Colloidal Precursor-Induced Growth of Ultra-Even ... - ACS Publications

Oct 14, 2016 - Yinglong Yang,. ‡ and Shihe Yang*,‡. †. School of Materials Science and Engineering, Beihang University, No. 37 Xueyuan Road, Hai...
0 downloads 0 Views 6MB Size
Download PDF
Research Article www.acsami.org

Colloidal Precursor-Induced Growth of Ultra-Even CH3NH3PbI3 for High-Performance Paintable Carbon-Based Perovskite Solar Cells Xiaowen Chang,† Weiping Li,† Haining Chen,*,†,‡ Liqun Zhu,† Huicong Liu,† Huifang Geng,† Sisi Xiang,† Jiaming Liu,† Xiaoli Zheng,‡ Yinglong Yang,‡ and Shihe Yang*,‡ †

School of Materials Science and Engineering, Beihang University, No. 37 Xueyuan Road, Haidian District, Beijing 100191, People’s Republic of China ‡ Department of Chemistry, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong S Supporting Information *

ABSTRACT: Carbon-based hole transport material (HTM)free perovskite solar cells (PSCs) have attracted intense attention due to their relatively high stability. However, their power conversion efficiency (PCE) is still low, especially for the simplest paintable carbon-based PSCs (C-PSCs), whose performance is greatly limited by poor contact at the perovskite/carbon interface. To enhance interface contact, it is important to fabricate an even-surface perovskite layer in a porous scaffold, which is not usually feasible due to roughness of the crystal precursor. Herein, colloidal engineering is applied to replace the traditional crystal precursor with a colloidal precursor, in which a small amount of dimethyl sulfoxide (DMSO) is added into the conventional PbI2 dimethylformamide (DMF) solution. After deposition, PbI2(DMSO) adduct colloids (which are approximately tens of nanometers in size) are stabilized and dispersed in DMF to form a colloidal film. Compared with PbI2 and PbI2(DMSO) adduct crystal precursors deposited from pure DMF and DMSO solvents, respectively, the PbI2(DMSO) adduct colloidal precursor is highly mobile and flexible, allowing an ultra-even surface to be obtained in a TiO2 porous scaffold. Furthermore, this ultra-even surface is well-maintained after chemical conversion to CH3NH3PbI3 in a CH3NH3I solution. As a result, the contact at the CH3NH3PbI3/carbon interface is significantly enhanced, which largely boosts the fill factor and PCE of C-PSCs. Impressively, the achieved champion PCE of 14.58% is among the highest reported for C-PSCs. KEYWORDS: carbon-based perovskite solar cells, hole transport material-free, interface contact, dimethyl sulfoxide, dimethylformamide

■

INTRODUCTION

distinct reasons for the different structure-specific devices of CPSCs. The complicated device structure of the multilayer meso C-PSCs developed by Han’s group16,17,25 makes the deposition of high-quality perovskite difficult. The embedment method for producing C-PSCs developed by Yang’s group19,26−28 has simplified the structure of these devices. However, the carbon electrode on the PbI2 layer is weak, which may induce many cracks and decrease the device’s performance. The development of paintable C-PSCs further simplified the fabrication process and enhanced the mechanical properties of the electrode, but the postdeposition of a carbon electrode on perovskite leads to poor contact at the interface.29−33 Recently, we partially solved this problem by fabricating an evenperovskite surface to enhance the interfacial contact through solvent engineering (introducing cyclohexane into a MAI IPA solution) during the conversion of PbI2 to MAPbI3, which

Organic−inorganic hybrid perovskite solar cells (PSCs) have attracted much attention due to their rapid increase in power conversion efficiency (PCE) from 3.8% in 2009 to 22.1% in 2016.1−7 However, the low stability of these devices has strongly prohibited their practical application. This low stability results not only from the characteristic properties of hybrid organic−inorganic materials but also from unstable and airsensitive organic hole transport materials (HTMs; such as spiro-OMeTAD).8,9 Fortunately, researchers have proved that perovskite (e.g., CH3NH3PbI3 or MAPbI3) can serve as both a light harvester and hole transporter. 10−15 This unique ambipolar nature makes HTM-free PSCs possible. Promisingly, low-cost and highly stable carbon materials have been demonstrated to serve as an efficient hole extraction electrode in HTM-free PSCs. Carbon-based HTM-free PSCs (C-PSCs) have been widely accepted as the most stable PSCs and have the most potential to be used in commercialized devices.16−24 Unsatisfactorily, the PCEs of C-PSCs still largely lag behind those of traditional PSCs with HTMs. Additionally, there are © 2016 American Chemical Society

Received: August 8, 2016 Accepted: October 14, 2016 Published: October 14, 2016 30184

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

Figure 1. Morphology of the DMF (H), DMF, DMSO, and DMF/DMSO Pb−I precursors: (a1−d1) solvent composition and optical photographs, (a2−d2) SEM images, and (a3−d3) AFM images.

