Investigation of Growth Sectors in Gallium Nitride Substrate Wafers from Ammonothermal and Patterned Hvpe Growth Methods

Abstract

Gallium Nitride (GaN) is a wide band gap semiconductor material that has the potential to replace silicon (Si) in power electronics. However, the quality and availability of bulk GaN substrates is currently limiting its use in high-performance power electronics. Several methods have been developed to grow bulk GaN crystals, including physical vapor transport (PVT) [1, 2], Na-flux [3], ammonothermal [4, 5], and hydride vapor phase epitaxy (HVPE) [6]. Additionally, patterned HVPE growth methods have been developed to produce GaN substrates with reduced dislocation densities [7]. Ammonothermal and patterned HVPE growth methods have shown promise in producing high-quality GaN crystals for use in power devices [8-10]. Ammonothermal-grown GaN has an overall low dislocation density of 1000 cm-2, while patterned growth can concentrate dislocations into small areas, leaving the rest of the crystal relatively dislocation-free (< 102 cm-2). Besides dislocations, the overall homogeneity of these wafers is also important for devices. Recent synchrotron X-ray topography studies revealed domain features in both ammonothermal and patterned HVPE GaN wafers. Using synchrotron X-Ray Plane-Wave Topography (SXPWT, previously known as synchrotron X-Ray Rocking Curve Topography), these domain features that have enclosed diamond or quadrilateral shapes are found to have relative compressive strain values (~ 10-5) (Figure 1b). These features are visible on SEM micrographs using a Robinson detector (RBSD) (Figure 1c). The origin of these strains is likely the different impurity incorporation rates in different growth sectors.In the patterned HVPE growth method, growth sectors have been reported to exist due to growth along different directions [11]. It is likely that a similar situation exists for ammonothermal substrates, although no studies have been reported on this. Understanding the nature of these domain areas is important for device fabrication as the inhomogeneities will likely be replicated into active regions. SIMS measurements are currently being conducted to characterize the differences in impurity concentrations in these domains. These results will be used to confirm the origin of the strain.Figure 1. (a) Synchrotron monochromatic beam X-ray topograph of a selected area on a patterned HVPE grown GaN substrate; (b) Cropped strain map of the same area obtained by SXPWT; (c) SEM image of the same area with RBSD contrast. All three images show contrast from the domains defined by the growth sectors.[1] H. Wu, J. Spinelli, P. Konkapaka, M. Spencer, MRS Online Proceedings Library Archive 892 (2005).[2] D. Siche, D. Gogova, S. Lehmann, T. Fizia, R. Fornari, M. Andrasch, A. Pipa, J. Ehlbeck, Journal of crystal growth 318 (2011) 406-410.[3] M. Aoki, H. Yamane, M. Shimada, S. Sarayama, F.J. DiSalvo, Journal of crystal growth 242 (2002) 70-76.[4] R. Dwiliński, R. Doradziński, J. Garczyński, L. Sierzputowski, A. Puchalski, Y. Kanbara, K. Yagi, H. Minakuchi, H. Hayashi, Journal of Crystal Growth 310 (2008) 3911-3916.[5] T. Hashimoto, F. Wu, J.S. Speck, S. Nakamura, Journal of Crystal Growth 310 (2008) 3907-3910.[6] H.P. Maruska, J. Tietjen, Applied Physics Letters 15 (1969) 327-329.[7] T. Nakamura, K. Motoki, GaN substrate technologies for optical devices, IEEE, vol 101, 2013, p. 2221.[8] B. Raghothamachar, Y. Liu, H. Peng, T. Ailihumaer, M. Dudley, F.S. Shahedipour-Sandvik, K.A. Jones, A. Armstrong, A.A. Allerman, J. Han, H. Fu, K. Fu, Y. Zhao, Journal of Crystal Growth 544 (2020) 125709.[9] Y. Liu, B. Raghothamachar, H. Peng, T. Ailihumaer, M. Dudley, R. Collazo, J. Tweedie, Z. Sitar, F.S. Shahedipour-Sandvik, K.A. Jones, Journal of Crystal Growth 551 (2020) 125903.[10] Y. Liu, H. Peng, T. Ailihumaer, B. Raghothamachar, M. Dudley, Journal of Electronic Materials 50 (2021) 2981-2989.[11] K. Motoki, T. Okahisa, S. Nakahata, N. Matsumoto, H. Kimura, H. Kasai, K. Takemoto, K. Uematsu, M. Ueno, Y. Kumagai, Journal of Crystal Growth 237 (2002) 912-921. Figure 1