Abstract

The performance of bare TiO2 photoanodes in photoelectrochemical devices and other applications is suboptimal due to the narrow light absorption range and poor electron-hole separation. Here, heterostructured films of Ni-doped lead titanate (PbTiNiO3, PTN)-coated titania (TiO2) nanorod arrays were fabricated via a two-step process comprising hydrothermal and sol-gel methods. In the PTN/TiO2 composite films, the PTN component served as a visible light-responsive photosensitizer and the TiO2 nanorod was the electron transport layer. When applied as photoanodes, the PTN/TiO2 films achieved the maximum photoconversion efficiency of ~2.6% and a photocurrent intensity of about 4 times higher than that of the PbTiO3/TiO2 film. These results demonstrate that the ferroelectric material-coated TiO2 heterostructured films have high potential for application in photoelectric and optical devices.

1. Introduction

The applications of photosynthetic and photocatalytic reactions based on semiconductor materials using photoelectrochemical (PEC) cells have been recognized as ideal renewable systems leveraged on green energy technology [1, 2]. Since the PEC system was first developed with TiO2 thin film by Fujishima and Honda in 1972 [3], TiO2 has become the most widely studied photochemical material due to its advantages of high electron mobility, high photocorrosion resistance, low cost, and nontoxicity. However, low utilization efficiency of solar energy resulting from the wide bandgap of TiO2 (~3.0 eV for rutile and ~3.2 eV for anatase) and relatively high recombination rate of photogenerated charge carriers significantly disincentivize its application in PEC systems. To overcome these shortfalls, TiO2 is commonly dye sensitized, ion doped, or combined with other semiconductors to form heterojunctions. As a consequence, the doping process introduces absorption states in the energy gap, enabling absorption of lower-energy photons [46]. Semiconductor junctions can promote separation of charge carriers through favorable energy band alignments; this potentially reduces the occurrence of recombination losses [7]. For example, many strategies have been developed to control the synthesis of heterostructured composites, such as Bi2S3/TiO2 [8], Cu2O/TiO2 [9], and ZnO/TiO2 [10]. In particular, TiO2 modified by CdS quantum dots is much more exploited for PEC application and reached a maximum IPCE value of nearly 80% [11, 12]. This is aimed at improving their PEC activity by reducing the recombination rate of the photogenerated electron-hole pairs.

Since the discovery that ferroelectric oxides possess photovoltaic effect unlike the conventional semiconductors (such as CdS, Cu2O, and Bi2S3) [13], ferroelectric oxides are now increasingly applied in conjunction with TiO2. Previous studies have suggested that, in the heterostructured TiO2 and ferroelectric oxide composites, visible light is absorbed by the ferroelectric component, in which charge carriers are generated and possibly separated by its inherent electric field, which arises from spontaneous polarization and is subsequently transferred from the ferroelectric component to TiO2 [1416]. In a prior work, the composite films of BiFeO3 coating TiO2 nanorod arrays were prepared to study the enhanced photocatalytic performance in visible light [17]. Lead titanate (PbTiO3, PTO), a well-known solid ferroelectric oxide, exhibits excellent ferroelectric and piezoelectric properties and, thus, has been extensively investigated over the past several decades [1820]. Since light absorption and carrier concentration both depend on bandgap, polar materials with narrow bandgap are highly desirable. Furthermore, first-principles calculations revealed that the bandgap can be reduced by doping the TiO6 octahedron in PbTiO3 with oxygen vacancy- (O vacancy-) stabilized ions (where , Pd, and Pt) and, concurrently, a large polarization value is retained [21, 22].

On the other hand, for solar energy conversion based on narrow bandgap semiconductors, the controlled synthesis of vertically aligned one-dimensional (1D) nanostructures is an effective way to develop highly efficient energy conversion materials. Aligned 1D nanostructures, such as nanowires, nanotubes, and nanorods, offer a short diffusion length for holes along the radial direction, while the long axial direction of the nanostructures becomes the preferred electron channel and provides optimal length for light absorption [23, 24]. Numerous studies have reported that the aligned 1D nanostructures provide superior photocatalytic, photovoltaic, and PEC properties relative to their bulky counterparts and randomly shaped particles [2527]. For instance, pioneer work by Lindgren et al. [28] showed the use of hematite nanorod arrays as photoanodes in a PEC cell and revealed that 1D nanostructures could potentially address some of the fundamental issues in PEC systems and improve device performance.