increased the PCE to over 14%.34 However, although the use of a mixed solvent can avoid the generation of large MAPbI3 crystals, the rough morphology of the PbI2 crystal layer due to its ease of crystallization from N,N-dimethylformamide (DMF) still limits the evenness of MAPbI3, which suppresses the interfacial contact and photovoltaic performance of paintable C-PSCs. As demonstrated above, preparing an even PbI2 layer is necessary for a more even MAPbI3. Additionally, prohibiting the crystallization of PbI2 crystals after deposition should be a useful strategy to prevent the roughening of the surface. Dimethyl sulfoxide (DMSO) could retard the crystallization of PbI2, but the strong interaction between DMSO and Pb2+ would facilitate the formation of PbI2(DMSO) adduct crystals, which would also roughen the Pb−I precursor.7,35−39 Herein, in order to prevent the crystallization of PbI2 and PbI2(DMSO) adducts, a mixed solvent of DMF/DMSO was applied in which DMSO could retard the crystallization of PbI2; furthermore, the presence of DMF could also suppress the formation of PbI2(DMSO) adduct crystals. It is demonstrated that a PbI2 DMF/DMSO solution can achieve an ultra-even Pb−I precursor containing small PbI2(DMSO) adduct colloids (tens of nanometers in size). This precursor afforded MAPbI3 with an ultra-even surface that enhanced the interfacial contact with the carbon electrode, which significantly improved the fill factor (FF) of these paintable C-PSCs. A PCE of 14.58% was

achieved for the paintable C-PSCs, a value among the highest PCE levels for C-PSCs.5,20,22,23,34,40−45

■

RESULTS AND DISCUSSION To prepare the colloidal Pb−I precursor, a mixed solvent containing DMF and DMSO (9:1 volume ratio) was used as the solvent for the PbI2 solution. In order to make a reasonable comparison, pure DMF and DMSO were also used as solvents for the PbI2 solution. To deposit the Pb−I precursor on a TiO2 porous scaffold, spin-coating was applied. The conventional PbI2 layer was prepared by spin-coating the PbI2 DMF solution followed by heating at 100 °C for 5 min; this is termed the DMF (H) Pb−I precursor. Additionally, a PbI2 layer prepared from the PbI2 DMF solution without heating was also investigated and termed the DMF Pb−I precursor. Similarly, Pb−I precursor films prepared from PbI2 DMSO and DMF/ DMSO solutions without heating were termed the DMSO and DMF/DMSO Pb−I precursors, respectively. The four as-prepared Pb−I precursor layers show obviously different colors, as shown in Figure 1a1−d1, from brown yellow for DMF (H) Pb−I, bright yellow for DMF Pb−I, and almost transparent for DMSO Pb−I. The difference in color indicates possible variations in their morphologies and/or compositions. To evaluate the differences in their morphologies, SEM (Figure 1a2−d2) and AFM (Figure 1a3−d3) images were recorded. The conventional DMF (H) Pb−I precursor exhibits a mountain30185

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

Figure 2. Characterization of the different Pb−I precursor films: XRD patterns (a), FT-IR spectra (b), UV−vis diffuse reflectance spectra (c), and schematic of the formation (d) of the DMF (H), DMF, DMSO, and DMF/DMSO Pb−I precursors.

the small angle region are likely related to PbI2(DMF) adducts as the DMF molecules could be easily inserted into the gaps between PbI2 layers. When the Pb−I precursor was deposited from a DMSO solvent, many other new diffraction peaks at around 9.8, 15.8, 18.6, 21.7, and 29.5° appeared, which could be indexed to the previously reported diffraction peaks of PbI2(DMSO) adducts.7,39 Although DMSO has been proven to effectively prevent the crystallization of PbI2 due to the strong interaction between Pb2+ and DMSO molecules, this strong interaction has been widely observed to form PbI2(DMSO) adducts, in which DMSO molecules are inserted into the gaps between the PbI2 layers.7,39,46 Our present result indicates that PbI2(DMSO) adducts crystallize easily, which induces the formation of a rough surface morphology for the DMSO Pb−I precursor. The FT-IR spectra also confirmed the presence of a large amount of DMSO molecules in the asprepared DMSO Pb−I precursor. As a mixed solvent (DMF/ DMSO = 9:1) was applied, the XRD pattern still presented some diffraction peaks of PbI2(DMSO) adducts. However, the diffraction peaks become much weaker, implying that the crystallization of PbI2(DMSO) adducts is suppressed and prohibited in this mixed solvent system. The FT-IR spectra in Figure 2b indicate the presence of a large amount of DMF molecules in the DMF/DMSO Pb−I precursor, which should play a role in preventing the crystallization and growth of PbI2(DMSO) adducts. The small PbI2(DMSO) crystals or colloidal particles would be more mobile and flexible, allowing the formation of a compact and ultra-even Pb−I precursor film. In order to highlight the difference in grain size, we calculated the grain size D for the different Pb−I precursors, according to the Debye−Scherrer formula, D = Kγ/(B cos θ), where K is the Scherrer constant, γ is the X-ray wavelength, B is the full width at half-maximum (FMHM) of the diffraction peaks, and θ represents relevant diffraction angles. The grain sizes for the DMF (H), DMF, and DMSO Pb−I precursors were calculated to be about 359, 193, and 310 nm, respectively, whereas the grain size for the DMF/DMSO Pb−I precursor was significantly reduced to only about 59 nm. These results

