Tuning the self-assembly of lanthanide triple stranded heterobimetallic helicates by ligand design

Article abstract of DOI:10.1039/b715672c

The heterobitopic ligands L^AB4 and L^AB5 have been designed and synthesised with the ultimate aim of self-assembling dual-function lanthanide complexes containing either a magnetic and a luminescent probe or two luminescent probes emitting at different wavelengths. They react with lanthanide ions to form complexes of composition [Ln₂(L ^ABX)₃]⁶⁺ of which three (X = 4; Ln = Pr, Nd, Sm) have been isolated and characterised by means of X-ray diffraction. The unit cells contain triple-stranded helicates in which the three ligand strands are wrapped tightly around the two lanthanide ions. In acetonitrile solution the ligands form not only homobimetallic, but also heterobimetallic complexes of composition [Ln¹Ln²(L^ABX)₃] ⁶⁺ when reacted with a pair of different lanthanide ions. The yield of heterobimetallic complexes is analyzed in terms of both the difference in ionic radii of the lanthanide ions and of the inherent tendency of the ligands to form high percentages of head-head-head (HHH) helicates in which all three ligand strands are oriented in the same direction with respect to the Ln-Ln vector. The latter is very sensitive to slight modifications of the tridentate coordinating units. The Royal Society of Chemistry.

Full text of DOI:10.1039/b715672c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Tuning the self-assembly of lanthanide triple stranded heterobimetallic
helicates by ligand design†‡
Thomas B. Jensen, Rosario Scopelliti and Jean-Claude G. Bu¨nzli*
Received 11th October 2007, Accepted 13th November 2007
First published as an Advance Article on the web 4th December 2007
DOI: 10.1039/b715672c
The heterobitopic ligands LAB4 and LAB5 have been designed and synthesised with the ultimate aim of
self-assembling dual-function lanthanide complexes containing either a magnetic and a luminescent
probe or two luminescent probes emitting at different wavelengths. They react with lanthanide ions to
form complexes of composition [Ln2(LABX )3]6+ of which three (X = 4; Ln = Pr, Nd, Sm) have been
isolated and characterised by means of X-ray diffraction. The unit cells contain triple-stranded helicates
in which the three ligand strands are wrapped tightly around the two lanthanide ions. In acetonitrile
solution the ligands form not only homobimetallic, but also heterobimetallic complexes of composition
[Ln1Ln2(LABX )3]6+ when reacted with a pair of different lanthanide ions. The yield of heterobimetallic
complexes is analyzed in terms of both the difference in ionic radii of the lanthanide ions and of the
inherent tendency of the ligands to form high percentages of head-head-head (HHH) helicates in which
all three ligand strands are oriented in the same direction with respect to the Ln–Ln vector. The latter is
very sensitive to slight modifications of the tridentate coordinating units.
Introduction
The distinctive luminescent properties of trivalent lanthanide
ions, including sharp and easily recognisable emissions lines,
substantial Stokes’ shifts upon ligand excitation, and long lifetimes
of the excited states—allowing time-resolved detection for better
sensitivity—explain their attractiveness as versatile bioprobes
giving off light in the visible and/or near-infrared range.1–5 Associ-
ating two metal centres emitting different colours or having specific
luminescent and magnetic properties in a single molecular probe
is an attractive way of developing bimodal bioanalyses. While
luminescence-detectable contrast agents have been proposed with
success,6–11 dual lanthanide luminescent probes remain largely
unexplored, despite the recent development of double lanthanide
binding tags in which two identical LnIII ions are bound to a
specific peptide.12
While several synthetic strategies are available to produce bi- or
poly-metallic lanthanide edifices, e.g. selective crystallization from
a statistical mixture13 or successive complexation into kinetically
inert macrocyclic cavities,14–16 we are investigating the feasibility of
Scheme 1 Structures of symmetrical bitopic and monotopic ligands.
taking advantage of thermodynamically controlled self-assembly
processes. Initial experiments involved symmetric hexadentate
bitopic ligands, such as LA (Scheme 1) which forms triple
stranded homobimetallic helicates [Ln2(LA)3]6+ in acetonitrile.17,18
The design has been expanded to a whole series of homobitopic
ligands LB,19 H2LC,20 LE,21 LF,21 LG,22 H2LC2,23 and H2LC3,24 the
latter two leading to efficient probes for cell imaging. All these
ligands afford a 9-coordinate chemical environment for both LnIII
ions derived from an idealised D3 symmetry and similar to the
one found for aqua-ions. Increasing the number of tridentate
coordination sites gives ligands forming triple-stranded helicates
with three25 or four26 lanthanide ions.
For assembling heterometallic bifunctional edifices, the ligand
design is inspired by studies on monometallic [Ln(L1,2)3]3+ com-
plexes with tridentate monotopic ligands: L1 (Scheme 1)27 has a
preference for the larger and mid-size lanthanide ions, a behaviour
most likely associated with a supramolecular size-discriminating
effect, the coordination cavity being stabilised by interstrand
´
Laboratory of Lanthanide Supramolecular Chemistry, Ecole Polytechnique
Fe´de´rale de Lausanne, BCH 1402, CH-1015, Lausanne, Switzerland.
E-mail: jean-claude.bunzli@epfl.ch; Fax: +41 21 69 39 825; Tel: +41 21
69 39 821
† CCDC reference numbers 663485–663487. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b715672c
‡ Electronic supplementary information (ESI) available: Assignment of
NMR spectra of ligands, analysis of the coordination polyhedra and
aromatic stacking interactions, speciation of homo- and heterobimetallic
complexes in solution. See DOI: 10.1039/b715672c
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1027–1036 | 1027
©

View Article Online
p–p interactions. In contrast, other ligands, exemplified by L2,
form stronger complexes with the smaller lanthanide ions if they
display any selectivity at all.28 Based on this, the heterobitopic
ligand LAB1 (Scheme 2)29 was designed to form heterobimetallic
complexes of composition [Ln1Ln2(LAB1)3]6+ when reacted with
a heteropair of lanthanide ions, the yield of which depends on
the difference in ionic radii of the metal ions. For the heteropair
La/Lu, maximising this difference, the yield of the hetero species
exceeds 90%. It, however, decreases to about 50% for pairs of
lanthanide ions more similar in size.
Experimental
Spectroscopic measurements
MS spectra used for the characterization of organic compounds
were recorded in CH3OH or CH3CN with a Finnigan SSQ-710C
spectrometer. 1D H NMR spectra, as well as 2D COSY and
ROESY experiments, were performed on a Bruker Avance 400
(400 MHz) spectrometer. Chemical shift values are given in ppm
using TMS as reference; J values are given in Hz.
1
Preparation of ligands
Solvents and starting materials were purchased from Fluka, Acros
or Aldrich and used without further purification, unless otherwise
stated. Acetonitrile and dichloromethane were distilled from
CaH2; thionyl chloride was distilled from elemental sulfur. Silica
gel (Merck 60, 0.04–0.06 mm) was used for preparative column
chromatography. Duplicate elemental analyses were performed by
Dr H. Eder from the Microchemical Laboratory of the University
of Geneva. Compounds 1,30 2,31 521 and 1029 were prepared
according to published procedures.
