Metal-directed synthesis and photophysical studies of trinuclear V-shaped and pentanuclear X-shaped ruthenium and osmium metallorods and metallostars based upon 4'-(3,5-dihydroxyphenyl)-2,2':6',2"-terpyridine divergent units

Article abstract of DOI:10.1002/chem.200500119

A new series of V-shaped trinuclear metallorods and X-shaped pentanuclear metallostars has been prepared by the reaction of metal complexes bearing pendant phenolic functionalities with complexes containing electrophilic ligands. Specifically, {M(tpy)

Full text of DOI:10.1002/chem.200500119

Metal-Directed Synthesis and Photophysical Studies of Trinuclear V-Shaped
and Pentanuclear X-Shaped Ruthenium and Osmium Metallorods and
Metallostars Based upon 4’-(3,5-Dihydroxyphenyl)-2,2’:6’,2’’-terpyridine
Divergent Units
Edwin C. Constable,*[a] Robyn W. Handel,[a] Catherine E. Housecroft,[a]
Angeles Farrꢀn Morales,[b] Barbara Ventura,[b] Lucia Flamigni,[b] and
Francesco Barigelletti*[b]
Abstract: A new series of V-shaped tri-
nuclear metallorods and X-shaped pen-
tanuclear metallostars has been pre-
pared by the reaction of metal com-
plexes bearing pendant phenolic func-
tionalities with complexes containing
electrophilic ligands. Specifically, {M-
(tpy)2} motifs (M=Ru or Os; tpy=
2,2’:6’,2’’-terpyridine) bearing one or
two pendant 3,5-dihydroxyphenyl sub-
stituents at the 4-position of the central
ring of the tpy have been reacted with
the complexes [Ru(tpy)(Xtpy)]2+ (X=
Cl or Br) to form new ether-linked spe-
cies. The energy transfer from rutheni-
um to osmium in these complexes has
been investigated in detail and the effi-
ciency of transfer shown to be highly
temperature dependent; the energy
transfer is highly efficient at low tem-
perature, whereas at room temperature
nonradiative and nontransfer deactiva-
tion of the excited {Ru(tpy)2}* domains
is most significant.
Keywords: dendrimers
· nitrogen
heterocycles · osmium · photochem-
istry · ruthenium
Introduction
motifs (tpy=2,2’:6’,2’’-terpyridine, bpy=2,2’-bipyridine).[1–3]
Particular interest has centred upon systems derived from
multitopic ligands with multiple tpy metal-binding domains
that lead to multinuclear species with multiple {M(tpy)2}
motifs; these have an advantage over compounds with mul-
tiple stereogenic {M(bpy)3} motifs in that they give rise to
monodisperse species rather than mixtures of diastereoisom-
ers.[3–5] By controlling the relative spatial positioning and the
nature of the chemical connectivity between the metal cen-
tres, it is possible to devise systems that generate designed
structures showing controlled directional ruthenium-to-
osmium energy transfer or electron transfer.[6] We have also
drawn attention to the synthetic merits of metallostars (or
first-generation metallodendrimers) in which there is a
single branching point at the core.[1a,7] Such systems show
minimal interactions between surface functional groups and
are amenable to both divergent and convergent synthetic
strategies.
Metallodendrimers may incorporate metal centres at various
sites and a variety of synthetic methodologies have been de-
vised for the efficient preparation of species in which metal–
metal interactions may be fine-tuned, both with respect to
their magnitude and their directionality.[1–3] We and others
have been particularly interested in species in which lumi-
nescent ruthenium(ii) and osmium(ii) centres have been
built into dendrimers and related species and have devel-
oped strategies for the preparation of a wide variety of mul-
tinuclear compounds based upon {M(tpy)2} and {M(bpy)3}
[a] Prof. Dr. E. C. Constable, Dr. R. W. Handel,
Prof. Dr. C. E. Housecroft
Department of Chemistry, University of Basel
Spitalstrasse 51, 4056 Basel (Switzerland)
Fax : (+41)61-267-1005
E-mail: edwin.constable@unibas.ch
Although a range of strategies for the synthesis of metal-
lodendrimers have been developed, we and others have es-
tablished a metal-directed approach as an efficient, general
and facile method for the linking of metal-containing build-
ing blocks.[7,8] The basic strategy involves reactions at the or-
ganic component of metal complexes containing functional-
[b] Dr. A. Farrꢀn Morales, Dr. B. Ventura, Dr. L. Flamigni,
Dr. F. Barigelletti
Istituto per la Sintesi Organica e Fotoreattivitꢀ (ISOF-CNR)
Via P. Gobetti 101, 40129 Bologna (Italy)
Fax : (+39)051-639-9844
E-mail: franz@isof.cnr.it
4024
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200500119
Chem. Eur. J. 2005, 11, 4024 – 4034

FULL PAPER
ised heterocyclic ligands. The reaction of either free or coor-
dinated ligands containing nucleophilic substituents with a
complex containing an (activated) electrophilic ligand is
used for the covalent linking of two metal centres through a
new bridging ligand that is assembled in situ. This approach
has been particularly useful in the preparation of complexes
with ether bridges between the component metal-containing
building blocks[9] and is also effective in the preparation of
metallostars in which a core mononuclear electrophilic/nu-
cleophilic zeroth-generation complex is functionalised by re-
action with nucleophilic or electrophilic ligands or metal
complexes, respectively. Convergent approaches in which
bpy and tpy ligands with multiple metal-containing substitu-
ents are brought together by coordination to a metal at the
core of the zeroth generation have also proved success-
ful.[7,8]
The lowest-lying triplet states in ruthenium oligopyridine
complexes lie about 0.2 eV higher in energy than in the cor-
responding osmium complexes and light-driven Ru!Os
energy transfer is easily observed in heterometallic rutheni-
um(ii)–osmium(ii) systems.[6,9–12] In several such systems, the
direction of the vectorial energy flow may be controlled by
modifying the spatial arrange-
Our specific target was the bifurcating ligand 4’-(3,5-dihy-
droxyphenyl)-2,2’:6’,2’’-terpyridine (2). This can bind a metal
centre at the tpy domain and react with two electrophilic
centres at the pendant phenolic sites. The difference in hard-
ness of phenol/phenolate and pyridine nitrogen donors and
the chelating character of the tpy domain ensure that all re-
actions are highly regioselective.
Ligand synthesis: We adopted a variation on the Krçhnke
pyridine synthesis[14] for the preparation of 2 and considered
it optimal to protect the phenolic substituents as methyl
ethers through the ring synthesis. Subsequent deprotection
of the methoxy groups can be achieved in high yield in tpy
derivatives. The reaction of 2-acetylpyridine with 3,5-dime-
thoxybenzaldehyde proceeded smoothly in a sequential one-
pot process involving treatment with KOtBu in THF fol-
lowed by ammonium acetate/acetic acid to give a satisfacto-
ry yield of the dimethoxy compound 1.[15,16] Deprotection by
heating with pyridinium chloride[16,17] gave the bifurcating
ligand 2 in quantitative yield (Scheme 1). The new ligands 1
and 2 were fully characterised (see Experimental Section)
and had the expected spectroscopic properties.
ment of the Ru- and Os-based
units, although
a quantitative
clarification of the rate constants
is not always possible.[12] At
room temperature, the pentanu-
clear and trinuclear tpy-based
arrays reported herein are pho-
toinert towards Ru!Os photo-
induced energy transfer. Howev-
er, their study at low tempera-
ture provides deep insights into
the process: a practically com-
plete Ru!Os energy conversion
is observed at low temperature
and the measured rate constants
for energy transfer can be de-
scribed in terms of a dipole–
dipole transfer mechanism.[12]
Scheme 1. Synthesis of the bifurcating ligand 2 and the core mononuclear complexes for the trinuclear com-
plexes showing the ring-labelling scheme adopted. 1) KOtBu, THF; 2) NH4OAc, AcOH, EtOH, 54.5% yield;
3) pyridinium chloride, 2108C, quantitative yield; 4) [M(Ztpy)Cl3]; M=Os or Ru, Z=H (ligand=tpy) or 2-
thienyl (ligand=5). In the case of the complexes with ligand 5, the thienyl ring is labelled F.
Results and Discussion
Strategy: We planned a diver-
gent approach to the synthesis
of polynuclear complexes in which branching occurred in
the zeroth generation at bifurcating substituents. In previous
studies, we have used pentaerythritol cores and branching
sites to prepare high nuclearity metallodendrimers based
upon a connectivity of four.[5,13] In this paper, we describe
the preparation of metallocentric systems in which the
zeroth generation contains tpy metal-binding domains with
pendant multiple phenolic substituents. The key expansion
step in the synthesis involves the reaction of the pendant
phenolate functionality with electrophilic metal complexes.
Synthesis of trinuclear complexes: One of our long-term
aims is the construction of light-driven machines and for this
we require bi- or polyfunctional systems in which light-col-
lecting components are connected through energy- or elec-
tron-transfer processes to motifs that can subsequently per-
form chemical or electrochemical transformations. To this
end, the {M(tpy)2} motif is attractive as one of the tpy li-
gands may be further functionalised with the light-absorbing
components and the other with the conversion system (for
example, a carboxylic acid for anchoring to a semiconductor
Chem. Eur. J. 2005, 11, 4024 – 4034
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4025

E. C. Constable, F. Barigelletti et al.
in Grꢂtzel-type photovoltaic devices).[18] The key building
blocks are the mononuclear complexes [M(Ztpy)(2)]2+
denote a tpy ligand ready for use in the metal-directed as-
sembly step (usually 4’-chloro-2,2’:6’,2’’-terpyridine 4 or 4’-
bromo-2,2’:6’,2’’-terpyridine 3).
