Effect of a halogenide substituent on the stability and photophysical properties of lanthanide triple-stranded helicates with ditopic ligands derived from bis(benzimidazolyl)pyridine

Article abstract of DOI:10.1039/b001818j

Bis {1-ethyl-2-[6-(N,N-diethylcarbamoyl)-4-halogenopyridin-2-yl]benzimidazol-5-yl} methane (halogeno = chloro, L^E; bromo, L^F) have been synthesized as ditopic receptors for the development of lanthanide-containing helicates able to couple with biological material and to test the influence of the halogeno substituent on the wrapping process, the structure of the resulting dimetallic edifices, and the photophysical properties of the encapsulated ions. The stability of the [Eu₂(L)₃]⁶⁺ helicates, as determined by NMR competitive titrations, decreases by respectively one (L^F) and three (L^E) orders of magnitude compared to the value found for the unsubstituted ligand (L^B) although it remains large, log β₂₃ = 23.8 (L^F) and 21.8 (L^E) in acetonitrile. The [Ln₂(L^E)₃]⁶⁺ helicates are shown to be isostructural in acetonitrile over the lanthanide series (Pr to Yb) and the crystal structure of [Tb₂(L^B)₃]⁶⁺ appears to be a good model for their solution structure, as demonstrated by paramagnetic NMR measurements (lanthanide induced shift method) and relaxation time determination. Ligand L^E appears to be a fair sensitiser of Eu^III, the quantum yield of [Eu₂(L^E)₃]⁶⁺ being 25% larger than that found for [Eu₂(L^B)₃]⁶⁺, but the ligand ³ππ* state and Tb(⁵D₄) excited level are in resonance, which limits the sensitisation of Tb^III. High resolution luminescence spectra of [Eu₂(L^E)₃]⁶⁺, both in solution and in the solid state, are presented and discussed in terms of site symmetry and vibronic coupling mechanisms. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/b001818j

View Article Online / Journal Homepage / Table of Contents for this issue
Effect of a halogenide substituent on the stability and
photophysical properties of lanthanide triple-stranded helicates
with ditopic ligands derived from bis(benzimidazolyl)pyridine†
Carlos Platas Iglesias,a Mourad Elhabiri,a Marcel Hollenstein,a Jean-Claude G. Bünzlia* and
Claude Piguetb
a Institute of Inorganic and Analytical Chemistry, BCH, University of Lausanne,
CH-1015 Lausanne, Switzerland. E-mail: jean-claude.bunzli@icma.unil.ch
b Department of Inorganic, Analytical and Applied Chemistry, University of Geneva,
30 quai E. Ansermet, CH-1211 Geneva 4, Switzerland
Received 7th March 2000, Accepted 17th May 2000
Published on the Web 15th June 2000
Bis{1-ethyl-2-[6-(N,N-diethylcarbamoyl)-4-halogenopyridin-2-yl]benzimidazol-5-yl}methane (halogeno = chloro,
LE; bromo, LF) have been synthesized as ditopic receptors for the development of lanthanide-containing helicates
able to couple with biological material and to test the influence of the halogeno substituent on the wrapping process,
the structure of the resulting dimetallic edifices, and the photophysical properties of the encapsulated ions. The
stability of the [Eu2(L)3]6ϩ helicates, as determined by NMR competitive titrations, decreases by respectively one
(LF) and three (LE) orders of magnitude compared to the value found for the unsubstituted ligand (LB) although it
remains large, log β23 = 23.8 (LF) and 21.8 (LE) in acetonitrile. The [Ln2(LE)3]6ϩ helicates are shown to be isostructural
in acetonitrile over the lanthanide series (Pr to Yb) and the crystal structure of [Tb2(LB)3]6ϩ appears to be a good
model for their solution structure, as demonstrated by paramagnetic NMR measurements (lanthanide induced shift
method) and relaxation time determination. Ligand LE appears to be a fair sensitiser of EuIII, the quantum yield of
[Eu2(LE)3]6ϩ being 25% larger than that found for [Eu2(LB)3]6ϩ, but the ligand 3ππ* state and Tb(5D4) excited level
are in resonance, which limits the sensitisation of TbIII. High resolution luminescence spectra of [Eu2(LE)3]6ϩ, both
in solution and in the solid state, are presented and discussed in terms of site symmetry and vibronic coupling
mechanisms.
LIII allow the isolation of heterodimetallic non-covalent lan-
Introduction
thanide() podates [LnM(L)3]5ϩ (M = Zn,16 Fe 17 or Co18)
Lanthanide co-ordination compounds1 are the subject of
which are building blocks for luminescent (Zn) and magnetic
intense research efforts due to their applications as contrast
(Fe, Co) materials. On the other hand, the symmetrical ligands
agents for NMR imaging,2 as catalysts in RNA hydrolysis,3 or
LA,19 LB,20 and LC 21 have been synthesized for the self-assembly
as active agents in cancer radiotherapy.4 Moreover, lanthanide-
of homodimetallic cationic helicates [Ln2(LA,B)3]6ϩ in aceto-
containing luminescent stains5 are valuable for the develop-
nitrile or highly stable and neutral [Ln2(LC Ϫ 2H)3] edifices in
ment of fluoroimmunoassays6 and of complexes which signal
water.
changes in pH, pO2 or chloride concentration.7 A unique com-
Probes for bioanalytical studies require assemblies that are
bination of features must be realised to design a lanthanide
stable under physiological conditions and which can easily
luminescent sensor: (i) presence of multiple absorbing groups
couple with biological material. A logical place to introduce a
suitable for energy transfer (antenna effect8), (ii) high thermo-
coupling group is the pyridine 4 position of ligands LB,C. In this
dynamic and kinetic inertness, and (iii) protection of the metal
paper we propose an easy strategy to introduce a halogenide
ion from various quenching or back-transfer processes.9 Several
substituent in this position, which opens the way for the prepar-
classes of receptors have been designed to meet these require-
ation of more elaborate ligands by substitution of the halogen
ments, e.g. preorganised10 and predisposed11 macrocyclic
atom by, for instance, alkyne derivatives.22 Lamture et al. have
ligands or multidentate podands.12 However, self-assembly pro-
noted that a 4-Cl substituent increases the energy transfer effi-
cesses13 appear to be best suited to achieve the required ultra
ciency in terbium() dipicolinates (pyridine-2,6-dicarboxyl-
fine tuning of the lanthanide() co-ordination sphere.14 In
ates)23 and therefore we devote special attention to the effect of
recent years we have developed a research program taking
the substitution on the photophysical properties and on the
thermodynamics of the complexes. Moreover, we take advan-
advantage of the induced-fit concept15 and using ligands
derived from bis(benzimidazolyl)pyridine in order to prepare
tage of the crystal-field independent method applied by Platas
lanthanide-containing mono- and di-metallic supramolecular
et al.24 to monometallic cryptates and recently adapted to
precursors for functional devices.14 Dimetallic edifices in which
dimetallic f–f complexes25 for the analysis of lanthanide
lanthanide ions lie at a fixed distance are interesting because
they combine two probes in one molecule and are useful for
imaging purposes, or they may be used as precursors for doped
induced paramagnetic NMR shifts.
functional materials requiring the presence of metal ions at a
Results and discussion
Synthesis of ligand LE
specific distance. The unsymmetrical hexadentate ligands LI–
Bis(1-ethyl-2-[4-chloro-6-(N,N-diethylcarbamoyl)pyridin-2-yl]-
† Electronic supplementary information (ESI) available: analytical, IR
and emission spectral data, contact and dipolar shifts, longitudinal
benzimidazol-5-yl)methane (LE) was obtained in good yield
according to the previously described two-step procedure
J. Chem. Soc., Dalton Trans., 2000, 2031–2043 2031
relaxation times. See http://www.rsc.org/suppdata/dt/b0/b001818j/
DOI: 10.1039/b001818j
This journal is © The Royal Society of Chemistry 2000

View Article Online
Table 1 1H NMR shifts (δ, with respect to TMS) for LE and the [Ln2(LE)3]6ϩ helicates in CD3CN at 298 K
Compound
H3
H4Ј
H5
H7Ј
H6Ј
H10Ј
b-CH2
H11Ј
H7
H9
H8
H10
1.08
0.83
Ϫ3.97
Ϫ1.41
0.27
LE a
La
Pr
Nd
Sm
Eu
Tb
Dy
Ho
Er
8.42
7.67
7.52
7.35
7.52
7.77
8.01
7.43
6.47
1.11
2.00
4.09
7.04
8.61
7.93
7.53
7.24
7.27
6.32
6.81
7.15
7.82
0.40
Ϫ0.96
3.03
8.69
4.72
4.27
3.77
2.49
3.13
3.64
4.55
1.42
1.53
3.33
2.37
1.69
3.59
3.35
1.26
0.96
1.42
1.28
1.05
0.48
5.04
6.26
3.51
Ϫ0.32
Ϫ1.72
Ϫ0.04
1.11
8.21
11.10
10.07
8.40
5.95
Ϫ8.05
Ϫ0.30
4.26
7.67
9.38
8.88
7.76
5.89
13.08
14.60
12.87
7.10
6.10
7.00
7.67
4.80, 4.68
6.43, 5.54
5.56, 5.10
4.95, 4.76
3.47, 3.40
4.28, 3.33
3.96, 3.54
3.54, 3.37
3.07, 2.93
2.73, 1.00
2.99, 2.14
2.94, 2.67
5.52
12.14
3.09, 2.73
0.63
4.20, 3.99
3.78, 3.06
3.27
b
b
b
17.79
21.45
16.10
7.75
4.79
6.67
Ϫ86.37
Ϫ99.73
Ϫ38.64
14.31
37.16
19.62
Ϫ3.49
13.28
15.55
9.16
Ϫ0.19
Ϫ3.33
Ϫ0.63
1.53
Ϫ29.90
Ϫ34.92
Ϫ19.78
3.19
11.18
5.61
b
b
b
b
b
b
b
b
b
b
b
b
Ϫ4.68
Ϫ0.74
4.87
6.55
4.99
Tm
Yb
Lu
10.52
8.65
7.25
1.76, 1.17
4.90, 4.76
4.13, 3.14
3.44
3.51
2.85
8.29
5.46
3.76
0.74
a In CDCl3. b Not assigned.
involving a modified Phillips type coupling reaction as the
key step in achieving the desired benzimidazole units.16b The
solution structure of 10Ϫ2 M LE in CD3CN–CDCl3 (5:1)
conformation of the two ligand arms, with the N-ethyl substitu-
ent of the benzimidazole moiety on the same side as the N atom
of the pyridine ring. Similar arrangements have been observed
previously with tridentate receptors based on bis(benzimid-
azolyl)pyridine units26 as well as for LC.21
1
and CDCl3 was investigated by H and 13C NMR, including
{1H–1H} COSY experiments (Table 1). The spectra are typical
of a species with C2 symmetry since (i) only 12 proton and 20
carbon signals are observed for the 42 protons and 39 carbon
atoms of LE and (ii) the seven pairs of methylene protons dis-
play enantiotopic protons. NOE effects are evidenced between
the bridging methylene protons and H4Ј, H6Ј as well as between
H7Ј and the methylene protons of the ethyl substituent H10Ј,
while no such effect is detected between the latter and the aro-
matic protons of the pyridine unit. This points to a transoid
Preparation and stability of the complexes
Reaction of stoichiometric amounts of LE and Ln(ClO4)3ؒ
xH2O (Ln = La, Eu, Gd, Tb or Lu) in acetonitrile–dichloro-
methane mixture gives pale yellow solutions from which the
complexes [Ln2(LE)3][ClO4]6ؒnH2O (n = 4–10) can be crystal-
lised in 65–80% yield (Table S1, supporting information). These
2032
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
complexes present a number of identifying IR bands including:
(i) an intense carbonyl stretching vibration which is red shifted
by 49 cmϪ1 with respect to the “free” ligand, (ii) unco-ordinated
H2O molecules at 3400 cmϪ1 and (iii) typical vibrations from
ionic perchlorate at 1090 and 624 cmϪ1.27
In order to get a relative stability scale for the [Ln2(L)3]6ϩ
helicates with LB, LE and LF, we have performed competitive
titrations of the europium complexes with 1,4,10,13-tetraoxa-
7,16-diazacyclooctadecane ([2,2]) in a dry CD3CN–CDCl3
(2.3:1) mixture, since the benzimidazole ligands are only spar-
ingly soluble in acetonitrile. The solutions were heated for one
week at 45 ЊC in order to reach thermodynamic equilibrium.
The concentrations of the [Eu2(L)3]6ϩ helicates and of “free”
Fig. 1 360 MHz 1H NMR spectrum of [Nd2(LE)3]6ϩ in CD3CN at
295 K.
1
ligand L were estimated from the H NMR signals of the aro-
of (i) the co-ordination of the pyridine N atom to the metal
ions and (ii) the conformational change from a transoid to a
cisoid arrangement of the two arms of the ligand. The latter
conformation is confirmed by the NOE effect detected for the
lanthanum helicate between H10Ј and H3. The spectra of the
moderately paramagnetic helicates (Ln = Pr, Nd, Sm, Eu or
Yb) also display an A2 spin system for the b-CH2 protons and
an AB system for the H10Ј protons, corresponding to a time-
averaged D3 symmetry in solution (Fig. 1). For the samarium
helicate, which displays very small isotropic shifts and sharp
signals, assignments could be made referring to the spectrum
of the diamagnetic lanthanum helicate while the spectra of
the complexes of Nd and Eu were assigned with the help of
{1H–1H}-COSY and {1H–1H}-ROESY spectra. For strongly
paramagnetic lanthanide complexes resonance assignments
were made on the basis of signal integration, linewidth
analyses and the results obtained for the complexes of Pr
and Eu.
matic protons, the total concentrations of L and [2,2] being ca.
5ؒ10Ϫ3 and 2ؒ10Ϫ4 M, respectively and the total concentration
in metal ranging between 2.8ؒ10Ϫ3 and 3.1ؒ10Ϫ3 M. It is
noteworthy that signals from [Ln2(L)2]6ϩ complexes were not
detected under these conditions.20 Defining [L]f as the concen-
tration of “free” ligand and [L]t as the total concentration
of ligand, we have found [L]f/[L]t = 0.30, 0.62 and 0.47 for LB,
LE and LF, respectively. These data demonstrate that the
stability of the helicates decreases gradually in the following
sequence: [Ln2(LB)3]6ϩ > [Ln2(LF)3]6ϩ > [Ln2(LE)3]6ϩ. That is,
the [Ln2(L)3]6ϩ helicates become less stable with increasing
electronegativity of the substituent introduced in the pyridine 4
position.
The absolute value of the stability constants may be esti-
mated by using log K(Eu[2,2]) = 9.7, as determined in pure
CH3CN,28 and taking into account a model involving both 2:2
and 2:3 helicates since these species were shown to form with
LA, LB, and LC. We find log β22(Eu) = 20.9 and log β23(Eu) = 25.4
for LB, log β22(Eu) = 18.3 and log β23(Eu) = 21.8 for LE, and
log β22(Eu) = 19.9 and log β23(Eu) = 23.8 for LF, where log
β22(Eu) and log β23(Eu) are defined by equilibria (1) and (2).
Finer structural information can be gained from the separ-
ation of the contact (δcij) and pseudocontact contributions (δ ) to
pc
ij
para 29
the isotropic paramagnetic shift (δ
)
of a nucleus i induced
ij
by a lanthanide ion j (LIS), eqn. (3) where the diamagnetic
2 Eu3ϩ ϩ 2 LB,E,F
2 Eu3ϩ ϩ 3 LB,E,F
[Eu2(LB,E,F)2]6ϩ log β22(Eu) (1)
[Eu2(LB,E,F)3]6ϩ log β23(Eu) (2)
pc
δpijara = δiejxp Ϫ δdi ia = δcij ϩ δ
(3)
ij
contribution δdi ia is obtained by measuring chemical shifts for
isostructural diamagnetic complexes: [La2(LE)3]6ϩ was used for
the earlier (Pr to Dy) and [Lu2(LE)3]6ϩ for the later (Ho to Yb)
members of the lanthanide series. Equations have recently been
developed for homodimetallic complexes with large Ln ؒ ؒ ؒ Ln
separation (8–9 Å) in which pseudocontact contributions affect
the LIS of protons which are significantly remote from the
paramagnetic centre.21b For axial lanthanide complexes30 the
pseudocontact contribution is given by eqn. (4) where A20 〈r2〉 is
Although these data can only be taken as an estimate (evaluated
error: 1.5 to 2), the experimental conditions employed
not being entirely the same, the results obtained for LB are in
very good agreement with those determined previously by a
direct spectrophotometric method (log β22(Eu) = 19.9 and log
β23(Eu) = 24.1).20
Solution structure of [Ln2(LE)3]6؉ in acetonitrile
(A20 〈 r2 〉)n 1 Ϫ 3 cos2 θin
2
2
δpijc = Σ
ؒ
ͩ
ͪ
ؒCnj = Σ Gni Cnj
(4)
T2
(rni )3
The 1H NMR spectra of the lanthanum and lutetium
diamagnetic complexes (Table 1) show an A2 spin system for the
bridging b-CH2 group, which is typical for C2-related enantio-
topic protons, while the AB spin system observed for methylene
protons H10Ј arises from two symmetrically non-equivalent
diastereotopic protons, leading to the conclusion that the heli-
cates adopt a time-averaged D3 symmetry on the NMR time-
scale.16,20 The complete assignment of the spectra was made on
the basis of homodinuclear {1H–1H}-COSY and {1H–1H}-
ROESY (rotating frame Overhauser enhancement spectro-
scopy) experiments. Analysis of the spectra reveals that the
ionic radius contraction does not modify substantially the
overall structure of the helicates. However, the signal of the H4Ј
protons is shifted by ϩ0.48 ppm in going from La to Lu,
pointing to some distortion in the co-ordination sphere of the
heavier metal ion.20 This proton experiences a large shift upon
complexation (∆δ = ϩ1.74 and ϩ2.22 ppm for La and Lu,
respectively) because the peculiar conformation of the helicates
brings it in the shielding zone of the benzimidazole units.19,20
The pyridine protons also undergo substantial shifts as a result
n = 1
n = 1
the crystal field parameter, Cjn the anisotropic part of the axial
magnetic susceptibility tensor, rni and θin are the internal axial
coordinates of nucleus i with respect to the threefold axis of
site n and Gni is the pseudocontact term originating from ion n at
a given temperature. On the other hand, the contact contribu-
tions δcij only affect nuclei relatively close to the paramagnetic
centre and it can be assumed that in the helicates [Ln2(LE)3]6ϩ
they result from the interaction with a single metal ion. Under
these conditions, and assuming that the paramagnetic centres
do not interact in view of their large separation, the linearised
equations (5)–(7) hold21b where 〈Sz〉j is the spin expectation
para
ij
δ
Cj
= Fi ϩ Ggi lobal
(5)
〈 Sz 〉j
〈 Sz 〉j
para
δ
〈 Sz 〉j
ij
= Ggi lobal ϩ Fi
(6)
Cj
Cj
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2033

View Article Online
1
Table 2 Computed values for contact (Fi) and pseudocontact (Ggi lobal) terms and agreement factors AFi and AFj for H nuclei in [Ln2(LE)3]6ϩ in
CD3CN solutiona
b-CH2
H4Ј
H6Ј
H7Ј
H3
H5
H11Ј
H10
H8
Pr to Dy
θ1i /Њa
r1i /Åa
θ2i /Њa
r2i /Åa
53.15
6.634
46.28
39.84
3.917
22.41
53.13
7.019
49.12
58.81
6.755
46.04
81.55
5.476
33.16
9.862
0.245(6)
0.199(1)
1.23
0.0263
0.412
0.261
0.000
109.21
5.420
24.13
11.974
0.170(5)
0.116(3)
1.47
0.0940
0.382
0.195
0.000
83.14
6.618
38.53
10.513
0.039(3)
0.153(1)
0.25
0.0124
0.000
140.28
5.515
14.26
13.605
Ϫ0.116(6)
Ϫ0.399(3)
0.29
0.02036
0.126(8)
Ϫ0.307(7)
0.000
111.38
6.613
26.84
12.998
0.031(7)
0.049(5)
0.63
0.2237
0.09
0.083(2)
0.000
7.349
6.556
7.427
8.015
b
Fi
Ϫ0.049(1)
Ϫ0.101(1)
0.49
0.0239
0.000
Ϫ0.078
0.0757
Ϫ0.23(3)
Ϫ1.17(2)
0.19
0.0346
0.000
Ϫ1.06
0.0018
Ϫ0.025(3)
Ϫ0.085(2)
0.29
0.0519
0.000
Ϫ0.053
0.3377
0.144(6)
Ϫ0.017(2)
Ϫ8.47
0.0557
0.156
Ϫ0.013
0.000
Ggi lobal b
Fi/Ggi lobal b
b
AFi
Fi
c
Ggi lobal c
0.149(3)
0.0801
c
AFi
Ho to Yb
b
Fi
0.080(4)
Ϫ0.066(1)
Ϫ1.21
0.0361
Ϫ0.045
0.1(2)
Ϫ0.73(2)
Ϫ0.14
0.3551
Ϫ0.61
0.062(1)
Ϫ0.0711(4)
Ϫ0.87
0.0125
Ϫ0.03
0.098(4)
Ϫ0.031(3)
Ϫ3.16
0.0635
Ϫ0.01
Ϫ0.20(1)
0.097(1)
Ϫ2.06
Ϫ0.143(3)
0.049(2)
Ϫ2.92
Ϫ0.138(2)
0.1142(9)
Ϫ1.21
0.0118
0.09
0.434(9)
Ϫ0.270(4)
Ϫ1.61
Ϫ0.0164
0.055(1)
Ϫ0.30
0.0447
0.048
Ggi lobal b
Fi/Ggi lobal b
b
AFi
0.0949
0.0351
0.0175
Ggi lobal c
Pr
Nd
Eu
Tb
Dy
Ho
Er
Tm
Yb
b
AFj c
0.0322
0.0245
0.136
0.0474
0.0314
0.113
0.0681
0.0516
0.251
0.0160
0.0256
0.190
0.0442
0.0270
0.1746
0.0223
0.1882
0.0315
0.0588d
0.187
0.0209
0.0523d
0.187
AFj e
AFj
0.203
0.381
a Averaged axial coordinates rin and θin tabulated from the crystal structure of the [Tb2(LB)3]6ϩ helicate. b According to Reilley’s method. c According to
Kemple’s method. d According to the dipolar model excluding H3, H5 and H10 from the fits. e According to the dipolar model.
AFi = [Σ(δiejxp Ϫ δ )2/Σ(δeijxp)2]1/2 and
cal
ij
j
j
cal
AFj = [Σ(δiejxp Ϫ δ )2/Σ(δiejxp)2]1/2
(7)
ij
i
i
value and Fi the contact term at fixed temperature. Assuming
further that 〈Sz〉j and Cj values are the same for the complexes
and the free ions,31 for which they are tabulated,32,33 plots of
δpijara/〈Sz〉 against Cj/ 〈Sz〉 and of δipjara/Cj against 〈Sz〉/Cj should be
linear with a slope equal to Ggi lobal = Gi1 ϩ Gi2 and Fi, respectively,
if the complexes are isostructural and possess comparable
crystal field parameters. The pseudocontact and contact con-
tributions to the observed LIS separated by using eqns. (5) and
(6) are reported in Table 2 along with the agreement factors AFi
and AFj (see also Table S2, supporting information). The dia-
stereotopic methylene protons of the ethyl substituents have
been excluded from the analysis because a reliable assignment is
not possible for an AB spin system.21b We have also not used
samarium data in our calculations because the shifts induced by
this ion are small and extremely temperature dependent. The
resulting plots fall into two groups (Ln = Pr to Dy and Ln = Ho
to Yb) with a break near the middle of the series as illustrated
in Fig. 2 for H10. In principle, such a situation points to the
[Ln2(LE)3]6ϩ helicates being not isostructural. However, devi-
ations from linearity sometimes occur because of variation in
the crystal field parameter34 or in both this parameter and the
hyperfine coupling constant Fi.24 The agreement factors calcu-
lated for both series, 0.012 < AFi < 0.22 (Pr to Dy) and
0.01 < AFi < 0.36 (Ho to Yb), are comparable to those found
for the 3d–4f helicates [LnZn(L)3]5ϩ with L = LI (0.08–0.27),16a
LII (0.07–0.20),16b or for [Ln2(LC Ϫ 2H)3] (0.0015 < AFi <
0.35).21b The contact contributions are relatively small for
most of the protons in the first half of the series (Ln = Pr to
Tb), Fi reaching significant values only for H4Ј, H3, H5, H7Ј, and
H10. The large Fi values calculated for the H3 and H5 protons for
both series point to an important spin density delocalisation
onto the pyridine ring. However, these values are smaller than
those obtained previously for [LnZn(L)3]5ϩ, with L = LI (0.34,
0.34)16a or LII (0.35, 0.22),16b and for [Ln2(LC Ϫ 2H)3] (0.32,
Fig. 2 Plots according to eqns. (5) and (6) for the H10 protons.
0.36 for Ce to Tb and Ϫ0.47, Ϫ0.46 for Er to Yb),21b pointing to
a reduced spin density delocalisation on the pyridine rings, as a
consequence of the electronegative substituent.
Recently, Platas et al.24 proposed a crystal-field independent
method, which allows one to rationalise the origin of the breaks
2034
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
Table 3 Geometric ratios Rik and intercepts (Fi Ϫ RikFk) for [Ln2(LE)3]6ϩ (CD3CN, 298 K)
b-CH2–H4Ј b-CH2–H6Ј b-CH2–H11Ј b-CH2–H10 H4Ј–H11Ј H4Ј–H10 H4Ј–H6Ј
H11Ј–H10 H6Ј–H10Ј H6Ј–H11
Pr to Yb
a
Rik
0.087(2)
0.94(4)
0.01(1)
Ϫ0.61(2)
Ϫ0.008(9)
0.249(3)
Ϫ0.018(3)
Ϫ7.9(4)
Ϫ0.2(2)
2.86(8) 10.8(5)
Ϫ0.41(1)
0.03(1) Ϫ0.02(2) Ϫ0.01(1)
0.25(2) Ϫ0.61(2)
(Fi Ϫ RikFk)a Ϫ0.013(7)
Ϫ0.11(9) H4Ј–H11Ј
Pr to Dy
b
Rik
0.086
1.19
Ϫ0.66
0.253
Ϫ7.65
2.93
13.76
Ϫ0.38
0.21
Ϫ0.56
Ho to Yb
b
Rik
0.090
0.93
Ϫ0.58
0.244
Ϫ6.39
2.70
10.27
Ϫ0.42
H5–H3
0.26
Ϫ0.62
b-CH2–H3 b-CH2–H5 H3–H4Ј
H5–H4Ј
H3–H10
H5–H10
H11Ј–H3
H11Ј–H5
H6Ј–H5
H6Ј–H3
Pr to Dy
a
Rik
Ϫ0.507(7)
0.069(4)
Ϫ0.508
Ϫ0.87(2)
0.087(6)
Ϫ0.87
Ϫ0.170(3)
0.18(1)
Ϫ0.17
Ϫ0.099(3)
0.127(9)
Ϫ0.099
Ϫ0.499(7)
0.174(7)
Ϫ0.499
Ϫ0.290(7) 0.767(4) 0.58(1)
0.121(7) Ϫ0.136(2) 0.020(6) Ϫ0.162(7) 0.010(4)
1.32(2) Ϫ0.73(1)
Ϫ2.34(5)
0.20(1)
Ϫ2.34
(Fi Ϫ RikFk)a
b
Rik
Ϫ0.291
0.769
0.58
1.32
Ϫ0.73
Ho to Yb
a
Rik
Ϫ0.67(6)
Ϫ1.34(7)
Ϫ0.11(1)
Ϫ1.35
Ϫ0.134(6)
Ϫ0.11(2)
Ϫ0.133
Ϫ0.067(6)
Ϫ0.09(2)
Ϫ0.067
Ϫ0.36(4)
0.07(5)
Ϫ0.36
Ϫ0.18(1)
Ϫ0.07(2)
Ϫ0.18
1.16(8)
0.11(4) Ϫ0.04(3)
1.18 0.51
0.51(3)
2.3(1)
1.20(2) Ϫ0.15(1)
2.3 Ϫ1.45
Ϫ1.45(6)
Ϫ1.37(9)
Ϫ0.13(3)
Ϫ1.36
(Fi Ϫ RikFk)a Ϫ0.07(2)
b
Rik
Ϫ0.68
a According to eqn. (8). b According to eqns. (5) and (6).
responsible for the break observed between Dy and Ho (Fig. 2),
rather than structural variation, and that these protons have
similar hyperfine coupling constants with all the studied lan-
thanide ions, consistent with a single structure along the series.
(ii) On the other hand, plots for the pairs involving the H8 and
H7Ј protons do not show obvious correlation, which can be
explained by the low Gglobal values (Table 2) inducing large
errors in Giglobal/Gkglobal. (iii) Finally, plots for pairs involving H3
and H5 clearly show two approximately parallel straight lines,
one for the larger (Pr to Dy) and one for the smaller (Ho to Yb)
ions which implies a significant change in the hyperfine coup-
ling constants for these protons in the two series. This intricate
behaviour can be explained if one compares the contact contri-
butions obtained by Reilley’s method (Table 2), which display
important variations between the first and second half of
the lanthanide series. However, the contact contribution and/or
the contact to dipolar quotient (Fi/Ggi lobal) are relatively small
for most of the protons, except for H3 and H5. Thus, the plots
according to eqn. (8) for the latter protons are very sensitive to
changes in the hyperfine coupling constant.
Turning our attention to the Rik values (Table 3) we note that
for pairs generating a single straight line the values obtained
according to both Reilley’s and the crystal-field independent
methods are in excellent agreement. For the pairs generating
two lines, the Rik values for the first and second half of the
series are in qualitative good agreement, pointing to a minor
structural change in the pyridine units between the helicates of
Dy and Ho. Furthermore, a plot of Ggilobal (Ho to Yb) vs. Ggi lobal
(Pr to Dy) for all protons generates a straight line going
through the origin with a slope of 0.62(2) (R2 = 0.997), indicat-
ing that the Ggi lobal values for the first and second part of the
lanthanide series differ only by a proportionality constant
assigned to a change in the crystal field parameter between Dy
and Ho. We conclude that the solution structure of the helicates
is essentially maintained along the series and that the breaks
found between Dy and Ho in the plots according to eqns. (5)
and (6) are due to changes in the crystal field parameter and
in the hyperfine coupling constant as similarly discussed for
[Ln2(LC Ϫ 2H)3] in water.25
Fig. 3 Plots according to eqn. (8) for the pairs b-CH2–H10 (top) and
H4Ј–H3 (bottom).
found in the plots according to eqns. (5) and (6), by simul-
taneously solving eqn. (3) for two different nuclei i and k. Its
extension to dimetallic helicates leads to eqn. (8).25 The corre-
para
para
δ
δ
ij
kj
Rik = Ggi lobal/Gkglobal
(8)
= (Fi Ϫ RikFk) ϩ R
,
〈Sz〉j
ik 〈Sz〉j
sponding plots for the studied helicates present three different
behaviours depending on the protons. (i) Pairs involving H4Ј,
H6Ј, H11Ј, H10 and b-CH2 display a single straight line (Pr to Yb,
Fig. 3) indicating that changes in the crystal field parameter are
Substitution of Bleaney’s coefficients Cj by a direct
determination of the magnetic susceptibility tensor35 consider-
ably improves the separation of the contact and dipolar contri-
butions to the LIS. To calculate the latter, the axial coordinates
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2035

View Article Online
rni and θin (n = 1 or 2) of the protons b-CH2, H4Ј, H6Ј, H7Ј, H3, H5,
H11Ј, H8 and H10 need to be determined. As it is reasonable
to assume that the introduction of a chloride substituent does
not affect substantially the structure of the helicates, we have
calculated the axial coordinates from the crystal structure of
[Tb2(LB)3]6ϩ 20 using the Tb–Tb pseudo-threefold axis as the
magnetic z axis. LIS values have been first fitted by eqn. (9)
method (Table 2). We conclude that the solid state structure of
[Tb2(LB)3]6ϩ is also a good model for the solution structure of
the [Ln2(LE)3]6ϩ helicates with the heavier lanthanide ions and
that no drastic structural change occurs for the LE complexes
along the lanthanide series.
Further structural information can be obtained by measuring
the NMR longitudinal relaxation times T1. The effects of the
paramagnetic centre on T1 are accounted for by the Solomon–
Bloembergen–Morgan theory,37 which reduces to dipolar and
Curie-spin contributions, depending on riϪ6, for lanthanide
complexes.29 The use of a nucleus sufficiently remote from the
paramagnetic centre as an internal reference gives eqn. (10) in
1 Ϫ 3 cos2 θi1 1 Ϫ 3 cos2 θ2i
δpijara = ξχzj z
ͩ
ϩ
ͪ
ϩ Σ δcij
(9)
(r1i )3
(r2i )3
i
without including the contact contributions. As expected, these
fits were rather poor (cf. the large agreement factors, Table 2). A
much better agreement (0.025 < AFj > 0.052) was obtained for
the first half of the series when contact contributions for H7Ј,
H3, H5, H8 and H10 were treated as fitting parameters. Thus,
the mathematical treatment employed required the calculation
of five parameters for each lanthanide ion in addition to the
experimental axial anisotropic susceptibility parameter ξχjzz
(9 × 6 fit). Although the improvement obtained should be con-
sidered with caution because adding new fitting parameters
lowers the AFj values,36 the calculated contact and dipolar con-
tributions obtained from both Reilley’s and Kemple’s methods
are in qualitative agreement (Table 2, Tables S2 and S3, sup-
porting information), except for the contact contribution for
H10. We conclude that the solid state structure of [Tb2(LB)3]6ϩ is
a good model for the solution structure of the helicates of Pr to
Dy with LE. This statement is confirmed by the good match
observed between the theoretical Cj values and the experimental
axial component of the anisotropic susceptibility (Table 4).
For the second half of the series (Ln = Ho to Yb) the contact
contributions are dominant for most of the protons and this
mathematical treatment could not successfully be employed.
