Deep levels doping effects

Systematic investigations on the dependence of the PPC properties on different growth conditions are still needed to elucidate the nature of the deep level defects which are responsible for PPC. Needless to say, the future development of GaN devices depends critically on the improvements in impurity doping, which would rely heavily on the full understanding of the physics of doped impurities. For many device applications, it is important to eliminate (or minimise) effects of deep level impurities through improved crystal growth processes and device designs. 

The presence of a shallow acceptor level in GaN has been attributed to C substituting on an N site by Fischer et al [7], In luminescence experiments on GaN from high temperature vapour phase epitaxy in a C-rich environment donor-acceptor and conduction-band-to-acceptor transitions have been distinguished in temperature dependent experiments. From the separation of both contributions an optical binding energy of 230 meV close to the value of effective mass type acceptors was obtained. Hole concentrations up to 3 x 1017 cm 3 were achieved by C doping with CCU by Abernathy et al [10], In addition Ogino and Aoki [17] proposed that the frequently observed yellow luminescence band around 550 nm should be related to a deep level of a C-Ga vacancy complex. The identification of this band, however, is still very controversial. 

In a study that addressed the effect of doping on quantum dots, the donor and acceptor levels were found to be practically independent of particle size [De3]. In other words, shallow impurities become deep ones if the dot size is reduced. Experimental observations show that the luminescence is not affected by doping if a thermal diffusion process, for example using a POCl3 source, is used [Ell]. Implantation, in contrast, is observed to effectively quench the PL [Tal4]. If the pores are filled with a medium of a large low-frequency dielectric constant, such as water or any other polar solvent, it is found that deep impurity states still exist,... 

In high resistivity GaN films, Glaser and co-workers observed ODMR on a broad band peaked at 3 eV and observed a second deep donor state at g = 1.978 [29]. Because the g value is between the effective mass donor and that of the first deep donor, they argued that the energy level was shallower than the first deep donor. Koschnick and co-workers [30] have also resolved a donor resonance atg - 1.96 in undoped fihns and a possibly different donor (also at g 1.96) in Mg-doped fihns using high frequency (70 GHz) ODMR. The existence of multiple donor levels was first proposed by Gotz and co-workers [31] based on their electrical measurements and these results seem to confirm that idea. 

---