Experimental tests of the classical model

The simplicity of the two-component equivalent circuit can be expected to be reflected in practice provided 

Under these circumstances, we might expect that the whole of the a.c. component of the potential should appear in the bulk of the semiconductor, across the depletion layer. In addition, we may replace — by ps — (p, where is the flat-band potential and pB is the actual d.c. potential applied to the semiconductor. 

a plot of (Csc) 2 vs. / B should be linear with an intercept at the flat-band potential. This linearity can be used as an immediate test of the theory and an example is shown in Fig. 20 for p-GaAs [62]. The capacitance derived from the two-component circuit is almost independent of frequency for potentials more than 0.5 V from the flat-band potential and the intercept is clearly defined. Similar results for both n-GaAs and p-GaAs are shown in Fig. 21 [63] and an important check on the theory is that the flatband potentials for the n- and p-type electrodes differ by ca. 1.4 V, corresponding to a bandgap of 1.4 eV for GaAs. This is expected as the Fermi level will be very close to the band edges in the bulk for these wide band materials. 

Regrettably, such ideal behaviour as that exhibited in Figs. 20 and 21 is very rare. In almost all materials studied, deviations from the simple theory are encountered. There are four commonly observed cases  

A large number of corrective treatments has been derived for the semiconductor-electrolyte interface and we shall consider these below. The complexity of some of the models considered has, however, led to considerable doubt as to the appropriateness of some of the treatments. 

---