Robust and Efficient Constrained DFT Molecular Dynamics Approach for Biochemical Modeling.

ABSTRACT

Constrained density functional theory (cDFT) is a powerful tool to investigate the dynamics of the electrons accompanying various physical-chemical processes. In this article we present our recent progresses in the implementation of the method in the parallelized version of the DFT program deMon2k. We take advantage of the possibility to express atomic densities in terms of linear combination of Hermite Gaussian functions to improve the computation of the cDFT integration weights within the Hirshfeld and Voronoi deformation density electronic population approaches. The efficiency of the method is illustrated on the computation of the average electronic coupling for an electron transfer (ET) through a glycine polypeptide of increasing length. The sampling is based on cDFT and hybrid cDFT/molecular mechanics molecular dynamics simulations. We also report the first estimations of quantum decoherence times from cDFT-based simulations for an ET reaction.