Selected Electrochemical Properties of 4,4’-((1E,1’E)-((1,2,4-Thiadiazole-3,5-diyl)bis(azaneylylidene))bis(methaneylylidene))bis(N,N-di-p-tolylaniline) towards Perovskite Solar Cells with 14.4% Efficiency

Planar perovskite solar cells were fabricated on F-doped SnO2 (FTO) coated glass substrates, with 4,4’-((1E,1’E)-((1,2,4-thiadiazole-3,5-diyl)bis(azaneylylidene))bis(methaneylylidene))bis(N,N-di-p-tolylaniline) (bTAThDaz) as hole transport material. This imine was synthesized in one step reaction, starting from commercially available and relatively inexpensive reagents. Electrochemical, optical, electrical, thermal and structural studies including thermal images and current-voltage measurements of the full solar cell devices characterize the imine in details. HOMO-LUMO of bTAThDaz were investigated by cyclic voltammetry (CV) and energy-resolved electrochemical impedance spectroscopy (ER-EIS) and were found at −5.19 eV and −2.52 eV (CV) and at −5.5 eV and −2.3 eV (ER-EIS). The imine exhibited 5% weight loss at 156 °C. The electrical behavior and photovoltaic performance of the perovskite solar cell was examined for FTO/TiO2/perovskite/bTAThDaz/Ag device architecture. Constructed devices exhibited good time and air stability together with quite small effect of hysteresis. The observed solar conversion efficiency was 14.4%.


Introduction
Perovskite solar cells (PSC) are attractive due to their record solar conversion efficiency near 25%, approaching already the efficiencies of the best solid-state photovoltaics (Si, GaAs, CIGS, CdTe) [1,2]. The PSC is sometimes called a 'young sister' of dye-sensitized solar cell [3,4] to highlight their common principles: The generic perovskite, CH 3 NH 3 PbI3 w as actually disclosed for the first tim e (in 2009) using a liquid-junction photoelectrochem ical device [1,3,4]. Yet, the electrochem ical approaches are still rather scarce in PSC research [4]. In the regular (and so far the most efficient) n-i-p architecture of PSC, the oxide-based electron-selective layer (usually TiO2 ) provides a negative electrical contact to the perovskite photo-absorber. The second (positive) electrical contact is fabricated from hole transporting m aterial (HTM ), like CuSCN [5] or spiro-OMeTAD, the latter being ubiquitously used in alm ost all highly-efficient PSCs [1].
In another w ork [16]  Petrus et al. [15] proposed interesting small azomethine molecules: thiophene or benzenothiazolounits as core and tw o TPA units at the end of m olecules, for application in organic solar cells. They conducted an in-depth investigation of all crucial param eters for application in organic photovoltaics. The best device w as obtained by using the donor to acceptor ratio equal to 1 Herein, we report the selected electrochemical properties of a symm etrical imine, 4,4'-((1E,1'E)-((1,2,4-thiadiazole-3,5-diyl)bis(azaneylylidene))bis(m ethaneylylidene))bis(N ,N -di-p-tolylaniline) (abbreviated bTAThDaz) with the aim of its use in perovskite solar cells as hole transporting material (HTM). Our im ine was characterised by cyclic voltam m etry and energy-resolved electrochem ical impedance spectroscopy to assess its applicability in perovskite solar cell construction. Morphology, composition and structure of the obtained layers were investigated by scanning electron microscopy (SEM) together w ith X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD). Thermal properties of the investigated im ine were checked by therm ogravim etric analyses in two different atm ospheres, synthetic air and Ar. FTIR spectra were measured in w ide tem perature range to investigate possible inter-or intram olecular interactions in our imine. The therm ographic camera w as used to detect the location of defects in the fabricated devices and the electrical behavior of the investigated imine. To check the crucial electrical properties, the hysteresis effect in the investigated compound, the electrical and thermal response to forward and reverse bias was also studied. Finally perovskite solar cells w ith our im ine as HTM w as constructed and com pared w ith a control device with spiro-OMeTAD. New materials, that can replace the currently used HTMs such as spiro-OMeTAD represent the current trends in organic and perovskite solar cells, taking into consideration price and some synthesis steps of the currently used com pounds. Here we present, for the first time, the perovskite solar cells with imine bTAThDaz as HTM with efficiency exceeding 14%.

