Inverted Planar Perovskite Solar Cells with a High Fill Factor and

Dec 6, 2017 - Support. Get Help · For Advertisers · Institutional Sales; Live Chat. Partners. Atypon · CHORUS · COPE · COUNTER · CrossRef · CrossCheck...
1 downloads 0 Views 4MB Size
Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

www.acsami.org

Inverted Planar Perovskite Solar Cells with a High Fill Factor and Negligible Hysteresis by the Dual Effect of NaCl-Doped PEDOT:PSS Lijun Hu,† Kuan Sun,*,† Ming Wang,‡ Wei Chen,† Bo Yang,‡ Jiehao Fu,† Zhuang Xiong,† Xinyi Li,† Xiaosheng Tang,‡ Zhigang Zang,‡ Shupeng Zhang,§ Lidong Sun,∥ and Meng Li*,† †

Key Laboratory of Low-Grade Energy Utilization Technologies and Systems, Ministry of Education, School of Power Engineering, Chongqing University, Chongqing 400044, China ‡ Key Laboratory of Optoelectronic Technology & Systems, Ministry of Education, Chongqing 400044, China § School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing, Jiangsu 210094, China ∥ School of Materials Science and Engineering, Chongqing University, Chongqing 400044, PR China S Supporting Information *

ABSTRACT: The performance of inverted perovskite solar cells is highly dependent on hole extraction and surface properties of hole transport layers. To highlight the important role of hole transport layers, a facile and simple method is developed by adding sodium chloride (NaCl) into poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS). The average power conversion efficiency of the perovskite solar cells prepared on NaCldoped PEDOT:PSS is 17.1% with negligible hysteresis, compared favorably to the control devices (15.1%). Particularly, they exhibit markedly improved Voc and fill factor (FF), with the best FF as high as 81.9%. The enhancement of photovoltaic performance is ascribed to two effects. Better conductivity and hole extraction of PEDOT:PSS are observed after NaCl doping. More intriguingly, the perovskite polycrystalline film shows a preferred orientation along the (001) direction on NaCl-doped PEDOT:PSS, leading to a more uniform thin film. The comparison of the crystal structure between NaCl and MAPbCl3 indicates a lattice constant mismatch less than 2% and a matched chlorine atom arrangement on the (001) surface, which implies that the NaCl crystallites on the top surface of PEDOT:PSS might serve as seeds guiding the growth of perovskite crystals. This simple method is fully compatible with printing technologies to mass-produce perovskite solar cells with high efficiency and tunable crystal orientations. KEYWORDS: perovskite, PEDOT:PSS, NaCl doping, crystal orientation, perovskite solar cells



INTRODUCTION

Despite that world record, PSCs usually adopt n−i−p device configurations, also known as normal architecture, and the inverted PSCs have a simpler architecture allowing for low temperature processing, and thus they are closer to commercialization.20−23 In general, surface properties such as roughness and composition of hole transport layers (HTLs) in inverted PSCs can have a great impact on crystal growth as well as on the crystal orientation of the perovskite layer by influencing nucleation at an early stage of crystal formation.7,23 Among the many reported hole transport materials, poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is the most widely used hole transport material in inverted PSCs because of its excellent optical transparency, high thermal stability, good mechanical flexibility, and suitable energy levels for organic−inorganic hybrid lead halide PSCs.24−31 However, the surface properties of PEDOT:PSS

Since methylammonium lead iodide (CH3NH3PbI3) was first introduced as a light harvesting material by Miyasaka in 2009, the organic−inorganic hybrid lead halide perovskite solar cells (PSCs) have attracted great attention.1−5 In general, the power conversion efficiency (PCE) of the PSCs are highly dependent on the active layer film morphology, crystal structure, and quality, including grain size and distribution, surface coverage, and crystal orientation of perovskite films.6−8 Besides, the defects can also affect the cell performance, especially on open circuit voltage (Voc) and fill factor (FF), by acting as nonradiation recombination centers.9−11 To circumvent these issues, great effort has been paid, such as additive engineering, postsolvent annealing or control of grain nucleation and growth, aiming to produce a uniform and crystalline perovskite layer with full coverage and fewer grain boundaries.12−18 Up to now, considerable development has been made, and a certified PCE of 22.1% was achieved in the laboratory through decreasing the concentration of deep-level defects.19 © XXXX American Chemical Society

