Recently we presented a theory' which proposed that the most chemically reactive diol epoxide on a benxo-ring of a polycyclic aromatic hydrocarbon would have the highest biological activity as a mutagen or carcinogen. Studies on benxo[a]pyrene2, benxo[a]anthracene3, substituted benzo[a]-4 anthracenes , chrysene 5 , and dibenzo[a,h]anthracene 6 , have provided support for the predictions of this theory. In addition, Eisenstadt and Cold7 have shown that the environmental contaminant cyclopenta[c,d]pyrene (21, found in carbon black' and automobile exhaust 9 , is highly mutagenic on metabolic activation. Our calculations of chemical reactivity' provide support for their suggestion7 that an epoxide at the 3,4-position of 1 could be responsible for the mutagenicity of this hydrocarbon on metabolic activation. Because of the interest in 1, we have sought an improved synthesis 10 in order to make the hydrocarbon and its derivatives more readily available for biological studies.
Two obvious pathways for the synthesis of 1 consist of reduction of ketones 2 or 3 to alcohols followed by dehydration to 1.
Potential precursors of these ketones are 1-pyrenylacetic acid 11 and 4-pyrenylacftic acid12, respectively. Since 1-pyrenylacetic acid was readily accessible via acetylation of pyrene, cyclization of this acid was attempted first. Interestingly, treatment of 1-pyrenylacetic acid with either polyphosphoric acid at 120% or with HP at room temperature overnight, or attempted cycliaation of 1-pyrenylacetyl chloride with SnC14 in boiling benzene failed to produce 2. In