Increase in basin sediment yield from landslides in storms following major seismic disturbance

(1)

Solar Energy Materials & Solar Cells 87 (2005) 357–367

On the structural variations ofRu(II) complexes

for dye-sensitized solar cells

Ying-Chan Hsu

a

, Hegen Zheng

b

, Jiann T’suen Lin

c

,

Kuo-Chuan Ho

a,d,

a

Department of Chemical Engineering, National Taiwan University, Taipei 106, Taiwan

b

State Key Laboratory of Coordination Chemistry, Nanjing University, Nanjing 210093, PR China

c

Institute of Chemistry, Academia Sinica, Nankang, Taipei 115, Taiwan

d

Institute of Polymer Science and Engineering, National Taiwan University, Room 110A, 1, Sec. 4, Roosevelt Road, Taipei 10617, Taiwan

Received 15 May 2004; received in revised form 12 July 2004; accepted 15 July 2004 Available online 23 November 2004

Abstract

Ruthenium(II) complexes with new phenanthrenyl ligand (TAPNB) have been synthesized and examined. The spectroscopic and electrochemical measurements showed that the excited states ofthose complexes matched the conduction band oftitanium dioxide. The overall power conversion efﬁciencies of the solar cells utilized these new complexes as sensitizers for TiO2 ﬁlms were less than that ofN3-sensitized cell. Although the open-circuit voltage was similar to that ofN3-sensitized cell, the short-circuit current was one order lower. Such outcome may be attributed to the less amount ofdyes adsorbed due to the steric congestion ofthe complex. When NCS ligand was replaced by pyridyl ligand, the energy ofmetal-to-ligand charge transfer (Ru(II)-TAPNB) increased and resulted in blue shift of the absorption band. Anchoring ofcarboxylic acid at the surface ofTiO2slightly lowered the energy ofRu(II)-TAPNB charge transfer band. As carboxylic acid anchor was replaced by acetyl ester, the weaker interaction between the semiconductor and the ligand led to diminishing amount ofthe complex adsorbed and less photocurrent was detected.

Keywords: Dye-sensitized solar cell; Ruthenium dye; Phenanthrenyl ligand; Titanium dioxide

www.elsevier.com/locate/solmat

Corresponding author. Tel.: +886 2 2366 0739; fax: +886 2 2362 3040. E-mail address: kcho@ntu.edu.tw (K.-C. Ho).

(2)

1. Introduction

Dye sensitization oflarge band gap semiconductors has been investigated for many years[1–3]. An efﬁciency up to 10%[4,5]has been achieved on dye-sensitized solar cells (DSSCs) based on bis(bipyridyl) ruthenium complexes-coated TiO2.

Recently, several organic dyes possessing intense charge transfer character were found to be promising sensitizers for solar cells[6–8]. Such dipolar type compounds can be effectively tuned to absorb in the longer wavelength region via incorporation ofa methine unit in the conjugation spacer. However, oligomethine moiety is not particularly stable in general. Low-band gap conjugated oligomers have been demonstrated to be alternate for photovoltaic cells [9,10]. Among these, cis-Ru(NCS)2(dcbpy) (dcbpy=2,20-bipyridyl-4,40-dicarboxylate) and Ru(NCS)3(terpy)

(terpy=2,20_,200_{-terpyridyl-4,4}0_,400_{-tricarboxylate) appear to be most promising}

[11–14].

Ideal dyes should show increased absorption in the red region and retain a high photopotential and a quantitative incident monochromatic photo-to-current

(3)

conversion efﬁciency (IPCE) at shorter wavelength in the device. Approaches toward this aim include: (1) use ofligands with a lower p* level than 4,40_{-dicarboxy-2,2}0

-bipyridine; (2) raising the energy ofthe ruthenium t2g orbitals; (3) increasing the

metal-to-ligand charge transfer (MLCT) coefﬁcient[15,16]. Appropriate elongation ofthe conjugation length or incorporation ofelectron-withdrawing segments to terpyridine (or bipyridine) ligands is expected to lower the p* energy level ofthe terpyridines. Consequently, metal terpyridine complexes may have a lower MLCT energy. On the other hand, ruthenium ground state tuning may be approached by using a better electron-donating ligand such as benzimidazole [17]. Therefore, dedicate balance of electronic factors among different ligands is important.

