Open Access
Add to BibSonomyAbstract
In this paper we analyze the interplay between transparency and efficiency in dye sensitized solar cells by varying fabrication parameters such as the thickness of the nano-crystalline TiO2 layer, the dye loading and the dye type. Both transparency and efficiency show a saturation trend when plotted versus dye loading. By introducing the transparency-efficiency plot, we show that the relation between transparency and efficiency is linear and is almost independent on the TiO2 thickness for a certain thickness range. On the contrary, the relation between transparency and efficiency depends strongly on the type of the dye. Moreover, we show that co-sensitization techniques can be effectively used to access regions of the transparency-efficiency space that are forbidden for single dye sensitization. The relation found between transparency and efficiency (T&E) can be the general guide for optimization of Dye Solar Cells in building integration applications.
©2013 Optical Society of America
1. Introduction
Despite moderate power generation with respect to Silicon solar cell technology, Dye Solar Cells (DSCs) [1] offer additional features such as transparency and color control over the entire device area. These features are undoubtedly very engaging for Building Integrated Photo-Voltaic (BIPV) applications. The BIPV concept was born from the need to combine the production of renewable energy and architectural features of elements that are part of the human being environment. Glass becomes an active element in the production of energy while preserving the feature of a construction element. The requirements for a photovoltaic module architecturally integrated in a glass façade are: i) adequate power generation and ii) good light transmission [2–4]. Such requirements imply different strategies in module optimization. In fact, efficiency of the module should be optimized whilst maintaining transparency at acceptable levels.
The light transmission with opaque crystalline silicon PV is achieved by spacing the cells in the module. Amorphous silicon (a-Si) technology can provide semi-transparency by reducing the thickness of the deposited a-Si film (0.3μm) or through micro perforation obtained by laser post-processing. Typical transparency for commercial a-Si panel is of the order of 20% [5, 6].
DSCs, on the other hand, are based on semi-transparent elements. In fact, the basic structure of a DSC is composed of two conductive transparent glasses coated with a Transparent Conductive Oxide (TCO), few micrometers of transparent mesoporous TiO2, a monolayer of dye molecule that covers the titania surface, an electrolyte and a transparent film of platinum catalyst. The actual color of the cell is mainly due to the superposition of the optical spectra of the electrolyte and the dye.
All the elements that constitute a DSC can, however, influence the transparency (or light transmittance, LT). Precise formulation of the titania paste and the thickness of the titania layer is crucial if transparency is required: TiO2 paste containing particles of the order of 15-20 nanometers and a total thickness of a few microns typically provides a very good transmittance in the visible range (around 75-80%); on the contrary a paste containing larger particles (hundreds of nm), normally used for diffusive scattering layers, enhances photocurrent but greatly reduces transparency. Moreover, the type and amount of electrolyte solution, platinum and dye in the cell are other important aspects that affect the calculation of the LT.
Concerning DSC optimization, co-sensitization has been reported to be effective for panchromatic response, hence to improve the efficiency of the cells [7–11]. However very little attention has be given to the influence of co-sensitization on the transparency of the cell [12].
In this work we show that carefully tailoring the immersion time in both dyes the optical response can satisfy the compromise between efficiency and transparency.
The aim of this study is to perform a thorough investigation of cell fabrication parameters influencing the relation between efficiency and transparency. In order to limit the parameter space, we will only consider the TiO2 layer thickness, the dye loading, and the dye type, including the co-sensitization, as variables in the analysis we will perform, fixing all other parameters. Since transparency is an important target to be considered in DSC technology [13], in this work we emphasized this fundamental property for BIPV.
2. Material and methods
Nano-crystalline titania layers (nc-TiO2) were prepared by screen printing the Dyesol 18NR-T paste onto transparent conductive glass substrates (Pilkington, TEC 8, 8 Ω/□) followed by sintering at 480°C for 30min. The film final thicknesses (i.e. 6, 9, 12 µm) were measured with “Veeco Dektak 150” profilometer. The size of the photoelectrodes was 5mm x 5mm (0.25cm2). After sintering, the photoelectrodes were immersed in the dye solution for a given immersion time at RT which will be used to control the dye loading into the TiO2 layer. The dye used was, Cis-bis(isothiocyanato)bis(2,2’-bipyridyl-4,4’-dicarboxylato)-ruthenium(II)bis-tetrabutylammonium (N719, Red Dye in the following), 0.3mM in ethanol and cis-dicyano-bis(2,2’-bipyridyl-4,4’-dicarboxylic acid) ruthenium(II) (Orange Dye in the following) at 0.3 mM in ethanol. Transparent counterelectrodes were prepared by firing, at 420°C for 15min, a solution of Pt-paste (CHIMET) deposited by screen printing on the conductive glass. These two electrodes were assembled by using 60µm-thick surlyn gaskets (Dupont 1702). The electrolyte solution (HSE Dyesol) was introduced by vacuum backfilling through a hole previously realized on the counter-electrode which was then closed by a glass patch.
