Synthesis and characterization of rigid +2 and +3 heteroleptic dinuclear ruthenium(II) complexes

Article abstract of DOI:10.1016/j.poly.2010.06.012

Synthesis and characterization of the dinuclear ruthenium coordination complexes with heteroleptic ligand sets, [Cl(terpy)Ru(tpphz)Ru(terpy)Cl](PF ₆)₂ (7) and [(phen)₂Ru(tpphz)Ru(terpy)Cl] (PF₆)₃ (8), are reported. Both structures contain a tetrapyrido[3,2-α:2′,3′-c:3″,2″-h:2″, 3″-j]phenazine (tpphz) (6) ligand bridging the two metal centers. Complex 7 was obtained via ligand exchange between, RuCl₂(terpy)DMSO (5) and a tpphz bridge. Complex 8 was obtained via ligand exchange between, [Ru(phen)₂tpphz](PF₆)₂ (4) and RuCl ₂(terpy)DMSO (5). Metal-to-ligand-charge-transfer (MLCT) absorptions are sensitive to ligand set composition and are significantly red-shifted due to more electron donating ligands. Complexes 7-9 have been characterized by analytical, spectroscopic (IR, NMR, and UV-Vis), and mass spectrometric techniques. The electronic spectral properties of 7, 8, and [(phen) ₂Ru(tpphz)Ru(phen)₂](PF₆)₄ (9), a previously reported +4 analog, are presented together. The different terminal ligands of 7, 8, and 9 shift the energy of the MLCT and the π-π* transition of the bridging ligand. These shifts in the spectra are discussed in the context of density functional theory (DFT). A model is proposed suggesting that low-lying orbitals of the bridging ligand accept electron density from the metal center which can facilitate electron transfer to nanoparticles like single walled carbon nanotubes and colloidal gold.

Full text of DOI:10.1016/j.poly.2010.06.012

Polyhedron 29 (2010) 2696–2702
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Synthesis and characterization of rigid +2 and +3 heteroleptic dinuclear
ruthenium(II) complexes
*
J.R. Alston, S. Kobayashi, T.J. Younts, J.C. Poler
Chemistry Department, The University of North Carolina at Charlotte, 9201 University City Blvd., Charlotte, NC 28223, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthesis and characterization of the dinuclear ruthenium coordination complexes with heteroleptic
Received 30 April 2010
Accepted 18 June 2010
Available online 25 June 2010
ligand sets, [Cl(terpy)Ru(tpphz)Ru(terpy)Cl](PF6)2 (7) and [(phen)2Ru(tpphz)Ru(terpy)Cl](PF6)3 (8),
are reported. Both structures contain a tetrapyrido[3,2-
ligand bridging the two metal centers. Complex 7 was obtained via ligand exchange between, RuCl2-
(terpy)DMSO (5) and tpphz bridge. Complex was obtained via ligand exchange between,
a
:20,30-c:300,200-h:200,300-j]phenazine (tpphz) (6)
a
8
Keywords:
Multinuclear
Coordination complex
Polypyridine ligand
Charge transfer
MLCT
[Ru(phen)2tpphz](PF6)2 (4) and RuCl2(terpy)DMSO (5). Metal-to-ligand-charge-transfer (MLCT) absorp-
tions are sensitive to ligand set composition and are significantly red-shifted due to more electron donat-
ing ligands. Complexes 7–9 have been characterized by analytical, spectroscopic (IR, NMR, and UV–Vis),
and mass spectrometric techniques. The electronic spectral properties of 7, 8, and [(phen)2Ru(tpphz)-
Ru(phen)2](PF6)4 (9), a previously reported +4 analog, are presented together. The different terminal
ligands of 7, 8, and 9 shift the energy of the MLCT and the p–p* transition of the bridging ligand. These
shifts in the spectra are discussed in the context of density functional theory (DFT). A model is proposed
suggesting that low-lying orbitals of the bridging ligand accept electron density from the metal center
which can facilitate electron transfer to nanoparticles like single walled carbon nanotubes and colloidal
gold.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
tions due to the electron density distributions and net molecular
charge on the compounds.
Understanding directed self-assembly of supramolecular sys-
tems is critical to the development of nanoscale devices. Dendri-
mer chemistry continues to provide both building blocks and
scaffolds for supramolecular and nanodevice engineering [1–4].
Metallodendrimers are of broad interest because of their potential
use in light harvesting, sensing, signal amplification, and electron-
transfer reagents in water splitting applications [4–6]. Optically
and electrochemically active ruthenium metallodendrimers
possess several properties which enable their versatility in supra-
molecular systems. By changing the morphology and the local
charge density of the ruthenium complexes, we have studied
how these mononuclear and decanuclear metallodendrimers
(decamers) [7] interact with and coordinate to single walled car-
bon nanotubes (SWNT) [2,3,8]. Towards this end, novel dinuclear
ruthenium coordination complexes (dimers) have been synthe-
sized. The results presented below will enable further studies of
the optoelectronic properties of ruthenium coordination com-
pounds and their interactions with nanoparticles. Specifically these
new compounds will help us discriminate between the supramo-
lecular interactions due to molecular morphology and the interac-
We have synthesized two structurally similar ruthenium di-
mers with differing electrostatic charges. Both complexes are com-
posed of two ruthenium monomer units bridged by the aromatic
tetrapyrido[3,2-a
:20,30-c:300,200-h:200,300-j]phenazine (tpphz) ligand.
