Boosting efficiency and stability of perovskite solar

G Model JMST-1213; No. of Pages 7

ARTICLE IN PRESS Journal of Materials Science & Technology xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Journal of Materials Science & Technology journal homepage: www.jmst.org

Boosting efficiency and stability of perovskite solar cells with nickel phthalocyanine as a low-cost hole transporting layer material Mustafa Haider a,b,1 , Chao Zhen a,1 , Tingting Wu a,c , Gang Liu a,c,∗ , Hui-Ming Cheng a,d,e a

Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang, 110016, China University of Chinese Academy of Sciences, 19(A) Yuquan Road, Beijing, 100049, China School of Materials Science and Engineering, University of Science and Technology of China, 72 Wenhua Road, Shenyang, 110016, China d Low-Dimensional Material and Device Laboratory, Tsinghua-Berkeley Shenzhen Institute, Tsinghua University, 1001 Xueyuan Road, Shenzhen, 518055, China e Center of Excellence in Environmental Studies, King Abdulaziz University, Jeddah, 21589, Saudi Arabia b c

a r t i c l e

i n f o

Article history: Received 4 February 2018 Received in revised form 2 March 2018 Accepted 3 March 2018 Available online xxx Keywords: Solar cells Perovskite Hole transfer material

a b s t r a c t The efficiency of perovskite solar cells (PSCs) has increased from around 4% to over 22% following a few years of intensive investigation. For most PSCs, organic materials such as 2,2 ,7,7 -tetrakis(N,Npdimethoxyphenylamino)-9,9 -spirobifluorene (spiro-OMeTAD) are used as the hole transporting materials (HTMs), which are thermally and chemically unstable and also expensive. Here, we explored nickel phthalocyanine (NiPc) as a stable and cost-effective HTM to replace the conventionally used spiroOMeTAD. Because of its high carrier mobility and proper band alignments, we achieved a PCE of 12.1% on NiPc based planar device with short-circuit current density (Jsc ) of 17.64 mA cm−2 , open circuit voltage (Voc ) of 0.94 V, and fill factor (FF) of 73%, outperforming the planar device based on copper phthalocyanine (CuPc) that is an outstanding representative of metal phthalocyanines (MPcs) reported. Moreover, the device with NiPc shows much improved stability compared to that based on the conventional spiroOMeTAD as a result of NiPc’s high stability. Photoluminescence (PL) and Impedance spectroscopy analysis results show that thermally deposited NiPc has good hole-extraction ability. Our results suggest that NiPc is a promising HTM for the large area, low cost and stable PSCs. © 2018 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & Technology.

1. Introduction Metal halide perovskites have emerged as promising photovoltaic materials due to the superb optoelectronic properties possessed in this family of materials, such as appropriate bandgap for capturing solar light, high absorption coefficient, long carrier diffusion lengths [1–4]. Photovoltaic cells based on these materials have been denoted as perovskite solar cells (PSCs), delivering skyrocketed efficiency and holding the promise of accessible scalability with low-cost solution processability [5–12]. For most PSCs, the perovskite active layer is sandwiched between an electron transport material (ETM) and a hole transport material (HTM), which perform tasks of collecting photogener-

∗ Corresponding author at: Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang, 110016, China. E-mail address: [email protected] (G. Liu). 1 These authors equally contributed to this work.

ated electrons (blocking holes) and holes (blocking electrons), respectively. The HTM materials commonly used in PSCs are p-type organic semiconductors such as poly(3,4-ethylenedioxythiophene) polystyrene sulphonate (PEDOT:PSS) [13,14], 2,2 ,7,7 -tetrakis(N,Npdimethoxyphenylamino)-9,9 -spirobifluorene (spiro-OMeTAD) [15,16], poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (PTAA) [17,18] and poly(3-hexylthiophene-2,5-diyl) (P3HT) [19]. Although devices based on conventional organic hole transport materials (HTMs) such as spiro-OMeTAD or PTTA show high-efficiency, their high cost and thermal/chemical instability constrain the commercialization of this emerging PSC technology. Moreover, thermo- and moisture-labile metal halide perovskites need a stable HTM layer performing as a protection shell on the top of the cells. Consequently, it is imperative to explore new suitable HTMs with desired features of proper band edge positions, high charge carrier mobility, low cost and thermo- and moisture-stability. To date, a sequence of inorganic HTMs (e.g. NiO, Cu2 O, V2 O5 , MoO3 CuGaO2 etc.) has been reported to address the limiting issues of high-cost and thermal/chemical instability of PSCs [20–26].

https://doi.org/10.1016/j.jmst.2018.03.005 1005-0302/© 2018 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & Technology.

