Dye-sensitized solar cells (DSSCs) based on poly(3-hexylthioacetate thiophene)

This work details the synthesis of a functionalized all-donor polymer, poly(3-hexylthioacetate thiophene) (P3HTT). The spectrophotometry analysis of the polymer showed absorption bands at 450 and 536 nm in the liquid and solid states, respectively with optical band gap ( ) of 1.9 eV. In addition, the polymer emits in the green region of the emission spectrum with photoluminescence (PL) of 579 nm and 26% photoluminescence quantum yield (PL Φ ). The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of the polymer were found to be -5.05 and -2.9 eV, respectively. Dye-sensitized solar cells (DSSCs) devices based on this polymer were fabricated in combination with nano-crystalline titanium dioxide (nc-TiO 2 ) and ruthenium 535-bis tetrabutylammonium (TBA) dye to give devices with the configuration, glass/SnO 2 :F/nc-TiO 2 /dye/polymer/Au. The best device showed power conversion efficiency (PCE) of 0.035% with V oc , J sc and FF of 610 mV, 1.7 mA/cm 2 and 34%, respectively. 535-bisTBA, photovoltaic.


INTRODUCTION
Dye-sensitized solar cells (DSSCs) like polymer photovoltaic solar cells are another promising technology that is highly promising and cost-effective alternative for the photovoltaic energy sector. O' Regan and Grätzel (1991) introduced the first DSSC based on ruthenium(II)polypyridyl complex as the active material in conjunction with I -/I 3 liquid electrolyte and nano-crystalline titanium dioxide (nc-TiO 2 ) affording an overall power conversion efficiency (PCE) approaching 11% under standard AM1.5G illumination. This high PCE was attributed to the wide absorption range of the ruthenium(II)-polypyridyl complex, which extends from the visible to the nearinfrared (NIR) regime. As a result, a lot of effort has been devoted to the synthesis and investigation of materials for DSSCs. The synthesis can be grouped into two broad areas: a) functional ruthenium(II)-polypyridyl complexes *Corresponding author. E-mail: ajmanji2000@yahoo.com. Tel: +234(0) 8052 8080 21.  (Figure 1b). The former class contains expensive ruthenium metal and requires careful synthesis and tricky purification steps. On the other hand, the second category can be prepared cheaply, and their absorption and chemical properties can easily be tuned through suitable molecular design (Mishra et al., 2009).
Despite the relatively high PCE of DSSCs, they suffer from operational drawback. The electrolyte and its solvent are prone to degradation and evaporation, which has posed a problem with long-term performance. As a result, there have been several research efforts to replace the liquid electrolyte with semi-conducting polymers or other organic materials with the aim to solving these problems (Watanabe and Kasuya, 2005). For instance, Bach et al. (1998) reported a promising solid-state DSSC based on heterojunction of nc-TiO 2 with an amorphous organic hole transport material (HTM), 2,2′,7,7′-tetrakis(N,N-di-pmethoxyphenyl-amine) 9,9′-spirobifluorene (OMeTAD). They reported a incident PCE (IPCE) of 33% and PCE of 0.74% for the devices to which additives [N(PhBr) 3 SbCl 6 and Li(CF 3 SO 3 ) 2 N)] were added. Dittmer et al. (1999) have also reported the effect of perylenebis(phenethylimide) (PPEI) dye in DSSCs based on poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV). They investigated the photoluminescence (PL) quenching in a blend of MEH-PPV/PPEI. The PL efficiency quenching they observed for PPEI/MEH-PPV blends was more efficient than in blends of MEH-PPV and cyano-polyphenylene vinylene (CN-PPV). From the efficient PL quenching, they deduce that virtually all excitons generated in MEH-PPV were affected by the presence of PPEI. This then indicated that a continuous interpenetrating network of the polymer/dye with interface within the exciton diffusion range was formed. Therefore, charge separation was improved by introducing PPEI into MEH-PPV thin film devices. It was also observed that the external quantum efficiency (EQE) of the blend was more than fourfold for a 1:60 PPEI:MEH-PPV blend when compared to pristine MEH-PPV as a monolayer in a sandwich photovoltaic cell which shows an EQE of 0.04% under similar conditions.
