Blue copper model complexes with distorted tetragonal geometry acting as effective electron-transfer mediators in dye-sensitized solar cells

Article abstract of DOI:10.1021/ja0506814

The electron self-exchange rate constants of blue copper model complexes, [(-)-sparteine-N,N′](maleonitriledithiolato-S,S′)copper ([Cu(SP)(mmt)])^0/-, bis(2,9-dimethy-1,10-phenanthroline)copper ([Cu(dmp)₂]^2+/+), and bis(1,10-phenanthroline)copper ([Cu(phen)₂]^2+/+) have been determined from the rate constants of electron transfer from a homologous series of ferrocene derivatives to the copper(II) complexes in light of the Marcus theory of electron transfer. The resulting electron self-exchange rate constant increases in the order: [Cu(phen)₂]^2+/+ < [Cu(SP)(mmt)]^0/- < [Cu(dmp)₂]^2+/+, in agreement with the order of the smaller structural change between the copper(II) and copper(I) complexes due to the distorted tetragonal geometry. The dye-sensitized solar cells (DSSC) were constructed using the copper complexes as redox couples to compare the photoelectrochemical responses with those using the conventional I ₃^-/I^- couple. The light energy conversion efficiency (η) values under illumination of simulated solar light irradiation (100 mW/cm²) of DSSCs using [Cu(phen)₂] ^2+/+, [Cu(dmp)₂]^2+/+, and [Cu(SP)(mmt)] ^0/- were recorded as 0.1%, 1.4%, and 1.3%, respectively. The maximum η value (2.2%) was obtained for a DSSC using the [Cu(dmp)₂] ^2+/+ redox couple under the light irradiation of 20 mW/cm² intensity, where a higher open-circuit voltage of the cell was attained as compared to that of the conventional I₃^-/I^- couple.

Full text of DOI:10.1021/ja0506814

Published on Web 06/10/2005
Blue Copper Model Complexes with Distorted Tetragonal
Geometry Acting as Effective Electron-Transfer Mediators in
Dye-Sensitized Solar Cells
Shigeki Hattori, Yuji Wada, Shozo Yanagida, and Shunichi Fukuzumi*
Contribution from the Department of Material and Life Science, Graduate School of
Engineering, Osaka UniVersity, SORST, Japan Science and Technology Agency, Suita,
Osaka 565-0871, Japan
Received February 2, 2005; E-mail: fukuzumi@ap.chem.eng.osaka-u.ac.jp
Abstract: The electron self-exchange rate constants of blue copper model complexes, [(-)-sparteine-
N,N′](maleonitriledithiolato-S,S′)copper ([Cu(SP)(mmt)])0/-, bis(2,9-dimethy-1,10-phenanthroline)copper
([Cu(dmp)2]2+/+), and bis(1,10-phenanthroline)copper ([Cu(phen)2]2+/+) have been determined from the rate
constants of electron transfer from a homologous series of ferrocene derivatives to the copper(II) complexes
in light of the Marcus theory of electron transfer. The resulting electron self-exchange rate constant increases
in the order: [Cu(phen)2]2+/+ < [Cu(SP)(mmt)]0/- < [Cu(dmp)2]2+/+, in agreement with the order of the smaller
structural change between the copper(II) and copper(I) complexes due to the distorted tetragonal geometry.
The dye-sensitized solar cells (DSSC) were constructed using the copper complexes as redox couples to
compare the photoelectrochemical responses with those using the conventional I3-/I- couple. The light
energy conversion efficiency (η) values under illumination of simulated solar light irradiation (100 mW/cm2)
of DSSCs using [Cu(phen)2]2+/+, [Cu(dmp)2]2+/+, and [Cu(SP)(mmt)]0/- were recorded as 0.1%, 1.4%, and
1.3%, respectively. The maximum η value (2.2%) was obtained for a DSSC using the [Cu(dmp)2]2+/+ redox
couple under the light irradiation of 20 mW/cm2 intensity, where a higher open-circuit voltage of the cell
was attained as compared to that of the conventional I3-/I- couple.
-
Introduction
reported to find electron-transfer mediators other than the I3
/
I- couple.6-10 Although the redox potentials of some mediators,
Nanocrystalline dye-sensitized solar cells (DSSC), pioneered
by Gra¨tzel and co-workers, have attracted increasing attention
as an inexpensive alternative to the conventional p-n junction
solar cells.1-5 The I3-/I- redox couple has so far been employed
as the best electron-transfer mediators working in DSSC.1-5
However, the I3-/I- couple has the following undesirable
-
employed as rivals of I3-/I-, are more positive than that of I3
/
I-, no significant improvement on the open-circuit potential of
the cells has so far been achieved.6-10
In the natural electron transport systems, copper complexes,
known as blue copper proteins, act as efficient electron-transfer
mediators.11-13 Plastocyanin is involved in electron transfer
between Photosystems I and II in the photosynthetic chain, and
azulin transfers electrons between cytochrome c551 and cyto-
chrome c oxidase in respiratory chain.11 Electron transfer is
chemical properties.4 First, I3 at high concentrations, which
-
are required to attain the good performance, absorbs a significant
amount of the visible light. Second, the low triiodide ion redox
potential prevents attaining a high open-circuit voltage. A large
polarization loss results from the potential mismatch between
the redox couples of the sensitizers (S+/S) and the mediator
(I3-/I-). Last, I2, which is corrosive, has a substantial vapor
pressure and thus would escape from the cell unless it is
hermetically sealed. A number of attempts have thereby been
(6) (a) Nusbaumer, H.; Moser, J.-E.; Zakkeruddin, S. M.; Gra¨tzel, M. J. Phys.
Chem. B 2001, 105, 10461. (b) Nusbaumer, H.; Zakeeruddin, S. M.; Moser,
J.-E.; Gra¨tzel, M. Chem.-Eur. J. 2003, 9, 3756.
