Photoexcitation of semiconductor electrodes

Semiconductor electrodes exhibit electron photoemission into the solution, like metal electrodes, but in addition they exhibit further photoelectrochemical effects due to excitation of the electrode under illumination. The first observations in this area were made toward the middle of the twentieth century. At the end of the 1940s, 

Vladimir I. Veselovsky studied the photoelectrochemical behavior of metals covered with oxide layers having semiconductor properties. In 1955, Walter H. Brattain and Charles G. B. Garrett published a paper in which they established the connection between the photoelectrochemical properties of single-crystal semiconductors and their electronic structure. 

FIGURE 29.2 Energy bands in an -type semicondnctor in contact with an electrolyte solution for two values, j and E2, of electrode potential E2 E.  

When semicondnctors are irradiated with photons of high energy, electron photoemission is possible, as in the case of metals. When the photon energy is lower than the electron work function in the solution, under given conditions, but is still higher than the semiconductor s bandgap W,  

The difference between metals and semiconductors becomes apparent when the further fate of these excited charges is considered. In metals an excited electron will very quickly (within a time on the order of 10 s) return to its original level, and the photon s original energy is converted to thermal energy. Photoexcitation has no other consequences. 

---

Fig. 10-S. Photopotential and band bending in a space chaige layer of semiconductor electrodes (a) in the dark and (b) in a photoexcited state - Ae e - pbotopotential.

Photoelectrochemistry (PEC) is emerging from the research laboratories with the promise of significant practical applications. One application of PEC systems is the conversion and storage of solar energy. Chapter 4 reviews the main principles of the theory of PEC processes at semiconductor electrodes and discusses the most important experimental results of interactions at an illuminated semiconductor-electrolyte interface. In addition to the fundamentals of electrochemistry and photoexcitation of semiconductors, the phenomena of photocorrosion and photoetching are discussed. Other PEC phenomena treated are photoelectron emission, electrogenerated luminescence, and electroreflection. Relationships among the various PEC effects are established. 

FIGURE 29.3 Charge separation during photoexcitation of a semiconductor electrode. 

The boundary between effects thus defined is, however, not sharp. If, for instance, light is absorbed by a layer of a photoactive adsorbate attached to the semiconductor electrode, it is apparently difficult to identify the light-absorbing medium as a solution or electrode material . Photoexcited solution molecules may sometimes also react at the photoexcited semiconductor electrode this process is labelled photogalvanovoltaic effect. 

Under light illumination, semiconductor electrodes absorb the energy of photons to produce excited electrons and holes in the conduction and valence bands. Compared with photoelectrons in metals, photoexcited electrons and holes in semiconductors are relatively stable so that the photo-effect on electrode reactions manifests itself more distinctly with semiconductor electrodes than with metal electrodes. 

Fig. 10-9. Photoexcited reaction current (photocurrent) at semicon ductor electrodes (a) photoexcited reaction of cathodic electron transfer (OX + e - RED) at p-type semiconductor electrode, (b) photoexcited reaction of anodic hole transfer (RED - OX + e) at n-type semiconductor electrode, iph = photocurrent.

Fig. 10-11. Anodic photoexcited dissolution current of an n-type semiconductor electrode of gallium arsenide as a function of electrode potential in a 0.6 M sulfuric add solution lo - photon intensity = diotocurrent. [From Memming-Kelly, 1981.]...

With n-type semiconductor electrodes, the anodic oiQ en reaction (Euiodic hole transfer) will not occur in the dark because the concentration of interfacial holes in the valence band is extremely small whereas, the same reaction will occur in the photon irradiation simply because the concentration of interfadal holes in the valence band is increased by photoexcitation and the quasi-Fermi level pEp of interfadal holes becomes lower than the Fermi level the o en redox... 

Similarly, the cathodic photoexcited electron transfer of the h3 rogen reaction shown in Eqn. 10-19 can occur at p-type semiconductor electrodes at which the cathodic hydrogen reaction is thermodynamically impossible in the dark. 