The transmittance measurements were performed by spectrophotometer SHIMADZU UV-Vis 2550 in the range 380-780 nm. IV characterization was carried out with a sun simulator KHS Solar Constant 1200 at AM1.5G and a source meter Keitley 2400.
2.1 Sensitization of photo-electrodes
The actual amount of the adsorbed dye on titania depends on the thickness of TiO2 layer [14], its immersion time, the temperature and the dye concentration. In our work, control on the dye loading is obtained by varying the immersion time up to the adsorption saturation point. After saturation, no more dye can be adsorbed onto the TiO2 layer due to the lack of free anchoring sites on the titania. Accordingly to the Lambert-Beer law the larger is the quantity of dye attached, larger will be the absorption of the light and smaller will be the LT and vice versa.
In our experiment, we considered cells with different titania thicknesses, namely 6, 9 and 12μm and, for each thickness, several cells were realized (see Fig. 1 ) with different immersion times (see Table 1 and Table 2 ). Some devices were sensitized with the Red or Orange Dye while other were co-sensitized with both dyes. Following the procedure of Inazaku and associates [15], co-sensitization were performed by immersion first the photoelectrode in the Red Dye for a given (short) time and then in the Orange Dye up to saturation.
Fig. 1 Sub-set of fabricated photoanodes with Red-Dye (N719). Each glass substrate contains three cells sensitized with the same immersion time.
Table 1. TiO2 thickness and immersion times in the Red or in the Orange Dye
Table 2. TiO2 thickness and immersion time firstly in the Red (RD) and then in the Orange Dye (OD)
2.2 Measurements
In order to determine luminous and solar characteristics, required for the study of transparency, we refer to ISO 9050 normative [16]. The light transmittance τv, also known as transparency, is calculated using the following formula:
where Dλ is the spectral distribution of light source D65 (in our case AM1.5); τ(λ) is the spectral transmittance of the illuminated object (transmittance); V(λ) is the eye sensitivity factor of the photopic view’s observer; Δλ is the wavelength step. The eye sensitivity factor together with the absorbance of photoanodes sensitized with Red Dye, Orange Dye and co-sensitized with both dyes is shown in Fig. 2 . Transparency of the cell is mainly influenced by the overlap between absorbance and V(λ). Larger is the overlap, smaller will be the transparency.Fig. 2 Absorbance spectra of photoanodes sensitized with Red Dye, Orange Dye and co-sensitized with both Red and Orange Dye. Photoanodes were fabricated with a TiO2 thickness of 9μm and dipped up to saturation. The co-sensitized cell were dyed into Red Dye for 25min and in Orange Dye up to saturation. Figure also show the eye sensitivity factor V(λ).
3. Results and discussion
3.1 Photovoltaic characterization
We performed a photovoltaic characterization of fabricated DSCs to correlate the cell efficiency with the TiO2 thickness and the immersion time. Figure 3 shows the results for the Red Dye. Here, each efficiency value is the average between a set of six samples and the standard deviation bar is also plotted. As clearly depicted in Fig. 3, for large immersion times we observe a saturation of the dye loading.
Fig. 3 Efficiency as a function of the immersion time in the Red Dye solution for cells with 6, 9, 12μm thick TiO2 layers.
For a TiO2 thickness of 9 micrometers we observe that the efficiency of the cell dipped for 320min is greater than the efficiency obtained for longer immersion times. When the adsorption saturation point of Dye on to TiO2 is reached, other dye molecules can attach on the dye layer by Van der Waals forces. This phenomenon, known as molecular stacking, quenches electron transfer to the TiO2 hence reducing the efficiency of the cell [17]. Similar curves were obtained for the Orange Dye (Fig. 4 ).
Fig. 4 Efficiency as a function of the immersion time in the Orange Dye solution for cells with 6, 9, 12μm thick TiO2 layers.
3.2 Transparency
Figure 5 shows the trend of cell transparency versus the photo-electrodes immersion time for each set of thicknesses for the Red Dye. For a fixed titania thickness, the transparency decreases as a function of the immersion time until the saturation level. Thicker titania layers permit to increase the dye load, hence a more opaque cell is fabricated. Obviously, the efficiency of DSCs will depend on the TiO2 thickness, as already seen in Figs. 3 and 4.
Fig. 5 Transparency of Red Dye solar cells as a function of the immersion time and for several TiO2 thicknesses.
