ISSN 0003 701X, Applied Solar Energy, 2013, Vol. 49, No. 1, pp. 4045. Allerton Pr ess, Inc.

., 2013. HELIOTECHNICAL MATERIALS SCIENCE Effect of Organic Dyes on the Performance of ZnO Based Dye Sensitized Solar Cell s1 I. B. Karkia, J. J. Nakarmia, P. K. Mandalb, and S. Chatterjeeb b Tribhuvan University, Nepal University of North Bengal, India Received July 25, 2012 a AbstractThe present paper reports the synthesis and performance study of ZnO base d Dye sensitized solar cells (DSSC). ZnO nanorods were fabricated by solgel techn ique using. Three types of DSSCs were fabri cated by two different classes of or ganic dyes using, namely xanthenes, anthocyanins and a mixed dye with equal prop ortion mixture of xanthenes and anthocyanins. The absorbance spectra of dyes wer e compared with performances of the cells. The parasitic resistances of the cell s were determined from the IV character istics of the cells and various parameter s of DSSCs like Ideality factor, Fill factor and Energy conversion effi ciency w ere determined. Among the three kinds of dyes tested, the efficiency of cell wit h mixed dye was inter mediate but its value was found to be higher than the aver age efficiency of cells with two individual dyes. This is due to the improvement in ideality factor of the cell. DOI: 10.3103/S0003701X13010052 1 INTRODUCTION The solar energy conversion and electric energy storage are becoming most import ant techniques towards issues on energy crisis and sustainable use. For the sola r energy conversion, as mostly by the form of crystalline solar cell, dye sensit ized solar cells (DSSCs) with nanocrystalline ceramics, dye mole cules, and elec trolytes are recently developed for light harvesting [1, 2]. Solar cell producti on has grown at about 30% per annum over the past 15 years. In recent years, gre at attention has been paid to DSSC's due to their low fabrication cost as a viable alternative technology for renewable energy [1]. It has attracted much attentio n due to their environment friendly nature, high effi ciencies and potentially l ow production costs. Organic dyes have been intensively investigated due to thei r potential using in various low cost, large area DSSCs device application [3, 4 ]. Normally TiO2 nanostructures were used to fabri cate DSSC cell. However, ZnO has shown a great deal of research interest in DSSCs due to some of its fasci na ting properties. ZnO is often utilized as a photo anode material in DSSCs due to its attractive proper ties such as large exciton binding energy, wide direct ba nd gap (3.27 eV), which is generally used in organic solar cells [2]. ZnO is als o highly transparent, which allows for greater light penetration. Finally, 1D sin gle crystal structure formation is possible with ZnO, enabling a higher surface to volume ratio for greater dye loading. 1 The article is published in the original. To increase the conversion efficiency of ZnO nan orod based DSSCs, it would be d esirable to eliminate the interface between ITO and the ZnO nanorods [47]. Some r esearchers [8] have recently taken up this concept by growing ZnO nanorods on a ZnO film using a two step method. In this work we grew ZnO nanorods on an ITO co ated glass using solgel spin coating technique. DSSCs are a promising low cost, g reen energy source [9, 10]. A power conversion efficiency of 10.6% has been achi eved [1]. DSSC is a device for the con version of visible light into electricity . The absorption spectrum of the dye and the anchorage of the dye to the surface

of ZnO are important parameters deter mining the efficiency of the cell. Genera lly, transition metal coordination compounds (ruthenium poly pyridyl complexes) are used as the effective sensitizers, due to their intense charge transfer abso rption in the whole visible range and highly efficient metal to ligand charge tr ansfer [1113]. However, ruthenium polypyridyl complexes contain a heavy metal, wh ich is undesirable from point of view of the environmental aspects and high cost [10]. Moreover, the process to synthesize the complexes is complicated and cost ly. Alternatively, natural dyes can be used for the same purpose with an accepta ble effi ciency [3, 1417]. The advantages of natural dyes include their availabil ity and low cost [17]. The sensiti zation of wide bandgap semiconductors by natu ral pig ments using is usually ascribed to anthocyanins. This makes electron tra nsfer from the anthocyanins mole cule to the conduction band of ZnO [15]. As rep orted [3, 17], anthocyanins from various plants gave different 40

