Multicomponent supramolecular devices: Synthesis, optical, and electronic properties of bridged bis-dirhodium and -diruthenium complexes

Article abstract of DOI:10.1002/ejic.200600466

Four ruthenium- and rhodium-based metal-metal-bonded multicomponent systems have been synthesized, and their absorption, redox, spectroelectrochemical and structural properties have been studied. The absorption spectra of the four bis-dimetallic compounds

Full text of DOI:10.1002/ejic.200600466

FULL PAPER
DOI: 10.1002/ejic.200600466
Multicomponent Supramolecular Devices: Synthesis, Optical, and Electronic
Properties of Bridged Bis-dirhodium and -diruthenium Complexes
Anne Petitjean,[a][‡] Fausto Puntoriero,[b] Sebastiano Campagna,*[b] Alberto Juris,[c] and
Jean-Marie Lehn*[a]
Dedicated to Professor Vincenzo Balzani
Keywords: N ligands / Electronic coupling / Electrochemistry / Metal–metal interactions / Optical properties
Four ruthenium- and rhodium-based metal–metal-bonded
multicomponent systems have been synthesized, and their
absorption, redox, spectroelectrochemical and structural
properties have been studied. The absorption spectra of the
four bis-dimetallic compounds M2LM2, where L is a bridging
ligand and M is rhodium or ruthenium, exhibit very strong
bands in the UV, visible and, for the diruthenium species,
near-IR region. The low-energy absorption bands are as-
signed to charge-transfer transitions involving a metal–metal
bonding orbital as the donor and an orbital centered on the
bis-tetradentate aromatic ligands as the acceptor (metal–
metal to ligand charge transfer, M2LCT). Each compound ex-
hibits reversible bridging-ligand-centered reductions at mild
potentials and several reversible oxidation processes. The
oxidation signals of the two equivalent dimetallic centers of
each bis-dimetallic compound are split, with the splitting − a
measure of the electronic coupling − depending on both the
metal and bridging ligand. The mixed-valence species of the
dirhodium species was investigated, and the electronic coup-
ling matrix element calculated from the experimental inter-
valence band parameters for one of them (86 cm–1) indicates
a significant inter-component electronic interaction which
compares well with good electron conducting anionic brid-
ges such as cyanides. Although none of these compounds
is luminescent, the M2LCT excited state of one of the bis-
dirhodium complexes is relatively long-lived (about 6 µs) in
degassed acetonitrile at room temperature. The results pre-
sented here are promising for the development of linear poly-
dimetallic complexes built on longer naphthyridine-based
strands, with significant long-range electronic coupling and
molecular-wire-like behavior.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
rich redox properties, transition metal ions connected by a
bridging spacer have been the focus of much work in this
field, although recent examples have also proved the tre-
mendous potential of organic redox-active species con-
Introduction
Since the discovery of mixed-valence complexes by
Creutz and Taube[1] over 30 years ago, architectures includ-
ing several redox centers of the same[2] or different[3] nature
and capable of exhibiting inter-component electronic inter-
action have been the center of much attention. Due to their
nected through bridging transition metal complexes.[4]
A
variety of systems connecting mononuclear metal ions have
been explored, where the spacer is itself a ligand[1,5] or
where spacer and ligands are decoupled and involve single
bonds[6] and multiple bonds (alkynyl for 1D,[7] 2D,[7b,7c]
and 3D[7c] architectures; alkenyl,[5b] cumulenes,[7a,8]
aryl[5b,7b,7d,9a,9b]). These systems may incorporate other
inert[10a] or labile[7b,10b] complexes and have the electronic
communication between the two sites modulated by pH,[11]
light,[12a,12b] or auxiliary ligands.[13] The field has more re-
cently been extended to metal–metal-bonded redox units
because they themselves can include multiple bonds,[14] sim-
ilar to carbon-based units but energetically more accessible,
and may therefore play the role of a conductor. The dissym-
metry around the metal sites allows for testing of the com-
munication along the metal–metal bond through axial li-
gands and spacers similar to their mononuclear counter-
[a] ISIS, Université Louis Pasteur, CNRS UMR 7006,
B. P. 70028, 67083 Strasbourg, France
Fax: +33-3-90245140
E-mail: lehn@isis.u-strasbg.fr
[b] Dipartimento di Chimica Inorganica, Chimica Analitica e Chi-
mica Fisica, Università di Messina,
Via Sperone 31, 96166 Villaggio S. Agata, Messina, Italy
Fax: +39-090-393-756
E-mail: campagna@unime.it
[c] Dipartimento di Chimica “G. Ciamician”, Università di Bolo-
gna,
Via F. Selmi 2, 40126 Bologna, Italy
Fax: +39-051-209-9456
E-mail: alberto.juris@unibo.it
[‡] Present address: Department of Chemistry, Queen’s University,
90 Bader Lane, Kingston ON, K7L 3N6, Canada
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
3878
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
1
Figure 1. Formulae and abbreviation of the complexes. Numbering is related to the discussion of the H NMR spectroscopic data.
parts (e.g., triple bonds[15] and double bonds[16]), but also ether[20] and submitted to acidic hydrolysis to yield the cor-
successfully along the equatorial direction.[17]
responding diketones 5Ph and 5Ant, ready for Friedländer
We report here the synthesis and characterization, ab- condensation. Once again, different routes were used for
sorption spectra, redox behavior, and spectroelectrochemis- the butyl- and anthracenyl-substituted ligands. The butyl-
try of four bi-component systems in which each “compo- substituted ligand was obtained by two consecutive Fried-
nent” or subunit is made of a metal–metal-bonded dimet- länder condensations: a base-catalyzed condensation with
allic site (Figure 1). The excited-state properties of one such 4-aminopyrimidine-5-carboxaldehyde[21] yielded pyrimi-
species are also investigated. As we will see later, each di- dine-ended intermediate 6Ph, which was then converted
metallic site can indeed be viewed as a single component of into the bis(aminocarbaldehyde) 7Ph under acidic condi-
a “dinuclear” system from a redox viewpoint.
tions. The latter was then condensed with commercially
available 2-acetylpyridine to yield the final ligand LPh.
Bis(ketone) 5Ant, on the other hand, was directly con-
densed with 6-amino-2,2Ј-bipyridinyl-5-carbaldehyde (9) to
yield ligand LAnt. Compound 9 was also synthesized by
Friedländer condensation of 2-acetylpyridine and 4-amino-
pyrimidine-5-carboxaldehyde (8) followed by acidic hydro-
lysis. LAnt displays the fluorescence emission typical of
anthracene derivatives.
Results and Discussion
Synthesis
Ligands
The naphthyridine-based ligands 2-(4-tert-butylphenyl)-
4,6-bis(7-pyridin-2-yl[1,8]naphthyridin-2-yl)pyrimidine
(LPh) and 2-anthracen-9-yl-4,6-bis(7-pyridin-2-yl[1,8]naph-
thyridin-2-yl)pyrimidine (LAnt) were obtained by a Fried-
länder condensation methodology applied to a pyrimidine
scaffold, in a very similar fashion as previously reported[18]
(Schemes 1 and 2). Two synthetic routes were used to access
the central pyrimidine unit. In the case of the 4-tert-bu-
tylphenyl substituent leading to LPh (Scheme 1), the pyr-
imidine ring was built by condensation of the correspond-
ing amidine 2, resulting from ammonia addition to the cor-
responding nitrile 1, with diethyl malonate. The obtained
dihydroxypyrimidine 3 was then chlorinated with a combi-
nation of phosphoryl chloride and N,N-dimethylaniline to
yield the dichloropyrimidine 4Ph. The 9-anthracenyl-substi-
tuted dichloropyrimidine 4Ant, on the other hand, was pre-
pared directly from unsubstituted 4,6-dichloropyrimidine
by deprotonation at the 2-position, reaction with 9-bromo-
anthracene, and oxidation with DDQ, as published earlier
Metal Complexes
Both dirhodium complexes were obtained by treatment
of a suspension of the ligand with commercially available
dirhodium tetraacetate in methanol in the presence of two
equivalents of protons (1 equiv. per acetate bridge to be
substituted).[22a] Heating a suspension of the ligand with
the mixed-valence complex [Ru2(OAc)4Cl][23] in methanol
yielded the desired diruthenium(II) complexes without ad-
ditional reductant.[22b] All complexes were isolated as their
hexafluorophosphate salts.
Characterization
1H NMR Spectroscopy
The diamagnetic dirhodium(II) complexes show a simple
(Scheme 2).[19] The dichloropyrimidines 4Ph and 4Ant were set of signals, thus revealing that the C2 symmetry of the
then coupled with 1-(tri-n-butylstannyl)-1-vinyl ethyl ligands is maintained in the complexes and therefore that
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3879

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
1
Scheme 1. Synthesis of LPh (numbering relates to the discussion of the H NMR spectroscopic data).
1
Scheme 2. Synthesis of LAnt (numbering relates to the discussion of the H NMR spectroscopic data).
both dinuclear binding sites are occupied (in agreement due to the ring effect provided by the anthracene moiety,
with elemental analysis and mass spectrometry measure- thua confirming the involvement of the pyrimidine nitrogen
ments). The two sets of signals from the acetate ligands con- atoms in the coordination of the Rh2 units and the rotation
firm the dissymmetry introduced by a single L ligand (two of the C–C bonds between the heteroaryl units compared
different signals for the 12 equatorial and the six axial CH3 to the initial ligands. Although the spectra of free ligands
acetate protons). The six axial acetate protons are shielded and metal complexes were recorded in different solvents
3880
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
Table 1. Chemical shifts [ppm] of the protons in ligands LPh and LAnt (in CDCl3) and their dirhodium complexes (in CD3CN).
FULL PAPER
Compound
LPh
[(Rh2)2LPh] 9.71
H1
H3 H3Ј
H3py
H4 H4Ј
H4py H5py H6py
Ha
Hb
tBu
10.06
8.81–9.0
8.8–9.0
8.42 8.47
7.97
8.58
7.43
8.29
8.8
9.67
8.73
9.24
7.64
7.75
1.44
1.49
Compound
LAnt
[(Rh2)2LAnt] 10.1
H1
H3 H3Ј
H3py
H4 H4Ј
H4py H5py H6py
H5
H6 H7
H8
H9
10.3
8.7–8.85
9.47
9.03
8.8–9.1
8.3–8.4
8.8–9.1
7.99
8.52
7.4–7.6 8.7–8.9
8.2–8.3 9.50
7.88
7.92
7.4–7.6
7.31, 7.47
8.14
8.2–8.3 8.8–9.1
8.67
and therefore the chemical shifts cannot be accurately com- naphthyridines [Rh(1)–N(2) 2.00 and Rh(2)–N(3) 1.98 Å].
pared (cf. tBu signal as a probe of solvent effect), the intro- As anticipated by the difference in basicity, the strong coor-
duction of the Rh2 units has a significant deshielding effect dinative contribution of the pyridine [Rh(1)–N(1) 2.16 Å]
on the protons para to the nitrogen atoms of the bound stands in contrast to the weaker binding of the pyrimidine
heterocycles (ca. +0.5 ppm for H4, H4Јand H4py; Table 1) ring [Rh(2)–N(4) 2.31 Å]. Overall, the quasi-planar, cres-
and on the pyridine protons. The substituents on the pyr- cent-shaped ligand[25] holds the two bis-dirhodium metallic
imidine ring also undergo a downfield shift due to the prox- units in a linear alignment that is bridged by the central
imity of the acetate ligands (ca. +0.5 ppm for Ha in LPh pyrimidine and slightly bent by the coordination-induced
and ca. +0.2 ppm for H9 in LAnt).
pinching of the subunits.
The diruthenium(II) complexes are paramagnetic, as ex-
pected for the distribution of the 12 d-electrons of the two
ruthenium(II) centers in a σ2π4δ2δ*2π*2 configuration. The
1H NMR spectra nevertheless show well-defined peaks
ranging from δ = –50 to 80 ppm (Figure 2) whose number
is in accordance with the expected symmetry.
IR Spectra
The IR spectra of all complexes show many common
features (number and location of bands), which is further
evidence for their similar functional and structural environ-
ments as well as identical symmetry.
Absorption Spectra
All complexes are deeply colored. The optical properties
are represented in Figure 4 and reported in Table 2, to-
gether with those of the simple [M2bpnp]+ dimetallic com-
plexes reported earlier by other groups[22a,22b] (M = RuII or
RhII; these complexes were synthesized again for compari-
son purposes). The absorption spectra are very rich, as ex-
pected because of the presence of several chromophoric
units. The spectra of all the complexes exhibit an intense
absorption feature between 320 and 400 nm. This feature
should therefore receive contribution from spin-allowed π–
πء transitions involving the common L bridging ligand. In
[(Rh2)2LAnt] and [(Ru2)2LAnt], the spin-allowed, anthra-
Figure 2. 1H NMR spectra of the paramagnetic complexes
[(Ru2)2LAnt]2+ (top) and [(Ru2)2LPh]2+ (bottom) (CD3CN,
200 MHz, 25 °C).
X-ray Structure Determination
1
cene-centered La transitions are expected within the same
Single-crystal X-ray analysis of [(Rh2)2LPh][24] confirmed
the anticipated structure of the bis-dirhodium complex
(Figure 3). All nitrogen sites are bound to the rhodium
centers. The distance between the two rhodium centers
(2.40 Å) is the same as that found in the parent [Rh2bpnp]+
complex [Rh2 is Rh2(OAc)3 and bpnp is bis(pyridyl)naph-
thyridine],[22a] as are the distances involving the coordinated
wavelength range,[26c,27,28] and their contribution is clear
when looking at the vibrational progression appearing in
the spectra of both species. For these latter species, the very
intense absorption band around 254 nm can be straightfor-
1
wardly assigned to the Ba transition of the anthryl sub-
units.[26c,27,28]
As far as the visible absorption is concerned, the lowest-
energy band of the rhodium species is attributed to a
charge-transfer transition from each dirhodium subunit
(most likely from a πء orbital extending over the two metal
centers) to the bridging ligand. Such a transition is a form
of metal–metal to ligand charge transfer (M2LCT) transi-
tion because of the strong interaction between the metal–
metal-bonded units of each component. The red shift of the
lowest-energy absorption band in the present compounds
{λmax = 622 and 634 nm for [(Rh2)2LPh] and [(Rh2)2LAnt],
respectively} compared to that of the parent species
[Rh2bpnp] (λmax = 578 nm) supports the CT assignment,
since in the dicomponent species the LUMO of the polypyr-
Figure 3. Solid-state structure of [(Rh2)2LPh][24] showing the align-
ment of the two bis-dirhodium units bridged by the central pyrid-
imine (hydrogen atoms, solvent, and counter anions have been
omitted for clarity).
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3881

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
idine-type bridging ligand (i.e., the acceptor orbital of the
CT transition) is lower in energy than the LUMO of bpnp
(see below).
For the ruthenium compounds, the absorption bands,
which have a similar origin to those of the rhodium ana-
logues, are displaced to lower energies than the rhodium
species, with the lowest-energy band in the near-IR region
(Figure 4). This is also in agreement with a M2LCT assign-
ment, since the metal-based HOMOs of the ruthenium
compounds are higher in energy than those of the corre-
sponding rhodium species (see redox behavior).
Redox Behavior
The four bis-dimetallic species investigated here exhibit a
rich redox behavior. Figure 5 shows the cyclic voltammog-
rams of the compounds and Table 3 reports the main redox
data, together with data relative to the free ligands (bpnp
and LPh) and parent [Rh2bpnp]+ complex for comparison.
To better evidence the bi-component nature of the studied
compounds, these latter are labeled (M2)L(M2) instead of
(M2)2L (M = Rh or Ru; L = LPh or LAnt) in this para-
graph.
As already discussed,[24] the easier reduction of the LPh
ligand compared to the bpnp ligand (∆E1/2first-red = –0.28 V)
is mirrored by the reduction properties of the complexes,
with [(Rh2)LPh(Rh2)] being easier to reduce than
first-red
[Rh2bpnp]+ (∆E1/2
= –0.43 V). This observation is in
accordance with earlier work describing the LUMO of
[Rh2(acetate)3L]+ complexes (L = chelating pyridine/naph-
thyridine-type ligand) as being mainly ligand-based.[29] Sub-
stituting a pyridine for a more easily reduced pyrimidine
yields a new ligand and its derived complexes with less
negative reduction potentials. The lowering of the LUMO
of polypyridyl-type ligands upon complexation has also al-
ready been observed.[26b] Such considerations allow us to
assign the reduction potential of all the bis-dimetallic com-
plexes studied here to bridging-ligand-based processes.
Table 3 also indicates that both bis-dirhodium complexes
are easier to reduce than their bis-diruthenium counterparts
(ca. 0.1 V difference). This may be due to an enhanced π
back-donation in the bis-diruthenium complexes, which
have metal-centered HOMOs of higher energy (see below).
Figure 4. Absorption spectra of [(Rh2)LPh(Rh2)] (a), [(Rh2)-
LAnt(Rh2)] (b), [(Ru2)LPh(Ru2)] (c), and [(Ru2)LAnt(Ru2)] (d) in
acetonitrile solution.
Table 2. Absorption data in CH3CN at room temperature.
λmax [nm] (log10 ε [–1 cm–1])
[Rh2bpnp]·PF6
248
(4.69)
250
(4.9)
248
(3.1)
246
(4.5)
248
(4.8)
283
(4.33)
292
(4.6)
295
(4.5)
287
346
363
(4.63)
371
(4.6)
364
(4.6)
408 sh
(3.34)
412 sh
(4.2)
382
(4.8)
440 sh
(3.08)
468
(4.0)
478
540 sh
(3.40)
543
(3.7)
545
(3.8)
520 sh
(3.4)
600
(4.0)
602
578
(3.52)
622
(3.7)
634
(3.8)
586 sh
(3.6)
695
(4.46)
347 sh
(4.5)
347 sh
(4.5)
351
(4.5)
370
(4.6)
369
[(Rh2)LPh(Rh2)]·2PF6
[(Rh2)Ant(Rh2)]·2PF6
[Ru2bpnp]·PF6
254
(5.2)
274
(4.4)
270 sh
(4.6)
(4.1)
440 sh
(3.35)
400 sh
(4.4)
612
(3.7)
729
(3.8)
917
(3.2)
(4.3)
[(Ru2)LPh(Ru2)]·2PF6
(4.2)
701
(4.2)
[(Ru2)LAnt(Ru2)]·2PF6 254
387 (4.6)
(4.7)
(4.5)
(4.0)
3882
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
Table 3. Half-wave potentials (vs. SCE) for the oxidation and first
reduction of the four bis-dimetallic complexes and of reference spe-
cies in acetonitrile at room temperature (0.5 m, 0.05  NBu4-
ClO4). Number of exchanged electrons in square brackets. The pro-
cesses are reversible unless otherwise stated.
Reduction, E1/2
[V] vs. SCE
Compound
Oxidation, E1/2 [V] vs. SCE
bpnp[a]
no oxidation at V Ͻ +1.80
no oxidation at V Ͻ +1.80
+1.30 [1]
–1.64 [1]
–1.36 [1]
LPh[a]
[Rh2bpnp][a]
[(Rh2)LPh(Rh2)]
–0.64 [1]
+1.43 [1], +1.50 [1]
–0.21 [1][b]
–0.18 [1][b]
–0.31 [1][b]
[(Rh2)LAnt(Rh2)] +1.32 [1], +1.53 [2][c]
[(Ru2)LPh(Ru2)]
+0.76 [1], +0.85 [1]
[(Ru2)LAnt(Ru2)] +0.73 [1], +0.86 [1], +1.51 [1][c] –0.30 [1][b]
[a] Data taken from ref.[24] For the free ligands, the solvent is
dichloromethane and the reported values were obtained at 0 °C.
[b] Further reduction processes take place for all the complexes but
they are ill-behaved and will not be discussed here. [c] Irreversible
process.
The contribution of the diruthenium unit to the LUMO
destabilization thus results in a more negative reduction po-
tential.
Several oxidation waves are observed for all complexes,
as evidenced by the cyclovoltammograms reported in Fig-
ure 5. [(Rh2)LPh(Rh2)] shows two very close oxidation
waves (resolved by differential voltammetry), each of which
corresponds to a one-electron oxidation and is related to
the sequential removal of one electron from each Rh2 unit.
Therefore, the first oxidation process leads to the mixed-
valence, bi-component system [(Rh25+)LPh(Rh24+)],
whereas [(Rh25+)LPh(Rh25+)] is obtained upon further oxi-
dation. The first one-electron oxidation of [(Rh2)-
LAnt(Rh2)] is followed by a two-electron irreversible oxi-
dation at 1.53 V. To assign the various processes occurring
in this latter compound to specific subunits, it is useful to
consider that an irreversible oxidation at +1.51 V is ob-
served for [(Ru2)LAnt(Ru2)] (attributed to oxidation of the
anthryl center, see below) and the second oxidation of the
dimetallic subunit of [(Rh2)LPh(Rh2)] occurs at +1.50 V
(see above). It is therefore likely that the first oxidation pro-
cess of [(Rh2)LAnt(Rh2)] corresponds to the oxidation of
one bimetallic subunit and the subsequent irreversible bi-
electronic oxidation involves both the one-electron oxi-
dation of the second Rh2 unit and the irreversible oxidation
of the anthracene unit.
Both the bis-diruthenium complexes are much easier to
oxidize than their bis-dirhodium counterparts. This indi-
cates that the HOMOs are higher in energy for the diruthe-
nium centers than for the dirhodium ones. Once again, two
consecutive one-electron oxidations occur for the bis-ruthe-
nium species (Table 3, Figure 5), which are assigned to the
consecutive oxidation of the interacting dimetallic subunits.
The additional irreversible oxidation at 1.51 V in [(Ru2)-
LAnt(Ru2)], mentioned above, is attributed to the anthra-
cene oxidation.
Figure 5. Cyclic voltammograms (vs. SCE) of the four complexes
[(Rh2)LPh(Rh2)] (a), [(Rh2)LAnt(Rh2)] (b), [(Ru2)LPh(Ru2)] (c),
and [(Ru2)LAnt(Ru2)] (d) in acetonitrile at room temperature.
Scanning rate: 500 mVs–1. The process at about 0.4 V is the ferro-
cene oxidation, used as an internal reference.
Intercomponent Interactions
The observation of two distinct oxidation potentials for
the two M2 units in [(M2)LPh(M2)] and [(M2)LAnt(M2)]
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3883

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
suggests that a non-negligible electronic interaction exists redox properties of the bis-dimetallic compounds studied
between the two metal–metal-bonded dimetallic subunits of here with those of closely related dinuclear complexes in
the four complexes. Actually, the difference between the two which each component is made by a RuII polypyridine sub-
oxidation potentials is a function of the degree of interac- unit[26b–26d] reveals striking similarities. The structure of the
tion between the two redox centers.[30] The splittings be- naphthyridine-based ligands LPh and LAnt is indeed de-
tween the first two oxidation waves reported in Table 4 rived from that of the pyridine-based parents LЈPh and
show that [(Ru2)LPh(Ru2)] displays a slightly greater degree LЈAnt, which accommodate mononuclear ruthenium(II)
of electronic interaction than its bis-dirhodium counterpart complexes (see Table 4, right, for the structure of the RuII
(for the LAnt series, overlap between the second metal- complexes). Interestingly, the enhancement of the electronic
based oxidation and the anthracene-based irreversible oxi- interaction provided by the anthracene pyrimidine substitu-
dation in [(Rh2)LAnt(Rh2)] makes the comparison mean- ent in the bis-dimetallic complexes and the “classic” dinu-
ingless). To rationalize such a difference, it is useful to recall clear RuII polypyridine complexes[26b–26d] is the same
that, beside Coulombic factors, which are expected to be (0.04 V), thus suggesting that similar processes could be oc-
identical for both the dirhodium and diruthenium com- curring.
pounds studied here, contribution to the electronic interac-
Two main contributions can explain the anthracene ef-
tion can be related to the separation between the energetic fect. On the one hand, the anthracene moiety moves the
levels of the HOMOs, which are centered on the metallic electronic levels of the bridging ligand to lower energies, as
unit, and the LUMO, which is centered on the bridging also suggested by the reduction potentials {[(Rh2)-
ligand, assuming that an effective mechanism for inter-com- LAnt(Rh2)] is easier to reduce than [(Rh2)LPh(Rh2)], see
ponent electronic interaction is superexchange via an elec- Table 3, with the reduction mainly based on the bridging
tron-transfer pathway involving the bridging ligands.[30] ligand}, probably because the anthryl substituent can con-
Hence, on the basis of the Koopman theorem, which ap- tribute to a better delocalization of the added electron. As
proximates the energy of the orbitals to the redox potential a consequence, the LUMO–HOMO gap becomes smaller
values, a smaller difference between first oxidation and re- and the interaction is enhanced. However, according to this
duction potentials corresponds to a larger inter-component mechanism, the effect on the HOMO(s) should be negligi-
interaction. From the redox data in Table 3, the HOMOs ble, but this is not the case, since the data in Table 3 indicate
of the bis-diruthenium compounds are located at higher en- that on passing from the LPh to the LAnt series the
ergy than those of the bis-dirhodium species, and this justi- HOMOs are even more affected than the LUMO by the
fies why the interaction is larger in the former compounds presence of the anthracene group. On the other hand, the
(the effect on the HOMO is only partially compensated by anthracene moiety may contribute to the electronic coup-
a reversed effect on the LUMO).
ling between the redox sites by opening a new pathway
As highlighted in Table 4, it also appears that the anthra- for superexchange, for example by mediating a through-
cene-derived ligand promotes a larger electronic coupling space interaction involving anthracene-based orbitals.
than the 4-tert-butylphenyl-substituted ligand: the half-po- The redox data indeed indicate that the anthracene-
tential difference for [(Ru2)LAnt(Ru2)] is 0.13 V, compared based HOMO is close in energy to the metal-centered
with 0.09 V for [(Ru2)LPh(Ru2)], and, although the poten- orbitals, so that a hole-transfer pathway involving the
tial of the metal-based oxidation of [(Rh2)LAnt(Rh2)] is not anthracene HOMO is quite possible. It is likely that both
known exactly, the same appears also to be valid for the contributions come into play. This situation is schematized
dirhodium couple of complexes. Comparison between the in Figure 6.
Table 4. Differences in half-wave potentials between the first two one-electron oxidations of the four bis-dimetallic complexes and the
two related ruthenium-based bis-mononuclear analogues (the structure of the latter is presented on the right).
[a] In acetonitrile. [b] Not determined due to overlap of the second oxidation signal with the anthracene peak. [c] DMF.
3884
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
Figure 6. Schematic representation of the interaction between the subunits. In the scheme, A and B represent the dimetallic subunits,
with one of them oxidized (so the mixed-valence species is considered, for clarity). (a) refers to species with LPh as the bridge and (b)
to LAnt as the bridging ligand. Orbital interactions are shown on the left and interactions between states are represented on the right.
In (b), the anthracene (identified as “C”) HOMO contributes to the overall coupling; the bridge is here labeled LC on the “state” diagram
to highlight the contribution of anthracene. S indicates overlap integrals, Ψ the wavefunctions of the various states, and H the electron
coupling matrix elements. The superscript 0 indicates zero-order quantities (i.e. quantities that derive from pure, localized electronic
configurations). The subscripts i, f, e, and h indicate the initial, final, electron- and hole-transfer states, respectively (these latter two are
“virtual” states).[30]
The electronic information gathered through the electro- based “mononuclear” to the LPh- and LAnt-based “dinu-
chemical (and electronic absorption) studies reported above clear” complexes, mainly due to a significant LUMO lower-
is summarized in Scheme 3. In both the metal–metal- ing (LUMO centered on the ligand where a pyridine is sub-
bonded dirhodium and diruthenium complexes, the stituted by a pyrimidine, plus bimetallic coordination of the
LUMO–HOMO gap is reduced on going from the bpnp- second chelating site). An additional gap reduction is
caused by changing the substitution of the pyrimidine sub-
stituent from a phenyl to an anthryl moiety, which results
in a small further decrease of the LUMO energy and an
increase of the HOMO level.
Spectroelectrochemistry
The splitting of the oxidation processes of the four bis-
dimetallic compounds, which indicates a relatively high sta-
bility of the mixed-valence species with respect to the isoval-
ent forms,[31] prompted us to investigate the properties of
their mixed-valence compounds in more detail by per-
forming spectroelectrochemistry experiments. However, re-
sults were poor for the ruthenium complexes: featureless
absorption spectra were obtained, probably due to the very
Scheme 3. Compared electronic levels of the one and two dimetallic
site complexes of rhodium and ruthenium.
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3885

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
broad and unresolved absorption spectra of the starting first oxidation process: as for the former compound, a new
compounds (see Figure 4), particularly in the otherwise di- absorption (assigned to the IT transition) appears in the
agnostic visible and near-IR region. For this reason, our 680–900-nm region even for [(Rh2)LAnt(Rh2)], with con-
discussion will be limited to [(Rh2)LPh(Rh2)] and [(Rh2)- comitant decrease of the M2LCT band and changes in the
LAnt(Rh2)].
330–450-nm ligand-centered region. Removing the applied
Figure 7 shows the changes in the absorption spectrum potential leads back to the starting spectrum quantitatively,
of [(Rh2)LPh(Rh2)] on applying increasingly positive poten- whereas the increase of the positive applied potential leads
tials. At the beginning (potential: +1.45 V), a slight increase to the disappearance of both IT and M2LCT bands and
of absorption in the range 680–900 nm occurs, with simul- further changes in the ligand-centered region (not shown)
taneous decrease of the absorption of the M2LCT band at also in this case.
about 640 nm. A slight absorption decrease of the 350-nm
band also takes place, with formation of a new absorption
in the range 380–450 nm. The absorption feature in the red
is attributed to an intervalence transfer (IT) transition from
4+
5+
the Rh2 subunit to the Rh2 component of the mixed-
valence “dinuclear” [(Rh25+)LPh(Rh24+)] species obtained
upon first oxidation, whereas the decrease of the M2LCT
band is a consequence of the disappearance of one of the
donor orbitals of the corresponding CT transition. The
changes in the 340–440-nm region are probably the conse-
quence of an asymmetry of the coordination sites of the
otherwise symmetric bridge induced by the mixed-valence
nature of the species. On increasing the applied potential, a
new process starts, as indicated by the isosbestic points at
about 440 and 345 nm (not visible in Figure 7, where only
a limited numbers of spectra are shown for clarity): the IT
band disappears, as does the M2LCT band, and the ligand-
centered bands continue to change. The final spectrum is
constant upon successive increase of applied potential, and
the initial spectrum is recovered when the potential is elim-
inated, in agreement with the reversibility of the oxidation
processes.
Figure 8. Changes in the absorption spectrum of [(Rh2)LAnt(Rh2)]
in acetonitrile on applying positive potentials: (a) starting spec-
trum; (b) applied voltage of +1.40 V; (c) applied voltage of +1.55 V.
The possibility of following the formation and disappear-
ance of the IT band in [(Rh2)LAnt(Rh2)] allowed us to cal-
culate the electronic interaction parameters from Equa-
tion (1).[32]
2
εmax = (2380 r2/Eop ∆ν1/2) HAB
(1)
According to Hush,[32] Equation (1) correlates the optical
inter-valence transfer band with the magnitude of the elec-
tronic coupling matrix element between the two subunits A
4+
and B (in the present case, the Ru2 subunits), HAB, εmax
is the maximum molar absorption coefficient of the IT
band, Eop and ∆ν are the band maximum energy and half-
width (in cm–1), respectively, and r is the inter-component
distance (in Å). For [(Rh2)LAnt(Rh2)], εmax is
1900 –1 cm–1, Eop is 13700 cm–1, ∆ν is 3000 cm–1, and r is
9.0 Å (r is taken as the center-to-center distance). By using
the above experimental parameters, a value of 86 cm–1 is
calculated for HAB. This value compares well with the elec-
tronic coupling matrix elements of dinuclear RuII polypyri-
dine complexes bridged by good electron coupling mediator
spacers such as cyanide ligands,[33] thus confirming the sig-
nificant inter-component interaction mediated by the LAnt
bridge.
Figure 7. Changes in the absorption spectrum of [(Rh2)LPh(Rh2)]
in acetonitrile on applying positive potentials: (a) starting spec-
trum; (b) applied voltage of +1.45 V; (c) applied voltage of +1.60 V.
Figure 8 shows the changes of the absorption spectrum
of [(Rh2)LAnt(Rh2)] upon oxidation at relatively mild po-
tentials. In this case, probably due to the increased separa-
tion between the two oxidation potentials (Table 4), it is
possible to follow the first oxidation process in more detail,
as evidenced by the several isosbestic (or quasi-isosbestic)
Excited-State Properties
None of the four bis-dimetallic compounds studied here
points. Indeed, Figure 8 only shows changes related to the (including the two species containing the LAnt ligand,
3886
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
Figure 9. Transient absorption spectrum and (inset) decay of [(Rh2)LPh(Rh2)] in acetonitrile. Excitation wavelength: 532 nm.
which shows the typical anthracene fluorescence when not metal–metal-bonded dirhodium complexes.[35] The reason
coordinated to metals) exhibits any luminescence, either at for the absence of such long excited-state lifetimes in the
room temperature or at 77 K. This agrees with the excited- other compounds studied here could be the nature of the
state properties of metal–metal-bonded acetate species, metal (for the diruthenium species, for which the M2LCT
which are known to deactivate by nonradiative transi- state is probably so low in energy that nonradiative transi-
tions.[34] However, it has recently been reported[35] that di- tions are very fast) and the presence of the anthracene moi-
rhodium compounds of this class, although nonlumines- ety {for [(Rh2)LAnt(Rh2)]; the anthracene moiety could in-
cent, can exhibit lifetimes on the microsecond timescale and troduce other low-lying excited states, such as an anthra-
can therefore be easily involved in bimolecular photoin- cene-to-bridging ligand CT level, which can accelerate the
duced electron- and energy-transfer processes. This is the decay to the ground state}. Nevertheless, the relatively long-
case, for example, for the complexes [Rh2(O2CCH3)4(L)2] lived excited state of [(Rh2)LPh(Rh2)] indicates that this
(L = CH3OH, tetrahydrofuran, triphenylphosphane, pyri- class of compounds also has a good potential to be involved
dine).[35] However, the nature of the long-lived excited state in processes based on light as the energy input.
was not clarified.
The similarity between the dirhodium complexes re-
ported to exhibit the long-lived excited state[35] and the bis-
dimetallic species studied here prompted us to perform
Conclusion
nanosecond transient absorption spectroscopy (in degassed
The naphthyridine-based ligands reported here are able
acetonitrile at room temperature) on the present systems. to accommodate two sets of dimetallic redox active units
With our nanosecond-limited equipment, no transient spec- that show electronic communication with one another. Bis-
trum could be recorded for [(Rh2)LAnt(Rh2)], [(Ru2)- dirhodium and bis-diruthenium complexes of similar struc-
LAnt(Ru2)], or [(Ru2)LPh(Ru2)]. This suggests that the ex- ture have been synthesized and characterized. Among the
cited states of these latter species decay faster than the laser two families of species, electronic communication is
flash (10 ns). On the contrary, a clear absorption feature stronger in the diruthenium complexes than in their dirho-
appeared in the transient absorption spectrum of [(Rh2)- dium counterparts and is further enhanced by the replace-
LPh(Rh2)]. This transient absorption exhibits a broad ment of the phenyl group born by the bridging pyrimidine
maximum around 440 nm that extends over a large part of with an anthryl moiety. The positive effect of the anthra-
the visible region (Figure 9). It resembles the absorption cene substituent on the electronic communication between
spectrum of the reduced [(Rh2)LPh(Rh2)], so it can be as- dimetallic units is similar to that observed earlier with anal-
signed to the absorption of the radical anion of the bridging ogous mononuclear ruthenium complexes and is proposed
ligand, in agreement with the M2LCT assignment of the to be mainly due to the involvement of an anthracene-based
lowest-energy excited state for the complex.
HOMO in the interaction mechanism. Moreover, one of the
The transient spectrum of [(Rh2)LPh(Rh2)] disappears compounds exhibits a long-lived (microsecond timescale)
with a monoexponential decay, yielding a lifetime of 5.8 µs, excited state. The results presented here are promising for
the same timescale reported for the excited state of other the development of poly-dimetallic complexes built on
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3887

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
ammonia (pH ca. 8–9), extracted with ethyl acetate (3×50 mL),
and the combined organic layers were dried (Na2SO4), concen-
trated, and the brown residue purified by column chromatography
(basic alumina, CH2Cl2/EtOAc) to give 180 mg (82%) of the de-
longer naphthyridine-based strands. Furthermore, the re-
placement of the bridging pyrimidine by a pyrazine group
should lead to a linear (not curved) structure with stronger
electronic coupling between subunits, as is the case in mo-
nonuclear bi-component complexes.[26e] The introduction of
dimetallic units characterized by single metal–metal (such
as the those described here) but also by multiple metal–
metal bonds (e.g., Mo2, Re2) in larger ligands is currently
in progress to extend the work on mononuclear ruthenium
racks[26] towards nanoscale molecular wires.[36]
1
sired amino aldehyde as a yellow powder (m.p. 143 °C). H NMR
3
(CDCl3): δ = 9.90 (s, 1 H), 8.70 (br. d, J = 4.3 Hz, 1 H), 8.36 (d,
3
3
3J = 8.0 Hz, 1 H), 7.94 (d, J = 7.9 Hz, 1 H), 7.85 (d, J = 7.7 Hz,
3
3
3
1 H), 7.81 (br. t, J = 8.0 Hz, 1 H), 7.33 (ddd, J5Ј,4Ј = 7.5, J5Ј,6Ј
= 4.8, J5Ј,3Ј = 1.4 Hz, 1 H) ppm. 13C NMR (CDCl3): δ = 192.3,
4
160.2, 158.0, 155.1, 149.3, 145.0, 136.8, 124.5, 122.1, 113.8,
110.6 ppm. FAB+: m/z 200.1 [MH+]. C11H9N3O (199.21) + 0.064
EtOAc: calcd. C 66.00, H 4.68, N 20.51, O 8.81; found C 66.00, H
4.62, N 20.50.
Experimental Section
2,7-Dipyridin-2-yl[1,8]naphthyridine (bpnp): Two drops of a 10%
methanolic potassium hydroxide solution were added to a hot solu-
tion of 6-amino-2,2Ј-bipyridinyl-5-carbaldehyde (3; 75 mg,
3.76×10–4 mol) and 2-acetylpyridine (51 mg, 4.21×10–4 mol,
1.1 equiv.) in absolute ethanol (15 mL). The solution was further
refluxed for 6 h under dinitrogen. The solvent was evaporated and
the solid taken up in dichloromethane (35 mL) and washed with
water (10 mL). The aqueous layer was extracted with dichloro-
methane (5 mL) and the combined organic layers washed with
brine (15 mL), dried (Na2SO4), filtered, concentrated, purified by
column chromatography (basic Al2O3, CH2Cl2), and washed with
acetone and diethyl ether to yield an off-white solid (88 mg, 82%).
1H NMR (CDCl3): δ = 8.89 (d, 3J = 8.0 Hz, 2 H), 8.75 (d, 3J =
Materials and Methods: All reagents were used as received. Dry
solvents (dichloromethane, toluene, THF) were distilled over dry-
ing agents (calcium hydride, Na, Na/benzophenone, respectively)
under argon. 4-Amino-5-cyanopyrimidine,[37] 4-aminopyrimidine-
5-carboxaldehyde,[21] (1-ethoxyvinyl)tri(n-butyl)stannane,[20] 2-(an-
thracen-9-yl)-4,6-dichloropyrimidine,[26b] and tetra-µ-acetodiru-
thenium(II,III) chloride [Ru2(OAc)4Cl][23] were prepared according
to published protocols. 1H and 13C NMR spectra were recorded
with a Bruker AC 200 (200 and 50 MHz respectively) at 25 °C.
Electronic and luminescence spectra were recorded with a Varian
CARY 13e and an AMINCO Bowman Series 2 spectrometer,
respectively, in the reported spectroscopic grade solvents. Mass
spectrometry was performed at the Laboratoire de Spectrométrie
de masse Bio-organique (LSMBO, Strasbourg, France) and ele-
mental analyses at the elemental analysis service, Louis Pasteur
University (Strasbourg, France). IR spectra: (f) = weak, (F) =
strong, (m) = medium.
3
3
8.5 Hz, 2 H), 8.8 (m, 2 H), 8.36 (d, J = 8.5 Hz, 2 H), 7.92 (td, J
3
3
4
= 7.7, 4J = 1.7 Hz, 2 H), 7.41 (ddd, J = 7.5, J = 4.8, J = 1.0 Hz,
2 H) ppm.
4-tert-Butylbenzamidine Hydrochloride (2):[38] Gaseous hydrogen
chloride was bubbled into a solution of 4-tert-butylbenzonitrile (1;
15.94 g, 0.10 mol) in a mixture of dry benzene (25 mL) and abso-
lute ethanol (20 mL) until saturation was attained. The solution
was stirred for an hour and left to stand at room temperature for
three days. The solution was then concentrated to half its volume
in vacuo until white crystals appeared. The flask was cooled in ice
and diethyl ether was added. The white solid was then filtered and
washed with diethyl ether (3×15 mL) and dried in vacuo. It was
then suspended in absolute ethanol (40 mL) and 60 mL of ammo-
nia-saturated ethanol was added. A white precipitate formed imme-
diately and the reaction mixture was stirred for a day and left to
stand for four days. The white suspension was filtered and the fil-
trate concentrated in vacuo. The solid residue was then collected
and washed with diethyl ether and dried in vacuo to yield 16.71 g
[78%; m.p. 155 °C (dec.)] of the desired amidine hydrochloride as
The electrochemical equipment and methods have been described
previously.[26e] Absorption spectroscopy in the near-IR region were
recorded with a JASCO V570 spectrophotometer. For spectroelec-
trochemical measurements, the latter spectrophotometer was used
in connection with an EG&G 273A potentiostat. Nanosecond
transient absorption experiments were performed in argon-purged
acetonitrile solutions. A Continuum Surelite SLI-10 Nd:YAG laser
was used to excite the sample with 10-ns pulses at 355 nm. The
monitoring beam was supplied by a Xe arc lamp, and the signal
was detected by a red-sensitive photodiode after passing through a
high radiance monochromator. Differential absorption spectra
were recorded point-by-point, while kinetic measurements were
made at a fixed wavelength. Sixty four individual laser shots were
averaged to improve the reliability of each acquisition. The signals
were stored and analyzed on a dedicated PC.
1
a white solid. H NMR ([D6]DMSO): δ = 9.5 (br. s, 2 H), 9.3 (br.
3
3
s, 2 H), 7.85 (d, J = 8.6 Hz, 2 H), 7.62 (d, J = 8.6 Hz, 2 H), 1.30
7-(Pyridin-2-yl)pyrido[2,3-d]pyrimidine (8): Two drops of 10%
methanolic sodium hydroxide were added to a hot solution of 4-
aminopyrimidine-5-carboxaldehyde[21] (198 mg, 1.61 mmol) and 2-
acetylpyridine (220 mg, 1.82 mmol, 1.13 equiv.) in 23 mL of abso-
lute ethanol. The solution was then refluxed for 4 h under argon
and concentrated in vacuo. The residue was recrystallized from eth-
(s, 9 H) ppm. 13C NMR ([D6]DMSO): δ = 165.4, 156.9, 127.9,
125.7, 124.9, 34.8, 30.6 ppm. IR (KBr): ν = 3061 (F), 1674 (m),
˜
1486 (m), 852 (f), 734 (f), 674 (f), 557 (f), 210 (m) cm–1. FAB+: m/z
177.2 [C11H17N2]+. C11H17ClN2 (212.72): calcd. C 62.11, H 8.06,
N 13.17; found C 62.20, H 7.92, N 13.14.
1
anol to yield 229 mg of an off-white solid (68%, m.p. 205 °C). H
2-(4-tert-Butylphenyl)-4,6-dihydroxypyrimidine (3):[39] A freshly pre-
pared solution of sodium methoxide (6.6 g sodium in 75 mL abso-
lute methanol) was added dropwise, at room temperature, to a
suspension of 4-tert-butylbenzamidine hydrochloride (16.71 g,
79 mmol) and diethyl malonate (12.8 mL, 84 mmol, 1.07 equiv.) in
absolute ethanol (150 mL). The mixture was refluxed under dini-
trogen for 7 h. The volatile solvents were removed and the pinkish
residue taken up in water (120 mL) and acidified to pH 3–4 with
concentrated hydrochloric acid (ca. 20 mL). The pale-yellow pre-
cipitate was filtered and washed with water (2×20 mL), dried in
NMR (CDCl3): δ = 9.57 (s, 1 H), 9.50 (s, 1 H), 8.92 (d, 3J = 8.5 Hz,
1 H), 8.88 (d, 3J = 7.9 Hz, 1 H), 8.8 (br. d, 1 H), 8.44 (d, 3J =
3
4
8.5 Hz, 1 H), 7.93 (td, J = 7.9, J = 1.7 Hz, 1 H), 7.45 (m, 1 H)
ppm. FAB+: m/z 209.1 [MH+]. C12H8N4 (208.22) + 0.025 CHCl3:
calcd. C 68.38, H 3.83, N 26.53; found C 68.42, H 3.85, N 26.53.
6-Amino-2,2Ј-bipyridinyl-5-carbaldehyde (9): A solution of 7-(pyri-
din-2-yl)pyrido[2,3-d]pyrimidine (229 mg, 1.1 mmol) in 2  aque-
ous hydrochloric acid (80 mL) was refluxed for 2.5 h and then co-
oled in ice. The solution was neutralized with concentrated aqueous
3888
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
air and in vacuo to yield 18.4 g of 2-(4-tert-butylphenyl)-4,6-dihy-
tylphenyl)pyrimidin-4-yl] ethyl ketone (5Ph; 1.00 g, 3.41 mmol) and
4-aminopyrimidine-5-carboxaldehyde (890 mg, 7.23 mmol,
droxypyrimidine [92%, m.p. Ͼ 260 °C (dec.)]. 1H NMR ([D6]-
3
DMSO): δ = 8.8 (br. s, 1 H), 8.05 (d, 3J = 8.5 Hz, 2 H), 7.54 (d, J 2.1 equiv.) in absolute ethanol (100 mL) at reflux under argon. The
= 8.6 Hz, 2 H), 5.32 (s, 1 H), 1.31 (s, 9 H) ppm. 13C NMR ([D6]-
solution rapidly turned brown and a beige precipitate appeared.
The mixture was refluxed for 8 h, cooled to room temperature, and
the precipitate was centrifuged and dried in vacuo. It was then sus-
DMSO): δ = 167.3, 157.2, 154.6, 129.3, 127.5, 125.3, 88.1, 34.6,
30.8 ppm. IR (KBr): ν = 2962 (m), 1637 (F), 1325 (m), 1268 (m),
˜
1197 (f), 844 (f), 520 (f) cm–1. FAB+: m/z 245.2 [MH+]. pended in 2  aqueous HCl (340 mL) and vigorously stirred at re-
C14H16N2O2 (244.29): calcd. C 68.83, H 6.60, N 11.47, O 13.10; flux for 5 h. At room temperature, the mixture was neutralized with
found C 68.79, H 6.60, N 11.33.
concentrated ammonia. The orange suspension was filtered,
washed with water (2 ×15 mL), dried in vacuo, and purified by
flash chromatography (SiO2; CH2Cl2/EtOAc, 400:15 to 400:40) to
yield the desired bis(aminoaldehyde) [1.1 g, 70%, m.p. Ͼ 260 °C
(dec.)] as a yellow solid. A sample was recrystallized from dichloro-
2-(4-tert-Butylphenyl)-4,6-dichloropyrimidine (4Ph):[39] Phosphoryl
chloride (21 mL, 0.23 mol, 5.3 equiv.) was added dropwise followed
by 2-(4-tert-butylphenyl)-4,6-dihydroxypyrimidine (10.34 g, 42.3
mmol) portionwise to N,N-dimethylaniline (9.2 mL, 73 mmol,
1.7 equiv.) at room temperature under argon. The brown mixture
was allowed to warm up whereupon it turned reddish. The mixture
was refluxed for 1.25 h, then allowed to cool and cautiously poured
onto ice (ca. 80 g). The pinkish solid was filtered, washed with
water until the filtrate was colorless, dried in air and in vacuo, and
purified by flash chromatography (SiO2; hexane/CH2Cl2, 80:6 to
40:7) to yield 21.6 g of a crystalline white solid (92%, m.p. 93 °C).
1H NMR (CDCl3): δ = 8.35 (d, 3J = 8.6 Hz, 2 H), 7.51 (d, 3J =
8.7 Hz, 2 H), 7.22 (s, 1 H), 1.37 (s, 9 H, CH3) ppm. 13C NMR
(CDCl3): δ = 165.9, 162.0, 156.0, 132.3, 128.8, 125.8, 118.4, 35.1,
1
methane/hexane for analysis. H NMR (CDCl3): δ = 9.99 (s, 2 H),
9.11 (s, 1 H), 8.60 (d, 3J = 8.5 Hz, 2 H), 8.16 (d, 3J = 7.4 Hz, 2 H),
3
3
8.01 (d, J = 7.6 Hz, 2 H), 7.59 (br. s), 7.59 (d, J = 8.5 Hz, 2 H),
6.86 (br. s, 4 H), 1.43 (s, 9 H) ppm. 13C NMR ([D6]DMSO): δ =
193.4, 163.3, 163.0, 157.8, 157.0, 153.9, 146.2, 134.1, 127.9, 125.5,
114.6, 111.8, 110.0, 34.6, 30.9 ppm. Rf (SiO2; CH2Cl2/EtOAc,
2.0:0.3) = 0.46. IR (KBr): ν = 3349 (m), 2961 (m), 1671 (m), 1619
˜
(m), 1581 (m), 1534 (F), 1379 (m), 1211 (m), 795 (m), 210 (F) cm–1.
FAB+: m/z 453.0 [MH+]. C26H24N6O2 (452.51) + 0.37hexane:
calcd. C 69.97, H 6.07, N 17.35, O 6.61; found C 69.97, H 6.07, N
16.53.
31.3 ppm. R (SiO ; hexane/CH Cl , 2.0:0.3) = 0.47. IR (KBr): ν =
˜
f
2
2
2
2-(4-tert-Butylphenyl)-4,6-bis(7-pyridin-2-yl[1,8]naphthyridin-2-yl)-
pyrimidine (LPh): Two drops of a 10% methanolic potassium hy-
droxide solution were added to a hot solution of bis(aminoal-
dehyde) 7Ph (150 mg, 3.31 × 10–4 mol) and 2-acetylpyridine
(103 mg, 8.5×10–4 mol, 2.6 equiv.) in pyridine (26 mL) and the
solution stirred at 75 °C for 22 h. The solvent was removed in
vacuo and the solid taken up in chloroform (60 mL), which was
successively washed with water (20 mL), brine (15 mL), dried
(Na2SO4), filtered, and concentrated. The residue was washed with
acetone to yield an off-white solid (184 mg, 90%; m.p. Ͼ 260 °C),
which was recrystallized from chloroform and acetone and then
purified by flash chromatography (SiO2; CH2Cl2/CH3OH, 1.0:0.1).