Experimental Section
The electronic structure of the imine thin film was measured by energy-resolved electrochemical im pedance spectroscopy (ER-EIS) [21,22]. The im ine and spiro-OMeTAD films were spun coated on indium tin oxide (ITO) substrate in a glove box under N 2. The microcells for electrochem ical m easurem ents was m ade of plastic cone and consisted of disc w orking electrode (area of 12 m m 2), Ag/AgCl reference electrode and Pt w ire as counter electrode. The measurements were controlled by potentiostat (home-made) and perform ed in 0.1 M TBAPFg (Sigma Aldrich, Bratislava, Slovakia) solution in anhydrous acetonitrile. To calculate the local vacuum level a potential of -4 .6 6 eV for Ag/AgCl reference electrode energy vs. vacuum was assumed. The im pedance measurements were performed with AC harmonic voltage signal frequency equal 0.5 Hz, the rms value of 100 mV, and the sweep rate of the DC voltage ramp 10 mV-s-1 . The results of impedance experiment were controlled by an impedance/gain-phase analyzer, Solartron analytical, model 1260 (Ametek, Berwyn, IL, USA). Other electrochem ical m easurem ents were carried out, as described elsewhere [23], using a Metrohm Autolab PGSTAT M204 potentiostat (Barendrecht, The Netherland), glassy carbon electrode (diam. 2 mm) and a platinum rod and Ag/AgCl. As reference redox system ferrocene (Fc) was used. The optical (UV-Vis) spectra were measured using a PerkinElm er Lam bda 19 UV-VIS-NIR spectrophotometer (PerkinElmer, Waltham, MA, USA), either in chlorobenzene solution (quartz optical cell, 1 cm length) or on thin films supported by quartz plates. To fabricate high-quality films on quarts, the solvents'-purified substrate (see below) was subjected to additional cleaning in plasma using the piezo-brush PZ2 (Relyon plasma, GmbH, Regensburg, Germany).
Pow der X-ray diffraction (XRD) was studied on a Bruker D8 Advance diffractometer (Bruker, Billerica, MA, USA) using C u K a radiation.
The conductivity w as measured by using the van der Pauw contact configuration of quartzsupported thin film. The gold contacts were deposited in the corners of the square film samples.
Two of the contacts served as current input/output; the other two contacts were used for voltage m easurem ent. A Keithley 6430 current source (Tektronix Inc., Solon, OH, USA) and two Keithley 6514 electrom eters (Tektronix Inc., Solon, OH, USA) m onitoring potentials at voltage contacts were used; the resulting voltage drop was measured by connecting a Keithley 2182A nanovoltmeter to the analogue low-impedance outputs of the electrometers. The conductivity across the film was measured by two-points probe using the Au-supported thin film.
XPS spectra were recorded using a Thermo Scientific K-Alpha XPS system (K-Alpha 2010, Thermo Fisher Scientific Inc., East Grinstead, UK) equipped with a microfocused, monochromatic Al K a X-ray source (1486.68 eV). An X-ray beam of 400 pm size was used at 6 mA X 12 kV. The spectra were acquired in the constant analyser energy mode with pass energy of 200 eV for the survey. Narrow regions were collected with pass energy of 50 eV. The Thermo Scientific Avantage software, version 5.9915 (Thermo Fisher Scientific Inc., East Grinstead, UK), w as used for data acquisition and processing. Spectral calibration was achieved by using the automated calibration routine and the internal Au, Ag and Cu standards supplied w ith the K-A lpha system. The surface com positions (in at%) were determ ined by considering the integrated peak areas of the detected atoms and the respective sensitivity factors. The fractional concentration of a particular element A was computed using: where I and s are the integrated peak areas and the Scofield sensitivity factors corrected for the analyzer transm ission, respectively. (The subscripts 'A ' or 'n' m ean the elem ent A or any elem ent detected in the surface, respectively). The experim ent was designed to apply voltage in a programmed w ay as follows: the potential was applied in range from the 0 V to 10 V with 0.5 V step between different values during three minutes for each voltage. The current response w as recorded during this three m inute-intervals and each step was separated with 10 s window, when the IR image was collected, while the current was still passing through the sample (see Figure S1). Both camera and power source were digitally controlled. For this experiment the samples were prepared by spin-coating technique using 5000 rpm/20 s for two solutions 15 mg/mL w/w and 10 mg/mL giving ca. 40 nm and 20 nm, respectively. The Fourier transform infrared (FTIR) spectra of the im ine in the region of 4000-400 cm -1 were measured at 4 cm -1 resolution w ith accum ulation of 256 scans on a Bruker Tensor 37 spectrom eter (Bruker Optics, Ettlingen, Germany) with an MCT detector using the KBr pellet technique.
Therm ogravim etric (TGA) m easurem ents were perform ed on TA Q5000 IR therm obalance (New Castle, U.S) using Pt holders. Behaviour of the im ine w as exam ined under two different atm ospheres of synthetic air and Ar (5N). The gas flow was 100 m L m in -1 , supplied directly in the vicinity of the studied sample. Each TGA test included heating in a temperature range of 30-800 °C with a rate of 1 0°-min-1 .
Alternative substrates for thin-film deposition were optical quartz or m icroscope glass; they were cleaned in the same way. For conductivity measurements, the substrate was 90 nm gold layer which was vacuum -deposited on top of 20 nm Ti layer on glass. A freshly made Au/Ti/glass substrate was used as is, i.e., without cleaning.
A thin film of bTAThDaz was deposited from 70 mM chlorobenzene solution. The solution was carefully homogenized by ultrasonic treatment (bath-type cleaner as above) as long as the SEM images of the final film contained no visible particles (see Figure S2, exhibiting the typical m orphology of a non-optimized film containing undissolved particles). The deposition was carried out by spin-coating method (Laurel Technologies Corp. spin-coater, Laurel Technologies Corp., North Wales, PA, USA) at 6000 rpm/20 s. The dynam ic m ode of operation w as em ployed (the solution was dropped onto the substrate under rotation).