Coordinating various ligands to the ruthenium monomer precur-
sors allows us to specify the charge state of the resulting dimer
as either +2, +3, or +4. The new ruthenium complexes produced
through this method are [Cl(terpy)Ru(tpphz)Ru(terpy)Cl](PF6)2
(7) and [(phen)2Ru(tpphz)Ru(terpy)Cl](PF6)3 (8). These complexes
are similar in structure to the +4 complex [(phen)2Ru(tpphz)Ru-
(phen)2](PF6)4 (9), previously reported by MacDonnell et al. [9].
In addition to the ruthenium dimer synthesis we report a mixed
solvent synthesis method for their precursor, Ru(DMSO)4Cl2 (3).
By using a mixed solvent, product readily precipitates from solu-
tion as it is formed, simplifying collection while maintaining high
purity with yields of approximately 90% [10,11].
By comparing our new compounds to 9, we were able to inves-
tigate how different electrostatic charges affect the optical proper-
ties of structurally similar ruthenium complexes. Our results show
that the MLCT absorption is more red-shifted as the charge state is
lowered. DFT level calculations of the highest occupied molecular
orbital (HOMO) and the lowest unoccupied molecular orbital
(LUMO) of the +2, +3, and +4 dimers suggest that electron density
* Corresponding author. Tel.: +1 704 687 8289; fax: +1 704 687 3151.
E-mail address: jcpoler@uncc.edu (J.C. Poler).
0277-5387/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2010.06.012

J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
2697
shifts from the ruthenium centers to the tpphz bridge during the
MLCT transition. The dimensions and molecular rigidity of 7, 8,
and 9 are configured to interact effectively with nanoparticles. Cat-
alytic metal nanoparticles, quantum dots and rods, and nanotubes
are targets for non-covalent, site specific functionalization and sen-
sitization by these coordination compounds. Because we can effec-
tively manipulate the optoelectronic properties of these ruthenium
complexes and the overall charge-state while maintaining their ri-
gid structure, we aim to use complexes like 7, 8, and 9 as potential
building blocks and integral components in the design of supramo-
lecular nanoscale systems.
at 0 °C for 12–15 h. The resulting purple/black solid precipitate
was collected by filtration. The product was washed with acetone
(50 mL, 2ꢂ) and then several times with ice cold water (30 mL,
3ꢂ), yielding a dark purple solid. This compound is used best with-
out further purification due to degradation during recrystallization.
Yield: 6.85 g, 65%; UV–Vis (DMF) kMLCT 552 nm
(e = 3400
Mꢁ1 cmꢁ1).
2.2.2. [Ru(phen)2phendione](PF6)2 (2)
A mixture of absolute ethanol (50 mL) degassed with Ar(g), com-
plex 1 (0.408 g, 0.80 mmol), and phendione (0.210 g, 1.0 mmol)
was stirred and refluxed under an inert atmosphere of Ar(g) for
24 h. After the mixture was cooled to RT, a saturated KPF6(aq) solu-
tion was added dropwise to precipitate the brown product. The
solution was kept at 4 °C for 12–15 h and the brown solid was col-
lected by filtration. The solid was washed with cold ethanol (95%,
20 mL, 3ꢂ), then with cold absolute ethanol (20 mL, 2ꢂ), and
finally with diethyl ether (20 mL, 2ꢂ). The yellow brown solid
was air dried. Yield: 0.490 g, 66%; UV–Vis (MeCN) kMLCT 430 nm
2. Experimental
2.1. Materials and methods
1,10-Phenanthroline (phen), RuCl3ꢀ3H2O, dimethyl sulfoxide
(DMSO), 2,20;60,200-terpyridine (terpy), hydrazine hydrate, and pal-
ladium/carbon catalyst were purchased from Sigma–Aldrich or
Alfa and were used without further processing or purification
unless otherwise noted. 1,10-Phenanthroline-5,6-dione (phendi-
one) [12,13], 1,10-phenanthroline-5,6-diamine (phendiamine),
and tpphz (6) [14] were prepared via literature procedures with
any variations noted herein. cis-RuCl2(phen)2 (1) [15] was prepared
via literature procedures substituting phen for bipyridine (bipy).
[Ru(phen)2phendione](PF6)2 (2) [13] and RuCl2(DMSO)4 (3)
[10,11,16,17] were prepared through a slight adaptation of litera-
ture procedures. RuCl2(terpy)DMSO (5) [18] was prepared with
slight variations to literature procedures with changes noted be-
low. [(phen)2Ru(tpphz)Ru(phen)2](PF6)4 (9) was prepared follow-
ing literature methods [9,14]. All solvents were analytical grade
and dried on molecular sieves when applicable.
(e
= 14,000 Mꢁ1 cmꢁ1).
