Lanthanide triple helical complexes with a chiral ligand derived from 2,6-pyridinedicarboxylic acid

Article abstract of DOI:10.1039/b102728j

The ligand 3-[2,6-bis(diethylcarbamoyl)pyridine-4-yl]-N-(tert-butoxycarbonyl)alanine methyl ester (L⁴) bearing a chiral group in the 4-position of the pyridine ring has been synthesised and its interaction with trivalent lanthanide ions studied. L⁴ yields stable [Ln(L⁴)₃]³⁺ complexes in acetonitrile, with logβ₃ in the range 19-20. The specific rotary dispersion of the complexes is about ten times as large as that of the ligand alone, consistent with triple helical structures in solution. NMR data show the presence of only one time-averaged species in acetonitrile solution with a trigonal symmetry, that is, the helical P ? M interconversion is fast on the NMR time scale. Circularly polarised luminescence on the Eu and Tb triple helical complexes displays weak effects, pointing to only a small diastereomeric excess in solution. High resolution luminescence spectra of the Eu complex in the solid state reveal a local symmetry derived from a trigonal arrangement around the metal ion and the long ⁵D₀ lifetime (1.58 ms) is indicative of the absence of water molecules bound in the inner co-ordination sphere. Energy transfer processes in the luminescent Eu and Tb triple helical complexes are discussed.

Full text of DOI:10.1039/b102728j

View Article Online / Journal Homepage / Table of Contents for this issue
Lanthanide triple helical complexes with a chiral ligand
derived from 2,6-pyridinedicarboxylic acid†
Gilles Muller,a Boris Schmidt,b,c Jan Jiricek,c Gérard Hopfgartner,d James P. Riehl,e
Jean-Claude G. Bünzli*a and Claude Piguetf
a Institute of Inorganic and Analytical Chemistry, BCH, University of Lausanne,
CH-1015 Lausanne, Switzerland. E-mail: jean-claude.bunzli@icma.unil.ch
b Novartis Pharma AG, Nervous System, WKL 136.5.82, CH-4002 Basel, Switzerland
c Department of Organic Chemistry, University of Hannover, Schneiderberg 1b,
D-30167 Hannover, Germany
d F. Hoffmann-La Roche Ltd, Pharmaceuticals Division, PRNS 68/142, CH-4070 Basel,
Switzerland
e Department of Chemistry, University of Minnesota Duluth, Minnesota, 55812-2496, USA
f Department of Inorganic, Analytical and Applied Chemistry, University of Geneva,
30 quai E. Ansermet, CH-1211 Geneva 4, Switzerland
Received 26th March 2001, Accepted 18th June 2001
First published as an Advance Article on the web 22nd August 2001
The ligand 3-[2,6-bis(diethylcarbamoyl)pyridine-4-yl]-N-(tert-butoxycarbonyl)alanine methyl ester (L4) bearing a
chiral group in the 4-position of the pyridine ring has been synthesised and its interaction with trivalent lanthanide
ions studied. L4 yields stable [Ln(L4)3]3ϩ complexes in acetonitrile, with logβ3 in the range 19–20. The specific rotary
dispersion of the complexes is about ten times as large as that of the ligand alone, consistent with triple helical
structures in solution. NMR data show the presence of only one time-averaged species in acetonitrile solution with
a trigonal symmetry, that is, the helical P
M interconversion is fast on the NMR time scale. Circularly polarised
luminescence on the Eu and Tb triple helical complexes displays weak effects, pointing to only a small diastereomeric
excess in solution. High resolution luminescence spectra of the Eu complex in the solid state reveal a local symmetry
derived from a trigonal arrangement around the metal ion and the long 5D0 lifetime (1.58 ms) is indicative of the
absence of water molecules bound in the inner co-ordination sphere. Energy transfer processes in the luminescent
Eu and Tb triple helical complexes are discussed.
Dipicolinic acid (H2dipic) and its derivatives are also good
candidates for the design of triple-stranded Ln() building
Introduction
The considerable interest in lanthanide co-ordination chem-
blocks with interesting luminescence properties. For instance,
istry, which has steadily risen during the last two decades, is
the optically-active [Ln(dipic)3]3Ϫ complexes are strongly lumin-
essentially motivated by applications in biology and medicine,
escent in water and circularly polarised luminescence shows
mainly for analysis1 and diagnosis purposes2 but, also, for
that the enantiomeric excess in the ground state can be modu-
therapeutic uses.3 One challenge pertaining to the design of
lated by the addition of chiral sugars.9 On the other hand, we
functional lanthanide-containing co-ordination compounds
lies in the need for a precise control of the Ln() inner co-
have shown that the weakly stable [Ln(L2)3]3ϩ complexes exist as
a mixture of rapidly interconverting P
M conformers with
ordination sphere: these ions present non-directional bonding
and large and variable co-ordination numbers.4 Various
strategies have been developed to master this difficulty, for
instance the encapsulation of the spherical Ln() ions into
pre-organised macrocyclic ligands or into cavities built from
pre-disposed molecules or through the self-assembly of care-
fully tailored multidentate ligands and podands.5,6 In the latter
strategy a finer recognition can be achieved if weak non-
covalent inter-strand interactions are used to stabilise the func-
tional edifices. Working along these lines, we have shown that
bis(benzimidazole)pyridines (bzimpy, Scheme 1) are versatile
tridentate aromatic donors acting as building blocks for various
monotopic and ditopic receptors aimed at designing triple
helical monometallic and bimetallic (4f–4f and 3d–4f ) func-
tional molecular edifices with light-converting or spin-crossover
capability,5,7 or as candidates for rod-like extended materials.8
Ln = Eu, Tb displaying fair quantum yields in acetonitrile,
while [Ln(L1)3]3ϩ complexes are more stable but less lumin-
escent.10,11 The triple helical complexes are inherently chiral
but they usually appear as racemic mixtures both in the solid
state and in solution (often as fast interconverting isomers).
However, a small excess of one helical enantiomer can be
induced in the excited state via the use of circularly polarised
excitation light.12 Related helical induction in the ground state
requires diastereomeric resolution and we recently have intro-
duced a chiral substituent in the R2 position of bzimpy to give
L*. However, the bulky neopentyl substituent precludes the
helical wrapping of the third ligand, resulting in a very low
stability of the [Ln(L*)3]3ϩ complexes.13 In this paper we turn to
the substituted pyridine-2,6-dicarboxamide ligand L4, where
a chiral group is grafted onto the 4-position of the pyridine
ring, which should lead to (i) less steric hindrance and (ii) easier
helical wrapping of the three ligand strands around the Ln()
ions, thus providing pairs of diastereoisomers for the triple
helical complexes [Ln(L4)3]3ϩ.