Scheme 2 Structures of heterobitopic ligands with the proton numbering
used in NMR assignments.
Synthesis of compound 3
A solution of 1 (3.18 g; 10.8 mmol), 25 mL of SOCl2 and 5 drops of
DMF in 100 mL of dry CH2Cl2 was refluxed under N2 for 2 h. The
solvents were evaporated and the residue was dried in a vacuum
for 1 h and re-dissolved in 50 mL of dry CH2Cl2. A solution of 2
(3.09 g; 18.6 mmol) and 10 mL of NEt3 in 50 mL of dry CH2Cl2 was
added dropwise. The solution was refluxed for 1 h, evaporated to
dryness and partitioned between 120 mL half saturated aqueous
NH4Cl solution and 120 mL CH2Cl2. The aqueous phase was
extracted with CH2Cl2 (3 × 100 mL) and the combined organic
phases were dried over Na2SO4, filtered and evaporated to dryness.
The crude product was purified on a column (silica gel, CH2Cl2–
CH3OH = 100 : 0 → 96 : 4) to give pure 3 (1.79 g, 37%). 1H NMR
In the initial heterobitopic ligand LAB1 (Scheme 2) the
benzimidazole-pyridine-benzimidazole (bpb) moiety, based on
ligand L1, complexes preferentially with the larger LnIII ions
when reacted with a pair of different trivalent lanthanide ions.
In solution as well as in the solid state the benzimidazole-
pyridine-carboxamide (bpa) moiety is always found to bind to
the smaller LnIII ion.29 The electron density of the nitrogen
donor atom of the bpa pyridine was modified in ligands LAB2
and LAB3 by introducing NEt2 (electron donating) and Cl (electron
withdrawing), respectively, in the para position of the pyridine.30
The increased electron density in LAB2 was expected to improve
the preference of the bpa moiety of the ligand for the smaller and
harder LnIII ions, leading to an improved overall selectivity of the
ligand; the opposite effect was expected for LAB3. Unexpectedly, the
substituents also influenced the yield of HHH (head-head-head)
isomer in which the three ligand strands of the [Ln1Ln2(LABX )3]6+
complex are pointing in the same direction. This eventually led
to lower selectivity of LAB2 for heteropairs of LnIII ions, since
this ligand preferentially forms the HHT (head-head-tail) isomer,
in which one ligand strand is oriented oppositely with respect
to the two others. Consequently, we explore here the effect of
substitution in the para position of the bpb pyridine leading to
ligands LAB4 and LAB5. The aim of the electron withdrawing Cl
in LAB5 is to reduce the electron density of the nitrogen donor
atom, thus increasing the preference of this complexation unit for
larger and softer over smaller and harder LnIII ions. LAB4, with an
electron donating NEt2 group, is designed as a control ligand for
which the final selectivity is not expected to be improved.
3
4
3
(CDCl3): d 7.94 (dd, 1H, J = 8.3, J = 1.5), 7.46 (td, 1H, J =
7.7, 4J = 1.5), 7.35 (td, 1H, 3J = 7.8, 4J = 1.3), 7.19 (dd, 1H, 3J =
8.0, 4J = 1.2), 6.88 (d, 1H, 4J = 2.7), 6.44 (d, 1H, 4J = 2.6), 4.35
(sextet, 1H, J = 7.1), 3.55 (sextet, 1H, J = 7.1), 3.41 (sextet, 1H,
J = 7.1), 3.32 (q, 4H, J = 7.1), 3.25 (sextet, 1H, J = 7.1), 3.13
(sextet, 1H, J = 7.3), 2.97 (sextet, 1H, J = 7.3), 1.25 (t, 3H, 3J =
7.2), 1.19 (t, 3H, 3J = 7.2), 1.11 (t, 6H, 3J = 7.0), 0.91 (t, 3H, 3J =
7.0). MS (CH3CN): m/z: 442.4 ([M + H]+, calcd 442.2).
Synthesis of compound 4
0.99 g of 3 (2.24 mmol), 1.8 g of Fe powder, 5 mL of 25%
HCl, 10 mL of H2O and 45 mL of EtOH was refluxed under
N2 overnight. Excess Fe was filtered off and EtOH was removed
by evaporation. The solution was mixed with 100 mL of CH2Cl2
and 70 g of Na2H2edta·2H2O in 200 mL H2O. Solid KOH was
added up to a pH value of 7. 10 mL of 30% H2O2 was added
dropwise and the pH was adjusted to 8.5 with solid KOH. After
stirring for 30 min the phases were separated and the aqueous
phase was extracted with CH2Cl2 (4 × 75 mL). The combined
organic phases were dried over Na2SO4, filtered and evaporated to
dryness. The crude product was purified on a column (silica gel,
CH2Cl2–CH3OH = 100 : 0 → 97 : 3) to yield 4 (407 mg, 46%). 1H
In this paper, we modify the coordination strength of the
bpb moiety, reporting the synthesis of ligands LAB4 and LAB5
(Scheme 2). The speciation of the resulting helicates is investi-
1
gated in acetonitrile by means of H NMR. Furthermore, three
homobimetallic complexes were crystallised and characterised by
X-ray diffraction.
1028 | Dalton Trans., 2008, 1027–1036
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
3
4
NMR (CDCl3): d 7.82 (dd, 1H, J = 7.0, J = 2.2), 7.50 (d, 1H,
4J = 2.7), 7.44 (dd, 1H, 3J = 7.0, 4J = 2.0), 7.31 (td, 1H, 3J = 7.3,
4J = 1.6), 7.28 (td, 1H, 3J = 7.2, 4J = 1.6), 6.73 (d, 1H, 4J = 2.7),
3J = 7.0), 3.62 (q, 4H, 3J = 7.0), 1.62 (t, 3H, 3J = 6.9), 1.28 (t, 3H,
3J = 7.0). MS (CH3CN): m/z: 339.3 ([M + H]+, calcd 339.2).
3
3
3
4.74 (q, 2H, J = 7.1), 3.58 (q, 2H, J = 7.1), 3.48 (q, 4H, J =
7.1), 3.39 (q, 2H, 3J = 7.1), 1.44 (t, 3H, 3J = 7.1), 1.27 (t, 3H, 3J =
7.1), 1.22 (t, 6H, 3J = 7.1), 1.08 (t, 3H, 3J = 7.1). MS (CH3CN):
m/z: 394.3 ([M + H]+ calcd 394.3).
Synthesis of compound 9
A solution of 7 (429 mg; 1.30 mmol) in 30 mL of 1 M KOH was
refluxed overnight. After cooling down, the solution was extracted
with 2 × 50 mL CH2Cl2 and the pH was adjusted to 1 with 25%
HCl. A white precipitate of 9 formed and was filtered off, washed
with aqueous HCl (pH = 2) and dried in a vacuum (257 mg, 66%).
