiN [Si(CH3)3]2 ; R=CH3(CH2)3CHC2H5
SCHEME 7. Method for the preparation of Schering penem SCH 34343, from intermediate
as adding manpower resources, upgrading laboratories and pilot plants, acquiring
modern instrumentation, and concurrently intensifying our dialogue with the Analytical
Development operation. There was a short respite as we mourned the demise of
our first penem, Sch 29482,26 partly owing to the significant social problems associated
with its use.27 By the time the second penem (Sch 34343, XI) was identified,
Chemical Development was well advanced in the transition from a largely empirical
experimental culture to one based on process innovation and seeking the best possible
understanding of the chemistry being scaled up.
In searching for a better synthesis of penems free of the previous need to start with
a penicillin, we looked for sources of a starting material that would be relatively low
in cost, which would avoid the penicillin cross contamination concerns in the early
steps of the synthesis and which would allow us to incorporate as much of the starting
material as possible in the final penem. We looked at an l-threonine route devised
26(a) Girijavallabhan, V. M., Ganguly, A. K., McCombie, S. W., Pinto, P., and Rizvi, R. Tetrahedron
Letters, 1981, 22, 3485. (b) Afonso, A., Ganguly, A. K., Girijavallabhan, V. M. and McCombie, S. W. In
Recent Advances in the Chemistry of ?-Lactam Antibiotics, Brown, A. G. and Roberts, S. M., Eds., S.
Special publication No. 52, The Royal Society of Chemistry, London; 1984, p. 266.
27SCH 29482 carried a 2-ethylmercapto side chain in place of the 2-carbamoyloxyethylmercapto group in
Sch 34343. Volunteers enrolled in the clinical trials developed a foul odor, attributed to ethylmercaptan,
which contaminated their clothing, offices, and homes. There was one anecdotal story of a volunteer in
the American South who was chased by dogs when his wife sent him out to the woods behind their house
to relieve himself. Schering tried, unsuccessfully, to mask the problem by formulating Sch 29482 with
chlorophyll. Lack of success was predictable. I recall a BBC radio talk given by an expert in the 1950s,
debunking the use of chlorophyll as a mouth freshening deodorant in toothpaste. He ended his talk with
the lines: “The goat that reeks on yonder hill has browsed all day on chlorophyll!”
O Si(CH3)3
1. Base
2. H+
O Si(CH3)3
SCHEME 8. Method for the conversion of O-acetyl-l-threonine into and intermediate suitable
for Sch 34343 preparation.
by our research colleagues, notably Drs. Sam Chackalamannil, Adrian Afonso, and
Ashit Ganguly. This is outlined in Scheme 8.
The major virtues of this route to penem XI lie in the low cost of l-threonine and in
avoiding a penicillin starting material. However, the route intersects with the Scheme
7 chemistry at too early a stage in the synthesis, with the need to introduce sulfur
at C-4. Despite these disadvantages, we did begin an evaluation of the l-threonine
We also became interested in the possibility of evaluating sourced
phenylacetylanhydro-penicillin (XII), itself obtainable from Penicillin G, since it
seemed to us to be outside the usual definition of a penicillin, and may be amenable
to elaboration of the 6-hydroxyethyl side chain and conversion of the five-membered
ring to the desired penem, retaining the sulfur atom.
Unfortunately at this time, toxicity and marketing concerns overtook Sch 34343 and
finally obliged Schering to abandon its penem program. As so often happens in
development work, the end came quickly with manpower and equipment resources
rapidly reallocated to other projects. It is usually difficult for all personnel, especially
chemists and engineers who have devoted themselves and their energies so totally
(often for years) to a project, to suddenly stop thinking and working on the intellectual
and company challenges associated with it and be expected to immediately pick
up the next API candidate with the same commitment. Management needs to be
especially aware of the need to express thanks and provide explanations, support, and
encouragement to employees during such transitions.
Back to Classical Cephalosporins
Several years later (1989), Schering–Plough licensed the oral cephalosporin,
Ceftibuten dihydrate (XIII), from Shionogi, Japan.
The contract included a sourcing agreement requiring Schering to purchase the bulk
Ceftibuten from Shionogi. Within a year of commencement of the project, the cost
of XIII had to be raised by Shionogi. This was partly due to Shionogi recognizing
the cost realities of their fairly long synthesis and partly due to amortization of their
$100 million investment in a Ceftibuten manufacturing plant in Kanegasaki. Our
Chemical Development group was called in to help in finding a better (lower cost)
process for producing Ceftibuten in collaboration with Shionogi. Clearly, we could
do little with the plant depreciation costs unless Shionogi could find a use for the
chemical processing equipment that might be idled by moving to a shorter synthesis.
The Shionogi process is outlined in Scheme 9.
The thrust of the Chemical Development program was to find a more economical
route starting with a low-cost cephalosporin instead of a penicillin. Such an
approach would allow us to eliminate the several steps taken by Shionogi to create
the cephalosporin ring from their penicillin G starting material (Scheme 9). A further
objective was to intersect the Shionogi synthesis by producing the key Shionogi
intermediate, diphenylmethyl 7-aminoceph-3-em-4-carboxylate (XIV). By employing
precisely the same last few process steps as used by Shionogi, we anticipated
eliminating any regulatory concerns regarding new impurities that could arise from
a different synthesis.
Our exploratory program, searching for a low-cost cephalosporin starting material,
ended with the selection of II derived from the process sequence developed
by Antibioticos S.p.A., Milan, Italy, starting with fermented cephalosporin C broth
(Schemes 1 and 2). To reiterate, Antibioticos had gained considerable commercial
advantage by creating and industrializing their Scheme II process to convert
cephalosporin C in filtered fermentation broth into the important cephalosporin starting
material, 7-amino-3-acetoxymethylceph-3-em-4-carboxylate (III, 7-ACA). By
the judicious application of macroreticular resins in the chromatographic purification
of cephalosporins in aqueous solution and the use of immobilized enzymes to carry
Penicillin G
1. Esterification
2. Oxidation
1. O3/CH2Cl2/
2. (MeO)3P
1. TsCl /
1. Br2/Pyr
< ?30°C
2. H2SO4
1. NaBH4
2. MsCl/Et3N
1. PCl5/Pyr
2. NaHCO3
CbzHN 1. AlCl3/anisole
2. Aq. HCl
3. Malic acid
4. Recrystalliztion
SCHEME 9. Shionogi synthesis of Ceftibuten.
out efficient transformations without isolating any intermediates, Antibioticos succeeded
in industrializing an all aqueous process for the manufacture of 7-ACA (see
footnote 8).
Antibioticos agreed to undertake a joint R&D program with Schering to explore
chemical synthesis options based on utilizing II, preferably in aqueous solution,
to produce XIV. We reasoned that by continuing to work in water for as long as
possible without isolating intermediates, we would maintain Antibioticos’ low-cost
processing philosophy. The question then became, What should the new derivative
of II be?
Initially we evaluated potential routes via the desacetyl derivative of II, namely
7-glutaroylamino-3-hydroxymethylceph-3-em-4-carboxylic acid. It was known that
several microorganisms deacetylate cephalosporins in high yield. However, evaluation
of the literature and review of the production plant requirements to adopt a route
to convert 3-hydroxymethyl cephalosporins to 3-H cephalosporins dissuaded us from
this option.28
In analyzing the literature, it seemed to me that a shorter and more economical
route for the conversion of II to XIV might be created if we could
improve on the 25-year-old electrochemical reduction method for converting 3-
acetoxymethylcephalosporins to 3-exomethylene-cephalosporins (e.g., XVI) and
thence to 3-hydroxycephalosporins (XV) via ozonolysis. Shionogi’s Dr. Mitsuru
Yoshioka and others pointed out that many people, notably at Takeda and Eli Lilly,
had published the results of their extensive efforts to achieve such an electrochemical
transformation but without commercial success.29 A review of this literature quickly
revealed that the major impediments to commercialization were as follows:
 The use of a mercury cathode
 The low substrate concentration (generally approximately 5 g/liter)
 The low current density
 The inefficiency of the process [the yield of their 3-exomethylene compound
was generally <70% with a ring-opened thiazoline as the primary byproduct
(up to 30%) and 3-methylcephalosporin (XVIII) (up to approximately 7%) as a
minor, but difficult to remove, secondary byproduct.
Although the prospects for success appeared daunting, we were encouraged to undertake
further evaluation by senior management (Dr. Hal Wolkoff). Our consultant,
Professor Sir Derek Barton, suggested that we engage Professor Charles R. Martin
[Colorado State University (CSU)] as an electrochemical consultant. Professor
Martin’s enthusiastic endorsement of our proposed program led us into three years
of fruitful collaboration with him through a funding program for several excellent
postdoctoral students, Drs. Haiyan Zhang, Vinod Menon, and Piotr Zelenay. Later on
as the project developed, we engaged the services of the Electrosynthesis Company
(ESC) (primarily Drs. David Genders and Guillermo Zappi, with further input from
Dr. NormanWeinberg). They carried out the definitive pilot plant work to validate the
28See Bernasconi, E., Genders, D., Lee, J., Longoni, D., Martin, C. R., Menon, V., Roletto, J., Sogli, L.,
Walker, D., Zappi, G., Zelenay, P., and Zhang, H. Org. Proc. Res. Dev., 2002, 6, 158.
29(a) Ochiai, M., Aki, O., Morimoto, A., Okada, T., and Shimadzu, H. J. Chem. Soc., Chem. Commun.,
1972, 800. (b) Ochiai, M., Aki, O., Okada, T., Shinozaki, K., and Asahi, Y., J. Chem. Soc., Perkin Trans.
I, 1974, 258. (c) Ochiai, M., Aki, O., Morimoto, A., Okada, T., Shinozaki, K., and Asahi, Y., Tetrahedron
Lett., 1972, 2341. (d) Ochiai, M., Aki, O., Morimoto, A., Okada, T., Shinozaki, K., Asahi, Y., and Masuda,
K. U.S. Patent 3,792,995, 1974 (to Takeda). (e) Hall, D. A. J. Pharm. Sci. 1973, 62, 980. (f) Hall, D. A.
U.S. Patent 4,042, 472, 1977 (to Eli Lilly). (g) Hall, D. A., Berry, D.M., and Schneider, C. J. J. Electroanal.
Chem., 1977, 80, 155.
S GluN
S GluN
S GluN
OH O2C(CH2)3
CO2H 0
XIX Glu = HO2C(CH2)3CO.
Path A
Path B
Path C
SCHEME 10. Electrochemical reduction of 7-glutaroylamino-3-acetoxymethylceph-3-em-
4-carboxylic Acid (II).
laboratory process. Most of the technical work carried out in Schering laboratories
was undertaken with great skill and dedication by Dr. Junning Lee. Dr. Lee was
also responsible for the day-to-day liaison with all outside laboratory operations (in
CSU, Antibioticos and ESC), a task carried out with thoroughness, imagination and
In beginning work in CSU laboratories, using II sourced from Antibioticos,
we quickly confirmed the process yield and byproduct profile results obtained by
Takeda’s Ochiai et al. and Eli Lilly’s Hall et al.29 (Scheme 10).
I reasoned that it may be possible to change the pathway for breakdown of radical
ion intermediates such as XVII by either changing the acetoxy leaving group in
II to one which would leave more readily (e.g., halide, SCN or pyridinium) or
by changing the electronic character of the sulfur atom (e.g., by oxidation to the
sulfoxide). The use of the sulfoxide of II as the electrochemical reduction substrate
proved to be most successful, eliminating both of the unwanted pathways B and
C in Scheme 10. Work at CSU on the other major practical objectives (increasing
1. XAD-16 purification
2. Extractive
with DDM
XX in water
1. Desulfoxidation
2. Reduction 1. Mesylation/demesylation
2. Amide cleavage
DPM = (C6H5)2CH ; DDM = (C6H5)2CN2 ; Glu = HO2C(CH2)3CO
SCHEME 11. Conversion of the sulfoxide of XVI (compound XX) to key intermediate XIV.
the substrate concentration and current density, and finding an alternative to the
mercury cathode) enabled us to identify tin as a promising cathode, and to increase the
substrate concentration to 50–100 g/liter and the current density to 120–200 mA/cm2.
Finally, ESC (suggestion of Dr. David Genders) showed that a high-surface-area tin
mesh cathode would give virtually the same result as mercury, thereby completing
the picture and enabling us to create a very efficient process (>90% yield) for
the production of the needed intermediate, the sulfoxide of XVI (i.e., XX) free
of previously process-compromising byproducts (the reduced solution contained no
XIX or XVIII sulfoxide). The process was validated by ESC on a pilot plant scale
using solutions of II sulfoxide obtained from Antibioticos’ manufacturing plant in
Italy.30We were thus able to keep Antibioticos’ low-cost processing philosophy going
for one more step by carrying out the electrochemical reduction very efficiently in
water. The stable sulfoxide XX may have been amenable to ozonolysis in water,
but we did not test this, preferring instead to undertake the reaction steps to desired
intermediate XIV according to Scheme 11.
In practical terms, the purification of aqueous solutions of XX on Rohm and
Haas macroreticular resin XAD-16 proved to be very efficient (95–98% recovery
of material with a purity of approximately 95%) and economical to carry out. This
chromatographic purification step removes salts (especially phosphates) introduced
at earlier processing steps in Antibioticos’ plant (Schemes 1 and 2), thereby enabling
us to minimize the amount of DDM needed to fully esterify both carboxyl groups of
XX; 2.3 to 2.5 moles DDM per mole of XX generally proved sufficient.31 Although
DDM is a relatively stable molecule (see footnote 20), its separate preparation does
introduce some safety concerns. In practical terms, it should be possible to prepare
30Chai, D., Genders, D., Weinberg, N., Zappi, G., Bernasconi, E., Lee, J., Roletto, J., Sogli, L., Walker,
D., Martin, C. R., Menon, V., Zelenay, P., and Zhang, H. Org. Process Res. Dev., 2002, 6, 178.
31Bernasconi, E., Lee, J., Sogli, L., and Walker, D. Org. Proc. Res. Dev., 2002, 6, 169.
1. (PhO)3PCl2
2. BuiOH
3. HCl
S Cl H.H2N
CO2DPM 85%
SCHEME 12. Outline for a process for the manufacture of a key intermediate for Cefaclor.
and use DDM in situ,32 but this option was not pursued in this initial phase of the
development of our ceftibuten process.
The ozonolysis of XXIwas straightforward and very high yielding. The indications
were that this step could have been carried out at a much higher temperature than
the –65?C temperature we used, since the substrate XXI was already in the sulfoxide
The sodium borohydride reduction of the C-3 enol XXII did not give good yields,
necessitating that the sulfoxide be first reduced to the sulfide before reduction. Several
desulfoxidation reagents were used successfully, notably Lawesson’s reagent, acyl
chloride/iodide, and PCl3. From this point we intersected the established Shionogi
chemistry, the only query being whether the trace impurity (XXIIIa) in the key
intermediate XIV prepared by the electrochemically based route would be as readily
removed as XXIIIb produced in the Shionogi synthesis (Scheme 9).
XXIIIa (C6H5)2CHO2C(CH2)3 , DPM = (C6H5)2CH
XXIIIb R = C6H5CH2 , DPM = (C6H5)2CH
It was, and the ceftibuten produced using the electrochemical route proved to be of
equivalent quality to that produced by Shionogi.
Despite the technical success, the process was a commercial failure in that it was
never adopted by Shionogi.33
We went on to show that the electrochemical reduction process afforded an elegant
route to a key cefaclor intermediate diphenylmethyl 7-amino-3-chloroceph-4-
carboxylate (Scheme 12). (see footnote 31).
32Bywood, R., Gallagher, G., Sharma, G. K., and Walker, D. J. Chem. Soc, Perkin Trans. I, 1975, 2019.
33Despite requiring only approximately $1.5 million in capital investment in electrochemical reduction
equipment, such expenditure was not considered worthwhile because the capacity of Shionogi’s existing
ceftibuten plant was considered adequate for future market projections, and the new process would require
additional resources for registration with the FDA and other agencies. In addition, Shionogi was not
convinced by our savings projections on the cost of goods. We discussed this matter many times over
lunch in the Gas Building in Osaka. When I asked for a change of lunch venue to the Electricity Building
(I did not know if one existed), Eichii Yamaguchi only smiled!
(Bun)4N F
CH2 3
(Pri)2 N Et
Me2S BH3
SCHEME 13. Schering–Plough synthesis35 of the ?-lactam Ezetimibe.
The project slowly died, a suitable analogy being the last passages of Haydn’s
Farewell Symphony.34
This excursion covers scientific “wanderings” through the mostly conventional
?-lactam antibiotic field—we did not participate in the even more exotic adventures
in the carbapenem field, which is undoubtedly worthy of historical review.
The ?-lactam antibiotics’ field continues to employ thousands of scientists worldwide,
if not the tens of thousands during its heyday in the 1960s through the 1980s.
34In which the musicians each snuff out the candle over their music stands when they complete their piece!
By the time I finished, even the audience had left!!!!
Undoubtedly, further ?-lactam antibioticswill be marketed to continue what is widely
perceived as a historic 70-year contribution to the welfare of mankind.
Nor are the contributions to be made by ?-lactam structures to the pharmaceuticals
field over. The extraordinary developments of recent years in cholesterol absorption
inhibitors (CAIs) based on the ?-lactam ring deserve some mention, especially
since Dr. T. K. Thiruvengadam in our chemical process development organization in
Schering–Plough is the brilliant architect of one of the patented syntheses of this novel
class of CAIs. His synthesis35 of Ezetimibe (Zetia), outlined in Scheme 13, serves
to (a) introduce this new direction in the development of novel ?-lactam medicinals
and (b) end this excursion in the field of ?-lactams.
This section describes the early history leading to creation of the steroid industry,
provides an account of key processes used in the manufacture of the oral contraceptives
and anti-inflammatories currently marketed in the United States and highlights
the creation of diverse biological activities by molecular manipulation of the steroid
molecule. The section concludes with an outline of a few of the synthesis challenges
faced and overcome in the manufacture of betamethasone anti-inflammatories.
Over the course of history, the ancients discovered extraordinarily diverse medicinal
and poison applications for extracts from the natural world. In the last hundred years,
most of the active principles of these extracts have been isolated and identified and
many shown to be structural variants of the steroid molecule (I).
4 5 6
14 15
The cucurbitacins (II), found in plants of the Cucurbitaceae family (gourds,
cucumbers, etc.) and also in Begoniaceae, Euphorbaceae, and others, have been used
35Thiruvengadam, T. K., Fu, X., Tann, C.-H., McAllister, T. L., Chiu, J. S., and Colon, C. U.S. Patent
6,207,822, 2001 (to Schering Corp.).
as emetics, narcotics, antimalarials, and anthelmintics.
Curcurbitacin B
Batrachotoxin A (III), just one of many steroidal alkaloids, is one of the most
lethal substances known (LD50 = 2µg/kg subcutaneously in mice). It is found in the
skin secretions of the brightly colored tropical frog Phyllobates aurotaenia and is
used by Colombian Indians to prepare poison darts.
Batrachotoxin A
Digoxin (IV) is in widespread medical use today as a cardiotonic. It is extracted
from the Foxglove, Digitalis lanata.
The steroid molecules II–IV illustrate the diverse biological activity which nature
has produced through biosynthetic processes. During the 20th century, man gained
a greater appreciation of the biological activity associated with the steroid molecule
and created an extraordinary range of new activities through structural manipulation
(see later). It is pertinent to add here that scientific investigation of biological processes
reached across many other fields including amino acids, peptides and proteins,
carbohydrates, alkaloids, and so on.
As knowledge grew, earlier more empirical approaches to treating disease gave
way to a more rational effort directed at understanding the biochemical processes
taking place in the human body. The isolation and purification of biologically active
molecules from human and other animal organs enabled scientists and physicians to
undertake painstaking degradative studies to determine their structure and begin to
piece together structure–activity relationships. In regard to steroids, in the 1920s and
1930s work on the sex hormones secreted by adrenal glands, gonads, and placenta
identified the principal estrogens (V–VII) and androgens (VIII–XI). Studies on the
ovarian organs identified the gestagens, with the principal one being progesterone
(XII) secreted during the menstrual cycle. Also, urine from pregnant animals was
found to contain (in addition to the estrogens) several pregnanes, notably pregnanediol
(XIII), pregnanedione (XIV), and related compounds.
Early in the 20th century, and continuing today, exploration of the mechanisms of
biochemical transformation occurring in man and other animals became an important
Enzyme-H Enzymes Enzyme-induced
shifts of H & CH3
Enzymes O2/NADPH
Squalene oxide Squalene
SCHEME 1. The biosynthesis of cholesterol.
branch of academic work. Understanding the pathways by which the steroidmolecule
is assembled provided a humbling appreciation of nature’s supreme elegance in
molecular assembly (Scheme 1). The biosynthesis of lanosterol and cholesterol attracted
the most attention. Many hypotheses proposed by major scientists (Robinson,
Woodward, Cornforth, Eschenmoser, Arigoni, and Stork) stimulated work that culminated
in a series of papers defining the steps occurring in cholesterol biosynthesis.36
The further degradation of cholesterol in the liver gives the bile acids:
Cholesterol inter alia +
Cholic Acid Deoxycholic Acid
36(a) Stork, G., and Burgstahler, A. W. J. Am.Chem.Soc., 1955, 77, 5068. (b) Eschenmoser, A., Ruzicka,
L., Jeger, O., and Arigoni, D. Helv. Chim. Acta, 1955, 38, 1890. (c) Van Tamelen, E. E., Willett, J. D.,
Clayton, R. B., and Lord, K. E. J. Am. Chem. Soc., 1966, 88, 4572. (d) Van Tamelen, E. E., Leopold, E.
J., Marson, S. A., and Waespe, H. R. J. Am. Chem. Soc., 1982, 104, 6479. (e) Van Tamelen, E. E. J. Am.
Chem. Soc., 1982, 104, 6480.
Ox bile, which contains cholic acid as its principal constituent, provided one of the
earliest mammalian sources of steroid raw materials for the commercial manufacture
of the androgens. In nature, cholesterol itself is the mammalian precursor of the
androgens, the biosynthesis passing through progesterone (XII).
Progesterone Testosterone
In the 1930s, recognition of the role progesterone was playing in the development
of a fetus and in inhibiting ovulation, thereby preventing pregnancy, stimulated only
weak commercial interest. Themain reason lay in the great difficulty and astronomical
cost of producing progesterone. Early supplies were made from pregnanediol (ex
mammalian pregnancy urine)37 and stigmasterol (XVI, ex soy and calabar beans).38
Some progesterone was also produced from cholic acid39 and some from cholesterol,
available from nonsaponifiable animal matter, including wool fat. However, at
the time, nature’s exquisite enzyme-mediated transformation of the C-17 side chain
of cholesterol could not be reproduced or mimicked using chemical means. Oxidative
techniques, including oxidative degradation, were widely employed in producing
mixtures of steroids from which progesterone was recovered in low yield. The work
needed can only be described as heroic. As a result, the cost of progesterone was
commercially prohibitive.
It took several years and a maverick chemist with new ideas on other steroid
sources, and a towering commitment to prove them, to achieve a breakthrough. In
37Butenandt, A., and Schmidt, J. Berichte, 1943, 67, 1901.
38Butenandt, A., Westphal, U., and Cobler, H. Berichte, 1934, 67, 1611, 1903, and 2085.
39Zondek, B., and Bergman, E. U.S. Patent 2,314,185 (1943).
Ac2O CrO3
Boiling Ac2O
SCHEME 2. The Marker degradation41a,b of sarsasapogenin.
1938, Russell Marker, working at State College, Pennsylvania, with funding from
the Parke Davis Co., proposed an alternative formula40(a) for the steroid sapogenin,
sarsasapogenin (XVII, isolated from the Sarsaparilla root by hydrolysis of the C-3
glycosidic side chain), to the structures proposed by Power40(b) and Jacobs.40(c,d)
Based on his new structure, Marker reasoned that the C-17 side chain should be
much more amenable to chemical degradation than that of cholesterol. He went on to
prove his hypothesis creating the “Marker degradation,” a process still in commercial
use. The Marker degradation is outlined in Scheme 2. Unfortunately, at the time,
obtaining sarsasapogenin proved an expensive proposition leading Marker to search
for another plant source of steroids of the sapogenin class. This soon led him to
identify a wild Mexican yam of the Dioscoreaceae family (known as cabeza de negro
in Mexico), which proved to be a useful source of diosgenin (XVIII). [Later (1949)
a yam (barbasco) with a higher diosgenin content (up to 5% on a dried basis) was
discovered and commercialized.]Marker collected cabeza de negro roots in Veracruz
state which were “lost” in transport. One 50-pound root was recovered by dint of
bribing a policeman and surreptitiously spirited back to Pennsylvania where he used
his degradation process41(c) to access the intermediate needed to refine his synthesis
of progesterone (XII) – (Scheme 3).
40(a) Marker, R. E., and Rohrmann, E. J. Am. Chem. Soc., 1939, 61, 846. (b) Power, F. B., and Salway, A.
H. J. Chem. Soc., 1914, 105, 201. (c) Jacobs, W. A., and Simpson, J. C. E. J. Biol. Chem. 1934, 105, 501.
(d) idem ibid, 1935, 109, 573.
41(a) Marker, R. E., and Rohrmann, E. J. Am. Chem. Soc., 1939, 61, 3592. (b) idem ibid., 1940, 62, 518,
521, 896, and 898. (c) Marker, R. E., Tsukamoto, T., and Turner, D. L. J. Am. Chem. Soc., 1940, 62, 2525.
reduces the
double bond
and C-20
to protect
CrO3 Zn/HOAc
SCHEME 3. Marker process for the manufacture of progesterone.
Although the sponsor of Marker’s research at State College obtained U.S.-only
patents on his work,42 Marker was unable to get support from the pharmaceutical
industry to undertake the commercialization of his process for the manufacture of
progesterone. He eventually found backing in Mexico City, founded Syntex, and
began production in 1944. Following a financial dispute with his backers in 1945, he
quit, taking his process secrets with him. Syntex hired Dr. George Rosenkranz, who
was able to rediscover the process and restart production in a fewmonths. Rosenkranz
subsequently was the driving force in building Syntex into a scientifically powerful
steroid research company in Mexico City. Syntex was responsible for the discovery
and development of some of the most important progestogens, which, when mixed
with an estrogen, were marketed as oral contraceptives—for example, structures
XIX and XX. The 3-methyl ether of estrogen XX, Mestranol, is also used in some
formulations. Indeed a Mestranol and Norethinodrel [the 5(10) double-bond isomer
of XIX] formulation was the first, albeit short-lived, oral contraceptive on the market
(Enovid, G.D. Searle).
17?-Ethinyl estradiol
Several other companies were quickly attracted into the field, notably G. D. Searle,
Roussel–UCLAF, Wyeth, Philips–Duphar, Organon, Upjohn, and Merck, each pursuing
the synthesis of new molecules. The most important consideration was to find
novel patentable structures derived from their positions in their chosen starting raw
42(a) Marker, R. E. U. S. Patent 2,223,377, 1940 (to Parke Davis & Co.). (b) Marker, R. E. U. S. Patent
2,352,852, 1944 (to Parke Davis & Co.) Marker’s failure was also despite his prescient observation that
Mexican women had been eating yams of the Dioscorea genus for centuries for contraception.
materials. Roussel–UCLAF and Merck founded their work on the bile acids. Ingeniously,
Philips–Duphar found a niche based on lumisterol, a so-called retrosteroid
prepared by UV irradiation of a benzene–alcohol solution of ergosterol, the most
important of the provitamins D. Ergosterol was itself obtained from yeast. Ultraviolet
irradiation of ergosterol inverts C-10 methyl and C-9H.43
Ergosterol Lumisterol
275–300 nm
Vitamin D2
Retroprogesterone Dydrogesterone
Dydrogesterone44 still has a small market today.Wyeth, being a later starter, founded
its business on total synthesis (see later).
Other producers of steroid contraceptives started with diosgenin, tigogenin (a 5?,
6-dihydroderivative of diosgenin) or hecogenin (a 12-keto derivative of tigogenin).
The last two compounds are both obtained from numerous species of the Agave plant.
More recently, sitosterol (from the soy bean industry) and cholesterol have become
attractive as starting raw materials, largely due to improvements in removing the
C-17 side chain using methods that create a 17-carbonyl functionality.
The contraceptive field received the most attention in the 1940s and early 1950s.
A second major development in the steroid field was triggered by the 1949 announcement
by Hench et al.45 at the Mayo Clinic that cortisone greatly relieved the ravages
of rheumatoid arthritis.
Merck and Co. had produced the cortisone used by Hench in research quantities using
a 36-step synthesis from bile acids. Hench’s revelation rapidly escalated interest in
steroidal anti-inflammatories. A brief outline of ongoing developments in each of
these fields follows.
43Askew, F. A., Bourdillon, R. B., Bruce, H. M., Callow, R. K., Philpot, J. St. L. and Webster, T. A. Proc.
Roy. Soc. (London), 1932, B109, 488.
44(a) Reerink, E. H., Westerhof, P., and Scholer, H. F. L., U.S. Patent 3,198,792,1962 (to North American
Philips). (b) Westerhof, P. and Reerink, E. H., Rec. Trav. Chem., 1960, 79, 771.
45Hench, P. H., Kendall, E. C., Slocumb, C. H., and Polley, H. F., Ann. Rheum. Dis., 1949, 8, 97.
1. LiAlH4
2. Li/NH3
3-Methyl ether
of Estrone
Overall yield from Estrone
= 70–77%
SCHEME 4. Conversion of the 3-methyl ether of estrone to 19-nortestosterone.
Contraceptives. The realization that progesterone was only weakly active when
administered orally, stimulated vigorous programs both to overcome this disadvantage
and also to find more active progestational steroids.46 One of the earliest successes
was the discovery47 by the Syntex group that 19-norprogesterone was four to eight
times more active than progesterone as a progestational hormone. Shortly thereafter,
this observation led to the identification of 19-nor-17?-ethinyltestosterone (XIX,
The key chemical step in the creation of starting materials for the production
of the 19-nor series of compounds resulted from the pioneering work of Birch and
Mukherji49(a) who reduced the 3-glyceryl ether of estradiol to 19-nortestosterone
with sodium or potassium. Wilds and Nelson49(b) improved the original procedure
and created the basis of a commercial process by utilizing the 3-methyl ether of estrone
and reducing the aromatic ring using lithium and liquid ammonia (Scheme 4).
46However, progesterone, in a finely micronized form, is marketed (Prometrium) in large-dose capsules
(100 mg and 200 mg) for the treatment of somemenstrual conditions.Micronization increases the bioavailability
of insoluble APIs (see Chemical Engineering).
47(a) Miramontes, L., Rosenkrantz, G., and Djerassi, C. J. Am. Chem. Soc., 1951, 73, 3540. (b) idem
ibid., 1953, 75, 4440. (c) Tullner, W. W., and Hertz, R., J. Clin. Endocrinology Metab., 1952, 12, 916. (d)
Interestingly, early work by Dirscherl et al. (Dirscherl,W., Z. Physiol. Chem., 1936, 239, 53 and Dirscherl,
W., Kraus, J., and Voss, H. E., Z. Physiol. Chem. 1936, 241, 1) demonstrated that hydrogenated products
of estrone (i.e.19-nor steroids) possessed some androgenic activity. Marker, R. E., and Rohrmann, E. J.
Am. Chem. Soc., 1940, 62, 73 also described the hydrogenation of estrone and the preparation of relatives
of 19-nor-testosterone and 19-nor-androstenedione.
48Djerassi, C., Miramontes, L., Rosenkranz, G., and Sondheimer, F. J. Am. Chem. Soc., 1954, 76, 4092.
Almost simultaneously, the G. D. Searle group identified Norethynodrel [the 5(10) double-bond isomer of
Norethindrone]: Colton, F. B. U.S. Patents, 2,691,028, 1954 and 2,725,389, 1955 (to G. D. Searle).
49(a) Birch, A. J., and Mukherji, S. M. J. Chem. Soc., 1949, 2531. (b) Wilds, A. L., and Nelson, N. A. J.
Am. Chem. Soc., 1953, 75, 5366.
Me O
600oC in
mineral oil
H2/Pd to
reduce ?6
Marker degradation
Scheme 3
H2NOH Beckmann
Me2SO4 Li/NH3 CrO3
O Me
K OAmyl/HC
SCHEME 5. Outline of Syntex Process for the Manufacture of Norethindrone.
In the early days, estrone was recovered from the urine of pregnant mares, but this
source proved unattractive because of the difficulty of separating it from related
compounds, notably equilenin (XXI) and equilin (XXII).
Later the estrone produced for commercial purposes was obtained from diosgenin.
Diosgenin was first oxidized to its 1,4,6-triene-3-one derivative, which was aromatized
by pyrolysis at 500–600?C,50 subjected to the Marker degradation and the
17-acetyl group removed by Beckmann rearrangement of the 20-oxime. The overall
Syntex process51 for the manufacture of Norethindrone from diosgenin (XVIII) is
outlined in Scheme 5.
Today, the major oral contraceptives on themarket in the United States are mixtures
of the estrogen, 17?-ethinylestradiol (XX) and one of the progestogens listed in
Table 1. The first two progestogens have also been formulated with mestranol, the
3-methyl ether of XX.
50The aromatization reaction was first applied to steroids by Inhoffen, H. H., U.S. Patent 2,361,847, 1944
(to Schering Corp.) and later greatly improved by Hershberg, E. B., Rubin, M. and Schwenk, E. J. Org.
Chem. 1950, 15, 292. The Hershberg work provided the foundation for the industrial process.
51Sondheimer, F., Neumann, F., Ringold, H. J., and Rosenkranz, G. J. Am. Chem. Soc., 1954, 76, 2230. (b)
Rosenkranz, G., Mancera, O., Sondheimer, F., and Djerassi, C. J. Org. Chem., 1956, 21, 520. (c) Djerassi,
C., Miramontes, L., and Rosenkranz, G. U.S. Patent 2,744,122, 1956 (to Syntex).
TABLE 1. Progestogen Components of Most-Prescribed Oral Contraceptives in the United
Compound Name Mixture Namea Original Company
Brevicon Syntex13
Demulen G. D. Searle52
(–) form
Norgestrel Alesse Wyeth53
Desogestrel Desogen Organon54
Norgestimate Ortho Cyclen Ortho55
Source: Physicians Desk Reference, 55th edition, Medical Economics Company, Inc. Montvale, New
Jersey, 2001.
aMany other market names exist for the mixture, depending on the formulation, cross-licensing, and so
It is pertinent to note that market forces eventually favored the 19-nor progestogens
of Table 1 over the 19-methylprogestogens, at least in the United States.56
52Colton, F. B. U.S. Patent 2,843,609, 1958 (to G.D. Searle).
53(a) Smith, H. Belgian Patent 623,844,1963 (Chem. Abstr., 1964, 61, 4427). (b) Smith, H., Hughes, G.
H., Douglas, G. H., Wendt, G. R., Buzby, G. C., Edgren, R. A., Fisher, J., Foell, T., Gadsby, B., Hartley,
D., Herbst, D., Jansen, A. B. A., Ledig, K., McLoughlin, B. J., McMenamin, J., Pattison, T. W., Phillips,
P. C., Rees, R., Siddall, J., Siuda, J., Smith, L. L., Tokolics, J., and Watson, J. H. P. J. Chem. Soc., 1964,
4472. (c) Hughes, G. H., and Smith, H. U.S. Patent 3,959,322, 1976 (to Herchel Smith)—patent original
priority date August 1964!
54(a) van den Broek, A. J.; van Bokhoven, C., Hobbelen, P.M. J., and Leemhuis, J. Rec. Trav. Chim. 1975,
94, 35. (b) van den Broek, A. J. U.S. Patent 3,927,046, 1975 (to Akzona, Inc.).
55Schroff, A. P. U.S. Patent 4,027,019, 1977 (to Ortho).
56Several 19-methyl compounds, derived from plant starting materials, were marketed in combination
with an estrogen such as XX. Some of the best known were Provest (Medroxyprogesterone
(Example 43*)
(Example 68*)
(Example 81*)
Li/NH3 in
(Example 103*)
NH3/Li Foil
(Example 107*)
Al(OPr i)3/Toluene
Cyclohexan one
(Example 116*)
(Example 121*)
(Example 170*)
( ) Norgestrel ±
Et Et
*Example numbers are from U.S. Patent 3,959,322,1976 [see footnote 53 (c)].
SCHEME 6. Outline of Smith and Hughes’ (Wyeth’s) synthesis of Norgestrel (see
footnote 53).
The early successes of the Syntex group in producing orally active contraceptives
from the 19-nor steroids attracted several companies into the field. All the competing
compounds introduced by the rival companies also carry the 17?-hydroxy-17?-
ethinylmoiety and owe theirmarketing to their patented distinctions from the original
Norethindrone. Probably the most heroic achievement in the 19-norsteroid field resulted
from the work of Dr. Herchel Smith (University of Manchester and laterWyeth
Laboratories, Inc.), who, with his co-workers and particularly Dr. Gordon Hughes,
pioneered the commercialization of steroid manufacture by total synthesis. Dr. Smith
was uniquely placed in Manchester being a prot?eg?e of Professor A. J. Birch, whose
Birch reduction process49(a) instigated the work leading to the Syntex process for the
manufacture of Norethindrone.
Total synthesis, although disadvantageous in requiring optical resolution to create
the most biologically active compounds, has the advantage of enabling a wide range
of structural changes to be made which could not be easily done startingwith naturally
occurring steroids. Total synthesis thus increased the opportunity to create enhanced
biological activity and, very important, to create an essential patent portfolio. Dr.
Smith brilliantly and opportunistically rose to the occasion with a synthesis of the
progestogen, Norgestrel (XXIV), marketed by Wyeth as Alesse. An outline of this
synthesis is given in Scheme 6.
acetate, or 17?-acetoxy-6?-methyl-progesterone), Gestafortin (Chlormadinone acetate, or 17?-acetoxy-
6-chloro-6,7-dehydroprogesterone), and Ervonum (Megestrol acetate, or 17?-acetoxy-6,7-dehydro-6-
From the outset, Dr. Smith and his large team of co-workers57 focused on alternatives
to the C-13 methyl group in natural steroids and on introducing the acetylene
moiety at C-17 following the leads of G. D. Searle and Syntex (see footnote 48).
In producing the (?) form of Norgestrel optical resolutions were carried out as
early in the synthesis as possible, with such as compound XXVII being a favored
starting point.Both chemical resolution (of the hemisuccinate58) and dehydrogenation
methods59 were used for resolution.
Adecade later, Organonworkers discovered the 13?-ethyl compound, Desogestrel
XXV (see footnote 54). This molecule is interesting in demonstrating that significant
structural change can be made without affecting progestogen activity (XXV lacks a
3-oxo substituent and carries an 11-exomethylene group; however, the biologically
active metabolite is the 3-oxo compound). Because of its unique structural features,
the Desogestrel molecule is more costly to produce than others in Table 1. Starting
materials are such as Birch’s 19-nortestosterone or Smith and Hughes’ 13-ethyl
analogue of 3-0-methyl estradiol (XXVII). In order to introduce Desogestrel’s unique
features, the Organon workers employed both standard and some novel chemistry.
(a) Conversion of 10-methyl to 10-ethyl:
MeMgX Wolff
(b) Conversion of 3-keto to 3-CH2:
(c) Conversion of 11?-HO to 3-exomethylene:
57Smith, H., Hughes, G. A., Douglas, G. H., Hartley, D., McLoughlin, B. J., Siddall, J. B., Wendt, G.
R., Buzby, G. C., Herbst, D. R., Ledig, K. W., McMenamin, J. R., Pattison, T. W., Suida, J., Tokolics, J.,
Edgren, R. A., Jansen, A. B. A., Gadsby, B., Watson, D. H. R., and Phillips, P. C. Experientia, 1963, 19,
58Buzby, G. C., Hartley, D., Hughes, G. A., Smith, H., Gadsby, B.W., and Jansen, A. B. A. J. Med. Chem.,
1963, 10, 199.
59Smith, L. L., Greenspan, G., Rees, R., Foell, T., and Alburn, H. E. J. Org. Chem. 1966, 31, 2512.
Those seeking more depth on the total synthesis of steroids are referred to the very
informative book by Blickenstaff et al.60
In concluding this section, it is pertinent to point out that improvements made in
the total synthesis and optical resolution of steroids over the last three decades have
greatly reduced the costs of manufacturing all the steroids in Table 1 made via the
total synthesis route. Today, the cost of most steroids made by the total synthesis
route are comparable with the costs of those produced from plant sources.
Anti-inflammatories. Identification of the field as worthy of commercial pursuit
arose from findings in academia, beginning in the 1930s, that extracts from the
adrenal glands possessed hormone activity. One of these, found by Kendall (Mayo
Foundation),was known asKendall’s compound E (later cortisone). Several industrial
organizations, notably Organon, Upjohn, Schering A.G., and Merck, sensed commercial
potential and were helpful to academia in producing extracts from animal organs
using their large-scale equipment. In 1942, the National Research Council (NRC) in
theUnited States sponsored a research program with the visionary objective of finding
a method for the synthesis of enough cortisone to evaluate its possible application to
medicine. Although the war hampered European collaboration, a corresponding program
was also going on in the ETH in Zurich under the guidance of Professor Tadeus
Reichstein. The NRC program led to the finding by Hench and co-workers that cortisone
was effective in relieving the symptoms of arthritis (see footnote 45). As already
indicated, the bile acids were the major source of raw materials for the cortisone
supplied to Hench by Merck. Immediately following Hench’s revelations, Roussel in
Paris and Schering in New Jersey61 took out licenses from the Research Corporation
(set up to manage the intellectual property generated during NRC studies) to utilize
cholic acids in their own programs for the production of cortisone.
Bile acids were the logical starting point for the manufacture of cortisone in the
late 1940s, but steroids from plant sources, and particularly diosgenin ex yams, were
gaining credibility as time passed. The chemical synthesis challenges associated with
each of these raw materials are outlined in Scheme 7.
From a commercial point of view, one of the most important breakthroughs favoring
the route from diosgenin resulted from the work of Durey Peterson in Upjohn.
Peterson and co-workers (following the precedent set by Alexander Fleming?) collected
a culture of Rhizopus arrhizus on an agar plate left out on a window sill
in Kalamazoo and found that it would hydroxylate progesterone in the 11? position.
62 This led them to Rhizopus nigricans which hydroxylated the 11? position
of progesterone in over 80% yield.62(b) Other companies found 11? –hydroxylating
60Blickenstaff, R. T., Ghosh, A. C., and Wolf, G. C. Total Synthesis of Steroids (Organic Chemistry, Vol.
30), Academic Press, New York, 1974.
61Schering (U.S.) became a separate U.S. Company resulting from American seizure of Schering A.G.
assets during World War II.
62(a) Peterson, D. H., and Murray, H. C. J. Am. Chem. Soc., 1952, 74, 1871. (b) Peterson, D. H., Murray,
H. C., Eppstein, S. H., Reineke, L. M., Weintraub, A., Meister, P. D., and Leigh, H. M. J. Am. Chem. Soc.
1952, 74, 5933. (c) Eppstein, S. H., Meister, P. D., Peterson, D. H., Murray, H. C., Leigh, H. M; Lyttle, D.
A., Reineke, L. M., and Weintraub, A. J. Am. Chem Soc., 1953, 75, 408.
3 7
Selective conversion of 7-OH to 7-H
Use of 12-OH to create 11-OH and 11-keto
Conversion of C-20 side chain to C-20 keto
Hydroxylation at C-17
Hydroxylation at C-21
Conversion of 3-OH to?4-3-keto
Cholic Acid
Bile Acids
Marker degradation to Progesterone (see Scheme 3)
Hydroxylation at C-17
Hydroxylation at C-21
conversion to11-keto
Selective hydroxylation at C-11 and
SCHEME 7. Structural manipulations for the conversion of bile acids and diosgenin.63
microorganisms of their own (Searle discovered Aspergillus strains and Schering
discovered Metarrhizium). New technology, as well as increased plant capacity and
competition, drove the cost of cortisone down by a factor of almost 100 in the next
Vigorous R&D programs were initiated by many companies in Europe and the
United States with the objective of finding and patenting compounds with greater
activity and fewer side effects than cortisone or its acetate. Oral administration of these
two compounds was found to cause undesirable side effects such as salt retention.
Ointment and other topical dosage forms looked to be preferred ways of delivering
these drugs, but considerable variability in activity was observed: Cortisone acetate,
for example, was inactive in topical dosage whereas hydrocortisone acetate (carrying
an 11?-carbinol group in place of cortisone’s 11-keto group and the same 21-acetate)
was active in both topical and oral forms.
Further evaluation of minor compounds in the extracts of animal organs did not
provide any leads to suggest that structural variants with greater activity or fewer side
effects than observed with cortisone might be found. It appeared that conformation
63For a detailed account of the many years of work taken to achieve a working process to produce
cortisone from bile acids, see Fieser, L. F., and Fieser, M. Steroids, Reinhold Publishing Co., New York,
1959, Chapter 19.
of the cortisone and hydrocortisone molecules was vital to their activity. However,
more active compounds did emerge as a result of peripheral observations.64 Fried,
working in Squibb, in the course of studies on the conversion of epicortisol (the 11?-
hydroxy analogue of hydrocortisone) into hydrocortisone, prepared the 9?-bromo
11?-hydroxy compound, XXVIII, by the addition of hypobromous acid to the 9,11
olefine, XXIX. Fried speculated that the addition reaction should have given the 9?-
bromo-11?-hydroxy compound and thought that if the orientation of the hydroxyl
group was indeed 11?, as in hydrocortisone acetate, then compound XXVIII might
show some weak antiarthritic activity.65
To the Squibb group’s surprise, XXVIII proved to have almost one-third of the activity
of cortisone acetate. This revelation set off an examination of other halohydrins. The
9?-fluoro analogue of XXVIII proved to be almost 11 times as active as cortisone
acetate, and the 9?-chloro analogue was almost five times as active.
However, the new compounds still possessed salt retention disadvantages. Nevertheless,
the Squibb groups findings, coupled with other beneficial discoveries (see
below), set off an enthusiastic stampede for superior anti-inflammatory steroids.
One of the other major contributions leading to improved biological properties
emerged from work carried out by Hershberg and co-workers in Schering Corporation,
New Jersey. They were interested in finding microorganisms that would
selectively hydrolyze the diacetate ester of hydrocortisone. Attempts to achieve this
using Corynebacteria led to a new compound proven to be the 1 dehydro derivative
of hydrocortisone diacetate. Insignificant degradation of the C-17 side chain was
observed. The Schering workers promptly subjected cortisone and hydrocortisone to
the new microbial technique, creating prednisone (XXX) and prednisolone (XXXI)
(Scheme 8).66 Prednisone and prednisolone proved to be superior to cortisone and
64"Chance," Pasteur once said, “only visits the prepared mind.”
65(a) Fried, J. and Sabo, E. F., J. Am. Chem. Soc 1953, 75, 2273. (b) idem ibid., 1954, 76, 1455. (c) Fried,
J., Thoma, R. W., Perlman, D., Herz, J. E., and Borman, A. Recent Progress in Hormone Research, 1955,
11, 149.
66(a) Herzog, H. L., Nobile, A., Tolksdorf, S, Charney, W., Hershberg, E. B., and Perlman, P. L., Science,
1955, 121, 176. (b) Nobile, A., Charney, W., Perlman, P. L., Herzog, H. L., Payne, C. C., Tully, M. E.,
Jevnik, M. A., andHershberg, E.B. J. Am.Chem. Soc., 1955, 77, 4184. (c)Nobile, A. U.S. Patent 2,837,464,
1958 (to Schering Corporation). (d) Nobile, A. U.S. Patent 3,134,718, 1964 (to Schering Corporation).
It should be noted that Schering’s publications and patents were delayed because of the time required
Cortisone X, Y = O
Hydrocortisone X = H, Y = OH
Prednisone (XXX) X, Y = O
Prednisolone(XXXI)X = H, Y = OH
SCHEME 8. Microbiological dehydrogenation of cortisone and hydrocortisone.
hydrocortisone in both antiarthritic activity and particularly in greatly minimizing
salt retention (mineralocorticoid activity).67
Many other structural variants were pursued, the most beneficial being the introduction
of fluorine or methyl at the 6? position and, later, methyl at the 16? or 16?
position. Considerable improvements in biological activity were also created by esterification
of hydroxyl substituents at C-17 and C-21. Some aspects of the synthesis
of the 16-methyl glucococorticoids are detailed later in a summary of some of the
work undertaken by my colleagues in Schering.
Table 2 lists the most-prescribed anti-inflammatory steroids on the U.S. market
today. Table 2 shows that some of the first compounds to be marketed still have a
place in the treatment of inflammatory disease 50 years and more after their discovery.
Table 2 also illustrates that continued effort to improve potency, to increase safety and
to find better drug delivery systems (e.g., inhalers) succeeded in creating marketable
products into the 1990s. Today very little, if any, research is going on to find improved
steroid anti-inflammatory drugs. Some work continues in the pharmaceutical
development area mostly to improve the formulation of existing molecules and to
find better delivery systems.
It should be noted that many market names are used for all of the antiinflammatories
listed in Table 2, depending on market liaisons, generic competition,
derivatives such as esters, and the form of the drug product (e.g., topical, inhaler, nasal
spray, etc.). For example, Schering’s betamethasone 17,21-dipropionate is marketed
as Diprosone, and Glaxo’s betamethasone-17-valerate was marketed as Betnovate.
With a few exceptions the main structural features needed for anti inflammatory
activity are the 1,4-3-ketone structure, hydroxyl at C-11? and C-17? and hydroxyacetyl
at C-17?. Esterification of either or both C-17 and C-21 hydroxyl groups
is a common feature of all the anti-inflammatories produced after the mid-1960s.
Discovery of the importance of esterification was largely due to work at Glaxo. This
work is worth highlighting since it provides an early example of the importance of
to establish that Schering had the priority in invention over competing claims by others, notably Merck,
Upjohn, Syntex, Squibb, and Pfizer.
67Prednisone and prednisolone are some four times more active at glucocorticoid than at mineralocorticoid
TABLE 2. Most-prescribed Anti-inflammatories on the U.S. Market
Compound Name Market Name
Cortisone acetate Cortone acetate Merck68
Prednisone Prednisone Schering66
Prednisolone Prednisolone Schering66
68Sarett, L. H. J. Biol. Chem., 1946, 162, 601.
69Wendler, N. L., Graber, R. P., Jones, R. E., and Tishler, M. J. Am. Chem. Soc. 1950, 72, 5793.
