Unveiling the role of short-range exact-like exchange in the optimally tuned range-separated hybrids for fluorescence lifetime modeling

Abstract

We propose and validate several variants of the optimally tuned range-separated hybrid functionals (OT-RSHs) including different density functional approximations for predicting the fluorescence lifetimes of different categories of fluorophores within the time-dependent density functional theory (TD-DFT) framework using both the polarizable continuum and state-specific solvation models. Our main idea originates from performing the optimal tuning in the presence of a contribution of the exact-like exchange at the short-range part, which, in turn, leads to the small values of the range-separation parameter, and computing the fluorescence lifetimes using the models including no or small portions of the short-range exact-like exchange. Particular attention is also paid to the influence of the geometries of emitters on fluorescence lifetime computations. It is shown that our developed OT-RSHs along with the polarizable continuum model can be considered as the promising candidates within the TD-DFT framework for the prediction of fluorescence lifetimes for various fluorophores. We find that the proposed models not only outperform their standard counterparts but also provide reliable data better than or comparable to the conventional hybrid functionals with both the fixed and interelectronic distance-dependent exact-like exchanges. Furthermore, it is also revealed that when the excited state geometries come into play, more accurate descriptions of the fluorescence lifetimes can be achieved. Hopefully, our findings can give impetus for future developments of OT-RSHs for computational modeling of other characteristics in fluorescence spectroscopy as well as for verification of the related experimental observations.

Stacked Ensemble Machine Learning for Range-Separation Parameters

“…We selected molecules with corresponding experimental measurements as the test set of optical properties and categorized them into subsets based on their structural features, excited-state properties, and real-life applications. These subsets include OPV, , TADF, polycyclic aromatic hydrocarbons (PAH), aggregation-induced emission molecules (AIE), fluorophores (FL), and bio-organic molecules (BIO) . The majority of molecules in the present test set of optical properties, such as all AIE compounds and many OPV and TADF structures, are not included in the above-mentioned test set of ω ML , so that they serve as great benchmarks for the generalizability of our model.…”

Excited-state properties of organic semiconductor dyes as electrically pumped lasing candidates from new optimally tuned range-separated models

“…As another important point, the optimal values of μ for the considered systems are close to those recently reported for the emitters exhibiting thermally activated delayed fluorescence, sensitizers and annihilators in the triplet–triplet annihilation mechanism, and other different types of fluorophores. 40,42,43,49 This, in turn, hints that it is possible to recommend a more general applicable range-separation parameter for describing the related photophysical phenomena. However, we also find that the optimally tuned values of the range-separation parameter are smaller, by a factor of about 2–3, than those empirically determined for the corresponding standard LC models, such as 0.33 Bohr −1 for first- to third-row atoms and 0.47 Bohr −1 for the larger sets of small molecules.…”

“…It has recently been illustrated that more reliable accounting for the emission energies of chromophores necessitates performing the corresponding computations using the combinations of a and b parameters with no portions (small contributions) of the SR HF exchange. 49 Although we have considered the latter findings in our study, namely, using the case of a = 0.0, b = 1.0 for all the excited-state computations, to examine the effects of the SR HF exchange we have also included different percentages of the SR HF exchange during the optimal tuning processes. Even though the SR HF exchange values may be obtained from the first-principles arguments, [50][51][52] they can also be determined by empirical fittings.…”

Stacked Ensemble Machine Learning for Range-Separation Parameters

“…and categorized them into subsets based on their structural features, excited-state properties, and real-life applications. These subsets include OPV, 141,146 TADF, 147 polycyclic aromatic hydrocarbons (PAH), 141 aggregation-induced emission molecules (AIE), 148 fluorophores (FL) 149 and bio-organic molecules (BIO). 141 The majority of molecules in the present test set of optical properties, such as all AIE compounds and many OPV and TADF structures, were not included in the above-mentioned test set of ω ML , so that they served as great benchmarks for the generalizability of our model.…”