A gradient engineered hole-transporting material for monolithic series-type large-area perovskite solar cells

Yongguang Tu a, Jihuai Wu *a, Xin He a, Panfeng Guo a, Tongyue Wu a, Hui Luo a, Quanzhen Liu a, Qihui Wu b, Jianming Lin a, Miaoliang Huang a, Zhang Lan a and Sizhong Li a
aEngineering Research Center of Environment-Friendly Functional Materials for Ministry of Education, Key Laboratory of Functional Materials for Fujian Higher Education, College of Material Science and Engineering, Huaqiao University, Xiamen 361021, Fujian, People's Republic of China. E-mail: jhwu@hqu.edu.cn; Fax: +86-595-22692229
bDepartment of Materials Chemistry, College of Chemical Engineering and Materials Science, Quanzhou Normal University, Quanzhou 362000, Fujian, People's Republic of China

Received 5th May 2017 , Accepted 18th July 2017

First published on 18th July 2017


Further efficiency enhancement mainly relies on decreasing the interface losses between the active layers in perovskite solar cells. The design of a gradient engineered hole-transporting material is expected to tune the interface losses in perovskite solar cells. In this work, we reported gradient engineering that afforded the hole-transport material (spiro-OMeTAD) dispersed in the upper part of the perovskite layer. Photoluminescence measurements indicated an enhanced hole extraction from the perovskite–spiro-OMeTAD gradient film. And a maximum PCE of 19.16% and a steady-state efficiency of 18.01% were obtained for the small-area device. Furthermore, we assembled monolithic series-type large-area perovskite solar cells based on gradient engineering. The large-area perovskite solar cell with an active area of 1.01 cm2 obtained a PCE of 16.61%. Moreover, monolithic series-type large-area perovskite solar cells showed a Voc of 2.095 V for the binary module and a Voc of 3.104 V for the ternary module, respectively.


Introduction

Perovskite solar cells (PSCs) have aroused burgeoning interest for causing a transformative surge in the next-generation photovoltaic field1–5 in view of their many merits such as broad and strong light absorption,6 longer carrier lifetimes,7 and low exciton binding energy,8 as well as exceptional processability.9 The power conversion efficiency (PCE) of PSCs has been boosted from 3.8% (ref. 1) in 2009 to exceeding 22% now (http://www.nrel.gov/ncpv/images/efficiency_chart.jpg), which would lead to substantial advances in energy conversion technology.

To achieve high performance, a lot of creative technologies and methodologies have been developed, in particular, solution-chemistry methodologies including solvent engineering with anti-solvents,10,11 interfacial engineering,12–17 bandgap engineering,18–22 heterojunction engineering,23etc. Among these methods, heterojunction engineering was recently adopted in perovskite solar cells. For typical heterojunction engineering, a perovskite–fullerene graded heterojunction structure was prepared by the formation of a gradient distribution of n-type electron-collection materials (PCBMs) in a perovskite layer, and the device showed a significantly improved photoelectron collection and reduced recombination loss in its inverted structure.23 The idea may stem from the functional gradient design. Functional gradient materials (FGMs) were first designed to solve the problem of material instability in the combustion chamber of space shuttle engines in the last century. People always focus on homogeneous perovskite materials, whose properties are evenly distributed without any change at the macroscopic level. For non-homogeneous materials, the performance may also continuously vary as the composition and structure of materials change continuously in the geometric space. As for a normal-structured device assembled with mesoporous TiO2, 2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD)24 and poly[bis(4-phenyl)-(2,4,6-trimethylphenyl)amine] (PTAA)10 are well-known hole transport materials (HTMs) for efficient perovskite solar cells. The ideal HTMs typically require a compatible energy level and sufficient charge carrier extraction and transferability.25–29 In general, a solution processed HTM layer was fabricated on top of the prepared perovskite film, possessing a clear interface between the perovskite layer and HTM layer. Nevertheless, to the best of our knowledge, there is no report on the functional gradient structure design of hole-transporting materials. Meanwhile, the unique properties of the functional gradient design provide the possibility to tune the interface losses between the active layers.

In this work, we reported gradient engineering that afforded the hole-transporting material (spiro-OMeTAD) dispersed in the upper part of the perovskite layer with a gradient prepared by the antisolution-dripping method for perovskite solar cells. The gradient engineered hole-transporting material provided an efficient pathway for hole extraction confirmed by PL measurements. The perovskite–spiro-OMeTAD gradient film device exhibited a maximum PCE of 19.16% and a steady-state efficiency of 18.01%. Furthermore, we assembled monolithic series-type large-area perovskite solar cells based on gradient engineering. The large-area perovskite solar cell with an active area of 1.01 cm2 obtained a PCE of 16.61%. The monolithic binary module device with an active area of 1.40 cm2 obtained a PCE of 12.39% with a Voc of 2.095 V. The monolithic ternary module device with an active area of 1.26 cm2 obtained a PCE of 12.80% with a Voc of 3.104 V.

