Study on the Photocathodic Protection of Q235 Steel by CdIn 2 S 4 Sensitized TiO 2 Composite in Splash Zone

: In this work, the photo-catalytic activity of TiO 2 is considerably enhanced via sensitization with CdIn 2 S 4 , and its application for protecting Q235 from corrosion in splash zones is examined. TiO 2 nanotube arrays (NTAs) are prepared on a ﬂat Ti substrate via two-step anodization. CdIn2S4 is deposited on the surface of TiO 2 NTAs by hydrothermal reaction. TiO 2 NTAs with enormous speciﬁc surface areas and large-diameter hollow nanostructures are found to beneﬁt the immobilization of CdIn 2 S 4 . As a narrow band gap semiconductor, CdIn 2 S 4 is able to extend the light absorption range of TiO 2 , and the construction of an n–n type hetero-junction accelerates the separation of carriers. Strong solar light, which accelerates the corrosion of Q235 in the splash zone area, is converted into the necessary condition for protecting Q235 from corrosion. In this work, TiO 2 is sensitized with MoS 2 microspheres (MoS 2 / TiO 2 nanocomposites), which were prepared on a ﬂat Ti substrate via a two-step anodization and hydrothermal method, sequentially.


Introduction
With the increasing development and utilization of marine resources, metallic materials have been extensively used in ports, cross-sea bridges, and offshore oil platforms over the past decades [1]. As low-carbon steel, Q235 steel (with a yield strength of 235 MP) displays ultra-high cost performance and well strength, and is low-cost, which makes it suitable for bridges, platforms, and other engineering structures [2][3][4]. However, corrosion poses a challenge to the application of Q235 in the marine environment. Marine corrosion areas are generally divided into mud, immersion, tidal, splash, and atmospheric zones [5][6][7][8][9]. Among these areas, the splash zone experiences the most serious level of marine corrosion due to its alternating dry and wet seasons, abundant amount of sunshine, high salt concentration, and other factors [10][11][12]. In particular, the corrosion rate of carbon steel could be increased by the corrosion product, which contains a p-type semiconductor, under strong solar light. Specifically, the corrosion rate in the splash zone is approximately 3 to 10 times higher than that recorded in other marine corrosion areas. Therefore, the protection of steel in the splash zone must be examined.
Photocathodic protection, as a new cathodic protection method, has been widely examined by scholars due to its lack of power and anode material consumption and theoretically permanent validity [13][14][15][16]. Photocathodic protection is based on the transfer of photo-induced electrons from semiconductors to the surface of the protected metal under light to replace the loss of electrons in the metal. The requirements for effective photocathodic protection are that (1) the semiconductor should be n-type, (2) the semiconductor should possess high photo-induced carrier separation efficiency, and (3) the semiconductor should have a more negative Fermi level than the self-corrosive potential of the protected metal. The performance of photocathodic protection can be improved by sufficient solar light. Hence, this approach is particularly effective for corrosion protection in splash zones with an abundant amount of solar light.
TiO 2 is the most important semiconductor in this area [17,18]. TiO 2 has been widely used in photocracking water and photodegrading organic pollutants and solar cells due to its excellent photoelectric properties and costly performance [19][20][21]. As a 1D semiconductor material, TiO 2 nanotube arrays (NTAs) have been extensively utilized in photocathodic protection due to their excellent electronic transmission properties [22,23]. However, TiO 2 also demonstrates some defects. For example, the large band gap of TiO 2 can limit its absorption of ultraviolet light below 380 nm, and its high carrier recombination rate leads to a low photon-to-electron conversion efficiency. In addition, the conduction band of TiO 2 is insufficiently negative for Q235. Consequently, electrons from Q235 can be transferred to TiO 2 to accelerate the corrosion. A strategy has been proposed to improve the light absorption of TiO 2 , accelerate the separation of photo-induced carriers, and make the Fermi level negative. Loading a narrow band semiconductor with a negative conductance band on the surface of TiO 2 NTAs and constructing heterojunction simultaneously could be an effective method.
As a narrow band gap (2.1 eV) semiconductor material, CdIn 2 S 4 is an ideal material for improving the visible light response of semiconductor composites [24][25][26][27][28][29]. The conduction band (−0.76 eV) of CdIn 2 S 4 is more negative than that of TiO 2 (−0.29 V versus NHE), which could cause the negative shift of the Fermi level of TiO 2 . As a ternary semiconductor, CdIn 2 S 4 is much more stable due to the presence of In 3+ along with Cd 2+ [27,30]. It effectively avoids the occurrence of photo-corrosion due to photo-induced holes and •OH in CdS-based systems. In addition, similarly to TiO 2 , CdIn 2 S 4 is an n-type semiconductor. The recombination of photo-induced carriers could be inhibited due to the construction of an n-n heterogeneous type when CdIn 2 S 4 is coupled with TiO 2 . Hence, the proposed loading of CdIn 2 S 4 on the surface of TiO 2 may improve the photoelectrochemical properties of TiO 2 , and Q235 could be protected by this composite via photocathodic protection. Na 2 S/Na 2 SO 3 would be used as hole trapping agents which could rapidly absorb photo-induced holes and •OH on the surface of the composite. The hole trapping agents could further prevent the occurrence of photo-corrosion. By contrast, the alkaline environment provided by Na 2 S/Na 2 SO 3 could negatively shift the Fermi level of the synthetic materials, which is good for the protection of Q235.
In this paper, photocatalytic active CdIn 2 S 4 /TiO 2 NTAs are successfully prepared by using the two-step anodization process and facile hydrothermal method. The photocathodic protection of CdIn 2 S 4 /TiO 2 NTAs for Q235 is examined. XRD, SEM, X-ray photoelectron spectroscopy (XPS), and TEM are used to confirm the success of material preparation. UV-VIS diffuse reflectance and photoluminescence spectra analyses are conducted to characterize the samples. Photochemical experiments are used to investigate the photocathodic protection properties of Q235.

