A Multireference Approach to Electron and Electron-Nuclear Dynamics in Nanomaterials

In recent years, there has been great progress toward the accurate and efficient modeling of nonadiabatic molecular dynamics on a few (10) electronic states, e.g. carrier dynamics in semiconductors, localized surface plasmons, strong field processes, and radiation damage.  In this regime, many standard simulation approaches become impractical due to the sheer number of potential energy surfaces (PESs) that must be computed, while others are limited to very approximate levels of electronic structure theory.  In this work, we address the critical need to develop efficient, general, and definitive ab initio molecular dynamics (AIMD) tools suitable for modeling nonadiabatic dynamics spanning many electronic states.  Here we use the word “definitive” to denote the fact that these methods will allow us to draw definitive conclusions about experimentally observed dynamics.  This will be achieved not only through the development of highly accurate simulation methods, but also by developing strategies to unambiguously connect simulation with real experimental data. We are working to develop a hierarchy of AIMD approaches for modeling dynamics on many electronic states.  Toward this end, we will address three key challenges:  1) We will develop efficient time-dependent electronic structure simulation approaches that achieve high accuracy by incorporating dynamic electron correlation.  2)  We will accurately incorporate light into ab initio nonadiabatic dynamic simulations by careful reformulation of the Ehrenfest with Collapse To A Block (TAB) method that our group has previously developed to simulate nonadiabatic molecular dynamics on many electronic states.  3)  We will close the gap between theoretically computable quantities and experimentally observable ones by develop tools to explicitly compute the signal of ultrafast x-ray absorption spectra from TAB-DMS simulations to enable the unambiguous assignment of these cutting-edge experiments.Throughout the proposed work, we will apply the developed methods to model important physical phenomena, including hot carrier cooling in semiconductor nanocrystals, photocatalysis, and spin dynamics in transition metal complexes.  The outcomes of this work will be novel methods and software tools for modeling nonadiabatic molecular dynamics of systems on many electronic states.