Numerical investigation of Aloe Vera-mediated green synthesized CuAlO₂ as HTL in Pb-free perovskite solar cells

Abstract

This study explores green-synthesized delafossite CuAlO₂ as a hole transport layer (HTL) in a CH₃NH₃SnI₃-based perovskite solar cells (PSCs) with an FTO/CuAlO₂/CH₃NH₃SnI₃/WO₃/Au structure. The performances of the cell have been theoretically investigated using SCAPS-1D. Interface defects significantly impact cell efficiency, as defect density increases from 10¹⁴ cm⁻³ to 10²⁰ cm⁻³ efficiency reducing from 25.3% to 24.45% in the HTL/perovskite junction and from 25.2% to 17.8% in the perovskite/electron transport layer (ETL) interface. PCBM as buffer layer at perovskite/ETL interface compensates for power conversion efficiency (PCE) losses. Optimizing parameters reveals the delafossite CuAlO₂-based, lead-free perovskite solar cell achieving peak efficiency at 26.74%, with V_OC, J_SC, and FF values of 0.99 V, 33.43 mA cm⁻², and 81.05%, respectively. This research underscores delafossite CuAlO₂ as a promising HTL for eco-friendly, stable CH₃NH₃SnI₃-based perovskite solar cells, emphasizing its potential in enhancing device performance.

KEYWORDS:

• Delafossite
• green synthesis
• HTL
• perovskite
• CuAlO₂
• SCAPS-1D

1. Introduction

Over the last few years, much time and efforts have been given in improving the efficiency and stability of perovskite solar cells (PSCs). PSC is a challenger to the most commercially established silicon solar cell due to its easy processing and fabrication technique, low cost, variation in composition, wide range of spectrum absorption ability [1–6]. Environmental issues such as moisture, oxygen, light, heat, intrinsic molecular dissociation, and ion movement inside perovskites are all important factors that impact the stability of PSCs [7,8]. The main restriction of PSCs before industrialization is device durability. Due to the features such as long-term durability, cost-effectiveness, low-temperature processability, and decreased hysteresis properties, inverted PSCs have been explored extensively. Compared to the record 25.7% power conversion efficiency (PCE) of regular perovskite solar cell structure, inverted PSC declares 22.1% highest PCE [9–12]. Pb toxicity is another major problem; thus, researchers have explored device performance by substituting Pb with Sn, Ge, Cu, Bi, and so on, but the findings are not satisfactory. Among them, tin (Sn) based perovskites are a potential candidate in this area because of their appropriate bandgap, outstanding photovoltaic characteristics, low manufacturing cost, great performance, and environmental friendliness [13–15]. The oxidation of Sn²⁺ to Sn⁴⁺ in tin-based perovskites is a critical challenge that affects the stability and performance of these materials in solar cells. Encapsulation and passivation techniques are vital strategies to counteract this issue and improve the stability of tin perovskite-based devices. Despite several efforts, the maximum efficiency is found at 13.24% for SnX₃-based PSCs and stability at 3800 h, which are both lower than the Shockley–Queasier (SQ) limit [16–19]. There is still plenty of space for advancement in SnX₃-based PSCs.

Material engineering regarding the hole transport material (HTM) in inverted PSCs plays a vital role in cell performance because light penetrates through the HTM before absorbed by the perovskite in this structure. So, optoelectronic properties of the HTM should match with the perovskite layer. Organic HTMs such as Spiro-OMeTAD, PTAA, PEDOT: PSS, P3HT, etc. are commonly used HTMs in perovskite solar cells [20]. Improving the electrical properties of organic hole transport layer (HTL) materials require doping or the addition of chemicals, which causes the perovskite layer to degrade rapidly. To resolve these restrictions, scientists have used the most robust, inexpensive, and abundant inorganic materials as appropriate HTMs for PSCs. Delafossite oxides (ABO₂ where, “A” monovalent, “B” trivalent cation) are attractive HTL for perovskite solar cells due to their wide band gap, excellent energy band alignment relative to the perovskite layer, and ease of manufacturing [21,22]. For p-i-n type PSC structures, CuAlO₂ has higher optical transparency in the visible band while maintaining excellent electrical conductivity [23].

