pubs.acs.org/Langmuir © 2009 American Chemical Society
Regenerative PbS and CdS Quantum Dot Sensitized Solar Cells with a Cobalt Complex as Hole Mediator Hyo Joong Lee, Peter Chen, Soo-Jin Moon, Frederic Sauvage, Kevin Sivula, Takeru Bessho, Daniel R. Gamelin,† Pascal Comte, Shaik M. Zakeeruddin, Sang Il Seok,‡ Michael Gr€atzel,* and Md. K. Nazeeruddin* Laboratory for Photonics and Interfaces, Institute of Chemical Sciences and Engineering, School of Basic Sciences, Swiss Federal Institute of Technology, CH-1015 Lausanne, Switzerland. † On sabbatical leave from the Department of Chemistry, University of Washington, Seattle, Washington. ‡ KRICT-EPFL Global Research Laboratory, Advanced Materials Division, Korea Research Institute of Chemical Technology, 19 Sinseongno, Yuseong, Daejeon, 305-600, Korea Received January 20, 2009. Revised Manuscript Received May 13, 2009 Metal sulfide (PbS and CdS) quantum dots (QDs) were prepared over mesoporous TiO2 films by improved successive ionic layer adsorption and reaction (SILAR) processes. The as-prepared QD-sensitized electrodes were combined with a cobalt complex redox couple [Co(o-phen)3]2+/3+ to make a regenerative liquid-type photovoltaic cell. The optimized PbS QD-sensitized solar cells exhibited promising incident photon-to-current conversion efficiency (IPCE) of over 50% and an overall conversion efficiency of 2% at 0.1 sun in a regenerative mode. The overall photovoltaic performance of the PbS QD-sensitized cells was observed to be dependent on the final turn of the SILAR process, giving a better result when the final deposition was Pb2+, not S2-. However, in the case of CdS QD-sensitized cells, S2- termination was better than that of Cd2+. The cobalt complex herein used as a regenerative redox couple was found to be more efficient in generating photocurrents from PbS QD cells than the typical hole scavenger Na2S in a three-electrode configuration. The CdSsensitized cell with this redox mediator also showed better defined current-voltage curves and an IPCE reaching 40%.
Introduction Dye-sensitized solar cells (DSSCs) have attracted much attention throughout the world from both academic and industrial fields as a promising alternative to conventional solid-state photovoltaic devices since a report by O’Regan and Gr€atzel in 1991.1 For further improvement of the overall efficiency and long-term stability in DSSCs, various dye molecules have been designed and tested as sensitizers for absorbing incident solar radiation and injecting photogenerated electrons into the conduction band of metal oxides.2 The oxidized dye cations are then regenerated by electron donation from the electrolyte or, alternatively, by hole injection into an organic hole transporting material for the solid-state counterpart.3 Although the results obtained so far are very impressive, further improvements in both efficiency and stability by introducing new materials and engineering their interfaces can be anticipated. Inorganic semiconducting materials are attractive candidates for solar cell sensitization because of the facile tunability of their effective band gaps by using proper precursor combinations, as demonstrated in thin-film solar cells.4 Recently, analogous semiconductor quantum *To whom correspondence should be addressed. E-mail: michael.graetzel@ epfl.ch (M.G.);
[email protected] (M.K.N.). Telephone: +41-21693-6124. Fax: +41-21-693-4311. (1) O’Regan, B.; Gr€atzel, M. Nature 1991, 353, 737. (2) (a) Nazeeruddin, M. K.; Gr€atzel, M. In Encyclopedia of Electrochemistry; Licht, S., Ed.; Wiley-VCH: 2002; Vol. 6, p 407. (b) Robertson, A. Angew. Chem., Int. Ed. 2006, 45, 2338. (c) Wang, P.; Zakeeruddin, S. M.; Moser, J.-E.; Nazeeruddin, M. K.; Sekiguchi, T.; Gr€atzel, M. Nat. Mater. 2003, 2, 402. (d) Kuang, D.; Klein, C.; Ito, S.; Moser, J.-E.; Humphry-Baker, R.; Evans, N.; Duriaux, F.; Gr€atzel, C.; Zakeeruddin, S. M.; Gr€atzel, M. Adv. Mater. 2007, 19, 1133. (e) Sapp, S. A.; Elliott, C. M.; Contado, C.; Caramori, S.; Bignozzi, C. A. J. Am. Chem. Soc. 2002, 124, 11215. (3) (a) Hagfelt, A.; Gr€atzel, M. Acc. Chem. Res. 2000, 33, 269. (b) Gr€atzel, M. Nature 2001, 414, 338. (c) Snaith, H. J.; Schmidt-Mende, L. Adv. Mater. 2007, 19, 3187. (4) Chopra, K. L.; Paulson, P. D.; Dutta, V. Prog. Photovoltaics 2004, 12, 69.
