Electrochimica Acta

(1)

Enhancing the ef ﬁ ciency of low-temperature planar perovskite solar cells by modifying the interface between perovskite and hole

transport layer with polymers

Yangyang Cai

^a

, Zongbao Zhang

^a

, Yang Zhou

^a

, Hui Liu

^a

, Qiqi Qin

^a

, Xubing Lu

^a

, Xingsen Gao

^a

, Lingling Shui

^a

, Sujuan Wu

^a^,^*

, Junming Liu

^b

aInstitute for Advanced Materials and Guangdong Provincial Key Laboratory of Quantum Engineering and Quantum Materials, South China Academy of Advanced Optoelectronics, South China Normal University, Guangzhou 510006, China

bLaboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China

a r t i c l e i n f o

Article history:

Received 26 October 2017 Received in revised form 15 December 2017 Accepted 19 December 2017 Available online 22 December 2017

Keywords:

Interface modiﬁcation by polymers Low-temperature TiO2compact layer Photoelectric properties

Planar perovskite solar cells

a b s t r a c t

In this work, planar perovskite solar cells (PSCs) based on CH3NH3PbI3 perovskite layer and low- temperature processed TiO2 have been fabricated. Polymers including poly(methylmethacrylate) (PMMA), poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) and polyethylene glycol (PEG) in chlorobenzene solution have been selected to modify the interface between perovskite and hole transport layer (HTL), respectively. The concentrations of the three polymer solutions have been optimized. The effect of interfacial modification by different polymer solutions on the photoelectric properties of perovskite layer and the performance of PSCs has been systematically investigated. The microstructure and photoelectric properties of the modified perovskitefilms has been systematically studied by X-ray diffraction (XRD), scanning electron microscopy (SEM), atomic force microscopy (AFM), conducting force microscopy (CFM) and Kelvin probe force microscopy (KPFM). The results reveal that the modified perovskitefilms with tetrahedral perovskite structure have lager grain size, lower roughness and better photoelectric properties compared with the reference sample. The electron trap state density (Dtrap), charge extraction, carrier transfer and recombination process in the PSCs have been investigated by current-voltage (I-V) characteristic curves, steady-state photoluminescence (PL), photo- voltage decay and electrochemical impedance spectroscopy (EIS). The results indicate that the polymeric interface modification at the optimum concentration can reduce the Dtrap, promote the charge transfer and suppress carrier recombination, resulting in the improved performance of PSCs. All of the modified PSCs at an optimum concentration exhibit the improvedfill factor (FF) and open circuit voltage (Voc), thus the power conversion efficiency (PCE) is enhanced to over 17% from 15.49%.

1. Introduction

In recent years, organic-inorganic hybrid perovskite solar cells (PSCs) have attracted considerable attentions because of their simple process and high power conversion efﬁciency (PCE) [1e4].

The PCE of PSCs have increased from 3.8% to more than 22% in the just seven years [5,6]. Due to the simple device structure and the compatible process with theﬂexible substrates, low-temperature processed planar PSCs are widely studied [7,8]. It is well known

that a good perovskitefilm is necessary for the high performance of PSCs [9]. The microstructure of perovskite layer has an significant impact on exciton separation, charge transfer and recombination process of PSCs [10]. High-quality perovskite films need to be dense, high surface coverage and lager grain size [11]. Lots of methods have been applied to control the morphology of perovskite layer such as solvent extraction [12e15], interface modification [16e20] and additive engineering [21e26]. Among them, the solvent extraction is an effective and widely used method to adjust the microstructure of perovskite layer [13e15]. On the one hand, chlorobenzene is one of the most popular extraction solvent which can induce a fast precipitation of CH3NH3PbI3and lead to a uniform perovskitefilms with high surface coverage [15].

*Corresponding author.

E-mail address:[email protected](S. Wu).

Contents lists available atScienceDirect

Electrochimica Acta

j o u r n a l h o me p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e l e c t a c t a

(2)

On the other hand, polymers are one of the important materials.

They have been widely used in dye-sensitized solar cells (DSSCs), organic solar cells (OSCs) and PSCs. Natural cellulose fibers were used to fabricate bioderived photoanodes and polymer electrolytes of DSSCs. The PCE of the DSSC based on cellulose-based electrodes and electrolytes retained 96% of its initial value after 1000 h of accelerated aging test [27]. The OSCs based on the polymer:fullerene blend of poly[9,9-dioctylfluorene-4,7-alt-(5,6- bis(octyloxy)-4,7 -di(2,2⁰-bithiophen-5-yl)benzo[c] [1,2,5]thiadia- zole)-5,5-diyl]:phenyl-C71-butyric acid methyl ester (PFDT2BT- 8:PC71BM) could retain about 50% of their initial efficiency after over 1.4 years stored in outdoor circumstance [28]. The PCE of DSSCs based on cobalt polymer redox had reached over 14% [29]. It was found that the introduction of a polymer-based luminescent down-shifting layer to OSCs significantly improved device efficiency and lifetime [30]. A new copolymer synthesized by S.

Ramesh group had been identiﬁed as a promising candidate to be used as a base polymer in electrolyte formulation for DSSCs [31]. It can be seen that polymeric materials are very important for 3rd generation solar cells.

