Periodic Acid Modification of Chemical‐Bath Deposited SnO2 Electron Transport Layers for Perovskite Solar Cells and Mini Modules

Abstract Chemical bath deposition (CBD) has been demonstrated as a remarkable technology to fabricate high‐quality SnO2 electron transport layer (ETL) for large‐area perovskite solar cells (PSCs). However, surface defects always exist on the SnO2 film coated by the CBD process, impairing the devices’ performance. Here, a facile periodic acid post‐treatment (PAPT) method is developed to modify the SnO2 layer. Periodic acid can react with hydroxyl groups on the surface of SnO2 films and oxidize Tin(II) oxide to Tin(IV) oxide. With the help of periodic acid, a better energy level alignment between the SnO2 and perovskite layers is achieved. In addition, the PAPT method inhibits interfacial nonradiative recombination and facilitates charge transportation. Such a multifunctional strategy enables to fabricate PSC with a champion power conversion efficiency (PCE) of 22.25%, which remains 93.32% of its initial efficiency after 3000 h without any encapsulation. Furthermore, 3 × 3 cm2 perovskite mini‐modules are presented, achieving a champion efficiency of 18.10%. All these results suggest that the PAPT method is promising for promoting the commercial application of large‐area PSCs.


Introduction
Perovskite solar cells (PSCs) have made great progress over the past ten years, [1][2][3] while the highest certified power conversion efficiency (PCE) exceeded 25% recently. [4] A typical planar PSC device includes an electron transport layer (ETL), perovskite layer, hole transport layer (HTL), and electrodes. ETL plays a critical role in extracting and transporting the photogenerated electrons as well as blocking holes to reduce charge recombination. [5] Therefore, the ETL's intrinsic optical and electronic properties will influence the photovoltaic performance and stability of a PSC. TiO 2 is one of the most popular ETLs in PSCs due to its favorable electronic and optical properties. For a long period, the compact TiO 2 (c-TiO 2 )/mesoporous TiO 2 (mp-TiO 2 ) stack occupies an important position in high-efficiency PSCs. [6,7] However, TiO 2 still has a few disadvantages, such as relatively low bulk electron mobility (<1 cm 2 V −1 s −1 ), high processing temperature, and high photocatalytic activity, which may decrease the long-term stability of PSCs. [8] Recently, SnO 2 has been considered as a promising ETL material owing to its outstanding bulk electron mobility (≈250 cm 2 V −1 s −1 ), low-temperature processability (<200°C), and superb long-term operational stability. [9] Spin coating is the most typical technique to deposit SnO 2 thin films. In 2015, Fang and co-workers used solution-processed SnO 2 as ETL material of PSCs for the first time, achieving a champion PCE of 17.21%. They synthesized SnO 2 films by a facile sol-gel approach: spin coating SnCl 2 ·2H 2 O precursor solutions and followed by thermal annealing at 180°C. [10] Later, much work reported the deposition of synthesized SnO 2 nanoparticles (NPs) or quantum dots (QDs) to form SnO 2 films with better crystallization and less recombination centers. [11] You and co-workers used commercial SnO 2 colloidal precursor solutions to fabricate a dense and pinhole-free SnO 2 film. The PSCs based on this low-temperature solutionprocessed SnO 2 film achieved a certified PCE of 19.9%. [12] Chemical bath deposition (CBD) is another good technique for fabricating SnO 2 thin films. Different from the spin-coating method, the CBD approach is extraordinarily suitable for depositing large-area, dense and uniform SnO 2 films without any area limits. [13] It is very suitable for the future commercialization of perovskite solar cells. In 2016, Correa-Baena and co-workers firstly introduced the CBD technique for depositing the SnO 2 layer in planar PSCs, which yielded a high open-circuit voltage of 1.21 V at a bandgap of 1.62 eV, and achieved a stabilized PCE of 20.7%. [14] Recently, Seo and co-workers fabricated highly efficient PSCs based on CBD-coated SnO 2 , with