70(a) Bernstein, S., Lenhard, R. H., Allen, W. S., Heller, M., Littell, R., Stolar, S. M., Feldman, L. I., and
Blank, R. H. (American Cyanamid/Lederle). J. Am. Chem. Soc., 1956, 78, 5693. (b) Fried, J., Borman, A.,
Kessler, W. B., Grabowich, B., and Sabo, E. F. (Squibb). J. Am. Chem. Soc., 1958, 80, 2338. (c) Lederle
(American Cyanamid) became the patent holder for the joint marketing venture through its U.S. Patents,
as follows: Bernstein, S., Lenhard, R. H., and Allen,W. S. U.S. Patent 2,789,118, 1957; Bernstein, S., and
Allen, G. R., Jr., U.S. Patent 2,990,401,1961 and Allen, G. R. Jr., Marx, M., and Weiss, H. J. U.S. Patent
TABLE 2. (Continued)
Compound Name Market Name
Methylprednisolone Medrol Upjohn71
Synalar Syntex72
Dexamethasone Decadron
Betamethasone Celestone
71(a) Spero, G. B., Thompson, J. L., Magerlein, B. J., Hanze, A. R., Murray, H. C., Sebek, O. K., and
Hogg, J. A. J. Am. Chem. Soc., 1956, 78, 6213. (b) Sebek, O. K., and Spero, G. B. U.S. Patent 2,897,218,
1959 (to Upjohn).
72Mills, J. S., Bowers, A., Djerassi, C., and Ringold, H. J. J. Am. Chem. Soc. 1960, 82, 3399. Bowers, A.,
and Mills, J. S. U.S. Patent 3,014,938, 1961 (to Syntex).
73(a) Arth, G. E., Fried, J., Johnston, D. B. R.; Hoff, D. R., Sarett, L. H., Silber, R. H., Stoerk, H. C., and
Winter, C. A. (Merck). J. Am. Chem. Soc. 1958, 80, 3161. (b) Oliveto, E. P., Rausser, R., Nussbaum, A.
L., Gebert, W., Hershberg, E. B., Tolksdorf, S., Eisler, M., Perlman, P. L., and Pechet, M. M. (Schering).
J. Am. Chem. Soc., 1958, 80, 4428. (c) Oliveto, E. P., Rausser, R., Weber, L., Nussbaum, A. L., Gebert,
W., Coniglio, C. T., Hershberg, E. B., Tolksdorf, S., Eisler, M., Perlman, P. L., and Pechet, M. M. J. Am.
Chem. Soc., 1958, 80, 4431.
74(a) Taub, D., Hoffsommer, R. D., Slates, H. L., andWendler, N. L. (Merck). J. Am. Chem. Soc., 1958, 80,
4435. (b) Oliveto, E. P., Rausser, R., Herzog, H. L., Hershberg, E. B., Tolksdorf, S., Eisler, M., Perlman,
P. L., and Pechet, M. M. (Schering). J. Am. Chem. Soc., 1958, 80, 6688.
TABLE 2. (Continued)
Compound Name Market Name
Desonide Locapred
Flunisolide Nasalide Syntex76
Desoximetasone Topicort Roussel77
Beconase Glaxo78
Budesonide Rhinocort Bofors79
75(a) Bernstein, S., Littel, R., Brown, J. J., and Ringler, I. J. Am. Chem. Soc., 1959, 81, 4573. (b) Bernstein,
S., and Allen, G. R., Jr. U.S. Patent 2,990,401, 1961 (to American Cyanamid/Lederle).
76Ringold, H. J., and Rosenkranz, G. U.S. Patent 3,124,571, 1964 (to Syntex).
77Jolly, R., Warnant, J. and Goffinet, B. U.S. Patent 3,099,654 (to Roussel—UCLAF).
78Elks, J., May, P. J., and Weir, N. G. U.S. Patent 3,312,590, 1967 (to Glaxo).
79(a) Thalen, B. A., and Brattsand, R. L. Arzneimittel-Forsch, 1979, 29, 1787. (b) Brattsand, R. L.,
Thuresson, B., Claeson, K. G., and Thalen, B. A. U.S. Patent 3,929,768, 1975 (to Bofors).
TABLE 2. (Continued)
Compound Name Market Name
Temovate Glaxo80
Cloderm Schering A. G.81
Aclovate Schering82
Cutivate Flonase Glaxo83
80Elks, J., May, P. J., and Weir, N. G. U.S. Patent 3,312,590, 1967 (to Glaxo).
81Kasper, E., and Phillipson, R. U.S. Patent 3,729,495, 1973 (to Schering A.G.).
82(a) Green, M. J., Berkenkopf, J., Fernandez, X., Monahan, M., Shue, H-J., Tiberi, R. L., and Lutsky, B.
N. J. Steroid Biochem., 1979, 11, 61. (b) Green, M. J., Shue, H.-J.; Shapiro, E., and Gentles, M. A. U.S.
Patent 4,076,708, 1978 (to Schering).
83(a) Phillipps, G. H., Bailey, E. J., Bain, B.M., Borella, R. A., Buckton, J. B., Clark, J. C., Doherty, A. E.,
English, A. F., Fazakerley, H., Laing, S. B., Lane-Allman, E., Robinson, J. D., Sandford, P. E., Sharratt, P.
J., Staples, I. P., Stonehouse, R. D., and Williamson, C. J. Med. Chem., 1994, 37, 3717. (b) Phillipps. G.
H.; Bain, B. M., Steeples, I. P., and Williamson, C. U.S. Patent 4,335,121, 1982 (to Glaxo).
TABLE 2. (Continued)
Compound Name Market Name
O Mometasone
Alrex Otsuka85
Source: Physicians Desk Reference, 55th edition, Medical Economics Company, Inc., Montvale, New
Jersey, 2001.
finding an assay to identify the effects of structural change on biological activity—
but first a little background on Glaxo’s steroid interests.
Glaxo was one of several European companies involved in finding a route to
cortisone. Initially, Glaxo worked with ergosterol as a starting material but soon
switched to hecogenin (XXXIV) found in juices discarded by the manufacturers of
sisal fiber in East Africa.
O Me
Hecogenin, like the cholic acids, possesses a C-12 oxygen function and was regarded
as the best available prospect for the preparation of cortisone. Glaxo’s relationship
with Schering86 and its position in hecogenin encouraged Schering to arrange a
84(a) Shapiro, E. L., Gentles, M. J., Tiberi, R. L., Popper, T. L., Berkenkopf, J., Lutsky, B., and Watnick,
A. S. J. Med. Chem., 1987, 30, 1581. (b) Shapiro, E. L. U.S. Patent 4,472,393, 1984 (to Schering).
85Bodor, N. S. U.S. Patent 4,996,335,1991 (to Otsuka).
86Glaxo was already a licensee of Schering’s prednisone-prednisolone patents.
working relationship with Glaxo in which Glaxo agreed to develop a process for the
manufacture and supply of betamethasone to Schering in exchange for marketing
rights to betamethasone and derivatives. A growing interest in topical dosage forms
and the knowledge that betamethasone was not particularly active topically against
eczema led Glaxo to the hypothesis that a less polar form of betamethasone such as
an ester might penetrate the skin better. To test this hypothesis, an assay was needed
which would evaluate the relative potency of betamethasone derivatives in topical
situations. Glaxo heard of the work of a Dr. A. W. McKenzie, a dermatologist in St.
John’s Hospital, London, who claimed that the relative potency of topical steroids
could be measured by their ability to create vasoconstriction and skin “bleaching”
after overnight contact with the skin under a protective dressing. The McKenzie test
proved crucial87 in the rapid evaluation of a variety of esters and the selection of the
valerate (Betnovate)88 as the best. Betnovate was found to be three times as active
as Synalar (Table 2), the best compound on the market at the time. Schering later
developed the 17,21-dipropionate of betamethasone (Diprolene/Diprosone) which
proved to be even more potent than Betnovate.
Today, the markets for anti-inflammatory and contraceptive steroids amount to
billions of dollars a year. Although these two markets are by far the major ones for
steroids, it is worth closing this historical review with an indication of other biological
activities based on the steroid molecule.
Arguably, no other molecule has had more biological diversity built into its basic
four-ring structure than the steroid molecule. By way of illustration, Table 3 provides
an incomplete list of several steroid structures and their very different biological
activities. Most of the compounds in Table 3 are still on the market.
The diversity of biological activities identified in Table 3 and the known role of
steroids in the biochemical processes taking place in the human body suggest to the
optimist that additional useful steroid products might emerge in the future. However,
no one anticipates that another blockbuster industry to rival the anti-inflammatory
and oral contraceptives fields will arise. The illegal use of anabolic steroids such
as stanozolol (Table 3) to enhance “sporting” performance has spawned a small
“underground” industry to design “undetectable” anabolic steroids. One example is
tetrahydrogestrinone (THG).
87It was reported by Sir David Jack and Dr. Tom Walker, in their tribute to Dr. B. A. Hems (1912–1995),
Royal Society BiographicalMemoirs, 1997, 228, that Dr.McKenzie was given a radiogram (a combination
radio and record player) for his invaluable contribution.
88Elks, J., and Bailey, E. J. U.S. Patent 3,376,193, 1968 (to Glaxo).
TABLE 3. A Selection of Steroids with Diverse Biological Activity
Compound Name
Biological Activity and
Original Company
Roussel–UCLAF, 1982 (often
in combination with a
Stanozolol Anabolic agent, Sterling Drug,
1962, Controlled substance
Alfadolone acetate
Combination is
Anesthetic, Glaxo, 1971
Structure IV
Structure IV Digoxin Cardiotonic, Ancient origins
Finasteride Proscar Treatment of benign prostatic
hypertrophy, Merck, 1985;
also marketed to treat male
hair loss
TABLE 3. (Continued)
Compound Name
Biological Activity and
Original Company
Spironolactone Diuretic/Antihypertensive,
G. D. Searle, 1961
Treatment of advanced
breast cancer.
Zemuron (also
Pancuronium, and
Marketed as a family of
similar structures for
muscle relaxation in
surgeries, Organon, 1990
Antimigraine, L?ovens
Komiske Fabrik, 1962
This compound, accessible from commercial intermediates such as XXVII, was
added to the list of banned substances following its identification in a spent syringe
sent anonymously to the Olympic Analytical Laboratory in Los Angeles in June
2003. Despite the risks, clandestine efforts continue. A second designer steroid, desoxymethyltestosterone,
carrying a 17?-methyl-17?-hydroxy functionality, 3-desoxy,
and the A-ring double bond at 2,3 was identified in 2005. Perversely, it seems likely
that the ethical medical community will benefit from knowledge gained on the adverse
health effects arising from the illegal use of banned substances.
In conclusion, negative aspects aside, one can only look back with amazement and
marvel at the astounding achievements of the many thousands of people who created
the steroid industry.
Adventures in Steroid Chemistry
Schering Chemical Development’s intense involvement in the business end of the
company’s steroid manufacturing operations proved to be an adventure because of its
broad open-ended scope.We were invited, and funded, by theManufacturingDivision
to help its offshore manufacturing operations improve their technical performance
(process safety, process yields, and plant throughput) so as to reduce manufacturing
costs and, in addition, to enable them to avoid the need to make capital investments
in offshore plants, notably in Puerto Rico and Mexico City. To this end, they added
vigorous support to the efforts being promoted by the Research Division to upgrade
the Chemical Development operation. This comprised reorganization, investment in
modern facilities, the addition of highly qualified scientists and engineers, and the
introduction of modern instrumentation. The family of people in Chemical Development
who were mostly concerned with Manufacturing Division’s technical problems
came from Taiwan, China, India, the United States, Ireland, Italy, Poland, Romania,
and the United Kingdom.
We established a close rapport with colleagues in the offshore operations. Through
exchanges of people, frequent meetings, a focus on science, enthusiasm, and management
support, much was achieved. In local vernacular, “we had a ball.”89
I will close this section with a description of a couple of general projects, one each
with our colleagues in Mexico City and Puerto Rico. These were:
 Overcoming a Health/Safety issue in the manufacture of a betamethasone intermediate
from Diosgenin in Mexico City, and the evaluation of newer raw
materials derived from plants.
 The early steps for converting the 11?-hydroxylated betamethasone intermediate
into betamethasone alcohol precursors in Manati, Puerto Rico.
From Plant Saponins to 16?-Methyl Intermediates. With the Bhopal disaster in
India in mind, the initial program of work with our colleagues in Mexico City
was directed at reexamining an alternative process to the existing process using
diazomethane as a reagent for the introduction of the 16?-methyl group (Scheme 9,
Routes B and A, respectively).
Perversely, our earliest work was to determine what could be done to ensure
continuation of the diazomethane-based operation in light of DuPont’s decision to
terminate nitrosan production. Production of nitrosan in Mexico City was established
using DuPont’s process. Understandably, the perceived carcinogenic nature of nitrosan
and the continuing, albeit relatively rare, occurrence of minor diazomethane
89Our consultant, Professor Ronald Breslow, once famously remarked “I cannot believe you people get
paid so much money for having so much fun!"
Me OAc
Ac2O CrO3/aq. H2SO4/(ClCH2)2
Separate layers
H2O2 (16,17-epoxidation)
Glycol (C-20 ketalization)
H2/Pd (?5,6-reduction)
3 M MeMgBr/b.THF/H+
Route B
Route A
SCHEME 9. Conversion of Diosgenin to 16?-methyl intermediates.
explosions only added to the anxiety and prompted questions on what might be done
to enhance safety in the short term.90
Ing. Miguel Escobar, General Manager of our Mexico City facility, suggested that
one way around using diazomethane would be to reexamine the possibility of starting
with the 16?, 17?-epoxide of XXXV. He had saved approximately 100 kg of this
compound from a failed earlier program of work. He reported that the epoxide was
easy to product. In adopting this suggestion, we engaged Professor Eugene Braetoff,
University of Mexico City, to investigate the sequence of reactions outlined in Route
B, Scheme 9. Professor Bratoeff demonstrated that the sequence of reactions was
feasible, but found that process yields at both the ketalization and Grignard reaction
steps were poor. In particular, substantial byproduct formation was observed during
the ketalization step.
Dr. Donal Maloney, temporarily based in our Mexico City plant, was given responsibility
for understanding and developing a process for the ketalization step and
Dr. David Tsai (Union, New Jersey) was assigned corresponding responsibility for
the Grignard reaction step.
Dr.Maloney, in collaboration with Professor Bratoeff and Dr. Richard Draper (visiting
fromUnion, NewJersey), quickly found, usingNMR, that the BF3 catalyst being
used to promote the ketalization at high temperaturewas causing aMeerwein–Wagner
rearrangement along the lines of the reaction reported by Schering’s Dr. H. Herzog
years earlier.91
Lewis acid
90TheMexico City diazomethane plant suffered minor damage with each explosion. Fortunately, the plant
was set up in the open and no personnel injuries were sustained. Glaxo kindly invited us to visit their own
diazomethane plant in Montrose, Scotland. This facility was sited in a bunker in a remote on-site location.
One of several reasons for Glaxo’s excellent safety record in operating this plant was due to their plant
design, with polished internal surfaces—rough surfaces are known to trigger detonations. In today’s world,
those needing to use diazomethane would seek the services of custom manufacturers skilled in producing
and using this compound in situ— for example, Phoenix Chemicals Ltd. (see Proctor, L. D., and Warr, A.
J. J. Org. Proc. Res. Dev., 2002, 6, 884) and Aerojet General Corporation (see Archibald, T. G., Huang,
D.-S., Pratton, M. H., and Harlan, R. F. U.S. Patent 5,817,778, 1998; and Archibald, T. G., Barnard, J. C.,
and Harlan, R. F. U.S. Patent 5,854,405, 1998).
91Herzog, H., Joyner, C. C., Gentles, M. J., Hughes, M. T., Oliveto, E. P., Hershberg, E. B., and Barton,
D. H. R J. Org. Chem., 1957, 22, 1413.
Rearrangement was avoided by ketalization at lower temperature with a mixture of
ethylene glycol, trimethylorthoformate and p-toluenesulfonic acid as a catalyst.
Glycol, (MeO)3CH
p-TSA, toluene
Catalytic hydrogenation of the 5,6 double bond proved straightforward. Reaction
with an excess of Grignard reagent at high temperature followed by work-up gave
XXXVI (free 3-hydroxy form). Key factors in forcing the Grignard reaction to completion
were reaction temperature, the molar excess of methyl magnesium bromide,
and the reaction concentration. An important consideration in operating in Mexico
City (elevation approximately 7000 feet) was the boiling point of reaction solvents.
In our case, the THF (used as a Grignard reaction solvent) had to be substantially
replaced by toluene to enable us to increase the reaction temperature to a practical
The replacement of RouteA(Scheme 9) by Route Bwas a triumph of collaboration
and the application of science.
Shortly thereafter, we were visited by Marketing/Technical people from Gistbrocades
in Delft offering us 9?-hydroxyandrost-4-ene-3,17-dione (9?-OH AD,
XXXVIII) as a new steroid raw material.
Gist had succeeded in building on the pioneering work of others92 by finding their
own, industrially attractive, microbiological system for the efficient degradation of
sitosterol, available in large quantities from soya bean byproducts to XXXVIII.93
9? -OH AD
92(a) Sih, C. J., andWeisenborn, F. L. U.S. Patent 3,065,146, 1962 (to Olin Mathieson Chemical Corporation)
set the stage with their finding that C-9 unsubstituted steroids can be microbiologically hydroxylated
to C-9?-hydroxysteroids. (b) Later, Upjohn workers pioneered the use of a mutant of Mycobacterium
fortuitum to degrade sitosterol to XXXVIII (Wovcha, M. G., Antosz, F. J., Knight, J. C., Kominek, L. A.,
and Pyke, T. R. Biochim. Biophys. Acta, 1978, 539, 308.)
93The beauty of the microbiological method lies in its ability to oxidize the structurally similar C-17
side chains of other steroids which contaminate sitosterol (e.g., campesterol, dihydrobrassicasterol, and
stigmasterol—all of which carry the 3?-hydroxy-5-ene structure) at both C-17 and 9?sites to compound
The quality of the XXXVIII obtained by the microbiological degradation process
was surprisingly good.
The main reasons for interest in a totally different starting material were the likelihood
of greater process efficiency, lower costs, and freedom from dependence on
diosgenin from the Barbasco root. Cost projections using the Gist-brocades price
idea for XXXVIII, our own cost calculations based on the literature, the downstream
process simplifications we envisaged, and the demonstrated elimination of a major
source of impurities in the Schering synthesis of a key 9,11-olefine94 (see also later
section entitled “Conversion of 11?-Hydroxylated Betamethasone Intermediate into
Betamethasone Alcohol in Puerto Rico”) were major driving forces in the Schering
decision to evaluate the Gist-brocades startingmaterial. This decision recognized that
considerable changes in the Scheringmanufacturing plantwould be needed to accommodate
the new chemistry, that much of the existing plant (e.g., the fermentation unit
for introducing the 11?-hydroxy substituent) would be idled, and that there would
be a significant involvement of Quality Assurance and Regulatory Affairs personnel
if Schering made the decision to adopt the Gist-brocades starting material.
Given a supply of XXXVIII from Gist-brocades, the main thrust of our work
became finding ways of introducing the dihydroxy acetone side chain at C-17. The
elegant work published by Upjohn chemists Van Rheenen and Shephard94 provided
a foundation for our own work. Over the next three years, our Dr. David Andrews,
Dr. Nick Carruthers, and Dr. Anantha Sudhakar identified two promising options for
further development (Options A&B, respectively, in Scheme 10).
By the beginning of 1990, the reactions in Scheme 10 had mostly reached the
Method stage of development (see Chapter 8 for the definition of Method). In short,
considerable process development and scale-up work, in collaboration with manufacturing
and others, was still needed to create a process.
The year 1990 proved to be a turbulent one in which many uncertainties needed to
be resolved, not least of which was the Gist-brocades announcement that they were in
negotiation with several companies to sell their 9?-hydroxyandrost-4-ene-3,17-dione
(XXXVIII) technology. We subsequently learned that Roussel–UCLAF had bought
the Gist-brocades technology and would be abandoning their long-standing position
using bile acids as a starting material. Roussel also expressed interest in continuing
to supply XXXVIII to Schering. A few in Schering manufacturing were unnerved
by the thought of relying on a competitor to supply starting material, but it was other
considerations that dictated the decision not to develop the route via XXXVIII.
Schering’s New Drug Application for Mometasone Furoate (see Table 2) was
nearing approval by the FDA, and plant capacity projections were such as to require
that we increase plant capacity more rapidly than could be achieved by developing
94The Schering synthesis of a 9,11-intermediate, essentially by dehydration of an 11?-hydroxy intermediate,
always resulted in the formation of approximately 10% (frequently more) of an unwanted
11,12-olefine. Formation of this impurity not only diverted valuable starting steroid into useless product,
but also caused purification problems in later process steps. Although “dehydration” of 9?-hydroxysteroids
can lead to formation of a 8,9-olefine impurity (see (footnote 92(a)), this unwanted reaction has been
shown to be avoidable [see Beaton, J. M., Huber, J. E., Padilla, A. G., and Breuer, M. E. U.S. Patent
4,127,596, 1978 (to Upjohn); and VanRheenen, V., and Shephard, K. P. J. Org. Chem., 1979, 44, 1582.]
Option A
Oxone ClSO3H
Option B
Pd(O) / Bu3P
aq. HCl
Oxone (gives bisepoxide)
As 21-acetate
R = CO2Me
SCHEME 10. Potential routes to betamethasone intermediates from 9?-hydroxyandrost-4-
one of the options in Scheme 10. Essentially this amounted to directing efforts into
increasing the yield of existing plant processes and to developing a long-standing
idea, promoted by our Mexico City General Manager, Ing. Miguel Escobar, to develop
Sarsasapogenin (XVII) as the starting material.
The Sarsasapogenin idea did not come out of the blue. As often happens, thinking
scientists and engineers often quietly pursue ideas in a low-key way, outside the
mainstream of work, and launch them when they feel they have a workable prospect.
Indeed, just as Ing. Escobar had been quietly evaluating opportunities to develop
sarsasapogenin as a starting material, we in Chemical Process development had been
working for years on ideas for improving the yield of the current manufacturing
process (see later section entitled “Conversion of 11?-Hydroxylated Betamethasone
TABLE 4. Principal Commercial Sources of Steroid Raw Materials from Plants
Plant Main Countries of Origin Main Steroid Component
Barbasco (yam) Mexico, China Diosgenin (XVIII)
Agave East Africa, Mexico, China Hecogenin/Tigogenin (XXXIV,95)
Soyabean Numerous Sitosterol/Stigmasterol (XXXVII/XVI)
Yucca Mexico, United States Sarsasapogenin (XVII)
Intermediate into Betamethasone Alcohol in Puerto Rico”). These low-key efforts
later coalesced into the lead process (see later).
The principal issues in promoting the idea of a new starting material in the steroid
field are the cost, practicality, quantity, reliability, and quality of the source and, in
this case, what the chemical plant requirements would have to be to process it. Ing.
Escobar had been aware for some time of the decline in the collection of Barbasco
roots for his diosgenin production. For the most part, this was a consequence of
a movement started by President Echeverria years earlier. This movement led to
government efforts to maintain the Barbasco natural resource for the country rather
than allow the multinational pharmaceutical companies to harvest at will. This only
resulted in most multinationals seeking alternative steroid sources. Reduction of
the supply of roots led Ing. Escobar to develop sources of diosgenin (XVIII) and
tigogenin95 in China to guard against possible disruption of the domestic supply of
roots. At this point, it is worth summarizing the principal commercial plant sources
of steroid raw materials (Table 4).
The success of the multinationals in exploiting alternative steroid sources exposed
the limitations of the wild yam. Yams were laboriously dug from the ground exclusively
for those involved in the manufacture of steroids. In contrast, all the other plant
sources of steroids identified in Table 4 were harvested for other industries; these
industries generally regarded the steroid-containing waste as worthless to them.
Thus, hecogenin and tigogenin are byproducts of the sisal (hemp fiber) industry,
sitosterol and stigmasterol are byproducts of the large soyabean industry, and, in the
late 1980s, Sarsasapogenin appeared, to Ing. Escobar, likely to become a byproduct
of a fledgling Northern Mexico industry in cattle feed, Yucca oil, and other minor
The possible use of sarsasapogenin, originally suggested by the redoubtable Russell
Marker (see footnote 41), had been explored by Schering in the late 1970s following
the Mexican government’s efforts to create a steroid starting material business.96
Schering’s Hershel Herzog, working with a botanist (P. Simpson), surveyed large
stands of Yucca brevifolia plants in the American southwest and in 1978 publicized
his findings that the fruits of this plant could yield viable quantities of sarsasapogenin.
The initiative was not followed up because the Mexican government did not pursue
its diosgenin business venture and there was later indication that the U.S. government
95Structurally this is hecogenin (XXXIV) with the 12-keto group reduced to CH2.
96Herzog, H., and Oliveto, E. P. Steroids, 1992, 57, 617.
had expressed some concern about exploiting the fruits of the plant in the American
southwest in case it became endangered in the wild.
Ing. Escobar quietly followed up the Herzog initiative and explored potential
sources of sarsasapogenin from Yucca plants in Mexico. He found that there was
already an established business in northern Mexico built around utilizing the fruits
of Yucca filifera in the cattle feed business. The sweet fruits that develop from a
panicle of white flowers 10–15 ft above the ground were snipped off the plant with
long-handled clippers by Mexican farmers in their winter season, enabling them to
greatly supplement their income or simply just work to meet their needs! The fruits
were chopped and the black seeds, comprising almost 25% by weight of the total fruit,
were separated. The sweet chopped fruit was mixed with dried chicken manure (still
rich in protein since chickens do not process all the protein present in their feed) and
supplements and the product sold as cattle feed. The seeds contain a useful oil that is
extracted for domestic purposes. The de-oiled milled seeds cannot be fed to animals
because of their steroid content. Analysis of de-oiled seeds showed them to contain
approximately 6% sarsasapogenin (versus only approximately 4–5% diosgenin in
dried Barbasco root).
Though there are many thick stands of Yucca filifera, the distribution of the plant
is mainly sparse, spread across approximately 800,000 hectares (2 million acres) in
four states (Durango, Coahuila, Nuevo Leon and San Luis Potosi). However, judging
by this author’s experience gardening in New Jersey, the seed germinates readily and
could undoubtedly be grown commercially in a warm climate.97 The only downside
would be a 12 to 15-year wait for the Yucca to commence flowering and setting
fruit! There is also a need to ensure proper pollination—apparently this is achieved
naturally in the arid areas west of Saltillo by a small flying beetle.
The production of sarsasapogenin from de-oiled seeds in our Mexico City plant
proved straightforward using essentially the same equipment as previously used in
the production of diosgenin from dried Barbasco roots. Indeed, virtually all of the
subsequent chemical processing steps leading to the final intermediate produced in
Mexico (viz. compound XLI) were implemented in the same plant as was used
starting with diosgenin.
97In New Jersey, it is necessary to bring the potted plant into a warm building and place it under lights
for the winter. After 10 years, this is becoming a serious challenge: One needs to dress in body armor to
prevent impalement on its formidable array of bayonets, and that’s just to get your arms round the pot!
ClCO2Et / Et3N
CH3SO2Cl / pyr.
?, NaOAc / HOAc
HF Epoxidation
via bromohydrin
and cyclization/
As above
SCHEME 11. Conversion of Mexican intermediate XLI into betamethasone alcohol.
Although there is an advantage (versus diosgenin) in starting with sarsasapogenin,
since no hydrogenation step is needed to reduce the 5,6 double bond as in Scheme
9, (compounds XXXV to XXXVI), there proved to be some small disadvantage in
utilizing the sarsasapogenin-derived compound equivalent to XXXVI, which carried
a 5?-hydrogen. Further work was needed to optimize the chemistry for introducing
the 1,4 -3-ketone into the 5?-hydrogen equivalent of XXXVI.
All the work needed to switch from diosgenin to sarsasapogenin as the starting
material was carried out much faster than the program seen as needed if we were
to adopt 9?-hydroxyandrost-4-ene-3,17-dione (XXXVIII) as the starting steroid.
A major factor in the decision making was the time projected for all the plant
modifications thatwould have been necessary startingwith XXXVIII. In short, all the
process selection and development work, and particularly all the engineering design,
plant installation, and process testing in a new or adapted plant, was eliminated by
deciding to use sarsasapogenin in the existing plant.
The success of the Mexico City program, largely inspired by Ing. Escobar, imaginatively
and enthusiastically established by chemists and engineers in Union, New
Jersey, and Mexico, and unswervingly supported by senior management, provided
an excellent example of what can be achieved when groups with different objectives
and skills are teamed to work together toward a common goal.
Conversion of 11?-Hydroxylated Betamethasone Intermediate into Betamethasone
Alcohol in Puerto Rico. As indicated in the section entitled “From Plant Saponins to
16?-Methyl Intermediates,” the most inefficient steps in processing the final intermediate
fromMexico (i.e., XLI) into betamethasone alcohol were the steps to introduce
the 9,11 double bond into XLI. The process sequence is outline in Scheme 11.
The formation of >10% of the 11,12-olefine byproduct (XLV) not only
represented a significant loss of desired steroid, but introduced a further loss by
requiring wasteful purifications of both the epoxide XLIV and betamethasone
alcohol. Formation of XLV occurred because the 11?-hydroxy group (as its
mesylate) in XLII is not anti-coplanar with the 9?-H—elimination from a position
of true anti-coplanarity would be expected to greatly favor formation of the desired
9,11 olefine (XLIII). Indeed, it is known that 11?-hydroxy steroids corresponding
to XLII, which possess favorable anti-coplanarity with 9?-H, can be “dehydrated”
to the 9,11 olefine in 90% yield.
The problem of improving the yield of the9,11 olefine had been tackled, with little
success, by several of our scientists over many years.98 The Schering fermentation
scientists, several with past experience in trying to find a practical biological system
for introducing the 11?-hydroxy group into XLI, did not encourage us to try to
reopen this project—they thought that it would be more productive to put resources
into improving the existing strain, which produced the 11?-hydroxy steroid (XLII)
in Puerto Rico.
Not surprisingly, given the long history of failure, few of our chemists seemed
willing to take on the 9,11-olefine project. One who asked was Dr. Chou Tann, who
had already established a record of success on other projects. He was given the task at
a time of increased awareness that there was an urgent need to improve process yield,
product quality, and plant throughput in the manufacture of betamethasone alcohol.
Dr. Tann’s proposals, to investigate the use of reactive phosphorus compounds to form
(Cl)x(O)y P–O bonds with the 11?-hydroxy group and to find ways of “arranging” a
cis eliminationwith the 9?-H,were greeted with skepticism. The skeptics thought that
compounds such as POCl3 and PCl5, suggested byDr. Tann asworthy of investigation,
would give high levels of the 11?-chloro compound through an SNi type of reaction.
However, Dr. Tann pointed to old literature which showed that 11?-hydroxysteroids
would react at room temperature with POCl3 to give 9,11 steroids, albeit in only
15% yield,99 and with PCl5 to give 9,11 steroids in 65% yield.100 The ratio of 9,11
to 11,12 olefines was 98:2 in the POCl3 case and not reported in the PCl5 case. The
skeptics pointed out that the PCl5 reaction100 gave 9?, 11?-dichloride as a major
byproduct and that the very low yield using POCl3 did not augur well.
Examination of the molecular mechanics energies of elimination to 9,11 versus
11,12 by Dr. B. E. Bauer of the Schering Research computational chemistry group
revealed an energy difference between the two eliminations of only 1.51 kcal/mol in
favor of the 9,11-elimination.
98From time to time, I assigned the “dehydration” problem, for a few months, to incoming scientists, as
well as a few of our experienced ones who were “between projects.”
99Bernstein, S., Lenhard, R. H., and Williams, J. H., J. Org. Chem., 1954, 19, 41.
100Shoppee, C. W., and Nemorin, J. J. Chem. Soc., Perkin Trans., 1973, 542.
7.4% Cl
SCHEME 12. PCl5-mediated “dehydration” of 11?-hydroxysteroids.
In spite of all the negatives, and in keeping with the principle that one should never
use theory to abort an experiment, Dr. Tann took on the project.
In exploring the 11?-chlorophosphate idea further, Dr. Tann’s group concentrated
on solvent choice and temperature effects, with PCl5 as the reactant. Tetrahydrofuran
was selected as the solvent following comparison of reactions carried out in this
solvent versus pyridine and methylene chloride. A reaction at room temperature in
THF gave an encouraging result. Three products were produced, with the desired
9,11 olefine being the major one (Scheme 12 summarizes the findings and outlines
the speculative mechanism for the main reaction). An alternative mechanism could
be via dehydrochlorination of an initially formed 11?-chlorosteroid (XLVI). Against
this it was observed that when pyridine was used as the solvent at room temperature,
the PCl5 “dehydration” actually produced 44% of XLVI!
Interestingly, no 9?,11?-dichloride was produced under the conditions employed.
In developing the chemistry further, it was reasoned that the small energy difference
between the 9,11- and 11,12-eliminations would be enhanced in favor of the
9,11 olefine at low temperatures. This led to a study of the effects of temperature
with the results summarized in Table 5.
Despite the very low levels of 11?-chlorosteroid impurity (XLVI) produced in the
low-temperature reactions, it was thought prudent to find a process for reducing the
level of the 11?-chloro compound or eliminating it altogether. This was because the
betamethasone alcohol produced from PCl5-derived 9,11 olefine could contain up
to 0.1% of a new impurity, found to be the 21-alcohol corresponding to XLVI. Since
we preferred to avoid dealing with the potential regulatory issues associated with
introducing a new impurity into an API, we sought a simple method for eliminating
the 11?-chloro compound. Based on observations that XLVI gave predominantly
the 9,11 olefine (XLIII) when heated at 80?C in aqueous acetonitrile, Dr. Tann’s
group experimented with other polar solvents and found that simply heating XLVI in
dimethyl sulfoxide at 90?C until the dehydrochlorination was complete gave >97%
TABLE 5. Influence of Temperature on the “Dehydration" of XLII (as its 21-Cathylate) by
PCl5 in THFa
Temperature % 9,11 % 11,12 % 11?-Chloro (XLVI)
Room temperature 81 7.4 11.3
?20?C 92.3 4.9 2.8
?45?C 95.5 3.3 1.2
?78?C 97.5 1.8 0.7
?85?C 98.4 1.3 0.3
aAll figures are HPLC area %.
of XLIII. In practice, this evolved into heating in DMSO for 1 hr at approximately
100?C. The PCl5-mediated “dehydration” process was patented and published.101
Dr. Tann’s group went on to study the subsequent reactions for converting XLIII
into XLIV (Scheme 11). The first step was to isolate, identify, and quantify the
impurities and then determine how best to avoid their formation. Themain impurities,
generally produced in low amounts in the manufacturing plant, were compounds
XLVII was speculated as being formed by the debromination of a 9?,11?-dibromo
derivative of XLIII during the epoxidation process, whereas the 21-cathylate XLVIII
was believed to have arisen due to occlusion in the relatively insoluble crystals
of desired epoxide (XLIV) formed during the hydrolysis of this 21-cathylate in
Formation of XLVII was avoided by carrying out the epoxide formation step in
DMF (as solvent) with a low level of 70% HClO4. It was reasoned, correctly, that a
bromoformate would be the preferred intermediate over a bromohydrin or dibromo
intermediate and that this would cyclize to the ?-epoxide more cleanly. This proved
to be the case, eliminating 9?,11?-dibromo intermediate formation and with it the
impurity XLVII.
101(a) Fu, X., Thiruvengadam, T. K., Tann, C.-H., Lee, J., and Colon, C. U.S. Patent 5,502,222, 1996 (to
Schering Corp.). (b) Fu, X., Tann, C.-H., Thiruvengadam, T. K., Lee, J., and Colon, C. Tetrahedron Lett,
2001, 42, 2639.
Formation ofXLVIIIwas avoided by utilizing a solvent mixture (methanol/methylene
chloride versus methanol/tetrahydrofuran) in which the epoxide (XLIV) was more
soluble, thus avoiding the occlusion problem. This work was published.102
The new PCl5-mediated “dehydration” reaction coupled with the improved epoxidation
reaction created major benefits for the manufacturing division. The overall
yield in taking the a 11?-hydroxy compound (XLII) to the epoxide (XLIV) (Scheme
11) increased from 60% to 85% and the quality of the epoxide improved to 99%
purity from 92% previously. Although capital investment in low-temperature plants
was necessary,wasteful purifications were eliminated, plant throughputwas increased
substantially, and regulatory concerns were greatly diminished. As a result, the cost
of goods for betamethasone alcohol dropped dramatically.
The historical development of the steroid industry summarized, albeit briefly, in this
excursion provides perspective both on the medical importance of steroids and on
the astonishing contributions made by scientists and engineers of many disciplines,
from many countries, which led to today’s industry worth many billions of dollars a
year. Tragically, World War II prevented German and Swiss chemists from making
greater contributions to this industry.
The relatively high cost of steroids, along with the small scale of operations,
allowed chemists working in the steroids field the luxury of taking an unconfined
approach to developing chemical syntheses and processes. Thus, it has been possible
to consider the use of many avant garde chemical reactions and chemical reagents,
biological transformations, and process conditions that the larger chemical industry
generally does not consider because of cost or hazards. Such contributions have in
many cases brought relatively exotic chemistry of all types into the mainstream of
chemical process development.
The development of process chemistry in the steroid field has played an important
role in challenging chemists to find ways of dealing with specific transformations in
polyfunctional molecules. This has enhanced appreciation of the role of protecting
groups, reaction selectivity, chiral chemistry, and the need for high-yielding chemical
reactions. Chemical engineers have also played a vital role in translating new chem-
102Fu, X., Tann, C.-H., and Thiruvengadam, T. K. J. Org. Proc. Res. Dev., 2001, 5, 376.
istry into plant practice; without their ingenuity, much of the new chemistry would
not have reached plant operations. All the new chemistry has challenged process
analysts to find and apply analytical techniques, enabling them to quantify chemical
reactions and identify and quantify impurities. Inevitably, in the highly regulated
world of today, everyone needs to be conscious of the needs in meeting the requirements
of Regulatory Affairs (e.g., API impurities), Safety (e.g., finding alternatives
to hazardous reagents, etc.), and the Environment (e.g., tracking and dealing with the
disposition of such as oral contraceptives).
As indicated by Table 3 in this presentation, there is still life left in the steroid
industry, now more than 70 years old. Ongoing research to understand the biological
workings of the human body will undoubtedly uncover more opportunities for
creating steroidal drugs.
For those of us who have worked in the steroid field for some time, the progress
made, looking back, can only be regarded with a deep sense of satisfaction, humility,
and awe. Looking forward, there is still much to be learned and done.
Development work is always rewarding, even when bittersweet.
A long and demanding program of work is needed to take a potential API from
the identification stage to the market. Process development chemists and engineers
working in the pharmaceutical industry initially reproduce or adopt research chemists’
Recipes to supply the small quantities of APIs that enable research colleagues to
progress their biological evaluations. Later,when a final candidate structure is selected
for further work, they engage both in further improving the research recipe and/or
quickly searching for and developing a synthesis Method suitable for safe scale-up.
Scale-up of this Method provides large quantities of the API for further development.
Chemical Process development follows as part of the enormous comprehensive effort
that a company undertakes to harness all of the many disciplines and resources (in
chemical and biological research, toxicology, clinical, chemical, and pharmaceutical
development, etc.) needed to create an NDA submission and gain FDA approval.
In reality, almost all of the molecules identified in the torrent of enthusiasm
attending the discovery phasewill fail to become commercialAPIs. Most are “weeded
out” during the preliminary evaluation phase. Even then, very few of the survivor
molecules will succeed in becoming commercialAPIs. The development climate, like
theweather, is unpredictable—becomingmore certain only as information frommany

Copyright C 2008 John Wiley & Sons, Inc.
sources coalesces. Nevertheless, candidate API molecules selected for development
are all progressed as though they will all succeed.
Chemists are fortunate in that the discipline of the chemistry profession enables
them to rise above the uncertainties. As a group, they are easily “transported” above
the fray by the intellectual challenges, the flow of ideas, and the molecular manipulations
they see as needed for a practical synthesis. Successful management of such
wonderful professionalism is an important requirement in pursuing a development
program. It is not easy to be “carried away” and yet keep your feet on the ground.
It is not easy to accommodate all the disciplines needed to succeed and adjust to
broader points of view. It is not easy to handle project failure even though failure
is often for reasons that are outside one’s control—for example, a toxicity issue.
Although mourning losses is difficult (mourning is harder the longer one has spent
on a project), the chemist and engineer can frequently salvage considerable scientific
satisfaction from the technical successes within the chemistry of the project.
Publication, if possible, also helps to bring closure.
Under the heading of “Case Studies,” I am providing two examples of chemical
process development projects. One of these, the development of a commercial process
for the manufacture of the cardiovascular drug, Dilevalol hydrochloride, reached
the manufacturing and marketing stage, albeit briefly. The second case summarizes
the early process research work carried out to identify a safe, low-cost process for the
manufacture of the brain cancer drug, Temozolomide. This work never got beyond
the laboratory phase of identifying a better process option for the manufacture of an
intermediate—our avant-garde proposals for building the Temozolomide molecule
itself were never tested. Despite the unfinished state of the project, it is worth recounting
the ideas and their progression. Sad to say, many projects end like this—in
a suspended state, perhaps one day to be resurrected!
The turbulent nature of active pharmaceutical ingredient (API) development is widely
recognized. Only I in 10 or so new entities identified for development will reach the
marketplace.1 Pharmacological, Toxicological and Clinical findings can halt development
at any time. The rate of progression of a new entity can be greatly changed
by toxicology issues, by metabolic findings, formulation difficulties, the availability
of bulk supplies, readouts from both clinical studies and the FDA, changing market
conditions and so on. In short, priorities are constantly shifting. The process of
1DiMasi, J. A. Success rates for new drugs entering clinical testing in the United States, Pharmacology
and Therapeutics, July 1995, 58, 1. See also Scrip, 1995, August 18, p. 18. Overall, the significance of
prior testing of an API outside the United States is apparent: For licensed-in compounds the success rate
was 29.4%; for compounds first tested abroad, 14.6%; for those originated and first tested in the United
States, 10.4%.
CASE 1 269
API development also runs so fast that, from a chemical development standpoint,
conflicts between keeping the IND synthesis and improving or changing to a better
synthesis provide endless challenge and, frequently, much frustration. In an API development
respect, Chemical Development organizations act mostly in a service role,
providing high-quality APIs for toxicology, clinical, pharmaceutical development,
and analytical programs in a timely manner. Beyond the service role, chemical development’s
contribution to the identification and development of a commercial process
is generally crucial. In this area, involvement with manufacturing organizations is
essential—indeed the most successful chemical development organizations have a
close link with manufacturing.
In achieving its mission, a Chemical Development organization relies upon
chemists, engineers, and analysts as the primary professionals and works to forge important
collaborations with other groups. Above all, Chemical Development ensures
the safety of its chemical processes from the beginning and increasingly concerns
itself with creating as environmentally sound a process as possible. Beyond Safety,
the groups of most importance, especially at the start of a project, are Quality Control
and Regulatory Affairs—a continuous dialogue addresses issues and expedites the
filing of IND’s, IND updates, and NDAs.
The development of a commercial process for the manufacture of dilevalol hydrochloride
was not much affected by toxicology, pharmacology, or formulation
considerations since experience with labetalol hydrochloride had provided a knowledge
base on which to build. The greatest challenges were provided by the need
to produce large quantities very quickly, by the cost-of-goods (COG) target and by
chiral synthesis requirements—we assumed that a chiral synthesis would lead to the
lowest COG.
The development of dilevalol hydrochloride2 for use as a vasodilator and competitive
antagonist at ?-adrenergic receptor sites was an outgrowth of efforts in
Schering–Plough, initiated in the late 1970s, to determine whether a single enantiomer
of the racemic ?- and ?-antagonist labetalol would offer an advantageous
marketing situation. It was speculated that one of the four enantiomers which comprise
labetalol would carry enhanced ?-adrenergic receptor blocking activity and
fewer side effects.
Testing the concept required the preparation of pure samples of each of the four
enantiomers. These were prepared by Gold et al.3 and Chemical Development using
classical resolution and chiral synthesis methods. It was quickly found that the
RR enantiomer, later named dilevalol, was virtually free of ?-adrenergic receptor
blocking activity and also possessed superior vasodilator properties versus labetalol.
2U.S. Patent 4,619,919 to Schering Corporation, Nov. 14, 1986; U.S. Patent 4,950,783 to Schering Corporation,
Aug. 21, 1990.
3Gold, E. H., Chang, W., Cohen, M., Baum, T., Ehrreich, S., Johnson, G., Prioli, N., and Sybertz, E. J. J.
Med. Chem., 1982, 25, 1363.
The receptor blocking properties of the four enantiomers were published by Gold
et al.3 and Hartley.4 A summary of Gold’s figures is presented in Table 1.
TABLE 1. Summary of Comparative Cardiovascular Effects of Labetalol
and Its Stereoisomers Relative Potenciesa
Compound ?1-Receptor Blockade ?-Receptor Blockade Vasodilation
Labetalol 1 1 1
RR isomer 3.5 <0.2 7
RS isomer <0.06 0 –
SR isomer <0.05 5.1 –
SS isomer 0 1.5 –
aPotencies normalized to labetalol = 1. ?1-Blockade and ?-blockade are on different
absolute scales (see Gold et al.3 for detail and qualifications).
These results demonstrate that the RR and SR enantiomers are most responsible
for the ?- and ?-blocking activities, respectively. Hartley4 also showed that a 1:1
mixture of RR and SR enantiomers was twice as active a ?-blocker as labetalol and
about 1.3 times as potent an ?-blocker.
Based on knowledge of the activity of the RR enantiomer and other marketing
considerations, the Schering Cardiovascular Therapy Team decided to pursue the
development of a 100- to 200-mg maintenance dose twice daily. This situation raised
cost-of-goods (COG) questions which were addressed to Chemical Development
through the marketing representative in the Cardiovascular Therapy Team. By the
time of the COG request, Chemical Development had taken over the implementation
and development of the chiral synthesis of dilevalol outlined by the Schering–Plough
Research organization. Although this chiral synthesis was improved and shown to be
workable from an initial supply standpoint (indeed it was scaled up to meet urgent
bulk drug supply needs), it was not considered a good candidate for commercial
operation. Nevertheless, in the spirit of reaching to achieve the lowest COG, we
in Chemical Development projected that considerable simplification of the Research
chiral synthesis should be possible (see later). Since both the Research chiral synthesis
and the projected simplified synthesis of dilevalol were based on the original labetalol
synthesis, we “guesstimated”—following discussion with Schering Manufacturing,
who produced labetalol—that a fully absorbed manufacturing cost (raw materials,
4Hartley, D., Chem. Ind., 1981, 551.
CASE 1 271
labor and overhead) at a 50-tonne/annum scale (a figure provided by Marketing for
3 years after launch) should be in a range as follows:
Cost of dilevalol = [3 ? cost of labetalol] ± 25%
Not surprisingly, given these figures,Marketing promptly set the COG target at [3
? cost of labetalol] minus 25%!
It is worth adding a cautionary note in regard to COG projections. In the early
phases of a project, such as was the case with dilevalol, there is a danger that
COG projections might be used to justify termination of a project, rather than serve
to challenge the creativity of process R&D Chemists to invent a better synthesis.
Fortunately, in the dilevalol case, the Cardiovascular Therapy Team and particularly
Marketing and Manufacturing aggressively supported the development chemists and
engineers in their efforts to create a simpler synthesis. COG projections were used, as
the project developed, to validate that the core simpler, lower-cost synthesis strategy
was viable and to identify those features and components of the synthesis most in
need of improvement.
Early Considerations in Selecting a Synthesis Route for
Further Development
The possibility of separating dilevalol from labetalol was considered as an option at
the commencement of the dilevalol project. However, it quickly became clear, from
work carried out in both Research and Development, that this approach might not be
a viable option. Although the racemic pairs (RR + SS and RS + SR) were separable
by crystallization, and although the optical resolution of the RR and SS enantiomers
could be achieved through salt formationwith a chiral acid, the direct yield of dilevalol
was less than 20%. Nevertheless, it was recognized that if the recovery and recycling
of the waste streams from the physical and optical resolutions could be carried out
efficiently, considerable economies would be obtained (Scheme 1).
Acid-catalyzed racemization of the benzylic alcohol in the waste was shown to
be quite straightforward. However, racemization of the carbon carrying the amino
group appeared likely, from probing experiments, to prove difficult especially on
a manufacturing scale. Moreover, COG concerns haunted the prospects of creating
an efficient separation and recycling process, especially one starting with the most
expensive molecule in the labetalol synthesis, labetalol itself. In summary, labetalol
was considered unsatisfactory as a starting material on account of the logistics of
the initial separations, the low one-pass yield of dilevalol, the need for two steps
to racemize the very large quantity of waste, and the excessive solids handling
requirements. Simple calculations showed that theMarketingCOGtarget for dilevalol
would not be attainable starting with labetalol. Furthermore, it was quickly evident
that a large labor-intensivemanufacturing plantwould be needed to develop a dilevalol
process based on labetalol.
The above realities, along with the realization that large quantities of dilevalol
would be needed quickly for the Toxicology, Clinical, and Pharmaceutical Development
programs, led Research to propose a chiral synthesis for the initial supplies.
The process identified was based on the use of a labetalol intermediate and analogous
chemistry to that used in the subsequent manufacturing steps for labetalol.
Synthesis of Initial Supplies of Dilevalol for Cardiovascular Therapy
Team Programs
The process for the manufacture of labetalol is outlined in Scheme 2.
It was clear that a dilevalol synthesis strategy based on Scheme 2 would be advantageous.
Use of the same intermediates and synthesis scheme as used for labetalol
introduces operating economies. In addition, faster implementation and lower costs
were anticipated by building on existing operations. Furthermore, although it was
recognized that dibenzylamine in the labetalol synthesis was an expensive way of introducing
the NH2 group needed in labetalol, it was reasoned that use of a secondary
amine, with the desired chirality already built in, may lead to induction of chirality
in the subsequent reduction step. The Research synthesis was based on the O-benzyl
derivative of 5-ASA (I) and is outlined in Scheme 3.
CASE 1 273
In this approach the R-amine moiety in III was considered likely, in view of the
work of Yamada and Koga5 and later Kametani et al.,6 to provide some inductive
control in the sodium borohydride reduction of IV. Moreover, the R-amine moiety is
a necessary component of dilevalol. Desired inductive control was quickly demonstrated
by Gold et al. (internal communication, October 25, 1979). However, a broad
study of process conditions, particularly of solvent and temperature effects, only
gave, at best, a ratio of RR to SR of
75% 25%
A later publication by Hartley (see footnote 4) validated these figures.
The above result provided a basis for the idea that if both alkyl substituents on
the amine moiety were R-configuration inductive control in the sodium borohydride
reduction of the keto group might be greatly increased. Since ?-methylbenzylamines
are known to hydrogenolyze relatively easily (cf. benzylamine itself),?-methylbenzyl
substitution was considered a good choice. The ready availability of both RS- and
R-?-methylbenzylamine prompted investigation of this proposal.
The RR-secondary amine (VI) was synthesized as follows:
5Yamada, S., and Koga, K. Tetrahedron Lett. 1967, No. 18, 1711; Koga, K., and Yamada, S. Chem. Pharm.
Bull., 1972, 20, 526.
6Kametani, T., Kigasawa, K., Hiiragi, M., Wagatsuma, N., Kohagizawa, T., and Inoue, H. J. Pharm. Soc.
Japan, 1980, 100, 839.
