Optimizing the photovoltaic performance of thiocyanate-free ruthenium photosensitizers by structural modification of C^N cyclometalating ligand in dye-sensitized solar cells

Article abstract of DOI:10.1016/j.poly.2014.04.059

Five new thiocyanate-free ruthenium(II) complexes with different electron-donating functionalized cyclometalating ligands C^N were synthesized, characterized and applied as photosensitizers in dye-sensitized solar cells (DSSCs). Their photophysical, electrochemical, thermal and photovoltaic properties have been investigated and density functional theory (DFT) calculations have been carried out on these dyes. These dyes exhibit good thermal stability with the onset decomposition temperature at 5% weight-loss (T_d) of around 330 °C. The DSSC device using the Ru(II) dye with the 9-tolylcarbazole chromophore exhibited the highest power conversion efficiency (η) up to 3.39%, with a short-circuit photocurrent density (J_sc) of 8.06 mA cm^-2, an open-circuit photovoltage (V_oc) of 0.62 V and a high fill factor (ff) of 0.68 under illumination of an AM 1.5 solar cell simulator.

Full text of DOI:10.1016/j.poly.2014.04.059

Accepted Manuscript
Optimizing the Photovoltaic Performance of Thiocyanate-free Ruthenium Pho-
tosensitizers by Structural Modification of C^N Cyclometalating Ligand in Dye-
Sensitized Solar Cells
Chi-Ho Siu, Cheuk-Lam Ho, Jian He, Tao Chen, Poulomi Majumda, Jianzhang
Zhao, Hua Li, Wai-Yeung Wong
PII:
S0277-5387(14)00286-1
DOI:
http://dx.doi.org/10.1016/j.poly.2014.04.059
POLY 10703
Reference:
Polyhedron
To appear in:
Received Date:
Accepted Date:
25 March 2014
29 April 2014
Please cite this article as: C-H. Siu, C-L. Ho, J. He, T. Chen, P. Majumda, J. Zhao, H. Li, W-Y. Wong, Optimizing
the Photovoltaic Performance of Thiocyanate-free Ruthenium Photosensitizers by Structural Modification of C^N
Cyclometalating Ligand in Dye-Sensitized Solar Cells, Polyhedron (2014), doi: http://dx.doi.org/10.1016/j.poly.
2014.04.059
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Polyhedron (Special Issue on Molecular Materials for Solar Energy Conversion)
Optimizing the Photovoltaic Performance of Thiocyanate-free Ruthenium
Photosensitizers by Structural Modification of C^N Cyclometalating Ligand in
Dye-Sensitized Solar Cells
Chi-Ho Siu,a Cheuk-Lam Ho,a,b* Jian He,c Tao Chen,c,* Poulomi Majumda,d
Jianzhang Zhao,d Hua Li,a,e Wai-Yeung Wong,a,b*
a
Department of Chemistry and Institute of Advanced Materials, Hong Kong Baptist University,
Waterloo Road, Kowloon Tong, Hong Kong, PR China
b
HKBU Institute of Research and Continuing Education, Shenzhen Virtual University Park,
Shenzhen, 518057, PR China
c Department of Physics, The Chinese University of Hong Kong, Shatin, Hong Kong, PR China
d State Key Laboratory of Fine Chemicals, School of Chemical Engineering,158 Zhongshan Road,
Dalian University of Technology, Dalian 116012, PR China
e College of Life and Environmental Sciences, Minzu University of China, Beijing, 100081, PR
China
_______________
*
Corresponding authors. Tel.: +852 3411 7074 (W.-Y. Wong); fax: +852 3411 7348 (C.-L. Ho,
W.-Y. Wong), Tel.: +852 3943 6278; fax: +852 2603 5204 (T. Chen).
E-mail addresses: clamho@hkbu.edu.hk (C.-L. Ho), rwywong@hkbu.edu.hk (W.-Y. Wong),
taochen@phy.cuhk.edu.hk (T. Chen).
1

Abstract
Five new thiocyanate-free ruthenium(II) complexes with different electron-donating
functionalized cyclometalating ligands C^N were synthesized, characterized and
applied as photosensitizers in dye-sensitized solar cells (DSSCs). Their photophysical,
electrochemical, thermal and photovoltaic properties have been investigated and
density functional theory (DFT) calculations have been carried out on these dyes.
These dyes exhibit good thermal stability with the onset decomposition temperature at
5% weight-loss (Td) of around 330 oC. For the DSSC device using the Ru(II) dye with
the 9-tolylcarbazole chromophore exhibited the highest power conversion efficiency
(η) up to 3.39%, with a short-circuit photocurrent density (Jsc) of 8.06 mA cm–2, an
open-circuit photovoltage (Voc) of 0.62 V and a high fill factor (ff) of 0.68 under
illumination of an AM 1.5 solar cell simulator.
2

1. Introduction
Because of the ascending demand for energy and the limited amount of fossil fuel
available in our Earth, solar energy is expected to play a strategic role to deal with the
energy crisis as it is abundant, unlimited in supply and clean [1−9]. Sunlight can be
harvested and directly converted into electricity by using photovoltaic devices [3−13].
Dye-sensitized solar cell (DSSCs) have drawn tremendous research attention due to
the possibility that mechanically flexible and aesthetically pleasing modules can be
developed and deployed on a large scale for relatively low overall cost [14-23].
Ruthenium(II) complexes incorporated with thiocyanate ligands such as N3 [24],
N719 [25,26], N907 [27−29] and black dye (N749) [30,31] have maintained a
significant lead in the photovoltaic performance of DSSCs over the past two decades.
However, the thiocyanate-containing ruthenium(II) complexes are facing stability
problems since the NCS ligand would deteriorate the lifespan of the cell, making the
photosensitizers unstable under thermal stress and light soaking that may hinder the
commercialization of DSSCs [32,33]. Recently, Grätzel and co-workers reported an
impressive finding on thiocyanate-free cyclometalated Ru(II) photosensitizers which
showed excellent power conversion efficiency and stability for practical application
[32]. This kind of Ru(II) complexes possesses lower molecular symmetry and results
in lowering the degeneracy of the energy levels in order to produce broader absorption
3