Femi Igbari et al. investigated the CuAlO₂ hole interfacial layer and obtained a maximum of 14.52% PCE [24]. Achilleas Savva et al. researched CuAlO₂/Cu–O hole-selective contact and obtained 16.3% PCE with negligible hysteresis effect. They also found near 80% transparency for CuAlO₂ in the visible wavelength range. Mona Shasti et al. simulated MAPbI₃-based perovskite solar cells using CuAlO₂ as HTM and obtained 19.82% PCE [25]. Fanliang Bao et al. fabricated PSCs with CuAl(M)O₂ HTL in CsPbBr₃-based perovskite solar cells [26]. They declared that the oxygen movement and structure in CuAl(Sc)O₂ boosted electric charge flow, making solar cells using this material 10.13% efficient, compared to 6.5% with HTL-free device. Shayan Tariq Jan et al. investigated the n-i-p and p-i-n structures for MAGeI₃ PSCs, using CuAlO₂ as the HTL, and found that the p-i-n structure showing a 24.32% of efficiency, while the n-i-p structure provides a 14.82% PCE [27].

The primary novelty of this study lies in the synthesis methodology employed, specifically, the utilization of green Aloe-vera extract to synthesize CuAlO₂. In addition to providing a distinctive, environmentally friendly method, this novel approach also addresses the rising significance of sustainable practices in material synthesis. The p-i-n structure with Pb-free perovskite and inorganic charge transport layers (CTLs) was studied in this work. The objective of this work is to utilize green synthesized CuAlO₂ as HTL in the PSC device structure in the configuration of FTO/CuAlO₂/CH₃NH₃SnI₃/WO₃/Au. The materials used in the proposed device are environmentally benign, abundant, and inexpensive. Delafossite CuAlO₂ as HTL and WO₃ as ETL [28–31] are both promising materials for dramatically improving device performance; specifically, inorganic CTLs enhance stability. By the sol–gel method, synthesized CuAlO₂ has been investigated by XRD and FTIR studies, which verified the material properties. The optical properties of the prepared CuAlO₂ were also determined. The device performance was then simulated in preparation for actual device fabrication. Variations in Absorber thickness, HTL and ETL thickness, carrier concentration, interfacial defect, series and shunt resistance, Mott–Schottky capacitance studies, and temperature effect have been investigated to optimize the device.

2. Methodology

This study was conducted in two parts: (i) experimental and (ii) numerical simulation. In the experimental part, synthesized CuAlO₂ was analysed using XRD, UV-Vis, and FTIR techniques. In the second part, a perovskite solar cell (p-i-n) has been designed with characterized CuAlO₂ as hole transport material, and its PV properties have been studied by the SCAPS-1D simulation programme.

2.1. Materials

CuAlO₂ was synthesized by a green sol–gel process. Aloe vera-water extract, which acted as a green chelating/complexing agent, was used [32]. Other chemicals Cu(NO₃)₂.3H₂O, Al(NO₃)₃.9H₂O were used as the sources of Cu and Al, respectively. All chemicals were AR grade, collected from Sigma-Aldrich, and used directly without further purification.

2.2. Collection of plant extract

Initially, the collected Aloe Vera stems were washed several times with DI water before being dried in normal conditions and sliced. Each 15 g Aloe Vera slice was placed in a glass beaker with 100 mL DI water and stirred for 1 h on a hot plate at 60°C. After the extraction was finished, the mixture was filtered, and the filtrate was stored in the refrigerator to be used as a chelating or complexing agent during the synthesis process.

2.3. Preparation of CuAlO₂

Millimole equivalents of 0.241 g Cu(NO₃)₂.3H₂O and 0.38 g Al(NO₃)₃.9H₂O were weighed out and mixed into 100 mL Aloe Vera extract separately in two conical flasks while gently shaking to mix up homogeneously. In a 500 mL glass beaker, both coloured mixtures were warmed at 80°C with continuous stirring until the mixture was transferred into a clear solution. The solution was evaporated at 100–120°C until it became a light greenish gelatinous mass. The mass in gel form was dried completely in an oven at 150°C for 12 h. Finally, the dried powder material was calcined for 6 h at 500°C to yield the desired product. The entire synthesis process is depicted schematically in Figure 1.

Figure 1. Flowchart of the synthesis process.