In addition, various polymers have been introduced into perovskite precursor solutions to control the morphology of perovskite films [22e26]. Colella et al. have found that some polymers as cooperative additives can establish hydrogen bond with perovskite precursors and drive thefilm growth by acting as a three-dimensional template to enhance the morphology of perovskitefilm [22]. It was found that adding polyethylene glycol (PEG) into the perovskite precursor solutions can improve the coverage of perovskite films [23,24]. PEG molecules can establish hydrogen bonds with CH3NH3þto promote the growth of perovskite grains and improve the stability of PSCs to humidity [24]. Using polymer polyethylenimine (PEI) as additive into the perovskite precursor solution also can improve the morphology and crystallinity, resulting in the promoted charge extraction and transfer [25]. Gigli et al. have reported that the poly[2-methoxy-5-(2-ethylhexyloxy)- 1,4-phenylenevinylene] (MEH-PPV) as additive can control the self- assembly of CH3NH3PbI3crystalline domains and obtain a smooth and homogenous perovskitefilm, but the obtained PSCs show a PCE of only 3% [26].

Except for the chlorobenzene solvent extraction and polymer additive enhancement, the polymer interfacial modification also is one of the important methods to promote charge transfer and collection in PSCs. It is well-known that the perovskite/HTL interface also plays an important role on the charge transport and collection and has a significant impact on the performance of PSCs [9,11]. Researches on the polymeric interface modification between perovskite and HTL have also attracted much attention [16e20].

Matsuda et al. have found that inserting a thin poly(methyl methacrylate) (PMMA) layer between the annealed perovskitefilm and HTL can reduce charge recombination, leading to the improved performance of PSCs [17]. Moreover, Gr€atzel's group have found that PMMA can be used as a template to control nucleation and crystal growth of perovskite layer, resulting in a smooth perovskite film with excellent electronic characteristics [18]. Lin et al. have reported that depositing a thin polystyrene layer on the perovskite film can promote the separation of photo-generated electrons and holes, resulting in the reduced charge recombination rates at the interface [19]. Miyano et al. have found that depositing a thin (poly [9,9-bis(3⁰-(N,N-dimethylamino)-propyl)-2,7-fluorene)-alt-2,7- (9,9-dioctylfluorene)]) (PFN-P1) layer on perovskite films can obtain a more uniform distribution of grain size and the improved the stability of PSCs [20]. Polymers such as PMMA, MEH-PPV and PEG all have long molecular chain structure and can establish chemistry interactions with perovskite precursors to modify the

microstructure of perovskiteﬁlms [22,24,26]. Until now, it is not found that MEH-PPV and PEG have been used to optimize the perovskite/HTL interface and improve the performance of PSCs.

Moreover, there are no reports to simultaneously investigate and compare the effect of PMMA, MEH-PPV and PEG interfacial modification on the photoelectric properties of perovskite layer and the modified PSCs. In order to simultaneously control the microstructure of perovskite layer and optimize the CH₃NH₃PbI₃/HTL interface, three easily-accessible polymers solutions including PMMA, MEH-PPV and PEG chlorobenzene solutions were selected to modify the as-spun perovskite film. This method not only incorporates the triple advantages of chlorobenzene solvent extraction, polymer additive enhancement and interface modification, but also combines the three advantages into one step to simplify the process. Here polymers not only can act as the template for perovskite growth but also modify the CH3NH3PbI3/HTL interface.

Based on these considerations, planar PSCs based on CH₃NH₃PbI₃ perovskite ﬁlm, 2,2⁰,7,7⁰-tetrakis(N,N-di-4-methoxy- phenylamino)-9,9⁰-spirobiﬂuorene (spiro-OMeTAD) HTL and low- temperature processed TiO2 compact layer have been fabricated.

Our PSCs consist of FTO/TiO2/CH3NH3PbI3/spiro-OMeTAD/Ag structure. The PMMA, MEH-PPV and PEG chlorobenzene solutions were used to modify the CH3NH3PbI3/spiro-OMeTAD, respectively.

The concentrations of polymer chlorobenzene solutions have been optimized. The effect of these polymer solutions on the microstructure and photoelectric properties of perovskiteﬁlm, and the performance of PSCs have been systematically investigated. The related mechanism has been explored. The results demonstrate that the interfacial modiﬁcation can improve the microstructure of perovskite layer, decrease electron trap state density (Dtrap), promote carrier transfer and suppress charge recombination in PSCs.

Finally, all of the modified PSCs demonstrate the enhanced PCE from 15.49% to over 17.0%. For convenience of presentation, we discuss four types offilms/PSCs: the referencefilms/PSCs that the interface between CH₃NH₃PbI₃ and spiro-OMeTAD has not been modified by polymer, the PMMA, MEH-PPV and PEG-films/PSCs that the interfaces are modified by PMMA, MEH-PPV and PEG chlorobenzene solution, respectively.

2. Experimental section 2.1. Material and methods

CH3NH3I was synthesized by the reported method [32]. PbI2

(99.9985%) and PMMA (Molecular weight (M) ¼ 43982) were purchased from Alfa Aesar. Dimethyl sulfoxide (DMSO) (99.9%) and 4-hydroxybutyric acid lactone (GBL) (99.9%) were bought from Sigma-Aladrich and Aladdin, respectively. Spiro-MeOTAD, [6,6]-phenyl-C61-butyric acid methyl ester (PCBM) and PEG (M¼20000) were bought from the company of Feiming in Shengzhen, Banhe in Taiwan and Beilian of Tianjing, respectively.

MEH-PPV (M¼100000e1000000) and ﬂuorine-doped tin oxide (FTO, 15 Ohm/square) were purchased from luminescence technology of Taiwan and Asahi Glass Company Limited of Japan, respectively.

FTO substrates were ultrasonically cleaned with deionized water, acetone, isopropanol and ethanol successively, and dried with nitrogen (N₂)flow. After oxygen UV treatments for 15 min, the cleaned FTO substrates were immersed in a TiCl4 solution (200 mM) in a closed vessel at 70C for 1 h. After washed with deionized water and ethanol, thefilms were dried with N₂flow and annealed at 200C for 30 min in air. CH3NH3I and PbI2with a molar ratio of 1:1 were dissolved in a mixed solution with DMSO and GBL

(3)

(volume ratio of DMSO and GBL is 3:7) to obtain a 40 wt% perovskite precursor solution. After the TiO₂ films were treated with oxygen UV for 10 min, perovskite precursor solution was spin- coated onto the TiO2 films at a speed 4000 rpm for 40 s in the glove box. Then the polymers (PMMA, MEH-PPV, PEG) chlorobenzene solution with different concentrations were dropped on the as-spun perovskitefilms during spin-coating at 20 s, respectively.