,
which are readily prepared by the reaction of ligand 2 with
[M(Ztpy)Cl3] (M=Ru or Os; Z=H or thienyl) under reduc-
ing conditions (Scheme 1). As representative compounds,
we describe herein complexes with tpy and 4’-(2-thienyl)-
2,2’:6’,2’’-terpyridine (5) auxiliary ligands decorating the sur-
face (ligands as depicted in Scheme 2). The choice of 5
We have previously shown that complexes containing 4’-
halopyridine moieties are activated towards attack at the C-
4’ position by nucleophiles,[20,21] and we have used this phe-
nomenon for the preparation of metallorods, metallostars
and metallodendrimers[5,7,13] by reaction with oxygen-based
nucleophiles. The complexes [M(Ztpy)(2)]2+ (M=Ru or Os,
Ztpy=tpy or 5) react smoothly with the electrophilic com-
plexes [Ru(Ztpy)(Xtpy)]2+ (M=Ru or Os; Ztpy=tpy or 5,
Xtpy=3 or 4) in MeCN in the presence of K2CO3 to give
the desired homo- or heterometallic trinuclear complexes
[(Ztpy)M(6){Ru(Ztpy)}2]6+ containing the new bifurcated
tritopic bridging ligand 6 (Scheme 3).
The trinuclear complexes are V-shaped and intermetallic
distances may be estimated by molecular modelling. We
have modelled these complexes using molecular mechanics
methods (MMFF in Spartan ’04), and although this approach
ignores intercation repulsion, interactions within and be-
tween ligands are handled well. By constraining the metric
quantities within the {M(tpy)2} units to crystallographically
ꢀ
ꢀ
observed values (M Nterminal 2.07 ꢃ; M Ncentral 1.98 ꢃ), relia-
ble structures may be obtained (Figure 1). In particular, we
Scheme 2. Structures and ring-labelling scheme for the ligands.
arose from the unusual chemical and photophysical proper-
ties of complexes with this ligand.[16,19] All new complexes
were fully characterised and exhibited the expected spectro-
scopic properties. 1H NMR spectroscopy is particularly
useful for the characterisation of these complexes and the
singlets due to the H3 protons of the central pyridine rings
of the tpy domains make these useful spectator groups as
the polynuclear systems are developed.
Figure 1. MMFF (Spartan ’04) minimised structure of the trinuclear V-
shaped complex [(tpy)Ru(6){Ru(tpy)}2] showing the expected conforma-
tion of the ether-linked moieties; the Rucentral···Ruouter and Ruouter···Ruouter
distances are in the ranges 12.3–12.6 ꢃ and 14.0–14.2 ꢃ, respectively. Hy-
drogen atoms have been omitted for clarity.
Herein, we describe systems with up to 11 magnetically
independent aromatic rings and make a note here of the no-
menclature used. In ligand 2 and functionalised derivatives
thereof, the terminal rings of the tpy are always denoted as
A and the central tpy ring as B, with the phenyl substituent
being the C ring. The constituent rings of the Ztpy ligand in
[M(Ztpy)(2)]2+ (Scheme 1) and derivatives are described as
rings D (terminal), E (central) and F (thienyl substituent).
For mononuclear complexes not containing 2, A, B and C
note that the chemically sensible and sterically favoured ori-
entation of one {(Otpy)Ru(tpy)} motif above and the other
below the phenyl ring reproduces well that found in the
only structurally characterised noncyclic 3,5-diaryloxyar-
ene.[22] The Mcentral···Ruouter and Ruouter···Ruouter distances are
in the ranges 12.3–12.6 ꢃ and 14.0–14.2 ꢃ, respectively.
Spectroscopic and other properties will be discussed togeth-
er with those of the pentanuclear species.
4026
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4024 – 4034

Ruthenium and Osmium Metallorods and Metallostars
FULL PAPER
the coupling reaction is that
due to proton B3, which shifts
downfield
from
d=8.92ꢁ
0.04 ppm to d=9.35ꢁ0.07 ppm.
The ES mass spectra of the
complexes provide compelling
evidence for the coupling reac-
tions and the nuclearity of the
products; in addition to the ex-
pected [MꢀPF6]+ peaks, series
of peaks arising from more
highly charged ions with the ap-
propriate isotopomer distribu-
tions and peak separations are
observed. For example, the ES
mass spectrum of [Ru((6){Ru-
(tpy)}2)2][PF6]10 exhibits clusters
of peaks with m/z correspond-
ing
to
[Mꢀ2PF6]2+
,
,
,
[Mꢀ3PF6]3+
,
[Mꢀ4PF6]4+
[Mꢀ5PF6]5+ and [Mꢀ6PF6]6+
all of which show the expected
isotopomer distributions for
pentaruthenium species.
Electrochemical properties of
the complexes: All of the metal
complexes are redox-active,
showing both metal-centred
and ligand-centred processes
(Table 1). Those complexes
Scheme 3. Synthesis of the bifurcated trinuclear complexes showing the ring-labelling scheme adopted.
1) K2CO3, MeCN; M=Os or Ru, Z, X=H (tpy) or 2-thienyl (5). In the case of the complexes with 5, the
thienyl ring is labelled F when it is in the central position attached to ring E, but is labelled L when attached
to the terminal sites on ring K.
Synthesis of the pentanuclear complexes: The same strategy
as described above was adopted for the preparation of the
X-shaped pentanuclear species, with the key intermediates
being the tetraphenolic complexes [M(2)2]2+. Reaction of
[M(2)2]2+ with [M(Ztpy)(Xtpy)]2+ (M=Ru or Os; Ztpy=
tpy or 5, Xtpy=3 or 4) in MeCN in the presence of K2CO3
gave the homo- or heterometallic pentanuclear complexes
[M((6){Ru(Ztpy)}2)2]10+ in moderate but acceptable yields
(Scheme 4).
1
Characterisation of the complexes: The H NMR spectra of
all the trinuclear and pentanuclear complexes are sharp and
well-resolved and have been fully assigned by a combination
of NOESY and COSY techniques. Figure 2 shows the
500 MHz COSY 90 spectrum of a solution of the V-shaped
complex [(tpy)Ru(6){Ru(tpy)}2][PF6]6 in CD3CN, illustrating
how all 24 pyridine proton environments are readily ob-
served and how the connectivity within each of the magneti-
cally independent subspectra may be established. The most
noticeable feature on going from the complexes of 2 to
those of 6 is a downfield shift of the signals of the C2 (d=
7.11ꢁ0.02 ppm) and C4 (d=6.55ꢁ0.04 ppm) protons by
Figure 2. 500 MHz COSY 90 spectrum (room temperature) of a solution
of [(tpy)Ru(6){Ru(tpy)}2][PF6]6 in CD3CN.
1.36 and 1.44 ppm, respectively. The other indicator peak for
Chem. Eur. J. 2005, 11, 4024 – 4034
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4027

E. C. Constable, F. Barigelletti et al.
Scheme 4. Synthesis of the dendritic pentanuclear complexes showing the ring-labelling scheme adopted. 1) RuCl3·3H2O or K2[OsCl6], ethane-1,2-diol,
N-ethylmorpholine, microwave; 2) K2CO3, MeCN; M=Os or Ru, X=H (tpy) or 2-thienyl (5). In the case of the complexes with 5, the thienyl ring is la-
belled L.
thenium and pentaruthenium
nuclear building blocks and metallostars in MeCN (1.0m [nBu4N][PF6], Pt or glassy carbon working electrode, processes by cyclic voltamme-
Table 1. Redox potentials (E (EaꢀEc) in volts) measured by cyclic or square-wave voltammetry of the mono-
Pt wire counter-electrode, Ag wire, referenced against internal Fc/Fc+). All metal-based processes were rever-
try, differential pulse voltamme-
try or square-wave voltamme-
try, and a single, apparently re-
versible wave was observed.
However, in the heterometallic
complexes [(5)Os(6){Ru(tpy)}2]-
[PF6]6, [(5)Os(6){Ru(5)}2][PF6]6
sible;[a] the ligand reductions were irreversible and potentials are taken from the forward scan. All complexes
were used as hexafluorophosphate salts.
Compound
Ligand
Ru
Os
tpy
[(tpy)Ru(3)]2+
0.920
0.896
0.901
0.889
ꢀ1.55, ꢀ1.67, ꢀ1.93
ꢀ1.56, ꢀ1.80
[(5)Ru(4)]2+
[(tpy)Ru(2)]2+
1.34
1.77
1.78, 1.56
ꢀ1.64, ꢀ1.85
[(5)Ru(2)]2+
ꢀ1.27, ꢀ1.72, ꢀ1.94
ꢀ1.56, ꢀ1.79
[(5)Os(2)]2+
0.535
0.516
and
[Os((6){Ru(5)}2)2][PF6]10,
[Os{2}2]2+
ꢀ1.38, ꢀ1.66
both osmium and ruthenium
processes in the correct current
ratio are observed, indicating
that the observation of a single
signal for the tri- and pentaru-
thenium complexes arises from
an unfortunate degeneracy of
the redox processes for the
inner and outer centres, rather
[(tpy)Ru(6){Ru(tpy)}2]6+
[(5)Ru(6){Ru(tpy)}2]6+
[(5)Os(6){Ru(tpy)}2]6+
[(5)Os(6){Ru(5)}2]6+
[Ru((6){Ru(tpy)}2)2]10+
[Os((6){Ru(5)}2)2]10+
0.854
0.846
0.840
0.824
0.892
0.835
ꢀ1.62, ꢀ1.83, ꢀ1.99
ꢀ1.60, ꢀ1.80, ꢀ1.93
ꢀ1.62, ꢀ1.77, ꢀ1.95
ꢀ1.60, ꢀ1.79, ꢀ1.98
ꢀ1.63, ꢀ1.81
1.72
1.45
1.47
0.535
0.543
1.48
0.563
ꢀ1.61, ꢀ1.78, ꢀ1.95
[a] Reversible peaks showed the same DE=EaꢀEc as ferrocene in the same solvent.
containing ligand 5 also exhibit thiophene-based oxidation
waves. In acetonitrile solution, the ruthenium and osmium
complexes exhibit reversible metal(iii)/metal(ii) processes at
+0.8 to +0.9 and +0.5 to +0.6 V (versus Fc/Fc+), respec-
tively, typical of the {M(tpy)2} centres present. It was not
possible to resolve separate metal-centred processes corre-
sponding to the central and the outer metals for the triru-
than a shifting of one of the two processes out of the observ-
able window subsequent to the oxidation of either the inner
or outer metal centres.
Absorption and emission spectroscopy: Table 2 lists absorp-
tion maxima and absorption coefficients for solutions of the
tri- and pentanuclear complexes in MeCN as well as those
4028
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4024 – 4034

Ruthenium and Osmium Metallorods and Metallostars
FULL PAPER
Table 2. Absorption spectroscopic data for the polynuclear complexes and related mononuclear building
blocks. All data are for solutions of [PF6]ꢀ salts in MeCN at room temperature.
volving ligand 5, for both the
absorption and emission fea-
tures. This resulted in a modu-
lation of the Ru!Os energy
Compound
MLCT
Ligand centred
lmax [nm] (emax [104 dm3 molꢀ1 cmꢀ1])
lmax [nm] (emax [104 dm3 molꢀ1 cmꢀ1])
[(tpy)Ru(3)]2+
480 (1.5)
490 (2.6)
480 (1.7)
495 (2.1)
670 (0.78), 494 (2.9)
490 (2.0)
670 (0.50), 489 (2.0)
480 (5.0)
324 sh, 307 (5.9), 272 (4.0)
326 sh, 305 (6.3), 275 (4.6)
330 sh, 306 (5.6), 272 (3.8)
330 sh, 311 (4.9), 283 (4.4)
314 (2.24), 284 (5.7)
329 sh, 310 (5.1), 284 (4.3)
331 sh, 314 (6.0), 285 (5.0)
324 sh, 306 (16.0), 272 (12.0)
329 sh, 306 (13.5), 275 (10.7)
306 (17.0), 276 (13.0)
transfer
rate
between
and
[(5)Ru(4)]2+
[(5)Os(6){Ru(5)}2]6+
[(tpy)Ru(2)]2+
[(5)Os(6){Ru(tpy)}2]6+
.
In
[(5)Ru(2)]2+
[(5)Os(2)]2+
points 1)–4) below, we describe
the results obtained at room
temperature. The glassy matrix
data at 77 K are discussed later.
[Ru(2)2]2+
[Os(2)2]2+
[(tpy)Ru(6){Ru(tpy)}2]6+
[(5)Ru(6){Ru(tpy)}2]6+
[(5)Os(6){Ru(tpy)}2]6+
[(5)Os(6){Ru(5)}2]6+
[Ru((6){Ru(tpy)}2)2]10+
[Os((6){Ru(5)}2)2]10+
490 (4.8)
670 (0.8), 490 (6.0)
670 (0.8), 500 (7.0)
485 (10.5)
330 sh, 306 (18.0), 285 (16.0)
306 (26.0), 275 (24.0)
331 sh, 306 (22.0), 286 (19.0)
1) Each member of the series
of V-shaped trinuclear and X-
shaped pentanuclear complexes
can be considered in terms of
clearly defined building blocks.
This is evident from the absorp-
670 (0.71), 495 (10.5)
for the relevant mononuclear building blocks. All of the
complexes exhibit intense UV absorption bands correspond-
ing to ligand-centred (1LC) transitions, and absorptions in
the visible region attributable to metal-to-ligand charge-
transfer (1MLCT) transitions. These latter bands fall in the
range 480–490 nm (with absorption coefficients e=1.4–2.6ꢄ
104 mꢀ1 cmꢀ1 per {M(tpy)2} chromophore for both RuII and
OsII centres). An additional absorption at ꢂ670 nm (e=
(6ꢁ2)ꢄ103 mꢀ1 cmꢀ1) is observed for the OsII-containing
complexes and is assigned to the formally spin-forbidden
tion spectra of the polynuclear species, which appear as su-
perimpositions of the spectra of the component mononu-
clear model complexes. This is consistent with a substantial
ground-state electronic decoupling of the components, as a
consequence of the insulating character of the ether connec-
tions.[9,11] Similar conclusions follow from the electrochemi-
cal data (Table 1). With regard to the pentanuclear species,
the absorption spectra presented in Figure 3 provide an il-
lustration of this point, with the additional consideration
that the presence of the thienyl group in ligand 5 stabilises
the MLCT levels involving this ligand.
3
transition directly to the {Os(tpy)2} MLCT state.[23] Table 3
lists the luminescence properties of model mononuclear
complexes along with those of the new trinuclear and penta-
nuclear complexes, as observed at room temperature and at
77 K in air-equilibrated acetonitrile.
The spectroscopic properties of the trinuclear complexes
[(5)Os(6){Ru(tpy)}2]6+ and [(5)Os(6){Ru(5)}2]6+ have been
briefly discussed in a previous report.[9] The pendant thienyl
group was found to stabilise (redshift) the MLCT levels in-
2) The mononuclear and homometallic polynuclear com-
plexes of ruthenium(ii) are all very weak emitters at room
temperature (F<10ꢀ4, lmax =630–670 nm, lexc =480 nm),
with lifetimes in the range of few nanoseconds or less
(Table 3). These features are in general agreement with the
wealth of results from other complexes based on {Ru(tpy)2}
chromophores.[23]
Table 3. Luminescence properties of model mononuclear complexes and those of the new trinuclear and pentanuclear complexes.[a]
298 K 77 K
Ru-based region
Os-based region
Ru-based region
Os-based region
lmax [nm]
105
f
t [ns]
lmax [nm]
102
f
t [ns]
lmax [nm]
t [ms]
lmax [nm]
t [ms]
[(tpy)Ru(2)]2+[b]
[Ru(2)2]2+
630
634
670
<3.0
4.5
15.0
0.6
1.4
6.4
615
622
654
11.0
10.5
11.5
[(5)Ru(2)]2+[b]
[Os(2)2]2+
732
744
1.8
2.0
110
150
722
745
2.0
2.0
[(5)Os(2)]2+[b]
[(tpy)Ru(6){Ru(tpy)}2]6+
[(5)Ru(6){Ru(tpy)}2]6+
[(5)Os(6){Ru(tpy)}2]6+[b]
640
<3.0
ꢃ0.6
630
650
630
11.0
10.0
670
<3.0
ꢃ0.6
[c]
[c]
[c]
745
745
(0.6)
(0.5)
120
120
1.8ꢄ10ꢀ3
750
750
1.6ꢄ10ꢀ3[d]
2.0[d]
[(5)Os(6){Ru(5)}2]6+[b]
630
>5ꢄ10ꢀ3[b]
36.0ꢄ10ꢀ3[d]
2.0[d]
[c]
[c]
[c]
[Ru((6){Ru(tpy)}2)2]10+
[Os((6){Ru(5)}2)2]10+
662
3.6
~1
630
656
10.3
[c]
[c]
[c]
740
0.5
110
34ꢄ10ꢀ3
724
39ꢄ10ꢀ3[d]
2.0[d]
[a] In air-equilibrated acetonitrile solvent, lexc =480 nm; [(tpy)Ru(3)]2+ and [(5)Ru(4)]2+ are both very weak emitters at room temperature, f<3ꢄ10ꢀ5
,
and have not been investigated in detail. [b] See also ref. [9]. [c] Too weak to detect. [d] Dual exponential behaviour; the indicated values refer to t1 and
t2 of I(t)=b1exp(ꢀt/t1)+b2exp(ꢀt/t2), where b1 and b2 are pre-exponential factors, and b1 is a negative value (see text).
Chem. Eur. J. 2005, 11, 4024 – 4034
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4029

E. C. Constable, F. Barigelletti et al.
served osmium-based luminescence intensities. This is illus-
trated by the inset in Figure 3, which displays the room tem-
perature luminescence spectra obtained by excitation of iso-
absorbing
(480 nm)
solutions
of
[Os(2)2]2+
and
[Os((6){Ru(5)}2)2]10+, the luminescence intensities of which
are in an approximate 4:1 ratio. We stress the fact that upon
480 nm excitation of isoabsorbing solutions followed by hy-
pothetical complete Ru!Os energy transfer, identical Os-
based luminescence intensities would be expected for
[Os(2)2]2+ and [Os((6){Ru(5)}2)2]10+
.
[24,25]
With regard to measurements at 77 K in frozen solvent,
Figure 4 shows the normalised luminescence profiles of the
pentanuclear complex [Os((6){Ru(5)}2)2]10+
, along with
Figure 3. The ground-state absorption spectra of [Os((6){Ru(5)}2)2]10+
(c), [Os(2)2]2+ (b), [Ru(2)(5)]2+ (g) and [Ru((6){Ru(tpy)}2)2]10+
(d). The inset shows the room temperature luminescence spectra of
[Os((6){Ru(5)}2)2]10+ (c) and [Os(2)2]2+ (b), as obtained from iso-
absorbing solutions at lexc =480 nm.
3) The mononuclear osmium(ii) complexes, [Os(2)2]2+ and
[(5)Os(2)]2+, are more strongly emitting than their rutheni-
um(ii) analogues and exhibit luminescence quantum yields
F~2ꢄ10ꢀ3 (lmax =732 and 744 nm, respectively, lexc
=
480 nm) with luminescence lifetimes of 110 and 150 ns, re-
spectively. These wavelengths and lifetime features are typi-
cal, and indeed characteristic, of {Os(tpy)2} luminophores in
fluid solution at room temperature.[23]
Figure 4. Normalised
luminescence
spectra
at
77 K
for
4) For the heterometallic species [(5)Os(6){Ru(tpy)}2]6+
,
[Os((6){Ru(5)}2)2]10+ (c), [Os(2)2]2+ (b) and [Ru(2)(5)]2+ (g);
lexc =480 nm in all cases.
[(5)Os(6){Ru(5)}2]6+ and [Os((6){Ru(tpy)}2)2]10+, excitation
at room temperature at 480 nm (at which both Ru-based
and Os-based chromophores are isoabsorptive and absorb
light in approximately 2:1 and 4:1 ratios for the trinuclear
and pentanuclear species, respectively), no ruthenium-based
emission (below 700 nm) is detected. Instead, only osmium-
based emissions at greater than 700 nm are observed for
these heterometallic complexes. This conclusion is based on
a comparison of the lmax and t values listed in Table 3. In
addition, for the heterometallic species, the luminescence
yield is f=0.5–0.6ꢄ10ꢀ2, that is, a third to a quarter of that
of the reference mononuclear complexes [Os(2)2]2+ and
[(5)Os(2)]2+. This can be rationalised by the fact that, in the
heterometallic complexes at room temperature, the fraction
of light absorbed by the ruthenium-based chromophores
does not result in Ru!Os energy transfer (which is energet-
ically allowed, with the osmium states lying ~0.25 eV lower
in energy, as estimated from the ruthenium- and osmium-
based emission band maxima (Table 3)) and no sensitisation
of the {Os(tpy)2} luminescence occurs. On this basis, and
given that the lifetimes of the ruthenium-based lumino-
phores investigated here are in the range 0.6 to 6 ns (at
room temperature, see Table 3), one concludes that the
Ru!Os intramolecular energy transfer must be slower than
the intrinsic deactivation, ken ꢃkd, with kd =1.6–10ꢄ108 sꢀ1.
Conversely, only the fraction of light directly absorbed by
the osmium-based chromophores is responsible for the ob-
those of [(5)Ru(2)]2+ and [Os(2)2]2+, the mononuclear com-
plexes that can be regarded as the reference component
units. It can be seen that for the pentanuclear complex, the
ruthenium-based emission region at 650 nm exhibits
a
strong reduction in intensity compared to the mononuclear
reference compound. Evidence for quenching and sensitisa-
tion phenomena can be gained from time-resolved experi-
ments, as illustrated for [Os((6){Ru(5)}2)2]10+ in Figure 5. In
particular, while the luminescence of [(5)Ru(2)]2+ (observed
at 650 nm) is rather long-lived (t=11.5 ms), the Ru-based
emission for [Os((6){Ru(5)}2)2]10+ (observed at 650 nm) ex-
hibits tRu =34 ns. In addition, for this complex, the analysis
of the luminescence decay detected at 730 nm (osmium-
based emission region) required a dual exponential law,
I(t)=b1exp(ꢀt/t1)+b2exp(ꢀt/t2), where b1 and b2 are pre-ex-
ponential factors, and b1 is a negative value. In this case, we
found t1 =39 ns and t2 =2.0 ms. Further results relating to
the time-dependent luminescence properties of the com-
plexes are collected in Table 3.
Photoinduced processes: The results obtained at 77 K (Fig-
ures 4 and 5) are consistent with effective intramolecular
Ru!Os energy transfer in [Os((6){Ru(5)}2)2]10+. This causes
quenching of the donor luminescence (tRq u =34 ns, as op-
4030
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4024 – 4034

Ruthenium and Osmium Metallorods and Metallostars
FULL PAPER
geometric factor (for statistical reasons, K2 =2= ), f and t are
3
the luminescence quantum yield and lifetime of the donor,
respectively; h is the refractive index of the solvent and dcc
is the interchromophore distance. By using the 77 K lumi-
nescence properties of the donor model complex
[(5)Ru(2)]2+, t=11.5 ms and f=0.30,[9] and the absorption
features of the acceptor model [Os(2)2]2+, for the pentanu-
clear case one obtains JF =6.7ꢄ10ꢀ14 cm3 mꢀ1 and R0 =39 ꢃ.
Based on Equation (2c), dcc =15 ꢃ (cf. 16.5 ꢃ for
[(5)Os(6){Ru(5)}2]6+[9]), consistent with the molecular-mod-
elling results.
In summary, for the heterometallic trinuclear and penta-
nuclear species at 77 K (Table 3), the Ru!Os energy trans-
fer is quite an efficient process, at variance with what is ob-
served at room temperature. This can be understood on the
basis of the fact that Ru!Os energy transfer for tpy-type
complexes is little affected by the temperature or the state
(fluid or frozen) of the solvent.[27] Thus, at room tempera-
ture, the intrinsic deactivation at the Ru-based donor unit
(Table 3) is much faster than energy transfer, kRd T >ken; the
reverse is true at 77 K, k7d7 K <ken.
Figure 5. Time-resolved luminescence properties of [Os((6){Ru(5)}2)2]10+
at 77 K, as observed at 650 nm (ruthenium-based region, single-exponen-
tial decay, t=34 ns) and 730 nm (osmium-based region, dual-exponential
decay, t1 =39 ns and t2 =2.0 ms); see text and Table 3. The time profile of
the 465 nm NanoLED excitation source is also shown (norm.=normal-
ised.).
posed to tRu =11.5 ms in the reference complex [(5)Ru(2)]2+),
and concomitant sensitisation of the osmium-based lumines-
cence with t1 =39 ns (a rise time, see Figure 5) and t2 =
2.0 ms. The Ru!Os energy-transfer rate constant ken can
then be calculated according to ken =1/tRq uꢀ1/tRu. This yields
ken =2.6ꢄ107 sꢀ1, practically identical to the value found pre-
Conclusion
We have demonstrated that the use of metal-activated or-
ganic electrophiles is an efficient and controllable method
for the preparation of V-shaped metallorods and X-shaped
metallostars. In particular, [(5)Os(6){Ru(tpy)}2]6+ and
[Os((6){Ru(5)}2)2]10+ are of special interest as the rutheni-
um-to-osmium energy transfer is temperature dependent. At
low temperatures, the energy transfer is efficient and
osmium-centred emission is observed, whereas at room tem-
perature the ruthenium luminophores are efficiently deacti-
vated and no energy transfer is observed. These observa-
tions introduce an additional parameter that may be adjust-
ed in order to tune the energy-transfer processes in antenna
devices.
viously for the trinuclear complex [(5)Os(6){Ru(5)}2]6+
,
ken =2.8ꢄ107 sꢀ1.[9] This is consistent with an energy-transfer
efficiency, fen ~unity, defined in Equation (1):
ken
ken þ kd
ð1Þ
ꢀen
¼
in which kd is the intrinsic deactivation rate constant for the
unquenched Ru-based donor determined for the model;
kd =1/tRu, with tRu =11.5 ms (Table 3). Given the insulating
character of the ether linkage,[9,11] the Ru!Os energy-trans-
fer step within the trinuclear and pentanuclear species can
be described using the Fçrster dipole–dipole energy-transfer
approach[12,26] [Eqs. (2a)–(2c)]:
Experimental Section
General methods: Infrared spectra were recorded on Mattson Genesis
and Shimadzu FTIR8300 Fourier-transform spectrophotometers with
samples in compressed KBr discs or as solids using a Golden Gate ATR
accessory. 1H and 13C NMR spectra were recorded on Bruker AC300,
AV300 and DRX500 spectrometers at room temperature; the ring-label-
ling scheme adopted for the ligands is shown in the structure diagrams;
chemical shifts are referenced with respect to TMS (d=0 ppm). EI mass
spectra were recorded on a Kratos MS50 instrument. Electrospray mass
spectra (ES-MS) were recorded by using Finnigan MAT LCT and LCQ
mass spectrometers. Electrochemical measurements were performed with
an Autolab PGSTAT20 system using platinum or glassy-carbon working
and auxiliary electrodes with an Ag/AgCl electrode as reference, using
purified acetonitrile as solvent and 0.1m [nBu4N][BF4] as the supporting
electrolyte. Potentials are quoted versus the ferrocene/ferrocenium
couple (Fc/Fc+ =0.0 V) and were referenced to internal ferrocene added
at the end of each experiment. All glassware was flame-dried prior to use
and flushed with N2; K2CO3 was dried at 1708C for at least 24 h before
use.
R
4
~
~
~
~
FðnÞeðnÞ=n dn
R
ð2aÞ
ð2bÞ
ð2cÞ
JF ¼
~
FðnÞ
R0 ¼ 9:79 ꢄ 103ðK2hꢀ4FJFÞ
1=6
ꢀ
ꢁ
R0
dcc
6
ken ¼ kd
Here, JF is the Fçrster overlap integral between the lumines-
cence spectrum of the donor, F(n˜), and the absorption spec-
trum of the acceptor, e(n˜), on an energy scale (cmꢀ1); R0 is
the critical transfer radius (the interchromophore distance
for which the energy-transfer efficiency is fen =0.5); K2 is a
Chem. Eur. J. 2005, 11, 4024 – 4034
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4031

E. C. Constable, F. Barigelletti et al.
Absorption spectra of dilute solutions in acetonitrile (1–2ꢄ10ꢀ5 m) were
measured with a Perkin-Elmer Lambda 45 UV/Vis spectrophotometer.
Luminescence spectra were obtained with a Spex Fluorolog II spectro-
fluorimeter equipped with a Hamamatsu R928 phototube. Air-equilibrat-
ed sample solutions were excited at 480 nm and their concentrations
were adjusted so as to obtain absorbance values ꢃ0.15. Low-temperature
measurements were performed using capillary tubes immersed in liquid
nitrogen. Whilst uncorrected luminescence band maxima are referred to
throughout the text, corrected spectra were employed for the determina-
tion of the luminescence quantum yields (the correction procedure was
based on the use of software that takes care of the wavelength-dependent
phototube response). From the wavelength-integrated area of the cor-
rected luminescence spectra, we obtained luminescence quantum yields f
for the samples with reference to [Ru(bpy)3]Cl2 (r, fr =0.028 in air-equili-
brated water[28]) through the use of Equation (3):[29]
(dt, 2H; A5), 6.84 (d, J=2.2 Hz, 2H; C2), 6.41 ppm (t, 1H; C4); 13C
NMR (100 MHz, [D6]DMSO): d=147.4 (A6), 140.1 (A4), 125.4 (A5),
122.2 (A3), 118.9 (B3), 104.8 (C2), 103.8 ppm (C4); due to low solubility,
the quaternary C could not be observed; IR (solid): n˜ =3055 (m), 2827
(m), 2727 (m), 1593 (s), 1523 (s), 1473 (w), 1419 (m), 1353 (m), 1330 (m),
1292 (m), 1234 (m), 1215 (m), 1153 (s), 1095 (w), 1026 (m), 1002 (m), 956
(m), 879 (w), 844 (s), 783 (s), 729 (m), 702 (m), 678 (m), 648 (m), 617
(m), 567 (s), 532 cmꢀ1 (m); MS (ESI, high resolution): m/z calcd for
C21H16N3O2 [M+H]+: 342.124; found: 342.123; MS (ESI): m/z: 364
[M+Na]+, 342 [M+H]+.