Results and discussion

Fig. 1 schematically illustrates the fabrication process that we have studied. A cesium-containing triple cation perovskite precursor was first dispensed onto the mesoporous TiO2 film. In the first-step program, removing the excess precursor solution was a major process.11 In the second-step program, most DMF could be extracted by rapid evaporation of the anti-solution and then the intermediate would be crystallized by supersaturation from which a smooth and uniform film was formed.30,31 To form the perovskite–spiro-OMeTAD gradient film, 0 μL (as pristine group), 2.5 μL, 5 μL, 10 μL, 20 μL and 40 μL of 60 mM spiro-OMeTAD stock solutions without dopants were pre-dissolved in 1 mL anti-solvent of chlorobenzene (the spiro-OMeTAD content as listed in Table S1 in the ESI). During the process of antisolution-dripping, due to the volatilization of chlorobenzene, the small molecule spiro-OMeTAD would penetrate into the wet precursor film, resulting in a gradient distribution in the vertical direction. The inset in Fig. 1 shows spiro-OMeTAD molecules dispersed in the intermediate layer with a certain gradient during the quick washing process.
image file: c7ta03890a-f1.tif
Fig. 1 Schematics of the process of fabricating perovskite films.

To understand the growth process of perovskite films using the anti-solution (spiro-OMeTAD in chlorobenzene) instead of the pure anti-solvent (chlorobenzene), we studied the surface morphology of perovskite films versus the concentration by SEM. As shown in Fig. 2, the grain sizes were greatly influenced by the concentration of the anti-solution. As we all know, antisolvent-dripping is an effective method to improve the nucleation process by triggering homogeneous nucleation at the surface of the formed layer (solid/antisolvent/air interface).32 The anti-solution (spiro-OMeTAD in chlorobenzene) may cause heterogeneous nucleation because of the solute (spiro-OMeTAD) remaining in the upper part of the formed film along with the quick volatilization of chlorobenzene. Due to the decrease of the nucleation free energy barrier, heterogeneous nucleation is considered to be several orders of magnitude faster than homogeneous nucleation.33 In Fig. 2a, the pristine group showed a compact and pinhole-free layer, the grains with sizes of ca. 260 nm were closely packed, as reported in other literature studies.19,34,35 When a smaller concentration (2.5 μL mL−1) of anti-solution was used, the morphology of the perovskite film changed less with similar grain sizes. The average grain size was determined to be 260 nm for 0 μL mL−1, 280 nm for 2.5 μL mL−1, 290 nm for 5 μL mL−1, 325 nm for 10 μL mL−1, 330 nm for 20 μL mL−1 and 225 nm for 40 μL mL−1, respectively (as shown in Fig. S1). When the concentration was increased from 5 μL mL−1 to 20 μL mL−1, it presented an enlarged average grain size. Cross-sectional SEM images of the resulting perovskite films in Fig. 3 also verified the difference in the morphology. The pristine perovskite film contained a plurality of grains with small sizes, but the grain size prepared by the antisolution-dripping method increased as the concentration increased from 0 μL mL−1 to 20 μL mL−1 and the grain size was commensurate with its thickness among samples of 10 μL mL−1 and 20 μL mL−1, which would minimize the grain boundary energy and be beneficial for the charge transmission.31 However, there were some incomplete grains in the film based on 20 μL mL−1. Upon increasing the anti-solution concentration to 40 μL mL−1, there was a stark contrast versus other groups, possessing decreased grain sizes and thinner thicknesses. The higher concentration of anti-solution would lead to a high spiro-OMeTAD content covering the surface of the wet precursor film. During the annealing process (100 °C for 40 min), the annealing treatment induced high crystallization of spiro-OMeTAD.36,37 There is a competition between the perovskite and spiro-OMeTAD in the crystallization behaviour. For a sample of 40 μL mL−1, a large amount of spiro-OMeTAD remaining in the wet precursor film may inhibit the growth of perovskite, resulting in a defective film. The perovskite films also affect the light absorption properties. The UV-vis absorption spectra of perovskite films are shown in Fig. S2. The resulting perovskite films had different absorption values in the short wave region. Accordingly, stronger absorption was observed for the perovskite films based on the low concentration of the anti-solution, most likely ascribed to the uniform and compact surface coverage.


image file: c7ta03890a-f2.tif
Fig. 2 Top-view SEM images of perovskite films on top of mesoporous TiO2 with the anti-solution (spiro-OMeTAD solution in chlorobenzene) of 0 μL mL−1, 2.5 μL mL−1, 5 μL mL−1, 10 μL mL−1, 20 μL mL−1 and 40 μL mL−1.

image file: c7ta03890a-f3.tif
Fig. 3 Cross-sectional SEM images of perovskite films fabricated with anti-solutions of 0 μL mL−1, 2.5 μL mL−1, 5 μL mL−1, 10 μL mL−1, 20 μL mL−1 and 40 μL mL−1.