1H NMR (DMSO-d6): d 8.52 (d, 1H, 4J = 2.0), 8.13 (d, 1H, 4J =
Synthesis of compound 6
A solution of 5 (1.50 g; 6.96 mmol), 20 mL of SOCl2 and 0.1 mL
of DMF in 75 mL of freshly distilled CH2Cl2 was refluxed under
N2 for 75 min. After evaporation of the solvent, the residue was
dried in a vacuum for 25 min and then re-dissolved in 75 mL of
dry CH2Cl2. To this solution was added a solution of 2 (1.16 g;
6.96 mmol) and 8 mL of NEt3 in 75 mL of dry CH2Cl2 before
refluxing for 60 min. The solvent was evaporated and the residue
partitioned between 100 mL of CH2Cl2 and 100 mL of half
saturated aqueous NH4Cl. The aqueous phase was extracted with
CH2Cl2 (2 × 100 mL) and the combined organic extracts were
dried over Na2SO4, filtered and evaporated to dryness. The product
was purified on a column (silica gel, CH2Cl2–CH3OH = 100 : 0 →
98 : 2) to give 6 (2.18 g, 86%). 1H NMR (CDCl3): d 8.10 (d, 1H,
4J = 1.8), 8.04 (dd, 1H, 3J = 8.2, 4J = 1.5), 7.91 (d, 1H, 4J = 1.9),
7.57 (td, 1H, 3J = 7.7, 4J = 1.5), 7.41 (td, 1H, 3J = 7.9, 4J = 1.5),
3
3
1.9), 7.75 (d, 1H, J = 8.0), 7.70 (d, 1H, J = 8.1), 7.36 (td, 1H,
3J = 7.6, 4J = 1.2), 7.29 (td, 1H, 3J = 7.6, 4J = 1.2), 4.89 (q, 2H,
3J = 7.1), 1.41 (t, 3H, J = 7.0). H NMR (CD3OD): d 8.50 (d,
3
1
4
4
1H, J = 1.8), 8.28 (d, 1H, J = 1.8), 7.75 (m, 2H), 7.50 (t, 1H,
3J = 7.4), 7.45 (t, 1H, J = 7.8), 4.96 (q, 2H, J = 7.2), 1.57 (t,
3H, 3J = 7.0). 1H NMR (CD3CN): d 8.58 (d, 1H, 4J = 1.6), 8.24
3
3
4
3
3
(d, 1H, J = 1.8), 7.83 (d, 1H, J = 7.8), 7.71 (d, 1H, J = 8.0),
7.48 (t, 1H, 3J = 7.7), 7.42 (t, 1H, 3J = 7.5), 4.87 (q, 2H, 3J = 7.1),
1.51 (t, 3H, 3J = 7.2). MS (CH3CN): m/z: 302.4 ([M + H]+ calc.
302.1). Anal. calcd for C15H12ClN3O2·1.5H2O: C, 54.8; H, 4.6; N,
12.8. Found: C, 55.1; H, 4.1; N, 12.6.
3
4
7.35 (dd, 1H, J = 7.8, J = 1.4), 4.23 (sextet, J = 7.1), 3.80 (s,
3H), 3.72 (sextet, J = 7.1), 1.25 (t, 3H, 3J = 7.2). MS (CH3CN):
m/z: 364.3 ([M + H]+, calcd 364.1).
Synthesis of compound 11
A solution of 8 (150 mg; 0.443 mmol), 5 mL of SOCl2 and 0.05 mL
of DMF in 50 mL of dry CH2Cl2 was refluxed under N2 for 1 h.
The solvent was evaporated and the residue was dried in vacuum
for 30 min and re-dissolved in 20 mL of dry CH2Cl2. A solution of
10 (550 mg; 1.00 mmol) and 1 mL of NEt3 in 30 mL of dry CH2Cl2
was added dropwise and the reaction mixture was refluxed for
30 min and then left stirring overnight at rt. After evaporation of
the solvent the residue was partitioned between 100 mL of CH2Cl2
and 100 mL of half saturated aqueous NH4Cl. The phases were
separated and the aqueous phase was extracted with 3 × 30 mL
CH2Cl2. The combined organic phases were dried over Na2SO4,
filtered and evaporated to dryness. The residue was purified on a
column (silica; CH2Cl2–CH3OH = 100 : 0 → 97 : 3) to give 11
(153 mg, 40%). MS (CH3CN): m/z: 869.3 ([M + H]+, calcd 869.4);
435.2 ([M + 2H]2+, calcd 435.2).
Synthesis of compound 7
A mixture of 6 (1.84 g; 5.02 mmol), 6 g of Fe powder, 150 mL
of EtOH, 40 mL of H2O and 30 mL of 25% HCl was refluxed
overnight under N2. Excess Fe was filtered off and EtOH was
evaporated on rotavapor. The solution was added to a mixture
of 100 mL of CH2Cl2 and 40 g of Na2H2edta in 150 H2O and
pH was adjusted to 7 with 5 M KOH. After addition of 3 mL
of 30% H2O2, 5 M KOH was added to pH = 9 and the mixture
was stirred for 40 min. After filtration, the phases were separated
and the aqueous phase was extracted with CH2Cl2 (6 × 100 mL).
The combined organic extracts were washed with 150 mL H2O,
dried over Na2SO4, filtered and evaporated to dryness. The crude
product was purified on a column (silica gel, CH2Cl2–CH3OH =
100 : 0 → 98 : 2) to yield pure 7 (429 g, 26%). 1H NMR (CDCl3):
d 8.68 (d, 1H, 4J = 1.7), 8.12 (d, 1H, 4J = 1.7), 7.84 (d, 1H, 3J =
7.9), 7.49 (d, 1H, 3J = 7.8), 7.38 (t, 1H, 3J = 7.0), 7.33 (t, 1H, 3J =
7.2), 4.92 (q, 2H, 3J = 7.1), 4.48 (q, 2H, 3J = 7.1), 1.58 (t, 3H, 3J =
7.1), 1.46 (t, 3H, 3J = 7.1). MS (CH3CN): m/z: 330.3 ([M + H]+,
calcd 330.1).
Synthesis of compound 12
A solution of 9 (291 mg; 0.964 mmol), 5 mL of distilled SOCl2
and 0.05 mL of DMF in 50 mL of dry CH2Cl2 was refluxed for
80 min. The solvent was evaporated and the residue was dried in
a vacuum for 60 min and dissolved in 20 mL of dry CH2Cl2. To
this solution was added a solution of 10 (1.00 g; 1.82 mmol) and
1.5 mL of distilled NEt3 in 30 mL of dry CH2Cl2. The reaction
mixture was refluxed for 2 h and evaporated to dryness. The residue
was partitioned between 75 mL of CH2Cl2 and 75 mL of half
saturated aqueous NH4Cl solution. After separation of the phases,
the aqueous phase was extracted with CH2Cl2 (3 × 75 mL) and
the combined organic phases were dried over Na2SO4, filtered and
evaporated to dryness. The residue was purified on column (silica;
CH2Cl2–CH3OH = 100 : 0 → 97 : 3) to yield 12 (357 mg, 44%).
MS: m/z = 832.3 ([M + H]+, calcd 832.3).
