REVIEWS SCIENCE CHINA Materials Published online 24 June 2016 | doi: 10.1007/s40843-016-5048-y Sci China Mater 2016, 59(6): 495–506 mater.scichina.com link.springer.com SPECIAL ISSUE: Flexible and Stretchable Energy Flexible organic-inorganic hybrid perovskite solar cells Henry Halim and Yunlong Guo* ABSTRACT The area of organic-inorganic hybrid perovskite has undergone especially intense research and transformation over the past seven years. Although most of the focus is on achieving high power eficiencies (>20%) on rigid substrates by solution process, the lexible version has not been neglected and has also gone through vast improvements in terms of eficiencies and durability. In this review paper, the most recent three years’ developments in lexible perovskite solar cells are covered, showcasing the key of strategies used to transform these cells from rigid to lexible. Future outlook will be presented at the end exhibiting the potential problems that need to be solved in order to send these novel lexible power generators into the future market. Keywords: inorganic-organic hybrid material, perovskite solar cell, lexible device, power conversion eficiency INTRODUCTION The area of inorganic-organic hybrid perovskite solar cell has rapidly developed over the past seven years and power conversion eficiencies (PCE) of over 20% have been achieved [1,2]. The high performance of these cells is due to appropriate band gaps (~1.5 eV) [3,4], high absorption coeficients (~ 103–106 cm−1) [5,6], low exciton binding energies (35–75 meV) [7], long charge diffusion lengths (> 1 μm) [5,8] and high carrier mobilities (>10 cm2 V−1 s−1) of hybrid perovskite materials [9]. Through the use of the material (MA/FA) PbX3 (where MA is CH3NH3, FA is NH2CH=NH2 and X is I, Cl or Br), research developments have gone far enough to the point where issues regarding out of laboratory applications and large-scale productions become major considerations. Unlike the traditional thin-ilm solar cells of silicon, copper indium gallium selenide (CIGS) or CdTe [10] perovskite thin ilm solar cells can be solution processed; an important factor as it means that if all the fabrication steps are conducted at low temperatures, the solar cells can be made on cheap and lexible plastics [11]. Flexibility will allow the use of large-scale production methods such as roll-to-roll printing to lower the costs and also at the same time provide the possibility of power generation for wearable electronics [12]. In addition to cost eficient manufacturing methods, perovskite solar cells also use very cheap and widely available starting materials such as methyl ammonium iodide (MAI) and lead iodide (PbI2) [13,14]. In this review, notable methodologies of achieving lexible perovskite solar cells (FPSCs) are described, with respect to the developments that lead them to their current state. This review will start from the structure of lexible perovskite solar cell, and then proceed with covering the new preparation methods for these cells. Then, based on our own experience of perovskite solar cell based on glass and lexible substrates [15], we will give an outlook on the potential research points required before the eventual application of this technology and inally what possible exciting future applications of FPSCs that await us. STRUCTURE OF FPSC In order to avoid confusion, throughout this paper we will deine and distinguish FPSCs mainly from the viewpoint of their device structures, namely as forward type or inverted type (Fig. 1). For perovskite solar cells, the type of charge collected by the transparent bottom electrode (commonly indium tin oxides (ITO) or luorine tin oxides (FTO)) can be used as a way to distinguish the device’s type. If the electron-transporting layer (ETL) is in contact with the bottom electrode to collect electrons, it is generally called as the forward type structure. In this case, sunlight enters the perovskite layer Department of Chemistry, University of Tokyo, Tokyo 113-0033, Japan * Corresponding author (guoyunlong@chem.s.u-tokyo.ac.jp) © Science China Press and Springer-Verlag Berlin Heidelberg 2016 495 REVIEWS SCIENCE CHINA Materials Figure 1 Two general device structures of perovskite solar cells, left: forward type and right: inverted type. In the middle is the perovskite crystal cell structure, where A is FA+ or MA+; B is Pb2+ or Sn2+; X is I−, Br−, or Cl−. after passing through the transparent bottom electrode and ETL layer. The non-transparent top electrode collects the holes, which are generated from the perovskite layer and are transported through the hole-transporting layer (HTL). On the other hand, if the HTL is in contact with the bottom electrode to collect holes, it can generally be called as the inverted type structure. In this case, sunlight enters the perovskite layer by irst going through the bottom transparent electrode and HTL layer. The top electrode collects electrons, which are generated from the perovskite and transported through the ETL. Forward type Polyethylene terephthalate or polyethylene 2,6-naphthalate substrate During the initial development of perovskite solar cells the vast majority of the perovskite solar cells used forward type structure with metal oxides as electron transporting layers, which required high (approx. 450°C) processing temperatures. Soon after, scientists realized the lowering of processing temperature comes with several important advantages; production costs can be reduced and substrate choices can be widened. The implication of these two advantages will most notably allow the use of plastic substrates such as polyethylene terephthalate (PET) or polyethylene 2,6-naphthalate (PEN) (Fig. 2), which will bestow on perovskite solar cells the aforementioned advantages stated in the introduction. This will serve perovskite solar cells well as it possesses competitive advantage over organic solar cells due to its higher PCE, while at the same time capable of doing things that eficient silicon solar cells cannot 496 Figure 2 Chemical structures of plastic substrates PET and PEN. do such as being semi transparent and lightweight. As perovskite solar cells started with forward type structures, it was logical to simply adopt the same structure with small modiications to processing conditions to fabricate the early lexible cells. The irst lexible perovskite solar cell was developed in 2013, when Kumar et al. [11] successfully fabricated zinc oxide (ZnO) nanorods via a low temperature process. These nanorods were used to substitute the typical TiO2 electron-transporting layer but unfortunately the performance of such device (Fig. 3a) was not so satisfactory. In that report rigid devices based on zinc oxide nanorods only reached about half the performance (PCE = 8.9%) of the best titanium oxide based counterparts (PCE~15%) [11]. In addition, the PCE of the solar cell was even lower when the same zinc nanorod strategy was applied to lexible PET (PCE = 2.62%). This might be due to the hydrophobic properties of PET, which caused the formation of low quality of ZnO layer, further lowering the absorbed photon-to-current conversion eficiency (APCE) of the solar cells (Fig. 3b). In 2014, Liu et al. [16] used an extremely small ZnO nanoparticles (~5 nm) to get a high density and uniform © Science China Press and Springer-Verlag Berlin Heidelberg 2016 REVIEWS SCIENCE CHINA Materials Figure 3 (a) Schematic illustration of the architecture for the perovskite devices fabricated by Kumar et al. [11]. (b) Absorbed photon-to-current conversion eficiency (APCE) for the devices. Reprinted with permission from Ref. [11], Copyright 2013, the Royal Society of Chemistry. (c) Photograph of FPSC with the structure of PET/ITO/ZnO/CH3NH3PbI3/spiro-OMeTAD/Ag [16]. (d) J-V characteristics measured under 100 mW cm−2 AM1.5G illumination (red line) and in the dark (black line) for the best lexible device. Reprinted with permission from Ref. [16], Copyright 2013, Macmillan Publishers Limited. metal oxide layer. The breakthrough of this approach is that it required no calcination or sintering step. Then, a solution processable two-step method [17] was performed on it for the fabrication of a high quality perovskite layer. Under these conditions, a PCE of 15.7% was achieved at low temperature and a corresponding PCE of 10.2% for a lexible device (Figs 3c and d). With regard to TiO2, scientists developed several low-temperature processes, such as using nanocomposite of TiO2 with graphene [18], nanosize rutile type TiO2 [19,20] and atom layer depositing TiO2 [21]. In 2015, Kim et al. [12] used atomic layer deposition to deposit the TiOx layer on PEN surface. A void free TiOx layer was achieved through this method and they were able to fabricate a lexible device exhibiting 12.2% PCE (Fig. 4). In addition to the powerful performance, this device also retained 95% of its original PCE after 1000 bending cycles. In contrast with earlier works in this ield, this cell was much more durable as previous high durability cells had bending test survivability of only one or two hundred cycles. In this work the authors also mentioned the important observation that the cause of the failure in their devices after the bending tests originated from fracture in the ITO layer. They have backed up this statement by fabricating a test sample without ITO that showed no fracture after bending [12], demonstrating that the lexibility of perovskite solar cells can be vastly improved if ITO can somehow be substituted. Meanwhile other alternative methods of fabricating the TiO2 layer continued to be developed. An example of this is a method employing an electron beam to deposit an amorphous TiO2 layer onto the ITO/PET substrate [22]. This method