Chapter 3: Toxic Agents
The Chemical Warfare Service came into existence because the armed forces needed a branch to deal with the problems arising from the use of poison gas, and although the service acquired the responsibility for other areas of warfare, such as incendiaries and smokes, its major concern during World War II remained the research, production, and neutralization of toxic agents. The first chemical used in World War I was chlorine, a heavy green gas. As the war progressed liquid and solid compounds were also used to launch chemical attacks.1
One of the first steps by the CWS just before World War II was to expand research on the classes of substances that might be suitable for toxic agents. In this program the National Defense Research Committee did much work.2 Soon after the committee came into existence in 1940, the CWS submitted to it six projects, four of which were concerned wholly or partially with toxic agents. To screen compounds synthesized by hundreds of chemists in universities and industry, the NDRC established in April 1941 a toxicity laboratory at the University of Chicago.3 In its four years of existence this laboratory screened about seventeen hundred compounds.4 The most promising of these, including sulphur fluorides,
nitrogen mustards, arsenicals, sulphur mustards, aromatic carbamates, fluoro-acetates, and aliphatic nitrosocarbamates were studied in more detail by the CWS at’ the Edgewood laboratories.
Of the vast number of compounds investigated, the CWS and NDRC found not one new standard agent. The difficulty lay in finding substances that met a large number of varying conditions. The compound had to be highly toxic so that a small amount would contaminate a large area. It had to be available in large quantities from the chemical industry, or it had to be of such a nature that it could be synthesized on a large scale at a reasonable price. It had to be stable during storage and not decompose into harmless materials. It was desirable that the density of the gas or vapor be heavier than air so that the compound would linger over the target. The vapor had to be nonflammable so that it would not be ignited by the flash of the burster.5 It could not react unduly with air or moisture. It could not corrode the container or evolve a gas that might burst the container. If it were a liquid, the freezing point had to be low, else it would freeze in a cold climate or in airplanes at high altitudes. Finally, the chemical, physical (i.e., color) and physiological (i.e., odor) properties had to be such that the enemy would be unable to detect the gas quickly and would have difficulty in providing protection.
These conditions were difficult to meet. Of the thousands of compounds considered by the CWS between 1917 and 1940, and by the CWS and NDRC during World War II, not one was found that could come up to the standbys of World War I. This was also the experience of Great Britain and the other Allied nations; and only the Germans through an accidental industrial discovery made while investigating insecticides, came upon a new group of agents, the so-called nerve gases or G-agents.6
In addition to seeking new agents the CWS spent much time improving the methods of preparing the standard agents and of overcoming such undesirable properties in the agents as instability. The service erected new plants using the improved processes at CWS arsenals and at other locations, and renovated older plants. It advanced the design of
chemical munitions, and obtained considerable information on their potential usefulness through exhaustive field tests. Arsenals and depots conducted large-scale surveillance tests to determine the storage life of agents. During World War II the CWS devoted most of the time spent on the research and development of toxics to the standard agents.
Phosgene
Phosgene, or carbonyl chloride (CWS symbol, CG), is a colorless liquid, slightly denser than water. It boils at 47° F. and hence in warm weather is in the form of vapor, unless under slight pressure as in a cylinder or shell. The vapor dissipates into the air in a few minutes, and for this reason CG is known as a nonpersistent agent. The vapor smells like green corn or new mown hay, and is extremely toxic. When inhaled, phosgene damages the capillaries in the lungs, allowing watery fluid to seep into the air cells. If the quantity inhaled is less than the lethal dose the injury is slight, the fluid is reabsorbed, the cell walls heal, and the patient eventually recovers; but if a large amount is inhaled, the air cells become flooded and the patient dies from lack of oxygen. It is difficult to estimate the severity of poisoning since the full effect is usually not apparent until three or four hours after exposure.
Phosgene was the second major agent to appear in World War I. The Germans first employed it in a cloud gas attack against the British in Flanders in December 1915 when 88 tons of the gas released from 4,000 cylinders caused more than 1,000 casualties.7 The Allies quickly adopted it and used it in enormous quantities throughout the war. It was an extremely dangerous agent, causing more than 80 percent of all chemical fatalities. After the war the CWS surveyed all of the nonpersistent agents, but could not find any that were more effective than phosgene.8 In 1928 the service classified CG as a substitute standard agent and in 1936 as a standard.9
In World War I Edgewood Arsenal and several chemical companies
produced phosgene for the AEF.10 The compound was prepared by combining chlorine and carbon monoxide in the presence of a catalyst. There were two methods of carrying out the reaction; in one concentrated carbon monoxide was used, in the other, dilute, The American plants adopted the concentrated gas process, but after the war the CWS weighed the relative merits of the two processes and concluded that the dilute gas method was more practical because it required simpler equipment that would be more readily available in an emergency.11 The CWS, however, was unable to construct a dilute gas plant because of lack of funds.
