24 



Report of Schimmel $ Co. April 1913. 



Congress of applied Chemistry which assembled in September 1912 at Washington. 

 Dodge has made control-tests of some of the methods which have been recommended 

 for this estimation and has found that Sadtler's method 1 ), which has been incorporated in 

 the Pharmacopoeia of the U. S. does not give satisfactory results. The method devised 

 by Ripper 2 ) gave better values, although these, too, are only approximately accurate. 

 Ripper shakes up the aldehyde with a bisulphite-solution of known strength and titrates 

 back the excess of bisulphite with iodine solution at a low temperature. According 

 to Dodge, the most suitable manner of carrying out this method is as follows: 0,15 g. 

 aldehyde or bitter almond oil is weighed in a flask containing exactly 25 cc. of a, 

 say, Vfi n. bisulphite solution, and brought to solution by gentle shaking. The flask is 

 then closed, left standing in ice-water for \ x j% to 2 hours, after which the ice-cold 

 solution is titrated with decinormal iodine-solution, starch being used as an indicator. 

 A blank test is carried out in the same way, the benzaldehyde content being calculated 

 from the difference (1 cc. decinormal iodine solution = 0,0053 g. benzaldehyde). The 

 reaction is completed as soon as the blue colour persists for a few seconds. By this 

 method Dodge, in the case of pure benzaldehydes, found from 94,8 to 99,0 p. c. 



Determination with hydroxylamine (according to Walther, as modified by A. H. Bennett, 

 comp. Reports April 1909, 1 10 and October 1909, 153) offered no advantages as compared 

 with the iodometric process, inasmuch as in the case of pure benzaldehydes only 

 93,5 and 95,1 p. c. was found. On the other hand, the colorimetric estimation with 

 phenylhydrazine recommended by Denner 3 ), as well as by Denis and Dunbar 4 ) gives 

 useful, although at times rather too high, values (95,1 to 100,9 p. c), but in this method 

 the unstable character of the reagent is a disadvantage. 



It should further be stated that in natural bitter almond oil all these methods only 

 permit of the estimation of the free benzaldehyde, and not of that which is present 

 in combination with hydrocyanic acid. This is shown by the following table: — 





Sample No. 1 



Sample No. 2 



Sample No. 3 



HCN 



Equivalent quantity of cyanohydrin . 

 Calc. quantity of free benzaldehyde . 

 Estim. by Ripper's method .... 

 Estim. with hydroxylamine .... 

 Estim. with phenylhydrazine (Denner's 

 or Denis' and Dunbar's method)' 5 ) 



3,6 p.c. 



17,6 p.c. 



82,4 p. c. 



81,3 to 82,1 p.c. 



78,9 p.c. 



83,2 p.c. 



2,8 p.c. 



86,2 p.c. 

 80 p.c; 81,4 p.c. 



83,6 p.c; 87,2 p.c. 



2,4 p.c. 

 88,2 p.c. 



85,2 p. c 



x ) Volumetric estimation of the separated sodium hydroxide, with the aid of neutral sodium sulphite. 

 Comp. our Reports April 19(H, 48 and October 1904, 119. — 2 ) Monatsh. f. Chem. 21 (1900), 1079. — *) Zeitschr. 

 f. anal. Chem. 29 (1890), 228. Also comp. Chem. Zentralbl. 1887, 1411. — *) Compt. Report October 1909, 155. 

 — 5 ) We may add to the above that this also applies to Kleber's phenylhydrazine method, which is likewise 

 adapted for the estimation of benzaldehyde (comp. p. 45). In the case of three samples of bitter almond oil 

 containing hydrocyanic acid, of which we estimated the benzaldehyde-content in this manner, we obtained the 

 values given below, which clearly show that only the aldehyde wich was present in the oil in the free state 

 has entered into reaction: — 



di5Q 



HCN 



Equiv. 



quant. 



cyanohydrin 



Acid 

 v. 



Benzoic 

 acid 



Free Benzaldehyde 

 c , Est. by 



KlebeT's method 



Sample a 

 Sample b 

 Sample c 



1 ,0678 

 1 ,0603 

 1,0848 



4,95 p. c. 

 2,79 p. c. 

 9,5 p. c. 



24,4 p. c. 

 13,3 p.c. 

 46,8 p. c. 



2,7 

 5,3 

 14,0 



0,6 p. c. 

 1,2 p.c. 

 3,1 p.c. 



ca. 75 p. c. 

 „ 85,5 p. c. 

 „ 50,1 p.c. 



73 p. c. 

 83 p. c. 

 50 p. c. 



