Linear multinuclear RuII photosensitizers

Article abstract of DOI:10.1002/ejic.200300321

Four ditopic bridging ligands, containing 2,2′:6′,2″- terpyridine and 2,6-dipyrazinylpyridine (dpp) metal-binding units attached through p- or m-phenylene linkages, have been incorporated into eight mono-, di- and trinuclear linear Ru^II complexes. These were characterized by UV/Vis spectroscopy and cyclic voltammetry, and by their ability to undergo light-induced electron transfers to methylviologen. The dpp-bearing complexes were more difficult to prepare but were superior sensitizers, a fact attributable to longer excited state lifetimes and an electrostatically favored reductive quenching pathway. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejic.200300321

FULL PAPER
Linear Multinuclear RuII Photosensitizers
Simona Vaduvescu[a] and Pierre G. Potvin*[a]
Keywords: Ruthenium / Tridentate ligands / Bridging ligands / Electron transfer / Photochemistry
Four ditopic bridging ligands, containing 2,2Ј:6Ј,2ЈЈ-terpyrid-
ine and 2,6-dipyrazinylpyridine (dpp) metal-binding units at-
tached through p- or m-phenylene linkages, have been in-
bearing complexes were more difficult to prepare but were
superior sensitizers, a fact attributable to longer excited state
lifetimes and an electrostatically favored reductive quench-
corporated into eight mono-, di- and trinuclear linear RuII ing pathway.
complexes. These were characterized by UV/Vis spectro-
scopy and cyclic voltammetry, and by their ability to undergo
light-induced electron transfers to methylviologen. The dpp-
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
above-mentioned report, the assessment used measure-
ments of the rates of formation of the methylviologen rad-
ical cation (MV·ϩ) from methylviologen (MV2ϩ) in CH3CN
in the presence of a sacrificial reductant (triethanolamine)
according to a set protocol.[4] Comparisons were also made
with [Ru(ttpy)2]2ϩ (ttpy ϭ 4Ј-p-tolyl-2,2Ј:6Ј,2ЈЈ-terpyridine),
a related complex of known τ (0.95 ns in CH3CN).[5] This
paper provides full details of that work and extends it by
exploring the photoactivity of several new complexes of re-
lated bridging ligands, namely the meta isomer of A (ligand
B) and the dipyrazinylpyridine (dpp) analogues of A and B
(ligands C and D). It was anticipated that the meta linkages
Ruthenium()Ϫpolypyridine complexes act as light ab-
sorption sensitizers,[1] a characteristic that can be put to use
in solar energy conversion and storage applications. A criti-
cal limitation on the photo-responsiveness is the concen-
tration of the sensitizer’s excited state that can be achieved,
and this notably depends on the excited state lifetime (τ)
and on the extinction coefficient (ε) for the metal-to-ligand
charge transfer (MLCT) absorption that leads to the excited
state. One approach to improving the photoactivity is to
increase the number of metal atoms in the sensitizer and
the probability of light absorption. Dendritic assemblies ep-
itomize this idea,[2] but electrostatic considerations make
them unattractive because their globular structures and
high charges would impede electron transfers to positively
charged acceptors, such as methylviologen (MV2ϩ), and/or,
in the case of neutral or anionic acceptors, would facilitate
the energy-wasting reverse electron transfer that regenerates
the initial states. The effect of charge repulsion can be mini-
mized with linear multinuclear arrays, where an end-on ap-
proach by an electron acceptor is minimally affected by re-
pulsions from the more distal metal centers. Hence, the elec-
trostatics can be relatively insensitive to the sensitizer chain
length while the light absorption would increase with
chain length.
We have recently communicated the preparation and
characterization of a linear trinuclear RuII complex in
which the ‘‘back-to-back’’ bis(terpy) ligand A bridges the
metal centers. This material was assessed as a photosensit-
izer relative to the mono- and dinuclear analogues.[3] In the
[a]
Department of Chemistry, York University,
4700 Keele Street, Toronto, ON M3J 1P3, Canada
Fax: (internat.) ϩ 1-416-736-5936
E-mail: pgpotvin@yorku.ca
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769
DOI: 10.1002/ejic.200300321
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1763

S. Vaduvescu, P. G. Potvin
FULL PAPER
of B and D would enable electrostatically easier access, by field migration of the doublets from 6-H, which juts out
MV2ϩ, to the areas between metal atoms in di- and trinu- above the plane of the perpendicular ligand and experiences
clear complexes and to the central metal atoms in trinuclear the effect of its ring current. Two sets of terpy signals were
cases, while the dpp ligands would be expected to promote obtained with [Ru(A)2](PF6)2, including singlets for 3Ј/5Ј-
longer τ values.[6]
H at δ ϭ 9.14 ppm and 8.98 ppm for the complexed and
free terpy units, respectively, and the phenylene singlet of A
was replaced by a pair of doublets in this less symmetrical
species. Though [(ttpy)Ru(A)](PF6)2 has been previously re-
ported,[15] it was characterized only by NMR spectroscopy.
Here, three sets of partially overlapping terpy signals were
evident, including three distinct 3Ј/5Ј-H singlets at δ ϭ 9.12,
Results and Discussion
Synthesis
Bis(terpyridine) A is known, and has been prepared by 9.02 and 8.97 ppm for bound A, ttpy and free A terpyridine
two routes.[7,8] We prepared it by the one-pot method devel- units, respectively, as well as phenylene doublets. With the
oped for the synthesis of dpp ligands,[6] and recently applied trinuclear complex [(ttpy)Ru(A)Ru(A)Ru(ttpy)](PF6)6, three
1
this to other terpy-based materials.[9] In this remarkable re- sets of terpy H NMR signals were expected but the ma-
action, a total of seven reactant molecules condense and jority overlapped considerably, reflecting the similarities in
expel six molecules of H2O, then undergo two cycles of aer- the three ligand environments, including those from the
ial dehydrogenation. The isolated yield (39%) was entirely phenylene spacer, which appeared together as one large sin-
similar to those of the previous preparations but with less glet. However, the three distinct central pyridine 3Ј/5Ј-H
effort and in less time. The linkage isomer B was produced singlets at lowfield positions (δ ϭ 9.23, 9.21 and 9.06 ppm)
under the same conditions from isophthalaldehyde and in confirmed the lack of symmetry and metal coordination.
a similar yield (37%).[10] Its X-ray crystal structure showed [(ttpy)Ru(A)Ru(ttpy)](PF6)4 showed only two sets of terpy
a lack of co-planarity consistent with the absence of a res- signals reflecting the higher symmetry, including a phenyl-
onance pathway of communication between terpy units. Its ene singlet and 3Ј/5Ј-H singlets at δ ϭ 9.19 and 9.05 ppm
NMR spectra showed a single set of pyridine signals, which for metal-bound terpy units of A and ttpy, respectively.
implies free rotation about the terpy-to-linker bonds. A use-
An entirely analogous sequence from the m-linked ligand
Ϫ
ful spectroscopic marker for the complexes of A and B was B led to PF6 salts of [(ttpy)Ru(B)]2ϩ, [(ttpy)Ru(B)Ru(B)-
1
the most downfield H NMR signal, i.e. the singlet from Ru(ttpy)]6ϩ and [(ttpy)Ru(B)Ru(ttpy)]4ϩ. The yields were
the central pyridine 3Ј/5Ј-Η (at δ ϭ 8.83 ppm in free A; δ ϭ slightly lower than in the p-linked cases (65%, 70% and
8.84 ppm in free B).
