Enhanced Conversion Efficiency in MAPbI₃ Perovskite Solar Cells through Parameters Optimization via SCAPS-1D Simulation

Abstract

In this study, various factors affecting the efficiency of the MAPbI₃ perovskite solar cell (PSC) were analyzed using the SCAPS-1D simulation program. The basic device analyzed in this study had a structure of ITO/TiO₂/MAPbI₃/Cu₂O/Au. The thickness of each layer (electron transport layer (ETL), perovskite absorption layer (PAL), and hole transport layer (HTL)), PAL defect density and interface defect density were investigated as parameters. The optimized parameters that yielded the highest light conversion efficiency were an ETL (TiO₂) thickness of 100 nm, a PAL (MAPbI₃) thickness of 1300 nm, an HTL (Cu₂O) thickness of 400 nm, a PAL defect density of 10¹⁴ cm⁻³, and an interface defect density of 10¹³ cm⁻³ for both absorber/ETL and absorber/HTL interfaces. The optimized PSC exhibited a maximum efficiency of 19.30%. These results obtained in this study are expected to contribute considerably to the optimization and efficiency improvement of perovskite solar cells using inorganic charge-carrier transport layers.

1. Introduction

Organic–inorganic perovskites have recently attracted considerable attention as absorbers in solar cells; they have a chemical formula of ABX3 and have a three-dimensional structure with organic cations, such as CH₃NH₃, CH(NH2)₂, Cs, and Rb at A, cations such as Pb, Sn, and Ge at B, and halogen anions such as I, Br, Cl, and F at X [1,2]. Perovskites exhibit different performances and properties in solar cells depending on their constituent elements [3]. Since the first study on CH₃NH₃PbX₃ perovskites in 2009, their efficiency has considerably improved to 25.7% or more, resulting in the maximum power conversion efficiency (PCE) [4,5,6,7,8]. Representative perovskites include CH₃NHPbI₃ (MAPbI₃) and CH(NH₂)₂PbI₃ (FAPbI₃). The advantages of perovskites include the possibility of a low-cost production through a low-temperature solution process and a high efficiency because of a high light-absorption coefficient [9]. However, some issues need to be solved; in particular, Pb is harmful to the environment; thus, Pb-free materials should be developed [10].

In addition, existing structures using organic materials for the common layers, Electron Transport Layer (ETL) and Hole Transport Layer (HTL), are sensitive to moisture [11,12]. Therefore, in order to overcome these disadvantages, research on using inorganic materials instead of organic materials for common layers such as ETL and HTL is being actively conducted [13].

Among the widely studied inorganic Electron Transport Layers (ETL) and Hole Transport Layers (HTL) in the device structure of perovskite solar cells (PSCs), looking first at ETL materials, substances such as SnO₂, ZnO, and TiO₂, which exhibit n-type semiconductor characteristics, are commonly utilized [14,15].

SnO₂ is one of the most widely researched transparent conducting oxides (TCOs) as a transparent electrode since the early days of TCOs research, due to its transparent properties and high conductivity [16]. Furthermore, by controlling oxygen deficiency, its electrical characteristics can be adjusted from conducting to semiconductor, enabling extensive application research on thin film transistors as channel layers [17]. In particular, significant research has been conducted to control the electrical properties and improve reliability by doping SnO₂ with other materials, such as Al, In, Ta, etc. [18,19,20].

Additionally, ZnO is one of the most abundant materials on earth and is currently being evaluated in various multi-layer device structures as an Electron Transport Layer (ETL) material in perovskite solar cells. ZnO has a wide bandgap energy (Eg = 3.26 eV) and excellent mobility [21]. With its wide bandgap and significant exciton binding energy, ZnO is considered one of the most interesting ETL materials [22]. In particular, there has been active research in doping with various other materials to create p-type ZnO [23,24]. However, a reliable p-type ZnO has not yet been reported [25]. Nonetheless, in the case of IGZO (a mixture of In₂O₃, Ga₂O₃, and ZnO), although it exhibits n-type characteristics, it is commercially utilized as the channel layer of the thin film transistor (TFT) in large organic light-emitting devices, due to its reliability under operating voltage and current conditions [26,27].

TiO₂ is the most researched material for the electron transport layer in the multi-layer structure of perovskite solar cells. This preference is due to TiO₂’s high chemical and optical stability, non-toxicity, and low cost, making it well-suited for applications in the optical field [28,29]. Among TiO₂ structures, anatase is preferred in the photovoltaic field due to its higher conduction band edge energy and lower electron–hole recombination rate [30]. These various excellent properties led to the selection of TiO₂ as the ETL material in this study.

