Ultrasmooth Perovskite Film via Mixed Anti-Solvent ... - ACS Publications

Research Article www.acsami.org

Ultrasmooth Perovskite Film via Mixed Anti-Solvent Strategy with Improved Efficiency Yu Yu,†,‡ Songwang Yang,*,† Lei Lei,† Qipeng Cao,†,‡ Jun Shao,†,‡ Sheng Zhang,†,§ and Yan Liu*,† †

CAS Key Laboratory of Materials for Energy Conversion, Shanghai Institute of Ceramics, Chinese Academy of Sciences, 588 Heshuo Road, Shanghai 201899, P. R. China ‡ University of Chinese Academy of Sciences, Beijing 100039, P. R. China § School of Environmental and Materials Engineering, College of Engineering, Shanghai Second Polytechnic University, Shanghai 201209, P. R. China S Supporting Information *

ABSTRACT: Most antisolvents employed in previous research were miscible with perovskite precursor solution. They always led to fast formation of perovskite even if the intermediate stage existed, which was not beneficial to obtain high quality perovskite films and made the formation process less controllable. In this work, a novel ethyl ether/n-hexane mixed antisolvent (MAS) was used to achieve high nucleation density and slow down the formation process of perovskite, producing films with improved orientation of grains and ultrasmooth surfaces. These high quality films exhibited efficient charge transport at the interface of perovskite/hole transport material and perovskite solar cells based on these films showed greatly improved performance with the best power conversion efficiency of 17.08%. This work also proposed a selection principle of MAS and showed that solvent engineering by designing the mixed antisolvent system can lead to the fabrication of high-performance perovskite solar cells. KEYWORDS: mixed antisolvent, transformation process, nucleation density, ultrasmooth film, perovskite solar cell

■

INTRODUCTION Recently, a power conversion efficiency (PCE) over 22% has been certified for perovskite solar cells.1 Compared to the PCE of 9.7% first achieved by a solid-state perovskite solar cell in 2012,2 there was a surge in the device performance. Among all the factors that drive the surging of organo-lead halide perovskite (APbX3; A = CH3NH3+, HC(NH2)2+; X = Br−,I−) based photovoltaics, the contribution of film deposition technology is indispensable.3−16 MAPbI3 absorber layer for perovskite solar cell can be produced by simple one-step solution process, which is low-cost and less time-consuming.2,9,10,17,18 A conventional one-step solution method always produces a perovskite film with poor coverage because of the undesirable crystallization caused by the solubility difference between MAI and PbI2, which is detrimental to the device performance, as previously reported.2,18−21 Therefore, many researchers aimed at obtaining a high quality film by improving the film deposition method. Additives such as 1,8-diiodooctane (DIO), HI, N-cyclohexyl-2-pyrrolidone (CHP), or the cosolvent of dimethyl sulfoxide (DMSO) were added in the precursor solution to increase the solubility of PbI2,7,22,23 which made the crystallization process more controllable and resulted in films with better coverage. However, pinholes were still unavoidable.7 Solvent annealing or a vapor-assisted annealing process employing N,N-dimethylformamide (DMF) were also introduced to promote the crystal growth and played an important role in increasing the grain size and smoothing the surface of the film, however, the DMF vapor could redissolve © 2017 American Chemical Society

the perovskite and produce cavities if without delicate control.11,24,25 Recently, dripping an antisolvent such as ethyl ether, toluene, or chlorobenzene (CBZ) during the spin coating has become one of the most effective ways to achieve a dense film, because PbI2, MAI, or perovskite cannot be dissolved in these antisolvents and hence they can be precipitated quickly by antisolvents.9,10,17,26 The antisolvent dripping method is simple to operate and can be easily repeated. Perovskite films applying this method have achieved great improvements in device performance. Many kinds of antisolvents are miscible with precursor solution9,10,17 and hence an abrupt increase of supersaturation can be achieved when the antisolvent was dripped into the presaturated precursor solution.27 Consequently, high nucleation density and small grain size are more likely to be obtained according to the classical crystal growth theory and the fact that the number of grains is inversely proportional to the grain size.11,28 Small grain size leads to large amount of grain boundaries which should be responsible for the low VOC of the device because of the severe charge recombination at grain boundaries.29 Also, the surfaces of films with high roughness may be achieved, which increases their probability of piercing through the hole transport material layer and results in the direct contact between perovskite and Ag electrode. This kind Received: November 8, 2016 Accepted: January 6, 2017 Published: January 18, 2017 3667

DOI: 10.1021/acsami.6b14270 ACS Appl. Mater. Interfaces 2017, 9, 3667−3676

Research Article

ACS Applied Materials & Interfaces

MASs (defined by n-hexane proportion). Chloroform and n-hexane were mixed at various volume ratios of 5:95, 10:90, 25:75, 50:50, and 75:25 to obtain 5%, 10%, 25%, 50%, and 75% chloroform MASs (defined by chloroform proportion). MASs were then shaken to make sure the solvent was homogeneous. All the perovskite films were deposited in dry room with a humidity level of less than 1%. To prepare perovskite films, 20 μL of perovskite precursor solution (A or B) was spin-coated on top of the m-TiO2 substrate at room temperature at 5000 rpm, and then 500 μL of MAS was dropped on the substrate at the sixth s and lasted for 6 s to accelerate the crystallization, and the substrate was spun for another 8 s. The asdeposited films were then annealed at 100 °C on a hot plate for 25 min. The solution for hole transport material (HTM) was prepared by dissolving 72.3 mg of sprio-MeOTAD (Merck) in 1 mL of chlorobenzene, in which 21.8 μL of (lithium bis (trifluoromethylsulfonyl) imide) in acetonitrile (Merck) (520 mg/mL) and 17.6 μL of 4tert-butylpyridine (Sigma-Aldrich) were added. HTM layer was deposited by spin coating HTM solution at 4000 rpm for 30 s. Finally, a 120 nm-thick Ag layer was thermally evaporated on top of the device as back contact at a rate of 0.2 nm/s under vacuum of 10−5 ∼ 10−6 Torr. Characterization. The top-view and cross-section SEM images of perovskite films were obtained with field emission scanning electron microscope (FESEM, FEI, Magellan 400). A thin chromium layer was introduced to improve the quality of cross-section SEM images. The XRD patterns of as-deposited films and annealed films were measured by an Ultima IV X-ray diffractometer using Cu Kα radiation under operation condition of 40 kV and 40 mA at room temperature (step width of 0.02°, scan speed of 4°/min, 2θ range of 5°−60°). The light absorptions of films were recorded on a Vis−NIR spectrophotometer (HITACHI U-3010) in the wavelength range of 400−900 nm at room temperature. Steady photoluminescence (PL) measurements were conducted at room temperature on a NanoLog iHR320 device with an excitation wavelength of 583 nm. The surface properties of perovskite films were analyzed by Atomic Force Microscopy (CEKO, SPI3800N). Solar cell performance was measured using a class AAA solar simulator equipped with a 450 W xenon lamp which was calibrated to give simulated AM 1.5 sunlight, at an irradiance of 100 mW/cm2. The irradiance was calibrated using a Si-reference cell (Oriel-91150). J-V curves were recorded with a Keithley-2420 source meter and the mask area was 0.07 cm2. The step voltage was fixed at 10 mV, and the delay time was 50 ms, which is a set delay at each voltage step before measuring each current. Hysteresis was measured by reverse (forward bias (1.1 V) → short circuit (0 V)) or forward (short circuit (0 V) → forward bias (1.1 V)) scan. The IPCE of perovskite solar cell from 300 to 900 nm was measured with a SM-250 system (Bunkoh-keiki, Japan) and a Si photodiode (S1337−1010BQ).

