Correction to “Excited-State Absorption: Reference Oscillator Strengths, Wave Function, and TDDFT Benchmarks”

Read full text Summarise with VIRAT
Authors: Jakub Širůček, Boris Le Guennic, Yann Damour, Pierre‐François Loos, Denis Jacquemin
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

InformationT here were some inconsistencies concerning the ESA oscillator strengths of carbon monoxide in the original work.First, for the doubly degenerate Π → Π transition, the published values are correct for all methods.However, unlike in the case of the doubly degenerate benzene transition, they should be multiplied by 2 instead of 4 for carbon monoxide to allow comparisons with experiment, since among the four possible combinations, only two have significant f.Hence, the statement made on p. 4692 just before Section 3.1 is not fully correct.Second, for all TD(A)-DFT calculations, the three singly degenerate transitions of carbon monoxide, i.e., 2Σ + → 2Π, 1Π → 2Σ + , and 1Π → 3Σ + were treated incorrectly.The originally published values were incorrectly multiplied by 2 for TD(A)-DFT results during the statistical analysis.Therefore, while the wave function statistics of the original paper, as well as the energy statistics for all methods, are correct, the TD(A)-DFT statistics are slightly altered (see below).We provide new Supporting Information with corrected data, as well as the corrected Figures 5678and 10 below.The key changes are 1.TD-CAM-B3LYP is even more accurate than originally reported and becomes very competitive with respect to QR-CC2 and, in some cases, even QR-CCSD.2. The overall trends among the tested functionals remain the same, with BH&HLYP swapping place with LC-BLYP47 by a small margin.3. The errors of all functionals for small molecules (1-3 non-hydrogen atoms) become smaller.The general trends and qualitative conclusions regarding the impact of the system size are, however, unaffected.4. The largest changes are logically obtained for the small subset of transitions used for the estimate of the geometry relaxation (Section 3.4).The error metrics of all TD(A)-DFT determined on the optimal S 0 drops significantly, and they all become very competitive with QR-CC2 and even QR-CCSD.Their order changes as well, with CAM-B3LYP moving, gratifyingly, to the top spot in terms of MAE. 5.In contrast, the statistical indicators for the same subset but using the excited-state geometries do not change significantly, and the original functional order is conserved.This means that the deterioration of f accuracy

Fields of Study & Subjects

Advanced Chemical Physics Studies Spectroscopy and Quantum Chemical Studies Photochemistry and Electron Transfer Studies

Publication details

Journal / source
Journal of Chemical Theory and Computation
Volume
22
Issue
7
Pages
3775–3778
Publication date
27 March 2026
Language
EN