In this study, we have synthesized new phenanthrenyl ligand, 4-(1H-1,3,7,8-tetraaza-cyclopenta[l]phenanthren-2-yl)-benzoic acid (abbreviated as TAPNB), possessing elongated conjugation for lower p* orbital energy level. Ruthenium complexes containing the new ligand were synthesized and subjected to photovoltaic studies. Scheme 1 shows the structural variations oftwo dyes: (1) Ru-1 (Ru(TAPNB)2(NCS)2) with two SCN ligands and two COOH anchors; (2) Ru-2

([Ru(TAPNB)(bipy)2][PF6]2) with two pyridine ligands and one COOH anchor. A

comparison with the ruthenium complex containing 4,40_{-dicarboxy-2,2}0_-bipyridine

will be discussed.

2. Experimental

2.1. Chemicals and substrate

All the chemicals were ACS reagent grade and not further puriﬁed before using. Fluoride-doped tin oxide (FTO) conducting glass substrates, serving as the optically transparent electrodes (OTEs), were obtained from a local supplier (Sinonar Corporation, Hsinchu, Taiwan), and the sheet resistance was ca. 30 O/sq.

2.2. Preparation and characterization of the Ru complexes 2.2.1. Phenanthrenyl ligands

TAPNB and 4-(1H-1,3,7,8-tetraaza-cyclopenta[l]phenanthren-2-yl)- benzoic acid methyl ester (abbreviated as TAPNBE) were prepared by a similar procedure. Only the synthesis ofTAPNB will be described in detail.

TAPNB: A mixture ofaldehyde (3.5 mmol), 1,10-phenanthroline-5,6-dione (2.5 mmol), ammonium acetate (50 mmol) and glacial acetic acid (15 ml) was reﬂuxed for about 6 h, then cooled to room temperature and diluted with water (ca. 25 ml). Yellow precipitates were collected and washed with water. The crude products were puriﬁed by recrystillization with methanol to produce TAPNB as yellow–white solids (720 mg, 85%). The proton NMR spectrum was recorded in d6

-DMSO on a 300 MHz NMR spectrometer (Bruker AC300). 1H NMR (d, ppm, TMS): 7.84–7.94 (m, 2H, phen), 8.17 (d, J=8.44, 2H, C6H4), 8.40 (d, J=8.44, 2H,

(4)

Mass spectra (EI) were recorded on a VG70-250S mass spectrometer. FAB MS (m/e): 341.1 (M+H).

TAPNBE: Yellow solid. Yield=73%.1H NMR (d6-DMSO): d 3.9 (s, 3H, CH3),

7.81–7.84 (m, 2H, phen), 8.19 (d, J=8.18, 2H, C6H4), 8.43 (d, J=8.15, 2H, C6H4),

8.92–8.95 (m, 2H, phen), 9.03–9.05(m, 2H, phen). FAB MS (m/e): 355.1 (M+H). 2.2.2. Preparation of (Ru(TAPNB)2(NCS)2) (Ru-1)

RuCl32H2O (262 mg, 1.0 mmol), TAPNB (681 mg, 2.0 mmol) and LiCl (1/15

mmol, 2.0 mg) were added to dimethylformamide (DMF) (60 ml). The mixture was reﬂuxed for 3 h, then cooled to room temperature, and DMF was evaporated in N2.

The residue was added acetone and solid formed was collected and dried for next step reaction. TAPNB2RuCl2(768 mg, 0.9 mmol) and KNCS (1.75 g, 18 mmol) were

added to DMF (80 ml), and the solution was reﬂuxed for 6 h. The solvent DMF was evaporated under vacuum, and the resulting solid was collected in a sintered glass, washed with methanol and ether. Brownish product was obtained in 65% yield after puriﬁcation with column chromatography (DMF:EtOH=1:20 as eluent).1H NMR (d, ppm, TMS): 7.88–7.92 (m, 2H, phen), 8.15 (d, J=8.40, 2H, C6H4), 8.42 (d,

J=8.40, 2H, C6H4), 8.95–8.99 (m, phen, 2H), 9.06–9.10 (m, 2H, phen), 14.22 (s, 1H,

COOH). FAB MS (m/e): 898.1. Anal. Calc. for C42H24N10O4RuS2: C, 56.18; H, 2.69;

N, 15.60. Found: C, 57.07; H, 2.95; N, 15.22.

