Exploration of Computational Methods for Mixed Quantum-Classical Solvation Models

Poster Number

15

Lead Author Major

Computer Science

Format

Poster Presentation

Faculty Mentor Name

Anthony Dutoi

Faculty Mentor Department

Chemistry

Abstract/Artist Statement

Solvent and solute interactions are important chemical processes, but the details of such interactions are not well understood. As an example, the acidity of a particular residue of a protein cannot be measured individually since there are many such sites on a protein. Since measurement is not possible, it is useful to calculate the acidity based on the interactions between solvent and solute. This requires accurate position data on the solvent and solute molecules. To acquire this position data we will use an intermolecular potential from quantum mechanical calculations, to run a molecular simulation. We will take advantage of the fact that the behavior of many small groups can be used to describe the behavior of bulk solution. The Monte Carlo method will be used to sample configurations of small groups of molecules with a quantum force field, from which we will obtain position distributions. This sampling requires knowledge of the effects of entropy on the molecules, which is gained from a molecular simulation using an ad hoc potential that is computationally inexpensive. Presently, the code to provide the initial position data required to create the position functions is being worked on. The poster will present the theoretical development of this topic.

Location

DeRosa University Center, Ballroom

Start Date

25-4-2015 2:00 PM

End Date

25-4-2015 4:00 PM

This document is currently not available here.

Share

COinS
 
Apr 25th, 2:00 PM Apr 25th, 4:00 PM

Exploration of Computational Methods for Mixed Quantum-Classical Solvation Models

DeRosa University Center, Ballroom

Solvent and solute interactions are important chemical processes, but the details of such interactions are not well understood. As an example, the acidity of a particular residue of a protein cannot be measured individually since there are many such sites on a protein. Since measurement is not possible, it is useful to calculate the acidity based on the interactions between solvent and solute. This requires accurate position data on the solvent and solute molecules. To acquire this position data we will use an intermolecular potential from quantum mechanical calculations, to run a molecular simulation. We will take advantage of the fact that the behavior of many small groups can be used to describe the behavior of bulk solution. The Monte Carlo method will be used to sample configurations of small groups of molecules with a quantum force field, from which we will obtain position distributions. This sampling requires knowledge of the effects of entropy on the molecules, which is gained from a molecular simulation using an ad hoc potential that is computationally inexpensive. Presently, the code to provide the initial position data required to create the position functions is being worked on. The poster will present the theoretical development of this topic.