TiC>2 surface

In principle, the 3.2 eV (309 kJ mol-1) electron donor/electron acceptor pairs in Ti02 should have more than enough energy to decompose water into hydrogen and oxygen (1.23 V), but the evolution of both O2 and H2 on TiC>2 surfaces is hindered by very high overpotentials. The phenomenon of overpotential is considered at length in Section 15.4, but for present pur-... 

The chemisorption of ions plays another role in determining the properties of PEC devices. The adsorbed ions may create the chemical intermediates or specific reaction sites necessary for charge transfer and chemical product formation. The rapid kinetics in photoelectrolysis and wet photovoltaic cells are in no small part due to the fact that in most of these systems the redox species are strongly adsorbed. Knotek(10,ll) and others (12,13) have shown that the nature of the TiC>2 surface and the species adsorbed on it greatly affect its catalytic properties. 

Large adsorbates, such as bi-isonicotinic acid, may bind to a surface at several sites which are sufficiently far apart not to interact strongly in a direct way. This kind of system is by necessity large and complex, and few detailed studies have been reported on such systems. Various structural aspects of bi-isonicotinic acid adsorption on rutile and anatase TiC>2 surfaces have been presented in several recent studies [68, 77, 78]. Bi-isonicotinic acid adsorption on TiC>2 surfaces is not only taken as a problem of direct interest to the photoelectrochemical applications, but also serves as a model system for surface science investigations of phenomena connected to the adsorption of large organic adsorbates on metal oxide surfaces. 

With the semiconductor oxidation catalyst, however, the surface becomes activated only upon photoexcitation. At low light intensities, the possibility that many holes are formed in the valence band is remote, so that the irradiated semiconductor powder becomes an effective one-electron oxidant. Now if the same chemistry ensues on the photochemically activated TiC>2 surface, then the reaction will proceed as in the bottom route of eqn 9. Thus, the carboxy radical is formed, producing an alkyl radical after loss of carbon dioxide. Since the semiconductor cannot continue the oxidation after the first step, the radical persists, eventually recapturing the conduction band electron, either directly or through the intervention of an intermediate relay, perhaps superoxide. The resulting anion would be rapidly protonated to product. 

The other important aspect in dye-sensitized solar cells is water-induced desorption of the sensitizer from the TiC>2 surface. Extensive efforts have been made in our laboratory to overcome this problem by introducing hydrophobic properties in the ligands 13-17. The absorption spectra of these complexes show broad features in the visible region and display maxima around 530 nm. The performance of these hydrophobic complexes as charge transfer photosensitizers in nanocrystalline TiC>2-based solar cells shows excellent stability towards water-induced desorption [36]. 

In order to obtain high overall light to electric power conversion efficiencies, optimization of the short circuit photo current (z sc) and open circuit potential (Voc) of the solar cell is essential. The conduction band of the TiC>2 is known to have a Nernstian dependence on pH [55,67]. The fully protonated sensitizer 2, upon adsorption transfers most of its protons to the TiC>2 surface, charging it positively. The electric field associated with the surface dipole generated in this fashion enhances the adsorption of the anionic Ru complex and assists electron injection from the excited state of the sensitizer into the titania conduction band, favoring high photocurrents (18-19 mA cm-2). However, the open-circuit potential (0.65 V) is lower due to the positive shift of the conduction band edge induced by the surface protonation. 

Making the reasonable assumption that the regeneration of the oxidized redox couple at the counter electrode is faster than its recombination with the TiC>2 surface (7c6[e-] electron collection in the external circuit, r, the last term in the equation given above, can be simplified as follows ... 

Figure 6.15 Photocurrent-voltage characteristics for a nanocrystalline TiC>2 surface modified with [Ru(tcterpy)(NCS)3]. Reprinted with permission from A. Hagfeldt and M. Griitzel, Acc. Chem. Res., 33, 269 (2000). Copyright (2000) American Chemical Society...

Photoinduced electron injection is by no means a new development. This process has already been applied in areas such as silver halide photography. In this discussion, only sensitized TiC>2 surfaces will be considered. Many experiments have shown that the charge injection into the semiconductor surface is very fast. In order to study these processes, fast spectroscopic techniques are preferred. Whether or not charge injection takes place can be studied conveniently on the nanosecond time-scale by using transient absorption spectroscopy. However, to address the injection process directly, experiments are carried out on the femtosecond time-scale, while recombination and charge separation require the nanosecond to microsecond range. 