(7) Oskam, G.; Bergeron, B. V.; Meyer, G. J.; Searson, P. C. J. Phys. Chem.
B 2001, 105, 6867.
(8) Gregg, B. A.; Pichot, F.; Ferrere, S.; Fields, C. L. J. Phys. Chem. B 2001,
105, 1422.
(9) Sapp, S. A.; Elliott, M.; Contado, C.; Caramori, S.; Bignozzi, C. A. J. Am.
Chem. Soc. 2002, 124, 11215.
(1) (a) O’Regan, B.; Gra¨tzel, M. Nature 1991, 353, 737. (b) Gra¨tzel, M. Nature
2001, 414, 338.
(2) (a) Hagfeldt, A.; Gra¨tzel, M. Chem. ReV. 1995, 95, 49. (b) Hagfeldt, A.;
Gra¨tzel, M. Acc. Chem. Res. 2000, 33, 269. (c) Gra¨tzel, M. Pure Appl.
Chem. 2001, 73, 459.
(3) (a) Nazeeruddin, M. K.; Pe´chy, P.; Renouard, T.; Zakeeruddin, S. M.;
Humphry-Baker, R.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.; Shklover,
V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gra¨tzel, M. J. Am. Chem.
Soc. 2001, 123, 1613. (b) Wang, P.; Klein, C.; Humphry-Baker, R.;
Zakeeruddin, S. M.; Gra¨tzel, M. J. Am. Chem. Soc. 2005, 127, 808.
(4) Gra¨tzel, M.; Moser, J.-E. In Electron Transfer in Chemistry; Balzani, V.,
Ed.; Wiley-VCH: Weinheim, Germany, 2001; Vol. V, pp 589-644.
(5) Hara, K.; Arakawa, H. In Handbook of PhotoVoltaic Science and Engineer-
ing; Luque, A., Hegedus, S., Eds.; John Wiley & Sons Ltd.: Chichester,
UK, 2003; pp 663-700.
(10) Cameron, P. J.; Peter, L. M.; Zakeeruddin, S. M.; Gra¨tzel, M. Coord. Chem.
ReV. 2004, 248, 1447.
(11) (a) Gray, H. B.; Solomon, E. I. In Copper Proteins; Spiro, T. G., Ed.;
Wiley: New York, 1981; pp 1-39. (b) Sykes, A. G. AdV. Inorg. Chem.
1991, 107, 377.
(12) Gross, E. L. Photosynth. Res. 1993, 37, 103.
(13) (a) Groeneveld, C. M.; Canters, G. W. J. Biol. Chem. 1988, 263, 167. (b)
Groeneveld, C. M.; Ouwerling, M. C.; Erkelens, C.; Canters, G. W. J. Mol.
Biol. 1988, 200, 189. (c) Groeneveld, C. M.; Dahlin, S.; Reinhammar, B.;
Canters, G. W. J. Am. Chem. Soc. 1987, 109, 3247. (d) Canters, G. W.;
Hill, H. A. O.; Kitchen, N. A.; Adman, E. T. J. Magn. Reson. 1984, 57, 1.
(e) Dahlin, S.; Reinhammar, B.; Wilson, M. T. Biochem. J. 1984, 218,
609. (f) Kyritsis, P.; Dennison, C.; Ingledew, W. J.; McFarlane, W.; Sykes,
A. G. Inorg. Chem. 1995, 34, 5370.
9
9648
J. AM. CHEM. SOC. 2005, 127, 9648-9654
10.1021/ja0506814 CCC: $30.25 © 2005 American Chemical Society

Blue Copper Model Complexes
A R T I C L E S
triledithiolato-S,S′)copper ([Cu(SP)(mmt)]),25 as a blue copper
model complex. Although the (-)-sparteine and the maleoni-
triledithiolate ligands do not accurately represent active site of
the blue copper(II) proteins, the model complex closely mimics
spectral and redox behaviors in the blue copper centers.25 The
intense bands in the region of 500-1000 nm are assigned to S
f Cu(II) charge transfer (CT) bands.25 A distorted tetragonal
geometry is also obtained for bis(2,9-dimethy-1,10-phenanthro-
line)copper ([Cu(dmp)2](CF3SO3)2)26-29 by introducing methyl
groups in the 2,9 positions of phenanthroline ligand as a result
of the steric hindrance provided by two methyl groups.30 We
have examined incident photon-to-current conversion efficiency
and the photoelectrochemical responses of DSSCs containing
[Cu(SP)(mmt)]0/-, [Cu(dmp)2]2+/+, and Cu(phen)2]2+/+ in com-
parison with the electron-transfer properties in solution.
Figure 1. Structures of copper complexes used in this study.
known to proceed via transition state structures, which are
intermediate between those of the reactant and product, accord-
ing to the Franck-Condon principle as clarified by Marcus.14
Thus, electron transfer does not occur until the metal center is
vibrationally excited to match a geometry appropriate for that
of the product complex.14 This is accomplished by a simple
adjustment in the bond lengths, but for copper such adjustment
in the geometry usually requires a large energy because copper-
(I) and copper(II) have different preferred geometries, that is,
tetrahedral and tetragonal, respectively.15,16 In the natural system,
the difference in bond lengths and geometries between copper-
(I) and copper(II) is cleverly minimized by the protein structure,
which provides a copper site optimized for fast electron
transfer.17,18 In plastocyanin, a distorted tetragonal geometry,
which is an intermediate geometry between copper(I) and
copper(II), with an unusually long copper-methionine interac-
tion is found for each oxidation state.17,18 Plastocyanin, classified
as the type I blue copper protein, compared to tetragonal copper
complexes exhibits intense absorption at ca. 600 nm, distinctive
electron spin resonance spectra, and unusually high redox
potentials.19,20 Considerable efforts have been devoted to the
design of blue copper model complexes.21-25 However, blue
copper model complexes with favorable properties such as
efficient electron transfer and high redox potentials have yet to
be employed as effective electron mediators to attain the high
energy conversion efficiency and high open-circuit voltages of
the cells.