2.2.3. Preparation of (Ru(TAPNBE)2(NCS)2) (Ru-12)

The complex Ru-12 was synthesized by a similar procedure as described for Ru-1 except that TAPNBE was used instead ofTAPNB. Brownish product was obtained in 60% yield after puriﬁcation with column chromatography (DMF:EtOH=1:20 as eluent).1H NMR (d6-DMSO): d 3.9 (s, 3H, CH3), 7.68–7.76 (m, 2H, phen), 8.11 (d,

J=8.18, 2H, C6H4), 8.35 (d, J=8.15, 2H, C6H4), 8.93–9.02 (m, 2H, phen), 9.24–9.34

(m, 2H, phen). FAB MS (m/e): 926.1. Anal. Calc. for C44H28N10O4RuS2: C, 57.07;

H, 3.05; N, 15.13. Found: C, 57.87; H, 3.45; N, 14.75. 2.2.4. Preparation of [Ru(TAPNB)(bipy)2][PF6]2(Ru-2)

To a stirred, deaerated solution ofligands TAPNB (0.5 mmol) in ethanol (20 ml) was added [Ru(bpy)2Cl2] 2H2O (260 mg, 0.5 mmol). The reaction mixture was

heated under reﬂux for 6 h in an argon atmosphere. After cooling to room temperature, an aqueous solution ofNH4PF6was added until no further precipitate

was formed. After the suspension was stored for 2 h at 0 1C, the precipitate was ﬁltered, washed successively with H2O (10 ml), ethanol (10 ml) and diethyl ether

(20 ml) to furnish the metal complexes as analytically pure orange to red solids in 78% yield (390 mg).1H NMR (d6-DMSO): d 7.34 (t, J=6.7, 2H, bpy), 7.57–7.62 (m,

4H, bpy), 7.84 (d, J=5.4, 2H, bpy), 7.95 (m, 2H, phen), 8.07–8.12 (m, 4H, bpy), 8.19–8.23 (m, 4H, bpy, C6H4), 8.46 (d, J=8.2, 2H, C6H4), 8.83–8.89 (m, 4H, phen,

bpy), 9.1–9.2 (m, 2H, phen), 14.8 (s, 1H, COOH). FAB MS (m/e): 753.3 (M-2PF6).

Elemental analyses were performed on a Perkin-Elmer 2400 CHN analyzer. Anal. Calc. for C40H28F12N8O2P2Ru: C, 46.03; H, 2.70; N, 10.74. Found: C, 46.08; H,

(5)

2.2.5. Preparation of [Ru(TAPNBE)(bipy)2][PF6]2(Ru-22)

The complex Ru-22 was synthesized by a similar procedure as described for Ru-2 except that TAPNBE was used instead ofTAPNB. Red solid. Yield=70%. 1H NMR (d6-DMSO): d 3.92 (s, 3H, CH3), 7.32–7.36 (m, 2H, bpy), 7.57–7.63 (m, 4H,

bpy), 7.83–7.84 (m, 2H, bpy), 7.90–7.98 (m, 2H, phen), 8.07–8.12 (m, 4H, bpy), 8.20–8.24 (m, 4H, bpy, C6H4), 8.52 (d, J=8.13, 2H, C6H4), 8.83–8.89 (m, 4H, phen,

bpy), 9.10–9.28 (m, 2H, phen). FAB MS (m/e): 912.9 (M-PF6). Anal. Calc. for

C41H30F12N8O2P2Ru: C, 46.56; H, 2.86; N, 10.59. Found: C, 46.12; H, 3.36; N,

10.27.

