Photocatalyst

Definition

Photocatalysts are materials that induce photocatalytic reaction under photoirradiation. A general definition of photocatalysis, a conceptual name of photocatalytic reactions, is a chemical reaction induced by photoabsorption of a solid material, or “photocatalyst,” which remains unchanged during the reaction. Therefore, “photocatalyst” should act catalytically, i.e., without change, under light. Although molecules or metal complexes dissolved in solution or in the gas phase, not solid materials, can drive such photoinduced reactions without change during the reactions, they are called “photosensitizer” but not “photocatalyst.” photoinduced reactions are, in principle, initiated by photoabsorption, i.e., excitation of electrons in a material, followed by electron transfer from/to a reaction substrate. The photoexcited electrons can reduce a substrate adsorbed on the surface of a photocatalyst, and positive holes, electron deficiencies, can oxidize an adsorbed substrate, if the electronic levels of photoexcited electrons and positive holes are higher and lower than the redox levels of substrates, respectively. This is one of the necessary conditions for photocatalysis and will be discussed later.

For photocatalysts, there are several modes of photoexcitation as shown below.

Band-Gap Excitation

For conventional photocatalysts such as titanium(IV) oxide (TiO₂; titania), band-gap excitation occurs when irradiated. In a schematic representation of the electronic structures of semiconducting (or insulating) materials, a band model, an electron in an electron-filled valence band (VB) is excited by photoabsorption to a vacant conduction band (CB), which is separated by a band gap (a forbidden band), from the VB, leaving a positive hole in the VB. An important point is that photoabsorption and electron-positive hole generation are inextricably linked; a VB electron is not excited after photoabsorption. These electrons and positive holes can induce reduction and oxidation, respectively, of chemical species adsorbed on the surface of a photocatalyst, unless they recombine with each other so as not to induce a redox reaction but to produce heat and/or photoemission. Such a mechanism accounts for the photocatalytic reactions of semiconducting (or insulating) materials absorbing photons by the bulk of materials.

Since light of energy greater than band gap can excite electrons in VB to CB, light of wavelength shorter than that corresponds to the band gap, i.e., longer-wavelength limit for photoabsorption (photoexcitation) is fixed by the band structure of a photocatalyst. For metal oxides, it has been known that VB is mainly composed of O_2p orbitals and thereby the position (top) of VB is independent of the kind of metal, i.e., only the CB (bottom) position is changed depending on the kind of metal. Therefore, in order to narrow the band gap, i.e., to use light of longer wavelength, metal oxides having lower CB level should be used. Considering the VB top level seems enough low to oxidize most of organic/inorganic compounds, the band-gap narrowing for metal oxides leads to disadvantageous lowering of the CB bottom level.

Transition Between an Electronic Level and a Band

Photocatalysts that can use visible light included in sunlight and indoor light have been looked for since the conventional photocatalysts such as titania can absorb, i.e., be excited only by ultraviolet light. The strategy to utilize visible light in photocatalysis can be roughly divided into three categories. The first one is to use the semiconductor with a narrow band gap. Metal nitride or sulfides often have colors and are expected to work as photocatalysts under visible-light irradiation when used. However, they tend to be oxidized and possibly lose their photocatalytic activity. Then, it was proposed to raise the top of VB of stable metal oxides to reduce the band-gap energy by doping various elements as the second strategy. Many papers have reported the band-gap narrowing by doping nitrogen, carbon or sulfur, etc., to titania and the introduction of visible-light responsibility. Actually, shift of the photoabsorption spectrum to the longer-wavelength side is observed by doing a variety of elements. It has been suggested in the recent studies, however, that the electronic structure of such doped material was not an expected one; levels of doped elements are separated from the VB to form an independent sub-bands.

In recent studies, clusters of ions such as copper or iron and their oxide were deposited (grafted) to induce photoexcitation of electrons in VB of base metal oxide such as titania to the electronic level of these loaded clusters or electrons in the electronic level to CB, i.e., interfacial charge transfer. When the electronic states of the grafted clusters are located between the CB bottom and the VB top of titania, the interfacial charge transfer can be driven under visible-light irradiation, and a new photoabsorption band appears in a wavelength region longer than that for band-gap excitation.

Excitation Through Surface-Plasmon Resonance Absorption

Another example of visible light-driven reaction through non-band-gap excitation is a photocatalytic reaction that uses surface-plasmon resonance (SPR) absorption of small metal particles loaded on base metal oxides. For example, gold particles of the size of ten to several ten nanometers, presenting purplish red color by the SPR absorption, loaded on titania particles have been used for photocatalytic reactions under visible-light irradiation at the wavelength of ca. 600 nm. Based on the results that titania or a related material is necessary for this visible light-driven reaction and that SPR absorption cannot induce electronic excitation of electrons, the mechanism of this kind of reaction seems complicated and is now under discussion.

