In recent years, perovskite solar cells have emerged as a promising technology for next-generation photovoltaics due to their exceptional optoelectronic properties, including high carrier mobility, tunable bandgaps, and large absorption coefficients. Among various perovskite materials, all-inorganic CsPbIBr2 has gained significant attention for balancing stability and efficiency, with a bandgap of approximately 1.90 eV. This makes it suitable for applications in semi-transparent devices, such as building-integrated photovoltaics or as top cells in tandem configurations. However, challenges like high defect density, non-radiative recombination, and inefficient charge extraction limit their performance. Electron transport layers (ETLs) play a critical role in addressing these issues, and titanium dioxide (TiO2) has been widely used due to its favorable energy level alignment, chemical stability, and cost-effectiveness. In this study, we focus on the synthesis of nano-sized TiO2 via a solvothermal method and its application as an ETL in carbon-based CsPbIBr2 perovskite solar cells. We investigate the influence of TiO2 concentration on the photovoltaic performance and environmental stability, providing insights into optimization strategies for high-efficiency devices.
The synthesis of nano TiO2 was carried out using a solvothermal approach, which allows for precise control over particle size and morphology. We employed cyclohexane as the organic solvent, with oleic acid and oleylamine serving as capping ligands to prevent aggregation and ensure uniform dispersion. Specifically, 7 mL of oleic acid and 20 mL of cyclohexane were mixed in a conical flask under magnetic stirring. Titanium tetrabutoxide was then added dropwise, followed by the addition of oleylamine. The mixture was heated to 70°C with continuous stirring for 30 minutes before being transferred to an autoclave and maintained at 180°C for 24 hours. After cooling, the product was precipitated with ethanol, centrifuged, and redispersed in toluene to form stable colloidal solutions of varying concentrations (30, 40, 50, and 60 mg mL−1). This method yielded TiO2 nanocrystals (NCs) with a narrow size distribution, as confirmed by subsequent characterization.
Structural analysis of the synthesized TiO2 NCs was performed using X-ray diffraction (XRD). The diffraction pattern exhibited peaks corresponding to the anatase phase, with the most intense peak at the (101) plane. Comparison with standard PDF cards (#21-1272) confirmed the phase purity and high crystallinity, as no impurity peaks were detected. The crystallite size was estimated using the Debye-Scherrer formula:
$$D = \frac{K \lambda}{\beta \cos \theta}$$
where \(D\) is the crystallite size, \(K\) is the Scherrer constant (approximately 0.9), \(\lambda\) is the X-ray wavelength (0.154 nm for Cu Kα radiation), \(\beta\) is the full width at half maximum (FWHM) of the diffraction peak, and \(\theta\) is the Bragg angle. By applying Lorentzian fitting to the (101) peak, the average crystallite size was calculated to be 2.78 nm. Transmission electron microscopy (TEM) images revealed well-dispersed spherical particles with a size range of 1.8 to 4.2 nm, following a normal distribution centered around 3 nm. The organic ligand shell facilitated this dispersion, which is crucial for forming uniform ETL films.
Optical properties of the TiO2 NCs were evaluated using UV-visible absorption spectroscopy. The absorption spectrum showed strong absorption in the UV region, with a bandgap determined from the Tauc plot method. For direct bandgap semiconductors like TiO2, the relationship is given by:
$$(\alpha h \nu)^2 = A (h \nu – E_g)$$
where \(\alpha\) is the absorption coefficient, \(h \nu\) is the photon energy, \(A\) is a constant, and \(E_g\) is the optical bandgap. The calculated bandgap was 3.37 eV, consistent with literature values for anatase TiO2, indicating its suitability as an ETL for blocking low-energy photons and facilitating electron transport.
To fabricate the perovskite solar cells, fluorine-doped tin oxide (FTO) substrates were cleaned sequentially with acetone, isopropanol, ethanol, and deionized water. The TiO2 NC colloids were spin-coated onto the FTO and annealed at 500°C for 30 minutes to form compact ETLs. The CsPbIBr2 perovskite layer was deposited by dissolving stoichiometric amounts of PbBr2 (0.5 M), PbI2 (0.5 M), and CsBr (1.0 M) in dimethyl sulfoxide (DMSO). The precursor solution was spin-coated onto the TiO2 ETL, pre-annealed at 35°C for 3–5 minutes, and then thermally annealed at 160°C for 10 minutes to form crystalline films. Finally, a carbon electrode was applied to complete the device structure. The entire process is summarized in the flowchart below, illustrating the steps from synthesis to device assembly.
The crystal structure of the CsPbIBr2 perovskite films was analyzed by XRD, showing prominent peaks at (100) and (200) planes, indicating preferred orientation along the [100] direction and high crystallinity. Scanning electron microscopy (SEM) images of the films revealed dense, pinhole-free surfaces with grain sizes around 300–400 nm, essential for minimizing recombination losses. Cross-sectional SEM confirmed the layered structure: FTO/TiO2 (≈40 nm)/CsPbIBr2 (≈400 nm)/carbon, with good interfacial contact between layers.
