Chemistry of ternary monocarboxyterpyridine-bipyridine-trimercaptotriazine ruthenium complexes and application in dye sensitized solar cells

Article abstract of DOI:10.1016/j.poly.2020.114513

The synthesis of the [Ru^II(Hmctpy)(H₂O)(Cl₂)], [Ru^II(Hmctpy)(dmbpy)Cl]Cl and [Ru^II(Hmctpy)(dmbpy)(H₃tmt)](PF₆)₂ complexes (mctpy = monocarboxyterpyridine, dmbpy = dimethylbipyridine, tmt = trimercaptotriazine) was carried out based on a vectorial design for energy transfer in dye sensitized solar cells (DSSC). The [Ru^II(mctpy)(dmbpy)(H_xtmt)]ⁿ complexes were fully characterized and the electrochemical and absorption/emission properties corroborated their potentially as TiO₂ photosensitizing agents. However, the observed current density and open circuit voltage were rather small, imparting a low efficiency for the assembled DSSCs. The results indicated a unexpected contribution of the charge recombination effects at the electrode surface. Such effects increased with the number of deprotonated groups in the tmt ligand as they interact with the nanocrystalline TiO₂ surface, capturing electrons from the conduction band.

Full text of DOI:10.1016/j.poly.2020.114513

Polyhedron 182 (2020) 114513
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Chemistry of ternary monocarboxyterpyridine-bipyridine-
trimercaptotriazine ruthenium complexes and application in dye
sensitized solar cells
⇑
Juan S. Aguirre-Araque, Robson R. Guimaraes, Henrique E. Toma
Instituto de Química, Universidade de São Paulo, 05508-000 São Paulo, Brazil
a r t i c l e i n f o
a b s t r a c t
The synthesis of the [RuII(Hmctpy)(H2O)(Cl2)], [RuII(Hmctpy)(dmbpy)Cl]Cl and [RuII(Hmctpy)(dmbpy)
(H3tmt)](PF6)2 complexes (mctpy = monocarboxyterpyridine, dmbpy = dimethylbipyridine, tmt = trimer-
captotriazine) was carried out based on a vectorial design for energy transfer in dye sensitized solar cells
(DSSC). The [RuII(mctpy)(dmbpy)(Hxtmt)]n complexes were fully characterized and the electrochemical
and absorption/emission properties corroborated their potentially as TiO2 photosensitizing agents.
However, the observed current density and open circuit voltage were rather small, imparting a low effi-
ciency for the assembled DSSCs. The results indicated a unexpected contribution of the charge recombi-
nation effects at the electrode surface. Such effects increased with the number of deprotonated groups in
the tmt ligand as they interact with the nanocrystalline TiO2 surface, capturing electrons from the con-
duction band.
Article history:
Received 8 November 2019
Accepted 11 March 2020
Available online 17 March 2020
Keywords:
Ruthenium dyes
Photoelectrochemistry
Trimercaptotriazine dyes
Carboxypolyimine dyes
Charge recombination
Ó 2020 Elsevier Ltd. All rights reserved.
1. Introduction
Considering the current efforts to improve the DSSC yields,
alternative photoinjecting complexes can be devised, for instance,
The synthesis of heteroleptic or ternary ruthenium complexes
encompassing mixed ligands [1–5] has been pursued aiming new
applications in chemistry, including alternative routes for improv-
ing the photoelectrochemical response of TiO2 dye solar cells
(DSSC) inspired on the [Ru(dcbpy)2(SCN)2]nꢀ and [Ru(tctpy)
(SCN)3]nꢀ complexes. Such complexes have also been referred as
N719 and black dyes, (dcbpy = dicarboxybipyridine, tctpy = tricar-
boxyterpyridine), respectively [6], and are in the top list of pho-
by replacing the thiocyanate ligands by trimercaptotriazine (tmt).
This particular ligand is a trimeric species from thiocyanate,
(SCN)3, exhibiting an aromatic triazine ring and three thiol ligands
to bind metal ions. As a matter of fact, the [Ru(dcbpy)2(H2tmt)]3ꢀ
complex has already been investigated in our laboratory [12]
exhibiting a significant performance in DSSC, with quantum yields
comparable to those observed for the N3 and N719 thiocyanate
dyes. Along this line, we are directing efforts to obtain the ternary
complex [Ru(tctpy)(tmt)(SCN)]2ꢀ, inspired on the black dye com-
plex. Unfortunately, however, its synthesis has proved rather cum-
bersome and has not yet been successful in our Laboratory. For this
reason, we are investigating the best routes for obtaining this type
of complex, and in a particular attempt we have replaced the tri-
carboxyterpyridine ligand by the less expensive monocarboxyter-
pyridine (mctpy) analogue. This ligand can be synthesized with
good yields, allowing to generate new ruthenium–tmt mixed com-
plexes capable of anchoring onto TiO2 surfaces and promoting
excited electron transfer in DSSC devices.
toinjecting
complexes
[7].
In
their
design,
the
carboxypolypyridine ligands promote the binding and photoelec-
tron injection to the TiO2 nanoparticles, while the thiocyanate
ancillary ligands help improving the cell efficiency, by increasing
light harvesting from the red shift of the absorption maxima in
the visible and also increasing the optical oscillator strength. In
addition, steric hindrance and electrostatic repulsion diminish
the dark current, precluding recombination from the Iꢀ3 species in
the electrolyte media. Ancillary ligands [8] have also been
employed to improve the electron-hole charge separation distance
from the TiO2 surface and the oxidized dye, and increase the DSSC
efficiency [9–11].
Accordingly, in this work we focused our attention on the tern-
ary [Ru(mctpy)(dmbpy)(Hxtmt)]n complexes (Fig. 1), encompass-
ing the Ru(mctpy) moiety and two ancillary ligands: dmbpy and
tmt. The tmt ligand exhibits suitable HOMO-LUMO levels for pro-
moting vectorial energy transfer into the central ruthenium(II)
ion. Similarly, the 4,40-dimethylbipyridine ligand was also selected
⇑
Corresponding author.
