Photo-catalyzed surface hydrolysis of iridium(III) ions on semiconductors: a facile method for the preparation of semiconductor/IrO_x composite photoanodes toward oxygen evolution reaction

Abstract

We previously reported that the hydrolysis of Ir³⁺ in homogeneous solution could be triggered by irradiation with light whose energy was larger than a threshold value. In this work, we demonstrated that, by introducing Fe₂O₃ particles into solution, the incident light energy-restriction for the photo-catalyzed hydrolysis could be broken and the hydrolysis occurred at the Fe₂O₃/solution interface. The photo-generated holes on the Fe₂O₃ surface played a key role in oxidizing Ir(III) to Ir(IV) species and triggered the deposition of IrO_x. We showed that this photo-catalyzed surface hydrolysis is a universal phenomenon that takes place on the surface of many n-type semiconductors such as Fe₂O₃, TiO₂, and Ag₃PO₄. As IrO_x is an efficient catalyst for oxygen evolution reaction, surface hydrolysis is a general, facile and efficient strategy to prepare semiconductor/IrO_x composites, which can be used as anodic materials for photoelectrochemical water splitting.

---

1. Introduction

Hydrolysis of metal cations is a well-studied and documented reaction, which is not only fundamental to the chemistry of metal cations in aqueous solution, but also important for the wet chemical process for metal-based materials. Hydrolysis has long been employed as a successful strategy for the preparation of metal oxides.^1–5 For example, in many wet chemical approaches, such as forced hydrolysis synthesis,^1,4 hydrothermal growth,^5–7 and sol–gel method,^8,9 hydrolysis of metal cations is the key process for the formation of metal hydroxides and oxides. During a hydrolysis approach to prepare metal oxides, the metal inorganic salts and alkoxides usually undergo firstly a hydrolysis reaction and then a polycondensation, resulting in the formation of colloids or precipitates of metal oxides.² Many transition metal oxides, such as manganese,^10,11 copper,^6,12 titanium,^13,14 zinc,^15,16 nickel,¹⁷ cobalt^18,19 and iron^20,21 oxides, have been prepared by hydrolysis.

The rate of many hydrolysis reactions is very slow at room temperature under 1 atm,^3,22 making these reactions impractical for the synthesis of metal oxides at ambient temperature and pressure. Raising the temperature and pressure can provide the additional energy to trigger or accelerate these reactions.^22,23 Light, another form of energy that can drive chemical reactions at room temperature, should also be effective in inducing and accelerating hydrolysis reactions. However, few works have reported the photo-driven hydrolysis of metal salts even though irradiation has been successfully used in the reductive preparation of noble metal nanoparticles (NP).^24–26 Recently, we demonstrated that light could drive the hydrolysis of iridium(III) chloride,²⁷ whose rate was extremely slow at room temperature in the dark.²² We showed that UV-vis irradiation played a crucial role not only in the formation of the hydroxide complex [Ir(OH)₆]³⁻ but also in the polymerization and oxidation of Ir(III) species to form nanoparticles of iridium oxide (IrO_x). More importantly, there existed a critical wavelength of ca. 500 nm for the photo-catalyzed hydrolysis reaction, and the hydrolysis occurred only when the wavelength of incident light was much shorter than 500 nm.

Iridium oxides exhibit catalytic activity for many reactions.^28–30 For example, they have long been regarded as highly efficient catalysts for electrochemical oxygen evolution reaction (OER),^28,31,32 which makes them a good candidate as a co-catalyst of photoanodes for OER in solar water splitting.^28,31,33 This application requires iridium oxide to be deposited on the surface of semiconductors. Therefore, as a co-catalyst, it is desirable that iridium oxide is deposited on the photoactive sites of the semiconductor surface, where photoinduced holes can be easily captured by iridium oxide and then used to oxidize water. Photo-assisted deposition of co-catalysts on semiconductors is an ideal strategy to prepare semiconductor/co-catalyst systems.^34–38 This is because the photo-assisted deposition was usually caused by an oxidation or reduction process, which was induced by photogenerated holes or electrons. As a result, the deposition occurred at the sites where photoinduced holes or electrons were mostly available on the semiconductor surface.^34–36 Inspired by these works, we believe that the photo-catalyzed hydrolysis of Ir³⁺ may be an effective way to deposit IrO_x on the photo-active sites of the semiconductor surface, especially if we can confine the hydrolysis at the semiconductor/solution interface.

To examine the feasibility of depositing IrO_x on the semiconductor surface by using the strategy of photo-catalyzed surface hydrolysis, hematite (Fe₂O₃) was selected as a model semiconductor support in this work for the following two reasons. First, to generate IrO_x nanoparticles from homogeneous solution via photo-catalyzed hydrolysis of Ir³⁺, the oxidation of Ir(III) to Ir(IV) by dissolved oxygen is crucial.²⁷ On the other hand, for the generation of IrO_x on the surface of the semiconductor, the oxidation process can be accomplished by photoinduced holes. Therefore, n-type semiconductors whose valence band edges are more positive than the potential of the Ir(IV)/Ir(III) redox couple (ca. 0.50 V vs. NHE, 1.21 V vs. RHE at pH 12, the pH value of the deposition solution)^39,40 can be used as supports. Hematite (Fe₂O₃) meets this requirement because it is an n-type semiconductor,^41,42 with the valence band edge located at 2.48 V vs. NHE (3.19 V vs. RHE, pH 12).⁴³ This potential is more positive than that of the Ir(IV)/Ir(III) redox couple.^39,40 As a result, photoinduced holes in the valence band of Fe₂O₃ have enough energy to oxidize Ir(III) to Ir(IV), and then trigger the hydrolysis. Second, Fe₂O₃ is a promising photoanode material for photoelectrochemical (PEC) OER, which suffers from a high electron–hole recombination rate.⁴³ Deposition of OER catalysts on the surface of Fe₂O₃ is an effective strategy to solve this problem because OER catalysts facilitate the transfer of photo-generated holes from the semiconductor to the O₂/H₂O redox couple, thus retarding the recombination rate.^33,44,45 As IrO_x is one of the most effective OER catalysts,^31,33,44 the combination of Fe₂O₃ with IrO_x may improve the efficiency of the PEC OER.

