Photochemically induced isomerisation of ruthenium polypyridyl complexes

Article abstract of DOI:10.1002/1099-0682(200102)2001:2<529::AID-EJIC529>3.0.CO;2-V

The synthesis and characterisation of a series of ruthenium polypyridyl complexes containing pyridyltriazole ligands in different coordination modes are described. The electrochemical and electronic properties of the compounds with respect to the coordination mode of the pyridyltriazole ligand are reported. Upon photolysis of the complex containing the 1-methyl-3-(pyridin-2-yl)-1,2,4-triazole ligand, irreversible ligand isomerisation is observed.

Full text of DOI:10.1002/1099-0682(200102)2001:2<529::AID-EJIC529>3.0.CO;2-V

FULL PAPER
Photochemically Induced Isomerisation of Ruthenium Polypyridyl Complexes
Stefano Fanni,[a] Frances M. Weldon,[a] Leif Hammarström,[b] Emad Mukhtar,[b]
Wesley R. Browne,[a] Tia E. Keyes,[c] and Johannes G. Vos*[a]
Keywords: Ruthenium / Coordination modes / Photochemistry / Electrochemistry / Heterocycles
The synthesis and characterisation of a series of ruthenium
polypyridyl complexes containing pyridyltriazole ligands in
respect to the coordination mode of the pyridyltriazole ligand
are reported. Upon photolysis of the complex containing the
different coordination modes are described. The electro- 1-methyl-3-(pyridin-2-yl)-1,2,4-triazole ligand, irreversible
chemical and electronic properties of the compounds with
ligand isomerisation is observed.
Introduction
In this method, the N2 isomer of a 3-(pyridin-2-yl)-1,2,4-
triazole (L1) ruthenium polypyridyl precursor complex
(RuL1N2) was directly methylated yielding complexes
which then contain the ligands 1-methyl-3-(pyridin-2-yl)-
1,2,4-triazole (L2) and 4-methyl-3-(pyridin-2-yl)-1,2,4-tri-
azole (L3). Surprisingly, the sterically least favoured com-
plex, RuL2N2, methylated at the N2 position; was obtained
as the main product, with a yield of about 90%, whereas in
the reaction between L2 and [Ru(bpy)2Cl2]·2H2O only the
N4 isomer of the complex, RuL2N4, could be obtained in
appreciable amounts. This new synthetic strategy provided
us with the first opportunity to compare their photophys-
ical and photochemical properties and to investigate the
role played by steric hindrance. In this paper we report the
photochemical and photophysical properties of ruthenium
bis(bipyridyl) complexes with the N-substituted triazole li-
gands L2, L3 and 1,4-bis[1-methyl-3-(pyridin-2-yl)-1,2,4-tri-
During the last number of years we have been actively
involved in the study of ruthenium and osmium polypyridyl
complexes containing ligands with 1,2,4-triazole moieties
such as pyridyl-[1] and pyrazyltriazoles.[2] These studies have
shown that for dinuclear compounds containing anionic tri-
azolate bridges, strong intercomponent interactions are ob-
served.[3] In addition, mononuclear complexes containing
negatively charged triazole rings were found to be pho-
tostable.[4] Of further interest in these studies is the inherent
asymmetry of the ligands. The nitrogen atoms of the pyri-
dine or pyrazine rings have significantly different electronic
properties to the triazole-based nitrogen atoms. In addition
the N1 (or N2) and N4 atoms of the triazole ring are non-
equivalent, as shown by differences between the acidϪbase
properties of complexes containing N2- and N4-bonded li-
gands.[1,2] With N-substituted triazole ligands, either N2- or
N4-coordination of the ligands is observed, depending on
the position of the substituent.[5] Furthermore, it has been
shown that for complexes containing neutral pyridyltriazole
ligands, photochemically induced isomerisation of the pyri-
dyltriazole ligand can occur.[4,6]
Recently we developed a new synthetic route towards
complexes containing N-substituted triazole ligands as
shown in Scheme 1.[7]
Scheme 1
Figure 1. Ligand structures
[a]
National Centre for Sensor Research, School of Chemical
Sciences, Dublin City University
Dublin 9, Ireland
azol-5-yl]benzene (L4) (see Figure 1). In the latter ligand,
both the N2 and N4 sites are sterically hindered. The effect
of the benzene and the methyl groups on the composition
of the compounds and on their photophysical and photo-
chemical properties were investigated. Emission lifetime,
UV/Vis and NMR studies indicate that an irreversible
E-mail: han.vos@dcu.ie
[b]
Department of Physical Chemistry, University of Uppsala,
Box 532, 75121, Uppsala, Sweden
E-mail: leif.hammarstrom@fki.uu.se
[c]
School of Chemistry, Dublin Institute of Technology,
Dublin 8, Ireland
E-mail: tia.keyes@dit.ie
Eur. J. Inorg. Chem. 2001, 529Ϫ534
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ1948/01/0202Ϫ0529 $ 17.50ϩ.50/0
529