Please cite this article in press as: M. Haider, et al., J. Mater. Sci. Technol. (2018), https://doi.org/10.1016/j.jmst.2018.03.005

G Model JMST-1213; No. of Pages 7 2

ARTICLE IN PRESS M. Haider et al. / Journal of Materials Science & Technology xxx (2018) xxx–xxx

Metal phthalocyanines (MPcs) as an attractive organic semiconductor have been frequently used as HTMs in organic electronics such as organic field effect transistors and diodes, and also as absorber layer in dye-sensitized solar cells (DSSCs) [27,28]. The highest occupied molecular orbitals (HOMOs) and lowest unoccupied molecular orbitals (LUMOs) of MPcs are well matched with the band alignment of perovskite absorbers. Pristine MPcs, which are little soluble in most solvents, could be deposited on substrates by vacuum thermal evaporation method due to their thermal and chemical stability [29]. The room temperature vacuum process is a promising option for the HTM deposition in n-i-p type PSCs because it avoids damaging the underlying perovskite materials and also produces good surface coverage and morphology of HTMs. Some researchers also suggest that photovoltage of PSCs mainly depends on the morphology of perovskite and selective contact materials [30]. In 2014, copper phthalocyanine (CuPc) was explored as an alternative HTM in PSCs by Kumar and coworkers, and the maximum power conversion efficiency (PCE) of 5.0% was achieved [31]. Ke et al. obtained the maximum efficiency of 15.4% by use of CuPc as the HTM in the fully-vacuum-processed PSCs [32]. Tuning the morphology of thermally evaporated CuPc in the form of nanorods further improves the device efficiency up to 16.1% [33]. Besides CuPc, other MPcs (e.g. ZnPc and PbPc) have also been explored as HTMs in PSCs but show relatively low hole-collection efficiency compared to CuPc. Meanwhile, functional group substituted MPcs that become soluble in solvents have also been used as HTMs in all-solution-processed PSCs for reducing production costs. Cu(II)–methyl phthalocyanine and Zn(II) octa(2,6-diphenylphenoxy) phthalocyanine have been adopted as HTMs in PSCs, and efficiencies of 5.2% and 6.7% were obtained, respectively, with CH3 NH3 PbI3 (MAPbI3 ) as the absorbing layer [34,35]. By incorporating with mixed-ion perovskite light harvester such as [FAPbI3 ]0.85 [MAPbBr3 ]0.15 , the devices based on substituted MPcs have shown efficiencies of 17.5% for tetra-5hexylthiophene-based ZnPc, 9.9% for nickel (II) 1, 4, 8, 11, 15, 18, 22, 25-octabutoxy-29H, 31H-phthalocyanine (NiPc-(OBu)8) and 18.8% for copper (II) 2, 9, 16, 23-tetra-tert-butyl-29H, 31Hphthalocyanine (CuPc under AM1.5G illumination [36–38]. Above mentioned studies indicate that the MPcs can be used as efficient HTMs in PSCs, and vacuum thermal evaporation is better than the solution process method for high quality MPc HTM film deposition. Here, we introduced commercially available pristine Nickel phthalocyanine (NiPc) as a new efficient HTM in PSCs. Compared with the other reported MPcs (7.6 × 10−5 cm2 V−1 s−1 for ZnPc [39], 3.0 × 10−4 cm2 V−1 s−1 for CoPc [40], 5.0 × 10−3 cm2 V−1 s−1 for CuPc [41]), NiPc has a relatively high mobility of 1.0 × 10−1 cm2 V−1 s−1 [42], which is even higher than that of undoped spiro-OMeTAD (4.0 × 10−5 cm2 V−1 s−1 ) [43]. Moreover, NiPc has a much better thermal, chemical stability and lower cost than the spiro-OMeTAD. Due to these superior properties of NiPc, we made planar structured PSCs with vacuum thermally deposited NiPc as the HTM, TiO2 compact layer as the ETM and gold as the counter electrode. A PCE of 12.1% has been demonstrated on NiPc based device with short-circuit current density (Jsc ) of 17.64 mA cm−2 , open circuit voltage (Voc ) of 0.94 V, and fill factor (FF) of 73%, outperforming the device based on CuPc that is an outstanding representative of reported MPcs. The big size device (0.5 cm2 ) still shows a PCE of 6.2% with Jsc of 13.34 mA cm−2 , Voc of 0.88 V, and FF of 53.1%. Moreover, the device with NiPc shows much improved stability compared to that based on the conventional spiro-OMeTAD as a result of NiPc’s high stability. The efficiency of device based on NiPc remains over 80% of the initial value after 38 day storage in air without encapsulation, while the efficiency of device based on commercial spiro-OMeTAD decay to almost half of the initial value only after 12 day storage. Our

