Wetting-induced formation of void-free metal halide perovskite films by green ultrasonic spray coating for large-area mesoscopic perovskite solar cells

A void-free metal halide perovskite (MHP) layer on a mesoscopic TiO2 (m-TiO2) film was formed via the wetting-induced infiltration of MHP solution in the m-TiO2 film via a green ultrasonic spray coating process using a non-hazardous solvent. The systematic investigation of the behavior of ultrasonic-sprayed MHP micro-drops on the m-TiO2 film disclosed that the void-free MHP layer on the m-TiO2 film can be formed if the following conditions are satisfied: (1) the sprayed micro-drops are merged and wetted in the mesoscopic scaffold of the m-TiO2 film, (2) the MHP solution infiltrated into the m-TiO2 film by wetting is leveled to make a smooth wet MHP film, and (3) the smooth wet MHP film is promptly heat treated to eliminate dewetting and the coffee ring effect by convective flow in order to form a uniform void-free MHP layer. A void-free MHP layer on the m-TiO2 film was formed under optimal ultrasonic spray coating conditions of substrate temperature of ∼30 °C, spray flow rate of ∼11 mL h−1, nozzle to substrate distance of ∼8 cm, and MHP solution-concentration of ∼0.6 M under a fixed scan speed of 30 mm s−1 and purged N2 carrier gas pressure of 0.02 MPa. The mesoscopic MHP solar cells with an aperture area of 0.096, 1, 25, and 100 cm2 exhibited 17.14%, 16.03%, 12.93%, and 10.67% power conversion efficiency at 1 sun condition, respectively.


Introduction
][12][13][14] Among them, the MHP SCs such as CH 3 NH 3 PbX 3 (MAPbX 3 : X ¼ Cl, Br, I, and their mixture), CH(NH 2 ) 2 PbX 3 (FAPbX 3 ), CsPbX 3 , and their mixtures have attracted signicant attention because of their excellent properties such as strong absorptivity due to the direct bandgap, high open-circuit voltage due to the small exciton binding energy, convenient bandgap tunability by structure and composition engineering, and the mild/low temperature solution processability. 15][18] Besides the development of highly efficient MHP SCs, it is also important to develop a scalable coating process for the deposition of large-area MHP thin lms for commercial applications as high-performance MHP SCs are fabricated via the spin coating process.Therefore, slot-die coating, 19,20 blade coating, [21][22][23] bar coating, [24][25][26] spray coating, 27 and ink-jet printing processes 28,29 have been developed for the deposition of MHP thin lms.Among them, the spray coating process is a unique process combining the advantage of a solution process and vapor deposition process and can form conformal MHP thin lms even on rough surfaces with a stoichiometrically dened MHP solution.In addition, spray coating can also be used to form certain patterns by spraying MHP solution on a substrate with a patterned mask.Hence, the spray coating process, including air brush, [30][31][32] ultrasonic spray, 33,34 megasonic spray, 35 and electrospray [36][37][38] coating, has been studied intensively as a scalable process to replace the conventional spin coating process.
For instance, Heo et al. demonstrated 18.3% MAPbI 3Àx Cl x MHP SCs composed of F-doped tin oxide (FTO)/dense TiO 2 (d-TiO 2 )/MAPbI 3Àx Cl x /poly(triarylamine) (PTAA)/Au by depositing an MHP thin lm via air brush coating. 32Park et al. reported 16.9% MAPbI 3 MHP SCs constructed using indium tin oxide (ITO)/poly (3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS)/MAPbI 3 /C 60 /bathocuproine (BCP)/Cu via the megasonic spray coating process. 35Su et al. fabricated 13.5% MAPbI 3Àx Cl x MHP SCs composed of ITO/d-SnO 2 /MAPbI 3Àx Cl x / 2,2 0 ,7,7 0 -tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9 0 -spirobi-uorene (spiro-MeOTAD)/Au via an ultrasonic spray coating process under air blowing. 34Hong et al. demonstrated 13.27% ITO/PEDOT:PSS/MAPbI 3 /C 60 /BCP/LiF/Al via an electrospray coating process, which could control the size of the MHP solution droplets through the modulated applied electric eld. 37hus far, spray coating has been performed on dense at substrates such as d-TiO 2 , d-SnO 2 , and PEDOT:PSS, which do not have any issue of pore lling because voids are oen formed in the m-ETL/MHP layer due to the incomplete inltration of MHP drops during the spray coating process. 38However, the m-ETL such as an m-TiO 2 layer is important to reproducibly fabricate high performance MHP SCs with small current density-voltage (J-V) hysteresis because it prevents direct contact between the hole transporting material/Au and d-TiO 2 /FTO electrode.However, pinholes are formed in the MHP layer, which reduce the J-V hysteresis with respect to the scan direction due to the extended interface area balancing electron and hole ux, once the sprayed MHP micro-drops are wetted in the m-TiO 2 ETL during the process.Recently, Heo et al. reported that the problem of void formation in m-TiO 2 due to the incomplete inltration of the MHP lm during the electrospray coating process can be solved by using a vertically aligned TiO 2 nanorod electrode with straight macropores. 38][40] Herein, we used an ultrasonic spray coating process because it formed uniform MHP solution micro-drops.To achieve the perfect inltration of the MHP solution micro-drops sprayed on top of the m-TiO 2 ETL, we attempted to determine the optimum wetting conditions via systematic control of the ultrasonic spray processing parameters such as temperature, ow rate, nozzle height, and solution concentration.Consequently, we fabricated dense MHP thin lms without voids via the wettinginduced perfect inltration of the MHP solution micro-drops in the m-TiO 2 ETL.In addition, we only used the nonhazardous g-butyrolactone (GBL) solvent for the green ultrasonic spray coating process because the solvent in the atomized MHP solution micro-drops from the ultrasonic sprayer are more quickly evaporated than that in the bulk MHP solution, and thus it is desirable to avoid the use of hazardous solvents such as N-methyl-2-pyrrolidone, N,N-dimethylformamide, dimethylacetamide, and dimethyl sulfoxide. 41

