Novel Dopant-Free D-π-D-π-D Conjugated Hole Transport Materials with Tunable Energy Levels for Efficient and Stable Perovskite Solar Cells

Three novel hole transporting materials (HTM) using the 4-Methoxytriphenylamine (MeOTPA) core were designed and synthesized.The HTMs energy levels were tuned to match with perovskite by introducing electron donating groups symmetrically linked with olefinic bonds as the π bridge. The CH3NH3PbI3-based perovskite solar cells exhibited a remarkable overall power conversion efficiency (PCE) of 16.1 % without any dopants and additives based on 4-((E)-4-(bis(4-methoxyphenyl)amino)styryl)-N-(4-((E)-4-(bis(4methoxyphenyl)amino) styryl)phenyl)-N-(4-methoxyphenyl)aniline (Z34) , which is comparable to 16.7 % obtained by p-doped spiro-OMeTAD-based device. Importantly, the devices based-on three HTMs show relatively better stability compared to devices based on spiro-OMeTAD when aged under ambient air of 30% relative humidity in the dark. 2 Increasing energy demands and concerns about global warming drive the exploration of clean, inexpensive and renewable energy sources. Recently, the photovoltaic community has witnessed a rapid emergence of a new class of solid-state heterojunction solar cells based on solution-processable organometal halide perovskite absorbers . The power conversion efficiency (PCE) of solid-state perovskite solar cells (PSCs) has been quickly increased to over 20% [4-6] because of their unique characteristics, such as a broad spectral absorption range, large absorption coefficient, high charge carrier mobility and long diffusion length. [2] In the configuration of a PSC, the hole transporting material (HTMs) plays the key role of promoting hole migration, as well as preventing internal charge recombination. [7] A great number of HTMs have been developed and applied in PSCs, including various newly designed, inorganic p-type semiconductors, conducting polymers, as well as small molecule hole conductors . 2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD) that was well studied in solid state DSCs, continues to exhibit high performance in PSCs. Due to its relatively low hole mobility and its tedious synthesis strongly correlated to its production cost, numerous alternative HTMs have been explored with an aim to replace such a material . Although inorganic HTMs (CuI and CuSCN) have drawn much attention due to their high hole mobilities and low production cost ,polymeric HTMs, such as conjugated polytriarylamine (PTAA) , poly(3-hexylthiophene-2,5-diyl) (P3HT) [23] etc., have also shown competitive performances in PSCs, small molecular HTMs have advantages of their convenient purification, controllable molecular structures and relatively high efficiency .Regarding small molecule HTMs incorporated with dopants, the PCE of the dopant-free HTM based devices are consistently lying between 10% to 13% , few are over 15%. Herein, we report the synthesis and characterization of three novel dopant-free 4Methoxytriphenylamine (MeOTPA) -based HTMs as shown in Figure 1a, as well as their application in perovskite solar cells. The energy levels of the HTMs can be tuned by attaching 3 to the MeOTPA core with different electron donating groups by olefinic bonds. The device, fabricated with Z34 as HTM, achieves a PCE of 16.1 %without doping under AM 1.5G (100 mW cm) illumination. This result is comparable to that obtained using the well-known ptype doping spiro-OMeTAD (16.7 %). Moreover, the three HTMs based devices presented a better stability than that based on spiro-OMeTAD under ambient air condition of 30% relative humidity without encapsulation after 1000 h in the dark. The cost of the new HTMs is around 1/10 of that of spiro-OMeTAD. The MeOTPA -core HTMs were synthesized by Wittig reaction with cheap starting materials. The synthetic route for the HTMs is depicted in Figure 1b and experimental details are given in the Experimental Section.† The new MeOTPA derivatives (Z33,Z34 and Z35) were fully characterized by H NMR spectroscopy, high resolution mass spectrum, and elemental analysis. All the analytical data are consistent with the proposed structures. All of them show good solubility in common organic solvents, such as dichloromethane, chloroform, tetrahydrofuran and toluene, etc. We also roughly estimated the synthesis cost of 1 gram Z33, Z34 and Z35 and the details are shown in the supporting information. The estimated synthesis cost of Z33, Z34 and Z35 is 70 $/g, 66 $/g and 54 $/g, respectively which is much cheaper than that of spiro-OMeTAD (598 $/g). The UV-Vis absorption spectra of Z33, Z34 and Z35 in THF solution and on thin film state are shown in Figure 2a, and the corresponding data are summarized in Table 1. As shown in Figure 2a, all of them show two absorption bands at 300-320 nm and 400-450 nm regions. The absorption bands in 300-320 nm region can be assigned to the n-π* transition of the TPA moieties. The absorption in 400-450 nm is attributed to the intramolecular charge transfer (ICT) of π-π* transition. [41] Due to the smaller conjugated degree, [42] the ICT peak (λabs/max) of Z35 shows a blue shift compared with Z33 and Z34. Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) measurements show that these three HTMs have high decomposition temperatures (Td, 432.3 4 oC,417.2 oC and 402.5 oC for Z33, Z34 and Z35, respectively) and glass transition temperatures (Tg, 96.2 oC,106.1 oC and 89.1 oC for Z33,Z34 and Z35, respectively) (Figure S1a, ESI†). To understand the charge-carrier transport properties of these HTMs, their hole mobilities (μ) were determined from transit times (tT, ESI Figure S3) with equation of μ= d / VtT. Here d is the organic film thickness, V applied voltage. At room temperature, hole mobilities of Z33, Z34 and Z35 are 4.67×10 cm V s,7.46×10 cm V s and 2.76×10 cm V s at the electric field of 1.0×10 V cm, respectively, all of them are higher than that of pristine spiro-OMeTAD (2×10 cm V s at the electric field of 2.6×10 V cm). Density functional theory (DFT) calculations were carried out in order to understand the electronic structure and the energy levels of the HTMs. The optimized molecular geometries, the highest occupied molecular orbital (HOMO) levels and the lowest unoccupied molecular orbital (LUMO) energy levels were shown in Figure S2. The LUMO distribute mainly on the part of the peripheral units close to the triphenylamine core while the HOMO energy levels distribute mainly on the central triphenylamine core and the extended vinyl bridge. The calculated HOMO levels of three HTMs are estimated to be -4.35 eV,-4.23 eV and -4.38 eV for Z33, Z34 and Z35, respectively. Furthermore, their energy levels are experimentally determined by photoemission yield spectroscopy (PYS). According to PYS result (Figure S1b,ESI), the HOMO energy level of Z33 is -5.34 eV while that of Z35 is -5.42 eV , slightly lower than that of spiro-OMeTAD (5.22 eV) , whereas the HOMO energy level of Z34 is -5.14 eV ,higher than that of spiroOMeTAD, as shown in Figure 2b. Moreover, the optical band gap (Eg) is calculated from the absorption onset wavelength (Eg=1240/λonset) of the corresponding absorption spectrum, indicating that the Eg is 2.72, 2.71 and 2.81 eV for Z33, Z34 and Z35 respectively. The LUMO levels of HTMs are calculated to be -2.62 eV,-2.43 eV and -2.61 eV, which are more positive than that of CH3NH3PbI3 (-3.91 eV). These results agreed well with the trend 5 derived from DFT calculations. Thus, these three HTMs can not only act as a hole transporting layer, but also play as an electron blocking layer in the PSCs. The steady-state PL spectra are shown in Figure 2c. Strong PL quenching was observed when the HTM materials were coated on perovskite films. For the three HTMs coated perovskite films, the PL intensity was reduced to roughly 26%, 20% and 46% of that from pristine films for Z33, Z34 and Z35 respectively, suggesting that Z34 can extract charge carrier more efficiently than the other two HTMs. The hole extraction capacities of Z33-Z35 at glass/CH3NH3PbI3/HTMs interfaces have been investigated by time-resolved photoluminescence (TRPL) measurement. Figure 2d presents the measured PL decay spectra and the corresponding decay time are obtained by fitting the data with biexponential decay function. The PL decay lifetime is reduced to 82.8 ns, 63.1 ns and 107.9 ns for devices with Z33, Z34 and Z35 in comparison with 121.8 ns for the device without HTM layer. From these observations, we conclude that hole injection from the valence band of perovskite into the HOMO of Z34 is more efficient than the other two HTMs. The schematic diagram of the perovskite solar cells applied in this study are shown in Figure 3a. The perovskite solar cells were fabricated by sequential deposition using a similar method as reported in our recent paper. Figure 3b presents a cross-sectional scanning electron microscopy (SEM) image of the perovskite solar cell indicating clearly the infiltration of CH3NH3PbI3 perovskite into the TiO2 pores forming a perovskite/TiO2 nanocomposite which is covered by a perovskite capping layer. We evaluated the photovoltaic performance of perovskite solar cells based on the three HTMs and spiro-OMeTAD with or without doping. The photocurrent densitiy-voltage (J–V) curves under AM 1.5G irradiation of 100 mW cm are presented in Figure 3c and Figure S4a and the photovoltaic parameters are summarized in Table 2. The average PCE of the devices based on three HTMs varies from 10.8% for Z35 to 15.9% for Z34. The lower performance shown by Z35 is mainly related to an insufficient driving force for hole injections related to its deeper HOMO energy level (– 6 5.42 eV) as compared to the valence band of MAPbI3 (–5.43 eV). As expected, Z33 gives a higher Voc (1.087 V ) than spiro-OMeTAD (1.078 V) for the device, which is commensurate with its lower HOMO level, while Z34 gives a lower Voc (1.055 V ) than spiro-OMeTAD due to its higher HOMO level.The best device based on Z34 affords an open-circuit voltage (Voc) of 1.055 V, a short-circuit current density (Jsc) of 21.2 mA cm and a fill factor (FF) of 0.70, leading to a PCE of 16.1 % under AM 1.5G (100 mW cm) illumination. This result is comparable to that of spiro-OMeTAD (16.7 %) doped with lithium bis(trifluoromethylsulfonyl)imide (LiTFSI) and 4-tert-butylpyridine (tBP). However, devices based on doped three HTMs exhibit lower photovoltaic performance compared to dopant-free HTMs, especially in terms of FF and Jsc. It is partly because of that the dopants that work well with spiro-OMeTAD might not be suitable for these three HTMs. [49] Moreover, the dopants seem to have a negative impact on film morphology whcih can be seen from the SEM image in Figure S5. Similar behavior was observed with other reports. In the absence of dopants, spiro-MeOTAD based devices generated a PCE of only 3.92% due to the significant lowering of Voc and FF compared to the doped devices. Hysteresis behavior is frequently observed in perovskite solar cells. A small hysteresis was observed in the J–V curves, the measured PCEs differences are 1%, 3%, 10% and 1% for Z33, Z34, Z35 and spiro-OMeTAD, respectively. The stabilized power outputs from devices based on spiro-OMeTAD and Z33Z35 are 16.6%, 15.2%, 16.1% and 11.2% respectively (Figure S6), consistent with the obtained PCE. The Jsc of the PSC devices shows linear relationship with light intensities (Figure S4 b), indicating that the charge collection ability of HTM containing devices are independent of light density. According to previous investigations, it can be inferred that no space charge limited photocurrent occurs in the devices with Z33-Z35 because of faster electrons and higher hole mobility .