High-performance hematite-integrated perovskite solar cells

Abstract

We report high-performance organometallic perovskite solar cells (PSCs) utilizing a thermodynamically stable hematite (α-Fe₂O₃) as an electron transport layer (ETL). The incorporation of an α-Fe₂O₃ layer could provide suitable energy band alignment with the perovskite, as well as induce a deep valence band maximum, which promotes electron extraction from the conduction band of the perovskite and facilitates hole blocking. Using the solar cell capacitance simulator (SCAPS-1D) software for the optimized PSC under the typical AM 1.5G light spectrum, it was predicted to achieve a competitive power conversion efficiency (PCE) of 25.62% with a high short circuit current (J_SC) of 23.58 mA cm⁻², an open circuit voltage (V_OC) of 1.286 V, and a fill factor (FF) of 84.39%. Moreover, high thermal stability of PSCs with exposure to a high temperature of 85 °C can be attained. Through a series of optimization processes, we conclude that a thinner α-Fe₂O₃ layer (10 nm) improves charge extraction and the transmittance of the visible light, while the decreased defect density significantly reduces recombination rates, thereby enhancing V_OC and PCE. An optimum perovskite thickness of 800 nm was found to maximize light absorption. Additionally, controlled acceptor doping concentration (N_A) enhanced carrier extraction and quasi-Fermi level splitting (QFLS), while high series resistance (R_S) and low shunt resistance (R_SH) were demonstrated to limit FF and efficiency.

---

1. Introduction

Growing industrialization and population growth are accelerating the demand for alternative energy sources. Solar energy's abundance and sustainability make it a viable replacement for fossil fuels.¹ Power conversion efficiency (PCE) and the practical application of photovoltaic technologies have advanced significantly over the past ten years.^2,3 With 27% efficiency,⁴ organometallic halide perovskite solar cells (PSCs) have garnered significant interest in the optoelectronic field in the last decade. Organic–inorganic hybrid perovskites have emerged as promising light absorbers due to their adjustable bandgap (E_g), ambipolar charge transport, high absorption coefficient, defect tolerance, and long charge-carrier diffusion lengths.⁵ Additionally, the substantial improvement of charge transporting materials (CTMs), namely the ETL and hole transporting layer (HTL), as well as their interfaces with the perovskite, is responsible for high efficiency.⁶ Regulating the properties of these CTMs is essential because they govern the physico-chemical characteristics of their interfaces with the perovskite and control the collection and transfer of photo-induced carriers to the appropriate terminals.⁷ To realize efficient solar cells, the CTMs require minimal charge transfer resistance, low optical absorption, a proper energy level matching with the perovskite, and high mobility for effective charge extraction.^8,9

Several n-type metal oxides, such as titania (TiO₂), zinc, niobium, cerium, and tin, have been explored as ETLs in planar n–i–p PSCs.^5,10 The ETL used most often is TiO₂, and solar cells fabricated with this material have shown a PCE approaching 26%.¹¹ However, in addition to its defect-rich surface, the TiO₂ layer degrades PSCs when exposed to ultraviolet (UV) light, which impairs the operational stability of PSCs. Also, the fabrication of a TiO₂ layer demands a high annealing temperature beyond 450 °C, which hampers its use in mass manufacturing.¹² As a thermodynamically stable n-type iron oxide semiconductor, hematite (α-Fe₂O₃) has attracted much attention in optoelectronics owing to its low cost, abundance, non-toxicity, high chemical stability, and desirable energy band position.¹³ Fe₂O₃ has been utilized as a UV-stable photoelectrode in PSCs due to its UV photocatalytic resistance.¹⁴ Hu et al. applied α-Fe₂O₃ as the ETL in methylammonium lead triiodide (MAPbI₃)-based PSCs and achieved an efficiency of 10.78% with enhanced durability as compared to a TiO₂-based device. This enhancement is due to a high built-in potential (V_bi) across MAPbI₃ in α-Fe₂O₃ PSCs, which results in improved charge transport and lower accumulation at the interface.¹⁵ Later, Qureshi et al. reported different solvents to improve the crystallinity and morphology of α-Fe₂O₃ ETLs. Among all, the ethanol-processed α-Fe₂O₃ layer leads to a champion efficiency of 13% with a suppressed hysteresis index of 0.04 in an n–i–p device.¹⁶ Recently, Bilal et al. demonstrated that dual interfacial modification of α-Fe₂O₃ enhances the efficiency and stability of planar PSCs. The 4-trifluorophenylammonium iodide dual modifier resulted in a performance improvement, yielding a champion efficiency of 15.63%.¹⁷

The present work reports for the first time, through systematic SCAPS-1D simulations, the potential of hematite-based ETLs for designing high-performance perovskite photovoltaics. Due to its deeper conduction band minimum than that of the perovskite, α-Fe₂O₃ can accelerate electron extraction from the perovskite absorber. We have tuned the layer thickness, trap-state density (N_Trap), defect density, doping concentration of perovskite, resistive losses, and operating temperature. Numerically optimized α-Fe₂O₃-based PSCs let us achieve a PCE of 25.62%. Simultaneously, the optimized PSC exhibits high thermal stability, maintaining 93% of its initial performance at 85 °C and under AM 1.5G solar irradiation. The suggested approach to ETL optimization may represent a significant advancement for reliable and effective PSCs.

