Wavelets for Density-Functional Theory and Post-Density-Functional-Theory Calculations

TL;DR

This paper reviews wavelet applications in density-functional theory and time-dependent DFT, highlighting their advantages in multiresolution problems and demonstrating their effectiveness through comparisons with traditional methods in quantum chemistry.

Contribution

It introduces a wavelet-based approach for DFT and LR-TD-DFT, including implementation details and comparison with Gaussian-type orbital methods, showing improved convergence.

Findings

Wavelet basis converges faster than GTO basis in DFT calculations.

BigDFT code effectively implements wavelet-based DFT and LR-TD-DFT.

Wavelet-based LR-TD-DFT is feasible and promising for future applications.

Abstract

We give a fairly comprehensive review of wavelets and of their application to density-functional theory (DFT) and to our recent application of a wavelet-based version of linear-response time-dependent DFT (LR-TD-DFT). Our intended audience is quantum chemists and theoretical solid-state and chemical physicists. Wavelets are a Fourier-transform-like approach which developed primarily in the latter half of the last century and which was rapidly adapted by engineers in the 1990s because of its advantages compared to standard Fourier transform techniques for multiresolution problems with complicated boundary conditions. High performance computing wavelet codes now also exist for DFT applications in quantum chemistry and solid-state physics, notably the BigDFT code described in this chapter. After briefly describing the basic equations of DFT and LR-TD-DFT, we discuss how they are solved in BigDFT and present new results on the small test molecule carbon monoxide to show how BigDFT results compare against those obtained with the quantum chemistry gaussian-type orbital (GTO) based code deMon2k. In general, the two programs give essentially the same orbital energies, but the wavelet basis of BigDFT converges to the basis set limit much more rapidly than does the GTO basis set of deMon2k. Wavelet-based LR-TD-DFT is still in its infancy, but our calculations confirm the feasibility of implementing LR-TD-DFT in a wavelet-based code.