On the basis of the introduction above, composite heterostructured films of PbTiNiO3 (PTN, , 0.06, and 0.15, abbreviated as PTO, PTN6, and PTN15, respectively)-coated TiO2 nanorod arrays were fabricated via a two-step process: First, the TiO2 nanorod arrays were grown hydrothermally on F:SnO2 (FTO) conducting substrate. This was followed by the formation of a PTN ferroelectric layer around the TiO2 nanorods. The phase structures and morphologies of the composite films, as well as optical adsorption properties, were systematically investigated. Finally, the PEC performance was characterized under visible light irradiation.

2. Methods

2.1. Materials

All chemicals were of analytical grade and were used without further purification. Hydrochloric acid (37% mass fraction), acetic acid (CH3COOH, 99.5%), acetylacetone (CH3COCH2COCH3, 99.0%), tetrabutyl titanate (Ti(OC4H9)4, 98.0%), and lead acetate (Pb(CH3COO)2·3H2O, 99.5%) were purchased from Sinopharm Chemical Reagent Co., Ltd., China. Nickel nitrate (Ni(NO3)2·6H2O, 99.99%) was purchased from Sigma-Aldrich (Shanghai) Trading Co., Ltd., China. FTO substrates were purchased from Kejing (Hefei) Materials Technology Co., Ltd., China. Deionized water was homemade at the laboratory.

2.2. Growth of TiO2 Nanorod Arrays

TiO2 nanorod arrays were grown on FTO conductive glass via a hydrothermal method, as previously reported [17]. Briefly, the FTO substrates were cut into rectangle pieces and ultrasonically cleaned with deionized water, acetone, and ethanol, in that order. A mixture of hydrochloric acid and deionized water (mixed at volume ratio of 1 : 1), containing a suitable amount of Ti(OC4H9)4, was used for the hydrothermal reaction. Subsequently, the FTO was put into the autoclave containing the mixture solution and kept in an oven at 180°C for 6 h. Finally, the TiO2 nanorod arrays were grown on the surface of the FTO substrates.

2.3. Fabrication of Pb(TiNi)O3/TiO2 Thin Films

Precursor solutions of PTO, PTN6, and PTN15 were prepared via a sol-gel technique. Pb(CH3COO)2·3H2O, Ni(NO3)2·6H2O, and Ti(OC4H9)4 were used as the starting materials. Acetic acid and acetylacetone served as the solvent and chemical stabilizer, respectively. The concentration of the final solutions was adjusted to 0.3 mol/L, while the pH value was set at 3–4 with acetic acid. Subsequently, the precursor solutions were spin coated three times onto the as-prepared TiO2 nanorod arrays to create uniform wet films. After each spin coating step, the wet films were initially dried at 180°C for 5 min to evaporate the solvent and then rapidly heated to 400°C for 10 min to remove the residual organic materials. Finally, the films were annealed at 600°C for 30 min to allow the titanate to crystallize, forming the PTN/TiO2 composite films.

2.4. Physical Characterization

The phase structures and morphologies of the as-prepared TiO2 nanorod arrays and PTN/TiO2 films were characterized by X-ray diffraction (XRD, D8 Advance, with Cu radiation, Bruker Corp., Germany) and scanning electron microscopy (SEM, JSM-7100F, Hitachi Corp., Japan), respectively. The ultraviolet-visible absorption spectra were recorded with an ultraviolet-visible light (UV-Vis) spectrophotometer (UV2600, Shimadzu Corp. Japan).

