The electronic structure of semiconducting materials at the electrode/electrolyte interface plays a vital role in the process of photoelectrochemical (PEC) water splitting. In this work, we could reliably tune the surface defect density of ZnO films through changing the annealing temperature, thereby optimizing the PEC performance. The surface photovoltage (SPV) of ZnO films could be obtained by Kelvin probe force microscopy and compared to insightfully understand the effect of the annealing temperature on the performance of the electrode in PEC water splitting. The minimum SPV annealed at 450°C indicated low surface defect density, eventually resulting in an enhanced photoelectrochemical performance. The applied bias photo-to-current efficiency of ZnO films annealed at 450°C reached 0.237%, about 7.4 times that of unannealed ZnO photoanode. This work provides an effective method for the rational fabrication of efficient photoelectrodes for the realization of high-performance photoelectrochemical water splitting.

1. Introduction

Hydrogen generated from photoelectrochemical (PEC) water splitting has shown significant potential for solar energy conversion to meet the steady increase of energy demands [1]. In principle, the efficiency of PEC water splitting mainly depends on the following factors: (1) light absorption, (2) photogenerated charge separation and transport, and (3) surface chemical reaction [2, 3]. For a PEC system, the light absorption is determined by the bandgap of photoelectrodes, and the driving force of charge separation originates from the electric field in the depletion region at the photoelectrode/electrolyte interface. Thus, the interfacial electronic structure of semiconducting materials plays a vital role in the process of photoelectrochemical water splitting [4]. Therefore, the understanding of the change in electronic structure of semiconductor materials is favourable for the rational design of efficient photoanode. Thus, the key in a PEC system is the selection of high-performance photoelectrodes.

Although narrow-bandgap semiconductors, such as silicon and III–IV group of composites, show high light harvesting efficiency, the surface corrosion hinders their lifetime even in a sacrificial electrolyte [5, 6]. Among various N-type metal oxides [712], zinc oxide (ZnO) has been extensively investigated as photoanode for photoelectrochemical water splitting ascribed to its low cost, excellent electron mobility (115–155 cm2·V−1·s−1), and appropriate conduction/valence band edges for water splitting [1317]. As a postprocessing method, annealing has been proved to be a valid method to change the interfacial electronic structure, resulting in a significant impact on the properties of ZnO nanomaterials [18, 19]. Wu et al. [20] demonstrated that the work function of ZnO nanorod arrays exhibited a gradual increase after a decrease with the increase in the annealing temperature. Wang et al. [21] found that the improved PEC performance of ZnO nanowires (NWs) was mainly attributed to the larger depleted width in ZnO NWs caused by annealing. Further, the carrier concentration of ZnO films by electron beam annealing is about 2~3 orders of magnitude higher than that of unannealed ZnO thin films [22]. However, there is still a lack of in-depth understanding about the change in interfacial electronic structure for the enhanced PEC performance caused by different annealing temperatures.

Previously, we have constructed various photoanodes based on ZnO nanomaterials. And surface modification materials, doping, and structure optimization were sequentially taken into combination to form ZnO-based composite nanostructures, in which each component worked synergistically to compensate each other [1316]. In conclusion, there are some significant differences between previous works and current work as for the following aspects. Previous works put emphasis on the realization of highly efficient photoanode based on ZnO nanomaterials to raise the light harvesting efficiency, reduce the charge recombination, and accelerate the surface reaction. However, multistep and complex processes are always introduced into these strategies. Thus, this work adopts a one-step process and posttreatment to tune the PEC performance, and we mainly focused on the change of interfacial electronic structure by the KPFM. The surface photovoltage (SPV) of ZnO films under different annealing temperatures was obtained through calculating the difference between the surface potentials under UV illumination and in the dark measured by KPFM. After annealing at 450°C, the minimum SPV of ZnO films indicated low defect density, which contributed to the improved photocurrent density and enhanced photoelectrochemical performance. This simple and novel method also allows us to have more fundamental understanding of the effects of ZnO itself on the PEC performance.

2. Materials and Methods

Commercially available indium tin oxide- (ITO-) coated glass (2 × 2 cm2) was used as the substrate. ITO substrates were sequentially sonicated in acetone, isopropanol, ethanol, and deionized water, each for 10 minutes, and then dried with nitrogen gas flow. The ZnO film was deposited on the ITO substrate via the radiofrequency (RF) magnetron sputtering method as reported [23]. After deposition, ZnO films were annealed in a vacuum drying oven at 350°C, 450°C, and 550°C for 2 hours, respectively. The corresponding samples were named ZnO-350, ZnO-450, and ZnO-550.

