Uniform ZnO nanorod/Cu₂O core–shell structured solar cells by bottom-up RF magnetron sputtering

Abstract

Cu₂O is a good candidate material for use as a p-type absorber in solar cells. Here, uniform ZnO nanorod (NR)–Cu₂O core–shell structures are fabricated and their diode performances are studied. ZnO NRs are grown on fluorine-doped tin oxide (FTO) glass using a hydrothermal method. Cu₂O is then deposited on the ZnO NRs using bottom-up RF magnetron sputtering. The crystal structures of the deposited ZnO NRs and Cu₂O are characterized using X-ray diffraction. From secondary electron microscopy analysis, the uniform core–shell structure and its size are identified. UV-vis spectroscopy measurements show that the optical bandgap of the Cu₂O in this structure is 2.3 eV. The diode characteristics of the fabricated nanostructures depend on the thickness of Cu₂O; 2.7 μm-thick Cu₂O on ZnO NRs shows diode properties. Lastly, we propose a band alignment model based on X-ray photoelectron spectroscopy analysis and demonstrate a possible approach for fabricating Cu_xO–ZnO nanohybrids for further improvements to device efficiency, highlighting a need for interfacial band offset medication for oxide heterojunction solar cells.

---

1. Introduction

Renewable energy technology has become increasingly important for solving current issues pertaining to environmental pollution and energy resource usage. Among the current renewable energy motifs, solar cells are quite appealing because they utilize the unlimited and eco-friendly energy provided by sunlight.¹ Current Si-based solar cells still have high production costs in spite of recent technical developments. Therefore, alternative solar materials with low cost and high efficiency are needed and are being developed for future solar cells, including those involving the highly abundant Cu.^2–5 Cu_xO (x: 1–2) is nontoxic, has good visible light absorption with a bandgap of 1–2 eV, and can be used for p-type light absorber layers.^6–9 ZnO can be used as a counterpart of Cu_xO, as it is a suitable n-type semiconductor material able to form a p/n junction with Cu_xO. ZnO has a large bandgap (∼3.3 eV) and therefore is being utilized as a window layer for thin film solar cells.^10,11 Previous studies on other oxide solar cells utilized planar structure junctions,^12–14 but recently there has been an increasing interest in three-dimensional (3D) nanostructure junctions that increase the interfacial area for improving the efficiency of solar cells.^15–18

In this study, uniform ZnO nanorod (NR)–Cu₂O core–shell structured solar cells are fabricated and characterized. ZnO NRs are grown on fluorine-doped tin oxide (FTO) using a hydrothermal method, and Cu₂O is deposited on the ZnO NRs using RF magnetron sputtering. Atomic layer deposition (ALD) is generally used to make uniform core shell structures,^19,20 but it is difficult to deposit thicker Cu₂O films (>1 μm) as absorber layers using ALD. Therefore, in this study, bottom-up sputtering is used to obtain thick uniform Cu₂O shell structures on ZnO NRs in an effort to create 3D interfacial p/n junctions. The diode characteristics and solar cell properties are evaluated, and the interfacial band alignment is analyzed.

2. Experimental

ZnO NR/Cu₂O core–shell structures were fabricated by a two-step process involving ZnO NR hydrothermal synthesis and Cu₂O sputtering. Vertically aligned ZnO NRs were synthesized on FTO (Sigma-Aldrich, 7 Ω), which was used as the transparent conductive substrate. First, a 200 nm-thick ZnO seed layer was deposited using RF magnetron sputtering and annealed at 400 °C for 1 h. This seed layer was then used to grow NRs using a hydrothermal method. 25 mM Zn(NO₃)₂ (Sigma-Aldrich, 99%) and hexamethylenetetramine (HMT) (Sigma-Aldrich, 99.5%) were stirred for 1 h to obtain a transparent solution. This solution was autoclaved with a clean FTO substrate placed vertically along the Teflon wall of the autoclave. Deposition was carried out at 90 °C for 4 h. After deposition, the ZnO-deposited FTO was rinsed with deionized water,^21,22 then annealed at 400 °C for 1 h. Bottom-up RF magnetron sputtering was used to deposit Cu₂O on the ZnO NRs. Cu₂O layers were deposited at 200 °C using a 2-inch copper target (99.9% purity) in various Ar : O₂ (50 : 0–50 : 1.5) reactive mixtures. The applied RF power was 150 W under 10 mTorr working pressure.²³

