Heterojunction photoanodes for solar water splitting using chemical-bath-deposited In₂O₃ micro-cubes and electro-sprayed Bi₂WO₆ textured nanopillars

Abstract

In₂O₃ micro-cubes were fabricated via chemical bath deposition (CDB) on an indium tin oxide (ITO) substrate, over which highly textured Bi₂WO₆ nanopillars were grown via a diffusion-and-aggregation phenomenon by electrostatic spraying deposition (ESD). The resulting heterojunction bilayer exhibited a photocurrent density value of 1.03 mA cm⁻² under visible light of 100 mW cm⁻² intensity. Widening the light absorptivity spectrum and reducing the charge recombination enhanced the performance of the heterojunction bilayer films for solar water splitting.

---

Introduction

Photoelectrochemical (PEC) cells are widely used for hydrogen production via solar water splitting. In order to improve the performance of PEC cells, various semiconductor-based photoelectrodes^1,2 have been developed with high activities in the visible light spectrum. Bi-based compounds, which are photoactive in the visible light range, have been widely used because they are chemically stable and non-toxic.^3,4 Bi is a p-block metal with d¹⁰ electronic configuration; hence, the Bi 6s orbital can hybridize with O 2p levels to form a hybrid valance band that enhances the mobility of photogenerated carriers and increases the photocatalytic activity of the material.⁵ Among various Bi oxides, Bi₂WO₆ has recently attracted attention for the relatively narrow bandgap (2.6–2.8 eV) and unique layered Aurivillius oxide structure,⁶ which efficiently uses visible light for photocatalysis. However, the actual utilization of this material is limited because of the propensity for recombination of charge carriers in devices made with it.⁷

Various approaches have been attempted to overcome the problem of carrier recombination, such as doping,⁷ porous films,⁸ and bilayer structures.⁹ In particular, the use of heterojunction structures can promote the separation of electrons and holes and the absorption of photons, while further reducing the width of the bandgap. Heterojunction bilayer films using Bi₂WO₆ with Bi₂O₃,⁹ TiO₂,¹⁰ WO₃,¹¹ and ZnWO₄ (ref. 12) have been studied for use in the photodegradation of organic dyes and pollutants. However, to the best of our knowledge, the use of these bilayer films for solar water splitting has not yet been studied.

In₂O₃ is also an important p-block d¹⁰ metal oxide,¹³ suitable for photocatalysis and water splitting.¹⁴ In₂O₃ has a bandgap value of E_g ∼ 3.5 eV and has been used in solar cells, photovoltaics, optoelectronic devices, liquid crystal displays, and gas sensors.¹⁵ Recently, some studies regarding the nanostructure of In₂O₃ reported that the bandgap of the material could be lowered to 2.8 eV,^16,17 which widens the absorption range of visible light for maximal water splitting and photocatalysis.

However, the water splitting performance of pure In₂O₃ is not very competitive, yielding a photocurrent density below 1 mA cm⁻² in most cases. For this reason, the heterojunction structure with additional layers of other metal oxides has been commonly studied.^18,19 There are two major advantages accompanying the introduction of the heterojunction bilayer structure. First, the heterojunction can widen the light absorption range by tuning the bandgap of the base material. Second, charge carrier recombination can be reduced by tuning the alignment of the bands. These synergetic effects can enhance the water splitting performance and hence the photoconversion efficiency.²⁰ Liu et al.²¹ fabricated a bilayer film of In₂O₃/Bi₂WO₆ using electrospinning, but the application of the film was limited to the photodegradation of methyl-orange dye.

Herein, we demonstrate the fabrication of In₂O₃ micro-cubes via chemical bath deposition (CBD) on an indium tin oxide (ITO) substrate. Highly textured Bi₂WO₆ nano-pillars were grown on the surfaces of the In₂O₃ micro-cubes, creating an In₂O₃/Bi₂WO₆ heterojunction film. The process parameters were optimized to yield a heterojunction film with the best performance for water splitting.