[Ru(tpy)(3)][PF6]2:
A mixture of 4’-bromo-2,2’:6’,2’’-terpyridine (3;
65.0 mg, 0.209 mmol), [Ru(tpy)Cl3] (95.0 mg, 0.216 mmol) and N-ethyl-
morpholine (4 drops) in EtOH (20 mL) was heated at reflux for 3 h.
Aqueous NH4PF6 was then added and the red precipitate that formed
was collected by filtration. Purification of the crude product by column
chromatography (SiO2; MeCN/saturated aqueous KNO3/H2O, 7:1:0.5)
gave [Ru(tpy)(3)][PF6]2 as an orange solid. Yield: 142 mg (72.5%); 1H
NMR (300 MHz, CD3CN): d=8.97 (s, 2H; B3), 8.73 (d, J=8.1 Hz, 2H;
E3), 8.47 (d, J=7.4 Hz, 4H; A3+D3), 8.40 (t, 1H; E4), 7.91 (m, 4H;
A4+D4), 7.35 (m, 4H; A6+D6), 7.15 ppm (m, 4H; A5+D5); IR
(solid): n˜ =1600 (w), 1446 (m), 1423 (m), 1392 (m), 1338 (m), 1288 (m),
1242 (w), 1161 (w), 1107 (w), 1053 (w), 902 (w), 817 (s), 786 (s), 763 (s),
648 (m), 621 (m), 551 cmꢀ1 (s); MS (ESI): m/z: 793 [MꢀPF6]+, 324
ꢀ
Arh2ðareaÞ
Ah2rðareaÞr
¼
ð3Þ
ꢀr
in which A and h refer to the absorbance values and refractive indices of
the two solutions. Band maxima and relative luminescence intensities
were measured with estimated uncertainties of 2 nm and 20%, respec-
tively. Luminescence lifetimes were measured using an IBH 5000F
single-photon counting spectrometer; excitation was performed at
465 nm (with a NanoLED source); single- or double-exponential analyses
of the decays were performed with the help of software provided by
IBH. The estimated uncertainty for the lifetime values is 8%.
[Mꢀ2PF6]2+
.
[Ru(5)Cl3]: A mixture of RuCl3·3H2O (261 mg, 1.00 mmol) and 4’-(2-
thienyl)-2,2’:6’,2’’-terpyridine (5; 315 mg, 0.100 mmol) in EtOH (200 mL)
was heated at reflux for 2 h. The reaction mixture was cooled and the
brown solid that formed was collected by filtration. The crude [Ru(5)Cl3]
was washed with ethanol, dried and used without further purification.
All metal complexes were repeatedly recrystallised prior to use in photo-
physical studies and were judged to be pure on the basis of chromato-
graphic and spectroscopic studies. In our hands, combustion microanaly-
ses of multinuclear ruthenium or ruthenium/osmium oligopyridine com-
plexes have been found to result in data with anomalously low carbon
figures due to carbide formation.
[Ru(4)(5)][PF6]2: This complex was obtained as a red solid by following
the method described for [Ru(tpy)(3)][PF6]2 but using 4’-chloro-2,2’:6’,2’’-
terpyridine (4; 54.0 mg, 0.202 mmol) and [Ru(5)Cl3] (108 mg,
0.207 mmol). Yield: 120.7 mg (61.3%); 1H NMR (300 MHz, CD3CN):
d=8.92 (s, 2H; E3), 8.83 (s, 2H; B3), 8.65 (d, J=8.1 Hz, 2H; D3), 8.49
(d, J=8.1 Hz, 2H; A3), 8.18 (dd, J=1.1, 3.7 Hz, 1H; F3), 7.94 (dt, J=
1.1, 8.1 Hz, 4H; A4+D4), 7.83 (dd, J=1.1, 5.1 Hz, 1H; F5), 7.46 (d, J=
5.5 Hz, 2H; A6), 7.42 (dd, 1H; F4), 7.38 (d, J=5.5 Hz, 2H; D6), 7.19
(ddd, 2H; A5), 7.17 ppm (ddd, 2H; D5); IR (solid): n˜ =1595 (w), 1539
(w), 1427 (m), 1342 (s), 1288 (m), 1245 (m), 1114 (w), 1029 (w), 825 (s),
786 (s), 756 (m), 702 (m), 651 (m), 555 (s), 528 cmꢀ1 (m); MS (ESI): m/z:
4’-(3,5-Dimethoxyphenyl)-2,2’:6’,2’’-terpyridine
(1):
2-Acetylpyridine
(690 mL, 5.8 mmol) was added dropwise to a solution of KOtBu (1.04 g,
9.3 mmol) in THF (50 mL) and the mixture was stirred at room tempera-
ture for 30 min to give a pale yellow suspension containing the enolate.
A solution of 3,5-dimethoxybenzaldehyde (508 mg, 3.06 mmol) in THF
(5 mL) was then added, whereupon the reaction mixture immediately
became clear and bright orange. After stirring for 17 h at room tempera-
ture, a suspension of dried NH4OAc (5 g) in 2:1 EtOH/AcOH (60 mL)
was added and the reaction mixture was heated to reflux for 5 h. It was
then cooled to room temperature, poured onto ice and after 2 h water
(300 mL) was added, resulting in the deposition of 1 as an off-white pre-
cipitate, which was collected by filtration. Extraction of the filtrate with
CH2Cl2 (100 mL) gave a brown oil from which a further small amount of
1 was obtained after chromatographic separation (Al2O3, toluene/5%
Et2NH). Yield: 545 mg (50.9%); 1H NMR (300 MHz, CDCl3): d=8.73
(d, J=4.8 Hz, 2H; A6), 8.69 (s, 2H; B3), 8.67 (d, J=8.1 Hz, 2H; A3),
7.88 (dt, J=1.8, 7.7 Hz, 2H; A4), 7.36 (ddd, 2H; A5), 7.01 (d, J=2.2 Hz,
2H; C2), 6.56 (t, 1H; C4), 3.69 ppm (s, 6H; OCH3); 13C NMR (125 MHz,
CDCl3): d=161 (C3), 156 (B2), 155 (A2), 150 (B4), 149 (A6), 140 (C1),
136 (A4), 123 (A5), 121 (A3), 119 (B3), 105 (C2), 101 (C4), 55 ppm
(CH3); IR (solid): n˜ =2939 (w), 2839 (w), 2360 (w), 1581 (s), 1465 (m),
1388 (s), 1334 (w), 1296 (w), 1203 (s), 1149 (s), 1064 (m), 987 (w), 894
(w), 786 (s), 732 (s), 694 (m), 655 cmꢀ1 (s); MS (EI): m/z: 369 [M]+, 339
[Mꢀ2Me]+.
4’-(3,5-Dihydroxyphenyl)-2,2’:6’,2’’-terpyridine (2): Concentrated hydro-
chloric acid (17.6 mL) and pyridine (16 mL) were heated under N2 at
2108C for approximately 2 h with continuous removal of water until the
solution reached a constant internal temperature of 2108C. After cooling
to 1508C, 1 (1.1 g, 3.0 mmol) was added and the mixture was heated at
2108C for 3 h. It was then cooled to 1008C, whereupon warm water
(60 mL) was added. The pale precipitate that formed was collected by fil-
tration and dried over P2O5 in vacuo. Yield: quantitative; 1H NMR
(300 MHz, [D6]DMSO): d=8.92 (d, J=8.1 Hz, 2H; A3), 8.86 (d, J=
4.8 Hz, 2H; A6), 8.73 (s, 2H; B3), 8.30 (dt, J=1.5, 7.7 Hz, 2H; A4), 7.75
829/831 [MꢀPF6]+, 343/342 [Mꢀ2PF6]2+
.
[Ru(2)2][PF6]2:
A mixture of 2 (63.6 mg, 0.186 mmol), RuCl3·3H2O
(24.3 mg, 0.093 mmol) and N-ethylmorpholine (4 drops) in ethane-1,2-
diol (10 mL) was heated in a modified domestic microwave oven at
600 W for 4 min to give a clear red solution. The product was precipitat-
ed as a red solid by the addition of aqueous NH4PF6. Yield: 87.9 mg
(88.2%); 1H NMR (300 MHz, CD3CN): d=8.93 (s, 4H; B3), 8.63 (d, J=
7.7 Hz, 4H; A3), 7.72 (t, J=7.7 Hz, 4H; A4), 7.40 (d, J=5.1 Hz, 4H;
A6), 7.15 (t, J=7.0 Hz, 4H; A5), 7.12 (d, J=2.2 Hz, 4H; C2), 6.58 ppm
(t, 2H; C4); IR (solid): n˜ =3078 (w), 1604 (m), 1535 (w), 1461 (m), 1411
(m), 1323 (m), 1215 (m), 1149 (m), 1026 (m), 972 (m), 833 (s), 786 (s),
752 (m), 729 (m), 690 (m), 648 (m), 551 (m), 528 cmꢀ1 (m); MS (ESI):
m/z: 929 [MꢀPF6]+, 392 [Mꢀ2PF6]2+
.