In typical investigations on the behavior of modified TiC>2 surfaces upon irradiation, measurements are carried out in acetonitrile containing 0.1 M UCIO4. Under these conditions, very fast electron injection into the semiconductor surface is observed. However, it has been noted that this injection process depends strongly on the lithium concentration of the contacting acetonitrile solution [10,11]. In the absence of lithium, no injection is observed. This is an important observation since this opens up the possibility of modulating the photophysical behavior of the interfacial supramolecular assembly by external manipulation of the conditions. In this section, this observation is discussed in more detail, and, in addition, the possibility to use the surface potential of the semiconductor surface as a driving force will be considered. 

This method of analysis will now be applied to investigate charge injection under various conditions. First of all, the observation made by Meyer and co-workers [10] that in pure acetonitrile no injection is taking place will be considered. They reported that upon excitation of a TiC>2 surface modified with [Ru(deebpy)(bpy)2] (for the structure, see Figure 6.7 above) in pure acetonitrile, a transient absorption... 

The investigation of modified TiC>2 surfaces is one of the best studied subjects in the area of surface-bound supramolecular assemblies. There are, however, quite a number of problems which still need to be addressed. The understanding of the recombination reactions is at present very limited. Time-resolved measurements have shown that the recombination reaction is a multi-exponential process. The reason for this is not yet understood, but is most likely related to the nanocrystalline nature of the semiconductor substrate. In order to gain more information about this issue, systematic studies on the effects of surface roughness and particle size are needed. 

Modified TiC>2 surfaces have also found application in the design of fast elec-trochromic devices. The influence of the substrate on the behavior of interfacial assemblies is well illustrated in this book. However, it is important to realize that the electrochromic behavior observed for modified TiC>2 surfaces was not expected. The oxidation and reduction of attached electrochromic dyes are not mediated by the semiconductor itself but by an electron-hopping process, not unlike that observed for redox polymers, where the electrochemical reaction is controlled by the underlying indium-tin oxide (ITO) contact. These developments show that devices based on interfacial assemblies are a realistic target and that further work in this area is worthwhile. 

There are also some very important scientific conclusions to be drawn from this work. The attachment of photoactive molecular components to TiC>2 surfaces has shown that the photophysical properties of the resulting interfacial ensemble are significantly different from those observed for the individual components. The semiconductor surface clearly acts as a rectifying one. In order to fully understand the balance between the intramolecular and the interfacial processes, more experiments are needed where the nature and connectivity of the molecular components, as well as the nature of the contacting solution, are systematically varied. 

The detailed discussion of the roles the continuous donor clusters play and the mechanisms of their functioning is beyond the scope of this paper therefore we shall restrict our consideration to the brief inspection of their influence on the formation and properties of the deposited metal nanoparticles. It is clear that during the contact deposition of metal (in the dark open circuit conditions) the metal particles are formed on the TiC>2 surface at the sites of donor location and, to the most extent, at the continuous donor clusters. 

Photons enter through the photoactive electrode and can be absorbed by sensitizer molecules (S) at various depths in the film. The sensitizer molecules (S ) excited in this way inject electrons into the conduction band of the adjacent TiC>2 particles (eCB), leaving an oxidised sensitizer molecule (S+) on the TiC>2 surface ... 

There was already a considerable body of knowledge on catalysts of this type [29]. For those used for selective oxidations, there was much evidence to show that the active phase was a monolayer of oxovanadium species chemically bonded to the TiC>2 surface such a material would have about 1 wt% V2O5 for a TiC>2 area of 10m2g l, but technical catalysts usually contained substantially larger amounts. The excess appeared to be in the form of V2O5 microcrystals which neither helped nor hindered in selective oxidation it seemed to serve as a reserve supply to replenish the monolayer, should it become depleted. There was also evidence that uncovered TiC>2 surface was harmful, in that it could cause deep oxidation to carbon oxides. In these applications, the anatase form of TiC>2 was generally used, and unless the contrary is stated the formula TiC>2 will imply this form. 

---