Iodide ion pairing with highly charged ruthenium polypyridyl cations in CH3CN

Article abstract of DOI:10.1021/acs.inorgchem.5b00344

A series of three highly charged cationic ruthenium(II) polypyridyl complexes of the general formula [Ru(deeb)_3-x(tmam)_x](PF₆)_2x+2, where deeb is 4,4′-diethyl ester-2,2′-bipyridine and tmam is 4,4′-bis[(trimethylamino)methyl]-2,2′-bipyridine, were synthesized and characterized and are referred to as 1, 2, or 3 based on the number of tmam ligands. Crystals suitable for X-ray crystallography were obtained for the homoleptic complex 3, which was found to possess D₃ symmetry over the entire ruthenium complex. The complexes displayed visible absorption spectra typical of metal-to-ligand charge-transfer (MLCT) transitions. In acetonitrile, quasi-reversible waves were assigned to Ru^III/II electron transfer, with formal reduction potentials that shifted negative as the number of tmam ligands was increased. Room temperature photoluminescence was observed in acetonitrile with quantum yields of φ ～ 0.1 and lifetimes of τ ～ 2 μs. The spectroscopic and electrochemical data were most consistent with excited-state localization on the deeb ligand for 1 and 2 and on the tmam ligand for 3. The addition of tetrabutylammonium iodide to the complexes dissolved in a CH₃CN solution led to changes in the UV-vis absorption spectra consistent with ion pairing. A Benesi-Hildebrand-type analysis of these data revealed equilibrium constants that increased with the cationic charge 1 < 2 < 3 with K = 4000, 4400, and 7000 M^-1. ¹H NMR studies in CD₃CN also revealed evidence for iodide ion pairs and indicated that they occur predominantly with iodide localization near the tmam ligand(s). The diastereotopic H atoms on the methylene carbon that link the amine to the bipyridine ring were uniquely sensitive to the presence of iodide; analysis revealed that an iodide "binding pocket" exists wherein iodide forms an adduct with the 3 and 3′ bipyridyl H atoms and the quaternized amine. The MLCT excited states were efficiently quenched by iodide. Time-resolved photoluminescence measurements of 1 revealed a static component consistent with rapid electron transfer from iodide in the "binding pocket" to the Ru metal center in the excited state, k_et > 10⁸ s^-1. The possible relevance of this work to solar energy conversion and dye-sensitized solar cells is discussed.

Full text of DOI:10.1021/acs.inorgchem.5b00344

Article
pubs.acs.org/IC
Iodide Ion Pairing with Highly Charged Ruthenium Polypyridyl
Cations in CH3CN
Wesley B. Swords, Guocan Li, and Gerald J. Meyer*
Department of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3290, United States
S
* Supporting Information
ABSTRACT: A series of three highly charged cationic
ruthenium(II) polypyridyl complexes of the general formula
[Ru(deeb)3−x(tmam)x](PF6)2x+2, where deeb is 4,4′-diethyl
ester-2,2′-bipyridine and tmam is 4,4′-bis[(trimethylamino)-
methyl]-2,2′-bipyridine, were synthesized and characterized
and are referred to as 1, 2, or 3 based on the number of tmam
ligands. Crystals suitable for X-ray crystallography were
obtained for the homoleptic complex 3, which was found to
possess D3 symmetry over the entire ruthenium complex. The
complexes displayed visible absorption spectra typical of metal-
to-ligand charge-transfer (MLCT) transitions. In acetonitrile,
quasi-reversible waves were assigned to RuIII/II electron transfer, with formal reduction potentials that shifted negative as the
number of tmam ligands was increased. Room temperature photoluminescence was observed in acetonitrile with quantum yields
of ϕ ∼ 0.1 and lifetimes of τ ∼ 2 μs. The spectroscopic and electrochemical data were most consistent with excited-state
localization on the deeb ligand for 1 and 2 and on the tmam ligand for 3. The addition of tetrabutylammonium iodide to the
complexes dissolved in a CH3CN solution led to changes in the UV−vis absorption spectra consistent with ion pairing. A
Benesi−Hildebrand-type analysis of these data revealed equilibrium constants that increased with the cationic charge 1 < 2 < 3
with K = 4000, 4400, and 7000 M−1. 1H NMR studies in CD3CN also revealed evidence for iodide ion pairs and indicated that
they occur predominantly with iodide localization near the tmam ligand(s). The diastereotopic H atoms on the methylene
carbon that link the amine to the bipyridine ring were uniquely sensitive to the presence of iodide; analysis revealed that an
iodide “binding pocket” exists wherein iodide forms an adduct with the 3 and 3′ bipyridyl H atoms and the quaternized amine.
The MLCT excited states were efficiently quenched by iodide. Time-resolved photoluminescence measurements of 1 revealed a
static component consistent with rapid electron transfer from iodide in the “binding pocket” to the Ru metal center in the excited
state, ket > 108 s−1. The possible relevance of this work to solar energy conversion and dye-sensitized solar cells is discussed.
after excited-state injection involves a putative ion pair between
the oxidized dye and iodide.7 This paper describes the synthesis
and characterization of ruthenium(II) polypyridyl complexes
with 4+, 6+, and 8+ charges that display enhanced ion pairing
and excited-state reactivity with iodide.