[Ru(tpy)(2)][PF6]2: This complex was prepared by following the method
described for [Ru(tpy)(3)][PF6]2 using 2 (35.3 mg, 0.103 mmol) and [Ru-
(tpy)Cl3] (45.6 mg, 0.103 mmol). The product was precipitated as an
orange solid by the addition of aqueous NH4PF6; chromatographic purifi-
cation was not required. Yield: 61.2 mg (61.6%); 1H NMR (300 MHz,
CD3CN): d=8.88 (s, 2H; B3), 8.71 (d, J=8.1 Hz, 2H; E3), 8.58 (d, J=
7.7 Hz, 2H; A3), 8.46 (d, J=7.7 Hz, 2H; D3), 8.37 (t, J=8.1 Hz, 1H;
E4), 7.88 (t+t, J=1.8, 8.1 Hz, 4H; A4+D4), 7.37 (d, J=5.9 Hz, 2H;
A6), 7.30 (d, J=5.5 Hz, 2H; D6), 7.12 (m, 4H; A5+D5), 7.09 (d, J=
1.8 Hz, 2H; C2), 6.55 ppm (t, 1H; C4); IR (solid): n˜ =3325 (w), 1600
(m), 1535 (m), 1465 (m), 1415 (m), 1245 (m), 1161 (m), 1006 (w), 817 (s),
783 (s), 767 (s), 690 (m), 644 (m), 551 (m), 516 cmꢀ1 (s); MS (ESI): m/z:
821 [MꢀPF6]+, 338 [Mꢀ2PF6]2+
.
[Ru(2)(5)][PF6]2: This complex was obtained as a red solid by following
the method described for [Ru(tpy)(2)][PF6]2 using
2
(32.5 mg,
4032
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4024 – 4034

Ruthenium and Osmium Metallorods and Metallostars
FULL PAPER
0.095 mmol) and [Ru(5)Cl3] (50.2 mg, 0.096 mmol). Yield: 78.3 mg
(78.6%); 1H NMR (300 MHz, CD3CN): d=8.92 (s, 4H; B3+E3), 8.65
(d, J=7.7 Hz, 2H; D3), 8.62 (d, J=7.7 Hz, 2H; A3), 8.18 (dd, J=1.1,
3.7 Hz, 1H; F3), 7.94 (dt, J=1.5, 7.7 Hz, 2H; D4), 7.92 (dt, J=1.5,
7.7 Hz, 2H; A4), 7.83 (dd, J=1.1, 5.2 Hz, 1H; F5), 7.43 (d, J=5.5 Hz,
2H; D6), 7.42 (dd, 1H; F4), 7.41 (d, J=5.5 Hz, 2H; A6), 7.17 (m, 4H;
A5+D5), 7.13 (d, J=2.2 Hz, 2H; C2), 6.59 ppm (t, J=2.2 Hz, 1H; C4);
IR (solid): n˜ =1604 (m), 1539 (w), 1465 (m), 1411 (m), 1323 (w), 1288
(w), 1245 (w), 1149 (m), 1087 (w), 1006 (w), 817 (s), 786 (s), 752 (m), 713
(m), 651 (m), 621 (m), 551 cmꢀ1 (m); MS (ESI): m/z: 903 [MꢀPF6]+, 379
2H; D3), 8.55 (d, J=8.1 Hz, 4H; G3), 8.47 (d, J=2.2 Hz, 2H; C2), 8.47
(d, J=7.0 Hz, 4H; J3), 8.38 (t, J=8.1 Hz, 2H; K4), 8.20 (dd, J=1.1,
3.7 Hz, 1H; F3), 8.03 (t, 1H; C4), 7.95 (dt, J=1.5, 7.7 Hz, 2H; D4), 7.92
(dt, J=1.5, 8.1 Hz, 2H; A4), 7.85 (dt, J=1.5, 7.7 Hz, 4H; J4), 7.85 (dd,
J=1.1, 5.1 Hz, 1H; F5), 7.72 (dt, J=1.5, 7.7 Hz, 4H; G4), 7.55 (d, J=
5.5 Hz, 4H; J6), 7.48 (d, J=5.5 Hz, 2H; A6), 7.47 (d, J=5.5 Hz, 2H;
D6), 7.43 (dd, J=3.7, 5.1 Hz, 1H; F4), 7.33 (d, J=5.5 Hz, 4H; G6), 7.19
(m, 4H; A5+D5), 7.07 ppm (m, 8H; G5+J5); IR (solid): n˜ =1604 (w),
1585 (w), 1446 (w), 1404 (m), 1353 (m), 1288 (w), 1242 (w), 1195 (m),
1164 (w), 1026 (w), 999 (w), 829 (s), 786 (s), 767 (s), 651 (m), 555 cmꢀ1
[Mꢀ2PF6]2+
.
(m); MS (ESI): m/z: 1234 [Mꢀ2PF6]2+
,
774 [Mꢀ3PF6]3+
, 545
[Mꢀ4PF6]4+
.
[Os(2)2][PF6]2: This complex was obtained as a purple solid by following
the method described for [Ru(2)2][PF6]2 using 2 (58.7 mg, 0.172 mmol),
K2OsCl6 (27.8 mg, 0.089 mmol) and N-ethylmorpholine (9 drops) in
ethane-1,2-diol (10 mL). Yield: 85.9 mg (86.0%); 1H NMR (300 MHz,
CD3CN): d=8.95 (s, 4H; B3), 8.61 (d, J=8.1 Hz, 4H; A3), 7.77 (dt, J=
1.5, 7.7 Hz, 4H; A4), 7.26 (d, J=5.9 Hz, 4H; A6), 7.08 (ddd, 4H; A5),
7.10 (d, J=1.8 Hz, 4H; C2), 6.51 ppm (t, 2H; C4); IR (solid): n˜ =3052
(w), 1600 (m), 1523 (m), 1461 (m), 1419 (m), 1280 (m), 1245 (m), 1153
(w), 1130 (m), 1087 (m), 1026 (w), 1002 (m), 937 (m), 833 (s), 783 (s), 729
(m), 644 (m), 536 (m), 513 cmꢀ1 (m); MS (ESI): m/z: 1018 [MꢀPF6]+,
[Ru((6){Ru(tpy)}2)2][PF6]10
: A solution of [Ru(2)2][PF6]2 (12.0 mg,
0.011 mmol) and [Ru(tpy)(4)][PF6]2 (45.0 mg, 0.048 mmol) in MeCN
(10 mL) was treated with dry K2CO3 (200 mg, 1.35 mmol) and the mix-
ture was heated to reflux for 18 h. Complexes were precipitated by the
addition of aqueous NH4PF6. The pentanuclear product was found to
bind irreversibly to silica; attempted purification by column chromatog-
raphy on silica or Sephadex or by HPLC was unsuccessful. The precipi-
tated salts were dissolved in MeCN and very slowly reprecipitated by the
addition of further aqueous NH4PF6. The first orange-red precipitated
material was collected by filtration and was found to be pure [Ru((6){Ru-
(tpy)}2)2][PF6]10. Yield: 8.9 mg (18.0%); 1H NMR (500 MHz, CD3CN):
d=9.34 (s, 4H; B3), 8.80 (d, J=8.1 Hz, 4H; A3), 8.71 (d, J=8.1 Hz, 8H;
K3), 8.69 (s, 8H; H3), 8.56 (d, J=8.1 Hz, 8H; G3), 8.48 (d, J=2.2 Hz,
4H; C2), 8.45 (d, J=7.0 Hz, 8H; J3), 8.40 (t, 4H; K4), 8.05 (t, 2H; C4),
7.93 (dt, J=1.5, 8.1 Hz, 4H; A4), 7.83 (dt, J=1.5, 7.7 Hz, 8H; J4), 7.70
(dt, J=1.5, 7.7 Hz, 8H; G4), 7.56 (d, J=5.5 Hz, 8H; J6), 7.50 (d, J=
5.5 Hz, 4H; A6), 7.31 (d, J=5.5 Hz, 8H; G6), 7.09 (dt, 4H; A5), 7.07
(ddd, 8H; G5), 7.03 ppm (ddd, 8H; J6); IR (solid): n˜ =1604 (w), 1542
(w), 1454 (w), 1404 (m), 1357 (w), 1288 (w), 1199 (m), 1154 (w), 1002
(w), 825 (s), 786 (s), 763 (s), 651 (m), 551 cmꢀ1 (s); MS (ESI): m/z: 2104
[Mꢀ2PF6]2+, 1353 [Mꢀ3PF6]3+, 979 [Mꢀ4PF6]4+, 754 [Mꢀ5PF6]5+, 604
437 [Mꢀ2PF6]2+
.
[Os(2)(5)][PF6]2: This complex was prepared by following the method
described for [Ru(2)(5)][PF6]2 using
2 (30.0 mg, 0.088 mmol) and
[Os(5)Cl3] (54.0 mg, 0.088 mmol) in ethane-1,2-diol (10 mL). Purification
of the crude product by column chromatography (SiO2; MeCN/saturated
aqueous KNO3/H2O, 7:2:2) gave [Os(2)(5)][PF6]2 as a brown solid. Yield:
77.2 mg (77.3%); 1H NMR (300 MHz, CD3CN): d=8.95 (s, 4H; B3+
D3), 8.63 (d, J=7.4 Hz, 2H; D3), 8.61 (d, J=7.4 Hz, 2H; A3), 8.10 (dd,
J=1.1, 3.7 Hz, 1H; F3), 7.80 (dt, J=1.5, 7.7 Hz, 2H; D4), 7.79 (dt, J=
1.5, 8.08 Hz, 2H; A4), 7.72 (dd, J=1.1, 5.2 Hz, 1H; F5), 7.42 (dd, J=3.7,
5.1 Hz, 1H; F4), 7.30 (d, J=5.1 Hz, 2H; A6), 7.29 (d, J=5.1 Hz, 2H;
D6), 7.10 (m, 4H; A5+D5), 7.10 (d, J=2.2 Hz, 2H; C2), 6.52 ppm (t,
1H; C4); IR (solid): n˜ =3066 (w), 1604 (m), 1581 (m), 1523 (m), 1465
(m), 1427 (m), 1396 (m), 1361 (m), 1334 (m), 1284 (m), 1245 (m), 1153
(w), 1076 (w), 1026 (w), 825 (s), 783 (s), 752 (m), 713 (m), 651 (m), 621
[Mꢀ6PF6]6+
.
