Enhanced Performance of Planar Perovskite Solar Cells Using Low-Temperature Solution-Processed Al-Doped SnO2 as Electron Transport Layers

Lead halide perovskite solar cells (PSCs) appear to be the ideal future candidate for photovoltaic applications owing to the rapid development in recent years. The electron transport layers (ETLs) prepared by low-temperature process are essential for widespread implementation and large-scale commercialization of PSCs. Here, we report an effective approach for producing planar PSCs with Al3+ doped SnO2 ETLs prepared by using a low-temperature solution-processed method. The Al dopant in SnO2 enhanced the charge transport behavior of planar PSCs and increased the current density of the devices, compared with the undoped SnO2 ETLs. Moreover, the enhanced electrical property also improved the fill factors (FF) and power conversion efficiency (PCE) of the solar cells. This study has indicated that the low-temperature solution-processed Al-SnO2 is a promising ETL for commercialization of planar PSCs.


Background
The solar energy has attracted much attention since it is a renewable and clean energy source [1][2][3][4]. In recent years, a large amount of research groups have focused on organic-inorganic lead halide perovskite solar cells as it have the advantages of a lower manufacturing cost and a simpler process compared with Si solar cells. Moreover, PSCs have a great potential for providing an alternative to conventional photovoltaic devices. The PCE of PSCs has increased from 3.8 to 22.1% in a few years [5][6][7][8][9]. However, the efficiency and stability of PSCs strongly depend on some crucial factors, for instance, the morphology of perovskite films and the preparation of electron/hole transport material [10][11][12][13][14][15].
Both electron transport layers (ETLs) and hole transport layers (HTLs), which can extract electrons and holes from the light harvesting layers, respectively, are essential for the high-efficiency PSCs. Most of the high performance, PSCs were accomplished using compact TiO 2 layer or mesoporous TiO 2 layer as the ETLs [8,16]. However, the processes of both the compact TiO 2 layer and the mesoporous TiO 2 layer generally require a high sintering temperature (>450°C), which is an obstacle for the stretchable device fabrication and the commercial development of PSCs [17,18]. Previously, SnO 2 has shown up as an effective electron transport layer in perovskite solar cells due to the wider band gap (about 3.6 eV) and higher mobility (100 to 200 cm 2 V −1 s −1 ) compared with TiO 2 [19][20][21]. Furthermore, the temperature of forming SnO 2 thin films (<200°C) is helpful for widespread implementation and large-scale commercialization of PSCs [22]. Therefore, SnO 2 is a promising candidate for ETLs used in high-performance PSCs.
It has been reported that doping ETLs with metal aliovalent cations are an effective method to improve properties of both ETLs and ETLs/perovskite interfaces. Other groups have already doped Y 3+ and Li + in SnO 2 to improve carrier transport and optical abilities [23,24]. In addition, doping can also improve the conductivity of ETLs and optimize the energy level matching between ETLs and the perovskite film, resulting in the enhanced performance of device [23].
Here, we present a low-temperature synthesized Aldoped SnO 2 as an ETL in the n-i-p structure perovskite solar cells. Al-doped SnO 2 thin films prepared at a low temperature (190°C) show a better charge transport and electron extraction behavior than the pristine SnO 2 . The enhanced electrical property of SnO 2 improves the PCE, V OC , J SC , and fill factor (FF) of the PSCs. The champion cell based on Al-doped SnO 2 reaches a PCE of 12.10% with a V OC of 1.03 V, a J SC of 19.4 mA/cm 2 , and a FF of 58%, while the PSCs based on undoped SnO 2 achieves a PCE of 9.02% with a V OC of 1.00 V, a J SC of 16.8 mA/ cm 2 , and a FF of 53%.

Methods
The fluorine-doped tin oxide (FTO) glass substrates were sequentially cleaned with acetone, ethyl alcohol, and deionized (DI) water in the ultrasonic bath for 15 min. Then the substrates were dried by a N 2 flow and further cleaned by UV-ozone for 10 min before the deposition of SnO 2 .
SnO 2 and Al-SnO 2 ETLs were deposited by a spincoating method. The solution was prepared by dissolving SnCl 4 •5H 2 O in isopropyl alcohol at a concentration of 0.075 M and subsequently stirred for 60 min at room temperature. For Al-doping, we dissolved AlCl 3 •6H 2 O in isopropyl alcohol. Then this aluminum precursor was added to the antecedent solution at a series of molar ratio and stirred until the solution became clear. Afterwards, the two different kinds of solution were separately deposited on cleaned FTO substrates at 3000 rpm for 30 s. The substrates were then pre-dried at 100°C for 10 min and annealed at 190°C for 1 h.
After the deposition of SnO 2 and Al-SnO 2 electron transport layers, the samples were treated by UV-ozone for 10 min again. The CH 3 NH 3 PbI 3 film was fabricated by a one-step spin-coating method. The CH 3 NH 3 PbI 3 solution (45 wt%) was deposited on the treated SnO 2 at 5000 rpm for 30 s. 0.5 mL chlorobenzene was dropped onto the substrate when spin-coating the CH 3 NH 3 PbI 3 solution. All the perovskite layers were annealed at 100°C for 10 min. The hole transport layers were deposited by spin-coating the 2,29,7,79-tetrakis(N,Ndi-p-methoxyphenylamine)-9,9-spirobifluorene (spiro-OMeTAD) solution at 4000 rpm for 30 s. Finally, 100-nm-thick gold top electrode was deposited on the HTL via thermal evaporation.

Device Characterization
The J-V characteristics of perovskite solar cells were measured by Keithley 2400 source measuring unit under the AM 1.5 G solar-simulated light (Newport Oriel Solar 3 A Class AAA, 64023 A). The sun light (100 mw/cm 2 ) was calibrated by a standard Si-solar cell (Oriel, VLSI standards). X-ray photoelectron spectroscopy (XPS) was measured using the Kratos XSAM 800 X-Ray Photoelectron Spectrometer.

Result and Discussion
In our work, SnO 2 layers have been deposited on FTO substrates by a low-temperature solution method. Top view scanning electron micrographs (SEM) of the SnO 2 and bare FTO are shown in Fig. 1a-b. A dense and pinhole-free film is formed by spin coating SnO 2 ETLs solution on FTO substrates, suggesting that the FTO substrates have been fully covered. Dense ETLs are known as an essential element of high-performance PSCs. Thus, a compact SnO 2 layer deposited on FTO can enhance the interfacial contact with perovskite layers and improve the performance of the solar cells [25].
To examine the efficiency of PSCs based on the lowtemperature solution-processed SnO 2 and Al-SnO 2 ETLs, we have fabricated the planar solar cells with the structure of FTO/(Al-)SnO 2 /MAPbI 3 /Spiro-OMeTAD/Au shown in Fig. 1d. In addition, Fig. 1c shows a cross-sectional SEM image of a PSC based on Al-doped SnO 2 layer without the Au electrode, and the energy band diagram of PSCs is shown in Fig. 1e.
To confirm that the employment of Al-doping SnO 2 as ETLs has no effect on the perovskite films, we measured the UV-vis absorbance spectra and the corresponding X-ray diffraction (XRD) of the MAPbI 3 film on FTO/SnO 2 and FTO/Al-SnO 2 substrates. The results are shown in Fig. 2a and b, respectively. The MAPbI 3 films deposited on SnO 2 and Al-SnO 2 ETLs exhibit almost identical absorbance spectra, suggesting that the absorption of the perovskite films is nearly not affected by Al-doping in SnO 2 ETLs.
As to the XRD patterns, several strong peaks are located at 14.05, 23.44, 24.25, 28.18, 31.88, 34.93, and 40.16°. All these peaks can be assigned to orthorhombic crystal of the perovskite with high crystallization [26][27][28]. The XRD patterns show negligible difference between the samples of FTO/SnO 2 /MAPbI 3 and FTO/Al-SnO 2 /MAPbI 3 , indicating the dopant of Al in SnO 2 film does not affect on the structure property of MAPbI 3 film. Furthermore, the main PbI 2 peak is absent from the XRD patterns, which indicates PbI 2 has sufficiently reacted with MAI.
As the evidences were obtained from the UV-vis absorbance spectra and XRD patterns of the devices, the dopant of Al in SnO 2 does not cause any obvious changes on the structure and optical properties in the perovskite layers. Therefore, the performance enhancement induced by Al-doping in SnO 2 as ETLs is most likely due to the improvement of the ETLs/perovskite interfacial properties. In other words, the charge transport and electron extraction are improved. Figure 2c illustrates the external quantum efficiency (EQE) spectrum of the best-performance solar cells based on SnO 2 and Al-SnO 2 . Obviously, the EQE of Al-doped device is higher than the device based on pristine SnO 2 over the entire wavelength range. The higher EQE means superior electron extraction capability of the ETLs [29]. The calculated J SC (≈19.0 mA/cm 2 ) based on Al-SnO 2 from the EQE spectra is consistent with the measured value of the current density-voltage (J-V) curves measured under the one-sun light. As for undoped SnO 2 , the calculated J SC is approximately equal to 16.6 mA/cm 2 .
The J-V curves of the best-performance PSCs based on SnO 2 and Al-SnO 2 ETLs are shown in Fig. 2d. The PCE increases from 9.02 to 12.10% by doping SnO 2 with Al. Al-doping may cause an improvement on the charge transport and electron extraction behavior of the SnO 2 , leading to the increment of the J SC (16.8 to 19.4 mA/cm 2 ). Furthermore, the V OC of the best PSC based on Al-SnO 2 (1.03 V) is a little higher than that of the best cell based on SnO 2 (1.00 V), indicating less energy loss of electrons [30]. Therefore, the enhanced parameters mentioned above leads to the improvement of FF (53 to 58%).