EFFECT OF ORGANIC DYES ON THE PERFORMANCE OF ZnO 41 sensitizing performances. However, there is no accept able explanation behind th ese results, so far. A natural organic dye, pomegranate juice (Punica Granatum L .) is a common source of Anthocyanin and is commonly used for fabricating DSSC's. Rose Bengal dye is one of the best photo sensitizer for ZnO photoanode to date a nd is much cheaper than Rucomplex dyes [18]. It is in xanthene class which absorb s wide spectrum of solar energy and energeti cally matches the ZnO and usual KII2 redox couple for DSSCs applications [11]. The performance of DSSCs by the mixed xanthene dyes using was also investigated by some earlier work ers [19, 20]. Tw o xanthene having different absorption characteristics would give even more syne rgistic effect compared to the mixed xanthene chlorophyll dye was reported [4]. This is because xan thene has advantages over chlorophyll as DSSC sensitizer [17 ]. Two kinds of dyes, namely, metal organic complexes (e.g., Ru complexes) [2122] and metalfree organic dyes, [19, 20, 23] have been widely used as sensitizers of DSSCs. Organic dyes have several advantages as photo sensitizers for DSSCs: (1) Larger absorption coeffi cients than metal complex photo sensitizers; (2) vari e ty in their structures provides possibilities for molec ular design, e.g., the i ntroduction of substituent, and thus allows for easy control of their absorption spectra; (3) inexpensive because they do not contain noble metals like rutheniu m. This reduces the overall cost of the cell production. In this study, a natura l organic dye of class Antho cyanin, pomegranate juice was investigated in fabri cating ZnO nanorods based DSSC's as these dyes are abundant in tropical countries [18] resulting in low fabrication cost. Another cell is fabricated with Xan the ne class of dye, Rose Bengal, and a third cell was fabricated with a mixed Dye p repared by mixing equal proportions of Rose Bengal and Pomegranate juice. Finall y the qualities of the fabricated cells were inves tigated in terms of standard cell parameters including its efficiency. The effects of dyes are discussed in t erms of their UVVis absorption spectra. EXPERIMENTAL ZnO nanorod was synthesized by solgel method using and fabricated by Spin coating technique using. Spin coati ng is a simple method for preparing ZnO nanoseed from Zinc acetate solution. In this process, we prepared 5 mM solution of Zinc acetate dehydrate, (CH3COO)2Zn, 2H2O, (98% Merck) with methanol. The solution was spun on indium tin oxide (ITO) coated glass substrates at 1000 rpm for 30 s. The Zinc acetate solution is sprea d on a rotating substrate [3, 17]. The substrates were heated to 350C in con vent ional oven for 30 min to yield layers of ZnO islands with their (100) plane para llel to the substrate surface. After evaporation of solvent, a thin ZnO film was APPLIED SOLAR ENERGY Vol. 49 No. 1 2013 formed. Repetition of the above process 5 to 7 times is carried out to control t he thickness of the film. Con centration of the solution and spinning speed of t he substrate also play important roles in adjusting the thickness of the fabrica ted film. The counter electrode (cathode) is prepared on another ITO coated glas s by amorphous carbon dust using. Characterization of Dye and Dye Deposition The device was then immersed in a solution of sen sitized dye for 24 h to allow the dye molecules to covalently bond to the surface of the ZnO. The sample were the n rinsed with ethanol to remove excess dye on the surface and air dried at room temperature. The dye was dissolved in distilled water medium with a concentratio n in the order of 1.00 105 g/mL and the absorption spectra of dyes were recorded using a UVVis spectrophotometer (Perkin Elmer Lamda 35 model UVVis). DSSC Assemblin g DSSCs were assembled following the procedure described in the literature [3]. The carbon dust coated counter electrode was placed on the top so that the condu ctive side of the counter electrode faces the ZnO film with a Teflon spacer (thi ckness around 100 m) in between of them. The iodide based solution as the liq uid electrolyte (0.5 M potassium iodide mixed with 0.05 M iodine in waterfree ethyle ne glycol) was placed at the edges of the plates. The liquid was drawn into the space between the electrodes by capillary action. Two binder clips were used to hold the elec trodes together. The effective cell area of the cell was 2 2 cm2.