For the reference samples, only the chlorobenzene was dropped on the perovskite layer. After spin-coating, the perovskitefilms were annealed at 100C for 10 min immediately. Subsequently, 0.08 M spiro-OMeTAD in chlorobenzene solution was spin-coated onto the perovskitefilm at 5000 rpm for 30 s. For the devices of electron trap state density study, 20 mg/ml PCBM chlorobenzene solution was spin-coated onto the perovskitefilm at 3000 rpm for 30 s. These samples were left in dry air overnight in the dark. Finally, silver (Ag) electrode with thickness of ~100 nm was evaporated on the sample surface through a shadow mask under a vacuum of 110⁴Pa. The sample size is 0.045 cm².

2.2. Characterizations

The morphology of the perovskitefilms was characterized by scanning electron microscopy (SEM, ZEISS ULTRA 55). The local electrical properties such as photo-current and contact potential difference (CPD) were characterized by the measurements of conductive atomic force microscopy (CFM) and Kelvin probe atomic force microscopy (KPFM) using a white light-emitting diode (LED) with an irradiance of 10 mW/cm²by an atomic force microscope (AFM) (Asylum Research, Cypher). The crystallinity of perovskite films was characterized by XRD (X'Pert PRO, CuKa radiation). The current density-voltage (J-V) characteristics of these PSCs were measured by Keithley 2400 source meter under an illumination of 100 mW/cm² (Newport 91160, AM1.5G). The radiation intensity was calibrated by a standard silicon solar cell. The current-voltage (I-V) curves were measured by Keithley 2400 source meter in the dark. The external quantum efficiency (EQE) was measured using EQE system (Newport 66902). The UVevis absorption spectra of the perovskitefilms were measured by spectrophotometer (SHIMADZU UV-2550). The electrochemical impedance spectroscopy (EIS) and photo-voltage decay measurements were performed on the Zahner Zennium electrochemical workstation in the dark. For the EIS measurements, a 5 mV ac-sinusoidal signal source was employed over a1.0 V constant bias with the frequency ranging from 1 Hz to

1 MHz. For the photo-voltage decay measurements, a LED (l¼526 nm) with an illumination 300 W/m²was using as light source. The steady-state photoluminescence (PL) spectrum was measured by aﬂuorescence spectrophotometer (HITACHI F-5000) exited at 515 nm.

3. Results and discussion

In this work, the CH3NH3PbI3/spiro-MeOTAD interface has been modified by different polymer chlorobenzene solution including PMMA, MEH-PPV and PEG, respectively. The polymer solutions are respectively spin-coated on the as-spun perovskite films. Fig. 1 shows the schematic drawing of the interfacial modification by polymer solutions. The photovoltaic performance of the modified PSCs has been investigated.Fig. S1eS3show the J-V curves and the detailed photovoltaic parameters as a function of concentrations, respectively. The reference PSC achieves a PCE of 15.49% with a short-circuit current density (J_sc) of 22.64 mA/cm², an open circuit voltage (Voc) of 0.967 V and a fill factor (FF) of 70.75%. For the PMMA-, MEH-PPV- and PEG-PSCs, the concentrations vary from 0.2 to 0.8 mg/ml, 1e4 mg/ml and 0.04e0.08 mg/ml, respectively. The PMMA-PSC with an optimum concentration of 0.6 mg/ml achieves the highest PCE of 17.15%, yielding a Jscof 21.86 mA/cm², a Vocof 1.008 V and a FF of 77.76%. The MEH-PPV-PSC with an optimum concentration of 2 mg/ml also indicates an enhanced PCE of 17.09%

with a Jscof 22.01 mA/cm², a Vocof 0.988 V and a FF of 78.5%. For the PEG-PSC with an optimum concentration of 0.06 mg/ml, it gives a PCE of 16.96% with a Jscof 22.10 mA/cm², a Vocof 0.994 V and a FF of 77.14%. For all of the modified PSCs, a higher concentration than the optimum value can result in the dramatic decrease of Jsc, Vocand FF, thus a lower PCE. Because the interfacial modification by polymer solutions with the higher concentration can reduce the quantum tunneling of holes to the anode, which will reduce the carrier collection efficiency, resulting in the decreased PCE [17,18]. There- fore the optimum concentrations for PMMA, MEH-PPV and PEG are 0.6, 2 and 0.06 mg/ml, respectively. These are the conditions for the three different polymer modification in all of our other experiments shown from Figs. 2e9,Figs. S4 and S5.Fig. 2(a) displays the J-V curves of the PSCs without and with polymer modification.Table 1 lists the corresponding photovoltaic parameters. It can be noted that the three polymer modification all improve the performance of PSCs. The PMMA-PSC shows the best PCE. The PCE of PMMA-PSC increases to 17.15% from 15.49% of the reference PSC. The reason

Fig. 1.Schematic drawing of the fabrication of the polymeric chlorobenzene solution modiﬁed perovskiteﬁlm.