Received: September 26, 2017 Accepted: November 27, 2017

A

DOI: 10.1021/acsami.7b14592 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. (a) Schematics of the cell architecture and (b) corresponding energy diagram of each layer of MAPbI3−xClx PSCs. (c) Typical J−V curves of the cells prepared with pristine (black) or NaCl-doped (red) PEDOT:PSS, which were measured under a forward scan from −0.2 to 1.3 V under 1 sun illumination. (d) Forward and reverse scans of the NC-5 PEDOT:PSS-based PSC. Then, the substrates were moved into a glovebox filled with dried nitrogen (O2 < 10 ppm; H2O < 1 ppm). A perovskite precursor solution with a composition of 0.14 M (39 mg) PbCl2, 1.26 M (581 mg) PbI2, and 1.3 M (209 mg) MAI in 1 mL co-solvent of DMSO/ GBL (3:7 volume ratio) was spin-coated on the PEDOT:PSS layer. Specifically, the substrate and the precursor were maintained at 70 °C before spin-coating at 1000 rpm for 20 s and then 3500 rpm for 40 s. One milliliter of chlorobenzene was injected onto the spinning film at 15 s before the end of the spin-coating programme.32 Subsequently, the obtained perovskite films were put in low vacuum (0.2 bar) for 15 s and then annealed at 100 °C for 20 min. PCBM with a concentration of 20 mg/mL was deposited on the top of the perovskite layer by spincoating at 2500 rpm for 40 s. After that, 0.6 mg/mL RhB101 was spincoated onto the PCBM layer at 1500 rpm for 40 s.33,34 The films were then transferred to a thermal evaporation chamber at approximately 1 × 10−6 mbar, where 1.3 nm LiF and 100 nm Ag were deposited through a shadow mask with the active area of 0.11 cm2. Devices Characterizations. X-ray diffraction (XRD) patterns were acquired using a PANalytical Empyrean diffractometer equipped with Cu Kα radiation (λ = 1.5406 Å). Transmission of a HTL and optical absorption spectra of an active layer were collected by a Shimadzu UV-1800 UV−vis spectrophotometer. X-ray photoelectron spectroscopy (XPS) was measured by an Axis Ultra DLD X-ray photoelectron spectrometer with an Al Kα X-ray source (1486.6 eV). The steady state photoluminescence (PL) spectra were obtained by a fluorescence spectrophotometer (Cary Eclipse, Agilent) with an excitation wavelength of 530 nm. Scanning electron microscopic (SEM) images were measured by a JEOL JSM-7800F field emission scanning electron microscope. Photocurrent−voltage (J−V) curves were achieved with an AM 1.5G solar simulator (Newport, 2612A) which calibrated by a silicon reference cell under a light intensity of 100 mW/cm2.

and its influence on perovskite growth and PSC performances are often ignored. In this work, we present a facile method to obtain highly efficient inverted PSCs with an impressive FF and Voc by doping sodium chloride (NaCl) into PEDOT:PSS solution (abbreviated as NaCl-PEDOT:PSS). Studies show that NaCl not only influences the electrical properties of PEDOT:PSS but also impacts the nucleation and growth of perovskites on top of PEDOT:PSS, resulting in an uniform MAPbI3−xClx perovskite layer with a slightly different crystal orientation. The average PCE increased from 15.1% of control samples with pristine PEDOT:PSS to 17.1% of PSCs with NaCl-doped PEDOT:PSS. The champion cell achieved a PCE of 18.2%, with a Voc of 1.08 V and an FF of 80.0%.



EXPERIMENTAL SECTION

Materials. PEDOT:PSS aqueous solution (Clevios PVP Al 4083) was purchased from Heraeus. The concentration of PEDOT:PSS was 1.3−1.7% by weight, and the weight ratio of PSS to PEDOT was 6. Methylammonium iodide (MAI, >99.5% purity) and phenyl-C61butyric acid methyl ester (PCBM, 99.5% purity) were acquired from Xi’an Polymer Light Technology Corp. PbI2 (>99.99% purity), PbCl2 (>99.99% purity), and rhodamine 101 (RhB101) were supplied by Sigma-Aldrich. All other chemicals, including NaCl (99% purity), dimethyl sulphoxide (DMSO, >99.9% purity), isopropanol (99.7% purity), and γ-butyrolactone (GBL, >99% purity) were supplied by Admas. All reagents were used directly without further purification. Device Fabrication. Inverted PSCs were fabricated on indium tin oxide (ITO)-coated glass substrates (10 Ω/sq, AE Tech.). The substrates were cleaned sequentially with detergent, de-ionized water, ethyl alcohol, acetone, and isopropyl alcohol. They were dried in nitrogen flow and treated with UV ozone for 30 min. The HTL was fabricated on the ITO substrates by spin-coating the PEDOT:PSS aqueous solution with or without NaCl doping at 8000 rpm for 40 s, followed by thermal annealing on a hot plate at 120 °C for 10 min.