of direct contact could accelerate the degradation of perovskite and the invalidation of the electrode due to the redox reaction, which is detrimental to device performance.30 Furthermore, the fast crystallization process makes it hard to control the morphology of perovskite films.19 Therefore, employing an appropriate antisolvent is indispensable to obtain films with large grain size and smooth surface, and could also make the crystallization control easier. There is no doubt that solvent design has played a significant role in preparing high quality perovskite films via solution process. Kim et al. utilized a mixture of γ-butyrolactone (GBL) and DMF as solvent to prepare precursor solution and found that it could produce films with better coverage than pure GBL or pure DMF and reduce the recombination loss at the interface of electron transport material (ETM)/perovskite and perovskite/hole transport material (HTM).31 Wu et al. employed a mixture of DMF and CBZ to achieve a vaporassisted annealing process which could dissolve small grains and facilitate recrystallization, hence greatly improving the quality of films and the performance of PSCs.25 Recently, Noel et al. have used the composite acetonitrile (ACN)/methylamine (MA) solvent system to prepare CH3NH3PbI3 precursor salts, resulting in the crystallization of CH3NH3PbI3 at room temperature and films exhibiting long carrier lifetime.32 Herein, a novel mixed antisolvent (MAS) has been designed and applied to deposit a perovskite film by a one-step solution method. A mixture of ethyl ether and n-hexane in varied proportions is employed to obtain films with adjustable morphologies and various qualities. Optimized blend of ethyl ether and n-hexane produces films with ultrasmooth surface due to the high nucleation density achieved by ethyl ether and the retarded formation of perovskite achieved by n-hexane. Perovskite films deposited by optimized MAS also possess improved orientation of grains, reduced charge recombination, and improved transport capability of charge carrier on perovskite/HTM interface. Solar cells based on optimized films show excellent performance with efficiency boosting to 17.08% under one sun illumination. In addition, the effect mechanism of MAS on the film quality and device performance have been clarified, and a MAS selection principle have also been proposed based on this work.

■

■

EXPERIMENTAL SECTION

Substrate Preparation. FTO glasses (Pilkington,8Ω/□) were cleaned and sonicated with cleaning agent, deionized water, acetone, and ethanol in an ultrasonic bath successively, each for 15 min. The compact TiO2 layer was formed through the spin-coating of a precursor sol at 3000 rpm for 20 s and subsequently sintering at 510 °C for 30 min. The substrates were further treated with 40 mM TiCl4 solution at 70 °C for 40 min, then cleaned with deionized water and sintered at 510 °C for 30 min after drying. TiO2 paste diluted by ethanol with weight ratio of 1:4 was used to prepare mesoporous TiO2 layers by spin-coating at 3000 rpm for 20 s, and the prepared samples were then sintered at 510 °C for 30 min again. Finally, the substrates were exposed to UV-ozone for 15 min before using. Perovskite Solar Cell Fabrication. To generate the MAPbI3 perovskite precursor solution A, PbI2 (Sigma-Aldrich), MAI (TCI98%) and DMSO (Sigma-Aldrich) were dissolved in an anhydrous DMF (Sigma-Aldrich) solvent at a molar ratio of 1:1:1 with final concentration of 1.42 mM. To produce a perovskite precursor solution B, PbI2 and MAI were dissolved at a molar ratio of 1:1 in a mixture of DMF and GBL (GBL and DMF were blended at a weight ratio of 3:1) with the final precursor concentration of 45 wt %. Ethyl ether and nhexane were mixed at volume ratios of 70:30, 50:50, 30:70, 10:90, 5:95, and 2:98 to obtain 30%, 50%, 70%, 90%, 95%, and 98% n-hexane

RESULTS AND DISCUSSION MAI, PbI2, and DMSO dissolved in DMF solvent with 1:1:1 stoichiometric ratio was employed as the precursor solution A in this research. The preparation process of perovskite was described as following: the precursor solution A was deposited onto the mesoporous TiO2 substrate by spin-coating at 5000 rpm and 500 μL of antisolvent was dropped at the sixth s and lasted for 6 s. Another 8 s later the spin-coating was finished, and then the as-deposited film was annealed at 100 °C for 25 min on a hot plate. Ethyl ether/n-hexane mixed antisolvent (MAS) was obtained by mixing ethyl ether and n-hexane at a certain volume ratio which is expressed by the volume percent of n-hexane. For example, 5% n-hexane MAS means that nhexane and ethyl ether are mixed with a volumetric proportion of 5:95. All of the ethyl ether/n-hexane MASs are homogeneous solution because ethyl ether is completely miscible with n-hexane.33 0% and 100% n-hexane MASs represent ethyl ether and n-hexane to simplify the expression, respectively. 3668


