War of Nerves

by Jonathan Tucker

  The transester process had the advantage of generating a highly pure product that did not need to be distilled. To test the new process on an industrial scale, the Edgewood engineers built two pilot plants: one capable of making twenty pounds of VX per eight-hour shift, and a larger version that produced 250 pounds. Based on experience with the two pilot plants, plans were drawn up for a full-scale VX production facility with an output of ten tons of agent per day.

  In April 1957, Dugway Proving Ground organized a “V-Agent Team” to handle testing and evaluation of new delivery systems for VX, including self-dispersing submunitions dropped from aircraft or guided missiles. Engineers sought to design bomblets that would generate VX droplets of optimal size for military purposes, taking into account the downwind transport of the agent cloud, absorption of the droplets through clothing and skin, and other factors. From the standpoint of chemical defense, the V agents posed difficult technical challenges because of the need for complete protection of the skin and decontamination of vehicles and equipment.

  WHILE ONE TEAM of chemical engineers at Edgewood was working on a manufacturing process for VX, another team began to develop “binary” chemical weapons, in which two relatively nontoxic ingredients were combined inside a bomb or shell to yield a lethal nerve agent. The basic principle of binary weapons dated back to 1885, when military chemists had attempted to produce nitroglycerin, a dangerously unstable explosive, by combining its two chemical constituents (nitric acid and glycerin) inside an artillery shell. In the late 1940s, scientists at Edgewood Arsenal had studied binary munitions as a possible way to reduce the hazards associated with the production, storage, and handling of unitary chemical weapons. Leading this research effort was a hulking German chemist named Fritz Hoffmann. He had worked in the Nazi nerve agent program and, after the war, had been recruited by the U.S. government under Project Paperclip. After first experimenting with a design for a binary mustard bomb, Hoffmann switched to nerve agents in the early 1950s. Because the last step in the synthesis of Sarin, Soman, and VX involved the reaction of two precursors, he realized that it would be possible to store these chemicals in separate compartments inside a bomb or shell and cause them to mix and react to produce the lethal agent while the munition was still in flight to the target. (For technical reasons, Tabun could not be produced in a binary system.)

  In 1954, building on Hoffmann’s work, three engineers in the Weapons Research Division at Edgewood—Ted Tarnove, Gene Bowman, and Marty Sichel—developed a concept for a binary VX bomb in which powdered sulfur would be injected into a liquid solution of QL. The technology looked promising, but Colonel Jim Hebbeler, the head of the Chemical Warfare Laboratories, and Dr. Ben Witten, the director of weapons research and development, were skeptical about its military utility. Although it was true that the storage and transport of binary chemical munitions would entail fewer risks, that difference meant little to Chemical Corps officers, who prided themselves on their ability to handle unitary weapons safely. Furthermore, because the reaction of two chemical precursors inside a binary bomb or shell would inevitably be incomplete, such a munition would actually deliver less nerve agent to the target than a unitary weapon, reducing its military effectiveness. Given the lack of institutional support for binary weapons, Sigmund Eckhaus, the director of the VX pilot plant at Edgewood, had to “smuggle” samples of QL to the binary development team. Ultimately, the lack of interest from senior Army officials led to the cancellation of the binary R&D program after a few years of work.

  IN PARALLEL with the development of V agents at Edgewood, scientists in U.S. government laboratories and academic institutions worked on improved medical defenses against nerve agents. This area became a high priority when it was discovered that atropine, which could counter the effects of Tabun and Sarin, was far less effective at treating exposures to Soman or VX. In the mid-1950s, the Chemical Corps awarded a secret contract to Dr. David Nachmansohn, a professor of biochemistry at Columbia University’s School of Physicians and Surgeons in New York, to investigate the toxicology of the nerve agents and develop new antidotes.

  Nachmansohn chose to work with electric eels from the Amazon basin, which generate high voltages to stun their prey. These animals provide an excellent experimental model of the human nervous system because the electrochemical phenomena are greatly amplified and therefore much easier to study. Moreover, because the nervous system of the electric eel is highly enriched in cholinesterase, it is possible to purify the enzyme in relatively large quantities. To obtain enough cholinesterase for his experiments, Nachmansohn asked the Army to procure a hundred electric eels. The Army contracted with the New York Aquarium, which in turn gave the assignment to J. Auguste Rabaut, a middle-aged Frenchman and expert fisherman who lived in the upper reaches of the Amazon basin. Rabaut trapped more than a hundred electric eels and kept them alive in waterfilled tanks, which were shipped by plane to New York City.

  At Columbia University, Nachmansohn and his coworkers purified cholinesterase from the nervous tissue of the eels. Because samples of Sarin or VX were not available for experimental use, the scientists used the less potent nerve agent DFP, which was easier to synthesize and safer to work with. In test-tube experiments, the investigators found that DFP completely inactivated eel cholinesterase but did not destroy it. This observation suggested that it might be possible to develop an effective treatment for nerve agent poisoning by restoring the activity of the enzyme.

  Subsequent experiments by Nachmansohn and his colleague Irwin Wilson determined that the nerve agent molecule attaches to the “active site” of cholinesterase, a groove in the surface of the enzyme where catalysis occurs, blocking the further breakdown of acetylcholine. Once the Columbia researchers had elucidated this mechanism, they sought to develop a drug that would displace the nerve agent molecule from the active site of cholinesterase. In this way, it would be possible to restore the normal activity of the enzyme and counteract the effects of the poison. Through a long and tedious process of trial and error, Nachmansohn and his team synthesized dozens of novel compounds and screened them for the ability to reactivate cholinesterase. Finally, a compound called PAM (pyridine aldoxime methiodide), a member of the class of drugs known as oximes, proved to be highly effective at displacing DFP from the active site of cholinesterase and restoring the normal function of the enzyme, at least in the test tube.

  Excited by this finding, Nachmansohn devised an experiment to test the activity of PAM in living animals. He divided forty white mice into two groups of twenty and injected all of them with DFP. The first group of mice was then given a prompt injection of PAM, while the twenty “control” mice received only a saline solution. Within five minutes, all of the control mice were dead, but the treated mice continued to scamper about unharmed. Subsequent studies on laboratory animals with more potent nerve agents, such as Sarin and VX, showed that PAM was of limited benefit when given alone but highly effective when administered together with atropine. Whereas treatment with PAM raised the lethal dose of VX two- or threefold, PAM plus atropine increased the level of protection tenfold. The explanation for this difference was that the two antidotes worked synergistically through different but complementary mechanisms. By blocking the receptors for acetylcholine and thereby counteracting its physiological effects, atropine served to tide the animals over until PAM restored the normal activity of their cholinesterase.

  Based on the knowledge that the basic operation of the nervous system is similar in all mammals, the Columbia scientists expected that PAM would also be effective in humans. To test their prediction, they administered the antidote to human volunteers exposed to low concentrations of Sarin vapor. These experiments demonstrated that PAM itself was relatively nontoxic and that it was highly effective at counteracting the symptoms of nerve agent exposure, especially when given together with atropine.

  The next challenge was to develop an effective means of administering the two antidotes into the thigh muscle of a soldier withi
n minutes of exposure. Although metal syrettes containing atropine had been issued to U.S. troops in the 1950s, they posed numerous problems. Not only did soldiers wearing gas masks and thick rubber gloves have difficulty manipulating the metal tubes, but many individuals were understandably reluctant to plunge a needle into their own body. To solve these problems, British scientists at Porton Down developed an automatic syringe called the “Autoject.” In the United States, Stanley J. Sarnoff, a professor of physiology at Harvard University, invented a similar device called the “Ace autoinjector,”which the Army standardized in 1959. Both systems consisted of a cigar-shaped syringe containing a premeasured dose of atropine or PAM and a recessed, spring-loaded needle. A soldier who had been exposed to nerve agent simply released the safety catch and pressed the end of the device against his thigh. The pressure triggered the spring mechanism, driving the needle through layers of protective clothing and injecting the correct dose of antidote.

  A U.S. soldier self-administers nerve-agent antidote with an autoinjector during a chemical warfare exercise.

