Utilization of self-sorting processes to generate dynamic combinatorial libraries with new network topologies

Article abstract of DOI:10.1002/chem.200500621

The synthesis of water-soluble, organometallic macrocycles is described. They were obtained by self-assembly in reactions of the half-sandwich complexes [{Ru(C₆H₅Me)Cl₂}₂], [{Ru(p-cymene)Cl₂}₂

Full text of DOI:10.1002/chem.200500621

DOI: 10.1002/chem.200500621
Utilization of Self-Sorting Processes To Generate Dynamic Combinatorial
Libraries with New Network Topologies
Isabelle Saur, Rosario Scopelliti, and Kay Severin*[a]
Abstract: The synthesis of water-solu-
ble, organometallic macrocycles is de-
scribed. They were obtained by self-as-
sembly in reactions of the half-sand-
wich complexes [{Ru(C6H5Me)Cl2}2],
[{Ru(p-cymene)Cl2}2], [{Rh(Cp)Cl2}2],
and [{Ir(Cp*)Cl2}2] with the ligand 5-di-
methylaminomethyl-3-hydroxy-2-
methyl-4-(1H)-pyridone in buffered
aqueous solution at pH 8. The structure
of the Ru–(p-cymene) complex was de-
termined by single-crystal X-ray crys-
tallography. Upon mixing, these com-
plexes undergo scrambling reactions to
give dynamic combinatorial libraries.
In combination with structurally relat-
ed complexes based on amino-meth-
ylated 3-hydroxy-2-(1H)-pyridone li-
gands, an exchange of metal fragments
but no mixing of ligands was observed.
This self-sorting behavior was used to
construct dynamic combinatorial libra-
ries of macrocycles, in which two four-
component sub-libraries are connected
by two common building blocks. This
type of network topology influences
the adaptive behavior of the library as
demonstrated in selection experiments
with lithium ions as the target.
Keywords: combinatorial chemis-
try · iridium · macrocycles · recep-
tors · ruthenium
Introduction
mainly focused on two types of libraries. In the first type of
DCL, a symmetrical coupling chemistry is employed. Conse-
quently, each building block can assemble with each other
building block and with itself (Scheme 1a). Systems of this
Starting with seminal publications by the groups of Sand-
ers,[1] Lehn,[2] and others[3] in the mid 1990s, dynamic combi-
natorial chemistry has emerged as a powerful tool for the
discovery of new receptors, drugs, catalysts, and materials.[4]
The adaptive behavior of dynamic combinatorial libraries
(DCLs) has received particular attention in this context.
DCLs are formed by combinatorial assembly[5] of molecular
building blocks under thermodynamic control. They repre-
sent chemical networks that are able to respond to changes
in their environment.[6] Upon addition of a target molecule
that selectively interacts with some members of the library,
a re-equilibration occurs. This adaptation can be used to
identify library members with a high affinity for the respec-
tive target.[4]
Scheme 1. Schematic representation of chemical networks based on four
different monofunctional molecules. a) In a fully connected network,
each molecule can be connected to each other molecule and to itself.
b) Directional chemical reactivity may lead to subgroups, for which intra-
group connections are not possible.
A key characteristic of a particular DCL is its network
topology, which is controlled by the chemical reactivity of
the constituent building blocks and other factors such as
steric and geometric restraints. So far, experiments have
kind have been realized with thiols, which were connected
by oxidation to give exchange-labile disulfides,[7] or with the
help of cross-metathesis reactions.[8] A second type of DCL
is based on building blocks that display a directional chemi-
cal reactivity. This includes libraries that are formed by reac-
tions between aldehydes and amines,[2c,9] or aldehydes and
hydrazides,[10] among various others.[4] For such libraries,
[a] Dr. I. Saur, Dr. R. Scopelliti, Prof. K. Severin
Institut des Sciences et IngØnierie Chimiques
École Polytechnique FØdØrale de Lausanne (EPFL)
1015 Lausanne (Switzerland)
Fax : (+41)21-693-9305
E-mail: kay.severin@epfl.ch
1058
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1058 – 1066

