Nanoscale View Article Online
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
PAPER
Cite this: Nanoscale, 2014, 6, 8914
View Journal | View Issue
Solar-driven hydrogen evolution using a CuInS2/ CdS/ZnO heterostructure nanowire array as an efficient photoanode Youngwoo Choi, Minki Beak and Kijung Yong* Photoanodes prepared using CuInS2/CdS/ZnO nanowires were fabricated by a solution-based process for constructing a photo-driven hydrogen generation system. For efficient light harvesting and photoexcited charge collection, ZnO nanowire (NW) photoanode arrays were co-sensitized with CdS and CuInS2 (CIS). A CdS layer was deposited on the ZnO NW via successive ion layer adsorption and reaction (SILAR), and the CIS layer was prepared by depositing a molecular precursor solution onto the CdS/ZnO NW. The generated anodic photocurrent was increased with the subsequent deposition of the CIS and CdS layers.
Received 25th March 2014 Accepted 19th May 2014
Ultraviolet photoelectron spectroscopy analysis revealed cascade type-II band alignments for the CIS/ CdS/ZnO NW photoanodes, which enabled efficient electron collection. Our heterostructure
DOI: 10.1039/c4nr01632g
photoelectrode has generated a greatly improved photocurrent density of 13.8 mA cm2 at 0.3 V vs. SCE
www.rsc.org/nanoscale
under 1 sun illumination.
Introduction Hydrogen production by photoelectrochemical (PEC) water splitting using semiconductor materials is a promising solution to the global energy problem. Since Fujishima and Honda rst reported the photoelectrolysis of water splitting using a TiO2 electrode,1 there have been many studies conducted on the effect of various structures and materials on the performance of PEC.2–4 In particular, one dimensional (1D) metal oxide (e.g., ZnO and TiO2) nanostructures have been widely investigated as photoelectrodes due to their unique properties, such as efficient charge separation and transport, high surface-to-volume ratio, and light scattering and trapping.5–7 However, the band gap of these metal oxides is too wide to absorb the visible light. To overcome the limited absorption region, narrow band gap absorbers can be deposited onto the wide band gap metal oxides. Simultaneously, appropriate band positions are required to provide the overall photovoltage (>1.23 V) for direct water splitting.8 Recently, Cu(In,Ga)(SeS)2 (CIGS) chalcogenide materials have been described as attractive candidates for solar-driven water splitting because they have high absorption coefficients (104 to 105), tunable band gap energies (1.0–1.7 eV), and can produce a considerable portion of the required photovoltage.9–13 Especially because CuInS2 (CIS) does not require a highly toxic Se source and non-toxic solution synthesis is possible, various
Surface Chemistry Laboratory of Electronic Materials, Department of Chemical Engineering, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Korea. E-mail: [email protected]; Fax: +82-54-279-8298; Tel: +82-54279-2278
8914 | Nanoscale, 2014, 6, 8914–8918
deposition methods of CIS have been studied for PEC applications.14–17 Most of the efficient CIS electrodes were fabricated through vacuum methods, including co-evaporation,18 sputtering,19 and pulsed laser deposition (PLD).20 However, because these vacuum methods have drawbacks, such as a high production cost and difficulty in scaling-up, non-vacuum methods have attracted much attention in recent years. Matsumura et al. reported on the fabrication of CIS-based solar cells and PEC water splitting systems by using a facile nonvacuum technique, spray pyrolysis.21 Amal et al. reported on the fabrication of CuInS2–TiO2 photoelectrochemical cells using square wave pulse-assisted electrodeposition.22 However, the synthesized CIS photoelectrode through non-vacuum systems has shown relatively low photocurrent values. In most CIGS-based PEC water splitting, the p-type CIGS layer forms a p–n junction with an n-type semiconductor to improve its photoelectrochemical properties. Domen et al. demonstrated signicant improvements of PEC performances with p-type CuGaSe2 photocathodes by formation of a heterojunction with n-type CdS.23 This may be due to efficient charge separation caused by the surface p–n junction. Compared to the p-type CIS photocathode, only a few studies have investigated the use of an n-type CIS photoanode in PEC water splitting. It is known that the type of CIS can be determined according to stoichiometry.22 Through the development of n-type CIS, tandem PEC cells, consisting of a p-type CIS cathode and an n-type CIS anode, can be realized. In this study, we report on the synthesis of an n-type CIS absorber deposited onto a CdS/ZnO NW via a solution-based method and applied to a photoanode in PEC cells. From the point of scaling-up, a solution-based process has the advantage
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
of being easy to perform and having a low cost. Furthermore, because the CIS/CdS/ZnO NW has a cascade band alignment, which facilitates photogenerated charge carrier collection, the photocurrent density was drastically improved. Our photoelectrode has greatly improved the saturated photocurrent density (13.8 mA cm2) at 0.3 V vs. SCE under 1 sun illumination.