From 1922 onward the phosgene plant at Edgewood lay idle as the War Department forbade the manufacture of toxic agents. In 1937 the CWS rehabilitated and operated the plant for a brief period to produce phosgene and to provide the Technical Division with engineering data for a larger plant. The design was ready in 1939, and the new plant constructed and placed in operation in July 1941.12
Between 1940 and 1945 the CWS studied the manufacture of phosgene along four lines: improvement of the Edgewood process, pilot plant studies of the dilute gas process, erection of a by-product plant in Tennessee, and investigation of the diphosgene process,
Improvement of the Edgewood plant began in 1942 when the Technical Division carried out experiments that increased the efficiency of the process at an annual savings of $65,000.13 In 1944 the division established a pilot plant for further improvement of the process.14 The plant at Edgewood served as model for a plant of thirty tons’ capacity a day that the CWS erected at the Huntsville Arsenal and began operating in 1944.15
The concentrated gas process used at the Edgewood plant required solid carbon dioxide, pure oxygen, refrigeration equipment, and gas compressors, all classified as critical materials. In July 1942 the CWS
Development Laboratory at MIT began to investigate a dilute gas process using ordinary producer gas, containing from 15 to 20 percent carbon monoxide, in place of concentrated carbon monoxide. After eight months of labor the laboratory perfected a pilot plant capable of producing ten pounds of liquid phosgene an hour. Using the data obtained from the trials the laboratory drew up plans for a plant having a capacity of twenty-five tons a day and requiring for the most part standard industrial equipment. The CWS found it unnecessary to construct a dilute gas plant, but the plans were on hand for use in an emergency.16
In 1944 the CWS added yet another process. It erected a plant near Columbia, Tenn., to take advantage of the tremendous quantity of carbon monoxide available as a by-product from the Monsanto Chemical Co. phosphate works. This carbon monoxide contained impurities, particularly phosphorus and sulphur compounds, which had to be removed before the gas could be used. The Monsanto Co., under CWS contract, set up two pilot plants for the development of a large-scale method of purifying carbon monoxide and manufacturing phosgene. These pilot plants and those at Edgewood furnished the CWS with information for the design of a large plant with capacity of thirty-six tons a day. Construction began in May 1944, and the first phosgene was produced in February of the following year. Monsanto operated the process until the CWS closed the plant in April.17
The fourth and most unusual method of producing phosgene was based on the use of trichloromethyl chloroformate or diphosgene. This compound is less volatile than phosgene and is therefore less dangerous and troublesome to load into bombs and shells. By means of a catalyst it is quickly converted into phosgene. Taking these facts into consideration the CWS conceived the possibility of filling munitions with diphosgene and enclosing a catalyst which would convert the material into phosgene. In 1942 Morris S. Kharasch at the University of Chicago and in 1943 S. Temple at E. I. du Pont de Nemours & Co. investigated the reaction under NDRC contracts. Kharasch also studied the catalysts.18 The scheme
appeared practicable, but the CWS finally decided that the advantages of the method did not compensate for the higher cost of diphosgene and the changes that would have been necessary in the design of munitions.
During the war the CWS manufactured and purchased from industry more than forty million pounds of phosgene for use in various munitions.19 Two munitions for phosgene, the chemical mortar shell and the portable cylinder, had descended from World War I. In the event of gas warfare, the mortar would have been the chief weapon of the ground forces for laying down concentrations of phosgene on caves, dugouts, bunkers, and artillery and machine gun emplacements. From 1941 to 1944 the CWS filled almost half a million 4.2-inch mortar shells with CG. Each shell held almost seven pounds of CG, about 25 percent of the total weight of the filled munition.
The cylinder had been a standard weapon in the static trench warfare of World War I, but it was scarcely suited for the blitz tactics of World War II. It could have been used, however, to overcome resistance within caves or bunkers on Japanese-held islands. It contained thirty-one pounds of phosgene, about 56 percent of the total weight. The cylinder also held about two pounds of carbon dioxide to expel the phosgene in the form of a mist. In view of the possible employment of cylinders, the service retained the final model M1A2, standardized in 1936, until World War II was over.20
New phosgene weapons were the 7.5-inch rocket, the AN-M78 500-pound bomb, and the AN-M79 1000-pound bomb. The rocket, which was the World War II counterpart of the World War I Livens projectile, was readied by 1944. The Navy took almost eight thousand of these, the Army more than twenty-three thousand.
Development of phosgene bombs started in early 1942 when the CWS asked the Ordnance Department for a series of chemical bombs of approximately the same shape as general purpose bombs. The new munitions were produced in 1943 and sent to Dugway Proving Ground for testing and evaluation.21 The 1000-pound bomb holding 415 pounds of CG
turned out to be an extremely effective munition. When it hit the ground and burst open a large amount of liquid phosgene was freed. The evaporating liquid cooled the vapor and caused it to flatten out against the ground in a pancake-shaped cloud instead of rising as had been expected. This cloud always formed, regardless of the weather. The 500-pound bomb containing 205 pounds of CG was not quite half as effective as the 1000-pound bomb, but it was still a useful munition because American planes could carry more than twice as many 500-pound bombs as 1000-pound bombs. The CWS filled twenty-five thousand 500-pound bombs in 1944, and sixty-three thousand 1000-pound bombs from 1943 to 1945. The Air Forces could, in case of chemical warfare, have used these chemical bombs against targets beyond mortar range, against fortifications on Iwo Jima and other islands before amphibious assaults were made, and against strategic targets, such as war plants during working hours.
After the war examination of stocks of gas weapons captured in Germany showed that the German Army had on hand thousands of 250-and 500-kilogram phosgene bombs.22 These bombs, however, had been largely superseded by bombs containing the nerve gas tabun, which the Germans began producing in 1942.23 The Germans did not favor the use of phosgene in shells. Italy had phosgene bombs, and shells ranging in size from 149-mm. to 305-mm.24 Phosgene shells, from 75-mm. up to 150-mm. were captured from the Japanese, who also had bombs in sizes up to 200 kilograms.25
Had gas warfare started early in World War II, phosgene would probably have been used widely by the Allied and the Axis armies wherever the tactical situation called for the employment of a nonpersistent, delayed-action agent. Sometime in 1942 or thereafter, evidence indicates that as a stockpile accumulated the Germans would have introduced tabun, and phosgene would then have had to share the field with the new nerve gas.
