Facile Synthesis of Spherical TiO2 Hollow Nanospheres with a Diameter of 150 nm for High-Performance Mesoporous Perovskite Solar Cells

The electron transport layer (ETL) of organic–inorganic perovskite solar cells plays an important role in their power conversion efficiency (PCE). In this study, TiO2 hollow nanospheres with a diameter of 150 nm were prepared by a facile synthesis method. The synthesized TiO2 hollow nanospheres had a highly porous structure with a surface area of 85.23 m2 g−1, which is significantly higher than commercial TiO2 (P25) (54.32 m2 g−1), indicating that they can form an ideal mesoporous layer for Formamidinium iodide-based perovskite solar cells (PSCs). In addition, the nanospheres achieved a remarkable perovskite performance, and the average PCE increased from 12.87% to 14.27% with a short circuit current density of 22.36 mAcm−2, an open voltage of 0.95 V, and a fill factor of 0.65. The scanning electron microscopy images revealed that the enhanced PCE could be due to the improved carrier collection and transport properties of the nanosphere, which enabled efficient filtration of perovskite into the TiO2 mesoporous ETL. The TiO2 hollow nanospheres fabricated in this study show high potential as a high-quality ETL material for efficient (FAPbI3)0.97(MAPbBr3)0.03-based PSCs.


Introduction
Hybrid organic-inorganic perovskite solar cells (PSCs) that employ formamidinium lead halide as a light-absorbing material have been applied to various photovoltaic devices owing to their excellent optoelectronic properties [1][2][3][4][5][6][7]. Over the years, the photovoltaic conversion efficiency (PCE) of PSCs has been improved to obtain efficient and affordable PSC devices, with the PCE recently reaching 25.2% [8]. In mesoscopic PSCs, the perovskite is prepared on a mesoporous TiO 2 layer, which accepts photoexcited electrons from the absorbers and transports them to the fluorine-doped tin oxide (FTO) substrate. Mesoporous layer-free PSCs have been demonstrated to exhibit high PCE in planar-structured devices; however, planar-structured devices show hysteresis behavior and trap the photogenerated free electrons, thus hindering further enhancement of the PCE [9][10][11][12][13][14][15]. Therefore, it is important to develop ideal mesoporous materials that possess sizeable pores, extended contact areas, and defect-free nanostructures with negligible boundaries to prevent the charge recombination [16][17][18]. Small-sized TiO 2 nanoparticles (NPs) possess a large surface area; however, their narrow pore size reduces the infiltration of the perovskite absorber to the electron transport layer (ETL) [19][20][21]. In contrast, larger-sized TiO 2 NPs possess a smaller surface area, which affects efficient charge injection through the hole transport layer (HTL). Therefore, it is important to develop an optimized mesoporous material with relatively large TiO 2 NPs to form a scaffold structure for the collection of electrons from the perovskite layer.
Hui Wang et al. [22] fabricated hollow TiO 2 nanospheres with a shell thickness of approximately 30 nm and applied them in PSCs. The PSC achieved a PCE of 15.87%, which could be due to the improved carrier transport properties and matching of the hollow TiO 2 to the perovskite layer. In addition, hollow spheres TiO 2 with a radius of 200-300 nm using titanium isopropoxide as precursor material were fabricated in a previous study. The TiO 2 was applied in a PSC, and the PSC achieved a PCE of 14.2% owing to the suppression of carrier recombination at perovskite/TiO 2 interfaces [23]. These results indicate that the PCE of PSCs can be further enhanced using hollow TiO 2 nanospheres with a uniform shape as the mesoporous ETL.
In this study, a large spherical TiO 2 hollow structure with a diameter of 150 nm was successfully synthesized by controlling the sol-gel reaction. The synthesized nanoporous spherical TiO 2 hollow nanospheres was a monodispersed material with a porous structure and high surface area and was thus highly suitable for fabricating mesoporous PSCs. The PSC fabricated with the as-synthesized TiO 2 hollow nanospheres exhibited enhanced photovoltaic properties, and the PCE increased from 12.87% to 14.27% under AM 1.5G irradiation (100 mW·cm −2 ).

