The optical centers in AlN can frequently exist in various charge states and can be accompanied by many coexisting defect species, creating a complex environment where mutual interactions are inevitable. Therefore, it is an immediate quest to design AlN crystal growth protocols that can target a specific optical center of interest and tune its concentration while preventing the formation of other unwanted point defects. Here, we provide a powerful workflow for point defect engineering in wide band gap, binary semiconductors that can be readily used to design optimal crystal growth protocols through combining CALPHAD-based phase analysis, and ab initio defect calculations. We investigate technologically relevant chromium- and manganese-induced optical centers in AlN, followed by studying the impact of oxygen that can be unintentionally incorporated during crystal growth. We present the dominant defects in all three cases as a function of process parameters along with the optical signatures. In the case of both Cr and Mn doping, the CrAl and MnAl defects are most likely, and increasing nitrogen partial pressure tends to enhance their concentration. We show that it is possible to use nitrogen fugacity as a tool for tuning the intensity of optical signatures. We calculate the CrAl charge transition levels with respect to the valence-band maximum at 2.60 eV (E+/-), 3.83 eV (E0/-), and 5.41 eV (E-/2-) and electron and hole capture transitions with luminescence bands centered at 2.82, 1.91, and 3.15 eV. Unlike the Cr doping, Mn aggregation is unlikely, and the MnAl-VN is the most abundant defect after MnAl under most synthesis conditions. Oxygen tends to form complexes with VAl, and ON-VAl is a prominent defect following ON, with near-UV emission bands at 3.17, 3.26, and 3.81 eV. Our results agree with the available experimental optical signatures of Cr-, Mn-, and O-related centers and provide pathways on how to tune the luminous intensity of these centers through changes in growth conditions.
Read full abstract