Large-Scale Benchmarking of Multireference Vertical-Excitation Calculations via Automated Active-Space Selection

King, Daniel S.; Hermes, Matthew R.; Truhlar, Donald G.; Gagliardi, Laura

doi:10.26434/chemrxiv-2022-9g7fd

Cited by 2 publications

(17 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Methods for automatically selecting active spaces continue to be an active research topic, and several approaches exist. 20,21,[29][30][31]34,[36][37][38] Here, we employ approximate pair coefficient (APC) selection, 12,32 in which candidate Hartree-Fock orbitals are ranked for the active space by means of their approximate pair coefficient interaction with other orbitals. Given doubly occupied orbitals i and virtual orbitals a, approximate pair coefficients are calculated as…”

Section: Methodsmentioning

confidence: 99%

“…Interactions with singly occupied orbitals are left uncalculated, and singly occupied orbitals are automatically given the highest possible entropy. Finally, due to the observed biasing of APC entropies towards doubly occupied orbitals 12,32 a series of virtual orbital removal steps are employed N times in which the highest-entropy virtual orbital is removed from the sums in equations 6 and 7 and the entropies are recalculated; these highest-entropy virtual orbitals are then assigned the highest entropy at the end of the calculation. For small-to-medium sized organic systems we have found good results with N = 2, 12 which we have used here.…”

Section: Methodsmentioning

confidence: 99%

“…Implementation of APC is now available in PySCF. 39,40 Candidate HF orbitals are then ranked in importance by their orbital entropies, with this ranking used to choose an active space meeting some user-defined size requirement (e.g., (12,12)). Here, to select consistent active space sizes between geometries we employ a simple size requirement in which for an (A,B) active space, the A/2 highest-entropy doubly occupied orbitals and the B − A/2 highest-entropy virtual orbitals are added to the active space; we refer to these active spaces as APC-(A,B).…”

Section: Methodsmentioning

confidence: 99%

“…With a good choice of active space, all static correlation can be addressed with far fewer configurations than FCI and comparable expense to DFT. 12 However, despite the many academic applications of these approaches in the literature, [13][14][15][16][17][18][19] the widespread adoption of these methods for reactivity has been hindered by the challenge of choosing a consistent and adequate active space along the reaction surface. 20,21 The CASSCF energy expression is given by…”

Section: Introductionmentioning

confidence: 99%

“…3 with KS-DFT, MC-PDFT appears promising for attenuating part of the active space inconsistency error, especially when paired with automated methods for choosing the active space in a reliable and consistent fashion. 12,20,21,[29][30][31][32] While MC-PDFT has been tested on a wide variety of systems and excitations, 12,16,33,34 it has yet to be tested in a high-throughput fashion for reactivity.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Organic Reactivity Made Easy and Accurate with Automated Multireference Calculations

Wardzala,

King,

Ogunfowora

et al. 2023

Preprint

View full text Add to dashboard Cite

In organic reactivity studies, quantum chemical calculations play a pivotal role in serving as the foundation of understanding and machine learning model development. While prevalent black-box methods like density functional theory (DFT) and coupled-cluster theory (e.g., CCSD(T)) have significantly advanced our understanding of chemical reactivity, they frequently fall short in accurately describing multiconfigurational transition states and intermediates. Achieving a more accurate description necessitates the use of multireference methods. However, these methods have not been used at scale due to their often-faulty predictions without expert input. Here, we overcome this deficiency through the use of automated multiconfigurational pair-density functional theory (MC-PDFT) calculations. We apply this method to 908 automatically generated main-group organic reactions. We find that 68% of these reactions present significant multiconfigurational character, in which the automated multiconfigurational approach often provides a more accurate and/or efficient description than DFT and CCSD(T). This work presents the first high-throughput application of automated multiconfigurational methods to chemical reactivity, and is enabled by automated active space selection algorithms and the computation of electronic correlation with the on-top functionals of MC-PDFT. The combination of these two new developments can be used in a black-box fashion, avoiding significant active space inconsistency error in both single- and multireference cases and providing accurate multiconfigurational descriptions when needed.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Organic Reactivity Made Easy and Accurate with Automated Multireference Calculations

Wardzala,

King,

Ogunfowora

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Linearized Pair-Density Functional Theory for Vertical Excitation Energies

Hennefarth

King

Gagliardi

2023

Preprint

View full text Add to dashboard Cite

Multiconfiguration pair-density functional theory (MC-PDFT) is a computationally efficient method that computes the energies of electronic states in a state specific or state average framework via an on-top functional. However, MC-PDFT does not include state-interaction among these statets, since the final energies do not come from the diagonalization of an effective model-space Hamiltonian. Recently, multi-state extensions such as linearized (L-) PDFT have been developed to accurately model the potentials near conical intersections and avoided crossings; however, there has not been any systematic study evaluating their performance for predicting vertical excitations at the equilibrium geometry of a molecule, when the excited states are generally well separated. In this letter, we report the performance of L-PDFT on the extensive QUESTDB data set of vertical excitations using a database of automatically selected active spaces. We show that L-PDFT performs well on all these excitations and successfully reproduces the performance of MC-PDFT. These results further demonstrate the potential of L-PDFT, as its scaling is constant with the number of states included in the state-average manifold, whereas MC-PDFT scales linearly in this regard.

show abstract

Large-Scale Benchmarking of Multireference Vertical-Excitation Calculations via Automated Active-Space Selection

Cited by 2 publications

References 71 publications

Organic Reactivity Made Easy and Accurate with Automated Multireference Calculations

Organic Reactivity Made Easy and Accurate with Automated Multireference Calculations

Linearized Pair-Density Functional Theory for Vertical Excitation Energies

Contact Info

Product

Resources

About