Thermal Stability-Enhanced and High-Efficiency Planar ...wdxy.hubu.edu.cn/__local/B/56/2C/462C847E781FB2C738C37C5D53… · of perovskite solar cells based on ZnO and PCBM-passivated

Thermal Stability-Enhanced and High-Efficiency Planar PerovskiteSolar Cells with Interface PassivationWeihai Zhang,†,‡,§ Juan Xiong,*,‡ Li Jiang,† Jianying Wang,† Tao Mei,† Xianbao Wang,†

Haoshuang Gu,‡ Walid A. Daoud,§ and Jinhua Li*,†,‡

†Hubei Collaborative Innovation Center for Advanced Organic Chemical Materials, Key Laboratory for the Green Preparation andApplication of Functional Materials, Ministry of Education, Hubei Key Laboratory of Polymer Materials, School of Materials Scienceand Engineering and ‡Hubei Key Laboratory of Ferro & Piezoelectric Materials and Devices, Faculty of Physics & ElectronicSciences, Hubei University, Wuhan 430062, China§School of Energy and Environment, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong

*S Supporting Information

ABSTRACT: As the electron transport layer (ETL) ofperovskite solar cells, oxide semiconductor zinc oxide (ZnO)has been attracting great attention due to its relatively highmobility, optical transparency, low-temperature fabrication,and good environment stability. However, the nature of ZnOwill react with the patron on methylamine, which woulddeteriorate the performance of cells. Although many methods,including high-temperature annealing, doping, and surface modification, have been studied to improve the efficiency and stabilityof perovskite solar cells with ZnO ETL, devices remain relatively low in efficiency and stability. Herein, we adopted a novelmultistep annealing method to deposit a porous PbI2 film and improved the quality and uniformity of perovskite films. The cellswith ZnO ETL were fabricated at the temperature of <150 °C by solution processing. The power conversion efficiency (PCE) ofthe device fabricated by the novel annealing method increased from 15.5 to 17.5%. To enhance the thermal stability ofCH3NH3PbI3 (MAPbI3) on the ZnO surface, a thin layer of small molecule [6,6]-phenyl-C61-butyric acid methyl ester (PCBM)was inserted between the ZnO layer and perovskite film. Interestingly, the PCE of PCBM-passivated cells could reach nearly19.1%. To our best knowledge, this is the highest PCE value of ZnO-based perovskite solar cells until now. More importantly,PCBM modification could effectively suppress the decomposition of MAPbI3 and improve the thermal stability of cells.Therefore, the ZnO is a promising candidate of electron transport material for perovskite solar cells in future applications.

KEYWORDS: perovskite solar cells, crystallinity, zinc oxide, electron transport material, interface passivation

1. INTRODUCTION

Organometal halide perovskite solar cells (PSCs) have attractedmuch attention in recent years due to their high photovoltaicefficiency and low fabrication cost.1−6 Their power conversionefficiency (PCE) jumped from 3.8% in 20097 to a certified22.1% in 20168 via film crystal growth control, band gapadjustment, and interface engineering.9−12 In typical meso-porous PSCs, compact TiO2 as well as porous TiO2 films werecommonly used as the electron transport layer (ETL).13−15

However, a high sintering temperature (>450 °C) was requiredto form the highly crystallized TiO2 films, which made itimpossible to fabricate low-cost and flexible solar cells.To fabricate low-temperature and high-efficiency PSCs, a

mesoporous-free “planar” configuration has been developedrecently.16−20 The planar PSCs are often fabricated by using theelectron transport materials, such as SnO2, TiO2, andZnO,21−26 which can be deposited by low-temperature solutionprocessing. These materials have a relatively high carriermobility and superior stability, which made them idealcandidates as transport layers. The planar perovskite solarcells based on a SnO2 ETL have been reported to have an

efficiency over 20%.27,28 When a modified TiO2 film wasadopted as the ETL, the efficiency of cells could reach19.62%.23 Compared to TiO2 and SnO2, the ZnO has thehighest carrier mobility. However, the highest efficiency of thecell based on the ZnO ETL was only 17.7%, reported by Im etal.29 The poor performances could be mainly ascribed to severalfactors. First, the hydroxyl groups, organic residuals, and oxygenvacancies on the surface of the ZnO layer can easily lead tothermal decomposition.30 Second, without a porous structure, itis always difficult to acquire a pure CH3NH3PbI3 film withoutsome PbI2 residuals, which plays a negative effect on thetransportation of the charges.31,32 Third, ZnO itself causes thedeprotonation of the methylammonium cation, which results inthe loss of methylamine and the formation of PbI2.

33,34

Therefore, it is difficult to obtain high-efficiency PSCs by usingthe ZnO ETL. Several strategies were addressed to resolve theabove-mentioned problems, including high-temperature anneal-

Received: July 25, 2017Accepted: October 13, 2017Published: October 13, 2017

Research Article

www.acsami.org

© 2017 American Chemical Society 38467 DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

Cite This: ACS Appl. Mater. Interfaces 2017, 9, 38467-38476

ing to remove the residuals on the ZnO surface,34 adoptingaluminum-doped ZnO (AZO) as the ETL to reduce the badeffect,35 and depositing an intermediate layer to prevent thedirect interaction between ZnO and the perovskite layer.36,37

