Visible-light responsive BiNbO4 nanosheet photoanodes for stable and efficient solar-driven water oxidation

Maheswari Arunachalam a, Kwang-Soon Ahn b and Soon Hyung Kang *c
aDepartment of Chemistry, Chonnam National University, Gwangju 500-757, Republic of Korea
bSchool of Chemical Engineering, Yeungnam University, Gyeongsan 712-749, Republic of Korea
cDepartment of Chemistry Education and Optoelectronic Convergence Research Center, Chonnam National University, Gwangju 500-757, Republic of Korea. E-mail: skang@jnu.ac.kr; Tel: +82-62-530-2497

Received 17th April 2020 , Accepted 5th June 2020

First published on 8th June 2020


Herein, we report bismuth niobate (BiNbO4), which is regarded as an emerging photoanode material for sustainable photoelectrochemical (PEC) solar energy conversion. BiNbO4 possesses a direct bandgap (Eg) of ∼2.6 eV, and shows an appropriate band alignment for the water oxidation/reduction reaction. In this study, a simple sol–gel route followed by a spin coating method was applied to develop BiNbO4 nanosheets under the optimum annealing conditions. It is known that the annealing temperatures of 500 and 550 °C influence the crystallinity and PEC properties of BiNbO4 films. In particular, the 550 °C annealed film exhibited sharply improved crystalline properties, and rapidly enhanced PEC performance, which were accompanied by a photocurrent density of 0.45 mA cm−2 at 1.23 V vs. the reversible hydrogen electrode (RHE) (briefly abbreviated as 1.23 VRHE) in a strong alkaline solution of 1 M NaOH, compared with 0.26 mA cm−2 at 1.23 VRHE of the 500 °C annealed film. This may be attributed to the main increase of the crystallinity, as well as the improvement of the electronic properties. In addition, the BiNbO4 (550 °C) film showed an incident photon-to-current efficiency of 20% at 425 nm, and produced a stable photoresponse under light illumination in a strong alkaline solution over 5 h, compared with a BiVO4 electrode.


Introduction

Solar energy has been an essential resource for self-conversion to high-valued chemical fuels, such as water splitting to H2 and O2, CO2 reduction to hydrocarbon fuels, and N2 reduction to ammonia products, because the use of clean, unlimited and strong solar energy corresponds to no cost or regional limits, except for the polar regions1–5 Besides, the efficiency of energy producing devices is critically important in terms of technical and research points. In general, the photoelectrochemical (PEC) water splitting consists of a photoanode for water oxidation, a photocathode for water reduction, and an electrolyte to be used under solar illumination are important. Herein, the multi-step complex reactions have usually been carried out together, and a lot of loss of charge occurs during the several stages in these multiple reactions.6,7 Accordingly, an effective photoelectrode must have fast carrier separation and transfer of the photogenerated carriers to reduce the charge losses from the charge recombination reaction. Numerous semiconducting oxides (e.g., α-Fe2O3, BiVO4, or WO3), which partially fulfill these conditions in terms of the photostability, abundance, efficient light harvesting in the visible spectrum, favorable band alignment, fast interfacial water oxidation kinetics, and nontoxicity, have been investigated to date.8–10 Although for PEC devices, highly competitive and practical photoelectrodes have still been limited, most of the candidate metal oxides have exhibited an inappropriate band position, weak photostability, insufficient visible light harvesting ability, and inadequate charge transfer yield.11–13 Therefore, the pursuit of new materials having novel and fascinating characteristics can be a crucial and challenging task. Sometimes, a certain semiconducting photoelectrode showing excellent photoabsorption properties and an appropriate band alignment can be thermodynamically unstable, owing to photocorrosion.14 Photostability can mainly depend on materials’ surface chemical properties, cocatalyst activity, charge carrier density, and electrolyte acidity. In this regard, ternary semiconducting photoelectrode materials show enhanced visible light absorption and photo-catalytic properties at affordable cost.15–20 In addition, these rely on their chemical composition ratio, and their chemical and electronic characteristics can be facilely modulated. In particular, MNbO4-type metal niobates (M = Bi, Ta, and Fe) display energetic electronic properties, because niobium possesses electronic transition states with 4d orbitals, a suitable band structure responsive to visible-light irradiation, hybridized O 2p and Bi 6s2 energy bands, and exhibit high activity under visible light, which contribute to the optical absorption arising from the M element–O transitions for the energy conversion processes.21–25

Among them, we chose the bismuth orthoniobate (BiNbO4) materials prepared by the simple solid-state method, due to their easy structural phase transformation that is highly dependent on the annealing temperature, and they are available in cubic, orthorhombic, and triclinic phases.26 Their enhanced PEC activities were reported to be closely associated with the reduction of the Nb 4d level, along with the distorted octahedral (NbO6) units, improving the visible-light absorption. According to other findings, the band structure of BiNbO4 consists of Bi3+ of 6s and 2p of oxygen at the valence band (VB) and 4d of Nb5+ at the conduction band (CB). Interestingly, the position of VB atoms can be altered by the growth or annealing temperature, enabling the making of several phase structures. That is, in the case of β-BiNbO4, the 2p level is higher than the 6s2 level, which is just the reverse of α-BiNbO4. For this reason, α-BiNbO4 displays a narrower bandgap than that of the high-temperature β-BiNbO4, as well as high photostability for a wide range of pH solutions, becoming a potential candidate for practical applications. In order to closely investigate BiNbO4 (briefly marked as BNO), herein, BNO was synthesized at different annealing temperatures of 500 and 550 °C. The basal structure of BNO consists of nanosheets, which enable improving the light harvesting yield. Furthermore, BNO has great potential to be used in four-electron oxidation for water photoelectrolysis to evolve oxygen without anodic photocorrosion, and is regarded as an attractive ternary photoanode for photostable PEC water oxidation. This indicates that BiNbO4 without any co-catalyst holds more promise for practical utility, compared with other metal oxide electrodes with smaller band gap energies.

