Back to search

University of Minnesota

Absolutely Localized Huzinaga Projection Based Embedding for Efficient and Accurate Molecular Modeling

Abstract

dc:description.abstract

Within computational chemistry, Kohn-Sham density functional theory has become invaluable for low computational cost quantum mechanical calculations.However, computationally expensive post-Hartree-Fock wave function quantum methods are still required to accurately model many chemical systems. We are able to recreate post-Hartree-Fock wave function levels of computational accuracy while only incurring computational costs on par with density functional theory using our absolutely localized Huzinaga level-shift projection based wave function in density functional theory embedding method. Computationally accurate gas adsorption energies on transition metal clusters of metal organic frameworks, spin transition energies of iron models, full configuration interaction (FCI) level energies of small molecules on surfaces and analytical nuclear gradients of our embedding method have been developed and are reported here. We have demonstrated many valuable features of our embedding method such as systematic improvability, applicability to a breadth of chemical problems, and low computational cost for highly accurate calculations. Our Huzinaga embedding method pushes the boundaries of computational chemistry by enabling the calculation of accurate molecular energies for chemical systems previously beyond the scope of existing computational methods. We anticipate our method will be of interest to anyone studying large, complex systems that cannot be accurately modeled using density functional theory and are too large for traditional post-Hartree-Fock methods.

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Graham, Daniel

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Citation
Graham, Daniel. (2022). Absolutely Localized Huzinaga Projection Based Embedding for Efficient and Accurate Molecular Modeling. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/243176.

Chain of custody

source
Harvested from
University of Minnesota
Base URL
conservancy.umn.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Graham, Daniel. Absolutely Localized Huzinaga Projection Based Embedding for Efficient and Accurate Molecular Modeling. 2022. https://hdl.handle.net/11299/243176