Cu₂O Heterojunction Solar Cell with Photovoltaic Properties Enhanced by a Ti Buffer Layer

Abstract

In this study, semiconductor oxide cuprite (Cu₂O) and indium tin oxide (ITO) heterojunction solar cells with and without a 10 nm thick titanium (Ti) thin film as the buffer layer were fabricated and characterized for comparison. The Cu₂O film was formed by low-cost electrodeposition, and Ti and ITO layers were deposited on a glass substrate by sputtering. The interfacial microstructures, surface topology, and electrical and photovoltaic properties of both solar cells were investigated. The test results showed that the Ti buffer layer changed the surface morphology, resistivity, and contact potential of the electrodeposited Cu₂O film. With these changes, the photovoltaic performances of the Cu₂O/Ti/ITO solar cell including open-circuit voltage (V_OC) and short-circuit current (I_SC) were all enhanced compared to the Cu₂O/ITO solar cell, and the power conversion efficiency was improved from 1.78% to 2.54%. This study offers a promising method to improve the efficiency of Cu₂O-based solar cells for sustainability in material resource, environment and eco-system, and energy production.

1. Introduction

As a renewable, safe and clean source of power, solar energy plays an important role in sustainability by reducing greenhouse gas emissions and water usage and protecting the environment and ecosystems from energy production, and mitigating climate change and improving air quality. Solar cell device technologies based on a variety of materials [1,2,3,4,5] have been developed to harvest solar energy, including silicon, cadmium telluride, copper indium gallium selenide, III-V semiconductors, copper oxide, organic materials, quantum dots, perovskites, etc. Among these materials, cuprous oxide (Cu₂O, cuprite) has been recognized as a promising semiconductor for device applications since the 1920s, which was long before silicon, germanium and other potential semiconducting materials were discovered. However, due to the difficulty in n-type doping, the interest in Cu₂O declined after single crystal silicon and germanium dominated in the field since they can be doped in both n- and p-types. Research interests in Cu₂O for photovoltaics have revived since the 1970s [6,7] due to the need to develop terrestrial photovoltaic devices for solar energy conversion driven by the first oil crisis of 1973. The research was at first focus on Schottky junction PV devices with a variety of techniques to produce Cu₂O films on copper such as thermal oxidation, electrodeposition, and anodic and chemical oxidation. However, the power conversion efficiency has been limited to no more than 2% due to dangling bonds or the reduction of Cu₂O to form a Cu-rich region at the interface because of oxygen diffusion into the Schottky metal [8], which keeps the Schottky barrier height in the range of 0.7 to 0.9 eV regardless of the choice of Schottky metal. Therefore, later the focus was shifted to the heterojunction structure to improve the performance of Cu₂O PV devices. The interest in Cu₂O PV devices has grown after the 1990s to find suitable transparent n-type semiconductors for heterojunction formation and develop low damage junction formation techniques. As a native p-type semiconducting oxide [9], Cu₂O has a bulk direct bandgap energy of 1.9~2.1 eV [8,10] which is favorable for good solar spectral absorption and energy conversion. Other important characteristics [11] include high hole-mobility of 50~100 cm²/Vs at room temperature [12,13,14], a large carrier diffusion length of 2~12 μm [8], a high career concentration of 10¹⁶~10¹⁹ cm⁻³ [15,16,17], chemically stable, non-toxic, abundant in nature, a simple and inexpensive fabrication process, etc.