Preparation of MAI and MACl powder
Briey, 35 mL of methylamine solution was charged in a roundbottom ask (RBF), which was immersed in an ice bath.Then, 50 mL of HI solution in a dropping funnel was slowly dropped into the RBF.Aer reacting for 1 h, the color of the solution changed to a yellowish color.The solution was magnetically stirred during the entire reaction.To collect the MAI powder, the product solution was dried using a rotary evaporator.To obtain pure MAI, the dried MAI powder was recrystallized by dissolution in ethanol and consecutive precipitation in diethyl ether.The recrystallized MAI powder was fully dried in a vacuum oven at 60 C for 24 h.MACl was also synthesized following the above procedure except HCl solution was used instead of HI solution.

Device fabrication
To fabricate mesoscopic MAPbI 3Àx Cl x MHP SCs via the ultrasonic spray coating process, we deposited a 50 nm-thick d-TiO 2 ETL on a patterned FTO (Pilkington, TEC7) substrate via the spray pyrolysis deposition of titanium diisopropoxide bis(acetylacetonate)/ ethanol (1/9 vol/vol) solution at 500 C with a hand brush (DH-125, Sparmax).Then an m-TiO 2 paste/ethanol (1/5 wt/wt) solution was spin coated on the FTO/d-TiO 2 substrate at 5000 rpm (revolution per min) for 30 s.The spin-coated FTO/d-TiO 2 /m-TiO 2 lm was then calcined at 500 C for 1 h.For the deposition of the wetting-induced void-free MAPbI 3Àx Cl x MHP lm on the m-TiO 2 layer, we prepared MAPbI 3Àx Cl x MHP solutions by dissolving MAI : PbI 2 : PbCl 2 : MACl (1 : 0.95 : 0.05 : 0.1 mole ratio) powder in GBL solvent to obtain a specic concentration (0.2, 0.4, 0.6, 0.8, and 1.0 M).For the systematic studies to determine the wetting condition of MHP solution, we controlled the processing parameters such as substrate temperature (20 C, 30 C, and 60 C), spray ow rate of MHP solution (7, 9, 11, 13, and 15 mL h À1 ), nozzle to substrate distance (4, 6, 8, 10, and 12 cm), and concentration of MHP solution (0.2, 0.4, 0.6, 0.8, and 1.0 M).Then, the MHP solution was deposited on the temperature-controlled FTO/d-TiO 2 / m-TiO 2 substrate using an ultrasonic spray coater (S120, CERA-TORQ, 120 kHz) combined with a 3D printer (Creality, CR-10S Pro), where the MHP solution was delivered by a syringe pump (KD-Scientic, KDS100).Aer spraying the MHP solution, we waited for 25 s to ensure the wetting of the MHP solution on the m-TiO 2 lm and heat treated it on a hot plate pre-heated to 120 C for 2 min.The scan speed of the ultrasonic sprayer and purged N 2 carrier gas pressure for the sprayer was xed to 30 mm s À1 and 0.02 MPa, respectively.We then spin coated a PTAA/toluene (15 mg mL À1 ) solution with 7.5 mL Li-TFSI/acetonitrile (170 mg mL À1 ) and 7.5 mL t-BP/acetonitrile (1 mL mL À1 ) on the FTO/d-TiO 2 /m-TiO 2 / MHP substrate at 3000 rpm for 30 s. Finally, the Au counter electrode was deposited on the FTO/d-TiO 2 /m-TiO 2 /MHP/PTAA substrate via thermal evaporation.The active area of small-sized MHP SC was xed to 0.16 cm 2 .The large area MHP SCs were fabricated following our previous report. 42All experiments were conducted under an ambient air atmosphere and controlled relative humidity of $20%.

