Improved Reproducibility for Perovskite Solar Cells with 1 cm2 Active

Jan 31, 2017 - With rapid progress in recent years, organohalide perovskite solar cells (PSC) are promising candidates for a new generation of highly ...
0 downloads 0 Views 7MB Size
Research Article www.acsami.org

Improved Reproducibility for Perovskite Solar Cells with 1 cm2 Active Area by a Modified Two-Step Process Heping Shen,* Yiliang Wu, Jun Peng, The Duong, Xiao Fu, Chog Barugkin, Thomas P. White, Klaus Weber, and Kylie R. Catchpole Centre for Sustainable Energy System, Research School of Engineering, The Australian National University, Canberra, Australia S Supporting Information *

ABSTRACT: With rapid progress in recent years, organohalide perovskite solar cells (PSC) are promising candidates for a new generation of highly efficient thinfilm photovoltaic technologies, for which up-scaling is an essential step toward commercialization. In this work, we propose a modified two-step method to deposit the CH3NH3PbI3 (MAPbI3) perovskite film that improves the uniformity, photovoltaic performance, and repeatability of large-area perovskite solar cells. This method is based on the commonly used two-step method, with one additional process involving treating the perovskite film with concentrated methylammonium iodide (MAI) solution. This additional treatment is proved to be helpful for tailoring the residual PbI2 level to an optimal range that is favorable for both optical absorption and inhibition of recombination. Scanning electron microscopy and photoluminescence image analysis further reveal that, compared to the standard two-step and one-step methods, this method is very robust for achieving uniform and pinhole-free largearea films. This is validated by the photovoltaic performance of the prototype devices with an active area of 1 cm2, where we achieved the champion efficiency of ∼14.5% and an average efficiency of ∼13.5%, with excellent reproducibility. KEYWORDS: perovskite solar cells, large area, modified two-step method, photoluminescence image, uniformity, repeatability approximately 20% based on 1 cm2 active area.17 From these attempts, we can see that one of the most important and challenging factors to achieve high performance PSCs is to make continuous, pinhole-free, and uniform perovskite films over a large area. One of the most commonly used deposition techniques for perovskite films on different small-area substrates is the convenient and inexpensive two-step (sequential) deposition method,19−22 for which there has also been progress on large-scale device fabrication. Laxman Gouda et al.15 used MAI vapor to heal the perovskite films (CH3NH3PbI3−xClx) and optimized the treatment temperature, achieving uniform films and a cell efficiency of 15.6%, although this was for small area devices. Yuanyuan Zhou et al. developed a successive spin coating/annealing (SSCA) process to ensure uniform solidstate conversion repeatedly infiltrating MAI into a nanoporous PbI2 film, resulting in maximum power conversion efficiency (PCE) close to 15% based on small area.23 Yang Zhou et al. developed an up-scalable method based on a vapor-assisted solution process, achieving the highest efficiency of 12.6% and an average efficiency of 10.9% over 1 cm2 size.12 A two-step ultrasonic spray method has also been developed to deposit high-quality and large-area perovskite films, resulting in an efficiency of 13.1% for a 1 cm2 cell.24 Recently, Hongseuk Lee

1. INTRODUCTION Since the pioneering work in 2009 by Miyasaka et al.,1 organic− inorganic lead halide perovskite solar cells (PSCs) have received a great deal of attention due to the simple fabrication processes and rapidly improving efficiency reaching 22.1% (certified value) in early 2016.2 However, both the certified and most of the reported efficiencies3−6 in the literature for PSCs are based on relatively small active areas, usually around 0.2 cm2 or even smaller. As a competitive candidate for commercialization, it is urgent that more effort is devoted to large-scale PSC device fabrication. There have been a few investigations into fabrication methods for large area PSCs, including doctorblading,7 inkjet printing,8 slot-die coating,9 spray deposition,10 electrodeposition,11 and spin-coating.12−18 Huang et al.7 improved the doctor-blading method by controlling the formulation of the precursor inks to remove impurities and employing a high deposition temperature to guide the nucleation and grain growth process, achieving continuous, pinhole free and phasepure perovskite films on large area substrates with the highest efficiency of 15.1%. Kai Zhu et al.13 reported a solution route that achieves better control over the decoupling of the nucleation/crystallization and the grain-growth processes through a nonstoichiometric CH3NH3PbI3 precursor, and obtained the highest efficiency of 16.3% for a planar PSC with a 1.2 cm2 active area. Recently, another breakthrough was achieved based on the spin-coating method where vacuum flash is used to assist the crystallization, enabling efficiencies of © 2017 American Chemical Society

Received: October 30, 2016 Accepted: January 31, 2017 Published: January 31, 2017 5974

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

Figure 1. Schematic of the synthesis procedure of perovskite films based on the modified two-step method, where a new process, step 3, is introduced into the standard two-step method. (Here, In-TiOx represents compact indium doped TiOx layer and mp-TiO2 represents mesoporous TiO2 layer.)