Construction o f Perovskite Solar Cells with bTAThDaz as HTM
Etched FTO/glass (15 Q sq-1 ) was cleaned w ith acetone, isopropyl alcohol, deionized w ater (15 m in in each solvent). A com pact TiO2 layer (cp-TiO2) was deposited onto FTO by spin-coating of 55 pL titanium diisopropoxide bis(acetylacetonate) (75wt% in isopropanol, Sigm a-Aldrich, Saint Louis, MO, USA) dissolved in 1 mL 1-butanol (99.8%, Sigm a-Aldrich, Saint Louis, MO, USA) at 1500 rpm for 30 s, w hich w as dried at 120 °C for 10 min. A mesoporous TiO2 layer (ms-TiO2 ) was form ed on the cp-TiO2 by spin-coating the diluted TiO2 paste (0.25 g m L -1 ) at 3000 rpm for 30 s. After the substrate was dried at 120 °C for 10 min, it w as then sintered at 450 °C for 1 h. After Seoul, Korea). All devices were m easured by m asking the active area w ith a thin mask (0.14 cm 2).
The J -V characteristics of all the devices were measured at a voltage scan rate of 0.1 V-s-1 .

Synthesis and Purification o f bTAThDaz
The im ine w as synthesized in a one-step condensation reaction in solution betw een of 4-(di-p-tolylam ino)benzaldehyde) and diamine such as l,2,4-thiadiazole-3,5-diam ine, w here w ater w as the m ain by-product (see Figure 1). Detaiis aboul synthesis are presented in [24,25] and in Supplem entary M aterials. Tire fact that ths synthesis path of this imine proposed by us w as perform ed in a single step reaction followed by simple purification significantly lowers the production cost. The simplicity of the whyle process allows solvent recovery, i.e., ethanol, acetone or NdV-dimethylacetamine (DMA). Moreover, in the precess are not inclu det any expensive catalysts or inorganic compounds and the main by-product -water does no harm to the environment and can be reused [26].
Special attention w as paid to purification of the obtained im ine by applying various solvents [ethanol, acetone), com bined w fih crystallizstion from acetone-heeane mixture. The purification progress was monitored by thin layer chromatography (TLC) and proton nuclear magnetic resonance (1H NMR). After the reaction, the raw product contained unreacted aldehyde. The second purification, including recrystallization from acetone-hexane resulted in pure prsduct.
The symmetrical imine was characterized by XPS and XRD. Additional characteristics by 1H NMR and FTIR spectroscopy are presented in Supplementary Materials and in [24,25]. All the experimental data w ere consistent w ith else; proposed structure. The XPSi spectra of the investigated im ine are depicted in Snpplem entary M aterials ( Figure S3), w hile concentrations and binding energies of the identified functional groups are listed in Table S1. Powder X-ray diffractogram of bTAthDaz showed amorphous nature of the investigated imine with a broad peak at 2 © = 20° ( Figure S4a Shape of the TGA curves of imine indicates two main reaction stages occurring in both atmospheres. Assuming 5 wt.% loss as a criterion for assessing thermal stability of the material, this values is reached at ca. 156-159 °C, being alm ost independent on the atmosphere. Consequently, it can be stated that TGA analysis proved good therm al stability of the m aterial in both gasses. A t higher tem petatures, eeceeding ca. 300 °C in ihis partitular case, some differences can be observed between both recorded curves. Different behaoior in 300-550 °C range seems to reflect different oxygen partial pressures, suggesting that further decompoeition proceeds e asier at low p O2. On the other hand, at tire final stage of the process, above 600 °C, leek of oxygen in Ar causes some am ount of carbonized residue to lie present after whole TGA cycle, while in air, thi s residue is oxidized, with the sample being completely decomposed at ca. 650 °C (Figure 2 ). O bviously good thermal stability of the considered material is crucial concerning usage in organic solar cells.
DSC thermogrsm registered at heating for bTAThDaz is p resented in Figure S4b in Supp lementary Materials. Becausa the atudied compound is stable during cooling up to 100 °Ct which is why onty the intereating temperature range is presented. As can be seen DSC curve shows two very small anomalies at 35.1 °C and 53.8 °C w ith enthalpy changes of -0 .3 2 J/g and -0 .6 6 J/g, respectively w hich m ay be probably responsible ior crystallization process in two stages. tnfrered spactra and their assignm ent were reported in Raf. [25] t We followed the tem perature depandence of ihe epectra in nitrogen atm osphere to see it any structural transitions occur prior to therm al deguadation shat would indicate the presence of intra-or interm olecular secondary bonds.
"The spectra diet not change upon heating from room tem perature to 180 °C and being kept at this temperatute for 5 h (Figure 3 and Figure S5), showing that both the molecular and crystal structure are thermally stable.

