skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: High-Performance First-Principles Molecular Dynamics for Predictive Theory and Modeling

Abstract

This project focused on developing high-performance software tools for First-Principles Molecular Dynamics (FPMD) simulations, and applying them in investigations of materials relevant to energy conversion processes. FPMD is an atomistic simulation method that combines a quantum-mechanical description of electronic structure with the statistical description provided by molecular dynamics (MD) simulations. This reliance on fundamental principles allows FPMD simulations to provide a consistent description of structural, dynamical and electronic properties of a material. This is particularly useful in systems for which reliable empirical models are lacking. FPMD simulations are increasingly used as a predictive tool for applications such as batteries, solar energy conversion, light-emitting devices, electro-chemical energy conversion devices and other materials. During the course of the project, several new features were developed and added to the open-source Qbox FPMD code. The code was further optimized for scalable operation of large-scale, Leadership-Class DOE computers. When combined with Many-Body Perturbation Theory (MBPT) calculations, this infrastructure was used to investigate structural and electronic properties of liquid water, ice, aqueous solutions, nanoparticles and solid-liquid interfaces. Computing both ionic trajectories and electronic structure in a consistent manner enabled the simulation of several spectroscopic properties, such as Raman spectra, infrared spectra, and sum-frequency generation spectra. Themore » accuracy of the approximations used allowed for direct comparisons of results with experimental data such as optical spectra, X-ray and neutron diffraction spectra. The software infrastructure developed in this project, as applied to various investigations of solids, liquids and interfaces, demonstrates that FPMD simulations can provide a detailed, atomic-scale picture of structural, vibrational and electronic properties of complex systems relevant to energy conversion devices.« less

Authors:
 [1];  [2];  [3]
  1. Univ. of California, Davis, CA (United States). Dept. of Computer Science
  2. Univ. of Chicago, IL (United States)
  3. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Publication Date:
Research Org.:
Univ. of California, Davis, CA (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC-22). Materials Sciences & Engineering Division
OSTI Identifier:
1410963
Report Number(s):
DOE-UCDAVIS-0008938-1
DOE Contract Number:  
SC0008938
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND SUPERFLUIDITY; 97 MATHEMATICS AND COMPUTING; First-principles molecular dynamics; Density Functional Theory; Hydrogen-bonded liquids; Solid-liquid interfaces; Vibrational spectroscopy

Citation Formats

Gygi, Francois, Galli, Giulia, and Schwegler, Eric. High-Performance First-Principles Molecular Dynamics for Predictive Theory and Modeling. United States: N. p., 2017. Web. doi:10.2172/1410963.
Gygi, Francois, Galli, Giulia, & Schwegler, Eric. High-Performance First-Principles Molecular Dynamics for Predictive Theory and Modeling. United States. doi:10.2172/1410963.
Gygi, Francois, Galli, Giulia, and Schwegler, Eric. Sun . "High-Performance First-Principles Molecular Dynamics for Predictive Theory and Modeling". United States. doi:10.2172/1410963. https://www.osti.gov/servlets/purl/1410963.
@article{osti_1410963,
title = {High-Performance First-Principles Molecular Dynamics for Predictive Theory and Modeling},
author = {Gygi, Francois and Galli, Giulia and Schwegler, Eric},
abstractNote = {This project focused on developing high-performance software tools for First-Principles Molecular Dynamics (FPMD) simulations, and applying them in investigations of materials relevant to energy conversion processes. FPMD is an atomistic simulation method that combines a quantum-mechanical description of electronic structure with the statistical description provided by molecular dynamics (MD) simulations. This reliance on fundamental principles allows FPMD simulations to provide a consistent description of structural, dynamical and electronic properties of a material. This is particularly useful in systems for which reliable empirical models are lacking. FPMD simulations are increasingly used as a predictive tool for applications such as batteries, solar energy conversion, light-emitting devices, electro-chemical energy conversion devices and other materials. During the course of the project, several new features were developed and added to the open-source Qbox FPMD code. The code was further optimized for scalable operation of large-scale, Leadership-Class DOE computers. When combined with Many-Body Perturbation Theory (MBPT) calculations, this infrastructure was used to investigate structural and electronic properties of liquid water, ice, aqueous solutions, nanoparticles and solid-liquid interfaces. Computing both ionic trajectories and electronic structure in a consistent manner enabled the simulation of several spectroscopic properties, such as Raman spectra, infrared spectra, and sum-frequency generation spectra. The accuracy of the approximations used allowed for direct comparisons of results with experimental data such as optical spectra, X-ray and neutron diffraction spectra. The software infrastructure developed in this project, as applied to various investigations of solids, liquids and interfaces, demonstrates that FPMD simulations can provide a detailed, atomic-scale picture of structural, vibrational and electronic properties of complex systems relevant to energy conversion devices.},
doi = {10.2172/1410963},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Sun Dec 03 00:00:00 EST 2017},
month = {Sun Dec 03 00:00:00 EST 2017}
}

Technical Report:

Save / Share: