Importance of the chromophore orientation to the ligand-to-metal energy transfer in lanthanide complexes with pendant-arm fitted cyclen derivatives

Article abstract of DOI:10.1039/b101312m

Comparative solid state structural studies of three lanthanide macrocyclic complexes derived from 1,4,7,10-tetraazacyclododecane with pendant arms bearing amide co-ordinating groups have been performed in order to evaluate the parameters influencing the co-ordination polyhedron and to assess the importance of the geometric factor in energy transfer processes. In all the investigated structures the co-ordination geometry is a mono-capped twisted square antiprism, a situation commonly observed for similar compounds. High resolution luminescence spectra of the europium complexes are consistent with a tetragonal site symmetry for the metal ion. An analysis of the presented crystal structures and of previously reported ones indicates that (i) the relative orientation of the O4 and N₄ planes is not determined by the co-ordinated solvent molecule and (ii) the twist angle between them is mainly dictated by theflexibility of the pendant arms. Interpretation of the luminescence properties of the complexes of Sm(III), Eu(III) and Tb(III) can be made from the structural parameters found in the solid state and in solution (by NMR spectroscopy). Both the energy parameter (i.e. the gap between the ligand triplet state and the metal ion excited state) and the geometric parameter (i.e. the donor-acceptor distance and the orientation of the choromophore) have to be taken into account to explain the results obtained in terms of the efficiency of the L→Ln intra-molecular energy transfer. Furthermore, the correlative comparison between structural and luminescent properties shows how inter-molecular interactions in the solid state can be a prominent factor in the effectiveness of this transfer.

Full text of DOI:10.1039/b101312m

View Article Online / Journal Homepage / Table of Contents for this issue
Importance of the chromophore orientation to the ligand-to-metal
energy transfer in lanthanide complexes with pendant-arm fitted
cyclen derivatives†
Gaël Zucchi, Rosario Scopelliti and Jean-Claude G. Bünzli
Institute of Inorganic and Analytical Chemistry, University of Lausanne, CH-1015 Lausanne,
Switzerland. E-mail: Jean-Claude.Bunzli@icma.unil.ch
Received 9th February 2001, Accepted 11th May 2001
First published as an Advance Article on the web 18th June 2001
Comparative solid state structural studies of three lanthanide macrocyclic complexes derived from 1,4,7,10-
tetraazacyclododecane with pendant arms bearing amide co-ordinating groups have been performed in order to
evaluate the parameters influencing the co-ordination polyhedron and to assess the importance of the geometric
factor in energy transfer processes. In all the investigated structures the co-ordination geometry is a mono-capped
twisted square antiprism, a situation commonly observed for similar compounds. High resolution luminescence
spectra of the europium complexes are consistent with a tetragonal site symmetry for the metal ion. An analysis of
the presented crystal structures and of previously reported ones indicates that (i) the relative orientation of the O4
and N4 planes is not determined by the co-ordinated solvent molecule and (ii) the twist angle between them is
mainly dictated by the flexibility of the pendant arms. Interpretation of the luminescence properties of the
complexes of Sm(III), Eu(III) and Tb(III) can be made from the structural parameters found in the solid state and
in solution (by NMR spectroscopy). Both the energy parameter (i.e. the gap between the ligand triplet state and
the metal ion excited state) and the geometric parameter (i.e. the donor–acceptor distance and the orientation of
the choromophore) have to be taken into account to explain the results obtained in terms of the efficiency of the
L→Ln intra-molecular energy transfer. Furthermore, the correlative comparison between structural and luminescent
properties shows how inter-molecular interactions in the solid state can be a prominent factor in the effectiveness
of this transfer.
We are interested in designing stable luminescent complexes
and in unravelling the relationship between structural and
Introduction
Ligands based on the cyclen framework (cyclen is 1,4,7,10-
photophysical properties of lanthanide-containing edifices.
tetraazacyclododecane) are ideal complexation agents for tri-
Along these lines, we have recently described the properties of a
valent lanthanide ions. Besides stabilising the complexes by the
tetra-amide ligand bearing benzyl moieties (L1, see Scheme 1)
macrocyclic effect, they fulfil the requirement of Ln(III) for
large co-ordination numbers.1 The resulting edifices, strongly
stable even in aqueous media,2 are therefore particularly
appropriate for analytical and/or medical uses, for instance as
NMR shift reagents,3 magnetic resonance imaging contrast
agents,4–6 radiolabels,7 or radiotracers.8 Some of the complexes,
especially with the earlier lanthanides, have demonstrated a
catalytic activity in the specific cleavage of RNA and DNA.9,10
Several luminescent lanthanide-containing chemosensors have
recently been developed, based on cyclen complexes and in
which modulation of the emission occurs via ligand- or metal-
centred processes.11 The design of such edifices requires a good
command of the structural factors influencing their stability
and photophysical properties and many structural studies
have been reported on complexes with cyclen derivatives
bearing pendant arms functionalized by carboxylate,12–15
phosphinate16–18 and amide19–24 co-ordinating groups. In these
compounds the ion is the common vertex of two square
pyramids, defined by O4, Ln and by N4, Ln and the resulting co-
ordination geometry is a twisted square antiprismatic arrange-
ment, often completed by a solvent molecule, which caps the
plane defined by the oxygen atoms.
Scheme 1
† Electronic supplementary information (ESI) available: selected bond
and of its lanthanide complexes.25 In this paper we pursue the
lengths and angles, intermolecular hydrogen bonds in [Eu(L2)(H2O)]3ϩ
excitation, emission and NMR spectra. See http://www.rsc.org/
suppdata/dt/b1/b101312m/
,
analysis of similar systems by varying the length of the arms
and/or the nature of the chromophoric groups in L2, L3 and L4
DOI: 10.1039/b101312m
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
This journal is © The Royal Society of Chemistry 2001
1975

View Article Online
Fig. 1 Molecular structure and atom-numbering scheme for [Eu(L2)-
(H2O)]3ϩ
.
in order to determine the parameters influencing and control-
ling the photophysics of the complexes. We also hope to gain a
better understanding of the energy migration processes occur-
ring between the ligand and the lanthanide centre by comparing
geometrical parameters with luminescence data.
Fig. 2 Molecular structure and atom-numbering scheme for [Eu(L3)-
(CH3OH)]3ϩ
.
macrocycle adopts the stable [3333] conformation usually
observed for twelve-membered rings with the nitrogen atoms at
the corners.27 The co-ordination polyhedron displays a twisted
square antiprismatic arrangement of the four oxygen atoms
and the four nitrogen atoms, with an average twist angle τ of
38.1Њ (compared to 0 and 45.0Њ for ideal square prismatic and
antiprismatic arrangements, respectively). This leads to a
distorted mono-capped antiprismatic geometry around the
Eu(III). The amide hydrogen atoms are involved in hydrogen
bonds with two triflate counter ions, one disordered half-triflate
and one free water molecule. There is a network of hydrogen
bonds connecting two neighbouring molecules via the co-
ordinated water molecule (O5), the free water molecules Ow2,
Ow3, Ow4 and the two ordered triflates (Table 1; Fig. S1, ESI).
Thus, the crystal packing consists of a series of dimers.
Results and discussion
Ligands L2–4 have been obtained in 70–80% yield by treating the
appropriate N-substituted bromoacetamide with cyclen in THF
(L2) or DMF (L3,4) in the presence of an excess of triethylamine.
The complexes were prepared in the usual way25 by reaction of
equimolar amounts of anhydrous lanthanide triflate (trifluoro-
methanesulfonate) and ligand. They were recrystallised from
acetonitrile (L2) or methanol (L3,4) and gave satisfactory elem-
ental analyses. Complexation of the four arms through the
carbonyl functions is reflected by the ν(C᎐O) vibration which is
᎐
shifted towards lower energy by ca. 20 cmϪ1 upon complex-
ation. Crystals suitable for structure determination could only
be obtained with L2 and L3.
The complex [Eu(L3)(CH3OH)][CF3SO3]3ؒ2CH3OH, EuL3,
crystallises with three triflate counter ions and two molecules of
Solid state structures
¯
methanol. The crystals belong to the P1 space group in the
The complex [Eu(L2)(H2O)][CF3SO3]3ؒ4H2O, EuL2, crystallises
with four interstitial water molecules, two ordered triflate
counter ions and two disordered half-triflates in the asymmetric
unit. Ligand L2 acts as an octadendate ligand and the metal ion
completes its co-ordination sphere with one water molecule
thus generating a nine-co-ordinate geometry. The atom num-
bering scheme is shown in Fig. 1. The Eu(III) is sandwiched
between two almost parallel planes (interplanar angle = 0.4(1)Њ)
defined by the four oxygen and the four nitrogen atoms. The
average bond distance between the Eu(III) and the four amide
oxygen atoms amounts to 2.379(5) Å, while the mean Ln–
N(macrocycle) distance is longer, at 2.680(6) Å; the Ln–
O(water) bond length is standard, at 2.431(5) Å, being very
close to the one observed for similar tetra-amide cyclen deriv-
atives.22,23,26 The europium ion lies closer to the plane defined by
the oxygen atoms, as indicated by the distances to the mean O4
and N4 planes: 0.708(3) and 1.673(3) Å, respectively. This situ-
ation is partly due to the small cavity of the macrocycle and,
possibly, to the harder nature of the O donor atoms. The dis-
tances between the trans nitrogen atoms, defining the size of the
cavity, are 4.161(8) and 4.21(1) Å for N1 ؒ ؒ ؒ N3 and N2 ؒ ؒ ؒ N4,
respectively. These values are too short for “in-plane”
co-ordination of the europium ion.‡ The average values of the
torsion angles NCCN in the macrocycle and NCCO in the
pendant arms are ϩ59.4(9) and Ϫ35.1(10)Њ, respectively, show-
ing that the complex has crystallised in the Λ(δδδδ) form. The
triclinic system, as do the europium,12 gadolinium,28 holmium,14
yttrium,29 and lutetium30,31 complexes formed with DOTA
(1,4,7,10-tetraazacyclododecanetetraacetate),
a
europium
tetra(carboxyethyl) derivative of cyclen,13 and the dysprosium
complex of DTMA (1,4,7,10-tetrakis(methylcarbamoyl-
methyl)-1,4,7,10-tetraazacyclododecane).24 One methanol mol-
ecule is co-ordinated to the Eu(III) by capping the square plane
of the co-ordinated oxygen atoms (Fig. 2). The metal ion is
positioned 0.831(5) Å under the mean O4 plane whereas it lies
1.624(6) Å above the N4 plane. The two planes are almost paral-
lel, the angle between them being 0.6(2)Њ. The average Eu–O
and Eu–N distances are respectively 2.414(9) and 2.636(12) Å,
while the Eu–O (CH3OH) distance is 2.451(9) Å. The angle
between the O4 plane and the O13–Eu1 bond of the co-
ordinated methanol molecule is 91.2Њ. The average values of the
NCCN and NCCO torsion angles are both negative, Ϫ56.9(15)Њ
and Ϫ22.6(18)Њ, pointing to a Λ(λλλλ) conformation. The dis-
tances between the trans nitrogen atoms are 4.14(2) and 4.16(2)
for N1 ؒ ؒ ؒ N3 and N2 ؒ ؒ ؒ N4, respectively, while the twist angle
between the four oxygen and the four nitrogen atoms (25.5Њ) is
indicative of a twisted square prismatic arrangement. As for
EuL2, the crystal packing consists of dimers (Fig. 3) and is
mainly determined by three types of hydrogen bonds (Table 1):
(i) the first interaction occurs between interstitial methanol
molecules and the NH groups of two amide functions, (ii) the
two other amide H atoms interact with triflate counter ions, and
(iii) two [Eu(L3)(CH3OH)]3ϩ molecules, related by an inversion
centre, are linked by one nitro group of one cationic complex
and interacting with a co-ordinated methanol molecule of the
second one.
‡ 4.18 Å < (2r(Eu3ϩ) ϩ 2r(N)), with r(Eu3ϩ) = 1.12
Å for a co-
ordination number of 9 and r(N) = 1.32 Å for a co-ordination number
of 4.
1976
J. Chem. Soc., Dalton Trans., 2001, 1975–1985

View Article Online
Table 1 Hydrogen bonds in the investigated crystal structuresa
D–H ؒ ؒ ؒ A
d(D–H)/Å
d(H ؒ ؒ ؒ A)/Å
d(D ؒ ؒ ؒ A)/Å
α(DHA)/Њ
EuL2
N5–H5 ؒ ؒ ؒ O12
N6–H6 ؒ ؒ ؒ Ow1
N7–H7 ؒ ؒ ؒ O6
N8–H8 ؒ ؒ ؒ O9
Ow1 ؒ ؒ ؒ O14#1
Ow1 ؒ ؒ ؒ O10#2
Ow1 ؒ ؒ ؒ Ow3#3
Ow2 ؒ ؒ ؒ Ow4#4
Ow2 ؒ ؒ ؒ O5
0.88
0.88
0.88
0.88
2.12
2.26
2.11
2.26
2.89(2)
3.01(1)
2.922(9)
2.825(9)
2.80(2)
2.98(1)
2.99(1)
2.66(2)
2.731(9)
2.88(2)
2.97(2)
2.92(1)
2.65(2)
145.6
143.2
152.7
121.5
Ow2 ؒ ؒ ؒ Ow3#5
Ow2 ؒ ؒ ؒ O13#6
Ow3 ؒ ؒ ؒ O8
Ow4 ؒ ؒ ؒ O11
EuL3
N(5)–H(5) ؒ ؒ ؒ O(20)
N(7)–H(7) ؒ ؒ ؒ O(24)
N(9)–H(9) ؒ ؒ ؒ O(23)
N(11)–H(11) ؒ ؒ ؒ O(18)
O(13)–H(13A) ؒ ؒ ؒ O(3)#7
O(23)–H(23A) ؒ ؒ ؒ O(22)#8
O(24)–H(24A) ؒ ؒ ؒ O(15)#9
0.88
0.88
0.88
0.88
0.84
0.84
0.84
1.96
1.96
1.87
1.93
2.00
2.01
2.15
2.825(17)
2.774(18)
2.74(2)
2.786(17)
2.696(18)
2.792(19)
2.773(18)
168.2
152.3
167.2
162.6
139.0
155.5
131.3
LuL3
N(5A)–H(5A) ؒ ؒ ؒ O(13)#10
N(7A)–H(7A) ؒ ؒ ؒ O(28)#10
N(9A)–H(9A) ؒ ؒ ؒ O(19) #11
N(11A)–H(11A) ؒ ؒ ؒ O(34)
O(13A)–H(13B) ؒ ؒ ؒ O(3A)
N(5B)–H(5B) ؒ ؒ ؒ O(16)#12
N(7B)–H(7B) ؒ ؒ ؒ O(39)
0.88
0.88
0.88
0.88
0.84
0.88
0.88
0.88
0.88
1.98
1.89
2.01
1.93
2.40
1.98
1.92
2.05
1.83
2.854(14)
2.76(2)
2.861(16)
2.800(2)
2.885(16)
2.854(14)
2.80(2)
176.2
176.5
162.2
168.4
117.4
171.0
172.1
171.0
169.0
N(9B)–H(9B) ؒ ؒ ؒ O(23)
N(11B)–H(13B) ؒ ؒ ؒ O(26)
2.92(2)
2.70(2)
a Symmetry operations: #1 2 Ϫ x, 2 Ϫ y, 2 Ϫ z; #2 x, 2 Ϫ y, 2 ϩ y; #3 2 Ϫ x, y, 1 Ϫ z; #4 Ϫx, Ϫy, Ϫz; #5 x Ϫ 2, Ϫy, z; #6 2 ϩ x, Ϫy, z; #7 Ϫx, Ϫy,
1
1
1
1
1
1
1
1
–
–
–
–
–
–
–
–
1
1
1
1
1
1
–
–
–
–
–
–
Ϫz ϩ 1; #8 x Ϫ 1, y, z; #9 x, y Ϫ 1, z; #10 x ϩ 1, y, z; #11 x Ϫ 2, Ϫy ϩ 2, z Ϫ 2; #12 Ϫx ϩ 2, y ϩ 2, Ϫz ϩ 2.
Fig. 3 Intermolecular hydrogen bond occurring in [Eu(L3)(CH3-
OH)]3ϩ linking two neighbour molecules.
The asymmetric unit of [Lu(L3)(CH3OH)][Lu(L3)(H2O)]-
[CF3SO3]6ؒ10CH3OH, LuL3, contains two different complexes,
one with a methanol molecule co-ordinated ion (molecule A),
while the other Lu(III) bears one water molecule (molecule B,
Fig. 4). They are connected by only one hydrogen bond involv-
ing the co-ordinated water molecule and one oxygen atom of a
nitro group. Therefore, as in the two complexes described above,
the crystal packing consists of a series of dimers. The average
Lu–O and Lu–N distances are 2.318(10) and 2.567(11) Å in A
and 2.310(10) and 2.558(12) Å in B, respectively. The Lu–O
(MeOH) distance is 2.347(9) Å, while the Lu–O (H2O) is
2.368(10) Å. The absolute conformation of the two molecules is
different: Λ(δδδδ) for A, with average NCCN and NCCO
torsion angles of ϩ56.5(15) and Ϫ26.7(14)Њ, and ∆(λλλλ) for
Fig. 4 Atom-numbering scheme for [Lu(L3)(CH3OH)]3ϩ (molecule A)
and [Lu(L3)(H2O)]3ϩ (molecule B).
B, with average torsion angles of Ϫ56.4(17) and ϩ28.1(16)Њ for
the NCCN and NCCO angles, respectively. The O atom of the
water or methanol molecule lies on the C4 axis going through
Lu and Y, where Y is the centre of the O4 plane. As in EuL2 and
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
1977

View Article Online
Table 2 Parameters characterising the square antiprismatic arrangements found in the investigated crystal structures (distances in Å and angles
in Њ)a
[Eu(L2)(H2O)]3ϩ
[Eu(L3)(CH3OH)]3ϩ
[Lu(L3)(CH3OH)]3ϩ
[Lu(L3)(H2O)]3ϩ
α
90.0
90.0
90.0
90.0
β
τ
103.9(3)
38.1
102.3(2)
25.5
107.8(2)
39.7
108.2(4)
39.6
δ
51.9
64.5
50.3
50.4
ε
90.0
90.0
90.0
90.0
ꢀ
177.9
179.0
177.7
177.9
φ
51.4
52.0
53.9
54.1
π
θ
139.7(3)
72.7
145.4(2)
69.9
141.8(4)
70.9
140.0(4)
70.0
γ
66.4(2)
1.723
63.0(3)
1.621
66.7(4)
1.587(1)
2.93(2)
3.10(1)
1.511(6)
0.759(5)
4.118
67.0(4)
1.576(1)
2.93(2)
3.07(2)
1.499(6)
0.791(5)
4.142
Y–Ocoord
a
b
c (Ln1–X)
d (Ln1–Y)
N1 ؒ ؒ ؒ N3
N2 ؒ ؒ ؒ N4
Ln–O
Ln–N
Ln–Ocoord
2.962(9)
3.212(7)
1.673(3)
0.708(3)
4.161(8)
4.21(1)
2.379(5)
2.680(6)
2.431(5)
2.93(2)
3.21(1)
1.624(6)
0.831(5)
4.14(2)
4.16(2)
2.414(9)
2.636(12)
2.451(9)
4.182
4.145
2.318(10)
2.567(11)
2.347(9)
2.310(10)
2.558(12)
2.368(10)
a X and Y are respectively the centres of the N4 and O4 planes, Ocoord is the supplementary co-ordinated molecule.
EuL3, the Lu(III) occupies the common vertex of two square
pyramids and both co-ordination polyhedra can be described as
distorted monocapped antiprisms, with very similar average
twist angles of 39.7 (A), and 39.6Њ (B). For each sub-unit, four
hydrogen bonds take place involving the amide hydrogen atoms
(Table 1). No stacking interaction is evidenced between the
two closest aromatic moieties of each sub-unit, the distance
between the two centres being 4.79 Å and the angle between the
two aromatic rings amounting to 23.5Њ.
Analysis of the co-ordination polyhedra
We have performed a quantitative analysis of the co-ordination
polyhedra found in the investigated structures using the dis-
tances and angles presented in Fig. 5 and listed in Table 2 for
each compound. In all cases, X, the centre of the N4 plane, and
Y, the centre of the O4 plane, lie on the same line and define a
C4 axis which goes through the O atom of the solvent molecule
co-ordinated in the ninth site (H2O or MeOH). Ln–O distances
are shorter than Ln–N distances with, as a consequence, φ
being smaller than θ. Thus, the O4Ln square pyramid is more
flattened. The value of the angle β and the distances between
the trans nitrogen atoms are too small to allow encapsulation of
the Ln(III) in the cavity of the macrocycle; on the other hand, π
is large enough to allow co-ordination of an additional small
molecule. This angle is largest in EuL3 to accommodate the
bound methanol molecule.
Comparing [Eu(L3)(CH3OH)]3ϩ with [Lu(L3)(CH3OH)]3ϩ
gives some insight into the influence of the ionic radius on the
first co-ordination sphere. The obvious fact is the shorter Lu–O
and Lu–N distances due to the smaller ionic radius of Lu(III).
As a consequence, a tightening of the arms around the Lu(III)
is observed and β becomes greater. However, the difference in
the Ln–O distances is more pronounced than for the Ln–N
distances because of the relative rigidity of the macrocycle and
the ability of the flexible pendant arms to wrap more or less
closely around the metal ion. Furthermore, we note that in both
complexes the average Ln–O distances are smaller than the sum
r(Ln3ϩ) ϩ r(O) = 2.36 Å, and the Ln–N distances are larger
than r(Ln3ϩ) ϩ r(N) = 2.44 Å. This arises from the fact that the
tertiary amine nitrogen atoms of the macrocycle are in a more
rigid environment than the oxygen atoms which can rearrange
and then adopt a favourable co-ordination geometry.32 There-
fore, the a value is independent of the ionic radius of the
metal ion. The comparison between [Lu(L3)(CH3OH)]3ϩ and
Fig.
5
Parameters defining
a
twisted square antiprismatic co-
ordination polyhedron: relevant distances, spatial (
) angles; X and Y are the centres of the N4 and O4 planes,
respectively.
) and projection
(
[Lu(L3)(H2O)]3ϩ does not show any drastic difference. Thus the
relative orientation of the O4 and N4 planes is not governed by
the supplementary co-ordinated molecule. The most remark-
able difference found between the reported structures is the
value of the angle τ which is 25.5Њ in [Eu(L3)(CH3OH)]3ϩ
whereas it varies between 38.1 and 39.7Њ in the other com-
1978
J. Chem. Soc., Dalton Trans., 2001, 1975–1985

View Article Online
Table 3 Absorption characteristics of compounds L2–4 and their complexes with Eu and Tb in dry and degassed acetonitrile at 293 K, and the
energy of the singlet (S*) and triplet (T*) states at 77 K
Compound
π → π* transitions, Ea/cmϪ1
E(S*)a/cmϪ1
E(T*)/cmϪ1
c
L2
41 700 (4.72)
22 800,b 24 400c
26 100,b 19 800b
22 350,b 24 100d
L3
44 840 (4.80), 40 650 sh (4.59), 33 300 (4.54)
47 100 (4.57), 36 500 (4.62)
41 500 (4.74)
35 000
35 250
34 600
L4
EuL2
TbL2
EuL3
TbL3
EuL4
TbL4
25 700b
c
41 150 (4.75)
34 600
c
c
c
45 250 (4.75), 34 000 (4.80)
45 400 (4.65), 34 120 (4.71)
47 170 (4.56), 37 450 (4.58)
47 200 (4.45) 37 550 (4.48)
c
35 250
22 300b
22 300b
a Maximum of band envelope; log ε is given within parentheses. The π–π*(CO) band is seen between 50 000 and 54 000 cmϪ1 (log ε = 5.1–5.3) for the
“free” ligands. b Maximum of emission c Not observed. d 0–Phonon transition.
pounds. This is why the co-ordination polyhedron of
[Eu(L3)(CH3OH)]3ϩ is closer to a twisted square prismatic
arrangement.
pseudo C4v symmetry. At room temperature, the energy of the
5D0 ← 7F0 transition amounts to 17 233 cmϪ1 (fwhh = 4
cmϪ1), closer to the calculated value of 17 228 cmϪ1 for a CN
of nine (δN = –17.8 cmϪ1),34 taking into account the bound
methanol molecule, than to the calculated value of 17 244 cmϪ1
for CN = 8, without a co-ordinated methanol molecule.
Probing the local environment of the metal ion by high resolution
luminescence
When dried under vacuum, the isolated complexes tend to lose
their solvation molecule, as indicated by the vibrational spectra.
This prompted us to use the europium(III) ion as a probe of its
local environment by resorting to high resolution luminescence
measurements on microcrystalline samples of dried EuL2 and
of undried EuL3. Non-selective excitation (5L6 ← 7F0,1 tran-
sitions, 10 K) of EuL2 yields an emission spectrum similar to
the one obtained for EuL1.25 The relative and corrected inten-
Photophysical properties of ligands L2–4 and of the complexes
with L2,3 in the solid state
The absorption spectral data in acetonitrile are reported in
Table 3. They indicate that complexation induces minor shifts
in the energy of the transitions and small variation in their
intensity. We note that molar absorption coefficients of the
lanthanide complexes are large, a favourable condition for an
efficient antenna effect.
5
sities of the D0 → 7FJ transitions (J = 0–4) are 0.21, 1.00,
1.33, 0.07, 1.52, respectively. The sizeable oscillator strength of
At room temperature, no emission from the “free” ligands is
seen, neither in the solid state nor in solution. On the other
hand, the ligands display a weak luminescence at 77 K (Fig. S3,
ESI). The maximum of the singlet state emission occurs at
35 000 cmϪ1 for L3 and 35 250 cmϪ1 for L4, but no fluorescence
band appears in the emission spectrum of L2. The maximum of
the triplet state emission is found at 24 400 cmϪ1 for L2 and
24 100 cmϪ1 for L4, while two emission bands are observed for
L3, the most intense having a maximum at 26 100 cmϪ1, while
the second one appears at 19 800 cmϪ1 with a very weak inten-
sity. The excitation spectrum recorded upon monitoring the
latter displays a maximum at 27 800 cmϪ1 which is also seen for
the complexes (see below). We think that the triplet state with
the highest energy is due to electronic density mainly localised
on the aromatic rings, as for L2 and L4, while the lowest energy
one may be explained by the presence of a molecular orbital
involving the nitro groups. We have no explanation why such a
triplet state is not seen for L4. The energies of the observed
excited states of the “free” and complexed ligand are reported
in Table 3.
