Enhanced Performance of Perovskite Solar Cells by Using Ultrathin

Oct 1, 2018 - Efficiency promotion has been severely constrained by charge recombination in perovskite solar cells (PSCs). Interface modification has ...
1 downloads 0 Views 1MB Size
Subscriber access provided by Lancaster University Library

Energy, Environmental, and Catalysis Applications

Enhanced performance of perovskite solar cells by using ultrathin BaTiO3 interface modification Jianqiang Qin, Zhenlong Zhang, Wenjia Shi, Yuefeng Liu, Huiping Gao, and Yanli Mao ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 01 Oct 2018 Downloaded from http://pubs.acs.org on October 1, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Enhanced performance of perovskite solar cells by using ultrathin BaTiO3 interface modification a

a,*

Jianqiang Qin , Zhenlong Zhang , Wenjia Shi a, Yuefeng Liu a, Huiping Gao a, Yanli Mao a,b,* a b

School of Physics and Electronics, Henan University, Kaifeng 475004, China

Institute of Micro/Nano Photonic Materials and Applications, Henan University, Kaifeng 475004, China *

Corresponding author: +86-371-23881602; E-mail: [email protected]; [email protected]

Abstract Efficiency

promotion

has

been

severely

constrained

by

charge

recombination in perovskite solar cells (PSCs). Interface modification has been proved to be an effective way to reduce the interfacial charge recombination. In this work, mesoporous TiO2 (mp-TiO2) layer was modified by an ultrathin BaTiO3 layer to suppress charge recombination in PSCs. The ultrathin BaTiO3 modification layer was prepared by spin-coating method using barium salt solution. The concentration of barium salt solution was optimized, and the effect of BaTiO3 modification layer on the performance of the cells was also investigated. The modification layer can not only successfully retard charge recombination but also effectively boost the rate of electron extraction at interface, resulting in enhanced open-circuit voltage (Voc), short-circuit current density (Jsc), and fill factor (FF), respectively. Furthermore, the hysteresis of PSCs was also significantly reduced after modification. By optimizing and employing the BaTiO3 modification layer, the power conversion efficiency (PCE) of the cells was increased from 16.13% to 17.87%. Keywords: Perovskite solar cells; interface modification; BaTiO3; optimized concentration; spin-coating method

1. Introduction Since the first report in 2012, perovskite solar cells (PSCs) have attracted enormous interest due to its low-cost and high power conversion efficiency (PCE).1 In 1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

the past few years, the PCE of the PSCs have been rapidly enhanced from 9% to 22.1%.2 Research shows that organic-inorganic halide perovskite materials are excellent light absorbers for solar cell due to its large light absorption coefficient, ambipolar charge transport, and long carrier diffusion length.3 In general, high-performance PSCs use perovskite materials as light absorption layer located between electron transfer layer (ETL) and hole transfer layer (HTL). Typical ETL consist of compact TiO2 layer (c-TiO2) and mesoporous TiO2 layer (mp-TiO2) to collect and transport the electrons generated from perovskite layer.4 However, the electrons generated from perovskite layers may recombine with the holes at interface between mp-TiO2 and perovskite layers, resulting in reduced PCE. Similarly, dye-sensitized solar cells (DSSCs) are also confronted with the same problem. In DSSCs, interface modification has been proved to be one of the most effective ways to reduce the interfacial charge recombination.5-8 Recently, metal oxides are widely used to modify TiO2 layer to reduce charge recombination in PSCs, including MgO, Al2O3, La2O3, and In2O3.9-12 Meanwhile, some other materials have been employed as modification layer coated on TiO2 layer in PSCs, which can effectively enhance electron extraction rate and reduce interfacial recombination, such as ZnS, Cs2CO3, C60, and PCBA ([6,6]-phenyl-C61-butyric acid).13-16 Suzuki et al. have reported using 50-100nm BaTiO3 layer as the second ETL on TiO2 in PSCs, which was prepared by spin-coating the BaTiO3 paste.17 The results suggest that it is an effective way to suppress charge recombination. In addition, previous studies on DSSCs have shown that dipping mp-TiO2 in saturated barium nitrate solution can form a BaTiO3 modification layer,18 which can effectively suppress electrons transfer from TiO2 to I3-. Similarly, Tsujimoto et al. have used the same method to fabricate BaTiO3 modification layer in Sb2S3-based extremely thin absorber solar cells,19 resulting in enhanced photovoltaic performance. However, study on modifying mp-TiO2 with BaTiO3 has been rarely reported in PSCs. The effect of Ba2+ concentration on the performance of PSCs is also unclear. In this study, we coated barium nitrate solution on mp-TiO2 to synthesize an ultrathin BaTiO3 modification layer by spin-coating method. 2

ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

550°C Ba(NO3 )2 +TiO 2  → BaTiO3 + 2NO 2 + 1 2 O 2

(1)

The PSCs were fabricated using mp-TiO2 and mp-TiO2/BaTiO3 as scaffold layer, respectively. The performance of the PSCs was optimized by adjusting the concentration of the barium salt solution. Furthermore, the characteristics of the ultrathin BaTiO3 modification layer were investigated, and its effect on the performance of the PSCs was systematically discussed. By optimizing the BaTiO3 modification layer, a high PCE of 17.87% was obtained from the cells based on mp-TiO2/BaTiO3 (0.9 wt%). 2. Experimental 2.1 Materials The FTO glass substrates (~15Ω/Sq) were purchased from NSG (Japan). TiO2 paste was purchased from Dyesol (30NR-D) with an average particle size about 30nm. Lead (II) Iodide (PbI2, 99.99%) and lead (II) Bromide (PbBr2, 99.99%), Formamidinium Iodide (FAI, 99.5%), Methylammonium Bromide (MABr, 99.5%), and

2,2',7,7'-Tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9'-spiro-bifluorene

(spiro-OMeTAD) were obtained from Polymer Light Technology Corp (Xi'an, China). All the other materials were purchased from Aladdin (Shanghai, China), including acetonitrile (99.9%), N, N-Dimethylformamide (DMF, 99.9%), dimethyl sulfoxide (DMSO, 99.9%), chlorobenzene (99.5%), titanium butoxide (99%), barium nitrate (Ba(NO3)2, 99.99%). 2.2 Device fabrication Firstly, the FTO glass substrates were patterned, washed and UV-O3 treated in turn. Then, the compact TiO2 layer was fabricated on FTO glass with titanium precursor by spin-coating at 3500 rpm for 30 s, which is consist of 0.25 mL tetrabutyl titanate, 5 mL ethanol, 1 mL nitric acid, and 0.5 mL deionized water, followed by annealing at 500 ℃ for 30 min. Secondly, the mesoporous TiO2 layer was deposited on the compact layer by spin-coating using TiO2 paste (Dyesol 30 NR-D) diluted in ethanol (1:6 weight ratio) with a speed of 4000 rpm for 30 s. Substrates were then heated at 100 ℃ for 10 min and sintered at 500 ℃ for 30 min, respectively. Thirdly, the 3

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

substrates were modified by barium nitrate solution. Barium nitrate was pre-dissolved in deionized water with a series of concentrations (0.3, 0.6, 0.9, 1.2 and 2.4 wt%, respectively) by stirring overnight, resulting in transparent and homogeneous solution. The prepared solution was coated on the mesoporous TiO2 layer by spin-coating at a speed of 5000 rpm for 30 s. The films were annealed at 550 ℃ for 30 min in air. Subsequently, the perovskite films were formed using solvent-engineering method. The mixed-perovskite precursor solution was prepared by dissolving FAI (1 M), PbI2 (1.1 M), MABr (0.2 M) and PbBr2 (0.2 M) in a mixed solution of DMF and DMSO (4:1 volume ratio) in a glovebox. The resulting solution was deposited on the substrates by two-step spin-coating at 1000 rpm for 10 s and 4000 rpm for 30 s, respectively. Before the end of the second step, 100 µL chlorobenzene was poured on the substrates. Then, the substrates were immediately heated 100 ℃ for 1 h in glovebox. After cooling down to room temperature, the hole transport material (HTM) solution was spin-coated on the substrates at 4000 rpm for 30 s. The HTM solution was prepared by dissolving 72.3 mg spiro-MeOTAD in 1 mL chlorobenzene and doping with 17.5 µL lithium bis (trifluoromethanesulfonyl) imide (Li-TFSI) solution (520 mg Li-TFSI in 1 mL acetonitrile), 28.8 µL 4-tert-butylpyridine (TBP). Finally, 80 nm Au electrodes were successfully deposited on the substrates by thermal evaporation. 2.3 Characterization The morphology of the TiO2 films with and without modification was observed by field emission scanning electron microscope (FE-SEM, JEM-7001F, JEOL) and scanning probe microscope (Multimode 8, Bruker, America), respectively. The phase structure of the TiO2 with and without modification were analyzed by X-ray diffraction (XRD) from a diffractometer with Cu-Kα source (λ = 0.1542 nm) (D8-Advance, Bruker, Germany). X-ray photoelectron spectroscopy (XPS) of the samples were performed by photoelectron spectrometer (ESCALAB 250XI, Thermo Fisher Scienti, America) with Al-Kα (hν =1486.6 eV) and calibrated with C1s (284.8 eV). Current density-voltage (J-V) curves of cells were measured by using a source meter (Keithley 2440) under standard illumination (AM 1.5G, 100 mW cm-2) 4

ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

simulated by a solar simulator (Oriel, Newport). During test, the active area of the cells was defined by a shadow mask (0.1 cm2). The UV-vis absorption spectra were investigated by a UV-vis spectrophotometer (Cary 5000 UV-vis-NIR). The incident photon-to-current efficiency (IPCE) spectrum was examined by a solar cell IPCE measurement system (Crowntech Qtest Station 500ADX, America). Steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) spectra were obtained from FLS 980E fluorometer (Edinburgh Photonics). Electrochemical impedance spectroscopy (EIS) was tested by an electrochemical workstation (CHI660e, Shanghai CHI Co., Ltd) under one illumination, applying 10 mV perturbation amplitude in the frequency range of 0.1 Hz to 1M Hz at 0.7 V. 3. Results and discussion

Figure 1. Top-view SEM images of (a) FTO/c-TiO2/mp-TiO2 (without modification) and

(b) FTO/c-TiO2/mp-TiO2/BaTiO3. Cross-sectional SEM

images

of

(c)

FTO/c-TiO2/mp-TiO2 and (d) FTO/c-TiO2/mp-TiO2/BaTiO3. Figure 1 shows the top-view SEM images of mp-TiO2 with and without modification, respectively. Figure 1a exhibits the nanoparticulate morphology of 5

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

mp-TiO2 with an average particle size about 30 nm. In Figure 1b, the nanoparticulate morphology is almost invariant after modification, but the particles are slightly bigger than that without modification. In addition, atomic force microscope (AFM) measurements were also carried out. As presented in Figure S1, the morphology of two samples has no significant difference, which is in good agreement with the SEM results. The root mean square (RMS) roughness values of mp-TiO2 with and without modification on a 1 µm ×1 µm scale are listed in Table S1. Obviously, the RMS increases a little from 5.568 to 6.067 nm after modification, which may arise from the formation of BaTiO3. It can ensure that the perovakite precursor solution can well adhere to the ETL. Figure 1c and d display the cross-sectional SEM images of FTO/c-TiO2/mp-TiO2 and FTO/c-TiO2/mp-TiO2/BaTiO3, respectively. The thickness of mp-TiO2 is about 230 nm. The thickness of the sample has no quite different after modification, which indicates that the modification layer is ultrathin.

