Effects of Conformational Sampling on Computing Redox Properties Using Linear Response Approach

Redox processes are an important step in many chemical and biochemical reactions. One simple approach to calculate the free energy change of a redox process is linear response approximation (LRA). However, variability in conformational and energy-gap sampling poses a challenge in balancing computational cost and accuracy. Herein, we calculate the redox properties of the one-electron oxidation processes for small, biologically relevant redox-active molecules (e.g., phenol, phenolate, benzene, indole, lumiflavin) in aqueous solution using two conformational sampling strategies. We sampled the conformations using molecular mechanics (MM) and hybrid quantum mechanics/molecular mechanics (QM/MM) simulations to investigate how these techniques affect redox properties. We also performed QM/MM energy-gap sampling while varying the QM region to investigate its impact on overall redox behavior. We observed free energy of oxidation, and consequently, oxidation potential differs consistently by ∼0.2-0.4 V between QM/MM and MM sampling for the molecules under investigation. Overall, we infer that computationally cheaper MM sampling would be adequate for computing the redox properties of small molecules when corrected by a system-specific correction factor.