2.2.3. cis,fac-RuCl2(DMSO-S)3(DMSO-O) (3)
DMSO (10 mL) was degassed with Ar(g) for 20 min. RuCl3ꢀ3H2O
(3.247 g, 12.0 mmol) was added to the DMSO and stirred under
Ar(g) until dissolved. Isopropanol (35 mL) was then added, and
the light orange solution was heated at 85 °C under an inert atmo-
sphere of Ar(g) for 30 h. During heating, the product readily precip-
itated from solution. After the reaction was cooled to RT, the bright
yellow precipitate was broken up with a glass rod and collected by
vacuum filtration, washed liberally with dry acetone, rinsed liber-
ally with toluene, and finally dried on the filter. Yield: 5.240 g, 87%;
UV–Vis (MeCN) kmax 356 nm (
e
= 380 Mꢁ1 cmꢁ1). Selected FT-IR
(DMSO-O) 923.4 (s).
absorptions, cmꢁ1
:
m
(DMSO-S) 1083.1 (s), m
C, H, and N elemental analysis was performed by Atlantic
Microlab, Inc. 1H and 13C NMR spectra of 7, 8, and 9 in DMSO-d6
were obtained at room temperature on a JEOL 500 MHz Fourier
Transform (FT) spectrometer unless otherwise noted. Chemical
shifts are reported in parts per million (ppm) using TMS as an
internal reference for 1H and 13C. Assignment of 1H NMR peaks
for 7 and 8 were accomplished by high resolution phase sensitive
COSY 2D 1H NMR. ESI mass spectrometry was performed in posi-
tive ion mode on a Mariner Biospectrometry Workstation in aceto-
nitrile. All m/z are reported for the most abundant isotope and
[Ru2ꢁXPF6ꢁ] represents the dinuclear species without X of its
associated PF6ꢁ counterions. Electronic absorption spectra were re-
corded on a Varian Cary 5000 spectrometer in dimethylformamide
(DMF) or acetonitrile (MeCN). Infrared (IR) spectra were obtained
on a Perkin–Elmer Spectrum One FT-IR with a universal diamond
attenuated total reflectance top plate. Melting/degradation points
(MPdeg) of 7 and 8 were measured on a Mettler-Toledo Thermo-
gravimetric Analysis TGA/SDTA 851e under N2(g) atmosphere. Pho-
todegradation studies were done in air saturated DMF under a
5 mW illumination of a 450 ( 5) nm source for 12 h. Limiting molar
2.2.4. [Ru(phen)2tpphz](PF6)2 (4)
A mixture of 2 (0.200 g, 0.20 mmol) in acetonitrile (10 mL) was
stirred and brought to a boil. A separate volume of methanol
(35 mL) was heated to
a boil then phendiamine (0.051 g,
0.20 mmol) was dissolved into the hot methanol. The hot phendi-
amine solution was added to the boiling acetonitrile solution con-
taining 2 and brought to reflux while stirring under Ar(g) for 5 h.
The reaction was cooled to RT, and then saturated KPF6(aq) was
added dropwise until the brown precipitate stopped forming. The
precipitate was collected by vacuum filtration and the solid
washed with ice cold water (20 mL, 2ꢂ), cold ethanol (95%,
20 mL, 2ꢂ), and finally with diethyl ether (20 mL, 2ꢂ). The
brown solid was dried on the filter. Yield: 0.168 g, 71%; UV–Vis
*
(MeCN) ktpphz
385 nm (
e
= 28,000 Mꢁ1 cmꢁ1), kMLCT 455 nm
p–p
ꢁ 2+
(
e
= 19,700 Mꢁ1 cmꢁ1).
ESI-MS:
[Ruꢁ2PF6
]
m/z = 423.1,
ꢁ +
[Ruꢁ1PF6
]
m/z = 192.2. 1H NMR: 300 MHz (CD3CN) d (ppm):
9.80 (d, J = 8.25 Hz, 2H), 9.71 (d, J = 7.32 Hz, 2H), 9.01 (d, 2H),
8.65 (d, J = 2.76 Hz, 2H), 8.62 (d, J = 2.76 Hz, 2H), 8.40 (d, J =
4.77 Hz, 2H), 8.23 (s, 4H), 8.19 (d, J = 5.52 Hz, 2H), 8.06 (d,
J = 5.13 Hz, 2H), 7.92 (dd, J = 8.43, 2H), 7.81 (dd, J = 8.04, 2H),
7.68 (m, 4H).
conductivity (K
°m, S cm2 molꢁ1) in DMF was determined from
Kohlrausch’s law using conductivity data from an Accumet AR20
calibrated with a KCl(aq) standard.
2.2.5. RuCl2(terpy)DMSO (5)
2.2. Synthesis of ruthenium coordination complexes
A mixture of ethanol (95%, 20 mL) and methanol (6 mL) was
slowly stirred as 3 (1.263 g, 2.20 mmol) was added. The mixture
was stirred and refluxed under an inert Ar(g) atmosphere for
15 min. A separate mixture of ethanol (95%, 10 mL) and terpy
(0.604 g, 2.6 mmol) was prepared and slowly transferred under
Ar(g) to the refluxing solution of 3. The resulting brown mixture
was refluxed and stirred under Ar(g) for another 8.5 h. The brown
precipitate was collected by vacuum filtration. The solid was
washed with cold water (20 mL, 3ꢂ), and then cold ethanol (95%)
2.2.1. cis-RuCl2(phen)2 (1)
LiCl and RuCl3ꢀ3H2O were dried under vacuum at 125 °C for
15 h.
