Photocurrent onset

The (photo)electrochemical behavior of p-InSe single-crystal vdW surface was studied in 0.5 M H2SO4 and 1.0 M NaOH solutions, in relation to the effect of surface steps on the crystal [183]. The pH-potential diagram was constructed, in order to examine the thermodynamic stability of the InSe crystals (Fig. 5.12). The mechanism of photoelectrochemical hydrogen evolution in 0.5 M H2SO4 and the effect of Pt modification were discussed. A several hundred mV anodic shift of the photocurrent onset potential was observed by depositing Pt on the semiconductor electrode. 

Photocurrent onset, pH = 10.1, and quasi-Fermi level, 633 nFf - 0.85 V vs SCE, were determined from pH-dependent photocurrent measurements in the presence of methylviologen and glucose... 

The photocurrent onset potential is often taken as the flatband potential, since the measurement of the flatband potential is typically only good to 100 mV and the onset of photocurrent is often observed with less than 100 mV of band bending. This practice is dangerous, however, since the onset potential is actually the potential at which the dark cathodic current and the photoanodic current are equal. Even though in the case of the p-GaP illustration, the observation of an anodic current and a photocathodic current are separated by several hundred millivolts, in many systems these two currents overlap. In those cases, the relationship between the flatband potential and the onset potential becomes unclear. 

One interpretation presumes that the photocurrent onset in the absence of sulfide is determined by electron-hole recombination. The sulfide ions on the surface are then supposed to be bound to these surface recombination levels rendering them unavilable for recombination reactions. The charge transfer reactions could then proceed at lower voltages. In this case the corrosion suppression role of the sulfide ions would be to reduce the oxidized corrosion site before a cadmium ion could go into solution. A variation on this theme is to consider the corrosion site to be the recombination state, i.e., the site on the surface that normally leads to corrosion when oxidized by a photoexcited hole can be... 

When an n-CdS electrode is suddenly illuminated with light capable of producing holes in the CdS, j, would almost immediately reach some large value (equal to or less than the saturation current of curve 1) and then decay to the steady state value of curve 1 as the steady state value of N is approached according to equation 3. If such a transient does not occurthe oxidized corrosion site acting as a recombination state is not the controlling factor in the photocurrent onset. 

Regardless of the nature of the surface state it is clear that it can capture an electron from the conduction band producing cathodic current. This cathodic current balances the anodic current produced when the photoexcited holes produced the oxidized surface state. The net result of these two processes is electron-hole recombination leading to no net current. This recombination process is what controls the voltage of photocurrent onset as can be seen in curve 2 of Figure 5. 

Below, we discuss one more (photoelectrochemical) method for determination of the flat-band potential [40, 172]. The flat-band potential can be determined (i) as the photocurrent onset potential Eomsi (ii) from the dependence of electrode open-circuit photopotential Eocph on light intensity J, as the limiting value of Eoc at a sufficiently high J and (iii) by extrapolating, to zero photocurrent, the potential dependence of the photocurrent jvh squared (see Section 7). These methods are based on the concept... 

To compare these methods, in Fig. 40 we show (a) EIX vs. J and (b) /Ph vs. E plots for a CVD single crystal thin-film electrode. We see that with increase in illumination intensity J, the open-circuit potential E(X approaches a limit of 0.7 V, which is close to the photocurrent onset potential (0.75 V). [The photocurrent density squared vs. potential dependence for this electrode, although far from linear (unlike that of Fig. 38), by the extrapolation to yph -> 0 gives the potential value of approx. 0.65 V.] It is concluded that, on the whole, methods (i) and (ii) are in a good agreement and can be used in the determination of the flat-band potential. Similar results were obtained with HTHP single crystals. 

For all the derivatized electrodes the photocurrent onset is at a wavelength greater than 600 nm, where the photon energy is 2.0 eV in contrast to the 3.1 eV band gap of the semiconductor. For untreated Ti02 electrodes the photocurrent action spectrum follows the absorption edge of the semiconductor, with little response in the visible. The onset potential in all cases is the same as with the untreated electrode (Figure 6). 

