Blisters as Weapons of War: The Vesicants of World War I

By Joel A. Vilensky and Pandy R. Sinish

In World War I two vesicant (blister-causing) compounds, mustard gas and lewisite, received much notoriety for their toxicity. Lewisite in particular was hailed as what would now be called a “weapon of mass destruction.” Whereas mustard was discovered some 90 years earlier, lewisite, the “dew of death,” was developed primarily during the war. Both the Allies and the Central Powers pursued research on these compounds, and various countries continued to produce them throughout most of the 20th century—and sometimes to use them. But on the Allied side during World War I the United States led the way in producing these vesicant agents in massive quantities.

Resorting to Gas

Although poison agents were used in combat before World War I, that is the war most associated with the use of poisons as weapons. Gas was deployed in response to the stalemate reached by early 1915, as both the Allies and the Central Powers realized that their own high-explosive artillery shells were ineffective at dislodging soldiers from their defensive trenches, and that opposing machine guns effectively prevented offensive actions without devastating losses. The belligerents thus sought a weapon to drive their opponents from the relative safety of their trenches. Poison gas became that weapon.

France had actually initiated plans to use tear gas, but the German forces used chlorine first, on 22 April 1915. The man who designed the chlorine attack was the German chemist Fritz Haber, director of the Kaiser Wilhelm Institute for Physical Chemistry in Dahlem, near Berlin, and head of the German gas program. Haber was then known for the process he had developed with Carl Bosch of BASF in 1910, of fixing nitrogen from the atmosphere for use in fertilizer, for which the two would win the Nobel Prize in 1918.

The German army’s initial use of chlorine succeeded, but the Allies rapidly developed gas masks, rendering chlorine attacks much less effective. They also began using chlorine against German troops. The Germans then retaliated with a more toxic agent, phosgene, on 19 December 1915. But the Allies again quickly developed protective gas masks and retaliated with the same agent. Thus gas was not proving to be the decisive weapon the German High Command had envisioned, and they were on the lookout for a new one. Mustard seemed a promising candidate.

The History of Mustard Gas

Mustard, a liquid that is vaporized to produce a battlefield agent, is chemically 2,2'-dichlorodiethyl sulfide. Probably the first to produce mustard was César-Mansuète Despretz, a Belgium-born chemist and physicist who on combining ethylene and sulfur chloride in 1822 observed the synthesis of a foul-smelling liquid. Similarly, Alfred Riche, a French chemist and colleague of Auguste Cahours, produced it in 1854 by reacting chlorine with ethyl sulfide. Slightly later, in 1860, the British chemist Frederick Guthrie bubbled ethylene through sulfur dichloride to produce a liquid with “a not unpleasant but indescribable smell; its taste is intensely sweet and pungent.” Furthermore, he observed in a footnote, “A drop placed beneath the tongue destroys the epidermis and causes a soreness which lasts for days” (Guthrie, 1860). Also in 1860 the German chemist Albert Niemann, best known for isolating cocaine from the coca plant leaf, described mustard’s vesicant properties: “[T]he minutest trace which may accidentally come in contact with any portion of the skin, though at first it causes no pain, produces in the course of a few hours, a reddening and on the following day, a severe blister which suppurates for a long time and is very difficult to heal” (Niemann, 1860).

The renowned 19th-century German chemist Victor Meyer was the first to characterize the compound definitely, in 1886. Meyer was working in his Göttingen laboratory with ethyl chlorohydrin, which he combined with sodium sulfide to produce the compound thiodiglycol.

Thiodiglycol was then chlorinated with phosphorous dichloride to produce mustard. Meyer’s assistant was seriously injured by the final product. Meyer wondered whether his assistant’s blisters and conjunctivitis might be the result of a mental problem instead, because there was nothing in mustard’s precursors to suggest it would be so toxic. Accordingly, Meyer decided to have the product tested further at a medical school, on rabbits. The rabbits exposed to the vapors developed conjunctivitis and then died. Meyer wrote: “The intended work with this chloride was not continued—on account of the extremely poisonous qualities of the compound. It is very striking that this apparently harmless substance . . . should exert a specific toxic effect. Its chemical properties would never lead one to expect its aggressive properties.”

No further published work exists for this compound until 1912, when the British-born chemist Hans T. Clarke, then studying under Emil Fischer in Berlin, improved Meyer’s process by substituting hydrochloric acid for phosphorous chloride as the chlorinating agent. When a flask containing mustard broke, Clarke suffered burns on his leg and consequently was hospitalized for nearly two months. He later made a report on his injuries to the German Chemical Society. A letter he wrote in 1947 describes the accident and his suspicion that the report spurred German use of this agent in World War I. (For more on Meyer and Clarke, see Senior [1958].)

