Spiro-Phenylpyrazole/Fluorene as Hole-Transporting Material for Perovskite Solar Cells

Spiro-OMeTAD with symmetric spiro-bifluorene unit has dominated the investigation of hole-transporting material (HTM) for efficient perovskite solar cells (PSCs) despite of its low intrinsic hole conductivity and instability. In this study, we designed and synthesized three asymmetric spiro-phenylpyrazole/fluorene base HTMs, namely: WY-1, WY-2 and WY-3. They exhibit excellent electrochemical properties and hole conductivities. Moreover, the PSC based on WY-1 exhibits the highest power conversion efficiency (PCE) of 14.2%, which is comparable to the control device employing spiro-OMeTAD as HTM (14.8%). These results pave the way to further optimization of both molecular design and device performance of the spiro-based HTMs.

Optical and electrochemical properties. Figure 2 showed the UV-vis absorption of WY-1, WY-2 and WY-3 in both solution and solid states, together with those of spiro-OMeTAD for comparison, while the corresponding numerical data are depicted in Table 1. The lowest-energy absorption peak in solution occurred at 386 nm for all new HTMs and spiro-OMeTAD, which can be attributed to π−π* electronic transition at the diphenylamine substituted fluorene moiety 27 . Their optical energy gap (E g,opt ), estimated from absorption onset, was calculated to be ca. 2.98 eV, which is identical to that of spiro-OMeTAD (3.00 eV) documented in literature 36 . Going from solution to solid state, their absorption profiles are essentially unchanged, indicating no obvious π-π stacking interaction in the solid state due to the orthogonal spiro geometry. Cyclic voltammetry (CV) was next performed to determine the relative energy of the highest occupied molecular orbital (HOMO) ( Table 1). As shown in Figure S2, all samples display reversible multi-oxidative behaviours. Particularly, the numbers of reversible CV wave are proportional to the number of diphenylamino substituents, confirming excellent electrochemical stability. In addition, their oxidative onsets versus FcH/FcH + were found to be slightly more positive than spiro-OMeTAD, giving the more stabilized HOMO (−5.15 ∼ −5.16 eV) compared with spiro-OMeTAD (−5.10 eV), and were expected to have higher V OC values for the PSCs. The stabilized HOMOs may be attributed to the reduced π-conjugation and greater electron withdrawing effect exerted by the orthogonal phenyl pyrazole. This is confirmed by theoretical calculation based on density functional theory (DFT) method using B3LYP hybrid functional and 6-31 g* basis set ( Figure S3), for which the electron density of HOMO and LUMO for WY-1, WY-2 and WY-3 are all residing on both the diphenylamino substituents and adjacent fluorene unit. In sharp contrast, the frontier orbitals of spiro-OMeTAD are delocalized over both of the bifluorene cores. Such different electronic density distributions result in reduced HOMO levels for WY-1 to WY-3 which is in accordance with the experimental tendency.
Mobility measurement. The hole mobilities were studied by measuring the electrochemical impedance measurements on a symmetrical, hole-only device comprising of Au/HTM/Au architecture (Fig. 3a) under different voltage bias over 1 MHz to 100 Hz in the dark. A typical Nyquist plot of Au/WY-1/Au measured at −0.4 V bias was shown in Fig. 3b; data points were fitted using an equivalent circuit provided in the insert of Fig. 3b. The equivalent circuit contains a series resistance (R 0 ), a charge transfer resistance (R 1 ) and a constant phase element (CPE 1 ) which associates with charge accumulation in the depletion layer near the interface of metal electrode (gold) and semiconductor (HTM) 37,38 . Because the HTM is an organic semiconductor itself; therefore, R 1 and CPE 1 decreases when forward bias increases 39 . The hole mobility could be estimated using the equation (1): where L is the width of the depletion 40,41 zone and is approximately equal to the film thickness at such resistive and thin domain, k b is Boltzmann constant, and T is the absolute temperature. CPE 1 and R 1 were obtained by curve fitting. The relationship of hole mobility with different applied bias is shown in Fig. 3c. It can be seen that WY-1 has identical mobility to spiro-OMeTAD except at high bias over 0.5 V, while the mobilities of WY-2 and WY-3 are both lower than WY-1 and spiro-OMeTAD. The lower mobility for WY-2 and WY-3 may originate from their greater size and higher rigidity 33 , both are expected to jeopardize the hole transporting property.   (Fig. 4a). The perovskite layer (CH 3 NH 3 PbI 3 ) was sequentially deposited, following the method developed by Burschka et al. 42 Fig. 4b reveals that the energy levels of new HTMs are suitable for making efficient PSCs. HTMs were spin-coated with typical additives such as bis(trifluoromethane) sulfonimide lithium salt (Li-TFSI) and 4-tert-butylpyridine (tBP) (detailed information can be found in the experimental section). The photovoltaic performances of PSCs made with all studied HTMs and benchmark spiro-OMeTAD are presented in Fig. 5. Figure 5a-d summarizes photovoltaic performance of multiple devices from different batches using the same recipe and under 100 mW cm −2 AM1.5 G solar illumination. It can be seen that WY-1 exhibits competitive PCE vs. spiro-OMeTAD, while WY-2 and WY-3 show lowered short-circuit current density (J SC ) and   Fig. 5f, the higher J SC obtained in WY-1-based device is derived from its higher IPCE at 700 to 750 nm, which may be due to the superior optical transmittance of WY-1 at around 700 nm ( Figure S4), allowing better IPCE over that of spiro-OMeTAD. But owing to the lowered hole mobility vs. spiro-OMeTAD, WY-1 device suffers from notable ohmic losses in FF due to the slightly higher resistivity. In addition, WY-2 and WY-3 based PSCs exhibit very similar photovoltaic performance with J SC of 18.50 and 18.27 mA cm −2 , V OC of 1.06 and 1.05 V, FF of 0.70 and 0.67 and PCE of 13.60 and 12.89%, respectively. The inferior performance of WY-2 and WY-3 than WY-1 may be attributed to the larger molecular volume vs. WY-1. Hence, the greater steric hindrance in WY-2 and WY-3 renders poor intermolecular π-stacking interaction (i.e. carrier transport) in the deposited thin film and thus lowers the J SC and FF of the PSC.

