THE ALUMNI JOURNAL. 



183 



THE ASSAY OF NITROGLYCERIN. 



By CHARLES RICE, Ph. D. 



(From Research Committee D. of the Committee of Re- 

 vision, etc., of U. S. Pharmacopoeia.) 



The attention of the writer has recently- 

 been drawn to the surprisingly variable 

 effects produced by certain nitroglycerin 

 preparations of the market. When it is 

 remembered that probably all of the nitro- 

 glycerin used in medicine is obtained in 

 form of so-called " 10 percent." alco- 

 holic solution from the manufacturer's 

 works, which the purchaser never visits, 

 and that no reliable and simple process 

 of assay has hitherto been available, the 

 variation is to some extent accounted 

 for. In the hope of finding a way out of 

 this difficulty, the writer examined the 

 methods heretofore proposed for the assay 

 of nitroglycerin, and found that at least 

 one of them, when suitably modified and 

 properly executed, will yield trustworthy 

 results. 



When nitroglycerin is heated with al- 

 coholic potassa or soda, it is split up 

 into glycerin and a mixture of nitrate, 

 nitrite and other salts of the alkali. The 

 glycerin, according to Hay {Journ. 

 Chem. Soc. 48, 74.2), is oxidized at the 

 expense of the N0 3 groups, about two- 

 thirds of which are reduced to the nitrous 

 condition, and the final solution appears 

 t© contain no glycerin at all. Besides 

 nitrate and nitrite, some acetate, lormate 

 and oxalate is formed, together with a 

 small amount of ammonia, and a reddish- 

 brown resinous substance, probably alde- 

 hyde resin, which- gives a dark color to 

 the liquid. Allen {Commerc. Org. An- 

 alysis II., 307) remarks to this: "Com- 

 plex as the reaction is, it appears to 

 occur in a fairly definite manner. Thus 

 Hay found the proportion of nitrous an- 

 hydride (N 2 3 ) produced by the saponifi- 

 cation of 100 parts of nitroglycerin to 

 range between 34.14 and 35.24, the 

 theoretical yield corresponding to the 



equation given by Hay being 33.48." 

 Allen also states that he attempted to 

 apply Koettsdorffer's principle (saponifi- 

 cation with alcoholic potassa) to the as- 

 say of nitroglycerin, but though the re- 

 sults were fairly concordant, the dark 

 color of the liquid prevented the point of 

 neutrality from being ascertained with 

 accuracy by any of the indicators tried. 

 And in a foot note he adds that aqueous 

 potassa solution acts in a manner similar 

 to the alcoholic alkali, but very slowly, 

 owing to the sparing solubility of nitro- 

 glycerin in water. 



Now it is this very process of saponifi- 

 cation which the writer finds to be quite 

 suitable for the assay. If alcoholic po- 

 tassa is employed and alcoholic solution 

 of nitroglycerin be added to it, saponifi- 

 cation will occur at once, or, at least, 

 very rapidly in the cold, and instantly, 

 if the alcoholic potassa is previously 

 heated. With an aqueous solution of 

 potassa the saponification of nitro- 

 glycerin (which must be added in alco- 

 holic solution) is materially slower in 

 the cold, but nearly as rapid, a$ with 

 the alcoholic, if a little heat be applied. 



As far as the assay of nitroglycerin is 

 concerned, the products of the reaction 

 do not at present concern us, since none 

 of them have any influence upon the titer 

 of a volumetric alkali. 



To decompose one molecule of nitro- 

 glycerin, there are required three mole- 

 cules of potassium hydrate: 



(theoretically) 

 C 3 H 5 (N0 3 )3+3KOH = C 3 H 5 (OH) s -f3KN0 3 

 nitroglycerin potassa glycerin potassium 

 226.58 167.97 nitrate. 



Each molecule of KOH. therefore, 

 corresponds to }i molecule or 75.5267 

 parts of nitroglycerin. 



fience each Cc. of decinormal potas- 

 sium hydrate V. S. (containing o 05599 

 Gm. of KOH) corresponds to 0.0755267 

 Gin. of nitroglycerin. 