As discussed above, the stacking of molecules for long immersion times influences also the relation between transparency and efficiency. To investigate this phenomena further, we consider the 9μm thick devices in Figs. 5 and 6 . While the efficiency reaches the maximum after 160min of dyeing and undergoes a slight reduction for longer immersion times, the transparency of the cells has a continuous monotonic decreases also for immersion times larger than 160min. This phenomena is clearly related to the presence of molecular stacking which induces excitonic exchange between omologue molecules on the titania surface [12]. This phenomena is not desirable for the applications we are discussing in this work and for this reason, in the following analysis, we will only consider immersion times which prevents stacking phenomena.
Fig. 6 Transparency of Orange Dye solar cells as a function of the immersion time and for several TiO2 thicknesses.
Similar transparency plot are shown in Fig. 6 for the Orange Dye.
3.3 Transparency vs. efficiency
It is possible to correlate efficiency and transparency of Figs. 3-4 and 5-6 by plotting, for all the cells, the cell transparency as a function of the cell efficiency as shown in Fig. 7 . The relation between efficiency and transparency is quite linear and is almost independent on TiO2 thickness. This relation, on the other hand, is strongly related to the dye type. With the Red Dye it is possible to realize cells with efficiency higher than those with Orange Dye. However the transparency of the cells is always lower than the cells with Orange Dye. As an example, Red Dye cells have an efficiency of 4% with a transparency of around 30%, while Orange Dye cells with the same efficiency have a transparency of 40%. If, for a given application, the required transparency is set to 40% then a Red Dye cell has an efficiency of 3.15% while the one with the Orange Dye has an efficiency of 4.3% (around 35% higher than the Red Dye). The better transparency properties of the Orange Dye based cells with respect to the Red Dye ones is easily explained considering that the Red Dye has a strong absorption peak at 536 nm very close to the maximum of the V(λ) function in Eq. (1) (eye responsivity) which occurs at 555 nm. The main absorption peak of the Orange Dye is blu-shifted by 50nm. For this reason the numerator in Eq. (1) is larger for cells with Orange Dye with respect to the one with Red Dye.
Fig. 7 Transparency-Efficiency plot. Transparency versus Efficiency of all dye sensitized solar cells for several TiO2 thicknesses (□, 6μm, ○, 9μm, Δ, 12μm). Results for Red Dye (solid line) and Orange Dye (dashed line) cells are reported.
The curves presented in Fig. 7 can represent a calibration diagram for transparent DSC technology. If the target application requires efficiency irrespective of transparency, then the better DSC structure will be the one with the Red Dye. On the other hand, if a target transparency larger than 35% is required, the better structure will be the one with the Orange Dye.
As shown in Fig. 7, cells sensitized with Orange Dye cannot access the efficiency space above 4.5%. To push further the efficiency of the cell without losing the benefit of the Orange Dye concerning transparency, we have considered co-sensitized cell with both Red and Orange dyes as described in Experimental Section. We have found that, for a TiO2 thickness of 9 μm it is possible to increase efficiency even above 5% with an overall transparency well above the one of the Red Dye cells. This clearly shows that the co-sensitization strategy can be used to improve the relation between efficiency and transparency and is very indicated for BIPV applications. All the co-sensitization experiments performed with different TiO2 thicknesses are shown in Fig. 8 together with the trends of efficiency/transparency for the Red and Orange dyes already shown in Fig. 7.
Fig. 8 Transparency-Efficiency plot for several TiO2 thicknesses (□, 6μm, ○, 9μm, Δ, 12μm; see Table 2 for details) of the co-sensitized cells. The Red (solid line) and Orange Dye (dashed line) transparency-efficiency trends, already depicted in Fig. 7, are also reported.
For 6μm the results of the co-sensitization structures, independently of the immersion time, are all between the Orange Dye line and Red Dye line. This shows that there is no particular benefit in the co-sensitization. The same is true for TiO2 thicknesses of 12µm, where the data overlap those of the Red Dye cells.
4. Conclusions
The relation between light transmission and efficiency of DSCs was investigated studying the effect of TiO2 thickness variation and the dye loading of the photoelectrode for two dyes (Red and Orange) and for devices co-sensitized with these two dyes. As expected, we observed that, when the TiO2 thickness increases, light transmission decreases, according to Lambert-Beer law, whilst the efficiency of the corresponding cells rises. To point-out the correlation between these two quantities, a transparency-efficiency plot was introduced. We show that the relation between transparency and efficiency is linear and almost independent on the TiO2 thicknesses for the range of values chosen in this work. On the other hand, a strong dependence on the dye type was observed. Dyes with absorption peaks shifted with respect to the eye sensitivity factor of the photopic view’s observer (eye responsivity) can perform better in term of transparency vs. efficiency. We have also shown that co-sensitization can be effectively used to improve the relation between efficiency and transparency. We found that for cells with 5% efficiency, the transparency can be increased by 30% with respect to the that of Red Dye cells, if a Red-Orange co-sensitization is used. The transparency-efficiency plot we have introduced in this work is an important instrument to consider for a proper design of BIPV applications where transparency and efficiency both represent design parameters.