5
the D0 → 7F0 transition is consistent with a Cn or Cnv point
group of symmetry. The laser-excited excitation spectrum of
the 5D0 ← 7F0 transition displays one large band when
5
monitored on the D0 → 7F2 transition (full width at half
height, fwhh = 14 cmϪ1), which is consistent with the presence
of a single averaged environment for the Eu(III). Part of the
broadening arises from vibronic components with energies simi-
7
lar to that of the F2 state, as demonstrated by the excitation
spectrum obtained upon monitoring the highest energy com-
ponent of the 5D0 → 7F1 transition. In this case the excitation
5
band is narrower: fwhh = 9 cmϪ1. The energy of the D0 ←-
7F0 transition recorded at room temperature (ν = 17 244 cmϪ1
)
allows us to assess the co-ordination number of the Eu(III) on
the basis of the nephelauxetic effect: ν Ϫ ν0 = CCNΣiniδi; CCN
=
1.06 for a co-ordination number of 8, ni is the number of donor
atoms of type i, the δi parameters represent the nephelauxetic
effect induced by each donor atom, and ν0 is equal to 17 374
cmϪ1 33 The calculated position of the 5D0 level for a co-
.
ordination number (CN) of eight, using δC᎐O = Ϫ16.6 cmϪ1 and
᎐
δN = Ϫ12.8 cmϪ1 33
,
is 17 249 cmϪ1, in good agreement with
When Li are complexed to luminescent Ln(III) the ligand
luminescence is strongly affected by the energy transfer to the
metal ion. LnL2 complexes have a similar luminescence
behaviour to that previously described for complexes formed
with L1. In both series of complexes the ligand fluorescence
remains relatively important, pointing to a not so efficient inter-
system crossing. The first excited singlet state is found at 34 600
cmϪ1 (maximum of the emission band). Furthermore, emission
from the triplet state (25 700 cmϪ1) is also important in the
europium() complex, partly because of a poor triplet-to-
Eu(III) energy transfer (Fig. 6).
the experimental one, showing that there is no bound water
molecule in the solid state after drying the compound and con-
5
sistent with the decay rate constants of the D0 excited state
obtained for dried and undried crystals (see below). Recalcu-
lating the nephelauxetic parameter from our data, we find
δN-cyclen = Ϫ14.1 cmϪ1, a value close to those reported for other
complexes with cyclen derivatives (Ϫ13.6 to Ϫ14.3 cmϪ1).34
The emission spectrum of EuL3 displays much sharper bands
than those of EuL1 and EuL2, but the relative intensities of the
5D0 → 7FJ transitions (J = 0–4) are similar: 0.19, 1.00, 1.46,
0.07, 1.75, respectively at 10 K (Fig. S2, ESI). The 5D0 ← 7F0
excitation spectrum presents a very narrow symmetrical band
(fwhh = 1 cmϪ1) pointing to a single well defined europium()
site. Under selective laser excitation, the 5D0 → 7F1 transition
displays two components while the 5D0 → 7F4 transition
presents four components assigned to two E-type (each display-
ing a narrow splitting) and two A-type, consistent with a
Excitation spectra of the LnL3 complexes in the solid
state (Ln = Sm, Eu or Tb) display only one broad ligand-
centred band with a maximum at 27 850 cmϪ1, indicating
that the L→Ln energy transfer goes through the triplet state
involving the nitro groups (Fig. 7). The Eu(III) emits relatively
strongly in the solid state at room temperature, which is not the
case of Tb(III), pointing to a main energy transfer path going
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
1979

View Article Online
5
11
–
–
transitions (J = – , Fig. 7), and no luminescence from the
2
2
ligand is observed.
To assess the importance of the non-radiative de-excitation
processes, we have measured the decay rate constants of solid
state samples of the complexes with Sm, Eu and Tb (Table 4).
All the luminescence decays could be fitted satisfactorily with
single exponential decay laws. As expected from the small
energy gap between G5/2 and F11/2 (ca. 7 200 cmϪ1), the
de-activation rate constant of SmL3 is large, lifetimes of the
excited state level being in the range 22–27 µs only. De-
activation rate constants for the dried sample of EuL2 are small,
only slightly temperature-dependent and comparable to those
found for EuL1,25 reflecting that no water is co-ordinated to the
metal ion, while the rate constant for the undried sample is
substantially larger. In the case of the dried sample of EuL3 the
decay constants are about double those for dried EuL2, and
they are even larger than for undried EuL2, reflecting the
presence of additional de-activation paths. Back transfer to
the ligand triplet state does not seem to play a large role since
the rate constants are not very temperature dependent, but
photo-induced electron transfer (PET) processes could be
operative with this ion. In particular, we suspect such a process
involving the ligand singlet state and the Eu(III) because of the
higher absorption coefficients found in the absorption spectrum
of EuL3, which could be indicative of the presence of a charge
transfer. The de-activation rate constant for TbL2 lies indeed in
the expected range for this type of compound (cf. 0.53 msϪ1 for
TbL1 25). In the case of TbL3, the larger decay constant can be
explained by an efficient back transfer process (see above),
while the rate constant measured for TbL4 is closer to that
observed for the compounds with L1,2. Generally speaking, the
rate constants for the terbium() edifices are smaller than those
for the europium() complexes in view of the larger energy gap
between the lowest sub-level of the excited state and the highest
sub-level of the ground state in the former ion.
4
6
Fig. 6 Fluorescence (a) and phosphorescence (b) spectra of EuL2, and
phosphorescence (c) spectrum of TbL2 (solid state, 77 K). The following
excitation energies, time delays and bandpaths were used: (a) 42 373
cmϪ1, 0 µs, 4 nm; (b) 42 553 cmϪ1, 20 µs, 10 nm; (c) 35 714 cmϪ1, 20 µs,
2.5 nm.
Structural properties of the complexes in solution
In order to rationalise the photophysical behaviour of the com-
plexes in solution, we have first attempted to establish their
structure in solution. We have taken the complexes of Eu(III)
and Lu(III) as models since their NMR spectra can easily be
assigned and since we have solid state structures for the com-
plexes with L2 and L3. The NMR spectra of the complexes are
consistent with an average C4-symmetry at 293 K (see Experi-
mental section) which leads us to conclude that the complexes
adopt in solution a conformation similar to the one found in
the crystal structures. However, upon lowering the temperature,
spectra are obtained for Eu(III) which are characteristic of an
equilibrium between isomers, a commonly observed fact for
such systems.35,36 In the spectra of the lutetium() complexes
the methylenic protons display broad signals at room temper-
ature, pointing to this ion being in a somewhat more labile
environment but only one isomer is present, even at low tem-
perature (ESI, Fig. S5).
To render the comparison between solid state and solution
structure more quantitative, we have evaluated the distance
r between the aromatic protons and Eu(III) in EuL2 in D2O by
determining the longitudinal relaxation times T1 and using the
equation:37 (1/T1)p Ϫ (1/T1)d = k/r6 where the indices p and d
stand for paramagnetic and diamagnetic complexes, respec-
tively. The values obtained for the europium() complex
amount to 1.321, 1.139 and 0.182 s for the para, meta and ortho
protons, respectively. The corresponding values could not be
determined for the lutetium() complex because the signals are
not separated. Therefore we have chosen the proton in the
para position as a reference because it is less influenced by the
paramagnetism of the Eu(III), which allowed us to estimate
k and the distances between Eu(III) and the ortho and meta
protons: 5.87 and 7.97 Å, respectively. The respective average
Fig. 7 Emission spectra of TbL3 (a), SmL3 (b) and EuL3 (d) in the
solid state at 77 K, and excitation spectrum of EuL3 (c, Ean = 16 234
cmϪ1). The following excitation energies, time delays and bandpaths
were used: 29 155 cmϪ1, 60 µs, 10 nm (a), 28 165 cmϪ1, 10 µs, 5 nm
(b), 25 316 cmϪ1, 60 µs, 5 nm, 1% attenuator (d). The asterisk refers to
the Rayleigh diffusion band.
