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

Title: mBEEF-vdW: Robust fitting of error estimation density functionals

Abstract

Here, we propose a general-purpose semilocal/nonlocal exchange-correlation functional approximation, named mBEEF-vdW. The exchange is a meta generalized gradient approximation, and the correlation is a semilocal and nonlocal mixture, with the Rutgers-Chalmers approximation for van der Waals (vdW) forces. The functional is fitted within the Bayesian error estimation functional (BEEF) framework. We improve the previously used fitting procedures by introducing a robust MM-estimator based loss function, reducing the sensitivity to outliers in the datasets. To more reliably determine the optimal model complexity, we furthermore introduce a generalization of the bootstrap 0.632 estimator with hierarchical bootstrap sampling and geometric mean estimator over the training datasets. Using this estimator, we show that the robust loss function leads to a 10% improvement in the estimated prediction error over the previously used least-squares loss function. The mBEEF-vdW functional is benchmarked against popular density functional approximations over a wide range of datasets relevant for heterogeneous catalysis, including datasets that were not used for its training. Overall, we find that mBEEF-vdW has a higher general accuracy than competing popular functionals, and it is one of the best performing functionals on chemisorption systems, surface energies, lattice constants, and dispersion. We also show the potential-energy curve of graphene onmore » the nickel(111) surface, where mBEEF-vdW matches the experimental binding length. mBEEF-vdW is currently available in gpaw and other density functional theory codes through Libxc, version 3.0.0.« less

Authors:
 [1];  [1];  [1];  [2];  [3]
  1. Stanford Univ., Stanford, CA (United States); SLAC National Accelerator Lab., Menlo Park, CA (United States)
  2. Technical Univ. of Denmark, Lyngby (Denmark)
  3. SLAC National Accelerator Lab., Menlo Park, CA (United States)
Publication Date:
Research Org.:
SLAC National Accelerator Lab., Menlo Park, CA (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC-22)
OSTI Identifier:
1349403
Alternate Identifier(s):
OSTI ID: 1259991
Grant/Contract Number:
AC02-76SF00515
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Physical Review B
Additional Journal Information:
Journal Volume: 93; Journal Issue: 23; Journal ID: ISSN 2469-9950
Publisher:
American Physical Society (APS)
Country of Publication:
United States
Language:
English
Subject:
36 MATERIALS SCIENCE

Citation Formats

Lundgaard, Keld T., Wellendorff, Jess, Voss, Johannes, Jacobsen, Karsten W., and Bligaard, Thomas. mBEEF-vdW: Robust fitting of error estimation density functionals. United States: N. p., 2016. Web. doi:10.1103/PhysRevB.93.235162.
Lundgaard, Keld T., Wellendorff, Jess, Voss, Johannes, Jacobsen, Karsten W., & Bligaard, Thomas. mBEEF-vdW: Robust fitting of error estimation density functionals. United States. doi:10.1103/PhysRevB.93.235162.
Lundgaard, Keld T., Wellendorff, Jess, Voss, Johannes, Jacobsen, Karsten W., and Bligaard, Thomas. Wed . "mBEEF-vdW: Robust fitting of error estimation density functionals". United States. doi:10.1103/PhysRevB.93.235162. https://www.osti.gov/servlets/purl/1349403.
@article{osti_1349403,
title = {mBEEF-vdW: Robust fitting of error estimation density functionals},
author = {Lundgaard, Keld T. and Wellendorff, Jess and Voss, Johannes and Jacobsen, Karsten W. and Bligaard, Thomas},
abstractNote = {Here, we propose a general-purpose semilocal/nonlocal exchange-correlation functional approximation, named mBEEF-vdW. The exchange is a meta generalized gradient approximation, and the correlation is a semilocal and nonlocal mixture, with the Rutgers-Chalmers approximation for van der Waals (vdW) forces. The functional is fitted within the Bayesian error estimation functional (BEEF) framework. We improve the previously used fitting procedures by introducing a robust MM-estimator based loss function, reducing the sensitivity to outliers in the datasets. To more reliably determine the optimal model complexity, we furthermore introduce a generalization of the bootstrap 0.632 estimator with hierarchical bootstrap sampling and geometric mean estimator over the training datasets. Using this estimator, we show that the robust loss function leads to a 10% improvement in the estimated prediction error over the previously used least-squares loss function. The mBEEF-vdW functional is benchmarked against popular density functional approximations over a wide range of datasets relevant for heterogeneous catalysis, including datasets that were not used for its training. Overall, we find that mBEEF-vdW has a higher general accuracy than competing popular functionals, and it is one of the best performing functionals on chemisorption systems, surface energies, lattice constants, and dispersion. We also show the potential-energy curve of graphene on the nickel(111) surface, where mBEEF-vdW matches the experimental binding length. mBEEF-vdW is currently available in gpaw and other density functional theory codes through Libxc, version 3.0.0.},
doi = {10.1103/PhysRevB.93.235162},
journal = {Physical Review B},
number = 23,
volume = 93,
place = {United States},
year = {Wed Jun 15 00:00:00 EDT 2016},
month = {Wed Jun 15 00:00:00 EDT 2016}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 5works
Citation information provided by
Web of Science

Save / Share:
  • Cited by 5
  • In recent years, the popularity of density functional theory with periodic boundary conditions (DFT) has surged for the design and optimization of functional materials. However, no single DFT exchange–correlation functional currently available gives accurate adsorption energies on transition metals both when bonding to the surface is dominated by strong covalent or ionic bonding and when it has strong contributions from van der Waals interactions (i.e., dispersion forces). Here we present a new, simple method for accurately predicting adsorption energies on transition-metal surfaces based on DFT calculations, using an adaptively weighted sum of energies from RPBE and optB86b-vdW (or optB88-vdW) densitymore » functionals. This method has been benchmarked against a set of 39 reliable experimental energies for adsorption reactions. Our results show that this method has a mean absolute error and root mean squared error relative to experiments of 13.4 and 19.3 kJ/mol, respectively, compared to 20.4 and 26.4 kJ/mol for the BEEF-vdW functional. For systems with large van der Waals contributions, this method decreases these errors to 11.6 and 17.5 kJ/mol. Furthermore, this method provides predictions of adsorption energies both for processes dominated by strong covalent or ionic bonding and for those dominated by dispersion forces that are more accurate than those of any current standard DFT functional alone.« less
    Cited by 3
  • In recent years, the popularity of density functional theory with periodic boundary conditions (DFT) has surged for the design and optimization of functional materials. However, no single DFT exchange–correlation functional currently available gives accurate adsorption energies on transition metals both when bonding to the surface is dominated by strong covalent or ionic bonding and when it has strong contributions from van der Waals interactions (i.e., dispersion forces). Here we present a new, simple method for accurately predicting adsorption energies on transition-metal surfaces based on DFT calculations, using an adaptively weighted sum of energies from RPBE and optB86b-vdW (or optB88-vdW) densitymore » functionals. This method has been benchmarked against a set of 39 reliable experimental energies for adsorption reactions. Our results show that this method has a mean absolute error and root mean squared error relative to experiments of 13.4 and 19.3 kJ/mol, respectively, compared to 20.4 and 26.4 kJ/mol for the BEEF-vdW functional. For systems with large van der Waals contributions, this method decreases these errors to 11.6 and 17.5 kJ/mol. Furthermore, this method provides predictions of adsorption energies both for processes dominated by strong covalent or ionic bonding and for those dominated by dispersion forces that are more accurate than those of any current standard DFT functional alone.« less