Achieving 27.20% efficiency for a lead-free double perovskite solar cell with all inorganic Cs₂BiAgI₆ using AZO UTL as a passivation layer

Abstract

A major challenge in the commercialization of perovskite solar cells (PSCs) is the presence of toxic metals, like lead, in their composition. Compared with conventional lead halide perovskites, double halide perovskites have garnered significant interest owing to their reduced toxicity, adjustable bandgap, structural flexibility, and enhanced stability. This study focuses on evaluating a lead-free Cs₂BiAgI₆-double perovskite solar cell (DPSC) using a one-dimensional solar cell capacitance simulator (SCAPS-1D) with a bilayer ZnO/AZO electron transport layer (ETL) and ZnO ETL, along with various hole transport layers (HTLs) for the first time. The selected HTLs included CBTS, Cu₂O, CuAlO₂, CZTS, CuSCN, spiro-OMeTAD, MoO₃, and V₂O₅. Various factors, such as energy band alignment, recombination and generation rates, absorber thickness, defect and doping densities for all layers, energy levels of ETLs and HTL, interfacial defect densities, back metal contact, and operating temperature, were examined for improving the performance of DPSC. This study was aimed at enhancing the efficiency and deepening our understanding of the electron transport mechanisms in Cs₂BiAgI₆-DPSCs. The research findings suggested that V₂O₅ and ZnO/AZO were the most suitable materials for the HTL and ETL, respectively, among the various options considered. Therefore, we utilized ITO/ZnO/AZO/Cs₂BiAgI₆/V₂O₅/Au as the required DPSC. To boost the performance of the DPSC, electron–hole pair handling at the ETL/perovskite interface was optimized by adding a 10 nm AZO UTL, thereby enhancing the ZnO/double perovskite interface properties. The bilayer structure of ZnO/AZO offered advantages such as efficient electron extraction and minimal interfacial recombination owing to its enhanced energy level alignment and defect passivation. After optimizing these parameters, the system with the ZnO/AZO bilayer ETL achieved an efficiency of 27.20%, along with a V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², and FF of 86.28%. Thus, this work presents a straightforward and promising approach for fabricating photovoltaic devices, particularly for various types of double perovskites, with favorable charge transport layers and recombination properties. Furthermore, these findings offer theoretical guidance to improve the efficiency of Cs₂BiAgI₆-based photovoltaic solar cells (DPSCs) and facilitate the widespread adoption of eco-friendly and stable perovskites.

---

1. Introduction

To meet the energy demands of daily life, we use fossil fuels, such as natural gas and coal, which are harmful to society and the environment.^1,2 Hence, the focus of global research is shifted to the development of alternative energy sources and sustainable energy options that address both economic and environmental concerns. Solar cells are a key renewable energy source, especially perovskite ones, which have attained a significant increase in efficiency from 3.8% to 26.1% since 2009.^1,2 The ABX₃ formula of perovskite solar cells (PSCs) has a monovalent cation (MA⁺ = CH₃NH₃⁺, FA⁺ = H₂NCHNH₂⁺, and Cs⁺) as A, a divalent metal cation (lead (Pb²⁺), tin (Sn²⁺), titanium (Ti²⁺), bismuth (Bi²⁺), and silver (Ag²⁺)) as B, and a halogen anion (Cl⁻, Br⁻, and I⁻) as X. Lead–halide and organic–inorganic hybrid PSCs have gained attention owing to their high power conversion efficiency (PCE) of 26.1%, which is comparable to top silicon solar cells.^3–5 Despite this remarkable efficiency of PSCs, the presence of toxic lead materials and volatile organic cations in their composition hinders their widespread application.^6–11 Organic–inorganic PSCs are unstable owing to the presence of oxygen, moisture, and heat generated by organic cations.¹² Moreover, lead-based perovskites are very toxic and pose health risks.^13–15

The demand for non-toxic PSCs has led to the research on lead-free materials. To replace lead-based perovskites, Sn²⁺ and Ge²⁺ (germanium) were used instead of Pb²⁺ in previous studies. However, Sn²⁺ and Ge²⁺ in lead-free PSCs had limited stability owing to oxidation.¹⁶ Pb²⁺ were also replaced with non-toxic Bi³⁺ ions, which are isoelectronic and stable.¹⁷ Bismuth PSCs showed longer charge carrier diffusion owing to their reduced trap densities.¹⁸ Introducing Bi³⁺ in A¹⁺B²⁺X₃ resulted in optoelectronic properties that were worse than those of lead perovskites.^19,20 To recover from these inferior characteristics, Bi³⁺ anion was added to an elpasolite structure or a double perovskite structure.²¹ Elpasolite structure has the formula A₂B¹⁺B³⁺X₆, where A represents a monovalent cation, X represents a halide anion, B¹⁺ represents an inorganic cation, and B³⁺ represents an organic or inorganic cation. According to recent investigations, double perovskites incorporating Bi³⁺ and Ag¹⁺ ions exhibit considerable promise for application in photovoltaic technologies because of their advantageous bandgaps, comparable charge carrier effective masses, exceptional photoluminescence lifetimes, extended carrier recombination lifetimes, and robust stabilities.^22–27 McClure et al. have stated that Cs₂AgBiBr₆ and Cs₂AgBiCl₆ show a noticeable bandgap and exhibit higher stability than MAPbX₃.²¹ However, Cs₂AgBiBr₆ and Cs₂AgBiCl₆ exhibit low efficacy owing to their high charge carrier effective masses, restricted charge carrier transport capabilities, and substantial band gap,^28–30 making them unsuitable for integration into solar cells. The absorber Cs₂AgBiI₆ exhibits a band gap of 1.12 eV, excellent light absorption capability and enhanced efficacy compared to Cs₂AgBiBr₆ and Cs₂AgBiCl₆, indicating its potential for double perovskite solar cell (DPSC).^31,32

The physical properties of a material in a device are key to understanding its state and potential practical applications. Researchers have used density functional theory (DFT) to study material properties, such as halide perovskites. These materials have special properties spotless for optoelectronic and photovoltaic applications.^33–35 Hadi et al.³⁶ studied Cs₂AgBiBr₆ using DFT to explore its properties by inducing disorder in the compound through the creation of an antisite defect in the sublattice; the indirect band gap of Cs₂AgBiBr₆ was reduced and converted to a direct band gap. This modification enhanced the optical absorption in the visible region, making Cs₂AgBiBr₆ suitable for solar cell applications.

Here, we conduct a literature review focusing on the ZnO/AZO bilayer structure, exploring its various properties, applications, potential advancements, and efficiency enhancement in the field of PSCs.³⁷

Dong et al.³⁸ reported that ZnO (zinc oxide) nanorods modified with aluminum-doped ZnO (AZO) are utilized in PSCs containing MAPbI₃. This modification has demonstrated a beneficial impact on both the V_oc (open-circuit voltage) and the PCE. The average PCE is enhanced from 8.5% to 10.07%, with the maximum efficiency reaching 10.7%. Tseng et al.³⁹ reported that Al doping proved to be effective in altering the physicochemical characteristics of ZnO to enhance its performance. A high-quality, fully coated thin film of AZO on an ITO (indium tin oxide) substrate was successfully fabricated using a sputtering technique. When compared to a cell utilizing ZnO, a perovskite cell incorporating AZO as the ETL exhibits superior stability. The most efficient PSC based on AZO achieves a PCE of 17.6% for the ZnO ETL-based PSC. Wu et al.⁴⁰ reported that the ZnO/perovskite interface has several drawbacks, including the decomposition of MAPbI₃ and misaligned energy levels. To solve these issues, we suggest a new design using a low-temperature ZnO/AZO bilayer thin film with band alignment as electron transport layers in PSCs. This enhances PSC efficiency. The PCE increases from 12.3% to 16.1% with the incorporation of the AZO thin film. In addition, some researchers have worked experimentally on the performance of PSCs to improve efficiency using ZnO/AZO bilayer.^41–44

To the best of our knowledge, for the first time, the evaluation of lead-free Cs₂BiAgI₆-DPSC using a one-dimensional solar cell capacitance simulator (SCAPS-1D) with bilayer ZnO/AZO ETL and ZnO ETL with various HTLs is carried out and compared. Various HTLs, including CBTS (copper barium thiostannate), Cu₂O (copper(I) oxide), CuAlO₂ (copper aluminum oxide), CZTS (copper zinc tin sulfide), CuSCN (copper(I) thiocyanate), Spiro-OMeTAD (2,2′,7,7′-tetrakis[N,N-di(4-ethoxyphenyl)amino]-9,9′-spirobifluorene), MoO₃ (molybdenum trioxide), and V₂O₅ (vanadium oxide) are selected. Next, we investigated the impact of various factors to enhance the performance of the DPSC. These factors included the energy band alignment, total recombination and total generation rate, thickness of the absorber, defect and doping densities of all layers, the energy levels of both ETLs and HTL, the interfacial defect densities of both the ETL and HTL sides, the back metal contact, and operating temperature. This study has the potential to enhance efficiency and deepen our understanding of the electron transport mechanism in Cs₂BiAgI₆-DPSCs. Additionally, simulation work was carried out on the most efficient cells to gain a more comprehensive understanding of the device's electrical properties. Ultimately, it was revealed that these interfacial bilayers significantly enhanced the photovoltaic characteristics and overall performance of the DPSC.

2. Materials and methodology

2.1. Methodology

Various simulation software, such as SCAPS-1D, PC-1D, AMPS-1D, wxAMPS, COMSOL, and Silvaco, are utilized to simulate the device structures and evaluate the photovoltaic performance of different kinds of solar cells.⁴⁵ In this particular study, SCAPS-1D version-3.3.08, developed by the Department of Electronics and Information Systems (ELIS) at the University of Gent, Belgium, is employed to simulate and model the DPSC with Cs₂BiAgI₆ as the active layer.⁴⁶ The simulation process involves using three semiconductor equations: eqn (1) represents the Poisson equation, which establishes the relationship between carrier concentrations and electrostatic potential, while eqn (2) and (3) correspond to the continuity equations for electrons and holes, respectively. These equations enable the analysis of the charge carrier generation and recombination mechanisms in the semiconductor.where ψ denotes electric potential; ξ denotes the electric field; ε denotes the permittivity; q denotes the electronic charge; p(x) and n(x) are concentrations of charge carriers; τ_n,p represents the carrier lifetime, μ_n,p denotes charge mobilities; N_A⁻ and N_D⁺ denote shallow acceptor and shallow donor concentrations, respectively; D_n,p denotes the diffusion coefficient of the charge carrier; G_n or G_p denotes the charge carrier generation rate, and p_t(x) denote the defect densities of electrons and holes, respectively. At the (x) and p_t(x) equilibrium state, we have the following equation:

2.2. Device structure and material parameters

In the formation of a double perovskite solar cell, the Cs₂BiAgI₆ absorber layer is accompanied by ETL, HTL, and back contact to create the DPSC structure, as it captures photons effectively owing to its double heterostructure, ensuring charge and photon confinement. The performance of various double perovskite solar cell structures (DPSCs) is investigated using SCAPS-1D software, with the ambient temperature set at 300 K and the AM 1.5G sunlight spectrum. Different structures are studied by incorporating two ETLs and eight HTLs, with a gold back contact.

