QUININE ESTERS OF PHENYLARSINIC ACID DERIVATIVES. 31 



formation, would react in an intermoleeular or intramolecular manner on the 

 secondary amido-group, to form, for instance: 



2AsCl, . CeH, . NH . CH3 . COCl — ^ AsCL . CeH, • N < qo~CH. > -^ ' ^"^* ' ^^^^" 

 Phosphorus trichloride of course will not only react upon the carboxyl group, 

 but at the same time reduce the acid and form the benzarsinic dichloride 

 derivative. 



Therefore different methods were used in the attempt to obtain a compound 

 which on decomposition with water and alkali would give a substance with 

 the properties to be expected of the reduction product of phenylglycinarsinie acid, 

 namely, phenylglycinarsinious acid. 



Dry phenylglycinarsinie acid, added to phosphorus trichloride or phosphoras 

 oxychloride, reacts almost immediately and is charred or largely changed into 

 compounds which are yellow or insoluble in ammonia. A slight reaction takes 

 place in the cold if the acid is suspended in chloroform and phosphorus oxychloride 

 added. . After this addition, cold phosphorus trichloride-chloroform solution was 

 added, and the reaction started by gentle heating, much hydrochloric acid being 

 given off. A considerable rise in temperature takes place on adding a chloroform 

 solution of quinine, but on isolating the products, the arsenic derivative and 

 quinine were found to be uncombined. When, on the other hand, phosphorus 

 trichloride and oxychloride are for the greater part distilled off in vacuo, decom- 

 position takes place resulting in compounds which are insoluble in ammonia. 

 Phenylglycin itself, or nearly related derivatives like methylphenylglycin, are 

 known to be rather unstable in as far as they easily give off carbon dioxide, and 

 it is therefore not surprising to find a compound like AsCL • C9H4 . NH • CHj • COCl 

 also to be unstable. Without investigating whether this compound existed or 

 not, it was set aside as not being suited for the preparation of larger quantities 

 ■of the quinine ester. 



Instead, I attempted to prepare the ester of its acetyl derivative, this 

 change in any event excluding the action of the carboxychloride group 

 on the secondary amido group. 



When the carefully dried phenylglycinarsinie acid (10 grams) was added 

 to cold acetylchloride ( 35 grams ) , a slow reaction set in which was completed 

 by slightly warming on a reflux condenser. The solution had become dark red, 

 and only part of the substance had gone in solution. After the reaction had 

 ceased, the solution was allowed to stand for one hour and was then evaporated 

 imder diminished pressure by means of a dry current of air. On addition of ice 

 water and gradual and careful addition of sodium hydroxide, avoiding an excess, 

 the reaction product easily and completely dissolved to a brown solution. This 

 was acidified with acetic acid which at once gave a yellow precipitate; 

 after standing one hour, addition of acetic acid had no further effect. The 

 precipitate was then filtered (3.5 grams), hydrochloric acid afterwards at once 

 precipitated the remainder of the substance as an almost white compounds 

 (6.5 grams). The instantaneous precipitation, and the form of the crystals, 

 which differs widely from that of the leaf-shaped ones of phenylglycinarsinie 

 acid, were a first indication that the substance was not the original acid. Its 

 sodium salt is decomposed by acetic acid, whereas the sodium salt of phenylglycin- 

 arsinie acid is not. The compound is an acid and it was found to contain arsenic. 

 Since acetylphenyl sodium arsinate is not decomposed by acetic acid, this first 

 compound mentioned above and separated by acetic acid, is very probably an inter- 



-molecular amide, perhaps: AsOjH, • CH, . N^<coLcH,> ^"^ • '^^^' ' AsOjH,, formed 