2.5. PEC Measurements

The PEC performance of the TiO2 nanorod arrays and PTN/TiO2 composite films was measured in a three-electrode configuration on an electrochemical workstation (CHI660E, CH Instruments, Inc., Shanghai, China). The samples, a Pt foil, and Hg/HgO in aqueous KOH, served as the working, counter, and reference electrodes, respectively. A solution of 0.1 M KOH was used as the electrolyte for all of the electrochemical measurements. The effective area of the photoelectrodes was calculated to be after encapsulation. The photocurrent-potential and photocurrent response curves were recorded under visible light irradiation using an AM 1.5 solar simulator with a 300 W Xe lamp (100 mW/cm2). Electrochemical impedance spectroscopy (EIS) experiments were performed in the absence of light by applying an initial potential of 1.0 V, and the amplitude of the ac potential was kept at 5 mV over the frequency range of 0.1 Hz to 1 MHz.

3. Results and Discussion

Figure 1 shows the XRD patterns of TiO2 nanorod arrays, PTO/TiO2, and PTN/TiO2 composite films. In Figure 1(a), all of the diffraction peaks, in addition to those representing the FTO substrate, can be indexed to the standard patterns of tetragonal perovskite PTO (JCPDS Card No. 70-0746) and rutile TiO2 (JCPDS Card No. 21-1276). The XRD spectra of the composites were found to be consistent with the superposition of the two component phases, and no diffraction peak from the impurity and lead-deficient pyrochlore phases were recorded. This indicated that there was no chemical reaction between PTO and TiO2. In Figure 1(b), it is evident that by increasing the Ni content, the (101) and (110) peaks of the PTN components become partially overlapped in the case of PTN6 and completely overlapped in the case of PTN15. The analysis reveals that the PTN6 film exhibits a multiphase coexistence of pseudotetragonal and cubic symmetries, whereas the PTN15 film contains a single phase of cubic perovskite [29, 30]. The origin of the phase transition in PTN can be attributed to the inner stress field caused by the substitution of Ti4+ by the larger Ni2+, leading to the disruption of the Ti-O orbital hybridization balance [31].

Figure 2 displays the surface and cross-section morphologies of TiO2 nanorod arrays and the PTO/TiO2 and PTN/TiO2 composite films characterized by SEM. In Figures 2(a) and 2(b), the nanorods were grown independently on the surface of FTO and were quasialigned along the plane vertical to the substrate. TiO2 nanorods display a tetragonal cross-section with smooth side facets and rough square top facets. After being synthesized at 180°C for 6 h, the average diameter and length were estimated to be ~100 and 850 nm, respectively. Figures 2(c)2(h) depict the surface and cross-section morphologies of the PTO/TiO2 and PTN/TiO2 films. Although the nanorods cannot be identified on the surface images, PTN crystalline particles were observed. Indeed, the cross-section image shows that the voids between the nanorods were filled by the PTN components, and the thickness of the films was equivalent to the length of the nanorods. The image reveals that the pores on the surface of the films decreased gradually, indicating that the average grain size of PTN decreases with increasing Ni content. The decrease in grain size can be ascribed to the lower grain growth rate in PTN. This can be explained by Pauling’s theory of electronegativity; that is, the bond energy of Ni-O is weaker than that of Ti-O because of the smaller difference in electronegativity between Ni-O (1.53) and Ti-O (1.90). This results in less heat being released during the formation of the Ni-doped PTO, leading to a slower grain growth rate and smaller grain size [32].

The optical absorption spectrum of the semiconductor is required to determine its bandgap and the corresponding photoelectrical performance. Generally, the corresponding optical bandgaps can be estimated by the tangent lines in the plot of versus according to Tauc’s law for the direct bandgap materials, where is the absorbance and is the photon energy. However, some researchers believe that Tauc transformation should not be applied directly to the spectrum of both components together but to the spectrum of the semiconductor alone [33]. Thus, the TiO2 nanorod arrays, PTO, and PTN thin films were prepared on FTO separately for optical absorption spectra analysis, as presented in Figure 3(a). As expected, TiO2 nanorod arrays exhibited little absorbance at wavelengths longer than 400 nm. In contrast, the onset of the absorption occurred at a longer wavelength and exhibited a redshift when Ti in PTO was partially replaced by Ni. The absorption edge of PTN15 thin film begins at ~450 nm, which is consistent with the bandgap of 2.78 eV according to the linear fitting of the Tauc plot, as presented in Figure 3(b). These results point to the conclusion that PTN is an effective visible light-absorbing component of PTN/TiO2 composite films.