Thermodynamics

As thermodynamics says, if ΔG is negative (ΔG < 0 as in the case, e.g., in oxidative decomposition of organic compounds under aerated conditions) and if ΔG is positive (DG > 0, as in the case, e.g., in splitting of water into hydrogen and oxygen), energy is released and stored, respectively. Therefore, if the standard electrode potential of the compound to be reduced by electrons is higher, i.e., more negative (cathodic), than that of the compound to be oxidized by positive holes, ΔG is positive, i.e., the reaction stores energy and vice versa. A notable point is that both situations, energy release and storage, are possible in photocatalysis, while thermal catalyses are limited to only reactions of negative ΔG, i.e., spontaneous reactions. The reason why photocatalysts can drive even a reaction of positive ΔG, which does not proceed spontaneously, is that an overall redox reaction can proceed, even if its ΔG is positive, in a system in which reduction and oxidation steps are spatially or chemically separated; otherwise, the reaction between reduction and oxidation products proceeds to give no net products. Under these conditions, both partial Gibbs energy change for reactions of photoexcited electrons with oxidant (ΔG _e) and positive holes with reductant (ΔG _h) are required to be negative, i.e., reactions by photoexcited electrons and positive holes proceed spontaneously by photoexcitation. In other words, for the reaction through band-gap excitation, the CB bottom and VB top positions must be higher (more cathodic) and lower (more anodic) than standard electrode potentials of an electron acceptor (oxidant) and an electron donor (reductant), respectively, to make Gibbs energy change of both half reactions negative, as has often been pointed out as a necessary condition for photocatalysis.

Photocatalytic Activity

The widely used scientific term “activity” often appears in papers on photocatalysis as “photocatalytic activity.” Although the author does not know who first started using this term in the field of photocatalysis, researchers in the field of conventional catalysis were using this term even before the 1980s, when photocatalysis studies had begun to be promoted by the famous work of the so-called Honda–Fujishima effect on photoelectrochemical decomposition of water into oxygen and hydrogen using a single-crystal titania electrode, as mentioned above. Most authors, including the present author, prefer to use the term “photocatalytic activity,” but in almost all cases, the meaning seems to be the same as that of absolute or relative reaction rate. A possible reason why the term “photocatalytic activity” is preferably used is that the term may make readers think of “photocatalytic reaction rate” as a property or ability of a photocatalyst, i.e., photocatalysts have their own activity. On the other hand, “reaction rate” seems to be controlled by given reaction conditions including a photocatalyst. In the field of conventional catalysis, “catalytic activity” has been used to show a property or performance of a catalyst, since an “active site,” substantial or virtual, on a catalyst accounts for the catalytic reaction. The estimated reaction rate per active site can be called “catalytic activity.” In a similar sense, the term “turnover frequency,” i.e., number of turnovers per unit time of reaction, is sometimes used to show how many times one active site produces a reaction product(s) within unit time. On the contrary, it is clear that there are no active sites, as in the meaning used for conventional catalysis, in which rate of catalytic reaction is predominantly governed by the number or density of active sites, on a photocatalyst. The term “active site” is sometimes used for a photocatalytic reaction system with dispersed chemical species, e.g., metal complexes and atomically adsorbed species, on support materials. However, even in these cases, a photocatalytic reaction occurs only when the species absorb light, and therefore, species not irradiated cannot be active sites. A possible mechanism of photo induced reaction is that photoirradiation induces production of stable “active sites” that work as reaction centers of conventional catalytic reactions, though this is different from the common mechanism of photocatalysis by electron–positive hole pairs. Anyway, photocatalytic reaction rate strongly depends on various factors such as the irradiance of irradiated light that initiates a photocatalytic reaction. Considering that the dark (nonirradiated) side of a photocatalyst or suspension does not work for the photocatalytic reaction, the use of the term “active site” seems inappropriate.

Design of Active Photocatalysts

Since an ordinary photocatalysis is induced by photoexcited electrons and positive holes, rate of photocatalytic reaction must depend on photoirradiation irradiance (light flux) and efficiencies of both photoabsorption and electron–positive hole utilization. The efficiency of electron–positive hole utilization is called quantum efficiency, i.e., the number (or rate) ratio of product(s) and absorbed photons, and even if quantum efficiency is high, the overall rate should be negligible when the photocatalyst does not absorb incident light. This is schematically represented as

$$ \left[ {{{\mathrm{ Rate}} \left/ {{\mathrm{ mol}\ {{\mathrm{ s}}^{-1 }}}} \right.}} \right] = \left[ {{{\mathrm{ Irradiance}} \left/ {{\mathrm{ mol}\ {{\mathrm{ s}}^{-1 }}}} \right.}} \right]\,{} \times \left[ {\mathrm{ Photoabsorption}\ \mathrm{ efficiency}} \right]\,{} \times \left[ {\mathrm{ Quantum}\ \mathrm{ efficiency}} \right]. $$

Since all of the parameters in this equation must be functions of light wavelength, the overall rate can be estimated by integration of a product of spectra of photoirradiation, photoabsorption, and quantum efficiencies. When we discuss activity of a photocatalyst, it seems reasonable to evaluate a product of photoabsorption and quantum efficiencies, i.e., apparent quantum efficiency. Assuming that quantum efficiency does not depend on the irradiation (absorption) irradiance, the actual reaction rate can be estimated by multiplying with the irradiance. On the basis of these considerations, enhancement of photocatalytic activity can be achieved by increase in both efficiencies. For example, preparing visible-light absorbing photocatalysts, as a recent trend in the field of photocatalysis, and depositing noble metal particles onto the surface of photocatalysts lead to the improvement of these efficiencies, respectively. In this sense, the design of active photocatalysts seems simple and feasible, but we encounter the problem that both efficiencies are related to each other, and we do not know how we can improve the quantum efficiency since correlations between physical/structural properties and photocatalytic activity have only partly been clarified.

Future Perspectives

Since most of researchers in the field of photocatalysis came from different fields of chemistry, catalysis chemistry, electrochemistry, materials chemistry, photochemistry, etc., there seemed no common concepts shared by them. It is necessary to understand photocatalysis appropriately considering thermodynamics and kinetics of photocatalysis introduced in this section.

Cross-References

TiO₂ Photocatalyst