Work function engineering derived all-solid-state Z-scheme semiconductor-metal-semiconductor system towards high-efficiency photocatalytic H₂ evolution

Abstract

An artificial semiconductor-metal-semiconductor Z-scheme photosynthetic system has recently been developed. Herein we suggest a general strategy for fabricating an effective Z-scheme system. The work function measured by a Kelvin probe and surface photovoltage spectroscopy were employed to obtain direct evidence for the photogenerated charge separation direction at metal-semiconductor interfaces. The suitable work function values of the components enabled the heterostructure of WO₃/Au/CdS to exhibit more unobstructed charge carrier transfer. As a result, the WO₃/Au/CdS Z-scheme system shows a better photocatalytic H₂ evolution efficiency than WO₃/Pt/CdS. These findings illuminate that the work function engineering of the intermediate metal plays a key role in Z-scheme photosynthetic systems.

---

The H₂ evolution from water splitting using solar energy conversion is one of the promising solutions to address the global energy crisis and environmental pollution.¹ Recently, a smart strategy of coupling semiconductor-metal-semiconductor analogous to the so-called Z-scheme mechanism in green plants has been represented as semiconductor-based biomimetic artificial photosynthetic systems. In analogy to the natural Z-scheme, semiconductors with different band structures are used as half-cell substitutes for photosystem I (PSI) and photosystem II (PSII), and the metal is used as an electron transport chain, respectively. The all-solid-state Z-scheme system was first introduced by Tada et al., which highlighted the metal Au as an electron transport mediator to promote vectorial photoinduced charge carrier separation between semiconductors TiO₂ (PSI) and CdS (PSII).² Since then the construction of all-solid-state Z-scheme systems has attracted considerable efforts due to their potential application in efficient visible light photocatalytic reactions. So far, various kinds of Z-scheme systems have been reported, such as CdS–Au–TiO₂, CdS–Au–TiO_1.96C_0.04, WO₃–ITO–CaFe₂O₄, AgCl–Ag–TaON, CdS–Au–ZnO, BiVO₄–PRGO–Ru/SrTiO₃:Rh, Ag₃PO₄–Ag–AgI, g-C₃N₄–TiO₂ and Pt/MoO₃–TiO₂, etc.³ These heterostructures exhibit high activity in photocatalytic water splitting or photocatalytic degradation compared with the single component systems. However, to our knowledge, due to the lack of universally applicable strategies, the design and construction of a Z-scheme system remains a rather challenging and blind process.

Charge carrier transport is one of the central issues in solar energy conversion. It is of great importance to control the separation direction of photoinduced charge carriers for improving their separation efficiency. Surface photovoltage (SPV) spectroscopy and SPV transient measurements are powerful techniques to study the transport behavior of photogenerated charge carriers at the interface of the nanoscaled semiconductor materials due to their high sensitivity.⁴ Gross etc. used SPV spectroscopy to demonstrate the directionality of charge separation in multilayered type II CdTe and CdSe NCs interface.⁵ In our previous study, it has been demonstrated that the intensity of the interfacial electric field and the separation direction of photoinduced charge carriers in p–n Cu₂O homojunction, p–n TiO₂/CoO_x and Cu₂O/ZnO heterojunction film can be detected by SPV spectroscopy and SPV transient measurements.⁶ These findings demonstrate that work function matching principle can be a general strategy for the construction of efficient photocatalyst. More recently, we reported a new all-solid-state Z-scheme WO₃/Pt/CdS system, which shows a high activity for photocatalytic H₂ generation under visible radiation.⁷ Beyond the experimentally determined photoinduced charge transport process, the Fermi levels for many components remain unclear, which make the selection of the semiconductors and metals for optimizing those Z-scheme systems are aimless.

In this communication, a general strategy of fabricating more efficient artificial all-solid-state Z-scheme semiconductor-metal-semiconductor photosynthetic system is proposed based on work function engineering, whose validity was confirmed by probing the photoinduced charges and H₂ evolution. The work function values of metal Pt, Au and semiconductor CdS, WO₃ are determined by Kelvin probe first. The relative positions of Fermi level of the components in metal/semiconductor induced depletion layer between Pt/CdS, Pt/WO₃, Au/CdS interface and accumulation layer between Au/WO₃ interface upon contact, which have been identified according to the results of SPV spectroscopy. It is revealed that the Z-scheme charge carriers transfer process could be more unobstructed in WO₃/Au/CdS, which could significantly enhance the efficiency of photocatalytic H₂ evolution. We highlight that the work function of the intermediate metal plays an important role in the construction of unobstructed Z-scheme electron transport process.