In a similar work, Dittmer et al. (2000) blended poly(3hexylthiophene) (P3HT) (as a donor and hole conductor) with perylene dye [perylenetetracarboxyldiimide N,Nbis(1-ethylpropyl)-3,4:9,10-perylene bis(tetracarboxyldiimide)(EP-PTC)] (as an acceptor and electron transport material). This dye can form crystals within a polymer matrix, and such crystals are expected to have higher electron mobilities than amorphous composites and should allow much higher exciton diffusion ranges as a result of the higher degree of order in the crystals. Films of the pristine polymer and that of the dye and their blends with different weight ratios were investigated in a sandwich structure between an Aluminum (Al) top electrode and indium tin oxide (ITO) coated glass. The EQE in the pristine materials lie below 0.2%, whereas in the blend devices it reached maximum of ca. 7% at 495 nm for device containing 80 wt % of EP-PTC. Thus, there is an improvement in the EQE by a factor of 40 compared to the pristine dye and 250 compared to the pristine polymer device. This device had also shown open circuit voltage (V oc ) of 350 mV and fill factor (FF) of 41% under illumination at a wavelength of 540 nm through the ITO contact at an incident light intensity of 0.16 mW/cm 2 . This afforded a PCE ca. 0.4% at 540 nm. Furthermore, the PL was quenched by a factor of more than 10 3 for a 30:70 EP-PTC:P3HT blend with respect to pristine dye and more than 200 with respect to pristine P3HT. According to them, the observed EQE enhancement was due to additional charge separation and the possibility of energy transfer followed by radiative recombination was ruled out. They concluded that photo-induced charge transfer took place between the polymer and the dye. Gebeyehu et al. (2002) have also demonstrated the effect of dye on the performance of hybrid solar cell based on nano-porous TiO 2 and conjugated polymers; [poly(3-octylthiophene), P3OT] and three different thiophene-isothianaphthene based copolymers. The photovoltaic properties of devices which did not contain the dye cis-bis[(isothiocyanato)bis(2,2′-bipyridyl-4,4′dicarboxylato)-ruthenium(II)bis-tetrabutyl-ammonium, and those that contain it were compared. It was observed that the devices based on P3OT that contained the dye showed a PCE of 0.16%, which is two-fold higher than those without it, while the PCE of the devices based on the copolymers depended on the band gap of the respective copolymers. Zafer et al. (2005) have also reported solid-state dye sensitized nc-TiO 2 solar cells based on perylenediimide (PDI) derivative dye; N,N′ -bis-2-(1-hydoxy-4methylpentyl)-3,4,9,10-perylene bis (dicarboximide) (HMPER) with P3OT and P3HT polymers. Devices containing both polymers gave similar values of short circuit current (J sc ), V oc , FF and PCE; ca. 0.08 mA/cm 2 , 0.7 V, 0.26 and 2%, respectively at 80 mW/cm 2 AM 1.5 light intensity. Tsekouras et al. (2007) have reported the doping of poly(terthiophene) with anionic dyes during electrodeposition for the fabrication of solid-state DSSCs. The device based on sulforhodamine B dye doped polymer showed the best result with J sc = 0.178 mA/cm 2 , V oc = 318 mV, FF = 29.6% and PCE = 0.033% under white light illumination with an intensity of 500 Wm −2 . Komiya et al. (2004) have reported a highly efficient quasi-solid state dye-sensitized solar based on ion conducting polymer electrolyte, ruthenium dye [Ru (2,2′bipyridine-4,4′-dicarboxylic acid) 2 ]and porous TiO 2 . Poly(ethylene-oxide-co-propylene oxide)trimethacrylate, (oligomer) having three polymerizable reactive groups was used to form a stable quasi-solid three-dimensional polymer structure, which acted as the ion conducting polymer. According to the authors, the conductivity values of the polymer electrolyte in different organic solvents showed that the ionic conductivity increased with decreasing viscosity of the solvent and a high ionic conductivity of 9 mS/cm was observed for the polymer electrolyte. The cells were fabricated by sandwiching these materials between fluorine doped SnO 2 conducting glass and platinum coated counter electrode. By optimizing the electrolyte composition, polymer concentration and thickness of the porous TiO 2 , efficiency close to the highest value reported for liquid DSSCs was achieved. A J sc of 14.8 mA/cm 2 , V oc of 0.78 V, a FF of 70% and an overall PCE of 8.1% under AM 1.5 irradiation (100 mW/cm 2 ) were reported for the best cell. It was also observed that this cell showed higher V oc than those of the liquid cells, a phenomenon the authors attributed to the possibility of the suppression of back electron transfer between the conduction band of the TiO 2 electrode and the triiodide ion in the electrolyte of liquid DSSC.
In this study, we report the synthesis of side chain functionalized poly [3-(6-thioacetatehexyl) thiophene], (P3HTT), using 3-(6-bromohexyl)thiophene intermediate, via the post-polymerisation method. The thioacetate functional group was introduced with the aim of lowering the band gap of the polymer.