E-mail address: henetoma@iq.usp.br (H.E. Toma).
https://doi.org/10.1016/j.poly.2020.114513
0277-5387/Ó 2020 Elsevier Ltd. All rights reserved.

2
J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
2.2. Synthesis of [RuII(Hmctpy)(H2O)Cl2].H2O
This complex was synthesized by dissolving 1.10 g of RuCl3-
ꢂ3H2O (4.23 ꢁ 10ꢀ3 mol) in dinitrogen purged ethanol, in a
125 mL reaction flask equipped with a magnetic bar. Under a
stream of dinitrogen, 1.17 g of mctpy (4.24 ꢁ 10ꢀ3 mol) were
added and the reaction mixture was let to reflux for 4 h. Upon cool-
ing down to room temperature, a red-brown solid precipitated.
This solid was centrifuged and washed three times with cold etha-
nol, yielding 1.95 g of product (82% yield). Analysis: Exp(Calc) C
39.56(39.60); H 3.19(3.12);, N 7.91(8.66). 1H NMR (500 MHz,
dmso-d6) d ppm 7.58 (2H, ddd, J = 7.40, 5.72, 1.22 Hz) 8.00 (2H,
td, J = 7.78, 1.53 Hz) 8.79 (2H, d, J = 7.32 Hz) 8.98 (2H, s) 9.34
(2H, dd, J = 5.49, 0.92 Hz). Mass spectra of [RuII(mctpy)Cl2(H+)]+
ion (449.93 m/z) exhibited 449.2 m/z signal.
2.3. Synthesis of [RuII(Hmctpy)(dmbpy)Cl]Cl
Fig. 1. Structural representation of the [Ru(Hmctpy)(dmbpy)(H2tmt)]+ complex.
In a 50 mL round bottom flask equipped with a magnetic bar,
174.9 mg (4.1 mmol) LiCl, 200 mg (0.41 mmol) [RuII(Hmctpy)
(H2O)Cl2].H2O and 0.5 mL 4-ethylmorpholine were dissolved in
15 mL of methanol/water 5:1 mixture. 76 mg (0.41 mmol) dmbpy
were added to the stirred solution, and brought to reflux for three
hours in the dark. The solution was let to cool down to room tem-
perature and the crude sample was dried in a rotary evaporator.
The purple solid was dissolved and loaded into a silica gel column
and eluted with acetone/methanol/water (3:1:1) saturated with
LiCl. The purple fraction was collected and concentrated in the
rotary evaporator and 3 mL of concentrated HCl were added. The
purple solid precipitated was centrifuged and washed three times
with 1 mol Lꢀ1HCl. The solid was left to dry under vacuum. Yield
45%. 1H NMR (500 MHz, methanol d4) d ppm 2.37 (s, 3H) 2.81 (s,
3H) 6.86 (d, J = 7.05 Hz, 1H) 7.09 (d, J = 5.76 Hz, 1H) 7.33–7.37
(m, 2H) 7.70 (d, J = 5.55 Hz, 2H) 7.86 (d, J = 6.83 Hz, 1H) 7.94 (t,
J = 8.43 Hz, 2H) 8.37 (s, 1H) 8.57 (d, J = 7.90 Hz, 2H) 8.66 (s, 1H)
9.03 (s, 2H) 9.99 (d, J = 5.76 Hz, 1H).
according to a vectorial planning, in order to achieve efficient
energy transfer to the ruthenium(II) center.
Pursuing this idea, we here report the sequential synthesis and
characterization of the monocarboxyterpyridine ligand (mctpy)
and the corresponding [Ru(mctpy)Cl3], [Ru(mctpy)(H2O)2Cl], [Ru
(mtcpy)(dmbpy)Cl]Cl (dmbpy = 4,40dimethyl,2,20bipyridine) and
[Ru(mctpy)(dmbpy)(Hxtmt)]n complexes. The photoelectrochem-
istry of the target complex, [Ru(mctpy)(dmbpy)(H2tmt)] (Fig. 1)
has been investigated in detail, exhibiting significant photoinjec-
tion response in TiO2/DSSCs while providing important clues for
understanding the limitations and for improving the efficiency of
this type of dye.
2. Materials and methods
All reagents were obtained from commercial sources including
Sigma-Aldrich and Synth, and used as supplied, unless stated
otherwise.
2.4. Synthesis of [RuII(Hmctpy)(dmbpy)(H2O)](PF6)2
150 mg (0.24 mmol) of [RuII(Hmctpy)(dmbpy)Cl]Cl were dis-
solved in deionized water and 81.5 mg (0.48 mmol) of AgNO3 were
added slowly over the course of four hours under reflux. AgCl
formed was centrifuged and discarded after cooling down the reac-
tion mixture. Supernatant was concentrated and the complex pre-
cipitated with the addition of NH4PF6. The solid was filtered out,
washed carefully with cold water and dried under vacuum over-
night. This complex was used without further characterization.
2.1. Synthesis of [2,20:60,20’-terpyridine]-40-carboxylic acid (Hmctpy)
The monocarboxyterpyridine ligand was obtained by adapting a
literature procedure [13]: in a 125 mL round bottom reaction flask
equipped with a magnetic bar 3.44 mL of furfural (0.042 mol) and
9.44 mL of 2-acetylpyridine (0.084 mol) were poured together.
Then, 5.2 g of KOH and 60 mL of NH4OH (28–30%) were added
and the reaction was left to proceed for 19 h under reflux. The dark
brown solid was collected on a filter, and recrystallized from an
ethanol–water mixture giving a white solid with about 50% yield.
