Carbazole‐Based Hole‐Transport Materials for Efficient Solid‐State Dye‐Sensitized Solar Cells and Perovskite Solar Cells

Two carbazole-based small molecule hole-transport materials (HTMs) are synthesized and investigated in solid-state dye-sensitized solar cells (ssDSCs) and perovskite solar cells (PSCs). The HTM X51-based devices exhibit high power conversion efficiencies (PCEs) of 6.0% and 9.8% in ssDSCs and PSCs, respectively. These results are superior or comparable to those of 5.5% and 10.2%, respectively, obtained for the analogous cells using the state-of-the-art HTM Spiro-OMeTAD. With the recent rise in awareness of environmental and energy issues, renewable energy sources are given more and more attention. Looking at photovoltaics, dye-sensitized solar cell (DSC) is considered to be a promising candidate as a large-area and low-cost renewable energy source.1 In this rapidly developing field, the identification of efficient electrolytes to improve efficiency is of great importance for the promotion of DSCs.2, 3 So far, the best conversion efficiency of 13% have been obtained with prophyrin-based dyes used together with the Co(II/III)-tris(bipyridyl) complexes as redox couple in liquid electrolytes.4 Despite the fact that considerable progress has been made in this field of research, the risk of leakage associated with the volatile nature of a liquid electrolyte, together with fundamental stability issues represent significant limitations for the commercialization of these devices.3 A significant part of the efforts made in this field has been devoted to the optimization of the fabrication of solar cells based on solid-state hole-transport materials (HTMs) as the formal electrolyte system responsible for the regeneration (reduction) of the sensitizing dye molecules.5, 6 To date, the highest efficiency reported amounts to 7.2% for solid-state dye-sensitized solar cells (ssDSCs)7 and 15.4% for perovskite solar cells (PSCs);8 both champion results have been achieved using Spiro-OMeTAD as the HTM. However, previous studies have demonstrated that the onerous synthesis9 and low charge-carrier mobility10 of Spiro-OMeTAD significantly limit its potential of up-scaling for applications of ssDSCs and PSCs.11, 12 Therefore, the development of a new generation of HTMs13 with low cost, facile synthesis and high charge-carrier mobility is of high priority.6, 14 Carbazole-based derivatives have attracted much attention because of their interesting photochemical properties.15 Recent interest in the carbazole derivatives has been caused by its good charge-transport function, which can be exploited in the molecular design of new types of HTMs in organic light-emitting diodes (OLED)16 and as electron donor group in organic sensitizers of D-π-A-type in DSCs.17 Another fascinating advantage is the versatility of the carbazole reactive sites that can be substituted with a wide variety of functional groups, allowing fine-tuning of its optical and electrical properties.18 Herein, we present on the design and synthesis of two novel carbazole-based, small-molecule organic HTMs denoted X51 and X19 (Figure 1) characterized by a straightforward synthetic route and high hole mobility, together with their successful application in ssDSCs and PSCs. One of the HTM- based, X51, devices exhibit high power conversion efficiencies (PCEs) amounting to 6.0% and 9.8% in dye sensitized and perovskite solar cells, respectively. These results are comparable to the PCEs of 5.5% and 10.2% obtained by using the well-studied champion HTM, Spiro-OMeTAD, under standard AM 1.5 solar light illumination of 100 mW·cm−2 intensity. To the best of our knowledge, this is the first report of a carbazole-based, small-molecule organic HTM that can challenge the state-of-the-art HTM Spiro-OMeTAD both in ssDSCs and PSCs. Results from DFT calculations give insights into the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of the HTMs, as are shown in Figure S1. The energies of the calculated HOMO and LUMO of these HTMs have been summarized in Table S1, and these data make it clear that the energy of the HOMO levels of all HTMs are very similar. The reorganization energy represents the relaxation of the molecule after oxidation or reduction, typically modeled by withdrawing