17%, respectively). In the latter preparation, we found that
By similar reactions with acetylpyrazine, the bis(dpp) li- [Ru(ttpy)2]2ϩ was a contaminant, presumably arising from a
gands C and D were obtained in isolated yields of 42% and ligand exchange process. This contaminant and the reduced
37%, respectively. Those materials are soluble only in tri- yields imply some resistance to coordinate to B, in compari-
fluoroacetic acid (TFA) and could not be properly recrys- son with A, perhaps due to steric encumbrance or increased
tallized. Their NMR spectra in [D]TFA were inconclusive, repulsions between metal centers. The dinuclear
but their identities were confirmed by EI-MS and, for C, [(ttpy)Ru(B)Ru(ttpy)]4ϩ also appeared as a side-product in
by conversions to RuII complexes (see below).
the preparation of [(ttpy)Ru(B)]2ϩ and necessitated chroma-
The monotopic ttpy was used as a terminal ligand. It tographic purification. The 1H NMR spectra of the B com-
could also be prepared by the one-pot method from acetyl- plexes were very similar to those of the A series, except for
pyridine and p-tolualdehyde in 46% yield after recrystalliza- the singlet-doublet-triplet pattern from the m-phenylene
tion, while a 70% yield was obtained by our earlier, two- linkage, and showed only one set of pyridine signals per
step method.[11]
[Ru(ttpy)Cl3][12] was activated with AgBF4,[13] then terpy units about the linkage.
distinct terpy unit, again consistent with free rotation of the
treated with 1 equiv. of A to form [(ttpy)Ru(A)](PF6)2 in
The reactions with the pyrazine-containing ligands
70% yield, after chromatography and anion exchange. In were less straightforward. Attempted preparation of
contrast, [Ru(A)2](PF6)2 was obtained in only 40% yield, by [(ttpy)Ru(C)]2ϩ from (ttpy)RuCl3 and C in ethane-1,2-diol
treatment of RuCl3 with 2 equiv. of A; 2 equiv. of [(ttpy)Ru- produced only the ligand-exchange product [Ru(ttpy)2]2ϩ
,
(A)](PF6)2 and 1 equiv. of RuCl3 were heated in 1,2-ethane- owing to the poor solubility of C. Using TFA as a co-sol-
diol to form the trinuclear complex [(ttpy)Ru(A)Ru(A)- vent was helpful, and the best solvent found was a TFA/
Ru(ttpy)](PF6)6 in 87% yield, after anion exchange and ethane-1,2-diol (5:1) mixture that led to a 70% yield of
recrystallization. The dinuclear analogue [(ttpy)Ru(A)- [(ttpy)Ru(C)](PF6)2 after anion exchange and chromatogra-
Ru(ttpy)](PF6)4 was needed for comparative purposes. It is phy. The 1H NMR spectra differ from those of the all-terpy
a known species[14] which we prepared from 2 equiv. of series in that the lowest field signals are now due to the
AgBF4-activated [(ttpy)RuCl3] and A in a yield of 19% (not pyrazinyl 3/3ЈЈ-H instead of the pyridine 3Ј/5Ј-H. The pyra-
optimized), after anion exchange, chromatography and a se- zinyl 5-H and 6-H are weakly coupled but COSY spec-
cond cycle of anion exchange.
troscopy revealed a long-range W-coupling between 3-H
1H and COSY NMR spectroscopy were used to confirm and 5-H. The dinuclear species [(ttpy)Ru(C)Ru(ttpy)](PF6)4
the structures. Complexation was indicated by a strong up- was obtained in 20% yield from Agϩ-activated [(ttpy)RuCl3]
1764
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769

Linear Multinuclear RuII Photosensitizers
FULL PAPER
and C in a mixture of TFA and ethane-1,2-diol. Chroma- toward the red with an increasing number of metal centers,
tography was required to separate the desired material from while λmax. remained constant with the m-linked ligand B.
the now abundant [Ru(ttpy)2]2ϩ impurity. The 1H NMR Relative to [Ru(ttpy)2]2ϩ, each added dpp or terpy unit and
spectrum was similar to that of the mononuclear analogue, each added metal center constitutes an electron-with-
simplified by symmetry, and COSY spectroscopy was used drawing group. The drift with a p-linkage is likely a result
to identify one pyrazinyl 3-H singlet, two pyridine 3Ј/5Ј-H of electron withdrawal on the ligands’ π systems and stabil-
singlets and one phenylene singlet, all consistent with the ization of the ligand π* orbitals, relative to the metal-cen-
structure. Unfortunately, all attempts at preparing the trinu- tered orbitals. The lack of such a drift with the m-linkage
clear analogue failed, producing only a brown-black, para- is probably owing to the absence of a resonance path for
magnetic material. Similarly, all complexation reactions electronic transmission, and the 2 nm shift in λmax. with
with ligand D failed as well. Treatment of [(ttpy)RuCl3] with [(ttpy)Ru(B)]2ϩ then represents the inductive contribution
1 equiv. of D also produced a paramagnetic product. With of an added terpy unit. At the same time, the ε values grew
0.5 equiv., [Ru(ttpy)2]2ϩ was the major product and the de- upon increasing the number of metal atoms, as expected. It
sired dinuclear species was only a minor constituent detect- is interesting to note that the ε values of all the mononu-
able by NMR spectroscopy and ESI-MS. It became obvious clear complexes [(ttpy)RuL]2ϩ (L ϭ AϪC) are larger than
that the m-phenylene linkage and the lower basicity of dpp for the homoleptic case, where L ϭ ttpy. This is even more
units, relative to terpy, decreased the reactivity of the li- pronounced with the homoleptic [Ru(A)2]2ϩ (λmax.
ϭ
gand. An alternate route was attempted to circumvent the 496 nm, ε ϭ 42700 Ϫ1·cmϪ1). Because the MLCT λmax.
latter problem. Treatment of RuCl3 with 0.5 equiv. of D in values are all of lower energies than with [Ru(ttpy)2]2ϩ, it is
EtOH provided
a brown-black precipitate, putatively reasonable to conclude that the bridging ligands have lower
[Cl3Ru(D)RuCl3], but treatment of this with ttpy in ethane- π* orbital energies than does the terminal ligand ttpy, and
1,2-diol to produce the dinuclear compound led to are therefore the repositories of the excited electron density.