Results and Discussion
A two-step anodization process was performed to prepare TiO 2 NTAs with large tube diameters. A facile hydrothermal method was also used to load CdIn 2 S 4 nanoparticles on the surface of TiO 2 NTAs.  Figure 1, all diffraction peaks of the standard card (ICDD 00-027-0060) can be found in the curve of pure CdIn 2 S 4 , and the intensity matches well. This finding indicates the preparation of highly crystalline and pure CdIn 2 S 4 . For the pristine TiO 2 NTAs, the main peaks appeared at 25.4 • , 37.0 • , 37.8 • , 48.1 • , 53.9 • , 55.2 • , 70.5 • , and 76.2 • ; these can be assigned to the (101), (103), (004), (200), (105), (211), (220), and (301) crystalline planes of the anatase structure, respectively (ICDD 01-89-4921). The diffraction peak of Ti (ICDD 01-089-3725) is stronger and sharper than that of TiO 2 , suggesting that TiO 2 forms only a thin layer on the Ti sheet. Simultaneously, the weak peaks of the monomer are found in the curve of the synthetic CdIn 2 S 4 /TiO 2 NTA composites. This finding indicates that a small amount of CdIn 2 S 4 particles had been successfully loaded onto the surface of the TiO 2 NTAs.  Figure 1, all diffraction peaks of the standard card (ICDD 00-027-0060) can be found in the curve of pure CdIn2S4, and the intensity matches well. This finding indicates the preparation of highly crystalline and pure CdIn2S4. For the pristine TiO2 NTAs, the main peaks appeared at 25.4°, 37.0°, 37.8°, 48.1°, 53.9°, 55.2°, 70.5°, and 76.2°; these can be assigned to the (101), (103), (004), (200), (105), (211), (220), and (301) crystalline planes of the anatase structure, respectively (ICDD 01-89-4921). The diffraction peak of Ti (ICDD 01-089-3725) is stronger and sharper than that of TiO2, suggesting that TiO2 forms only a thin layer on the Ti sheet. Simultaneously, the weak peaks of the monomer are found in the curve of the synthetic CdIn2S4/TiO2 NTA composites. This finding indicates that a small amount of CdIn2S4 particles had been successfully loaded onto the surface of the TiO2 NTAs.

SEM Analysis
A morphology image of the prepared TiO2 NTAs is shown in Figure 2a. The internal diameter of the tubes was approximately 150 nm, and their wall thickness was 2 nm. These factors could lead to a large load of CdIn2S4 due to the appearance of load on the surface of the NTAs and the wall of tubes. This finding is due to the relatively large internal diameter, which facilitates the entry of the solution into the tubes. As shown in Figure 2b, the length of TiO2 NTAs was found to be 3.9 μm, and independent and ordered tubes were observed. Figure 2c shows the SEM image of the resultant CdIn2S4, which was supported on TiO2 NTAs. The synthetic CdIn2S4 shows a relatively symmetrical octahedral or truncated octahedral shape with a smooth surface, grain boundaries, and tip cones. The presence of CdIn2S4 on the surface and inner part of TiO2 NTAs indicates the successful entry of the electrolyte into the tubes. As shown in Figure 2d, the octahedral structure inside the tubes could also be observed in sections.

