A Measure of Success
A Finnigan Instrument Corporation Model 1015 GC/MS/DS. From left to right: a minicomputer from Digital Equipment; part of the quadrupole mass spectrometer; the remainder of the mass spectrometer electronics console and gas chromatograph. Image courtesy of Robert Finnigan.
In 1970 many Americans, dismayed by pollution in the air, water, and soil, participated in the first Earth Day and contributed to the political momentum that led to the creation of the Environmental Protection Agency (EPA). To protect the environment the EPA needed legislation, and to create legislation it needed the ability to precisely identify pollutants and measure their concentrations. Enter an unlikely entrepreneur, ex–cold war engineer Robert E. Finnigan.
Computers played an important role in waging the cold war. In the 1950s, at the Lawrence Livermore National Laboratory scientists and engineers used some of the most advanced digital computers of the time to design nuclear weapons. Finnigan, an Air Force officer with a Ph.D. in electrical engineering, led a Livermore effort to develop a computerized control system for a nuclear reactor intended to power a thermonuclear missile that could prowl the skies for days. By 1962, though his team had successfully run a prototype reactor using a computer, Finnigan was convinced the missile program was doomed.
Finnigan and collaborator Mike Uthe wondered what they would do with their expertise in computer control of advanced nuclear reactors until the Stanford Research Institute (SRI) recruited them to start a controls group. At the time another team at SRI was developing quadrupole mass spectrometers, in which particular combinations of voltages are applied to four parallel rods, allowing selected ions to pass between them and reach a detector. Such an instrument can identify specific substances and measure their quantities. Finnigan and Uthe speculated that such a spectrometer would have broad applications.
Two years later Finnigan and Uthe joined Electronic Associates, Inc. (EAI), a leading U.S. supplier of analog computers, and continued working on quadruple mass spectrometers. The two wanted to contract the quadrupole research out to their former SRI colleagues but SRI shut down development, telling its researchers to buy the instruments rather than produce the devices themselves. However, no quadrupole mass spectrometer manufacturers existed, so Finnigan decided to become a commercial manufacturer, with SRI as his first customer.
Jack Jennings and Charles Rosen at SRI decided to simply give Finnigan what he needed to get into commercial production as quickly as possible: know-how and names. From 1964 to 1966 Finnigan and Uthe’s EAI division sold over 500 quadrupole residual gas-analyzer instruments. Researchers working in microelectronics and materials science were major consumers. With orders increasing, Finnigan and Uthe raised the price of their instruments, and their upstart division provided most of EAI’s profit.
While marketed as residual gas analyzers, the instruments that EAI produced from 1964 to 1966 were bona-fide mass spectrometers. They were not, however, adequate for the demanding analyses required in chemical research. Much would be required to transform them into analytical instruments. Finnigan became increasingly interested in the idea of a computer-controlled instrument in which a gas chromatograph (GC) would be used to separate the constituents of complex samples, and a robust quadrupole mass spectrometer (MS) would determine the nature and quantity of these constituents: a computerized GC/MS. With this idea Finnigan was developing a vision of the next phase of the instrumentation revolution in chemistry, a phase associated with the minicomputer revolution.
The first minicomputers appeared in the early 1960s. These were refrigerator-sized digital computers that sold for tens of thousands of dollars. Laboratory scientists first used them to process the data from a single laboratory’s instruments. Palo Alto was home to several laboratories leading the application of GC/MS to biochemistry, including those of Carl Djerassi and Joshua Lederberg, who worked in pharmaceuticals and genetics, respectively. But making sense of the data produced in a single GC/MS run was extremely time-consuming. The measurement could be performed in a few hours, but organizing the data from a single run and then manually interpreting it could take weeks, even months. Both groups were looking at minicomputers as a solution to this glacially slow pace of data handling and interpretation. Their ultimate goal—real-time speed where data interpretation happens as a measurement takes place.