Oblique Electrostatic Inkjet-Deposited TiO2 Electron Transport Layers for Efficient Planar Perovskite Solar Cells

In this study, a new, simple, and novel oblique electrostatic inkjet (OEI) technique is developed to deposit a titanium oxide (TiO2) compact layer (CL) on fluorine-doped tin oxide (FTO) substrate without the need for a vacuum environment for the first time. The TiO2 is used as electron transport layers (ETL) in planar perovskite solar cells (PSCs). This bottom-up OEI technique enables the control of the surface morphology and thickness of the TiO2 CL by simply manipulating the coating time. The OEI-fabricated TiO2 is characterized tested and the results are compared with that of TiO2 CLs produced by spin-coating and spray pyrolysis. The OEI-deposited TiO2 CL exhibits satisfactory surface coverage and smooth morphology, conducive for the ETLs in PSCs. The power-conversion efficiencies of PSCs with OEI-deposited TiO2 CL as the ETL were as high as 13.19%. Therefore, the present study provides an important advance in the effort to develop simple, low-cost, and easily scaled-up techniques. OEI may be a new candidate for depositing TiO2 CL ETLs for highly efficient planar PSCs, thus potentially contributing to future mass production.

thermally decompose simultaneously to form the TiO 2 CL 20,21 . High-quality ETLs can be produced through either SC or SP top-down techniques, but these are very sensitive to the control parameters, and hence the PCE of such fabricated PSCs may differ significantly, even for TiO 2 CLs made by the same procedure. In addition, sputtering 22 and CVD 23 require a vacuum environment and have a slow deposition rate, which present challenges for producing TiO 2 thin films. Scalable bottom-up ALD 24 requires a relatively long time and is rather costly for preparing TiO 2 thin films. In addition, although CBD 25 and ED 26 enable TiO 2 layers to be produced at low temperature, it is quite difficult to control the morphology and thickness of the TiO 2 CL, thus limiting its reproducibility. Therefore, fabricating high-quality, reproducible TiO 2 CLs for the efficient performance of planar PSCs via a scalable, controllable, and cost-effective technique remains a significant challenge.
Electrostatic inkjet deposition (EI) is a bottom-up strategy that produces TiO 2 film by discharging in the form of a spray via electrostatic force. In comparison to other bottom-up techniques, EI offers a cost-effective, simple, and promising way to obtain high-quality TiO 2 film with an easy-to-control thickness as well as large area, multi-stack thin films with high reproducibility. Previously, Umezu et al. 27 reported efficient dye-sensitized solar cells (DSSC) with a PCE of 6.6% by using EI-deposited thicker TiO 2 thin films. Conventional EI suffers from large droplets that are stacked on the FTO substrate due to gravity. To remove the large droplets, this work modifies conventional EI such that thinner TiO 2 film can be patterned by changing the substrate position (an ejection angle of 45° with respect to the substrate), which is known as oblique electrostatic inkjet (OEI) deposition. Although it may be implied that OEI enables the fabrication of a smooth surface morphology while allowing for easy control of the TiO 2 CL thickness, both of which are required for enhancing the performance of PSCs, it remains a challenge to use OEI-deposited TiO 2 CLs in planar PSCs to achieve good electron injection and extraction.
In this study, we report for the first time a high quality and easily controllable OEI technique that enables patterning of a TiO 2 CL on the FTO substrate (without the need for a vacuum environment) for efficient planar PSCs. The morphology and optimum thickness of TiO 2 CL can be controlled in this bottom-up technique by simply manipulating the coating time. Herein, we test and compare the effectiveness of SC-, SP-, and OEI-deposited TiO 2 CLs s ETLs in PSCs.

Results and Discussion
Higher potential TiO 2 paste was patterned for use in a DSSC by using conventional EI, as described by Umezu et al. 27 This technique has proven to be a very promising technology for printing at high resolution and its capability to eject TiO 2 from highly viscous liquid. Conventionally, EI discharges TiO 2 paste in the form of a spray via electrostatic force in the direction perpendicular to the FTO substrate, resulting in large droplets stacked among the thick layers on the resultant substrate due to the gravity acting upon the droplet. The creation and presence of such droplets present limitations in conventional EI to produce thinner and smooth TiO 2 film for the next generation of PSCs. To overcome this, we modify conventional EI by changing the substrate position to 45° with respect to the substrate, which results in the deposition of high quality TiO 2 CLs whose thickness is easily controlled. An ejection angle of 45° with respect to the substrate was optimized for smooth morphology and full surface coverage. The CLs prepared by OEI are implemented as ETLs in planar PSCs. Figure 1a shows the schematic illustration of the OEI experimental setup for pattering TiO 2 CLs. On the lower left, inset a photograph of the precursor shows it is a transparent yellow-orange TiO 2 solution (concentration of 0.30 M), and the photograph in the upper center of the illustration shows the drop ejection of the spray mode. Overall, one can see the TiO 2 layer deposition mechanism based on the OEI setup and phenomena. The deposition of large droplets is avoided as these are removed by gravity due to the angle of the spray nozzle. A smooth TiO 2 CL was deposited based on the electrostatic force. The OEI technique produces TiO 2 CLs with a smooth morphology and allows the thickness to be controlled by simply optimizing the coating time. Figure 1b reveals a schematic illustration of the device configuration of OEI-TiO 2 CL-based PSCs.
Top-view scanning electron microscope (SEM) images of bare FTO and samples SC-TiO 2 CL, SP-TiO 2 CL, OEI-TiO 2 CL-60 sec, and OEI-TiO 2 CL-30 + 30 sec are shown in Fig. 2a-e, respectively. The bare FTO grains show are rough and range from tens to hundreds of nanometers in size (Fig. 2a). From Fig. 2b, it can be seen that SC-deposited TiO 2 CL (60 nm thickness) covers smoothly the entire FTO substrate. In addition, as shown in Fig. 2c, the TiO 2 CL (Optimum thickness of 70 nm) deposited by using SP-holds few cracks (indicated by blue circles), which correlates to the surface of FTO-substrate, followed by linking between FTO and perovskite and facilitates to crucial charge recombination. As shown in Fig. 2d, the OEI-deposited TiO 2 CL feature a denser smooth scaffold for 60 sec (referred as OEI-TiO 2 CL-60 sec). The optimized thickness of the OEI-TiO 2 CL was 100 nm. Some cracks (specified by the white rectangle) occur in the OEI-TiO 2 CL-60 sec, which may correlate to the FTO grain boundaries, providing possible links between the FTO and perovskite and facilitating extensive charge recombination. To overcome such cracks, the coating time of the OEI-deposited TiO 2 CL is changed (referred as OEI-TiO 2 CL-30 + 30 sec), resulting in a uniform, dense scaffold that satisfactorily covers the surface and contains no visible cracks (Fig. 2e). The smooth OEI-TiO 2 CL-30 + 30 sec samples offer more efficient charge separation and recombination rates, owing to the smooth and entirely crack-free surface coverage. This newly developed OEI technique enables precise control of the morphology, thickness, and deposition rate of the materials for patterning TiO 2 CLs without a vacuum environment. A cross-sectional SEM image of the OEI-TiO 2 CL-30 + 30 sec film is shown in Fig. 2f, which was used to measure the TiO 2 CL thickness.
The surface morphology of perovskite on the OEI TiO 2 CL is shown in Fig. 3a. The homogeneous surface morphology with large grains induce fewer grain boundaries and thus fewer traps and charge carrier losses. Generally, trap states generated at grain boundaries should be minimized for efficient PSC operation 28 . A cross-sectional SEM image of a complete PSC device made with the OEI-TiO 2 CL-30 + 30 sec sample as the ETL is shown in Fig. 3b.
The photovoltaic performance of PSCs processed with SC-TiO 2 CL, SP-TiO 2 CL, OEI-TiO 2 CL-60 sec, and OEI-TiO 2 CL-30 + 30 sec is investigated and demonstrated. The J-V characteristics of the best-performing PSC devices are presented in Fig. 4a. The corresponding J-V parameters are summarized in Table 1. A comparison of the forward scan (FS) and reverse scan (RS) results are summarized in Table S1. Outline of the OEI-TiO 2 CL based PSCs performance characteristics with different concentration of the TiO 2 precursor solution (Table S2). The thickness of the OEI-TiO 2 CL was tuned by changing (from 0.05 to 0.50 M) precursor solution concentration of the TiO 2 and fabricated the resultant devices, as shown in Fig. S1. Ultimately, optimized and PSCs with 0.30 M    www.nature.com/scientificreports www.nature.com/scientificreports/ that has no visible cracks, thus facilitating more efficient carrier flow, as compared to that of the PSC made using the OEI-TiO 2 CL-60 sec sample. In contrast, the OEI-TiO 2 CL-30 + 30 sec-based PSCs exhibit PCEs as high as 13.19%, which is also better than that of PSCs formed using the SC-TiO 2 CL, SP-TiO 2 CL, and OEI-TiO 2 CL-60 sec samples (10.27%, 12.19%, and 7.86%, respectively). The OEI-TiO 2 CL-30 + 30 sec deposited-film strongly influences the TiO 2 CL morphology, which significantly affects the resulting solar cell performance. However, the device characteristic J-V curves have hysteresis in all single-layer TiO 2 CLs, which underestimates the actual performance of the PSCs (Table S1). In general, the hysteresis phenomenon impacted the rich number of both oxygen vacancies and electron traps on the surface of TiO 2 30 . The incident photon-to-electron conversion efficiency (IPCE) is measured to further reveal the correctness of the output J sc , as shown in Fig. 4b. The IPCE coverage also significantly improved, especially in the range of ~380 nm to ~700 nm, with the OEI-TiO 2 CL-30 + 30 sec spans compared to the counterparts. This significantly evidence the improved charge collection efficiency of OEI-TiO 2 CL-30 + 30 sec-based perovskites, which could be solely related to high surface coverage (less shorting paths). The lower coverage for the devices made with SC-TiO 2 CL and SP-TiO 2 CL indicates that the perovskite quality and its interfacial connection with charge transport layers is poor. Figure 5a-d show the statistical distribution of the four main parameters of the PSCs devices (J sc , V oc , FF, and PCE) as a function of the type of TiO 2 CL as the ETL. To ensure the performance reproducibility of our results, we fabricated multiple PSCs with SC-TiO 2 CL (9 devices), SP-TiO 2 CL (15 devices), OEI-TiO 2 CL-60 sec (11 devices), and OEI-TiO 2 CL-30 + 30 sec (14 devices) as ETLs. The best control and highest reproducibility in terms of all the parameters are obtained from the uniform and full surface coverage OEI-deposited TiO 2 CL-for 30 + 30 sec film. This implies that the OEI technique enables the control of the surface morphology and thickness of the TiO 2 CL by simply modifying the coating time, which is crucial in inhibiting carrier recombination and enhancing the interface between the TiO 2 CL and perovskite of the PSCs.

conclusions
For the first time, we applied a simple, novel, and promising OEI technique that does not require a vacuum environment to fabricate uniform TiO 2 CL on FTO substrate; subsequently, we use the film as the ETL in planar PSCs. We tuned the surface morphology of the film OEI-TiO 2 by simply changing the coating times to achieve a smooth and denser scaffold covering the entire substrate without visible cracks. The PSCs made with the optimized film as the ETL showed enhanced efficiency up to 13.19%. Therefore, OEI-deposited TiO 2 CLs are candidates for use as ETLs in planar PSCs. Owing to further enhance the performance, the OEI-TiO 2 CL-based PSCs, we suggest that surface engineering or elemental doping in the TiO 2 CL will ensure a good interface between the TiO 2 CL and perovskite, reducing the hysteresis. www.nature.com/scientificreports www.nature.com/scientificreports/ In order to develop large-scale printing through OEI-approach, we will further apply multi-nozzle system as the single nozzle approach will be covered only 5 mm width. When each nozzle is set closely, OEI-approach will not be stable, and uniform printing will not be observed due to cross-talk of electric field around each nozzle. As the shield around each nozzle is set, the cross-talk will not be generated, and uniform and large-size printing will be achieved. However, by benchmarking our results, OEI-approach will be thoroughly explored to coat