(4)

to the improved performance after the interfacial modification will be explored below. In order to better compare the performance of the modified PSCs, a statistical histogram of the PCE for each batch of 40 individual devices is shown inFig. 2(b). The average PCE for reference, PMMA-, MEH-PPV- and PEG-PSCs are 14.42, 16.56, 16.38 and 16.29%, respectively. Obviously, the modified PSCs demonstrate the higher PCE than that of the reference PSC.Fig. 2(c) and (d) show the detailed photovoltaic parameter variation including J_sc, V_oc, FF and PCE of the reference and modified PSCs. It can be seen that all the modified PSCs exhibit the increased Vocand FF compared with reference PSC, thus the enhanced PCE.

In order to explore the reason of the improved performance, the UVevis absorbance and EQE measurements were carried out.

Fig. 3(a) and (b) present the UVevis absorbance spectra and EQE curves of the PSCs, respectively. As shown in Fig. 3(a), all the samples demonstrate similar absorbance. It can be seen that the interfacial modification by PMMA, MEH-PPV and PEG does not change the absorption of PSCs. Compared to the reference PSC, the modified PSCs display the reduced EQE in the range of 400e750 nm, which is consistent with the decreased Jsc in J-V characteristics. The integrated J_scvalues are listed inTable 1, which are close to the measured J_sc. It was reported that the appearance of polymer in perovskitefilms can increase the series resistance of PSCs, resulting in the decrease of Jsc because of the insulating characteristics of polymer [18,23]. Thus the decreased J_sc in the modified PSCs can be attributed to the insulating characteristics of the three polymers. The value of shunt resistance (Rsh) and series resistance (R_s) can be obtained from J-V measurement. Fig. 3(c) shows the variation of Rsh/Rsand FF of the PSCs with and without modification. As seen inFig. 3(c), the FF increases with the increase

Fig. 2.(a) J-V curves; (b) Statistical histogram of the PCE; (c) and (d) Voc, Jsc, FF and PCE for reference, PMMA-, MEH-PPV- and PEG-PSCs.

Table 1

Photovoltaic parameters of reference, PMMA-, MEH-PPV- and PEG-PSC at the optimum concentration.

Polymer Voc(V) Jsc(mA/cm²)^c Jsc(mA/cm²)^d FF (%) PCE (%) reference 0.967 22.64 22.14 70.75 15.49^a(14.42)^b

PMMA 1.00 21.86 21.22 77.76 17.15^a(16.56)^b

MEH-PPV 0.988 22.01 21.45 78.50 17.09^a(16.38)^b

PEG 0.994 22.10 21.67 77.14 16.96^a(16.29)^b

aBest PCE.

bAverage PCE from 40 devices.

c Measured Jsc.

d Integrated Jsc from the EQE curves.

Fig. 3.(a) UVevis absorbance spectra; (b) EQE spectra, and (c) Rsh/Rsand FF for reference, PMMA-, MEH-PPV- and PEG-PSC.

(5)

of Rsh/Rs. Because the FF depends on the value of Rsh/Rs[33,34], the higher FF in the modified PSCs can be attributed to the larger R_sh/R_s. To study the effect of polymeric interface modification on the electrical properties of perovskitefilm at nanoscale level, the local electrical properties such as photo-current and contact potential difference (CPD) were characterized by CFM and KPFM, respectively [35,36].Fig. 4(a)e(d) show the CFM images of the reference film, PMMA-, MEH-PPV- and PEG-film, respectively. The corresponding average photo-current values for the four samples are respectively shown inFig. 4(i). It is clear that the average photo- current decreases from 22.1 pA of reference film to 2.1 pA of PMMA-film, 6.6 pA of MEH-PPV-film and 9.0 pA of PEG-film. It was reported that the measured average photo-current from CFM is agreement with the current of bulk devices and can be correlated with the device performance as well [36]. The Jsc variation dis- cussed above is consistent with the photo-current result from CFM, which can explain the reason about the modified PSCs with lower J_scthan that of reference PSC.Fig. 4(e)e(h) demonstrates the KPFM images of perovskitefilms with and without modification, respectively. The average CPD for the four samples are respectively shown inFig. 4(j). There is an obvious rise in the average CPD from 242 mV of referencefilm to 645 mV of PMMA-film, 418 mV of MEH-PPV- film and 595 mV of PEG-film. It was reported that the average CPD at nanometer scale was in accordance with the bulk devices [37]. Thus the higher Vocis associated with the larger CPD in the modified PSCs, which corresponds to the effective photo-generated charge separation [37].

The XRD patterns are used to investigate the effect of polymeric interface modiﬁcation on the crystal structure of perovskiteﬁlms.

Fig. 5shows the XRD patterns of the reference, PMMA-, MEH-PPV- and PEG-film. The peaks at 23.39 and 26.45are ascribed to the TiO2/FTO substrate. The peaks at 14.05, 19.97, 24.57, 28.41 and 31.85correspond to the (001), (112), (202), (004) and (310) crystal planes of CH₃NH₃PbI₃ with tetrahedral perovskite structure, respectively [38]. It can be seen that the reference, PMMA-, MEH- PPV- and PEG-film show the comparative XRD intensity and no other phases appear. This indicates that the polymer modifications do not change the crystalline structure of the perovskitefilms.

To study the effect of interfacial modiﬁcation on the micrograph of perovskiteﬁlms,Fig. 6(a)e(d) show the top-view SEM images of

Fig. 4.CFM images of CH3NH3PbI3film without and with interface modification: (a) referencefilm; (b) PMMA-film; (c) MEH-PPV-film; (d) PEG-film. KPFM images of CH3NH3PbI3

film without and with interface modification: (e) referencefilm; (f) PMMA-film; (g) MEH-PPV-film; (h) PEG-film. The scale bar is 1mm. (i) Histogram of average photo-current obtained from the CFM measurement; (j) Histogram of average CPD obtained from the KPFM measurement.

Fig. 5.XRD patterns of reference, PMMA-, MEH-PPV- and PEG-ﬁlm. The red“#”corresponds to the TiO2/FTO.

