Photocorrosion anodic

Suppression of Photocorrosion of Photoanodes and Manipulation of Kinetics for Anodic Processes... 

Several groups have recently shown (36,42,43,44) that photoanode materials can be protected from pRotoano3ic corrosion by an anodically formed film of "polypyrrole".(45) The work has been extended (46) to photoanode surfaces first"Treated with reagent that covalently anchors initiation sites for the formation of polypyrrole. The result is a more adherent polypyrrole film that better protects n-type Si from photocorrosion. Unlike the material derived from polymerization of I, the anodically formed polypyrrole 1s an electronic conductor.(45) This may prove ultimately important in that the rate of ionTransport of redox polymers may prove to be too slow... 

Photocorrosion can be prevented by adding a redox couple to the electrolyte whose potential is more favourable than the decomposition potential such that the redox reaction occurs preferentially. When n-CdS is used as photoanode in aqueous electrolytes, the electrode is photocorroded since the reaction, CdS -1- 2h - S -1- Cd, occurs readily. By adding NaOH and sodium polysuphide to the electrolyte (Ellis et al, 1976), photocorrosion is prevented. The /S redox couple preferentially scavenges the photoholes. At the anode, sulphide is oxidized to polysulphide (free sulphur) and free sulphur is reduced back at the dark cathode. Similarly n-Si anodes have been stabilized by using a nonaqueous electrolyte containing a ferricinium/ferrocene redox couple (Legg et al, 1977 Chao et al, 1983). Unfortunately, a similar stabilization technique cannot be applied to photoelectrolysis cells. Some examples of electrode... 

KC1, which bathed the CdS film. This system was also investigated by cyclic voltammetry both in the dark and under illumination. Starting at about — 0.9 V, the dark cathodic current exhibited a peak at — 1.15 V due to Cd2+ reduction and then rose to — 1.4 V as a result of hydrogen production. The observed anodic peak at — 0.85 V was attributed to the stripping of cadmium deposits in the lattice (Cd ). Cyclic voltammetry subsequent to illumination resulted in the appearance of cathodic waves at — 1.0 V and — 1.3 V, at the expense of that at — 1.15 V. The anodic peak broadened, as is indicative of photocorrosion. 

Fig. 15. Diagram illustrating the thermodynamic stability of a semiconductor against corrosion and photocorrosion (a) semiconductor is absolutely stable, (b) stable against cathodic decomposition, (c) stable against anodic decomposition, and (d) unstable. [From Gerischer (1977a).]...

Consider now the processes caused by the formation of quasilevels. As was noted above, the shift of Fn relative to F is very small for majority carriers (electrons) and can usually be neglected precisely, this was done in constructing Fig. 16b. But for minority carriers (holes) the shift of Fp can be very large. The shifts of both Fnx F and Fp increase with the growing intensity of semiconductor illumination, so that for a certain illumination intensity Fp may reach the level of the electrochemical potential of anodic decomposition Fdec, p, and Fn—the level of a certain cathodic reaction (for example, reduction of water with hydrogen evolution FHljH20). These reactions start to proceed simultaneously, and their joint action constitutes the process of photocorrosion. 

Fig. 17. Prevention of photocorrosion of a semiconductor by a redox couple in the solution I—the semiconductor is stable, and II—the semiconductor is unstable against anodic photodecomposition.

Cadmium sulfide, CdS (Fig. 19), is, on the contrary, liable to intensive anodic photocorrosion in aqueous solutions, according to the equation... 

An interesting example of the kinetic effect in semiconductor photocorrosion is photopassivation and photoactivation of silicon (Izidinov et al., 1962). Silicon is an electronegative element, so it should be dissolved spontaneously and intensively in water with hydrogen evolution. But in most of aqueous solutions the surface of silicon is covered with a nonporous passivating oxide film, which protects it from corrosion. The anodic polarization curve of silicon (dashed line in Fig. 20) is of the form characteristic of electrodes liable to passivation as the potential increases, the anodic current first grows (the... 

Anodic oxidation of pyrrole and N-substituted pyrroles results in the formation of polypyrroles in an oxidized state, which can be useful for the preparation of conducting organic polymers.185-188 Oxidation of 2,5-di-substituted pyrroles produces soluble products and no layer of polymers.187 One of the proposed applications of such a layer of conducting polymer is the protection of semiconductor electrodes from photocorrosion.189-191... 

Finally, we note that the photocorrosion process is strongly pH-dependent, occurring most readily in strongly acid solutions, and that the presence of a carboxylic acid is required for the occurrence of severe photocorrosion. In Table II we present analytical results, based on inductively coupled argon plasma (ICP) emission spectroscopy, for representative electrolyte solutions after 6-8 hr. of photo-Kolbe electrolysis with n-SrTiC anodes. It can be seen that the formation of soluble strontium and titanium species is... 

The mechanism of photocorrosion seems to be the same for zinc and cadmium sulfide. For the latter it was investigated in detail for colloids [26] and crystals as discussed in the following [56]. In the absence of air, anodic photocorrosion (Eq. 15) ... 

The anodic process consists of reactions (22)-(25), which are based on detailed comparative photocorrosion studies at single-crystal electrodes and at powder suspensions ... 

There is also an etched layer of Si on the surface of the anodized Si sample under illumination as illustrated in Fig. 8.45. This etched layer, which unlike that in the dark is required for the surface roughening for the initiation of pores, is mainly due to photocorrosion. As a result of the photoinduced dissolution the top surface of the PS layer recedes with time. The rate of dissolution depends on doping, HF concentration, current density, and illumination intensity. Figure 8.46 shows the variation of the three layers with the amount of charges passed the etched layer on a highly doped sample is thicker than that on a lowly doped material. The thickness of the micro PS decreases while pore diameter and etched layer increase with increasing light intensity. Table 8.5 shows... 

Photovoltage spectra are measured at open circuit using chopped light of low intensity. It might appear that an advantage of photovoltage spectroscopy over photocurrent spectroscopy is that no photocorrosion occurs. However, this is not necessarily correct, because anodic photocorrosion in the illuminated areas may be balanced by cathodic reduction of solution species such as oxygen or protons. 

---