† Electronic supplementary information (ESI) available: further analy-
sis of the structures Ln(ClO4)3ؒxH2O and [Ln(L4)3]3ϩ. See http://
www.rsc.org/suppdata/dt/b1/b102728j/
DOI: 10.1039/b102728j
J. Chem. Soc., Dalton Trans., 2001, 2655–2662
This journal is © The Royal Society of Chemistry 2001
2655

View Article Online
Scheme 2
ics calculations using the AM1 package (Fig. S1†).21 At room
temperature, the H- and 13C-NMR spectra display 11 and 13
1
Scheme 1
signals, respectively, pointing to a local two-fold symmetry for
the pyridine dicarboxamide portion of the molecule: X3,5, X7,7Ј,
X8,8Ј, X9,9Ј and X10,10Ј (X = H, C) and C2,6 give rise to only one
signal each.
Results and discussion
Synthesis and characterisation of the ligand and its complexes
The 1 : 3 complexes were obtained in 70–80% yield by adding
L4 dissolved in dichloromethane to solutions of lanthanide per-
chlorates in a mixture of CH3CN and CH2Cl2. Complexation is
evidenced by the shift of the νCO vibration to lower energy
(∆ν = Ϫ29 cmϪ1 for La to Ϫ22 cmϪ1 for Lu), reflecting a lower-
ing in the electron density of the carbonyl double bond and/or a
larger reduced mass associated with the complexation to LnIII.
All perchlorate ions are ionic, so that the complexes are most
probably 9-co-ordinate. Single crystals for X-ray structure
determination could not be obtained.
Ligand L4 was synthesised from 4-bromo-N,N,NЈ,NЈ-tetra-
ethylpyridine-2,6-dicarboxamide 2 via a palladium mediated
reaction with Jackson’s zinc reagent 3 (Scheme 2). Despite
previous discouraging reports on the activity of (Ph3P)2-
PdCl2,14,15 this catalyst proved to be better than (PPh3)4Pd
or Pd(OAc)2 and gave L4 in excellent yield (95%). Takalo’s
16
approach17,18 was used to prepare 2 by treatment of chelidamic
acid with phosphorous pentabromide and hydrolysis of the
intermediate acid bromide with diethylamine. Jackson’s zinc
reagent 3 was synthesised from N-Boc-serine14,15 using a single-
step procedure to get iodoalaninemethyl ester by a combination
of I2/PPh3/imidazole reagents. This method gave a better yield
(61%)19 and made isolation of the final product easier than with
the originally proposed method.14 This was then followed by a
Knochel’s zinc activation reaction20 (BrCH2CH2Br, trimethyl-
silylchloride) which was revealed to be the method of choice,
provided that the recommended zinc dust was replaced by zinc
turnings to allow reagent transfer via cannulation. Slow reac-
tion rates and excess of zinc have to be avoided to prevent
palladium(0) precipitation and decomposition of iodoalanine
into di-dehydroalanine. The best results were obtained with zinc
amounts reduced to one third of the reported reaction con-
ditions.20 Other methods of zinc activation such as ultrasound,
CuCN, HgCl2, DMA or Rieke zinc led to inferior results. The
replacement of THF by DMF, DMA, toluene or benzene or
combinations thereof was also unsuccessful.
Interaction between L4 and Ln(III) ions
To quantify this interaction, we first measured ES-MS spectra
in order to identify the various species in solution. Solutions
of L4 (10Ϫ4 M in anhydrous acetonitrile) were titrated with
Ln(ClO4)3ؒxH2O (x = 0.3–0.6) at 298 K for ratios R = [L4]t/
[Ln()]t = 0–3. The data show that the Boc substituent is easily
split (FW = 101 Da) in the gas-phase, giving rise to a primary
amine L4Ј. As a result, the spectra reflect the presence of
[Ln(L4)n]3ϩ (n = 1–4), [Ln(L4Ј)m]3ϩ (m = 2–4) and [Ln(L4)p-
(L4Ј)q]3ϩ (p = 1, q = 1–2; p = 2, q = 1) species, some of them with
solvation molecules and/or in the form of perchlorate adducts.
All the identified species are listed in Table S1†. These data
clearly support the successive formation of 1 : 1, 1 : 2 and 1 : 3
species, as well as of a species with four ligands. Such species
have been observed for similar systems and were assigned to
outer-sphere association of a fourth ligand with a 1 : 3 edifice.10
It is noteworthy that the proportion of the various species is
dependent upon the ionic radius of the metal ion (Table 1). For
R = 3, the 1 : 3 complex is the most abundant one for the three
investigated ions, while solutions containing the larger La()
ions contain sizeable amounts of the 1 : 4 species, in line with
the explanation given above. On the other hand, the perchlorato
adducts [Ln(L4)3(ClO4)]2ϩ form more easily with cations which
The specific rotary dispersion of L4 in anhydrous and
degassed acetonitrile, [α]2D5 = Ϫ3.5 0.4 deg dm2 molϪ1, con-
firms the chirality arising from the asymmetric carbon. 1H- and
13C-NMR spectra in the same solvent, including COSY, DEPT-
135, HSQC and NOE experiments, clearly point to L4 having a
trans–trans conformation of the terdentate binding units in
agreement with that predicted by gas-phase molecular mechan-
2656
J. Chem. Soc., Dalton Trans., 2001, 2655–2662

View Article Online
Table 1 Relative intensities of the ES-MS peaks for solutions with [L4]t/[Ln()]t = 3 (CH3CN, 298 K)
Species
La
Eu
Lu
Species
La
Eu
0.0
19.1
0.0
0.0
0.0
6.5
Lu
0.0
27.0
0.0
0.0
8.1
0.0
[Ln(L4)2]3ϩ
55.6
100
40.7
7.4
7.4
1.8
4.7
100
8.8
0.0
35.3
1.5
0.0
100
10.8
1.3
37.8
0.0
[Ln(L4)(L4Ј)]3ϩ
14.8
0.0
16.7
5.9
0.0
0.0
[Ln(L4)3]3ϩ
[Ln(L4)(L4Ј)2]3ϩ
[Ln(L4)4]3ϩ
[Ln(L4)2(L4Ј)]3ϩ
[Ln(L4)2(ClO4)]2ϩ
[Ln(L4)3(ClO4)]2ϩ
[Ln(L4)4(ClO4)]2ϩ
[Ln(L4)2(L4Ј)2]3ϩ
[Ln(L4)(L4Ј)(ClO4)]2ϩ
[Ln(L4)(L4Ј)2(ClO4)]2ϩ
Table 2 Cumulative stability constants logβn (298 K,
2σ) for
[Ln(Li)n]3ϩ species (i = 1, 3, 4; n = 1–3) as determined under anhydrous
conditions in acetonitrile by spectrophotometry
Ligand
L3
Ln
logβ1
logβ2
logβ3
La
Eu
Lu
La
Eu
Lu
La
Eu
Lu
7.4 0.4
8.2 0.4
8.7 0.4
7.4 0.4
8.2 0.4
8.7 0.4
7.4 0.3
8.3 0.3
8.1 0.3
14.0 0.5
14.5 0.5
15.3 0.5
13.9 0.5
14.6 0.5
15.2 0.5
14.8 0.3
15.3 0.3
15.2 0.3
19.0 0.5
19.8 0.5
20.3 0.5
19.0 0.5
19.7 0.5
20.5 0.5
21.0 0.3
22.3 0.3
22.9 0.3
L4
L1 a
a From ref. 11.
