On the geometry dependence of tuned‐range separated hybrid functionals

Abstract

Molecules and materials that absorb and/or emit light form a central part of our daily lives. Consequently, a description of their excited-state properties plays a crucial role in designing new molecules and materials with enhanced properties. Due to its favorable balance between high computational efficiency and accuracy, time-dependent density functional theory (TDDFT) is often a method of choice for characterizing these properties. However, within standard approximations to the exchangecorrelation functional, it remains challenging to achieve a balanced description of all excited states, especially for those exhibiting charge-transfer (CT) characteristics. In this work, we have applied two approaches, namely, the optimal tuning and triplet tuning methods, for a nonempirical definition of range-separated functionals to improve the description of excited states within TDDFT. This is applied to study the CT properties of two thermally activated delayed fluorescence emitters, namely, PTZ-DBTO 2 and TAT-3DBTO 2 . We demonstrate the connection between the two methods, the performance of each in the presence on multiple excited states of different characters and the geometry dependence of each method especially relevant in the context of developing size-consistent potential energy surfaces.

Assessing challenging intra‐ and inter‐molecular charge‐transfer excitations energies with double‐hybrid density functionals

“…For this kind of luminophores, the fine-tuning of the ω parameter for range separation [105][106][107] is known to deliver very accurate values, although at a considerable computational effort and with some geometry dependence. 108 As expected, our results…”

Using the Mechanical Bond to Tune the Performance of a Thermally Activated Delayed Fluorescence Emitter**

“…Although the 3 LE triplet state lies ~0.3 eV higher in energy, it is expected that this gap is overestimate due to the challenges associated with TD-DFT calculations predicting the absolute energy of CT states. [36] The oscillator strength (f) for the S1-S0 transition was calculated to be 0.316, 0.272 and 0.271 for 1, 1⊂2 and 1⊂22 respectively at the ground state geometry but fell to 0 for all structures in their computed excited state geometry. Given that we experimentally observed photoluminescence with a µs lifetime in all cases, we set out to explore the role of molecular flexibility on the photophysical properties using ab initio MD simulations.…”

Theoretical Insights for Materials Properties of Cyclic Organic Nanorings

“…The calculation of and excitation energies, or in general and , normally benefits from TD‐DFT methods (e.g., range‐separated functionals) able to cope with intramolecular charge‐transfer excitations, as those arising from the localization of the HOMO and LUMO orbitals on different spatial molecular fragments or monomers, [ 298 ] and the Tamm–Dancoff approximation [ 299 ] to improve the accuracy of the excitation energy and band shapes. [ 300,301 ] The use of double‐hybrid density functionals for excited‐states, [ 302,303 ] incorporating a fraction of long‐range perturbative correlation effects, [ 304 ] offers an alternative to robustly describe excitonic effects too.…”