The theoretical values for the ratio of the contact to dipolar
contributions to the isotropic shifts should be of the order of
Ϫ0.58:0.47:0.15:0.12 for an isostructural series of complexes
of trivalent Ho, Er, Tm and Yb. Thus, good fits, according to
the dipolar model, are expected for Tm and Yb only, which
is indeed the case when the H3, H5 and H10 protons having the
largest contact to dipolar ratio (Table 2) are excluded from the
fits. Furthermore, the dipolar contributions calculated by this
method are in good agreement with those obtained by Reilley’s
dia
ref
krteoftϪ k
ri
6
=
ͩ ͪ
rref
(10)
ktiot Ϫ kdi ia
which ktiot and kdi ia are the longitudinal relaxation rates measured
for the nucleus i in the paramagnetic complex and its dia-
magnetic analogue, respectively. For dimetallic helicates the T1
of each nucleus is the sum of two dipolar paramagnetic contri-
butions,21b eqn. (11). The relaxation times of the paramagnetic
krteoftϪ k
(r1ref)6 ϩ (r2ref)6
(r1i)6 ϩ (r2i)6
(r1i)6(r2i)6
(r1ref)6 (r2ref)6
dia
ref
=
ͩ
ͪ
ϩ
ͩ
ͪ
(11)
ktiot Ϫ kdi ia
[Ln2(LE)3]6ϩ complexes (Ln = Pr or Yb), corrected for dia-
magnetic contribution using the values for the helicates of La
and Lu, are reported in Table 5 (uncorrected values; Table S4,
supporting information). Taking the rni distances from the crys-
tal structure of [Tb2(LB)3]6ϩ (Table 2) and H8 as an internal
reference, we have estimated the expected quotients (krteoftϪ k )/
dia
ref
(ktiot Ϫ kdi ia). The differences between the two approaches (eqns.
(10) and (11)) are minor for protons H3, H5, H8, H10 and H4Ј
(<2.5%) because r1i and ri2 are sufficiently different and the con-
tribution of one paramagnetic centre dominates. Eqn. (10) can
thus be used to extract the approximate Ln ؒ ؒ ؒ H distances
collected in Table 5. The experimental distances obtained from
the relaxation data of the complexes of Pr and Yb are in excel-
lent agreement, which confirm again that there is no important
structural change along the series. Moreover, a comparison of
dia
the experimental [(krteoftϪ k )/(kitot Ϫ kdi ia)]exp values obtained
ref
from the relaxation measurements with those calculated from
the crystal structure of the [Tb2(LB)3]6ϩ helicate shows a satis-
fying concordance for all the protons except for H10.
Table 4 Comparison of theoretical (Cj) and experimental (ξχzj z/ppm
Å3) values of the axial component of the anisotropic susceptibility for
[Ln2(LE)3]6ϩ
Photophysical properties
Ligand-centred transitions. The emission spectrum of LE in
acetonitrile solution (295 K) under excitation at 31 546 cmϪ1
presents a single band (Table 6) whose intensity quickly dimin-
ishes when a short time delay (0.1 ms) is enforced and therefore
zz
Pr
Ln
Cj
ξχzj z
Cj/CPr
ξχzj z/ξχ
Pr
Ϫ11.00
Ϫ4.20
4.00
Ϫ86.00
Ϫ100
Ϫ761
Ϫ342
339
Ϫ5022
Ϫ5753
1.00
0.38
Ϫ0.364
7.82
1.000
0.45
Ϫ0.44
6.60
7.56
1
Nd
Eu
Tb
Dy
has been attributed to the ππ* state. This band is shifted by
ca. 1550 cmϪ1 to higher energy with respect to that of LB. The
absolute fluorescence quantum yield is large and amounts to
QL = 92%. The emission spectrum recorded in frozen solution
9.09
Table 5 Ln ؒ ؒ ؒ H distances in CD3CN [Ln2(LE)3]6ϩ solutions calculated from corrected experimental relaxation times (T1) and normalised to the
Tb ؒ ؒ ؒ H8 distance
Ln
b-CH2
H6Ј
H7Ј
H11Ј
H4Ј
H3
H5
H10
H8
a
a
a
Pr
T1/ms
r1i /Å
623.4
259.4
25.91
3.76
8.41
3.75
3.91
0.034
0.033
0.042
190.0
180.7
156.1
766.5
6.61
257.43
6.61
6.61
1
1
1
5.25
84.4
5.50
5.48
0.25
0.33
0.32
5.21
72.2
5.35
5.42
0.24
0.28
0.31
5.06
52.7
5.06
5.52
0.20
0.20
0.34
Yb
T1/ms
226.9
259.5
189.1
r1i /Å
Tbb
Pr
r1i /Å
6.63
7.02
6.76
6.62
a
dia
c
c
a
a
[(krteoftϪ k )/(ktiot Ϫ kidia)]exp
0.81
1.01
0.85
ref
dia
Yb
[(krteoftϪ k )/(ktiot Ϫ kidia)]exp
0.88
0.67
1.01
0.85
0.73
0.96
ref
Tbb
[(krteoftϪ k )/(ktiot Ϫ kdi ia)]calc
dia
c
ref
a Overlapping of some peaks prevents the T1 measurement. b From the crystal structure of [Tb2(LB)3]6ϩ c According to eqn. (11) with H8 as an
.
internal reference, see text.
2036
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
Table 6 Ligand-centred absorption and emission properties of LE and [Ln2(LE)3]6ϩ
Compound
LE
π → π*a
π → π*b
1ππ*c
3ππ*c
τ(3ππ*)c
29 420 (sh)
28 484
41 298
25 883
16 806
18 957
21 199
18 860
19 860
670 20
30 841 (4.77)
40 160 (4.51)
29 710 (5.10)
40 095 (5.03)
47 820 (5.45)
29 480 (5.10)
39 940 (5.04)
48 080 (5.41)
La
Eu
Gd
Tb
Lu
26 232
40 955
47 567
25 976
41 857
48 121
26 438
41 857
47 463
26 193
40 838
47 633
26 407
41 088
47 592
22 730
22 821
22 245
587 30
20 860
d
d
17 946
19 219
5.3 0.1
20 455
d
d
29 520 (5.10)
39 900 (5.02)
47 820 (5.42)
18 388
18 999
20 337
146
5
a Electronic spectral data in acetonitrile at 295 K; energies are given for the maximum of the band envelope in cmϪ1, and log ε is given within
parentheses. b Reflectance spectra recorded at 295 K. c Luminescence data and lifetimes (ms) in frozen CH3CN solutions at 77 K. d 3ππ* luminescence
quenched by transfer to the lanthanide ion.
[Eu2(LE)3]6ϩ helicate in acetonitrile solution. The excitation
spectrum of a 7ؒ10Ϫ4 M solution of [Eu2(LE)3]6ϩ in acetonitrile
produces a broad and intense band with a maximum at 25 200
cmϪ1, corresponding to excitation through the 1ππ* ligand state.
The corresponding emission spectrum points to the complexes
having an averaged trigonal symmetry in solution, in agreement
with the NMR data. There is a single, symmetrical and broad
5
(full width at half-height fwhh = 21 cmϪ1) D0 → 7F0 transi-
tion at 17 226 cmϪ1, which is consistent with the presence of
a single europium() co-ordination environment. Using the
phenomenological relationship of Frey and Horrocks38 corre-
lating the energy of the 0–0 transition and parameters
describing the ability δ of co-ordinating atoms to produce a
nephelauxetic effect, v Ϫ v0 = CCNΣiniδi, and δCO = Ϫ15.7, we
find δC᎐N(LE) = Ϫ16.8 cmϪ1, a value somewhat larger than the
᎐
averaged one deduced from our previous studies of complexes
with heterocyclic imines, δC᎐N = Ϫ15.3.26 The δ parameters are
᎐
however very sensitive to bonding distances,39 as shown for
[Eu2(LB)3]6ϩ in the solid state for which three different metal ion
sites were evidenced whose 0–0 transitions correspond to δC᎐N
parameters ranging from Ϫ13.7 to Ϫ18 cmϪ1.20 The large
δ value obtained for the LE helicate is nevertheless consistent
with the large Fi values calculated for the H3 and H5 protons
(see above) and pointing to an important spin delocalisation
Fig. 4 Corrected phosphorescence spectra of LE and the [Ln2(LE)3]6ϩ
helicates at 77 K (νexc = 26 316 cmϪ1). Vertical scale: arbitrary units.
onto the pyridine ring.
The emission spectrum (cf. Table 7) may be interpreted in
terms of a pseudo-D3 symmetry, with a very weak 5D0 → 7F0
transition (forbidden in D3 symmetry), two components for
the transition to 7F1 (allowed A1 → A2 and A1 → E
transitions, the latter being split by distortion from the trigonal
symmetry into two components spaced by 38 cmϪ1) and two
(77 K) presents a second more structured band with a maxi-
mum at 18 957 cmϪ1 and with a number of low- and high-
energy shoulders (Fig. 4). This band has a single exponential
time decay with a lifetime of 670 20 ms and is therefore
3
7
assigned to the ππ* state. This lifetime is considerably longer
components for the transition to F2 (two allowed A1 → E
5
than the one reported for LA (4.4 ms19) and LB (4.2 ms20),
which should promote efficient ligand-to-metal energy transfer
processes.
transitions). The D0 lifetime, 1.62 0.01 ms (Table 8), is long
and identical to the one found for the helicate with LB
(1.67 0.06 ms), consistent with fairly rigid metal ion sites.
Reinhoudt and co-workers40 have concluded from their work
on modified EuIII-containing calix[4]arenes that the antenna
effect is improved when the 3ππ* 0-phonon transition lies 3500
cmϪ1 above the lanthanide excited state. They also observed
that the 1ππ* → 3ππ* intersystem crossing is maximised when
the energy difference between these states amounts to ca. 5000
cmϪ1. A similar conclusion was reached by Latva et al.,39 find-
ing that the best efficiency in energy transfer is obtained when
the 0-phonon band of 3ππ* lies at 21 000–22 000 cmϪ1. Compar-
ing the relative ability of LB and LE to sensitise europium()
In acetonitrile the [Ln2(LE)3]6ϩ helicates display three main
absorption bands at 48 000, 40 000 and 29 500 cmϪ1 assigned
respectively to π → π* transitions mainly located on the
carbonyl, pyridine and benzimidazole moieties. The low
energy band undergoes a red shift of 1360 cmϪ1 upon com-
plexation. Similarly, the ligand-based singlet state emission
band shifts to lower energies in the helicates of LaIII, GdIII
and LuIII and the energy of the 0-phonon transition of
the 3ππ* state undergoes an increasing bathochromic shift
with increasing charge density of the metal ion: 339 (La), 744
(Gd) and 858 (Lu) cmϪ1.
1
luminescence, we note the following facts: (i) the ππ*–3ππ*
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2037

View Article Online
Table 7 Energy (cmϪ1) of the 5D0 and of the identified 7FJ crystal field
sublevels (origin: 7F0) in the [Eu2(LE)3]6ϩ helicate from luminescence
spectra at 10 (Eu-H10, Eu-H0, solid state) and 295 K (Eu-H0 in MeCN)
Eu-H10
site Ia
site Ib
Eu-H0
10Ϫ3 M in MeCN
5D0
7F1
17 212
318
374
17 227
298
386
17 218
308
383
17 226
333
390
436
970
1 002
458
966
1 003
442
428
7F2
963 a
999 b
1 005 b
1 053
1 078
1 103
1 833
1 866
1 004
1 094
1 044
1 081
1 106
1 830
1 045
1 094
1124
1 837
7F3
7F4
1 824
2 655
2 677
2 724
2 811
2 880
2 983
3 006
2 694
2 739
2 813
2 897
2 988
3 016
2 701
2 722
2 809
2 880
2 990
2 699
2 735
2 831
2 900
Fig. 5 Corrected emission spectra of anhydrous [Eu2(LE)3]6ϩ at 295 K:
3 001
(a) 7ؒ10Ϫ3 M in CH3CN, νexc = 25 316 cmϪ1; (b) νexc = 25 000 cmϪ1, solid
state; (c) νexc = 17226 cmϪ1, solid state. Vertical scale: arbitrary units.
a Vibronic transition. b Splitting due to Fermi resonance (see text).
studied in order to unravel the influence of the hydration of the
complex, one of them being anhydrous (Eu-H0) and the other
one containing ten hydration water molecules (Eu-H10).