INTRODUCTION
■
The yields and dynamics of photoinduced electron-transfer
reactions between donors and acceptors in a fluid solution are
known to be influenced by Coulombic forces.1 It has become
increasingly apparent that ionic charges also play critical roles in
the efficiency of dye-sensitized solar cells based on mesoporous
thin films of anatase TiO2 nanocrystallites.2,3 Cations present in
the electrolyte are known to be important for excited-state
injection, regeneration of the oxidized dye, and transport of the
injected electron to the external circuit.2 Because these ions
influence so many different aspects of the solar cell, it is often
difficult to isolate and study a specific phenomenon. For
example, surface adsorption of potential determining Lewis
acidic cations such as Li+ is known to influence the energy
levels of the acceptor states in TiO2, and the same cations have
also recently been shown to screen the electric fields generated
by excited-state injection.4−6 Much less is known about how
solar conversion efficiencies are influenced by the overall charge
of the dye molecules, termed sensitizers. This is remarkable
when one considers the vast research efforts in this area and the
fact that the accepted mechanism for sensitizer regeneration
In one of the very few studies to address the importance of
sensitizer charge, Nazeeruddin and co-workers reported careful
titration experiments that enabled the characterization and
isolation of three different protonation states of N3, cis-
Ru(dcb)2(NCS)2, where dcb is 2,2′-bipyridine-4,4′-dicarboxylic
acid.8 It was found that the sensitizer with two of the four
carboxylic acid groups deprotonated, termed N719, gave rise to
the highest solar energy conversion efficiency. If the 2− charge
of N719 was maintained upon surface binding, then the
oxidized form would be a 1− anion unlikely to ion pair with
iodide. As was pointed out by these authors, the presence of
acidic protons shifts the TiO2 conduction band away from the
vacuum level, which can lower the open-circuit photovoltage
Received: February 11, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Ammonium hexafluorophosphate, dichloro(p-cymene)ruthenium(II)
dimer (Ru-dimer), 48% hydrobromic acid, lithium perchlorate, silver
hexafluorophosphate, silver nitrate, sodium borohydride, sulfuric acid,
tetrabutylammonium iodide (TBAI), tetrabutylammonium perchlo-
rate, and 45% aqueous trimethylamine were purchased from Aldrich
and used as received. 4,4′-Diethyl ester-2,2′-bipyridine (deeb) was
prepared according to the literature procedure.14
NMR. Characteristic NMR spectra were obtained using Bruker
Avance III 400 MHz (1H) and 600 MHz (13C) spectrometers. NMR
spectra were referenced to the central line of the solvent (CD3CN)
and processed using MNOVA.
Mass Spectrometry. Electrospray ionization mass spectrometry
(ESI-MS) data were collected on a Micromass Triple Quadrupole
mass spectrometer with a Z-spray nanoelectrospray source and
sampled by an Advion TriVersa NanoMate sampling system.
Measurements were made on complexes dissolved in acetonitrile.
The complexes were identified by their multiple ionization peaks.
X-ray Diffraction. A suitable crystal was selected and mounted on
a MITIGEN holder in paratone oil on a Bruker APEX-II CCD
diffractometer. The crystal was kept at 100 K during data collection.
Using Olex2, the structure was solved with the olex2.solve structure
solution program using charge flipping and refined with the XL
refinement package using least-squares minimization.15−17
UV−Vis Absorption. UV−vis absorption spectra were recorded
using Varian Cary 50 and 60 UV−vis spectrophotometers with a
resolution of 1 nm.
Steady-State PL. Steady-state PL spectra were recorded using an
ISS K2 fluorimeter. Samples were sparged with argon for 20 min and
excited at the MLCT absorption maximum (typically λ ∼ 460 nm).
The intensity was integrated for 0.5 s at 4 nm resolution. PL quantum
yields were measured through comparative actinometry using
[Ru(bpy)3][PF6]2 in acetonitrile (ϕem = 0.062) as a quantum yield
standard.18
and solar energy conversion efficiency. Therefore, while N3
could transfer four protons to TiO2, N719 transferred only two.
Hence, Brønsted acid−base surface chemistry complicated
analysis of the role of charge in this study, and it is likely that
N3 and all of the conjugate bases derived from it were fully
deprotonated when anchored to the TiO2 surface.8
Indeed, to our knowledge, all of the attenuated total
reflectance Fourier transfer infrared studies of transition-metal
sensitizers with dcb ligand(s) reveal complete deprotonation
when anchored to a TiO2 surface from organic solvents.2 An
asymmetric CO stretch is observed near 1610 cm−1, which is
most consistent with carboxylate binding.9−11 Therefore,
surface-anchored cis-Ru(dcb)2(NCS)2 possesses a 4− total
charge in the ground and excited states and a 3− charge after
excited-state injection regardless of which protonation state is
initially anchored to the surface. Unfortunately, in this
literature, the oxidized sensitizers are often referred to as
“dye cations” even though the best available data indicate that
champion RuII sensitizers remain anionic after excited-state
injection.12,13 It is not clear whether an oxidized sensitizer
anchored to TiO2 has ever held a cationic charge. As a result, it
is of interest to characterize sensitizers with high cationic
charges to more fully understand the importance of the
sensitizer charge and the possible interfacial ion pairing with
iodide. This paper describes the first studies directed toward
this goal.
Herein a series of three complexes of the general formula
[Ru(deeb)3−x(tmam)x](PF6)2x+2 were synthesized and charac-
terized and are referred to by the number of 4,4′-
bis[(trimethylamino)methyl]-2,2′-bipyridine (tmam) ligands,
x = 1 (1), 2 (2), or 3 (3). The ligands are shown in Scheme
1. The 4,4′-diethyl ester-2,2′-bipyridine (deeb) ligand was
Time-Resolved PL. Lifetimes and time-resolved PL single-
wavelength decays were acquired on a nitrogen dye laser with
excitation centered at 500 nm. Decays were monitored at the PL
maximum and averaged over 180 scans.
Scheme 1. tmam and deeb Ligands Utilized in This Paper
Electrochemistry. Cyclic voltammetry (CV) experiments were
performed with a BASi CV-50W voltammetric analyzer using a
standard three-cell setup. Cells used consisted of platinum, gold, glassy
carbon, or mercury working electrodes with platinum mesh, disk, or
wire auxiliary electrodes and aqueous silver/silver chloride reference
electrodes from Pine Instrumentation and BASi Analytical Instru-
mentation (mercury electrode only). A BASi controlled-growth
mercury electrode on the standing drop setting was used as the
working mercury electrode. Supporting electrolytes of 300 mM LiClO4
or TBAClO4 in acetonitrile were used for all measurements. The
reference electrodes were referenced to an external ferrocene standard
(630 mV vs NHE) before and after each series of measurements.19
Iodide Titrations. Titrations were performed on complexes 1−3
in CH3CN with a fixed ruthenium concentration and variable
concentrations of TBAI. 1H NMR titrations were performed on a
chosen because the ethyl ester groups can later be saponified
for studies at TiO2 interfaces. The tmam ligand has a 2+ charge
by virtue of the quaternary alkylated amine substituents in the 4
and 4′ positions of the bipyridine ligand. This cationic ligand is
shown to influence the redox properties of the complexes as
well as the metal-to-ligand charge-transfer (MLCT) excited
states. The tmam ligand also enhances ion pairing with iodide,
Bruker Avance III 500 MHz spectrometer. Titrations were performed
1
with ∼2 mM ruthenium and
/ equiv additions of TBAI. Each
2
spectrum was averaged over 32 scans. UV−vis titrations were
performed with ∼50 μM ruthenium titrated with /4 equiv of TBAI.