2. Methodology

The device simulations were carried out using SCAPS-1D, a solar cell simulator developed by Burgelman and his colleagues at Ghent University, Belgium.¹⁸ The operating concept relies on solving continuity and Poisson equations. The number of layers can be extended to seven, and the calculation can be run in both light and dark environments. It also enables defects in bulk and at interfaces.¹⁹ The radiative, Auger, and Shockley–Read–Hall (SRH) recombination models were added with interface recombination incorporated via surface defect states at layer boundaries. Among these, SRH recombination is usually the dominant mechanism in PSCs because of trap-assisted charge losses, while radiative and Auger recombination are considered in materials with high carrier densities or direct bandgaps.²⁰

R_rad = B(np − n_i²) (1)

R_Auger = C_nn(np − n_i²) + C_pp(np − n_i²) (2)

where n_i is the intrinsic carrier concentration; n₁ and p₁ stand for carrier concentrations at the trap energy level; τ_n and τ_p denote the carrier lifetime for electrons and holes, respectively; and B, C_n, and C_p represent radiative and Auger recombination coefficients, respectively.

The E_g and electron affinity of the α-Fe₂O₃ film were fixed at 2.13 eV and 3.96 eV, respectively. α-Fe₂O₃ was prepared by the high pressure hydrothermal method, as reported by Khurram et al.²¹ The carrier-charge mobilities and dielectric constant were adopted from density functional theory-based studies,^22,23 describing intrinsic properties of α-Fe₂O₃. A more accurate PSC can be developed by incorporating ETL/MAPbI₃ and MAPbI₃/HTL interfaces. Both interfaces are assumed to have a neutral defect density of 10¹⁰ cm⁻² and a characteristic energy of 0.1 eV (see Table S1). In the initial stage, R_S and R_SH are considered to be 4 Ohm cm² and 300 Ohm cm², which are close to the experimentally obtained values. The key parameters are taken from previously published experimental and theoretical reports, as summarized in Table 1. The standard AM 1.5G illumination and a temperature of 300 K were utilized for the simulations.

Table 1 The starting input parameters of the layers implemented in PSC simulation. The optimized parameters can be found in the results section

Parameters (units)	α-Fe2O3	MAPbI3	Spiro-OMeTAD	TiO2	SnO2	ZnO
Thickness [nm]	40	550	100	10	10	10
Bandgap Eg [eV]	2.13	1.59	3.01	3.2	4.1	3.3
Electron affinity [eV]	3.96	3.9	2.18	4.05	4.39	4.17
Dielectric permittivity	26	25	3	33	7.25	3.7
Effective density of states in the conduction band [cm^−3]	1.0 × 10^17	2.0 × 10^19	2.2 × 10^18	1.0 × 10^17	1.0 × 10^17	1.0 × 10^17
Effective density of states in the valence band [cm^−3]	1.8 × 10^18	1.0 × 10^19	2.0 × 10^18	1.8 × 10^18	1.8 × 10^18	1.8 × 10^18
Electron mobility [cm^2 V^−1 s^−1]	0.098	25	7.9 × 10^−3	18.5	265	8.4 × 10^−1
Hole mobility [cm^2 V^−1 s^−1]	0.031	105	7.9 × 10^−3	1.0 × 10^−4	7.6	91.9
Electron thermal velocity [cm s^−1]	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7
Hole thermal velocity [cm s^−1]	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7	1.0 × 10^7
Shallow donor density [cm^−3]	1.0 × 10^18	—	—	1.0 × 10^18	1.0 × 10^18	1.0 × 10^18
Shallow acceptor density [cm^−3]	—	1.0 × 10^16	1.2 × 10^17	—	—	—
Trap-state density [cm^−3]	1.0 × 10^15	1.03 × 10^16	1.0 × 10^15	1.0 × 10^15	1.0 × 10^15	1.0 × 10^15
Electron capture cross-section [cm^2]	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15
Hole capture cross-section [cm^2]	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15	1.0 × 10^−15
Defect position above the valence band edge [eV]	0.6	0.6	0.6	0.6	0.6	0.6
Energetic distribution	Single	Single	Single	Single	Single	Single
Ref.	21–23	24–26	27–29	30–32	33–35	36 and 37

---

3. Results and discussion

Fig. 1a shows the PSC architecture that was proposed for this study, in which iron oxide (α-Fe₂O₃) was used as an ETL, while conventional MAPbI₃ and 2,2′,7,7′-tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9′-spirobifluorene (spiro-OMeTAD) were utilized as a light absorber and an HTL, respectively. Fig. 1b displays the energy band diagram for the structure of α-Fe₂O₃/MAPbI₃/spiro-OMeTAD. Under the conditions of thermal equilibrium, the Fermi level is homogeneous across the structure. Nevertheless, the light irradiation in the device causes a disruption in this alignment, resulting in the formation of QFLS. It can be seen that the valence band offset of the MAPbI₃/spiro-OMeTAD interface and the conduction band offset of the α-Fe₂O₃/MAPbI₃ interface are small, which promotes the easy transfer of electrons and holes to the respective electrodes (inset of Fig. 1c). As shown in Fig. 1c, the electric field distribution reveals an intensive peak at the ETL/perovskite interface, confirming its role in driving charge separation. The simulated PSC structure was identical to the solar cell fabricated by Chen and co-workers.²⁴ In that work, the α-Fe₂O₃, MAPbI₃, and spiro-OMeTAD films were deposited by the spin-coating method, followed by heat treatment. The current density–voltage (J–V) plots in Fig. 1d show a small mismatch between the simulated and fabricated PSCs, with a high correlation (R² = 0.89). The initial simulated device achieved a PCE of 14.05% with low deviations in J_SC and V_OC. The incident photon-to-current efficiency (IPCE) spectra (Fig. 1e) and the Nyquist plots from impedance analysis (Fig. 1f) show that the modeled curves correlate with the overall patterns observed in the experiments. Unfortunately, the R² values of 0.58 and 0.59 are relatively low, indicating differences that are not fully addressed by the simulation, such as interfacial defects or optical scattering effects. The fabricated films may differ in morphology, roughness, and defect distribution, which influence carrier kinetics and the recombination process, while SCAPS-1D simulations imply idealized films. Fig. 1g shows that the perovskite layer attained the highest absorption coefficient (α), fulfilling its role as a sunlight harvester. The Mott–Schottky plot in Fig. 1h displays that V_bi is 1.1 V, implying good junction formation. Fig. 1i illustrates the recombination rate profile across the PSC. This suggests that non-radiative losses are decreased at the α-Fe₂O₃/perovskite interface.