These methods are effective to a certain extent and indeed havea positive effect on enhancing the performance of the device.However, the efficiency of perovskite solar cells with the ZnOETL remains inferior compared to the efficiency of those withthe TiO2 and SnO2 ETLs.[6,6]-Phenyl-C61-butyric acid methyl ester (PCBM) is an

important organic electron transport material in PSCs.18−20

The lowest unoccupied molecular orbital (LUMO) energy levelof PCBM is 3.9 eV that can match the conductive band level ofMAPbI3 well.38,39 Electron transport can be very efficientbetween the interface of MAPbI3 and PCBM. Nie et al.40

reported MAPbI3/PCBM inverted planar hybrid PSCs, and thehigh PCE of 17.7% was achieved. Heo et al.19 also reported thatthe PCE of the same-structured PSCs could reach 18.1%. Thehigh efficiency was due to the high conductivity of PCBMcompared to that of TiO2. Very recently, Qiu et al.41 usedcross-linked PCBM as the interlayer between perovskite andTiO2 to fabricate planar PSCs with a high efficiency of 18.4%.This suggests that it is possible to obtain the high-performancePSCs when PCBM-passivated ZnO is used as the ETL. Herein,we adopt a novel multistep annealing (SA) method to prepare aporous PbI2 film for the formation of a pure CH3NH3PbI3 film.The results reveal that the SA method is helpful to fabricate

high-quality CH3NH3PbI3 film. Compared with the commonannealing (CA) method, the PCE of the cells fabricated by theSA method increased from 15.5 to 17.5%. Furthermore, a seriesof perovskite solar cells based on ZnO and PCBM-passivatedZnO ETL were fabricated to investigate the thermal stability ofthe devices. When a buffer layer of PCBM is inserted betweenthe MAPbI3 film and ZnO film, the degradation of theCH3NH3PbI3 film is effectively suppressed. More interestingly,an enhancement of PCE is observed. The efficiency of thedevice increased from 17.5 (without PCBM) to 19.07% bychoosing a suitable thick PCBM. To the best of our knowledge,this is the highest record ever reported for perovskite solar cellsbased on the ZnO ETL. This implies that ZnO will be a greatlypromising ETL material for future applications of PSCs.

2. EXPERIMENTAL SECTION2.1. Preparation of ZnO Nanoparticles. The ZnO nanoparticles

(NPs) were synthesized according to our former work.42 Zinc acetatedihydrate (ZnAc·2H2O) (0.59 g) and potassium hydroxide (KOH)(0.296 g) were dissolved in 25 and 13 mL of methanol, respectively.Before the KOH solution was added, the zinc acetate dihydratesolution was processed with a constant 65 °C water bath for 10 min.Then, the mixed solution was stirred for 2.5 h at 65 °C. Thereafter, thesolution was cooled down and stored overnight to precipitate theproduct. Finally, the transparent ZnO precursor was produced bydispersing the precipitate in the n-butanol (14 mL), methanol (1 mL),and chloroform (2 mL), with a concentration of 6 mg mL−1. Before

Figure 1. Properties of ZnO NPs. (a) Transmission electron microscopy (TEM) image of ZnO nanoparticles. (b) High-resolution TEM image ofZnO nanoparticles. (c) Electron diffraction pattern of ZnO nanoparticles. (d) X-ray diffraction pattern of ZnO nanoparticles. (e) Scanning electronmicroscope (SEM) image of ZnO nanoparticles deposited on a glass/FTO substrate. (f) Optical properties of ZnO films with and without PCBM.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38468

the usage, the ZnO precursor was filtered through a 0.45 μmpoly(vinylidene fluoride) (PVDF) syringe filter.2.2. Perovskite Solar Cell Fabrication. The fluorine-doped tin

oxide (FTO) substrate was thoroughly cleaned by acetone, ethanol,and deionized water, respectively. Then, it was treated by an O2

plasma for 3 min to increase the hydrophilic nature of the surface. Forthe ZnO layer, the prepared ZnO precursor was spin-coated on well-cleaned FTO substrates at 3000 rpm for 30 s and then annealed at 150°C for 10 min to remove the organic solvent in ambient air. Thisprocedure was repeated several times to obtain a continuous smoothZnO film with an appropriate thickness. The PCBM film wasdeposited by spin coating PCBM chlorobenzene solution on the ZnOlayer with a rate of 3000 rpm for 40 s and annealed on a hotplate at 85°C for 5 min in a glovebox. The perovskite layer was deposited by atwo-step dipping method. First, the PbI2 layer was deposited by spincoating 1.1 M PbI2, which dissolved in dimethylformamide/dimethylsulfoxide (DMF/DMSO) (90/10 v/v) at 3000 rpm for 30 s. The as-prepared PbI2 film was annealed by a multistep annealing method (60,80, and 100 °C for 5 min, respectively) to obtain the uniform andporous PbI2 film. Then, the porous PbI2 film was dipped into asolution of CH3NH3I in 2-propanol, with a concentration of 10 mg/mL for 5 min. The initial perovskite film was dried under a flow ofclean air at 100 °C for 5 min to form a well-crystallized CH3NH3PbI3layer. Subsequently, the spiro-OMeTAD hole transport layer (72.3 mgof spiro-OMeTAD, 28.8 μL of 4-tert-butyl pyridine, and 17.5 μL oflithium-bis (trifluoromethanesulfonyl) imide (Li-TFSI) solution (520mg Li-TFSI in 1 mL of acetonitrile) all dissolved in 1 mL ofchlorobenzene) was prepared by spin coating at 4000 rpm for 30 s.Finally, a 80 nm thick Au layer was deposited by thermal evaporation.The area of the Au electrode was 6 mm2.

2.3. Device Characterization. A transmission electron micro-scope (TEM) (Tecnai G20, FEI) was used to analyze the ZnOnanoparticles. The optical transmittance and absorbance spectra wereobtained by an ultraviolet−visible diffuse reflection spectroscopy(UV−vis−DRS, Shimadzu UV-3600). The roughness of the ZnO layerwith and without PCBM was analyzed by an atomic force microscope(AFM) (Ntegra upright, NT-MDT, Russia). A field-emission scanningelectron microscope (FESEM, JEOL 7100 F) was used to acquiresurface and cross-sectional morphologies. The X-ray diffraction (XRD)patterns (2θ scans) were obtained by Bruker Advanced D8 X-raydiffractometer using Cu Kα (λ = 0.154 nm) radiation. Currentdensity−voltage (J−V) curves of the perovskite solar cells weremeasured by Keithley 237 under AM 1.5 G one-sun illuminationprovided by a solar simulator (Newport Oriel Sol 3 A Class AAA,64023 A). The external quantum efficiency (EQE) spectrum wasmeasured under a constant white light supplied by an array of whitelight-emitting diodes. The excitation beam coming from a 300 Wxenon lamp was focused through a double monochromator andchopped at 10 Hz. The monochromatic light intensity for themeasurement was calibrated by a reference silicon photodiode. Time-resolved photoluminescence (TRPL) spectra were measured by a PLsystem (DW-PLE03). The electrochemical impedance spectroscopy(EIS) measurements were carried out under 1 sun light illumination byusing a computer-controlled electrochemical workstation (ZenniumZahner, Germany). A bias of the open-circuit voltage and a smallvoltage perturbation of 10 mV were applied during the measurementat frequencies ranging from 1 MHz to 1 Hz.