Experimental

Materials

Bismuth(III) nitrate pentahydrate (Bi(NO3)3·5H2O), niobium(V) chloride, nitric acid (HNO3, 60%), and ethanol (C2H5OH) were purchased from Sigma-Aldrich. All chemicals were used without any further purification.

Sample fabrication: synthesis of BiNbO4 films

Briefly, the respective concentration and final volume of the precursor solutions were fixed to be 0.1 M and 20 mL, respectively. Then, 0.1 M NbCl5 was dispersed in a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 vol ratio of deionized water (DI) and the HNO3 (60%) solution, and the obtained mixture was stirred over a long period of time, until the precursor solution changed to transparent. Subsequently, the required amount of Bi(NO3)3·5H2O was added to the above solution under continuous stirring. Finally, 10 mL of ethanol was mixed with this solution. Then, the fluorine-doped tin oxide (FTO) substrates (2 cm × 2 cm) were cleaned with a liquid detergent, and rinsed several times with DI water. To improve the adhesion of the BiNbO4 layer to the substrates, the FTO substrates were immersed in a solution of DI water, H2O2, and conc. H2SO4, (1[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]3 vol ratio) for 10 min. Afterwards, the FTO substrates were thoroughly cleaned with DI water, and finally dried with a N2 stream. To develop the BiNbO4 layer by a simple spin-coating tool, 50 μL of the corresponding precursors were spun at 1500 rpm for 30 s onto the FTO substrate; and subsequently, the films were dried at 150 °C on a hotplate. This procedure was repeated to get the optimum thickness. In order to choose the optimum annealing temperature, as-prepared BiNbOx was subjected to Thermal Gravimetric Analysis (TGA) as shown in Fig. S2 (ESI). Finally, the obtained amorphous film was annealed at 500 or 550 °C for 3 h in a box furnace maintaining a heating rate of 2 °C min−1, to obtain the targeted products. Samples are denoted as BNO (500 °C) and BNO (550 °C), respectively.

Characterization

Thermal stability of the as-synthesized materials was determined by TGA using a thermal analyzer (PerkinElmer TGA7), the measurement was performed from room temperature to 800 °C at a scan rate of 10 °C min−1 under an air atmosphere. A High-Resolution X-Ray Diffraction (HR-XRD, PANalytical, X’Pert PRO) instrument operating at 40 kV and 30 mA was used to obtain detailed information on the structural and crystalline properties of the BNO films. Field Emission-Scanning Electron Microscopy (FE-SEM) images were recorded on a S4800 (Hitachi Inc. instrument) at a 20 kV acceleration voltage. UV-vis spectra were recorded in the wavelength range of 400–700 nm using a PerkinElmer UV-vis Lambda 365 spectrometer. UPS spectra were recorded using He I excitation (21.2 eV) with a constant pass energy of 5 eV in the ultrahigh vacuum (UHV) chamber of the XPS instrument. UPS binding energies were referred to the Fermi edge of Au, which was sputtered onto the sample in the UPS chamber. The chemical composition was analyzed by X-ray photoelectron spectroscopy (XPS, K-Alpha 1063) with monochromatic Al Kα irradiation.

Linear Sweep Voltammetry (LSV) with chopping mode (on/off interval time = 2 s) was carried out at a scan rate of 20 mV s−1 from 0.3 to 1.8 VRHE in a standard three-electrode configuration: the working, platinum counter, and homemade saturated Ag/AgCl electrodes (0.11 V vs. normal hydrogen electrode). This potential can be converted to the RHE (= NHE at pH = 0), using the following eqn (1):

 
ERHE = EAg/AgCl + 0.0591 V·pH + 0.11 V(1)

An aqueous electrolyte containing 1 M NaOH (pH = 13.5) was used for the PEC test. A Newport Oriel Xenon 150 W solar light simulator (100 mW cm−2, AM 1.5G) was used as the light source. LSVs were conducted at a scan rate of 20 mV s−1 with chopped light alternating between the dark and light every 2 s. A UV cutoff filter (λ > 420 nm) was used to simulate the visible light. Incident photon-to-current conversion Efficiency (IPCE) was measured as a function of wavelength from 300 to 600 nm at an applied potential bias of 1.23 VRHE, using an IPCE system designed for PEC water splitting. Herein, a 150 W Xe lamp was employed to produce the monochromatic beam. The calibration of the 1 sun illumination was in compliance with NREL-certified Si photodiodes. Electrochemical impedance spectroscopy (EIS) of all the samples was conducted to estimate the cell resistance of each component. The working conditions were measured using a standard potentiostat (Autolab/PGSTAT, 128N) equipped with an impedance-spectra analyzer (Nova). The measured frequency ranged from 0.1 Hz to 10 kHz at an amplitude of ±10 mV. Moreover, to measure the flat-band potentials (VFB) and donor concentrations of the films, Mott–Schottky plots (Autolab/PGSTAT, 128N) at frequencies of 100–250 Hz were obtained using a standard potentiostat equipped with an impedance spectra analyzer (Nova), using the same electrochemical configuration and electrolyte, under dark conditions. The double-layer capacitance of the electrodes (Cdl) was derived from the cyclic voltammograms recorded in 1 M NaOH (pH 13.5) within the potential range of 0.2 to 0.4 VRHE with various scan rates of 25–200 mV s−1 under dark conditions.