The purposes of this work are two-fold: (1) to clarify if the photo-catalyzed hydrolysis of Ir(III) ions can be realized at the semiconductor/solution interface, and (2) to develop a method for the preparation of semiconductor/IrO_x composites by using photo-catalyzed surface hydrolysis. We demonstrated that the photo-catalyzed hydrolysis of Ir³⁺ could occur at the Fe₂O₃/solution, Ag₃PO₄/solution, and TiO₂/solution interfaces when they were illuminated with both polychromatic and monochromatic light. We also demonstrated that Fe₂O₃/IrO_x and TiO₂/IrO_x composites could be prepared by using photo-catalyzed surface hydrolysis. All the obtained composite photoanodes exhibited enhanced photoactivity toward PEC OER. These results confirmed that photo-catalyzed hydrolysis of Ir³⁺ is a universal strategy to prepare n-type semiconductor/IrO_x composites, which can be used as active photoanodes for PEC OER.

2. Experimental

2.1 Chemicals and materials

Iridium chloride trihydrate (IrCl₃·3H₂O) and tetraisopropyl titanate were purchased from Alfa Aesar company. Iron(III) chloride hexahydrate was purchased from Sinopharm Chemical Reagents Company. Silver nitrate (AgNO₃), sodium nitrate (NaNO₃), disodium hydrogen phosphate dodecahydrate (Na₂HPO₄·12H₂O) and potassium hydroxide (KOH) were purchased from Beijing Chemical Reagents Corporation. All chemicals were used as delivered and without further purification. Fluorine-doped tin oxide substrates (F:SnO₂, 8 Ω sq⁻¹) were purchased from Asahi Glass, Japan.

2.2 Preparation of Fe₂O₃, TiO₂, and Ag₃PO₄ electrodes

Before use, fluorine-doped tin oxide (FTO) coated glass was firstly cleaned by ultrasonication in acetone, ethanol and deionized water, consecutively, for 15 min each and then dried under an N₂ stream at room temperature.

The Fe₂O₃ electrodes were prepared by a hydrothermal method that was a modification of a traditional approach.⁴⁶ Firstly, FTO substrates were immersed in an aqueous solution of 1 M NaNO₃ and 50 mM FeCl₃, with the conducting side facing downward. The hydrothermal reaction was carried out at 100 °C for 18 h, during which the FeOOH films were formed on both sides of the FTO substrates. Secondly, the FeOOH film coated FTO substrates were rinsed with high purity water, placed on a stainless steel slide on the top of an Al₂O₃ crucible, and then annealed at 850 °C in air for 5 min. After annealing, the yellow FeOOH films changed to red Fe₂O₃ films. The Fe₂O₃ powders were obtained by a procedure similar to that used for obtaining the Fe₂O₃ films. After the hydrothermal reaction, the FeOOH particles in the solution were separated by centrifugation, dried in air, and annealed at 850 °C in air for 0.5 h.

The TiO₂ electrodes were also fabricated by a hydrothermal method, according to a previous work.⁴⁷ In detail, 15 ml of concentrated hydrochloric acid was diluted with 15 ml of high purity water, and mixed with 0.5 ml of tetraisopropyl titanate. This solution and FTO substrates were sealed in an autoclave and maintained at 150 °C for 5 h. The obtained TiO₂ substrates were then annealed at 550 °C in air for 3 h to enhance the crystallinity.

The Ag₃PO₄ powders were fabricated following a previous work.⁴⁸ The Na₂HPO₄ aqueous solution (20 ml, 0.2 M) was added dropwise to a AgNO₃ aqueous solution (20 ml, 0.6 M) with continuous stirring in a 100 ml flask, and the mixture was kept under stirring for 0.5 h. The yellow-colored precipitate was separated by centrifugation and dried in air at 60 °C.

2.3 Photo-catalyzed hydrolysis and photo-catalyzed surface hydrolysis of Ir(III) ions

In a typical procedure, an aqueous solution of 2 mM IrCl₃ was prepared in a beaker, and the pH of the solution was adjusted to 12 with concentrated KOH. Then, the beaker was placed under polychromatic or monochromatic light irradiation. The reaction system was kept at room temperature by using a circulating water bath during irradiation. UV and visible light was generated from a PS-SXE300UV xenon lamp. A UV reflection filter (Beijing Perfect Light Science and Technology Co., Ltd) was used to generate polychromatic UV light with wavelength ranging from 260 nm to 410 nm. Bandpass filters with wavelengths 365 nm and 498 nm were used to produce monochromatic light of required wavelength. The intensity of polychromatic light was 150 mW cm⁻², while the intensity of monochromatic light was 4.0 mW cm⁻².

For photo-catalyzed surface hydrolysis, 1.5 g of Fe₂O₃ or Ag₃PO₄ powder was dispersed in 25 ml of IrCl₃ precursor solution with continuous stirring, and the solution was irradiated with monochromatic light with λ = 498 nm. After irradiation, the powder was separated by centrifugation and the UV-vis absorption spectra of the solution were recorded.

2.4 Photo-catalyzed deposition of IrO_x on Fe₂O₃ and TiO₂ semiconductors

In a typical procedure, the Fe₂O₃ or TiO₂ electrodes prepared by the hydrothermal method (1 cm × 3 cm) were immersed in a 10 ml IrCl₃ aqueous solution (2 mM IrCl₃, pH 12) in a 25 ml beaker. The solution was irradiated with polychromatic or monochromatic irradiation for 2 h. After irradiation, Fe₂O₃ and TiO₂ electrodes were rinsed with high purity water and dried in air before use. Bandpass filters (365 nm and 498 nm) were used to provide monochromatic light. The intensities of polychromatic and monochromatic light were 150 mW cm⁻² and 4.0 mW cm⁻², respectively. The deposition of IrO_x on Fe₂O₃ was also carried out in an O₂-free system. In this experiment, the IrCl₃ solution (2 mM, pH 12) was deoxygenated by bubbling high purity N₂ for 0.5 h, and the reaction system was continuously purged with N₂ during photo-catalyzed deposition.

2.5 Characterization

X-ray diffraction (XRD) measurements were performed on a Rigaku, rint2000 advance theta-2theta powder diffractometer with Cu Kα radiation. The UV-vis absorption spectra were recorded using a UV2600 spectrophotometer (Tianmei Co., China). The morphology of Fe₂O₃ was characterized by using a JEOL JSM7500 field emission scanning electron microscope (SEM), operating at an accelerating voltage of 10 kV. Transmission electron microscopy (TEM) characterization was performed on a JEM-2100F TEM (JEOL, Japan, accelerating voltage: 200 kV). X-ray photoelectron spectroscopic (XPS) measurements were carried out on a PHI Quantera SXM scanning X-ray microprobe (ULVAC-PHI, Japan), with a monochromatic Al source at a power of 25 W. The XPS measurements were carried out at a pressure of 4.5 × 10⁻⁷ Pa. The binding energies were calibrated to the C1s line at 284.8 eV, and Shirley background subtraction was applied to the raw data before deconvolution.