Experimental Section
Materials. [Cu(SP)(mmt)], [Cu(dmp)2](CF3SO3)2, [Cu(dmp)2](CF3-
SO3), [Cu(phen)2](CF3SO3)2, and [Cu(phen)2](CF3SO3) were prepared
and characterized according to the literature.25-29 [Cu(SP)(mmt)]Na was
prepared by the reduction of [Cu(SP)(mmt)] using NaSCN aq. The Ru-
dye, cis-dithiocyanato-N,N′-bis(4-carboxylato-4-tetrabutylammonoium-
carboxylate-2,2′-bipyridine)ruthenium(II) (N719), was synthesized from
cis-dithiocyanato-N,N′-bis(4,4′-dicarboxylate-2,2′-bipyridine)ruthenium-
(II) as described in the literature.31 Dimethylferrocene ([Fe(C5H4Me)2]),
benzoylferrocene ([Fe(C5H5)(C5H4COOBzl)]), and decamethylferrocene
(Fe(C5Me5)2) were purchased from Wako Pure Chemical Industries Ltd.
All solvents and chemicals were of reagent grade quality, and they
were purchased and used without further purification unless otherwise
noted.
Electron-Transfer Kinetics of Copper Complexes. Kinetic mea-
surements for electron transfer from ferrocene derivatives to copper-
(II) complexes were carried out using a UNISOKU RSP-601 stopped-
flow rapid scan spectrometer in deaerated acetonitrile at 298 K. In each
case, it was confirmed that the rate constants of electron transfer (ket)
derived from at least five independent measurements agreed within an
experimental error of (5%. Pseudo-first-order rate constant k1 was
determined by a least-squares curve fit using a Macintosh personal
computer. The pseudo-first-order plots of ln(A∞ - A) versus time were
linear for three or more half-lives with the correlation coefficient F >
0.999. Second-order rate constants were calculated from the slope of
k1 versus the concentration of excess substrate by the least-squares
analysis.
We report herein the use of blue copper model complexes
with distorted tetragonal geometry (Figure 1) as electron-transfer
mediators in DSSC in comparison with a tetragonal copper
complex, [Cu(phen)2](CF3SO3)2 (phen ) 1,10-phenanthroline),
in which the four N-donor atoms are expected to be nearly
coplanar. First, we employed [(-)-sparteine-N,N′](maleoni-
Electrochemical Measurements. Electrochemical measurements
were performed on a BAS 100 W electrochemical analyzer in deaerated
(14) (a) Marcus, R. A.; Eyring, H. Annu. ReV. Phys. Chem. 1964, 15, 155. (b)
Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265.
(15) Rorabacher, D. B. Chem. ReV. 2004, 104, 651.
acetonitrile containing 0.10 M NBun ClO4 as a supporting electrolyte
4
at 298 K. A conventional three-electrode cell was used with a platinum-
working electrode (surface area of 0.3 mm2), and a platinum wire was
used as the counter electrode. The working electrode was polished with
BAS polishing alumina suspension and rinsed with acetone before use.
The measured potentials were recorded with respect to an Ag/AgNO3
(0.01 M) reference electrode and converted to vs SCE by adding 0.29
V.32
(16) (a) Hathaway, B. J.; Billing, D. E. Coord. Chem. ReV. 1970, 5, 143. (b)
Hathaway, B. J. Coord. Chem. ReV. 1981, 35, 211.
(17) Guss, J. M.; Bartunik, H. D.; Freeman, H. C. Acta Crystallogr. 1992, B48,
790.
(18) Adman, E. T. AdV. Protein Chem. 1991, 42, 145-197.
(19) (a) Solomon, E. I.; Lowery, M. D. Science 1993, 259, 1575. (b) Randall,
D. W.; Gamelin, D. R.; LaCroix, L. B.; Solomon, E. I. J. Biol. Inorg. Chem.
2000, 5, 16. (c) Penfield, K. W.; Gewirth, A. A.; Solomon, E. I. J. Am.
Chem. Soc. 1985, 107, 4519.
(20) (a) Solomon, E. I.; Baldwin, M. J.; Lowery, M. D. Chem. ReV. 1992, 92,
521. (b) Solomon, E. I.; Szilagyi, R. K.; George, S. D.; Basumallick, L.
Chem. ReV. 2004, 104, 419.
(26) Hoffmann, S. K.; Corvan, P. J.; Singh, P.; Sethulekshmi, C. N.; Metzger,
R. M.; Hatfield, W. E. J. Am. Chem. Soc. 1983, 105, 4608.
(27) Pawlowski, V.; Kunkely, H.; Zabel, M.; Vogler, A. Inorg. Chim. Acta 2004,
357, 824.
(28) Liu, S.-J.; Huang, C.-H.; Chang, C.-C. Mater. Chem. Phys. 2003, 82, 551.
(29) Tran, D.; Skelton, B. W.; White, A. H.; Laverman, L. E.; Ford, P. C. Inorg.
Chem. 1998, 37, 2505.
(30) Although [Cu(SP)(mmt)] has a relatively large extinction coefficient at 736
nm (ꢀ ) 2000 M-1 cm-1) due to the S f CuII charge transfer (CT) band,25
[Cu(dmp)2]2+ has a much smaller extinction coefficient at 740 nm (ꢀ )
100 M-1 cm-1), see: Sundararajan, S.; Wehry, E. L. J. Phys. Chem. 1972,
76, 1528.
