Long-term stabilized high-density CuBi2O4/NiO heterostructure thin film photocathode grown by pulsed laser deposition

Jongmin Lee a, Hongji Yoon a, Seungkyu Kim a, Sehun Seo a, Jaesun Song a, Byeong-Uk Choi b, Seung Yo Choi c, Hyunwoong Park c, Sangwoo Ryu d, Jihun Oh be and Sanghan Lee *a
aSchool of Materials Science and Engineering, Gwangju Institute of Science and Technology, Gwangju 61005, Republic of Korea. E-mail: sanghan@gist.ac.kr
bDepartment of Materials Science and Engineering, Korea Advanced Institute of Science and Technology, Daejeon 34113, Republic of Korea
cSchool of Energy Engineering, Kyungpook National University, Daegu 41566, Republic of Korea
dDepartment of Advanced Materials Engineering, Kyonggi University, Suwon, Gyeonggi-do 16227, Republic of Korea
eGraduate School of Energy, Environment, Water, and Sustainability (EEWS), Korea Advanced Institute of Science and Technology, Daejeon 34141, Republic of Korea

Received 6th August 2019 , Accepted 9th September 2019

First published on 9th September 2019


Harvesting sustainable hydrogen through water-splitting requires a durable photoelectrode to achieve high efficiency and long lifetime. Dense, uniform CuBi2O4/NiO thin film photocathodes grown by pulsed laser deposition achieved photocurrent density over 1.5 mA cm−2 at 0.4 VRHE and long-term chronoamperometric stability for over 8 hours.


Copper-based p-type oxides are attractive photocathode materials owing to their relatively high charge carrier mobility and narrow bandgap for solar energy conversion through photoelectrochemical (PEC) water splitting.1 However, Cu-based p-type oxides demonstrate weak photostability and severe photocorrosion owing to the redox potential position of CuO and Cu2O within their bandgaps in the electrolyte.2 To overcome these disadvantages of binary copper oxide photocathodes, Cu-based ternary oxide materials, such as CuAlO2 and CuFeO2, are suggested.3,4

Among Cu-based ternary oxides, CuBi2O4 (CBO) is the most promising p-type photocathode material in the p/n-PEC tandem cell because of its higher photocurrent onset potential of ∼1.0 VRHE and narrow bandgap of ∼1.8 eV.5–7 In particular, the photocurrent onset potential of CBO is relatively higher than that of conventional photocathodes (such as Si and Cu2O).8,9 However, the poor internal charge transport properties of CBO owing to low hole mobility (∼1.2 × 10−3 cm2 V−1 s−1) hinder efficient photo-induced carrier separation and result in the recombination of photogenerated charge carriers in the CBO photocathode.7 To overcome sluggish hole transport, various strategies are employed in CBO thin film photocathodes. For example, the presence of Ag ions increases hole concentration in the CBO thin film,10 and the Cu-doped NiO (Cu:NiO) layer has a significant role as hole transport layer in the CBO/Cu:NiO heterojunction thin film.11 Nevertheless, the CBO photocathodes do not reach the theoretical photocurrent density of 18–20 mA cm−2 and demonstrate severe electron/hole photocorrosion.12

For a systematic investigation with the aim of overcoming the poor charge transport properties of the CBO thin film photocathode, it is also essential to fabricate a reliable CBO thin film. Because it is well known that the point defects (e.g., pores and voids) within the grains obstruct carrier conduction in oxide thin films,13 structural defects such as porous and non-uniform structure,14 which reduce the photocurrent of photoelectrodes, must be minimized. In other words, a reliable CBO thin-film synthesis technique is necessary for the fabrication of durable p-type photocathodes.

Efforts to fabricate compact, dense, and homogeneous thin films in CBO photocathodes using solution-based deposition techniques such as spray pyrolysis with slow deposition rate15 and polymer-assisted spin coating have been reported.16 Nevertheless, CBO thin film photocathodes grown by solution-based thin film deposition still exhibit declining PEC performance of photoelectrodes such as increase in the dark current through exposure of the electrode to electrolytes or diminishing chronoamperometric spectra.11,15 Because the annealing process is indispensable for removing the solvents of precursors in the solution-based thin-film deposition technique, defects such as secondary phases or voids occur in these photocathodes. Thus, it is essential to fabricate dense and crystalline CBO thin films without defects.

The pulsed laser deposition (PLD) technique is a promising physical vapor deposition (PVD) method that can be employed to fabricate dense thin films without requiring the post-annealing process. Recently, dense complex metal oxide thin films grown using PLD have been investigated for fundamental PEC properties of the photoelectrode such as optical absorption, charge carrier transport, and chemical stability.17,18

In addition to the formation of dense thin films, PLD also offers the advantage of depositing sharp interfaces between heterojunctions. In particular, fabricating staggered (type-II) heterojunctions using PLD has been perceived as an effective strategy to enhance the charge transport efficiency in photoelectrodes. Various type-II heterojunction of photoelectrodes such as BiVO4/WO3,19,20 BiVO4/Bi4V2O11,21 Fe2O3/ZnO,22 and MoS2/WS2/WSe2 thin-film catalysts8 grown using PLD demonstrate enhanced photocurrent. This suggests that high interfacial quality contributes to suitable band alignment between multilayers.

In this work, we fabricated single-layer CBO and CBO/NiO heterostructure thin film photocathodes grown on a fluorine-doped SnO2 (FTO) glass substrate using PLD. High-density (≥85%) bulk CBO and NiO ceramic targets were used for dense and homogeneous CBO and CBO/NiO thin film growth (for further experimental details, see the ESI). Fig. 1(a) shows X-ray diffraction (XRD) patterns for the CBO/NiO and CBO thin films deposited at a growth temperature of 450 °C on the FTO glass substrate. The XRD patterns of both CBO/NiO and CBO thin films show only CBO peaks without secondary phases such as those of Bi2O3, CuO, and Cu2O. It is also confirmed that the metal phases such as Cu, Bi and CuBi alloy are not observed in CBO/NiO, CBO thin films and bulk CBO ceramic target in Fig. S1(a) (ESI). The polycrystalline CBO phases are formed at a temperature range 400 to 450 °C. In particular, the β-Bi2O3(201) peak appears at a growth temperature of 500 °C with decrease in the CBO peak intensity (Fig. S1(b), ESI). The NiO peaks are not visible owing to low peak intensity when the layer of NiO is thin (∼50 nm). However, the NiO(111) peak appears clearly at 2θ ∼ 37.15° with layer thickness of ∼200 nm in Fig. S1(c) (ESI). Fig. 1(b) and (c) show the cross-sectional scanning electron microscopy (SEM) images of the CBO (800 nm)/NiO (50 nm) and CBO (800 nm) thin films, respectively. Both the CBO/NiO and CBO thin films grown by PLD are dense, uniform, and have flat surfaces in contrast to those grown using solution-based thin film deposition techniques such as drop casting,7,23 spray pyrolysis,11,15,24 and electrodeposition.10,25 A uniform and flat surface is advantageous in forming well-connected CBO/NiO heterojunctions between CBO and NiO without significant amounts of pores and voids, as shown in Fig. 1(b).


image file: c9cc06092h-f1.tif
Fig. 1 (a) X-ray diffraction θ–2θ scans. (b), (c) cross-sectional scanning electron microscopy (SEM) images of CuBi2O4 (800 nm)/NiO (50 nm) thin film and CuBi2O4 (800 nm) thin film on FTO (fluorine-doped SnO2) glass, respectively. (d) Schematic of CuBi2O4/NiO heterostructure thin film photocathode.

To investigate the optical properties depending on the surface morphology of both CBO/NiO and CBO thin films, we acquired top-view SEM and atomic force microscopy (AFM) topography images of the thin films and FTO glass substrate, as shown in Fig. S2(a)–(d) (ESI). The optical absorbance characteristics of CBO and CBO/NiO thin films are almost the same owing to their similar surface morphology, as shown in Fig. S2(i), (j) (for the detailed explanation, see the page S6, ESI).

To further realize the mass density of CBO and CBO/NiO thin films, we measured the critical angle (θc) to perform quantitative analysis using the X-ray reflectivity (XRR) technique (Fig. S3 in the ESI). We found that the CBO and CBO/NiO thin films exhibit the same θc = 0.77°. The mass density (ρ) is derived using the following equation:26

image file: c9cc06092h-t1.tif
where A is the atomic weight, Z is the atomic number, and θc is the critical angle. The mass density of the samples calculated from θc using equation nearly corresponds to the theoretical mass density of kusachiite CBO.27 The calculated mass density of the NiO thin film is also quite comparable to the theoretical density of the rock-salt NiO (Table S1, ESI).28 This indicates that the CBO thin films grown by PLD have much higher density compared to those of solution-based CBO thin films.7 Hence, it was inferred that the dense CBO and CBO/NiO thin films might exhibit improved durability as photocathodes. In addition, we performed ultraviolet photoelectron spectroscopy (UPS), X-ray photoelectron spectroscopy (XPS), and ultraviolet-visible spectroscopy (UV-vis) measurements to investigate the energy band structure of the CBO/NiO heterojunction in Fig. 2. Note that p-type NiO, which has a wide bandgap of 3.7 eV and relatively low electron affinity,29 is a good candidate material for the hole transport and electron blocking layer to form staggered, type-II band alignment with the CBO main layer.11


image file: c9cc06092h-f2.tif
Fig. 2 (a) Work function as a function of kinetic energy and (b) difference between Fermi energy and valence band maximum as a function of binding energy. (c) Tauc plots derived from absorption data of NiO and CuBi2O4 thin films. Note that the indirect (r = 2) and direct (r = 1/2) transition of Tauc plots of (αhν)1/r are used in CuBi2O4 and NiO thin films, respectively. (d) Schematic of energy-band diagrams in equilibrium. CB and VB indicate the conduction band and valence band, respectively.

In Fig. 2(a), the work functions (from vacuum energy level to Fermi energy level) of the CBO (4.63 eV) and NiO (4.67 eV) thin films are determined using the x-intercepts for the extrapolated lines for spectra with low kinetic energy cut-off values (Vcut-off) using UPS measurement. Moreover, the difference between the Fermi energy level and valence band maximum is also determined from the x-intercepts of the extrapolated lines of the spectra at low binding energies in Fig. 2(b) (0.62 eV and 0.34 eV for the CBO and NiO thin films, respectively). To estimate the conduction band minimum of the CBO and NiO thin films, Tauc plots are used to identify the optical bandgap of the films from absorption data (For the absorption coefficients, see Fig. S4 in the ESI). The optical bandgaps of the CBO and NiO thin films are 1.7 (indirect) and 3.7 eV (direct), respectively, as shown in Fig. 2(c). Combining with UPS, XPS, and UV-vis data, the energy band structures of the CBO and NiO layers are presented, respectively, in Fig. S5 (ESI) As demonstrated by a previous report,11 we confirmed the constructing type-II band alignment between the CBO and NiO layers. NiO template layer serves as a blocking layer of photogenerated electrons in CBO layer because conduction band edge position of NiO template layer is negatively higher than that of CBO layer. In addition, the valence band edge position of NiO template layer is also negatively higher than that of CBO layer as indicated in the Fig. 2(d). Therefore, efficient charge separation and suppression of electron–hole recombination rate are expected in the bulk CBO main layer.

To examine the overall PEC performance of uniform and high-density CBO and CBO/NiO thin film photocathodes, we first measured photocurrent as a function of potential. Fig. 3(a) shows chopped (light/dark) linear sweep voltammetry (LSV) curves with respect to the reversible hydrogen electrode (RHE, VRHE) for the CBO and CBO/NiO thin film photocathodes. The photocurrent densities of the CBO and CBO/NiO thin film photocathodes are 1.0 and 1.5 mA cm−2 at 0.4 VRHE, respectively. It should be noted that the usage of H2O2 as electron scavenger give rise to current-doubling effect due to carrier injection by following chemical reaction.30

H2O2 + 2H+ + e → OH + OH* + 2H+ → 2H2O + h+


image file: c9cc06092h-f3.tif
Fig. 3 (a) Chopped (light/dark) linear sweep voltammetry (LSV) curves. Photocurrent density (0.4 VRHE) of photocathodes depending on (b) single-layer CuBi2O4 thin film thickness and (c) NiO template layer thickness. (d) Electrochemical impedance spectroscopy (EIS) spectra in the frequency range 100 kHz–0.1 Hz. Inset shows the equivalent Randles circuit for the fitting of the EIS data. (e) Incident-photon-to-current conversion efficiency (IPCE) spectra of CuBi2O4 (800 nm) and CuBi2O4 (800 nm)/NiO (50 nm) photocathodes, respectively. (f) Chronoamperometry stability measurement. Note that the measurements were performed at 0.4 VRHE in 0.1 M potassium phosphate buffered solution with H2O2 (pH = 8.55).

The difference of photocurrent by addition of H2O2 indicate that the surface charge recombination dominate reaction kinetics of CBO photocathode. The thickness of each layer of CBO and NiO is optimized as 800 nm and 50 nm, respectively, by photocurrent density depending on layer thickness as shown in Fig. 3(b) and (c). (For LSV curves, see the Fig. S6 in ESI) We also confirmed that the photocurrent enhancement of CBO/NiO thin film photocathode do not come from NiO layer as shown in Fig. S7 (ESI). An increase of photocurrent density depending on the CBO thickness is deeply associated with improvement of crystalline quality at 600 nm thick CBO thin film in contrast to that of 300 and 400 nm thick CBO thin film (for the details, see Fig. S8 in the ESI). From electrochemical impedance spectroscopy (EIS) measurements, the fitting of EIS spectra for CBO and CBO/NiO thin film photocathodes in Fig. 3(d), and Table S2 in the ESI, it was confirmed that the charge transfer kinetics are improved in the CBO/NiO thin film photocathode, which corresponds to the LSV result and type-II band alignment between the CBO and NiO layer. Note that the equivalent circuits of both photocathodes are composed of the series resistance (Rs), constant phase element (CPE), and charge-transfer resistance (Rct) at the interface between the photocathode and the electrolyte. In addition, the efficiency of photon to electron conversion as a function of the illumination of wavelength was measured from the incident-photon-to-current conversion efficiency (IPCE) spectra using the following equation.

image file: c9cc06092h-t2.tif
where j is the photocurrent density and Iinc is the power density at a specific wavelength (λ). The CBO/NiO thin film photocathode exhibits high photoconversion efficiency compared to that of the CBO thin film photocathode, whereas the IPCEs of CBO and CBO/NiO thin film photocathodes are higher in the ultraviolet wavelength region.31 Note that an abrupt peak is observed at a wavelength of approximately 620 nm by second harmonic generation. Finally, we performed chopped light and dark chronoamperometry measurements for both CBO and CBO/NiO thin film photocathodes at constant 0.4 VRHE, as shown in Fig. 3(f). Interestingly, both CBO and CBO/NiO thin film photocathodes show a photocurrent of over 1 mA cm−2 without significant decay for 8 h, whereas the photocurrent density of the CBO/NiO thin film photocathode is higher than that of the CBO thin-film photocathode. To the best of our knowledge, most studies suggest that the stability test time scale is approximately 3 h owing to severe decay in the photocurrent density (see Table S3 in the ESI). We also performed stability test of CBO and CBO/NiO thin film photocathodes in the potassium phosphate buffered solution without the electron scavengers of H2O2 (see Fig. S9 in the ESI). The transient spike and overall poor photocurrent density indicate the surface charge trapping or sluggish surface reaction kinetics in CBO and CBO/NiO thin film photocathode. It should be noted that the usage of H2O2 enable improvement of poor surface reaction of CBO photocathode through carrier injection. In reality, CBO photocathodes demonstrate severe photocorrosion or dissolution.10 We believe that surface engineering to protect photoelectrode are challenging issue in CBO photocathodes. In terms of photoelectrode stability with H2O2, dark current of the CBO/NiO thin film photocathode did not increase, whereas that of the CBO thin film photocathode increased slightly. Such a discrepancy might arise from the existence of the NiO intermediate layer, which hinders the exposed FTO electrode from the electrolyte. Furthermore, the oxidation states of Cu and Bi in the CBO and CBO/NiO thin film photocathodes were examined before and after the stability test using X-ray photoelectron spectroscopy (XPS), as shown in Fig. S10 (ESI). The predominance of Cu2+ peaks was maintained in the PEC-tested samples, implying that chemically stable CBO and CBO/NiO thin films were grown by PLD. We have found that high-density CBO and CBO/NiO thin film photocathodes sustain the oxidation state of Cu2+ of CBO, thereby improving the photostability.

In conclusion, we fabricated high-density CBO and CBO/NiO thin film photocathodes grown by PLD, which is a useful PVD technique to grow dense and crystalline thin films. The PLD-grown CBO thin film photocathodes exhibit very long-term stabilized PEC performance at 0.4 VRHE without any substantial decay for 8 h. In particular, the photocurrent density (1.5 mA cm−2) of the CBO/NiO thin film photocathode is higher than that (1.0 mA cm−2) of the CBO thin film, which was demonstrated by the type-II band alignment. We expect that the formation of copper-based p-type oxide type-II heterojunction with other semiconducting materials via PLD or other PVD methods would enhance the efficiency of p-type photocathodes.

This research was supported by the Creative Materials Discovery Program (2017M3D1A1040834) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT. The authors acknowledge H.-J. Jung and Prof. S.-Y. Chung (Korea Advanced Institute of Science and Technology) for fabrication of the CuBi2O4 target.

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. K. Zhang, M. Ma, P. Li, D. H. Wang and J. H. Park, Adv. Energy Mater., 2016, 6, 1600602 CrossRef.
  2. A. Paracchino, V. Laporte, K. Sivula, M. Grätzel and E. Thimsen, Nat. Mater., 2011, 10, 456 CrossRef CAS PubMed.
  3. I. Sullivan, B. Zoellner and P. A. Maggard, Chem. Mater., 2016, 28, 5999 CrossRef CAS.
  4. Y. J. Jang and J. S. Lee, ChemSusChem, 2019, 12, 1835 CrossRef CAS PubMed.
  5. T. Arai, M. Yanagida, Y. Konishi, Y. Iwasaki, H. Sugihara and K. Sayama, J. Phys. Chem. C, 2007, 111, 7574 CrossRef CAS.
  6. N. T. Hahn, V. C. Holmberg, B. A. Korgel and C. B. Mullins, J. Phys. Chem. C, 2012, 116, 6459 CrossRef CAS.
  7. S. P. Berglund, F. F. Abdi, P. Bogdanoff, A. Chemseddine, D. Friedrich and R. Van De Krol, Chem. Mater., 2016, 28, 4231 CrossRef CAS.
  8. S. Seo, S. Kim, H. Choi, J. Lee, H. Yoon, G. Piao, J. C. Park, Y. Jung, J. Song, S. Y. Jeong, H. Park and S. Lee, Adv. Sci., 2019, 6, 190031 Search PubMed.
  9. W. Septina, R. R. Prabhakar, R. Wick, T. Moehl and S. D. Tilley, Chem. Mater., 2017, 29, 1735 CrossRef CAS.
  10. D. Kang, J. C. Hill, Y. Park and K. S. Choi, Chem. Mater., 2016, 28, 4331 CrossRef CAS.
  11. A. Song, P. Plate, A. Chemseddine, F. Wang, F. F. Abdi, M. Wollgarten, R. Van De Krol and S. P. Berglund, J. Mater. Chem. A, 2019, 7, 9183 RSC.
  12. Z. Chen, T. F. Jaramillo, T. G. Deutsch, A. Kleiman-Shwarsctein, A. J. Forman, N. Gaillard, R. Garland, K. Takanabe, C. Heske, M. Sunkara, E. W. McFarland, K. Domen, E. L. Miller, J. A. Turner and H. N. Dinh, J. Mater. Res., 2010, 25, 3 CrossRef CAS.