To further study the crystal structure, we conducted XRD measurements for the resulting perovskite films, as shown in Fig. 4. The peak at 12.7° can be indexed to the (001) lattice plane of crystallized PbI2, which is beneficial for the device performance due to the passivation of surface defects.38 The two main peaks for all perovskite films approximately located at 14.4° and 28.8° can be indexed to the (110) and (220) planes,39–42 confirming the presence of the tetragonal perovskite phase in all films. And the diffraction intensity for the latter gradually strengthened with the anti-solution concentration increased from 0 μL mL−1 to 20 μL mL−1, consistent with the surface morphology.


image file: c7ta03890a-f4.tif
Fig. 4 XRD patterns of perovskite films fabricated with anti-solutions of 0 μL mL−1, 2.5 μL mL−1, 5 μL mL−1, 10 μL mL−1, 20 μL mL−1 and 40 μL mL−1.

In order to explore the distribution of spiro-OMeTAD, we employed energy dispersive X-ray spectroscopy (EDX) measurements and secondary ion mass spectrometry (SIMS) for the sample based on 10 μL mL−1. The C and O element contents in the spiro-OMeTAD molecule were 79.39% and 10.44%, respectively. Considering that the C element in spiro-OMeTAD would be superimposed on the C element in the perovskite and there was no O element in the perovskite material, we selected C and O elements as characteristic elements in EDX measurements. As shown in Fig. S3, the red line in the figure shows the C element, and the blue line shows the O element. The C element decreased gradually alongside the vertical direction of the perovskite film. However, the O element remained stable without an obvious gradient, which may be ascribed to the low content and incomplete volatilization of DMSO during the annealing process. Secondary ion mass spectrometry (SIMS) has a high sensitivity of element detection and high spatial resolution in both the surface and depth directions. As the signals of the spiro-OMeTAD fragments are very weak in mass spectroscopy, we selected the 10 μL mL−1 (0.72 mg mL−1) spiro-OMeTAD solution only doped by bis(trifluoromethylsulfonyl)imide lithium salt (Li-TFSI) as the anti-solution and the Li element (not be detected by EDX) in Li-TFSI as the collection element in the SIMS measurement. As shown in Fig. 5, changes in the spatial distribution of Li were monitored by measuring the depth profile of the perovskite film. With the continuous etching with Ar, the content of the Li element in the perovskite film was gradually reduced. The change in the depth profiles of the C and Li elements associated with the perovskite film was monitored to evaluate the gradient distribution of spiro-OMeTAD.


image file: c7ta03890a-f5.tif
Fig. 5 Secondary ion mass spectroscopy depth profile (Li as the characteristic element) of the perovskite film based on 10 μL mL−1 (0.72 mg mL−1).

The perovskite film not only plays the role of a light absorber, but it is also the medium for charge transport. In order to understand carrier extraction based on the perovskite–spiro-OMeTAD gradient structure, we carried out steady-state photoluminescence (PL) spectroscopy and time-resolved photoluminescence (TRPL) intensity decay measurements of the resulting perovskite films with the configuration of the non-conducting glass/perovskite–spiro-OMeTAD gradient film. Fig. 6 shows the steady state PL spectra. A clear difference is observed in the PL spectra among the perovskite films prepared using the anti-solution. Only the pristine group had an emission peak at 760 nm. Other samples presented significant quenching of PL intensity. As shown in Fig. S4, the magnified PL spectra had no obvious signal of the emission peak. The significant quenching of PL intensity in the perovskite–spiro-OMeTAD gradient film confirmed the injection of holes from the valance band of the perovskite into the HOMO of spiro-OMeTAD.25,43,44Fig. 7 shows the time-resolved photoluminescence intensity decay, and the curves were fitted with a two-component exponential decay function. The sample of 0 μL mL−1 (pristine group) exhibits a fast and slow phase lifetime of τ1 = 6.130 ns (fraction A1 = 55.69%) and τ2 = 54.572 ns (fraction A2 = 44.31%), respectively while the samples of anti-solution show short τ1 and τ2. For example, the sample of 10 μL mL−1 exhibits a fast and slow phase lifetime of τ1 = 4.693 ns (fraction A1 = 57.98%) and τ2 = 33.713 ns (fraction A2 = 42.02%), respectively, as listed in Table S2. From these observations, we conclude that an enhanced hole injection was obtained. In addition, we further carried out steady-state photoluminescence (PL) spectra and time-resolved photoluminescence (TRPL) intensity decay measurement of the resulting perovskite films with the configuration of the non-conducting glass/PEDOT:PSS/perovskite–spiro-OMeTAD gradient film. PEDOT:PSS is a frequently-used hole extraction layer in inverted perovskite solar cells. Significant quenching of PL intensity and a clear difference in TRPL response are presented in Fig. S5 and S6. The pristine perovskite film had the longest slow phase lifetime (τ2 = 29.151 ns) versus experimental groups (as listed in Table S3), indicating that the charge injection of the perovskite–spiro-OMeTAD gradient film to the hole transport layer was more efficient than the reference perovskite layer.45,46 From the above results, we obtained relatively more efficient hole injection of the perovskite–spiro-OMeTAD gradient film, indicating a decreased interface loss, as shown in Fig. 8. The negative circles and positive circles represent electrons and holes, respectively, and the arrows indicate their flow directions.


