homeresources & publications › analysis of the objectionable n-background conductivity in zno crystals


Analysis of the objectionable n-background conductivity in ZnO crystals


Due to its physical properties (e.g., high excitonic binding energy, high saturation velocity, high piezoelectric constant) the wide-band-gap semiconductor ZnO is a very promising material for electronic and optoelectronic devices. Such devices include high power, high speed or high temperature electronics, chemical sensors and blue or UV light emitters. However, most of these applications are currently hampered by the lack of control over the electrical conductivity. As-grown ZnO usually exhibits n-type conductivity and the cause for the background doping is heavily debated. Possible origins for the n-background doping of ZnO are unintentionally incorporated impurities (e.g., F, Cl on the O site, B,Al,Ga on the Zn site; or hydrogen), intrinsic defects (e.g., antisite, interstitial, vacancy), or extended structural defects. We have performed temperature dependent Hall measurements and materials characterization (SIMS) on nominally undoped bulk ZnO crystals as well as material doped with potential candidates for p-doping (e.g. N, P, Li, Cu, K, Ag). The main goal of these investigations is to explore the cause of the background doping and to study the impact of possible candidates for p-doping on the ZnO material. Hall measurements provide the carrier concentration, mobility and the activation energy of the defects responsible for the background doping. Many of the samples doped with possible p-type dopants exhibited an n-type background that is even higher than what we find in the undoped material. Surprisingly, we obtained almost the same activation energy for various doped samples, suggesting that the cause of the background doping is the same in all samples. We used SIMS to check for unintentionally incorporated impurities. We particularly looked for a species with a high concentration that could be correlated with the electron concentrations in the various samples. To date no systematic correlation between electrical and SIMS data has been found. This negative result, combined with the observation that p-dopants increase often the n-background, suggest that defect associated with structural defects (native defects or extended defects) are responsible for the n-type background. We also found a second category of doped samples which show an essentially different electrical behavior. These samples have high resistivity, and Hall measurements are strongly influenced by the sample ambient. We explain this behavior in terms of a surface conducting channel that exists reproducibly under certain ambient conditions and dominates the electrical properties of these ZnO samples. These results may shed light on the general difficulty to achieve reproducible p-type ZnO. #This work was supported by AFOSR under grant # F49620-02-1-1163, monitored by Lt. Col. Dr. T. Steiner.


Schmidt, O. ; Kiesel, P. ; Van de Walle, C. G. ; Johnson, N. M. Analysis of the objectionable n-background conductivity in ZnO crystals. 47th Electronic Materials Conference; 2005 June 22-24; Santa Barbara; CA; USA.