Methods to Improve Light Harvesting Efficiency in Dye-SensitizedSolar Cells

(1)

Vol. 1, No. 2, 2010, 69-74 DOI:10.5229/JECST.2010.1.2.069

− 69 −

Journal of Electrochemical Science and Technology

Methods to Improve Light Harvesting Efficiency in Dye-Sensitized Solar Cells

Nam-Gyu Park^†

School of Chemical Engineering and Department of Energy Science, Sungkyunkwan University, Suwon 440-746, Korea

ABSTRACT :

Methodologies to improve photovoltaic performance of dye-sensitized solar cell (DSSC) are reviewed. DSSC is usually composed of a dye-adsorbed TiO₂ photoanode, a tri-iodide/iodide redox electrolyte and a Pt counter electrode. Among the photovoltaic parameters of short-circuit photocurrent density, open-circuit voltage and fill factor, short-circuit photocurrent density is the collective measure of light harvesting, charge separation and charge collection efficiencies. Internal quantum efficiency is known to reach almost 100%, which indicates that charge separation occurs without loss by recombination. Thus, light harvesting efficiency plays an important role in improvement of photocurrent. In this paper, technologies to improve light harvesting efficiency, including surface area improvement by nano-dispersion, size-dependent light scattering efficiency, bi-functional nano material, panchromatic absorption by selective positioning of three different dyes and transparent conductive oxide (TCO)-less DSSC, are introduced.

Keywords : Dye-sensitized solar cell, Light harvesting, Light scattering, Panchromatic, TCO-less Received November 15, 2010 : Accepted December 3, 2010

1. Introduction

Since the discovery of photovoltaic effect by Becquerel in 1839,¹⁾ tremendous efforts have been made to improve photovoltaic performance. Becquerel’s photovoltaic cell was based on electrochemical junction. However, this cell was not commercialized because of dissolution of pho- toactive material into electrolyte, which was studied by Williams in 1960.²⁾ The first long-term stable electrochemical junction solar cell was invented by O’Regan and Graetzel in 1991, where dye molecules were attached on the surface of semiconductor to be used as a light absorber instead of excitation of electron from semiconductor.³⁾ The first reported efficiency was about 7%. Recently, the conversion efficiency has been improved to 10-11%.^4-7) The certi- fied efficiency of dye-sensitized solar cell reaches 11.2%.⁸⁾ Mesoporous and nanosized oxide layer, dye and redox

electrolyte are key components comprising dye-sensitized solar cell (DSSC). Photovoltaic performance is affected by several factors in relation to light harvesting, charge separation and charge collection. In this paper, methods to control such parameters are reviewed from the opto-electronic points of view.

2. Semiconductor pn junction verses electro- chemical junction

Photovoltaic action is in general as follows. Active material absorbs light being equal to or greater than band gap energy, which leads to electron excitation from valence band (or ground state) to conduction band (or excited state).

Photo-excited electrons are separated from the absorber and finally those electrons are collected at transparent conductive or metallic charge collector. This consecutive process is the basic working principle of solar cell. The first step is related to light harvesting efficiency, the second one depends on charge separation efficiency and the last step

†Corresponding author. Tel.: +82-31-290-7241 E-mail address: [email protected]

(2)

has influence on charge collection efficiency. To separate electron and hole, pn junction is essential. There are two types of junction: solid/solid pn junction and liquid/solid pn junction. The latter is called electrochemical junction.

Dye-sensitized solar cell is a typical example of electrochemical junction. Fig. 1 compares electrochemical junction type with solid pn junction type.