RESULTS AND DISCUSSION Cell architecture of this study was illustrated schematically in Figure 1a. Different functional layers were stacked in the following configuration: ITO/PEDOT:PSS or NaCl-PEB

DOI: 10.1021/acsami.7b14592 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

increase of the FF and Voc. In a wide concentration range between 1 and 10 mg/mL, an enhanced average PCE was observed. However, further increasing the NaCl doping concentration in PEDOT:PSS led to the deterioration of photovoltaic performance; for example, an average PCE of only 9.8% was obtained when the doping concentration reached 100 mg/mL, which was due to hindered charge transport by the insulating NaCl. To understand the underlying mechanism for the observed photovoltaic performance, XPS was first utilized to probe the presence of NaCl in the PEDOT:PSS films. As depicted in Figure 2a, a clear signal between 196 and 202 eV was observed in the NC-5 PEDOT:PSS sample but was missing in the pristine PEDOT:PSS film. This doublet peak is associated with Cl 2p orbitals,35 suggesting the successful doping of NaCl into the PEDOT:PSS films. The work function values were obtained by ultraviolet photoelectron spectroscopy (UPS) measurements to evaluate the changes in the band alignment before and after doping. The work function shifts from 5.0 eV of pristine PEDOT:PSS to 5.2 eV of NaCl-doped PEDOT:PSS as depicted in Figure S2. The increased work function ensures a better Ohmic contact between the PEDOT:PSS HTL and the perovskite, resulting in more efficient hole extraction and a higher Voc.36 It should be noted that low concentration NaCl doping does not affect the optical properties of the PEDOT:PSS films. As illustrated in Figure 2b, both pristine and NC-5 PEDOT:PSS films exhibited high transmittance and indistinguishable spectrum profiles in the visible light range. Excellent transmittance of the HTL indicated a small optical loss in the front side of the inverted PSC, ensuring more photons to be absorbed by the active layer. However, the transmittance decreased when the doping concentration increased beyond 5 mg/mL, and it eventually reached 79.3% at the maximum doping concentration of 100 mg/mL (Figure S3). The diminished transmittance of PEDOT:PSS at a high NaCl doping concentration explains the low Jsc observed in PSCs with NC-10 and NC-100 PEDOT:PSS HTLs (Table 1). The morphology of PEDOT:PSS with and without NaCl doping was investigated. Atomic force microscopy (AFM) images (Figure S4) demonstrate both PEDOT:PSS films show very uniform and smooth surface morphology; the roughness was only 0.97 and 1.05 nm for the films with and without NaCl doping, respectively. Besides, the NaCl-doped film showed a stronger contrast in its phase image. This might be due to the presence of NaCl on the surface, which is much harder than the

DOT:PSS/MAPbI3−xClx/PCBM/RhB101/LiF/Ag. The PEDOT:PSS HTL was doped by different concentrations of NaCl, which was denoted as NC-x, where x is the doping concentration with a unit of mg/mL. The corresponding energy diagram of the devices was presented in Figure 1b. The energylevel alignment ensured that the photo-generated electrons and holes could be transported effectively in the device. Figure 1c described the current density−voltage (J−V) curves of the best control device and the optimized NaCl-PEDOT:PSS-based PSCs under standard simulated solar radiation (AM 1.5G, 100 mW/cm2). As for the control device with pristine PEDOT:PSS as the HTL, the highest PCE reached 16.4% with a Voc of 1.01 V, an FF of 75.5%, and a short circuit current density (Jsc) of 21.5 mA/cm2. Device performance was improved when using NaCl-doped PEDOT:PSS as the HTL, showing a PCE of 18.2% with a comparable Jsc of 21.1 mA/cm2 but a much higher Voc and FF of 1.08 V and 80.0%, respectively. It is worth noting that an FF as high as 81.9% was observed in one of those devices, as shown in Figure S1, suggesting that the charge transport in our optimized devices was well-balanced. Besides an improved PCE performance, another important PSC evaluation parameter, hysteresis, is almost negligible. As shown in Figure 1d, a PCE of 18.2% under a reverse scan and 17.9% under a forward scan were recorded. Statistical analysis of photovoltaic performance with different NaCl doping concentrations was obtained from more than 10 cells and is summarized in Table 1. It showed a clear evolution of Table 1. Photovoltaic Performance of MAPbI3−xClx Perovskite Solar Cells Prepared on NaCl-Doped PEDOT:PSS Films of Various Concentrations (mg/mL)a PSCs

Voc [V]

Jsc [mA/cm2]