Synthesis
The key building block was synthesized according to a method developed by Bauerle et al. (1990). Briefly, 6-(4-methoxyphenoxy) hexylbromide [7], prepared from 1,6-dibromohexane and 4methoxyphenol, was reacted with Magnesium to give a Grignard compound which was coupled with 3-bromothiophene in the presence of NiDPPPCl2 catalyst producing a terminally protected hexylthiophene [8]. The intermediate, 3-(6-bromohexyl)thiophene [9] was then obtained by reacting [8] with hydrogen bromide in acetic anhydride giving an 80% yield after column chromatography. The precursor polymer, poly[3-(6-bromohexyl) thiophene] (PBHT), and its functionalization to the thioacetate derivative was done via the GRIM method (Zhai et al., 2003). Compound [9] was reacted with n-Bromosuccinimide in tetrahydrofuran (THF) and acetic acid to give 2,5-dibromo-3-bromohexylthiophene [10], as a colourless oil. Reacting [10] with Grignard reagent in freshly distilled THF in the presence of NiDPPPCl2 catalyst produced the precursor PBHT [11]. The polymerisation was terminated using methanol and the solid polymer was washed with methanol and acetone in Soxhlet extractor. The washed polymer was dissolved in chloroform in the same apparatus, and the removal of the solvent gave a 60% yield of the polymer. Reacting [11] with potassium thioacetate in THF at the reflux temperature of the solvent afforded the thioacetate functionalised derivative, P3HTT [12], with a yield of 93% (Scheme 1).

Fabrication of DSSC using Compound [12]
The DSSCs were fabricated on glass substrate coated with a transparent conducting layer of Fluorine-doped tin dioxide (SnO2:F) covered with a layer of compact TiO2∼500 nm thick. Porous layer of nc-TiO2, which was purchased in the form of sol-gel, was then applied on the compact layer using the doctor blade method (Fukuri et al., 2006). This layer served as the n-type semiconductor, while the polymer served as the p-type semiconductor and light absorber. Ruthenium 535-bisTBA dye (same structure as [2]) was used as sensitizer. Gold electrode was deposited on the active (polymer) layer through shadow mask evaporation method to form a solar cell of the type SnO2:F/nc-TiO2/dye/polymer/Au. A schematic representation of a typical DSSC is illustrated in Figure 2.
The substrate, nc-TiO2 and the dye were all purchased from Solaronix SA and the dye and the nano-crystals were used without further treatment, while the substrate was cleaned before use.

Porous layer of nc-TiO2
The nc-TiO2, which comes in the form of sol-gel, was stirred (1 min) using a clean glass rod before use. A non-stick 3 M Scotch Magic tape (50 µm thick) was used to define the area of the substrate to be coated with the sol-gel. The sol-gel (50 µl) was deposited on the edge of the substrate using a micropipette and was uniformly spread out with a clean glass rod over the surface defined by the tape (doctor blade), to make a 3 µm thick layer. The substrate was dried in air for 10 min and the tape was removed. The substrate is   now ready for the next stage.