The white 40-(2-furyl)-2,20:60,20’-terpyridine (futpy) product was
filtered and washed with water, diethyl ether and dried under vac-
uum. In the next step, 1.0 g of futpy (3.3 ꢁ 10ꢀ3 mol) were sus-
pended in 100 mL of deionized water and the pH was adjusted
to 13 using NaOH. Then, 2.12 g of KMnO4 (1.34 ꢁ 10ꢀ2 mol) were
added to the slurry and let to reflux for four hours. After this time,
1.66 g Na2S2O3 were added to reduce unreacted permanganate and
the MnO2 was filtered off. The supernatant pH was adjusted to 4
with HCl (36%) to precipitate the [2,20:60,20’-terpyridine]-40-car-
boxylic acid which was centrifuged and washed three times with
pH 4–5 deionized water. The white solid was dried under vacuum,
yielding 0.61 g (67%) of product. 1H NMR (300 MHz, D2O /NaOD) d
ppm 7.41 (2H, ddd, J = 7.62, 4.91, 1.25 Hz) 7.87 (2H, td, J = 7.76,
1.76 Hz) 8.09 (2H, d, J = 7.91 Hz) 8.22 (2H, s) 8.50 (2H, d,
J = 4.98 Hz).
2.5. Synthesis of [RuII(Hmctpy)(dmbpy)(H3tmt)](PF6)2
100 mg (0.12 mmol) of [RuII(Hmctpy)(dmbpy)(H2O)](PF6)2
were dissolved in 15 mL ethanol and 21.3 mg (0.12 mmol) of
H3tmt were added to the stirred solution and refluxed for 4 h in
the dark. After cooling down, the solvent was removed in the
rotary evaporator and the solid dissolved in the minimum amount
of DMF. This solution was added dropwise to a concentrated NH4-
PF6 aqueous solution, and the precipitated solid was filtered off and
washed carefully with cold water and ether and dried under vac-
uum. Further purification was carried out using Sephadex LH-20
column chromatography using methanol as eluent. The solid was
loaded and four fractions were eluted from the column; the last
one containing the desired complex. Such fraction was dried under
reduced pressure in a rotary evaporator and the hygroscopic solid
kept overnight under high vacuum. 1H NMR (500 MHz,
methanol d4) d ppm 2.38 (s, 3H) 2.79 (s, 3H) 6.94 (dd, J = 5.95,
1.07 Hz, 1H) 6.98–7.00 (m, 1H) 7.34 (ddd, J = 7.32, 5.80, 1.22 Hz,

J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
3
2H) 7.80 (dd, J = 5.80, 1.22 Hz, 1H) 7.86 (dd, J = 5.65, 0.76 Hz, 2H)
5. Theoretical calculations
7.91 (td, J = 7.78, 1.53 Hz, 2H) 8.39 (s, 1H) 8.57 (d, J = 7.93 Hz, 2H)
8.63 (s, 1H) 9.04 (s, 2H) 9.79 (d, J = 5.80 Hz, 1H). Calcd(Found) for
[Ru(Hmctpy)(dmbpy)(H3 tmt)](PF6)2ꢂ12H2O, C31H50F12N8O14P2-
RuS3: C 29.88 (29.43); H 4.05 (3.87); N 8.99 (9.02).
Theoretical calculations were carried out on a preliminary basis,
for help understanding of the electronic structure and spectra.
Although DFT and TDTFD theoretical calculations were initially
performed using the ORCA software, the molecular complexity of
the complexes demanded more than 10 days of computational pro-
cessing, and were frequently aborted for technical problems after
such a long computational time. For this reason, the ZINDO-S
semiempirical method from the Hyperchem 8.05 computational
package was here preferred.13 As demonstrated by Gorelski and
Lever [14] and according to our own experience, there is a reason-
able agreement between the two methods; however, the last one
requires only few seconds to perform on a lap top computer. As a
normal procedure, atomic dipoles were initially used to start the
MM+ geometry optimization, and then replaced by the atomic
charges obtained from the ZINDO-S method, with a convergence
limit of about 10ꢀ5 kcal Åꢀ1 molꢀ1. The electronic distribution
was generated from single CI excitations in an active space involv-
ing 20 frontier molecular orbitals (10 highest occupied and 10 low-
est unoccupied MOs). After many cycles of MM+/ZINDO-S
geometry optimization up to the convergence point, the final elec-
tronic spectra and the molecular orbitals involved were plotted as
2D energy contours.
2.6. Spectrophotometric titration
The complex [RuII(Hmctpy)(dmbpy)(H3tmt)](PF6)2ꢂ12H2O was
spectrophotometric titrated employing a H2O/MeOH 20:1 solution
containing 0.5 mol Lꢀ1 of NaCl supporting electrolyte in order to
maintain and control the ionic force. All of the spectra were col-
lected at room temperature (25 °C) by starting at pH 13 and acid-
ifying stepwise with HCl (3 mol Lꢀ1) solution until pH 2.
2.7. Preparation of [RuII(mctpy)(dmbpy)(Hxtmt)]n deprotonated
species
Spectrophotometric titration afforded three pKa values in the
pH window employed. In this way, the deprotonated species,
[RuII(mctpy)(dmbpy)(Hxtmt)]n were prepared by adjusting the pH
with (Bu4N)OH and precipitated upon addition of diethyl ether.
3. DSSC assembly
6. Results and discussion
DSSCs were assembled by employing a sandwich format. In the
anode, TiO2 nanoparticles paste were deposited by spin coating
forming working areas averaging 0.25 cm2 onto FTO/glass optically
transparent electrode and then sintered at 450 °C for 30 min. The
dyes were adsorbed on TiO2 surface by immersing the anode in
dye solution for 24 h before mounting the cells. The FTO/glass
counter electrode was coated with Pt by treating with H2PtCl6
solution and heating at 450 °C for 30 min. The cells were put
together using SurlynÒ as spacer and gluer, and sealed at 110 °C
for ten minutes. Through a predrilled hole in the counter electrode,
the Iꢀ/Iꢀ3 based electrolyte (0.5 M tert-butylpyridine, 0.6 M tetra-
butylammonium iodide, 0.1 M LiI, 0.1 M I2 in methoxypropioni-
trile) was injected after applying vacuum, and then the hole was
sealed with a coverslip glass using SurlynÒ as glue. The electrolyte
was let to accommodate on the DSSC unity for 24 h before making
the measurements.