(31) Nazeeruddin, M. K.; Zakeeruddin, S. M.; Humphry-Baker, R.; Liska, J.
P.; Shklover, N. V.; Fischer, C.-H.; Gra¨tzel, M. Inorg. Chem. 1999, 38,
6298.
(21) (a) Kitajima, N.; Fujisawa, K.; Moro-oka, Y. J. Am. Chem. Soc. 1990, 112,
3210. (b) Kitajima, N.; Fujisawa, K.; Tanaka, M.; Moro-oka, Y. J. Am.
Chem. Soc. 1992, 114, 9232.
(22) (a) Holland, P. L.; Tolman, W. B. J. Am. Chem. Soc. 1999, 121, 7270. (b)
Holland, P. L.; Tolman, W. B. J. Am. Chem. Soc. 2000, 122, 6331.
(23) (a) Randall, D. W.; DeBeer George, S.; Holland, P. L.; Hedman, B.;
Hodgson, K. O.; Tolman, W. B.; Solomon, E. I. J. Am. Chem. Soc. 2000,
122, 11632. (b) Randall, D. W.; DeBeer George, S.; Hedman, B.; Hodgson,
K. O.; Fujisawa, K.; Solomon, E. I. J. Am. Chem. Soc. 2000, 122, 11620.
(24) (a) Anderson, O. P.; Perkins, C. M.; Brito, K. K. Inorg. Chem. 1983, 22,
1267. (b) Sugiura, Y. Inorg. Chem. 1978, 17, 2176.
(25) Kim, Y.-J.; Kim, S.-O.; Kim, T.-I.; Choi, S.-N. Inorg. Chem. 2001, 40,
4481.
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9649

A R T I C L E S
Hattori et al.
Preparation of Working Electrodes for Dye-Sensitized Solar
Cells. Nanoporous TiO2 electrode consisted of a TiO2 film with a triple-
layer structure. A compact blocking underlayer of coated TiO2 was
deposited onto an optically transparent electrode (OTE) (Nippon Sheet
Glass, SnO2:F, 8 Ω/cm2). A solution of titanium tetraisopropoxide in
an ethanol-water mixed solution was spin-coated on the OTE surface.
The resulting electrodes were annealed at 450 °C in air for 30 min. A
transparent layer was prepared by dropping and spreading the colloidal
suspensions of TiO2 nanoparticles (P25, d ) 21 nm, Nippon Aerogel)
on OTE with a glass rod and adhesive tapes as a spacer. The resulting
electrodes (OTE/TiO2) were annealed at 450 °C in air for 30 min. The
thickness of the electrodes ranged between 4 and 5 µm as measured
by a profiler. Post-treatment with an aqueous solution of TiCl4 was
then carried out according to a previously published procedure.6
A
scattering layer was prepared by using TiO2 nanoparticles (d ) 400
nm, Nippon Shokubai) under the same experimental conditions as
described above.
Coloration of the TiO2 surface with the Ru dye (N719)30 was carried
out by immersing the film (still warm, i.e., 80-100 °C) for 12 h in a
tert-butyl alcohol-acetonitrile (1:1) mixture containing 3.0 × 10-4
M
of N719. After completion of the dye adsorption, the electrode was
withdrawn from the solution, washed with acetonitrile, and dried.
Electrolyte solutions were prepared by dissolving a copper complex
redox mediator (0.2 M, CuII/(CuI+CuII) ) 0.4), tert-butylpyridine (0.5
M), and LiClO4 (0.5 M) in anhydrous methoxyacetonitrile.
Photoelectrochemical Measurements. Photocurrent-voltage char-
acteristics of solar cells were measured using a voltage-current source
monitor (Advantest R6246) under AM 1.5 simulated light source
(Yamashita Denso, YSS-80). Measurements of incident photon-to-
photocurrent efficiency (IPCE) were carried out using a setup for IPCE
measurements (PV-25DYE, JASCO) under 5 mW/cm2 monochromatic
light illuminations. Solar cells were made by placing the Pt sputtered
optically transparent electrode (F-doped SnO2 sheet resistance 10 Ω/cm2,
Asahi Glass) on the dye-coated TiO2 film. The electrolyte was injected
from one hole made on the counter electrode. The gap between the
porous dye coated-TiO2 electrode and the Pt sputtered-counter electrode
was adjusted to be a few tens micrometers and sealed using thermal
adhesive films (HIMILAN 1652: Mitsui-Dupont Polychemical), and
the electrodes were sealed by heat. The injection holes were sealed by
cover glass (Iwaki Glass), HIMILAN film, and epoxy resin (Torr seal:
Varian Vacuum Products).
Results and Discussion
Kinetics of Electron Transfer from Ferrocene Derivatives
to Copper Complexes. To compare the electron-transfer
reactivities of all of the copper complexes employed in this
study, the kinetic studies on the reduction of copper complexes
by ferrocene derivatives were performed. The one-electron
Figure 2. Kinetic curves for reduction of (a) [Cu(SP)(mmt)] (3.0 × 10-5
M) by [Fe(C5H4Me)2] (6.0 × 10-4 M) monitored by a decrease in
absorbance at 738 nm due to [Cu(SP)(mmt)], (b) reduction of [Cu(dmp)2]-
(CF3SO3)2 (1.25 × 10-6 M) by [Fe(C5H5)(C5H4COOBzl)] (2.5 × 10-5 M)
monitored by an increase in absorbance at 456 nm due to [Cu(dmp)2]+,
and (c) reduction of [Cu(phen)2](CF3SO3)2 (5.0 × 10-5 M) by [Fe(C5Me5)2]
(1.0 × 10-3 M) monitored by an increase in absorbance at 438 nm due to
[Cu(phen)2]+ in MeCN at 298 K.