The Germans Use Mustard

Two German chemists, W. Lommel at Bayer and Wilhelm Steinkopf, who worked under Haber at the Kaiser Wilhelm Institute, had advocated the use of mustard gas in the war as early as 1916. The German name for mustard, “Lost,” was formed from the first two letters of their last names. Haber also advocated its use—but only if the war could be won in a year, because he presciently feared massive Allied retaliation in kind.

The Germans launched their mustard attack on 12 July 1917. Mustard’s primary advantage over previous agents used in the war was its vesicant action. Since contact of mustard vapor or liquid with any part of the body resulted in painful, debilitating blisters, gas masks could no longer defeat this agent, which also penetrated most clothing. Mustard was also persistent and sticky. Soldiers feared not only every breath they took, but also every step and every branch or leaf they touched. The Germans had a poison gas that became known as the “king of war gases.”

All German production of mustard was based on the process developed by Meyer and modified by Clarke using ethylene chlorohydrin as the precursor compound. German companies had ample facilities to make this compound because it was manufactured on a large scale for use in the German dye industry. The Allies, on the contrary, although they quickly identified mustard’s chemical formula, did not have a strong dye industry and therefore were not ready to manufacture chlorohydrin in large quantities. They did not retaliate with mustard until almost a year after Germany’s first use: the French in June 1918, and the British not until September of that year, shortly before the war ended in November.

During that time massive, similar research and development efforts were undertaken in England, France, and the United States. Here we will discuss the U.S. effort primarily, because by the end of the war the United States was producing 30 tons of mustard per day—more than England, France, and Germany combined.

Scaling Up Mustard in the United States

The U.S. effort to produce mustard gas in large quantities was headed by an unsung hero of American chemical prowess: Frank M. Dorsey. When the war broke out, Dorsey was a chemical engineer at the National Lamps Works Company, a division of the General Electric Company in a suburb of Cleveland, Ohio. Once the United States declared war, Dorsey became involved with engineers from the nearby National Carbon Company in the design and production of gas masks for American soldiers (the National Carbon Company engineers were consulted because of their knowledge of charcoal, the filtering agent used in the masks). This led to the militarization of some of the engineers who worked on this project: Mr. Dorsey became Colonel Dorsey and head of the Development Division of the Chemical Warfare Service (CWS), which was initially organized under the Bureau of Mines but transferred to the U.S. Army in June 1918.

New poison gases were first studied at the CWS Research Division on the grounds of American University—the American University Experimental Station, or AUES—in Washington, D.C., which was set up in July 1917. Gases considered effective for battlefield use were then assigned to Dorsey’s division. Dorsey’s job was to find ways to scale up the small-scale production techniques developed in Washington so as to produce the agents in the thousands of tons necessary for use in war. The job of developing processes to manufacture mustard had actually been given to teams both in England and at the AUES. Sir William Pope, a well-known Cambridge University professor of chemistry, was primarily responsible for the English effort; he sent a cable to the CWS announcing the success of his technique in January 1918. In the United States, James Bryant Conant, a former Harvard University chemist now employed by the CWS, developed the same process virtually simultaneously: Pope’s cable arrived just as Conant was preparing to announce his own success. Both used essentially the same process as those used earlier by the 19th-century chemists; sulfur chloride and ethylene gas were the raw materials.

The project now fell to Dorsey, who established a small experimental mustard plant in Cleveland. In solving the problem of scaling up, he faced three challenges: he had to produce large quantities of ethylene efficiently; to design appropriate equipment to facilitate the absorption of ethylene by sulfur chloride; and to develop purification techniques.

To solve the first problem, Dorsey’s crew developed an ethylene generator—the Dorsey ethylene furnace—that used kaolin (clay) as a catalyst to produce ethylene from alcohol. Once perfected, 40 of these units were installed at Edgewood Arsenal in Maryland, enabling the United States to produce 30 tons of mustard per day by the time the armistice was signed. To solve the second problem, Dorsey’s unit developed a procedure in which dry ethylene gas was reacted with the sulfur chloride until absorption stopped. This procedure involved introducing steam as well as alcohol into the reaction, facilitating much better temperature control (for efficient ethylene absorption the reaction had to be kept at 131–140°F). A one-ton reactor based on this technique was designed in Cleveland for the mustard operation at Edgewood. For the third problem, Dorsey developed a procedure for flash-still distillation.