3. RESULTS AND DISCUSSION3.1. Characterization of ZnO Nanoparticles and Thin

Films. The prepared ZnO NPs are very uniform and show an

Figure 2. Schematic illustration of the (a) CA method and (b) SA method. (c) XRD patterns of the CH3NH3PbI3 film derived from differentannealing methods. (d) J−V curves of the devices fabricated by different annealing methods, the inset table lists the detail parameters of the cells.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38469

approximate diameter of 10 nm revealed by the transmissionelectron microscope (TEM), as shown in Figure 1a. TheHRTEM image in Figure 1b indicates that the ZnO NPs arehighly crystallized, which is extremely important to reduce thetraps in the ZnO ETL and improve the ability of chargetransfer. The particles exhibit a hexagonal wurtzite structureand a polycrystalline structure as confirmed by the electrondiffraction pattern in Figure 1c and X-ray diffraction (XRD)pattern in Figure 1d. A dense and pinhole-free ZnO film isobserved by a field-emission scanning electron microscopy(FESEM), as shown in Figure 1e. Atomic force microscopeimages shown in Supporting Information reveal a smoothsurface of the ZnO film with a root mean square (RMS) at 2.31nm (Figure S1a). Moreover, a smoother surface with RMS =1.35 nm can be observed when a thin layer of PCBM is coatedonto the ZnO layer, as shown in Figure S1c. The opticaltransmittance and absorbance of the ZnO layer with andwithout PCBM are shown in Figure 1f. The transmittance ofthe layer with and without PCBM are both greater than 90% inthe visible region. The absorbance spectra reveal that theabsorption of the PCBM/ZnO bilayer film just decreasesslightly compared to that of the pure ZnO film.3.2. Multistep Annealing Method. To fabricate a high-

quality CH3NH3PbI3 film without any PbI2 residual, we adopt anovel multistep annealing (SA) method to prepare a porousPbI2 film, which is helpful in the formation of pureCH3NH3PbI3. Figure 2a,b illustrates the schematic of thecommon annealing (CA) and SA procedure of the PbI2 films,

respectively. As shown in Figure 2a, after PbI2 is deposited, thefilm is directly annealed at a constant temperature of 100 °C for15 min before it is dipped into CH3NH3I solution. For the SAmethod, PbI2 film is annealed successively at 60, 80, and 100 °Cfor 5 min respectively, as shown in Figure 2b. The XRDpatterns of the perovskite films derived from different PbI2annealing methods are shown in Figure 2c. Clearly, there issome PbI2 residual in the perovskite film when the CA methodis used, although the SA method tends to obtain the pureperovskite film. It suggests that the formation of pureperovskite film can be mainly attributed to the porous PbI2film (shown in Figure 3b), which is beneficial to the easypermeation of the CH3NH3I solution into the PbI2 film and infacilitation of the formation of CH3NH3PbI3. The absorbancespectra of the CH3NH3PbI3 film fabricated by differentannealing methods are shown in the Supporting Information,Figure S2. It is clear that the film fabricated by the CA methodhas a lower absorbance than that of the film fabricated by theSA method, which would be mainly due to the PbI2 residual.Figure 2d shows the current density−voltage (J−V) curves ofthe solar cells fabricated by different annealing methods. Thedetail parameters are shown in the inset table. Obviously, thedevice fabricated by the CA method shows a relatively low PCEof 15.5%. The poor performances are primarily due to thelower current density (Jsc) and fill factor (FF), which could becaused by the excess PbI2.

31,32 However, PCE of cells increasesto 17.5% when the SA method is used.

Figure 3. Top viewed SEM images of PbI2 films prepared by the (a) CA method and (b) SA method, and the corresponding SEM images ofCH3NH3PbI3 film prepared by the (c) CA method, (d) SA method.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38470

The morphology control of the PbI2 film is the key factor forforming the high-quality CH3NH3PbI3 film. As shown in Figure3a, the PbI2 film prepared by the CA method indicates anextremely dense and pinhole-free morphology, whereas a PbI2film prepared by the SA method indicates a uniform and porousmorphology, as shown in Figure 3b. The remarkable differenceof the morphology would be mainly due to the different growthrates of the PbI2 crystal. When the SA method was adopted, itseemed impossible to vaporize away all of the DMSO duringthe first half of the annealing process at the relatively lowtemperature, which is lower than the boiling point of DMSO. Ithas been reported that DMSO shows stronger capability thanDMF to coordinate with PbI2 to form a complex PbI2−DMSOintermediate, which would retard the crystallization of the PbI2film and form a porous structure.43,44 Besides, the intermediatephase is proven to be helpful to form a better quality perovskitefilm, as reported by Han et al.45 As a consequence, thecorresponding CH3NH3PbI3 film shows apparently differentmorphologies as well. Figure 3c shows an uneven grain size anda rough CH3NH3PbI3 film derived from the CA method.However, a very uniform and relatively smooth perovskite filmprepared by the SA method is observed, as shown in Figure 3d.Therefore, the SA method is effective to control themorphology of the PbI2 film and further to control theformation of a better quality CH3NH3PbI3 film.3.3. Stability Enhancement. Although a relatively high

efficiency of 17.5% is achieved by the SA method, we find outthat the perovskite film is easily decomposed when the film isannealed at 100 °C. The main reason may be the deprotonationof the methylammonium cation caused by ZnO, which results

in the loss of methylamine and the formation of PbI2, asreported by Kelly et al.33,34