Research Article

ACS Applied Materials & Interfaces Here we divide the formation process of perovskite films deposited by antisolvent dripping method into two stages simply. The first stage is the nucleation of the intermediate of MAI·PbI2·DMSO during the spin-coating induced by solvent evaporation and antisolvent dripping, while the second stage is the structure transformation from the intermediate of MAI· PbI2·DMSO to perovskite (also the formation of perovskite) and the crystal growth of perovskite during the annealing. Figure 1a shows the XRD patterns of the as-deposited films

with 0% n-hexane MAS, suggesting that the complete conversion from the intermediate of MAI·PbI2·DMSO to perovskite was almost achieved by ethyl ether dripping without annealing. When the proportion of n-hexane is 50%, intensities of peaks belonging to perovskite structure decreased sharply, indicating that n-hexane can help to slow down the fast transformation process caused by ethyl ether. When the proportion of n-hexane exceeds 70%, intensities of peaks at 6.80° and 9.44° attributed to the intermediate of MAI·PbI2· DMSO increased and signals belonging to perovskite almost disappeared. For the films prepared with 90%, 95%, 98% and 100% n-hexane MASs, diffraction peaks belonging to the perovskite structure could not be detected, suggesting the intermediate of MAI·PbI2·DMSO would not turn to perovskite by n-hexane rich MASs during spin-coating. The previous research believed that nucleation and crystal growth taking place in separate processes was beneficial to obtain high quality films and well device performance.27 Obviously, ethyl ether dripping accelerated the structure transformation process and perovskite was formed in the first stage before annealing, indicating that nucleation of MAI·PbI2·DMSO and crystal growth of perovskite could not be well separated. By contrast, n-hexane in MAS was able to suppress the effect of ethyl ether and delayed the formation of perovskite. As for the annealed films, XRD patterns are shown in Figure 1b. Peaks located at 14.20°, 28.51°, 31.92° assigned to the tetragonal perovskite phase of (110), (220), and (310) crystal planes were detected for all samples. It can be seen that the structure transformation process and the following crystal growth of perovskite did not take place until annealing for the films prepared by 70%, 90%, 95%, 98%, and 100% n-hexane MASs (n-hexane rich). The intensity ratios of (310) and (110) for various films were listed close to the corresponding (310) peak in Figure 1b. It is clear that the ratio varied between different samples. The intensity ratio was 0.49 for the film prepared by 0% n-hexane MAS, which was the highest ratio among the films, indicating the lowest grain-orientation. It dropped to 0.36, 0.19, 0.20, and 0.12 for films produced by 70%, 90%, 95% and 98% n-hexane MASs respectively, indicating that the growth of grains was more organized for the films prepared with n-hexane-rich MASs. Here, it can be found that the dripping of ethyl ether not only accelerated the transformation of the intermediate but also led to disordered crystal growth, while n-hexane added in MASs effectively delayed the structure transformation and the following crystal growth of perovskite until annealing and hence films with improved orientation of grains were achieved. It is known that films consisting of highly oriented grains could promote more effective charge transport (low recombination) and are beneficial to high cell efficiency.3,19,35,36 Figure 2 shows the top-view SEM images of films obtained by dripping 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% nhexane MASs, respectively. The corresponding cross-section SEM images are also presented in Figure 2, and all the films indicates a perovskite layer with the thickness of 300−350 nm. It can be seen that 0% n-hexane MAS resulted in dense but uneven perovskite layers with many small grains attaching to the larger one from the top-view SEM images, which also contributed to too many grain boundaries in different directions as shown in the corresponding cross-section images. The bright grains and dark grains appeared in the top-view SEM image also presented the rough surface because they were obviously not in the same plane. The rough surface probably

Figure 1. XRD patterns of (a) as-deposited and (b) annealed films prepared with with 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% nhexane MASs, the intensity ratios of (310) and (110) are listed close to the corresponding (310) peak. The same color of the curves in (a) and (b) indicates the samples with the same MAS.

without annealing. For the films prepared with 0%, 30% nhexane MASs, sharp peaks at 14.20°, 28.51°, 31.92°, 40.57°, and 43.18° correspond to the (110), (220), (310), (224), and (314) crystal planes of tetragonal perovskite structure, respectively. However, the signals at 6.80° and 9.44° attributed to the intermediate of MAI·PbI2·DMSO were too weak to be observed, indicating that the structure conversion from the intermediate of MAI·PbI2·DMSO to perovskite have already taken place in these two films before annealing, which is probably related to the solvent extraction capacity of ethyl ether34 (extraction of DMSO). It is also clear that the peak supposed to belong to (220) was actually composed of two strong diffraction peaks at 28.23° and 28.51° corresponding to (004) and (220) crystal planes as shown in the inset in Figure 1a. This phenomenon was only observed in the as-deposited films prepared with 0% and 30% n-hexane MASs, which should be related to the initial formation of small perovskite grains with (004) and (220) planes exposed during the spin-coating. However, the (004) plane grew much faster than the (220) plane, and this is why the peak belonging to the (004) plane disappeared in the annealed films. The fitted peaks at 14.12° and 14.22° corresponding to (002) and (110) also indicated the same phenomenon as shown in Figure S1 of the Supporting Information. Furthermore, the high angle diffraction peak of (440) plane was also detected at 58.92° for the film prepared 3669


Research Article


Figure 2. Top-view and the corresponding cross-section SEM images of films prepared with 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% nhexane MASs. The scale bars correspond to 1 μm in length.

resulted from the disordered crystal growth caused by ethyl ether dripping. 30%, 50%, and 70% n-hexane MASs also resulted in dense perovskite layers. However, the grains packed orderly without any grain boundaries parallel to the substrate only for the films produced by 50% and 70% n-hexane MASs according to the cross-section SEM images, but not for the film derived from 30% n-hexane MAS. When the percentage of nhexane is up to 90%, the film was dense and smooth with no cavity in the interior, however, few voids existed on the surface in Figure 2. The number of voids increased with the increase of the proportion of n-hexane in MAS. For the films obtained from 98% and 100% n-hexane MASs, a large amount of isolated voids were observed on the surface and large cavities were located in the interior of the films. Grain size analysis is also conducted with the top-view SEM images at low magnification. In-plane grain size distributions are shown in Figure 3, and the average grain size versus nhexane proportion is also summarized. The perovskite film fabricated by dripping 0% n-hexane MAS presents the smallest average grain size of 194 nm, while the 95% n-hexane MAS shows the largest average grain size of 302 nm. The statistical measurements are based on 1195 grains and 675 grains for 0% and 95% n-hexane MAS, respectively. Although the change range in grain size was not very large (about 100 nm), the average grain sizes of the films prepared by 0%, 30%, 50%, and 70% n-hexane MASs were lower than those of the films produced by 90%, 95%, 98%, and 100% n-hexane MASs. The difference in average grain size should be related to the different nucleation density, considering that high nucleation density always leads to small grain size.11,28 Both ethyl ether and nhexane are conducive to crystallization of perovskite because the reactant and perovskite cannot be dissolved in them, which is favorable for depositing films with good coverage and densely packed grains in the previous reports.9,26 Ethyl ether can be miscible with DMF33 and consequently it can promote the nucleation process in the whole precursor solution, however, the effect of n-hexane is limited due to its immiscibility with

Figure 3. (a) Grain size distribution of perovskite films prepared with 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% n-hexane MASs, and (b) average grain size versus the proportion of n-hexane in the MASs.