3. Structure and working principle of DSSC

Fig. 2 shows the structure of DSSC. DSSC is composed of the dye-adsorbed nanocrystalline TiO₂ film formed on a transparent conductive oxide (TCO) substrate, the I₃⁻/I⁻ redox electrolyte and the Pt counter electrode. F-doped SnO₂ (FTO) is usually used as the TCO material. In order

to induce good contact between TiO₂ and electrolyte, TiO₂ film should have mesoporous structure. TiO₂ acts as electron acceptor and electron transport layer, while electrolyte acts as hole transport layer. When dye absorbs photons, electrons on the ground state of dye are excited to the excited state, which are injected into the conduction band of TiO₂ within pico- to femto-second. The photo-injected electrons are collected at FTO via the porous TiO₂ network by diffusion process with rate of about micro- to mili-second. The oxidized dyes are regenerated by oxidation of iodide with nano-second time scale. Photovoltage is generally deter- mined by the energy difference between the Fermi level of TiO₂ and redox potential of electrolyte.

Contrary to the electron transport by drift in semiconductor pn junction, photo-injected electrons are transported by diffusion process in DSSC since electric field is negligible in the entire TiO₂ film. At short-circuit condition, the Fermi energy level decreases in the direction of the back contact (charge collector) due to difference in electron concentra- tion with TiO₂ film distance. On the other hand, at open- circuit condition the Fermi energy level is constant across the entire film.⁹⁾ The photovoltaic action in DSSC is quite different from the conventional solid/solid pn junction solar cell,¹⁰⁾ where no significant space charge region is found in DSSC. Moreover, carrier of only one type is presented in the oxide semiconductor in DSSC compared with coex- istence of electron and hole in n-type and p-type semi- conductors in solid/solid pn junction solar cell.

Solar-to-electrical conversion efficiency (η) can be estimated from short-circuit photocurrent density (JSC), open-circuit voltage (VOC) and fill factor (FF). Among the three factors, JSC is resulted from the collective measure of light harvesting efficiency (ηLH), charge separation efficiency (ηCS) and charge collection efficiency (ηCC). Since internal quantum efficiency approaches 100% in DSSC, which indicates that there is little problem in charge separation, photovoltaic performance of DSSC is expected to be strongly dependent on light harvesting and charge collection efficiencies.

4. Methods to improve light harvesting effi- ciency

4.1. Nanodispersion method

In DSSC, the dye-adsorbed TiO₂ film plays an important role because it serves as a pathway for photo-injected electrons. It is no doubt that the nanocrystalline TiO₂ film with high surface area is able to utilize more photons due to a large amount of the adsorbed dye. Nanocryatlline TiO₂ Fig. 1. (Left) Solid/solid pn junction and (Right) electrochemical

junction structures. In the solid/solid pn junction electrons move to n-type material and holes are transported in p-type material.

In electrochemical junction, electrons are transported in n-type material and holes are regenerated by redox reaction of electrolyte.

Fig. 2. DSSC structure consisting of mesoporous TiO₂ film on a TCO substrate, dye, redox electrolyte and Pt-coated TCO.

Arrows indicate electron flow.

(3)

particles can be prepared by hydrothermal technique, which usually offers high photovoltaic performance. However, commercially available nanoparticles do not usually show high performance due to aggregation. Micro-bead milling of such aggregates can improve the photocurrent density because nano-dispersion leads to increase in surface area.

As a result, the conversion efficiency was significantly improved by about 44% from nano-dispersion technique (Fig. 3).¹¹⁾

4.2. Light scattering layer

Increase in film thickness will be one of possible ways to enlarge the total surface area. However, this method is restricted because the increased surface states induced by the increased surface area could act as recombination centers.^12,13) In addition, the use of only nanocrystalline TiO₂ particles with about 20 nm in diameter allows light to pass through the film, which deteriorates light harvesting efficiency. To address this problem, the bi-layer structure has been proposed, where a nanocrystalline TiO₂ layer (under-layer) is deposited on a TCO substrate and a light scattering layer (over-layer) is deposited on the top of the nanocrystalline TiO₂ layer. This bi-layer structure can