1
PL titrations were completed using the same sample for both steady-
state and time-resolved measurements. The solutions were excited at
500 nm, where absorbance changes were minimal. Solutions of ∼25
μM ruthenium were titrated with 1/4 equiv of TBAI. Data analysis for
all experiments was performed using OriginLab, version 9.0.
4,4′-Bis[(trimethylamino)methyl]-2,2′-bipyridine Bis-
(hexafluorophosphate) (tmam). The following synthesis of tmam
was modified from Li et al., shown in Scheme 2.20 Reduction of deeb
with sodium borohydride in refluxing ethanol gave 4,4′-dihydrox-
ymethyl-2,2′-bipyridine in 81% yield. Continued refluxing in 48%
hydrobromic acid with catalytic sulfuric acid produced 4,4′-
dibromomethyl-2,2′-bipyridine in 72% yield. To a stirring solution of
4,4′-dibromomethyl-2,2′-bipyridine (207 mg, 0.660 mol) in 10 mL of
ethanol was added 1.5 mL (excess) of 45% aqueous trimethylamine
1
as evidenced by H NMR, UV−vis, and photoluminescence
(PL) titration studies. Significantly, the tmam ligand appears to
provide a specific “binding pocket” for iodide that facilitates
rapid photooxidation. To our knowledge, complex 3 is the most
highly charged mononuclear ruthenium(II) complex reported.
EXPERIMENTAL SECTION
■
Materials. Argon gas (Airgas, 99.998%) was passed through a
Drierite drying tube before use. Acetonitrile (Burdick and Jackson,
99.98%), acetone (Aldrich, 99.5%), dichloromethane (DCM; Fisher,
99.5%), diethyl ether (Aldrich, 99.5%), ethanol (Fisher, 200 proof,
99.5%), and methanol (Aldrich, 99.5%) were used as received.
B
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
[Ru(tmam)2(deeb)][PF6]6 (2). To a 25 mL round-bottomed flask
were added [Ru(p-cymene)(deeb)Cl]Cl (10.1 mg, 0.0169 mmol), 2
equiv of tmam (19.6 mg, 0.0339 mmol), and a slight excess of silver
nitrate (7.0 mg, 0.0412 mmol) along with 6 mL of ethanol. The
mixture was sparged with nitrogen for 20 min and refluxed over 2 days,
during which the color changed from orange to brown. The reaction
mixture was filtered over a fine frit, yielding a brown-orange
precipitate, which was dissolved in acetonitrile and filtered over a
fine frit. Removal of the solvent under vacuum gave a red-orange
powder. This powder was redissolved in acetonitrile, and a large excess
of ammonium hexafluorophosphate was added. Acetonitrile was
removed under vacuum, and the remaining solid was slurried in
ethanol, filtered, and dried under vacuum. Recrystallization by vapor
diffusion of methanol into a concentrated acetonitrile solution yielded
15 mg (47%) of 2. 1H NMR: δ 9.07 (d, 2H, J = 0.8 Hz), 8.62 (dd, 4H,
J = 5.6 and 1.2 Hz), 7.91 (d, 4H, J = 4.4 Hz), 7.88 (dd, 2H, J = 4.8
Hz), 7.81 (d, 2H, J = 4.8 Hz), 7.55 (dd, 2H, J = 4.8 and 1.6 Hz), 7.50
(dd, 2H, J = 4.8 and 1.6 Hz), 4.55 (s, 4H), 4.53 (s, 4H), 4.47 (q, 4H, J
= 5.6 Hz), 3.14 (s, 18H), 3.12 (s, 18H), 1.41 (t, 6H, J = 5.6 Hz). 13C
NMR: δ 30.9, 38.0, 54.1, 54.2, 67.65, 67.67, 124.8, 127.9, 129.2, 129.3,
132.2, 132.4, 139.1, 139.1, 140.2, 153.8, 154.1, 154.2, 158.0, 158.2,
158.3, 164.9. ESI-MS. Calcd (found) for RuC52H72N10O4P5F30: m/z+
1727.08 (1727.15). Calcd (found) for RuC52H72N10O4P4F24: m/z2+
791.09 (791.03). Elem anal. Calcd for RuC52H72N10O4P6F36
(1872.05): C, 33.36; H, 3.88; N, 7.48. Found: C, 31.88; H, 3.97; N,
7.30.
[Ru(tmam)3][PF6]8 (3). To a 25 mL round-bottomed flask was
added [Ru(p-cymene)(tmam)Cl][PF6]2Cl (20 mg, 0.224 mmol), 2
equiv of tmam (26.5 mg, 0.0448 mmol), and a small excess of silver
hexafluorophosphate (17.2 mg, 0.0493 mmol) along with 6 mL of
ethanol and 10 mL of acetone. The mixture was refluxed over 2 days,
changing from yellow to red. The reaction mixture was filtered over a
fine frit to isolate a brown powder. The powder was washed with
acetonitrile to dissolve only the product. Removal of the solvent under
vacuum yielded 17 mg of 3 (35%) as an orange powder.
An alternative synthesis was performed using an Anton Parr
Monowave 300 microwave reactor. To a 10 mL microwave tube was
added [Ru(p-cymene)(tmam)Cl][PF6]2Cl (68 mg, 0.0758 mmol),
along with tmam (93.0 mg, 0.158 mmol) and 6 mL of DI water. The
slurry was reacted at 150 °C for 2 h, during which the yellow slurry
turned into a red solution. The reaction mixture was filtered through a
fine frit, and an excess of ammonium hexafluorophosphate was added
to precipitate orange solid 3. The precipitate was isolated on a fine frit,
washed with water and ethanol, and dried under vacuum, yielding 118
mg (72%). X-ray-quality crystals were grown out of acetonitrile
through vapor diffusion of diethyl ether. 1H NMR: δ 8.38 (d, 6H, J = 2
Hz), 7.88 (d, 6H, J = 6 Hz), 7.55 (dd, 6H, J = 6 and 2 Hz), 4.54 (s,
12H), 3.14 (s, 54H). 13C NMR: δ 54.1, 67.6, 129.0, 132.3, 38.8, 154.0,
158.1. ESI-MS. Calcd (found) for RuC54H84N12P6F36: m/z2+ 936.11
(936.19). Calcd (found) for RuC54H84N12P5F30: m/z3+ 575.75
(575.87). Elem anal. Calcd for RuC54H84N12P8F48 (2162.15): C,
30.00; H, 3.92; N, 7.78. Found: C, 30.04; H, 4.01; N, 7.63.