Stereoisomer Analysis of
Crude Dilevalol Acetate
Analysis of
Yield of
Dilevalol DBTA
Salt Based on
Starting RR
Conformation of
RR:RS* = 52:48 — 21.3 77.5 1.2 — 4.3 94.6 1.2 50.5%
RR > 98% 0.7 10.2 86.9 2.1 0.2 2.3 96.2 1.3 59.0%
?S component derives from RS-?-methylbenzylamine.
Reaction of VI with II gave a high yield of aminoketone VII, which when reduced
first with sodium borohydride and then with hydrogen Pd/C gave crude dilevalol
isolated as its acetate (Scheme 4). The acetate salt was purified by dissolution and
crystallization as its DBTA salt.
In order to determine whether or not added steric effects in using ?-methylbenzyl
contributed to the induction of R configuration in the sodium borohydride reduction,
the same sequence of reactions was carried out employing RS-?-methylbenzylamine.
Analysis of the products from both reaction sequences gave the results summarized
in Table 2.
Table 2 indicates that there is little difference between benzyl and RS-?-methyl
benzyl in terms of inductive effect in the ketone reduction. Furthermore, as expected,
the use of the secondary amine VI, in which both amine substituents were R in
conformation did give a desirable increase in the yield of dilevalol. As an aside it
is known that hydrogenation of the keto group in compound IV gives a 1:1 mixture
of the RR and SR enantiomers (see footnote 4). From this it is clear that the most
important factors in the induction of maximum chirality in the ketone reduction step
CASE 1 275
are the complexation of borohydride with the amine function (see footnote 5) and the
like chirality of the N-alkyl substituents.
The above Scheme 4 process utilizing RR-amine (VI), referred to as the “RRamine
process,” became the IND process for producing hundreds of kilos of dilevalol
hydrochloride needed for the early clinical, toxicological, and pharmaceutical sciences
work. The Scheme 4 process was patented.7 It should be noted that, despite a
great deal of work on process conditions, solvents and reducing agents the ratio of
RR to SR could not be improved. Thus, in a practical sense, it proved impossible
to eliminate the DBTA resolution as a step for creating desired chiral purity. Minor
changes in the process and improving the optical purity of VI gave an acceptablequality
dilevalol with an RR assay of generally greater than 97%. Process changes
were filed, as IND updates, with the FDA as they were validated. The full analytical
specification set for dilevalol, as its hydrochloride salt, was
 Description: White to off-white powder.
A. I.R.—Agrees with reference standard specification.
B. Chloride—Responds to test.
C. TLC—Sample spot migrates at the same rate (Rf) as the reference standard
 Related Compounds: Maximum 1% total with not more than 0.5% of any one
 Stereoisomer Content: Maximum 3% total of other stereoisomers (SS + RS + SR).
 Specific Rotation [?]26
D = ?26.5? to ?30.5?
 Loss on Drying: Maximum 0.5%.
 Residue on Ignition: Maximum 0.1%.
 Heavy Metals: Maximum 0.002%.
 Assay (HPLC): 97–102% (calc. on dry basis).
The impurity profile (synthesis related impurities) was actually better for dilevalol
than for the parent compound, labetalol, reflecting the benefits of further purification
during the DBTA resolution step. Thus, although dilevalol hydrochloride contained
traces of DBTA itself, no tertiary amine impurities or brominated dilevalol could be
detected—both types of impurity are present at very low levels in labetalol.
Selecting the NDA Process and Reducing the Cost-of-Goods
The above “RR-amine process,” based on RR-amine (VI), was used for the first
approximately two-year production program supplying most of the early requirements
of bulk drug for the Clinical, Toxicology, and Pharmaceutical Development
7U.S. Patent 4,658,060 to Schering Corporation, Apr. 26, 1982.
programs. During this time, efforts were undertaken to improve the process and to
assess its commercial potential. At the same time, process research was going on in
Chemical Development to evaluate the simpler synthesis, referenced earlier, which
was projected as likely, if successfully developed, to meet the Marketing COG targets.
Other ideas for improving the simpler synthesis, as well as ideas for radically
different synthesis, were “championed” during this period.
Initially, only a small effort was disposed to assess the feasibility of the simpler
synthesis. This grew at the expense of the “RR-amine process” because it became
evident that this process was, like its forerunner (the separation of labetalol isomers),
unlikely to achieve the Marketing COG target. Ideas for radically different syntheses
were given an even lower priority and were often left to “bootleg efforts” by the
The simpler synthesis was built on the same premise as the “RR-amine process”,
namely that it should be based on the existing labetalol process, particularly in terms
of using the same or similar raw materials and intermediates, wherever possible, and
also using similar plant equipment. The simpler synthesis grew out of a critique of
the disadvantages of the “RR-amine process” (Scheme 4).
Main Disadvantages of Scheme 4 Improvements
1. Too many steps. Avoid steps or combine them.
 Is benzylation of the phenol
Test elimination of Bz protection.
 RR-Amine is a new compound—
low-cost source needed.
Go to third party—minimize
 Chiral reduction not 100%—
DBTA resolution unavoidable.
Minimum is to recycle DBTA.
 Two reductions necessary (BH?4
and H2/Pd:C)
One reduction if Bz eliminated
 High solvent and reagent usage. Increase reaction
2. Costs are high.
 Expensive chiral
?-methylbenzylamine is lost as
Since DBTA resolution is
unavoidable, eliminate
enhancement of chiral induction.
 Recovering/recycling wastes adds
Minimize wastes.
 Considerable capital investment
needed—high depreciation
 Simplify process to
reduce/avoid these costs.
Based on the above critique the simpler process was defined as follows (Scheme 5).
The simplified process concept was itself the subject of criticism and doubt:
 Could R-1-methyl-3-phenylpropylamine (R-amine, VIII) be sourced at low
enough cost?
CASE 1 277
 Would the likely dialkylation of R-amine (VIII) introduce new impurities which
are difficult to remove?
 Would R-amino ketone (IX) be isolable relatively free of dialkylated impurities,
thereby serving as a purification step if needed?
 Would the DBTA resolution of an expected 1:1 SR: RR mixture be efficient?
 Would the dilevalol hydrochloride obtained by this Scheme contain any new
impurities which would complicate the Regulatory registration process?
 Could the work needed to demonstrate and prove that the simplified process
gives dilevalol hydrochloride acceptable to Regulatory Affairs and the FDA be
done in the time frame needed to update the IND prior to NDA filing?
Raw Materials Sourcing. There was relatively little problem in sourcing 5-
bromoacetylsalicylamide (in-house) or dibenzoyl-(+)-tartaric acid (large tonnage
Italian source). Although RS-1-methyl-3-phenylpropylamine was available at low
cost ($10–12/kg) in tonnage quantities (Germany and Holland) no supplier of the
R-amine VIII was known.
R-1-Methyl-3-phenylpropylamine. Many chiral acids were evaluated, with water as
the solvent, before N-formyl-L-phenylalanine (FPA) was selected as the best acid for
resolving RS-1-methyl-3-phenylpropylamine. The resolving acid and process were
patented.8 An outline of the commercial process implemented in Germany is given
in Scheme 6.
The resolution process step worked very well on a commercial scale. In an early
version, the R-amine was obtained as a methylene chloride solution. This had not
posed a problem in the Chemical Development plant.Methylene chloride was rapidly
distilled using a condenser system which efficiently liquefied the distilled methylene
chloride. when this process was transferred to Germany, the rate of distillation of the
8European Patent 320898 to Schering Corporation, June 21, 1989.
methylene chloride in the equipment available was greatly extended (from 3–4 hours
to >30 hours) to enable the German plant to stay within its environmental emission
permit for methylene chloride. Under these conditions the R-amine was alkylated
In hindsight, the effects of using methylene chloride as a solvent for a primary amine
should have been predictable. The reality is that the possibility of adverse reactions
occurring was lost with repeated successful use of methylene chloride under the rapid
distillation conditions prevailing in the Chemical Development plant.
Various solvents were evaluated as methylene chloride replacements. Toluene was
selected as best meeting all the needs. Thus, the R-amine was sufficiently soluble,
water phases were readily separable at 25?C, and toluene contamination was of
no consequence since the reaction of R-amine and 5-bromoacetylsalicylamide was
already conducted in the presence of traces of toluene.
The chemical resolution process of Scheme 6 was developed into an economically
favorable one by the ready recycle of both the FPA resolving agent and the combined
(mostly S) amine fractions. FPA, which was prepared by methyl formate reaction
with l-phenylalanine, was shown to be stable (no racemization and no hydrolysis)
when the resolution and work-up processes were conducted within the pH range
of 2–12 at temperatures below 25?C (higher temperatures were not studied). FPA
meeting specification was isolated in yields of ca. 95% by simple acidification of its
aqueous salt solutions and filtration.
The 1-methyl-3-phenylpropylamine containing fractions of largely S conformation
were found to be readily racemized without degradation by heating at 150?C
and 150 psi hydrogen in the presence of Raney nickel.9 Thus the manufacturers of
9Finding by Dr. N. Carruthers and R. DeVelde in our laboratories.
CASE 1 279
RS-1-methyl-3-phenylpropylamine, who produced this compound by the reductive
amination of benzylacetone, were well able to racemize the byproduct S-containing
fractions, thereby providing some additional cost reduction and also avoiding a waste
disposal problem.
Although the above commercial process succeeded in providing R-1-methyl-3-
phenylpropylamine (VIII) for significantly less than $100/kg several other companies
carried out research to find even lower cost processes based on the RS raw material.
A few of these processes will be described later.
Development of the New NDA Process to a Commercial Scale
Early Considerations. A degree of nervousness developed in the early stages of
evaluation and promotion of the simpler synthesis. Many were concerned by the
risks associated with process change. In particular QC and Regulatory Affairs raised
questions on quality, especially the impurity profile and the equivalence of dilevalol
hydrochloride from the simpler process. By this time the ‘biobatch’ had already been
produced via the “RR amine process” for toxicological and pharmacological work.
Understandably Pharmaceutical Development, in harmony with QC and Regulatory
Affairs, wanted assurance that such parameters as crystal size, bulk density, particle
size distribution and tablet dissolution rates would not change and that tablets from
the simpler synthesis would be bioequivalent to the dilevalol hydrochloride from
the “RR-amine process” already registered in the IND application. In short, urgent
evaluations of the product of the simpler synthesis were needed in order to gather the
data required for an IND update. Regulatory Affairs proposed that, since the NDA
filing was only about 2 years away, the data needed to validate the simpler synthesis
should be obtained as quickly as possible and presented to the FDA Cardiovascular
Division reviewer, as much in advance of the NDA filing as we could manage.
In defense of the process change, it was pointed out that the R-amine and RRamine
processes were based on the same chemistry. Nevertheless, we still needed
to accommodate the views of those who queried whether the hydrogenolysis step
in the “RR-amine process” was somehow also introducing a purge of “something”
during the debenzylatoin step in Scheme 4! Those with concerns did, however,
concede that by intersecting the “RR amine process” before the DBTA resolution and
purification steps, the simpler R-amine process did maintain considerable “clean-up”
The Schering Manufacture Division was, at the same time, also looking to the
future by initiating outside evaluation of the “RR-amine process.” This effort was
intended to determine whether others, with under-used manufacturing plant capacity,
could take on the production of dilevalol hydrochloride, or a late intermediate, and
generate cost projections, at a 50-metric tonne/annum production rate, which would
be advantageous versus in-house projections. The primary objective was to source
a late intermediate, thereby allowing Schering Manufacturing to undertake the final
steps itself under the strictest GMP control. Another scenario was also initiated, with
both Chemical Development and Manufacturing determining the capital investments
which would be required to manufacture dilevalol hydrochloride in-house. The figure
for a 50 tonne/annum manufacturing plant on the Schering manufacturing site in
Ireland via the “RR-amine process”, assuming raw material outsourcing, was
estimated at $48–50 million. Chemical Development proposed that since it had much
unused plant equipment in Union, New Jersey (bequeathed by Manufacturing when
it moved operations to Puerto Rico), that an alternative would be to carry out early
launch manufacture of a late stage intermediate in Union by using the simpler synthesis.
Ireland would then take on the manufacture of the dilevalol hydrochloride
in mostly existing plant. The reasoning was that it would be advantageous to limit
capital spending and delay major investment in manufacturing plant until the process
was better defined and the market needs were better known. Since the Union capital
investment was projected at only $5–6 million, this strategy was adopted. This approach
also made best use of Chemical Development’s chemical engineers who were
closely involved in the design of the process as well as in the testing and selection of
process equipment.
Process Development Leading to FDA Review. One of the core premises in the
simpler process as outlined in Scheme 5 was that it would be highly desirable,
especially from a cost reduction standpoint, to avoid isolating solids. In thisway all the
equipment and labor needs associated with solids handling would be avoided, thereby
reducing the product cost byminimizing labor and overhead costs and also by boosting
the process throughput and plant capacity. The main disadvantage of such a strategy
is that the process loses an outlet for byproducts (impurities) and generally requires
that high reaction yields are obtained to minimize the need for purging reaction
byproducts. The following describes the successful efforts to combine process steps
such that the first isolated product from the reaction sequence is dilevalol as its DBTA
From Raw Materials to Dilevalol DBTA Salt. In order to provide the best chance of
success in the Scheme 5 sequence, great emphasis was placed on starting with the
highest-quality raw materials.
Appearance: White to cream solid.
Identity: IR agrees with standard.
Purity (HPLC): >95%
The major impurity is 5-acetylsalicylamide
Small amounts (<0.3%) of ring brominated
impurity are also present.
Appearance: Clear, colorless to light yellow
Identification: IR agrees with standard.
Specific Rotation: [?]20
D = ?19.0 ± 1.5? (c= 5 in cyclohexane).
Enantiomeric Purity: Minimum 98% R
(Moshers acid method)
Residual Solvent: ?0.5% (gc).
Related Compounds: ?1% (gc).
Assay: 98–102% (HClO4 titration)
CASE 1 281
*CH2Cl2 used in an early version of the process
Typical HPLC chromatogram
Column: C18 Novapak (or equivalent)
Element: 0.004 M 1-DSA Na in. CH3OH:H2O
(60:40) with 1% HOAc
Detection: UV at 254 nm
Quality criteria and analytical release specifications were set for all the solvents and
reagents used in the process. Operating conditions for the process were established
to ensure the minimum decomposition of intermediates and reagents, such as DBTA,
at the same time as maximizing product yield.
In order to minimize dialkylation of the amino group, 5-BrASA was added to a
large excess of the R-amine. This necessitated the engineering of a simple countercurrent
toluene extraction system for recycle of the excess R-amine. This first reaction
stepwas studied at a variety of temperatures, concentrations, excesses of R-amine and
extraction conditions. The process was monitored by HPLC. In summary, the following
outlines the optimum process to the sodium salt of R-aminoketone (X)—Scheme
7. The plant equipment layout is shown in Figure 1.
Although R-amine (VIII) forms a phenolate salt with phenols, no evidence of
phenolate reaction with the bromoketone was found. As expected the only byproduct
of consequence was the dialkylated R-amine (up to ca. 2%). The toluene solution
containing the unreacted R-amine was extracted with aqueous acetic acid and the
aqueous layer treated with aqueous sodium hydroxide to give neat R-amine for
recycle. The color of the R-amine did increase with multiple recycles but this did not
appear to affect the quality or yield of IX. As a precaution however, the R-amine was
distilled after about every tenth batch. Although toluene could be recovered for re-use
by distillation it was found that simple washing with c. sulfuric acid, separating the
layers and washing with water gave toluene suitable for re-use.
In a further investigation, it was found that the aqueous solution of the sodium salt
of R-aminoketone (X) could be acidified, the R-aminoketone extracted into a solvent
(for recycle)
(for recycle)
Toluene / R-Amine
Aq. Acetic
Aq. Phenolate
to reduction step
H2SO4 and Water
waste for
4. H2SO4
NaOH/NaOAc waste
for neutralization
3. Aq. NaOH
5-Bromo ASA
1. R-Amine
10% n-Butanol
Toluene +
Slurry Tank
Karr Column
Aq. NaOH
c. H2SO4
FIGURE 1. Plant Equipment for R-Amine reaction with 5-Bromo ASA (Scheme 7).
and precipitated as a pure hydrochloride. Although this step never proved necessary in
commercial operation substantial quantities of the hydrochloride of R-aminoketone
IX were produced for work on alternative routes to dilevalol (qv).
The optimum process for converting aqueous solutions of the sodium salt of Raminoketone
(X) to dilevalol DBTA salt resulted from a study of such parameters
as reaction solvent, temperature, mole equivalents of sodium borohydride, mole
equivalents of DBTA and crystallization conditions. The reduction process, which
essentially yielded a 1:1 mixture of RR and SR compounds, was monitored by
HPLC (Scheme 8). The plant equipment layout is shown in Figure 2. The process
containment equipment (Krauss Maffei Titus system) used for the filtration, washing
and drying steps is shown in Figure 3.
The purpose of the acidification step (to pH 4.5 with sulfuric acid) was to destroy
the borate esters and complexes which compromised the distribution of the RR/SR
mixture (XI) into the n-butanol layer. Considerable work was carried out on the
crystallization of the dilevalol DBTA salt in an effort to avoid oiling and to crystallize
a salt with an RR content of ca. 97%. This could be achieved by dissolution with
DBTA at ca. 55?C and crystallizing at ca. 45?C. Higher yields of lower purity proudct
(ca. 95% RR to 5% SR) were obtained by cooling to 0?C.
For day-to-day process monitoring and assay of the synthesis related impurity
levels it proved more convenient to use a thin layer chromatographic assay [with
an elution system comprising ethyl acetate (100), isopropanol (60), water (32), and
ammonia (8)] than to use HPLC.
CASE 1 283
Typical HPLC chromatogram
Column: C18 Novapak (or equivalent)
Element: 0.004 M 1-DSA Na in. CH3OH:H2O
(60:40) with 1% HOAc
Detection: UV at 254 nm
Overall yield 5-bromo ASA to dilevalol DBTA salt = 30–40%
Aq. Phenolate
Aq. Borate/Sulfate
salts to waste
n-Butanol +
(RS)R Mixture XI
Resolution Mother Liquor for
recovery of DBTA and
Racemization of SR/RR Mixture
Wet Dilevalol-DBTA
Salt Slurry to Titus
Filter Dryer
Wet Crude
Dilevalol-DBTA salt
Receiver DBTA
FIGURE 2. Plant equipment for reduction/resolution step (Scheme 8).
For assaying the enantiomeric purity of dilevalol in the DBTA salt, the Schering
Research Analytical Department worked out an efficient glc procedure, utilizing
methylboronic acid. Although the method did not separate RR and SS enantiomers or
RS and SR enantiomers, it served to quickly indicate the efficiency of the resolution
process since no racemization of the R-amine moiety was ever found. A typical glc
trace was as follows:
The specification set for the dilevalol DBTA salt was
Appearance: White to off-white solid
Chemical Purity (tlc): <3% related substances
Enantiomeric Purity: RR + SS > 97%, SR + RS < 3%
As indicated by the above enantiomeric purity assay, the quality of the DBTA
salt prepared according to Scheme 8 was often borderline (ca. 97%). A process was
registered (as an IND update) for occasional use in which the salt was split back to
the base (in n-butanol) and the DBTA salt formation step repeated. As time passed, it
became apparent that itwould be better, from an operations and economics standpoint,
to seek a higher first crop yield (by cooling to ca. 0?C) and to recrystallize the wet
first crop routinely. This position became the subject of criticism during the FDA’s
pre-approval inspection (q.v.).
Mother Liquor for recovery
of DBTA and Racemization
of SR/RR Mixture
Filter Socks
FIGURE 3. Plant equipment for filtration, washing, and drying of dilevalol DBTA salt.
CASE 1 285
Racemization and Recycle of the n-Butanol Mother Liquors from the DBTA Resolution.
The aqueous base extraction of DBTA followed by acidification of the aqueous
layer and filtration of the DBTA proved relatively straight forward. On the other hand,
much work was needed to identify and develop the most cost-effective system for
racemizing the SR/RR mixture (approximately 70:30 respectively in composition)
and to recycle the racemate produced.
Initially the SR/RR mixture was precipitated as an oxalate salt, with azeotropic
drying of the n-butanol to enhance the yield. This oxalate salt was filtered and the
washed crystals racemized by heating with aqueous sulfuric acid at 40–60?C. The
racematewas extracted into n-butanol at pH 8.3 to 8.8 for recycle to the resolution step.
This process, although workable, proved cumbersome, necessitating the handling of
large quantities of the oxalate salt, adding to equipment and labor requirements and
alsowaste disposal problems. An elegant solution to the problemwas generated by the
finding that water wet n-butanol solutions of the 70:30/SR:RR mixture, as obtained
after extracting the DBTA from the resolution mother liquors, could be racemized
utilizing a strong cation exchange resin—Dowex XFS 43279 (H+) was particularly
effective. This operation was conducted in a batch fashion using a 2000-gallon
jacketed glass-lined vessel equipped with a filter in the bottom valve for drawing
off liquids. After loading the SR/RR mixture, the n-butanol was sucked away; the
loaded resign was washed successively with one bed volume of n-butanol, followed
by water, aq 5% sulfuric acid, and water. The racemization was carried out by heating
the water resin slurry at ca. 90?C for 2 hours. The racemized SR/RR mixture was
removed from the resin by treatment with aq. sodium hydroxide/n-butanol.
The pH was adjusted to 8.5 to retain carboxylic acid (ca. 7% hydrolysis of the
amide) in the aqueous layer. The n-butanol layer containing the 1:1/SR:RR mixture
(88–90% recovery) was recycled to the resolution step. The resin was regenerated
via sulfuric acid treatment. This process was patented.10 The equipment layout for
this step is shown in Figure 4.
Dilevalol DBTA Salt to Dilevalol Hydrochloride. Very little improvement in the
RR composition results from the transformation of dilevalol DBTA salt to dilevalol
hydrochloride. Ethyl acetate was used as the solvent vehicle in early work. In a
search for a more stable solvent methyl isobutyl ketone (MIBK) was selected as the
best alternative. DBTA was removed by extraction into water with sodium hydroxide
(DBTA of excellent quality was recovered from the aqueous phase in high yield
>90%). The MIBK solution of dilevalol was then treated with hydrochloride acid
to precipitate dilevalol hydrochloride (Scheme 9). The pH needed for maximum
efficiency in the crystallization of dilevalol hydrochloride was 0.5 (this is in sharp
contrast to the pH required for maximum efficiency in the crystallization of labetalol
hydrochloride—pH 3.0). It should also be noted that dilevalol hydrochloride could
not be handled in stainless steel equipment. Hastelloy, plastic or ceramic equipment
was employed to eliminate the risk of coloration of dilevalol hydrochloride by traces
of iron compounds. The plant equipment layout for this step is shown in Figure 5.
10International Patent Application, WO 91/08196 to Schering Corporation, June 13, 1991.
DOWEX XFS 43279 (H+)
n-Butanol + mostly SR waste
n-Butanol Solution of
(SR)R mixture XI for recycle
Aq. H2SO4
Aq. NaOH
FIGURE 4. Equipment for Racemization of n-Butanol Solution of Mostly SR Waste.
The conditions used in the hydrochloride formation/crystallization stepwere somewhat
different when MIBK/water was used in place of ethyl acetate. In particular,
hydrochloride formation needed to be carried out by adding concentrated hydrochloric
acid to the MIBK solution of dilevalol base at ca. 55?C (versus ca. 25?C for
ethyl acetate). In this way, oiling out of the hydrochloride salt was avoided. A small
amount of citric acid was included in the crystallization system to chelate any traces
of iron which may be introduced. The amount of water in the system is more than
sufficient to dissolve the small amount of citric acid—in early versions of the process,
using much less water, precipitation of some citric acid caused a slight discoloration
of the dilevalol hydrochloride. The crystallization conditions were carefully chosen
to produce a crystal which filtered and washed well, which dried well (to MIBK
<0.5%) and which gave a bulk density (ca. 0.3 g/ml) which met Pharmaceutical
Development’s criteria for operation of their tabletting process.
The work carried out on the dilevalol hydrochloride step was undertaken in close
collaboration with Schering Manufacturing in Ireland, who contributed greatly to
the establishment of the IND/NDA process conditions for final API preparation. The
more aggressive process conditions of temperature, coupled with the use of a pH <
1, as employed in the MIBK-based production process, were examined in depth. It
CASE 1 287
Aq. H2SO4
Aq. NaOH
Vacuum Dryer
c. HCl
for recycle
DI Water
DBTA for recycle
Aq. Waste
FIGURE 5. Equipment for conversion of dilevalol DBTA salt to dilevalol hydrochloride
(Scheme 9).
was shown that the 50–60?C process condition in the crystallization step at low pH
did not lead to detectable recemization at the carbinol center.
The product of the simpler synthesis was compared in detail with the product of
the “RR-amine process.” In particular, the Research Quality Control Unit searched
for the presence of different polymorphs and new impurities (e.g., the dialkylation
byproduct from the first step). They compared the stabilities of both products and
also compared the hardness and dissolution rates of tablets made from both products.
Since the DBTA resolution, crystallization, and product isolation steps, as wll as the
final dilevalol hydrochloride preparation step, were the same for both the “RR-amine
process” and the simpler synthesis, it was anticipated that these steps should protect
against the introduction of new impurities or changed physical parameters in the final
crystalline product. Such proved to be the case.
Process Engineering. Chemical Development’s chemical engineers worked closely
with the chemists and analysts in the internal team created to progress the dilevalol
hydrochloride project. The engineering input and sharing of points of viewcontributed
greatly to the speedy simplification of the process and the early focus on cost reduction
through minimizing isolations and recycling solvents, as well as utilizing waste
streams. Safety issues were identified and overcome. Emissions control needs were
met. An existing plant was adapted to the requirements of the simpler synthesis.
Additional needed process equipment was evaluated, selected, purchased, and set up.
Automation opportunities were defined, and process control instruments were tested,
purchased, and installed. An existing clean (HEPA-filtered) area was upgraded for the
final isolation of the dilevalol hydrochloride made in New Jersey—this was needed
to serve the requirements for the parenteral dosage form.
The equipment flow sheets (Figures 1–5) outline vessel needs for the process.
Equipment is mostly conventional. The only equipment purchases were a 2000-
gallon glass-lined vessel for the racemization of fractions containing S-carbinol, a
countercurrent extraction column (Karr column), and the automated Krauss Maffei
Titus system. This latter piece of equipment (Figure 3) is designed for closed system
crystallization, filtration,washing, and drying. It provides nitrogen blanketing, solvent
capture, and drying capabilities under totally contained conditions. The only exposure
of operators to the hypotensive dilevalol DBTA salt is during the step of offloading the
dry powder; protective clothing is worn during this procedure. It is pertinent to add
that process containment equipment of the type of the Titus system is invaluable in the
processing of solids where dusts have explosion potential—dilevalol hydrochloride
dust, for example, was found to be more explosive than coal dust.
FDA Review and Compliance Activities. A package of information detailing the
above simpler synthesis and the definitive work carried out in Chemical Development,
Research QC and Pharmaceutical Development to show equivalence versus
the original “RR-amine process” was approved by the Review Branch of the FDA’s
Cardiovascular Division at a meeting in Rockville. This package provided the basis
for the NDA filing.
Approval of several additional process changes was sought post the NDA filing.
Although it is considered risky to request FDA approval of process changes after
NDA filing (because of the potential that changes may set back the NDA review),
Schering–Plough Regulatory Affairs was able to review additional changes, and all
the supporting data, with the Cardiovascular Division and gain agreement that the
changes were of a noncritical nature such that there was no risk of compromising the
quality of dilevalol hydrochloride. The NDA chemistry section was updated without
penalty and the changes adopted. The changes were:
1. The registration of toluene or n-butanol as the slurry solvent for adding 5-bromo
ASA to the R-amine in the first step—minimizes solids handling.
2. The substitution of sulfuric acid for hydrochloric acid to reduce the pH to 4.5
after the borohydride reduction step—reduces the risk of plant corrosion and
product contamination by iron.
3. The use of 55–60?C in the DBTA crystallization step with cooling to 25–45?C
to replace the 50?C and cooling to 0–5?C—gave more consistent enantiomeric
purity results (and slightly lower yields). It should be noted that we later reverted
to 55–60?C in the crystallization steps with cooling to 0–5?C, followed by
routine recrystallization.
CASE 1 289
4. The use of a wet n-butanol recrystallization for reprocessing out-ofspecification
DBTA salt to replace the original split back to the base and
repeating the DBTA salt formation. The recrystallization process gave a product
with higher enantiomeric purity.
5. Detail of the 50–60?C crystallization of dilevalol hydrochloride from MIBK/
water. A comprehensive comparison report with the earlier process using 25?C
hydrochloride crystallization temperatures was provided.
In today’s more formalized review climate, it seems unlikely that such initiatives
would be attempted. As a result, many companies in the Pharmaceutical Industry
have taken on the challenge of accelerating definition of the NDA process, and today
they essentially freeze the process by the start of the Phase III program.
In regard to compliance with FDA Regulations for bulk drug substance manufacture,
the Pharmaceutical Industry has, over the years, built a strong formalized
program to meet GMP requirements. The industry continues an energetic dialogue
with Regulatory Administrations around the world, primarily with United States, European,
and Japanese Agencies. Harmonization of Regulatory guidelines is a major
interest at this time.
For the manufacture of dilevalol hydrochloride (and other APIs), Schering created
needed process and control documentation and set up formal compliance programs
to ensure GMP guidelines were met. Major programs for ensuring GMP compliance
Operator and Management Training
Batch Sheet Preparation and Change Control
Materials Management and Control
Process Operation and Control
Equipment and Instrument Calibration
Equipment Monitoring and Maintenance
Validation (process chemistry, plant operation, and cleaning)
Facility and Equipment Cleaning Program
Quality Assurance Auditing and Continued Monitoring
These programs are also matched in the areas of Safety and Industrial Hygiene as
well as Environmental Compliance.
At the time of the FDA’s pre-approval inspection (our first) of the Chemical Development
dilevalol DBTA manufacturing operation (dilevalol hydrochloride itself
was manufactured in Schering’s Ireland facility), Chemical Development received an
FDA 483 notification stating that the full-time use of the butanol/water recrystallization
process for dilevalol DBTA salt was in violation of the NDA. The NDA stated
that the recrystallization process was registered for use only when the first crystallization
of the DBTA salt gave a product outside specification; the FDA interpreted
this to mean no more than about 10% of the time. Since the process had evolved
to taking a higher first crop yield of lower enantiomeric purity, followed by routine
recrystallization, the FDA criticism was justified. The fact that higher-purity dilevalol
DBTA was being produced by the change was subordinate to the wording of the NDA
for which approval was given. An NDA supplement providing detail of the reasoning
for the change, along with analytical comparison of batches made before and after
the change, was filed with the FDA and approved.
Ongoing Process Development and Alternative Routes to dilevalol
Once the recycle operations for R-amine, for DBTA, for the racemization of SR
byproduct, and for solvent recovery were in place, the simpler synthesis as described
above met the Marketing COG targets [3? cost of Labetalol minus 25%). Several
additional cost reduction programs were in hand at the time of the NDA filing. The
ones which were significant in terms of laboratory and pilot plant effort are worthy of
brief reviews. Efforts on these programs illustrate the diversity of ideas and individual
endeavorwhich flourished in the challenging climate created to solve the cost of goods
Cost reduction efforts were undertaken both inside and outside the company.
They covered the preparation of the raw materials, particularly R-1-methyl-3-
phenylpropylamine and derivatives, and also several exciting programs for the direct
preparation of dilevalol from chiral intermediates.
RawMaterials. The Celgene Company,Warren, New Jersey, building on the knowledge
that RS-1-methyl-3-phenylpropylamine costs only $12/kg, proposed an enantiomeric
enrichment process. This process, utilizing Celgene technology, is based on
the ability of omega-amino acid transaminases to preferentially convert one of the
two chiral forms of the racemic amine, in our case the S-enantiomer, to a ketone.11
In this approach the S-enantiomer acts as the preferred nitrogen source (Scheme 10).
The process evolved to one in which the converting enzyme was isolated and
used in a batch process with a small amount of pyridoxal 5-phosphate as co-factor
and pyruvate as the amine acceptor.12 This process was in a pilot plant phase when
dilevalol hydrochloride was withdrawn from the market.
11U.S. Patent 4,950,606 to Celgene Corporation, Aug. 21, 1990.
12U.S. Patent 5,300,437 to Celgene Corporation, Apr. 5, 1994.
CASE 1 291
Another initiative in the Schering Manufacturing Division was based on the idea
the N-benzylated RS-1-methyl-3-phenylpropylamine may be readily resolved, and
used advantageously in a process analogous to Scheme 3 without the phenol blocking
group. This process did indeed work well giving a high RR enantiomer yield (ca. 80–
85%) in the borohydride reduction step. The cost of the benzylation/debenzylation
steps was not worked out (for comparison with the NDA process costs) by the time
dilevalol hydrochloride was withdrawn.
Alternative Routes to Dilevalol Hydrochloride. Themajor disadvantage of the NDA
process lies in the need for the classical resolution of the RS-carbinols XI using
DBTA. Although the recycle of DBTA and the waste SR/RR mixture (ca. 70:30) did
enable the COG target to be met, it would be far more elegant and potentially lower
in cost if a more direct process could be found which would eliminate the classical
resolution and recycling operations. The following proposals were evaluated.
Scheme 11 Option. A great deal of work was carried out on the preparation of Repoxides
(XII) and their reaction with R-amine (VIII). It has long been known that
styrene oxides react with primary amines at either of the epoxide carbons, and also
that neat amines appear to favor the desired reaction, attack at the methylene carbon
atom of XII.13
The epoxide (XII, Y = C6H5CH2) was prepared in high yield and high ee(>98%)
by the enantioselective reduction of bromoketone (II), using Itsuno chemistry14 [with
R-diphenylvalinol borane complex], followed by cyclization of the bromohydrin.
Epoxide XII readily formed the desired aminocarbinol XVIII (Y = C6H5CH2) with
R-amine VIII which yielded dilevalol after hydrogenolysis of the benzyl group.
The corresponding series with the free phenol (XII, Y = H) gave a poor result.
The greatest problem with the epoxide sequence, apart from the extra blocking and
deblocking steps, lies in the need for a large excess of the expensive R-diphenylvalinol
13Parker, R. E., and Isaacs, N. S. Chem. Rev., 1959, 59, 737.
14Itsuno, S., Hirao, A., Nakahama, S., and Yamazaki, N. J. Chem. Soc., Perkin Trans. I, 1983, (8), 1673.
Itsuno, S., Ito, K., Hirao, A., and Nakahama, S. J. Chem. Soc. Chem. Commun., 1983 (8), 469. Itsuno, S.,
Ito, K., Hirao, A., and Nakahama, S. J. Org. Chem. 1984, 49, 555.
borane complex. The Coreymodification15 of Itsuno’smethod, in which only catalytic
amounts of a chiral auxiliary are needed, failed probably owing to complexation of
borane with the ring amide and phenol (or protected phenol) groups, leading to chirally
uncontrolled reduction of the keto group. Several other routes for preparing the chiral
epoxide were pursued without success. These included biological approaches to the
reduction of bromoketone (II) as well enzyme mediated selective hydrolysis of RS
halohydrin esters.
The epoxide route was eventually abandoned on the grounds that cost reduction
versus the NDA process did not appear to be attainable.
Scheme 12 Option. The oxynitrilase-catalyzed HCN addition to the aldehyde XIII
appeared to offer an attractive prospect presuming that the R-cyanohydrin (XIV)
could be formed, and this then converted to dilevalol via intermediates XV and
XVI. Although the oxynitrilase-catalyzed formation of chiral aromatic and aliphatic
cyanohydrins and their reduction to chiral aminoalcohols has been known for some
time,16 the selective reduction of XIV to XV and the likelihood of 100% induction
in the reduction of the Schiff base XVI raised many questions.
Work by our Swiss Chemical Development group demonstrated that when the
hydroxyl group of the racemic form of XV (Y=C6H5CH2) was blocked by tbutyldimethylsilyl,
selective reduction of the nitrile to CH2NH2 could be achieved
(NaBH4/CoCl2/CH3OH). However, when the free cyanohydrin was reduced with the
same reagent, only the hydroxymethyl compound could be obtained, presumably
owing to cyanohydrin conversion to the aldehyde prior to the reduction step.
In addition, our Swiss group found, in probing experiments, that the racemic form
of XV (Y = C6H5CH2) did not give the Schiff base corresponding to XVI.
Since the Scheme 12 option appeared likely to require the extensive use of blocking
groups, it lost its simple appeal and was abandoned.
15Corey, E. J., Bakshi, R. K., and Shibata, S. J. Am. Chem. Soc., 1987, 109, 5551. In this paper, Corey et
al. used S-diphenylprolinol as the chiral auxiliary.
16Becker, W., Freund, H., and Pfeil, E., Angew. Chem. Int. Ed., 1965, 4, 1079.
CASE 1 293
Scheme 13 Option. A great deal of work was carried out to find a reduction procedure
for the chiral reduction of the readily available R-aminoketone (IX). In addition,
despite the benzyl blocking group, the R-aminoketone (XVII) was also the subject
of chiral reduction work.
It quickly became clear that the use of Itsuno chemistry for reducing the carbonyl
group of XVII [with R-diphenylvalinol borane complex] would not be economic,
again owing to the need for excesses of the borane complex. The catalytic elaborations
of Itsuno’s chemistry also failed.
Biotransformations have attracted increasing attention as more chiral APIs are being
created by pharmaceutical companies. The use of microorganisms and enzymes
is particularly attractive in that systems can often be “engineered” to achieve desirable
goals. Moreover, biotransformations are generally carried out in water. It was
logical therefore to screen the microorganisms in ATCC banks which are known to
reduce ketones to carbinols. Some 50 microorganisms were screened, including bacteria
(such as Schizomycetes) and fungi (such as Ascomycetes, Basidiomycetes, and
Phycomycetes). Unfortunately, none of these was active in reducing the keto group of
R-aminoketone (IX) to the desired R-carbinol, dilevalol. A major breakthrough occurred
when Dr. William Charney of Schering’s Biotechnology Development group
observed a large underground oil storage tank being removed near his laboratory.
Dr. Charney, who was about 70 years old at the time, clambered to the bottom of
the approximately 15-foot deep pit for a soil sample. He isolated a novel fungus
from this sample which rapidly carried out the desired transformation of IX to dilevalol.
The organism was separated by the soil enrichment method, wherein the soil
sample is mixed with a compound which restricts the growth to those organisms
which can use that compound. In this case the compound was 5-methoxyacetyl-
2-hydroxybenzamide. Incubation was carried out for several days and the mixture
sampled using standard microbiological techniques and plated out. The active pure
culture was a white mold, characterized as belonging to the genus Aspergillus and
was further identified as Aspergillus niveus. An investigation of other members of
the family, Aspergillus niger (ATCC 11488) Aspergillus orxyae (ATCC 1454), and
Aspergillus oryzae (ATCC 11488), failed to provide chiral reduction of the keto group
of R-aminoketone (IX). Aspergillus niveus, ATCC 20922, was the subject of patent
An outline of the process for the biotransformation of R-aminoketone (IX) to
dilevalol at the point the project was canceled is as follows:
17U.S. Patent 4,948,732 to Schering Corporation, Nov. 7, 1989.
Since this process involved an Aspergillus fermentation as the last chemical transformation
step, considerable concern was expressed concerning possible contamination
of the dilevalol hydrochloride with such as citrinin or allergenic proteins. All
test carried out by the time dilevalol hydrochloride was withdrawn were negative.
However, a full testing program had not been completed.
It is clear from the conversion yield that the biological process was worthy of
further development for potential use in the longer term.Work was especially needed
to improve the concentration (8 g/liter at the time) and to deal with the slightly
different impurity profile (total 0.3% with 0.1% identified as the R-amine VIII). Also,
technologies (e.g. ultrafiltration) needed to be evaluated to ensure that proteinaceous
material did not contaminate the product.
Withdrawal of Dilevalol Hydrochloride from the Market
Dilevalol hydrochloride was on the market in Japan and Portugal when it was withdrawn.
Approximately 34 hepatic events were recorded in a population of 176,000
patients taking dilevalol hydrochloride. Most of the hepatic events were reversible,
but there were two deaths. During an extensive clinical research and development
program (ca. 10,000 patients), there had been no significant evidence to indicate
that hepatotoxicity would become a problem once the drug was marketed. Since
labetalol hydrochloride (which contains about 25% dilevalol hydrochloride) was
being continually compared with dilevalol hydrochloride, more detailed evaluation
and comparison of the hepatic data accumulated on both compounds was undertaken.
Interestingly, in the first year of marketing labetalol hydrochloride, 15 hepatic
events were recorded/million prescriptions. This rate dropped such that in the 7
years of marketing labetalol hydrochloride to the time of dilevalol hydrochloride
withdrawal, only 80 reports of hepatic reactions were recorded for labetalol hydrochloride.
In line with this data, it was quickly shown that the patterns of hepatic
injury associated with the two medications were not similar. In the case of dilevalol
hydrochloride, those patients affected demonstrated a fairly rapid onset of
hepatic events, expressed as showing jaundice, dark urine, nausea, vomiting, and
The withdrawal of dilevalol hydrochloride represents a unique milestone. As far
as the author is aware, this is the first case wherein a deliberately produced chiral API
may have demonstrated more toxic liability than the racemic mixture.
CASE 2 295
I am indebted to the many Chemists, Engineers, Analysts, and Regulatory people
whose dedication and hard work led to the technical success of the dilevalol hydrochloride
project. In particular, special thanks go to those who worked tirelessly to
produce the needed supplies and to engineer the technology to meet the COG target
(Messrs. Bruce Shutts, Raymond Werner, and their staffs), to those who “imagined”
and carried out the exploratory work on future options [Dr. Richard Draper (chiral
epoxides), Dr. Ingrid Mergelsberg (chiral cyanohydrins), the late Dr. William Charney
(biological reductions), Dr. Maurice Fitzgerald (N-benzyl R-amine approach)],
to senior management for their support and encouragement, to those who made suggestions
to improve the manuscript, and to Lavonne Wheeler who did all the typing.
In 1992, Schering–Plough was granted a license by Britain’s Cancer Research Campaign
Technology Ltd. (CRCT), to develop and market Temozolomide (XIX)—a
promising anticancer drug, seen as especially useful in treating gliomas (brain
1,2,3,5-Tetrazines were first synthesized in the late 1970s, and bicyclic compounds
based on pyrazoles, triazoles, and indazole (analogous to XIX) were synthesized
shortly thereafter.18 Later interest in the imidazole series centered on the chloroethyl
compound (XX) known as Mitozolomide, by structural analogy with the well-known
18(a) Ege, G., and Gilbert, K. H. Tetrahedron Lett., 1979, 4253. (b) Ege, G., and Gilbert, K. H. German
Patent 2,932,305,1990 (filed August 9, 1979). Azolo-[5,1-d]-[1,2,3,5]-tetrazine-4-ones were shown to
possess biological activity of potential value in agriculture and medicine.
antitumor agents dacarbazine (XXI) and the nitrosoureas (XXII).
N N(CH3 )2
Mitozolomide Dacarbazine
R = Cl(CH2)2 (carmustine)
R = Cyclohexyl
All are powerful alkylating agents. The development of Mitozolomide was progressed
by May and Baker in Britain. Although it showed clinical activity and a
marked advantage over Dacarbazine in crossing the blood–brain barrier, its development
was terminated when it was found to cause severe thrombocytopenia (decrease
of the blood platelet count).
Temozolomide was first synthesized by Professor Malcolm Stevens and coworkers
at Aston University, Birmingham, essentially using the same elegant but
hazardous chemistry19 described by Ege and Gilbert18 (Scheme 14). Temozolomide
is a pro-drug. Although stable under physiologically acidic conditions—enabling
the molecule to survive oral administration—it opens to the triazene XXV prior to
alkylating needed sites:
Several safety issues associated with rawmaterials, intermediates, and Temozolomide
itself needed to be addressed in manufacturing the drug. These were as follows:
19Stevens, M. F. G., Hickman, J. A., Stone, R., Gibson, N. W., Baig, G. U., Lunt, E., and Newton,
C. G. J. Med. Chem., 1984, 27, 196. (b) Lunt, E., Stevens, M. F. G., Stone, R., Wooldridge, K. R. H., and
Newlands, E. S. U.S. Patent 5,260,291,1993, to Cancer Research Campaign Technology Ltd. (CRCT).
CASE 2 297
SCHEME 14. Synthesis of Temozolomide.
1. The diazonium intermediate XXIV is unstable and is explosive under dry
2. Methyl isocyanate (b.p. 39?C—often referred to as “Bhopal gas”) is flammable
and poisonous.
3. Temozolomide itself is regarded as a carcinogen.
Creating a GMP manufacturing plant in order to protect against all the above
hazards, including the in situ manufacture and containment of methyl isocyanate,
was clearly an expensive proposition, especially since the scale of operation was
projected to eventually reach only a few thousand kilograms a year. An outside
partner was found who was both an expert in explosives manufacture and willing to
accept all the manufacturing risks in creating a production plant. The small scale and
the need to harness specialist technology and explosives experts in a GMP plant led
to a cost-of-goods that was far greater than would ordinarily be the case for a drug of
such small molecular weight (194).
Given the above analysis and also the high cost of purchasing relatively small quantities
of 5-aminoimidazole-4-carboxamide (XXIII, AIC), we in Chemical Development
were given the backing to explore other possibilities and identify a safer, lowercost
route to Temozolomide that would avoid the hazards and preferably require
conventional rather than specialist equipment. Another requirement, made necessary
by the heavy workload in our New Jersey Chemical Development operation, was that
we find an outside partner to carry out the process exploration work with us.
Exploratory Program to Identify Chemistry for the Preparation
of Temozolomide
Several initiatives essentially led us to starting a small exploratory program with
Fachhochschule Nordwest Schweiz (FHNS) in Switzerland. In beginning a working
relationship with FHNS, we decided to focus initially on finding a better (shorter and
less expensive) route for producing AIC.
SCHEME 15. Outline of a commercial route to AIC.20 .
SCHEME 16. Possible process for producing AIC from hypoxanthine.
Chemistry for Preparing AIC. This seemingly simple molecule is prepared commercially
by a multistep synthesis from cyanoacetamide (Scheme 15). The need for
several synthesis steps carried out on a small scale inevitably led to a high cost for
AIC. Recognizing these weaknesses, our senior chemical engineer, Bruce Shutts,
suggested that a shorter, simpler synthesis was needed and proposed that we look at
producing AIC by hydrolytically extruding the methine fragment from hypoxanthine
(XXVI), a compound that he thought might be low in cost (Scheme 16). Picking up
on this suggestion, our Dr. Ernst Vogel, head of our satellite chemical development
operation near Lucerne, Switzerland, learned (1997) that low-cost ($28/kg in tonne
lots) hypoxanthine, of 99% purity, was indeed available, being offered by Wenzhou
No. 3 Pharmaceutical Factory in China. Samples from China validated this information.
Dr. Vogel was also instrumental in introducing us to FHNS. Secrecy agreements
were signed and we participated with Professors Ernst Hungerb?uhler and Beat Zehnder,
and Dr. Uta Scherer in a fruitful program of work on the AIC project, essentially
focusing on Scheme 16.
An additional plus was the engagement of our Swiss consultant, Dr. Jacques
Gosteli, to aid us in the project. Probably the most indispensable resource, however,
was the agreement with Dr.Birendra (Ben) Pramanik, the head of ScheringResearch’s
Structural Chemistry group, that he and his team would aid us in analyzing the
products of our FHNS experiments. In effect, Dr. Pramanik provided the weight of
his considerable experience, especially in mass spectroscopy and nuclear magnetic
resonance, to unravel the complexities revealed by FHNS chemists and analysts.
20Shaw, G., Warrener, R. N., Butler, D. N., and Ralph, R. K. J. Chem. Soc., 1959, 1648.
CASE 2 299
The following provides an outline of the (unpublished) work undertaken to define
a route to AIC from hypoxanthine and also ideas for the conversion of AIC to Temozolomide
to avoid most of the hazards associated with the established manufacturing
Preparation ofAIC from Hypoxanthine. Asearch of the chemical literature revealed
a paper by Friedman and Gots,21 who found that hypoxanthine was stable to heating
in 1 N sulfuric acid. However, the same authors showed that when hypoxanthine was
heated in 1.5 N sulfuric acid, with zinc dust added, extensive degradation occurred.
The imidazole moiety of hypoxanthine was shown to be stable to the reducing
conditions and the product of degradation was found to be a mixture of AIC and a
structurally related compound. However, the mixture could not be separated, nor was
the unknown compound identified.
Interestingly, the literature also revealed22 that when inosinic acid was subjected
to reductive hydrolysis, the formation of 5-amino-1?-d-ribofuranosylimidazole-4-
carboxamide-5’-phosphate (AICAR) was detected. Again, spectroscopic and chromatographic
techniques revealed that an approximately equal amount of an unidentified
but structurally related compound was also produced. Although the two compounds
could be separated using paper chromatography, efforts to separate them
using ion exchange chromatography failed.
(HO)2P O
Zn/aq. HCl AICAR
Inosinic acid +
Structurally related compound
The alkaline hydrolysis of hypoxanthine to AIC has been described in the literature,23
but seemed unattractive to us. Even under rigorous reaction conditions (150?C in 0.4
N NaOH for 4 hr in a sealed tube!) only approximately 30% of AIC was produced,
with most of the hypoxanthine remaining unchanged.
We started our FHNS program by repeating the reductive hydrolysis of hypoxanthine
described by Friedman and Gots, confirming their findings. Our interest in the
reductive hydrolysis approachwas piqued by the thought that if the unwanted reaction
21Friedman, S., and Gots, J. S. Arch. Biochem. Biophys., 1952, 39, 254.
22(a) Miller, R. W., and Buchanan, J. M. J. Biol. Chem., 1962, 237, 485. (b) Wilson, D. W. Bradford
University Ph.D. Thesis, 1967, pp. 59 and 60.
23Suzuki, Y. Bull. Chem. Soc. Japan, 1974, 47, 898.
SCHEME 17. Reductive hydrolysis of hypoxanthine.
could be avoided, we might improve the yield to AIC. This led us, with the help of Dr.
Pramanik et al., to determine the structure of the byproduct in the expectation that,
knowing its structure, we would perhaps be enabled to find conditions that would
avoid byproduct formation.
The FHNS group prepared a lyophilized sample of the byproduct by chromatographic
separation from the crude mixture obtained as outlined in Scheme 17.
The byproduct was analyzed by HPLC (Spherisorb ODS-2 column) using the
following processing conditions:
CH3OH (18)/CH3CO2H containing 0.5%
H2O (82) and 5 mM sodium
Flow Rate: 1 ml/min
UV Detector: 265 nm
The HPLC chart (Figure 6) revealed the byproduct to be a mixture of a major (90.5%)
and minor (approximately 6.9%) product, with similar polarities, along with traces
(approximately 2.6%) of other substances.
At this point, Dr. Pramanik and his associates, notably Drs. T. M. Chan (NMR)
and P. Shipkova (MS), undertook the task of identifying the structures of the major
and minor byproducts.
