1 INTRODUCTION

Metallic corrosion can cause huge economic losses and great harm to human's living environments. Traditional anti-corrosion technologies, such as organic coating and cathodic protection, cause not only waste of resources and energy, but also environmental pollution. The photoelectrochemical cathodic protection (PECCP) technology couples semiconductor materials with metallic materials and transfers the photoinduced electrons generated by the semiconductor materials under light illumination to the metallic materials to protect them. Therefore, the PECCP technology is considered an environmentally friendly novel corrosion protection technology (Bu et al., 2018). Semiconductor materials that can be used for PECCP mainly include TiO₂ (Ohko et al., 2001; Sun et al., 2013; Momeni et al., 2019a), ZnO (Sun et al., 2014; Yang and Cheng, 2018), SrTiO₃ (Yang and Cheng, 2017), and g-C₃N₄ (Bu et al., 2013; Jing et al., 2019b). Among them, TiO₂ has been widely used in photocatalytic degradation of organic pollutants (Zhuang et al., 2010), dye-sensitized solar cells (Roy et al., 2010; Sauvage et al., 2010), supercapacitors (Lu et al., 2012, 2013), photocatalytic sterilization (Foster et al., 2011; Yadav et al., 2014), photocatalytic hydrogen production (Kho et al., 2010; Xiang et al., 2012), and in other fields since the photoelectric conversion effect was found by Fujishima and Honda (1972). In addition, TiO₂ is of low cost, easy preparation, and good stability, and can control the directional migration of photogenerated electrons through the regulation of one-dimensional (1D) nanorods and nanotubes. However, as a wide-bandgap semiconductor, TiO₂ can only response to ultraviolet light, which limits its applications in the areas of photocatalysis (Momeni, 2015; Momeni et al., 2015), PECCP and dye-sensitized solar cell. Therefore, the main methods to broaden the application of TiO₂ are to extend the light absorption of TiO₂ to visible region and to make it use solar energy effectively.

At present, the main methods for expanding the light absorption range of a semiconductor material are metallic elements doping (Kočí et al., 2010; Tian et al., 2012; Momeni et al., 2018; Momeni et al., 2019b), non-metallic elements doping (Zhang et al., 2011; Momeni et al., 2019c), introduction of defects (Dong et al., 2014; Tian et al., 2018), and compounding with other semiconductors possessing narrow bandgap (Yan and Yang, 2011; Zhang et al., 2012; Chen et al., 2014a; Sun et al., 2018). The metallic and non-metallic elements doping will introduce doping levels in the bandgap of a semiconductor, thereby narrowing the forbidden bandwidth of the semiconductor and broadening the response range of light. The introduction of defects will also reduce the forbidden bandwidth of the semiconductor through introducing the defect levels, thereby broadening the light response range of the semiconductor. The compounding with other semiconductors having narrow bandgap increases the light response capability. Meanwhile, the heterojunction system will form at the interface of different semiconductors, which can increase the separation efficiency of the photogenerated carriers. Therefore, more photogenerated carriers can be separated and participate in the chemical reactions. Zhang et al. (2012) prepared Bi₂S₃/Bi₂WO₆ composite photocatalyst by one-step hydrothermal method. The coupling of narrow-bandgap Bi₂S₃ extended the light absorption threshold of Bi₂WO₆ from 450 nm to 800 nm, and the heterojunction formed at the interface of Bi₂S₃/Bi₂WO₆ efficiently separated the photogenerated electrons and holes, which made the phenol degradation rate of Bi₂S₃/Bi₂WO₆ be 6.2 times higher than that of pure Bi₂WO₆. Gou et al. (2017) prepared zeolite-supported Ag₂O/TiO₂ composites with excellent photocatalytic degradation performance of norfloxacin by improved sol-gel method. The high photocatalytic performance is attributed to the enhanced visible light absorption and high charge separation efficiency because of the combination of the narrow-band gap semiconductor.