Synthesis of compound 8
A mixture of 4 (248 mg; 0.63 mmol), 20 g of KOH, 20 mL of EtOH
and 30 mL of H2O was refluxed for 2 d. The EtOH was evaporated
and the solution was extracted with CH2Cl2 (2 × 50 mL). The pH
value was adjusted to 2 with 25% HCl and a white precipitate
formed. This was filtered of, washed with aqueous HCl (pH = 2)
and dried in vacuum to give 8 (150 mg, 70%). 1H NMR (CDCl3):
d 8.12 (d, 1H, 3J = 7.3), 8.00 (s, 1H), 7.61 (d, 1H, 3J = 7.9), 7.53
(t, 1H, 3J = 7.5), 7.52 (t, 1H, 3J = 7.4), 7.47 (s, 1H), 4.90 (q, 2H,
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1027–1036 | 1029
©

View Article Online
Table 1 Crystallographic data for [Ln2(LAB4)3](ClO4)6·6CH3CN·(CH3)3COCH3
Pr2
Nd2
Sm2
Formula
Mol weight
Temp/K
C158H186Cl6N36O28Pr2
3531.95
140(2)
C158H186Cl6N36Nd2O28
3538.61
100(2)
C158H186Cl6N36O28Sm2
3550.83
140(2)
Crystal system
Space group
Monoclinic
P21/c
21.2566(15)
24.799(2)
31.792(2)
90.00
101.684(6)
90.00
16 412(2)
7320
Monoclinic
P21/c
21.324(4)
24.860(5)
31.875(6)
90.00
101.72(3)
90.00
16 545(6)
7328
Monoclinic
P21/c
21.3200(16)
24.6739(11)
31.934(2)
90.00
102.178(6)
90.00
16 420.6(18)
7344
˚
a/A
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
3
˚
V/A
F(000)
Z
4
4
4
Dc/Mg m−3
1.429
0.766
1.421
0.798
1.436
0.887
l(Mo-Ka)/mm−1
Crystal size/mm
Reflections measured
Unique reflections
R(int)
0.15 × 0.10 × 0.07
0.13 × 0.10 × 0.06
0.16 × 0.12 × 0.10
97 229
48 343
96 524
26 317
0.2211
961
0
0.714
0.0700
0.1763
14 755
0.1600
961
0
1.081
0.1046
0.2530
28 163
0.2021
961
0
0.838
0.1101
0.1858
No. of parameters
Constraints
b
GoF on F2
R1 [I > 2r(I)] a
wR2 a
a R = RꢀFo| − |Fcꢀ/R|Fo|, wR2 = [R [w(Fo − Fc2)2]/R [w(Fo2)2]]1/2
.
b GoF = [R [w(Fo − Fc2)2]/(n − p)]1/2
.
2
2
Synthesis of ligand LAB4
column (silica; CH2Cl2–CH3OH = 100 : 0 → 96 : 4) to yield
pure ligand LAB5 (154 mg, 50%). MS: m/z = 736.3 ([M + H]+,
calcd 736.3). Anal. calcd for C43H42ClN9O·H2O: C, 68.5; H, 5.9;
N, 16.7. Found: C, 68.8; H, 5.9; N, 16.7. See Table S2 (ESI) for the
assignment of the 1H NMR spectrum in CDCl3.
A mixture of 11 (147 mg; 0.169 mmol), 0.6 g of Fe powder, 30 mL
of EtOH, 10 mL of H2O and 5 mL of 25% HCl was refluxed under
N2 for 8 h. The EtOH was evaporated after removal of excess Fe.
The solution was mixed with 15 g of Na2H2edta·2H2O, 100 mL
of H2O and 75 mL of CH2Cl2 and the pH was adjusted to 7 with
solid KOH. Following addition of 4 mL of 30% H2O2, solid KOH
was added to pH = 8.5 and the mixture was stirred for 2 h. The
phases were separated and the aqueous phase was extracted with
3 × 40 mL CH2Cl2. The organic extracts were dried over Na2SO4,
filtered and evaporated to dryness. The crude product was purified
on a column (silica; CH2Cl2–CH3OH = 100 : 0 → 96 : 4) to yield
pure ligand LAB4 (58 mg, 44%). MS (CH3CN): m/z: 387.4 ([M +
2H]2+ calc. 387.2); 773.3 ([M + H]+ calc. 773.4). Anal. calcd for
C47H52N10O·H2O: C, 73.0; H, 6.8; N, 18.1. Found: C, 73.2; H, 6.7;
Synthesis of [Ln2(LABX )3](ClO4)6 and [Ln1Ln2(LABX )3](ClO4)6
(X = 4 or 5) complexes
Partially dehydrated perchlorate salts Ln(ClO4)3·xH2O (Ln =
La–Lu, x ≈ 2–4) were prepared from the corresponding oxides
(Rhoˆne-Poulenc, 99.99%) in the usual way.32 Caution! Perchlorate
salts combined with organic ligands are potentially explosive and
should be handled in small quantities and with adequate precau-
tions.33,34 Stock solutions of Ln(ClO4)3·xH2O (x ≈ 2–5) in CH3CN
were prepared by weighting. The concentrations of the solutions
were determined by complexometric titrations with Na2(H2edta)
in the presence of urotropine using xylene orange as the indicator.
NMR samples of homobimetallic [Ln2(LABX )3](ClO4)6 com-
plexes were prepared by reacting a weighed amount of LABX (3–
15 mg) dissolved in CH2Cl2 with 2/3 equivalents of Ln(ClO4)3·
xH2O in the form of a CH3CN solution. After stirring for 1–3 h
the solution was evaporated to dryness, the residue was dried in
vacuum at 50 ◦C and re-dissolved in 0.6 mL CD3CN. Samples of
heterobimetallic helicates [Ln1Ln2(LABX )3](ClO4)6 were prepared in
an analogous way using 1/3 equivalent [Ln1(ClO4)3·xH2O] and1/3
equivalent [Ln2(ClO4)3·xH2O] and stirring the sample overnight
before evaporation.
1
N, 18.2. See Table S1 (ESI) for the assignment of the H NMR
spectrum in CDCl3.
Synthesis of ligand LAB5
A mixture of 12 (350 mg; 0.421 mmol), 0.6 g of Fe powder,
30 mL of EtOH, 10 mL of H2O and 5 mL of 25% HCl was
refluxed under N2 for 7 h. After removal of excess Fe, EtOH was
removed by evaporation. The solution was mixed with 16.7 g of
Na2H2edta·2H2O, 150 mL of H2O and 75 mL of CH2Cl2 and the
pH value was adjusted to 7 with solid KOH. Following addition
of 4 mL of 30% H2O2, solid KOH was added to pH = 9 and
the solution was stirred for 30 min. The phases were separated
and the aqueous phase was extracted with CH2Cl2 (3 × 75 mL).
The combined organic extracts were dried over Na2SO4, filtered
and evaporated to dryness. The crude product was purified on
Crystals of homobimetallic triple helicate complexes of LAB4
were obtained by reacting 2/3 equivalents of Ln(ClO4)3 (Ln =
Pr, Nd, Sm) with LAB4 (5–15 mg) in CH3CN solution. After
evaporation to dryness and drying in vacuum the solid residues
1030 | Dalton Trans., 2008, 1027–1036
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
were re-dissolved in a 1 : 1 CH3CN : CH3CH2CN mixture
(≈0.5 mL) and precipitated by slow diffusion of tert-BuOMe at
T = −18 ◦C.
