Thermal Desorption of Deuterium from GaN(0001)
- PDF / 106,243 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 56 Downloads / 194 Views
Thermal Desorption of Deuterium from GaN(0001) Y. Yang, J. Lee, B. D. Thoms, Georgia State University, Atlanta, GA D. D. Koleske, R. L. Henry, Naval Research Laboratory, Washington, DC Abstract The recombinative desorption of deuterium from GaN(0001) has been investigated using temperature programmed desorption (TPD) with variable heating rates. With a heating rate of 1 °C/s, molecular deuterium desorption peaks at 410 °C in agreement with related previous work. However, the shape of the curve indicates a secondary peak at around 280 °C which is merged into the lower temperature shoulder of the dominant peak. By changing linear heating rate from 0.05 °C/s to 8 °C/s desorption peak temperatures from 380°C to 570°C were observed. Fitting to a pseudo-first-order desorption model results in a hydrogen desorption barrier, Ed, from surface of 1.1eV and a pre exponential factor, ν, of 2 x 106 s-1. Both are below expected values and are assumed to be due to a variation of desorption barrier heights. If a typical pre-exponential factor of 1 x 1013 s-1 is assumed, reanalysis of the desorption data produce a desorption barrier of 2.0 eV, in agreement with the existence of the surface adsorption barrier at room temperature. Introduction GaN, a direct, wide bandgap semiconductor, shows advantages for use in light emitting devices and high temperature, high power, and high frequency transistors [1-3]. The reaction between hydrogen and GaN is important since it alters growth and processing. Hydrogen affects growth rate and quality, both in MBE [4] and MOCVD [5,6,7]. Several researchers [4,6,7] have suggested that hydrogen bonds to the surface during growth thereby affecting deposition reactions. Also p-type GaN growth studies show that the resistivity of samples grown by hydrogen-rich methods like MOCVD [8] can be millions of times higher than those grown by hydrogen-free processes like MBE [9]. Researchers conclude that hydrogen forms Mg-H complexes on anti-nitrogen sites causing passivation and that it will be removed to second closest nitrogen sites by 400 °C annealing [8,10,11,12]. It is still not clear how hydrogen is eventually liberated from the neighborhood of the Mg site restoring dopant activation. The rate and quality of GaN dry-etch processing [13,14,15] are affected by hydrogen too. Pearton et al. suggest a two-fold effect of hydrogen in dry etching of GaN. It either diffuses into the sample bulk and passivates dopants or defects causing an increase in resistivity or it produces a preferential loss of nitrogen on the GaN surface enhancing the etching process by removing nitrogen as ammonia [14]. Studying more about hydrogen reactions with the GaN surface would be helpful in understanding its effects during dry etching. The effects of hydrogen on GaN growth, doping, and processing motivate the present study of hydrogen desorption from the GaN surface. Previous work in the author’s lab used high resolution electron energy loss spectroscopy (HREELS) and energy loss spectroscopy (ELS) to study the adsorption and desorptio
Data Loading...