7602 DOI: 10.1021/la900247r
dots (CdS,5 CdSe,6 InP,7 InAs,8 etc.9) and thin absorbers (CdS,10 CdSe,11 CdTe,12 CuInS2,13 etc.14) have been explored as new sensitizers over mesoporous metal oxide layers in efforts to merge the advantages of both DSSCs and thin-film solar cells. Unfortunately, the well-known and most efficient I -/I3- redox couple is not compatible with low band gap semiconducting materials, leading to a rapid corrosion process of the semiconductor, and both welldefined and efficient regenerative redox couples appropriate for either quantum dot or thin absorber-sensitized solar cells have not yet been identified. Consequently, in most cases, such cells have been tested with hole scavengers using a three-electrode configuration. With such three-electrode photoelectrochemical cells, the evaluation of the real photovoltaic parameters appears to be not straightforward because of the irreversible character of the redox (5) Peter, L. M.; Riley, D. J.; Tull, E. Z.; Wijayantha, K. G. U. Chem. Commun. 2002, 1030. (6) (a) Lee, H. J.; Yum, J.-H.; Leventis, H. C.; Zakeeruddin, S. M.; Haque, S. A.; Chen, P.; Seok, S. I.; Gr€atzel, M.; Nazeeruddin, Md. K. J. Phys. Chem. C 2008, 112, 11600. (b) Robel, I.; Subramanian, V.; Kuno, M.; Kamat, P. V. J. Am. Chem. Soc. 2006, 128, 2385. (c) Leschkies, K. S.; Divakar, R.; Basu, J.; Enache-Pommer, E.; Boercker, J. E.; Carter, C. B.; Kortshagen, U. R.; Norris, D. J.; Aydil, E. S. Nano Lett. 2007, 7, 1793. (d) Okazaki, K.; Kojima, N.; Tachibana, Y.; Kuwabata, S.; Torimoto, T. Chem. Lett. 2007, 36, 712. (7) Zaban, A.; Micic, O. I.; Gregg, B. A.; Nozik, A. J. Langmuir 1998, 14, 3153. (8) Yu, P.; Zhu, K.; Norman, A. G.; Ferrere, S.; Frank, A. J.; Nozik, A. J. J. Phys. Chem. B 2006, 110, 25451. (9) (a) Plass, R.; Pelet, S.; Krueger, J.; Gr€atzel, M.; Bach, U. J. Phys. Chem. B 2002, 106, 7578. (b) Vogel, R.; Hoyer, P.; Weller, H. J. Phys. Chem. 1994, 98, 3183. (10) Larramona, G.; Chone, C.; Jacob, A.; Sakakura, D.; Delatouche, B.; Pere, D.; Cieren, X.; Nagino, M.; Bayon, R. Chem. Mater. 2006, 18, 1688. (11) Levy-Clement, C.; Tena-Zaera, R.; Ryan, M. A.; Katty, A.; Hodes, G. Adv. Mater. 2005, 17, 1512. (12) Ernst, K.; Engelhardt, R.; Ellmer, K.; Kelch, C.; Muffler, H.-J.; Lux-Steiner, M.-Ch.; Konenkamp, R. Thin Solid Films 2001, 387, 26. (13) Kaiser, I.; Ernst, K.; Fischer, Ch.-H.; Konenkamp, R.; Rost, C.; Sieber, I.; Lux-Steiner, M.-Ch. Sol. Energy Mater. Sol. Cells 2001, 67, 89. (14) Belaidi, A.; Bayon, R.; Dloczik, L.; Ernst, K.; Lux-Steiner, M.-Ch.; Konenkamp, R. Thin Solid Films 2003, 431-432, 488.
