Solar cells are photovoltaic (PV) devices, which directly convert light energy into electricity. The development of solar cell applications and other renewable energy resources will be an excellent solution for the future energy crisis. Amongst available solar cells, Si-based polycrystalline devices are still dominating the photovoltaic market due to their high efficiencies and long-term stability [1] . Other types of solar cells such as thin-film, organic material, inorganic material, and dye-sensitized solar cells are currently under intense research to achieve higher efficiencies and stability [2] [3] [4] . Many researches have been devoted to improving the thin-film solar cell (second-generation solar cells) performances due to their low-cost fabrication, long-term stability, and excellent power conversion (PCE) efficiencies. Research on thin film is of great interest and with a very active research topic in the photovoltaic community, and the study of the CIGS-based solar cells is a topic under intense research. Recently worldwide research interest in cadmium telluride (CdTe) and chalcopyrite Cu(In, Ga)Se₂ (CIGS) based solar cells has caused considerable interest within the photovoltaic community due to their excellent PCE and cost-effectiveness. Because Cd is toxic to living beings and causes environmental concerns, investigations on CIGS-based solar cells have become more popular over CdTe-based solar cells. CIGS solar cells have shown excellent power conversion efficiencies of over 20% on lab-scale experiments [5] [6] . In addition to low cost and higher efficiencies, CIGS-based photovoltaic devices offer higher radiation resistance, durability, and lightweight, which is essential in space applications than Si-based solar cells. Therefore, CIGS has become a promising semiconductor material in the thin-film solar cell industry. Even though conventional CIGS solar cells show excellent performances, many research groups are investigating making a new device by replacing the CdS buffer layer, which produces fewer environmental concerns such as In₂S₃, ZnS, and ZnSnO₄, etc. Recently we have shown that n-In₂S₃ can be used as an alternative to CdS with a remarkable efficiency of 26%, which offers less health and environmental effects [7] . However, further theoretical and experimental investigations are necessary to explore CIGS-based solar cells’ working principles to achieve higher efficiencies and become more popular in the photovoltaic market. Solar cells also encounter efficiency losses due to material properties and device engineering limitations like all other electronic devices. Carrier recombination is the potential drawback observed in all types of solar cells [8] [9] .

Optimizing the absorber layer’s physical parameters and other engineering developments will help to suppress the carrier recombination. Back-contact recombination becomes more dominant with a thinner absorber layer. Hence, the absorber layer’s thickness and bandgap optimization are crucial parameters for lab-scale experiments and theoretical works to achieve better solar cell performance. In thin-film solar cell studies, p-CIGS absorber material has become more promising, because it possesses a tunable bandgap with a bandgap range of ~1.06 - 1.7 eV. One can find more details about the practical and commercial level fabrication process of CIGS-based solar cells from the available literature [10] . A conventional CIGS-based solar cell consists of Substrate/Mo/p-CIGS/n-CdS/Al-ZnO/metal(Ni) grids. Many lab-scale and theoretical investigations have been conducted and modified in such a way as to reach different innovative ideas to achieve higher efficiencies [11] . Based on available data, most lab and commercial-based fabrications have achieved efficiencies of more than 12% [12] - [17] .

To simulate realistic experimental conditions, this work extensively analyzed impacts of the absorber layer’s bandgap, thickness, and defect density on output solar cell performances viz short circuit current density (J_sc), open-circuit voltage (V_oc), fill factor (FF%), power conversion efficiency (PCE) and quantum efficiency(QE%). Finally, the optimized structure was simulated by varying the operating temperature.

A deep understanding of electronic device fabrication and its working principle is vital to elevating the devices’ performance and making them available at the industrial level. Implementing new device architecture, devices’ engineering developments and numerical simulations provide a robust platform for that. Experimental time-scale and fabrication costs can significantly be reduced by employing numerical simulations before starting the lab-scale experiments. Hence, a large amount of theoretical research based on numerical simulations is carried out worldwide. The following sections performed a numerical simulation using one dimensional Solar Cell Capacitance Simulator (1D-SCAPS) software package (version 3.3.07) developed at the University of Gent, Belgium. Initially, SCAPS software was designed only for CdTe-based thin films. After implementing several modifications by the developers, SCAPS can now simulate Si, GaAs-based thin-film solar cells, and more details can be found in the literature [18] [19] [20] [21] . 1D-SCAPS is a powerful simulation software tool that could solve basic semiconductor equations such as Poisson’s and continuity equations. Also, the SCAPS simulation can produce steady-state band diagrams, which are essential for semiconductor physics. Poisson’s equation mainly describes the relationship between charge density and the electrostatic potential, and continuity equations describe the electrons and holes moving in and out of the material. These equations are given below.