[Ru(ttpy)2]2ϩ as the only detectable product.
Electrochemistry
As is common with aromatic materials and Ru com-
plexes, elemental analyses often return correct H compo-
sitions but low C and/or N compositions, even in the pres-
ence of extra oxidation catalyst, presumably due to incom-
plete combustion and/or the formation of stable carbides
and nitrides. Mass and isotopic distribution analyses (ESI,
MALDI or high resolution MALDI) were used to confirm
the structural assignments. Fortunately, the precipitation of
PF6Ϫ salts enabled purification, which was made evident by
chromatographic homogeneity and the absence of ex-
traneous NMR peaks.
The results of cyclic voltammetry are also reported in
Table 1. As is typical for complexes of this nature,[14,17] the
new complexes show single reversible oxidation waves at
potentials that are practically insensitive to chain length
and at least two reduction waves that are somewhat less
constant. Owing to their additional nitrogen atoms, the two
complexes of C show particularly positive E(RuIII/II) poten-
tials, lying between the value for the homoleptic dpp com-
plex [Ru(E)2]2ϩ (ϩ1.62 V)[6] and that for the bis(terpy) com-
plex [Ru(ttpy)2]2ϩ (ϩ1.25 V).[18] The first reduction poten-
tials of the new complexes are all positive compared with
that of [Ru(ttpy)2]2ϩ. Along with the spectroscopic evi-
UV/Vis Spectroscopy
The electronic spectra of all complexes in CH3CN dence, this suggests that reduction adds an electron to the
showed intense π-π* bands below 350 nm and MLCT bridging ligands, rather than to the terminal ttpy whose π*
bands (λmax.) at 490Ϫ500 nm, reported in Table 1, in com- orbital is evidently at a higher energy.
plete analogy with [Ru(ttpy)2]2ϩ [16]
.
In addition, the com-
In accordance with the more negative reductions of the
plexes of C showed shoulders near 450 nm. With the p- terpy-based ligands, we can assign the first and second re-
linked ligands A and C, the position of λmax. drifted slightly duction waves shown by the complexes of C to reductions
Table 1. Key spectroscopic, electrochemical and photochemical data
Compound[a]
λmax.
[nm]
ε
Eox
Ered
kf
kq
χ
∆G‡
[eV]
∆G‡
[eV]
FET
RET
[104 Ϫ1·cmϪ1] [V vs. SCE] [V vs. SCE]
[10Ϫ5
s
Ϫ1] [10Ϫ3
s
Ϫ1] [m]
[(ttpy)2Ru]2ϩ
490[b]
496
502
1.55[b]
3.27
5.98
7.33
2.61
3.70
6.96
1.25[c]
1.27
1.27[g]
1.27
1.28
1.28
1.28
1.48
1.49
Ϫ1.24,[c] Ϫ1.42[d]
Ϫ1.18, Ϫ1.42
Ϫ1.18, Ϫ1.45[g]
Ϫ1.03, Ϫ1.23
Ϫ1.22, Ϫ1.44
Ϫ1.18, Ϫ1.45
Ϫ1.18, Ϫ1.37
1.12(1)[e] 8.8(6)[e]
1.93(18) 7.4(9)
1.086(2) 2.40(16) 45(3) 0.29
11(4)[e] 0.24
26(4) 0.25
0.22
0.23
0.25
0.26
0.23
0.25
0.27
0.32
0.34
[(ttpy)Ru(A)]2ϩ [f]
[{(ttpy)Ru}2A]4ϩ [f]
[{(ttpy)Ru(A)}2Ru]6ϩ [f] 504
1.339(5) 3.67(1)
36.3(2) 0.32
40(5) 0.25
47(3) 0.30
[(ttpy)Ru(B)]2ϩ
492
492
492
1.01(5)
1.98(4)
0.95(3)
2.5(3)
4.2(3)
[{(ttpy)Ru}2B]4ϩ
[{(ttpy)Ru(B)}2Ru]6ϩ
[(ttpy)Ru(C)]2ϩ
3.79(11) 25(1) 0.33
7.0(3)
456, 504 1.48, 2.01
454, 506 2.54, 3.33
Ϫ0.86, Ϫ1.25, Ϫ1.65 3.48(2)
Ϫ0.85, Ϫ1.25, Ϫ1.65 2.31(13) 4.7(13)
50(2) 0.38
49(13) 0.42
[{(ttpy)Ru}2C]4ϩ
[a]
[b]
[d]
Uncertainties in least significant digits appear in brackets.
From Mikel and Potvin.[16] [c] From Kalyanasundaram et al.[18]
In
DMSO, from Constable et al.[31] [e] From Potvin et al.[4] [f] From Vaduvescu and Potvin.[3] [g] From Collin et al.[14]
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1765

S. Vaduvescu, P. G. Potvin
FULL PAPER
of the relatively electron-poor bridging and electron-rich results should confirm or deny. The constant kf incorpor-
terminal ligands, respectively. Since E(RuIII/II) Ϫ E(L0/Ϫ) is ates the concentration of excited state species (RuII*) and
known to correlate with the MLCT energy,[19] the higher- will depend on ∆G‡FET, which contains the electrostatic bar-
energy MLCT bands of the C complexes can then be as- rier to the forward electron transfer which increases with
signed to RuǞttpy transitions, and the lower-energy bands increasing N, while kq incorporates the concentration of the
to Ru Ǟ C transitions. With these pairings, the correlation RuIII photoproduct and depends on ∆G‡RET, which contains
is fairly good (r2 Ͼ 0.92).
the corresponding electrostatic barrier. The constants kq
and kf, however, are linked because a higher kf leads to a
higher concentration of RuIII and therefore a higher kq.