In addition to the abovementioned Bi₂S₃ and Ag₂O, Cu₂O, as a narrow-bandgap semiconductor (2– 2.2 eV), has good visible light absorption performance and is widely used in the field of photocatalysis. Meanwhile, Cu₂O is a p-type semiconductor (Yang et al., 2010; Wang et al., 2013). When the p-type Cu₂O contacts with an n-type semiconductor, a built-in electric field will be formed at the interface of these two semiconductors, accelerating the directional movements of the photogenerated electrons and holes. Therefore, the preparation of p-n heterojunction can effectively separate the photogenerated electronhole pairs (Zhao et al., 2016), leading to the promotion of the photoelectrochemical (PEC) properties and the photocatalytic performance of the composite photocatalysts. Although the Cu₂O/TiO₂ system has been studied in photocatalytic degradation of organic pollutants (Peng et al., 2014), CO₂ reduction (Li et al., 2015c; Aguirre et al., 2017), and photocatalytic hydrogen production (Yao et al., 2018), the studies of the Cu₂O/TiO₂ composite in the area of PECCP have never been reported.

In this paper, TiO₂ nanotube arrays were prepared by anodic oxidation, and Cu₂O was subsequently deposited on the surface of TiO₂ by constant current deposition technology to obtain the Cu₂O/TiO₂ p-n heterojunctions. The morphologies and crystal structures of the prepared Cu₂O/TiO₂ photoelectrodes were observed in scanning electron microscopy (SEM), high resolution transmission electron microscopy (HRTEM), and X-ray diffraction (XRD). The effect of Cu₂O deposition amount characterized by its deposition time on the PECCP of 316L stainless steel (SS) was investigated under visible light in 3.5 wt% NaCl solution. Furthermore, the mechanism of the PECCP of the Cu₂O/TiO₂ composites was studied by the photoinduced volt-ampere characteristic curves (i-V curves), photoluminescence spectroscopy (PL), and electrochemical impedance spectroscopy (EIS).

2 MATERIAL AND METHOD 2.1 Material

Ti foil (10 mm×10 mm×0.3 mm, purity>99.6%) was purchased from Aladdin Reagent Co., Ltd. Ethanol (C₂H₆O, 99.7%), acetone (C₃H₆O, 99.5%), ethylene glycol (C₂H₆O₂, 99.5%), ammonium fluoride (NH4F, 96%), copper sulfate (CuSO₄·5H₂O, 99%), sodium hydroxide (NaOH, 96%), sodium sulfate (Na₂SO₄, 99%), sodium chloride (NaCl, 99.5%), hydrochloric acid (HCl, 36%–38%), and lactic acid (C₃H₆O₂, 85%) were purchased from China Pharmaceutical Group Reagents Co., Ltd. All reagents were used directly without further purification.

2.2 Preparation of TiO₂ nanotubes

First, the Ti foils were ultrasonically cleaned in detergent to remove the oil stains on surfaces. Second, they were ultrasonically degreased in a mixture of ethanol, acetone, and deionized water at volume ratio of 1꞉1꞉1 for 10 min. Finally, they were washed with distilled water and dried naturally. The TiO₂ nanotubes were prepared by a two-step anodic oxidation and the specific steps are described as follows: Firstly, a twoelectrode system (Ti foil as anode and Pt as cathode) was used for the first anodic oxidation. The electrolyte used for anodic oxidation was ethylene glycol containing 0.5 wt% ammonium fluoride and 2.5 wt% deionized water, and the potential and time of anodic oxidation were 60 V and 1 h, respectively. The Ti foil after the first anodic oxidation was ultrasonically vibrated for 10 min in 10 wt% hydrochloric acid solutions to remove the superfluous products on the surface of Ti foil, and the Ti foil with smooth surface was obtained. Afterwards, a second anodic oxidation was performed for 1 h on the Ti foil, after which the surface sediments were washed with deionized water. Finally, the TiO₂ nanotube arrays were obtained by annealing the Ti foil with two-step anodic oxidation in air at 450℃ for 3 h at a heating rate of 5℃/min, and the obtained sample was marked as TiO₂.

