Scaling up penicillin

From tender-minded laboratory chemists to tough-minded chemical engineers.

Ben Recht writes about the case studies on the efficacy of penicillin, done in 1941 by the team of William H. Florey at Oxford. Ben’s post is primarily about the value of case studies as an evidentiary tool. However, there is an important historical aspect that touches on the profound difference between laboratory synthesis of chemicals vs. large-scale industrial production. This was an urgent matter in the early days of World War II, as Florey’s team realized after their heroic efforts of growing the penicillin mold in hundreds of pots failed to yield enough of the drug to treat the first patient. (Recycling the penicillin secreted in the patient's urine did not provide sufficient quantities either, and the patient died after a relapse of the infection.) Sunny Auyang’s book Engineering—An Endless Frontier¹ discusses how a broad cooperative effort between laboratory chemists and chemical engineers quickly led to large-scale wartime production of penicillin:

An ample supply of penicillin was ready for Allied soldiers who stormed the Normandy beaches on June 6, 1944. Three months earlier, the first commercial plant for penicillin had started operation. By D-day, monthly production exceeded 100,000 doses. Wounded soldiers were injected with penicillin as a precaution against infection, and 95 percent of those treated recovered. No longer would infectious bacteria ravage battlefields and hospitals.

A potent medicine has little impact if it is more precious than fine diamonds. Increasing yield and recovery, scaling up production, and making penicillin a lifesaver was an achievement of chemical engineering. In June 1941 Florey took his research results to America. A project was promptly set up under the direction of Robert Coghill of the U.S. national laboratory at Peoria, Illinois. What proceeded was one of the greatest international collaborations in research, development, and production, involving the free flow of information among governments, industries, and universities. Scientific research improved the mold strands, feed media, and chemical processes tremendously. However, despite an elevated yield, scientists still took eighteen months to produce enough penicillin for tests on 200 patients. Then engineers swung into action, mobilizing their knowledge about chemical processing. Working closely with microbiologists, engineers at Pfizer introduced methods for deep submerged fermentation in 1,000-gallon tanks. Merck engineers invented violent circulation for oxygen transfer, a process that became the core of stirred-tank reactors suitable for growing air-breathing microbes. Racing against time, what is now called concurrent engineering was practiced although not named. When, nine months before D-day, penicillin moved beyond the pilot-plant stage, some firms had one team of engineers building production plants and another developing fermentation and recovery processes for the plants. The first plant to operate, Pfizer’s in Brooklyn, took only five months from design to operation.

The broader context for this is the emergence of chemical engineering as a discipline through the work of George Davis in Great Britain and of William Walker and Warren Lewis in the United States in the early 1900s, despite the fact that Germany was the world leader in scientific and industrial chemistry at the time. Through the efforts of Walker and Lewis at MIT, the idea of unit operations, i.e., the processes common to many types of chemical reactions that could be combined in a modular way in reactor design, became a core concept in the chemical engineering curriculum, which before then was mostly a mixture of chemistry and mechanical engineering. This brought to the fore the difference between laboratory synthesis and industrial production. While Fritz Haber received the 1919 Nobel Prize in Chemistry for synthesizing ammonia, Carl Bosch, who was a pioneer of high-pressure industrial chemistry, shared the 1931 prize for inventing the industrial process for large-scale production of synthetic ammonia. As Auyang writes,

Industrial chemical processes are not simply larger versions of laboratory chemical reactions. Besides constraints imposed by cost, production volume, and environmental impact, many hurdles obstruct scaling up chemical reactions from the test tube to the industrial tank. A chemist can shake a flask over a flame to heat and mix up the reactants for a desired reaction, but a similar method applied to a thousand-gallon tank could end in a deadly explosion. Large containers, with their smaller surface-to-volume ratio, are more inimical to heat distribution. To heat their contents, close attention must be paid to processes of fluid motion and heat transfer to ensure the proper heating and mixture required for the chemical reaction. The physical processes relevant to industrial chemical reactors can be systematically discovered and their general principles studied. Then they can be readily used to scale up a wide variety of chemical reactions for efficient industrial production. It is in these tasks that chemical engineering distinguishes itself from chemistry.

To me, this serves as a good illustration of the distinction William James made between the “tender-minded” rationalists, who go by principles, and the “tough-minded” empiricists, who go by facts. While James’ intent was to single out two philosophical temperaments, the distinction applies more broadly to other complementary modes of inquiry, such as science and engineering.

1. The book’s title is a reference to Vannevar Bush’s 1945 report “Science the endless frontier”.