2.2.6. Spectroscopic and electrochemical studies

Electronic absorption spectra were obtained on a Perkin-Elmer Lambda 9 spectrometer. Emission spectra were recorded by a Hitachi F-4500 fluorescence spectrometer. The details on the experimental set-up for fluorescence lifetime measurements have been described in the previous report[18]. Cyclic voltammetry experiments were performed with a BAS-100 electrochemical analyzer. All measurements were carried out at room temperature with a conventional three-electrode configuration consisting ofplatinum working and auxiliary three-electrodes and a nonaqueous Ag/AgNO3reference electrode. All potentials are reported relative to

Ag/AgNO3and are not corrected for the junction potential. Fc+/Fc was measured

to be 0.22 V relative to Ag/AgNO3. The details on the experimental set-up for

electrochemical measurements have been described in the previous report[19]. 2.3. Preparation of the TiO2thin film and dye adsorption

The TiO2thin ﬁlm, serving as the photoanode in this work, was prepared through

the general sol–gel method. The precursor solution was made according to the following procedure: 135 ml of 0.1 M nitric acid solution under vigorous stir was dropped with 22.5 ml pure Ti(C3H7O)4 slowly to form a mixture. After the

hydrolysis, the mixture was heated at 8575 1C in a water bath and stirred vigorously for 12 h in order to achieve the peptization. When the mixture cooled down to room temperature, it was ultrasonically vibrated for 10 min, and then 30 wt% of polyethylene glycol (molecular weight of20 000) was added in a proportion ofthe TiO2weight. The precursor solution with an equivalent TiO2concentration ofca.

4 wt% was thus obtained and ready for the subsequent dip-coating process. During the dip-coating operation, a cleaned FTO glass substrate with a dimension of4 2 cm2was halfdipped into the solution. After 10 min, the substrate was slowly drawn out with a speed of6 cm/min. The dipping-drawing out procedure was performed ten times totally, and the substrate was maintained as vertically as possible. After exposing to air for 30 min, the substrate coated with the precursor solution was dried at 50 1C for 15 min. Then it was heated to 450 1C at a rate of 20 1C min1and then sintered at 450 1C for 30 min to form the TiO2thin ﬁlm. The

ﬁlm was cooled down naturally to room temperature. The thickness ofthe TiO2ﬁlm

was estimated as ca. 5 mm from a side-view scanning electron microscopic (SEM, Hitachi, model S800) image and a proﬁlometer (Sloan technology, Dektak 3030).

(6)

The dye adsorption on the sintered TiO2thin ﬁlm was performed as follows: After

heating the TiO2thin ﬁlm to 80 1C, the ﬁlm was taken out from the oven and dipped

into the solution containing 3 104M dye sensitizers. The solvents used were dimethylformamide for Ru-1 and Ru-12, acetonitrile for Ru-2 and Ru-22, and ethanol for cis-di(thiocyanato)bis(2,20_{-bipyridyl-4,4}0_{-dicarboxylate)ruthenium(II)}

(N3, Solaronix) for at least 12 h, respectively. 2.4. Assembly and characterization of the DSSCs

The dye-sensitized photoanode was rinsed with acetonitrile and dry. A platinized FTO with 1 mm-thick Pt by sputtering was used as a counter electrode and was controlled an active area of1 1 cm2by adhered polyester tape (3 M) with thickness of60 mm. The photoanode was placed on top ofthe counter electrode and tightly clipping them together to form a cell. Electrolyte was then injected into the cell space through one ofthe two open holes present in the tape and then sealing the cell with the Torr Sealscement (Varian, MA, USA). The electrolyte was composed of0.5 M tetrabutylammonium iodide (TBAI), 0.02 M lithium iodide (LiI), 0.05 M iodine (I2),

and 0.5 M 4-tert-butylpyridine that was dissolved in acetonitrile.

