Abstract
Determining the multi-reference character of a molecular system and its impact on the limits within which its properties may be calculated accurately by different quantum chemical methods remains a difficult yet important task in computational chemistry. Especially, transition metal compounds continue to frequently provide a challenge to quantum chemists in this regard. In this work, we construct, analyze, and evaluate different computational protocols to determine the impact of the multi-reference character of transition metal compounds on their bond dissociation energies using a set of reference data for 60 diatomic molecules. We find that the fractional orbital density approach allows to determine two global indicators on a physically sound basis. These can subsequently be used to classify the assessed set of molecules with high accuracy into categories of systems for which their multi-reference character matters substantially for their bond dissociation energies and for which it does not. A comparison with earlier suggested thresholds for classification of molecular systems due to their multi-reference character suggests that our approaches yield substantially better performance.
Highlights
In order to calculate molecular properties accurately, the consideration of electron correlation plays a crucial role
Note further that such an extreme classification is accompanied by true positive rate (TPR) = 0 and true negative rate (TNR) = 1, i.e., the overall accuracy of 72% is achieved by sacrificing the TPR, which is not favorable for a diagnostic
The overall accuracies using these predictors are all less or equal to 68% with TPRs ranging from 6% to 41% on the cost of TNRs ranging from 74% to 88%
Summary
In order to calculate molecular properties accurately, the consideration of electron correlation plays a crucial role. Post Hartree–Fock (HF) approaches such as coupled cluster theory usually incorporate dynamic correlation sufficiently, typically leading to accurate results for systems for which electron correlation is dominated by its dynamic contribution. In order to determine if the description of a certain system requires a single- or multi-configurational ansatz, many diagnostics have been developed, such as the T18 and D1,9 and the %TAE, M,11 and c02 12 approaches. In the T1 and D1 diagnostics, the Frobenius norm and the matrix 2-norm, respectively, of the coupled-cluster calculations with single and double substitutions (CCSD) single excitation amplitude are taken as measures for the MR character of the considered system, which probably appear as the most direct approaches to estimate the MR character of a molecular system based on a single-determinant ansatz. The M diagnostic is based on the occupation number obtained from the one-particle reduced density of the CASSCF wavefunction, while scitation.org/journal/jcp in the case of the c02 diagnostic, the dominant coefficient of the CASSCF wavefunction is determined
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