DMF,33 only the precursor/n-hexane interface being involved, which should be responsible for the lower nucleation density achieved by n-hexane-rich MASs than that by ethyl ether-rich MASs. The average grain size was stable at around 225 nm for 30%, 50%, 70% n-hexane MASs, also indicating the strong effect of ethyl ether on the nucleation process in MASs. The voids existed in films formed with 95%, 98%, and 100% MASs that are n-hexane rich should be due to the relatively low nucleation density. Although the low nucleation density is able to prepare 3670


Research Article


Figure 4. Schematic representation of the formation process of perovskite films by MASs with 0%, 30%, 70%, and 100% of n-hexane.

Figure 5. AFM images in 1 × 1 μm2 of films prepared with 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% n-hexane MASs.

films with slightly large grain size, it would be easy to leave voids on the surface of films. Therefore, it can be seen that high nucleation density is crucial to obtain dense films for the MAS system. The schematic diagrams of depositing perovskite films deposited by dripping various MASs are illustrated in Figure 4. Four representative MASs of 0%, 30%, 70%, and 100% were employed to present the formative process. As discussed above, MASs could control not only the nucleation density but also the intermediate transformation process by the proportion of nhexane. For 0% and 30% n-hexane MASs, high nucleation density, and fast transformation were achieved at the first stage during the spin-coating, leading to dense films with rather rough surface. For 50% and 70% n-hexane MAS, the high nucleation density was also obtained while the transformation from the intermediate to perovskite was slowed down, hence resulting in dense and smooth films with improved orientation of grains. When the proportion of n-hexane was above 90% in the MAS, the quite slow transformation of the intermediate contributed to films with improved orientation of grains. However, low nucleation densities introduced voids and cavities, which were considered to be detrimental to device performance. In many previous reports,4,25,31,32 a smoother surface was believed to be crucial for films to achieve better performance.

To further characterize the surface roughness of the perovskite films, AFM measurements were conducted. The measurements were made in two different scanning areas, i.e., 1 ×1 μm2, 5 × 5 μm2. To make sure the results statistically informative, the scanning area was randomly chosen from a large and uniform perovskite film with a size of 1 cm2. Figure 5 shows the AFM images in 1 × 1 μm2 scanning areas of perovskite films. AFM images in 5 × 5 μm2 can be found in Figure S2. The corresponding root-mean-square (RMS) roughness was obtained in 5 × 5 μm2 scanning area as listed in Table S1 to estimate the surface roughness of the film. Film prepared by dripping 100% n-hexane MAS showed the largest RMS roughness of 8.22 nm. Roughnesses of 7.56, 7.75, and 6.41 nm were obtained for films prepared by 98%, 95%, and 90% nhexane MASs, respectively. Separate voids on the film surface should be responsible for these rough surfaces. By contrast, quite low roughness of 4.34 and 5.12 nm were detected for the films fabricated with 50% and 70% n-hexane MASs, which is in accordance with the ultrasmooth surface of film in SEM images. Although there was no void existing in the film prepared with 0% n-hexane MASs, the RMS roughness increased to 6.07 nm, indicating that the fast transformation from the intermediate of MAI·PbI2·DMSO to perovskite and the following crystal growth of perovskite before annealing leads to films with unsmooth surfaces. Rough surface is believed to be harmful to 3671


Research Article


Table 1. Summarized RMS Roughness of Perovskite Films Deposited by Six Typical and Significant Solution Methods and MAS Strategy in the Present Work method

area (μm2)

roughness (nm)

substrate

vacuum flash−assisted solution process toluene drop-casting DMSL/GBL solvent system acetonitrile (ACN)/MA solvent system DMF/GBL solvent system mixed vapor-assisted annealing (CBZ/DMF vapor system) vapor-assisted solution process MAS strategy ethyl ether/n-hexane (A)a ethyl ether/n-hexane (B)b chloroform/n-hexane (A)c

10 × 10 3×3 5×5

30 ± 5 8.3 7.8 6.6 30

m-TiO2 m-TiO2 c- TiO2 PEDOT:PSS TiO2 nanoparticles (TNPs)

5×5 5 5 5 1

× × × ×

5 5 5 1

23.2 4.34 4.88 4.80

c- TiO2 m-TiO2 m-TiO2 m-TiO2

reference 8 17 32 31 25 4 this work this work this work

a Employing precursor solution A and 50% ethyl ether/50% n-hexane MAS. bEmploying precursor solution B and 70% ethyl ether/30% n-hexane MAS. cEmploying precursor solution A and 25% chloroform/75% n-hexane MAS.

Figure 6. (a) PL spectra for m-TiO2/perovskite samples, perovskite layers were prepared by dripping 0%, 50%, 70%, 95%, and 100% n-hexane MASs and the peak between 765 and 775 nm presents the radiative recombination of free carriers. (b) Absorbance spectra of perovskite films prepared with 0%, 30%, 50%, 70%, 90%, 95%, 98%, and 100% n-hexane MASs. (c) Dependence of JSC, VOC, FF, and PCE on the proportion of n-hexane in MASs including 0%, 30%, 50%, 70%, 90%, 95%, and 100%. (d) PL spectra for m-TiO2/perovskite/sprio samples, the diminished PL intensity comparing to that in (a) presents the efficient injection of free carriers into sprio. (e) IPCE spectrum of PSC derived from 50% n-hexane MAS, and the integrated current density is 21.64 mA/cm2. (f) Stabilized photocurrent and efficiency of the device derived from 50% n-hexane MAS, measured by holding the device at J−V determined maximum power point for 300 s.

based on the films prepared with 0% n-hexane MAS and 100% n-hexane MAS exhibited comparable efficiency, there were significant differences in VOC, JSC, and FF. Devices derived from 0% n-hexane MAS possessed higher VOC and FF while PSCs derived from 100% n-hexane MAS possessed higher JSC. The PL spectra carried on the m-TiO2/perovskite samples in Figure 6a shows that the PL intensities of samples derived from 95%, 100% n-hexane MASs were lower than those of films with dense morphology, indicating that serious nonradiative recombination40 took place by a large population of trapped carriers for samples with isolated voids, which should be responsible for the relatively low VOC and low FF. Considering the films prepared with 95%, 98%, and 100% n-hexane MASs presented excellent light absorption in the range of 500−800 nm in Figure 6b, the efficient utilization of light through

the device performance due to its high probability of piercing through the hole transport material and thus many researchers have aimed at improving the fabrication method to overcome the uneven surface.37−39 The surface roughnesses of the films prepared by other significant solution methods and MAS strategy in this work were summarized in Table 1. It is clear that the film produced by 50% n-hexane MASs exhibited extremely low roughness, suggesting that the ultrasmooth surface could be obtained through both reaching high nucleation density and slowing down the transformation process from the intermediate to perovskite. The effect of ethyl ether/n-hexane MASs dripping method on cell performance was investigated via mesoporous perovskite solar cells. The devices were structured as FTO/c-TiO2/ m-TiO2/perovskite/sprio-MeOTAD/Ag. J-V curves can be found in Figure S3. Although perovskite solar cells (PSCs) 3672


