The structure shown in Figure 1 corresponds to a nip-type PSC ³³ with five layers, which is the configuration studied in this work. The first and last layers are the electrical contacts, while the internal layers are responsible for the device’s energy conversion. For the top contact, fluorine-doped tin oxide (FTO) was used since it is ideal to function as a transparent electrode. Titanium oxide (TiO 2 ) and tin oxide (SnO 2 ) were used for the electron transport layer (ETL) due to their electronic properties and proven use in the scientific literature. ^{17, 34– 38} For the perovskite absorber layer, methylammonium lead iodide (MAPbI 3 ), methylammonium tin iodide (MASnI 3 ) and formamidinium lead iodide (FAPbI 3 ) were used because these materials have the highest reported energy efficiencies and have been extensively studied by the scientific community. ^{10, 25, 34, 39– 41} For the hole transport layer (HTL), Spiro-OMeTAD and copper(I) thiocyanate (CuSCN) were used because configurations with these materials have demonstrated remarkable performance in hole mobility and effective energy alignment. ^{41– 43} Finally, the last layer is generally made of gold (Au) since it has high electrical conductivity. ⁴³ These materials were chosen for their optoelectronic properties, energy compatibility, and the high performance demonstrated in experimental and simulated studies available in the literature. ^{11, 20, 21, 35, 41, 43– 47} To convert solar energy into electrical energy, the PSC absorbs photons from solar radiation in the perovskite layer, generating electron-hole pairs, which are separated and transported by the ETL layer, which extracts the electrons, while the HTL layer collects the holes. The top electrode, usually made of a transparent conductive oxide like FTO, allows the entry of light and the collection of carriers, while the bottom metallic electrode completes the circuit, allowing the flow of external current under load conditions.

SCAPS-1D analyzes the electrical response of a PSC solving a coupled set of differential equations that include the Poisson equation (1), the continuity equations for electrons (2) and holes (3), and the performance metrics equations (4)-(7). The Poisson equation is presented below: ∂ ∂ x ( ε 0 ε r ∂ ψ ∂ x ) = − q ( p − n + N D + − N A − + ρ def q ) (1) where ψ is the electrostatic potential, q is the elementary charge, ε 0 and ε r are the vacuum and the relative permittivity, p and n are hole and electron concentrations, N D + and N A − are charge impurities of donor and acceptor type, and ρ def is the defect density. The continuity equations for electrons (2) and holes (3) are given by: − ∂ J n ∂ x − U n + G = ∂ n ∂ t , (2) − ∂ J p ∂ x − U p + G = ∂ p ∂ t . (3) where J n and J p are the electron and hole current densities, G is the electron-hole generation rate, and U n and U p are the electron and hole recombination rates. To calculate the performance metrics Voc (4), Jsc (5), FF (6), and PCE (7), SCAPS-1D uses the equations presented below: J SC = ∫ 0 λ max q ϕ ( λ ) EQE ( λ ) dλ (4) V OC = kT q ln ( J SC J 0 + 1 ) , (5) FF = V mp ⋅ J mp V OC ⋅ J SC , (6) PCE = V OC ⋅ J SC ⋅ FF P in (7) where λ is the wavelength, ϕ ( λ ) is the incident solar spectrum and EQE ( λ ) is the external quantum efficiency, k is the Boltzmann constant, T is the absolute temperature, J 0 is the reverse saturation current, V mp and J mp are the voltage and current at the point of maximum power, and P in is the incident power.

The data set was generated through numerical simulation using the freely available software SCAPS-1D version 3.3.09 and the evaluation of the electrical performance of the device was carried out by studying J-V curves under the standard working conditions of AM1.5G illumination (1000 W/m 2 ), temperature of 300 K and maximum applied voltage of 1.2 V. ⁴⁸ From these curves, the main electrical parameters that characterize the performance of the system were determined, including the Voc, Jsc, FF and PCE. These parameters were obtained directly from the software after simulating the optoelectronic behavior of the device, allowing a precise evaluation of the expected performance and allowing comparative analysis in terms of efficiency, stability, and robustness against variations in the materials, properties or thickness of the studied layers.