was added dropwise into solution A with strong stirring. Second, the indium source precursor was prepared as follows: 90 mg of indium acetate was added to a round-bottom flask and degassed for 6 h, followed by addition of 6 mL of anhydrous pyridine. The solution was heated at 40 °C and stirred for 48 h. The resulting transparent solution was then filtered with a 0.45 μm pore-sized PTFE filter. Finally, 3% (v/v) of this indium source precursor solution was added to 97% (v/v) TiO2 precursor solution, and the resulting solution was stirred at room-temperature for 6 h before use. The spin-coated films were sintered at 500 °C for 30 min and left to cool to room temperature. An 80 nm-thick mesoporous TiO2 layer was deposited on the cp-TiO2 film by spin-coating the TiO2 paste solution in ethanol (1:12 of weight ratio) at 5000 rpm for 25 s with an acceleration rate of 5000 rpm/s. The films were then sintered at 500 °C for 30 min again and left to cool to room temperature. 2.2.1. One-Step Method to Prepare Perovskite Film. We used the optimized parameters that have been developed in our lab for the one-step preparation method for perovskite.30 In detail, 1 M equimolar of MAI and PbI2 were mixed with 7/3 (v/v) of N,N-dimethylformamide (DMF) and dimethyl sulfoxide (DMSO). Inside the glovebox with N2 atmosphere, the solution was spin-coated on the substrate at 5000 rpm/s, and chlorobenzene was dropped on the substrate for fast crystallization 10 s after the spin coating commences.31 The substrate was then annealed at 100 °C for 10 min. 2.2.2. Two-Step Method and Modified Two-Step Method to Prepare Perovskite Film. For the standard two-step method, we used parameters commonly used in the literature.19−22 1.1 M PbI2 solution in DMF was deposited onto the mp-TiO2 film by spin-coating at 5000 rpm for 60 s, followed by drying at 100 °C for 30 min. After cooling to room temperature, the films were dipped in a 10 mg/mL MAI solution in IPA and rinsed with pure IPA for 15 s. To check the effect of dipping time on the amount of residual PbI2, 5 min, 15 and 25 min of dipping were also used. The films were then sintered on the hot plate at 100 °C for 30 min. We refer to the above process as the standard two-step method. For the modified two-step method, samples were first dipped in the 10 mg/mL MAI solution for 1 min, and then a higher concentration MAI solution (20 mg/mL) was spin-coated onto the substrate with a speed of 5000 rpm/s before the film was sintered at 100 °C on a hot plate for 30 min. The hole transport material (HTM) solution was prepared by dissolving 72.3 mg Spiro-OMeTAD, 17.5 μL of a solution of 520 mg/mL (trifluoromethane)sulfonimide lithium salt (LiTFSI) in acetonitrile and 28.8 μL of 4-tert-butylpyridine (TBP) in 1 mL of anhydrous chlorobenzene. The as-prepared perovskite films were then spin-coated with HTM solution at 3000 rpm

reported using NaCl in precursor solution to get a large areauniform film by retarding the perovskite crystallization.25 It is apparent that, despite progress to date, efforts devoted to improving the film quality and performance for large-scale solar cells based on either the two-step method26 or the other alternatives lag significantly behind those for the small areas, and the photovoltaic performance is still far from satisfactory. In this work, we introduce a modified two-step method that leads to pinhole free and much more uniform large-scale films than the standard two-step process. The detailed process for this method is illustrated in Figure 1, where the photoanode was first coated with PbI2 (step 1), then dipped into 10 mg/mL MAI solution (step 2), followed by spin-coating with a more concentrated MAI solution (20 mg/mL; step 3). The combination of step 1 and step 2 is the above-mentioned standard two-step method. XRD patterns reveal that this modified twostep method is favorable for converting the superfluous PbI2 crystals remaining after the standard two-step method into perovskite, which is also beneficial for the light absorption. The as-prepared films on the centimeter scale are not only pinhole free, but also have much better uniformity and repeatability compared to both the standard two-step and the widely used solvent engineered one-step method,27−29 as demonstrated by PL image analysis. After being incorporated into solar cells, impressive power-conversion efficiency (PCE) for the 1 cm2 device based on this modified two-step is obtained, with an average efficiency of ∼13.5% and the highest efficiency reaching 14.5%.

2. MATERIALS AND METHODS 2.1. Materials. All of the chemicals were used as purchased without further purification. MAI powder and mesoporous TiO2 paste (30 NR-D) were purchased from Dyesol. 2,2′,7,7′Tetrakis(N,N-di-4-methoxyphenylamino)-9,9′-spirobifluorene (Spiro-OMeTAD) was purchased from Luminescence Technology. All the other materials were purchased from SigmaAldrich. 2.2. Device Fabrication. FTO substrates were cleaned sequentially with detergent, acetone, iso-propyl alcohol (IPA), and ethanol in ultrasonic bath, with each process lasting for 15 min. They were then treated with UV-ozone for 15 min immediately before all the films were deposited. Next, ∼50 nm indium-doped TiO2 (denoted as In-TiOx) compact layers were prepared by spin coating the In-TiOx precursor solution onto the substrates at 2000 rpm for 30 s. The precursor solution was prepared as follows.30 First, the TiO2 precursor solution was prepared. A total of 369 μL of titanium isopropoxide (TTIP) was diluted with 2.53 mL of anhydrous IPA (solution A); 35 μL of 2 M HCl was added into 2.53 mL of anhydrous IPA (solution B); and then, solution B 5975

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

Figure 2. (a) Illustration of the device structure used in our work. (b) XRD patterns fabricated with different methods, including one-step method, standard two-step method by dipping PbI2 film in MAI solution for different times (1, 5, 15, and 25 min), and modified two-step method. (c) Magnified view of the low angle range (10−30 degree) for three typical films based on one-step, standard two-step with 25 min dipping, and modified two-step methods. (d) Absorbance of films fabricated with different methods.