2.6 Photoelectrochemical (PEC) measurements

The PEC measurements were carried out in a traditional 3-electrode configuration on a CHI 660D work station (CH Instruments Co.). The Fe₂O₃ or TiO₂ photoanodes (with or without IrO_x) were used as working electrodes, with Pt foil and a saturated calomel electrode (SCE) as the counter and reference electrodes, respectively. All of the PEC experiments were conducted in 1 M KOH solution (pH = 13.6). The relationship between the potential with respect to the SCE and that with respect to a reversible hydrogen electrode (RHE) is described by the following equation:

E_(vs. RHE) = E_(vs. SCE) + 0.0592 × pH + E⁰_(SCE) (1)

where E⁰_(SCE) is the standard potential of the SCE, which is equal to 0.24 V. For PEC measurements on the Fe₂O₃ and Fe₂O₃/IrO_x composite photoanodes, the incident light was generated by a 300 W xenon lamp, and the incident light intensity on the electrode surface was 100 mW cm⁻². On the other hand, for the TiO₂ and TiO₂/IrO_x composite photoanodes, the incident light was generated by a PS-SXE300UV xenon lamp equipped with a UV reflection filter, and the incident light intensity on the electrode surface was 150 mW cm⁻².

2.7 Detection of oxygen evolution

The quantity of O₂ evolved from the Fe₂O₃ and Fe₂O₃/IrO_x composite photoanodes was measured using an O₂ fluorescence detector (Ocean Optics, R-sensor, SN#J497).^49,50 A three-compartment PEC cell was designed for simultaneously measuring both the photocurrent and the O₂ concentration. Each electrode was placed in one compartment. The compartment containing the working electrode has a quartz window on its side for illumination. The illumination window was a circle with a diameter of 1.0 cm. The intensity of incident light on the electrode surface was 100 mW cm⁻². The oxygen sensor was immersed in the solution to detect the concentration of dissolved O₂. Before PEC measurements, the solution was purged with nitrogen for 40 minutes to remove O₂ from both the electrolyte and the gas phase in the headspace of the PEC cell, and then the system was sealed and left undisturbed for 20 min to check for leakage. The detection of evolved O₂ was conducted at a bias of 0.25 V vs. SCE (1.29 V vs. RHE, pH 13.5), and the concentration of dissolved O₂ was continuously recorded at an interval of 20 s throughout the measurement.

3. Results and discussion

3.1 Photo-catalyzed surface hydrolysis of Ir(III) ions

As demonstrated previously,²⁷ the photo-catalyzed hydrolysis of Ir³⁺ in homogeneous solution occurred under the condition that the wavelength of incident light is much shorter than 500 nm. In other words, there exists a threshold energy (ca. 2.49 eV) of the incident light to trigger the hydrolysis of Ir³⁺ in aqueous solution. Only when the energy of the incident light is larger than the threshold value, can the hydrolysis take place. However, to our surprise, we found that the threshold-wavelength-restriction ceased to hold when Fe₂O₃ nanoparticles were introduced into the reaction system.

As shown in Fig. 1a and b, after the IrCl₃ solution was illuminated for 2 h with polychromatic UV light (260 nm < λ < 410 nm) or with monochromatic UV light (λ = 365 nm), the color of the solution changed from yellow to dark blue, indicating the occurrence of photo-catalyzed hydrolysis and the formation of the IrO_x colloid.²⁷ Moreover, when the IrCl₃ solution was irradiated with monochromatic visible light (λ = 498 nm) that is very close to the threshold wavelength of ca. 500 nm, the color of the solution did not change even after 4 hour irradiation, suggesting that the hydrolysis did not occur (see Fig. 1c). All these observations are in good agreement with the previous results, confirming the existence of the threshold wavelength (ca. 500 nm) of light to trigger the hydrolysis of Ir³⁺ ions in homogeneous solution.²⁷ However, when Fe₂O₃ powder was introduced into the IrCl₃ solution to form an aqueous suspension under continuous stirring, the color of the mixed solution changed to blue upon illumination with monochromatic visible light (λ = 498 nm). Fig. 1d shows the photograph of the IrCl₃ solution after the Fe₂O₃ powder was separated by centrifugation. This observation suggests that the incident light of 498 nm, which cannot drive the hydrolysis of Ir³⁺ in homogeneous solution, triggered the reaction with the help of Fe₂O₃ particles in solution. Fig. 1e shows the UV-vis absorption spectra of the IrCl₃ solution after irradiation with 498 nm light for 2 h in the absence (red line) and presence of Fe₂O₃ powder (blue line). The shoulder peak which appeared at 320 nm after irradiation suggests the formation of [Ir(OH)₆]³⁻ in solution,^51,52 while the broad absorption peak at 580 nm implies the appearance of IrO_x nanoparticles (NPs).^22,27 The UV-vis absorption spectra provide further evidence that, in the presence of Fe₂O₃ particles, the incident light of 498 nm initiated the hydrolysis reaction and resulted in the formation of IrO_x NPs.

Fig. 1 The photographs of IrCl₃ solution (pH 12) before and after irradiation under polychromatic and monochromatic incident light with different wavelengths. (a) Polychromatic UV (260 nm < λ < 410 nm) irradiation for 2 h, (b) monochromatic UV irradiation (λ = 365 nm) for 2 h, (c) monochromatic visible irradiation (λ = 498 nm) for 4 h, and (d) monochromatic visible irradiation (λ = 498 nm) for 2 h, with Fe₂O₃ powder diffused in the solution. (e) The UV-vis absorption spectra of the solution before and after irradiation under the conditions of (c) and (d). The inset in (e) shows the local magnification of the absorption spectra.