Experimental methods

Precursors

Indium(III) nitrate (In(NO₃)₃, 98%, Sigma-Aldrich), urea (NH₂CONH₂, 99%, Samchun Chemical), bismuth(III) nitrate pentahydrate (BiN₃O₉·5H₂O, ≥98%, Sigma-Aldrich), and tungsten(VI) ethoxide (C₁₂H₃₀O₆W, 99.8%, 5% w/v in ethanol, Alfa Aesar), were used as precursors. Deionized (DI) water and propylene glycol (PG, C₃H₈O₂, 99.0%, Duksan Chemical) were used as solvents for the deposition of In₂O₃ and Bi₂WO₆ films, respectively. All chemicals were used without further purification.

Deposition of In₂O₃

The In₂O₃ films were formed by chemical bath deposition (CBD). Fifty mL of solution was prepared by mixing aqueous solutions of 0.05 M In(NO₃)₃ and 0.15 M urea in equal volumes. This solution was placed in a falcon tube; ITO-coated glass of size 2.5 × 2.5 cm² was immersed in this tube and the tube was sealed. Thereafter, the tube was kept in a water bath maintained at 80 °C. Prior to the deposition of the films, the ITO-coated glass substrates were cleaned by ultrasonication in an ethanol bath and then dried. A schematic of the CBD process is shown in Fig. 1. The deposition of In₂O₃ films was performed for different deposition times as listed in Table 1. The films were air-annealed at 450 °C for 30 min to obtain crystallized and impurity-free In₂O₃ films.

Fig. 1 Schematics of CBD and ESD used for deposition of In₂O₃ and Bi₂WO₆, respectively.

Table 1 Operating conditions for CBD and ESD

Parameters	Conditions
	CBD	ESD
Substrate	ITO	In2O3 films on ITO
Nozzle to substrate distance [cm]	—	4.5
Bath temperature [°C]	80	—
Substrate temperature [°C]	—	120
Flow rate [μl h^−1]	—	40
Deposition time	8, 12, 24, 36 [h]	20, 40, 60, 80, 100, 120 [min]
Annealing temperature [°C]	450	500
Annealing time [min]	30	10

---

Deposition of Bi₂WO₆

The Bi₂WO₆ films were formed by electrostatic spray deposition (ESD). The spray solution was prepared by mixing 0.1 M BiN₃O₉·5H₂O in PG and adding an amount of C₁₂H₃₀O₆W necessary to maintain a 2 : 1 Bi : W ratio. The solution was stirred at a room temperature of 25 °C and appeared to be transparent with no precipitation for one week. The solution was further atomized by electrical forces to deposit the Bi₂WO₆ films. A schematic of ESD is shown in Fig. 1; the operating conditions used for the deposition of the films are presented in Table 1. The other details of ESD were reported in earlier papers.^22,23 The bilayers were formed by electro-spraying the Bi₂WO₆ films for 20, 40, 60, 80, 100, and 120 min (henceforth Bi20, Bi40, Bi60, Bi80, Bi100, and Bi120) on CBD In₂O₃ films synthesized for different times of 8, 12, 24, and 36 h (henceforth In8, In12, In24, and In36).

The surface morphologies of the fabricated films were investigated using scanning electron microscopy (SEM, S-5000, Hitachi, Ltd). The crystalline structures of the films were characterized by X-ray diffraction (XRD, Rigaku Japan, D/MAX-2500, with Cu Kα X-rays). X-ray photoelectron spectroscopy (XPS, PHI5600, Physical Electronics) measurements were carried out to evaluate the chemical states of the films.

Photoelectrochemical measurements

Photocurrent measurements were performed under backside illumination and in a single-cell setup with a three-electrode configuration. In the setup, the bilayer film of the 0.5 × 0.5 cm² irradiated area was the working negative electrode (anode), Ag/AgCl was the reference electrode, and Pt wire was the counter positive electrode (cathode). All electrodes were positioned near each other; the distances between them were maintained for all measurements. The electrolyte was a solution of 1 M KOH. Artificial sunlight from a Xe arc lamp (Newport, Oriel Instruments, USA) was used to illuminate the photoelectrode at a light intensity of 100 mW cm⁻². An AM 1.5 filter was used to provide light with an intensity equivalent to one sun.