In I/E curves the onset of photocurrent is expected from classical theories to occur near the Hatband potential as measured in the dark (Efb (d)), i.e. where the majority carrier current starts too. However, a large shift of the onset potential is seen especially if no additional redox couple is present in the aqueous electrolyte, in cathodic direction for p-, in anodic direction for n-type materials (Fig. 1). This shift depends on the light intensity but saturates already at relatively low intensities (Memming, 1987). If minority carrier acceptors (oxidants for p- and reductants for n-type semiconductors) are added to the solution, the onset can be shifted back to Efb (d) if they have the appropiate redox potential. In principal two types of redox couples can be found those which lead to a shift of the photocurrent onset potential and those which don t. The transition between the two classes occurs at a specific redox potential. 

It should be mentioned further, that the shift of energy bands upon creation of minority carriers does not only occur upon light excitation. Holes can also be produced in the dark via hole transfer from a hole donor into valence band. This kind of process occurs for instance by using the oxidized species of the redox couple [Ru(bipy)3]2+/3+. A corresponding cathodic dark current starts at the same potential at which the photocurrent onset was found in the presence of the reduced form of the same couple. This shift of Efb by hole injection from the electrolyte has been found also with Ce4+ at n-WSe2 (McEvoy et al., 1985) and n-GaAs (Schroder et al., 1985). 

Adding halides to the electrolyte, the shift of bands under illumination and thereby the photocurrent onset potential decreased in accordance with the change in the redoxpotential of the halide couple used. The results can only be understood in terms of Fermi level pinning by the redox couple about 0.5 to 0.8 V below... 

The experimental observations217 of an apparent light intensity threshold for the photocurrent onset have been rationalized218 on the basis that a critical photon flux must be exceeded to counteract the dark current of opposite polarity flowing through the cell. Thus, there appears to be confusion between alternate definitions of a light intensity threshold a threshold for incipient product (say H2) formation and a threshold for product formation in a specific (e.g., standard) state,218... 

Clearly, analytical solutions for cases (1) and (2) are no longer practicable, but numerical solutions can be obtained and some examples are shown in Fig. 75 for a species that can be oxidised both from the valence band and from the surface state [151]. One interesting feature is that the dependence of the photocurrent onset on (f>0 is much reduced another effect that can be observed is that, if there is a redox couple present in solution that can only be oxidised via surface states, the rise of the photocurrent from flat-band potential becomes very sensitive to the value of ksC2 as shown in Fig. 76. 

They checked this mechanism by using TiOj-electrodes instead of particles and found that the 2-phenylindazole was only formed in a small potential range around the photocurrent onset, i.e. at potentials at which also a comparable cathodic current occurs. At more anodic potentials, the hydroxymethyl radical is further oxidized [175]. It should be mentioned here that the oxidation of CH3OH (Eq. (83)) or other organic compounds must not necessarily occur via direct hole transfer. There are strong indications that at first an OH -radical is formed at the surface of Ti02-particles by hole transfer, and that this radical oxidizes the organic molecule in a second step, as proved in the case of acetate oxidation [176]. 

At potentials near the photocurrent onset (roughly Fn,), a spiked response is seen with a characteristic overshoot when the light is turned off. At positive potentials near the plateau regime (again for the specific illustrative case of an n-type semiconductor), the response reverts to a rectangular profile that mimics the excitation waveform. Intermediate response patterns manifest at potentials in between. 

M n (111) photocurrent onset (fig. 1 (n) (b) region I) hole supply, controlled by e — h photogeneration and recombination etch hillocks and ridges... 

The photocurrent in nonfluoride solutions is affected by the amount of preanodic current passed through the sample as shown in Fig. 5.10. It is also seen that the photocurrent onset potential is shifted to more anodic values with formation of an oxide film and the amount of shift is related to the thickness of the film. This shift is due to the potential drop across a growing oxide layer and is one of the reasons for the difference between the photocurrent onset potential and the flatband potential. ... 

---