Scheme 2. Synthetic Procedure for the tmam Ligand
dropwise. The cloudy solution was stirred at room temperature where
it transitioned to clear and back to cloudy over 1 h. Deionized (DI)
water was added dropwise until the solution became clear. A large
excess of ammonium hexafluorophosphate was added, precipitating
tmam. Collection over a fine frit, washing with DI water, and drying
under vacuum gave a light-pink-gray powder, yielding 280 mg (78%).
The overall yield from deeb to tmam was 45%. 1H NMR (CD3CN): δ
8.84 (d, 2H, J = 4.0 Hz), 8.58 (d, 2H, J = 1.0 Hz), 7.56 (dd, 2H, J = 4.0
and 1.6 Hz), 4.50 (s, 4H), 3.09 (s, 9H). 13C NMR (CD3CN): δ 53.9,
68.7, 125.3, 128.9, 138.3, 151.5, 157.1.
[Ru(p-cymene)(deeb)Cl]Cl. Modified from Yu et al., [Ru(p-
cymene)(deeb)Cl]Cl was synthesized by the addition of Ru-dimer
(201 mg, 0.329 mmol) and deeb (198 mg, 0.657 mmol) to a 25 mL
round-bottomed flask.21 After the addition of 1:1 DCM/acetone (8
mL), the slurry was sparged with nitrogen for 20 min and refluxed
under dinitrogen for 3 h with little color change. Removal of solvent
yielded an orange product, which was slurried in DI water and filtered.
Removal of DI water under vacuum gave 323 mg of the desired
product (81% yield). 1H NMR (CD3CN): δ 9.87 (dd, 2H, J = 6.0 and
0.8 Hz), 8.84 (d, 2H, J = 1.6 Hz), 8.13 (dd, 2H, J = 8.0 and 1.6 Hz),
6.22 (d, 2H, J = 6.4 Hz), 6.03 (d, 2H, J = 6.4 Hz), 4.46 (q, 4H, J = 7.2
Hz), 2.67 (hept, 1H, J = 6.8 Hz), 1.42 (t, 6H, J = 6.8 Hz), 1.01 (d, 6H,
J = 7.2 Hz). 13C NMR (CD3CN): δ 14.4, 19.0, 22.2, 31.8, 63.7, 83.2,
88.3, 105.6, 107.01, 124.1, 127.6, 141.8, 155.9, 158.3, 164.0.
[Ru(p-cymene)(tmam)Cl][PF6]2Cl. A total of 2 equiv of tmam
(150 mg, 0.254 mol) and Ru-dimer (78 mg, 0.127 mmol) were
dissolved in acetone (8 mL) in a 25 mL round-bottomed flask and
sparged with nitrogen for 20 min. The solution was refluxed for 3 h
under dinitrogen, changing from red to yellow. After cooling to room
temperature, a yellow precipitate formed. Filtering over a fine frit and
drying under vacuum yielded 181 mg (80%) of the desired product.
1H NMR (CD3CN): δ 9.49 (d, 2H, J = 5.6 Hz), 9.19 (s, 2H), 7.89 (dd,
2H, J = 5.6 and 1.6 Hz), 5.99 (d, 2H, J = 6.4 Hz), 5.84 (d, 2H, J = 6.4
Hz), 4.80 (m, 4H), 3.23 (s, 18H), 2.72 (hept, 1H, J = 6.8 Hz), 2.16 (s,
3H), 1.08 (d, 6H, J = 6.8 Hz). 13C NMR (CD3CN): δ 18.8, 22.2, 31.9,
54.5, 67.9, 86.3, 87.2, 129.9, 130.1, 132.0, 140.4, 141.2, 156.38, 156.96.
[Ru(tmam)(deeb)2][PF6]4 (1). To a 25 mL round-bottomed flask
was added [Ru(p-cymene)(tmam)Cl][PF6]2Cl (19.5 mg, 0.0223
mmol), 2 equiv of deeb (13.4 mg, 0.0446 mol), and a slight excess
of silver hexafluorophosphate (13.4 mg, 0.0480 mmol). Ethanol (6
mL) and acetone (2 mL) were added and sparged with nitrogen for 20
min. The mixture was refluxed under dinitrogen over 2 days, while the
color changed from yellow to red. The reaction was filtered over a fine
frit. Removal of solvent under vacuum yielded an orange solid, which
was dissolved in DI water and filtered. The DI water was removed
under vacuum, yielding 15 mg (43%) of 1 as a red powder. 1H NMR:
δ 9.07 (m, 4H), 8.56 (d, 2H, J = 1.2 Hz), 7.93 (d, 2H, J = 6.0 Hz), 7.88
(dd, 2H, J = 6.0 and 1.6 Hz), 7.86 (d, 2H, J = 6.0 Hz), 7.83 (m, 4H),
7.51 (dd, 2H, J = 5.6 and 1.6 Hz), 4.52 (s, 4H), 4.47 (dq, 8H, J = 4.0
and 7.2 Hz), 3.13 (s, 18H), 1.41 (dt, 12H, J = 4.0 and 7.2 Hz). 13C
NMR (CD3CN): δ 14.4, 54.2, 63.8, 63.8, 67.7, 124.8, 124.9, 127.7,
127.9, 129.2, 132.3, 139.2, 140.6, 140.7, 153.9, 154.0, 154.3, 158.1,
158.2, 158.3, 164.3, 164.4. ESI-MS. Calcd (found) for
RuC50H60N8O8P3F18: m/z+ 1437.02 (1437.35). Calcd (found) for
RuC50H60N8O8P2F12: m/z2+ 646.03 (646.23). Elem anal. Calcd for
RuC50H60N8O8P4F24 (1581.98): C, 37.96; H, 3.82; N, 7.08. Found: C,
37.12; H, 3.84; N, 7.00.
RESULTS
■
A modified literature procedure was used to obtain tmam in
45% yield.5 Ligation of ruthenium proceeded through reaction
with a known ruthenium dimer and gave the [Ru(LL)(p-
cymene)Cl]Cl intermediate in high yield (LL = tmam or deeb).
Further reaction with 2 equiv of tmam or deeb under reflux or
high-temperature microwaving led to the isolation of 1−3 in
43%, 47%, and 72% yield, respectively. The identity of the
complexes was confirmed using 1H and 13C NMR, mass
spectrometry, and elemental analysis.
Crystals of 3 that were of sufficient quality for character-
ization by single-crystal X-ray crystallography were isolated
−
(Figure 1). The expected stoichiometry of eight PF6 anions
−
per Ru center was observed, although the PF6 anions were
omitted from Figure 1 for clarity. The average Ru−N bond
C
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. Steady-state absorption (solid line) and PL (dashed line)
spectra for the indicated complexes in CH3CN.
Light excitation into the MLCT absorption bands resulted in
orange-red PL, which was visible to the dark adapted eye
(Figure 2). Both 1 and 2 displayed maxima that were red-
shifted ∼260 cm−1 from that of 3. Quantum yields were
calculated through the optically dilute technique with [Ru-
(bpy)3][PF6]2 in CH3CN as a quantum yield standard.18
Pulsed-laser excitation resulted in time-resolved PL data that
were well described by a first-order kinetic model. Both the
lifetime measurements and quantum yields were performed in
triplicate and averaged to ensure accuracy, with uncertainty
expressed in the last significant digit. Nonradiative and radiative
rate constants were calculated using the corresponding
quantum yields, ϕ = kr/(kr + knr) and lifetimes, τ0 = 1/(kr +
knr). The photophysical properties of the complexes are given
in Table 2.
Figure 1. Displacement ellipsoid plot (50% probability level) for the
cation 3, [Ru(tmam)3]8+, obtained from single-crystal structure
determination. The eight PF6 anions have been masked for clarity.
−
Color code: pink, Ru; purple, N; gray, C.
distance was 2.063 Å. The average bite angle was 79.21°, and
average N−Ru−N angles were 93.69 and 173.63°. The
trimethylamine groups were oriented on opposite faces of the
bipyridine rings, and the complex still maintained an overall D3
point group, with the C3 axis intersecting two faces of the
octahedron and three C2 axes through each bipyridine.
Additional X-ray crystallographic data are included in Table 1.
Table 2. Photophysical Properties for 1−3 in CH3CN
Table 1. Crystal Parameters for 3
MLCT abs
max (nm)
PL max
(nm)
τ0
(μs)
kr (×104 knr (×104
empirical formula
C58H90F48N14P8Ru
complex
ϕ
s−1 s−1
)
)
fw
2244.26
1
2
3
465
463
461
631
635
625
2.25
2.06
1.67
0.12
0.14
0.14
5.4
6.6
7.0
39.0
41.0
53.0
temperature/K
100
cryst syst
monoclinic
space group
C2/c
a/Å
14.3741(7)
b/Å
30.0594(13)
Cyclic voltammetry was performed in an argon-sparged 0.3
M LiClO4/CH3CN solution in a standard three-electrode cell
with a platinum disk working electrode. Quasi-reversible waves
were found at positive potentials and assigned to the RuIII/II
reduction potential (Table 3). Complex 3 was found to have
the lowest potential at 1.65 V vs NHE, and 1 had the highest
potential at 1.75 V vs NHE. The electrochemistry is termed
quasi-reversible because the cathodic and anodic currents were
approximately equal while the peak-to-peak separation was
c/Å
21.7725(9)
α/deg
90
β/deg
92.048(3)
γ/deg
volume/Å3
90
9401.4(7)
Z
4
ρcalc/(g/cm3)
μ/mm−1
1.586
3.920
cryst size/mm3
0.354 × 0.12 × 0.057
Cu Kα (λ = 1.54178)
R1 = 0.0517, wR2 = 0.1342
R1 = 0.0691, wR2 = 0.1423
radiation
Table 3. Electrochemical Data for 1−3 in a CH3CN
Electrolyte
final R indexes [I > 2σ(I)]
final R indexes [all data]
E° (V vs NHE)
complex
ΔGES
RuII/+
RuII*/+
RuIII/II
RuIII/II
*
All characterization was performed in acetonitrile, under an
argon atmosphere, and at room temperature, except where
otherwise stated. The UV−vis spectra displayed absorption
bands at 460 and 440 nm, which were assigned as MLCT
transitions (Figure 2). An increase in the extinction coefficients
was observed as the number of tmam ligands increased from 1
→ 2 → 3. The intense band at 300 nm was assigned as a ligand-
localized π−π* transition.
a
b
1
2
3
2.16
2.16
2.21
−0.68
−0.69
−0.71
1.54
1.75
1.72
1.65
−0.41
−0.44
−0.56
a
a
b
1.53
a
Irreversible reductions were observed under all conditions inves-
tigated; potentials reported are at the peak reductive current.
b
Calculated using the literature value for the first reduction of
[Ru(deeb)3][PF6]2 dissolved in a CH3CN electrolyte.24
D
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
greater than 59 mV.22 Attempts to reduce the complexes
resulted in irreversible reduction chemistry at platinum, gold,
glassy carbon, or hanging mercury drop working electrodes
(Figure S1 in the Supporting Information, SI). A linear region
on the high energy edge of the steady-state PL spectra was
extrapolated to zero, and the intercept provided the free energy
stored in the excited state, ΔGES.23 The reducing power of the
excited state was calculated as E°(RuIII/II*) = E°(RuIII/II) −
ΔGES. The irreversible nature of the ligand reductions leads to
some uncertainty in the oxidizing power of the MLCT excited
state. When the literature value for the first reduction of
[Ru(deeb)3][PF6]2 was used as an approximate value for 1 and
2, the oxidizing potentials of the excited-state complexes were
calculated as E°(RuII*/+) = E°(RuII/+) + ΔGES. The electro-
chemical data are summarized in Table 3.
both static and dynamic quenching mechanisms are operative.
