Selective Area Growth and Structural Characterization of GaN Nanostructures on Si(111) Substrates

Abstract

Selective area growth (SAG) of GaN nanowires and nanowalls on Si(111) substrates with AlN and GaN buffer layers grown by plasma-assisted molecular beam epitaxy was studied. For N-polar samples filling of SAG features increased with decreasing lattice mismatch between the SAG and buffer. Defects related to Al–Si eutectic formation were observed in all samples, irrespective of lattice mismatch and buffer layer polarity. Eutectic related defects in the Si surface caused voids in N-polar samples, but not in metal-polar samples. Likewise, inversion domains were present in N-polar, but not metal-polar samples. The morphology of Ga-polar GaN SAG on nitride buffered Si(111) was similar to that of homoepitaxial GaN SAG.

Eutectic Formation, V/III Ratio, and Controlled Polarity Inversion in Nitrides on Silicon

“…Focused ion beam (FIB) milling was used to prepare cross‐sectional lamellae from each of the samples for scanning transmission electron microscopy (STEM). Details of the FIB preparation were given in a previous study . In this study, none of the samples were Ar‐ion milled, but instead all received a final mill at ±3° with a 5 kV:20 pA Ga beam.…”

“…Details of the FIB preparation were given in a previous study. [29] In this study, none of the samples were Ar-ion milled, but instead all received a final mill at AE3 with a 5 kV:20 pA Ga beam.…”

“…[11,12,30] We have previously reported defects at the Si/AlN interface under Al-rich growth conditions, which are consistent with eutectic formation. [29,30] In addition, we have shown that the Si in the eutectic layer on the AlN surface can be incorporated into the film if the growth is changed from metal-rich to either stoichiometric or N-rich, and, if sufficient Si is incorporated, it can lead to polarity inversion. [30] This is in agreement with reports that silicon on the order of an ML causes polarity inversion in GaN.…”

“…In fact, rather than an SiN x layer, the formation of Al–Si eutectic has been reported for PAMBE growth of AlN on both Si(111) and SiC substrates, and it has been proposed that the eutectic is responsible for a liquid Al–Si layer floating on the AlN surface during growth under Al‐rich conditions …”

“…[19] Similarly, no SiN x was found at the Si/AlN interface when a Si(111) substrate was exposed to active nitrogen prior to PAMBE AlN deposition. [27] In fact, rather than an SiN x layer, the formation of Al-Si eutectic has been reported for PAMBE growth of AlN on both Si(111) and SiC substrates, [11,12,28,29] and it has been proposed that the eutectic is responsible for a liquid Al-Si layer floating on the AlN surface during growth under Al-rich conditions. [11,12,30] We have previously reported defects at the Si/AlN interface under Al-rich growth conditions, which are consistent with eutectic formation.…”