With the urgency of both the energy crisis and the aggravation of environmental pollution, the search for an alternative fuel source that is clean, environmental-friendly and reproducible has received increasing attention. As the cleanest energy, solar cells are a renewable energy that is being developed as an alternative source to traditional fossil fuel-based sources. During the last ten years, the most successful oxide solar cells were based on heterostructures of Cu₂O. In 2011, interest was renewed due to the conversion efficiency increasing up to 3.8% of Cu₂O/ZnO solar cells in the material for photovoltaic applications.^[1,2] Cu₂O is a promising p-type semiconductor with a direct band gap (E_g) of 2.17 eV. Its common p-type conductivity is due to copper vacancies with an ionization energy of 0.28 eV.^[3] Zinc oxide (ZnO) is an intrinsically doped n-type semiconductor with a direct band gap of 3.37 eV.^[4] In the past years, the conversion efficiency of Cu₂O/ZnO-based cells has been improved significantly. So far, the highest achieved conversion efficiency of Cu₂O/ZnO-based cells is only 4.13%,^[5] even though an upper limit of 18% is theoretically predicted.^[6] Generally, the efficiency enhancement is associated with the suppression of the recombination current, which is due to a large conduction band discontinuity at the interface.^[7] In the case of the Cu₂O/ZnO heterojunction, the small conduction band offset (CBO), which is required by high-efficiency solar cells, has been a research focus.^[8] Several groups have reported on regulating the band alignment of Cu₂O/ZnO interfaces.^[9–11] However, defects leading to a considerable modification of energy band alignment are often formed during the fabrication of the interface. This is especially true for the Cu₂O/ZnO heterointerface, where the presence of metallic Zn, metallic Cu or CuO can lead to a modification of the energy band alignment. Several methods have been employed for fabricating Cu₂O/ZnO heterointerfaces, including the electrochemical deposition method,^[12,13] radio-frequency magnetron sputtering,^[14] an ion beam sputtering method,^[15] and low-damage deposition methods.^[16,17] But up to now, few studies have focused on the fabrication of a Cu₂O/ZnO heterojunction with laser-molecular beam epitaxy (L-MBE). In this study, we report on the growth of Cu₂O/ZnO heterojunction by L-MBE. The energy band alignment of the Cu₂O/ZnO heterojunction is systematically investigated by in situ x-ray photoelectron spectroscopy.

Cu₂O/ZnO films were grown on c-Al₂O₃ substrates using L-MBE. Sintered ZnO (purity, 99.9995%) and CuO (purity, 99.9%) ceramic targets were ablated by a KrF excimer laser (248 nm, 350 mJ, 1 Hz, and 25 ns). The substrates were mounted 5 cm away from the target in an ultra-high vacuum chamber with a base pressure prior to 2×10⁻⁷ Pa. The investigation of the energy band alignment was carried out using in situ x-ray photoelectron spectroscopy (XPS). The XPS spectra were performed in an ultra-high vacuum chamber (5×10⁻⁸ Pa) equipped with an unmonochromatized Al source (1486.6 eV). The layer thickness was precisely controlled by the product of the deposition time and deposition rate, which were predetermined by single-layer deposition experiments.

In this study, three sets of samples were prepared in order to clarify the electronic band configuration of the Cu₂O/ZnO heterojunction: (i) 100-nm-thick ZnO film grown on a c-Al₂O₃ substrate to measure the valence band maximum (VBM) and Zn 2p core level (CL), named the ZnO sample; (ii) 100-nm-thick Cu₂O/100-nm-thick ZnO films grown on a c-Al₂O₃ substrate to measure the VBM and Cu 2p CL, called the Cu₂O sample; (iii) 10 nm-thick Cu₂O/100-nm-thick ZnO films grown on a c-Al₂O₃ substrate to determine the difference in the CLs of Cu 2p and Zn 2p, referred to as the Cu₂O/ZnO heterojunction sample. The x-rays could go through the 10-nm-thin Cu₂O film but could not pass through the 10-nm-thick Cu₂O film. The XPS spectra were accurately calibrated by the C 1s peak (284.6 eV) to compensate for the charge effect, and the total energy resolution of this XPS system was less than 0.05 eV, which usually affects the measured kinetic energy of photoelectrons.^[18]

Krautʼs method is a direct and useful tool to determine the valence band offsets (VBOs), and has been used to measure the valence band discontinuities of heterojunctions.^[19–21] This CL photoemission-based method was employed for semiconductor/semiconductor heterojunctions.^[22] However, due to the insulating properties of dielectrics, neutralization differential charging and Fermi edge decoupling effects were not negligible for heterostructures with dielectrics.^[23,24] Based on Krautʼs method, the VBO of the Cu₂O/ZnO interface can be described by the following formula^[22] 1 where 2 is the energy difference between Zn 2p and Cu 2p, which are measured in the heterojunction sample; and are the values of the CL energy and VBM from the Cu₂O sample; and are the values of the CL energy and VBM from the ZnO sample, respectively.

