Efficient implementation of the continuous-time hybridization expansion quantum impurity solver

TL;DR

This paper presents an efficient, comprehensive implementation of the continuous-time hybridization expansion quantum impurity solver, enhancing capabilities for strongly correlated systems within dynamical mean-field theory.

Contribution

It introduces a complete, optimized implementation of the hybridization expansion quantum impurity solver with new features like retarded interactions and advanced measurement routines.

Findings

Supports retarded interactions in impurity problems

Includes improved estimators for self-energy and vertex functions

Embedded in the ALPS-DMFT package for broader use

Abstract

Strongly correlated quantum impurity problems appear in a wide variety of contexts ranging from nanoscience and surface physics to material science and the theory of strongly correlated lattice models, where they appear as auxiliary systems within dynamical mean-field theory. Accurate and unbiased solutions must usually be obtained numerically, and continuous-time quantum Monte Carlo algorithms, a family of algorithms based on the stochastic sampling of partition function expansions, perform well for such systems. With the present paper we provide an efficient and generic implementation of the hybridization expansion quantum impurity solver, based on the segment representation. We provide a complete implementation featuring most of the recently developed extensions and optimizations. Our implementation allows one to treat retarded interactions and provides generalized measurement routines based on improved estimators for the self-energy and for vertex functions. The solver is embedded in the ALPS-DMFT application package.