Acknowledgments
We gratefully acknowledge the financial support of Polo Solare Organico – Regione Lazio, GREAT project of Lazio Region and PRIN 2008 project of the Italian Ministry of Education, University and Research (MIUR).
References and links
1. B. O'Regan and M. Grätzel, “A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films,” Nature 353(6346), 737–740 (1991). [CrossRef]
2. A. V. Shah, H. Schade, M. Vanecek, J. Meier, E. Vallat-Sauvain, N. Wyrsch, U. Kroll, C. Droz, and J. Bailat, “Thin-film silicon solar cell technology,” Prog. Photovolt. Res. Appl. 12(23), 113–142 (2004). [CrossRef]
3. K. Yamamoto, A. Nakajima, M. Yoshimi, T. Sawada, S. Fukuda, T. Suezaki, M. Ichikawa, Y. Koi, M. Goto, T. Meguro, T. Matsuda, M. Kondo, T. Sasaki, and Y. Tawada, “A high efficiency thin film silicon solar cell and module,” Sol. Energy 77(6), 939–949 (2004). [CrossRef]
4. http://www.schottsolar.com/global/products/building-integrated/
5. http://www.taiyokogyo.com/tss/
6. http://www.egl-energy.biz/downloads/EGLSoluxBIPVbrochure.pdf
7. J.-H. Yum, S.-R. Jang, P. Walter, T. Geiger, F. Nüesch, S. Kim, J. Ko, M. Grätzel, and M. K. Nazeeruddin, “Efficient co-sensitization of nanocrystalline TiO2 films by organic sensitizers,” Chem. Commun. (Camb.) (44): 4680–4682 (2007). [CrossRef] [PubMed]
8. R.Y. Ogura, S. Nakane, M. Morooka, M. Orihashi, Y. Suzuki and K. Noda, “High-performance dye-sensitized solar cell with a multiple dye system,” Appl. Phys. Lett. 94, 073308–073308–073303 (2009).
9. J.-J. Cid, J.-H. Yum, S.-R. Jang, M. K. Nazeeruddin, E. Martínez-Ferrero, E. Palomares, J. Ko, M. Grätzel, and T. Torres, “Molecular Cosensitization for Efficient Panchromatic Dye-Sensitized Solar Cells,” Angew. Chem. 119(44), 8510–8514 (2007). [CrossRef]
10. K. Hara, T. Sato, R. Katoh, A. Furube, Y. Ohga, A. Shinpo, S. Suga, K. Sayama, H. Sugihara, and H. Arakawa, “Molecular Design of Coumarin Dyes for Efficient Dye-Sensitized Solar Cells,” J. Phys. Chem. B 107(2), 597–606 (2003). [CrossRef]
11. Y. Chen, Z. Zeng, C. Li, W. Wang, X. Wang, and B. Zhang, “Highly efficient co-sensitization of nanocrystalline TiO2 electrodes with plural organic dyes,” New J. Chem. 29(6), 773–776 (2005). [CrossRef]
12. D. Colonna, V. Capogna, A. Lembo, T. M. Brown, A. Reale, and A. Di Carlo, “Efficient Cosensitization Strategy for Dye-Sensitized Solar Cells,” Apex 5(2), 022303 (2012). [CrossRef]
13. S. Yoon, S. Tak, J. Kim, Y. Jun, K. Kang, and J. Park, “Application of transparent dye-sensitized solar cells to building integrated photovoltaic systems,” Build. Environ. 46(10), 1899–1904 (2011). [CrossRef]
14. M. Dürr, A. Schmid, M. Obermaier, A. Yasuda, and G. Nelles, “Diffusion Properties of Dye Molecules in Nanoporous TiO2 Networks,” J. Phys. Chem. A 109(17), 3967–3970 (2005). [CrossRef] [PubMed]
15. F. Inakazu, Y. Noma, Y. Ogomi, and S. Hayase, “Dye-sensitized solar cells consisting of dye-bilayer structure stained with two dyes for harvesting light of wide range of wavelength,” Appl. Phys. Lett. 93(9), 093304 (2008). [CrossRef]
16. http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=35062
17. J. A. Pollard, D. Zhang, J. A. Downing, F. J. Knorr, and J. L. McHale, “Solvent Effects on Interfacial Electron Transfer from Ru(4,4′-dicarboxylic acid-2,2′-bipyridine)2(NCS)2 to Nanoparticulate TiO2: Spectroscopy and Solar Photoconversion,” J. Phys. Chem. A 109(50), 11443–11452 (2005). [CrossRef] [PubMed]