J. G. Vos et al.
FULL PAPER
photoisomerisation takes place, in which the sterically clearly indicating that both centres are bonded in the same
hindered N2-bonded L2 complex is transformed into the manner, namely through N2.
N4 species. For the mono- and dinuclear L4 complexes, a
Redox Properties
less clear picture was obtained, but photoinduced reactions
were also observed. The results obtained are discussed with
respect to the electronic properties of the compounds.
All complexes show well-behaved electrochemistry (see
Table 2). The redox processes are reversible, with peak-to-
peak separations of 60Ϫ100 mV. RuRuL4N2N2Ј exhibits a
single, two-electron, metal-based oxidation, without any
sign of splitting, indicating that the interaction between the
metal centres is at best very weak. The N4-coordinated
complexes exhibit somewhat lower RuII/III potentials than
the N2 compounds possibly indicating improved π-acceptor
properties of ligands coordinated in the N2 mode. The sim-
ilarity of the reduction potentials of the compounds re-
ported in this work and those observed for [Ru(bpy)3]2ϩ
suggests that the first reduction process is bpy based. This
is further confirmed by the first reduction potential ob-
served for [Ru(L2N4)3]2؉ (L2 coordinated through N4) of
Ϫ1.75 V vs. SCE,[9] approximately 400 mV more negative
than that observed for [Ru(bpy)3]2ϩ. This value suggests
that the lowest unoccupied molecular orbital (LUMO) in
the L2 complex is about 3300 cmϪ1 higher than in the
homoleptic bpy complex.
Results and Discussion
General
An important issue to be addressed first is the coordina-
tion mode of the triazole rings in the obtained products, in
particular in the L4 complexes. For L2 and L3, the coor-
dination of the triazole ring is known from earlier stud-
ies.[1,7] From the reaction mixture of the mononuclear L4
complex, two isomers were isolated by semi-preparative
HPLC. In the dinuclear complex, coordination of the two
metal units may occur to N2/N2Ј, N4/N4Ј or N2/N4Ј of the
triazole ring. HPLC analysis of the reaction mixture
showed the presence of different isomers. One main isomer
was isolated in a pure form by recrystallisation.
1H NMR Spectroscopy
1
The position of the methyl signals in the H NMR spec-
tra is diagnostic for the type of isomer obtained (see
Table 1). For L2 and L4, the methyl substituent is expected
to resonate significantly further upfield than in the free li-
gand when the ligand is bonded through the N2 atom. This
is a consequence of the steric interaction between the
methyl group and a neighbouring bpy ligand. For the N4
isomer a smaller shift is expected since no interaction be-
tween the methyl group and the bpy ligands is possible.[1,5,8]
The methyl resonances recorded for the complexes fall cle-
arly into two groups. The first set of signals at δ ഠ 3.2 is
indicative of the presence of steric hindrance, since the
values are shifted upfield by about 0.7 ppm, with respect to
the free ligand. Such a shift indicates N2 coordination for
L2 and L4 complexes. The resonances found at δ ഠ 4 point
to coordination modes where the methyl group does not
interact with the neighbouring bpy ligands and are indica-
tive of N4 coordination in the L2 and L4 ligands and of
N2 coordination in the L3 complex. For the dinuclear L4
complex, only one methyl signal was obtained at δ ϭ 3.23,
Table 2. Electrochemical data for the triazole complexes; all meas-
urements were carried out in acetonitrile with 0.1  TEAP; all
redox potentials measurements were performed using ferrocene as
an internal reference
Complex
Oxidation potential
[V vs. SCE]
Reduction potential
[V vs. SCE]
RuL2N2
1.30
1.20
1.19
1.31
1.20
1.29
1.26
Ϫ1.36, Ϫ1.57
Ϫ1.39, Ϫ1.61
Ϫ1.39, Ϫ1.61
Ϫ1.35, Ϫ1.56
Ϫ1.35, Ϫ1.58
Ϫ1.38, Ϫ1.59
Ϫ1.35, Ϫ1.55, Ϫ1.80
RuL2N4
RuL3N2
RuL4N2
RuL4N4
RuRuL4N2N2Ј
[Ru(bpy)3]2؉
Electronic Properties
The absorption spectra bands observed in the
400Ϫ460 nm region are associated with dπϪπ* MLCT
bands. The MLCT bands for complexes containing a N4-
coordinated triazole ring are found at about the same en-
ergy as that of [Ru(bpy)3]2ϩ. In agreement with the electro-
chemical data, the N2-bonded isomers have an MLCT
band at higher energy suggesting that in this coordination
mode the σ-donor properties of the L2 and L4 ligands are
reduced.