Fig. 1 1H-NMR spectra of L4 and [Ln(L4)3]3ϩ (5 × 10Ϫ3 M in CD3CN
at 298 K); asterisk and small circle (•) correspond to signals from the
solvent and water, respectively.
have a larger charge density (Eu, Lu) than with La. This
speciation has been confirmed in solution by monitoring the
titration of L4 with La() by 1H-NMR: the spectra of the three
inner-sphere complexes could be identified and when R = 3, the
1 : 3 species is the main complex in solution (>92%, see Fig.
S2†). The apparent easy formation of the 1 : 3 species contrasts
with what was observed with L* and points to the substitution
in the R1 position of the central pyridine not preventing the
formation of the helical complexes.
hindrance, which is comparable in both L3 and L4, destabilises
somewhat the complexes with the smaller lanthanide ions.
Solution structure and dynamic behaviour of the [Ln(L4)3]3؉
complexes
To quantify further the Ln()–L4 interaction, we performed
spectrophotometric titrations of both L3 and L4 against
Ln(Otf )3 (Ln = La, Eu, Lu; Otf = trifluoromethanesulfonate) in
MeCN under anhydrous conditions ([H2O] < 30 ppm). The
ligand L3 is an achiral analogue of L4. Factor analysis indicated
the presence of three or four absorbing molecules and the data
were fitted by a least-squares procedure using either set (1)–(3)
of the following equilibria, where solvation and anion co-
ordination has been omitted, or a model including [Ln(Li)4]3ϩ
chosen since this species was found in the ES-MS spectra.
Increasing the number of complex species however did not
improve the fit, in line with the postulated outer-sphere nature
of this 1 : 4 species.
1
The H-NMR spectra of the 1 : 3 complexes are displayed in
Fig. 1 and chemical shifts are reported in Table 3. At room
temperature, these spectra point to a trigonal symmetry of the
complexes with equivalent ligand strands. The chemical shifts
induced by the complexation are in line with those reported for
complexes with L1 and, together with NOE effects observed
between H3–H9 are characteristic of a meridional co-ordination
of the tridentate binding unit of each ligand strand. Moreover,
the large downfield shifts of the signals of H3,5 and C3,5 indicate
co-ordination of the pyridine N atom to the Ln() ions.
Analysing in more detail the spectra of the diamagnetic com-
plexes with La and Lu, we find substantial differences for the
protons of the N-ethyl substituents and of the Boc group. For
instance, H7 and H9 appear as a single broadened quadruplet
(4H) for La while they give rise to two resolved quadruplets
with similar intensities (2H) for Lu. An immediate conclusion
is that the rotation around the C–N amide bond is slower in
the Lu structure, as previously observed for [Ln(L1)3]3ϩ and
ascribed to the larger charge density of Lu() which hinders
the amide rotation process by reducing the electron density of
the co-ordinated carbonyl group.10 Moreover, the equivalence
of the geminal protons of the methylene probes implies that
Ln3ϩ ϩ Li
[Ln(Li)]3ϩ logβ1
[Ln(Li)2]3ϩ logβ2
[Ln(Li)3]3ϩ logβ3
(1)
(2)
(3)
Ln3ϩ ϩ 2 Li
Ln3ϩ ϩ 3 Li
Electronic spectra of the three complexes were correlated,
which rendered the fitting process difficult and generated rel-
atively large uncertainties in the logβi values. The results are
reported in Table 2, along with comparative data for L1. Stab-
ility constants for L3 and L4 are identical, within experimental
errors, indicating that the introduction of the NH Boc on C12
does not affect complex formation. If logβ1 and logβ2 are the
same, within experimental errors, for the three ligands L1,3,4, a
different pattern is observed for logβ3. The triple helical species
with ligands L3 and L4 are somewhat less stable than those with
L1, the difference increasing from La to Lu, as shown by
∆logβ3(L1 Ϫ Li) = 2 for La and 2.4–2.6 for Lu (i = 3, 4). There-
fore, the poor electron attracting group in R1 has a minor
destabilising effect as measured for Ln = La, and its steric
fast P
M helical interconversion occurs at room temper-
ature on the NMR time scale along the complete lanthanide
series. Differences appear also for the H14 protons, one signal
being observed for La and three for Lu; this is also indicative of
a slowing down of the rotation process of the Boc group in the
case of the smaller lanthanide ion.
A variable temperature study of the Lu solution (234–340 K,
Fig. S3†) shows that the H14 protons appear as two singlets
of approximately the same intensity at elevated temperature
while they give rise to at least 6–7 signals at 242 K, two of them
being substantially more intense than the others. The hindered
rotation about the OC–N bond of the carbamate groups in
J. Chem. Soc., Dalton Trans., 2001, 2655–2662
2657

View Article Online
Table 3 1H and 13C chemical shifts (ppm) for L4 and its [Ln(L4)3]3ϩ complexes with Ln = La, Pr, Eu and Lu (CD3CN, 5 × 10Ϫ3 M, 298 K)
H3,5
H7,7
Ј
H8
H9,9
Ј
H10
H11,11
Ј
H12
H13
H14
H15
L4
La
Pr
Eu
Lu
7.35
7.84
10.14
5.87
3.23
3.29
1.80, 2.49
3.67, 4.15
3.12, 3.29
1.07
0.98
Ϫ0.74
1.27
3.49
3.68
3.70, 4.30
3.67
3.74
1.19
1.37
1.52
1.68
1.42
3.00, 3.22
3.19, 3.45
3.26, 3.46
3.01, 3.44
3.20, 3.49
4.46
4.64
5.58
4.15
4.62
5.70
5.84
6.55
5.52
5.87
1.34
1.30
1.35
1.21
1.25
1.30
1.34
3.68
3.74
4.04
3.59
3.74
7.90
0.89
C3,5
C7
C8
C9
C10
C11
C12
C14
C15
Cquat.