Device Characterization and Measurement Apparatus setup. Keithley model 2400 dig ital source pico ammeter was used to measure the dark light and illuminated IV cu rves of the DSSC under white light illumination (Xenon lamp) condition were reco rded during the efficiency measurement experi ments. The position of the light s ource was adjusted such that the light intensity is 100 mW/cm2 (the equiv alent of one sun) at AM 1.5 of power was delivered to the surface of the measured DSSC solar cell. The cur rent voltage characteristics of DSSCs were obtained. Charac teristics of dye sensitized solar cells: The currentvoltage characteristics of a cell in the dark and under illumination permit an evaluation of most of its phot ovoltaic performances as well as its electric behavior [24]. The short circuit c urrent Isc is the one which crosses the cell at zero applied voltage and it is a func

42 I Rs Id Rsh V Iph Ish KARKI et al. 100 nm Fig. 2. Scanning electron microscope (SEM) picture of ZnO nanorod used in the dy e sensitized solar cell. Fig. 1. Equivalent circuit diagram of DSSC. tion of illumination. Charges travel under an internal potential difference typi cally equal to Voc. The open circuit voltage is measured when current in the cel l is 0, corresponding to almost flat valence and conduction bands; Imax and Vmax values are defined in order to maximize the power |Imax Vmax|. This is the maxi mum power Pmax delivered by the cell. The fill factor FF is the ratio of the max imum power to the external short and open circuit values: FF = P max I V = max m ax . V oc I sc I sc V oc (1) RESULTS AND DISCUSSION Structure Characterization The samples morphology was obs erved by a scan ning electron microscope (SEM) using with a field emission gun o perating at 200 kV. Figure 2 displays ZnO nanowire arrays in a wide surface area . The nan orods have an average length of 600 nm, diameter ranging from 100 to 2 00 nm and they are mostly verti cally aligned with the substrate having hexagona l shapes. The thickness of ZnO film was around ~2 m. The nanowires, which make ba rrier free contact with the substrates, exhibit resistivity around 0.3 to 2.0 cm along the long axis [23]. The nanowire used in the present work had a resistivi ty 0.7 cm at 0 V. Due to absence of interfaces in nanowires, resistivity of a na nowire should be lower than nanoparticle thin films. Moreover, conductivity in t he nanowire arrays increases by 5 to 20% when they are soaked in standard DSSC e lectrolytes [23]. In general, the energy conversion efficiency of ZnO DSSCs is l ower than that of TiO2 DSSCs [2123]. However due to the barrier free contact, ZnO nanow ire is expected to give higher efficiency. Because only about 4% of the s olar spectrum falls in the UV region, ZnO semiconductor absorbs solar radiation while sen sitized by natural dye molecules with absorption spec tra at visible r egion. Anthocyanin dyes present in Pomegranate Fruits are responsible for absorp tion of solar energy [26]. Anthocyanin dye is responsible for several colors in the redblue range depending on pH value. The red anthocyanin absorbs at 530 nm (b and gap 2.3 eV). Whereas, Rose Bengal has got a Xanthene class of dye with a ban d gap 1.9 eV and maximum absorption peak at 555 nm. After absorbing photon energ y from the illumi nated white light, the dye molecules in the DSSCs become excit ed and inject electrons to the ZnO APPLIED SOLAR ENERGY Vol. 49 No. 1 2013 The external photovoltaic yield or efficiency h is defined as the ratio of the max imum electric power extracted to the illumination times the surface area, i.e. P in of the cell: h = V oc I sc FF P in (2)

