Frozen-Density Embedding Theory (FDET) was originally formulated for obtaining the ground-state energy and density of an embedded system by means of variational-principle based treatments of interacting NA electrons [1,2]. The key element of FDET is the functional EV[ΨA,ρB] for the total energy which depends on two independent variables: a) the quantum chemical descriptor for NA electrons (ΨA - embedded wavefunction for instance) and b) the non-negative function ρB(r) representing the remaining Ntot-NA electrons. Using the functional EV[ΨA,ρB] in Euler-Lagrange equation results in the Schrödinger-like equation with the total Hamiltonian differrinig from the Hamiltonian of the environment-free NA-electron system by the embedding operator. This operator is a multiplicative potential by construction. It depends, however, on both ρB(r) and the density of the embedded system (see Figure). It is, therefore, state dependent. FDET can be practically implemented using any method from the toolbox of quantum chemistry in order to optimize ΨA. Concerning the other independent variable - ρB(r) - any physical model/method can be used to generate it. Various computational protocols are thus possible for applications of FDET in multi-level simulations. In this paper, we will discuss our recent results [3,4] concerning: 1) Generalizations of FDET for: a) non-variational methods to treat embedded NA electrons and b) excited states. 2) Correspondence between FDET expression for the embedding potential and energy with the corresponding terms in widespread QM/MM methods. 3) The sensitivity of the FDET results on the choice of ρB(r) in case of non-covalently bound environments.

Frozen-Density Embedding Theory (FDET) based multi-level simulations for electronic structure of embedded species: approximations, procedures, and benchmarking.

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.