Techniques to prepare Cu₂O films on conducting substrates to form a heterojunction structure include thermal oxidation [18], anodic oxidation [19] and chemical oxidation [20] of Cu films, reactive magnetron sputtering [21,22], electrochemical deposition [23], epitaxial grown via pulsed laser deposition [24], etc. Recently, a 8.4% power conversion efficiency [25] of a Cu₂O solar cell was reported with the Cu₂O thin film deposited by reactive sputtering in a mixed O₂ and Ar atmosphere on a transparent conducting oxide, which is composed of double layers of SnO₂:Sb (ATO) and ITO on a glass substrate. Among these techniques, considerable interest has been shown in the electrochemical preparation of Cu₂O thin films by electrodeposition, and this process has been applied on metals or glass substrates coated with high conducting and transparent semiconductors such as ITO, SnO₂, In₂O₃, etc. [26,27,28]. Compared to other techniques, electrodeposition by dissolving the precursor of copper ions in deionized water for controlled deposition of the Cu₂O thin films is versatile, environmental benign, cost-effective, operated at room temperature, and flexible for depositing on both rigid and flexible conducting substrates of a large area. The keys to control the growth, texture, grain size and shape, film thickness and surface morphology [29,30,31,32] in electrodeposition are by adjusting process parameters such as current density and voltage between the electrodes, deposition temperature and time, the composition of the electrolytic solution, etc. Solar cells fabricated by employing the electrodeposition of Cu₂O films have been reported on Cu₂O/ZnO/ITO, glass/Cr/ITO/Cu₂O/Zn_1−xMg_xO/AZO, FTO/Cu₂O/AZO, etc. [33,34,35,36,37]. It was also reported that a thin gold layer in 80–150 nm thickness deposited before the electrodeposition of Cu₂O [38] can resolve reproducibility issues for an electrodeposited Cu₂O thin film on ITO or Si substrates.

In this paper, we developed a novel method that is feasible and simple to improve photovoltaic properties and power conversion efficiency by adding a Ti buffer layer in the solar cell. Photovoltaic properties of Cu₂O and ITO heterojunction solar cells with the Cu₂O film deposited by electrodeposition and with and without a Ti buffer layer between the Cu₂O and ITO film were investigated. The Cu₂O/Ti/ITO and Cu₂O/ITO junction interface, surface morphology of the Cu₂O film on Ti/ITO and ITO substrate, and photovoltaic properties of Cu₂O/Ti/ITO and Cu₂O/ITO solar cells were characterized. ITO is a transparent n-type semiconductor with a large bandgap of around 4 eV [39]. The p-type Cu₂O and n-type ITO form a heterojunction diode. Test results showed that the Ti buffer layer added in between the Cu₂O and ITO film improved the growth of a uniform Cu₂O film to form a better heterojunction, which reduced the carrier diffusion and recombination at the junction interface, and therefore improved the photovoltaic properties and conversion efficiency of the Cu₂O/Ti/ITO solar cell. The solar cells in this study were fabricated on a glass slide, with ITO and Ti films deposited by sputtering and a Cu₂O film by electrodeposition; both are inexpensive processes. Such a simple and low-cost fabrication process combined with improved photovoltaic properties make Cu₂O/Ti/ITO a promising heterojunction solar cell for photovoltaic applications for sustainable energy production.

2. Method

Both the Cu₂O/Ti/ITO and Cu₂O/ITO heterojunction structures are illustrated in Figure 1. Two glass microscope slides (2.5 cm × 2.5 cm) were used as the substrates. They were cleaned in acetone, isopropyl alcohol and DI water for 10 min in each step by immersing in an ultrasonic bath. After cleaning and drying by nitrogen gas, the glass substrates were loaded in a Kurt Lesker Nano36 sputter system for deposition of a 0.5 µm thick indium tin oxide (ITO) film as the anode electrode using an ITO target (In₂O₃/SnO₂, 90/10 wt%, 99.99% purity). After ITO film formation, one of the two substrates was saved for further processing of Cu₂O electrodeposition, while the other substrate was loaded in the sputter system again and deposited a 10 nm thick Ti film on the ITO film using a Ti target (99.995% purity) in Ar ambient. The ITO surface was partially covered with aluminum foil during Ti deposition to leave an area open for cathodes. At this point, these two glass substrates, with one covered by an n-type ITO semiconductor layer and the other covered by a Ti/ITO bilayer, were prepared for comparison.