Fig. 1 Schematic illustration and calibration of the designed PSC structure. (a) Layered structure of the PSC. (b) Band diagram of the simulated device. (c) Electric field distribution. Inset: energy level alignment. (d) J–V plots of the fabricated and simulated cells under AM1.5 solar light.²⁴ (e) IPCE spectra. (f) Nyquist plots extracted from AC impedance analysis. (g) Absorption coefficient spectra. (h) C–V and Mott–Schottky plots under dark conditions. (i) Recombination and generation rate profiles.

Fig. 2a presents the J–V curves of PSCs with varying α-Fe₂O₃ ETL thicknesses. Fig. 2b shows that V_OC and J_SC reduce as the α-Fe₂O₃ thickness increases from 10 nm to 50 nm, indicating limited charge collection performance in thicker ETLs. A thinner α-Fe₂O₃ ETL could facilitate enhanced transmittance of the visible light, which contributes to more photon absorption of the MAPbI₃ layer. The variations in FF and PCE are exhibited in Fig. 2c, implying that an optimum performance is obtained with thinner ETLs (10–20 nm); instead, thicker layers cause transfer barriers that deteriorate photovoltaic properties. The findings are further supported by the IPCE spectra in Fig. 2d, which demonstrate that PSCs with thinner α-Fe₂O₃ layers exhibit enhanced quantum efficiency across the visible spectrum. A thick ETL absorbs/scatters incoming light, thereby suppressing the photon flux reaching MAPbI₃ and consequently lowering J_SC. Fig. 2e shows that the device with an ETL with a thickness of 10 nm has a stronger electric field at the ETL/perovskite interface than the 50-nm ETL. This allows for separate carriers and minimizes recombination losses. Finally, Nyquist curves (Fig. 2f) show enlarged recombination resistance (R_rec) in the thinner α-Fe₂O₃, aligning well with enhanced charge extraction and mitigated recombination pathways. The 10-nm α-Fe₂O₃-based 2DRP device achieved a PCE of 14.47%.

Fig. 2 Optimization of the α-Fe₂O₃ electron transport layer in 2DRP-based PSCs. (a) J–V characteristics. (b) Dependency of V_OC and J_SC on ETL thickness. (c) Evaluation of FF and PCE with ETL thickness. (d) IPCE spectra. (e) Electric field across PSCs with 10 nm and 50 nm ETL thicknesses. (f) Nyquist plots extracted from C–f measurements under AC perturbation.

Fig. S1 shows the J–V curves of 10-nm Fe₂O₃-based devices with varying defect densities in Fe₂O₃ ETLs. Using different defect concentrations in the 10-nm Fe₂O₃ layer induced ignorable changes in device performance due to the ultrathin ETLs, where charge-carrier transfer occurs by drift rather than bulk diffusion. Consequently, the light harvesting efficiency of PSCs is unchanged with increasing defects in ETLs (Fig. S2). Furthermore, the donor doping concentration (N_D) in the 10-nm Fe₂O₃ layer was varied from 10¹⁴ cm⁻³ to 10¹⁸ cm⁻³, as shown in Fig. S3 and S4. Increasing the N_D level led to a slight enhancement in efficiency (0.55%) owing to suppression in series resistance inside Fe₂O₃, implying that the ETLs perform as electron-selective rather than recombination sites.

The effect of perovskite N_Trap on the PSC performance was systematically assessed while changing the values from 10¹⁵ cm⁻³ to 10¹⁷ cm⁻³, as shown in Fig. 3a. It is found that the device's J_SC does not change while varying N_Trap, while V_OC gradually reduces to 0.91 V (Fig. 3b). High traps force photogenerated carriers to recombine before extraction, which narrows the QFLS.³⁸ The IPCE responses (Fig. S5) show nearly identical patterns across different N_Trap values, further corroborating the unchanged trend of J_SC. On the other hand, both FF and PCE are markedly decreased with increasing defect concentration, as shown in Fig. 3c. The power lose in PSCs occurs because of trap-assisted non-radiative. This process is known as SRH recombination. According to the SRH model, a reduction in N_Trap leads to an extension in the charge carrier's lifetime. This, in turn, results in higher diffusion lengths and a decrease in recombination. The Nyquist patterns shown in Fig. 3d prove this behavior by demonstrating smaller semicircle diameters at higher N_Trap, indicating lower R_rec and higher carrier losses. As shown in Fig. 3e, when N_Trap reduces, the recombination rate inside MAPbI₃ and at the interfaces decreases. Additionally, the electric field alignment across the PSC (Fig. 3f) slightly improves by decreasing N_Trap, providing better charge extraction. From the band diagram structure (Fig. 3g and h), the QFLS variations with increasing N_Trap were calculated and are plotted in Fig. 3i. Increasing N_Trap led to noticeable declines in QFLS magnitude due to increased nonradiative recombination losses. The same pattern is observed in the calculated V_OC, indicating voltage losses.