worked well giving a device with PCE 13.5% on lexible PET substrates, however the bending durability was not reported in the paper. Scientist also used magnetron sputtering as a method to deposit a dense layer of TiO2 on PET, which worked especially effectively as an electrontransporting layer offering fast charge transport [23]. This work increased the PCE of FPSCs to 15.07% thanks to the improved electron injection. Recently, Seok group reported a method to use Zn2SnO4 as a novel ETL for lexible solar cells [24]. The fabrication process (shown in Fig. 5a) used temperatures only at or below 100°C. In addition, the formation of these nanopar- © Science China Press and Springer-Verlag Berlin Heidelberg 2016 497 REVIEWS SCIENCE CHINA Materials Figure 4. (a) Cross-sectional SEM image of the inorganic-organic halide perovskite planar heterojunction lexible solar cell and schematic of the lexible device structure. Scale bar: 200 nm. (b) A photograph of an FPSC under bending. (c) J-V characteristics measured under the simulated solar light (100 mW cm−2 AM 1.5G) for the best performing PEN/ITO/TiOx/CH3NH3PbI(3−x)Clx/spiro-MeO-TAD/Ag lexible device. Reprinted with permission from Ref. [12], Copyright 2015, the Royal Society of Chemistry. Figure 5. (a) Schematic illustration of the low-temperature process for fabricating lexible device with ZSO NPs. (b) Cross-sectional SEM image and photograph of the ZSO-based lexible perovskite solar cell (scale bar, 500 nm). (c) Energy levels of the materials used in that study. (d) Photocurrent density-voltage (J-V) curve measured by reverse scan using 10 mV voltage steps and 40 ms delay times under AM 1.5 G illumination. Reprinted with permission from Ref. [24]. ticles did not use high pressures, which made them highly suitable for the application on plastic substrates and also potentially cost effective, large-scale production. After de498 position of perovskite by fast-washing method, following up with HTL as well as gold electrode deposition, a high performance lexible solar cell was achieved with PCE of © Science China Press and Springer-Verlag Berlin Heidelberg 2016 REVIEWS SCIENCE CHINA Materials 15.3% on PEN substrate [24] (see Figs 5b–d). The bending durability (95% of original PCE after 300 bending cycles) did not exactly hold the record at that time, but this work demonstrated the possibility of reaching high PCEs on lexible substrates. Metal substrate Apart from plastic substrates, high temperature resistant metallic substrates (>500°C), used for lexible CIGS or CdTe solar cells, are available and can also be a good choice for FPSCs. However due to the fact that metals are actually not transparent, the top electrode has to be transparent. A couple of attempts have been made using this approach, each with their own reasons for the selection of metal substrates. The irst perovskite solar cell that used metals as the bottom layer was a cell fabricated on lexible titanium substrate [25]. This method allows one to bypass the requirement of a low temperature process in exchange for using an ultra-thin transparent metal electrode on the top. With this, a reasonable PCE of 6% can be achieved with good bending durability since ITO was not necessary and was not used. However because this approach did not show signiicant advantages over low temperature processes, it had been seemingly abandoned. In addition, low temperature processes have also become more common and much more effective in the past few years making plastic substrates much more viable. Moreover, it is also important to remember that low temperature processes will save production costs of the perovskite solar cells. Furthermore, the use of 12 nm silver as transparent electrode may not be a good choice for lead halide perovskite cells due to absorption and stability issues. Years later, metal substrates were selected once again [26] just for the sole goal of ITO substitution. By keeping the forward structure instead of using an inverted structure, the characteristically higher Voc can be retained and allowed this device to reach a PCE of 10.3%. The major improvement of this method is that they used a PET with embedded Ni mesh as a transparent top electrode, which enhanced the Jsc of solar cell greatly when compared to the 12 nm Ag as top transparent electrode. Metallic fiber In terms of fabrication, solar cells are indeed most easily manufactured in a two dimensional architecture. However, solar cells in other formats could also ind novel applications in the future. An interesting concept was proposed by Peng et al. [27], who attempted to fabricate perovskite solar cells on lexible metallic iber (Fig. 6). The iber-shaped perovskite solar cell exhibits a PCE of 3.3%. Later on in 2016, they used TiO2 nanotubes as ETL and deposited the perovskite layer by electrochemical deposition process. The greater density