Hydrogen Cyanide
At the battle of the Somme in July 1916 French artillery fired shells filled with hydrogen cyanide (CWS symbol, AC).26 The compound had
been familiar to chemists for a century but this was the first time it was used in warfare.27 It is a colorless liquid which evaporates quickly at room temperature and boils at 78° F. The liquid and vapor interfere with normal processes in body cells, particularly in the respiratory center of the nervous system, and if present in more than a certain small concentration quickly causes death. But if cyanide is present in less than the lethal concentration the cells can convert it into a harmless compound and the body is uninjured. In this respect AC is different from phosgene, mustard, and other toxic agents which are harmful even when present in less than the lethal dose.
The French had some difficulty in using hydrogen cyanide as an agent because AC vapor is light and therefore has a tendency to diffuse instead of lying close to the ground. Also, AC has a tendency to decompose—sometimes so violently that the container exploded.28
In an attempt to cut down the rate of diffusion the French mixed AC with stannic chloride. To prevent AC from decomposing the French added arsenic trichloride. To keep the mixture from crystallizing and to make soldiers more susceptible to the agent they added chloroform. The addition of these compounds diluted the AC so much that the final mixture contained only 50 percent of the cyanide. This meant that twice as many shells, or shells with twice the capacity, were needed to deliver the same weight of the cyanide—a rather wasteful procedure.
In addition to employing a dilute agent the French used small shells holding only about a pound of filling. Furthermore, their artillery fired at a slow rate. As a result the French were not able to place a lethal concentration of gas on an enemy area. Other nations observed the apparent failure of hydrogen cyanide and came to the conclusion that it was not suitable as a war gas, but the French never lost faith in it and continued to use it until the end of the war.
Despite its drawbacks, hydrogen cyanide was inexpensive, commercially available, and had several of the other properties that have been mentioned as being necessary or desirable for a toxic agent. After the war the opinion gained ground in the CWS that the agent had not been given a fair trial.29 In the 1930s chemists made laboratory and field studies,
including firing tests with 155-mm. howitzer shells, and came to the conclusion that the compound was potentially an effective lethal, nonpersistent agent.30 They overcame the old problem of decomposition with the aid of Du Pont and American Cyanamid, manufacturers of hydrogen cyanide, who gave information which finally enabled the CWS to stabilize the cyanide in munitions.31
After the United States entered World War II the CWS extended its work and tested AC bombs of the 100-, 115-, 1000-, and 2000-pound size. The 1000-pound bomb, holding approximately 200 pounds of agent, proved particularly suitable as a munition. With this large quantity of the cyanide the cooling effect brought about by evaporation of the liquid produced a cloud of gas whose density was greater than air and which hovered close to the ground. Under favorable meteorological conditions the cloud was fatal hundreds of yards from the point of impact.32
The bomb was unquestionably an efficient munition for use in a cyanide gas attack, but the tests uncovered a serious problem. The vapor which billowed outward from the bomb was easily ignited by the flash of the burster. In some tests, practically all the bombs caught fire as they split open. There were three ways of preventing the burning of AC: one was to devise a “cold” bursting charge that would not ignite the vapor, the second was to use a more powerful bursting charge that would push the vapor cloud away from the bomb faster than the flame could follow, and the third was to add a substance that would make AC more difficult to ignite. Since the first two methods would have required too much time and field work, the third was followed. Anton B. Burg and his associates conducted the research under an NDRC contract at the University of Southern California. They discovered that hydrocarbons such as those in gasoline were the best flame inhibitors. Dugway Proving Ground tested AC protected with hydrocarbons and found that it did not take fire as readily as pure AC, but bombs still burned occasionally, and the problem was never completely solved.33
The 1000-pound bomb would have been the chief means of dumping hydrogen cyanide on the enemy if gas had been used in the latter part of the war. It was standardized for use with AC in 1943, and about 5,000
bombs were filled and stored.34 This munition accounted for almost all of the 1,132,000 pounds of AC procured by the CWS from July 1940 to the end of 1945.35
An unusual AC weapon was a glass bottle holding about a pint of liquid.36 This grenade was produced from 1942 onward as a possible last-ditch weapon against tanks or in overcoming bunkers. It was finally dropped from the approved munitions in 1944 because of the danger of breakage during shipment, either through accident or enemy action, and because tests had proven that it would not always break on soft jungle underbrush or if it glanced off log bunkers.37
In view of the fact that the Germans did not regard hydrogen cyanide as highly as some other agents, they did not procure large quantities or fill shells, bombs, or grenades. They did, however, think that AC might be useful in the form of a spray, and the Luftwaffe carried out extensive field trials with aerial spray tanks.
The Japanese, on the other hand, felt as the Americans did about the value of AC, but they planned to use it in shells and grenades rather than in bombs. Their AC munitions ranged from mortar shells—light, medium, and heavy—to 150-mm. howitzer shells. Japanese glass grenades containing hydrogen cyanide were captured on Guadalcanal, in Burma, and on the upper Chindwin River.
Hydrogen cyanide was not as important as some of the other toxic agents, but if gas warfare had broken out, both sides would certainly have employed it in tactical situations where its rapid action and lack of persistence would have been of advantage to the attacking force.
Cyanogen Chloride
Cyanogen chloride (CWS symbol, CK) is a colorless liquid slightly denser than water.38 It boils at a temperature of 55° F., giving off a vapor which is approximately twice as dense as air and which irritates the eyes
and nasal passages. When air containing a high concentration of the vapor is inhaled the compound quickly paralyzes the nervous system and causes death. When a low concentration is inhaled the reaction is not so rapid, but the compound accumulates in the body until a lethal concentration is reached.