Table 1. Chemical shifts (ppm) for methyl groups; measurements
were carried out in DMSO unless otherwise stated
Compound
CH3
All the compounds show emission at room temperature
and at 77 K. The emission lifetimes of the complexes in
acetonitrile, measured by time-correlated single-photon
counting, are given in Table 3. For RuL2N4 and RuL4N4,
the major decay component shows a lifetime of 9.0 ns and
35 ns, respectively. A minor (Ͻ 5%), faster component has
a lifetime of approximately 1 ns. In contrast, for RuL2N2
and RuL4N2 the 1 ns component was dominant. In these
samples, a second slower component was present, and its
magnitude varied between 15% and 50% for different meas-
RuL2N2
RuL2N4
RuL3N4
RuL4N4 (f)
(b)
3.15 (3.54[a]
3.97 (4.17[a]
4.21
)
)
4.07
3.82
RuL4N2 (f)
(b)
4.10
3.27
RuRuL4N2N2
(in CD3CN)
3.23
3.17
[a]
Measured in acetone; (f) and (b) refer to the free and bound
pyridyltriazole ring, respectively.
530
Eur. J. Inorg. Chem. 2001, 529Ϫ534

Photochemically Induced Isomerisation of Ruthenium Polypyridyl Complexes
Table 3. UV/Vis absorption and emission data for the complexes; all measurements were carried out in acetonitrile unless otherwise stated
FULL PAPER
Complex
Absorption λmax [nm] (ε [104 Ϫ1cmϪ1])
Emission λmax [nm]
300 K
τ300 K [ns]
77 K[a]
RuL2N2
435 (1.15)
452 (1.07)[a]
440 (1.44)[a]
415 (1.33)
453 (0.90)
414 (2.60)
452 (1.30)
615
566
585
584
570
584
573
582
1.0
9.0
14.0
1
RuL2N4
600[a]
600[a]
611
RuL3N2
RuL4N2
RuL4N4
614
35
RuRuL4N2N2Ј
603
1
[Ru(bpy)3]2؉
615
950[b]
[a]
[b]
Measured in ethanol/methanol (4:1, v/v). Ϫ Deaerated solution.
1
urements. The lifetime of this component matched that of were observed in the H NMR spectra of RuL2N4 after
the corresponding N4 isomer. The dimeric RuRuL4N2N2Ј irradiating for 24 h. These results suggest that irreversible
also has a dominant 1 ns component, with a minor slower photoinduced isomerisation of RuL2N2 to RuL2N4 is
(32 ns) component that matches the lifetime of RuL4N4.
The results obtained from the lifetime measurements for
the N2 isomers of L2 and L4 were at first hard to explain,
since an impurity level as high as 50% would certainly have
been detected by the other characterisation techniques. The
formation of a photoproduct seems a more likely explana-
tion for the behaviour observed. For the compounds under
investigation, the occurrence of photoinduced ligand pro-
cesses is not unexpected. It is generally assumed that upon
excitation of ruthenium polypyridyl complexes, an electron
is transferred from the metal-based ground state to a singlet
metal-to-ligand charge-transfer (1MLCT) state. From this
taking place as shown in Scheme 2.