Research Article


Table 2. Photovoltaic Parameters for Devices Derived from Dripping Various Ethyl Ether/n-Hexane MASs (0%, 30%, 50%, 70%, 90%, 95%, and 100%)a cell (%) 0 30 50 70 90 95 100 a

JSC (mA/cm2) 21.32 21.46 21.78 21.88 21.83 22.12 22.20

± ± ± ± ± ± ±

0.20 0.13 0.28 0.22 0.34 0.03 0.29

VOC (V) 1.061 1.067 1.078 1.070 1.065 1.050 1.053

± ± ± ± ± ± ±

FF (%)

0.004 0.004 0.009 0.007 0.006 0.003 0.006

68.42 69.51 70.78 69.47 69.10 66.54 66.20

± ± ± ± ± ± ±

0.80 0.53 0.92 0.44 1.52 1.30 0.57

PCE (%)

PCEBEST (%)

± ± ± ± ± ± ±

15.74 16.10 17.08 16.48 16.28 15.74 15.89

15.47 15.92 16.62 16.26 16.06 15.45 15.48

0.18 0.20 0.27 0.15 0.16 0.29 0.28

Reverse scans were measured for all devices and five cells were fabricated for each type of devices.

Figure 7. Top-view SEM images of films prepared with 5%, 10%, 25%, 50%, 75%, and 100% chloroform MASs. The scale bars correspond to 1 μm in length.

21.64 mA/cm2, consistent with the JSC from the J−V measurement. This indicates an excellent incident photon collection capability of the cell derived from 50% n-hexane MAS. The J−V curves in Figure S4 was obtained with forward and reverse scanning directions, showing that there was a little hysteresis for device derived from 50% n-hexane MAS. Furthermore, stabilized power output under load near the point of maximum power was monitored. As shown in Figure 6f, the photocurrent rose to a relatively stable value, and the steady-state efficiency was close to that measured from the photocurrent scanning, yielding a stabilized PCE of 16.33%. From the above discussion, we can see that relatively high nucleation density and retarded transformation from the intermediate of MAI·PbI2·DMSO to perovskite (retarded formation of perovskite) are crucial to obtain dense and smooth films with greatly improved device performances. In order to confirm the universality of the MAS strategy, we conducted experiments with varied MAS system and varied precursor system, respectively. For the varied MAS system, chloroform/n-hexane MAS could also work well in improving the quality of films and the device performances. Chloroform and n-hexane were mixed at the volume ratios of 5:95, 10:90, 25:75, 50:50, and 75:25 to prepared 5%, 10%, 25%, 50%, and 75% chloroform MASs (defined by chloroform proportion), and all the MASs were homogeneous. The method of producing perovskite films employing this type of MASs was similar to ethyl ether/nhexane MASs. Figure 7 shows the top-view SEM images of films prepared with various chloroform/n-hexane MASs. Chloroform is miscible with DMF and is able to lead to fast crystallization like ethyl ether. Perovskite films produced by dripping 5% and 10% chloroform MASs presented porous

scattering effect caused by existing voids should be responsible for the higher JSC. Furthermore, a batch of 5 cells were also tested for each type of MAS to investigate the reproducibility. The photovoltaic parameters of these PSCs were statistically analyzed in Figure 6c and Table 2. It is worth noting that the average PCE of devices derived from 50%, 70%, and 90% n-hexane MASs exceeded 16% possessing superior VOC and FF. PSCs derived from 50% n-hexane MAS had an average PCE of 16.62% with a VOC of 1.078 V, a FF of 70.78%, and a JSC of 21.78 mA/cm2, comparing to the 15.47% and 15.48% PCE of the devices from 0% n-hexane MAS and 100% n-hexane MAS with VOC of 1.061 and 1.053 V, JSC of 21.32, and 22.20 mA/cm2, FF of 68.42% and 66.20%, respectively. From the PL spectra carried on mTiO2/perovskite/sprio-MeOTAD samples in Figure 6d, it can be found that the perovskite (prepared with 50%, 70%, 95% nhexane MASs)/HTM interfaces were more effective at separating free holes than the perovskite/HTM interfaces derived from 0%, 100% n-hexane MASs. It is the isolated voids that hinder the effective transportation of carriers for 100% nhexane MAS. However, 0% n-hexane MAS was able to get films without any voids, suggesting that dense films would not guarantee effective separation of charges if without orderly packed grains and smooth surface. Here, for 50%, 70%, and 90% n-hexane MASs, dense and smooth films with improved orientation of grains contributed to the enhancement of VOC, FF, and PCE. The best PCE of 17.08% was achieved by 50% nhexane MAS with a VOC of 1.091 V, a JSC of 22.06 mA/cm2, and a FF of 70.99%. The incident photon-to-current conversion efficiency (IPCE) spectrum of the best performance device is shown in Figure 6e, the integrated JSC obtained from the IPCE for this cell was 3673


Research Article


Figure 8. AFM images in 1 × 1 μm2 of films prepared with 5%, 10%, 25%, 50%, 75%, and 100% chloroform MASs.