A mixture of LiCl (8.89 g, 210 mmol), RuCl3 (5.33 g,
25.0 mmol), phen (7.11 g, 40.0 mmol), and DMF (40 mL) was stir-
red and refluxed for 15 h. The purple/black solution was cooled
to room temperature (RT) and then rinsed from the reaction flask
and diluted with acetone (500 mL). This solution was then kept

2698
J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
until the filtrate was clear and colorless. The brown product was
N
12–Ru2P2F12ꢀ5H2O: C, 43.12; H, 2.95; N, 11.18. Found: C, 43.13;
vacuum dried at 90 °C. Yield: 0.916 g, 88%; UV–Vis (MeCN) kMLCT
H, 2.60; N, 10.87%. Limiting molar conductivity,
K
°m: 230
ꢁ 2+
ꢁ +
513 nm (
e
= 3000 Mꢁ1 cmꢁ1).
S cm2 molꢁ1
.
ESI-MS: [Ru2ꢁ2PF6
]
m/z = 562.1, [Ru2ꢁ1PF6
]
m/z = 1269.3. 1H NMR (DMSO-d6) d (ppm): 10.57 (d, J = 5.15 Hz,
tpphz), 10.53 (d, J = 4.60 Hz, tpphz), 10.38 (d, J = 8.30 Hz, tpphz),
10.21 (d, J = 5.70 Hz, tpphz), 10.18 (d, J = 9.15 Hz, tpphz), 9.82 (d,
J = 8.05 Hz, tpphz), 9.62 (d, 8.00 Hz, tpphz), 8.89 (d, J = 8.05 Hz, ter-
py 2H), 8.84 (d, J = 8 Hz, tpphz), 8.37–8.24 (m, tpphz and terpy 4H),
8.37–8.24 (m, tpphz and terpy 4H), 8.04–7.90 (m, tpphz and terpy
4H), 7.73 (d, J = 6.30 Hz, terpy 4H), 7.77–7.62 (m, tpphz), 7.44
(apparent dd, J = 6.30 Hz, tpphz), 7.32–7.27 (m, tpphz and terpy
4H); 13C NMR (DMSO) d 158.24, 157.51, 152.48, 151.60, 149.54,
137.11, 127.21, 123.65, 122.73.
2.2.6. [Cl(terpy)Ru(tpphz)Ru(terpy)Cl](PF6)2ꢀ5H2O (7)
A mixture 0.166 g (0.40 mmol) of 6 and 0.415 g (0.90 mmol) of
complex 5 were added to a mixed solvent solution of ethanol (95%,
15 mL) and water (15 mL). The resulting solution was stirred and
refluxed for 24 h. After the reaction was cooled to RT, the dark pur-
ple mixture was brought to dryness under vacuum, and then the
residue was dissolved in a minimal amount of water (ꢃ30 mL).
The dark purple solution was added dropwise to a saturated
KPF6(aq) solution to yield a precipitate. The purple solid was col-
lected by vacuum filtration, and then washed with water until
the filtrate was clear and colorless. The solid was dried under vac-
2.2.7. [(phen)2Ru(tpphz)Ru(terpy)Cl](PF6)3ꢀ2H2O (8)
uum at 90 °C. Yield: 0.468 g, 77%; UV–Vis (DMF) ktpphz
*
375 nm
A mixture of 4 (0.105 g, 0.10 mmol), 5 (0.046 g, 0.10 mmol), and
DMF (15 mL) were stirred and refluxed under Ar(g) for 23 h. The
reaction was cooled to RT, and then saturated KPF6(aq) was added
p–p
(e
= 33,000 Mꢁ1 cmꢁ1), kMLCT 518 nm (
e
= 28,000 Mꢁ1 cmꢁ1). Melt-
ing point of degradation (MPdeg): 356 °C. Anal. Calc. for C54H34
-
Scheme 1. Reaction sequence for the synthesis of +2 Ru dimer (7) and +3 Ru dimer (8).

J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
2699
dropwise, producing a reddish precipitate. The mixture was then
held at 4 °C for 12 h. The dark orange-red solid was collected by
vacuum filtration, washed with cold water (20 mL, 2ꢂ), and then
cold ethanol (95%, 20 mL, 2ꢂ), and finally with diethyl ether
(20 mL, 2ꢂ) and dried under vacuum at 90 °C for 12 h. Yield:
as a reactant and the primary solvent. In these cases the cis product
is very soluble in DMSO while the trans product can precipitate as
the reaction mixture cools [16,17,20,21]. Temperature tends to be
the isoform determining factor during these reactions. During our
synthesis procedure we chose to use DMSO as a reactant in slight
excess and use isopropanol (IPA) as the primary solvent. Isopropa-
nol was selected as a solvent due to the limited solubility of
RuCl2(DMSO)4 which had the intended consequence of precipitat-
ing product as it was formed, simplifying the collection of product.
The solubility of cis isoform in IPA is 0.6 mg/mL while the solubility
of the trans isoform is 5.0 mg/mL.
0.087 g, 52%; UV–Vis (DMF) ktpphz
*
375 nm
(e = 35,000
p–p
Mꢁ1 cmꢁ1), kMLCT 453 nm (
e
= 26,000 Mꢁ1 cmꢁ1). MPdeg: 370 °C.