Characterization
The SEM surface and cross-sectional images were measured using a high-resolution eld emission scanning electron microscope (FE-SEM, FEI, Quanta 250, FEG).The external quantum efficiency (EQE) was measured using a power source (150 W xenon lamp, 13014, ABET) with a monochromator (MonoRa-500i, DongWoo Optron Co., Ltd.) and potentiostat (IviumStat, Ivium).The current density-voltage (J-V) curves were measured using a solar simulator (PEC-L01, Peccell) with a potentiostat (IviumStat, Ivium) under the illumination of 1 sun (100 mW cm À2 AM 1.5G), which was calibrated using a certied Si-reference cell (Japanese Industrial Standards).The J-V curves of the small MHP SCs were measured by masking the active area with a metal mask having an aperture of 0.096 cm 2 .

Results and discussion
Fig. 1(a) presents a schematic of the structure of the ultrasonic spray coater.The ultrasonic sprayer was connected to a 3D printer, which has a motorized x-y stage.Here, we deposited an MHP solution on an m-TiO 2 lm via single-pass spraying, and thus we only controlled the ultrasonic sprayer to the xdirection.The moving speed of the ultrasonic sprayer to the xdirection is dened to scan speed, which was xed at 30 mm s À1 .Accordingly, the ultrasonic spray coater can fabricate large-sized MHP solar sub-modules by multi-pass spraying.The MHP solution was fed to the ultrasonic sprayer by a syringe pump, and the feed (spray) ow rate was controlled to 7-15 mL h À1 .The MHP solution in the nozzle end was atomized by 120 kHz ultrasound and the atomized MHP solution micro-drops were sprayed using N 2 carrier gas whose pressure was xed to 0.02 MPa.The sprayed volume of MHP solution per unit area of the m-TiO 2 lm could be controlled by the spray ow rate and the concentration of the MHP solution.The coating coverage and the sprayed volume of MHP solution per unit area could be simultaneously controlled by the nozzle to substrate distance.The evaporation rate of solvent in the MHP solution sprayed on the m-TiO 2 lm was controlled by the temperature of the plate.To conrm the formation of MHP micro-drops by the ultrasonic sprayer, we calculated the size of the MHP drops (D) formed by the ultrasonic sprayer using Lang's equation. 43¼ 0.34(8ps/rf 2 ) 1/3 (1) where s, r, and f represent the surface tension of the MHP solution, density of the MHP solution, and ultrasound frequency (120 kHz), respectively.The s of the MHP solution was calculated via the pendent drop method, as shown in This journal is © The Royal Society of Chemistry 2020