a certified PCE of 25.2%, indicating that CBD-SnO 2 could potentially improve photovoltaic performances. [15] However, surface defects, such as oxygen vacancies and hydroxyl groups, always exist on the surface of CBD-coated SnO 2 , which not only increase the nonradiative recombination and impair the device's performances but also accelerate the degradation of the devices. [16] Sargent and co-workers reported surface modification of CBD-SnO 2 by ammonium fluoride, leading to reduced defects sites and better energy level match. This treatment achieved higher open circuit voltages and a champion PCE of 23.2%. [17] Qi and co-workers reported the incorporation of KMnO 4 into the CBD process of SnO 2 films. The strong oxidizing nature of KMnO 4 promoted the conversion from Sn(II) to Sn(IV), leading to reduced trap defects and higher carrier mobility of SnO 2 . In addition, K + ions and Mn 2+ ions improved perovskite films' crystallinity and phase stability. [16] In this work, we developed a periodic acid post-treatment (PAPT) method to modify the CBD-SnO 2 films. Periodic acid (H 5 IO 6 ), with acidity and strong oxidizing properties, shows a bifunctional effect on SnO 2 layers. It can not only react with the hydroxyl groups on the surface of SnO 2 films but also oxidize the Tin(II) oxide (SnO) to Tin(IV) oxide (SnO 2 ). With the help of periodic acid, a better energy level alignment between the SnO 2 and perovskite layers and reduced defect sites in the SnO 2 layer were achieved simultaneously. Furthermore, we investigated the effect of periodic acid on the photovoltaic performances and stability of the corresponding perovskite devices. Figure 1 shows the schematic illustration of SnO 2 films fabricated by the chemical bath deposition and PAPT method. Etched FTO glasses were immersed in dilute SnCl 2 aqueous solution with urea, thioglycolic acid, and hydrochloric acid, and then heated at 70°C for 4 h. For PAPT-modified substrates, they were subsequently immersed in periodic acid solution with different concentrations for 5 min. After that, the substrates were thoroughly washed with deionized water to remove the residual reagents and annealed at 150°C for 40 min.

Results and Discussions
To study the treatment mechanism of periodic acid to SnO 2 films, we conducted X-ray photoelectron spectroscopy (XPS) measurements. For the Sn 3d XPS spectra (Figure 2a), the two peaks of Sn 3d 5/2 and 3d 3/2 in pristine SnO 2 shifted from 487.2 and 495.6 eV to the higher binding energies of 487.6 and 496.1 eV, with the spin-orbit splitting energy of 8.4 eV. [18] The higher binding energy of Sn 3d indicates more Sn (II)'s conversion to Sn (IV), [16,19] potentially widening the band gap and decreasing the number of oxygen vacancies. [20,21] In terms of O 1s spectra (Figure 2b), two peaks of around 532.5 and 531.1 eV refer to the hydroxyl group (-OH) on the SnO 2 surface and the saturated lattice oxygen in SnO 2 film, respectively. [17,20] The calculated areas of the -OH in pristine SnO 2 reduced from 82.9% to 35.3% in the PAPT-modified SnO 2 , indicating more chemisorbed non-lattice oxygen transferred to lattice oxygen and existed in the form of SnO 2 . It is known that surface -OH introduces a deep energy level into the bandgap that is near the valence band, bringing the generation of non-radiative recombination and energy losses in the devices. [22] As shown in Figure 2c, two typical peaks of I 3d orbitals appeared in the PAPT-modified SnO 2 sample. In contrast, the pristine SnO 2 sample showed no signal, indicating iodide remained on the PAPT-modified SnO 2 surface by chemical doping. The two peaks of I 3d 5/2 and 3d 3/2 located at 619.8 and 631.3 eV refer to iodine anion (I − ), which were consistent with the referenced SnI 2 signals in the previous reports. [23,24] According to the XPS results of Sn 3d and O 1s, we believe that the PAPT treatment urges the transformation of Sn(II) to Sn(IV) and the conversion of chemically absorbed hydroxyl to lattice oxygen, which lead to reduced trap states and potentially higher V OC and FF of the PSCs.
As previously reported, [17,22] four kinds of oxygen defects exist on the surface of the CBD-SnO 2 films, including two kinds of hydroxyl defects and two kinds of oxygen vacancy defects (V O ): 1) terminal hydroxyl (OH T ), which is bonded to a single Sn atom on the surface (Figure S1a, Supporting Information); 2) bridge hydroxyl (OH B ), including a bridge oxygen and an absorbed hydrogen atom (Figure S1b, Supporting Information); 3) bridgeoxygen vacancy (V OB ), i.e., the missing of the bridge oxygen atom (Figure S1c, Supporting Information); 4) deep oxygen vacancy (V OD ), i.e., the missing of another oxygen atom located at a deeper layer than bridge oxygen ( Figure S1d, Supporting Information). To figure out if the iodine atom will remain at the hydroxyl site or fill the V O , we conducted density functional theory (DFT) calculations. The stoichiometric (110) lattice plane was chosen as the studied surface due to that it is the most thermodynamically stable configuration. [12,25] According to the results of DFT calculations, the absorption energies (E a ) of the iodine atom absorbed on The negative value of E a indicates that the iodine atom is prone to be absorbed in the same site after the hydroxyl groups were eliminated by periodic acid. Also, the more negative E a implies iodine atoms are more apt to be absorbed in the bridge oxygen sites. As for V O , the E a of the iodine atom substituted the bridge V OB is exactly the same as previously calculated results, which is −2.732 eV. And the E a of the iodine atom substituted at V OD (−1.973 eV) showed a similar result (Figure 2f), indicating that iodine atoms can effectively passivate the V O defects on the SnO 2 surface. Above all, the schematic image of the PAPT method is shown in Figure 2g.
From the above analysis, we found that periodic acid has the following three functions: 1) H + can decrease the surface hydroxyl groups; 2) The strong oxidation of IO 5− 6 can oxidize Sn 2+ into Sn 4+ ; 3) The reduced I − can also reduce the surface -OH and fill in the oxygen vacancies. To figure out which one plays a more critical role, we treated SnO 2 films with sulfuric acid (H 2 SO 4 ) and hydroiodic acid (HI), respectively. It should be noted that the concentrations of H 2 SO 4 and HI are 5 and 2 mm, respectively, which are consistent with the H + and I − concentrations in the optimal concentration of periodic acid (2 mm, which will be discussed later).
XPS spectra of SnO 2 films treated by H 2 SO 4 and HI are shown in Figure S2 (Supporting Information). The double peaks of the Sn 3d of both samples are located at 487.2 and 495.6 eV ( Figure  S2a, Supporting Information), which is consistent with the position of the control sample (Figure 2a), indicating that a small amount of Sn 2+ may exist in the SnO 2 films. It is because dilute H 2 SO 4 and HI cannot oxidize Sn 2+ into Sn 4+ . As shown in Figure  S2b (Supporting Information), the O 1s peak can be divided into two peaks corresponding to -OH and lattice oxygen, respectively. It can be seen that the -OH peak area of H 2 SO 4 -modified SnO 2 samples reduces from 82.9% to 40.1%, while that of HI-modified samples only reduces it to 53.2%, indicating that H + is more effective for eliminating the surface -OH than I − . Finally, we observed weak I 3d double peaks on the surface of the HI-treated sample ( Figure S2c, Supporting Information), the intensity of which is much lower than that of the H 5 IO 6 -treated sample, suggesting that more iodine atoms in periodic acid stay on the SnO 2 films. Above all, periodic acid has a stronger ability to eliminate surface oxygen defects than H 2 SO 4 and HI. Furthermore, we carried out XPS depth analysis measurements to study the oxygen defects inside the SnO 2 film. The etching depth is around 10 nm. It can be seen that the peak of O 1s is located around 531.1 eV ( Figure S3, Supporting Information), corresponding to the position of lattice oxygen, while no -OH peaks are observed. In addition, I 3d double peaks are observed in PAPT-modified SnO 2 samples, indicating that only the periodic acid can fill the oxygen vacancies inside the SnO 2 films.
Ultraviolet photoelectron spectroscopy (UPS) measurements were performed to investigate the influence of energy level by periodic acid. The valence band maximum (VBM) values of samples can be calculated by the formula of VBM = 21.22 eV − E cutoff + E onset . [26] As shown in Figure 3a,  Figure S4, Supporting Information) and 1.56 eV ( Figure S5, Supporting Information), respectively. Therefore, the schematic energy level diagram of FTO, the pristine SnO 2 , PAPT-modified SnO 2 , and perovskite is shown in Figure 3c. A better energy band alignment between the PAPT-modified SnO 2 and perovskite layers was achieved due to the upshift of the conduction band and Fermi level of SnO 2 . As a result, a reduced energy barrier and faster electron extraction at the SnO 2 and perovskite interface could be realized. [27] Grazing incidence X-ray diffraction (GIXRD) was conducted to investigate the change in the crystal structure of the SnO 2 films. As shown in Figure 4a, several peaks of the PAPT-modified SnO 2 film are slightly higher than that of the pristine sample, indicating a better crystallinity was obtained after PAPT treatment. We carried out electron conductivity and space-chargelimited current (SCLC) measurements of devices based on the FTO/SnO 2 /Au structure, as shown in Figure 4b and Figure S6 (Supporting Information). The PAPT-modified SnO 2 film exhibit higher conductivity and electron mobility than those of the pristine SnO 2 film, suggesting the PAPT treatment could improve the electron transport of the SnO 2 films. As shown in According to the UV-visible light transmittance spectra of SnO 2 films ( Figure S7, Supporting Information), the PAPT treatment would not influence the high visible light transmission of SnO 2 films. Atomic force microscopy (AFM) measurements were carried out to investigate the surface morphology of the chemicalbath deposited SnO 2 layers ( Figure S8, Supporting Information). The PAPT-modified SnO 2 film showed a relatively lower roughness with a root mean square (RMS) of 9.91 nm compared to the pristine SnO 2 film (RMS = 11.08 nm), indicating the PAPT treatment would not impair the crystal growth process of the upperlevel perovskite layers. Scanning electron microscope (SEM) of perovskite films ( Figure S9, Supporting Information) deposited on the surface of the pristine and PAPT-modified SnO 2 films could also prove this point of view.
Steady-state photoluminescence (PL) and time-resolved PL (TRPL) measurements were carried out to study the charge www.advancedsciencenews.com www.advancedscience.com   transfer dynamics at the SnO 2 /perovskite interfaces. As shown in Figure 4d, the perovskite film on the PAPT-modified SnO 2 film exhibited only 23.1% PL intensity of that of the perovskite on the pristine SnO 2 film. It indicates more efficient charge extraction of PAPT-modified SnO 2 film, owing to the better energy level match. For the TRPL spectra, both samples showed biexponential decay curves (Figure 4e). The average carrier lifetime ( avg ) was fitted and listed in Table S1 (Supporting Information). The perovskite film deposited on the PAPT-modified SnO 2 film exhibited the avg of 50.55 ns, which is about 60% shorter than that of the perovskite film on the pristine SnO 2 film, 82.70 ns. It also suggests faster electron collection and transportation.
We also estimated the trap density by measuring the dark current density-voltage (J-V) curves based on electron-only devices (FTO/SnO 2 /perovskite/PCBM/Ag). The typical dark J-V curves can be divided into three regions: ohmic region, trapfilling limited (TFL) region, and trap-free space charge limited current region. The electron defect density (N t ) can be estimated by the trap-filling limit voltage (V TFL ). [25,28] The electron-only devices based on the PAPT-modified SnO 2 film obtained a much lower V TFL (0.334 V) and N t (3.51 × 10 15 cm −3 ) compared to those based on the pristine SnO 2 film (V TFL = 0.574 V and N t = 6.04 × 10 15 cm −3 ), as shown in Figure 4f. The fitting parameters (R 2 ) were listed in Table S2 (Supporting Information). Obviously, the PAPT treatment helps reduce the electron defect density and facilitates the charge transport between the SnO 2 and perovskite films.
To investigate the photovoltaic performances of PSCs with the PAPT-modified SnO 2 films, we fabricated devices with a structure of FTO/SnO 2 /perovskite/OAI/Spiro-OMeTAD/Au, as shown in Figure 5a,b. After careful optimization of the periodic acid concentration (1-4 mm) (Table S3, Supporting Information), an optimal concentration of the PAPT treatment (2 mm) led to a champion PCE of 22.25% (Figure 5c  The PCE enhancement is mainly attributed to the increment of FF and V OC , which benefited from the interface modification by the periodic acid treatment. In addition, we investigated the role of periodic acid on the transport properties in PSCs by dark I-V characterization. As shown in Figure 5h, PAPT-modified devices exhibited a significantly lower dark current, indicating the reduction of leakage current between the SnO 2 /perovskite interface, which could be attributed to the higher compactness and uniformity of the PAPT-modified SnO 2 films. [25] Moreover, we conducted electrochemical impedance spectroscopy (EIS) to investigate the charge transfer and recombination processes in the PSCs. Figure 5i shows the Nyquist plots of devices based on the pristine SnO 2 and PAPT-modified SnO 2 films, measured at a bias potential of −1.0 V under dark conditions. It is worth mentioning that the semicircle at the high-frequency region representing the charge transfer resistance (R ct ) was too small to be identified. [29] Therefore, only one semicircle at the low-frequency region was observed from the Nyquist plots, reflecting the interfacial recombination resistance (R rec ) between the SnO 2 layer and the perovskite layer. According to the fitting results (Table S4, Supporting Information) employing the equivalent circuit (inset of Figure 5h), the R rec value increased from 240.6 to 399.7 Ω after the PAPT modification. The higher R rec indicates a lower charge re-combination rate, [29] thus contributing to the improved FF and V OC . In addition, the fitted series resistance (R S ) values are 29.49 and 23.96 Ω for PSCs based on the pristine SnO 2 and PAPTmodified SnO 2 films, respectively. The smaller R S is ascribed to increased conductivity of the SnO 2 layer, consistent with the previous measurement. We also performed the electric capacityvoltage (C-V) measurement to analyze the built-in potential (V bi ) by the Mott-Schottky equation. As shown in Figure S11 (Supporting Information), the PAPT-modified device displayed a higher V bi (0.94 V) than the control device (0.77 V), indicating more efficient charge separation and collection and, thus, higher V OC . [30] Drive-level capacitance profiling (DLCP) measurement was conducted to obtain spatial distributions of trap densities in SnO 2 films. [31,32] As shown in Figure S12 (Supporting Information), the trap density in PAPT-SnO 2 is relatively lower than that of the control-SnO 2 film, indicating PAPT modification is beneficial for reducing the defect density in the SnO 2 layer. All these characterization results suggest the PAPT modification optimizes the interfacial charge transportation and inhibits the interfacial recombination between the ETL and perovskite layers, leading to better photovoltaic performances.
Finally, we fabricated the laser-etched perovskite solar minimodules (PSMs) with an area of 3 × 3 cm 2 using a slot-die coating method, as shown in Figure 6a. The champion PSM based on the PAPT-modified SnO 2 film achieved an excellent PCE of 18.10% (J SC = 7.16 mA cm −2 , V OC = 3.45 V, and FF = 73.42%) with an aperture area of 2.29 cm 2 , while the control one only achieved a PCE of 16.13% (J SC = 7.08 mA cm −2 , V OC = 3.35 V, and FF = 68.02%). These results reveal that the CBD process with the periodic acid treatment is quite suitable for large-area production of PSCs and PSMs. Next, we evaluated the stabilities of PSCs according to the International Summit on Organic PV Stability (ISOS) protocols. [33] As shown in Figure 6b, the unencapsulated device based on the PAPT-modified SnO 2 film remained at 93.32% of its initial PCE after 3000 h of exposure to a relative humidity of 20-25% at 20°C in an ambient atmosphere (conform to ISOS-D-1 procedure). The control device only kept 29.79% of its original PCE under the same conditions. The excellent stability of PAPT-modified devices could be ascribed to the more compact and smoother interface between the SnO 2 and perovskite films, with fewer defects and channels for moisture.

Conclusions
In summary, we developed a powerful periodic acid posttreatment (PAPT) method to modify the SnO 2 films fabricated by the CBD process. According to XPS results, it is found that periodic acid promoted the transformation of Sn(II) to Sn(IV) and the conversion of chemically absorbed hydroxyl to lattice oxygen, which could lead to reduced trap states and potentially higher V OC of the PSCs. In addition, a better energy band alignment between the PAPT-modified SnO 2 and perovskite layers was achieved due to the upshift of the conduction band and Fermi level of SnO 2 . As a result, a reduced energy barrier and faster electron extraction at the SnO 2 and perovskite interface could be realized. PSCs based on PAPT-modified SnO 2 layer achieved a champion PCE of 22.25%, with J SC , V OC , and FF of 25.02 mA cm −2 , 1.09 V, 81.55%, respectively. Meanwhile, PSCs based on pristine SnO 2 layer only achieved a champion PCE of 19.77% with J SC , V OC , and FF of 25.09 mA cm −2 , 1.06 V, 74.28%, respectively. The PCE enhancement is mainly attributed to the increment of FF and V OC , benefited from the interface modification by the periodic acid treatment. In addition, the devices based on PAPT-modified SnO 2 exhibited excellent stability, which remain 93.32% of its initial efficiency stored at 20°C, 20-25% RH after 3000 h without any encapsulation. Furthermore, we fabricated 3 × 3 cm 2 perovskite mini-modules based on PAPT-modified SnO 2 layers, achieving a high efficiency of 18.10% with the aperture area of 2.29 cm 2 , while the control one only achieved PCE of 16.13%. These results indicate that PAPT method efficiently boost the large-scale production of SnO 2 ETL and perovskite modules.

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