3
singlet state, the emitting MLCT state is populated effici-
ently by intersystem crossing. Population of the nearby
antibonding triplet metal-centred state (3MC) then explains
the photoinduced reactivity of the compounds, and in the
literature both ligand exchange and rearrangements have
been reported.[4,10,11]
Figure 2. 1H NMR spectra of the methyl region taken during
photolysis of RuL2N2 in [D6]acetone
To investigate this hypothesis further, a photolysis was
carried out in which 1H NMR spectra were taken after vari-
ous irradiation times. After irradiating for 6 h in acetoni-
trile, analysis of the CH3 resonance suggested that up to
70% of the complex underwent photoinduced isomerisation
to the corresponding RuL2N4 complex. Further irradiation
(up to 12 h) resulted mainly in decomposition of the com-
plex RuL2N2. This decomposition probably occurs by the
formation of a photoinduced intermediate where the pyri-
dyltriazole ligand is coordinated in a monodentate fashion.
Such behaviour in strongly coordinating solvents has been
reported before for other complexes containing neutral pyr-
idyltriazole ligands.[6b] Under the same conditions, RuL2N4
shows a remarkable photostability and decomposition be-
Scheme 2
This photoisomerisation process can also be observed us-
comes noticeable only after 24 h of irradiation. The same ing UV/Vis absorption spectroscopy. As can be seen in Fig-
experiment was carried out in a weakly coordinating sol- ure 3, irradiation of RuL2N2 in acetone causes a gradual
vent, namely acetone, in an effort to reduce the formation decrease in intensity of the band at 415 nm and a concomi-
of secondary photoproducts (see Figure 2). Analysis of the tant growth of bands at 355 nm and 455 nm, which are as-
methyl region of the spectra shows that the resonance ori- sociated with the RuL2N4 complex (Table 3). The UV/Vis
ginally obtained for the N2 isomer at δ ϭ 3.54 is replaced spectra show isosbestic points at 448 nm and 372 nm, indi-
by a signal at δ ϭ 4.17 after photolysis. This is indicative cating that photoproducts are formed directly without the
of the formation of the N4 isomer, and in acetone typically production of long-lived intermediates. Upon complete
a 100% photoisomerisation of RuL2N2 to RuL2N4 was ob- photolysis, no further changes in the absorption spectrum
tained after irradiation for 30 h. No evidence for the forma- occur, which further indicates the photostability of
tion of a solvated metal complex was observed. No changes RuL2N4.
Eur. J. Inorg. Chem. 2001, 529Ϫ534
531

J. G. Vos et al.
FULL PAPER
Since RuL4N2 and RuRuL4N2N2Ј also showed a double-
exponential emission decay, NMR experiments were also
carried out for the L4 compounds. For all three complexes,
irradiation in acetone resulted in a complex mixture of
products, which could be attributable to partial isomer-
isation and decomposition of the starting material and
products. The results obtained from these experiments were
not further analysed.
Concluding Remarks
The results obtained in this investigation demonstrate the
importance of steric factors in the photochemical properties
of ruthenium polypyridyl compounds. Earlier studies[4] have
shown that for the protonated L1 complexes, a reversible
N2/N4 photoinduced isomerisation is taking place. The re-
sults obtained for the L2 complex indicate that sterically
hindered complexes (i.e. RuL2N2), have the tendency to re-
lease the steric congestion upon photolysis by converting
into the corresponding unhindered isomer (i.e. RuL2N4).
The presence of a steric constraint in the compound there-
fore leads to an irreversible photoisomerisation.
Figure 3. UV/Vis absorption spectra of RuL2N2 in acetone upon
irradiation for (a) 0 min, (b) 2.5 min, (c) 5 min, (d) 10 min, (e) 17.5
min, (f) 27.5 min, (g) 120 min
The conclusions drawn from the NMR and UV/Vis stud-
ies were further corroborated by emission lifetime measure-
ments on RuL2N2 in acetonitrile exposed to varying
amounts of light. The results are shown in Figure 4. To
minimise exposure of the sample to light during the lifetime
measurements, these were continued for only 2Ϫ3 min (giv-
ing 1000 counts in the peak channel). In the first measure-
ment, the contribution from the longer lived N4 isomer
emission was only 15%. However, after only a few minutes
of subsequent irradiation with a 450-W Xe lamp (using a
water heat filter and a λ Ͼ 385 nm cut-off), this contribu-
tion had increased to 80%, indicating an almost complete
conversion into the N4 isomer. When irradiation was con-
tinued, an instrument-response limited component (τ Ͻ 150
ps) started to appear (not shown), and after 100 min of
irradiation very little of the N4 isomer remained. We attrib-
ute the very short-lived component to the photolysed com-
plex in which the L2 ligand had presumably been replaced
by acetonitrile.