Conclusions
In summary, we reported three HTMs with asymmetric spiro-phenylpyrazole/fluorene architecture, which renders the effective modification on conventional symmetric spiro-OMeTAD. These new HTMs possess suitable HOMO level which could be beneficial to the high V OC . Moreover, they exhibit adequate thermal, optical and electrochemical properties but a slightly different hole-transporting ability, while WY-1 shows the best hole mobility relative to that of WY-2 and WY-3. As a result, WY-1 based device exhibits the highest PCE of (f) IPCE diagrams of the corresponding devices; the total currents derived from the integrated curves are: 18.05 mA/cm 2 , 18.16 mA/cm 2 , 18.08 mA/cm 2 and 17.32 mA/cm 2 for the spiro-OMeTAD, WY-1, WY-2, and WY-3-based device, respectively.
14.20% with a J SC of 19.48 mA/cm 2 , a V OC of 1.05 V and a FF of 0.69 which are very close to the benchmark HTM spiro-OMeTAD (14.84%). On the other hand, higher series resistances originate from lowered hole mobility are found in these WY-based PSCs, rendering lowered J SC and FF. Currently, modification on molecular structure and optimization on concentration of additives are undergoing in our laboratory to solve this issue. We believe that further modulation on such spiro-phenylpyrazole/fluorene structure will push the device performance to an even higher level.

Methods
General Information and Materials. All reactions were performed under a nitrogen atmosphere and solvents were distilled from appropriate drying agents prior to use. Commercially available reagents were used without purification unless otherwise stated. The starting materials (A) 43 , (B) 44 and (C) 33 were prepared according to the literature methods.

3-Tert-butyl-5-(2-bromophenyl)-1-(tetrahydro-2H-pyran-2-yl)-1H-pyrazole (3).
A mixture of compound 2 (2.09 g, 7.5 mmol), 3,4-dihydro-2H-pyran (2.52 g, 30 mmol), p-toluenesulfonic acid monohydrate (140 mg, 0.75 mmol) and CH 2 Cl 2 (50 mL) was refluxed overnight under nitrogen. The mixture was washed with NaHCO 3 , brine and water and dried over anhydrous Na 2 SO 4 . The solid was next purified by column chromatography using hexane: EA = 20:1 as the eluent. Yield: 77%. 1 (4). Following the same procedure described for 2, treatment of A and dimethylbutan-2-one afforded a white product 4 in 75% yield. 1 (5). Following the procedure described for 3, combination of 4 with 3,4-dihydro-2H-pyran afforded a white product 5 in 85% yield. 1 Table 2. Photovoltaic parameters of PSCs with different hole transporting layer. Photomasks (0.2 × 0.5 cm 2 ) made of thin metal sheet were applied before all measurements. The hysteresis index (HI) is defined as (P max(forward scan) /P max(backward scan) )−1, which P max is the maximum power in I-V measurement. *These values were obtained from 6-8 devices.  (6). Following the procedure described for 2, condensation of compounds A and B afforded a white product 6 in 63% yield. 1 (7). Following the procedure described for 3, treatment of 6 with 3,4-dihydro-2H-pyran afforded a white product 7 in 90% yield. 1  After stirring for 1 h at RT, the mixture was quenched with ice water and neutralized with NaHCO 3 (aq). The crude product was extracted with CH 2 Cl 2 and washed with water. The organic layer was dried over anhydrous Na 2 SO 4 and purified by silica gel column chromatography eluting with a mixture of hexane: EA = 8:1, giving a pale green product in 45% yield. 1  were recorded on a JEOL SX-102A instrument operating in electron impact (EI) or fast atom bombardment (FAB) mode. The thermogravimetric analysis and differential scanning calorimetry were measured on a Seiko SSC 5000 instrument. UV-Vis spectra were recorded on a Hitachi U-3900 spectrophotometer. Cyclic voltammetry was conducted on a CHI 621 A Electrochemical Analyzer with a conventional three-electrode system consisting of a platinum working electrode, a platinum counter electrode and an Ag/AgCl reference electrode, and using 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF 6 ) as the supporting electrolyte. All potentials were calibrated against the ferrocene/ferrocenium couple assuming FcH/FcH + = −5.1 eV 47 . Cross-sectional view of PSC was investigated using a Hitachi SU-8010 field emission scanning electron microscope. The conductivity of HTM was investigated by a computer-controlled Autolab PGSTAT30 Potentiostat/Galvanostat. The I-V characteristic was measured using a Keithley 2400 digital source meter under AM1.5 G illumination at intensity of 100 mW/cm 2 (Peccell Technologies, PEC-L15). A KG3 monocrystalline silicon photodiode (Oriel, USA) was used to calibrate the light intensity. A 0.2 × 0.5 cm 2 photo-mask was attached to the front side of the PSC to accurately control the illuminating area. The IPCE was measured using monochromatic light illumination (Peccell Technologies, PEC-20).