The Cs₂BiAgI₆-DPSC is composed of multiple layers, including indium tin oxide (ITO), ZnO as ETL, Cs₂BiAgI₆ as the double perovskite absorbing (DPPVK) layer, and CBTS as HTL.⁴⁷ Notably, the DPPVK layer, Cs₂BiAgI₆, exhibits p-type carrier characteristics, with a bulk defect density of 1 × 10¹⁵ cm⁻³. This parameter significantly affects the charge carrier lifetime of electrons and holes, resulting in a value of 100 ns for both. The diffusion lengths of the electron and hole charge carriers, denoted by L_n and L_p, respectively, are measured to be 0.72 μm. For simulation purposes, the DPPVK layer is assumed to have a thickness of 800 nm, as shown in Table 1. The thermal velocity of electrons and holes is confirmed to be 1 × 10⁷ cm s⁻¹. Additionally, the defect energy level is positioned at the center of E_g, following a Gaussian distribution with a characteristic energy of 0.1 eV. These defects are electrically neutral and exhibit optimized capture cross sections for electrons and holes, with a value of 1 × 10⁻¹⁵ cm². The pre-factor for all the layers, denoted as A_α, is 1 × 10⁵ (cm⁻¹ eV^−0.5).⁴⁷ These values are determined ×using the following equation: α = A_α(hν − E_g)^0.5, where hν represents the photon energy. Fig. 1(a) depicts a schematic representation of the Cs₂BiAgI₆-DPSC device configuration. The defect densities of each layer are specified in parentheses. Fig. 1(b) shows the energy band alignment for two ETLs, eight HTLs, and the DPPVK layer.

Table 1 Material parameters used in the simulation

Material parameters	ITO	ZnO	Cs2BiAgI6	CBTS
Thickness (nm)	500	50	800	100
Energy band gap, Eg (eV)	3.5	3.3	1.6	1.9
Electronaffinity, χ (eV)	4	4	3.9	3.6
Relativedielectricpermittivity, εr	9	9	6.5	5.4
Conduction band effective density of state, NC (cm^−3)	2.2 × 10^18	3.7 × 10^18	1 × 10^19	2.2 × 10^18
Valence band effective density of state, NV (cm^−3)	1.8 × 10^19	1.8 × 10^19	1 × 10^19	1.8 × 10^19
Electronmobility, μn (cm^2 V^−1 s^−1)	20	100	2	30
Holemobility, μp (cm^2 V^−1 s^−1)	10	25	2	10
Shallow uniform donor doping density, ND (cm^−3)	1 × 10^21	1 × 10^18	0	0
Shallow uniform acceptor doping density, NA (cm^−3)	0	0	1 × 10^15	1 × 10^18
Defect density, Nt (cm^−3)	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15
Electron thermal velocity, Ve (cm s^−1)	1 × 10^7	1 × 10^7	1 × 10^7	1 × 10^7
Hole thermal velocity, Vh (cm s^−1)	1 × 10^7	1 × 10^7	1 × 10^7	1 × 10^7
Ref.	47	47 and 48	47	47

---

Fig. 1 (a) Device structure of the Cs₂AgBiI₆-DPSC. The defect densities of each layer (N_t) are shown in the parentheses. (b) Energy band alignment of ETLs, HTLs, and the DPPVK layer.

The validation of the model is confirmed through a comparison of the results obtained from Hossain et al.⁴⁷ with our model for the ITO/ZnO/Cs₂BiAgI₆ (800 nm)/CBTS/Au device. In Table 1, the material parameters for this device are presented. Table 2 shows the interface defect density parameters used in our simulation.

Table 2 Interface defect density parameters⁴⁷

Interface parameter	ZnO/Cs2BiAgI6	Cs2BiAgI6/CBTS
σ n,p: capture cross section for electron and hole; Et: defect energy level; EV: valence band minimum energy; Nt: interface defect density.
Defect type	Neutral	Neutral
σ n (cm^2)	1 × 10^−17	1 × 10^−18
σ p (cm^2)	1 × 10^−18	1 × 10^−19
Distribution of energy	Single	Single
E t–EV	Above the VB maximum	Above the VB maximum
Energy level w.r.t. reference (eV)	0.6	0.6
N t (cm^−2)	1 × 10^10	1 × 10^10

---

Fig. 2(a) shows our simulation of the current density–voltage curve of DPSC, ITO/ZnO/Cs₂BiAgI₆ (800 nm)/CBTS/Au, with the PV output parameters of V_oc of 1.0890 V, J_sc of 24.18 mA cm⁻², FF of 81.87%, and PCE of 21.56%, and Fig. 2(b) demonstrates the simulation of the EQE curve against the wavelength of the corresponding device. The EQE curve commences from 27.922% at 300 nm and reaches a peak of 99.807% at 360 nm. Then, it remains almost constant up to 600 nm and finally decreases to zero at 780 nm. Additionally, Fig. 2(c) shows the energy band alignment of this device.

Fig. 2 Simulation results for the (a) current density–voltage (J–V) curve, (b) external quantum efficiency (EQE) curve, and (c) energy band alignment of the ITO/ZnO/Cs₂BiAgI₆ (800 nm)/CBTS/Au device.⁴⁷

Table 3 displays the comparison between the Hossain et al.⁴⁷ result and our simulated performance parameters of ITO/ZnO/Cs₂BiAgI₆ (800 nm)/CBTS/Au.

Table 3 Comparison of the results of Hossain et al.⁴⁷ with our simulation of performance parameters for ITO/ZnO/Cs₂BiAgI₆ (800 nm)/CBTS/Au

PV parameters	V oc (V)	J sc (mA cm^−2)	FF (%)	PCE (%)
V oc (open-circuit voltage), Jsc (short-circuit current density), FF (fill factor), and PCE (power conversion efficiency).
Hossain et al.^47	1.085	23.76	83.78	21.59
Simulation	1.089	24.18	81.87	21.56

---

In this step, we examined AZO ETL and various HTLs to enhance efficiency. It is presumed that the N_D (cm⁻³) and N_t (cm⁻³) values of AZO ETL are the same as those of ZnO. It is assumed that the thickness of all HTLs is the same as the value of CBTS, and the values of N_A (cm⁻³) and N_t (cm⁻³) are equal to those of CBTS. Table 4 presents the input material parameters of AZO ETL and different HTLs used in our simulation.

Table 4 Input material parameters of AZO (ETL) and different HTLs used in our simulation

Material parameters	ETL				HTL
	AZO	Cu2O	CuAlO2	CuSCN	CZTS	MoO3	Spiro-OMeTAD	V2O5
Thickness (nm)	10	100	100	100	100	100	100	100
E g (eV)	3.4	2.17	3.46	3.6	1.45	3	3	2.2
χ (eV)	4	3.2	2.5	1.7	4.5	2.5	2.45	3.4
ε r	9	7.1	60	10	9	12.5	3	10
N C (cm^−3)	2 × 10^18	2.02 × 10^17	2.2 × 10^18	2.2 × 10^19	2.2 × 10^18	2.2 × 10^18	2.2 × 10^18	9.2 × 10^17
N V (cm^−3)	1.8 × 10^19	1 × 10^19	1.8 × 10^19	1.8 × 10^18	1.8 × 10^19	1.8 × 10^19	1.8 × 10^19	5 × 10^18
μ n (cm^2 V^−1 s^−1)	150	200	2	100	60	25	2 × 10^−4	3.2 × 10^2
μ p (cm^2 V^−1 s^−1)	25	80	8.6	25	20	100	2 × 10^−4	4 × 10^1
N D (cm^−3)	1 × 10^18	0	0	0	0	0	0	0
N A (cm^−3)	0	1 × 10^18	1 × 10^18	1 × 10^18	1 × 10^18	1 × 10^18	1 × 10^18	1 × 10^18
N t (cm^−3)	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15	1 × 10^15
Ref.	49–51	52	37	47	53	51	54 and 55	56

---

2.2.1. Influence of various ETLs and HTLs. Herein, we explored a DPSC based on Cs₂BiAgI₆ with an initial configuration using two kinds of ETLs, eight kinds of HTLs, and Au as a back contact. For this study, we used a total of eight HTLs comprising Cu₂O, CuSCN, V₂O₅, Spiro-OMeTAD, MoO₃, CuAlO₂, CZTS, CBTS, and two ETLs (ZnO and ZnO/AZO) to compare the PV performance for these 16 configurations. During the performance evaluation, the appropriate Cs₂BiAgI₆ absorber seems to be with HTLs, such as CBTS, CuAlO₂, CZTS, MoO₃, and V₂O₅ (Fig. 3). Compared with these HTLs, the performance of Cu₂O, CuSCN, and Spiro-OMeTAD was reduced when paired with any of the ETLs. As a p-type layer, the HTL must be thinner than the n-type ETL to avoid recombination and allow for the fast exchange of sufficient charge carriers in the structure. In Fig. 3(a), each of ZnO or ZnO/AZO ETLs and V₂O₅ HTL showed the highest V_oc of 1.894–1.897 V, while CuSCN HTL showed the lowest V_oc of 1.0845–1.0847 V. The J_sc of the CBTS HTL was the highest, 24.18 mA cm⁻², among all of the studied HTLs, as illustrated in Fig. 3(b). ZnO/AZO ETL and CuSCN HTL showed the lowest J_sc of 24.16 mA cm⁻². Fig. 3(c) depicts that the V₂O₅ HTL had the highest FF of 81.91–81.93%, whereas the Spiro-OMeTAD HTL demonstrated the lowest FF of 81.19–81.21% for all the studied transport layers. Fig. 3(d) illustrates the lowest performance of CuSCN HTL, 21.34–21.35%, while V₂O₅ HTL illustrates the highest performance of 21.58–21.59%. The V₂O₅ is characterized by several features, including excellent climate stability and strong optical and electrical properties. The V₂O₅ film can be fabricated by applying an inexpensive and simple spin-coating technique;⁵⁷ additionally, it was used as an HTL and interlayer between the perovskite absorber/HTL interface to achieve high solar cell performance.^58–61 V₂O₅ layer has a direct band gap of about 2.20 eV with good thermal stability, optical absorption coefficient, and long-term performance and has been recently put into the high priority production list because of its low environmental impact and low-cost fabrication techniques.⁶² Furthermore, V₂O₅ nanoparticles were suitable for the modification of PEDOT:PSS (poly(3,4-ethylenedioxythiophene) polystyrene sulfonate owing to their high stability in acid, good dispersibility in polar solvents, and excellent photoelectric properties, the). Therefore, the conductivity of the resulting anode interfacial layer (AIL) improved. It exhibited different AIL conductivities, where the V₂O₅:PEDOT:PSS had the highest conductivity owing to the V₂O₅ nanoparticles filling the pinholes, exposing more PEDOT:PSS chains to the surface core–shell structure.⁶³

Fig. 3 Photovoltaic output parameters of (a) V_oc (V), (b) J_sc (mA cm⁻²), (c) FF (%), and (d) PCE (%). Contour mapping plots of the device of ITO/ETLs/Cs₂BiAgI₆ (800 nm)/HTLs/Au for two ETLs and eight HTLs with different materials.