L4
La
Pr
Eu
Lu
123.20
127.47
142.60
95.25
39.48
42.45
42.41
41.79
43.52
12.14
11.43
9.49
12.80
11.45
42.85
44.68
46.07
42.48
44.85
13.59
13.35
12.84
14.70
13.35
36.45
37.01
39.81
33.33
36.83
53.82
53.33
52.99
54.87
53.33
27.44
27.48
27.48
27.39
27.39
51.97
52.32
52.67
52.18
52.35
149.05, 154.31, 167.77, 171.80
149.77, 153.62, 169.28, 171.27
152.66, 157.77, 169.28, 171.89
143.62, 162.68, 163.23, 171.03
147.77, 153.53, 168.30, 171.35
128.07
[Ln(L4)3]3ϩ provides four different conformers for each dia-
stereoisomer, defined by the cis or trans position of the H13
atom with respect to the neighbouring O carbonyl atom. For a
given diastereoisomer, we predict that H14 produces one signal
in the 1H-NMR spectrum for the cis–cis–cis (symmetry C3), one
signal for the trans–trans–trans (symmetry C3), three signals for
the cis–trans–cis (symmetry C1) and three signals for the trans–
trans–cis conformer (symmetry C1) in the statistical ratio
1 : 3 : 3 : 1 if no steric or electronic constraints are present.
Therefore the H14 protons can theoretically generate 8 signals
for each diastereoisomer, that is a total of 16 signals if both
diastereoisomers are present. However, in our case, examin-
ation of the signals arising from the other nuclei (H and C)
clearly points to the solutions containing one major diastereo-
isomer only, as several other studies on complexes between
d- and f-transition ions and potentially chiral ligands have
shown.22–24 We were however unable to assign the observed sig-
nals and to exploit quantitatively the spectral changes as a func-
tion of temperature. The rotation of the N,N Ј-diethylamide and
Boc groups is also seen in the variable temperature spectra of
the Pr and Eu complexes. At 332 K, proton H14 gives rise to one
signal only whereas 5 peaks are observed at 234 K. However
two of the conformers have much larger populations than the
other ones, as demonstrated by the 13C-NMR spectra which
evidence the splitting of the C14 signal at 298 K into two com-
ponents at 244 K (Fig. S4†). In the case of Pr, the rotation
process is considerably slowed down with respect to La, and we
were able to determine the kinetic parameters for the inter-
conversion between the two major conformers (275–298 K)
rotary dispersion measured for the ligand alone. Therefore,
there is no doubt that [α]2D5 contains a large structural contribu-
tion, pointing to the formation of triple helical complexes
in solution. As a comparison, [α]2D5 is approximately equal to
one or two times the rotary dispersion of the ligand in 1 : 1 and
1 : 2 complexes with L*, for which no (1 : 1) or little, (1 : 2)
structural contribution is expected.12 Circularly polarised
luminescence (CPL) spectra for the Eu and Tb complexes are
reported in Fig. 2. They display very small ∆I values for the
25
25
5
5D0 7F1(Eu) and D4 7F5(Tb) transitions. The CPL spectrum
D
D
of the Eu complex depends upon the polarisation of the excita-
tion light, indicating the presence of two species in solution,24 in
line with the speciation that can be calculated from the stability
constants: 10Ϫ3 M solutions contain 92% of the 1 : 3 species and
8% of the 1 : 2 species, which is expected to generate a smaller
chiral effect. The CPL spectrum of the 5D4 7F5(Tb) transition
displays several peaks corresponding to crystal-field splitting of
the electronic levels.24 The calculated glum factor amounts to
0.02 only, a value substantially smaller than those reported for
chiral complexes with trigonal symmetry.23 These CPL data are
consistent with the presence of only a small excess of one dia-
stereoisomer in solution on the short time scale of optical
measurements.
Photophysical properties
The electronic spectrum of the ligand in solution displays an
intense band centred around 51000 cmϪ1 corresponding to a
transition mainly centred on the carbonyl groups. Its intensity
considerably increases upon complexation, indicating a strong
interaction between the carbonyl functions and the metal ions
(Table 4, Fig. S5†). In addition, two weak and broad bands are
seen around 41150 and 37460 cmϪ1, assigned to transitions
mainly centred on the pyridine groups.11 Upon complexation,
the energy gap between these two transitions increases from
3700 cmϪ1 in the free ligand to ca. 6500 cmϪ1 in the complexes.
using the Eyring equation: ∆G‡ = 57.7
0.6 kJ molϪ1
,
,
∆H‡ = 45.8 0.3 kJ molϪ1, ∆S‡ = Ϫ40.0 1.0 J molϪ1 KϪ1
and k(298 K) = 479 11 sϪ1. These parameters correspond to
the most energetic process going on in solution. For T < 275 K,
the spectra point to the presence of other conformers the inter-
conversion of which is faster. Although the introduction of
the chiral group in R1 together with the use of an amide-
type spacer between the asymmetric carbon and the Boc
group result in complications in the 1H-NMR spectra, the
In the solid state, the gap becomes larger, being ca. 10000 cmϪ1
.
No significant emission is detected upon irradiation of L4 in the
UV region (solution and solid state), pointing to the existence
of efficient non-radiative de-activation pathways in the free
ligand. On the other hand, ligand emission is observed for the
solution behaviour of the [Ln(L4)3]3ϩ complexes is very similar
to that previously reported for [Ln(L1)3]3ϩ
except for the
11
increased rate of the dynamic process involving P
M
helical interconversion in [Ln(L4)3]3ϩ
.