To investigate the interface of the heterostructure sample, a cross-sectional scanning electron microscope (SEM) image and energy dispersive spectrometer (EDS) analysis were obtained and are displayed in Fig. 1. It is obvious that the ZnO film has been grown on a c-Al₂O₃ substrate, followed by Cu₂O film deposition. The element analysis of the region marked by the white rectangle in the SEM image is carried out by EDS, and shown in Figs. 1(c)–1(e). Figure 1(b) displays the Cu₂O/ZnO interface region observed in a high-resolution transmission electron microscope (HRTEM). Although the interface between Cu₂O and ZnO is not very clear, two different compounds are clearly distinguished by the red dashed line. The Cu₂O crystallizes into a cubic structure with parameter , while the interplanar distance is 0.30 nm corresponding to the distance between the two adjacent Cu₂O (110) crystal planes. The interplanar distance is 0.26 nm corresponding to the distance between the two adjacent (0 0 0 2) crystal planes of ZnO with hexagonal wurtzite structure.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (a) Cross-section SEM image, (b) HRTEM of Cu2O/ZnO interface, (c)–(e) cross-section element mapping by EDS of Cu2O/ZnO heterojunction sample.

Figure 2 shows the XRD scan pattern of 100 nm-Cu₂O/100 nm-ZnO/c-Al₂O₃ grown on a c-Al₂O₃ substrate. Besides the c-plane Al₂O₃ substrate diffraction peak at 41.71°, only the pronounced (0 0 0 2) and a (0 0 0 4) peak of ZnO in the XRD pattern are observable at 34.35° and 72.30°, which shows that the ZnO film is of hexagonal wurtzite structure, and highly c-axis oriented without any impurity phases. The peak at 61.23° can be assigned to Cu₂O (220), which reveals a preferred orientation of (110) for Cu₂O. Therefore, it can be concluded that the Cu₂O/ZnO heterojunction is grown on a c-Al₂O₃ substrate with the orientation relationship of Cu₂O(110) ZnO (0002) Al₂O₃ (0001).

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. XRD scan patterns of 100-nm-Cu2O/100 nm-ZnO/c-Al2O3.

All the core-level peaks are fitted using a Shirley background and Voigt (mixed Lorentzian–Gaussian) line shapes. The binding energy for each of the CL peaks is precisely taken as the energy corresponding to the maximum intensity. Figure 3(a) shows the XPS survey scan of the Zn 2p_3/2 CL of the ZnO sample, while the peaks located at 1021.35 eV and 1044.13 eV are assigned to the electronic states of Zn 2p_3/2 and Zn 2p_1/2, respectively. The Zn element exists only in the oxidized state, which confirms that no evidence of the metallic Zn peak is observed. In addition, the peak of Zn 2p_3/2 CL has a symmetric shape, implying a uniform Zn–O bonding state. The valence band spectrum of 100 nm-ZnO/c-Al₂O₃ is shown in Fig. 3(b) and is used to measure the VBM position. The VBM position with respect to the surface Fermi level is obtained by the intersection of linear fits to the leading edge of the valence band photoemission and the ground.^[25–27] This linear method has already been widely used to determine the VBM of semiconductors with an accuracy of about ± 0.05 eV.^[28,29] Focusing on the inset of Fig. 3(b), it is worth noting that the peak at ∼0 eV in the inset is a distinct satellite feature induced by Al , which is marked by the red solid square. The VBM value of the ZnO sample relative to E_F is determined to be 2.44 ± 0.05 eV.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. XPS survey scan of 100 nm-ZnO/c-Al2O3 (a) Zn 2p3/2 CL spectrum; (b) valence band spectrum, with inset showing VBM determination.