Anal. Calc. for C63H39N13–Ru2P3F18ꢀ2H2O: C, 44.86; H, 2.57; N,
10.80. Found: C, 44.47; H, 2.46; N, 10.87%. Limitinꢁg 3m+ olar conduc-
tivity,
K
°m: 310 S cm2 molꢁ1. ESI-MS: [Ru2ꢁ3PF6
]
m/z = 405.4,
ꢁ 2+
ꢁ +
[Ru2ꢁ2PF6
]
m/z = 680.5, [Ru2ꢁ1PF6
]
m/z = 1506.1. 1H NMR
(DMSO-d6)
d
(ppm): 10.54 (d, J = 5.20 Hz, 1H), 10.26 (d,
J = 8.05 Hz, 1H), 10.18 (d, J = 7.45 Hz, 1H), 9.97 (d, J = 8.00, 1H),
9.70 (d, J = 8.00 Hz, 1H), 8.91 (d, J = 8.05 Hz, 1H), 8.81 (app. t, J =
8.00 Hz, 4H), 8.76 (d, J = 8.00 Hz, 2H), 8.72 (dd, 8 Hz, 1H), 8.43 (d,
J = 6.90 Hz, 4H), 8.34–8.26 (m, 4H), 8.12–8.06 (m, 5H), 8.01–7.95
(m, 3H), 7.85–7.77 (m, 5H), 7.71 (app. d, 2H), 7.30 (dd, 12.6 Hz,
3.2. Synthesis and stability of 7 and 8
Reaction Scheme 1 illustrates the two synthetic routes through
which we manipulated the final charge state and optical properties
of 7 and 8. The phen derivatives, 1,10-phenanthroline-5,6-dione
[12,13], 1,10-phenanthroline-5,6-diamine [14], and tpphz [14]
were produced via previously reported synthetic procedures. Com-
plex 7 was synthesized from a ligand exchange reaction of two mo-
lar equivalents of 5 and one molar equivalent of 6. Complex 8 was
synthesized from a ligand exchange reaction between 5 and the
PF6 salt of 4. Following synthesis, compounds 7 and 8 were charac-
terized by EA, UV–Vis, SDTA, ESI-MS, 1H NMR, and 13C NMR. ESI-
MS was essential to identifying and characterizing the compounds.
There are seven ruthenium isotopes with large enough natural
abundances to be measured in ESI-MS. The signature of a mononu-
clear complex is easily distinguished (7 isotope peaks) from the
isotope patter of a dinuclear species (13 easily identified peaks
for one complex). There was no observable monomer in the ESI-
MS of any of the dimers reported here. The dimers show excellent
stability in solution. Each dimer has a unique mass to charge (m/z)
pattern. The calculated isotope patterns of 7, 8, and 9 accurately
match the measured mass spectrum (Supplementary Fig. S2). The
difference in m/z within the isotope pattern allows us to determine
the absolute charge on the species. When the resulting isotope
peaks are separated by 1.0 m/z unit the overall electrostatic charge
on the complex is +1. Correspondingly, the separation is 0.50 m/z
units for +2 and 0.33 m/z units for +3, and 0.25 m/z units for a +4
charge.
3H); 13C NMR (DMSO-d6)
d 158.26, 157.51, 154.42, 153.17,
152.50, 152.49, 151.71, 150.33, 149.64, 147.05, 139.69, 137.16,
136.92, 130.40, 129.40, 128.63, 127.99, 127.21, 126.31, 123.68,
122.77.
3. Results and discussion
3.1. Synthesis of 3
Addition of RuCl3ꢀ3H2O in DMSO to give 3 can yield either cis or
trans isomers [16]. X-ray crystallography has previously shown the
cis product coordinates one of the DMSO ligands through the oxy-
gen and the other three through the sulfur while the trans product
coordinates DMSO exclusively through sulfur [17]. The kinetically
stable isomer, trans-RuCl2(DMSO-S)4, should be the most
likely product at low temperatures (<100 °C) [19], whereas the
thermodynamically stable isomer, cis,fac-RuCl2(DMSO-S)3(DMSO-
O), should be the primary product of RuCl3 in hot (120–150 °C)
DMSO [11,17]. The synthesis procedure described above was per-
formed at low temperature (85 °C) in mixed solvent. Under low
temperature, mixed solvent conditions, we observe the synthesis
of pure cis,fac-RuCl2(DMSO-S)3(DMSO-O). Using FT-IR spectros-
copy, the as-produced product was analyzed to determine the ori-
entation of the coordinated DMSO from our mixed solvent
procedure. The S@O stretching frequency of uncoordinated DMSO
is approximately 1055 cmꢁ1 [20]. This transition is not found in
our IR therefore we do not observe any uncoordinated DMSO in
this sample. The stretching frequency of Ru–S-bonded DMSO is re-
ported to resonate at 1080 cmꢁ1. While the Ru–O-bonded DMSO
stretching frequency is reported to appear between 890 and
950 cmꢁ1 [20,21], we experimentally detected stretching frequen-
cies at 923.4 cmꢁ1 and 1083.1 cmꢁ1 from our as-produced sample
suggesting the cis product. Following procedures reported by Ales-
sio et al. [16] we converted samples of RuCl2(DMSO)4 to the pure
cis and pure trans isoforms. IR spectra of the pure samples were
overlaid with that of the as-produced sample and compared (Sup-
plementary Fig. S1). The spectra of the pure cis,fac-RuCl2(DMSO-
S)3(DMSO-O) is identical to that of our as-produced product. The
spectra of the pure trans-RuCl2(DMSO-S)4 not only lacks the
stretching frequency of Ru–O-bonded DMSO, the frequency of
Ru–S-bonded DMSO is at a lower energy,
m
m
1076.7 cmꢁ1, versus
1083.1 cmꢁ1 in the cis spectra. The lower vibrational transition
energy is likely caused by the decrease of ruthenium electron
donation due to the additional S-bonded DMSO competition [20].