like surface, suggesting a rough surface. This morphology was confirmed from the AFM image, in which the roughness factor was measured to be about 37 nm. In addition to having a similar mountain-like surface, many small pinholes also exist on the DMF Pb−I precursor, which was not heat treated; this is well-reflected in its AFM image, with the roughness factor increasing to 54 nm. Differing from the above two precursors, the DMSO Pb−I precursor shows a completely different morphology with two significantly different parts. One part shows an even surface but with many obvious cracks, whereas the other part shows many vivid grain-like embossments, partially suggesting strong crystallinity. These two different regions could also be well-observed from the AFM images, with a low roughness factor (14 nm) (Figure S2) for the even part and a considerably higher roughness factor (107 nm) for the strong crystallinity region. Amazingly, a significantly even surface was obtained with the DMF/DMSO Pb−I precursor, which is too even to be observed from an SEM image, showing no obvious grain features. This even morphology was further confirmed from an AFM image, and the roughness was measured to be only about 6.7 nm, much smaller than the above three precursors. As observed, a significant difference in morphology was obtained among the precursors, which suggests possible compositional variation. To explain this phenomenon, XRD patterns, FT-IR spectra, and UV−vis spectra were recorded. As presented in Figure 2a, in addition to the diffraction peaks of the FTO/TiO2 substrates, DMF(H) Pb−I presents many intense diffraction peaks at 12.7, 26.0, 28.3, 34.3, 38.7, and 52.4°, which are consistent with the characteristic peak positions of PbI2. Therefore, the DMF (H) Pb−I precursor is composed of a pure phase, high-crystallinity PbI2 crystals, and DMF molecules are almost non-observable from the FT-IR spectra (Figure 2b). Because no heat treatment process was used for the DMF Pb−I precursor, the characteristic peaks of PbI2 crystals become weaker and two additional weak diffraction peaks appear in the region from 5 to 10°. Since the FT-IR spectra exhibit an obvious CON peak for the DMF Pb−I precursor, these two additional diffraction peaks in 30186

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

Figure 3. Morphology of different MAPbI3 films fabricated from the DMF (H), DMF, DMSO, and DMF/DMSO Pb−I precursors: optical photographs (a1−d1), SEM images (a2−d2), grain size distribution histograms (a3−d3), and AFM images (a4−d4) of DMF(H), DMF, DMSO, and DMF/DMSO MAPbI3.

crystallization was obtained for the DMF Pb−I precursor, accompanied by a small number of PbI2(DMF) adducts. Since DMSO has a low volatility and strong coordination with Pb2+, no PbI2 crystals would form in the DMSO Pb−I precursor, but it tends to form large PbI2(DMSO) crystals.7,39,46 When the DMF/DMSO mixed solvent was applied, DMF molecules thoroughly dilute the DMSO molecules. The generation of PbI2(DMSO) adduct crystals may terminate at the initial stage after nucleation due to the limited number of DMSO molecules. As a result, only small grains or colloidal particles were produced for the DMF/DMSO Pb−I precursor. These small grains or colloidal particles are proposed to stabilize and disperse in DMF, which favors the high mobility and flexibility of the Pb−I precursor that allows it to obtain an even surface. To convert the Pb−I precursors to MAPbI3, a MAI IPA/ CYHEX solution was applied, in which the addition of CYHEX promotes the formation of MAPbI3 and suppresses the Ostwald ripening process to obtain a high-quality perovskite film. Figure 3a1−d1 displays optical photographs of the MAPbI3 films obtained from the DMF (H), DMF, DMF/DMSO, and DMSO Pb−I precursors, indicating that all of the films appear dark brown after conversion. An optical photograph of all four MAPbI3 films from a different direction (Figure S3) vividly depicts that the DMF/DMSO MAPbI3 film is more shiny than the other MAPbI3 films, demonstrating a mirror-like, even property for the MAPbI3 film prepared from DMF/DMSO solvent. SEM (Figure 2a2−d2,a3−d3) and AFM (Figure 2a4−d4) images were taken to further evaluate the surface morphologies of the different MAPbI3 films. Obvious cubic crystals were observed for all four MAPbI3 films. The DMF (H) MAPbI3 film

vividly confirm the presence of much smaller grains in the DMF/DMSO Pb−I precursor. UV−vis reflectance spectra were also recorded to study the different Pb−I precursors. As shown in Figure 2c, the DMF (H) and DMF Pb−I precursors exhibited a considerably higher baseline reflectance than the DMSO and DMF/DMSO Pb−I precursors, partially suggesting that the former two precursors have surfaces that are more rough. In addition, the main reflectance onset for the DMF (H) and DMF Pb−I precursors is similar at about 520 nm, consistent with the characteristic absorption onset for a crystallized PbI2 semiconductor due to its band gap of 2.3 eV, which further confirms that the main composition is PbI2 in these two precursors. In a distinctive comparison, no reflectance onset at 520 nm was observed for the DMSO Pb−I precursor, but an onset at a shorter wavelength region of about 450 nm appears, reflecting the absorption property of PbI2(DMSO) adducts.36,46 Compared with the reflectance onset of the DMSO Pb−I precursor, a slight red shift phenomenon from 450 to 470 nm is observed for the DMF/DMSO Pb−I precursor, which could be partially attributed to the quantum confinement effect of small crystals or colloidal particles.36 Therefore, our UV−vis results further confirmed our above analysis of the XRD and FT-IR spectra. To better understand the above analysis, we have illustrated the features of the different film compositions in Figure 2d. Heat treatment easily evaporates most of the DMF molecules and PbI2 strongly crystallizes, resulting in large PbI2 crystal grains for the DMF(H) Pb−I precursor. If no post heat treatment process was applied, then some DMF molecules remain in the precursor and some would insert into the gaps between the PbI2 layers. As a result, relatively weak 30187