To enhance the thermal stability of the device, we fabricatedthe cells by the above-mentioned SA method, with a structureof FTO/ZnO/PCBM/CH3NH3PbI3/Spiro-OMeTAD/Au. Athin layer of PCBM was inserted as a buffer layer to avoidthe direct contact between ZnO and the perovskite film. Toexamine the effect of PCBM passivation, we systemicallyinvestigated the thermal stability of perovskite films depositedon FTO/ZnO, FTO/ZnO/PCBM, and the FTO/PCBM layer,respectively. Figure 4a shows the photographs of the perovskitelayers deposited on different substrates with different annealingtimes varying from 5 to 40 min. Apparently, the perovskite filmfabricated on the FTO/ZnO layer is easily decomposed anddemonstrates yellow color after annealing for 20 min at 100 °Cin air. In contrast, the color of the perovskite layer deposited onFTO/ZnO/PCBM and the FTO/PCBM layer remains darkafter 40 min. The XRD patterns in Figure 4b−d show theevolution of crystalline phases of the CH3NH3PbI3 filmsfabricated on different substrates under the different annealingtimes. As shown in Figure 4b (FTO/ZnO/perovskite), a PbI2peak at 12.6° is observed when the film is annealed for 10 minand the peak is stronger with a longer annealing time. After 40min of annealing, it is hard to detect the peak of theCH3NH3PbI3 film, whereas the perovskite film deposited onthe FTO/ZnO/PCBM layer remains pure phase until theannealing time is longer than 15 min, as shown in Figure 4c.Besides, the peaks of the CH3NH3PbI3 film are still dominantafter annealing for 40 min. For the perovskite film deposited onthe FTO/PCBM layer, it seems that it has an excellent thermal

Figure 4. (a) Photographs of perovskite layers deposited on different substrates annealed for different times at 100 °C in air. XRD patterns of theCH3NH3PbI3 films deposited on the (b) ZnO layer, (c) ZnO/PCBM layer, and (d) PCBM layer.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38471

stability, as shown in Figure 4d. The appearance of the PbI2peak would be mainly attributed to the effect of water for thefact that the annealing experiment is done in an ambientatmosphere. The different intensity of the PbI2 peak betweenFigure 4c and Figure 4d is ascribed to the fact that the PCBMlayer is too thin to cover the whole area of the ZnO layer, asshown in the Supporting Information, Figure S1d, and theexposed ZnO NPs would deteriorate the quality of theCH3NH3PbI3 film when an annealing process is conducted.The results suggest that CH3NH3PbI3 is indeed sensitive to theZnO layer and a buffer layer of PCBM has an undeniablepositive effect on the enhancement of the thermal stability ofthe perovskite film.In addition to thermal stability, long-term stability is also very

important for the application of the perovskite solar cells.Therefore, to ensure whether PCBM passivation has a positiveor negative effect on the long-term stability of the device, weconducted a stability test. Figure 5a shows the results of thestability test of the devices with and without PCBM passivationin an ambient environment without encapsulation (40−50%humidity, T = 23 °C). Clearly, the PCEs of the devices withoutas well as with PCBM passivation decreased with the increaseof the exposure time. The instability of the devices would bemainly attributed to the intrinsic properties of the perovskitematerial (sensitive to water), the nonpacked devices, and thepinhole perovskite film that allows water to penetrate into thefilm easily, as shown in Figure 3. In spite of the degradation, wecan notice that the device with PCBM passivation (maintainsover 60% of the initial efficiency after 10 days) shows a higherstability than that without PCBM (only about 40% of the initialefficiency after 10 days), as shown in Figure 5b. The betterstability performance of the PCBM-passivated device would be

mainly ascribed to two factors: (1) The inserting layer ofPCBM between ZnO and CH3NH3PbI3 film is helpful tosuppress the penetration of water because of the fact thatPCBM shows a poor hydrophilicity. (2) The CH3NH3PbI3 filmdeposited on the PCBM-passivated ZnO layer shows a densermorphology with fewer pinholes than that without PCBMpassivation, as shown in the Supporting Information, Figure S3.Better morphology and fewer pinholes and voids allow alimited path for water penetration, which is essential for long-term stability, as is reported by Wu et al.51 Figure 5c,d showsthe J−V curves of the devices without and with PCBMpassivation varying with the exposure time, and thecorresponding detailed parameters are summarized in theSupporting Information, Table S1. It is easily concluded thatthe degradation of the device performance is determined by theoverall deterioration of the Voc, Jsc, and FF. Nevertheless, it canbe noted that a PCBM-passivated device shows a betterstability, indicating that PCBM passivation plays a positiveeffect on the long-term stability of the device.

3.4. Photovoltaic Performance. Figure 6a shows thecross-sectional SEM image of the completed device with a layerof PCBM. It can be seen that the perovskite layer has anapproximate thickness of 500 nm. The top viewed SEM imagesof CH3NH3PbI3 films grown on the ZnO layer with andwithout PCBM passivation are shown in the SupportingInformation, Figure S3. Clearly, the film deposited on thePCBM-passivated ZnO layer shows a denser and more uniformmorphology, which is helpful to enhance the performances ofthe device, as reported by Petrozza et al.46,47 The energydiagram of the device with a layer of PCBM is shown in Figure6b. It is clear that the LUMO energy level of PCBM (−3.9 eV)is consistent with the band of the CH3NH3PbI3 film, which

Figure 5. (a) PCEs of the devices with different exposure times in an ambient environment (40−50% humidity, T = 23 °C). (b) Long-term stabilitytest of the different types of devices. Corresponding J−V curves of the devices without (c) and with (d) PCBM passivation varying with the exposuretime.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38472

indicates that there is no barrier between the interface forextracting electrons, whereas the holes can be blocked.38,39