or adding one electron, and a small reorganization energy implies, based on the anatomy of a Marcus-type electron-transfer mechanism between the mole­cules in the molecular HTMs, fast transfer rate for electrons or holes. As indicated in Table 1, the molecule X51 has the lowest calculated reorganization energy. This implies, on a relative scale, faster hole-transport than the other two HTMs. It is not unlikely that these properties can be assigned to the larger conjugation length of X51 as compared to Spiro-OMeTAD and X19 from an intrinsic structural point of view. The synthetic routes of X19 and X51 are shown in Scheme. S1. Inexpensive starting materials were utilized to synthesize the desired materials. X19 was synthesized by a bromination, a copper-catalyzed Ullman reaction and a palladium-catalyzed Buchwald-Hartwig reaction step. Compared to X19, the synthesis of X51 requires one further step of protection and deprotection. All of the materials were characterized by 1H/13C NMR spectroscopy and high-resolution mass spectrometry (HR-MS) (Figures S2–S7). The role of Spiro-OMeTAD is as a reference HTM in this study. Cyclic Voltammetry (CV) (Figure S8) and Differential Pulsed Voltammetry (DPV) (Figure 2b) applied to each of the HTMs were recorded in order to determine the ground-state oxidation potential (EOX). The corresponding data are collected in Table 1. The EOX values were determined to 0.61 V and 0.64 V vs. NHE in DCM solution for X19 and X51, respectively. Considering the oxidation potential (0.88V vs. NHE) of the LEG4 dye,19 all hole conductors have an overpotential between 270 mV and 240 mV, which should be sufficient for efficient regeneration of the oxidized dye. The optical absorption and fluorescence spectra (Figure 2a) in DCM were recorded at a concentration of 10−5 M. All photophysical data of the HTMs are listed in Table 1. X19 and X51 exhibit strong absorption at 310 nm and a weak spectral feature at 370 nm, while the absorption of Spiro-OMeTAD shows a weak feature at 306 nm and strong absorption at 386 nm. The weak absorption in the visible light region indicates that all of these HTMs will not obstruct sunlight from efficiently reaching the sensitizing dye molecules. In addition, the photoluminescence spectra show that the X-type HTMs display an emission peak around 460 nm. This is slightly red-shifted as compared to Spiro-OMeTAD. From the intersection of the emission and absorption spectra we can obtain the E0–0 transition energy and thus estimate the optical band gap (Table 1). As previously highlighted, charge transport is another significant parameter to consider in the design of new HTMs for ssDSCs. Herein, the hole mobility of the HTMs were determined by use of space-charge-limited currents (SCLCs) according to literature.11, 20 Fitting the J–V curves (Figure 2c) for each material to this expression gives the mobility data listed in Table 1. The value obtained here for Spiro-OMeTAD is similar to the data reported in the literature and our former work.21 The hole mobility of X51 is slightly higher than that of X19, and both of these carbazole-based HTMs showed almost three times higher hole mobility than of Spiro-OMeTAD. These results are in good agreement with the implications from the computational data showing that the molecule X51 has the lowest calculated reorganization energy. The conductivities of the films were determined by using a two-contact electrical conductivity set-up, which were performed by following a published procedure. The curves obtained and values recorded for the LiTFSI-doped HTMs are depicted in Figure 2d and Table 1. The conductivity of pure Spiro-OMeTAD closely matches that reported in the literature.11, 20 We note that the conductivity of X51 is slightly higher than that of Spiro-OMeTAD, but much higher than that of X19. This might linked to the larger conjugated system in the former molecular material, leading to more efficient π–π stacking in HTMs films. In this case, the hole mobility and conductivity of X51 are superior to those of X19. Consequently, it is a realistic assumption that the hole transfer of the HTM X51 could contribute to the better photovoltaic parameters observed as compared to those of X19-, as well as Spiro-OMeTAD-based devices; as will