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Experimental Materials All chemicals were purchased from Sigma-Aldrich and were used as received. Zinc nitrate hexahydrate (Zn(NO3)2$6H2O, 98%), ammonium hydroxide (28 wt% NH3 in water, 99.99%), cadmium sulfate (CdSO4, 99%), sodium sulde (Na2S, 98%), copper iodide (CuI, 99.999%), indium acetate (In(OAc)3, 99.99%), thiourea (CH4N2S, 99%), 1-butylamine (C4H9NH2, 99.5%), and 1-propionic acid (C2H5COOH, 99.5%) were used. Preparation of the CIS/CdS/ZnO NW The CIS/CdS/ZnO NW arrays were prepared by an all-solution process. Firstly, ZnO NW arrays were grown on a FTO substrate by the hydrothermal method according to a previous report.24 A 50 nm ZnO seed layer was deposited on the cleaned FTO glass by sputtering a ZnO target at room temperature. Aer that, the ZnO sputtered FTO substrates were immersed in a 10 mM Zn(NO3)3$6H2O (98%, Aldrich) aqueous solution and 3.8 mL of ammonium hydroxide (28–30 wt%) and kept in an oven at 95 C overnight. Aer the NW growth process, the substrate was removed from the solution, rinsed with deionized (DI) water, and then dried using blown nitrogen. The ZnO nanowires were modied with CdS deposition using successive ionic layer adsorption and reaction (SILAR). Briey CdS was deposited with 200 mM CdSO4 solution and 200 mM Na2S aqueous solution with 20 repeated cycles. Next, CIS was deposited by molecular precursor solution onto the ZnO/CdS NW. CuI (0.5715 g), In(OAc)3 (0.971 g), thiourea (0.628 g), 1-butylamine (20 mL), and 1-propionic acid (1.3 mL) were mixed together and stirred for 10 min. The transparent sky-blue precursor solution was spincoated onto the CdS/ZnO NW samples at 1300 rpm for 30 s. The sample was then placed on a 150 C preheated hot plate for 10 min and immediately moved to a 250 C preheated hot plate and maintained at this temperature for 10 min. Aer heat treatment, the samples were allowed to cool to room temperature in an air environment.
Nanoscale
(SHIMADZU) spectrometer with an ISR-2200 integrating sphere attachment for the diffuse reection measurement. UPS measurements were carried out using a KRATOS AXIS Nova instrument. He I hn ¼ 21.22 eV was used as a light source and sputtered Au substrates were used as a reference. The photocurrent–voltage (I–V) measurements were performed using a typical three-electrode potentiostat system (Potentiostat/Galvanostat, model 263A, EG&G Princeton Applied Research) with a Pt counter electrode and a saturated calomel reference electrode (SCE). The electrolyte was a 0.25 M Na2S and 0.35 M Na2SO3 aqueous solution (pH 12.5) through which nitrogen was bubbled. The working electrode was illuminated from the front side with a solar-simulated light source (AM 1.5G ltered, 100 mW cm2, 91160, Oriel).
Results and discussion Fig. 1 shows scanning electron microscopy (SEM) images of the bare ZnO NW and the CIS/CdS/ZnO NW electrodes. Dense vertically aligned ZnO nanowire arrays have an average diameter and length of 150 nm and 17 mm, respectively. The deposition of CIS/CdS leads to a rough surface of ZnO NWs compared to bare ZnO NWs. In particular, the small nanoparticles were found at the tips of the nanowires. For a more detailed atomic structure analysis, we conducted a transmission electron microscopy (TEM) analysis. The TEM images in Fig. 2a and b clearly show that the CIS nanoparticles, which have a mean diameter of 3 nm, were deposited on the CdS/ ZnO NW. The interplanar spacings of the lattice fringes were ˚ 3.4 A ˚ and 2.5 A ˚ which corresponded to the measured to be 3.2 A, (112) planes of CIS (JCPDS-750106), the (111) planes of CdS (JCPDS-800019) and the (001) planes of ZnO (JCPDS-792205), respectively. Because process temperatures were relatively low (250 C), small CIS crystal grain sizes were obtained. In general, CIGS is sintered by sulfurization or selenization at 500 C,17 but in our study, we used a lower annealing temperature because at
Characterization The morphologies of the nanostructures and lms were investigated using eld emission scanning electron microscopy (FESEM, XL30S, Philips) and the detailed microscopic structure was observed by high-resolution scanning transmission electron microscopy (Cs-corrected HR-STEM, JEOLJEM-2200FS with electron energy loss spectroscopy (EELS) elemental mapping). The crystal structural properties were analyzed by X-ray diffraction studies (XRD, Max-2500 V, RIGAKU). The optical absorbance of the sample was analyzed using a UV2501PC
This journal is © The Royal Society of Chemistry 2014
Fig. 1 30 tilted view SEM images of (a) bare-ZnO NWs and (b) CIS/ CdS/ZnO NWs. (c) and (d) are high-magnification images of the samples (a) and (b), respectively.
Nanoscale, 2014, 6, 8914–8918 | 8915
View Article Online
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Nanoscale
Paper
on the substrate. The broad CdS peaks were also veried at 26.5 , corresponding to CdS (111) (JCPDS-800019). Because the CdS layer is thin (7 nm) and consists of nano-sized particles, a relatively broad peak was observed. Aer the CIS deposition via solution-based spin coating, the CIS (112) peak appeared at 27.9 . With the deposition of CIS on the CdS/ZnO NW surface, CdS and ZnO peak intensities were decreased. Because of the nanometer size of the CIS particles as identied in TEM, the CIS peak is broad with low intensity, which is due to the decient thermodynamic driving force for the sintering. This low temperature process yielded also residual In(OH)3, which was identied in the XRD pattern. To conrm the absorption properties of CIS and CdS as sensitizers on the ZnO NWs, the optical absorption spectra of various combinations of nanowires were recorded (Fig. 4(a)). The optical absorption spectra were gathered over the range 300–800 nm. Because ZnO has a wide band gap (approximately 3.4 eV) as observed in Fig. 4(a), the as-grown ZnO NWs only absorbed light in the UV region below 370 nm. When the CdS
(a) TEM and (b) HRTEM images of the CIS/CdS/ZnO NW. (c–h) EELS elemental images of Cu, In, S, Zn, O, and Cd.