6.1. NMR spectra
The first hint about the low spin Ru(II) center in [Ru(mctpy)Cl2-
H2O] was given by its 1H NMR spectrum (Fig. 2 top). Five well-
defined, symmetric signals can be observed, and the lack of broad-
ening is typical of a diamagnetic metal center, exhibiting slower
nuclear relaxation times in contrast with the paramagnetic species
[15]. This supported the assignment of a highly symmetrical Ru(II)
complex and the singlet in 8.98 ppm enabled the integration of
every signal acting as reference. The coupling constants permitted
the correlation between nuclei thanks to the -orto, -meta and -para
values.
Although the 1H NMR spectra gave us a hint about the oxidation
state of the ruthenium center, spin delocalization across the whole
complex may be poor enough to enable a good NMR signal [16].
The complex formed by the coordination of the tridentate ligand
terpy and its analogues is commonly reported as being RuIII(tpy)
Cl3. It is assumed that due to the lack of a reducing agent in the
reaction media, the ruthenium center does not undergo a redox
process during the complex formation [17–20]. However, EPR mea-
surement provided no signals of unpaired electrons as expected for
4d5 Ru(III) [21], indicating a low spin Ru(II) center
The quantification of the loaded complexes on the DSSC anode
was performed by treating with 0.1 mol Lꢀ1 NaOH aqueous solu-
tion, and monitored spectrophotometrically after diluting with
deionized water. Based on the absorption spectra of the complex,
the surface coverage was estimated as 9.3 ꢁ 1016 molecules cmꢀ2
.
4. Instrumentation
1H NMR spectrum of the [RuII(Hmctpy)(dmbpy)Cl]Cl complex
afforded 13 signals between 2.37 and 9.99 ppm expected for a
symmetric terpy with 5 signals and a unsymmetrical dmbpy ligand
due to the electronic trans influence of the chloride to one of the
pyridyl rings, making them magnetically distinct. Therefore, every
ring in dmbpy gives rise to four signals each, as shown in Fig. 2
middle. In the figure, the labeling has been kept for mctpy for com-
parison purposes and another labeling was adopted for dmbpy
hydrogens.
The electronic spectra of the complexes were obtained using a
Hewlett-Packard, model HP-8453-A diode array spectrophotome-
ter in the 200–1100 nm range. 1H NMR spectra were recorded on
a Bruker AIII 500 MHz equipment. Low resolution mass spectra
were obtained using a Esquire 3000 Plus Bruker Daltonics, and high
resolution spectra were collected on a MicroTof Bruker Daltonics
spectrometer. Cyclic voltammetry was carried out with an Autolab
PGStat30 instrument, using a glassy carbon working electrode. An
Oriel Spectral Luminator, whose power at cell position was con-
trolled (1 and 2 mW cmꢀ2) with a standard Si photodiode (1830-
C Newport Optical Power Meter), was used for the IPCE measure-
ments. The I-V curves were registered using an ABB class Oriel
solar simulator (AM 1.5, IEC, JIS, ASTM) calibrated with a Si cell
(VLSI standards, Oriel P/N 91150 V), interfaced to a computer-con-
trolled Keithley 2400 instrument.
The spectrum of the [RuII(Hmctpy)(dmbpy)(H3tmt)](PF6)2 com-
plex is shown in Fig. 2 bottom. Due to the use of a protic solvent
(MeOD-d4), isotopic exchange takes place with the thiol residues
making the hydrogens non visible in the spectrum. Nevertheless,
chemical shifts for the eleven signals found between 6.94 and
9.79 ppm differ substantially from those found for the
[RuII(Hmctpy)(dmbpy)Cl]+ precursor as observed in Fig. 2, allowing

4
J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
Fig. 2. 1H NMR spectra of the synthesized ruthenium complexes in DMSO d6 (top) and MeOD-d4 (middle and bottom).
us to infer the coordination of the tmt is exerting a magnetic effect,
which is more pronounced in the nearest hydrogens although
more analyses were performed to confirm this hypothesis.
line) exhibits two absorption bands of pp ? pp* nature at 285
and 325 nm with a sharper and simpler profile than that of the
[RuII(mctpy)(H2O)Cl2]ꢀ complex due to the contribution of both
polypyridine ligands. On the other hand, the MLCT (d
p ? pp*)
transitions afforded a notorious change upon dmbpy coordination
with a broader band centered in 523 nm displaying a stronger
absorption when compared with the [RuII(mctpy)(H2O)Cl2]ꢀ spec-
6.2. Electronic spectra
The electronic spectra of the [RuII(mctpy)(dmbpy)(H2tmt)]
complex and its precursor species can be seen in Fig. 3. The bands
at 275 and 326 nm for the [RuII(mctpy)(H2O)Cl2]ꢀ complex (black
line) are assigned to IL (p
the band located at 396 and 489 nm are ascribed to charge transfer
trum, indicating the influence of low lying p* orbitals from dmbpy.
The electronic spectrum of the [RuII(mctpy)(dmbpy)(H2tmt)]
complex (blue line) exhibits typical IL transitions in the region
below 400 nm. At 281 nm is located an intense band with a shoul-
der in 317 nm which are ascribed to both dmbpy and mctpy IL
p ? pp*) from the mctpy ligand and
(MLCT) from the Ru(II) center to the terpy ligand (dp ? pp*). The
electronic absorption spectrum of [RuII(mctpy)(dmbpy)Cl] (red
transitions (pp ? pp*). However, the bands profile differs notably
from those in the precursors, indicating a major influence or pres-
ence of underlying absorptions from the tmt ligand. In the visible
region there are located two MLCT bands at 435 and 497 nm and
a shoulder at 630.
