CHAPTER X

ARSENIC DERIVATIVES

Since arsenic is well known as an insecticide in the form of lead arsenate, arsenic acid etc., and in pharmacy, specially in the form of salvarsan and neosalvarsan, it is not surprising that the Germans should have endeavored to discover an arsenic derivative which would be of value from the point of view of chemical warfare. Very early in the war persistent rumors were circulated that the Germans were to use arsine. These rumors led to the use of sodium permanganate in the canister, but as far as is known, no arsine was actually used. Another suggestion which received considerable attention from American workers was the use of arsenides, which might decompose under the influence of the atmospheric moisture with the liberation of arsine. Calculation of the amount of arsenide necessary to establish a lethal concentration of arsine showed, however, that there was no possibility of using the material on the field.

Because of the use of arsenic trichloride in the manufacture of organic arsenic compounds, a method of preparation was developed from arsenic trioxide and sulfur chloride or hydrogen chloride. It was also shown experimentally that the phosgene of the tail gas of phosgene plants might be converted into arsenic trichloride by reaction with arsenic trioxide. Charcoal is the catalyzer of this reaction.

Arsenic trichloride is also of interest because it was one of the constituents of the mixture vincennite, early used by the French. This was a mixture of hydrocyanic acid, stannic chloride, arsenic trichloride and chloroform. While extensively used at first, it was gradually replaced by phosgene.

Arsenic triflouride was also prepared by the action of sulfuric acid upon a mixture of calcium flouride and arsenic trioxide. The compound is very easily decomposed by the moisture of the air, and furthermore is not very toxic.

Organic arsenic derivatives are the most important compounds from the military point of view. The first substance used was diphylchloroarsine, a white solid, which readily penetrated the canister and caused sneezing. This was used alone, and in solution in phenyl dichloroarsine. Later methyl and ethyl dichloroarsines were introduced.

[Illustration: FIG. 38.—Apparatus for the Manufacture of Methyldichloroarsine.]

METHYLDICHLOROARSINE

The Germans apparently used ethyldichloroarsine because they had no suitable method for the preparation of methyl dichloroarsine, which is a more satisfactory material. The Chemical Warfare Service developed the following method of preparation of the methyl derivative. Sodium arsenite (Na₃AsO₃) is prepared by dissolving arsenic trioxide in sodium hydroxide solution. The action of methyl sulfate at 850 C. gives disodium methyl arsenite, Na₂CH₃AsO₃. Sulfur dioxide reduces the arsenite to methyl arsine oxide, CH₃AsO, which is then reacted with hydrochloric acid to give methyl dichloroarsine. The final product is distilled from the mixture and condensed. This material costs from two to two and a half dollars per pound for chemicals (war prices).

Methyldichloroarsine is a colorless liquid of powerful burning odor, which boils at 132° C. It is somewhat soluble in water and is soluble in organic solvents. The specific gravity is 1.838 at 20° C. The vapor pressure at 25° was found to be 10.83 mm. mercury. Not only is the material toxic but it has remarkable vesicant properties, comparing favorably with mustard gas in this respect.

Ethyldichloroarsine, which was used by the Germans, was prepared by the method given above, using ethyl sulfate, but the yield was never over 20 per cent. In general this has properties similar to the methyl derivative.

DIPHENYLCHLOROARSINE

The best known of the arsenicals, however, is diphenylchloroarsine or sneezing gas. Although this is an old compound (having been prepared by German chemists in 1885), there was no method for its preparation on a large scale when, first introduced into chemical warfare. It was finally discovered that the interaction of triphenyl arsine with arsenic trichloride was fairly satisfactory and a plant was erected for its manufacture.

When pure, diphenylchloroarsine is a colorless solid, melting at 44°. Because of this, it was always used in solution in a toxic gas or in a shell which contained a large amount of explosive so that on the opening of the shell the material would be finely divided and scattered over a wide territory.

Its value lay in the fact that the fine particles readily penetrated the ordinary mask and caused the irritation of the nose and throat, which resulted in sneezing. This necessitated the perfection of special smoke filters to remove the particles, after which the other toxic materials were removed by the absorbent in the canister.

It causes sneezing and severe burning sensations in the nose, throat and lungs in concentrations as slight as 1 part in 10 million. In higher concentrations, say 1 in 200 to 500 thousand it causes severe vomiting. While neither of these effects are dangerous or very lasting, still higher concentrations are serious, as in equal concentrations diphenylchloroarsine is more poisonous than phosgene.