Photochemistry
Under visible light irradiation and in the presence of tri- Both, then, also depend on the photophysical properties
ethanolamine (TEOA) as a sacrificial reductant, RuII com- that govern the production of RuII*, including ε and τ.
plexes are able to generate the methylviologen radical cation However, the ratio kq/kf (ϭ [MV2ϩ]0/χ Ϫ 1) was previously
(MV·ϩ) from methylviologen (MV2ϩ) to different degrees. shown to equal kr[MV2ϩ]0/kt[TEOA]0,[4] where kr and kt are
In real-life photoredox applications, there would be no sac- the second-order rate constants for the reactions of the
rificial donor present, but the excited state could be re- RuIII photoproduct with the MV·ϩ co-product (reverse elec-
ductively quenched at an electrode surface or by a reversibly tron transfer) and with TEOA (trapping), respectively. This
reduced electron relay molecule such as MV2ϩ at an illumi- ratio is therefore a measure of the competition between the
nated organic-aqueous interface. Comparisons between non-productive and productive fates of the RuIII form of
sensitizers in terms of the kinetic parameters governing this the sensitizer. The first constant, kr, will vary within a series
process have previously enabled us to probe steric, electro- according to exp(Ϫ∆G‡RET), while kt involves encounters
static and solvent effects on activity,[16,20,21] and we wished, with the neutral TEOA and should remain constant within
in the present work, to probe the effect of increasing the Ru a series. Hence, if the sensitizers within a series only differed
chain length N. Using our standard protocol,[4] we meas- in electrostatic terms, kq/kf should steadily decrease as
ured the time courses of MV·ϩ development at room tem- ∆G‡RET steadily increases with N. The values in Table 1
perature in CH3CN at 600 nm with a sensitizer concen- reveal that this is not so, with [{(ttpy)Ru}2A]4ϩ and
tration of 4 ϫ 10Ϫ5 . The key results presented in Table 1 [{(ttpy)Ru(B)}2Ru]6ϩ being particularly outstanding. More-
are the first-order rate constants kf and kq for MV·ϩ forma- over, kf should vary with ε exp(Ϫ∆G‡FET), but ln(kf /ε) does
tion and anaerobic quenching, respectively, and the anaer- not uniformly decrease as ∆G‡FET increases. These findings
obic yields χ, given by kf[MV2ϩ]0/(kf ϩ kq). The usual imply that the sensitizers within a series also differ in their
mechanism is oxidative quenching of the excited state of the photophysical characteristics, such as τ, orbital overlap, ap-
sensitizer (RuII* ϩ MV2ϩ Ǟ RuIII ϩ MV·ϩ) followed by proach geometry or approach distance, which this study
reverse electron transfer to (quenching of) MV·ϩ (RuIII
ϩ
ϩ
cannot probe. Nevertheless, Table 1 identifies the most ef-
MV·ϩ Ǟ RuII ϩ MV2ϩ), or competitive trapping (RuIII
fective sensitizers [{(ttpy)Ru(A)}2Ru]6ϩ, [{(ttpy)Ru}2B]4ϩ
,
TEOA Ǟ RuII ϩ TEOA·ϩ) which allows MV·ϩ to accumu- and especially the two complexes of C, as meriting further
late. Additionally, Table 1 reports estimates of ∆G‡FET and study.
∆G‡RET, the Marcus activation energies for the forward and
The two C-containing complexes show higher activities
reverse electron transfers, respectively, for end-on ap- in spite of having the highest barriers ∆G‡FET. They prob-
proaches by MV2ϩ/·ϩ to the terminal metals. Their compu- ably benefit from longer τ values than the all-terpy com-
tations and the values for electrostatically less favorable ap- plexes of A and B because the first dpp complex,
proaches are detailed in the Supporting Information (see [Ru(E)2]2ϩ, showed a longer τ (18 ns at 667 nm)[6] than its
also the footnote on the first page of this article).
All of the new complexes are better sensitizers than dition, the redox potentials for the C-containing complexes
[Ru(ttpy)2]2ϩ [4]
with higher kf and χ values. This may re- lie about 0.2 V more positive of those of the other com-
terpy analogue [Ru(ttpy)2]2ϩ (0.95 ns at 640 nm).[22] In ad-
,
flect generally longer τ values owing to a greater π-delocal- plexes and this is probably sufficient to enable reductive
ization and π* stabilization in the bridging ligands which, quenching (formally RuII* ϩ TEOA Ǟ RuI ϩ TEOA·ϩ), as
as discussed earlier, are the preferred sites for the excited was the case in other complexes with high Eox values.[23Ϫ25]
electrons. Clearly, there is variation between the A and B Reductive quenching has an electrostatic advantage over
series, in spite of similarities in structures, properties (Eox, oxidative quenching because it involves initial encounters
Ered, λ, ε/N) and even ∆G‡ values. Part of the success of with neutral TEOA instead of cationic MV2ϩ, and because
[{(ttpy)Ru}2B]4ϩ relative to the A analogue may be due to the wasteful reverse electron transfer (formally RuI
ϩ
easier access to the areas between the metal atoms, but this TEOA·ϩ Ǟ RuII ϩ TEOA) is electrostatically disfavored
advantage does not appear to carry forward to the trinu- relative to the forward reaction, whereas the opposite is true
clear case.
in oxidative quenching (RuIII ϩ MV·ϩ Ǟ RuII ϩ MV2ϩ).
Because of similar properties within a ligand series and Moreover, the TEOA·ϩ by-product can be deprotonated in
virtually identical structures at the metal termini, we can an alkaline medium (TEOA·ϩ Ǟ Hϩ ϩ HϪ1TEOA·) to gen-
make the a priori assumption that the differences in sensit- erate a reducing species[26] that is depleted by O2.[4] This
izer activities within a series should be largely electrostatic consumption of TEOA·ϩ further reduces the probability of
and not photophysical in origin, an assumption which our reverse electron transfer, thereby leaving more of the re-
1766
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769

Linear Multinuclear RuII Photosensitizers
FULL PAPER
duced form of the sensitizer able to produce MV·ϩ (RuI ϩ
MV2ϩ Ǟ RuII ϩ MV·ϩ) in electrostatically more favorable
encounters than is the case in oxidative quenching. These
combined electrostatic advantages would contribute to a
faster net accumulation of the ultimate photoproduct,
MV·ϩ, as witnessed with ligand C.
agents and spectroscopic-grade solvents (SigmaϪAldrich) were
used directly. Anhydrous solvents were purchased in
SigmaϪAldrich Sure-seal bottles. All other solvents were reagent-
grade quality (Caledon Chemical Ltd.) and used directly.
[Ru(ttpy)Cl3] was prepared in 71% yield by a literature method[12]
,
with its identity verified by ESI-MS [m/z (%) ϭ 531 (100)]. Thin-
layer chromatography was used to monitor reactions and chroma-
tography and was carried out on MachereyϪNagel Polygram Sil
G/UV silica gel plates and Alox N/UV alumina plates. Column
chromatography was performed employing silica gel (ca. 200Ϫ400
Conclusion
˚
˚
mesh, 60 A) or neutral alumina (ca. 150 mesh, 58 A), both from
SigmaϪAldrich. Elemental analyses were performed by Guelph
Chemical Laboratories Ltd., Guelph, ON, Canada.
Ditopic bridging ligands can be assembled in a one-pot
reaction and incorporated into short, linear multinuclear
assemblies that act as photosensitizers in spite of electro-
static repulsions. The photochemical experiments have
identified [{(ttpy)Ru(A)}2Ru]6ϩ, [{(ttpy)Ru}2B]4ϩ and both
complexes of C as candidates for photophysical studies, in
order to elucidate the origins of the difference in behavior
between the A series and the B series and the origin of the
4Ј-p-Tolyl-2,2Ј:6Ј,2"-terpyridine (ttpy): 2-Acetylpyridine (1.00 g,
8.3 mmol) was added to a solution of p-tolualdehyde (0.50 g,
4.2 mmol) in CH3OH (32 mL). 15% aq. KOH (2.9 mL) and conc.
