2015
DOI: 10.1021/jp511275e
|View full text |Cite
|
Sign up to set email alerts
|

Performance of Frozen Density Embedding for Modeling Hole Transfer Reactions

Abstract: We have carried out a thorough benchmark of the frozen density-embedding (FDE) method for calculating hole transfer couplings. We have considered 10 exchange-correlation functionals, 3 nonadditive kinetic energy functionals, and 3 basis sets. Overall, we conclude that with a 7% mean relative unsigned error, the PBE and PW91 functionals coupled with the PW91k nonadditive kinetic energy functional and a TZP basis set constitute the most stable and accurate levels of theory for hole transfer coupling calculations… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
89
0

Year Published

2015
2015
2024
2024

Publication Types

Select...
6
1

Relationship

2
5

Authors

Journals

citations
Cited by 59 publications
(89 citation statements)
references
References 120 publications
0
89
0
Order By: Relevance
“…In addition, the problem of the electron delocalisation error in DFT electronic structure calculation of adiabatic electronic states is largely suppressed. Several DFT methods are available that yield diabatic electronic states, e.g., constrained DFT, [25][26][27][28][29] frozen density embedding, 30,31 and DFTB. [13][14][15] In the context of FSSH, a drawback of diabatic states is that all energies and nuclear gradients need to be transformed to the adiabatic representation as the nuclear dynamics should be run on the adiabatic states in this method.…”
Section: Introductionmentioning
confidence: 99%
“…In addition, the problem of the electron delocalisation error in DFT electronic structure calculation of adiabatic electronic states is largely suppressed. Several DFT methods are available that yield diabatic electronic states, e.g., constrained DFT, [25][26][27][28][29] frozen density embedding, 30,31 and DFTB. [13][14][15] In the context of FSSH, a drawback of diabatic states is that all energies and nuclear gradients need to be transformed to the adiabatic representation as the nuclear dynamics should be run on the adiabatic states in this method.…”
Section: Introductionmentioning
confidence: 99%
“…This framework has recently been applied [130,131] with great success to systems of biological interest and realistic size (such as portions of the DNA double helix), yielding qualitative and quantitative agreement with experimental data.…”
Section: Note On Self-interactionmentioning
confidence: 92%
“…The subsystem DFT method has been exploited to constrain spin and excess electron densities on specific fragments [127][128][129][130][131] even when KS-DFT fails to do so due to self-interaction error. This behavior of FDE arises from the fact that the FDE theory does not impose orthogonality between the orbitals of different subsystems, thus the typical delocalization of the KS orbitals originating from the orbital's hybridization is completely avoided.…”
Section: Note On Self-interactionmentioning
confidence: 99%
“…In addition, there are several extensions which require to combine results from different subsystem calculations in special ways. Examples are the calculation of excitonically coupled excitations or response properties, where time‐dependent DFT (TDDFT) results for embedded subsystems need to be combined, or pragmatic, physically motivated computational protocols for the description of quasi‐diabatic states as in calculations of transfer integrals in charge transport …”
Section: Introductionmentioning
confidence: 99%
“…Examples are the calculation of excitonically coupled excitations [28,29,[38][39][40] or response properties, [41] where timedependent DFT (TDDFT) results for embedded subsystems need to be combined, or pragmatic, physically motivated computational protocols for the description of quasi-diabatic states as in calculations of transfer integrals in charge transport. [42][43][44] On the one hand, this multitude of options offers great possibilities for a potential user in tailor-made descriptions of molecular processes; in fact, there are probably many interesting extensions that one could easily think of. On the other hand, the large number of variants of approximate FDE/sDFT computations may create confusion among non-expert users.…”
Section: Introductionmentioning
confidence: 99%