Pyridyltriazoles are mainly σ-donor ligands and have
only limited π-acceptor properties. Therefore it seems likely
that the small differences observed in the electronic spectra
and in the RuII/III redox potentials of the different isomers
can be explained by a reduced σ-donor capability of the
N2 isomer owing to steric hindrance. This leads to a lower
3
excited-state MC level in the N2 isomer than in the N4
isomer and facilitates the photochemically induced iso-
merisation of the N2 isomer by a photochemically driven
process.
When substituents are present at both the N1 and C5
positions (as in the three L4 complexes), neither of the pos-
sible isomers is sterically unhindered. Hence, irradiation of
these complexes results in a mixture of different isomers of
similar stability (or, in other words, ‘‘photochemical insta-
bility’’).
Experimental Section
Materials: All synthetic reagents were of commercial grade and no
further purification was employed, unless otherwise stated. All
solvents employed in spectroscopic and electrochemical measure-
ments were of HPLC or spectroscopic grade.
Experimental Methods: 1H NMR spectra were recorded with a
Bruker AC400 (400 MHz) instrument. For NMR-scale photochem-
ical experiments, 5Ϫ10 mg of a given complex was dissolved in
[D3]acetonitrile or [D6]acetone (0.5 mL), placed in a borosilicate
NMR tube and irradiated using a 150-W halogen lamp. NMR
spectra were taken at regular intervals and the degree of isomeri-
sation checked by integration of the CH3 resonance signals. Ϫ UV/
Vis spectra (accuracy Ϯ 2 nm) were obtained using a Shimadzu
UV3100 UV/VisϪNIR spectrophotometer interfaced to an Elonex
PC433 personal computer. The estimated error of the extinction
coefficients is 5%. Ϫ Emission spectra (accuracy Ϯ 5 nm) were ob-
tained with a PerkinϪElmer LS50B luminescence spectrometer
Figure 4. Emission decays for the same sample of RuL2N2 in ace-
tonitrile; curve A: sample prepared in the dark; curve B: after 1.5
min of irradiation with an Xe lamp; curve C: after 6.5 min of irradi-
ation; the lines are double-exponential fits to the data, with life-
times of 1.0 ns and 9.0 ns; the contribution from the slower com-
ponent increased from 15% in A to 40% in B and 80% in C
532
Eur. J. Inorg. Chem. 2001, 529Ϫ534

Photochemically Induced Isomerisation of Ruthenium Polypyridyl Complexes
FULL PAPER
equipped with a red-sensitive Hamamatsu R928 detector, inter-
faced with an Elonex PC466 personal computer employing
PerkinϪElmer FL WinLab custom-built software. At room tem-
perature, unless otherwise indicated, acetonitrile was the solvent
used, and excitation and emission slit widths of 10 nm, were em-
ployed. At 77 K measurements were carried out in ethanol/meth-
anol (4:1, v/v) using excitation and emission slit widths of 5 nm.
The spectra were not corrected for the photomultiplier response. Ϫ
For electrochemical measurements, HPLC grade acetonitrile dried
with molecular sieves was employed. Potentials are Ϯ 50 mV. The
Step 1. Synthesis of N2-methyl-2-pyridylamidrazone: 2-Cyanopyri-
dine (36 g, 0.35 mol) and excess methylhydrazine (18.43 g, 0.40
mol) were mixed in a minimum amount of ethanol and allowed to
stir overnight. The pale yellow crystalline product was filtered,
washed with diethyl ether and air dried. Yield 26.7 g (50%). Ϫ M.p.