Results and discussion

Fig. 1 shows the X-ray diffraction (XRD) patterns that confirmed the structural features of the BNO (500 °C) and BNO (550 °C) films. The main diffraction peaks of the samples (Fig. 1(a) and (b)) can be assigned to the (111), (121), (130), (040), (002), (060) and (232) planes, located at 2θ of 26.44, 28.18, 29.15, 30.50, 37.73, 47.95 and 55.14°, respectively. The XRD pattern of BiNbO4 well matches with that of the phase-pure orthorhombic α-BiNbO4 (JCPDS No. 82-0348), which is consistent with the previous literature.22 Also, the additional diffraction peaks are observed at 33.62, 37.84, and 51.60°, which come from the FTO substrate. No significant difference in the XRD pattern is observed between the BNO (500 °C) and BNO (550 °C) films. The average crystallite size ranges 14 to 20 nm for BNO (500 °C) to BNO (550 °C), estimated from the major diffraction peak (2θ = 28.18°) using Scherrer's equation, and can probably be ascribed to the slight growth of the composed grains under the higher temperature annealing conditions. Also, the intensities of the peaks are significantly enhanced as the calcination temperature increases, indicating the improved crystallization of BNO (550 °C) at higher temperature. To confirm this phenomenon, the crystallinity (%) of the films was calculated to be approximately 81% for 500 °C and 88% for 550 °C, respectively. Moreover, the strongest (121) diffraction peak of the orthorhombic structure suggests that the nanosheets on the FTO substrate prefer to grow in the (121) orientation, and it appears that quite mechanically stable BNO (550 °C) phases form in the high-temperature processes.
image file: d0cp02071k-f1.tif
Fig. 1 Typical XRD patterns of the (a) BNO (500 °C) and (b) BNO (550 °C) films.

Fig. 2(a–f) show the top-view FE-SEM images of the representative BNO (500 °C) and BNO (550 °C) films. The synthesized BiNbO4 films have a distinct nanosheet morphology, meaning that the annealing temperature barely alters the growth process of BiNbO4. Fig. 2(a–c) display the slightly anisotropic shape and asymmetrical morphology of the composed grains in the BNO (500 °C) film. This is intimately related to the anisotropic properties at the relatively low temperature, and it can be expected that most of the photogenerated carriers within these large particles cannot even diffuse to the surfaces before the charge recombination occurs. Then, as the annealing temperature was increased to 550 °C, the aggregated sheets disappear, and uniform vertical nanosheets are established, which cross with each other. The BNO (550 °C) film shows a clearly textured characteristic with the preferential growth direction (the same as the XRD results). It can be observed from a larger magnification image that the nanosheets possess a sheet-like morphology, of which the width of the nanosheets is found to be approximately 80–110 nm (Fig. 2(f)). Additionally, the wall thickness of the nanosheets can be estimated to be about 30 nm and it approaches the approximately 1 μm range. This indicates that the vertically aligned and nanosized nanosheet morphology can be beneficial for shortening the diffusion distance of the photogenerated charge carriers and the large exposed surface area for electrocatalytic reactions. Thus, the nanosheet structure with a high surface area can be considered as the morphology for efficient charge transfer, finally improving the PEC water-splitting performance.


image file: d0cp02071k-f2.tif
Fig. 2 Surface FE-SEM views of (a–c) the BNO (500 °C) and (d-f) BNO (550 °C) photoanode films.

To understand the optical properties of the BiNbO4 films, the UV-Vis absorption spectra were acquired, as shown in Fig. 3(a). Overall, the absorbance of the BNO (550 °C) film is higher than that of BNO (500 °C), presenting high light absorption in the 350–500 nm wavelength range, probably ascribed to the high absorption probability of charges in the environment of the highly crystalline BNO (550 °C) film. A similar phenomenon was found in the recent reports.23–27,43 Also, compared with the BNO (500 °C) film, the absorption onset of BNO (550 °C) shows a slight red shift around 495 nm, indicating its meaningful photoresponse to visible light. To exactly estimate the optical band gap of each film, Tauc plots (inset of Fig. 3(a)) were calculated using the following eqn (2):28

 
αhν = A(Eg)n(2)
where α is the absorption coefficient; h is the Planck constant (eVs); A is a constant; Eg is the band gap energy; and BiNbO4 film has a direct band gap, providing 2 of n herein for direct transition. The optical band gap is estimated to be 2.65 and 2.56 eV for the BNO (500 °C) and BNO (550 °C) films, respectively. The slight narrowing of the band gap in the BNO (550 °C) film is due to the unique nanomaterial physical properties, e.g. large surface-to-volume ratio and nanosheet structures showed higher absorption, a light trapping effect to increase the optical path of photons inside the nanostructures formed by the scattering effect. As well as, the interaction between light and the photoelectrode is enhanced with an oriented nanostructure.


image file: d0cp02071k-f3.tif
Fig. 3 (a) UV-Vis absorption spectra, (b) UPS valence band spectra, (c) secondary electron cut-off region of the BNO (500 °C) and BNO (550 °C) films and (d) the representative band alignment diagram of the BNO (550 °C) photoanode.

The electronic band position of the valence band (VB) and the conduction band (CB) was certified using the ultraviolet photoelectron spectra (UPS).29Fig. 3(b) and (c) show the valence band spectra and the secondary electron cut-off region (Ecut-off) of the BNO films. Combining the band gap obtained from the UV-vis spectra (Fig. 3(a)) and the valence band spectra (Fig. 3(b)), the work function (Φ), the locations of Ef, the valence band maximum (VBM), and the conduction band minimum (CBM) of the BNO films can be obtained. In general, the work function is a surface characteristic rather than a property of the bulk material, and the surface work function depends on various factors, such as the morphology, composition, and contamination in the surface region.30 The work function is generally derived by subtracting the binding energy (BE) of the Ecut-off onset from the He(I) excitation energy (21.2 eV) as the light source. Ecut-off for both the BNO (500 °C) and BNO (550 °C) films can be extrapolated from the linear part intercept to the x-axis, as shown in Fig. 3(c).31 The calculated work functions of BNO (500 °C) and BNO (550 °C) are 4.73 and 4.94 eV, respectively. Note that the work function may significantly affect the magnitude of the band bending. That is, the upshift of Ef in the BNO (550 °C) film that was achieved probably resulted from the increased free carrier density, and the formation of a more favorable energy level. This can be deduced to show efficient electron extraction with the suppressed recombination loss, and to finally enhance the charge transport ability. The changes in the valence band maximum of the films were derived from the UPS data using the equation:32

 
VBM = − (Ecut-offEonset)(3)
where, is equal to 21.2 eV, Ecut-off is the secondary electron cut-off value and Eonset is the valence band onset value obtained from Fig. 3(b and c). The obtained results show that the VBM of the BNO (550 °C) film has a positive up-shift with respect to the vacuum scale representing the bandgap narrowing.