Cyanogen chloride was first used as a toxic agent by the French in October 1916. In 1917 and 1918 the CWS investigated the manufacture, the chemical, physical, and physiological properties, and the effectiveness in shells and Livens projectiles of cyanogen chloride.39 The Research Division found that the gas passed rapidly through the German but not through the American mask. This was an important discovery and might have led to the adoption of the compound as an American chemical warfare agent had not the density of the vapor been so low that the CWS felt it was impossible to place a lethal concentration of cyanogen chloride on enemy positions.40 The same decision, apparently, was also reached by the French and other European armies, for cyanogen chloride was never used to any extent.
Between the wars the CWS conducted few trials with CK. The compound’s chief test came in 1933 when the Technical Division, searching for an agent that would act more rapidly than phosgene, the standard non-persistent agent, examined CK and decided it was not acceptable.41 But early in World War II the CWS, while examining captured Japanese and German masks, obtained data that indicated that CK would penetrate enemy canisters in harassing concentrations if the humidity of the air was high—a condition common to the tropics.42 This discovery opened the way for the adoption of CK as a standard agent. As a prelude to standardization technicians had to learn if a lethal concentration could be laid down over enemy positions, to see if CK was available in quantities sufficient for military use, to find means of overcoming the instability of the compound, and to modify the canister of the mask for greater protection to American soldiers.
The CWS and NDRC assessed CK at Dugway Proving Ground in
February and March 1943.43 Testers placed a 100-pound chemical bomb containing 67 pounds of CK and a 500-pound bomb containing 280 pounds of CK in shallow craters and split them open with tetryl bursters. They estimated the strength of the gas cloud by means of vapor sampling devices and goats placed downwind from the burst. The trials showed that the 500-pound bomb released a low-hanging cloud that was lethal for a considerable distance and that the flash from a tetryl burster would not ignite the compound.
Even though CK was shown to be suitable as an agent, the CWS still might not have standardized it if the protective properties of the American mask had not been improved. The mask carried by soldiers at the start of the war gave excellent protection against chloropicrin, phosgene, mustard, and lewisite but only fair protection against hydrogen cyanide, and cyanogen chloride. The CWS in 1943 adopted Type ASC charcoal, treated with chromium, which was more effective in removing CK. Thus at the time when the investigators were uncovering evidence of the usefulness of CK on the offense, the technicians were developing better protection for defense.
Another hurdle that remained was the chemical instability of cyanogen chloride which had a tendency to polymerize. That is, the short molecules of the compound would join together spontaneously to form large molecules of a new compound. Sometimes the reaction took place so rapidly that the container exploded. Polymerization within bombs or shells also meant a wastage of the munition, since the new compounds were relatively harmless as agents.
The task of preventing or retarding polymerization was undertaken by Division 10 of the NDRC in 1942. A group of chemists headed by Wendell M. Latimer of the University of California made a preliminary search for stabilizing compounds. Later researchers of American Cyanamid Co., working under CWS contract, took up the quest and uncovered additional information. Dugway Proving Ground contributed to these studies by setting up a large-scale surveillance test of munitions filled with CK. In August 1943 the NDRC started an additional experimental program under Anton B. Burg of the University of Southern California. Burg’s group ran nearly two thousand tests on cyanogen chloride. This work expanded the knowledge of the chemistry of CK, particularly the reactions which took place during storage, but still did not provide the complete
answer. In 1944 Division 9 of the NDRC entered the field with a group of men under Kharasch of the University of Chicago. This group observed the retarding power of inorganic compounds on the polymerization of CK and finally found that a small amount of sodium pyrophosphate would preserve CK under normal storage conditions for many years. From then on sodium pyrophosphate was used to stabilize CK.
In order to obtain sufficient CK for chemical munitions, the CWS had to erect a plant. Before the war the only plant in the country was owned by the American Cyanamid Co. at Warners, N.Y. This plant produced sufficient cyanogen chloride for industry, but could not turn out the large quantity needed for chemical warfare. In October 1943 the War Department approved the construction of a CWS plant with a capacity of fifteen tons, later increased to sixty tons, per day.44 The Chemical Construction Co. broke ground for the “Owl” plant, as it was called, on 27 November at a site adjacent to the American Cyanamid Co.’s hydrogen cyanide plant at Azusa, Calif. This location thus assured the “Owl” plant with the hydrogen cyanide needed in the process. American Cyanamid, which operated the plant under contract, started the first unit in April 1944.
The CWS chose two types of munitions for cyanogen chloride-4.2-inch mortar shells and bombs. The mortar shell was made the official CK munition for ground forces in 1945, but was not filled. Instead, almost all of the twenty-five million pounds of CK procured by the CWS went into 33,347 M78 500-pound bombs, each holding 165 pounds of agent, and 55,851 M79 1000-pound bombs, each holding 332 pounds.45
Cyanogen chloride bombs, in event of chemical warfare, would probably have been used early against the Japanese, particularly in the tropics, where the humidity would have assisted the vapor in passing through the canister. The soldier then would have been forced to tear off his mask, exposing himself to other lethal agents dropped simultaneously. In time the Japanese and Germans could have treated the charcoal in such a way that CK would no longer pass through their canisters. The agent would then have lost its chief usefulness as a war gas.
Mustard Gas
In World War I the protection experts on each side tried to devise means of neutralizing enemy agents as soon as new agents appeared.
Chlorine, the first gas used, was soon parried by an adequate mask. As new gases appeared, the masks were improved. Soon the mask furnished full protection and men were gassed only when they were careless, panicky, or caught by surprise. But in July 1917 the German Army brought out a new type of agent, mustard gas, that not only attacked the respiratory system but also the skin, soaking through clothes and shoes and raising painful blisters. It was almost impossible to shield soldiers completely against mustard. It became the king of battle gases and caused four hundred thousand casualties before the armistice.46
Crude mustard gas (CWS symbol, H) was a mixture of approximately 70 percent β,β´-dichloroethyl sulfide and 30 percent of sulphur and other sulphur compounds. It was an oily, brown liquid that evaporated slowly, giving off a vapor five times heavier than air. It was almost odorless in ordinary field concentrations but smelled like garlic or mustard in high concentrations—hence the name. It irritated and poisoned body cells, but generally several hours passed before symptoms appeared.