through this low-lying 3T*(NO2) state. The energy gap 5D4(Tb)–
3T*(NO2) could not be estimated because no emission from the
lowest triplet state is observed in the complexes, even with non-
luminescent ions. A proof of this proposed sensitisation mech-
anism is further given by the fact that terbium() emission
is decreasing exponentially with increasing temperature, a
5
phenomenon characteristic of a thermally activated D4(Tb)-
to-3T*(NO2) back transfer. The activation energy of the latter
process, as determined by measuring the 5D4 de-activation rate
versus temperature, amounts to 1076 37 cmϪ1 for a solid
state sample, a value that should approximately match the
5D4(Tb)–3T*(NO2) energy difference (Fig. S4, ESI). The
relatively intense emission of Eu(III) in the solid state can be
partially explained by the proximity of 3T*(NO2) with 5D0(Eu),
the energy gap being smaller than in EuL1,2. We have also
4
studied the samarium() complex because its G5/2 acceptor
5
level lies around 550 cmϪ1 higher in energy than the D0(Eu)
level, so a sizeable L→Sm energy transfer was expected. Its
4
emission spectrum displays the characteristic G5/2 → 6HJ
1980
J. Chem. Soc., Dalton Trans., 2001, 1975–1985

View Article Online
Table 4 Observed decay rate constants of the Eu(5D0), Tb(5D4) and Sm(4G5/2) excited levels in Ln(L2,3,4) (Ln = Eu, Tb or Sm) under various
experimental conditions
LnLi
Conditions
T/K
Eexc/cmϪ1
k/msϪ1
EuL2
Dried crystals
Dried crystals
Dried crystals
Dried crystals
Undried crystals
1.1 × 10Ϫ3 /CH3CNa
Dried crystals
77
293
77
293
293
293
293
77
293
293
293
77
17 258
17 258
42 553
42 553
42 553
42 373
35 714
37 037
34 483
34 483
34 247
17 238
25 974
25 707
26 667
26 667
26 667
28 011
28 011
28 169
28 011
41 667
41 667
20 530
0.89 0.02
0.96 0.02
0.94 0.07
1.09 0.07
1.67 0.11
1.05 0.04
0.54 0.01
0.55 0.07
0.42 0.01
0.62 0.02
0.37 0.02
1.92 0.07
2.04 0.07
2.22 0.10
1.12 0.05
1.30 0.03
2.27 0.10
0.87 0.03
20.8 1.7
TbL2
EuL3
Dried crystals
1.1 × 10Ϫ3 /CH3CNa
1.3 × 10Ϫ3 /Water
1.2 × 10Ϫ3 /D2O
Dried crystals
Dried crystals
77
Dried crystals
293
293
293
293
77
293
77
293
77
77
1.0 × 10Ϫ3 /CH3CNa
8.6 × 10Ϫ4 /D2O
1.0 × 10Ϫ3 /Water
Dried crystals
TbL3
Dried crystals
Dried crystals
SmL3
SmL3
EuL4
TbL4
TbL4
45.2 4.1
37.5 0.3
1.12 0.05
0.72 0.04
0.71 0.02
Dried crystals
5 × 10Ϫ5 /CH3CNa
5 × 10Ϫ5 /CH3CNa
Dried crystals
30–290
a Dried and degassed CH3CN (≈15 ppm of water).
Table 5 Quantum yields measured for LnLi, Ln = Eu or Tb; i = 1–4,
in dry and degassed acetonitrile at 293 K
LnLi
Φabs (%)
LnLi
Φabs (%)
EuL1
EuL2
EuL3
EuL4
0.06a
0.02
TbL1
TbL2
TbL3
TbL4
6.4a
0.22
0.003
5 × 10Ϫ3
5.6 × 10Ϫ2
b
a Ref. 25. b Too low to be measured.
values obtained by X-ray diffraction are 5.63 and 7.43 Å. We
therefore conclude that the average structure in solution is close
to the one observed in the solid state and that this is most
probably true for all the complexes we are investigating.
Fig. 8 Schematic representation of the Li→Ln energy transfer in LnLi
complexes (left, i = 1; centre, i = 2; right, i = 3).
space dipole–dipole (Förster) mechanism,38 the improvement
expected on the sole basis of the shorter R distance in EuL1 is
calculated to be 1.3. The remaining factor (1.9) is therefore
attributable to the better orientation of the chromophore in
EuL1, which is almost perpendicular to the C4 axis of the
molecule.
Photophysical properties of the complexes in solution
The absolute quantum yields of the complexes in water are
listed in Table 5. All the europium() complexes are poorly
luminescent, which can be traced back, at least in part, to the
relatively high energy of the triplet state in the complexes with
L1,2,4. The values found for LnL1 and LnL2 may be rationalised
as follows. The efficiency of the L→Ln energy transfer depends
on energetic and geometric factors. In fact, the former are not
very different in the two complexes, ∆E(3ππ*–5D0) being
equal to 9 000 and 8 750 cmϪ1 in EuL1 25 and EuL2, respectively.
Therefore the 2.5 fold larger quantum yield exhibited by EuL1
must arise from a better positioning of the chromophoric
groups. Comparing the crystal structures of EuL5 (taken as
model for EuL1)22 and EuL2 we indeed evidence the importance
of the spacer methylenic group between the amide function and
the aromatic moiety. It allows the arms to bend and the
chromophoric groups adopt a much better conformation for an
efficient through space transfer onto the metal ion (Fig. 8). We
have defined R as the average distance between the centre of the
aromatic ring and the lanthanide() ion while the average angle
between the metal ion, the amide nitrogen atom and the centre
of the aromatic moiety and the nitro group has been labelled α
and αЈ, respectively. These parameters amount to R = 5.93 and
6.23 Å and α = 92.0 and 115.2Њ for EuL5 and EuL2, respectively.
Assuming that the energy transfer is operative via a through
EuL3 exhibits surprising luminescence properties. Its quan-
tum yield is very low, but it displays a strong europium() solid
state luminescence intensity upon ligand excitation. The former
can be explained by the long distance RЈ between the centre of
the NO2 group (where the orbital describing the low-lying trip-
let state is mostly located) and the metal ion (9.82 Å): a calcul-
ation similar to the one above predicts a quantum yield 15 times
smaller than for EuL2 (αЈ is similar to the corresponding α angle
in EuL2, 117.2Њ) an unfavourable factor which more than over-
comes the energetic factor, expressed by a smaller, favourable,
energy gap ∆E(3T*–5D0) estimated to 2 500 cmϪ1 (from the
3
maximum of the T* emission in the “free” ligand data). On
the other hand, we think that the relatively strong luminescence
emitted in the solid state (and visible to the naked eye) can be
explained by an inter-molecular energy transfer from the
3T*(NO2) state of one molecule to the europium centre of a
neighbouring one, the distance between the sensitiser and the
metallic acceptor (4.79 Å) being much shorter than the intra-
molecular one. This process is favoured by the presence of a
hydrogen bond between one nitro group and the co-ordinated
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
1981

View Article Online
ments confirm the tetragonal site symmetry of the europium()
ion evidenced by crystallographic data. Our photophysical
investigations show that both energy and geometric parameters
must be taken in consideration to describe precisely the L→Ln
energy transfer efficiency. The latter can be very different for the
same compound, depending on the solvation, as shown for
EuL3. We conclude that both sets of parameters must be well
characterised to explain and understand the photophysical
properties of the complexes with ligands derived from cyclen.
In addition, a sensitive design of a lanthanide-containing
luminescent probe will also have to take into account the effi-
ciency of the inter-system crossing from singlet to triplet states,
a matter that is not documented here, but that we have
discussed recently.40
Experimental
Synthesis and characterisations
Fig. 9 Phosphorescence spectra of TbL4 (top) and EuL4 (bottom),
5 × 10Ϫ5 M in CH3CN at 77 K, Eexc = 41 667 cmϪ1, time delay = 50 µs,
bandpath = 5 nm.
Solvents and chemicals were purchased from Fluka AG (Buchs,
Switzerland). Acetonitrile was treated by CaH2 and P2O5.41
Dichloromethane and tetrahydrofuran were distilled from
CaH2. Lanthanide trifluoromethanesulfonates (triflates) were
prepared from the oxides (Rhône-Poulenc, 99.99%) and triflic
acid.42 Elemental analyses were performed by Dr H. Eder of
the Microchemical Laboratory of the University of Geneva.
1,4,7,10-Tetraazacyclododecane43 and L1 were synthesized as
previously described.25
N-(Phenyl)bromoacetamide was prepared in 92% yield using
the method of Hamada et al.44 with bromoacetyl bromide
instead of chloroacetyl chloride. TLC: 1 spot, Rf = 0.76 (5%
MeOH in CH2Cl2). δH (CDCl3) 4.01 (s, 2H, BrCH2), 7.16 (t,
2J = 7.5, 1H, Ar), 7.35 (t, 2J = 7.5, 2H, Ar), 7.52 (d, 2J = 7.5 Hz,
2H, Ar). Mp 133.5–134.5 ЊC.
methanol molecule of a neighbouring complex (Fig. 3). Ligand
L4 and its complexes have been synthesized to support this
explanation, with respect to the properties of LnL3. It turned
out that the europium() complex exhibits an extremely poor
luminescence, both in the solid state and in solution. This is
explained by the existence of only one triplet state occurring at
22 300 cmϪ1 for the complexes (maximum of the emission band,
Fig. 9). The resulting ∆E(3ππ*–5D0) gap of 5 070 cmϪ1 (a value
which is in reality larger if we take into account the 0–phonon
transition) is usually not considered to be extremely favourable
to an efficient energy transfer from the ligand to Eu(III);
emission from the triplet state is seen in the luminescence
spectrum, the ratio between the intensity of the metal-centered
luminescence and that from the triplet state being around
0.7 : 1.