The current−voltage characteristics of the cells were measured using a Solar Simulator model no. SS150 with a Nova Scan current/voltage source. The light intensity was calibrated at one Sun (100 mW/cm2, AM1.5G) using a certified FraunhoferCalLab reference cell. The perovskite cell was illuminated and maintained at open-circuit condition until it reached a steadystate prior to the measurement.

for 60 s. Devices were then placed in a drybox (air humidity in the range of 1% to 5%) for 12 h prior to thermal evaporation of 100 nm Au electrodes (under vacuum of ∼10−6 Torr, at a rate of ∼0.1 nm/s) to complete the solar cells. 2.3. Characterization. XRD measurements of thin films were done using a PANalytical X’Pert Pro system with grazing incidence angle configuration, which is operated at 30 kV, 10 mA at 2θ (Cu Kα) 10−80°, step 0.02°, and scan speed 2.3° min−1. Absorption spectra were measured with a PerkinElmer Lambda 1050 spectrophotometer with an integrating sphere detector. A FEI Verios scanning electron microscope (SEM) was used to investigate the surface morphology of samples. A Helios Nanolab 600 FIB system was used to prepare cross-sectional SEM images of the cells. For PL imaging: the as-prepared films were held in a nitrogen-filled, temperature controlled jig. The jig is mounted in a home-built PL imaging system and uniformly illuminated with two 430 nm royal-blue LED chips, both filtered by bandpass filters (451/106 nm). Following illumination, PL images of the films were recorded at 3 s intervals with 1 s exposure time using a Peltier-cooled (−70 °C) Si CCD camera (Princeton Instruments Pixis 1024) with a long-pass filter (750 nm). A 10 mg/mL poly(methyl methacrylate) (PMMA) solution in chlorobenzene was spin-coated onto the perovskite film with a speed of 3000 rpm for 1 min before the PL images were taken.

3. RESULTS AND DISCUSSION At present, most of the reported highest efficiencies are based on mesoporous TiO2, with the layer structure FTO/ cp-TiO2/mp-TiO2/perovskite/Spiro-OMeTAD/Au (shown in Figure 2a).4,32 More importantly, the mesoporous TiO2 is also found to be favorable for perovskite crystallization.19 Therefore, in our work we employed this structure to fabricate solar cells. The films investigated in the following are also based on this structure (except that the top HTM and Au layers are not deposited), so that the crystallization conditions for the perovskite will be the same. It is also important to note that in our cell fabrication, we employed In-TiOx as the compact layer instead of pure TiO2 developed in our group, which is favorable for achieving highly efficient solar cells. First, indium doping increases the carrier density and hall mobility, and hence the conductivity of the In-TiOx electron transport layer (ETL) compared to the standard TiO2 ETL. This is favorable 5976

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

Consistent with the XRD results, the optical absorption is also progressively improved by extending the dipping time in the two-step method (Figure 2d). It reaches a maximum when the modified two-step method is employed, thanks to a more thorough reaction between PbI2 and MAI to form perovskite. Better light absorption properties will lead to higher current yield in the solar cell, which is discussed later. The modified two-step method also showed higher absorption than that based on the one-step method, which is due to a slightly thicker perovskite film, shown in the cross-section SEM picture (Figure 6b and Figure S4). The morphology and uniformity of the as-prepared films revealed by SEM images are illustrated in Figure 3. We observe