d d x ( − ε ( x ) d ψ d x ) = q [ p ( x ) − n ( x ) + N d + ( x ) − N A − ( x ) + p t ( x ) − n t ( x ) ] (1)

d p n d t = G p − p n − p n 0 τ p − p n μ p d ξ d x − μ p ξ d p n d x + D p d 2 p n d x 2 (2)

d n p d t = G n − n p − n p 0 τ n + n p μ n d ξ d x + μ n ξ d n p d x + D n d 2 n p d x 2 (3)

where ε is the permittivity, D p and D n are carrier diffusion coefficients for holes and electrons respectively, q is the electronic charge, µ is the carrier mobility, the n p (electrons) and p n (holes) are minority carrier concentrations, G is the generation rate, n t and p t are trapped electrons and holes, respectively, N d + and N A − are ionized donors and acceptors, respectively, ψ is the electrostatic potential, and ξ is the electric field.

In the conventional CIGS cell structure, the depletion region (active region) is made with p-CIGS and the n-CdS, and the window layer is made from Al-ZnO with a bandgap of 3.30 eV. The device should have a p-n junction with a healthy depletion region to reach a better output performance. The formation of a weak depletion region will directly affect the power conversion efficiency. In this device, the n-Si layer was employed between CIGS and the buffer layer to obtain a healthy depletion region in the p-n junction and a higher conductivity in the structure. Therefore, this study proposed a new device structure with substrate/Mo/p-CIGS/n-Si/i-In₂S₃/Al-ZnO/metal grids, as depicted in Figure 1. Defect density was implemented in the CIGS, n-Si, and i-In₂S₃ layers to keep the device close to an actual device. In this study, CdS-buffer layer has been replaced by i-In₂S₃ because Cd is toxic to living beings and creates environmental concerns. Before the simulation, material parameters should be entered into software as physical input parameters. Additionally, test conditions such as operating temperature, illumination, bias voltage, etc., were set before starting the simulation. Table 1 shows the optimized and required material parameters of each layer used in this simulation.

Robust and precise simulation software will play a vital role during new device modeling. Herein, numerical simulation was performed for a new device architecture of CIGS-based solar cell using the 1D-SCAPS simulator. The proposed structure contains a p-CIGS absorber layer, n-Si layer, and In₂S₃ as a buffer layer. In this simulation, first, each layer thickness was optimized using a pre-selected model, and 2 µm, 0.5 µm, 0.05 µm, and 0.1 µm thicknesses were selected for CIGS, n-Si, In₂S₃, and Al-ZnO layers, respectively. The CIGS layer’s optimum bandgap value was determined by varying bandgap values from 1.1 eV to 1.7 eV before investigating the absorber layer thickness’s impacts. Secondly, the proposed structure was investigated at various absorber layer thicknesses with the simulated optimum bandgap. In this study’s final stage, the optimized cell was then simulated at multiple operating temperatures from 298 K to 400 K and various defect densities from 1.77 × 10¹³ to 1.77 × 10¹⁷ cm⁻³. The incident power density was maintained at AM 1.5 (Power = 1000 Wm⁻²).

Figure 2 depicts that efficiency increases rapidly from 1.1 eV to 1.4 eV, then remains constant until 1.7eV. Therefore, it is apparent that a thin-film solar cell’s efficiency significantly depends on the selected semiconductor material’s bandgap. This work used 1.4 eV as the CIGS layer’s optimized bandgap throughout the experiment based on the simulation.

After optimizing the CIGS absorber layer’s bandgap, thickness has been numerically investigated in the ranges of 0.2 μm to 2 μm while keeping other material parameters of different layers constant. This study’s main objective was to optimize the absorber layer thickness to achieve maximum output performance with minimal cell’s power conversion efficiency (PCE) losses. A heterojunction solar cell is essentially a p-n junction device. Therefore, it is crucial to confirm the formation of the p-n junction in the thin-film structure.

The energy band diagram (EBD) is the most evident signature of the thin film’s p-n junction building. Figure 3 presents the energy band diagram of the CIGS-based cell structure of this study. The energy band diagram shows different recombination regions named A, B, C & D. Regions A & B are assigned to recombination at back contact and bulk recombination. Thus the energy band diagram confirms that by increasing absorber layer thickness, back contact recombination current density significantly reduces and ensures the enhancement of cell output performances. Also, the formation of region C (depletion layer) in EBD confirms that the proposed structure has a healthy, active region essential for higher performance.