image file: c7ta03890a-f6.tif
Fig. 6 Steady state photoluminescence spectra of perovskite films on non-conducting glass detected at 460 nm excitation wavelength.

image file: c7ta03890a-f7.tif
Fig. 7 Time-resolved photoluminescence intensity decay spectra of perovskite films on non-conducting glass detected at 760 nm emission wavelength.

image file: c7ta03890a-f8.tif
Fig. 8 Schematic description of charge dynamics and band levels of the perovskite solar cell.

The perovskite solar cells were fabricated from the antisolution-dripping of 0 μL mL−1, 2.5 μL mL−1, 5 μL mL−1, 10 μL mL−1, 20 μL mL−1 and 40 μL mL−1. The photocurrent–voltage (JV) curves of the devices were measured under AM 1.5 G (100 mW cm−2) as shown in Fig. 9 and the corresponding photovoltaic parameters are listed in Table 1. For the pristine device, a PCE of 16.84% was achieved with a short-circuit current density (Jsc) of 22.13 mA cm−2, an open-circuit photovoltage (Voc) of 1.072 V and a fill factor (FF) of 0.71. The perovskite solar cell based on 10 μL mL−1 achieved a highest PCE value of 19.16% with a Jsc of 23.02 mA cm−2, a Voc of 1.095 V and a FF of 0.76. The increased FF and Voc were explained in terms of the lower interface losses, which was confirmed by PL measurements. And with the further increase of the anti-solution concentration, the performance began to decline, which mainly stemmed from defective perovskite films.


image file: c7ta03890a-f9.tif
Fig. 9 JV curves of the perovskite solar cells under AM 1.5 G illumination.
Table 1 The photovoltaic parameters of perovskite solar cells
Samples V oc (V) J sc (mA cm−2) FF PCE (%)
0 μL mL−1 1.072 22.13 0.71 16.84
5 μL mL−1 1.082 22.62 0.77 18.84
10 μL mL−1 1.095 23.02 0.76 19.16
20 μL mL−1 1.083 22.35 0.74 17.91
40 μL mL−1 1.065 19.55 0.63 13.11


It is well known that the photocurrent density–voltage (JV) hysteresis often exists in perovskite solar cells when measured under different scan directions, which makes accurate efficiency determination challenging. Fig. S7 shows the JV curves of the devices based on 0 μL mL−1 and 10 μL mL−1 under both reverse and forward scan directions, and there is still a hysteresis effect. For the forward scan, the pristine device shows a PCE of 11.41%, with a short-circuit current density (Jsc) of 22.13 mA cm−2, an open-circuit photovoltage (Voc) of 0.991 V and a fill factor (FF) of 0.52; the device based on 10 μL mL−1 shows a PCE of 13.42%, with a Jsc of 23.02 mA cm−2, a Voc of 0.971 V and a FF of 0.60.

Fig. S8 shows the steady-state photocurrent and output PCE of the devices based on 10 μL mL−1 at the maximum power points with a stabilized current density output of 19.25 mA cm−2 (at the voltage of 0.935 V), yielding a PCE of 18.01%. Statistical results of the cell performance are provided in Fig. S9 as box charts. It can be seen that the devices based on 10 μL mL−1 show better performance (each team is calculated from a batch of 25 cells). Long-term stability is a critical concern for practical application of perovskite solar cells. As shown in Fig. S10, two kinds of devices based on 0 μL mL−1 and 10 μL mL−1 had a similar stability, and the devices could retain over 70% of their initial performance after 360 hours.

Subsequently, four kinds of large-area devices based on the anti-solution-dripping method (10 μL mL−1) were fabricated on a 2 × 2 cm2 substrate. The first form was a unary cell with an active area of 1.00 cm2; the second form was a unary cell with an active area of 2.00 cm2; the third form was a binary module with a total active area of 1.4 cm2, composed of two serial-connected sub-cells of an aperture area of 0.7 cm2 each; the fourth form was a ternary module with a total active area of 1.26 cm2, provided with an aperture area of 0.42 cm2 each, first cell, second cell and third cell, respectively. The sub-cells were combined by a standard serial interconnection geometry. Fig. 10 shows the schematics of the monolithic series-type module. Fig. 11 shows photos of backsides of the monolithic perovskite cell or module with different areas and forms.


image file: c7ta03890a-f10.tif
Fig. 10 Schematics of the monolithic series-type module.

image file: c7ta03890a-f11.tif
Fig. 11 Photos of backsides of the monolithic perovskite cell or module with different areas and forms. (a) Unary cell with an active area of 1.00 cm2; (b) unary cell with an active area of 2.00 cm2; (c) binary module with a total active area of 1.4 cm2; (d) ternary module with a total active area of 1.26 cm2.