reduction potentials (E0 vs SCE) of [Cu(SP)(mmt)], [Cu-
red
(dmp)2](CF3SO3)2, and [Cu(phen)2](CF3SO3)2 in acetonitrile
(MeCN) were determined to be 0.29, 0.66, and -0.10 V,
respectively (see Supporting Information S1). Thus, ferrocene
derivatives with less positive oxidation potentials than the E0
red
values of copper complexes were chosen as reductants to
examine the kinetics of the electron transfer from ferrocene
derivatives to copper complexes: [Fe(C5H4Me)2] (E0red ) 0.26
As shown in Figure 2, the reduction of [Cu(SP)(mmt)], [Cu-
(dmp)2](CF3SO3)2, and [Cu(phen)2](CF3SO3)2 was detected by
monitoring a decrease in the CuII absorption band at 738 nm,
an increase in the CuI absorption band at 456 nm, and an
increase in the CuI absorption band at 438 nm, respec-
tively.25,34,35 The rates obeyed pseudo-first-order kinetics under
the present experimental conditions where concentrations of
V),33 [Fe(C5H5)(C5H4COOBz)] (E0 ) 0.57 V), and [Fe(C5-
red
Me5)2] (E0 ) -0.08 V)33 for the electron-transfer reduction
red
of [Cu(SP)(mmt)], [Cu(dmp)2](CF3SO3)2, and [Cu(phen)2]-
(CF3SO3)2, respectively.
(32) Mann, C. K.; Barnes, K. K. Electrochemical Reactions in Nonaqueous
Systems; Marcel Dekker: New York, 1970.
(33) (a) Fukuzumi, S.; Mochizuki, S.; Tanaka, T. Inorg. Chem. 1989, 28, 2459.
(b) Fukuzumi, S.; Okamoto, K.; Gros, C. P.; Guilard, R. J. Am. Chem.
Soc. 2004, 126, 10441.
(34) Doine, H.; Yano, Y.; Swaddle, T. W. Inorg. Chem. 1989, 28, 2319.
9
9650 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Blue Copper Model Complexes
A R T I C L E S
Table 1. Electron Self-Exchange Rate Constants (k22) of Copper
Complexes
0 a
4
b
1
∆
Get
10- k12
k22
1
copper complex
ferrocene derivative
(eV)
(M-1 s-
)
(M-1 s-
)
[Cu(SP)(mmt)]
[Cu(dmp)2](CF3SO3)2 [Fe(C5H5)(C5H4COOBzl)] -0.09
[Cu(phen)2](CF3SO3)2 [Fe(C5Me5)2]
[Fe(C5H4Me)2]
-0.03
1.4
6.1
0.19
7.7
23
0.20
0.02
a Free energy change of electron transfer from ferrocene derivatives to
copper(II) complexes. b Rate constants (k12) of electron transfer from
ferrocene derivatives to copper complexes.
0
energy change of electron transfer is slightly positive (∆Get
)
+0.02 eV), when the electron-transfer reaction is reversible (in
equilibrium) at 298 K. In such a case, k1 is expressed by eq 1,
k1 ) k21 + k12[Fc]
(1)
where k12 and k21 are the forward and back electron-transfer
rate constants, respectively.
From the slope and intercept were determined the k12 and
k21 values, from which the electron-transfer equilibrium constant
K ()k12/k21) was obtained as 310 M-1. This value agrees with
the K value (320 M-1) determined independently from the
spectral titration (see Supporting Information S2). The k12 values
are obtained from the slope of the plots. The k12 values of [Cu-
(SP)(mmt)], [Cu(dmp)2](CF3SO3)2, and [Cu(phen)2](CF3SO3)2
are listed in Table 1.
Electron Self-Exchange Rate Constants of Copper Com-
plexes. The electron self-exchange rate constants of [Cu(phen)2]-
(CF3SO3)2 and [Cu(dmp)2](CF3SO3)2 have been reported pre-
viously.34,37 However, the self-exchange rate constants of copper
complexes determined by NMR analyses were sensitively
influenced by their experimental conditions, and there has been
controversy on the self-exchange rate constants of copper
complexes.15 Here, we compare the electron-transfer reactivities
of copper complexes under the same experimental conditions
using a homologous series of ferrocene derivatives. Ferrocene
derivatives are known to react by an outer-sphere mechanism
in a variety of electron-transfer processes especially with
inorganic oxidants.38 A quantitative discussion on electron-
transfer characteristics of copper complexes can be achieved
by comparing the observed rates of outer-sphere electron transfer
to those predicted by the Marcus theory for the rates of outer-
sphere electron-transfer reactions.14,15 The Marcus relation for
the rate constant of the electron transfer from the reductant to
the oxidant, k12, is given by eq 2,14
Figure 3. Plot of pseudo-first-order rate constants k1 versus concentrations
of (a) [Fe(C5H4Me)2] for the reduction of [Cu(SP)(mmt)], (b) [Fe-
(C5H5)(C5H4COOBzl)] for the reduction of [Cu(dmp)2](CF3SO3)2, and (c)
[Fe(C5Me5)2] for the reduction of [Cu(phen)2](CF3SO3)2 in MeCN at 298
K.
k12 ) (k11k22K12f)1/2
(2)
where k11 and k22 are the rate constants of the corresponding
self-exchanges, and K12 is the equilibrium constant for the
electron-transfer reaction. The K12 value is obtained from the
oxidation potential of the reductant, Eox0, and the reduction
potential of the oxidant, Ered0, by using eq 3.14,39 The parameter
f in eq 2 is given by eq 4, in which the frequency factor Z is
ferrocene derivatives were greater than a 20-fold excess of the
copper complex. The pseudo-first-order rate constants (k1)
increase linearly with increasing concentrations of ferrocene
derivatives (Figure 3a-c). Thus, there is no “gated” behavior
that exhibits limiting first-order rate, independent of the reduc-
tant concentration due to the conformational “gating”.15,36 In
the case of electron transfer from Fe(C5Me5)2] (-0.08 V vs
SCE)33 to [Cu(phen)2](CF3SO3)2 (-0.10 V vs SCE), the free
taken as 1 × 1011 M-1 s-1 14
.