Published on Web 06/05/2009
Langmuir 2009, 25(13), 7602–7608
Lee et al.
Article
reaction (sacrificial) used to regenerate the sensitizer. Furthermore, using reversible redox mediators, the utilization of three-electrode photoelectrochemical cell overestimates the conversion efficiency,15 since diffusion and electron transfer overvoltage losses are neglected. After an extensive search for efficient, stable, and compatible redox mediators for use in quantum dot (QD)-sensitized solar cells, a few encouraging candidates have recently been identified. For the liquid-type cell, alternate redox couples based on polypyridyl cobalt complexes2e,6a,8 or polysulfide (S2-/Sx)16 appear to work quiet well, although the latter remains poorly defined because of the complex chemistry of sulfur in the aqueous solvent used. Recently, the well-known I -/I3- redox couple also was tested with CdS- or CdSe-sensitized cells while the corrosive reaction was observed to progress gradually.6c,17 On the other hand, CuSCN and spiroOMeTAD [2,20 ,7,70 ,- tetrakis(N,N-di-p-methoxyphenylamine)-9, 90 -spirobifluorene], basically used in solid state DSSCs, were moderately successful when combined with QD sensitizers.9a,10,11 It appears then urgent to develop better redox mediators to keep maintaining progress in QD-sensitized solar cells that will allow accurate evaluation of functional parameters of such cells operating in regenerative mode. Our recent results6a have identified a new redox system based on cobalt complexes as a promising hole carrier for testing liquid type QD-sensitized cells. The [cobalt(o-phen)3]2+/3+ has shown efficient hole carrier properties, giving overall efficiencies of over 1% at full-sun intensity and a high incident photonto-current conversion efficiency (IPCE) value (over 35% at broad ranges) as well as reasonable stability for QD-sensitized solar cells prepared from CdSe colloids.6a Using this redox relay, it was possible to characterize interfacial charge transfer kinetics in a regenerative QD-sensitized solar cell successfully. In this paper, we now apply the [Co(o-phen)3]2+/3+ redox couple to develop regenerative PbS and CdS QD-sensitized solar cells. Although PbSand CdS-sensitized electrodes have been extensively investigated so far using sacrificial hole scavengers9b and ill-defined redox couples,16,17 those QD sensitizers were combined for the first time with a well-defined redox couple based on a cobalt complex to make regenerative solar cells and then investigate their photovoltaic performance.
Experimental Method Chemicals. Cadmium nitrate tetrahydrate (Fluka, g99.0%), lead nitrate (Aldrich, 99.99%), sodium sulfide (Aldrich), titanium diisopropoxide bis(acetylacetonate) (Aldrich), and NH4F (Fluka, 98%) were used as received. Ethanol and methanol were of HPLC grade. Liquid Electrolyte Preparation. The cobalt(II) complex, [Co (o-phen)3](TFSI)2 (o-phen=1,10-phenanthroline and TFSI=bis (trifluoromethanesulfonyl)imide), was synthesized according to literature procedures.18,19 The cobalt electrolytes were prepared at concentrations of 0.5 M Co2+ complex, 0.05 M Co3+ complex, (15) (a) Sun, W.-T.; Yu, Y.; Pan, H.-Y.; Gao, X.-F.; Chen, Q.; Peng, L.-M. J. Am. Chem. Soc. 2008, 130, 1124. (b) Hodes, G. J. Phys. Chem. C 2008, 112, 17778. (16) (a) Diguna, L. J.; Shen, Q.; Kobayashi, J.; Toyoda, T. Appl. Phys. Lett. 2007, 91, 023116. (b) Mora-Sero, I.; Gimenez, S.; Moehl, T.; Fabregat-Santiago, F.; Lana-Villareal, T.; Gomez, R.; Bisquert, J. Nanotechnology 2008, 19, 424007. (c) Tachibana, Y.; Akiyama, H. Y.; Ohtsuka, Y.; Torimoto, T.; Kuwabata, S. Chem. Lett. 2007, 36, 88. (17) (a) Shen, Y.