Sintering
The dried substrate was transferred to a hot plate pre-heated to 100°C for 30 min, and was held at this temperature for 15 min. The temperature was then increased to 150°C and was kept for another 15 min. The temperature was further ramped to 250, 350 and 450°C and was maintained for 15, 30 and 30 min, respectively. During the heating process, the substrate was observed to first turn brownish and later yellowish-white due to the temperature dependent band-gap narrowing in the pure titanium dioxide (anatase) (supplier). The sintered substrate was then cooled to 70°C.

Sensitizer impregnation
The sensitizer (10 mg) was dissolved in pure ethanol (50 ml), giving a wine-red colour. The sintered substrate, kept at 70°C was slowly put into the sensitizer solution with its face up. This was stored in the dark for 2 days for proper impregnation. After this period, the substrate was removed, rinsed with pure ethanol and dried with stream of nitrogen for 3 min.

Spin coating of polymer layer
Various concentrations of P3HTT (10, 20 and 30 mg/L) were prepared in anhydrous chloroform. The solutions were sonicated at 50°C for 30 min to enhance the solubility of the polymer. The solutions were filtered through a 0.45 µm syringe filter into clean vials. The treated substrate was placed on the vacuum chuck of a spinner and enough drop of the polymer solution was placed on it. The coating was done at 1000 rpm for 60 s. The substrate was removed and spilled polymer solution was cleaned off the conducting surface of the substrate that was not covered with the porous nc-TiO2 and the dye using cotton wool soaked in chloroform.
The substrate was then kept in the dark until a gold electrode was deposited on the active layer to complete the device.

Deposition of gold electrode
A gold wire was cleaned with Decon 90® and rinsed with warm and ultrapure water and then dried in a stream of hot air. The clean gold wire was then placed in a tungsten boat in an Edwards AUTO 306 Turbo Evaporation System. A shadow mask was placed on the substrate held in place with tape and the set up was placed above the tungsten boat in the evaporation system. The evaporator was pumped to ∼10 -6 torr and a gold film, 50 nm thick was deposited through the shadow mask to form ∼2 mm 2 circular electrodesto complete the DSSC device ( Figure 2).

Nuclear magnetic resonance (NMR)
The proton NMR ( 1 H NMR) spectra of 3bromohexylthiophene [9] show signals at δ 7.28 and 6.96 ppm for the 2α and γ-hydrogen atoms on the thiophene ring (inset Figure 3), while those of Compounds [11] and [12] show only one signal at δ 6.98 ppm for the γ-H (inset a, Figures 4 and 5). This could suggest that coupling had   occurred on the α-carbon of the thiophene ring giving rise to head-to-tail (HT) regioregular polymer (inset b, Figure  4). The protons of the side chain (6-hexylbromide) in Compound [11] had remained unchanged in Compound [12] suggesting that the side chain is stable to the postpolymerization conditions. This result is similar to those reported by other workers for similar polymers (Zhai et al., 2003;Xu et al., 2008). The presence of a singlet at δ 2.32 ppm due to 3-hydrogen atoms in the spectra of Compound [12] could indicate the presence of methyl group attached to a carbonyl group ( Figure 5). Thus, the bromine atom at ω-position in Compound [11] has been successfully replaced by the acetate group. The HT or the regioregularity of P3HTT [12] was found to be 98% based on the ratio of the peaks at δ 2.58 and 2.83 ppm (inset b, Figure 4). Other workers have published similar result for P3HT which has similar structure with Compound [12] (Amou et al., 1999).

Polymerization and thermal stability of Compound [12]
The spectrum of the molecular weight and polydispersity index of the polymer which was determined using gel permeation chromatography (GPC) in THF is shown in Figure 6, while the signals for the thermal analysis are shown in Figure 7; Table 1 gives a summary of the results of the GPC and the thermogravimetric analysis (TGA). It can be inferred from these values that Compound [12] has relatively high molecular weight and good thermal stability.

Optical properties of Compound [12]
The UV-Vis spectra of Compound  chloroform solution and solid state as thin films cast from chloroform onto ITO glass (Figure 8). The absorption spectrum of the film showed a red-shift of 86 nm compared to the spectrum of the polymer in solution. This bathochromic shift suggests that the polymer exhibits more ordered molecular configuration in the solid state than in the liquid state. The optical band-gap ( ) was calculated from the band-edge of the absorption spectrum of the film. The emission maxima of the polymer lies in the green region of the emission, suggesting that it is a green emitter. The optical and PL properties of this polymer are summarized in Table 2.

