Linear fractional charge behavior in density functional theory through dielectric tuning of conductor-like polarizable continuum model

Understanding Photophysical Properties of Molecules Relevant in Organic Semiconductor Laser Diodes from Electron Localization Function-Tuned and Solvent-Tuned Range-Separated Functionals

Organic semiconductor laser diodes (OSLDs) are prevalent in optoelectronics because of their sustainable energy applications. Organic molecules used in such diodes are usually large; hence, their studies are computationally challenging with high-end benchmark methods. Computational methods with reliable accuracy and efficiency are always indispensable. In the present work, we have applied our computationally inexpensive, nonempirically tuned [electron localization function (ELF*) and solvent (Sol*)] range-separated (RS) functionals to study five molecules used in OSLDs. The emission energies in three different environments [toluene, CBP (4,4'-bis(n-carbazolyl)-1,1'-biphenyl) film, and gas] have been computed with the tuned functionals and compared with the experimental emission energies. ELF* and Sol* functionals can accurately reproduce emission energies in toluene and CBP film environments. On the other hand, both ELF* and IP-tuned functionals with excited-state geometry (IP*) perform better in the gas phase. In addition, a comparative study is performed between time-dependent density functional theory and the Tamm-Dancoff approximation. Along with the emission energy, oscillator strength values have also been reported. Different IP-tuned RS parameters were obtained with the ground- and excited-state geometries. Interestingly, it has been observed that the optimally tuned RS parameter with excited-state geometry (IP*) performs better compared to that with ground-state geometries (IP). Fractional occupation calculations show that the tuned functionals exhibit less localization and delocalization error. The study envisages that ELF* and Sol* functionals can be used to design future candidates for OSLDs.

Excited-State Properties of Some Thermally Activated Delayed Fluorescence Emitters: Quest for an Accurate and Reliable Computational Method

Thermally activated delayed fluorescence (TADF) finds application in organic light-emitting diodes. The molecules exhibiting TADF are characterized by small singlet-triplet energy gaps that help reverse intersystem crossing. Recently, ionization potential (IP)-tuned range-separated (RS) density functionals have been well accepted for studying excited-state properties. In the present work, two efficient descriptor-based tuning schemes [electron localization function (ELF) and Sol] of RS density functionals have been used to accurately reproduce the excited-state properties of TADF emitters by performing a single self-consistent field calculation. The lowest singlet vertical excitation energies (E_VA(S₁)) and the vertical singlet-triplet energy gaps (ΔE_VST) are computed with ELF-, Sol-, and IP-tuned RS functionals (LC-BLYP, ωB97, ωB97X, and ωB97XD). Encouraging mean absolute deviations from the experimental values with ELF*-, Sol*-, and IP-tuned functionals are observed. Consistent performance of the non-empirical tuned functionals is noted in different solvent dielectrics. In addition to these, fractional occupation calculations have shown that our tuned functionals almost satisfy the energy linearity curve. Thus, ELF*- and Sol*-tuned functionals are promising and reliable alternatives in computing the excited-state properties. Considering the small experimental singlet-triplet gap, we recommend ELF* to calculate E_VA(S₁) and Sol* to calculate ΔE_VST.

Poisoning density functional theory with benchmark sets of difficult systems

Large benchmark sets like GMTKN55 [Goerigk et al., Phys. Chem. Chem. Phys., 2017, 19, 32184] let us analyse the performance of density functional theory over a diverse range of systems and bonding types. However, assessing over a large and diverse set can miss cases where approaches fail badly, and can give a misleading sense of security. To this end we introduce a series of 'poison' benchmark sets, P30-5, P30-10 and P30-20, comprising systems with up to 5, 10 and 20 atoms, respectively. These sets represent the most difficult-to-model systems in GMTKN55. We expect them to be useful in developing new approximations, identifying weak points in existing ones, and to aid in selecting appropriate DFAs for computational studies involving difficult physics, e.g. catalysis.