Results and discussion

The X-ray diffraction spectra of pristine In₂O₃, Bi₂WO₆, and a bilayer of In₂O₃/Bi₂WO₆ are presented in Fig. 2. The XRD pattern in Fig. 2(a) clearly corresponds to the cubic In₂O₃ crystal structure. The diffraction peaks appearing at 21.52, 30.61, 35.59, 51.08, and 60.5° correspond to the (221), (222), (400), (440), and (622) planes, respectively, in good agreement with the JCPDS file 06-0416 for cubic In₂O₃. The strong diffraction peaks in Fig. 2(b) at 2θ = 28.36, 32.84, 47.10, and 57.12° correlate to the orthorhombic crystal planes of Bi₂WO₆; the pattern matches well with the JCPDS file 39-0256. Fig. 2(c) of the bilayer film clearly shows that the In₂O₃/Bi₂WO₆ heterostructure is well-integrated, reflecting phases of both orthorhombic Bi₂WO₆ and cubic In₂O₃. No peaks corresponding to traces of metal compounds mixed with Bi, In, and O are detected in the XRD patterns. Thus, the spectrum confirms that the In₂O₃/Bi₂WO₆ heterostructure is composed of separate phases of Bi₂WO₆ and In₂O₃. The ITO substrate peaks overlap with the peaks of In₂O₃.

Fig. 2 X-ray diffraction patterns of (a) pristine In₂O₃, (b) pristine Bi₂WO₆, and (c) In₂O₃/Bi₂WO₆ bilayer.

The SEM images of pristine In₂O₃ films deposited by CBD shows the cubic morphology of the material, as seen in Fig. 3(a) and (d). Pillar-like structures (Fig. 3(b)) are observed in the cross-sectional image of the pristine Bi₂WO₆ film deposited by ESD, whereas the surface image of the same material (Fig. 3(e)) has a porous appearance, as the pillar spacing is large.^23,24 Such pillar formation is unique to ESD fabrication, because of the simultaneous diffusive, thermophoretic, and electrophoretic transport of ESD particles. As the film grows, the electrostatic forces parallel to the substrate attract particles to the tips of pillars and repel particles in the valleys between pillars. Similar results were also observed in an earlier examination of BiVO₄ films fabricated by ESD.²⁴ Thus, it is clear that pillar formation by ESD is uniquely characteristic of Bi-based oxides. The cross-sectional SEM image of the bilayer film shows that the surface is composed of many uniform structures of pillars on cubes (Fig. 3(c)), similar to the pristine films. The surface-view SEM image shows the Bi₂WO₆ pillars coated over different cubes of In₂O₃. The image also shows an empty area resulting from repulsion between Bi₂WO₆ pillars, which excluded Bi₂WO₆ particles from filling the empty area, as mentioned above. This unique cube and pillar morphology allows light absorption deep in the In₂O₃ film and increases both the surface area and charge separation.

Fig. 3 Cross-sectional and surface SEM images of (a and d) pristine In₂O₃, (b and e) pristine Bi₂WO₆, (c and f) bilayer In24–Bi80.

The chemical states of constituent elements for the bilayer, investigated by XPS, are presented in Fig. 4. The XPS survey spectrum of the bilayer film, presented in ESI Fig. S1,† confirms the coexistence of In₂O₃ and Bi₂WO₆ in the obtained sample. The binding energies for Bi 4f_7/2 and Bi 4f_5/2, shown in Fig. 4(a), are 159.1 and 164.4 eV, attributed to Bi³⁺ and Bi⁴⁺. The peaks centered at 35.4 and 37.54 eV (Fig. 4(b)) are assigned to W 4f_5/2 and W 4f_7/2 respectively.²⁵ The presence of In³⁺ is confirmed by the peaks at 444.6 and 452.2 eV, corresponding to the 3d_5/2 and 3d_3/2 spin orbits of In shown in Fig. 4(c), which confirms the formation of In₂O₃.²⁶ The peak located at 530.2 eV is assigned to O 1s in Fig. 4(d).

Fig. 4 XPS (a) Bi 4f, (b) W 4f, (c) Si 2p, and (d) C 1s spectra for bilayer In24–Bi80.