Exclusive production of the thermodynamically favored cis iso-
form even at low reaction temperatures can be explained by the
solubility of the different isomers in the mixed solvent environ-
ment. To produce RuCl2(DMSO)4 it is common to use DMSO both
Fig. 1. UV–Vis absorption spectra of complexes 7, 8, and 9. Complex 7 kMLCT 518 nm
(e e
= 28,000 Mꢁ1 cmꢁ1), complex 8 kMLCT 453 nm ( = 26,000 Mꢁ1 cmꢁ1), and com-
plex 9 kMLCT 444 nm (e
= 41,000 Mꢁ1 cmꢁ1).

2700
J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
Thermal stability of the ruthenium coordination complexes 7, 8,
2D spectroscopy, the proton signals of 7 and 8 were assigned.
The large distance between the ancillary ligands of 9 and the tpphz
bridge make it so the 1H NMR spectra of all eight magnetically
inequivalent protons belonging to the same ancillary phens over-
lap identically [22]. The magnetic patterns of 7 and 8 are complex
(Supplementary Fig. S4). The protons of the terpy ligands are easily
distinguished. However, the asymmetric heteroleptic arrangement
of 8 as well as the presence of syn and anti forms of 7 produce
anisotropic magnetic effects on the pyridine cycle of tpphz. The
magnetic inequivalence of the tpphz protons presents 12 distinct
proton signals in the spectra of 8. For 7, there are two isomers.
The chloride ligand on either ruthenium core can be in syn and anti
arrangements. These isomers are likely formed in equal amounts.
The presence of the isomers in solution can be observed by the
shielding effect of the chloride ligand on the proximal H2 protons
of tpphz. The shielding reduces the J coupling to approximately
5 Hz between the H2 and H3 protons of the tpphz. Complex 8
has two diastereomers which produce similar chloride shielded
H2 signals at 10.54 ppm. The two isomers of 7 will have slight dif-
ferences in H2 resonances and produce three shielded signals at
and 9 were measured using scanning differential thermal analysis,
SDTA (Supplementary Fig. S3). The dimers were thermally stable
under N2(g) up to 356 °C. Complex 7 began decomposing at
356 °C, noted as the melting point onset temperature of degrada-
tion (MPdeg) for the PF6 salt. Complexes 8 and 9 followed suit,
beginning decomposition at 370 °C and 380 °C, respectively. Photo-
degradation of 7, 8, and 9 evaluated with UV–Vis showed no mea-
surable decomposition after 12 h of 5 mW/cm2 of 450 ( 5) nm
illumination. A series of 1H NMR spectra as a function of time
showed that the ruthenium coordination complexes are air stable.
Uncoordinated phendiamine was prone to degradation. We use the
phendiamine immediately after it was prepared to synthesize 6.
The other uncoordinated polypyridine ligands such as phen-5,6-
dione, 5-nitro-phen and terpy are air and light stable.
3.3. 1H NMR of 7, 8, and 9
On the basis of previous reports [7,14,22,23] and the spectra of
9, used in conjunction with high resolution phase sensitive COSY
Fig. 2. Frontier molecular orbitals for dinuclear Ru complexes. (A, C, E) illustrate the +4, +3, and +2 dimers, respectively looking along the tpphz edge, and (B, D, F) illustrate
+4, +3, and +2 dimers, respectively looking perpendicular to the tpphz ligand. The HOMO is depicted on the Ru center and colored green–red; the LUMO is located across the
tpphz bridge and colored blue–yellow. Notice the LUMO is similar across species whereas the HOMO at the Ru centers varies depending on the presence of chloride ligand.
The only significant difference between the enhanced dz2 character on the Ru center of the +4 dimer (A, B) as compared to the orbital on the chloride on the +2 dimer (E, F).
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
2701
10.57, 10.53, and 10.21 ppm. COSY 2D NMR aided the assignment
of the tpphz protons. To distinguish between the protons, the desh-
ielding effects of the ancillary ligands were considered. Those
tpphz protons opposite phen ligands would be furthest downfield,
followed by those opposite terpy and then chloride [24]. For sim-
plicity, proton peaks presented for 7 are integrated only for the ter-
py signals. These observations are in complete agreement with the
data reported for the mononuclear [Ru(L)2tpphz]2+ complexes
[L = phen or bipy] [22,23].
the metal center and the chloride ligand. In contrast, the HOMO
through HOMO-5 orbitals on 8 all resided on the metal center with
the chloride ligand. We have also completed excited state calcula-
tions using configuration interaction singles with the semi-empir-
ical Zerner’s Independent Neglect of Differential Overlap (ZINDO)
Hamiltonian. The transition energies, determined using this model,
are consistent with our observation that the MLCT transition is red-
shifted as the charge state is lowered. These computational models
are consistent with our spectrometric observations.