Thin films of the Pt and Au on FTO glass were obtained by cyclic voltammetry method in 20 mM HAuCl₄ and H₂PtCl₆ solution. The CdS and WO₃ thin films on FTO glass were obtained by coating method with commercial CdS and WO₃ powder. The XRD patterns of the four thin films are shown in Fig. S1.† The X-ray diffraction peaks of prepared films can be well assigned to the metal Pt, Au and semiconductor CdS, WO₃. Work functions (ϕ) of the Pt, Au, CdS, and WO₃ films are measured by Kelvin probe, which directly provides the contact potential differences (CPDs) between the samples and the Au probe.⁸ Fig. 1a shows the work function mapping results of Pt, Au, CdS and WO₃ loaded on ITO glass. The ϕ values are calculated by using the following equation:

ϕ_sample = ϕ_Au + eCPD, (ϕ_Au = 5.1 eV)

Fig. 1 (a) The work function mapping of Pt, Au, CdS, and WO₃ loaded on ITO glass measured by Kelvin probe in air. Side FESEM view of Au/WO₃ (b), Pt/WO₃ (c), Au/CdS (d), Pt/CdS (e). SPV spectra of FTO/Pt/WO₃, FTO/Au/WO₃ (f) and FTO/Pt/CdS, FTO/Au/CdS (g) irradiated from the FTO side under low light intensity.

Therefore, the working function values of Au, Pt, CdS, and WO₃ films are about 5 eV, 5.2 eV, 4.9 eV and 5.05 eV, respectively. These data imply that the Fermi level of CdS is higher than those of the metals Au and Pt, and the Fermi level of WO₃ is between those of the two metals.

In order to study the metal (Pt, Au) and semiconductor (CdS, WO₃) contact induced band bending, we further construct the metal/semiconductor heterostructure by coating commercial CdS and WO₃ powder on Pt/FTO and Au/FTO glass. The cross sectional FESEM images of the FTO/Au/CdS, FTO/Pt/CdS, FTO/Au/WO₃ and FTO/Pt/WO₃ heterostructure films are shown in Fig. 1b–e, respectively. All the composite films exhibit obvious hierarchical heterostructure. The metals Au and Pt show good coverage of the FTO substrate and both the metal films of the Pt and Au are almost have the same thickness of about 100 nm. It is shown obvious that the semiconductor CdS and WO₃ are well attached on the surface of Pt and Au films.

Fig. 1f shows the SPV spectra of FTO/Au/WO₃ film and FTO/Pt/WO₃ film, respectively. The details of the signal sampling and SPV measurement configurations are illustrated in Fig. S2.† The frequency-modulated light is incident from the FTO side of the films. It is interesting to observe that a 180° phase difference existed between their SPV phase spectra, which indicates that the transport directions of photoinduced charge carriers are opposite to each other for the two samples probed.⁹ Furthermore, the SPV spectra of the two samples respectively reveal opposite-sign responses dependent on spectrum. For Au/WO₃ film, with monochromatic light irradiation from 2.6 to 4 eV, the positive photovoltage signal indicates that photoinduced electron will transfer from WO₃ toward the Au layer and holes transfer to the non-irradiation side of WO₃, which is corresponding to a SPV phase towards 0°. Similarly, for Pt/WO₃ film, the negative photovoltage signal indicates photoinduced holes will transfer toward the Pt layer and electron transfer to the non-irradiation side of WO₃, corresponding to a SPV phase towards 180°. Therefore, we can determine that the orientations of the two interfacial electric fields at the Au/WO₃ and Pt/WO₃ interfaces are opposite.