prepared by 0%, 30% and 50% MAS exhibited improved device performances with the average PCE of 8.48%, 11.49%, 9.89%, respectively. It is clear that 30% and 50% n-hexane MAS could achieve better quality of films and also the greatly improved PCEs than pure ethyl ether and pure n-hexane antisolvents, with the best PCE of 11.81% and 11.02%, respectively. However, PSCs employing precursor solution B (a mixture solvent of DMF and GBL) achieved lower performance than those utilizing precursor A (a mixture solvent of DMF and DMSO). Grains in the films derived from precursor solution B (Figure S6) were much smaller than those from precursor solution A (Figure 5), which should be responsible for the inferior device performances. Furthermore, DMSO could result in the Lewis base adduct of MAI·PbI2·DMSO, which possibly led to highly regulated crystal growth according to the previous work.9 Therefore, MAS strategy is suitable for a solution method containing the intermediate stage, which is usually achieved by coordination between polar solvent and Pb cation and needs further structure transformation to form perovskite.28,41 The MAS strategy should meet the following conditions to obtain dense and smooth films. One is achieving relatively high nucleation density and the other is the retarded formation of perovskite during the spin-coating. Initially, both antisolvents should be able to promote crystallization of perovskite (cannot dissolve MAI, PbI2 and CH3NH3PbI3) and they should be mutually soluble. Also, one type of antisolvent should be miscible with the solvent of precursor which is usually polar solvent to guarantee high nucleation density (this kind of antisolvent should be polar solvent like ethyl ether, chloroform, toluene, and chlorobenzene). And another antisolvent should not accelerate the transformation of the intermediate during the spin-coating through polar solvent extraction (this kind of antisolvent should be nonpolar solvent like n-hexane).

morphology and nearly dense morphology with some isolated voids on the surface, respectively. For 25% and 50% chloroform MASs, dense and smooth films were obtained. Dense films were also produced by 75% and 100% chloroform MASs, however, much small grains were observed attaching to the large ones, indicating the relatively rough surfaces. AFM images are also shown in Figure 8, and the film roughnesses calculated from the AFM measurements were listed in Table S2. It is clear that the films with voids possessed high roughnesses of 24.0 nm (5% chloroform MAS) and 20.2 nm (10% chloroform MAS), and the dense films prepared by 75% and 100% chloroform MASs also achieved high roughnesses of 13.29 and 9.43 nm, while 25% and 50% chloroform MASs produced quite smooth films with extremely low roughnesses of 4.80 and 5.14 nm, respectively. Photovoltaic parameters of devices based on chloroform/n-hexane MASs are listed in Table S3. It can be seen that the films produced by 25% chloroform MAS showed the greatly improved device performance with an average PCE of 15.79% (the best PCE of 16.77%), and the film prepared by 50% chloroform MAS also possessed good performance with the average PCE of 14.53%, however, 12.62% and 12.12% PCE of the devices from 75% and 100% chloroform MASs were obtained, respectively. Therefore, high quality films with ultrasmooth surfaces indeed have boosted the device performance greatly, and both types of antisolvents in MAS have played significant roles in obtaining ultrasmooth films. For the varied precursor system, 20 μL of precursor solution B was dropped on the substrate and spun coated at 5000 rpm. 500 μL of ethyl ether/n-hexane MAS was dripped onto the precursor at the sixth s and lasted for 6 s, and then the substrate was spun for another 8 s. Figure S5 shows the AFM images in 5 × 5 μm2 scanning areas of perovskite films. It is clear that the films produced by 70%, 90%, and 100% n-hexane MASs were porous, which should be responsible for the rather rough surfaces with roughness of 10.74, 15.37, and 41.5 nm, respectively (Table S4). While the grains were packed densely in the films produced by 0%, 30%, and 50% n-hexane MASs. Thirty% and 50% n-hexane MASs resulted in quite smooth films with the roughness of 4.88 and 5.47 nm compared to 6.87 nm obtained by 0% n-hexane MAS. Device performances were also measured and listed in Table S5. Films with voids and high roughness surfaces showed inferior performances, with the average PCE of 1.23%, 4.67%, and 7.12% for 100%, 90%, and 70% n-hexane MASs, respectively. By contrast, dense films

■

CONCLUSIONS

In summary, we have developed a novel mixed antisolvent (MAS, e.g., a mixture of ethyl ether and n-hexane) deposition method to prepare perovskite films. Formative processes of films with different morphologies obtained by dripping various MASs were analyzed. Ethyl ether in MAS is beneficial to achieving high nucleation density, which is helpful to produce dense films. And n-hexane in MAS is able to slow down the transformation from the intermediate of MAI·PbI2·DMSO to 3674