Fig. 3 Optimization of the perovskite trap-state density in PSCs. (a) J–V plots. (b) Variations of V_OC and J_SC. (c) Evaluation of FF and PCE. (d) Nyquist plots. (e) Recombination rate profiles. (f) Electric field distribution. (g) Band structure of PSC at 10¹⁵ cm⁻³ defect density. (h) Band structure of PSC at 10¹⁷ cm⁻³ defect density. (i) Variations of V_OC and QFLS with defect density.

The device performance using the FTO/α-Fe₂O₃/MAPbI₃/spiro-OMeTAD/Au configuration is analyzed with perovskite thicknesses ranging from 400 nm to 800 nm in increments of 100, and the corresponding J–V plots are illustrated in Fig. 4a. According to Fig. 4b and c, the J_SC and PCE parameters notably improve as the MAPbI₃ thickness increases, while V_OC remains almost constant, and FF shows a minor improvement. The enhancement in FF is due to improved carrier collection performance, which decreases resistive losses and mitigates recombination processes. The highest J_SC (23.61 mA cm⁻²) and PCE (17.36%) are realized at a thickness value of 800 nm owing to improved photon absorption.³⁹ The IPCE plots indicate that thicker perovskites show a higher spectral response for visible light, consistent with improved absorption and J_SC (Fig. 4d). Collectively, the impedance analysis shown in Fig. 4e and the recombination rate profiles shown in Fig. 4f exhibit limited changes in R_rec and recombination dynamics, particularly near interfaces.

Fig. 4 Optimization of the perovskite layer thickness in PSCs. (a) J–V curves. (b) Variation of V_OC and J_SC. (c) Dependence of FF and PCE. (d) IPCE spectra. (e) Nyquist plots. (f) Recombination rate profiles.

Fig. 5a shows the J–V performances of α-Fe₂O₃-based PSCs at different perovskite doping densities of shallow acceptor (N_A), ranging from 10¹⁵ to 10¹⁹ cm⁻³. As displayed in Fig. 5b and c, V_OC and FF showed an improved pattern with increasing N_A concentration, leading to a higher PCE value of 21.17% at 10¹⁹ cm⁻³. It was found that an increase in N_A concentration enhances device performance by increasing the V_bi and depletion width at the Fe₂O₃/MAPbI₃ junction, thereby promoting faster charge-carrier separation.⁴⁰ However, the J_SC revealed a slight reduction due to the carrier scattering effect. The bulk recombination in perovskites is significantly minimized at high N_A values due to suppressing trap-mediated SRH processes, as confirmed in Fig. 5d. The electric field distribution in Fig. 5e exhibits an intensified field at a high N_A concentration, facilitating more efficient carrier collection. Also, high N_A doping can improve band bending and result in an increased QFLS, as shown in the energy band diagram (Fig. 5f), which matches well with the observed V_OC increase.

Fig. 5 Effect of the perovskite shallow acceptor density on the photovoltaic performance of PSCs. (a) J–V curves. (b) Variation of V_OC and J_SC. (c) Dependence of FF and PCE. (d) Recombination rate profiles. (e) Electric field distribution. (f) Energy band diagram of the PSC with an acceptor doping of 10¹⁹ cm⁻³.

Individual R_SH and R_S resistances significantly affect the efficiency of the solar cells, as they control the slopes and shapes of the J–V features, which are produced by stacking various materials, metal contacts, the semiconductor–metal interface, and an inadequate PSC fabrication approach. To evaluate the impact of R_S on the PSC, the value of R_S is altered from 2.0 to 10 Ohm cm², while the value of R_SH is kept constant at 300 Ohm cm², as depicted in Fig. 6a. It can be seen that V_OC and J_SC (Fig. 6b) are almost independent of variation in R_S, while FF severely drops, leading to a reduction in PCE to 18.67%, with an increase in R_S to 10 Ohm cm² (Fig. 6c), which stems from resistive power losses and impeded charge extraction. After that, the value of R_SH was varied from 500 Ohm cm² to 4000 Ohm cm² but the R_S value was kept fixed at 2 Ohm cm² and the changes in J–V curves are exhibited in Fig. 6d. The results indicate that the photovoltaic parameters tend to improve with the increase in R_SH until 3000 Ohm cm² (Fig. 6e and f), and after that, they slightly change with further increase in R_SH. These findings corroborate well with those reported in ref. 41. According to the modified diode equation,⁴¹ a low R_SH induces leakage current through recombination routes for photogenerated carriers, leading to decreased device performance, whereas higher R_SH values suppress unwanted leakage currents, as supported by increasing R_rec (Fig. S6). Therefore, the device achieved an optimized PCE of 25.56% with a J_SC of 23.60 mA cm⁻², a V_OC of 1.28 V, and an FF of 84.30% at an R_S of 2 Ohm cm² and an R_SH of 4000 Ohm cm².

Fig. 6 Optimization of the parasitic resistances in PSCs. (a) J–V plots at different series resistances. (b) Variation of V_OC and J_SC with series resistance. (c) Dependence of FF and PCE on series resistance. (d) J–V plots at different shunt resistances. (e) Evaluation of V_OC and J_SC with shunt resistance. (f) Dependence of FF and PCE on shunt resistance.

