Band gaps of halide perovskites from a Wannier-localized optimally tuned screened range-separated hybrid functional

Abstract

The accurate prediction of the band gaps of halide perovskites within density functional theory is known to be challenging. The recently developed Wannier-localized optimally tuned screened range-separated hybrid functional was shown to be highly accurate for fundamental band gaps of standard semiconductors and insulators. This was achieved by selecting the parameters of the functional to satisfy an ansatz that generalizes the ionization potential theorem to the removal of charge from a state that corresponds to a Wannier function. Here, we present applications of the method to the band gaps of typical halide perovskites. We find a satisfyingly small formal mean absolute error of ∼0.1 eV with respect to experimental band gaps and very good agreement with previous many-body perturbation theory calculations.

Halide perovskites from first principles: from fundamental optoelectronic properties to the impact of structural and chemical heterogeneity

“…This result is consistent with other studies employing RSH functionals for computing QP band gaps in standard semiconductors [104,106]. More recently, reference [108] employed the newly developed WOT-SRSH functional to calculate band gaps of organic-inorganic halide perovskites based on Pb and Sn, and obtained agreement with experiment with a mean absolute error of approximately 0.1 eV. Furthermore, the meta-GGA TASK [111] was found to result in band gaps on par with those calculated using HSE, at significantly reduced computational cost [112].…”

“…Most recently, reference [106] used localized Wannier functions to enforce the ionization potential theorem and formulate an inexpensive and non-empirical method of optimally tuning the SRSH parameters. The Wannier OT-RSH functional (WOT-SRSH) developed in reference [106] has since been successfully used to calculate quasiparticle band gaps for standard III-V and II-VI semiconductors and insulators [106], oxides [107] and halide perovskites [108], as well as a starting point for MBPT calculations with improved accuracy [107,109]. Finally, meta-GGAs such as the SCAN [110] and TASK [111] functionals, include a dependence on the KS kinetic energy density and/or the Laplacean of the charge density, and have been constructed to satisfy other analytical properties such as the derivative discontinuity of the exact exchange-correlation functional than typical (semi)local approximations.…”

First-Principles Studies of the Electronic and Optical Properties of Zinc Titanium Nitride: The Role of Cation Disorder

“…The so-called range separation parameter, which dictates the length scale at which the Coulomb potential transitions from its short-range form to its long-range form, is tuned to satisfy an ionization potential (IP) ansatz; at long range, the fraction of exact exchange is set by the averaged high-frequency dielectric constant, here 6.81. WOT-SRSH has been shown to predict the fundamental band gap accurately for standard semiconductors, metal oxides, and halide perovskites, and it has been shown to be an optimal starting point for one-shot GW calculations . Using α = 0.25 (the fraction of exact exchange in short range) and α + β = 0.147 (the fraction of exact exchange in long range, determined by 1 ϵ ∞ ), we determine that a range separation parameter of γ = 1.07 Bohr –1 satisfies the IP ansatz for cation-ordered ZnTiN 2 , following the procedure outlined in prior work (see Supporting Information for details).…”

Accurate and Efficient Computation of the Fundamental Bandgap of the Vacancy-Ordered Double Perovskite Cs2TiBr6

“…KS-DFT methods that correct for (or minimize) self-interaction offer a new approach that is complementary to MBPT and have been found to provide accurate spectral properties of many solids at reduced computational cost. − The motivation for these methods is that the total energy of a system, E ( N ), as a function of the total number N of electrons in that system, ought to be piecewise linear. This is the well-known piecewise linearity condition (PWL) …”