complexes. Fluorescence from a short-lived ππ* state occurs
1
To assess whether or not the ligand strands are helically
for all complexes while ligand phosphorescence from a long-
3
wrapped around the metal ion, we measured the specific rotary
lived ππ* state is seen for the La and Lu complexes only, an
dispersion of solutions of the [Ln(L4)3]3ϩ complexes (10Ϫ3
M
efficient 3ππ*-to-Ln energy transfer process taking place for Eu
and Tb complexes and resulting in metal-centred emission
exclusively (Fig. 3). The 0-phonon transitions observed for
emissions from both the singlet and triplet states are slightly
shifted toward lower energy when going from La to Lu, consist-
ent with the larger charge density of the heavier Ln() ion. The
in anhydrous acetonitrile). This experiment occurs on a
time scale where P
M helical interconversion is blocked
in [Ln(L4)3]3ϩ. The specific rotary dispersion amounts to
[α]2D5 = Ϫ35, Ϫ30, Ϫ28, and Ϫ31 deg dm2 molϪ1 for La, Eu,
Tb, and Lu, respectively and is about ten times larger than the
2658
J. Chem. Soc., Dalton Trans., 2001, 2655–2662

View Article Online
Table 4 Ligand-centred absorptions in acetonitrile and in the solid state (293 K), ligand-centred singlet- and triplet-state energies as determined
from emission spectra of the solids (77 K) and of solutions (10Ϫ3 M in acetonitrile, 293 K) for the ligand L4 and its triple helical complexes
[Ln(L4)3]3ϩ a
b
E/cmϪ1
E(1ππ*)/cmϪ1
E(3ππ*)/cmϪ1
Compound
L4
Solutionc
Solid state
Solution
Solid stated
Solid stated
e
e
e
50010 (4.36)
41150 (3.71)
37460 (3.65)
49750 (4.92)
42320 (4.49) sh
35820 (4.26)
35060 (4.23) sh
49140 (4.91)
41270 (4.50) sh
35850 (4.31)
34850 (4.26) sh
49140 (4.89)
42220 (4.45) sh
35630 (4.25)
34940 (4.24) sh
49260 (4.86)
42000 (4.46) sh
35760 (4.21)
34900 (4.17) sh
40330 sh
36900
45870 sh
36360
[La(L4)3]3ϩ
[Eu(L4)3]3ϩ
[Tb(L4)3]3ϩ
[Lu(L4)3]3ϩ
31990
25575
25190
18345
26110
23150
24180
23140
24210
f
f
43860 sh
37175
25445
f
f
46510 sh
35970
25445
44385 sh
35000
25000
30745
25760
23200
17610
a
b
c
sh = shoulder. Ligand-centred transitions. At the concentration used, 92% of the metal ions are in the form of the 1 : 3 complex and 8% in
the form of the 1 : 2 complex; log ε values are given in parentheses. d The 0-phonon transition is given in italics for frozen solutions in acetonitrile.
e Too weak to be observed. f Not observed because of the L4-to-Ln energy transfer process.
5
Fig. 2 CPL spectra of [Ln(L4)3]3ϩ 10Ϫ3 M in anhydrous CH3CN at 298 K, upon excitation at 320 nm. Left: D0 7F1(Eu) transition, right:
5D4 7F5(Tb) transition.
5
phosphorescence decay is mono-exponential and the corre-
sponding lifetimes amount to 240 2 and 181 5 ms for La
and Lu, respectively.
the D0 7F0 transition to the total nephelauxetic effect gener-
erated by the ligands bound to the EuIII ion, we predict a value
of 17230 cmϪ1 for this transition if the following parameters
are taken into account: δN(py) = Ϫ16.726 and δCO = Ϫ15.7.25 The
discrepancy is relatively large and may arise from the fact that
δN(py) has been estimated for polyaminocarboxylates.26 More-
over, the experimental trend observed in the nephelauxetic
effect between the triple helical complexes with L1 and L4 is not
consistent with the presence of an electron attracting group in
the 4-position of the pyridine, which induces less stable 1 : 3
complexes and is expected to produce a somewhat smaller
nephelauxetic effect. Nephelauxetic parameters are extremely
sensitive to bond lengths and it might well be that the smaller
interaction between N(py) and the metal ion in [Eu(L4)3]3ϩ causes
the Ln–O distances to become longer so that the increased
Despite the low intensity metal-centred emission in [Eu-
(L4)3]3ϩ (see Fig. S6†), we have been able to use Eu() as a
structural probe to investigate the chemical environment of this
metal ion in a solid sample of the triple helical complex by high
resolution luminescence. At 13 K, the 5D0 7F0 excitation
spectrum displays one broad and slightly asymmetrical band
located at 17211 cmϪ1 with a full width at half height (fwhh)
equal to 16.6 cmϪ1. This band shifts to 17221 cmϪ1 at room
temperature while becoming sharper (fwhh = 11.6 cmϪ1). As a
comparison, the same transition has been observed at 17227
cmϪ1 for the corresponding complex with L1.11 Using the corre-
lation proposed by Frey and Horrocks25 relating the energy of
J. Chem. Soc., Dalton Trans., 2001, 2655–2662
2659

View Article Online
Fig. 4 Schematic energy diagram for [Ln(L4)3]3ϩ complexes. Data for
S1 and T1 are those of the La() complex in the solid state at 77 K
(0-phonon transitions).
To get a better insight into the various energy conversion
processes occurring in the [Ln(L4)3]3ϩ complexes, we deter-
mined the quantum yields of the ligand- and metal-centred
luminescence of 10Ϫ3 M solutions at 295 K. The ligand itself is
essentially non-luminescent and is only weakly luminescent in
the complexes with La (QF = 0.055%) and Lu (QF = 0.065%).
The slightly larger quantum yield of the Lu complex with
respect to the La complex (factor 1.2) may be explained by a
better intersystem crossing (isc) efficiency in the La complex
due to a more favourable energy difference: ∆E(1ππ* Ϫ 3ππ*) =
6800 cmϪ1 (as determined from the energy of the 0-phonon
transitions) as compared to 7545 cmϪ1 for the Lu complex,
Fig. 3 Top: Time-resolved emission spectra of [Ln(L4)3]3ϩ in solid
state at 77 K and recorded with time delays of 0.01 (La, Lu) and 0.1 ms
(Eu, Tb). Bottom: Emission spectra of [Ln(L4)3]3ϩ 10Ϫ3 M in CH3CN at
295 K.
while it is known that an efficient isc process takes place when
3
∆E(1ππ* Ϫ ππ*) is around 5000 cmϪ1 27
.
This in line with the
observed decrease in the phosphorescence intensity in going
from La to Lu, as exemplified by the ratio I(3ππ*)/I(1ππ*)
which decreases from 8 × 10Ϫ3 in [La(L4)3]3ϩ to 4.7 × 10Ϫ3 in
[Lu(L4)3]3ϩ, that is by a factor 1.7. The quantum yields of the
metal-centred luminescence obtained upon ligand excitation
nephelauxetic effect of the six carboxamide groups largely
overcomes the decreased effect of the three pyridine moieties.
The emission spectrum recorded at 13 K upon excitation
through the ligand bands displays relatively broad bands and is
typical of a low symmetry environment for the metal ion (Fig.
