메뉴 건너뛰기




Volumn 7, Issue 5, 2011, Pages 1443-1449

First principles simulations of the infrared spectrum of liquid water using hybrid density functionals

Author keywords

[No Author keywords available]

Indexed keywords


EID: 79955903974     PISSN: 15499618     EISSN: 15499626     Source Type: Journal    
DOI: 10.1021/ct2000952     Document Type: Article
Times cited : (162)

References (82)
  • 27
    • 79955896700 scopus 로고    scopus 로고
    • In a so-called rigid water model, the O-H bond length and the H-O-H bond angle are constrained to be fixed during the simulation; in a flexible water model, all ionic degrees of freedom are allowed to vary.
    • In a so-called rigid water model, the O-H bond length and the H-O-H bond angle are constrained to be fixed during the simulation; in a flexible water model, all ionic degrees of freedom are allowed to vary.
  • 37
    • 79955917904 scopus 로고    scopus 로고
    • (accessed March 2011).
    • Li, Z.; Wu, X.; Car, R. http://meetings.aps.org/Meeting/MAR10/Event/ 119032 (accessed March 2011).
    • Li, Z.1    Wu, X.2    Car, R.3
  • 44
    • 79955911374 scopus 로고    scopus 로고
    • Reference 35 reports simulations within the NVT ensemble, and ref 36 reports simulations within the NVE ensemble. In ref 36, a multiple time step scheme was used.
    • Reference 35 reports simulations within the NVT ensemble, and ref 36 reports simulations within the NVE ensemble. In ref 36, a multiple time step scheme was used.
  • 45
    • 46949097364 scopus 로고    scopus 로고
    • The results presented in this paper differ from earlier PI simulations within DFT reported by Chen et al. (82) However, they are consistent with a large body of PI simulations carried out with classical or ab initio derived potentials. It is possible that the results of Chen et al. may suffer from inaccuracies in the comparison between classical and quantum simulations deriving from the choice of technical parameters in their simulations
    • Morrone, J. A.; Car, R. Phys. Rev. Lett. 2008, 101, 017801. The results presented in this paper differ from earlier PI simulations within DFT reported by Chen et al. (82) However, they are consistent with a large body of PI simulations carried out with classical or ab initio derived potentials. It is possible that the results of Chen et al. may suffer from inaccuracies in the comparison between classical and quantum simulations deriving from the choice of technical parameters in their simulations
    • (2008) Phys. Rev. Lett. , vol.101 , pp. 017801
    • Morrone, J.A.1    Car, R.2
  • 54
    • 79955910870 scopus 로고    scopus 로고
    • Qbox code. (accessed March 2011).
    • Qbox code. http://eslab.ucdavis.edu/software/qbox (accessed March 2011).
  • 57
    • 79955915589 scopus 로고    scopus 로고
    • For 32 molecule cells, the ratio of the CPU time per MD step using PBE0 and PBE is about a factor of 25.
    • For 32 molecule cells, the ratio of the CPU time per MD step using PBE0 and PBE is about a factor of 25.
  • 63
    • 79955917404 scopus 로고    scopus 로고
    • Pseudopotential Table. (accessed March 2011).
    • Pseudopotential Table. http://fpmd.ucdavis.edu/potentials (accessed March 2011).
  • 69
    • 79955893275 scopus 로고    scopus 로고
    • Although the HSCV PP yields results in slightly better agreement with AE calculations than the Hamann PP, we used the latter here, in order to be able to compare them to our previous results (33) for IR spectra, obtained with the Hamann PP.
    • Although the HSCV PP yields results in slightly better agreement with AE calculations than the Hamann PP, we used the latter here, in order to be able to compare them to our previous results (33) for IR spectra, obtained with the Hamann PP.
  • 71
    • 79955891723 scopus 로고    scopus 로고
    • In refs 35 and 36, a cutoff function of the type erfc (α r)/ r was adopted for the screened Coulomb operator; different values of α were used in the two papers. The functionals used in refs 35 and 36 are therefore different from the PBE0 one as commonly defined. Another difference between our simulation and those of refs 35 and 36 is the length of the trajectories (10(5) ps and 7 ps in refs 35 and 36, respectively, and 17 ps in our studies) and the equilibration procedure. The simulations of ref 35 contain 32 water molecules and are performed within the NVT ensemble at 350 K, while those of ref 36 contain 64 water molecules and are carried out within the NVE ensemble at 325 K.
    • In refs 35 and 36, a cutoff function of the type erfc (α r)/ r was adopted for the screened Coulomb operator; different values of α were used in the two papers. The functionals used in refs 35 and 36 are therefore different from the PBE0 one as commonly defined. Another difference between our simulation and those of refs 35 and 36 is the length of the trajectories (10(5) ps and 7 ps in refs 35 and 36, respectively, and 17 ps in our studies) and the equilibration procedure. The simulations of ref 35 contain 32 water molecules and are performed within the NVT ensemble at 350 K, while those of ref 36 contain 64 water molecules and are carried out within the NVE ensemble at 325 K.


* 이 정보는 Elsevier사의 SCOPUS DB에서 KISTI가 분석하여 추출한 것입니다.