Ergodicity at large couplings with the determinant Monte Carlo algorithm

Richard T. Scalettar, Reinhard M. Noack, and Rajiv R. P. Singh
Phys. Rev. B 44, 10502 – Published 15 November 1991
PDFExport Citation

Abstract

It is known that ergodicity problems are encountered in simulating the Hubbard model at large ratios of the Coulomb interaction strength U to the hopping matrix element t using the ‘‘determinant Monte Carlo’’ algorithm with single-spin-flip moves. We discuss the potential barriers that cause this difficulty in sampling, and introduce global moves of a line of Hubbard-Stratonovich variables on a single site and all time slices to equilibrate the system. We show that at half filling these moves allow the system to explore fully the singly occupied subspace without having to move out of it, in contrast to the single-spin-flip moves of the Hubbard-Stratonovich variables where the system must pass through configurations with appreciable weight for doubly occupied or empty sites in order to go from one state in the singly occupied sector to another. We use these global moves to obtain the antiferromagnetic structure factor at large U/t.

  • Received 1 August 1991

DOI:https://doi.org/10.1103/PhysRevB.44.10502

©1991 American Physical Society

Authors & Affiliations

Richard T. Scalettar

  • Physics Department, University of California, Davis, California 95616

Reinhard M. Noack

  • Physics Department, University of California, Santa Barbara, California 93106

Rajiv R. P. Singh

  • Physics Department, University of California, Davis, California 95616

References (Subscription Required)

Click to Expand
Issue

Vol. 44, Iss. 19 — 15 November 1991

Reuse & Permissions
Access Options
Author publication services for translation and copyediting assistance advertisement

Authorization Required


×
×

Images

×

Sign up to receive regular email alerts from Physical Review B

Log In

Cancel
×

Search


Article Lookup

Paste a citation or DOI

Enter a citation
×