Very slow decay of a defect related emission band at 2.4 eV in AlN: Signatures of the Si related shallow DX state

Abstract

We report on a defect related luminescence band at 2.4 eV in aluminum nitride bulk crystals, for which we find strong indications to be related to silicon DX centers. Time resolved photoluminescence spectroscopy using a sub-bandgap excitation reveals two different recombination processes with very long decay times of 13 ms and 153 ms at low temperature. Based on the results of temperature and excitation dependent photoluminescence experiments, the process with the shorter lifetime is assigned to a donor-acceptor pair transition involving the shallow silicon donor state, which can be emptied with a thermal dissociation energy of 65 meV. The slower process with a thermal quenching energy of 15 meV is assigned to the slightly deeper Si DX state known from electron paramagnetic resonance experiments, which is transferred back to the shallow donor state.

Slow decay of a defect‐related emission band at 2.05 eV in AlN: Signatures of oxygen‐related DX states

“…Figure shows the resultant PLE spectrum. We find an onset for pumping wavelengths nm (3.5 eV), and a maximum at around 310 nm (4.0 eV), similar to the PLE spectrum we found for the defect band centered at 2.4 eV in a recent publication . We consider this as a strong suggestion, that in both bands the same deep (acceptor) state must be involved, from which free carriers (electrons) are excited.…”

“…We consider this as a hint, that the recombination characteristics are similar as those presented in Ref. . In the model presented there, the carriers recombine from a state spatially indirect to a deep acceptor for low temperatures, and via a donor‐acceptor pair transition for higher temperatures.…”

“…For comparison, k of the generically similar PL band at 2.4 eV studied in Ref. is equal to (K) and to (K). The errors represent the uncertainty of the fitting process, only.…”

“…Besides the high symmetry lattice position, shallow donors can capture a second electron and undergo a so‐called large lattice relaxation and move into a position with lower total energy. This state is labeled “DX center.” Unlike the intensively studied case of silicon in AlN , for oxygen the energy for both its shallow and state are not experimentally known up to now. Theoretical calculations on oxygen DX centers were published by Silvestri et al.…”

“…Based on a combination of doping studies and density functional theory (DFT) calculations, they assigned it tentatively to a transition involving the neutral deep state of substitutional carbon, ; . A band centered at slightly lower energy of 2.46 eV found in similar bulk crystals was recently assigned to donor‐acceptor‐pair (DAP) transitions between the shallow and related DX states of the silicon donor and a deep acceptor, the latter also likely related to carbon .…”