In addition, p-type inorganic semiconductor materials that can serve as a hole transport layer (HTL) include MoO₃, CuI, NiO_x and Cu₂O [31,32,33]. MoO₃ has a wide bandgap, typically around 3.0 eV, which contributes to its transparency in the visible part of the spectrum [34]. This makes it useful in applications requiring transparent conducting layers, as well as in optoelectronic devices where light absorption or transmission is critical. Also, MoO₃ has a high work function, which makes it an excellent hole-injection layer in organic electronic devices [35]. This property facilitates the efficient transfer of holes from the anode to the organic layer in organic-emitting diodes (OLEDs) and organic photovoltaics (OPVs). Copper Iodide (CuI) is another p-type inorganic semiconductor material known for its distinctive properties that make it suitable for a range of electronic and optoelectronic applications [36]. CuI has a direct bandgap that is typically reported to be around 3.1 eV, making it transparent to visible light [37]. One of the significant advantages of CuI is its relatively high hole mobility compared to other p-type semiconductors [38]. CuI is chemically stable under a wide range of conditions and can be easily doped with various elements to fine-tune its electrical and optical properties [39].

P-type NiO_x demonstrates a significant bandgap (greater than 3.5 eV) coupled with exceptional transparency within the visible spectrum [40]. This characteristic effectively reduces losses, including charge recombination, improves charge transport, and ensures ideal energy level alignment with diverse photoactive absorbers, attributed to its adequate conductivity and chemical stability [41,42,43]. These properties have led to extensive research on using these materials as hole injection layers or hole transfer layers in Quantum-Organic Light Emitting Diodes (QLED) device structures [44,45].

In this paper, Cu₂O was selected as the hole transport layer (HTL) material. Because Cu₂O has a low bandgap energy of 1.8~2.5 eV, the efficient conversion of light into electrical energy is possible, and it has a high hole mobility, enabling the photovoltaic solar cell efficiency to be improved [46,47]. Furthermore, Cu₂O is a stable material and does not undergo significant changes in its electronic properties, even under harsh conditions, and it is a non-toxic and environmentally friendly material, making it a desirable choice for applications in the fields of energy and environment [48,49]. Therefore, in this study, TiO₂ and Cu₂O were selected as the ETL and HTL materials, respectively, based on these excellent characteristics and the band alignment of the energy band diagram. In summary, this study optimized the power conversion efficiency (PCE) of perovskite solar cells using TiO₂ as an Electron Transport Layer (ETL), MAPbI₃ as an absorber layer, and Cu₂O as a Hole Transport Layer (HTL), respectively.

We analyzed the effects of the following parameters on the power conversion efficiency (PCE) of perovskite solar cells (PSCs) through parameters optimization via the SCAPS-1D simulation tool [50,51]: HTL thickness, perovskite absorber layer (PAL) thickness, ETL thickness, the defect density in the perovskite absorber layer (PAL), the defect density at the interface between the absorber and electron transport layers (ETLs), and that between the absorber and hole transport layers (HTLs). Hence, herein, each key factor on PCE was optimized through a SCAPS-1D simulation, and the effects of each factor on the PCE were clarified.

2. Materials and Methods

In this study, the current density–voltage (J–V) plot, V_oc, J_sc, fill factor (FF), and PCE of the ITO/TiO₂/MAPbI₃/Cu₂O/Au solar cell with MAPbI₃ as an absorber were calculated using the numerical simulation program SCAPS-1D [50,51]. In the SCAPS-1D program, each output value is derived through Equations (1)–(5) [52,53,54].

Poisson’s equation is commonly used for simulating semiconductor devices:

$$\frac{d^2}{{dx}^2}\psi \left(x\right)=\frac{e}{{\epsilon }_0{\epsilon }_r}(p\left(x\right)-n\left(x\right)+N_D-N_A+{\rho }_P-{\rho }_n)$$

(1)

Here, $\psi$ is the electrostatic potential, e is the electrical charge, ε_r is the relative permittivity, ε₀ is the vacuum permittivity, p and n are the hole and electron concentrations, N_D and N_A are the charged impurities of the donor and acceptor types, and ρ_P and ρ_n are the distributions of holes and electrons, respectively.