Figure 10-17 shows the polarization ciirves for the cathodic hydrogen reaction (cathodic electron transfer) on a p-type semiconductor electrode of galliiun phosphide. The onset potential of cathodic photoexcited hydrogen reaction shifts significantly from the equilibrium electrode potential of the same hydrogen reaction toward the flat band potential of the p-type electrode (See Fig. 10-15.). 

A shift of the flat band potential due to photoexcitation of the type shown in Fig. 10-18 results from the capture of holes in the surface state level, e , on the electrode as shown in Fig. 10-19. We now consider a dissolution reaction involving the anodic transfer of ions of a simple elemental semiconductor electrode according to Eqns. 10-24 and 10-25 ... 

This conclusion is valid regardless whether the electrode is n-fype or p-fype. Consequently, if the quasi-Fermi level of interfacial holes in a photoexcited n-type semiconductor electrode equals the quasi-Fermi level of interfacial holes pEp, (eq ial to the Fermi level pEp., of the interface) in a p-type electrode of the same semiconductor in the dark, the current due to anodic holes will be the same on the two electrodes and, hence, the curves of the anodic reaction current as a function of the quasi-Fermi level of interfacial holes will be the same for the two electrodes as suggested in Fig. 10-21. The curves of the anodic reaction current represented as a function of the electrode potential (the Fermi level of the electrode), instead of the quasi-Fermi level of interfacial holes, are not the same for the two electrodes, however. 

Fig. 10-21. Quasi-Fermi levels of holes in transfer reaction of anodic holes, from the valence band (a) of a photoexcited n-type electrode and (b) of a dark p-type electrode of the same semiconductor, to redox particles pCp, = quasi-Fermi level of interfacial holes in a photoexcited n-type electrode where pCp, is lower than the Fermi level cp and in a dark n-type electrode where pCp, equals the Fermi level sp.

Fig. 10-22. Overvoltages in an anodic hole transfer (a) at a photoexcited n-type electrode and (b) at a p-type electrode of the same semiconductor iih = overvoltage for hole transfer across an interface = inverse overvoltage due to generation and transport of photoexcited holes in an n>type electrode.

Since the overvoltage iip,sc for the generation and transport of holes is a negative quantity, the total overvoltage becomes negative when the magnitude of Ti p, sc exceeds t) h the condition usually occims with photoexcited n-type electrodes. This provides the basis for the fact that the potential for the onset of anodic hole transfer at photoexcited n-type electrodes is more cathodic (n ative) than the potential for the onset of anodic hole transfer at p-type electrodes of the same semiconductor or at metal electrodes. 

In photoexcited n-type semiconductor electrodes, photoexcited electron-hole pairs recombine in the electrodes in addition to the transfer of holes or electrons across the electrode interface. The recombination of photoexcited holes with electrons in the space charge layer requires a cathodic electron flow from the electrode interior towards the electrode interface. The current associated with the recombination of cathodic holes, im, in n-type electrodes, at which the interfadal reaction is in equilibrium, has already been given by Eqn. 8-70. Assuming that Eqn. 8-70 applies not only to equilibrium but also to non-equilibrium transfer reactions involving interfadal holes, we obtain Eqn. 10-43 ... 

The current i flowing in photoexcited n-type semiconductor electrodes equals the sum of the photoexcited hole current i >h, the limiting current of hole diffusion ip. itB, and the current of hole recombination inc as shown in Eqn. 10—44 ... 

Figure 10-23 shows the electron levels and the polarization curves for the transfer of anodic redox holes both at a photoexcited n-fype electrode and at a dark p-type electrode of the same semiconductor. The range of potential where the anodic hole current occurs at the photoexcited n-type electrode is more cathodic (more negative) than the range of potential for the anodic hole current at the dark p-type electrode. The difference between the polarizatitm potential aE(i) (point N in the figure) of the photoexcited n-type electrode and the polarization potential pE(i) (point P in the figme) of the dark p-type electrode at a constant anodic current i is equivalent to the difference between the quasi-Fermi level pej of interfacial holes and the Fermi level bEf of interior holes (electrons) in the photoexcited n-type electrode this difference of polarization potential, in turn, equals the inverse overvoltage rip.sc(i) defined in Eqn. 10-46 ... 