Crystal structure determination of [Ln2(LAB4)3](ClO4)6·6CH3CN·
(CH3)3COCH3 (Ln = Pr, Nd, Sm)†
Data collections for all compounds have been performed on an
Oxford Diffraction Sapphire/KM4 CCD equipped with a kappa
geometry goniometer. Data were treated for cell refinement and
integration using CrysAlis RED.35 No absorption correction was
applied to the obtained data sets. Structure solutions and refine-
ments have been carried out by SHELXTL.36 Crystal structures
have been refined using the full-matrix on F2. H atoms have been
placed in calculated positions by means of the “riding” model.
The entire structures (aside from the metal and chlorine atoms)
have been retained as isotropic because the crystals were very
weakly diffracting and any attempt to refine them as anisotropic
(in combination with suitable restraints) failed. A summary of the
general crystal data and refinement parameters is given in Table 1.
Results and discussion
Ligand design, synthesis and properties
The preparation of the two ligands is based on principles
developed previously for the synthesis of a rapidly growing family
of monotopic28 as well as homo-17–23,24 and heterobitopic29,30,37
ligands (Schemes 1 and 2). The key steps are the formation of the
asymmetric secondary amine 10, which reacts easily with the acid
chlorides of 8 or 9 to form 11 and 12. Closing of the benzimidazole
rings by a modified Phillips reaction38 leads to the targeted ligands
in good yields (Scheme 3).
The solution structure of the ligands was investigated by means
1
of H-NMR in CDCl3. NOE signals are observed between H24
and H26, H25 and H26, H24 and H27, H25 and H27, H4 and
H6, H12 and H13, H18 and H20 (see Scheme 2 for proton
numbering). The NOE signals observed between the hydrogen
atoms of the bridging methylene group (H9) and all neighbouring
benzimidazole hydrogen atoms (H7, H8, H10 and H11) are
indicative of free rotation about the methylene group. The absence
of signals between pyridine hydrogen atoms (H1, H3, H15 and
H17) and ethyl groups (e.g. H4, H5, H24 and H25) indicate
that the nitrogen atoms of the pyridine groups are oriented in
a transoid fashion with respect to the other potentially ligating
nitrogen atoms. The same kind of structure (free rotation around
the central CH2 group and transoid conformation of the pyridine
N atoms with respect to the neighbouring pair of potential ligating
atoms) has also been observed for the other ligands of this type.
Scheme 3 Synthesis of LAB4 and LAB5
.
are wrapped tightly around the two lanthanide ions in a helical
fashion (Fig. 2 and 3). The complex cations are present in a
racemic mixture of P and M isomers with the helix being either
left- or right-handed. The configuration of the ligand strands
is HHH meaning that all three carboxamide oxygen atoms are
coordinated to the same LnIII ion. It is noteworthy that HHT
helicates could not be crystallized with any of the synthesized
ligands so far, even with those yielding a high proportion of these
isomers in solution. Very weak aromatic stacking interactions
contribute to the stability of the complexes. The strongest of
Solid state structure of homobimetallic [Ln2(LAB4)3](ClO4)6
helicates
Both LAB4 and LAB5 react with 2/3 equivalent of LnIII ions to
form triple-stranded bimetallic helicates in acetonitrile solution.
We were able to isolate crystals of X-ray quality of the three
complexes [Ln2(LAB4)3](ClO4)6·solvent (Ln = Pr, Nd, Sm) (Table 1).
The three compounds are isostructural and the unit cell (Fig. 1)
contains solvent molecules, perchlorate ions and complex cations
of composition [Ln2(LAB4)3]6+. In the latter the three ligand strands
◦
˚
these interactions (≈3.5 A) is between two almost parallel (10–13 )
imidazole groups on adjacent ligand strands of the benzimidazole-
pyridine-benzimidazole moiety of the ligand (See Tables S3, S4 in
the ESI‡ for a full analysis of aromatic interactions). The helical
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1027–1036 | 1031
©

View Article Online
The coordination polyhedra around the lanthanide ions can
best be described as tricapped trigonal prisms in which bridging
benzimidazole nitrogen atoms and either carboxamide oxygen
atoms or terminal benzimidazole nitrogen atoms define the upper
and lower triangular faces of the prism. Pyridine nitrogen atoms
cap the rectangular faces of the prisms (Fig. 4). The most
significant deviation from completely regular tricapped trigonal
prisms is the twist angle x of the upper triangular face with respect
to the lower triangular face of the prism. In the coordination
compartment formed by the benzimidazole-pyridine-carboxamide
moieties of the three ligand strands this is found to be 15(3)◦
for the Pr2 and Nd2 complexes and 14(3)◦ for the Sm2 complex.
In the benzimidazole-pyridine-benzimidazole compartment the
angles found for the three complexes are 14(2)◦, 13(2)◦ and 12(1)◦,
respectively. The values are similar to what has been determined
29
30,37
for bimetallic complexes of LAB1 and LAB3
;
the decrease in
Fig.
1
Cell packing of [Pr2(LAB4)3](ClO4)6·6CH3CN·(CH3)3COCH3
viewed along the b axis.
twist angle with decreasing ionic radius of the LnIII ion is also
typical of this type of complex. A full analysis of the coordination
polyhedra including a graphical presentation of the twist angle is
given in the ESI‡ (Table S6, Chart S1).
Fig. 4 Two views of the bpa coordination polyhedron in the [Pr2(LAB4)3]6+
ion.
Metal–ligand Ln–X distances are listed in Table 2. Comparing
Fig. 2 The [Pr2(LAB4)3]6+ ion in the solid state (Pr green, O red, N blue; C
atoms of the three ligand strands are white, grey and black, respectively).
29
30,37
with the distances found in complexes of LAB1 and LAB3
it
is concluded that the introduction of a substituent (Cl in LAB3
,
NEt2 in LAB4) does not significantly change the Ln–X distances or
indeed any other structural parameter measured.
Speciation in CD3CN solution
Homobimetallic complexes
Reacting two equivalents of [Ln(ClO4)3·xH2O] with three equiv-
alents of LABX in acetonitrile solution gives the corresponding
[Ln2(LABX )3](ClO4)6 (X = 4 or 5) complexes, which after evap-
oration of solvent and drying can be dissolved in CD3CN in
preparation for NMR experiments. The spectra contain two
sets of signals with different intensities. One set consists of the
expected number of signals for a helicate with a set of three
equivalent ligand strands; this set of signals is assigned to the
HHH-[Ln2(LABX )3]6+ isomer evidenced in the solid state by X-ray
diffraction (Table S7; ESI‡). Diastereotopic methylene protons
rule out a linear structure and confirm that the helical structure
is maintained in solution. The other set of resonances contains
a larger number of signals; overlap of lines and the smaller
intensity of this set prevent a full analysis, except in a few cases.