For wiser decision of results, Fig. 3 illustrates the PV output parameters (V_oc (V), J_sc (mA cm⁻²), FF (%), and PCE (%)) of the device of ITO/ETLs/Cs₂BiAgI₆ (800 nm)/HTLs/Au, for two ETL and eight HTL materials, as contour mapping plots.

Table S1 (ESI†) presents the PV output parameters of 16 devices for two ETLs and eight HTLs of Cs₂BiAgI₆-DPSC, and ITO/ETLs/Cs₂BiAgI₆ (800 nm)/HTLs/Au. Fig. S1a–d (ESI†) displays the J–V and EQE curves with the corresponding ETLs (a) and (b) ZnO/AZO and (c) and (d) ZnO for the device of ITO/ETLs/Cs₂BiAgI₆ (800 nm)/HTLs/Au. The eight HTLs are shown in the inset of Fig. S1a–d (ESI†).

According to Table S1 (ESI†) and Fig. 3, it can be concluded that the Cs₂BiAgI₆-DPSC with ETL (ZnO/AZO) and HTL (V₂O₅) has the maximum PCE, attaining 21.59%, together with a V_oc of 1.0897 V, J_sc of 24.17 mA cm⁻², and FF of 81.93%. We employ an ultra-thin layer of AZO (10 nm) on a ZnO, ITO/ZnO/AZO/Cs₂BiAgI₆ (800 nm)/V₂O₅/Au device in our simulation. By incorporating an AZO UTL layer, the extraction of electrons is improved at the interface, resulting in efficient carrier extraction, minimal leakage loss,⁶⁴ and reduced energy loss.⁶⁵ Additionally, it enhances the alignment of energy levels, facilitates electron transport, and enhances resistance to recombination.⁶⁶ A bilayer ETL can be designed to have superior film quality, lower trap density, and decreased charge accumulation at the ETL/perovskite interface in the devices. The bilayer structure also displays a uniform, smooth surface with fewer defects, which enhances charge extraction at the ETL/DPPVK layer junction compared to a single layer.⁶⁷ This strategy presents a promising and efficient approach for producing cost-effective, high-performance, and reliable planar DPSCs.⁴⁰

Fig. 4(a) presents the current density–voltage (J–V) curve, and Fig. 4(b) illustrates the external quantum efficiency (EQE) curve of ITO/ZnO/AZO/Cs₂BiAgI₆ (800 nm)/V₂O₅/Au device. The EQE curve originates from 27.782% at 300 nm and reaches a peak of 99.711% at 360 nm. Then, it remains constant until 600 nm and finally decreases to zero at 780 nm. Table S2 (ESI†) shows the interfacial defect density parameters of ITO/ZnO/AZO/Cs₂BiAgI₆ (800 nm)/V₂O₅/Au device.

Fig. 4 (a) J–V curve and (b) EQE curve for ITO/ZnO/AZO/Cs₂BiAgI₆ (800 nm)/V₂O₅/Au device.

3. Results and discussion

3.1. Energy band alignment

The energy band alignment between the passivation layer and the DPSC plays a crucial role in determining the performance of the DPSCs. The conduction band (CB) alignment between the passivation layer and the DPSC is particularly important for efficient electron extraction from double perovskite material. It is necessary to have a minimal offset between the CB of the passivation layer and the CB of the DPSC to ensure effective electron extraction. However, the valence band (VB) alignment between the passivation layer and the DPSC is important for blocking the movement of holes. A considerable difference in the VBs is required to prevent holes from migrating towards the ETL and causing recombination. Similarly, proper alignment between the valence band of the HTL and the double perovskite material is crucial for facilitating hole separation. Additionally, a significant offset in the conduction bands of the HTL and DPSC is necessary to prevent electron migration towards the HTL and minimize recombination. To achieve optimal band alignment in DPSCs, a minimal offset at the CB and a maximal offset at the VB between the double perovskite and the ETL are essential, enabling electron flow while blocking hole transmission.⁶⁸ Similarly, minimal VBO (valence band offset) and maximal CBO (conduction band offset) affect HTL and double perovskite characteristics. These offsets enable seamless transmission of holes from the absorber to the HTL while impeding electron mobility. Fig. S2 (ESI†) depicts the energy band alignment of the Cs₂BiAgI₆-DPSCs, while Table 5 presents the VBO and CBO assessed by the respective layers. The VBO and CBO values are calculated based on the material's electron affinity (χ) and energy band gap (E_g), as follows:

CBO = χ_DPPVK − χ_CTL, (5)

VBO = (χ_CTL + E_g,CTL) − (χ_DPPVK + E_g,DPPVK). (6)

Table 5 CBO and VBO at interfaces

Interface	CBO (eV)	VBO (eV)
Double perovskite/ETL
Cs2BiAgI6/ZnO	−0.1	1.8
Double perovskite/passivation
Cs2BiAgI6/AZO	−0.1	1.9
Passivation/ETL
AZO/ZnO	0	0.1
Double perovskite/HTL
Cs2BiAgI6/CBTS	0.3	0
Cs2BiAgI6/Cu2O	0.7	−0.13
Cs2BiAgI6/CuAlO2	1.4	0.46
Cs2BiAgI6/CuSCN	2.2	−0.2
Cs2BiAgI6/CZTS	−0.6	0.45
Cs2BiAgI6/MoO3	1.4	0
Cs2BiAgI6/Spiro	1.45	−0.05
Cs2BiAgI6/V2O5	0.5	0.1

---

The implementation of an interlayer (IL) is noted to modify the arrangement of energy levels within the films and hinder ion migration. The alignment of interface energy levels significantly affects the rate at which electrons are injected, particularly between the DPPVK layer and the ETL. The presence of an energy barrier at the interfaces results in the recombination of charge carriers, thereby restricting the efficiency of charge transfer. Conversely, the lack of an energy barrier across the interface promotes efficient charge transfer and injection, leading to reduced recombination rates. The addition of an AZO IL, which acts as a passivation layer at the interface of the DPPVK layer and the ETL, improves electron transport through the interface and could potentially reduce interface charge recombination.

3.2. Influence of total generation and recombination rate

Owing to the light incident on the double perovskite solar cell, Fig. 5a–d illustrates the generation and recombination rates for different ETLs. Electron transfers from valence to the conduction band, leaving a hole. The release of electrons and holes contributes to carrier generation. The SCAPS-1D software utilizes the incident photon flux N_phot(λ, x) to determine the creation of electron–hole pairs G(λ, x). By analyzing this photon flux at different positions and wavelengths, the value of G(λ, x) can be calculated as follows:

G(λ, x) = α(λ, x)·N_phot(λ, x), (7)

where α(λ, x) is the absorption coefficient and N_phot(λ, x) is the photon incident number. Recombination is the opposite of generation, involving the pairing and annihilation of electrons and holes in the conduction band. The recombination rate of a solar cell is determined by the lifespan and density of the charge carriers. Initially, electron–hole recombination is reduced owing to defect states in the absorber layer. Subsequently, the creation of energy states affects the electron–hole recombination profile in the structure of the solar cell. Recombination rates alter owing to defects, as shown in Fig. 5b and d.⁶⁹ Recombination rate in ZnO/AZO is a little smaller than that of ZnO for V₂O₅, as illustrated in Fig. 5b and d.

Fig. 5 Total generation and recombination rate of (a) and (b) ZnO/AZO and (c) and (d) ZnO ETLs.

3.3. Influence of the thickness of the double perovskite absorber (DPPVK) layer on the Cs₂BiAgI₆-DPSC performance

Herein, we determine the optimum parameters for the ITO (500 nm)/ZnO (50 nm)/AZO (10 nm)/Cs₂BiAgI₆ (800 nm)/V₂O₅ (100 nm)/Au device. The impact of varying the thickness of the double perovskite absorber (DPPVK) layer on the cell parameters was calculated. The DPPVK layer thickness ranged from 50 to 1000 nm in 20 increments, with the parameters shown in Fig. 6. In Fig. 6(a), the current density–voltage curves for DPPVK layers from 50 to 1000 nm in 20 steps are shown. As depicted in Fig. 6(b), increasing the DPPVK layer thickness from 50 to 1000 nm in 20 steps, up to 650 nm, boosts EQE; after that, it rises slightly. This increase is due to more charge carriers resulting from better light absorption.⁷⁰Fig. 6(c) illustrates the device's performance parameters in proportion to the thickness of the DPPVK layer changing from 50 to 1000 nm in 20 steps. The thickness of the DPPVK layer significantly influences the quality and performance of the DPSCs.

Fig. 6 (a) J–V curve, (b) EQE curve, and (c) performance parameters for different double perovskite absorber layer thicknesses ranging from 50 to 1000 nm in 20 steps.

In accordance with Fig. 6(c), the V_oc and FF of the DPPVK layer diminish as the thickness of the layer augments. This is because the excitons generated from photon absorption are unable to overcome the barrier potential of the depletion layer, resulting in a higher recombination rate of charge carriers and an increased reverse saturation current, which ultimately leads to a decrease in V_oc.

Increasing the thickness of the DPPVK layer also enhances photon absorption, resulting in the production of more excitons, which, in turn, increases the J_sc and the overall PCE of the DPSC.⁷¹ However, further increasing the thickness leads to higher series resistance, which introduces more defects and increases the recombination rate, thereby reducing the FF.^72,73

Therefore, when determining the optimal absorber thickness, it is crucial to consider all factors, including V_oc, J_sc, FF, and PCE. Experimental studies have demonstrated that the performance of DPSCs is heavily influenced by the morphology of the DPPVK layer, which directly affects the photogenerated lifetime and diffusion length.^74,75

Fig. 6(c) demonstrates the performance parameters to Cs₂BiAgI₆ thickness (50–1000 nm in 20 steps). The optimal value occurs at a thickness of 650 nm, with the corresponding PV output parameters of V_oc at 1.1005 V, J_sc at 23.78 mA cm⁻², FF of 82.78%, and PCE of 21.67%. The V_oc decreases from 1.2101 V to 1.0790 V, J_sc increases from 9.85 mA cm⁻² to 24.45 mA cm⁻², FF decreases from 85.07% to 80.76%, PCE increases from 10.14% to 21.67% at 650 nm, and then decreases to 21.31% at 1000 nm.