perovskite grains induced a higher PCE of iber solar cell to 7.1% with a Voc of 0.85 V [28]. Figure 6 (a) Structure and (b) energy-level diagram of the iber-shaped perovskite solar cell. MAPbI3 = CH3NH3PbI3, OMeTAD = 2,2',7,7'-tetrakis(N,N-di-para-methoxyphenyl-amine)-9,9-spirobiluorene. (c) Photograph of a sample textile. Reprinted with permission from Ref. [27], Copyright 2014, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim. © Science China Press and Springer-Verlag Berlin Heidelberg 2016 499 REVIEWS SCIENCE CHINA Materials Willow glass Around one year ago, an interesting approach combining stability and lexibility was reported by Tavakoli et al. [29]. A lexible willow glass (~50 μm) was used as substrate and an antirelection and superhydrophobic layer (polydimethylsiloxane, PDMS) was placed on top of the glass (Fig. 7). With this, some water can be avoided in addition to a surface ‘self cleaning’ mechanism, where dirt deposited on the lexible glass can be easily removed by water due to the superhydrophobicity of the layer. Although the lexible glass substrate is temperature resistant, the process used here is also a low temperature process, which is attractive for industrial production. The bending durability was similar to other FPSC benchmarks at that time (96% of original PCE after 200 bending cycles), in addition to good PCE boosted by the antirelection support (13.14% PCE in total). Inverted type Alongside the forward type architecture, the inverted version of the FPSCs has been attempted almost as early as when the irst lexible forward type was developed. During those stages, the development of the inverted type was faster due to the non-use of high temperature processes for this architecture. This means that it was possible to improve the inverted solar cells without waiting for some kind of breakthrough while the forward type somewhat struggled to ind a good way to lower their processing temperatures. Although currently this type does not offer the highest PCE, it is certainly attractive in other signiicant aspects as shown in this section. After the paper regarding the irst forward type perovskite solar cells was published, signiicant improvements in lexibility and PCE was shown to be possible by Bolink et al. [30] who adopted the inverted structure. In 2014 [31], they provided an evaporation method to deposit CH3NH3PbI3 on poly[N,N'-bis(4-butylphenyl)-N,N'bis(phenyl)benzi-dine] (polyTPD)/PEDOT:PSS surface which they applied for this report on a lexible substrate. Through this method, the maximum processing temperature required is only 90°C and the fabricated device can reach a PCE of 7%. This is more than double of the PCE compared to the irst FPSC and it certainly gave inverted type cells a solid spot to start in the ield. The application of PEDOT:PSS as HTL on lexible cells continued to develop steadily and as a result, in early 2014 PEDOT:PSS inverted type solar cells were already able to reach power conversion eficiencies of 9.2% based on CH3NH3PbI3−xClx perovskite [32], most likely due to the better understanding of thin ilm fabrication. However, the bending durability was lacking in these studies. As mentioned in the forward type section, for FPSCs, it would be possible to gain additional lexibility by not using ITO altogether since it is the part that is most susceptible to cracking [12]. Kelly et al. [33] irst tried this approach by substituting the ITO electrode with conducting PEDOT:PSS. Through this method, the fabricated device can survive longer bending cycles without shorting, but the device performance still deteriorated after a few hundred bends. Instead of the electrodes breaking this time, the limiting factor for this device is the formation of cracks in the perovskite layer. These cracks decreased the ill factor quickly as it occurred. In the recent advances of the inverted structure, lexible devices using the same plastic, PEDOT:PSS, perovskite Figure 7 (a) Schematic structure of the perovskite solar cell device with nanocone polydimethylsiloxane (PDMS) ilm attached on the top. (b) J-V measurements of perovskite solar cell devices with and without PDMS nanocone ilm (inset image is the schematic of light scattering in the device with a nanocone layer). Reprinted with permission from Ref. [29], Copyright 2015, American Chemical Society. 