coverage as compared to the standard dye N719 [32]. This discovery provides a good
opportunity to develop highly efficient dyes with long term stability for
commercializing the DSSCs [34−38].
Optimizing the energy levels of the photosensitizers is also a very important factor
for high photovoltaic performance of DSSCs because the energy levels of the
semiconductor anode and the redox electrolyte need to show a good match as they
would affect the electron injection and dye regeneration [36]. By varying the
cyclometalating ligand (C^N) on the Ru(II) complexes, it can have a profound impact
on the photovoltaic performance of DSSCs because the structure of the C^N ligand
has a close relationship with the highest occupied molecular orbital (HOMO) level
[38]. Therefore, modifying the structure of C^N ligand provides an avenue to
modulate the HOMO level and the efficiency of DSSCs. To date, most of the research
efforts along this line are to use 2-phenylpyridine derivatives as the cyclometalating
ligands for the ruthenium core [39–44], while those containing strongly
electron-donating
moieties
such
as
carbazole,
triphenylamine
and
diphenylaminofluorene in place of phenyl ring have seldomly been investigated in
such Ru(II) complexes.
The aim of this study is to design and synthesize a series of thiocyanate-free
cyclometalated Ru(II) complexes (D1−D5), [Ru(dcbpy)2(LH)]+PF6 (LH
=
4

3-(5-(trifluoromethyl)pyridin-2-yl)-9-p-tolyl-9H-carbazole
2-(5-(trifluoromethyl)pyridin-2-yl)-9-phenyl-9H-carbazole
9-butyl-3-(5-(trifluoromethyl)pyridin-2-yl)-9H-carbazole
N-(4-(5-(trifluoromethyl)pyridin-2-yl)phenyl)-N-phenylbenzenamine
(D1),
(D2),
(D3),
(D4),
9,9-diethyl-7-(5-(trifluoromethyl)pyridin-2-yl)-N,N-diphenyl-9H-fluorenyl-2-amine
(D5)), for DSSC application (Figure 1). Their photophysical properties,
electrochemistry, computational analyses on frontier orbitals, thermal properties and
photovoltaic performance in DSSCs will be investigated. The structure-photovoltaic
performance relationship will also be studied.
2. Experimental Section
All the reactions were carried out under a nitrogen atmosphere. Solvents were
dried by standard methods and distilled prior to use. Commercially available reagents
were used without further purification unless otherwise stated. All reagents for the
chemical syntheses were purchased from Aldrich or Acros Organics. All reactions
were monitored by thin-layer chromatography (TLC) with Merck pre-coated glass
plates. Column chromatography was carried out using silica gel (230−400 mesh) or
basic aluminium oxide (63−200 mesh) from Merck. All the mass spectra were
recorded on an Autoflex Bruker in matrix-assisted laser desorption ionization time of
flight (MALDI-TOF) system. NMR spectra were measured in CDCl3, CD3CN or
5

CD3OD on Bruker AM 400MHz FT-NMR spectrometer and chemical shifts were
quoted relative to Me4Si standard for 1H and 13C nuclei. Electronic absorption spectra
were obtained with a Hewlett Packard 8453 spectrometer. Cyclic voltammograms
were obtained with CH Instrument CV. Thermal gravimetric analyses were performed
with a Perkin-Elmer TGA6 thermal analyzer. L1 [45a], 2c [45b], 2d [45c], 3c [38], 4b
[38] and 5a/5b [45f] were prepared as previously reported.
2.1 Synthesis of 2-(5-(trifluoromethyl)pyridin-2-yl)-9-phenyl-9H-carbazole (L2)
Pd(PPh3)4 (40 mg) was added to a mixture of 2d (230 mg, 0.801 mmol) and
2-chloro-5-trifluoromethylpyridine (772 mg, 0.667 mmol) in toluene (40 mL) and 2
M aqueous Na2CO3 (4 mL) under an inert atmosphere of nitrogen. The reaction was
heated at 110 oC for 48 hr. After the mixture was cooled to room temperature, water
was added and the mixture was extracted with ethyl acetate. The combined organic
layer was dried over MgSO4, filtered and evaporated under reduced pressure. The
residue was purified by preparative TLC plates on silica by gradient elution using a
mixture of hexane/CH2Cl2 (2:1, v/v) to afford a white solid of L2 in moderate yield
(750 mg, 67 %). 1H NMR (400 MHz, CDCl3): δ = 8.91 (m, 1H, Ar), 8.21–8.10 (m, 3H,
Ar), 7.92–7.82 (m, 3H, Ar), 7.63–7.56 (m, 4H, Ar), 7.49–7.38 (m, 3H, Ar), 7.31–7.27
13
(m, 1H, Ar) ppm. C NMR (400 MHz, CDCl3): δ = 161.24, 146.55, 141.97, 141.37,
6

137.38, 135.80, 133.83, 130.08, 127.79, 127.30, 126.73, 125.21, 124.89, 124.60,
122.80, 122.51, 120.72, 120.27, 110.21, 110.03, 108.86 (Ar), 123.95 (CF3) ppm.
MALDI-TOF: m/z 388.32 [M]+. Calculated [M]+: 388.12.
2.2 Synthesis of 9-butyl-3-(5-(trifluoromethyl)pyridin-2-yl)-9H-carbazole (L3)
A
mixture of 2-chloro-5-trifluoromethylpyridine (0.905 g, 4.99 mmol),
9-butyl-9H-carbazol-3-yl-3-boronic acid (2.0 g, 7.48 mmol), and 2 M K2CO3 (10 mL)
in tetrahydrofuran (50 mL) were degassed with N2 for 5 min. Pd(PPh3)4 (100 mg) was
added and the resulting mixture was heated to reflux for two days. After being cooled
to room temperature, water was added and the mixture was extracted with ethyl
acetate. The combined organic fractions were dried over Na2SO4, filtered and
concentrated under reduced pressure. The crude compound was purified by silica gel
column chromatography using CH2Cl2/n-hexane (1:1, v/v) as eluent to afford L3 as a
white solid (1.02 g, 55%). 1H NMR (400 MHz, CDCl3): δ = 8.96 (m, 1H, Ar), 8.82 (d,
J = 1.4 Hz, 1H, Ar), 8.20–8.16 (m, 2H, Ar), 7.97 (m, 2H, Ar), 7.52–7.48 (m, 2H, Ar),
7.45 (d, J = 8.1 Hz, 1H, Ar), 7.30–7.26 (m, 1H, Ar), 4.37 (t, J = 7.1 Hz, 2H, alkyl),
1.93–1.85 (m, 2H, alkyl), 1.45–1.40 (m, 2H, alkyl), 0.90–0.86 (m, 3H, alkyl) ppm. 13C
NMR (400 MHz, CDCl3): δ = 161.47, 146,52, 146,48, 141,66, 141.03, 133.80, 133.77,
128.77, 126.15, 125.38, 125.00, 123.42, 123.07, 120.66, 119.67, 119.44, 109.07 (Ar),
7