This bandgap narrowing effect is due to the formation of new states in the electronic structure of both the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), which are associated with the presence of Ni stabilized by an accompanying O vacancy [32]. The mechanism of bandgap narrowing due to Ni doping is shown by the inset of Figure 3(b). In PTO, the HOMO is localized around O, wherein a weak interaction occurs between the O-2p orbital and the Ti-3d and Pb-6s orbitals. The LUMO is localized around Ti and resembles the 3d state. Since the Ni and O vacancies replace those of Ti, the Ni-3d state falls in the bandgap of PTO, thereby reducing the bandgap of the compound. In this case, the HOMO is localized around Ni and resembles the 3d state, whereas the LUMO is localized around the Ni-3d state, resembling the 3d state. The tailing of the conduction band edge into the bandgap increases because the energy of the Ni-3d state is lower than that of the Ti-3d state [34]. The decrease in the bandgap improves the light absorption by the Ni-modified PTO in ultraviolet and visible regions, implying that the Ni-modified PTO can potentially increase the solar absorption efficiency of perovskite-type ferroelectrics in solar photovoltaic devices.

To assess the performance of this type of composite film in photoelectric conversion, current-voltage characteristics were investigated. The photocurrent () versus applied voltage () curves (linear voltage sweep) shown in Figure 4(a) revealed that the PTN15/TiO2 nanorod electrode had a significantly enhanced unsaturated PEC photocurrent in the high-voltage region (above 1.5 , RHE represents reversible hydrogen electrode) when compared to the other electrode materials. In the low-bias region (inset of Figure 4(a)), the PTN/TiO2 nanorod electrodes exhibited a higher photocurrent intensity compared to the values measured under dark conditions. This indicates a higher visible light-driven PEC activity of the TiO2 nanorod electrodes modified by the PTN photosensitizer. However, no significant difference was observed between the intensities of photocurrent and dark current of the bare TiO2 nanorod photoanode.

To further evaluate the performance of the PTN/TiO2 photoelectrodes, the photoconversion efficiency () was calculated from the photocurrent curves using the following equation [35]: where is the applied bias versus RHE, is the photocurrent density at the measured bias, and is the irradiance intensity of 100 mW·cm−2. Figure 4(b) shows the photocurrent efficiency versus bias potential for the PTN/TiO2 and PTO/TiO2 electrodes. The plot reveals that the PTN15/TiO2 electrode exhibits significantly higher photoconversion efficiency compared with the PTN6/TiO2 and PTO/TiO2 electrodes, and a maximum efficiency of ~2.6% was obtained at a bias potential of 2 V, which is ~10 times higher than the efficiency value of the pristine PTO/TiO2 electrode. For the PTO/TiO2 photoelectrode, the photoconversion efficiency is consistent with that reported by Chandrasekaran et al. [35] and Tabari et al. [36].

EIS is an important tool used to characterize the migration of charge carriers. The charge transport performance of the PTN-modified TiO2 nanorod arrays is depicted in Figure 5(a), which represents Nyquist plots of PTN15/TiO2 electrodes and the reference bare TiO2 nanorod arrays. Generally, the arc at lower frequency of the Nyquist plot reflects the charge transfer impedance at the electrode/electrolyte interface, whereas a smaller arc radius stands for a faster charge transfer process with lower recombination [37]. Comparison of the EIS data revealed that PTN15/TiO2 electrodes were considerably superior to the film of TiO2 nanorod arrays with a smaller semicircle, demonstrating an effective charge separation and faster charge carrier transport [38, 39]. In addition, an equivalent circuit model was proposed by fitting the experimental data with analysis software (ZSimpWin, Princeton Applied Research, USA) [40], as shown in the inset of Figure 5(a). In the circuit model, is the resistance of the solution, and and represent the resistance to charge transfer in the semiconductor/electrolyte and semiconductor/substrate interfaces, respectively. and represent the constant phase elements (CPEs) of the semiconductor/electrolyte and semiconductor/substrate interfaces, respectively [41]. The values of , , and were calculated as , 47.15, and 0.01 Ω·cm2 for the bare TiO2 and , 39.9, and 0.01 Ω·cm2 for the PTN15/TiO2 films, respectively. These results further prove that the TiO2-PTN composite decreases the carrier transfer resistance, thereby improving the effective electron-hole separation and photoelectric conversion performance of PTN-modified TiO2 nanorods.