Therefore, the recombination of the carriers can be reduced atthis interface, which is well demonstrated by the results of time-resolved photoluminescence (TRPL) measurement and elec-trical impedance spectroscopy (EIS) characterization. TheTRPL spectra of the FTO/CH3NH3PbI3, FTO/ZnO/CH3NH3PbI3, FTO/ZnO/PCBM/CH3NH3PbI3, and FTO/ZnO/PCBM/CH3NH3PbI3/Spiro in Figure 6c indicate that thelifetime of the carriers can be significantly reduced when a layerof PCBM is applied. It implies that an efficient electron transferoccurred from CH3NH3PbI3 to PCBM.47−49 To further studythe mechanism of the charge transport process, we carried outEIS characterization. The Nyquist plots of the devices with andwithout PCBM as measured at open-circuit voltage under 1 sunAM 1.5G illumination are shown in Figure S4, SupportingInformation. The inset image reveals the correspondingequivalent circuit. It is clear that the data in the plot areseparated into two semicircles. The first semicircle refers to theR1 for the charge-transfer resistance, and the second semicirclerefers to R2 for the charge-recombination resistance.23,33,50 Thecorresponding fitting values of the Rs, R1, and R2 aresummarized in Table S2. As expected, the PCBM-passivateddevice shows a reduced value of R1 (365.1 Ω), much lower thanthe resistance (470.3 Ω) obtained using the ZnO layer withoutPCBM passivation, in keeping with the energy level alignmentbetween PCBM and CH3NH3PbI3 film. A lower value of R1allows a more efficient electron extraction at the PCBM/CH3NH3PbI3 interface because the CH3NH3PbI3/Spiro-

OMeTAD interface is identical in both cases, which wouldcertainly enhance the current density of the device.Figure 6d shows the J−V curves of the best performing

devices with and without PCBM. The device without PCBMhas the short circuit current density (Jsc) of 21.4 mA cm−2, Voc

of 1.08 V, and FF of 75.35%, producing a PCE of 17.5%.Interestingly, when the optimal thickness PCBM is depositedon the ZnO surface, the device shows an outstandingenhancement of Jsc from 21.4 to 22.8 mA cm−2 and thechampion PCE of 19.07% is achieved (the performance of thedevice based on PCBM ETL without the ZnO layer is shown inFigure S5, Supporting Information). The detailed parameters ofthe devices for different concentrations of PCBM aresummarized in Table 1, and the corresponding J−V curvesare shown in Figure S6. The external quantum efficiency(EQE) spectra and integrated Jsc of the best performing devicesare shown in Figure 6e. It is clear that the EQE of the devicewith PCBM is higher than that of the device without PCBM.

Figure 6. (a) Cross-sectional SEM image of the perovskite solar cell. (b) Energy level diagram of each layer of the device. (c) Time-resolvedphotoluminescence (TRPL) spectra of the perovskite based on different substrates. (d) J−V curves and (e) EQE spectra as well as correspondingintegrated current density of the PSCs with and without PCBM passivation.

Table 1. Detail Parameters of the Devices with DifferentPCBM Concentrations

concentration (mg mL−1) Voc (V) Jsc (mA cm−2) FF (%) PCE (%)

0 1.08 21.4 75.35 17.511 1.10 22.1 73.35 18.02 1.08 22.8 77.3 19.073 1.07 22.3 76.3 18.24 1.09 21.86 75.8 18.03

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38473

The corresponding integrated Jsc of the devices with andwithout PCBM are in good agreement with the J−V measuredresults. It is considered that the higher Jsc of the PCBM-passivated device is mainly due to the better performingCH3NH3PbI3 film as well as the band alignment betweenPCBM and CH3NH3PbI3 film, leading to a nonbarrierenvironment for electron transportation.The histograms in Figure 7 show the distribution of Voc, Jsc,

FF, and PCE for the devices without and with PCBMpassivation. As shown in Figure 7a, the average Voc ofPCBM-passivated devices is about 1.06 V, which is similar tothat of non-PCBM-passivated devices. It implies that PCBMpassivation has little influence on the Voc of the device.Interestingly, more than 90% of devices have current densityover 20.5 mA cm−2 and average fill factor over 75% when theZnO surface is passivated by a thin PCBM layer, as shown inFigure 7b,c. Compared with devices without PCBM passiva-tion, the PCBM-passivated devices demonstrate significantenhancements of Jsc and FF. Finally, the elevated PCEs ofPCBM-passivated devices are achieved. The overall enhance-ment of the performances of devices can be ascribed to the pureand uniform perovskite film, very smooth interfaces, and bandalignment between PCBM and the perovskite layer.

4. CONCLUSIONS

In summary, we employed a novel SA method to deposit aporous PbI2 film that enabled the formation of a highlyuniform, relatively smooth, and pure perovskite film. Theefficiency of the device made by the SA method increased from15.5 to 17.5%. Furthermore, a thin layer of PCBM was insertedbetween the ZnO and perovskite film to passivate the ZnOsurface. Experimental results demonstrated that a layer ofPCBM can not only effectively suppress the thermaldecomposition of the perovskite film through insulating thedirect contact of the ZnO and CH3NH3PbI3 but also reduce theinterface barrier for the band alignment between PCBM andthe perovskite layer. Combining with the novel SA method andthe modification of PCBM, the PCE of device can reach19.07% and thermal stability of the device can be enhanced.This suggests that the ZnO is a promising candidate of electrontransport material for high-performance perovskite solar cellsfor future applications.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.7b10994.

Additional data for AFM images of the ZnO layer andZnO/PCBM layer (Figure S1), absorbance spectra of the

Figure 7. Histograms of the parameters measured for 100 separate devices with and without PCBM, respectively. (a) Voc, (b) Jsc, (c) FF, and (d)PCE.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38474

perovskite layer made by different annealing methods(Figure S2), SEM images of the CH3NH3PbI3 films(Figure S3), Nyquist plots of the devices (Figure S4), J−V curve and EQE spectrum of the PCBM-based device(Figure S5), J−V curves of the devices with differentconcentrations of PCBM (Figure S6), parameters of thedevices varying with the exposure time (Table S1), andthe fitting parameters of the EIS (Table S2) (PDF)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: [email protected].*E-mail: [email protected] Wang: 0000-0001-9612-7257Xianbao Wang: 0000-0001-7765-4027Haoshuang Gu: 0000-0003-1232-2499Jinhua Li: 0000-0002-5226-0272NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was financially supported by the National NaturalScience Foundation of China (Grant Nos. 11574075,61106070, 51673060, 21401049, 51272071, and 11304088),Natural Science Fund for Distinguished Young Scholars ofHubei Province, China (No. 2016CFA036), Hubei ProvincialDepartment of Science & Technology (2015CFB266,2016CFB199, and 2014CFA096), Hubei Provincial Depart-ment of Education (Q2016010 and D201602).