The morphology and core–shell structures of the samples were measured by field emission scanning electron microscopy (SEM) (JEOL JSM-6700F Japan, Hitachi S-4300SE Japan). X-ray diffraction (XRD) measurements were obtained using a Rigaku MiniFlex-II desktop X-ray diffractometer (Japan). Ultraviolet-visible (UV-vis) absorption spectra and optical bandgaps were recorded using an Avantes spectroscopy system with an AvaLight-DH-S-BAL balanced power source and AvaSpec-ULS2048 Starline versatile fiber-optic spectrometer detector (Netherlands). Surface chemical binding states, valence band-edge states, and depth profiles were measured by X-ray photoelectron spectroscopy (XPS) (theta probe base system, Thermo Fisher Scientific Co. USA). Diode properties of the samples were characterized in the probe station by measuring electrical conductivity using a Keithley 4200-SCS apparatus (USA). Diode properties were measured with linear sweep from 0 V to −10 V and from 0 V to 10 V consecutively by the step of 0.1 V. Diode cells were made by depositing Ni in 100 μm dots on the ZnO NR/Cu₂O using an e-beam evaporator. IPCE analysis was performed at the probe station using a homemade optical setup using monochromatic light from a 300–1000 nm form xenon lamp. A standard Si solar cell (BS-500, BUNKOUKEIKI Co., Japan) was used to calculate the IPCE. The photocurrent response under 1 sun conditions was measured using an Ivium technologies Ivium-n-Stat (Netherlands) with a constant applied potential of 0 V by blocking and passing the light.

3. Results and discussion

During sputtering, the oxygen stoichiometry x in Cu_xO is easily tunable. When using bottom-up RF magnetron sputtering, controlling the Ar : O₂ ratio (50 : 0–50 : 1.5) forms different copper oxides with varying x values. Fig. 1 shows XRD patterns of the synthesized ZnO NRs on FTO and sputtered Cu_xO samples with varying x ratios. Cu₂O was obtained at a 50 : 0.5 sccm Ar : O₂ ratio (JCPDS-01-071-3645), and this ratio was used to make our nanostructures.

Fig. 1 XRD results of (a) ZnO NRs and (b) Cu₂O film as a function of various Ar : O₂ ratios during RF sputtering (c) ZnO NRs/Cu₂O structure.

The surface morphology and thickness of the deposited Cu₂O film were analyzed using SEM. Fig. 2 shows a cross-sectional SEM image of Cu₂O deposited on FTO at different deposition times. Cu₂O thickness is dependent on sputtering time based on these data, and the growth rate was estimated to be ∼30 nm min⁻¹. In addition, the deposited films have no pores or voids, and hence are suitable for use as diode junctions. Fig. 3 shows cross sectional SEM images of Cu₂O deposited on ZnO nanorods. From these, the amount of Cu₂O on ZnO can be easily controlled by the sputtering process. After 30 min of deposition time, a thick layer of Cu₂O is already formed. The overall thickness of the nanostructure is 816 nm. At this deposition time, the deposited Cu₂O is uniform across the entire ZnO nanorod. When the deposition time is increased to 1.5 h, a much thicker Cu₂O layer is formed (approximately 2.7 μm). Under these conditions, Cu₂O is deposited more heavily on top of the nanorod compared to the bottom. These deposition conditions preferentially fill the upper side of the nanorod structure, leaving empty regions at the bottom. This demonstrates the advantage of the present method in controlled deposition of Cu₂O by RF magnetron sputtering.

Fig. 2 Cross-sectional SEM images of Cu₂O films as a function of deposition time at (a) 10 min (286 nm), (b) 30 min (816 nm), (c) 1 h (1.8 μm), and (d) 1 h 30 min (2.7 μm).

