Abstract

PEDOT-based material counter electrodes (CEs) are recently given an enormous attention as new renewable energy sources due to their cost-effectiveness and accessibility, coupled with the simplicity of production. The existing dye-sensitized solar cells (DSSCs) are expensive because they are made using platinum-based glass electrode. In this work, a new CE, that is, poly(3,4-ethylenedioxythiophene)-graphene oxide/titanium dioxide (PEDOT-GO/TiO2) with a low charge transfer resistance ( Ω), was fabricated. In addition, PEDOT-GO/TiO2 CE possesses a good electrocatalytic activity (ECA) toward the tri-iodide ions reduction and an improved efficiency of 1.166% was reached in DSSC.

1. Introduction

Since the most recent decades, dye-sensitized solar cells (DSSCs) have received great interest, as a substitute toward renewable energies due to the severe energy crisis [1]. A typical DSSC counter electrode (CE) which is platinum (Pt) metal is an expensive metal and thus limits its practical application [2]. Thus, it is critical to find new materials, which are well-appointed for replacing Pt that possesses good electroconductivity and electrocatalytic activity (ECA) toward the reduction of tri-iodide ions and cost-effective [1, 3]. Recently, researchers are interested in the utilization of Pt-free materials such as conducting polymers (CPs) [4], carbon materials [5], and transition metals based inorganic materials [6] as new CE materials in DSSCs due to their remarkable properties that can enhance the efficiency of the solar cell [7]. Poly(3,4-ethylenedioxythiophene) (PEDOT) is a conducting polymer suitable to be utilized as a CE due to its good stability, high electrochemical activity, and excellent transparency [810]. Reference [11] reported that PEDOT CE displayed the highest power conversion efficiency (PCE) of 1.35% compared to polypyrrole CE (0.41%) and polythiophene CE (0.49%) due to the fact that PEDOT CE has high ECA and low charge transfer resistance (). In addition, the high surface roughness of PEDOT observed from the FESEM images compared to other polymers also contributed to higher ECA and hence increases the PCE of PEDOT CE [11].

Graphene possesses high carrier mobility [12], high specific surface area [13], excellent thermal conductivity [14], high optical transparency [15], and high Young’s modulus [16]. Reference [17] reported that graphene is a good material to substitute Pt in DSSC because of its outstanding conductivity and high surface area that reduce . However, graphene oxide (GO) CE only yielded a PCE of 0.03% in DSSC due to the low conductivity of GO that leads to a higher series resistance ( = 46.5 Ω) [18]. Moreover, good thermal stability, high catalytic activity, and low cost are the reasons why metal oxides are also labeled as potential substitutes for Pt CE [17]. Reference [19] reported that nickel oxide doped Pt as CE yielded a PCE of 3.51%. Another work reported by Zhang et al. [20] reported that incorporation of MnO2 on PEDOT/GO and PEDOT yielded a higher catalytic activity compared to PEDOT/GO and pure PEDOT.

Herein, we present a new CE that consists of a combination of three different materials which are PEDOT, GO, and TiO2. The synergistic effect of these materials increases the efficiency of the DSSC. This new CE is Pt-free and does not require high-temperature treatment. It can be simply produced by simple doctor blade technique to coat TiO2 on ITO followed by deposition of PEDOT-GO onto it. A PCE of 1.166% was obtained for a DSSC with the PEDOT-GO/TiO2 CE compared to PEDOT (0.015%), TiO2 (0.001%), PEDOT-GO (0.683%), and Pt (0.727%) CEs. Moreover, PEDOT-GO/TiO2 CE achieved the lowest value ( = 9.0 Ω) compared to other Pt-free CEs.

2. Experimental

2.1. Chemicals and Reagents

3,4-Ethylenedioxythiophene (EDOT) (stored at 2–4°C), lithium perchlorate (LiClO4), chloroplatinic acid hexahydrate, titanium isopropoxide (TTIP), and titanium dioxide (TiO2, Degussa P25) were obtained from Sigma Aldrich. Graphene oxide (GO) was purchased from Graphenea and Sigma Aldrich, respectively. Potassium chloride (KCl), acetone, tert-butyl alcohol, acetonitrile, and ethanol were obtained from Fisher Chemical, Merck KGaA, J. T. Baker, ChemAR®, and HmbG®, respectively. Ruthenizer 535-bis TBA (N719) and Iodolyte Z-100 were purchased from Solaronix SA. Deionized water from Millipore (Mili- Q, 18.2 MΩ·cm) was utilized throughout the experiments. Indium tin oxide (ITO) glasses (7 Ω/sq) were purchased from Xinyan Technology Ltd.