According to the theoretical calculations, the highest occupied
molecular orbitals (MO110, 111, 112) in Fig. 4 have a major contri-
bution of the ruthenium center, in addition to the mctpy and
dmbpy ligands. The tmt ligand has a predominant participation
in MO107. On the other hand, the lowest unoccupied molecular
orbitals, MO111, 112, exhibited a majori contribution from the
mctpy ligand, while MO113 seems mainly localized on the dmbpy
ligand. MO 114 involves contributions of both ligands. It seems
that tmt has little participation in the accessible LUMO levels.
Accordingly, the main electronic band observed at 497 nm in
the [RuII(mctpy)(dmbpy)(H2tmt)] complex, is consistent
with the theoretical band calculated at 495 nm, corresponding to
a
predominant Ru(II)
?
mctpy charge-transfer transition
(MO108 ? MO111). On the other hand the second band at
435 nm is compatible with the theoretical band at 477 nm,
predominantly of Ru(II)
(MO110 ? MO113). The shoulder around 630 nm corresponds to
? dcbpy charge transfer nature
Fig. 3. UV–vis spectra in MeOH of the synthesized ruthenium species (1.3 ꢁ 10ꢀ5
mol Lꢀ1).

J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
5
Fig. 4. 2D contour orbital plots for the HOMO and LUMO levels in the [RuII(mctpy)(dmbpy)(H2tmt)] complex and the corresponding theoretical electronic transitions
indicated by the dotted arrows.
the theoretical band expected at 636 nm involving
a
Ru
noticeable variations are in the pp ? pp* transitions, with a bath-
(II) ? mctpy charge-transfer transition (MO110 ? MO112). It is
interesting to note that the theoretical transition from the occu-
pied tmt level (MO107) to the empty mctpy level (MO111) only
occurs at 302 nm, and does not contribute to the electronic spec-
trum in the visible.
The mctpy ligand provides not only the anchoring groups to
TiO2 but also the injecting levels (MO 111) to the conducting band
(Fig. 4). It should be noted that the charge-transfer transition to the
dmbpy ligand (MO 113) occurs at a higher energy in relation to
mctpy, and in this sense it can promote vectorial energy transfer
to the electrode. Although the tmt LUMO level seems not accessi-
ble, this ligand can have a significant role as a donor species
through the HOMO level, MO 107. Therefore, both dmbpy and
tmt ligands can contribute for energy transfer to the injecting orbi-
tals of the dye, as expected from the vectorial design.
ochromic shift owed to the stabilization of the
upon protonation.
p* mctpy orbital
In order to interpret the acid-base equilibria, a plot of the absor-
bance at 302 nm (which was the band with the greater variation
along the experiment) versus pH is presented in Fig. 5C. It displays
three distinguishable regions at pH 2–7, 7–10 and 11–13, which
can be fitted by a sigmoidal function with r2 = 0.99979, providing
the pKa values shown in Table 1, in comparison with the reported
pKa for the free tmt ligand. Presumably the first pKa of 5.35 encom-
passes the protonated Hmctpy and H3tmt ligands, while the pKa of
8.49 and 12.32 belongs to the coordinated H2tmt and Htmt ligands.
6.4. Electrochemistry
Cyclic voltammetry for the [RuII(mctpy)(dmbpy)(H2tmt)] com-
plex was carried out as shown in Fig. 6, exhibiting five redox pro-
cesses in the range of ꢀ2 to +1.5 V vs SHE. The anodic region
presents one monoelectronic process assigned to the RuII/RuIII oxi-
dation at E1/2 = 1.03 V. The cathodic region is much richer since it
displays all of the electrochemical processes of the ligands. The
6.3. Acid-base titration
The [RuII(mctpy)(dmbpy)(H2tmt)] complex exhibits several
functional groups displaying acid-base equilibria. Determining
their pKa values is quite important, since the coordination chem-
istry (for anchoring onto TiO2 nanoparticles), electronic properties
(HOMO/LUMO energies) and electrostatic interactions depend on
the degree of protonation in the complex, as previously shown
by Guimarães et al. [12].
E
pc = ꢀ0.98 V can be ascribed to the reduction of the mctpy ligand,
and the voltammogram is followed by the reduction of tmt at
Epc = ꢀ1.31 V and the stepwise reduction of the dmbpy rings at
Epc = ꢀ1.51 and ꢀ1.77 V, these four processes match their current
intensities for a one electron process each in agreement with
previously reported ruthenium complexes [23–25].
In Fig. 5A it is shown a collection of spectra in the 13–7 pH win-
dow. It is noteworthy the bathochromic shift of the p
p ? pp* tran-
sitions, due to the stabilization of the * molecular orbital of the
p
6.5. Spectroelectrochemistry
tmt ligand upon protonation, whereas the MLCT bands in the visi-
ble range exhibit minimum variation. Fig. 5B shows the spectra
collected between pH 7–2 and once again the main and most
The spectroelectrochemical behavior illustrated in Fig. 7 is con-
sistent with the Ru(III)/(II) process observed at 1.07 V in the cyclic

6
J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
Fig. 5. Spectrophotometric titration of [RuII(mctpy)(dmbpy)(H2tmt)] complex (1 ꢁ 10ꢀ5 mol Lꢀ1) in basic (A) and acid (B) pH window. Absorbance at 302 nm vs pH (C).
Table 1
pKas for [RuII(mctpy)(dmbpy)(H2tmt)] complex in comparison with free tmt ligand.
Compound
pKa1
pKa2
pKa3
TMT
5.71b
5.35a
8.36b
8.49
11.38b
12.32
[RuII(Hmctpy)(dmbpy)(H3tmt)]2+
a) Average value for the protonated mctpy and tmt ligands, b) pKas extracted from
ref. [22].
Fig. 7. Spectroelectrochemistry response of [RuII(mctpy)(dmbpy)(H2tmt)] associ-
ated with the Ru(III)/(II) process in DMF.
6.6. Emission and excitation studies
In Fig. 8 one can compare the electronic spectrum (blue) for the
[RuII(mctpy)(dmbpy)(H2tmt)] complex and its fluorescence spec-
trum (red) in methanol. Upon excitation in the MLCT band at
496 nm, a composite emission profile has been detected at
583 nm. The energy gap between HOMO and LUMO orbitals (E0-
0) was calculated in the point where the two spectra practically
cross, at 516 nm, affording a 2.39 eV difference.