Figure 2. (a) XRD patterns of pure TiO2 and TiO2/BaTiO3. (b) Details of the XRD patterns for (101) plane. XPS spectra of (c) Ti2p peaks, (d) Ba3d peaks of pure TiO2 and TiO2/BaTiO3. 6

ACS Paragon Plus Environment

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

It is difficult to get XRD patterns of TiO2 and TiO2/BaTiO3 films on FTO substrate due to the strong peak intensity of FTO and the thin thickness of TiO2 and TiO2/BaTiO3 films. In this work, the XRD patterns were obtained from TiO2 and TiO2/BaTiO3 powders. The TiO2 powders referred as TiO2-500 was prepared by drying the TiO2 paste and annealing at 500 ℃ for 30 min, and TiO2/BaTiO3 powders was prepared by treating the TiO2-500 powder with Ba(NO3)2 aqueous solution with the optimized concentration of 0.9 wt% and annealing at 550 ℃ for 30 min. Figure 2a shows the XRD patterns of the TiO2 and TiO2/BaTiO3 powders. The TiO2 exhibits a series of diffraction peaks at 25.3°, 37.8°, 48.0°, 53.9° and 55.1°, which are assigned to the (101), (004), (200), (105) and (211) planes, respectively, confirming the formation of anatase phase (JCPDS card no.21-1272).20 Intriguingly, XRD pattern of TiO2/BaTiO3 shows an extra peak at 2θ=31.5°, which corresponds to the characteristic peak of cubic structure BaTiO3.21,22 No more peaks of BaTiO3 can be detected due to its ultrathin thickness of the film. As shown in Figure 2b, the diffraction peak of TiO2/BaTiO3 shifts to a higher degree about 0.2° compared with the pure TiO2. In order to explain the shift, a sample referred as TiO2-550 was also prepared by annealing the TiO2-500 powders at 550 ℃ for 30 min, which is the same as the condition of TiO2/BaTiO3 preparation without treating with Ba(NO3)2 aqueous solution. The XRD patterns of TiO2-500, TiO2-550, and TiO2/BaTiO3 powders are presented in Figure S2. There is no peak shift in the XRD pattern of TiO2-550 compared with that of TiO2-500. While a peak shift was observed in the XRD pattern of TiO2/BaTiO3 compared with that of TiO2-550. The results indicate that the peak shift could be due to the effect of BaTiO3 formation.23 To further clarify the chemical compositions of TiO2 powder with and without modification, XPS measurements were carried out. Figure S3 displays the XPS survey spectra of TiO2 with and without modification. The extra Ba peak can be clearly observed, indicating the existence of Ba element in the sample of TiO2/BaTiO3. As shown in Figure 2c, both of the samples reveal the Ti2p peaks at 458.5 and 464.3 eV, which are assigned to Ti 2p3/2 and Ti 2p1/2, respectively. Figure 2d exhibits the XPS spectra of Ba peaks for two samples. The modified TiO2 displays two peaks at binding 7

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

energies of 780.1 and 795.4 eV, which correspond to Ba 3d5/2 and Ba 3d3/2, respectively. This result is consistent with previous literature about BaTiO3.24 As discussed above, the XRD and XPS results can confirm that an ultrathin BaTiO3 film has been successfully coated on the TiO2 nanoparticles surface.