NH4OH (29 mL) were then added to the solution and vigorous
stirring was maintained for 3 d. The yellow precipitate that formed
was removed by vacuum filtration, washed with water to neutral
better activity for the C series. However, the dpp-containing pH and dissolved in CHCl3 (40 mL). The solution obtained was
washed with H2O (3 ϫ 30 mL). The organic fraction was dried
with MgSO4 and freed of solvent in vacuo. The residue was recrys-
tallized from CHCl3/Et2O to yield 0.61 g (46%) of ttpy, m.p. 168
ligands C and D were fraught with solubility and reactivity
problems that need to be addressed. An obvious improve-
ment to the general design is to add negatively charged
groups to the peripheries which can constitute binding sites
for positively charged electron acceptors such as MV2ϩ
since this has proven to be of great benefit in mononuclear
species.[21] However, the ideal linear multinuclear sensitizer
will absorb light anywhere along the chain and relay the
excited electron to the ends (the so-called antenna effect),
where electrostatic repulsions with acceptors can be mini-
mized. Another improvement would therefore be to use the
most electron-poor ligands at the termini.
3
°C. 1H NMR (CDCl3): δ ϭ 8.77 (2 s, 4 H, a ϩ e), 8.71 (d, JH,H ϭ
3
8 Hz, 2 H, d), 7.92 (dd, 2 H, c), 7.87 (d, JH,H ϭ 7.9 Hz, 2 H, f),
3
3
7.39 (dd, JH,H ϭ 5 Hz, 2 H, b), 7.35 (d, JH,H ϭ 7.9 Hz, 2 H, g),
2.46 (s, 3 H, CH3) ppm. EI-MS: m/z (%) ϭ 324 (100) [M·ϩ], in
agreement with the literature data.[11,29,30]
1,4-Bis(2,2Ј:6Ј,2"-terpyridin-4Ј-yl)benzene (A): The same procedure
as for ttpy was applied, using 2-acetylpyridine (1.00 g, 8.3 mmol)
and terephthalaldehyde (0.2764 g, 2.1 mmol) at reflux for 48 h.
Recrystallization from CHCl3 and CH3OH yielded 0.4427 g of A
(39%), m.p. Ͼ 300 °C. 1H NMR (CDCl3): δ ϭ 8.83 (s, 4 H, e), 8.79
3
3
(d, JH,H ϭ 4.2 Hz, 4 H, a), 8.74 (d, JH,H ϭ 7.0 Hz, 4 H, d), 8.09
3
3
(s, 4 H, f), 7.92 (dd, JH,H ϭ 7.0 Hz, 4 H, c), 7.39 (dd, JH,H
ϭ
4.2 Hz, 4 H, b) ppm. MALDI-MS: m/z (%) ϭ 541 (100) [Mϩ].
C36H24N6 (540.6): calcd. C 79.98, H 4.47, N 15.54; found C 79.57,
H 4.73, N 12.31.
Experimental Section
General Remarks: NMR spectroscopic data were obtained with a
Bruker AMX 400 MHz spectrometer in CD3CN (unless indicated
otherwise) at 25 °C and are reported relative to SiMe4. The signal
assignments were made on the basis of COSY spectra and use the
position letters appearing in the structural diagrams. A Kratos Pro-
file mass spectrometer was used for EI-MS acquisitions. MALDI-
MS data were acquired with a Voyager-DE STR MALDI/TOF in-
strument and ESI-MS data with a Sciex Targa 6000E MS/MS sys-
tem. High resolution MALDI-MS (HR-MALDI-MS) was per-
formed by the McMaster Regional Center for Mass Spectrometry,
Hamilton, ON, Canada with a Micromass TofSpec 2E instrument.
1,4-Bis(2,6-dipyrazinylpyridin-4-yl)benzene (C): A mixture of acetyl-
pyrazine (500 mg, 4.1 mmol) and terephthalaldehyde (135 mg,
1.0 mmol) in CH3OH (50 mL) containing 15% aq. KOH (1.2 mL)
and conc. NH4OH (11.5 mL) was vigorously stirred while heating
to reflux for 48 h. The precipitate formed was removed by vacuum
filtration and washed with water to neutral pH. The compound
proved to be soluble only in TFA and insoluble in all common
organic solvents. It was dried under vacuum and a yield of 42%
(0.2285 g) was obtained, m.p. Ͼ 300 °C. EI-MS: m/z (%) ϭ 545
(100) [Mϩ]. C32H20N10 (544.6): calcd. C 70.58, H 3.70, N 25.72;
found C 69.90, H 3.80, N 25.10.
UV/Vis spectra were obtained at room temperature with
a
HewlettϪPackard 8452A diode array spectrophotometer, with 10Ϫ5
 solutions in spectroscopic-grade solvents. Cyclic voltammetry
was performed with a Cypress System three-electrode configuration
consisting of a reference Ag/AgCl wire electrode and platinum wire
working and auxiliary electrodes, using CH3CN distilled from P2O5
1,3-Bis(2,6-dipyrazinylpyridin-4-yl)benzene (D): The same pro-
cedure as for C with isophthalaldehyde produced 0.2013 g (37%
yield) of D, which was similarly insoluble, m.p. Ͼ 300°C. EI-MS:
m/z (%) ϭ 545 (100) [Mϩ]. C32H20N10 (544.6): calcd. C 70.58, H
3.70, N 25.72; found C 69.93, H 3.81, N 25.13.