FF [%]

control NC-1 NC-5 NC-10 NC-50 NC-100

1.01/1.02 1.11/1.04 1.08/1.06 1.07/1.05 0.99/1.03 1.09/1.03

21.5/20.1 19.6/20.4 21.1/20.5 21.4/20.2 20.6/19.0 17.0/15.2

75.5/74.2 79.6/78.0 80.0/79.2 77.3/75.3 75.7/72.6 65.8/63.5

PCE [%] 16.4/15.1 17.4/16.5 18.2/17.1 17.7/16.0 15.4/14.1 12.2/9.8

± ± ± ± ± ±

1.9 0.6 0.6 1.2 1.0 1.1

a

The former value is the best result, whereas the latter value is the average result. Statistical analysis is based on 10 devices.

photovoltaic performance with the increment of the doping concentration; for example, the average PCE increased from 15.1 to 17.1% with the doping concentration increasing from 0 to 5 mg/mL. The PCE improvement was mainly due to the

Figure 2. (a) High-resolution XPS profiles for the Cl element in the PEDOT:PSS films. (b) Transmittance of PEDOT:PSS films (without substrate) prepared w/o 5 mg/mL NaCl. C

DOI: 10.1021/acsami.7b14592 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. (a) J−V curves of devices with the structure of ITO/PEDOT:PSS or NaCl-PEDOT:PSS/Ag. (b) Time-resolved PL spectra of perovskite films prepared on PEDOT:PSS or NaCl-PEDOT:PSS.

Figure 4. (a) XRD diffraction pattern of the perovskite film prepared on the PEDOT:PSS or NaCl-PEDOT:PSS film. (b,c) Zoom-in images of XRD spectra in the range of 13.8°−14.7° and 28°−29°, respectively.

XPS was performed to understand the improvement in conductivity and hole extraction of NaCl-PEDOT:PSS from a molecular perspective. Figure S6 presented the XPS S 2p fine scans of a pristine PEDOT:PSS film and a NaCl-PEDOT:PSS film. The XPS signal in the range of 166−172 eV can be assigned to sulfur atoms of PSS, whereas the doublet peaks between 162 and 166 eV belong to the sulfur atoms of PEDOT.39−41 It is interesting to observe that NaCl doping results in an obvious red shift of the PEDOT S 2p peak. A similar peak shift was reported in H2PtCl6-doped PEDOT:PSS films.42 Such a peak shift implies that the chemical environment of PEDOT is slightly changed after the addition of NaCl. PEDOT chains might not be wrapped by insulating PSS chains, resulting in better charge hopping between conductive PEDOT chains, and thus the improved conductivity and hole extraction capability of PEDOT:PSS. To examine the effect of PEDOT:PSS on the crystal formation of perovskites, XRD was carried out on the perovskite films deposited on pristine or doped PEDOT:PSS films. As shown in Figure 4a, except for the diffraction peak at 12.6° that belongs to the (001) face of PbI2, all other diffraction peaks can be assigned to the typical MAPbI3−xClx phase; for example, the peaks at 14.2°, 28.5°, and 31.9° are assigned to (110), (220), and (310) crystal planes of the perovskite phase, respectively.43,44 The excess PbI2 in the perovskite film is served as a defect passivation material located at the grain boundary and the interface.45 Interestingly, differences appear when more details are revealed in the 2θ ranges of 13.8°−14.8° and 28.0°−29.0° (Figure 4b,c). It was evident that a splitting of diffraction peaks in these ranges occurred when the perovskite film was prepared

PEDOT:PSS polymers. Moreover, the NaCl-doped PEDOT:PSS exhibited nanofiber-like structures, implying that phase separation between PEDOT and PSS might have occurred.37 The hole transport property of the PEDOT:PSS HTL with or without NaCl doping was compared. The pristine or doped PEDOT:PSS films of the same thickness were sandwiched between an ITO and an Ag electrode, and then J−V curves were recorded. Figure 3a shows that the NaCl-doped PEDOT:PSS leads to a higher current density than pristine PEDOT:PSS under the same bias, indicating a better hole transport property of the NaCl-PEDOT:PSS than the pristine one. For a quantitative comparison, the electrical conductivity of the PEDOT:PSS films with different doping concentrations was evaluated. As depicted in Figure S5, NaCl doping led to enhancement of PEDOT:PSS film conductivity, which was peaked at the 5 mg/mL doping concentration. This was in accordance with the PCE performance. The hole extraction capability of HTLs was further assessed by time-resolved PL decay spectra as shown in Figure 3b. The curves were fitted by the biexponential function I(t ) = I0 + A1 exp(−t − t0/ τ1) +A 2 exp(−t − t0/ τ2) to calculate the average decay time (τ avg ), which was obtained from