Figure 3. Photovoltaic parameters of the devices with varying Ba2+ concentration (a)Voc, (b) Jsc, (c) PCE, and (d) FF. To optimize the performance of the cells, mp-TiO2 modified with different concentrations barium nitrate solution varying from 0 to 2.4 wt%, labeled as mp-TiO2 and mp-TiO2/BaTiO3 (0.3, 0.6, 0.9, 1.2 and 2.4 wt%), respectively. Figure 3 shows the photovoltaic parameters of the devices plotted as a function of the Ba2+ concentration, including Voc, Jsc, PCE, and FF, respectively, which were obtained from the J-V curves measured under AM 1.5G shown in Figure S3. Furthermore, the average photovoltaic parameters of the devices are also summarized in Table 1. Obviously, the photovoltaic performance of the devices is strongly influenced by the content of Ba2+. The devices based on mp-TiO2 exhibit an average Voc of 1.06±0.02 V, a Jsc of 21.88±0.34 mA cm-2, and a FF of 69.30±1.45%, yielding an average PCE of 16.13±0.64%. With the increase of Ba2+ content, Voc gradually increase until the 8

ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

content of Ba2+ reach 0.9 wt % and then almost keep constant. Interestingly, Jsc and FF exhibit the same tendency. As the concentration of Ba2+ increase, Jsc and FF increase firstly and then gradually decrease, which reach the maximum value at 0.9 wt%. Consequently, the champion PCE is obtained from the cells based on mp-TiO2/BaTiO3 (0.9 wt%), with an average PCE of 17.87±0.61%, a Voc of 1.10±0.01 V, a Jsc of 22.52±0.28 mA cm-2, and a FF of 71.73±1.05%, respectively.

Table 1. Photovoltaic parametersa of mp-TiO2/BaTiO3 based perovskite solar cells with varying Ba2+ concentrations (0, 0.3, 0.6, 0.9, 1.2 and 2.4 wt%) Sample

Jsc (mA cm-2)

Voc (V)

Control device (mp-TiO2)

21.88±0.34

1.06±0.02

16.13±0.64 69.30±1.45

mp-TiO2/BaTiO3 (0.3 wt%)

22.29±0.28

1.09±0.02

17.21±0.62 69.94±1.33

mp-TiO2/BaTiO3 (0.6 wt%)

22.43±0.36

1.10±0.02

17.53±0.39 71.05±1.07

mp-TiO2/BaTiO3 (0.9 wt%)

22.52±0.28

1.10±0.01

17.87±0.61 71.73±1.05

mp-TiO2/BaTiO3 (1.2 wt%)

22.41±0.35

1.10±0.01

17.41±0.56 70.49±0.99

mp-TiO2/BaTiO3 (2.4 wt%)

21.62±0.24

1.10±0.01

16.36±0.53 68.55±1.29

a

PCE (%)

FF (%)

Average values and standard deviation were obtained from 18 devices (3 batches)

measured under AM 1.5G (100 mW cm-2).

9

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

Figure 4. UV-vis absorption spectra of (a) mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%), (b) perovskite deposited on mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%). (c) Conductivity measurements of ETL with and without BaTiO3 modification. (d) Plots of -dV/dJ vs (Jsc-J)-1 derived from the illuminated J-V curves and the fitting line. Figure 4a shows the UV-vis absorption spectra of the mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%). Notably, both of the samples display low absorption in the range of 350 to 800 nm, allowing most of the lights deliver to light harvesting layer.25 Moreover, the absorption of the sample increases slightly after modification. Figure 4b exhibits the UV-vis absorption spectra of perovskite deposited on the mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%), respectively. The perovskite layers on both substrates reveal almost identical and wide absorbance in visible region due to the transparent and thin modification layer, indicating the similar light harvesting ability. To evaluate the effect on conductivity of ETL caused by the introduction of BaTiO3 modification layer and concentration of Ba2+, DC I-V measurement was performed and shown in Figure 4c. The inset figure depicts the structure of devices 10

ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

composed of FTO/c-TiO2/mp-TiO2/BaTiO3/Au. It is apparent that the resistivity of the ETL changes varying with the Ba2+ concentration. The sample based on mp-TiO2/BaTiO3 (0.9 wt %) exhibits the highest conductivity among these samples. According to early literatures, the J-V characteristic of the solar cells can be described as an equivalent circuit and Equation 2.26,27  e(V + JRs )   V + JRs  J = J L − J o exp   − 1 − Rsh  AK BT   