SEM Analysis
A morphology image of the prepared TiO 2 NTAs is shown in Figure 2a. The internal diameter of the tubes was approximately 150 nm, and their wall thickness was 2 nm. These factors could lead to a large load of CdIn 2 S 4 due to the appearance of load on the surface of the NTAs and the wall of tubes. This finding is due to the relatively large internal diameter, which facilitates the entry of the solution into the tubes. As shown in Figure 2b, the length of TiO 2 NTAs was found to be 3.9 µm, and independent and ordered tubes were observed. Figure 2c shows the SEM image of the resultant CdIn 2 S 4 , which was supported on TiO 2 NTAs. The synthetic CdIn 2 S 4 shows a relatively symmetrical octahedral or truncated octahedral shape with a smooth surface, grain boundaries, and tip cones. The presence of CdIn 2 S 4 on the surface and inner part of TiO 2 NTAs indicates the successful entry of the electrolyte into the tubes. As shown in Figure 2d, the octahedral structure inside the tubes could also be observed in sections.

XPS Analysis
A total survey spectrum of the synthesized composite is shown in Figure 3a. The peaks of S, In, Cd, O, and Ti elements can be observed from the survey spectrum. The existence of S (II) in the fresh CdIn2S4/TiO2 NTA composite is indicated by the closely spaced spin-orbit components (Δ = 1.16 eV and intensity ratio = 0.511) at approximately 161.3 eV ( Figure 3b). Meanwhile, the In (IV) 3d spectrum exhibits two broad peaks, namely, 3d5/2 and 3d3/2 (resulting from the spin-orbit splitting) without any loss features in the high binding energy side. The two peaks are located at 444.6 and 452.2 eV ( Figure  3c) and can be assigned to CdIn2S4. For Cd, two broad peaks appear near 405 eV with well-separated spin-orbit components (Δ = 6.8 eV) on the spectrum. The loss features, which cannot be observed in the high binding energy side of the 3d3/2 spin-orbit component, indicate that the Cd was trivalent. The O (II) and Ti (IV) in TiO2 are presented in Figure 3e,f, respectively. Figure 4 indicates the presence of hydroxyl groups on the surface of TiO2.
In addition, the valence band of the monomers was determined by two monomer XPS testing. In Figure 3g,h, the straight line near 0 eV in the obtained graph was extrapolated to intersect with the horizontal extension line. The intersection point (1.39 eV) denotes the valence band (Ev) of the synthetic material, and the Ev of the synthetic TiO2 was found to be 2.81 eV.

XPS Analysis
A total survey spectrum of the synthesized composite is shown in Figure 3a. The peaks of S, In, Cd, O, and Ti elements can be observed from the survey spectrum. The existence of S (II) in the fresh CdIn 2 S 4 /TiO 2 NTA composite is indicated by the closely spaced spin-orbit components (∆ = 1.16 eV and intensity ratio = 0.511) at approximately 161.3 eV ( Figure 3b). Meanwhile, the In (IV) 3d spectrum exhibits two broad peaks, namely, 3d 5/2 and 3d 3/2 (resulting from the spin-orbit splitting) without any loss features in the high binding energy side. The two peaks are located at 444.6 and 452.2 eV ( Figure 3c) and can be assigned to CdIn 2 S 4 . For Cd, two broad peaks appear near 405 eV with well-separated spin-orbit components (∆ = 6.8 eV) on the spectrum. The loss features, which cannot be observed in the high binding energy side of the 3d 3/2 spin-orbit component, indicate that the Cd was trivalent. The O (II) and Ti (IV) in TiO 2 are presented in Figure 3e,f, respectively. Figure 4 indicates the presence of hydroxyl groups on the surface of TiO 2 .
In addition, the valence band of the monomers was determined by two monomer XPS testing. In Figure 3g,h, the straight line near 0 eV in the obtained graph was extrapolated to intersect with the horizontal extension line. The intersection point (1.39 eV) denotes the valence band (Ev) of the synthetic material, and the Ev of the synthetic TiO 2 was found to be 2.81 eV.

TEM and HRTEM Analysis
A typical low-magnification Transmission electron microscopy (TEM) image of the synthesized CdIn2S4/TiO2 NTAs is shown in Figure 4a,b. A batch of nanotubes was clearly developed, forming orderly NTAs; some octahedrons existed on the walls of the TiO2 nanotubes, which agrees with the SEM results. High resolution transmission electron microscopy (HRTEM) was used to characterize the catalysts after the reaction, as shown in Figure 4c,d. The synthesized sample shows the characteristic 0.351 nm spacing of the (101) lattice plane of anatase TiO2. Lattice spacings of around 0.62 nm, which correspond to the distances of the (111) plane of CdIn2S4, were also observed. These findings indicate that the CdIn2S4 nanoparticles were bound to the TiO2 NTAs.