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

precursors. The slower conversion for the DMF(H) and DMF Pb−I precursors is likely due to the difficulty of inserting MAI molecules into the highly crystallized PbI2. This would also suppress the number of nucleation sites (known as localized nucleation) and induce the growth of large crystals,47 which explains the previous SEM and AFM results. The exchange between DMSO and MAI would promote the conversion from the DMSO Pb−I precursor to MAPbI3.7 However, the poor morphology of the high-crystallization DMSO Pb−I precursor still leads to a rough MAPbI3 film with lots of pinholes. More importantly, the exchange between DMF/DMSO and MAI not only maintains the ultra-even surface morphology of the Pb−I precursor but also promotes the conversion, allowing for a pure MAPbI3 film to be obtained. UV−vis reflectance spectra in the 300−850 nm wavelength region, as shown in Figure 4b, demonstrated much weak reflection from 350 to 780 nm, which is identical to the light absorption range of MAPbI3. The large baseline reflectance for DMF (H) and DMF MAPbI3 indicate their relatively rough surface morphologies. DMF/DMSO MAPbI3 showed the weakest baseline reflectance, suggesting that it has the most even surface morphology. The largest baseline reflectance was obtained for DMSO MAPbI3, which is consistent with it having the roughest surface morphology, as indicated in the SEM and AFM images in Figure 3c2,c4. Therefore, the UV−vis reflectance spectra for the different MAPbI3 films exhibit the same trends as the different Pb−I precursors, which further confirms the close relation between the morphologies of the Pb−I precursors and those of the final MAPbI3. As demonstrated in our previous work,34 the morphology of MAPbI3 perovskite layer greatly influences the electric contact and charge transfer at the perovskite/carbon interface in paintable carbon-based PSCs, which directly impacts the performance of the cell. To evaluate the effects of perovskite layers with different morphologies on cell performance, paintable carbon-based PSCs were fabricated by directly printing a commercial carbon paste on perovskite layers (as illustrated in Figure 5a1,a2), followed by annealing at 100 °C for 60 min. In this cell structure, photogenerated electrons on the conduction band (CB) of MAPbI3 are injected into the CB of TiO2, whereas the holes on the valence band (VB) of MAPbI3 are extracted by the carbon electrode. As shown in Figure S5, totally different interfacial contact features were observed from the cross-sectional SEM images. The more rough MAPbI3 layer exhibits a weaker interfacial contact, and the DMF/DMSO MAPbI3 film affords a very intimate contact with the carbon electrode. As demonstrated in previous work, a better interfacial contact promotes more fluent charge transfer at the interface and leads to a higher FF value for C-PSCs.34 Therefore, current density−voltage (J−V) curves were recorded for the C-PSCs fabricated from different MAPbI3 films, and the results are displayed in Figure 5b and Table S1. Clearly, the photovoltic performance, especially FF values (0.61, 0.52, 0.50, and 0.66 for DMF (H), DMF, DMSO, and DMF/DMSO, respectively), is closely related to the evenness of MAPbI3, and the most even film, DMF/DMSO MAPbI3, achieves a significantly higher performance (Voc = 1.00 V, Jsc = 20.80 mA/cm2, FF = 0.66, and PCE = 13.71%) than that of other MAPbI3 films. As a side note, DMF MAPbI3 yielded a lower FF and PCE than DMF (H) MAPbI3, although the former has a lower roughness value. This may be attributed to a more obvious fluctuation that forms on the macroscale for DMF MAPbI3, which would weaken the interfacial contact with

presents a lot of pinholes among the cubic crystals, and the grain size ranges from dozens of nanometers to several hundreds of nanometers, with a mean size of 114.7 nm, which affords a roughness of 51 nm. The cubic crystal size for the DMF MAPbI3 film is reduced, with a mean size of 86.0 nm, and the roughness slightly decreases to 47 nm. However, as observed in the AFM image (Figure 3b4), more obvious fluctuations occur on the macroscale. A significantly different morphology was observed for the DMSO MAPbI3 film. Cubic crystals are loosely dispersed with an obvious increase in the crystal grains, with a considerably large mean size of 147.4 nm. The AFM image indicates that the surface is much rougher, with the roughness increasing up to 68 nm. Amazingly, a very compact surface is obtained for the DMF/DMSO MAPbI3 film, and its grain size became smaller, with the mean size decreasing to 79.4 nm. The compact surface and smaller grain size led to an ultra-even and uniform surface with a significantly smaller roughness of 13 nm, as confirmed by the AFM image. Obviously, the surface roughness and uniformity trends of the MAPbI3 films are almost consistent with those of the Pb−I precursors, indicating that the Pb−I precursor has an effect on MAPbI3 films. XRD patterns (Figure 4a) were further recorded to study the compositions of the prepared MAPbI3 films. All of the MAPbI3

Figure 4. Composition and light absorption properties of different MAPbI3 films: (a) XRD patterns and (b) UV−vis diffuse reflectance spectra.

samples exhibited intense diffraction peaks at 14.1, 19.9, 24.5, 28.4, 31.9, 34.9, 40.6, and 50.3°, which are consistent with those characteristic planes of MAPbI3, suggesting a high degree of conversion by immersing the Pb−I precursors into the MAI solution. However, there still were diffraction peaks of PbI2 at 12.7° for DMF (H) and DMF MAPbI3, indicating that the DMF(H) and DMF Pb−I precursors could not be completely converted. Clearly, the complete conversion of Pb−I precursors to MAPbI3 was achieved for the DMSO and DMF/DMSO Pb−I precursors, suggesting a faster conversion for these two 30188