improve photocurrent density because of the improved light scattering effect.^14,15) The scattering efficiency in DSSCs^16-20) has been investigated theoretically, in which light scattering efficiency correlates with the size of the scattering particle. The size-dependent scattering efficiency in dye-sensitized solar cell was reported.²¹⁾ Two kinds of large scattering particles with rutile phase, G1 (about 0.3µm in diameter) and G2 (about 0.5 µm in diameter), were compared. The conversion efficiency of 7.55% for the 7µm-thick TiO2 film composed of only nanocrystalline TiO₂ was improved to 8.94% and 8.78% after introduction of the scattering layers comprising G1 and G2 particles, respectively. The relatively smaller sized G1 showed higher conversion efficiency than the larger G2 and the improved efficiency was mainly caused by the increase in photocurrent density. Fig. 4 shows the IPCE spectra as a function of wavelength. The bi-layer structured film having G1 over-layer shows better IPCE than G2 in the entire wavelength, which is due to higher reflectance.

Scattering efficiency (h_S) calculated from the relation ηS= (TML) × (RSL) × (IPCE at under-layer), where TML and RSL represent transmittance of the under-layer and reflectance of the scattering layer, respectively, was found to be

Fig. 4. Bi-layer structure comprising G1 (~0.3µm)- and G2 (~0.5µm) light scattering layers and IPCE spectra of G1- and G2-contained nanocrystalline TiO₂ films along with pure nanocrystalline TiO₂ film.

Fig. 3. (Left) size distribution of P-25 TiO₂ particles and (Right) photocurrent-voltage curves before and after micro bead milling at a rate of 45 Hz for 30 min.

(4)

maximized at around 600 nm. This indicates that scattering efficiency is expected to be maximized when particle size approaches l/2.

4.3. Metal oxide with dual functions

For the purpose of light scattering, large TiO₂ particles with spherical shapes or flat surfaces have been normally used as mentioned previously. However, except for the role in scattering of light, electron generation is hardly expected because of poor dye adsorption property due to the relatively low surface area. To improve electron generation in such a normal scattering particle, high surface area is required. Recently, nano-embossed hollow spherical TiO₂ (NEHS TiO₂) was developed to show bi-functional behav- ior.⁶⁾ Fig. 5 shows TEM images of the as-synthesized NEHS TiO₂ particles. The diameter of the spheres is in the range of 1-3µm and a wall thickness is about 0.25 µm.

The surface of NEHS TiO₂ is made up of TiO₂ nanoparticles with an average diameter of about 18 nm. When using NEHS TiO₂ as an over-layer (1L-NEHS), the conversion efficiency was improved by about 21% from 7.79% to 9.43%. The amount of adsorbed dye for the NEHS particles was about 0.58 × 10⁻⁷mol/cm², which was about 5 times larger than the amount of 0.12 × 10⁻⁷mol/cm² for the 400 nm-sized flat surface scattering particles, which led to higher short-circuit photocurrent. From the UV-vis reflectance spectra in Fig. 5, in the short wavelength rang- ing from 400 to 600 nm, the reflectance of the NEHS TiO₂ film is close to that of the nanocrystalline TiO₂ film, which is mainly due to light absorption by the dye molecules.

Whereas, in the long wavelength region, the dye-adsorbed NEHS film exhibits a substantially higher reflectance than the dye adsorbed nanocrystalline TiO₂ film and its

reflectivity is close to that of the CCIC film. Thus, it is confirmed that the improved JSC in NEHS-contained electrode is likely to be due to not only the improved light scattering but also charge generation.

4.4. Panchromatic Absorption

Instead of using a single dye, multi dyes may have advantage in terms of utilizing broad absorption of light.

A new method for selective positioning of dye molecules with different absorption ranges in a mesoporous TiO₂ film was recently developed.²²⁾ The key technology in this method is to control the desorption depth. The polystyrene- filled mesoporous TiO₂ film and a Bronstead base-contained polymer solution were developed to control the desorption depth. Three dyes, P5 (yellow), N719 (red), and N749 (balck), were successfully aligned by considering absorption range of each dye. Since the longer wavelength light can reach the TiO₂ layer far away from the FTO substrate, the dye alignment was designed to be the following order: FTO/P5/N719/N749. In Fig. 6, the IPCE spectral shape of the resulting three dyed cell shows superimposi- tion of each IPCE spectrum from a single dyed cell.

