Effect of high-temperature annealing on the residual strain and bending of freestanding GaN films grown by hydride vapor phase epitaxy

Abstract

The effect of high-temperature high-pressure annealing on the residual strain, bending, and point defect redistribution of freestanding hydride vapor phase epitaxial GaN films was studied. The bending was found to be determined by the difference in the in-plane lattice parameters in the two faces of the films. The results showed a tendency of equalizing the lattice parameters in the two faces with increasing annealing temperature, leading to uniform strain distribution across the film thickness. A nonmonotonic behavior of structural parameters with increasing annealing temperature was revealed and related to the change in the point defect content under the high-temperature treatment.

Lattice parameters of bulk GaN fabricated by halide vapor phase epitaxy

“…Finally, in the case of HVPE FS GaN substrates some residual strain may be present as a result of the heteroepitaxial nature of the growth. This small residual strain, often demonstrated by a bending of the FS GaN materials [17,18], was shown to vary depending on the thickness of the FS substrate and the particular nucleation scheme used [19].…”

Introduction and recovery of Ga and N sublattice defects in electron-irradiated GaN

“…N vacancies that are also produced in the 2 MeV electron irradiation, although at a lower rate, and hydrogen are possible candidates as it would be difficult to distinguish Ga vacancy complexes with either ͑or both͒ of them from the isolated Ga vacancies. 20 In addition, a significant part of the Ga vacancies cannot be complexed with oxygen as the O concentration is too low in our samples, and the breakup of V Ga -O N complexes has been shown to occur only above 1500 K. [29][30][31] Interestingly, recent results of ODEPR measurements 7 ͑in high-purity HVPE GaN samples irradiated with 2.5 MeV electrons to a similar fluence as in our work͒ were interpreted as isolated Ga vacancies surviving the annealing at 600 K and disappearing above 800 K. However, the effect of hydrogen and Ga interstitials, which are likely to be positively charged and thus invisible to positrons, should be considered in more detail, and further experiments are needed here as well to clarify the situation ͑note that the H and Ga interstitials will be partly charged and partly neutral, since the Fermi level is likely to be close to their energy levels͒.…”

Freestanding 2-in GaN layers using lateral overgrowth with HVPE

“…The defective interface region with a high dislocation density [3,4] and microcracks [4] were proposed to be the origin of the bow due to a change in lattice constants between the backside of the freestanding layer to the defect-reduced surface. Another proposed explanation relies in the inhomogeneous distribution of point defects like gallium vacancies V Ga and complexes V Ga ÀO N within the layer [11,12]. There is no apparent and straightforward way to avoid these potential causes for bowing of the separated layers.…”