The XPS scan spectrum of 100 nm-Cu₂O/100 nm-ZnO/c-Al₂O₃ is shown in Fig. 4(a). The Cu 2p_3/2 and 2p_1/2 peak are observed at 932.4 eV and 952.3 eV, respectively. To identify the chemical state of the Cu element in the Cu₂O sample, XPS and x-ray-excited Auger spectra (XAES) are measured. The absence of characteristic shakeup satellites of CuO located at ∼9 eV, which is higher than the binding energy of the main 2p_3/2 and 2p_1/2 peak, indicates that the Cu species should not be Cu²⁺.^[30] Then, Cu⁰ and Cu¹⁺ could be easily distinguished from the XAES of Cu L₃VV. The characteristic kinetic energy values of the different oxidation states of Cu are shown in Ref. [30]. The major Cu LVV Auger peak with kinetic energy of 916.3 eV is presented in Fig. 4(b), which is attributed to Cu¹⁺. It has been reported that the Cu Auger spectrum has a distinct satellite feature at a kinetic energy of ∼2.5 eV higher than the main peak.^[30–32] The absence of the satellite peak of Cu in the presented curves also implies that the Auger peaks are associated with Cu¹⁺, which is consistent with the XRD result in Fig. 2. Besides, this conclusion is further supported by the appearance of a weak peak at around 965 eV (marked by a red rectangle in Fig. 4(a)), which is in good agreement with the satellite peak position of Cu₂O.^[31] Like the ZnO sample, the VBM value of the Cu₂O sample relative to E_F is determined to be 0.14 ± 0.05 eV, which is illustrated by Figs. 4(c) and 4(d).

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. (a) XPS survey scan of Cu 2p1/2 and Cu 2p3/2 CL spectrum, (b) XAES of Cu LVV, (c) valence band spectrum, and (d) VBM determination of 100 nm-Cu2O/100 nm-ZnO/c-Al2O3.

Figure 5(a) and 5(b) show the Zn 2p_3/2 CL spectrum and Cu 2p_3/2 CL spectrum of 10 nm-Cu₂O/100 nm-ZnO/c-Al₂O₃. The peak of the Zn 2p_3/2 CL of the Cu₂O/ZnO heterojunction located at 1021.00 eV is shifted by 0.35 eV compared with that of the ZnO sample. Correspondingly, the Cu 2p_3/2 peak located at 932.19 eV is shifted by 0.02 eV compared with that of the Cu₂O sample.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. (a) Zn 2p3/2 CL spectrum and (b) Cu 2p3/2 CL spectrum of 10 nm-Cu2O/100 nm-ZnO/c-Al2O3.

For clarity, all of the peak energy values of CL and VBM achieved from 100 nm-ZnO/c-Al₂O₃, 100 nm-Cu₂O/100 nm-ZnO/c-Al₂O₃, and 10 nm-Cu₂O/100 nm-ZnO/c-Al₂O₃ are summarized in Table 1. By inserting the relevant peak parameters into Eq. (1), the VBO of the Cu₂O/ZnO heterojunction is calculated to be 1.97 eV. Given the optical band gap of 3.37 eV for ZnO and 2.17 eV for Cu₂O, we obtain a type-II staggered band alignment with a CBO of 0.77 eV as shown schematically in Fig. 6. Different values for the CBO in the Cu₂O/ZnO heterojunction have been reported.^[33–35] The main improvement is due to the reduction of the CBO, which leads to an increase in the open circuit voltage.^[7] In our study, a CBO of 0.77 eV is achieved, which is smaller than these earlier results. As is well known, the open circuit voltage (V_oc) is determined by the difference in electrochemical potential between the hole and electron. In the energy band alignment of the pCu₂O-nZnO heterojunction, the thermodynamic limit for V_oc is approximately deduced by subtracting the CBM of ZnO from the VBM of Cu₂O. In our study, the thermodynamic limit for V_oc is calculated to be 1.4 eV, much higher than the experimental result at a low range of 0.2 V–0.8 V,^[36–38] which is also of benefit to the efficiency enhancement of Cu₂O/ZnO solar cells.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. Band alignment for the heterostructure p-Cu2O/n-ZnO.

Table 1.

Table 1.

Table 1. Peak positions of core level and VBM used to calculate VBO of p-Cu2O/n-ZnO heterojunction. . Sample Region Binding energy n-ZnO Zn 2p3/2 1021.35 VBM 2.44 p-Cu2O Cu 2p3/2 932.21 VBM 0.14 p-Cu2O/n-ZnO Zn 2p3/2 1021.00 Cu 2p3/2 932.19

Table 1. Peak positions of core level and VBM used to calculate VBO of p-Cu2O/n-ZnO heterojunction. .

In this work, we successfully fabricated a high-quality p-Cu₂O/n-ZnO heterojunction. By in situ XPS, the VBO of the Cu₂O/ZnO heterojunction is determined to be 1.97 eV. Given the band gap difference of 1.20 eV between two materials, this translates into a staggered interface band alignment with a CBO of 0.77 eV. The determination of the band alignment is beneficial to understanding the electronic transport mechanism and improving the photoelectrical properties of Cu₂O/ZnO solar cells.