123.52 (CF3), 43.05, 31.16, 20.60, 13.93 (alkyl) ppm. MALDI-TOF: m/z 368.16 [M]+.
Calculated [M]+: 368.15.
2.2 Synthesis of 4-(5-(trifluoromethyl)pyridin-2-yl)-N,N-diphenylbenzenamine (L4)
A
mixture of 2-chloro-5-trifluoromethylpyridine (0.919 g, 5.07 mmol),
4-(diphenylamino)phenylboronic acid 4b (2.2 g, 7.61 mmol) and 2 M K2CO3 (10 mL)
in tetrahydrofuran (50 mL) were degassed with N2 for 5 min. Pd(PPh3)4 (100 mg) was
added and the resulting mixture was heated to reflux for two days. After being cooled
to room temperature, water was added and the mixture was extracted with ethyl
acetate. The combined organic fractions were dried over Na2SO4, filtered and
concentrated under reduced pressure. The crude compound was purified by silica gel
column chromatography using CH2Cl2/n-hexane (1:1, v/v) as eluent to afford L4 as a
1
yellow solid (1.30 g, 66%). H NMR (400 MHz, CDCl3): δ = 8.88 (s, 1H, Ar),
7.93–7.89 (m, 3H, Ar), 7.77 (d, J = 8.4 Hz, 1H, Ar), 7.31–7.27 (m, 4H, Ar), 7.16–7.13
(m, 6H, Ar), 7.10–7.06 (m, 2H, Ar) ppm. 13C NMR (400 MHz, CDCl3): δ = 159.14,
148.70, 146.10, 145.53, 145.49, 145.45, 145.41, 132.74, 132.71, 129.93, 128.40,
128.21, 128.16, 128.12, 127.08, 126.90, 124.20, 124.13, 123.79, 122.69, 121.49,
121.35, 117.99 (Ar), 123.26 (CF3) ppm. MALDI-TOF: m/z 389.78 [M]+. Calculated
[M]+: 390.13.
8

2.3
Synthesis
of
9,9-diethyl-7-(5-(trifluoromethyl)pyridin-2-yl)-N,N-diphenyl-9H-fluoren-2-ylamine
(L5) A mixture of 2-chloro-5-trifluoromethylpyridine (0.965 g, 5.32 mmol),
7-(diphenylamino)-9,9-diethylfluoren-2-yl-2-boronic acid 5b (3.0 g, 6.93 mmol), and
2 M K2CO3 (10 mL) in tetrahydrofuran (50 mL) were degassed with N2 for 5 min.
Pd(PPh3)4 (120 mg) was added and the resulting mixture was heated to reflux for two
days. After being cooled to room temperature, water was added and the mixture was
extracted with ethyl acetate. The combined organic fractions were dried over Na2SO4,
filtered and concentrated under reduced pressure. The crude compound was purified
by silica gel column chromatography using CH2Cl2/n-hexane (1:1, v/v) as eluent to
afford L5 as a white solid (0.88 g, 31%). 1H NMR (400 MHz, CDCl3): δ = 8.95 (s, 1H,
Ar), 8.01–7.98 (m, 3H, Ar), 7.96 (m, 1H, Ar), 7.73–7.71 (m, 1H, Ar), 7.62–7.60 (m,
1H, Ar), 7.28–7.14 (m, 4H, Ar), 7.12–7.11 (m, 5H, Ar), 7.06–7.00 (m, 3H, Ar), 2.10
(m, 4H, alkyl), 0.43 (m, 6H, alkyl) ppm. 13C NMR (400 MHz, CDCl3): δ = 161.31,
152.43, 149.92, 148.71, 148.30, 141.03, 140.41, 139.67, 136.47, 135.54, 135.47,
130.58, 129.75, 129.30, 128.99, 126.31, 125.65, 123.26, 122, 78, 122,19, 120.78 (Ar),
123.46 (CF3), 48.3 (sp3 C), 35.41, 8.91 (alkyl) ppm. MALDI-TOF: m/z 534.15 [M]+.
Calculated [M]+: 534.23.
9

–
2.4 General procedures for the synthesis of [Ru(CH3CN)4(L)]+PF6 (D1a–D5a)
To a flask containing [Ru(C6H6)Cl2]2, 2 molar equivalents of the corresponding
cyclometalating ligands (L1–L5) and KPF6 were added into degassed CH3CN (10
mL). Then, 2 molar equivalents of NaOH were added to the reaction mixture, and the
mixture was heated at 45 °C for 20 h. The solution was filtered by aluminum oxide
–
(basic) column chromatography. Yellow solids of [Ru(CH3CN)4(L)]+PF6 were drawn
out of the solution by the slow addition of n-hexane and dried under vacuum
completely. The ruthenium intermediate (D1a–D5a) was used without further
purification.
2.5 General procedures for the synthesis of ruthenium dyes (D1–D5)
A typical example is given for D1.
–
2.5.1 Synthesis of [Ru(dcbpy)2(L1)]+PF6 (D1)
–
To a flask containing [Ru(CH3CN)4(L1)]+PF6 (D1a) (100 mg, 0.123 mmol) in
methanol (15 mL), 2,2’-bipyridine-4,4’-dicarboxylic acid (dcbpy) (60 mg, 0.246
mmol) and NaOH (20 mg, 0.560 mmol) were added. The solution was refluxed
overnight at 100 °C. The dark purple mixture was dissolved in distilled water and
acidified with a few drops of 0.2 M HNO3, until a precipitate was formed. The dark
10