2.2. Fabrication of TiO2 Thin Film Photoanode

2 g of commercial TiO2 powder (Degussa P25) was mixed with 8 ml of ethanol. The mixture was then stirred until all the TiO2 powder was dissolved. 0.16 ml TTIP was added to intensify the viscosity of the TiO2 paste. The mixture was stirred for 30 minutes and proceeded by sonication in an ultrasonic bath for 30 minutes to remove any impurity present. The prepared pastes were coated onto a conductive ITO glass using doctor blade technique. The film was then heated for 2 hours at 100°C using a hot plate. The as-prepared TiO2 thin film was then cooled. After the cooling process, the photoanode was immersed in the 0.2 mM dye solution of N719 dissolved in the equivalent ratio of acetonitrile: tert-butyl alcohol for 24 hours in order for the dye molecules to be fully absorbed. The active area was 0.25 cm2.

2.3. Fabrication of Platinum, GO, PEDOT, PEDOT-GO, and PEDOT-GO/TiO2 Counter Electrode

Pt electrode was fabricated by spin coated 50 μL of 0.2 mM chloroplatinic acid hexahydrate on ITO at 1500 rpm for 15 s and was baked at 400°C for 4 hours. The GO CE was prepared by depositing 1 mL of GO (1 mg/mL) on ITO by drop casting whereas PEDOT and PEDOT-GO CEs were prepared potentiostatically at 1.2 V for 100 s with a three-electrode system. The electropolymerization of PEDOT/GO was conducted in an aqueous solution containing 10 mM EDOT and 1 mg/mL GO. For comparison, PEDOT CE was also prepared in 0.1 M LiClO4 without the presence of GO. The silver/silver chloride (Ag/AgCl), TiO2 coated ITO, and Pt wire were used as a reference, working, and CEs, respectively.

2.4. Characterization Techniques

The electrodeposited CEs were analyzed using Shimadzu XRD Diffractometer with Cu Kα radiation ( Å) to obtain the crystalline phase of the materials. The scan range was set from 8° to 60°. Each diffractogram was compared to the reference pattern from the Joint Committee on Powder Diffraction Standards (JCPDS). Perkin-Elmer FTIR spectrophotometer coupled with attenuated total reflectance (UATR) accessory was used to identify the functional groups. The surface morphology of the CEs was evaluated using field emission scanning electron microscopy (FESEM, JEOL JSM-7600F). Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were used to study the ECA and of the CEs toward the reduction of tri-iodide to iodide. CV measurements were performed using a three-electrode system in a solution containing 1 mM tri-iodide and 0.1 M LiClO4 in acetonitrile. The reference and CE were Ag/Ag+ and platinum wire, respectively. The potential range applied was −1.0 V to 2.0 V with a scan rate of 0.05 Vs−1. EIS measurements were performed in a dark condition using a two-electrode system at open circuit potential (OCP) with AC amplitude of 0.8 V between the frequency range of 100 kHz and 1 Hz in the presence of Iodolyte Z-100. Tafel polarization measurements were conducted at a scan rate of 10 mVs−1. A symmetric cell containing two identical CEs and an electrolyte Iodolyte Z-100 was used. All electrochemical measurements were carried out using a potentiostat (Autolab PGSTAT204) equipped with NOVA software. Current-voltage measurements were performed using SCS 4200, Keithley Instruments, USA, under 1.5 A.M. (100 mWcm−2) irradiation.

