2019
DOI: 10.1021/acs.jpca.8b11569
|View full text |Cite
|
Sign up to set email alerts
|

Analytic Gradients for the Spin-Flip ORMAS-CI Method: Optimizing Minima, Saddle Points, and Conical Intersections

Abstract: Analytic nuclear gradients are derived and implemented for the recently introduced SF-ORMAS-CI (spin-flip occupation restricted multiple active space CI) method. Like most SF methods, SF-ORMAS-CI successfully describes bond breaking, diradical systems, transition states, and low-lying excited states, without suffering from spin contamination. The availability of analytic gradients now enables the efficient optimization of equilibrium structures in both ground and excited electronic states, as well as the compu… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
13
0

Year Published

2019
2019
2024
2024

Publication Types

Select...
5
1

Relationship

0
6

Authors

Journals

citations
Cited by 13 publications
(13 citation statements)
references
References 73 publications
(176 reference statements)
0
13
0
Order By: Relevance
“…144 ORMAS has also been used with spin-flip (SF) in order to solve the problem of spin contamination in SF approaches as will be discussed later. 145,146 Several other variations on CASSCF/MCSCF have been discussed in the literature. Optimizing both the CI coefficients and the orbitals is a computationally intensive problem.…”
Section: Multireference Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…144 ORMAS has also been used with spin-flip (SF) in order to solve the problem of spin contamination in SF approaches as will be discussed later. 145,146 Several other variations on CASSCF/MCSCF have been discussed in the literature. Optimizing both the CI coefficients and the orbitals is a computationally intensive problem.…”
Section: Multireference Methodsmentioning
confidence: 99%
“…145,153,338 SF-ORMAS is a single reference CI method that uses a high-spin restricted open shell determinant, where the wave function that has a given multiplicity is generated by an arbitrary amount of spin-flip excitations. 145,146 SF-ORMAS has been developed further with gradients and derivative couplings in order to focus on CoIns. 145,146,343 It has been shown to produce MECI geometries that are similar to MRCI.…”
Section: Topography Of Coinsmentioning
confidence: 99%
“…It is worth mentioning the existence of other electronic structure methods closely related to RAS‐SF. This is the case of the spin‐flip non‐orthogonal CI, 37 the spin‐flip version of the occupation‐restricted multiple active space (SF‐ORMAS), 38,39 which generalizes the orbital partition in RAS‐SF to any number of orbital sets, or quasidegenerate second‐order perturbation approximations 40 . Although these methods are extremely interesting from the fundamental point of view and can potentially provide valuable insight in the characterization of molecular electronic states, they deserve a separate treatment and will not be discussed here.…”
Section: Introductionmentioning
confidence: 99%
“…pqst pqst (5) where the one-and two-body excitation operators are given as (7) where σ indicates α or β spin, and E ot is the on-top density functional with a dependence on the electronic density ρ and the on-top pair density Π (9) as well as the magnitudes of their gradients…”
Section: Introductionmentioning
confidence: 99%
“…In order to obtain accurate energies, theoretical methods must be employed that are capable of recovering both dynamic and static correlation energy . For this reason, multireference methods are often employed. In most such approaches, a multiconfiguration self-consistent field (MCSCF) step is performed first to serve as a reference calculation to recover predominantly static correlation energy, and an additional calculation is performed to recover primarily dynamic correlation energy. Some methods commonly used include complete active space second-order perturbation theory (CASPT2), multireference Møller–Plesset second-order perturbation theory (MRMP2), or one of their variants and multireference configuration-interaction (MRCI). , One drawback of these methods is that the cost of the post-SCF step often scales poorly with system size, and thus, applications of these methods are often limited to systems of small-to-medium size.…”
Section: Introductionmentioning
confidence: 99%