To answer the question why the presence of Fe₂O₃ helps to trigger the reaction, we need to review the mechanism of photo-catalyzed hydrolysis of Ir³⁺ in homogeneous solution. This reaction involves two consecutive steps, and the first step is the photo-driven generation of the [Ir(OH)₆]³⁻ complex, which has no restriction concerning the wavelength of incident light, as can be confirmed from the shoulder absorption peak at 320 nm in UV-vis spectra (see Fig. 3 in ref. 27 and the red line in Fig. 1e of this work).²⁷ The second step is the polymerization of the [Ir(OH)₆]³⁻ complex, which involved the oxidation of Ir(III) species by dissolved oxygen to form IrO_x nanoparticles.²⁷ In the absence of Fe₂O₃ particles, only the light whose wavelength was much shorter than 500 nm could trigger the polymerization and oxidation of Ir(III). This is the reason for the existence of the threshold wavelength of incident light for the hydrolysis reaction in homogeneous solution. However, in the presence of Fe₂O₃ particles in solution, the incident light could be absorbed by Fe₂O₃ and generated electron–hole pairs as long as the light energy is larger than 2.1 eV (corresponding to 591 nm), which is the bandgap energy of Fe₂O₃ (see Fig. S1 in the ESI†).^43,53,54 Moreover, Fe₂O₃ is an n-type semiconductor with its valence band edge located at ca. 2.48 V vs. NHE (3.19 V vs. RHE, pH 12),⁴³ which is more positive than the redox potential of Ir(IV)/Ir(III) at ca. 0.50 V vs. NHE (1.21 V vs. RHE, pH 12).^39,40 Therefore, the incident light with wavelength in the range of 498 ≤ λ ≤ 590 nm could generate holes in the valence band of Fe₂O₃. These holes were highly oxidative and played the same role as the dissolved oxygen in homogeneous solution. They could oxidize Ir(III) species to trigger the formation of IrO_x NPs at the Fe₂O₃/solution interface. In other words, the incident light with wavelength in the range of 498 ≤ λ ≤ 590 nm can trigger the formation of IrO_x NPs at the Fe₂O₃/solution interface though it cannot in homogeneous solution. Accordingly, the hydrolysis of Ir can be easily localized to the Fe₂O₃/solution interface by using an incident light with wavelength in the range of 498 ≤ λ ≤ 590 nm. The solution in Fig. 1d shows a relatively light blue color than that in Fig. 1b, though the incident light intensity was the same. The lighter color implies that the concentration of IrO_x NPs of the solution in Fig. 1d is lower. This is because the photo-catalyzed surface hydrolysis of Ir³⁺ is a more complicated heterogeneous reaction, which involves absorption of light, generation of electron–hole pairs, transfer of holes and reaction on the surface of Fe₂O₃. On the other hand, the photochemical hydrolysis of Ir³⁺ at λ = 365 nm is a homogeneous reaction. Therefore, the reaction rate in Fig. 1d is much slower than that in Fig. 1b, resulting in a relatively light blue color.

The photo-catalyzed surface hydrolysis of Ir³⁺ on the surface of Fe₂O₃ is of great significance. (1) The presence of Fe₂O₃ breaks the restriction of the threshold wavelength of the incident light that can trigger the hydrolysis. (2) The photo-catalyzed hydrolysis can be limited to the Fe₂O₃/solution interface as long as the wavelength of incident light falls in the range 498 ≤ λ ≤ 590 nm. (3) The surface hydrolysis of Ir³⁺ leads to the deposition of IrO_x on the surface of Fe₂O₃. Inspired by the above results, we developed a facile method to prepare the Fe₂O₃/IrO_x composite OER catalyst by using the photo-catalyzed hydrolysis reaction.

3.2 The preparation of the Fe₂O₃/IrO_x OER catalyst by photo-catalyzed hydrolysis of Ir(III)

To confirm that the photo-catalyzed surface hydrolysis of Ir³⁺ can be used to deposit IrO_x on the surface of Fe₂O₃, the FTO supported Fe₂O₃ films prepared by a hydrothermal method were used as substrates. Fig. 2a shows the typical SEM image of a Fe₂O₃ film in which a porous structure composed of nanograins is clearly seen. The porous structure ensures a high specific area for Fe₂O₃, which is beneficial for the catalytic activity especially when Fe₂O₃ is used as the photoanode material for PEC OER. High resolution TEM (HR-TEM) was used to characterize the microstructure of the Fe₂O₃ nanograins, and the typical result is shown in Fig. 2b. The Fe₂O₃ film was composed of single crystalline nanograins, and the lattice spacing is 0.25 nm, which corresponds to the (110) plane of rhombohedral Fe₂O₃, indicating that the Fe₂O₃ film has a rhombohedral structure. The crystal structure of the Fe₂O₃ film was also characterized by XRD, and the diffraction patterns are shown in Fig. 2c. The diffraction peaks located at 35.44° and 63.70° are indexed to the (110) and (300) planes of rhombohedral Fe₂O₃ (JCPDS 79-7), in good agreement with the TEM results.

Fig. 2 SEM image (a), high-resolution TEM image (b), and XRD pattern (c) of the FTO supported Fe₂O₃ film.

X-ray photoelectron spectroscopy (XPS) can provide information of the elemental composition and chemical state of elements on the surface of a material. In this work, XPS was employed to examine the presence of Ir on the surface of Fe₂O₃ after photo-catalyzed surface hydrolysis, and the results are shown in Fig. 3a. The appearance of the Ir 4d and Ir 4f peaks after illumination provides solid evidence that Ir species were successfully deposited on the surface of Fe₂O₃. Fig. 3b shows the high-resolution spectrum of the Ir 4f peaks, which could be deconvoluted into two pairs of peaks. One pair of peaks was attributed to Ir(IV) (the pink line), while the other pair belonged to Ir(III). These results confirmed that the iridium oxide deposited on the surface of Fe₂O₃ was composed of Ir(III) and Ir(IV) oxides. Moreover, the high intensity of the Ir(IV) 4f peaks suggests that a large part of Ir(III) was oxidized to Ir(IV) during the photo-catalyzed surface hydrolysis. This result implies that photo-generated holes in Fe₂O₃ played a key role in the deposition of IrO_x.

Fig. 3 (a) XPS survey scans of Fe₂O₃ and Fe₂O₃/IrO_x composite electrodes. (b) XPS spectrum detailing the Ir 4f region of the Fe₂O₃/IrO_x composite electrode.