studies suggest that NiPc is a promising HTM for the large area, low cost and stable PSCs. 2. Experimental 2.1. Materials and methods All the chemicals were directly used without any further purification. Chemicals includes PbCl2 (99.999%, Aladdin), HI (57 wt% in water, Sigma-Aldrich), CH3 NH2 (33 wt. in absolute ethanol, Sigma-Aldrich), titanium isopropoxide (99.995%, Sigma-Aldrich), dimethylformamide/DMF (Sigma-Aldrich), diethyl ether, acetone, isopropanol, ethanol and detergent. 2.2. Synthesis of CH3 NH3 I (MAI) The synthesis of MAI was conducted by reacting of 24 ml CH3 NH2 and 10 ml of HI in a round-bottom flask at 0 ◦ C with stirring for 2 h. The precipitates were obtained when carefully remove the solvents at 50 ◦ C by using a rotary evaporator. The precipitates were re-dissolved in absolute ethanol and again made the precipitates by adding diethyl ether in order to obtain clear white powder. This process was repeated three times. The product MAI was collected by drying at 60 ◦ C in a vacuum oven for 24 h. 2.3. Thin film and solar cell fabrication FTO glass was sequentially cleaned in detergent mixed water, in ethanol after rinsing with deionized water and in acetone and isopropanol, each for 15 min under ultrasonic condition. Prior to the deposition of a TiO2 compact layer, the FTO glass was treated with plasma oxidation for 5 min to remove organic contaminants on surface. The compact layer was deposited on the FTO substrate by spin coating method. The precursor solution used for coating was prepared by dissolving 0.9 ml of titanium isopropoxide in 15 ml of ethanol with 75 ␮l of HCl (37 wt%) as additive that was stirred overnight. The coating was conducted by spinning at 5000 rpm for 30 s. The coated FTO glass was annealed at 500 ◦ C for 30 min. For the perovskite layer, PbCl2 and CH3 NH3 I with the respective concentration of 0.88 M and 2.64 M in Dimethylformamide (DMF) were mixed stirred at room temperature overnight. Spin coating of the perovskite layer on the FTO/TiO2 substrate without pre-heating was conducted at 3000 rpm for 30 s. After the coating, the substrate was annealed at 100 ◦ C for 1.5 h. NiPc as HTM was thermally deposited by vacuum evaporation using quartz crystal monitor to determine deposition rate and thickness of the film on the top of perovskite layer. Gold film with a thickness of 110 nm as a top electrode was finally deposited using vacuum evaporator. 2.4. Characterization UV–vis absorption spectra of NiPc on the FTO-TiO2 compact layer, MAPbI3 on the FTO-TiO2 compact layer and FTO-TiO2 MAPbI3 -NiPc were measured with a UV–vis spectrophotometer (JASCO-770). Steady-state and time-resolved photoluminescence spectra measurements were performed with fluorescence spectrophotometer (Edinburgh Instruments, FLSP-920). The excitation was performed with pump pulse at 471 nm provided by a nanosecond tunable OPOLett-355II laser. The data were fitted with bi-exponential function of A1 exp(−t/ 1 ) + A2 exp(−t/ 2 ). Raman spectra (532 nm) were collected with LabRam HR 800. X-ray diffraction patterns were recorded on a Rigaku diffractometer using Cu K␣ irradiation. SEM was performed with FEI (Field Emission Instruments: Nova Nano SEM 450).


G Model

ARTICLE IN PRESS

JMST-1213; No. of Pages 7

M. Haider et al. / Journal of Materials Science & Technology xxx (2018) xxx–xxx

3

Fig. 1. (a) Device architecture and (b) energy level diagram of the device: FTO/compact TiO2 /CH3 NH3 PbI3 /NiPc/Au.

2.5. Solar cell characterization The photocurrent–voltage (J-V) characteristics of solar cells were measured by using an electrochemical station (biologic, VMP300) under illumination of a simulated sunlight (AM 1.5G, 100 mW cm−2 ) provided by Oriel Sol3A solar simulator (Newport USA, Model: 94023A) with an AM 1.5 filter in ambient air. Light intensity was calibrated with a Newport calibrated standard Si reference cell (SER. No: 506/0358). The J-V curves were obtained at a sweep rate of 200 mV s−1 . The measurement of the incident photo–to–current conversion efficiency was obtained by QTest Station 1000AD (CROWNTECH). Prior to measurement, a standard silicon solar cell was used as reference. 3. Results and discussion Device fabrication procedure is listed as follows. Cleaned fluorine doped SnO2 (FTO) substrate was firstly coated with a thin compact TiO2 as an electron transporting (hole blocking) layer. The CH3 NH3 PbI3 (MAPbI3 ) film was prepared via one step spin coating method. The p-type HTM NiPc was then introduced by in-situ thermal deposition under high-vacuum condition. Finally, the photovoltaic devices were completed by depositing gold (Au) as the counter electrode. The structure of resultant device is schematically shown in Fig. 1(a). The detailed device fabrication procedure is described in the experimental section. On the basis of the HOMO level of NiPc of −5.0 eV [44] and its bandgap of 1.6 eV determined by UV–vis absorption spectrum analysis (Fig. S1), its LOMO level is estimated to be −3.4 eV. The energy band diagram of the complete device together with the possible separation and transport of photo-generated electrons and holes towards the corresponding electrodes is shown in Fig. 1(b). Under illumination, the holes are transported from the perovskite active layer to NiPc (HTM) layer and then to gold electrode based on the proper HOMO position of NiPc (a little higher than that of perovskite). Meanwhile, the electrons are effectively blocked away from NiPc as the LUMO level of NiPc (−3.4 eV) is higher than that of the perovskite (−3.7 eV), hindering the undesired recombination. To characterize the crystal structures of components of each layer in the device, X-ray diffraction (XRD) technique was carried out on three FTO/TiO2 , FTO/TiO2 /MAPbI3 and FTO/TiO2 /MAPbI3 /NiPc films (Fig. 2(a)). All diffraction peaks in XRD pattern of the FTO/TiO2 film are assigned to rutile phase SnO2 and no peaks of anatase TiO2 are formed due to very thin TiO2 layer deposited on FTO. The presence of anatase TiO2 on FTO is indicated by the active mode of 144 cm−1 (Eg ) in Raman spectrum (Fig. S2). Compared to FTO/TiO2 , FTO/TiO2 /MAPbI3 gives the additional peaks at 14.05◦ , 28.42◦ , 31.78◦ and 43.37◦ , which are assigned to the diffraction peaks of (110), (220), (310) and (330) planes of MAPbI3 crystal. A weak peak at 6.68◦ , which is attributed to (110) plane of