2.3 Preparation of Cu₂O

Cu₂O was prepared on Ti foil by constant current electrodeposition method with a conventional threeelectrode system. The Ti foil worked as the working electrode, a Pt foil acted as the counter electrode and the Ag/AgCl electrode was the reference electrode. The electrodeposition process was carried out in copper lactate basic aqueous solution at 45℃ water bath with the constant current density of 0.05 mA/cm² and the deposition time of 60 min. After electrodeposition, the Ti foil was cleaned several times with deionized water and ethanol, and Cu₂O was obtained by drying in a vacuum drying chamber at 60℃ for 4 h. The copper lactate basic aqueous solution was prepared according to the previous work (De Jongh et al., 1999; Inguanta et al., 2008). Firstly, 75 mL of 5 mol/L sodium hydroxide was slowly added to 25 mL lactic acid solution containing 23.5 g CuSO₄·5H₂O and stirred for 12 h. Finally, a dark blue copper sulfate electrolyte was obtained by adjusting the pH value of the mixed solution to 9 with 5 mol/L sodium hydroxide solution.

2.4 Preparation of Cu₂O/TiO₂ composite

The Cu₂O/TiO₂ composite was also prepared by a constant current electrodeposition method using a three-electrode cell system, which is the same as that of Cu₂O. The only difference from the preparation method of Cu₂O was that the working electrode was the prepared TiO₂ nanotube in the three-electrode cell system, and the deposition time were 5, 10, 20, and 30 min, respectively. It is worth noting that the TiO₂ nanotubes were immersed in the electrolyte for 2 h before the electrodeposition to make sure that the Cu2+ ions and the nanotubes were in good contact. After the electrodeposition, the prepared samples were washed several times with deionized water and ethanol, and dried in a vacuum oven at 60℃ for 4 h. The prepared samples were labeled as Cu₂O/TiO₂-5, Cu₂O/TiO₂-10, Cu₂O/TiO₂-20, and Cu₂O/TiO₂-30 according to the Cu₂O electrodeposition time.

2.5 Characterization of the prepared photoeletrodes

Scanning electron microscopy (SEM, JSM-6700F; JEOL, Tokyo, Japan) and field emission highresolution transmission electron microscopy (FEHRTEM, Tecnai G2 F20, FEI Company, USA) were used to characterize the microscopic morphologies of the prepared samples. The crystal structure of the prepared samples was analyzed by X-ray diffraction (XRD, D/MAX-2500/PC; Rigaku Co., Tokyo, Japan). The elemental composition and distribution of the samples were characterized by energy scatter spectroscopy (EDS, FEI Tecnai G20, FEI Company, USA). The bonding information of the prepared samples was analyzed by X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical Ltd., England). The optical absorption properties of the prepared samples were investigated by UV-visible diffuse reflectance spectroscopy (U-41000; HITACHI, Tokyo, Japan). The photoluminescence intensity of the samples was tested using microscopic fluorescence spectroscopy (PL, Fluro Max-4, HORIBA Jobin Yvon, France).

2.6 PECCP performance measurements

The PECCP performance was characterized by the photoinduced open circuit potential (OCP) and photoinduced current densitites with time (i-t curves) of the photoelectrode coupled with 316L SS under visible light illumination. OCP and i-t curves were measured by a CHI660D electrochemical workstation, and the tests were carried out in 3.5 wt% NaCl solution. The light source used in the test was a 300- W xenon lamp. A 420-nm filter was used to filter out the ultraviolet light below 420 nm, and the intensity of the light was adjusted to 100 mW/cm² using an optical power meter. During the OCP test, the intermittent light illumination was performed every 50 s for 3 cycles. The bias of i-t tests was 0 V and the intermittent light illumination was also applied in every 50 s for 3 cycles.

2.7 PEC and electrochemical performance measurements

All of the PEC and electrochemical performance measurements were performed in 0.1 mol/L Na₂SO₄ using CHI660D electrochemical workstation and three-electrode cell system. The photoelectrode was used as the working electrode, a Pt foil was acted as the counter electrode, and the Ag/AgCl electrode was used as the reference electrode. The light source used in the test was a 300-W xenon lamp. A 420-nm filter was used to filter out the ultraviolet light below 420 nm, and the intensity of the light was adjusted to 100 mW/cm² using an optical power meter. The voltampere characteristic curves (i-V curves) were measured from -0.7 V to 1 V with a scan rate of 20 mV/s, and the gap between the switching on and turning off the light was 2 s. Electrochemical impedance spectroscopy (EIS) tests were performed in the dark at OCP over the frequency range between 10⁵ and 10^-1 Hz, with an AC voltage magnitude of 5 mV, using 12 points/decade.

