A method of preparing pure dihydrochloride of diaminodioxyarsenobenzene

Abstract

The synthesis of a pure arsphenamine or salvarsan, in spite of the excellent published work of Ehrlich and Bertheim 1 and their collaborators, is still a vital problem. It is known by those who have given attention to the subject that the toxicity of arsphenamine varies and that batches from individual manufacturers vary more than they can account for in their procedures. Furthermore, as it seems fairly well proven that even Ehrlich's own manufacturers are unable to keep up a uniformly high standard, 2 it is evident there are some factors which are not understood or under control.

In studying this subject, I came to the idea that the toxicity of arsphenamine is largely due to the use of methyl alcohol and ether in the precipitation of the dihydrochloride and therefore made experiments to prepare the arsphenamine in aqueous solutions, free from any extraneous and objectionable substances:

Finding that the dihydrochloride of the salvarsan base was insoluble in excess of chlorides, as might be expected from the Law of Mass Action, an excess of hydrochloride acid was tried in saiting out the drug. When first tried by making an aqueous solution of the dihydrochloride directly from the base, by dissolving in two normal sodium hydroxide and adding a slight excess of hydrochloride acid, and pouring the solution of hydrochloride into a strong solution of hydrochloric acid (1-1), a white precipitate was formed which, however, turned to a dark-colored gum. This transformation of the white precipitate into the black gum, as will be shown later, was due simply to coalescence of the particles. To prevent this coalescence three factors were changed: (1) The precipitation was conducted at a low temperature and (2) under more dilute conditions and (3) with vigorous stirring.