Solar cell output performances, viz V_oc, J_sc, FF, and PCE, are shown in Figure 4, and the simulation results are summarized in Table 2. The results clearly show that J_sc, V_oc, FF, and PCE have increased as the absorber layer’s thickness increases. The maximum efficiency value of 22.57% was achieved at 2 μm. The rapid rise in output parameters, V_oc, J_sc, FF & PCE, can be observed in the thickness ranging from 0.2 μm to 0.8 μm, and 0.8 μm to 2 μm will only increase slightly, which means saturation phase has been reached. Thickness optimization of the absorber and the buffer layer is essential for an efficient photovoltaic

device. Recently we have shown that the In₂S₃ buffer layer is a promising and robust alternative for CdS. Also, buffer layer thickness variation from 20 nm to 100 nm had a minimum effect on the cell’s performance [22] . Thus the extensive investigation of how In₂S₃ buffer layer thickness impacts the output performance has been excluded from this study.

It is evident that CIGS layer thickness is directly proportional to the number of absorbed photons, contributing to generating more electron-hole pairs (e-h) before recombination with the back contact and, consequently, enhancing the cell’s overall efficiency. According to Figure 5, back contact recombination is drastically reduced, while in Figure 6, bulk recombination current density increases as the absorber layer’s thickness increases.

However, obtaining higher CIGS absorber layer thickness is still challenging because of the high fabrication cost during mass production due to the high indium content (In) and gallium (Ga). Hence, it was evident from the results that the optimum absorber layer thickness should be around 2 μm to minimize the material consumption without any significant impact on solar cell efficiency. Also, cells with thinner absorber layers will negatively affect the cell’s performance by increasing back-contact recombination. Because the back contact is located near the depletion region of the p-n junction, this will cause higher recombination with the back contact and result in a downward trend in output performance. Therefore, more experimental research needs to be performed to determine the CIGS absorber layer’s optimum thickness to obtain higher efficiencies. Output performances of photovoltaic devices are highly dependent on the spectrum of the incident light. Hence, the proposed device was studied within the spectral range of 300 nm (4.13 eV) - 1200 nm (1.03 eV). The spectral response of the device is measured in terms of quantum efficiency (QE%). QE (λ) of a photoactive material is the number of carriers measured at the output detector per number of incident photons per unit time. It estimates the device’s ability to convert a photon flux to a photogenerated electric current and can be expressed by Equation (4).

QE ( λ ) = J p h q ϕ (4)

where J p h is the photogenerated current density, q is the electronic charge, and ϕ is the number of incident photons.

The thickness of the absorber layer affected the QE% of the cell, and Figure 6 shows the variation of QE% under AM 1.5 as a function of CIGS absorber layer thickness. As expected, by increasing CIGS layer thickness, many photons are absorbed towards the higher thickness range, resulting in more photogenerated carriers collected before recombination and directly enhancing the QE%. There are three distinct regions in Figure 6; above 600 nm, 440 nm to 520 nm, and below 400 nm. Above 600 nm, one can observe that QE% has progressively increased from 0.2 - 2 μm due to the spectrum’s high absorption by thicker layers, thus lowering the recombination, ensuring the maximum QE%. Nevertheless, below 400 nm, one can observe the blue response mainly due to front surface recombination. The change in QE% from 440 nm to 520 nm is apparent in Figure 7 due to the buffer layer’s influence.

However, the defect density of polycrystalline material significantly affected the performance of the cell. It is well known that defects in the absorber layer influence the device output solar cell parameters. Hence, to get a deep understanding of defect density in the absorber layer, PCE and FF parameters were simulated. Meanwhile, the defect density was changed from 1.77 × 10¹³ to 1.77 × 10¹⁷ cm⁻³. The effects of defect density on PCE & FF% are shown in Figure 8. As the defect density increases, PCE decreases from 22.17% to 10.57%, and FF% decreases from 84.30 to 69.35 which means defects at the absorber layer in the cell are unfavorable for the efficiency fill factor. This observation can be attributed to higher defect density leading to a higher recombination rate, which reduces the cell’s efficiency.

In summary, this novel study shows that n-Si is an excellent candidate to build a pn junction with a healthy depletion layer. Furthermore, it was found out that the CIGS layer’s optimum bandgap energy was 1.4 eV and the optimized thickness was 2 µm. Thickness optimization of the absorber layer is vital to reduce material consumption and consequently relieve the fabrication cost. Additionally, at higher operating temperatures (above 310 K) and higher defect densities (above 1.77 × 10¹³) in the absorber layer, the optimized cell shows a downward trend in power conversion efficiency; however, further lab-scale experiments are needed to implement the outcome of this simulation work.

The author thanks the University of Gent in Belgium for providing the SCAPS 1D simulation software.

The authors declare no conflicts of interest regarding the publication of this paper.

Mohottige, R.N., Farndale, M., Coombs, G.S. and Saburhhojayeva, S. (2024) Numerical Analysis on the Effect of n-Si on Cu(In, Ga)Se₂ Based Thin-Films for High-Performance Solar Cells by 1D-SCAPS. Open Journal of Applied Sciences, 14, 1315-1329. https://doi.org/10.4236/ojapps.2024.145086