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

Figure 5. Device architecture and performance of the paintable carbon-based PSCs with different MAPbI3 films. Schematic illustration of the (a1) deposition process of the carbon electrode, (a2) cross-sectional structure, and (a3) working principle; (b) J−V curves; (c) steady PCE at voltages close to the maximum output point (0.75, 0.75 0.60, and 0.82 V for DMF (H), DMF, DMSO, and DMF/DMSO, respectively); and (d) photovoltaic parameter distribution for the different devices.

values are around 75−95% in the absorption range of MAPbI3 from 400 to 760 nm, which affords a corresponding integrated Jsc of 20.23 mA/cm2, consistent with the Jsc calculated from the J−V curve.

the carbon electrode containing macro graphite sheets (as shown in Figure S5). These results indicate the importance of the evenness of MAPbI3, which enhances the interfacial contact and consequently the FF and PCE for C-PSCs. The IPCE spectra in Figure S6 show a good consistency between the Jsc values from the J−V curves, and the J−V curves with forward and reverse scans in Figure S7 demonstrate a higher hysteresis for the DMSO device but a smaller hysteresis for the other devices. The steady PCE at a voltage close to the maxium output point further confirms the real performance demonstration from the J−V results (Figure 5c). Photovoltaic parameter distributions were also collected and are presented in Figure 5d to evaluate the performance reproducibility of the different MAPbI3 films in C-PSCs. As indicated, the more even DMF (H) and DMF/DMSO MAPbI3 films also showed higher performance reproducibility in C-PSCs than the rougher DMF and DMSO MAPbI3 films. After preliminary optimization, the best carbon-based paintable PSCs fabricated from the IPA/CYHEX solvent yielded a PCE of 14.58%, resulting from a Voc of 1.0 V, Jsc of 21.83 mA/cm2, and FF of 0.67 (Figure 6a). So far, this is the highest reported PCE value for a carbon-based HTM-free PSC.25−32,34,40,48 The IPCE spectrum of the best performing paintable C-PSCs in Figure 6b indicates that most of the IPCE

■

CONCLUSIONS We have systematically demonstrated the effects of the precursor solvent on the composition and morphology of Pb−I precursors and their final MAPbI3 films. The Pb−I precursors (with or without heat treatment) deposited from DMF solvent were mainly composed of PbI2 crystals, whereas Pb−I precursors deposited from DMSO tended to generate PbI2(DMSO) adduct crystals. Both PbI2 and PbI2(DMSO) adduct crystals demonstrated rough surfaces, resulting in rough MAPbI3 films, which led to poor interfacial contacts with the carbon electrode and low performances in paintable C-PSCs. After exploiting a mixed precursor solvent (DMF/DMSO), PbI2(DMSO) adduct colloidal particles with a small grain size (tens of nanometers in size) constructed the Pb−I precursor, which was more mobile and flexible, forming an ultra-even surface. As a result, an ultra-even MAPbI3 film was obtained after chemical conversion in MAI IPA/CYHEX solution, which afforded a significantly enhanced interfacial contact with the carbon electrode and hence boosted both the FF and PCE. The 30189

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces

Deposition of a Carbon Counter Electrode. A commercial carbon paste produced by Guangzhou Seaside Technology Co., Ltd. was used for the carbon electrode, which was composed of carbon black and graphite.32,34 The carbon paste was painted on the perovskite layer under ambient conditions followed by heating at 100 °C for 60 min. Characterizations. Surface and cross-section morphologies were observed by a JSM-7500F field-emission scanning electron microscope (FESEM). X-ray diffraction (XRD) patterns were recorded by a Rigaku D/MAX-2500 X-ray diffractometer with Cu Kα radiation. Diffuse reflectance spectra were measured by a UV-3000 ultraviolet and visible (UV−vis) spectrophotometer. FT-IR spectra were recorded by a Thermo-Nicolet Nexus 470 FTIR. Surface morphologies and roughnesses were characterized by a Bruker Dimension ICON atomic force microscope (AFM). Optical photographs of the samples were taken by a digital camera (Olympus E-PL1, Japan). A solar light simulator (Newport solar simulator, model number 6255, 150 W Xe lamp, AM 1.5 global filter) was calibrated to 1 sun (100 mW cm−2) using a silicon reference solar cell equipped with a KG-5 filter. The active cell area was fixed at 5 mm2 with a metal mask, and J−V curves were recorded on an IM6x electrochemical workstation (ZAHNER-Elektrik GmbH & Co., KG, Germany). IPCE spectra were recorded using the IPCE kit developed by ZAHNER-Elektrik in AC mode with a frequency of 1 Hz.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b09925. SEM images and AFM images of Pb−I precursor and MAPbI3, cross-sectional SEM images and performance parameters of C-PSCs devices (PDF)

Figure 6. Performance of the champion device based on DMF/DMSO MAPbI3. (a) J−V curve and (b) IPCE spectrum and the corresponding integrated Jsc.

■

achieved champion PCE of 14.58% is among the highest reported for C-PSCs. Therefore, our present work has highlighted an effective colloidal precursor strategy to prepare ultra-even perovskite layers for high-performance paintable CPSCs.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (H.C.). *E-mail: [email protected] (S.Y.). Notes

The authors declare no competing financial interest.