Synthesis of Carbonaceous Nanospheres (CNs)
Carbonaceous nanospheres (CNs) were synthesized according to the method developed by Sun et al. and reported in other researches [24,25]. Briefly, 89 g of glucose (Sigma-Aldrich, St. Louis, MO, USA) was firstly dissolved in water (250 mL) to form a clear solution, and the solution was transferred into a 500 mL Teflon autoclave and heated at 180 • C for 5.5 h. The mixture was collected by centrifugation, washed 3-4 times with ethanol and water, and oven-dried at 80 • C for over 4 h for further experiments.

Synthesis of the TiO 2 Hollow Nanospheres
To synthesize the mesostructured TiO 2 hollow nanospheres, 0.02 g of the as-synthesized CNs was homogeneously dispersed in ethanol (10 mL) by ultrasonication and stirred for 30 min at room temperature. Then, 0.04 mL of titanium isopropoxide (TIIP) (Sigma-Aldrich) was slowly injected into the dispersion under vigorous stirring for 30 min. CNs-TiO 2 nanospheres were collected by centrifugation; they were washed with a mixture of ethanol/water several times and then dried in air for a day. Finally, the as-synthesized CNs-TiO 2 nanospheres were sintered at 450 • C for 2 h in a muffle furnace under air to obtain the TiO 2 hollow nanospheres.

Fabrication of the Perovskite Solar Cells
Briefly, a glass/FTO substrate (the sheet resistance of 15 Ωsq −1 ) was cleaned with isopropanol, acetone, distilled water, and ethanol. The cleaned FTO substrate was dried under a N 2 stream and treated with ultraviolet ozone for 15 min to remove any organic contamination. To prepare the compact TiO 2 layer on the substrate, the compact TiO 2 solution was spin-coated on the cleaned substrate using a titanium diisopropoxide bis(acetylacetone) (75% in 1-butanol) solution (0.15 M, 2000 rpm, 20 s), after which the substrate was dried at 125 • C for 20 min. The substrate was cooled down to room temperature, and the porous layer was spin-coated on the compact TiO 2 at 4000 rpm for 20 s using the TiO 2 paste (Supplementary Materials Figure S1) diluted in ethanol (1:6, wt:wt), and finally annealed at 480 • C for 30 min. After cooling down to room temperature, the samples were moved into a nitrogen-filled glovebox (water and oxygen content below 1 ppm) for fabricating perovskite and HTL layers. The perovskite layer was deposited by a two-step deposition method. First, the PbI 2 precursor solution was spin-coated onto the ETL at 2000 rpm for 20 s using 1.3 M PbI 2 (600 mg PbI 2 in 900 µL of N, N-Dimethylformamide (DMF, Sigma-Aldrich) and 100 µL of Dimethyl sulfoxide (DMSO, Sigma-Aldrich). Subsequently, a mixture perovskite precursor solution of FAI:MABr:MACl (60:6:6, mg) in 1 mL of 2-propanol was loaded onto the PbI 2 layer for 20 s (loading time) and then spin-coated at 4000 rpm for 20 s. The perovskite films were heated at 150 • C for 15 min. Next, a solution containing 72.3 mg of spiro-MeOTAD (Sigma-Aldrich), 1 mL of chlorobenzene (Sigma-Aldrich), 28.8 µL of 4-tert-butylpyridine, and 17.5 µL of Li-TFSI solution (52 mg Bis(trifluoromethane)sulfonimide lithium salt (Li-TSFI, Sigma-Aldrich) in 100 µL acetonitrile (Sigma-Aldrich) was deposited on top of the perovskite layer. Finally, 80 nm gold electrodes were thermally evaporated on the spiro-MeOTAD film.

Device Characterization
The ultraviolet-visible (UV-vis) light absorption spectra of the films were measured using a UV-vis spectrophotometer (Agilent 8453, Agilent technologies, Santa Clara, CA, USA) to evaluate the absorption property of the films. The X-ray diffraction (XRD) patterns of the samples were obtained using a XRD Rigaku DMAX 2200 system (Rigaku, Tokyo, Japan) with Cu Kα (λ = 0.15406 nm) as the X-ray source. The specific Brunauer-Emmett-Teller (BET) surface areas of the samples were investigated using an ASAP 2020 (Micromeritics, Atlanta, GA, USA) apparatus. An infrared spectrometric analyzer (Vertex 70, Bruker, Ettlingen, Germany) was used to record the Fourier transform infrared (FTIR) spectra. The top and cross-sectional morphologies of the samples were examined by field emission scanning electron microscopy (FESEM, Hitachi S-4700, Tokyo, Japan) operated at 10 kV. The steady-state photoluminescence (PL) spectra were determined using a QuantaMaster TM 50 PTI (Birmingham, New Jersey, NJ, USA). A sun simulator (Polaromix K201, Solar simulator LAB 50, McScience K3000, McScience, Gyeonggi-do, Korea) with an irradiance of 100 mW cm −2 (AM 1.5G) was used to simulate solar irradiation. The external quantum efficiency (EQE) was measured using McScience K3100 measurement system (McScience, Gyeonggi-do, Korea).