500 © Science China Press and Springer-Verlag Berlin Heidelberg 2016 REVIEWS SCIENCE CHINA Materials device structure were able to attain PCE of 12.25% based on CH3NH3PbI3−xClx. After bending 1000 times, the solar cells kept a PCE of 11.9%. This was achieved by improving the perovskite layer through layer-by-layer deposition to make a highly compact and optimally thick layer for great performance [34]. In this research it was found that the PCE of the device increases as the perovskite layer gets up to a certain thickness and then decreases to lower PCEs when the perovskite layer is too thick. The most recent lexible device using PEDOT:PSS inverted structure combined a hexagonal silver mesh along with PH 1000 for an ultra lexible bottom electrode (see Fig. 8a) [35]. It worked very successfully giving a device with 14.2% PCE along with a bending durability of 5000 cycles, bent over the radius of 5 mm. This bending test resulted in only 4.6% degradation of the original PCE (Fig. 8b). The device also exhibits a long-term stability comparable with the ones made on ITO/glass surface. (Fig. 8c). Through this work, the importance of not using transparent metal oxide electrodes on perovskite solar cells was highlighted, if high lexibility were to be achieved. For a more extreme example, Kaltenbrunner et al. [36] took advantage of the availability of ultra-thin PET (1.4 μm) and ultra light materials to create an ultra lexible and ultra light solar cell (Fig. 9a). Their results showed that perovskite solar cells gave the best power-per-weight values when compared to all types of solar cells (Fig. 9b). This type of cell can be used for many additional and perhaps several other niche applications, which they proposed in their paper (Figs 9c and d). Their paper also demonstrated that perovskite solar cell on elastomeric substrates and perovskite powered model planes are possible interesting uses of these light solar cells. Apart from PET or PEN substrate, a shape memory substrate [Noland Optical Adhesive 63 (NOA 63)] has also been once tried for FPSC preparation [37]. In this work the authors’ motive was to retain the integrity of the interface between the layers, which they believed was highly important in perovskite solar cells. With this method the solar cells fabricated were able to not only be bent but also be crumpled (see Fig. 10). A newly prepared sample gave a PCE of 10.2% and after crumpling it can recover its shape Figure 8 (a) Device architecture of the hybrid electrode/PEDOT:PSS (35 nm)/MAPbI3 (B280 nm)/PCBM (B60 nm)/Al (100 nm) cells (left). Crosssection SEM images of the complete perovskite device showing both Ag-mesh area (scale bar, 1 mm) and Flat PET area (scale bar, 500 nm). SEM top-view images of perovskite ilms (right). (b) PCEs of lexible pero-SCs based on both PET/Ag-mesh/PH1000 and PET/ITO electrodes as a function of bending cycles at a radius of 5 mm. (c) Stability of FPSCs based on both PET/Ag-mesh/PH1000 and glass/ITO substrates under room temperature in N2-illed glove box in a timescale of a few hundred hours. PCE values are obtained from statistical distribution of six devices for each condition represented by error bars. Reprinted with permission from Ref. [35]. © Science China Press and Springer-Verlag Berlin Heidelberg 2016 501 REVIEWS SCIENCE CHINA Materials Figure 9 (a) Schematic of the solar cell stack. 1.4-μm-thick PET foils serve as substrate, PEDOT:PSS as the transparent hole selective electrode. Using DMSO as an additive to their PEDOT:PSS solution promotes pinhole-free perovskite layer formation. The MAPI absorber is formed by a one-step solution precursor deposition method. PTCDI or PCBM were tested to be suitable electron-transport layers for this strategy. A chromium layer with accompanying Cr2O3 stabilizes the metal top contact for operation in ambient air. Low-resistivity metals, for example, gold, copper and aluminium, complete the device. Optionally, PU serves as a 1-μm-thick capping layer for mechanical protection. (b) Power-per-weight of ultrathin perovskite solar cells. (c) Close-up photograph of the horizontal stabilizer with integrated solar panel. Scale bar, 2 cm. (d) Snapshot of the model plane during solar-powered outdoor light. Scale bar, 10 cm. Reprinted with permission from Ref. [36], Copyright 2015, Macmillan Publishers Limited. Figure 10 (a) Photograph of the perovskite on shape memory substrate. (b) The performance of perovskite solar cells on shape recoverable polymers before and after crumpling. (c and d) SEM images of perovskite layer after crumpling. Reprinted with permission from Ref. [37], Copyright 2015, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim. 502 © Science China Press and Springer-Verlag Berlin Heidelberg 2016 REVIEWS SCIENCE CHINA Materials and some of its function shown by the decreased PCE of 6.1%. The decrease in PCE was thought to be due to the broken surfaces within the cell (Figs 10b–d). Very recently, a low-temperature processed NiOx was used as HTL for FPSCs. With ITO-PEN substrate, the best device showed a PCE of 13.43% [38]. CONCLUSIONS AND OUTLOOK The area of perovskite solar cells has been a research ield only for seven years and the emergence of FPSCs appeared even more so recently in the past three years but as shown in this review paper, there have already been several very promising approaches that can make future power generation go