purple precipitate was filtered through a silica gel column using methanol as the
eluent. The solution was concentrated under reduced pressure. The compound was
1
then washed with diethyl ether to give D1 as a dark purple solid (95 mg, 68%). H
NMR (400 MHz, CD3OD): δ = 8.97 (s, 1H, Ar), 8.86 (s, 1H, Ar), 8.84 (s, 1H, Ar),
8.75 (s, 1H, Ar), 8.65 (s, 1H, Ar), 8.27 (d, J = 8.8 Hz, 1H, Ar), 8.01–7.98 (m, 3H, Ar),
7.89 (d, J = 6.0 Hz, 1H, Ar), 7.83–0.79 (m, 2H, Ar), 7.62 (d, J = 5.9 Hz, 1H, Ar),
7.54–7.50 (m, 3H, Ar), 7.46 (dd, J = 5.9, 1.7 Hz, 1H, Ar), 7.18–7.13 (m, 5H, Ar),
7.08–7.04 (m, 1H, Ar), 6.99–6.97 (m, 2H, Ar), 2.32 (s, 3H, CH3) ppm. MALDI-TOF:
– +
– +
m/z 990.89 [M–PF6 ] . Calculated [M–PF6 ] : 990.88. Elemental analysis calcd (%)
for C49H32N6F9O8PRu: C 51.81, H 2.84, N 7.40; found: C 51.90, H 3.02, N 7.65.
–
2.5.2 Synthesis of [Ru(dcbpy)2(L2)]+PF6 (D2)
1
Yield: 87 mg, 62%; dark purple solid. H NMR (400 MHz, CD3OD): δ = 9.09 (s, 1H,
Ar), 9.01 (s, 1H, Ar), 8.99 (s, 1H, Ar), 8.93 (s, 1H, Ar), 8.20–8.15 (m, 2H, Ar),
8.04-8.01 (m, 2H, Ar), 7.95–7.93 (m, 3H, Ar), 7.90 (d, J = 8.6 Hz, 1H, Ar), 7.82 (d, J
= 5.8 Hz, 1H, Ar), 7.70–7.65 (m, 5H, Ar), 7.62–7.58 (m, 4H, Ar), 7.52–7.50 (m, 1H,
Ar), 7.29–7.21 (m, 2H, Ar), 7.07 (t, J = 7.6 Hz, 1H, Ar) ppm. MALDI-TOF: m/z
– +
– +
977.64 [M–PF6 ] . Calculated [M–PF6 ] : 976.85. Elemental analysis calcd (%) for
C48H30N6F9O8PRu: C 51.39, H 2.70, N 7.49; found: C 51.54, H 2.98, N 7.44.
11

–
2.5.3 Synthesis of [Ru(dcbpy)2(L3)]+PF6 (D3)
1
Yield: 73 mg, 52%; dark purple solid. H NMR (400 MHz, CD3OD): δ = 9.12 (s, 1H,
Ar), 9.01 (s, 1H, Ar), 8.92 (s, 1H, Ar), 8.76 (s, 1H, Ar), 8.38 (d, J = 8.8 Hz, 1H, Ar),
8.17 (d, J = 5.8 Hz, 1H, Ar), 8.09–8.06 (m, 2H, Ar), 8.01–7.99 (m, 1H, Ar), 7.96 (dd,
J = 5.5, 1.5 Hz, 1H, Ar), 7.93–7.89 (m, 2H, Ar), 7.81 (dd, J = 5.9, 1.7 Hz, 1H, Ar),
7.69–7.67 (m, 2H, Ar), 7.61 (dd, J = 5.9, 1.7 Hz, 1H, Ar), 7.33–7.30 (m, 2H, Ar),
7.16–7.12 (m, 1H, Ar), 6.32 (s, 1H, Ar), 3.96 (t, J = 7.3 Hz, 2H, alkyl), 1.58–1.51 (m,
2H, alkyl), 1.15–1.10 (m, 2H, alkyl), 0.82–0.78 (m, 3H, alkyl) ppm. MALDI-TOF:
– +
– +
m/z 957.03 [M–PF6 ] . Calculated [M–PF6 ] : 956.86. Elemental analysis calcd (%)
for C46H34N6F9O8PRu: C 50.14, H 3.11, N 7.63; found: C 50.23, H 3.34, N 7.90.
–
2.5.4 Synthesis of [Ru(dcbpy)2(L4)]+PF6 (D4)
1
Yield: 76 mg, 54%; dark purple solid. H NMR (400 MHz, CD3OD): δ = 8.95 (s, 1H,
Ar), 8.86 (s, 1H, Ar), 8.74 (s, 1H, Ar), 8.57 (s, 1H, Ar), 8.14 (d, J = 5.8 Hz, 1H, Ar),
7.91–7.87 (m, 2H, Ar), 7.80–7.78 (m, 2H, Ar), 7.72 (d, J = 8.6 Hz, 1H, Ar), 7.67 (d, J
= 8.7 Hz, 1H, Ar), 7.62–7.60 (m, 2H, Ar), 7.47 (s, 1H, Ar), 7.41 (d, J = 5.9 Hz, 1H,
Ar), 7.32 (d, J = 5.7 Hz, 1H, Ar), 7.14–7.08 (m, 4H, Ar), 6.95–6.91 (m, 2H, Ar), 6.78
(d, J = 8.0 Hz, 4H, Ar), 6.48 (dd, J = 8.5, 2.1 Hz, 1H, Ar), 5.69 (d, J = 2.2 Hz, 1H, Ar)
12

– +
– +
ppm. MALDI-TOF: m/z 979.89 [M–PF6 ] . Calculated [M–PF6 ] : 978.87. Elemental
analysis calcd (%) for C48H32N6F9O8PRu: C 51.30, H 2.87, N 7.48; found: C 51.42, H
3.10, N 7.70.
–
2.5.5 Synthesis of [Ru(dcbpy)2(L5)]+PF6 (D5)
1
Yield: 69 mg, 51%; dark purple solid. H NMR (400 MHz, CD3OD): δ = 8.81 (s, 1H,
Ar), 8.72 (s, 2H, Ar), 8.69 (s, 1H, Ar), 7.92 (d, J = 5.8 Hz, 1H, Ar), 7.88 (d, J = 7.9 Hz,
1H, Ar), 7.74–7.71 (m, 2H, Ar), 7.68–7.63 (m, 3H, Ar), 7.51–7.43 (m, 4H, Ar),
7.34–7.31 (m, 2H, Ar), 7.01–6.99 (m, 4H, Ar), 6.88 (d, J = 8.2 Hz, 1H, Ar), 6.81–6.78
(m, 4H, Ar), 6.76–6.73 (m, 3H, Ar), 6.63 (dd, J = 8.2, 1.8 Hz, 1H, Ar), 1.85–1.59 (m,
– +
4H, alkyl), 0.22–0.18 (m, 3H, alkyl) ppm. MALDI-TOF: m/z 1122.99 [M–PF6 ] .
– +
Calculated [M–PF6 ] : 1123.08. Elemental analysis calcd (%) for C59H44N6F9O8PRu:
C 55.88, H 3.50, N 6.63; found: C 55.64, H 3.39, N 6.54.
2.6 Experimental set up for DSSC fabrication
The devices are fabricated according to our reported methods with minor
modifications [46]. Briefly, transparent conducting glass (SnO2:F, FTO glass, 15 Ω/□)
was cleaned by sonication in acetone, ethanol and deionized water for 20 min during
each washing step. The anode is composed of 12 µm of transparent layer of TiO2
13

nanoparticles (diameter: about 20 nm) and 6 µm of scattering layer (diameter of the
TiO2 nanoparticle at around 200 nm). The Ru(II) dye molecules (0.2 mM) were
dissolved in acetonitrile/t-butanol (1/1, v/v), and chenodeoxycholic acid (CDCA) was
also used as coadsorbent in the experiments. Platinum counter electrodes were
prepared by sputtering method at 15 mA for 60 s at a power of 150 W. Two holes
(0.75 mm in diameter) were pre-drilled in the FTO glass for introducing the
electrolyte. After that, dye-adsorbed TiO2 electrode and counter electrode were
stacked and sealed with 60 mm diameter thermal-plastic parafilm spacers at about 120
–
oC. I–/I3 liquid electrolyte (HL-HPE, 0.1 M LiI, 50 mM I2 and 0.6 M
dimethylpropylimidazolium iodide) was introduced into the sandwich cells through
the drilled holes in the Pt electrode. After that, the holes were sealed by parafilm and
covering glass at elevated temperature. The effective areas of all the photoanodes are
0.24 cm2. The photocurrent density–voltage (J−V) measurements were measured by a
semiconductor characterization system (Keithley 236) at room temperature in air
under the spectral output from solar simulator (Newport) using an AM 1.5G filter
−2
with a light power of 100 mW cm . The IPCEs of DSSCs were recorded in Solar
Cells QE/IPCE Measurement System (Zolix Solar Cell Scan 100) using DC mode.
2.7 Molecular computational calculations
14

The ground state structures of the Ru(II) complexes were optimized using density
functional theory (DFT) with B3LYP functional and 6-31G (d)/LanL2DZ basis set.
The excited-state related calculations (UV-Vis absorptions) were carried out using the
time-dependent DFT (TDDFT) method with the optimized structure of the ground
state (DFT/6-31G(d)/LanL2DZ). There are no imaginary frequencies in the frequency
analysis of all calculated structures. All of these calculations were performed with
Gaussian 09W program [47].
3. Results and discussion
3.1 Synthesis and characterization
A series of cyclometalating ligands (L1–L5) were synthesized and the synthetic
pathways are outlined in Scheme 1. All the ligands were synthesized from
2-chloro-5-(trifluoromethyl)pyridine and the corresponding boronic acid via Suzuki
coupling reaction. For the synthesis of L1, the tolylcarbazole was made by the
modified Ullmann condensation using CuI/phen/KOH as the catalyst combination
[45a]. Then one equivalent of N-bromosuccinimide (NBS) was used to produce
3-bromocarbazole precursor. Its boronic acid was then obtained by the reaction
between the precursor, n-butyllithium and trimethylborate. The ligand L1 was
synthesized from Suzuki coupling reaction between the boronic acid and the
15

commercially available 2-chloro-5-(trifluoromethyl)pyridine. For L2, the compound
4-bromo-2-nitrobiphenyl was obtained from the Suzuki-Miyaura coupling of
phenylboronic acid with 2,5-dibromonitrobenzene in the presence of a catalytic
amount of Pd(PPh3)4. It was then heated in triethylphosphite under the Cadogan’s
reductive cyclization conditions to afford 2-bromocarbazole. Ullmann condensation
between 2-bromocarbazole and iodobenzene gave 2-bromo-N-phenylcarbazole in
good yield, and the resulting compound reacted with n-butyllithium and
trimethylborate to afford the corresponding boronic acid. The cyclometalating ligand
L2 was then completed by Suzuki coupling of 9-phenyl-9H-carbazol-2-yl-2-boronic
acid with 2-chloro-5-(trifluoromethyl)pyridine. L3 and L4 can be obtained by similar
procedures of L1 and L2. L5 was obtained from the Suzuki coupling of
7-(diphenylamino)-9,9-diethyl-9H-fluoren-2-yl-2-boronic
acid
and
2-chloro-5-(trifluoromethyl)pyridine. Photosensitizers D1–D5 were then synthesized
in two steps (Scheme 2) [38]. Firstly, the ruthenium precursors (D1a–D5a) were
obtained by the reaction of the corresponding ligands (L1–L5) and the ruthenium
dimer [RuCl2(p-cymene)]2 in basic acetonitrile solution. The obtained Ru precursors
were then filtered through basic aluminum oxide and washed with diethyl ether to
remove the excess ligand. The Ru intermediates were eventually converted into the
cyclometalated Ru dyes D1–D5 without further purification through the reaction with
16

4,4’-dicarboxylic acid-2,2’-bipyridine (dcbpy) in refluxing methanol. The target dyes
are composed of a Ru metal core coordinated with two dcbpy as the anchoring groups
and one functionalized C^N ligand as the cyclometalating chelate.
All the ligands and Ru complexes were characterized by NMR spectrometry and
1
mass spectrometry and elemental analysis. The H NMR spectra of D1–D5 exhibit
multiple peaks because of the two dcbpy ligands that are magnetically non-equivalent.
The proton signals of aromatic rings of D1–D5 are located in the downfield region
(about 6.00−9.00 ppm), and the α-protons near the nitrogen atom in the dcbpy are at
the very downfield position at about 8.50–9.00 ppm. Those signals of the alkyl group
were observed in the upfield region (below 5.00 ppm).
3.2 Photophysical properties
The light-harvesting capacities at the molecular level were first compared by
recording the UV-Vis absorption spectra of the dyes D1–D5 in methanol. Their
normalized absorption spectra are depicted in Figure 2 and the detailed absorption
parameters are compiled in Table 1. Generally, for all the complexes, intense π→π*
transitions of the cyclometalating C^N ligands were observed below 390 nm. While
the broad absorption coverage originated from metal-to-ligand charge transfer (MLCT)
was detected at around 400, 490 and 560 nm, which are commonly observed in Ru(II)
17

dyes in the literature [32]. The lowest energy MLCT bands of our dyes are red shifted
as compared to the standard N719 dye (535 nm) and this indicates that more sunlight
can be harvested in the near infrared region [32]. Furthermore, the striking feature of
these complexes lies in the presence of a new band at about 490 nm which is absent in
N719 and the molar extinction coefficients of these cyclometalated Ru(II) complexes
are much higher than those of N719. This likely allows the use of thinner
nanocrystalline films in the devices and enhances the electrolyte diffusion in the film
and reduces the recombination of the light-induced charges during transportation
[45d].
Upon the introduction of carbazole, triphenylamine or diphenylaminofluorene
moiety to the cyclometalating ligand, the light-harvesting ability can be further
improved. These are reflected by the presence of an additional π→π* band at about
350 to 390 nm in our dyes which is not observed in phenylpyridine-based Ru(II)
complexes [42]. D1 has the lowest energy MLCT absorption wavelength at 556 nm,
which is almost the same as that observed in D2, and the absorption profiles of D1
and D2 are rather independent of the nature of the substitution position of
5-(trifluoromethyl)pyridine unit on carbazolyl ring [45e]. There was no notable
energy difference in the absorption bands between D1 and D3, but the absorption
coefficient was clearly affected by the variation of N-substituent on the carbazole unit.
18

D3 exhibited a smaller absorption coefficient as compared to D1 in the whole
absorption range. On the other hand, D5 was red-shifted in its absorption spectrum as
compared to D4 due to the extension of electronic delocalization over the whole C^N
ligand caused by the introduction of an additional fluorene moiety in the former.
3.3 Electrochemical properties
While a good absorption profile is a preliminary requirement for photovoltaic
cells, the highest occupied molecular orbital (HOMO) and lowest unoccupied
molecular orbital (LUMO) levels of the dyes also determine the efficient device
operation, as they govern the possibility of electron injection from the excited dye
complexes to the conductive band (Ecb) of TiO2 and the ease of dye regeneration
[46,48,49]. To investigate the energy levels of the dye molecules, a conventional
three-electrode configuration consisting of a glassy carbon working electrode and
Pt-wire as the counter and reference electrodes was used. Cyclic voltammetry
experiments of the dye complexes were performed in thin films on glassy carbon
electrode measured in 0.1 M tetrabutylammonium hexafluorophosphate in acetonitrile
solution at a scan rate of 50 mV s–1. Under these conditions, the E1/2 of ferrocene was
0.07 V versus Ag/Ag+. The oxidation (Eox) potential was used to calculate the HOMO
energy level while the LUMO can be calculated from the reduction (Ered) potential.
19

The LUMO levels of the dyes D1–D5 are almost the same because the dcbpy with
carboxylate anchoring group would govern the LUMO which is in agreement with the
computational results of DFT studies. The LUMO levels lie between –3.55 to –3.56
eV with respect to the energy level of the conduction band of TiO2 (–4.4 eV), which
supply a sufficient driving force for charge injection. Therefore, efficient charge
injection can occur from the excited photosensitizer molecule to TiO2 conduction
band [48,49]. The HOMO levels of the sensitizers range between –4.91 and –5.06 eV,
–
relative to that of the redox couple I–/I3 (–4.9 eV) [46] which avoid the geminate
charge recombination between oxidized dye molecules and photoinduced electrons in
the nanocrystalline titania film. This can successfully provide ample driving force for
dye regeneration to complete the circuit and restore the dyes [46]. However, some of
the HOMO levels of these dyes are close to the potential of the redox couple in the
electrolyte, making them less favorable for fast dye generation and thus resulting in
lower power conversion efficiency of DSSCs. The electrochemical properties of all
the photosensitizers are summarized in Table 2.
3.4 Computational studies
To gain further insight into the electronic properties of our dyes, the geometries of
the complexes at the ground state were optimized with DFT approach (Figure 3). The
20

frontier molecular orbitals are presented and the other relevant computational results
are tabulated in the Supplementary Information. The HOMO of the complexes is
generally localized on the electron-donating carbazole or triphenylamine unit.
Conversely, the LUMO of the complexes is generally localized on the dcbpy ligand.
Therefore, the photoexcitation induced electron transfer from the C^N ligand to the
dcbpy moiety is in agreement with the good potential of our complexes as photoactive
materials for solar cell application.
In order to study the transitions involved in the UV-Vis absorption of the
complexes, the UV-Vis absorption spectra of the complexes were calculated with the
TDDFT method. Generally, TDDFT calculations predict that the carbazole-containing
complexes give strong absorption in the visible spectral region as compared with the
triphenylamine-containing complexes. This information should be helpful for the
design of the photosensitizers in DSSCs.
3.5 Thermal properties
The thermal properties of D1–D5 were examined by thermal gravimetric analysis
(TGA) under nitrogen. All of them exhibited good thermal stability with the onset
decomposition temperature for 5% weight-loss (Td) at 329, 319, 345, 333 and 347 oC,
respectively for D1–D5. These results indicate that the thiocyanate-free
cyclometalated Ru(II) complexes are promising candidate for thermally stable DSSC
21

application.
3.6 Photovoltaic properties
DSSCs were fabricated using the synthesized dyes according to the methods
described in the experimental section and their photovoltaic properties were measured
−2
under standard AM 1.5G irradiation conditions (100 mW cm ). The photocurrent
action spectrum, i.e. the incident photon-to-current conversion efficiency (IPCE) as a
function of wavelength, was measured to evaluate the photoresponse of
photoelectrode in the entire spectral region. Figure 4 shows the IPCE spectra obtained
with D1–D5 as photosensitizers for DSSCs. All of the dyes can convert light to
photocurrent in the spectral region from 400 to 750 nm satisfactorily. Notably, the
IPCE of D1 is the highest as compared to the other dyes, and it reaches a maximum of
56% at 560 nm and remains high throughout the near infrared region of the IPCE
curve. This indicates that D1 has the highest conversion yield from the sun. The main
reason for this observation is probably due to its highest extinction coefficient among
all the dyes.
All the DSSC parameters of these dyes, such as short-circuit current density (Jsc),
open-circuit voltage (Voc), fill factor (ff) and power conversion efficiency (η) are
summarized in Table 3. Figure 5 shows the J–V curves for all the devices. There are
two major structural factors worthy of attention for improving the photovoltaic
22

performance of the solar cells: Firstly, the effect of varying the electron-donating
moiety such as 3-carbazole, 2-carbazole, triphenylamine or diphenylaminofluorene of
the cyclometalating ligand, and secondly, the influence induced by the substituent at
the N-position of carbazole in the bidentate ligand. The Ru(II) photosensitizer D1
with 9-p-tolyl-9H-carbazole moiety enhances the DSSC efficiency via an obvious
improvement of Jsc. However, there is a significant influence on η when the
N-substituted tolyl group of the carbazole is changed to the butyl chain in D3. The
photovoltaic performance for D3 dramatically decreases due to the lowering of the Jsc
value. This finding shows that not only the electron-donating moiety, but also the
identity of N-substituent on the carbazole unit of the C^N ligand can largely affect the
overall photovoltaic performance. Under the same fabrication conditions, D1 shows
the highest Jsc value of 8.06 mA cm–2 and the value is almost double of those other
four dyes, causing D1 to exhibit the highest η of 3.39%. The broader IPCE spectrum
of D1 among all the dyes promotes higher Jsc value. This implies that D1 with broad
spectrum could inject more electrons effectively into the TiO2 conduction band.
Therefore, D1 with 3-substituted tolylcarbazole ligand is the most promising dye with
good solar cell performance among this series of photosensitizers and hence the η of
the dyes can be improved by the structural modification of the cyclometalated C^N
ligand in such system of Ru(II) thiocyanate-free complexes. Although the IPCE of D5
23

is higher than those of D2–D4, the similarity of the η is ascribed to the trade-off
relationship between Voc and Jsc in their solar cells.
4. Conclusions
Five new thiocyanate-free ruthenium(II) cyclometalated complexes with different
electron-donating moieties attached to 5-trifluoromethylpyridine were successfully
synthesized and fully characterized by spectroscopic and computational methods. The
nature of cyclometalating ligands effectively tunes the properties of the metal
complexes and the resulting DSSC performance. The photosensitizer D1 gave the
highest η of 3.39% under AM 1.5 irradiation. These findings draw our attention to the
fact that the photovoltaic performance of the DSSC device can be controlled by the
structural modification of the cyclometalating ligand.
Acknowledgements
C.-H. Siu and C.-L. Ho contributed equally to this work. We thank Hong Kong
Baptist University (FRG2/12-13/083 and FRG1/12-13/046), Hong Kong Research
Grants Council (HKBU203011) and Areas of Excellence Scheme, University Grants
Committee of HKSAR, China (project No. AoE/P-03/08), the National Natural
Science Foundation of China (project number 91222201), the National Basic
24

Research Program of China (973 Program) (2013CB834702), the Science,
Technology and Innovation Committee of Shenzhen Municipality
(JCYJ20120829154440583) and Dalian University of Technology. H.L. thanks the
985 Program and 111 Project from Minzu University of China (CUN985-07-08 and
111 Project B08044) for financial support.
References
[1] B.L. Feringa, Molecular Switches, Wiley-VCH, Weinheim, (2001).
[2] V. Balzani, M. Venturi, A. Credi, Molecular Devices and Machines, Wiley-VCH,
Weinheim (2003).
[3] W.-Y. Wong, C.-L. Ho, Coord. Chem. Rev. 253 (2009) 1709.
[4] W.-Y. Wong, C.-L. Ho, J. Mater. Chem. 19 (2009) 4457.
[5] Y. Chi, P.-T. Chou, Chem. Soc. Rev. 39 (2010) 638.
[6] Y.-J. Cheng, S.-H. Yang, C.-S. Hsu, Chem. Rev. 109 (2009) 5868.
[7] S. Gunes, H. Neugebauer, N.S. Sariciftci, Chem. Rev. 107 (2007) 1324.
[8] B.C. Thompson, J.M. Frechet, Angew. Chem. Int. Ed. 47 (2008) 58.
[9] G. Vilaca, B. Jousseaume, C. Mahieux, C. Belin, H. Cachet, M. C. Bernard, V.
Vivier, T. Toupance, Adv. Mater. 18 (2006) 1073.
[10] B.C. O’Regan, J.R. Durrant, Acc. Chem. Res. 42 (2009) 1799.
25

[11] W.-Y. Wong, C.-L. Ho, Acc. Chem. Res. 43 (2010) 1246.
[12] W.-Y. Wong, Macromol. Chem. Phys. 209 (2008) 14.
[13] J.K. Rath, R.E.I. Schropp, Sol. Energy Mater. Sol. Cells 53 (1998) 189.
[14] B. O’Regan, M. Grätzel, Nature 353 (1991) 737.
[15] S. Ardo, G.J. Meyer, Chem. Soc. Rev. 38 (2009) 115.
[16] M. Grätzel, Acc. Chem. Res. 42 (2009) 1788.
[17] M.K. Nazeeruddin, M. Grätzel, Struct. Bonding 123 (2007) 113.
[18] L.M. Goncalves, V. Bermudez, H.A. Ribeiro, A.M. Mendes, Energy Environ. Sci.
1 (2008) 655.
[19] A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem. Rev. 110 (2010)
6595.
[20] M. Grätzel, Nature 414 (2001) 338.
[21] A. Hagfeldt, M. Grätzel, Acc. Chem. Res. 33 (2000) 269.
[22] M.K. Nazeeruddin, Coord. Chem. Rev. 248 (2004) 1161
[23] (a) L.-F. Lai, C. Qin, C.-H. Chui, A. Islam, L. Han, C.-L. Ho, W.-Y. Wong, Dyes
Pigments 98 (2013) 428;
(b) L.-F. Lai, C.-L. Ho, Y.-C. Chen, W.-J. Wu, F.-R. Dai, C.-H. Chui, S.-P. Huang,
K.-P. Guo, J.-T. Lin, H. Tian, S.-H. Yang, W.-Y. Wong, Dyes Pigments 96 (2013)
516.
26

[24] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Mueller, P. Liska,
N. Vlachopoulos, M. Grätzel, J. Am. Chem. Soc. 115 (1993) 6382.
[25] M.K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska,
S. Ito, B. Takeru, M. Grätzel, J. Am. Chem. Soc. 127 (2005) 16835.
[26] M.K. Nazeeruddin, R. Splivallo, P. Liska, P. Comte, M. Grätzel, Chem. Commun.
(2003) 1456.
[27] P. Wang, S.M. Zakeeruddin, I. Exnar, M. Grätzel, Chem. Commun. (2002) 2972.
[28] P. Wang, S.M. Zakeeruddin, J.E. Moser, M.K. Nazeeruddin, T. Sekiguchi, M.
Grätzel, Nat. Mater. 2 (2003) 402.
[29] P. Wang, B. Wenger, R. Humphry-Baker, J.E. Moser, J. Teuscher, W.
Kantlehner, J. Mezger, E.V. Stoyanov, S.M. Zakeeruddin, M. Grätzel, J. Am.
Chem. Soc. 127 (2005) 6850.
[30] M.K. Nazeeruddin, P. Pechy, 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. Grätzel, J. Am. Chem. Soc. 123 (2001) 1613.
[31] M.K. Nazeeruddin, P. Péchy, M. Grätzel, Chem. Commun. (1997) 1705.
[32] T. Bessho, E. Yoneda, J.-H. Yum, M. Guglielmi, I. Tavernelli, H. Imai, U.
Rothlisberger, M. Nazeeruddin, M. Grätzel, J. Am. Chem. Soc. 131 (2009) 5930.
[33] K. Wu, H, Hsu, K. Chen, Y. Chi, M. Chung, W. Liu, P. Chou, Chem. Commun.
27

46 (2010) 5124.
[34] C. Dragonetti, A. Valore, A. Colombo, D. Roberto. V. Trifiletti, N. Manfredi, M.
M. Salamone, R. Ruffo, A. Abbotto, J. Organomet. Chem. 714 (2012) 88.
[35] A. Abbotto, C. Coluccini, E. Dell’Orto, N. Manfredi, V. Trifiletti, M.M.
Salamone, R. Ruffo, M. Acciarri, A. Colombo, C. Dragonetti, S. Ordanini, D.
Roberto, A. Valore, Dalton Trans. 41 (2012) 11731.
[36] P.G. Bomben, K.D. Thériault, C.P. Berlinguette, Eur. J. Inorg. Chem. (2011)
1806.
[37] P.G. Bomben, K.D. Robson, C.P. Berlinguette, Inorg. Chem. 48 (2009) 9631.
[38] C.-H. Siu, C.-L. Ho, J. He, T. Chen, X. Cui, J. Zhao, W.-Y. Wong, J. Organomet.
Chem. 748 (2013) 75.
[39] P.G. Bomben, J. Borau-Garcia, C.P. Berlinguette, Chem. Commun. 48 (2012)
5599.
[40] P.G. Bomben, T.J. Gordon, E. Schott, C.P. Berlinguette, Angew. Chem. Int. Ed.
123 (2011) 10870.
[41] S.H. Wadman, J.M. Kroon, K. Bakker, M. Lutz, A.L. Spek, G.P.M. van Klink, G.
van Koten, Chem. Commun. (2007) 1907.
[42] S.H. Wadman, M. Lutz, D.M. Tooke, A.L. Spek, F. Hartl, R.W.A. Havenlth,
G.P.M. van Kllnk, G. van Koten, Inorg. Chem. 48 (2009) 1887.
28

[43] S.H. Wadman, Y.M. van Leeuwen, R.W.A. Havenith, G.P.M. van Klink, G. van
Koten, Organometallics 29 (2010) 5635.
[44] S.H. Wadman, J.M. Kroon, K. Bakker, R.W.A. Havenith, G.P.M. van Klink, G.
van Koten, Organometallics 29 (2010) 1569.
[45] (a) C.-L. Ho, Q. Wang, C.-S. Lam, W.-Y. Wong, D. Ma, L. Wang, Z.-Q. Gao,
C.-H. Chen, K.-W. Cheah, Z. Lin, Chem. Asian. J. 4 (2009) 89;
(b) C.-L. Ho, L.-C. Chi, W.-Y. Hung, W.-J. Chen, Y.-C. Lin, H. Wu, E. Mondal,
G.-J. Zhou, K.-T. Wong, W.-Y. Wong, J. Mater. Chem. 22 (2012) 215;
(c) S.M. Aly, C.-L. Ho, W.-Y. Wong, D. Fortin, P.D. Harvey, Macromolecules
42 (2009) 6902;
(d) W. Wu, J. Zhang, H. Yang, B. Jin, Y. Hu, J. Hua, C. Jing, Y. Long, H. Tian, J.
Mater. Chem. 22 (2012) 5382;
(e) W.-Y. Wong, C.-L. Ho, Z. Q. Gao, B. X. Mi, C. H. Chen, K. W. Cheah, Z.
Lin, Angew. Chem. Int. Ed. 45 (2006) 7800;
(f) W.-Y. Wong, G.-J. Zhou, X.-M. Yu, H.-S. Kwok and B.-Z. Tang, Adv. Funct.
Mater. 16 (2006) 838.
[46] (a) S. Chang, H. Wang, Y. Hua, Q. Li, X. Xiao, W.-K. Wong, W.-Y. Wong, X.
Zhu, T. Chen, J. Mater. Chem. A 1 (2013) 11553;
(b) Y. Hua.,S. Chang, D. Huang, X. Zhou, X. Zhu, J. Zhao, T. Chen, W.-Y. Wong,
29