The morphology and crystalline structure of the ZnO film were characterized by field-emission scanning electron microscopy (FESEM, FEI QUANTA 3D) and X-ray diffraction (XRD) (Rigaku DMAX-RB with Cu Kα radiation), respectively. The light absorption and photoluminescence spectrum were recorded by a UV-vis spectrophotometer (UV-2500, Shimadzu, Japan) and a continuous He-Cd (325 nm) laser as an excitation source, respectively. KPFM measurements were tested under a commercially available atomic force microscope (ICON, Bruker Inc.) in lift mode at about 25°C. PtIr-coated silicon cantilevers (Olympus, OSCM-PIT) with a nominal resonance frequency of 75 kHz were used as conductive probes.

A copper wire was secured on the exposed conductive parts of ITO with a silver conducting paint, and the rest of the substrate was subsequently encapsulated with epoxy resin except the active working area, confirming the successful preparation of the photoanode. A three-electrode system used for various electrochemical measurements on an electrochemical workstation (Solartron SI 1287/SI 1260) consists of a photoanode as the working electrode, a coiled Pt wire as the counter electrode, and an Ag/AgCl reference electrode. Neutral aqueous Na2SO4 solution (0.5 M) was chosen as the electrolyte. The simulated sunlight provided by a solar simulator (Oriel, 91159A) is incident from the front of the photoanode, and the intensity of the simulated sunlight was 60 mW/cm2 measured by a Si diode (Newport). The applied bias photo-to-current efficiency () was calculated from curves, assuming 100% faradaic efficiency through the following equation [24]: where is the applied external potential versus the reversible hydrogen electrode (RHE). is the externally measured current density at . is the power density of the incident light. The potentials were measured versus the Ag/AgCl reference electrode and converted to the RHE scale using the Nernst function: where is the converted potential versus RHE. is the external potential measured against the Ag/AgCl reference electrode. is the standard electrode potential of the Ag/AgCl reference electrode (0.1976 V versus RHE at 25°C).

3. Results and Discussion

Figure 1(a) exhibits a representative cross-sectional scanning electron microscope (SEM) image of the ZnO film on the ITO prior to annealing, indicating that the film is composed of multiple ZnO columnar grains. The grain size calculated from the top view image in the inset of Figure 1(a) is in the range of 30–100 nm. Furthermore, the corresponding XRD pattern is used to characterize the structure of an unannealed ZnO film. As shown in Figure 1(b), XRD of ZnO films exhibits that all diffraction peaks can be corresponded to the planes of highly crystalline ZnO (JCPDS file no. 36-1451). And the apparent diffraction peak of (002) at 34.4° demonstrates that the ZnO film is grown preferentially along -axis direction on the ITO [25].

The schematic illustration of the ZnO film on the ITO upon the bandgap illumination is shown in Figure 2(a). Under simulated sunlight, electrons and holes are generated in the conduction band () and valence band (), respectively. The photogenerated electrons transfer to the Pt electrode for the hydrogen evolution in the PEC cell device. On the other hand, the photogenerated holes in of ZnO participate in the oxidation evolution to finish the whole process.

In order to study the photoelectrochemical properties of the prepared photoanodes, the PEC measurements of various as-prepared photoanodes were characterized by linear sweep voltammetry (LSV) curves in the dark and under simulated sunlight illumination. A very small dark photocurrent in the scale of 10−3 mA/cm2 is shown in Figure 2(b), implying that the observed photocurrent is related to the photogenerated charge carriers. Under illumination, a significant increase in photocurrent density was observed, and the ZnO-450 photoanode exhibited the highest photocurrent density in the whole investigated potential range, ascribed to the lowest density of trap states at the surface among those of other samples. The photocurrent density of unannealed ZnO, ZnO-350, ZnO-450, and ZnO-550 reached 0.19, 0.559, 0.586, and 0.457 mA/cm2 at 1 V vs. Ag/AgCl, respectively. According to (1) and (2), the ABPE of different photoanodes was calculated from the LSV curve. As shown in Figure 3(c), the ZnO-450 photoanode exhibited the highest efficiency of 0.237%, about 7.4 times that of the unannealed ZnO photoanode (0.032%).