Electrical properties of polymer Compound [12]
The electrochemical data of low band gap (LBG) polymer can give valuable information regarding the intrinsic material stability and allow the estimation of the relative positions of its highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels. The knowledge of these values is required for compiling the energetically compatible donoracceptor pairs in bulk heterojunction (BHJ) devices. Cyclic voltammetry (CV) is the method used to calculate these energy levels. Therefore, the cyclic voltammograms of thin film of Compound [12] drop-cast from chloroform solution were obtained and the energy levels were calculated from the values of the onset oxidation potential ) and the onset reduction potential vs. half-wave potential of ferrocene ( , 0.098 eV) using Equations 1 and 2, respectively. (1) The difference between the HOMO and LUMO energy levels gave the band-gap of the polymer. Figures 9a and  b show the cyclic voltammograms, while Table 3 presents a summary of the electrical data of the polymer.
The values of the and are in close agreement and are within the experimental error limit of 0.2 to 0.5 eV. The of the polymer is similar to that of P3HT (1.9 eV) as reported by Andersson et al. (2007) (Table 2).  This may suggest that the thioacetate functional group attached to the hexyl side chain did not lower the bandgap of the polymer below that of P3HT as expected.

Photovoltaic properties of DSSC devices based on Compound [12]
The photovoltaic properties of the devices were investigated using a Sun 2000 Solar Simulator, with light intensity of 1 KW/m 2 equivalent to 1 Sun and AM1.5 which was used to illuminate the devices. Connecting the testing device electrodes to a Keithley 2400 and sweeping the applied bias from -1 to 1 V, measured the I-V characteristics. All measurements, both in the dark and under illumination were done in air. Linear plots of typical current versus voltage (I-V) characteristic of a cell showing the best performance is shown in Figure 10.
The devices showed good rectification in the dark and photovoltaic properties were exhibited when illuminated. The devices were fabricated using 10, 20 and 30 mg polymer solutions and the I-V characteristics data is given in Table 4. A slight increase in the values of J sc and FF, thus the PCE, was observed as the polymer concentration increased from 10 to 20 mg. This could be due to increased photon absorption by the polymer layer   or an increased polymer/nano-crystal contact, which might have enhanced the dissociation of excitons at the polymer/nano-crystal interface. The decrease in these properties in the devices based on 30 mg polymer solution could be due to increased shunts and defects as a result of increased polymer thickness. These might have led to decreased charge mobility and transport, thus, the decreased efficiency. A significant decrease in the values of V oc was also observed as the polymer concentration was increased.
This may be as a result of charge recombination due to the short diffusion length of the charges. Mishra et al. (2009) and Hagfeldt et al. (2010) have stated that charge recombination lowers the V oc of DSSCs.
The overall photovoltaic performance of these devices is less than what has been reported for similar devices based on P3OT and P3HT by other workers (Gebeyehu et al., 2002). The poor performance of these devices may be as a result of poor penetration of the polymer into the nc-TiO 2 mesopores due to the bulky nature of the thioacetate functional group, in addition, the polymer and the dye-absorbed surface might not have established a good contact, thus, reducing charge separation and subsequent transport (Beaupre et al., 2010). However, Tsekouras et al. (2007) have reported similar result in solid-state DSSCs using sulforhodamine B dye.

Conclusion
The polymer (P3HTT) has shown good solubility in common organic solvents together with stable optical properties and good thermal stability. Its PL Φ is in agreement with those of its analogues, the homothiophene polymers and its emission λ max indicates that it is a good green light emitter. The result of the preliminary test of the photovoltaic properties of the polymer in DSSC solar cells has shown good photovoltaic response with the cell having 20 mg polymer affording 0.035% PCE, 34% FF, 1.7 mA/cm 2 J sc and 610 mV V oc . The low PCE could be attributed to low values of the FF and J sc .