τavg =

∑ A i τi 2 , ∑ A i τi

where τi was the decay time and Ai is the weight

factor of each decay channel.38 The extracted parameters are summarized in Table S1. Notably, NC-5 PEDOT:PSS showed a τavg of 17.4 ns, which is smaller than 20.3 ns of the control sample, suggesting a faster charge transfer from the perovskite to NaCl-PEDOT:PSS. D

DOI: 10.1021/acsami.7b14592 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 5. (a) Crystal structure of MAPbCl3 (top) and NaCl (bottom) and (b) atomic arrangement on the (001) surface. (c) ⟨010⟩ direction view of the (001) face of NaCl and MAPbCl3 crystals.

Figure 6. SEM images of perovskite films deposited on (a) pristine PEDOT:PSS and (b) on NaCl-PEDOT:PSS.

lattice constant provides the possibility of the epitaxial growth of the perovskite film on NaCl. In fact, the slightly larger lattice constant of NaCl could induce a strain in the perovskite crystal that shifts its (110) diffraction peak to a smaller angle. Besides the lattice constant, the spatial structure is another key factor for the epitaxial growth. Both NaCl and MAPbCl3 exhibit a polyhedral structure composed of metal cations and six coordinated chlorine atoms. More specifically, the atomic arrangement in the (001) face of both NaCl and MAPbCl3 is shown in Figure 5b. Because of a similar lattice constant and atomic arrangement, NaCl is likely to direct the growth of perovskites through chlorine atoms at the interface, which can coordinate with sodium atoms in NaCl and with perovskites simultaneously. The effect of combining the (001) surfaces of a NaCl crystal and a MAPbCl3 crystal is depicted in Figure 5c, showing a perfect match between the two crystals. The morphologies of the perovskite film prepared on PEDOT:PSS or NaCl-PEDOT:PSS were investigated by SEM. Figure 6a,b shows that the surface of both samples was completely covered by perovskite crystalline grains of submicron sizes, and the average size of perovskite crystals was slightly larger on NaCl-PEDOT:PSS than on pristine PEDOT:PSS. Some rod-like bright particles located at the grain boundaries were the excessive PbI 2 grains with low conductivity. It is worth noting that plenty of small perovskite grains aggregated on the surface in the control sample, whereas they were seldom observed on the perovskite deposited on NaCl-PEDOT:PSS. Small grains can dramatically increase the grain boundaries, which in addition are not passivated by PbI2. Therefore, these grain boundaries can act as recombination centers and result in low Voc and FF. Optical absorption and steady state PL spectra of half cells, i.e. ITO/HTL/perovskites, were recorded to provide insight into the improvement of photovoltaic performance with NaCl doping of PEDOT:PSS. As depicted in Figure 7a, both

on NaCl-PEDOT:PSS. The single diffraction peak assigned to (110) of the perovskite crystal was split into two in the range of 13.8°−14.8°, which could be assigned to (110) and (002) faces of the perovskite crystal. This phenomenon was more conspicuous in the second-order diffraction peak, i.e. the intensity of the peak at 28.3° belonging to the (004) face was greatly enhanced (Figure 4c). The peak splitting may be caused by a change in the crystal orientation or structure transition, and such changes have been observed in the Co2+- or Cs+doped perovskite films.7,43,46 The (002) orientation of the perovskite exhibits an efficient hole transfer between the perovskite and the HTL.47 The structural change was further evidenced by a subtle shift of (110) and (220) diffraction peaks to a smaller diffraction angle relative to the film prepared on pristine PEDOT:PSS (Figure 4b,c). Generally, the shift of the diffraction peak to a smaller diffraction angle is caused by an increase of interplanar spacing, which was mainly induced by the decrease of the grain size or the nonuniform strain.43 It should be noted that the grain size-induced diffraction peak shift and broadening became significant only when the grain size was less than 100 nm. However, in our system, the grain size of the perovskite crystal is much larger than the threshold. Therefore, the diffraction peak shift should be attributed to the presence of the nonuniform strain in the perovskite crystal, which might be related to the epitaxial growth of the perovskite on NaCl. To further understand the additional crystal orientation and the (110) diffraction peak shift of the perovskite film on NaClPEDOT:PSS, we compared carefully the crystal structure of NaCl and the perovskite. The schematics shown in Figure 5a are the typical crystal structure of NaCl (bottom) and the cubic phase of MAPbCl3 (top). It is interesting to find that the two materials have a similar crystal structure and a nearly identical lattice constant of 0.58 and 0.57 nm for NaCl and MAPbCl3, respectively.48,49 The small mismatch (