Influence of C, N and O Ion-Implantation on Yellow Luminescence
- PDF / 481,404 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 89 Downloads / 137 Views
YL. A. Polyakov et al.[13], A. Ishibashi et al.[14] and R. Niubuhr et al.[15] observed the correlation of the YL in GaN grown by MOVPE to residual carbon impurities in GaN films. Other chemical and physical origins, such as a NGa antisite[16], a Ga vacancy[17], a N vacancy[10,18] , a Ga interstitial[19], chemical impurities[20,21], and other physical defects[l 1,22,23] have been suggested to lead to the presence of the YL band. HVPE does not use a C-containing growth source to synthesis GaN and HVPE-grown GaN, under conventional growth conditions, shows negligible or very weak YL[24,25,26]. The HVPE technique has been used to study the influence of intentional introduction of specific impurities or growth conditions on the YL in GaN[27,28,29]. Both intentional C-doping with C3H 8[27] and H2 -addition into the ambient [28,29] during HVPE growth can induce a strong YL in GaN, indicating that there are multiple origins for the spectral features in the wavelength regime associated with the YL band. In each case, multiple transition channels involved within the YL band have been observed. In this paper, we investigate the luminescence properties of the C-, N- and O-implanted HVPEand MOVPE-grown GaN. For HVPE-grown GaN, a substantial enhancement of YL can be observed after carbon implantation. A variable-temperature photoluminescence (PL) measurement reveals that the intensity of this YL band monotonically decreases and the fullwidth-half-maximum (FWHM) of the YL band linearly increases with the increasing measurement temperature from 20K to 293.5 K. The YL peak energy changes little over that same measurement temperature range. These results can be explained by a two-channel transition model. Implantation of 0 or N does not change the YL appreciably but severely decreases the BE. For MOVPE-grown GaN, all ion-implanted samples exhibit a significant loss of overall integrated PL intensity. Residual C contamination in the MOVPE samples could be large compared to the additional ion-implanted carbon leading to a reduced impact of the ionimplanted carbon[13,14,15]. EXPERIMENTS AND RESULTS All HVPE-grown GaN samples were fabricated on (0001) sapphire substrates in a horizontal atmospheric pressure HVPE system. The details of the reactor and growth technique were previously described[25,26]. NH 3 was used as the nitrogen source and the mole fraction of NH 3 was kept approximately constant at XNH 3 _ 0.09. HC1 was reacted at a high temperature with the liquid Ga to form GaCI. The GaC1 was transported to the substrate, which was held at a higher temperature. Temperature of the Ga boat was kept at 850'C while the substrate temperature was 1050TC. The input ratio of NH 3-to-HCl was 30. No intentional dopant was used during growth. The growth rate under that condition was -2.0 iim/min. A 3 jim thick GaN/sapphire sample was used in the ion-implantation experiment. The surface morphology of that sample is generally flat mirror-like. The full-width-half-maximum (FWHM) of the X-ray rocking curve (RC) is 1020 arcsec. Room-temperature Ha
Data Loading...
