Thin-Film Carbon Nitride (C₂N)-Based Solar Cell Optimization Considering Zn_1−xMg_xO as a Buffer Layer

Abstract

Carbon nitride (C₂N), a two-dimensional material, is rapidly gaining popularity in the photovoltaic (PV) research community owing to its excellent properties, such as high thermal and chemical stability, non-toxic composition, and low fabrication cost over other thin-film solar cells. This study uses a detailed numerical investigation to explore the influence of C₂N-based solar cells with zinc magnesium oxide (Zn_1−xMg_xO) as a buffer layer. The SCAPS-1D simulator is utilized to examine the performance of four Mg-doped buffer layers (x = 0.0625, 0.125, 0.1875, and 0.25) coupled with the C₂N-based absorber layer. The influence of the absorber and buffer layers’ band alignment, quantum efficiency, thickness, doping density, defect density, and operating temperature are analyzed to improve the cell performance. Based on the simulations, increasing the buffer layer Mg concentration above x = 0.1875 reduces the device performance. Furthermore, it is found that increasing the absorber layer thickness is desirable for good device efficiency, whereas a doping density above 10¹⁵ cm⁻³ can degrade the cell performance. After optimization of the buffer layer thickness and doping density at 40 nm and 10¹⁸ cm⁻³, the cell displayed its maximum performance. Among the four structures, C₂N/Zn_0.8125Mg_0.1875O demonstrated the highest PCE of 19.01% with a significant improvement in open circuit voltage (V_oc), short circuit density (J_sc), and fill factor (FF). The recorded results are in good agreement with the standard theoretical studies.

1. Introduction

The excessive use of fossil fuels is a major contributor to the global warming phenomenon, making renewable energy a better alternative to meet our energy needs. Solar energy is one of the most abundant, environmentally friendly and sustainable non-conventional sources [1]. The incident sunlight can be converted directly into useful electrical energy using photovoltaic (PV) technology. The price of PV modules have declined by 80%, and their installation capacity has increased rapidly over the past decade, from 40 GW to over 600 GW [2]. This breakthrough in the PV industry has occurred due to the advancement in fabrication processes and materials to be implemented in solar cell structures. Currently, silicon is the most common material used in the fabrication of PV modules. Although silicon is a widely available element in nature in the form of silicon dioxide (SiO₂), processing it requires intensive energy [3]. Alternative solar cell materials, such as CdTe [4], GaAs [5], CIGS [6], perovskite, and organic materials [7] have also demonstrated high efficiencies, but they still have some disadvantages, including a high manufacturing cost, toxic nature, low thermal and chemical stability, and a shorter outdoor life [8,9]. Developing novel materials has always been a driving factor for economic, social, and cultural progress. The current focus in the research on sustainable energy is on preparing materials that are cyclic and omnipresent.

In recent years, the use of two-dimensional (2D) materials in electronics, optoelectronics, batteries [10,11,12], and photovoltaic cells [13,14] has become increasingly popular. Among 2D materials, graphene has grabbed the attention of the worldwide research community due to its promising properties [15]. In the modern age of nanotechnology, the vast potential of graphene has opened new ways to develop other 2D materials for future technologies. For the first time in 2015, an academic research team in South Korea synthesized nitrogenated holey-two graphene (C₂N-h2D) by using a simple wet-chemical reaction [16]. The C₂N-h2D crystal was prepared by introducing nitrogen (N) atoms into a dense network of covalently bonded carbon atoms. The crystal network of carbon atoms can easily accommodate N-atoms due to their atomic size and five-electron valence structure (sp² hybridization), as shown in Figure 1a,b. The addition of nitrogen to the carbon lattice results in a significant improvement in the properties of carbon materials, such as thermal conductivity, band gap, and chemical conductivity. Additionally, the phonon modes of C₂N indicate high structural stability, which can be useful for generating high energy for photovoltaic cells [17]. To the best of our knowledge, C₂N as a layer has not yet featured in the fabrication of standard PV technologies; however, theoretical studies are limited to highlighting their basic physical properties. Therefore, it is necessary to simulate solar cells before fabricating them in the laboratory. Zhou et al. have performed a numerical study using five window layer materials coupled with the C₂N absorber layer, among which the CdS/C₂N heterojunction structure displayed the best efficiency of above 17% [18]. Tsoeu et al. found a 2D-based GaS/C₂N heterostructure to be a suitable alternative to lead-containing materials and reported their proposed device efficiency as 17.8% [12]. Yasin et al. performed a numerical study on C₂N-based solar cells using four different buffer layers and attained a best efficiency of 18.57% for an IGZO-based device [19].