The reported values of self-
log K12 ) [F/(-2.3RT)](Eox0 - Ered
)
(3)
(4)
0
log f ) (log K12)2/[4 log(k11k22/Z2)]
(35) Goldstein, S.; Czapski, G. J. Am. Chem. Soc. 1983, 105, 7276.
(36) Koshino, N.; Kuchiyama, Y.; Funahashi, S.; Takagi, H. D. Chem. Phys.
Lett. 1999, 306, 291.
exchange rate constants (k11) for [Fe(C5H4Me)2] (8.3 × 106 M-1
s-1) and [Fe(C5Me5)2] (3.8 × 107 M-1 s-1) constitute a
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9651

A R T I C L E S
Hattori et al.
satisfactory basis for accounting for the experimental kinetic
data when considered in light of the Marcus relation (eqs 2-4).39
The electron self-exchange rate constant (k11) of [Fe(C5H5)(C5H4-
COOBzl)] was taken as that of [Fe(C5H5)2] (5.3 × 106 M-1
s-1).40 The reduction of [Cu(SP)(mmt)] by [Fe(C5H4Bun)2] (k11
) 4.9 × 106 M-1 s-1) was also examined to determine the k22
value for the electron self-exchange between [Cu(SP)(mmt)]
and [Cu(SP)(mmt)]- (see Supporting Information S3). The k22
value of [Cu(SP)(mmt)]/[Cu(SP)(mmt)]- determined from
electron transfer from [Fe(C5H4Me)2] to [Cu(SP)(mmt)] (7.7
M-1 s-1) agrees with that determined independently from
electron transfer from [Fe(C5H4Bun)2] to [Cu(SP)(mmt)] (10.7
M-1 s-1). Such an agreement demonstrates that the electron
transfer between copper complexes and ferrocene derivatives
proceeds by an outer-sphere mechanism.
Figure 4. Comparison of photocurrent action spectra (IPCE values) of
DSSCs containing (a) [Cu(SP)(mmt)]0/-, (b) [Cu(dmp)2]2+/+, and (c)
[Cu(phen)2]2+/+ as the redox couples.
Table 1 lists all of the k22 values of copper complexes
determined in this study. The self-exchange rate constant of a
blue copper model complex, [Cu(SP)](mmt)], is 40 times larger
than that of a reference copper complex, [Cu(phen)2](CF3SO3)2.
The k22 of [Cu(dmp)2](CF3SO3)2 is even larger, and it is 100
times larger than that of [Cu(phen)2](CF3SO3)2. This indicates
that S ligands of blue copper are not essential to attain the
efficient electron-transfer properties of copper complexes. The
order of k22 found in this study is well reflected as the difference
in geometries between copper(I) and copper(II) (vide infra). The
X-ray crystal structures of [Cu(SP)](mmt)] and [Cu(dmp)2]2+/+
have been reported previously.25,27,41 The CuII atom of [Cu-
(SP)](mmt)] adopts a distorted tetragonal geometry with a
dihedral angle of 68° between the N2Cu and the S2Cu planes.25
The CuI atom of [Cu(SP)(mmt)]- may adopt a normal tetrahe-
dral geometry. On the other hand, both CuII and CuI atoms of
[Cu(dmp)2]2+/+ adopt distorted tetragonal geometries with
dihedral angles of 70° and 78°, respectively.27,41 Thus, the
coordination geometry change between CuII and CuI of
copper complexes. The monochromatic incident photon-to-
photocurrent conversion efficiency (IPCE), defined as the
number of electrons generated by light in the outer circuit
divided by the number of incident photons, was determined by
using eq 5,42
IPCE (%) ) 100 × 1240 × Isc(mA/cm2)/[λ(nm) ×
Pin(mW/cm2)] (5)
where Isc is the short circuit photocurrent generated by
monochromatic light, λ is the wavelength of incident mono-
chromatic light, and whose light intensity is Pin.
Figure 4 shows the photocurrent action spectra of DSSCs
using [Cu(SP)(mmt)]0/-, [Cu(dmp)2]2+/+, and [Cu(phen)2]2+/+
.
The IPCE values of DSSCs increase in order: DSSC using
[Cu(phen)2]2+/+ < [Cu(dmp)2]2+/+ < [Cu(SP)(mmt)]0/-. The
maximum of IPCE value reaches 40% in the case of the DSSC
[Cu(dmp)2]2+/+ is certainly smaller than that of [Cu(SP)(mmt)]0/-
leading to the larger k22 value for [Cu(dmp)2]2+/+ than that for
[Cu(SP)(mmt)]0/-
,
using [Cu(SP)(mmt)]0/-
.
The current-voltage (I-V) characteristics of DSSCs using
copper complexes were examined using a thin-layer sandwich-
type solar cell under illumination of simulated AM 1.5 solar
light (100 mW/cm2 from a Xe lamp) to determine the light
energy conversion efficiency (η). The dye-coated TiO2 film as
a working electrode was placed directly on top of an OTE as a
counter electrode, on which the Pt was sputtered. The redox
electrolyte containing a copper complex was injected from a
hole made on the counter electrode to the inter-electrode space.
The solar cells were illuminated in the front through a
conducting glass substrate. Light energy conversion efficiency,
η, is calculated by eq 6,43
.