3. Results and Discussion

3.1. Fourier Transformed Infrared

Figure 1 shows the FTIR spectra of TiO2, GO, PEDOT-GO, and PEDOT-GO/TiO2 CEs. The TiO2 spectrum shows a band at 545 cm−1 which corresponds to the Ti-O-Ti stretching [21]. The peak centered at 936 cm−1 belongs to O-Ti-O bonding in anatase TiO2 [22]. The GO spectrum displays a broad O-H absorption band centered at 3393 cm−1 [23]. The peaks at 1717, 1424, 1169, and 1063 cm−1 are attributed to the C=O carboxylic group, C-OH deformation, C-O-C, and C-O epoxy, respectively [24]. The FTIR spectrum of PEDOT shows main characteristic bands at 1510, 1310, and 1155 cm−1 which are attributed to the asymmetric stretching mode of C=C, C-C and the stretching mode of C-O-C, respectively [25]. The C-S bond stretching vibrations in the thiophene ring can be observed at 762 cm−1 [26]. The FTIR spectrum of PEDOT-GO shows a combination of GO and PEDOT spectra. All the peaks are seen in PEDOT and GO spectra also present in PEDOT-GO/TiO2; however, the peaks of TiO2 (Ti-O-Ti and O-Ti-O bending at 545 and 936 cm−1, resp.) are overlapped with the peak of PEDOT (C-S vibration at 634 cm−1) and GO (C-O epoxy at 838 cm−1).

3.2. Morphological Study

The morphology of PEDOT, GO, TiO2, PEDOT-GO, and PEDOT-GO/TiO2 CEs were examined using FESEM. Figure 2(a) shows the FESEM images of electropolymerized PEDOT on ITO glass substrate. The electropolymerized PEDOT shows similar morphology as reported by [27] which is a rough and dense granular shaped morphology was observed. The FESEM image of PEDOT/GO (Figure 2(c)) reveals a uniform wrinkled surface in which resembles paper-like sheet, contributing to an increase in the surface area [28]. This FESEM image is similar to the morphology of GO (Figure 2(b)); however, the wrinkled morphology of PEDOT/GO (Figure 2(c)) is more pronounced than GO (Figure 2(b)) indicating an increase in surface area of the material. As shown in Figure 2(d) the FESEM image of TiO2 reveals similar morphology as observed by [29] which is porous spherical nanoparticles that is important in the adsorption of the tri-iodide ion. The incorporation of electropolymerized PEDOT-GO onto TiO2 coated ITO to produce PEDOT-GO/TiO2 CE (Figure 2(e)) shows a mixture of porous spherical nanoparticles and paper-like sheet morphology. However, the porous spherical nanoparticles morphology of TiO2 (Figure 2(d)) is more pronounced than PEDOT-GO (Figure 2(c)). This result indicates a synergistic effect of both materials resulting in more active sites being exposed for tri-iodide reduction and subsequently achieving enhanced ECA to yield high PCE.

3.3. XRD Analysis

XRD analysis was performed to examine the crystal structure of GO, TiO2, PEDOT, PEDOT-GO, and PEDOT-GO/TiO2 and the XRD patterns are shown in Figure 3. The characteristic peaks at 25.4° (101), 37.9° (004), 48.0° (200), 55.0° (211), 64.5° (204), and 77.7° (215) are associated with the crystal planes of anatase TiO2 (JCPDS 01-073-1764) [30]. The (001) peak of GO at 10.6° is originated from the stacked graphene sheets in the GO [31]. All conducting polymers show a characteristic peak assigned to the interchain planar ring stacking at approximately 2~26° [3234]. In this work, PEDOT also shows a peak at 25.2° indexed by (107) which is due to interchange planar ring stacking (JCPDS 00-048-1449). PEDOT-GO composite possesses peaks at 10.6° and 24.8°, indicating that PEDOT-GO was successfully deposited on ITO glass whereas PEDOT-GO/TiO2 which possesses all the peak belongs to anatase TiO2, stacked graphene sheets of GO, and interchain planar ring stacking of carbon in PEDOT that indicates that all three materials were present in the PEDOT-GO/TiO2 composite.