-J.; Lee, Y.-L. Nanotechnology 2008, 19, 045602. (b) Lee, W.; Lee, J.; Lee, S.; Yi, W.; Han, S.-H.; Cho, B. W. Appl. Phys. Lett. 2008, 92, 153510. (c) Lee, Y. L.; Chang, C. H. J. Power Sources 2008, 185, 584. (18) (a) Shklover, V.; Eremenko, I. L.; Berke, H.; Nesper, R.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Gr€atzel, M. Inorg. Chim. Acta 1994, 219, 11. (b) Dekorte, J. M.; Owens, G. D.; Margerum, D. W. Inorg. Chem. 1979, 18, 15381. (c) Szalda, D. J.; Creutz, C.; Mahajan, D.; Sutin, N. Inorg. Chem. 1983, 22, 2372. (d) Nusbaumer, H. Ph.D. Thesis, EPFL, 2004. (19) Nusbaumer, H.; Zakeeruddin, S. M.; Moser, J.-E.; Gr€atzel, M. Chem.; Eur. J. 2003, 9, 3756.
Langmuir 2009, 25(13), 7602–7608
and 0.2 M LiClO4 in acetonitrile/ethylene carbonate (4:6/v:v). The Co3+ complex itself was synthesized and isolated as reported earlier.18d
Successive Ionic Layer Adsorption and Reaction (SILAR) Deposition of PbS and CdS QDs. For in situ deposition of PbS QDs from their precursor solutions, four different beakers were prepared; one contained 0.02 M Pb(NO3)2 in methanol, with a few hours of stirring being necessary for complete dissolution, a second contained 0.02 M Na2S in methanol, and the other two contained pure methanol to rinse the samples from the excess of each precursor solution. The TiO2-modified electrode (see below) was dipped into the Pb2+ solution, pure methanol (then dried in air), the S2- solution, and then pure methanol (then dried in air) successively for 1 min each. Such an immersion cycle was repeated several times (five or six times). The electrode became darker as the number of SILAR cycles was increased. For CdS QDs, the SILAR process was the same as for PbS QDs except for minor differences in the concentrations of precursors and cycle numbers; 0.05 or 0.4 M Cd(NO3)2 dissolved in ethanol and 0.05 or 0.1 M Na2S dissolved in methanol, respectively. When the 0.4 M Cd2+ solution was used, four SILAR cycles were performed, whereas nine cycles were performed when the 0.05 M solution was used. Assembly of QD-Sensitized Cells. Photoelectrodes consisted of a TiO2 film with a double or triple layer structure. A compact blocking underlayer of spray-pyrolyzed TiO2 (∼80 nm thick) was deposited onto a cleaned conducting glass substrate (NSG, F-doped SnO2, resistance 10 Ω sq-1). A solution of titanium diisopropoxide bis(acetylacetonate) in ethanol (0.02 M) was sprayed 16 times over the conducting glass surface, which was maintained at 450 °C. The treated glass plates were fired at 450 °C for 30 min to remove remaining organic traces. Successive depositions of a 2.8 μm thick transparent layer and a 5.6 μm thick 60 nm light-scattering layer by screen-printing, and final posttreatment with an aqueous solution of TiCl4, were then carried out according to typical procedures done in our laboratory for dye cells.20 After preparing the QD-sensitized electrode by the SILAR process described above, the cell was assembled using a transparent hot-melt 25 μm thick Surlyn ring (DuPont) as a spacer between the QD-sensitized electrode and the counter electrode (Pt on FTO glass, chemical deposition of 0.05 M hexachloroplatinic acid in 2-propanol at 400 °C for 20 min). The electrolyte was injected into the interelectrode space from the counter electrode side through a predrilled hole, and the hole was then sealed with a Bynel sheet and a thin glass slide cover by heating. All of the procedures in preparing electrodes