RSC Advances Paper
that the MHP micro-drops are deposited on the m-TiO 2 lm via the ultrasonic spray coater as illustrated in Fig. 1(b).
To determine the proper processing parameters for the ultrasonic spray coating of MHP solution, we checked the effect of plate temperature on the wetting behavior, as shown in Fig. 2. The ultrasonic spray coating conditions are summarized in Table 1.Here, we deposited the MHP solution under a nozzle to substrate distance of 8 cm, spray ow rate of 11 mL h À1 , and concentration of MHP solution of 0.6 M as a reference condition.The front, transmitted, and back side photographs of the FTO/d-TiO 2 /m-TiO 2 /MHP substrate formed by the ultrasonic spray coater at a plate temperature of 20 C, 30 C, and 60 C are shown in Fig. 2(a).Apparently, the front and back side photographs of the 30 C and 60 C samples have uniform morphologies, whereas the 20 C sample does not have a uniform morphology due to dewetting and coffee ring effect. 44he obtained photographs provide distinctive images of the coating uniformity.Although the coating uniformity of MHP lm depends on the temperature of the plate, the UV-visible absorption spectra at different temperatures were similar, as shown in Fig. 2(b), because the mass of the deposited MHP lm was same.The optical microscopy images in Fig. 2(c)-(e) indicate that the 30 C sample has the largest MHP lm domain size and the 60 C sample has a non-uniform morphology due to the relatively quick evaporation of the GBL solvent.The corresponding scanning electron microscopy (SEM) cross-sectional images in Fig. 2(f)-(h) indicate that the 30 C sample has an MHP lm with a uniform thickness, whereas the others do not.For a better understanding of the pore lling of MHP in m-TiO 2 , the magnied SEM image is presented in the inset in Fig. 2(h).Unlike Fig. 2(f) and (g), Fig. 2(h) reveals that the m-TiO 2 layer is not fully inltrated with MHP because the sprayed micro-drops dried quickly.This imperfect inltration of MHP in the m-TiO 2 layer will seriously deteriorate the device efficiency because the generated electrons in MHP cannot be effectively transported to the m-TiO 2 ETL.In addition, the uniform thickness of the MHP layer is also very important to obtain high efficiency.The uniformity of the lm thickness is related to the wetting and dewetting phenomena and leveling of the sprayed drops.At high temperature, the sprayed drops dry quickly, and consequently form a rough MHP lm.At a low temperature, the sprayed drops are fully inltrated in the m-TiO 2 layer via a wetting and leveling process, but the wetted MHP lms are dewetted.Accordingly, a macroscopically non-uniform MHP layer is formed.To reveal the macroscopic and microscopic uniformity of the MHP lms, we obtained photographs, optical microscopic images, and SEM images.Based on the above experimental results, subsequently we xed the plate temperature to 30 C because it resulted in the formation of the most uniform morphology in the MHP lm.
Fig. 3 shows the effect of spray ow rate on the morphologies of the MHP lms during the ultrasonic spray coating process.The front, transmitted, and back-side photographs in Fig. 3(a) indicate that the uniformity of the ultrasonic-sprayed MHP lm was good at a spray ow rate of 9-11 mL h À1 , not bad at both 7 and 13 mL h À1 , and not good at 15 mL h À1 .The 7 mL h À1 sample had not sufficient volume of MHP solution to fully wet and inltrate the m-TiO 2 lm, and thus some of the sprayed MHP micro-drops le spots upon drying.The 9-11 mL h À1 samples formed uniform MHP lms because there was a sufficient volume of MHP solution to fully wet and inltrate the m-TiO 2 lms.The 13-15 mL h À1 samples had non-uniform MHP lms due to the inhomogeneous convective ow of the sprayed MHP solution by the Marangoni effect and coffee ring effect of the drying MHP solution via the pinning and depinning process.The UV-visible absorption spectra of the MHP/m-TiO 2 lms in Fig. 3(b) show that their absorptions gradually increased with an increase in the spray ow rate.The 11 mL h À1 sample had stronger absorptivity than the 7 mL h À1 sample, and thus the spray ow rate condition of 11 mL h À1 is desirable to make better MHP SCs.Their corresponding optical microscopy images in Fig. 3(c)-(g) indicate that the gradually intensi-ed contrast with an increase in the ow rate is consistent with the absorption spectra and the domain sizes increased with the ow rate.
Fig. 4 shows the effect of the nozzle to substrate distance on the morphologies of the MHP lms.The front, transmitted, and back-side photographs in Fig. 4(a) show that a uniform MHP lm was obtained only with a nozzle to substrate distance of 8 cm.The MHP lm uniformities of the 6 cm and 10 cm samples were not bad and the lm uniformity of 4 cm and 12 cm samples were not good.These results imply that the MHP lm uniformity is sensitive to the nozzle to substrate distance because the spray coated area is proportional to the square of the nozzle to substrate distance, and consequently the volume of sprayed MHP solution in a unit area has the same correlation.The UV-visible absorption spectra in Fig. 4(b) exhibit an inverse correlation between the nozzle to substrate distance and absorptivity.The corresponding optical microscopy images in Fig. 4(c)-(g) clearly indicate that the MHP domain sizes are signicantly more dependent on the nozzle to substrate distance than the other processing parameters.Similar to the absorption spectra, the MHP domain sizes were inversely proportional to the nozzle to substrate distance.
The effect of the concentration of the MHP solution on the uniformities of the MHP lms formed by the ultrasonic spray coating process was investigated, as shown in Fig. 5.The front,     6(a) and (b), respectively.Upon illumination with light, the MHP lm generates loosely bonded electron-hole pairs or free charge carriers due to its small exciton binding energy. 15,16The electrons are either injected into m-TiO 2 or transported through the MHP layer, and are then transported to the d-TiO 2 /FTO electrode. 17Simultaneously, the holes are transferred/transported to the PTAA/Au electrode.By connecting the FTO and the Au electrode with an external circuit, the electrons continuously ow through the external circuit and recombine with holes by generating electricity.
The external quantum efficiency (EQE) spectra of the mesoscopic MHP SCs prepared with different MHP solutionconcentrations of 0.2, 0.4, 0.6, 0.8, and 1.0 M are shown in Fig. 6(c).Apparently, the EQE values gradually increased as the concentration of the MHP solution increased from 0.2 to 0.6 M, and then the EQE values decreased with a further increase in the concentration of the MHP solution.The EQE value can be expressed as follows: where h lhe , h cs , and h cc represent the light harvesting efficiency, charge separation (charge injection) efficiency, and charge collection efficiency, respectively.Thus, h lhe monotonically increased with an increase in the concentration of the MHP solution, as shown in Fig. 5(b).Accordingly, the decrease in the EQE value for the 1.0 M sample can be attributed to the deteriorated h cs and/or h cc .The morphologies of the MHP lms prepared with different concentrations of MHP solution, as shown in Fig. 5 and S2, † indicate that the decline in the EQE value for the 1.0 M MHP solution concentration can be attributed to the formation of a relatively thick and rough MHP lm on m-TiO 2 owing to the quick drying of the solvent before complete wetting of the MHP solution during the ultrasonic spray coating process.The calculated short-circuit current density (J sc ) values from the integration of the EQE spectra are    2. Finally, large-area MHP SCs were fabricated and their photovoltaic properties are shown in Fig. 7.The best device PCE with an aperture area of 1 cm 2 was 16.79% under 1 sun condition and the average V oc , J sc , FF, and PCE of the 30 samples were 1.04 AE 0.02 V, 20.96 AE 0.47 mA cm À2 , 67.39% AE 0.02%, and 14.64% AE 0.92%, respectively.The MHP SCs with an aperture area of 25 and 100 cm 2 exhibited a PCE of 12.93% and 10.67%, respectively.