ProtonNMRin CD3OD as solvent revealed that the byproduct contained two large
peaks at 2.83 and 2.92 ppm, consistent with methyl groups, along with two singlets
at 7.13 and 7.22, consistent with methine protons (Figure 7). This result suggested
that each of the impurities may be carrying a methyl group and an imidazole ring.
NMR also revealed the presence of smaller quantities of other impurities, including
impurities carrying methyl groups.
The NMR results are in agreement with earlier work21–23 showing that the pyrimidine
ring of hypoxanthine and its glycosides is far more susceptible to cleavage than
the imidazole ring.
This information, and recognition that the only source of the methyl groups was
through reductive hydrolysis of the methine group in the pyrimidine ring of hypoxanthine,
led to the postulation of structures XXVII and XXVIII, respectively, for the
CASE 2 301
Absorbance (AU)
2 4 6 8
Retention Time (min)
10 12 14
Chrom Type: Fixed WL Chromatogram, 265 nm
FIGURE 6. HPLC analysis of byproduct from the reductive hydrolysis of hypoxanthine.
FIGURE 7. Proton NMR of the byproduct (mixture) obtained from the reductive hydrolysis
of hypoxanthine.
major and minor products.
(MW = 140) (MW = 140)
FIGURE 8. EI–MS of AIC itself.
FIGURE 9. EI–MS of byproduct mixture.
CASE 2 303
FIGURE 10. APCI–MS of Peaks A and B.
Dr. Petia Shipkova undertook a rigorous mass spectral evaluation of the byproduct
mixture and comparison with the MS of AIC (XXIII) itself. The EI–MS fragmentation
of AIC (Figure 8) and the byproduct mixture (Figure 9) revealed consistent mass
ion peaks (126 and 140, respectively) and interesting revelations in the fragmentation
Dr. Shipkova followed the EI–MS analysis with an APCI–MS evaluation of the
mixture showing that the two peaks representing compounds (XXVII) and (XXVIII)
(Peaks B and A, respectively) both gave fragmentation patterns that were very similar
to each other (Figure 10).
Dr. Pramanik and co-workers, Drs. Shipkova and Guodong Chen,24 reasoned that
the virtually identical fragmentations and peak intensities of Peaks A and B could
24We are indebted to ProfessorMichael Gross,Washington University, St. Louis, for his guidance regarding
the structures of the mass spectral fragments.
only be accounted for by the two samples being mixtures of XXVII and XXVIII.
Despite this, they proposed that the fragmentation patterns of AIC and byproducts
XXVII and XXVIII could be explained by such breakdown losses as assigned in
Table 3.
They went on to suggest that the structural fragmentations needed to account for
the major mass peaks could be occurring as follows:
m/z 126
m/z 81
Fragment m/z 54
m/z 82 m/z110
m/z 55
+. +.
N-Methylamine (XXVII)—APCI-MS
m/z 141
m/z 124
m/z 124
m/z 141
m/z 110
CASE 2 305
N-Methylamide (XXVIII)—APCI–MS
m/z 141
m/z 110
m/z 110
m/z 141
m/z 124
m/z 124
The information from mass spectrometry, although incomplete from a mass spectroscopist’s
viewpoint, was enough to justify the chemical development search for a
process that would avoid byproduct formation.
The structural revelations for the major products from the reductive hydrolysis
of hypoxanthine, coupled with the earlier observation that hypoxanthine does
not degrade when heated with 1 N sulfuric acid alone, suggested that the dissolving
zinc caused addition of hydrogen to the methine nitrogen double bond in the
pyrimidine ring of hypoxanthine. It then followed that the dihydrohypoxanthine
produced underwent hydrolytic cleavage and reduction, perhaps along the lines of
Scheme 18.25
Confirmatory, if incomplete, support for the structures of the N-methyl-amine
(XXVII) and N-methyl-amide (XXVIII) byproducts was obtained from reactions
with AIC as follows26:
25We have no rigorous proof for the intermediacy of the two N-hydroxymethyl compounds versus the
alternative of direct addition of hydrogen in two ways.
26This work, and especially the MS/MS analyses, revealed that the peak intensities in the fragmentation
of XXVII and XXVIII were different and more consistent with expectation.
TABLE 3. MS Fragmentation Assignments for XXIII, XXVII, and XXVIII
AIC (XXIII) N-Methylamine (XXVII) N-Methylamide (XXVIII)
Mass Ions Losses Mass Ions Losses Mass Ions Losses
126 M+ — 140 M+ — 140 M+ —
123 [M ? 17]+ NH3 123 [M ? 17]+ NH3
110 [M ? 16]+ ·NH2 110 [M ? 30]+ ·NHCH3 110 [M ? 30]+ ·NHCH3
109 [M ? 17]+ NH3 109 [M ? 31]+ CH3NH2 109 [M ? 31]+ CH3NH2
95 [M ? 45]+ NH3 + CO 95 [M ? 45]+ NH3 + CO
83 [M ? 43]+ HNCO
82 [M ? 44]+ CONH2
81 [M ? 45]+ CO + NH3
68 [M ?72]+ NH3 + CO
+HCN  68 [M?72]+ NH3 + CO
54 [M?72]+ NH3 + CO
CASE 2 307
*This structure may dehydrate to the formaldehyde Schiff base prior to the
hydrogenation step leading to (XXVII).
aq. H2SO4
SCHEME 18. The reductive hydrolysis of hypoxanthine.
 AIC was reacted with formaldehyde under reducing conditions (zinc/sulfuric
acid) and the reaction mixture (after “purification” along the lines of Scheme
17) subjected to HPLC/MS and MS/MS analysis. This analysis showed that the
main product was the N-methylamino compound (XXVII). Not surprisingly,
the N,N-dimethylamino compound (Mass 154) was found to be present in
significant amount; it is pertinent to also note the presence of traces of N,Ndimethyl
compound in the byproduct mixture from the reductive hydrolysis of
hypoxanthine (Figure 9).
 AIC was heated with methylamine under pressure at 70–80?C for several hours,
and the reaction mixture was subjected to HPLC/MS and MS/MS analysis.
This analysis showed that the expected N-methylamide, (XXVIII), had been
produced, along with a number of other products (not identified).
 In a related exercise we also reacted formaldehyde with AIC under acidic
conditions (no zinc). HPLC/MS and MS/MS analysis of the crude complex
product indicated the presence of a small amount of a compound (mass 138)
consistent with dihydrohypoxanthine (XXIX), or the formaldehyde Schiff base
of AIC.
The identification of the N-methylamine (XXVII) and the N-methylamide (XXVIII)
during the reductive hydrolysis of AIC raised the prospect that these byproducts
might be avoided if the reduction step was separated from the hydrolysis step.
Reduction of Hypoxanthine to Dihydrohypoxanthine. The decision to separate
the reduction step from the hydrolysis step led to a further search of the literature
H+ H2O
SCHEME 19. Polarographic behavior of hypoxanthine (XXVI).27
pertaining to the reduction of purine and its derivatives. Bendich et al.27 reported that
neutral or acidic solutions of hypoxanthine did not absorb hydrogen at room temperature,
under one atmosphere of hydrogen pressure in the presence of palladiumcharcoal.
Later, however, Smith and Elving28 described the electrochemical reduction
of purine and a few of its derivatives (adenine, hypoxanthine, and guanine).
Prophetically, they interpreted the polarographic behavior of hypoxanthine in terms
of a 2-electron reduction leading to dihydrohypoxanthine, which then hydrolyzes
(Scheme 19).
None of the compounds in the sequence was isolated or synthesized, but their
work validated the idea that the best approach to preparing AIC from hypoxanthine
was likely to result from initially reducing hypoxanthine to its dihydro derivative and
then hydrolyzing the XXIX produced to AIC.
In view of the widespread adoption, by industry, of catalytic hydrogenation, it was
deemed most appropriate that, in the first instance, we focus effort on the catalytic
hydrogenation of hypoxanthine rather than the electrochemical reduction approach.29
We also thought it better to avoid the presence of inorganic salts, as used in the buffers
employed in the polarographic work, in order to avoid the complication of separating
AIC from salts. The main thrust thus became to hydrogenate hypoxanthine in an
anhydrous solvent, to give dihydrohypoxanthine (XXIX), and then to separately
hydrolyze XXIX to AIC. A solvent that could be readily removed by distillation was
considered desirable.
Dr. Gosteli pointed out that Schering had earlier patented the hydrogenation of
indoles in trifluoroacetic acid (TFA) using platinum oxide as a catalyst.30 Based on
this information, an evaluation of the hydrogenation of hypoxanthine over platinum
oxide, in TFA and a few other acid solvents, was undertaken—partly to determine
whether the pKa of the acidwas a factor in the hydrogenation. No significant hydrogen
uptake was observed when hydrogenations were carried out (1 atmosphere and room
temperature) in formic acid (pKa 3.75), dichloroacetic acid (pKa 1.48), sulfuric acid
27Bendich, A., Russell, P. J., and Fox, J. J. J. Am. Chem. Soc., 1954, 76, 6073.
28Smith, D. L., and Elving, P. J. J. Am. Chem. Soc., 1962, 84, 1412.
29Later, Dr. M. H?urzeler-M?uller andMr. F. Stapf (FHNS) demonstrated that the electrochemical reduction
of hypoxanthine could be achieved using a mercury cathode and pH 4.65 acetate buffer. The yield of
XXIX was 83%.
30Neustadt, B. R., Smith, E. M., Magatti, C. V., and Gold, E. H. South African Patent 8600083, 1986 to
Schering Corporation.
CASE 2 309
4 6 8 10
Retention Time (min)
12 14 16 18 20
Chrom Type: Fixed WL Chromatogram, 265 nm
N-Meth 0 N-Meth 0
Seno 10
FIGURE 11. HPLC analysis of product of hydrogenation of hypoxanthine.
(pKa ?3 and 1.96), and triflic acid (pKa ?8?). Some hydrogenation was observed
in methanesulfonic acid (pKa ?1.2), with the product being primarily AIC (due
to small amounts of water?). Trifluoroacetic acid (pKa 0.5) was the only acid that
allowed substantial hydrogen uptake. This “solvent” was considered to be attractive
because of its low boiling point (72?C), allowing for easy recycle—a necessary
requirement owing to the relatively high cost of TFA($12–15/kg in tonne quantities in
Although in the few months before financial support was terminated there was
little time to optimize the hydrogenation and hydrolysis steps, the best hydrogenation
conditions, giving a yield of dihydrohypoxanthine of >90%, were found to be as
Hypoxanthine (2 g) in trifluoroacetic acid (40 ml) was stirred and hydrogenated
(100 bar/35?C/21 hr) over platinum oxide (0.1 g).31 The catalyst was filtered and the
trifluoroacetic acid stripped under vacuum. The product was shown (potentiometric
titration) to contain 1.9 moles of TFA.
The product was analyzed by HPLC (Spherisorb ODS-2 column) under the following
process conditions (Figure 11).
CH3OH (9)/Phosphate buffer (pH 3) with 50
mM/liter potassium dihydrogen phosphate and 5
mM/liter sodium heptanesulfonate adjusted to pH
3 with 85% H3PO4.
Flow Rate: 1 ml/min.
UV Detector: 265 nm
31Later, Dr. Uta Scherer used 0.2 g PtO2 in this experiment and obtained dihydrohypoxanthine in a yield
of 96.8% in 8.5 hr at ambient temperature.
FIGURE 12. Proton NMR spectrum of dihydrohypoxanthine (XXIX) trifluoroacetate.
FIGURE 13. EI–MS of dihydrophypoxathine (XXIX) trifluoroacetate.
The NMR spectrum (Figure 12) was consistent with the dihydrohypoxanthine
structure, though minor unidentified impurities are also present.
The mass spectrum (Figure 13) corresponds in most ways with those of AIC and
its N-methyl derivatives, with a closely related pattern of fragmentation (Table 4).
The structural fragmentation corresponding to the major mass peaks may be
accounted for as follows for the TFA salt of dihydrohypoxanthine:
CASE 2 311
TABLE 4. Mass fragmentation assignments for dihydrohypoxanthine (XXIX)
Mass Ions Losses
139 (weak) [M+1]+ Protonated M
138 M+ —
137 [M?1]+ ·H
122 [M+1?17]+ or [M?16]+ NH3 from protonated M or NH2 from M
110 [M+1?29]+ or [M?28]+ CH2=NH from protonated M
XXIX Trifluoroacetate —EI–MS
m/z 139
m/z 122
m/z 122
m/z 139
m/z 110
The large MS peak at 69 may be the CF3 radical derived from trifluoroacetic acid.
The peak at 126 may represent AIC from partial hydrolysis of the sample.
Hydrolysis of Dihydrohypoxanthine to AIC. There was sufficient time to demonstrate
that dihydrohypoxanthine could be hydrolyzed to AIC but not enough time to
provide the basis of a workable process to isolate AIC.
20 40 60 80 100
Time (minutes)
120 140
dihydrohypoxanthine in TFA/water
AIC in TFA/water
dihydrohypoxanthine in water
AIC in water
160 180 200 0
Relative area
FIGURE 14. Hydrolysis of dihydrohypoxanthine to AIC.
As illustrated in Figure 14, dihydrohypoxanthine (containing ?1.9 moles of tri-
fluoroacetic acid) is hydrolyzed to AIC by refluxing in water or a 1:1 mixture of
water and trifluoroacetic acid. Figure 14 (HPLC monitoring) indicates that some loss
of the AIC produced is occurring in the trifluoroacetic acid–water case (perhaps by
hydrolysis to the aminoimidazole carboxylic acid or through reaction with released
formaldehyde, or both?).
The fate of the released formaldehyde was not determined. For best results it may
be necessary to distill the formaldehyde or scavenge it, preferably using such as a
phenol immobilized in polymer form to allow ready separation from the aqueous
solution of AIC.
This is essentially where the project ended, but it is pertinent to outline some
of our work on the preparation of temozolomide itself, especially our efforts to
avoid preparation of the potentially explosive diazonium intermediate (XXIV in
Scheme 14) and the use of methyl isocyanate.
NewChemistry for the Preparation of Temozolomide. The three main disadvantages
of the present synthesis (Scheme 14) were outlined earlier. Of these only the first
two, the potential explosivity of the diazonium intermediate (XXIV) and the toxicity
of methyl isocyanate, can be addressed. In the first, the handling of the potentially
explosive intermediate is best accommodated by ensuring that this material remains
wet at all times. Thus, avoiding the preparation and use of methyl isocyanate became
the primary objective.
The use of masked methyl isocyanates has been described,32 but this seemed to
us merely replacing a known hazardous compound with an isocyanate of unknown
toxicity which could well be equally hazardous. In this spirit our Dr. Shen-chun Kuo
32Shutts, B. P., Stevens, M. F. G., Thomson, W. T., and Wang, Y. J. Chem. Soc. Perkin I, 1995, 21.
(Step 1)
where X is an "appropriate" leaving group
(Step 2)
.Step 3
SCHEME 20. Speculative synthesis of temozolomide.
speculated that it may be possible to avoid isocyanates by using a synthesis sequence
such as outlined in Scheme 20.33
Although methylhydrazine has a higher boiling point than methyl isocyanate (88?C
versus 39?C), and would therefore be expected to be easier to contain, it is a toxic
and hazardous compound (rocket fuel). However, methylhydrazine does not have the
emotive baggage associated with the Bhopal disaster caused by methyl isocyanate.
On these grounds, work to determine the feasibility of Scheme 20 was considered a
worthwhile objective.
In the first instance, p-nitrophenyl chloroformate was selected for the acylation of
AIC (Step 1 in Scheme 20). Reaction ofAIC hydrochloride (0.154 mole) inmethylene
chloride first with an excess of triethylamine (0.323 mole) and then with a solution of
p-nitrophenylchloroformate (0.169 mole) in methylene chloride at room temperature
for 24 hr gave compound XXX (X = p-nitrophenoxy) in ? 93% yield.
Reaction of XXX (0.144 mole) in dimethylformamide with methylhydrazine
(0.188 mole) at 0?C for 1 hr and the reaction mixture quenched into ethyl acetate
gave compound XXXI in ?95% yield.
The cyclization of XXXI was studied only briefly but shown to be feasible. Thus,
a very dilute solution of XXXI (2.5 mmol), along with tetrabutylammonium iodide
(0.25 mmol) in 500 ml 1:1 tetrahydrofuran and acetonitrile, was heated at 60?C for
20 min, cooled to room temperature, treated with periodic acid (5 mmol), and stirred
vigorously for 1 hr. The excess periodic acid was destroyed by adding saturated
aqueous sodium thiosulfate (5 ml) and the solution concentrated to dryness. The
residue was treated with acetonitrile (200 ml), filtered, and chromatographed on a
silica gel column (1.5–2% CH3CO2H/EtOAc) to afford temozolomide in 58% yield.
In the decade following the above feasibility studies, much has happened to drive
the pharmaceutical industry in new directions and, in particular, to affect the pursuit
33Kuo, S. C. U.S. Patent 6,844,434,2005, to Schering Corporation.
of chemical process development in the pharmaceutical industry. Some of the major
events are identified below:
1. Because the process of obtaining regulatory (FDA) approval for an API has
become more exacting, more paper intensive, and therefore longer, the pharmaceutical
industry has tried to find and overcome sources of delay in gaining
FDA approval. In chemical process development terms, this has led to efforts
to minimize process changes post the IND filing, and particularly in
Phase III—significant process changes raise the risk of introducing new impurities
and cause delays in the filing of the NDA. As a result, free wheeling
(innovation-driven) efforts to identify, develop, and introduce the most costeffective
process for the manufacture of an API are mostly restricted to the
pre-IND phase and post-NDA approval. API manufacturing sites thus usually
receive only Methods or, at best, partially developed synthesis processes for
the production of new API’s. By avoiding major process changes between the
IND filing and the NDA approval, companies advance their NDAs faster and
accept the heavier analytical, regulatory, and manufacturing burdens and the
higher cost of goods that go with using partially developed processes.34 The
product of any innovative process work which can be carried out post the IND
is generally never qualified for use in any Phase III studies that will appear in
the NDA.
2. The majority of pharmaceutical companies have moved away from the manufacture
of their own intermediates, and in many cases their own APIs, with the
result that outsourcing has become a major program in chemical process development
organizations. With more parties involved, and the precise transfer
of technology requiring great effort and vigilant follow-up, API costs and the
time frame needed to develop the best process have escalated.
3. The trend to outside involvement can compromise a company’s control over its
own destiny as far as intellectual property and sundry regulatory obligations
are concerned. Thus chemical process patents, which often extend a company’s
exclusivity on an API, may be more in the hands of others, requiring that
legal protection of the company’s position be given due attention. Outside
involvement also brings concerns as to whether or not a partner can meet all the
safety, environmental, FDAregulatory affairs, and confidentiality requirements,
thereby creating a need for additional legal contracts.
As a result of the above trends, many innovative opportunities for reducing the
cost of APIs are put on the back burner awaiting developments which, in those APIs
that reach the market place, may bring resurrection. The growth of the market for
temozolomide may be just such a case.
34There is also a knock-on effect in the pharmaceutical sciences area since they need both to validate that
changes in the API synthesis do not affect the dosage form, and also introduce beneficial improvements
of their own.
From the foregoing, it appears that relatively little work would be needed to
develop a practical process for the manufacture of AIC (XXIII) from hypoxanthine
(XXVI). Much more workwould be needed to develop a process based on speculative
Scheme 16, but the cost reduction opportunities appear considerable. If successful,
significant reduction of the API cost of goods would be achieved by developing a
safer process using conventional equipment. Such a process, allowing escape from
manufacturing in an explosion-proof and contained environment, would reduce plant
depreciation overheads and allow ready development of a second source for insurance
of supply. In addition, the patenting of a superior process33 essentially extends patent
protection for the API by restricting eventual generic competitors to the use of the
current hazardous and expensive process, a situation that may even deter them from
competing at all.
The significant problems we face today cannot be solved at the same level of thinking
we were at when we created them.
——Albert Einstein
One cannot think of the global future without it stirring anguished feelings of excitement,
tempered by trepidation; optimism, tarnished by cynicism; despair, lifted
by hope; and nightmare confusions that cannot be reduced to words. It may seem
impossible to imagine the diverse forces now driving the world, coalescing and collaborating
for the good of the planet and its occupants, without our first going through
some cataclysmic event humbling us into creating a mostly harmonious, constantly
evolving steady state. But imagine, and encourage collaboration, we must.
Globally, the inability of some populations to develop,1 leading to diversity, disagreements
and the unhealthy evolution of the “me-phenomenon,” set cultures apart.
Achieving world intra- and intersocial harmony now only seems possible by generating
a global educational curriculum to create understanding and formulae for
reform of the world’s antagonistic driving forces. Creating new interpretations of
such driving forces as democratic capitalism, socialism and religion demands a much
1One eloquent explanation is provided by Jared Diamond in Guns, Germs and Steel, W. W. Norton and
Co., New York, 1999.

Copyright C 2008 John Wiley & Sons, Inc.
more highly educated population. Melding of these reformed driving forces through
a harmonized curriculum seems to offer the most likely path to long-term survival.
This chapter, however, is primarily concerned with creating action strategies for
pursuing relatively small specific missions applicable in the chemical process development
field. My views merely add to the existing dialogue to create a consensus
for moving forward. Since, to paraphrase John Donne, “no single discipline is an
island. . .,” moving forward requires a multidisciplinary assessment of the potential
value of proposed future directions in order to provide an agreed basis for initiating
action to create an end product, in our case a holistic commercial chemical
process operation. The incorporation of some global issues, such as promoting an
interdisciplinary education (chemistry, embracing interacting disciplines), reducing
government bureaucracy, recognizing that the world’s organic feedstock is finite, and
global warming, essentially reflect the need to consider global issues in everyday
thinking, even in small endeavors.
The subjects I have addressed in the form of short essays are outlined below,
concluding with a fantasy that to many may seem a far-fetched approach to solving
one of the world’s major “chemistry” problems:
 Bureaucracy Reduction
 Trends and Technologies:
 Notes on synthesis
 Water and enzymes in chemical process development
 Thoughts on other “solvents”
 Polymer-supported synthesis and reagents
 Microwave–assisted chemistry
 Sustainable development
First I would like to express a few “global thoughts” about education. All of us need to
regard education as a spirit of pursuit, not only in one specific discipline but as a route
to diverse matrix style development—say along the lines that the Therapeutic Teams
in the pharmaceutical industry use to promote drug development (see Chapter 3). It
seems to me that up until the early decades of the 20th century, education was better
regarded than it is today—the teacher, minister, doctor, and lawyer were thought of
as largely equal in importance as pillars of the community. In the twentieth century,
industry and business became more dominant and lawyers and doctors managed to
leverage their professions into more visible high-standing roles in society. Western
capitalism ushered in great material wealth leading to enormous social freedom and
liberalism. Education lost ground as decades of industrial growth enticed people into
jobs enabling them to gain immediate material gratification in the course of pulling
themselves out of the poverty of more socially challenged times. To some extent
this 20th-century trend, into feeding the needs of industry and its growth, skewed
education toward short-term practical development. Although unprecedented social
good was realized, few in power anticipated or dealt with the unhealthy consequences
of the subsequent excessive growth of the “me-phenomenon” and the discord and
friction generated in both the have-not communities at home as well those in the non-
Western world. In different ways, both felt disenfranchised and exploited. In short,
education was unwittingly hijacked by industrial, business, and attendant political
and social development and accorded a shabbier role than was needed of it.
Fortunately, education is still an evolving and revolutionary force as families often
slowly recognize that their children would benefit from getting as much education
as they can possibly take. Wider recognition of the vital role of education in shaping
the future would allow the education profession to leverage its importance to
individuals, families, captains of industry, politicians, and others, thus establishing
education as the prime vehicle for improving the social system. In order to contribute
more effectively in shaping a better future, the education profession needs
massive politically driven funding to reach out, not only to those with a hunger for
education, but also to those it loses, often at the grass roots, that is, the downtrodden
and forgotten. People need to be lifted to a level where they can more fully express
themselves, be appreciated for their contribution, and realize a satisfactory life. In
this way they preserve dignity through feelings of being useful and being well regarded,
even as they become educated enough to recognize and deal with their own
limitations. Supporting education is not something the Western world needs to do
merely through money and physical resources. Education needs to be placed on a
war footing through a major unrelenting effort to create a permanent self-sustaining
hunger for learning. It is just as important (arguably more important) to “educate” the
disenfranchised (those unable to get the most out of today’s formal education system)
as it is to educate tomorrow’s “elite.” Such a commitment needs to be aimed, inter
alia, at overcoming enormous social problems (including compliance, top to bottom,
with law and order2) and creating dignity in all forms of work. Over time, such a
2The political dialogue in Britain in 2004 between Prime Minister Blair and opposition leader, Michael
Howard, bears on this point:
Blair (July): “A society of different lifestyles spawned a group of young people brought up without parental
discipline, without proper role models and without any sense of responsibility to or for others. All of this
was then multiplied in effect by the economic and social changes that altered the established pattern of
community life in cities, towns and villages throughout Britain and throughout the developed world. Here,
now, today, people have had enough of this part of the 1960’s consensus. People do not want a return to
old prejudices and ugly discrimination. But they do want rules, order and proper behaviour. They know
there is such a thing as society. They want a society of respect. They want a society of responsibility. They
want a community where the decent law-abiding majority is in charge; where those that play by the rules
do well and those that don’t, get punished.”
commitment might be expected to create better understanding of the great social
needs, to enable adaptation to the changes needed to adjust to emerging economies,
and to deal with the challenges that a burgeoning developing population poses to the
developed populations, to the environment, and, eventually, to planetary stability.
Present-day educators are either not capable of meeting the challenges or do not
have the needed support and encouragement to revolutionise the educational system.
Today’s forms of education cannot be really fully effective unless they are integrated
with the wider social and survival needs. Many young minds are lost by failed
anemic institutions ineffectively dealing with dysfunctions in society. A few of those
who acquired the least formal education rely on their wits, and often extraordinary
“testeronic zeal.” Many pursue criminal careers. These are the lost clever minds in
the greatest danger of creating something that leaves an atrocity in the wake of their
“profit.” Education needs to provide for them (and their family “mentors”) while at
the same time minimizing the crimping effects of formalization on their creativity.
Themore educated we are, the easier it is for us to gain amodicum of understanding
of the problems we all face; to be accepted as participants in the dialogue with
those shaping the future (primarily politicians, captains of industry and social and
religious leaders); to be effective in contributing to the solution of problems; and
to be successful in overcoming the worldwide frictional forces which create social
discord everywhere.
Today the capitalist economies are far too soft and comfortable and are not desperate
enough to rebuild decaying social foundations. We have led ourselves deep
into our land of plenty by neglecting the consequences of our journey to get there.
Dealing with consequences was not considered relevant—they did not seem to matter
when we started out. But it has now been evident for some time that, in John Muir’s
words, “When we try to pick out anything by itself, we find it hitched to everything
else in the universe.”
Our problems only seem to be exacerbated by today’s efforts to maintain and
increase our well-being by increasing the speed of our ongoing exploitation of available
physical and human resources. We are on a treadmill going faster. The need for
changes in behavior becomes both more imperative and equally readily postponed.
John Kenneth Galbraith once put it that “Faced with the choice between changing
one’s mind, and proving there is no need to do so, almost everyone gets busy on the
Dealing with consequences will be enormously difficult and expensive and will
require bitter personal, social, industrial, and religious revolution in the way humans
coexist with everything on the planet to reach something approaching a “constantly
evolving steady state.” Since today’s industrial, political, social, and religious systems
Howard (August): “Most damaging of all has been the decline in personal responsibility. Many people
now believe that they are no longer wholly responsible for their actions. It’s someone else’s, or something
else’s fault—the environment, society, the Government. The decline of responsibility and the proliferation
of rights have left us in an ethical quagmire, which is undermining our fight against crime. The clear
distinction between right and wrong has been lost in sociological mumbo-jumbo and politically correct
nonsense. Conservatives will stand up for the silent law-abiding majority who play by the rules and pay
their dues. We will put their rights first.”
cannot, at present and by themselves, be relied on to work without regulators, getting
to the steady-state level becomes an almost incomprehensibly difficult and forbidding
task. Transcending the present missions of today’s major forces can only be done
through a massive program of education for survival. Most present-day educators
cannot succeed in this on their own.
It is no easy matter to live up to the political slogan “no child left behind.” The fact
that many are behind is mostly a consequence of our social systems losing their way.
The solution is not simply an education matter. Revolutionary investment is needed
to overcome the dysfunctions in families, communities, and countries at their source,
to correct the lack of fairness in society, the corruption and crime, and so on. One way
may be to deal with one of the greatest challenges, to greatly weaken and restructure
the “me-phenomenon” in society (and its attendant malaise, delegating philosophical
and social change to the next generation) at the same time as minimizing impairment
of the search for new initiatives to enable the Western world to meld with the social
and physical challenges posed by the emerging economies. Further education is vital
to harmonize the emerged and emerging economies through preparing the rich for
the more frugal material but richer cerebral and physical lifestyles needed to adapt
to a steady state. The social philosophies emerging under monarchies, dictatorships,
communism, and socialism were all corrupted into self-preserving philosophies that
intensified their inability to develop the kind of missions needed for permanent
survival. Today’s democracies seem in danger of following suit. As far as “no child
left behind” is concerned, many are trying to move now.3
If the developed nations do not deal with their excesses and do not progress
education, innovation, and change, they risk foundering, like the old empires, giving
way to an even more perilous repetition of the same cycle of rise and fall—and
eventually, perhaps, Armageddon. Unfortunately, present-day inertias suggest that
it may take a cataclysmic event, perhaps taking the form of a slowly developing
“environmental catastrophe,” rather than a flash viral pandemic, making it more
difficult to deal with, before anything is done. Would it be too late?
From time to time small reassurances appear, often only to evaporate in the heat
of short term budget priorities.4 Politically, leaders with staying power are needed to
3They do not want to wait until social programs are integrated into educational ones. William L. Taylor, a
retired lawyer, lobbyist, and U.S. Government official, one of the leading spokesmen in urging government
progression of the “no child left behind” program, put it: “To say just give more money is not an answer. it
is not good enough to say this is a societal problem, though that certainly is the case. If you say everybody
is responsible, nobody is responsible.”
4A recent report, commissioned for the National Academy of Sciences (Rising above the Gathering Storm:
Energising and Employing America for a Brighter Future, 2005), and a bipartisan Congressional Bill to
“Protect America’s Competitive Edge” generated a flurry of initiatives among Senators and Congressmen
to deal with the need to invest more in enhancing education and supporting more basic research. One of
the major concerns was to find ways of dealing with evidence of decline in the achievement of 12th-grade
students in mathematics and science—U.S. mathematics and science 12th-graders rank in the bottom 10%
among their international peers (Markoff, J., New York Times, February 2, 2006). Congressional legislators,
aware of the need to update their science knowledge, also organized a scientific briefing by The Center for
Health and the Global Environment at Harvard in January 2006—perhaps a harbinger of further education
and action proposals to come? (See Dean, C., New York Times, January 31, 2006.)
define and progress solutions and to encourage the population to persevere with their
further education.
Progressing Science Education
Science is only one root feeding a complete education. The “magic” of science is
appreciated by the general public in those areas that touch their everyday lives, particularly
the publicized components that create the “news.” These include medicine
(e.g., health, diet, disease, surgery, and treatment), chemistry (pharmaceuticals, polymers,
chemical safety and toxicity, explosives, analytical chemistry), environmental
sciences (pollution, global warming, declining biological diversity), food sciences
(agriculture, fishing, brewing), biological sciences (evolution, genetic engineering,
virology), material sciences (forestry, glass, ceramics), physics (particle physics,
atomic fission and fusion, Einstein’s theories, astronomy, computer sciences), geology
(earthquakes, volcanoes, mining, oil and gas), engineering (construction, prosthetic
devices, flood levels), and so on. Mathematics, on the other hand, is the language
of the universe (all other sciences are merely dialects as a Connecticut mathematics
professor once put it!). Although mathematics seems remote and austere, simplified
connections are made by the general public through computer sciences, economics,
and statistics. In reality, we need to appreciate that all of the above disciplines are
closely integrated, indeed embedded with “rules” from the others, in creating the
“magic” of sciences.
Outside the “magic,” people conveniently “opt out” and separate benefits from consequences.
This leads to science being widelymisunderstood and misinterpreted—its
dominant role in daily life seems to become dominating and resented. Most people,
it seems, cannot be encouraged to give up a little of the time spent on addictive
“diversions” to grasp and deal with all the uncertainties at the interfaces where the
“magic” of science, nature, and factual understanding come together. The realization
of wider public support and broader understanding can only come if more people can
be stimulated into wanting to learn more about science in a way that enables them
to gain satisfaction from understanding the problems and goals of science as applied
for social good. Not everyone will be willing, nor should they be, to sacrifice the
qualitative feeling approach to their lives. No doubt many find mystical comfort, as
well as fear, in the unknown and seemingly unknowable. Without detracting from
this, there is much to be gained by greater efforts to increase everyone’s curiosity in
the “magic” of science.
To make science more digestible and to provide a modern foundation, many
schools and universities are already working to define and teach a canon of basic
knowledge and fundamental principles which integrates the above components,
stimulates curiosity, and thereby provides a basis for a rewarding career.
Chemistry is one of the relatively poorly regarded branches of science, despite
providing the architectural foundation of our everyday lives. To make the subject
more mainstream, a broad universal redefinition of what scientific material should
be included in chemistry courses is needed, building on what has been started. The
knowledge base today is so vast that selecting the chemistry to be taught, enabling
students to grasp the fundamentals and at the same time to encourage curiosity, is
a daunting task. Many inputs are needed to agree on content, at the same time as
avoiding rigidity. My thoughts are that a wide range of other disciplines needs to be
incorporated from the biological sciences (biochemistry, botany, and microbiology,
for example), from physics (e.g., fundamental matter, thermodynamics, kinetics), and
from inorganic chemistry (noncarbon elements/compounds in life processes, in catalysis,
in energy generation). More specifically, regarding the biological sciences, the
relevance of pre-biotic chemistry, evolution, electrochemistry, and photochemistry
to life, and the chemistry/biology associated with amino acids, peptides, proteins,
enzymes, and DNA seem to offer opportunities to stimulate curiosity. Again, specifi-
cally, an introduction to modern analytical instruments and their application to quantitative
analysis and an understanding of what is going on in chemical reactions would
provide students with a sound basis for quantitative thinking and for experimentation.
Analytical chemistry should cover a body of basic knowledge from potentiometric
titration all the way to spectroscopy.5 Perhaps, above all, an education in chemistry
needs to emphasize the experimental nature of the subject. Experimentation enables
students to intellectually engage and feel the “magic.” In short, the classical courses
of time past need to continuously evolve and to be continually revitalized in order to
raise chemistry to the level of an essential discipline that will enhance the lives of all
who participate.
There is, within the broad subject matter of chemistry, the opportunity to be absorbed
and stimulated by the esoteric and/or to find equal satisfaction in a devotion to
the practical interdisciplinary applications. A chemistry education in either a prestigious
university or a school of technology can lead students either way, depending on
themselves. In Europe, universities are usually seen as classical while the technology
schools lean more to the practice of chemistry. In Europe there is a distinguishing balance
between the two kinds of education, with, perhaps, a nod in the technical to the
old apprenticeship system of qualification. America, for the most part, hardly distinguishes
the classical (Ivy League) schools, as well as the large number of second-tier
(definitely not second-class) schools, from the great Institutes of Technology. America
embraces more of its graduating high school population in a third-tier community
college system than does Europe, and many of the graduates from these schools find
their way into jobs all across the business spectrum. However, while there are many
good technical and trade schools in America, there is no nationally standardized
technical education in chemistry of the sort available in the major science countries
of Europe. In particular, few promote a coherent long-term interaction with the industries
and businesses in their areas. This is partly because there are fewer places
in America where businesses are really concentrated. There are no equivalents to
the concentration of the Swiss pharmaceutical industry around Basel (New Jersey is
similar but more dispersed), or the huge chemical/pharmaceutical complexes around
Ludwigshafen, Leverkusen, and Frankfurt in Germany, orManchester/Liverpool and
the Tees-side in England.
5An excellent place for undergraduates to start is Organic Spectroscopic Analysis,Anderson, R. J., Bendell,
D. J., and Groundwater, P. W. Royal Society of Chemistry, Cambridge, England, 2004.
Ph.D. chemists joining the pharmaceutical industry in America come, literally,
from all over the world (one of the great strengths in America). The support and
technical personnel usually come from more local universities and technical schools.
Most need quite a bit of training in the practical chemistry techniques and most have
had no, or little, experience in working, like the old-style apprentice, in industrial laboratories,
particularly with the kind of thinking and modern instrumentation needed
to address problems in a multidisciplinary and quantitative way.
Building a broader foundation for technical training in the United States, especially
in places where there is a fairly heavy concentration of a specific industry
(e.g., the pharmaceutical industry in New Jersey), could be achieved through sustained
industrial-university collaborations in pertinent scientific disciplines. A few
such collaborations go on already, but not with the permanence or depth to be
found in European models. One such model I have seen at work in Switzerland
might be advantageously adapted to enhance the education of scientifically and
technically oriented students to the mutual benefit of both universities and industries
in appropriate regions. Switzerland has set up institutes of higher learning,
all around the country, which have strong programs of teaching usually greatly
relevant to meeting the technical needs of businesses/industries all across the country.
In the Basel area the Fachhochschule Nordwest Schweiz (FHNS) provides a
technical education in the disciplines of chemistry, mechanical/chemical, and some
biological engineering and electronics/computer sciences. FHNS’s location draws
on the strengths of the region. FHNS is very well equipped with analytical instruments,
computers, and a pilot plant for practical chemical and chemical engineering
The chemistry department provides a basic education in the conventional disciplines
of organic, physical, inorganic, and analytical chemistry and, in keeping with
its focus on technical education, provides a great deal of practical training as well
as instruction in safety/environment technology, economics, law, organization and
management. FHNS draws on the technical expertise of specialists from industry
in its teaching activities. FHNS introduces its students to practical chemical work
through programs on synthesis and practical topics, where possible of some industrial
significance. It also seconds students into industrial laboratories to gain experience
in technical problem solving in the industrial world.
The chemistry department emphasizes teaching in the acquisition of practical
skills. Students learn how to go about:
 Searching for information in the literature and electronic databases
 Handling chemicals and running chemical reactions safely
 Acquiring analytical instrumentation skills.
 Understanding quality and gaining skills in efficiently producing chemicals of
standard quality—they also learn the basics of GLP and GMP
 Learning how to deal with unsuccessful experiments, how to think about their
problems, how to improve on their results, and how to guide and motivate others
in these areas
 Dealing with “customers,” taking responsibility and taking the iniative to repeat
experiments when “customers” are not satisfied
 Collaborating with others and working in teams
 Communicating with and presenting work to others, including writing reports
 Becoming especially aware of the importance of safety, environmental, and
economic matters in creating and running chemical processes
One way FHNS courses help students to appreciate the importance of the above is
through creating teams of four students per team, each team being asked to find a
chemical reaction in the literature which should be relatively low in cost to reproduce
and also safe to run. The objective in reproducing the reaction is to develop it
for scale-up. One student is appointed team leader, responsible for communicating
with others and keeping everyone informed on the work program. Each student is
allowed to change one parameter in the published reaction (e.g., a solvent to improve
regulatory compliance). After the first student has completed the reaction, the results
are discussed within the team. Each student justifies to other members of the team
the reasons for making a proposed change. The students add “know-how” to each
experiment and make a final recommendation, in the form of a short report, on what
further work they think should be done on their modified literature procedure to make
it suitable for scale-up. Staff members are always available for answering questions
and for general guidance.
FHNS’s industrial liaisons are of enormous value to the students, enabling them
to gain some understanding of the importance of organization, partnerships, innovation,
and commercial interests and particularly the place of their technical work
in advancing the goals of an increasingly responsible industrial sector. Industrial
liaisons promoted by FHNS and local industry sometimes leads to the industrial
partner working with FHNS in the FHNS pilot plant.
In the above, I have tried to illustrate the importance of education in creating the
interdisciplinary interactions needed to progress chemistry and increase appreciation
of its importance in the modern world. More national publicity of all sciences is
essential to increase realization of the vitally important role that science will play
in the future. More early exposure to the “magic” of experimentation would help
to inspire people to follow a science career. People need to be galvanized to want
to learn more. Chemical Process Development is one of the more practical areas of
chemistry and, as the thrust of the earlier chapters illustrates, one which particularly
requires that practitioners embrace many other disciplines in order to create success
in any mission.
The bottom line is that educators everywhere would benefit from an International
dialogue with their counterparts to create the chemistry courses needed to equip
students for a future career in chemistry.
Many have written, sounding the alarm, on the subject tof overregulation and its
impact on the future. First it is important to say that a strong regulatory framework is
needed to establish standards, to deal with theirmaintenance, with unwittingmistakes
and particularly with the excesses of the many individuals and businesses around the
world who work to create advantage for themselves by bending the rules or criminally
exploiting perceived opportunities to profit, and so on. If all people were honest and
socially responsible (i.e., by being broadly better educated), we would have no need
for the excessive regulations of today. Unfortunately, in developing regulations to
build and enforce standards and to counter unwitting mistakes and irresponsible
opportunists, the regulatory “industry” has gone too far.
Overregulation is now a fact of life. The legal profession grew dramatically over
the last 40 years or so and is now in danger of inhibiting progress, becoming like
the Trade Unions! Trade Unions once provided, and particularly in the developing
world they still do, the much needed power base to protect industrial working people
from raw exploitation by industrial and bureaucratic barons who had little concern
for their workers. Despite all the good they did in the past, today’s Trade Unions, at
least in the developed world, still carry too much adversarial and self-preservation
baggage and are more seen as societies for the promulgation of mediocrity—in short
they have not fully evolved with the times. To be fair, Unions are still needed in some
cases where managements have not done enough to foster better relationships with
their workers. In parallel with Trade Union developments, today’s regulatory/legal
profession has grown to the point of promoting and defending the introduction of
new laws, many of which are grossly more costly than the value they provide.6
In looking at past major documents (“laws”) that established social standards and
provided social guidance, one is struck by their brevity and simplicity. The words
needed to convey the spirit of social development were relatively few. Granted,
dealingwith infractions, each on its ownmerits, in a common lawsetting undoubtedly
introduced more complexity (and words) in particular cases, but the generally lawabiding
honest citizen was allowed to run his/her life with relative ease compared
with today. To illustrate, consider the number of words in a few of the major guidance
 The Ten Commandments—297 words in the King James version, Exodus 20.7
 The Magna Carta (English translation of the Latin original)—4482 words.
 The American Declaration of Independence—1332 words.
 The Bill of Rights (the first ten amendments to the American Constitution)—461
 President Lincoln’s Gettysburg Address—269 words.
6In addition, in Europe, EU laws are often introduced on top of member country laws without guidance
on the melding of the new laws with existing laws or creating “sunset provisions.”
7The 100-Minute Bible reduces the Ten Commandments to 59 words.
Compare the simple beauty of these documents with the mountains of paper issueing,
for example, from the European Union; for instance, the European Initiative on
Caramel Products, issued in the 1980s, uses about 250,000 words (about one-third
of the words in the entire Bible)!! Such excesses become the source of ridicule and
an embarrassment to the legal profession. Even the legal profession itself recognizes
that there are many idiocies in the law. For readers interested in this area, I suggest
they read a lawyer’s view on the subject.8
Bureaucracy in Chemical Process Development
Getting down to the basic issue in the present-day pursuit of chemical process development,
most practitioners are appalled by the growth of restrictions on the initiatives
of chemists and chemical engineers to change chemical processes for themanufacture
of APIs. Granted it is not too difficult to make changes before the IND, or even in the
early steps of a process post the IND (i.e., before Phase III—see Chapter 3, Figure 2),
provided that one can demonstrate sufficient advantage and no effect, or a beneficial
effect, on API quality. However, post the IND, process changes, especially late in
the synthesis sequence, are very difficult to introduce and it is virtually impossible
to promote a revolutionary new synthesis. This is largely because of regulatory concerns,
often most zealously applied by a company’s internal regulatory watchdogs
(who seek to keep the company “whiter than white”). Restricting change is perfectly
understandable to those working to avoid delays since they see restrictions as virtually
eliminating the chances that the quality of the API or its biological effect will
be adversely affected (e.g., by new impurities); they eliminate the need for the validation
of changes (which can be enormously resource- and time-consuming), they
eliminate the need to divert often limited regulatory resources into the documentation
of changes, and they eliminate the need to go back to the FDA for approval (both
also resource-consuming and time-consuming). By avoiding all of these activities,
companies feel that they are eliminating the chances of delay in developing, filing,
and gaining approval of their NDAs.
The major consequence of virtually eliminating post-IND changes is that API
processes essentially become no more than expensive methods, refined into often
inadequate processes. Such method processes consume enormous time and present
often great difficulty and expense in technology transfer and manufacture since they
use sub optimal chemistry (e.g., the employment of chemicals you would rather not
use), they require more and sometimes specialized manufacturing equipment, they
introduce unwanted waste-disposal and environmental challenges, and they employ
far more labor/support resources than a real developed process would use. As a
result, the cost of manufacturing APIs is generally far greater than it should be.
Pharmaceutical companies are willing to pursue this strategy to get their drugs to the
marketplace as quickly as possible, thereby realizing profit as soon as possible. The
argument is that this is a reasonable strategy since, generally speaking, the cost of the
API is only a relatively minor component of the cost of the marketed drug.
8The Death of Common Sense—How Law Is Suffocating America, Phillip K. Howard, Random House,
New York, 1994.
There are twomajor downsides to this strategy, in addition to the lost cost-reduction
opportunity. First, process development chemists and engineers are greatly diverted
from undertaking the sort of intellectual/innovatory work needed to find the best
process chemistry and to develop it into a real commercial process. Second, once an
NDA process is approved, the time taken to fully resume innovation and prove a new
process, as well as to undertake all the testing, validation, and regulatory approval
work and to invest in new plant, transfer technology, and so on, adds yearsmore to the
time frame.My two Case Studies neatly illustrate the situation (q.v.). In the Dilevalol
Hydrochloride case, in the 1980s (see Case Study 1), we were able to demonstrate,
late on in the program to file an NDA, that a better manufacturing process had been
proven, in scientific terms.We shared the datawith the FDAas we were implementing
the new process. They agreed with our initiatives. In complete contrast, 10 years later,
a seemingly better process for the manufacture of Temozolomide was not pursued
despite the promise of dramatically lower costs and replacing the use of specialized
equipment with conventional equipment (see Case Study 2). Granted the seemingly
better process had not been developed to the point of demonstrating that it was better,
largely because the future market projections at the time did not justify the effort.
Since that time, market growth has exceeded all expectations, but the seemingly better
process system remains in limbo.9 I should make it clear that I am not suggesting we
return fully to the Dilevalol Hydrochloride situation. Nor do I think that we should
simply adopt the thinking, which seems current with some companies, namely that
they establish, with the regulatory authorities, a chemistry/manufacturing/controls
(CMC) section that qualifies a very late intermediate as the starting material for the
API synthesis. Qualifying a late intermediate does not free the company from all
regulatory control in early intermediate manufacture since the company still needs
to build quality into the intermediate by using an approved sequence of reactions
and process conditions, all overseen by an exacting analytical control system (see
Chapter 6). In addition, qualification of a very late intermediate does not give the
synthesis chemist or engineer a carte blanche to change the synthesis sequence, or
introduce a revolutionary synthesis. This is because process changes, especially those
attendant on a revolutionary synthesis, might reasonably be expected to introduce
new impurities. Just the very threat of new impurities would require considerable
analysis and lead to process qualification and validation work to prove that the new
process intermediate, and the API made from it, gave the same quality drug product
as registered in the NDA. From this it can be seen that generating the creative
chemistry component of a chemical process takes much less time than the sum of
the analytical work (providing assurance that quality has not been compromised), all
the pharmaceutical development work (particularly to show that the drug product’s
physical parameters and stability are unaffected), all the additional validation work,
and all the supporting documentation. Add to this any plant modifications, along with
9Comparison with the fires that destroyed La Fenice opera house in Venice, 170 years apart, also illustrates
the impact of time on getting things done. In 1836 the Austrian government, then in power, took less than
a year to rebuild the opera house. After the 1996 fire, started by two misguided electricians, it took the
Italian government nine years to rebuild.
the manufacture of three new batches to enable demonstration of API equivalence,
and one quickly sees that the time and cost of introducing a new process becomes
greatly extended. Clearly an enormous amount of effort, time, and money is needed,
under present rules, to introduce new chemistry post the NDA.
Companies engaged in new ventures would gain some relief if suggestions by
the late Peter F. Drucker had been adopted. He advocated allowing new venture
companies to charge the government for all the efforts taken to meet the regulations.
Unfortunately, such an approach does not get to the heart of the problem but it would
certainly have encouraged new thinking!
We need to find a way, acceptable to the regulatory authorities, of allowing process
innovation to continue, without interruption, not only during the IND to NDA
phase of API process development, but continuously throughout the life of a product.
Before I retired from the Schering–Plough Research Institute, I proposed a simple direct
chromatography-based alternative to the present “no-process-changes-allowed”
strategy with the objective of keeping process innovation and change going continually
throughout the IND–NDA time frame and beyond. Nothing was done, perhaps
because there were downsides I did not think of. Nevertheless, I thought it would be
worthwhile summarizing my reasoning and suggestions for greatly reducing regulatory
bureaucracy in the hope of stimulating debate enabling process development
chemists and engineers to promote innovation overcoming the present discontinuities
and costs.
The world of manufacturing APIs is essentially built from two components. One
component is the chemical/biological synthesis scheme for producing the intermediates
used in the final construction of the API. The other is the API synthesis itself,
governed by cGMPs. Today cGMPs greatly affect much of the first component as well
as dominating all of the second component. Rather than provide for split activities,
companies (trying to be “whiter than white”) err on the side of operating virtually
everything under cGMP control. In my view they should not. We need a way of
operating that restricts the heavy documentation/validation aspects of cGMPs to the
last API steps and, in exceptional cases, to the penultimate step. I wish to stress that I
am not advocating adoption of a free-for-all in the manufacture of intermediates and
APIs and therefore leaving all purification to a final chromatography step. Nor am
I advocating making synthesis changes on the basis of good science, adopting them
and then gaining FDA approval (as with the Case Study on Dilevalol Hydrochloride).
We have to recognize, as scientists and engineers, that the FDA changed the landscape
for producing APIs. Today responsible chemists and engineers have refined
and formalized their approach to chemical process development to firmly and irrevocably
build it on sound science. They have also adopted a collegiate, collaborative
approach to process development, working closely with all those who have a voice
in creating a commercial process—principally personnel in analytical departments,
pharmaceutical development, safety, environmental, manufacturing, and regulatory
affairs disciplines. Companies, especially the major pharmaceutical companies (and
most of the fine chemical companies which serve them), have built solid reputations
with the FDA for creating sound processes. There is no doubt in my mind that the
FDA should take a great deal of the credit for raising and maintaining professional
standards; they also need to be continually involved in auditing to ensure that high
standards continue to be maintained in qualifying newcomerswho seek to be suppliers
of intermediates and APIs.
