SEMITIP V6, UniInt2, Example 1: n-type GaAs(110), with intrinsic surface states

Click here for input/output files for Example 1

The first voltage point of the FORT.9 input file and FORT.16 output file here are the same ones as with the VERSION 4 source files. This example illustrates the so-called "dopant induced" component in the current for highly doped n-type GaAs(110) (1018 cm doping concentration). A Gaussian-shaped band of intrinsic surface states (i.e. associated with the Ga dangling bonds on the surface) are placed at 1.75 eV above the valence band maximum, with width (FWHM) of the band of 0.25 eV. These surface states restrict the formation of an accumulation layer, so that essentially all of the current throughout the gap originates from extended states in the conduction band that are occupied due to the n-type doping. Output for the conductance goes to FORT.93, FORT.94, and FORT.15 for the valence band (VB), conduction band (CB), and their sum, respectively. When we plot the total conductance (sum of conductance from extended states plus localized states) in each case, we find:

where the black circles show the sum of VB and CB, red x-marks show the VB, and green x-marks show the CB. This result is the same as in Fig. 2(a) of Phys. Rev. B, 075320 (2009). It should be noted that in this paper the temperature dependence of the surface state occupation was neglected. This same approximation is made in the present example, i.e., the indicator on line 22 for FORT.9 is set to 0. If we include this temperature dependence by setting the indicator to 1, then the results for the conductance change noticably (since significant occupation of the surface band occurs even when the Fermi-level is kT in energy below the band). However, if we furthermore shift the band edge slightly, then nearly identical results as in the graph are obtained. For example, with the indicator set to 1 we can shift the center of the band to 1.775 eV above the valence band maximum and maintain the same width of 0.25 eV, or change the width to 0.22 eV and maintain the center at 1.75 eV, and in both cases the results are practically indistinguishable from those in the above plot.