Fig. 3 Cross-sectional SEM images of uniformly distributed Cu₂O on ZnO nanorods at (a) and (b) with 816 nm-thick Cu₂O deposition for 30 min and (c) and (d) with 2.7 μm-thick Cu₂O deposition for 90 min.

The optical absorbances of the prepared heterojunctions were measured. UV-vis analysis was conducted to measure the light absorbance of the ZnO NRs and Cu₂O-deposited samples (Fig. 4). ZnO NRs on glass show a bright gray color and highly visible light transmittance. The Cu₂O deposited sample, however, shows a brown color and has high absorbance in the ultraviolet-visible range, with a wide band from 250 nm to 800 nm. This shows that the present heterojunction can effectively absorb visible light. As shown in the inset image in Fig. 4, Cu₂O-deposited ZnO NR samples deposited at 30 min with 286 nm thickness effectively absorb visible light and with relatively low transmission. From the optical absorbance data, the optical bandgap was plotted using Tauc plots (Fig. 5) as expressed by versus hν curves for the band alignment model. Tauc plots were plotted with n = 1/2 (in the case of direct bandgap) where α is the absorption coefficient calculated from absorption spectra and photon energy.^24,25 The optical bandgaps of ZnO and Cu₂O were 3.25 eV and 2.25 eV, respectively.

Fig. 4 UV-vis spectra of ZnO NRs and Cu₂O film (thickness = 286 nm) with digital images of the samples.

Fig. 5 The extracted optical bandgap of (a) ZnO and (b) Cu₂O from Tauc plots.

The fabricated ZnO NR/Cu₂O heterojunction was further analyzed to determine its I–V characteristics. From I–V measurements (Fig. 6(a)), it was confirmed that the device forms p–n junctions, showing diode rectifying ZnO as an n-type semiconductor,^26,27 and Cu₂O as a p-type semiconductor formed by controlling the oxygen stoichiometry during sputtering deposition. Because the properties of the final NR structure are dependent on the thickness of the deposited Cu₂O (for the diode), Cu₂O deposited for 10 min to a thickness of 286 nm has a limited local electrical connection in the upper region of the ZnO NR. Here, the current rises at positive potentials to a similar magnitude as the negative potential, suggesting that diode behavior was not achieved with 286 nm-thick Cu₂O. However, for the diode with a Cu₂O layer 2.7 μm thick, regular diode properties emerged. This diode characteristic of the device was mainly dependent of Cu₂O thickness which is investigated in this study and found out that the Cu₂O which the less than 1 micrometer did not showed convinced p–n junction characteristic while thickness more than 1 micrometer showed significant p–n junction property. Fig. 6(b) shows a rectification factor that is calculated by the ratio of reverse current (+bias polarity) to forward current (−bias polarity) as a function of applied voltage. Here, the rectification factor has a value between 0.14 and 90.57, and incrementally improved at high voltages. This bias-related change to the rectification factor is possibly due to the different charge emission mechanism as a function of voltage bias under a forward current. In particular, a potential difference predicted by a band-offset of the Cu₂O/ZnO conduction band and valence band is approximately −5 V at the interface, which matches with the forward current onset bias at −5 V in Fig. 6(a).²⁸

Fig. 6 (a) I–V curves of ZnO/Cu₂O core–shell structures with 286 nm and 2.7 μm Cu₂O thickness and (b) rectification factor as a function of applied bias.

Solar cell properties of the ZnO/Cu₂O nanostructures were evaluated by measuring the transient photocurrent and IPCE (incident photon-to-current efficiency) at 1 sun conditions (Fig. 7). Its properties were measured by on/off switching every 10 s in 1 sun conditions, in order to determine the reversibility of photo-carrier generation and recombination. From Fig. 7(a), no noticeable residual photocurrent lag was present during the on/off transition step, meaning that there is no significant photocurrent trapping in the diode. The approximate efficiency of the solar cells was calculated from IPCE measurements. From the IPCE (Fig. 7(b)) and absorbance (Fig. 4) data, we can define the effective visible light absorbance range (380–510 nm), generating a relatively high photocurrent with a maximum IPCE of approximately 4% at 406 nm. The comparison of IPCE values of the reported methods and our result are tabulated in Table 1.