Static quenching was indicated through the decreases in the
initial time-resolved PL amplitude (I0) measured after pulsed-
laser excitation. Stern−Volmer analysis of I0 provided an
estimate of the ground-state ion pairing equilibrium constant
(Ks) for 1. This value, KS = 14000 M−1, was considerably larger
than that estimated by the Benisi−Hildebrand-type analysis of
the ground-state absorption. A Stern−Volmer plot of the
excited-state lifetime for 1 as a function of free iodide was also
linear and revealed the dynamic quenching constant KD (Figure
4). However, the same analysis for 2 and 3 resulted in nonlinear
Stern−Volmer plots for both static and dynamic components
(Figure S3 in the SI). The reasons for this nonlinearity are
uncertain, but it may result from an inability to determine the
free iodide concentration with these more highly charged
complexes. A bimolecular quenching rate constant (kq) of 6.27
× 1010 M−1 s−1 (kq = KD/τ0) was calculated from the dynamic
quenching constant measured for 1 and the excited-state
lifetime. This value is very close but slightly less than that
Iodide titrations were performed for 1−3 in CH3CN and
monitored by UV−vis absorption. Representative data for 1 are
given in Figure 3 and those for 2 and 3 in Figure S2 in the SI.
previously reported for diffusion-limited electron transfer, kdiff
=
6.4 × 1010 M−1 s−1.24
Iodide titrations were also monitored by 1H NMR in
CD3CN. Representative data for 1 are included in Figure S4 in
the SI. Large downfield shifts were observed for the H atoms on
the tmam ligand after the addition of 1/2 equiv of TBAI to 1−3.
The magnitudes of the shifts were dependent on the iodide
concentration but saturated after the addition of 10 equiv of
iodide. The 3 and 3′ H atoms on the tmam ligands showed the
largest shifts (∼1.0 ppm) for all three complexes. Representa-
tive data for 1 are given in Figure 5. In neat acetonitrile, the
resonances from the H atoms on the methylene C atom that
separates the amine from the bipyridine ring were singlets. As
iodide was added to the solution, the individual proton
resonances appeared, as was indicated by the appearance of two
roofed doublets. The coupling constants for each doublet, J =
14 Hz, aligned with known methylene proton constants.26
Relatively large shifts were also seen in the 5, 5′, 6, and 6′ tmam
H atoms. In contrast, resonances associated with the deeb
ligands in 1 or 2 were essentially independent of the iodide
concentration. The magnitudes of the downfield shifts at 6
equiv of I− for 1 are shown in Figure 6.
Figure 3. Absorption spectra of 1 with the indicated number of iodide
equivalents. The arrows indicate the direction of the absorption
change with increasing [I−]. The inset displays the absorption change
at 353 nm with an overlaid fit to a nonlinear modified Benesi−
Hildebrand model from which an ion pairing equilibrium constant of
K = 4000 400 M−1 was abstracted.
All three complexes displayed a small red shift in the low-
energy MLCT absorption upon the addition of iodide. A more
significant absorption growth was observed near 360 nm and
was analyzed by a nonlinear modified Benesi−Hildebrand
equilibrium model, which was previously used to quantify ion
pairing equilibrium constants.25 The abstracted equilibrium
constants are reported in Table 4. The concentration of
solvated iodide, i.e., “free” or non-ion paired iodide, was
determined from the equilibrium constant abstracted from
Benesi−Hildebrand analysis.
Iodide was also found to quench the PL intensity and
excited-state lifetimes of complexes 1−3 in CH3CN.
Representative data are shown for complex 1 in Figure 4.
Stern−Volmer plots of the steady-state PL data were nonlinear
with upward curvature, behavior that is often observed when
DISCUSSION
■
The synthesis of new highly charged ruthenium complexes
based on the tmam ligand was successful. To our knowledge,
complex 3 is the most highly charged mononuclear ruthenium-
(II) complex that has been synthesized, with an overall charge
of 8+. Interestingly, the homoleptic 3 maintains a D3 symmetry
over the entire complex including the amines in the solid state.
Complexes 1 and 2 have ethyl ester groups that can be
hydrolyzed for binding to TiO2 in future studies. The
quaternary amine groups were found to influence the redox
and photophysical properties of the complexes as well as the
interactions with iodide. This behavior is described below in the
context of the relevant literature.
Table 4. Equilibrium Constants for Ion Pairing in CH3CN
Redox and Photophysical Properties. The redox,
structural, and photophysical properties of the newly
synthesized and highly charged complexes can largely be
understood by consideration of the electronic influence of the
substituents in the 4 and 4′ positions of the bipyridine ligands
tmam and deeb. The methylene bridge that links the bipyridine
to the quaternary N atom in tmam isolates the cationic
trimethylamine group from the π system of the bipyridine
a
complex
UV−vis (K, M−1
)
PL (KS, M−1
)
1
2
3
4000
4400
7000
14000
b
b
a
Determined from Benesi−Hildenbrand modeling of the UV−vis
b
absorption changes. Nonlinear Stern−Volmer plots.
E
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. Steady-state (A) and time-resolved (B) PL spectral changes with increasing [I−] for 1. Inset: Lifetime (black) and intensity (blue) Stern−
Volmer plots with overlaid best fits (red lines) for 1.
1
Figure 5. H NMR resonances for 3 and 3′ (A) and methylene (B) tmam H atoms of 1 in CD3CN with the indicated equivalents of iodide.
nm.27 The complexes displayed room temperature PL with
excited-state lifetimes of a few microseconds. Pioneering
resonance Raman studies by Woodruff and co-workers have
provided compelling evidence that the excited state of
Ru(bpy)3 is localized on a single ligand.28 For heteroleptic
2+
ruthenium(II) diimine complexes, Blakley and DeArmond have
shown that the excited state localizes upon the ligand that is
most easily reduced.29 The irreversible nature of the ligand
reductions for these complexes precluded such a determination.
However, solely on the basis of the mesomeric electron-
withdrawing effect of the ester group relative to the inductive
effect of the tmam ligands, the deeb ligand should be reduced
first and therefore be the location for excited-state localization.
However, one could envision that Coulombic interaction on
the tmam ligand would stabilize the MLCT excited state further
by virtue of the cationic charge present. Hence, there was some
concern that complexes 1 and 2 could be an exception to
DeArmond’s rule. While the data do not provide definitive
proof of where the excited state is localized, there is strong
evidence that it is localized on a deeb ligand in the heterolpetic
ruthenium(II) complexes 1 and 2.
Figure 6. Downfield shifts in ppm for the indicated positions of the H
atoms measured at high iodide concentrations, 6 equiv, relative to neat
acetonitrile. The deeb ligands were equivalent on the NMR time scale
at ambient temperature, and only one is depicted here for clarity.
The free energy stored in the MLCT excited state is
proportional to the energy separation between the metal-based
E°(RuIII/II) and the first ligand reduction E°(RuII/+). To a very
good approximation, the first ligand reduction potentials are
insensitive to the identity of the other diimine ligands or even
the metal ion that it is coordinated to.27 With this in mind,
consider complex 3, whose excited state is well formulated as
[RuIII(tmam−)(tmam)2]8+* and stores about 2.21 eV of free
energy in the excited state with E°(RuIII/II) = 1.65 V vs NHE.
