Mixing Stochastic-Deterministic Density Functional Theory In The PAW Formalism To Tackle Extreme Conditions Physics

In computational materials modeling, density functional theory (DFT) is a powerful tool for studying systems ranging from just a few molecules to much more condensed phases. However, the finite temperature extension to DFT formulated by Mermin has a cubic scaling with system size and temperature, limiting its applicability for studying the physics of materials in extreme environments. In this talk, I will describe a novel proposal for a mixed DFT (mDFT) formalism that combines the stochastic and deterministic Kohn-Sham algorithms of DFT to study matter at any temperature [1]. We incorporate projector-augmented wave (PAW) potentials that improve the overall scaling of stochastic and mixed DFT toward a universal method across temperatures. Furthermore, we show that mDFT with PAW drastically reduces the computational effort without compromising the accuracy of purely deterministic DFT for studying the ground-state properties of materials. The time-dependent extension to mDFT enables us to simulate the dynamics within the Born-Oppenheimer approximation and beyond.

LA-UR-22-31081

[1] White, A.J., Collins, L.A., "Fast and Universal Kohn-Sham Density Functional Theory Algorithm for Warm Dense Matter to Hot Dense Plasma,'' Phys. Rev. Lett., 2020, 125, 055002.