■ REFERENCES(1) Saliba, M.; Matsui, T.; Seo, J. Y.; Domanski, K.; Baena, J. C.;Nazeeruddin, M. K.; Zakeeruddin, S. M.; Tress, W.; Abate, A.;Hagfeldt, A.; Gratzel, M. Cesium-containing Triple Cation PerovskiteSolar Cells: Improved Stability, Reproducibility and High Efficiency.Energy Environ. Sci. 2016, 9, 1989−1997.(2) Son, D. Y.; Lee, J. W.; Choi, Y. J.; Jang, I. H.; Lee, S.; Yoo, P. J.;Shin, H.; Ahn, N.; Choi, M.; Kim, D.; Park, N. G. Self-Formed GrainBoundary Healing Layer for Highly Efficient CH3NH3PbI3 PerovskiteSolar Cells. Nat. Energy 2016, 1, No. 16081.(3) Xu, B.; Bi, D. Q.; Hua, Y.; Liu, P.; Cheng, M.; Gratzel, M.; Kioo,L.; Hagfeldt, A.; Sun, L. C. A Low-Cost Spiro[fluorene-9,9′-xanthene]-Based Hole Transport Material for Highly Efficient Solid-State Dye-Sensitized Solar Cells and Perovskite Solar Cells. Energy Environ. Sci.2016, 9, 873−877.(4) Bi, D.; Tress, W.; Dar, M. I.; Gao, P.; Luo, J.; Renevier, C.;Schenk, K.; Abate, A.; Giordana, F.; Baena, J. C.; Decoppet, J. D.;Zakeeruddin, S. M.; Nazeeruddin, M. K.; Gratzel, M.; Hagfeldt, A.Efficient Luminescent Solar Cells Based on Tailored Mixed-CationPerovskites. Sci. Adv. 2016, 2, No. e1501170.(5) Deng, Y. H.; Dong, Q. F.; Bi, C.; Yuan, Y. B.; Huang, J. S. Air-Stable, Efficient Mixed-Cation Perovskite Solar Cells with CuElectrode by Scalable Fabrication of Active Layer. Adv. Energy Mater.2016, No. 1600372.(6) Mei, A.; Li, X.; Liu, L. F.; Ku, Z. L.; Liu, T. F.; Rong, Y. G.; Xu,M.; Hu, M.; Chen, J. Z.; Yang, Y.; Gratzel, M.; Han, H. W. A Hole-Conductor−Free, Fully Printable Mesoscopic Perovskite Solar Cellwith High Stability. Science 2014, 345, 295−298.(7) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. OrganometalHalide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells. J.Am. Chem. Soc. 2009, 131, 6050−6051.(8) Research Cell Efficiency Records. http://www.nrel.gov/ncpv/images/efficiencychart.

(9) Li, X.; Bi, D. Q.; Yi, C. Y.; Decoppet, J. D.; Luo, J. S.;Zakeeruddin, S. M.; Hagfeldt, A.; Gratzel, M. A Vacuum Flash-AssistedSolution Process for High-Efficiency Large-Area Perovskite Solar Cells.Science 2016, 353, 58−62.(10) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.;Seok, S.Il. High-Performance Photovoltaic Perovskite Layers Fab-ricated Through Intramolecular Exchange. Science 2015, 348, 1234−1237.(11) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok,S.Il. Solvent Engineering for High-Performance Inorganic-OrganicHybrid Perovskite Solar Cells. Nat. Mater. 2014, 13, 897−903.(12) Xing, Y.; Sun, C.; Yip, H. L.; Bazan, G. C.; Huang, F.; Cao, Y.New Fullerene Design Enables Efficient Passivation of Surface Trapsin High Performance p-i-n Heterojunction Perovskite Solar Cells.Nano Energy 2016, 26, 7−15.(13) Tai, Q.; You, P.; Sang, H. Q.; Liu, Z. K.; Hu, C. L.; Chan, H. L.W.; Yan, F. Efficient and Stable Perovskite Solar Cells Prepared inAmbient Air Irrespective of the Humidity. Nat. Commun. 2016, 7,No. 11105.(14) Yi, C.; Li, X.; Luo, J. S.; Zakeeruddin, S. M.; Gratzel, M.Perovskite Photovoltaics with Outstanding Performance Produced byChemical Conversion of Bilayer Mesostructured Lead Halide/TiO2