NiPc [45,46] is present in XRD pattern of FTO/TiO2 /MAPbI3 /NiPc. In order to further confirm this assignment, XRD pattern of NiPc film deposited on a quartz substrate was examined and gives a much more intensive peak at 6.68◦ (Fig. S3). The morphology of the top surface of FTO/TiO2 /MAPbI3 and FTO/TiO2 /MAPbI3 /NiPc films was studied by scanning electron microscopy (SEM) as shown in Fig. 2(b) and (c). The pine holes-free MAPbI3 layer with the grain size of several micrometers fully covers the FTO/TiO2 substrate. After the deposition of NiPc layer on the top of MAPbI3 layer, the original smooth top becomes coarse. This indicates that the NiPc layer consists of small particles. Obvious grain boundaries are formed in the NiPc modified MAPbI3 layer largely because small-size NiPC molecules easily permeate into MAPbI3 grain boundaries during thermal evaporation deposition of NiPc. The filling of NiPc into the grain boundaries of MAPbI3 may improve the efficiency of collecting photogenerated holes as a result of the increased interface contact. The cross-section SEM image of the cell device demonstrates a well-defined layer-by-layer structure with sharp interfaces (Fig. S4). The absorption spectrum of FTO/TiO2 /MAPbI3 in Fig. 2(d) shows that the absorbance at the wavelength range between 550 and 800 nm is obviously lower than that before 550 nm because of the sharply decreased absorption coefficient of MAPbI3 with the increase of wavelength beyond 550 nm, limiting the utilization of photons at the long wavelength range. With the deposition of NiPc on FTO/TiO2 /MAPbI3 , the FTO/TiO2 /MAPbI3 /NiPc film has an increased absorbance at the wavelength range from 550 to 800 nm as a result of a wide Q band absorption of NiPc itself from 500 to 750 nm. This Q band absorption has two feature peaks at 624 and 676 nm in visible light region as indicated by UV–vis absorption spectrum of NiPc film deposited on FTO substrate (Fig. S1). It is anticipated that the low energy photons penetrating MAPbI3 layer can be absorbed by the NiPc HTM layer. The photogenerated electrons in NiPc can inject into the MAPbI3 layer to contribute complementary solar energy conversion. In order to achieve the best efficiency, the thickness of HTM layer needs be finely tuned for an effective collection of photogenerated holes (and blocking photogenerated electrons) from MAPbI3 perovskite active layer. A series of devices with different thicknesses of NiPc HTM layers were fabricated and their J-V curves were recorded under AM 1.5 G sunlight irradiation with power density of 100 mW cm−2 (Fig. 3(a)). The corresponding data of different parameters was given in Table S1. The device with a thin NiPc layer (100 nm measured from the detector) shows a relatively low efficiency of 8.92% with Jsc of 15.32 mA cm−2 , Voc of 0.91 V and FF of 64%. With the thickness increase of NiPc layer to 120 nm, a maximum efficiency of 12.1% was achieved with Jsc of 17.64 mA cm−2 , Voc of 0.94 V and FF of 73%. The device performance decreases when the thickness of the NiPc layer is beyond 150 nm. The dependence of the conversion efficiency on the NiPc layer thickness is


G Model JMST-1213; No. of Pages 7 4


Fig. 2. (a) XRD patterns of FTO/TiO2 (black line), FTO/TiO2 /MAPbI3 (red line) and FTO/TiO2 /MAPbI3 /NiPc (blue line). (b, c) Top-view SEM images of FTO/TiO2 /MAPbI3 and FTO/TiO2 /MAPbI3 /NiPc. (d) UV–vis absorption spectra of FTO/TiO2 /NiPc (black dot), FTO/TiO2 /MAPbI3 (red line) and FTO/TiO2 /MAPbI3 /NiPc (blue line).