In thin-film solar cells (TFSCs), the buffer layer contributes to form a p-n junction with the absorber layer while allowing maximum light penetration into the junction and the absorber layer. It improves the band alignment and assists in the flow of photogenerated charge carriers from the absorber layer to the back electrode [20]. Moreover, it plays an essential role in minimizing the absorption of short wavelengths and the recombination at the interface between the buffer and the absorber layers [21]. In recent studies on C₂N-based solar cells, CdS and IGZO have been utilized as electron transport layers (ETLs) [18,19]. The presence of cadmium in CdS makes it undesirable to be used in solar cells, while IGZO leads to higher manufacturing costs because it contains rare earth metals such as indium and gallium. Therefore, selecting an appropriate buffer layer is essential for efficient and cost-effective C₂N-based solar cells. In terms of cost and efficiency, Zn-based compounds are considered the most promising alternatives. According to a research group based at Colorado State University, losses due to short wavelengths can be largely eliminated via the ZMO window layer [22]. Ke et al. have reported that the concentration of Mg in Zn_1−xMg_xO directly affects the bandgap, conduction band alignment, and Fermi-level position of the alloy’s structure [23]. Bahfir et al. developed a CZTS-based solar cell by replacing the CdS and ZnO ETLs with ZnMgO as a single-layer buffer material and achieved a better PCE of 11.50% [24]. Ren et al. constructed a CdTe-based thin solar cell by depositing a bilayer of zinc magnesium oxide (ZnMgO) on tin oxide (SiO₂) and reported an improved device efficiency of 16.76% [25]. Ren et al., in their separate study, used RTA-treated ZnMgO for CdTe solar cells and obtained a PCE of 15.7% [26]. Shahram et al. proposed a kesterite solar cell structure with a 50 nm-thick layer of Zn_0.81Mg_0.19O with an achieved efficiency of 17.05% [27].

In this work, we aimed to investigate C₂N-based solar cells structured with four Mg-doped zinc magnesium oxide (Zn_1−xMg_xO) buffer layers in the SCAPS-1D simulation software. The doping concentration of Mg varied from x = 0.0625 to x = 0.25 to achieve the best band alignment. The absorber and buffer layers’ thickness, doping concentration, and defect concentration are investigated with a deep understanding of the device’s internal architecture. This study also focuses on the impact of interface defects and temperature on solar cell performance.

2. Numerical Modeling and Simulation Parameters

The basic architecture of the proposed C₂N-based solar cell structure (Al/TCO/Zn_1-xMg_xO/C₂N)/Ni) is shown in Figure 2. Simulations were conducted via the one-dimensional solar cell capacitance simulator SCAPS-1D (version 3.3.10). The SCAPS-1D was developed by Marc Burgelman and his colleague at the University of Gent, Belgium [28]. This software solves three basic equations for semiconductor devices, namely the Poisson equation and continuity equations for electrons and holes, as shown in Equations (1)–(3) [29,30], respectively.

$$\mathrm{Poisson}\mathrm{equation}:\frac{\partial }{{\partial }_x}\left({\epsilon }_o{\epsilon }_r\frac{\partial \Psi }{{\partial }_x}\right)=-q\left(p-n+N_D^{+}+N_A^{-}+\frac{\rho }{q}\right)$$