2.4. Material characterization

Phase structure of the synthesized CuAlO₂ was recorded by X-ray diffraction (XRD: Diffractometer; Bruker AXS D8 ADVANCE) equipped with Cu Kα radiation. The Fourier transmittance infrared (FTIR) spectroscopy is equipped with a Bruker Vertex-70 spectrometer, and a Hyperion 3000 IR microscope was used to observe the transmittance spectral response of the sample. To study the optical nature of the material, UV-Vis spectroscopy was carried out by a spectrophotometer (Perkin Elmer Lambda-35).

3. Results and discussion

3.1. XRD study

Figure 2 illustrates the XRD outcomes of CuAlO₂ prepared by the green synthesis route. The diffraction peaks of the film have been recognized as (006), (101), (102), (012), (104), (009), and (108) delafossite structures with the R$\overline{3}$m space cluster. Measured crystallite size, dislocation densities, strain, and lattice parameters are 23.40 nm, 1.60 × 10¹¹ cm⁻³, 4.79 × 10⁻³, and 3.51 Å, respectively. Figure 2(b) represents the elemental arrangement in the lattice of a unit cell. The green colour sphere is an Al atom, which is octahedrally connected with the deep ash-coloured sphere oxygen atoms and reddish sphere Cu ions are stacked inside by the layered of (Al₂O₆)⁻¹ [33].

Figure 2. (a) XRD analysis of prepared CuAlO₂, (b) Crystal structure of CuAlO₂.

3.2. Optical properties

The UV-Vis absorbance profile of the pure phase CuAlO₂ is exposed in Figure 3(a). The direct E_g can be determined by using absorption onset (355 nm) as in the below equation. (1) $E_g=\frac{1240}{\lambda }\text{eV}$(1) The direct optical E_g is assessed from the Tauc plot in Figure 3(b) around 3.5 eV. This E_g is reliable with E_g (3.5 eV) estimated for CuAlO₂ thin film grown by pulsed laser deposition (PLD) method [34].

Figure 3. Optical Properties of CuAlO₂; (a) UV-Visible absorption spectra, (b) E_g estimation from Tauc plot.

3.3. FTIR study

The FTIR study provides important information about metal–oxygen stretching, ion sites, and the presence of functional groups in the sample [35]. The spectra exhibit a broad peak at 604 cm⁻¹, which covers the transmittance ranging frequency wave number from 550 cm⁻¹ to 900 cm⁻¹ (Figure 4). The next three peaks are obtained at 1066, 1242, and 1404 cm⁻¹, respectively. Another two close signals are found at around 2901, and 2987 cm⁻¹ and the last peak appeared at 3665 cm⁻¹. The broad peak at 604 cm⁻¹ is the result of the convolution of several smaller peaks. This peak is assigned to the transmittance peaks of Cu-O, Al–O bond vibrations. Peaks in the 800–900 cm⁻¹ range may be attributed to short Al–O stretching vibrations in distorted AlO₆ octahedral CuAlO₂ [36]. Peak around 1000 cm⁻¹ may be assigned to Al–O-C, or Cu-O–C [37]. The peak appears at 2901 cm⁻¹ due to C−H stretching and peaks at 2987 cm⁻¹ wavenumber or above due to O-H bond stretching, which may have been incorporated from the atmosphere or phytochemicals used in the precursor [38].

Figure 4. Represents the FTIR spectra of the CuAlO₂ synthesized by the phytochemical-assisted sol-gel method.

4. Device modelling for simulation

The 2nd part of the study has been presented in this section. The proposed device, in which synthesized CuAlO₂ was considered as HTL, was studied using SCAPS-1D simulation software. From the experimental part, it was found that the CuAlO₂ exhibited an energy bandgap of 3.5 eV, which is higher than that of some promising inorganic HTMs such as CuO₂ (2.45 eV), CuSCN (3.40 eV), CuCrO₂ (3.30 eV), SrCu₂O₂ (3.30 eV) [39,40]. As a higher energy bandgap of the window layer in PSCs can pass more light to the absorber layer by minimizing the charge leakage, charge separation, and charge transfer [41], experimentally synthesized CuAlO₂ would be considered a suitable HTL, through which solar radiation would travel to the perovskite absorber layer for the proposed p-i-n structured PSC. Table 1 displays the reference data. Regarding holes and electrons, SCAPS-1D adheres to Poisson's calculation and the continuity equation. To get the simulated results realistic enough, SCAPS-1D considers the different types of defects, both in bulk and at the interfaces. Using the defect density values, the simulation provides cell performance parameters that match the experimental performances, confirming the validation of accepting those defect values [42,43]. In this simulation, a single energy distribution with standard lighting of 100 mW/cm² (T = 300 K (K) or 27°C) has been used to provide the illumination (Figure 5).