The chief problem concerning mustard had to do with its purification. In World War I the CWS adopted the Levinstein process of the English in which ethylene reacted with sulphur monochloride under carefully controlled conditions.47 The reaction at first glance seems simple, but actually it was rather complex and defied the efforts of CWS chemists to chart its course. The impurities were of such a nature that they could not be isolated and analyzed. They resisted separation from the main ingredient, β,β´-dichloroethyl sulfide, and caused or hastened decomposition of the sulfide. Decomposition was a disadvantage, first, because some of the resulting products corroded the storage container, making storage unsafe; secondly, other products settled out as a sludge that could change the ballistic properties of shells or prevent the liquid from dispersing in the most favorable pattern; thirdly, a gas was evolved which built up pressure and threatened to burst containers; and, later, after airplane spray tanks were devised, the decomposition products made it impossible to thicken mustard for use in airplane spray attacks.
Chemists of the research and development division investigated methods of purifying mustard, but the processes proved to be impractical for large-scale use.48 After the armistice the CWS disposed of the mustard
plants at Cleveland, Ohio, Buffalo, N.Y., Midland, Mich., and Hastings-on-Hudson, N.Y., and closed the plant at Edgewood. Research on mustard practically ceased until the early 1930s when the plant at Edgewood was restored. In 1937 this plant was put into production for a two-week period but not until 1940 was it opened for large-scale production.49
After Edgewood Arsenal began producing mustard again the CWS, assisted by the NDRC, examined a number of purification methods including distillation under low pressure, distillation using steam and organic liquids, extraction with solvents, treatment with ammonia, flash distillation, and crystal fractionation. Of these processes only vacuum distillation, steam distillation, and solvent extraction proved to be feasible for use on a large scale.
Purification by extraction dated back to 1918 when the CWS carried out laboratory and pilot plant investigations to see if β,β´-dichloroethyl sulfide could be separated from impurities by dissolving it in gasoline or other solvents. The insoluble impurities remained in the residue and the sulfide was recovered from the solvent by distillation.50 In 1942 this line of research was resumed at the CWS-MIT Development Laboratory. The chemists first obtained data on the solubility of the constituents of crude mustard in various solvents, and on rates of solution. Then, using glass extraction apparatus, they determined the data necessary for designing a large-scale extractor.51 The NDRC assisted by awarding a contract to the Texas Co. for pilot plant studies. Texas Co. engineers proved that large-scale extraction was practical, but they found that the product was less pure than steam distilled mustard and that the process required complex, expensive equipment.
Steam distillation, in which a current of steam was passed into the still to help carry away mustard, leaving the impurities behind as a tarry residue, had also been tested by the CWS back in 1918. In 1943 the CWS-MIT Development Laboratory re-examined this method and found that it produced a sulfide of high purity and fair stability, and that only simple equipment was required.52 The Texas Co. then made a pilot plant
investigation to obtain data for the construction and operation of a full size plant.53 It is possible that this process would have been the one utilized, as it seemed the most promising at the time, had not the CWS and NDRC come across a superior method, vacuum distillation.
The CWS obtained the clue which led them to vacuum distillation in November 1943 when Capt. J. W. Eastes visited the University of Illinois to confer with NDRC chemists. He learned that they had distilled at low pressure mustard which had been washed with water, and that the temperatures in the distillation column indicated that fairly pure β,β´-dichloroethyl sulfide could be prepared in this way.54 In other words, water removed certain impurities, and distillation removed the remainder. The CWS had investigated vacuum distillation earlier, but had never washed the crude mustard before distilling.55 The Technical Division investigated the process and found that it produced a purer and more stable β,β´-dichloroethyl sulfide than the other methods and that it was quite practical so far as apparatus was concerned. A pilot plant was first set up and then a full-scale plant.56 In 1945 the service switched to the new process at Edgewood and at Rocky Mountain Arsenal.57 By the end of the year 9,218,357 pounds of distilled mustard (symbol, HD) had been produced. With the successful production of HD, production of the old Levinstein mustard was halted.
Mustard, in terms of the quantity that the CWS stockpiled, was the most important American toxic agent. The plants at the Edgewood, Huntsville, Pine Bluff, and Rocky Mountain arsenals produced 174,610,000 pounds, exclusive of the nine million pounds of the new distilled mustard.58
Since mustard evaporated slowly and thus remained effective from several hours to several days, depending upon the weather and terrain, its use was indicated on strategic targets or on enemy positions that would not be taken immediately by American troops. Thus, it could be used to “seal off” an enemy area into which American troops were advancing, and
to hamper enemy lines of communication, airfields, landing beaches, artillery emplacements, and observation points. In withdrawals it could be used to contaminate the routes of enemy advance.