Conclusions
We successfully formed a void-free MHP layer on an m-TiO 2 lm via the wetting-induced perfect inltration of MHP solution in the m-TiO 2 lm during the green ultrasonic spray coating process using the non-hazardous GBL solvent.To nd the optimal process conditions for the formation of a void-free MHP layer on the m-TiO 2 lm, we controlled the substrate temperature, nozzle to substrate distance, and concentration of MHP solution under the xed scan speed and N 2 carrier gas pressure conditions of 30 mm s À1 and 0.02 MPa, respectively.Through systematic studies, we found that the possible processing windows of substrate temperature, spray ow rate, nozzle to substrate distance, and MHP solution-concentration for the formation void-free MHP lms are 30-60 C, 7-13 mL h À1 , 6-12 cm, and 0.2-0.8M, and the optimal process conditions are 30 C, 11 mL h À1 , 8 cm, and 0.6 M, respectively.Unlike the spin-coating process, the ultrasonic sprayer formed micronsized MHP solution drops, and consequently the sprayed microdrops on m-TiO 2 lm required wetting time for perfect inltration of the MHP solution in the m-TiO 2 lm.For the wetting of the sprayed MHP micro-drops, the micro-drops should merge before solidication by quick solvent evaporation.Simultaneously, the wetted MHP micro-drops were leveled by forming a smooth surface.Here, the wetted lm was heattreated in order to quench the morphology of the wetted lm.If the solvent is in excess in the wetted MHP lm, the wetted lm makes the MHP lm non-uniform owing to the dewetting and coffee ring effect by convective ow during heat treatment.If the solvent is too low in the wetted MHP lm, it quickly solidies before leveling, and consequently produces a rough MHP lm.Therefore, a void-free MHP layer on the m-TiO 2 lm could be obtained by optimizing the process conditions.Similarly the MHP SCs exhibited the best PCE of 17.14% under the optimal processing conditions.Using the optimal processing conditions, the large-area MHP SCs with an aperture area of 1, 25, and 100 cm 2 exhibited a PCE of 16.03%, 12.93%, and 10.67%, respectively.