■

EXPERIMENTAL SECTION

Preparation of the TiO2 Blocking Layer and Porous Scaffold. First, the TiO2 blocking layer was deposited on cleaned FTO glass by spin-coating 0.15 M titanium diisopropoxide bis(acetylacetonate) in 1butanol solution at 2000 rpm for 20 s; afterward, it was heated at 125 °C for 5 min. Then, the TiO2 porous scaffold was spin-coated at 5000 rpm for 30 s using a commercial TiO2 paste (Dyesol 30 NRD, Dyesol) that was diluted by ethanol at a weight ratio of 1:2.5 and heated at 100 °C for 5 min. Afterward, the temperature was gradually heated to 550 °C and held for 2 h. After that, the TiO2 scaffold was cooled to room temperature.32,34 Preparation of CH3NH3I. CH3NH3I was synthesized by reacting methylamine and hydroiodic acid at a molar ratio of 1.6:1 under a N2 atmosphere in an ice bath for 2 h with stirring. After rotary evaporation at 50 °C for 1 h, the obtained precipitate was washed by normal hexane three times and then dried at 60 °C in vacuum for 24 h to acquire CH3NH3I. Deposition of the PbI2 Film. Three kinds of PbI2 solutions were prepared by dissolving 1.4 M PbI2 into DMF, DMSO/DMF (VDMSO/ VDMF = 1:9), and DMSO at 80 °C, respectively. PbI2 layers with different parameters were deposited into TiO2 scaffolds maintained at 80 °C by spin-coating at 2000 rpm for 20 s (some of them needed to be annealed at 100 °C for 5 min). Conversion of the PbI2 Film to a CH3NH3PbI3 Film. First, 1 mg/mL CH3NH3I was dissolved into IPA/CYHEX (VIPA/VCYHEX = 1:9) to make the CH3NH3I solution. Then, PbI2 samples of different parameters were immersed into the forward solution for 12 h to be converted into CH3NH3PbI3. Finally, the as-prepared CH3NH3PbI3 samples were heated at 100 °C for 15 min.

ACKNOWLEDGMENTS This work was financially supported by the Young Talent of “Zhuoyue” Program of Beihang University, National Natural Science Foundation 268 of China (No. 51371020), HK-RGC General Research Funds (GRF No. HKUST 606511), and the HK Innovation and Technology Fund (ITS/004/14).

■

REFERENCES

(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 (17), 6050−6051. (2) Im, J. H.; Lee, C. R.; Lee, J. W.; Park, S. W.; Park, N. G. 6.5% efficient perovskite quantum-dot-sensitized solar cell. Nanoscale 2011, 3 (10), 4088−4093. (3) Kim, H. S.; Lee, C. R.; Im, J. H.; Lee, K. B.; Moehl, T.; Marchioro, A.; Moon, S. J.; Humphry-Baker, R.; Yum, J. H.; Moser, J. E.; Gratzel, M.; Park, N. G. Lead iodide perovskite sensitized all-solidstate submicron thin film mesoscopic solar cell with efficiency exceeding 9%. Sci. Rep. 2012, 2, 591. (4) Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. Nature 2013, 499 (7458), 316−319. (5) Zhou, H.; Chen, Q.; Li, G.; Luo, S.; Song, T.-b.; Duan, H.-S.; Hong, Z.; You, J.; Liu, Y.; Yang, Y. Interface engineering of highly efficient perovskite solar cells. Science 2014, 345 (6196), 542−546.