The continuity equations for electrons and holes are:

$$\frac{d{j}_n}{dx}=G-R$$

(2)

$$\frac{d{j}_p}{dx}=G-R$$

(3)

Here, j_n is the electron current density, j_p is the hole current density, G is the generation rate, and R is the recombination rate.

$$J_n=D_n\frac{d_n}{d_x}+{\mu }_{n}n\frac{d\mathsf{\phi }}{d_x}$$

(4)

$$J_p=D_p\frac{d_p}{d_x}+{\mu }_{p}p\frac{d\mathsf{\phi }}{d_x}$$

(5)

In the above drift-diffusion model, ${\mu }_n$ is the mobility of electrons, ${\mu }_p$ is the mobility of holes, D_n is the electron diffusion coefficient, and D_p is the hole diffusion coefficient. Figure 1a shows the basic structure of the PSC considered in this study: ITO/TiO₂/MAPbI₃/Cu₂O/Au. ITO was used as the front contact, TiO₂ as the ETL, MAPbI₃ as the PAL, Cu₂O as the HTL, and Au as the back contact. The energy band alignment of this structure is shown in Figure 1b. The work function (W/F) of ITO is −4.6 eV, the lowest unoccupied molecular orbital (LUMO) in TiO₂ is at −3.9 eV, and the LUMO in MAPbI₃ are at −3.9 eV. Thus, the transfer of electrons to ITO was expected to proceed well. In addition, the highest occupied molecular orbital (HOMO) in MAPbI₃ is at −5.45 eV, and the HOMO in Cu₂O are at −5.37 eV [55,56,57,58]. The W/F of Au is −5.1 eV [59]. Therefore, the migration of holes from the absorption layer to the Au electrode, which is the lower electrode, was also expected to proceed well.

Table 1. The initial physical parameters of TiO₂, MAPbI₃, and Cu₂O used in the ITO/TiO₂/MAPbI₃/Cu₂O/Au PSC.

Parameters	TiO2 [55,56]	MAPbI3 [57]	Cu2O [58]
Thickness (nm)	100	600	400
Bandgap (eV)	3.26	1.55	2.17
Electron affinity (eV)	4.2	3.9	3.2
Dielectric permittivity (relative)	10	32	6.6
CB effective density of states (cm−3)	${2.2\times 10}^{18}$	${2.8\times 10}^{20}$	${2.5\times 10}^{20}$
VB effective density of states (cm−3)	${1.8\times 10}^{18}$	${3.9\times 10}^{20}$	${2.5\times 10}^{20}$
Electron mobility (cm2/Vs)	20	11.8	80
Hole mobility (cm2/Vs)	10	11.8	80
Shallow uniform acceptor density NA (cm−3)	0	${1\times 10}^{13}$	${3\times 10}^{18}$
Shallow uniform donor density ND (cm−3)	${1\times 10}^{17}$	${1\times 10}^{13}$	0
Defect density Nt (cm−3)	${1\times 10}^{15}$	${3\times 10}^{14}$	${1\times 10}^{15}$

shows the parameters of each layer used in the simulation here. Variables not specified in Table 1 are the thermal velocity of electrons/holes of 10⁷ cm/s, the interface defect density of 1 × 10⁻¹³ cm⁻³, the hole capture cross-section of the absorber/ETL of 10⁻¹⁸ cm², the hole capture cross-section of the absorber/HTL of 10⁻¹⁹ cm², the electron capture cross-section of the absorber/ETL of 10⁻¹⁹ cm², and the electron capture cross-section of the absorber/HTL of 10⁻¹⁸ cm². A series resistance of 1 Ω-cm², shunt resistance of 4200 Ω-cm², constant illumination of 1000 W/m², and operating temperature of 300 K at AM 1.5G were used [55].

3. Results and Discussion

3.1. Performance of the Perovskite Solar Cell under Initial Conditions

Prior to the optimization of each factor affecting the efficiency of PSCs, the initial conditions for the PSC device were set: an ETL TiO₂ layer thickness of 100 nm, a PAL MAPbI₃ layer thickness of 600 nm, and an HTL Cu₂O layer thickness of 400 nm. Figure 2 shows the current density–voltage (J–V) graph of the initial PSC. The maximum optical conversion efficiency was calculated using Equations (6) and (7) [60] while considering the initial conditions as 13.79%.

$$FF=\frac{V_{max}{J}_{max}}{V_{oc}{J}_{sc}}$$

(6)

$$PCE=\frac{V_{oc}{J}_{sc}FF}{P_{in}}$$

(7)

Here, FF is the fill factor, V_max is the maximum voltage, J_max is the maximum current density, V_oc is the open-circuit voltage, J_sc is the short-circuit current density, PCE is the power conversion efficiency, and P_in is the incident power of the sun (

Table 2. Solar cell main parameters for the ITO/TiO₂/MAPbI₃/Cu₂O/Au PSC structure.

Voc	Jsc	FF	PCE
0.83 V	21.54 mA/cm2	76.68%	13.79%

).

