Effect of Deoxycholic Acid on the Performance of Liquid Electrolyte Dye-Sensitized Solar Cells Using a Perylene Monoimide Derivative

The effect of coadsorption with deoxycholic acid (DCA) on the performance of dye-sensitized solar cell based on perylene monoimide derivative (PCA) as sensitizer and liquid electrolyte had been investigated. The current-voltage characteristics under illumination and incident photon to current efficiency (IPCE) spectra of the DSSCs showed that the coadsorption of DCA with the PCA dye results in a significant improvement in short circuit photocurrent and slight increase in the open circuit photovoltage, which lead to an overall power conversion efficiency. The enhancement of short circuit current was attributed to the increased electron injection efficiency from the excited state of PCA into the conduction band of TiO2 and charge collection efficiency. The current-voltage characteristics in dark indicates a positive shift in the conduction which also supports the enhancement in the photocurrent. The coadsorption with DCA suppressed charge recombination as indicated from the electrochemical impedance spectra and thus improved the open circuit photovoltage.


Introduction
and black dye have achieved PCEs up to 11%, due to their intense and wide range absorption of visible light [7][8][9][10]. Dye-sensitized solar cells (DSSCs) represent one of the most However, due to the relatively high cost and environmental promising candidates for developing new renewable energy concerns associated with the use of Ru-based dyes have sources due to their low cost and high efficiency [1][2][3][4][5][6].
inspired strenuous efforts to develop metal-free dyes as Typical DSSCs consist of a dyed semiconductor TiO 2 photoethe alternative sensitizers [11][12][13][14]. Moreover, high molar lectrode, a Pt counterelectrode, and redox electrolyte. Among extinction coefficients of metal-free dyes allow the use of these components, the sensitizer, that is, dye, is one of the thinner TiO 2 film, which is beneficial for charge separation key components that affects the overall power conversion [15,16]. Another important advantage over Ru-based dyes efficiency (PCE) of the DSSC. The dye must absorb sunlight is the availability of versatile functional molecules associated over a wide region of visible and near infrared (NIR) and with tuning the electronic and optical properties. Recently, inject the photoexcited electrons into the conduction band of the DSSCs performance based on organic dye has been the semiconductor in the photoelectrode. The typical dyes at remarkably improved and reached impressive PCE more present times are Ruthenium-based dyes, such as N3, N719, than 9% [15] and 11% by porphyrin dyes [17,18]. Perylene dyes are well known as the key chromophores among the metal-free dyes [19][20][21][22][23][24][25][26]. Their solubility, absorp tion, and emission behavior can be efficiently controlled using a variety of synthetic procedures, which include func tionalization of peri-or bay-positions of perylene core. Moreover due to their outstanding photophysical and photo chemical stability as well as their high fluorescence quantum yields, perylene derivatives have been used as active compo nents for application in organic electronic devices [27][28][29][30][31][32][33][34][35]. Systematic tuning of HOMO and LUMO levels of perylene dyes improve both light harvesting properties and electron injection capabilities to TiO 2 conduction band for obtaining the high PCE. Several perylene dyes have been used as sensitizers in DSSCs [36][37][38][39][40][41][42][43][44][45]. Edvinsson et al. reported that by attaching a bulky nonplanar di-p-tert-octylphenylamino and an anhydride moiety to the perylene core, which not only suppresses the aggregation of dye molecules but also improves the stability of the organic sensitizer [46]. With two additional phenylthio groups in the 1,6-positions of the perylene unit, a new perylene sensitizer yields 87% incident monochromatic photon-to-current conversion effi ciency (IPCE) and 6.8% power conversion efficiency under standard AM 1.5 solar conditions [47].
Another strategy for enhancing the PCE is to incor porate donor-linker-acceptor (D-π-A) into the framework. Introducing D-π-A groups increases intramolecular charge transfer from donor to acceptor side of the molecule by π linker, which results a strong electron transfer from excited state of dye molecule to the conduction band of TiO 2 .