UV-Vis Diffuse Reflectance Spectra Analysis
As shown in Figure 5a, the TiO2 NTAs exhibited a high absorption intensity in the ultraviolet region; its absorption band edge was approximately 385 nm, and the corresponding band gap (Eg) was 3.22 eV. These characteristics are typical of anatase. Similarly, CdIn2S4 exhibited effective light absorption at wavelengths less than 590 nm, and the Eg of CdIn2S4 was 2.1 eV. The absorption of the composite material was much better than that of TiO2, and the absorption boundary shifted to 492 nm, thereby extending the absorption range of light. The apparent band gap of the corresponding composite material was 2.52 eV. The introduction of CdIn2S4 slightly narrowed the apparent band gap of TiO2, thereby allowing the semiconductor to absorb visible light with a large percentage of solar energy.
In combination with the valence band position of the composite material, the position of the conduction band (Ec) can be calculated using the following formula.
The Ec of the synthetic TiO2 was −0.41 eV, whereas that of the CdIn2S4 was −0.71 eV. These factors indicate that the prepared TiO2 NTAs and CdIn2S4 were as expected. Moreover, the synthetic composite showed excellent absorption performance of visible light and provided photocathodic protection for Q235.

TEM and HRTEM Analysis
A typical low-magnification Transmission electron microscopy (TEM) image of the synthesized CdIn 2 S 4 /TiO 2 NTAs is shown in Figure 4a,b. A batch of nanotubes was clearly developed, forming orderly NTAs; some octahedrons existed on the walls of the TiO 2 nanotubes, which agrees with the SEM results. High resolution transmission electron microscopy (HRTEM) was used to characterize the catalysts after the reaction, as shown in Figure 4c,d. The synthesized sample shows the characteristic 0.351 nm spacing of the (101) lattice plane of anatase TiO 2 . Lattice spacings of around 0.62 nm, which correspond to the distances of the (111) plane of CdIn 2 S 4 , were also observed. These findings indicate that the CdIn 2 S 4 nanoparticles were bound to the TiO 2 NTAs.

UV-Vis Diffuse Reflectance Spectra Analysis
As shown in Figure 5a, the TiO 2 NTAs exhibited a high absorption intensity in the ultraviolet region; its absorption band edge was approximately 385 nm, and the corresponding band gap (Eg) was 3.22 eV. These characteristics are typical of anatase. Similarly, CdIn 2 S 4 exhibited effective light absorption at wavelengths less than 590 nm, and the Eg of CdIn 2 S 4 was 2.1 eV. The absorption of the composite material was much better than that of TiO 2 , and the absorption boundary shifted to 492 nm, thereby extending the absorption range of light. The apparent band gap of the corresponding composite material was 2.52 eV. The introduction of CdIn 2 S 4 slightly narrowed the apparent band gap of TiO 2 , thereby allowing the semiconductor to absorb visible light with a large percentage of solar energy. 2.1.6. Photoluminescence (PL) Spectra Analysis PL spectra of CdIn2S4/TiO2 and TiO2 NTAs at the same excitation wavelength are shown in Figure 5b. The same shapes of the two curves indicate that the supported CdIn2S4 did not increase the recombination of photo-induced electrons and holes within the wavelength range of less than approximately 590 nm. The spectral intensity of the CdIn2S4/TiO2 NTAs composite was stronger than that of pure TiO2 NTAs. This finding that the amount of photo-induced electrons or holes for recombination increased in TiO2 NTAs is due to the supported CdIn2S4 nanoparticles. This increase is due to the transfer of photo-induced electrons from the conduction band of CdIn2S4 to that of TiO2. In the large wavelength range, the spectral intensity increased with different intensifications due to the recombination of photo-induced carriers in CdIn2S4.