Fig. 1 (
b) presents a schematic illustration of the morphologies of the MHP/m-TiO 2 lms formed via the conventional and new method.When the MHP solution is sprayed on the m-TiO 2 lm by the ultrasonic spray coater, initially the MHP microdrops are positioned on top of the m-TiO 2 lm.If the m-TiO 2 lm is heated to a high temperature ($120 C), the MHP microdrops on top of the m-TiO 2 lm will be quickly solidied by immediate solvent evaporation and crystallization occurs before

Fig. 1
Fig. 1 (a) Schematic structure of ultrasonic spray coater for the deposition of MHP solution and (b) schematic illustration of the morphologies of MHP/m-TiO 2 films formed by the conventional and new wetting induced infiltration method.
Fig. 2 (a) Front, transmitted (Trans.),and back side photographs, (b) UV-vis absorption spectra, (c-e) optical microscopic images, and (f-h) cross-sectional scanning electron microscopy (SEM) images of MHP/m-TiO 2 films formed by ultrasonic spray coater at different plate temperatures of 20 C (c and f), 30 C (d and g), and 60 C (e and h), respectively.

) 3 0
This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 33651-33661 | 33655 Paper RSC Advances transmitted, and back-side photographs in Fig. 5(a) indicate that the 0.4, 0.6, and 0.6 M samples have relatively more uniform MHP lms than the other samples.The 0.2 M sample had excess solvent, and consequently exhibited inhomogeneous convective ow patterns.The 1.0 M sample was quickly solidi-ed because the sprayed MHP micro-drops were easily saturated by solvent evaporation and immediately crystallized before wetting the m-TiO 2 lm.Accordingly, it produced

Fig. 6
Fig. 6 (a) Energy band diagram of MHP SC and (b) representative SEM cross-sectional image of MHP SC fabricated with 0.6 M MHP solution.(c) EQE spectra and calculated J sc and (d-l) photovoltaic properties of MHP SCs prepared with different solution concentrations of MHP.(d-h) J-V curves of the best devices prepared with (d) 0.2, (e) 0.4, (f) 0.6, (g) 0.8, and (h) 1.0 M MHP solution and (i-l) average photovoltaic parameters of 30 samples for (i) V oc , (j) J sc , (k) FF, and (l) PCE.

Fig. 7
Fig. 7 (a-e) Photovoltaic properties of MHP SCs with an aperture area of 1 cm 2 : (a) J-V curves of best device and (b-e) average photovoltaic parameters of 30 samples: (b) V oc , (c) J sc , (d) FF, and (e) PCE.(f) J-V curves of MHP SCs with an aperture area of 25 and 100 cm 2 .

Table 1
Summary of the MHP film uniformity with the processing parameters of ultrasonic spray coating (O: good, O: not bad, and x: not good)

Table 2
Summary of photovoltaic parameters of MHP SCs