Critical By-products in the Synthesis of Polychlorobiphenyls

Abstract

PURE samples of polychlorobiphenyls are desirable for toxicity testing in preference to the randomly substituted products¹ of commercial synthesis. Substrates required (by G. Denton) for specific tests with the American hardshelled clam Mactra mercenaria were obtained by chlorination of N,N-diacetylbenzidine². This method has the advantage that the starting material can be obtained in a pure state and, following chlorination in acetic acid solution, the derivative (Fig. 1; X, Y=NHAc, Z=H) can be diazotized and reduced to the tetrachlorobiphenyl (Fig. 1; X, Y, Z= H; Fig. 2, peak A) or further chlorinated by the Sandmeyer method to the hexachloro compound (Fig. 1; X,Y=Cl, Z=H; Fig. 2, peak B). The method has been used by a number of groups³, who have not, however, commented on the many by-products formed. Gas chromatography using an electron capture detector revealed eight significant components and seven minor products from the Sandmeyer reaction, while deamination of the same starting material gave two major and and three other products (Fig. 2). None of these peaks was detected in laboratory solvents. We have been able to show that these complex mixtures arise from two known sidereactions which have not been detected before in PCB chemistry.