(it is often expressed as a percentage). Conversion yield is the key parameter a s concerns cells productiv ity [25]. THE EQUIVALENT CIRCUIT MODEL OF DSSC A sola r cell is generally characterized by the equiv alent circuit of the single diode model using as shown in Fig. 1 and the relation between the current I and the v oltage V is given by q ( V + Rs I ) V + Rs I I = I ph I s exp , 1 R sh Ak B T where Iph, Is, Rs, Rsh, q, A, kB, and T are the photocur rent, the saturation cu rrent of the diode, the series resistance, the shunt resistance, the electron ch arge, the ideality factor, the Boltzmann constant, and abso lute temperature, re spectively.

EFFECT OF ORGANIC DYES ON THE PERFORMANCE OF ZnO 43 nanowires (Fig. 3). Due to favorable energy differ ence, electron transfer occur s between the lowest unoccupied molecular orbital (LUMO) of the dye and the cond uction band of ZnO. The photogenerated electrons percolate rapidly through the Z nO nanowire and are collected by the conducting glass support. The highest occup ied molecular orbital (HOMO) of the dye is energetically lower than the redox po ten tial, Eredox of the iodine/triiodide couple. The energy difference provides the driving force for hole injection into the electrolyte. Recombination of char ge carrier is also minimized in such devices since transport of only one type of carrier (electron, in general) is ener getically possible from the dye to the s emiconductor. For Rose Bengal dye (Xanthenes), the energy differ ence is 1.9 eV [3] and for pomegranate (Anthocyanins) the energy difference is 2.3 eV [26]. Cha racterization of Dyes Three organic dyes Rose Bengal, Pomegranate and Mixed Dye were used to sensitize the semiconductor layer in DSSC. Absorption spectra of th ese dyes are showing absorption peak at 555, 525, and 545 nm, respectively Fig. 4. These dyes suffer from the fact that individually they absorb very small port ion of the visi ble spectrum of solar radiations giving rise to a low efficiency of DSSCs. The peaks of the absorption spectra of the dyes are in tune with HOMOL UMO energy difference of Xanthene and Anthocyanin class of dye. Since no sin gle dye can perform efficiently in the entire visible spectrum, the present work is focused on the perfor mance study of IV curve of DSSCs with various dyes having broadband absorption in the visible spectrum for efficient harvesting of light b y DSSCs. The Rose Bengal dye absorbs a larger fraction of the solar spec trum in the visible region 460650 nm. The Rose Ben gal dye has been used to sensitize Zn O electrode of the DSSCs. On the other hand Pomegranate is a natural dye with wi der absorption peak at lower wavelength range of solar spectrum. The spectrum of mixed dye shows even wider absorption band and is expected to harvest more sola r energy. Performance of the DSSC We have recorded IV characteristics of ZnO nan orods based DSSCs by Rose Bengal, Pomegranate (Punica Granatum) and Mixed dye us ing with equal proportion mixture of these two, with varied external resistances under illumination. Figure 4 shows the current voltage curves of various dyes b ased on DSSCs solar cells. The Rose Bengal dye shows highest short circuit curre nt whereas Pome granate has lower value of both open circuit voltage and short c ircuit current. However, the mixed dye showed the highest open circuit voltage w ith apprecia ble short circuit current. APPLIED SOLAR ENERGY Vol. 49 No. 1 2013 ntho ITO ZnO A an Dye Fig. 3. Schematic band diagram showing the working prin ciple of DSSC with vario us dyes. Absorbance, a.u. 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 450 475 500 525 550 5 75 600 Wave length, nm Rose bengal Pomegranate Mixed Dye Fig. 4. Absorption spectra of Rose Bengal, Pomegranate and Mixed dye. Current, mA 2.0 Rose bengal 1.5 Mixed Dye Pomegranate

1.0 0.5 Model equation for fitting deta I = I I e s q ( V + R s I ) /AkT 1 ( V + IR ) / R s sh ph 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Voltage, V Fig. 5. Current voltage curves of various dyes based DSSC.

