Benchmarking coupled cluster methods on singlet excited states of nucleobases

Dániel Kánnár, P. Szalay

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

In this paper, coupled cluster methods CC2, CCSD, CCSDR(3) and EOM-CCSD(T) have been benchmarked against CC3 for the transition energies of nucleobases. Beside presenting vertical excitation energies for about 30 singlet transitions of four molecules, the results are analyzed statistically and problematic cases have been discussed in detail. It is concluded that the mean deviation of the CC2 results is smaller than that of the CCSD. However, the latter seems to be more systematic, i.e. it usually overestimates excitation energies by about 0.2 eV but with somewhat smaller standard deviation. Unfortunately, with decreasing single excitation contribution in the wave function CCSD gives large error, which can not be corrected by the non-iterative triples methods CCSDR(3) and EOM-CCSD(T).

Original languageEnglish
Pages (from-to)2503
Number of pages1
JournalJournal of Molecular Modeling
Volume20
Issue number11
DOIs
Publication statusPublished - Nov 1 2014

Fingerprint

Benchmarking
Excitation energy
Electron transitions
Excited states
Wave functions
excitation
Molecules
energy
standard deviation
wave functions
deviation
molecules

ASJC Scopus subject areas

  • Medicine(all)

Cite this

Benchmarking coupled cluster methods on singlet excited states of nucleobases. / Kánnár, Dániel; Szalay, P.

In: Journal of Molecular Modeling, Vol. 20, No. 11, 01.11.2014, p. 2503.

Research output: Contribution to journalArticle

@article{96fb5a9f77e44f1683fbe5e78e759e19,
title = "Benchmarking coupled cluster methods on singlet excited states of nucleobases",
abstract = "In this paper, coupled cluster methods CC2, CCSD, CCSDR(3) and EOM-CCSD(T) have been benchmarked against CC3 for the transition energies of nucleobases. Beside presenting vertical excitation energies for about 30 singlet transitions of four molecules, the results are analyzed statistically and problematic cases have been discussed in detail. It is concluded that the mean deviation of the CC2 results is smaller than that of the CCSD. However, the latter seems to be more systematic, i.e. it usually overestimates excitation energies by about 0.2 eV but with somewhat smaller standard deviation. Unfortunately, with decreasing single excitation contribution in the wave function CCSD gives large error, which can not be corrected by the non-iterative triples methods CCSDR(3) and EOM-CCSD(T).",
author = "D{\'a}niel K{\'a}nn{\'a}r and P. Szalay",
year = "2014",
month = "11",
day = "1",
doi = "10.1007/s00894-014-2503-2",
language = "English",
volume = "20",
pages = "2503",
journal = "Journal of Molecular Modeling",
issn = "1610-2940",
publisher = "Springer Verlag",
number = "11",

}

TY - JOUR

T1 - Benchmarking coupled cluster methods on singlet excited states of nucleobases

AU - Kánnár, Dániel

AU - Szalay, P.

PY - 2014/11/1

Y1 - 2014/11/1

N2 - In this paper, coupled cluster methods CC2, CCSD, CCSDR(3) and EOM-CCSD(T) have been benchmarked against CC3 for the transition energies of nucleobases. Beside presenting vertical excitation energies for about 30 singlet transitions of four molecules, the results are analyzed statistically and problematic cases have been discussed in detail. It is concluded that the mean deviation of the CC2 results is smaller than that of the CCSD. However, the latter seems to be more systematic, i.e. it usually overestimates excitation energies by about 0.2 eV but with somewhat smaller standard deviation. Unfortunately, with decreasing single excitation contribution in the wave function CCSD gives large error, which can not be corrected by the non-iterative triples methods CCSDR(3) and EOM-CCSD(T).

AB - In this paper, coupled cluster methods CC2, CCSD, CCSDR(3) and EOM-CCSD(T) have been benchmarked against CC3 for the transition energies of nucleobases. Beside presenting vertical excitation energies for about 30 singlet transitions of four molecules, the results are analyzed statistically and problematic cases have been discussed in detail. It is concluded that the mean deviation of the CC2 results is smaller than that of the CCSD. However, the latter seems to be more systematic, i.e. it usually overestimates excitation energies by about 0.2 eV but with somewhat smaller standard deviation. Unfortunately, with decreasing single excitation contribution in the wave function CCSD gives large error, which can not be corrected by the non-iterative triples methods CCSDR(3) and EOM-CCSD(T).

UR - http://www.scopus.com/inward/record.url?scp=84937001864&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84937001864&partnerID=8YFLogxK

U2 - 10.1007/s00894-014-2503-2

DO - 10.1007/s00894-014-2503-2

M3 - Article

C2 - 25394400

AN - SCOPUS:84912073631

VL - 20

SP - 2503

JO - Journal of Molecular Modeling

JF - Journal of Molecular Modeling

SN - 1610-2940

IS - 11

ER -