On the Interplay between Electronic Structure and Polarizable Force Fields When Calculating Solution-Phase Charge-Transfer Rates.

We present a comprehensive analysis of the interplay between the choice of an electronic structure method and the effect of using polarizable force fields vs. nonpolarizable force fields when calculating solution-phase charge-transfer (CT) rates. The analysis is based on an integrative approach that combines inputs from electronic structure calculations and molecular dynamics simulations and is performed in the context of the carotenoid-porphyrin-C60 molecular triad dissolved in an explicit tetrahydrofuran (THF) liquid solvent. Marcus theory rate constants are calculated for the multiple CT processes that occur in this system based on either polarizable or nonpolarizable force fields, parameterized using density functional theory (DFT) with either the B3LYP or the Baer-Neuhauser-Livshits (BNL) density functionals. We find that the effect of switching from nonpolarizable to polarizable force fields on the CT rates is strongly dependent on the choice of the density functional. More specifically, the rate constants obtained using polarizable and nonpolarizable force fields differ significantly when B3LYP is used, while much smaller changes are observed when BNL is used. It is shown that this behavior can be traced back to the tendency of B3LYP to overstabilize CT states, thereby pushing the underlying electronic transitions to the deep inverted region, where even small changes in the force fields can lead to significant changes in the CT rate constants. Our results demonstrate the importance of combining polarizable force fields with an electronic structure method that can accurately capture the energies of excited CT states when calculating charge-transfer rates.