Research Article www.acsami.org
A Triarylamine-Based Anode Modifier for Efficient Organohalide Perovskite Solar Cells Qianqian Lin, Wei Jiang, Shanshan Zhang, Ravi Chandra Raju Nagiri, Hui Jin, Paul L. Burn,* and Paul Meredith* Centre for Organic Photonics & Electronics, School of Chemistry and Molecular Biosciences, and School of Mathematics and Physics, The University of Queensland, Brisbane, Queensland, Australia 4072 S Supporting Information *
ABSTRACT: Organohalide lead perovskite solar cells have emerged as a promising next-generation thin-film photovoltaic technology. It has been clearly recognized that interfacial engineering plays a critical role in cell performance. It has been also proposed that the opencircuit voltage is dependent on the ionization potential of the hole transport layer at the anode. In this communication, we report a simple modification of the anode with a triarylamine-based small molecule (1), which avoids the need to use standard hole transport materials and delivers a relatively high open-circuit voltage of 1.08 V and a power conversion efficiency of 16.5% in a simple planar architecture. KEYWORDS: organohalide lead perovskite, solar cell, work function, triarylamine, interlayer
■
2′,1′,3′-benzothiadiazole)]19 (PCDTBT), as well as other material combinations that have been developed for p-i-n or n-i-p structured planar perovskite solar cells.17,20,21 However, these materials are relatively expensive and they add complexity and constraints to cell fabrication. For example, SpiroMeOTAD is often doped with lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) and processed with tert-butylpyridine (tBP) as an additive. For inverted solar cells, the hole transport/interlayer material needs to modify the anode such that its work function is close to the ionization potential of the organohalide perovskite junction to maximize Voc. Poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS)22,23 can act as an anode work function modifier and is a sufficiently hydrophilic hole transport material to enable the deposition of good quality perovskite films. However, PEDOT:PSS has a relatively low work function, which limits the open-circuit Voc. In contrast, there are many “neutral” hole transport materials that have ionization potentials suitable for yielding a high Voc, but they are hydrophobic in nature. It is difficult to form a high-quality OHP film on top of hydrophobic organic interlayers, such as PCDTBT, via solution processing because of the mismatch of the surface energies.17 As such, there have been only a few reports of OHP solar cells that use thin polymer interlayers for work function modification, e.g., Lin et al.17 and Huang et al.24 Self-assembled monolayers (SAMs) have also been reported to improve the interfaces between the junction and transport
INTRODUCTION Over the past 5 years, an extraordinary level of effort has been expended to understand and utilize the optoelectronic properties of organohalide lead perovskites (OHPs).1−5 In particular, the application of organohalide perovskites in photovoltaics has driven the field, because the lead-based materials can be solution-processed6 or deposited by lowtemperature evaporation techniques7 and their optoelectronic properties can be tuned.8 With the optimization of thin-film processing and device architectures, there are now reports of small laboratory OHP test solar cells that have a power conversion efficiency (PCE) of >20%.9−11 More recently, largearea cells or modules12,13 and efficient tandem solar cells14,15 have also been demonstrated. A critical factor that has led to the improvement in OHP solar cell efficiency has been the use of interfacial engineering.16,17 It has been found that the open-circuit voltage (Voc) is dependent on the energetics of the (inter)layers between the electrodes and organohalide perovskite junction.3 In principle, OHPs can work efficiently in a very simple device architecture: a high-quality absorbing layer sandwiched between work function modified anodes and cathodes.3,18 In the conventional form of this architecture, electron extraction occurs via the transparent conducting electrode (TCE), whereas in an inverted architecture, the holes are collected at the TCE. The hole transport/interlayer materials most commonly used include [2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′spirobifluorene]14 (Spiro-MeOTAD), poly[N,N′-bis(4-n-butylphenyl)-N,N′-bisphenylbenzidine]18 (polyTPD), poly[bis(4phenyl)(2,4,6-trimethylphenyl)amine)]6 (PTAA), and poly[N9′-heptadecanyl-2,7-carbazole-alt-5,5-(4′,7′-di(thien-2-yl)© XXXX American Chemical Society
Received: November 28, 2016 Accepted: February 23, 2017
A
DOI: 10.1021/acsami.6b15147 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 1. (a) Schematic illustration of the device architecture using the organohalide perovskite CH3NH3PbI3. (b) Chemical structure of 1. (c) Proposed energy level diagram of this device structure. The work function of the 1 modified ITO was measured using SKPM, with the other energy levels taken from refs 19 or 33.