Figure 5. Schematic devise structure of FTO/CuAlO₂/CH₃NH₃SnI₃/WO₃/Au solar cell with band alignment.

Table 1. Summarized data for simulation.

Parameters	WO3 [22]	CH3NH3SnI3 [22]	CuAlO2 [44]
Thickness (nm)	50	700	50
Eg (eV)	2.92	1.3 [45]	3.5 (Exp.)
χ (eV)	4.59	4.17	2.5
ϵr	5.76	10	60
Nc (cm^−3)	1.96 × 10^19	1 × 10^19	2.2 × 10^18
Nv (cm^−3)	1.96 × 10^19	1 × 10^18	1.8 × 10^19
Electron thermal velocity (cm/s)	1 × 10^7	1 × 10^7	1 × 10^7
Hole thermal velocity (cm/s)	1 × 10^7	1 × 10^7	1 × 10^7
µe (cm^2/Vs)	10	1.6	2
µh (cm^2/Vs)	10	1.6	8.6
NA (cm^−3)	0	3 × 10^17	3.6 × 10^18
ND (cm^−3)	3.68 × 10^19	3 × 10^17	0

4.1. Band diagram

The photocurrent is caused by an excessive amount of minority carriers in the absorber layer. The process of drift and diffusion regulates the transport of holes and electrons from the absorber to the HTL and ETL. A mismatch in the lattice between two interfaces leads to the creation of interface defects, which in turn cause carrier recombination [46]. Cliff and spike are the two types of band structures that may be found at the interface. The principal recombination process takes place at the interface when the band structures reveal cliff-like forms. Figure 6 shows that a spike is observed in both interfaces, leading to a decrease in carrier recombination. The bending in the conduction band is due to the existence of a cliff structure at the interface. Band offset at the pn-junction must be positive (spike-like) in order to attain the optimal device performance [47]. However, negative band offset with suitable band alignment can show better performance. This is attributable to the fact that the activation energy of recombination (E_a) at the absorber layer plays an important role in device performance [48]. When the E_a value is increased, the degree of interface recombination diminishes. This band offset may reduce the amount of charge recombination by preventing injected electrons/holes from reaching the interface [49]. The band gap (E_g) of the absorber layer was considered 1.3 eV. The conduction band offset (CBO) of −0.43 eV (χ_Abs − χ_ETL) [27] creates a cliff type barrier [27,50] that facilitates electron passage from the perovskite layer to the ETL, thereby preventing recombination. Similarly, the valence band offset (VBO) of −0.53 eV ((χ_Abs+ E_g,Abs) − (χ_HTL + E_g,HTL)) [25] also establishes a cliff-type barrier at the valence band edge, enabling holes to travel from the perovskite layer to the HTL, thereby mitigating excessive recombination. This barrier at the interfaces reduces the probability of carriers returning to the region where they originated and decreases the chance of recombination. While a larger band offset might typically be associated with increased carrier recombination, in this case, the alignment of energy levels is more favourable for charge transport, enhancing the overall device performance. This alignment could improve charge extraction, thereby reducing recombination rates and leading to better device efficiency.

Figure 6. Band diagram generated during the simulation process.