To observe the influence of operating temperature on the solar cells, the value of temperature is changed from 25 to 85 °C, depicted in Fig. 7a. Increasing the temperature does not have an impact on J_SC, indicating that light harvesting and carrier generation are not affected. It is found that V_OC and FF are progressively reduced with increase in the temperature of PSCs, attributed to the induced reverse saturation current in the PSCs. Interestingly, an increase in the temperature from 25 to 85 °C resulted in a minor reduction in the overall PCE (Fig. 7b), where the PSC retained about 93% of its initial performance, demonstrating its high thermal stability. An increase in temperature creates deformation stress, resulting in a forward movement that increases N_Trap at the interfaces. This influences the diffusion length, thereby limiting device performance. To address this hypothesis, we performed temperature-dependent interface defect-density simulations at low (10¹⁰ cm⁻²) and high (10¹⁵ cm⁻²) defect concentrations of Fe₂O₃/MAPbI₃ and MAPbI₃/spiro-OMeTAD interfaces. The simulated J–V plots of PSCs with different temperatures are shown in Fig. S7, and the corresponding efficiencies are shown in Fig. S8. Increasing the temperature results in a reduction in V_OC for both low- and high-defective devices due to increased charge recombination. A PSC with 10¹⁵ cm⁻² concentration reveals high drops in PCE under both temperature conditions. Therefore, interfacial defect passivation is important to achieve efficient and stable PSCs. Without modification, defective interfaces hinder carrier transfer, leading to energy loss and low power output.¹⁷ The decreasing interfacial defects are demonstrated to enhance the IPCE spectrum (Fig. S9), suggesting that the modification can increase charge carrier generation and minimize the recombination rate (Fig. S10).

Fig. 7 Effect of operating temperature on device performance, along with a comparative analysis against literature-reported Fe₂O₃-based PSCs. (a) J–V plots at different temperatures. Inset: Zoomed-in image of the V_OC region. (b) Tracking the variation of PCE as a function of temperature. (c) The J–V curves of PSCs with various ETL materials. (d) The dependency of V_OC and J_SC on the ETL type. (e) Variations of FF and PCE. (f) Statistical diagram of PCEs extracted from the experimentally reported Fe₂O₃-based PSCs.

Then, the performance of Fe₂O₃-based PSCs was compared with that of other typical semiconductors at the same optimized MAPbI₃ and spiro-OMeTAD layers. Fig. 7c shows the J–V characteristics of different devices simulated at room temperature. All calculations were conducted with optimized series and shunt resistances and an ETL thickness of 10 nm. As depicted in Fig. 7d, all devices yielded close values of J_SC because the perovskite layer mainly governs the light absorption process. A device with the Fe₂O₃ ETL achieved a V_OC value higher than that of a tin oxide (SnO₂) device, and the value closely matched that of titanium dioxide (TiO₂) and zinc oxide (ZnO) devices. In addition, the Fe₂O₃-based PSC demonstrates strong FF (Fig. 7e), thus surpassing the performance of PSCs with ZnO and SnO₂ ETLs due to better conduction band alignment and suppressed recombination losses. A comparison of our results with those of experimentally reported α-Fe₂O₃-based PSCs (Fig. 7f) demonstrates that this study attains a markedly superior PCE (25.62%), exceeding prior reports from 2017 to 2025 (Table 2), thereby illustrating the benefits of the optimized device design, even under thermal stress conditions.

Table 2 Summary of the experimentally reported PSCs with the Fe₂O₃ ETL tested under AM 1.5G illumination

Structure	V OC (V)	J SC (mA cm^−2)	FF (%)	PCE (%)	Year	Ref.
FTO/Ni–Fe2O3/MAPbI3/spiro-OMeTAD/Au	0.92	22.35	69.10	14.20	2017	42
FTO/Fe2O3/MAPbI3/spiro-OMeTAD/Au	0.65	20.4	16.58	10.78	2017	15
FTO/Fe2O3/MAPbI3/spiro-OMeTAD/Au	1.55	11.27	33.00	5.70	2018	43
FTO/Fe2O3/PCBM/MAPbI3/spiro-OMeTAD/Au	0.91	23.2	68.00	14.2	2018	44
FTO/Ti–Fe2O3/MAPbI3/spiro-OMeTAD/Ag	1.10	21.56	75.00	17.85	2018	45
FTO/Fe2O3/PbI2/Cs0.05(FA0.83MA0.17)0.95Pb(I0.83Br0.17)3/spiro-OMeTAD/Au	1.06	23.48	77.10	19.32	2022	46
ITO/Fe2O3/(FAPbI3)0.97(MAPbBr3)0.03/spiro-OMeTAD/Au	0.99	24.79	58.31	14.33	2022	47
ITO/Fe2O3/FA1−xMAxPb(I1−yBry)3/spiro-OMeTAD/Au	0.98	23.45	54.74	12.61	2022	14
FTO/Fe2O3/MAPbI3/spiro-OMeTAD/Au	0.70	21.10	56.70	8.60	2023	48
ITO/Fe2O3/Cs0.05(FA0.83MA0.17)0.95Pb(I0.83Br0.17)3/spiro-OMeTAD/WO3/Au	1.05	19.21	64.48	13.01	2023	16
FTO/TiO2/Fe2O3@SnO2/CsPbBr3/carbon	1.60	7.88	80.85	10.23	2023	49
ITO/Fe2O3/SnO2/MAPbI3/spiro-OMeTAD/Au	1.14	21.29	75.13	18.24	2024	50
ITO/Fe2O3/CF3PhAI/Cs0.05(FA0.83MA0.17)0.95Pb(I0.83Br0.17)3/CF3PhAI/spiro-OMeTAD/Au	1.12	19.01	72.90	15.63	2025	17
FTO/Fe 2 O 3 /MAPbI 3 /spiro-OMeTAD/Au	1.28	23.58	84.39	25.62	2025	This work