(6)

reference and modified perovskitefilms. The calculated grain size from the SEM images and Gaussianfitting of the statistical data are shown inFig. 6(e)e(h), respectively. The mean grain size increases from ~150 nm of referencefilm to ~200 nm of PMMA-film, ~220 nm of MEH-PPV-film and ~250 nm of PEG-film. The increased grain size

reduces the density of grain boundaries, resulting in the improved performance of the modified PSCs [10,11]. In addition, the surface characteristics of perovskitefilms also have an important impact on the optoelectronic properties of PSCs [11]. The surface morphology and roughness of the reference, PMMA-, MEH-PPV- and PEG-film Fig. 6.Top-view SEM images of the CH3NH3PbI3film without and with modification: (a) referencefilm; (b) PMMA-film; (c) MEH-PPV-film; (d) PEG-film. The scale bar is 200 nm.

The histogram of the calculated grain size and Gaussianfitting of the statistical data from SEM images: (e) reference; (f) PMMA-film; (g) MEH-PPV-film; (h) PEG-film.

Fig. 7.I-V curves of the electron-only devices displaying the VTFLkink point behavior: (a) referencefilm; (b) PMMA-film; (c) MEH-PPV-film; (d) PEG-film; (e) Histogram of VTFL

obtained from the I-V curves; (f) Histogram of the calculated Dtrap.

(7)

are investigated by AFM, as shown inFig. S4. The values of root mean square roughness (RMS) for the reference, PMMA-, MEH- PPV- and PEG-ﬁlm are 11.94, 9.63, 8.93 and 8.12 nm, respectively.

It can be seen that the polymeric modification decreases the RMS of perovskite layer, resulting in the smoother perovskitefilms. The smoother perovskite surface will greatly benefit to charge transfer and collection [9,17]. The modified perovskite films with better microstructure can be attributed the facts that polymers not only act as a skeleton to support perovskite crystal, but also establish weak chemical interactions with perovskite which are in favor of perovskite crystal growth [18,22,24,26]. The SEM and AFM results indicate that the polymeric modification can effectively regulate the microstructure of perovskitefilm. The improved microstructure will benefit to promote charge transfer and improve PCE.

It is noted that the PEG-PSC with the biggest grain size and lowest RMS does not demonstrate the highest PCE. It was reported that electron trap state density (D_trap) is an important parameter which can determine the performance of PSCs [39,40]. In order to study the reason of PCE variation for the modified PSCs, the Dtrapof perovskitefilms with and without interfacial modification has been investigated. A series of devices with the structure of FTO/TiO2/ perovskite/PCBM/Ag are fabricated [41]. The I-V curves of the four devices are shown inFig. 7(a)e(d), respectively. The linear relation in the I-V indicates the ohmic response of the electron-only device at low bias voltage. When the bias voltage is over the kink point, the current will increases nonlinearly, indicating the trap states are completelyfilled. The D_trapcan be calculated by the trapfilled limit voltage (VTFL) using theflowing equation [41]:

V_TFL¼eD_trapL² 2εε0

where the e is the electron elementary charge (e¼1.610¹⁹C), L is the thickness of perovskite film as shown in Fig. S5,ε is the relative dielectric constant of CH₃NH₃PbI₃(ε¼28.8),ε0is the vacuum permittivity (ε0¼8.85410¹²F/m). The obtained VTFLfrom Fig. 7(aed) and calculated Dtrapvalues for the four perovskitefilms are illustrated inFig. 7(e) and (f), respectively. The V_TFLvalues of reference, PMMA-, MEH-PPV-, PEG-film are 0.227, 0.106, 0.121 and 0.167 V, respectively. Thus the Dtrap values of reference, PMMA-, MEH-PPV-, PEG-film are 1.810¹⁶/cm³, 0.6110¹⁶/cm³, 0.6710¹⁶/cm³, 1.1510¹⁶/cm³, respectively. It is clear that D_trap deceases after the interfacial modification. Although the PEG-film has larger grain size (250 nm) compared to the PMMA-film (200 nm), the PMMA-film has the lowest D_trap, thus the highest PCE [39,40]. The result is agreement with the J-V result.

To investigate the effect of polymeric modiﬁcation on the charge extraction and transfer process, the steady-state PL and photo- voltage decay spectra are measured and displayed in Fig. 8.

Fig. 8(a) presents the PL spectra of the reference and modiﬁed

perovskitefilms deposited on the TiO₂coated FTO (FTO/TiO₂). It can be seen that the intensity of the emission peak drastically decreases in the modified samples compared to the reference sample. This indicates that the charge recombination rates are drastically Fig. 8.(a) PL spectra of reference, PMMA-, MEH-PPV- and PEG-film; (b) Photo-voltage decay curves of reference, PMMA-, MEH-PPV- and PEG-PSC.

Fig. 9.(a) The Nyquist plots of reference, PMMA-, MEH-PPV- and PEG-PSC; (b) The equivalent circuit diagram used toﬁt the Nyquist plots; (c) Fitted Rrecof reference, PMMA-, MEH-PPV- and PEG-PSC.

(8)

reduced, suggesting the enhanced charge extraction ability in the modified samples [42]. Fig. 8(b) shows the photo-voltage decay curves of the reference and modified PSCs. The photo-voltage decay is related to the decrease of electron density, which is mainly caused by the charge recombination [42e45]. As shown inFig. 8(b), the photo-voltage decay in the modified PSCs is slower than that of reference PSC. This result confirms that the recombination rates in the modified PSCs are lower than that of reference PSC [44,45]. This is consistent with the result of PL measurement. It is well-known that lower charge recombination rates contribute to higher FF and V_oc[34,42,43]. This can explain the higher FF and V_ocin the modified PSCs.