1H NMR (CDCl3): δ = 10.06 (s, 1 H), 9.0 (br. d, 3J = 8.4 Hz, 4 H),
2964 (f), 1551 (F), 1519 (F), 1387 (m), 1248 (f), 1097 (f), 830 (m),
642 (f), 218 (F) cm–1. FAB+: m/z 281.1 [MH+]. C14H14Cl2N2
(280.05): calcd. C 59.96, H 5.04, N 10.05, Cl 25.31; found C 59.91,
H 5.02, N 9.85.
1-[6-Acetyl-2-(4-tert-butylphenyl)pyrimidin-4-yl]ethyl
Ketone
(5Ph):[20] Dichlorobis(triphenylphosphane)palladium(II) (450 mg)
was added to a solution of 2-(4-tert-butylphenyl)-4,6-dichloropyr-
imidine (4Ph; 6.0 g, 21,3 mmol) and (1-ethoxyvinyl)tri(n-butyl)-
stannane[20] (17.68 g, 49 mmol, 2.3 equiv.) in dry DMF (90 mL).
The solution was heated at 80 °C under argon for 12 h. At room
temperature, the black solution was poured into an aqueous solu-
tion of potassium fluoride (20 g KF in 200 mL of water). The
brown precipitate was filtered and washed with diethyl ether
(300 mL and 4×80 mL). The organic layer was washed with brine
(6×40 mL), dried (Na2SO4), filtered, and concentrated. The re-
sulting beige solid was purified by flash chromatography (SiO2;
hexane/diethyl ether, 95:5) to yield 7.22 g of the bis(vinyl) ether as
a pale-yellow solid (97%). The latter (5.11 g, 14.5 mmol) was dis-
solved in acetone (80 mL), 2  aqueous HCl (15 mL) was added,
and the solution stirred at room temperature for 6 h. The volatile
solvent was removed and the white residual solid taken up in
dichloromethane (200 mL) and washed with an aqueous saturated
solution of sodium hydrogen carbonate (60 mL). The aqueous layer
was extracted with dichloromethane (3×20 mL) and the combined
organic layers dried (Na2SO4), filtered, concentrated, and purified
by flash chromatography (SiO2; hexane/CH2Cl2, 1.0:2.0) to yield
the desired bis-acetylated arylchloropyrimidine (4.1 g, 95%, m.p.
3
8.81 (d, J = 8.4 Hz, 2 H), 8.8 (m, 2 H), 8.73 (d, 3J = 8.5 Hz, 2 H),
3
3
3
8.47 (d, J = 8.5 Hz, 2 H), 8.42 (d, J = 8.6 Hz, 2 H), 7.97 (td, J
4
3
3
= 7.7, J = 1.8 Hz, 2 H), 7.64 (d, J = 8.5 Hz, 2 H), 7.43 (ddd, J
= 7.5, J = 4.8, J = 1.1 Hz, 2 H), 1.44 (s, 9 H) ppm. IR (KBr): ν
3
3
˜
= 1604 (F), 1525 (F), 1469 (f), 1427 (m), 1371 (f), 862 (f), 792 (m),
409 (f) cm–1. FAB+: m/z 623.1 [MH+], 645.3 [MNa+]. C40H30N8
(622.72) + 0.32 CHCl3: calcd. C 74.55, H 4.70, N 17.30; found C
74.49, H 4.83, N 17.09. UV/Vis (CH2Cl2): λmax (logε) = 253 nm
(4.95), 273 (sh, 4.8), 293 (sh, 4.6), 331 (sh, 4.4), 344 (4.56), 363
(4.48).
1-[6-Acetyl-2-(anthracen-9-yl)pyrimidin-4-yl]ethyl Ketone (5Ant):
2-Anthracen-9-yl-4,6-dichloropyrimidine 4Ant (986 mg, 3.03
mmol)[26b] and 1-(tri-n-butylstannyl)-1-vinyl ethyl ether (2.40 g,
6.66 mmol, 2.2 equiv.) were suspended in 15 mL of DMF. After
purging three times with argon, dichlorobis(triphenylphosphane)-
palladium(II) (64 mg, 9.1×10–5 mol, 0.03 equiv.) was added and
the mixture heated to 80 °C under argon for 12 h whilst being pro-
tected from light. The solvent was removed in vacuo and the brown
residue taken up in dichloromethane (50 mL) and washed with
water (3×40 mL). The organic layer was dried (Na2SO4), filtered,
concentrated, and purified by flash chromatography (SiO2; hexane/
CH2Cl2, 1.0:1.0) to yield 1.078 g of the desired bis(vinyl ether) as
1
136 °C). H NMR (CDCl3): δ = 8.50 (d, 3J = 8.6 Hz, 2 H), 8.25 (s,
3
1 H), 7.58 (d, J = 8.6 Hz, 2 H), 2.83 (s, 6 H), 1.40 (s, 9 H) ppm.
13C NMR (CDCl3): δ = 199.2, 165.4, 161.4, 155.4, 133.6, 128.4,
125.9, 110.5, 35.1, 31.3, 25.8 ppm. Rf (SiO2; hexane/CH2Cl2,
1.0:2.0) = 0.30. IR (KBr): ν = 2963 (m), 1712 (F), 1551 (F), 1419
˜
(F), 1355 (F), 1239 (F), 1182 (F), 854 (m), 790 (m), 565 (m), 210
(F) cm–1. FAB+: m/z 297.4 [MH+]. C18H20N2O2 (296.36): calcd. C
72.94, H 6.81, N 9.46, O 10.8; found C 72.84, H 6.79, N 9.61.
1
a yellow solid (90%). H NMR (CDCl3): δ = 8.55 (s, 1 H), 8.05 (s,
3
3
4,6-Bis(2-amino-3-formylpyridyl)-2-(4-tert-butylphenyl)pyrimidine 1 H), 8.05 (d, J = 8.3 Hz, 2 H), 7.70 (d, J = 8.6 Hz, 2 H), 7.3–
(7Ph):[40] Five drops of a 10 % methanolic potassium hydroxide
solution were added to a solution of 1-[6-acetyl-2-(4-tert-bu-
7.6 (m, 4 H), 5.66 (d, J = 2.0 Hz, 2 H), 4.49 (d, 2J = 2.0 Hz, 2 H),
2
4.05 (q, 3J = 7.0 Hz, 4 H), 1.52 (t, 3J = 7.0 Hz, 6 H) ppm. 13C
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3889

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
NMR (CDCl3): δ = 165.5, 162.1, 157.1, 134.2, 131.5, 129.9, 128.4,
128.0, 126.1, 125.9, 125.1, 108.2, 88.5, 63.8, 14.5 ppm. Rf (SiO2;
hexane/CH2Cl2, 1.0:1.0) = 0.40.
1604 (f), 1564 (F), 1437 (f), 1158 (f), 1021 (f), 846 (F), 705 (f), 560
(f), 405 (m) cm–1. ES-MS (CH3CN, 100 V): m/z 694.0 [(Rh2)2-
LPh]2+, 1532.8 [(Rh2)2LPh, PF6]+. UV/Vis (CH3CN): λmax (logε)
= 250 nm (4.9), 292 (4.6), 347 (sh, 4.5), 371 (4.6), 412 (sh, 4.2), 468
(4.0), 543 (3.7), 622 (3.7). E1/2 (V vs. SCE, CH2Cl2, 0 °C): –1.72
(rev.), –1.41 (rev.), –1.20 (rev.), –0.60 (rev.), –0.22 (rev.), 1.46 (rev.).
2  Hydrochloric acid (6 mL) was added to a mixture of this bis(vi-
nyl ether) in acetone/THF (25/10 mL) and the mixture stirred at
room temperature for 12 h. The volatiles were evaporated and the
residue taken up in dichloromethane (50 mL) and washed with an
aqueous saturated sodium hydrogen carbonate solution
(2 × 10 mL). The combined aqueous layers were extracted with
dichloromethane (2 ×10 mL), and the combined organic layers
dried (Na2SO4), filtered, concentrated, and purified by flash
chromatography (SiO2; hexane/CH2Cl2, 1.0:2.0) to yield 828 mg of
a yellow solid [90%; m.p. 228 °C (dec.)], which was recrystallized
from chloroform/diethyl ether. 1H NMR (CDCl3): δ = 8.66 (s, 1
H), 8.57 (s, 1 H), 8.12 (d, 3J = 8.4 Hz, 2 H), 7.4–7.6 (m, 6 H), 2.72
(s, 6 H) ppm. 13C NMR (CDCl3): δ = 199.0, 168.0, 161.6, 132.3,
131.3, 129.9, 129.1, 128.8, 126.6, 125.4, 125.1, 111.2, 25.9 ppm. Rf
(SiO2; hexane/CH2Cl2, 1.0:2.0) = 0.36. EI-MS: m/z 340.3 [M–],
297.3 [M – COCH3]–, 255.3 [M – 2 COCH3]–. C22H16N2O2 (340.37)
+ 0.075 CH2Cl2: calcd. C 76.46, H 4.69, N 8.08, O 9.23; found C
76.50, H 4.62, N 8.06.
[{(CH3COO)3Rh2}2LAnt](PF6)2 [(Rh2)2LAnt]:[22a] The same proto-
col as above starting from ligand LAnt yielded a brown-green prod-
uct with a similar yield after precipitation of the hexafluorophos-
1
phate salt from a methanolic solution. (m.p. Ͼ 260 °C) H NMR
3
(CD3CN): δ = 10.13 (s, 1 H), 9.50 (br. d, J = 5 Hz, 2 H), 9.47 (d,
3J = 8.8 Hz, 2 H), 8.8–9.1 (m, 9 H), 8.52 (td, 3J = 7.7, 4J = 1,4 Hz,
3
3
2 H), 8.2–8.3 (m, 4 H), 7.92 (d, J = 8.4 Hz, 2 H), 7.47 (br. t, J =
3
8.4 Hz, 2 H), 7.31 (br. t, J = 7.3 Hz, 2 H), 1.33 (s, 6 H), 1.21 (s,
12 H) ppm. IR (KBr): ν = 3450 (F), 1605 (f), 1565 (F), 1437 (m),
˜
1414 (m), 1154 (f), 1017 (f), 845 (F), 704 (f), 559 (f), 406 (m) cm–1.