lexible. Table 1 summarizes the lexible devices mentioned in this review, including both forward type and inverted type. The lexibility of perovskite solar cells promises to bring very important advantages to energy technology, such as cheap manufacture through roll to roll printing, lightweight, potentially color tuning [39,40] and wearable power generation. Out of all the methods provided here, the ‘winner’ of the FPSCs will ultimately strongly depend on the inal application of the perovskite solar cells. If the lexible solar cells are needed just for the sake of making roll to roll production possible, then the forward type PET based substrates seem more appropriate for that role based on the PCE alone. However if durability and proper lexibility is desired, strategies adopting the inverted type structure have already set foot on novel and durable lexible electrodes while so far the forward type has not gained significant progress outside ITO based electrodes. The compatibility of the forward type structure with the more lexible electrodes is still not extensively studied, but right now it seems clear that the inverted type has the greater advantage in this ield. In addition to that, other very serious concerns still need to be addressed. These concerns are mainly related to the stability of these inorganic-organic hybrid perovskite solar cells, making stability alone one of the largest roadblocks to the commercialization of any kind of existing perovskite solar cell technology. Recently, more and more basic chemical pathways of perovskite growth [41] and degradation have been established [42], but solutions have not. Right alongside this major issue, the dificulty of producing good perovskite solar cells with large areas is still challenging today (Fig. 11). As for safety, although elemental lead and most of its compounds are toxic, life cycle and environment impacts of lead shows a better behavior than tin atom. Thus, if we can recover waste lead from perovskite solar cell, it will not become such a serious problem [43,44]. With the progress of the current research, it is safe to believe that all these pro- Table 1 Summary of the lexible devices mentioned in this review, forward type and inverted type Jsc (mA cm−2) Device structure Forward Inverted Voc (V) FF PCE (%) Ref. PET/ITO/ZnO layer/ZnO nanorods/perovskite/spiro-MeOTAD/Au contact 7.52 0.8 0.43 2.62 [11] PEN/ITO/TiOx/perovskite/spiro-MeOTAD/Ag 21.4 0.95 0.6 12.2 [12] PET/ITO/TiO2/perovskite/spiro-MeOTAD/Au 20.9 1.03 0.7 15.1 [21] PET/ITO/TiO2/perovskite/PTAA/Au 21.3 0.91 0.69 13.5 [22] PEN/ITO/Zn2SnO4/perovskite/PTAA/Au 21.6 1.05 0.67 15.3 [24] Ti/TiO2 blocking layer/TiO2/perovskite/spiro-MeOTAD/thin Ag 9.5 0.89 0.73 6.15 [25] Ti/TiO2/Al2O3/perovskite/spiro-MeOTAD /PEDOT:PSS/conductive adhesive/PET+Ni 17 0.98 0.61 10.3 [26] stainless steel/compact TiO2/mesoporous TiO2/perovskite/spiro-MeOTAD/CNT sheet 10.2 0.66 0.49 3.3 [27] titanium wire/TiO2 nanotube array/perovskite/CNT sheet/Ag 14.5 0.85 0.56 7.1 [28] PDMS/willow glass/ITO/ZnO/perovskite/spiro-MeOTAD/Au 19.3 0.98 0.69 13.4 [29] PET/Al doped ZnO/Ag/Al doped ZnO/PEDOT:PSS /PolyTPD/perovskite/PCBM/Au 14.3 1.04 0.47 7.0 [30] PET/ITO/PEDOT:PSS/perovskite/PCBM/Al 16.5 0.86 0.64 9.2 [32] 15 0.8 0.6 7.6 [33] 17.19 0.99 0.72 12.25 [34] PET/conductive PEDOT/PEDOT:PSS/perovskite/PCBM/Al PET/ITO/PEDOT:PSS/perovskite/PCBM/Ca+Al PET/Ag mesh/PH1000/PEDOT:PSS/perovskite/PCBM/Al 19.5 0.91 0.8 14.2 [35] PET/PEDOT:PSS/perovskite/PCBM/PTCDI/CrOx/Au/polyurethane 17.5 0.93 0.76 12.0 [36] NOA63/PEDOT:PSS/perovskite/PCBM/EGaIn 16.56 0.939 0.7 10.9 [37] PEN/ITO/NiOx/perovskite/PCBM/Ag 18.74 1.04 0.69 13.43 [38] © Science China Press and Springer-Verlag Berlin Heidelberg 2016 503 REVIEWS SCIENCE CHINA Materials Received 8 April 2016; accepted 25 May 2016; published online 24 June 2016 1 2 3 4 5 6 Figure 11. One of the proposed perovskite decomposition pathways. Other than water, other factors are also speculated to decompose these solar cells such as silver/gold electrodes, UV light or high temperature. Reprinted with permission from Ref. [42], Copyright 2014, American Chemical Society. 7 8 9 blems will eventually clear away some day, but for now it is with little doubt true to say that the success of perovskite solar cells ultimately hinges on the question of whether organic-inorganic hybrid perovskite can be stabilized or not. For stability issues of FPSCs, we consider three factors to deine stability: bending durability (mechanical stability), long-time storage ability (chemical stability) and long-time power output ability (electro/photochemical stability). Current research points out that perfect perovskite crystal grains [15], stable organic ligands in perovskite crystal [2] and stable interface layers [36] are desired or even necessary for FPSCs. Based on irst aim, using lexible and non-brittle electrodes