Fig. 12 shows the IV curves of the monolithic perovskite cell or module with different areas and forms. The PCE of unary cell with an active area of 1.01 cm2 was 16.61%, whereas the PCE of unary cell with an active area of 2.00 cm2 was 9.56%. The main contributing factor to decreasing PCE was FF, which caused poor uniformity on a larger dimension by antisolution-dripping. The binary module with an active area of 1.4 cm2 yielded a short circuit current (Isc) of 12.54 mA, a Voc of 2.095 V, and a FF of 0.66, resulting in a PCE of 12.39%. The ternary module with an active area of 1.26 cm2 yielded a short circuit current (Isc) of 7.64 mA, a Voc of 3.104 V, and a FF of 0.68, resulting in a PCE of 12.80%. The corresponding photovoltaic parameters are listed in Tables 2 and 3. This finding demonstrates a novel method for fabricating large-area perovskite solar cells.


image file: c7ta03890a-f12.tif
Fig. 12 IV curves of the monolithic perovskite cell or module with different areas and forms. (a) Unary cell with an active area of 1.01 cm2; (b) unary cell with an active area of 2.00 cm2; (c) binary module with a total active area of 1.4 cm2; (d) ternary module with a total active area of 1.26 cm2.
Table 2 The photovoltaic parameters of the binary module and its two sub-cells
Sample V oc (mV) I sc (mA) FF PCE (%)
Binary module 2095 12.54 0.66 12.39
First cell 1050 13.28 0.66 13.15
Second cell 1069 13.33 0.71 14.45


Table 3 The photovoltaic parameters of the ternary module and its three sub-cells
Sample V oc (mV) I sc (mA) FF PCE (%)
Ternary module 3104 7.64 0.68 12.80
First cell 1060 8.00 0.70 14.13
Second cell 1049 8.04 0.72 14.46
Third cell 1.053 8.02 0.67 13.47


Conclusion

In summary, we reported a perovskite–spiro-OMeTAD gradient structure to enhance the hole extraction efficiency of perovskite films by the antisolution-dripping method. Detailed studies revealed that the gradient engineered hole-transporting material was beneficial for enhancing the PCE due to which it provided an efficient pathway for hole extraction. The perovskite–spiro-OMeTAD gradient film small-area device exhibited a maximum PCE of 19.16% and a steady-state efficiency of 18.01%. Furthermore, we assembled a large-area monolithic series-type perovskite solar module based on gradient engineering. The large-area perovskite solar cell with an active area of 1.01 cm2 obtained a PCE of 16.61%. The monolithic binary module device with an active area of 1.40 cm2 obtained a PCE of 12.39% with a Voc of 2.095 V. The monolithic ternary module device with an active area of 1.26 cm2 obtained a PCE of 12.80% with a Voc of 3.104 V.