for better charge collection efficiency in the device. Second, the indium-doping makes it possible to tune the work function of the ETL, and it has been proven that the doping concentration used in this work supplies good band alignment at the ETL/ perovskite interface.30 The phase compositions in the as-prepared films using different fabrication methods are illustrated in the XRD patterns (Figure 2(b, c)). The typical peaks for PbI2 and MAPbI3 are located at ∼12.6° (corresponding to (001) face) and ∼14° (corresponding to (110) face), respectively. In the one-step method, we used a stoichiometric ratio for PbI2 and MAI (1:1 molar ratio) in the precursor, and therefore the resulting film consisted of almost pure perovskite phase. The small amount of PbI2 detected for the one-step method is due to decomposition of perovskite during heat treatment.33 However, for the standard two-step method, much more residual PbI2 was observed. We first adopted the widely used dipping time (1 min) to prepare the film and observed a sizable amount of unreacted PbI2 in the as-prepared film. Therefore, we tried extending the dipping time to 5, 15, and 25 min to trace the XRD evolution of the perovskite film. With increasing dipping time, the intensity of the PbI2 peaks (such as ∼12.6°) is inhibited gradually, and correspondingly the perovskite peaks are enhanced (such as ∼14° for (110) and ∼ 50.4° for (404)).34 Nevertheless, the peak for the residual PbI2 after dipping for 25 min is still relatively high. It is well-known that MAI molecules have a limitation for diffusion through solid films, thus typically some residual PbI2 is observed in the as-prepared films.35,36 It should be pointed out that we consistently observe relatively more PbI2 than reported by some other groups. One of the likely reasons is that we use a glovebox to prepare the films, and there is always some IPA vapor generated during the dipping process. It is also noteworthy that there is some evidence that this unreacted remnant PbI2 could have a passivation effect for perovskite films resulting in inhibited recombination, thus being helpful for solar cell performance.37,38 Especially, it has been clearly proved that PbI2 with controllable amount locating at grain boundary could effectively passivate the perovskites with a longer PL lifetime and enhanced Voc.39 However, the optimum amount of PbI2 is critical because too much excess PbI2 may partially block the electron transport and reduce the absorption efficiency of the as-prepared films because of its wide band gap (2.34 eV). Some work on the effect of excess PbI2 by intentionally adding extra PbI2 into the precursors revealed that the optimal amount of excess PbI2 in the precursor ranges from 5% to 10%, depending on the perovskite material system (pure MAPbI3 or mixed perovskite) and the fabrication method.40−42 For the present two-step method, even if 25 min dipping is used, the residual PbI2 is still far above the optimum range.37 This issue can be avoided by the modified two-step step we introduce here. By dipping for 1 min and further treating the sample with highly concentrated MAI, the residual PbI2 is much reduced, to an amount that is comparable with that in the one-step method (Figure 2c). A long dipping time is detrimental to the film sintering process in that it leads to increased IPA vapors in the glovebox that decompose the perovskite film. Therefore, to reduce the effect caused by IPA vapor exposure, in the following results, a dipping time of 1 min is used to fabricate the standard two-step perovskite films. Skipping the 1 min dip and spin-coating the highly concentrated MAI directly onto PbI2 was also tried, but leads to visibly nonuniform films (shown in Figure S1).

Figure 3. SEM images of films fabricated by different methods: (a) one-step method, (b) standard two-step method with 1 min dipping, and (c) modified two-step method; SEM images of perovskite films similar to those in (a) and (c) in a larger view (∼5 μm scale). (d) One-step method (pin-holes are pointed out with red arrows) and (e) modified two-step method.

that both one-step and modified two-step methods yield relatively higher quality films with no pin-holes observed at around 1 μm scale, although there is a clear difference in the morphology. The modified two-step method produces perovskite crystals of a relatively smaller size, with an average value of ∼200 nm (compared to ∼310 nm for that from the one-step method). However, the morphology and crystal size differences observed here do not significantly affect the cell performance, as discussed later. It should be noted that in a larger view (∼5 μm scale) some pin-holes are observed in the one-step method films, while none were found in the modified two-step method films (Figure 3e,d). This demonstrates that the modified two-step method is superior for preparing high quality pinhole free perovskite films over large areas. We next investigated the film uniformity at a larger (millimeter) scale using PL image analysis (Figure 4). The PL images were normalized to the maximum intensity in each image in order to make comparisons among the different methods. It can be seen that the one-step method resulted in substantial spatial variation across the film from the middle to the edges. This is due to the use of the antisolvent method: the application of chlorobenzene in the center of the sample during the spin coating process creates a difference in the nucleation/ crystallization behavior of the film from the center to the edge, sometimes resulting in a round mark on the film.43 The film prepared with the standard two-step method was spatially more uniform, but still results in some small extremely bright spots against a much darker background. The modified two-step 5977

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

Figure 4. PL images of films fabricated by different methods: (a) one-step method, (b) standard two-step method with 1 min dipping, and (c) modified two-step method. The intensities have been normalized to the brightest pixel in each image.

Figure 5. (a) Statistical distribution of the standard deviation of the PL obtained from seven normalized PL images for each different fabrication method. The dots represent the spatial variation in PL intensity of a single film. (b) SEM cross-sectional image of a device based on the modified two-step method. Scale bar in the right corner is 100 nm. (c) Photo of the large-area solar cells based on one-step method, two-step method, and modified two-step method from the top line to the bottom line, respectively. (d) J-V curves for the champion devices based on different methods using reverse scanning with a rate of 50 mV/s.

method gave the most uniform film with the smallest variation across the sample. We note that the PL images shown here are intended to provide evidence of film uniformity rather than to identify pinholes which have already been clearly shown by the SEM images in Figure 3. Quantitative analysis of the above-mentioned PL images is shown in Figure 5a. In this plot, the normalized PL standard deviation (y-axis) defines the spatial uniformity of a single film (represented by a single dot in Figure 5a), calculated from the pixel-by-pixel PL intensity data of an entire image, where the size of an individual pixel is 14.6 μm. The same analysis is

repeated on multiple films (7 films for each method) to check the reproducibility, represented by the statistical distribution of dots in Figure 5a. The modified two-step method gave the smallest standard deviation with an average value of ∼0.3, compared to >0.4 for the one-step method and ∼0.7 for the standard two-step method. This further confirms that the modified two-step method produces the most uniform films. In addition, the reproducibility of the modified two-step method is also very impressive, with the standard deviation distribution much narrower than that of the standard two-step method and comparable to that of the one-step method despite the 5978