Various other arsenical chemicals were developed in the laboratory, but with one or two exceptions they were not as valuable as diphenylchloroarsine and methyldichloroarsine and were therefore discarded.

GERMAN METHODS FOR MANUFACTURING ARSENICALS[23]

DIPHENYLCHLOROARSINE

[Footnote 23: Norris, _J. Ind. Eng. Chem._, =11=, 825 (1919).]

“This substance (Blue Cross) was a famous gas of the Germans and was made in large quantities. The method used by the Germans was different from the one worked out by the Allies, and on account of the fact that the German method could be carried out without specially designed apparatus and required as raw materials substances readily obtainable, it was probably preferable. It is doubtful, however, whether the Allies would have made this gas, for as the result of its use no fatalities were reported. The German process consisted in preparing phenylarsenic acid by condensing benzene diazonium chloride with sodium arsenite. The acid was next reduced by sulfur dioxide to phenylarsenous acid, which was, in turn, condensed with the diazonium compound to form diphenylarsenic acid. This acid was reduced to diphenylarsenous oxide, which with hydrochloric acid yielded diphenylchloroarsine. The chemical equations for the reactions will make clearer the steps involved.

C₆H₅N₂Cl + Na₃AsO₃ = C₆H₅AsO₃Na₂ + NaCl + N₂ C₆H₅AsO₃Na₂ + 2HCl = C₆H₅AsO₃H₂ + 2NaCl C₆H₅AsO₃H₂ + SO₂+H₂O = C₆H₅AsO₂H₂ + H₂SO₄ C₆H₅N₂Cl + C₆H₅AsO₂Na₂ = (C₆H₅)₂AsO₂Na + NaCl + N₂ (C₆H₅)₂AsO₂Na + HCl = (C₆H₅)₂AsO₂H + NaCl 2(C₆H₅)₂AsO₂H + 2SO₂ + H₂O = [(C₆H₅)₂As]₂O + 2H₂SO₄ [(C₆H₅)₂As]₂O + 2HCl = 2(C₆H₅)₂AsCl + H₂O.

“The entire process was carried out at Höchst. The method used at Höchst was as follows: In preparing the diazonium solution, 3 kg.-mols of aniline were dissolved in 3000 liters of water and the theoretical quantity of hydrochloric acid. The temperature of the solution was reduced to between 0° and 5° and the theoretical amount of sodium nitrite added. The reaction was carried out in a wooden tank of the usual form for the preparation of diazonium compounds. A solution of sodium arsenite was prepared which contained 20 per cent excess of oxide over that required to react with the aniline used. The arsenous oxide was dissolved in sodium carbonate, care being taken to have enough of the alkali present to neutralize all of the acid present in the solution of the diazonium salt. To the solution of the sodium arsenite were added 20 kg. of copper sulfate dissolved in water, this being the amount required when 3 kg.-mols of aniline are used. The solution of the diazonium compound was allowed to flow slowly into the solution of the arsenite while the temperature was maintained at 15°. The mixture was constantly stirred during the addition which requires about 3 hrs. After the reaction was complete, the material was passed through a filter press in order to remove the coupling agent and the tar which had been formed. Hydrochloric acid was next added to the clear solution to precipitate phenylarsenic acid, the last portions of which were removed by the addition of salt.

“The phenylarsenic acid was next reduced to phenylarsenous acid by means of a solution of sodium bisulfite, about 20 per cent excess of the latter over the theoretical amount being used. For 100 parts of arsenic acid, 400 parts of solution were used. The reaction was carried out in a wooden vessel and the mixture stirred during the entire operation. A temperature of 80° was maintained by means of a steam coil. Phenylarsenous acid separated as an oil. The aqueous solution was decanted from the oil, which was dissolved in a solution of sodium hydroxide, 40° Bé. The solution of the sodium salt of phenylarsenous acid was treated with water so that the resulting solution had a volume of 6 cu. m. when 3 kg.-mols of the salt were present. Ice was next added to reduce the temperature to 15° and a solution of benzene diazonium chloride, prepared in the manner described for the first operation, was slowly added. After the coupling, diphenylarsenic acid was precipitated by means of hydrochloric acid. The acid was removed by means of a filter press and dissolved in hydrochloric acid, 20° Bé. For one part of diphenylarsenic acid, 3 parts of hydrochloric acid were used. Into this solution was passed 5 per cent excess of sulfur dioxide over that required for the reduction. The sulfur dioxide used was obtained from cylinders which contained it in liquid condition.