As is well known, hematite is a promising photoanode material for solar water splitting.^43,54 However, Fe₂O₃ suffers from a high electron–hole recombination rate^43,55 due to the slow water oxidation kinetics at the Fe₂O₃/solution interface. The deposition of IrO_x on Fe₂O₃ may greatly improve the photo-to-chemical conversion efficiency of the Fe₂O₃ photoanode because of the following two reasons. (1) IrO_x is an efficient OER catalyst^40,56 that may significantly improve the OER rate by accelerating the transfer of photo-generated holes from Fe₂O₃ to H₂O. (2) Photo-catalyzed deposition ensured that IrO_x was deposited on the surface sites where photo-generated holes were most available, and as a result, the photo-generated holes were easily captured by IrO_x. To examine the effect of IrO_x on the photoactivity of Fe₂O₃ toward PEC OER, the current density vs. potential (J–V) curves were recorded at the Fe₂O₃ photoanode under chopped illumination before and after the deposition of IrO_x, and the results are shown in Fig. 4. The onset potential of photocurrent at the Fe₂O₃ photoanode was ca. 0.85 V vs. RHE and the photocurrent density (J_ph) reached 0.59 mA cm⁻² at 1.23 V vs. RHE. According to literature reports,^44,45,57 this J_ph is an average value among those obtained at the Fe₂O₃ photoanodes prepared by the hydrothermal method. We did not optimize the preparation conditions of Fe₂O₃ to further improve its J_ph because this work focuses on the strategy of combining semiconductor materials with an OER catalyst by photo-catalyzed surface hydrolysis. Fig. 4 clearly shows that, after deposition of IrO_x, the resulting Fe₂O₃/IrO_x composite exhibited a significantly improved J_ph within the entire potential sweep region. Compared to the Fe₂O₃ photoanode, the onset potential of the Fe₂O₃/IrO_x composite was negatively shifted to 0.60 V vs. RHE and J_ph reached 0.98 mA cm⁻² at 1.23 V vs. RHE, which is an increase of ca. 66.1%. These results indicate that the deposited IrO_x acted as an OER catalyst, which efficiently captured the photo-induced holes and transferred them from Fe₂O₃ to H₂O, and thus accelerated the OER rate.

Fig. 4 The J–V curves of the Fe₂O₃ (black line) and the Fe₂O₃/IrO_x composite (red and green lines) photoanodes in 1 M KOH under chopped illumination (100 mW cm⁻²) at a scan rate of 5 mV s⁻¹. Both the Fe₂O₃/IrO_x composite photoanodes were prepared by photo-catalyzed hydrolysis of IrCl₃ solution (2 mM, pH 12) under 150 mW cm⁻² polychromatic UV illumination, with one of them (green line) obtained in IrCl₃ solution that was deoxygenated by purging with high purity N₂.

Moreover, as a control experiment, J_ph was also recorded at the Fe₂O₃ photoanode that was first immersed in 2 mM IrCl₃ (pH 12) for 2 h but without UV irradiation, and then rinsed with distilled water. The J–V response of this photoanode showed no significant difference compared to that of the pure Fe₂O₃ electrode (Fig. S2 in the ESI†). This result confirms that illumination is the key factor for the deposition of IrO_x.

As mentioned above, the oxidation of Ir(III) to Ir(IV) by dissolved O₂ was crucial for the photo-catalyzed hydrolysis of Ir³⁺ and the formation of IrO_x in homogeneous solution.²⁷ On the other hand, for the photo-catalyzed surface hydrolysis of Ir³⁺, photo-generated holes were believed to act as the oxidizing agent instead of dissolved O₂. To examine this assumption, a control experiment was carried out in an O₂-free IrCl₃ solution (2 mM, pH 12), which was deoxygenated by bubbling high purity N₂ for 0.5 h before reaction and purging with N₂ during the two hour reaction. The J–V response (green line) of the resulting Fe₂O₃/IrO_x composite is also shown in Fig. 4. This Fe₂O₃/IrO_x composite photoanode exhibited nearly the same J–V characteristics as that of the Fe₂O₃/IrO_x composite prepared in IrCl₃ solution with dissolved O₂. This observation provided direct evidence that photo-generated holes participated in the photo-deposition process by oxidizing Ir(III) to Ir(IV), and then triggered the formation of IrO_x.

All the above-mentioned Fe₂O₃/IrO_x composite photoanodes were prepared by using polychromatic UV light (150 mW cm⁻²). In fact, the composite photoanode could also be prepared by using 498 nm monochromatic visible light even though the energy density of the incident light was only 4 mW cm⁻². It should be pointed out here that, though the deposition of IrO_x could be localized on the surface of Fe₂O₃ (with IrO_x formed in solution) by using the 498 nm incident light, the deposition rate was much smaller than that achieved by using polychromatic UV light. As a result, the amount of IrO_x deposited on the surface of Fe₂O₃ during the same deposition time was very small. We believe that the use of a monochromatic visible light (498 ≤ λ ≤ 590 nm) source with high energy density (such as a laser) can solve this problem. On the basis of the above discussion, we believe that photo-catalyzed deposition of IrO_x by using polychromatic UV light was a simple and efficient way to prepare the Fe₂O₃/IrO_x composite.

The mechanism of the photo-catalyzed deposition can be described as follows. Firstly, irradiation provided the activation energy for the Ir(H₂O)₃Cl₃ complex, which was the product of aquation of IrCl₃,^27,58 to convert into the [Ir(OH)₆]³⁻ complex. Secondly, part of [Ir(OH)₆]³⁻ was oxidized to [Ir(OH)₆]²⁻ by photo-generated holes at the Fe₂O₃/solution interface, and the polymerization of [Ir(OH)₆]²⁻ with another [Ir(OH)₆]²⁻ or with [Ir(OH)₆]³⁻ triggered the formation of IrO_x. It should be pointed out here that, when the reaction system was irradiated with polychromatic UV light, we could not exclude the possibility that the IrO_x NPs generated in homogeneous solution contributed to the deposition of IrO_x on the surface of Fe₂O₃.