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

for the 1 cm2 devices than those that are otherwise identical but of smaller area (Figure S5). There is no significant difference between the hysteresis for the devices based on the present method and the one-step method (Figure S5), although it is true that there is some difference observed in the steady-state evolution between the two different methods. However, compared to the effect caused by the size of the cell, this difference is much smaller. We attribute this difference in steady-state value to the different hysteresis behavior of these two methods. It is notable that another widely used protocol to assess the hysteresis which is based on the loop size of the reverse and forward scanning curve under moderate scanning rate was also carried out and shown in Figure S 5 (a). It is observed that the modified two-step shows slightly smaller loop indicating smaller hysteresis. Overall, we think this slightly different hysteresis behavior which is linked to the fabrication method could be aroused by the composition and energy level differences, which is worth further study.44 In detail, the modified-two step method gave the best PCE reaching 14.5% with Voc = 1.07 V, Jsc = 19.6 mA cm−2, and FF = 0.690. The standard two-step method resulted in significantly poorer cell performance with Jsc = 17.8 mA cm−2, which is partially ascribed to the smaller absorption shown in Figure 2d. The pinholes existing in the films also lead to more serious recombination which is detrimental for charge collection, thus contributing to both smaller current and FF. Too much PbI2 existing in the perovskite film also tends to lead to smaller shunt resistance and thus more severe recombination.40 Despite the comparable photovoltaic performance of the champion cells produced by the one-step method, the repeatability and correspondingly the average values of all the photovoltaic parameters are poorer than the modified two-step method, as discussed below. We next checked the repeatability of cell performance from different methods, for which the statistical distributions of the photovoltaic parameters are shown in Figure 6. Consistent with

additional processing steps. This indicates that the modified two-step method is the superior process, enabling both low spatial variation across each film, and low variation between films. The films were then incorporated into solar cells with the structure shown in Figure 2a, with identical deposition conditions for every layer except the perovskite layer. Figure 5b shows an SEM image of the cross-section of a modified two-step method solar cell, where the thicknesses of the layers are ∼60 nm for In-TiOx, ∼70 nm for mp-TiO2, ∼300 nm for perovskite capping layer, ∼180 nm for Spiro-OMeTAD, and ∼100 nm for Au, respectively. A cross-section of a device based on the one-step method is also shown in Figure S4. All of the other layers have similar thickness except that the perovskite capping layer thickness is slightly smaller (∼260 nm) for the one-step method. It should be noted that the thicknesses for the perovskite in the device based on the one-step method has been optimized in our lab.24 The cells were fabricated into large-scale devices with an active area of around 1 cm2, with the design shown in Figure 5c. The current−voltage (I-V) curves for the champion cells are shown in Figure 5d, and all the parameters including open-circuit voltage (Voc), short-circuit current (Jsc), fill factor (FF), and PCE are listed in Table 1. We Table 1. Photovoltaic Parameters for the best 1 cm2 Sized Perovskite Solar Cells Based on Different Perovskite Fabrication Methods perovskite preparation method

Voc/V

Jsc/mA cm−2

FF

PCE/%

one-step standard two-step modified two-step

1.07 1.07 1.07

19.5 17.8 19.6

0.686 0.530 0.690

14.2 10.1 14.5

measured the cell performance by reverse scanning with a reasonably slow scanning rate (50 mV/s) without prebiasing, which enables a fair comparison between the different methods. The as-prepared devices show some hysteresis, more pronounced

Figure 6. Statistical distributions of the photovoltaic parameters for perovskite solar cells based on different perovskite deposition methods. 5979

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces

Table 2. Photovoltaic Parameters for the 1 cm2 Sized Perovskite Solar Cells Based on Different Perovskite Deposition Methods (Average Performance and Standard Deviation) perovskite preparation method

Voc/V

Jsc/mA cm−2

FF

PCE/%

one-step standard two-step modified two-step

0.991 ± 0.149 0.936 ± 0.365 1.062 ± 0.007

18.9 ± 0.34 15.0 ± 1.65 19.5 ± 0.55

0.58 ± 0.11 0.41 ± 0.08 0.65 ± 0.02

11.2 ± 4.1 6.0 ± 2.4 13.5 ± 0.84

Notes

the SEM and PL image analysis, where the modified two-step method gave the highest quality perovskite film with no pinholes and high uniformity, the photovoltaic performance of the modified two-step cells is also the highest and showed the smallest variation from sample to sample. The photovoltaic parameters for the 1 cm2 sized perovskite solar cells based on different perovskite deposition methods including both average performance and the standard deviation are listed in Table 2. Specifically, the devices based on the modified two-step method showed an average Voc = 1.062 ± 0.007 V, an average Jsc = 19.5 ± 0.55 mA cm−2, an average FF = 0.65 ± 0.02, and an average PCE of 13.5 ± 0.84%. Notably, compared to reports on various deposition methods in the literature, the present modified twostep method shows superior repeatability.17,45 Not surprisingly, the one-step method showed lower average efficiency of 6.0% and higher standard deviation of 2.4%, with all the other parameters exhibiting larger variations. It is noteworthy that average Jsc and FF values of the devices based on the one-step method are also poorer, with Jsc = 18.9 ± 0.34 mA cm−2 and FF = 0.58 ± 0.11. The corresponding average efficiency of the one-step method is 11.2% with a standard deviation of 4.1%.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work has been supported by the Australian Government through the Australian Renewable Energy Agency (ARENA). Responsibility for the views, information or advice expressed herein is not accepted by the Australian Government. Part of the experiment was performed at Australian National Fabrication Facility (ANFF) ACT Node and Centre for Advanced Microscopy (CAM) at the Australian National University. K.R.C. acknowledges the support of ARC Future Fellowship.