4.2. Thickness optimization of HTL, ETL, and the Absorber layer

This section contains mainly simulated investigation on perovskite absorber layer thickness and their corresponding photovoltaic performances deploying WO₃ (ETL) and CuAlO₂ (HTL) as charge transporting layers. Besides, the effects of ETL and HTL thickness on the cell efficiency are also mentioned here. Maintaining other factors constant, CH₃NH₃SnI₃ film thickness is scaled from 100 to 1000 nm, and photovoltaic outputs are shown in Figure 7. The variance of active layer thickness shows a very small influence on V_oc (Figure 7(a))_, but the J_sc (Figure 7(b)) has been intensely influenced since the J_sc reached 35.00 mAcm⁻² from 15.00 mAcm⁻² because of increased thickness. The reduction of Voc outside of 200 nm happens as recombination results from sluggish carrier movement. Positively, the thicker active layer is responsible for carrier generation as well as related recombination governing carrier lifetimes. Therefore, J_sc increases; as additional photons are captivated by the absorber layer when it is thicker. The promotion of photocarriers increased J_sc. An augmented V_oc is apparent with cumulative depth till 300 nm. Outside these widths, V_oc twitches lessen. The downshifting of V_oc in respect to width because of incremental J_o is referred to as dark saturation current that is liable for higher carrier recombination. The reliance of V_oc on J_sc and J_o follows as [51]. (2) $V_{\text{oc}}=\frac{nKT}{q}\text{ln}\left(\frac{J_{sc}}{J_o}+1\right)$(2) where K and T denote Boltzmann constant, temperature (Kelvin), correspondingly. Based on Equation (2), a short J_o is obligatory to attain greater V_oc.

Figure 7. Effect of absorber layer thickness on cell performance, (a) V_oc, J_sc vs thickness; (b), PCE, FF vs thickness; (c), QE (%) as a function of absorber thickness; (d) PCE for ETL, HTL and Without HTL Vs Thickness.

The FF outcomes in an unceasing decrement (Figure 7(b)) with cumulative enhanced thickness. Recombination incidences are considered for the typical model. The advanced PCE might be recognized through robust photon engagement along with long diffusion length (L_d) [52]. Absorber thickness of a few hundred nm is satisfactory to engross supreme incident sunlight. All photogenerated carriers ought to gather, which are generated at minor spaces rather than scattering distance from CH₃NH₃SnI₃/WO₃ and CuAlO₂/CH₃NH₃SnI₃ interfaces. Variation of quantum efficiency (QE) graph for different values of absorber layer thickness is shown in Figure 7(c) representing that the thicker absorber layer confirms higher values of QE. Figure 7(d) shows the influence of ETL and HTL (with or without) thickness on the PCE of the PSC. The width difference of ETL, HTL, and HTL-free PSCs has been examined from 50 to 150 nm. The PCE remains nearly unaltered for thickness variation of ETL and HTL. The PCE decreases very insignificant values, specifically for higher HTL thickness. It is highly essential to have a decent pact among simulations and tentative studies to authenticate the correctness and applied understanding of PSC.

Therefore, it is expected that projected PV factors of the proposed structure would continue inside scope when leading to actual investigational demos. The presented performance of HTL-free PSC has to be examined by opposing HTL width in the same range. A PSC device without HTL displays the maximum 3% PCE at any thickness value smaller than that of a PSC with HTL. Therefore, it could be remarked that ETL and HTL are obligatory in device structure to attain high performance.

4.3. Study of the effect of carrier density of HTL (N_A)

The simulations are performed at different N_A in HTL to observe the doping influence. The N_A concentration varied from 1 × 10¹⁴ cm⁻³ to 1 × 10²⁰ cm⁻³. The effect of N_A of CuAlO₂ on J_sc and V_oc of SC is presented in Figure 8.

Figure 8. Effect of carrier density of HTL (N_A), (a) V_oc, J_sc vs N_A; (b) FF, PCE vs N_A.

The J_sc and V_oc tend to increase at a very insignificant value and decline, as seen in Figure 8(a). The highest and the lowest J_sc are observed at 33.44 mA cm⁻² and 33.42 mA cm⁻², individually. After an N_A value of 1 × 10¹⁷ cm⁻³, J_sc remains to drop by a small amount. V_oc displays an inspiring stance of 0.966 V. Figure 8(b) indicates the gained PCE and FF. At lower N_A (1 × 10¹⁸ cm⁻³) agreements, the highest FF and highest PCE are 78.1% and 25.25%, respectively. PCE increases by N_A mainly due to V_oc. Amplified doping stages at HTL increase the interfacial electric field, requiring distinct excitons after amplified device presentation [53].