To attain a deep understanding of the PEC performance, electrochemical impedance spectroscopy (EIS) was used to explore the electrochemical behavior at the photoelectrode/electrolyte interface at 0 V vs. Ag/AgCl under white light illumination. Figure 2(d) reveals the typical EIS Nyquist plots for four different photoanodes, and the equivalent circuit in Figure 2(d) was used to fit the Nyquist plots, in which , and CPE represented series resistance, interfacial charge-transfer resistance at the photoelectrode/electrolyte interface, and constant phase angle element, respectively. The smaller size of the arc in an EIS Nyquist plot indicated a faster charge transfer and less recombination at the photoelectrode/electrolyte interface. The results obtained by fitting revealed that the corresponding of unannealed ZnO, ZnO-350, ZnO-450, and ZnO-550 was 5.18, 2.6, 2.18, and 3.98 kΩ, respectively. The smallest of the ZnO-450 photoanode could result in the highest photocurrent density and ABPE consistent with the LSV and ABPE curve shown in Figures 2(b) and 2(c).

To explore the photocurrent improvement, light absorption properties of ZnO films annealed under different temperatures were investigated by the UV-vis absorbance spectra. As shown in Figure 4(a), all the samples showed similar optical absorption in the UV region. As the annealing temperature increases, the optical absorption edges showed a slight red shift. Further, the optical bandgap of different samples was estimated based on the following equation [26]: where is the absorption coefficient of the material, is the energy of photon, represents the index which depends on the electronic transition of the semiconductor (for direct bandgap semiconductor ZnO, ), and is the proportionality constant related to the material. Therefore, the bandgap energy could be obtained from the intercept of the tangent line in the plot of versus energy. As shown in Figure 4(b), the optical bandgap decreased from 3.11 eV to 3.09 eV as the annealing temperature increased. Thus, the negligible deviation (~0.02 eV) meant that the light absorption is not the dominating reason for the improvement of the photocurrent.

To reveal the effect of annealing temperature on the charge separation, the room temperature photoluminescence (PL) spectra were investigated to study the change of the surface defect density of ZnO films annealed at different temperatures. As shown in Figure 4(c), the samples showed two main peaks: the sharp one around 390 nm was ascribed to the near-band-edge (NBE) emission and the other broad one was in the range of 440–660 nm corresponding to the intrinsic defect-related energy [27, 28]. In Figure 4(d), a red shift of NBE emission peak was observed with the increase in annealing temperature, which is consistent with the change of optical bandgap in Figure 4(b). When the annealing temperature was no more than 450°C, the intensity of broad deep-level (DL) emission decreased ascribed to the improved crystalline quality after annealing. However, when the annealing temperature reached 550°C, the intensity of DL emission revealed a significant increase due to the increased probability of the radiative recombination which is associated with the deep-level emission. Therefore, the low intensity of DL emission of ZnO-450 indicated reduced electron-hole recombination.

To further characterize the surface defect density of ZnO films annealed at different temperatures, Kelvin probe force microscopy (KPFM) was employed under dark and light conditions to measure the surface potential difference [2931]. As shown in Figure S1 in Supplementary Materials, the contact potential difference () between a sample and a tip was acquired for different samples in the dark and under UV illumination, respectively, and the corresponding average was statistically analysed as shown in Figures 3(a)3(d). As shown in Figure 4(f), the difference in work function in the dark and under UV illumination is defined as the surface photovoltage (SPV), which could be acquired based on the following equations [4, 32, 33]: where and are the work functions of the tip and sample, respectively, and is the electronic charge. Figure 4(e) exhibits the average and SPV of different samples, and the SPV of unannealed ZnO, ZnO-350, ZnO-450, and ZnO-550 was 95.8 mV, 73.3 mV, 71.6 mV, and 89.9 mV, respectively. As the SPV was used to demonstrate the change of the band bending at the photoelectrode/electrolyte surface, the SPV could be used to indicate the density of surface states [30, 34]. Thus, the minimum SPV value of ZnO-450 suggested the low density of trap states at the surface, which was in agreement with the results of photoluminescence spectra in Figure 4(c) and contributed to the enhancement of the PEC performance.

4. Conclusions

In summary, the photoelectrochemical performances of ZnO films on the ITO substrate were investigated through the work function study at the electrode/electrolyte interface under different annealing temperatures. After annealing at 450°C, the minimum SPV of ZnO films indicated low defect density, which contributed to the improved photocurrent density and enhanced photoelectrochemical performance. The ZnO films annealed under 450°C exhibited an efficiency of 0.237%, which was much higher than that of unannealed ZnO films (0.032%). As a whole, this work provided a potential method to design innovative photoanodes for photoelectrochemical water splitting.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

There is no conflict of interest regarding the publication of this paper.


The authors acknowledge the funding support from the Project of Thousand Youth Talents and National Natural Science Foundation of China (NNSFC) (21705122).

Supplementary Materials

The images of contact potential differences in the dark and under UV illumination measured by KPFM are provided. (Supplementary Materials)