3.3 Stability and Faraday efficiency of the Fe₂O₃/IrO_x composite for PEC OER

Fig. 5a shows the current density–time curves obtained in 1 M KOH at a constant potential of 0.25 V (1.29 V vs. RHE, pH 13.5) under 100 mW cm⁻² chopped illumination. Both the Fe₂O₃ and the Fe₂O₃/IrO_x composite photoanodes exhibited very good stability in alkaline solution. The photocurrent density of the Fe₂O₃ photoanode stabilized at ca. 0.70 mA cm⁻² and showed no decay throughout the 10 hour measurement. On the other hand, for the Fe₂O₃/IrO_x composite photoanode, the photocurrent density even slowly increased in the first 8.5 h and stabilized at ca. 0.91 mA cm⁻², which is 30% higher than the photocurrent density of the Fe₂O₃ photoanode. Although the reason for the gradual increase of photocurrent density is not quite clear, we believe that this observation was related to the activation of the IrO_x co-catalyst.⁴⁰

The high stability of both the Fe₂O₃ and the Fe₂O₃/IrO_x composite photoanodes ensured that, during PEC OER, we could measure the amount of O₂ evolution, which is very important to evaluate the photoanodes and to obtain the Faraday efficiency. In this work, the amount of O₂ evolution from both photoanodes was measured by using an O₂ fluorescence probe immersed in solution.^49,50Fig. 5b shows the variation of the detected and calculated amount of O₂ evolution as a function of time at both the Fe₂O₃ and the Fe₂O₃/IrO_x photoanodes under continuous illumination (100 mW cm⁻²). The point of zero time was taken as the time at which illumination started. The negative time values on the abscissa refer to the time duration within which a bias of 1.29 V vs. RHE was applied to the photoanodes but with no illumination. No O₂ was detected in this period, confirming that (1) electrochemical OER did not occur at 1.29 V vs. RHE on both electrodes and (2) the PEC cell had good gas tightness. When illumination was turned on, O₂ started to evolve from both photoanodes, as can be seen in Fig. 5b. After 40 min PEC OER, O₂ production from the Fe₂O₃/IrO_x and the Fe₂O₃ photoanodes was detected to be 2.70 μmol and 1.92 μmol, respectively. The high O₂ evolution rate on the Fe₂O₃/IrO_x composite provides solid evidence that the deposited IrO_x acted as an OER co-catalyst and greatly accelerated the PEC OER. The amount of O₂ evolved from both photoanodes could also be calculated by integrating the photocurrent and assuming that the photocurrent arose from the 4-electron OER. The calculated amount of O₂ is 2.71 μmol for the Fe₂O₃/IrO_x and 1.97 μmol for the Fe₂O₃ photoanodes, corresponding to a Faraday efficiency of 99.6% and 97.4%, respectively. These results indicate that nearly all of the photocurrents were engaged in OER on both the Fe₂O₃/IrO_x and the Fe₂O₃ photoanodes.

Fig. 5 (a) The current density–time response of the Fe₂O₃ and Fe₂O₃/IrO_x composite photoanodes in 1 M KOH at a constant bias of 0.25 V vs. SCE (1.29 V vs. RHE) under chopped illumination (100 mW cm⁻²). (b) The detected and calculated amount of oxygen evolution on the Fe₂O₃ and the Fe₂O₃/IrO_x composite photoanodes in 1 M KOH at a constant bias of 0.25 V vs. SCE (1.29 V vs. RHE) under continuous illumination (100 mW cm⁻²). The illumination window was a circle with a diameter of 1.0 cm.

3.4 Photo-catalyzed surface hydrolysis of Ir(III) as a universal strategy to prepare n-type semiconductor/IrO_x composites for PEC OER

We believe that the photo-catalyzed surface hydrolysis of Ir(III) described in this work can be realized on the surface of many n-type semiconductors, as long as the valence band edge of the semiconductor is located at a more positive potential than the redox potential of Ir(IV)/Ir(III). As the redox potential of O₂/H₂O (1.23 V vs. RHE) is more positive than the redox potential of Ir(IV)/Ir(III), a semiconductor that is able to catalyze water oxidation is certainly suitable for the photo-catalyzed deposition of IrO_x. Herein, we use two kinds of n-type semiconductors Ag₃PO₄ and TiO₂ as examples to show that the photo-catalyzed hydrolysis is a universal strategy to prepare n-type semiconductor/IrO_x composites.

Ag₃PO₄ has been demonstrated as a possible photoanode material for PEC OER.^59,60 The band-gap energy of Ag₃PO₄ is 2.36 eV,⁵⁹ which corresponds to a band-edge absorption of λ = 530 nm. The valence band edge of Ag₃PO₄ is located at 2.84 V vs. NHE (3.55 V vs. RHE, pH 12),^59,60 ensuring that the photo-induced holes have an energy high enough to oxidize Ir(III). In our experiments, Ag₃PO₄ powder was added to the IrCl₃ solution (2 mM, pH 12) and the mixture was irradiated under 498 nm light (4 mW cm⁻²) for 4 h. The photographs and the UV-vis absorption spectra of the solution before and after irradiation are shown in Fig. 6a. The color of the solution changed from yellow to blue, implying the formation of IrO_x. In addition, the peak that was characteristic of IrO_x nanoparticles was observed at 580 nm in UV-vis spectra, confirming that photo-catalyzed surface hydrolysis of Ir³⁺ could also take place on the surface of Ag₃PO₄.

Fig. 6 (a) Photographs and UV-vis absorption spectra of the IrCl₃ solution (2 mM, pH 12) before and after irradiation under 498 nm light (4 mW cm⁻²) for 4 h with Ag₃PO₄ powder in the solution, the inset shows the local magnification of the absorption spectra. (b) The J–V curves of TiO₂ and TiO₂/IrO_x composite photoanodes in 1 M KOH, under chopped UV light (150 mW cm⁻²) at a scan rate of 5 mV s⁻¹.

TiO₂ has long been studied as a photo-anode material for water splitting.^61,62 The valence band edge of TiO₂ is located at ca. 3.0 V vs. NHE (3.7 V vs. RHE, pH 12),⁶¹ which is more positive than the Ir(IV)/Ir(III) potential. To test whether or not IrO_x can be deposited on the surface of TiO₂ under illumination, an FTO supported TiO₂ electrode was immersed in IrCl₃ solution (2 mM, pH 12) and illuminated under polychromatic UV light (150 mW cm⁻²) for 0.5 h. Then, the obtained electrode was used as a photoanode for water splitting. Fig. 6b shows the J–V curves of the TiO₂ photoanode before and after the photo-catalyzed deposition of IrO_x. It can be seen clearly in Fig. 6b that, after illumination in IrCl₃ solution, the resulting photoanode exhibited a significantly improved photoactivity toward PEC OER within the entire potential sweep region. This result suggests that IrO_x was successfully deposited on the surface of TiO₂ and acted as an OER co-catalyst on the surface of TiO₂. Similar results were obtained on the Ag₃PO₄/IrO_x electrode. But the improvement was not obvious compared with the results obtained on Fe₂O₃ and TiO₂. We guess this is because of the mismatch of the energy levels of Ag₃PO₄ and IrO_x or high electrical resistance between Ag₃PO₄ and IrO_x. Further study is needed to find out the reason and suitable conditions to deposit IrO_x effectively on Ag₃PO₄ using the photo-catalyzed hydrolysis method.