1
108Ϫ110 °C (ref.[12] 109Ϫ110 °C). Ϫ H NMR (CDCl3): δ ϭ 7.24
(1 H, dd, pyridyl H5), 7.67 (1 H, dd, pyridyl H4), 8.08 (1 H, d,
pyridyl H3), 8.50 (1 H, d, pyridyl H6), 5.24 (2 H, br. s, NH), 2.98
(3 H, s, CH3). Ϫ 13C NMR (CDCl3): δ ϭ 38.48, 119.60, 123.32,
136.02, 146.41, 147.77, 150.72.
electrolyte used was 0.1
 tetraethylammonium perchlorate
Step 2. Synthesis of N,NЈ-Terephthaloylbis[(N2-methyl-2-pyridyl)hy-
drazidine]: This intermediate was prepared by the dropwise addition
of a THF solution (30 cm3) of terephthaloyl dichoride (5.08 g,
0.025 mol) to a solution of N2-methyl-2-pyridylamidrazone (7.50
g, 0.05 mol) and triethylamine (10 cm3) in THF while maintaining
the reaction mixture at 0°C. The reaction mixture was then reduced
to approximately 25 mL and an equal volume of water was added.
The yellow product was filtered, washed with water, hot methanol
and diethyl ether, dried under vacuum and then left in an oven at
60 °C for 24 h. Yield 7.00 g (65%). Ϫ M.p. 178Ϫ180 °C. Ϫ 1H
NMR (CDCl3): δ ϭ 8.15 (2 H, s, phenyl), 8.10 (1 H, d, pyridyl
H3), 7.65 (1 H, dd, pyridyl H4), 7.43 (1 H, dd, pyridyl H5), 8.60 (1
H, d, pyridyl H6), 10.30 (1 H, br. s, NH), 4.11 (3 H, s, CH3). Ϫ 13C
NMR (CDCl3): δ ϭ 38.50, 120.82, 124.67, 128.85, 129.10, 136.89,
148.24, 150.26, 153.79, 161.2.
(TEAP). The electrochemical cell used was a conventional three-
compartment cell with glass frits. The reference electrode used was
a saturated calomel electrode. The working electrode was a 3-mm
diameter teflon shrouded glassy carbon electrode and a platinum
gauze was employed as the counter electrode. Prior to reduction
measurements, the solutions were degassed for 15 min with nitro-
gen. Cyclic voltammetry (100 mVsϪ1) was carried out using a CH
instruments Model 660 electrochemical workstation interfaced to
an Elonex 486 PC. Analytical HPLC experiments were carried out
using a Waters HPLC system, consisting of a model 501 pump, a
20-µL injector loop, a Partisil SCX radial PAK cartridge mounted
in a radial compression Z module and a Waters 990 photodiode
array detector. An NEC APCIII computer controlled the system.
The detection wavelength used was 280 nm and the flow rate was
2.0 mL/min. The mobile phase was CH3CN/H2O (80:20) con-
taining 0.1  LiClO4. Semipreparative HPLC was carried out using
an ACS pump, a 1-mL injection loop and a Waters Partisil SCX
10 µm cation exchange column (25 ϫ 100 mm). The mobile phase
used was CH3CN:H2O (80:20) containing KNO3 (0.12Ϫ0.20 ).
The flow rate used varied between 1.50 and 2.0 mL/min.
Step 3. Cyclization of N,NЈ-Terephthaloylbis[(N2-methyl-2-pyrid-
yl)hydrazidine] to Form L3: N,NЈ-Terephthaloylbis[(N2-methyl-2-
pyridyl)hydrazidine] (7.00 g, 0.017 mol) was suspended in a mini-
mum volume (ca. 30 mL) of ethylene glycol and heated at reflux
for 2 h. A white crystalline precipitate was obtained upon cooling
of the solution. Further precipitation of the product was induced
by the addition of a small amount of water to the mother liquor.
The ligand was recrystallised from boiling methanol, filtered and
dried under vacuum overnight. Yield 4.20 g (66%). Ϫ M.p.