such as Ag mesh might just solve the bending stability of FPSCs [35]. Secondly, it is known that moisture permeates PEN or PET easily relative to glass, which makes long-time storage and long-time power output of FPSCs seriously dificult to achieve. Hence encapsulation is necessary for all reported FPSCs if they were to last long [29] and such methods might be learned from current experiences from lexible organic photovoltaic devices [45]. Finally we mention that although the long-time power output is an important parameter for FPSCs’ application, the reported papers do not explicitly mention such problem since the same problem exists for perovskite solar cells on rigid substrate. Ways to get a stable perovskite crystals and understanding entropic stabilization will become more and more important for the future of this ield [46]. 504 10 11 12 13 14 15 16 17 18 19 20 21 Yang WS, Noh JH, Jeon NJ, et al. High-performance photovoltaic perovskite layers fabricated through intramolecular exchange. Science, 2015, 348: 1234–1237 Saliba M, Matsui T, Seo JY, et al. Cesium-containing triple cation perovskite solar cells: improved stability, reproducibility and high eficiency. Energy Environ Sci, 2016, 9: 1989–1997 Hodes G. Perovskite-based solar cells. Science, 2013, 342: 317–318 Kojima A, Teshima K, Shirai Y, et al. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J Am Chem Soc, 2009, 131: 6050–6051 Xing G, Mathews N, Sun S, et al. Long-range balanced electronand hole-transport lengths in organic-inorganic CH3NH3PbI3. Science, 2013, 342: 344–347 Sun S, Salim T, Mathews N, et al. The origin of high eficiency in low-temperature solution-processable bilayer organometal halide hybrid solar cells. Energy Environ Sci, 2014, 7: 399–407 D’innocenzo V, Grancini G, Alcocer MJP, et al. Excitons versus free charges in organo-lead tri-halide perovskites. Nat Commun, 2014, 5: 3586 Dong Q, Fang Y, Shao Y, et al. Electron-hole diffusion lengths > 175 mm in solution-grown CH3NH3PbI3 single crystals. Science, 2015, 347: 967–970 Wehrenfennig C, Eperon GE, Johnston MB, et al. High charge carrier mobilities and lifetimes in organolead trihalide perovskites. Adv Mater, 2014, 26: 1584–1589 Green MA, Ho-baillie A, Snaith HJ. The emergence of perovskite solar cells. Nat Photon, 2014, 8: 506–514 Kumar MH, Yantara N, Dharani S, et al. Flexible, low-temperature, solution processed ZnO-based perovskite solid state solar cells. Chem Commun, 2013, 49: 11089–11091 Kim BJ, Kim DH, Lee YY, et al. Highly eficient and bending durable perovskite solar cells: toward a wearable power source. Energy Environ Sci, 2015, 8: 916–921 Susrutha B, Giribabu L, Singh SP. Recent advances in lexible perovskite solar cells. Chem Commun, 2015, 51: 14696–14707 Wang Y, Bai S, Cheng L, et al. High-eficiency lexible solar cells based on organometal halide perovskites. Adv Mater, 2016, 28: 4532–4540 Guo Y, Sato W, Shoyama K, et al. Sulfamic acid-catalyzed lead perovskite formation for solar cell fabrication on glass or plastic substrates. J Am Chem Soc, 2016, 138: 5410–5416 Liu D, Kelly TL. Perovskite solar cells with a planar heterojunction structure prepared using room-temperature solution processing techniques. Nat Photon, 2013, 8: 133–138 Burschka J, Pellet N, Moon SJ, et al. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. Nature, 2013, 499: 316–319 Wang JTW, Ball JM, Barea EM, et al. Low-temperature processed electron collection layers of graphene/TiO2 nanocomposites in thin ilm perovskite solar cells. Nano Lett, 2014, 14: 724–730 Kim HS, Lee JW, Yantara N, et al. High eficiency solid-state sensitized solar cell-based on submicrometer rutile TiO2 nanorod and CH3NH3PbI3 perovskite sensitizer. Nano Lett, 2013, 13: 2412–2417 Lee JW, Lee TY, Yoo PJ, et al. Rutile TiO2-based perovskite solar cells. J Mater Chem A, 2014, 2: 9251–9259 Tan ZK, Moghaddam RS, Lai ML, et al. Bright light-emitting diodes based on organometal halide perovskite. Nat Nanotech, 2014, 9: 687–692 © Science China Press and Springer-Verlag Berlin Heidelberg 2016 REVIEWS SCIENCE CHINA Materials 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Qiu W, Paetzold UW, Gehlhaar R, et al. An electron beam evaporated TiO2 layer for high eficiency planar perovskite solar cells on lexible polyethylene terephthalate substrates. J Mater Chem A, 2015, 3: 22824–22829 Li Y, Meng L, Yang YM, et al. High-eficiency robust perovskite solar cells on ultrathin lexible substrates. Nat Commun, 2016, 7: 10214 Shin SS, Yang WS, Noh JH, et al. High-performance lexible perovskite solar cells exploiting Zn2SnO4 prepared in solution below 100 °C. Nat Commun, 2015, 6: 7410 Lee M, Jo Y, Kim DS, et al. Flexible organo-metal halide perovskite solar cells on a Ti metal substrate. J Mater Chem A, 2015, 3: 4129–4133 Troughton J, Bryant D, Wojciechowski K, et al. Highly eficient, lexible, indium-free perovskite solar