and assembling cells were the same as in our typical dye-sensitized cells,20 except the step of QD deposition onto the TiO2 films. The active area (0.159 cm2) was defined with a mask in the same way as typical dye cells are tested in our laboratory. Photocurrent-Voltage Measurements. The light source for the photocurrent-voltage (I-V) measurement is a 450 W xenon lamp (Osram XBO 450). The incident light intensity was calibrated with a standard Si solar cell. The spectral output of the lamp matched precisely the standard global AM 1.5 solar spectrum in the region of 350-750 nm (mismatch < 2%) by the aid of a Schott K113 Tempax sunlight filter (Pr€azisions Glas & Optik GmbH, Germany). Various irradiance intensities from 0.01 to 1.0 sun can be provided with a neutral density wire mesh. The current-voltage curves were obtained by measuring the photocurrent of the cells using a Keithley model 2400 digital source meter under an applied external potential scan. The transient current dynamics characterization of the cell was obtained using the same system of I-V measurements. The measurement of IPCE was performed by using a similar data (20) (a) Kuang, D.; Ito, S.; Wenger, B.; Clein, C.; Moser, J.-E.; Humphry-Baker, R.; Zakeeruddin, S. M.; Gr€atzel, M. J. Am. Chem. Soc. 2006, 128, 4146. (b) Ito, S.; Chen, P.; Comte, P.; Nazeeruddin, M. K.; Liska, P.; Pechy, P.; Gr€atzel, M. Prog. Photovoltaics 2007, 15, 603.
DOI: 10.1021/la900247r
7603
Article collection system but under monochromatic light obtained by passing the output of a 300 W xenon lamp (ILC Technology) through a Gemini-180 double monochromator (Jobin Yvon Ltd., U.K.). Transmission Measurements. To check the optical properties of as-prepared electrodes, transmission spectra of both PbSand CdS-sensitized electrode were measured by using a Varian Cary 5 spectrophotometer fitted with an 11 cm diameter integrating sphere coated with polytetrafluoroethylene (PTFE). Optical characteristics were checked for transparent TiO2 films as well as the ones including light scattering layers. Photoelectrochemical Characterization. Photocurrents were measured in a three-electrode configuration with 0.1 M Na2S (Aldrich, pro analysis in Milli-Q water, 25 °C) as electrolyte, Ag/AgCl/sat. KCl as reference, and platinum wire as counter electrode, separated by glass frits. The potential of the photoelectrode was controlled by using a potentiostat (PAR 273A). The TiO2/PbS film was illuminated through 6 mm of electrolyte and a fused silica window with a 0.5 cm2 circular mask. Sunlight was simulated with a 450 W xenon lamp (Osram, ozone free) with the light intensity adjusted so that a Si photodiode gave the same photocurrent as that in global AM 1.5 sunlight of 1000 W/m2. Photocurrent action spectra were obtained under light from a 300 W xenon lamp with an integrated parabolic reflector (Cermax PE 300 BUV) passing through a monochromator (Bausch & Lomb, bandwidth 10 nm fwhm). The wavelength was scanned at 1 nm/s, and the monochromatic photocurrent of the TiO2/PbS electrode compared with that of a UV enhanced Si photodiode (Oriel 71883) of known IPCE spectrum. Electrochemical impedance spectroscopy (EIS) was carried out by using a FTO-Pt symmetric configuration. The real and imaginary part of the impedance of the cell was recorded using an Autolab potentiostat/galvanostat PGSTAT 30 equipped with a FRA impedance generator module. The EIS spectrum was recorded at 20 °C by scanning frequencies from 60 kHz to 10 mHz with a sinusoidal potential of 10 mV amplitude.