44 Solar cell parameters of the three DSSCs ZnO DSSC Dyes Rose Bengal Pomegranat e Mixed Dye Voc (V) 0.6 0.56 0.65 Isc (mA) 2.1 1.18 1.57 Is (A) 13.26 9.88 7.96 KARKI et al. Rs () 0.00308 0.00489 0.00171 Rsh () 3461 5011 3934 A 4.55 4.57 4.73 FF 0.49 0.56 0.54 h (%) 1.56 1.01 1.41 Photovoltaic Properties The open circuit voltage (Voc), short circuit cur rent ( Isc), maximum voltage (Vmax), maximum current (Imax), values of DSSC cells with three different dyes were calculated from Fig. 3 and fitted with the equa tion ( 3) for one diode equivalent circuit model (Fig. 1) by Newton Raphson's method. Res ults of Curve fit ting are given in table. The parasitic resistances (series and shunt) of DSSCs are important parameters that affect on its effi ciency. The re sulting parasitic resistances, series (Rs) and shunt resistance (Rsh) were evalu ated from these results. Pomegranate has high series resistance, whereas by mixe d dye using we get subsequent reduc tion in series resistance, which results in substantial improvement in cell performance. Form the data of table, it was conf irmed that for all kinds of cells, higher the shunt resistance, the efficiency i s lowered and effi ciency raises with lower series resistances. The fill fac tor (FF) for all the cells by different dyes using are also evaluated from the IV ch aracteristics by equation (1) using and finally the energy conversion efficiency (h) is calculated by equation (2) using. All these results are presented in table . All cell parameters like Ideality Factor (A), Fill fac tor (FF) and Energy con version efficiency (h) of ZnO nanorod based DSSCs with different dyes are calcu la ted and presented in Table 1, where values are calcu lated from current voltage curves of DSSC cells fabri cated with various dyes. The Rose Bengal dye shows hi ghest efficiency and lowest fill factor, whereas Pomegranate shows lowest effici ency and highest fill factor. Though the mixed dye has an intermediate value of efficiency and fill fac tor, its values are very close to that of Rose Bengal. S o, by two kinds of dyes mixing we achieved very good effi ciency, which may be d ue to exploitation of wider band of energy in the solar spectrum as found out by the UVVis. spectroscopic results of the mixed dye. This improvement in efficienc y of the mixed dye is due to the improvement in ideality factor. Ideality factor indicates perfectness of the diode in the equivalent cir cuit, and it is 1.0 fo r a perfect diode. In our observation an improvement in ideality factor was obse rved, 4.73 for mixed dye, compared with 4.55 and 4.57 of pure dyes. CONCLUSIONS ZnO nanorod based DSSC solar cells structure was fabricated on an IT O coated glass substrate and the cell performance of ZnObased DSSCs was found out for two types of dyes and also for their mixture. The dyes differ in their abso rption spectra and absorb sun light in different frequency range. The mixed dye with larger band of frequency spectrum shows improve ment in efficiency than the average efficiency of the two dyes. This improvement in efficiency of the mixed dye is due to the improvement in ideality factor. ACKNOWLEDGMENTS We thank Cent ral Department of Physics, Tribhu van University, Kathmandu; North Bengal Univer sity, Darjeeling, Siliguri; for allowing us to carry out research work. We are g rateful to University Grants Commission (UGC), Nepal; National Academy of Scienc e and Technology (NAST), Nepal; and Indian National Science Academy (INSA), Indi a, for provid ing financial and technical support for this work. REFERENCES 1. Gratzel, M., J. Photochem. Photobiol. A, 2003, vol. C4, p. 145. 2. Tennakone, K., Kumara, G.R.R.R.A., Kumara singhe, A.R., et al., J. Photochem. Photobiol. A , 1996, vol. 94, p. 217. 3. Pradhan, B., Kumar, S., Amal, B., and Pal, A.J., Sol

