The Business of Detection

Diagram of the Precipitation Apparatus, Smoke Particle-Size Project. approx. 1943.

Our Section L-1 on Aerosols has been set up to handle problems dealing with both offense and defense against toxic smokes. In connection with that program they have naturally run into the old problem of measurement of particle size and particle-size distribution, and have employed two or three of the more promising optical and microscope techniques in this connection… I wonder if you would give some thought to possible new methods of attacking this problem…

James B. Conant, letter to Linus Pauling, June 13, 1941

Following his work with the oxygen meter, Pauling and his right hand man, J. Holmes Sturdivant, were asked to create a carbon monoxide (CO) detection device. This new meter was intended to measure the levels of carbon monoxide in the air.

Carbon monoxide, a colorless, odorless, and tasteless gas, bonds to hemoglobin in the human bloodstream. In the process, carboxyhemoglobin (also known as carbonmonoxyhemoglobin) is created, preventing delivery of oxygen to body tissue, a complication eventually resulting in brain damage and death.

In high temperature environments where there exists an abundance of carbon, the combustion of carbon releases nitrogen and carbon monoxide. Certain tank cabins and airplane cockpits, in which CO could be released by the repeated firing of the vehicle’s weapons systems, were particularly subject to high levels of CO accumulation, sometimes leading to poisoning of the crew. As a result, the U.S. military needed a means of quick-testing air samples for CO saturation.

Pauling and Sturdivant realized early on that traditional chemical indicators like iodine simply weren’t well suited to the apparatus’ requirements. After only a few days of research, they discussed the possibility of using an organic substance as an indicator – hemoglobin seemed a likely candidate. After some calculations and a small amount of experimentation, they were ready to build a prototype.

J. Holmes Sturdivant, 1950.

The CO measurement apparatus contained a sample of hemoglobin molecules bonded with oxygen, otherwise known as oxyhemoglobin. When this sample came into contact with carbon monoxide, carboxyhemoglobin was created. The sample was then measured by a spectrophotometer, providing the user with a reading of the immediate environment’s carbon monoxide saturation in parts per million. Because the conversion of oxyhemoglobin to carboxyhemoglobin was reversible, the apparatus was capable of making consecutive readings without maintenance.

Moving from theory to practice proved a challenge however, as the duo soon found it more difficult than expected to build a working spectrometer that could accurately read carboxyhemoglobin levels. Between October 1942 and November 1943, Pauling and Sturdivant built several spectrometers, attempting to calibrate them in such a way that other environmental changes, including the addition of pure oxygen, would not disrupt the readings.

In December 1943, the pair had taken the project as far as possible. On many levels, the apparatus had proved itself inadequate for use in the field. For one, it was bulky and thus unsuitable for use in confined spaces like tank cabins and cockpits. Worse, it was too fragile to survive transport to the Pacific or European theaters, much less the strain of battlefield conditions. Finally, the device was hopelessly inaccurate unless used under specific, controlled conditions. All of these factors rendered it useless for its intended purpose, and the project was stopped dead in its tracks.

Disappointed, Pauling released his final report and quickly retired the project. In an attempt to find some justification for the hundreds of hours of labor that went into the device, Pauling noted that the apparatus was well-suited for use in the laboratory. Indeed, when deployed in a stable environment, the device was highly accurate and very useful for rapid measurements. As a result, the few meters that were produced before the closure of the project were distributed among researchers at Caltech for their own implementations.

Thanks to his work with the oxygen and carbon monoxide meters, Pauling was considered something of an expert on testing gaseous mixtures. In July 1942, while he was still working with both the oxygen and CO devices, he was asked to develop a field-use apparatus for identifying specific toxins in air samples. The task was complex, but Pauling thought that air contaminants could be identified by their representative particle size. Particle sizes could be determined by electrically charging a group of particles and then drawing them to a condenser plate. According to Stokes’ Law, the rate at which these particles are attracted to the plate is inverse to the radius of the particle. By examining this grouping, Pauling argued that it would be possible to estimate the composition of the smoke being analyzed.

Encouraged by some early calculations, Pauling set a lab assistant, Charles Wagner, to the task of making preliminary measurements. His results were positive and Pauling chose to move the program forward. By 1943 he was overseeing a group of men making calculations, building the apparatus, and creating stable smokes for the testing process, and by early 1944 the team was ready to put the device through its paces.

Diagram of the Filament Charging Device, Smoke Particle-Size Project. approx. 1943.

The initial tests were not good. The condenser plates were causing a bizarre phenomenon in which the largest particles were being grouped with the smallest, resulting in a highly inaccurate reading. And that was just the beginning of the team’s problems. They soon found that unfiltered air, such as that found in standard field conditions, contained a vast range of particles. In addition to the smoke or fog meant to undergo analysis, a typical sample could also contain dust, industrial pollutants, and natural contaminants like pollen. The distribution of precipitated particles was already making analysis difficult with clean samples; adding a host of impurities to the sample so complicated the results that an accurate determination was impossible for professional scientists. It was clear that a soldier in the field would be unable to operate the instrument effectively.

On March 28, 1944, Pauling filed his final report on the Particle Size Measurement Apparatus, number OEMsr-103. The project, he surmised, was a failure. In his final write-up for the Office of Scientific Research and Development, Pauling suggested that the apparatus might be reworked to give a more accurate reading under controlled laboratory conditions. While he frankly admitted that the instrument could “hardly be perfected for field use,” he hoped that his work and that of his fellow researchers had not been conducted in vain.

In hindsight, Pauling’s work with testing gaseous samples was doomed by the available technology. Complex instruments capable of operating in field conditions are difficult to engineer even today. Without access to microchips, solid-state computing equipment, or high-tech manufacturing processes, certain precision instruments were simply unachievable during World War II.