The photoelectrochemical characterizations on the solar cells were carried out using an AM 1.5 simulated light radiation. The light source was emitted from a 450 W Xe lamp (Oriel, #6266) equipped with a water-based IR filter and AM 1.5 filter (Oriel, #81075). Light intensity attenuated by neutral density filter (Optosigma, #078-0360) at the measuring (cell) position was estimated to be ca. 10 mW cm2 upon the reading from a radiant power meter (Oriel, #70260) connected to a thermopile probe (Oriel, #70263). Photoelectrochemical characteristics ofthe DSSCs, including their on–off responses of open-circuit voltages and photocur-rent–voltage curves, were recorded through the potentiostat/galvanostat.

3. Results and discussion

3.1. Spectroscopic and electrochemical properties of Ru complexes

Figs. 1, 2 and 3show the normalized absorption spectra ofRu-1, Ru-2 and N3 on TiO2 ﬁlms and in organic solutions, respectively. The spectral difference between

adsorbed complex on TiO2ﬁlm and dissolved molecule in solution is resulted from

p* bond formed by Ti 3d orbital and p* orbital ofligand electronic coupling. A larger red shift was observed in Ru-2 containing bipyridine. The same phenomenon was reported and was suggested to be due to increasing delocalization ofelectrons from Ru complex containing pyridyl ligand to TiO2 [20]. Absorption and

electrochemical data ofRu complexes were listed in Table 1. From the oxidation potential ofRu2+, Eox0

0

, the highest occupied molecular orbital (HOMO) ofeach complex was calculated. The presence ofelectron-withdrawing bipyridine likely lowers the electron density ofthe ruthenium center which will lower the energy ofRu t2g orbital and increase the transition energy from the Ru center to TAPNB.

(7)

Inefﬁcient energy conversion of solar cells (vide infra) using Ru-1 and Ru-2 may be due to absence ofabsorbance beyond 600 nm. The absorption spectra and oxidation potentials ofthe Ru-12 and Ru-22 with phenyl acetate anchor are almost the same as Ru-1 and Ru-2, respectively.

3.2. Working principle of the TiO2DSSCs

According to the spectroscopic and electrochemical measurements, the energy level ofeach Ru complex can be calculated and was shown in Fig. 4.

400 500 600 700 800 900 Wavelength / nm 0 0.2 0.4 0.6 0.8 1 Normalized absorbance Ru-2 TiO₂/Ru-2

Fig. 2. Normalized absorption spectra ofRu-2 on TiO2ﬁlm (solid line) and in acetonitrile solution (dash

line). 400 500 600 700 800 900 Wavelength / nm 0 0.2 0.4 0.6 0.8 1 Normalized absorbance Ru-1 TiO₂/Ru-1

(8)

HOMO was calculated from Eox0 0

in Table 1; while lowest unoccupied mole-cular orbital (LUMO) was obtained from the subtraction Eox0

0

by energy gap estimated from electronic absorption edge. The transition energy of Ru-1 or Ru-2 was shifted to higher energy than N3, indicating that TAPNB may be less efﬁcient electron acceptor compared to the bipyridyl ligand in N3. The energy levels ofRu-1 are very similar to those ofRu-11, and so are Ru-2–Ru-22. It can be concluded here that elongation ofthe conjugation length

400 500 600 700 800 900 Wavelength / nm 0 0.2 0.4 0.6 0.8 1 Normalized absorbance N3 TiO₂/N3

Fig. 3. Normalized absorption spectra ofN3 on TiO2ﬁlm (solid line) and in ethanol solution (dash line).

Table 1

Absorption and electrochemical data ofRu complexes Absorption

Complex lmax/nm (e/104M1cm1) Eox0 0_c (V) Ered0 0 (V) Ru-1 507 (0.55)a _1.20 _1.02 Ru-12 507 (1.12)a _1.25 _1.03 Ru-2 460 (0.64)b _1.40 _1.29 Ru-22 460 (1.60)b _1.33 _1.30 N3d _{534 (1.42)} _1.09 _— a Measured in DMF solution. b

Measured in acetonitrile solution.

c

All E00data are reported relative to NHE, which were calibrated from ferrocene (0.55 V vs. NHE) in CH2Cl2solution, The concentration ofthe complexes used in this experiment was 103M containing

0.1 M TBAP (tetrabutylammonium hexaﬂuorophosphate) in CH2Cl2solution at 25 and the scan rate was

100 mV s1_.