variety of substrates on the large scale for high-throughput large-area perovskite solar modules in our upcoming study. The rapid development toward scaling-up and large-area uniformity of OEI-substrates, coupled with low-cost fabricating capability that has a high material utilization rate and might reduce of production cost of resultant devices. This facile fabrication technique may be a new TiO 2 CL ETL deposition candidate that helps to further improve the photovoltaic performance of planar small-scale and large-PSCs modules. experimental Materials. FTO-patterned glass substrates were purchased from Asahi glass (Tokyo, Japan). Lead iodide (PbI 2 ), methylammonium bromide (MABr), lead bromide (PbBr 2 ), formamidinium iodide (FAI), and cesium iodide (CsI) were purchased from Tokyo Chemical Industry (Tokyo, Japan). The N,N-dimethylformamide (DMF; 99.5% purity) and dimethyl sulfoxide (DMSO; 99.5% purity) were purchased from Wako Pure Chemical (Osaka, Japan). Titanium diisopropoxide bis(acetylacetonate) (75 wt% in isopropanol) and 2-isopropanol (99.9% purity) were supplied from Sigma-Aldrich, Wako Pure Chemical, respectively. fabrication of oei-deposited tio 2 cL. The FTO-patterned glass substrates were cut into 25 mm squares and washed consecutively with a soap solution, distilled water, acetone, ethyl alcohol, and once more distilled water. Then, the UV ozone treatment was applied to FTO substrates for 15 min. The tip of a Terumo syringe (10 ml) was fitted with a San-ei Tech Ltd (Tokyo, Japan) TT taper nozzle (0.21 mm internal diameter), and the solution was filled up to 2.0 ml with the TiO 2 solution. A clear yellow-orange precursor solution [from 0.05 to 0.50 M titanium diisopropoxide bis(acetylacetonate) in isopropanol], was filled into the ink tank. The optimized precursor solution of the OEI-TiO 2 CL was 0.30 M. The nozzle was installed at the end of the tank. To maintain the electric field around the nozzle tip, a holed plate electrode (outside 100 × 145 mm 2 , hole diameter of 50 mm) was used between the nozzle tip and the FTO electrode. The FTO glass electrode was set on the XY linear stage that can be rotated. The gap between the FTO substrate and the tip of the nozzle was fixed at 60 mm, and the discharge time was 60 s. The applied voltage was 8.5 kV, the nozzle angle was 45°, and the substrate position was set in the vertical direction. The rotation speed was 2000 rpm. After depositing, the substrates were dried at 105 °C for 5 min, came after by sintering at 450 °C for 30 min in a muffle furnace and the resultant samples were used as ETLs. The OEI-deposited TiO 2 CL surface coverage was evaluated for two different deposition conditions based on the duration of the discharge; samples made with 60 seconds and 30 + 30 seconds are referred to as OEI-TiO 2 CL-60 and OEI-TiO 2 CL-30 + 30, respectively. The OEI-TiO 2 CL-30 + 30 sample preparation procedure as follows. First, TiO 2 precursor solution was patterned by OEI for 30 seconds as usual. Then, the as-deposited substrate was dried at 105 °C for 5 min and allowed to cool down to room temperature. After that, the TiO 2 precursor solution was further deposited for 30 seconds onto the existing film. After the deposition, the substrates were dried at 105 °C for 5 min, followed by sintering at 450 °C for 30 min. Again, the samples were cooled slowly to room temperature and subsequently used as ETLs.
fabrication of pScs. The perovskite precursor solution with Cs 5% was made by combining with FA/MA in DMF and DMSO with the ratio of 4:1 mixed solvents. The descriptive solution preparation and spin-coating process is described elsewhere 21 . The FAI (1 M), PbI 2 (1.1 M), MABr (0.2 M), and PbBr 2 (0.2 M) were mixed in anhydrous DMF: DMSO (4:1). Since then, the Cs 5% solution was added in the perovskite precursor solution to achieve the Cs based perovskite precursor solution. As deposited samples were annealed at 100 °C for 60 min. For the hole-transport layer (HTL), the detailed precursor preparation is described elsewhere 21 . At last, vacuum deposited Au (100 nm) was formed as an electrode on the HTL to complete the device. characterization. Field emission scanning electron microscopy (S-4800, Hitachi High-Tech, Tokyo, Japan) was used to examine the surface morphology. The current density versus voltage (J-V) curves were measured and analyzed with the simulated solar conditions (100 mW⋅cm −2 , AM 1.5, 1 sun intensity) utilizing a Keithley 2401 digital source meter. A monochromatic xenon arc light scheme (Bunkoukeiki, SMI-250JA) was used to measure the IPCE of the resultant devices. The 0.09 cm 2 was the active area of the fabricated devices.