30190

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces (6) http://www.nrel.gov/ncpv/images/efficiency_chart.jpg. (7) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.; Seok, S. I. SOLAR CELLS. High-performance photovoltaic perovskite layers fabricated through intramolecular exchange. Science 2015, 348 (6240), 1234−1237. (8) Niu, G.; Guo, X.; Wang, L. Review of recent progress in chemical stability of perovskite solar cells. J. Mater. Chem. A 2015, 3, 8970. (9) Wang, D.; Wright, M.; Elumalai, N. K.; Uddin, A. Stability of perovskite solar cells. Sol. Energy Mater. Sol. Cells 2016, 147, 255−275. (10) Stranks, S. D.; Eperon, G. E.; Grancini, G.; Menelaou, C.; Alcocer, M. J. P.; Leijtens, T.; Herz, L. M.; Petrozza, A.; Snaith, H. J. Electron-Hole Diffusion Lengths Exceeding 1 Micrometer in an Organometal Trihalide Perovskite Absorber. Science 2013, 342 (6156), 341−344. (11) Xing, G.; Mathews, N.; Sun, S.; Lim, S. S.; Lam, Y. M.; Grätzel, M.; Mhaisalkar, S.; Sum, T. C. Long-Range Balanced Electron- and Hole-Transport Lengths in Organic-Inorganic CH3NH3PbI3. Science 2013, 342 (6156), 344−347. (12) Etgar, L.; Gao, P.; Xue, Z.; Peng, Q.; Chandiran, A. K.; Liu, B.; Nazeeruddin, M. K.; Grätzel, M. Mesoscopic CH3NH3PbI3/TiO2 Heterojunction Solar Cells. J. Am. Chem. Soc. 2012, 134 (42), 17396− 17399. (13) Laban, W. A.; Etgar, L. Depleted hole conductor-free lead halide iodide heterojunction solar cells. Energy Environ. Sci. 2013, 6 (11), 3249. (14) Aharon, S.; Gamliel, S.; El Cohen, B.; Etgar, L. Depletion region effect of highly efficient hole conductor free CH3NH3PbI3 perovskite solar cells. Phys. Chem. Chem. Phys. 2014, 16 (22), 10512−10518. (15) Etgar, L. Semiconductor Nanocrystals as Light Harvesters in Solar Cells. Materials 2013, 6 (2), 445. (16) Mei, A.; Li, X.; Liu, L.; Ku, Z.; Liu, T.; Rong, Y.; Xu, M.; Hu, M.; Chen, J.; Yang, Y.; Grätzel, M.; Han, H. A hole-conductor−free, fully printable mesoscopic perovskite solar cell with high stability. Science 2014, 345 (6194), 295−298. (17) Li, X.; Tschumi, M.; Han, H.; Babkair, S. S.; Alzubaydi, R. A.; Ansari, A. A.; Habib, S. S.; Nazeeruddin, M. K.; Zakeeruddin, S. M.; Grätzel, M. Outdoor Performance and Stability under Elevated Temperatures and Long-Term Light Soaking of Triple-Layer Mesoporous Perovskite Photovoltaics. Energy Technology 2015, 3 (6), 551−555. (18) Xu, X.; Liu, Z.; Zuo, Z.; Zhang, M.; Zhao, Z.; Shen, Y.; Zhou, H.; Chen, Q.; Yang, Y.; Wang, M. Hole selective NiO contact for efficient perovskite solar cells with carbon electrode. Nano Lett. 2015, 15 (4), 2402−2408. (19) Chen, H.; Wei, Z.; Yan, K.; Yi, Y.; Wang, J.; Yang, S. Liquid phase deposition of TiO2 nanolayer affords CH3NH3PbI3nanocarbon solar cells with high open-circuit voltage. Faraday Discuss. 2014, 176, 271−286. (20) Chan, C.-Y.; Wang, Y.; Wu, G.-W.; Wei-Guang Diau, E. Solventextraction crystal growth for highly efficient carbon-based mesoscopic perovskite solar cells free of hole conductors. J. Mater. Chem. A 2016, 4, 3872−3878. (21) Liu, Z.; Shi, T.; Tang, Z.; Sun, B.; Liao, G. Using a lowtemperature carbon electrode for preparing hole-conductor-free perovskite heterojunction solar cells under high relative humidity. Nanoscale 2016, 8 (13), 7017−7023. (22) Li, H.; Cao, K.; Cui, J.; Liu, S.; Qiao, X.; Shen, Y.; Wang, M. 14.7% efficient mesoscopic perovskite solar cells using single walled carbon nanotubes/carbon composite counter electrodes. Nanoscale 2016, 8 (12), 6379−6385. (23) Zheng, X.; Wei, Z.; Chen, H.; Zhang, Q.; He, H.; Xiao, S.; Fan, Z.; Wong, K. S.; Yang, S. Designing nanobowl arrays of mesoporous TiO2 as an alternative electron transporting layer for carbon cathodebased perovskite solar cells. Nanoscale 2016, 8 (12), 6393−6402. (24) Chen, H.; Zheng, X.; li, q.; Yang, Y.-L.; xiao, S.; hu, c.; Bai, Y.; Zhang, T.; Wong, K. S.; Yang, S. Amorphous Precursor Route to the Conformable Oriented Crystallization of CH3NH3PbBr3 in Mesoporous Scaffolds: Toward Efficient and Thermally Stable Carbonbased Perovskite Solar Cells. J. Mater. Chem. A 2016, 4, 12897−12912.