Fig. 2
higher temperatures, CdS can diffuse into the ZnO NWs or the CIS layer. Spatial elemental analysis using electron energy loss spectroscopy (EELS) was also performed (Fig. 2c–h). The results conrmed that the Cu, In, S elements were uniformly deposited on the CdS/ZnO nanowire. XRD measurements were used to investigate the crystal structures of CdS/ZnO NWs and CIS/CdS/ZnO NWs. As shown in Fig. 3, the X-ray diffraction q–2q scan shows the various ZnO peaks, including the main peak of ZnO (00l) at 34.4 . These results indicate that individual ZnO NWs were vertically aligned
Fig. 3 XRD patterns of the CdS/ZnO NW and CIS/CdS/ZnO NW structures.
8916 | Nanoscale, 2014, 6, 8914–8918
Fig. 4 (a) Light absorbance spectra of the as-prepared ZnO NW, CdS/ ZnO NW, CdS/CIS/ZnO NW and CIS/CdS/ZnO NW. (b) Photocurrent density–voltage curves obtained from various photoanodes; asprepared ZnO, CdS/ZnO, CIS/ZnO, CdS/CIS/ZnO and CIS/CdS/ZnO NW arrays.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Paper
layer was deposited onto ZnO NWs, its absorption range was extended up to 500 nm. Finally, as the CIS absorber layer was deposited, the whole UV-visible light region was absorbed. The optical band gap (Eg) can be estimated from the absorption spectra, and the calculated direct band gap of the CIS absorber is approximately 1.60 eV. This value is slightly larger than that of the bulk CIS (1.54 eV) caused by its quantum connement effect. In the case of reversing the order of deposition of CdS and CIS (CdS/CIS/ZnO NW), the absorption was decreased, which is believed to be due to a reduction in the amount of CIS deposition. CdS could act as a seed layer to facilitate the deposition of CIS and thus the amount of deposited CIS is larger on CdS than that on bare ZnO NWs. Fig. 4(b) shows a set of chopped sweep current density versus potential curves obtained from various NW photoanodes using a white light illumination of 100 mW cm2 (AM 1.5 G). All of the measurements were carried out in a three-electrode electrochemical cell, with a Pt wire and a saturated calomel electrode as the counter and the reference electrode, respectively. An aqueous solution of 0.25 M Na2S and 0.35 M Na2SO3 (pH 12.5) was used as an electrolyte and a sacricial reagent, respectively. The bare ZnO NW shows negligible photocurrent generation because of the narrow absorption region. The enhanced photocurrent was observed with the CdS/ZnO NW due to the improved visible light absorption via the CdS absorber layer, but the photocurrent density was still rather low. Aer the CIS layer deposition on the CdS/ZnO NW, the photocurrent value was signicantly increased. Furthermore, unlike typical CIGS PEC cells, which showed cathodic photocurrent under illumination, our CIS electrode generated an anodic photocurrent, indicating that the prepared CIS has an n-type character because the type of CIS can be determined according to stoichiometry.22 A maximum photocurrent density of 13.8 mA cm2 at 0.3 V bias vs. SCE was achieved from the CIS/CdS/ZnO NW sample. This result indicated that the band structure of the CIS/ CdS/ZnO NW is suitable for charge carrier transfer, which is discussed further below in more detail. To conrm the effect of the deposition of the CdS layer on the photocurrent, the photocurrent change was conrmed according to the deposition sequence. While n-type materials were employed on the surface of the electrode in most applications of CIGS to PEC water splitting to improve its photoelectrochemical properties, in our case of the CdS/CIS/ZnO NW, it generated a lower photocurrent compared to the CIS/CdS/ZnO NW, despite only reversing the deposition sequence. This result showed that the band structure of the CIS/CdS/ZnO NW has an advantage of charge transfer from the CIS absorber layer to the ZnO NW compared to that of the CdS/CIS/ZnO NW. Furthermore, the CdS layer can mediate the lattice mismatch between CIS and ZnO. The lattice mismatch between the CIS layers and CdS is smaller than that between the CIS layers and the ZnO NW.25 Thus the lattice mismatch, which can cause strain or stress at the interface, is minimized when the CdS layer is inserted between CIS and ZnO. In addition, as shown in UV-vis spectra, the absorption was enhanced due to the increase in CIS deposition caused by the CdS layer acting as a seed layer. This result can be conrmed by the I–V results of the CIS/ZnO NW
This journal is © The Royal Society of Chemistry 2014
Nanoscale
structure. It exhibited not only a lower photocurrent density but also a higher dark current than the CdS/ZnO NW. This indicates poor heterojunction formation of CIS/ZnO, which is mainly due to the absence of a seed layer; therefore, a large uncoated surface of the ZnO NW was le. To obtain a better understanding of the detailed band structure of heterostructure NWs, we recorded ultraviolet photoelectron spectra (UPS) of CIS/CdS/ZnO and CdS/ZnO NWs, as shown in Fig. 5. Fig. 5(a) and (b) present the secondary electron cutoff (Ecutoff) and valence band maximum (EVBM) in UPS of the CdS and CIS layers. The valence band maximum (VBM) energy levels are calculated from the UPS data using the following formula:27 VBM ¼ hn (Ecutoff EVBM) where hn (21.22 eV) is the incident photon energy and EVBM is the onset relative to the Fermi level (EF) of Au, in which EF is determined from the Au substrate. The VBM energy levels for CIS and CdS were 5.52, and 6.52 eV, respectively. The conduction band minimum (CBM) levels of CIS and CdS can be estimated from the VBM energy levels and the optical band gap. The calculated CBM energy levels of CIS and CdS were 3.92 eV and 4.08 eV, respectively. Energy band diagrams based on these data can be found in Fig. 5(d). The energy levels of CIS, CdS and ZnO are summarized in Table 1.
(a) Secondary edge region of ultraviolet photoelectron spectra (UPS) of CdS and CIS. (b) Low energy onset region of UPS spectra of CdS and CIS. (c) Schematic diagram of the CIS/CdS/ZnO NW array in the PEC cell. (d) Electronic structures of ZnO, CdS and CIS.