To further investigate the charge transfer and carrier recombination characteristics of these PSCs, EIS was carried out.Fig. 9(a) shows the Nyquist plots of PSCs measured at the applied voltage of1.0 V in the dark.Fig. 9(b) displays the model used tofit the data of Nyquist plots [46,47]. The solid lines inFig. 9(a) are thefitting results of experimental data. It can be seen that the Nyquist plot can be wellfitted by the model. As shown in theFig. 9(a), a significant difference between the reference and modified PSCs can be found.

The arc size in the modiﬁed PSCs is bigger than that of the reference PSC. The main arc at medium-frequency region is related to the charge recombination of TiO2/CH3NH3PbI3/spiro-OMeTAD [12,48].

The bigger arc for the modified PSCs indicates that the values of the recombination resistance (Rrec) in the modified PSCs are higher than that of the reference PSC [46,49]. The value of Rrec can be obtained from thefitting data.Fig. 9(c) displays the histogram of Rrecfor PSCs with and without modification. It can be seen that the modified PSCs all show higherRrecthan that of the reference PSC.

The carriers are injected from the supplied external voltage in the dark. The higherR_recin the modified PSCs can be related to the decreased electron trapping at the defect states which appear at the CH3NH3PbI3/spiro-OMeTAD interface [47]. It is reported that the value ofR_recis inversely proportional to charge recombination rates [50]. Thus the charge recombination rates in the modified PSCs are lower than that of the reference sample. This is consistent with the results of PL and photo-voltage decay measurement, which confirm that the recombination rates are reduced after interfacial modification. Thus the modified PSCs show better performance than that of the reference PSC. The PMMA-PSC corresponds to the maximalR_rec and the lowest Dtrap, thus the best performance.

4. Conclusions

In summary, the planar PSCs with the FTO/TiO2/CH3NH3PbI3/ spiro-OMeTAD/Ag structure have been fabricated. Polymer (PMMA, MEH-PPV and PEG) chlorobenzene solutions are used to modify the as-spun perovskite/spiro-OMeTAD interface and the concentrations have been optimized, respectively. It is found that the modified perovskitefilms show bigger grain size, smoother surface and lower Dtrap, resulting in the higher FF, Vocand PCE. The PL, photo- voltage decay and EIS measurements indicate that the polymeric interface modification reduces the charge recombination rates of the modified PSCs. It is worth noting that the highest processing temperature of PSCs is only 200C. This work makes it possible to developflexible devices by a simple low-temperature process.

Acknowledgements

The authors acknowledge theﬁnancial support of the National Key Research&Development Program of China (2016YFB0401502), the Natural Science Foundation of Guangdong Province (No.

2016A030313421), the Characteristic Innovation Project of Guang- dong Provincial Department of Education (Science 2016, 22), the National Natural Science Foundation of China (Grant Nos.

51431006, 51472093, 61574065), the Project for Guangdong Prov- ince Universities and Colleges Pearl River Scholar Funded Scheme (2016), the Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT, IRT13064), the Science and Technology Planning Project of Guangdong Province (No.

2015B090927006, 2016B090906004), the Guangdong Innovative Research Team Program (No. 2011D039), and the MOE Interna- tional Laboratory for Optical Information Technologies.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.electacta.2017.12.135.

References

[1] C. Zuo, H.J. Bolink, H. Han, J. Huang, D. Cahen, L. Ding, Advances in perovskite solar cells, Adv. Sci. 3 (2016), 1500324.

[2] M. Ye, X. Hong, F. Zhang, X. Liu, Recent advancements in perovskite solar cells:

ﬂexibility, stability and large scale, J. Mater. Chem. 4 (2016) 6755e6771.

[3] S. Prasanthkumar, L. Giribabu, Recent advances in perovskite-based solar cells, Curr. Sci. 111 (2016), 00113891.

[4] P. Wang, Y. Guo, S. Yuan, C. Yan, J. Lin, Z. Liu, Y. Lu, C. Bai, Q. Lu, S. Dai, C. Cai, Advances in the structure and materials of perovskite solar cells, Res. Chem.

Intermed. 42 (2015) 625e639.

[5] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, Organometal halide perovskites as visible-light sensitizers for photovoltaic cells, J. Am. Chem. Soc. 131 (2009) 6050e6051.

[6] W.S. Yang, B.-W. Park, E.H. Jung, N.J. Jeon, Y.C. Kim, D.U. Lee, S.S. Shin, J. Seo, E.K. Kim, J.H. Noh, Iodide management in formamidinium-lead-halideebased perovskite layers for efﬁcient solar cells, Science 356 (2017) 1376e1379.

[7] A. Yella, L.-P. Heiniger, P. Gao, M.K. Nazeeruddin, M. Gr€atzel, Nanocrystalline rutile electron extraction layer enables low-temperature solution processed perovskite photovoltaics with 13.7% efﬁciency, Nano Lett. 14 (2014) 2591e2596.

[8] J.W. Jung, S.T. Williams, A.K.Y. Jen, Low-temperature processed high- performanceﬂexible perovskite solar cells via rationally optimized solvent washing treatments, RSC Adv. 4 (2014) 62971e62977.

[9] A. Sharenko, M.F. Toney, Relationships between lead halide perovskite thin- ﬁlm fabrication, morphology, and performance in solar cells, J. Am. Chem.

Soc. 138 (2016) 463e470.

[10] L. Zheng, D. Zhang, Y. Ma, Z. Lu, Z. Chen, S. Wang, L. Xiao, Q. Gong, Morphology control of the perovskiteﬁlms for efﬁcient solar cells, Dalton Trans. 44 (2015) 10582e10593.

