Walukiewicz et al.1 have recently stated that previously reported contributions to the electron mobility of GaAs from space-charge and/or central-cell scattering are in fact insignificant, and that reports of a T−1/2 term in the mobility2,3 are artifacts due to the assumption of Mathiessen’s rule. This conclusion is an overstatement of their results and in fact demonstrably incorrect. First, an analysis of the data reported by Stringfellow2 and Stringfellow and Kuenzel3 as well as others has already been performed by Chattopadhyay et al.4 without assuming Mathiessen’s rule. Their conclusion is that central-cell scattering is indeed significant. Second, the particular data analyzed by Walukiewicz et al. are in fact acknowledged in Ref. 2 to have very little T−1/2 scattering. The magnitude of the scattering cross section for T−1/2 scattering, SCA, for other samples is more than an order of magnitude larger, too large to be ascribed to errors inherent in using Mathiessen’s rule. Experimental data convincingly demonstrate this. The mobility versus temperature curves are lower, especially at higher temperatures, for C as opposed to Zn- or Ge-doped samples where all have the same values of ND+NA (see Fig. 2 of Ref. 3). In addition, recently published data5 for MBE GaAs grown with different C doping levels show that for constant ND+NA, SCA is three times larger for the more highly C doped samples. This could not be due to errors inherent in the use of Mathiessen’s rule. For these samples C clearly introduces an increase in the T−1/2 scattering which is not observed for other acceptors. ufc15xr 1W. Walukiewicz, J. Lagowski, and H. C. Gatos, J. Appl. Phys. 52, 5853 (1981). 2G. B. Stringfellow, J. Appl. Phys. 50, 4178 (1979). 3G. B. Stringfellow and H. Kuenzel, J. Appl. Phys. 51, 3254 (1980). 4D. Chattopadhyay, H. J. Queisser, and G. B. Stringfellow, J. Phys. Soc. Jpn. 49, Suppl. A, 293 (1980). 5G. B. Stringfellow, R. Stall, and W. Koschel, Appl. Phys. Lett. 38, 156 (1981). UFTELTAL G. B. Stringfellow G. B. Stringfellow G. B. Stringfellow The T−1/2 term in electron mobility for GaAs and AlxGa1−xAs is shown not to be solely due to errors inherent in the use of Mathiessen’s rule.
Read full abstract