3.4. Cyclic Voltammetry

In order to study the ECA properties, CV measurements were carried out and the cyclic voltammograms of platinum, TiO2, PEDOT, PEDOT-GO, and PEDOT-GO/TiO2 CEs are shown in Figure 4. A pair of redox peaks was noticed for platinum, PEDOT-GO, and PEDOT-GO/TiO2 films. The peak in the range 0.75–1.5 V indicates the oxidation of iodide to tri-iodide (see (1)) and the peak in the range of −0.25 to 0.5 V indicates the reduction of tri-iodide to iodide (see (2)).The current densities of two pairs of oxidation and reduction peaks for PEDOT-GO/TiO2 are higher than PEDOT-GO and platinum, demonstrating excellent electrocatalytic behavior [35]. The reaction rate of the catalyst for the reduction of ions in the electrolyte can be influenced by the cathodic peak current density in a CV curve. An enhanced ECA for the catalytic material can be obtained by a higher absolute value. The value of PEDOT-GO/TiO2 (−5.49 mA·cm−2) is higher than the platinum (−3.57 mA·cm−2) and PEDOT-GO (−3.54 mA·cm−2), suggesting that the ECA toward the / redox reaction is greater for PEDOT-GO/TiO2 compared with PEDOT-GO/TiO2 and PEDOT-GO. These results are in agreement with obtained from EIS measurements and DSSC with higher efficiency.

3.5. Electrochemical Impedance Spectroscopy

In the DSSCs, the electrolyte of / plays a vital role to preserve the regeneration cycle of the electron. The TiO2 nanoparticles absorb dye molecules and the photon from the sunlight excites the electron of the dye molecules to the CE through an oxidation process. This oxidation process of dye produces a hole in TiO2; hence, the electrolyte must provide electrons to be transferred back to occupy the holes of TiO2, which allows the electron of the dye molecule to be generated [36]. Thus, the ions should continuously undergo reduction at the CE, as shown in reaction (2). A symmetric cell composed of two identical electrodes was used to perform the EIS analysis in order to study the ohmic series resistance () of the substrate and its catalytic layer and the charge transfer resistance () for the reduction of at the electrode/electrolyte interface. Figure 5 shows the Nyquist plots of platinum, TiO2, PEDOT, PEDOT-GO, and PEDOT-GO/TiO2 CE. The value of PEDOT-GO/TiO2 (22.2 Ω) is lower than PEDOT-GO (30.6 Ω) and platinum (175.8 Ω), implying better electric conductivity of the former. A smaller value of (9 Ω) is also observed for PEDOT-GO/TiO2 compared to platinum (19 Ω), indicating good ECA toward the reduction offor PEDOT-GO/TiO2 CE. This result is in agreement with the CV analysis. Figure 6 shows the equivalent circuit used for fitting the Nyquist plots of all CEs. The circuits include the ohmic series resistance (), constant phase element (CPE) that represents the irregularity of sample, and . It was observed that GO has the highest (Table 1) due to the low conductivity followed by TiO2 and PEDOT. However, after PEDOT is incorporated with GO, the value of PEDOT-GO drops to 21 Ω, indicating faster electron transfer rate due to good charge propagation behavior and low contact resistance [37].

3.6. Photovoltaic Performances of DSSCs

Figure 7 shows the current-voltage (-) curves of the DSSC prepared using platinum, TiO2, PEDOT, PEDOT-GO, and PEDOT-GO/TiO2 as CEs under AM 1.5 illumination (100 mW/cm2). The open circuit voltage (), short-circuit current (), maximum power (), fill factor (FF), and overall conversion efficiency () are summarized in Table 2. Apparently, TiO2 CE yielded the lowest power conversion efficiency (PCE) with 0.001% due to slow electron transport and low electron mobility causing recombination process to occur in the electrolyte, hence generating low efficiency [38]. Although PEDOT has excellent conductivity and chemical stability, it also produces a low efficiency (0.015%) that could be due to its small surface area [39] and subsequently resulted in low active site for reduction of iodide to tri-iodide. In order to overcome this problem a high surface area carbon-based material which is graphene oxide (GO) is incorporated with PEDOT to form a synergistic effect to obtain a higher efficiency of DSSC [40]. As a result, the PCE of PEDOT-GO CE is enhanced to 0.68%, which is 45.5 times higher than PEDOT CE. In addition, TiO2 as a source of metal oxide was introduced to PEDOT-GO to increase the performance of DSSC. Even though TiO2 itself has low efficiency (0.001%), the PCE of TiO2 incorporated with PEDOT-GO composite gives a promising result with the PCE of 1.166%. The PCE of PEDOT-GO/TiO2 CE is higher than platinum CE (0.727%) and comparable with conducting polymer based CEs reported in the literature [41, 42]. The novel PEDOT-GO/TiO2 CE produced in this work is a promising CE due to an enhance PCE observed in PEDOT-GO/TiO2 CE compared to PEDOT-GO, PEDOT, and TiO2 CEs. It gives a clear proof that PEDOT-GO/TiO2 is capable of producing a synergistic effect that utilizes the advantages of each material.