S6†). The relative intensities of the 5D0 7FJ transitions amount
to 0.09, 1.00, 6.7, <0.01, and 1.4 for J = 0, 1, 2, 3, and 4, respect-
ively. The splitting of the 7F1 level (346, 373 and 428 cmϪ1) with
two relatively close sublevels reflects a symmetry derived from a
are small, QLn = 0.22 and 1.2% for Eu and Tb, respectively.
L
They may be explained by the weak efficiency of the isc process,
the emission spectra being dominated by the fluorescence band
of the ligand, and by the relative energies of the triplet state of
5
5
the complexed ligand strands and of the excited D0 and D4
level (Fig. 4). Indeed, ∆E(3ππ* Ϫ 5D0) is very large (almost 8000
5
cmϪ1) while ∆E(3ππ* Ϫ D4), 4850 cmϪ1, is closer to the ideal
value of 2500–3500 cmϪ1 for an efficient ligand-to-metal energy
trigonal arrangement, but more distorted than in [Eu(L1)3]3ϩ
.
transfer process.27 With respect to [Ln(L1)3]3ϩ 11
the quantum
,
The Eu(5D0) lifetime determined upon broad-band (ligand) or
selective (5D0 7F0) excitation decreases from 1.45 0.02 ms
between 13 and 77 K to 1.15 0.02 ms at room temperature,
pointing to temperature-dependent quenching mechanisms we
may trace back to the presence of the fluxional substituent R1
since this lifetime is somewhat larger (1.8 ms) and temperature-
yield for Eu is 26 times larger which reflects the lowering of the
ligand triplet state by about 2250 cmϪ1 (as determined from
data for the La complexes) induced by the electron-attracting
substituent in R1.
independent in [Eu(L1)3]3ϩ 11
Nonetheless, the measured life-
.
Conclusion
time clearly points to the water molecules evidenced in the
elemental analysis and in the vibrational spectrum not being
co-ordinated to the metal ion. The lifetime of a 10Ϫ3 M solution
The introduction of a bulky chiral substituent in the pyridine
4-position of L1 to give L4 results in a tridentate ligand capable
of interacting strongly with trivalent lanthanide ions in
acetonitrile resulting in thermodynamically stable 1 : 3 com-
plexes with logβ3 in the range 19–20. Compared with complexes
with the unsubstituted ligand L1, the loss of stability due to the
introduction of the electron-attracting group in R1 is relatively
small and steric effect seems to be minimum, as indicated by
∆logβ3(L1 Ϫ L4) which increases only slightly from La (2.0) to
Lu (2.6). In solution, the NMR spectra are consistent with an
average pseudo-trigonal symmetry of the [Ln(L4)3]3ϩ com-
in acetonitrile, 1.58
0.02 ms, also indicates the absence of
inner-sphere interaction with water, that is partial decomplex-
ation of one ligand arm through binding of a water molecule
does not occur, consistent with the NMR data. The case of the
Tb() complex is somewhat puzzling in that the lifetime
observed in the solid state is quite short, 0.92 ms, but
temperature-independent in the range 13–295 K, whilst it is
much longer in solution, 1.55
0.01 ms. The latter value
is again consistent with no inner-sphere interaction with water
while the shorter value in the solid state may arise from
self-quenching processes in a closely packed solid.
plexes, but fast P
M helical interconversion on the NMR
time scale prevents diastereomeric segregation in solution.
2660
J. Chem. Soc., Dalton Trans., 2001, 2655–2662

View Article Online
However, chiro-optical data clearly demonstrate the helical
wrapping of the ligand strands around the Ln() ions on a
shorter time scale, contrary to what was observed when bulky
neopentyl substituents were grafted onto the non co-ordinating
N atoms of bis(benzimidazole)pyridine to yield L*.13 Although
only a very small excess of one diastereoisomer is induced
in solution by the chiral ligand L4, as shown by CPL data,
consequent design of ligands similar to L4, but producing
dynamically inert triple-helical complexes together with a larger
diastereomeric induction should lead to luminescent stains able
to probe the chirality of more complex molecules, e.g. bio-
molecules, for instance by preferential interaction with one
enantiomer. The photophysical properties of the [Ln(L4)3]3ϩ
complexes can be smoothly rationalised in terms of the energy
of the ligand and metal ion excited states and therefore modifi-
cation of the R1 substituent leading to a better antenna effect
should be within reach.
K from 10Ϫ4 M solutions in acetonitrile (spectroscopy grade)
containing Et4NClO4 0.1 M as inert electrolyte with a Perkin-
Elmer Lambda 7 spectrometer connected to an external com-
puter and using quartz cells of 0.1 cm path length. Solutions
were prepared in a thermostatted vessel (Metrohm 6.1418.220)
and the titrant solution was added with an automated burette
from Metrohm (6.1569.150 or .210) fitted with an anti-diffusion
device. In a typical experiment, 5–10 mL of L4 were titrated
with a solution of Ln() triflate 10Ϫ4 M in acetonitrile. After
each addition of 0.20 mL and a delay of 2 min, the spectrum
was measured and transferred to the computer. In some
instances, reverse titrations, i.e. with L4 as titrating species, were
also performed to confirm the values of the stability constants.
All the experiments were conducted in a dry box and the water
content of the solutions, measured at the end of the titration by
the Karl Fischer method, was always <30 ppm. Factor analysis
and stability constant determinations were carried out with the
program SPECFIT, version 2.10.34
Experimental
Preparation of the ligand (Scheme 2)
4-Bromopyridine-2,6-dicarbonyl dibromide (1). PBr3 (9.3 cm3,
98.9 mmol) was added slowly to a vigorously stirred solution of
bromine (4.2 cm3, 81.9 mmol) in petroleum ether (35 cm3). The
solvent was removed in vacuo after complete precipitation of
PBr5. Chelidamic acid (5 g, 27.3 mmol) was added and the
reaction mixture was heated to 90 ЊC under reflux for 4 h. The
residue was cooled to rt and extracted three times with CHCl3
(75 cm3). The combined organic layers were concentrated. The
residue was purified by Kugelrohr distillation (160–185 ЊC/0.5
mmHg) to give 8.8 g of 1 (21.8 mmol, 80%) as a white solid.
1H-NMR (400 MHz) in CDCl3: δ 8.38 (2H, s). 13C-NMR (100
Solvents and starting materials
Acetonitrile, dichloromethane, chloroform, tetrahydrofuran,
diethyl ether and petroleum ether were purified in the usual
way.28 Silica gel (Merk 60, 0.04–0.06 mm) was used for prepar-
ative column chromatography. Other products were purchased
from Fluka AG (Buchs, Switzerland) or Merck and used with-
out further purification. Lanthanide perchlorates were obtained
as previously described.29 CAUTION: dry perchlorates and their
complexes with aromatic amines may easily explode and should
be handled in small quantities and with extreme caution.30
MHz) in CDCl3: δ 130.78, 135.90, 150.54, 166.74. IR ν (cmϪ1
,
Spectroscopic and analytical measurements
KBr): 1763, 1246, 908.