All the above results indicate that photo-catalyzed surface hydrolysis of Ir³⁺ is a universal phenomenon that takes place on the surface of n-type semiconductors whose valence band edge is more positive than the redox potential of Ir(IV)/Ir(III). Moreover, the successful deposition of IrO_x on the surface of Fe₂O₃, TiO₂, and Ag₃PO₄ demonstrates that surface hydrolysis under illumination is a general route to couple IrO_x with n-type semiconductor photoanodes.

4. Conclusions

In summary, based on our previous work on photo-driven hydrolysis of IrCl₃ in homogeneous solution,²⁷ we demonstrated in this work that the photo-catalyzed surface hydrolysis could be realized on n-type semiconductors such as Fe₂O₃, TiO₂, and Ag₃PO₄. The photo-catalyzed hydrolysis follows a mechanism different from that in the homogeneous solution. For surface hydrolysis under illumination, the photo-generated holes played the key role of an oxidizing agent, which oxidized Ir(III) to Ir(IV) species and then triggered the formation of IrO_x. We proved that the hydrolysis could be localized at the Fe₂O₃/solution interface by using monochromatic visible light of 498 nm, which could not induce hydrolysis in homogeneous solution. As IrO_x is an efficient OER catalyst, the photo-catalyzed surface hydrolysis can be developed as a universal strategy to couple the OER co-catalyst IrO_x with semiconductor photoanode materials, and the resulting semiconductor/IrO_x composites could be employed as promising photoanodes for PEC OER. We demonstrated that, after photo-catalyzed deposition of IrO_x, the Fe₂O₃ and the TiO₂ photoanodes exhibited significantly improved photoactivity toward PEC OER. We believe that the surface hydrolysis under illumination may open a new avenue for the construction of semiconductor/co-catalyst composites for solar-powered water splitting.

Acknowledgements

We gratefully acknowledge the financial support of this work by National Natural Science Foundation of China (NSFC 51672017 and 21173016) and Beijing Natural Science Foundation (2142020 and 2151001).