For delivery of mustard by ground troops the CWS had 4.2-inch mortar shells, artillery shells, and land mines. The land mines were simply rectangular 1-gallon tin cans, such as were commonly used to hold varnish or syrup. They had a capacity of ten pounds of mustard. When exploded with a slow-burning fuze or by electrical means, the mines spread mustard over a considerable area. They were intended for use as booby traps or in contaminating fields, roads, and buildings. The CWS procured and stored (but did not fill) almost two million such mines.59 For possible use by troops, 540,746 4.2-inch mortar shells were filled and stored. For the artillery, 1,360,338 75-mm. Mk 64, 1,983,945 105-mm. M60, 784,836 155-mm. Mk 2A1, 290,810 155-mm. M110, and smaller quantities of other shells, were readied.60
For carrying out aerial mustard attacks the CWS had chemical bombs and spray tanks.61 The service procured 594,216 M70 and M70A1 115-pound bombs, developed by the Ordnance Department, and 539,727 M47A1 and M47A2 100-pound bombs, developed by the CWS in the 1930s.62 The bombs were slightly over 4 feet long, about 8 inches in diameter, and contained a cylindrical burster. The bombs held from 60 to 70 pounds of mustard, and when dropped contaminated an area of from 15 to 40 yards in diameter, depending upon the altitude of the plane, hardness of the ground, thickness of vegetation, and so on.63
In addition to bombs the service procured 92,337 M10 30-gallon airplane spray tanks. A plane flying at an altitude of 100 feet and carrying four of these tanks could spray mustard over an area 75 to 80 yards wide and 600 to 700 yards long. A larger tank, the M33 or M33A1, of which the service obtained 20,598, held more than twice as much mustard. A plane carrying two of these tanks could contaminate an area 75 to 100 yards wide and 700 yards long.64
In anticipation of the use of spray tanks the CWS expended much effort in trying to improve the spraying properties of mustard. In the 1930s the CWS had accepted the doctrine that mustard spray attacks would be
carried out by planes flying at low altitudes and moderate speeds. By 1941 plans called for planes flying a mile high and at speeds up to 350 miles an hour. At high altitudes and speeds the wind could easily carry small droplets beyond the target or spread them over too wide an area. Small droplets also evaporated so quickly that they either might not reach the ground at all, or else become so minute as to be practically ineffective. To obtain the desired large droplets chemists began to search for materials which would thicken mustard.65
After starting the project CWS learned that the British had already determined the best size for high altitude droplets and were adding various substances to mustard to increase the particle size. In cooperation with the NDRC the CWS tested more than seventy thickeners.66 Finally, the search narrowed down to polystyrene and methyl methacrylate. After methyl methacrylate sheet scrap (Plexiglas and Lucite) became available from aircraft factories, the CWS adopted it as a mustard thickener.67
As things turned out the work of the CWS and NDRC on thickeners went for naught. High and low altitude spray tests carried out by the CWS in cooperation with the Signal Corps and Army Air Forces at Dugway Proving Ground from 1943 onward finally proved that unthickened mustard was a better substance for spraying purposes than thickened mustard, and thickening agents were given up.68
Like the American Army, the German Army placed much reliance on mustard. An examination of captured documents and gas dumps showed that they had produced more than twice as much mustard as any other agent for use in artillery shells of all calibers, mortar shells, 250- and 500-kilogram bombs, rockets, and spray tanks.69 A notable feature was the tendency to use mustard in conjunction with thickening agents or with substances that would lower the freezing point. Arsenol, a mixture of arsenic compounds, mainly diphenylchloroarsine, was widely used for this purpose.
The Japanese, too, used mustard as a filling for shells and bombs. They
favored a 50-50 mixture of mustard and lewisite, the lewisite acting to lower the freezing point and also as a toxic agent in its own right.
In all probability if toxics had been called upon in World War II, mustard would have been used extensively whenever tactics pointed to the need of a persistent chemical agent.
Lewisite
In 1918 a group of organic chemists headed by Dr. Winford Lee Lewis prepared a highly vesicant substance, dichloro (2-chlorovinyl) arsine, which they named lewisite.70 The CWS leased the old Ben Hur Automobile Co. building at Willoughby, Ohio, installed equipment, and began to produce the agent.71 A shipment was on the seas headed for Europe when the war ended. The CWS kept the existence of lewisite and the site of its manufacture a strict secret during the war, but later revealed the information in scientific journals.72 After the armistice the service closed down the Willoughby plant and did not prepare the compound again except in laboratory quantities until 1941.
In the early method of manufacture, acetylene and arsenic trichloride were combined with the aid of a catalyst, aluminum chloride. The process was complicated, a large quantity of unwanted by-products were formed, and sometimes the crude product exploded as it was being distilled. In the early 1920s the CWS renewed its research on lewisite, but was unable to continue the investigation to any great length because of the small staff and projects of higher priority.73 In 1939 the service set out to design a pilot plant that would produce lewisite by a continuous process using the old aluminum chloride catalyst. Shortly thereafter reports from Great Britain told of the successful use of mercuric chloride as a catalyst.74
The CWS checked this work, found that the new catalyst was an improvement, and adopted it.75 But since the mercuric chloride was a batch process, unlike the aluminum chloride process which had been continuous, engineers had to modify the design of the pilot plant.76 Furthermore, with the new catalyst there was considerable corrosion of equipment.77 As a result of the problems attending the change in process the production of lewisite was held up until the end of 1942, when plants were opened at Huntsville and Pine Bluff Arsenals.78 In 1943 a larger plant was started at Rocky Mountain Arsenal.79
While the CWS was erecting and starting plants, evidence was accumulating that lewisite might have only limited use. The service had no World War I data to use in evaluating lewisite since the war ended before the agent reached France. The information gained from field tests between 1920 and 1940 was not sufficient for World War II.80 To obtain additional data the CWS conducted toxicological and field tests.81 Results indicated that lewisite was of less value than had been supposed because there was difficulty in setting up a high concentration in the field, the gas mask gave complete protection against the vapor, the vapor had a distinctive odor that made it readily recognizable, and the agent could be readily decontaminated. In addition, British chemists had come upon a powerful therapeutic agent, DTH or BAL (British Antilewisite), that destroyed lewisite on contact.82
Consideration of all these facts led the CWS to close the lewisite plants in 1943, after 20,000 tons had been produced.83 Because of the possible utility of lewisite under certain limited conditions, a supply was retained
throughout the war. Afterwards the CWS sank a large quantity at sea, and finally abandoned the agent completely.84
As has been mentioned, lewisite, because of its ability to lower the freezing point of mustard (which was only 58° F.), was used in the form of lewisite-mustard mixtures by the Japanese. The Russians also employed lewisite for this purpose. The Germans were familiar with L-H mixtures for cold weather, but they preferred to use other arsenical liquids in place of lewisite.