3 RESULT AND DISCUSSION 3.1 Analyses of the structures and morphologies of the photoeletrodes

XRD was used to characterize the crystal structure of the prepared photocatalysts (Fig. 1). The diffraction peaks at 35.09°, 38.42°, 40.17°, 53.00°, 62.95°, and 70.66° correspond to the (100), (002), (101), (200), (102), (110), and (103) crystal faces of Ti, respectively, which are consistent with the hexagonal phase of Ti (PDF #44-1294). The diffraction peaks at 25.28°, 37.80°, 48.05°, 53.89°, and 55.06° were assigned to the (101), (004), (200), (105), and (211) crystal planes of TiO₂, respectively, which are agreed with the XRD pattern of anatase TiO₂ (PDF#21-1271). Compared with the XRD pattern of pure TiO₂, the intensities and the positions of the diffraction peaks of Cu₂O/TiO₂-X composites did not change, indicating that the deposition of Cu₂O did not change the crystal structure of TiO₂. No diffraction peaks of Cu₂O were found in the XRD patterns of the prepared series Cu₂O/TiO₂-X composites, which could be attributed to the less deposition amount, smaller particle size, higher dispersion, and low crystallinity of Cu₂O (Chen et al., 2014b).

The micromorphologies of the prepared photocatalysts were observed using SEM (Fig. 2). As shown in Fig. 2a, TiO₂ prepared by the anodic oxidation method exhibited the morphology of nanotube array. The nanotube walls were thin. The pore sizes were not uniform, and the average diameters were approximately 90 nm. Figure 2b shows the SEM image of Cu₂O/TiO₂-5 composite, whose Cu₂O deposition time was 5 min. By comparing Fig. 2b with 2a, one can see a small amount of Cu₂O nanoparticles on the surface of TiO₂, and the thickness of the wall of the nanotube was slightly thicker than that of TiO₂. With the prolongation of Cu₂O deposition time to 10 min, the number of the deposited Cu₂O nanoparticles on the surface of Cu₂O/TiO₂-10 composite gradually increased (Fig. 2c). Prolonging the Cu₂O deposition time to 20 min, the surface of TiO₂ was evenly covered with the Cu₂O nanoparticles and no agglomeration of Cu₂O nanoparticles was observed (Fig. 2d). When the Cu₂O deposition time reached 30 min, the number of Cu₂O nanoparticles deposited on the surface of TiO₂ was further increased (Fig. 2e). In addition to being distributed on the tube wall of TiO₂, a part of Cu₂O nanoparticles were attached to the nozzle of the TiO₂ nanotube. Excessive deposition of Cu₂O nanoparticles might be detrimental to the transfer of the photogenerated electrons during PECCP. Figure 2f is the EDS results of Cu₂O/TiO₂- 20, and C, Ti, O and Cu elements were observed. Among these elements, Ti, O, and Cu came from Cu₂O/TiO₂-20 itself, while C was from the conductive adhesives used in the EDS test. The EDS results are consistent with the elemental composition of the Cu₂O/TiO₂ composite. The table in Fig. 2f is the elemental composition of Cu₂O/TiO₂-20. Cu₂O accounts for approximately 1.74% (mole ratio) of the Cu₂O/TiO₂-20. Figure 2g shows the SEM image and the corresponding element mapping of Cu₂O/TiO₂-20 composite. Ti, O, and Cu were uniformly distributed in Cu₂O/TiO₂-20, indicating that Cu₂O was uniformly deposited on the surface of TiO₂.