Electronic spectra in the UV-vis range were recorded at 293 K
with a Perkin-Elmer Lambda 900 spectrometer using 1.0 and
0.1 cm quartz cells. Reflectance spectra were recorded as finely
ground powders dispersed in MgO (5%), with MgO as the
reference, on the same spectrometer equipped with a Labsphere
PELA-1000 integration sphere. Specific rotary dispersion
values were measured from 10Ϫ3 M solutions in degassed
anhydrous acetonitrile at 298 K with the help of a JASCO DIP-
370 polarimeter (sodium D line). IR spectra were obtained
from KBr pellets with a Mattson α-Centauri FT-IR spectro-
meter. Pneumatically-assisted electrospray (ES-MS) mass
spectra were recorded from CH3CN solutions on API III or
API 3000 tandem mass spectrometers (PE Sciex) by infusion at
10 µl minϪ1. The spectra were recorded under low up-front
declustering or collision induced dissociation (CID) conditions,
typically ∆V = 0–30 V between the orifice and the first quad-
rupole of the spectrometer. 1H and 13C-NMR spectra were
recorded at 25 ЊC on Bruker AM-360 or Bruker AVANCE 400-
DRX spectrometers. Chemical shifts are reported in parts per
million with respect to TMS. Ligand excitation and emission
spectra were recorded on a Perkin-Elmer LS-50B spectrometer
equipped for low temperature (77 K) measurements. The
experimental procedures for high resolution laser excited lumin-
escence studies have been published previously.31 Emission
spectra were corrected for the instrumental function. Quantum
yields of the ligand-centred emission were measured relative to
quinine sulfate in 0.05 M H2SO4 (A347 = 0.05, absolute quantum
yield: 0.546).32 Quantum yields of the metal-centred emission
were determined as previously described33 at excitation wave-
lengths where (i) the Lambert–Beer law is obeyed and (ii) the
absorption of the reference [Ln(terpy)3]3ϩ closely matches that
of the sample. Elemental analyses were performed by Dr H.
Eder (Microchemical Laboratory, University of Geneva).
4-Bromo-N,N,NЈ,NЈ-tetraethylpyridine-2,6-dicarboxamide
(2). 1 (2.2 g, 5.9 mmol) was dissolved in diethyl ether (50 cm3)
followed by the addition of diethylamide (2.5 cm3, 23.9 mmol)
at 0 ЊC. After stirring at 0 ЊC for ten minutes the reaction mix-
ture was stirred at rt for 20 min. The solvent was removed in
vacuo and the residue was crystallised from acetylacetate to give
1.7 g of 2 (4.8 mmol, 81%) as a colourless solid. Mp 92–94 ЊC.
1H-NMR (400 MHz) in CDCl3: δ 1.17 (6 H, t, 3J = 7.2 Hz), 1.27
3
3
(6 H, t, J = 7.2 Hz), 3.34 (4 H, q, J = 7.2 Hz), 3.56 (4 H, q,
3J = 7.2 Hz), 7.81 (2 H, s). 13C-NMR (100 MHz) in CDCl3:
δ 12.72, 14.24, 40.27, 43.27, 127.10, 134.53, 154.50, 166.70.
IR ν(cmϪ1, KBr): 1628, 1564, 1460, 1444. MS (FAB, 70 eV)
m/z: 356, 358 (M ϩ 1).
3-[2,6-Bis(diethylcarbamoyl)pyridine-4-yl]-N-(tert-butoxy-
carbonyl)alanine methyl ester (L4). Two 10 cm3 round-bottom
flasks were dried, one of them charged with zinc (216.6 mg, 3.31
mmol). Dibromoethane (0.016 cm3, 0.23 eq) was added under
inert atmosphere and warmed to 60 ЊC for 3 min. Tetrahydro-
furan (3 cm3) and trimethylsilyl chloride (5 cm3, 0.05 eq) were
added after cooling the dibromethane mixture to 30 ЊC and
stirred for 30 min, prior to heating to 60 ЊC and addition of
iodoalanine (326.8 mg, 0.993 mmol). Stirring was continued
until complete consumption of iodoalanine as judged by thin
layer chromatography. The second flask was loaded with 2
(295 mg, 0.828 mmol) and (Ph3P)2PdCl2 (30.7 mg, 0.044 mmol)
under argon. The zinc reagent (3) was transferred with a syringe
into the second flask and the mixture was warmed to 50 ЊC and
stirred for 24 h. Evaporation of the solvent left a brown oil
which was purified by flash chromatography (SiO2, diethyl
ether : petroleum ether 2 : 1) to give 377 mg of L4 (0.788 mmol,
95%) as a colourless solid. Mp 87–90 ЊC. [Found: C, 60.0; H,
8.0; N 11.4. Calc. for C24H38O6N4: C, 60.2; H, 8.0; N, 11.7%].
Spectrophotometric titrations
3
1H-NMR (400 MHz) in CD3CN: δ 1.07 (6 H, t, J = 7.15 Hz),
The electronic spectra in the UV-vis range were recorded at 298
1.19 (6 H, t, 3J = 7.15 Hz), 1.34 (9 H, s), 3.00 (1 H, m), 3.22 (1 H,
J. Chem. Soc., Dalton Trans., 2001, 2655–2662
2661

View Article Online
6 C. Piguet, C. Edder, H. Nozary, F. Renaud, S. Rigault and J.-C. G.
Bünzli, J. Alloys Compd., 2000, 303/304, 94.
m), 3.23 (4 H, q, 3J = 7.15 Hz), 3.49 (4 H, q, 3J = 7.15 Hz), 3.68
(3 H, s), 4.46 (1 H, m), 5.70 (1 H, d, 3J = 8.3 Hz), 7.35 (2 H, s).
13C-NMR (400 MHz) in CD3CN: δ 12.14, 13.59, 27.44, 36.45,
39.48, 42.85, 51.97, 53.82, 123.20, 149.05, 154.31, 167.77,
171.80. IR ν(cmϪ1, KBr): 1744, 1712, 1628. MS m/z: 478 (Mϩ,
16%), 351 (M Ϫ Boc Ϫ C2H2ϩ, 100%).
7 C. Piguet, C. Edder, S. Rigault, G. Bernardinelli, J.-C. G. Bünzli
and G. Hopfgartner, J. Chem. Soc., Dalton Trans., 2000, 3999.