The emission spectrum of Eu-H0 recorded at 295 K under
excitation through the ligand ππ* state (Fig. 5) cannot simply
be analysed in terms of a pseudo-D3 symmetry, even if some of
Table 8 Lifetimes of the Eu(5D0) excited levels (ms) in the solid state
samples of [Eu2(LE)3]6ϩ under various excitation conditions
Compound
Eu-H10
T/K
νexc/cmϪ1
τ/ms
1
10
17 227
17 221
17 209
17 227
17 221
17 209
17 224
1.86 0.01
2.12 0.01
2.16 0.02
1.91 0.02
2.13 0.01
2.19 0.01
0.33 0.02a
1.64 0.01a
2.15 0.01
2.20 0.01
2.27 0.01
2.12 0.04
2.28 0.03
2.18 0.01
1.47 0.01
1.62 0.01
its characteristics are consistent with such a local symmetry at
5
the co-ordination centre: (i) the D0 → 7F0 transition, which
appears at 17 226 cmϪ1 is fairly broad (fwhh = 16 cmϪ1), indicat-
ing some statistical distribution of europium() sites,1 and it is
quite weak (Table S5, supporting information), consistent with
the fact that this transition is forbidden in D3 symmetry; (ii) the
77
295
10
7
transition to the F2 level is comprised of two main bands
Eu-H0
17 224
17 221
17 215
17 203
17 203
17 221
17 224
17 227
(allowed A1 → E transitions), but these bands appear to be
5
split; (iii) the D0 → 7F4 transition presents four lines, again
consistent with D3. On the other hand, the 5D0 → 7F1 transi-
tion comprises three almost equally spaced components, point-
ing to a low symmetry, since two transitions only are allowed in
D3. One of the latter being an A1 → E transition, slight devi-
77
295
295
b
7
ations from the idealised D3 symmetry lift the F1(E) degener-
a Decomposition of bi-exponential decay. b 7ؒ10Ϫ 4 M in dry deoxygen-
acy and often result in the observation of two lines for this
transition. Here the distortion appears to be quite large with a
splitting of 60–75 cmϪ1. As a comparison, splittings in the
range 20–30 cmϪ1 were observed for the helicate with LB. To try
to solve the ambiguity about the local symmetry at the euro-
pium() ion sites, we have recorded the 5D0 ← 7F0 excitation
spectrum. It consists of one relatively broad band (fwhh =
11–12 cmϪ1) with a maximum at 17 218 and 17 223 cmϪ1 at 10
and 295 K, respectively, and a shoulder on the high energy side.
Upon analysing at different emission wavelengths correspond-
ing to transitions to F1 and F2, two additional components
could be unravelled, at 17 228 and 17 209 cmϪ1. However, the
emission spectra recorded under excitation at the maximum of
the emission band and in both shoulders are nearly identical
(Fig. F1, Supporting Information), pointing to very similar
metal-ion environments and suggesting that the additional
shoulders appearing on the excitation spectra may be caused by
vibronic transitions.1,41 It could also be possible that the two
metal ion sites are not completely identical within the
same molecule or, even, as observed for [Ln2(LC Ϫ 2H)3],
that two different types of molecules, with differing Ln ؒ ؒ ؒ Ln
distances, are present in the solid state. However, the crystal
structures we have determined on several helicates have always
shown that the two metal ion sites are fairly identical, differing
ated CH3CN.
energy difference amounts to 4510 (LB) and 3600 cmϪ1 (LE) but
(ii) the 0-phonon transition of the ligand ππ* state (as meas-
3
ured for the gadolinium complexes at 77 K) lies at 20 930 and
20 445 cmϪ1 for LB and LE, respectively, leading to ∆E(3ππ*–
5D0) differences equal to 3700 (LB) and 3220 cmϪ1 (LE). From
these data one expects similar antenna effects in both helicates,
which is observed: the absolute quantum yield of the metal-
centred luminescence in acetonitrile amounts to 3.5 × 10Ϫ3 for
LB 20 and 4.4 × 10Ϫ3 for LE. We note that these data parallel the
results obtained for terbium() complexes with dipicolinates
where the 4-Cl substituted compound proved to have better
energy transfer efficiency than the unsubstituted complex.23
7
7
[Ln2(LE)3]6ϩ (Ln ؍ Eu or Tb) helicates in the solid state. In
complete contrast with the complexes of La, Gd and Lu, the
emission from the ligand 3ππ* state of helicates of Eu and Tb is
completely quenched and the characteristic emission bands
of the metal ions appear (Fig. 4). Since the europium helicate
is fairly luminescent, we have investigated its high resolution
emission spectra to gain information on the inner co-ordination
sphere around the EuIII. Furthermore, two samples were
2038
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
emission spectra (see Table 7), while excitation at the maximum
yields a spectrum which is a combination of the two other ones
(Fig. F1, supporting information). The lifetime of site Ib (1.86
ms) is shorter than that of site Ia (2.16 ms, identical to the
lifetime of the anhydrous sample). Both lifetimes are long,
pointing to the absence of inner-sphere water molecules. Taking
into account ∆k = 0.15 msϪ1 proposed by Beeby et al.44 for the
quenching effect of a closely diffusing unbound water molecule,
we calculate that the lifetime difference corresponds to the pres-
ence of 0.5 H2O in the outer co-ordination sphere of Eu(Ib).
Upon increasing the temperature the lifetime of this site
decreases much more drastically than the lifetime of site Ia,
confirming larger vibronic interaction. Therefore, hydration of
the helicate does not result in co-ordination of water in the
inner sphere of the metal ion, but it generates non-covalent
interactions with the ligand strands resulting in two slightly
different metal ion sites. We cannot determine whether we are
in the presence of two different types of molecules or of
one type of unsymmetrical molecule. In the crystal structure
of [Ln2(LC Ϫ 2H)3]ؒ21.5H2O, not only two different types
of molecules were observed, but within a dimetallic edifice,
outer-sphere interaction with water molecules was found to be
different for each metal ion site.21
The excitation spectrum of the terbium helicate at 77 K pro-
duces a broad band with a maximum at 25 814 cmϪ1, corre-
1
sponding to the excitation through the ππ* ligand state. The
corresponding emission spectrum displays D4 → 7FJ transi-
5
Fig. 6 Comparison between emission (solid line), IR (dotted line) and
Raman (dashed line) spectra of [Eu2(LE)3]6ϩ in the energy range of the
5D0 → 7F1 (bottom) and 5D0 → 7F2 (top) transitions.
tions at 20 362, 18 355, 17 101, 16 088 cmϪ1 for J = 6, 5, 4, and 3,
7
respectively. It is dominated by the transition to F5, as shown
by the integrated and corrected relative intensities: 0.54, 1.00,
5
only by small distortions and slightly dissimilar Ln–O and
Ln–N distances. This may cause a broadening but not a
splitting of the luminescence bands. Developing further the
hypothesis of an interaction with vibronic levels to explain
the emission spectrum of [Eu2(LE)3]6ϩ we note that the highest
energy 5D0 → 7F1 emission line undergoes a shift varying lin-
early with the excitation wavelength from 16 882 to 16 935 cmϪ1
(νexc = 17 209–17 227 cmϪ1), as has been previously observed for
a macrocyclic complex.42 A correlation between the vibrational
and emission spectra in the energy range of the transition to
7F1 clearly shows interference with vibrational levels (Fig. 6).
Therefore the large splitting observed for the A1 → E transi-
0.30 and 0.12 for J = 6, 5, 4 and 3. The lifetime of the D4(Tb)
state is short and sharply decreases from 1.10 0.05 ms at 77 K
to 0.122 0.001 ms at 295 K, which is characteristic of a com-
plex in which a TbIII→LE energy back transfer occurs as a result
of the close proximity of the 5D4 and 3ππ* states, ∆E = 83 cmϪ1,
as was also observed for the helicate with LB (∆E = 70 cmϪ1).20
The luminescence decay of [Tb2(LF)3][ClO4]6 is monoexponen-
tial with respect to temperature and an Arrhenius type plot
yields an activation energy of 1410 20 cmϪ1, confirming the
back-transfer mechanism.
Conclusion
7
tion to the F1 sublevel could arise from such an interaction.43
5
Indeed, with respect to the energy of the D0 level, the IR and
The introduction of a chloride substituent in the 4 position of
the pyridine rings of ligand LB is straightforward and the result-
ing compound LE is obtained in five steps with good overall
yield, 32%, as calculated from the starting 4-hydroxypyridine-
2,6-dicarboxylic acid. On the other hand, the bromide substi-
tuted ligand LF, which is a more interesting synthon for further
substitution at the 4 position of the pyridine, is more difficult to
synthesize as indicated by an overall yield of 4–5% only. When
treated with stoichiometric amounts of lanthanide perchlorates
in acetonitrile, both ligands yield triple-stranded dimetallic
helicates [Ln2(L)3]6ϩ the stability of which is large (log β23 in the
range 22–24) because the destabilising effect of the halogenide
remains small. The structure of these edifices is very similar to
the one determined for helicates with ligand LB, pointing to the
substitution not influencing the overall wrapping of the ligand
strands around the two lanthanide ions. The electro-attracting
chloride substituent in LE results in a somewhat improved
luminescence of the europium() ion over the helicate with the
unsubstituted ligand, which demonstrates that tuning the effi-
ciency of the ligand-to-metal energy transfer process may be
achieved by varying the substituent at this position of the lig-
and. Moreover, the developed synthetic technology will enable
us to graft coupling groups on this 4 position in order to pro-
duce bimetallic lanthanide-containing triple helicates able to
combine with biological material45 and therefore it opens the
way to a whole series of new ditopic ligands for the design of
luminescent (and magnetic) probes.
7
Raman peaks at ca. 412 cmϪ1 fall in between the two F1 sub-
levels with energy of 383 and 442 cmϪ1 which could therefore be
the result of an interaction between a vibrational and an elec-
tronic state having very similar energy. Furthermore, at least six
5
maxima can be identified in the D0 → 7F2 transition, but
again there is an extensive interaction with the phonon density
7
of states (Fig. 6), so that we interpret some closely spaced F2
sublevels (Table 7) as resulting from interaction with vibrational
levels, e.g. the levels at 999 and 1005 cmϪ1 (Raman band at 1005
cmϪ1) or at 1078 and 1103 cmϪ1 (broad IR band at 1100 cmϪ1,
Raman peak at 1075 and 1120 cmϪ1). Therefore, the emission
spectrum of sample Eu-H0 can be interpreted as arising from
two very similar europium() ions in a severely distorted
pseudo-D3 site symmetry, a conclusion reinforced by the fact
that the Eu(5D0) lifetime remains the same when the excitation
5
energy is scanned through the D0 ← 7F0 profile: 2.19 0.07
ms (Table 8). The lifetime is long, as expected for an euro-
pium() environment free of OH oscillator but it is temper-
ature dependent which points to vibrations interacting with the
electronic levels of the metal ion and facilitating non-radiative
de-excitation, in agreement with the above interpretation.