Fig. 7 (a) Dark and photocurrent on/off under 1 sun conditions and (b) IPCE of ZnO/Cu₂O core–shell structured solar cell.

Table 1 Method and IPCE of ZnO/Cu₂O based solar cells

Materials	Method	IPCE (%)	Reference
Patterned ZnO NRAs/Cu2O	Hydrothermal/electrodeposition	58	15
ZnO/Cu2O planar	ALD/thermally oxidize	78	12
AZO/Cu2O planar	DC sputter/electro deposition	90	5
Cu2O/ZnO planar	Electrodeposition/electrodeposition	58	8
ZnO nanotube/Cu2O	Electrodeposition/electrodeposition	25	18
Cu2O/ZnO nanorod	Electrodeposition/hydrothermal	—	11
ZnO nanorod/Cu2O	Hydrothermal/RF magnetron sputter	4	Present work

---

The XPS depth profiles of the ZnO NR/Cu₂O heterojunctions were measured (Fig. 8). 30 nm Cu₂O was deposited on ZnO NRs for XPS measurement. The Ar sputtering conditions were 0.2 nm s⁻¹ deposition rate (referenced to SiO₂) and a sputtering area of 1 mm². As the etch time increased, the Zn 3d_5/2 binding energy at 9.23 eV and Zn 3p_3/2 binding energy at 88.8 eV increased, suggesting an approach of depth position to ZnO interface. In addition, the Cu 3p binding energy at 74.83 eV decreased as the etch time increased. In this XPS depth profile, a 5 s etch time was considered as bulk Cu₂O. Because of the nanostructure's geometric characteristics, the Cu peak is present during the first 120 s of etching, denoting the interface of the ZnO NR/Cu₂O heterojunction.²⁹ For the band alignment model, the valence band maximum (VBM) change was analyzed, showing a slight shift to a higher binding energy as etching occurred. No significant change in the VBM was observed, as the ZnO NR/Cu₂O interface continues due to longitudinal interface effects along the ZnO nanorods sidewalls.

Fig. 8 XPS depth profiles of (a) atomic% and (b) valence band spectra with etch time.

VBM was obtained separately for the planar Cu₂O film and ZnO NR to determine the valence band offset by XPS analysis. Fig. 9(a) shows VBM values of bulk Cu₂O and ZnO NRs as 0.5 eV and 1.75 eV, respectively. Using similar experimental extraction and optical bandgap conditions (Fig. 5), a band alignment model with conduction band offset (CBO), valence band offset (VBO) and band bending structure is suggested. VBO (ΔE_v) is calculated using Kraout's theory as shown in eqn (1) and (2)

ΔE_v = E^B_VB − E^A_VB + [(E^A_CL − E^A/B_CL) + (E^B/A_CL − E^B_CL)] (1)

ΔE_C = E^B_g − E^A_g + ΔE_v (2)

E^A_CL and E^B_CL are the core level energies of bulk material A and B. E^A/B_CL and E^B/A_CL are the same core level energies of A and B at the interface. The band alignment sketch is suggested using calculated values (Fig. 9(b)).^30,31

Fig. 9 (a) Valence band-edge XPS spectra of ZnO NRs and bulk Cu₂O and (b) electronic band alignment of ZnO NR/Cu₂O heterostructure constructed by XPS data.

From this band alignment model, we determined a high spike barrier height between the ZnO NRs and Cu₂O at the conduction band and the valence band. This small spike barrier is more suitable for solar cells than cliff barriers or no offset, but too high of a spike can disrupt the carrier flow and decrease the photocurrent efficiency.^32,33 From this, we could assume that this band structure has too high of a spike at the valence band and conduction band, and photogenerated currents become blocked at this spike. Therefore, this result strongly suggests that a significant modification of the interfacial band offset is needed for Cu₂O/ZnO heterojunction stacking in solar cell applications. This task may be related to the surface control of ZnO (i.e., O- or H-polar surface termination)^34,35 and Cu₂O (i.e., CuO subphase formation at the surface).³⁶

4. Conclusions

In this work, ZnO/Cu₂O core–shell nanostructures were fabricated and their diode and solar cell characteristics were investigated. Well-aligned ZnO nanorods were grown on FTO using a hydrothermal method, followed by RF magnetron sputtering to deposit Cu₂O. By using bottom-up RF magnetron sputtering, thickness-controlled uniform deposition of Cu₂O was easily achieved by varying the deposition time. In addition, oxygen stoichiometry in the resulting film can be easily tuned by adjusting the deposition conditions, as confirmed by XRD measurements. ZnO/Cu₂O cells show diode rectifying properties as confirmed by I–V measurements, certifying p/n junction formation. Photocurrent and IPCE measurements under 1 sun conditions demonstrated suitable solar cell performance. We analyzed the band alignment using VB-edge XPS spectra and optical bandgaps to determine the cause of the low efficiency in sputtered Cu₂O layers in solar cells. From our results, we demonstrated that Cu₂O can be easily deposited with a desired oxygen stoichiometry using bottom-up RF magnetron sputtering. We suggest that there is significant room for improving the conversion efficiency by modifying the interfacial band alignments.

Acknowledgements

This research was supported by the Basic Science Research Program (NRF-2014R1A2A1A11053174) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning, Korea.