Optical and Electrochemical Studies
The development of novel electron/hole conducting materials for p erovskite photovoltaics [4,5,27] including; m olecular conductors to replace spiro-OMeTAD is surely one of the l eading chall enges in the field. The UV-Vis absorption spectra of the investigated imine were measured in chloroform and chlorobenzene solution. and in thin film. The absorption spectra of thc bTAThDaz in chloroform solution display three partially overlapped signals; w ith m axim a at 247 nm , 297 nm and 13713 nm.
They are observed in the same range as for spiro-OMeTAD (see Figure S6 ). Figure S7 shows the optical spectrum of bTAThDaz in chlorobenzene soCuticn at different concentrations. The m elar extinc tion coeffccitnts equal e297 = 154,000 M -1 -cm-1 and e070 = 149,000 M -1 -cm- 1 . They are about two times larger as com pared to 8he extinction coeffecients of spirc-OMeTAD in the same solvent.
The/ corresponding maxima of are blue-shifted by ca. 20 nm ae referenced to those of spzro-OMeTAD [28].
Honestly, the larger optical absorbance of bTAThDaz ir an issue for inve/ted perovskite solar cell architecture (p-i-n), but not for the regular one (n-i-p).
The optical spectrum of thin-fflm of bTAThDaz is s hown in F igure 4 . Obviously, the spectrum of solutiun ( Figure S6 Table 1 and Figure S8 . The cyclic voltammogram of bTAThDaz demonstrated three oxidation processes with maxima at 0.05, 0.29 and 0.82 V and only one reduction offset at -2 .0 5 V. The lowest original cathodic peak came from thiadiazole, due to its strong aromaticity [31]. For bTAThDaz there is no difference between the thiadiazole ring and imine signals which suggest that the aromaticity of 1,2,4-thiadiazole ring expands on neighboring imine bonds. To summarize: Though peak in CV is very broad the LUMO tail of the reference spiro-OMeTAD in ER-EIS is well defined and sufficiently (0.4 V) far from -2 .9 V (reduction of ACN in the glove box with Ar protective atmosphere).

Code no*™"* E offset HOMO (eV) LUMO (eV) Eg (eV) *abs [nm] in (V) (V) CV ER-EIS CV ER-EIS CV ER-EIS
The LUM O position of -2 .2 eV recalculated vs vacuum level is in agreem ent w ith com m only reported value.
The value of -2 .3 6 V vs Ag/AgCl observed for bTAThDaz is even 0.5 V above the -2 .9 V, which warrants reliability of the presented data.
The calculation of LUMO from Egopt and HOMO CV gives value ranging -2 .5 9 V -2 .7 9 V, which is in good agreement with LUMO data obtained from CV experiment (-2 .5 2 V).
Moreover, the voltammogram for acetonitrile with BU4NPF6 showed the reduction offset at -2 .2 7 V vs. Ag/AgCl, which is 0.22 V above value for bTAThDaz in solution, hence the influence of the solvent reduction process does not overlap w ith reduction of the studied imine.
The me asured ER-EIS spectra o f the bTAThDaz film along with spiro-OMeTAD film ate shown in  Table 1 together w ith the respective Cransport gaps. Unlike CV, the ER-EIS techn5que elim in stei the parasitic currents and im htoves the sensitivith by several orders 5 l magnitude. It is worth noting that the absorbance spectra oi both films show slightly different absorption edges; (5igure 4 and Figure S6 i [It] confirming smaller band g o . fot the imine bTAThDaz film. In general, we can note reasonable agreement of the; optical bard gap Egopt (see above) and EgCV measured in solution. Values from ER-EIS on thin films are similarl y consistent for spsro-OMeTAD, but nest for bTAThDaz. A difference between optical and transport band gaps of 2hin film is coacmoo characteristic of organic semiconductors. T h t optical band gap characterizes bound elec2ron-hole pair (exciton). The transport band gap, measured with electrochemical technieues, can lie larger because the exciton binding energy has to be overcome to athieve charge transport of free polarons [33], which is the case of the bTAThDaz film.