(1) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal Halide Perovskites as Visible-light Sensitizers for Photovoltaic Cells. J. Am. Chem. Soc. 2009, 131, 6050−6051. (2) NREL Solar Efficiency Chart: 2016 [accessed April 26th 2016]; Available from: http://www.nrel.gov/ncpv/images/efficiency_chart. jpg. (3) McMeekin, D. P.; Sadoughi, G.; Rehman, W.; Eperon, G. E.; Saliba, M.; Hörantner, M. T.; Haghighirad, A.; Sakai, N.; Korte, L.; Rech, B.; Johnston, M. B.; Herz, L. M.; Snaith, H. J. A Mixed-cation Lead Mixed-halide Perovskite Absorber for Tandem Solar Cells. Science 2016, 351, 151−155. (4) Seok, Y. W.; Hong, N. J.; Joong, J. N.; Chan, K. Y.; Seungchan, R.; Jangwon, S.; Seok, I. S. High-performance Photovoltaic Perovskite Layers Fabricated through Intramolecular Exchange. Science 2015, 348, 1234−1237. (5) Saliba, M.; Matsui, T.; Seo, J. Y.; Domanski, K.; Correa-Baena, J. P.; Nazeeruddin, M. K.; Zakeeruddin, S. M.; Tress, W.; Abate, A.; Hagfeldt, A.; Grätzel, M. Cesium-containing Triple Cation Perovskite Solar Cells: Improved Stability, Reproducibility and High Efficiency. Energy Environ. Sci. 2016, 9, 1989−1997. (6) Zhou, H. P.; Chen, Q.; Li, G.; Luo, S.; Song, T. B.; Duan, H. S.; Hong, Z. R.; You, J. B.; Liu, Y. S.; Yang, Y. Interface Engineering of Highly Efficient Perovskite Solar Cells. Science 2014, 345, 542−546. (7) Deng, Y. H.; Peng, E.; Shao, Y. C.; Xiao, Z. G.; Dong, Q. F.; Huang, J. S. Scalable Fabrication of Efficient Organolead Trihalide Perovskite Solar Cells with Doctor-bladed Active Layers. Energy Environ. Sci. 2015, 8, 1544−1550. (8) Li, S. G.; Jiang, K. J.; Su, M. J.; Cui, X. P.; Huang, J. H.; Zhang, Q. Q.; Zhou, X. Q.; Yang, L. M.; Song, Y. L. Inkjet Printing of CH3NH3PbI3 on a Mesoscopic TiO2 Film for Highly Efficient Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 9092−9097. (9) Hwang, K.; Jung, Y. S.; Heo, Y. J.; Scholes, F. H.; Watkins, S. E.; Subbiah, J.; Jones, D. J.; Kim, D. Y.; Vak, D. Toward Large Scale Rollto-roll Production of Fully Printed Perovskite Solar Cell. Adv. Mater. 2015, 27, 1241−1247. (10) Das, S.; Yang, B.; Gu, G.; Joshi, P. C.; Ivanov, I. N.; Rouleau, C. M.; Aytug, T.; Geohegan, D. B.; Xiao, K. High-performance Flexible Perovskite Solar Cells by Using a Combination of Ultrasonic Spraycoating and Low Thermal Budget Photonic Curing. ACS Photonics 2015, 2, 680−686. (11) Chen, H. N.; Wei, Z. H.; Zheng, X. L.; Yang, S. H. A Scalable Electrodeposition Route to the Low-cost, Versatile and Controllable Fabrication of Perovskite Solar Cells. Nano Energy 2015, 15, 216−226. (12) Yang, Z.; Cai, B.; Zhou, B.; Yao, T. T.; Yu, W.; Liu, S. Z.; Zhang, W. H.; Li, C. An Up-scalable Approach to CH3NH3PbI3 Compact

4. CONCLUSION In summary, we show that the modified two-step deposition method presented here leads to high quality perovskite films over large areas, with few pinholes and outstanding uniformity. The amount of residual PbI2 existing in the perovskite film prepared by this method is minimized to support both charge transport and light absorption in the device. Correspondingly, impressive photovoltaic performance and excellent repeatability were achieved based on this method. Our work revealing the beneficial effects caused by the post-treatment with high concentration MAI is not limited to spin coating, but could be used to improve the uniformity for a variety of deposition techniques. Therefore, it opens up more possibility to make highly efficient large area perovskite solar cells, thus paving the way for largescale fabrication of both single cells and also modules.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b13868. Cross-sectional image of the solar cell based on the one-step method and the hysteresis of the cells (PDF)



REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Heping Shen: 0000-0002-8409-8839 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. 5980