8 J.-M. Bénech, C. Piguet, G. Bernardinelli, J.-C. G. Bünzli and
G. Hopfgartner, J. Chem. Soc., Dalton Trans., 2001, 684; H. Nozary,
C. Piguet, J.-P. Rivera, P. Tissot, G. Bernardinelli, N. Vulliermet,
J. Weber and J.-C. G. Bünzli, Inorg. Chem., 2000, 39, 5286.
9 E. Huskowska and J. P. Riehl, Inorg. Chem., 1995, 34, 5615.
10 F. Renaud, C. Piguet, G. Bernardinelli, J.-C. G. Bünzli and
G. Hopfgartner, Chem. Eur. J., 1997, 3, 1660.
11 F. Renaud, C. Piguet, G. Bernardinelli, J.-C. G. Bünzli and
G. Hopfgartner, Chem. Eur. J., 1997, 3, 1646.
12 G. Muller, PhD Dissertation, University of Lausanne, 2000.
13 G. Muller, J.-C. G. Bünzli, K. J. Schenk, C. Piguet and
G. Hopfgartner, Inorg. Chem., 2001, 40, 2642.
14 R. F. W. Jackson, N. Wishart, A. Wood, K. James and M. Wythes,
J. Org. Chem., 1992, 57, 3397.
15 S. Gair, R. F. W. Jackson and P. A. Brown, Tetrahedron Lett., 1997,
38, 3059.
16 J. M. Fox, X. H. Huang, A. Chieffi and S. L. Buchwald, J. Am.
Chem. Soc., 2000, 122, 1360.
17 H. Takalo, P. Pasanen and J. Kankare, Acta Chem. Scand. Ser. B,
1987, 41, 219.
18 H. Takalo, P. Pasanen and J. Kankare, Acta Chem. Scand. Ser. B,
1988, 42, 373.
19 J. A. Bajgrowicz, A. El Hallaoui, R. Jacquier, C. Pigiere and
P. Viallefont, Tetrahedron Lett., 1985, 41, 1833.
20 S. AchyuthaRao and P. Knochel, J. Org. Chem., 1991, 56, 4591.
21 M. J. Dewar, E. V. Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am.
Chem. Soc., 1985, 107, 3902.
22 W. Zarges, J. Hall, J.-M. Lehn and C. Bolm, Helv. Chim. Acta, 1991,
74, 1843.
23 N. Coruh, G. L. Hilmes and J. P. Riehl, Inorg. Chem., 1988, 27, 3647.
24 J. P. Riehl and F. S. Richardson, Chem. Rev., 1986, 86, 1.
25 S. T. Frey and W. deW. Horrocks Jr., Inorg. Chim. Acta, 1995, 229,
383.
26 M. Latva and J. Kankare, J. Coord. Chem., 1998, 43, 121.
27 F. J. Steemers, W. Verboom, D. N. Reinhoudt, E. B. Vandertol and
J. W. Verhoeven, J. Am. Chem. Soc., 1995, 117, 9408.
28 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, Pergamon Press, Oxford, 1988.
Preparation of the complexes
The 1 : 3 complexes [Ln(L4)3](ClO4)3ؒsolv (Ln = La, solv = 0, 4;
Eu, solv = 2.5 H2O, 5; Tb, solv = 3 H2O, 1 CH3CN, 6; Lu,
solv = 0, 7) were prepared by mixing L4 with stoichiometric
amounts of Ln(ClO4)3ؒnH2O (Ln = La, Eu, Tb, Lu and n = 0.2–
0.6) in dichloromethane–acetonitrile. The complexes could be
directly isolated in 70–80% yield by removing the solvent. The
slow diffusion of a solvent in which the complexes are not
soluble, led to the formation of oils. 4; Found: C, 46.0; H, 6.3;
N, 9.0. Calc. for C72H114N12O30Cl3La: C, 46.2; H, 6.1; N, 9.0%.
5; Found: C, 44.8; H, 6.2; N, 8.7. Calc. for C72H114N12-
O30Cl3Euؒ2.5H2O: C, 44.8; H, 6.2; N, 8.7%. 6; Found: C, 44.7;
H, 6.2; N, 9.2. Calc. for C72H114N12O30Cl3Tbؒ3H2OؒCH3CN: C,
44.7; H, 6.3; N, 9.2%. 7; Found: C, 45.3; H, 6.0; N, 8.9. Calc. for
C72H114N12O30Cl3Lu: C, 45.3; H, 6.0; N, 8.8%. IR ν(cmϪ1, KBr):
1599 (La), 1601 (Eu), 1602 (Tb), 1606 (Lu) (νCO); 1089, 625
(ionic ClO4Ϫ).
Acknowledgements
This work is supported through grants from the Swiss National
Science Foundation. We thank Véronique Foiret for her help in
recording the high resolution luminescence spectra, Dennis
Ehlert for improving the ligand synthesis, Paula Gawryszewska
for measuring the CPL spectra, Prof. Pierre Vogel for the use of
his polarimeter and the Fondation Herbette (Lausanne) for the
gift of spectroscopic equipment.
29 J.-C. G. Bünzli, J.-R. Yersin and C. Mabillard, Inorg. Chem., 1982,
21, 1471.
References
30 W. C. Wolsey, J. Chem. Educ., 1973, 50, A335.
31 N. Martin, J.-C. G. Bünzli, V. McKee, C. Piguet and G. Hopfgartner,
Inorg. Chem., 1998, 37, 577.
1 D. Parker, Coord. Chem. Rev., 2000, 205, 109.
2 V. W. W. Yam and K. K. W. Lo, Coord. Chem. Rev., 1999, 184,
157.
32 S. R. Meech and D. C. Phillips, J. Photochem., 1983, 23, 193.
33 H.-R. Mürner, E. Chassat, R. P. Thummel and J.-C. G. Bünzli,
J. Chem. Soc., Dalton Trans., 2000, 2809.
34 H. Gampp, M. Maeder, C. J. Meyer and A. D. Zuberbühler, Talanta,
1985, 23, 1133.
3 Z. J. Guo and P. J. Sadler, Angew. Chem., Intl. Ed., 1999, 38,
1513.
4 J.-C. G. Bünzli, in Rare Earths, ed. R. Saez Puche and P. Caro,
Editorial Complutense, Madrid, 1998, pp. 223–259.
5 C. Piguet and J.-C. G. Bünzli, Chem. Soc. Rev., 1999, 28, 347.
2662
J. Chem. Soc., Dalton Trans., 2001, 2655–2662