[11] T. Salim, S. Sun, Y. Abe, A. Krishna, A.C. Grimsdale, Y.M. Lam, Perovskite-based solar cells: impact of morphology and device architecture on device performance, J. Mater. Chem. 3 (2015) 8943e8969.

[12] D. Liu, L. Wu, C. Li, S. Ren, J. Zhang, W. Li, L. Feng, Controlling CH3NH3PbI(3- X)Cl(X)ﬁlm morphology with two-step annealing method for efﬁcient hybrid perovskite solar cells, ACS Appl. Mater. Interfaces 7 (2015) 16330e16337.

[13] Y. Wu, W. Chen, Y. Yue, J. Liu, E. Bi, X. Yang, A. Islam, L. Han, Consecutive morphology controlling operations for highly reproducible mesostructured perovskite solar cells, ACS Appl. Mater. Interfaces 7 (2015) 20707e20713.

[14] N. Ahn, D.Y. Son, I.H. Jang, S.M. Kang, M. Choi, N.G. Park, 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. 137 (2015) 8696e8699.

[15] N.J. Jeon, J.H. Noh, Y.C. Kim, W.S. Yang, S. Ryu, S.I. Seok, Solvent engineering for high-performance inorganic-organic hybrid perovskite solar cells, Nat. Mater.

13 (2014) 897e903.

[16] Y. Wu, X. Yang, W. Chen, Y. Yue, M. Cai, F. Xie, E. Bi, A. Islam, L. Han, Perovskite solar cells with 18.21% efﬁciency and area over 1 cm²fabricated by heterojunction engineering, Nat. Energy 1 (2016), 16148.

[17] F. Wang, A. Shimazaki, F. Yang, K. Kanahashi, K. Matsuki, Y. Miyauchi, T. Takenobu, A. Wakamiya, Y. Murata, K. Matsuda, Highly efﬁcient and stable perovskite solar cells by interfacial engineering using solution-processed polymer layer, J. Phys. Chem. C 121 (2017) 1562e1568.

[18] D. Bi, C. Yi, J. Luo, J.-D. Decoppet, F. Zhang, Shaik M. Zakeeruddin, X. Li, A. Hagfeldt, M. Gr€atzel, Polymer-templated nucleation and crystal growth of perovskiteﬁlms for solar cells with efﬁciency greater than 21%, Nat. Energy 1 (2016), 16142.

[19] X. Wen, J. Wu, M. Ye, D. Gao, C. Lin, Interface engineering via an insulating polymer for highly efﬁcient and environmentally stable perovskite solar cells, Chem. Commun. 52 (2016) 11355e11358.

[20] N. Tripathi, Y. Shirai, M. Yanagida, A. Karen, K. Miyano, Novel surface passivation technique for low-temperature solution-processed perovskite PV cells, ACS Appl. Mater. Interfaces 8 (2016) 4644e4650.

[21] K.M. Boopathi, R. Mohan, T.-Y. Huang, W. Budiawan, M.-Y. Lin, C.-H. Lee, K.- C. Ho, C.-W. Chu, Synergistic improvements in stability and performance of lead iodide perovskite solar cells incorporating salt additives, J. Mater. Chem.

(9)

4 (2016) 1591e1597.

[22] S. Masi, A. Rizzo, F. Aiello, F. Balzano, G. Uccello-Barretta, A. Listorti, G. Gigli, S. Colella, Multiscale morphology design of hybrid halide perovskites through a polymeric template, Nanoscale 7 (2015) 18956e18963.

[23] C.Y. Chang, C.Y. Chu, Y.C. Huang, C.W. Huang, S.Y. Chang, C.A. Chen, C.Y. Chao, W.F. Su, Tuning perovskite morphology by polymer additive for high efﬁ- ciency solar cell, ACS Appl. Mater. Interfaces 7 (2015) 4955e4961.

[24] Y. Zhao, J. Wei, H. Li, Y. Yan, W. Zhou, D. Yu, Q. Zhao, A polymer scaffold for self-healing perovskite solar cells, Nat. Commun. 7 (2016), 10228.

[25] Q. Dong, Z. Wang, K. Zhang, H. Yu, P. Huang, X. Liu, Y. Zhou, N. Chen, B. Song, Easily accessible polymer additives for tuning the crystal-growth of perovskite thin-ﬁlms for highly efﬁcient solar cells, Nanoscale 8 (2016) 5552e5558.

[26] S. Masi, S. Colella, A. Listorti, V. Roiati, A. Liscio, V. Palermo, A. Rizzo, G. Gigli, Growing perovskite into polymers for easy-processable optoelectronic devices, Sci Rep. UK 5 (2015) 7725.

[27] F. Bella, D. Pugliese, L. Zolin, C. Gerbaldi, Paper-based quasi-solid dye-sensitized solar cells, Electrochim. Acta 237 (2017) 87e93.

[28] Y. Zhang, H. Yi, A. Iraqi, J. Kingsley, A. Buckley, T. Wang, D.G. Lidzey, Comparative indoor and outdoor stability measurements of polymer based solar cells, Sci Rep. UK 7 (2017) 1305.

[29] F. Bella, S. Galliano, C. Gerbaldi, G. Viscardi, Cobalt-based electrolytes for dye- sensitized solar cells: recent advances towards stable devices, Energies 9 (2016) 384.

[30] D. Pintossi, G. Iannaccone, A. Colombo, F. Bella, M. V€alim€aki, K.-L. V€ais€anen, J. Hast, M. Levi, C. Gerbaldi, C. Dragonetti, S. Turri, G. Grifﬁni, Luminescent downshifting by photo-induced sol-gel hybrid coatings: accessing multi- functionality onﬂexible organic photovoltaics via ambient temperature material processing, Adv. Electron Mater. 2 (2016), 1600288.