Fig. 3. (a) Current density–voltage (J–V) curves of the PSCs with different thicknesses of NiPc. (b) J–V curves of devices with NiPc and CuPc as HTM. (c) Efficiency histogram of devices based on NiPc (and CuPc) as HTM for 30 cells. (d) Incident photo-to-current conversion efficiency (IPCE) spectra of the best-performing cell using the NiPc HTM. (The J-V curves were recorded under reverse scan with scan rate of 200 mV s−1 ).

controlled by the balance between the photo-hole extraction and transport. A thin NiPc film cannot cause a full coverage of MAPbI3 perovskite layer to prevent the direct contact between Au electrode

and MAPbI3 perovskite layer, which causes serious interface recombination. Too thick NiPc HTM layer increases the series resistance of the device due to low conductivity of NiPc. This correspond-




5

Fig. 4. (a) Steady-state photoluminescence (PL) spectra of MAPbI3 on quartz (Quartz/MAPbI3 ) and MAPbI3 sandwiched between quartz and NiPc film (Quartz/MAPbI3 /NiPc). (b) Time-resolved photoluminescence (TRPL) decay curves of Quartz/MAPbI3 and Quartz/MAPbI3 /NiPc taken at the peak emission wavelength of 775 nm. Excitation is performed with a pulsed laser source at 471 nm with an excitation power density of 1 nJ cm−2 impinged on the quartz substrate side. Laser pulse length is 4 ns. Solid lines represent the fitting curve resulting from exponential fitting in the form of y = A1 exp(x/t1 ) + A2 exp(x/t2 ).

ingly limits the transport of the photo-holes in the NiPc layer and increases their recombination with photo-electrons. To evaluate the performance of NiPc as HTM in PSCs, we also fabricate PSCs under the same conditions based on CuPc HTM, an outstanding representative of reported MPcs, as the control devices. The best performance device with NiPc shows a PCE of 12.1% with Jsc of 17.64 mA cm−2 , Voc of 0.94 V, and FF of 73%, while the best performance device with CuPc only shows a PCE of 10.4% with Jsc of 16.8 mA cm−2 , Voc of 0.92 V and FF of 67.5% (Fig. 3(b)). Moreover, the reproducibility of these two kinds of PSCs were evaluated by making tens of devices with NiPc and CuPc, respectively. The efficiency statistical histogram of devices based on NiPc shifts to higher value in comparison with the histogram of devices based on CuPc (Fig. 3(c)). The distribution range of Jsc , Voc , FF and PCE and the corresponding average values of these two kinds of devices were summarized in Fig. S5 and Table S2, respectively. It is obvious that the average values of devices based on NiPc are higher than those of devices based on CuPc. These results indicates NiPc show better performance as HTM in PSCs than CuPc. The incident photo-to-current conversion efficiency (IPCE) spectrum of the best performing device was recorded to understand the role of HTM more clearly (Fig. 3(d)). The IPCE value gradually decreases up to ∼750 nm after reaching the peak value at ∼550 nm, which is in accordance with the absorption feature of MAPbI3 perovskite active layer. The main reason is the insufficient absorption of the perovskite layer to reduce photocurrent at this range. The photocurrent density (Jsc ) integrated from the IPCE is about 15.0 mA cm−2 that is lower than that (17.64 mA cm−2 ) extracted from J-V curves. This discrepancy on Jsc may result from the existence of interface capacity that contributes to the formation of hysteresis behavior of device. The interface could be charged/discharged with changing voltage scan directions, leading to the performance of device strongly depends on the scan direction (Fig. S6). Considering the practical application of NiPc HTM, large size PSCs with area of 0.5 cm2 were also assembled and tested (Fig. S7). The big size device shows a PCE of 6.2% with Jsc of 13.34 mA cm−2 , Voc of 0.88 V and FF of 53.1%. Noted that some hysteresis exists in J-V curve largely because of the charge accumulation at the interfaces of the planar cells. To evaluate photo-hole extraction ability of NiPc HTM, normalized steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) decay spectra were collected on MAPbI3 perovskite film deposited on quartz substrate (Quartz/MAPbI3 ) and MAPbI3 perovskite film sandwiched between quartz substrate and NiPc layer (Quartz/MAPbI3 /NiPc). The Quartz/MAPbI3 film shows a strong steady-state PL emission peak at 775 nm (Fig. 4(a)), indicat-