(2)

Where are JL, J0, A, KB, T, Rs, and Rsh are photocurrent density, reverse saturated current density, ideality factor, Boltzmann constant, absolute temperature, series resistance, and shunt resistance, respectively. Hence, Rs of the cells can be obtained from Equation 3.28 −

dV AK BT = ( J sc − J ) −1 + Rs dJ e

(3)

As shown in Figure 4d, the cells based on mp-TiO2/BaTiO3 (0.9 wt%) show a lower Rs of 2.1 Ω cm2 compared with that of without modification (3.9 Ω cm2), which is consistent with the results of resistivity measurement. A lower Rs, as is reported in early study, corresponds to a higher FF.28,29 Apparently, the cells based on mp-TiO2/BaTiO3 (0.9 wt%) have a lower Rs, which can account for the higher FF listed in Table 1.

Figure 5. (a) Steady-state photoluminescence spectra of glass/TiO2/perovskite and glass/TiO2/BaTiO3 (0.9 wt%)/perovskite, (b) time-resolved photoluminescence spectra of glass/TiO2/perovskite and glass/TiO2/BaTiO3 (0.9 wt%)/perovskite. To study the charge transfer kinetics at the ETL/perovskite interface, steady-state 11

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

photoluminescence (PL) and time-resolved photoluminescence (TRPL) measurements were performed.31-33 Figure 5a shows the normalized PL spectra of the sample of mp-TiO2/perovskite and mp-TiO2/BaTiO3 (0.9 wt%)/perovskite deposited on blank glass. Furthermore,PL spectra of pristine mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%) were also measured shown in Figure S5. It can be observed that two samples exhibit strong photoluminescence peak at 770nm after perovskite deposition, which can coincide with the previous reports.30 The sample of mp-TiO2/BaTiO3 (0.9 wt%)/perovskite displays more drastic PL quenching than that of without modification, which indicate the electrons could be more efficiently extracted from perovsktie layer to ETL.31 Figure 5b exhibits the TRPL spectra the samples of mp-TiO2/perovskite and mp-TiO2/BaTiO3 (0.9 wt %)/perovskite, which can be fitted with a biexponential decay function (Equation 4),

I (t ) = A1 exp(- τt1 ) + A2 exp(- τt2 )

(4)

where τ1, τ2 are the fast decay and slow decay, respectively. The fast decay could be associated with the quenching of free carriers transfer from perovskite to electron or hole contact. While, the slow decay would be related to the radiative recombination of the charge carries before the charge collection, respectively.31,32 The fitted parameters are summarized in Table 2. After modification, the fast and slow decay of the sample decrease from 6.01 to 2.26 ns, 30.6 to 26.1 ns, respectively, and the fraction of fast decay increases from 19.87% to 25.99%. This result implies the electron injection from perovskite layer to ETL of the sample mp-TiO2/BaTiO3 (0.9 wt %)/perovskite is faster than that of mp-TiO2/perovskite.33,34

Table 2. Parameters of the time-resolved photoluminescence (TRPL) spectra ETL

τ1 (ns)

Fraction 1 (%)

τ2 (ns)

Fraction 2 (%)

TiO2

6.01

19.87

30.6

80.13

TiO2/BaTiO3 (0.9 wt%)