Al-SnO 2 films deposited by a low-temperature solutionprocessed were further investigated by X-ray photoemission spectroscopy (XPS). Figure 3a displays the full XPS spectrum, which shows the presence of O, C, and Sn. The binding energies of 487.3 and 495.8 eV shown in Fig. 3b corresponds to Sn 3d 5/2 and Sn 3d 3/2 , respectively. The main binding energy of 531.0 eV shown in Fig. 3c corresponds to O 1 s, which reveals the O 2− state in SnO 2 [21]. The absence of Al in the full XPS spectrum can be attributed to the low concentration (5% molar ratio), while Al 2p peak can be observed in Fig. 3d with a relatively low content, indicating that the doping is truly practicable. Identically, the Cl 2p peak missed in the full XPS spectrum can also be observed in Fig. 2e. The lowcontent Cl suggests that both most of SnCl 4 and AlCl 3 have been oxidized.
To identify the reasons for the enhanced performance due to the Al-doping, we carried out the time-resolved photoluminescence (TRPL) to study the electron-extraction    Fig. 3f, are exponentially fitted, where τ1and τ2 represent the bulk recombination in perovskite bulk films and the delayed recombination of trapped charges, respectively [31]. For the FTO/SnO 2 /MAPbI 3 sample, τ1 is 1.07 ns and τ2 is 7.98 ns, while τ1 is 1.32 ns and τ2 is 5.13 ns for the doped SnO 2 sample (1% Al-doping content). Apparently, the perovskite film deposited on the Al-SnO 2 ETL has a lower τ2 with a lower ratio of τ2/τ1, indicating a better change transfer from perovskite to ETLs and more efficient extraction of the photo-induced electrons between the perovskite and ETLs, as compared to the film deposited on the pristine SnO 2 ETL [32,33]. In addition, the decay curves also confirm the remarkable enhancement of the electron extraction and charge transport induced by Al-doping in SnO 2 . These properties result in the improvement of current density and the power conversion efficiency.
We also compared four different parameters of the cell performance with a series of doping concentration. From the box charts in Fig. 4a, it is obvious that the PCE of the cells is strongly influenced by Al doping. The average PCE of the cells is improved with the increment of Al content before the concentration of 1%, while the average PCE is reduced with the higher Al content (3 and 5%). The change of J SC of these solar cells is shown in Fig. 4b, and the variation tendency is like the trend of PCE. The highest J SC is 23.82 mA/cm 2 , which confirms a good charge transportation of the cells. Regarding the change of V OC , Fig. 4c shows the variation of V OC , value is smallest with 1% Al-doping content. The results demonstrate that the solar cells with 1% Al-doping exhibit the best repeatability. As exhibited in Fig. 4d, the change tendency of FF is analogous to the trend of PCE. In addition, the average FF of the solar cells doped with 0.5 and 1% Al 3+ content is higher than the undoped solar cells impressively. The results mentioned above reveal that a suitable Al-doping is beneficial for the performance of the perovskite solar cells based on SnO 2 .

Conclusions
In summary, we studied the effect of Al-doping on SnO 2 as ETLs for planar perovskite solar cells. According to the results of UV-vis absorbance spectra and XRD patterns of perovskite films deposited on Al-SnO 2 and SnO 2 , the Al dopant in SnO 2 does not influence the structure and optical properties of the perovskite layers. Based on the TRPL test, the Al-dopant in SnO 2 enhances the charge transport and electron extraction behavior of the PSCs and then the J SC of the devices is improved. The champion cell based on Al-SnO 2 exhibited a higher efficiency of 12.10% than that using SnO 2 (9.02%) as ETLs. Our results suggest that efficient planar perovskite solar cells based on SnO 2 can be fabricated by doping SnO 2 with Al 3+ .

Funding
This work was supported by the National Natural Science Foundation of China under grant nos. 61421002, 61574029, and 61371046. This work was also partially supported by University of Kentucky.
Authors' Contributions HZ, DL, and FW designed and carried out the experiments. CW, TZ, PZ, and HS participated in the work to analyze the data and prepared the manuscript initially. SL and ZC gave equipment support. All authors read and approved the final manuscript.