Bibliography of Literature on GW Ab Initio Calculations which use Imaginary Time

Vincent Sacksteder IV

DOI: https://doi.org/10.48550/arXiv.1904.02565

2019-04-04

Abstract:The GW Approximation is an ab initio approach to calculating electronic structure which avoids using the Local Density (LDA) Approximation, the Generalized Gradient (GGA) Approximation, or similar density functionals. It goes beyond the Hartree-Fock approximation by including screening and excited state effects, and shares conceptual similarities with MP2 and RPA calculations. Because GW includes dynamics and time/frequency dependence of the system's screening and excited state behavior, a pivotal issue in any GW calculation is the question of how to numerically represent and manipulate time/frequency dependence. While earlier GW calculations generally used a representation in real time/frequency, many recent calculations have used a representation on the imaginary time axis. Imaginary time is important not only for numerics but also because it can enable additional physics such as systems where self-consistent GW is needed or that are strongly interacting, integration with DMFT, and scaling of GW to very large systems. This current text reviews the research literature on imaginary time GW codes and briefly discusses the possibilities for memory and CPU optimizations.

Mesoscale and Nanoscale Physics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: in GW approximation calculations, how to effectively represent and manipulate time/frequency dependence, especially the representation on the imaginary time axis. Specifically, the paper explores the advantages of the imaginary time axis representation in the following five aspects: 1. **Systematically solve the pole problem**: - In the real - time representation, numerical sensitivity and divergence (pole) problems often occur. The imaginary time axis representation provides a systematic method to deal with these poles, which is helpful for self - consistent GW calculations, not just a single calculation. 2. **Achieve self - consistent GW calculations**: - The imaginary time axis representation provides a systematic method to solve the pole problem, so it is one of the important means for self - consistent GW calculations. Another main method is the quasi - particle qsGW method of Kotani and Schilfgaarde. Part of the motivation for the author's literature search was to explore self - consistent GW calculations, and two research results of self - consistent GW calculations without using qsGW or the imaginary time axis were reported. 3. **Improve CPU efficiency and scalability**: - Through the space - time approach pioneered by Godby et al. and recently implemented by the VASP code, the imaginary time axis representation can significantly improve CPU efficiency and scalability. Using this method, GW calculations can be carried out on systems containing dozens of atoms on a small cluster. 4. **Combine with DMFT**: - Dynamic mean - field theory (DMFT) calculations are usually carried out on the imaginary time axis. Since there is ongoing work to combine GW (or constrained RPA) with DMFT to produce a fully ab - initio strongly correlated code, it is advantageous to use the same imaginary time axis representation in GW calculations. 5. **Include temperature effects**: - The imaginary time axis allows the fully self - consistent inclusion of temperature effects in ab - initio calculations. In addition, the paper also discusses possible memory and CPU optimization measures in GW codes, aiming to provide researchers with a comprehensive overview, encourage further research on the imaginary time method, and summarize existing research results and algorithms.

Bibliography of Literature on GW Ab Initio Calculations which use Imaginary Time

Accelerating Analytic-Continuation GW Calculations with a Laplace Transformation and Natural Auxiliary Functions

The $GW$ Approximation: A Quantum Chemistry Perspective

Accelerating Analytic-Continuation GW Calculations with a Laplace Transform and Natural Auxiliary Functions

Automated Workflow for Accurate High-Throughput GW Calculations

All-electron self-consistent GW in the Matsubara-time domain: implementation and benchmarks of semiconductors and insulators

Two-component $GW$ calculations: Cubic scaling implementation and comparison of vertex corrected and partially self-consistent $GW$ variants

AB-G$_0$W$_0$: A practical G$_0$W$_0$ method without frequency integration based on an auxiliary boson expansion

Coulomb-hole summations and energies for GW calculations with limited number of empty orbitals: a modified static remainder approach

Precision benchmarks for solids: G0W0 calculations with different basis sets

A numerical approach for calculating exact non-adiabatic terms in quantum dynamics

AB-G0W0: A practical G0W0 method without frequency integration based on an auxiliary boson expansion

Numerical integration for ab initio many-electron self energy calculations within the GW approximation

SternheimerGW: a program for calculating GW quasiparticle band structures and spectral functions without unoccupied states

$GW$ density matrix to estimate self-consistent $GW$ total energy in solids

BerkeleyGW: A massively parallel computer package for the calculation of the quasiparticle and optical properties of materials and nanostructures

Joint Approximate Diagonalization approach to Quasiparticle Self-Consistent $GW$ calculations

Accelerating Time-Dependent Density Functional Theory and GW Calculations for Molecules and Nanoclusters with Symmetry Adapted Interpolative Separable Density Fitting

Many-Body Green's Function Theory for Electronic Excitations in Complex Chemical Systems

Fully Dynamic G 3 W 2 Self-Energy for Finite Systems: Formulas and Benchmark

Efficient Full-Frequency GW Calculations Using a Lanczos Method