Given that high professional standards in the field of chemical process development
have been secured in responsible companies and given that FDA audits of
pharmaceutical companies continue to ensure the maintenance of high standards,
the quality of APIs only very rarely becomes an issue today. Such high standing
established with the FDA could, I believe, be taken to the point where the responsible
companies are given the freedom10 to continue innovatory process development work
and implementation of new processes on a continuous basis, including through the
IND toNDAphase, given that the responsible companies add one safeguard—namely
a validatable chromatographic purification step at the API formation step or the previous
step. I would not make this chromatography step a clean-up step for any kind of
chemistry done in preceding steps. Chemists and engineers in designing innovatory
new processes must keep a record, as now, showing that they really understand the
chemistry being introduced, including establishing a mass balance accounting for
byproducts and impurities and showing that they can be eliminated or reduced to
very low levels before the validatable chromatography step. The chromatography
step thus serves as a bulwark process step to provide a guarantee of quality. I see the
chromatography step as an aid to “building in quality” and not a means to remove
rubbish from sloppy chemistry at earlier steps. In this risk-based approach Iwould ask
that the FDA allows pharmaceutical company chemists and engineers the freedom to
undertake any chemistry they think can improve intermediate quality or process economics
without affecting API quality, at all steps up to the final or penultimate step,
without needing more than their professional notebooks and SOPs to prove they are
meeting the needed quality criteria in the final API. I would still insist on subjecting
the early step activities to independent (QA) audit but not subjecting process changes
to the bureaucratic approval requirements seen today.11 In reality, as in the Dilevalol
Hydrochloride case, responsible chemists and engineers do not need the draconian,
professional-morale-busting attention (however well-meaning) of bureaucrats trained
to drive compliance activities by a book of regulations. Nevertheless, for everyone’s
peace of mind, auditing of the early steps must be continued.
I recognize that the cost of introducing a chromatographic process step will, correctly,
be regarded as expensive,12 but I hypothesize (I have only a feeling based
on experience rather than a proven example) that the cost of creating and validating
a chromatographic step or steps (including capital costs, operating costs, solvent
costs, cGMP, and validation costs) will be far, far less than the costs, outlined earlier,
associated with developing and implementing a method-process and delaying
innovation to a later date. As seen in the case of Temozolomide, it takes a Herculean
10Perhaps such freedom could be accommodated under the refined FDA guidance role outlined in its
publication Pharmaceutical cGMPs for the 21st Century—A Risk-based Approach.
11In some respects, qualifying the process steps up to the chromatography step is similar to qualifying
plants that produce chemical products under the European ISO 9000 system (see Chapter 6).
12Except when chromatographic purification is successfully applied to aqueous solutions (see footnote 7
in Chapter 9).
effort (and cost) to replace even a poor and very expensive manufacturing process
post NDA approval—in my view the potential lost opportunity cost (i.e., savings in
manufacturing cost) that might have been realized has probably been enormous.
To fit with the philosophy expressed in the chapter on regulatory affairs, I would
argue that the proposed chromatographic purification step be built into the program
for working out the last step first (see Chapter 6). In short, the bulwark qualityassuring
chromatography step should be built in from the very beginning of the API
supply program.
The introduction of a chromatographic purification step, while at the same time
allowing well-worked-out process changes, or new chemistry, to be introduced cautiously
but continuously, places added requirements on validating the final step(s)
[from the chromatographic step to the final API]. There will need to be assurance
that the chromatography step itself is not introducing new factors (e.g., traces of
resins13) to the API. Work may also have to be done up front if new impurities
(which must never exceed the 0.1% level) are introduced into the API, despite chromatographic
purification. It is essential to ensure that no unwanted toxicity has been
introduced (as was the case when process changes were made in the l-tryptophan
process (see Chapter 6). Isolation/preparation of introduced new impurities and their
toxicological evaluation should, generally, be undertaken to provide assurance on this
The chromatographic separation and purification of organic compounds has been
widely practiced, particularly in the anti-infectives field, for several decades. Examples
include elution from resin-extracted fermentation broths (e.g., aminoglycosides),
absorption and selective elution from filtered broths (cephalosporin C and
derivatives), and the purification of proteins (interferons), inter alia. The merits of
chromatography as an industrial scale purification technology have been greatly extended
in the last few decades to other fields and APIs—for example, the cyclic
heptapeptide, Integrilin, an antithrombotic, and the synthetic substituted phthalan,
Citalopram, an antidepressant. In short, despite the perceived costs associated with
chromatography as a process step, I suggest that when considered in the same context
as the consequences of inhibiting process innovation post the IND phase of chemical
process development, as outlined above, the routine introduction of a chromatographic
process step in every API synthesis should be considered as a bargain.
Chromatographic purificationmight not be applicable to every API synthesis—for
example, when the API itself is very insoluble. However, in these circumstances,
ways might be found of introducing the technology, say at the penultimate step, in
the bulwark purification of more soluble fragments.
The range of chromatographic purification technologies is very broad, ranging
from the relatively low-cost water-based classical resin systems (see footnote 12) on
13Such might be removed by an additional step (e.g., ultrafiltration, removing particles in the 4- to 25- °A
range or reverse osmosis, removing particles in the 1- to 5- °A range) which will introduce its own validation
14Where the difference in the structure of the new impurity versus the original one is relatively small (as
in structures XIIIa and XIIIb in Chapter 9), an expert judgment may eliminate the need for additional
toxicology work.
to related solvent-based systems and on to high-performance liquid chromatography
(used in the production of tonnes of Integrilin) and simulated moving bed chromatography
(used in the production of tens of tonnes of citalopram). The separation
of optical enantiomers on a large scale represents one of the more valuable contributions
made by chromatography specialists to the pharmaceutical industry; this is
surely a growing field as single-enantiomerAPIs are increasingly being recommended
for development.
It behooves anyone interested in testing the above concept to take a proven system
to the FDA to gain their feedback, to examine their suggestions for fine-tuning
the specific case, and, ultimately, to gain FDA approval. It follows that ongoing
process changes, in terms of outlining new chemistry and providing the data showing
that API quality is unaffected, will need to be lodged with the FDA. These steps,
without the current regulatory bureaucracy governing early steps, not only have the
potential to greatly reduce API costs but also spur thinking chemists and engineers
to express themselves freely, to the benefit of companies and chemists/engineers
In the pharmaceutical industry, regulations are essential for the protection of the
public against egregious events (however innocent) that adversely affect API quality
and place the public at risk. During the last 30 years the FDA has been themajor force
in causing the pharmaceutical industry to quantify its science and to create systems
guaranteeing they produce quality APIs. In the slow progress toward “perfection,”
rules and guidelines, abetted by company failures and aided by the growth of often
draconian internal Quality Assurance and Audit groups, have generated a bureaucracy
that has greatly suffocated the innovatory spirit in chemical process development
as well as in other departments responsible for API production, at least beyond
phase III. In this 30-year time period, quantitative science has come to the fore, and
process understanding has enabled scientists to create processes that control quality to
exacting FDAstandards, even though the process chemistry has often been suboptimal
in terms of molecular elegance, plant and labor requirements, environmental, and
ultimately cost considerations.
Stifling innovation has had a significant adverse impact on the cost of drugs
and particularly on the creative scientists and engineers frustrated in their efforts to
express their creative talents. Recognition of the adverse effects of bureaucracy on
the rate of progress toward lower cost drugs and on company competitiveness has led
to the present proposal to try to keep innovation going through the simple device of
introducing a validatable chromatographic purification step at the final stage(s) of API
manufacture. By introducing such a step, without sacrificing scientific understanding
and control in the earlier process steps, the innovation process could continue and be
implemented on a continuous basis.
Readers may have other and better ideas and should be encouraged to express and
evaluate them.
Notes on Synthesis
Organic chemistry, on its own, is no longer the frontier discipline it was. Its enormous
contribution to the identification of the chemical structures of diverse natural products,
along with the enhancement of the properties of these products by structural
manipulation, led to the applications that created major new industries and social
prosperity. In those days, roughly the first half of the 20th century, organic chemistry
became the cornerstone of such industries as the dyestuff industry, the pharmaceutical
industry, the agricultural chemical (including insect control) industry, and the
fiber industry. These industries, and the explosives and polymers industries, were
greatly enhanced by experimental research, often simply empirical synthesis endeavors.
The rapid expansion and diversification of all these industries, which occurred
in the second half of the 20th century, led to greater appreciation of both the good
and the adverse effects of the organic chemicals being synthesized, which stimulated
the search for better molecules and ways of dealing with adversity. Hybrid
disciplines (e.g., chemical engineering, biotechnology, the ecological sciences) were
developed and industry, often prompted by social activists, began the process of integrating
practical organic chemistry with many other disciplines (see Figure 3 in
Chapter 3).
Schools and universities teaching chemistry never really accommodated the needs
of industry to integrate chemistry with other disciplines and lost ground as some of
the newer integrated disciplines, particularly environmental sciences,marine biology,
forensic sciences, and political sciences, emerged in the education curriculum and
the so-called pure sciences—chemistry, physics, and mathematics—attracted fewer
students. Chemistry departments in institutes of higher education need to integrate the
newdisciplines into their fundamental chemistry courses and also recognize that many
of their students will want to branch out into the newer disciplines. By not embracing
the practical new areas, chemistry has limited itself. This trend has undoubtedly been
a factor in the closing of a few university chemistry and mathematics departments,
or their merger into a general sciences curriculum. Understandably, left to their own
devices, the newer hybrid courses “cherry pick” by only incorporating those parts of
a pure chemistry course which they see as relevant to their own course. This leads
not only to loss of much of the fundamental rigor of the pure sciences but also to
a tragic loss of historical perspective and that indefinable “magic of science” which
are so important to future creativity. In synthesis terms, if the foundation for building
imaginative ideas is not there, we will not be able to reach the intellectual heights
needed to progress.
Organic synthesis is evolving rapidly and I am not close enough to current research
to identify those ideas that will create the best chemical process development
prospects; the ideas needed will also be dependent on the structure of the molecular
targets. I can only take shelter behind parts of an inspirational statement15 on the
15Danishefsky, S. J. Tetrahedron, 1997, 53, 8689–8714.
challenges facing synthesis chemistry made several years ago by one of America’s
great organic synthesis chemists, Professor Samuel J. Danishefsky, in paying tribute
to Dr. Sarah Jane Etheredge:
The opportunities for discovery are greater than ever for those who are willing to study
and practice synthesis with scholarly dedication and experimental exactitude.
The playing field of synthesis today encompasses all but the rarest elements of the
periodic table. The debt of total synthesis to methodology development goes well beyond
the convenient availability of many new methods, important as they are. The new
technologies liberate and, indeed, beckon the architects of synthesis to think in much
broader and sweeping terms about tomorrow’s problems. Clearly, the most dramatic
advances have been registered from the mobilization of transition metals and other
organometallic reagents to achieve specific transformations even in multifaceted contexts.
It is well to recognize that these breakthroughs were, on the whole, achieved
by scholars of chemistry and even curiosity seekers—unconcerned with any apparent
application to total synthesis. The synergism of methodology, mechanism, and strategy
constitutes the core of synthesis.
There is a diminishing need for the logistically intensive multistep assaults simply
because the mountains are “there.” The syntheses that will warrant the greatest interest
are those that convey new ideas and new chemistry arising from a willingness to explore
ambitious and risky propositions. It is in the context of dreaming such dreams—and,
above all, in the struggle to reduce them to a “do-able” state—that the power of our
science, as well as its beauty, flourishes.
The devotees of synthesis have good reason to be particularly optimistic about their
field. The opportunities in the design of high “value added” structures of either theoretical,
material science, or biological impact fire the imagination. Moreover, as bioassay
systems become more and more sophisticated, and as more lead compound types,
including structurally fascinating natural products, become increasingly amenable to
deduction at the level of mode of action, the number of potential projects of high
promise will continue to increase.
Quite properly, organic synthesis will be drawn to multidisciplinary undertakings.
I would urge that, in these ventures, the synthetic organic chemist assume a significant
leadership position. Those who accomplish the synthesis of a target are apt to
have gained a privileged vantage point as to its true molecular nature. However, it is
in a class by itself in terms of its capacity for creation. To fully exploit this power,
chemists must be particularly well informed and venturesome in the broader contexts
and applications of their accomplishments. Only through such activism can the
formidable heuristics inherent in organic chemistry find full expression in multifield
The future will be particularly bright for those who sort carefully and select wisely
from an ever expanding menu. Again, I urge the emerging leaders of tomorrowto conduct
their synthesis morewith daring and imagination and less with reflexive recourse to welltrodden
paths. In such settings, synthesis will surely provide more magic moments, first
to its creative enthusiasts and then to the larger scientific enterprise and, hopefully, to
the public we all seek to serve.
The outsourcing of many functions in the process of developing a drug to the marketplace
is a long-standing but still growing practice in the pharmaceutical industry.
Outsourcing enables companies to leverage their physical and intellectual assets, enabling
them to do more with their core resources and, at least partially, to transform
themselves into more of a guiding service organization. In the heavily regulated pharmaceutical
industry, outsourcing requires a major dedication of pharmaceutical company
resources to ensure the receiving company (RC) has the physical and intellectual
capabilities needed to take on the work to be outsourced. Once the credibility of the
RC has been established, legal agreements are usually reached, including on recompense
for and ownership of any RC discovery of advantageous (e.g., cost-reducing)
intellectual property, and the supervision of technology transfer begins. Following the
transfer of technology and associated regulatory responsibilities, the pharmaceutical
company becomes involved in monitoring and periodically auditing the RC to ensure
regulatory compliance and the achievement of yield/quality expectations.
By adopting outsourcing as a legitimate practice, pharmaceutical companies usually
gain significantly, especially from overseas transfer, where, for the moment,
much lower operating costs often prevail. They may also gain from reducing the internal
regulatory burden—most API manufacturers have more regulatory oversight,
to ensure they are “whiter-than-white,” than is generally required. Another major
gain can be in capital expenditure avoidance for the physical facilities needed to
produce the transferred intermediate or API. Even if the pharmaceutical company
has the facilities, it may prefer to deploy the scientists and engineers not involved in
technology transfer in the development of other experimental API candidates. Some
tax advantage may also accrue from transfer to another country.
There are of course many risks associated with outsourcing. The technical capabilities
of the RC may not be sustainable if key scientists/engineers leave them. In these
circumstances, API quality, and even intellectual property, may be compromised.
Safety and environmental commitments may also be endangered. The reliability of
supply (mostly quality, delivery times, and costs, in that order) from the RC will
depend on many parameters, including the personnel they assign to the work, the
RCs investment in needed capital equipment and analytical instrumentation, their
raw material sources and quality, their support resources and commitment, and their
professional training to undertake needed tasks. Outsourcing companies also need
to guard against changes in the RC’s status. For instance, takeover of the RC could
change legal agreements. For these reasons, pharmaceutical companies need to have
contingency plans to deal with cataclysmic events, such as the possibility of an Asian
disease pandemic. Cultivating a second source elsewhere is an essential component
of every technology transfer scenario that has developed, or appears to be developing,
to a production scale.
Outsourcing to reliable RCs in underdeveloped nations (e.g., India and China)
provides enormous benefits to these nations, greatly enhancing their economies and
stimulating the educational process needed to develop national culture—not only to
deal with introduced environmental and safety issues but, not incidentally, to fund
efforts to deal with their own “socio-politico-religious” imbalances. The outsourcing
nation faces different consequences, especially where people in the developed world
are laid off when their work moves overseas. Increasingly, all the developed world’s
populations need to be prepared for this. If education can be transformed into a
lifelong spirit of pursuit, those put out of one kind of work may be in a better position
to re-engineer themselves to do another. To support re-engineering, the governments
of outsourcer nations could impose a kind of tax on the outsourcer, say in the form
of obliging the outsourcer to spend part of the “outsourcing profit” on funding the
research needed to create “new industries” and/or improve existing ones (e.g., by
tackling environmental issues) and to help re-educate those thrown out of work.
Experience suggests that the outsourcer tax would best be spent in the private sector
rather than by big government. As an aside, few of the millions who are or will be put
out of work will benefit from such a task in the short term. They will need support
and training supplied by programs such as the U.S. government’s Trade Adjustment
and Assistance programs, originally developed in the 1960s to aid those affected by
tariff cuts. It will be a harrowing process for the displaced, but taxing to support
education and R&D will eventually lead to new industries and improvement in core
competencies in both industry and the workforce.
Returning to the main theme, the company personnel responsible for and undertaking
technology transfer need to be well-versed and trained in all the disciplines
required to ensure that a chemical process conducted elsewhere is carried out to meet
both the NDA requirements and local situations.16 In transfer overseas, personnel
also need to be educated in the customs and culture of the receiving country to ensure
that the technology transfer is not compromised by misunderstandings.
The above indicates that outsourcing, rather than being merely an instrument to
reduce short-term costs, should also be a spur for longer-term innovation to ensure
competitiveness. Since the companies overseas who do the work also see a new
opportunity for innovation, the onus on the outsourcer is to ratchet up the discovery
effort to a bigger picture, higher level—embracing, for example, more innovative
programs to identify “new businesses” and new patent protections. Another example
would be to adopt a spaceship earth mentality by supporting innovation to deal with
the future impact of more rigorous environmental and safety laws.
It is clear that outsourcing, and dealing with the risks associated with
outsourcing—ensuring quality, suffering time delays, losing intellectual property,
addressing the downside of domestic layoffs—will be in vogue for as long as cost advantages,
in terms of cheap labor and low-cost capital equipment (and, unfortunately,
relaxed environmental and safety laws), continue to hold in developing countries. It
16When Schering–Plough transferred a corrosive reaction to China, the Chinese RC said they would prefer
to work in a low-cost regular steel reactor available in China rather than go to the expense of purchasing a
Hastelloy reactor. This could only be agreed to by repeatedly running the reaction in the presence of the
Chinese regular steel and establishing that leached metals did not compromise process yield or product
quality. This proved to be the case, leading the Chinese to invest in a system whereby a second back-up
reactor was available on site to replace the corroded reactor when it failed. The turn-around time in
removing the corroded reactor and installing and testing the new one proved to be remarkably fast.
will probably take decades for outsourcing harmonization to be achieved—that is, to
the point that carbon dioxide emissions and “chemical miles” (the distance molecules
have to travel from source to final API) and the application of worldwide standard
safety/environmental laws will enter into the cost equation. Until that time, outsourcing
nations need to spend that outsourcing “tax” on innovation (and patents) and
education to maintain competitiveness, and the receiving countries need to pass and
enforce safety and environmental laws and promote (also through education) those
facets of their social, cultural, and economic life needed to transform themselves into
responsible, developed nations. In conclusion, it seems to be particularly necessary
for both outsourcer and “outsourcee” to aggressively address the many opportunities
and consequences associated with outsourcing in order to reach a constantly evolving
steady state where invention benefits all and where waste and antisocial, self-serving
activities are addressed to the satisfaction of all parties involved.
Water and Enzymes in Chemical Process Development
Water is the “solvent” for most of the organic chemical processes of life itself. In
contrast, in the last 150 years the birth and growth of the organic chemicals processing
industry has been largely dependent on the use of organic solvents as vehicles
for carrying out chemical transformations. These solvents became available as a
result of the growth of the coal-fired energy industry in the 19th century. The organic
chemicals industry continued its inextricable link with the energy industry as
coal gave way to oil and gas in the 20th century. Although the organic chemicals
industry is only a minor offshoot of the energy industry, in terms of its production
volume, it is nevertheless linked with it as the world begins to deal with the substantial
environmental, economic, and social consequences associated with the use and
disposal of the world’s organic feedstock. The energy industry’s caused-emission of
greenhouse gases (particularly carbon dioxide) and particulates, as well as the not
infrequent coal mine or oil tanker disasters, have had substantial impact on public
health, wildlife, and the climate. The public recognizes that the energy industry is
promoting work to harness alternative sources of energy to the existing finite organic
feedstock. Yet, despite this, all recently promoted “clean” forms of energy (harnessed
from hydroelectric, solar, wind, waves and tides, geothermal, and nuclear sources)
provide only a small proportion of the world’s energy needs; and all of them, especially
nuclear sources, have their detractors. Paradoxically, the public themselves,
as consumers, are only now reluctantly recognizing that their “addiction” to energy
is driving the slow-moving energy industry to a socio-political tsunami, with,
likely, immense adverse consequences for the planet. Although no one can predict
the enormity of the consequences, few in power appear willing to take a proactive
stance to deal with trends. In short, no leaders have yet emerged to marshal the social
and political will needed to promote energy conservation and to “treat” energy
The organic chemicals industry, although operating at a small fraction of the scale
of the energy industry, nevertheless makes a substantial contribution to planetary
problems, seen in the creation of polar “ozone holes” and in the damage inflicted,
from time to time, by chemical spills and accidents involving noxious, flammable,
and explosive chemicals.
As for using water-based systems for the manufacture of organic chemicals, it is
useful to recall that the fermentation industry once provided part of the base of the
organic chemicals industry, with the fermentation of n-butanol, acetone, acetic acid,
lactic acid, and monosodium glutamate, inter alia.Chemical methods have superseded
fermentation for the first three of these compounds through interest in fermenting
n-butanol for fuel purposes has been rekindled. The recent growth of fermentation
for the production of ethanol from corn, to provide a replacement for t-butyl methyl
ether as a gasoline additive, has highlighted the merits of the water-based production
of organic chemicals. Although the production of ethanol by the fermentation of corn
has serious downsides,17 its development has helped to raise interest in the use of
enzymes for the manufacture of some organic chemicals. In addition, plain interest
in the safety, low cost, and non-flammability of water has spurred many to build on
revelations on the use of water, some published decades ago.18
Today, the drive to consider water as a solvent has been greatly stimulated by
growth of interest in so-called “green chemistry,” representing a desire to employ
sustainable renewable resources for the safe and environmentally advantageous production
of organic chemicals. Chemists doing so also realize that there is still a need
to deal with aqueous process wastes and to think in terms of conservation and recycling.
Water will not, of course, be universally applicable, and a substantial shift to
its optimal use will take time to achieve.
Dealing with water as a solvent for conventional organic chemical transformations
first, many have picked up on the earliest observations18 to create a solid and growing
body of literature validating the movement to water as a solvent for conventional
organic chemistry.19 The Kobayashi reviews19(d) describe many remarkable achievements
in the area of developing catalysts for use in aqueous systems (particularly for
enantioselective hydrogenation, carbon–carbon bond formation, hydroformylation,
17The use of ethanol as a fuel additive has several worrisome features. First, it is based on government
subsidies. Second, making a gallon of ethanol from corn in the United States is said to require 29%
more energy from fossil fuels than a gallon of ethanol can provide (see Patzek, T. W., and Pimentel D.
Natural Resources Res., 2005, 14, 65); biological production from cellulose would change this. Third,
crop diversity is being destroyed in the state of Iowa, which is now almost exclusively a corn-growing
18(a) Jo?o, F., and Beck, M. T. (Magyar K?emiai Foly?oirat, 1973, 79, 189) the first to sulfonate phosphine
hydrogenation catalysts for aqueous hydrogenation reactions. (b) Breslow, R., and Rideout, C. J. (Am.
Chem. Soc., 1980, 102, 7816) showed that cyclopentadiene and methyl vinyl ketone underwentDiels–Alder
cycloaddition 700 times faster in water than in isooctane, even though, on its own, cyclopentadiene has
no solubility in water. Chemists are now more aware that sparing solubility is often all that is needed for
reactions to occur. Indeed it is recognized that water can have a profound effect in biphasic systems by
promoting hydrophobic interactions.
19(a) Lubineau, A. Chem. Ind., 1996, 123. (b) Li, C. J., and Chan, T. H. Organic Reactions in Aqueous
Media, John Wiley & Sons, New York, 1997. (c) Organic Synthesis in Water, Grieco, P. A., Ed., Blackie
Academic and Professionals, London, 1998. (d) Kobayashi, S., Ed. Advanced Synthesis and Catalysis,
2002, 344, 219–451.
alcohol oxidation, inter alia). Reviews by Sinou20 and by Dwars and Oehme21 are of
particular interest to the pharmaceutical industry in their description of catalysts for
inducing the enantioselective hydrogenation of C=C, C=O, and C=N bonds. The
Kobayashi process for preparing certain esters from carboxylic acids and alcohols,
using tailor-made polymer-supported sulfonic acids as catalysts, in water, is conceptually
intrigueing.22 Interest has been extended by chemical engineers, exploring
organic reactions in supercritical water23 and in a creative combination of microwave
activation, of mostly intramolecular reactions, in supercritical water.24
The research chemist’s successful demonstration of a synthesis based on water
as the solvent does not necessarily mean that the reaction can be shaped into a
superior, safer process. For example, probably most homogenous enantioselective
reduction processes are actually carried out in water/organic solvent systems. The
process development chemist, in considering whether the water-based process should
be developed to a commercial scale, in a pharmaceutical industry setting, needs to
ask many questions (see Chapter 6). If the process is different from that previously
used for preparing the API batches used in the toxicology studies, have new or
greater levels of impurities been introduced? Is the physical form and stability of
the API (or even the intermediate) acceptable? What is needed to deal with solvent
handling if the enantioselective reduction uses a solvent, often, say, to make catalyst
recycling easier? (Catalyst turnover—the number of times the catalyst can be usefully
recycled—becomes a critical factor when both the water-soluble ligand and the
rare metal forming the catalyst are very expensive.) What process is needed to
eventually recover and purify spent catalyst components for recycle? What equipment
does the manufacturer need to carry out the process, recycle catalyst and solvents,
dispose of wastes, and so on? How does the cost of manufacture by the homogeneous
enantioselective reduction process compare with, say, an existing process or other,
more conventional ones?
It is obvious that considerable effort is needed to make a good enough case to
justify investment, even when, on the surface, the merits of a water-based process
appear so appealing.
As an aside, decades of work have gone into the study of enantioselective homogeneous
hydrogenation processes in both organic and aqueous systems. There is
increasing commercial interest in this field spurred by the spectacular, time-encrusted
development of a complex catalyst for the enantioselective hydrogenation of an imine
to a chiral amine needed for manufacture of the important herbicide, (S)-Metolaclor.
The technical success of this program (Scheme 1)25 owed much to the perserverance
20Sinou, D. Adv. Synth. Catal., 2002, 344, 221–237.
21Dwars, T., and Oehme, G. Adv. Synth. Catal., 2002, 344, 239–260.
22Kobayashi, S., and Manabe, K. Adv. Synth. Catal., 2002, 344, 270–273.
23Savage, P. E. Chem. Rev., 1999, 99, 601.
24Strauss, C. A. Australian J. Chem., 1999, 52, 83.
25(a) Bader, R., Flatt, P., and Radimerski, P. X. Y. U.S. Patent, 5430188, 1995 (to Ciba-Geigy). (b) Blaser,
H. U., Buser, H.-P., Coers, K., Hanreich, R., Jalett, H.-P., Jelsch, E., Pugin, B., Schneider, H. D., Spindler,
F., andWegman, A. Chimia, 1999, 53, 275. (c) Blaser, H.-U., Adv. Synth. Catal., 2002, 344, 17. (d) Blaser,
H2/Iridium complex–
Xylophos catalyst
80 bar/50°C
CH (S)
Xylophos = 1-[1-(di-3,5-dimethylphenylphosphino)ethyl]-2-(diphenylphosphino) ferrocene.
SCHEME 1. Enantioselective homogeneous hydrogenation in the manufacture of (S)-
and courage of adventurous people in different chemistry cultures who designed the
Efforts to use water as a solvent in conventional organic chemical transformations
will continue to grow with the increasing interest of chemists and engineers
in cleaner, more efficient, lower-cost and “greener” chemistry. However, the
well-established, very large parallel field of endeavor, utilizing enzymes in various
forms to produce needed chemicals in the bulk chemical, commodity chemical,
and fine chemicals fields, is also growing very rapidly—this includes the production
of both achiral and chiral molecules, the latter being of prime interest to
the fine chemicals, and particularly the pharmaceutical and agricultural chemicals,
Man has long been harnessing enzymes in “processing” organic chemicals, notably
in sewage treatment, garden composting, the brewing industry, starch hydrolysis and
isomerization, cheese manufacture, the baking industry, and even the manufacture
of detergents. Most of these are of ancient origin, long preceding the fermentation
of butanol, acetone, and so on, mentioned earlier. For instance, the Japanese have
been exploiting fermentation, quite apart from the fermentation of rice for sake, for
hundreds of years with soybean fermentation for soy sauce, having led them to create
enormous fermentation businesses, to produce amino acids and proteins,most of them
with a link to the food industry. Building on their ancient fermentation industry, over
the last few decades the Japanese have achieved considerable success in producing
bulk chemicals as well as fine chemicals. One recent notable achievement in the bulk
chemical field is Mitsubishi Rayon’s elegant process26 for producing concentrated
acrylamide solutions by the selective enzyme-mediated partial hydrolysis of acrylonitrile
(Scheme 2). This process promises to completely replace the mineral-acid-based
partial hydrolysis of acrylonitrile which suffers the disadvantage of requiring separation
of the mineral acid from the water-soluble acrylamide product. It is worth
pointing out, at this point, that enzyme-based processes often require a longer time to
H.-U., Pugin, B., Spindler, F., and Togni, A. Comptes Rendu Chim., 2002, Vol.5, 379. (e) Blaser, H.-U.,
and Schmidt, E. Asymmetric Catalysis on Industrial Scale: Challenges, Approaches and Solutions, John
Wiley & Sons, New York, 2004.
26Ishi, K., and Murao, K. U.S. Patent 6,043,061, March 28, 2000, and references cited therein.
Rhodococcus rhodochrous
48% aqueous solution
SCHEME 2. Enzyme-mediated partial hydrolysis of acrylonitrile to acrylamide.
develop to their maximum potential compared with chemical processes, and for this
reason it has usually taken many years for enzyme-mediated processes to supersede
chemical ones.27 This may well change, especially in the pharmaceutical industry, as
more chiral APIs are identified.
Growth of the chiral chemical industry will require greater emphasis on the education
of chemists, chemical engineers, and others in the fundamentals and applications
of biotransformation. Such an education, involving a sort of hybridization with microbiologists,
geneticists, biochemists, and biochemical engineers, will need, also,
to accommodate new manufacturing requirements, including new regulatory and
environmental considerations.
In the last two to three decades the growth in applications of enzymes in the
pharmaceutical industry (both in isolated form and in whole cell fermentations) has
been little short of phenomenal. This short essay can do no more than stimulate
classically trained chemists and chemical engineers to gain further education by
referring them to a few of the many books and reviews that have been published in the
biotransformation field.28 These publications provide, respectively, comprehensive
overviews on hydrolases and accounts of what has worked previously to provide a
guide for (a) applying enzymes to other synthesis problems, (b) mechanisms used by
enzymes to catalyze particular organic transformations, (c) source information for
successfully using enzymes, (d) enzyme mimics, and (e) combining catalytic process
steps without intermedicate recovery.
The search for new sources of drugs via both greater biochemical understanding
of disease and continued screening of nature’s encyclopedia of natural products will
create new APIs and new challenges for chemical process development practitioners.
The main challenges lie in finding practical ways to simulate nature’s “enzymic
sleights of hand” in manufacturing complex molecules. Consider, for example, four
of the most successful APIs of the last 20 years, namely,
27Examples of this, in addition to the above acrylamide example and the related enzymatic conversion of
3-cyanopyridine to nicotinamide, are the amidase cleavage of penicillins to produce 6-aminopenicillanic
acid (6-APA) and the corresponding (more complex) process for converting cephalosporin C to 7-
aminocephalosporanic acid (7-ACA) (see Chapter 9, footnote 8.)
28(a) Bornscheuer, U. T., and Kazlauskas, R. J. Hydrolases in Organic Synthesis, John Wiley & Sons,
New York, 2006. (b) Silverman, R. B. The Organic Chemistry of Enzyme-Catalysed Reactions, Academic
Press, New York, 2002. (c) Drauz, K., and Waldman, H., Ed. Enzyme Catalysis in Organic Synthesis
(3 Volumes), John Wiley & Sons, New York, 2002. (d) Breslow, R. Artificial Enzymes, John Wiley &
Sons, New York, 2005. (e) Bruggink, A., Schoevart, R., and Kieboom, T. Concepts of nature in organic
synthesis (Review). J. Org. Proc. Res. Dev., 2003, 7, 622.
Oseltamivir (Tamiflu) is derived from shikimic
acid, itself extracted from various plant sources (e.g.,
Quinic acid ex the bark of various Cinchona trees and
Shikimic acid ex the Star Anise plant).
10-Deacetylbaccatin III (isolated from leaves of
Taxus baccata L.)
Intermediate for the synthesis of Paclitaxel
(marketed as the antineoplastic drug, Taxol).
Lovastatin (fungal metabolite from Aspergillus
Marketed as the antihypercholesterolemic drug,
Mevacor. (The closely related statin, Zocor,
is produced from lovastatin by replacing the
2-methylbutanoic acid side chain with the
2, 2-dimethylbutanoic acid side chain).
Artemesinin (Isolated from Artemesia annua L. -
Wormwood bush found in China)
Now marketed, pure, as Artemesinin, but known
for almost 2000 years in China as a component of
the Chinese antimalarial, Qinghao.
In developing practical processes for producing structures of the above complexity,
process development chemists and engineers need to increasingly collaborate
with biotechnology partners, indeed often to play an important supporting role to the
biotechnologist. Support is essential in the form of analysis, establishing stable process
conditions, aiding in isolation and purification efforts, and so on. Manufacturing
complex natural products, such as the above, eventually requires that “intermediates”
are sourced which protect the wild source.29 In this spirit, biotechnology strategies
are emerging for the manufacture of intermediates for Tamiflu,30 Artemisinin,31 and
Vinblastine analogues.32 Discovering and understanding the processes that Nature
uses to create the unique chemical structures in complex molecules such as the above
is an ongoing endeavor that will undoubtedly enable access to completely novel
structures some of which will provide drugs of the future.
The once-obscure field of exploring potential uses of bacteria and their enzyme
systems that thrive in hot or boiling water (extremophiles)—from geysers and sea bed
vents in tectonically active areas—is another instance of nature generating practically
valuable resources. The use of hyperthermophilic enzymes to speed the fermentation
of ethanol from starch is a case in point.33
Adapting existing natural systems to undertake chemical transformation is now
well established, although this appears to be only the beginning. The field is going
beyond evolution, which cannot do more than work with what nature provides. Thus,
on the periphery of biological development, two avant-garde engineering professors
at the Massachussetts Institute of Technology, Drs. Drew Endy and Tom Knight, are
working on “Bio-bricks” to link strands of DNA with one set of desired functions in
one “brick” to another strand with other functions in another “brick.” These may be
linked to the DNA of a cell to control its activity in the production of new things. At
least that is the idea, but, in this arena, the propensity of living organisms to evolvemay
mete against survival—tinkering with the unknowns in modifying genetic material
needs to be handled with great care.
Although substrate insolubility inwater need not always be a deterrent to the use of
water as a solvent in organic reactions [see footnote 18(b)], viscerally, the probabilities
would seem to preclude thewide use ofwater for carrying out reactions with insoluble
substrates.However, in the spirit of ProfessorBreslow’s initiative [see footnote 18(b)],
water must be tried. Furthermore, the reality is that water is such an attractive solvent
from a commercial point of view that, under some circumstances, it would be worth
undertaking “artificial” steps to bend the required sequence of reactions to enable
29Thus, in the case of Taxol, it was eventually found that the destruction of yew trees themselves could be
avoided by harvesting yew tree needles as a source of intermediates.
30Shikimic acid (for Tamiflu) can be produced from genetically engineered bacteria as a less threatening
alternative to harvesting from Star Anise (work of Professor JohnW. Frost and co-workers,Michigan State
University, New York Times, Business Section, November 5, 2005.)
31The limited supply of wormwood is encouraging efforts to engineer genes, from the wormwood bush
and yeast, into E. coli with the objective of getting their genes to work together to produce precursors to
Artemisinin which can be readily converted, chemically, to Artemesinin itself. Jay Keasling, University
of California, Berkeley.
32McCoy, E., and O’Connor, S. E. J. Am. Chem. Soc., 2006, 128, 14276.
33Corfield, R. Chemistry World, September 2005, p. 51.
them to be carried out in water (“process hydration”). The concept is no more than an
extension of the already well-established field of protecting functional groups during
transformations that would otherwise change that functional group. Thus in order to
solubilize a given substrate in water, the insoluble substrate would have to be readily
substituted at a reactive site by a water-solubilizing group capable of being readily
removed later in the reaction sequence, without introducing undesirable impurities
to the final API. Water-solubilizing protecting groups can be imagined wherein the
water-solubilizing moiety carries, for example, a carboxylic, sulfonic, or phosphoric
group to introduce water solubility.
The circumstances wherein process hydration would be justified will depend on
both the ease of introduction and removal of the water-solubilizing group and the
cost of both manipulations. Process hydration would seem likely to be more readily
justified in a situation where a sequence of high-yielding process steps could be
arranged; the longer the sequence, the less the impact of the cost of process hydration.
Process hydration may be of particular value in enabling a great variety of chemical
transformation technologies to be considered, including those mediated by enzymes,
homogeneous catalysis, electricity, microwaves, and photochemical technologies.
As in all other endeavors, a movement toward using water as a solvent may be
promoted by an individual chemist or chemical engineer, but its success will depend
on harnessing all the other disciplines needed to create a practical process. In the
particular case of enzyme-mediated transformation, many chemists and engineers
have been stimulated by the results of biological science work to team up with
biochemists, microbiologists, geneticists, biochemical engineers, inter alia, to drive
chemical process development in new directions. In ending on a speculative note, it
seems more than likely that enzyme-mediated transformation will be shown to do
more than is generally thought and that the interdisciplinary dialogue will increasingly
showthat needed chemical transformations will find a biological system that can meet
the need. It behooves all chemists and chemical engineers, and especially those in the
pharmaceutical and agricultural industries, to educate themselves in the life science
disciplines, at least to the level of understanding needed to appreciate how they may
need to work with the microbiological disciplines, among others, to get things done.
Clearly, there is a healthy drive, already going on, to promote the wider use of
water as a solvent for organic chemical reactions. Thismight even extend to looking at
reactions in sea water for coastal manufacturing plants! The literature also suggests
[see footnote 28(d)] that the use of enzymes could well grow outside the box of
conventional biological systems.
For those of us at the end of our careers, it seems a pity to have to leave such an
exciting future.
Thoughts on Other Solvents
In keeping with the experimental nature of chemistry, many individuals have championed
the use of alternative solvents to the ubiquitous organic solvent option. As
indicated in “Water and Enzymes in Chemical Process Development,” water has
its limitations as a solvent for insoluble substrates, though, as stated, water always
needs to be tried. Even sea water may have a place in situations where sea salts are
tolerable—for example, in utilizing extremophile microorganisms or enzymes from
deep-sea tectonic plate vents.
From a practical chemical process development point of view, other solvents, or,
indeed, no solvent at all, are considered, because their use may be advantageous over
all other options.
Supercritical Fluids. These solvents, particularly supercritical carbon dioxide, have
been used for the safe extraction of food components (e.g., hop extracts for beer
production), in arranging selective extraction (selectivity can be manipulated by controlling
pressure and temperature), and in ensuring that extracts are essentially free of
solvent residues. They have also found use in supercritical chromatography and associated
analytical applications. Supercritical fluids, again particularly carbon dioxide,
are, presently, mostly used as extraction solvents rather than reaction solvents. The
largest use is in the extraction of caffeine from coffee beans, but the technique is
also used in the extraction of oils from soybeans, inter alia, and in the production of
small quantities of perfumes and flavors. These uses all depend on another advantageous
property of supercritical fluids, namely that they are generally superior to
conventional liquids in being able to penetrate themicropores of a solid structure; this
phenomenon has even been applied in coal extraction. Other formidable advantages
associated with using carbon dioxide are its low toxicity, nonflammability, ready
availability, relatively simple removal and recycle, the ease of isolating extracted
products, and its low cost. One surmountable disadvantage is the capital expenditure
requirement for pressure equipment and overcoming the usual reservations concerning
new technology.34 Wider adoption needs the ingenuity and leadership of chemical
engineers, with data from successful pilot plant work, to make the economic case for
the needed investment. Acceptance will also be enhanced by the “greening” of those
involved in decision making, especially if there are credible projections that other
applications of the technology may be achievable.
In the last 20 years, meaningful successes in projects outside classical extraction
applications, especially in using supercritical carbon dioxide as a reaction solvent
for hydrogenation, are increasing industrial interest in the technology. If there is
one intrinsic property of supercritical fluids which has changed the landscape of
hydrogenation technology, it is their ability to dissolve hydrogen, and most organic
substrates, in high concentration. This property, coupled with the use of noble metal
catalysts in many varieties, has enabled the creation of efficient high mass transfer
hydrogenation systems with significant advantages over the classical heterogeneous
organic solvent/noble metal catalyst systems. Thus hydrogen and the substrate move
directly to the catalyst surface and reduced products move off easily, avoiding the
hydrogen transport rate issues associated with classical hydrogenations. Tailoring
34Supercritical carbon dioxide was considered by Schering–Plough for the extraction of diosgenin from
Barbasco roots inMexico, but, despite being an efficient extractor and less energy intensive versus continuing
to use solvent extraction, the capital cost of the high-pressure equipment and operating considerations
meted against its adoption.
reaction parameters such as hydrogen concentration, pressure, temperature, and flow
rates (in a continuous system) is more facile than in classical hydrogenation systems,
enabling mild or extreme hydrogenation conditions to be engineered.35 Enormous
increases in reaction throughput have been engineered by developing continuous-
flow reaction equipment. The recovery of product is also simplified. Hydrogenations
in supercritical carbon dioxide are inherently somewhat safer as well as being more
environmentally friendly.
Although supercritical carbon dioxide is a poor solvent for many highly polar
substances, and the reactivity of carbon dioxide will always limit its use as a reaction
solvent, university-based champions of the use of supercritical carbon dioxide, working
with adventurous industrialists, initially in the area of a range of hydrogenation
reactions, will help to pave the way for other applications.
Poliakoff et al., University of Nottingham, England, have built on their earlywork36
demonstrating the hydrogenation of aromatic rings, olefins, aldehydes, and ketones
in supercritical carbon dioxide, and of nitro and imino compounds in supercritical
propane, using a laboratory scale stainless steel continuous reactor, to create aworking
relationship with Thomas Swan and Co. Ltd., Consett, Co. Durham, England.37
Hydrogenations in supercritical carbon dioxide, have, inevitably, been explored as
routes into chiral molecules. For example, given identification of an effective chiral
catalyst that can be easily recycled, the use of supercritical carbon dioxide as a solvent
in the enantioselective reduction of the prochiral imine leading to the intermediate
for the commercially useful herbicide, Metolachlor (see Scheme 1), would seem to
be an attractive target.38
In the United States, several academic groups are collaborating with industry,
especially in the area of hydrogenations (both hydrogenolysis and hydrogen addition
to double bonds) in supercritical carbon dioxide. DuPont is perhaps the leading
company in the drive to adopt “green chemistry” in its manufacturing base, with particular
interest in working with several universities in the fields of hydrogenation39
and polymerization.40 In the latter case, DuPont has already built a supercritical carbon
dioxide pilot plant unit capable of manufacturing over two million pounds/year
35Mild conditions will reduce only the double bond and not the keto group of isophorone. Extreme
conditions will reduce nitrobenzene to cyclohexane and ammonia. However, it should be noted that carbon
dioxide itself can be reduced to carbon monoxide and water under extreme conditions; see footnote 36
and Chem. Eng. News, 2001, May 28, 32.
36Hitzler, M. G., Smail, F. R., Ross, S. K., and Poliakoff, M. J. Org. Process Res. Dev., 1998, 2, 137.
37Thomas Swan and Co. Ltd. built a 1000-tonne/annum multipurpose supercritical carbon dioxide demonstration
plant at its Consett site to explore hydrogenation and Friedel–Crafts type acylations and alkylations,
inter alia.
38The enantioselective hydrogenation of imines, such as the Metolachlor intermediate, in supercritical
carbon dioxide has been investigated by Walter Leitner et al., Max Planck Institute for Coal Research,
M?ulheim, Germany; but as always, replacing an established technology is a difficult task.
39Chemical engineering professors B. Subramaniam, University of Kansas, and J. Brennecke, University
of Notre Dame, along with their groups, are two collaborators.
40Chemistry/chemical engineering professor, J. DeSimone, University of North Carolina, has long been a
protagonist of “green” polymerizations in supercritical carbon dioxide. A major achievement has been his
patented findings on the use of alkylsilyl and perfluoroalkyl polymer fragments as “surfactants” in creating
high-molecular-weight polymers previously unachievable in this solvent.
of Teflon co-polymers using carbon dioxide as the solvent. Other areas of interest
include oxidations,41 hydroformylations, and Friedel–Crafts reactions—that is,
reactions that utilize process conditions on the acid side.
As with all new technological applications, much pioneering experimental work
needs to be done. A start could be made if chemical process development groups
would all acquire a piece of equipment, such as that described by Poliakoff (see
footnote 36) to undertake the evaluation of the use of supercritical carbon dioxide as
an integral component of chemical process development programmes. Alternatively,
and probably advantageously, liaisons with university groups already working in the
area would be beneficial.
Liquid Sulfur Dioxide. Although sulfur dioxide has a pungent choking odor, it has
been of interest as a solvent, especially in academia, for over 100 years. Its boiling
point (?10?C) allows that it can be handled safely under its own vapor in a sealed
glass vessel without special precautions.42 Its reactivity with water requires that
liquid sulfur dioxide be generally used as a solvent under anhydrous conditions.
Sulfur dioxide also forms solvates, charge transfer complexes, or adducts with many
compounds—for example, tertiary amines, ethers, alcohols, and primary amines. In
part, this behavior undoubtedly contributes to some of the unique character of liquid
sulfur dioxide as a solvent. Liquid sulfur dioxide has also proved to be a useful
medium for carrying out electrochemical reactions.
Despite its reactivity toward nucleophilic substances, liquid sulfur dioxide has
found minor use as a solvent in many reactions. This is largely because it has strong
ionizing powers and is a very good solvent for most organic compounds, including
amines, acids, alcohols, esters, and aromatic hydrocarbons. The ease of stripping
and recycling sulfur dioxide at the end of a reaction is also an ecologically positive
attribute. Indeed, another major attribute is the availability of liquid sulfur dioxide
in tank car quantities for, very approximately, 35 cents/liter (2007). In this regard,
it seems rather odd that organic chemists, inured as they are to the use of organic
solvents for organic chemical reactions, have essentially ignored sulfur dioxide as a
potential solvent. In today’sworld,with process emissions a serious process chemistry
concern, it is surprising that more chemists and engineers have not picked up on the
notion that to properly use sulfur dioxide, it is essential to contain it safely. Building
a relatively simple, safe, general-purpose laboratory containment unit, suitable for
handling those kinds of reaction where sulfur dioxide might be used, would bring
sulfur dioxide into the main stream as a solvent for organic chemical reactions. A
unit with the capability of handling reactions up to, say, one to five atmospheres
would increase the evaluable range of temperature/pressure conditions. Extending
process conditions in this way may reveal other process benefits worthy of further
exploration and optimization. The general-purpose containment unit referred to above
41The oxidation of hydrogen to hydrogen peroxide in supercritical carbon dioxide is a case in point
Hancu, D., and Beckman, E. J. Green Chemistry, 2001, 3, 80.
42Waddington, T. C., in Non-aqueous Solvent Systems,Waddington, T.C., Ed., Academic Press, New York,
1965, p. 253.
would also be valuable in enabling the evaluation of other noxious reactions and more
“adventurous” process conditions.43
A couple of published reaction types serve to illustrate uses and some benefits in
employing sulfur dioxide as the solvent.
(a) Esterification of cephalosporins under conditions44 that avoid 3-bond
Base/Alkyl Cl
Liquid SO2
Alkyl can be benzhydryl (89% yield), methoxymethyl (98% yield), inter alia.
(b) Solvent effects in the bromination of alkylbenzenes.45
Alkylbenzene T = ?9?C (Reflux) T = 25?C 1,2-C6 H4Cl2
Ortho Meta Para Ortho Meta Para Ortho Meta Para
Toluene 11.4 0 88.6 17.6 0 82.4 22.8 0 77.2
Ethylbenzene 8.9 0 91.1 13.0 0 87.0 15.3 0 84.7
i-Propylbenzene 2.6 0 97.4 8.1 0 91.9 10.3 0 89.7
t-Butylbenzene 0 0 100 0 0 100 1.6 0 98.4
It seems that the noxious properties of sulfur dioxide and its limitations as a
general-purpose solvent have been largely responsible for the continued lack of
interest in its use in the organic chemical manufacturing industry. To change the
perception requires that entrepreneurial people put together the laboratory and pilot
plant equipment, undertake experiments, and publish their results.
Ionic Liquids. In the last 20 years, much attention has been given to the use of certain
ionic liquid salts, compounds comprising an organic cation, and usually, an inorganic
anion as “solvents.”46 The most common classes of ionic liquids have the following
general structures:
43The unit might also incorporate the capability to accommodate reactions in supercritical carbon dioxide.
44Seki, S., Nakabayashi, S., Nishata, K., Ito, N., and Fukatsu, S. Tetrahedron Lett., 1977, 2915.
45Canselier, J.-P. Bull. Soc. Chim. France, 1972, Number 2, 762. See also Canselier, J.-P. Bull. Soc. Chim.
France, 1971, Number 5, 1785.
46Wasserschied, P., and Welton, T., Eds. Ionic Liquids in Synthesis, Wiley-VCH, Weiheim, 2002.
R-N N-R' X R-N X
R, R?, R??, and R??? are usually lower alkyl
X is a variety of cations, including AlCl4 , BF4 , PF6 , R PF 3 3 , etc.
Structures can be tailored to provide a range of liquidity from?100?C to+200?C.
Interest in ionic liquids, early on mostly in academia, centered around their remarkable
properties as solvents and catalysts for supporting and enhancing reactions,
many of which are often difficult to carry out using conventional process conditions.
The perceived advantages of ionic liquids have been widely publicized, and
enthusiasm has encouraged scientists to look upon these advantages sometimes with
disproportionate favor. The main advantages that attract attention are as follows:
 They are nonvolatile (no air pollution issues).
 They are nonflammable (many have high thermal stability).
 They provide a highly ionizing medium, which has enabled scientists to enhance
reaction rates and increase reaction selectivity. The ionizing properties of ionic
liquids have also proved to be valuable in electrochemical transformations and
in battery uses.
 They often offer processing advantages, in terms of easier separation of products
or catalysts, greatly facilitating recycle and reuse. However, as often happens
in separation operations, any solubility of the ionic liquid in another phase can
compromise the efficiency of recycling these expensive liquids.
A few examples of the broad applicability of ionic liquids are as follows:
 As vehicles for the electropolymerization of benzene to form the conducting
polymer poly (p-phenylene).47 Ionic liquids are said to be good alternatives to
liquid sulfur dioxide.