The micromorphology of Cu₂O/TiO₂-20 was further characterized by HRTEM (Fig. 3). Figure 3a shows the HRTEM image of Cu₂O/TiO₂-20 under low magnification. The diameters of the TiO₂ nanotubes were not uniform, and the average diameter was approximately 90 nm, which are consistent with the SEM results. Figure 3b shows the HRTEM image of Cu₂O/TiO₂-20 under high magnification. Clear lattice fringes and heterojunction interface could be observed. The lattice spacing of 0.352 nm corresponds to the (101) crystal plane of TiO₂, which agreed with the XRD result. The lattice spacing of 0.246 nm corresponds to the (111) crystal plane of Cu₂O, which is consistent with the reports from literature (Wang et al., 2013; Li et al., 2019). Figure 3c is the HRTEM image of the cross-section view of Cu₂O/TiO₂-20 under low magnification and Fig. 3d is that under the high magnification. The lattices spacing of 0.213 nm and 0.310 nm correspond to the (200) and (110) crystal planes (Sui et al., 2010; Chen et al., 2014a) of Cu₂O, respectively. The lattice spacing of 0.352 nm corresponds to the (101) crystal plane of TiO₂. This indicates that Cu₂O nanoparticles were also deposited on the wall of the TiO₂ nanotubes in addition to the surface of the TiO₂ nanotubes. The HRTEM results shown in Fig. 3 confirm the successful synthesis of the Cu₂O/TiO₂ heterojunction composite.

3.2 Chemical composition and optical absorption properties of the photoeletrodes

XPS was further used to investigate the chemical states of Ti and Cu elements of the prepared TiO₂ and Cu₂O/TiO₂-20 (Fig. 4). The total survey spectrum in Fig. 4a shows that Ti, O, and C were in the prepared TiO₂, while Ti, Cu, O, and C were in the prepared Cu₂O/TiO₂-20. C was derived from the calibration C during the XPS test. In addition to the Ti, O, and C elements in the XPS spectrum of Cu₂O/TiO₂-20, Cu element was also presented. Figure 4b shows the Ti 2p XPS core level spectra of TiO₂ and Cu₂O/TiO₂-20. The binding energies at 458.3 eV and 463.9 eV were observed, which correspond to Ti 2p_3/2 and Ti 2p_1/2 (Li et al., 2015b), respectively. However, for Cu₂O/TiO₂- 20, the binding energy peaks assigned to Ti 2p_3/2 and Ti 2p_1/2 were detected at 458.5 eV and 464.1 eV, which were shifted to higher binding energies compared with those of TiO₂. This could be due to strong interaction between Cu₂O and TiO₂ (Wang et al., 2017). Figure 4c shows the Cu 2p XPS core level spectrum of Cu₂O/TiO₂-20. The binding energy peaks at 931.9 eV and 951.6 eV were detected, which corresponded to the characteristic binding energies of Cu 2p_3/2 and Cu 2p_1/2 in the Cu₂O crystal (Xi et al., 2014). Figure 4d shows the XPS results for Cu LMM of Cu₂O/TiO₂-20, and the binding energy peak of 570 eV corresponding to Cu⁺ in Cu₂O was observed (Zhu et al., 2013), which is consistent with the previous report (Zhu et al., 2013). The XPS results shown in Fig. 4 confirm the successful preparation of Cu₂O/TiO₂ heterojunction again.

To understand the effect of Cu₂O deposition on the light absorption properties of TiO₂, the UV-visible diffuse reflectance spectra of the prepared pure TiO₂, Cu₂O/TiO₂-5, Cu₂O/TiO₂-10, Cu₂O/TiO₂-20, Cu₂O/ TiO₂-30, and pure Cu₂O were tested (Fig. 5a). The light absorption threshold of TiO₂ nanotubes was approximately 387 nm, which is consistent with the previous report (Sun et al., 2018). In addition, the absorption of TiO₂ at 400–800 nm was caused by the Ti foil, which appeared blue (Li et al., 2015a). The absorption threshold of pure Cu₂O obtained by electrochemical deposition was about 563 nm. The light absorption bandedges of Cu₂O/TiO₂-5, Cu₂O/ TiO₂-10, Cu₂O/TiO₂-20, and Cu₂O/TiO₂-30 did not show significant shift, indicating that the deposition of Cu₂O did not change the light absorption performance of TiO₂. This might be due to the small amount of Cu₂O deposited on the surface of TiO₂ and on the wall of the nanotubes. The Kubelka-Munk functions of TiO₂ and Cu₂O transformed from Fig. 5a are shown in Fig. 5b, in which the intersection with the X-axis was the forbidden bandwidth of the semiconductor materials (Jing et al., 2019a). As shown in Fig. 5b, the forbidden bandwidths of TiO₂ and Cu₂O were 3.2 and 2.2 eV, respectively. The XPS valence band spectrum was used to characterize the valence band position of the prepared materials. Figure 5c & dshow the XPS valence band (VB) spectra of TiO₂ and Cu₂O, respectively. The intercept of the slope with the horizontal axis in the XPS valence band spectrum is the VB value. As shown in Fig. 5c & d, the VB values of TiO₂ and Cu₂O were 2.65 and 1.56 eV, respectively. According to the following formula:

where E_g is the forbidden bandwidth, E_VB and E_CB are the potentials of the VB and the conduction band (CB) relative to the standard hydrogen electrode (SHE), respectively. Based on the results shown in Fig. 5, the E_CB of TiO₂ and Cu₂O was calculated as -0.55 and -0.64 eV (vs. SHE), respectively.

3.3 PEC and PECCP performances of the prepared photoeletrodes

The PEC performance of the prepared photoelectrodes was studied using a photoinduced volt–ampere characteristic curve (i–V curve) and a photoinduced current density curve (i–t curve), which were tested in 0.1 mol/L Na₂SO₄ solution under visible light illumination (Fig. 6). Figure 6a depicts the i-t curves of the prepared photoelectrodes, which were obtained at an applied bias potential of 0 V. The currents of TiO₂, Cu₂O/TiO₂-5, Cu₂O/TiO₂-10, Cu₂O/ TiO₂-20, and Cu₂O/TiO₂-30 showed positive shifts when visible light was switched on, indicating that they all exhibit the characteristics of n-type semiconductors (Jing et al., 2019b). However, Cu₂O exhibited a negative photocurrent under visible light illumination, indicating that Cu₂O prepared at this condition was a p-type semiconductor (Jing et al., 2019b). The pure TiO₂ photoelectrode exhibited a small photocurrent density due to its narrower light absorption range. After Cu₂O deposition, the photoinduced current densities of the Cu₂O/TiO₂-X photoelectrodes significantly increased. This could be due to the formation of Cu₂O/TiO₂ heterojunctions, which accelerated the separation of the photoinduced electron-hole pairs. Thus, the Cu₂O/TiO₂-X photoelectrodes had higher photochemical conversion performance. The photoinduced current densities increased with the prolongation of the Cu₂O deposition time. When the Cu₂O deposition time was 20 min, the Cu₂O/TiO₂-20 had the largest photoinduced current density. With the further prolongation of the Cu₂O deposition time to 30 min, the photoinduced current densities of Cu₂O/TiO₂-30 decreased again. When less electrodeposition time was used, only a small amount of Cu₂O deposited and it could not provide sufficient photogenerated electron-hole pairs under visible light. When the Cu₂O deposition time was long, a thicker layer of Cu₂O was deposited on the surface of TiO₂. The over-thick Cu₂O coating could remarkably reduce the specific surface area of the photosensitive surface, and the excessive Cu₂O could inhibit the effective combination of Cu₂O and TiO₂; thus the photogenerated electron-hole pairs could not be effectively separated, resulting in a worse photoelectric conversion performance.

Figure 6b shows the i-V curves of the prepared photoelectrodes under intermittent visible light. Pure TiO₂ had a small photocurrent density under visible light illumination, while the photocurrent density of Cu₂O/TiO₂-X increased first and then decreased with the increase of Cu₂O deposition time. Cu₂O/TiO₂-20 had the highest photoinduced current density, which is similar to the change rules of i-t curves shown in Fig. 6a. Cu₂O would be oxidized to CuO under more positive voltage, therefore, the i-V curve of Cu₂O was scanned to 0 V only. In the whole scanning voltage range, the current density of Cu₂O was negative (Fig. 6b).