Methods

Small-area device fabrication

Preparation of the photoanode. The FTO coated substrates were etched using a laser to form the desired electrode patterns. The FTO patterned substrates were cleaned by UV–ozone treatment for 15 min, followed by cleaning with detergent and ethanol consecutively. Ultrathin blocking TiO2 layers based on highly crystallized TiO2 quantum dots (QDs) are described in our previous reports.47 The contact area was cleaned with toluene to remove the extra compact TiO2 layer. A homemade TiO2 paste was deposited on the compact TiO2 layer and sintered at 500 °C for 30 min, providing a thickness of ca. 150 nm. To remove the extra TiO2 layer from the contact area, the film was cleaned with ethanol before the sintering process. The mesoporous TiO2 film was then immersed in 0.02 M aqueous TiCl4 solution at 80 °C for 20 min. After rinsing with deionized water and ethanol, the film was sintered at 500 °C for 30 min.
Cs/MA/FA perovskite precursor solution and film deposition. The perovskite layer was deposited on the top of the mesoporous TiO2 film by a two-step spin coating program (10 s at 1000 rpm and 20 s at 6000 rpm) with dripping of spiro-OMeTAD solution in chlorobenzene as the anti-solution during the second step, 5 s before the end. In detail, the mixed-cation lead mixed-halide perovskite solution was prepared by dissolving FAI (1 M, Baolaite), PbI2 (1.1 M, TCl), MABr (0.2 M, Baolaite) and PbBr2 (0.22 M, TCl) in a 4[thin space (1/6-em)]:[thin space (1/6-em)]1 (v/v) mixture of anhydrous DMF[thin space (1/6-em)]:[thin space (1/6-em)]DMSO (Sigma-Aldrich). Then 1.5 M stock solution of CsI (TCI) in DMSO was added to the above solution in the 5[thin space (1/6-em)]:[thin space (1/6-em)]95 volume ratio. To prepare the gradient perovskite layer, 0 μL (as pristine group), 2.5 μL, 5 μL, 10 μL, 20 μL, 40 μL of 60 mM spiro-OMeTAD stock solutions without dopants were pre-dissolved in 1 mL anti-solvent of chlorobenzene. 20 μL perovskite precursor solution was loaded on the top of the TiO2 film and 100 μL anti-solution was dripped onto the rotating substrate. All the perovskite layers were annealed at 100 °C for 40 min.
Hole transporting layer and Au electrode deposition. 20 μL of 60 mM 2,2′,7,7′-tetrakis (N,N-di-p-methoxyphenylamine)-9,9-spirobifluorene (spiro-OMeTAD, luminescence) solution was spin-coated on the perovskite layer at 4000 rpm for 30 s. A standard spiro-OMeTAD solution was prepared by dissolving 72.3 mg of spiro-OMeTAD in 1 mL of chlorobenzene, to which 28.8 μL of 4-tert-butyl pyridine and 17.5 μL of lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) solution (520 mg LI-TSFI in 1 mL acetonitrile, Sigma-Aldrich, 99.8%) were added. All devices were stored in a desiccator (humidity < 15%) in the dark for 12 h. Before the Au electrode contact was deposited, the perovskite/spiro-OMeTAD layer was removed using a surgical blade and a methanol–DMF blend (9[thin space (1/6-em)]:[thin space (1/6-em)]1 volume ratio) in order to ensure a clean blank electrode. Finally, 80 nm of gold was deposited under vacuum through a shadow mask. The small-area devices had an active area of 0.125 cm2.

Large-area module fabrication

For the large-area unary perovskite solar cell, 40 μL perovskite precursor solution was loaded on the top of the TiO2 film (the substrate size was 2 × 2 cm2) and 200 μL anti-solution was dripped onto the rotating substrate. The large-area unary devices had two types of area. One type was adjusted to be 1.00 cm2 and the other type was adjusted to be 2.00 cm2. All the layers and the contact areas were prepared according to the same process as for the fabrication of the small-area device.

For the large-area binary perovskite solar cell module, the patterned FTO-glass substrate (2 × 2 cm2) has two FTO strips interconnected in series. The active area of one cell was adjusted to be 0.5 × 1.4 cm2. All the layers and the contact areas were prepared according to the same process as for the fabrication of the small-area device.

For the large-area ternary perovskite solar cell module, the patterned FTO-glass substrate (2 cm × 2 cm) has three FTO strips interconnected in series. The active area of one cell is adjusted to be 0.3 × 1.4 cm2. All the layers and the contact areas are prepared according to the same process as for the fabrication of the small-area device.

Characterization

The current density–voltage (JV) curves were measured using a Keithley 2420 source-measure unit under AM 1.5 G illumination at 100 mW cm−2 provided with an Oriel Sol 3A solar simulator in an ambient environment. The light intensity was adjusted using a NREL-calibrated Si solar cell equipped with a KG-2 filter. The pre-sweep delay time was 40 ms, the dwell time at each voltage step was 30 ms, and 100 data points were measured between −0.1 and 2.0 V. Surface morphologies were characterized by field-emission scanning electron microscopy (SEM, Hitachi S-8000, Japan). The crystalline structures were examined using X-ray diffraction (XRD, Bruker AXS, D8 Advance). The steady state fluorescence spectrum was acquired using a fluorescence spectrophotometer (Thermo Scientific Lumina). The time-resolved photoluminescence spectrum was acquired using an Omin-λ Monochromator/Spectrograph. The secondary ion mass spectrometry (SIMS) was performed with a PHI 5000VersaProbe III.

Author contributions

J. W. conceived the project; J. W. and Y. T. designed the device and experiments; and Y. T. carried out most of the experiments. Y. T. wrote the first draft of the manuscript, and J. W. finished the submitted paper. All the authors discussed the results and approved publishing this paper.

Acknowledgements

The authors acknowledge the financial joint support by the National Natural Science Foundation of China (No. 91422301 and 51472094) and the Cultivation Program for Postgraduate in Scientific Research Innovation Ability of Huaqiao University (No. 1400102002).