Fig. 6. Cyclic voltammetry of [RuII(mctpy)(dmbpy)(H2tmt)] in CH3CN, at room
temperature, 50 mV sꢀ1
.
voltammogram. The major changes reflect the predominant contri-
bution of the ruthenium center to the electronic structure of the
complex, since all of the absorption’s intensities in the visible are
diminished when the potential sweep is set between 0.89 and
1.15 V. The MLCT bands in 400–600 nm interval are replaced by
a new band at 425 nm, consistent with a LMCT from the sulfur
in the tmt ligand to the Ru(III) center, as previously reported by
Guimaraes et al. [12].
7. Photoaction performance
After obtaining the energy gap it is necessary to calculate either
HOMO or LUMO energy for comparison with the energy levels of
the conduction band (CB) in nanocrystalline anatase TiO2 and
redox potential of the Iꢀ/I3ꢀ couple found in literature [26]. Accord-
ing to the CV experiments the first oxidation potential is related to
the ionization potential (Ip) for the RuII/RuIII process, and the

J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
7
energy of the HOMO orbital can be expressed by equation (1)
[27,28].
Ip ¼ ꢀðEox þ 4:44ÞeV
ð1Þ
where Eox is the onset potential for the RuII/RuIII process versus an
SHE reference electrode and the value 4.44 is the vacuum level
potential of the normal hydrogen electrode. The calculated values
shown in Fig. 9 indicate that the HOMO energy in the complex is
in good position with respect to the Iꢀ/I3ꢀ redox potential for the
proper regeneration of the oxidized dye after the photoelectron
injection into the TiO2 CB. Furthermore, the energy of the LUMO
is placed well above the TiO2 CB, thus favoring the electron
injection.
The energy profiles support a correct planning of the electronic
structure for the [RuII(mctpy)(dmbpy)(tmt)] complex as a photo-
sensitizer for DSSCs.
7.1. Incident photon-to-current efficiency (IPCE) measurements
Three DSSCs were assembled for the series of deprotonated
ruthenium complexes, [RuII(mctpy)(dmbpy)(H2tmt)], [RuII(mctpy)
(dmbpy)(Htmt)]ꢀ, [RuII(mctpy)(dmbpy)(tmt)]2ꢀ and a reference
cell was also mounted with the classical N719 sensitizer for com-
parison purposes. Fig. 10 exhibits the monochromatic photo injec-
tion properties of the three ruthenium species prepared. It can be
observed that all three species exhibit IPCE profiles consistent with
the shape of their absorption spectra, with an injection efficiency
Fig. 8. (A) Absorption (blue) and (B) emission spectra (red) of [RuII(mctpy)(dmbpy)
(H2tmt)] in MeOH. ((Colour online.))
peak around 490 nm where the MLCT dp ? pp* bands are located.
Light harvesting seems not particularly improved by the presence
of the tmt ligand in the complex, in contrast to a previous paper
[12], but in agreement with the theoretical calculations indicating
the tmt ? Ru(mctpy)* band only at 302 nm (Fig. 4). It should be
noticed that in the previous paper the tmt ligand was bidentate,
and the tmt ? Ru(dcbpy)* transition was indeed observed at
465 nm. In the series, the [RuII(mctpy)(dmbpy)(H2tmt)] complex
exhibited the most efficient photoaction response, but with a max-
imum IPCE of only 23%. The efficiency decreased systematically
with the proton removal from the Hxtmt ligand in the (N-Bu4)
([RuII(mctpy)(dmbpy)(Htmt)] and (N-Bu4)2[RuII(mctpy)(dmbpy)
(tmt)] complexes.
In Fig. 11 it is shown the I-V curves obtained for the
[RuII(mctpy)(dmbpy)(H2tmt)] (left) and N719 (right) DSSCs. The
H2tmt complex exhibited much lower values of Voc and Jsc in rela-
tion to N719, reflecting undesired recombination processes in the
Fig. 9. Energy diagram comparison for TiO2 (anatase), [RuII(mctpy)(dmbpy)
(H2tmt)] and Iꢀ/Iꢀ3 redox couple.
Fig. 10. IPCE measurements of [RuII(mctpy)(dmbpy)(H2tmt)] (a), (N-But4) [RuII(mctpy)(dmbpy)(Htmt)] (b) and (N-But4)2[RuII(mctpy)(dmbpy)(tmt)](c) complexes.

8
J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
Fig. 11. I–V curves for [RuII(mctpy)(dmbpy)(H2tmt)] (left) and N719 (right) assembled DSSCs.
Jpmax ꢃ Vpmax
FF ¼
ð2Þ
ð3Þ
Jsc ꢃ Voc
Pdssc
g
¼
ꢃ 100
Plamp ꢃ A
From Table 2, it is noticeable the low values for the current den-
sity and the open circuit voltage, responding for the low power and
efficiency of the assembled DSSCs. In relation to the reference
N719 dye, the poor performance of the [RuII(mctpy)(dmbpy)(Hx-
tmt]n complexes should be ascribed to their smaller light harvest-
ing properties and the large recombination effects at the electrode
surface. On the other hand, the monocarboxyterpyridine ligand has
only an anchoring group to the TiO2 surface, and this might also
decrease its photoinjecting efficiency in relation to the reference
N719 and tricarboxyterpyridine black dyes.
Fig. 12. Vectorial electron transfer and recombination effects in the TiO2 /Ru
(mctpy)(dmbpy)(Hxtmt)]nꢀ DSSC.