(9)

or pyridine ligand without electron-withdrawing would shift the ground state downward.

3.3. Photocurrent–voltage characteristics

Photocurrent–voltage (I–V) characteristics ofthe DSSCs fabricated with Ru-1 and Ru-2 dyes were measured under illumination by a simulated AM 1.5 solar light. The I–V curves ofDSSCs with Ru-1 and Ru-2 complexes were shown in Fig. 5. The values ofopen-circuit voltage were similar, however, the short-circuit current ofRu-2 was twice as large as that ofRu-1. The low current density ofRu-1 was mainly caused by the small amount ofdye adsorbed, measured and presented as DA in

Table 2. This may be due to the greater steric hindrance ofRu-1. The overall conversion efﬁciencies for DSSCs made with Ru-1, Ru-2, and N3 were 0.25%, 0.79%, and 4.54%, respectively. The low efﬁciency and current density were partially caused by the small amount ofdyes adsorbed, only ca. 1/4 (estimated from DA/emax)

for Ru-1 when compared with that of N3. Another reason may be due to lack of absorption beyond l4600 nm. Energy gaps ofboth dyes are large, 1.9 eV for Ru-1 and 2.2 eV for Ru-2, as compared with N3 of 1.7 eV. It is inefﬁcient in harvesting solar energy. Direct bonding between the titanium ion and the complex is not possible when benzoic acid anchor was replaced by phenyl acetate anchor (Ru-12 and Ru-22). Instead, the interaction between the complex and TiO2may involve the

hydrogen bond with surface hydroxyl group of TiO2 or Van der Waals force

adsorption. The weaker interaction will result in less amount ofRu-12 (or Ru-22) adsorbed by TiO2 when compared to Ru-1 (or Ru-2). Less effective electronic

coupling between Ti 3d orbital and p* orbital on ligand will also lead to smaller photocurrents in Ru-12 (or Ru-22) than that ofRu-1 (or Ru-2).

E (V vs. NHE) -1 0 1 TiO₂ CB HOMO LUMO N3 Ru-1 Ru-2 I-_{/ I} 3

-Fig. 4. Schematic illustration ofthe energy band diagram ofRu dyes, TiO2, and redox couple ofiodide/tri

(10)

4. Conclusions

Two Ru complexes with phenanthrenyl ligand have been synthesized and their spectroscopic and electrochemical properties were studied. The compound with thiocyanate ancillary ligand absorbes at longer wavelength than that with bipyridine ancillary ligand. The larger steric congestion of the former results in less efﬁcient adsorption by TiO2, however. The lower efﬁciency of the photovoltaic devices than

that ofN3 device is attributed to the absence ofthe absorption at longer wavelength and poorer adsorption by TiO2 in the former. When benzoic acid anchor was

replaced by phenyl acetate anchor, the linkage between Ti 3d orbital and p* orbital on ligand was even less effective, and smaller photocurrent was detected. Further modiﬁcation ofthe phenanthrenyl ligand is in progress.

Table 2

Performance parameters of TiO2solar cells sensitized with Ru complexes

Complex Voc(V) Isc(mA cm2) Za(%) FF IPCEb(%) DAc

Ru-1 0.57 0.09 0.25 0.49 11 0.29

Ru-12 0.55 0.07 0.17 0.46 — 0.28

Ru-2 0.58 0.22 0.79 0.62 18 1.92

Ru-22 0.56 0.10 0.30 0.54 — 0.85

N3 0.58 1.26 4.54 0.62 40 1.42

a_{Efﬁciency ofthe test solar cells with simulated AM 1.5 solar light of10 mW cm}2_. b_{IPCE=% ¼}1240

l=nm

Isc=mA cm2

PLight=mW cm2; where l=500 nm and PLlight=2.8 mW cm 2_. c_{Absorbance difference of a TiO}

2photoanode before and after adsorbed the complex at the maximum

wavelength. 0 0.2 0.4 0.6 0.8 Cell voltage / V -0.1 0 0.1 0.2 Current density / mA cm -2 Ru-1 Ru-2 X η = 0.25 % η = 0.79 % X

(11)

Acknowledgements

The authors want to thank Pi-Tai Chou ofthe Department ofChemistry at the National Taiwan University for spectroscopic measurements. This work was ﬁnancially supported through a grant from the CTCI Foundation, Taipei, Taiwan. Useful discussions with Dr. Jin-Guu Wang and Dr. Teh-Ming Tao, of the CTCI Foundation, are deeply appreciated.