4.5. TCO-less technique

Regarding light harvesting issue, a certain fraction of incoming light may be reflected at TCO layer, which diminishes number of photons. In case of wafer-based sili- con solar cell, back-contact structure has been proposed in order to improve light harvesting efficiency by min- imizing the loss from light reflection at front electrode.

Similarly, back contact concept has been applied to dye- sensitized solar cell. Efficiency of 3.6% was first reported for the back contact structure in dye-sensitized solar cell.²³⁾ Fig. 5. (Left) TEM image of nano-embossed hallow spherical (NEHS) TiO₂ and (Right) reflectance of the conventional light scattering particle (CCIC) with flat surface and the NEHS particle, along with pure TiO₂ film

(5)

Improved efficiencies of about 7-8% were reported.^24-26) According to the reports, Ti metal was used as a back contact electrode. Transparent ITO was reported for use as back contact electrode in TCO-less DSSC.²⁷⁾ Fig. 7 shows the TCO-less DSSC structure and I-V property.

An ITO back-contact DSSC with a plain glass substrate shows higher photocurrent than the ITO-coated substrate.

However, open-circuit voltage is found to be deteriorated, which is due to electron recombination from ITO back contact to electrolyte. The lowered voltage is recovered by deposition of about 30 nm-thick TiO₂ film on the surface of the back contact indium tin oxide (ITO) electrode.

The transient photovoltage spectroscopy measurement confirmed the improved electron life time after deposition of thin TiO₂ layer, which is related to the voltage recovery.

5. Summary

Dye-sensitized nanostructured TiO₂ films were investigated in terms of methodologies for enhancing light harvesting efficiency. Light scattering method was found to a useful way to utilize long wavelength light, which was dependent on the size of scattering particle, associated with reflectance of scattering particle. Nano-embossed hallow sphere was one of good candidates for dual functions of both efficient light scattering and charge generation. To cover wide range of incoming light, method of selective positioning of dyes with different absorption character- istics was a useful way to realize panchromatic absorption.

Selective desorption was realized by controlling retention time of desorption solutions by means of downsizing pore of TiO₂ film and exploring viscous NaOH solution. TCO- less technology seems to be beneficial for optical gain. The combined methods are expected to lead to high photo- conversion efficiency due to an improved light harvesting efficiency, associated with improved photocurrent.

Acknowledgements

This work was supported by the National Research Fig. 6. From top to bottom: a comparison of single dye for

narrow absorption and multi dyes for panchromatic absorption, a conventional method being unable to position two different dyes, a new method being able to position three different dyes and IPCE spectra of three dyed TiO₂ film along with each dyed TiO₂ film

Fig. 7. TCO-less DSSC structure and I-V curves of TCO-less substrates along with ITO-coated glass substrate. TF-TiO₂ represents thin film TiO₂ layer on the top of back contact ITO in TCO-less DSSC.

(6)

Foundation of Korea (NRF) grant funded by the Ministry of Education, Science and Technology (MEST) of Korea under contracts 2010-0028821, 2010-0014992, 2010- 0001842 (ERC) and R31-2008-000-10029-0 (WCU). This work was also supported by the Korea Institute of Energy Technology Evaluation and planning (KETEP) grant funded by the Ministry of Knowledge Economy under contract 20103020010010.