“The reduction was carried out in an iron tank lined with tiles and a temperature of 80° was maintained. About 8 hrs. were required for the reaction. The diphenylarsenic acid on reduction by the sulfur dioxide was converted into diphenylarsenous oxide which, in the presence of the hydrochloric acid, was converted into diphenylchloroarsine, which separated as an oil. The oil was next removed and heated in the best vacuum obtainable until it was dry and free from hydrochloric acid. The compound melted at 34°. It was placed in iron tanks for shipment. The yield of diphenylchloroarsine calculated from the aniline used was from 25 to 30 per cent of the theoretical. No marked trouble was observed in handling the materials and no serious poisoning cases were reported.

DIPHENYLCYANOARSINE

“This compound was prepared by treating diphenylchloroarsine with a saturated aqueous solution of potassium or sodium cyanide.

(C₆H₅)₂AsCl + NaCN = (C₆H₅)₂AsCN + NaCl.

Five per cent excess of the alkaline cyanide was used. The reaction was carried out at 60° with vigorous stirring. The yield was nearly theoretical.

ETHYLDICHLOROARSINE

“This compound was prepared at Höchst from ethylarsenous oxide which was obtained from the Badische Anilin und Soda Fabrik.

“PREPARATION OF ETHYLARSENOUS OXIDE—The compound was prepared by treating sodium arsenite with ethyl chloride under pressure. The resulting sodium salt of ethylarsenic acid was converted into the free acid and reduced by sulfur dioxide. The ethylarsenous acid formed in this way lost water and was thereby transformed into ethylarsenous oxide. The reactions involved are as follows:

C₂H₅Cl + Na₃AsO₃ = C₂H₅AsO₃Na₂ + NaCl C₂H₅AsO₃Na₂ + 2 HCl = C₂H₅AsO₃H₂ + 2 NaCl C₂H₅AsO₃H₂ + SO₂ + H₂O = C₂H₅AsO₂H₂ + H₂SO₄ 2C₂H₅AsO₂H₂ = (C₂H₅As)₂O + H₂O.

“The ethyl chloride used in the preparation was in part made in this factory, and in part received from other sources. As ethyl chloride is an important product used in peace time, it is not, therefore, essentially a war product and its preparation was not described.

“In preparing the solution of sodium arsenite, one molecular weight of arsenous oxide was dissolved in a solution containing 8 molecular weights of sodium hydroxide. The solution of the base was prepared from a 50 per cent solution of sodium hydroxide to which enough solid alkali was added to make the solution a 55 per cent one. In one operation 660 kg. of arsenous oxide were used. For 100 parts of arsenous oxide, 130 parts of ethyl chloride were used, this being the theoretical amount of the latter.

“The reaction was carried out in a steel autoclave of about 300 liters capacity. The temperature was maintained at between 90° and 95°. The ethyl chloride was pumped in, in 3 or 4 portions, and the pressure in the autoclave was kept at from 10 to 15 atmospheres. The several portions of ethyl chloride were introduced at intervals of about 1½ hrs. During the entire reaction, the contents of the autoclave were vigorously stirred. After all the ethyl chloride had been added, the material was stirred from 12 to 16 hrs., at the end of which time the pressure had fallen to about 6 atmospheres. The excess of ethyl chloride and the alcohol formed in the reaction were next distilled off. At this point a sample of the solution was drawn off for testing. This was done by determining the amount of arsenite present in the solution. If not more than 20 per cent of sodium arsenite had not reacted, the preparation was considered satisfactory. Water was then added to the contents of the autoclave in sufficient amount to dissolve the solid material. The product was next drawn over into a tank and neutralized with sulfuric acid. It was then treated with sulfur dioxide gas until there was an excess of the latter present. The mixture was then heated to about 70° when the ethylarsenous oxide precipitated as a heavy oil. This was readily separated and shipped without further purification. The yield of ethylarsenous oxide, from arsenic oxide, was from 80 to 82 per cent of a product which contained about 93 per cent of pure ethylarsenous oxide.

“PREPARATION OF ETHYLDICHLOROARSINE—The compound was prepared by treating ethylarsenous oxide with hydrochloric acid. The reaction is as follows:

C₂H₅AsO + 2HCl = C₂H₅AsCl + H₂O.

The operation was carried out in an iron kettle lined with lead, which was cooled externally by means of water and which was furnished with a lead covered stirrer. To the kettle, which contained from 500 to 1000 kg. of hydrochloric acid left over from the previous operation, were added 4000 kg. of ethylarsenous oxide. The gaseous hydrochloric acid was next led in. The kettle was kept under slightly diminished pressure in order to assist in the introduction of hydrochloric acid. The temperature during the reaction must not rise above 95°. When the hydrochloric acid was no longer absorbed and was contained in appreciable quantities in the issuing gases, the operation was stopped. This usually occurred at the end of from one to two days. The product of the reaction was drawn off by means of a water pump and heated in a vacuum until drops of oil passed over. The residue was passed over to a measuring tank and finally to tank-wagons made of iron. The yield of the product was practically the theoretical.

On account of the volatility of the compound and its poisonous character, the apparatus in which it was prepared was surrounded by an octagonal box, the sides of which were fitted with glass windows. Through this chamber a constant supply of air was drawn. This was led into a chimney where the poisonous vapors were burned. The gases given off during the distillation of the product were passed through a water scrubber.”

“LEWISITE”

The one arsenical which created the most discussion during the War, and about which many wild stories were circulated, was “Lewisite,” or as the press called it, “Methyl.” Its discovery and perfection illustrate the possibilities of research as applied to Chemical Warfare, and points to the need of a permanent organization to carry on such work when the pressure of the situation does not demand such immediate results.

The reaction of ethylene and sulfur chloride, which led to the preparation of mustard gas, naturally led the organic chemists to investigate the reaction of this gas and other unsaturated hydrocarbons, such as acetylene, upon other inorganic chlorides, such as arsenic, antimony and tin. There was little absorption of the gas, either at atmospheric or higher pressures, and upon distilling the reaction product, most of the gas was evolved, showing that no chemical reaction had taken place. However, when a catalyser, in the form of aluminium chloride, was added, Capt. Lewis found that there was a vigorous reaction and that a highly vesicant product was formed. The possibilities of this compound were immediately recognized and the greatest secrecy was maintained regarding all the details of preparation and of the properties of this new product. At the close of the War, this was considered one of the most valuable of Chemical Warfare secrets, and therefore publication of the reactions involved were withheld. Unfortunately or otherwise, the British later decided to release the material for publication, and details may be found in an article by Green and Price in the _Journal of the Chemical Society_ for April, 1921. It must be emphasized that the credit for this work belongs, not to these authors, but to Capt. W. Lee Lewis and the men who worked with him at the Catholic University branch of the American University Division (the Research Division of the C. W. S.).

On a laboratory scale, acetylene is bubbled through a mixture of 440 grams of anhydrous arsenic trichloride and 300 grams of anhydrous aluminium chloride. Absorption is rapid and much heat is developed. After six hours, about 100 grams of acetylene is absorbed. The reaction product was dark colored and viscid, and had developed a very powerful odor, suggestive of pelargoniums. Attempts to distill this product always led to violent explosions. (It may be stated here that Lewis was able to perfect a method of distillation and separation of the products formed, so that pure materials could be obtained, with little or no danger of explosion.) The English chemists therefore decomposed the product with ice-cold hydrochloric acid solution of constant boiling point (this suggestion was the result of work done by Lewis). The resulting oil was then distilled in a current of vapor obtained from constant boiling hydrochloric acid and finally fractionated into three parts.

The first product obtained consist in the addition of one acetylene to the arsenic trichloride molecule, and, chemically, is chlorovinyldichloroarsine, CHCl: CH·AsCl₂, a colorless or faintly yellow liquid, boiling at 93° at a pressure of 26 mm. A small quantity, even in very dilute solution, applied to the skin causes painful blistering, its virulence in this respect approaching that of mustard gas. It is more valuable than mustard gas, however, in that it is absorbed through the skin, and as stated on page 23, three drops, placed on the abdomen of a rat, will cause death in from one to three hours. It is also a very powerful respiratory irritant, the mucous membrane of the nose being attacked and violent sneezing induced. More prolonged exposure leads to severe pain in the throat and chest.

The second fraction (β, β′-dichlorodivinylchloroarsine) is a product resulting from the addition of two acetylene molecules to one arsenic trichloride, and boils at 130° to 133° at 26 mm. It is much less powerful as a vesicant than chlorovinyldichloroarsine, but its irritant properties on the respiratory system are much more intense.

The third fraction, β, β′, β″-trichlorotrivinylarsine, (CHCl: CH)₃As, is a colorless liquid, boiling at 151° to 155° at 28 mm., which solidifies at 3° to 4°. It is neither a strong vesicating agent nor a powerful respiratory irritant. At the same time, its odor is pungent and most unpleasant and it induces violent sneezing.

##