Research Article


(3) Huang, F.; Dkhissi, Y.; Huang, W.; Xiao, M.; Benesperi, I.; Rubanov, S.; Zhu, Y.; Lin, X.; Jiang, L.; Zhou, Y.; Gray-Weale, A.; Etheridge, J.; McNeill, C. R.; Caruso, R. A.; Bach, U.; Spiccia, L.; Cheng, Y.-B. Gas-Assisted Preparation of Lead Iodide Perovskite Films Consisting of a Monolayer of Single Crystalline Grains for High Efficiency Planar Solar Cells. Nano Energy 2014, 10, 10−18. (4) Chen, Q.; Zhou, H.; Hong, Z.; Luo, S.; Duan, H. S.; Wang, H. H.; Liu, Y.; Li, G.; Yang, Y. Planar Heterojunction Perovskite Solar Cells via Vapor-Assisted Solution Process. J. Am. Chem. Soc. 2014, 136, 622−625. (5) Nie, W.; Tsai, H.; Asadpour, R.; Blancon, J.-C.; Neukirch, A. J.; Gupta, G.; Crochet, J. J.; Chhowalla, M.; Tretiak, S.; Alam, M. A.; Wang, H.-L.; Mohite, A. D. High-Efficiency Solution-Processed Perovskite Solar Cells with Millimeter-Scale Grains. Science 2015, 347, 522−525. (6) Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Sequential Deposition as a Route to High-Performance Perovskite-Sensitized Solar Cells. Nature 2013, 499, 316−319. (7) Liang, P. W.; Liao, C. Y.; Chueh, C. C.; Zuo, F.; Williams, S. T.; Xin, X. K.; Lin, J.; Jen, A. K. Additive Enhanced Crystallization of Solution-Processed Perovskite for Highly Efficient Planar-Heterojunction Solar Cells. Adv. Mater. 2014, 26, 3748−3754. (8) Li, X.; Bi, D.; Yi, C.; Decoppet, J.-D.; Luo, J.; Zakeeruddin, S. M.; Hagfeldt, A.; Grätzel, M. A Vacuum Flash-Assisted Solution Process for High-Efficiency Large-Area Perovskite Solar Cells. Science 2016, 353, 58−62. (9) Ahn, N.; Son, D. Y.; Jang, I. H.; Kang, S. M.; Choi, M.; Park, N. G. Highly Reproducible Perovskite Solar Cells with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct of Lead(II) Iodide. J. Am. Chem. Soc. 2015, 137, 8696−8699. (10) Xiao, M.; Huang, F.; Huang, W.; Dkhissi, Y.; Zhu, Y.; Etheridge, J.; Gray-Weale, A.; Bach, U.; Cheng, Y.-B.; Spiccia, L. A Fast Deposition-Crystallization Procedure for Highly Efficient Lead Iodide Perovskite Thin-Film Solar Cell. Angew. Chem., Int. Ed. 2014, 53, 9898−9903. (11) Xiao, Z.; Dong, Q.; Bi, C.; Shao, Y.; Yuan, Y.; Huang, J. Solvent Annealing of Perovskite-Induced Crystal Growth for PhotovoltaicDevice Efficiency Enhancement. Adv. Mater. 2014, 26, 6503−6509. (12) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.; Seok, S. I. High-Performance Photovoltaic Perovskite Layers Fabricated through Intramolecular Exchange. Science 2015, 348, 1234−1237. (13) Cao, Q.; Yang, S.; Gao, Q.; Lei, L.; Yu, Y.; Shao, J.; Liu, Y. Fast and Controllable Crystallization of Perovskite Films by Microwave Irradiation Process. ACS Appl. Mater. Interfaces 2016, 8, 7854−7861. (14) Ji, F.; Wang, L.; Pang, S.; Gao, P.; Xu, H.; Xie, G.; Zhang, J.; Cui, G. A Balanced Cation Exchange Reaction toward Highly Uniform and Pure Phase FA1−xMAxPbI3 Perovskite Films. J. Mater. Chem. A 2016, 4, 14437−14443. (15) Zhou, Y.; Yang, M.; Pang, S.; Zhu, K.; Padture, N. P. Exceptional Morphology-Preserving Evolution of Formamidinium Lead Triiodide Perovskite Thin Films via Organic-Cation Displacement. J. Am. Chem. Soc. 2016, 138, 5535−5538. (16) Xiao, Z.; Bi, C.; Shao, Y.; Dong, Q.; Wang, Q.; Yuan, Y.; Wang, C.; Gao, Y.; Huang, J. Efficient, High Yield Perovskite Photovoltaic Devices Grown by Interdiffusion of Solution-Processed Precursor Stacking Layers. Energy Environ. Sci. 2014, 7, 2619−2623. (17) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S. I. Solvent Engineering for High-Performance Inorganic-Organic Hybrid Perovskite Solar Cells. Nat. Mater. 2014, 13, 897−903. (18) Etgar, L.; Gao, P.; Xue, Z.; Peng, Q.; Chandiran, A. K.; Liu, B.; Nazeeruddin, M. K.; Grätzel, M. Mesoscopic CH3NH3PbI3/TiO2 Heterojunction Solar Cells. J. Am. Chem. Soc. 2012, 134, 17396− 17399. (19) Huang, L.; Hu, Z.; Yue, G.; Liu, J.; Cui, X.; Zhang, J.; Zhu, Y. CH3NH3PbI(3‑x)Cl(x) Films with Coverage Approaching 100% and with Highly Oriented Crystal Domains for Reproducible and Efficient

perovskite during spin-coating, making the following crystal growth of perovskite quite organized and mainly take place under annealing. As a result, dense perovskite films with ultrasmooth surfaces were produced by 50% and 70% n-hexane MASs. The significantly improved device performances were achieved by the films prepared by 50%, 70%, and 90% n-hexane MASs and the best efficiency of 17.08% was obtained by 50% nhexane MAS. Chloroform/n-hexane mixed antisolvent systems also worked well and ultrasmooth films prepared by 25% chloroform MASs achieved the best PCE of 16.77%. The selection principle of MAS based on this work was proposed. It also presents that solvent engineering by designing the mixed antisolvent system could lead to the fabrication of highperformance perovskite solar cells.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b14270. Peak overlap phenomenon for XRD patterns of the films derived from 0% and 30% ethyl ether/n-hexane MASs, AFM images in 5 × 5 μm2 of perovskite films prepared by ethyl ether/n-hexane MASs, J−V curves of PSCs derived from ethyl ether/n-hexane MASs, hysteresis level for the PSC derived from 50% ethyl ether/n-hexane MAS, AFM images in 5 × 5 μm2 and 1 × 1 μm2 of films derived from precursor solution B and ethyl ether/nhexane MASs, RMS roughness of perovskite films prepared with ethyl ether/n-hexane MASs, RMS roughness of films prepared with chloroform/n-hexane MASs, photovoltaic parameters for devices derived from chloroform/n-hexane MASs, RMS roughness of films derived from precursor solution B and ethyl ether/n-hexane MASs, photovoltaic parameters for devices derived from precursor solution B and ethyl ether/n-hexane MASs (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (S.Y.). *E-mail: [email protected] (Y.L.). ORCID

Songwang Yang: 0000-0001-6304-5941 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National High Technology Research and Development Program of China (Grant No. 2014AA052002), Shanghai Municipal Natural Science Foundation (Grant No. 16ZR1441000), Shanghai Municipal Sciences and Technology Commission (Grant No. 12DZ1203900), and the Shanghai High & New Technology’s Industrialization Major Program (Grant No. 2013-2).

■

REFERENCES

(1) NREL, www.nrel.gov/ncpv/images/efficiency_chart.jpg 2016. (2) Kim, H. S.; Lee, C. R.; Im, J. H.; Lee, K. B.; Moehl, T.; Marchioro, A.; Moon, S. J.; Humphry-Baker, R.; Yum, J. H.; Moser, J. E.; Gratzel, M.; Park, N. G. Lead Iodide Perovskite Sensitized AllSolid-State Submicron Thin Film Mesoscopic Solar Cell with Efficiency Exceeding 9%. Sci. Rep. 2012, 2, 591. 3675