Furthermore, the distances between the VBM and Ef of the BNO (500 °C) and BNO (550 °C) films are calculated to be 2.7 and 2.4 eV, respectively; and consequently, the CBM energies could be calculated to be −0.05 and −0.16 eV by the subtraction of the optical bandgaps of the BNO (500 °C) and BNO (550 °C) films, respectively. The calculated CBM of the BNO films is advantageous to form an appropriate energy level for PEC water reduction. Fig. 3(d) shows the representative band diagram of the BNO (550 °C) film, revealing the proper band alignment for the PEC water oxidation/reduction reaction.

High-resolution X-ray photoelectron spectroscopy was performed to analyze the state surface elemental composition in the BNO (500 °C) and BNO (550 °C) films, as shown in Fig. 4(a–c). The Bi 4f core-level XPS spectrum in Fig. 4(a) shows two peaks positioned at 158.78 and 164.08 eV, corresponding to Bi 4f7/2 and Bi 4f5/2 for BNO (500 °C), whereas these peaks are identified at around 159.09 and 164.34 eV, in the BNO (550 °C) film. The BE difference of 5.3 eV confirms that Bi exists in the trivalent (3+) oxidation state in both samples.33 The XPS spectrum of Nb 3d (Fig. 4(b)) displays well-resolved peaks of Nb 3d5/2 and Nb 3d3/2 of the BNO (500 °C) film at 205.51 and 209.28 eV, respectively, and those of the BNO (550 °C) film at 206.70 and 209.46 eV, respectively, indicating the existence of Nb5+ in the film.34 Also, the spin-energy separation of 2.73 eV between the Nb 3d5/2 and Nb 3d3/2 peaks confirms the pentavalent state (5+). Note that in Fig. 4(a and b), the BNO (550 °C) film shows a higher BE shift, indicating that metal ions are connected to a more electronegative O atom, or there is different chemical bonding between the metal ions, due to the different annealing environments, and the effective interaction between the Bi–O–Nb chemical bond environments as well. After deconvolution, the O 1s core level spectra (Fig. 4(c)) can be fitted into two individual peaks at 529.62 and 530.10 eV for the BNO (500 °C) and BNO (550 °C) films, which correspond to lattice oxygen (O2−) and oxygen vacancies (OV) respectively. In the higher annealing environment, the lattice O content receives a large amount of OV. The higher proportion of OV is advantageous for efficient charge transport via higher donor density charge carriers, improving the PEC performance. Additionally, according to the element content analysis of XPS, the atomic ratio of 14[thin space (1/6-em)]:[thin space (1/6-em)]15[thin space (1/6-em)]:[thin space (1/6-em)]60 is approximately equal to 1[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]4, indicating the confirmation of the BiNbO4 formation.


image file: d0cp02071k-f4.tif
Fig. 4 XPS spectra of (a) Bi 4f, (b) Nb 3d, and (c) O 1s of the BNO (500 °C) and BNO (550 °C) photoanode films.

Fig. 5 shows the photocurrent–voltage (JV) curves of the BNO (500 °C) and BNO (550 °C) photoanodes under the full sun light or a UV-cutoff filter (>420 nm), depending on the presence or the absence of hole scavenger containing electrolytes. Fig. 5a shows the linear sweep voltammograms (LSVs) of the BNO (500 °C) and BNO (550 °C) photoanodes in 1 M NaOH (pH = 13.5) solution at a scan rate of 20 mV s−1. The photocurrent is found to be generated in the anodic direction, confirming the n-type semiconducting properties for the PEC water oxidation. The J value of the BNO (550 °C) film was significantly higher than that of the BNO (500 °C) film throughout the entire scanned potential region. The onset potential (Von) of the BNO films was calculated from which the photocurrent is generated to be 0.34 and 0.29 V for the BNO (500 °C) and BNO (550 °C) photoanodes, respectively, revealing that the BNO (550 °C) film facilitates the survival of photogenerated charges, enabling the driving of efficient water oxidation kinetics in the low potential region. Also, the achievable J is 0.26 and 0.45 mA cm−2 at 1.23 VRHE for the BNO (500 °C) and BNO (550 °C) photoanode, respectively. These results demonstrate that the annealing temperature significantly affects the crystallinity and light absorption of the BNO films, and in particular, increases the J values of the BNO (550 °C) film. Meanwhile, the obtained JV curves of both samples show a similar shape, even in the low potential region where relatively low J values were obtained, whereas scanning the positive potential from 0.7 VRHE, the J values show an abrupt increase. This means that the relatively high resistance in the low potential region exists in the films, mainly caused by incomplete crystallization or the presence of defect or trap sites, inducing the slow interfacial hole transfer kinetics of the photoelectrodes.


image file: d0cp02071k-f5.tif
Fig. 5 (a) Linear sweep voltammograms curves, (b) chopped LSV curves of the BNO (500 °C) and BNO (550 °C) films with and without hole scavenger containing electrolytes under 1 sun illumination, (c) the typical LSV curves of the BNO (500 °C) and BNO (550 °C) films with and without hole scavenger containing electrolytes under a UV-cutoff filter, and (d) chopped LSV curves of the BNO (500 °C) and BNO (550 °C) films with and without hole scavenger containing electrolytes under a UV-cutoff filter.