 As a medium to facilitate the acid-catalyzed transfer of the acetyl group from
acetylmesitylene to anisole.48
 As a solvent for the asymmetric epoxidation of 2,2-dimethylchromene mediated
by Jacobsen’s chiral (salen)-manganese catalyst.49
 In tandem with the use of supercritical carbon dioxide, in ester synthesis employing
Candida Antarctica lipase B, adsorbed on silica gel, as the esterification
catalyst under minimum water conditions.50
47Endres, F., Zein El Abadin, S., and Borissenko, N., Electrochemistry Commun., 2004, 6, 422.
48Laali, K. K., and Sarca, V. D. Green Chem., 2004, 6, 245.
49Song, C. E., and Roh, E. J. Chem. Commun., 2000, 837.
50Lozano, P., de Diego, T., Gmouh, S., Vaultier, M., and Iborra, J. L. Biotechnol. Prog., 2004, 20, 669.
Commercially, the most elegant use of ionic liquids, perhaps discovered serendipitously,
51 is the BASF application for acid scavenging in their manufacture of the
photoinitiator intermediate diethoxyphenylphosphine.52
2 EtOH + Cl2 PPh (EtO)2P Ph + 2HCl (scavenged) 80 °C
Acid scavenger
BASF’s original acid scavenger, triethylamine, created viscosity and work-up problems
because of the need to maintain anhydrous conditions. BASF’s use of 1-
methylimidazole instead of triethylamine, at their reaction temperature of 80?C,
led to the formation of two liquid phases, an upper diethoxyphenylphosphine phase,
and a lower methylimidazolium chloride phase (fortuitously, this ionic liquid melts
at 75?C!). Moreover, methylimidazolium chloride proved to be a nucleophilic catalyst.
The processing revolution generated by these discoveries enabled creation of a
high-productivity continuous process for diethoxyphenylphosphine manufacture and
created a whole new business in acid scavenging technology.
An analogous technology commercialized by IFP, Paris, is their continuous,
chloroaluminate ionic liquid dimerization of n-butene to isooctane, promoted by
a Ziegler–Natta-type homogeneous catalyst. The poorly miscible isooctane product
is readily separated.
Ionic liquids are also being evaluated as alternatives to the use of hydrofluoric acid
and sulfuric acid—for example, in simplifying aromatic alkylations by directly using
olefins instead of alkyl halides to carry out the alkylation.
Despite the impressive achievements of those working in the ionic liquids field,
chemical process development scientists and engineers, being aware of their disadvantages,
generally take a more cautious view of their potential value. The main
disadvantages are as follows:
 Their high cost, which both limits their appeal as solvents and highlights the
need for safe, efficient recycle.
 The instability of some anionic components of ionic liquids (e.g., AlCl4
?, PF6
types) to water limits the applicability of these types in processes involving
 Concerns re biotoxicity and biodegradability.
Many quaternary ammonium compounds are known to possess bactericidal properties,
a factor taken into account in the marketing of quaternary ammonium fabric
softeners. However, in the spirit of the European REACH initiative (See Chapter 5),
many of the ionic liquids in common use are being scrutinized for toxicity potential.
For instance, in the aquatic world there are indications that exposure to as little
as 10 mg/liter causes skin and gill hyperplasia in zebrafish, leading to respiratory
51But, to repeat Pasteur, “Chance only visits the prepared mind.”
52Vagt, U., and Emanuel, C., Chem. Processing, June 2006, 45.
problems, behavioral effects, and some deaths.53 Long-chain alkylammonium salts
appear to be the most toxic. Despite these concerns, BASF continues to progress
its ionic liquid businesses with the understanding that toxicity issues need to be
addressed and accommodated. To this end, BASF, in collaboration with Degussa,
has published toxicology data on three of the ionic liquids marketed by them (see
footnote 52).
There is little doubt that intrepid entrepreneurs will continue to champion ideas for
the application of ionic liquids. Without a doubt, further uses of these materials will
be found and justified for commercial use. However, the perceived disadvantages,
and particularly the toxicity issues, will temper enthusiasm until the data obtained
provide reassurances that an ionic-liquid-based process satisfies all the economic,
toxicological, safety, and (in the pharmaceutical industry) FDA regulatory criteria to
justify implementation.
When process development chemists consider supercritical fluids and nontraditional
solvents in their process research, the unconventional aspects enhance the
importance of integrating the chemical engineering discipline and all the regulatory
disciplines into their evaluation. In the case of supercritical fluids, chemical engineering
becomes a core discipline with respect to creating the physical equipment and
guiding the work program of unit operations in systems that are outside the chemist’s
normal glassware world, including evaluating prospects for a continuous process.
In evaluating ionic liquids for practical operations, the chemical engineering discipline
is not only a necessary partner in the more conventional aspects of equipment
utilization, process engineering, and so on. In new fields the engineering purview
may encourage consideration of unconventional technologies in ways that make the
difference between success and failure. For instance, since the field of using ionic
liquids in the pharmaceutical industry is relatively new, the chemist may be stymied
if toxicity concerns create potential FDA issues (e.g., trace contamination) or environmental
issues (e.g., eliminating toxic materials from process wastes). Such issues
could derail a potentially promising technology. The chemical engineer, open to the
use of solventless manufacture, might offer the possibility of using analogous alternatives,
including, say, the reevaluation of even older technologies such as the use of
urea/choline chloride melts, or even the use of liquid sulfur dioxide.
Polymer-Supported Synthesis and Reagents
Process development chemists, seeking simplicity in creating manufacturing
processes, have long been intrigued by the seeming elegance of polymer-supported
synthesis, and also the benefits to be gained by harnessing polymer-supported
reagents. On paper, undertaking the multistep synthesis of a molecule on a polymer
support, despite its lack of synthesis convergence, looks as though it could revolutionize
chemical manufacturing. Thus, combining a sequence of reactions on a single
polymer, using only one reaction vessel—or, better, one reaction column—would
53Chiappe, C. et al., News@Nature.com (doi: 10.1038/news 051031-8) and Chemistry World, 2005,
December, p. 19.
contain and massively reduce the unit operations associated with using conventional
multiple batch reactors and associated work-up equipment.
Such a concept offers practical advantages over and above the ready separation
of reactants from the substrate being manipulated. First, process containment embodies
intrinsically safer operation, decreasing the exposure of workers to chemicals
and minimizing the leakage of volatile chemicals into the atmosphere. Second, the
need for investment in conventional capital equipment (reactors, filters, driers, etc.)
would be substantially reduced, leading to fewer building requirements, including
warehouses for equipment and chemicals. This, in turn, leads to lower requirements
for plant services, including maintenance, chemicals handling, and energy, as well
as providing opportunities to simplify process automation. The astute reader will of
course appreciate that many of these advantages can be, and frequently are, realized
by working to combine very efficient reaction sequences in one conventional
reactor—it is the work-up between reaction steps which still represents the major
equipment burden associated with the use of conventional batch processes. Thus the
key factor, for best realizing the benefits of polymer-supported synthesis, is to achieve
high efficiency in each step of the series of reactions on the polymer in order to avoid
perhaps insurmountable purification problems at the end of the synthesis.
The related polymer-supported reagents field affords useful if lesser advantages
than polymer-supported synthesis in that the polymer support allows the ready removal
of a reagent promoting only a single reaction step. Polymer-supported reagents,
and especially catalysts, have nevertheless become very important in numerous chemical
manufacturing situations.
Without a doubt, many of the above advantages were recognized in awarding the
Nobel prize to Rockefeller University’s Professor R. Bruce Merrifield for his seminal
work on polymer-supported peptide synthesis.54
Polymer-Supported Synthesis. The passage of time, the publication of thousands
of papers, several books,55 and many commercial successes, especially in smallscale
applications of Merrifield-type polymers in the peptide field, have led to the
realization that harnessing polymer supports for large-scale synthesis work has many
hurdles to overcome to become more widely accepted for use in practical large-scale
From a chemical process development point of view the main issues/limitation are
as follows:
 Determining the best resin to use to maximize the loading of reactive groups,
to allow efficient complete reactions and to permit some recycle of the spent
54(a) Merrifield, R. B. J. Am. Chem. Soc., 1963, 85, 2149. (b) Merrifield, R. B. J. Am. Chem. Soc., 1964,
86, 304.
55For example, Dorwald, F. Z. Organic Synthesis on Solid Phase: Supports, Linkers, Reactions, John
Wiley & Sons, New York, 2000, and publications cited therein.
 Accommodating the likelihood of lower productivity when concentrations of
the substrate on the polymer are low.
 Ensuring that the slow diffusion of reactants to reactive sites in the polymer
matrix reaches as many buried sites as possible.
 Overcoming the problems generated by incomplete reactions, especially the
impurity burden created when both the desired product and byproducts are
unhitched from the polymer.
 Working out and optimizing the synthesis sequence to accommodate the known
vagaries of working on a polymer, especially minimizing the need to change
solvents (this can be a time-consuming process needing large volumes of solvent).
 Finding ways of reducing solvent usage if reaction solvents have to be changed.
 Creating analytical monitoring methodologies to follow reactions on the polymer.
 Persuading management, in a commercial, project-oriented, time-managed setting,
to allowthe thinking andR&Dtime needed for evaluation and development.
 Judging the probabilities of commercial success.
Advancing a field with such formidable hurdles (and yet such practical promise)
requires people of vision, enthusiasm, and experience to champion the technology.
Peptide manufacturers have shown that polymer-supported synthesis can be made
practical, albeit at a price necessitated by the small scale of operation. High costs and
the limitations outlined above have no doubt inhibited the field but in my view, the
development of polymer-supported synthesis has been more inhibited by the absence
of inputs from chemical engineers. Practical people recognize that the chemistry and
engineering disciplines need to work together in order to comprehensively evaluate
the perceived potential of technologies such as this. I can illustrate the point through
my own all-too-brief flirtation with polymer-supported synthesis when working in
Glaxo 40 years ago!
My colleagues and I backed into the field as a result of our commercially successful
development of the use of the diphenylmethyl (DPM) group as a temporary
protecting group in the synthesis of cephalexin (see Chapter 9). Initially,we employed
a separately prepared solution of diphenyldiazomethane (DDM) to prepare diphenylmethyl
esters until we demonstrated an esterification process using DDM prepared
in situ. Polymer-supported DDM seemed to us to offer a much safer prospect, especially
since we reasoned that the benzene rings of the polystyrene backbone could
become the benzophenone starting material. We quickly demonstrated the following
high-yielding reaction sequence:56
56(a) Chapman, P. H., and Walker, D. J. Chem. Soc., Chem. Commun., 1975, 690. (b) Walker, D., and
Chapman, P. H. U.S. Patent 4,038, 469, 1977.
2% DVB beads*
(200–400 mesh)
Cl2C = CCl2
C Ph P
2H4 CH3CO3H/I2
The polymer-DDM was used by us as it was prepared, but it appeared reasonably
stable.57 Polymer-DDM reacted rapidly with penicillin G sulfoxide acid, with the
nitrogen evolution perhaps helping to “stir” the reactionmixture(?).58 The subsequent
ring expansion of the penicillin sulfoxide moiety to a 3-methylcephalosporin and the
cleavage of the phenylacetyl side chain to give polymer-supported 7-ADCA (Scheme
1) gave surprisingly good yields59 (Scheme 3).
The ring expansion and cleavage reactions revealed two of the major hurdles
that face scientists trying to use polymer supports. First, in adapting the solution
chemistry used for ring expansion of the monomer to the polymer-supported penicillin
G sulfoxide, we found out that the switch from methylene chloride, used for the
immobilization of the penicillin, to dioxane, needed for the ring expansion, exposed
some of the limitations of working on the polymer. Thus, when the cephalosporin
resulting from the ring expansion was removed from the polymer, the yield was
somewhat lower and the productwas of poorer quality than the corresponding product
obtained using conventional solution chemistry. Second, the PCl5-mediated amide
cleavage reaction, in this case between two insoluble compounds, would not, a
priori, be expected to occur. However, the partial solubilization of PCl5, by the
pyridine used as the base, was sufficient for conversion of the side-chain amide in
the polymer-supported cephalosporin to an iminochloride. The iminochloride side
chain was then easily converted, via the iminoether, to the desired polymer-supported
aminocephalosporin. Third, the need to thoroughly remove all the debris from each
reaction necessitated extensive washing before the next reaction could be undertaken.
Large volumes of solvent were required to do this effectively.
There was not enough time during my career with Glaxo (1966–1975) to address
the above hurdles. We had plenty of ideas but little time to explore them. Reasoning,
as protagonists of working on glass bead surfaces had done, that diffusion problems
?It should be noted that the evaluation of beads from different sources showed those from Dow Chemicals
to be the best.
57Our expectation was that polymer-DDM would have the same stability as the monomer (see Chapter 9,
footnote 20). Polymer-DDM is now available commercially in small quantities from Bachem, CH-4416
Bubendorf, Switzerland. In addition, Bachem scientists have published data on the stability and uses of
their polymer: (a) Mergler,M., and Nyfeler, R. In Innovation and Perspectives in Solid Phase Synthesis and
Combinatorial Libraries, 1998. Collected Papers, 5th International Symposium, London, England, Epton,
R., Ed., Mayflower Scientific Ltd., Birmingham, 1999, pp. 351–354. (b) Mergler, M., Dick, F., Gosteli, J.,
and Nyfeler, R. Tetrahedron Lett., 1999, 40 4863–4864. (c) Gramberg, C. W., Dick, F., and Vorherr, T. In
Peptides 2000, Proceedings of the Twenty-sixth European Peptide Symposium, September 2000, Martinez,
J., and Fehrentz, J.-A., Eds., EDK Publishers, (Probably Santarcangelo di Romagna, Italy, according to
the Internet! EDK is a partner of the Karnak group and linked with Pragma 2000 an Automated Publishing
Solutions service), 2001, pp. 289–290. Publications (b) and (c) cover reactions with Fmoc-aminoalcohols.
58Walker, D., and Chapman, P. H. U.S. Patent 4,067,858, 1978 (to Glaxo).
59However, the yield across these two steps was somewhat lower than that obtained using solution
chemistry. It takes time and patience to adopt and reoptimize efficient solution reactions for use in polymer
Penicillin G
Ring expansion
80% yield from
Pen G 1(S)-oxide
SCHEME 3. Polymer-supported synthesis of Cephalexin59
would be minimized by working as much as possible on the surface of such as a
200–400 mesh bead, I proposed that it may be possible to double the loading of
the diazomethylene groups on the surface of the bead. Thus, it should be possible
to undertake a mild Friedel–Crafts acylation, to minimize deep penetration of the
beads, by acylating with a bulky reagent such as p-benzoylbenzoyl chloride. Such an
approach may also allow increased loading of a desired substrate on the polymer:
200–400 mesh
C Ph C Ph P
Despite the prospect of some productivity gains, the idea was never tried. We also
never had a chance to determine whether a single phenylene ring separating immobilized
substrate molecules would have other consequences in terms of neighboring
molecule interactions, especially if applied to peptide synthesis where tangling of
elongated chains may be problematic.
The problem of switching solvents and minimizing solvent usage between reaction
steps, especially if almost complete removal of the original reaction solvent is necessary,
poses one of the biggest challenges impeding the adoption of polymer-supported
synthesis in commercial synthesis operations on a larger scale than polypeptide manufacture.
From my vantage point the problem needed more involvement of chemical
engineers. One seemingly promising avenue of exploration was demonstrated in a
chromatography column by my colleague, Dr. Eric Martlew. Dr. Martlew fitted the
column with a vertical plunger and showed that the solvent swelling the resin beads
could be squeezed out to a large extent by lowering the plunger under increasing pressure.
Dr. Martlew demonstrated that reduced solvent useage was achievable. In the
hands of chemical engineers, working at higher pressures, perhaps with a pulsating
plunger carrying perforations with one way valves, could provide a dynamic flushing
effect in the beads. There was no time to determine whether repeated mechanical
compression and relaxation of solvent swollen beads would have any adverse effect
on the physical stability of the beads.
We made another unpublished (unfinished) modification of our polymer diphenyldiazomethane
to try to expand the scope of reaction possibilities to include aqueous
systems.We could greatly improve the hydrophilicity of our beads by sulfonating the
benzoylated polystyrene before carrying out hydrazone formation and subsequent oxidation
of the hydrazone groups to diazo groups. The resulting polymer, as a sodium
salt, did exhibit hydrophilic behavior, but there was insufficient time to investigate
any of the uses we envisaged for it (see later).
It appears that the field of polymer-supported synthesis would be worth further
exploration with a view to exploiting possibilities for applications beyond the peptide
synthesis field. Its amalgamation with chemical engineering to evaluate combinations
of other technologies, such as using microwave energy to promote more complete
reactions,60 say in concert with the use of pulsating pressure plungers, would seem
to be well worthy of further exploration.
One of the greatest advances that could be made to promote the use of compounds
such as polymer DDM would be to find ways of reducing its cost and also to recycle
the spent beads by oxidation of the polymer benzhydrol waste back to the polymer
benzophenone and repeating the reaction sequence needed to produce polymer DDM.
Regarding manufacture, the development and scale-up of existing processes to an
industrial scale would obviously reduce polymer-DDM costs. The opportunity for
recycle has also been demonstrated (see Example 24 in footnote 58) using 30%
aqueous nitric acid as the oxidant. However, the polymer DDM produced possessed
only?78% of the original activity.Other cost reduction opportunitiesmay be feasible.
Thus for those reactions not needing the very reactive diazomethylene group (e.g., in
the preparation of polymer benzhydryl esters), it may be possible to use the polymer
benzhydrol resin to directly produce the corresponding ester using chemistry already
established for producing benzhydryl esters from benzhydrol.61
60Yu, H.-M., Chen, S.-T., and Wang, K.-T. J. Org. Chem., 1992, 57, 4781 (see Section 4.6).
61Yoshioka, M. Pure Appl. Chem. 1987, 59, 1041.
The application of polymer-supported synthesis methodology in the peptide field
continues to grow. Refinements in the area of reducing the impurity burden in both
peptide synthesis62 and in applying polymer-supported synthesis techniques in the
field of combinatorial chemistry63 will undoubtedly be useful in influencing the
further exploration of the use of polymer supports in the broader field of chemical
Polymer-Supported Reagents. This field has grown enormously in the last few
decades as chemists and engineers have harnessed the obvious merits associated
with carrying out a reaction using a readily removable reagent. The concept is essentially
an elaboration of the immobilized catalyst field that has been widely applied in
industry, especially the petrochemicals industry, for over 50 years.
In recent times the polymer-supported reagents field has grown enormously, re-
flecting its application to the generation of novel “chemical libraries” for the pharmaceutical
and agrochemical industries. An excellent review of this field by Ley
et al.64 provides a comprehensive overview and also “an extensive listing of known
supported reagents, catalysts and scavenging agents . . . as an aid in the future design
of synthesis programmes.”
In regard to chemical process development, the polymer supported reagents field
offers legion opportunities for the use of almost any reagent imaginable, limited only
by what can be immobilized and still be effective. This definition allows that even
hazardous and noxious reagents, made safer by being immobilized on a polymer,
can be considered. The scope may, again, be limited by cost, difficulties in the
immobilization process, or the need to recycle the polymer-supported reagent or its
Use of a readily separable, polymer-supported reagent to facilitate a chemical
transformation is not always a panacea. Yields should preferably be high, and the
recovery and purification of the desired product from the reaction mixture should be
practical. Polymer-supported scavenging agents to scavenge byproducts have been
employed, as also have polymer-supported reagents that capture the desired product
from a reaction mixture.64
Capturing desired products from aqueous reaction mixtures (in a form useful
for further synthesis steps) using monomeric agents is well known. We successfully
applied such a technology to the extractive esterification of cephalosporin
derivatives from a filtered cephalosporin fermentation broth65 and a solution of a 3-
62For instance, it has long been known that missed sequences can be avoided by acetylating amino terminals
of incompletely reacted sites of a growing polymer peptide – (see Bayer, E., Eckstein, H., Hagele, K.,
Konig, W. A., Bruning, W., Hagenmaier, H., and Parr, W. J. Am. Chem. Soc., 1970, 92, 1735) and using
fragment synthesis methods for large peptides.
63Dolle, R. E., Le Bourdonnec, B., Morales, G. A., Moriarty, K. J., and Salvino, J. M. J. Comb. Chem.,
2006, 8, 597. (b) Parlow, J. J., Devraj, R. M., and South, M. S. Curr. Opin. Chem. Biol. 1999, 3, 320.
64Ley, S. V., Baxendale, I. R., Bream, R. N., Jackson, P. S., Leach, A. G., Longbottom, D. A., Desi, M.,
Scott, J. S., Storer, R. I., and Taylor, S. J. J. Chem. Soc., Perkin Trans. 1, 2000, 3815–4195.
65(a) Bywood, R., Robinson, C., Stables, H. C., Walker, D., and Wilson, E. M. In Recent Advances in
the Chemistry of ?-Lactam Antibiotics, Elks, J., Ed., Special Publication No. 28, The Royal Society of
Chemistry, London, 1977, p. 139. (b) Robinson, C., andWalker, D. U.S. Patent 4,059,573, 1977 (to Glaxo).
exomethylene-7R-glutaroylaminocepham-4-carboxylic acid 1(S)-oxide obtained by
the electrochemical reduction of the corresponding 3-acetoxymethylcephalosporin.66
Although it was never tried, it seems likely that polymer-DDM beads, swollen in
methylene chloride, would undertake the same extraction reactions.
We envisaged that the aforementioned sulfonated polymer-DDM may prove useful
for the extractive esterification of those water-soluble acidic materials from aqueous
solution that would not dissolve in methylene chloride for reaction with swollen
polymer-DDM beads. Such compounds could include enzymes with some acidic
character, but the risk is that desired enzyme activity may be lost by reaction with the
DDM group.
From a chemical process development point of view, there seems little doubt that
pursuit of the fields of polymer-supported synthesis and reagents, especially with
chemical engineers, will result in new applications of these fields in the future.
Microwave-Assisted Chemistry
The use of microwave ovens for cooking and for the rapid heating of foods and
beverages has been one of the most important domestic success stories of the last
30 years. The successful application of microwave energy to promote organic synthesis
was described by Gedye67 and Giguere68 and their co-workers in 1986, using
closed Teflon vessels to contain the reactants. The safety issues raised by the use of
solvents, along with the occasional reports of explosions in early laboratory work
carried out in domestic microwave ovens, provided cautionary restraint but did not
dampen the enthusiasm raised by findings that extraordinary reductions in reaction
times, often with cleaner reactions, could be achieved using microwave irradiation to
promote desired chemical transformations.69
Several small firms market more sophisticated equipment than the domestic microwave
oven, and research departments in a number of pharmaceutical companies
continue to work with the technology to aid their synthesis of APIs and their intermediates.
Several microwave equipment companies70 are developing larger-scale
equipment to produce kilogram quantities. Their systems are evolving to meet some
of the needs identified by the earlier practioners such as:
66Bernasconi, E., Lee, J., Sogli, L., and Walker, D. J. Org. Process Res. Dev., 2002, 6, 169.
67(a) Gedye, R. N., Smith, F. E., Westaway, K. G., Ali, H., Balderisa, L., Laberge, L., and Roussell, J.
Tetrahedron Lett., 1986, 27, 279. (b) Gedye, R. N., Smith, F. E., andWestaway, K. G. Can. J. Chem., 1988,
66, 17.
68Giguere, R. J., Bray, T. L., Duncan, S. N., and Majetich, G. Tetrahedron Lett., 1986, 27, 4945.
69For reviews see (a) Mingos, D. M. P., and Baghurst, D. R. Chem. Soc. Rev., 1991, 20, 1. (b) Caddick,
S. Tetrahedron, 1995, 51, 10403. (c) Deshayes, S., Liagre, M., Loupy, A., Luche, J.-L., and Petit, A.
Tetrahedron, 1999, 55, 10851.
70Equipment vendors include Prolabo Co., France; Personal Chemistry Ltd., Cambridge, England; Milestone
Inc., Monroe, Connecticut, USA and CEM Microwave Technology, Matthews, North Carolina,
TABLE 1. Solvents that Are Transparent or Absorb Microwave Energy and their
Dielectric Constants
Transparent Absorb
Solvent  Solvent  Solvent 
n-Hexane 1.9 Water 80.1 Acetone 21.0
Cyclohexane 2.0 Methanol 33 Ethyl acetate 6.1
Benzene 2.3 Ethanol 25.3 Methylene chloride 8.9
Toluene 2.4 Glycol 41.4 Chloroform 4.8
p-Xylene 2.3 Formic acid 51.1 Dimethyl formamide 38.3
Carbon tetrachloride 2.2 Acetic acid 6.2 Methyl formamide 189.0
PTFE (Teflon R )a 2.0 Chlorobenzene 5.7 N-Methylpyrrolidone 32.6
1, 2-Dichloroethane 10.4 Dimethyl sulfoxide 47.2
Trifluoroacetic acid 8.4
aOften used for the construction of pressure vessels for small-scale microwave reactors. Gedye (see
footnote 67) has noted that, owing to microwave power requirements, scale-up is not feasible using
large Teflon vessels.
1. Homogeneous energy supply.
2. Reaction temperature determination and feedback control.
3. Provision for open vessel systems as well as pressure systems.
4. Pressure determination and feedback mechanisms.
5. Continuous-flow reactors with residence time control for larger-scale preparation.
As in domestic applications, the microwave heating of chemical reactions provides
a way of rapidly obtaining high temperatures. It is this rapid attainment of high
temperatures, rather than some special microwave effect, which is believed to be
responsible for most observed results.
The literature indicates that a significant proportion of microwave reactions have
been carried out in an organic solvent. The solvent choice is dependent on the dipole
properties of the reactants. If none of the reactants couples with microwaves, a solvent
which does is needed. Some of the most used solvents that can be categorized as those
essentially transparent to microwaves and those which absorb, and thus heat rapidly,
are identified in Table 1.
Many microwave-assisted reactions have been described using neat conditions;
such conditions can be attractive from an environmental and process productivity
point of view. Neat reactions may be assisted by the presence of, or by support on, a
microwave active solid, such as a zeolite (e.g., a molecular sieve) or Montmorollinite
clay. A great deal of processing flexibility is possible using combinations of the above
with microwave irradiation inputs of varying intensity.
The use of microwave energy to speed analytical sample dissolution and chemical
reactions can be of great value in analytical departments, in laboratory synthesis work
in research departments, and in chemical process development, but is not necessarily
worthwhile in chemical production operations for various reasons. Each case has to
be considered on its own merits. As in all scale-up situations to an eventual industrial
scale, the volume requirement of the product is one of the more important factors,
along with safety, environmental considerations, and the prospects for outsourcing,
inter alia. Product volume, multiplied by the potential savings/kilogram obtainable
usingmicrowave chemistry versus conventional chemistry, provides the dollar amount
that is used in cost calculations needed to justify investment in new equipment. If a
conventional plant suitable for running the conventional thermal alternative is already
in place, or a third party can take on the process, the economicsmay not favor in-house
investment in equipment for large-scale operation (particularly for manufacturing)
unless other important supporting circumstances apply—for example, a patentable
surprise in the microwave chemistry result which encourages confidentiality, a vision
of future applications, a safety or environmental advantage, inter alia. Comment on
a few published reactions illustrates the above:
b.p. 31°C
Microwave (MW)/10 min
Neat (pressure)
325°C +
EXAMPLE 1. Diels–Alder reaction.68
The same reaction carried out (neat) at 100?C (pressure) for 4 hr gave a yield of
Comment. Even if development work improved the microwave reaction to 95+%,
the volume requirement of the product would have to be quite large to justify the
cost of developing the microwave reaction and investing in a high-pressure plant to
capture the potential plant productivity gains achievable with microwave chemistry.
In short, time is not the only important factor in adopting a particular technology.
MW/2 min
96% Formic acid
EXAMPLE 2. Fischer indole synthesis.71
The cyclization did not occur in boiling formic acid alone.
Comment. Although a 20% solution of the nitrohydrazone in polyphosphoric acid
could be cyclized by heating,71(c) the use of polyphosphoric acid is not particularly
desirable. The microwave-assisted process could be worthy of development
if there was no other, more economical, way of accessing this specific indole
derivative—for example, by nitration of the more readily prepared 1,2,3,4-tetrahydro-
MW/6 min
2 K2CO3/0.3 Bu4NBr
HO2C CO2H + 2.5 OctnBr
b.p 198°C
OctnO2C CO2Octn
EXAMPLE 3. Esterification/alkylation.72
The identical reaction without microwave-assistance, again carried out without
solvent and heating for 6 min at 175?C, gave a yield of only 20%.
Comment. This bisesterification (bisalkylation) reaction demonstrates a principle. It
would not be industrially feasible for dioctylterephthalate production since the cost
of the octyl fragment in the n-octyl bromide is at least three times the cost of the
same fragment in 1-octanol, the classical esterifying agent. The large-scale use of
quaternary ammonium compounds, such as tetrabutylammonium bromide, also raises
concerns because their biocidal, algicidal, and fungicidal properties make them toxic
to some sewage systems; however, despite these reservations, quaternary ammonium
compounds are produced on a large scale for use as fabric softeners, bactericides,
and phase transfer catalysts.
Aryl CHO
MW/15 min
H2NOH.HCl (1.5)/N-Methylpyrrolidone (NMP)
12% concentration/100°C
Aryl CN
EXAMPLE 4. Aldehydes to nitriles.73
Comment. Superficially, this one-pot microwave-assisted process looks very attractive.
However, in this case, development of a microwave-assisted process will
depend on comparison with classical procedures.74 For instance, Ganboa and
71(a) Abramovitch, R. A., and Bulman, A. Synlett, 1992, 795. (b) Abramovitch, R. A., and Shapiro, D.
J. Chem. Soc., 1956, 4589. (c) Abramovitch, R. A. J. Chem. Soc., 1956, 4593.
72(a) Loupy, A., Pigeon, P., and Ramdani, M. Tetrahedron, 1996, 52, 6705. (b) A much more promising
approach has been described by Shieh and co-workers (Tetrahedron Lett., 2002, 43, 5607), who advantageously
produced methyl esters in high yield by the microwave irradiation of carboxylic acids using
dimethyl carbonate.
73Chakraborti, A. K., and Kaur, G. Tetrahedron, 1999, 55, 13265. See also Villemin, D., Lalaoui, M., and
Ben Alloum, A. Chem. Ind., 1991, 76.
74March, J. Advanced Organic Chemistry, 4th edition, Wiley-Interscience, New York, 1992, p. 907.
Palomo75 report that various aromatic aldehydes can be converted to nitriles in
94–97% yield by refluxing the aromatic aldehyde, hydroxylamine hydrochloride,
and magnesium sulfate in toluene or xylene, with p-toluenesulfonic acid as catalyst
for 1.5 to 3 hr. The microwave-assisted process may prove better for aliphatic aldehydes
and may be made even more attractive if the above process conditions could be
refined to reduce or eliminate NMP—for instance, if both aldehyde and nitrile form
a homogeneous liquid at the reaction temperature.
MW/5 min
2 Parts acid clay (Bentonite)
350–400°C CO2H
1 part
EXAMPLE 5. Lewis acid-type cyclization.76
This microwave-assisted reaction was carried out on a 3-g scale in a glass vessel
placed in a “bath” of alumina/magnetite. The anthraquinone (m.p. 284?C) produced
was collected as it sublimed from the reactor. Further, o-benzoylbenzoic acid was
added and the reaction repeated. The main advantage of the microwave-assisted
reaction lies in the recycling of the catalyst. The yield in the conventional heating
process falls to 50% after four uses of catalyst, whereas in the microwave-assisted
process the yield is still 84% after fifteen uses.
Comment. The classical industrial process comprises cyclizing o-benzoylbenzoic
acid (from reaction of phthalic anhydride with benzene) with acid using a solvent
or a ball mill process. It is difficult to see how the Bram process could be made
competitive. All solid support processes would seem to require pumping a solid
slurry through a tubular microwave reactor.
R R' R R'
EXAMPLE 6. Polymer-supported peptide synthesis.60
Comment. There may be considerable advantage in commercial polymer-supported
synthesis by employing microwave heating techniques; see earlier section entitled
“Polymer-Supported Synthesis and Reagents.”
75Ganboa, I. and Palomo, C. Synth. Commun., 1983, 13, 219.
76Bram, G., Loupy, A., Majdoub, M., and Petit, A. Chem. Ind., 1991, 396.
47.5 mmol 38 mmol
R=PhCH2, Ph, CH3Ph or MeOPh
TEA (75.5 mmol)
Ethylene glycol
EXAMPLE 7. ?-Lactam synthesis and transformation.77
The authors described the preparation of 25 g of the chiral cis-hydroxylactam
using these two microwave-assisted reactions in one day. Purification of the product
was achieved using chromatography on silica gel.
Comment. The Bose group at Stevens Institute of Technology, New Jersey, has been
especially active in applying microwave-assisted chemistry to the preparation and
further transformation of ?-lactam synthons into other lactams, Taxol R  precursors,
amino sugars, and hydroxyamino acids.78 Some steric control has been observed.78(c)
The Bose group’s prolific applications of the microwave-assisted ketene–imine
annelation process suggest that research laboratories everywhere should be able to
justify acquisition of microwave equipment to accelerate research programs via the
rapid preparation of a variety of new structures. In a process development setting, the
microwave technique should find application in the rapid evaluation and optimization
of process parameters, such as solvent, reaction concentration (and neat reactions),
temperature, time, pressure, the structure of catalyzing bases/acids, rates of addition,
and so on.
Recombinant human
interferon ?-2b
(rh-IFN ?-2b)
Dissolve in 90 µl of
NH4HCO3 buffer
Trypsin/MW 5–10 min
Up to 70%
digestion depending
on reaction temperature
[10 µl of a 3µ g/ µl solution]
EXAMPLE 8. Accelerating enzyme reactions.79
Comment. This remarkable study79 of an enzymic cleavage enabled the authors to
establish that considerable enhancement of the reaction rate could be achieved at
temperatures (>60?C) above those normally used (?37?C) in conventional enzyme
cleavage reactions. Simulating the rapid temperature elevation in the same trypsin
77Banik, B. K., Manhas, M. S., Kaluza, Z., Barakat, K. J., and Bose, A. K. Tetrahedron Lett., 1992, 33,
78(a) Bose, A. K., Banik, B. K., Mathur, C., Wagle, D. R., and Manhas, M. S. Tetraheddron, 2000, 56,
5603. b) Manhas, M. S., Banik, B. K., Mathur, A., Vincent, J. E., and Bose, A. K. Tetrahedron, 2000, 56,
5587. (c) Reference (a), p. 5611.
79Pramanik, B. N., Mirza, U. A., Ing, Y. H., Liu, Y-H., Bartner, P. L.,Weber, P. C., and Bose, A. K. Protein
Sci., 2002, 11, 2676.
digestion using a preheated block at 60?C gave the same result as obtained under
microwave conditions. This novel observation suggests that microwave enhancement
(superheating) of enzyme reactions deserves exploration in all enzyme-mediated
cleavages, at least of proteins. The authors postulate that protein unfolding at the
higher temperatures may be enabling enzyme access to the sites of cleavage. In
the above case, relatively large quantities of enzyme are needed, indicating that
the enzyme is also being deactivated—addition of fresh enzyme after 10 min of
microwaving enhances the enzyme reaction.
Despite more than 20 years of study, the application of microwave irradiation
to chemical process development is still in relative infancy. Microwave equipment
companies continue to address the requirements for large-scale continuous flow and
other reactors.80 The availability of versatile equipment, and preferably a “champion”
in a chemical process development department, would encourage evaluation of the
technology to identify those reactions where the main advantage, enormous reduction
in reaction times (often with cleaner reactions and yield increases beyond those
achievable using conventional conditions), can be harnessed in practical terms.
Evaluation ofmicrowave technology has added merits.Organic synthesis chemists
working in the chemical process development field generally limit themselves to (a)
working at atmospheric pressure in conventional laboratory glassware and (b) carrying
out their reactions using conventional heating techniques. By screening reactions
under microwave activation conditions, chemists expand the range of process parameters
they normally consider. In essence, they move toward the world of the
chemical engineer whose training familiarizes him/her with chemical operations at
higher pressures and temperatures. Bringing chemistry and engineering disciplines
together generally creates invaluable synergy in the chemical process development
field. In this case the chemical engineer may encourage the process development
chemist to collaborate in the design and evaluation of a continuous flow microwave
heated pressure tube reactor to study the effects of “instant” superheating on chemical
If increased reaction yields and/or cleaner reactions are demonstrable, to add to
large reductions in reaction time, one has the best of all possible opportunities to
profitably partner the process development chemist and chemical engineer in the development
of a pilot plant scale microwave reactor as an alternative to conventionally
heated plant equipment. As with any relatively new technology, sober-minded and
realistic evaluation is needed in order to gain credibility with one’s supporters and to
overcome the inertia, often born of skepticism, which is frequently associated with
excursions into the unknown.
The future of microwave technology thus lies with the identification of reactions
with a potential for advantageous results vs. other means. An industrial success would
undoubtedly bring microwave technology into mainstream thinking as a chemical
process development tool.
Our own appreciation of the possibilities of applying microwave technology arose
from a single experiment carried out by Dr. Y.-H. Ing, working in Dr. Birendra
80Following the lead of Peterson, C. New Scientist, 1989, 123, 44, working at CSIRO, Australia.
50 100 150
200 250
297 225 201 185 158 115 93
75 57
2 minute 0 minute
Relative Abundance
Relative Abundanc
50 100 150
200 250
FIGURE 1. Electrospray–MS of microwave-assisted hydrolysis of dihydrohypoxanthine tri-
Pramanik’s group. He reasoned that the hydrolysis of dihydrohypoxanthine to
5-aminoimidazole-4-carboxamide (AIC) (see Scheme 19 and Figure 14 in Case Study
2) might be conveniently carried out using microwave irradiation. He dissolved dihydrohypoxanthine
trifluoroacetate salt (81.1 mg) in 1:1 trifluoroacetic acid:water
(2 ml) and micronized the solution for 2 min at ambient pressure. After this time the
electrospray–MS spectrum of the solution indicated that substantial AIC formation
had occurred (Figure 1).
Should the economics of the above-described new process favor a switch from the
present AIC process (see Case Study 2) to one based on dihydrohypoxanthine new
investment might be justified in new production equipment.81 A switch to the microwave
option versus the conventional heating option would depend on the outcome
of process optimization work on both. If a rapid clean reaction in near-quantitative
yield were realized by optimizing the microwave conditions, then it would prove
worthwhile to engineer a simple continuous flow reactor for further development
work. The case in favor of adopting the microwave-assisted process would be further
enhanced if the solution from the microwave reactor could be used directly in the
next process step (thus eliminating the need for isolation equipment).
Conclusion. From a chemical process development perspective, the principal advantage
favoring the use of microwave irradiation in promoting chemical reactions lies
in achieving shorter reaction times. If cleaner reactions giving higher yields are also
demonstrated, they add greatly to the attractiveness.
As indicated by comments on a few published microwave-assisted reactions, microwave
technology is no panacea and appears likely to be applicable only to selected
81New situations generally provide the best opportunity for the evaluation and introduction of new technology.
cases. Much work will also be needed to validate the technology for commercial
Nevertheless, the availability of a microwave oven, and a champion of the technology
in the laboratory, would allow chemical process development scientists and
engineers to satisfy themselves on questions concerning the limitations of conventional
heating. The rapid evaluation of superheating effects may well reveal opportunities
missed by restriction to conventional heating, including demonstrating
transformations that cannot be achieved under conventional conditions. Also, in process
development terms, a given microwave-induced reaction might be developed
more efficiently by the rapid screening and evaluation of process parameters.
The bottom line is that the broader application of microwave technology is worth
pursuing. Any eventual commercial adoption of technology such as a microwaveirradiated
continuous reactor system generating attractive economics, both in capital
and operating terms, would stimulate broader interest in evaluating the technology.
Electricity has generally been recognized as a probable promoter of organic synthesis
going back billions of years. A report by Sutherland and Whitfield summarizes some
of the evidence for this.82 Thus, electrical spark discharges in the presence of simple
molecules believed to be present in the planet’s prebiotic atmosphere have been
shown to produce a few of the building blocks needed for the creation of life. The
sun also provided other forms of energy (e.g., infrared and ultraviolet radiation, inter
alia). Some of the processes proved by laboratory experiments are:
H2 + N2 + CO
Electric spark discharge
Electric spark discharge
(CN)2 +
CH4 + NH3
The melting of ice around 4 billion years ago introduced the water needed for the
beginnings of the organic chemical processes essential for the production of the amino
acids, purines, and pyrimidines required for the evolution of early life forms. Thus
CH4 + NH3 + H2 + H2O Amino acids Electric discharge
Interestingly, the proportions of the mostly racemic amino acids produced in this
mixture are approximately the same as those found in fragments of the Murchison
meteorite, which fell near the village of Murchison, Victoria, Australia in 1969.82,83
The self-condensation of hydrogen cyanide has been shown to lead to the formation
of the purines adenine and guanine, two of the bases needed for the formation ofDNA.
82Sutherland, J. D., and Whitfield, J. N. Tetrahedron, Report Number 425, 1997, 53, 11493–11527,
Prebiotic Chemistry: A Bioorganic Perspective.
83Some N15 enrichment in individual Murchison amino acids (versus terrestrial counterparts) suggests an
extraterrestrial origin for an l-enantiomer excess-Engel, M. H., and Macko S. A. Nature, 1997, 389, 265.
TABLE 2. Comparison of the Cost of Electricity and the Costs of Chemicals Commonly
Used in Oxidation and Reduction
Reagent Cost/kgb ($) Moles/kg Cents/Mole Cents/Equivalent
Electrons at 6c/kWh, 3.5 va 0.6
Sulfur dioxide 0.54 15.63 3.5 1.75
Chlorine 0.65 14.08 4.6 2.3
Hydrogen peroxide 2.27 29.4 7.8 3.9
Zinc (metal) 1.80 15.2 11.9 5.95
Chromic acid (CrO3) 4.10 10.0 41.0 6.8
Sodium (metal) 3.10 43.4 7.2 7.2
Hydrazine 5.85 31.2 19.17 9.58 (4.79)c
Sodium hydrosulfite 1.85 5.74 32.1 16.05
Potassium permanganate 3.45 6.33 54.47 18.1
Magnesium (metal) 19.84 41.6 47.9 23.95
Sodium borohydride 74.00 26.3 299.5 159.75
aInformation kindly provided by the Electrosynthesis Company, Lancaster, New York.
bFigures from Chemical Marketing Reporter, April 2007, and Dr. Prashant Savle, Schering–Plough.
cBracketed figure assumes that four hydrogens are available.
It has also been found that, of the two pyrimidines needed for the preparation of DNA,
cytosine can be produced from cyanoacetylene.82
However, despite its apparently fundamental role in creating the building blocks
of life, electricity has not become one of the foremost means of manipulating organic
molecules to useful end.84 Nevertheless, where it can be applied to remove
or add electrons in specific ways, electricity offers unique and often advantageous
approaches to carrying out oxidation and reduction reactions. The interested reader
is referred to several of the books published on the subject for further education.85
Obviously, air (oxygen) is the lowest-cost reagent for carrying out certain oxidation
reactions. Added to this the electrolysis of water affords a convenient, low-cost
means of producing oxygen, and hydrogen, in concentrated form, for a large range
of oxidation and reduction reactions practiced commercially today. It follows, in
going back to fundamentals, that the removal or addition of electrons could provide
the lowest cost route for carrying out oxidation or reduction reactions. This can
be illustrated by comparing the cost of electricity with the cost of some of the
common chemicals used in oxidation and reduction in the organic chemistry field
(Table 2).
84Electrochemistry is, of course, commercially very well established in the inorganic chemicals
industry—for example, for the manufacture of chlorine, aluminum, sodium, and sodium hydroxide, inter
85(a) Torii, S. Electroorganic Synthesis, Methods and Applications, Kodansha Ltd., Tokyo; VCH, Weinheim,
Germany; VCH Publishers, Deerfield Beach, Florida, 1985. (b) Pletcher, D., and Walsh, F. C.
Industrial Electrochemistry, 2nd edition, Chapman and Hall, London, 1990. (c) Lund, H., and Baizer, M.
M. Organic Electrochemistry, An Introduction and Guide, 3rd edition, Marcel Dekker, New York, 1991.
Although relatively few organic chemicals are produced on a commercial scale
using electrochemical means, those that are produced owe their existence to the
commitment and perseverance of individual scientists, and no doubt to the wisdom
of the visionary leaders who supported these scientists’ inspiration and provided the
There are always setbacks which damage the credibility of a technology.86 Again,
Nalco’s elegant electrochemical process for the manufacture of tetraalkyl lead, installed
in the mid-1960s, was highly successful until lead was phased out of gasoline
in the 1980s. Probably the most successful electroorganic process, which has been
running commercially for more than 40 years, is the Monsanto process for the hydrodimerization
of acrylonitrile to adiponitrile (reference 85(c), p. 1317).
Emulsion with aq. phosphate buffer, lead alloy, or
steel anode/lead or cadmium cathode, 55°C
This reaction is conducted on a scale of >200,000 tonnes/annum. More recently,
another elegant application of electrochemistry on a multi-thousand-tonne scale
has been HydroQuebec’s investment in a cerium (IV)-mediated oxidation of naphthalene
to naphthaquinone, a process licensed from W.R. Grace.87 HydroQuebec
then uses the naphthaquinone in a Diels–Alder reaction with butadiene to produce
Pt Anode/5 M MeSO3H
This reaction is noteworthy in its use of the expensive cerium IVsalt, which is recycled
very efficiently in the anodic oxidation process thus reducing its contribution to the
cost of the anthraquinone product.
A further illustration of the variety of reactions that can be carried out
by electrochemical processes is BASF’s production, on a small scale, of phydroxybenzaldehyde.
86The Atlas Powder Company’s large plant for the manufacture of mannitol and sorbitol by the cathodic
reduction of glucose was rendered obsolete within a few years by a high-pressure catalytic hydrogenation
87Kreh, R. P., Spotnitz, R. M., and Lundquist, J. T. J. Org Chem., 1989, 54, 1526.
88Barl, M., Degner, D., Siegel, H., and Hoffmann, W. European Patent, 0025883, 1983 (to BASF).
Carbon anode/CH3OH-KF
electrolyte (?e, ?H+, ?e)
CH3OH Fast
ex p-cresol
The above are but a few of the many reactions that have been carried out by electrochemical
means. Reference 85(c) describes these and several other processes that
have been conducted on a semicommercial or pilot plant scale, in greater depth.89 A
complementary review of organic electrosyntheses in industry, by Degner (BASF),90
provides an extensive review, with 633 references, of the patent literature into 1987.
Degner expresses the opinion that themost likely successeswith electroorganic chemistry
in the future will come from continuing work in the areas outlined above—that
is, from cathodic hydrodimerization (e.g., the C–C couplings such as the abovedescribed
approach to adiponitrile manufacture), the electrochemical regeneration of
expensive reagents (see the above naphthaquinone example), and the functionalization
of alkenes and aromatic compounds, as in the BASF p-hydroxybenzaldehyde
process. Other reviews91 provide further examples of such reactions and suggest that
electrochemistry has a larger role to play in the pharmaceutical industry. Indeed all
three reviews provide examples of the use of electrochemical methods especially in
the deprotection of a variety of amino protecting groups and in the transformation and
functionalisation opportunities afforded by electrochemistry. It also appears that electrochemistry
has a valuable part to play in chiral synthesis—for example, through the
use of electrochemical systems to control the oxidation state of enzymes. A review
by Steckhan92 describes many opportunities for the application of electroenzymic
oxidations and reductions, opening up a new field in bioelectrosynthesis.
Our own interest in applying electrochemistry grew from the highly successful
process developed for the manufacture of the third-generation cephalosporin antibiotic,
Ceftibuten.93 The electrochemical reduction component of the process and the
product extraction step are, in outline, as follows:
89See Danly, D. E., and King, C. J. H. Chapter 31 in reference 85(c), p. 1285.
90Degner, D. Topics Curr. Chem., 1988, 148, 1–95.
91(a) Utley, J. Chem. Ind., 1994, 215. (b) Genders, J. D., and Pletcher, D. Chem. Ind., 1996, 682. (c) Ban,
Y. Chapter 19, Natural Products and Pharmaceuticals, in reference 85(c), p. 765.
92Steckhan, E. Topics Curr. Chem., 1994, 170, 83.
93Detail of the entire process is provided in Chapter 9 (q.v.) and in the papers by (a) Bernasconi, E., Lee,
J., Roletto, J., Sogli, L., and Walker, D. J. Org. Proc. Dev., 2002, 6, 152. (b) Bernasconi, E., Genders, D.,
Lee, J., Longoni, D., Martin, C. R., Menon, V., Roletto, J., Sogli, L., Walker, D., Zappi, G., Zelenay, P.,
and Zhang, H. J. Org. Proc. Res. Dev., 2002, 6, 158. (c) Bernasconi, E., Lee, J., Sogli, L., and Walker,
D. J. Org. Proc. Res. Dev., 2002, 6, 169. (d) Chai, D., Genders, D., Weinberg, N., Zappi, G., Bernasconi,
E., Lee, J., Roletto, J., Sogli, L., Walker, D., Martin, C. R., Menon, V., Zelenay, P., and Zhang, H. J. Org.
Proc. Res. Dev., 2002, 6, 178.
Tin mesh
+e, -AcO , + H
+2H+, Ph2CN2 in
CO2CH (Ph)2
(Ph)2CH GluN
This process grew out of earlier efforts by several pharmaceutical companies, Eli
Lilly, Takeda and Shionogi, to electrochemically reduce 3-acetoxycephalosporins,
such as cephalosporin C and its derivatives to 3-exomethylenecephalosporins.94
3-Exomethylenecephalosporin Cephalosporin C, R= HO2C CH(NH2) (CH2)3
Cephalothin, R=
Evaluation of the process as a candidate for scale-up and further development evidently
did not meet the Lilly or Takeda requirements. It could be that a champion of
the new technology never emerged. Indeed, when Schering–Plough took up the idea
for the manufacture of Ceftibuten intermediates, our partner, Shionogi, based on their
earlier knowledge of Eli Lilly and Takeda work, were only lukewarm supporters. The
main objections were as follows:
94As happens from time to time, the beginnings of this electrochemical process emerged from work
undertaken by another pharmaceutical company, Glaxo Laboratories for quite different reasons. Thus,
Glaxo analytical chemists, investigating the use of polarographic methods for the quantitative analysis
of cephalosporins, especially cephaloridine, were the first to observe that these compounds underwent
electrochemical reduction (Jones, I. F., Page, J. E., and Rhodes, C. T. J. Pharm. Pharmacol., 1968, 20,
455). Fundamental investigations by Lilly and Takeda workers led to further exploration of the above
electrochemical process (Hall, D. A. J. Pharm. Sci., 1973, 62, 980; Hall, D. A., Berry, D. M., and
Schneider, C. J. J. Electroanal.Chem., 1977, 80, 155; and Ochiai, M., Aki, O., Morimoto, A., Okada, T.,
Shinozaki, K., and Asahi, Y. J. Chem. Soc., Perkin 1, 1974, 258).
