lower value of indium density was obtained by C-V and junction breakdown methods, while the higher value was obtained by Hall measurements and curve fitting. Because of uncertainties in the value of the Hall scattering factor, Schroder et al consider the lower value of density more reliable. Note that our theoretical calculation falls between the two values of dopant density reported by Schroder et al [64]. Values of resistivity of indium-doped silicon reported by Backenstoss [16] are about 25 percent higher than our calculated values. The work of Backenstoss [16], however, was done in the high doping region where dopant densities approach the limit of solid solubility. Backenstoss found that for dopant densities greater than 4xlO17 cm-3 there was a considerable amount of indium precipitation. Thus it is possible that part of the discrepancy between our theoretical calculations and the data of Backenstoss is due to the low solid solubility limit of indium in silicon. Recent theoretical results of Sclar [14] for In-doped silicon also agree very closely with our theoretical calculations at 300 K.

     To find out the adequacy of our theoretical model for temperatures other than 300 K, we compared the calculated and measured values of resistivity for silicon samples doped with boron, gallium and indium for temperatures ranging from 100 to 400 K. Figure 7.4 shows the comparison between the theoretical and measured resistivities for borondoped silicon. Except for a couple of data points, agreement between the theoretical and measured values was within 8 percent over the entire range of temperatures. Figures 7.5 and 7.6 show the comparison between theoretical and measured resistivities for gallium- and indium-doped silicon respectively. Agreement here was not as good as in the borondoped case, but except for a couple of data points, agreement between