Hence, the factors affecting the J values at these potentials are examined in the hole scavenger containing electrolyte. Herein, the electrolyte composed of 1 M NaOH solution including the H2O2 solution as a hole scavenger was explored, and Fig. 5(c) shows the results. Both BNO films exhibit an obvious improvement over the measured potential range, and the reasonable J values of 0.3 and 0.47 mA cm−2 at 1.23 VRHE for the BNO (500 °C) and BNO (550 °C) films, respectively, were obtained. These results signify that the water oxidation kinetics at the surface was dramatically improved with the aid of a fast surface charge transfer mediator, minimizing the surface carrier's recombination process. In particular, its dramatic effect occurred in the low potential region, and the enhancement of PEC activity can be achieved via interfacial engineering.

Fig. 5(b) represents the J–V curves of the BNO photoelectrodes obtained under the same conditions in the on–off transmittance mode with a constant interval of 2 s. This indicates that the BNO films do not experience any detrimental effects or undergo self-oxidation processes under a strong alkaline electrolyte, and show agreeable J values of the LSVs (Fig. 5(a)), which confirms that the photocurrents are generally stable without photoinduced charging effects through the scanned potential range, and the fast charge generation/decay resulted from the fast reaction kinetics under the negligible dark current was in agreement. To understand the extent of the visible-light induced J values of the BNO films, the LSVs were acquired under the same measurement conditions, except that the ultra-violet illumination was blocked using a UV-cutoff filter (>420 nm). It is noteworthy that Von shifted to a more negative value under visible light illumination, and the anodic sharp photocurrent spike is totally eliminated for all samples. This means that the production of more stable and long-lived photogenerated charges can be comparable to full sun-based photogenerated charges. Moreover, the photocurrents steadily increase with the applied potential, fulfilling the J values of 0.05 and 0.45 mA cm−2 at 1.23 VRHE for the BNO (500 °C) and BNO (550 °C) films, respectively.

The contribution of the photocurrent generated in visible light to the total photocurrent is about 15%, and it needs to be considered that the cut-off wavelength (<420 nm) contains near visible light. Furthermore, the JV curves of the hole-scavenger containing electrolyte are shown together in Fig. 5(c), revealing the double enhancement of J values for the entire scanned potential region, indicating that the interfacial issue is also important, even in visible light. From the chopped on/off LSV under visible light, the stable photocurrents with negligible dark currents were recorded in the scanned potential region for the same J values on the LSV curve (Fig. 5(c)), implying the fast reaction kinetics.

Fig. 6(a) shows the IPCE of the BNO (500 °C) and BNO (550 °C) films, to calculate the external and internal quantum efficiency, at a potential of 1.23 VRHE in 1 M NaOH solution. In general, IPCEs were measured under monochromatic light irradiation, and plotted as a function of the wavelength. The IPCE value can be described by the following eqn (4):35

 
IPCE (%) = (1240 × J)/(λ × Plight)(4)
where, J is the photocurrent density (mA cm−2), and λ is the wavelength of the illuminating light. In the case of the BNO (550 °C) film, IPCE of approximately 50% was observed at 350 nm, whilst only 25% was detected for the BNO (500 °C) film at the same wavelength. In the visible light region, the IPCE value for the BNO (500 °C) and BNO (550 °C) electrodes is 6 and 15% at 425 nm, respectively. This result indicates that the electron–hole pairs excited in BNO (550 °C) can be quickly separated and collected more efficiently than those in BNO (500 °C), in agreement with the values obtained from the J–V curves in Fig. 5(a), thanks to the higher conduction band edge and high crystallinity of BNO (550 °C), in comparison with those of BNO (500 °C).

Furthermore, the interfacial electrical characteristics of the BNO (500 °C) and BNO (550 °C) films were surveyed by Mott–Schottky (M–S) analysis, as shown in Fig. 6(b). Herein, the M–S measurements were performed in 1 M NaOH solution at different applied AC frequencies of 100 and 250 Hz under the dark conditions, to verify that the slope and intercept do not change. To exactly calculate the quantitative EFB and ND values, the following eqn (5) and (6) can be adapted:36,37

 
image file: d0cp02071k-t1.tif(5)
 
image file: d0cp02071k-t2.tif(6)


image file: d0cp02071k-f6.tif
Fig. 6 (a) IPCE spectra, (b) Mott–Schottky plots, (c) EIS spectra, and (d) open circuit photovoltage curves of the BNO (500 °C) and BNO (550 °C) films.

From Fig. 6(b), the positive slopes were achievable, representing an n-type semiconducting property, and the estimated Efb values were about 0.3 and 0.22 V for BNO (500 °C) and BNO (550 °C) films, respectively, consistent with the Efb values calculated from the UPS spectra. Furthermore, the steepness of the slope in the curve can determine the carrier densities (ND), closely associated with the electronic conductivity in the film. From the slope of the respective M–S plot obtained using eqn (6), ND is calculated to be approximately 1.02 × 1022 and 6.85 × 1023 cm−3 for the BNO (500 °C) and BNO (550 °C) films, respectively, revealing the much higher free carrier densities in the BNO (550 °C) film. A higher ND may cause a larger up-shift of the local Efb, resulting in a significant upward band bending at the surface, facilitating the fast charge separation at the solid/electrolyte interface. Furthermore, the increased carrier densities above one order of magnitude can promote the electronic conductivity in terms of charge transportation, increasing the J value for the PEC performance.

Moreover, in order to intimately survey the interfacial working conditions between the BNO film and electrolyte, the EIS measurements under AM 1.5G illumination at open-circuit potential (OCP) were carried out to provide an intimate understanding of the charge transfer phenomenon. In general, a photoanode presents a minimum radius of the semicircle in a high-frequency region, which means that the surface recombination of photoproduced holes is greatly reduced. The Nyquist plots as exhibited in Fig. 6(c) were fitted by an equivalent circuit model in the inset of Fig. 6(c), and the fitted curve is marked as a solid line in Fig. 6(c). This model includes the series resistance (RS) between the FTO substrate and the photoanode, charge transfer resistance at the electrode–electrolyte interface (RCT) associated with the hole transfer process from the surface states, and constant phase element (CPE). The (RS and RCT) values of the BNO (500 °C) and BNO (550 °C) films are estimated from the fitted results to be (150, 2900) and (95, 1780) Ω, respectively. The great reduction of photo-induced charge transfer resistance across the BNO (550 °C)/electrolyte interface also promotes the survival probability of the photogenerated holes.