References

1. P. V. Kamat, J. Phys. Chem. C, 2007, 111, 2834–2860.
2. J. J. Loferski, J. Appl. Phys., 1956, 27, 777–784.
3. C. Wadia, A. P. Alivisatos and D. M. Kammen, Energy Environ. Sci., 2009, 43, 2072–2077.
4. J. Luo, L. Steier, M.-K. Son, M. Schreier, M. T. Mayer and M. Grätzel, Nano Lett., 2016, 16, 1848–1857.
5. M. Abdelfatah, J. Ledig, A. El-Shaer, A. Wagner, V. Marin-Borras, A. Sharafeev, P. Lemmens, M. M. Mosaad, A. Waag and A. Bakin, Sol. Energy Mater. Sol. Cells, 2016, 145(3), 454–461.
6. A. Mittiga, E. Salza, F. Sarto, M. Tucci and R. Vasanthi, Appl. Phys. Lett., 2006, 88, 163502.
7. I. Sullivan, B. Zoellner and P. A. Maggard, Chem. Mater., 2016 DOI:10.1021/acs.chemmater.6b00926.
8. A. S. Zoolfakar, R. A. Rani, A. J. Morfa, S. Balendhran, A. P. O'Mullane, S. Zhuiykov and K. Kalantar-zadeh, J. Mater. Chem., 2012, 22, 21767–21775.
9. A. S. Zoolfakar, R. A. Rani, A. J. Morfa, A. P. O'Mullane and K. Kalantar-zadeh, J. Mater. Chem. C, 2014, 2, 5247–5270.
10. K. P. Musselman, A. Wisnet, D. C. Iza, H. C. Hesse, C. Scheu, J. L. MacManus-Driscoll and L. Schmidt-Mende, Adv. Mater., 2010, 22, E254–E258.
11. N. G. Elfadill, M. R. Hashim, K. M. Chahrour, M. A. Qaeed and M. Bououdina, Superlattices Microstruct., 2015, 85, 908–917.
12. Y. Ievskaya, R. L. Z. Hoye, A. Sadhanala, K. P. Musselman and J. L. MacManus-Driscoll, Sol. Energy Mater. Sol. Cells, 2015, 135, 43–48.
13. S. S. Jeong, A. Mittiga, E. Salza, A. Masci and S. Passerini, Electrochim. Acta, 2008, 53, 2226–2231.
14. M. Izaki, T. Shinagawa, K.-T. Mizuno, Y. Ida, M. Inaba and A. Tasaka, J. Phys. D: Appl. Phys., 2007, 40, 3326.
15. X. Chen, P. Lin, X. Yan, Z. Bai, H. Yuan, Y. Shen, Y. Liu, G. Zhang, Z. Zhang and Y. Zhang, ACS Appl. Mater. Interfaces, 2015, 7, 3216–3223.
16. Y.-K. Hsu, H.-H. Lin, J.-R. Wu, M.-H. Chen, Y.-C. Chen and Y.-G. Lin, RSC Adv., 2014, 4, 7655–7659.
17. X. Jiang, Q. Lin, M. Zhang, G. He and Z. Sun, Nanoscale Res. Lett., 2015, 10, 30.
18. M. Abd-Ellah, J. P. Thomas, L. Zhang and K. T. Leung, Sol. Energy Mater. Sol. Cells, 2016, 152, 87–93.
19. V. Ganapathy, B. Karunagaran and S.-W. Rhee, J. Power Sources, 2010, 195, 5138–5143.
20. H. Kim, S.-H. Kim, K. Y. Ko, H. Kim, J. Kim, J. Oh and H.-B.-R. Lee, Electron. Mater. Lett., 2016, 12, 404–410.
21. S.-Y. Liu, T. Chen, J. Wan, G.-P. Ru, B.-Z. Li and X.-P. Qu, Appl. Phys. A, 2009, 94, 775–780.
22. Z. Kang, X. Yan, Y. Wang, Z. Bai, Y. Liu, Z. Zhang, P. Lin, X. Zhang, H. Yuan, X. Zhang and Y. Zhang, Sci. Rep., 2015, 5, 7882.
23. S. H. Lee, S. J. Yun and J. W. Lim, ETRI Journal, 2013, 35, 1156–1159.
24. Y. Nakano, S. Saeki and T. Morikawa, Appl. Phys. Lett., 2009, 94, 022111.
25. B. D. Viezbicke, S. Patel, B. E. Davis and D. P. Birnie, Phys. Status Solidi B, 2015, 252, 1700–1710.
26. Ü. Özgür, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Doğan, V. Avrutin, S.-J. Cho and H. Morkoç, J. Appl. Phys., 2005, 98, 041301.
27. V. K. Sahu, P. Misra, R. S. Ajimsha, A. K. Das and B. Singh, Mater. Sci. Semicond. Process., 2016, 54, 1–5.
28. M. Yang, L. Zhu, Y. Li, L. Cao and Y. Guo, J. Alloys Compd., 2013, 578, 143–147.
29. C. Zhu, A. Osherov and M. J. Panzer, Electrochim. Acta, 2013, 111, 771–778.
30. K. Eom, S. Kim, D. Lee and H. Seo, RSC Adv., 2015, 5, 103803–103810.
31. E. A. Kraut, R. W. Grant, J. R. Waldrop and S. P. Kowalczyk, Phys. Rev. B: Condens. Matter Mater. Phys., 1983, 28, 1965–1977.
32. I. Konovalov and R. Szargan, Appl. Phys. Lett., 2003, 82, 2088–2090.
33. S. Sharbati and J. R. Sites, IEEE Journal of Photovoltaics, 2014, 4, 697–702.
34. J. V. Lauritsen, S. Porsgaard, M. K. Rasmussen, M. C. R. Jensen, R. Bechstein, K. Meinander, B. S. Clausen, S. Helveg, R. Wahl, G. Kresse and F. Besenbacher, ACS Nano, 2011, 5, 5987–5994.
35. S. Blumstengel, N. Koch, S. Sadofev, P. Schäfer, H. Glowatzki, R. L. Johnson, J. P. Rabe and F. Henneberger, Appl. Phys. Lett., 2008, 92, 193303.
36. S. W. Lee, Y. S. Lee, J. Heo, S. C. Siah, D. Chua, R. E. Brandt, S. B. Kim, J. P. Mailoa, T. Buonassisi and R. G. Gordon, Adv. Energy Mater., 2014, 4, 1301916.

---