The situation for the terbium() complexes is similar to that
discussed for Eu(III) as far as the complexes with L1 and L2 are
concerned: both the energy factor (∆E(3ππ*–5D4) = 5600 and
5500 cmϪ1, respectively, considering the maxima of triplet state
emission) and the geometric factor are more favourable for L1,
which explains the 30-fold smaller quantum yield for TbL2. The
40-fold decrease in quantum yield upon going from TbL2 to
TbL3 is easily explained by the back transfer process quantified
in the solid state (see above) while the tenfold larger quantum
yield of TbL4 with respect to TbL3 may be explained both by a
better orientation of the chromophoric group and a better
energy factor. Finally, the measured radiative rates in water and
D2O for the TbL2 and EuL3 complexes are consistent with one
co-ordinated water molecule: applying the proposed formula
for this type of compounds,39 we find hydration numbers
q = 0.95 and 0.86, respectively.
1,4,7,10-Tetrakis[N-(phenyl)carbamoylmethyl]-1,4,7,10-tetra-
azacyclododecane (L2). A mixture of 1 equivalent of cyclen
(0.517 g, 3.0 mmol), 4 equivalents of N-(phenyl)bromo-
acetamide (2.569 g, 12.0 mmol) and 5 equivalents of triethyl-
amine (2.091 cm3, 15.0 mmol) in 50 cm3 of dry tetrahydrofuran
was refluxed for 20 h under nitrogen. After evaporation of the
solvent, the residue was partitioned between 100 cm3 of water
and 100 cm3 of CH2Cl2. The organic phase was evaporated and
the remaining solid dissolved in 6 cm3 of hot dimethylform-
amide; 20 cm3 of water were added and the cloudy solution was
heated until boiling. Upon cooling, a white solid precipitated
and was filtered off. After drying (8 h, 1 mbar, 333 K), 1.640 g
of L2 were collected (yield 78%). Found: C, 68.07; H, 6.92; N,
15.65. Calc. for C40H48N8O4: C, 68.15; H, 6.87; N, 15.90%;
δH (CD3CN, 295 K) 2.90 (s, 4H, NCH2), 3.23 (s, 2H, NCH2CO) ,
7.01 (t, 1H, Ar), 7.19 (t, 2J = 7.5, 2H, Ar), 7.49 (d, 2J = 8.5 Hz,
2H, Ar), 9.20 (s, 1H, NH). mp 201–203 ЊC. νmax 3290 (NH),
1634 cmϪ1 (C᎐O).
᎐
N-(4-Nitrophenyl)bromoacetamide. A mixture of p-nitro-
aniline (5.525 g, 40.0 mmol) and dry pyridine (4.0 cm3, 50.0
mmol) in 250 cm3 of dry tetrahydrofuran was ice-cooled. A
solution of bromoacetyl bromide (8.074 g, 40.0 mmol) in 15
cm3 of dry tetrahydrofuran was slowly added and the resulting
mixture stirred for 1 h at 273 K, then 1 h at 298 K. The solution
was filtered and concentrated to 50 cm3. Then 50 cm3 of water
were added and the solution was allowed to stand overnight at
274 K. The orange-yellow solid was filtered and recystallised in
100 cm3 of acetone–water (2 : 3, v/v). After drying (8 h, 1 mbar,
328 K), 7.190 g were obtained (yield: 69%). Rf = 0.20 (CH2Cl2).
δH (DMSO-d8, 295 K) 4.11 (s, 2H, BrCH2), 7.83 (d, 2J = 9.0, 2H,
Ar), 8.24 (d, 2J = 9.0 Hz, 2H, Ar), 11.00 (s, 1H, NH). mp 174.0–
175.5 ЊC.
Conclusion
In all the complexes discussed here the macrocyclic nitrogen
atoms and the amide oxygen atoms define two squares, the
latter being capped by a solvent molecule, leading to a co-
ordination number of nine. Thus, all the compounds adopt a
twisted square antiprismatic arrangement around the metallic
centre. This co-ordination is imposed by the design of the
ligand and its co-ordinating atoms. For instance, when the
amide function is replaced by phosphinate, the latter creates a
steric hindrance which prevents co-ordination of an additional
molecule with the same Ln(III).17 Comparing the solid state
structures of two complexes with the same ligand but two dif-
ferent ions, [Eu(L3)(CH3OH)]3ϩ and [Lu(L3)(CH3OH)]3ϩ, we do
not observe any important changes in the first co-ordination
sphere, except for smaller distances between the Lu(III) and the
co-ordinating atoms. High-resolution luminescence measure-
1,4,7,10-Tetrakis[N-(4-nitrophenyl)carbamoylmethyl]-1,4,7,
10-tetraazacyclododecane (L3). A mixture of 1 equivalent of
1982
J. Chem. Soc., Dalton Trans., 2001, 1975–1985

View Article Online
cyclen (0.517 g, 3.0 mmol), 4 equivalents of N-(4-nitrophenyl)-
bromoacetamide (3.109 g, 12.0 mmol) and 8 equivalents of tri-
ethylamine (3.345 cm3, 24.0 mmol) in 150 cm3 of DMF was
heated for 14 h at 358 K. 500 cm3 of water were added and the
solution was heated at 373 K during 0.5 h. After cooling, the
beige precipitate was filtered, washed with 20 cm3 of H2O and
dried (8 h, 1 mbar, 298 K). The solid was suspended in 35 cm3
of DMF and 25 cm3 of water and the mixture heated at 373 K
during 0.5 h. After cooling, the solid was filtered off and dried
(8 h, 1 mbar, 298 K) to give 1.830 g of L3 (yield 69%). Found: C,
54.02; H, 5.28; N, 18.68. Calc. for C40H44N12O12: C, 54.30; H,
5.01; N, 19.00%; δH (DMSO-d8, 295 K) 3.33 (s, 4H, NCH2), 3.36
(d, 2J = 7.5, 2H, Ar), 7.01 (d, 2J = 7.5 Hz, 2H, Ar), Ϫ2.89 (s, br,
1H, eq. ring CH2), Ϫ4.52 (s, br, 1H, eq. ring CH2), Ϫ7.12 (s, br,
1H, ax. ring CH2), Ϫ8.65 (s, br, 1H, NCH2CO), Ϫ9.79 (s, br,
1H, NCH2CO); δC{1H} (CD3CN, 293 K) 192.1 (1C, CO), 145.4
(1C, Ar), 138.9 (1C, Ar), 124.1 (2C, Ar), 121.5 (2C, Ar), 103.4
(br, 1C, ring CH2), 92.0 (br, 1C, ring CH2), 80.6 (br, 1C,
NCH2CO). νmax 3482 (OH), 3293 (NH), 1629 (C᎐O), 1530 and
᎐
1351 cmϪ1 (NO2).
[Tb(L3)(CH3OH)][CF3SO3]3ؒH2O. Found: C, 34.10; H, 3.48;
N, 11.05. Calc. for C44H50F9N12O23S3Tb: C, 34.45; H, 3.29; N,
10.96%. νmax 3483 (OH), 3272 (NH), 1632 (C᎐O), 1530 and
᎐
1351 cmϪ1 (NO2).
2
2
(s, 2H, NCH2CO), 7.76 (d, J = 9.1, 2H, Ar), 8.11 (d, J = 9.1
[Lu(L3)]3ϩ. δH (CD3CN, 233 K, ESI, Fig. S5) 10.45 (s, br, 1H,
NH), 8.24 (d, 2J = 9.1, 2H, Ar), 7.83 (d, 2J = 9.1, 2H, Ar), 3.96
(d, 2J = 18.2, 1H, NCH2CO), 3.88 (d, 2J = 18.2, 1H, NCH2CO),
Hz, 2H, Ar), 10.50 (s, 1H, NH). mp > 230 ЊC. νmax 3284 (NH),
1651 (C᎐O), 1523 and 1351 cmϪ1 (NO2).
᎐
3.50 (t, 1H, 2Jgem(ax-eq) = 3Jax-eq = 14.1, ax. ring CH2), 3.00 (d, 1H,
2
N-(4-Nitrobenzyl)bromoacetamide. A mixture of 0.419 g
(10.5 mmol) of NaOH in 10 cm3 of water and 1.000 g (5.3
mmol) of 4-nitrobenzylamine hydrochloride in 100 cm3 of
CH2Cl2 was ice-cooled. A solution of bromoacetyl bromide
(1.070 g, 5.3 mmol) in 10 cm3 of CH2Cl2 was slowly added and
the resulting mixture stirred for 1 h at 273 K, then 1 h at 298 K.
The organic phase was washed twice with 100 cm3 of water,
dried over Na2SO4 and evaporated. The white solid was
recrystallised in 30 cm3 of acetone–water (2 : 3, v/v). After dry-
ing (8 h, 1 mbar, 328 K), 1.170 g were obtained (yield: 81%).