ing the formation of high-quality perovskite film on the substrate. The emission peak is effectively quenched after the NiPc layer deposition on top of MAPbI3 film (Quartz/MAPbI3 /NiPc). The significant emission quenching indicates that the photo-holes can be efficiently extracted from MAPbI3 perovskite active layer into NiPc layer, hindering their radiative recombination with photoelectrons due to the spatial separation. The extraction rate of photo-holes at MAPbI3 /NiPc interface was evaluated by comparing the lifetimes of charge carriers in Quartz/MAPbI3 /NiPc and Quartz/MAPbI3 in TRPL spectra (Fig. 4(b)). The TRPL decay curve of Quartz/MAPbI3 shows a linear shape, suggesting that the decay follows one rate. The decay lifetime of 88.41 ns is obtained by fitting the dynamic curve. In contrast, Quartz/MAPbI3 /NiPc gives a biexponential decay curve, indicating two recombination rates of photo-carriers. The much shorter lifetimes of 1.94 and 13.12 ns in the film suggest the effective extraction of photo-holes from MAPbI3 perovskite active layer. A comparison of photo-hole extraction ability between NiPc and CuPc has been carried out by comparing their PL quenching efficiency (Fig. S8). The PL quenching efficiency of NiPc is better than that of CuPc, indicating the superior photo-hole extraction ability of NiPc to CuPc. TRPL spectra of MAPbI3 perovskite films sandwiched between TiO2 and NiPc (TiO2 /MAPbI3 /NiPc) and sandwiched between TiO2 and CuPc (TiO2 /MAPbI3 /CuPc) were recorded to further evaluate and compare the photo-hole extraction rates of NiPc and CuPc (Fig. S9). Due to the existence of photo-carrier extraction layer, both samples show biexponential TRPL decay curves. The fitted photo-carrier decay lifetimes (2.13 ns and 39.91 ns) in TiO2 /MAPbI3 /CuPc are longer those (1.74 ns and 13.60 ns) in TiO2 /MAPbI3 /NiPc, again suggesting the superiority of NiPc to CuPc. All these PL results are in accordance with the performance characterization of devices. The impedance spectra of the complete device were carried out at varied positive voltage (from 0 to 0.8 V) to evaluate carrier interfacial dynamics (Fig. 5). The Nyquist plot of a PSC usually consists of two semi-arcs in different frequency ranges. The small semi-arc at the high frequency range is related to the hole transfer resistance across perovskite/HTM interface, and the semi-arc at the low frequency range reflects the charge recombination resistance (Rrec ) at ETM/perovskite interface [47,48]. The smaller the first semi-arc is, the faster the hole transfer at perovskite/HTM interface is. The Nyquist plots of device based on NiPc only show one semi-arc at the whole frequency range and the arc becomes small with the increase of positive voltage applied. This semi-arc represents the second one in the normal plots, reflecting the charge recombination resistance (Rrec ) at ETM/perovskite interface. The disappearance of the first




6

The best-performing device (its area, 0.09 cm2 ) with NiPc as HTM exhibits a PCE of 12.1% with Jsc of 17.64 mA cm−2 , Voc of 0.94 V and FF of 73%. A device with large area of 0.5 cm2 demonstrates a conversion efficiency of about 6.2%. The comparison of NiPc with reported CuPc in the devices shows the superiority of NiPc to CuPc as HTM as a result of the excellent extraction ability of holes from perovskite active layer. Furthermore, the device based on NiPc shows a much improved air stability in comparison with the control device based on commercial Spiro-OMeTAD. Over 80% of initial efficiency is maintained after 38 day storage in air. Our results demonstrate that NiPc is a promising candidate of HTM for the fabrication of large size, low cost and air stable PSCs. Acknowledgements

Fig. 5. Nyquist plot of the perovskite solar cells with NiPc as HTMs measured in dark at different applied bias ranged from 0 V to 0.8 V.

The authors thank the Major Basic Research Program, Ministry of Science and Technology of China (2014CB239401), the National Natural Science Foundation of China (Nos. 51402306, 51422210, 51629201, 51521091), the Key Research Program of Frontier Sciences CAS (QYZDB-SSW-JSC039) for the financial support. G. L. is grateful for the award of the Newton Advanced Fellowship. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jmst.2018.03.005. References

Fig. 6. Normalized efficiency decay curves of devices based on NiPc and commercial spiro-OMeTAD in air. The devices were not encapsulated.

semi-arc should be caused by the excellent transfer ability of holes into the NiPc HTM. As the NiPc has a good chemical and thermal stability, the long-term stability test of the device based on NiPc HTM was also conducted in air (humidity: ∼30%) to evaluate the effectiveness of NiPc as a protection shell of perovskite. In parallel, the device based on commercial spiro-OMeTAD HTM was tested as the control device. The device with NiPc shows a much superior stability to that with OMeTAD as shown in Fig. 6. Specifically, the efficiency of the device with NiPc can retain over 80% of the initial value after the storage of 38 days. However, only half of the initial efficiency of the device with spiro-OMeTAD was retained after 12 day storage. Besides the chemical and thermal stability of NiPc itself, the stability improvement for NiPc based devices is also contributed by both the hydrophobic nature of NiPc and the high quality NiPc film prepared by the thermal evaporation method. The good coverage of the hydrophobic NiPc film on the device top may effectively prevent the moisture-labile metal halide perovskite from being attacked by water. The largely improved efficiency stability of the device based on NiPc indicates that NiPc outperforms spiro-OMeTAD as the protection shell of perovskite. 4. Conclusion Phthalocyanine NiPc was investigated as a new type of stable high-efficiency hole transfer material in perovskite solar cells.