Furthermore, for more insights into the surface charge transfer phenomenon, the OCP decay was measured as a function of time as soon as the light illumination is abruptly blocked, as shown in Fig. 6(d). Ideally, for n-type semiconducting materials, the OCP will be shifted towards a more negative potential with respect to the RHE under the illumination. In the experiment, the photovoltage (Vph) estimated through OCPdark–OCPlight is 0.26 and 0.36 V for the BNO (500 °C) and BNO (550 °C) films, respectively. This shows that the nanostructured BNO (550 °C) film eliminates the Ef pinning caused by the hole trapping state, which directly induces the higher Vph and the greater band bending at the photoanode/electrolyte interface, thus effectively restraining the surface carrier recombination, and accelerating the surface charge transfer capability. Furthermore, the BNO (500 °C) film exhibits low photovoltage due to the high surface recombination rates, which might be caused by the high bulk recombination rates coming from the highly disordered states.38 The average lifetimes of each Vt profile by fitting to a biexponential function with two-time constants were calculated using the following eqn (7) and (8):39

 
y(t) = A0 + A1et/τ1 + A2et/τ2(7)
 
τn = (τ1τ2)/(τ1 + τ2)(8)
where, τn is the average lifetime, and the total half-life is log(2 × τn). τn of the photocarrier is estimated to be 0.28 s and 1.24 s for BNO (500 °C) and BNO (550 °C), respectively. The BNO (550 °C) film holds a higher OCP and exhibits a slower Vph decay behavior, corresponding to a longer average lifetime of the photoexcited carrier. This indicates the highly efficient formation of the photogenerated charges against the fast charge recombination surrounding the BNO (550 °C) film.

To understand the surface area effect related to the charge transfer process, the PEC behavior was further assessed with the measurement of the electrochemically active surface area (ECSA). Herein, the capacitive currents were measured in the non-faradaic region through cyclic voltammetry (CV) analysis in the potential range of 0.2 to 0.4 VRHE. The CVs were measured in the 1 M NaOH electrolyte at different applied scan rates from 25 to 200 mV s−1 under dark conditions. Moreover, from Fig. 7(a), the anodic and cathodic capacitive current differences (ΔJ) at a fixed potential of 0.29 VRHE were plotted as a function of the scan rate (Fig. 7(b)), and the double layer capacitance (Cdl) calculated from the slope of the linear plot can be directly correlated with the ECSA, or the number of active sites on the surfaces of photoanodes. Finally, the ECSA can be calculated from the double-layer capacitance according to the following eqn (9):40,41

 
ic ∝ (v × AECSA)(9)
where the capacitive current (ic) is proportional to the product of the scan rate (v) and the electrochemical active surface of the electrode. Hence, a plot of ic as a function of v led to a straight line with a slope corresponding to Cdl. The larger slope of BNO (550 °C) indicates that the double-layer capacitance at the interface between the BNO (550 °C) nanosheets and electrolyte was larger, due to the larger contact area. The Cdl values for the BNO (500 °C) and BNO (550 °C) films are quantitatively calculated to be 0.61 and 12 μF cm−2, respectively. Cdl of BNO (550 °C) is almost 20-fold higher than that of BNO (500 °C), which indicates that the BNO (550 °C) nanosheet morphology exhibits a larger surface area, which will enable efficient charge transfer pathways for a higher PEC reaction through the nanostructured film surface. This result is in agreement with the results of interfacial electrical characteristics obtained from M–S and EIS shown in Fig. 6(b and c) in the previous section. In addition, it can be concluded that each of the factors involved in the overall PEC system can be improved, such as (1) the strong crystallinity with the nanosheet morphology of a higher electro-active surface area, (2) the higher band bending that occurred via the effective formation of carrier density, and lower resistive interfacial working conditions for fast reaction kinetics, respectively.


image file: d0cp02071k-f7.tif
Fig. 7 (a) Cyclic voltammetry recorded in the non-faradaic region under dark conditions, and (b) the double layer capacitance (Cdl) slope of the BNO (500 °C) and BNO (550 °C) films.

Finally, to examine the photostability of the BNO based PEC cells, the time-dependent photoresponse was measured under the constant illumination at 1.23 VRHE, and Fig. 8 summarizes the results. Usually, the photoelectrode stability is mainly affected by the surface defect density. In particular, the surface bonds and reaction intermediates can be considered as the surface states.42 Sometimes, the surface atoms of unstable weak bonds may lead to chemical decomposition. Additionally, the absorption of electrolyte species and the formation of the intermediate free radicals substantially influence the photogenerated hole transfer rate in the surface interfacial region. In this PEC system, it can be found that an enhanced photostability in the BNO (550 °C) film indicates that the efficient hole extraction can possibly influence the stability of the PEC devices. However, the initial photocurrent degradation until 30 min steadily occurred, afterwards reaching a stable photocurrent value. After completing the stability test in the BNO (550 °C) film, the morphological modification was checked, shown as the inset of Fig. 8. The original morphology of the BNO (550 °C) film can be preserved even after PEC measurements over 5 h, disclosing the strong structural rigidity. Additionally, the interesting and emerged photoelectrodes are summarized and their PEC behaviour is listed in Table 1.


