Theoretical comparison of ketene dimerization in the gas and liquid phase

We present the first theoretical comparison between ketene dimerization in gas phase and ketene dimerization in solution. Density functional theory (DFT) calculations on the ketene dimerization were carried out considering the following product dimers: diketene (d-I), 1,3-cyclobutanedione (d-II), 2,4-dimethylene-1,3-dioxetane (d-III), and 2-methyleneoxetan-3-one (d-IV). All structures were optimized at the PW86x+PBEc/DZP level of theory. Based on these geometries, a total of 58 meta and hybrid functionals were used to evaluate the heat of dimerization. The MPW1K functional was found to fit the experimental data best and subsequently used in the final analyses for all energy calculations. It was found on both kinetic and thermodynamic grounds that only d-I and d-II are formed during ketene dimerization in gas phase and solution. In gas phase, d-I is favored over d-II by 2 kcal/mol. However, the dimerization barrier for d-I is 1 kcal/mol higher than for d-II. Solvation makes dimerization more favorable. On the enthalpic surface this is due to a favorable interaction between the dimer dipole moment and solvent molecules. The dimer is stabilized further on the Gibbs energy surface by an increase of the dimerization entropy in solution compared to gas phase. The species d-I remains the most stable dimer in solution by 1 kcal/mol. Kinetically, the dimerization barriers for the relevant species d-I and d-II are cut in half by solvation, due to both favorable dimer-dipole/solvent interactions (DeltaH++, DeltaG++) and an increase in the activation entropies (DeltaS++). While the dimerization barrier for d-II is lowest for the gas phase and toluene, the barrier for d-I formation becomes lowest for the more polar solvent acetone by 1 kcal/mol as d-I dimerization has the most polar transition state.