The photocurrent value obtained in the bilayer electrode is positive, indicating the formation of an n-type photoanode. Thus, the generated photocurrent correlates directly to the oxygen evolution. To investigate the influence of individual layer thicknesses on the photocurrent generated by In₂O₃/Bi₂WO₆ heterojunction films, linear scan voltammograms were recorded of the different samples. Fig. 5 shows a plot of photocurrent density versus the applied potential for various Bi₂WO₆ thicknesses coated on 24 h CBD-coated In₂O₃. The samples were illuminated with one sun (100 mW cm⁻²) intensity of light. The plot clearly reveals that the photocurrent increases as the Bi₂WO₆ deposition time increases, reaching a maximum at 80 min and thereafter decreases for further increases in coating time. The value of photocurrent density (PCD) measured for the maximum case of In24–Bi80 is 1.03 mA cm⁻², which is almost double than that of pristine In₂O₃ and 20 times greater than that of pristine Bi₂WO₆, as shown in ESI Fig. S2.† The higher photocurrent density of the In24–Bi80 film relative to those of the pristine materials results from the lower-energy bandgap of the bilayer film, which increases the light absorption. In addition, the formation of the In₂O₃/Bi₂WO₆ heterojunction increases the effective electron/hole separation. However, further increase in Bi₂WO₆ spray time above 80 min lowers the photocurrent generated from the bilayer. The increased height of the Bi₂WO₆ pillars increases the transport length of the charge carriers. The electric field also decreases in the bulk and the interface of the bilayers with increasing pillar length, which leads to the recombination of charge carriers. Thus, for higher Bi₂WO₆ deposition times of 100 or 120 min, the photocurrent falls to 0.8 mA cm⁻² and 0.6 mA cm⁻², respectively. The onset potential for all bilayer films is observed to be −0.1 V, which is much lower than that of pristine Bi₂WO₆ (shown in ESI Fig. S2†). The inset image shows the photocurrent at 0.7 V (vs. Ag/AgCl) for different Bi₂WO₆ spray times.

Fig. 5 Current–voltage curve of 24 h-coated In₂O₃ layered with various Bi₂WO₆ thicknesses.

The photocurrent density (PCD) of bilayer films with variation in the thickness of the In₂O₃ base layer is shown in Fig. 6. In all cases, the Bi₂WO₆ was sprayed for 80 min. The PCD value is observed to increase from 8 h to 24 h of In₂O₃ deposition. The maximum PCD achieved is 1.03 mA cm⁻² from the In24–Bi80 bilayer film. However, further increase in the In₂O₃ thickness results in a decrease in the PCD. This decrease results from the recombination process, which becomes dominant after the optimal thickness is exceeded. Thus, films deposited for 36 h have higher thicknesses, which increase the transport path lengths of the photogenerated charge carriers; simultaneously, the applied electric field becomes weak in the bulk material, driving the recombination of electron/hole pairs and reducing the PCD.

Fig. 6 Photocurrent density curve of bilayer with 80 min deposited Bi₂WO₆ on varied thicknesses of In₂O₃.

As shown, the bilayer films exhibit higher photocurrents as compared to the pristine materials. In the pristine films, In₂O₃ and Bi₂WO₆ are limited in light absorption capacity; additionally, photogenerated hole/electron pairs recombine quickly. However, in the bilayer films, better results are observed because of the relative energy band positions of the component semiconductor materials. The bandgap energy values of the pristine Bi₂WO₆ and In₂O₃ films were extrapolated from the Tauc plot, which yielded 2.7 and 2.5 eV, respectively.^27,28 The bandgap energy of the In24–Bi80 bilayer film was 2.4 eV, which is in agreement with the value reported by Liu et al.²¹ This reduced bandgap is favorable to absorb the increased range of the light spectrum. Note that the conduction band (E_cb) and valance band (E_vb) edge positions can be determined by using the empirical formula such as, E_cb = E_vb − E_g and E_vb = X − 4.5 + 0.5E_g, where X is the absolute electronegativity, 4.5 eV is the energy of free electrons on a hydrogen scale.^29,30 As shown in Fig. 7, under the illuminated condition both oxides absorb light to generate electron/hole pairs in the different layers. The conduction band edge potential of In₂O₃ is more negative than that of Bi₂WO₆, which forbids the direct transport of photogenerated electrons from Bi₂WO₆ to In₂O₃. Meanwhile, the valance band maximum of In₂O₃ is close to the conduction band minimum of Bi₂WO₆; thus, the photogenerated holes in In₂O₃ and electrons in Bi₂WO₆ are recombined. Hence, all holes generated in Bi₂WO₆ are used for water oxidation, while the electrons in In₂O₃ are transferred to the counter electrode. In₂O₃/Bi₂WO₆ bilayer films such as those synthesized in this study may provide improved energetics for the reduction of electrons on the counter electrode. Thus, the overall electron/hole separation efficiency is enhanced. Introducing the heterojunction film structure also decreases the onset potential of the photocurrent and increases the degree of charge separation. A similar mechanism is observed in the case of BiVO₄ deposited on an n-type Si substrate.³¹ Thus, the deposited n/n heterojunction bilayer film shows enhancement in photocurrents at the optimized film layer thicknesses.