---

4. Conclusion

In summary, an efficient and eco-friendly electron-transporting layer based on α-Fe₂O₃ is reported for typical n–i–p PSCs. The small conduction band offset toward the MAPbI₃ absorber and the deep valence band maximum facilitate the extraction and collection of photo-induced electrons from MAPbI₃ with suppression in the recombination losses. Therefore, carrier accumulation at the α-Fe₂O₃/MAPbI₃ interface is blocked, thus making the PSC less sensitive to degradation. We show that 10-nm Fe₂O₃ mainly operates as a tunneling and selective contact rather than a bulk extraction layer. The primary power losses in PSCs take place in the MAPbI₃ layer and at the interfaces, not in the thin ETLs. It was found that the reduction in N_Trap led to decreased recombination processes, improved V_OC, and boosted device efficiency. A nearly unchanged V_OC and fluctuations in J_SC were detected at various perovskite thicknesses, indicating that thickness variations mainly affect J_SC through absorption rather than tuning the QFLS. Moreover, the suppression in the recombination rate at higher acceptor doping concentrations originates from the improved V_bi and QFLS, which promote faster charge injection and reduce trap-mediated SRH recombination paths. Consequently, these advantages lead to substantial improvement of the photovoltaic parameters, and an impressive PCE for the optimized 10-nm-thick α-Fe₂O₃-based PSC of 25.62% (J_SC = 23.58 mA cm⁻², V_OC = 1.286 V, and FF = 84.39%) at 25 °C was achieved. Finally, the optimized device with the α-Fe₂O₃ layer retained 93% of its original efficiency at 85 °C, demonstrating a high thermal stability.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5cp04354a.

The data will be available from the corresponding author on reasonable request.