The 5D0 ← 7F0 transition for the hydrated Eu-H10 sample
presents a maximum at 17 221 cmϪ1 (10 K, fwhh between 13
and 17 cmϪ1 depending on the analysis wavelength) and two
shoulders at 17 212 (site Ia) and 17 227 (site Ib) cmϪ1. Excit-
ation onto these two shoulders generates two slightly different
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2039

View Article Online
(ν(C᎐O) acid) and 1575 (ν(C᎐C) py). ESI-MS(CH OH): m/z
Experimental
Solvents and starting materials
᎐
᎐
3
201.9, [M ϩ H]ϩ; and 401.0, [2M ϩ H]ϩ.
4-Bromopyridine-2,6-dicarboxylic acid 1b. The compound
Acetonitrile,
dichloromethane,
N,N-dimethylformamide
was obtained by hydrolysis of diethyl 4-bromopyridine-2,6-
(DMF), and triethylamine were distilled from CaH2, thionyl
chloride from elemental sulfur. Silica gel (Merck 60, 0.04–0.06
mm) was used for preparative column chromatography. Other
products were purchased from Fluka AG (Buchs, Switzerland)
and used without further purification, unless otherwise stated.
1
dicarboxylate as described above for 1a (90% yield). H NMR
in DMSO-d6: δ 8.36 (2 H, s). 13C NMR in DMSO-d6: δ 134.15,
138.20, 153.64 and 168.42. ESI-MS (MeOH): m/z 266.0,
[M ϩ H ϩ H O]ϩ. IR ν(cmϪ1, KBr): 1732 (ν(C᎐O) acid) and
᎐
2
1571 (ν(C᎐N) py).
᎐
Spectroscopic and analytical measurements
4-Chloro-6-methoxycarbonylpyridine-2-carboxylic acid 2a. To
a solution of 4.0 g of compound 1a (19.8 mmol) in water–
MeOH (1:1) (70 mL), 2.5 mL of concentrated H2SO4 were
added, and the resulting solution was refluxed for 30 min and
then stirred for 12 h at room temperature. A saturated potas-
sium bicarbonate aqueous solution (250 mL) was added and
the mixture extracted with chloroform to remove all the diester
(Rf = 0.81; CH2Cl2–MeOH 95:5). The aqueous phase was acidi-
fied to pH 2 with aq. HCl (25%) and extracted again with
chloroform (4 × 150 mL). The combined organic phases were
dried over Na2SO4 and evaporated to give 2.6 g of 2a as a white
solid (61% yield). 1H NMR in CDCl3: δ 4.04 (3 H, s, CH3), 8.31
(1 H, d, 4J = 1.94) and 8.39 (1 H, d, 4J = 1.98 Hz). 13C NMR in
CDCl3: δ 43.75, 130.81, 132.71, 149.44, 153.12, 154.53, 167.88
Electronic spectra in the UV-vis range were recorded at 20 ЊC
with a Perkin-Elmer Lambda 7 spectrometer using 1.0 and 0.1
cm quartz cells, reflectance spectra as finely ground powders
dispersed in MgO (5%) with MgO as the reference on a Perkin-
Elmer Lambda 900 spectrometer equipped with a Labsphere
PELA-1000 integration sphere, IR spectra from KBr pellets
with a Mattson α-Centauri FT-IR spectrometer and ES-MS
spectra of the complexes on a Finnigan SSQ 710C spectrometer
on 10Ϫ4 M solutions in acetonitrile (capillary temperature set to
200 ЊC and acceleration potential to 4.5 kV). 1H and 13C NMR
spectra were recorded at 25 ЊC on Bruker AM-360 or AVANCE
400-DRX spectrometers. Chemical shifts are reported in parts
1
per million with respect to TMS. Longitudinal H relaxation
times T1 were measured by the inversion–recovery pulse
sequence. The experimental procedures for high resolution,
laser excited luminescence studies have been published previ-
ously.20,46 Emission spectra were corrected for the instrumental
functions, but high resolution excitation spectra were not.
Quantum yields of the ligand centred emission were measured
relative to quinine sulfate in diluted acidic solution (absolute
quantum yield: 0.546).47 Quantum yields of the metal-centred
emission were determined as previously described9c at excit-
ation wavelengths where (i) the Lambert–Beer law is obeyed
and (ii) the absorption of the reference [Eu(terpy)3]3ϩ (absolute
quantum yield 0.013)24 closely matches that of the sample.
Ligand excitation and emission spectra were recorded on a
Perkin-Elmer LS-50B spectrometer equipped for low temper-
ature (77 K) measurements. Elemental analyses were performed
by Dr H. Eder (Microchemical Laboratory, University of
Geneva).
and 168.68. IR ν(cmϪ1, KBr): 1707 (ν(C᎐O) ester), 1727
᎐
(ν(C᎐O) acid) and 1586 (ν(C᎐C) py). ESI-MS (MeOH): m/z
᎐
᎐
215.6, [M ϩ H]ϩ.
4-Bromo-6-methoxycarbonylpyridine-2-carboxylic acid 2b.
9.46 g (38 mmol) of compound 1b were dissolved in 200 mL of
a 1:1 water–MeOH mixture and 3 mL of concentrated H2SO4
were added. The resulting solution was refluxed until complete
dissolution of the carboxylic acid, stirred overnight at room
temperature and then poured into 250 mL of a saturated
NaHCO3 solution. The mixture was then extracted with CHCl3
to remove all the diester (Rf = 0.81, CH2Cl2–MeOH 95:5). The
aqueous phase was acidified slowly to pH 2 with aq. HCl (37%)
and extracted again with chloroform (5 × 100 mL). The com-
bined organic phase was dried over Na2SO4 and evaporated to
1
give 3.66 g of 2b as a white solid (37% yield). H NMR in
4
CDCl3: δ 4.05 (3 H, s, CH3), 8.05 (1 H, d, J = 1.94) and 8.32
(1 H, d, 4J = 1.98 Hz). ESI-MS (MeOH): m/z 259.8, [M ϩ H]ϩ.
IR ν(cmϪ1, KBr): 1706 (ν(C᎐O) ester), 1721 (ν(C᎐O) acid) and
Preparation of the ligands (Scheme 1)
᎐
᎐
4,4Ј-Bis(N-ethylamino)-3,3Ј-dinitrodiphenylmethane,48 diethyl
4-bromopyridine-2,6-dicarboxylate49 and LB 20 were prepared
as previously described.
1572 (ν(C᎐C) py).
᎐
4-Chloro-6-(N,N-diethylcarbamoyl)pyridine-2-carboxylic acid
3a. A 1.6 g (7.4 mmol) amount of compound 2a, freshly dis-
tilled thionyl chloride (35.35 g, 297 mmol), and DMF (100 µL)
were refluxed for 1 h in dry dichloromethane (30 mL) under an
inert atmosphere. The yellow solid obtained after evaporation
was pumped for 1 h (10Ϫ2 Torr) and dissolved in 30 mL of dry
dichloromethane. 5.43 g (74.2 mmol) of diethylamine were
added dropwise under an inert atmosphere. The solution
was refluxed for 90 min and, after cooling, 150 mL of a half
saturated NH4Cl solution were added. The organic phase was
separated and the aqueous phase extracted again with dichloro-
methane (2 × 100 mL). The combined organic phases were
washed twice with 100 mL of a saturated KHCO3 aqueous
solution, dried over Na2SO4 and evaporated. The residue was
dissolved in 100 mL of a 1 M KOH aqueous solution and
the mixture refluxed for 30 min. The aqueous solution was
extracted with dichloromethane (2 × 100 mL), acidified to pH 2
with aq. HCl (25%), and cooled at 0 ЊC for 12 h. The white
crystals were collected by filtration to give 1.6 g of 3a (6.2
mmol, yield 84%). 1H NMR in CDCl3: δ 1.26 (3H, t, 3J = 7.00),
4-Chloropyridine-2,6-dicarboxylic acid 1a. The product was
prepared by the method of Robison.50 Phenylphosphonic
dichloride (50.41 g, 259 mmol) was added to 4-hydroxy-
pyridine-2,6-dicarboxylic acid (13 g, 65 mmol), and the reaction
mixture heated at 120 ЊC for 2 h under an inert atmosphere.
After cooling, 200 mL of anhydrous methanol were added. The
mixture was stirred vigorously for 1 h, concentrated to dryness
and the residue dissolved in chloroform (200 mL). The solution
was washed twice with water and twice with a half-saturated
NaHCO3 aqueous solution. The organic phase was dried over
Na2SO4, filtered and evaporated to dryness. The residue was
re-crystallised from methanol to obtain 12.78 g of dimethyl
4-chloropyridine-2,6-dicarboxylate (86% yield). 1H NMR in
CDCl3: δ 4.00 (6 H, s) and 8.27 (2 H, s). 13C NMR in CDCl3:
δ 54.05, 128.74, 147.31, 149.94 and 164.65. IR ν (cmϪ1, KBr):
1728 (ν(C᎐O) ester). This ester (12.78 g, 55.6 mmol) was
᎐
refluxed for 1 h in 250 mL of 5 M aqueous NaOH. After
cooling, the solution was acidified to pH 2 with 25% aq. HCl.
The white precipitate formed was filtered off and dried under
vacuum to obtain 11 g of compound 1a (98% yield). 1H NMR
in CDCl3: δ 8.24 (2 H, s). 13C NMR in DMSO-d6: δ 130.39,
132.35, 149.38, 154.10 and 168.68. IR ν(cmϪ1, KBr): 1722
3
3
1.31 (3 H, t, J = 7.00), 3.39 (2 H, q, J = 7.00), 3.61 (2 H, q,
3J = 7.00), 7.90 (1 H, d, J = 1.76) and 8.24 (1 H, d, J = 1.76
Hz). 13C NMR in CDCl3: δ 13.21, 14.62, 42.00, 45.19, 127.18,
127.42, 147.95, 150.52, 157.39, 166.52 and 168.78. IR ν(cmϪ1,
4
4
2040
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
Scheme 1
KBr): 1607 (ν(C᎐O) amide), 1653 (ν(C᎐O) acid) and 1580
dried over anhydrous Na2SO4. The resulting product was dried
᎐
᎐
(ν(C᎐C) py). ESI-MS (MeOH): m/z 256.6, [M ϩ H]ϩ.
under vacuum (10Ϫ2 Torr) at 50 ЊC to get 2.5 g (7.9 mmol)
of methyl 4-bromo-6-(N,N-diethylcarbamoyl)pyridine-2-carb-
oxylate (yield 79%). 1H NMR in CDCl3: δ 1.25 (6 H, t,
3J = 7.02), 3.38 (2 H, q, 3J = 7.02), 3.53 (2 H, q, 3J = 7.02), 3.98
(3 H, s), 8.00 (1 H, d, 4J = 1.72) and 8.28 (1 H, d, 4J = 1.44 Hz).
The product was dissolved in 100 mL of a 1 M KOH aqueous
solution and the mixture refluxed for 2 h and acidified to pH
2 with aq. HCl (37%). A white precipitate formed which was
collected by filtration to give 2.27 g (7.53 mmol) of 3b (overall
᎐
4-Bromo-6-(N,N-diethylcarbamoyl)pyridine-2-carboxylic acid
3b. A 2.59 g (9.97 mmol) amount of compound 2b was dis-
solved in 60 mL of dry toluene under an inert atmosphere (N2)
and 6.45 g (30 mmol) of (COBr)2 and 100 µL of DMF were
added slowly. The resulting solution was vigorously stirred for
1 h and then heated to 70 ЊC for 2 h. The brown solid obtained
after evaporation was pumped for 1 h (10Ϫ2 Torr) and dissolved
in 30 mL of dry dichloromethane. 7.03 g (99.7 mmol) of diethyl-
amine were added dropwise under an inert atmosphere. The
solution was refluxed under nitrogen for 90 min and, after cool-
ing, 100 mL of CH2Cl2 were added. The solution was washed
with water (2 × 100 mL). The combined organic phases were
1
3
yield 76%). H NMR in CDCl3: δ 1.22 (3 H, t, J = 6.96), 1.29
(3 H, t, 3J = 7.16), 3.32 (2 H, q, 3J = 7.09), 3.58 (2 H, q,
3J = 7.09), 8.00 (1 H, d, J = 1.84) and 8.40 (1 H, d, J = 1.84
Hz). 13C NMR in CDCl3: δ 13.38, 15.10, 41.12, 43.99, 128.70,
131.10, 137.0, 146.6, 155.3, 163.3 and 166.2. ESI-MS (MeOH):
4
4
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2041

View Article Online
3
m/z 302.7, [M ϩ H]ϩ. IR ν(cmϪ1, KBr): 1598 (ν(C᎐O) amide),
3J = 6.97), 4.27 (1 H, s), 4.72 (2 H, q, J = 6.72), 7.24 (1 H, d,
᎐
3
4
1713 (ν(C᎐O) acid) and 1566 (ν(C᎐C) py).