Notes and references

1. V. Fierro, G. Muñiz, G. Gonzalez-Sánchez, M. L. Ballinas and A. Celzard, J. Hazard. Mater., 2009, 168, 430–437.
2. D. W. Lee and B. R. Yoo, J. Ind. Eng. Chem., 2014, 20, 3947–3959.
3. J. Lee, S. Zhang and S. Sun, Chem. Mater., 2013, 25, 1293–1304.
4. S. Musić, S. Krehula, S. Popović and Ž. Skoko, Mater. Lett., 2003, 57, 1096–1102.
5. Q. Yang, Z. Lu, J. Liu, X. Lei, Z. Chang, L. Luo and X. Sun, Prog. Nat. Sci., 2013, 23, 351–366.
6. J.-Y. Li, S. Xiong, J. Pan and Y. Qian, J. Phys. Chem. C, 2010, 114, 9645–9650.
7. B. Liu and H. C. Zeng, J. Am. Chem. Soc., 2003, 125, 4430–4431.
8. Z. Miao, D. Xu, J. Ouyang, G. Guo, X. Zhao and Y. Tang, Nano Lett., 2002, 2, 717–720.
9. M. Niederberger and G. Garnweitner, Chem. – Eur. J., 2006, 12, 7282–7302.
10. C. Cheng, Y. Huang, N. Wang, T. Jiang, S. Hu, B. Zheng, H. Yuan and D. Xiao, ACS Appl. Mater. Interfaces, 2015, 7, 9526–9533.
11. T. D. Schladt, T. Graf, O. Köhler, H. Bauer, M. Dietzsch, J. Mertins, R. Branscheid, U. Kolb and W. Tremel, Chem. Mater., 2012, 24, 525–535.
12. G. Zhu, H. Xu, Y. Xiao, Y. Liu, A. Yuan and X. Shen, ACS Appl. Mater. Interfaces, 2012, 4, 744–751.
13. X. Hong, Z. Wang, W. Cai, F. Lu, J. Zhang, Y. Yang, N. Ma and Y. Liu, Chem. Mater., 2005, 17, 1548–1552.
14. J. Zhong, F. Chen and J. Zhang, J. Phys. Chem. C, 2010, 114, 933–939.
15. D. A. Schwartz, N. S. Norberg, Q. P. Nguyen, J. M. Parker and D. R. Gamelin, J. Am. Chem. Soc., 2003, 125, 13205–13218.
16. C. X. Guo, J. Xie, H. Yang and C. M. Li, Adv. Sci., 2015, 2, 1500135.
17. C. Yuan, X. Zhang, L. Su, B. Gao and L. Shen, J. Mater. Chem., 2009, 19, 5772–5777.
18. C. Justin Raj, B. C. Kim, W.-J. Cho, S. Park, H. T. Jeong, K. Yoo and K. H. Yu, J. Electroanal. Chem., 2015, 747, 130–135.
19. M. R. Buck, A. J. Biacchi and R. E. Schaak, Chem. Mater., 2014, 26, 1492–1499.
20. C. W. Wang, S. Yang, W. Q. Fang, P. Liu, H. Zhao and H. G. Yang, Nano Lett., 2016, 16, 427–433.
21. F. Ooi, J. S. DuChene, J. Qiu, J. O. Graham, M. H. Engelhard, G. Cao, Z. Gai and W. D. Wei, Small, 2015, 11, 2649–2653.
22. Y. Zhao, E. A. Hernandez-Pagan, N. M. Vargas-Barbosa, J. L. Dysart and T. E. Mallouk, J. Phys. Chem. Lett., 2011, 2, 402–406.
23. L. Chen, J. Xu, D. A. Tanner, R. Phelan, M. van der Meulen, J. D. Holmes and M. A. Morris, Chem. – Eur. J., 2009, 15, 440–448.
24. F. Kim, J. H. Song and P. Yang, J. Am. Chem. Soc., 2002, 124, 14316–14317.
25. L. Maretti, P. S. Billone, Y. Liu and J. C. Scaiano, J. Am. Chem. Soc., 2009, 131, 13972–13980.
26. M. Sakamoto and T. Majima, Bull. Chem. Soc. Jpn., 2010, 83, 1133–1154.
27. D. Xu, P. Diao, T. Jin, Q. Wu, X. Liu, X. Guo, H. Gong, F. Li, M. Xiang and Y. Ronghai, ACS Appl. Mater. Interfaces, 2015, 7, 16738–16749.
28. S. Cherevko, S. Geiger, O. Kasian, N. Kulyk, J.-P. Grote, A. Savan, B. R. Shrestha, S. Merzlikin, B. Breitbach, A. Ludwig and K. J. J. Mayrhofer, Catal. Today, 2016, 262, 170–180.
29. S. W. Lee, M. Janyasupab, C.-C. Liu and R. M. Sankaran, Catal. Today, 2013, 211, 137–142.
30. J. Park, M. Kim and S. Kim, Sens. Actuators, B, 2014, 204, 197–202.
31. E. Fabbri, A. Habereder, K. Waltar, R. Kötz and T. J. Schmidt, Catal. Sci. Technol., 2014, 4, 3800–3821.
32. D. Chandra, D. Takama, T. Masaki, T. Sato, N. Abe, T. Togashi, M. Kurihara, K. Saito, T. Yui and M. Yagi, ACS Catal., 2016, 6, 3946–3954.
33. S. D. Tilley, M. Cornuz, K. Sivula and M. Grätzel, Angew. Chem., Int. Ed., 2010, 122, 6549–6552.
34. E. M. Steinmiller and K. S. Choi, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 20633–20636.
35. T. Jin, P. Diao, Q. Wu, D. Xu, D. Hu, Y. Xie and M. Zhang, Appl. Catal., B, 2014, 148–149, 304–310.
36. T. Jin, P. Diao, D. Xu and Q. Wu, Electrochim. Acta, 2013, 114, 271–277.
37. A. Irshad and N. Munichandraiah, J. Electrochem. Soc., 2015, 162, H235–H243.
38. Y. Liu, Y. Zhou, G. Chen, T. Guo, L. Wang, X. Huang and W. Zeng, Mater. Lett., 2015, 148, 155–158.
39. T. Kuwabara, E. Tomita, S. Sakita, D. Hasegawa, K. Sone and M. Yagi, J. Phys. Chem. C, 2008, 112, 3774–3779.
40. T. Nakagawa, C. A. Beasley and R. W. Murray, J. Phys. Chem. C, 2009, 113, 12958–12961.
41. J. Sun, D. K. Zhong and D. R. Gamelin, Energy Environ. Sci., 2010, 3, 1252–1261.
42. L. Xi, P. D. Tran, S. Y. Chiam, P. S. Bassi, W. F. Mak, H. K. Mulmudi, S. K. Batabyal, J. Barber, J. S. C. Loo and L. H. Wong, J. Phys. Chem. C, 2012, 116, 13884–13889.
43. K. Sivula, F. Le Formal and M. Grätzel, ChemSusChem, 2011, 4, 432–449.
44. W. Li, S. W. Sheehan, D. He, Y. He, X. Yao, R. L. Grimm, G. W. Brudvig and D. Wang, Angew. Chem., Int. Ed., 2015, 54, 11428–11432.
45. Y.-S. Hu, A. Kleiman-Shwarsctein, A. J. Forman, D. Hazen, J.-N. Park and E. W. McFarland, Chem. Mater., 2008, 20, 3803–3805.
46. L. Vayssieres, N. Beermann, S.-E. Lindquist and A. Hagfeldt, Chem. Mater., 2001, 13, 233–235.
47. Y. C. Pu, G. Wang, K. D. Chang, Y. Ling, Y. K. Lin, B. C. Fitzmorris, C. M. Liu, X. Lu, Y. Tong, J. Z. Zhang, Y. J. Hsu and Y. Li, Nano Lett., 2013, 13, 3817–3823.
48. X. Yang, H. Cui, Y. Li, J. Qin, R. Zhang and H. Tang, ACS Catal., 2013, 3, 363–369.
49. D. Hu, P. Diao, D. Xu and Q. Wu, Nano Res., 2016, 9, 1735–1751.
50. W. D. Chemelewski, H. C. Lee, J. F. Lin, A. J. Bard and C. B. Mullins, J. Am. Chem. Soc., 2014, 136, 2843–2850.
51. T. Ioroi, N. Kitazawa, K. Yasuda, Y. Yamamoto and H. Takenaka, J. Electrochem. Soc., 2000, 147, 2018.
52. J.-Y. Chen, Y.-M. Chen, Y. Sun, J.-F. Lee, S.-Y. Chen, P.-C. Chen and P.-W. Wu, Ceram. Int., 2014, 40, 14983–14990.
53. N. Beermann, L. Vayssieres, S.-E. Lindquist and A. Hagfeldt, J. Electrochem. Soc., 2000, 147, 2456–2461.
54. A. Kay, I. Cesar and M. Gratzel, J. Am. Chem. Soc., 2006, 128, 15714–15721.
55. B. C. Fitzmorris, J. M. Patete, J. Smith, X. Mascorro, S. Adams, S. S. Wong and J. Z. Zhang, ChemSusChem, 2013, 6, 1907–1914.
56. S. Siracusano, N. van Dijk, E. Payne-Johnson, V. Baglio and A. S. Aricò, Appl. Catal., B, 2015, 164, 488–495.
57. E. S. Cho, M. J. Kang and Y. S. Kang, Phys. Chem. Chem. Phys., 2015, 17, 16145–16150.
58. C. S. G. James and C. Chang, Inorg. Chem., 1965, 4, 209–215.
59. Z. Yi, J. Ye, N. Kikugawa, T. Kako, S. Ouyang, H. Stuart-Williams, H. Yang, J. Cao, W. Luo, Z. Li, Y. Liu and R. L. Withers, Nat. Mater., 2010, 9, 559–564.
60. Q. Wu, P. Diao, J. Sun, D. Xu, T. Jin and M. Xiang, J. Mater. Chem. A, 2015, 3, 18991–18999.
61. A. Fujishima, X. Zhang and D. Tryk, Surf. Sci. Rep., 2008, 63, 515–582.
62. M. Ni, M. K. H. Leung, D. Y. C. Leung and K. Sumathy, Renewable Sustainable Energy Rev., 2007, 11, 401–425.

---

Footnote

† Electronic supplementary information (ESI) available: UV-vis absorption spectrum of Fe₂O₃ and J–V curves of the Fe₂O₃ electrode before and after immersion in IrCl₃ solution. See DOI: 10.1039/c6cp06821a

---