3.4. Influence of defect density in the double perovskite absorber (DPPVK) layer on the Cs₂BiAgI₆-DPSC performance

Various processes involving generation, recombination, and absorption occur within the DPPVK layer, directly affecting the quality of the layer and the defect density.⁷⁶ The efficiency of Cs₂BiAgI₆-DPSC is significantly affected by high defect densities in the DPPVK layer. These defects act as recombination centers, accelerating the recombination rate and shortening the lifespan of charge carriers, which ultimately leads to a decrease in device performance. In Cs₂BiAgI₆-DPSC, Shockley–Read–Hall (SRH) recombination is prevalent, and the trap-assisted SRH recombination model can be utilized to calculate the diffusion length.⁷⁷ The SRH recombination formula is represented as follows:where R_SRH denotes SRH recombination; n and p denote the electron and hole concentration density, respectively; n_i denotes intrinsic density; τ_n,p denotes carrier lifetime; E_i denotes the intrinsic energy level; and E_t denotes trap energy level. Thus, we havewhere σ_n,p denotes the capture cross section of electrons and holes, υ_th denotes the thermal velocity of charge carriers and N_t denotes the total bulk defect density of the DPPVK layer of the Cs₂BiAgI₆-DPSC. Moreover, the diffusion length L and the diffusion coefficient D are as follows:where k, T, q, and μ_n,p represent Boltzmann's constant, temperature, charge carriers, and carrier mobility, respectively.

The maximum distance that charge carriers can travel is determined by these equations, which are primarily influenced by the lifetimes of the carriers. In principle, the diffusion length of charge carriers depends mainly on their lifetimes. Furthermore, the defects found in the DPPVK layer serve as non-radiative SRH recombination centers.

Fig. 7a and b display the PV parameters against the defect density of the DPPVK layer, N_t,DPPVK, which ranges from 1 × 10¹³ cm⁻³ to 1 × 10¹⁷ cm⁻³. As demonstrated in Fig. 7, the optimal value occurs at a defect density of 1 × 10¹³ cm⁻³, with corresponding PV output parameters of V_oc at 1.321 V, J_sc at 23.83 mA cm⁻², FF of 84.50%, and PCE of 26.61%. As depicted in Fig. 7(a), V_oc decreases from 1.321 V to 0.8248 V as N_t,DPPVK increases from 1 × 10¹³ cm⁻³ to 1 × 10¹⁷ cm⁻³, and J_sc remains fixed at 23.83 mA cm⁻² for both 1 × 10¹³ cm⁻³ and 1 × 10¹⁴ cm⁻³, and 23.78 mA cm⁻² at 1 × 10¹⁵ cm⁻³ and then decreases to 2.10 mA cm⁻² at 1 × 10¹⁷ cm⁻³. As illustrated in Fig. 7(b), FF decreases from 84.82% to 58.15% as N_t,DPPVK increases from 1 × 10¹⁴ cm⁻³ to 1 × 10¹⁷ cm⁻³, and the PCE value is 24.39% at 1 × 10¹⁴ cm⁻³ and then decreases to 1.01% at 1 × 10¹⁷ cm⁻³.

Fig. 7 (a) V_oc and J_sc, and (b) FF and PCE curves for Cs₂BiAgI₆-DPSC against defect density of absorber.

3.5. Influence of acceptor doping concentration in the absorber layer on the Cs₂BiAgI₆-DPSC performance

The performance of the Cs₂BiAgI₆-DPSC device is greatly affected by the doping in the DPPVK layer. To explore the effect of acceptor doping density on the Cs₂BiAgI₆-DPSC performance, we initially model the device performance by adjusting the acceptor doping concentration of the DPPVK layer (N_A,DPPVK) from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³, as shown in Fig. 8.

Fig. 8 (a) V_oc and J_sc, and (b) FF and PCE curves for Cs₂BiAgI₆-DPSC against acceptor density of absorber.

Enhancing the density of dopants can increase the concentration of charge carriers, thus improving the performance of the Cs₂BiAgI₆-DPSC.^78,79 Nevertheless, excessively high levels of doping in the DPPVK layer can lead to reduced device performance, as high doping introduces more traps in the DPPVK layer that affect the mobility of charge carriers and result in increased recombination. Excessive doping densities also influence the semiconductor behavior of Cs₂BiAgI₆-DPSC, making it more metallic.

Fig. 8a and b show how the values of V_oc, J_sc, FF and PCE vary with the acceptor density of the DPPVK layer, which alters from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³. Regarding the acceptor density in the DPPVK layer, the highest efficiency of 27.18% was achieved at an acceptor density of 1 × 10¹⁶ cm⁻³, with the V_oc of 1.3213 V, J_sc of 23.83 mA cm⁻², and FF of 86.31%. As depicted in Fig. 8(a), V_oc is 1.321 V when N_A,DPPVK increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁵ cm⁻³ and then increases to 1.3213 V at 1 × 10¹⁶ cm⁻³, and J_sc remains unchanged at 23.83 mA cm⁻² from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³. As illustrated in Fig. 8(b), FF is 84.42% when N_A,DPPVK increases from 1 × 10¹² cm⁻³ and 1 × 10¹⁴ cm⁻³, marginally increases up to 84.50% at 1 × 10¹⁵ cm⁻³, and then increases to 86.31% at 1 × 10¹⁶ cm⁻³; the PCE value is 26.58% as N_A,DPPVK increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁴ cm⁻³, increases to 26.61% at 1 × 10¹⁵ cm⁻³, and increases to 27.18% at 1 × 10¹⁶ cm⁻³.

3.6. Influence of defect density of V₂O₅-HTL on the Cs₂BiAgI₆-DPSC performance

Deficiencies in the V₂O₅ (HTL) film lead to an increase in defect density, causing a higher rate of electron and hole recombination within the film. Consequently, the PCE of Cs₂BiAgI₆-DPSC decreases.^80–82Fig. 9 demonstrates the variation of V_oc, J_sc, FF and PCE for defect density of V₂O₅ and N_t,V2O5, altering from 1 × 10¹² cm⁻³ to 1 × 10¹⁸ cm⁻³. The optimum value increases at 1 × 10¹⁷ cm⁻³, and the PV output parameters of Cs₂BiAgI₆-DPSC are V_oc of 1.3213 V, J_sc of 23.83 mA cm⁻², FF of 86.31%, and PCE of 27.18%. As shown in Fig. 9, V_oc is 1.3213 V when N_t,V2O5 increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁷ cm⁻³ and then decreases to 1.3203 V at 1 × 10¹⁸ cm⁻³, J_sc is 23.83 mA cm⁻² when N_t,V2O5 increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁸ cm⁻³, FF is 86.31% when N_t,V2O5 increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁸ cm⁻³, and PCE value is 27.18% when N_t,V2O5 increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁷ cm⁻³ and then moderately falls to 27.17%.

Fig. 9 PV output parameters for Cs₂BiAgI₆-DPSC against the defect density of V₂O₅-HTL.

3.7. Influence of conduction and valence band offset of V₂O₅-HTL based on the Cs₂BiAgI₆-DPSC performance

In the Cs₂BiAgI₆-DPSC, a comprehensive numerical analysis is conducted with device simulation to understand the device operation mechanism and develop an optimal layer configuration for enhanced efficiency. It is essential to examine the electronic properties of HTL and ETL, such as their electron affinity and band gap, when designing highly efficient solar cells. In this study, we aim to determine the ideal energy levels for efficient Cs₂BiAgI₆-DPSC design. To achieve this goal, careful selection of CBO₁ and VBO₁ is necessary. In this study, we investigate how modifying the affinities and band gaps of the V₂O₅ can lead to optimal values of CBO₁ and VBO₁. To achieve a higher V_oc and reduce recombination, CBO₁ or VBO₁ must be positive and below 0.3 eV. Negative CBO₁ or VBO₁ values increase electron and hole accumulation, causing more recombination and lower charge output.⁸³ A positive CBO₁ or VBO₁ restricts electron movement, while a negative one facilitates it.⁸³ The charge polarity of CBO₁ and VBO₁ is determined by the barrier height of the charge carriers generated by light, as follows:

CBO₁ = χ_HTL − χ_DPPVK, (12)

VBO₁ = (χ_HTL + E_g,HTL) − (χ_DPPVK + E_g,DPPVK). (13)

The electron affinity of HTL changes from 1.5 eV to 2.5 eV, and VBO₁ is altered from −0.6 eV to +0.4 eV. Therefore, according to Table S3 (ESI†), when the energy band gap of HTL is concerned, the optimum value increases at 2.1 eV or VBO₁ is 0.0 eV, with the PV output parameters obtained as V_oc of 1.321 V, J_sc of 23.84 mA cm⁻², FF of 86.32%, and PCE of 27.19%. Fig. S3(a) (ESI†) shows the current density-volt xage curve, Fig. S3(b) (ESI†) illustrates the energy band diagram, and the inset of the figures demonstrates the variation of the energy level of V₂O₅-HTL based versus VBO₁ values. Fig. 10a–d illustrates the PV output parameters with a variation of VBO₁ between −0.6 eV and +0.4 eV. According to our calculations, there is no change in CBO₁, see Table S4 (ESI†).

Fig. 10 (a)–(d) PV output parameters with the variation of VBO₁ of V₂O₅/Cs₂BiAgI₆ between −0.6 eV and +0.4 eV.

3.8. Influence of defect density of AZO UTL on the Cs₂BiAgI₆-DPSC performance

Defects in the AZO UTL film lead to more electron–hole recombination. Consequently, the PCE of the Cs₂BiAgI₆-DPSC decreases.^84,85Fig. 11 shows the variation of V_oc, J_sc, FF and PCE for defect density of AZO UTL, N_t,AZO, increasing from 1 × 10¹² cm⁻³ to 1 × 10¹⁹ cm⁻³. The optimum value increases at 1 × 10¹⁶ cm⁻³, and the PV output parameters of Cs₂BiAgI₆-DPSC are V_oc of 1.321 V, J_sc of 23.84 mA cm⁻², FF of 86.32%, and PCE of 27.19%. As depicted in Fig. 11, V_oc is 1.321 V when N_t,AZO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³, slightly decreases to 1.3189 V at 1 × 10¹⁸ cm⁻³, and then increases to 1.3224 V at 1 × 10¹⁹ cm⁻³. J_sc is 23.84 mA cm⁻² when N_t,AZO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁸ cm⁻³ and then decreases to 22.96 mA cm⁻² at 1 × 10¹⁹ cm⁻³. FF is 86.32% as N_t,AZO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁷ cm⁻³, increases to 86.39% at 1 × 10¹⁸ cm⁻³, and then decreases to 81.90% at 1 × 10¹⁹ cm⁻³. PCE value is 27.19% as N_t,AZO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³; to some extent, increases to 27.18% at 1 × 10¹⁷ cm⁻³ and 27.17% at 1 × 10¹⁸ cm⁻³; and then decreases to 24.87% at 1 × 10¹⁹ cm⁻³.

Fig. 11 PV output parameters for Cs₂BiAgI₆-DPSC against the defect density of AZO UTL.

3.9. Influence of valence and conduction band offset of AZO UTL on the Cs₂BiAgI₆-DPSC performance

To determine the optimal energy level, it is crucial to carefully select the values of CBO₂ and VBO₂ for the AZO UTL.⁸³ The barrier heights (CBO₂ and VBO₂) can be influenced by photo-generated charge carriers, either positive or negative, as follows:

CBO₂ = χ_DPPVK − χ_AZO, (14)

VBO₂ = (χ_DPPVK + E_g,DPPVK) − (χ_AZO + E_g,AZO). (15)

Eqn (14) represents the electron affinity of the DPPVK layer (χ_DPPVK) and the electron affinity of the AZO UTL (χ_AZO). Eqn (15) presents the band gap energy of the DPPVK layer (E_g,DPPVK) and the band gap energy of the AZO UTL (E_g,AZO).