Fig. 7 Mechanism of charge separation in bilayer film.

The stability of the In24–Bi80 bilayer film was tested using the chronoamperometry technique where the photocurrent was measured at a potential of 0.2 V (vs. Ag/AgCl). In Fig. 8, the photocurrent value of 0.5 mA cm⁻² remains constant for 60 min under one-sun illumination of 100 mW cm⁻². This constant photocurrent value confirms the photostability of the bilayer film. This result is also consistent with the photocurrent obtained using the linear voltammetry scan for the In24–Bi80 film shown in Fig. 6. This consistent and stable photocurrent confirms that Bi₂WO₆ nanopillars on top remains intact with the In₂O₃ layer below even after 1 h light illumination.

Fig. 8 Photocurrent stability curve of bilayer film at a potential of 0.2 V (vs. Ag/AgCl).

Conclusions

In₂O₃/Bi₂WO₆ bilayer films were prepared successfully by utilizing CBD and ESD methods. The deposited films showed unique pillar-coated cubic morphologies with increased surface area. The XRD and XPS spectra confirmed the presence of the individual component phases in the heterojunction film; no other intermetallic phases were observed. The optimized thickness of both films was reflected in the highest photocurrent of 1.03 mA cm⁻² at 0.7 V. The mechanism described for n/n bilayer heterojunction films clarifies the reason that better electron/hole separation occurs in films that are narrow in the pristine state.

Acknowledgements

This research was supported by NRF-2013R1A2A2A05005589, Special International Collaboration and Post-Doctoral Fellowship funded by Korea University. This research was also supported by the Commercializations Promotion Agency for R&D Outcomes (COMPA) funded by the Ministry of Science, ICT and Future Planning (MISP). The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for its funding this Prolific Research group (PRG-1436-03).