3.2. Optimization of HTL Thickness

The PCE of PSCs is affected by the properties of ETL and HTL materials. Furthermore, the thickness of the thin film is closely related to its electrical properties, which determine the transport of electrons and holes. Thickness is also directly related to PCE. Thus, as shown in Figure 3, we first investigated the effects of thickness on the V_oc, J_sc, FF, and PCE of the Cu₂O HTL layer. Extremely thin HTL film forms micro-pinholes, resulting in direct contact between the absorption layer and the rear electrode, and an extremely thick HTL can increase the series resistance in the interface between the metal electrode and the HTL [61].

Notably, as shown in Figure 3, the PCE value does not change with an increasing HTL thickness at the initial acceptor density of the HTL layer from Table 1. To understand such behavior from the PCE, the effects of HTL thickness on V_oc, J_sc, and FF were examined. Similar to PCE, all these parameters remain almost constant with an increasing HTL thickness. In general, as the thickness of the charge transport layer increases, the physical distance that the charge can migrate across increases. However, the results in Figure 3 demonstrate that the PCE remains almost constant with an increasing HTL thickness. To clarify these results further, we examined the factors that had the most significant effect by changing the level of each parameter specified in Table 1. Among them, the impact of the HTL acceptor density was most significant, as shown in Figure 4, with the acceptor density of Cu₂O being adjusted from 10¹¹ to 10²⁰ cm⁻³.

Figure 4 shows the PCE values according to the thickness and acceptor density of Cu₂O. Figure 4 confirms that the efficiency decreases with an increasing thickness when the acceptor density of HTL is 10¹⁵ cm⁻³ or less. In contrast, when the acceptor density is 10¹⁶ cm⁻³ or higher, the PCE changes insignificantly due to the increase in thickness. To investigate the cause of this effect, the relationship between the PCE and acceptor density was examined for a 100 nm HTL thickness, where we obtained a remarkable change. Figure 5 shows the PCE as a function of the acceptor density at a Cu₂O thickness of 100 nm. This result indicates that the higher the acceptor density, the greater the efficiency. In particular, the rapid increase in the values of 10¹⁶ cm⁻³ or higher appears to be due to the improvement of sheet resistance and conductivity at a high doping rate [62,63].

3.3. Optimization of Perovskite Absorber Layer Thickness

The absorption layer critically affects solar cell efficiency because it absorbs solar light and produces electron–hole pairs. Therefore, the optimization of PAL thickness is most important.

Figure 6 shows the current density–voltage (J–V) plot for various PAL thicknesses, and Figure 7 shows V_oc, J_sc, FF, and PCE values as functions of PAL thickness (100–1300 nm). As shown in Figure 7d, with an increasing PAL thickness, the PCE gradually increases until the thickness of the PAL reaches 1300 nm with the highest conversion efficiency of 19.30%. Afterwards, the PCE tends to decrease with an increasing PAL thickness. This decrease in PCE was attributed to the increasing electron–hole recombination in the absorption layer, rather than the generation of electron–hole pairs [64,65,66]. The J_sc value increases as the thickness of PAL increases, which is due to improved light absorption [67]. It increases sharply from 15.18 mA/cm² to 23.70 mA/cm² for the range of 100–700 nm. Beyond this thickness, it increases slowly until it saturates at 24.91 mA/cm². Additionally, as V_oc increases from 100 nm to 1200 nm, PCE also increases. Afterward, from 1300 nm to 2000 nm (not shown in Figure 6), V_oc slightly decreases. The initial increase in V_oc indicates a reduction in charge carrier recombination within the thin absorber layer, but as the thickness increases beyond 1300 nm, the decrease in V_oc could be due to a reduction in the effective bandgap and increased recombination [68].

3.4. Optimization of ETL Thickness

ETL is the first layer in the PSC through which light passes, and its thickness has a significant impact on efficiency. To find out the effect of ETL on PSC, the thickness was changed from 50 to 200 nm for the simulation.

Figure 8 exhibits the current density-voltage (J–V) graphs for different ETL (TiO₂) thicknesses, and Figure 9 shows the effects of ETL thickness on V_oc, J_sc, FF, and PCE. As shown in Figure 9d, PCE increases with ETL thickness until the thickness of the ETL layer reaches 100 nm and decreases afterwards. The maximum PCE of 19.30% is observed at an ETL thickness of 100 nm. An overly thin ETL leads to leakage current and device failure due to discontinuity. Therefore, as the thickness increases up to 100 nm, the efficiency also increases. After reaching the maximum of 19.30% at 100 nm, the PCE efficiency decreases with an increasing ETL thickness because, similar to the HTL, the charging resistance and series resistance rapidly increase with the ETL thickness, resulting in a decrease in PCE due to absorption loss [69,70].