image file: d0cp02071k-f8.tif
Fig. 8 Amperometric (it) curve of the BNO (500 °C) and BNO (550 °C) films, and the inset FE-SEM image was taken after the stability measurement.
Table 1 Summary of the preceding literature survey of emerging photoelectrodes
Sample Substrate/fabrication Electrolyte/pH PEC (mA cm−2) at 1.23 VRHE and IPCEmax (%) Stability Ref.
β-Cu2V2O7 FTO/spray pyrolysis 0.3 M K2SO4 + 0.2 M phosphate buffer (pH = 6.8) 0.05 and 1.2% 2 h 19
CaBi2O4 FTO/hydrothermal process 0.25 M Na2SO3 + 0.1 M Na2SO4 1.19 and 5% 1 h 43
Bi2WO6 FTO/template method 1 M Na2SO4 (pH = 6.6) 0.1 250 s 18
Fe2WO6 FTO/electrophoretic deposition 0.1 M KOH 0.1 and 9% 44
CaBi2Nb2O9 Molten salt method 0.1 M Na2SO4 O2 content: 29.8 μmol h−1 g−1 10 h 45
Bi3TaO7 FTO/hydrothermal process 0.1 M Na2SO4 (pH = 7.1) 0.1 1.5 h 23
Fe2TiO5 FTO/electrospray deposition 1 M NaOH (pH = 13.6) 0.25 46
BiNbO4 FTO/spin-coating 1 M NaOH (pH = 13.6) 0.45 and 50% 5 h This work


Conclusions

In summary, we developed BiNbO4 nanosheet films on the FTO substrate by a facile spin coating method. Depending on the post-annealing temperature, the BNO films exhibited remarkably unique characteristics. First of all, compared with the BNO (500 °C) film, the BNO (550 °C) film disclosed enhanced crystallinity in the same crystal structure of the phase-pure orthorhombic α-BiNbO4 phase, as well as increased light absorption. The UPS analysis showed that the variations in the work function significantly altered the magnitude of the band-bending height, and resulted in the upshift of the Fermi level, forming a more favorable band diagram, finally resulting in efficient electron extraction with suppressed recombination loss in the BNO (550 °C) film. On the basis of these properties, the J values obtained were more than two times higher in the BNO (550 °C) film, compared to those in the BNO (500 °C) film at 1.23 VRHE, accompanying a ∼50 mV cathodic shift in Von. To understand the interfacial issue, the hole-scavenger containing electrolyte was tested to prove the considerable interfacial resistances or troubles. More importantly, the prolonged photocurrent stability of the BNO (550 °C) film as a function of time evidenced the structural and mechanical rigidity of BiNbO4 itself for a PEC cell. For more enhanced J values, the introduction of a heterojunction for efficient charge separation or a selective hole acceptor co-catalyst to improve the slow kinetics of water oxidation would be suggested in the near future.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) (grant number: 2019R1A2C1007637). This work was supported by the Priority Research Centers Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science and Technology (2018R1A6A1A03024334). The authors are grateful to the Center for Research Facilities at the Chonnam National University for their assistance in the analysis of the BNO films (FE-SEM and XRD).