(1)

$$\mathrm{Continuity}\mathrm{equation}\mathrm{for}\mathrm{electrons}:-\left(\frac{1}{q}\right)\frac{\partial J_n}{{\partial }_x}-u_n+G=\frac{{\partial }_n}{{\partial }_t}$$

(2)

$$\mathrm{Continuity}\mathrm{equation}\mathrm{for}\mathrm{holes}:-\left(\frac{1}{q}\right)\frac{\partial J_p}{{\partial }_x}-u_p+G=\frac{{\partial }_p}{{\partial }_t}$$

(3)

where Ψ is the wave function for electrostatic potential, and ε_o and ε_r are the permittivity of free space and the relative permittivity, respectively. The density of the defect charge is denoted by ρ, symbols n and p represent the free carrier concentrations for electrons and holes, the generation rate is denoted by G, and the ionized donor and acceptor densities are labelled as $N_D^{+}$ and $N_A^{-}$, respectively. In addition, J_n and J_p represent the current densities of the electron and hole, respectively. The SCAPS-1D is capable of performing in a variety of conditions, such as light and dark, different device operating temperatures and rear-contact voltages, and integrating various interfaces states. The program solves the above semiconductor equations by using input parameters that are appropriate for the working conditions and generating output parameters that represent the solar cell performance. This research focus on the investigation of key performance parameters represented by open-circuit voltage (V_oc), short-circuit current density (J_sc), fill factor (FF), and power conversion efficiency (PCE) to evaluate the performance of the solar cell.

The solar cell input parameters depend on the accuracy of the results generated by the simulator. Therefore, the material parameters for each layer are selected carefully from authentic literature, as summarized in

Table 1. Initial key input parameters for the SCAPS-1D simulation.

Parameters	TCO [19,32]		Zn1−xMgxO [31,32,33]			C2N [18,19]
		x = 0.0625	x = 0.125	x = 0.1875	x = 0.25
Thickness (nm)	30	30	30	30	30	300
Band gap (eV)	3.5	3.41	3.48	3.65	3.82	1.8
Electron affinity (eV)	4	4.44	4.37	4.21	4.03	4.42
Dielectric Permittivity	9	9	9	9	9	4.5
CB-effective density of states (cm−3)	2.2 × 1018	1 × 1017	9 × 1016	1 × 1016	1 × 1015	1 × 1019
VB-effective density of states (cm−3)	1.8 × 1019	1 × 1018	9 × 1017	1 × 1017	1 × 1016	1 × 1015
Electron thermal velocity (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Hole thermal velocity (cm/s)	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107	1 × 107
Electron mobility (cm2/(V·s))	20	50	50	50	50	13
Hole mobility (cm2/(V·s))	10	20	20	20	20	20.6
Donor doping concentration (cm−3)	1 × 1014	1 × 1014	1 × 1014	1 × 1014	1 × 1014	0
Acceptor doping concentration (cm−3)	0	0	0	0	0	1 × 1014
Defect density Nt (cm−3)	1 × 1014	1 × 1014	1 × 1014	1 × 1014	1 × 1014	1 × 1014
Defect type	Neutral	Neutral	Neutral	Neutral	Neutral	Neutral
Defect distribution	Single	Single	Single	Single	Single	Gaussian

[18,19,31,32,33]. The simulation parameters for interfacial defects are presented in

Table 2. Interface layer settings for the SCAPS simulation.

Parameters	TCO/ZMO	ZMO/C2N
Defect type	Neutral	Neutral
Capture cross section of electrons (cm2)	1 × 10−19	1 × 10−19
Capture cross section of holes (cm2)	1 × 10−19	1 × 10−19
Energetic distribution	Single	Single
Energy level with respect to Ev (eV)	0.6	0.6
Major carrier barrier height relative to Ef (eV)	SCAPS	SCAPS
Major carrier barrier height relative to Ev (eV)	SCAPS	SCAPS

, while the metal contact electrodes and operating temperatures are listed in

Table 3. Input parameters used for front and back contacts in the simulation.

Parameters	Back-Contact Metal Electrode	Front-Contact Metal Electrode
Surface recombination velocity of electrons (cms−1)	105	107
Surface recombination velocity of holes (cms−1)	107	105
Work function (eV)	5.35 Ni (111) [34]	4.06 (AI) [34]
Working temperature (K)	300

. In the first stage of optimization, at a defect density of 10¹⁴ cm⁻³, the C₂N layer thickness was changed from 50 nm to 1000 nm while keeping the acceptor density, thickness, and donor density of C₂N and ZMO constant at 10¹⁵ cm⁻³, 30 nm, and 10¹⁵ cm⁻³, respectively. At the optimum thickness of C₂N, the acceptor concentration was changed from 10¹³ cm⁻³ to 10¹⁸ cm⁻³. In the second stage, the buffer layers were optimized following the same procedure. Finally, the optimized cells are examined at different levels of bulk and interface defects, and their stability is studied by changing the operating temperature.

3. Results and Discussion

3.1. Energy Band Alignment

The band alignment of semiconductor materials is crucial in determining solar cells’ performance. When two dissimilar semiconductor materials combine in a heterojunction structure, a discontinuity in the conduction band (CB) forms. These misalignments are represented as the conduction band offset (CBO). The values of the CBO at the interface of the absorber layer and the buffer layer are listed in

Table 4. CBO of C₂N/ZMO heterostructures.

Buffer Layer	CBO (eV)
C2N/Zn1−xMgxO (x = 0.0625)	−0.02
C2N/Zn1−xMgxO (x = 0.125)	0.05
C2N/Zn1−xMgxO (x = 0.1875)	0.21
C2N/Zn1−xMgxO (x = 0.25)	0.39

, and are calculated using the following equation:

$$\mathrm{CBO}=\left({\chi }_{{\mathrm{C}}_2\mathrm{N}}-{\chi }_{\mathrm{ZMO}}\right)$$

(4)

Generally, the band offset is of two types: (i) cliff and (ii) spike. Cliff (negative CBO) will be generated if the CB level of the C₂N layer is above the CB of the ZMO; otherwise, spike (positive CBO) will be formed, as illustrated in Figure 3a–d. The cliff offset reduces the built-in potential of the cell, while the spike offset increases it. In a C₂N/ZMO heterostructure, as the Mg concentration increases from x = 0.0625 to x = 0.1875, the band gap of Zn_1−xMg_xO increases, which improves the energy band alignment and enhances the solar cell performance. Whereas at x = 0.25, the increase in band gap causes a large CBO, which reduces the efficiency. The interface between the C₂N and ZMO (0.0625) forms a small cliff (negative CBO) of −0.02 eV. Although cliff does not hinder the flow of charge carriers from the absorber to the buffer layer, it decreases the built-in potential, which subsequently reduces the charge carriers’ separation at the interface. A minute spike (CBO of 0.05 eV) at the C₂N/ZMO (x = 0.125) interface causes an increase in the built-in potential, which increases the flow of electrons and improves the cell’s performance as compared to the first structure. A moderate spike (CBO of 0.21 eV) at the C₂N/ZMO (x = 0.1875) further enhances the built-in potential at the heterojunction, which hinders the photogenerated charge carriers from going back to the junction interface and suppresses charge recombination, hence improving the device performance. Moreover, a large spike at the interface between C₂N and ZMO (x = 0.25) creates a large built-in potential; it also significantly hinders the electrons from flowing through the junction. Thus additional drift fields are needed to ensure the flow of electrons through the junction, consequently degrading the solar cell performance.

Furthermore, the absorption behavior of the investigated buffer layers indicates that the QE improves as the Mg doping of the ZMO layer increases from x = 0.0625 to x = 0.1875, as shown in Figure 4. Below 700 nm, the QE response to the incident light spectrum displays a noticeable variation for various buffer layers. On the other hand, when the Mg concentration increases to x = 0.25 a CBO of 0.39 eV is formed which produces a large spike at the heterojunction of the two layers. This large spike causes huge hurdles in the movement of electrons from the absorber to the buffer layer, which in turn increases recombination. The increase in the recombination rate reduces the amount of photogenerated charge carriers collected at the back electrode of the solar cell, which drastically affects the QE and reduces it.

Therefore, the energy band alignment of C₂N/ZMO structures demonstrates that a spike (positive CBO) interface helps in improving cell performance. However, a large spike (overly positive CBO) creates a barrier for the photogenerated charge carriers, lowering the short circuit current (J_sc) Figure 5 and efficiency.

3.2. Thickness Optimization of the C₂N Layer

The absorber layer thickness significantly influences the overall device efficiency [9,18,35]. In solar cells, an absorber layer with a thickness of a few hundred nanometers is sufficient for efficient photocurrent generation. We performed our simulations by varying the absorber layer thickness from 50 nm to 1000 nm. For a C₂N/ZMO (x = 0.1875) structure, the essential PV parameters, V_oc, J_sc, and PCE, improved, as shown in Figure 6a,b,d. Above a thickness of 500 nm, the photogenerated charge carriers increased; however, as we increased the thickness to a maximum of 1000 nm, a partial recombination of the charge carriers occurred in the C₂N layer, causing saturation in the performance. On the other hand, the fill factor (FF) demonstrates unusual variation when increasing the thickness of the C₂N layer, as shown in Figure 6c. It starts with a downtrend but gradually increases from a 700 nm thickness. Similar behavior was observed in the previously reported literature on C₂N and perovskite [9,19]. It can be ascribed to the charge carriers’ diffusion length, which changes the series resistance and decreases the internal power. From 50 nm to 700 nm, the absorber layer thickness is larger than the diffusion length, hence recombination takes place, which causes the series resistance to increase and the internal power to deplete, leading to a drop in FF. Above 700 nm, the absorption increases as the diffusion length exceeds the absorber layer thickness, and the FF rises again.

In addition, the positive impact of the absorber layer on solar cell performance in terms of quantum efficiency (QE) and short circuit current (J_sc) can be seen in Figure 6e,f, where increasing C₂N thickness improves the absorption process and increases electron-hole pairs generation. Thus, the optimum device performance was obtained at a thickness of 1000 nm with the following parameters: PCE of 18.61 %, V_oc of 1.23 V, J_sc of 18.25 mA/cm², and FF of 83.16 %.

3.3. Impact of C₂N Defect Density

As a member of the graphene family, C₂N may contain defect points, such as vacancies, interstitial atoms, and impurities [36]. Some deep defect points in C₂N could contribute to optical absorption and reduce the band gaps, while others negatively impact the C₂N band structure [37]. The presence of deep defect points in the absorber layer may trap the photogenerated charge carriers and block them from reaching the back contact, thereby reducing the device’s performance. Experimental studies have found graphene-like materials to have defect densities varying from 10¹⁰ cm⁻³ to 10¹⁴ cm⁻³ [38,39]. Moreover, several articles in the literature reported on the impact of defect density in SCAPS-simulation ranges from 10¹³ cm⁻³ to 10²⁰ cm⁻³ [40,41]. In this study, the C₂N layer is examined at three defect levels: low, average, and deep, from 10¹³ cm⁻³ to 10²⁰ cm⁻³. According to the simulated results, the solar cell parameters sharply decrease by increasing the defect density above 10¹⁵ cm⁻³, as illustrated in Figure 7a–d. The sharp decrease in V_oc and J_sc with increasing defect density is ascribed to high series resistance and poor fill factor [18].

