Semiconductor anode

The photo-Kolbe reaction is the decarboxylation of carboxylic acids at tow voltage under irradiation at semiconductor anodes (TiO ), that are partially doped with metals, e.g. platinum [343, 344]. On semiconductor powders the dominant product is a hydrocarbon by substitution of the carboxylate group for hydrogen (Eq. 41), whereas on an n-TiOj single crystal in the oxidation of acetic acid the formation of ethane besides methane could be observed [345, 346]. Dependent on the kind of semiconductor, the adsorbed metal, and the pH of the solution the extent of alkyl coupling versus reduction to the hydrocarbon can be controlled to some extent [346]. The intermediacy of alkyl radicals has been demonstrated by ESR-spectroscopy [347], that of the alkyl anion by deuterium incorporation [344]. With vicinal diacids the mono- or bisdecarboxylation can be controlled by the light flux [348]. Adipic acid yielded butane [349] with levulinic acid the products of decarboxylation, methyl ethyl-... 

Fig. 5.63 Scheme of a photoelectrochemical cell with sensitized semiconductor anode... 

Fig. 8-24. Redox reaction currents via the conduction and the valence bands of semiconductor electrode as functions of electrode potential of semiconductor anodic polarization corresponds to Figs. 8-20, 8-21 and 8-22. i (i )= anodic (cathodic) current in (ip) = reaction crnrent via the conduction (valence) band BLP = band edge level pinning FLP = Fermi level pinning.

An example of the effect of photon irradiation on the flat band potential is shown in Fig. 10-18 this figure compares a Mott-Schott plot with the anodic polarization curve of the dissolution reaction of a semiconductor anode of n-type molybdeniun selenide in an acidic solution in the dark and in the photoexcited conditions. In this example photoe dtation shifts the flat band potential from Em in the dark to pii) in the photoexcited state is about 0.75 V more positive than Em. This photo-shift of the flat band potential, Emi )-Em, corresponds to the change in the potential, of the compact layer due to photoexcitation as defined in Eqn. 10-23 ... 

Figure 10-25 shows an energy diagram of a photoelectrol3ddc cell for decomposing water this cell is composed of a metallic cathode and an n-type semiconductor anode, on which the following anodic and cathodic reactions, Eqns. 10-53(a) and 10-53(b), proceed ... 

In order for the photoelectrolytic decomposition of liquid water to proceed, the Fermi levels of the redox reactions in Eqns. 10-53a and 10- 3b need to be located within the band gap of the n-type semiconductor anode. In Fig. 10-26(a), we have assumed that the Fermi level ep(ac) of the n-type semiconductor anode at the flat band potential is higher than the Fermi level ep(h-/h2) of hydrogen redox reaction we have also assumed that the Fermi level e,(M) of the metallic cathode is lower than ekh /Hj)- Further, we have assiuned that the edge level of the conduction band is higher than the Fermi level of hydrogen redox... 

Fig. 10-26. Energy diagrams of a cell for photoelectrolytic decomposition of water consisting of a metal cathode (M) and an n-type semiconductor anode (n-SC) of which the Fermi level is higher than the Fermi level of hydrogen redox reaction ( R8o>ep(H /H2)) (a) cell circuit is open in the dark, (b) cell circuit is closed in the daric, (c) cell circuit is closed in a photoezdted state (cell reaction proceeds.). potential hairier of a space charge layer.

When the cell circuit is closed in the dark, as shown in Fig. 10-25(b), the Fermi level is equilibrated between the metallic cathode and the n-lype semiconductor anode. As a result, a depletion layer of space charge (potential barrier, is formed in the semiconductor anode, thereby shifting the potential of the anode from the flat band potential to a more anodic (more positive) potential (= + ). In the dark, however, the anodic hole transfer... 

When the n-type semiconductor anode is photoexcited, as shown in Fig. 10-25(c), the Fermi level of the anode is raised (the potential of the anode is lowered) by an energy equivalent to the photopotential at the same time, the Fermi... 