8. Conclusions
DSSC device. This may be due to a) the smaller number of anchor-
ing groups in the tmtH2 complex, decreasing the electron photoin-
jection in TiO2 and also Jsc and b) the possible interaction of the
monodentate H2tmt ligand (see Fig. 12) with the nanocrystalline
TiO2 surface, facilitating the electron transfer from the CB to the
Vectorial planning has been successfully applied for the synthe-
sis of the new [RuII(mctpy)(dmbpy)(H2tmt)] complex, pursuing
suitable properties for performing as a DSSC dye. Its electrochem-
ical and spectroscopic properties confirmed a good potentially for
DSSC application, while the determination of the pKa’s of the
acid-base processes allowed to test the behavior of the deproto-
nated species in the IPCE response and cell performance. In com-
parison with the N719 DSSC, rather low values for current
density and open circuit voltage have been observed, accounting
for low power and efficiency of the assembled DSSCs. Such results
reflected their smaller light harvesting properties and large recom-
bination effects occurring at the electrode surface. The recombina-
tion effects increased with the number of deprotonated groups in
the tmt ligand, facilitating the interactions with the nanocrys-
talline TiO2 surface and promoting back-electron transfer from its
CB level.
adsorbed photooxidized complex, thus significantly reducing Voc
.
Such interaction would be facilitated in the deprotonated com-
plexes, because as the sulfur atoms become more available for
interacting with the titanium ions, the energy transfer from the
conducting band is facilitated, leading to the decay of efficiency
and voltage parameters observed along the series. This conclusion
is also consistent with our previous report [12] for a bidentate tmt
ligand exhibiting a much better photoaction response in relation to
the monodentate complex.
FF and efficiency (g) were calculated employing equations (2)
and (3) and the results for all of the three dyes and the reference
(N719) appear in Table 2.
Table 2
Comparative performance of [RuII(mctpy)(dmbpy)(Hxtmt)]n complexes in their assembled DSSCs.
Compound
Power (mW)
FF
Jsc (mA cmꢀ2
)
Voc (V)
g (%)
[RuII(mctpy)(dmbpy)(H2tmt)]
[RuII(mctpy)(dmbpy)(Htmt)]ꢀ
[RuII(mctpy)(dmbpy)(tmt)]2ꢀ
N719
0.175
0.49
0.58
0.50
0.60
3.12
1.71
1.33
14.14
0.44
0.36
0.33
0.69
0.7
8.59 x10ꢀ2
4.79 x10ꢀ2
1.32
0.35
0.19
5.28

J.S. Aguirre-Araque et al. / Polyhedron 182 (2020) 114513
9
position in ruthenium dye molecule on performance of dye-sensitized solar
cells (vol 150, pg 335, 2018), Dyes Pigm. 170 (2019), https://doi.org/10.1016/j.
dyepig.2019.107571.
Author contributions
Juan S. Aguirre-Araque has performed the investigation as part
of his PhD thesis. Robson R. Guimaraes has also performed the
investigation, focusing on dye solar cells, Henrique E. Toma has
been involved in conceptualization, supervision and writing.
[12] R.R. Guimaraes, A.L.A. Parussulo, H.E. Toma, K. Araki, New tunable ruthenium
complex dyes for TiO2 solar cells, Inorg. Chim. Acta 404 (2013) 23–28, https://
doi.org/10.1016/j.ica.2013.04.016.
[13] M.W. Cooke, P. Tremblay, G.S. Hanan, Carboxy-derived (tpy)2Ru2+ complexes
as sub-units in supramolecular architectures: the solubilized ligand 4’-(4-
carboxyphenyl)-4,4‘‘-di-(tert-butyl)tpy and its homoleptic Ru(II) complex,
Inorg. Chim. Acta 361 (2008) 2259–2269.
[14] S.I. Gorelsky, A.B.P. Lever, Electronic structure and spectral, of ruthenium
diimine complexes by density functional theory and INDO/S. Comparison of
the two methods, J. Organometall. Chem. 635 (2001) 187–196, https://doi.org/
10.1016/S0022-328X(01)01079-8.
[15] G.A. Webb, Nuclear magnetic resonance spectroscopy of paramagnetic species,
Annu. Rep. NMR Spectrosc. 3 (1970) 211–259.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
[16] K.L. Bren, Nuclear magnetic resonance (NMR) spectroscopy of
metallobiomolecules, in: Encyclopedia of Inorganic and Bioinorganic
Acknowledgements
Chemistry, John Wiley
9781119951438.eibc0296.
& Sons, Inc., 2008, pp. 1–27. doi:10.1002/
The support from Fundação de Amparo à Pesquisa do Estado de
São Paulo, grant FAPESP 2018/21489-1 is gratefully acknowledged.
[17] V.J. Catalano, T.J. Craig, Monometallic and dimetallic ruthenium (II)
ꢀ
terpyridine complexes employing the Tetradentate ligands
dipyridylpyrazolyl, dipyridyloxadiazole, and their dimethyl derivatives,
Inorg. Chem. 42 (2003) 321–334.
References
[18] N. Chanda, S.M. Mobin, V.G. Puranik, A. Datta, M. Niemeyer, G.K. Lahiri,
Stepwise synthesis of [Ru(trpy)(L)(X)]n+ (trpy = 2,2’:6’,2’’-terpyridine; L = 2, 2 0
-dipyridylamine; X = Cl -, H2O, NO2 -, NO+, O2-). Crystal structure, spectral,
electron-transfer, and photophysical aspects, Inorg. Chem. 43 (2004) 1056–
1064.
[19] A. Garza-Ortiz, P.U. Maheswari, M. Siegler, A.L. Spek, J. Reedijk, A new family of
Ru(II) complexes with a tridentate pyridine Schiff-base ligand and bidentate
co-ligands: synthesis, characterization, structure and in vitro cytotoxicity
studies, New J. Chem. 37 (2013) 3450–3460, https://doi.org/10.1039/
c3nj00415e.
[20] P.M. Van Vliet, S.M.S. Toekimin, J.G. Haasnoot, J. Reedijk, O. Novakova, O.
Vrana, V. Brabec, mer-[Ru(terpy)Cl3] (terpy = 2,2’:6’,2’’-terpyridine) shows
biological activity, forms interstrand cross-links in DNA and binds two guanine
derivatives in a trans configuration, Inorg. Chim. Acta 231 (1995) 57–64.
[21] M.M. Kahn Taqui, D. Srinivas, R.I. Kureshy, N.H. Khan, Synthesis,
characterization, and EPR studies of stable ruthenium(III) Schiff base chloro
and carbonyl complexes, Inorg. Chem. 29 (1990) 2320–2326.
[22] K.R. Henke, A.R. Hutchison, M.K. Krepps, S. Parkin, D.A. Atwood, Chemistry of
2,4,6-trimercapto-1,3,5-triazine (TMT): acid dissociation constants and group
2 complexes, Inorg. Chem. 40 (2001) 4443–4447.
[1] J. Seo, M. Jeong, S.Y. Na, E. Lee, Y.-R. Kim, B. Park, B.H. Kim, Synthesis and
characterization of novel heteroleptic Ru(II) bipyridine complexes for dye-
sensitized solar cell applications, Monastshefte Fur Chemie 150 (2019) 1445–
1452, https://doi.org/10.1007/s00706-019-02480-w.
[2] T. Jakusch, T. Kiss, Characterisation of ternary complex formation between {[}
Ru-III(edta)](-) and various bidentate ligands, Z. Anorg. Allg. Chem. 639 (2013)
1640–1647, https://doi.org/10.1002/zaac.201300111.
[3] D. Saccone, C. Magistris, N. Barbero, P. Quagliotto, C. Barolo, G. Viscardi,
Terpyridine and quaterpyridine complexes as sensitizers for photovoltaic
applications, Materials 9 (2016), https://doi.org/10.3390/ma9030137.
[4] M.W. Cooke, M.-P. Santoni, B. Hasenknopf, G.S. Hanan, Heteroleptic ruthenium
(II) chromophores based on tunable polytopic 4 ‘-(benzamidinato)-2,2 ‘: 6 ‘, 2
‘’-terpyridines, Dalton Trans. 45 (2016) 17850–17858, https://doi.org/10.1039/
c6dt03169b.
[5] M.R.E. Da Silva, T. Auvray, B. Laramée-Milette, M.P. Franco, A.A.C. Braga, H.E.
Toma, G.S. Hanan, Unusual photooxidation of S-bonded mercaptopyridine in a
mixed ligand ruthenium(II) complex with terpyridine and bipyridine ligands,
Inorg. Chem. 57 (2018), https://doi.org/10.1021/acs.inorgchem.7b02965.
[6] C. Hora, F. Santos, M.G.F. Sales, D. Ivanou, A. Mendes, Dye-sensitized solar cells
for efficient solar and artificial light conversion, ACS Sust. Chem. Eng. 7 (2019)
13464–13470, https://doi.org/10.1021/acssuschemeng.9b02990.
[23] K. Araki, L. Angnes, C.M.N. Azevedo, H.E. Toma, Electrochemistry of
a
tetraruthenated cobalt porphyrin and its use in modified electrodes as
sensors of reducing analytes, J. Electroanal. Chem. 397 (1995) 205–210,
https://doi.org/10.1016/0022-0728(95)04193-3.
[7] K. Sharma, V. Sharma, S.S. Sharma, Dye-sensitized solar cells: fundamentals
and current status, Nanoscale Res. Lett. 13 (2018), https://doi.org/10.1186/
s11671-018-2760-6.
[24] K. Araki, H. Toma, Synthesis and characterization of a multibridged porphyrin
complex containing peripheral bis(bipyridine)- ruthenium(II) groups, J. Coord.
Chem. 30 (1993) 9–17, https://doi.org/10.1080/00958979308022742.
[25] K. Araki, H.E. Toma, Luminescence, spectroelectrochemistry and
[8] S. Ashraf, R. Su, J. Akhtar, H.M. Siddiqi, A. Shuja, K.A. Al-Saad, S.Y. Al-Qaradawi,
A. El-Shafei, Investigations into structure-property relationships of novel Ru(II)
dyes with N, N ‘-Diethyl group in ancillary ligand for dye-sensitized solar cells,
Dyes Pigm. 171 (2019), https://doi.org/10.1016/j.dyepig.2019.107754.
[9] A.L.A. Parussulo, T.A. Matias, R.R. Guimaraes, S.H. Toma, K. Araki, H.E. Toma,
Accessing the charge separation effects in dye-sensitized solar cells based on a
vectorial planning of supramolecular ruthenium dyes, Inorg. Chim. Acta 453
(2016) 764–770, https://doi.org/10.1016/j.ica.2016.09.039.
[10] C.J.P. Monteiro, P. Jesus, M.L. Davies, D. Ferreira, L.G. Arnaut, I. Gallard, M.M.
Pereira, C. Serpa, Control of the distance between porphyrin sensitizers and the
TiO2 surface in solar cells by designed anchoring groups, J. Mol. Struct. 1196
(2019) 444–454, https://doi.org/10.1016/j.molstruc.2019.06.074.
photoelectrochemical properties of
a tetraruthenated zinc porphyrin, J.
Photochem. Photobiol., A 83 (1994) 245–250, https://doi.org/10.1016/1010-
6030(94)03821-X.
[26] M. Gratzel, Photoelectrochemical cells, Insight Rev. Art. 414 (2001) 338–344,
https://doi.org/10.1038/35104607.
[27] E. Kucur, J. Riegler, G.A. Urban, T. Nann, Determination of quantum
confinement in CdSe nanocrystals by cyclic voltammetry, J. Chem. Phys. 119
(2003) 2333–2337, https://doi.org/10.1063/1.1582834.
[28] U. Retter, Electrified interfaces in physics, chemistry and biology, Zeitschrift
Für Physikalische Chemie 177 (1992) 120–121, https://doi.org/10.1524/
zpch.1992.177.part_1.120a.
[11] M. Zalas, B. Gierczyk, A. Bossi, P.R. Mussini, M. Klein, R. Pankiewicz, M.
Makowska-Janusik, L. Popenda, W. Stampor, The influence of anchoring group