3.4. Electronic spectral studies
4. Conclusions
The absorption spectra of 7, 8, and 9 in DMF are shown in Fig. 1.
As observed for uncoordinated polypyridine ligands and mononu-
clear ruthenium complexes the ruthenium dimer complexes retain
similar spectral features. The absorption spectra for compounds 7,
8, and 9 show an absorption at approximately 375 nm with a high-
er energy shoulder for 9 which evolves into another peak for 7 and
In summary, in this work we have presented the synthesis of
two new heteroleptic dinuclear ruthenium(II) coordination com-
plexes, [Cl(terpy)Ru(tpphz)Ru(terpy)Cl](PF6)2 (7) and [(phen)2Ru-
(tpphz)Ru(terpy)Cl](PF6)3 (8). It has been demonstrated that the
two compounds are stable and have been produced in good yield.
The optical absorption of these complexes, specifically the MLCT
transitions have been shown to be influenced by the electron with-
drawing nature of ligands opposite the tpphz bridging ligand. Both
complexes exhibit characteristic electronic absorption spectra that
correlate well with the morphologically similar +4 dimer (9) and
with the calculated HOMO and LUMO of both complexes. It is
important to note that regardless of the varying terminal ligands
or the resulting charge state of the dimer, the calculated molecular
orbitals show that the LUMO maintains a similar shape on all three
complexes while predominantly residing on the bridging ligand
between the two ruthenium centers. Our ZINDO calculated transi-
tion energies lack accuracy, but are consistent with our observed
trends in the shifting MLCT bands of 7, 8, and 9. Time dependent
DFT studies are required to further analyze the differences in these
spectra. Presently, our data supports previous work suggesting that
ruthenium complexes bridged by tpphz could interact with nano-
scale materials through the extension of that complex’s electron
density [1–3]. This phenomena can be stimulated by optical excita-
tion of a MLCT band [3] which we have shown resides solely on the
bridging ligand and in the nanoscale pocket formed by the rigid
structure.
The complexes synthesized herein are new members of a nas-
cent family of rigid ruthenium complexes that have optical proper-
ties that could be useful for controlling nanoparticle interactions as
well as being utilized in charge transfer applications. Using the
synthetic strategies described above we are able to shift the MLCT
band to longer wavelengths enabling a more efficient overlap with
the solar spectrum, which implies more utility for involvement in
photovoltaic and photocatalytic applications. Preliminary results
show strong interactions with nanoparticles such as carbon
nanotubes and metal nanoparticles and show great potential for
becoming building blocks and scaffold for technologically viable
supramolecular structures with specified shape and charge
[1,3,8]. The key to this goal is a better understanding of how mor-
phology and charge state of supramolecular systems affects their
optoelectronic properties. Further study and synthesis of photoac-
tive ruthenium coordination complexes as well as study of nano-
structures that integrate rigid coordination complexes will
hopefully contribute to this goal.
8. These two bands correspond to the (tpphz)n–p* and (tpphz)p–
p
* transitions [14]. This * transition is red-shifted compared
p–p
to the corresponding kmax of 365 nm for uncoordinated tpphz
due to the back-donation from the ruthenium (Supplementary
Fig. S5).
Each ruthenium dimer displays a strong and characteristic
MLCT band at 400–600 nm. For 7 the MLCT bands are attributed
to the overlap of Ru ? tpphz(
bands of 8 appear more broad which is attributed to the overlap
of Ru ? phen( *) and Ru ? tpphz( *) from one half of the dimer
combining with the overlap of Ru ? tpphz( *) and Ru ? terpy( *)
p*) and Ru ? terpy(p*). The MLCT
p
p
p
p
from the other half. For comparison the absorption spectra of 9 in
Fig. 1 and the spectra of polypyridine coordinated ruthenium
monomers (Supplementary Fig. S6) are included. The MLCT bands
of 9 have been reported and assigned to the corresponding transi-
tions, Ru ? phen(p*) and Ru ? tpphz(p*), which correlate well
with our observations [7,14,22,25]. A red-shift from 9 of the MLCT
bands for 7 and 8 is observed, with decreasing energy relative to
their formal charge. The electron density over ruthenium, the loca-
tion of the HOMO (vide infra), is increased due to the coordination
of chloride ligands and also by the smaller back-donation from
ruthenium to terpy versus ruthenium to two phens. The increased
electron density narrows the energy difference between the HOMO
and LUMO of the molecule, shifting the MLCT transition to the red
[14].
3.5. Molecular modeling
DFT calculations were completed with Materials Studio 4.4
(Accelrys) using a gradient corrected Becke, Lee, Yang, and Parr
functional. All core electrons were treated fully and with relativis-
tic effects. A self consistent field tolerance of 10ꢁ6 Ha and geometry
optimization tolerance of 10ꢁ5 Ha were imposed with a maximum
atom displacement of 0.005 Å (or a maximum force gradient of
0.002 Ha/Å). All optimizations were treated with a conductor-like
screening model with a dielectric constant of 36.7 to simulate
DMF, the solvent used for our electronic spectral studies. Frontier
molecular orbitals for the three dimers are shown (Fig. 2).
