CI Calculations on Didehydrobenzenes Predict Heats of Formation for the Meta and Para Isomers That Are Substantially Higher than Previous Experimental Values

Abstract

CI calculations with the 6-3110** basis set on o-didehydrobenzene (1) predict a singlet-triplet energy difference of 34.4 kcal/mol, which is about 3 kcal/mol smaller than the value measured by Leopold, Miller, and Lineberger. Calculations at this level also predict energy differences between the singlet ground state of 1 and those of the meta (2) and para (3) isomers of respectively 15.8 and 28.4 kcal/mol. These values are both larger by about 6 kcal/mol than the values for the differences between the heats of formation obtained from the experiments of Wenthold, Paulino, and Squires. Computational evidence is presented that calculations which increased ΔEST in 1 by 3 kcal/mol would also increase the energy differences computed between 1 and 2 and 1 and 3 by about the same amount. In addition, the calculated values of ΔEST in 2 and 3 are both at least 8 kcal/mol smaller than the values estimated from two different types of experimental data for ΔH°f for the triplets and the experimental values of ΔH°f reported for the singlets. Calculations of the bond strengths in singlet and triplet 1 support the experimental value of ΔH°f = 106 ± 3 kcal/mol for singlet 1, but the calculations predict bond strengths in 2 and 3 that are about 8 kcal/mol smaller than the values obtained from their experimental heats of formation. Since the bicyclic isomers (4 and 5) of 2 and 3 are both calculated to be higher in energy than their monocyclic counterparts, the formation of 4 and 5 in the experiments of Wenthold, Paulino, and Squires cannot reconcile the heats of formation, measured by them, with the energies for the lowest singlet states of 2 and of 3, calculated by us. On the basis of our computational results, it is proposed that ΔH°f for 2 and 3 is higher than that of 1 by respectively ≥18 and ≥30 kcal/mol. © 1993, American Chemical Society. All rights reserved.