Notes and references

  1. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051 CrossRef CAS PubMed.
  2. J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316–319 CrossRef CAS PubMed.
  3. N. J. Jeon, J. H. Noh, W. S. Yang, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Nature, 2015, 517, 476–480 CrossRef CAS PubMed.
  4. A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang, M. Gratzel and H. Han, Science, 2014, 345, 295–298 CrossRef CAS PubMed.
  5. Q. Tang, Q. W. Zhu, B. He and P. Yang, ACS Nano, 2017, 11, 1540–1547 CrossRef CAS PubMed.
  6. Z. Xiao, C. Bi, Y. Shao, Q. Dong, Q. Wang, Y. Yuan, C. Wang, Y. Gao and J. Huang, Energy Environ. Sci., 2014, 7, 2619–2623 CAS.
  7. C. Wehrenfennig, G. E. Eperon, M. B. Johnston, H. J. Snaith and L. M. Herz, Adv. Mater., 2014, 26, 1584–1589 CrossRef CAS PubMed.
  8. W. Zhang, M. Saliba, S. D. Stranks, Y. Sun, X. Shi, U. Wiesner and H. J. Snaith, Nano Lett., 2013, 13, 4505–4510 CrossRef CAS PubMed.
  9. Y. Wu, A. Islam, X. Yang, C. Qin, J. Liu, K. Zhang, W. Peng and L. Han, Energy Environ. Sci., 2014, 7, 2934–2938 CAS.
  10. N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu and S. I. Seok, Nat. Mater., 2014, 13, 897–903 CrossRef CAS PubMed.
  11. M. Xiao, F. Huang, W. Huang, Y. Dkhissi, Y. Zhu, J. Etheridge, A. Gray-Weale, U. Bach, Y. B. Cheng and L. Spiccia, Angew. Chem., Int. Ed. Engl., 2014, 53, 9898–9903 CrossRef CAS PubMed.
  12. Z. Gu, L. Zuo, T. T. Larsen-Olsen, T. Ye, G. Wu, F. C. Krebs and H. Chen, J. Mater. Chem. A, 2015, 3, 24254–24260 CAS.
  13. M. Ma, Q. Tang, H. Chen, B. He and P. Yang, Sol. Energy Mater. Sol. Cells, 2017, 160, 67–76 CrossRef CAS.
  14. B. Li, Y. Chen, Z. Liang, D. Gao and W. Huang, RSC Adv., 2015, 5, 94290–94295 RSC.
  15. X. Shi, Y. Hu, B. Wang, L. Zhang, Z. Wang and L. Liao, Adv. Mater., 2015, 27(42), 6696–6701 CrossRef CAS PubMed.
  16. Y. Li, Y. Zhao, Q. Chen, Y. Yang, Y. Liu, Z. Hong, Z. Liu, Y.-T. Hsieh, L. Meng, Y. Li and Y. Yang, J. Am. Chem. Soc., 2015, 137, 15540–15547 CrossRef CAS PubMed.
  17. H. Zhou, Q. Chen, G. Li, S. Luo, T. B. Song, H. S. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542–546 CrossRef CAS PubMed.
  18. H. Y. Ye, W. Q. Liao, C. L. Hu, Y. Zhang, Y. M. You, J. G. Mao, P. F. Li and R. G. Xiong, Adv. Mater., 2016, 28, 2579–2586 CrossRef CAS PubMed.
  19. J. P. Correa Baena, L. Steier, W. Tress, M. Saliba, S. Neutzner, T. Matsui, F. Giordano, J. Jacobsson, A. R. Srimath Kandada, S. M. Zakeeruddin, A. petrozza, A. Abate, N. K. Mohammad, M. Grätzel and A. Hagfeldt, Energy Environ. Sci., 2015, 8, 2928–2934 CAS.
  20. T. Todorov, T. Gershon, O. Gunawan, Y. S. Lee, C. Sturdevant, L.-Y. Chang and S. Guha, Adv. Energy Mater., 2015, 1500799 CrossRef.
  21. M. R. Filip, G. E. Eperon, H. J. Snaith and F. Giustino, Nat. Commun., 2014, 5, 5757 CrossRef CAS PubMed.
  22. L. D. Zhao, J. He, S. Hao, C. I. Wu, T. P. Hogan, C. Wolverton, V. P. Dravid and M. G. Kanatzidis, J. Am. Chem. Soc., 2012, 134, 16327–16336 CrossRef CAS PubMed.
  23. Y. Wu, X. Yang, W. Chen, Y. Yue, M. Cai, F. Xie, E. Bi, A. Islam and L. Han, Nat. Energy, 2016, 1, 16148 CrossRef CAS.
  24. S. Wang, W. Yuan and Y. S. Meng, ACS Appl. Mater. Interfaces, 2015, 4, 24791–24798 Search PubMed.