Although 7 and 8 have electron donating chlorides on at least
one of their ruthenium centers, the shape and distribution of the
LUMO across the tpphz bridge is almost identical between com-
pounds7, 8, and 9. Interestingly, the most prominent difference be-
tween each compound was observed in the HOMO. Orbitals on the
Acknowledgement
Acknowledgement is made to the Donors of the American
Chemical Society Petroleum Research Fund for partial support of
this research. Financial support to JRA through the Nanoscale Sci-
ence Ph.D. program and the Center for Optoelectronics and Optical
Communications at UNC Charlotte is appreciated. Financial sup-
port to SK through a Thomas Walsh Fellowship and UNC Charlotte
2
ruthenium centers of 9, the +4 complex, had very strong dz char-
acter. The HOMO through HOMO-5 orbitals of 9 were similar on
the two metal centers and nearly degenerate (<7 ꢂ 10ꢁ4 Ha). The
HOMO through HOMO-5 orbitals of 7 were also similar and nearly
degenerate (<7 ꢂ 10ꢁ4 Ha). These molecular orbitals overlap both

2702
J.R. Alston et al. / Polyhedron 29 (2010) 2696–2702
[9] F.M. MacDonnell, S. Bodige, Inorg. Chem. 35 (1996) 5758.
[10] B.R. James, E. Ochiai, G.I. Rempel, J. Inorg. Nucl. Chem. 7 (1971) 781.
[11] I.P. Evans, A. Spencer, G. Wilkinso, J. Chem. Soc.-Dalton Trans. (1973) 204.
[12] J.E. Dickeson, L.A. Summers, Aust. J. Chem. 23 (1970) 1023.
[13] C. Hiort, P. Lincoln, B. Norden, J. Am. Chem. Soc. 115 (1993) 3448.
[14] J. Bolger, A. Gourdon, E. Ishow, J.P. Launay, Inorg. Chem. 35 (1996) 2937.
[15] B.P. Sullivan, D.J. Salmon, T.J. Meyer, Inorg. Chem. 17 (1978) 3334.
[16] E. Alessio, Chem. Rev. 104 (2004) 4203.
is appreciated. The authors would also like to thank Dr. Schme-
dake, Dr. Cliff Carlin, Dr. Jon Merkert, Dr. Michael Murphy, and
Dr. John Pickett for instrumental assistance and discussion.
Appendix A. Supplementary data
[17] E. Alessio, G. Mestroni, G. Nardin, W.M. Attia, M. Calligaris, G. Sava, S. Zorzet,
Inorg. Chem. 27 (1988) 4099.
[18] R. Ziessel, V. Grosshenny, M. Hissler, C. Stroh, Inorg. Chem. 43 (2004)
4262.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.poly.2010.06.012.
[19] J.S. Jaswal, S.J. Rettig, B.R. James, Can. J. Chem.-Rev. Can. Chim. 68 (1990) 1808.
[20] M. Calligaris, Coord. Chem. Rev. 248 (2004) 351.
References
[21] J.A. Davies, The coordination chemistry of sulfoxides with transition metals, in:
H.J. Emeléus, A.G. Sharpe (Eds.), Advances in Inorganic Chemistry, Academic
Press, 1981, pp. 115–187.
[22] E. Ishow, A. Gourdon, J.P. Launay, P. Lecante, M. Verelst, C. Chiorboli, F.
Scandola, C.A. Bignozzi, Inorg. Chem. 37 (1998) 3603.
[23] S. Bodige, A.S. Torres, D.J. Maloney, D. Tate, G.R. Kinsel, A.K. Walker, F.M.
MacDonnell, J. Am. Chem. Soc. 119 (1997) 10364.
[24] T. Norrby, A. Borje, A. Akermark, L. Hammarstrom, J. Alsins, K. Lashgari, R.
Norrestam, J. Martensson, G. Stenhagen, Inorg. Chem. 36 (1997) 5850.
[25] E. Ishow, A. Gourdon, J.P. Launay, C. Chiorboli, F. Scandola, Inorg. Chem. 38
(1999) 1504.
[1] H. Chaturvedi, A.N. Giordano, M.J. Kim, F.M. MacDonnell, S.S. Subaran, J.C.
Poler, J. Phys. Chem. C 113 (2009) 11254.
[2] H. Chaturvedi, J.C. Poler, J. Phys. Chem. B 110 (2006) 22387.
[3] H. Chaturvedi, J.C. Poler, Appl. Phys. Lett. 90 (2007) 3.
[4] V. Balzani, G. Bergamini, P. Ceroni, Coord. Chem. Rev. 252 (2008) 2456.
[5] S.M. Arachchige, J.R. Brown, E. Chang, A. Jain, D.F. Zigler, K. Rangan, K.J. Brewer,
Inorg. Chem. 48 (2009) 1989.
[6] V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi, Acc. Chem.
Res. 31 (1998) 26.
[7] F.M. MacDonnell, M.J. Kim, S. Bodige, Coord. Chem. Rev. 185–186 (1999) 535.
[8] A.N. Giordano, H. Chaturvedi, J.C. Poler, J. Phys. Chem. C 111 (2007) 11583.