An example of this is the 5.2–6.3 ppm range of the spectra of
the diamagnetic [La2(LABX )3]6+, [Y2(LABX )3]6+ and [Lu2(LABX )3]6+
homobimetallic helicates (Fig. 5). In the spectra of the free ligands
no signals are found in this region, but the helical wrapping of
Fig. 3 The [Pr2(LAB4)3]6+ ion viewed along the pseudo C3 axis. Atom
colouring as in Fig. 2.
pitch (the distance for the helix to do a full 360◦ twist) is 13.3–
˚
13.4 A in the three compounds; the Ln–Ln distance is 9.093–
˚
9.156 A (Table S5, ESI‡). Both of these values are in line with
what was found for complexes of LAB1 and LAB3
vary significantly with the size of the LnIII ion or the substituent
on the ligands.
and do not
29
30,37
1032 | Dalton Trans., 2008, 1027–1036
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 2 LnIII–X distances in [Ln2(LAB4)3](ClO4)6·6CH3CN·(CH3)3COCH3
Terminal
benzimidazole
Bpb unit
pyridine
Bridging
benzimidazole
Bridging
benzimidazole
Bpa unit
pyridine
Carboxamide
Pr2
N1
2.68(1)
N3
2.58(1)
2.62(1)
2.66(1)
2.62(4)
N4
2.66(1)
N6
2.72(1)
N8
2.59(1)
2.64(1)
2.68(1)
2.63(5)
O1
O2
O3
Mean
2.45(1)
2.42(1)
2.416(9)
2.43(2)
N11
N21
Mean
2.65(1)
2.63(1)
2.65(3)
N13
N23
Mean
N14
N24
Mean
2.64(1)
2.69(1)
2.66(3)
N16
N26
Mean
2.63(1)
2.60(1)
2.65(6)
N18
N28
Mean
Nd2
Sm2
N1
2.69(2)
2.67(2)
2.62(2)
2.66(3)
N3
2.56(1)
2.60(1)
2.61(2)
2.59(3)
N4
2.64(2)
2.64(2)
2.68(2)
2.65(2)
N6
2.71(2)
2.67(2)
2.65(2)
2.67(3)
N8
2.62(2)
2.67(1)
2.71(1)
2.67(5)
O1
O2
O3
Mean
2.45(1)
2.42(1)
2.44(1)
2.44(2)
N11
N21
Mean
N13
N23
Mean
N14
N24
Mean
N16
N26
Mean
N18
N28
Mean
N1
2.62(1)
2.63(1)
2.59(1)
2.61(2)
N3
2.53(1)
2.54(1)
2.57(1)
2.55(2)
N4
2.57(1)
2.57(1)
2.65(1)
2.60(5)
N6
2.68(1)
2.59(1)
2.59(1)
2.62(5)
N8
2.53(1)
2.61(1)
2.62(1)
2.59(5)
O1
O2
O3
Mean
2.42(1)
2.39(1)
2.37(1)
2.39(3)
N11
N21
Mean
N13
N23
Mean
N14
N24
Mean
N16
N26
Mean
N18
N28
Mean
discussion of the data, we take the percentages of the HHH-
[Ln2(LAB1)3]6+ helicates as a reference. Comparing the weakly
electron withdrawing Cl substituent with a Hammett coefficient
r = +0.23 with the strongly electron donating NEt2 substituent
with r = −0.73 (LAB2 vs. LAB3; LAB4 vs. LAB5) reveals opposite effects
for the two substituents: LAB2 lowers HHH yields, while LAB3
increases HHH yield. Moreover, the effect is stronger for NEt2
than for Cl. Finally, a substituent induces the opposite effect when
introduced on the bpb instead of the bpa moiety of the ligand
(LAB2 vs. LAB4; LAB3 vs. LAB5). The yields of HHH complexes thus
follow systematically from the sign of the Hammett coefficient of
the substituent and the position of the latter on either the bpb
or the bpa moiety of the ligand. Remarkably, most percentages
deviate from the statistical distribution of 25% HHH and 75%
HHT.
Fig. 5 Partial 1H NMR spectrum of [La2(LAB5)3]6+ in CD3CN. Signals of
the HHT isomer are indicated with *.
the ligand strands brings protons H8 and H10 in the vicinity
of aromatic benzimidazole groups of an adjacent ligand strand,
causing the signals to be shifted away from their usual location in
the spectrum (7.7–7.8 ppm). Apart from the two signals from
the three equivalent ligand strands of the HHH isomer, this
region contains six additional signals assigned to the H8 and H10
protons of the three non-equivalent ligand strands of the head-
head-tail (HHT) isomer in which one ligand strand is oriented in
the opposite direction of the other two. Spectra of paramagnetic
[Ln2(LABX )3]6+ complexes (Ln = Y, La, Lu) are less straightforward
to interpret since signals are shifted up- or downfield depending
on their magnetic interaction with the unpaired electrons of the
LnIII ions. Where it is possible to accurately count the number of
signals the result is invariably that the less intense set contains
approximately three times as many signals as the more intense
signal, again leading to the conclusion that the solution contains
a mixture of HHH and HHT isomers.
Heterobimetallic complexes
1
CD3CN solutions for H NMR experiments of overall compo-
sition [Ln1Ln2(LABX )3](ClO4)6 (Ln1 = Ln2; X = 4 or 5) were
obtained like the corresponding homobimetallic complexes from
the reaction of one equivalent of each of the lanthanide salts with
three equivalents of LABX . In the stoichiometric solutions of total
1
complex concentration ≈10−2 M the signals observed in the H
NMR spectra can be attributed to a mixture of different species,
all of which are bimetallic triple-stranded complexes composed of
two LnIII ions and three ligand strands. Furthermore, the chemical
shifts of the protons H8 and H10 confirm that the helical wrapping
of the ligands observed in the solid state is maintained in solution.
In the diamagnetic complexes the signals of these protons are
invariably found in the 5.2–6.3 ppm region, which (as discussed
for the homobimetallic complexes) can only be explained by a
tight wrapping of the ligand strands around the LnIII ions. In the
complexes containing paramagnetic LnIII ions the signals of these
protons are shifted considerably, in accordance with their close
proximity to the paramagnetic centres. Relevant assignments of
the spectra are given in Tables S9 and S10 (ESI‡), while parts of
typical spectra are shown on Fig. 6. A complete analysis of the
lanthanide induced paramagnetic shifts has also been carried out39
and will be published separately.
From the integrated intensities of peaks of the two sets of
signals the percentages of the two isomers can easily be calculated;
the results are given in Table S8 (ESI‡) for complexes of LnIII
ions spanning the whole lanthanide series; values for complexes
of LAB1, LAB2 and LAB3 are included for comparison. Two obser-
vations are of particular interest here. Firstly, the percentage of
HHH isomer is remarkably constant for a given ligand regardless
of the LnIII ion (LAB1: 63–73%; LAB2: 6–20%; LAB3: 79–87%; LAB4
:
93–96%; LAB5: 53–61%). Secondly, the Cl and NEt2 substituents
have a significant influence on these percentages. For the following
In total, eight different complexes can be formed from the
stoichiometric mixture Ln1 : Ln2 : 3LABX (Fig. 7, left). The first
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1027–1036 | 1033
©

View Article Online
four are the HHH and HHT isomers of the homobimetallic
complexes containing either two Ln1 or two Ln2 ions. For the
hetero complexes, the HHH configuration can occur for both
[Ln1Ln2(LABX )3]6+ and [Ln2Ln1(LABX )3]6+ complexes; in each case,
the LnIII ion mentioned first is in the bpb nonadentate compart-
ment formed by the three ligand strands. By inverting one of the lig-
and strands, these two helicates give rise to HHT-[Ln1Ln2(LABX )3]6+
and HHT-[Ln2Ln1(LABX )3]6+ isomers, respectively (Fig. 7, right).