The maximum power densities of CEs are displayed in Figure 8. The value can be calculated from the photocurrent-voltage graph based on From the graph, PEDOT-GO-TiO2 clearly shows the domination on the ability to produce maximum power compared to other CEs which is 1.166 mW/cm2 due to excellent ECA of PEDOT-GO/TiO2 toward the reduction of iodide to tri-iodide. The maximum power produced by platinum, PEDOT-GO, PEDOT, and TiO2 CEs is 0.727, 0.683, 0.015, and 0.001 mW/cm2, respectively. Due to the high conductivity of PEDOT and high surface area of GO, the maximum power yielded by PEDOT-GO composite is higher than a single compound PEDOT and GO. It is obvious that the maximum power and of the PEDOT-GO/TiO2 CE are higher than those of PEDOT-GO CE, which can be corroborated by the data from Tafel polarization curves.

3.7. Tafel Polarization Analysis

Tafel polarization curves were analyzed using the dummy cells to study the interfacial charge transfer behavior of CE/electrolyte and the corresponding results are shown in Figure 9. is inversely proportional to exchange current density (); thus in order to yield a higher a good CE should have lower . can be determined from the intersection of the tangent line of the polarization curve and the extension of the linear segment to the zero bias using [43]where and are constant, is the room temperature, and is the number of electrons involved in the reaction. The limiting diffusion exchange current density () can also be analyzed from Tafel plot using [44]where is the coefficient of tri-iodide, is the spacer thickness, is the number of electrons involved in the reduction of tri-iodide at the electrode, is the Faraday constant, and is the tri-iodide concentration.

depends on the diffusion coefficient of / redox couple. From the Tafel plot, PEDOT-GO/TiO2 exhibits a higher (1.625 mA/cm2) compared to PEDOT-GO (1.023 mA/cm2) and platinum (0.4627 mA/cm2). A lower value of PEDOT-GO and platinum compared to PEDOT-GO/TiO2 reveals that PEDOT-GO and platinum have a marginal edge over PEDOT-GO/TiO2 in terms of improving ECA; nonetheless, PEDOT-GO/TiO2 CE provides sufficient ECA. values also show a similar trend as of the values where PEDOT-GO/TiO2, PEDOT-GO, and platinum yielded 0.332, −0.280, and −1.015 mA/cm2, respectively. A higher reveals that PEDOT-GO/TiO2 possesses high ECA toward the reduction of tri-iodide to iodide, whereas a higher indicates good contact and better diffusion of electrolyte within the CE [43]. A comparison of the Tafel curves between PEDOT-GO/TiO2 and PEDOT-GO CEs indicates that the former has a higher catalytic activity compared to the latter. Note that the / diffusion capability in the PEDOT-GO/TiO2 CE is more prominent than that in the PEDOT-GO CE.

4. Conclusion

In summary, a novel PEDOT-GO/TiO2 CE for DSSC was successfully fabricated using a facile technique. TiO2 incorporated with PEDOT-GO showed that the electrochemical properties of the final device were affected by catalytic activity layer due to the synergistic effect produced by the combination of three different materials. By applying the PEDOT-GO/TiO2 composite film as the CE, an efficiency of 1.166% was achieved for the DSSC, which is 58.6% higher than that of a cell based on the PEDOT-GO CE. An excellent enhancement in the conductivity and good electrochemical conductivity for the reduction of to were obtained for PEDOT-GO/TiO2 CE due to the high conductivity of PEDOT, high surface area of GO, and high active area of TiO2 nanoparticles. Hence, PEDOT-GO/TiO2 CE is a promising material as a CE for DSSC.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors gratefully acknowledge the financial support by the Fundamental Research Grant Scheme (01-02-13-1388FR) from the Ministry of Education Malaysia.