DFT studies on the mechanism of the cycloaddition reaction between methyleneketene and 5-methylene-1,3-dioxan-4,6-dione: regioselectivity and solvent effect

Three different reaction schemes for the cycloaddition reaction between methyleneketene and 5-methylene-1,3-dioxan-4,6-dione were studied by means of the B3LYP/6-31G* methods both in the gas phase and in a solvent of dichloromethane. All the geometries of the stationary points on the reaction paths were optimized by energy gradient technique, and all transition states optimized by the Berny technique. Transition states were ascertained by frequency analysis. The results can be summed up as follows: there are three possible sites of cycloaddition in methyleneketene, i.e., 2,3-C=C, 1,2-C=C and C=O. In the gas phase, the cycloadditions of different double bonds of methyleneketene with 5-methylene-1,3-aioxan-4,6-dione are all concerted but asynchronous, taking place through twisted transition states. The activation barriers for reactions (1), (2) and (3) are calculated to be 2.96, 0.25, and 21.81 kcal/mol at the B3LYP/6-31G* level, respectively. The computational results show that the energy barrier for the reaction leading to 1,2-adduct is the lowest one, which is in consistence with the regioselectivity of the reactions observed by experiments. For comparison, the solvent effect was also studied in the solvent of dichloromethane using self-consistent reaction field model. (C) 1999 Elsevier Science B.V. All rights reserved.