Deciphering 3d Transition Metal Dopants in GaN (and Wide Band Gap Materials) using Ground State and Excited State Density Functional Theory

Doping of gallium nitride (GaN) and (ultra-)wide band gap, (U)WBG, materials is used to control the Fermi level and create functional defects for optoelectronic and spintronic applications. Harnessing the greater temperature tolerance, faster switching speeds, and larger breakdown voltages possible with III-nitrides would profoundly improve the performance of power electronics. To exploit the technological potential of 3d-doped (U)WBG requires greater understanding of 3<i>d</i> defects than currently exists. Firm identification and characterization of defects in (U)WBG remains fraught. The mismatch between what experiment can measure and definitively identify and what theory can reliably compute is a particularly acute challenge in 3<i>d</i>-doped GaN. Using a generalized gradient approximation (GGA) to density functional theory (DFT) along with a non-jellium local moment countercharge (LMCC) treatment of the charge boundary conditions, we demonstrate accurate predictions of defect levels of magnetic 3d dopants in GaN unhindered by a band gap problem and in agreement with available experimental data. We implement an occupation-constrained DFT (occ-DFT) to self-consistently obtain excited state total energies, and thereby predict vertical and adiabatic optical excitation energies corresponding to photoluminescence (PL) data. These GGA excited-state results agree quantitatively with PL measurements, whereas literature hybrid functional calculations fail qualitatively. In concert, ground state (defect levels) and excited state (PL) calculations mandate wide reinterpretation of GaN experimental literature. The Mn dopant is shown to take charge states as deep as (2+), with corroborating evidence anchored at one end by accurate prediction of the 1.4 eV PL for the Mn(0) and the other end by accurate prediction of PL for a <i>d</i>² Mn(2+) defect. The 1.19 eV PL in n-type GaN, attributed to Cr⁺⁴ and touted as a promising candidate for a quantum color center, cannot be the Cr(1+)–the Cr(1+) is stable only in midgap. A combined ground-state/excited-state modeling approach enables more confident chemical identification of defects in (U)WBG, aiding the search to identify, characterize, and design doping regimens to create functional defects with auspicious electronic properties.<br/>— Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA0003525. Air Force Research Laboratory gratefully acknowledges the support of the Air Force Office of Scientific Research (AFOSR) through Contracts No. FA9550-17RVCOR505 and No. FA9550-21RVCOR503.