References

[1] H. Meier, J. Phys. Chem. 69 (1965) 719.

[2] H. Tsubomura, M. Matsumura, Y. Nomura, T. Amamiya, Nature 261 (1976) 402. [3] B. O’Regan, M. Gra¨tzel, Nature 353 (1991) 737.

[4] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mu¨ller, P. Liska, N. Vlachopoulos, M. Gra¨tzel, J. Am. Chem. Soc. 115 (1993) 6382.

[5] C.J. Barbe´, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann, V. Shklover, M. Gra¨tzel, J. Am. Ceram. Soc. 80 (1997) 3157.

[6] Z.-S. Wang, F.-Y. Li, C.-H. Huang, Chem. Commun. (2000) 2063.

[7] K. Hara, M. Kurashige, S. Ito, A. Shinpo, S. Suga, K. Sayama, H. Arakawa, Chem. Commun. (2003) 252.

[8] K. Hara, M. Kurashige, Y. Dan-oh, C. Kasada, A. Shinpo, S. Suga, K. Sayama, H. Arakawa, New J. Chem. 27 (2003) 783.

[9] A. Dhanabalan, K.J. van Duren, P.A. van Hal, J.L.J. Dongen, R.A.J. Janssen, Adv. Funct. Mater. 11 (2001) 255.

[10] C.J. Brabec, C. Winder, N.S. Sacriciftic, J.C. Hummeden, A. Dhanabalan, P.A. van Hal, R.A.J. Janssen, Adv. Funct. Mater. 12 (2002) 709.

[11] A. Hagfeldt, M. Gra¨tzel, Acc. Chem. Res. 33 (2000) 269.

[12] Md.K. Nazeeruddin, P. Pe´chy, T. Renouard, S.M. Zakeeruddin, R. Humphry-Baker, P. Comte, P. Liska, L. Cevey, E. Costa, V. Shklover, L. Spiccia, G.B. Deacon, C.A. Bignozzi, M. Gra¨tzel, J. Am. Chem. Soc. 123 (2001) 1613.

[13] Z.-S. Wang, C.-H. Huang, Y.-Y. Huang, B.-W. Zhang, P.-H. Xie, Y.-J. Hou, K. Ibrahim, H.-J. Qian, F.-Q. Liu, Sol. Energy Mater. Sol. Cells 71 (2002) 261.

[14] Md.K. Nazeeruddin, R. Humphry-Baker, P. Liska, M. Gra¨tzel, J. Phys. Chem. B 107 (2003) 8981. [15] S. Ruile, O. Kohle, C. Klemenz, P. Pe´chy, M. Gra¨tzel, Inorg. Chim. Acta 261 (1997) 129. [16] A. Islam, H. Sugohara, H. Arakawa, J. Photochem. Photobiol. A: Chem 158 (2003) 131.

[17] M.K. Nazeeruddin, E. Mu¨ller, R. Humphry-Baker, N. Vlachopoulos, M. Gra¨tzel, J. Chem. Soc. Dalton Trans. (1997) 4571.

[18] P.-T. Chou, Y.-C. Chen, W.-S. Yu, Y.-H. Chou, C.-Y. Wei, Y.-M. Cheng, J. Phys. Chem. A 105 (2001) 1731.

[19] K.R.J. Thomas, J.-T. Lin, H.-M. Lin, C.-P. Chang, C.-H. Chuen, Organometallics 20 (2001) 557. [20] D.A. Gulino, H.G. Drickamer, J. Phys. Chem. 88 (1984) 1173.