discussed in more detail later. The photovoltaic properties of these HTMs were investigated by fabricating ssDSCs using the organic dye LEG4 known to possess a high molar extinction coefficient. In order to properly evaluate the photovoltaic properties, devices containing Spiro-OMeTAD as HTM were used as reference. It has been widely demonstrated that lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) is essential in ssDSCs to render a low series resistances and to facilitate photocurrent generation.20, 22 Our previous work has also manifested that different HTMs need different concentration of Li+ doping.9 Therefore, we firstly optimized the amount of LiTFSI that was added into the HTM solutions used for spin coating. The final optimal Li+ concentration used for X19- and X51-based ssDSCs was 30 mM; for J–V curves and photovoltaic parameters obtained at different concentrations of LiTFSI doping, see Figure S9 and Table S2. The optimal concentration of LiTFSI used in combination with Spiro-OMeTAD was found to be 20 mM; the same as frequently reported in literature. The J–V curves of the best devices containing different HTMs are shown in Figure 3a, and the corresponding parameters are stated in Table 2. The X51-based devices yielded the highest efficiency among these HTMs, up to 6.0%, with a short-circuit photocurrent density (JSC) of 9.27 mA·cm−2, an open-circuit photovoltage (VOC) of 0.92 V and a fill factor (FF) of 0.70. Under the optimal conditions, the X19-based devices offered a lower efficiency, 4.5%, with a VOC of 0.75 V, a JSC of 9.62 mA·cm−2 and an FF of 0.62. The standard HTM Spiro-OMeTAD-based devices reached a reference efficiency of 5.5% with a VOC of 0.90 V, a JSC of 9.22 mA•cm−2 and an FF of 0.66. We can note that the photocurrent densities in all the devices are very similar after optimization of the concentration of LiTFSI. The incident photon-to-current conversion efficiency (IPCE) spectra of these homologous devices are shown in Figure 2b. For the sensitizer used in this study, LEG4, all the HTMs generate highly similar IPCE of 65% at the maximum absorption wavelength of 475 nm. However, the open-circuit voltage of the X19-based devices was 170 mV and 150 mV lower than that of the Spiro-OMeTAD- and X51-based devices, respectively. One reason may be that the HOMO energies of X51 and Spiro-OMeTAD are more positive than that of X19; another factor might be due to the smaller molecular size of X19 induces faster electron recombination in the device, an explanation was proposed in our former work.21 Figure S11 show the τe plotted vs. VOC from transient photovoltage decay measurements of these devices. At a certain VOC value, the τe values from X51-based device was obviously the highest among all of the HTMs (indicating the slowest recombination). Especially, the τe value for the X19-based device were much lower than that of X51 and Spiro-OMeTAD-based devices, which is in good agreement with our previous discovery21 that the HTM-X19 with small molecule size results in a facilitated approach to the TiO2 interface, leading to a fast recombination process in the devices. Subsequently, we also fabricated devices based on the HTMs using perovskite as light-harvesting material under the same preparation conditions as used before without any further device optimization; thus, the concentrations of LiTFSI and t-BP were the same as of the ones used for the ssDSC devices. The configuration of this typical solid-state mesoscopic solar cell is as follows: compact TiO2/nc-TiO2/CH3NH3PbI3-XClX/HTM/Ag. A 350 nm thick nc-TiO2 layer was deposited by spin-coating of a diluted paste, and CH3NH3PbI3-XClX was employed as the light harvesting material in virtue of the long electron-hole diffusion lengths in the mixed halide perovskite.23 The J–V curves obtained from the best performing solar cells are showed in Figure 4a, and the corresponding photovoltaic parameters are summarized in Table 2. The devices prepared containing Spiro-OMeTAD exhibited the highest conversion efficiency of 10.2%, with a VOC of 0.83 V, a JSC of 16.82 mA·cm−2, and an FF of 0.73, while X51-based devices gave rise to an essentially comparable conversion efficiency of 9.8%, with a VOC of 0.88 V, a JSC of 16.79 mA·cm−2, and an FF of 0.66. The devices