3J = 8.32), 7.35 (1 H, d, J = 8.4), 7.52 (1 H, d, J = 1.72), 7.67
(1H, s), 8.42 (1 H, d, 4J = 1.8 Hz). 13C NMR in CDCl3: δ 13.42,
14.98, 16.07, 40.30, 41.26, 42.84, 43.48, 110.78, 120.86, 123.37,
125.64, 126.06, 135.62, 137.35, 143.71, 146.64, 148.73, 151.53,
᎐
᎐
Compound 4a. A 1.14 g (4.7 mmol) amount of compound 3a,
thionyl chloride (16.78 g, 141 mmol), and DMF (100 µL) were
refluxed for 90 min in dry dichloromethane (50 mL) under an
inert atmosphere. The solution was concentrated to dryness and
the residue dried under vacuum for 30 min. This was dissolved
in 30 mL of dry dichloromethane and a 30 mL solution of 4,4Ј-
bis(N-ethylamino)-3,3Ј-dinitrodiphenylmethane (0.80 g, 2.4
mmol) and triethylamine (47 mmol) was added dropwise. The
resulting solution was stirred at room temperature for 2 h and
refluxed during 4 h under an inert atmosphere. Upon evapor-
ation a brown residue was obtained. This was partitioned
between dichloromethane (3 × 100 mL) and half-saturated
NH4Cl (100 mL) solution. The combined organic phases were
dried over Na2SO4 and evaporated, and the resulting crude
residue was purified by column chromatography (silica gel;
CH2Cl2 to CH2Cl2–MeOH 99:1), to yield 1.50 g of 4a (78%
156.29 and 167.96. IR ν (cmϪ1, KBr): 1561 (ν(C᎐C)) and 1635
᎐
(ν(C᎐O)). ESI-MS (MeOH): m/z 725.3, [M ϩ H]ϩ.
᎐
Bis{1-ethyl-2-[4-bromo-6-(N,N-diethylcarbamoyl)pyridin-2-
yl]benzimidazol-5-yl}methane (LF). Freshly activated iron pow-
der (1.08 g, 19.3 mmol) and HBr (47%, 8.32 g, 48.7 mmol) were
added to a solution of compound 4b (0.59 g, 0.65 mmol) in
EtOH–water (111:28 mL). The mixture was refluxed for three
days under an inert atmosphere. The resulting solution was
cooled, filtered and evaporated to remove the ethanol.
Na2H2EDTAؒ2H2O (13.2 g) in 100 mL of water and 100 mL of
CH2Cl2 were added and the pH of the aqueous phase was
adjusted to 7 using aqueous NH4OH (12%). H2O2 (30%, 1 mL)
was added to the aqueous phase and the pH adjusted to 8.5.
The two phases were vigorously stirred before separation and
the aqueous phase was extracted with CH2Cl2 (6 × 100 mL).
The combined organic phases were dried over Na2SO4, filtered,
and evaporated to dryness, resulting in a yellow-orange solid
which was purified by column chromatography (silica gel;
CH2Cl2 to CH2Cl2–MeOH 99:1; Rf = 0.32 in CH2Cl2–MeOH
97:3) to give 0.35 g (0.43 mmol) of LF as a pale green solid
(yield 66%) [Found: C, 57.51; H, 5.25; N, 13.65. Calc. for
1
yield). H NMR in CDCl3: δ 0.78–1.42 (18 H, m), 2.81–3.68
(12 H, m), 4.01–4.42 (2 H, m) and 6.98–7.98 (10 H, m).
ESI-MS (MeOH): m/z 821.6, [M ϩ H]ϩ; and 843.5, [M ϩ Na]ϩ.
IR ν(cmϪ1, KBr): 1533 (ν(C᎐C)), 1349, 1568 (NO ), 1645
᎐
2
(ν(C᎐O)).
᎐
Compound 4b. A 1.60 g (5.32 mmol) amount of compound 3b
and 40 mL of dry toluene were mixed under an inert atmos-
phere (N2) and 3.48 g (16 mmol) of (COBr)2 and 100 µL of
DMF were slowly added. The resulting solution was vigorously
stirred for 1 h and then heated to 70 ЊC for 2 h. The solid
obtained after evaporation was pumped for 30 min (10Ϫ2 Torr)
and dissolved in 30 mL of dry dichloromethane. A solution of
4,4Ј-bis(N-ethylamino)-3,3Ј-dinitrodiphenylmethane(0.80g, 2.4
mmol) and triethylamine (47 mmol) in the same solvent (40
mL) was added dropwise. The resulting solution was stirred at
room temperature for 2 h, refluxed during 4 h under an inert
atmosphere and evaporated. The brown residue was partitioned
between dichloromethane (3 × 100 mL) and a half-saturated
NH4Cl (100 mL) solution. The combined organic phases were
dried over Na2SO4 and evaporated, and the resulting crude
residue was purified by column chromatography (silica gel;
CH2Cl2 to CH2Cl2–MeOH 98.5:1.5, Rf = 0.40 in CH2Cl2–
MeOH 97:3), to yield 0.60 g (0.66 mmol) of 4b (yield 24.8%).
1H NMR in CDCl3: δ 0.75–1.38 (18 H, m), 2.85–3.62 (12 H, m),
4.05–4.45 (2 H, m) and 7.03–8.01 (10 H, m). ESI-MS (MeOH):
m/z 911.2, [M ϩ H]ϩ; and 456.5, [M ϩ 2H]2ϩ. IR ν(cmϪ1, KBr):
1
C39H42Br2N8O2: C, 57.50; H, 5.20; N, 13.76%]. H NMR in
3
3
CDCl3: δ 1.10 (3 H, t, J = 7.04), 1.28 (3 H, t, J = 7.08), 1.43
(3 H, t, 3J = 6.98), 3.36 (2 H, q, 3J = 6.96), 3.59 (2 H, q,
3
3J = 6.97), 4.29 (1 H, s), 4.74 (2 H, q, J = 6.72), 7.26 (1 H, d,
3J = 8.32), 7.35 (1 H, d, J = 8.4), 7.68 (1 H, d, J = 1.72), 7.79
3
4
4
(1 H, s) and 8.60 (1 H, d, J = 1.8 Hz). ESI-MS (MeOH): m/z
815.2, [M ϩ H]ϩ; and 408.4, [M ϩ 2H]2ϩ. IR ν(cmϪ1, KBr):
1558 (ν(C᎐C)) and 1639 (ν(C᎐O)).
᎐
᎐
Preparation of the complexes
The perchlorate salts Ln(ClO4)3ؒnH2O (Ln = La to Lu, except
Pm; n = 6–8) were prepared from the corresponding oxides
(Rhône-Poulenc, 99.99%) according to the literature method.51
Eu(ClO4)3 was prepared from Eu(ClO4)3ؒnH2O (n ≈ 2) by
slowly increasing the temperature from 30 to 80 ЊC under
vacuum.52 CAUTION: perchlorate salts combined with organic
ligands are potentially explosive and should be handled in small
quantity and with the necessary precautions.53
1533 (ν(C᎐C)), 1345, 1566 (ν(NO )), 1644 (ν(C᎐O)).
᎐
᎐
2
[Ln2(LE)3][ClO4]6ؒnH2O (Ln ؍ La, Eu, Gd, Tb or Lu). A
22.97 µmol amount of Ln(ClO4)3ؒnH2O in dry acetonitrile (1.5
mL) was slowly added to a solution of LE (25 mg, 34.45 µmol)
in dry dichloromethane (1 mL). After stirring at room tem-
perature for 2 h the solution was evaporated, the solid residue
re-dissolved in acetonitrile (1 mL), solution was filtered, and
diethyl ether diffused into the solution for 2–3 days. The result-
ing pale yellow microcrystalline aggregates were collected
by filtration and dried to give the complexes [Ln2(LE)3]-
[ClO4]6ؒnH2O (Ln = La (n = 10), Eu (10), Gd (4), Tb (7) or Lu
(8) in 65–76% yields. The [Ln2(LE)3]6ϩ (Ln = Pr, Nd, Sm, Dy,
Ho, Er, Tm or Yb) complexes for NMR measurements were
prepared by the same procedure but not recrystallised.
Bis{1-ethyl-2-[4-chloro-6-(N,N-diethylcarbamoyl)pyridin-2-
yl]benzimidazol-5-yl}methane (LE). Freshly activated iron
powder (2.85 g, 51 mmol) and HCl (25%, 16.53 g, 118.5 mmol)
were added to a solution of compound 4a (1.30 g, 1.58 mmol)
in EtOH–water (220:56 mL). The mixture was refluxed over-
night under an inert atmosphere. The golden coloured solution
was cooled, filtered and evaporated to remove the ethanol.
Na2H2EDTAؒ2H2O (31.76 g, 85.32 mmol) in 100 mL of water
and 100 mL of CH2Cl2 were added to the mixture and the pH
of the aqueous phase was adjusted to 7 using aq. NaOH (25%).
H2O2 (30%, 1 mL) was added to the aqueous phase and the pH
adjusted to 8.5. The two phases underwent vigorous stirring
before separation and the aqueous phase was extracted with
CH2Cl2 (5 × 100 mL). The combined organic phases were dried
over Na2SO4, filtered, and evaporated to dryness, resulting in a
pale brown solid which was purified by column chromato-
graphy (silica gel; CH2Cl2 to CH2Cl2–MeOH 98:2, Rf = 0.48 in
CH2Cl2–MeOH 95:5) to give 0.95 g of LE as a pale yellow solid
(yield 83%) [Found: C, 64.35; H, 5.87; N, 15.29. Calc. for
Anhydrous [Eu2(LE)3][ClO4]6.
A 25 µmol amount of
anhydrous Eu(ClO4)3 in dry acetonitrile (1.5 mL) was slowly
added to a solution of LE (27.2 mg, 37.5 µmol) in dry dichloro-
methane (1 mL). After being stirred at room temperature for 2 h
the solution was evaporated to dryness. The solid residue was
dissolved in dry acetonitrile (1 mL), the solution filtered in a
glove-box, and dry THF diffused into the solution for 24 h. The
resulting pale yellow microcrystalline aggregates were collected
by filtration and dried to give [Eu2(LE)3][ClO4]6 (yield 80%).
1
C39H42Cl2N8O2: C, 64.55; H, 5.83; N, 15.44%]. H NMR in
3
3
CDCl3: δ 1.08 (3 H, t, J = 7.04), 1.26 (3 H, t, J = 7.08), 1.42
(3 H, t, 3J = 6.98), 3.35 (2 H, q, 3J = 6.96), 3.59 (2 H, q,
2042
J. Chem. Soc., Dalton Trans., 2000, 2031–2043

View Article Online
21 M. Elhabiri, R. Scopelliti, J.-C. G. Bünzli and C. Piguet, (a)
Acknowledgements
Chem. Commun., 1998, 2347; (b) J. Am. Chem. Soc., 1999, 121,
10747.