Results and Discussion A simplified schematic diagram of the SILAR process for preparing PbS- and CdS-sensitized electrodes is presented in the inset of Figure 1. This SILAR process looks facile and straightforward for depositing the wanted QDs and any additional coatings over the mesoporous metal oxide films if their appropriate precursors are available. After a few cycles of SILAR from the precursor solutions of Pb2+ (or Cd2+) and S2-, the color changes of the electrodes were clear, as shown in the Figure 1 inset. Based on such observations, the SILAR process has been used for a long time to make photoactive layers over various substrates, and it is considered a very effective way to prepare QDs inside the mesopores of metal oxide films.9,10,21 As confirmed in recent reports by another group22 and ours,23 alcoholic medium-based SILAR gave more reproducible and better results in QD photoelectrochemical cells than the classical aqueous solvent-based SILAR procedure used so far because alcohols have better wetting and faster drying characteristics than water, which could lead to formation of better-defined QDs through narrow pores onto mesoporous metal oxides. Therefore, Pb2+ and S2- precursors were dissolved in methanol while Cd2+ precursors were dissolved in ethanol owing to their different solubility. Their concentrations were adjusted to reach optimal photovoltaic performance. (21) Pathan, H. M.; Lokhande, C. D. Bull. Mater. Sci. 2004, 27, 85. (22) Chang, C.-H.; Lee, Y.-L. Appl. Phys. Lett. 2007, 91, 053503. (23) Lee, H. J.; Leventis, H. C.; Moon, S.-J.; Chen, P.; Ito, S.; Haque, S. A.; Torres, T.; N€uesch, F.; Geiger, T.; Zakeeruddin, S. M.; Gr€atzel, M.; Nazeeruddin, Md. K. Adv. Func. Mater. DOI: adfm.200900081.
7604 DOI: 10.1021/la900247r
Lee et al.
Figure 1. Transmission (%) spectra of PbS- and CdS-sensitized electrodes obtained after applying the SILAR process as described in the text. Inset: schematic diagram of the SILAR process and pictures of the QD-sensitized photoelectrodes.
Transmission spectra of typical PbS- and CdS-sensitized electrodes are shown in Figure 1. As expected, the transmission of the CdS-TiO2 film decreases at wavelengths shorter than about 510 nm due to absorption by the CdS deposited (bulk energy gap = 2.5 eV). The PbS-sensitized film shows a very broad feature throughout the visible spectrum that also tails at least down to 1000 nm (bulk energy gap ∼ 0.4 eV). These transmission spectra of PbS- and CdS-sensitized films agree well with the IPCE curves obtained from the corresponding solar cells (vide infra). After trying to optimize the thickness and composition of the TiO2 films used, we have come to the conclusion that a rather thin transparent layer (2-3 μm) should be maintained for better overall performance. The improvement of light harvesting within the film was provided by a second layer of ∼5.6 μm that acts to backscatter unabsorbed photons and consequently improve further overall efficiency. As for the SILAR PbS deposition, a low concentration precursor solution in methanol (20 mM) was found to be better than the higher concentration aqueous solutions (>1 M or saturated) described more typically in the literature. It was demonstrated recently that the SILAR process in alcoholic medium with precursors of low concentration produces better defined QDs over mesoporous TiO2 films, thus resulting in better photovoltaic performances.10,22,23 These new findings imply that the solvent in which SILAR is carried out and the concentration of precursors are both important parameters in the preparation of superior QD sensitizers within mesoporous films of small pore size. A reliable delivery of the Pb2+ and S2- precursors to the surface of TiO2 particles deep within the film, and then fast elimination of solvent during the drying process, could help the sequential growth of homogeneous particles or thin layers. Those advantages could help avoid clogging of the mesopores as the number of SILAR cycles is increased. Following those improved procedures in the SILAR process, we could construct a PbS-sensitized solar cell showing about 2% overall efficiency in a regenerative mode with the cobalt redox couple, [Co(o-phen)3]2+/3+. As can be seen in Figure 2a, the obtained IPCE value was over 50% throughout a large part of the visible range, and its tail extended up to 900 nm or more. The shape of the IPCE curve reflects the pattern of untransmitted light of the same film very well (Figure 1), indicating that most of the untransmitted light is ultimately absorbed by the PbS. To the best of our knowledge, this is the best IPCE value from PbS QD-sensitized liquid cells working in a regenerative mode. The integrated photocurrent from the IPCE curve was calculated to be about Langmuir 2009, 25(13), 7602–7608
Lee et al.