We systematically investigated the effect of TiO2 NC concentration on the performance of the perovskite solar cells. Current density-voltage (J-V) measurements under standard AM 1.5G illumination showed that the photovoltaic parameters varied significantly with ETL thickness. The key parameters, including open-circuit voltage (\(V_{OC}\)), short-circuit current density (\(J_{SC}\)), fill factor (FF), and power conversion efficiency (PCE), are summarized in Table 1 for different TiO2 concentrations.
| TiO2 Concentration (mg mL−1) | \(V_{OC}\) (V) | \(J_{SC}\) (mA cm−2) | FF (%) | PCE (%) |
|---|---|---|---|---|
| 30 | 1.19 | 11.19 | 57.54 | 7.867 |
| 40 | 1.211 | 11.75 | 60.76 | 8.645 |
| 50 | 1.275 | 11.83 | 61.83 | 9.326 |
| 60 | 1.223 | 11.65 | 56.42 | 8.039 |
The data indicate that the optimal performance was achieved at a TiO2 concentration of 50 mg mL−1, with a PCE of 9.326%, \(V_{OC}\) of 1.275 V, \(J_{SC}\) of 11.83 mA cm−2, and FF of 61.83%. This improvement is attributed to enhanced charge extraction and reduced recombination at the ETL-perovskite interface. At lower concentrations, incomplete coverage of FTO led to shunt paths and increased non-radiative recombination. Conversely, higher concentrations resulted in thicker ETLs, impeding electron transport and increasing series resistance. Additionally, excess organic ligands from the synthesis could act as insulating barriers, further reducing efficiency.
To assess reproducibility, we fabricated 25 independent devices for each TiO2 concentration and analyzed the PCE distribution using box plots. The narrow distributions for all concentrations confirmed high reproducibility, with the 50 mg mL−1 group showing the highest median PCE and minimal outliers. Statistical parameters, including mean and standard deviation, are provided in Table 2.
| TiO2 Concentration (mg mL−1) | Mean PCE (%) | Standard Deviation (%) | Maximum PCE (%) |
|---|---|---|---|
| 30 | 7.65 | 0.21 | 7.87 |
| 40 | 8.52 | 0.18 | 8.65 |
| 50 | 9.20 | 0.15 | 9.33 |
| 60 | 7.95 | 0.23 | 8.04 |
Environmental stability is a critical factor for perovskite solar cells, especially under operational conditions. We evaluated the long-term stability of devices with the optimal TiO2 concentration (50 mg mL−1) by storing them in ambient air (relative humidity ≈30%) without encapsulation. The normalized PCE was monitored over 800 hours, and the devices retained over 90% of their initial efficiency, demonstrating excellent environmental stability. This can be attributed to the robust interface between the TiO2 ETL and CsPbIBr2 layer, which mitigates degradation pathways such as ion migration or moisture ingress. The stability performance follows a logarithmic decay model, expressed as:
$$PCE(t) = PCE_0 \times e^{-kt}$$
where \(PCE_0\) is the initial efficiency, \(t\) is time, and \(k\) is the degradation rate constant. For our devices, \(k\) was calculated to be approximately \(1.25 \times 10^{-4}\) h−1, indicating slow degradation compared to other perovskite solar cell configurations.
Further analysis of the charge transport mechanisms revealed that the TiO2 ETL facilitates efficient electron extraction while blocking holes, reducing recombination losses. The energy level alignment between TiO2 and CsPbIBr2 is crucial; the conduction band minimum (CBM) of anatase TiO2 is around -4.2 eV, which aligns well with the CBM of CsPbIBr2 (-3.8 eV), promoting electron injection. The hole-blocking ability arises from the deep valence band maximum (VBM) of TiO2 (-7.4 eV) compared to CsPbIBr2 (-5.7 eV). This alignment minimizes interface recombination, as described by the Shockley-Read-Hall model:
$$R = \frac{np – n_i^2}{\tau_p (n + n_t) + \tau_n (p + p_t)}$$
where \(R\) is the recombination rate, \(n\) and \(p\) are electron and hole concentrations, \(n_i\) is the intrinsic carrier density, \(\tau_n\) and \(\tau_p\) are carrier lifetimes, and \(n_t\) and \(p_t\) are trap densities. By optimizing the TiO2 ETL, we reduced the trap density, leading to lower \(R\) and higher \(V_{OC}\).
In conclusion, we successfully synthesized anatase-phase TiO2 NCs with an average size of 3 nm and a bandgap of 3.37 eV using a solvothermal method. These NCs were employed as ETLs in carbon-based CsPbIBr2 perovskite solar cells, and the concentration was optimized to 50 mg mL−1 for peak performance (PCE of 9.326%). The devices exhibited high reproducibility and exceptional environmental stability, retaining over 90% of initial PCE after 800 hours. This work highlights the importance of ETL engineering in perovskite solar cells and provides a straightforward strategy for enhancing efficiency and durability. Future studies could explore doping or surface modification of TiO2 to further improve charge transport and interface properties for advanced perovskite solar cell applications.