22 K. Sonogashira, Y. Tohda and N. Hagihara, Tetrahedron Lett., 1975,
50, 4467; H. A. Dieck and F. R. Heck, J. Organomet. Chem., 1975,
93, 259.
23 J. B. Lamture, Z. Hong Zhou, A. S. Kumar and T. G. Wensel, Inorg.
Chem., 1995, 34, 864.
This work is supported by grants from the Swiss National
Science Foundation and by the Swiss Office for Education and
Science (COST action D18). We thank Veronique Foiret
and Gaël Zucchi for their help in recording high resolution
luminescence spectra and lifetimes.
24 C. Platas, F. Avecilla, A. de Blas, C. F. G. C. Geraldes, T. Rodriguez-
Blas, H. Adams and J. Mahia, Inorg. Chem., 1999, 38, 3190.
25 S. Rigault, C. Piguet and J.-C. G. Bünzli, J. Chem. Soc., Dalton
Trans., 2000, DOI: 10.1039/b000147n.
26 S. Petoud, J.-C. G. Bünzli, K. J. Schenk and C. Piguet, Inorg. Chem.,
1997, 36, 1345.
27 K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Co-ordination Compounds, J. Wiley, New York, 3rd edn., 1972,
pp. 142–154.
28 I. Marolleau, J. P. Gisselbrecht, M. Gross, F. Arnaud-Neu and
M. J. Schwing-Weill, J. Chem. Soc., Dalton Trans., 1990, 1285.
29 J. A. Peters, J. Huskens and D. J. Raber, Prog. NMR Spectrosc.,
1996, 28, 283.
References
1 J.-C. G. Bünzli, in Rare Earths, eds. R. Saez-Puche and P. Caro,
Editorial Complutense, Madrid, 1998, p. 223ff.
2 R. B. Lauffer, in MRI Clinical Magnetic Resonance Imaging,
eds. R. R. Edelman, M. B. Zlatkin and J. R. Hesselink, W. B.
Saunders Co., Philadelphia, 2nd edn., 1996, vol. 1, ch. 5; P. Caravan,
J. J. Ellison, T. J. McMurry and R. B. Lauffer, Chem. Rev., 1999, 99,
2293.
3 T. C. Bruice, A. Tsubouchi, R. O. Dempcy and L. P. Olson, J. Am.
Chem. Soc., 1996, 118, 9867; P. Hurst, B. K. Takasaki and J. Chin,
J. Am. Chem. Soc., 1996, 118, 9982.
4 G. L. DeNardo, G. R. Mirik, L. A. Kroger, R. T. O’Donnel,
C. F. Meares and S. J. De Nardo, J. Nucl. Med., 1996, 37, 451;
M. J. Heeg and S. S. Jurisson, Acc. Chem. Res., 1999, 32, 1053.
5 J.-C. G. Bünzli, in Lanthanide Probes in Life, Chemical and Earth
Science, eds. J.-C. G. Bünzli and G. R. Choppin, Elsevier Science
Publ. B.V., Amsterdam, 1989, ch. 7.
6 Biomedical Applications of Labelling Technologies, eds. I. Hemmilä,
T. Sta˚hlberg and P. Mottram, Wallac Oy, Turku, 2nd edn., 1995;
G. Mathis, in Rare Earths, eds. R. Saez-Puche and P. Caro, Editorial
Complutense, Madrid, 1998.
7 S. Aime, A. Barge, M. Botta, J. A. K. Howard, R. Kataky, M. P.
Lowe, J. M. Moloney, D. Parker and A. S. de Sousa, Chem.
Commun., 1999, 1047.
8 N. Sabbatini, S. Perathoner, V. Balzani, B. Alpha and J.-M. Lehn,
in Supramolecular Chemistry, ed. V. Balzani, Reidel Publishing Co.,
Dordrecht, 1987.
30 A. D. Sherry and C. F. G. C. Geraldes, in Lanthanide Probes in Life,
Chemical and Earth Sciences, eds. J.-C. G. Bünzli and G. R.
Choppin, Elsevier Science Publ. B.V., Amsterdam, 1989, ch. 4.
31 C. N. Reilley, B. W. Good and J. F. Desreux, Anal. Chem., 1975, 47,
2110.
32 R. M. Golding and M. P. Halton, Aust. J. Chem., 1972, 25, 2577;
A. A. Pinkerton, M. Rossier and S. Stavros, J. Magn. Reson., 1985,
64, 420.
33 B. Bleaney, J. Magn. Reson., 1972, 8, 91.
34 J. Ren and D. Sherry, J. Magn. Reson., 1996, B111, 178.
35 M. D. Kemple, B. D. Ray, K. B. Lipkowitz, F. G. Prendergast and
B. N. D. Rao, J. Am. Chem. Soc., 1988, 110, 8275.
36 J. Lisowski, J. L. Sessler, V. Lynch and T. D. Moody, J. Am. Chem.
Soc., 1995, 117, 2273; J. Lisowski, Magn. Reson. Chem., 1999, 37,
287.
37 V. Clementi and C. Luchinat, Acc. Chem. Res., 1998, 31, 351;
S. Aime, M. Botta, D. Parker and J. A. G. Williams, J. Chem. Soc.,
Dalton Trans., 1995, 2259.
9 (a) G. Blasse, Phys. Status Solidi A, 1992, 130, K85; (b) J.-C. G.
Bünzli, P. Froidevaux and J. M. Harrowfield, Inorg. Chem., 1993, 32,
3306; (c) S. Petoud, T. Glanzmann, J.-C. G. Bünzli, C. Piguet, X. Qin
and R. P. Thummel, J. Luminesc., 1999, 82, 69.
38 S. T. Frey and W. de W. Horrocks, Jr., Inorg. Chim. Acta, 1995, 229,
384.
10 V. Alexander, Chem. Rev., 1995, 95, 273; J.-C. G. Bünzli, F. Ihringer
and F. Besançon, ACS Symp. Ser., 2000, 14.
11 J.-C. G. Bünzli, N. André, M. Elhabiri, G. Muller and C. Piguet,
J. Alloys Compd., 2000, 303/304, 66.
12 P. L. Jones, A. J. Amoroso, J. C. Jeffery, J. A. McCleverty,
E. Psillakis, L. H. Rees and M. D. Ward, Inorg. Chem., 1997, 36, 10;
N. Armaroli, G. Accorsi, F. Barigelletti, S. M. Couchman, J. S.
Fleming, N. C. Harden, J. C. Jeffery, K. L. V. Mann, J. A.
McCleverty, L. H. Rees, S. R. Starling and M. D. Ward, Inorg.
Chem., 1999, 38. 5769; F. Renaud, C. Piguet, G. Bernardinelli,
G. Hopfgartner and J.-C. G. Bünzli, Chem. Commun., 1999, 457;
C. Piguet, C. Edder, H. Nozary, F. Renaud, S. Rigault and J.-C. G.
Bünzli, J. Alloys Compd., 2000, 303/304, 94.
39 M. Latva, H. Takalo, V. M. Mukkala, C. Matachescu, J.-C.
Rodriguez-Ubis and J. Kankare, J. Luminesc., 1997, 75, 149.
40 F. J. Steemers, W. Verboom, D. N. Reinhoudt, E. B. Vandertol and
J. W. Verhoeven, J. Am. Chem. Soc., 1995, 117, 9408.
41 C. Platas, F. Avecilla, A. de Blas, T. Rodriguez-Blas, C. F. G. C.
Geraldes, E. Toth, A. E. Merbach and J.-C. G. Bünzli, J. Chem. Soc.,
Dalton Trans., 2000, 611.
42 G. Zucchi, R. Scopelliti, P.-A. Pittet, J.-C. G. Bünzli and R. D.
Rogers, J. Chem. Soc., Dalton Trans., 1999, 931.
43 J.-C. G. Bünzli, B. Klein, D. Wessner and N. W. Alcock, Inorg. Chim.
Acta, 1982, 59, 269; J.-C. G. Bünzli, G. A. Leonard, D. Plancherel
and G. Chapuis, Helv. Chim. Acta, 1986, 69, 288.
44 A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker,
L. Royle, A. S. de Sousa, J. A. G. Williams and M. Woods, J. Chem.
Soc., Perkin Trans. 2, 1999, 493.
13 J.-M. Lehn, Supramolecular Chemistry. Concepts and Perspectives,
VCH, Weinheim, 1995.
14 C. Piguet and J.-C. G. Bünzli, Chem. Soc. Rev., 1999, 28, 347.
15 D. E. Koshland, Jr., Angew. Chem., Int. Ed. Engl., 1994, 33, 2375.
16 (a) C. Piguet, E. Rivara-Minten, G. Hopfgartner and J.-C. G.
Bünzli, Helv. Chim. Acta, 1995, 78, 1541; (b) C. Piguet, J.-C. G.
Bünzli, G. Bernardinelli, G. Hopfgartner, S. Petoud and O. Schaad,
J. Am. Chem. Soc., 1996, 118, 6681; (c) C. Edder, C. Piguet, J.-C. G.
Bünzli and G. Hopfgartner, J. Chem. Soc., Dalton Trans., 1997, 4657.
17 C. Piguet, E. Rivara-Minten, G. Hopfgartner and J.-C. G. Bünzli,
Helv. Chim. Acta, 1995, 78, 1651; C. Piguet, E. Rivara-Minten,
G. Bernardinelli, J.-C. G. Bünzli and G. Hopfgartner, J. Chem. Soc.,
Dalton Trans., 1997, 421.
18 S. Rigault, C. Piguet, G. Bernardinelli and G. Hopfgartner, Angew.
Chem., Int. Ed. Engl., 1988, 27, 169.
19 C. Piguet, J.-C. G. Bünzli, G. Bernardinelli, G. Hopfgartner and
A. F. Williams, J. Am. Chem. Soc., 1993, 115, 8197.
20 N. Martin, J.-C. G. Bünzli, V. McKee, C. Piguet and G.
Hopfgartner, Inorg. Chem., 1998, 37, 577.
45 M. Hollenstein, Diploma Thesis, University of Lausanne, 1999.
46 C. Piguet, A. F. Williams, G. Bernardinelli, E. Moret and J.-C. G.
Bünzli, Helv. Chim. Acta, 1992, 75, 1697.
47 S. R. Meech and D. Phillips, J. Photochem., 1983, 23, 193.
48 C. Piguet, G. Bernardelli, B. Bocquet, A. Quattropani and A. F.
Williams, J. Am. Chem. Soc., 1992, 114, 7440.
49 H. Takalo and J. Kankare, Acta Chem. Scand., Ser. B, 1987, 41, 219;
B. Schmidt and D. Ehlert, Tetrahedron Lett., 1998, 39, 3999.
50 M. H. Robison, J. Am. Chem. Soc., 1958, 80, 5481; B. Pitteri,
L. Cattalini, G. Chessa, G. Marangoni, N. Stevanato and M. L.
Tobe, J. Chem. Soc., Dalton Trans., 1991, 3049.
51 J. F. Desreux, in Lanthanide Probes in Life, Chemical and Earth
Sciences, eds. J.-C. G. Bünzli and G. R. Choppin, Elsevier Science
Publ. B.V., Amsterdam, 1989, ch. 2.
52 J.-C. G. Bünzli, J.-R. Yersin and C. Mabillard, Inorg. Chem., 1982,
21, 1471.
53 W. C. Wolsey, J. Chem. Educ., 1973, 50, A335.
J. Chem. Soc., Dalton Trans., 2000, 2031–2043
2043