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981

Research Article

ACS Applied Materials & Interfaces Films for High-performance Perovskite Solar Cells. Nano Energy 2015, 15, 670−678. (13) Yang, M. J.; Zhou, Y. Y.; Zeng, Y. N.; Jiang, C. S.; Padture, N. P.; Zhu, K. Square-Centimeter Solution-processed Planar CH3NH3PbI3 Perovskite Solar Cells with Efficiency Exceeding 15%. Adv. Mater. 2015, 27, 6363−6370. (14) Chen, W.; Wu, Y. Z.; Yue, Y. F.; Liu, J.; Zhang, W. J.; Yang, X. D.; Chen, H.; Bi, E. B.; Ashraful, I.; Grätzel, M.; Han, L. Y. Efficient and Stable Large-area Perovskite Solar Cells with Inorganic Charge Extraction Layers. Science 2015, 350, 944−948. (15) Gouda, L.; Gottesman, R.; Tirosh, S.; Haltzi, E.; Hu, J. G.; Ginsburg, A.; Keller, D. A.; Bouhadana, Y.; Zaban, A. Vapor and Healing Treatment for CH3NH3PbI3−xClx Films toward Large-area Perovskite Solar Cells. Nanoscale 2016, 8, 6386−6392. (16) Qiu, W.; Merckx, T.; Jaysankar, M.; Huert, C. M.; Rakocevic, L.; Zhang, W.; Paetzold, U. W.; Gehlhaar, R.; Froyen, L.; Poortmans, J.; Cheyns, D.; Snaith, H. J.; Heremans, P. Pinhole-free Perovskite Films for Efficient Solar Modules. Energy Environ. Sci. 2016, 9, 484−489. (17) Li, X.; Bi, D. Q.; Yi, C. Y.; Décoppet, J. D.; Luo, J. S.; 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. (18) Wu, Y. Z.; Yang, X. D.; Chen, W.; Yue, Y. F.; Cai, M. L.; Xie, F. X.; Bi, E. B.; Islam, A.; Han, L. Y. Perovskite Solar Cells with 18.21% Efficiency and Area over 1 cm2 Fabricated by Heterojunction Engineering. Nature Energy 2016, 1, 16148. (19) Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Grätzel, M. Sequential Deposition as a Route to High-performance Perovskite-sensitized Solar Cells. Nature 2013, 499, 316−319. (20) Im, J. H.; Jang, I. H.; Pellet, N.; Grätzel, M.; Park, N. G. Growth of CH3NH3PbI3 Cuboids with Controlled Size for High-efficiency Perovskite Solar Cells. Nat. Nanotechnol. 2014, 9, 927−932. (21) Ip, A. H.; Quan, L. N.; Adachi, M. M.; McDowell, J. J.; Xu, J. X.; Kim, D. H.; Sargent, E. H. A Two-step Route to Planar Perovskite Cells Exhibiting Reduced Hysteresis. Appl. Phys. Lett. 2015, 106, 143902. (22) Ko, H. S.; Lee, J. W.; Park, N. G. 15.76% Efficiency Perovskite Solar Cells Prepared under High Relative Humidity: Importance of PbI2 Morphology in Two-step Deposition of CH3NH3PbI3. J. Mater. Chem. A 2015, 3, 8808−8815. (23) Zhou, Y. Y.; Yang, M. J.; Vasiliev, A. L.; Garces, H. F.; Zhao, Y. X.; Wang, D.; Pang, S. P.; Zhu, K.; Padture, N. P. Growth Control of Compact CH3NH3PbI3 Thin Films via Enhanced Solid-State Precursor Reaction for Efficient Planar Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 9249−9256. (24) Huang, H. B.; Shi, J. J.; Zhu, L. F.; Li, D. M.; Luo, Y. H.; Meng, Q. B. Two-step Ultrasonic Spray Deposition of CH3NH3PbI3 for Efficient and Large-area Perovskite Solar Cell. Nano Energy 2016, 27, 352−358. (25) Lee, H. S.; Kim, A.; Kwon, H. C.; Yang, W. S.; Oh, Y. J.; Lee, D. H.; Moon, J. Retarding Crystallization during Facile Single Coating of NaCl-Incorporated Precursor Solution for Efficient Large-Area Uniform Perovskite Solar Cells. ACS Appl. Mater. Interfaces 2016, 8, 29419−29426. (26) Zhao, Y. X.; Zhu, K. Three-step Sequential Solution Deposition of PbI2-free CH3NH3PbI3 Perovskite. J. Mater. Chem. A 2015, 3, 9086−9091. (27) Liu, D. Y.; Kelly, T. L. Perovskite Solar Cells with a Planar Heterojunction Structure Prepared Using Room-temperature Solution Processing Techniques. Nat. Photonics 2013, 8, 133−138. (28) Zhao, X. Y.; Shen, H. P.; Zhang, Y.; Li, X.; Zhao, X. C.; Tai, M. Q.; Li, J. F.; Li, J. B.; Li, X.; Lin, H. Aluminum-doped Zinc Oxide as Highly Stable Electron Collection Layer for Perovskite Solar Cells. ACS Appl. Mater. Interfaces 2016, 8, 7826−7833. (29) Bae, S.; Han, S. J.; Shin, T. J.; Jo, W. H. Two Different Mechanisms of CH3NH3PbI3 Film Formation in One-step Deposition and Its Effect on Photovoltaic Properties of OPV-type Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 23964−23972.

