ITR: Cross-Correlation Filter-Diagonalization with Parallel-Computation Monte-Carlo Approaches for Extraction of Eigenvalues of He Clusters in Confined Spaces

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

Dan Neuhauser of UCLA is supported by the Chemistry Division through the Information Technology Program to combine cross-correlation filter diagonalization and Monte Carlo methods to investigate low-lying excitation dynamics of helium clusters confined within aromatic structures or fullerene molecules. Three Monte Carlo approaches will be employed: trajectory dependent cellularization, imaginary-time diffusion Monte Carlo, and auxiliary-field Monte Carlo. The rise in computational power and the reduced cost of high-speed Linux computer clusters motivate this development of new approaches for quantum dynamics based on parallel computations. The development of new methods to study spectra of large chemical systems on clusters of inexpensive, parallel computers demonstrates that formerly intractable quantum mechanical problems can benefit from the next generation of computing. Novel technical solutions such as those developed here will be useful in the more general context of signal processing and information extraction. Further, the investigations of fullerene-based encapsulation of molecules are of import to hydrogen storage applications.