References

1. M. A. Ismail, G. K. A. Sada, A. Amari, N. Elboughdiri, A. A. H. Kadhum, I. Elbadawy, A. Umarov and S. Madaminov, Design and optimization of a modified solar-driven energy system utilizing advanced heat recovery methods for electricity and hydrogen production in sustainable urban applications, Process Saf. Environ. Prot., 2025, 195, 106720.
2. H. Togun, A. Basem, M. J. Jweeg, N. Biswas, A. M. Abed, D. Paul, H. I. Mohammed, A. Chattopadhyay, B. K. Sharma and T. Abdulrazzaq, Advancing organic photovoltaic cells for a sustainable future: The role of artificial intelligence (AI) and deep learning (DL) in enhancing performance and innovation, Sol. Energy, 2025, 291, 113378.
3. H. Togun, A. Basem, A. A. H. Kadhum, A. M. Abed, N. Biswas, F. L. Rashid, R. A. Lawag, H. M. Ali, H. I. Mohammed and D. K. Mandal, Advancing photovoltaic thermal (PV/T) systems: Innovative cooling technique, thermal management, and future prospects, Sol. Energy, 2025, 291, 113402.
4. Best Research-Cell Efficiency Chart, 2025. https://www.nrel.gov/pv/cell-efficiency.html.
5. A. Aftab and R. Murshed, Recent progress in electron transport layers for enhancing the performance of tin-based perovskite solar cells, Sol. Energy, 2025, 300, 113866.
6. Y. Zheng, Y. Li, R. Zhuang, X. Wu, C. Tian, A. Sun, C. Chen, Y. Guo, Y. Hua and K. Meng, Towards 26% efficiency in inverted perovskite solar cells via interfacial flipped band bending and suppressed deep-level traps, Energy Environ. Sci., 2024, 17(3), 1153–1162.
7. G. Qu, L. Zhang, Y. Qiao, S. Gong, Y. Ding, Y. Tao, S. Cai, X.-Y. Chang, Q. Chen and P. Xie, Self-assembled materials with an ordered hydrophilic bilayer for high performance inverted Perovskite solar cells, Nat. Commun., 2025, 16(1), 86.
8. A. I. Shimul, B. C. Biswas, A. Ghosh, H. Alrafai and A. A. Hassan, A study on optoelectronic properties and charge transport layer influence in novel Sr3PF3-Based perovskite solar cells using numerical simulation and machine learning, Sol. Energy Mater. Sol. Cells, 2025, 293, 113838.
9. M. A. Fakhri, E. T. Salim, M. R. Ketab, H. D. Jabbar, O. A. Ibrahim, A. S. Azzahrani, M. J. AbdulRazzaq, R. A. Ismail, A. Basem and F. H. Alsultany, Optimizing charge transport in hybrid GaN-PEDOT: PSS/PMMADevice for advanced application, Sci. Rep., 2024, 14(1), 12841.
10. L. Zang, C. Zhao, X. Hu, J. Tao, S. Chen and J. Chu, Emerging trends in electron transport layer development for stable and efficient perovskite solar cells, Small, 2024, 20(26), 2400807.
11. H. Yan, X. Zhao, H. Huang, D. Wu, P. Zhu, D. Li, B. Fan, Y. Qiu, Y. Yang and Q. Geng, Flocculating-regulated TiO₂ deposition enables the synergistic effect of doping for perovskite solar cells with efficiency exceeding 25.8%, Adv. Energy Mater., 2025, 15(8), 2403200.
12. T. Chen, J. Xie and P. Gao, Ultraviolet photocatalytic degradation of perovskite solar cells: progress, challenges, and strategies, Adv. Energy Sustainability Res., 2022, 3(6), 2100218.
13. B. Gurudayal, D. Sabba, M. H. Kumar, L. H. Wong, J. Barber, M. Grätzel and N. Mathews, Perovskite–hematite tandem cells for efficient overall solar driven water splitting, Nano Lett., 2015, 15(6), 3833–3839.
14. B. Gu, Y. Du, S. Fang, X. Chen, X. Li, Q. Xu and H. Lu, Fabrication of UV-stable perovskite solar cells with compact Fe₂O₃ electron transport layer by FeCl₃ solution and Fe3O4 nanoparticles, Nanomaterials, 2022, 12(24), 4415.
15. W. Hu, T. Liu, X. Yin, H. Liu, X. Zhao, S. Luo, Y. Guo, Z. Yao, J. Wang and N. Wang, Hematite electron-transporting layers for environmentally stable planar perovskite solar cells with enhanced energy conversion and lower hysteresis, J. Mater. Chem. A, 2017, 5(4), 1434–1441.
16. A. A. Qureshi, S. Javed, M. A. Akram, L. Schmidt-Mende and A. Fakharuddin, Solvent-assisted crystallization of an α-Fe2O3 electron transport layer for efficient and stable perovskite solar cells featuring negligible hysteresis, ACS Omega, 2023, 8(20), 18106–18115.
17. A. S. S. Bilal, M. M. Bilal, M. F. Zia, S. Feroz, M. N. Ullah, M. A. Khan, N. Bano, I. Hussain and R. Fatima, Dual interfacial modification of hematite electron transport layer for efficient and stable perovskite solar cells, Results Eng., 2025, 25, 103954.
18. M. Burgelman, P. Nollet and S. Degrave, Modelling polycrystalline semiconductor solar cells, Thin Solid Films, 2000, 361, 527–532.
19. H. R. Abdul Ameer, A. N. Jarad, K. H. Salem, H. S. Hadi, M. A. Alkhafaji, R. S. Zabibah, K. A. Mohammed, K. Kumar Saxena, D. Buddhi and H. Singh, A role of back contact and temperature on the parameters of CdTe solar cell, Adv. Mater. Process. Technol., 2024, 10(2), 497–505.
20. M. K. Mohammed, Boosting efficiency in carbon nanotube-integrated perovskite photovoltaics, Langmuir, 2024, 40(51), 27114–27125.
21. R. Khurram, Z. Wang, M. F. Ehsan, S. Peng, M. Shafiq and B. Khan, Synthesis and characterization of an α-Fe₂O₃/ZnTe heterostructure for photocatalytic degradation of Congo red, methyl orange and methylene blue, RSC Adv., 2020, 10(73), 44997–45007.
22. C. S. Ahart, K. M. Rosso and J. Blumberger, Electron and hole mobilities in bulk hematite from spin-constrained density functional theory, J. Am. Chem. Soc., 2022, 144(10), 4623–4632.
23. R. A. Lunt, A. J. Jackson and A. Walsh, Dielectric response of Fe2O3 crystals and thin films, Chem. Phys. Lett., 2013, 586, 67–69.
24. H. Chen, Q. Luo, T. Liu, M. Tai, J. Lin, V. Murugadoss, H. Lin, J. Wang, Z. Guo and N. Wang, Boosting multiple interfaces by co-doped graphene quantum dots for high efficiency and durability perovskite solar cells, ACS Appl. Mater. Interfaces, 2020, 12(12), 13941–13949.