It should be emphasised that the statistical percentages given in
Fig. 7 are not theoretical predictions. Instead, they refer to the
hypothetical situation of a heterobitopic ligand exhibiting no
inherent selectivity and reacting with a pair of non-distinguishable
LnIII ions, far from the present circumstances; these percentages
are only included as references for further comparison.
Actually, the speciation of the LABX complexes departs from
the hypothetical statistical distribution in a remarkable manner
(Fig. 8; Tables S11, S12, and S13; ESI‡). For LAB1 the percentage
of hetero complexes varies from ≈50% for neighbouring pairs of
lanthanide ions to ≈95% for the La/Lu pair. Furthermore, the
relative percentages of HHH and HHT are far from statistical; no
HHT isomers of hetero complexes have been observed. Finally,
of the two possible HHH hetero complexes, the only isomer
usually observed is the one with the larger LnIII ion in the
bpb compartment of the helicate. Exceptions occur when the
pair of lanthanide ions have very similar sizes, but even for
the adjacent LaIII/CeIII pair, the HHH-[LaCe(LAB1)3]6+ isomer
largely dominates over the HHH-[CeLa(LAB1)3]6+ helicate. In the
ligand design for LABX (X = 2–5), introduction of an electron
donating NEt2 group on the bpa moiety of LAB2 and of a electron
withdrawing Cl substituent on the bpb moiety of LAB5 was expected
to modify the electron density of the corresponding pyridine
nitrogen donor atom and improve the yield of hetero complexes.
With the same reasoning, the yield of the hetero species was
predicted to decrease with ligands LAB3 and LAB4 in which the
same two substituents are introduced on the opposite moieties of
the ligands. As can be seen from Fig. 8 in which the percentages
of the heterobimetallic species are plotted against the differences
in ionic radii Dri, no important improvement in the yield of the
hetero species compared to LAB1 is observed for any of the four
substituted ligands. The behaviour of the three ligands LABX (X =
3–5) deviates somewhat compared to LAB1, but the general trend
Fig. 6 1H NMR spectra of mixtures of overall composition La : Pr :
3LAB4 (top) and La : Lu : 3LAB5 (bottom) with partial assignment used to
calculate the percentages of the species in solution.
Fig. 7 Schematic representation of the eight possible isomers in a mixture
of overall composition Ln1 : Ln2 : LABX = 1 : 1 : 3; the percentages are not
meant as predictions of the actual distribution of the species.
Fig. 8 Percentages of heterobimetallic [Ln1Ln2(LABX )3]6+ complexes in
CD3CN.
1034 | Dalton Trans., 2008, 1027–1036
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
is maintained, in that selectivity increases with Dri. While keeping
up with this trend, LAB2 displays a markedly different behaviour
in that the concentration of the hetero species remains very low
(20–65%). Differences between the four ligands with X = 1, 3,
4 and 5 are maximum for Dri ≈8–15 pm and minimum for both
smaller and larger values of Dri.
tuned in a predictable and conceptually straightforward manner,
while others still escape complete control. However, progress
in the understanding of the factors leading to the stability of
helical polymetallic assemblies in solution,40,41 particularly of the
weaker interactions contributing to the stability of supramolecular
edifices, should soon lead to improved handling of these problems.
This unexpected behaviour can be explained with reference
to the concentrations of the HHH isomer determined for the
homobimetallic helicates. Indeed, for a large selectivity of the
ligand towards a heteropair of lanthanide ions, it is important
that the ligand has a tendency to organise in the HHH config-
uration, which is consistent with the non-observation of HHT
Acknowledgements
This project is supported through grants from the Swiss National
Science Foundation.
isomers for heterobimetallic helicates. Both ligands (LAB2 and LAB5
)
expected to give improved yields of hetero complexes compared
to LAB1 eventually gave lower yields. LAB2 forms very low HHH
yields of homobimetallic helicates (6–20%) compared to 63–73%
for LAB1, while LAB5 induces only slightly smaller yields (53–61%).
Concomitantly, the La/Lu solution with LAB2 contains only 65%
of the heterobimetallic species, as compared to 96% for LAB1 and
92% for LAB5. On the other hand, an opposite behaviour is observed
for LAB3 (79–87% of HHH homobimetallic isomer) and LAB4 which
leads almost exclusively to the HHH isomers (93–96%). While the
former ligand results in concentrations of hetero species consistent
with the proportion of HHH isomers (e.g. 87% for the La/Lu
pair), LAB4 deviates from expectations. It is only slightly more
selective than LAB1 for Dri ≈ 2–5 pm and less for Dri > ≈ 5 pm:
the proportion of hetero helicate amounts to only 79% for the
La/Lu pair. The results obtained for LAB3 and LAB4 confirm that
the ligand design strategy is well-founded in the sense that these
two ligands as predicted give lower yields of heterobimetallic
complexes, despite higher HHH yields.
References
1 J.-C. G. Bu¨nzli and C. Piguet, Chem. Soc. Rev., 2005, 34, 1048–1077.
2 S. Comby and J.-C. G. Bu¨nzli, in Handbook on the Physics and
Chemistry of Rare Earths, ed. K. A. Gscheidner, Jr., J.-C. G. Bu¨nzli
and V. K. Pecharsky, Elsevier Science B. V., Amsterdam, 2007, vol. 37.
3 I. Hemmila¨ and V. Laitala, J. Fluoresc., 2005, 15, 529–542.
4 K. Nishioka, K. Fukui and K. Matsumoto, in Handbook on the Physics
and Chemistry of Rare Earths, ed. K. A. Gscheidner, Jr., J.-C. G. Bu¨nzli
and V. K. Pecharsky, Elsevier Science B. V., Amsterdam, 2007, vol. 37.
5 S. Pandya, J. Yu and D. Parker, Dalton Trans., 2006, 2757–2766.
6 R. J. Aarons, J. K. Notta, M. M. Meloni, J. Feng, R. Vidyasagar, J.
Narvainen, S. Allan, N. Spencer, R. A. Kauppinen, J. S. Snaith and S.
Faulkner, Chem. Commun., 2006, 909–911.
7 M. M. Hu¨ber, A. B. Staubli, K. Kustedjo, M. H. B. Gray, J. Shih, S. E.
Fraser, R. E. Jacobs and T. J. Meade, Bioconjugate Chem., 1998, 9,
242–249.
8 H. C. Manning, T. Goebel, R. C. Thompson, R. R. Price, H. Lee and
D. J. Bornhop, Bioconjugate Chem., 2004, 15, 1488–1495.
9 I. Nasso, C. Galaup, F. Havas, P. Tisne`s, C. Picard, S. Laurent,
L. Vander Elst and R. N. Muller, Inorg. Chem., 2005, 44, 8293–8305.
10 C. Picard, N. Geum, I. Nasso, B. Mestre, P. Tisne`s, S. Laurent, R. N.
Muller and L. Vander Elst, Bioorg. Med. Chem. Lett., 2006, 16, 5309–
5312.
