Gooch and Blake — The Per chlorate Method. 381 



Art. XXXIII. — The Perchlorate Method for the Determina- 

 tion of the Alkali Metals; by F. A. Gooch and G. R. 

 Blake. 



[Contribution from the Kent Chemical Laboratory of Yale Univ. — ccxciv.] 



In the work of which an account is here given, the object at 

 the outset was the examination of the perchlorate precipitation 

 of rubidium and caesium, as proposed bj Montemartini and 

 Matucci,"^ for the estimation of those elements ; but the recent 

 paper of Baxter and Kobayashi, which appeared while this work 

 w^as in progress, upon the perchlorate determination of potas- 

 sium, suggested the desirability of including that element also 

 within the scope of the investigation. In the work of Baxter 

 and Kobayashif careful attention is paid to the relations of 

 solubility of potassium perchlorate in the washing media, the 

 use of alcohol containing about O'l per cent of perchloric acid 

 (as first proposed by Wense:|:) and saturated with potassium 

 perchlorate (as suggested by Davis and advocated by Thin 

 and Cumming§) is adopted, and further recommendations are 

 made that the washing alcohol be of absolute strength and of a 

 temperature as near as possible to 0°. The authors state that 

 there is no danger of the deposition of potassium perchlorate 

 from the saturated alcoholic solution owing to upward change 

 of temperature during the manipulation and make record of the 

 observation that the addition of a large amount of sodium per- 

 chlorate failed to induce the precipitation of potassium perchlo- 

 rate from the solution saturated with the latter salt. It is obvi- 

 ous, however, that if a condition of supersaturation in respect 

 to potassium perchlorate were brought about by the addition of 

 sodium perchlorate to the solution, the contact of the sapersat- 

 urated solution with potassium perchlorate already precipitated 

 might induce a further precipitation of that salt from the wash- 

 ing liquid; and this is a point of much importance in the applica- 

 tion of the perchlorate method to the determination of rubidium 

 and caesium. We have therefore tested the matter by adding 

 sodium perchlorate to a saturated solution of potassium ])er- 

 chlorate in alchohol (97 per cent), made and used at the wui'k- 

 ing temperature of the laboratory, shaking the solution in con- 

 tact with a weighed amount of solid potassium perchlorate, 

 filtering and weighing upon asbestos the insoluble precipitate. 

 In the first two experiments solid sodium perchlorate was 

 dissolved in the saturated alcholic solution of potassium per- 

 chlorate to wdiicli a weighed amount of solid potassium perchlo- 

 rate had been added. In the last two experiments the sodium 



* Gaz. Chem., xxxiii, 189, 1903. 

 t Jour. Am. Chem. Soc, xxxix, 249, 1917. 

 X Zeitschr. anal. Chem., v, 691, 1891. 

 § Jour. Chem. Soc, cvii, 361, 1915. 



