Multiconfiguration pair-density functional theory: Barrier heights and main group and transition metal energetics

Rebecca K. Carlson, Giovanni Li Manni, Andrew L. Sonnenberger, Donald G. Truhlar, Laura Gagliardi

Research output: Contribution to journalArticle

34 Citations (Scopus)

Abstract

Kohn-Sham density functional theory, resting on the representation of the electronic density and kinetic energy by a single Slater determinant, has revolutionized chemistry, but for open-shell systems, the Kohn-Sham Slater determinant has the wrong symmetry properties as compared to an accurate wave function. We have recently proposed a theory, called multiconfiguration pair-density functional theory (MC-PDFT), in which the electronic kinetic energy and classical Coulomb energy are calculated from a multiconfiguration wave function with the correct symmetry properties, and the rest of the energy is calculated from a density functional, called the on-top density functional, that depends on the density and the on-top pair density calculated from this wave function. We also proposed a simple way to approximate the on-top density functional by translation of Kohn-Sham exchange-correlation functionals. The method is much less expensive than other post-SCF methods for calculating the dynamical correlation energy starting with a multiconfiguration self-consistent-field wave function as the reference wave function, and initial tests of the theory were quite encouraging. Here, we provide a broader test of the theory by applying it to bond energies of main-group molecules and transition metal complexes, barrier heights and reaction energies for diverse chemical reactions, proton affinities, and the water dimerization energy. Averaged over 56 data points, the mean unsigned error is 3.2 kcal/mol for MC-PDFT, as compared to 6.9 kcal/mol for Kohn-Sham theory with a comparable density functional. MC-PDFT is more accurate on average than complete active space second-order perturbation theory (CASPT2) for main-group small-molecule bond energies, alkyl bond dissociation energies, transition-metal-ligand bond energies, proton affinities, and the water dimerization energy.

Original languageEnglish (US)
Pages (from-to)82-90
Number of pages9
JournalJournal of Chemical Theory and Computation
Volume11
Issue number1
DOIs
StatePublished - Jan 13 2015

Fingerprint

Wave functions
Transition metals
Density functional theory
transition metals
density functional theory
metals
Dimerization
wave functions
Kinetic energy
Protons
energy
Molecules
Water
dimerization
Coordination Complexes
determinants
Metal complexes
self consistent fields
affinity
Chemical reactions

Cite this

Multiconfiguration pair-density functional theory : Barrier heights and main group and transition metal energetics. / Carlson, Rebecca K.; Li Manni, Giovanni; Sonnenberger, Andrew L.; Truhlar, Donald G.; Gagliardi, Laura.

In: Journal of Chemical Theory and Computation, Vol. 11, No. 1, 13.01.2015, p. 82-90.

Research output: Contribution to journalArticle

@article{47cbca7b13f04148995c5011a31c5ddf,
title = "Multiconfiguration pair-density functional theory: Barrier heights and main group and transition metal energetics",
abstract = "Kohn-Sham density functional theory, resting on the representation of the electronic density and kinetic energy by a single Slater determinant, has revolutionized chemistry, but for open-shell systems, the Kohn-Sham Slater determinant has the wrong symmetry properties as compared to an accurate wave function. We have recently proposed a theory, called multiconfiguration pair-density functional theory (MC-PDFT), in which the electronic kinetic energy and classical Coulomb energy are calculated from a multiconfiguration wave function with the correct symmetry properties, and the rest of the energy is calculated from a density functional, called the on-top density functional, that depends on the density and the on-top pair density calculated from this wave function. We also proposed a simple way to approximate the on-top density functional by translation of Kohn-Sham exchange-correlation functionals. The method is much less expensive than other post-SCF methods for calculating the dynamical correlation energy starting with a multiconfiguration self-consistent-field wave function as the reference wave function, and initial tests of the theory were quite encouraging. Here, we provide a broader test of the theory by applying it to bond energies of main-group molecules and transition metal complexes, barrier heights and reaction energies for diverse chemical reactions, proton affinities, and the water dimerization energy. Averaged over 56 data points, the mean unsigned error is 3.2 kcal/mol for MC-PDFT, as compared to 6.9 kcal/mol for Kohn-Sham theory with a comparable density functional. MC-PDFT is more accurate on average than complete active space second-order perturbation theory (CASPT2) for main-group small-molecule bond energies, alkyl bond dissociation energies, transition-metal-ligand bond energies, proton affinities, and the water dimerization energy.",
author = "Carlson, {Rebecca K.} and {Li Manni}, Giovanni and Sonnenberger, {Andrew L.} and Truhlar, {Donald G.} and Laura Gagliardi",
year = "2015",
month = "1",
day = "13",
doi = "10.1021/ct5008235",
language = "English (US)",
volume = "11",
pages = "82--90",
journal = "Journal of Chemical Theory and Computation",
issn = "1549-9618",
publisher = "American Chemical Society",
number = "1",