There is no change in VBO₂; see Table S5 (ESI†). In Table S6 (ESI†), PV output parameters are shown, the electron affinity of AZO ranges from 3.5 eV to 4.1 eV, and the CBO₂ values change from +0.4 eV to −0.2 eV. Fig. S4(a) (ESI†) shows the plots of current density–voltage curves for different values of CBO₂, while Fig. S4(b) (ESI†) illustrates the energy band diagram with various CBO₂ values as indicated by different colors. Fig. 12a–d presents the PV output parameters corresponding to several CBO₂ values. The optimal value for the DPPVK layer/AZO UTL interface occurs when the electron affinity of AZO UTL is 3.9 eV. The optimum value increases at CBO₂ = +0.0 eV because it has better energy alignment and an efficiency of 27.20%, while the PV output parameters of Cs₂BiAgI₆-DPSC are as follows: V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², FF of 86.29%, and PCE of 27.20%.

Fig. 12 (a)–(d) PV output parameters with the variation of CBO₂ of Cs₂BiAgI₆/AZO UTL between −0.2 eV and +0.4 eV.

3.10. Influence of defect density of ZnO layer on the Cs₂BiAgI₆-DPSC performance

Because of ZnO layer flaws, the defect density increases, leading to more charge carrier recombination. This leads to a decrease in the PCE of the Cs₂BiAgI₆-DPSC.⁸⁶ As depicted in Fig. 13, the defect density of ZnO changes from 1 × 10¹² cm⁻³ to 1 × 10¹⁸ cm⁻³. As illustrated in Fig. 13, V_oc is a constant value at 1.3222 V as N_t,ZnO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁷ cm⁻³, and then marginally decreases to 1.3220 V at 1 × 10¹⁸ cm⁻³. J_sc remains fixed at 23.84 mA cm⁻² as N_t,ZnO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³ and slightly decreases to 23.79 mA cm⁻² at 1 × 10¹⁸ cm⁻³. FF value is 86.28% when N_t,ZnO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁷ cm⁻³ and then fairly decreases to 86.27% at 1 × 10¹⁸ cm⁻³. PCE value is 27.20% when N_t,ZnO increases from 1 × 10¹² cm⁻³ to 1 × 10¹⁶ cm⁻³ and then scarcely decreases to 27.14% at 1 × 10¹⁸ cm⁻³.

Fig. 13 PV output parameters for Cs₂BiAgI₆-DPSC against the defect density of the ZnO layer.

3.11. Influence of interfacial defect density on the Cs₂BiAgI₆-DPSC performance

To optimize the efficiency of DPSCs, the quality of the junction plays a crucial role in determining the capabilities of the interface. Our focus in this stage was on adjusting the defect densities at the interface to emphasize the significance of interface quality and recombination rate in actual Cs₂BiAgI₆-DPSCs. Interactions between double perovskites and transport materials often contain multiple defect states, highlighting the importance of the quality of these interfaces for the PCE of Cs₂BiAgI₆-DPSC. Notably, the interface defect density near the illuminated side exerts a more pronounced influence on Cs₂BiAgI₆-DPSC performance compared to the back interface.⁸⁷ With an increase in these defects, additional trap levels emerge at the interface, consequently elevating the resistance of Cs₂BiAgI₆-DPSC. The heightened resistances impede the movement of charge carriers, promoting recombination and ultimately diminishing the PCE of DPSC.⁸⁸

Herein, there are three categories of interface defect densities: the first one is between the HTL and DPPVK layer, the second one is between AZO UTL and DPPVK layer, and the third one is between AZO UTL and ZnO layer.

In Fig. 14(a), it is demonstrated that the optimum value for the interface defect density (N_it) of the HTL/DPPVK layer (N_it changes from 1 × 10⁵ cm⁻² to 1 × 10¹⁵ cm⁻²) is 1 × 10¹⁰ cm⁻²; the PV output parameters are as follows: V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², FF of 86.28%, and PCE of 27.20%. Regarding HTL/DPPVK interface defect density, V_oc is 1.3222 V from 1 × 10⁵ cm⁻² to 1 × 10⁹ cm⁻² and then decreases to 1.1852 V at 1 × 10¹⁵ cm⁻². The parameter of J_sc remains constant, 23.84 mA cm⁻², as N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹² cm⁻² and then slightly decreases to 23.72 mA cm⁻² at 1 × 10¹⁵ cm⁻². The FF remains constant at 86.28% from 1 × 10⁵ cm⁻² to 1 × 10¹⁰ cm⁻², increases to 89.36% at 1 × 10¹⁴ cm⁻², and then decreases to 88.47% at 1 × 10¹⁵ cm⁻². The PCE remains immutable, 27.20%, from 1 × 10⁵ cm⁻² to 1 × 10¹⁰ cm⁻² and then decreases to 24.88% at 1 × 10¹⁵ cm⁻².

Fig. 14 PV output parameters for interface defect density of (a) V₂O₅ layer/DPPVK layer, (b) DPPVK layer/AZO UTL, and (c) AZO UTL/ZnO layer of Cs₂BiAgI₆-DPSC.

In Fig. 14(b), the DPPVK layer/AZO UTL interface defect density is changed from 1 × 10⁵ cm⁻² to 1 × 10¹⁵ cm⁻². V_oc is fixed at 1.3222 V as N_it increases from 1 × 10⁵ cm⁻² to 1 × 10⁹ cm⁻² and then decreases to 1.1750 V at 1 × 10¹⁵ cm⁻². J_sc remains unchanged at 23.84 mA cm⁻² as N_it enhances between 1 × 10⁵ cm⁻² and 1 × 10¹² cm⁻². At this point, it marginally dwindles to 23.83 mA cm⁻² at both 1 × 10¹³ cm⁻² and 1 × 10¹⁴ cm⁻² and then decreases to 23.80 mA cm⁻² at 1 × 10¹⁵ cm⁻². The FF value is fixed at 86.28% when N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹⁰ cm⁻² and subsequently increases to 89.17% at 1 × 10¹⁵ cm⁻². The PCE value is constant at 27.20% as N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹⁰ cm⁻² and subsequently decreases to 24.94% at 1 × 10¹⁵ cm⁻². Concerning the optimal value of this interface defect density, it occurs at 1 × 10¹⁰ cm⁻², with the following PV output parameters: V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², FF of 86.28%, and PCE of 27.20%.

AZO UTL/ZnO layer interface defect density is varied from 1 × 10⁵ cm⁻² to 1 × 10¹⁵ cm⁻². As depicted in Fig. 14(c), V_oc remains fixed at 1.3221 V as N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹⁵ cm⁻². J_sc remains constant at 23.84 mA cm⁻² as N_it enhances between 1 × 10⁵ cm⁻² and 1 × 10¹³ cm⁻², subsequently slightly decreases to 2383 mA cm⁻² at 1 × 10¹⁴ cm⁻², and then slightly decreases to 23.82 mA cm⁻² at 1 × 10¹⁴ cm⁻². FF remains constant at 86.28% when N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹⁵ cm⁻². PCE has a fixed value of 27.20% as N_it increases from 1 × 10⁵ cm⁻² to 1 × 10¹³ cm⁻²; at that point, it decreases to 27.19% at 1 × 10¹⁴ cm⁻²; and 27.18% at 1 × 10¹⁵ cm⁻². The optimum value of this interface defect density occurs at 1 × 10¹³ cm⁻², and the PV output parameters are as follows: V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², FF of 86.28%, and PCE of 27.20%.

In this article, we achieved an efficiency of 27.18% by optimizing the thickness, defect density, and acceptor doping concentration of the Cs₂BiAgI₆ absorber layer. Here, we want to find the optimum value of the defect density for ETL and HTL. Fig. 9 shows that the optimal defect density for the V₂O₅-HTL is 1 × 10¹⁷ cm⁻³ (PCE = 27.18%). In Fig. 11, the optimal defect density for the AZO UTL-ETL is 1 × 10¹⁶ cm⁻³ (PCE = 27.19%), while Fig. 13 depicts that the ZnO-ETL layer's optimal defect density is also 1 × cm⁻³ (PCE = 27.20%). Finally, Fig. 14(c) illustrates the optimal interface defect density for AZO/ZnO as 1 × 10¹³ cm⁻² (PCE = 27.20%).⁸⁹ The J–V–T (current density vs. voltage for various temperatures) and C–f–T (capacitance vs frequency for various temperatures) measurements can help researchers to determine the defect properties in the perovskite solar cell. Using the J–V–T technique, the J–V curves can be measured at different temperatures. After obtaining the amount of V_oc, and plotting it against different temperatures, this linear curve intersection with the V_oc-axis gives E_t/q (E_t: defect energy, q: electron charge); the amount of E_t is required in the SCAPS program.⁹⁰ We can also calculate the C–f curves for different temperatures and C–f–T (thermal admittance analysis, TAS method) using the SCAPS package, which might allow us to calculate the defect density as a function of defect energy N_t(E_t) from a set of C(f, T) experimental measurements. More information regarding the parameters (such as V_bi, built-in voltage) could be obtained by referring to the Mott–Schottky analysis, specifically (1/C²vs. V).^91,92 Moreover, some other techniques are available for considering the defect properties, such as temperature-dependent space charge limited current (SCLC); deep-level transient spectroscopy (DLTS); Laplace current DLTS (I-DLTS); steady-state photoluminescence (SSPL); time-resolved photoluminescence (TRPL); PL mapping; time-resolved microwave conductivity (TRMC); thermally stimulated current (TSC); transient photocapacitance (TPC); surface photovoltage (SPV) spectroscopy; and time-resolved spectroscopies such as transient absorption and reflection techniques, ultraviolet photoemission spectroscopy (UPS), and scanning tunneling microscopy (STM).⁹³

3.12. Influence of the work function of back metal contact on Cs₂BiAgI₆-DPSC performance

For the enhanced PCE of Cs₂BiAgI₆-DPSC, the back metal contact's work function is a critical factor. The value of ϕ_BC (work function) directly impacts the presence of an appropriate built-in electric field in Cs₂BiAgI₆-DPSC,⁹⁴ affecting the electric field that aids in hole transport and collection; thus, ϕ_BC significantly influences V_oc.⁹⁵ To study the effect of different back electrode ϕ_BC values on Cs₂BiAgI₆-DPSC performance, we explored ϕ_BC ranging from 4.65 eV to 5.9 eV.⁹⁶ Our findings from simulations are presented in Fig. 15.

Fig. 15 (a) J–V curve. Inset of the figure shows C (Carbon), Au (Gold), W (Tungsten), Ni (Nickel), Pd (Palladium), Pt (Platinum), and Se (Selenium) work functions. (b) PV parameters as a bar chart against different work functions of Cs₂BiAgI₆-DPSCs ranging from 4.65 eV to 5.9 eV.