References

  1. S. Bonke, M. Wiechen and D. R. MacFarlane, Energy Environ. Sci., 2015, 8, 2791–2796 RSC.
  2. M. Grätzel, Nature, 2001, 414, 338–344 CrossRef PubMed.
  3. J. R. McKone, N. S. Lewis and H. B. Gray, Chem. Mater., 2014, 26, 407–414 CrossRef CAS.
  4. T. Arai, Chem. Commun., 2010, 46, 6944–6946 RSC.
  5. Y. Zhao, Y. Zhao, R. Shi, B. Wang, G. I. N. Waterhouse, L. Z. Wu, C. H. Tung and T. Zhang, Adv. Mater., 2019, 1806482 CrossRef PubMed.
  6. S. Chu, W. Li and Y. Yan, Nano Futures, 2017, 1, 022001 CrossRef.
  7. J. Li and N. Wu, Catal. Sci. Technol., 2015, 5, 1360–1384 RSC.
  8. H. Dotan, K. Sivula and M. Grätzel, Energy Environ. Sci., 2011, 4, 958–964 RSC.
  9. S. Wang, H. Chen and G. Gao, Nano Energy, 2016, 24, 94–102 CrossRef CAS.
  10. Y. Zhang, D. Wang, X. Zhang, Y. Chen, L. Kong, P. Chen, Y. Wang, C. Wang, L. Wang and Y. Liu, Electrochim. Acta, 2016, 195, 51–58 CrossRef CAS.
  11. J. W. Ager, M. R. Shaner, K. A. Walczak, I. D. Shar and P. S. Ardo, Environ. Sci., 2015, 8, 2811–2824 CAS.
  12. P. Peerakiatkhajohn, J. H. Yun, H. Chen, M. Lyu, T. Butburee and L. Wang, Adv. Mater., 2016, 28, 6405–6410 CrossRef CAS PubMed.
  13. M. Hannula, H. A. Löytty, K. Lahtonen, E. Sarlin, J. Saari and M. Valden, Chem. Mater., 2018, 30, 1199–1208 CrossRef CAS PubMed.
  14. S. Hu, M. R. Shaner, J. A. Beardslee, M. Lichterman, B. S. Brunschwig and N. S. Lewis, Science, 2014, 344, 1005–1009 CrossRef CAS PubMed.
  15. W. Li, K. Jiang, Z. Li, S. Gong, R. L. Z. Hoye, Z. Hu, Y. Song, C. Tian, J. Kim, K. H. L. Zhang, S. Cho and J. L. MacManus-Driscoll, Adv. Energy Mater., 2018, 8, 1801972 CrossRef.
  16. X. Zhu, N. Guijarro, Y. Liu, P. Schouwink, R. A. Wells, F. L. Formal, S. Sun, C. Gao and K. Sivula, Adv. Mater., 2018, 1801612 CrossRef PubMed.
  17. M. Kölbach, I. Jordão Pereira, K. Harbauer, P. Plate, K. Höflich, S. P. Berglund, D. Friedrich, R. Krol and F. F. Abdi, Chem. Mater., 2018, 30, 8322–8331 CrossRef.
  18. L. Zhang and D. Bahnemann, ChemSusChem, 2013, 6, 283–290 CrossRef CAS PubMed.
  19. A. Song, A. Chemseddine, I. Y. Ahmet, P. Bogdanoff, D. Friedrich, F. F. Abdi, S. P. Berglund and R. Krol, Chem. Mater., 2020, 32(6), 2408–2419 CrossRef CAS.
  20. Y. Gao, Y. Huang, Y. Li, Q. Zhang, J. Cao, W. Ho and S. C. Lee, ACS Sustainable Chem. Eng., 2016, 4, 6912–6920 CrossRef CAS.
  21. Y. L. Min, F. J. Zhang, W. Zhao, F. C. Zheng, Y. C. Chen and Y. G. Zhang, Chem. Eng. J., 2012, 209, 215–222 CrossRef CAS.
  22. G. Tang, H. Zhu, H. Yu, X. Cheng, R. Zheng, T. Liu, J. Zhang, M. Shui and J. Shu, J. Electroanal. Chem., 2018, 823, 245–252 CrossRef CAS.
  23. Q. Song, P. Wu, S. Sarkar, Y. Zhao and Z. Liu, Dalton Trans., 2010, 49, 147 RSC.
  24. H. Zhang, K. Kim, H. Young, J. Jae and S. Lee, ACS Catal., 2019, 2, 1289–1297 CrossRef.
  25. B. C. Wang, J. Nisar, B. Pathak, W. Kang and R. Ahuj, Phys. Lett., 2012, 100, 182102 Search PubMed.
  26. F. Litimein, R. Khenata, S. K. Gupta, G. Murtaza, A. H. Reshak, A. Bouhemadou, S. B. Omran, M. Yousaf and P. K. Jha, J. Mater. Sci., 2014, 49, 7809–7818 CrossRef CAS.
  27. B. Y. AlfaifiaAsif, A. Tahira and K. G. UpulWijayantha, Sol. Energy Mater. Sol. Cells, 2019, 195, 134–141 CrossRef.
  28. M. Arunachalam, G. Yun, K. S. Ahn, D. S. Jung and S. H. Kang, Int. J. Hydrogen Energy, 2018, 43, 16458–16467 CrossRef CAS.
  29. W. J. Chun, A. Ishikawa, H. Fujisawa, T. Takata, J. N. Kondo, M. Hara, M. Kawai, Y. Matsumoto and K. Domen, J. Phys. Chem. B, 2003, 107, 1798–1803 CrossRef CAS.
  30. A. Kahn, Mater. Horiz., 2016, 3, 7–10 RSC.
  31. J. Liu, J. Zhang, D. Wang, D. Li, J. Ke, S. Wang, S. Liu, H. Xiao and R. Wang, ACS Sustainable Chem. Eng., 2019, 7, 12428–12438 CAS.
  32. H. Lin, X. Long, J. Hu, Y. Qiu, Z. Wang, M. Ma, Y. An and Y. Yang, ACS Appl. Mater. Interfaces, 2018, 10(13), 10918–10926 CrossRef CAS PubMed.
  33. N. Myung, S. Ham, S. Choi, Y. Chae, W. G. Kim, Y. J. Jeon, K. J. Paeng, K. Chanmanee, N. R. de Tacconi and K. Rajeshwar, J. Phys. Chem. C, 2011, 115, 7793–7800 CrossRef CAS.
  34. B. Luo, Y. Hong, D. Li, Z. Fang, Y. Jian and W. Shi, ACS Sustainable Chem. Eng., 2018, 6(11), 14332–14339 CrossRef CAS.
  35. Y. Gun, B. Maheswari, H. S. Kim, K. S. Ahn and S. H. Kang, J. Phys. Chem. C, 2016, 120, 5906–5915 CrossRef.
  36. W. S. Santos, M. Rodriguez, J. M. O. Khoury, L. A. Nascimento, R. J. P. Ribeiro, J. P. Mesquita, A. C. Silva, F. G. E. Nogueira and M. C. Pereira, ChemSusChem, 2018, 11, 589–597 CrossRef PubMed.
  37. M. Chhetri, S. Dey and C. N. R. Rao, ACS Energy Lett., 2017, 2, 1062–1069 CrossRef CAS.
  38. M. Arunachalam, G. Yun, K. S. Ahnn and S. H. Kang, J. Phys. Chem. C, 2018, 122, 9255 CrossRef.
  39. K. H. Ye1, H. Li, D. Huang, S. Xiao, W. Qiu, M. Li, Y. Hu, W. Mai, H. Ji and S. Yang, Nat. Commun., 2019, 10, 3687 CrossRef PubMed.
  40. A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, New York, 2001 Search PubMed.
  41. J. D. Benck, Z. Chen, L. Y. Kuritzky, A. J. Forman and T. F. Jaramillo, ACS Catal., 2012, 2, 1916–1923 CrossRef CAS.
  42. F. Nandjou and S. Haussener, ChemSusChem, 2018, 12, 984–1994 Search PubMed.
  43. X. Wang, Z. Li and Z. Liu, ChemCatChem, 2017, 9, 4029–4034 CrossRef CAS.
  44. F. F. Abdi, A. Chemseddine, S. P. Berglund and R. Krol, J. Phys. Chem. C, 2017, 121, 153–160 CrossRef CAS.
  45. Y. Zhang, J. Yuan, H. Gong, Y. Cao, K. Liu, H. Cao, H. Yan and J. Zhu, ACS Sustainable Chem. Eng., 2018, 6, 3840–3852 CrossRef CAS.
  46. S. Kuang, M. Wang, Z. Geng, X. Wu, Y. Sun, W. Ma, D. Chen, J. Liu, S. Feng and K. Huang, ACS Sustainable Chem. Eng., 2019, 17, 14347–14352 CrossRef.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cp02071k

This journal is © the Owner Societies 2020