3.5. Optimization of PAL Defect Density

The defects of the absorption layer can interfere when the carriers in the device move to electrodes [71]. The PAL defect density was changed from 10¹⁰ to 10²⁰ cm⁻³ to determine its effect on the PSC characteristics. Figure 10 and Figure 11 show the current density–voltage (J–V) graphs and changes in PCE according to the PAL defect densities. With an increase in defect density up to 10¹⁵ cm⁻³, the PCE remains at ~17%; however, with a further increase in defect density, PCE decreases to 13%. The Shockley–Read–Hall model shows that an increase in carrier life can reduce recombination [72,73,74].

$$\tau =\frac{1}{\sigma V_{th}{N}_t}$$

(8)

$$\mathrm{L}=\sqrt{D\tau }$$

(9)

$$\mathrm{D}=\frac{\mu kT}{q}$$

(10)

Here, τ, σ, V_th, D, and L are the minority carrier lifetime, capture cross-section, thermal velocity, diffusion coefficient, and diffusion length, respectively.

A low defect density prolongs carrier life and extends diffusion length, thereby reducing recombination (Equations (8) and (9)). Therefore, the defects in the absorption layer should be minimized to inhibit recombination.

3.6. Optimization of Interface Defect Density

The effect of the interface defect density on the PCE of PSC was investigated for the absorber/ETL and absorber/HTL interface with interface defect densities of 10¹³ to 10²⁰ cm⁻³ (Figure 12). As shown in Figure 12a, the PCE steadily decreases as the interface defect density between the absorber and ETL layer increases. The interface defect density decreases from 18.45% to 8.26%. As shown in Figure 12b, when increasing the interface defect density of the absorber/HTL up to 10¹⁴ cm⁻³, the PCE also remains almost constant (18%). The PCE then decreases, reaching 17.28% at 10¹⁸ cm⁻³. The decrease in PCE when increasing the interface defect density was attributed to the increase in the number of traps [55].

Comparing the range of change in Figure 12a,b, it can be seen that the change in the interface defect density between the absorber and the ETL has a greater effect on the PCE than that in the interface defect density between the absorber and the HTL. As shown in Figure 1a, upon irradiation, carriers started to be generated at the interface between absorber/ETL before those at the absorber/HTL interface. Thus, the change in the interface defect density between the absorber/ETL layer has a greater effect on the loss rate. This seems to have led to a rapid decrease in PCE [75].

3.7. Optimized PSC Device

Efficiency optimization of the ITO/TiO₂/MAPbI₃/Cu₂O/Au structure was performed using the optimal level of each factor. Figure 13 shows the current density–voltage (J–V) graph of the optimized PSC. The initial structure is ITO/TiO₂ (100 nm)/MAPbI₃ (600 nm)/Cu₂O (400 nm)/Au, whereas the optimized structure is ITO/TiO₂ (100 nm)/MAPbI₃ (1300 nm)/Cu₂O (400 nm)/Au. The PCE of the optimized structure is higher by 5.51% than that of the initial structure (19.30% and 13.79%, respectively). The key factor for this increase in efficiency is considered to be the increase in PAL thickness. At a PAL thickness of 1300 nm, the PCE showed the highest value of 19.30%. The more sunlight was absorbed, the more hole–electron pairs could be created [76]. Therefore, it is thought that these increased hole–electron pairs have contributed to the enhancement of PCE.

4. Conclusions

This study explored the influence of various parameters on the efficiency of MAPbI₃ perovskite solar cells (PSCs) through SCAPS-1D simulation analysis. The examined PSC featured a layered configuration of ITO/TiO₂/MAPbI₃/Cu₂O/Au, with a focus on the electron transport layer (ETL) thickness, perovskite absorption layer (PAL) thickness, and hole transport layer (HTL) thicknesses, as well as the PAL defect density and interface defect densities between the absorber and ETL layer and between the absorber and HTL layer. Optimal conditions achieving the highest photovoltaic conversion efficiency included an ETL (TiO₂) thickness of 100 nm, PAL (MAPbI₃) thickness of 1300 nm, and HTL (Cu₂O) thickness of 400 nm, with a PAL defect density at 10¹⁴ cm⁻³ and interface defect densities at 10¹³ cm⁻³ for both the absorber/ETL interfaces and absorber/HTL interfaces. These optimizations enhanced the PSC’s power conversion efficiency (PCE) to 19.30%, a significant increase from the baseline of 13.79% by 5.51%. The findings of this study are expected to contribute significantly to the optimization and efficiency enhancement of perovskite solar cells, particularly those employing inorganic charge carrier transport layers.