Low‐Temperature Synthesis of SnO2 Nanocrystals as Electron Transport Layers for High‐Efficiency CsPbI2Br Perovskite Solar Cells

Perovskite solar cells (PSCs) have recently become a hot topic in photovoltaics due to their high power conversion efficiency (PCE) and low‐cost processing. As a key component of PSCs, electron transport layers (ETLs) that play a vital role in efficient PSCs generally require high charge extraction ability, mobility, and easy fabrication. Herein, a simple route to obtain dispersion of tin oxide nanocrystals (SnO2 NCs) with uniform diameters and high stability as efficient ETLs for CsPbI2Br solar cells is demonstrated. The champion device achieves a remarkable PCE of 16.22% with an open‐circuit voltage of 1.30 V. This work offers a facile and effective way to fabricate high‐performance ETL nanocrystals in PSCs.

Here, we developed a facile solution synthetic method of SnO 2 colloid NCs using urea (CO(NH 2 ) 2 ) to control the formation kinetics of SnO 2 nanocrystals. The as-prepared SnO 2 NCs can be fabricated as compact ETL films with superior conductivity, mobility, low trap density, and improved energy-level alignment, which can dramatically promote the improvement of charge extraction and suppress nonradiative recombination of PSCs. The CsPbI 2 Br solar cells with the as-synthesized SnO 2 ETLs obtained an open-circuit voltage (V OC ) of 1.30 V and a PCE of 16.22%, respectively, which are much higher than that based on commercial SnO 2 ETLs (V OC = 1.21 V, PCE = 14.00%).

Results and Discussion
We propose a synthetic mechanism of SnO 2 NCs in Figure 1a and S1, Supporting Information. First, the Sn(OH) þ intermediate species are produced after the exposure of SnCl 2 ·2H 2 O in distilled water and the solution becomes milky. A sharp decrease in pH in the first 6 h reflects the hydrolysis of SnCl 2 ·2H 2 O ( Figure S2a, Supporting Information). Sn(OH) þ can be partially transformed back to Sn 2þ due to the increase of H þ as the hydrolysis goes on. Notably, urea is decomposed into ammonium (NH 4 þ ) and cyanate (CNO À ) ion, which is completely converted to NH 4 þ ions and carbon dioxide (CO 2 ) with the consumption of hydrogen ion (H þ ), [46] which explains the mild decrease of pH in the second stage of the reaction ( Figure S2a, Supporting Information). The generated NH 4 þ ions with positive charge surround the as-synthesized SnO 2 NCs, which would contribute to the extreme stability of the colloid. After stirring for a few days, Sn(OH) þ is oxidized to Sn 4þ by the dissolved oxygen in the solution. [47] Figure S2b, Supporting Information shows the changes in dissolved oxygen levels as a function of reaction time. Due to the decomposition of urea, the existing hydroxide ion (OH À ) reacts with Sn 4þ to form Sn(OH) 4 , which is finally converted to SnO 2 through dehydration reaction. However, due to the approximate ionic radii of Sn 2þ (0.62 Å) and Sn 4þ (0.69 Å), Sn 2þ from the first step may be incorporated into the hexa-coordinated SnO 2 lattice. [48] The presence of urea can control the reaction route and suppress the self-doping of Sn 2þ in SnO 2 NCs. For convenience, we denote commercial SnO 2 colloid dispersion and as-synthesized SnO 2 colloid dispersion as c-SnO 2 and u-SnO 2 , respectively, in the following discussion. The X-ray diffraction (XRD) pattern of u-SnO 2 NCs powder ( Figure 1b) show diffraction peaks at 26.6, 33.8, 37.8, 51.8, 61.7, and 65.7°, which can be ascribed to tetragonal rutile SnO 2 . In Figure 1d, the transmission electron microscopy (TEM) image shows the extremely small nanocrystals, and most of them are between 3 and 5 nm in diameter. This result is in agreement with the dynamic light scattering (DLS) of u-SnO 2 NCs dispersion (2.19 nm), which is obviously smaller than c-SnO 2 (4.94 nm) (Figure 2c,f ). The zeta potentials of SnO 2 NCs dispersions obtained by DLS analysis are À17.1 and 22.4 mV for the c-SnO 2 and u-SnO 2 NCs dispersion, respectively. The negative zeta potential for c-SnO 2 NCs corresponds to the presence of potassium hydroxide (KOH) stabilizer. [49] In contrast, the positive zeta potential for u-SnO 2 should be caused by the hydrolysis of SnCl 2 precursor. The larger absolute value of zeta potential for u-SnO 2 indicates more intense electrostatic repulsion between the charged SnO 2 nanocrystals, which suggests better stability than c-SnO 2 . High-resolution TEM (HRTEM) confirms the high crystallinity of u-SnO 2 NCs and shows a nanocrystal with (110) lattice spacing (0.333 nm) of SnO 2 with the tetragonal rutile structure. The selected-area electron diffraction (SAED) pattern of u-SnO 2 NCs shows distinct electron diffraction circles indicating a polycrystalline nature (Figure 1c). X-ray photoelectron spectroscopy (XPS) was implemented to analyze the surface chemical states of SnO 2 film. The separation of 8.45 eV between the Sn3d5/2 and Sn3d3/2 levels shows that the SnO 2 films only exists in a tetravalent oxidation state, [50] and the shift of the Sn 3d5/2 state toward lower binding energy indicates that the u-SnO 2 has lower oxygen vacancy than c-SnO 2 ( Figure S5, Supporting Information).