Research Article

ACS Applied Materials & Interfaces Planar Heterojunction Perovskite Solar Cells. Phys. Chem. Chem. Phys. 2015, 17, 22015−22022. (20) Eperon, G. E.; Burlakov, V. M.; Docampo, P.; Goriely, A.; Snaith, H. J. Morphological Control for High Performance, SolutionProcessed Planar Heterojunction Perovskite Solar Cells. Adv. Funct. Mater. 2014, 24, 151−157. (21) Jeng, J. Y.; Chiang, Y. F.; Lee, M. H.; Peng, S. R.; Guo, T. F.; Chen, P.; Wen, T. C. CH3NH3PbI3 Perovskite/Fullerene PlanarHeterojunction Hybrid Solar Cells. Adv. Mater. 2013, 25, 3727−3732. (22) Heo, J. H.; Song, D. H.; Han, H. J.; Kim, S. Y.; Kim, J. H.; Kim, D.; Shin, H. W.; Ahn, T. K.; Wolf, C.; Lee, T. W.; Im, S. H. Planar CH3NH3PbI3 Perovskite Solar Cells with Constant 17.2% Average Power Conversion Efficiency Irrespective of the Scan Rate. Adv. Mater. 2015, 27, 3424−3430. (23) Jeon, Y. J.; Lee, S.; Kang, R.; Kim, J. E.; Yeo, J. S.; Lee, S. H.; Kim, S. S.; Yun, J. M.; Kim, D. Y. Planar Heterojunction Perovskite Solar Cells with Superior Reproducibility. Sci. Rep. 2014, 4, 6953. (24) Zhu, W.; Yu, T.; Li, F.; Bao, C.; Gao, H.; Yi, Y.; Yang, J.; Fu, G.; Zhou, X.; Zou, Z. A Facile, Solvent Vapor-Fumigation-Induced, SelfRepair Recrystallization of CH3NH3PbI3 Films for High-Performance Perovskite Solar Cells. Nanoscale 2015, 7, 5427−5434. (25) Wu, W.-Q.; Chen, D.; Huang, F.; Cheng, Y.-B.; Caruso, R. A. Optimizing Semiconductor Thin Films with Smooth Surfaces and Well-Interconnected Networks for High-Performance Perovskite Solar Cells. J. Mater. Chem. A 2016, 4, 12463−12470. (26) Lin, N.; Qiao, J.; Dong, H.; Ma, F.; Wang, L. MorphologyControlled CH3NH3PbI3 Films by Hexane-Assisted One-Step Solution Deposition for Hybrid Perovskite Mesoscopic Solar Cells with High Reproductivity. J. Mater. Chem. A 2015, 3, 22839−22845. (27) Zhao, Y.; Zhu, K. Organic-Inorganic Hybrid Lead Halide Perovskites for Optoelectronic and Electronic Applications. Chem. Soc. Rev. 2016, 45, 655−689. (28) Li, L.; Chen, Y.; Liu, Z.; Chen, Q.; Wang, X.; Zhou, H. The Additive Coordination Effect on Hybrids Perovskite Crystallization and High-Performance Solar Cell. Adv. Mater. 2016, 28, 9862−9868. (29) Kim, H. D.; Ohkita, H.; Benten, H.; Ito, S. Photovoltaic Performance of Perovskite Solar Cells with Different Grain Sizes. Adv. Mater. 2016, 28, 917−922. (30) Kato, Y.; Ono, L. K.; Lee, M. V.; Wang, S.; Raga, S. R.; Qi, Y. Silver Iodide Formation in Methyl Ammonium Lead Iodide Perovskite Solar Cells with Silver Top Electrodes. Adv. Mater. Interfaces 2015, 2, 1500195. (31) Kim, H. B.; Choi, H.; Jeong, J.; Kim, S.; Walker, B.; Song, S.; Kim, J. Y. Mixed Solvents for the Optimization of Morphology in Solution-Processed, Inverted-Type Perovskite/Fullerene Hybrid Solar Cells. Nanoscale 2014, 6, 6679−6683. (32) Noel, N. K.; Habisreutinger, S. N.; Wenger, B.; Klug, M. T.; Hörantner, M. T.; Johnston, M. B.; Nicholas, R. J.; Moore, D. T.; Snaith, H. J. A Low Viscosity, Low Boiling Point, Clean Solvent System for the Rapid Crystallisation of Highly Specular Perovskite Films. Energy Environ. Sci. 2017, DOI: 10.1039/C6EE02373H. (33) Liu, G.; Ma, L.; Liu, J. Handbook of Physical Property Data in Chemistry and Chemical Industry-Organic, 1st ed; Chemical Industry Press: Beijing, 2002. (34) Zhou, Y.; Yang, M.; Wu, W.; Vasiliev, A. L.; Zhu, K.; Padture, N. P. Room-Temperature Crystallization of Hybrid-Perovskite Thin Films via Solvent-Solvent Extraction for High-Performance Solar Cells. J. Mater. Chem. A 2015, 3, 8178−8184. (35) Docampo, P.; Hanusch, F. C.; Giesbrecht, N.; Angloher, P.; Ivanova, A.; Bein, T. Influence of the Orientation of Methylammonium Lead Iodide Perovskite Crystals on Solar Cell Performance. APL Mater. 2014, 2, 081508. (36) Bi, D.; Yi, C.; Luo, J.; Décoppet, J.-D.; Zhang, F.; Zakeeruddin; Shaik, M.; Li, X.; Hagfeldt, A.; Grätzel, M. Polymer-Templated Nucleation and Crystal Growth of Perovskite Films for Solar Cells with Efficiency Greater than 21%. Nat. Energy 2016, 1, 16142. (37) Zhang, W.; Saliba, M.; Moore, D. T.; Pathak, S. K.; Horantner, M. T.; Stergiopoulos, T.; Stranks, S. D.; Eperon, G. E.; AlexanderWebber, J. A.; Abate, A.; Sadhanala, A.; Yao, S.; Chen, Y.; Friend, R.

H.; Estroff, L. A.; Wiesner, U.; Snaith, H. J. Ultrasmooth OrganicInorganic Perovskite Thin-Film Formation and Crystallization for Efficient Planar Heterojunction Solar Cells. Nat. Commun. 2015, 6, 6142. (38) Wang, S.; Ono, L. K.; Leyden, M. R.; Kato, Y.; Raga, S. R.; Lee, M. V.; Qi, Y. Smooth Perovskite Thin Films and Efficient Perovskite Solar Cells Prepared by the Hybrid Deposition Method. J. Mater. Chem. A 2015, 3, 14631−14641. (39) Zhang, M.; Yu, H.; Yun, J. H.; Lyu, M.; Wang, Q.; Wang, L. Facile Preparation of Smooth Perovskite Films for Efficient Meso/ Planar Hybrid Structured Perovskite Solar Cells. Chem. Commun. 2015, 51, 10038−10041. (40) Pascoe, A. R.; Meyer, S.; Huang, W.; Li, W.; Benesperi, I.; Duffy, N. W.; Spiccia, L.; Bach, U.; Cheng, Y.-B. Enhancing the Optoelectronic Performance of Perovskite Solar Cells via a Textured CH3NH3PbI3 Morphology. Adv. Funct. Mater. 2016, 26, 1278−1285. (41) Li, W.; Fan, J.; Li, J.; Mai, Y.; Wang, L. Controllable Grain Morphology of Perovskite Absorber Film by Molecular Self-Assembly toward Efficient Solar Cell Exceeding 17%. J. Am. Chem. Soc. 2015, 137, 10399−10405.

3676
