Beyond Explored Functionals: A Computational Journey of Two-Photon Absorption

Abstract

We present a thorough investigation into the efficacy of 19 density functional theory (DFT) functionals, relative to RI-CC2 results, for computing two-photon absorption (2PA) cross sections (σ 2PA ) and key dipole moments (|μ 00 |, |μ 11 |, |Δμ|, |μ 01 |) for a series of coumarin dyes in the gas-phase. The functionals include different categories, including local density approximation (LDA), generalized gradient approximation (GGA), hybrid-GGA (H-GGA), range-separated hybrid-GGA (RSH-GGA), meta-GGA (M-GGA), and hybrid M-GGA (HM-GGA), with 14 of them being subjected to analysis for the first time with respect to predicting σ 2PA values. Analysis reveals that functionals integrating both short-range (SR) and long-range (LR) corrections, particularly those within the RSH-GGA and HM-GGA classes, outperform the others. Furthermore, the range-separation approach was found more impactful compared to the varying percentages of Hartree−Fock exchange (HF E x ) within different functionals. The functionals traditionally recommended for 2PA do not appear among the top 9 in our study, which is particularly interesting, as these top-performing functionals have not been previously investigated in this context. This list is dominated by M11, QTP variants, ωB97X, ωB97X-V, and M06-2X, surpassing the performance of other functionals, including the commonly used CAM-B3LYP.

Comment on: “Toward Accurate Two-Photon Absorption Spectrum Simulations: Exploring the Landscape beyond the Generalized Gradient Approximation”

“…The planes of the donor and acceptor π systems in molecule 3 are slightly twisted, which has been identified as a contributing factor to the magnitude of the paramagnetic current density response upon π → π* excitation in ref . Indeed, in a recent study of the 2PA strengths of essentially perfectly planar coumarin derivatives, MGGA and cMGGA results obtained using the implementation from GF23 were in much closer agreement than found for the push–pull systems studied here . Importantly, Figure shows no correlation of deviations due to gauge dependence and other method errors (e.g., errors from the density functional approximations or basis sets), emphasizing the fundamentally erratic and arbitrary nature of the former.…”