The PECCP properties of the prepared photoelectrodes were characterized by the photoinduced OCP variations after coupling with 316L SS. Figure 7a & b are the results of the photoinduced OCP variations of the TiO₂, Cu₂O/TiO₂- 5, Cu₂O/TiO₂-10, Cu₂O/TiO₂-20, Cu₂O/TiO₂-30, and Cu₂O photoelectrodes coupled with 316L SS, which were measured in 3.5 wt% NaCl solution under intermittent visible light. As shown in Fig. 7a, the photoinduced OCP of the TiO₂ nanotubes showed only a small negative shift under visible light because of the narrow light absorption range of TiO₂. However, as shown in Fig. 7b, Cu₂O exhibited a positive shift of the OCP under visible light illumination, indicating that the electrons transfer from 316L SS to Cu₂O and Cu₂O alone could not provide PECCP for the coupled 316L SS. However, for Cu₂O/TiO₂-X, the OCP rapidly shifted to negative direction when the visible light was on, indicating that the photoinduced electrons generated on the CB of Cu₂O could be rapidly transferred to the CB of TiO₂ under visible light and be further to the coupled 316L SS to provide PECCP. The magnitude of the negative shift of the potential can be used to characterize the PECCP performance of the photoelectrode (Jing et al., 2017). The larger the magnitude of the negative shift of the potential is, the better the PECCP performance is. The photoinduced OCP variation of Cu₂O/TiO₂-5, Cu₂O/ TiO₂-10, Cu₂O/TiO₂-20, and Cu₂O/TiO₂-30 were 40, 48, 60, and 50 mV, respectively. The Cu₂O/TiO₂-20 photoelectrode has the largest photoinduced OCP variation under visible light, indicating that 20 min was the optimal Cu₂O deposition time and the Cu₂O/ TiO₂-20 photoelectrode had the best PECCP performance for 316L SS.

In addition to the photoinduced potential drop, the PECCP performance of the prepared photoelectrodes could also be characterized by the photoinduced current densities. Figure 8a & b show the variations of the galvanic current densities of the prepared TiO₂, Cu₂O/TiO₂-5, Cu₂O/TiO₂-10, Cu₂O/TiO₂-20, Cu₂O/ TiO₂-30, and Cu₂O photoelectrodes coupled with 316L SS in 3.5 wt% NaCl solution under intermittent visible light. A positive photoinduced current density indicates that the photogenerated electrons could be transferred to the coupled 316L SS (Jing et al., 2019b), resulting in the PECCP of the coupled 316L SS. The photoinduced current densities of the TiO₂ and Cu₂O photoelectrodes were positive and negative, respectively, indicating that TiO₂ can provide PECCP while Cu₂O cannot provide PECCP for 316L SS, which is consistent with the results of the photoinduced OCP variations shown in Fig. 7. Meanwhile, comparing the photoinduced current densities of Cu₂O/TiO₂-X composites with different Cu₂O deposition times, Cu₂O/TiO₂-20 had the highest photoinduced current density, indicating that it had the best PECCP performance. These results are also consistent with those of photoinduced OCP variations shown in Fig. 7.

The ideal PECCP material needs not only good PECCP performance, but also good stability. Figure 9 shows that the photoinduced potential drop of the Cu₂O/TiO₂-20 photoelectrode coupled with 316L SS in 3.5 wt% NaCl solution under long-term visible light. As shown in Fig. 9, after several cycles of intermittent visible light illumination, the Cu₂O/TiO₂- 20 photoelectrode coupled with 316L SS electrode was illuminated by visible light for 3.5 h. The potential slowly shifted to positive direction under the visible light. After 3.5 hours of visible light illumination, the positive displacement of the potential was lowered below 20 mV, and the Cu₂O/TiO₂-20 photoelectrode remained very good PECCP performance. When the light was switched off after a long-term light illumination, the potential rose immediately. When the light was switched on again, the potential decreased immediately, showing good PECCP stability.

3.4 Analyses of the mechanism for improving PEC and PECCP performance

To understand the role of deposited Cu₂O on improving the PECCP performance of TiO₂, the photoluminescence spectroscopy of TiO₂ and Cu₂O/ TiO₂-20 was used to analyze the separation efficiency of the photogenerated electron-hole pairs. The electrons on the VB of the semiconductor material were excited to the CB under light illumination, while leaving the photogenerated holes on the VB. Some of the photogenerated electrons and holes would participate in the PEC reactions; however, most of them would recombine and release energy, which could be detected in photoluminescence spectroscopy. The stronger the peak intensity of the photoluminescence spectrum was, the easier the photogenerated electrons and holes recombine could be. Figure 10a shows the photoluminescence spectra of TiO₂ and Cu₂O/TiO₂-20. The peak intensity of photoluminescence of Cu₂O/TiO₂-20 was much lower than that of pure TiO₂, indicating that Cu₂O/TiO₂-20 had a much higher separation efficiency of the photogenerated electrons and holes than pure TiO₂, which confirms that the Cu₂O/TiO₂ p-n heterojunctions can effectively separate the photogenerated electronhole pairs.