  25. K. Rakstys, A. Abate, M. I. Dar, P. Gao, V. Jankauskas, G. Jacopin, E. Kamarauskas, S. Kazim, S. Ahmad, M. Gratzel and M. K. Nazeeruddin, J. Am. Chem. Soc., 2015, 137, 16172–16178 CrossRef CAS PubMed.
  26. Q. Tang, Y. Duan, B. He and H. Chen, Angew. Chem., 2016, 128, 14624–14628 CrossRef.
  27. N. J. Jeon, H. G. Lee, Y. C. Kim, J. Seo, J. H. Noh, J. Lee and S. I. Seok, J. Am. Chem. Soc., 2014, 136, 7837–7840 CrossRef CAS PubMed.
  28. M. Li, Z. Wang, Y. Yang, Y. Hu, S. Feng, J. Wang, X. Gao and L. Liao, Adv. Energy Mater., 2016, 1601156 CrossRef.
  29. Z. Wang, M. Li, Y. Yang, Y. Hu, H. Ma, X. Gao and L. Liao, Adv. Mater., 2016, 28, 6695–6703 CrossRef CAS PubMed.
  30. N. Sakai, S. Pathak, H.-W. Chen, A. A. Haghighirad, S. D. Stranks, T. Miyasaka and H. J. Snaith, J. Mater. Chem. A, 2016, 4, 4464–4471 CAS.
  31. Y. Tu, J. Wu, X. He, P. Guo, H. Luo, Q. Liu, J. Lin, M. Huang, Y. Huang, L. Fan and Z. Lan, Appl. Surf. Sci., 2017, 403, 572–577 CrossRef CAS.
  32. D. Bi, C. Yi, J. Luo, J.-D. Décoppet, F. Zhang, S. M. Zakeeruddin, X. Li, A. Hagfeldt and M. Grätzel, Nat. Energy, 2016, 1, 16142 CrossRef CAS.
  33. A. Cacciuto, S. Auer and D. Frenkel, Nature, 2004, 428, 404 CrossRef CAS PubMed.
  34. M. Saliba, T. Matsui, J.-Y. Seo, K. Domanski, J.-P. Correa-Baena, N. K. Mohammad, S. M. Zakeeruddin, W. Tress, A. Abate, A. Hagfeldt and M. Grätzel, Energy Environ. Sci., 2016, 9, 1989–1997 CAS.
  35. E. H. Anaraki, A. Kermanpur, L. Steier, K. Domanski, T. Matsui, W. Tress, M. Saliba, A. Abate, M. Grätzel, A. Hagfeldt and J.-P. Correa-Baena, Energy Environ. Sci., 2016, 9, 3128–3134 CAS.
  36. T. Leijtens, I. Ding, T. Giovenzana, J. T. Bloking, M. D. Mcgehee and A. Sellinger, ACS Nano, 2014, 6, 1455 CrossRef PubMed.
  37. Y. Fang, X. Wang, Q. Wang, J. Huang and T. Wu, Phys. Status Solidi A, 2014, 211, 2809–2816 CrossRef CAS.
  38. Y. C. Kim, N. J. Jeon, J. H. Noh, W. S. Yang, J. Seo, J. S. Yun, A. Ho-Baillie, S. Huang, M. A. Green, J. Seidel, T. K. Ahn and S. I. Seok, Adv. Energy Mater., 2015, 1502104 Search PubMed.
  39. B. Ding, L. Gao, L. Liang, Q. Chu, X. Song, Y. Li, G. Yang, B. Fan, M. Wang, C. Li and C. Li, ACS Appl. Mater. Interfaces, 2016, 8, 20067–20073 CAS.
  40. E. Zheng, X. F. Wang, J. Song, L. Yan, W. Tian and T. Miyasaka, ACS Appl. Mater. Interfaces, 2015, 7, 18156–18162 CAS.
  41. J. He and T. Chen, J. Mater. Chem. A, 2015, 3, 18514–18520 CAS.
  42. J. W. Lee, D. J. Seol, A. N. Cho and N. G. Park, Adv. Mater., 2014, 26, 4991–4998 CrossRef CAS PubMed.
  43. G. Xing, N. Mathews, S. Sun, S. S. Lim, Y. M. Lam, M. Grätzel, S. Mhaisalkar and T. C. Sum, Science, 2013, 342, 344–347 CrossRef CAS PubMed.
  44. C. Zuo and L. Ding, Small, 2015, 11, 5528–5532 CrossRef CAS PubMed.
  45. M. C. Kim, B. J. Kim, D. Y. Son, N. G. Park, H. S. Jung and M. Choi, Nano Lett., 2016, 16, 5756–5763 CrossRef CAS PubMed.
  46. S. H. Im, J.-H. Heo, H. J. Han, D. Kim and T. Ahn, Energy Environ. Sci., 2015, 8, 1602–1608 Search PubMed.
  47. Y. Tu, J. Wu, M. Zheng, J. Huo, P. Zhou, Z. Lan, J. Lin and M. Huang, Nanoscale, 2015, 7, 20539–20546 RSC.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ta03890a

This journal is © The Royal Society of Chemistry 2017