Nitrogen Mustards
In 1935 there appeared an article by Kyle Ward, Jr., describing the preparation of a new compound, 2,2’,2” trichlorotriethylamine, and calling attention to its marked vesicant action.85 The CWS prepared and studied a sample of the substance, but found that it was less vesicant than mustard.86 Early in World War II the CWS learned through intelligence that the Germans were working with the same compound and with related compounds—which by now had gained the name of the “Nitrogen Mustards” because of their analogy to mustard gas.87
These reports led the CWS, NDRC, and British laboratories to synthesize and test a large number of nitrogen mustards. Three compounds known as HN-1 (N-ethyl(2,2´dichloro)diethylamine), HN-2 (N-methyl-(2,2´dichloro)diethylamine), and HN-3 (2,2´,2″ trichlorotriethylamine) were most promising because of their vesicant action and their lack of odor, and these the CWS studied intensively.
The British concentrated mainly on HN-2 and HN-3, but the Americans felt that HN-1 would be the most useful. Edgewood Arsenal set up a small pilot plant in 1942. Using data obtained from this plant, the CWS in 1943 set up a larger plant at Pine Bluff Arsenal capable of producing one ton per day.
In the meantime field evaluation showed that the 1936 estimate had been correct, and that the nitrogen mustards were not as potent as mustard gas. The plant at Pine Bluff turned out about 100 tons of HN-1 over
a period of four months, mainly to mislead German intelligence, and then closed down.88
The Germans had much more faith in nitrogen mustards than the Americans, and during the war turned out about 2,000 tons of HN-3. At the end of the war they had on hand 105-mm. and 150-mm. artillery shells, and 150-mm. rockets, filled with HN-3.89
Chloroacetophenone
Both sides used tear gas early in World War I to harass opposing troops. Troops exposed to tear gas had to wear masks for long periods of time and were very uncomfortable in the old-fashioned, heavy, bulky devices.
During the war the French, Germans, and British introduced a greater variety of tear gases than any other class of agents. After the United States entered the conflict, American chemists investigated chloroacetophenone (CWS symbol, CN) and found that it had the advantage of being cheaper and less corrosive to the inside of shells than other tear gases. The CWS developed methods of producing the agent, but the war ended before large-scale manufacture got underway.
After the armistice the service selected chloroacetophenone as the standard American tear gas. It erected a manufacturing plant at Edgewood Arsenal (1922), and developed a number of munitions for dispersing solid CN or solutions of CN in the field. The solid could be scattered from shells and grenades by means of high explosives, and volatilized from pots and candles by means of heat. Solutions of CN in chloroform (CNC), with chloropicrin in chloroform (CNS), and in carbon tetrachloride and benzene (CNB) could be thrown out by grenades, shells, and high pressure cylinders.
In 1941 the CWS erected a modern plant with a rated capacity of one ton of CN per day. This became the sole CWS plant in 1943 when the old 1922 plant was dismantled.90 The manufacture of chloroacetophenone involved three steps: the production of monochloroacetic acid, the chlorination of the acid to chloroacetyl chloride, and condensation of chloroacetyl chloride with benzene in the presence of a catalyst. The CWS was aware of another method that was potentially capable of being adapted to large-scale manufacture, the chlorination of acetophenone. If this could be
done satisfactorily the service was assured of a dependable source of low-cost CN.
In 1944 the Solvoy Process Co. contracted to develop the process and obtain data necessary for construction and operation of a plant. As events turned out the new plant was not needed, but the technical information was at hand in case of an emergency.91
In addition to the development work on CN the CWS tested new compounds for their lacrimatory effect. The Universal Oil Products Co. and Du Pont suggested other possibilities. But none of the new tear producing chemicals proved superior to the standard agent.
During the war the CWS produced at Edgewood Arsenal and purchased from the Pennsylvania Salt Manufacturing Co. and the Lake Erie Chemical Co. a total of 1,281,560 pounds of chloroacetophenone.92 A portion of this went to make up 5,282,000 pounds of CNB tear gas solution, another portion went into 3,309,000 pounds of CNS solution.93 Almost all of this solution was stored, but some was used to fill 4.2-inch chemical mortar shells, and 75-mm., 105-mm., and 155-mm. artillery shells,
Solid chloroacetophenone went into pots and grenades. The tear gas pot was a modified version of the ordinary tin can and was filled with 1.2 pounds of a CN-powder mixture. When the pot was ignited by a match-head, heat from the burning powder volatilized the CN. The pot gave off CN smoke for several minutes. Edgewood Arsenal filled a total of 785,383 pots for possible use in the war.94
The tear gas grenade was one of the first munitions developed by the CWS after World War I. The body was a small tin can having six holes to let out smoke. The filling contained CN and a powder that provided the heat to volatilize the CN. During World War II, the CWS filled 689,610 M7 grenades, each containing about six-tenths of a pound of tear gas mixture, at the Edgewood and Huntsville Arsenals.95
Since the CWS had learned back in the 1920s that tear gas grenades were not enough to drive determined men from their posts, they adopted the practice of adding another agent that would cause vomiting and other reactions. Gradually they perfected a mixture of chloroacetophenone and adamsite (CWS symbol, DM), as a filling for grenades. Adamsite caused
nausea, pain in the chest, sneezing, coughing, headache, and other disorders. It had another advantage—it acted so rapidly that the victims were unable to pick up the grenades and throw them back, as they did occasionally with ordinary tear gas grenades. Between 1941 and 1944, the Edgewood and Huntsville Arsenals filled 582,327 M6 grenades, each holding about six-tenths of a pound of chloroacetophenone-adamsite mixture.96
In 1943 the Provost Marshal General requested the CWS to develop a tear gas grenade with the size, shape, and weight of a baseball for military police to use in breaking up riots. The finished grenade, standardized in February 1945, was a plastic ball holding about two-tenths of a pound of CN, fused to burst 2-3 seconds after it left the hand and before it could be picked up by a rioter and thrown back. The service produced more than 10,000 of these riot grenades by the end of the year.97
Tear gas rockets, for possible use as antitank missiles, were first investigated by the CWS in 1942 with assistance from the California Institute of Technology and the National Bureau of Standards. While this work was in progress the Ordnance Department developed and standardized the antitank high explosive rocket, M6. The CWS thereupon turned to the Ordnance rocket and developed heads to carry chemical agents. The antitank tear gas rocket was finally dropped, but the idea took a new turn in 1943 when the Provost Marshal General’s office requested a rocket for use, like the tear gas grenade, in controlling riots. The CWS modified the rocket head to meet the new requirements, but the problems associated with ballistics, bursters, and size proved so difficult that in 1944 the project was canceled.
