TRIPHENYLARSINE AND DIPHENYLARSENIOIJS SALTS ]L 447 CLX.- Triphenylarsine and Dipheszylursenious Salts. By WILLIAM JACKSON POPE and EUSTACE EBENEZER TUBNEB. THE intro'duction of aromatic arsenic compounds as materials for chemical warfare rendered necessary the working out of satis-factory methods for preparing triphenylarsine (C,H,),As and diphenylchloroarsine (C,H,),AsCl. The former of these two' corn-pounds is conveniently prepared by Michaelis and Reess's method (Ber. 1882 15 2876) which consists in treating an ethereal solution of arsenic trichloride and bromobenzene with sodium ; i 1448 POPE AND TURNBBi a later paper (Ber. 1886 19 1031) Philips showed that chloro-benzene can be substituted for bromobenzene and that the reac-tion is stimulated by the addition of a small proportion of ethyl acetate but more recently (Annulen 1902 321 160) Michaelis claimed that a cleaner product is obtained by the use of bromo-benzene than of chlorobenzene.The several workers on this reaction used ether as a solvent, and itl was important to ascertain whether this could be replaced by some less volatile and less inflammable diluent. Preliminary experiments showed that the reaction proceeds better in benzene than in ethereal solution and that contrary to Michaelis’s sugges-tion chlorobenzene gives a cleaner product than does bromo-benzene. It was thus shown that triphenylarsine is readily p r e pared by the action of sodium on a mixture of arsenic trichloride and chlorobenzene in benzene solution to which a little ethyl acetate had been added; this method was described in a report to the Chemical Warfare Department dated January 28th 1918 and formed the basis of the larger-scale work done by Morgan and others (this vol.p. 777) in the conversion of the laboratory method into a works process. It was observed that the use of ether as a diluent is dis-advantageous in that the violent reaction tends to pass out of control and that when this happens pyrophoric sodium remains after the evaporation of the ether causing dangerous fires. The repetition of the method of Philips gave a yield of 71 per cent. of the theoretical; Michaelis’s later method gave a yield of 67 per cent,. of the theoretical. In the absence of a diluent sodium acts on a mixture of arsenic trichloride and chlorobenzene causing incandescence.Experiments were next carried out for the purpose of ascertain-ing how the yield of triphenylarsine is influenced by the propor-tions of the reacting materials and by the conditions. The general method adopted was to weigh outl the sodium in slices (s) granules (g) powder ( p ) or wire into a large flask cover with benzene containing I or 2 per cent of ethyl acetate allow to remain for half an hour to activate the metal and then slowly run in the arsenic trichloride and chlorobenzene. After a few minutes, a vigorous reaction sets in which when sodium wire is used must be controlled by the use of a freezing mixture; when sliced sodium is used no external cooling is necessary and indeed once the reaction is checked by cooling it can only be started again with considerable difficulty.The mixture is then left overnight, filtered and the inorganic residue well washed with hot benzene; the filtrate and washing are distilled until a thermometer place TRIPHENY LARSJNE AND DIPHENYLARSENIOUS SALTS. 1449 in the liquid registers 200c’. crystalline mass and is almost pure triphenylarsine ; in general it melts a t above 56O. The following table gives the results of a series of experiments, in each of which 136 grams of chlorobenzene were used; this quantity requires theoretically 60 grams of arsenic trichloride and 46 grams of sodium for complete conversion into triphenylarsine. The residue solidifies on cooling to No. of experi -ment. I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 It) 19 20 21 Volume of benzene.C.C. 400 7 9 9 9 7 9 9 9 9 7 . 7 9 9 9 9 1 7 ) G O 350 9 9 7 1 7 9 9 , 9 9 , Sodium. Grams. 68 7 7 9 7 9 9 7 7 7 9 3 . 1 . 9 9 69 G.3 60 60 67 1 9 9 9 55.5 9 9 9 9 ASCl,. Grams. 72 7 7 7 9 9 , 9 7 9 9 i b 85 90 80 9 9 $0 $0 85 85 90 95 90 Percentage of theoretical yield of C,H,Cl used. 67.5 69 66 64 62.5 71 83 91.5 95.5 93.5 80 75 83.5 81 88 93 89 80 83 74.5 84 Experiments 1 to 3 were carried o u t just as described above; 4 and 5 were carried out as rapidly as possible and the product was worked up immediately without remaining overnight. In numbers 6 and 7 the product was gently boiled under reflux for several hours after the spontaneous reaction had ended.It is thus shown that the yield is not improved by hastening the reac-tion but that itl does increase when the reaction is continued further by boiling; the operation of boiling under reflux was there-fore introduced in all the later experiments. A consideration of the further experiments shows that the yield of triphenylarsine, calculated on the amount of chlorobenzene used is appreciably raised by increasing the amount of arsenic trichloride to 85 grams, and is practically unaffected by diminishing the weight of sodium to 57 grams; it appears further that the volume of benzene used can be reduced to 300 C.C. without ill-effect on the yield. The most satisfactory results seem to be obtained by using 300 C.C.of benzene for each 136 grams of chlorobenzene 85 grams of arsenic trichloride and 57 grams of sodium but since the VOL. CXVIT. 3 1450 POPE AND TURNER: reaction proceeds well with considerable fluctuations of the pro-portions the appropriate quantities of materials to be used depend on the relative cost of the latter. The reaction between chlorobenzene and arsenic trichloride is not promoted by boiling with copper or aluminium powder the copper-zinc couple or magnesium or calcium turnings. The copper arsenide obtained by digesting a hydrochloric acid solution c,f arsenic trichloride collecting the black powder washing i t with water and acetone and drying is without action on chlorobenzene, but when heated with iodobenzene yields diphenyl.Conversion of Triphetiylnrsine into Di- and Jfono-phenylarsine Derivatives. Michaelis and Reese showed (Bcr. 1882 15 2876) that phenyl-arsenious dichloride is produced on heating triphenylarsine with arsenic trichloride under pressure; it is to be concluded that diphenylarsenious chloride is formed as an intermediate stage and particulars have been given by Morgan and Vining (this vol., p. 780) of a convenient means for preparing diphenylarsenious chloride by heating triphenylarsine with arsenic trichloride a t 250-280° under pressure. This method involves however the use of an autoclave and i t seemed of interest to ascertain whether the same reaction could be carried out under the ordinary atmo-spheric pressure.Triphenylarsine (30.6 grams) was maintained a t 350° while, arsenic trichloride (25.5 c.c.) was very slowly run in by means of a long capillary tube; the arsenic trichloride which distilled over was returned to the reaction vessel. The first addition of the arsenic trichloride occupied one and threequarter hours. On carefully distilling the product under 12-15 mm. pressure the following fractions were obtained a t above 1200 ( a ) 120-160°, 17.7 grams; ( b ) 160-200° 22.2 grams; ( c ) 200-250° 2.2 grams; (d) residue 2.2 grams. The fraction (a) is fairly pure phenyl-arsenious dichloride (ij) is pure diphenylarsenious chloride whilst ( c ) and (d) consist of nearly pure triphenylarsine; allowing for the recovery of the latter the yield of phenylarsenious dichloride and diphenylarsenious chloride is 97 per cent.of the theoretical. In the experiment just described the arsenic trichloride was added fairly rapidly and another may be quoted to show the effect of running it in more slowly. Using the same quantities as before but taking seven hours for the addition of the arsenic trichloride the following fractions were obtained on distilling the product under 12-15 mm. pressure ( a ) 120-160° 1 2 grams o TRIPHENYLARSINE AND DIPHENYLARSENIOUS SALTS. 1451 moderately pure pheiiylarsenious dichloride; ( b ) 160-205° 31.5 grams of practically pure diphenylarsenious chloride ; and a resi-due (c) of 7.2 grams of impure triphenylarsine. I n this case a larger proportion of diphenylarsenious chloride was produced. The general conclusion is drawn from the above and other experiments that under atmospheric pressure the following reac-tions occur (C,H,)3As + 2AsC13= 3(C6H5)AsClz and ~(C,H,)~AS + AsCl,= 3(C,H5)2AsCl.I n addition to the foregoing the reaction represented by the following equation may also occur: and experiment showed this reaction to take place almost quantitatively. On heating a mixture of triphenylarsine (15.3 grams) and phenylarsenious dichloride (11-2 grams) for four hours a t 300° in an open flask and distilling the resulting pasty mass under diminished pressure nearly pure diphenylarsenious chloride (20 grams) distilled a t 185O/15 mm.; this corresponds with a yield of about 80 per cent. of the theoretical; under the conditions stated a small amount of chlorobenzene was produced and practically no action occurs a t 250'.It is thus prsved that the reaction between arsenic trichloride and triphenylarsine proceeds very satisfactorily under atmospheric pressure and that the product is an equilibrium mixture resulting from the simultaneous occurrence of several reactions. In view of the possible importance of the observation that triphenylarsine can 'be converted into phenylarsenious dichloride and diphenyl-arsenious chloride by the action of arsenic trichloride under the ordinary pressure the above and analogous reactions were pro-tected by secret Patent No. 142880 of June l l t h 1918 of which the specification has now been published. (C,H,)AsC& + (CdHj)&s= 2(C,H,)&sCl, Dip h e uylnrs e nious Bromide (C,H&AsBr.Diphenylarsenious oxide [(C6H,)2As]20 was prepared by Michaelis and La Coste (AnnaZen 1880 201 229) but the follow-ing is a more expeditious method for obtaining it in a pure state. Potassium hydroxide (12 grams) dissolved in water (10 c.c.) is added to rectified spirit (200 c.c.) ; a solution of diphenylarsenious chloride (53 grams) in spirit (100 c.c.) is added and the mixture boiled for an hour. The solvents are then distilled off and the solid residue is extracted with chloroform ; on drying filtering, and evaporating the extract a quantitative yield of pure diphenyl-arsenious oxide remains as a colourless crystalline solid melting a t 8 9-9 lo. 3 a 1452 TRIPHENYLARSINE AND DIPHENYLARSENIOUS SALTS. On heating the oxide at looo with hydrobromic acid in a sealed tube and allowing to cool diphenylarsenious bromide (C&€,),AsBr, separates as a colourless crystalline solid melting a t 55-56O; this compound is described by Michaelis and La Coste as a yellow oily liquid.The bromide is also obtained by heating triphenylarsine (30.6 grams) with arsenic tribromide (15.8 grams) for three hours a t 300-350O; on distilling the product under 14 mm. pressure, the following fractions resulted below 170° 2 grams of a mixture of benzene and bromobenzene; 170-205O 26 grams of crude diphenylarsenious bromide and a residue of 15 grams of mixed diphenylarsenious bromide and triphenylarsine. By redistillation, pure diphenylarsenious bromide was readily obtained. Diphenylarsenious Iodide (Diphenpliodoarsine) (C&I&AsI. This previously undescribed substance is obtained by heating diphenylarsenious oxide (25 grams) with fuming hydriodic acid (30 grams) for two hours in a sealed tube a t looo; on cooling the crude iodide (29.5 grams) solidifies and melts at 4 2 4 5 O . On crystallisation from benzene the compound is obtained in yellow, crystalline scales melting a t 4 5 - 4 6 " (Found I = 35.6. C,,H,,IAs requires I=35*6 per cent.). On heating triphenylarsine (30.6 grams) with arsenic tri-iodide (22.8' grams) for six hours in an open flask a t 350-360° and dis-tilling the resulting mass under diminished pressure practically pure diphenylarsenious iodide (25 grams) distils a t 204-218O/ 10 mm.; the yield is less than 50 per cent. of the theoretical and the reaction does not proceed so satisfactorily as in the case of the corresponding bromo-derivative. The work described in the present paper was carried out for the purposes of the Chemical Warfare Department and permission for its publication has been given by the General Staff. THE CHEMICAL LABORATORY, UNIVEREITY OF CAMBRIDGE. [Received October 16th 1920.