[(5)Os(6){Ru(tpy)}2][PF6]6: This complex was obtained as an orange-
brown solid by following the method described for the analogous com-
plex [(5)Ru(6){Ru(tpy)}2][PF6]6 but using [Os(5)(2)][PF6]2 (35.0 mg,
0.030 mmol) and [Ru(tpy)(4)][PF6]2 (58.1 mg, 0.062 mmol). Yield:
32.7 mg (38.3%); 1H NMR (300 MHz, CD3CN): d=9.30 (s, 2H; B3),
8.96 (s, 2H; E3), 8.75 (d, J=8.1 Hz, 2H; A3), 8.72 (d, J=8.1 Hz, 4H;
K3), 8.67 (s, 4H; H3), 8.64 (d, J=8.1 Hz, 2H; D3), 8.55 (d, J=8.1 Hz,
4H; G3), 8.46 (d, J=8.1 Hz, 4H; J3), 8.44 (d, J=2.2 Hz, 2H; C2), 8.38
(4, J=8.1 Hz, 2H; K4), 8.10 (dd, J=1.1, 3.7 Hz, 1H; F3), 7.95 (t, 1H;
C4), 7.84 (dt, J=1.5, 7.7 Hz, 4H; J4), 7.80 (dt, J=1.5, 7.7 Hz, 2H; D4),
7.78 (dt, J=1.5, 8.1 Hz, 2H; A4), 7.73 (dd, J=1.1, 5.1 Hz, 1H; F5), 7.72
(dt, J=1.5, 7.7 Hz, 4H; G4), 7.56 (d, J=5.5 Hz, 4H; J6), 7.42 (dd, J=3.7,
5.1 Hz, 1H; F4), 7.34 (d, J=5.5 Hz, 2H; A6), 7.33 (d, J=5.5 Hz, 2H;
D6), 7.32 (d, J=5.5 Hz, 4H; G6), 7.12 (dt, 2H; A5), 7.10 (dt, J=1.1,
7.3 Hz, 2H; D5), 7.08 (dt, J=1.1, 7.3 Hz, 4H; G5), 7.05 ppm (dt, J=1.1,
7.3 Hz, 4H; J5); IR (solid): n˜ =1604 (w), 1434 (w), 1396 (m), 1357 (s),
1284 (m), 1245 (m), 1199 (m), 1164 (w), 1126 (w), 1029 (w), 999 (w), 825
(s), 786 (s), 763 (s), 651 (m), 551 cmꢀ1 (s); MS (ESI): m/z: 1279
(m), 551 cmꢀ1 (m); MS (ESI): m/z: 992 [MꢀPF6]+, 424 [Mꢀ2PF6]2+
.
[(tpy)Ru(6){Ru(tpy)}2][PF6]6: [Ru(tpy)(2)][PF6]2 (18.0 mg, 0.019 mmol),
dry K2CO3 (200 mg, 1.45 mmol) and [Ru(tpy)(4)][PF6]2 (39.0 mg,
0.044 mmol) were dissolved in dry acetonitrile (10 mL) and the mixture
was heated to reflux for 4 h. The product was precipitated by the addition
of aqueous NH4PF6 and was purified by column chromatography (SiO2;
MeCN/saturated aqueous KNO3/H2O, 7:1:0.5). The first orange fraction
to be eluted consisted of [Ru(tpy)(4)][PF6]2, and this was followed by a
slower moving red compound that was identified as [(tpy)Ru(6){Ru-
1
(tpy)}2][PF6]6. Yield: 17.5 mg (34.4%); H NMR (300 MHz, CD3CN): d=
9.28 (s, 2H; B3), 8.76 (d, J=8.1 Hz, 4H; A3+E3), 8.72 (d, J=8.1 Hz,
4H; K3), 8.67 (s, 4H; H3), 8.55 (d, J=8.1 Hz, 4H; G3), 8.50 (d, J=
8.1 Hz, 4H; D3), 8.46 (d, J=2.2 Hz, 2H; C2), 8.45 (d, J=7.0 Hz, 4H;
J3), 8.42 (t, J=8.1 Hz, 1H; E4), 8.37 (t, J=8.1 Hz, 2H; K4), 8.02 (t, 1H;
C4), 7.92 (dt, J=1.5, 7.7 Hz, 2H; D4), 7.89 (dt, J=1.5, 8.1 Hz, 2H; A4),
7.83 (dt, J=1.5, 7.7 Hz, 4H; J4), 7.71 (dt, J=1.5, 7.7 Hz, 4H; G4), 7.55
(d, J=5.5 Hz, 4H; J6), 7.44 (d, J=5.5 Hz, 2H; D6), 7.37 (d, J=5.5 Hz,
2H; A6), 7.31 (d, J=5.5 Hz, 4H; G6), 7.17 (dt, 2H; A5), 7.15 (dt, 2H;
D5), 7.08 (ddd, 4H; G5), 7.04 ppm (ddd, 4H; J5); IR (solid): n˜ =1604
(w), 1466 (w), 1404 (m), 1353 (m), 1288 (w), 1245 (w), 1199 (m), 1164
(w), 999 (w), 821 (s), 786 (s), 763 (s), 690 (m), 551 cmꢀ1 (m); MS (ESI):
[Mꢀ2PF6]2+, 804 [Mꢀ3PF6]3+, 567 [Mꢀ4PF6]4+
.
[(5)Os(6){Ru(5)}2][PF6]6: This complex was obtained as an orange-brown
solid by following the method described for [(5)Ru(6){Ru(tpy)}2][PF6]6
using [Os(5)(2)][PF6]2 (15.1 mg, 0.013 mmol) and [Ru(5)(4)][PF6]2
(30.5 mg, 0.031 mmol). Yield: 12.7 mg (32.4%); 1H NMR (500 MHz,
CD3CN): d=9.35 (s, 2H; B3), 8.97 (s, 2H; E3), 8.90 (s, 4H; K3), 8.78 (d,
J=8.1 Hz, 4H; A3), 8.72 (s, 4H; H3), 8.64 (d, J=8.8 Hz, 2H; D3), 8.62
(d, J=8.4 Hz, 4H; J3), 8.59 (d, J=8.1 Hz, 4H; G3), 8.46 (d, J=1.8 Hz,
2H; C2), 8.16 (dd, J=1.1, 3.7 Hz, 2H; L3), 8.10 (dd, J=1.1, 3.7 Hz, 1H;
F3), 7.99 (t, 1H; C4), 7.85 (dt, J=1.5, 8.1 Hz, 4H; J4), 7.81 (dd, J=1.1,
5.1 Hz, 2H; L5), 7.80 (dt, J=1.5, 7.7 Hz, 2H; D4), 7.77 (dt, J=1.5,
8.08 Hz, 2H; A4), 7.73 (dd, J=1.1, 5.1 Hz, 1H; F5), 7.71 (dt, J=1.5,
8.4 Hz, 4H; G4), 7.59 (d, J=5.1 Hz, 4H; J6), 7.42 (1H; F4), 7.41 (2H;
L4), 7.40 (d, J=5.1 Hz, 4H; G6), 7.34 (d, J=5.5 Hz, 2H; D6), 7.33 (d,
J=5.5 Hz, 2H; A6), 7.11 (ddd, 2H; A5), 7.09 (ddd, 2H; D5), 7.08 (dt,
4H; G5), 7.05 ppm (dt, 4H; J5); IR (solid): n˜ =1612 (w), 1454 (w), 1396
m/z: 1194 [Mꢀ2PF6]2+, 748 [Mꢀ3PF6]3+, 525 [Mꢀ4PF6]4+
.
[(5)Ru(6){Ru(tpy)}2][PF6]6: The method was as described for
[(tpy)Ru(6){Ru(tpy)}2][PF6]6
using
[Ru(5)(2)][PF6]2
(38.6 mg,
0.037 mmol) and [Ru(tpy)(4)][PF6]2 (69.8 mg, 0.078 mmol) in MeCN
(10 mL). Chromatographic separation (SiO2; MeCN/saturated aqueous
KNO3/H2O, 7:1:0.5) resulted in orange [Ru(tpy)(4)][PF6]2 as the first frac-
tion, red [Ru(5)(2)][PF6]2 as the second band, and a very slow moving
red band from which a red solid was isolated that was characterised as
[(5)Ru(6){Ru(tpy)}2][PF6]6. Yield: 52.0 mg (50.8%); 1H NMR (300 MHz,
CD3CN): d=9.30 (s, 2H; B3), 8.95 (s, 2H; E3), 8.77 (d, J=8.1 Hz, 2H;
A3), 8.73 (d, J=8.1 Hz, 4H; K3), 8.68 (s, 4H; H3), 8.67 (d, J=8.1 Hz,
Chem. Eur. J. 2005, 11, 4024 – 4034
www.chemeurj.org
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4033

E. C. Constable, F. Barigelletti et al.
(m), 1357 (m), 1288 (w), 1199 (m), 1154 (w), 1026 (w), 1002 (w), 825 (s),
Harverson, C. E. Housecroft, J. Chem. Soc. Dalton Trans. 1999,
3693–3700; f) E. C. Constable, P. Harverson, M. Oberholzer, Chem.
Commun. 1996, 1821–1822.
786 (s), 752 (m), 725 (m), 651 (m), 551 (s), 528 cmꢀ1 (m); MS (ESI): m/z:
1361 [Mꢀ2PF6]2+, 859 [Mꢀ3PF6]3+, 608 [Mꢀ4PF6]4+
.
[8] a) A. Juris, L. Prodi, New J. Chem. 2001, 25, 1132–1135; b) A.
Boerje, O. Koethe, A. Juris, New J. Chem. 2001, 25, 191–193; c) A.
Boerje, O. Koethe, A. Juris, J. Chem. Soc. Dalton Trans. 2002, 843–
848; d) D. Tzalis, Y. Tor, Chem. Commun. 1996, 1043–1045.