Results
In this study, to synthesize the TiO 2 hollow nanospheres, uniform CNs (diameter of around 240 nm, Supplementary Materials Figure S2) were synthesized by a hydrothermal method using glucose, which is easily disintegrated at low temperatures (under 450 • C), as the soft template. The schematic of the synthesis procedure is shown in Figure 1. Particularly, a titanium isoporoxide aqueous solution was used as the precursor solution during the reaction and was absorbed by the negatively charged CNs. propanol was loaded onto the PbI2 layer for 20 s (loading time) and then spin-coated at 4000 rpm for 20 s. The perovskite films were heated at 150 °C for 15 min. Next, a solution containing 72.3 mg of spiro-MeOTAD (Sigma-Aldrich), 1 mL of chlorobenzene (Sigma-Aldrich), 28.8 μL of 4-tert-butylpyridine, and 17.5 μL of Li-TFSI solution (52 mg Bis(trifluoromethane)sulfonimide lithium salt (Li-TSFI, Sigma-Aldrich) in 100 μL acetonitrile (Sigma-Aldrich) was deposited on top of the perovskite layer. Finally, 80 nm gold electrodes were thermally evaporated on the spiro-MeOTAD film.

Device Characterization
The ultraviolet-visible (UV-vis) light absorption spectra of the films were measured using a UV-vis spectrophotometer (Agilent 8453, Agilent technologies, Santa Clara, CA, USA) to evaluate the absorption property of the films. The X-ray diffraction (XRD) patterns of the samples were obtained using a XRD Rigaku DMAX 2200 system (Rigaku, Tokyo, Japan) with Cu Kα (λ = 0.15406 nm) as the X-ray source. The specific Brunauer-Emmett-Teller (BET) surface areas of the samples were investigated using an ASAP 2020 (Micromeritics, Atlanta, GA, USA) apparatus. An infrared spectrometric analyzer (Vertex 70, Bruker, Ettlingen, Germany) was used to record the Fourier transform infrared (FTIR) spectra. The top and cross-sectional morphologies of the samples were examined by field emission scanning electron microscopy (FESEM, Hitachi S-4700, Tokyo, Japan) operated at 10 kV. The steady-state photoluminescence (PL) spectra were determined using a Quan-taMaster TM 50 PTI (Birmingham, New Jersey, NJ, USA). A sun simulator (Polaromix K201, Solar simulator LAB 50, McScience K3000, McScience, Gyeonggi-do, Korea) with an irradiance of 100 mW cm −2 (AM 1.5G) was used to simulate solar irradiation. The external quantum efficiency (EQE) was measured using McScience K3100 measurement system (McScience, Gyeonggi-do, Korea).

Results
In this study, to synthesize the TiO2 hollow nanospheres, uniform CNs (diameter of around 240 nm, Supplementary Materials Figure S2) were synthesized by a hydrothermal method using glucose, which is easily disintegrated at low temperatures (under 450 °C), as the soft template. The schematic of the synthesis procedure is shown in Figure 1. Particularly, a titanium isoporoxide aqueous solution was used as the precursor solution during the reaction and was absorbed by the negatively charged CNs. The SEM images of the P25 TiO2 powders (Sigma-Aldrich).and hollow nanospheres are shown in Figure 2a,b. It can be seen that samples nanospheres in diameter were estimated to be ~150 nm; they were monodispersed during the entire synthesis process, while P25 are tiny with aggregated particles. The effect of the size of the TiO2 hollow nanosphered and P25 on the ETL was examined. The surface morphologies of P25 and the assynthesized TiO2 hollow nanospheres (Figure 2c,d) were analyzed by FESEM. As shown in the image, both TiO2 nanoparticles are uniformly distributed in the TiO2 films. However, the size of the TiO2 hollow nanospheres was significantly larger than that of the P25.  The SEM images of the P25 TiO 2 powders (Sigma-Aldrich).and hollow nanospheres are shown in Figure 2a,b. It can be seen that samples nanospheres in diameter were estimated to be~150 nm; they were monodispersed during the entire synthesis process, while P25 are tiny with aggregated particles. The effect of the size of the TiO 2 hollow nanosphered and P25 on the ETL was examined. The surface morphologies of P25 and the as-synthesized TiO 2 hollow nanospheres (Figure 2c,d) were analyzed by FESEM. As shown in the image, both TiO 2 nanoparticles are uniformly distributed in the TiO 2 films. However, the size of the TiO 2 hollow nanospheres was significantly larger than that of the P25. Figure 2e,f shows the surface morphologies of the perovskite layer on the mesoporous TiO 2 films. The perovskite deposited on the TiO 2 hollow nanospheres exhibited a smooth and flattened surface morphology with fewer pinholes than that deposited on the P25. TiO2 films. The perovskite deposited on the TiO2 hollow nanospheres exhibited a smooth and flattened surface morphology with fewer pinholes than that deposited on the P25.  Figure 3a shows the FTIR spectra of the P25 and TiO2 hollow nanospheres recorded from 4000 cm −1 to 400 cm −1 . The peaks at 3650 and 1628 cm −1 were ascribed to the C-H stretching region and H-O-H bending vibrations signature of water molecules. In addition, a strong absorption peak of Ti-O-Ti vibration was observed at 500 cm −1 in both samples, indicating that there was no shift in the main peak of Ti-O bonding compared to the TiO2 anatase phase of P25. Additionally, the characteristic peak of the C-H stretching vibration at 2920 cm −1 was not observed, confirming that the carbon nanosphere template was completely removed during the calcination process.
The crystal structure of the obtained powder was further investigated by XRD. The XRD patterns of P25 and the TiO2 hollow nanospheres were recorded from 20 to 80°, as shown in Figure 3b. The XRD peaks at 25.2°, 37.8°, 48.1°, 53.9°, 55.1°, 62.8°, 68.9°, 70.3°, and 75.1° were ascribed for the TiO2 anatase crystalline structure, which is consistent with the ICSD, No. 75-1537. In addition, the carbon reflection peak was not observed in the sample after calcination, which is consistent with the FTIR results [26]. Furthermore, the crystalline size of the TiO2 hollow nanospheres determined using the Debye-Scherrer equation from anatase (101) was approximately 11 nm.   The optical properties of the P25 and TiO2 hollow nanospheres were investigated by UV-vis spectrophotometry in the range of 300-800 nm (Figure 4a). Their direct bandgap energy was estimated by plotting (αhν) 2 versus the photon energy, Eg = 1240/λ, as shown in Figure 4b. As shown in the image, the bandgap energies of the P25 and TiO2 hollow nanospheres were 3.18 and 3.10 eV, respectively. The low bandgap energy of the hollow TiO2 nanospheres could be due to the higher average crystal size of the TiO2 hollow nanospheres because of the quantum size-effect [27,28]. Therefore, the decrease in the bandgap between the conduction band and the valence band was expected to enhance the charge injection from the perovskite layer to the ETL.
The specific surface areas and pore size distributions of the P25 and TiO2 hollow nanospheres were investigated by N2 adsorption/deposition analysis, and the results are shown in Figure 5a,b. The TiO2 hollow nanospheres exhibited a broad hysteresis loop of In addition, the carbon reflection peak was not observed in the sample after calcination, which is consistent with the FTIR results [26]. Furthermore, the crystalline size of the TiO 2 hollow nanospheres determined using the Debye-Scherrer equation from anatase (101) was approximately 11 nm.
The optical properties of the P25 and TiO 2 hollow nanospheres were investigated by UV-vis spectrophotometry in the range of 300-800 nm (Figure 4a). Their direct bandgap energy was estimated by plotting (αhν) 2 versus the photon energy, Eg = 1240/λ, as shown in Figure 4b. As shown in the image, the bandgap energies of the P25 and TiO 2 hollow nanospheres were 3.18 and 3.10 eV, respectively. The low bandgap energy of the hollow TiO 2 nanospheres could be due to the higher average crystal size of the TiO 2 hollow nanospheres because of the quantum size-effect [27,28]. Therefore, the decrease in the bandgap between the conduction band and the valence band was expected to enhance the charge injection from the perovskite layer to the ETL.
The specific surface areas and pore size distributions of the P25 and TiO 2 hollow nanospheres were investigated by N 2 adsorption/deposition analysis, and the results are shown in Figure 5a,b. The TiO 2 hollow nanospheres exhibited a broad hysteresis loop of type IV isotherms at a relative pressure of 0.5, showing that it is a mesoporous structure. However, the TiO 2 hollow nanospheres exhibited a relatively narrow pore size distribution of approximately 11.24 nm. In addition, the TiO 2 hollow nanospheres had a large surface area of 85.23 m 2 g −1 , which was 1.56 times higher than that of P25 (54.32 m 2 g −1 ). The higher surface area of the hollow TiO 2 nanospheres is expected to facilitate the efficient filtration of perovskite to the ETL, thus improving the photovoltaic properties of the PSC. The optical properties of the P25 and TiO2 hollow nanospheres were investigated by UV-vis spectrophotometry in the range of 300-800 nm (Figure 4a). Their direct bandgap energy was estimated by plotting (αhν) 2 versus the photon energy, Eg = 1240/λ, as shown in Figure 4b. As shown in the image, the bandgap energies of the P25 and TiO2 hollow nanospheres were 3.18 and 3.10 eV, respectively. The low bandgap energy of the hollow TiO2 nanospheres could be due to the higher average crystal size of the TiO2 hollow nanospheres because of the quantum size-effect [27,28]. Therefore, the decrease in the bandgap between the conduction band and the valence band was expected to enhance the charge injection from the perovskite layer to the ETL.
The specific surface areas and pore size distributions of the P25 and TiO2 hollow nanospheres were investigated by N2 adsorption/deposition analysis, and the results are shown in Figure 5a,b. The TiO2 hollow nanospheres exhibited a broad hysteresis loop of type IV isotherms at a relative pressure of 0.5, showing that it is a mesoporous structure. However, the TiO2 hollow nanospheres exhibited a relatively narrow pore size distribution of approximately 11.24 nm. In addition, the TiO2 hollow nanospheres had a large surface area of 85.23 m 2 g −1 , which was 1.56 times higher than that of P25 (54.32 m 2 g −1 ). The higher surface area of the hollow TiO2 nanospheres is expected to facilitate the efficient filtration of perovskite to the ETL, thus improving the photovoltaic properties of the PSC.  The commercial TiO2 and TiO2 hollow nanospheres were employed as ETLs in the mesoporous-structured perovskite solar cells consisting of FTO/c-TiO2/mp-TiO2/perovskite/Spiro/Au (Figure 6a). Figure 6b indicates an energy diagram of the PSCs with P25  The commercial TiO 2 and TiO 2 hollow nanospheres were employed as ETLs in the mesoporous-structured perovskite solar cells consisting of FTO/c-TiO 2 /mp-TiO 2 /perovskite/Spiro/Au (Figure 6a). Figure 6b indicates an energy diagram of the PSCs with P25 and TiO 2 hollow nanospheres. The cross-section image of the PSC fabricated using the TiO 2 hollow nanospheres (Figure 6c) revealed that the thicknesses of the Spiro-OMeTAD HTL and Au electrodes were approximately 250 nm and 80 nm, respectively. The enhanced physical properties of the TiO 2 hollow nanospheres are summarized in Table 1.  To investigate the applicability of the TiO2 hollow nanospheres as an efficient ETL for PSC, the perovskite layer of (FAPbI3)0.97(MAPbBr3)0.03 was fabricated by a two-step deposition method [4,29,30]. Thirty-six independent PSCs fabricated using P25 and TiO2 nanospheres as the ETLs were examined, the results are shown in Figure 7a To investigate the applicability of the TiO 2 hollow nanospheres as an efficient ETL for PSC, the perovskite layer of (FAPbI 3 ) 0.97 (MAPbBr 3 ) 0.03 was fabricated by a two-step deposition method [4,29,30]. Thirty-six independent PSCs fabricated using P25 and TiO 2 nanospheres as the ETLs were examined, the results are shown in Figure 7a   The PL of the samples was measured and analyzed to investigate the charge extraction dynamics at the perovskite/ETL interface. Figure 8a shows the PL spectra of the FTO/perovskite, FTO/cp-TiO2/P25/perovskite, and FTO/cp-TiO2/TiO2 hollow nanospheres/perovskite at wavelengths from 750 to 810 nm at a light excitation of 530 nm, where the emission peaks were obtained at approximately 780 nm for all the films. The PL intensity peak of the TiO2 hollow nanospheres decreased significantly compared to that of the P25 sample, which could be due to the efficient filtration of the perovskite precursor into the mesoporous structure and the reduced charge recombination behavior at the perovskite/ETL interfaces. As shown in Figure 8b, in comparison to the absorbance in the perovskite deposited on P25 and hollow TiO2 samples, an enhanced absorption value of perovskite/hollow TiO2/TFO sample from 550 nm to 800 nm can be realized for the efficient absorber layer of PSCs. The PL of the samples was measured and analyzed to investigate the charge extraction dynamics at the perovskite/ETL interface. Figure 8a shows the PL spectra of the FTO/perovskite, FTO/cp-TiO 2 /P25/perovskite, and FTO/cp-TiO 2 /TiO 2 hollow nanospheres/perovskite at wavelengths from 750 to 810 nm at a light excitation of 530 nm, where the emission peaks were obtained at approximately 780 nm for all the films. The PL intensity peak of the TiO 2 hollow nanospheres decreased significantly compared to that of the P25 sample, which could be due to the efficient filtration of the perovskite precursor into the mesoporous structure and the reduced charge recombination behavior at the perovskite/ETL interfaces. As shown in Figure 8b, in comparison to the absorbance in the perovskite deposited on P25 and hollow TiO 2 samples, an enhanced absorption value of perovskite/hollow TiO 2 /TFO sample from 550 nm to 800 nm can be realized for the efficient absorber layer of PSCs. Because the mesoporous scaffold plays an important role in the recombination behavior and hysteresis effect of PSCs, the current-voltage (J-V) curves of the forward and backward scans were obtained to investigate the hysteresis of the ETL-based PSCs (Figure 9a) [31][32][33]. The external efficiency quantum of the complete devices is shown in Figure 9b. The integrated current density of the TiO2 hollow nanospheres is 23 mAcm −2 with the external quantum efficiency of 89.5%, which is higher than that of P25. The high JSC could be due to the high quality of perovskite layer loaded on the TiO2 hollow nanospheres. Because the mesoporous scaffold plays an important role in the recombination behavior and hysteresis effect of PSCs, the current-voltage (J-V) curves of the forward and backward scans were obtained to investigate the hysteresis of the ETL-based PSCs (Figure 9a) [31][32][33]. The external efficiency quantum of the complete devices is shown in Figure 9b. The integrated current density of the TiO 2 hollow nanospheres is 23 mAcm −2 with the external quantum efficiency of 89.5%, which is higher than that of P25. The high J SC could be due to the high quality of perovskite layer loaded on the TiO 2 hollow nanospheres. Because the mesoporous scaffold plays an important role in the recombination behavior and hysteresis effect of PSCs, the current-voltage (J-V) curves of the forward and backward scans were obtained to investigate the hysteresis of the ETL-based PSCs (Figure 9a) [31][32][33]. The external efficiency quantum of the complete devices is shown in Figure 9b. The integrated current density of the TiO2 hollow nanospheres is 23 mAcm −2 with the external quantum efficiency of 89.5%, which is higher than that of P25. The high JSC could be due to the high quality of perovskite layer loaded on the TiO2 hollow nanospheres. As shown in the image, the TiO 2 hollow nanospheres exhibit relatively less hysteresis in the J-V measurement than P25. Furthermore, the TiO 2 hollow nanosphere-based PSCs exhibited a high reproductivity with the average PCEs of 18 cells in the range from 13.5 to 14.5% (Figure 9c). In conclusion, the TiO 2 hollow nanospheres with a diameter of 150 nm exhibited highly specific surface areas, thus making them ideal ETL materials for fabricating high-efficiency PSCs. The photovoltaic performance of the PSCs based on hollow TiO 2 and P25 in this study ware compared with those of the PSCs reported in previous studies (see Table 3).

Conclusions
In conclusion, TiO 2 hollow nanospheres with a diameter of 150 nm were fabricated by a simple and effective method and applied as mesoporous ETLs for PSCs. The fabricated PSCs based on the TiO 2 hollow nanospheres exhibited a PCE of 14.27% under AM 1.5G irradiation (100 mW·cm −2 ) with a J SC of 23.84 mAcm −2 , V OC of 0.94 V, and FF of 0.64, which were higher than those of the P25-based PSCs. In addition, the PL measurement revealed efficient electron injection from the perovskite to the TiO 2 layer, which could be due to the increase in the pore size of the hollow nanospheres.