Photochemical Analysis
Photoinduced volt-ampere characteristic curves under intermittent visible light (i-V curves) [31,32] were used to test the photoelectric chemical properties of the synthesized materials. Figure 6 shows the i-V curves of the samples. Negative photo-induced potential and positive photo-induced current density indicate that TiO2 is a typical n-type semiconductor. The photocurrent started to flow at −1.4 V (versus the saturated calomel electrode (SCE)) and the outset potential was negative compared with its Fermi level. This phenomenon was due to the negative shift of the Fermi level of TiO2 under alkaline conditions [20,33,34]. The i-V curves of the synthetic CdIn2S4/TiO2 NTA composite showed a higher photocurrent density than that of TiO2. This finding indicates that any composite other than TiO2 can easily inject electrons into Q235 in need of protection. In particular, at the corrosion potential of Q235 (from the Tafel curve), the photo-induced electrons were fully capable of transferring from CdIn2S4/TiO2 NTAs to Q235 due to the large positive photocurrent. This transfer is beneficial to the photocathodic protection of TiO2.  In combination with the valence band position of the composite material, the position of the conduction band (Ec) can be calculated using the following formula.
The Ec of the synthetic TiO 2 was −0.41 eV, whereas that of the CdIn 2 S 4 was −0.71 eV. These factors indicate that the prepared TiO 2 NTAs and CdIn 2 S 4 were as expected. Moreover, the synthetic composite showed excellent absorption performance of visible light and provided photocathodic protection for Q235.
2.1.6. Photoluminescence (PL) Spectra Analysis PL spectra of CdIn 2 S 4 /TiO 2 and TiO 2 NTAs at the same excitation wavelength are shown in Figure 5b. The same shapes of the two curves indicate that the supported CdIn 2 S 4 did not increase the recombination of photo-induced electrons and holes within the wavelength range of less than approximately 590 nm. The spectral intensity of the CdIn 2 S 4 /TiO 2 NTAs composite was stronger than that of pure TiO 2 NTAs. This finding that the amount of photo-induced electrons or holes for recombination increased in TiO 2 NTAs is due to the supported CdIn 2 S 4 nanoparticles. This increase is due to the transfer of photo-induced electrons from the conduction band of CdIn 2 S 4 to that of TiO 2 . In the large wavelength range, the spectral intensity increased with different intensifications due to the recombination of photo-induced carriers in CdIn 2 S 4 .