}

TY - JOUR

T1 - Multiconfiguration pair-density functional theory

T2 - Barrier heights and main group and transition metal energetics

AU - Carlson, Rebecca K.

AU - Li Manni, Giovanni

AU - Sonnenberger, Andrew L.

AU - Truhlar, Donald G.

AU - Gagliardi, Laura

PY - 2015/1/13

Y1 - 2015/1/13

N2 - Kohn-Sham density functional theory, resting on the representation of the electronic density and kinetic energy by a single Slater determinant, has revolutionized chemistry, but for open-shell systems, the Kohn-Sham Slater determinant has the wrong symmetry properties as compared to an accurate wave function. We have recently proposed a theory, called multiconfiguration pair-density functional theory (MC-PDFT), in which the electronic kinetic energy and classical Coulomb energy are calculated from a multiconfiguration wave function with the correct symmetry properties, and the rest of the energy is calculated from a density functional, called the on-top density functional, that depends on the density and the on-top pair density calculated from this wave function. We also proposed a simple way to approximate the on-top density functional by translation of Kohn-Sham exchange-correlation functionals. The method is much less expensive than other post-SCF methods for calculating the dynamical correlation energy starting with a multiconfiguration self-consistent-field wave function as the reference wave function, and initial tests of the theory were quite encouraging. Here, we provide a broader test of the theory by applying it to bond energies of main-group molecules and transition metal complexes, barrier heights and reaction energies for diverse chemical reactions, proton affinities, and the water dimerization energy. Averaged over 56 data points, the mean unsigned error is 3.2 kcal/mol for MC-PDFT, as compared to 6.9 kcal/mol for Kohn-Sham theory with a comparable density functional. MC-PDFT is more accurate on average than complete active space second-order perturbation theory (CASPT2) for main-group small-molecule bond energies, alkyl bond dissociation energies, transition-metal-ligand bond energies, proton affinities, and the water dimerization energy.

AB - Kohn-Sham density functional theory, resting on the representation of the electronic density and kinetic energy by a single Slater determinant, has revolutionized chemistry, but for open-shell systems, the Kohn-Sham Slater determinant has the wrong symmetry properties as compared to an accurate wave function. We have recently proposed a theory, called multiconfiguration pair-density functional theory (MC-PDFT), in which the electronic kinetic energy and classical Coulomb energy are calculated from a multiconfiguration wave function with the correct symmetry properties, and the rest of the energy is calculated from a density functional, called the on-top density functional, that depends on the density and the on-top pair density calculated from this wave function. We also proposed a simple way to approximate the on-top density functional by translation of Kohn-Sham exchange-correlation functionals. The method is much less expensive than other post-SCF methods for calculating the dynamical correlation energy starting with a multiconfiguration self-consistent-field wave function as the reference wave function, and initial tests of the theory were quite encouraging. Here, we provide a broader test of the theory by applying it to bond energies of main-group molecules and transition metal complexes, barrier heights and reaction energies for diverse chemical reactions, proton affinities, and the water dimerization energy. Averaged over 56 data points, the mean unsigned error is 3.2 kcal/mol for MC-PDFT, as compared to 6.9 kcal/mol for Kohn-Sham theory with a comparable density functional. MC-PDFT is more accurate on average than complete active space second-order perturbation theory (CASPT2) for main-group small-molecule bond energies, alkyl bond dissociation energies, transition-metal-ligand bond energies, proton affinities, and the water dimerization energy.

UR - http://www.scopus.com/inward/record.url?scp=84921395883&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921395883&partnerID=8YFLogxK

U2 - 10.1021/ct5008235

DO - 10.1021/ct5008235

M3 - Article

C2 - 26574206

AN - SCOPUS:84921395883

VL - 11

SP - 82

EP - 90

JO - Journal of Chemical Theory and Computation

JF - Journal of Chemical Theory and Computation

SN - 1549-9618

IS - 1

ER -