  Another top development priority for chemical defense was a portable battlefield detection system for nerve agents. Since Sarin was effectively odorless, an automatic detector and alarm system were needed to give troops advance warning of an approaching agent cloud so that they could don their gas masks in time. Early field detectors were slow and unreliable, and until the early 1960s, the British Army continued to use caged canaries—which are more sensitive to nerve agents than humans are—as their primary detection system. In the United States, the Chemical Corps developed the M6 automatic G agent field alarm, which was standardized by the Navy in 1958. When any G agent came in contact with a reagent in the device, it produced a color change in a paper tape that was detected by a photo cell, triggering a buzzer alarm. The M6 weighed twenty-five pounds, fit into a portable aluminum case, and could operate unattended for twelve hours, at which time it required fresh solutions and new tape. Nevertheless, the detector had major drawbacks: it ran off a battery, did not operate below freezing, and required frequent maintenance. Thus, although the Navy procured over five hundred of the units for its dockyards and ten for shipboard use, the Army rejected it for standardization.

  UNDER THE TRIPARTITE AGREEMENT, British, American, and Canadian scientists shared technology and coordinated research at biannual conferences on toxicological warfare. At the Eleventh Tripartite Conference in May 1957, for example, participants from the three countries agreed to intensify their work on the V-series agents. Research topics judged to be of highest priority were studies on the deposition and toxicity of agent droplets of various sizes on skin, clothing, and respiratory passages; the persistence of and residual hazard from agents in the field as influenced by climatic conditions; and the development of new agents or additives that could more readily penetrate the skin. At the Twelfth Tripartite Conference in the fall of 1957, the participating scientists agreed on a division of labor, with the Americans continuing to develop land and air munitions for the dispersal of V agents, the British evaluating the military potential of these agents with model systems, and the Canadians taking the lead in determining the secondary hazard from contaminated terrain.

  Under a second agreement called the U.S.-U.K. Mutual Weapons Development Program, British process engineers collaborated with their colleagues at Edgewood Arsenal to develop industrial production techniques for VX and other V agents. At Nancekuke, the British built a small pilot plant that produced kilogram quantities of VX with a method known as the “water process,” which they considered superior to the U.S. transester process. The bilateral collaboration was highly productive, but British scientists complained about what they perceived as a “one-way street”: they felt that they told the Americans everything and received relatively little information in return.

  Along with the expanded research and development on the V agents, the U.S. Army Chemical Corps began to plan for the large-scale production of VX. In 1957, the corps appointed a V-Agent Committee to study the problems of VX manufacture and to select a suitable site for the facility. For safety reasons, the plant could not be located near a densely populated area. The Chemical Corps also did not want to produce VX at an existing site, such as Muscle Shoals or Rocky Mountain Arsenal, on the principle of “not putting all of your eggs in one basket.”

  The V-Agent Committee finally decided to locate the VX plant at the Wabash River Ordnance Works near the town of Newport, Indiana, a small farming community in the western part of the state, about thirty miles from Terre Haute. The Ordnance Works had originally been built in 1941 to produce plastic explosive for World War II. It was also the site of an inactivated government facility, the Dana Heavy Water Plant, which the Atomic Energy Commission had built in 1952 to produce heavy water (deuterium oxide) for the U.S. nuclear weapons program. The abandoned AEC plant included sixty giant extraction columns that had served to refine heavy water from huge volumes of natural water drawn from the Wabash River. Although most of the columns were eventually torn down, the few left standing were incorporated into the design of the VX plant as structural support elements for pumps, pipes, and reactors.

  Major General William M. Creasy, the chief of the Chemical Corps, decided to contract with private industry for the design, construction, and operation of the VX factory and the associated munitions filling line. On May 9, 1958, the Chemical Corps asked the Army Corps of Engineers to solicit proposals from qualified industrial firms. A year later, on June 23, 1959, the Department of Defense signed contracts with the Lummus Company for the design of the plant and with the Food Machinery and Chemical (FMC) Corporation for its construction and operation. After two years of work and an expenditure of $8 million, the Newport Army Chemical Plant was completed in April 1961. Because of a number of technical problems, full-scale production and loading of munitions did not get under way until 1962, with a planned output of ten tons of VX per day.