Photochemical Analysis
Photoinduced volt-ampere characteristic curves under intermittent visible light (i-V curves) [31,32] were used to test the photoelectric chemical properties of the synthesized materials. Figure 6 shows the i-V curves of the samples. Negative photo-induced potential and positive photo-induced current density indicate that TiO 2 is a typical n-type semiconductor. The photocurrent started to flow at −1.4 V (versus the saturated calomel electrode (SCE)) and the outset potential was negative compared with its Fermi level. This phenomenon was due to the negative shift of the Fermi level of TiO 2 under alkaline conditions [20,33,34]. The i-V curves of the synthetic CdIn 2 S 4 /TiO 2 NTA composite showed a higher photocurrent density than that of TiO 2 . This finding indicates that any composite other than TiO 2 can easily inject electrons into Q235 in need of protection. In particular, at the corrosion potential of Q235 (from the Tafel curve), the photo-induced electrons were fully capable of transferring from CdIn 2 S 4 /TiO 2 NTAs to Q235 due to the large positive photocurrent. This transfer is beneficial to the photocathodic protection of TiO 2 .
TiO2 under alkaline conditions [20,33,34]. The i-V curves of the synthetic CdIn2S4/TiO2 NTA composite showed a higher photocurrent density than that of TiO2. This finding indicates that any composite other than TiO2 can easily inject electrons into Q235 in need of protection. In particular, at the corrosion potential of Q235 (from the Tafel curve), the photo-induced electrons were fully capable of transferring from CdIn2S4/TiO2 NTAs to Q235 due to the large positive photocurrent. This transfer is beneficial to the photocathodic protection of TiO2.  Tafel polarization curves of Q235 with and without composite semiconductor materials are shown in Figure 7.  On the curve of pure Q235, the self-corrosion current of pure Q235 was the current value extrapolated from the straight part of the anode area to the self-corrosion potential on the level of 10 −6 A. The left branch of the polarization curve shows a transient platform, indicating the existence of the cathode reaction of concentration polarization of O2. On the curve of Q235 coupled with CdIn2S4/TiO2 NTAs, the extrapolation values obtained by the same method in the cathode zone are similar to those obtained by i-V curves at −0.96 V (the equilibrium potential from the Tafel curve). This finding explains why the polarization curve of Q235 coupled with CdIn2S4/TiO2 NTAs was found to be above that without coupled CdIn2S4/TiO2. The equilibrium potential demonstrates a negative shift when Q235 was coupled with the CdIn2S4/TiO2 photo-anode due to the existence of the photoinduced electrons. Nonetheless, the dissolving current of Fe decreased under negative potential.
The variations in the potentials of the Q235 electrode coupled with the photo-anode and the current densities from Q235 to the synthetic composite with non-polarization under intermittent visible light were measured to characterize the photocathodic protection performance of the synthetic photo-anode. Figure 8a shows that the potential of Q235 coupled with TiO2 NTAs continued to decrease and slightly fluctuated when the light was turned on and off. This finding indicates that TiO2 NTAs have no photocathodic protection effect on Q235 and may even accelerate the corrosion of Q235. For CdIn2S4/TiO2 NTAs, the potential drop of Q235 coupled with CdIn2S4/TiO2 NTAs reached approximately 400 mV when visible light was switched on. Q235 was therefore polarized to −1.0 V. Ch. Barchiche et al. [35] have proven that Q235 is effectively resistant to corrosion from XRD, SEM, and electrochemical methods. Other studies [36][37][38] have also shown that Q235 might be effectively resistant to corrosion when polarized to −0.9~−1.2 V. Hence, the polarization of Q235 to −1.0 V could be considered effective protection from corrosion [39][40][41]. The dissolution of Fe was substantially inhibited in this potential. This inhibition is due to the response of CdIn2S4 to visible light, the n-n type heterojunctions, and the more negative Ec of CdIn2S4 compared with that of TiO2, contributing to the photocathodic protection effect of the composite on Q235. Figure 8b shows that the current density was negative with and without visible light for TiO2 NTAs. This result indicates that Q235 was able to accelerate corrosion when coupled with the TiO2 NTAs. For the CdIn2S4/TiO2 NTAs, current density sharply increased along the positive direction when the visible light was switched on On the curve of pure Q235, the self-corrosion current of pure Q235 was the current value extrapolated from the straight part of the anode area to the self-corrosion potential on the level of 10 −6 A. The left branch of the polarization curve shows a transient platform, indicating the existence of the cathode reaction of concentration polarization of O 2 . On the curve of Q235 coupled with CdIn 2 S 4 /TiO 2 NTAs, the extrapolation values obtained by the same method in the cathode zone are similar to those obtained by i-V curves at −0.96 V (the equilibrium potential from the Tafel curve). This finding explains why the polarization curve of Q235 coupled with CdIn 2 S 4 /TiO 2 NTAs was found to be above that without coupled CdIn 2 S 4 /TiO 2 . The equilibrium potential demonstrates a negative shift when Q235 was coupled with the CdIn 2 S 4 /TiO 2 photo-anode due to the existence of the photo-induced electrons. Nonetheless, the dissolving current of Fe decreased under negative potential.
The variations in the potentials of the Q235 electrode coupled with the photo-anode and the current densities from Q235 to the synthetic composite with non-polarization under intermittent visible light were measured to characterize the photocathodic protection performance of the synthetic photo-anode. Figure 8a shows that the potential of Q235 coupled with TiO 2 NTAs continued to decrease and slightly fluctuated when the light was turned on and off. This finding indicates that TiO 2 NTAs have no photocathodic protection effect on Q235 and may even accelerate the corrosion of Q235. For CdIn 2 S 4 /TiO 2 NTAs, the potential drop of Q235 coupled with CdIn 2 S 4 /TiO 2 NTAs reached approximately 400 mV when visible light was switched on. Q235 was therefore polarized to −1.0 V. Ch. Barchiche et al. [35] have proven that Q235 is effectively resistant to corrosion from XRD, SEM, and electrochemical methods. Other studies [36][37][38] have also shown that Q235 might be effectively resistant to corrosion when polarized to −0.9~−1.2 V. Hence, the polarization of Q235 to −1.0 V could be considered effective protection from corrosion [39][40][41]. The dissolution of Fe was substantially inhibited in this potential. This inhibition is due to the response of CdIn 2 S 4 to visible light, the n-n Catalysts 2019, 9, 1067 9 of 13 type heterojunctions, and the more negative Ec of CdIn 2 S 4 compared with that of TiO 2 , contributing to the photocathodic protection effect of the composite on Q235. Figure 8b shows that the current density was negative with and without visible light for TiO 2 NTAs. This result indicates that Q235 was able to accelerate corrosion when coupled with the TiO 2 NTAs. For the CdIn 2 S 4 /TiO 2 NTAs, current density sharply increased along the positive direction when the visible light was switched on for the photoelectrode. This finding suggests that a large number of electrons and holes were generated and that the photocurrent was maintained at a relatively stable level. Moreover, the photocurrent density of CdIn 2 S 4 /TiO 2 NTAs reached approximately 50 mA/cm 2 , which was larger than the current produced by the reduction of O 2 and H + (reference self-corrosion current). These findings suggested that Q235 was protected by the coupled photoelectrode under visible light. The high photocurrent can also polarize the coupled Q235 electrode to a much more negative potential, thereby providing an improved cathodic protection. Overall, CdIn 2 S 4 /TiO 2 NTAs showed an effective photocathodic protection effect on Q235 under visible light. A mechanism describing the enhanced photocathodic protection performance of the synthesized CdIn2S4/TiO2 NTA photoelectrode is illustrated in Figure 9. TiO2 and In2S3 can produce a large number of photo-induced charge carriers under bright visible light exposure. The separation of photo-induced carriers is accelerated by the construction of an n-n heterojunction considering the flow of photo-induced electrons from the conduction band of CdIn2S4 into the conduction band of TiO2. The holes from CdIn2S4 and TiO2 are then consumed by the hole trapping agent in the solution. Q235 is protected by the transfer of photo-induced electrons from TiO2 to Q235 because the dissolution of Fe is inhibited at a relatively negative potential.