In the present paper, perylene monoimide (PCA) has been employed as sensitizer for nanocrystalline TiO 2 DSSC. PCA contains an electron-donating cyclohexyl ring, which is connected with perylene core act as electron acceptor via imide nitrogen. Moreover, PCA carriers bulky alkylphenoxy groups at the 1,7 bay positions of perylene core, the alkyl chains of which enhance the solubility of the compound. Finally, PCA contains an acid anhydride as anchoring group for TiO 2 surface. We have used PCA as sensitizer for the DSSC based on TiO 2 nanocrystalline photoanode and liquid electrolyte. The effect of deoxycholic acid (DCA) as coad sorbent on the photovoltaic performance of DSSCs based on PCA had been investigated. The improvement in the overall PCE is attributed to enhancement in both J sc and V oc . The enhancement in the J sc of DSSC upon the DCA coadsorption was attributed to the increased efficiency of electron injection and/or charge collection efficiency. The coadsorption of DCA suppressed charge recombination and thus improved V oc .

Experimental Part
The synthesis of PCA has been reported by us in our earlier communication [48]. The chemical structure of the PCA is shown in Scheme 1. The UV-visible absorption spectra of the dye-loaded TiO 2 film was recorded on a Shimdzu UV-visible spectrophotometer. The cyclic voltammograms were mea sured with a three electrode electrochemical cell on poten tiostat/Galvanostat PGSTAT30 electrochemical analyzer. The dye-loaded TiO 2 film, platinum and Ag/Ag + (0.01 M AgNO 3 + 0.1 MTBAP in acetonitrile) were employed as working, counter, and reference electrodes, respectively. The sup porting electrolyte was 0.1 M LiClO 4 in acetonitrile. The potential of the reference electrode is 0.49 V versus normal hydrogen electrode (NHE) and is calibrated with ferrocene immediately after cyclic voltammogram measurement.
TiO 2 paste was prepared by mixing 1 g of TiO 2 powder (P 25 , Degussa), 0.2 mL of acetic acid, 1 mL of water. Then 60 mL of ethanol was slowly added while sonicating the mixture for 3 h. Finally, Triton X-100 was added and a well dispersed colloidal paste was obtained (TiO 2 ). The whole procedure is slow under vigorous stirring. The mixture was stirred vigorously for 2-4 h at room temperature and then stirred for 4 h at 100 • C to form a transparent colloidal paste. The TiO 2 paste was deposited on the F-doped tin oxide-(FTO-) coated glass substrates by the doctor blade technique. The TiO 2 coated FTO films were sintered at 450 • C for 30 min. After cooling to room temperature, the electrodes were impregnated in 0.05 M titanium tetrachloride aqueous solution, washed with distilled water, and again sintered at 450 • C for 30 min followed by cooling to room temperature. The thickness of the TiO 2 layer is about 12 μm. PCA dye solutions (0.5 mM) with and without DCA (10 mM) were prepared in THF and used to sensitize the TiO 2 electrodes by immersing the TiO 2 films in these dye solution for 24 h and then washed.
A thin Pt layer was deposited on an FTO conducting glass by thermal pyrolysis of H 2 PtCl 4 in isopropanol solution and then heated at 450 • C for 30 min in air. One drop of elec trolyte solution (0.6 M 1-propyl-2,3 dimethyl-imidazolium iodide DMPImI, 0.1 M LiI, 0.05 M iodine, and 0.5 M tetra butyl pyridine in acetonitrile) was deposited onto the surface of dye-sensitized TiO 2 electrode and penetrated inside the TiO 2 via capillary action. The Pt-coated FTO electrode was then clipped onto the top of TiO 2 working electrode to form the complete DSSC. The active area of each DSSC thus prepared was about 0.4 cm 2 .