Fig. 5
Nanoscale, 2014, 6, 8914–8918 | 8917
View Article Online
Nanoscale
Paper
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Table 1 Energy levels of the CdS and CIS layers determined using UVvis absorption spectra and UPS
Electrode
Eg (eV)
VBM (eV)
CBM (eV)
CIS CdS ZnO (ref. 26)
1.60 2.44 3.31
5.52 6.52 7.55
3.92 4.08 4.24
From the results, we can see that the CIS/CdS/ZnO NW electrode has a type-II cascade band structure that facilitates photogenerated carrier transport from CIS to the ZnO NW. This ideal band structure of the CIS/CdS/ZnO NW is suitable for improving the photocurrent generation of PEC cells. However, in the case of the CdS/CIS/ZnO NW, the photogenerated electrons in the CIS layer transfer to not only the ZnO NW but also the CdS layer and thus recombination increased between photogenerated electrons and the oxidizing species in the electrolyte.
Conclusions We have successfully developed a CIS/CdS/ZnO NW photoanode for PEC cells using a solution-based process. ZnO NWs provided a direct charge transfer path and an inserted CdS layer between CIS absorbers, and the ZnO NW yields a cascade of type-II band structures and serves as a seed layer to facilitate the deposition of the CIS layer. The synthesized CIS/CdS/ZnO NW photoelectrode produced a high photocurrent density of 13.8 mA cm2 at 0.3 V bias vs. SCE for hydrogen generation.
Acknowledgements This work was supported by the National Research Foundation of Korea (2013-R1A2A2A05-005344).
Notes and references 1 A. Fujishima and K. Honda, Nature, 1972, 238, 37–38. 2 M. Gratzel, Nature, 2001, 414, 338–344. 3 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori and N. S. Lewis, Chem. Rev., 2010, 110, 6446–6473. 4 K. Sivula, F. L. Formal and M. Gratzel, ChemPlusChem, 2011, 4, 432–449. 5 M. S. Seol, H. J. Kim, W. S. Kim and K. J. Yong, Electrochem. Commun., 2010, 12, 1416–1418.
8918 | Nanoscale, 2014, 6, 8914–8918
6 W. W. Zhan, Q. Kuang, J. Z. Zhou, X. J. Kong, Z. X. Xie and L.-S. Zheng, J. Am. Chem. Soc., 2013, 135, 1926–1933. 7 H. J. Kim and K. J. Yong, ACS Appl. Mater. Interfaces, 2013, 5, 13258–13264. 8 T. Bak, J. Nowotny, M. Rekas and C. C. Sorrell, Int. J. Hydrogen Energy, 2002, 27, 991–1022. 9 H. C. Ye, H. S. Park, V. A. Akhavan, B. W. Goodfellow, M. G. Panthani, B. A. Korgel and A. J. Bard, J. Phys. Chem. C, 2011, 115, 234–240. 10 J. Xia, Y. Liu, X. Qiu, Y. Mao, J. He and L. Chen, Mater. Chem. Phys., 2012, 136, 823–830. 11 L. Rovelli, S. D. Tilley and K. Sivula, ACS Appl. Mater. Interfaces, 2013, 5, 8018–8024. 12 D. Yokoyama, T. Minegishi, K. Maeda, M. Katayama, J. Kubota, A. Yamada, M. Konagai and K. Domen, Electrochem. Commun., 2010, 12, 851–853. 13 J. H. Kim, T. Minegishi, J. Kobota and K. Domen, Energy Environ. Sci., 2012, 5, 6368. 14 J. Suna, S. K. Batabyal, P. D. Tranb and L. H. Wong, J. Alloys Compd., 2014, 591, 127–131. 15 Y. Tang, Y. H. Ng, J. H. Yun and R. Amal, RSC Adv., 2014, 4, 3278. 16 A. Tapley, D. Vaccarello, J. Hedges, F. Jia, D. A. Loveb and Z. Ding, Phys. Chem. Chem. Phys., 2013, 15, 1431. 17 F. Yang, V. Kuznietsov, M. Lublow, C. Merschjann, A. Steigert, J. Klaer, A. Thomase and T. S. Niedriga, J. Mater. Chem. A, 2013, 1, 64072. 18 M. A. Contreras, B. Egaas, K. Ramanathan, J. Hiltner, A. Swartzlander, F. Hasoon and R. Nou, Prog. Photovoltaics, 1999, 7, 311. 19 J. F. Jiang, Thin Solid Films, 2006, 515, 1950. 20 S. Mandati, B. V. Sarada, S. R. Dey and S. V. Joshi, J. Renewable Sustainable Energy, 2013, 5, 031602. 21 S. Ikeda, M. Nonogaki, W. Septina, G. Gunawan, T. Harada and M. Matsumura, Catal. Sci. Technol., 2013, 3, 1849. 22 J. H. Yun, Y. H. Ng, S. Huang, G. Conibeerb and R. Amal, Chem. Commun., 2011, 47, 11288–11290. 23 M. Moriya, T. Minegishi, H. Kumagai, M. Katayama, J. Kubota and K. Domen, J. Am. Chem. Soc., 2013, 135, 3733–3735. 24 Y. J. Tak and K. J. Yong, J. Phys. Chem. B, 2005, 109, 19263– 19269. 25 K. A. Jones, J. Cryst. Growth, 1979, 47, 235–244. 26 K. Yuan, L. Chen, F. Lia and Y. Chen, J. Mater. Chem. C, 2014, 2, 1018. 27 J. H. Seo, R. Yang, J. Z. Brzezinski, B. Walker, G. C. Bazan and T. Q. Nguyen, Adv. Mater., 2009, 21, 1006–1011.
