barrier oxide layer serves as the resistance to the flow of reactive ions that are needed to be transported through the anodic oxide layer to sustain the oxidation. Formation of anodic oxide layer ceases when the resistance is increased too high (accompanied by thick barrier oxide layer). However, this issue can be overcome by increasing the applied potentials for a higher electric field or by using a higher concentration of fluoride ions in the electrolyte to allow the creation of internal channels. These channels facilitate the uneven internal resistance to maintain the ongoing oxidation process. The thickness of the barrier oxide layers underneath the pores is kept in an equilibrated state between the two competitive reactions, i.e. the thickness is reduced by dissolution but simultaneously regenerated by oxidation (Figure 3.2d). The electric field distribution at the pores boundaries causes anisotropic widening and deepening of pores. As the pores formed at deeper region, the electric field closer to the underlayer metallic regions increases, thus promoting the anodic oxide growth. Afterward, the formation of anisotropic nanostructures of oxide continues to grow in equilibrium. The growth of such nanostructure will continue until the oxidation rate at the metal–oxide interface equals the chemical dissolution rate at the oxide/electrolyte interface.
Figure 3.2 Schematic diagram of the growth mechanism for anodized metal foil: (a) growth of thin and compact oxide layer, (b) the initial formation of pores, (c) formation of patterned pores within oxide layer, (d, e) continue growth of the pores to form nanostructures.
As the schematic formation of nanostructure during anodization is explained in Figure 3.2, the presence of those processes can be observed in the measurement of current transient pattern during anodization. Figure 3.3 shows a typical transient current profile measured during the reaction under constant applied anodization voltage. The transient current consists of three main stages. In the first stage, the current decays rapidly during the first few minutes of the anodization. At the start of the reaction, electrolyte is in direct contact with the conductive metallic surface, and thus the starting current level is usually high. Under the applied anodization, the oxidation of metallic foil occurs almost instantaneously to form a compact thin oxide surface layer. As the initial oxide layer increases in thickness, so does the electrical resistance of the substrate. As a result, the oxidation of metallic foil slows down, which is reflected by the rapid drop of current density. Sequentially, chemical dissolution of oxide layer by fluoride ion in the electrolyte starts to take place and render more underlayer metal to oxidation; a temporary bounce‐back in current is therefore observed in the second stage. The third stage is the equilibrium stage in which the formation and dissolution of oxide happen at the same competitive rates, reflected by the steady‐state current generation. As the growth of anodic nanostructure is regulated by the formation and dissolution of oxides, the readers can logically expect the important synthetic conditions to be related to the magnitude of applied voltage (strength of oxidation process) and the type or concentration of etching agent (e.g. fluoride ions, which directly influence the strength of dissolution process). Indeed, most of the morphological controls over the anodic oxide thin films are a result of optimal modulation of these two parameters (voltage and electrolyte composition).
Figure 3.3 Typical current profile under a constant applied anodization voltage of metal foil in organic electrolyte containing fluoride ions.
A number of industrially important metal oxide thin films have been prepared through anodization method such as alumina, silica, and titania. Many more academically interesting simple metal oxide films can be afforded by this method. Figure 3.4 shows scanning electron microscopy (SEM) images of metal oxides with anisotropic nanostructures obtained through anodization [20–22]. Some of the most important ones as relevant to catalytic applications include TiO2 and α‐Fe2O3 nanotubes, Nb2O5 nanorods, WO3 nanoflowers, and MoO3 truncated rhombohedra. These oxide materials are intrinsic semiconductors with or without nanostructured morphologies, but in the case of the former, their efficiencies as (photo)electrodes can be further augmented. The effect stems from (i) the increased ratio of surface area to volume for electrolyte contact and where reaction takes place, (ii) enhanced light absorption due to trapping of photons within pores, and (iii) the vectorial charge transport.
Figure 3.4 SEM images of the simple metal oxides obtained through anodization: (a) titanium dioxide (TiO2), (b) molybdenum trioxide (MoO3), and (c) tungsten trioxide (WO3).
Source: (a) Reprinted with permission from Yun et al. [20]. Copyright 2011, American Chemical Society. (c) Reprinted with permission from Ng et al. [21]. Copyright 2010, American Chemical Society.
The vectorial charge transport is an intriguing phenomenon that relies on the vertically oriented 1D array of the oxide semiconductor and hence deserves special mention here. Under photoexcitation, i.e. when the semiconductor photoanode is exposed to photons with energy equal to or greater than its band gap (see Chapters 11, 31, and 36 on the basics of photocatalysis), the generated photoelectrons would need to diffuse to the back of the electrode within its charge carrier lifetime, or they will recombine with the photoholes, hence the loss of photocharge for surface reaction. The photocharge transport can be described by the following equation:
where Lc is the diffusion length or distance traveled by the charge carrier (electrons or holes) before recombination, Dc is the diffusion coefficient of the charge carrier, and τc is the lifetime of the charge carrier. It should be noted that the Dc (and hence Lc) is different for both electron and hole even on the same semiconductor material. For a photoanode (or photocathode) that is composed of irregular‐shaped or randomly packed particles, the photoelectrons (or photoholes) undergo the “random walk motion” that are rarely the most straightforward path to the back of the electrode. With the creation of 1D array of the oxide semiconductor, the diffusion of the photoelectrons is restricted to the shortest vertical path to the back of the electrode. This enables a large fraction of photoelectrons (or photoholes in the case of photocathode) to be collected within their τc. At the same time, it is important to restrict the wall thickness/diameter to not more than twice the Lc of the photoholes (or photoelectrons) such that majority of them could diffuse to the semiconductor surface to catalyze the oxidation (or reduction) reaction. Owing to their advantages, aligned nanotube and nanorod arrays have been used for water photoelectrolysis