DSSCs Using Copper Complexes as Redox Couples. The
electrolyte solutions were prepared by dissolving 0.2 M copper
complex, 0.5 M tert-butylpyridine, and 0.5 M LiClO4 in
methoxyacetonitrile. The CuII/(CuI+CuII) ratio was varied to
determine the optimized ratio as 0.4 (CuI of 0.12 M and CuII of
0.08 M, see Supporting Information S4). We used TiO2 films
having a triple-layered structure, in which the first layer is a
compact blocking underlayer, the second a transparent layer
(P25), and the third a scattering layer as in the case of Gra¨tzel’s
report.6b The compact blocking layer and scattering layer have
a better effect on photocurrent generation, respectively (Sup-
porting Information S5, S6). In addition, the film thickness of
the transparent layer was optimized as 4-5 µm because the
photochemical measurements of DSSCs consist of some thick-
ness (Supporting Information S7).
η ) FF × Isc × Voc/Pin
(6)
where the fill factor (FF) is defined as FF ) [IV]max/IscVoc,
where Voc is the open-circuit photovoltage, and Isc is the short
circuit photocurrent.
A series of photocurrent action spectra were recorded to
evaluate the photoelectrochemical response of DSSCs using
(41) Tran, D.; Skelton, B. W.; White, A. H.; Laverman, L. E.; Ford, P. C. Inorg.
Chem. 1998, 37, 2505.
(37) Lee, C.-W.; Anson, F. C. Inorg. Chem. 1984, 23, 837.
(38) (a) Pladziewicz, J. R.; Espenson, J. H. J. Am. Chem. Soc. 1973, 95, 56. (b)
Pelizzetti, E. Inorg. Chem. 1979, 18, 1386. (c) Borchardt, D.; Pool, K.;
Wherland, S. Inorg. Chem. 1982, 21, 93. (d) Borchardt, D.; Wherland, S.
Inorg. Chem. 1986, 25, 901.
(42) Khazraji, A. C.; Hotchandani, S.; Das, S.; Kamat P. V. J. Phys. Chem. B
1999, 103, 4693.
(43) (a) Kamat, P. V.; Barazzouk, S.; Thomas, K. G.; Hotchandani, S. J. Phys.
Chem. B 2000, 104, 4014. (b) Sudeep P. K.; Ipe, B. I.; Thomas, K. G.;
George, M. V.; Barazzouk, S.; Hotchandani, S.; Kamat, P. V. Nano Lett.
2002, 2, 29. (c) Kamat, P. V.; Barazzouk, S.; Hotchandani, S.; Thomas, K.
G. Chem.-Eur. J. 2000, 6, 3914.
(39) The work terms are neglected, because in the present case one of the
reactants and products is uncharged.
(40) Yang, E. S.; Chan, M.-S.; Wahl, A. C. J. Phys. Chem. 1980, 84, 3094.
9
9652 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005

Blue Copper Model Complexes
A R T I C L E S
Figure 5. Photocurrent-photovoltage curves of DSSCs containing (a)
[Cu(SP)(mmt)]0/-, (b) [Cu(dmp)2]2+/+, and (c) [Cu(phen)2]2+/+ as the redox
couples; input power: 100 mW/cm2.
Figure 6. Relationship between light intensity and short circuit photocur-
rents of DSSCs containing (a) [Cu(SP)(mmt)]0/-, (b) [Cu(dmp)2]2+/+, and
(c) [Cu(phen)2]2+/+ as the redox couples.
Table 2. Performance Characteristics of DSSCs Using Various
Copper Complexes under White Light Illumination; Input Power:
100 mW/cm2
/-
+/+
+/+
redox couple
[Cu(SP)(mmt)]0
[Cu(dmp)2]2
[Cu(phen)2]2
Isc/mA/cm2
Voc/V
FF
4.4
3.2
0.48
0.57
0.43
0.12
0.66
0.44
1.3
0.79
0.55
1.4
η/%
Figure 5 shows the I-V characteristics of DSSCs using
copper complexes. The short circuit photocurrents (Isc), open-
circuit photovoltages (Voc), and fill factors (FF) are summarized
in Table 2 together with the light energy conversion efficiencies
(η), which are determined using eq 6. The Voc values of copper
complexes increase in the order: [Cu(phen)2]2+/+ (0.57 V),
[Cu(SP)(mmt)]0/- (0.66 V), and [Cu(dmp)2]2+/+ (0.79 V), in
agreement with the order of the redox potentials of copper
Figure 7. Relationship between light intensity and energy conversion
efficiencies of DSSCs containing (a) [Cu(SP)(mmt)]0/-, (b) [Cu(dmp)2]2+/+
and (c) [Cu(phen)2]2+/+ as the redox couples.
,
complexes. The η values of DSSCs using [Cu(phen)2]2+/+
,
(mmt)]0/-. Thus, the potential mismatch between the redox
couples of the sensitizers (S+/S) and the mediator is least for
[Cu(dmp)2]2+/+, and [Cu(SP)(mmt)]0/- were determined as
0.1%, 1.4%, and 1.3%, respectively.
the [Cu(dmp)2]2+/+
.
The difference in the photochemical properties of DSSCs
using different copper complexes may result from both the
thermodynamics and the kinetics of electron-transfer processes
of [Cu(SP)(mmt)]0/-, [Cu(dmp)2]2+/+, and [Cu(phen)2]2+/+ in
the photocurrent generation in DSSCs (vide infra). First, the
excited state of N719 dye injects an electron to the conduction
band of TiO2 to produce the dye cation. The resulting N719
dye cation is reduced by the CuI complex to form the CuII
complex, which is reduced at the Pt counter electrode, leading
to the photocurrent generation. The worst photovoltaic properties
with the [Cu(phen)2]2+/+ mediator (the lowest Isc and Voc)
certainly result from the slowest electron self-exchange rate of
[Cu(phen)2]2+/+ and the lowest redox potential (-0.10 V vs
SCE). Both the Isc and the Voc values are improved in DSSC
with [Cu(SP)(mmt)]0/ - because of the much faster electron self-
exchange rate of [Cu(SP)(mmt)]0/- and the higher redox
potential (0.29 V). In the case of [Cu(dmp)2]2+/+, the Isc value
is smaller than the case of [Cu(SP)(mmt)]0/- despite the faster
The photovoltaic measurements were also performed under
the irradiation of various power lights. Figure 6 shows some Isc
values for various input light powers. The Isc values of DSSCs
using [Cu(SP)(mmt)]0/- and [Cu(dmp)2]2+/+ increase as the input
light power increases. In contrast, the Isc value of DSSC using
[Cu(phen)2]2+/+ reaches a constant value under the same
experimental conditions. This indicates that the back electron
transfer from TiO2 to the dye cation becomes faster than the
electron transfer from [Cu(phen)2]+ to the dye cation at the high
light power region above 10 mW/cm2 because of the slow
electron transfer from [Cu(phen)2]+.
