Electronic Structure Methods for Large Inhomogeneous Systems
Poster Number
4B
Introduction/Abstract
Coarsely considered, the bonds that hold atoms together in chemistry consist of mutual attractions between the nuclei of those atoms and the electrons that surround them. However, quantitatively useful methods for simulating (and therefore understanding) these interactions involve solving very detailed equations, which can consume a lot of computational resources. At present, there is generally a choice between choosing methods that provide a very detailed description or methods that can handle a very large system.
Purpose
This work is a continuation of a long-running project aimed at breaking this bottleneck.
Method
The details of local chemical interactions are repackaged in a way that it is transferable between chemical simulations, thus bringing down the computational costs of those simulations by taking advantage of previously computed quantities. Previous versions of the method under development were usable but it had some unexpected artifacts, unfortunately coming from things that were baked into the foundational formalism.
Results
Recently a path to circumvent these difficulties has been discovered. The outline of the path forward will be presented along with its context (the potential applications, previous developments, and the difficulties encountered).
Significance
As explained in the introduction, if successful, we will be able to do more detailed simulations of larger, more interesting systems.
Location
William Knox Holt Memorial Library and Learning Center, University of the Pacific, 3601 Pacific Ave., Stockton, CA 95211
Format
Poster Presentation
Poster Session
Afternoon
Electronic Structure Methods for Large Inhomogeneous Systems
William Knox Holt Memorial Library and Learning Center, University of the Pacific, 3601 Pacific Ave., Stockton, CA 95211
Coarsely considered, the bonds that hold atoms together in chemistry consist of mutual attractions between the nuclei of those atoms and the electrons that surround them. However, quantitatively useful methods for simulating (and therefore understanding) these interactions involve solving very detailed equations, which can consume a lot of computational resources. At present, there is generally a choice between choosing methods that provide a very detailed description or methods that can handle a very large system.