˚
and stored over 4-A molecular sieves, in solutions containing 0.1 
nBu4NPF6. Half-wave potentials were calculated as the average of
the cathodic and anodic peak potentials.[27] The current-scan rate [Ru(A)2](PF6)2: A solution of A (80.0 mg, 0.15 mmol) in DMF
dependence was evaluated for each sample to check the reversi-
(10 mL) was added to a solution of RuCl3·3H2O (19.4 g,
bility, and ferrocene was added at the end of each run as an internal 0.07 mmol) in DMF (10 mL) and the mixture was heated to reflux
reference (ϩ0.425 V vs. SCE in CH3CN).[28] Photochemical data
under Ar for 24 h. Addition of the cooled red solution to a satu-
rated aqueous solution of NH4PF6 precipitated the product as a
were obtained using the apparatus and data treatment described
earlier,[4] with stirred CH3CN solutions containing the sensitizer (4 red solid. Purification by column chromatography (alumina;
ϫ 10Ϫ5 ), MV(PF6)2 (0.01 ) and triethanolamine (0.05 ). Re-
CH3CN/satd. KNO3/H2O, 7:1:0.5, containing 1% N(C2H5)3] pro-
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1767

S. Vaduvescu, P. G. Potvin
FULL PAPER
1
1
duced 41.2 mg (40%). H NMR: δ ϭ 9.14 (s, 4 H, e), 8.98 (s, 4 H, a yield of 19% (18.7 mg) of [(ttpy)RuARu(ttpy)](PF6)4. H NMR:
3
3
eЈ), 8.82 (m, 8 H, aЈ, dЈ), 8.74 (d, JH,H ϭ 8 Hz, 4 H, d), 8.45 (d,
δ ϭ 9.19 (s, 4 H, A-e), 9.05 (s, 4 H, ttpy-e), 8.78 (d, JH,H ϭ 8 Hz,
3JH,H ϭ 7.8 Hz, 4 H, f), 8.36 (d, JH,H ϭ 7.8 Hz, 4 H, fЈ), 8.10 (t, 4 H, A-d), 8.71 (d, JH,H ϭ 8 Hz, 4 H, ttpy-d), 8.62 (s, 4 H, A-f),
3
3
3JH,H ϭ 7 Hz, 4 H, cЈ), 8.01 (t, JH,H ϭ 7 Hz, 4 H, c), 7.57 (m, 8 8.17 (d, JH,H ϭ 7.6 Hz, 4 H, ttpy-f), 8.03 (m, 8 H, A-c, ttpy-c),
3
3
3
3
H, bЈ, a), 7.24 (t, JH,H ϭ 7 Hz, 4 H, b) ppm. MALDI-MS: m/z 7.64 (d, JH,H ϭ 7.5 Hz, 4 H, ttpy-g), 7.51 (m, 8 H, A-a, ttpy-a),
(%) ϭ 593 (100) [M2ϩ]. C72H48F12N12P2Ru (1472.3): calcd. C
58.74, H 3.29, N 11.42; found C 59.15, H 3.49, N 10.89.
7.26 (m, 8 H, A-b, ttpy-b), 2.58 (s, 6 H, CH3) ppm. ESI-MS: m/z
(%) ϭ 348 (75) [M4ϩ].
[(ttpy)Ru(A)](PF6)2: A mixture of [Ru(ttpy)Cl3] (55.0 mg, 0.1 mmol)
and AgBF4 (62.5 mg, 0.3 mmol) was heated to reflux in acetone
[(ttpy)Ru(B)Ru(ttpy)](PF6)4: This complex was prepared by exactly
the same procedure as for [(ttpy)Ru(A)Ru(ttpy)](PF6)4, using
(25 mL) for 3 h. The AgCl precipitate was filtered off and the vol- [Ru(ttpy)Cl3] (52.6 mg, 0.1 mmol), AgBF4 (79.3 mg, 0.3 mmol) and
ume of the resultant violet solution was halved under reduced
pressure. DMF (15 mL) was added and the remaining acetone re-
moved. The resultant solution was added to a pre-heated (80 °C)
solution of A (84.0 mg, 0.15 mmol) in DMF (25 mL) and the mix-
ture was heated to reflux under Ar for 2 h. The cooled red solution
was added to a saturated aqueous solution of NH4PF6. The red
B[10] (28.0 mg, 0.05 mmol), affording a yield of 17% (16.7 mg) of
1
[(ttpy)Ru(B)Ru(ttpy)](PF6)4. H NMR: δ ϭ 9.27 (s, 4 H, B-e), 9.02
(s, 4 H, ttpy-e), 8.80 (d, 3JH,H ϭ 7.3 Hz, 4 H, B-d), 8.68 (d, 3JH,H ϭ
3
8.3 Hz, 4 H, ttpy-d), 8.47 (m, 4 H, B-f, B-g, B-h), 8.13 (d, JH,H
ϭ
3
7.6 Hz, 4 H, ttpy-f), 7.99 (m, 8 H, B-c, ttpy-c), 7.60 (d, JH,H
ϭ
7.6 Hz, 4 H, ttpy-g), 7.48 (m, 8 H, B-a, ttpy-a), 7.23 (m, 8 H, B-b,
precipitate was collected by vacuum filtration, washed with water ttpy-d), 2.55 (s, 6 H, CH3) ppm. ESI-MS: m/z (%) ϭ 347 (100)
(3 ϫ 30 mL) and Et2O (3 ϫ 30 mL), and freed of solvent under
vacuum. Column chromatography (alumina; CH3CN/satd. KNO3/
H2O, 31:2:1, containing 0.66% N(C2H5)3] and re-precipitation
(NH4PF6) afforded 87.8 mg (70% yield) of [(ttpy)Ru(A)](PF6)2. The
1H NMR spectroscopic data are in agreement with literature
data.[15] ESI-MS: m/z (%) ϭ 483 (100) [M2ϩ]. C58H41F12N9P2Ru
(1255.0): calcd. C 55.51, H 3.29, N 10.04; found C 55.86, H 3.50,
N 9.77.
[M4ϩ]. C80H58F24N12P4Ru2 (1969.4): calcd. C 48.79, H 2.97, N
8.53; found C 48.24, H 3.45, N 8.45.
[(ttpy)Ru(C)Ru(ttpy)](PF6)4: This was prepared by the same pro-
cedure as for [(ttpy)Ru(C)](PF6)2, using [Ru(ttpy)Cl3] (96.0 mg,
0.2 mmol) and AgBF4 (130.0 mg, 0.6 mmol) in acetone (25 mL),
ethane-1,2-diol (10 mL), and a solution of C (50.0 mg, 0.1 mmol)
in TFA (5 mL) and ethane-1,2-diol (10 mL), with heating to reflux
under Ar for 24 h in the presence of N-methylmorpholine (2 drops).