Based on the information of the work function mapping, the formation of the opposite interfacial electric fields can be explained. The concept of band bending was first presented by Mott and Schottky to explain the rectifying effect of metal-semiconductor contacts.¹⁰ According to the results of work function mapping, Fig. S3† shows the energy band diagrams of metal Au, Pt and semiconductor WO₃ before contacting, respectively. When Au and WO₃ come to contact, the free electron will transfer from the Au to the WO₃, as the metal work function ϕ_Au < ϕ_WO3, until the Fermi levels of metal and semiconductor are aligned. For semiconductor WO₃, the free electron concentration near the interface is accumulated compared with the bulk. The downward band bending toward the interface appears due to the existence of the accumulation layer, which is corresponding to the interfacial electric field from Au towards WO₃. When the Au/WO₃ interface is under irradiation, photoinduced electron will transfer to the interface and holes transfer to the semiconductor bulk under the influence of interfacial electric fields, which leads to the positive photovoltage signal as shown in Fig. S4.† Nevertheless, opposite process would occur when metal Pt and semiconductor WO₃ are in contact. The electron are depleted in the space charge region due to the free electron transferring from WO₃ to Pt. The depletion layer leads to the upward band bending toward the interface and the barrier at the Pt/WO₃ interface could be called as a Schottky barrier. Thus the negative photovoltage signal is obtained under irradiation.¹¹

Fig. 1g shows the SPV spectra of FTO/Au/CdS and FTO/Pt/CdS films. The similar SPV phase angles indicate that the two samples have the same transport direction of the photoinduced charge carriers under illumination. The approximate ±180° phase value, the negative SPV signal indicate the formation of Schottky barrier at both the Au/CdS and Pt/CdS interfaces. Additionally, there should be a larger magnitude of interfacial electric field at the Pt/CdS interface compared with Au/CdS, according to the stronger photovoltage intensity. This phenomenon can also be explained with Fig. S4.† Upon contact of the metal and semiconductor, the free electron will transfer from the CdS to Au (or Pt) films, because the metal work function ϕ_Au (or ϕ_Pt) > ϕ_CdS. The Schottky barrier height at the Au/CdS interface should be lower than Pt/CdS, as ϕ_CdS − ϕ_Au < ϕ_CdS − ϕ_Pt, which is corresponding to a weak photovoltage signal.

Based on the results above, the charge separation and transfer process of the WO₃/Au/CdS and CdS/Pt/WO₃ ternary composite structure can be concluded. As shown in Scheme 1, before contact, the Fermi level of Au is between semiconductor CdS and WO₃. When they come into contact, the free electron will transfer from CdS to Au, and also from Au to WO₃ until the Fermi levels of the three composites are aligned under thermal equilibrium. Therefore, the space-charge region formed at either side of metal Au are completely inverse in electric field orientation (accumulation layer at Au/CdS interface and depletion layer at Au/WO₃ interface, respectively). When the ternary WO₃/Au/CdS composite is under irradiation, photoinduced holes in CdS and photoinduced electron in WO₃ will transfer to metal Au under the influence of the interfacial electric field and they finally recombine. This process can be interpreted as a perfect framework of the Z-scheme. On the contrast, the Fermi level of Pt is below those of semiconductor CdS and WO₃ before contact. Upon contact of the components, the free electron will transfer from both CdS and WO₃ to Pt until the Fermi levels of the three composites are aligned. Depletion layer will be formed at both side of metal Pt. While the low height of the Schottky barrier at Pt/WO₃ make it possible for electron tunneling to metal Pt to recombine holes from CdS the Z-scheme framework in WO₃/Pt/CdS is imperfect, compared with the unobstructed charge carriers transfer process in WO₃/Au/CdS system.

Scheme 1 (a) Energy band diagrams of metal Au, Pt and semiconductor CdS, WO₃ before contacts. (b) Charge separation and transfer schematic of WO₃/Au/CdS and WO₃/Pt/CdS systems under light irradiation.

In a follow-up study, the metals Au and Pt were loaded on the surface of WO₃ nanorod by a NaBH₄ reduction method, and then the CdS was loaded on the metal by an uncomplicated precipitation method. The XRD patterns of the as-prepared WO₃/Au/CdS and WO₃/Pt/CdS ternary nanocrystal are shown in Fig. S5.† No diffraction peaks of metal Au and Pt have been found, due to the low loading content. Therefore, transmission electron microscopic (TEM) measurements were conducted to certify the formation of the two samples. As can be clearly seen in Fig. 2a, both of the two samples show obvious ternary composite structure at low resolution. Furthermore, the constituent of the ternary composite structure is detected by high-resolution TEM characterization collected from the interfacial region (square region marked in the figure). The lattice spacing of 0.365 nm can be assigned to the (110) planes of hexagonal structure WO₃ and the lattice spacing of 0.336 nm can be assigned to the (111) plane in cubic phase CdS. It is obvious that the interplanar distance of 0.235 nm and 0.266 nm could be identified as the (111) plane of Au and Pt, respectively. These results clearly reveal the metal Au and Pt are successfully loaded at the interface of the semiconductor CdS and WO₃, which make it possible for the formation of the Z-scheme charge separation and transfer framework.

Fig. 2 TEM images of WO₃/Au/CdS (a) and WO₃/Pt/CdS (b) ternary composite nanocrystal at low resolution and high resolution. (c) Rate of H₂ evolution of pure WO₃, CdS, Au/CdS, Pt/CdS, WO₃/CdS, WO₃/Au/CdS and WO₃/Pt/CdS samples. The inset shows the SPV transient measurements of CdS, WO₃/CdS, WO₃/Au/CdS and WO₃/Pt/CdS samples blended with 10 μL lactic acid sacrificial agent under 355 nm laser irradiated.

Lactic acid solution (5 vol%) was employed as a sacrificial reagent to evaluate the photocatalytic activity of H₂ evolution. This reaction is regarded as a half reaction of water splitting and is used as test of photocatalytic H₂ evolution.¹ The photocatalytic H₂ evolution efficiencies of WO₃, CdS, Au/CdS, Pt/CdS, WO₃/CdS, WO₃/Au/CdS and WO₃/Pt/CdS samples are shown Fig. 2c. The pure WO₃ sample shows no activity of H₂ production, which could be attributed to the conduction band position lower than the reduction potential of H⁺. The pure CdS sample shows a poor activity of H₂ evolution about 95 μmol h⁻¹, due to the low charge separation efficiency. The H₂ evolution rate of WO₃/Pt/CdS is significantly enhanced to 6.9 mmol h⁻¹, which is about 72 times higher than that of pure CdS and also more higher than that of Au or Pt modified CdS. Particularly, the WO₃/Au/CdS sample exhibits a H₂ evolution rate of 10.7 mmol h⁻¹, which is about 1.5 times higher compared with WO₃/Pt/CdS sample. The inset of Fig. 2c shows the SPV transient results of CdS, WO₃/CdS, WO₃/Au/CdS and WO₃/Pt/CdS samples. The negative SPV signal represent photogenerated electron accumulate at the surface of the samples, which indicates the Z-scheme charge transfer has happened and photoinduced electron transfer to the surface of WO₃/CdS, WO₃/Au/CdS and WO₃/Pt/CdS samples. At the same time, WO₃/Au/CdS has the longest lifetime of the charge carriers, which should be attributed to the more unobstructed charge carriers transfer process in WO₃/Au/CdS system. Therefore, the WO₃/Au/CdS sample shows the best photocatalytic activity of H₂ evolution and the photocatalyst also shows good stability after 7 cycles of photocatalytic reaction as shown in Fig. S7.†

In conclusion, we have proposed the charge separation and transfer mechanism of WO₃/Au/CdS and WO₃/Pt/CdS ternary composite systems combinative to the SPV and work function study. The accumulation layer and depletion layer are formed at Au/CdS and Au/WO₃ interface respectively, which make the Z-scheme framework charge carriers separation and transfer process more unobstructed in WO₃/Au/CdS system, compared with WO₃/Pt/CdS system. Therefore, the WO₃/Au/CdS sample reveals an amazing ability in photocatalytic H₂ evolution reaction. Our findings demonstrated that the work function of the intermediate metal plays an important role in the construction of artificial all-solid-state semiconductor-metal-semiconductor Z-scheme photosynthetic system. The work function matching principle proposed herein could be adopted as a general strategy in constructing of more efficient systems of artificial Z-scheme photosynthetic photocatalyst.

Acknowledgements

We are grateful to the National Natural Science Foundation of China (No. 51572106, 21173103 and 51172090), the National Basic Research Program of China (973 Program) (2013CB632403), and the Science and Technology Developing Funding of Jilin Province.