1. The use of a mercury cathode is environmentally unacceptable. (Our early
work was done using a conventional electrochemical cell using a mercury pool
cathode—see later).
2. The electrochemical reduction process applied to cephalosporins gave <70%
yield of desired exomethylene product, thereby introducing substantial purifi-
cation problems to add to the yield losses.
3. The productivity of the process (reaction concentration)was low(concentration
?5 g/liter).
4. The current density, again impacting productivity, was low.
5. No practical means of extracting the product was known.
6. In the early 1970s, electrochemical technology for processing organic compounds
was in its infancy and pharmaceutical manufacturers were inexperienced
in it and wary of it.
7. Concern was expressed regarding the possible impact of new technology on
product quality.
8. Why change? The production process operated by Shionogi was working well,
if not as cheaply as the projected costs (we were reminded that our projections
were unproven!) of the electrochemical route. However, limitations on Shionogi
plant capacity for larger Ceftibuten production volumes, originally anticipated,
were such that Shionogi supported evaluation of the electrochemical option
through the pilot plant phase of process development.
The events that followed resulted from confluence of several critical factors:
1. The vision, leadership, and support of senior Schering–Plough executives,
particularly Dr.HalWolkoff, SeniorV.P. of allDevelopment, andMr. John Nine,
Senior V.P. of all Manufacturing operations, encouraged further exploration of
the electrochemical process.
2. Continued and expanded funding of the Colorado State University program
under the enthusiastic direction of Professor Charles R. Martin was agreed to
be an essential component of the research effort.
3. Enlargement of the evaluation and development of the electrochemical component
of the project, by engaging the Electrosynthesis Company, Lancaster,
New York, and particularly harnessing the experience of their industrial electrochemistry
experts, Drs. Norman Weinberg and David Genders, was agreed
to offer the best means of gaining industrial perspective.
4. Commitments to purchase pilot plant quantities of key intermediates from
Antibioticos, our cephalosporin-producing partner in the development of the
electrochemical reduction process, was also considered essential in the securement
of a cost-effective supplier of raw materials.
5. Legal agreements covering objectives and intellectual property ownership approved
by all parties involved, were drawn up to ensure that all parties knew
where they stood.
6. Agreements on regular minuted meetings with all parties, including Shionogi,
to report progress and share information were considered the best vehicle to
ensure that progress was based on everyone’s understanding of the issues.
7. Day-to-day coordination of activities was agreed as best run by
Schering–Plough to ensure all the players in all the organisations involved
were working together as effectively as possible. Most of the Schering–Plough
component of this was very ably handled by Dr. Junning Lee.
As a result of this organization, rapid progress was made in overcoming the list of
objections (outlined earlier) which had not been overcome by Eli Lilly and Takeda
and which were, initially, legitimate concerns to Shionogi. In brief, the milestone
events were as follows.
Alternative Substrates. Although we in Schering–Plough did not find advantage in
replacing the acetoxy leaving group in the cephalosporin with better leaving groups
(e.g., pyridinium, chloro, chloroacetoxy), we quickly discovered that oxidation of
the cephalosporin sulfide atom to its sulfoxide produced a superior substrate for the
electrochemical reduction; yields to the desired exomethylene sulfoxide intermediate
exceeded 90%, and none of the thiazole-type impurities generated in the reduction
of the earlier cephalosporin sulfide were produced. Preparation of the sulfoxide was
easily carried out in situ by Antibioticos such that their preparation of glutaroylcephalosporanic
acid sulfoxide became the favored route to provide the substrate
of choice. We expected, and later proved, that the sulfoxide group could be readily
reduced to the sulfide later in the process (see Chapter 9 and previously cited
Replacing the Mercury Cathode. Professor Martin and his colleagues undertook
an entrepreneurial evaluation of alternative cathode materials and process conditions
using the new sulfoxide substrate. They identified tin as the closest to mercury.
Dr. David Genders proposed that a greatly increased surface area of the cathode
would help to increase the yield of the desired exomethylene sulfoxide and speed the
reaction. This indeed proved to be the case.
The work in both Colorado State University and the Electrosynthesis Company,
drawing on the free expression of university colleagues and the practical know-how
and experience of industrial Electrosynthesis experts, enhanced the rate of progress
leading to the technical success of the project.
Process Productivity. We quickly showed that, starting with the more stable sulfoxide,
the reaction concentration could be advantageously increased to 50g/liter, and
the current density to 100–200 mA cm?2.
Product Extraction. The little used but very efficient method of using diphenyldiazomethane
for the extractive esterification of acidic materials from aqueous solution
proved to be very effective [see footnote 93(c)].
Product Quality. Work was undertaken to show that the Ceftibuten product produced
via the electrochemical route was analytically acceptable (indeed at least as pure) as
that produced by the Shionogi process.
Why Change? The pilot plant work carried out by the Electrosynthesis company93(d)
enabled us to validate our earlier cost calculations. However, by this time the market
for Ceftibuten was declining and the justification for investment in new plant disappeared.
Thus although the process proved to be an enormous technical success, it was
a commercial failure.
In completing our R&D contract with Professor Martin, we tested two other potential
uses of electrochemistry in fields of interest. The first was to improve the
well-established Marker chromic acid (chromium VI) oxidation process for converting
diosgenin acetate to the 20-keto intermediate needed for the manufacture
of Schering–Plough’s line of betamethasone products (see Scheme 3 in Chapter 9).
The improvement sought was to use anodic oxidation to recycle the chromium III
waste back to chromium VI, for reuse on-site, thus avoiding or greatly reducing the
need to precipitate chromium hydroxide from the waste and pay to have it sent to
chromium waste disposers for recycle. Unfortunately, despite our demonstrating the
feasibility of the process, the return on investment (ROI) did not justify expenditure
in anodic oxidation equipment for the relatively small quantity of chromium III waste
generated in our Mexican plant.
The second project was to determine whether the anodic oxidation of benzophenone
hydrazone could be manipulated to produce diphenyldiazomethane (DDM) as
the end product in high yield. Literature evidence95 indicated that the anodic oxidation
of benzophenone hydrazone using either a platinum or graphite anode under
various conditions with a variety of electrolytes gave a number of products that were
shown to derive from the intermediacy of DDM:
Pt anode/RT/LiClO4-CH3CN
Pt anode/RT
Graphite anode/reflux
As above in presence of
methyl methacrylate
Ph2C(OMe)2+Ph2CH OMe + Ph2CH2
Chiba and co-workers gave no indication that the anodic oxidation of benzophenone
hydrazone could be stopped at the DDM stage. Professor Martin and Dr. John
Hulteen, Colorado State University, were, however, able to devise conditions, based
95Chiba, T., Okimoto, M., Nagai, H., and Takata, Y. J. Org. Chem., 1983, 48, 2968.
FIGURE 2. Electrochemical oxidation of Benzophenone Hydrazone to Diphenyldiazomethane.
on those described by Glaxo workers,96 which produced DDM, using a platinum
anode in the divided cell described in an earlier publication [see footnote 93(b)]. The
process conditions were as follows:
Benzophenone hydrazone (5.88 g, 20 mM) was dissolved in methylene chloride
(20 ml) and over-layered with 1 M sodium hydroxide (40 ml) containing, as phase
transfer catalyst, tetrabutylammonium sulfate (0.68 g) and sodium iodide (300 mg).
The cathode half cell contained 1 M sodium hydroxide (60ml). The whole cell was
cooled to 0?C, the anode compartment stirred and electrolysed at a current of 50 mA.
Formation of DDM was followed using the DDM absorption peak at 525 nm. The
chart obtained was as shown in Figure 2.
Conclusion. The above outline of successes achieved in using electrochemical methods
to carry out oxidation and reduction reactions in the field of organic chemical
transformations underlines the merits of electrochemistry.
These successes, the relatively low cost of electricity and especially the environmental
advantages that electrochemical methods afford, justify the wider evaluation
of electrochemistry as a first-line technology for oxidations and reductions in chemical
process development work. Today, the only factors standing in the way of this
are the lack of some education in the practical applications of electrochemistry, management
encouragement to reach out to university chemistry departments, as well
as to specialist industrial companies, working in the field, and the availability of
an electrochemical cell in chemical process development laboratories everywhere.
Regarding electrochemical cells, the reader is referred to the Lund and Baizer book
96(a) Adamson, J. R., Bywood, R., Eastlick, D. T., Gallagher, G., Walker, D., and Wilson, E. M. J. Chem.
Soc., Perkin Trans. I, 1975, 2030. (b) Bywood, R., Gallagher, G., Sharma, G. K., and Walker, D. J. Chem.
Soc., Perkin Trans. I, 1975, 2019.
[see footnote 85(c)] and specifically the chapter by Professor Lund, which describes
laboratory cell construction, electrode materials, operating parameters, solvents for
electrolysis, and electrolytes.97 A sketch of the cell used in our work in Colorado
State University is provided in reference 93(b).
Going back to the beginning, I hope that this somewhat superficial outline of
electrochemical process successes will be enough to stimulate the imaginations of
chemical process development scientists and engineers everywhere to explore the universe
of electron transfers which is out there and to adopt electrochemical technology
as a first-line endeavor.
Sustainable Development
Introduction. Since present human development is not sustainable the concept of
sustainable development requires that human expression has, in this century, to be
harmonized to control all the world’s runaway dangers, not only the profligate use of
the finite resources of the natural world, in order to create and maintain a stable slowly
evolving steady state designed to ensure long-term survival. A vast re-education of all
the world’s diverse cultures and social systems is becoming increasingly necessary
to create forms of human living that lead to harmonious development and “growth”
of a sugstainable kind.
Science, which has been the badly exploited core of development to this point,
will inevitably continue to be one of the main engines of the re-education process.
The power of current driving forces has to be urgently revolutionized, and more
sustainable forms of living have to rapidly emerge in order to secure greater world
harmony and stability. A re-education plan needs to be formulated and aggressively
funded, worldwide. In short, re-education for the near future, evolving into a longterm
program of education for long-term survival has to be placed on a war footing.
Self-preservation has always been the supreme motivating force of mankind, developing,
with little concern for the consequences, into the out-of-balance (excessive)
me-phenomenon of today. In various guises, “me” is a human phenomenon on a planetary
scale disrupting not only local social systems but, more ominously, national
and continental ways of life. In a word, “me” has set cultures into conflict with one
another. The problem has escalated rapidly in the last hundred years as liberalism
spawned self-determination in those who are not equipped to deal with the consequences
of free expression. Leaders of the world have seemed powerless to deal
with the problem. They are mostly seen as self-serving themselves, lacking in vision
and moral courage to seek the collaborations needed among the great world movements
and ideologies (democratic capitalism, socialism/communism and religion),
to address the problems.
Sustainable development might best be achieved by creating new social models
based on reexamining and reconstructing, from today’s vantage point, the ways
in which early families and village communities survived and advanced through
acquisition of the skills needed to guarantee their water and food supply, the provision
97Lund, H. Chapter 6, p. 253 in reference 85(c): Practical Problems in Electrosynthesis.
of shelter, the availability of energy sources, and the organizations they needed
in their own battles for self-preservation. Early life developed into larger, more
fractious societies, enabling scientific and other discoveries, leading to the industrial
revolution and the wartime disasters of the 20th century continuing! Going back
and remodeling history, by integrating and addressing the major consequences of
the process of human development, ignored at the time, may enable modern man to
determine better ways of utilizing water and other finite natural resources, as well as
creating sustainable social systems, continually educating the population intomore of
a selfless development mode, evolving leadership and still accommodating individual
creativity in a way that harmonises with the evolution of a steady state. Basically,
creativity has generated the wealth that sustains us all. But it is mostly the excesses
resulting from creativity, such as the abuse of power and profiteering, which have
produced the ignored consequences. The issue thus becomes how to fashion more
“WE phenomenon” while still preserving individual creativity.
Such modeling may identify how an ideal planetary world may evolve, but it
will undoubtedly be unimaginably difficult to reformulate the present-day human
condition at a time when the “me” world is irreversibly consumed by its accelerating,
consequence-ignoring, out-of-control growth. It hardly seems possible that modern
man can be persuaded to adopt a long-term holistic approach to life without first
precipitating some remorseless, unstoppable cataclysmic event with a catastrophic
loss of planetary life, leading to new foundations.
Everyone prepared to think about how to find ways of avoiding or minimising the
impact of a cataclysmic event quickly realizes that possible solutions have already
been identified. For me, pursuing the goal means more vigorously addressing only
three core matters. These are conservation, especially of energy use, much wider
education, especially enabling the have-not third-world countries achieve greater
health and wealth, and population stabilization, especially gained through education.
Energy conservation is only a minor component of the relatively small-scale
operations in the pharmaceutical industry. Nevertheless, it needs to be addressed. In
regard to the very large scale, those using organic chemical-based transport systems
need to be urgently obliged to become much more efficient consumers of fossil fuels.
To the chemist, seeing theworld’s organic chemical feedstock sowastefully converted
to carbon dioxide is almost a crime, whether or not you believe carbon dioxide also
contributes to global warming. The use of biologically derived ethanol and other biofuels—
preferably produced from abundant cellulosic raw materials, and even biogas
from animal slurry waste and manure98—is only a stopgap activity. Wind, waves,
tides, and sea current sources of electricity are being developed. They are not as limited
as hydroelectric power (available only from rivers and reservoirs), but they need
more research and development, including in better battery technology, to become
more mainstream. Solar power is better established and geothermal energy sources
could and will be increasingly harnessed as research and development improves their
contribution. These technologies are all likely to be only supplementary sources of
electricity for the near future. Fossil fuels, with greatly improved pollution controls in
98Doubtless, human foulwaste could also be used as a source of biogas, especially if it could be concentrated
by separating it from general domestic sewage in an economical way.
the case of coal,will be themajor sources of energy, especially electricity, over the next
several decades. Only nuclear energy, despite its fearful waste and terrorism baggage,
will become the major source of electricity for the foreseeable future.99 Much more
research and development is being undertaken, especially in waste recycle and also
in the cleaner nuclear fusion alternative, which will undoubtedly improve the already
quite favorable economics. Terror threats might be expected to decline given more
education. It would thus appear to this writer that electricity will become man’s
long-term source of clean energy, leaving fossil fuels for more specialist uses and,
particularly, for the world’s chemical industry.
As for education and population stabilization, the world only needs to watch and,
as requested, discretely help India, which seems to contain all the world’s problems
in “microcosm.” India, with its 700 million rural people in 600,000 villages, and its
educated class of more than 300 million, is already working, if painfully slowly, to
modernize its social and economic systems to raise itself to the level of the most
developed nations. Its success could provide a model for all the third-world nations.
However, as already mentioned, the so-called developed world needs, itself, to do far,
far more to educate and transform itself away from its excessive, me-phenomenon
obsession into a we-phenomenon state, providing an example for the underdeveloped
Education is the key requirement for all to undertake. It is probably too much to
ask all but a few to subscribe to Darwin’s view.100 But there is much to be said for
re-educating those populations, excessively given to seeking pleasure in unhealthful
diversions, to use their resources in more worthy social pursuits. The needs were no
doubt apparent before Roman times but were well expressed by John of Salisbury in
Sustainable Chemical Process Development in the
Pharmaceutical Industry
As in all fields of endeavor, the pursuit of excellence in education and re-education, in
any discipline, and particularly in the integration of disciplines that need to work together
is the prime requirement for progress and success. Specialization is inevitable,
but overspecialization and overfocus on narrow areas of endeavor (whether in a single
science discipline or, in general, obsessing on a single diversionary pleasure)
leads to narrow programming of the human mind. Excessive narrowing inhibits the
99The cost of uranium makes up between 5 and 10 percent of the total cost of electricity production in a
nuclear power plant. Strikingly, 5 g of uranium corresponds to 700 kg of coal, which also emits up to 1.7
tonnes of CO2! Vatanen, A. The Scotsman, March 9, 2007, p. 34.
100“A man who dares to waste one hour of life has not discovered the value of life.”
101He railed against people wasting their time enjoying performers (jugglers, musicians, actors, etc.):
“tedium steals upon unoccupied minds and they are not able to endure their own company unless they
are pampered by the solace of some pleasure. Therefore spectacles and the countless hosts of vanities by
which they who cannot endure to be entirely idle are occupied, but to their greater harm. Better it had been
for them to have idled away their time than to have busied themselves to their own ruin. . .” from John
of Salisbury’s work, “Policraticus.” What would he have said had he seen the excessive diversions that
consume today’s world.
ability to deal with the unexpected.102 Managements in any field therefore need to
ensure that the general philosophy in undertaking interdisciplinary work is founded
on preservation of the freedom and climate to enable their educated creative scientists
to express themselves in every collaboration.103 Working in this way, managements
in the chemical process development field create the best chance of discovering and
developing the best104 and most sustainable API manufacturing processes.
In terms of chemical process development, such a philosophy is not apparent
today. In my experience, most new commercial processes for API manufacture are
substantially suboptimal and therefore higher in cost than they should be. Generally,
progress toward better processes has been compromised by the overly zealous
compliance with regulatory guidance resulting from timid managements having succumbed
to excessive misinterpretation of regulatory intent. Rightly the regulatory
agencies (safety, environment, and FDA regulatory affairs) are continually striving
to ensure that the public is protected from harm. A major regulatory objective is
to prevent those entrepreneurs who have little regard for the consequences of their
self-serving activities from inflicting harm.105 Another objective is to raise standards
to prevent those with inadequate knowledge (which might trigger an adverse
event) from making mistakes. It is usually individuals and sometimes companies
who are excessively focused on personal and corporate profit who, neglecting or
ignoring consequences, spoil the field for the many reputable companies and individuals
who take the care to accommodate consequences and to protect public
One possible solution, enabling chemists and engineers to create optimal processes,
has been outlined in the earlier essay on Bureaucracy Reduction (q.v.).106
The proposed solution allows for high-integrity people and companies to become
qualified with special freedom to keep innovation going in practice throughout the
whole exercise of developing a chemical process for use in API manufacture. Such
an approach is in concert with the other great need, to raise education to a war
In terms of creating sustainability, enhancing education and reducing bureaucracy
need leadership and commitments from the most senior company executives. Such
102As an example, take the excessive use of the eyes only to tune manual dexterity to operate a computer
game as fast as possible. An addiction to such excess may have the adverse effect of stunting other senses.
As a result, language skills, people interactions, broad problem-solving abilities, general curiosity for the
natural world, personal health, and so on, may be affected to the point of eventually impacting on survival.
103This is not a carte blanche freedom but envisaged as one needing to gain the support of visionary people,
not bureaucrats. See comments on leadership in Chapter 2.
104Best in being the safest and most environmentally sound (including in waste minimization and recycling),
with superior economics (in labor, equipment, and materials useage). Best is, of course, a relative
term since frequently, even better processes emerge from ongoing research and development.
105Sadly, practices and laws that should be universal (not all laws should be) do not become so. It is
astonishing that following the 1938, FDA-driven outlawing of the use of diethylene glycol in medicinal
formulations (see Chapter 6), several countries have reported deaths caused by medicines contaminated
with diethylene glycol; most recently, 88 children died in Haiti (1995/1996), 33 children died in India
(1998), and more than 22 people died in Panama (2006).
106I can again challenge readers of this book to offer other practical alternatives to the chromatographybased
system proposed.
support is essential but will need better internal company structures than now exist.
An internal structure, which separates research and development from the time driven
systems currently in place everywhere, is essential for the sustainable discovery of
APIs and the development of processes to produce them. The current pressures
for faster progress usually result in incomplete and short-term thinking. Financial
investors and analysts play a large part in promoting company business strategies.
They are often seen as asking companies “What have you done for me lately.” They
seek quarterly reports from an industry that moves only slowly. Such attitudes can
play havoc with the selection and progression of a drug through the FDA approval
Excessive speed and disregard of the consequences of going very fast greatly
increase the chance of making mistakes and missing API opportunities, as well as
reducing costs. Despite this, stockholders and executives prefer to go for the best
chance. Powering those APIs (which look likely to be approved by the FDA) through
the system as fast as possible makes money for investors at the fastest possible rate.
But speed is at the mercy of the clinical data. Unexpected clinical findings, with no
clear understanding of the reasons, can greatly retard progression of an API through
the FDA approval process.107 Speed was also undoubtedly a factor in missing the
opportunity to patent a metabolite of Schering–Plough’s anti-androgen, Flutamide,
thereby limiting the drug’s market potential (see Chapter 6).
In chemical process development programs, discontinuities in introducing new
processes, due to regulatory constraints, frequently cause the selection of suboptimal
processes for development. Once established in manufacturing, even suboptimal
processes can be difficult to replace (see earlier essay on Bureaucracy Reduction).
Other agendas have also to be considered, as evidenced in our efforts to develop a
new process for albuterol manufacture (see Chapter 5). In another case, lack of time,
and perhaps also a lack of faith in the future market potential of Temozolomide, were
major factors retarding the development of new processes for the manufacture of this
API (see Case Study 2).
In the effort to pursue profit as rapidly as possible, short-term thinking is the
ascendant philosophy. Today it is speed from the discovery phase to the market that is
the accepted driving force of any pharmaceutical company. I know of no studies that
might challenge this philosophy. But it may well not be true for those APIs where
speed has caused potentially new APIs (e.g., metabolites) to be missed. And again,
developing suboptimal processes is not only expensive in its own right, but later
changing to a more efficient lower-cost process can take years at great expense (see
Bureaucracy Reduction). As a result of all the turbulence associated with developing
an API to the marketplace, including dealing with all the inhibitions to introducing
optimal processes for API manufacture, it is clear that achieving sustainable
development will be a difficult task. There is an urgent need for more enlightened
107The finding of cataracts in the clinical trials of Schering–Plough’s anti-inflammatory drug, Flunixin,
delayed the drug’s approval. It took years to resolve the issue by which time the patentwas near expiry. This
led to Flunixin being marketed only as a lower profit animal health drug. Schering–Plough’s nonsedating
antihistamine, Loratadine also suffered long delays in gaining FDA approval due to being caught up in a
liver enzyme induction issue raised by the FDA.
thinking on the structural changes needed to enable all of the research and development
components in any endeavor, including industrial, university, and government
components, to work together. Education, exploration, and experiment are all needed
to determine how the earth’s limited physical resources, from which all wealth is
drawn, can be harvested by individual stable populations for the benefit of all life
coexisting in a symbiotic way with the physical world. It seems, from the perspective
of today’s world scene, almost impossible to imagine sustainable development in a
slowly evolving steady state world,managed by a reformedmankind. A stepwise progression
is all we can promote starting with education, conservation, and population
I intend, in invoking fantasy, to provoke the wildest dreams and actions in those
who can make a difference. There are many enormous “chemistry” problems to
solve, with the human condition (the “chemistry” of the human brain, including brain
diseases) providing the greatest challenges. Over the millennia the human condition
has spawned many,many problems that grow larger with human development. Yet the
consequences of human development seem to attract, proportionately, diminishing
attention. Knowledge, to deal with consequences, is gathered only slowly and, sad to
say, no bold visionaries are emerging with the power to lead the proactive and painful
crusades required to restore balance. Of the many, many problems, educating people
enabling them to become socially responsible is surely the core endeavor on the planet
recognizing that education needs to be integrated with the reformulation of family and
social (including religious) agendas to create and implement new action plans—in
short an interdisciplinary endeavor. Unfortunately, problems increase as the planet
becomes bogged down by overly self-serving activities, plagued by intolerance and
befuddled by bureaucracy, leaving only a small-mindedness and a level of thinking
about the future which is far, far too trivial.
There are two major, seemingly insoluble, problems in the world today. The first
is the millennia old problem of population growth leading to organisations of people
creating dominating social systems. The second is the “spaceship earth” problem,
which is a major consequence of the first. Millennia ago, small consequences did not
The first problem developed as barbarism, tyranny, empires, monarchs, religious
zealots, and, more recently and more damaging, the ruling ideologies based on Socialism/
Communism and Capitalism/Democracy, all even more reliant on the tools
of power, gained dominance. Earlier, power was based on the ability to organize, in
many cases aided by the marshaling forces of religious doctrines. The major driving
forces resulted from the growth of populations leading to the take over of any fertile
lands they could reach and hold to promote their prosperity. This evolved as the
science-based creation of weaponry developed to sustain the ideologies euphemistically
referred to as “modern civilizations.” The development of the “spaceship earth”
problem, dealing with human physical emissions, from both the developed and
developing populations, adds greatly to the pressure on civilization to reform. World
leaders appear to recognize that the consequence of present expanding energy demands,
and particularly the seemingly mindless conversion of the world’s organic
feedstock into carbon dioxide, even if much of it could be captured by plants and
microorganisms to produce carbohydrates, is detrimental to the global environment
(and, not incidentally, to the organic chemical industry of future centuries).
The time for implementing vigorous global solutions is on us. The consequences of
power-imposed dominance and associated energy development have been neglected.
In reality, governments of the Western world are only a political mindset away from
overcoming the neglect of consequences. In the energy case, dominating narrow
economic arguments (although nowtempered by environmental issues) have impeded
efforts to aggressively enhance investment in science/engineering to effect the wider
use of solar, wind, wave, tide, hydroelectric (from glacial melting?), geothermal, and
nuclear sources of energy (given greater efforts to deal with nuclear wastes). And
although the narrow applications of science/engineering have enabled the powerful
to create their dominations, the powerful are now recognizing that scientists and
engineers need to be given a far more important role in determining the future of the
planet—better brains beget better behavior.
The Pharmaceutical Industry
Greater understanding of the chemistry and functions of the brain and its central
nervous system is perhaps the most challenging field of endeavor for scientists and
engineers in the 21st century. It is a field of endeavor involving the major disciplines
already integrated into pharmaceutical industry research and also into some university
At the genetic level, vigorous work programs are already in place (for instance,
on diseases such as Alzheimer’s) to find and map and thus gain knowledge of all the
genes involved in a disease with the expectation that this will provide a fundamental
basis for drug design.
At the molecular level the chemical sciences are already involved in the mechanistic
aspects and particularly in the analysis of brain processes in collaboration with
many other disciplines. And although the chemical process development field has no
direct link with chemical processes in the brain, except through the adaptation and
application of those exquisite analytical techniques, such as NMR and MS, so vital to
understanding chemical processes, the link is enough to provoke fantasies addressing
one of the major brain problems the world faces, namely how to deal with illegal
The developed, science-based nations spend billions of dollars worldwide to try
to eliminate their addicted populations’ use of narcotics, hallucinogens, hypnotics,
“controlled” analgesics, stimulants, and sedatives.108 The education, monitoring, and
rehabilitation of drug users has been going on for centuries and will undoubtedly grow
with increased investment in education generally. On the supply side, drug dealers
108Plant-based ones include Morphine ex Papaver somniferum, Cocaine ex Erythroxylon coca, Hashish
ex cannabis sative, Mescaline ex Lophophora Williamsii (source of peyote).
and transporters will continue to be pursued. Growers in Asia and South America
will continue to live with ongoing eradication of their crops. Efforts will continue to
provide incentives for growers to grow legitimate crops.
The ethical pharmaceutical industry will continue to look for profit in developing
pharmacotherapies against addiction. The use of nature’s basic plant structures as a
route into legitimate ethical drugs has yielded a few useful commercial products in
the case of the morphine alkaloids.109 There have so far been no major successes
along the line of the series of drugs produced from plant-derived steroid raw material
where an astonishing variety of biological activities has been created.110 This area
may be worthy of further investment but may be limited to new approaches to
creating anti-depression or other mood-altering and cognition–altering drugs. The
ability of addictive, narcotic biological drugs to cross the blood–brain barrier might
be harnessed by providing a handle for delivering, for example, anti-cancer drugs to
the brain, though their usefulness may also be compromised if the new combination
drugs lock onto brain receptors at their original principal locking sites.
The gravity and immense size of the world’s drug addiction problem continues to
attract enormous investment in the whole field of “brain chemistry,” especially in the
pharmaceutical industry, in universities, and by governments. Very large programs of
research and development are going on covering the whole field of addiction, such
as to alcohol, smoking, and illegal drugs.111 These programs only maintain a status
quo—larger programs might progress solutions.
The difficulties in treating addiction in general are compounded by the addict’s
willful lust for a short-term “fix,” leading to difficulties in ensuring treatment compliance,
and the all-too-frequent relapse and drop-out rates.
Currently, there are no options with guaranteed success for eliminating drug addiction,
only hope and the determination to create a “universal cure” which will, no
doubt, eventually happen, just as the world’s accessible oil resource will, eventually,
run out.112 In the meantime, one can only offer partial solutions to slowing down
the “drug trade.” This is happening on several fronts already, with restrictions on the
purchase and transport of certain chemicals used in making chemical drugs. For example,
intermediates used for manufacturing the hazardous methamphetamine, such
as pseudoephedrine (which illegal manufacturers extract from decongestion medications)
and methyl phenyl ketone (still available, with restrictions, from one supply
house), are being controlled, and in the first case replaced by other decongestants,
where possible. There is also a case for governments to direct some of the massive
and growing funding for combating drugs into selected existing programs such as
109A number of narcotic antagonists based on the morphinan structure have been marketed—for example,
Buprenorphine, Naloxone, Naltrexone, and Nalorfine. Nalmefene is being pursued for the treatment of
alcohol abuse. Oxycodone, and its precursor Codeine, are marketed, with restrictions, as analgesics.
110See Chapter 9, Table 3.
111For a good overview of these efforts see Thayer A. Chemical and Engineering News, 2006, September
25, pages 21–44.
112One draconian analogy lies in Aldous Huxley’s Brave New World: His imagined society promoted
immunization, at birth, of those who would serve in areas of the world where particular diseases are
prevalent! The complexity of gene functions in the brain may preclude altering genes, though it may be
possible to permanently alter particular cell types to affect gene activity.
some of those outlined in the Thayer overview (see footnote 111). A few of these
programs might be evaluated and funded through oversight by the National Institutes
of Health (NIH), say, initially, along the lines of support for orphan drug programs.
Other ideas should be solicited for review by panels of experts in the diverse fields
which one might expect to apply. I will offer a couple “starter” thoughts for the
botanically based illegal drug field—one for the chemical/pharmaceutical industry
and one for the biotechnology industry, specifically the seed industry.
The chemistry starter has two components. The first would ask for expert reviews
to assess whether the illegal botanicals (see footnote 108) might conceivably be
structurally manipulated to provide new activities, different from their controlled
substance role, along the lines of the successes realized in the steroid field (see
footnote 110). The second, which could be related to the first by providing new
structures, would involve a search for chemical reactions (e.g., ozonolysis) that would
radically alter the structure of the controlled substance, making it useful for a variety
of other uses—say, as resolving agents or chiral induction agents for chiral synthesis,
or as intermediates for agricultural applications (pesticides, herbicides, etc.), specialty
polymers, antioxidants, chelating agents, and so on.
The chemical drugs, such as methamphetamine, present somewhat more of a
problem since successes in dealing with the botanical illegal drugs seems likely to
drive the addict to switch to the chemical ones. It may take many years of relentless
education and constant public relations work, as in the case of tobacco addiction,
to enable addicts to recognize, and react to, the severe life and health impacts of
chemical drugs—sadly it seems unlikely that we will reform them all.
Finally, going back to the botanical illegal drugs, one has mostly to recognize that
the principal drugs, the morphine alkaloids, cocaine, and to some extent marijuana,
aremostly produced by people in alien cultures, where values and traditions differ and
where the illegal drug industry is often a major contributor to the country’s economy.
It will always be difficult to persuade the farmers in these poor countries to switch
to other crops when the drug lords offer so much more to the farmers for growing
illegal crops. It appears that even the security gained through a lower income from a
legal crop is not enough to offset losses in some years from government eradication
An action that might have a better chance of encouraging farmers to switch would
be to offer a higher price for a legal crop coupled with a threat to implement an
eradication program that would, forever, kill the illegal crop’s ability to produce
the offending addictive substance—much as the smallpox virus has been wiped out
worldwide—if the farmer does not accept the offer.
Others working in the botanical world must have made the same suggestion years
ago, from which it must be concluded that altering an offending plants genetics,
cellular or enzyme systems to achieve such an objective must be difficult.
Nevertheless, in view of the gravity of the illegal drug problem, it may be worth
re-appraising the idea from the platform of modern science. Take the poppy plant,
Papaver somniferum, as an example. The plant produces opium, which is the air-dried
milky exudate collected from incised unripe poppy-seed capsules. Opium contains a
mixture of about 20 alkaloids which amount to ca 25% by weight of the opium. Of
these 20 alkaloids the main ones are:
Morphine ~10–16% Noscapine ~ca 4–8%
Papaverine ~0.5–2.5%
Codeine ~0.8–2.5%
Thebaine ~0.5–2%
The amounts of these alkaloids in opium vary greatly with growing conditions and
regions. Of the above, only noscapine and papaverine have no narcotic properties. Papaverine
has found some medicinal use as an antispasmodic and cerebral vasodilator,
and noscapine found use a few decades ago in antitussive preparations.
No doubt plant breeding programs, even in countries such as Afghanistan, by
selection of seeds from high morphinan-producing plants, have enabled growers to
significantly increase their production of the narcotic components. The questions thus
arise, Can the biosynthesis of the morphinans be disrupted by interfering with the
biochemical synthesis steps, and could seeds be generated with the property of guiding
the biosynthesis of morphinans into useless products? The ultimate questions then
become, Could seeds so derived be grown in massive quantities for aerial distribution
over the growing fields of countries responsible for the heroin (diacetate ofmorphine)
trade? Then, would such seeds provide plants and flowers to dilute or destroy the
morphine-producing capability of plants already growing?
The taskmay not be such an easy one, although developments by commercial agricultural
and seed companies like Monsanto suggest that, given the “seed money” to
fund such an endeavor, seeds carrying the desired characteristicsmight be engineered.
Monsanto’s success with genetically modified soybean and tomato plants, to name
just two, would augur well for such a venture. In addition, continuing with the poppy
example, the known mechanism by which poppies biosynthesize the morphinans
suggests several possibilities for inhibiting narcotic formation.
There would appear, from Scheme 4, to be many opportunities to modify the
genetic material that creates the enzymes responsible for the biosynthesis of the
morphinans. Blocking the specific methylation of norlaudanosine to (?) retuciline
would be one worthwhile target, perhaps coupled with promoting the aromatization
and methylation sequences leading from norlaudanosine to papaverine.
Even given that the formation of (?) reticuline cannot be blocked, there may
be opportunities to block the biochemical sequences leading to thebaine, or even to
splice in mechanisms leading to nonaddictive alkaloids produced from (?) reticuline
by other plants—for example, berberine ex Hydrastis canadensis or protopine ex
Fumaria officinalis.
(-) - Reticuline
Tyrosine Hydroxylation
(?) Reticuline
quinone methide
+2H Reduction of CO
group of quinone
1. Aromatization
2. Methylation
Me N
HO Thebaine Codeine
HO Morphine
Cyclization to
SCHEME 4. The biosynthesis of the morphinans, papaverine and noscapine.
Aldous Huxley’s proactive solution apart, it may ultimately be recognized that the
addictive behaviors of some people cannot be dealtwith by any of the ingenious means
that can be imagined today. In the end, some horrific alternative to chemical pleasure
stimulation might need to be supported, perhaps a science-fiction-type implantation
of such as electrodes into addicted brains that can be coupled to stimulating inputs
supervised in some government-organized pleasure center— equivalent to Aldous
Huxley “Brave NewWorld” feelies. Such “warehousing,” coupled with an education
process, may enable addicts to also take on a function in life which is satisfying to
them and, preferably, useful.
I can conclude this fantasy by paraphrasing Einstein’s statement at the beginning
of this chapter: The problem of satisfying the pleasure requirements of the brain will
only be resolved when we have gained the understandings needed to invent processes
for dealing with them.
The bottom line, and the finale for this book, is that the world needs to fund
more interdisciplinary science, at the same time as figuring out how to deal with the
perceived consequences of all we have done and all we invent.
Abraham, Edward, 205
Abraham, Roy, 14
Abramovich, R. A., 361
Adamson, J. R., 210, 373
Adrio, J. L., 213
Afonso, Adrian, 25, 216, 218–219
Aki, O., 222, 370
Alburn, H. E., 239
Alcaide, Gina, 29
Ali, H., 358
Alimarante, L., 119
Allen, G. R. Jr., 244, 246
Allen, W. S., 244
Anderson, N. G., 1
Anderson, R. J., 323
Andrews, David, 26, 256
Andrews, S. L., 208
Angel, D., 19
Antosz, F. J., 255
Archibald, T. G., 254
Arigoni, D., 230
Arnold, B. H., 207
Arnstein, H. R. V., 204
Arth, G. E., 245
Asahi, Y., 222, 370
Askew, F. A., 234
August, Howard, 320
Avallone, Henry, 109
Babad, E., 148
Bader, R., 339
Baghurst, D. R., 358
Baig, G. U., 296
Bailey, E. J., 247, 249
Bain, B., 247
Baizer, M. M., 367, 374
Bakshi, R. K., 292
Balcom, Lewis, 29
Balderisa, L., 358
Baldwin, J. E., 205
Ban, Y., 369
Banik, B. K., 363
Barakat, K. J., 363
Barl, M., 368
Barnard, J. C., 254
Barrett, Stephen, 30
Bartlett, Paul A., 26, 31–32
Bartner, P. L., 363

Copyright C 2008 John Wiley & Sons, Inc.