The current-voltage characteristics of DSSCs were mea sured on a computer controlled source meter (Keithley, 2400). An AM 1.5 solar simulator with 150 W xenon lamp and an AM filter were used as the light source. Light intensity corresponding to AM 1.5 (100 mW/cm 2 ) was calibrated using a standard silicon solar cell. Monochromatic incident photon-to-current conversion efficiency (IPCE) for the solar cell, plotted as a function of excitation wavelength, was re corded on a monochromator system. The output current under short circuit condition at each wavelength was mea sured using the Keithley electrometer.
Electrochemical impedance spectra for DSSCs in dark and under illumination were measured with electrochemical analyzer, equipped with FRA software. The spectra were scanned in the frequency range of 0.01-10 5 Hz at room tem perature with applied potential set at open circuit voltage. The magnitude of modulation signal was set at 20 mV. Figure 1 depicts the optical absorption spectra of the PCA in THF solution and PCA adsorbed into the TiO 2 nanocrys talline film. The absorption curve in solution shows an optical absorption maximum around 510 nm and is broad extended up to 680 nm. Therefore the PCA exhibits very nice absorption coverage over the visible and near infrared (NIR) region, which is desirable for the photovoltaic application. The absorption onset is located around 645 nm that corre sponds to an optical band gap around 1.92 eV.

Results and Discussion
When PCA dye was sensitized on TiO 2 surface, the ab sorption maxima was redshifted by about 20 nm in compar ison to those in solution (Figure 1), due to the J aggregation on the semiconductor surface [23,49]. To evaluate the possibility of electron injection occurring from the excited state of dye molecules to the conduction band of TiO 2 , the electrochemical oxidation and reduction potentials of the dyes were measured by cyclic voltammetry (CV) using three electrode cells.
The oxidation potential of the ground state of PCA, which corresponds to their HOMO level is 1.05 V versus NHE. This value is sufficiently more positive than the I − /I 3 − redox potential (0.4 V versus NHE) and indicates that the oxidized PCA formed after the injection of electrons into the conduction band of TiO 2 could accept the electrons from I − ions thermodynamically. The reduction potential of PCA dye, which corresponds to their LUMO level is −0.95 V versus NHE, which is more negative that the TiO 2 conduction band edge (−0.5 V versus NHE). The energy gap between the LUMO level of PCA and the conduction band level of TiO 2 is 0. 0.45 V. Since the energy gap of around 0.2 V is necessary for efficient electron injection [50], the energy gap values of the dyes were above this value indicates that the excited electrons of the dye molecule can be injected into the conduction band of TiO 2 thermodynamically.
The current-voltage characteristics of the DSSCs based on PCA with and without DCA was measured at 100 mW/ cm 2 and shown in Figure 2(a). The photovoltaic parameters are complied in Table 1. The DSSC based on PCA without DCA coadsorbent gave a J sc of 8.7 mA/cm 2 , V oc of 0.63 V, and an FF of 0.56, corresponding to overall PCE of 3.07%. Under the same conditions, when DCA is incorporated into the dye solution, the DSSC showed a J sc of 10.8 mA/cm 2 , V oc of 0.68, and FF of 0.61, resulting an overall PCE of 4.48%. Figure 2(b) compares the photocurrent action spectra of the DSSCs before and after coadsorption with DCA, where incident photon to current efficiency (IPCE) is plotted as a function of wavelength. The IPCE values of DSSC based on PCA with CDA are higher than those for PCA without DCA, in the whole wavelength region. Because IPCE is a product of electron injection efficiency, light harvesting efficiency, and   band enlarge the driving force for electron injection from the excited state of dye into the conduction band of TiO 2 which results in the values of IPCE and J sc . In addition, coadsorp tion can break up the dye aggregation, and the nonaggre gated dye molecules are favorable for electron injection.