(25) Yang, Y.; Ri, K.; Mei, A.; Liu, L.; Hu, M.; Liu, T.; Li, X.; Han, H. The size effect of TiO2 nanoparticles on a printable mesoscopic perovskite solar cell. J. Mater. Chem. A 2015, 3 (17), 9103−9107. (26) Wei, Z.; Chen, H.; Yan, K.; Yang, S. Inkjet Printing and Instant Chemical Transformation of a CH3NH3PbI3/Nanocarbon Electrode and Interface for Planar Perovskite Solar Cells. Angew. Chem. 2014, 126 (48), 13455−13459. (27) Wei, Z.; Yan, K.; Chen, H.; Yi, Y.; Zhang, T.; Long, X.; Li, J.; Zhang, L.; Wang, J.; Yang, S. Cost-efficient clamping solar cells using candle soot for hole extraction from ambipolar perovskites. Energy Environ. Sci. 2014, 7 (10), 3326−3333. (28) Yan, K. Y.; Wei, Z. H.; Li, J. K.; Chen, H. N.; Yi, Y.; Zheng, X. L.; Long, X.; Wang, Z. L.; Wang, J. N.; Xu, J. B.; Yang, S. H. HighPerformance Graphene-Based Hole Conductor-Free Perovskite Solar Cells: Schottky Junction Enhanced Hole Extraction and Electron Blocking. Small 2015, 11 (19), 2269−2274. (29) Yang, Y. Y.; Xiao, J. Y.; Wei, H. Y.; Zhu, L. F.; Li, D. M.; Luo, Y. H.; Wu, H. J.; Meng, Q. B. An all-carbon counter electrode for highly efficient hole-conductor-free organo-metal perovskite solar cells. RSC Adv. 2014, 4 (95), 52825−52830. (30) Zhang, F.; Yang, X.; Wang, H.; Cheng, M.; Zhao, J.; Sun, L. Structure Engineering of Hole−Conductor Free Perovskite-Based Solar Cells with Low-Temperature-Processed Commercial Carbon Paste As Cathode. ACS Appl. Mater. Interfaces 2014, 6 (18), 16140− 16146. (31) Zhou, H. W.; Shi, Y. T.; Dong, Q. S.; Zhang, H.; Xing, Y. J.; Wang, K.; Du, Y.; Ma, T. L. Hole-Conductor-Free, Metal-ElectrodeFree TiO2/CH3NH3PbI3 Heterojunction Solar Cells Based on a Low-Temperature Carbon Electrode. J. Phys. Chem. Lett. 2014, 5 (18), 3241−3246. (32) Chen, H.; Wei, Z.; Zheng, X.; Yang, S. A scalable electrodeposition route to the low-cost, versatile and controllable fabrication of perovskite solar cells. Nano Energy 2015, 15 (0), 216− 226. (33) Zheng, X.; Wei, Z.; Chen, H.; Zhang, Q.; He, H.; Xiao, S.; Fan, Z.; Wong, K. S.; Yang, S. Designing nanobowl arrays of mesoporous TiO(2) as an alternative electron transporting layer for carbon cathode-based perovskite solar cells. Nanoscale 2016, 8 (12), 6393− 6402. (34) Chen, H.; Wei, Z.; He, H.; Zheng, X.; Wong, K. S.; Yang, S. Solvent Engineering Boosts the Efficiency of Paintable Carbon-Based Perovskite Solar Cells to Beyond 14%. Adv. Energy Mater. 2016, 6 (8), 1502087. (35) Rong, Y.; Tang, Z.; Zhao, Y.; Zhong, X.; Venkatesan, S.; Graham, H.; Patton, M.; Jing, Y.; Guloy, A. M.; Yao, Y. Solvent engineering towards controlled grain growth in perovskite planar heterojunction solar cells. Nanoscale 2015, 7 (24), 10595−10599. (36) Li, W.; Fan, J.; Li, J.; Mai, Y.; Wang, L. Controllable Grain Morphology of Perovskite Absorber Film by Molecular Self-Assembly toward Efficient Solar Cell Exceeding 17%. J. Am. Chem. Soc. 2015, 137 (32), 10399−10405. (37) Ahn, N.; Son, D. Y.; Jang, I. H.; Kang, S. M.; Choi, M.; Park, N. G. Highly Reproducible Perovskite Solar Cells with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct of Lead(II) Iodide. J. Am. Chem. Soc. 2015, 137 (27), 8696− 8699. (38) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.; Seok, S. I. High-performance photovoltaic perovskite layers fabricated through intramolecular exchange. Science 2015, 348 (6240), 1234− 1237. (39) 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 (9), 897−903. (40) Wei, H.; Xiao, J.; Yang, Y.; Lv, S.; Shi, J.; Xu, X.; Dong, J.; Luo, Y.; Li, D.; Meng, Q. Free-standing flexible carbon electrode for highly efficient hole-conductor-free perovskite solar cells. Carbon 2015, 93, 861−868. (41) Cao, K.; Cui, J.; Zhang, H.; Li, H.; Song, J.; Shen, Y.; Cheng, Y.; Wang, M. Efficient mesoscopic perovskite solar cells based on the 30191

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192

Research Article

ACS Applied Materials & Interfaces CH3NH3PbI2Br light absorber. J. Mater. Chem. A 2015, 3 (17), 9116−9122. (42) Zhang, L.; Liu, T.; Liu, L.; Hu, M.; Yang, Y.; Mei, A.; Han, H. The effect of carbon counter electrodes on fully printable mesoscopic perovskite solar cells. J. Mater. Chem. A 2015, 3 (17), 9165−9170. (43) Wang, F.; Endo, M.; Mouri, S.; Miyauchi, Y.; Ohno, Y.; Wakamiya, A.; Murata, Y.; Matsuda, K. Highly stable perovskite solar cells with an all-carbon hole transport layer. Nanoscale 2016, 8 (23), 11882−11888. (44) Chen, H.; Wei, Z.; He, H.; Zheng, X.; Wong, K. S.; Yang, S. Solvent Engineering Boosts the Efficiency of Paintable Carbon-Based Perovskite Solar Cells to Beyond 14%. Adv. Energy Mater. 2016, 6 (8), 1502087. (45) Zhou, H.; Shi, Y.; Dong, Q.; Zhang, H.; Xing, Y.; Wang, K.; Du, Y.; Ma, T. Hole-Conductor-Free, Metal-Electrode-Free TiO2/ CH3NH3PbI3 Heterojunction Solar Cells Based on a Low-Temperature Carbon Electrode. J. Phys. Chem. Lett. 2014, 5 (18), 3241−3246. (46) Wu, Y.; Islam, A.; Yang, X.; Qin, C.; Liu, J.; Zhang, K.; Peng, W.; Han, L. Retarding the crystallization of PbI2 for highly reproducible planar-structured perovskite solar cells via sequential deposition. Energy Environ. Sci. 2014, 7 (9), 2934−2938. (47) Im, J. H.; Jang, I. H.; Pellet, N.; Gratzel, M.; Park, N. G. Growth of CH3NH3PbI3 cuboids with controlled size for high-efficiency perovskite solar cells. Nat. Nanotechnol. 2014, 9 (11), 927−932. (48) Chan, C.-Y.; Wang, Y.; Wu, G.-W.; Diau, E. W.-G. Solventextraction Crystal Growth for Highly Efficient Carbon-based Mesoscopic Perovskite Solar Cells free of Hole Conductors. J. Mater. Chem. A 2016, 4, 3872.

30192

DOI: 10.1021/acsami.6b09925 ACS Appl. Mater. Interfaces 2016, 8, 30184−30192