Minnig — Separation and Estimation of Aluminium, 199 



In the experiments of Table II the iron was precipitated by 

 ]STH 4 OH after cautiously diluting the filtrate with water and 

 then determined gravimetrically as the oxide. The error of 

 the results ranged from +0-0004 grm. to +0*0013 grra. in 

 terms of ferric oxide. On investigation this was found to be 

 due to ferric phosphate derived from the phosphorous com- 

 pounds used in the preparation of the acetyl chloride. 



Since the precipitation of the aluminium was quantitative 

 and rapid, and iron could be determined easily other than as 

 the oxide, larger quantities of aluminium were used with 

 measured quantities of ferric chloride without determining the 

 iron in the filtrate. With the use of larger quantities of the 

 acetone-acetyl chloride solution, necessary for the increased 

 amount of the salts, the aluminium chloride precipitate also 

 began to include phosphate appreciably, as is shown in 

 Table III. 







Table III. 







Solution of 



A1 2 3 



Fe 2 3 







A1C1 3 .6H 2 



by 



present 



A1 2 3 





taken 



standard 



as chloride 



found 



Error 



cm 3 



grm. 



grm. 



grm. 



grm. 



15 



0-0592 



0-0373 



0-0597 



+ 0-0005 



15 



0-0592 



0-0373 



0-0602 



+ 0-0010 



15 



0-0592 



0-0746 



0-0600 



+ 0-0008 



15 



0-0592 



0-0746 



0-0602 



+ 0-0010 



15 



0-0592 



0-0373 



0-0603 



+ 0-0011 



15 



0-0592 



0-0373 



0-0606 



+ 0-0014 



Many efforts were made to purify the contaminated acetyl 

 chloride, by redistilling after refluxing (for six or seven hours) 

 with a return condenser over sodium acetate ;* over manganese 

 dioxide, or over lead dioxide ; but the product in every case 

 gave a test for phosphoric acid when hydrolyzed with water, 

 evaporated with nitric acid, and treated with the solution of 

 ammonium molybdate in nitric acid. Phosphorus-free acetyl 

 chloride was finally prepared by passing a rapid stream of 

 hydrogen chloride either through a mixture of glacial acetic 

 acid and phosphorus pentoxidef or, preferably, through purified 

 acetic anhydride kept at 100°4 Distillation over anhydrous 

 sodium acetate easily removes contaminating phosphorus com- 

 pounds from the acetic anhydride. 



Satisfactory results were obtained by the use of the pure 

 acetyl chloride when as much as 1*3 grm. of the A1C1 3 .6H 2 



* Richter, f, 301 (1909 edition). 



t Richter. i, 300. 



% Beilstein's Handbuch, i, 462. 



