Effect of Indium Surfactant on the Optical and Structural Properties of MBE-grown GaN

WK Fong,CF Zhu,C Surya,BH Leung,CC Cheng,B Sundaravel,EZ Luo,JB Xu,IH Wilson
DOI: https://doi.org/10.1117/12.392186
2000-01-01
Abstract:A small indium flux was used as a surfactant during the growth of gallium nitride films by rf-plasma assisted molecular beam epitaxy. The effects of the indium surfactant on the optical and structural properties of undoped GaN were studied by photoluminescence spectroscopy, high-resolution X-ray diffraction, atomic force microscopy, Rutherford backscattering spectroscopy, and low-frequency noise. Photoluminescence spectra show that GaN thin films grown in the presence of In surfactant exhibit suppressed yellow luminescence compared to films grown under the same experimental conditions but without In surfactant. The X-ray rocking curves demonstrate a 20% decrease in the full width at half maximum value for the films grown with In surfactant. AFM studies show that the root mean squared roughness for films grown with and without In surfactant are 5.86 and 6.99 nm respectively, indicating significant improvement in surface morphology. RES and ion channeling shows the presence of stacking faults and dislocations in GaN grown with In surfactant and stacking faults in GaN grown without In. For the characterization of defect properties in the films we conducted detailed studies of l/f and Generation-Recombination (C-R) noise on our samples. About 65% reduction in the Hooge parameter was observed in the Film grown with In surfactant, indicative of a corresponding reduction in defect states within the material. For f > 500 Hz, G-R noise was the dominating fluctuation process. Detailed characterization of the G-R noise over a wide range of temperatures enabled the determination of the energy levels of the traps responsible for the G-R noise. Three different trap levels were observed using noise measurement technique. For films grown without In surfactant, traps with activation energies 806 meV, 241 meV and 100 meV were observed, whereas for samples grown with In surfactant traps were observed at 666 meV, 208 meV and 90 meV. Such systematic reduction in the energy levels of the traps may arise from the relaxation of strains in the material when grown under the application of In surfactant.
What problem does this paper attempt to address?