  13. S. K. Choi and J. I. Lee, J. Vac. Sci. Technol., A, 2002, 19, 2043 CrossRef.
  14. F. E. Osterloh, Chem. Soc. Rev., 2013, 42, 2294 RSC.
  15. F. Wang, W. Septina, A. Chemseddine, F. F. Abdi, D. Friedrich, P. Bogdanoff, R. van de Krol, S. D. Tilley and S. P. Berglund, J. Am. Chem. Soc., 2017, 139, 15094 CrossRef CAS PubMed.
  16. N.-W. Kim, B.-U. Choi, H. Yu, S. Ryu. and J. Oh, Opt. Express, 2019, 27, 171 CrossRef PubMed.
  17. K. Harbauer, P. Plate, K. Ho, S. P. Berglund, D. Friedrich, R. Van De Krol and F. F. Abdi, Chem. Mater., 2018, 30, 8322 CrossRef.
  18. R. A. Henning, P. Uredat, C. Simon, A. Bloesser, P. Cop, M. T. Elm and R. Marschall, J. Phys. Chem. C, 2019, 123, 18240 CrossRef CAS.
  19. J. Song, J. Cha, M. G. Lee, H. W. Jeong, S. Seo, J. A. Yoo, T. L. Kim, J. Lee, H. No, D. H. Kim, S. Y. Jeong, H. An, B. H. Lee, C. W. Bark, H. Park, H. W. Jang and S. Lee, J. Mater. Chem. A, 2017, 5, 18831 RSC.
  20. C. Nguyen Van, T. H. Do, J.-W. Chen, W.-Y. Tzeng, K.-A. Tsai, H. Song, H.-J. Liu, Y.-C. Lin, Y.-C. Chen, C.-L. Wu, C.-W. Luo, W.-C. Chou, R. Huang, Y.-J. Hsu and Y.-H. Chu, NPG Asia Mater., 2017, 9, e357 CrossRef CAS.
  21. J. Song, K. S. Choi, M. J. Seo, Y. R. Jo, J. Lee, T. L. Kim, S. Y. Jeong, H. An, H. W. Jang, B. J. Kim, C. Jeon and S. Lee, Chem. Mater., 2018, 30, 5673 CrossRef CAS.
  22. L. T. Quynh, C. N. Van, W. Y. Tzeng, C.-W. Huang, Y.-H. Lai, J.-W. Chen, K.-A. Tsai, C. L. Wu, W.-W. Wu, C. W. Luo, Y.-J. Hsu and Y.-H. Chu, ACS Appl. Energy Mater., 2018, 1, 3900 CrossRef CAS.
  23. J. Li, M. Griep, Y. Choi and D. Chu, Chem. Commun., 2018, 54, 3331 RSC.
  24. F. Wang, A. Chemseddine, F. F. Abdi, R. Van De Krol and S. P. Berglund, J. Mater. Chem. A, 2017, 5, 12838 RSC.
  25. S. Liu, J. Zhou, Y. Lu and J. Su, Sol. Energy Mater. Sol. Cells, 2018, 180, 123 CrossRef CAS.
  26. Y.-W. Chung, Introduction to Materials Science and Engineering, CRC Press, 2006 Search PubMed.
  27. C. Henmi, Mineral. Mag., 1995, 59, 545 CrossRef CAS.
  28. S. A. Makhlouf, Thin Solid Films, 2008, 516, 3112 CrossRef CAS.
  29. M. D. Irwin, D. B. Buchholz, A. W. Hains, R. P. H. Chang and T. J. Marks, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 2783 CrossRef CAS.
  30. H. Dotan, K. Sivula, M. Grätzel, A. Rothschild and S. C. Warren, Energy Environ. Sci., 2011, 4, 958 RSC.
  31. G. Sharma, Z. Zhao, P. Sarker, B. A. Nail, J. Wang, M. N. Huda and F. E. Osterloh, J. Mater. Chem. A, 2016, 4, 2936 RSC.

Footnotes

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9cc06092h
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2019