Electrical Properties o f Imine and Simple Devices together with Surface Morphology Checked by SEM and Thermographic Camera
The surface conductivity of our thin film on quartz substrate (measured by 4-point probe) was poorly reproducible in the range of er « (10-8-1 0 -10) S/cm. This conductivity is lower than that of spiro-OMeTAD (in the undoped state) [32]. The latter is, how ever know n to increase by orders of magnitude upon doping, e.g., with LiTFSI [31] or Cu(II) pyridine complexes [34]. The conductivity of naturally doped CuSCN (which is another promising hole-conductor for perovskite photovoltaics) is ca. 10-4 S/cm but the experim ental values significantly depend on the sam ple environm ent [5]. The tw o-point conductivity m easured across the film (Au/bTAThDaz/Au) w as by several orders of magnitude better (see also below), but again poorly reproducible, perhaps due to inhomogeneity in the film (see Figure 6 and Figure S2). Obviously, doping of bTAThDaz is a challenge for future research. Therm al im aging w as carried out on devices m ade of spin-coated layer of bTAThDaz on ITO conductive glass; and with silver contact with following architecture: ITO/bTAThDaz/Ag/ITO. Figure 7 shows the obtained data. The therm al im ages prove reasonable hom ogeneity oF our organic layer, which is consisted with SEM images (Figure 6 ). Alsor it wea noticed that the organfc layer behaoed as conductor exhibiting the ohmic increase of current over increment of potential. Within tire measured range for bTAThDaz a d eisad ation was observed, w hen the temperatuse reached above 73 0C and 85 0C, respectively (see Figure 7). The resistance on 1 cm2 layer was similar and was equal 57.5 O, what is consistent with above mentioned conductivity measurements.

Perovskite Solar Cells with bTAThDaz
Finally, we tested the bTAThDaz im ine as HTM in perovskite solar cell. The photovoltaic performances of the perovskite solar cells (PSCs) using spiro-OMeTAD and bTAThDaz as hole transport materials are provided in Figure 9b . The characteristics of PSCs are summarized in Table 2

Conclusions
Selected electrochemical properties are presented for an imine based on TPA units and thiadiazole moieties as prospective, air-and thermally stable candidate of organic layer for perovskite solar cells.
Our study showed that bTAThDAz can be used as a component of HTM.
The synthesis of im ine w as perform ed in an one-step reaction in solution followed by simple purification which significantly lowers the production cost of pure product. Additionally, the simplicity of the process allows recovery of used solvent and reuse again in other processes. The absorption properties in the UV-Vis range of bTAThDaz are almost identical with the properties of spiro-OMeTAD, w hich is a prom ising characteristic. M oreover, the investigated im ine is stable in argon and air atm osphere and exhibited 5% of w eight loss at 156 °C. As for the layers based on bTAThDaz do not suffer from current direction changes, that makes this layer conductive in both directions equally and the imine does not form secondary bonds and is stable up to 180 °C. We have constructed perovskite solar cells with imine as HTM that showed PCE = 14.4%. Summarizing, constructed devices with new HTM are time stable and exhibit negligible hysteresis effect.
Taking into account all the reported results, the designed and synthesized bTAThDaz compound appear to be a promising material for use as a conductive organic layer in perovskite solar cell however additional m odifications including for exam ple doping of im ine are required to im prove charge transport properties of the investigated compound. The good photovoltaic performance of bTAThDaz means that bTAThDAz can be used as hole transport m aterial in perovskite solar cells. In this w ork the highest value of PCE was obtained for PSCs based on imine as HTM. In comparison w ith results presented in [1] w here PCE = 11.0% was found for their device w ith EDOT-OMeTPA, we increased the efficiency of PSCs w ith our im ine (bTAThDaz) by more than 30% and received PCE = 14.4%.
Furthermore, our HTM material is thermally and air stable with very small hysteresis effect. Conflicts of In terest: The au thors declare no conflict of interest.