26 OECHSLIN. 



slightly alkaline reaction; any uncombined acid would therefore remain in 

 solution. A slight excess of benzarsinic acid over phosphorus pentachloride is 

 necessary when beginning the reaction, because phosphorus pentachloride reacts 

 with quinine, forming chlorquinine, which latter is precipitated, together with 

 the ester, from the alcoholic solution by means of ether, and it requires a special 

 operation to separate the two. 



However, even if an excess of benzarsinic acid is present the products prepared 

 according to this method always contained noticeable quantities of chlorine. No 

 chlorine was formd in the filtrate upon dissolving in dilute acid and reprecipitat- 

 ing; so that substitution in the organic portion of the molecule had taken place, 

 probably because of the intermediate action of the tetrachlorarsinic portion of 

 the original product of the reaction. 



In view of the above difficulties another method was employed. Benz- 

 arsinic acid was first reduced to benzarsinious acid, so that no tetra- 

 chlorarsinic derivative would be produced while the acid chloride was 

 being formed. After a number of experiments with different reducing 

 agents such as SO2 or SnClj, and oxidizing agents for the subsequent 

 oxidation, the following method was found to give satisfactory results. 



Benzarsinic acid was suspended in chloroform and reduced to benzarsinic 

 dichloride by means of phosphorus trichloride." Both the benzarsinic acid and 

 the chloroform were dried carefully, since the presence of moisture caused the 

 reduction in part at least to proceed farther than to the dichloride. If great 

 care in this respect was not taken, a deep yellow product resulted, whereas it was 

 colorless when the substances were previously dried. 



La Coste believed it to be probable that the acid chloride of benzarsinic 

 dichloride ( AsClj • CsHj ■ COCl ) is formed, although analytical proofs, which would 

 have required the difficult isolation of the compound, were not given. However, 

 the product of the reaction does not combine with quinine when warmed with a 

 chloroform solution of the latter, but does so only after it has reacted with 

 phosphorus pentachloride, which proves that phosphorus trichloride has acted on 

 the acid only sufficiently to form the dichlorarsenic derivative of the earboxyacid 

 COOH • CeH, ■ AsCL. The same holds true if the reduction is not carried out in 

 chloroform, but in the presence of phosphorus oxychloride or by means of pure 

 phosphorus trichloride. 



After the above reduction in chloroform suspension by means of phosphorus 

 trichloride had taken place, the solution was evaporated to dryness in vacuo on 

 a water bath, by which means the excess of phosphorus trichloride was removed. 

 Chloroform was added to the residue and then phosphorus pentachloride ( 1-gram 

 molecule ) . A spontaneous reaction set in, hydrochloric acid was developed 

 and the residue went into solution. The whole was heated until the evolution 

 of gas ceased and then cooled. A chloroform-quinine solution was thereupon 

 added and the products again heated for one-half hour. The acid, aqueous so- 

 lution of the reaction product, prepared in the same manner as was given for the 

 benzarsinic acid ester, when neutralized gives a dense precipitate, which, differing 

 from the arsinic-acid ester, does not dissolve in an excess of sodium carbonate or 

 sodium hydroxide. It was filtered and washed, and since it still gave a strong 

 reaction for chlorine when tested by heating with calcium oxide it was twice redis- 

 solved and reprecipitated. Chlorine was not removed in this way so that the body 

 is the quinine ester of benzdichlorarsin : C20H23 • NjO ■ CO2 • CsH, . AsCI^. The chlo- 



''La Coste, Ann. d. Chem. (Liebig) (1881), 208, 16. 