11 Q. Zheng, H. Dai, M. E. Merritt, C. Malloy, C. Y. Pan and W.-H. Li,
J. Am. Chem. Soc., 2005, 127, 16178–16188.
Conclusions
12 L. J. Martin, M. J. Ha¨hnke, M. Nitz, J. Wo¨hnert, N. R. Silvaggi, K. N.
Allen, H. Schwalbe and B. Imperiali, J. Am. Chem. Soc., 2007, 129,
7106–7113.
The effect of electron donating and electron withdrawing sub-
stituents grafted on the pyridine of the bpb tridentate coordi-
nation unit of LAB1 have been compared to those induced by
similar substitutions of the bpa coordination site.30 The resultant
modification of the electron density of the pyridine N-donor
atoms induces systematic effects on the proportions of HHH
isomers in the homobimetallic solutions and of hetero helicates
13 J.-P. Costes, F. Dahan, A. Dupuis, S. Lagrave and J.-P. Laurent, Inorg.
Chem., 1998, 37, 153–155.
14 J. W. Buchler, A. de Cian, J. Fischer, M. Kihn-Botulinski and R. Weiss,
Inorg. Chem., 1988, 27, 339–345.
15 N. Ishikawa, T. Iino and Y. Kaizu, J. Am. Chem. Soc., 2002, 124, 11440–
11447.
16 V. Jacques and J. F. Desreux, Top. Curr. Chem., 2002, 221, 123–164.
17 G. Bernardinelli, C. Piguet and A. F. Williams, Angew. Chem., Int. Ed.
Engl., 1992, 31, 1622–1624.
18 C. Piguet, J.-C. G. Bu¨nzli, G. Bernardinelli, G. Hopfgartner and A. F.
Williams, J. Am. Chem. Soc., 1993, 115, 8197–8206.
19 N. Martin, J.-C. G. Bu¨nzli, V. McKee, C. Piguet and G. Hopfgartner,
Inorg. Chem., 1998, 37, 577–589.
20 M. Elhabiri, R. Scopelliti, J.-C. G. Bu¨nzli and C. Piguet, J. Am. Chem.
Soc., 1999, 121, 10747–10762.
21 C. P. Iglesias, M. Elhabiri, M. Hollenstein, J.-C. G. Bu¨nzli and C. Piguet,
J. Chem. Soc., Dalton Trans., 2000, 2031–2043.
22 R. Tripier, M. Hollenstein, M. Elhabiri, A.-S. Chauvin, G. Zucchi, C.
Piguet and J.-C. G. Bu¨nzli, Helv. Chim. Acta, 2002, 85, 1915–1929.
23 C. D. B. Vandevyver, A.-S. Chauvin, S. Comby and J.-C. G. Bu¨nzli,
Chem. Commun., 2007, 1716–1718.
24 A.-S. Chauvin, S. Comby, B. Song, C. D. B. Vandevyver, F. Thomas
and J.-C. G. Bu¨nzli, Chem.–Eur. J., 2007, 13(34), 9515–9526.
25 S. Floquet, N. Ouali, B. Bocquet, G. Bernardinelli, D. Imbert, J.-C. G.
Bu¨nzli, G. Hopfgartner and C. Piguet, Chem.–Eur. J., 2003, 9, 1860–
1875.
in Ln1/Ln2 heterobimetallic solutions for ligands LAB2, LAB3
,
and LAB4, compared to LAB1. The main factor here is the difference
in coordination strength of the two tridentate coordination units.
We have shown previously that if this difference is too large, a large
proportion of HHT isomer occurs, detrimental to the selectivity
for a hetero pair of lanthanide ions.29 The two substituents chosen
for this study have weak effects, in line with the very small
energy difference between HHT and HHH isomers (DG◦ ≈0–
7 kJ mol−1).37 Since the bpb coordinating unit is less coordinating
than the bpa unit, it is more affected by the substitution,
particularly by NEt2. In fact, LAB4 was predicted to yield less
HHH isomers since the coordination strength between the two
coordinating units is more equalized. This is by far not the case,
but we note that due to less difference in the coordination ability of
the bpb and bpa units, the proportion of the Ln2Ln1 isomer (with
respect to Ln1Ln2) is twice as large for LAB4, compared to LAB5 for
Dri < 5.3 pm. This shows how the self-assembly of supramolecular
structures depends on subtle effects. Some of them can be fine-
26 K. Zeckert, J. Hamacek, J.-M. Senegas, N. Dalla-Favera, S. Floquet,
G. Bernardinelli and C. Piguet, Angew. Chem., Int. Ed., 2005, 44, 7954–
7958.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1027–1036 | 1035
©

View Article Online
27 S. Petoud, J.-C. G. Bu¨nzli, F. Renaud, C. Piguet, K. J. Schenk and G.
Hopfgartner, Inorg. Chem., 1997, 36, 5750–5760.
28 T. Le Borgne, P. Altmann, N. Andre´, J.-C. G. Bu¨nzli, G. Bernadinelli,
P.-Y. Morgantini, J. Weber and C. Piguet, Dalton Trans., 2004, 723–733.
29 N. Andre´, T. B. Jensen, R. Scopelliti, D. Imbert, M. Elhabiri, G.
Hopfgartner, C. Piguet and J.-C. G. Bu¨nzli, Inorg. Chem., 2004, 43,
515–529.
33 W. C. Wosley, J. Chem. Educ., 1973, 50, A335.
34 K. N. Raymond, Chem. Eng. News, 1983, Dec 12, 2.
35 CrysAlis RED Version 1.7.0, Oxford Diffraction Ltd., Abingdon, OX14
4RX, Oxfordshire, UK2003.
36 SHELXTL Release 5.1, Bruker AXS, Madison, WI 53719, USA, 1997.
37 T. B. Jensen, R. Scopelliti and J.-C. G. Bu¨nzli, Chem.–Eur. J., 2007, 13,
8404–8410.
30 T. B. Jensen, R. Scopelliti and J.-C. G. Bu¨nzli, Inorg. Chem., 2006, 45,
38 M. A. Phillips, J. Chem. Soc., 1928, 172–177.
39 T. B. Jensen, PhD thesis, Ecole Polytechnique Fe´de´rale de Lausanne
´
7806–7814, (Inorg. Chem., 2007, 46, 1507).
31 B. Lamm, Acta Chem. Scand., 1965, 19, 2316–2322.
(EPFL), 2006.
32 J. F. Desreux, in Lanthanide Probes in Life, Chemical and Earth Sciences.
Theory and Practice, ed. J.-C. G. Bu¨nzli and G. R. Choppin, Elsevier
Science Publ. B. V., Amsterdam, The Netherlands, 1989.
40 G. Canard and C. Piguet, Inorg. Chem., 2007, 46, 3511–3522.
41 J. Hamacek, M. Borkovec and C. Piguet, Dalton Trans., 2006, 1473–
1490.
1036 | Dalton Trans., 2008, 1027–1036
This journal is
The Royal Society of Chemistry 2008
©