Furthermore, the time-dependent photocurrent generation process (amperometric curves) under intermittent light irradiation is presented in Figure 5(b). The three curves reveal the rapid change in photocurrent intensity with irradiation switching, along with excellent stability and reliability during the on-off irradiation cycles. Moreover, the photocurrent intensity significantly increases with the Ni content. The photocurrent intensity of PTN15/TiO2 films (~70 μA) is nearly 4 times higher than that of PTO/TiO2 films (~18 μA). It is worth noting that a significant decay with the light on and a burst with the light off were observed in the curve of PTO/TiO2. The main reason of photocurrent decay can be attributed to the recombination of photogenerated electron-hole [42]. The surface polarization charges formed on PTO might be a key reason for the photocurrent explosion with light off. In contrary, the recombination of photogenerated electron-hole on the surface of PTN/TiO2 is reduced with the doping of Ni. Meanwhile, ferroelectric polarization is also reduced in PTN with the introduction of Ni, which leads to the burst of photocurrent that was not observed.

4. Conclusions

In conclusion, Ni-doped PTO ferroelectric material-coated TiO2 nanorod films were fabricated by growing TiO2 nanorod arrays via a hydrothermal method and synthesizing ferroelectric PTN by a sol-gel method. After annealing, the PTO ferroelectric crystals partially substituted with Ni in B-sites were deposited on TiO2 nanorods as optical absorbers. The integration of TiO2 nanorods and PTN ferroelectric material dramatically extended the wavelength range for light absorption and reduced the charge transfer resistance. This, in turn, facilitated charge separation and electron transfer, resulting in an improved PEC activity under visible irradiation when compared to the individual component phases. Herein, the PTN/TiO2 composite displayed high potential as a photoanode agent, with high sensitivity to visible light, presenting a new approach of developing effective ferroelectric material-modified TiO2 assemblies for photoelectronic applications and solar photovoltaic device integration.

Abbreviations

PTN:Ni-doped lead titanate, PbTiNiO3
PTN6:PbTi0.94Ni0.06O3
PTN15:PbTi0.85Ni0.15O3
PTN/TiO2:PbTiNiO3/TiO2
PTO/TiO2:PbTiO3/TiO2
PEC:Photoelectrochemical
FTO:F-doped SnO2
JCPDS:Joint Committee on Powder Diffraction Standards
XRD:X-ray diffraction
SEM:Scanning electron microscopy
UV-Vis:Ultraviolet-visible light
HOMO:Highest occupied molecular orbital
LUMO:Lowest unoccupied molecular orbital
EIS:Electrochemical impedance spectroscopy
O vacancy:Oxygen vacancy.

Data Availability

All data supporting the conclusions of this article are included within this article.

Conflicts of Interest

DZ is a lecturer and a Ph.D. degree holder specializing in the investigation of photoelectric and nanometer materials. GZ is a professor and a Ph.D. degree holder specializing in the investigation of functional materials. ZLH is an associate professor and a Ph.D. degree holder specializing in the investigation of optical materials. YFT and YT are associate professors and Ph.D. degree holders specializing in the investigation and characterization of electrochemistry. YF and ZPM are graduate students major in the study of nanometer materials. The authors declare that they have no competing interests.

Authors’ Contributions

YF drafted the manuscript. YF and ZPM performed the experimental works. YFT, YT, and GZ carried out the series characterization of the samples and interpreted the data. DZ and ZLH conceived the idea of experiments and amended the manuscript. All authors read and approved the final manuscript.

Acknowledgments

Authors thank LetPub (https://www.letpub.com) for its linguistic assistance during the preparation of this manuscript. This work was financially supported by the National Natural Science Foundation of China (Grant Numbers: 51502114 and 61575085).