prepared with X19 as HTM yielded slightly lower energy output in the form of a VOC of 0.76 V, a JSC of 17.14 mA·cm−2, and an FF of 0.58, giving rise to a PCE of 7.6%, which is lower than the output from the devices prepared with the other HTMs of this study. In addition, we performed a comparative investigation using a higher concentration of LiTFSI (30 mM) doping in the Spiro-OMeTAD solution; however no significant change in conversion efficiency could be detected. The J–V curves and photovoltaic parameters obtained from DSCs prepared from HTM solutions containing different concentrations of LiTFSI doping are shown in Figure S10 and Table S2. It was also confirmed that the JSC values of the best performing devices are highly consistent with the results obtained from the IPCE measurements (Figure 4b). However, significant deviations were observed for the VOC and fill factor, where X51-based devices exhibited the highest VOC in agreement with the high HOMO energy level of X51. Moreover, the fill factor of devices based on X51 and Spiro-OMeTAD both in ssDSCs and perovskite solar cells were considerably higher than devices made with X19, which might cause by the higher conductivity of the former materials resulting in a lower overall device serial resistance. In conclusion, we have designed and synthesized two new carbozle-based, small molecules, organic HTMs of different molecular weight, oxidation potential, hole mobility, and electrical conductivity. Theoretical studies show that X51 clearly exhibits a lower reorganization energy implying faster hole transport than X19. This is in good agreement with the expected influence of a larger conjugation system in the HTM X51. By optimization of the conduction parameters of the HTMs and subsequent application in ssDSCs using LEG4 as photosensitizer, devices with the HTM X51 exhibit higher power conversion efficiencies (PCEs) of 6.0% than devices consisting of the smaller HTM X19 (4.5%). Devices based on the HTM X51 thus clearly outperform the state-of-the-art HTM Spiro-OMeTAD offering a PCE of 5.5%. Employing perovskite as light-harvesting material, devices containing the HTM X51 exhibited a higher power conversion efficiency (PCEs) of 9.8%, than that of devices based on the HTM X19 (7.6%), also comparable to the PCE of 10.2% obtained in devices based on the HTM Spiro-OMeTAD. These results clearly qualify the new family of HTMs as promising future materials for high-efficiency ssDSCs and PSCs. Moreover, X51 exhibits higher hole mobility and conductivity than X19, leading to better photovoltaic performance of the devices investigated. The results obtained show that the use of molecular engineering to develop new and efficient HTMs with high hole mobility and high conductivity is a feasible strategy for the ssDSCs and PSCs in the future. Supporting Information is available from the Wiley Online Library or from the author. Scheme S1: The synthetic routes of X19 and X51; Figure S1: Frontier orbitals of the HTMs; Figure S2–S7: Characterization of the new HTMs X19 and X51 (1H/13C NMR spectroscopy and HR Mass spectrometry); Figure S8: Cyclic Voltammograms; Figure S9: J–V curves of ssDSCs devices containing the HTMs with different concentrations of LiTSFI; Figure S10: J–V curves of perovskite solar cells containing Spiro-OMeTAD prepared using different solution concentrations of LiTSFI; Figure S11: Electron lifetime plots of ssDSCs based on different HTMs; Table S1: Calculated electrochemical properties of the HTMs; Table S2: Photovoltaic parameters determined from J–V measurements of ssDSCs and perovskite solar cells at different LiTFSI doping levels. This work was financially supported by the Swedish Research Council, the Swedish Energy Agency, the Knut and Alice Wallenberg Foundation, the National Natural Science Foundation of China (21120102036, 91233201), the National Basic Research Program of China (973 program, 2014CB239402) and the China Scholarship Council (CSC). We would like to thank Jianghua Zhao (Dalian University of Technology) for recording the HR-MS spectra. We also greatly thank Erik Gabrielsson (KTH) and Jing Huang (KTH) for their helpful discussions and Majid Safdari (KTH) for his kind help in laboratory experiments. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.