This journal is © The Royal Society of Chemistry 2014
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
PAPER
Cite this: Nanoscale, 2014, 6, 8914
View Journal | View Issue
Solar-driven hydrogen evolution using a CuInS2/ CdS/ZnO heterostructure nanowire array as an efficient photoanode Youngwoo Choi, Minki Beak and Kijung Yong* Photoanodes prepared using CuInS2/CdS/ZnO nanowires were fabricated by a solution-based process for constructing a photo-driven hydrogen generation system. For efficient light harvesting and photoexcited charge collection, ZnO nanowire (NW) photoanode arrays were co-sensitized with CdS and CuInS2 (CIS). A CdS layer was deposited on the ZnO NW via successive ion layer adsorption and reaction (SILAR), and the CIS layer was prepared by depositing a molecular precursor solution onto the CdS/ZnO NW. The generated anodic photocurrent was increased with the subsequent deposition of the CIS and CdS layers.
Received 25th March 2014 Accepted 19th May 2014
Ultraviolet photoelectron spectroscopy analysis revealed cascade type-II band alignments for the CIS/ CdS/ZnO NW photoanodes, which enabled efficient electron collection. Our heterostructure
DOI: 10.1039/c4nr01632g
photoelectrode has generated a greatly improved photocurrent density of 13.8 mA cm2 at 0.3 V vs. SCE
www.rsc.org/nanoscale
under 1 sun illumination.
Introduction Hydrogen production by photoelectrochemical (PEC) water splitting using semiconductor materials is a promising solution to the global energy problem. Since Fujishima and Honda rst reported the photoelectrolysis of water splitting using a TiO2 electrode,1 there have been many studies conducted on the effect of various structures and materials on the performance of PEC.2–4 In particular, one dimensional (1D) metal oxide (e.g., ZnO and TiO2) nanostructures have been widely investigated as photoelectrodes due to their unique properties, such as efficient charge separation and transport, high surface-to-volume ratio, and light scattering and trapping.5–7 However, the band gap of these metal oxides is too wide to absorb the visible light. To overcome the limited absorption region, narrow band gap absorbers can be deposited onto the wide band gap metal oxides. Simultaneously, appropriate band positions are required to provide the overall photovoltage (>1.23 V) for direct water splitting.8 Recently, Cu(In,Ga)(SeS)2 (CIGS) chalcogenide materials have been described as attractive candidates for solar-driven water splitting because they have high absorption coefficients (104 to 105), tunable band gap energies (1.0–1.7 eV), and can produce a considerable portion of the required photovoltage.9–13 Especially because CuInS2 (CIS) does not require a highly toxic Se source and non-toxic solution synthesis is possible, various
Surface Chemistry Laboratory of Electronic Materials, Department of Chemical Engineering, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Korea. E-mail: [email protected]; Fax: +82-54-279-8298; Tel: +82-54279-2278
8914 | Nanoscale, 2014, 6, 8914–8918
deposition methods of CIS have been studied for PEC applications.14–17 Most of the efficient CIS electrodes were fabricated through vacuum methods, including co-evaporation,18 sputtering,19 and pulsed laser deposition (PLD).20 However, because these vacuum methods have drawbacks, such as a high production cost and difficulty in scaling-up, non-vacuum methods have attracted much attention in recent years. Matsumura et al. reported on the fabrication of CIS-based solar cells and PEC water splitting systems by using a facile nonvacuum technique, spray pyrolysis.21 Amal et al. reported on the fabrication of CuInS2–TiO2 photoelectrochemical cells using square wave pulse-assisted electrodeposition.22 However, the synthesized CIS photoelectrode through non-vacuum systems has shown relatively low photocurrent values. In most CIGS-based PEC water splitting, the p-type CIGS layer forms a p–n junction with an n-type semiconductor to improve its photoelectrochemical properties. Domen et al. demonstrated signicant improvements of PEC performances with p-type CuGaSe2 photocathodes by formation of a heterojunction with n-type CdS.23 This may be due to efficient charge separation caused by the surface p–n junction. Compared to the p-type CIS photocathode, only a few studies have investigated the use of an n-type CIS photoanode in PEC water splitting. It is known that the type of CIS can be determined according to stoichiometry.22 Through the development of n-type CIS, tandem PEC cells, consisting of a p-type CIS cathode and an n-type CIS anode, can be realized. In this study, we report on the synthesis of an n-type CIS absorber deposited onto a CdS/ZnO NW via a solution-based method and applied to a photoanode in PEC cells. From the point of scaling-up, a solution-based process has the advantage
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
of being easy to perform and having a low cost. Furthermore, because the CIS/CdS/ZnO NW has a cascade band alignment, which facilitates photogenerated charge carrier collection, the photocurrent density was drastically improved. Our photoelectrode has greatly improved the saturated photocurrent density (13.8 mA cm2) at 0.3 V vs. SCE under 1 sun illumination.
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Experimental Materials All chemicals were purchased from Sigma-Aldrich and were used as received. Zinc nitrate hexahydrate (Zn(NO3)2$6H2O, 98%), ammonium hydroxide (28 wt% NH3 in water, 99.99%), cadmium sulfate (CdSO4, 99%), sodium sulde (Na2S, 98%), copper iodide (CuI, 99.999%), indium acetate (In(OAc)3, 99.99%), thiourea (CH4N2S, 99%), 1-butylamine (C4H9NH2, 99.5%), and 1-propionic acid (C2H5COOH, 99.5%) were used. Preparation of the CIS/CdS/ZnO NW The CIS/CdS/ZnO NW arrays were prepared by an all-solution process. Firstly, ZnO NW arrays were grown on a FTO substrate by the hydrothermal method according to a previous report.24 A 50 nm ZnO seed layer was deposited on the cleaned FTO glass by sputtering a ZnO target at room temperature. Aer that, the ZnO sputtered FTO substrates were immersed in a 10 mM Zn(NO3)3$6H2O (98%, Aldrich) aqueous solution and 3.8 mL of ammonium hydroxide (28–30 wt%) and kept in an oven at 95 C overnight. Aer the NW growth process, the substrate was removed from the solution, rinsed with deionized (DI) water, and then dried using blown nitrogen. The ZnO nanowires were modied with CdS deposition using successive ionic layer adsorption and reaction (SILAR). Briey CdS was deposited with 200 mM CdSO4 solution and 200 mM Na2S aqueous solution with 20 repeated cycles. Next, CIS was deposited by molecular precursor solution onto the ZnO/CdS NW. CuI (0.5715 g), In(OAc)3 (0.971 g), thiourea (0.628 g), 1-butylamine (20 mL), and 1-propionic acid (1.3 mL) were mixed together and stirred for 10 min. The transparent sky-blue precursor solution was spincoated onto the CdS/ZnO NW samples at 1300 rpm for 30 s. The sample was then placed on a 150 C preheated hot plate for 10 min and immediately moved to a 250 C preheated hot plate and maintained at this temperature for 10 min. Aer heat treatment, the samples were allowed to cool to room temperature in an air environment.
Nanoscale
(SHIMADZU) spectrometer with an ISR-2200 integrating sphere attachment for the diffuse reection measurement. UPS measurements were carried out using a KRATOS AXIS Nova instrument. He I hn ¼ 21.22 eV was used as a light source and sputtered Au substrates were used as a reference. The photocurrent–voltage (I–V) measurements were performed using a typical three-electrode potentiostat system (Potentiostat/Galvanostat, model 263A, EG&G Princeton Applied Research) with a Pt counter electrode and a saturated calomel reference electrode (SCE). The electrolyte was a 0.25 M Na2S and 0.35 M Na2SO3 aqueous solution (pH 12.5) through which nitrogen was bubbled. The working electrode was illuminated from the front side with a solar-simulated light source (AM 1.5G ltered, 100 mW cm2, 91160, Oriel).
Results and discussion Fig. 1 shows scanning electron microscopy (SEM) images of the bare ZnO NW and the CIS/CdS/ZnO NW electrodes. Dense vertically aligned ZnO nanowire arrays have an average diameter and length of 150 nm and 17 mm, respectively. The deposition of CIS/CdS leads to a rough surface of ZnO NWs compared to bare ZnO NWs. In particular, the small nanoparticles were found at the tips of the nanowires. For a more detailed atomic structure analysis, we conducted a transmission electron microscopy (TEM) analysis. The TEM images in Fig. 2a and b clearly show that the CIS nanoparticles, which have a mean diameter of 3 nm, were deposited on the CdS/ ZnO NW. The interplanar spacings of the lattice fringes were ˚ 3.4 A ˚ and 2.5 A ˚ which corresponded to the measured to be 3.2 A, (112) planes of CIS (JCPDS-750106), the (111) planes of CdS (JCPDS-800019) and the (001) planes of ZnO (JCPDS-792205), respectively. Because process temperatures were relatively low (250 C), small CIS crystal grain sizes were obtained. In general, CIGS is sintered by sulfurization or selenization at 500 C,17 but in our study, we used a lower annealing temperature because at
Characterization The morphologies of the nanostructures and lms were investigated using eld emission scanning electron microscopy (FESEM, XL30S, Philips) and the detailed microscopic structure was observed by high-resolution scanning transmission electron microscopy (Cs-corrected HR-STEM, JEOLJEM-2200FS with electron energy loss spectroscopy (EELS) elemental mapping). The crystal structural properties were analyzed by X-ray diffraction studies (XRD, Max-2500 V, RIGAKU). The optical absorbance of the sample was analyzed using a UV2501PC
This journal is © The Royal Society of Chemistry 2014
Fig. 1 30 tilted view SEM images of (a) bare-ZnO NWs and (b) CIS/ CdS/ZnO NWs. (c) and (d) are high-magnification images of the samples (a) and (b), respectively.
Nanoscale, 2014, 6, 8914–8918 | 8915
View Article Online
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Nanoscale
Paper
on the substrate. The broad CdS peaks were also veried at 26.5 , corresponding to CdS (111) (JCPDS-800019). Because the CdS layer is thin (7 nm) and consists of nano-sized particles, a relatively broad peak was observed. Aer the CIS deposition via solution-based spin coating, the CIS (112) peak appeared at 27.9 . With the deposition of CIS on the CdS/ZnO NW surface, CdS and ZnO peak intensities were decreased. Because of the nanometer size of the CIS particles as identied in TEM, the CIS peak is broad with low intensity, which is due to the decient thermodynamic driving force for the sintering. This low temperature process yielded also residual In(OH)3, which was identied in the XRD pattern. To conrm the absorption properties of CIS and CdS as sensitizers on the ZnO NWs, the optical absorption spectra of various combinations of nanowires were recorded (Fig. 4(a)). The optical absorption spectra were gathered over the range 300–800 nm. Because ZnO has a wide band gap (approximately 3.4 eV) as observed in Fig. 4(a), the as-grown ZnO NWs only absorbed light in the UV region below 370 nm. When the CdS
(a) TEM and (b) HRTEM images of the CIS/CdS/ZnO NW. (c–h) EELS elemental images of Cu, In, S, Zn, O, and Cd.
Fig. 2
higher temperatures, CdS can diffuse into the ZnO NWs or the CIS layer. Spatial elemental analysis using electron energy loss spectroscopy (EELS) was also performed (Fig. 2c–h). The results conrmed that the Cu, In, S elements were uniformly deposited on the CdS/ZnO nanowire. XRD measurements were used to investigate the crystal structures of CdS/ZnO NWs and CIS/CdS/ZnO NWs. As shown in Fig. 3, the X-ray diffraction q–2q scan shows the various ZnO peaks, including the main peak of ZnO (00l) at 34.4 . These results indicate that individual ZnO NWs were vertically aligned
Fig. 3 XRD patterns of the CdS/ZnO NW and CIS/CdS/ZnO NW structures.
8916 | Nanoscale, 2014, 6, 8914–8918
Fig. 4 (a) Light absorbance spectra of the as-prepared ZnO NW, CdS/ ZnO NW, CdS/CIS/ZnO NW and CIS/CdS/ZnO NW. (b) Photocurrent density–voltage curves obtained from various photoanodes; asprepared ZnO, CdS/ZnO, CIS/ZnO, CdS/CIS/ZnO and CIS/CdS/ZnO NW arrays.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Paper
layer was deposited onto ZnO NWs, its absorption range was extended up to 500 nm. Finally, as the CIS absorber layer was deposited, the whole UV-visible light region was absorbed. The optical band gap (Eg) can be estimated from the absorption spectra, and the calculated direct band gap of the CIS absorber is approximately 1.60 eV. This value is slightly larger than that of the bulk CIS (1.54 eV) caused by its quantum connement effect. In the case of reversing the order of deposition of CdS and CIS (CdS/CIS/ZnO NW), the absorption was decreased, which is believed to be due to a reduction in the amount of CIS deposition. CdS could act as a seed layer to facilitate the deposition of CIS and thus the amount of deposited CIS is larger on CdS than that on bare ZnO NWs. Fig. 4(b) shows a set of chopped sweep current density versus potential curves obtained from various NW photoanodes using a white light illumination of 100 mW cm2 (AM 1.5 G). All of the measurements were carried out in a three-electrode electrochemical cell, with a Pt wire and a saturated calomel electrode as the counter and the reference electrode, respectively. An aqueous solution of 0.25 M Na2S and 0.35 M Na2SO3 (pH 12.5) was used as an electrolyte and a sacricial reagent, respectively. The bare ZnO NW shows negligible photocurrent generation because of the narrow absorption region. The enhanced photocurrent was observed with the CdS/ZnO NW due to the improved visible light absorption via the CdS absorber layer, but the photocurrent density was still rather low. Aer the CIS layer deposition on the CdS/ZnO NW, the photocurrent value was signicantly increased. Furthermore, unlike typical CIGS PEC cells, which showed cathodic photocurrent under illumination, our CIS electrode generated an anodic photocurrent, indicating that the prepared CIS has an n-type character because the type of CIS can be determined according to stoichiometry.22 A maximum photocurrent density of 13.8 mA cm2 at 0.3 V bias vs. SCE was achieved from the CIS/CdS/ZnO NW sample. This result indicated that the band structure of the CIS/ CdS/ZnO NW is suitable for charge carrier transfer, which is discussed further below in more detail. To conrm the effect of the deposition of the CdS layer on the photocurrent, the photocurrent change was conrmed according to the deposition sequence. While n-type materials were employed on the surface of the electrode in most applications of CIGS to PEC water splitting to improve its photoelectrochemical properties, in our case of the CdS/CIS/ZnO NW, it generated a lower photocurrent compared to the CIS/CdS/ZnO NW, despite only reversing the deposition sequence. This result showed that the band structure of the CIS/CdS/ZnO NW has an advantage of charge transfer from the CIS absorber layer to the ZnO NW compared to that of the CdS/CIS/ZnO NW. Furthermore, the CdS layer can mediate the lattice mismatch between CIS and ZnO. The lattice mismatch between the CIS layers and CdS is smaller than that between the CIS layers and the ZnO NW.25 Thus the lattice mismatch, which can cause strain or stress at the interface, is minimized when the CdS layer is inserted between CIS and ZnO. In addition, as shown in UV-vis spectra, the absorption was enhanced due to the increase in CIS deposition caused by the CdS layer acting as a seed layer. This result can be conrmed by the I–V results of the CIS/ZnO NW
This journal is © The Royal Society of Chemistry 2014
Nanoscale
structure. It exhibited not only a lower photocurrent density but also a higher dark current than the CdS/ZnO NW. This indicates poor heterojunction formation of CIS/ZnO, which is mainly due to the absence of a seed layer; therefore, a large uncoated surface of the ZnO NW was le. To obtain a better understanding of the detailed band structure of heterostructure NWs, we recorded ultraviolet photoelectron spectra (UPS) of CIS/CdS/ZnO and CdS/ZnO NWs, as shown in Fig. 5. Fig. 5(a) and (b) present the secondary electron cutoff (Ecutoff) and valence band maximum (EVBM) in UPS of the CdS and CIS layers. The valence band maximum (VBM) energy levels are calculated from the UPS data using the following formula:27 VBM ¼ hn (Ecutoff EVBM) where hn (21.22 eV) is the incident photon energy and EVBM is the onset relative to the Fermi level (EF) of Au, in which EF is determined from the Au substrate. The VBM energy levels for CIS and CdS were 5.52, and 6.52 eV, respectively. The conduction band minimum (CBM) levels of CIS and CdS can be estimated from the VBM energy levels and the optical band gap. The calculated CBM energy levels of CIS and CdS were 3.92 eV and 4.08 eV, respectively. Energy band diagrams based on these data can be found in Fig. 5(d). The energy levels of CIS, CdS and ZnO are summarized in Table 1.
(a) Secondary edge region of ultraviolet photoelectron spectra (UPS) of CdS and CIS. (b) Low energy onset region of UPS spectra of CdS and CIS. (c) Schematic diagram of the CIS/CdS/ZnO NW array in the PEC cell. (d) Electronic structures of ZnO, CdS and CIS.
Fig. 5
Nanoscale, 2014, 6, 8914–8918 | 8917
View Article Online
Nanoscale
Paper
Published on 26 May 2014. Downloaded by University of Waikato on 14/07/2014 07:21:52.
Table 1 Energy levels of the CdS and CIS layers determined using UVvis absorption spectra and UPS
Electrode
Eg (eV)
VBM (eV)
CBM (eV)
CIS CdS ZnO (ref. 26)
1.60 2.44 3.31
5.52 6.52 7.55
3.92 4.08 4.24
From the results, we can see that the CIS/CdS/ZnO NW electrode has a type-II cascade band structure that facilitates photogenerated carrier transport from CIS to the ZnO NW. This ideal band structure of the CIS/CdS/ZnO NW is suitable for improving the photocurrent generation of PEC cells. However, in the case of the CdS/CIS/ZnO NW, the photogenerated electrons in the CIS layer transfer to not only the ZnO NW but also the CdS layer and thus recombination increased between photogenerated electrons and the oxidizing species in the electrolyte.
Conclusions We have successfully developed a CIS/CdS/ZnO NW photoanode for PEC cells using a solution-based process. ZnO NWs provided a direct charge transfer path and an inserted CdS layer between CIS absorbers, and the ZnO NW yields a cascade of type-II band structures and serves as a seed layer to facilitate the deposition of the CIS layer. The synthesized CIS/CdS/ZnO NW photoelectrode produced a high photocurrent density of 13.8 mA cm2 at 0.3 V bias vs. SCE for hydrogen generation.
Acknowledgements This work was supported by the National Research Foundation of Korea (2013-R1A2A2A05-005344).
Notes and references 1 A. Fujishima and K. Honda, Nature, 1972, 238, 37–38. 2 M. Gratzel, Nature, 2001, 414, 338–344. 3 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori and N. S. Lewis, Chem. Rev., 2010, 110, 6446–6473. 4 K. Sivula, F. L. Formal and M. Gratzel, ChemPlusChem, 2011, 4, 432–449. 5 M. S. Seol, H. J. Kim, W. S. Kim and K. J. Yong, Electrochem. Commun., 2010, 12, 1416–1418.
8918 | Nanoscale, 2014, 6, 8914–8918
6 W. W. Zhan, Q. Kuang, J. Z. Zhou, X. J. Kong, Z. X. Xie and L.-S. Zheng, J. Am. Chem. Soc., 2013, 135, 1926–1933. 7 H. J. Kim and K. J. Yong, ACS Appl. Mater. Interfaces, 2013, 5, 13258–13264. 8 T. Bak, J. Nowotny, M. Rekas and C. C. Sorrell, Int. J. Hydrogen Energy, 2002, 27, 991–1022. 9 H. C. Ye, H. S. Park, V. A. Akhavan, B. W. Goodfellow, M. G. Panthani, B. A. Korgel and A. J. Bard, J. Phys. Chem. C, 2011, 115, 234–240. 10 J. Xia, Y. Liu, X. Qiu, Y. Mao, J. He and L. Chen, Mater. Chem. Phys., 2012, 136, 823–830. 11 L. Rovelli, S. D. Tilley and K. Sivula, ACS Appl. Mater. Interfaces, 2013, 5, 8018–8024. 12 D. Yokoyama, T. Minegishi, K. Maeda, M. Katayama, J. Kubota, A. Yamada, M. Konagai and K. Domen, Electrochem. Commun., 2010, 12, 851–853. 13 J. H. Kim, T. Minegishi, J. Kobota and K. Domen, Energy Environ. Sci., 2012, 5, 6368. 14 J. Suna, S. K. Batabyal, P. D. Tranb and L. H. Wong, J. Alloys Compd., 2014, 591, 127–131. 15 Y. Tang, Y. H. Ng, J. H. Yun and R. Amal, RSC Adv., 2014, 4, 3278. 16 A. Tapley, D. Vaccarello, J. Hedges, F. Jia, D. A. Loveb and Z. Ding, Phys. Chem. Chem. Phys., 2013, 15, 1431. 17 F. Yang, V. Kuznietsov, M. Lublow, C. Merschjann, A. Steigert, J. Klaer, A. Thomase and T. S. Niedriga, J. Mater. Chem. A, 2013, 1, 64072. 18 M. A. Contreras, B. Egaas, K. Ramanathan, J. Hiltner, A. Swartzlander, F. Hasoon and R. Nou, Prog. Photovoltaics, 1999, 7, 311. 19 J. F. Jiang, Thin Solid Films, 2006, 515, 1950. 20 S. Mandati, B. V. Sarada, S. R. Dey and S. V. Joshi, J. Renewable Sustainable Energy, 2013, 5, 031602. 21 S. Ikeda, M. Nonogaki, W. Septina, G. Gunawan, T. Harada and M. Matsumura, Catal. Sci. Technol., 2013, 3, 1849. 22 J. H. Yun, Y. H. Ng, S. Huang, G. Conibeerb and R. Amal, Chem. Commun., 2011, 47, 11288–11290. 23 M. Moriya, T. Minegishi, H. Kumagai, M. Katayama, J. Kubota and K. Domen, J. Am. Chem. Soc., 2013, 135, 3733–3735. 24 Y. J. Tak and K. J. Yong, J. Phys. Chem. B, 2005, 109, 19263– 19269. 25 K. A. Jones, J. Cryst. Growth, 1979, 47, 235–244. 26 K. Yuan, L. Chen, F. Lia and Y. Chen, J. Mater. Chem. C, 2014, 2, 1018. 27 J. H. Seo, R. Yang, J. Z. Brzezinski, B. Walker, G. C. Bazan and T. Q. Nguyen, Adv. Mater., 2009, 21, 1006–1011.
This journal is © The Royal Society of Chemistry 2014