Compact SnO 2 films were deposited onto indium tin oxide (ITO) substrates by the spin coating method. Atomic force microscopy (AFM) was used to study the surface morphology of the SnO 2 films. Both films exhibit smooth surface with roughness of 1.574 and 1.678 nm for u-SnO 2 and c-SnO 2 films, respectively (Figure 2a,d). In addition, the scanning electron We used ultraviolet photoelectron spectroscopy (UPS) to identify the energy band structure of the two ETLs (Figure 3a,b). The deeper conduction band minimum (CBM) value of the u-SnO 2 film (À4.18 eV) ensures good electron extraction compared with that of the c-SnO 2 film (À3.61 eV). Meanwhile, the bandgaps of the c-SnO 2 film and the u-SnO 2 film were 3.98 and 4.06 eV according to the ultraviolet-visible (UV-vis) absorption spectra (Figure 3c), respectively. Owing to the larger bandgap of the u-SnO 2 film, its valence band maximum (VBM) (À8.24 eV) is much deeper compared to that of the c-SnO 2 film (À7.59 eV), which indicates the reduction of charge recombination in PSCs for u-SnO 2 . [51] The energy-level diagram of CsPbI 2 Br PSCs with the c-SnO 2 film and the u-SnO 2 film is shown in Figure 3d. The better energy-level alignment between the u-SnO 2 film and CsPbI 2 Br film would contribute to charge extraction. The CsPbI 2 Br film on u-SnO 2 ETL exhibits enhanced photoluminescence intensity (Figure 3e), indicating slow nonradiative recombination, which might be owing to fewer interfacial defects and lower surface roughness.
To obtain excellent photovoltaic (PV) performance, high carrier density and mobility for ETLs are crucial in PSCs. Electrical conductivity tests of SnO 2 ETLs are delivered based on structures of ITO/ETLs/Au. In Figure 4a, the conductivity (σ) of u-SnO 2 and c-SnO 2 is calculated to be 1.48 Â 10 À2 and 0.95 Â 10 À2 mS cm À1 . All detailed information is summarized in Table S1, Supporting Information. We fabricated a threeelectrode compartment with SnO 2 films, an Ag/AgCl/3.5 M KHCO 3 electrode and a platinum gauze as the working electrode, reference electrode, and counter electrode for C-V spectroscopy, respectively. Mott-Schottky (M-S) plots are shown in Figures 4b under a potential range of À1.5-0 V. The corresponding carrier density can be calculated using Equation (1) and (2) 1 where φ bi is the built-in potential, q is the elementary charge of an electron (1.6 Â 10 À19 C), ε 0 is the permittivity of free space (8.85 Â 10 À12 F m À1 ), ε r is the relative dielectric constant 9, [52] A is the area of SnO 2 film immersed in electrolyte (2.25 cm 2 ), C is the interfacial capacitance, V is the applied voltage (V ), and N d is the carrier density. The carrier density of u-SnO 2 film (2.72 Â 10 21 cm À3 ) increased by %17% in comparison with c-SnO 2 film (2.33 Â 10 21 cm À3 ), which contributes to the improvement in conductivity of the relevant SnO 2 ETLs. To estimate the trap density of ETLs, the space-charge-limited current (SCLC) curves were obtained using lateral structured devices, which were constructed with two 100 μm-gap Au interdigital where L is the thickness of the ETL (100 μm), and V TFL is the trapfilled limit voltage. [53] The trap density of the c-SnO 2 and u-SnO 2 films was calculated to be 1.68 Â 10 10 and 3.68 Â 10 9 cm À2 , respectively. In general, the superior conductivity, large carrier density, and low trap density of the u-SnO 2 film suggest the faster charge transport and suppression of carrier accumulation at the ETL/perovskite interface, resulting in the excellent