Workup, chromatography (CH3CN/satd. KNO3/H2O, 7:1:0.5), to
remove the minor impurity of [Ru(ttpy)2](PF6)2, and re-precipi-
[(ttpy)Ru(B)](PF6)2: This complex was prepared by the same pro-
cedure as with A, using [Ru(ttpy)Cl3] (55.0 mg, 0.1 mmol), AgBF4
(62.5 mg, 0.3 mmol), DMF (10 mL), ligand B[10] (84.0 mg, tation (NH4PF6) as before afforded a yield of 20% (39.5 mg) of
0.15 mmol) in DMF (15 mL), previously heated to 80 °C, and heat-
ing at reflux under Ar for 2 h. Workup, chromatography and re-
[(ttpy)Ru(C)Ru(ttpy)](PF6)4. 1H NMR: δ ϭ 9.84 (s, 4 H, C-d), 9.34
(s, 4 H, C-e), 9.04 (s, 4 H, ttpy-e), 8.70 (d, 3JH,H ϭ 8 Hz, 4 H, ttpy-
3
precipitation (NH4PF6) as before afforded a yield of 65% (0.0816 g)
d), 8.64 (s, 4 H, C-f), 8.41 (d, 4 H, C-a), 8.16 (d, JH,H ϭ 7.9 Hz,
1
3
of [(ttpy)Ru(B)](PF6)2. H NMR: δ ϭ 9.13 (s, 2 H, B-e), 9.00 (s, 4 4 H, ttpy-f), 8.03 (t, JH,H ϭ 7.5 Hz, 4 H, ttpy-c), 7.64 (m, 8 H, C-
3
H, B-eЈ, ttpy-e), 8.78 (m, 4 H, B-aЈ, B-dЈ), 8.70 (d, JH,H ϭ 4.5 Hz, b, ttpy-g), 7.42 (d, 3JH,H ϭ 5 Hz, 4 H, ttpy-a), 7.23 (t, 3JH,H ϭ 6 Hz,
3
2 H, B-d), 8.65 (d, JH,H ϭ 8 Hz, 2 H, ttpy-d), 8.33 (m, 2 H, B-f/
4 H, ttpy-b), 2.57 (s, 6 H, CH3) ppm. HR-MALDI-MS: m/z (%) ϭ
(60), requires 1393.509.
[Ru2C76H54N16]4ϩ
C76H54F24N16P4Ru2 (1973.4): calcd. C 46.26, H 2.76, N 11.36;
fЈ), 8.21 (m, 2 H, B-g, B-h), 8.12 (m, 4 H, ttpy-f, B-cЈ), 7.97 (m, 4 1393.297
H, B-c, ttpy-c), 7.60 (m, 4 H, ttpy-g, B-bЈ), 7.45 (m, 4 H, B-a, ttpy-
a), 7.21 (2 t, 4 H, B-b, ttpy-b), 2.54 (s, 3 H, CH3) ppm. ESI-MS: found C 46.75, H 3.07, N 11.55.
m/z (%) ϭ 482 (75) [M2ϩ]. C58H41F12N9P2Ru (1255.0): calcd. C
55.51, H 3.29, N 10.04; found C 55.86, H 3.49, N 9.68.
[(ttpy)Ru(A)Ru(A)Ru(ttpy)](PF6)6: To a solution of RuCl3·3H2O
(9.0 mg, 0.03 mmol) in ethane-1,2-diol (30 mL) was added [(ttpy)-
Ru(A)](PF6)2 (90.0 mg, 0.07 mmol) and N-methylmorpholine (9
drops). Ths solution was heated to reflux under Ar for 24 h. The
addition of the cooled red solution to a saturated aqueous solution
of NH4PF6 precipitated the product as a red solid. Recrystalliza-
tion using CH3CN/Et2O of the red solid yielded the desired pure
[(ttpy)Ru(C)](PF6)2: This complex was prepared by the same pro-
cedure as for A, using [Ru(ttpy)Cl3] (112.0 mg, 0.2 mmol) and
AgBF4 (125.0 mg, 0.6 mmol) in acetone (40 mL), complete removal
of the acetone, 1,2-ethanediol (15 mL) instead of DMF,
C
(170.0 mg, 0.3 mmol) in TFA (25 mL) and heating to reflux under
Ar for 3 h. Workup, chromatography (CH3CN/satd. KNO3/H2O,
1
product in 87% yield (75.8 mg). H NMR: δ ϭ 9.23 (s, 4 H, A-e),
50:2:1) and re-precipitation (NH4PF6) as before afforded a yield of
70% (0.1763 g) of [(ttpy)Ru(C)](PF6)2. H NMR: δ ϭ 9.99 (s, 2 H, 8.71 (d, JH,H ϭ 7.9 Hz, 4 H, ttpy-d), 8.64 (s, 8 H, A-f, A-fЈ), 8.17
9.21 (s, 4 H, A-eЈ), 9.06 (s, 4 H, ttpy-e), 8.79 (2 t, 8 H, A-d, A-dЈ),
1
3
3
C-dЈ), 9.91 (s, 2 H, C-d), 9.31 (s, 2 H, C-e), 9.03 (s, 2 H, C-eЈ), 8.97
(d, JH,H ϭ 7.4 Hz, 4 H, ttpy-f), 8.03 (m, 12 H, A-c, A-cЈ, ttpy-c),
3
(s, 2 H, ttpy-e), 8.79 (m, 4 H, C-bЈ, C-aЈ), 8.69 (d, JH,H ϭ 7.7 Hz, 7.64 (d, 3JH,H ϭ 7.4 Hz, 4 H, ttpy-g), 7.58 (d, 3JH,H ϭ 5.3 Hz, 4 H,
3
2 H, ttpy-d), 8.49 (d, JH,H ϭ 8.5 Hz, 2 H, C-f), 8.39 (m, 4 H, C-
A-a), 7.52 (m, 8 H, A-aЈ, ttpy-a), 7.28 (m, 12 H, A-b, A-bЈ, ttpy-b),
3
a, C-fЈ), 8.16 (d, JH,H ϭ 7 Hz, 2 H, ttpy-f), 8.02 (t, 2 H, ttpy-c), 2.58 (s, 6 H, CH3) ppm. ESI-MS: m/z (%) ϭ 339.2 (70) [M6ϩ].
7.62 (m, 4 H, ttpy-g, C-b), 7.41 (d, 3JH,H ϭ 5 Hz, 2 H, ttpy-a), 7.23
C116H82F36N18P6Ru3 (2901.0): calcd. C 48.03, H 2.85, N 8.69;
(t, 2 H, ttpy-b), 2.58 (s, 3 H, CH3) ppm. HR-MALDI-MS: m/z found C 47.52, H 2.89, N 7.89.
(%)
969.237 (100); [RuC54H37N13]2ϩ requires 969.044.
ϭ
[(ttpy)Ru(B)Ru(B)Ru(ttpy)](PF6)6: The same procedure used for
[(ttpy)Ru(A)Ru(A)Ru(ttpy)](PF6)6 was applied using [(ttpy)-
Ru(B)](PF6)2 (100 mg, 0.07 mmol) and RuCl3·3H2O (11 mg,
0.03 mmol), to produce the desired pure product in 70% yield
C54H37F12N13P2Ru (1259.0): calcd. C 51.52, H 2.96, N 14.46;
found C 52.20, H 3.56, N 13.94.