Found: C, 39.71; H, 3.56; N, 10.10. Calc. for C9H9BrN2O3: C,
39.58; H, 3.32; N, 10.26%. δH (CDCl3, 295 K) 3.97 (s, 2H,
2Jgem(ax-eq) = 14.1, eq. ring CH2), 2.80 (t, 1H, Jgem(ax-eq)
=
3Jax-eq = 14.1, ax. ring CH2), 2.74 (d, 1H, Jgem(ax-eq) = 14.1, eq.
ring CH2). Note: the triplets are in fact triplets of doublets and
the doublets, doublets of doublets, but the 3Jax-eq coupling con-
stant could not be determined. δC{1H} (CD3CN, 293 K) 175.9
(1C, CO), 144.6 (1C, Ar), 141.0 (1C, Ar), 124.7 (2C, Ar), 120.7
(2C, Ar), 55.5 (1C, ring CH2), 54.6 (1C, ring CH2), 63.4 (1C,
NCH2CO).
2
[Eu(L4)(CH3OH)][CF3SO3]3ؒH2O. Found : C, 36.05; H, 3.77;
N, 10.65. Calc. for C48H56EuF9N12O22NS3: C, 36.26; H, 3.68; N,
10.57%; νmax 3430 (OH), 3293 (NH), 1630 (C᎐O), 1522 and
᎐
1350 cmϪ1 (NO2).
2
BrCH2), 4.59 (d, 2H, J = 6.4, NCH2Ar), 6.92 (s, br, 1H, NH),
[Tb(L4)(CH3OH)][CF3SO3]3ؒ2H2O. Found: C, 35.07; H, 3.63;
N, 10.34. Calc. for C48H60F9N12O24S3Tb: C, 35.69; H, 3.74; N,
7.46 (d, 2H, 2J = 8.6, Ar), 8.21 (d, 2H, 2J = 8.6 Hz, Ar).
10.41%; νmax 3437 (OH), 3284 (NH), 1630 (C᎐O), 1520 and
᎐
1,4,7,10-Tetrakis[N-(4-nitrobenzyl)carbamoylmethyl]-1,4,7,
10-tetraazacyclododecane (L4). This was synthesized in 70%
yield using the same procedure as described for L3 and
recrystallised in dichloromethane–ethanol (1 : 1, v/v). Found:
C, 55.82; H, 5.39; N, 18.02. C44H52N12O12 requires: C, 56.16; H,
5.57; N, 17.86%. δH (DMSO-d8, 295 K) 2.62 (s, 4H, NCH2), 3.04
(s, 2H, NCH2CO) , 4.33 (d, 2H, NCH2Ar), 7.41 (d, 2J = 8.4, 2H,
Ar), 8.11 (d, 2J = 8.4 Hz, 2H, Ar), 8.58 (t, 1H, NH). νmax = 3270
1351 cmϪ1 (NO2).
Physico-chemical measurements
1H and 13C NMR spectra were recorded on a Bruker AM-360
spectrometer (360.16 and 90.6 MHz, respectively) or on a
Bruker Avance DRX-400 spectrometer (400.03 and 100.04
MHz). Proton chemical shifts are reported in parts per million
(ppm) with respect to TMS, and 13C chemical shifts are related
to CD3CN (δ 0.3, 117.5). UV-VIS Electronic spectra were
recorded at 293 K on a Perkin-Elmer Lambda 900 spectro-
meter, infrared spectra on a FT-IR Mattson Alpha Centauri
spectrometer (4000–400 cmϪ1, KBr pellets). Low-resolution
luminescence measurements (spectra and lifetimes) were made
on a Perkin-Elmer LS-50B spectrofluorimeter. High-resolution
spectra and lifetimes after direct excitation of the metal ion
were measured on a previously described instrumental set-up;45
lifetimes are averages of at least four determinations. Quantum
yields were determined in anhydrous and degassed MeCN with
the help of a Perkin-Elmer LS-50B spectrofluorimeter relatively
to EuL1 for the europium complexes, and to a solution of
quinine sulfate in 0.5 M H2SO4 (Φabs = 0.546) for terbium
complexes by using eqn. (1) where x refers to the compound,
(NH), 1650 (C᎐O), 1520 and 1345 cmϪ1 (NO2).
᎐
Complexes. [Ln(Li)][CF3SO3]3 (i = 2, 3 or 4) were prepared by
heating under reflux 1 equivalent of the lanthanide salt and 1
equivalent of Li in dry acetonitrile. After cooling, the solution
was filtered and concentrated, dichloromethane was added and
the resulting solution kept overnight at 277 K. The deposited
solid was recovered by filtration, washed with dichloromethane
and dried (2 h, 1 mbar, 313 K). Complexes with L2 were
re-crystallised from acetonitrile and with L3,4 from methanol
(yields 60–65%).
[Eu(L2)][CF3SO3]3. Found: C, 39.49; H, 4.00; N, 8.38. Calc.
for C43H48EuF9N8O13S3: C, 39.57; H, 3.71; N, 8.59%.
δH (CD3CN, 293 K) 20.58 (s, br, 1H, ax. ring CH2), 7.86 (s, 2H,
Ar), 7.52 (s, 1H, Ar), 7.20 (s, 2H, Ar), Ϫ3.25 (s, br, 1H, eq. ring
CH2), Ϫ5.38 (s, br, 1H, eq. ring CH2), Ϫ5.78 (s, br, 1H, ax. ring
CH2), Ϫ8.70 (s, br, 1H, NCH2CO), Ϫ9.63 (s, br, 1H, NCH2CO);
δC{1H} (CD3CN, 293 K) 194.6 (1C, CO), 134.9 (1C, Ar), 130.2
(2C, Ar), 128.4 (1C, Ar), 122.9 (2C, Ar), 100.4 (br, 1C, ring
CH2), 91.7 (br, 1C, ring CH2), 82.8 (br, 1C, NCH2CO).
Φx Ar(λr)Ir(λr)n2 Dx
x
=
(1)
Φr Ax(λx)Ix(λx)n2 Dr
r
r to the reference, Ai(λi) is the absorbance at the excitation wave-
length, I the intensity of excitation at λi, n the refractive index
(n(CH3CN) = 1.343, n (0.5 M H2SO4 in water) = 1.338), and D
the area of the emission spectrum.46
νmax 3286 (NH), 1630 cmϪ1 (C᎐O).
᎐
[Tb(L2)][CF3SO3]3. Found: C, 38.80; H, 3.80; N, 8.39. Calc.
for C43H48F9N8O13S3Tb: C, 38.86; H, 3.79; N, 8.43%. νmax 3287
(N–H), 1630 cmϪ1 (C᎐O).
᎐
[Sm(L3)(CH3OH)][CF3SO3]3ؒH2O. Found: C, 34.17; H, 3.48;
X-Ray crystallography
N, 11.07. Calc. for C44H50F9N12O23S3Sm: C, 34.49; H, 3.29; N,
10.97%. νmax 3479 (OH), 3288 (NH), 1631 (C᎐O), 1530 and
Single crystals of [Eu(L2)(H2O)][CF3SO3]3ؒ4H2O, [Eu(L3)-
(CH3OH)][CF3SO3]3ؒ2CH3OH and [Lu(L3)(CH3OH)][Lu(L3)-
(H2O)][CF3SO3]6ؒ10CH3OH were obtained from acetonitrile–
tert-butyl methyl ether for the former and methanol for the L3
complexes, and mounted in glass capillaries. Crystal data and
᎐
1351 cmϪ1 (NO2).
[Eu(L3)(CH3OH)][CF3SO3]3ؒH2O. Found: C, 34.14; H, 3.51;
N, 11.05. Calc. for C44H50EuF9N12O23S3: C, 34.45; H, 3.29; N,
10.96%. δH (CD3CN, 293 K) 19.75 (s, br, 1H, ax. ring CH2), 8.36
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
1983

View Article Online
Table
6
Crystal data and structure refinement for: [Eu(L2)(H2O)](CF3SO3)3ؒ4H2O, [Eu(L3)CH3OH][CF3SO3]3ؒ2CH3OH, and [Lu(L3)-
(CH3OH)][Lu(L3)(H2O)][CF3SO3]6ؒ10CH3OH
EuL2
EuL3
LuL3
Formula
M
C43H58EuF9N8O18S3
1394.11
C46H56EuF9N12O24S3
1580.17
C97H134F18Lu2N24O54S6
3384.58
Crystal system
Space group
Monoclinic
I2/a
Triclinic
P1
Monoclinic
P21/n
¯
a/Å
b/Å
c/Å
α/Њ
21.556(4)
23.387(5)
23.239(5)
12.897(2)
12.942(5
19.237(3)
87.85(3)
84.358(10)
86.51(3)
3187.8(14)
2
13.239(3)
35.095(7)
29.656(6)
βЊ
92.58(3)
91.13(3)
γ/Њ
V/Å3
11704(4)
8
0.600
34295
9529
13776(5)
4
1.636
32581
8387
Z
µ/mmϪ1
1.194
17821
10739
Reflections collected
Unique reflections
[I > 2σ(I)]
R [I > 2σ(I)]
wR2 (all data)
0.0656
0.2097
0.1055
0.2917
0.0929
0.2781
7 D. E. Reichert, J. S. Lewis and C. J. Anderson, Coord. Chem. Rev.,
1999, 184, 3.