Materials
All reagents involved in this work were purchased and used directly without any purification. NH4F, AR was obtained from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). A mechanism describing the enhanced photocathodic protection performance of the synthesized CdIn 2 S 4 /TiO 2 NTA photoelectrode is illustrated in Figure 9. TiO 2 and In 2 S 3 can produce a large number of photo-induced charge carriers under bright visible light exposure. The separation of photo-induced carriers is accelerated by the construction of an n-n heterojunction considering the flow of photo-induced electrons from the conduction band of CdIn 2 S 4 into the conduction band of TiO 2 . The holes from CdIn 2 S 4 and TiO 2 are then consumed by the hole trapping agent in the solution. Q235 is protected by the transfer of photo-induced electrons from TiO 2 to Q235 because the dissolution of Fe is inhibited at a relatively negative potential. A mechanism describing the enhanced photocathodic protection performance of the synthesized CdIn2S4/TiO2 NTA photoelectrode is illustrated in Figure 9. TiO2 and In2S3 can produce a large number of photo-induced charge carriers under bright visible light exposure. The separation of photo-induced carriers is accelerated by the construction of an n-n heterojunction considering the flow of photo-induced electrons from the conduction band of CdIn2S4 into the conduction band of TiO2. The holes from CdIn2S4 and TiO2 are then consumed by the hole trapping agent in the solution. Q235 is protected by the transfer of photo-induced electrons from TiO2 to Q235 because the dissolution of Fe is inhibited at a relatively negative potential.

Materials
All reagents involved in this work were purchased and used directly without any purification. NH4F, AR was obtained from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China).

Materials
All reagents involved in this work were purchased and used directly without any purification. NH 4 F, AR was obtained from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). In(NO 3 ) 3 ·4.5H 2 O, Na 2 S·9H 2 O, NaCl, Na 2 SO 3 , C 2 H 5 NS (thioacetamide), CH 3 CH 2 OH (ethyl alcohol), and (CH 2 OH) 2 (ethylene glycol) were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Cd(NO 3 ) 2 ·4H 2 O was acquired from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All the above chemicals used in the experiments were of analytical reagent grade. A custom-made H-type photochemical electrolytic cell was obtained from Tianjin Aida Heng Sheng Technology Development Co., Ltd. (Tianjin, China).