[(ttpy)Ru(A)Ru(ttpy)](PF6)4: This complex was prepared by the
same procedure as for [(ttpy)Ru(A)](PF6)2, using [Ru(ttpy)Cl3]
(53.0 mg, 0.1 mmol), AgBF4 (77.5 mg, 0.3 mmol) in acetone
1
(61 mg). H NMR: δ ϭ 9.29 (s, 4 H, B-e), 9.27 (s, 4 H, B-eЈ), 9.05
3
(s, 4 H, ttpy-e), 8.81 (2 t, JH,H ϭ 9.3 Hz, 8 H, B-d, B-dЈ), 8.71 (d,
(25 mL), DMF (10 mL), A (27.0 mg, 0.05 mmol) in DMF (10 mL), 3JH,H ϭ 8 Hz, 4 H, ttpy-d), 8.54 (m, 4 H, B-f, B-fЈ), 8.16 (d, 3JH,H ϭ
and heating at reflux under Ar for 3 h. Workup as before afforded
8 Hz, 4 H, ttpy-f), 8.02 (m, 16 H, B-c, B-cЈ, ttpy-c, B-g, B-h), 7.63
1768
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769

Linear Multinuclear RuII Photosensitizers
FULL PAPER
[14]
[15]
3
3
J.-P. Collin, P. Laine, J.-P. Launay, J.-P. Sauvage, A. Sour, J.
Chem. Soc., Chem. Commun. 1993, 434Ϫ435.
F. Barigelletti, L. Flamigni, V. Balzani, J.-P. Collin, J.-P. Sauv-
age, A. Sour, E. C. Constable, A. M. W. Cargill Thompson, J.
Am. Chem. Soc. 1994, 116, 7692Ϫ7699.
C. Mikel, P. G. Potvin, Polyhedron 2002, 21, 49Ϫ54.
A. El-ghayoury, A. Harriman, A. Khatyr, R. Ziessel, J. Phys.
Chem. A 2000, 104, 1512Ϫ1523.
K. Kalyanasundaram, M. K. Nazeeruddin, M. Grätzel, G. Vis-
cardi, P. Savarino, E. Barni, Inorg. Chim. Acta 1992, 198,
831Ϫ839.
A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser, A.
von Zelewsky, Coord. Chem. Rev. 1988, 84, 85Ϫ277.
C. Mikel, P. G. Potvin, Inorg. Chim. Acta 2001, 325, 1Ϫ8.
P. G. Potvin, P. U. Luyen, J. Bräckow, J. Am. Chem. Soc. 2003,
125, 4894Ϫ4906.
(d, JH,H ϭ 8 Hz, 4 H, ttpy-g), 7.58 (d, JH,H ϭ 5 Hz, 4 H, B-a),
7.52 (m, 8 H, B-aЈ, ttpy-a), 7.30 (m, 12 H, B-b, B-bЈ, ttpy-b), 2.58
(s, 6 H, CH3) ppm. ESI-MS: m/z (%) ϭ 338.5 (70) [M6ϩ].
C116H82F36N18P6Ru3·2H2O (2937.1): calcd. C 47.44, H 2.95, N
8.58; found C 47.60, H 2.94, N 8.25.
[16]
[17]
[18]
[19]
Acknowledgments
The authors thank A. B. P. Lever for access to instruments and the
Natural Sciences and Engineering Council of Canada for funding.
[20]
[21]
[1]
V. Balzani, F. Barigelletti, L. De Cola, Top. Curr. Chem. 1990,
[22]
158, 31Ϫ71.
F. Barigelletti, L. Flamigni, V. Balzani, J.-P. Collin, J.-P. Sauv-
age, A. Sour, E. C. Constable, A. M. W. Cargill Thompson, J.
Chem. Soc., Chem. Commun. 1993, 942Ϫ944.
D. R. Prasad, M. Z. Hoffman, J. Am. Chem. Soc. 1986, 108,
2568Ϫ2573.
[2]
S. Bodige, A. S. Torres, D. J. Maloney, D. Tate, G. R. Kinsel,
A. K. Walker, F. M. MacDonnell, J. Am. Chem. Soc. 1997,
[23]
[24]
[25]
[26]
[27]
[28]
119, 10364Ϫ10369.
[3]
S. Vaduvescu, P. G. Potvin, Inorg. Chem. 2002, 41, 4081Ϫ4083.
[4]
P. G. Potvin, P. U. Luyen, F. Al-Mutlaq, New J. Chem. 2001,
N. Kitamura, Y. Kawanishi, S. Tazuke, Chem. Lett. 1983,
25, 839Ϫ846.
1185Ϫ1188.
[5]
J.-P. Sauvage, J.-P. Collin, J.-C. Chambron, S. Guillerez, C. Co-
P. J. Delaive, J. T. Lee, H. Abruna, H. W. Sprintschnik, T. J.
Meyer, D. G. Whitten, Adv. Chem. Ser. 1977, 168, 28Ϫ43.
K. Kalyanasundaram, J. Kiwi, M. Grätzel, Helv. Chim. Acta
1978, 61, 2720Ϫ2730.
A. J. Bard, L. R. Faulkner, Electrochemical Methods, John
Wiley & Sons, New York, 1980.
B. J. Coe, J. A. Harris, K. Clays, A. Persoon, K. Wostyn, B. S.
Brunschweig, Chem. Commun. 2001, 1548Ϫ1549.
W. Spahni, G. Calzaferri, Helv. Chim. Acta 1984, 67, 450Ϫ454.
J.-P. Collin, S. Guillerez, J.-P. Sauvage, F. Barigelletti, L. De
Cola, L. Flamigni, V. Balzani, Inorg. Chem. 1991, 30,
4230Ϫ4238.
E. C. Constable, A. M. W. Cargill Thompson, D. A. Tocher,
M. A. M. Daniels, New J. Chem. 1992, 16, 855Ϫ867.
Received May 28, 2003
udret, V. Balzani, F. Barigelletti, L. De Cola, L. Flamigni,
Chem. Rev. 1994, 94, 993Ϫ1019.
[6]
R. Liegghio, P. G. Potvin, A. B. P. Lever, Inorg. Chem. 2001,
40, 5485Ϫ5486.
F. Kröhnke, Synthesis 1976, 1Ϫ24.
E. C. Constable, A. M. W. Cargill Thompson, J. Chem. Soc.,
[7]
[8]
Dalton Trans. 1992, 3467Ϫ3475.
[9]
[29]
[30]
L. Masciello, P. G. Potvin, Can. J. Chem. 2003, 81, 209Ϫ218.
[10]
S. Vaduvescu, P. G. Potvin, Acta Crystallogr., Sect. E 2003,
59, o483Ϫo484.
C. Chamchoumis, P. G. Potvin, J. Chem. Res. (M) 1998,
870Ϫ875.
B. P. Sullivan, J. M. Calvert, T. J. Meyer, Inorg. Chem. 1980,
19, 1404Ϫ1407.
M. Beley, J.-P. Collin, R. Louis, B. Metz, J.-P. Sauvage, J. Am.
[11]
[31]
[12]
[13]
Early View Article
Chem. Soc. 1991, 113, 8521Ϫ8522.
Published Online March 5, 2004
Eur. J. Inorg. Chem. 2004, 1763Ϫ1769
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1769