In August 1918 Dorsey’s crew worked on an improved mustard process that permitted manufacture at a lower temperature (86°F). This technique facilitated the removal of excess sulfur from the final product but needed refrigeration because the basic reaction was exothermic. Dorsey’s group nevertheless adapted this process for Edgewood, and it was being adopted when the armistice was signed on 11 November 1918. These reactors would have been capable of producing 100 tons per day by December 1918 and 200 tons per day by the following spring.

Lewisite: A “Better” Vesicant

The Allies were pleased to be able at last to retaliate against German offensive actions with mustard, but mustard itself was not an ideal offensive agent. It was typically not deadly, and it did not take effect immediately: soldiers exposed to it did not suffer symptoms until a few hours later. These characteristics were more suited to a defensive agent; the Allies were looking for a gas that would kill soldiers immediately and send the survivors into a panicked retreat.

Besides its main CWS unit, AUES, the Research Division of the CWS was using university laboratories throughout the country to develop new poison agents. One such unit, at Catholic University of America in Washington, D.C., was headed by Winford Lee Lewis, an associate professor of chemistry at Northwestern University before he volunteered for CWS research in Washington. One of the chemistry professors at Catholic University, John Griffin, had advised the department’s first doctoral student, Father Julius Arthur Nieuwland, who went to the University of Notre Dame after receiving his Ph.D. in 1904. Griffin one day suggested to Lewis that a passage in Nieuwland’s doctoral thesis might be useful to the CWS.

The passage in the thesis, “Some Reactions of Acetylene,” described the reaction of acetylene with arsenic trichloride:

“The contents of the flask turned black. When decomposed by pouring the substance into cold water, a black gummy mass separated out, and on standing for some time crystals appeared in the aqueous solution. The tarry substance possessed a most nauseating odor, and was extremely poisonous. Inhalation of the fumes, even in small quantity[,] caused nervous depression.”

Nieuwland did not report in his thesis that his exposure to the fumes sent him to the hospital for a few days. He did, however, remark that because of the compound’s poisonous nature, he would not investigate it further.

Lewis found Nieuwland’s thesis in the chemistry’s department library and liked what he read. He subsequently reproduced Nieuwland’s experiment and then tried to purify the mixture by distillation. Unfortunately, whenever he attempted it, the mixture exploded. Lewis finally sought help from his superior, none other than James Conant, who suggested he try pouring hydrogen chloride into the mixture before trying to distill it. This worked. Distillation revealed that the mixture was composed of three similar arsenic based compounds, which became known as lewisite 1, 2, and 3, after the number of acetylene molecules bonding with the arsenic trichloride.

Lewis’s unit tested the compounds and found that lewisite 1 was the most toxic. Very small amounts caused what appeared to be immediate pain in animals, followed by blister formation and often death. The CWS liked what Lewis told them of the compound and, after naming it after him, transferred further development to the AUES, where it was supervised by Conant.

Investigating Lewisite

Conant’s task was to shepherd lewisite through the various sections within the Research Division of the CWS. These sections investigated various methods of preparation, chemical properties, and physiological actions on animals and humans (as studied by the Pharmacological Section) of new agents.

A witness to the last process, Sergeant George Temple, was interviewed by American University’s student newspaper in 1965. Temple was in charge of maintenance for the 1,500 electric motors at the AUES. When a small drop of lewisite was applied to his forearm, his skin turned deep red and developed one-inch high blisters that did not heal for eight weeks. The scars on his forearm were still recognizable when he was interviewed in 1965, and he recalled that the silver-colored blisters were excruciatingly painful.

A profile of Conant in the New Yorker in 1936 referred to the accidents occurring during the development of lewisite at the AUES: “Pipes would frequently leak or vats would boil over. A vast tub of soapsuds awaited the frenzied plunges of men on whom the horrid stuff [lewisite] had settled.” Temple in the 1965 interview also described how “hundreds and hundreds” of stray dogs were gassed at the AUES as well as some monkeys. Soldiers tied the animals to stakes, exposed them to chemical bombs, and watched them struggle and usually die. The carcasses were then shaved and dissected to determine exactly how the gases affected the animals’ physiology.

The lewisite animal tests indicated that the first symptoms were blinking and tearing of the eyes followed by nasal secretion, retching, and vomiting. Next, the animals (generally dogs) salivated excessively, and their eyes became inflamed. Their nostrils clogged, and they coughed incessantly. Many died at this stage. If the dogs continued to survive, they sneezed violently with fluid continuously flowing from their nostrils. More dogs succumbed during this period. If an animal survived beyond the fifth day, it generally recovered by the tenth day.

