Photocurrent Spectroscopy of Interface States at a Semiconductor‐Electrolyte Junction

Abstract

We have investigated the photocurrent induced by subbandgap light irradiation at the n-Si-aqueous electrolyte junction., The observed effect is shown to arise from optical processes involving interface states, and mainly from electron transitions from the filled interface states to the conduction band. The dependence of the photocurrent upon anodic voltage is accounted for quantitatively by considering the Schottky barrier lowering effect, already measured from the dark I-V characteristics. The dependence of the photocurrent upon photon energy enables us to extract a density of states curve for the occupied interface states. It suggests that these states are bound states on the semiconductor side, due to the coulomb potential of adsorbed ions. Complementary observations on oxidized and bandgap irradiated junctions are discussed and are found to fit well in the above picture.Experimental evidence for the role of interface states at the semiconductor-electrolyte junction has been brought in many instances, mostly through capacitance and recombination studies (1, 2). These techniques however, hardly permit a systematic investigation of interface states through the semiconductor gap. A true spectroscopy of these states might rather be obtained by observing the photocurrent induced by subbandgap optical transitions between these levels and the semiconductor bands (3, 4). This technique however, suffers from two drawbacks: (i) the expected effects are rather small; (ii) larger effects may be present in the same photon energy range, for example onset of bandgap absorption, transitions involving deep levels in the bulk or the space charge region, and also thermal effects (5). We present here the results of an investigation of the photocurrent induced by subbandgap light at the n-siliconaqueous electrolyte junction, where interface states are known to be present (2, 6). We show that they are indeed responsible for the observed effect and their energy distribution can be attained from the experimental data.

Experimental study of the n-silicon/acetonitrile interface: Fermi level pinning and surface states investigation

“…This is verified by the rectifying /(F) characteristics obtained for nearly all of the (26) The present data might alternately be interpreted by invoking a filling (change of population) of preexisting surface states upon contact with the redox electrolyte. We think this alternate interpretation can be dismissed for the following reasons: (i) For a given redox system (e.g., NB/NB" or AQ/ AQ") the characteristic time for pinning seems to be rather insensitive to the concentrations of the redox species, (ii) Bandgap irradiation of oxidized surfaces causes hole trapping in the surface states, thereby decreasing the magnitude of the subbandgap photoeffect (see ref 17 and 25); turning the bandgap light on and off hence provides a technique for measuring directly the equilibration time of the surface states population. This time is found to be smaller than the lock-in response time (0.3 s) except for strongly oxidized surfaces.…”

Photoelectrochemical Charge Transfer Properties of Electrodeposited CdSe Quantum Dots

“…This sub- E g response is even more pronounced in the 30 nm film (at least 3 nanocrystal layers, containing some crystals which are no longer size-quantized (larger than 12 nm) and should have a band gap of 1.74 eV); it extends considerably beyond the bulk band gap value. Sub- E g response has been observed previously in semiconductor photoelectrodes, and attributed to absorption in intra-band-gap states. − …”

Photoelectrochemical Properties of CH3-Terminated p-Type GaP(111)A

“…The predominant mode for photocurrent generation in a semiconductor photoelectrode under illumination is through absorption of suprabandgap light, i.e., promotion of excited electrons and holes into the conduction and valence bands, respectively (Figure ). However, sub-bandgap light can also induce measurable photocurrent if there is an appreciable density of midgap states in which charge carriers can be promoted to/from. , For the commercial grade crystalline GaP used in this study, the prevalence of midgap states in the bulk and surface of GaP has been identified previously and shown to induce detectable levels of photocurrent . Since both the absorption cross-section and the total density of such defects are small, the expected sub-bandgap photocurrents can be vanishingly small.…”

“…However, sub-bandgap light can also induce measurable photocurrent if there is an appreciable density of midgap states in which charge carriers can be promoted to/ from. 44,45 For the commercial grade crystalline GaP used in this study, the prevalence of midgap states in the bulk and surface of GaP has been identified previously 40 and shown to induce detectable levels of photocurrent. 46 Since both the absorption cross-section and the total density of such defects are small, the expected sub-bandgap photocurrents can be vanishingly small.…”