Barton, Derek H. R. (Sir), 26, 31–32, 47,
222, 254
Bauer, B. E., 261–262
Baum, T., 269
Baxendale, I. R., 357
Bayer, E., 357
Beaton, J. M., 256
Beck, M. T., 338
Becker, W., 292
Beckman, E. J., 347
Beecham, 152–153
Beiner, Don, 29
Belsky, K. A., 28, 148
Ben Alloum, A., 361
Bendell, D. J., 323
Bendich, A., 308
Berenkopf, J., 247–248
Berge, S. M., 117
Bergman, E., 231
Bernasconi, E., 222, 224, 358, 369
Bernstein, J., 120
Bernstein, S., 244, 246, 261
Berry, D. M., 222, 370
Best, Arthur, 11–12, 37, 210–211
Bighley, L. D., 117
Birch, A. J., 235, 238
Blair, Tony, 319
Blank, R. H., 244
Blaser, H. U., 339
Blickenstaff, R. T., 240
Bodor, N. S., 248
Bomstein, J., 13
Bonaparte, Napoleon, 197
Boothroyd, Brian, 205
Borella, R. A., 247
Borissenko, N., 349
Borman, A., 242, 244
Bornscheuer, U. T., 341
Bose, A. K., 363
Bottomley, G., 18
Bourdillon, R. B., 234
Bowers, A., 245
Bradshaw, J., 118
Brady, Brian, 29–30, 46, 216
Braetoff, Eugene, 254
Bram, G., 362
Brattsand, R. L., 246
Bray, T. L., 358
Bream, R. N., 357
Brennan, Tom, 29
Brennecke, J., 346
Breslow, Ronald, 26, 31–32, 252, 338, 341
Breuer, M. E., 256
Brooks, T. J., 173
Brown, A. G., 208, 218
Brown, J. J., 246
Bruce, H. M., 234
Bruggink, A., 341
Brundidge, S., 20
Bruning, W., 357
Buchanan, J. M., 299
Buckley, Joe, 28, 70
Buckton, J. B., 247
Bulman, A., 361
Burgstahler, A., 230
Burke, M., 106
Buser, H.-P., 339
Butenandt, A., 231
Butler, D. N., 298
Buzby, G. C., 237, 239
Bywood, Roy, 17–18, 119, 155, 210–211,
225, 357, 373
Caddick, S., 358
Cai, Y., 105
Callow, R. K., 234
Cambiaghi, S., 102, 207
Camp, Howard, 28
Canselier, J.-P., 348
Cappelleti, L., 20
Carruthers, Nick, 26, 256, 278
Cerami, Joe, 28
Chackalamannil, Sam, 219
Chai, D., 224, 369
Chakraborti, A. K., 361
Chambers, Don, 28
Chan, T. H., 338
Chan, T. M., 299
Chang, W., 269
Chapman, P. H., 18, 146, 353–354
Charney, William, 242, 293
Chen, Guodong, 303–304
Chen, S.-T., 356
Chiappe, C., 351
Chiarino, D., 157, 160
Chiba, T., 373
Chiu, John, 28, 93, 227
Choi, Y. M., 146–147
Christensen, H., 10
Claeson, K. G., 246
Clark, J. C., 247
Clark, John, 28, 161–162
Clark, M., 18
Clayton, R. B., 230
Clegg, Brian, 17, 43
Coakley, Dan, 29
Cobban, R., 105
Cobler, H., 231
Coers, K., 339
Cohen, M., 269
Coleman, Ed, 29
Collet, A., 156
Colon, Cesar, 28, 93, 227, 263
Colton, F. B., 237
Conder, M. J., 213
Confroy, Jim, 24, 29
Coniglio, C. T., 245
Cook, M. C., 118
Cooper, Elvie, 29
Coppi, G., 119
Corey, E. J., 292
Corfield, R., 343
Coriolis, Gaspar Gustave de, 197
Crawford, L., 213
Cron, M. J., 20, 151
Cupps, T., 140
Cunningham, V. L., 107
Czerwinski, John, 29
Daggy, B. P., 107
Dale, J. K., 119
Danishefsky, S. J., 333–355
Danly, D. E., 368
Darwin, Charles, 377
Dealtry, C., 18
Dean, C., 321
DeAngelis, Nick, 28
DeCarneri, I., 119
de Diego, T., 349
Degner, D., 368–369
de Koning, J. J., 212
Demain, A. L., 213
Deshayes, S., 358
Desi, M., 357
DeSimone, J., 346
DeVelde, Ron, 29, 278
Devraj, R. M., 357
DeVroom, E., 212
Diamond, Jared, 317
DiBenedetto, D., 76
Dick, F., 354
Dietrich, George, 29
DiMasi, J. A., 268
Dinan, Noel, 28, 193–194
Dirnfeld, Liz, 28
Dirscherl, W., 235
DiSalvio, Al, 28
Djerassi, C., 235–236, 245
Djuhadi, Vince, 28
Dodds, David, 28
Doherty, A. E., 247
Dolle, R. E., 357
Dong, Jenny, 29
Doran, Henry, 29–30, 216
Dorwald, F. Z., 352
Douglas, G. H., 237, 239
Doyle, F. P., 118
Draper, Richard, 26, 216, 254, 295
Drauz, K., 341
Drucker, Peter F., 329
Duncan, S. N., 358
Dunitz, J. D., 120
Duran-Capece, Alicia, 28
Dwars, T., 339
Eastlick, D. T., 18, 210, 373
Eckstein, H., 357
Edgren, R. A., 237, 239
Ege, G., 295
Ehrreich, S., 269
Einstein, A., 317
Eisler, M., 245
Elander, R., 20
Elks, J., 18, 210, 246–247, 249, 357
Elving, P. J., 308
Emanuel, C., 350
Emblidge, R. W., 147
Emery, Bob, 28–29, 70
Endres, F., 349
Engel, J., 93, 156
Engel, M. H., 366
English, A. F., 247
Eppstein, S. H., 240
Epton, R., 354
Eschenmoser, A., 230
Escobar, Miguel, 31, 46, 254, 257–259
Etheredge, Sarah Jane, 334
Eulitz, K. D., 111
Farthing, J. E., 207
Fazakerley, H., 247
Fee, Bill, 29
Fehrentz, J.-A., 354
Feldman, L. I., 244
Fernandez, X., 247
Fernekess, H., 210
Fiers, Al, 29
Fieser, L. F., 241
Fieser, M., 241
Fildes, Robert A., 12–14, 38, 204, 207, 213
Fisher, J., 237
Fitzgerald, Maurice, 29–30, 216, 295
Flatt, P., 339
Flynn, E. H., 208
Foell, T., 237, 239
Fox, J. J., 308
Frederiksen, E. K., 118
Freund, H., 292
Fried, J., 242, 244–245
Friedman, S., 299
Fritschel, B., 140
Frost, J. W., 343
Frost, R., 49
Fu, X., 21, 25, 27, 227, 263–264
Fukatsu, S., 348
Gadamasetti, K., 1
Gadsby, B., 237, 239
Gala, Dinesh, 27, 76
Galbraith, John Kenneth, 320
Gallagher, G., 119, 155, 210–211, 225, 373
Ganboa, I., 361
Ganguly, Ashit, 25–26, 47, 216, 218–219
Gao, Y., 94
Gassner, U., 147
Gebert, W., 245
Gedye, R. N., 358
Genders, David, 222, 224, 369, 371
Gentles, M A., 247–248, 254
Gerber, R., 147
Ghosh, A. C., 240
Gibson, N. W., 296
Giguere, R. J., 358
Gil, L., 31
Gilbert, D. A., 207
Gilbert, K. H., 295
Girijavallabhan, Viyyor, 25, 216, 218
Giusto, Bob, 28
Gloor, Guy, 29
Gmouh, S., 349
Godtfredsen, W. O., 118
Goffinet, B., 246
Gold, E. H., 269, 308
Gorman, M., 208
Gosteli, J., 95, 153, 154, 209, 308, 354
Gots, J. S., 299
Goulden, Steve, 204
Graber, R. P., 244
Grabowich, B., 244
Gramberg, C. W., 354
Green, D. W., 165
Green, M. J., 25, 247
Green, Michael, 28, 93
Greenspan, G., 239
Gregory, Gordon, 18, 118, 208
Gregson, D., 105
Gregson, M., 118
Grieco, P. A., 338
Griskey, R. G., 1, 165
Gross, Michael, 303
Groundwater, P. W., 323
Gullo, V., 25
Gusev, A. J., 153
Hagele, K., 357
Hagenmaier, H., 357
Hall, D. A., 222–223, 370
Hall, N., 101
Hancu, D., 347
Hannah, R. E., 107
Hanreich, R., 339
Hanze, A. R., 245
Hardcastte, G., 20
Harlan, R. F., 254
Hart, H., 24
Hartley, D., 237, 239, 270, 273
Heller, M., 244
Hems, B. A., 18, 210, 211, 213, 249
Hench, P. H., 234, 240
Herbst, D., 237, 239
Herczeg, Lou, 27, 148
Hershberg, E. B., 236, 242, 245, 254
Hertz, R., 235
Herz, J. E., 242
Herzog, Hershel L., 242, 245, 254, 258
Heusler, K., 153, 154, 209
Hickman, J. A., 296
Hileman, B., 106
Hill, Henry, 29
Hiiragi, M., 273
Hirao, A., 291
Hirayama, Y., 212
Hitzler, M. G., 346
Hobbelen, P. M., 237
Hockenhull, Don, 204
Hoff, D. R., 245
Hoffmann, W., 368
Hoffsommer, R. D., 245
Hogg, J. A., 245
Holligan, J. R., 18, 146
Hong, Y., 94
Hood, Bill, 29
Horner, L., 210
Horsley, L. H., 174
Hou, Don, 26
Howard, Michael, 319–320
Howard, Phillip K., 327
Huang, D.-S., 254
Huber, J. E., 256
Hughes, M. T., 254
Hughes, Gordon, 237–239, 254
Hulteen, J., 373
Hultquist, Martin, 16–17, 41
Hungerb?uhler, E., 298
H?urzeler-M?uller, M., 308
Hurley, Aldous, 382, 387
Iborra, J. L., 349
Ing, Y., 363–365
Inhoffen, H. H., 236
Inoue, H., 273
Isaacs, N. S., 291
Ishi, K., 340
Ishikawa, N., 74–76, 161
Ishimaru, T., 146
Ito, K., 291
Ito, N., 348
Itsuno, S., 291–292
Iwagiri, H., 161
Jack, D. (Sir), 210, 249
Jackson, B. G., 208
Jackson, P. S., 357
Jacobs, W. A., 232
Jacobsen, Oscar, 9–11, 36
Jacquez, J., 156
Jallett, H.-P., 339
Jansen, A. B. A., 237, 239
Jaret, Bob, 10, 27, 216
Jeger, O., 230
Jelsch, E., 339
Jenkins, John, 26
Jevnik, M. A., 242
John of Salisbury, 371
Johnson, David, 14, 39, 153, 173
Johnson, F., 149–150
Johnson, G., 269
Johnson, Glenn, 19
Johnson, Samuel, 165
Johnston, D. B. R., 245
Johnston, Paul, 28
Johnston, J. B., 107
Jolly, R., 246
Jommi, G., 157, 160
Jones, E. R. H., 32
Jones, I. F., 370
Jones, R. E., 244
Joo, F., 338
Jost, Kurt, 216
Joyner, C. C., 254
Junio, John, 29
Kaluza, Z., 363
Kametani, T., 273
Kasper, E., 247
Kaur, G., 361
Kawano, S., 95
Kazlauskas, R. J., 341
Keasling, J., 343
Keil, J. G., 151
Kelly, R. M., 109
Kendall, E. C., 234
Kennedy, J. F. K., 49
Kennelley, E. L., 105
Kerridge, Ken, 163–164
Kessler, W. B., 244
Kieboom, T., 341
Kigasawa, K., 273
Kimmel, C. A., 66
Kimmerle, E., 164
King, C. J. H., 368
Kirilin, A. D., 153
Kleeman, A., 93, 156
Knight, J. C., 255
Kobayashi, E., 338–339
Kodama, Y., 146
Koga, K., 273
Kohagizawa, T., 273
Kohno, Y., 212
Komaki, T., 95
Komar, K. M., 105
Kominek, L. A., 255
Konig, W. A., 357
Kono, S., 95
Kooreman, H. J., 212
Kraus, J., 235
Kreh, R. P., 368
Kucharczyk, N., 147
Kugelman, M., 76
Kuo, Shen-chun, 26, 312–313
Kutscher, B., 93, 156
Kwasny, Rick, 28
Kwok, D.-I., 148
Laali, K. K., 349
Laberge, L., 358
Lagonikos, Perry, 28
Laing, S. B., 247
Laird, T., 164
Lalaoui, M., 361
LaMonica, F., 13
LaMorte, Janet, 29
Lane-Allman, E., 247
Lassalle, Yvonne, 30
Laundon, B., 18
Lavagnino, E. R., 208
Leach, A. G., 357
Le Bourdonnec, B., 357
Ledig, K., 237, 239
Lee, E., 224, 358
Lee, Junning, 21, 25, 27, 45, 93, 222–223,
263, 358, 369
Lee, S., 1
Leemhuis, J., 237
Leigh, H. M., 240
Leitner, W., 346
Lenhard, R. H., 244, 261
Leopold, E. J., 230
Lewis, R. J., 22, 66, 217
Ley, S. V., 357
Li, C. J., 338
Liagre, M., 358
Lide, D. R., 67, 82, 93
Lin, J. S., 20, 151
Littell, R., 244, 246
Liu, Y.-H., 363
Llaurador, Kenny, 31
Longbottom, D. A., 357
Longoni, D., 222, 369
Lord, K. E., 230
Loupy, A., 358, 361–362
Love, George, 20, 24–25, 45
Lowe, D., 20
Lozano, P., 349
Lubineau, A., 338
Luciano, Robert, 14, 20
Luche, J.-L., 358
Lund, H., 367, 374
Lundquist, J. T., 368
Lunt, E., 296
Lutsky, B., 247–248
Lyttle, D. A., 240
Ma, L. Q., 105
Machiavelli, Niccolo dei, 213
Macek, T. J., 119
Macko, S. A., 366
Magatti, C. V., 308
Magerlein, B. J., 245
Majdoub, M., 362
Majetich, G., 358
Maloney, Donald, 26, 254
Maloney, J. O., 165
Manabe, K., 339
Mancera, O., 236
Manhas, M. S., 363
March, J., 361
Marino, Jay, 38, 70
Marker, Russell, 232–233, 235
Markoff, J., 321
Marson, S. A., 230
Martin, Charles R., 222, 224, 369, 371–373
Martinez, J., 354
Martlew, E., 18, 356
Marx, M., 244
Mas, Jan, 28
Masuda, K., 222
Mathur, C., 363
Matsuura, P., 160
Mavroudakis, W., 140
May, P. J., 246–247
McAda, P. C., 213
McAllister, Tim, 28, 93, 227
McCombie, Stuart, 25, 216, 218
McConville, F. X., 1, 166
McCoy, E., 343
McGonigle, Gene, 28
McKenzie, A. W., 249
McLoughlin, B. J., 237, 239
McMenamin, J., 237, 239
McNamara, Paul, 25
Meinwald, J., 32
Meister, P. D., 240
Mendez, Paulo, 214
Menon, Vinod, 222, 224, 369
Mergelsberg, I., 24, 76, 295
Mergler, M., 354
Merrifield, R. B., 352
Meyer, Tony, 29
Miley, Michael, 216
Miller, Alan, 28
Miller, Charles, D., 8
Miller, R., 149–150
Miller, R. W., 299
Mills, J. S., 245
Mingos, D. M. P., 358
Miramontes, L., 235
Mironov, V., 153
Mirza, U. A., 363
Mitchell, M., 140
Monahan, M., 247
Monkhouse, D. C., 117
Monroe, Michael, 29
Morales, G. A., 357
Morales, Gilda, 31
Morgan, Brian, 28
Moriarty, K. J., 357
Morimoto, A., 222, 370
Morin, R. B., 208
Morris, D., 204
Moss, R., 24
Mueller, R. A., 208
Muir, John, 320
Mukherji, S. M., 235
Mulholland, S., 15
Mullins, J. D., 119
Murakami, M., 207
Murao, K., 340
Murawski, Rich, 30
Murphy, Bruce, 28, 161–162
Murray, H. C., 240, 245
Naegeli, P., 153, 154, 209
Nagabhushan, T. L., 156
Nagai, H., 373
Nakabayashi, S., 348
Nakahama, S., 291
Nayler, J. H. C., 118
Nelson, N. A., 235
Nemorin, J., 261
Neumann, D. A., 66
Neumann, F., 236
Neustadt, B. R., 308
Newlands, E. S., 296
Newton, C. G., 296
Nicholls, H., 106
Nine, John, 29, 371
Ning, Jian, 28
Nishata, K., 348
Nobile, A., 242
Nussbaum, A. L., 245
Nyfeler, R., 354
O’Brien, Conor, 30
Ochiai, M., 222–223, 370
O’Connor, S. E., 343
Oehme, G., 339
Okada, T., 222, 370
Okimoto, M., 373
Oliveto, E. P., 245, 254, 258
Oppolzer, W., 153, 154, 209
Ottiger, Phil, 216
Pachter, I., 13
Padilla, A. G., 256
Page, M. I., 205
Page, J. E., 370
Pagliarin, R., 157, 160
Palomo, C., 361
Parker, R. E., 291
Parlow, J. J., 357
Parr, W., 357
Partridge, K., 206
Pasteur, Louis, 350
Patterson, Jim, 18
Pattison, T. W., 237, 239
Patzek, T. W., 338
Payne, C. C., 242
Pearson, D., 105
Pechet, M. M., 245
Peer, Lydia, 29, 216
Perlman, D., 242
Perlman, P. L., 242, 245
Perry, R. H., 165
Peterson, C., 364
Peterson, Durey H., 240
Petit, A., 358, 362
Pfeil, E., 292
Phillipps, G. H., 247
Phillips, P., 105
Phillips, P. C., 237, 239
Phillipson, R., 247
Philpot, J. St. L., 234
Piete, Elaine, 29
Pigeon, P., 361
Pimentel, D., 338
Pinto, P., 218
Piwinski, J., 25
Pletcher, D., 367
Poliakoff, M., 346
Polley, H. F., 234
Popper, T. L., 248
Potts, J. R., 207
Power, F. B., 232
Pramanik, Birendra, 27, 46, 298, 300,
303–304, 363–364
Pratton, M. H., 254
Prenntzell, W., 147
Prioli, N., 269
Proctor, L. D., 254
Pugin, B., 339
Pyke, T. R., 255
Quinones, Iliana, 30–31
Radimerski, P. X. Y., 339
Ralph, R. K., 298
Ramage, R., 153, 154, 209
Rambosek, J. A., 213
Ramdani, M., 361
Ranganathan, S., 153, 154, 209
Rashid, Khalif, 29
Rausser, Richard, 28, 245
Read, A., 203
Reerink, E. H., 234
Rees, R., 237, 239
Reeves, C. D., 212
Regenye, Al, 29
Reichert, D., 93, 156
Reichstein, Tadeus, 240
Reif, Van, 28
Reineke, L. M., 240
Repic, O., 1
Rhodes, C. T., 370
Richardson, E. J., 173
Rideout, C. J., 338
Ridge, D., 140
Rief, Van, 28
Ringler, I., 246
Ringold, H. J., 236, 245–246
Rios, Luis, 31
Rivkin, S. M., 211
Rizvi, R., 218
Roberts, Fred, 28
Roberts, S. M., 208, 218
Robertson, D. N., 10
Robinson, C. H., 18, 212, 230, 357
Robinson, G., 1
Robinson, J. D., 247
Robinson, R. (Sir), 204
Rodionov, E. S., 153
Roh, E. J., 349
Rohrmann, E., 232, 235
Roletto, J., 222, 224, 369
Rombie, J. M., 173
Rosenhouse, Stan, 29, 216
Rosenkranz, George, 233, 235–236, 246
Ross, S. K., 346
Rouhi, A. M., 105
Roussell, J., 358
Rousche, N., 20
Rubin, M., 236
Ruffo, Pete, 29
Ruby, J., 20
Ruggeri, Mario, 20, 28, 43, 151, 216
Russell, P. J., 308
Ruzicka, L., 230
Sabo, E. F., 242, 244
Saijo, S., 212
Sakai, T., 212
Sakamoto, D., 212
Salvino, J. M., 357
Salway, A. H, 232
Samuel, Jeff, 29
Sanchez, Sergio, 31
Sandford, P. E., 247
Sandor, Paul, 28
Sapino, Chester, 19, 53
Sarca, V. D., 349
Sarett, L. H., 244–245
Savage, P. E., 339
Savle, P., 175
Scanlon, E. R., 208
Scherer, D., 76
Scherer, Uta, 309
Schmidt, E., 340
Schmidt, J., 231
Schneider, C. J., 222, 370
Schneider, H. D., 339
Schoevart, R., 341
Schofield, C., 205
Scholer, H. F. L., 234
Schroff, A. P., 237
Schumacher, Doris, 20, 24, 45, 161–162
Schwenk, E., 236
Schwenker, G., 147
Scott, J. S., 357
Sebek, O. K., 245
Seki, S., 348
Shapiro, D., 361
Shapiro, E., 25, 247–248
Sharapov, V. A., 153
Sharma, G. K., 119, 155, 210, 225,
Sharratt, P., 247
Shaw, G., 298
Sheludyakov, V. D., 153
Shephard, K. P., 256
Shibata, S., 292
Shih, K., 20
Shimadzu, H., 222
Shinozaki, K., 222, 370
Shipkova, Petia, 300, 303
Shiraishi, Y., 212
Shoppee, C. W., 261
Shue, H.-J., 247
Shutts, Bruce, 19, 20–21, 28, 44, 216, 295,
Siddall, J., 237, 239
Siegel, H., 368
Sih, C. J., 255
Silber, R. H., 245
Silverman, R. B., 341
Silvestri, H. H., 20, 153, 173
Simat, T. J., 111
Simonet, Dan, 29
Simpson, J. C. E., 232
Sinou, D., 339
Siuda, J., 237
Slates, H. L., 245
Slocumb, C. H., 234
Smail, F. R., 346
Smith, D. L., 308
Smith, E. M., 308
Smith, F. E., 358
Smith, H., 118
Smith, Herchel, 237–239
Smith, L. L., 237, 239
Smith, Tony, 30
Snodgrass Pilla, Caesar, 28
Sofia, R. D., 147
Sogli, L., 222, 224, 358, 369
Sondheimer, F., 235–236
Song, C. E., 349
South, M. S., 357
Sowa, T., 207
Spero, G. B., 245
Spindler, F., 339
Spotnitz, R. M., 368
Stables, H. C., 18, 357
Stacey, M., 32
Stapf, F., 308
Steckhan, E., 369
Steeples, I. P., 247
Steinhart, H., 111
Steinman, Martin, 20, 26, 44,
Stepan, A. M., 213
Stevens, M. F. G., 296, 312
Stiefel, F. J., 147
Stoerk, H. C., 245
Stolar, S. M., 244
Stone, R., 296
Stonehouse, R. C., 247
Stoodley, R., 32
Storer, R. I., 357
Stork, G., 230
Strack, Robert, 28
Strauss, C. A., 339
Sturm, F. J., 107
Subramaniam, B., 346
Sudhakar, Anantha, 27, 256
Suida, J., 239
Sutherland, J. D., 366
Suzuki, N., 207
Suzuki, Y., 299
Sybertz, E. J., 269
Sykes, R. B., 118
Szarek, W. A., 160
Tahbaz, Peter, 28, 216
Takaoka, A., 161
Takata, Y., 373
Tan, H.-S., 212
Tann, Chou H., 20, 21–23, 30, 42, 93, 148,
216, 227, 261–264
Tanner, M., 76
Taub, D., 245
Taylor, A. B., 76
Taylor, George, 18
Taylor, S. J., 357
Taylor, William L., 321
Thalen, B. A., 246
Thayer, A., 382
Thiruvengadam, T. K., 21, 22, 25, 27, 46,
93, 227, 263–264
Thoma, R. W., 242
Thompson, E., 18
Thompson, J. L., 245
Thomson, W. T., 312
Thuresson, B., 246
Tiberi, R. L., 247–248
Tishler, M., 244
Togni, A., 340
Tokolics, J., 237, 239
Tolksdorf, S., 242, 245
Tomaselli, S., 102, 207
Torii, S., 367
Tormos, W., 162
Torpey, Kathy, 29
Toto, Anthony, 29
Tsai, David, 26, 148, 254
Tsukamoto, T., 232
Tu, C., 105
Tullner, W. W., 235
Tully, M. E., 242
Turner, D. L., 232
Turner, K., 164
Uhlig, R., 103
Urben, P., 66
Utley, J., 369
Vagt, U., 350
van Bokhoven, C., 237
van den Broek, A. J., 237
Van Rheenen, V., 256
Van Tamelen, E. E., 230
Van Wickern, B., 111
Vatanen, A., 376
Vater, Gene, 28
Vaultier, M., 349
Verga, R., 102, 207
Verweij, J., 212
Villemin, D., 361
Vincent, J. E., 363
Vinci, V. A., 212
Virgilio, Abramo, 13
Visibelli, Ettore, 19, 43, 214–215
Vita, J., 13
Vogel, Ernst, 23–24, 45, 216
Von Linden, Dennis, 29
Vorbr?uggen, H., 153, 154, 209
Vorherr, T., 354
Voss, H. E., 235
Waddington, T. C., 347
Waespe, H. R., 230
Wagatsuma, N., 273
Wagle, D. R., 363
Waldman, H., 341
Walker, D., 10, 119, 148, 151, 154–155,
210–212, 222, 224–225, 353–354, 357,
358, 369, 373
Walker, Tom, 18, 210, 249
Walsh, F. C., 367
Wang, K.-T., 356
Wang, Y., 312
Warnant, J., 246
Warner, D., 19
Warr, A. J., 254
Warrener, R. N., 298
Wasserschied, P., 348
Watanabe, H., 95
Watnick, A., 248
Watson, D. H. R., 239
Watson, J. H. P., 237
Waugh, Richard, 9–11, 36
Waugh, Tom, 9–11, 36
Weber, L., 245
Weber, P. C., 363
Webster, T. A., 234
Wegman, A., 339
Weinberg, Norman, 222, 224, 369, 371
Weintraub, A., 240
Weir, N. G., 246–247
Weisburger, J. H., 217
Weisenborn, F. L., 255
Weiss, H. J., 244
Welton, T., 348
Wendler, N. L., 244–245
Wendt, G. H., 237, 239
Werner, Ray, 20, 23, 29, 44, 216, 295
Westaway, K. G., 358
Westerhof, P., 234
Westphal, U., 231
Wheeler, Lavonne, 29, 295
White, Al, 29
Whitfield, J. N., 366
Wilds, A. L., 235
Wilen, S., 156
Willett, J. D., 230
Williams, J. H., 261
Williams, W., 19, 20
Williamson, C., 247
Wilson, D. W., 299
Wilson, Ted, 17, 43, 76, 210, 357, 373
Winkelman, Al, 29
Winter, C. A., 245
Wolf, G. C., 240
Wolkoff, Hal, 14–15, 40, 57, 222, 371
Woodward, R. B., 153, 154, 204, 209,
Woolridge, K. R. H., 296
Wovcha, M. G., 255
Wu, George, 27, 161–162
Wyvratt, J., 140
Yamada, S., 273
Yamaguchi, Eichii, 225
Yamazaki, N., 291
Ye, Andy, 28
Yoshioka, Mitsuru, 222, 356
Yu, H.-M., 356
Yu, Steven, 20, 23, 28, 44, 216
Zaks, Alex, 28
Zappi, Guillermo, 107, 222, 224,
Zehnder, B., 298
Zein El Abadin, S., 349
Zelenay, Piotr, 222, 224, 369
Zepp, C. M., 94
Zhang, Haiyang, 222, 224,
Zhang, W., 105
Zimenoff, Steve, 29
Zondek, B., 231
Accelerating rate calorimeter (ARC):
nitro compound evaluation, 72–74
process safety and, 67–68
O-Acetyl-L-threonine, 218–219
Acid scavengers, ionic liquids and, 350–351
Acrylonitrile, electrochemical
hydrodimerization, 368
Active pharmaceutical ingredients (APIs):
assay and stability, 120–122
bureaucracy reduction and, 327–332
chemical development objectives
concerning, 53–56
chemical process development and, 3–4
Chemistry, Manufacturing, and Controls
(CMC) document requirements,
controlled environment for manufacture,
crystal form and particle size, 119–120
crystallization process, 124–126,
development report requirements,
dilevalol hydrochloride case study,
Good Manufacturing Practices and,
last process steps for, 122–126
organic process waste disposal, 100–103
outsourcing trends in, 335–337
particle size engineering, 190–195
patent considerations, 141–142
pilot plant production of, 186–190
quality control system for, 113–114
quality specification and last process
steps, 115–126
regulatory issues, 109–113
scale-up process, 167–169
structure selection, 116–118
sustainable discovery and development,
synthesis process selection, 13–14,
127–131, 268–294
technology transfer and, 135–137
therapeutic team development of,

Copyright C 2008 John Wiley & Sons, Inc.
Active pharmaceutical (cont.)
toxicology batch, 114–115
validation process, 133, 137–139
wastewater treatment, 103–107
workplace safety practices and, 84–86
ACV tripeptide, cephalosporin C
fermentation, 204–208
Adaptation, leadership skills and, 6–8
Addiction chemistry, future trends in,
Administrative procedures and SOP
manuals, 60
Absorption, Distribution, Metabolism, and
Excretion (ADME) studies, 54,
Adverse drug events (ADEs), regulatory
consequences, 111–113, 378
Agenda, for internal symposia, 58–61
Agitation equipment, 171
Air pollution, process emissions, 89, 92–96
Albuterol manufacture, process emissions
control and, 30, 93–96
fantasy, 382–387
opiates poppy, 383–387
steroidal, 228
Alkylation, microwave-assisted, 361
Amikacin case study, patent protection and,
carboxylate (7-ACA), 13, 19, 220,
patent protection in production of, 145–146
5-Aminoimidazole-4-carboxamide (AIC)
intermediate for temozolomide
manufacture, 295, 298–312
dihydrohypoxanthine hydrolysis,
preparation from hypoxanthine, 299–307
hypoxanthine reduction, 307–311
carboxylic acid
polymer synthesis, 353–355
quality control issues, 214–215
ring expansion research, 208–214
6-Aminopenicillanic acid (6-APA):
patent protection in production of,
penem synthesis and, 216–217
process patent case study involving,
quality control issues, 214–215
(AICAR), 299
Amorphous compounds, stability of,
Amoxicillin development, 14
process patents and, 152–154
Anabolic steroids, underground design of,
Analytical Research and Development
organization, 57
Androgens, natural sources, 229
Antibiotics, treatment in wastewater,
Anti-inflammatories commercialisation of,
API-hypochloride (API-HCl),
crystallization process, 178–181
“Applied common sense” principles,
Standard Operating Procedures and,
Arapahoe Chemicals, visionaries and
operating philosophy, 9–11
Artemesinin, preparation of, 341–343
Aspergillus family, dilevalol hydrochloride
case study, 293–294
Assays for quality control, active
pharmaceutical ingredients,
Audit of facilities, preparing for a
pre-approval inspection, 138–139
Auxoploses and auxochromes, 69–70
purpose and guidelines for, 32
safety award programs, 86
Bank security system, patent for, 164
Barbasco roots, status as a raw material
source, 258–260
Batrachotoxin A, toxicity of, 228
Beckmann rearrangement, oral
contraceptive production,
Beecham Amoxicillin process, patents case
study and, 152–154
Behavior-based safety programs,
characteristics of, 83–84
Benzophenone hydrazone, electrochemical
oxidation, 373–374
preparation of the alcohol, 260–264
intermediates and synthesis approaches
to, 255–264
Betnovate, structure identification, 249
Bile acids:
cholesterol degradation, 230–231
cortisone manufacture, 240–241
Bill of Rights, 326
Biological oxygen demand (BOD):
Clean Water Act provisions, 89
waste management, 199–200
wastewater treatment and, 103–107
Bioluminescence process, patents
involving, 162–163
Bioremediation techniques, wastewater
treatment, 105
Biotechnology Department, chemical
process development and, 57
Biotowers, wastewater treatment using, 105
Biotransformation department:
purpose of, 58
ring expansion research and, 213
Bisesterification, microwave-assisted
chemistry, 361
Boron intermediates, patent issues
surrounding, 147–150
Brain chemistry, a fantasy, 381–387
cephalosporin research and, 12–14
patents case study and, 152–154
scientists and engineers in, 18–20
N-Bromoamides, 10
Buchi-type filters, utility of, 181–182
Bureaucracy, chemical process development
and reduction of, 326–332
Business Development organization,
involvement of, 53
n-Butanol in dilevalol racemization and
recycle, 285
Calorimeters, description of types, 67–69
Capitalist economy, enhancing education in,
Carbon dioxide, supercritical, 345–347
Carbon oxygen demand (COD):
Clean Water Act provisions, 89
waste management, 199–200
wastewater treatment and, 103–107
Carboxyl protecting groups:
patent protection issues and, 145–146,
ring expansion and, 209–213
Catalytic hydrogenation, hypoxanthine
reduction to dihydrohypoxanthine,
in (S)-Metolaclor manufacture, 339–340
Catalytic oxidation, process emissions
control, 93
manufacturing process for, 225–226
Cefadroxil, 214
Ceforanide, 215
Cefotaxime, 215
Ceftibuten dihydrate, licensing and
development, 220–226
Ceftriaxone, 215
Cefuroxime axetil, 215
Center for Disease Control and Prevention
(CDC), 65–66
Centrifugal pumps, 195
basic equipment, 181–183
vapor explosions and, 81–82
active pharmaceutical ingredient
competing processes, 11–12
patent infringement issues, 154–155
polymer-supported synthesis, 353–356
quality control issues, 214–215
ring expansion research and, 210–214
cephalosporin C fermentation, 204–208
development of, 204–226
classical cephalosporins, 220–226
penem synthesis, 215–220
penicillin G/cephalosporin C
fermentation, 204–208
penicillin sulfoxide-cephalosporin ring
expansion, 208–211
semisynthetic compounds, product
quality, 214–215
electrochemistry and, 220–226,
Cephalosporins: (cont.)
esterification, 12, 118, 145–146, 152,
154, 209–212, 222–225, 348, 361,
extraction process, 171–174
extractive ecterification, 212, 369
penicillin sulfoxide ring expansion:
patent issue, 154–155
Cephapirin, 215
Cephradine, patent infringement by,
Chemical engineering:
agitation, 171
computer applications, 200–201
crystallization, working example,
distillation/evaporation, 174–177
education in, 321–325
extraction, 171–174
filtration, washing, and drying, 181–190
flow measurement, 196–198
heat exchange, 169–171
overview stages of process development,
pilot plant and plant maintenance,
particle size reduction, 190–195
process containment, 186–189, 282–284
process scale-up, 167–169
pumping systems, 195–196
reactor volume measurement, 198–199
waste management systems, 199–200
Chemical Engineering for Chemists, 165
Chemical explosion, chemical structure
classification, 69–70
Chemical libraries, polymer-supported
reagents, 357–358
Chemical process development:
bureaucracy reduction and, 326–332
computer technology and, 200–201
current Good Manufacturing Practices
for, 113–139
dilevalol hydrochloride case study,
education and, 318–325
evolution of, 166–167
fantasy on brain chemistry, 380–387
Florfenicol case study, 155–162
mission and structure, 53–62
objectives, 53–56
organization, problems, needs, structure,
functions and development, 49–63
overview, 1–4
patent protection and content, 141–143
public image of, 50– 51
sustainable chemistry and evolution
towards, 377–380
synthesis methodology for, 127–131
therapeutic teams for, 51–53
water and enzymes, 337–344
Chemical Safety and Hazard Investigation
Board (CSB), mission and duties, 66
Chemical safety organization, workplace
safety practices, 82–86
Chemicals handling, 96–103
Chemistry, Manufacturing, and Controls
(CMC) document:
bureaucracy reduction and, 327–332
FDA requirement for, 131–133
Chiral hydroxylation, safety issues with an
oxaziridine reagent, 75–76
Chloramphenicol stearate,
amorphous and crystalline forms of,
Chloroform, contaminant and carcinogenic
properties of, 217
Cholesterol, biosynthesis, 230–231
Cholesterol absorption inhibitors (CAIs),
Cholic acid, progesterone from, 231
Chromatographic purification, proposal of a
process for bureaucracy reduction,
Chromophores and plosophores, 69–70
Clean Air Act (CAA):
provisions of, 89–90
Clean Water Act (CWA), provisions of, 89
Cleavage reactions, polymer synthesis,
Clogging problems in micromise operation,
Collaboration, leadership and role of, 7–9
Commitment, leadership skills and role of, 8
leadership and role of, 7–9
organizational structure and, 58–60, 63
Competition, patent protection and,
Comprehensive Environmental Response,
Compensation and Liability Act
(CERCLA), overview of, 88
Computer technology:
chemical process development,
maintenance programs and, 201–202
Condition-based maintenance (CBM),
principles of, 201–202
Confetti patent, 162
Containment of processes, 186–189
Continuous-flow reactors, 171–173,
350–351, 359–366
Contraceptives, production of, 235–240
Controlled environment rooms (CERs),
requirements in, 186–190
Convergent synthesis, last process steps for
APIs, 122–124
Coriolis mass flowmeters, flow
measurement, 196–198
Cortisone, early research and improved
anti-inflammatories, 230, 240–249
Corynebacteria, cortisone research and
improved anti-inflammatories
Cost-of-goods (COG) projections:
dilevalol hydrochloride case study,
269–272, 275–279
process development and alternative
routes, 290–294
electrochemistry and, 220–225, 366–372
patent issues and, 146–150
Cross contamination, 21–22, 217
Crotonylation reaction, albuterol process
with lower emissions, 95–96
Crystalline structures effect on active
pharmaceutical ingredients,
Crystallization, basic principles of, 177–181
Cucurbitacins, medicinal properties,
Current Good Manufacturing Practice
bureaucracy reduction and, 327–332
chemical process development, 113–139
evolution of, 110–113
(AIC) preparation, 298
Dacarbazine (DTIC), 296
D-amino acid oxidase (DAAO),
cephalosporin research and, 12–13,
206–208, 220
DAST reagent, Florfenicol development
and, 157
DBTA salt formation:
dilevalol development and, 273–290
n-butanol racemization and recycling,
Deacetoxycephalosporin C synthase
(DAOCS), ring expansion process,
Deacetoxycephalosporin G (DAOG), ring
expansion of penicillin G to, 213
10-Deacetylbaccatin III, raw material for
taxol, 342–343
Dehydration reactions, steroid chemistry,
Dehydrogenation at C-1,2 to enhance
anti-inflammatory activity,
Delegation, as leadership skill, 6, 9
Desogestrel, synthesis of, 237, 239–240
Desoxymethyltestosterone, underground
production of, 251
Developing countries, outsourcing drug
development to, 335–337
Development organization, structure and
function, 56–61
Development reports:
components of, 133–135
pre-approval inspection and, 133
Dialogue, leadership and promotion of, 7–9,
Diazomethane, 16?-methyl intermediates
and, 252–260
Diels-Alder reaction:
electrochemistry and, 368
microwave-assisted chemistry, 360
water promoted, 338
Differential scanning calorimetry (DSC):
assessment of explosion potential,
process safety and, 67
Digoxin, medicinal use of, 228–229
hydrolysis to AIC, 311–312
hypoxanthine reduction to, 307–311
Dilevalol hydrochloride:
cardiovascular therapy applications, 269,
commercial process case study, 268–294
DBTA salt conversion to, 285–287
overview, 269–272
n-butanol racemization and recycling,
NDA process to commercial scale
development, 275–290
ongoing process development and
alternative routes, 290–294
process engineering for, 281–290
synthesis route research, 271–272
withdrawal from market, 294
Dimethyl-POPOP, development of, 10
norethindrone production from, 236
isolation of, 232
16?-methyl intermediates and,
Diphenyldiazomethane (DDM):
electrochemical route, 373–374
polymer supported, 353–356
ring expansion research and, 210–212
Diphenylmethyl (DPM) group:
cephalexin development and, 12
patent issues involving, 154–155
polymer synthesis, 353–358
ring expansion research and, 210–212
Dissolution rate, API-hypochloride
crystallization and, 179–181
Distillation, Evaporation, 174–177
Documentation SOP’s:
governing controlled environment rooms,
governing analytical instruments, 138
governing pilot plants, 60–61
governing process safety, 84–86
governing technology transfer, 136
last process steps for APIs and, 126
Doppler effect, flow measurement, 197
Drug discovery:
organizational structure matrix, 52–53
patent protection and, 141–142
Drug Master File (DMF):
Chemistry, Manufacturing, and Controls
(CMC) document requirements,
126, 132–133
IND/NDA applications, 126
Dryer systems:
basic equipment, 184–190
micronization consideration, 192–194
Drying times, API-hypochloride
crystallization and, 178–181
Dust explosions, safety procedures for,
chemical process development and,
sustainable development and, 376–377
synthesis technologies and, 333–335
redox reactions using, 366–374
use in wastewater treatment, 106–107
Electrodialysis, wastewater treatment using,
Electrospray-mass spectroscopy,
microwave-assisted hydrolysis of
dihydrohypoxanthine, 365–366
Embrittlement of solids, particle size
reduction and, 194–195
Emergency Planning and Community
Right-to-Know Act (EPCRA):
chemicals handling formalization,
provisions of, 90
Enantiomer structure, dilevalol
hydrochloride research and,
DBTA salt formation, 283–284
Enantioselective processes, water and
enzyme chemistry and, 339–344
Environmental issues:
cephalosporin C development and, 208
chemicals handling procedures, 96–100
organic process wastes, 100–103
overview of, 87–88
practical operations and, 91–107
regulatory acts concerning, 88–91
Environmental Protection Agency (EPA):
interaction with safety agencies, 66
wastewater treatment initiatives, 106–107
workplace safety practices and, 84–86
Environmental scientists, within the
chemical development organisation,
in chemical process development, 12–13,
206–208, 220, 290, 292–293,
microwave-assisted chemistry and,
Eosinophilia-myalgia syndrome (EMS),
Ergosterol, conversion to Dydrogesterone,
anti-inflammatory research, 243–248
cephalosporins, 12, 118, 145–146, 152,
154, 209–212, 224–225, 348, 361,
microwave-assisted chemistry, 361–369
Estrogens, identification of, 229
Estrone, production of, 236
European Inventory of Existing Chemical
Substances (EINECS), 91
European List of Notified Chemicals
Substances (ELINCS), 91
Evaporation, Distillation, 174–177
Exotherm reactions,
hydrochloride (POX-C) reduction,
centrifugation and, 182
chemical explosion, 69–70
differential scanning calorimetry
analysis, 70–72
dust explosion, 80–81
vapor explosion, 81–82
basic equipment, 171–174
extractive esterification, 212, 224–225
Extremely hazardous substances (EHSs),
classification of, 97–100
Schering-Plough synthesis for, 226–227
synthesis methodology for, 127
Failure, leadership skills in dealing with, 7
Felbamate, patent lessons, 146–150
Fermentation of penicillin G and
cephalosporin C, 204–208
basic equipment for, 181–190
Fischer indole synthesis,
microwave-assisted chemistry,
Flammable solvents, management of,
Flexibility, leadership and role of, 6–8
Florfenicol development:
patent aspects of, 155–162
process improvements for, 160–162
synthesis methodology for, 128
Flow measurement, instruments for,
Fluid bed dryers, 186
Flutamide, 117, 379
Fluorination reactions:
anti-inflammatory research and, 243–248
Florfenicol development and, 156–160
reactive system screening tool, 74–75
Food, Drug and Cosmetic Act, 110
Food and Drug Administration (FDA):
API-hypochloride crystallization and,
approval process for, 314
bureaucracy reduction and, 329–332
chemical process development and, 2–4
dilevalol hydrochloride case study,
chemistry, manufacturing, and controls
document requirements, 131–133
Compliance branch activities, 138–139
dilevalol hydrochloride review, 288–290
regulatory history of, 110–113
Review Branch requirements, 131–133
sustainable chemistry and, 378–380
validation process and, 138–139
workplace safety practices and, 84–86
Forced-air-heated dryers, environmental
requirements, 184–185
Friedel-Crafts reaction, polymer-supported
synthesis, 353–357
Fugitive emissions, 92
General Agreement on Trades and Tariffs
(GATT) treaty, 142
Generally Regarded as Safe (GRAS)
solvents, particle size reduction case
study, 192–195
Gestagens, identification of, 229
Gettysburg address, 326
Glaxo Laboratories:
antibiotics chemistry at, 11–14
cortisone and improved
anti-inflammatories, 234–248
patent protection issues and, 145–146,
scientists and engineers in, 17–18
em-4-carboxylic acid (II),
electrochemical reduction, 222–225
Good Manufacturing Practice (GMP):
evolution of, 110–113
maintenance practices, 201–202
Government inspections, organic process
waste disposal and, 102–103
Green chemistry:
organic process waste disposal and,
supercritical fluids and, 346–347
using water/enzyme systems, 337–344
Grignard reagents:
development of, 9–10
16?-methyl intermediates and, 253–255
Hazard and Operability Study (HAZOP),
runaway reactions, reactive system
screening tool, 74–76
Hazardous waste management, and
environmental laws, 88–91
Headcount, for Safety groups, 57–59
Heat exchange and control of, 169–171
raw material for anti-inflammatory
steroids, 234, 258
plant sources of, 258–260
Heinkel-type centrifuge, 183–184
Heptane, flammability characteristics,
Homogeneous energy supply,
microwave-assisted chemistry,
Horizontal spindle centrifuge, properties of,
Human resources, organization and function
of, 61–62
Hydrocortisone, structural foundation for
commercial anti-inflammatories,
AIC chemistry, 306–307
dihydrohypoxanthine to AIC, 311–312
Florfenicol development and, 161–162
9?-Hydroxyandrost-4-ene-3,17-dione, raw
material for steroids, 255–257
p-Hydroxybenzaldehyde, 368–374
11?-Hydroxylated betamethasone
intermediate, dehydration to C-9,
11-dehydrate steroid, 260–264
Hypoxanthine, 5-aminoimidazole-
4-carboxamide from, 299–307
dihydrohypoxanthine from, 307–311
Ignition, vapor explosions, 81–82
Impact sensitivity test, Nitro-dur scrubber
charcoal, 70–73
Impurities assay:
active pharmaceutical ingredients,
patent process and, 151–152
organic process waste disposal, 100–103
process emissions control and albuterol
case study, 93–96
Indefinite drug development program,
Industrial organization, some issues, 49–51
Industrial waste, government classification
of, 102–103
INOX-GLATT dryer, particle size and
milling issues, 192–195
Intermediate chemicals analysis group,
development of, 28, 57, 59
Internal symposia, importance of, 58–61
International Conference on Harmonization
(ICH), solvent guidelines from,
International Organization for
environmental standard (ISO 14000), 91
quality assurance standards (ISO 9000),
Investigational New Drug (IND) process:
API assay, impurities assay, and product
stability, 120–122
bureaucracy reduction and, 327–332
dilevalol hydrochloride case study,
filing procedures, 126
regulatory guidelines for, 112–113
synthesis methodology and, 130–131
toxicology batch, 114
Ion exchange, extraction process and,
Ionic liquids, properties and uses, 348–351
Iron contaminants, radex safety calorimeter
test, 76–78
Ishikawa reagent, runaway reactions,
reactive system screening tool test,
JANAF drop weight test, explosives
analysis, 70–73
Jet pulverizer micronizer, particle size
reduction, 192–195
Karr column, extraction using, 172
Keto aldehyde hydrate (KAH), process
emissions control, 93–96
Kraus-Maffei process containment unit,
187–189, 282–284
cardiovascular, properties vs. those of
dilevalol, 269–270
dilevalol hydrochloride separation from
labetalol vs. synthesis, 270–272
Labile reagent stability, radex safety
calorimeter evaluation, 76–78
future prospects, 226–227
microwave-assisted chemistry and, 363
overview, 203–204
penicillins and cephalosporins, 204–226
classical cephalosporins, 220–226
penem synthesis, 215–220
penicillin G/cephalosporin C
fermentation, 204–208
penicillin sulfoxide-cephalosporin ring
expansion, 208–211
semisynthetic compounds, product
quality, 214–215
Lagoons, wastewater treatment using,
Last process steps:
active pharmaceutical ingredients,
technology transfer and, 135–137
Late-stage intermediates, defining the
synthesis methodology,
Lawesson’s reagent:
cephalosporin development, 225–226
quazepam manufacture, 30
attributes of, 6–9
chemical process development and role
of, 5–9
role in organization of, 62–63
specifications and criteria for, 8–9
Lethal and toxic chemicals, extremely
hazardous substances classification,
Liquid-liquid extraction, 171–174
Liquid sulfur dioxide, properties and
potential, 347–348
Listening skills, leadership and role of,
Loratidine, 117, 379
Lovastatin, 342
Lumisterol, steroid raw material, 234
Magnetic flowmeters, flow measurement,
Maintenance procedures, pilot plants,
Management, internal symposia and role of,
disciplines managed by chemical process
development, 1–4
agreement an organisational structure,
14–15, 57–58
chemical process bridge to, 55–56
internal symposia and content of,
Marker degradation process, 232–233, 236,
258, 373
Marketing organization, place in drug
discovery, 53
Masked methyl isocyanates, temozolomide
preparation, 312
Mass spectrometry:
analytical involvement in AIC project,
AIC chemistry, 300–307, 310–312
Mass spectrometry: (cont.)
hypoxanthine reduction to
dihydrohypoxanthine, 308–311
Material Safety Data Sheets (MSDS):
contents of, 85
extremely hazardous substances
classification, 97–100
introduction to Safety/Health, 66
workplace safety practices and, 84–86
Matrix organization, drug discovery and
development, 51–53
Media, pharmaceutical industry coverage
by, 50–51
Medical organization in drug discovery and
development, 52–53
Mercury cathodes, replacement in
electrochemical cells, 222–224, 372
Mestranol/norethinodrel compound,
Metabolite development, API structure for,
116–118, 379
Metastable intermediates, chemical
explosion, 69–70
Method stage, chemical process
development and, 3–4, 165–167
16?-Methyl intermediates and, saponins to,
Methyl formate process to Felbamate
intermediate, patent issues and,
Methylhydrazine, temozolomide
preparation, 313
Methyl isobutyl ketone (MIBK), DBTA salt
to dilevalol hydrochloride, 285–289
Methyl (6S, 7R, 8R) 6-1(1-
penicillanate, 216–217
Methyl salicylate, raw material for albuterol
synthesis, 93–96
Methyl(S)-phenylglycinate hydrochloride
(POX-C), reduction, use of RC-1 for
heat of reaction study, 78–80
Methyl tropate reduction, for Felbamate
intermediate, patent issues
surrounding, 146–150
Metolaclor manufacture, 339–340
supercritical fluids and, 346–347
Metter RC-1 calorimeter, 68–69
Microbial contaminantion counts, quality
control assays, 120
overcoming clogging problems during,
particle size reduction, 190–195
progesterone, 235
unit designs, 191, 192
Microorganism, organic process waste
disposal and, 103–106
Microwave-assisted chemistry, trends and
technologies, 358–366
Milling process, particle size reduction,
Mitozolomide, 296
Mitsubishi Rayon acrylamide process,
Molecular foundations and designing round
process patents:
amoxicillin, 152–153
cephalosporin, 154–155
Molecular foundations of the steroid
industry, 227–229
Mometasone Furoate and raw material
selection, 256–257
Monsanto process, electrochemical
dimerisation of acrylonitrile, 368
Morphinans, biosynthesis inhibition,
National Academy of Sciences, 321
National Institute of Safety and Health
(NIOSH), mission, 65
National Research and Development
Council (NRDC), licensing process,
National Research Council (NRC),
cortisone program of, 240
New Drug Applications (NDA):
API-hypochloride crystallization and,
bureaucracy reduction and, 327–332
Chemistry, Manufacturing, and Controls
(CMC) document requirements,
dilevalol hydrochloride case study,
FDA review and compliance, 288–290
filing documentation, 126
quality control, impurities and stability,
regulatory guidelines for, 112–113
synthesis methodology and, 130–131
validation process, 133, 137–139
Nitric acid oxidation, recycling benzhydrol
waste, 211–214
Nitro-dur scrubber charcoal, differential
scanning calorimetry analysis,
Nitronates, accelerating rate calorimetry
analysis, 72–74
p-Nitrophenyl chloroformate,
temozolomide preparation, 313
Norethindrone, Syntex process for, 235–236
Norgestrel, synthesis of, 237–240
Notification of New Substances (NONS),
extremely hazardous substances
classification, 97–100
Nutrex unit, components of, 188–190
Occupational Safety and Health
Administration (OSHA):
carcinogenic chemicals list, 97
mission and activities, 65–66
process emissions protection, 93
workplace safety practices, 82–86
Olefin byproducts, steroid chemistry,
Operating structure, organization of, 58–61
Operational procedures, standard operating
procedures, 60–61
Oral contraceptives, early research on,
Orally absorbed antibiotics, ring expansion
and, 209
Organic chemical synthesis trends:
electrochemistry, 367–341
microwave-assisted chemistry and,
outsourcing, 335–337
other “solvents,” 344–351
use of polymer supports, 351–358
water and enzymes and, 337–344
Organic Process Research and
Development (journal), 164
Organic process wastes, management of,
in chemical development, 56–61
development through, 61–62
leadership role in, 6–9
mission and structure, 53–56
promotions in, 34–35
structure and function, 49–51, 59
therapeutic teams, 51–53
Orifice/Venturi meters, flow measurement
with, 196–197
Oseltamivir (Tamiflu), preparation of,
current trends in, 314, 335–337
last process steps for APIs and, 123–126
Oxaziridine stability, radex safety
calorimeter test, 76–77
Oxygen sensors, centrifuge explosion and,
cephalosporin synthesis and, 221–225
clean albuterol synthesis, 94–96
Papaverine, biosynthesis of, 385–386
Particle size:
active pharmaceutical ingredients,
milling, micronization, and precipitation
processes, 190–195
progesterone, 235
application content and process, 142–143
chemical process development and,
defense procedures for protection of,
history and exclusivity, 141–142
motivations for seeking, 143–144
odd patents, 162–164
timeliness of, 145–150
trade secrets vs., 155
worth of, 144–145
Penicillins, development of, 204–226
classical cephalosporins, 222–226
penem synthesis, 215–220
penicillin G/cephalosporin C
fermentation, 204–208
penicillin sulfoxide-cephalosporin ring
expansion, 208–211
semisynthetic compounds, product
quality, 214–215
synthesis of, 215–220
extraction equipment, 172–173
Penicillins: (cont.)
penicillin G development, 204–206,
6-APA quality problem, 214
Penicillin sulfoxides, ring transformation to
patent case, 154–155
process for, 208–214
People, vital importance of, 5–47
Perry’s Chemical Engineers Handbook,
Pharmaceutical Development scientists,
API collaboration with, 117–118,
177–181, 190–194
Pharmaceutical industry:
future trends in, 381–387
organizational structure in, 50–51
sustainable innovation, 327–332,
Pharmaceutical Research and
Manufacturers Association
organizational structure, 50–51
quality control and regulatory issues,
1-Phenyl-1,3-propanediol (PPD), felbamate
patent issues and, 146–150
Phosphorus compounds, steroid chemistry
and, 261–264
Phosphorus trichloride cleavage process,
patent protection for, 145–146
Pilot plants:
controlled environment rooms,
maintenance of, 201–202
multipurpose reactor package, 169–170
Piston pumps, 195–196
Plant-based chemistry:
future trends in, 105, 341–342, 381–387
steroids and, 256–260
Plant equipment and maintenance,
technology transfer and, 136–137
Plant failure analysis, runaway reactions,
reactive system screening tool,
Plosophore and chromophore, 69–70
p-Nitrobenzyl (PNB) protecting group:
cephalexin development and, 11–12
patent issues involving, 154–155
ring expansion research and, 209–211
Polychlorinated biphenyls (PCBs), Toxic
Substances Control Act provisions
concerning, 90
microwave-assisted chemistry and,
reagents, 357–358
synthesis methods and, 351–358
Positive displacement pumps, basic
principles, 195–196
Post-emissions calculations, air pollution
from process emissions, 92–93
Practical operations, environmental issues
in, 91–107
Pre-approval inspection (PAI):
development report, technology, transfer,
and validation, 133
dilevalol hydrochloride case study, 109,
new drug development and, 112–113,
Precipitation, particle size reduction,
Predictive maintenance, 201
Prednisolone, and improved
antiinflammatories, 242–249
Prednisone, and improved
antiinflammatories, 242–249
Pregnanediol issues, 231
Pressure issues, microwave-assisted
chemistry, 358–359
Preventive maintenance, 201–202
Process changes, technology transfer and,
Process development phase:
chemical process development and, 3–4,
computer technology and, 200–201
Process emissions:
to the air, 92
organic process waste, 100
wastewater, 103
Process engineering, dilevalol
hydrochloride case study,
Process hydration, opportunities, 102,
Process patents:
designing around patients, 152–155
process protection, 150–152
Process productivity, improving
electrochemistry, 372
Process safety:
calorimetric equipment for, 67–69
chemical structure with explosion risk,
chemical process development and, 2–4
runaway reactions, 74–76
overview of, 65–86
systems for, 57–61
workplace safety practices, 82–86
Process wastes, early management of, 92
Pro-drug development, active
pharmaceutical ingredients and,
Product stability assessment, active
pharmaceutical ingredients,
natural sources, 229–235
oral contraceptives based upon, 235–240
Progestogens, sources of, 233–237
Promotion system, chemical process
development and, 34–35
Pteris vittata, wastewater treatment using,
Publicly owned treatment works (POTW):
Clean Water Act provisions, 89
process emissions and, 92
wastewater treatment, 103
Pumping techniques and pump capabilities,
Purchasing chemicals and restrictions on,
Quality control (QC):
active pharmaceutical ingredients, 57, 59
API assay, impurities assay, and product
stability, 120–122
API quality specification and last process
steps, 115–116
API structure for, 116–118
chemical process development system
for, 113–114
crystal form and particle size, 119–120
Development report and, 133–135
dilevalol hydrochloride case study, 275,
electrochemical route to cephalosporins,
220, 225, 372
last process step guidelines, 122–126
semisynthetic penicillins and
cephalosporins, 214–215
synthesis methodology research and,
Radar detectors, reactor volume
measurement, 199
Radex safety calorimeter:
iron contaminants, labile reagent
stability, 76–78
process safety and, 68
R-aminoketone enzyme reduction, dilevalol
hydrochloride case study,
Reaction calorimeter (RC1):
optimisation of
reduction, 78–80
process safety and, 68–69
Reaction temperature enhancement using
microwave-assisted chemistry,
Reaction vessels, chemical engineering
perspective on scale-up,
Reactive system screening tool (RSST):
process safety and, 68
runaway reactions analysis, 74–76
Reactor jackets, pilot plant reactors,
Reactor volume measurement,
electricity, 367–374
enzymes, 337–344
polymer-supported, 357–358
Receiving company (RC), outsourcing,
Recipe stage, chemical process
development and, 3–4, 113,
Recrystallization, Florfenicol development
and, 160–161
Registration, Evaluation and Authorization
of Chemicals (REACH) program,
Regulatory Affairs department,
supplementary reinforcement of,
Regulatory issues:
bureaucracy reduction and, 326–332
overview of, 109–113
sustainable chemistry, 377–380
Research and development matrix
organization, characteristics of,
51–53, 62–63
Research organizational structure,
therapeutic team composition,
Residue-on-ignition (ROI) content, quality
control assays, 120–122
Resins, extraction process and, 173–174
Resource availability, utilization vs.,
Resource Conservation and Recovery Act
description of, 88–89
extremely hazardous substances
classification, 97–100
Reticuline in morphinon biosynthesis,
Reverse osmosis (RO), basic principles,
Rhizopus arrhizus:
progesterone hydroxylation, 240–241
Ring expansion:
penicillin sulfoxides to cephalosporins,
polymer synthesis, 353–357
Risk Management Program (RMP),
development of, 66
R-1-methyl-3-phenylpropylamine, dilevalol
hydrochloride and process for,
Rotameters, flow measurement with, 196
Rotary pumps, basic principles, 195–196
RR-amine process:
dilevalol hydrochloride research,
R,R-Amine preparation, 273
Runaway reactions:
methyl(S)-phenylglycinate hydrochloride
(POX-C) reduction, 78–80
preparation of Ishikawa reagent, 74–76
Safe Drinking Water Act (SDWA),
provisions of, 90
Safety and Environmental Departments:
environmental regulations and the
oversight of compliance, 87–108
safety in chemical process development
organisation, 57–61
workplace safety practices, 82–86
Safety award programs, workplace safety
and, 86
Salt formation:
of active pharmaceutical ingredients,
API-hypochloride crystallization and,
dilevalol DBTA salt development,
Saponins, 16?-methyl intermediates from,
intermediates research and, 255–260
isolation of, 232, 257–260
basic principles of, 167–169
electrochemistry and, 370–372
Schering penem Sch 29482, 218
Schering penem Sch 34343, 216–220
Schering-Plough Company:
awards for, 32–34
Ceftibuten development and, 220–227,
consultants with, 31–32
Florfenicol development and, 155–162
scientists and engineers at, 20–31
Science education, future needs in,
Scientists and engineers, chemical process
development and role of, 5, 15–47
Scrubber charcoal, safety procedures for,
Scrubbing systems, process emissions
control, 92–93, 183–185
Shionogi synthesis process, Ceftibuten,
Sodium dispersion chemistry:
development of, 10–11, 190–191
particle size, 190–191
capture equipment for, 92–93, 183–185
current research trends, 344–351
distillation and evaporation, 174–177
extraction process and, 172–174
ionic liquids, 348–351
last process steps for APIs and ranking
of, 124–126
liquid sulfur dioxide, 347–348
microwave-assisted chemistry, 358–366
organic process waste disposal and
recycling of, 100–101
recovery data for, 176–177
supercritical fluids, 345–347
Spherical dryers, 186
Spray dryers, 186
Standard Operating Procedures (SOPs):
controlled environment rooms, and, 186
chemicals handling process and, 99–100
maintenance practices, 201–202
organizational structure and, 60–61, 63
technology transfer, 136–137
workplace safety practices and, 84–86
anti-inflammatories, 240–249
betamethasone intermediatebetamethasone
alcohol conversion,
contraceptive synthesis development,
cucurbitacin molecular structure,
early history, 227–235
future research possibilities, 265
last process steps in manufacturing of,
molecular structure, 249–252
saponin-16?-methyl intermediates,
steroids with diverse biological activity,
structural chemistry, 227–235
Stigmasterol, raw-material, 231
Stirrer designs, sodium dispersion, 190–191
Sulfoxide, electrochemical reduction of a
cephalosporins, 224–226
Supercritical fluids as solvents, 345–347
Superfund Amendment and Reauthorization
Act (SARA), outline of, 88
Sustainable chemistry:
development of, 375–380
enzyme and fermentation processes,
Syntex oral contraceptive process, 235–238
Synthesis methodology:
electrochemical radar reactions, 366-375
microwave-assisted chemistry, 358–366
polymer-supported synthesis and
reagents, 351–358
research and development considerations,
trends and technologies for, 333–334
water and enzyme chemistry and,
Tank systems, wastewater treatment using,
Target molecule, stages of process
development, 166
Technology transfer:
and development report of, 133
chemical development and, 56
internal symposia discussion of, 57–61
pre-approval inspection and, 133
SOP for, 135–137
5-aminoimidazole-4-carboxamide (AIC)
chemistry, 298–312
dihydrohypoxanthanine hydrolysis,
preparation from hypoxanthine, 299
hypoxanthine reduction to
dihydrohypoxanthine, 307–311
original manufacturing process,
new preparation chemistry for, 312–313
Tetrahydrogestrinone, 249
Therapeutic teams, chemical development
and, 51–53
Thiamphenicol route to Florfenicol
intermediate, 156–160
Thin-film evaporation equipment, 176
Thin-layer chromatography, in improved
process to albuterol, 94–96
Third party manufacturers:
Chemistry, Manufacturing, and Controls
(CMC) document requirements,
initial process exploration with, 156–160
last process steps for APIs and, 123–126
patent agreements with, 155–156
qualifying a third party, 129–131
risks in outsourcing, 314, 335–337
Threshold limit values (TLVs), process
emissions protection, 66–67, 93
Tigogenin, plant sources of, 258–260
Time-weighted averages (TWAs), process
emissions protection, 66–67, 93
Toxicology batch, 114–115
Toxic Substances Control Act (TSCA):
extremely hazardous substances
classification, 97–100
provisions of, 90–91
Trade secrets, patents vs