[1] F. Deschler, M. Price, S. Pathak, L.E. Klintberg, D.D. Jarausch, R. Higler, S. Huttner, T. Leijtens, S.D. Stranks, H.J. Snaith, M. Atature, R.T. Phillips, R.H. Friend, J. Phys. Chem. Lett. 5 (2014) 1421–1426. [2] S.D. Stranks, G.E. Eperon, G. Grancini, C. Menelaou, M.J. Alcocer, T. Leijtens, L.M. Herz, A. Petrozza, H.J. Snaith, Science 342 (2013) 341–344. [3] C. Wehrenfennig, G.E. Eperon, M.B. Johnston, H.J. Snaith, L.M. Herz, Adv. Mater. 26 (2014) 1584–1589. [4] G. Xing, N. Mathews, S. Sun, S.S. Lim, Y.M. Lam, M. Grätzel, S. Mhaisalkar, T.C. Sum, Science 342 (2013) 344–347. [5] M.M. Lee, J. Teuscher, T. Miyasaka, T.N. Murakami, H.J. Snaith, Science 338 (2012) 643–647. [6] N.G. Park, J. Phys. Chem. Lett. 4 (2013) 2423–2429. [7] M. Grätzel, Nat. Mater. 13 (2014) 838–842. [8] H.J. Snaith, J. Phys. Chem. Lett. 4 (2013) 3623–3630. [9] M.A. Green, A. Ho-Baillie, H.J. Snaith, Nat. Photonics 8 (2014) 506–514. [10] Y. Li, B. Ding, G.J. Yang, C.J. Li, C.X. Li, J. Mater, Sci Technol. (2017), http://dx. doi.org/10.1016/j.jmst.2017.11.003. [11] B. Ding, Y. Li, S.Y. Huang, Q.Q. Chu, C.X. Li, C.J. Li, G.J. Yang, J. Mater. Chem. A 5 (2017) 6840–6848. [12] L.L. Gao, C.X. Li, C.J. Li, G.J. Yang, J. Mater. Chem. A 5 (2017) 1548–1557. [13] J. You, Z. Hong, Y. Yang, Q. Chen, M. Cai, T.B. Song, C.C. Chen, S. Lu, Y. Liu, H. Zhou, Y. Yang, ACS Nano 8 (2014) 1674–1680. [14] W. Nie, H. Tsai, R. Asadpour, J.C. Blancon, A.J. Neukirch, G. Gupta, J.J. Crochet, M. Chhowalla, S. Tretiak, M.A. Alam, H.L. Wang, A.D. Mohite, Science 347 (2015) 522–525. [15] W. Ke, G. Fang, J. Wan, H. Tao, Q. Liu, L. Xiong, P. Qin, J. Wang, H. Lei, G. Yang, M. Qin, X. Zhao, Y. Yan, Nat. Commun. 6 (2015) 6700–6707. [16] Y. Zhao, A.M. Nardes, K. Zhu, Appl. Phys. Lett. 104 (2014) 213906–213911. [17] N.J. Jeon, J.H. Noh, W.S. Yang, Y.C. Kim, S. Ryu, J. Seo, S.I. Seok, Nature 517 (2015) 476–480. [18] W.S. Yang, J.H. Noh, N.J. Jeon, Y.C. Kim, S. Ryu, J. Seo, S.I. Seok, Science 348 (2015) 1234–1237. [19] B. Conings, L. Baeten, C. De Dobbelaere, J. D’Haen, J. Manca, H.G. Boyen, Adv. Mater. 26 (2014) 2041–2046. [20] H. Nishimura, N. Ishida, A. Shimazaki, A. Wakamiya, A. Saeki, L.T. Scott, Y. Murata, J. Am. Chem. Soc. 137 (2015) 15656–15659. [21] C. Huang, W. Fu, C.Z. Li, Z. Zhang, W. Qiu, M. Shi, P. Heremans, A.K.Y. Jen, H. Chen, J. Am. Chem. Soc. 138 (2016) 2528–2531. [22] S. Paek, I. Zimmermann, P. Gao, P. Gratia, K. Rakstys, G. Grancini, M.K. Nazeeruddin, M.A. Rub, S.A. Kosa, K.A. Alamry, A.M. Asiri, Chem. Sci. 7 (2016) 6068–6075. [23] K. Rakstys, M. Saliba, P. Gao, P. Gratia, E. Kamarauskas, S. Paek, V. Jankauskas, M.K. Nazeeruddin, Angew. Chem. Int. Ed. 55 (2016) 7464–7468. [24] D. Bi, A. Mishra, P. Gao, M. Franckeviˇcius, C. Steck, S.M. Zakeeruddin, M.K. Nazeeruddin, P. Bäuerle, M. Grätzel, A. Hagfeldt, ChemSusChem 9 (2016) 433–438. [25] D. Bi, B. Xu, P. Gao, L. Sun, M. Grätzel, A. Hagfeldt, Nano Energy 23 (2016) 138–144.




