PRECIPITANTS FOR FURFURAL 41 



A COMPARISON OF BARBITURIC ACID, THIOBARBI- 

 TURIC ACID AND MALONYLGUANIDINE AS QUANTI- 

 TATIVE PRECIPITANTS FOR FURFURAL. 



ARTHUR W. DOX AND G. P. PLAISANCE. 



All of the methods for the quantitative determination of 

 pentoses and pentosans in agricultural products are based upon 

 the conversion of pentose into furfural by distillation with a 

 mineral acid, preferably hydrochloric, and subsequent estima- 

 tion of furfural in the distillate by means of a suitable reagent. 

 Giinther, Chalmot and Tollens 1 titrated the furfural with phen- 

 ylhydrazine, using aniline acetate paper as an indicator. 

 Stone 2 made use of the same reaction, but used Fehling's solu- 

 tion to determine the excess of phenylhydrazine. Later, Flint 

 and Tollens 3 showed that this titration method was not accurate, 

 on account of the levulinic acid resulting from the decomposition 

 of hexoses, as well as the instability of the standard phenylhy- 

 drazine acetate reagent used. Jolles 4 titrated the furfural with 

 potassium bisulphite and iodine. In the absence of other re- 

 ducing substances, the furfural could be determined directly 

 with Fehling's solution. Giinther and Tollens 5 precipitated the 

 furfural as hydrofurfuralimide by means of ammonia, while 

 Chalmot and Tollens 6 used phenylhydrazine and weighed the 

 resulting hydrazone. In both cases the condensation product 

 was somewhat soluble. 



Councilor 7 was the first to use phloroglucinol for the quanti- 

 tative determination of furfural. This method was later studied 

 and perfected by Tollens and his co-workers. The phloroglu- 

 cinol method, although known to be faulty in several respects, 

 is the one in general use today, having been adopted as provi- 

 sional by the Association of Official Agricultural Chemists. 8 It 

 is strictly empirical, since the nature of the reaction and the 

 constitution of the condensation product have not been deter- 

 mined. Krober 9 compiled a table in which the weight of fur- 

 furalphloroglucide obtained is interpreted in terms of furfural, 

 xylose, arabinose or pentose. This table is purely empirical, 

 being based on trial distillations and precipitations of the fur- 

 fural or the particular pentose employed, and not upon the 