As mentioned above, the E°(RuIII/II) potentials shift positively
when a tmam ligand is replaced by deeb. Therefore, if the
ligand. The methylene bridge is electron-donating relative to
the electron-withdrawing ethyl ester groups of the deeb ligand.
These inductive effects were clearly reported by the E°(RuIII/II
)
reduction potentials that shifted 20−70 mV negative when a
deeb ligand was replaced with a tmam ligand. Clearly, the tmam
ligand provides more electron density to the Ru metal center
than does the deeb ligand.
The absorption spectra of the complexes were typical of
ruthenium(II) diimine complexes with intense MLCT
absorption bands centered in the visible region near 460
F
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 3. Proposed Mechanism for Iodide Photooxidation by 1, Where N−N = deeb
excited state were localized on the tmam ligand in 1 or 2, the
corresponding PL spectra would be shifted to higher energy,
contrary to the experimental data. The lower-energy PL from 1
and 2 is, hence, most consistent with radiative relaxation from
an excited state localized on the deeb ligand. Additional, albeit
indirect, evidence for a deeb-localized excited state comes from
the significantly longer lifetimes of 1 and 2 (2.25 and 2.06 μs,
respectively) relative to 3 (1.67 μs). It has previously been
noted that [RuIII(deeb−)(LL)2]2+* complexes possess longer-
lived excited states than related complexes with similar energy
gaps, a behavior that has been attributed to delocalization over
the ethyl ester group in the MLCT excited state.30
are expected. However, in neat acetonitrile, they appeared as a
singlet, presumably because the average magnetic environment
for each H atom (on the same C) was equivalent. As iodide was
titrated into the solution, these singlets split into two roofed
doublets as the diastereotopic coupling emerged, indicating that
the magnetic environments were no longer equivalent. The
downfield shifts of the methyl (∼0.2 ppm) and methylene
(∼0.4 ppm) protons on the tmam ligand were significant and
provide clear evidence that iodide was indeed interacting with
the quaternary amine.
1
The results of the H NMR titrations indicated that iodide
was stabilized by the 3 and 3′ H atoms of the bipyridine ring
and the methylene and methyl groups of the tmam ligand. An
iodide “binding pocket” can be envisioned where iodide forms
an adduct by interacting with the acidic 3 and 3′ H atoms of the
tmam ligand and is further stabilized by the trimethylamine
groups, which can wrap around the iodide. An example of this is
shown in Scheme 3. This adduct gives rise to a nonemissive
excited state that appears as a static component in the excited-
state relaxation, consistent with rapid electron transfer from
iodide to the metal center, ket > 108 s−1. Given the accepted
E°(I0/−) = 1.23 V vs NHE reduction potential in CH3CN, the
reaction is endergonic by 0.31 eV.33 At low iodide
concentrations, a dynamic process was also observed for 1,
consistent with a second-order rate constant of 6.3 × 109 M−1
s−1 that is about an order of magnitude smaller than that
calculated for the diffusion limit, yet consistent with the known
rapid redox reactivity of iodide.33 The rapid static electron
transfer and identification of a iodide “binding pocket” on the
tmam ligand suggests that these complexes will be of use for
application in dye-sensitized solar cells and may indeed
enhance regeneration at the power point, where the electric
field is strong and regeneration by iodide is inhibited.34,35
Therefore, the thermally equilibrated excited states of
complexes 1 and 2 are well formulated as an oxidized Ru
center and a reduced deeb ligand. This is shown for complex 1
in eq 1.
RuII(deeb)2(tmam)4+ + hv
−
→ [RuIII(deeb )(deeb)(tmam)]4+*
(1)
Localization of the excited state on the ligand that interacts
with the TiO2 surface is likely important for future studies in
dye-sensitized solar cells. It also has significance for iodide
photooxidation because the excited state is localized away from
the quaternary amines that interact with iodide, as is discussed
below.
Iodide Interactions. Taken together, the results of NMR,
PL, and UV−vis absorption spectroscopy provide compelling
evidence that iodide forms ion pairs with these cationic
complexes in acetonitrile. Under the same conditions, there was
no evidence for ion pairing of RuII(deeb)(bpy)22+ with iodide.31
This suggests that the increased cationic charge of the tmam-
containing complexes was responsible for ion pairing. Indeed,
the tmam ligand was designed to facilitate ion pairing through
iodide’s interactions with the cationic charges of the quaternary
amine.
To investigate the details of iodide interactions with these
complexes, 1H NMR was utilized. The furthest downfield
resonances of the tmam ligand in all of the complexes
corresponded to the 3 and 3′ H atoms on the bipyridine
ring. These 3 and 3′ resonances shifted further downfield, by
more than 1 ppm, as iodide was titrated into the solution. This
downfield shift was previously reported for halide ion pairing in
related dicationic ruthenium(II) heteroleptic complexes in
DCM solutions.25,32 The electronegative iodide interacts with
the most acidic electropositive H atom, drawing it away from
the bipyridyl C atom and effectively deshielding it. This effect
was also “felt” by the other protons on the bipyridine rings of
tmam, which show significant yet smaller downfield shifts
(∼0.3−0.4 ppm). In contrast, the resonances associated with
the deeb ligand were largely insensitive to the addition of
iodide.
CONCLUSION
■
Three new highly charged cationic ruthenium(II) complexes
were successfully prepared and characterized based on the
tmam ligand. Two of these complexes had ester functional
groups that can be hydrolyzed for sensitization of metal oxide
electrodes. The third complex was ligated to three tmam
ligands with high symmetry in the solid state and an overall 8+
charge, making it the most highly charged mononuclear
ruthenium(II) complex ever prepared. The high cationic charge
resulted in significant ion pairing with low concentrations of
iodide in polar CH3CN solutions, a behavior that was absent
for related ruthenium(II) bipyridyl complexes with a 2+ charge.
NMR studies revealed that, by virtue of a unique “binding
pocket”, iodide preferentially interacts with the tmam ligand
over diethyl ester bipyridine ligands. Iodide photooxidation
occurred rapidly, k > 108 s−1, as was inferred by a static
component in excited-state quenching measurements. The data
demonstrate that electrostatics and acid−base chemistry can be
used to assemble well-characterized adducts of iodide and
transition-metal complexes in polar solvents. Such behavior is
important for optimizing the rates of photoinduced electron
transfer relative to excited-state relaxation and may help
A key insight into the interactions of iodide resulted from
analysis of the methylene C protons that link trimethylamine to
the bipyridine ring. Because the ruthenium complexes are
chiral, these methylene protons are diastereotopic, and doublets
G
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(23) Adamson, A. W.; Namnath, J.; Shastry, V. J.; Slawson, V. J.