4.4. Study of the effect of interface defect density

Recombination at edges can be condensed to raise PSC output [54]. To elude shunting loss, numerous approaches for interface modelling have lately been engaged keenly in attention [55]. Figure 9 embodies the impacts of boundary defects on PSC output, where defect concentrations are diverse from 10¹⁴ cm⁻³ to 10²⁰ cm⁻³. The effect of defect properties at CuAlO₂-MASnI₃ (IL1) and MASnI₃-WO₃ (IL2) were considered to study. Figure 9(a) proves that V_oc and J_sc for IL1 are almost constant till 10¹⁶ cm⁻³ and together decline by the additional upsurge of defect densities towards 10²⁰ cm⁻³. Likewise, Figure 9(b) specifies no quantifiable variations in FF, though some insignificant hills and valleys appear from 10¹⁶ cm⁻³ to 10²⁰ cm⁻³. FF suits the highest at defect density of 10¹⁸ cm⁻³ in IL1. The changes in series resistance due to increasing defect densities in PSC might be the reason for such variation [56]. But PCE remains constant till 10¹⁶ cm⁻³, and PCE subsequently drops from 25.3% to 24.45% for further rises of the defect. The key reason for PCE decrease is the lessening of both V_oc and J_sc after 10¹⁶ cm⁻³ in IL1 to some extent. Figure 9(c) displays that there are no variations in J_sc for defects in IL2. However, V_oc slowly declines nearly by 30% with higher interface defect density. Figure 9(d) displays that FF is intensely enriched with a cumulative defect in IL2 till 10¹⁷ cm⁻³. The association between FF and collective IL2 defect densities could be clarified through ideality factor (n) responsible for the charge carrier transportation pathways from creation to the load. This marvel is related to semiconductor characteristics. The n straight affects the FF of PSC [57]. The rise in FF is produced by lessening defects in IL2 powerfully distressing the PCE from 25.2% to 17.8%.

Figure 9. Effect of interface defect density (N_t) on device performance; (a) V_oc, J_sc vs N_t at IL1; (b) FF, PCE vs N_t at IL1; (c) V_oc, J_sc vs N_t at IL2; (d) FF, PCE vs N_t at IL2.

4.5. Study of the effect of series and shunt resistance on FF and PCE

Series and shunt resistance affect PV device performances significantly. The total resistance between terminals & the cell's right and rear contacts is series resistance. Series resistance is internal resistance, whereas shunt resistance is due to the reverse saturation current of the active junction [58]. A portion of the device's total electrical resistance is contributed by the back contact, HTL, perovskite layer, ETL, and front contact (FTO). In order to investigate how the series and shunt resistance influences the performance of PSCs, we simulate the performance by changing the series resistance from 0 to 5 Ω cm² and shunt resistance from 1000 to 5000 Ω cm². The higher the series resistance, the greater the power losses. The power lost through the PSC can be expressed as Equation (3), (3) $P_{\text{Loss}}=({I_{\text{sc}}}^{\ast }{)}^{\text{2}}\times R_s$(3) Figure 10 shows the impact of series and shunt resistance on the device's performance. Both series and shunt resistance greatly impact FF. FF is decreased from 78% to 64% and PCE is decreased from 25.5% to 20.5% with the increase of series resistance from 0 to 5 Ω cm² (Figure 10(a)). Shunt paths deflect the current away from the load that was allocated for it, and as a result, they have a detrimental effect on the performance of the module, particularly at lower illumination levels. High-quality cells have greater shunt resistance and lower shunt current. Shunt current may cause cell heating in modules, harming the encapsulating layer. Shunt current is usually caused by defects in solar cells or cell interconnection problems. Ideal shunt resistance is considered to be infinite [59]. In this simulation, FF is increased from 76.2% to 77.7% and PCE is increased from 24.6% to 25.13% with the increment of shunt resistance from 1000 to 5000 Ω cm² (Figure 10(b)).

Figure 10. Effect of series and shunt resistance, (a) FF, PCE vs R_s; (b) FF, PCE vs R_sh.

