The value of EV is generally clcse to EB at high illumination intensity.The finding of a constant E B fo,' n-CdTe, independent of Eredo x or s, leads to the conclusion that n-CdTe is Fermi level pinned. 2 ' 5 We take the term "Fermi level pinned" to refer to a semiconductor that is measured to have a constant EB, independent of the contacting medium for a wide range of Eredox or 0. The origin of a constant EB can be attributed to surface states situated between the top of the valence band, EVB, and the bottom of the conduction band, ECB. The density and distribution of surface states can control the value of EB as a function of E redox or 0.5,6 When the region between ECB and EVB is free of surface states, EB of an n-type semiconductor is expected to vary with Eredox according to equation (1)7for E redox situated between EVB and ECB whereEFB is the electrochemical potential of the semiconductor, Ef, when there is no band bending. When Ef = EFB there is no accumulation or depletion of the majority charge carrier at the surface and E B = 0. But even for surface-state-free semiconductors, equation (1) Pyrex tube and all surfaces but the exposed front surface of CdTe were sealed with ordinary epoxy. The exposed surface of the CdTe was the (111)face. Just prior to use all electrodes were etched and cleaned as described below.Etching Procedures. Two different types of etchants were used, oxidizing and reducing. An oxidizing etch was one of the following: (i) 4 g K 2 Cr 2 0 7 , 10 ml concentrated HNO 3 , and 20 ml H 2 0; 12 (ii) 1/1/l by volume saturated solution of K 2 Cr20 7 , concentrated HCl, and concentrated HNO 3;13 (11) 3/2/1 by volume solution of 48% HF, 30% H 2 0 2 , and H 2 0 and then briefly rinsed with concentrated HCl;14 (iv) 3/2/1/0.01 by volume of concentrated HNO 3 , concentrated H 2 SO 4 , glacial aceic acid, and concentrated HCl; (v) 2/1/l by volume solution of 48% HF, concentrated HNO 3 , and glacial' acetic acid. 15 For an oxidizing etch the CdTe electrode was immersed in the etchant for Abbreviations for redox couples are MV 2+ E N,N'-dimethyl-4,4'-bipyridinium; gun was used to produce a 2 keV Ar ion beam for sputtering. The pressure was maintained at 3 x -10 torr in the main vacuun, chamber and 1.5 x 10 -4 torr of Ar in the ionization chamber, while sputtering.X-ray photoelectron spectra (XPS) were obtained ona Physical Electronics model 548 spectrometer with a magnesium anode. The broad scans (0-1000 eV)were recorded with a pass energy of 100 eV and the narrow scans with a pass energy of 25 eV. The peak energies at the Cd and Te peaks were referenced to the Cls binding energy (284.6 eV) to correct for charging. Samples were mounted as above and sputtering was oune with a 5 keV Ar+ beam, after introducing Ar into the vacuum chamber to bring the pressure to -7 x 10 torr.Elements detected by Auger and XPS were identified by reference to data previously reported using these techniques. 17 ' 18. .
Results
a.Determination of EV vs. E redox for n-CdTe. Scheme I illustrdte', the interface enerqetLi...