Films. Adv. Mater. 2016, 28, 2964−2970.(15) Wang, Y. K.; Yuan, Z. C.; Shi, G. Z.; Li, Y. X.; Li, Q.; Hui, F.;Sun, B. Q.; Jiang, Z. Q.; Liao, L. S. Dopant-Free Spiro-Triphenyl-amine/Fluorene as Hole Transporting Material for Perovskite SolarCells with Enhanced Efficiency and Stability. Adv. Funct. Mater. 2016,26, 1375−1381.(16) Chen, Y. H.; Chen, T.; Dai, L. M. Layer-by-Layer Growth ofCH3NH3PbI3−xClx for Highly Efficient Planar Heterojunction Perov-skite Solar Cells. Adv. Mater. 2015, 27, 1053−1059.(17) Wu, C. G.; Chiang, C. H.; Tseng, Z. L.; Nazeeruddin, M. K.;Hagfeldt, A.; Gratzel, M. High Efficiency Stable Inverted PerovskiteSolar Cells Without Current Hysteresis. Energy Environ. Sci. 2015, 8,2725−2733.(18) Zhao, L. C.; Luo, D. Y.; Wu, J.; Hu, Q.; Zhang, W.; Chen, K.;Liu, T. H.; Liu, Y.; Zhang, Y. F.; Liu, F.; Russell, T. P.; Snaith, H. J.;Zhu, R.; Gong, Q. H. High-Performance Inverted Planar Hetero-junction Perovskite Solar Cells Based on Lead Acetate Precursor withEffciency Exceeding 18%. Adv. Funct. Mater. 2016, 26, 3508−3514.(19) Heo, J. H.; Han, H. J.; Kim, D.; Ahn, T. K.; Im, S. H. 18.1%Hysteresis-Less Inverted CH3NH3PbI3 Planar Perovskite Hybrid SolarCells. Energy Environ. Sci. 2015, 8, 1602−1608.(20) Kim, H. B.; Choi, H.; Jeong, J.; Kim, S.; Walker, B.; Song, S.;Kim, J. Y. Mixed Solvents for the Optimization of Morphology inSolution-Processed, Inverted-Type Perovskite/Fullerene Hybrid SolarCells. Nanoscale 2014, 6, 6679−6683.(21) Ke, W.; Fang, G. J.; Liu, Q.; Xiong, L. B.; Qin, P. L.; Tao, H.;Wang, J.; Lei, H. W.; Li, B. R.; Wan, J. W.; Yang, G.; Yan, Y. F. Low-Temperature Solution-Processed Tin Oxide as an Alternative ElectronTransporting Layer for Efficient Perovskite Solar Cells. J. Am. Chem.Soc. 2015, 137, 6730−6733.(22) Ke, W.; Xiao, C. X.; Wang, C. L.; Saparov, B.; Duan, H. S.;Zhao, D. W.; Xiao, Z. W.; Schulz, P.; Harvey, S. P.; Liao, W. Q.; Meng,W. W.; Yu, Y.; Cimaroli, A. J.; Jiang, C. S.; Zhu, K.; Jassim, M. A.;Fang, G. J.; Mitzi, D. B.; Yan, Y. F. Employing Lead ThiocyanateAdditive to Reduce the Hysteresis and Boost the Fill Factor of PlanarPerovskite Solar Cells. Adv. Mater. 2016, 28, 5214−5221.(23) Yang, D.; Zhou, X.; Yang, R. X.; Yang, Z.; Yu, W.; Wang, X. L.;Li, C.; Liu, S. Z.; Chang, R. P. H. Surface Optimization to EliminateHysteresis for Record Efficiency Planar Perovskite Solar Cells. EnergyEnviron. Sci. 2016, 9, 3071−3078.(24) Qiu, W.; Merckx, T.; Jaysankar, M.; Masse de la Huerta, C.;Rakocevic, L.; Zhang, W.; Paetzold, U. W.; Gehlhaar, R.; Froyen, L.;Poortmans, J.; Cheyns, D.; Snaith, H. J.; Heremans, P. Pinhole-FreePerovskite Films for Efficient Solar Modules. Energy Environ. Sci. 2016,9, 484−489.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38475

(25) Liu, D.; Kelly, T. L. Perovskite Solar Cells with a PlanarHeterojunction Structure Prepared Using Room-Temperature Sol-ution Processing Techniques. Nat. Photonics 2014, 8, 133−138.(26) Zhou, H. W.; Shi, Y. T.; Wang, K.; Dong, Q. S.; Bai, X. G.; Xing,Y. J.; Du, Y.; Ma, T. L. Low-Temperature Processed and Carbon-BasedZnO/CH3NH3PbI3/C Planar Heterojunction Perovskite Solar Cells. J.Phys. Chem. C 2015, 119, 4600−4605.(27) Anaraki, E. H.; Kermanpur, A.; Steier, L.; Domanski, K.; Matsui,T.; Tress, W.; Saliba, M.; Abate, A.; Gratzel, M.; Hagfeldt, A.; Correa-Baena, J. P. Highly Efficient and Stable Planar Perovskite Solar Cellsby Solution-Processed Tin Oxide. Energy Environ. Sci. 2016, 9, 3128−3134.(28) Jiang, Q.; Zhang, L. Q.; Wang, H. L.; Yang, X. L.; Meng, J. H.;Liu, M.; Yin, Z. G.; Wu, J. L.; Zhang, X. W.; You, J. B. EnhancedElectron Extraction Using SnO2 for High-Efficiency Planar-StructureHC(NH2)2PbI3-Based Perovskite Solar Cells. Nat. Energy 2016, 2,No. 16177.(29) Heo, J. H.; Lee, M. H.; Han, H. J.; Patil, B. R.; Yu, J. S.; Im, S. H.Highly Efficient Low Temperature Solution Processable Planar TypeCH3NH3PbI3 Perovskite Flexible Solar Cells. J. Mater. Chem. A 2016,4, 1572−1578.(30) Wang, P.; Zhao, J. J.; Liu, J. X.; Wei, L. Y.; Liu, Z. H.; Guan, L.H.; Cao, G. Z. Stabilization of Organometal Halide Perovskite Filmsby SnO2 Coating with Inactive Surface Hydroxyl Groups on ZnONanorods. J. Power Sources 2017, 339, 51−60.(31) Liu, T. H.; Hu, Q.; Wu, J.; Chen, K.; Zhao, L. C.; Liu, F.; Wang,C.; Lu, H.; Jia, S.; Russell, T.; Zhu, R.; Gong, Q. H. Mesoporous PbI2Scaffold for High-Performance Planar Heterojunction Perovskite SolarCells. Adv. Energy Mater. 2016, 6, No. 1501890.(32) 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-InducedPassivation of Hybrid Lead Iodide Perovskites Toward HighPerformance Solar Cells. Nano Lett. 2014, 14, 4158−4163.(33) Liu, D.; Yang, J. L.; Kelly, T. L. Compact Layer Free PerovskiteSolar Cells with 13.5% Efficiency. J. Am. Chem. Soc. 2014, 136, 17116−17122.(34) Yang, J. L.; Siempelkamp, B. D.; Mosconi, E.; Angelis, F. D.;Kelly, T. L. Origin of the Thermal Instability in CH3NH3PbI3 ThinFilms Deposited on ZnO. Chem. Mater. 2015, 27, 4229−4236.(35) Tseng, Z. L.; Chiang, C. H.; Chang, S. H.; Wu, C. G. SurfaceEngineering of ZnO Electron Transporting Layer via Al Doping forHigh Efficiency Planar Perovskite Solar Cells. Nano Energy 2016, 28,311−318.(36) Zuo, L.; Gu, Z. W.; Ye, T.; Fu, W. F.; Wu, G.; Li, H. Y.; Chen,H. Z. Enhanced Photovoltaic Performance of CH3NH3PbI3 PerovskiteSolar Cells Through Interfacial Engineering Using Self-AssemblingMonolayer. J. Am. Chem. Soc. 2015, 137, 2674−2679.(37) Cheng, Y.; Yang, Q. D.; Xiao, J. Y.; Xue, Q. F.; Li, H. W.; Guan,Z. Q.; Yip, H. L.; Tsang, S. W. Decomposition of Organometal HalidePerovskite Films on Zinc Oxide Nanoparticles. ACS Appl. Mater.Interfaces 2015, 7, 19986−19993.(38) Liu, T. H.; Chen, K.; Hu, Q.; Zhu, R.; Gong, Q. H. InvertedPerovskite Solar Cells: Progresses and Perspectives. Adv. Energy Mater.2016, 6, No. 1600457.(39) Fan, R. D.; Huang, Y.; Wang, L. G.; Li, L.; Zheng, G.; Zhou, H.P. The Progress of Interface Design in Perovskite-Based Solar Cells.Adv. Energy Mater. 2016, 6, No. 1600460.(40) 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-ProcessedPerovskite Solar Cells with Millimeter-Scale Grains. Science 2015, 347,522−525.(41) Qiu, W.; Bastos, J. P.; Dasgupta, S.; Merckx, T.; Cardinaletti, I.;Jenart, M. V. C.; Nielsen, C. B.; Gehlhaar, R.; Poortmans, J.;Heremans, P.; McCulloch, I.; Cheyns, D. Highly Efficient PerovskiteSolar Cells with Crosslinked PCBM Interlayers. J. Mater. Chem. A2017, 5, 2466−2472.(42) Zhang, W. H.; Xiong, J.; Wang, S.; Liu, W.; Li, J.; Wang, D. F.;Gu, H. S.; Wang, X. B.; Li, J. H. Highly Conductive and Transparent