[9] E. C. Constable, R. W. Handel, C. E. Housecroft, A. F. Morales, L.
Flamigni, F. Barigelletti, Dalton Trans. 2003, 1220–1222.
[10] V. Balzani, S. Campagna, G. Denti, A. Juris, S. Serroni, M. Venturi,
Acc. Chem. Res. 1998, 31, 26–34.
[11] A. Bçrje, O. Kothe, A. Juris, J. Chem. Soc. Dalton Trans. 2002, 843–
848.
[12] F. Barigelletti, L. Flamigni, Chem. Soc. Rev. 2000, 29, 1–12.
[13] E. C. Constable, C. E. Housecroft, M. Cattalini, D. Phillips, New J.
Chem. 1998, 22, 193–200.
[Os((6){Ru(5)}2)2][PF6]10
: This complex was obtained as an orange-
brown solid by following the method described for [Ru((6){Ru(tpy)}2)2]-
[PF6]10 using [Os(2)2][PF6]2 (14.0 mg, 0.012 mmol) and [Ru(5)(4)][PF6]2
(50.2 mg, 0.052 mmol). Yield: 12.8 mg (21.6%); 1H NMR (500 MHz,
CD3CN): d=9.43 (s, 4H; B3), 8.90 (s, 8H; K3), 8.82 (d, J=8.1 Hz, 8H;
A3), 8.74 (s, 8H; H3), 8.61 (d, J=8.1 Hz, 16H; J3+G3), 8.49 (d, J=
1.8 Hz, 4H; C2), 8.16 (d, J=3.7 Hz, 4H; L3), 8.03 (t, 2H; C4), 7.85 (dt,
J=1.5, 8.1 Hz, 8H; J4), 7.81 (dd, J=1.1, 5.1 Hz, 4H; L5), 7.77 (t, J=
7.7 Hz, 4H; A4), 7.69 (dt, J=1.1, 7.7 Hz, 8H; G4), 7.62 (d, J=5.5 Hz,
8H; J6), 7.40 (dd, J=3.7, 5.1 Hz, 4H; L4), 7.40 (d, J=5.5 Hz, 8H; G6),
7.38 (d, J=5.5 Hz, 2H; A6), 7.12 (t, 4H; A5), 7.07 (t, 4H; G5), 7.03 ppm
(dt, 4H; J5); IR (solid): n˜ =1608 (w), 1585 (w), 1465 (w), 1400 (m), 1353
(m), 1284 (w), 1199 (m), 1152 (w), 1087 (m), 1026 (m), 1002 (w), 829 (s),
786 (s), 752 (m), 729 (m), 651 (m), 555 (s), 520 cmꢀ1 (m); MS (ESI): m/z:
[14] F. Krçhnke, Synthesis 1976, 1–24.
1493 [Mꢀ3PF6]3+, 1083 [Mꢀ4PF6]4+, 837 [Mꢀ5PF6]5+, 674 [Mꢀ6PF6]6+
.
[15] W. Spahni, G. Calzaferri, Helv. Chim. Acta 1984, 67, 450–454.
[16] S. Encinas, L. Flamigni, F. Barigelletti, E. C. Constable, C. E. House-
croft, E. R. Schofield, E. Figgemeier, D. Fenske, M. Neuburger, J. G.
Vos, M. Zehnder, Chem. Eur. J. 2002, 8, 137–150.
Acknowledgements
[17] a) E. C. Constable, D. Morris, S. Carr, New J. Chem. 1998, 22, 287–
294; b) C. O. Dietrich-Buchecker, J. P. Sauvage, J. P. Kintzinger, P.
Maltese, C. Pascard, J. Guilhem, New J. Chem. 1992, 16, 931–942;
c) C. Dietrich-Buchecker, C. Hemmert, J. P. Sauvage, New J. Chem.
1990, 14, 603–605.
We thank the Schweizerischer Nationalfonds zur Fçrderung der Wissen-
schaftlichen Forschung, the University of Basel and the FIRB project
(No. RBNE019H9K, “Molecular Manipulation for Nanometric Devices”)
from MIUR, Italy, for financial support.
[18] M. Grꢂtzel, Nature 2001, 414, 338–344.
[19] a) J. Hjelm, E. C. Constable, E. Figgemeier, A. Hagfeldt, R. Handel,
C. E. Housecroft, E. Mukhtar, E. Schofield, Chem. Commun. 2002,
284–285; b) E. C. Constable, C. E. Housecroft, E. R. Schofield, S.
Encinas, N. Armaroli, F. Barigelletti, L. Flamigni, E. Figgemeier,
J. G. Vos, Chem. Commun. 1999, 869–870.
[20] a) E. C. Constable, Inorg. Chim. Acta 1986, 117, L33–L34; b) E. C.
Constable, T. A. Leese, Inorg. Chim. Acta 1988, 146, 55–58.
[21] E. C. Constable, Metals and Ligand Reactivity, VCH, Weinheim,
1995.
[22] A. S. Abd-El-Aziz, C. R. de Denus, M. J. Zaworotko, L. R. MacGil-
livray, J. Chem. Soc. Dalton Trans. 1995, 3375–3393.
[23] J. P. Sauvage, J. P. Collin, J. C. Chambron, S. Guillerez, C. Coudret,
V. Balzani, F. Barigelletti, L. De Cola, L. Flamigni, Chem. Rev. 1994,
94, 993–1019.
[24] According to a different approach, one might compare the behav-
iour of equimolar solutions; see, for instance, Serin et al. in ref. [25]
below. In this way, an apparent gain in luminescence intensity for
the dendritic case may be registered, as compared to what happens
in the component model, simply because of the increased absorb-
ance within the dendrimer, that is, because of absorption of a great-
er number of photons. However, given that the luminescence quan-
tum yield refers to photons emitted over photons absorbed (per unit
time), it seems more appropriate to deal with isoabsorbing condi-
tions.
[1] a) E. C. Constable, in Comprehensive Coordination Chemistry II,
Vol. 7 (Eds.: M. Fujita, A. Powell, C. Creutz; Series Eds.: J. A.
McCleverty, T. J. Meyer), Elsevier, Oxford, 2003, pp. 263–302;
b) C. N. Moorefield, E. He, G. R. Newkome, Chem. Rev. 1999, 99,
1689–1746.
[2] a) A. Juris, M. Venturi, P. Ceroni, V. Balzani, S. Campagna, S. Serro-
ni, Collect. Czech. Chem. Commun. 2001, 66, 1–32; b) J. F. Stoddart,
T. Welton, Polyhedron 1999, 18, 3575–3591; c) S. Serroni, S. Campa-
gna, G. Denti, A. Juris, M. Venturi, V. Balzani, in Advances in Den-
dritic Macromolecules, Vol. 3 (Ed.: G. R. Newkome), Jai Press,
Greenwich, 1996, pp. 193–228; d) M. Venturi, S. Serroni, A. Juris, S.
Campagna, V. Balzani, in Dendrimers (Ed.: F. Vçgtle), Springer-
Verlag, Berlin, 1998, pp. 193–228; e) A. Hearshow, A. T. Hutton,
J. R. Moss, K. J. Naidoo, in Advances in Dendritic Macromolecules,
Vol. 4 (Ed.: G. R. Newkome), Jai Press, Stamford, 1999, pp. 1–60.
[3] E. C. Constable, Chem. Commun. 1997, 1073–1080.
[4] a) C. Amatore, Y. Bouret, E. Maisonhaute, J. I. Goldsmith, H. D.
Abruna, Chem. Eur. J. 2001, 7, 2206–2226; b) H. F. Chow, I. Y. K.
Chan, P. S. Fung, T. K. K. Mong, M. F. Nongrum, Tetrahedron 2001,
57, 1565–1572; c) H. F. Chow, I. Y. K. Chan, D. T. W. Chan,
R. W. M. Kwok, Chem. Eur. J. 1996, 2, 1085–1091.
[5] D. Armspach, M. Cattalini, E. C. Constable, C. E. Housecroft, D.
Phillips, Chem. Commun. 1996, 1823–1824.
[25] J. M. Serin, D. W. Brousmiche, J. M. J. Frꢅchet, J. Am. Chem. Soc.
2002, 124, 11848–11849.
[6] a) V. Balzani, P. Ceroni, A. Juris, M. Venturi, S. Campagna, F. Pun-
toriero, S. Serroni, Coord. Chem. Rev. 2001, 219, 545–572; b) S.
Campagna, C. Di Pietro, F. Loiseau, B. Maubert, N. McClenaghan,
R. Passalacqua, F. Puntoriero, V. Ricevuto, S. Serroni, Coord. Chem.
Rev. 2002, 229, 67–74.
[7] a) E. C. Constable, M. Cattalini, O. Eich, C. E. Housecroft, L. A.
Johnston, I. Poleschak, Polym. Mater. Sci. Eng. 1999, 80, 268;
b) E. C. Constable, C. E. Housecroft, I. Poleschak, Inorg. Chem.
Commun. 1999, 2, 565–568; c) E. C. Constable, C. E. Housecroft,
M. Neuburger, I. Poleschak, M. Zehnder, Polyhedron 2003, 22, 93–
108; d) E. C. Constable, R. Frantz, C. E. Housecroft, J. Lacour, A.
Mahmood, Inorg. Chem. 2004, 43, 4817–4819; e) E. C. Constable, P.
[26] a) T. Fçrster, Discuss. Faraday Soc. 1959, 27, 7; b) J. R. Lakowicz,
Principles of Fluorescence Spectroscopy, Plenum, New York, 1999.
[27] L. Hammarstrçm, F. Barigelletti, L. Flamigni, N. Armaroli, A. Sour,
J. P. Collin, J. P. Sauvage, J. Am. Chem. Soc. 1996, 118, 11972–
11973.
[28] K. Nakamaru, Bull. Chem. Soc. Jpn. 1982, 55, 2967.
[29] J. N. Demas, G. A. Crosby, J. Phys. Chem. 1971, 75, 991–1024.
Received: February 2, 2005
Published online: April 28, 2005
4034
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 4024 – 4034