4.6. Study of the effect of temperature on cell performances

The stability of the device depends on the materials used in the cell, their properties (chemical inertness, thermal stability, non-hygroscopicity, etc.), defect-free manufacturing, encapsulation, etc. The absorber material (CH₃NH₃SnI₃) used in our proposed device has outstanding temperature stability and a high defect tolerance limit. There are some issues with the chemical stability of this material because it tends to oxidize in an open environment. Adding anti-oxidant and encapsulation to the device can solve the problem. Chemically inert inorganic metal oxides WO₃ and CuAlO₂ are more stable than most organic ETLs or HTLs. Temperature stability was simulated from 300 to 400 K, and it was found that PCE is lowered only by 4.3% (Figure 11(b)) due to the lower trend of V_oc (Figure 11(a)) and FF (Figure 11(b)) at a higher temperature. Temperature dependent V_oc can be written as Equations (4) and (5) [22]; (4) $\begin{array}{rl}{V}_{\text{oc}}& =\frac{E_g}{q}-\frac{n{k}_{B}T}{q}\text{ln}\left(\frac{I_{ph}}{I_o}\right)\end{array}$(4) (5) $\begin{array}{rl}{E}_g& =E_{g,0}-\frac{a{T}^2}{b+T}\end{array}$(5) where E_g,0 is the band gap at T = 0 K, and a and b are constants. So, the bandgap decreases with increasing temperature, which impacts the V_oc of the cell. With increasing temperature, J_sc remained constant. The series resistance of the solar cell is increased by the thermal scattering of electrons at high temperatures that lower the cell FF. Performance as a whole declined from a PCE of 25.3% to 19.8% with the increment of temperature 300 K to 400 K.

Figure 11. Effect of temperature, (a) V_oc, J_sc vs T; (b) FF, PCE vs T.

4.7. Mott–Schottky capacitance analysis

Temperature-dependent capacitance is seen in Figure 12. Low bias (< 0.4 V) does not affect capacitance, whereas high bias (> 0.4 V) does (Figure 12(a)). Temperature influences the dielectric characteristics of materials, increasing capacitance. Mott–Schottky investigation of temperature can be given in the following Equation (6) [60], (6) $\frac{1}{C^2}=\frac{2}{S^2.\epsilon .q}\cdot \frac{1}{N_A}\cdot (V-V_{bi}-\frac{k_{B}T}{q})$(6) Capacitance, surface area, and space permittivity are denoted by C, S, and ϵ, respectively, while k_B and T have their customary meanings, N_A stands for acceptor volume, V_bi indicates flat band potential, and q stands for an elementary charge. Temperature-dependent C-V and Mot-Schottky graphs are shown in Figure 12(b). At the donor–acceptor interface, where V_bi is independent of V_oc, recombination takes place. When V_bi goes below V_oc, FF decreases. Tangent intersecting at the voltage axis in the Mot-Schottky graphs yields the V_bi. As T rises from 300 K to 400 K, the V_bi decreases from 0.78 V to 0.69 V, indicating greater rates of carrier recombination and an increase in series resistance with temperature. The proposed structure can maintain 84% of its efficiency at higher temperatures because of the high V_bi value.

Figure 12. Mott-Schottky capacitance study; (a) Capacitance vs Voltage, (b) Mott-Schottky plot.

4.8. Study on the effect of the interfacial buffer layer

Figure 13 represents the impact of interfacial layer on device performances. Typically, PSC consists of four interfaces; for regular structure devices, these are TCO/ETL, ETL/perovskite, perovskite/HTL, and HTL/metal, and device for inverted structure; TCO/HTL, HTL/perovskite, perovskite/ETL, and ETL/metal. However, the two interfaces, ETL/perovskite, and HTL/perovskite, are the same for both devices except for the interfaces of TCO and metal electrodes [61]. These two common interfaces play a very effective role in transporting photogenerated charge from the perovskite layer to the electrode through respective CTL. Several factors are involved in the interfaces to reduce device performance, such as state of defect density, band alignment, the position of band edges, etc. [62]. In this study, IL1 (HTL/perovskite) does not have such a remarkable impact on device outcomes, but defects in IL2 (perovskite/ETL) have significant impacts that reduced the PCE by 32%, as indicated in Figure 9(d). The position of the conduction band edge of perovskite is at −4.11 eV, while that of ETL is at −4.38 eV. This energy gap could be one of the reasons for the efficiency losses [63]. To compensate for the losses, an interfacial buffer layer (PCBM with the thickness of 20 nm) has been incorporated into the device structure and simulated, which enabled the regain of the PCE by 5.8%. Figure 13 shows the J-V curve for presenting the impact of the interfacial buffer layer on device performances. The PCE of the modified device is 26.74%, and corresponding V_oc, J_sc, and FF are 0.99 V, 33.43 mA cm⁻², and 81.05%, respectively.

Figure 13. PCBM as an interfacial layer between Perovskite and WO₃ ETL.