Followed by the preparation of the substrate, n-type ITO, and Ti layer, the electrodeposition of Cu₂O layers on both substrates was carried out. The preparation of an aqueous precursor solution for electrodeposition of the Cu₂O layer started by mixing 100 mL of DI water, 50 mL of lactic acid (C₃H₆O₃, 90% concentration), and 20 g of copper sulfate (Cu₂SO₄). Cupric sulfate was slowly dissolved in the solvent with stirring and gentle heat at 50 °C. After the cupric sulfate was fully dissolved, the solution presented a light blue color. Once copper sulfate was fully dissolved, the solution was heated at 80 °C to evaporate DI water until the volume of the solution was reduced down to 85 mL with little or no crystallization in the solution. The solution was still a light blue color. Finally, the pH of the final bath was adjusted to 10 by adding a potassium hydroxide (KOH, 6 mol/L, 99.97%, Sigma Aldrich) solution to the above solution drop by drop and constantly stirring the solution. The solution appeared in the dark blue color at neutral when pH is at 7, and then became slightly purple when the pH reaches 10. In general, pH values greater than 9 are required to obtain a p-type Cu₂O, and the acidity of the solution is critical since it affects the native point defects and copper vacancies [17,40] and therefore the conductivity of the final electrodeposited Cu₂O film. Such pH dependence also makes the electrodeposition a flexible approach. The precursor solutions were always freshly prepared before each electrodeposition step.

Prior to each electrodeposition, cyclic voltammograms were performed to record the oxidation and reduction of the electrolyte species on the electrode surface and the potential window for the Cu₂O deposition. The electrodeposition experiment was conducted at room temperature and without stirring the solution. A Cu wire was used as the anode during the electrodeposition. The constant current density was set at 5 mA/cm² between electrodes. Electrodeposition of Cu₂O by cathodic reduction was performed for 5 min with the temperature of the solution kept at 70 °C. Electrodeposition of Cu₂O consists of two reduction reactions for Cu²⁺ ions: from Cu²⁺ to Cu⁺ ions and then from Cu⁺ to metallic Cu⁰, as in Equations (1) and (2), followed by the produced Cu⁺ ions reacting with OH⁻ ions to form Cu₂O [41] in the solution, as in Equation (3).

Cu²⁺ + e⁻ → Cu⁺

(1)

Cu⁺ + e⁻ → Cu

(2)

2Cu⁺ + 2OH⁻ → Cu₂O↓+ H₂O

(3)

After electrodeposition, the sample was rinsed with DI water and dried by N₂ gas, followed by annealing at 200 °C for 60 min in Ar ambient. After samples cool down to room temperature, silver paste was applied to form ohmic contacts on the Cu₂O film and ITO layer as electrodes to complete the solar cell fabrication.

Surface morphology of the electrodeposited Cu₂O on Ti/ITO and ITO substrate was characterized by a Nanosurf Flex-Axiom atomic force microscope (AFM) system in the contact mode. The interfacial structure and composition analysis was performed by inspecting the cross section of the samples in a FEI (currently Thermo Fisher Scientific, Hillsboro, OR, USA) TECNAI F20 TEM with a Gatan imaging filter 2001 and at an accelerating voltage of 200 kV. TEM cross-section samples were prepared using a FEI NanoLab 400S Dual-Beam FIB. A layer of black Sharpie marker was applied on the sample surface, followed by coating a thin Au film. An OMNIProbe 200 was used for in situ liftoff, and the cross-section laminas were cleaned at 8 kV and 25 pA. The current density vs. voltage (J-V) characteristics of Cu₂O/Ti/ITO and Cu₂O/ITO solar cells were recorded in darkness and under illumination by a Keithley 4200 semiconductor characterization system. An artificial light source of Oriel Sol1A solar simulator (Model 94021A) was used to mimic solar illumination. Since the n-type ITO layer has the wider energy gap in this heterojunction, light was incident through the glass slides to reach the junction as shown in Figure 1. The capacitance–voltage (C-V) measurements under reverse bias voltages were also performed by the Keithley 4200 semiconductor characterization system to characterize the contact potentials of the solar cells.

3. Results and Discussion

The electrodeposited Cu₂O film and sputtered Ti films on ITO were characterized by TEM and images of Cu₂O/Ti/ITO and Cu₂O/ITO heterojunction structures are illustrated in Figure 2, together with the elemental mapping images to show the distribution of all elements in both junctions. The thickness of the ITO layer was measured to be about 500 nm, and the electrodeposited Cu₂O films were about 900 nm thick on the Ti/ITO substrate and 700 nm on the ITO substrate. Elemental mapping analysis in Figure 2 shows that in both Cu₂O/Ti/ITO and Cu₂O/ITO structures, Si is confined in the glass substrate, Sn and In are confined in the ITO layers and Cu in the Cu₂O, while O spans across the ITO and Cu₂O films. As shown in the image for Ti mapping in Figure 2a, there is a continuous thin Ti buffer layer between ITO and Cu₂O. Also, there is no obvious inter-diffusion of elements in these heterojunction films.