Silver Grid/Metal Oxide Hybrid Electrodes for Low-TemperaturePlanar Perovskite Solar Cells. J. Power Sources 2017, 337, 118−124.(43) Li, W.; Fan, J. D.; Li, J. W.; Mai, Y. H.; Wang, L. D. ControllableGrain Morphology of Perovskite Absorber Film by Molecular Self-Assembly toward Efficient Solar Cell Exceeding 17%. J. Am. Chem. Soc.2015, 137, 10399−10405.(44) Li, L.; Chen, Y. H.; Liu, Z. H.; Chen, Q.; Wang, X. D.; Zhou, H.P. The Additive Coordination Effect on Hybrids PerovskiteCrystallization and High-Performance Solar Cell. Adv. Mater. 2016,28, 9862−9868.(45) 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 forHighly Reproducible Planar-Structured Perovskite Solar Cells viaSequential Deposition. Energy Environ. Sci. 2014, 7, 2934−2938.(46) Tao, C.; Velden, J. V. D.; Cabau, L.; Montcada, N. F.; Neutzner,S.; Kandada, A. R. S.; Marras, S.; Brambilla, L.; Tommasini, M.; Xu, W.D.; Sorrentino, R.; Perinot, A.; Caironi, M.; Bertarelli, C.; Palomares,E.; Petrozza, A. Fully Solution-Processed n-i-p-Like Perovskite SolarCells with Planar Junction: How the Charge Extracting LayerDetermines the Open-Circuit Voltage. Adv. Mater. 2017, 29,No. 1604493.(47) Tao, C.; Neutzner, S.; Colella, L.; Marras, S.; Kandada, A. R. S.;Gandini, M.; Bastiani, M. D.; Pace, G.; Manna, L.; Caironi, M.;Bertarelli, C.; Petrozza, A. 17.6% Stabilized Efficiency in Low-Temperature Processed Planar Perovskite Solar Cells. Energy Environ.Sci. 2015, 8, 2365−2370.(48) Wojciechowski, K.; Ramirez, I.; Gorisse, T.; Dautel, O.; Dasari,R.; Sakai, N.; Hardigree, J. M.; Song, S.; Marder, S.; Riede, M.; Wantz,G.; Snaith, H. J. Cross-linkable Fullerene Derivatives for Solution-Processed n-i-p Perovskite Solar Cells. ACS Energy Lett. 2016, 1, 648−653.(49) Ma, T.; Tadaki, D.; Sakuraba, M.; Sato, S.; Iwata, A. H.; Niwano,M. Effects of Interfacial Chemical States on the Performance ofPerovskite Solar Cells. J. Mater. Chem. A. 2016, 4, 4392−4397.(50) Giordano, F.; Abate, A.; Baena, J. P. C.; Saliba, M.; Matsui, T.;Im, S. H.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Hagfeldt, A.;Graetzel, M. Enhanced Electronic Properties in Mesoporous TiO2 viaLithium Doping for High-Efficiency Perovskite Solar Cells. Nat.Commun. 2016, 7, No. 10379.(51) Chiang, C. H.; Nazeeruddin, M. K.; Graetzel, M.; Wu, C. G. TheSynergistic Effect of H2O and DMF Towards Stable and 20%Efficiency Inverted Perovskite Solar Cells. Energy Environ. Sci. 2017,10, 808−817.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.7b10994ACS Appl. Mater. Interfaces 2017, 9, 38467−38476

38476