Reports: B6
46344-B6 Carbanion Intermediates in Proton-Transfer Reactions Between Carbon Acids and Alkoxides
The computational investigation has undergone significant modification. Our previous computational methods at the B3LYP/6-31+G(d,p) level in the gas phase and modeling bulk solvent effects using the IEFPCM method implemented in Gaussian03W at the same level were not reproducing known experimental results. Initially, it appeared when modeling the proton exchange reactions between methoxide in methanol with pentafluorobenzene, C6H5CHClCF3, C6H5CH(CF3)2, and 9-phenylfluorene using a purely bulk solvent model was the problem. We then added solvating methanol molecules and switched to a smaller proton donor, CHCl3, treating the system with both implicit and bulk solvent effects. This also did not reproduce experimental results.
In the Fall 2008 I "retooled" my computational lab to run Linux on all computers and switched to the latest versions of GAMESS and GAMESSPlus. We are currently modeling our systems with the a better DFT method, M06-2X to account for hydrogen bonding effects and using a fragment molecular orbital methods and effective fragment potentials to treat implicit solvent effects. We began investigating treatment of bulk solvent effects using SM8 which was an improvement over PCM models.
During the past year my research group has been exploring using EFP and FMO solvent models in the GAMESS package. These are showing the most promise in getting reasonable agreement between computational results and known experimental values.