8 J. Byegård, G. Skarnemark and M. Skålberg, J. Radioanal. Nucl.
Chem., 1999, 241, 281.
9 M. Komiyama, N. Takeda and H. Shigekawa, Chem. Commun.,
1999, 1443.
10 D. M. Epstein, L. L. Chappell, H. Khalili, R. M. Supkowski,
W. deW. Jr. Horrocks and J. R. Morrow, Inorg. Chem., 2000, 39,
2130.
11 D. Parker, Coord. Chem. Rev., 2000, 205, 109.
12 M. R. Spirlet, J. Rebizant, J. F. Desreux and M. F. Loncin, Inorg.
Chem., 1984, 23, 359.
13 J. A. K. Howard, A. M. Kenwright, J. M. Moloney, D. Parker,
M. Port, M. Navet, O. Rousseau and M. Woods, Chem. Commun.,
1998, 1381.
14 F. Benetollo, G. Bombieri, S. Aime and M. Botta, Acta Crystallogr.,
Sect. C, 1999, 55, 353.
15 S. Aime, A. S. Batsanov, M. Botta, J. A. K. Howard, M. P. Lowe and
D. Parker, New J. Chem., 1999, 23, 669.
16 S. Aime, A. S. Batsanov, M. Botta, J. A. K. Howard, D. Parker,
K. Senanayake and J. A. G. Williams, Inorg. Chem., 1994, 33, 4696.
17 S. Aime, A. S. Batsanov, M. Botta, R. S. Dickins, S. Faulkner,
C. E. Foster, A. Harrison, J. A. K. Howard, J. M. Moloney,
T. J. Norman, D. Parker, L. Royle and J. A. G. Williams, J. Chem.
Soc., Dalton Trans., 1997, 3623.
18 J. Rohovec, P. Vojtísˇek, P. Hermann, J. Ludvík and I. Lukesˇ,
J. Chem. Soc., Dalton Trans., 2000, 141.
19 J. R. Morrow, S. Amin, C. H. Lake and M. R. Churchill, Inorg.
Chem., 1993, 32, 4566.
structure refinement details are listed in Table 6. Diffraction
data have been collected on a mar345 Imaging Plate Detector
System at 143 K. Refinement of cell parameters, integration
and scaling of data was performed with marHKL release
1.9.1.47 No absorption correction was applied to any data
set. Structures were solved by ab initio direct methods.48 All
structures were refined using full-matrix least squares on F2
(except [Lu(L3)(CH3OH)][Lu(L3)(H2O)][CF3SO3]6ؒ10CH3OH
for which full-matrix-block least squares on F2 was used) with
all non-H atoms anisotropically defined. Some disorder prob-
lems arose during the refinement of [Eu(L2)(H2O)][CF3SO3]3ؒ
4H2O and [Lu(L3)(CH3OH)][Lu(L3)(H2O)][CF3SO3]6ؒ10CH3-
OH and they mostly dealt with CF3SO3Ϫ anions for which some
geometrical and rigid bond restraints were applied. Hydrogen
atoms were placed in calculated positions using the riding
model with a common isotropic displacement parameter
(Uiso = 0.08 Å2). Space group determination, structure solution,
refinement, molecular graphics and geometrical calculation
have been carried out with the SHELXTL software package,
release 5.1.49 Selected bond lengths and angles are reported in
Table S1 (ESI).
CCDC reference numbers 158148–158150.
See http://www.rsc.org/suppdata/dt/b1/b101312m/ for crys-
tallographic data in CIF or other electronic format.
20 S. Amin, J. R. Morrow, C. H. Lake and M. R. Churchill, Angew.
Chem., Int. Ed. Engl., 1994, 33, 773.
21 S. Amin, D. A. Voss, W. deW. Jr. Horrocks, C. H. Lake, M. R.
Churchill and J. R. Morrow, Inorg. Chem., 1995, 34, 3294.
22 R. S. Dickins, J. A. K. Howard, C. W. Lehmann, J. Moloney,
D. Parker and R. D. Peacock, Angew. Chem., Int. Ed. Engl., 1997,
38, 521.
23 R. S. Dickins, J. A. K. Howard, C. L. Maupin, J. M. Moloney,
D. Parker, J. P. Riehl, G. Siligardi and J. A. G. Williams, Chem. Eur.
J., 1999, 5, 1095.
Acknowledgements
This research is supported by grants from the Swiss National
Science Foundation and the Swiss Federal Office for
Science and Education (COST project D18). We thank the
Fondation Herbette (Lausanne) for spectroscopic equipment
and Dr P.-A. Pittet for synthesizing a first batch of ligands L2
and L3.
24 S. Aime, A. Barge, J. I. Bruce, M. Botta, J. A. K. Howard, J. M.
Moloney, D. Parker, A. S. de Sousa and M. Woods, J. Am. Chem.
Soc., 1999, 121, 5762.
25 G. Zucchi, R. Scopelliti, P.-A. Pittet, J.-C. G. Bünzli and R. D.
Rogers, J. Chem. Soc., Dalton Trans., 1999, 931.
References
1 J.-C. G. Bünzli, N. André, M. Elhabiri, G. Muller and C. Piguet,
J. Alloys Compd., 2000, 303–304, 66.
2 X. Wang, T. Jin, V. Comblin, A. Lopez-Mut, E. Merciny and
J.-F. Desreux, Inorg. Chem., 1992, 31, 1095.
3 A. D. Sherry, C. R. Malloy, F. M. H. Jeffrey, W. P. Cacheris and
C. F. G. C. Geraldes, J. Magn. Reson., 1988, 76, 528.
4 P. Caravan, J. J. Ellison, J. M. Murry and R. B. Lauffer, Chem. Rev.,
1999, 9, 2293.
26 L. Alderighi, A. Bianchi, L. Calabi, P. Dapporto, C. Giorgi, P. Losi,
L. Paleari, P. Paoli, P. Rossi, B. Valtancoli and M. Virtuani, Eur. J.
Inorg. Chem., 1998, 1581.
27 J. Dale, Isr. J. Chem., 1980, 20, 3.
28 J. P. Dubost, J. M. Leger, M.-H. Langlois, D. Meyer and
M. Schaefer, C. R. Acad. Sci., 1991, 312, 349.
5 R. B. Lauffer, MRI Clinical Magnetic Resonance Imaging, eds. R. E.
Edelman, M. B. Zlatkin and J. R. Hesselink, W. B. Saunders Co,
Philadelphia, 1996, vol. 1.
6 S. Aime, M. Botta, S. G. Crich, E. Terreno, P. L. Anelli and
F. Uggeri, Chem. Eur. J., 1999, 5, 1253.
29 D. Parker, K. Pulukkody, F. C. Smith, A. Batsanov and J. A. K.
Howard, J. Chem. Soc., Dalton Trans., 1994, 689.
30 S. Aime, A. Barge, M. Botta, M. Fasano, J. D. Ayala and
G. Bombieri, Inorg. Chim. Acta, 1996, 246, 423.
31 G. Bombieri, J. Alloys Compd., 1997, 249, 76.
1984
J. Chem. Soc., Dalton Trans., 2001, 1975–1985

View Article Online
32 K. O. A. Chin, J. R. Morrow, C. H. Lake and M. R. Churchill,
Inorg. Chem., 1994, 33, 656.
33 S. T. Frey and W. deW. Horrocks Jr., Inorg. Chim. Acta, 1995, 229,
383.
41 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, Pergamon Press, Oxford, 1988.
42 J.-C. G. Bünzli and F. Pilloud, Inorg. Chem., 1989, 28,
2638.
34 G. Zucchi, PhD Dissertation, University of Lausanne, 2000.
35 S. Aime, M. Botta and G. Ermondi, Inorg. Chem., 1992, 31, 4291.
36 F. A. Dunand, S. Aime and A. E. Merbach, J. Am. Chem. Soc., 2000,
122, 1506.
37 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.
38 T. Förster, Ann. Phys. (Leipzig), 1948, 2, 55.
39 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.
43 P. A. Pittet, D. Früh, V. Tissières and J.-C. G. Bünzli, J. Chem. Soc.,
Dalton Trans., 1997, 895.
44 T. Hamada, Y. Okuno, M. Ohmori, T. Niski and O. Yonemitsu,
Chem. Pharm. Bull., 1981, 29, 128.
45 J.-C. G. Bünzli and A. Milicic-Tang, Inorg. Chim. Acta, 1996, 252,
221.
46 S. Petoud, J.-C. G. Bünzli, K. J. Schenk and C. Piguet, Inorg. Chem.,
1997, 36, 1345.
47 Z. Otwinowski and W. Minor, Methods Enzym.: Macromolecular
Crystallogr., 1997, 276, 307.
40 H.-R. Mürner, E. Chassat, R. P. Thummel and J.-C. G. Bünzli,
J. Chem. Soc., Dalton Trans., 2000, 2809.
48 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467.
49 SHELXTL, I. Bruker AXS, Madison, WI, 1997.
J. Chem. Soc., Dalton Trans., 2001, 1975–1985
1985