ar Energy Mater. Solar Cells, 2007, vol. 91, pp. 769773. 4. Pradhan, B., Bandyopa dhyay, A., and Pal, A.J., Appl, Phys. Lett., 2004, vol. 85, p. 663. 5. Tsubomura , H., Matsumura, M., Nomura, Y., and Amamiya, T., Nature London, 1976, vol. 261, p. 402. 6. Terahara, N., Saito, N., Honda, T., et al., Phytochem., 1990, vol. 2 9, p. 949. 7. Bandyopadhyay, A. and Pal, A.J., J. Phys. Chem. B, 2003, vol. 107, pp. 25312536. 8. Chen, H., Du Pasquier, A., Saraf, G., et al., Semicon ductor Sc i. Tech., 2008, vol. 23, pp. 15. 9. Hao, S., Wu, J., Huang, Y., and Lin, J., Sola r Energy, 2006, vol. 80, p. 209. 10. Amao, Y. and Komori, T., Biosens. Bioelectr on., 2004, vol. 19, p. 843. 11. Duffy, N.W., Peter, L.M., Rajapakse, R.M.G., and Wijayantha, K.G.U.J., Phys. Chem., 2000, vol. 104, p. 8916. APPLIED SOLAR ENERGY Vol. 49 No. 1 2013

EFFECT OF ORGANIC DYES ON THE PERFORMANCE OF ZnO 12. Graetzel, M., Nature, 2001, vol. 114, pp. 338344. 13. Law, M., Greene, L.E., Johnson, J.C., et al., Nature M ater., 2005, vol. 4, pp. 455459. 14. Polo, A.S. and Iha, N.Y., Solar Energy Mater . Solar Cells, 2006, vol. 90, p. 1936. 15. Garcia, C.G., Polo, A.S., and Iha, N. Y., J. Photochem. Photobiol. A, 2003, vol. 160, p. 87. 16. Smestad, G.P., Solar Energy Mater. Solar Cells, 1998, vol. 55, p. 157. 17. Cherepy, N.J., Smestad, G. P., Gratzel, M., and Zang, J.Z.J., Phys. Chem. B, 1997, vol. 101, p. 9342. 18. N azeeruddin, M.K., Kay, A.I., Rodicio, R., et al., J. Am. Chem. Soc., 1993, vol. 115, p. 6382. 19. Gerischer, H., Michel Beyerle, M.E., Rebentrost, F., and Tribu tsch, H., Electrochim. Acta, 1968, vol. 13, p. 1509. 45 20. Watanabe, T., Fujishima, A., Tatsuoki, O., and Honda, H., Bull. Chem. Soc. J pn., 1976, vol. 49, no. 8. 21. Frank, T., J. Clin. Pharmacol., 2005, vol. 45, p. 203. 22. Nazeeruddin, M.K., Zakeeruddin, S.M., Humphry Baker, R., et al., Inorg . Chem., 1999, vol. 38, p. 6298. 23. Islam, A., Sugihara, H., Hara, K., et al., J. Photochem. Photobiol. A, 2001, vol. 145, p. 135. 24. Sicot, L., Etude et real ization de cellules photovol taiques en polymere, PhD Thesis, Orsay, 1999. 25. R ostalski, J. and Meissner, D., Solar Energy Mater. Solar Cells, 2000, vol. 61, p . 87. 26. Miguel, G., Fontes, C., Antunes, D., et al., J. Biomed. Biotechnol., 2 004, vol. 5, pp. 338342. APPLIED SOLAR ENERGY Vol. 49 No. 1 2013