Story of Pure Synthetics, Inc.
Latest update: April 1, 2002
All designs, texts & graphics of this website are © JSP Schwarz, San Diego, California, United States of America.
Permission is granted to reproduce anything within this website as long as it is not-for-profit & proper credit is given for the source.
Late in 1972, I had learned from Sue DeHoff, a friend, that a flavor
chemical called methyl cyclopentenolone (MCP) was in short
supply. She was employed by JayCee Laboratories, a flavor house in
Linden, NJ, which was also a reprocessor of this material, getting its supply
through channels from Lambiote in France. At Lambiote,
the MCP was a byproduct of hard-wood charcoal manufacture along
with perhaps dozens of other products including wood alcohol and
maltol (another important flavor chemical). The world supply of MCP
was short at that time because Cliffs-Dow a subsidiary of Dow
Chemical Company had just closed its Michigan plant that produced it
using the same wood-charcoal-byproduct route. The shortage was made up to
some extent by synthetic material from Pfizer, Inc. Pfizer's
product was made by a German subsidiary and marketed under the trade name
Apetizer. Pfizer's price was $32.50/LB whereas the wood-charcoal, by-product
material was marketed at about $28-29/lb. In addition, California Aromatics
of Sun Valley, CA was manufacturing MCP using a process developed by Paul
Pratter based on well-known German and French technology starting with dimethyl
adipate. However, I do not believe that California Aromatics
ever were able to find a purification method to match our ENTRAINMENT-CONDENSATION DISTILLATION.
GO TO THE CONCLUSION OF THIS STORY.
I felt that the best way to begin making this product was to use a tried-and-true method if I could find one in the chemical literature. After searching, I uncovered two well-known methods in the French and German journals principally:
dimethyl adipate - - - (base-catalyzed ring-closure) - - ->
2-carbomethoxycyclopentanone - - - (methylation) - - ->
2-carbomethoxy-2-methylcyclopentanone - - - (oxidation with chlorine) - - ->
2-carbomethoxy-2-methyl-5,5-dichlorocyclopentanone - - - (hydrolysis & decarboxylation) - - ->
3-methyl-cyclopentan-1,2-dione - This is MCP!
[In the crystal state, the 2-keto group enolizes toward
the 3-methyl to relieve dipolar repulsion of the keto groups]
Starts with Claison and Dieckmann condensations of diethyl oxalate with dimethyl glutarate to form 3,5-dicarboalkoxycyclpentan-1,2-dione (in enol form) which is methylated to 3,5-dicarboalkoxy-3-methylcyclopentan-1,2-dione which, after hydrolysis undergoes two-fold decarboxylation to the product, MCP. For all the pretty structures that depict this route clearly, see MAKING REAL MAPLE FLAVORING: For chemists! Far-out way to make MCP!
I chose to investigate the second route first since I didn't relish trying to work with chlorine at plant scale, and cheap dimethyl glutarate had just come on stream from El Paso Natural Gas via C.P. Hall Company of Chicago. This was a by-product of nylon-6 manufacture; thus, it was inexpensive. In planning my process-development effort, I called the intermediates from the oxalate-glutarate route A-1, A-2, etc. and the intermediates from the adipate route B-1, B-2, etc. After the successful implementation of the oxalate-glutarate route, we continued to refer to the various intermediates by my code names although we almost forgot the reason why as the years passed. The product was always A-3 (MCP) and was formed in the A-3 reaction (aqueous mineral-acid hydrolysis and decarboxylation).
We started building our first plant in Paterson in Nov. 1973 about the time of the Arab Oil Embargo and the resulting shortages of petrochemicals and of many other important chemicals. It was a pivotal time: suddenly with the plant about half-constructed, my sources dried up for 50% caustic soda (50% aqueous NaOH), inexpensive sodium metabisulfite (formerly called sodium bisulfite), as well as the necessary common solvents: methanol and toluene.
I finally found a supplier for NaOH pellets which we dissolved in water and cooled prior to use. I decided to use sodium sulfite in place of the metabisulfite. This proved to be an unfortunate choice since its use caused the production of methyl iodide to become violent - - - and, believe me, that is one reaction you want to have under control! Better yet, never even think about using methyl iodide for anything. It's very toxic as well as expensive. But read on!
Since my first few trials in the lab trying to methylate A-1 (3,5-dicarboalkoxycyclpentan-1,2-dione) to A-2 (3,5-dicarboalkoxy-3-methylcyclopentan-1,2-dione) using methyl sulfate didn't seem to be successful, I chose to use methyl iodide made from dimethyl sulfate and sodium iodide - - - in turn, derived from iodine, sodium metabisulfite, and sodium hydroxide. I knew that methyl iodide worked and I could recycle the iodide from the methylation reaction to make more methyl iodide. The recycling feature made the cost of methylating agent far less than $1/LB although the iodine was a little over $2/LB imported from Chile by Chilean Nitrate Sales Corp.
In the middle of January, 1974, I drove over to nearby Mallinckrodt
and bought 100 LBS of lab-grade sodium metabisulfite. It cost more than
a suitable industrial-grade material would have cost, had it been available (byproduct of Arab oil embargo); but, in my mind, the important thing was to get into production - - - I had always planned on replacing methyl iodide with dimethyl sulfate as the
methylating agent: one fewer step and less expense.
On January 20, 1974, Sunday, we made two plant batches
of methyl iodide at 72-l scale (19-gal), both going smoothly. I looked after
these preparations, the subsequent distillations and the drying of the
distilled methyl iodide. Helen, my wife and partner in this grand enterprise, worked in the same large room with me that day doing batches of the first-step
reaction (A-1 reaction) in a large flask (JPG: 45K).
In the midst of the bustle, I cut my finger on a glass tube when it broke
as I tried to insert it into a holed stopper. The finger bled profusely
and required three tight adhesives bandages to stop the bleeding. What happened
next, we surmised many weeks later, was that I spilled some methyl iodide
on the bandaged finger. Since methyl iodide is a local anesthetic, one feels
no pain as you would with alcohol on a cut. Furthermore, the methyl iodide
which boils a few degrees above body temperature cannot evaporate quickly
as it would if you merely spilled some on skin open to the air. This conjecture
is supported by the fact that when the physicians removed the bandage after
I was admitted to the hospital, they found a large vesicant blister on the
finger. But I am getting ahead of my story.
Just as we were preparing to close up and go home for dinner, I began to
lose my balance in a way that had never happened before. I, at first, thought
that I was just very over-tired because I had been feeling tired much of
The next morning as Helen prepared to go to her regular job (high-school-guidance
counselor), I told her that I was feeling a little off that morning and
that I would stay in bed. By the time she came back that evening, I was
really tired but wasn't bad enough to cause her tremendous alarm. But that
night I thrashed all over the bed with violent nausea. She made arrangements
to take me to the Neurological Institute at Columbia Presbyterian Medical
Center in New York City early the next morning.
By that time, I was pretty disoriented. My last rational thought as I sat
in a wheel-chair was that I must be very sick, but, if I could just tie
that shoelace which had just become untied, then perhaps things weren't
so bad after all. I couldn't do it! The next four weeks were a blur of confusing
fantasies with some hallucinations and misunderstandings of events transpiring
around me. I "woke up" four weeks later.