(30) Peng, J.; Duong, T.; Zhou, X. Z.; Shen, H. P.; Wu, Y. L.; Mulmudi, H. K.; Wan, Y. M.; Zhong, D. Y.; Li, J. T.; Tsuzuki, T.; Weber, K. J.; Catchpole, K. R.; White, T. P. Efficient Indium-Doped TiOx Electron Transport Layers for High-performance Perovskite Solar Cells and Perovskite-silicon Tandems. Adv. Energy Mater. 2016, 1601768. (31) Xiao, M. D.; Huang, F. Z.; Huang, W. C.; 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 Cells. Angew. Chem. 2014, 126, 10056− 10061. (32) Bi, D. Q.; Yi, C. Y.; Luo, J. S.; Décoppet, J. D.; Zhang, F.; Zakeeruddin, S. M.; Li, X.; Hagfeldt, A.; Grätzel, M. PolymerTemplated Nucleation and Crystal Growth of Perovskite Films for Solar Cells with Efficiency Greater than 21%. Nature Energy 2016, 1, 16142. (33) Sadoughi, G.; Starr, D. E.; Handick, E.; Stranks, S. D.; Gorgoi, M.; Wilks, R. G.; Bär, M.; Snaith, H. J. Observation and Mediation of the Presence of Metallic Lead in Organic−inorganic Perovskite Films. ACS Appl. Mater. Interfaces 2015, 7, 13440−13444. (34) Kosyachenko, L. A. Solar Cells-New Approaches and Reviews; Intech: Croatia, 2015. (35) Hao, F.; Stoumpos, C. C.; Liu, Z.; Chang, R. P. H.; Kanatzidis, M. G. Controllable Perovskite Crystallization at a Gas−solid Interface for Hole Conductor-free Solar Cells with Steady Power Conversion Efficiency over 10%. J. Am. Chem. Soc. 2014, 136, 16411−16419. (36) Wu, Y. Z.; Islam, A.; Yang, X. D.; Qin, C. J.; Liu, J.; Zhang, K.; Peng, W. Q.; Han, L. Y. Retarding the Crystallization of PbI2 for Highly Reproducible Planar-structured Perovskite Solar Cells via Sequential Deposition. Energy Environ. Sci. 2014, 7, 2934−2938. (37) Chen, Q.; Zhou, H. P.; Song, T. B.; Luo, S.; Hong, Z. R.; Duan, H. S.; Dou, L. T.; Liu, Y. S.; Yang, Y. Controllable Self-induced Passivation of Hybrid Lead Iodide Perovskites toward High Performance Solar Cells. Nano Lett. 2014, 14, 4158−4163. (38) Wang, S.; Dong, W.; Fang, X.; Zhang, Q.; Zhou, S.; Deng, Z. H.; Tao, R. H.; Shao, J. Z.; Xia, R.; Song, C.; Hub, L. H.; Zhu, J. Credible Evidence for the Passivation Effect of Remnant PbI2 in CH3NH3PbI3 Films in Improving the Performance of Perovskite Solar Cells. Nanoscale 2016, 8, 6600−6608. (39) Zhang, T. Y.; Guo, N. J.; Li, G.; Qian, X. F.; Zhao, Y. X. A Controllable Fabrication of Grain Boundary PbI2 Nanoplates Passivated Lead Halide Perovskites for High Performance Solar Cells. Nano Energy 2016, 26, 50−56. (40) Kim, Y. C.; Jeon, N. J.; Noh, J. H.; Yang, W. S.; Seo, J. W.; Yun, J. S.; Baillie, A. H.; Huang, S. J.; Green, M. A.; Seidel, J.; Ahn, T. K.; Seok, S. I. Beneficial Effects of PbI2 Incorporated in Organo-lead Halide Perovskite Solar Cells. Adv. Energy Mater. 2016, 6, 1502104. (41) Roldan-Carmona, C.; Gratia, P.; Zimmermann, I.; Grancini, G.; Gao, P.; Graetzel, M.; Nazeeruddin, M. K. High Efficiency Methylammonium Lead Triiodide Perovskite Solar Cells: the Relevance of Non-stoichiometric Precursors. Energy Environ. Sci. 2015, 8, 3550−3556. (42) Adhikari, N.; Dubey, A.; Khatiwada, D.; Mitul, A. F.; Wang, Q.; Venkatesan, S.; Iefanova, A.; Zai, J. T.; Qian, X. F.; Kumar, M.; Qiao, Q. Q. Interfacial Study to Suppress Charge Carrier Recombination for High Efficiency Perovskite Solar Cells. ACS Appl. Mater. Interfaces 2015, 7, 26445−26454. (43) 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. (44) Yu, H.; Lu, H. P.; Xie, F. Y.; Zhou, S.; Zhao, N. Native Defectinduced Hysteresis Behavior in Organolead Iodide Perovskite Solar Cells. Adv. Funct. Mater. 2016, 26, 1411−1419. (45) Yao, K.; Wang, X. F.; Xu, Y. X.; Li, F.; Zhou, L. Multilayered Perovskite Materials Based on Polymeric-ammonium Cations for Stable Large-area Solar cell. Chem. Mater. 2016, 28, 3131−3138.

5981

DOI: 10.1021/acsami.6b13868 ACS Appl. Mater. Interfaces 2017, 9, 5974−5981