Large-scale Modeling of Electron Dynamics and Excitations in Functional and Structural Materials

Understanding excited state phenomena is at the heart of many important materials problems, such as photovoltaics, photocatalysis, plasmonics, solid-state lighting, and biochemical sensing, to name a few. In recent years, time-dependent density functional theory (TDDFT) has become one of the most powerful, versatile and popular computational tools for probing electronic structure and excitations in molecular and solid state materials. However, despite tremendous success of TDDFT, the application of the theory to practical problems remains challenging, in a large part owing to its high computational demand. This challenge has confined the otherwise powerful theory to modest systems with no more than a few hundred atoms at most. In this project, we propose to develop two TDDFT methods that address the computational challenge headon and could potentially transform the landscape of the field. The objectives of the project are five-fold: (1) We will reformulate TDDFT based on time-dependent density functional perturbation theory (TD-DFPT) which circumvents the computational bottleneck of the conventional TDDFT formulation by projecting the complex Hamiltonian matrix to a substantially reduced sub-Hilbert space and leads to one order of magnitude saving in computational time and memory. As a result, the new subspace TD-DFPT method retains the same accuracy as the conventional TDDFT method, but with the computational time and memory similar to those of the ground state DFT calculations. (2) We will incorporate range-separated hybrid functionals into the subspace TD-DFPT method, which is necessary to deal with Rydberg states, charge-transfer states, and excitations in weakly bonded complexes. Furthermore, we will incorporate the new TDDFT formulation into the non-adiabatic molecular dynamics framework so that more rigorous coupled ion-electron dynamics can be performed for large systems. (3) We will develop a novel time-dependent orbital-free DFT (TD-OFDFT) method that is capable of treating electron dynamics in noble metals on a length-scale relevant to experiments. (4) We will validate the new methods against the experimental measurements and the conventional TDDFT method on smaller systems for which the latter is applicable, and make the software available to the research community. (5) We will bring the methods to bear on a number of important materials problems relevant to the Navy, including charge separation dynamics in organic and perovskite-based photovoltaics, photo-switching in organic compounds, plasmonic-enhanced photocatalysis and electroplasticity in metals.