[31] R. Shanti, F. Bella, Y.S. Salim, S.Y. Chee, S. Ramesh, K. Ramesh, Poly(methyl methacrylate-co-butyl acrylate-co-acrylic acid): physico-chemical character- ization and targeted dye sensitized solar cell application, Mater. Des. 108 (2016) 560e569.

[32] J.H. Im, C.R. Lee, J.W. Lee, S.W. Park, N.G. Park, 6.5% efﬁcient perovskite quantum-dot-sensitized solar cell, Nanoscale 3 (2011) 4088e4093.

[33] S.M. Sze, K.K. Ng, Physics of Semiconductor Devices, John wiley&sons, 2006.

[34] T.S. Gershon, A.K. Sigdel, A.T. Marin, M.F.A.M. van Hest, D.S. Ginley, R.H. Friend, J.L. MacManus-Driscoll, J.J. Berry, Improvedﬁll factors in solution- processed ZnO/Cu2O photovoltaics, Thin Solid Films 536 (2013) 280e285.

[35] J. Li, B. Huang, E. Nasr Esfahani, L. Wei, J. Yao, J. Zhao, W. Chen, Touching is believing: interrogating halide perovskite solar cells at the nanoscale via scanning probe microscopy, npj Quant. Mater. 2 (2017).

[36] B.P. Nguyen, G.Y. Kim, W. Jo, B.J. Kim, H.S. Jung, Trapping charges at grain boundaries and degradation of CH3NH3Pb(I1-xBrx)3 perovskite solar cells, Nanotechnology 28 (2017), 315402.

[37] M. Salado, R.K. Kokal, L. Calio, S. Kazim, M. Deepa, S. Ahmad, Identifying the

charge generation dynamics in Cs^þ-based triple cation mixed perovskite solar cells, Phys. Chem. Chem. Phys. 19 (2017) 22905e22914.

[38] Y.C. Shih, L.Y. Wang, H.C. Hsieh, K.F. Lin, Enhancing the photocurrent of perovskite solar cells via modiﬁcation of the TiO2/CH3NH3PbI3heterojunction interface with amino acid, J. Mater. Chem. 3 (2015) 9133e9136.

[39] A. Baumann, S. Vath, P. Rieder, M.C. Heiber, K. Tvingstedt, V. Dyakonov, Identiﬁcation of trap states in perovskite solar cells, J. Phys. Chem. Lett. 6 (2015) 2350e2354.

[40] X. Wu, M.T. Trinh, D. Niesner, H. Zhu, Z. Norman, J.S. Owen, O. Yaffe, B.J. Kudisch, X.Y. Zhu, Trap states in lead iodide perovskites, J. Am. Chem. Soc.

137 (2015) 2089e2096.

[41] G. Yin, J. Ma, H. Jiang, J. Li, D. Yang, F. Gao, J. Zeng, Z. Liu, S.F. Liu, Enhancing efﬁciency and stability of perovskite solar cells through Nb-Doping of TiO2 at low temperature, ACS Appl. Mater. Interfaces 9 (2017) 10752e10758.

[42] S.S. Mali, C.S. Shim, H. Kim, P.S. Patil, C.K. Hong, In situ processed gold nanoparticle-embedded TiO2nanoﬁbers enabling plasmonic perovskite solar cells to exceed 14% conversion efﬁciency, Nanoscale 8 (2016) 2664e2677.

[43] Y. Liu, X. Sun, Q. Tai, H. Hu, B. Chen, N. Huang, B. Sebo, X.-z. Zhao, Efficiency enhancement in dye-sensitized solar cells by interfacial modification of conducting glass/mesoporous TiO2using a novel ZnO compact blockingfilm, J. Power Sources 196 (2011) 475e481.

[44] Y.-J. Kim, D.-T. Tung, H.-J. Choi, B.-J. Park, J.-H. Eom, K.-S. Kim, J.-R. Jeong, S.- G. Yoon, Enhanced reproducibility of the high efﬁciency perovskite solar cells via a thermal treatment, RSC Adv. 5 (2015) 52571e52577.

[45] I. Hwang, M. Baek, K. Yong, Core/shell structured TiO2/CdS electrode to enhance the light stability of perovskite solar cells, ACS Appl. Mater. Interfaces 7 (2015) 27863e27870.

[46] W. Li, H. Dong, X. Guo, N. Li, J. Li, G. Niu, L. Wang, Graphene oxide as dual functional interface modiﬁer for improving wettability and retarding recombination in hybrid perovskite solar cells, J. Mater. Chem. 2 (2014) 20105e20111.

[47] M.A. Mahmud, N.K. Elumalai, M.B. Upama, D. Wang, M. Wright, K.H. Chan, C. Xu, F. Haque, A. Uddin, Single vs mixed organic cation for low temperature processed perovskite solar cells, Electrochim. Acta 222 (2016) 1510e1521.

[48] M. Yang, R. Guo, K. Kadel, Y. Liu, K. O'Shea, R. Bone, X. Wang, J. He, W. Li, Improved charge transport of Nb-doped TiO2nanorods in methylammonium lead iodide bromide perovskite solar cells, J. Mater. Chem. 2 (2014) 19616e19622.

[49] J. Zhang, Z. Hu, L. Huang, G. Yue, J. Liu, X. Lu, Z. Hu, M. Shang, L. Han, Y. Zhu, Bifunctional alkyl chain barriers for efﬁcient perovskite solar cells, Chem.

Commun. 51 (2015) 7047e7050.

[50] M. Park, J.-Y. Kim, H.J. Son, C.-H. Lee, S.S. Jang, M.J. Ko, Low-temperature solution-processed Li-doped SnO2as an effective electron transporting layer for high-performanceﬂexible and wearable perovskite solar cells, Nano En- ergy 26 (2016) 208e215.