Dynamical Cluster Algorithm for Highly Correlated Electron Systems
Abstract
We introduce a method to include short-ranged dynamical correlations in addition to the local dynamics of the dynamical mean field approximation (D=∞). The technique is based on an iterative self--consistency scheme on a finite size cluster with periodic boundary conditions. As the cluster size grows the technique approaches the infinite system correctly. The dynamical mean field approximation is obtained by reducing the cluster to a single site. At a given cluster size the methods includes correlations with range of roughly half the linear dimension of the cluster. The methods allows the study of nonlocal order parameters (e.g. d--wave superconductivity) by straightforward extension of the techniques which have been successful in the case of local dynamics in infinite dimensions. As a demonstration of feasibility we study the Falicov-Kimball model on a small cluster. We show that the spectral function is positive semi--definite and fulfills the sum rules, establishing that the method preserves causality for this model. This work was supported by the NSF grants DMR-9704021 and DMR-9357199.
- Publication:
-
APS March Meeting Abstracts
- Pub Date:
- March 1998
- Bibcode:
- 1998APS..MAR.X3103H