[26] K. Rakstys, A. Abate, M.I. Dar, P. Gao, V. Jankauskas, G. Jacopin, E. Kamarauskas, S. Kazim, S. Ahmad, M. Grätzel, M.K. Nazeeruddin, J. Am. Chem. Soc. 137 (2015) 16172–16178. [27] M.E. Ragoussi, M. Ince, T. Torres, Eur. J. Org. Chem. 2013 (2013) 6475–6489. [28] C.G. Claessens, U. Hahn, T. Torres, Chem. Rec. 8 (2008) 75–97. [29] G. de la Torre, C.G. Claessens, T. Torres, Chem. Commun. (2007) 2000–2015. [30] R.A. Belisle, P. Jain, R. Prasanna, T. Leijtens, M.D. McGehee, ACS Energy Lett. 1 (2016) 556–560. [31] G. Sfyri, C.V. Kumar, G. Sabapathi, L. Giribabu, K.S. Andrikopoulos, E. Stathatos, P. Lianos, RSC Adv. 5 (2015) 69813–69818. [32] W. Ke, D. Zhao, C.R. Grice, A.J. Cimaroli, G. Fang, Y. Yan, J. Mater. Chem. A 3 (2015) 23888–23894. [33] F. Zhang, X. Yang, M. Cheng, W. Wang, L. Sun, Nano Energy 20 (2016) 108–116. [34] G. Sfyri, C.V. Kumar, Y.L. Wang, Z.X. Xu, C.A. Krontiras, P. Lianos, Appl. Surf. Sci. 360 (2016) 767–771. [35] F. Javier Ramos, M. Ince, M. Urbani, A. Abate, M. Gratzel, S. Ahmad, T. Torres, M.K. Nazeeruddin, Dalton Trans. 44 (2015) 10847–10851. [36] K.T. Cho, O. Trukhina, C. Roldan-Carmona, M. Ince, P. Gratia, G. Grancini, P. Gao, T. Marszalek, W. Pisula, P.Y. Reddy, T. Torres, M.K. Nazeeruddin, Adv. Energy Mater. 7 (2017) 1601733–1601740.

7

[37] M. Cheng, Y. Li, M. Safdari, C. Chen, P. Liu, L. Kloo, L. Sun, Adv. Energy. Mater. 7 (2017) 1602556–1602563. [38] J. Liu, Y. Wu, C. Qin, X. Yang, T. Yasuda, A. Islam, K. Zhang, W. Peng, W. Chen, L. Han, Energy Environ. Sci. 7 (2014) 2963–2967. [39] G.D. Sharma, S.G. Sangodkar, M.S. Roy, Mater. Sci. Eng. B 41 (1996) 222–227. [40] S. Gravano, A.K. Hassan, R.D. Gould, Int. J. Electron. 70 (1991) 477–484. [41] R.D. Gould, J. Phys. D 19 (1986) 1785–1790. [42] T.G. Abdel-Malik, R.M. Abdel-Latif, A.E. El-Samahy, S.M. Khalil, Thin Solid Films 256 (1995) 139–142. [43] L. Yang, B. Xu, D. Bi, H. Tian, G. Boschloo, L. Sun, A. Hagfeldt, E.M.J. Johansson, J. Am. Chem. Soc. 135 (2013) 7378–7385. [44] F. Petraki, V. Papaefthimiou, S. Kennou, Org. Electron. 8 (2007) 522–528. [45] K.N. Narayanan Unni, C.S. Menon, Mater. Lett. 45 (2000) 326–330. [46] M. Neghabi, M. Zadsar, S.M.B. Ghorashi, Mater. Sci. Semicond. Process. 17 (2014) 13–20. [47] A. Dualeh, T. Moehl, N. Tetreault, J. Teuscher, P. Gao, M.K. Nazeeruddin, M. Graetzel, Acs Nano 8 (2014) 362–373. [48] H.S. Kim, J.W. Lee, N. Yantara, P.P. Boix, S.A. Kulkarni, S. Mhaisalkar, M. Graetzel, N.G. Park, Nano Lett. 13 (2013) 2412–2417.