Single-Photon Counting: Fluorescence lifetimes were determined
using time-correlated single-photon counting. The excitation
source had a tuneable output of a 200 kHz and a 120 femtosecond
optical parametric amplifier (OPA, Coherent Radiation Inc). This
device was pumped by a regenerative amplifier (700 µJ, 150 fs,
200 kHz) seeded by the 80 fs output of a Ti:Sapphire oscillator. The
frequency-doubled fundamental generated in the OPA was used to
excite the samples at 400 nm. A polariser was located before the
sample to define the vertical polarisation of the excitation light and
all fluorescence measurements were carried out with an analysing
polariser before the detector was oriented at the magic angle
(54.7°), relative to the excitation polarisation, to remove anisotropy
effects. Fluorescence was observed through an interference filter
centred at 600 nm (Oriel, USA) with a full-width at half-maximum
of 10 nm. A microchannel plate multiplier (Hamamatsu) was used
for the detection. Instrumental response curves were measured by
exchanging the 600 nm interference filter after the sample with a
400 nm filter and recording the scattering from a scatter solution.
The instrumental response (FWHM) was ca. 100 ps. The fluores-
cence decay data were analysed with a nonlinear least-squares pro-
gram using a modified LevenbergϪMarquardt algorithm with iter-
ative reconvolution. The reduced χ2 and residual plots were used
to judge the quality of the fits. The samples for emission lifetime
measurements were dissolved in spectroscopic grade acetonitrile
(Merck) to an optical density of ca. 0.3 at the excitation wave-
length. The solvents were not degassed since quenching by O2 had
negligible effect on the short lifetimes measured. Lifetimes are Ϯ
10%.
1
278Ϫ280 °C. Ϫ H NMR (CDCl3): δ ϭ 8.07 (2 H, s, phenyl), 8.10
(1 H, d, pyridyl H3), 7.92 (1 H, dd, pyridyl H4), 7.45 (1 H, dd,
pyridyl H5), 8.67 (1 H, d, pyridyl H6), 4.12 (3 H, s, CH3). Ϫ 13C
NMR (CDCl3): δ ϭ 37.65, 121.58, 124.15, 129.00, 129.13, 137.16,
149.45, 149.71, 154.26, 159.88. Ϫ MS; m/z: 395 [M ϩ 1]ϩ. Ϫ
C22H18N8: calcd. C 66.98, H 4.60, N 28.41; found C 66.69, H 4.61,
N 27.95.
Synthesis of Metal Complexes: cis-[Ru(bpy)2Cl2]·2H2O[13] was pre-
pared as reported in the literature. RuL2N4 and RuL3N2 were pre-
pared as reported before.[1] RuL2N2 was prepared by direct methyl-
ation[7] of RuL1N2.[1]
Synthesis of [Ru(bpy)2L4](PF6)2: L4 (0.789 g, 2 mmol) was dis-
solved in hot methanol/water (2:1, v/v). To the solution was added
cis-[Ru(bpy)2Cl2]·2H2O (0.312 g, 0.6 mmol) and the mixture was
heated at reflux for 6 h. Following the removal of most of the sol-
vent by rotary evaporation and the addition of saturated aqueous
NH4PF6 solution, the resulting precipitate was filtered and dried.
HPLC analysis revealed the presence of two isomers in a 30:70
ratio (isomer 1/isomer 2). These were obtained in a pure form by
semipreparative HPLC using acetonitrile/water (80:20) containing
0.11  KNO3 as the mobile phase and a flow rate of 1.7 mL/min.
Total yield after purification (isomer 1 ϩ isomer 2): 0.20 g (30%).
Isomer 1 (RuL4N4): C42H34F12N12P2Ru·3H2O: calcd. C 45.78, H
3.50, N 14.58; found C 45.24, H 3.12, N 14.14.
Ligand Synthesis
Synthesis of 1,4-Bis[1-methyl-3-(2-pyridyl)-1,2,4-triazol-5-yl]ben-
zene (L4)
Isomer 2 (RuL4N2): C42H34F12N12P2Ru·H2O: calcd. C 45.19, H
3.23, N 15.06; found C 45.41, H 3.15, N 14.93.
Eur. J. Inorg. Chem. 2001, 529Ϫ534
533

J. G. Vos et al.
FULL PAPER
P. Callaghan, J. J. McGarvey, J. G. Vos, J. Phys. Chem. A 1999,
103, 8915.