In DSSC the value of V oc is theoretically the difference between the Fermi level of TiO 2 under light and the redox potential of the redox couple I − /I 3 − , the positive shift of conduction band edge upon coadsorption with DCA will result in a decrease in V oc . However, we have observed a charge collection efficiency, it is interesting to analyze which factor affects IPCE and J sc . We have estimated the amount of adsorbed dye on the TiO 2 surface by desorbing the dye with basic solution and found that the surface concentration of PCA with and without DCA is about 4.6 × 10 −7 and 1.5 × 10 −7 mol/cm 2 , respectively. Upon coadsorption with DCA, although the amount of dye adsorbed on the TiO 2 surface is reduced, the IPCE spectra become broader with maximum IPCE improved from 60% to 71%. It is observed that the dye loading reduced on the coadsorption, resulted the reduction in LHE. Therefore, the IPCE improvement is attributed to the enhancements of electron injection efficiency and charge collection efficiency.
It has been already reported in literature that upon coadsorption with DCA, the TiO 2 surface is protonated and conduction band edge shifts positively due to the proton adsorption [12,50,51]. This positive shift of conduction slight increase in the V oc after coadsorption with DCA that is attributed to the suppression of charge recombination. To understand the role of DCA in improving performance parameters, effect of DCA on dark current was studied as shown in Figure 3. It is observed that the dark current onset potential shifted to a larger value, and also the dark current was also reduced upon the DCA adsorption. The addition of DCA in the dye solution, the surface properties of TiO 2 nanoparticle thin film layer by the passivation effect had similar result as also reported by De La Graza et al. for the modified TiO 2 layer by enediol ligands [52]. This effect of surface states leads to shifts the flat band potential and to decrease the charge recombination. The shift of the onset potential of dark current (Figure 3), in the presence of DCA in the dye solution, can also reflect less participation of surface state in the charge transfer and leads to suppression of charge recombination between the injected electrons and I − ions in the electrolyte, favorable for improvement in 3 V oc . The enhancement of V oc is usually associated with the negative shift on conduction band edge or suppression of charge recombination. Consequently, suppression of charge recombination may compensate for the V oc loss because of proton exchange from DCA to the TiO 2 surface, resulting in V oc improvement.
To further classify the DCA effect on the photovoltaic parameters, electrochemical impedance spectrum (EIS), which is a powerful tool to elucidate the electronic and ionic transport processes in DSSCs was measured in dark under a forward bias of −0.65 V. Three semicircles were observed in the Nyquist plots (Figure 4(a)). The large semicircle in the Nyquist plots located in the middle is attributed to the dark reaction impedance caused by charge transportation at the TiO 2 /dye/electrolyte interface, and other two small semicircles located in the low and high frequency regions are assigned to the charge transfer at counter electrode and diffusion of I 3 − in the electrolyte, respectively [12,[52][53][54][55]. It can be seen from Figure 4(a) that the radius of the semicircle in middle frequency region is larger for the DSSC based on DCA coadsorbent PCA that that for PCA only, indicating that the electron recombination resistance is higher for former than that for latter. The higher charge recombination resistance observed for the DSSC with DCA coadsorbent also indicated that there exists efficient suppression of back reaction of injected electron with I 3 − ions in the electrode, which is reflected from the improvement in the both V oc and J sc , yielding substantially enhanced overall power conversion efficiency. The electron lifetime derived from the Bode plots of EIS spectra (Figure 4(b)) is 12.7 ms and 18.3 ms for PCA-and DCA-PCA-based DSSCs, respectively. The longer electron lifetime in the TiO 2 film for the DSSCs sensitized with DCA coadsorbent as compared to that for with DCA also supports the higher value of IPCE and PCE.

Conclusion
We have investigated the photovoltaic response of the DSSC based on PCA dye using TiO 2 nanocrystalline TiO 2 photoanode and liquid electrolyte and achieved overall PCE of about 3.07%. The effect of coadsorption of the DCA on the photovoltaic performance of DSSC based on PCA dye was also studied. We found that DCA improves both J sc and V oc . As a consequence, overall efficiency has been enhanced up to 4.48%.