The surface morphology of the as-deposited Cu₂O thin film on the ITO substrate and Ti/ITO substrate by electrodeposition was characterized using AFM. Figure 3 displays the 3D and planar views of the Cu₂O layers on ITO and ITO/Ti substrates, respectively. In this study, the deposition of the Cu₂O layer onto both ITO and ITO/Ti substrates by electrodeposition were highly reproducible. Also, the Cu₂O films on both substrates were grown in a direction perpendicular to the substrate surface, but with different surface morphologies. The Cu₂O film directly deposited on the ITO substrate showed granular-shaped grains of the Cu₂O crystals with clear grain boundaries and a uniform size of 1 µm in average. The average and root mean square roughness, R_a and R_rms, of the Cu₂O film on the ITO substrate were 52.7 nm and 65.6 nm, respectively. In comparison, the Cu₂O film on the Ti/ITO substrate shows pyramid-like grains of the Cu₂O crystals with nanometer-size grains, but without clear grain boundaries; instead, it has a wavelike corrugated surface feature. The average and root mean square roughness, R_a and R_rms, of the Cu₂O film on the Ti/ITO substrate are 29.9 nm and 36.8 nm, respectively. Each roughness value was the average of four measurements at different locations. It is well known in the nanoparticle research field that crystals with a nanostructured morphology are capable of maintaining the stable equilibrium configuration with minimal surface energy in particular synthesis processes [11]. Therefore, it is expected that the nanometer-sized Cu₂O crystals in the wavelike surface of the Cu₂O film on the Ti/ITO substrate will minimize the overall surface energy. The results from the comparison of surface morphology show that the Ti buffer layer reduces the surface roughness and grain size of the electrodeposited Cu₂O film. It is clear that in addition to electrodeposition time, current density, solution temperature, pH, etc., the underlying materials (ITO and Ti) also affect the growth and morphology of Cu₂O film, which in turn affects the grain shape and size as well as surface roughness.

J-V characterizations of the solar cells were carried out using an artificial light source with a power density of P_in = 25 mW/cm² measured by an Oriel reference cell and meter (Model 91150V) and a digital solar power meter (DBTU1300). Figure 4 shows the acquired J-V curves from Cu₂O/ITO and Cu₂O/Ti/ITO solar cells in the dark and under illumination conditions. Both solar cells showed rectification properties in the dark, with a reverse leakage current much smaller than the forward bias current. Under illumination, both Cu₂O/ITO and Cu₂O/Ti/ITO solar cells showed photovoltaic properties, and they lost most rectification properties and changed to Ohmic behaviors, in which the current density increases proportionally with the applied voltage. When illumination is through the n-type ITO side, photons are absorbed within the depletion region between the Cu₂O and ITO, with photogenerated electron–hole pairs within the depletion region and within the minority carrier diffusion length separated by the built-in electric field and drifting to the neutral regions, which contribute to the photovoltaic effect. It is known that the depletion region around the metallurgical junction absorbs the overall potential difference between the neutral n-type and p-type sides. When the excess electrons drift and reach the neutral ITO side and excess holes drift and reach the neutral Cu₂O side, an open circuit voltage V_OC develops between the Cu₂O and ITO terminals with the p-type Cu₂O side positive with respect to the n-type ITO side. In the J-V characteristics in Figure 4, V_OC is obtained from the intersection of the recorded current density and applied voltage axis, i.e., when the current through the solar cell is zero or the external circuit is open. It is the maximum voltage available from the solar cell. When the terminals of the solar cell device are shorted, the excess electrons in the n-type ITO side can flow through the external circuit to neutralize the excess holes in the p-type Cu₂O side. This photocurrent due to the flow of the photogenerated carriers is the short circuit current J_SC, which is obtained from the intersection of the recorded current density and current axis when the applied voltage across the solar cell is zero.