  Meanwhile, several new delivery systems for nerve agents were under development. As its “workhorse” weapon for delivering VX or Sarin on the battlefield, the Army developed the M55 rocket. Made of rolled and welded sheet aluminum, the six-foot-long, fin-stabilized rocket had a range of six miles. It was an integrated package that included a solid-fuel motor, a warhead filled with five quarts of nerve agent, an impact fuse, and an explosive burster charge that would disperse the agent as a spray or vapor. In 1960, the Army Chemical Corps let a production contract for the M55 to the Norris Thermador Company of Los Angeles, which ultimately manufactured about 478,000 rounds. Each filled rocket weighed fifty-five pounds and was packed in a fiberglass shipping tube that also served as a launching tube. A single mobile launcher held up to forty-five rockets that could be fired in salvos, theoretically drenching a target with nerve agent. From the outset, however, the M55’s performance in testing was highly erratic and unreliable.

  A M55 rocket with a nerve-agent warhead is test-fired from a multiple-tube launcher system. Hundreds of thousands of these rockets were produced and filled with Sarin or VX in the early 1960s. Many of the Sarin-filled rounds soon began to leak, creating a huge disposal problem for the U.S. Army.

  In addition to the M55 rocket, new weapons in the U.S. chemical arsenal included a 750-pound aerial bomb containing 220 pounds of Sarin, a land mine that dispersed 11.5 pounds of VX, and VX-filled artillery shells. The Air Force, the Army, and the defense industry also developed self-dispersing cluster warheads for three mid-range tactical rockets: the Little John, the Honest John (range: 16 miles), and the Sergeant (range: 75 miles). Standardized in 1960, the Honest John warhead contained 356 spherical bomblets made of ribbed aluminum, each 4.5 inches in diameter and containing about a pound of Sarin. The warhead was designed to break open in flight over the target area and release the bomblets, which armed themselves by spinning as they fell to earth. Detonating on impact with the ground, the bomblets released clouds of Sarin vapor that merged to blanket a large area. Field tests indicated that a single Honest
John warhead could generate a lethal concentration of Sarin over a radius of 500 meters, not including downwind spread.

  An Honest John rocket designed to deliver a nerve-agent warhead is shown being test-fired around 1964.

  The Honest John rocket warhead contained 356 spherical aluminum bomblets, each of which contained about a pound of liquid Sarin.

  Cutaway of an M139 Sarin bomblet shows internal cavities that were filled with liquid Sarin. At its center are a fuse and an explosive burster that detonated on impact, generating a fine mist of nerve agent.

  A full-scale model of the 4.5-inch-diameter M139 Sarin bomblet, similar to the ones used in cluster warheads.

  ALTHOUGH FRANCE did not participate in the Tripartite Agreement, it launched an independent effort to develop and produce V-SERIES nerve agents. In January 1958, Defense Minister Jacques Chaban-Delmas transformed the Special Weapons Command, which had been subordinated to the General Staff of the French Army, into the Joint Special Weapons Command (Commandement Interarmées des Armes Spéciales), reporting to the Joint Chiefs of Staff. At the same time, the organization for chemical and biological weapons R&D was restructured so that it served all of the French armed services.

  Throughout the Algerian war of independence (1954–62), the French continued to test chemical weapons at B2-Namous. Despite the bloody colonial war taking place nearby, the testing campaigns in Algeria focused exclusively on the Warsaw Pact chemical threat to Western Europe. Indeed, the nerve agent trials were conducted by technical elements of the French Army that had no involvement in local combat operations.

  In 1959, chemists at the Centre d’Études du Bouchet synthesized VX, which they gave the code name “A4.” Three years later, in 1962, U.S. Army scientists transferred data on the V agents to their French colleagues after first obtaining permission from the British government. Over the next few years, French process engineers developed industrial production methods for Sarin and VX. During the mid-1960s, France manufactured several dozen tons of Sarin and 400 kilograms of VX for testing purposes at a pilot-scale (“semigrand”) production facility in Braqueville, near Toulouse in southern France.