Preparation of TiO 2 NTA and CdIn 2 S 4 /TiO 2 NTA Composites
A two-step anodization process [42,43] was performed to prepare the NTAs using Pt foil, Ti foil, and NH 4 F solution as a cathode, an anode, and an electrolyte, respectively, to produce TiO 2 . The Ti foil (99.99%, 0.30 mm) was cut into 10 × 30 mm 2 foil pieces and was ultrasonically washed with deionized water, ethyl ethanol, and acetone 10 min before its usage. The Pt foil was a 1.0 × 1.0 cm 2 commercial platinum electrode. The NH 4 F solution was obtained by dissolving 1.72 g NH 4 F in 50 mL deionized water and mixing well with 450 mL ethylene glycol. The voltage of anodic oxidation was set to 60.0 V. After the first step of anodizing, the Ti foil was placed in 1 M of HCl for ultrasonic cleaning until a glossy surface was realized. The conditions for the second oxidation were exactly the same as those for the first oxidation. The oxidized Ti sheet was rinsed with deionized water to remove the electrolyte solution from the surface and air dried afterward. The oxidized Ti foil was then placed in the muffle furnace, calcined at 450 • C for 120 min, and left to cool at room temperature. The TiO 2 NTAs were successfully prepared on the surface of the Ti foil.
A facile hydrothermal method was used to load CdIn 2 S 4 nanoparticles onto the surface of the TiO 2 NTAs. In this synthesis, 0.4 mmol of Cd(NO 3 ) 2 ·4H 2 O, 0.8 mmol of In(NO 3 )·4.5H 2 O, and 3.2 mmol of C 2 H 5 NS were dissolved in an ethanol solution containing 15.0 mL of deionized water and 15.0 mL of ethanol. After stirring for 20 min, the mixed solution and prepared TiO 2 NTAs were transferred to a 100 mL Teflon-lined autoclave and heated at 180 • C for 24 h. After the hydrothermal reaction, the Ti foil was rinsed with absolute ethanol and deionized water several times and dried at 80 • C for 12 h. A CdIn 2 S 4 /TiO 2 NTA composite was eventually synthesized.
XRD patterns were used to detect the crystalline structures of the synthesized CdIn 2 S 4 /TiO 2 NTAs and monomers using a Bruker D8 Advance diffractometer with Cu Kα radiation (λ = 0.154 nm). The monomer of CdIn 2 S 4 was collected from the bottom of the Teflon-lined autoclave and dried at 80 • C after washing in deionized water and absolute ethanol successively. The surface morphology of the samples was observed on a Hitachi S-4800 field-emission SEM, and the section was obtained by warping the edges caused by cutting. The surface chemical composition and chemical state of the samples were analyzed using XPS (ESCALAB 250XI, Thermo Fisher Scientific, Waltham, US) with Al Kα (hv = 1486.6 eV). TEM (JEM-F200, JEOL, Tokyo, Japan) was employed to characterize the microstructure of the CdIn 2 S 4 /TiO 2 NTAs. The UV-VIS diffuse reflectance (HITACHI U-3900H, HITACHI, Tokyo, Japan) and TiO 2 photoluminescence spectra (Edinburgh FLs980) were measured to explore the optical effect of loading CdIn 2 S 4 . Photoluminescence spectra were used to characterize the recombination rates of photo-induced carriers using a FLS980 Series fluorescence spectrometer (UK).
The photo-induced volt-ampere characteristic curves were realized in a three-electrode configuration in a 0.35 M Na 2 SO 3 and 0.25 M Na 2 S solution. The synthesized materials, SCE, and Pt electrode were used as the working, reference, and counter electrodes, respectively. The i-V curves were measured from −1.5 V to 0.5 V with a scan rate of 0.02 V/s and with visible light switched on and off every 2 s. The working, reference, and counter electrodes were the Q235 electrode, SCE, and Pt electrode, respectively.
Tafel plots, variations in the potentials, and the current densities under intermittent visible light were achieved in an H-type electrolytic cell. This cell was divided into the photocell and the corrosion cell was connected by a Nafion membrane. The electrolyte in the photocell was 0.35 M Na 2 SO 3 and was placed in a 0.25 M Na 2 S solution. The synthesized materials connected by a copper wire were placed in this cell. Meanwhile, the SCE, Pt electrode, and Q235 electrode were placed in the corrosion cell with a 3.5 wt.% NaCl solution. The Tafel plots for Q235 were obtained at a voltage range of −300 mV to 300 mV considering the open circuit at a scan rate of 0.166 mV/s, and Q235 was coupled with the synthesized materials under visible light.
The effect of photocathodic protection was determined by variations in the potential and current density tests. A Xe lamp and 3.5 wt.% NaCl solution were used to simulate intense solar light and seawater respectively. These properties were realized in an H-type electrolytic cell. In the variations in the potential test, Q235 coupled with the synthesized materials, SCE, and Pt electrode were used as the working, reference, and counter electrodes, respectively. The synthesized materials were irradiated with visible light every 200 s for 200 s at a time. In the current density test, Q235 was connected to the ground wire (GND wire), the synthesized materials were connected to the working electrode, and the counter and reference electrode wires were short connected. The synthesized materials were irradiated with visible light every 100 s for 100 s at a time.

Conclusions
In this work, CdIn 2 S 4 /TiO 2 NTA nanocomposites were successfully prepared via two-step anodization of Ti followed by a hydrothermal method. The CdIn 2 S 4 nanoparticles were distributed on the top surface of the NTAs and the tube walls with a relatively symmetrical octahedral structure. The optical absorption region was expanded and optical absorption in the visible light region was enhanced relative to TiO 2 NTAs, thereby facilitating the utilization of visible light. The PL curves indicate that the enhanced photocathodic protection of CdIn 2 S 4 /TiO 2 NTA nanocomposites can be ascribed to the formation of an effective n-n type heterojunction electronic field at the interface of CdIn 2 S 4 and TiO 2 . This electronic field remarkably improves the separation efficiency of the photo-induced electron-hole pairs. The feasibility of converting excellent optical properties to electrical properties was examined by i-V and Tafel curves. The results of the variations in the potentials and the current densities indicates that CdIn 2 S 4 /TiO 2 nanocomposites can provide a favorable photoelectrochemical cathodic protection for the coupled Q235 under visible light in the simulated splash zone.