Consequently, by measuring the polarization curves for the transfer reaction of anodic holes both at a photoexdted n-type electrode and at a dark p-type electrode of the same semiconductor, we obtain the relationship between the Fermi level of the electrode (polarization potential E) and the quasi-Fermi level of interfadal holes in the photoexcited n-type dectrode as a function of... 

Chapter 10 deals with photoelectrode reactions at semiconductor electrodes in which the concentration of minority carriers is increased by photoexcitation, thereby enabling the transfer of electrons to occur that can not proceed in the dark. The concept of quasi-Fermi level is introduced to account for photoenergy gain in semiconductor electrodes. Chapter 11 discusses the coupled electrode. mixed electrode) at which anodic and cathodic reactions occur at the same rate on a single electrode this concept is illustrated by corroding metal electrodes in aqueous solutions. 

Wilson RH (1977) A model for the current-voltage curve of photoexcited semiconductor electrodes. J Appl Phys 48 4292-4297... 

Photodiodes utilize principally the photophysical process of semiconductors. The most typical juctions to attain photoinduced charge separation are shown in Fig. 27 a c. If a photoexcited compound (P) is arranged with donor and/or acceptor on an electrode as shown in Fig. 25 (d), it must work as a kind of photodiode based on new principle of photochemical reaction. A polymer film must be most promising to construct such photoconversion element. 

In the case of a semiconductor electrode, the existence of the energy gap makes a qualitatively different location of energy levels quite probable (Figs. 23b, 23c). One of them, either the ground or excited, is just in front of the energy gap, so that the direct electron transition with this level involved appears to be impossible. This gives rise to an irreversible photoelectro-chemical reaction and, as a consequence, to photocurrent iph. The photoexcited particle injects an electron into the semiconductor conduction band... 

Fig. 37. Occurrence of electrogenerated luminescence with the participation of photoexcited reactants in the solution (a) semiconductor electrode, and (b) metal electrode.

At present there is a sufficiently complete picture of photoelectrochemical behavior of the most important semiconductor materials. This is not, however, the only merit of photoelectrochemistry of semiconductors. First, photoelectrochemistry of semiconductors has stimulated the study of photoprocesses on materials, which are not conventional for electrochemistry, namely on insulators (Mehl and Hale, 1967 Gerischer and Willig, 1976). The basic concepts and mathematical formalism of electrochemistry and photoelectrochemistry of semiconductors have successfully been used in this study. Second, photoelectrochemistry of semiconductors has provided possibilities, unique in certain cases, of studying thermodynamic and kinetic characteristics of photoexcited particles in the solution and electrode, and also processes of electron transfer with these particles involved. (Note that the processes of quenching of photoexcited reactants often prevent from the performing of such investigations on metal electrodes.) The study of photo-electrochemical processes under the excitation of the electron-hole ensemble of a semiconductor permits the direct experimental verification of the applicability of the Fermi quasilevel concept to the description of electron transitions at an interface. 

Fig. 1. Electron-Hole Generation upon Photoexcitation of a Semiconductor Electrode...

Figure 6.25 Outline of a cell for the photoelectrolysis of water. The photoexcited semiconductor electrode is made of SrTi02...

The Role of Interface States in Electron-Transfer Processes at Photoexcited Semiconductor Electrodes... 

The process by which the semiconductor carriers reach the surface to react with surface states must be considered. The case of greatest importance under photoexcitation is with the semiconductor biased to depletion as shown in Figure 1. While it is possible for semiconductor carriers to reach the surface of the semiconductor through tunneling, or impurity conduction processes, these processes have not been shown to be important in most examples of photoexcited semiconductor electrodes. Consequently, these processes will be ignored here in favor of the normal transport of carriers in the semiconductor bands. Furthermore, only carriers within a few kT of the band edges will be considered, i.e., "hot" carriers will be ignored. 

---