References

1. E. Becquerel, Compt. Rend. 9, 561 (1839).

2. R. Williams, J. Chem. Phys. 32, 1505 (1960).

3. B. O’Regan and M. Gratzel, Nature 353, 737 (1991).

4. M. Gratzel, Inorg. Chem. 44, 6841 (2005).

5. Y. Chiba, A. Islam, Y. Watanabe, R. Komiya, N. Koide and L. Han, Jpn. J. Appl. Phys. Part 2 45, L638 (2006).

6. H.-J. Koo, Y.J. Kim, Y.H. Lee, W.I. Lee, K. Kim and N.-G.

Park, Adv. Mater. 20, 195, (2008).

7. N.-G. Park and K. Kim, Phys. Stat. Sol., 205, 1895 (2008).

8. M.A. Green, K. Emery, Y. Hishikawa and W. Warta, Prog. Photovolt: Res. Appl. 18, 346 (2010).

9. J. van de Lagemaat, N.-G. Park and A.J. Frank, J. Phys.

Chem. B 104, 2044 (2000).

10. D. Cahen, G. Hodes, M. Gratzel, J.F. Guillemoles and I. Riess, J. Phys. Chem. B 104, 2053 (2000).

11. G.-W. Lee, S.-Y. Bang, C. Lee, W.-M. Kim, D. Kim, K. Kim and N.-G. Park, Curr. Appl. Phys. 9, 900 (2009).

12. S. Ito, T. Kitamura, Y. Wada and S. Yanagida, Sol. Energy Mater. Sol. Cells 76, 3 (2003).

13. J. Nissfolk, K. Fredin, A. Hagfeldt and G. Boschloo, J.

Phys. Chem. B 110, 17715 (2006).

14. S. Ito, S.M. Zakeerudiin, R. Humphry-Baker, P. Liska, P.

Charvet, P. Comte, M.K. Nazeeruddin, P. Pechy, M. Takata, H. Miura, S. Uchida and M. Gratzel, Adv. Mater. 18, 1202 (2006).

15. S. Hore, C. Vetter, R. Kern, H. Smit and A. Hinsch, Sol.

Energy Mater. Sol. Cells 90, 1176 (2006).

16. A. Usami, Chem. Phys. Lett. 277, 105 (1997).

17. A. Usami, Sol. Energy Mater. Sol. Cells 64, 73 (2000).

18. J. Ferber and J. Luther, Sol. Energy Mater. Sol. Cells 54, 265 (1998).

19. A. Usami, Sol. Energy Mater. Sol. Cells 59, 163 (1999).

20. W.E. Vargas and G.A. Niklasson, Sol. Energy Mater. Sol.

Cells 69, 147 (2001).

21. H.-J. Koo, J. Park, B. Yoo, K. Yoo, K. Kim and N.-G.

Park, Inorg. Chim. Acta 361, 677 (2008).

22. K. Lee, S.W. Park, M.J. Ko, K. Kim and N.-G. Park, Nature Mater. 8, 665 (2009).

23. J.M. Kroon, N.J. Bakker, H.J.P. Smit, P. Liska, K.R. Thampi, P. Wang, S.M. Zakeeruddin, M. Grätzel, A. Hinsch, S. Hore, U. Würfel, R. Sastrawan, J.R. Durrant, E. Palomares, H.

Pettersson, T. Gruszecki, J. Walter, K. Skupien and G.E.

Tulloch, Prog. Photovolt. Res. Appl. 15, 1 (2007).

24. N. Fuke, A. Fukui, Y. Chiba, R. Komiya, R. Yamanaka and L. Han, Jpn. J. Appl. Phys. 46, L420 (2007).

25. Y. Kashiwa, Y. Yoshida and S. Hayase, Appl. Phys. Lett.

92, 033308 (2008).

26. N. Fuke, A. Fukui, R. Komiya, A. Islam, Y. Chiba, M.

Yanagida, R. Yamanaka and L. Han, Chem. Mater. 20, 4974 (2008).

27. B. Yoo, K.-J. Kim, D.-K. Lee, K. Kim, M.J. Ko, Y.H. Kim, W.M. Kim and N.-G. Park, Optics Express, 18, A395 (2010).