Since the highest possible Fermi level of the photoexcited n-type anode corresponds to the flat band potential of the semiconductor anode, the Fermi level of the metallic cathode short-circuited with the photoexcited n-lype anode can also be raised up to the level equivalent to the flat band potential of the semiconductor anode. In order for the cathodic electron transfer of hydrogen redox reaction to proceed at the metallic cathode, the Fermi level 1 of the cathode needs to be higher than the Fermi level of hydrogen redox reaction. Consequently, in... 

The cathodic and anodic polarization potentials, Ec and E, in the stationary state of the cell for photoelectrolytic decomposition of water, in which the metallic cathode and the n-type semiconductor anode are short-circuited, are given, respectively, in Eqns. 10-55 and 10-56 ... 

La photovoltaic cells, the same redox reaction, OX + e = KED, may be used for both the anode and the cathode. Figure 10-33 shows an eneigy diagram of an operating photovoltaic cell this cell consists of a metallic cathode and a photoexcited n-type semiconductor anode. The electromotive force (the open cell voltage), ph > approximately equals the difference between the flat band potential of... 

Fig. 10-33. Energy diagram for a photovoltaic cell composed of a metal cathode and an n>type semiconductor anode Vpi, = cell voltage in operation at current <.

Fig. 10-34. Energy diagram for a photovoltaic ceU composed of an n-type semiconductor anode and a p-type semiconductor cathode.

Figure 6, Schematic showing energy correlations for photoassisted electrolysis of water using n-type TiOg as a photoanode and a metal cathode. Symbols as in Figures 3, 4, and 5, except Epis Fermi level for metal contact to TiO and E/ is higher Fermi level in metal cathode, polarized by an external source to a potential negative to the semiconductor anode. EF(Hi) and Ep(02) are abbreviated forms for Fermi energies for redox systems of Figure 3 (13j.

Figure 8. Schematic showing energy correlations at equilibrium for cell with two semiconductor electrodes in contact with aqueous solution and through an external circuit with each other. An n-type semiconductor anode and a p-type cathode are shown to left and right, respectively. In each case the minimum light energy to give rise to a photocurrent is indicated by hvmin (n) and hvmin (p), respectively. The energies available for oxidation and reduction are also indicated. and Ev(n) are conduction and valence band edges for the n-type material and Ec(p) and E fp) are those for the p-type material. Other symbols as in Figure 7.

A common photoelectrolysis cell structure is that of a semiconductor photoanode and metal cathode, the band diagrams of which are illustrated in Fig. 3.15 together with that of electrolyte redox couples. In Fig. 3.15(a) there is no contact between the semiconductor anode and metal cathode (no equilibrium effects communicated through the electrolyte). As seen in Fig. 3.15(b), contact between the two electrodes (no illumination) results in... 

The situation described so far with semiconductor electrode kinetics is the simplest case The semiconductor has no states for electrons or holes at the surface. More frequently met—particularly for semiconductors that evolve H2 or 02, are semiconductors with surface states. In such a case, the potential-distance relation inside the semiconductor becomes flatter, and the behavior of the semiconductor becomes more like that of a metal. Thus (see Fig. 7.27), for the high surface-state case for a /5-type semiconductor anode, there reappears a substantial p.d. in the solution the p.d. inside the semiconductor is reduced toward a small value. 

Fornarini, L., Stirpe, F., Scrosati, 8. and Razzini, G., Electrochemical Solar Cells with Layer-Type Semiconductor Anodes. Performance of n-MoSs Cells, Solar Energy Materials, 5, 107, (1981). 

L. Fomarini, F. Stirpe, and B. Scrosati, Electrochemical solar cells with layer-type semiconductor anodes. Nonaqueous electrolyte cells, J. Electrochem. Soc. 129, 1155, 1982. 

A variety of other systems have been suggested and some were also tested such as cells containing an n-semiconductor anode and a p-type cathode, photoelectrolysis cells integrated with photovoltaic configurations, and tandem and cascade type of cells (Nozik and Memming, 1996). 

Theoretical eflBciencies for the process are high ( 25 %) but, in practice, an order of magnitude lower is observed. A major problem with the process is the evolution of oxygen at the semiconductor anode surface. To date, very few materials of the required band-gap (1.5-2 eV) have proved suitable [36, 43]. 

---