The work function of the back metal contact is a critical parameter for increasing the PCE of the Cs₂BiAgI₆-DPSC. The value of ϕ_BC directly affects the presence of an appropriate built-in electric field in the Cs₂BiAgI₆-DPSC, which in turn influences the transport and collection of holes and ultimately determines the V_oc. To investigate the impact of different ϕ_BC values on the performance of the Cs₂BiAgI₆-DPSC, we conducted simulations with ϕ_BC ranging from 4.65 eV to 5.9 eV. The work functions of different materials for back electrodes are as follows: Cu (4.65 eV), Fe (4.81 eV), C (5 eV), Au (5.1 eV), W (5.22 eV), Ni (5.5 eV), Pd (5.6 eV), Pt (5.7 eV), and Se (5.9 eV).⁹⁶Fig. 15(a) illustrates that at low ϕ_BC values, the J–V curves display an S-shaped pattern because of the presence of elevated Schottky barriers at the interface between the HTL and back metal contact, leading to an impediment in hole transfer and a decrease in the FF.^97,98 Copper (4.65 eV) and iron (4.81 eV) showed this phenomenon in our study. Fig. 15(b) presents the PV parameters of the device ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V₂O₅/BC for different ϕ_BC values ranging from 4.65 eV to 5.9 eV. The highest efficiency of 27.20% was reached with Ni, Pd, Pt, or Se as the back metal contact, as depicted in Fig. 15(b).

As the work function of the back electrode decreases, the Schottky barrier gradually increases, leading to a higher energy requirement for holes to pass through the barrier. This hinders the transport of holes to the back metal contact,⁹⁴ which ultimately impedes the effective collection of holes and causes a gentle decrease in V_oc, thereby reducing efficiency. By utilizing a back electrode with a high work function, the Fermi level is lowered, facilitating the formation of improved Ohmic contacts.⁹⁹ Therefore, selecting the appropriate back contact material is crucial for achieving high efficiency in Cs₂BiAgI₆-DPSCs. It is essential to ensure that the V₂O₅-HTL and back contact form an ohmic contact at the interface to prevent the formation of a high Schottky barrier. Incorporating a low-cost metal, such as Se, as the back contact, as shown in Fig. 15(b), can significantly enhance the efficiency of Cs₂BiAgI₆-DPSCs.

3.13. Influence of working temperature on Cs₂BiAgI₆-DPSC performance

When Cs₂BiAgI₆-DPSCs are exposed to high temperatures, the performance parameters are affected by the temperature.¹⁰⁰ An increase in temperature results in a decrease in V_oc and a slight increase in J_sc of Cs₂BiAgI₆-DPSC.^101–103

Cs₂BiAgI₆-DPSCs are typically used in outdoor settings where temperatures exceed 300 K,¹⁰⁴ the outdoor temperature directly affects the performance of Cs₂BiAgI₆-DPSCs. Investigating temperature's impact on Cs₂BiAgI₆-DPSC, T ranges from 300 K to 500 K. As temperature increases, V_oc decreases, lowering Cs₂BiAgI₆-DPSC PCE.¹⁰⁵ As temperature increases, the defect density in the device increases, and mobility decreases, hurting Cs₂BiAgI₆-DPSC performance.¹⁰⁶ Changes in temperature greatly affect the performance of Cs₂BiAgI₆-DPSC through its impact on V_oc. The relationship between V_oc and T can be expressed as follows:¹⁰⁵

Eqn (16) reveals an inverse relationship between V_oc and T. An increase in T leads to a decrease in V_oc and an increase in the diode reverse saturation current.¹⁰⁴

The results from Fig. 16 indicate that the performance of the Cs₂BiAgI₆-DPSCs significantly decreases as the operating temperature increases. This is because R_S increases with temperature, which reduces diffusion length and increases recombination rate. Consequently, the FF and efficiency of the Cs₂BiAgI₆-DPSC decreases.^106,107

Fig. 16 PV output parameters: (a) V_oc, (b) J_sc, (c) FF, and (d) PCE for Cs₂BiAgI₆-DPSC against temperature.

In Fig. 16a–d, the PV output parameters are plotted against various temperatures, which vary from 300 K to 500 K in 11 steps. According to Fig. 16(a), V_oc diminishes from 1.3221 V to 1.0501 V. In Fig. 16(b), J_sc increases from 23.84011 mA cm⁻² at 300 K to 23.85634 mA cm⁻² at 500 K. As depicted in Fig. 16(c), FF decreases from 86.28% to 76.25%. In Fig. 16(d), PCE decreases from 27.20% to 19.10%. At 300 K, the optimum value is obtained, with the following PV output parameters: V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻², FF of 86.28%, and PCE of 27.20%.

3.14. Comparing the results of ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V₂O₅/Se device before and after optimization

Fig. 17 presents the results of before and after optimization for (a) the J–V curve (Table S7 and Fig. S5, ESI†) and (b) the EQE curve of ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V₂O₅/Se device. Regarding the EQE curve, prior to optimization, it initiates at 27.806% (300 nm), increases to 99.765% (360 nm), and then remains relatively constant until 550 nm (99.416%) before decreasing to zero at 780 nm. Following optimization, the EQE curve commences at 27.789% (300 nm), increases to 99.849% (355 nm), reaches its peak at 100% (377 nm), maintains a nearly constant value until 550 nm (99.662%), and eventually decreases to zero at 780 nm, as illustrated in Fig. 17(b). Table 6 illustrates the comparison of the results of the J–V curve before and after optimization for ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V₂O₅/Se device.

Fig. 17 Comparing the results of before and after optimization for ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V₂O₅/Se device: (a) J–V curve and (b) EQE curve.

Table 6 Comparing the results of the J–V curve before and after optimization for ITO/ZnO/AZO/Cs₂BiAgI₆ (650 nm)/V2O5/Se device

Parameter	V oc (V)	J sc (mA cm^−2)	FF (%)	PCE (%)
Before optimization	1.1005	23.78	82.78	21.67
After optimization	1.3221	23.84	86.28	27.20

---

3.15. Comparing our investigation results with those of previous studies

The electron transport layer plays a vital role in facilitating the movement of charge carriers within DPSCs to achieve the Shockley–Queisser limit. Employing a bilayer electron transport layer (ZnO/AZO) demonstrates superior charge transfer capabilities and enhanced charge collection, leading to a decrease in trap-assisted recombination at the interface. Comprehensive findings suggest that the utilization of a bilayer electron transport layer presents an efficient approach to improving the interface and fabricating high-performing planar double perovskite solar cells. Table 7 provides a comparison of the PV output parameters for different double perovskite solar cells and our results.

Table 7 Comparison of PV output parameters for different double perovskite solar cells

Device configuration	V oc (V)	J sc (mA cm^−2)	FF	PCE (%)	Procedure	Ref.
FTO/c-TiO2/m-TiO2/Cs2AgBiBr6/Spiro-OMeTAD/Au	0.98	3.93	0.63	2.43	Experiment	108
ITO/c-TiO2/Cs2AgBiBr6/Spiro-OMeTAD/Au	1.06	1.55	0.74	1.22	Experiment	109
ITO/SnO2/Cs2AgBiBr6/P3HT/Au	1.04	1.78	0.78	1.44	Experiment	110
FTO/c-TiO2/Cs2AgBiBr6/P3HT/Au	1.12	1.79	0.68	1.37	Experiment	111
ITO/c-TiO2/Cs2AgBiBr6/Spiro-OMeTAD/MoO3/Au	1.05	2.06	0.65	1.41	Experiment	112
ITO/Cu–NiO/Cs2AgBiBr6/C60/BCP/Ag	1.01	3.19	0.69	2.23	Experiment	113
FTO/c-TiO2/C–Chl m-TiO2/Cs2AgBiBr6/Spiro-OMeTAD/Ag	1.04	4.09	0.73	3.11	Experiment	114
FTO/c-TiO2/m-TiO2/D149/Cs2AgBiBr6-Ti3C2Tx/Spiro-OMeTAD/Ag	0.722	8.85	0.701	4.47	Experiment	115
ITO/SnO2/Cs2AgBiBr6/Zn–Chl/Ag	0.99	3.83	0.736	2.79	Experiment	116
FTO/c-TiO2/m-TiO2/Cs2AgSb0.25Bi0.75Br6/Spiro-OMeTAD/Au	0.64	1.03	0.38	0.25	Experiment	117
ITO/ZnO/AZO/Cs2BiAgI6 (650 nm)/V2O5/Se	1.322	23.84	0.863	27.20	This work

---

4. Conclusion

Toxic metals in perovskite solar cells hinder their commercialization. Double halide perovskites have garnered significant interest owing to their reduced toxicity, adaptable bandgap, structural versatility, and enhanced stability when compared to conventional lead halide perovskites. This study focused on assessing the performance of lead-free Cs₂BiAgI₆-double perovskite solar cells (DPSCs) using a one-dimensional solar cell capacitance simulator (SCAPS-1D) with bilayer ZnO/AZO ETL and ZnO ETL, in addition to various HTLs for the first time. The HTLs selected for evaluation included CBTS, Cu₂O, CuAlO₂, CZTS, CuSCN, spiro-OMeTAD, MoO₃, and V₂O₅. Several parameters, such as energy band alignment, recombination and generation rates, absorber thickness, defect and doping densities for all layers, energy levels of ETLs and HTL, interfacial defect densities, back metal contact, and operating temperature, were investigated to enhance the efficiency of the DPSC. Simulation works on the most effective cells were carried out to gain a comprehensive understanding of the electrical characteristics of the device, which demonstrated that these interfacial bilayers significantly enhanced the photovoltaic properties and overall performance of the device. This research also aimed to boost the efficiency and deepen our comprehension of electron transport mechanisms in Cs₂BiAgI₆-DPSCs. Results indicated that V₂O₅ and ZnO/AZO were the most appropriate materials for HTL and ETL, respectively, among the various options considered. Consequently, the required DPSC was opted as ITO/ZnO/AZO/Cs₂BiAgI₆/V₂O₅/Au. To achieve high performance in planar DPSCs, optimizing the extraction and recombination of electron–hole pairs at the ETL/perovskite interface is crucial. The main concept involved enhancing the ZnO/double perovskite interface properties by introducing a 10 nm ultra-thin layer (UTL) of AZO, which served as a passivation layer. The ZnO/AZO bilayer structure offered benefits, such as effective electron extraction and reduced interfacial recombination owing to its improved energy level alignment and defect passivation. The Cs₂BiAgI₆ absorber layer had a thickness of 650 nm, with a defect density of 1 × 10¹³ cm⁻³ and an acceptor density of 1 × 10¹⁶ cm⁻³. The V₂O₅ layer had a defect density of 1 × 10¹⁷ cm⁻³ and its VBO₁ was 0.0 eV. The AZO layer had a defect density of 1 × 10¹⁶ cm⁻³ and its CBO₂ was 0.0 eV. The defect density of ZnO (N_t,ZnO) was 1 × 10¹⁶ cm⁻³. The interface defect density achieved for the V₂O₅/Cs₂BiAgI₆ side was 1 × 10¹⁰ cm⁻² and for the Cs₂BiAgI₆/AZO side, it was 1 × 10¹⁰ cm⁻². For the AZO/ZnO side interface, the defect density was 1 × 10¹³ cm⁻². Selenium (Se) was chosen as the back metal contact and the temperature was set at 300 K. Upon fine-tuning these factors, the efficiency of the ZnO/AZO bilayer ETL system reached 27.20%, together with a V_oc of 1.3221 V, J_sc of 23.84 mA cm⁻² and an FF of 86.28%. Thus, this study introduces a direct and promising method for producing photovoltaic devices, especially for various double perovskite types, featuring advantageous charge transport layers and recombination characteristics. Furthermore, these results offer a theoretical framework for enhancing the efficiency of Cs₂BiAgI₆-based photovoltaic solar cells (DPSCs), promoting the widespread use of environmentally friendly and durable perovskites.

