Diphenyldiazomethane reaction with diphenylketene

Products which can be ascribed to the intermediate formation of radicals have long been observed in carbene reactions. In the gas phase these products could arise by homolytic decomposition of excited primary products before collisional deactivation rather than from radicals generated in the course of insertion. This is not so in solution. It is found that, in the thermal decomposition of diphenyldiazomethane (Bethell et al., 1965) or photolysis of diphenylketene (Nozaki et al., 1966) in toluene solution, the product of insertion of diphenylmethylene into the benzylic carbon-hydrogen bonds, 1,1,2-triphenylethane, is accompanied by substantial amounts of 1,1,2,2-tetraphenylethane and bibenzyl. This is a strong indication that discrete diphenylmethyl and benzyl radicals are formed, and, taken in conjunction with EPR (Section IIB) and other evidence (Etter et al., 1959) that diphenylmethylene is a ground-state triplet, would support the view that equation (20) is an adequate representation of triplet insertion. 

Diphenylketene also reacts with diphenyldiazomethane to form the [3+2] cycloadduct 369. In this reaction, the C=0 bond of the ketene participates in the cycloaddition reaction. 

The formation of acetanilide in low yield from the reaction of aniline and the pyrolysis product of a mixture of diazomethane and carbon monoxide was regarded as the most important argument for the existence of the highly reactive and shortlived methylene intermediate that couples with carbon monoxide to give the ketene intermediate [10]. However, the heating of diphenyldiazomethane in the presence of carbon monoxide has not led to the expected diphenylketene [llj. 

---