Furthermore, the bandgap tuning is caused by the increase in defect density which deteriorates the light absorption process [19], thereby adversely impacting the quantum efficiency (QE) and the current-voltage curve (J-V) of the cell, as can be seen from Figure 7e,f. Although simulating solar cell devices at an extremely low defect density exhibits high performance, it is challenging to develop such devices experimentally [42]. Therefore, we performed our simulations at a defect density of 10¹⁴ cm⁻³ with a recorded PCE of 18.61%.

3.4. Influence of C₂N-Doping Density

Numerous studies have indicated that doping in the absorber layer improves the overall performance of the solar cell [18,19,43,44]. However, doping above a certain level could enable the structure to develop a depletion region forced by an internal voltage and deteriorate the cell’s performance [43]. Such a depletion region potentially traps free charge carriers and changes the electronic behavior of the material, consequently degrading the solar cell efficiency. We varied the acceptor doping level of C₂N from 1 × 10¹³ cm⁻³ to 1 × 10¹⁸ cm⁻³. The variation in the solar cell performance parameters with an increase in the doping level is shown in Figure 8. The V_oc increases exponentially as the doping value increases. Above 10¹⁵ cm⁻³, J_sc declines drastically due to the high recombination rate and the increase in the series resistance of the cell, resulting in poor performance. According to our simulations, the optimum acceptor density level for C₂N is selected at 10¹⁵ cm⁻³, with an achieved PCE of 18.61%.

3.5. Thickness Optimization of the Buffer Layer

In a C₂N/ZMO structure, incident light passes through the buffer layer into the absorber layer. Therefore, optimizing the buffer layer thickness is an essential step for improving the photogeneration process in the cell. In this study, the performance of C₂N-based solar cells is investigated by varying the buffer layer thickness in the range of 10 nm to 100 nm while keeping the thickness of the absorber layer constant at 1000 nm. As shown in Figure 9, when the thickness of the ZMO (x = 0.1875) layer increases, the V_oc, J_sc, FF, and PCE decline after reaching the maximum point. This behavior is due to the high absorption of photons in the buffer layer region. Therefore, fewer photons reach the junction, which limits the generation of electron-hole pairs. Generally, a very thin buffer layer allows a high level of photon transmittance. However, two factors primarily prevent reducing the buffer layer thickness, i.e., the Schottky barrier structure, due to which the device performance could reverse, and the fabrication limitation [19]. Therefore, the optimum device performance is obtained at a thickness of 40 nm with a PCE of 18.67 %.

3.6. Optimization of the Buffer Layer Doping Density

We investigated the cell performance by varying the doping concentration of ZMO from 10¹³ cm⁻³ to 10²⁰ cm⁻³ while keeping the thickness of the absorber and the buffer layers constant at 1000 nm and 40 nm, respectively. The variation in the device performance parameters of the ZMO (x = 0.1875)-based structure is displayed in Figure 10, where the V_oc sharply declines after 10¹⁶ cm⁻³. When the doping concentration increased from 10¹⁵ cm⁻³ to 10¹⁸ cm⁻³, the J_sc, FF, and PCE improved. According to the Moss–Burstein effect, semiconductor bandgaps can slightly increase with high doping, providing sufficient charge carriers and increasing conductivity [44]. The optimum recorded PCE of 19.01% is found at a donor concentration of 10¹⁸ cm⁻³.

The simulation results for C₂N-based solar cells revealed that the performance parameters improved with a buffer layer Mg concentration from x = 0.0625 to x = 0.1875, as summarized in

Table 5. C₂N/ZMO-based solar cell parameters after optimization.

Simulated Parameters	Zn1−xMgxO
	x = 0.0652	x = 0.125	x = 0.1875	x = 0.25
Voc (V)	1.23	1.23	1.23	1.21
Jsc (mA/cm2)	18.24	18.25	18.26	18.23
FF (%)	84.89	84.94	84.96	79.25
PCE (%)	18.98	19	19.01	17.71

. Among the investigated structures, the C₂N/ZMO (x = 0.25) acquires the lowest PCE of 17.71%, while the C₂N/ZMO (x = 0.1875) yields the highest PCE of 19.01%. This value is higher than previously reported CdS and IGZO-containing C₂N-based solar cells [18,19].