Author contributions

Aminreza Mohandes: conceptualization, formal analysis, project administration, methodology, software, validation, data curation, writing original draft. Mahmood Moradi: supervision, conceptualization, writing, review and editing.

Data availability

All data that support the findings of this research are included in the article and its ESI.†

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. The authors express their gratitude to Professor Marc Burgeleman from the Gent University, Belgium for developing the SCAPS-1D program and permitting its use.

References

1. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
2. Y. Chen, L. Zhang, Y. Zhang, H. Gao and H. Yan, RSC Adv., 2018, 8, 10489–10508.
3. C. O. R. Quiroz, Y. Shen, M. Salvador, K. Forberich, N. Schrenker, G. D. Spyropoulos, T. Heumüller, B. Wilkinson, T. Kirchartz and E. Spiecker, J. Mater. Chem. A, 2018, 6, 3583–3592.
4. K. Yoshikawa, H. Kawasaki, W. Yoshida, T. Irie, K. Konishi, K. Nakano, T. Uto, D. Adachi, M. Kanematsu, H. Uzu and K. Yamamoto, Nat. Energy, 2017, 2, 17032.
5. N. Kumar, J. Rani and R. Kurchania, Sol. Energy, 2021, 221, 197–205.
6. M. Samiul Islam, K. Sobayel, A. Al-Kahtani, M. A. Islam, G. Muhammad, N. Amin, M. Shahiduzzaman and M. Akhtaruzzaman, Nanomaterials, 2021, 11, 1218.
7. S. M. Majeed, D. S. Ahmed and M. K. A. Mohammed, Org. Electron., 2021, 99, 106310.
8. M. K. A. Mohammed, S. Singh, A. K. Al-Mousoi, R. Pandey, J. Madan, D. Dastan and G. Ravi, RSC Adv., 2022, 12, 32611–32618.
9. Q. Wang, X. Zhang, Z. Jin, J. Zhang, Z. Gao, Y. Li and S. F. Liu, ACS Energy Lett., 2017, 2, 1479–1486.
10. J. Jiang, Z. Jin, J. Lei, Q. Wang, X. Zhang, J. Zhang, F. Gao and S. F. Liu, J. Mater. Chem. A, 2017, 5, 9514–9522.
11. A. Mathur and V. Maheshwari, J. Mater. Chem. C, 2022, 10, 7485–7493.
12. Y. Raoui, H. Ez-Zahraouy, S. Ahmad and S. Kazim, Sustainable Energy Fuels, 2021, 5, 219–229.
13. A. Babayigit, A. Ethirajan, M. Muller and B. Conings, Nat. Mater., 2016, 15, 247–251.
14. I. R. Benmessaoud, A. L. Mahul-Mellier, E. Horváth, B. Maco, M. Spina, H. A. Lashuel and L. Forró, Toxicol. Res., 2016, 5, 407–419.
15. A. Babayigit, D. Duy Thanh, A. Ethirajan, J. Manca, M. Muller, H.-G. Boyen and B. Conings, Sci. Rep., 2016, 6, 18721.
16. O. A. Lozhkina, A. A. Murashkina, M. S. Elizarov, V. V. Shilovskikh, A. A. Zolotarev, Y. V. Kapitonov, R. Kevorkyants, A. V. Emeline and T. Miyasaka, Chem. Phys. Lett., 2018, 694, 18–22.
17. R. E. Brandt, R. C. Kurchin, R. L. Z. Hoye, J. R. Poindexter, M. W. B. Wilson, S. Sulekar, F. Lenahan, P. X. T. Yen, V. Stevanovic and J. C. Nino, J. Phys. Chem. Lett., 2015, 6, 4297–4302.
18. H. Zhou, Q. Chen, G. Li, S. Luo, T. Song, H.-S. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542–546.
19. F. Giustino and H. J. Snaith, ACS Energy Lett., 2016, 1, 1233–1240.
20. F. Jiang, D. Yang, Y. Jiang, T. Liu, X. Zhao, Y. Ming, B. Luo, F. Qin, J. Fan and H. Han, J. Am. Chem. Soc., 2018, 140, 1019–1027.
21. E. T. McClure, M. R. Ball, W. Windl and P. M. Woodward, Chem. Mater., 2016, 28, 1348–1354.
22. P.-K. Kung, M.-H. Li, P.-Y. Lin, J.-Y. Jhang, M. Pantaler, D. C. Lupascu, G. Grancini and P. Chen, Sol. RRL, 2020, 4, 1900306.
23. K. Dave, M. H. Fang, Z. Bao, H. T. Fu and R. S. Liu, Chem. – Asian J., 2020, 15, 242–252.
24. C. Frangville, M. Rutkevičius, A. P. Richter, O. D. Velev, S. D. Stoyanov and V. N. Paunov, ChemPhysChem, 2012, 13, 4235–4243.
25. P. K. Nayak, D. T. Moore, B. Wenger, S. Nayak, A. A. Haghighirad, A. Fineberg, N. K. Noel, O. G. Reid, G. Rumbles and P. Kukura, Nat. Commun., 2016, 7, 13303.
26. F. Marchetti, E. Moroni, A. Pandini and G. Colombo, J. Phys. Chem. Lett., 2021, 12, 3724–3732.
27. A. H. Slavney, T. Hu, A. M. Lindenberg and H. I. Karunadasa, J. Am. Chem. Soc., 2016, 138, 2138–2141.
28. G. Volonakis, M. R. Filip, A. A. Haghighirad, N. Sakai, B. Wenger, H. J. Snaith and F. Giustino, J. Phys. Chem. Lett., 2016, 7, 1254–1259.
29. F. Igbari, Z. Wang and L. Liao, Adv. Energy Mater., 2019, 9, 1803150.
30. X.-G. Zhao, D. Yang, J.-C. Ren, Y. Sun, Z. Xiao and L. Zhang, Joule, 2018, 2, 1662–1673.
31. H. Absike, N. Baaalla, R. Lamouri, H. Labrim and H. Ez-zahraouy, Int. J. Energy Res., 2022, 46, 11053–11064.
32. H. Wu, A. Erbing, M. B. Johansson, J. Wang, C. Kamal, M. Odelius and E. M. J. Johansson, ChemSusChem, 2021, 14, 4507–4515.
33. M. I. Kholil, M. T. H. Bhuiyan, M. A. Rahman, M. S. Ali and M. Aftabuzzaman, RSC Adv., 2021, 11, 2405–2414.
34. M. Roknuzzaman, K. Ostrikov, H. Wang, A. Du and T. Tesfamichael, Sci. Rep., 2017, 7, 14025.
35. J. U. Rehman, M. Usman, S. Amjid, M. Sagir, M. B. Tahir, A. Hussain, I. Alam, R. Nazir, H. Alrobei and S. Ullah, Comput. Theor. Chem., 2022, 1209, 113624.
36. M. A. Hadi, M. N. Islam and J. Podder, RSC Adv., 2022, 12, 15461–15469.
37. A. Mohandes and M. Moradi, Phys. Scr., 2024, 99, 55951.
38. J. Dong, Y. Zhao, J. Shi, H. Wei, J. Xiao, X. Xu, J. Luo, J. Xu, D. Li and Y. Luo, Chem. Commun., 2014, 50, 13381–13384.
39. Z.-L. Tseng, C.-H. Chiang, S.-H. Chang and C.-G. Wu, Nano Energy, 2016, 28, 311–318.
40. S.-H. Wu, M.-Y. Lin, S.-H. Chang, W.-C. Tu, C.-W. Chu and Y.-C. Chang, J. Phys. Chem. C, 2018, 122, 236–244.
41. C. Bhoomanee, P. Ruankhama, S. Choopun, A. Prathan and D. Wongratanaphisan, Mater. Today Proc., 2019, 17, 1259–1267.
42. F. Khan and J. H. Kim, Solid State Electrochem., 2020, 164, 107714.
43. A. Guchhait, G. K. Dalapati, P. Sonar, S. Gopalan, F. Bin Suhaimi, T. Das, V. G. V. Dutt, N. Mishra, C. Mahata and A. Kumar, J. Electron. Mater., 2021, 50, 5732–5739.
44. J. Kruszyńska, J. Ostapko, V. Ozkaya, B. Surucu, O. Szawcow, K. Nikiforow, M. Hołdyński, M. M. Tavakoli, P. Yadav and M. Kot, Adv. Mater. Interfaces, 2022, 9, 2200575.
45. S. A. Moiz, A. N. M. Alahmadi and A. J. Aljohani, IEEE Access, 2021, 9, 54254–54263.
46. M. Burgelman, P. Nollet and S. Degrave, Thin Solid Films, 2000, 361–362, 527–532.
47. M. K. Hossain, A. A. Arnab, R. C. Das, K. M. Hossain, M. H. K. Rubel, M. F. Rahman, H. Bencherif, M. E. Emetere, M. K. A. Mohammed and R. Pandey, RSC Adv., 2022, 12, 35002–35025.
48. A. Mohandes, M. Moradi and M. Kanani, Opt. Quantum Electron., 2023, 55, 533.
49. J. C. Z. Medina, E. R. Andrés, C. M. Ruíz, E. C. Espinosa, L. T. Yarce, R. G. Isasmendi, R. R. Trujillo, G. G. Salgado, A. C. Solis and F. G. N. Caballero, Heliyon, 2023, 9(3), e14547.
50. K. T. Arockiya Dass, M. K. Hossain and L. Marasamy, Sci. Rep., 2024, 14, 1473.
51. L. R. Karna, R. Upadhyay and A. Ghosh, Sci. Rep., 2023, 13, 15212.
52. A. Mohandes, M. Moradi and H. Nadgaran, Opt. Quantum Electron., 2021, 53, 1–22.
53. M. Noman, I. N. Khan, A. Qamar, K. AlSnaie and H. M. H. Farh, ACS Omega, 2024, 9(10), 11398–11417.
54. A. Mohandes, M. Moradi and H. Nadgaran, Eur. Phys. J. Plus, 2021, 136, 113.
55. P. Rafieipour, A. Mohandes, M. Moaddeli and M. Kanani, Integrating transfer matrix method into SCAPS-1D for addressing optical losses and per-layer optical properties in perovskite/Silicon tandem solar cells, arXiv, 2023, preprint, arXiv:2308.01132 DOI:10.48550/arXiv.2308.01132.
56. S. Bhattarai, M. K. Hossain, G. F. I. Toki, R. Pandey, J. Madan, D. P. Samajdar, S. Ezzine, L. Ben Farhat, M. Z. Ansari and S. H. Ahammad, New J. Chem., 2023, 47, 17908–17922.