Chem. Educ. 1984, 61, 221.
(24) Farnum, B. H.; Jou, J. J.; Meyer, G. J. Proc. Natl. Acad. Sci. U. S.
overcome the mass-transport limitations that occur in dye-
sensitized solar cells when high solar fluxes are employed.
A. 2012, 109, 15628.
(25) Ward, W. M.; Farnum, B. H.; Siegler, M.; Meyer, G. J. J. Phys.
Chem. A 2013, 117, 8883.
(26) Friebolin, H. Basic One- and Two-Dimensional NMR Spectros-
copy; Wiley-VCH: Hiedelberg, Germany, 2011.
(27) Kalyanasundaram, K. Photochemistry of polypyridine and
porphyrin complexes; Academic Press: San Diego, CA, 1992.
(28) Bradley, P.; Kress, N. J. Am. Chem. Soc. 1981, 1441.
(29) Blakley, R.; DeArmond, M. J. Am. Chem. Soc. 1987, 4895.
(30) Taheri, A.; Meyer, G. J. Dalton Trans. 2014, 43, 17856.
(31) Marton, A.; Clark, C. C.; Srinivasan, R.; Freundlich, R. E.;
Narducci Sarjeant, A. a.; Meyer, G. J. Inorg. Chem. 2006, 45, 362.
(32) Beer, P.; Dent, S.; Wear, T. J. Chem. Soc., Dalton Trans. 1996, 1,
2341.
ASSOCIATED CONTENT
* Supporting Information
CIF file for 3, cyclic voltammograms for 1−3, absorption
spectra and nonlinear Stern−Volmer analysis for 2 and 3, and
full NMR titrations for 1. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: gjmeyer@email.unc.edu.
Notes
■
The authors declare no competing financial interest.
(33) Rowley, J. G.; Farnum, B. H.; Ardo, S.; Meyer, G. J. J. Phys.
Chem. Lett. 2010, 1, 3132.
ACKNOWLEDGMENTS
(34) Robson, K. C. D.; Hu, K.; Meyer, G. J.; Berlinguette, C. P. J. Am.
Chem. Soc. 2013, 135, 1961.
(35) Li, F.; Jennings, J. R.; Wang, Q. ACS Nano 2013, 7, 8233.
■
The Division of Chemical Sciences, Geosciences, and
Biosciences, Office of Basic Energy Sciences of the U.S.
Department of Energy, through Grant DE-FG02-96ER14662 is
gratefully acknowledged for support. Peter White assisted with
X-ray crystallography, and Marc ter Horst assisted with NMR
spectroscopy. Javier Grajeda and the laboratory of Prof. A. J. M.
Miller assisted with ESI-MS and synthetic discussion.
REFERENCES
(1) Sutin, N. Acc. Chem. Res. 1982, 15, 275.
(2) Ardo, S.; Meyer, G. J. Chem. Soc. Rev. 2009, 38, 115.
■
(3) Pelet, S.; Moser, J.-E.; Gratzel, M. J. Phys. Chem. B 2000, 104,
̈
1791.
(4) Redmond, G.; Fitzmaurice, D. J. Phys. Chem. 1993, 97, 1426.
(5) Kelly, C. A.; Farzad, F.; Thompson, D. W.; Stipkala, J. M.; Meyer,
G. J. Langmuir 1999, 15, 7047.
(6) O’Donnell, R. M.; Ardo, S.; Meyer, G. J. J. Phys. Chem. Lett. 2013,
4, 2817.
(7) Boschloo, G.; Hagfeldt, A. Acc. Chem. Res. 2009, 42, 1819.
(8) Nazeeruddin, M. K.; Zakeeruddin, S. M.; Humphry-Baker, R.;
Jirousek, M.; Liska, P.; Vlachopoulos, N.; Shklover, V.; Fischer, C.;
Gratzel, M. Inorg. Chem. 1999, 38, 6298.
̈
(9) Qu, P.; Meyer, G. J. Langmuir 2001, 17, 6720.
(10) Finnie, K.; Bartlett, J.; Woolfrey, J. Langmuir 1998, 7463, 2744.
(11) Umapathy, S.; Cartner, A. M.; Parker, A. W.; Hester, R. E. J.
Phys. Chem. 1990, 94, 8880.
(12) Nem
̌
ec, H.; Rochford, J.; Taratula, O.; Galoppini, E.; Kuzel, P.;
̌
Polívka, T.; Yartsev, A.; Sundstrom, V. Phys. Rev. Lett. 2010, 104,
197401.
̈
(13) Clifford, J. N.; Palomares, E.; Nazeeruddin, M. K.; Gratzel, M.;
Durrant, J. R. J. Phys. Chem. C 2007, 111, 6561.
(14) Gillaizeau-Gauthier, I.; Odobel, F.; Alebbi, M.; Argazzi, R.;
Costa, E.; Bignozzi, C. A.; Qu, P.; Meyer, G. J. Inorg. Chem. 2001, 40,
6073.
(15) Dolomanov, O.; Bourhis, L. J.; Gildea, R.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 339.
(16) Bourhis, L. J.; Dolomanov, O. V.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. Acta Crystallogr., Sect. A 2015, 59.
(17) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 112.
(18) Crosby, G. A.; Demas, J. N. J. Phys. Chem. 1971, 75, 991.
(19) Pavlishchuk, V. V.; Addison, A. W. Inorg. Chim. Acta 2000, 298,
97.
(20) Li, J.-H.; Wang, J.-T.; Hu, P.; Zhang, L.-Y.; Chen, Z.-N.; Mao,
Z.-W.; Ji, L.-N. Polyhedron 2008, 27, 1898.
(21) Yu, Z.; Najafabadi, H. M.; Xu, Y.; Nonomura, K.; Sun, L.; Kloo,
L. Dalton Trans. 2011, 40, 8361.
(22) Bard, A.; Faulkner, L. Electrochemical Methods Fundamentals and
Applications; John Wiley & Sons, Inc.: Hoboken, NJ, 2001.
H
DOI: 10.1021/acs.inorgchem.5b00344
Inorg. Chem. XXXX, XXX, XXX−XXX