The fill factor (FF) and power conversion efficiency (η) are metrics used to characterize the performance of solar cells. The fill factor, which is a figure of merit for the solar cell and an essential measure of the closeness of the measured solar cell J-V curve to the ideal shape, is defined as the ratio of maximum power (P_max) divided by the product of V_OC and I_SC, i.e., $FF=\frac{P_{max}}{V_{OC}{I}_{SC}}$. The maximum power P_max indicates the condition under which the solar cell generates its maximum power and is calculated by $P_{max}=V_m\times I_m$, in which V_m and I_m are the voltage and current at the maximum power point, respectively, as labelled in Figure 4. The power conversion efficiency η defines the percentage of the solar energy illuminated on the solar cell that is converted into usable electricity, i.e., $\eta =\frac{P_{max}}{P_{in}}=\frac{V_{OC}{I}_{SC}FF}{P_{in}}$. The values of V_OC, J_SC, V_m, J_m, P_max, FF, and η from both Cu₂O/ITO and Cu₂O/Ti/ITO solar cells are summarized in

Table 1. Summary of the key photovoltaic properties.

Solar Cell	n	ϕi (V)	VOC (V)	JSC (mA/cm2)	Vm (V)	Jm (mA/cm2)	Pm (mW/cm2)	FF (%)	η (%)
Cu2O/ITO	1.21	0.74	0.26	6.6	0.13	3.43	0.45	25.8	1.78
Cu2O/Ti/ITO	1.64	1.01	0.37	7.14	0.18	3.53	0.64	24.2	2.54

. Compared to other solar cell systems [5], such as silicon, cadmium telluride, copper indium gallium selenide, quantum dots, and perovskites, etc., the power conversion efficiency η reported in this research is still lower mainly due to the fabrication process such as the parameters for electrodeposition and post annealing not being optimized.

The ideality factor can be evaluated from the logarithmic plot of the forward current I versus bias voltage V, i.e., $n=\frac{q}{kT}\frac{dV}{dln\left(I\right)}$, where k, T and q are the Boltzmann constant, temperature and unit charge. The ideality factors of the Cu₂O/ITO and Cu₂O/Ti/ITO heterojunctions were derived as 1.21 and 1.64, respectively. Figure 5 compares the sheet resistance of the electrodeposited Cu₂O films on the ITO and Ti/ITO substrate. The values of 380 MΩ/sq and 550 MΩ/sq agree with other reported values [42], and they are dependent of electrodepositing conditions such as pH of the solution, deposition potential and temperature, etc. It is noticeable that the forward current of Cu₂O/Ti/ITO in the dark is lower than that of Cu₂O/ITO, indicating the resistance of Cu₂O film on Ti/ITO substrate is higher.

The contact potential ϕ_i of a heterojunction can be extrapolated from the capacitance–voltage (C-V) characteristics. The interface capacitance (per unit area) C_j is obtained from the series connection of the capacitance of each heterojunction layer [43] under the reverse bias voltage V_a:

$$C_j=\frac{1}{\frac{x_n}{{\epsilon }_{s,n}}+\frac{x_p}{{\epsilon }_{s,p}}}=\sqrt{\frac{q{\epsilon }_{s,n}{\epsilon }_{s,p}}{2}·\frac{N_a{N}_d}{(N_a{\epsilon }_{s,p}+N_d{\epsilon }_{s,n})({\mathrm{\varnothing }}_i-V_a)}}$$

(4)

where x_n, x_p, ɛ_s,n, ɛ_s,p, N_a and N_d are the width of the depletion region, dielectric constant, and doping concentration in the n-type and p-type layers. Equation (4) can be reformed by inverse of the capacitance squared $1/C_j^2$ as a function of reverse bias voltage V_a by:

$$\frac{1}{C_j^2}=\frac{2}{q}·(\frac{1}{N_d{\epsilon }_{s,n}}+\frac{1}{N_a{\epsilon }_{s,p}})({\mathrm{\varnothing }}_i-V_a)$$