References

1. J. Luo, S. D. Tilley, L. Steier, M. Schreier, M. T. Mayer, H. J. Fan and M. Grätzel, Nano Lett., 2015, 15, 1395–1402.
2. W. Sheng, B. Sun, T. Shi, X. Tan, Z. Peng and G. Liao, ACS Nano, 2014, 8, 7163–7169.
3. Y. Wang, Y. Wang, Y. Meng, H. Ding, Y. Shan, X. Zhao and X. Tang, J. Phys. Chem. C, 2008, 112, 6620–6626.
4. S.-J. Liu, Y.-F. Hou, S.-L. Zheng, Y. Zhang and Y. Wang, CrystEngComm, 2013, 15, 4124–4130.
5. C. Liao, Z. Ma, G. Dong, J. Qiu and R. J. Xie, J. Am. Ceram. Soc., 2015, 98, 957–964.
6. Y. F. Hou, S. J. Liu, J. H. Zhang, X. Cheng and Y. Wang, Dalton Trans., 2014, 43, 1025–1031.
7. R. Cao, H. Huang, N. Tian, Y. Zhang, Y. Guo and T. Zhang, Mater. Charact., 2015, 101, 166–172.
8. L.-W. Zhang, Y.-J. Wang, H.-Y. Cheng, W.-Q. Yao and Y.-F. Zhu, Adv. Mater., 2009, 21, 1286–1290.
9. Y. Peng, M. Yan, Q.-G. Chen, C.-M. Fan, H.-Y. Zhou and A.-W. Xu, J. Mater. Chem. A, 2014, 2, 8517–8524.
10. G. Colon, S. Murcia Lopez, M. C. Hidalgo and J. A. Navio, Chem. Commun., 2010, 46, 4809–4811.
11. M.-S. Gui, W.-D. Zhang, Y.-Q. Chang and Y.-X. Yu, Chem. Eng. J., 2012, 197, 283–288.
12. D. He, L. Wang, D. Xu, J. Zhai, D. Wang and T. Xie, ACS Appl. Mater. Interfaces, 2011, 3, 3167–3171.
13. N. Arai, N. Saito, H. Nishiyama, Y. Shimodaira, H. Kobayashi, Y. Inoue and K. Sato, J. Phys. Chem. C, 2008, 112, 5000–5005.
14. Y. Sun, C. Murphy, K. Reyesgil, E. Reyesgarcia, J. Lilly and D. Raftery, Int. J. Hydrogen Energy, 2008, 33, 5967–5974.
15. A. K. Chakraborty and M. A. Kebede, J. Cluster Sci., 2011, 23, 247–257.
16. A. Walsh, J. L. F. Da Silva, S.-H. Wei, C. Körber, A. Klein, L. F. J. Piper, A. DeMasi, K. E. Smith, G. Panaccione, P. Torelli, D. J. Payne, A. Bourlange and R. G. Egdell, Phys. Rev. Lett., 2008, 100, 167402.
17. D. V. Shinde, D. Y. Ahn, V. V. Jadhav, D. Y. Lee, N. K. Shrestha, J. K. Lee, H. Y. Lee, R. S. Mane and S.-H. Han, J. Mater. Chem. A, 2014, 2, 5490–5498.
18. J. Mu, B. Chen, M. Zhang, Z. Guo, P. Zhang, Z. Zhang, Y. Sun, C. Shao and Y. Liu, ACS Appl. Mater. Interfaces, 2012, 4, 424–430.
19. Z. Wang, B. Huang, Y. Dai, X. Qin, X. Zhang, P. Wang, H. Liu and J. Yu, J. Phys. Chem. C, 2009, 113, 4612–4617.
20. S. Choudhary, S. Upadhyay, P. Kumar, N. Singh, V. R. Satsangi, R. Shrivastav and S. Dass, Int. J. Hydrogen Energy, 2012, 37, 18713–18730.
21. X. Liu, Q. Lu, C. Wang, C. Zhu and S. Liu, J. Sol-Gel Sci. Technol., 2014, 73, 358–364.
22. H. Yoon, J. H. Woo, Y. M. Ra, S. S. Yoon, H. Y. Kim, S. Ahn, J. H. Yun, J. Gwak, K. Yoon and S. C. James, Aerosol Sci. Technol., 2011, 45, 1448–1455.
23. H. Yoon, M. G. Mali, J.-Y. Choi, M.-W. Kim, S. K. Choi, H. Park, S. S. Al-Deyab, M. T. Swihart, A. L. Yarin and S. S. Yoon, Langmuir, 2015, 31, 3727–3737.
24. M. G. Mali, H. Yoon, M.-W. Kim, M. T. Swihart, S. S. Al-Deyab and S. S. Yoon, Appl. Phys. Lett., 2015, 106, 151603.
25. M. Ge, Y. Li, L. Liu, Z. Zhou and W. Chen, J. Phys. Chem. C, 2011, 115, 5220–5225.
26. D. V. Shinde, V. V. Jadhav, D. Y. Lee, N. K. Shrestha, J. K. Lee, H. Y. Lee, R. S. Mane and S.-H. Han, J. Mater. Chem. A, 2014, 2, 5490–5498.
27. G. Zhang, Z. Hu, M. Sun, Y. Liu, L. Liu, H. Liu, C.-P. Huang, J. Qu and J. Li, Adv. Funct. Mater., 2015, 25, 3726–3734.
28. M. G. Mali, H. Yoon, H. Kim, B. N. Joshi, S. S. Al-Deyab and S. S. Yoon, ChemPhysChem, 2015 DOI:10.1002/cphc.201500636.
29. Y. Xu and M. A. Schoonen, Am. Mineral., 2000, 85, 543–556.
30. H. Cheng, B. Huang, P. Wang, Z. Wang, Z. Lou, J. Wang, X. Qin, X. Zhang and Y. Dai, Chem. Commun., 2011, 47, 7054–7056.
31. H. Jung, S. Y. Chae, C. Shin, B. K. Min, O.-S. Joo and Y. J. Hwang, ACS Appl. Mater. Interfaces, 2015, 7, 5788–5796.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra16833c

‡ Equal contribution.

---