The migration ability of the photogenerated carriers of the photoelectrodes could be characterized by EIS. Figure 10b shows the EIS results of TiO₂ and Cu₂O/TiO₂-20 measured in the dark at the OCP. The smaller the radius of the EIS impedance arc is, the larger the migration rates of the photogenerated carriers are. As shown in Fig. 10b, the radius of the impedance arc of Cu₂O/TiO₂-20 was much smaller than that of pure TiO₂, indicating that Cu₂O/TiO₂-20 had a larger carrier mobility than pure TiO₂. To understand why Cu₂O/TiO₂-20 had higher carrier mobility in detail, the EIS data of TiO₂ and Cu₂O/TiO₂-20 were further fitted by Zismwin software. The insert in Fig. 10b was the corresponding equivalent circuit used for fitting the EIS data. The measured data were fitted very well. Table 1 shows the fitted parameters and the errors of the fitting results of the EIS data. The film resistance and electron transfer resistance of Cu₂O/TiO₂-20 were greatly reduced compared with those of pure TiO₂, indicating that the deposited Cu₂O reduced the film resistance of TiO₂ and the transfer barrier of interface charge, thus accelerated the electron transfer process and led to a better PEC and PECCP performance.

Figure 11 is the proposed mechanism for improving the PEC and PECCP performance by depositing Cu₂O on the surface of TiO₂. TiO₂ is an n-type semiconductor and Cu₂O is a p-type semiconductor. The forbidden bandwidth, CB and VB positions of TiO₂ are 3.2 eV, -0.55 eV and 2.65 eV, respectively, and those of Cu₂O are 2.2 eV, -0.64 eV and 1.56 eV, respectively, which were verified by the results of UV-vis and XPS VB. Under visible light, the electrons on the VB of Cu₂O would be excited to the CB of Cu₂O, leaving the photogenerated hole on its VB. Since the CB potential of Cu₂O is more negative than that of TiO₂, the photogenerated electrons on the CB of Cu₂O would be transferred to that of TiO₂ under visible light, and be further transferred to 316L SS. Therefore, 316L SS was in an electron-rich state, and it was cathodically polarized and the cathodic protection of 316L SS was obtained. Meanwhile, the built-in electric field pointed from TiO₂ to Cu₂O at the interface of Cu₂O/ TiO₂ was formed by n-type TiO₂ and p-type Cu₂O, and this electric field could accelerate the transfer of the photogenerated electrons from the CB of Cu₂O to TiO₂, which was more favorable to the transfer of the photogenerated electrons to 316L SS. Thus, Cu₂O/ TiO₂-20 retained much better PEC and PECCP performance.

4 CONCLUSION

In this paper, TiO₂ nanotubes were prepared by anodic oxidation method. Cu₂O/TiO₂-X composites were prepared by galvanostatic deposition method, and the deposition amount of Cu₂O nanoparticles on the surface of TiO₂ nanotubes was controlled by the deposition time. The deposition of Cu₂O nanoparticles could not only increase the absorption capability of visible light, but also form heterojunctions with TiO₂ to improve the separation efficiency of the photogenerated charge carriers. The Cu₂O/TiO₂-20 photoelectrode had the best PEC and PECCP performance, which should be attributed to the establishment of the built-in electric field at the interface between the p-type Cu₂O and n-type TiO₂. The built-in electric field accelerated the separation of the photogenerated carriers and pushed more photogenerated electrons to be transferred to 316L SS for PECCP. In addition, the deposition of Cu₂O reduced the film resistance of TiO₂ and the transfer barrier of interface charge, thus accelerated the electron transfer process and led to a better PEC and PECCP performance. In summary, Cu₂O/TiO₂ is a feasible visible light-driven PECCP composite material, which provides a new idea for the deposition of Cu₂O nanoparticles onto nanotubes or other highly oriented n-type semiconductors to increase the number of photoactive sites, and to further enhance the PEC or PECCP performance under visible light.

5 DATA AVAILABILITY STATEMENT

The experimental and numerical data that support the findings of this study are available from the corresponding author on request.