Notes and references

1. (a) A. Fujishima and K. Honda, Nature, 1972, 238, 37–38; (b) A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278; (c) X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570; (d) T. Wang and J. Gong, Angew. Chem., Int. Ed., 2015, 54, 10718–10732; (e) C. G. Morales-Guio, M. T. Mayer, A. Yella, S. D. Tilley, M. Grätzel and X. Hu, J. Am. Chem. Soc., 2015, 137, 9927–9936; (f) F. Lei, L. Zhang, Y. Sun, L. Liang, K. Liu, J. Xu, Q. Zhang, B. Pan, Y. Luo and Y. Xie, Angew. Chem., Int. Ed., 2015, 54, 9266–9270; (g) C. Y. Zhai, M. S. Zhu, F. Z. Pang, D. Bin, C. Lu, M. C. Goh, P. Yang and Y. K. Du, ACS Appl. Mater. Interfaces, 2016, 8, 5972; (h) Y. J. Wu, Z. K. Yue, A. J. Liu, P. Yang and M. S. Zhu, ACS Sustainable Chem. Eng., 2016, 4, 2569.
2. H. Tada, T. Mitsui, T. Kiyonaga, T. Akita and K. Tanaka, Nat. Mater., 2006, 5, 782–786.
3. (a) H. J. Yun, H. Lee, N. D. Kim, D. M. Lee, S. Yu and J. Yi, ACS Nano, 2011, 5, 4084–4090; (b) Z. Liu and M. Miyauchi, Chem. Commun., 2009, 2002–2004; (c) J. Hou, C. Yang, Z. Wang, Q. Ji, Y. Li, G. Huang, S. Jiao and H. Zhu, Appl. Catal., B, 2013, 142–143, 579–589; (d) A. Iwase, Y. H. Ng, Y. Ishiguro, A. Kudo and R. Amal, J. Am. Chem. Soc., 2011, 133, 11054–11057; (e) Z. Chen, W. Wang, Z. Zhang and X. Fang, J. Phys. Chem. C, 2013, 117, 19346–19352; (f) J. Yu, S. Wang, J. Low and W. Xiao, Phys. Chem. Chem. Phys., 2013, 15, 16883–16890; (g) B. J. Ma, J. S. Kim, C. H. Choi and S. I. Woo, Int. J. Hydrogen Energy, 2013, 38, 3582–3587.
4. (a) L. Kronik and Y. Shapira, Surf. Sci. Rep., 1999, 37, 1–206; (b) V. Duzhko, V. Y. Timoshenko, F. Koch and T. Dittrich, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 075204.
5. D. Gross, I. Mora-Seró, T. Dittrich, A. Belaidi, C. Mauser, A. J. Houtepen, E. D. Como, A. L. Rogach and J. Feldmann, J. Am. Chem. Soc., 2010, 132, 5981–5983.
6. (a) T. Jiang, T. Xie, W. Yang, L. Chen, H. Fan and D. Wang, J. Phys. Chem. C, 2013, 117, 4619–4624; (b) T. Jiang, T. Xie, L. Chen, Z. Fu and D. Wang, Nanoscale, 2013, 5, 2938–2944; (c) S. Li, L. Hou, L. Zhang, L. Chen, Y. Lin, D. wang and T. Xie, J. Mater. Chem. A, 2015, 3, 17820–17826.
7. L. J. Zhang, S. Li, B. K. Liu, D. J. Wang and T. F. Xie, ACS Catal., 2014, 4, 3724–3729.
8. (a) P. Bergveld, J. Hendrikse and W. Olthuis, Meas. Sci. Technol., 1998, 9, 1801; (b) W. Li and D. Y. Li, J. Chem. Phys., 2005, 122, 064708; (c) C. Larisa-Emilia, J. Sherri, S. Saman and T. Michael, Meas. Sci. Technol., 2007, 18, 567; (d) K. Jakobi, Adsorbed Layers on Surfaces. Part 2: Measuring Techniques and Surface Properties Changed by Adsorption, ed. A. P. Bonzel, Springer, Berlin, Heidelberg, 2002, vol. 42A2, pp. 165–263.
9. (a) V. Donchev, K. Kirilov, T. Ivanov and K. Germanova, Mater. Sci. Eng., B, 2006, 129, 186–192; (b) I. Ts, V. Donchev, K. Germanova and K. Kirilov, J. Phys. D: Appl. Phys., 2009, 42, 135302.
10. (a) W. Schottky, Naturwissenschaften, 1938, 26, 843; (b) N. F. Mott, Proc. R. Soc. London, Ser. A, 1939, 171, 27–38.
11. Z. Zhang and J. T. Yates, Chem. Rev., 2012, 112, 5520–5551.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra14680e

---