undergone a 5% weight loss when the temperature reached 380 °C. The ionization potential of thin films of 1 was measured using photoelectron spectroscopy in air (PESA) and found to be −5.4 eV, which matches the valence band energy of CH3NH3PbI3. The wide optical gap (3.4 eV) was determined from the intersection of thin-film absorption and photoluminescence spectra (Figure S2), and the electron affinity was calculated to be −2.0 eV from the optical gap and the ionization potential. To prepare the 1/ITO-modified anodes, the ITO was first treated with UV-ozone for 10 min to activate its surface. A freshly prepared solution of 1 in chloroform was then immediately spin-coated onto the UV-ozone treated ITO. These substrates (henceforth referred to as 1/ITO) were then heated on a hot plate at 100 °C for 10 min. After cooling, the 1/ITO substrates were washed with chloroform twice to remove any 1 that was not strongly attached to the ITO electrode. The UV−vis absorption spectra of UV-ozone treated ITO, ITO coated with a thick film of 1, and the 1/ITO electrode after washing are shown in Figure 2a. The ITO covered with a thick 1 layer showed strong absorption at wavelengths shorter than 350 nm, which is the absorption tail of 1 (Figure S2). It is important to note that 1 does not absorb visible light (λ > 400 nm) and so should enable maximum transmission into the organohalide perovskite-absorbing layer, although the thick layer of 1 caused pronounced interference peaks in the visible regime. The 1/ITO absorption spectrum looked very similar to that of the UV-ozone treated ITO and hence did not provide conclusive evidence that there was an ultrathin layer of 1 on the ITO. X-ray photoelectron spectroscopy (XPS) was used to confirm the presence of the ultrathin layer of 1 on the ITO. Figure S3a−c shows wide XPS scans of the UV-ozone treated ITO, ITO coated with a thick film of 1, and the 1/ITO modified electrode. The bare UV-ozone treated ITO showed some carbon (either from incomplete removal or adsorption of hydrocarbons between the treatment and measurement) but no nitrogen. As might be expected, the ITO electrode covered with a thick layer of 1 showed strong signals and high percentage of carbon (C 1s) and nitrogen (N 1s). Importantly, the 1/ITO electrode showed a higher content of C 1s and N 1s than the bare ITO (see Figure 2b) indicating that there was 1 on the ITO surface. The presence of 1 on the ITO is also consistent with the small reduction in the relative intensities of the Sn 3d,
layers. For example, Abrusci et al. inserted a fullerene SAM between the organohalide perovskite absorbing layer and the TiO2 electron extraction layer, which improved the fill factor (FF) and reduced charge recombination.25 Li et al. introduced 4-aminobenzoic acid (PABA) SAMs and showed similar enhanced film quality and device performance.26 Zuo et al. employed a 3-aminopropanoic acid self-assembled monolayer (C3-SAM) on a ZnO layer to direct the OHP crystallization and hence film morphology. It was also proposed that the C3SAM helped to align the energetics and improve the charge extraction.27 To reduce the material costs and simplify the device fabrication, hole transport layer-free (HTL-free) organohalide perovskite solar cells have been proposed. Mei et al. reported efficient (PCE ∼ 12.8%) and stable (>1000 h) OHP solar cells using a carbon counter electrode as the anode.28 Shi et al. used gold as the anode and were able to achieve solar cells with PCEs of ∼10.5% without an HTL.29 Tsai et al. reported HTLfree OHP solar cells in which the junction was directly deposited onto the indium tin oxide (ITO) electrode giving PCEs of around 11%.30 The highest PCE for an HTL-free organohalide perovskite solar cell to date is ∼13.5%, with the cell based on a free-standing carbon electrode.31 In this communication, a triarylamine-based small molecule N1,N3,N5 -tris(4-n-butylphenyl)-N1,N3 ,N5-triphenylbenzene1,3,5-triamine (1), was used to modify the work function of the ITO anode. Using an ultraviolet (UV)-ozone pretreatment of the ITO electrodes, it was possible to deposit an ultrathin layer of 1 having the required surface energy to enable the formation of good quality organohalide perovskite films by solution processing. The small molecule modified ITO electrodes gave solar cells with improved Voc and PCE (up to 16.5%).
■
RESULTS AND DISCUSSION The triarylamine-based small molecule 1 (see Figure 1b) was synthesized by reaction of 1,3,5-tris(phenylamino)benzene and 1-n-butyl-4-iodobenzene under Buchwald conditions (see Experimental Section in the Supporting Information). The nbutyl groups imparted solubility to the small molecule. Differential scanning calorimetry at a heating rate of 10 °C/ min showed that on the first scan 1 underwent a melting transition at around 100 °C. On cooling, 1 became amorphous and had a glass transition temperature at 18 °C (Figure S1). Thermal gravimetric analysis (TGA) showed that 1 had B
DOI: 10.1021/acsami.6b15147 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
1/ITO anode had a similar surface roughness to the treated ITO due to the layer of 1 being ultrathin. The water contact angles of clean ITO, UV-ozone treated ITO, a thick film of 1 on glass, and 1/ITO were measured to determine the hydrophilicity of each of the layers (Figure S4). As expected the UV-ozone treated ITO was very hydrophilic with a contact angle