PV performance in the u-SnO 2 -based PSC. Dark current curves shown in Figure S6, Supporting Information indicate that the leakage current at the low-voltage scale was weakened dramatically in the PSCs with u-SnO 2 ETL, which suggests the better interfacial contact between u-SnO 2 ETL and the active layer. PSC devices were fabricated with the structure of ITO/ SnO 2 NCs/CsPbI 2 Br/poly(3-hexylthiophene2,5-diyl) (P3HT)/Ag (Figure 5a). The condition of perovskite films, which is relevant to crystallinity, grain size, and surface roughness, usually determines the performance of PSCs. XRD patterns of two perovskite films are shown in Figure S7, Supporting Information; two   Figure S8, Supporting Information, the SEM image of CsPbI 2 Br film coated on u-SnO 2 film shows larger grain size than that of c-SnO 2 , and both of them display smooth surface. Figure 5b and Table S2 Figure S10, Supporting Information. There is a significant increase in V OC and FF from c-SnO 2 to u-SnO 2 .
To evaluate the charge recombination of the PSCs, we conducted light intensity-dependent V OC and electrochemical impedance spectroscopy (EIS) measurement. Figure 5d shows the slopes of 2.61 and 1.50 kT/q for the PSCs with c-SnO 2 and u-SnO 2 ETLs, respectively. The smaller slope of the devices with u-SnO 2 indicates the lesser trap-assisted recombination than that of the device with c-SnO 2 . [54] As shown in Figure 5e, the recombination resistance (R rec ) of each PSC device is calculated from the diameter of the semicircle. Apparently, R rec of PSC device with u-SnO 2 is almost doubled compared to that with c-SnO 2 . In addition, it can be clearly seen that the PSC devices with u-SnO 2 exhibit higher R rec value than those of the control ones, which indicates the remarkable decrease in undesirable recombination ( Figure S11, Supporting Information). To further explore the relationship between carrier dynamics and device performance with different ETLs, we conducted the transient photovoltage (TPV) and transient photocurrent (TPC) measurements in PSC devices. TPV decay curves demonstrate that the PSC with u-SnO 2 ETL (76.1 μs) shows a longer lifetime than c-SnO 2 -based one (38.3 μs). The faster photocurrent decay response of u-SnO 2 -based device (0.9 μs) in Figure S12, Supporting Information reveals faster electron transport than c-SnO 2 -based one (2.0 μs).
Finally, we appraised the unencapsulated device stability with a device structure of ITO/SnO 2 /CsPbI 2 Br/P3HT/Au against heat. The thermal stability of CsPbI 2 Br devices was measured at 85°C in N 2 -filled glovebox in Figure S13, Supporting Information. In comparison with the %40% loss of PCE for the devices based on c-SnO 2 ETLs, the CsPbI 2 Br solar cells based on u-SnO 2 ETLs retained over 90% of its initial efficiency after 500 h, exhibiting excellent long-term stability.

Conclusion
We have reported a synthetic method for the preparation of colloidal tin oxide nanocrystals with uniform particle size and good crystallinity. The as-synthesized SnO 2 nanocrystalline films have superior electrical conductivity, optimized band edges, and low trap density compared to commercial ones. The champion www.advancedsciencenews.com www.small-science-journal.com device based on u-SnO 2 film attained a high PCE of 16.22% with a V OC of 1.30 V of CsPbI 2 Br cells. Our work will provide new insights into the design and synthesis of novel charge transport materials for ETLs in PSCs.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.