R. Hage, J. G. Haasnoot, H. A. Nieuwenhuis, J. Reedijk, D. J.
A. De Ridder, J. G. Vos, J. Am. Chem. Soc. 1990, 112, 9245.
R. Wang, J. G.Vos, R. H. Schmehl, R. Hage, J. Am. Chem. Soc.
1992, 114, 1964.
Synthesis of [(Ru(bpy)2)2L4](PF6)4·3H2O (RuRuL4N2N2Ј): cis-
[Ru(bpy)2Cl2]·2H2O (0.624 g, 1.2 mmol) and L4 (0.197 g,
0.5 mmol) were heated at reflux in ethanol/water (2:1, v/v) for 6 h.
Upon cooling of the reaction mixture, a few drops of a saturated
aqueous solution of ammonium hexafluorophosphate were added,
whereupon a bright orange precipitate was obtained. This was fil-
tered and recrystallised from acetone/water (2:1, v/v). HPLC anal-
ysis of the product showed the presence of only one isomer. Yield
0.30 g (33%). Ϫ C62H50F24N16P4Ru2·3H2O: calcd. C 40.11, H 3.04,
N 12.07; found C 40.27, H 3.08, N 12.02.
[3]
[4]
[5]
E. M. Ryan, R. Wang, J. G. Vos, R. Hage, J. G. Haasnoot,
Inorg. Chim. Acta 1993, 208, 49.
[6] [6a]
B. E. Buchanan, H. Hughes, J. H. van Diemen, R. Hage, J.
G. Haasnoot, J. Reedijk, J. G. Vos, J. Chem. Soc, Chem Com-
[6b]
mun. 1991, 300. Ϫ
B. E. Buchanan, P. Degn, J. M. Pavon
Velasco, H. Hughes, B. S. Creaven, C. Long, J. G. Vos, R. A.
Howie, R. Hage, J. H. van Diemen, J. G. Haasnoot, J. Reedijk,
J. Chem. Soc., Dalton Trans. 1992, 1177.
[7]
[8]
[9]
S. Fanni, S. Murphy, J. S. Killeen, J. G. Vos, Inorg. Chem. 2000,
39, 1320.
Acknowledgments
The authors thank the EC TMR programme (Grant number
P. J Steel, F. Lahousse, D. Lerner, C. Marzin, Inorg. Chem.
1983, 22, 1488.
CT96Ϫ0031) for financial assistance.
R. Hage, Ph. D Thesis, Leiden University, 1991.
[10] [10a]
S. Tachiyashiki, H. Ikezawa, K. Mizumachi, Inorg. Chem.
[10b]
[1] [1a]
1994, 33, 623. Ϫ
Anal. Chem. 1995, 67, 4138.
R.Arakawa, S. Tachiyaskiki, T. Matsuo,
R. Hage, R. Prins, J. G. Haasnoot, J. Reedijk, J. G. Vos, J.
Chem. Soc., Dalton Trans. 1987, 1381.[1b] B. E. Buchanan, J. G.
Vos, M. Kaneko, W. J. M. van der Putten, J. M. Kelly, R. Hage,
R. A. G. de Graaff, R. Prins, J. G. Haasnoot, J. Reedijk, J.
Chem. Soc., Dalton Trans. 1990, 2425.
[11]
[12]
[13]
A. Juris, V. Balzani, F. Barigelletti S. Campagna, P. Belser, A.
von Zelewsky, Coord Chem. Rev. 1988, 84, 85.
S. Kubota, M. Uda, T. Nakagawa, J. Heterocycl. Chem. 1975,
[2] [2a]
12, 855.
R. Hage, J. G. Haasnoot, H. A. Nieuwenhuis, J. Reedijk,
R. Wang, J. G. Vos, J. Chem. Soc. Dalton Trans. 1991, 3271. Ϫ
[2b] H. A. Nieuwenhuis, J. G. Haasnoot, R. Hage, J. Reedijk, T.
L. Snoeck, D. J. Stufkens, J. G. Vos, Inorg. Chem. 1991, 30, 48.
B. P. Sullivan, D. J. Salmon, T. J. Meyer, Inorg. Chem. 1978,
17, 3334.
Received June 15, 2000
[I00231]
Ϫ
[2c] T. E. Keyes, C. M. O’Connor, U. O’Dwyer, C. G. Coates,
534
Eur. J. Inorg. Chem. 2001, 529Ϫ534