3.7. Effect of the Interface Defect of C₂N/ZMO

In solar cells, interface defects create recombination centers and traps, which prevent charge carriers from reaching the back contact. In this paper, the impact of interface defects on C₂N-based solar cells is examined by changing their value from 10¹³ cm⁻² to 10²⁰ cm⁻². As illustrated in Figure 11, the efficiency of an x = 0.1875-based structure decreased from 19.01% to 15.84% at 10¹³ cm⁻² to 10²⁰ cm⁻², respectively. This drop in PCE is attributed to the Shockley––Read–Hall (SRH) recombination process, as shown in Equation (5) [45,46].

$${\Re }^{SRH}=\frac{\vartheta {\sigma }_n{\sigma }_p{N}_T\left[np-n_i^2\right]}{{\sigma }_p\left[p-p_1\right]+{\sigma }_n\left[n+n_1\right]}$$

(5)

where ϑ indicates the thermal velocity of an electron, and σ_n and σ_p represent the electrons and holes capture cross-sections, respectively. N_T symbolizes the number of defects per unit volume, n_i is the intrinsic number density, $n$ and $p$ denote the concentration of electrons and holes at equilibrium, and n₁ and p₁ are the concentrations of electrons and holes in the trap defect and valence band, respectively. According to Equation (5), it appears that increasing defects can cause trapping centers at the interface for charge carriers to recombine, resulting in the degradation of solar cell performance.

3.8. Impact of Temperature Variation on the Device Performance

Solar cells are temperature-sensitive devices, thereby high operating temperatures significantly lower their overall performance [30,47,48,49]. In TFSc, the rise in operating temperature leads to an increase the recombination rate, which heats up the cell and deteriorates its overall performance [50]. We examined the stability of the C₂N/ZMO (0.1875)-based cell by implementing an operating temperature in the range of 300 k to 500 k. The PV parameters, such as V_oc, J_sc, FF, and PEC sharply decreased with the rise in temperature, as shown in Figure 12a–d. At the given temperatures, the efficiency decreased from 19.01% to 10.17% at 300 k and 500 k, respectively. All other structures (not shown) exhibited similar behavior. The sharp drop in performance parameters is attributed to the high operating temperatures, which deformed the cell’s layers, shortened the diffusion length, and caused a weak interlayer connection. As a result, increasing the recombination rate and series resistance hinders the flow of photogenerated charge carriers from reaching the back contact and degrades solar cell efficiency.

4. Conclusions

This study presents the detailed numerical simulations of C₂N-based solar cells with the (AI/TCO/Zn_1−xMg_xO/C₂N)/Ni) device structure using the SCAPS-1D software. The influence of four Mg content levels on the performance of the Zn_1−xMg_xO/C₂N heterojunction has been investigated. The simulations revealed that increasing the Mg concentration beyond x = 0.1875 can deteriorate the solar cell performance. Among the investigated buffer layers, x = 0.1875 achieved the highest efficiency, whereas x = 0.25 performed the least well. The thickness of C₂N was found to have a significant influence on the PV parameters, with an optimum selected thickness of 1000 nm. When the acceptor density of C₂N was increased to 1 × 10¹⁵ cm⁻³, a stable performance was observed in the device. The same stability was found for the buffer layers when the thickness and doping density were set at 40 nm and 1 × 10¹⁸ cm⁻³, respectively. Additionally, the high defect density in the absorber layer and at the C₂N/ZMO interface accelerated the recombination rate, which caused the degradation of solar efficiency. Furthermore, the temperature analysis showed that the solar cell is susceptible to high operating temperatures. The findings of this contribution highlight the potential for C₂N-based heterojunction solar cells as an emerging thin-film PV technology.