ES-MS (CH3CN, 20 V): m/z 716.0 [(Rh2)2LAnt]2+, 1576.9 [(Rh2)2-
LAnt, PF6]+. C56H44F12N8O12P2Rh4·2.6 CH3OH: calcd. C 38.98,
H 3.04, N 6.21; found C 39.24, H 3.05, N 6.65. UV/Vis (CH3CN):
λmax (logε) = 248 nm (3.1), 254 (5.2), 295 (4.5), 347 (sh, 4.5), 364
(4.6), 382 (4.8), 478 (4.1), 545 (3.8), 634 (3.8). E1/2 (V vs. SCE,
CH3CN, 25 °C): –0.30 (rev.), 0.73 (rev.), 0.86 (rev.), 1.51 (irrev.).
[{(CH3COO)3Ru2}2LPh](PF6)2 [(Ru2)2LPh]:[22b] A suspension of li-
gand LPh (14 mg, 2.25 × 10–5 mol) and [Ru2(CH3COO)4Cl]
(21.8 mg, 4.60 × 10–5 mol, 2.05 equiv.) in methanol (5 mL) was
heated at 50 °C under argon for 48 h. After cooling to room tem-
perature, the hexafluorophosphate salt was precipitated by addition
of ammonium hexafluorophosphate (38 mg in 1 mL of degassed
water, 2.33×10–4 mol, 10 equiv.). The solid was washed with meth-
anol and diethyl ether and dried in vacuo to yield 28 mg of a dark
blue paramagnetic powder (75 %; m.p. Ͼ 260 °C). 1H NMR
(CD3OD, before addition of NH4PF6): δ = 72.0, 66.7, 29.9, 25.7,
16.9, 8.8, 4.1, 2.0, 1.25, –3.8, –7.3, –13.3, –15.4, –22.0, –39.8 ppm.
2-(Anthracen-9-yl)-4,6-bis(7-pyridin-2-yl[1,8]naphthyridin-2-yl)pyr-
imidine (LAnt): One drop of a 10% methanolic potassium hydrox-
ide solution was added to a hot solution (65 °C) of the diketone
5Ant (45 mg, 1.32×10–4 mol) and 6-amino-2,2Ј-bipyridinyl-5-car-
baldehyde (54 mg, 2.71×10–4 mol, 2.05 equiv.) in pyridine (10 mL).
The solution was stirred at this temperature for 11 h. The solvent
was then removed in vacuo and the solid taken up in dichlorometh-
ane (40 mL), washed with water (10 then 15 mL), dried (Na2SO4),
filtered, and concentrated. The crude was purified by column
chromatography (basic alumina) and recrystallized by diffusion of
acetone into a concentrated chloroform solution to yield a pale-
1
yellow solid (63 mg, 72%, m.p. Ͼ 260 °C). H NMR (CDCl3): δ =
IR (KBr): ν = 3450 (F), 1593 (f), 1536 (m), 1437 (F), 1329 (f), 1197
˜
3
3
(f), 1018 (f), 843 (F), 777 (f), 691 (f), 558 (f), 407 (m) cm–1. ES-MS
(CH3CN): m/z 691.0 [(Ru2)2LPh]2+, 1527 [(Ru2)2LPh, PF6]+, 460.4
[(Ru2)2LPh]3+. C52H48F12N8O12P2Ru4·2.65 H2O: calcd. C 36.33, H
3.13, N 6.52; found C 36.33, H 3.26, N 6.59. UV/Vis (CH3CN):
λmax (logε) = 248 nm (4.8), 270 (sh, 4.6), 370 (4.6), 400 (sh, 4.4),
600 (4.0), 695 (4.2). E1/2 (V vs. SCE, degassed CH3CN, 25 °C):
–0.31 (rev.), 0.76 (rev.), 0.85 (rev.).
10.3 (s, 1 H), 9.03 (d, J = 7.9 Hz, 2 H), 8.7–8.85 (m, J = 8.4 Hz,
6 H), 8.67 (s, 1 H), 8.39 (d, 3J = 9.1 Hz, 2 H), 8.34 (d, 3J = 8.7 Hz,
2 H), 8.14 (br. d, 3J = 7,6 Hz, 2 H), 7.99 (td, J = 7.8, J = 1.8 Hz,
2 H), 7.88 (br. d, 3J = 8.6 Hz, 2 H), 7.4–7.6 (m, 6 H) ppm. 13C
NMR (CDCl3): δ = 164.7, 164.6, 160.3, 158.0, 157.3, 155.6, 149.2,
138.1, 137.7, 137.2, 131.6, 130.1, 128.6, 128.0, 127.8, 126.3, 125.3,
3
4
124.8, 123.9, 123.0, 121.1, 120.9, 120.7, 114.7 ppm. IR (KBr): ν =
˜
1601 (F), 1523 (F), 1467 (f), 1426 (m), 1379 (f), 865 (f), 783 (m),
739 (f), 413 (f) cm–1. FAB+: m/z 667.3 [MH+]. C44H26N8 (666.73)
+ 0.90 CHCl3: calcd. C 72.55, H 3.63, N 15.18; found C 72.61, H
3.57, N 15.18. UV/Vis (CH2Cl2): λmax (logε) = 250 nm (sh, 5.0),
256 (5.14), 275 (sh, 4.5), 320 (sh, 4.1), 333 (sh, 4.3), 347 (4.46), 363
(4.47), 385 (3.88). Fluorescence (CH2Cl2): λmax (emission) = 331,
356, 376 (sh) nm.
[{(CH3COO)3Ru2}2LAnt](PF6)2 [(Ru2)2LAnt]:[22b] A suspension of
ligand LAnt (19.6 mg, 2.94 × 10–5 mol) and [Ru2(OAc)4Cl]
(30.2 mg, 6.37 × 10–5 mol, 2.17 equiv.) in degassed methanol
(6.5 mL) was heated at 50 °C under argon for 44 h. After cooling
to room temperature, the hexafluorophosphate salt was precipi-
tated by addition of ammonium hexafluorophosphate (53 mg in
1.3 mL of degassed water, 3.25×10–4 mol, 11 equiv.). The solid was
washed with methanol and diethyl ether, and dried in vacuo to
yield 40 mg of a deep blue paramagnetic compound (79%; m.p. Ͼ
260 °C). 1H NMR (CD3OD, before addition of NH4PF6): δ = 75.1,
69.8, 29.9, 36.3, 25.0, 17.7, 10.7, 9.8, 5.3, 2.0, –5.0, –8.3, –14.5,
[{(CH3COO)3Rh2}2LPh](PF6)2 [(Rh2)2LPh]:[22a] 1.0  Aqueous hy-
drochloric acid (66 µL, 6.6×10–5 mol, 2.0 equiv.) was added to a
suspension of ligand LPh (20.5 mg, 3.29×10–5 mol) and dirhodium
tetraacetate (29.9 mg, 6.6×10–5 mol, 2.0 equiv.) in methanol (7 mL)
under argon and the mixture was heated to 50 °C for 12 h. The
deep green solution was filtered and ammonium hexafluorophos-
phate (51 mg, 3.1×10–4 mol, 9.5 equiv.) in water (1 mL) was added.
The solvent was then evaporated and diethyl ether allowed to dif-
fuse into a concentrated solution of the residue in acetonitrile. The
precipitate was filtered and washed with water and diethyl ether to
yield a deep green solid (47 mg, 85%, m.p. Ͼ 260 °C). 1H NMR
–16.5, –16.9, –20.0, –35.6 ppm. IR (KBr): ν = 3448 (F), 1630 (f),
˜
1597 (f), 1535 (m), 1437 (F), 1329 (f), 1266 (f), 1213 (f), 844 (F),
776 (f), 691 (m), 559 (f), 406 (f) cm–1. ES-MS (CH3CN, 20 V): m/z
713.1 [(Ru2)2LAnt]2+, 1570 [(Ru2)2LAnt, PF6]+, 475.3 [(Ru2)2-
LAnt]3+. C56H44F12N8O12P2Ru4·2.2 H2O: calcd. C 38.33, H 2.78,
N 6.39; found C 38.33, H 2.92, N 6.37. UV/Vis (CH3CN): λmax
(logε) = 254 nm (4.7), 369 (4.5), 387 (4.6), 602 (4.0), 701 (4.2).
E1/2 (V vs. SCE, degassed CH3CN, 25 °C): –0.30 (rev.), 0.73 (rev.),
0.86 (rev.), 1.51 (irrev.).
3
(CD3CN): δ = 9.71 (s, 1 H), 9.67 (br. d, J = 4.4 Hz, 2 H), 9.24 (d,
3J = 8.8 Hz, 2 H), 8.8–9.0 (m, 10 H), 8.58 (td, 3J = 8.0, 4J = 1.5 Hz,
3
3
3
2 H), 8.29 (dd, J = 7.5, J = 4.7 Hz, 2 H), 7.75 (d, J = 8.8 Hz, 2 Supporting Information (see also the footnote on the first page of
H), 1.49 (s, 15 H), 1.27 (s, 12 H) ppm. IR (KBr): ν = 3450 (F), this article): Figures S1 and S2 show differential pulse voltammog-
˜
3890
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892

Multicomponent Supramolecular Devices
FULL PAPER
[15] Y. Shi, G. T. Yee, G. Wang, T. Ren, J. Am. Chem. Soc. 2004,
126, 10552–10553.
[16] a) D. Osella, R. Rossetti, C. Nevi, M. Ravera, M. Moretta, J.
Fiedler, L. Pospísil, E. Samuel, Organometallics 1997, 16, 695–
700; b) K.-T. Wong, J.-M. Lehn, S.-M. Peng, G.-H. Lee, Chem.
Commun. 2000, 2259–2260.
rams of complexes [(Rh2)LPh(Rh2)] and [(Ru2)LPh(Ru2)], respec-
tively, in acetonitrile.
Acknowledgments
[17] a) R. H. Cayton, M. H. Chisholm, J. C. Huffman, E. B. Lob-
kovsky, J. Am. Chem. Soc. 1991, 113, 8709–87324; b) F. A. Cot-
ton, C. Y. Liu, C. A. Murillo, D. Villagrán, X. Wang, J. Am.
Chem. Soc. 2004, 126, 14822–14831; c) B. E. Bursten, M. H.
Chisholm, J. S. D’Acchioli, Inorg. Chem. 2005, 44, 5571–5579;
d) M. H. Chisholm, A. M. Macintosh, Chem. Rev. 2005, 105,
2949–2976.