cells employing metallic substrates. J Mater Chem A, 2015, 3: 9141–9145 Qiu L, Deng J, Lu X, et al. Integrating perovskite solar cells into a lexible iber. Angew Chem Int Ed, 2014, 53: 10425–10428 Qiu L, He S, Yang J, et al. Fiber-shaped perovskite solar cells with high power conversion eficiency. Small, 2016, 12: 2419–2424 Tavakoli MM, Tsui KH, Zhang Q, et al. Highly eficient lexible perovskite solar cells with antirelection and self-cleaning nanostructures. ACS Nano, 2015, 9: 10287–10295 Roldán-carmona C, Malinkiewicz O, Soriano A, et al. Flexible high eficiency perovskite solar cells. Energy Environ Sci, 2014, 7: 994–997 Malinkiewicz O, Yella A, Lee YH, et al. Perovskite solar cells employing organic charge-transport layers. Nat Photon, 2013, 8: 128–132 You J, Hong Z, Yang YM, et al. Low-temperature solution-processed perovskite solar cells with high eficiency and lexibility. ACS Nano, 2014, 8: 1674–1680 Poorkazem K, Liu D, Kelly TL. Fatigue resistance of a lexible, eficient, and metal oxide-free perovskite solar cell. J Mater Chem A, 2015, 3: 9241–9248 Chen Y, Chen T, Dai L. Layer-by-layer growth of CH3NH3PbI3−xClx for highly eficient planar heterojunction perovskite solar cells. Adv Mater, 2015, 27: 1053–1059 Li Y, Meng L, Yang YM, et al. High-eficiency robust perovskite solar cells on ultrathin lexible substrates. Nat Commun, 2016, 7: 10214 Kaltenbrunner M, Adam G, Głowacki ED, et al. Flexible high power-per-weight perovskite solar cells with chromium oxide–metal contacts for improved stability in air. Nat Mater, 2015, 37 38 39 40 41 42 43 44 45 46 14: 1032–1039 Park M, Kim HJ, Jeong I, et al. Mechanically recoverable and highly eficient perovskite solar cells: investigation of intrinsic lexibility of organic-inorganic perovskite. Adv Energy Mater, 2015, 5: 1501406 Yin X, Chen P, Que M, et al. Highly eficient lexible perovskite solar cells using solution-derived NiOx hole contacts. ACS Nano, 2016, 10: 3630–3636 Lee KT, Fukuda M, Joglekar S, et al. Colored, see-through perovskite solar cells employing an optical cavity. J Mater Chem C, 2015, 3: 5377–5382 Guo Y, Shoyama K, Sato W, et al. Polymer stabilization of lead(II) perovskite cubic nanocrystals for semitransparent solar cells. Adv Energy Mater, 2016, 6: 1502317 Guo Y, Shoyama K, Sato W, et al. Chemical pathways connecting lead(II) iodide and perovskite via polymeric plumbate(II) iber. J Am Chem Soc, 2015, 137: 15907–15914 Frost JM, Butler KT, Brivio F, et al. Atomistic origins of high-performance in hybrid halide perovskite solar cells. Nano Lett, 2014, 14: 2584–2590 Serrano-lujan L, Espinosa N, Larsen-olsen TT, et al. Tin- and leadbased perovskite solar cells under scrutiny: an environmental perspective. Adv Energy Mater, 2015, 5: 1501119 Gong J, Darling SB, You F. Perovskite photovoltaics: life-cycle assessment of energy and environmental impacts. Energy Environ Sci, 2015, 8: 1953–1968 http://www.m-kagaku.co.jp/english/aboutmcc/RC/special/feature1.html (accessed: May, 2014) Yi C, Luo J, Meloni S, et al. Entropic stabilization of mixed A-cation ABX3 metal halide perovskites for high performance perovskite solar cells. Energy Environ Sci, 2016, 9: 656–662 Acknowledgments This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT, Japan) through the Strategic Promotion of Innovative Research. H.H is supported by the Graduate School of Science Fellowship, The University of Tokyo. We thank Prof. Eiichi Nakamura for discussion. Author contributions Halim H searched the literature. Halim H and Guo Y wrote the manuscript together. Conlict of interest interest. The authors declare that they have no conlict of © Science China Press and Springer-Verlag Berlin Heidelberg 2016 505 REVIEWS SCIENCE CHINA Materials Henry Halim received his BSc degree in chemistry with irst class honor from the University of Hong Kong in 2015. Currently he is a master course student of the University of Tokyo in Prof. Eiichi Nakamura’s group. His scientiic interest lies in solar energy and he is currently undertaking research in the ield of perovskite solar cells. Yunlong Guo received his BSc degree in chemistry from Hebei Normal University (2005) and PhD degree in physical chemistry from the Institute of Chemistry, Chinese Academy of Sciences (ICCAS) in 2010. From 2016, he is a project associate professor at the Department of Chemistry, University of Tokyo. His research interest includes fabrication, characterization, and optimization of organic-inorganic hybrid perovskite solar cells and functional organic ield-effect transistors. 柔性有机－无机杂化钙钛矿太阳能电池研究 Henry Halim, 郭云龙 * 摘要 在短短的7年中, 有机-无机杂化钙钛矿太阳能电池已经经历了深入的研究和巨大的变迁. 虽然在硬质基板上溶液加工的钙钛矿太阳 能电池的光电转化超过了20%, 但是科学家们并没有因此而忽视柔性器件领域的研究, 并且在高效率和耐用柔性器件方面取得了巨大进步. 本综述对近3年来柔性有机-无机钙钛矿太阳能电池的发展状况和如何制备柔性电池的关键技术与策略做了详尽的介绍和总结. 最后就柔 性有机-无机杂化钙钛矿太阳能电池研究和未来应用中存在的问题及如何解决这些问题做了展望. 506 © Science China Press and Springer-Verlag Berlin Heidelberg 2016