4.9. Comparative study

The simulation study has also been extended for some other prominent inorganic and organic HTMs in a similar device. The HTMs are NiO, CuI, Cu₂O, CuSCN, and PEDOT: PSS and their corresponding J-V curves are presented in Figure 14. All devices produce almost the same J_sc (33.15–33.44 mA cm⁻²), exceptional for organic HTM, PEDOT: PSS (31.26 mA cm⁻²) is the lowest one. The NiO-based device gives the lowest PCE at 18.64% due to its low FF (65.64%) compared to others. Cu₂O is a simple, promising semiconducting oxide that provides the 2nd highest PCE 24.23%, with V_oc 0.96 V, J_sc 33.15 mA cm^−2, and FF 76.20%. In this study, among all HTMs, the CuAlO₂-based device shows the highest PCE 25.29% without a buffer layer, and with a buffer layer, it is 26.74%. A comparative results are presented in Table 2, in which the best simulation results from this study considering CuAlO₂ as HTM are compared with the other studies found from different literature.

Figure 14. J-V curve for comparative simulation study of various HTLs.

Table 2. A comparison between this work and a related study published.

Device Structure	Type	Voc (V)	Jsc (mA/cm^2)	FF	PCE (%)	Ref.
ITO/CuAlO2/FASnI3/PCBM/Al	Simul.	1.133	25.200	86.197	24.611	[64]
ITO/CuAlO2: PEDOT: PSS /MAPbI3-xClx/PCBM	Exp.	0.88	21.98	75.00	14.52	[24]
ITO/CuAlO2: CuO /MAPbI3-xClx/ PC70BM: AZO/Al	Exp.	1.07	19.10	79.60	16.30	[65]
ITO/CuAlO2/MAPbI3-xClx/PCBM/Ag	Exp.	0.88	18.58	62.00	10.14	[66]
FTO/CuAlO2/Perovskite/TiO2/Au	Simul.	1.12	23.23	78.40	20.32	[25]
FTO/CuAlO2/CH3NH3SnI3/WO3/Au	Simul.	0.97	33.44	78.12	25.29	This Work

5. Conclusion

In the 1st part of the study, p-type CuAlO₂ has been prepared by the green synthesis process and characterized by XRD, FTIR, and UV-Vis spectroscopy. The spectroscopic data are used to calculate the bandgap (E_g = 3.5 eV) of delafossite CuAlO₂ material by the Tauc plot method. In the 2nd part, a SCAPS-1D simulation study has been carried out on Pb-free perovskite solar cells using synthesized CuAlO₂ as HTL and WO₃ as ETL. In this simulation, experimentally found E_g of CuAlO₂ was used, which matches with the previously used value in different studies, confirming the feasible use of green synthesized CuAlO₂ as a proper HTL in p-i-n structured PSC. Besides, this simulation confirmed the proper matching of CuAlO₂ as HTL in a lead-free perovskite-based solar cell. To the best of our knowledge, use of green synthesized CuAlO₂ as HTL in a eco-friendly PSC development was not analysed theoretically before. Furthermore, the device performances have been optimized with changing layer thickness, internal and interface defects, carrier density, shunt, and series resistance. The effects of temperature and buffer layer have also been evaluated. The performance of CuAlO₂ as HTL in the proposed device has been compared to a similar structured device with different HTLs such as NiO, Cu₂O, CuI, CuCNS, and PEDOT: PSS. The investigation represents that the CuAlO₂-based optimized device exhibits the highest PCE 26.74%, with relative V_oc, J_sc, and FF of 0.99 V, 33.43 mA cm^−2, and 81.05%, respectively. The results of the overall study identified in-depth insights into internal factors that encourage the fabrication of the real device.

Acknowledgements

This work is funded by the Deputyship of Research & Innovation, Ministry of Education in Saudi Arabia, through project number(959/1443). In addition, the authors would like to express their appreciation for the support provided by the Islamic University of Madinah. The authors would also like to express their gratitude for the financial assistance provided by the Dana Impact Research Award from the University Kebangsaan Malaysia (DIP 2021-026). Finally, the authors are grateful to Dr Marc Burgelman of the University of Gent in Belgium for supplying the SCAPS-1D simulation programme.

Disclosure statement

No potential conflict of interest was reported by the author(s).