We selected the varied parameters because they directly affect the device’s optical and electrical behaviour. The thickness of the perovskite layer (T_PVK) must be adjusted to absorb the largest amount of photons, maximizing Jsc without exceeding the carrier diffusion length, since excessive thicknesses increase recombination losses and degrade Voc and FF. ⁴⁹ Similarly, the thicknesses of the transport layers (T_ETL and T_HTL) must be optimized to ensure efficient electron and hole transport with low recombination and series resistance. ^{50, 51} Additionally, the properties of perovskite have a direct influence on the performance of the cell; the bandgap (EG_PVK) establishes the balance between the current density and voltage, based on the Shockley-Queisser limit ⁵² ; the dielectric permittivity (ER_PVK) influences exciton dissociation ^{53, 54} ; the acceptor density (NA_PVK) models the internal electric field, essential for charge separation and a high Voc ⁵⁵ ; and the defect density (NT_PVK) represents the main pathway for non-radiative recombination loss and limiting the carrier lifetime. ⁵⁶ The variation ranges were stablished based on physical models and publish data available in the literature, ^{11, 34, 43, 50, 51} and this are specified in Table 1 and parameter combinations leading to convergence errors in the software were discarded, as these typically arose from physically realistic ranges, disrupting the solution of Poisson’s equation, continuity equations, or boundary conditions. Figure 2 shows stuopti the workflow developed for the process of parameter selection, simulation development, parameter variation, result extraction, and data storage in the final file.

The accuracy of the simulation methodology was validated by successfully reproducing the results reported by research articles as shown in Table 2. For this purpose, over 50 recently published scientific articles were collected that included most of the parameters to simulate a nip-type PSC in SCAPS-1D and also reported the values of Voc, Jsc, FF, and PCE. After a systematic review, to avoid unrealistic results, articles reporting energy efficiencies above the Shockley-Queisser limit ⁵⁷ for cells based on MAPbI 3 , MASnI 3 and FAPbI 3 were excluded, as, according to experimental validations, ⁵⁸ the maximum efficiency achieved for these materials does not exceed 22.2%, 14.35% and 24.66%, respectively. It is important to note that, in the simulations of PSC, variable physical (surface roughness, grain size, and orientation), chemical (temperature and drying time, solvent and antisolvent engineering, and additives) and environmental factors (temperature variations, cloud cover, irradiance, etc.) are not incorporated, which can affect the actual performance of the cells. Therefore, it is expected that the values obtained through simulation will be higher than those observed experimentally maintaining consistency with realistic values. ^{34, 58}

A quantitative assessment was conducted to determine the agreement between simulations and literature reports for the cases that are summarised in Table 2. For the five reproduced studies the root-mean-square error (RMSE) and mean absolute percentage error (MAPE) were calculated using simulated and reported data (Voc, Jsc, FF and PCE). RMSE values are Voc = 0.018 V, Jsc = 0.454 mA/cm ², FF = 11.59%, and PCE = 0.474%. RMSE values corresponding to MAPE are Voc = 1.89%, Jsc = 1.72%, FF = 12.92%, and PCE = 2.23%. These results indicate agreement for Voc, Jsc and PCE, while the error in FF is driven by one case (reported FF = 54.19% vs simulated FF = 78.98%) where there are slight variations in the obtained values compared to the reported ones, which can be attributed to multiple causes, as very few authors report in detail all the simulation conditions or all the parameters used. Some studies include models for recombination, absorption or defects without specifying numerical values (defect density, defect type, or recombination coefficients); therefore, the exact replication of the simulation conditions is limited. This is crucial for a simulation since it considers recombination phenomena, losses due to defects of the layers or interfaces derived from manufacturing processes or impurities in the materials that compose the cell, which can significantly affect the performance of the system. Although reported parameter variability prevents exact replication, the simulated results remain consistent with published data and therefore support the robustness of our methodology.