The dependence of the η values on the input light power is
shown in Figure 7. The maximum η value (2.3%) is attained
for DSSC using [Cu(dmp)2]2+/+ under light irradiation of 20
mW/cm2 intensity. The photoelectrochemical responses of
DSSCs with [Cu(SP)(mmt)]0/- and [Cu(dmp)2]2+/+ redox couples
are compared to the corresponding DSSC with the I3-/I- redox
couple. The IPCE values at 510 nm of DSSCs using [Cu(SP)-
(mmt)]0/-, [Cu(dmp)2]2+/+, and I3-/I- as redox couples are
obtained as 40%, 31%, 60%, respectively (see Supporting
Information S8). The light energy conversion efficiencies of
DSSCs using [Cu(SP)(mmt)]0/-, [Cu(dmp)2]2+/+, and I3-/I- as
redox couples are determined as 1.9%, 2.2%, and 4.3% under
electron self-exchange rate than that of [Cu(SP)(mmt)]0/-
,
probably because of the significantly smaller driving force of
electron transfer from [Cu(dmp)2]2+/+ to the dye cation as
compared to [Cu(SP)(mmt)]0/-. However, this drawback is
compensated by the higher Voc value due to the higher redox
potential (0.66 V) of [Cu(dmp)2]2+/+ than that of [Cu(SP)-
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9653

A R T I C L E S
Hattori et al.
cm2 intensity. Although the light energy conversion efficiency
of DSSCs using blue copper model complexes with a distorted
tetragonal geometry as the redox couples is still lower than that
of the conventional I3-/I- couple, the higher Voc value is attained
for the DSSC with [Cu(dmp)2]2+/+ for the first time as compared
to that using the I3-/I- couple. This is particularly important
because the dmp ligand is commercially available and
[Cu(dmp)2]2+/+ complexes are readily made in one step. The
use of copper model complexes with a distorted tetragonal
geometry, in which the structural change between the copper-
(I) and copper(II) complexes is minimized, provides a promising
strategy to develop efficient and low cost photovoltaic cells,
because nature has successfully utilized such a strategy to
develop efficient electron-transport systems using blue copper
protein as electron mediators.
Figure 8. I-V curves of DSSCs using (a) [Cu(SP)(mmt)]0/-, (b)
[Cu(dmp)2]2+/+, and (c) I3-/I- as the redox couples; input power: 20 mW/
cm2. I3-/I-: 0.2 M LiI, 0.05 M I2, 0.3 M 1,2-dimethyl-3-propylimidazolium
iodide, and 0.5 M tert-butylpyridine in methoxyacetonitrile.
Acknowledgment. This work was partially supported by a
Grant-in-Aid (No. 16205020) from the Ministry of Education,
Culture, Sports, Science, and Technology, Japan. We are grateful
to Dr. Takayuki Kitamura, Osaka University, for helping with
the preparation of DSSC.
weak illumination of simulated AM 1.5 solar light (20 mW/
cm2, from a Xe lamp), respectively (Figure 8).
Conclusions
Supporting Information Available: Cyclic voltammograms
of copper complexes (S1), the titration of [Cu(phen)2](CF3SO3)2
by excess [Fe(C5Me5)2] (S2), plot of pseudo-first-order rate
We have evaluated the electron self-exchange rate constants
of [Cu(SP)(mmt)]0/-, [Cu(dmp)2]2+/+, and [Cu(phen)2]2+/+ based
on electron-transfer reactions from a homologous series of
ferrocene derivatives to the copper(II) complexes in light of the
Marcus theory of electron transfer. The resulting electron self-
exchange rate constant (k22) increases in the order: [Cu-
(phen)2]2+/+ < [Cu(SP)(mmt)]0/- < [Cu(dmp)2]2+/+, in agree-
ment with the order of the smaller structural change between
the copper(II) and copper(I) complexes. We then have con-
structed the dye-sensitized solar cells (DSSC) using the copper
complexes as redox couples to compare the photoelectrochemi-
cal responses with those using the conventional I3-/ I- couple.
The maximum η value was attained as 2.2% for DSSC using
[Cu(dmp)2]2+/+ under the weak solar light irradiation of 20 mW/
n
constants k1 versus [Fe(C5H4 Bu)2] concentration for the reduc-
tion of [Cu(SP)(mmt)] (S3), plot of η value versus CuII/CuI
concentrations ratio of DSSC using [Cu(SP)(mmt)] (S4), I-V
curves of DSSCs using [Cu(SP)(mmt)] with and without
compact blocking layer (S5), I-V curves of DSSCs using [Cu-
(SP)(mmt)] with and without scattering layer (S6), plot of η
value versus transparent layer film thickness (S7), and photo-
current action spectra (IPCE values vs wavelength) with input
power: 20 mW/cm2 (S8). This material is available free of
charge via the Internet at http://pubs.acs.org.
JA0506814
9
9654 J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005