When liquid lewisite was applied directly to the skin of dogs, they showed immediate irritation, very different from the delay following application of liquid mustard. Redness appeared in 4 to 6 hours and blisters in 16 to 48 hours, depending on the concentration. Non-lethal doses of liquid lewisite resulted in deep burns and death of skin cells. Lethal doses caused death in one to twelve days. Based on these tests, the pharmacologists concluded that a man of average weight (70 kilograms) would be killed by the application of about one-third teaspoon of lewisite to his skin. Since this evaluation, there has been much debate about how much lewisite would actually be required to kill a human, partially because susceptibility to the compound varies significantly among animals. (For details on this and other aspects of lewisite history not discussed here, see Dew of Death: The Story of Lewisite, America’s World War I Weapon of Mass Destruction.)

Scaling Up

Once the CWS was convinced of the merits of lewisite, in July 1918, the project and Conant were transferred to Willoughby, Ohio, where Dorsey had arranged for the construction of a top-secret plant to manufacture it. Conant had been ordered to have 3,000 tons of lewisite ready for a planned offensive against the Germans in the spring of 1919. At the time CWS researchers thought that lewisite would be the gas that fulfilled the initial beliefs about poison gas, that it would bring a quick end to the war. Conant and Dorsey worked frantically to design the plant and begin production of what was hoped would be approximately 50 tons per day. Amazingly, large-scale production was under way by early November. The plant was producing not only lewisite, but also its two precursors, arsenic trichloride and acetylene. Conant received a special commendation for accomplishing the impossible in making the plant operational in so short a period of time.

By the time the armistice was signed on 11 November, the plant was probably producing 10 tons of lewisite per day, and probably 150 tons had been produced by this date. (Because of the secrecy surrounding this operation, there are many conflicting accounts of these two figures.) After the war articles about lewisite “killing power” appeared in the popular press throughout the country, with one article boasting that ten airplanes carrying the “dew of death” would have “wiped out” all life in Berlin—a “weapon of mass destruction.” Fittingly, James Conant, who was instrumental in turning both mustard and lewisite into weapons, later became chair of the National Defense Research Committee during World War II, which oversaw the development of that war’s iconic weapon of mass destruction, the atomic bomb. Thus one Harvard chemist, Conant, supervised the production of the top-secret U.S. military projects of both world wars.

A Better Gas?

Whether the compound would have actually been effective in combat has been much debated since World War I. It has had little battlefield use, though the Japanese deployed it in China during World War II. One of lewisite’s drawbacks is that it degrades in the presence of moisture. Its supporters during World War I—and even during World War II, when the United States produced about 20,000 tons—did not believe that this hydrolysis would be a significant problem. Its detractors disagreed, and experiments during World War II confirmed that hydrolysis would indeed compromise lewisite’s effectiveness in most environments because of normal humidity. In fact, although the CWS believed during World War I that lewisite was unknown to the enemy, Haber’s scientists at the Kaiser Wilhelm Institute had evaluated it and concluded that hydrolysis would make it ineffective as a weapon. Nonetheless, some countries continued to produce lewisite until recently (e.g., the Soviet Union and probably Iraq), and North Korea may still be producing the agent.

The persistence of these poison agents developed during World War I in many arsenals today has its ironic side. After working flat out to match and to surpass German production, the United States now finds that its own cities potentially face attacks from terrorists using these same weapons, with the information on how to make them readily available and requiring only minimal chemical engineering skill.

For Further Reading

Büscher, Hermann. Green and Yellow Cross: Special Pathology and Therapy of Injuries Caused by the Chemical War Materials. . . . Trans. Nell Conway. Cincinnati: Kettering Laboratory of Applied Physiology, University of Cincinnati, 1944. Originally Grün- und Gelbkreuz (Leipzig: Barth, 1932).

Guthrie, Frederick. “On Some Derivations from the Olefins.” Quarterly Journal of the Chemical Society of London 13 (1860), 129–135.

Niemann, Alfred. “Über die Einwirkung des braunen Chlorschwefels auf Elaylgas” (On the effect of the brown chloride of sulfur on olefiant gas [ethylene]). Liebigs Annalen der Chemie und Pharmacie 113 (1860), 288–292.

Senior, James K. “The Manufacture of Mustard Gas in World War I.” Armed Forces Chemical Journal 12 (Sept./Oct. 1958), 12–14, 16–17, 29.

Tucker, Jonathan.War of Nerves: Chemical Warfare from World War I to al-Qaeda. New York: Pantheon, 2006.

Vilensky, Joel A., with Pandy R. Sinish. Dew of Death: The Story of Lewisite, America’s World War I Weapon of Mass Destruction. Bloomington: Indiana University Press, 2005.