2.26

25.99

26.1

74.01

12

ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 6. (a) EIS spectra of the control device and optimized device measured under AM 1.5G illumination at 0.7 V (the inset is the fitted equivalent circuit), (b) recombination resistance (Rrec) of the devices measured at different biases. To gain further insight into carrier transport behavior at perovskite interface, electrochemical impedance spectroscopy (EIS) measurements were carried out.36-38 Figure 6a shows the Nyquist plots of the devices based on mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%) measured under AM 1.5G illumination. The semicircle at mid-frequency region is related to the charge transfer and recombination at the heterojunction interface in PSCs.36,37 The Nyquist plots can be fitted with the simplified equivalent circuit shown in the inset of Figure 6a, including a series resistance (Rs), recombination resistance (Rrec), and resistance-constant phase element (CPE), respectively.35 The sample based on mp-TiO2/BaTiO3 (0.9 wt%) displays a larger semicircle which correspond to a larger Rrec than that of mp-TiO2, indicating less carrier recombination in the device.35-37 Figure 6b exhibits the Rrec plotted as a function of applied voltage. The value of Rrec gradually decreases with the increase of applied voltage, agreeing well with previous report.37,38 Meanwhile, the Rrec of the cells based on mp-TiO2/BaTiO3 (0.9 wt%) is notably larger than that of mp-TiO2 at different bias. It can be concluded that charge recombination of the cells treated with BaTiO3 modification layer has been effectively retarded.

13

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

Figure 7. Schematic energy diagram of the materials in the cells. To better understand the charge transfer and recombination in the devices discussed above, the energy band diagram of the materials in the cells are shown in Figure 7. According to the literature report, BaTiO3 is a semiconductor material with a band gap energy about 3.23 eV. The conduction band minimum of BaTiO3 is located at -3.82 eV, which is slightly higher than that of TiO2 (-4.05 eV) and lower than that of mix-perovskite (-3.79 eV).39,40 On the one hand, the band offset between TiO2 and BaTiO3 can help the photogenerated electrons transfer from perovskite to ETL. On the other hand, the introduction of BaTiO3 modification layer can successfully avoid the direct contact between TiO2 and the perovskite layer or HTL. The charge recombination can be effectively retarded at the TiO2/BaTiO3 or TiO2/HTL interface. Thus, the performance of the cells is remarkably improved.

Figure 8. (a) J-V curves and (b) IPCE spectra of the best performing cells based on mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%). 14

ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 8 exhibits the J-V curves and IPCE spectra of the best performing cells based on mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%), respectively. As shown in Figure 8a, J-V measurements were carried out by reverse and forward scan under AM 1.5G illumination. Obviously, the curve of the optimized device reveals lower hysteresis between reverse and forward scan than the control device. Previous studies show that the hysteresis could be associated with various processes, including ferroelectricity, ions rotation, carriers trap and charge accumulation at the perovskite interfaces, respectively.41,42 As discussed above, the optimized device exhibits efficient charge extraction and reduced charge accumulation at perovskite interfaces, which can account for the lower hysteresis. Figure 8b exhibits the IPCE spectra of the control device and optimized device. The integrated Jsc calculated from IPCE for the device based on mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%) are 21.1 mA/cm2 and 22.0 mA/cm2, respectively, which are close to the measured Jsc. Compared with the control device, the device based on mp-TiO2/BaTiO3 (0.9 wt%) reveals higher IPCE in the range of 450 to 750 nm. As is reported, IPCE depends on light harvesting efficiency, electron injection efficiency, and charge collection efficiency.43, 44 However, mp-TiO2 and mp-TiO2/BaTiO3 (0.9 wt%) show almost equal absorbance in visible wavelength range shown in Figure 4b. Therefore, the significant enchantment should be ascribed to electron injection and charge collection efficiency. Clearly, the device based on mp-TiO2/BaTiO3 (0.9 wt%) exhibits higher charge extraction and lower charge recombination than that of mp-TiO2, which can be demonstrated by the data of PL, TRPL, and EIS, respectively. Interestingly, the enhancement is mainly in long wavelength region after 450 nm. This could be due to the easy trapping for the lower-energy electrons prior to extraction when there is no the modification layer.45 When the modification layer is inserted, the trapping events could be reduced, which is demonstrated by the PL spectra of TiO2/perovskite and TiO2/BaTiO3/perovskite as shown in Figure 5a. Finally, we examined the stability of the devices in ambient atmosphere (temperature: 30 °C, humidity: