The design and synthesis of thiophene-based ruthenium(II) complexes as promising sensitizers for dye-sensitized solar cells

Article abstract of DOI:10.1016/j.dyepig.2017.01.030

A new series of promising synthetically facile cycloruthenated thiophene-based sensitizers have been developed and fully characterized by UV–vis spectroscopy, NMR and cyclic voltammetric studies. The synthesized dyes have broad MLCT bands spanning the visible spectrum, with high extinction coefficients. The energies of the molecular orbitals for the isolated molecules of the complexes and densities of occupied states were determined. The cycloruthenated compounds contains ortho-metallated thiophene moiety substituted by N-(methyliden)aniline or pyridine-2-yl at the ortho-position. Having 4,4’-dicarboxy-2,2’-bipyridine as the linker and auxiliary ligands and anchored to nanocrystalline TiO₂ films, they achieve efficient sensitization in the visible range and display an overall conversion efficiency of 5.3% under standard AM 1.5 sunlight.

Full text of DOI:10.1016/j.dyepig.2017.01.030

Accepted Manuscript
The design and synthesis of thiophene-based ruthenium(II) complexes as promising
sensitizers for dye-sensitized solar cells
A.V. Medved'ko, V.K. Ivanov, M.A. Kiskin, A.A. Sadovnikov, E.S. Apostolova, V.A.
Grinberg, V.V. Emets, A.O. Chizov, O.M. Nikitin, T.V. Magdesieva, S.A. Kozyukhin
PII:
S0143-7208(17)30071-2
10.1016/j.dyepig.2017.01.030
DYPI 5729
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 29 July 2016
Revised Date: 14 October 2016
Accepted Date: 12 January 2017
Please cite this article as: Medved'ko AV, Ivanov VK, Kiskin MA, Sadovnikov AA, Apostolova ES,
Grinberg VA, Emets VV, Chizov AO, Nikitin OM, Magdesieva TV, Kozyukhin SA, The design and
synthesis of thiophene-based ruthenium(II) complexes as promising sensitizers for dye-sensitized solar
cells, Dyes and Pigments (2017), doi: 10.1016/j.dyepig.2017.01.030.
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.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
The design and synthesis of thiophene-based ruthenium(II) complexes as prom-
ising sensitizers for dye-sensitized solar cells
✝
A. V. Medved’ko,*a,b V. K. Ivanov,c,d M. A. Kiskin,c A. A. Sadovnikov,c E. S. Apostolova,e V. A. Grin-
berg,f V. V. Emets,f A. O. Chizov,g O. M. Nikitin,a,h T. V. Magdesievaa and S. A. Kozyukhinc,d
a Department of Chemistry, Lomonosov Moscow State University, Moscow, 119991, Rus-
sia.*medved'ko@org.chem.msu.ru
b The Photochemistry Center, RAS, Moscow, 119421, Russia.
c Kurnakov Institute of General and Inorganic Chemistry, RAS, Moscow, 119991, Russia.
d Department of Chemistry, National Research Tomsk State University, Tomsk, 634050, Russia.
e Mendeleev University of Chemical Technology, Moscow, 125047, Russia
f Frumkin Institute of Physical Chemistry and Electrochemistry, RAS, Moscow, 119071, Russia
g Zelinsky Institute of Organic Chemistry, RAS, Moscow, 119991, Russia
h A.N. Nesmeyanov Institute of Organoelement Compounds, RAS, Moscow, Russia
Keywords: Cyclometallation; Ruthenium(II) sensitizer; Thiophene; Electronic spectra; DFT;Dye-
sensitized solar cells
Abstract
A new series of promising synthetically facile cycloruthenated thiophene-based sensitizers have been de-
veloped and fully characterized by UV-vis spectroscopy, NMR and cyclic voltammetric studies. The syn-
thesized dyes have broad MLCT bands spanning the visible spectrum, with high extinction coefficients.
The energies of the molecular orbitals for the isolated molecules of the complexes and densities of occu-
pied states were determined. The cycloruthenated compounds contains ortho-metallated thiophene moiety
substituted by N-(methyliden)aniline or pyridine-2-yl at the ortho-position. Having 4,4’-dicarboxy-2,2’-
bipyridine as the linker and auxiliary ligands and anchored to nanocrystalline TiO2 films, they achieve
1

ACCEPTED MANUSCRIPT
efficient sensitization in the visible range and display an overall conversion efficiency of 5.3% under
standard AM 1.5 sunlight.
1. Introduction
The search for sustainable sources of renewable energy is one of the greatest challenges to 21st century
civilization. Among existing energy sources, a conversion of solar energy to electric power (photovolta-
ics, PV) is one of the most attractive technologies. According to the International Energy Agency Photo-
voltaic Power Systems Programme (IEA-PVPS) report, energy production of PV systems increased up to
36.9 GW in the year 2013 as compared to 29 GW in the year 2012 [1]. Traditional monocrystalline Si-
based solar cells can reach 27% power conversion efficiency (PCE) [2]. The new type of multijuction cell
containing Ga, As, In, N, P elements gave 43.5% PCE [3]. However, the fabrication of such solar cells
requires very clean conditions (for Si-based cells) or the use of toxic elements (for SJ3-type cells). Thus,
nowadays traditional photovoltaic technology faces strong competition from third generation PV (3GPV)
technology, especially dye-sensitized solar cells (DSSC) and organic solar cells. PCE reached 13% for the
former [4] and 12.05% for the latter [5]. PCE of all-solid-state dye-sensitized solar cells based on lead-
halide organic-inorganic perovskites have quickly increased over the last six years [6], and the perfor-
mance of the cells with a National Renewable Energy Laboratory (NREL)-certified reached record effi-
ciency of 22.1%. DSSCs have a leading edge over other technologies due to their lower cost, simple fab-
rication methods and high power conversion efficiency. In these cells, the dye is one of the key compo-
nents for high PCE.
A typical ruthenium(II) dye for DCCSs based on mesoporous titania film utilizes three types of ligands –
a linker which connects Ru to the TiO2 surface [7]; an anionic ligand modulating HOMO energy; an aux-
iliary ligand modulating LUMO energy [8]. Linking and anionic ligands are usually 2,2’-bipyridine-4,4’-
dicarboxylic acid and two thiocyanate moieties, respectively. A lot of different auxiliary and linking lig-
2

ACCEPTED MANUSCRIPT
ands were probed to enhance the PCE of solar cells [9]. Maximal PCE value for NCS-containing equals
11.5% [10]. The main drawback of these dyes is their low photostability and thermal stability [11]. An-
other type of anionic ligand is 2-phenylpyridine and its derivatives. The PCE of corresponding DSSCs
reached 10.1% [12]. Attempts to replace 2-phenylpyridine with other similar ligands were also made [13–
16]. One very promising ligand is ortho-substituted thiophene. Recently it has been shown that thiophene
derivatives containing a coordinating ligand at the 2-position can be metallated to the 3-position to give
cycloruthenated compounds [17,18]. The usage of corresponding complexes as sensitizers in DSSCs was
considered in one report only [19]. In the present paper, the preparation of the simplest cycloruthenated
compounds containing ortho-metallated thiophene moiety substituted by N-(methyliden)aniline or pyri-
dine-2-yl at the ortho-position, with traditional 4,4’-dicarboxy-2,2’-bipyridine as the linker and auxiliary
ligands is reported. To get an insight of the substituent influence on the dyes electronic properties, their
electrochemical, spectral and DFT studies were carried out. To determine the solar energy conversion ef-
ficiency of new d6-ruthenium-based organometallic sensitizers, testing their performance in DSSCs was
carried out.
2. Experimental section
2.1. Materials and instruments
All starting reagents were obtained from commercial sources. The solvents were dried prior to the use by
conventional methods. Di-µ-chlorobis((η6-benzene)ruthenium(II) chloride) [20] and diethyl 2,2’-
bipyridine-4,4’-dicarboxylate [21] were prepared as described elsewhere. 1H and 13C NMR spectra were
obtained on Bruker DRX-400 and AVANCE-500 spectrometers. Elemental analysis was carried out on a
Eurovector EA 3000 automated analyzer. Electronic spectra were registered at room temperature on a
Cary5000 UV-vis-NIR spectrometer (Agilent) equipped with DRA2500. High resolution mass spectra
(HRMS) were measured on a Bruker micrOTOF II instrument using electrospray ionization (ESI) [22].
3

ACCEPTED MANUSCRIPT
The measurements were made in a positive ion mode (interface capillary voltage – 4500 V). Mass range
was from m/z 50 to m/z 3000; external or internal calibration was done with ESITuningMix, Agilent. A
syringe injection was used for solutions in acetonitrile (flow rate 3 ꢀL/min). For compounds 7 and 8, tri-
ethylamine was added to enhance solubility. Nitrogen was applied as dry gas; interface temperature was
set at 180 °C. The HRMS values are presented for the most abundant peak in the isotope cluster.
2.2. Synthesis and characterization of dyes
N-(thiophen-2-ylmethyliden)aniline (1). Synthesized as described previously [23]. Yield 72%. δН (400
MHz, CDCl3, 27 °С): 7.14-7.16 (dd, 1H, 4-H-anil, 3JH,H = 4.8, 3.8 Hz), 7.20-7.25 (m, 3H, 2-H-anil, 4-H-
th), 7.38-7.42 (m, 2H, 3-H-anil), 7.47-7.51 (m, 2Н, 3,5-H-th), 8.56 (s, 1Н, CH=N).
2-Bromothiophene. Synthesized as described previously [24]. Yield 58 %. Bp 53-56 °С (18 mmHg). δН
(400 MHz, CDCl3, 27°С): 6.87 (dd, 1Н, 4-H, 3JH,H = 5.5, 3.7 Hz), 7.05 (dd, 1Н, 3-H, 3JH,H = 1.7, 3.7 Hz),
7.23 (dd, 1Н, 5-H, 3JH,H = 5.5, 1.3 Hz).
2-(thiophen-2-yl)pyridine (2). Synthesized as described previously [25] and purified as follows. Sticky
solid obtained after washing with ammonium chloride solution was filtered off through Celite and washed
with 160 ml of diethyl ether until slight coloration of the washings. Then the water phase was extracted
with 30 ml of diethyl ether. Organic phases were combined and washed with 30 ml of brine. The organic
phase was dried over magnesium sulfate. The solvent was removed on a rotary evaporator. The residue
was dissolved in hot pentane. Upon cooling to room temperature red oil was precipitated. Pentane solu-
tion was decanted from oil. The solvent was removed on a rotary evaporator. The residue was crystallized
from the mixture of pentane and petroleum ether to give yellow needles of the product. The solvent from
mother liquor was removed on a rotary evaporator. The residue was purified with column chromatog-
raphy on silica gel (eluent: petroleum ether: ethyl acetate 4:1). 3.76 g (83%) of yellow needles was ob-
4

ACCEPTED MANUSCRIPT
tained. δН (400 MHz, CDCl3, 27 °С): 7.09-7.15 (m, 2Н, 4-H-th, 5-H-Py), 7.38 (m, 1Н, 5-H-th), 7.58 (m,
1Н, 3-H-th), 7.63-7.69 (m, 2Н, 3,4-H-Py), 8.55-8.57 (m, 1Н, 6-H-Py).
General procedure for synthesis of compounds 3 and 4. Synthesized analogously to [26] and purified
by column chromatography on silica gel (eluent DCM:MeCN 10:1). The product was dissolved in a min-
imum volume of DCM and precipitated with petroleum ether.
Tetrakis(acetonitrile)[N-((thiophene-2-yl-κC3)methyliden)ani-line-κN]ruthenium(II) hexafluoro-
phosphate (3). Orange powder. Yield 30%. Single crystals were grown by vapor diffusion of petroleum
ether into benzene/methanol solution of the product. δН (400 MHz, CD3CN, 27 °С): 1.96 (s, 3Н, CH3),
2.12 (s, 6Н, CH3), 2.51 (s, 3Н, CH3), 7.28-7.33 (m, 3Н, 2,4-H-anil), 7.41-7.45 (m, 2Н, 3-H-anil,) 7.62 (d,
1Н, th, 3JH,H = 4.7 Hz), 7.83 (d, 1Н, th, 3JH,H = 4.7 Hz), 8.40 (s, 1Н, CH=N).
Tetrakis(acetonitrile)[2-(thiophen-2-yl-κC3)pyridine-κN]ruthe-nium(II) hexafluorophosphate (4).
Yellow powder. Yield 45%. Single crystals were grown by vapor diffusion of diethyl ether into chloro-
form solution of the product. δН (400 MHz, CD3CN, 27 °С): 1.96 (s, CH3, 3Н), 2.03 (s, 6Н, CH3), 2.51
(s, 3Н, CH3), 6.98 (m, 1Н, Py), 7.43 (d, 1Н, 3-H-Py, 3JH,H = 8.1 Hz), 7.49 (d, 1Н, th, 3JH,H = 4.7 Hz), 7.56
(d, 1Н, th, 3JH,H = 4.7 Hz), 7.65 (td, 1Н, Py, 3JH,H = 8.9, 1.4 Hz), 8.73 (d, 1Н, 6-H-Py, 3JH,H = 5.5 Hz).
General procedure for synthesis of compounds 5 and 6. Synthesized analogously to [18] and purified
by column chromatography on silica gel (eluent DCM:MeOH 25:1). The product was dissolved in a min-
imum volume of DCM and precipitated with petroleum ether.
Bis(4,4’-diethoxycarbonyl-2,2’-bipyridine-κ2N,N’)[N-((thio-phen-2-yl-κC3)methyliden)aniline-
κN]ruthenium(II) hexafluorophosphate (5). Red solid. Yield 77%. δН (400 MHz, CDCl3, 27 °С): 1.39-
1.49 (m, 12Н, CH3), 4.41-4.53 (m, 8Н, CH2), 6.37 (m, 3H, 4-H-th, 2,6-H-anil), 6.90-6.93 (m, 2Н, 3,5-H-
anil), 6.94-6.97 (m, 1Н, 4-H-anil), 7.64 (d, 1Н, 5-H-th, 3JH,H = 4.6 Hz), 7.80 (dd, 1Н, 5-H-Py, 3JH,H = 6.0,
1.4 Hz), 7.86-7.89 (m, 3Н, 5,6-H-Py), 7.95 (d, 1Н, 6-H-Py, 3JH,H = 6.0 Hz), 8.00-8.03 (m, 2Н, 5,6-H-Py),
5

ACCEPTED MANUSCRIPT
8.47 (s, 1Н, 3-H-Py), 8.54 (s, 1Н, 3-H-Py), 8.59 (s, 1Н, CH=N), 8.74 (d, 1Н, 6-H-Py), 8.80 (s, 1Н, 3-H-
Py), 8.84 (s, 1Н , 3-H-Py). δC (100 MHz, CDCl3, 27 °С): 14.07 (CH3); 14.15 (CH3); 14.18 (CH3); 62.57
(CH2); 62.62 (CH2); 62.71 (CH2); 121.39 (2,6-CH-anil); 121.76 (3-CH-Py); 121.82 (3-CH-Py); 121.22
(3-CH-Py); 121.29 (3-CH-Py); 125.85 (5-CH-Py); 126.11 (5-CH-Py); 126.37 (6-CH-Py); 126.71 (4-CH-
anil, 5-CH-Py); 128.95 (3,5-CH-anil); 133.10 (4-CH-th); 135.78 (4-C-Py); 135.91 (5-CH-th); 136.06 (4-
C-Py); 136.75 (4-C-Py); 137.45 (4-C-Py); 137.86 (2-C-th); 149.74 (1-C-anil); 150.25; 151.39 (5-CH-Py);
151.63 (6-CH-Py); 153.16 (6-CH-Py); 154.58 (2-C-Py); 155.11 (6-CH-Py); 156.39 (2-C-Py); 157.53 (2-
C-Py); 157.68 (2-C-Py); 163.57 (COO); 163.71 (COO); 164.95 (CH=N); 206.65 (3-C-th). (Thiophene is
denoted as “th”, aniline - as “anil”, bipyridines - as “Py”, respectively). Found: C, 50.24; H, 4.20; N, 6.72;
S 2.72. Calc. for C43H40N5F6O8PRuS: C, 50.00; H, 3.90; N, 6.78; S, 3.10. High resolution ESI-MS: m/z
888.1637 (calc. for C43H40N5O8RuS+m/z 888.1647).
Bis(4,4’-diethoxycarbonyl-2,2’-bipyridine-κ2N,N’)[2-(thio-phen-2-yl-κC3)pyridine-
κN]ruthenium(II) hexafluorophosphate (6). Red solid. Yield 63%. δН (400 MHz, CDCl3, 27 °С): 1.44
(m, 12Н, CH3), 4.43-4.50 (m, 8Н, 4CH2), 6.16 (d, 1Н, 3JH,H = 4.6Hz, 4-H-th), 6.83-6.86 (m, 1Н, 4-H-
Py1), 7.31 (d, 1Н, 3JH,H = 6.0 Hz, 3-H-Py1), 7.35 (d, 1Н, 3JH,H = 4.6 Hz, 5-H-th), 7.50-7.52 (m, 1Н, 6-H-
Py1), 7.56-7.60 (m, 1Н, 5-H-Py1), 7.75-7.77 (m, 1Н, 6-H-Py), 7.79-7.81 (m, 1Н, 6-H-Py), 7.86-7.90 (m,
2Н, 5,6-H-Py), 8.05-8.09 (m, 2Н, 5,6-H-Py), 8.13 (d, 1Н, 3JH,H = 6.0 Hz, 5-H-Py), 8.15 (d, 1Н, 3JH,H = 6.0
Hz, 5-H-Py), 8.77 (s, 1Н, 3-H-Py), 8.77 (s, 1H, 3-H-Py), 8.84 (s, 1H, 3-H-Py), 8.92 (s, 1H, 3-H-Py). δC
(100 MHz, CDCl3, 27 °С): 14.11 (CH3); 14.17 (CH3); 62.43 (CH2); 62.54 (CH2); 62.63 (CH2); 62.69
(CH2); 118.49 (6-CH-Py1); 120.84 (4-CH-Py1); 122.02 (3-CH-Py); 122.05 (3-CH-Py); 122.42 (3-CH-
Py); 122.53 (3-CH-Py); 125.55 (6-CH-Py); 125.74 (6-CH-Py); 126.18 (5-CH-Py); 129.46 (6-CH-Py);
132.43 (5-CH-Py); 135.26 (2-C-th); 136.1 (4-C-Py); 136.45 (4-C-Py); 137.16 (4-C-Py); 138.15 (4-C-Py);
149.98 (3-CH-Py1); 150.68 (5-CH-Py); 150.80 (6-CH-Py); 151.69 (5-CH-Py); 155.20 (5-CH-Py); 155.42
(2-C-Py); 156.19 (2-C-Py); 157.61 (2-C-Py); 157.69 (2-C-Py); 162.31 (2-C-Py1); 163.55 (COO); 163.66
6

ACCEPTED MANUSCRIPT
(COO); 163.74 (COO); 163.85 (COO); 190.08 (3-C-th). (Thiophene is denoted as “th”, pyridine bound
with thiophene - as “Py1”, bipyridines - as “Py”, respectively). Found: C, 49.70; H, 4.18; N, 6.99; S, 2.87.
Calc. for C41H38N5F6O8PRuS: C, 48.91; H, 3.80; N, 6.96; S, 3.18. High resolution ESI-MS: m/z 862.1479
(calc. for C41H38N5O8RuS+m/z 862.1490).(4,4’-Dicarboxy-2,2’-bipyridine-κ2N,N’)(4-carboxy-4’-
carboxy-late-2,2’-bipyridine-κ2N,N’)[N-((thiophen-2-yl-κC3)methyli-den)aniline-κN]ruthenium(II)
(7). Synthesized analogously to [26] and purified as follows. The residue was suspended in hot DMF, in-
soluble fraction was removed by ultracentrifugation (12400 rpm, 1.5 min). The solvent was removed on a
rotary evaporator. The residue was dissolved in hot DMF and precipitated with diethyl ether. 18.5 mg
(70 %) of dark red solid was obtained. δН (400 MHz, CD3OD+NaOD+D2O, 27 °С): 6.45 (d, 1Н, 3JH,H
=
4.6 Hz), 6.50 (m, 2Н), 6.89-6.90 (m, 3Н), 7.63 (d, 1Н, 3JH,H = 6.0 Hz), 7.66 (br s, 1H), 7.68 (d, 1Н, 3JH,H
3
= 4.6 Hz), 7.73 (d, 1H, 3JH,H = 5.5 Hz), 7.82 (d, 1Н, 3JH,H = 6.0 Hz), 7.86 (d, 1Н, JH,H = 5.5 Hz), 7.91 (d,
1Н, 3JH,H = 6.0Hz), 8.62 (s, 1Н), 8.64 (d, 1Н, 3JH,H = 5.5 Hz), 8.69 (s, 1Н), 8.71 (s, 1Н), 8.88 (s, 2Н). δC
(150 MHz, CD3OD+NaOD+D2O, 27 °С): 123.10; 123.25; 123.37; 123.61; 126.80; 126.89; 127.24;
129.82; 134.64; 136.04; 139.31; 145.04; 145.38; 146.12; 146.73; 149.93; 151.53; 152.82; 153.98; 155.71;
156.62; 158.43; 159.22; 159.45; 166.38; 171.00; 171.15; 171.29; 171.48; 211.14. Found: C, 54.38; H
3.09; N 8.97; S 3.86. Calc. for C35H23N5O8RuS: C, 54.26; H, 2.99; N, 9.04; S, 4.14. High resolution ESI-
MS: m/z 776.0402 (calc. for C35H23N5O8RuS, [M+H]+, m/z 776.0392).
(4,4’-Dicarboxy-2,2’-bipyridine-κ2N,N’)(4-carboxy-4’-carboxy-late-2,2’-bipyridine-κ2N,N’)[2-
(thiophen-2-yl-κC3)pyridine-κN]ruthenium(II) (8). Synthesized analogously to [26] and purified as
follows. The residue was suspended in diethyl ether with addition of a small amount of DCM and MeOH,
filtered and dried. 26.4 mg (83%) of dark red solid was obtained. δН (400 MHz, CD3OD+NaOD+D2O, 27
°С): 6.29 (d, 1Н, 3JH,H = 4.8 Hz), 6.75-6.79 (m, 1Н), 7.37 (m, 1Н), 7.42 (d, 1Н, 3JH,H = 4.7 Hz), 7.56-7.60
(m, 2Н), 7.64-7.67 (m, 2Н), 7.69-7.71 (dd, 1Н, 3JH,H = 6.0, 1.6 Hz), 7.83-7.88 (m, 3Н), 7.93 (d, 1Н, 3JH,H
= 6.0 Hz), 8.04 (d, 1Н, 3JH,H = 5.9 Hz), 8.85 (s, 1H), 8.86 (s, 1H), 8.92 (s, 1H), 9.00 (s, 1H). δC (150
7

ACCEPTED MANUSCRIPT
MHz, CD3OD+NaOD+D2O, 27 °С): 119.08; 120.81; 123.35; 123.49; 123.89; 126.61; 127.67; 129.88;
134.32; 136.81; 137.46; 144.85; 145.23; 146.15; 147.57; 150.30; 151.22; 151.38; 151.83; 155.79; 157.42;
158.52; 159.11; 159.38; 161.62; 164.96; 171.09; 171.13; 171.31; 171.42; 195.93. Found: C, 52.69; H,
3.19; N, 9.29; S, 3.91. Calc. for C33H21N5O8RuS: C, 52.94; H, 2.83; N, 9.35; S, 4.28. High resolution ESI-
MS: m/z 750.0235 (calc. for C33H21N5O8RuS, [M+H]+m/z: 750.0196).
2.3. Voltammetric study
Electrochemical studies were carried out using an Autolab 100N potentiostat/galvanostat connected to a
PC. Voltammograms were recorded using cyclic voltammetry (CV) in DMF solution containing 0.05 M
Bu4NBF4 as a supporting electrolyte at 20°C in a 10 mL one-compartment electrochemical cell. A sta-
tionary Pt disk (d = 3.2 mm) was used as a working electrode; Pt wire served as an auxiliary electrode; a
saturated Ag/AgCl/KCl electrode was used as a reference electrode. The potential of the Fc/Fc+ redox
couple in the experimental conditions was +0.49 V vs. Ag/AgCl/KCl. Oxygen was removed from the cell
by purging with dry argon.
2.4. Calculations
The energy levels Е(LUMO+m) and Е(HOMO-n) of model cations of the complexes 7 and 8 were opti-
mized in DFT/BP calculations using an STO basis set of triple-ζ quality. The equilibrium structures of 7
and 8 were converged with accuracy of 0.01Å.All calculations were performed using the Amsterdam
Density Functional (ADF) program [27,28] version 2013 [29].
2.5. Cell fabrication
Photoanodes were assembled as follows. TCO22-15 fluorinated tin oxide coated 2.9×2.9 cm2 glasses (So-
laronix®, specific surface resistivity ~15 Ω/sq) were purified by aging in sulfochromic mixture followed
by ultrasonication in organic solvents (isopropanol and acetone) and distilled water, and then dried at
8

ACCEPTED MANUSCRIPT
50°C in air. The application of Ti-Nanoxide paste (D/SP, Solaronix®), comprising titania powder and α-
terpineol as a binding agent, was performed by the standard «doctor blade» technique using a stencil with
a 2.0 × 2.0 cm (~ 90 µm depth) square hole. After the paste application, raw photoanodes were dried at
50°C in air, and then calcined in a muffle furnace at 450°C for 1 hour (heating rate 3 °C min-1, in air). The
thickness of obtained titanium dioxide film was about 15 µm, and it was determined by Alpha-Step D-
100 profilometer (KLA-Tencor). Sensitizing of titanium dioxide was performed by soaking of pho-
toanodes in ~5·10-4 M methanol solutions of dyes 7 or 8 for 24 hours. A three-electrode photoelectro-
chemical cell PECC-2 (Zahner) cell was used for the photoanode potential measurements. The pho-
toanode served as a working electrode and a platinum wire with the surface area of 5 cm2 was used as an
auxiliary electrode, a silver wire was used as a reference electrode. The voltammetric measurements were
performed with IPC Pro MF potentiostat under AM 1.5 global one sun of illumination (100 mW cm-2)
provided by a solar simulator (Newport 96000). The illumination power at different distances was deter-
mined with Nova apparatus (OPHIR-SPIRICON Inc.). Current-voltage characteristic of DSSCs and pho-
tocurrent density at the short-circuit voltage were performed by the two-electrode scheme. The transients
of photoanode potential and photocurrent density at the short-circuit voltage were measured under irradia-
tion and in dark condition. The photoanode area was about of 1.0 cm2. The illuminated photoanode area
was restricted by a mask 0.196 cm2. The illumination was performed from the side of TiO2 photoanode
with the adsorbed dye.
3. Results and discussion
3.1. Ligands preparation and cyclometallation
We investigated the already known simplest thiophene-containing compounds that could undergo cy-
cloruthenation. The major advantage of the chosen derivatives 1 and 2 was their ability to be modified
easily with substituted anilines in 1 or pyridines in 2. Compound 1 was synthesized by condensation reac-
9

ACCEPTED MANUSCRIPT
tion with aniline [23], and compound 2 was synthesized by sequential thiophene bromination [24] and
Kumada coupling [25] with 2-bromopyridine in good yields. It is known that benzene in a η-coordinated
ruthenium precursor is easily substituted by anionic ligands [30]. Cyclometallation reaction proceeded
smoothly with [Ru(C6H6)Cl2]2 as a ruthenium source, but the yields of 3 and 4 after purification by col-
umn chromatography were quite moderate (Scheme 1).
Scheme 1. Synthesis of cyclometallated thiophenes. Conditions: a) EtOH, aniline, reflux; b)
[Ru(C6H6)Cl2]2; NaOH, KPF6; MeCN, 45°C, 48 h; c) H2O2, HBr, Et2O; -17-15°C; d) Mg, Et2O, reflux; e)
2-bromopyridine; Ni(dppp)Cl2; Et2O, rt.
In general, complexes 3 and 4 are light/oxygen sensitive powders, while they are more stable than 2-
phenylpyridine analogue. There are two possible ways to prepare target dyes: the direct reaction with salt
of dicarboxylic acid [31] or the sequential formation of ester followed by its hydrolysis [30]. To be able to
perform reliable experimental spectroscopic and electrochemical study, the second method was chosen.
Esters have much better solubility in organic solvents as compared to free acid complexes, even if the lat-
ter are in the sodium salt form.
Scheme 2. Preparation of ortho-metallated dyes. Conditions: a) diethyl 2,2’-bipyridine-4,4’-
dicarboxylate; EtOH, reflux, 3 h; b) DMF, Et3N, H2O, reflux, 15 h.
Ester moieties can be removed by alkaline hydrolysis with a DMF/water/Et3N mixture [30] (Scheme 2).
The structures of the ester 5 and 6 were proved by NMR (all proton and carbon chemical shifts were as-
signed by 2D-NMR techniques) and HRMS-ESI. The data obtained for free acids were in agreement with
that for the parent esters. Besides elemental analysis the absence of PF6 counteranion was also confirmed
19
by absence of fluoride signals in F-NMR. This fact allowed us to consider compounds 7 and 8 having
betain structures [16].
3.2. Crystallography
10

ACCEPTED MANUSCRIPT
The structures of complexes 3 and 4 were unambiguously proved by means of X-ray diffraction (Fig. 1).
Figure 1. Structures of [RuL(MeCN)4]+ in 3 (a) and 4*Et2O (b) (H atoms at carbon atoms and counteri-
ons are omitted for clarity, thermal ellipsoids are given with 30% probability).
These compounds adopted configurations of ligands typical to the previously known ruthenium(II) com-
plexes.[32] Complex 3 crystallizes in the P-1 space group. There were two crystallography independent
molecules (Z = 4) which had identical coordination geometry, so we described one of them. The rutheni-
um atom in 3 has distorted octahedral coordination geometry (Fig. 1a). The Ru1 atom coordinates four
nitrogen atoms of MeCN molecules (Ru1-N 2.013(4)-2.135(4) Å), nitrogen and σ-bound carbon atoms of
chelate anion (Ru1-N1 2.091(4), Ru1-C3 2.022(5) Å). The angle between the planes of C6 and C4S-C=N
fragments in L-H is 51.9°. The selected bond lengths (Å) and angles (degrees) in 3 were as follows: S1-C
1.708(6), 1.741(5), C2-C3 1.421(7), C3-C4 1.385(7), N1-C5 1.299(6), N1-C6 1.442(6), N-C (MeCN)
1.138(7)-1.130(7), C1-S1-C4 90.3(3), C2-C3-C4 109.8(5), C5-N1-C6 117.4(4), C3-Ru1-N1 79.4(2), C3-
Ru1-N2 90.6(2), C3-Ru1-N3 173.9(2), C3-Ru1-N4 90.7(2), C3-Ru1-N5 94.8(2).
Complex 4*Et2O crystallizes in the P-1 space group. The ruthenium atom in 4*Et2O has distorted octa-
hedral coordination geometry (Fig. 1b). The Ru1 atom in 4*Et2O coordinates four nitrogen atoms of
MeCN molecules (Ru-N 2.006(2)-2.048(2) Å), nitrogen and σ-bound carbon atoms of 2-(2-
thienyl)pyridine anion (Ru-N 2.060(5) and 2.037(12), Ru-C 2.028(5) and 2.14(3) for two forms of disor-
dered ligand, correspondingly). The selected bond angles (deg) in 4*Et2O are as follows: C1-Ru1-N5
175.0(2), C1-Ru1-N1 78.9(2), N1-Ru1-N4 172.8(2), N2-Ru1-N3 176.99(8), C1A-Ru1-N4 176.9(9),
C1A-Ru1-N1A 85.2(9), N1A-Ru1-N5 173.4(8).
3.3. Electrochemical study, absorption spectra and DFT/BP simulations
11

ACCEPTED MANUSCRIPT
To get estimation of the HOMO – LUMO levels as well as of the stability of the oxidized and reduced
species, electrochemical investigation of compounds 5-8 was performed. Half-wave potential values are
listed in Table 1.
Table 1. Half-wave potentials (E1/2) and direct/reverse peak separation values (∆Ep) for 5-8 complexes
(0.5mM, Pt, 0.05M Bu4NBF4, DMF, 100 mV/s, vs. Ag/AgCl/KClsat.)
*- the peak potential is indicated; the process is irreversible
As follows from the data given in Table 1, all the compounds are redox active in cathodic and anodic area
(Fig.2).
Figure 2. CV curves of 0.5 mM solutions of 5, 6 and 8 (Pt, 0.05M Bu4NBF4, DMF, 100 mV/s, vs.
Ag/AgCl (KClsat.)
Oxidation of 5-8 is completely reversible (direct/reverse peak separation values are of 60-70 mV at a po-
tential scan rate of 100 mV/s; the corresponding peak current ratios are close to unity), diffusion-
controlled and metal-centered for all compounds. The peak potential values are typical for Ru2+/Ru3+
transformation commonly observed in the potential region of 0.5-0.9 V (vs. Ag/AgCl/KCl) for various
pyridine-containing ruthenacycles [16,33–35]. Electrochemical reduction of complexes 5-8 is multi-
electron, stepwise and ligand-centered.
More likely, the first two reduction steps correspond to the consecutive reduction of the bipyridine lig-
ands as it has been previously observed for ruthenacycles with the similar coordination environment
[16,35]. The third electron might be expected to be transferred to the orbital mainly located on the thio-
phene-based moiety.
Electrochemical data obtained for compounds 5-6 and 7-8 are given in Table 2 allowing the comparison;
the CV curves for complexes 5-6and 8 are presented in Fig.2. It should be mentioned that for fabrication
12

ACCEPTED MANUSCRIPT
of DCCSs, the ruthenium(II) dye is based on mesoporous titania forming Ti-O bond, which is close to the
covalent one. Ester compounds 5-6 have covalent C-O-C bond, contrary to 7-8, which have ionized car-
boxy group. Thus, both types of compounds should be considered as models since they are not applied
directly in the photovoltaic cell. Meanwhile, the peripheral ester or carboxy group can be expected to
have not really much influence on the electronic properties of the redox centers. This follows, for in-
stance, from the UV-Vis spectral data which is almost the same for both series (see below). One can ex-
pect that the electrochemical potential values for 5-6 and 7-8 compounds also should not differ signifi-
cantly. The problem is in the presence of protons in compounds 7-8, which provides an additional cathod-
ic process of hydrogen evolution. The corresponding broad peak is superimposed to pyridine- and thio-
phene-centered redox processes, which are of interest for characterization of the material and occur in the
close potential region resulting in a poor peak resolution. The same situation was also observed in [36].
The other problem is much lower solubility of complexes 7-8 as compared to their 5-6 counterparts. The-
se two factors significantly degraded the accuracy of the measurement of the potential values for the re-
dox-transitions of interest for complexes 7-8, as compared to 5-6 models, which exhibit three consecutive
well distinguished one-electron redox transformations in the cathodic area (see Fig. 2). Thus, the CV
curves for complexes 7-8 are much less informative (given in Fig.2. and in SI, Figs. S15 and S16) than
that for 5-6. The electrochemical potential gaps (EOx – ERed values) measured for the compounds under
investigation are listed in Table 2. Evidently, the experimental estimation is influenced by the solvation
energy, as compared to the corresponding DFT calculated values (see below). However, for the ester de-
rivatives 5 and 6, in which the ligand-centered reduction is not influenced by the impact of hydrogen evo-
lution impeding the correct estimation of the formal potential values, the coincidence of the electrochemi-
cal, spectral and DFT-calculated energy gaps was quite reasonable. The UV-vis absorption spectra of the
prepared dyes in methanol solution are presented in Fig. 3 and the corresponding data are summarized in
Table 2.
13

ACCEPTED MANUSCRIPT
Figure 3. Comparison of experimental UV-vis spectra of compounds (a) 5 (solid) and 6 (dashed) (3.4*10-
5 M solution in methanol) and (b) 7 (solid) and 8 (dashed) (1*10-4 M solution in methanol). For calculated
spectra of 7 and 8 and alignment of the energy levels see Figure S17 and footnotes to Table S2. Tangent
line on Fig. 3a illustrates an estimation of optical band gap ꢁEoptical (Table 2).
Table 2. UV-vis spectral data and band gap energies. Molar extinction coefficients (M-1cm-1) for selected
maxima wavelengths in visible range are given in parentheses. The ꢁEoptical was obtained by extrapolation
of the spectral data (Fig. 3). ꢁEredox is the difference between the first oxidation and the first reduction
potentials of the complexes. ꢁEKT was calculated using the Koopmans’ theorem: ꢁEopticalKT as E(HOMO-
1) – E(LUMO+1) and E(HOMO-2) – E(LUMO+1) gap and ꢁEredoxKT as E(HOMO) – E(LUMO) and
E(HOMO) – E(LUMO+1) gap.
All dyes exhibited absorption bands at two distinct spectrum regions. The intense absorption (ε> 50000
M-1cm-1) occurs at around 270-350 nm that can be attributed to the localized π→π* electron transition
located on the ligands. Other broad absorption bands are in the visible region of 370-650 nm with molar
absorption coefficient of around 10000 M-1cm-1 are likely to originate from metal-to-ligand charge trans-
fer (MLCT) transition, which is common for similar Ru(II) compounds.[37–39] As it is well known,
MLCT in complexes arise from transfer of electrons from molecular orbitals (MO) with metal-like char-
acter to those with ligand-like character, therefore we used data from the spectra by extrapolation of the
absorption curve to evaluate the optical band gap of the studied complexes.[40]Having experimental
electrochemical and UV-Vis data, it was interesting to compare electrochemical and optical band gap
values. there is a relationship described generally by the equation[41]
∑
+ ꢁ ꢀꢁ
ꢁEoptical = ꢁEredox
where term ai collects factors not directly deriviable from either of these measurable variables. Various
factors contributing to this term have been discussed earlier [42]. In particular, this term must include a
14

ACCEPTED MANUSCRIPT
correction for the solvation energy differences between different species, and a term to account for the
differences in the orbital energies of the species. As it can be seen from comparison of the results of opti-
cal and electrochemical measurements (Table 2), this term varies from 0.14 to 0.49 eV for studied com-
plexes. These values are in agreement with literature data [41,42].
The calculated distribution of the density of electronic states (DOS) for models 7 and 8 (considered as
monovalent cations with fully protonated carboxylic groups) are given in Fig. 4.
Figure 4. Density of electronic states (DOS) and the energy levels of 7 (a) and 8 (b) sensitizers. The val-
ues of ꢁEopticalKT and ꢁEredoxKT are given in the Table 2.
The optical and electrochemical gaps ꢁEoptical and ꢁEredoxwere calculated using the Koopmans’ theorem
as
ꢁE11optical= E(LUMO+1) – E(HOMO-1)
ꢁE12optical= E(LUMO+1) – E(HOMO-2)
ꢁEm,nredox = E(LUMO+m) – E(HOMO-n)
where n = 0, 1, …, 15; m = 0, 1, …, 14
The calculated difference between the optical and the electrochemical band gaps is consistent with the
experimentally obtained values (Table 2). For the complexes 7 and 8 the ꢁEredoxvalues aresmaller than
those of the compounds5 and 6 by0.33 and 0.38 V respectively. These values are consistent with
E(HOMO)-E(LUMO) gap change in SCN-containing 2,2’-bipyridine Ru(II) complexes calculated previ-
ously by replacement of two of the carboxylic anchoring groups by 4,4’-methyl groups [43]. For com-
pound 7, the HOMO energy was estimated to be equal-7.49 eV, whereas for 8 it was equal to about -7.47
eV. The HOMO energy difference for these two complexes agrees well with the experimental trend of the
oxidation potentials (Table 1). Fig. S1 and Table S1 show that the theoretical results match with the ex-
15

ACCEPTED MANUSCRIPT
perimental data well. Our approach allows to partially compensate for the possible inaccuracy of the DFT
methods in treatment of the electron correlation effects in the transition metal compounds [44].
For compounds 7 and 8, the three weak bands located at 563 and 570, 498 and 495, 418 and 409, and two
strong bands located at 316 and 313, 305 and 304 nm in the UV-vis range, were observed experimentally.
The SCN-containing 2,2’-bipyridine Ru(II) complexes have similar UV-vis spectra [32,41–43]. We fo-
cused on the analysis of transitions that generated the peaks of longer wavelength absorptions for both
complexes. According to the electronic structure calculations of cations 7 and 8 (Fig. 5), the two most
long-wave bands are to be assigned to the n→π* absorption of Ru lone pairs delocalized on the 2-
thiophene substituent. In the green range of cation 7 and in the yellow of the 8 the HOMO→LUMO+6
transition is active (Table S2, Fig. S17).
Figure 5. Electronic transitions of cations defined absorption in visible range of spectrum for 7 (a) and 8
(b) dyes: HOMO→LUMO+6 and HOMO-1 and HOMO-2→LUМО+1 being the most long-wave bands
in the green (7) and the yellow (8) ranges, HOMO-1 and HOMO-2→LUМО+6 in the blue range. The
band gap HOMO-1 (HOMO-2) – LUMO+1 define ꢁEoptical, is consistent with the spectral data. The band
gap HOMO – LUMO (LUMO+1) defines ꢁEredox, is consistent with the electrochemical data. The forbid-
den HOMO – LUMO transition is not observed in the optical spectra. The numerical designations show
the Ru contribution to corresponding MO(%), elucidating more absorption intensity of dye 8 in yellow
and blue ranges.
In the blue range of both cations, HOMO-1→LUМО+1 and HOMO-2→LUМО+1 transitions are active.
In the yellow range of spectrum of cation 8, the absorption of the Ru1-N5π-bond of the 2-thiophene ligand
corresponding to HOMO-3→LUMO+1 transition appears. HOMO→LUMO+7 absorption is observed in
the blue range of the spectrum. The HOMO→LUMO+6 transition energy of cation 7 coincides with the
maximum of the most long-wave band. In cation 8, this transition is hypsochromically shifted by the val-
16

ACCEPTED MANUSCRIPT
ue of 65 nm. The similar shift of 68 nm was also found for the HOMO→ LUMO+7 transition. The ab-
sence of the same shift for the HOMO→LUMO+6 of cation 7 can be explained by the absence of the sol-
vent effect because of equivalence of the bond conjugation chain length of atoms, on which these MOs
are delocalized (HOMO: Ru1-N5-C24-C1a-C22-S21 and LUMO+6: Ru1-N5-C25-C1a-C22-S21, Fig. 5, Fig.
S17) and the same Mulliken atomic charges values on the C22-S21 atoms (Table S3), suitable for the for-
mation of the complex with solvent СН3ОН [44]. Thus, the influences of the solvent on HOMO and LU-
MO+6 energies of cation 7 are equal. However, in cation 8, the bond conjugation chain of atoms on
which theLUMO+6 is delocalized is longer than that for the HOMO. Therefore, the solvent effect results
into the decrease of the HOMO-LUMO+6(7) gap and bathochromic shift of this transition. The stabiliza-
tion of the ground state of the transition metal compounds, as a conjugation effect between the
metallaallene fragment and the co-planar phenyl ring [48], and abrupt change of dye properties by the de-
crease of the bond conjugation chain length were reported previously [49–51]. In cations 7 and 8, the dif-
ference between experimental and calculated values of wavelengths of the blue range transitions of 41
(44) and 67 (82) nm may be explained in the same way. The values of these transitions depend on the
lone pair localization on Ru, as well as on the Mulliken atomic charges distributions (Fig. 5 and Table
S3). The violet range band (Fig.S16, Table S1) is due to
(i) Absorption of: the 4,4’-dicarboxy-2,2’-bipyridine ligands with electron transitions from the HOMO-12
(7) or HOMO-11 (8) to the LUMO and from the HOMO-13 (7) or HOMO-10 (8) to the LUMO+1;
(ii) Absorption of the Ru1-N5 π-bond of 2-thiophene ligand with electron transition between the HOMO-3
and LUMO+6;
(iii) In cation 7, the HOMO-12→LUMO+1, HOMO-13→LUMO transitions are active, too. In cation 8
there is also the Ru lone pairs’ absorption corresponding to the transitions from the HOMO or HOMO-2
to LUMO+7. The absence of the similar LUMO of appropriate energy in cation 7 explains the higher in-
17

ACCEPTED MANUSCRIPT
tensity of the violet band of the 8 dye. Strong absorption in near UV is mostly due to π→π* transitions of
the 4,4’-dicarboxy-2,2’-bipyridine ligands. All electronic transitions in the visible and NIR ranges are
forbidden by the well-known selection rules; therefore, the intensity of absorption is weak. Fig. 5 shows
electronic transitions that define the long-wave absorption bands. These n→π* transitions are almost for-
bidden and have low intensity. In cation 8, comparing with 7, the contribution of Ru in HOMO and HO-
MO-2 is smaller, and it is larger in HOMO-1. The HOMO-2 (8) is more delocalized and HOMO-1 (8) is
less delocalized on the anionic ligand. The symmetry change from n to π increases probability of transi-
tion and its intensity. This determines higher intensity and a small bathochromic shift in the green range,
and a hypsochromic shift of bands in the blue range of UV-vis spectrum of the 8 dye in comparison to
dye 7.
3.4. Photovoltaic performance
After injecting the electron into the conduction band of TiO2, regeneration of the dye is an important pro-
cess within DSSC. To date, iodide/triiodide has been the most efficient redox couple used for high-
performance DSSCs. The physical diffusion of iodide/triiodide in the electrolyte is often a limiting pro-
cess. We used the method of photopotential and photocurrent transients to study the dye regeneration pro-
cess. The results are presented in Fig. 6.
Figure 6. (a) Photoanode potential versus time for the DSSC with the anodes stained with 7 (dash line)
and 8 (solid line) dyes in dark condition and under illumination with power of 100 mW cm-2. (b) Photo-
current density transient at the short-circuit voltage for the DSSCs with the anodes stained with 7 (dash
line) and 8 (solid line) dyes in dark condition and under illumination with power of 100 mW cm-2.
The potential rapidly shifted towards negative values upon illumination, which indicates that there is
high-rate accumulation of injected electrons in the TiO2 conduction band. This suggests that the excited
level of the dyes lay higher in energy than the bottom of the conduction band, and that there is a sponta-
18

ACCEPTED MANUSCRIPT
neous process of charge transfer from the excited state of dye into the band. It is to be noted that there is
no degradation of photopotential over the time. Fig. 6 shows the photocurrent response to the light on-off
sequence. One can see that the DSSCs with 7 and 8 dyes exhibit the current remaining constantly. These
results indicate that in the case of the iodide/triiodide system there is no diffusion limitation on the regen-
eration of the dyes, according to Zakeeruddin [52]. The photovoltaic performance is shown in Fig. 7 and
Table 3.
Figure 7. Current-voltage characteristic of DSSCs with the photoanodes made using the sensitizers 7
(dash line) and 8 (solid line) upon illumination (100 mW cm-2).
Table 3. DSSC performance data.
*Solaronix® N535 – bis(4,4’-dicarboxy-2,2’-bipyridine-κ2N,N’)ruthenium(II) cis-dithiocyanate
It was found that DSSCs have similar characteristics. However, a comparison of I-V characteristics shows
that dye 7 photocurrent density and open circuit voltage exceed those of the 8 dye. For comparison, the
corresponding parameters are also shown for the standard N535 dye and forthe known cycloruthenated
thiophenes(see Table 3). From the photovoltaic data it is apparent that the new synthetically facile ruthe-
nium(II) complexes are promising dyes for DSSCs overall conversion efficiency improvement. The study
of various 7 and 8 derivatives will be continued to find out the optimal configuration of the substituents
for better DSSCs performance.
Further improvement in the efficiency DSSC with the 7 and 8 sensitizers is possible by choosing of dop-
ing of titania with rare-earth metal complexes [53], tuning of electrolyte composition [4] and choosing
optimal configuration of substituents. The work directed towards this goal is in progress.
4. Conclusions
19

ACCEPTED MANUSCRIPT
Two bis-heteroleptic ruthenium(II) dyes containing different types of the simplest cyclometallated thio-
phene moieties were synthesized and characterized by NMR, UV-vis spectroscopy and cyclic voltamme-
try. Ortho-metallation of thiophenes was proved by means of single crystal X-ray data. To get an insight
of the influence of the substituents in the thiophene ligand on the electronic structure of the complexes,
DFT calculations of the energy and localization of their molecular orbitals were performed. Theoretical
and experimental estimations of the energy band gaps are in good agreement. Synthesized complexes are
of interest for the fabrication of dye-sensitized solar cells. The estimation of an overall conversion effi-
ciency range of new sensitizers gave values from 4.1 to 5.3% under standard AM 1.5 sunlight, which are
only slightly lower than that of the standard ruthenium N535 dye exhibiting 6.1% PCE but higher than of
other known cycloruthenated thiophenes.
Acknowledgements
The study was supported by RFBR (project 13-03-12415), RSF (project 14-23-00176), and by Tomsk
State University Competitiveness Improvement Program. We thank Prof. P. Dyachkov and Dr. D.
Krut’ko for fruitful and useful discussions.
Appendix A. Supplementary data
The following is the supplementary data related to this article:
References
[1] IEA-PVTS. PVPS Report Snapshot of Global PV 1992-2013. 2014.
[2] Green MA, Emery K, Hishikawa Y, Warta W. Solar cell efficiency tables (version 35). Prog
Photovoltaics Res Appl 2010;18:144–50. doi:10.1002/pip.974.
[3] Optimal Materials and Deposition Technique Lead to Cost-Effective Solar Cell with Best-Ever
Conversion Efficiency. Report of the National Renewable Energy Laboratory, 2012.
20

ACCEPTED MANUSCRIPT
[4] Mathew S, Yella A, Gao P, Humphry-Baker R, Curchod BFE, Ashari-Astani N, et al. Dye-
sensitized solar cells with 13% efficiency achieved through the molecular engineering of porphyrin
sensitizers. Nat Chem 2014;6:242–7. doi:10.1038/nchem.1861.
[5] Yella A, Lee H-W, Tsao HN, Yi C, Chandiran AK, Nazeeruddin MK, et al. Porphyrin-sensitized
solar cells with cobalt (II/III)-based redox electrolyte exceed 12 percent efficiency. Science (80- )
2011;334:629–34. doi:10.1126/science.1209688.
[6] Park N-G. Perovskite solar cells: an emerging photovoltaic technology. Mater Today 2015;18:65–
72.
[7] Nazeeruddin MK, Kay A, Rodicio I, Humphry-Baker R, Mueller E, Liska P, et al. Conversion of
light to electricity by cis-X2bis(2,2’-bipyridyl-4,4’-dicarboxylate)ruthenium(II) charge-transfer
sensitizers (X = Cl-, Br-, I-, CN-, and SCN-) on nanocrystalline titanium dioxide electrodes. J Am
Chem Soc 1993;115:6382–90. doi:10.1021/ja00067a063.
[8] Bomben PG, Robson KCD, Koivisto BD, Berlinguette CP. Cyclometalated ruthenium
chromophores for the dye-sensitized solar cell. Coord Chem Rev 2012;256:1438–50.
doi:10.1016/j.ccr.2012.02.005.
[9] Yin J-F, Velayudham M, Bhattacharya D, Lin H-C, Lu K-L. Structure optimization of ruthenium
photosensitizers for efficient dye-sensitized solar cells – A goal toward a “bright” future. Coord
Chem Rev 2012;256:3008–35. doi:10.1016/j.ccr.2012.06.022.
[10] Chen C-Y, Wang M, Li J-Y, Pootrakulchote N, Alibabaei L, Ngoc-le C-H, et al. Highly efficient
light-harvesting ruthenium sensitizer for thin-film dye-sensitized solar cells. ACS Nano
2009;3:3103–9. doi:10.1021/nn900756s.
[11] Zakeeruddin SM, Nazeeruddin MK, Humphry-Baker R, Grätzel M, Shklover V. Stepwise
Assembly of Tris-Heteroleptic Polypyridyl Complexes of Ruthenium(II). Inorg Chem
1998;37:5251–9. doi:10.1021/ic980357s.
21

ACCEPTED MANUSCRIPT
[12] Bessho T, Yoneda E, Yum J-H, Guglielmi M, Tavernelli I, Imai H, et al. New paradigm in
molecular engineering of sensitizers for solar cell applications. J Am Chem Soc 2009;131:5930–4.
doi:10.1021/ja9002684.
[13] Pogozhev D V., Bezdek MJ, Schauer PA, Berlinguette CP. Ruthenium (II) Complexes Bearing a
Naphthalimide Fragment: A Modular Dye Platform for the Dye-Sensitized Solar Cell. Inorg Chem
2013;52:3001–6.
[14] Brown DG, Schauer PA, Borau-Garcia J, Fancy BR, Berlinguette CP. Stabilization of ruthenium
sensitizers to TiO2 surfaces through cooperative anchoring groups. J Am Chem Soc
2013;135:1692–5. doi:10.1021/ja310965h.
[15] Schulze B, Escudero D, Friebe C, Siebert R, Görls H, Sinn S, et al. Ruthenium(II) photosensitizers
of tridentate click-derived cyclometalating ligands: a joint experimental and computational study.
Chem - A Eur J 2012;18:4010–25. doi:10.1002/chem.201103451.
[16] Bomben PG, Koivisto BD, Berlinguette CP. Cyclometalated Ru complexes of type [Ru(II)(N--
N)(2)(C--N)](z): physicochemical response to substituents installed on the anionic ligand. Inorg
Chem 2010;49:4960–4671. doi:10.1021/ic100063c.
[17] Cuesta L, Urriolabeitia EP. Cyclometallation of Heterocycles: A Reliable Strategy for Selective
Functionalization. Comments Inorg Chem 2012;33:55–85. doi:10.1080/02603594.2012.700971.
[18] Li X-H, Shi Z, Wang L, Cheng X, Li C, Zhang A. Chromogenic mercury ions recognition of a new
ruthenium(II) complex with cyclometalated 2-(2-thienyl)pyridine in CH3CN-aqueous system.
Inorg Chem Commun 2013;29:175–8. doi:10.1016/j.inoche.2012.12.004.
[19] Li C-Y, Su C, Wang H-H, Kumaresan P, Hsu C-H, Lee I-T, et al. Design and development of
cyclometalated ruthenium complexes containing thiophenyl-pyridine ligand for dye-sensitized
solar cells. Dye Pigment 2014;100:57–65. doi:10.1016/j.dyepig.2013.07.014.
[20] Vriamont N, Govaerts B, Grenuoilles P, de Bellefon C, Riant OXX. Design of a genetic algorithm
22

ACCEPTED MANUSCRIPT
for the simulated evolution of a library of asymmetric transfer hydrogenation catalysts. Chem - A
Eur J 2009;15:6267–78.
[21] Gillaizeau-Gauthier I, Odobel F, Alebbi M, Argazzi R, Costa E, Bignozzi CA, et al. Phosphonate-
Based Bipyridine Dyes for Stable Photovoltaic Devices. Inorg Chem 2001;40:6073–9.
doi:10.1021/ic010192e.
[22] Tsedilin AM, Fakhrutdinov AN, Eremin DB, Zalesskiy SS, Chizhov AO, Kolotyrkina NG, et al.
How sensitive and accurate are routine NMR and MS measurements? Mendeleev Commun
2015;25:454–6. doi:10.1016/j.mencom.2015.11.019.
[23] Wang D-L, Hwang W, Liang L, Wang L-I, Lee L, Chiang MY. Reaction of a Thienyl Schiff Base
with Diiron Nonacarbonyl: Characterization and Structures of [ꢀ- N -(((2,3-η 1ꢂ:η 2 )-5-Methyl-2-
thienyl)methyl)-η 1ꢂ:η 1 ( N )- anilino]hexacarbonyldiiron and [ꢀ- N -((Anilino(2-
thienyl)methyl)((2,3-η 1ꢂ:η 2 )-2-thienyl). Organometallics 1997;16:3109–13.
doi:10.1021/om961092x.
[24] Keegstra MA, Brandsma L. Convenient high-yield procedures for 2-bromothiophene and 2,5-
dibromothiophene. Synthesis (Stuttg) 1988:890–1.
[25] Novak I, Ng S, Mok C, Huang H, Fang J, Wang KK. New organic polymer precursors: synthesis
and electronic structure of thienylpyridines and thienylethynylpyridines. J Chem Soc Perkin Trans
2 1994:1771. doi:10.1039/p29940001771.
[26] Bomben PG, Gordon TJ, Schott E, Berlinguette CP. A trisheteroleptic cyclometalated Ru(II)
sensitizer that enables high power output in a dye-sensitized solar cell. Angew Chemie - Int Ed
2011;50:10682–5. doi:10.1002/anie.201104275.
[27] Guerra CF, Snijders JG, te Velde G, Baerends EJ. Towards an order-N DFT method. Theor Chem
Acc 1998;99:391–403. doi:10.1007/s002140050353.
[28] te Velde G, Bickelhaupt FM, Baerends EJ, Guerra CF, van Gisbergen SJA, Snijders JG, et al.
23

ACCEPTED MANUSCRIPT
Chemistry with ADF. J Comput Chem 2001;22:931–67. doi:10.1002/jcc.1056.
[29] ADF® molecular modeling suite 2013.
[30] Bomben PG, Thériault KD, Berlinguette CP. Strategies for Optimizing the Performance of
Cyclometalated Ruthenium Sensitizers for Dye-Sensitized Solar Cells. Eur J Inorg Chem
2011;2011:1806–14. doi:10.1002/ejic.201001345.
[31] Bomben PG, Robson KCD, Sedach P, Berlinguette CP. On the viability of cyclometalated Ru(II)
complexes for light-harvesting applications. Inorg Chem 2009;48:9631–43.
doi:10.1021/ic900653q.
[32] Cuesta L, Maluenda I, Soler T, Navarro R, Urriolabeitia EP. Novel thiophene-based
cycloruthenated compounds: synthesis, characterization, and reactivity. Inorg Chem 2011;50:37–
45. doi:10.1021/ic101946n.
[33] Yap JSL, Li B Bin, Wong J, Li Y, Pullarkat SA, Leung P-H. Development of a novel chiral
palladacycle and its application in asymmetric hydrophosphination reaction. Dalt Trans
2014;43:5777. doi:10.1039/c4dt00062e.
[34] Djukic J-P, Sortais J-B, Barloy L, Pfeffer M. Cycloruthenated Compounds - Synthesis and
Applications. Eur J Inorg Chem 2009;2009:817–53. doi:10.1002/ejic.200801016.
[35] Robson KCD, Bomben PG, Berlinguette CP. Cycloruthenated sensitizers: improving the dye-
sensitized solar cell with classical inorganic chemistry principles. Dalt Trans 2012;41:7814.
doi:10.1039/c2dt30825h.
[36] Dragonetti C, Valore A, Colombo A, Magni M, Mussini P, Roberto D, et al. Ruthenium
oxyquinolate complexes for dye-sensitized solar cells. Inorganica Chim Acta 2013;405:98–104.
doi:10.1016/j.ica.2013.05.006.
[37] Tobe ML, Burgess J. Inorganic Reaction Mechanisms. Addison Wesley Longman Ltd.; 1999.
[38] Ruile S, Kohle O, Péchy P, Grätzel M. Novel sensitisers for photovoltaic cells. Structural
24

ACCEPTED MANUSCRIPT
variations of Ru(II) complexes containing 2,6-bis(1-methylbenzimidazol-2-yl)pyridine. Inorganica
Chim Acta 1997;261:129–40. doi:10.1016/S0020-1693(97)05457-1.
[39] Zakeeruddin SM, Nazeeruddin MK, Humphry-Baker R, Grätzel M. Synthesis and photophysical
properties of trans-dithiocyanato bis(4,4′-dicarboxylic acid-2,2′-bipyridine) ruthenium(II) charge
transfer sensitizer. Inorganica Chim Acta 1999;296:250–3. doi:10.1016/S0020-1693(99)00359-X.
[40] Tarr DA, Miessler GL. Inorganic chemistry. 2nd ed. Englewood Cliffs, N.J: Prentice Hall; 1991.
[41] Vlcek A, Dodsworth ES, Pietro WJ, Lever BP. Excited State Redox Potentials of Ruthenium
Diimine Complexes; Correlations with Ground State Redox Potentials and Ligand Parameters.
Inorg Chem 1995;34:1906–13. doi:10.1021/ic00111a043.
[42] Dodsworth ES, Lever ABP. Correlation of electronic charge transfer transitions and
electrochemical potentials. Chem Phys Lett 1984;112:567–70. doi:10.1016/0009-2614(84)85780-
2.
[43] Azpiroz JM, De Angelis F. DFT/TDDFT study of the adsorption of N3 and N719 dyes on
ZnO(101̄0) surfaces. J Phys Chem A 2014;118:5885–93. doi:10.1021/jp501058x.
[44] Shorokhov D, Park ST, Zewail AH. Ultrafast electron diffraction: Dynamical structures on
complex energy landscapes. ChemPhysChem 2005;6:2228–50. doi:10.1002/cphc.200500330.
[45] Baryshnikov G V., Minaev BF, Minaeva VA. Quantum-chemical study of effect of conjugation on
structure and spectral properties of C105 sensitizing dye. Opt Spectrosc 2011;110:393–400.
doi:10.1134/S0030400X10061025.
[46] Nazeeruddin MK, De Angelis F, Fantacci S, Selloni A, Viscardi G, Liska P, et al. Combined
experimental and DFT-TDDFT computational study of photoelectrochemical cell ruthenium
sensitizers. J Am Chem Soc 2005;127:16835–47. doi:10.1021/ja052467l.
[47] Barolo C, Nazeeruddin MK, Fantacci S, Di Censo D, Comte P, Liska P, et al. Synthesis,
characterization, and DFT-TDDFT computational study of a ruthenium complex containing a
25

ACCEPTED MANUSCRIPT
functionalized tetradentate ligand. Inorg Chem 2006;45:4642–53. doi:10.1021/ic051970w.
[48] Ivanova-Shor EA, Nasluzov VA, Shor AM, Antonova AB, Rösch N. Vinylidene carbonylation at a
manganese-iron complex: A density functional study of mechanism. J Organomet Chem
2011;696:3445–53. doi:10.1016/j.jorganchem.2011.06.051.
[49] Apostolova ES, Eltsov A V. Quantum chemical study of complex of dianion forms of fluorescein
and its thioanalog with nucleophile. Zhurnal Obs Khimii 1991;61:1236–40.
[50] Apostolova ES, Eltsov A V. Quantum chemical computation of the effect of the nature of
heteroatom and phenyl substituent on electron characteristics of xanthene dye dianion forms.
Zhurnal Obs Khimii 1990;60:2606–2014.
[51] Apostolova ES, Smirnova NP, Eltsov A V. Influence of nucleophilic substituents and heteroatom
nature on electronic structure and properties of thiophen analogs of fluorescein. Zhurnal Obs
Khimii 1992;62:2750–6.
[52] Zakeeruddin SM, Grätzel M. Solvent-free ionic liquid electrolytes for mesoscopic dye-sensitized
solar cells. Adv Funct Mater 2009;19:2187–202. doi:10.1002/adfm.200900390.
[53] Wang X, Yang Y-L, Wang P, Li L, Fan R-Q, Cao W-W, et al. High efficiency co-sensitized solar
cell based on luminescent lanthanide complexes with pyridine-2,6-dicarboxylic acid ligands. Dalt
Trans 2012;41:10619–25. doi:10.1039/c2dt30430a.
26

ACCEPTED MANUSCRIPT
27

ACCEPTED MANUSCRIPT
Complex
-E1/2Red, V
∆Ep, V
0.06
E1/2Ox, V
∆Ep, V
0.99
1.22
1.75
0.07
0.80
0.06
5
0.09
28