The Germans and Japanese, like the Americans, used CN. The Germans employed it in two forms, one being solid CN, while the other was a mixture of CN, wax, and explosive. The explosive mixture was used in 250-kilogram and 500-kilogram air burst bombs, from which CN was dispersed in lumps.98 The Japanese had on hand an unusual tear gas candle containing a propellant charge capable of tossing a one-quarter pound chunk of CN a distance of 130 to 300 yards, depending on the angle of elevation. For closer quarters a grenade containing a solution of CN in carbon tetrachloride was available.99
In tactics, tear gases would probably not have been as useful in World
War II as they were in 1914-18. Positions did not remain static as they had in World War I, and the opportunities for harassment would not have been as great. In certain cases, however, such as attacks upon Japanese caves and bunkers, or upon isolated positions, in the Pacific Islands, the gases might have brought about surrender or have driven the enemy into the open.
Adamsite
The German Army introduced vomiting gases or sternutators into chemical warfare in July 1917, as an ingenious method of penetrating the canisters of Allied gas masks. They first used a solution of diphenylchloro-arsine (CWS symbol, DA), which evaporated and left minute particles of DA floating in the air. The canisters at that time were able to trap true gases, the particles of which were molecular in size, but they could not retain the larger particles of DA, which were colloidal in size. Therefore the DA passed through the canister into the mask and was inhaled by the soldier. It irritated his eyes, nostrils, throat, and chest, causing nausea and vomiting. The victim had to tear off his mask, exposing himself to lethal gases fired at the same time.
After the United States entered the war, American chemists investigated the possibility of manufacturing DA. The German process proved to be complicated. Still, the CWS might have gone into production if chemists had not found a related compound that could be manufactured more easily. This was diphenylaminechloroarsine, which was named adamsite after the chemist Roger Adams.
The United States did not produce vomiting gas in time for use by American troops. In the 1920s the CWS operated, for a brief time, a pilot plant for the production of DA and DM at Edgewood Arsenal, but it did not need a full size plant since DM could be purchased from the chemical industry.
The development work, instead, concentrated on means for spreading sternutators in the field. This could be done by explosion, which shattered the agent into a dust or mist, or by heat, which produced smoke. Engineers tested irritant smoke shells, ranging in size from 75-mm. to 105-mm., filled with solid agents and high explosives (HE), or with solutions of agents and HE, up to 1942, but the service did not produce any for combat use. Smoke candles, which were simply cans filled with a mixture of agent and fuel, proved to be much more efficient for dispersing DM. The early candle, dating from 1920–22, was displaced in 1941 by a
new model, M2, which differed in having a better fuel. The M2 weighed 9 pounds, held 2 pounds of DM, and burned from three to five minutes. Edgewood Arsenal filled 92,485 candles during the war, with a portion of the 644,589 pounds of DM purchased by the CWS.100
The Germans employed DM as a filling for base ejection and HE shells, candles, and bombs; and DA solution as a filling for rockets.101 The Japanese relied upon another arsenical vomiting gas, diphenylcyanoarsine (DC), as a filling for mortar shells, artillery shells, and candles.102
Undoubtedly vomiting gases would have found much less use in World War II than they had in 1917-18 because the canister of the gas mask had been improved by the addition of filters which held back fine particles. In certain limited situations, such as attacks upon isolated posts or upon surrounded caves or bunkers, the vomiting agents might have been employed to bring about surrender or weaken resistance.
An investigation of the status of chemical warfare within Germany made after V-E Day disclosed that it had produced a total of approximately 78,000 tons of agents—mustard, tabun, arsenol, chloroacetophenone, phosgene, adamsite, nitrogen mustard, and diphenylchloroarsine—during the Hitlerian period.103 In addition, the Germans had appreciable stocks of Italian, French, Greek, Polish, Hungarian, and Yugoslavian agents at their disposal. The quantity of war gases produced by Japan was placed at one-tenth of the German production. American production, on the other hand, amounted to more than 146,000 tons of chemicals—mustard, chloro-acetophenone, phosgene, adamsite, nitrogen mustard, hydrogen cyanide, cyanogen chloride, and lewisite—from 1940 to the end of 1945.104
Although the United States did not employ toxic agents during World War II, the money and time that went into the research, development, field tests, and production was not wasted. The armed forces had supplies of agents and equipment with which they could have waged warfare energetically if necessary. In this sense the work of the CWS was America’s insurance against chemical warfare.