(5)

in which a linear relationship of $\frac{1}{C_j^2}$ versus V_a can be observed and contact potentials ϕ_i can be extrapolated from the V-intercept when $\frac{1}{C_j^2}=$ 0. Room temperature C-V measurements on Cu₂O/ITO and Cu₂O/Ti/ITO heterojunction solar cells were carried out in the dark at a fixed frequency of 10 kHz and an AC voltage of 20 mV. The test results are shown in Figure 6. It is clear that the $\frac{1}{C_j^2}$-V_a relation for both solar cells is linear, with the extrapolated values of contact potentials ϕ_i to be 0.74 V and 1.01 V for Cu₂O/ITO and Cu₂O/Ti/ITO, respectively. They are consistent with the photovoltaic theory that a higher ϕ_i leads to a higher V_OC as shown in Table 1.

The enhancement of photovoltaic properties by adding the Ti buffer layer was attributed to prevented reactions near the junction, suppressed diffusion of chemical species, and reduced recombination of minority carriers at the heterojunction since charge carrier dynamics and free carrier lifetimes depend on the quality of materials, such as a smooth junction interface, surface defects and bulk impurities, the presence of interfaces and trap states, etc. Such enhancement has also been demonstrated by other solar cells with buffer layers [44]. For instance, the surface roughness of the Cu₂O film on the Ti buffer layer and ITO substrate was decreased to the nanometer-sized Cu₂O crystals in the wavelike surface with minimal overall surface energy as shown by AFM images in Figure 3. This suggests that electrodeposited Cu₂O film grows more uniformly with a smaller grain size on the smoother Ti surface than the ITO surface, and therefore forms a higher-quality heterojunction. Resistivity of the electrodeposited Cu₂O film on the Ti buffer layer was higher than on ITO as shown in Figure 5. However, this additional series resistance was compensated by the improvement of the junction. It needs to be pointed out that the fabrication process of the solar cells reported in this study is simple with low cost and Ti buffer layer enhanced photovoltaic performance. However, the power conversion efficiency η is still lower than the theoretical 20% value [25]. More experimental works are currently carried out for further improvement. J_SC can be increased by reducing the series resistance in ITO and Cu₂O films by techniques such as doping with donors and acceptors. V_OC is influenced by imperfection especially defects at the heterojunction interface, which causes reduction in the shunt resistance, higher ideality factor, and increased reverse saturation current. As a result, further reducing defects and improving heterojunction interface is expected to increase V_OC and therefore η. Other optimizations such as the Cu₂O film and Ti buffer layer thickness, electrodeposition and solar cell fabrication process are also under investigation. To investigate the stability, the fabricated Cu₂O/Ti/ITO and Cu₂O/ITO solar cell devices were characterized after 1 month, 3 months, and 6 months, without obvious degradation being observed.

4. Conclusions

Cu₂O/Ti/ITO and Cu₂O/ITO heterojunction solar cells with an electrodeposited Cu₂O film were fabricated and characterized for comparison. The elemental mapping images by TEM clearly showed the continuous Ti buffer layer between Cu₂O and ITO. The 3D and planar AFM scanning images showed that the surface roughness of the electrodeposited Cu₂O film was reduced by the Ti buffer layer. J-V characteristics showed that although the resistivity of the electrodeposited Cu₂O film was increased by the Ti buffer layer, which leads to a decreased forward current, V_m, V_OC and I_SC were all enhanced and therefore the power conversion efficiency of the Cu₂O/Ti/ITO solar cell. From C-V test results the contact potentials ϕ_i of 0.74 V and 1.01 V for Cu₂O/ITO and Cu₂O/Ti/ITO were extrapolated, respectively, which are consistent with the photovoltaic theory that a higher ϕ_i leads to a higher V_OC. The enhancement of photovoltaic performances by the Ti buffer layer was attributed to smoothing the junction interface, preventing unwanted reactions near the junction, suppressing the diffusion of chemical species, and reducing the recombination of minority carriers at the heterojunction. Compared to other solar cell systems, the efficiency η reported in this research is still lower mainly due to the fabrication process not optimized. However, considering the simple and low-cost fabrication process combined with the feasible method of adding the Ti buffer layer to improve photovoltaic properties, this research shows that Cu₂O/Ti/ITO would be a promising heterojunction solar cell for photovoltaic applications for sustainable energy production.