A. P. acknowledges the support the French ministère de l’Éducation
Nationale, de la Recherche et de la Technologie for a PhD fellow-
ship. The Italian team thanks the University of Bologna (“Funds
for selected topics”) and the University of Messina for financial
support.
[18]
[19]
[20]
[21]
[22]
A. Petitjean, L. A. Cuccia, J.-M. Lehn, H. Nierengarten, M.
Schmutz, Angew. Chem. Int. Ed. 2002, 41, 1195–1198.
G. S. Hanan, U. S. Schubert, D. Volmer, E. Rivière, J.-M. Lehn,
N. Kyritsakas, J. Fischer, Can. J. Chem. 1997, 75, 169–182.
L. A. Cuccia, E. Ruiz, J.-M. Lehn, J.-C. Homo, M. Schmutz,
Chem. Eur. J. 2002, 8, 3448–3467.
[1] a) C. Creutz, H. Taube, J. Am. Chem. Soc. 1969, 91, 3988–
3989; b) C. Creutz, H. Taube, J. Am. Chem. Soc. 1973, 95,
1086–1094.
[2] a) V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni,
Chem. Rev. 1996, 96, 759–833; b) D. Astruc, Acc. Chem. Res.
1997, 30, 383–391; c) M. Marcaccio, F. Paolucci, C. Paradisi,
S. Roffia, C. Fontanesi, L. J. Yellowlees, S. Serroni, S. Cam-
pagna, G. Denti, V. Balzani, J. Am. Chem. Soc. 1999, 121,
10081–10091.
[3] a) 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; b) L. De Cola, P. Belser, Coord.
Chem. Rev. 1998, 177, 301–346; c) A. Ceccon, S. Santi, L. Or-
ian, A. Bisello, Coord. Chem. Rev. 2004, 248, 683–724.
[4] a) K. Lee, H. Song, B. Kim, J. T. Park, S. Park, M.-G. Choi,
J. Am. Chem. Soc. 2002, 124, 2872–2873; b) K. Lee, Y. J. Choi,
Y.-J. Cho, C. Y. Lee, H. Song, Y. S. Lee, J. T. Park, J. Am.
Chem. Soc. 2004, 126, 9837–9844; c) P. H. Dinolfo, M. E. Wil-
liams, C. L. Stern, J. T. Hupp, J. Am. Chem. Soc. 2004, 126,
12989–13001.
[5] a) H. Vahrenkamp, A. Geiß, G. N. Richardson, J. Chem. Soc.,
Dalton Trans. 1997, 3643–3651; b) W. Kaim, A. Klein, M.
Glöckle, Acc. Chem. Res. 2000, 33, 755–763 and references
cited therein; c) K. D. Demadis, C. M. Hartshorn, T. J. Meyer,
Chem. Rev. 2001, 101, 2655–2683 and references cited therein.
[6] J. A. Mc Cleverty, M. D. Ward, Acc. Chem. Res. 1998, 31, 842–
851 and references cited therein.
[7] a) B. Hong, Comments Inorg. Chem. 1999, 20, 177–207; b) R.
Ziessel, M. Hissler, A. El-Ghayoury, A. Harriman, Coord.
Chem. Rev. 1998, 178–180, 1251–1298; c) F. Fabrizi de Biani,
M. Corsini, P. Zanello, H. Yao, M. E. Bluhm, R. N. Grimes, J.
Am. Chem. Soc. 2004, 126, 11360–11369; d) Y. Hoshino, Plati-
num Met. Rev. 2001, 45, 2–11.
[8] N. Mantovani, M. Brugnati, L. Gonsalvi, E. Grigiotti, F. Las-
chi, L. Marvelli, A. Peruzzini, G. Reginato, R. Rossi, P.
Zanello, Organometallics 2005, 24, 405–418.
[9] a) R. C. Rocha, H. E. Toma, Inorg. Chim. Acta 2000, 310, 65–
80; b) Y. Hoshino, T. Suzuki, H. Umeda, Inorg. Chim. Acta
1996, 245, 87–90.
[10] a) T. Sheng, H. Vahrenkamp, Eur. J. Inorg. Chem. 2004, 1198–
1203; b) M. Heitzmann, J.-C. Moutet, J. Pécaut, O. Reynes, G.
Royal, E. Saint-Aman, G. Serratrice, Eur. J. Inorg. Chem. 2003,
3767–3773.
H. Bredereck, G. Simchen, H. Traut, Chem. Ber. 1967, 100,
3664–3670.
a) W. R. Tikkanen, E. Binamira-Soriaga, W. Kaska, P. C. Ford,
Inorg. Chem. 1984, 23, 141–146; b) E. Binamira-Soriaga, N. L.
Keder, W. Kaska, Inorg. Chem. 1990, 29, 3167–3171.
R. W. Mitchell, A. Spencer, G. Wilkinson, J. Chem. Soc., Dal-
ton Trans. 1973, 846–854.
A. Petitjean, J.-M. Lehn, R. G. Khoury, A. DeCian, N. Kyrit-
sakas, C. R. Chimie 2002, 5, 337–340.
Weak repulsive steric interactions between protons H1 and H3
prevent the ligand from being perfectly planar, which explains
the 11° angle observed between the two naphthyridine planes.
a) G. S. Hanan, C. R. Arana, J.-M. Lehn, G. Baum, D. Fenske,
Angew. Chem. Int. Ed. Engl. 1995, 34, 1122–1124; b) G. S.
Hanan, C. R. Arena, J.-M. Lehn, G. Baum, D. Fenske, Chem.
Eur. J. 1996, 2, 1292–1302; c) A. Credi, V. Balzani, S. Cam-
pagna, G. S. Hanan, C. R. Arana, J.-M. Lehn, Chem. Phys.
Lett. 1995, 243, 102–107; d) P. Ceroni, A. Credi, V. Balzani, S.
Campagna, G. S. Hanan, C. R. Arana, J.-M. Lehn, Eur. J. In-
org. Chem. 1999, 9, 1409–1414; e) A.-M. Stadler, F. Puntoriero,
S. Campagna, N. Kyritsakas, R. Welter, J.-M. Lehn, Chem. Eur.
J. 2005, 11, 3997–4009; f) J. Ramirez, M. Stadler, A. DeCian,
J.-M. Lehn, work in progress.
a) M. Klessinger and J. Michl, Excited States and Photochemis-
try of Organic Molecules, VCH, New York, 1995; b) Handbook
of Photochemistry, 3rd Edition, Revised and Expanded (Eds.:
M. Montalti, A. Credi, L. Prodi, M. T. Gandolfi), CRC Press,
2005.
B. Maubert, N. D. McClenaghan, M. T. Indelli, S. Campagna,
J. Phys. Chem. A 2003, 107, 447–455, and refs. therein.
J. L. Bear, L. K. Chau, M. Y. Chavan, F. Lefoulon, R. P. Thum-
mel, K. M. Kadish, Inorg. Chem. 1986, 25, 1514–1516.
G. Giuffrida, S. Campagna, Coord. Chem. Rev. 1994, 135–136,
517–531 and references cited therein.
a) C. Creutz, Progr. Inorg. Chem. 1983, 30, 1–73; b) B. S.
Brunschwig, C. Creutz, N. Sutin, Chem. Soc. Rev. 2002, 31,
168–184.
a) N. S. Hush, Prog. Inorg. Chem. 1967, 8, 391–444; b) V. Balz-
ani, F. Scandola, Supramolecular Photochemistry, E. Horwood,
Chichester, 1991, chapter 3.
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[11] C. Di Pietro, S. Serroni, S. Campagna, M. T. Gandolfi, R. Bal-
lardini, S. Fanni, W. R. Browne, J. G. Vos, Inorg. Chem. 2002,
41, 2871–2878.
[12] a) G. Giuffrida, G. Calogero, V. Ricevuto, S. Campagna, Inorg.
Chem. 1995, 34, 1957; b) S. Encinas, F. Barigelletti, A. M.
Barthram, M. D. Ward, S. Campagna, Chem. Commun. 2001,
3, 277–278.
[13] G. García-Herbosa, N. G. Connelly, A. Muñoz, J. V. Cuevas,
A. G. Orpen, S. D. Politzer, Organometallics 2001, 20, 3223–
3229.
[14] a) F. A. Cotton, Chem. Soc. Rev. 1975, 4, 27–53; b) Multiple
Bonds between Metal Atoms, 3rd ed. (Eds.: F. A. Cotton, C. A.
Murillo, R. A. Walton), Springer Science and Business Media,
2005.
C. A. Bignozzi, R. Argazzi, C. Chiorboli, F. Scandola, R. B.
Dyer, J. R. Schoonover, T. J. Meyer, Inorg. Chem. 1994, 33,
1652–1659.
D. M. Roundhill, Photochemistry and Photophysics of Metal
Complexes, Plenum, New York, 1994.
P. M. Bradley, B. E. Bursten, C. Turro, Inorg. Chem. 2001, 40,
1376–1379.
a) A. Harriman, R. Ziessel, Coord. Chem. Rev. 1998, 171, 331–
339; b) F. Paul, C. Lapinte, Coord. Chem. Rev. 1998, 178–180,
431–509; c) J. M. Tour, Acc. Chem. Res. 2000, 33, 791–804; d)
C. M. Lieber, Sci. Amer. 2001, 285, 58–65; e) N. Robertson,
[34]
[35]
[36]
Eur. J. Inorg. Chem. 2006, 3878–3892
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3891

A. Petitjean, F. Puntoriero, S. Campagna, A. Juris, J.-M. Lehn
FULL PAPER
C. A. McGowan, Chem. Soc. Rev. 2003, 32, 96–103; f) A. C. [38] H. von Schubert, H. Zaschke, J. Prakt. Chem. 1970, 312, 494–
Benniston, Chem. Soc. Rev. 2004, 33, 573–578; g) T. Ren, Orga-
nometallics 2005, 24, 4854–4870; h) S. Flores-Torres, G. R.
Hutchinson, L. J. Soltzberg, H. D. Abruña, J. Am. Chem. Soc.
2006, 128, 1513–1522.
506.
[39] K. von Burdeska, H. Fuhrer, G. Kabas, A. E. Siegrist, Helv.
Chim. Acta 1981, 64, 113–152.
[40] G. Evens, P. Caluwe, J. Org. Chem. 1975, 40, 1438–1439.
Received: May 19, 2006
[37] K. R. Huffman, F. C. Schaefer, G. A. Peters, J. Org. Chem.
1962, 27, 551–558.
Published Online: August 15, 2006
3892
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 3878–3892