Article
Figure 2. (a) Photocurrent action spectra of the best PbS QD-sensitized cell working with a regenerative cobalt complex, [Co(o-phen)32+/3+], and (b) photocurrent density-voltage curves of the same device under various light intensities indicated. Table 1. Summary of Short-Circuit Currents, Open-Circuit Voltages, Fill Factors, and Overall Conversion Efficiencies Obtained from the Data Shown in Figure 2b
9.4% sun 29.8% sun 99.7% sun
Jsc (mA/cm2)
Voc (V)
FF
efficiency (%)
0.82 2.06 2.64
0.34 0.38 0.41
0.69 0.55 0.57
2.01 1.43 0.62
8-9 mA/cm2, which agrees well with the short-circuit current measured at low intensity (Table 1) and manifests the effectiveness of the charge generation and extraction from the interfaces between TiO2/PbS and [Co(o-phen)3]2+/3+ at low intensity condition. The overall efficiency of 2.01% at 9.4% sun intensity comes mostly from the high short-circuit current (Isc) despite a rather low open-circuit voltage of ∼0.3 V. Based on the optical band gap of the deposited PbS (Eg = 1.2-1.5 eV), this result suggests that further increasing of Voc can still be expected by further pursuing the efforts of understanding the overpotentials for charge transfers at the interfaces. While the QD regeneration efficiency leading to a high photocurrent by the regenerative cobalt complex redox couple looks promising here and in previous reported results,6a,8 being also somewhat competitive with I -/I3- at low intensity sun, the short-circuit current was not increased in a proportional way as light intensities increased as can be seen in Figure 2b. This effect results in a substantial reduction of overall efficiency from 2.01% at 9.4% sun intensity to 0.64% at full-sun intensity (Table 1). To better understand the origin of the deviancy between Jsc and light illumination, we further characterized the kinetic behavior of this [Co(o-phen)3]2+/3+ complex by means of electrochemical impedance spectroscopy to assess on the charge transfer resistance of this redox couple at the counter electrode as well as on its diffusion coefficient within the electrolyte. On the other hand, the chronoamperometric method in the complete device was also herein used under light illumination to evaluate the mobility of the complex within the mesoporous structure of the PbS-sensitized TiO2. Figure 3a shows the impedance response of the complex within the electrolyte when using Pt-catalyzed FTO symmetric cells. It depicts two main features. The first, in the high frequencies region, arises from the charge transfer resistance at the FTO-Pt electrode coupled with the double layer charge capacitance. The second phenomenon, at lower relaxation frequencies, traces the well-established semifinite Warburg impedance relative to the Langmuir 2009, 25(13), 7602–7608
diffusion control by the redox mediator. As a result from the simulation of the spectra using Zview software, the [Co(o-phen)3]2+/3+ exhibits a low charge transfer resistance Rct = 1.5 Ω 3 cm2 while displaying a diffusion coefficient of ∼2.1 10-7 cm2/s. Although the charge transfer resistance of this cobalt complex is relatively comparable to the one of the triiodide/iodide in acetonitrile-based solvent, the diffusion coefficient is herein found to be 3 orders of magnitude lower than that of triiodide in acetonitrile or more than 1 order of magnitude lower than that of the triiodide in the low volatile 3-methoxy propionitrile solvent. This low diffusion coefficient, that principally results from the low polarizability of the complex, confers to this QD-sensitized TiO2 device a kinetic control of the cell governed by the mass transport. This statement is notably supported by the chronoamperogram shown in Figure 3b where an exponential decrease of the current as a function of time is experienced. The current decreases perfectly follow a linear evolution as a function of t-1/2 according to the Cottrell equation from which the diffusion coefficient of the complex is extracted from the slope. Interestingly, the two experiments do not yield to the same value. The use of the Cottrell equation leads to an apparent diffusion coefficient of ∼7.5 10-11 cm2/s that is almost 4 orders of magnitude lower than that in the electrolyte. This difference clearly culprits the strong coulombic/steric interaction existing between the complex and the PbS-sensitized TiO2 with a significant concentration depletion of the redox species within the mesopore film. This diffusion control of the cell also gives an explanation about the necessity to keep a photoanode thickness as low as possible (