57. A. Kuddus, M. F. Rahman, S. Ahmmed, J. Hossain and A. B. M. Ismail, Superlattices Microstruct., 2019, 132, 106168.
58. Q. Guo, C. Wang, J. Li, Y. Bai, F. Wang, L. Liu, B. Zhang, T. Hayat, A. Alsaedi and Z. Tan, Phys. Chem. Chem. Phys., 2018, 20, 21746–21754.
59. X. Yao, J. Qi, W. Xu, X. Jiang, X. Gong and Y. Cao, ACS Omega, 2018, 3, 1117–1125.
60. Z. Liu, T. He, H. Wang, X. Song, H. Liu, J. Yang, K. Liu and H. Ma, RSC Adv., 2017, 7, 18456–18465.
61. H. Park, S. Jeong, E. Kim, S. Shin and H. Shin, ACS Appl. Mater. Interfaces, 2022, 14, 42007–42017.
62. M. F. Hossain, A. Ghosh, M. A. Al Mamun, A. A. Miazee, H. Al-lohedan, R. J. Ramalingam, M. F. I. Buian, S. R. I. Karim, M. Y. Ali and M. Sundararajan, Opt. Commun., 2024, 559, 130410.
63. J. Li, J. Qin, X. Liu, M. Ren, J. Tong, N. Zheng, W. Chen and Y. Xia, Sol. Energy, 2020, 211, 1102–1109.
64. J. Tian, Q. Xue, Q. Yao, N. Li, C. J. Brabec and H. Yip, Adv. Energy Mater., 2020, 10, 2000183.
65. J. Zhang, G. Hodes, Z. Jin and S. Liu, Angew. Chem., Int. Ed., 2019, 58, 15596–15618.
66. Q. Tai, K.-C. Tang and F. Yan, Energy Environ. Sci., 2019, 12, 2375–2405.
67. W. L. Lachore, D. M. Andoshe, M. A. Mekonnen, N. A. Tegegne and F. G. Hone, Highly crystalline metal oxide semiconductor electron transport bilayers to improve the interface between the electron transport and active layer in planar perovskite solar cells, Research Square, 2022, pp. 1–24 DOI:10.21203/rs.3.rs-2136665/v1.
68. P. Schulz, L. L. Whittaker-Brooks, B. A. MacLeod, D. C. Olson, Y.-L. Loo and A. Kahn, Adv. Mater. Interfaces, 2015, 2(7), 1400532.
69. M. H. K. Rubel, M. A. Hossain, M. K. Hossain, K. M. Hossain, A. A. Khatun, M. M. Rahaman, M. F. Rahman, M. M. Hossain and J. Hossain, Results Phys., 2022, 42, 105977.
70. M. Rai, L. H. Wong and L. Etgar, J. Phys. Chem. Lett., 2020, 11, 8189–8194.
71. T. Singh, A. Kulkarni, M. Ikegami and T. Miyasaka, ACS Appl. Mater. Interfaces, 2016, 8, 14542–14547.
72. W. Chang, H. Tian, G. Fang, D. Guo, Z. Wang and K. Zhao, Sol. Energy, 2019, 186, 323–327.
73. P. Zhao, Z. Liu, Z. Lin, D. Chen, J. Su, C. Zhang, J. Zhang, J. Chang and Y. Hao, Sol. Energy, 2018, 169, 11–18.
74. J. Barbé, M. L. Tietze, M. Neophytou, B. Murali, E. Alarousu, A. El Labban, M. Abulikemu, W. Yue, O. F. Mohammed, I. McCulloch, A. Amassian and S. Del Gobbo, ACS Appl. Mater. Interfaces, 2017, 9, 11828–11836.
75. H. Kim, K. G. Lim and T. W. Lee, Energy Environ. Sci., 2016, 9, 12–30.
76. D. Chen, Y. Wang, Z. Lin, J. Huang, X. Chen, D. Pan and F. Huang, Cryst. Growth Des., 2010, 10, 2057–2060.
77. Y. Chen, H. Shen and P. P. Altermatt, Sol. Energy Mater. Sol. Cells, 2014, 120, 356–362.
78. Q. Wang, Y. Shao, H. Xie, L. Lyu, X. Liu, Y. Gao and J. Huang, Appl. Phys. Lett., 2014, 105, 163508.
79. M. Abdi-Jalebi, M. I. Dar, A. Sadhanala, S. P. Senanayak, M. Grätzel and R. H. Friend, JoVE, 2017, e55307.
80. M. S. Jamal, S. A. Shahahmadi, M. A. A. Wadi, P. Chelvanathan, N. Asim, H. Misran, M. I. Hossain, N. Amin, K. Sopian and M. Akhtaruzzaman, Optik, 2019, 182, 1204–1210.
81. S. Bouazizi, W. Tlili, A. Bouich, B. M. Soucase and A. Omri, Mater. Res. Express, 2022, 9, 96402.
82. K. S. Nithya and K. S. Sudheer, Optik, 2020, 217, 164790.
83. T. Minemoto and M. Murata, Sol. Energy Mater. Sol. Cells, 2015, 133, 8–14.
84. K. D. A. Kumar, S. Valanarasu, S. R. Rosario, V. Ganesh, M. Shkir, C. J. Sreelatha and S. AlFaify, Solid State Sci., 2018, 78, 58–68.
85. D. Thakur, S.-E. Chiang, M.-H. Yang, J.-S. Wang and S. H. Chang, Sol. Energy Mater. Sol. Cells, 2022, 235, 111454.
86. D. K. Sarkar, A. K. M. Hasan, M. Mottakin, V. Selvanathan, K. Sobayel, M. A. Islam, G. Muhammad, M. Aminuzzaman, M. Shahiduzzaman and K. Sopian, Sol. Energy, 2022, 243, 215–224.
87. G. Pindolia, S. M. Shinde and P. K. Jha, Sol. Energy, 2022, 236, 802–821.
88. F. Izadi, A. Ghobadi, A. Gharaati, M. Minbashi and A. Hajjiah, Optik, 2021, 227, 166061.
89. A. S. Chouhan, N. P. Jasti and S. Avasthi, Mater. Lett., 2018, 221, 150–153.
90. S. Zahedi-Azad, M. Maiberg and R. Scheer, Prog. Photovoltaics Res. Appl., 2020, 28, 1146–1157.
91. M. Burgelman, Mott-Schottky analysis from C-V simulations, and Admittance Analysis from C-f simulations in SCAPS, Dept. of Electronics and Information Technology (ELIS), University of Gent, Belgium, 2017, 2–4.
92. T. Walter, R. Herberholz, C. Müller and H. W. Schock, J. Appl. Phys., 1996, 80, 4411–4420.
93. L. K. Ono, S. Liu and Y. Qi, Angew. Chem., Int. Ed., 2020, 59, 6676–6698.
94. L. Lin, L. Jiang, P. Li, B. Fan and Y. Qiu, J. Phys. Chem. Solids, 2019, 124, 205–211.
95. D. Qiu and E. K. Kim, Sci. Rep., 2015, 5, 13743.
96. K. D. Jayan and V. Sebastian, Sol. Energy, 2021, 217, 40–48.
97. L. Lin, L. Jiang, Y. Qiu and B. Fan, J. Phys. Chem. Solids, 2018, 122, 19–24.
98. J. Gao, J. M. Luther, O. E. Semonin, R. J. Ellingson, A. J. Nozik and M. C. Beard, Nano Lett., 2011, 11, 1002–1008.
99. J. Shewchun, J. Dubow, C. W. Wilmsen, R. Singh, D. Burk and J. F. Wager, J. Appl. Phys., 1979, 50, 2832–2839.
100. S. H. Liu, E. J. Simburger, J. Matsumoto, A. Garcia III, J. Ross and J. Nocerino, Prog. Photovoltaics Res. Appl., 2005, 13, 149–156.
101. M. Fathi, M. Abderrezek, F. Djahli and M. Ayad, Energy Procedia, 2015, 74, 1410–1417.
102. H. Heriche, Z. Rouabah and N. Bouarissa, Int. J. Hydrogen Energy, 2017, 42, 9524–9532.
103. M. Theelen, A. Liakopoulou, V. Hans, F. Daume, H. Steijvers, N. Barreau, Z. Vroon and M. Zeman, J. Renewable Sustainable Energy, 2017, 9, 21205.
104. E. Karimi and S. M. B. Ghorashi, J. Nanophotonics, 2017, 11, 32510.
105. N. Devi, K. A. Parrey, A. Aziz and S. Datta, J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.:Mater., Process., Meas., Phenom., 2018, 36, 04G105.
106. F. Behrouznejad, S. Shahbazi, N. Taghavinia, H.-P. Wu and E. W.-G. Diau, J. Mater. Chem. A, 2016, 4, 13488–13498.
107. S. R. Raga, E. M. Barea and F. Fabregat-Santiago, J. Phys. Chem. Lett., 2012, 3, 1629–1634.
108. E. Greul, M. L. Petrus, A. Binek, P. Docampo and T. Bein, J. Mater. Chem. A, 2017, 5, 19972–19981.
109. W. Ning, F. Wang, B. Wu, J. Lu, Z. Yan, X. Liu, Y. Tao, J. Liu, W. Huang and M. Fahlman, Adv. Mater., 2018, 30, 1706246.
110. C. Wu, Q. Zhang, Y. Liu, W. Luo, X. Guo, Z. Huang, H. Ting, W. Sun, X. Zhong and S. Wei, Adv. Sci., 2018, 5, 1700759.
111. M. Wang, P. Zeng, S. Bai, J. Gu, F. Li, Z. Yang and M. Liu, Sol. RRL, 2018, 2, 1800217.
112. F. Igbari, R. Wang, Z.-K. Wang, X.-J. Ma, Q. Wang, K.-L. Wang, Y. Zhang, L.-S. Liao and Y. Yang, Nano Lett., 2019, 19, 2066–2073.
113. W. Gao, C. Ran, J. Xi, B. Jiao, W. Zhang, M. Wu, X. Hou and Z. Wu, ChemPhysChem, 2018, 19, 1696–1700.
114. B. Wang, N. Li, L. Yang, C. Dall’Agnese, A. K. Jena, S. Sasaki, T. Miyasaka, H. Tamiaki and X.-F. Wang, J. Am. Chem. Soc., 2021, 143, 2207–2211.
115. L. Yang, P. Hou, B. Wang, C. Dall’Agnese, Y. Dall’Agnese, G. Chen, Y. Gogotsi, X. Meng and X.-F. Wang, Chem. Eng. J., 2022, 446, 136963.
116. B. Wang, L. Yang, C. Dall’Agnese, A. K. Jena, S. Sasaki, T. Miyasaka, H. Tamiaki and X.-F. Wang, Sol. RRL, 2020, 4, 2000166.
117. Y. Liu, L. Zhang, M. Wang, Y. Zhong, M. Huang, Y. Long and H. Zhu, Mater. Today, 2019, 28, 25–30.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ma01280a

---