25. S. Govinda, B. P. Kore, M. Bokdam, P. Mahale, A. Kumar, S. Pal, B. Bhattacharyya, J. Lahnsteiner, G. Kresse and C. Franchini, Behavior of methylammonium dipoles in MAPbX₃ (X = Br and I), J. Phys. Chem. Lett., 2017, 8(17), 4113–4121.
26. Q. Dong, Y. Fang, Y. Shao, P. Mulligan, J. Qiu, L. Cao and J. Huang, Electron-hole diffusion lengths >175 µm in solution-grown CH3NH3PbI3 single crystals, Science, 2015, 347(6225), 967–970.
27. S. Park, J. H. Heo, C. H. Cheon, H. Kim, S. H. Im and H. J. Son, A [2, 2] paracyclophane triarylamine-based hole-transporting material for high performance perovskite solar cells, J. Mater. Chem. A, 2015, 3(48), 24215–24220.
28. N. J. Jeon, H. G. Lee, Y. C. Kim, J. Seo, J. H. Noh, J. Lee and S. I. Seok, o-Methoxy substituents in spiro-OMeTAD for efficient inorganic–organic hybrid perovskite solar cells, J. Am. Chem. Soc., 2014, 136(22), 7837–7840.
29. H. Ashassi-Sorkhabi and P. Salehi-Abar, How the change of OMe substituent position affects the performance of spiro-OMeTAD in neutral and oxidized forms: theoretical approaches, RSC Adv., 2018, 8(33), 18234–18242.
30. J.-i Fujisawa, T. Eda and M. Hanaya, Comparative study of conduction-band and valence-band edges of TiO₂, SrTiO₃, and BaTiO₃ by ionization potential measurements, Chem. Phys. Lett., 2017, 685, 23–26.
31. S. Mondal and A. Kumar, Tunable dielectric properties of TiO₂ thin film based MOS systems for application in microelectronics, Superlattices Microstruct., 2016, 100, 876–885.
32. T. S. Krasienapibal, T. Fukumura, Y. Hirose and T. Hasegawa, Improved room temperature electron mobility in self-buffered anatase TiO₂ epitaxial thin film grown at low temperature, Jpn. J. Appl. Phys., 2014, 53(9), 090305.
33. M. Meskini and S. Asgharizadeh, Performance simulation of the perovskite solar cells with Ti₃C₂ MXene in the SnO₂ electron transport layer, Sci. Rep., 2024, 14(1), 5723.
34. Z. Li, P. Graziosi and N. Neophytou, Electron and hole mobility of SnO₂ from full-band electron–phonon and ionized impurity scattering computations, Crystals, 2022, 12(11), 1591.
35. L. Yang, Y. Dall'Agnese, K. Hantanasirisakul, C. E. Shuck, K. Maleski, M. Alhabeb, G. Chen, Y. Gao, Y. Sanehira and A. K. Jena, SnO₂–Ti₃C₂ MXene electron transport layers for perovskite solar cells, J. Mater. Chem. A, 2019, 7(10), 5635–5642.
36. A. Sarkar, S. Ghosh, S. Chaudhuri and A. Pal, Studies on electron transport properties and the Burstein-Moss shift in indium-doped ZnO films, Thin Solid Films, 1991, 204(2), 255–264.
37. J. Park, Y. S. Rim, P. Senanayake, J. Wu and D. Streit, Electrical defect state distribution in single crystal ZnO Schottky barrier diodes, Coatings, 2020, 10(3), 206.
38. M. K. Mohammed, E. Y. Salih, M. Alla, P. Kanjariya, A. Rajiv, M. F. Rahman, A. Shankhyan and S. Sundharam, Perovskitoid-Driven 1D@3D Dimensionality for High-Performance Perovskite Solar Cells, ACS Appl. Electron. Mater., 2025, 7(12), 5473–5484.
39. A. Usman and T. Bovornratanaraks, Modeling and optimization of modified TiO2 with aluminum and magnesium as ETL in MAPbI3 perovskite solar cells: SCAPS 1D frameworks, ACS Omega, 2024, 9(38), 39663–39672.
40. R. Teimouri, Z. Heydari, M. P. Ghaziani, M. Madani, H. Abdy, M. Kolahdouz and E. Asl-Soleimani, Synthesizing Li doped TiO2 electron transport layers for highly efficient planar perovskite solar cell, Superlattices Microstruct., 2020, 145, 106627.
41. M. Baro and P. Borgohain, Design and Performance Analysis of Perovskite Solar Cells: Insights from SCAPS 1D Device Simulations, Renewable Energy, 2025, 124225.
42. Y. Guo, T. Liu, N. Wang, Q. Luo, H. Lin, J. Li, Q. Jiang, L. Wu and Z. Guo, Ni-doped α-Fe2O3 as electron transporting material for planar heterojunction perovskite solar cells with improved efficiency, reduced hysteresis and ultraviolet stability, Nano Energy, 2017, 38, 193–200.
43. F. Bouhjar, M. Mollar, S. Ullah, B. Marí and B. Bessaïs, Influence of a compact α-Fe2O3 layer on the photovoltaic performance of perovskite-based solar cells, J. Electrochem. Soc., 2018, 165(2), H30.
44. Q. Hou, J. Ren, H. Chen, P. Yang, Q. Shao, M. Zhao, X. Zhao, H. He, N. Wang and Q. Luo, Synergistic hematite-fullerene electron-extracting layers for improved efficiency and stability in perovskite solar cells, ChemElectroChem, 2018, 5(5), 726–731.
45. W. Zhu, Q. Zhang, C. Zhang, D. Chen, L. Zhou, Z. Lin, J. Chang, J. Zhang and Y. Hao, A non-equilibrium Ti 4+ doping strategy for an efficient hematite electron transport layer in perovskite solar cells, Dalton Trans., 2018, 47(18), 6404–6411.
46. Y. Guo, T. Liu, H. He and N. Wang, Bifunctional interface modification for efficient and UV-robust α-Fe2O3-based planar organic–inorganic hybrid perovskite solar cells, Adv. Compos. Hybrid Mater., 2022, 5(4), 3212–3222.
47. S. Fang, B. Chen, B. Gu, L. Meng, H. Lu and C. M. Li, An ultrathin and compact electron transport layer made from novel water-dispersed Fe₃O₄ nanoparticles to accomplish UV-stable perovskite solar cells, Mater. Adv., 2021, 2(11), 3629–3636.
48. S. Yin, Y. Zou, M. A. Reus, X. Jiang, S. Tu, T. Tian, R. Qi, Z. Xu, S. Liang and Y. Cheng, Tailored fabrication of quasi-isoporous and double layered α-Fe₂O₃ thin films and their application in photovoltaic devices, Chem. Eng. J., 2023, 455, 140135.
49. R. Tui, H. Sui, J. Mao, X. Sun, H. Chen, Y. Duan, P. Yang, Q. Tang and B. He, Round-comb Fe₂O₃@ SnO₂ heterostructures enable efficient light harvesting and charge extraction for high-performance all-inorganic perovskite solar cells, J. Colloid Interface Sci., 2023, 640, 918–927.
50. M. A. Jan, A. A. Qureshi, H. M. Noman and F. Yang, α-Fe₂O₃/SnO₂ electron transport bilayer for efficient and stable perovskite solar cells, J. Mater. Sci.: Mater. Electron., 2024, 35(27), 1784.

---