FULL PAPER
connections between one subgroup (e.g., aldehydes) and an-
other subgroup (e.g., hydrazides) are possible, but connec-
tions within the subgroup (e.g., coupling of two aldehydes)
do not occur (Scheme 1b). When multifunctional building
blocks with two (or more) different connecting groups are
employed (e.g., a molecule with an aldehyde and a hydra-
zide group), homoaggregations become possible, but still the
complexity of the DCL is likely to be lower that which
would be found for a symmetrical coupling chemistry.
In recent years, DCLs with more complex network archi-
tectures have been explored. One approach is to simultane-
ously utilize several types of coupling chemistry for the con-
struction of the library. For example, noncovalent interac-
tions (hydrogen bonds) have been used in combination with
dynamic covalent bonds (acyl hydrazones),[11] metal–ligand
interactions have been employed in parallel to imine ex-
change reactions,[12] and thioester exchange reactions were
combined with disulfide exchange reactions.[13] In the follow-
ing we describe an alternative approach, which is based on
the utilization of a self-sorting process[14] in combination
with a metal–ligand assembly reaction. The resulting DCL
displays a network architecture, in which two sub-libraries
are connected by two common building blocks. As a conse-
quence, a unique behavior is observed in selection experi-
ments with lithium ions as the target.
Scheme 2. Synthesis of ligand 1.
reaction. The 3-hydroxy-2-methyl-4(1H)-pyridone core of
ligand 1, on the other hand, was known to support the for-
mation of macrocyclic halfsandwich complexes.[19]
Upon reaction of ligand 1 with half an equivalent of
[{M(L)nCl2}2] (M(L)n =Ru(C6H5Me), Ru(p-cymene)Ru,
Rh(Cp), Ir(Cp*)) in either D2O or D2O/CD3OD (7:3)[20]
containing phosphate buffer (pD=8.0), a single new com-
plex was formed in over 95% yield as evidenced by
1H NMR spectroscopy (Scheme 3). The spectra displayed
Results and Discussion
In previous publications we have shown that trinuclear met-
allamacrocycles can be obtained in water at neutral pH by
reaction of organometallic half-sandwich complexes of the
general formula [{M(L)nCl2}2] (M(L)n =Ru(h6-arene),
Rh(h5-C5HxMe5Àx), Ir(h5-C5HxMe5Àx)) with aminomethylated
3-hydroxy-2(1H)-pyridone ligands.[15,16] These complexes act
as potent and selective receptors for lithium ions, and we
used them to build a colorimetric sensor that allowed us to
detect Li+ in water in the pharmacologically relevant con-
centration range of 1mm with the “naked eye”.[16] Another
interesting feature of these complexes is that they easily un-
dergo exchange reactions. Upon mixing of aqueous solutions
of a Ru macrocycle with an Ir macrocycle, for example,
mixed-metal complexes are rapidly formed. This allowed us
to build simple model DCLs in order to study some very
basic phenomena, such as the influence of the target concen-
tration on the adaptive behavior of the DCL.[17]
In continuation of this work, we were interested in obtain-
ing organometallic macrocycles that are soluble in water,
but which display a structure different from what was found
for the 3-hydroxy-2(1H)-pyridone-based complexes. For
these investigations, we synthesized the new heterocyclic
ligand 1 by aminomethylation of 3-benzyloxy-2-methyl-
4(1H)-pyridone[18] and subsequent cleavage of the benzyl
group by hydrogenolysis (Scheme 2). The dimethylamino-
methyl group of 1 was expected to enhance the solubility of
the resulting complexes in aqueous solution (partial proton-
ation at neutral pH) without interfering with the assembly
Scheme 3. Self-assembly of organometallic macrocycles. The dimethyl-
amino groups are partially protonated under those conditions (not
shown).
only one set of signals for ligand 1 and for the metal frag-
ments M(L)n, indicating highly symmetrical structures. For
the NCH2 protons, two doublets at 3.6–4.0 ppm were ob-
served for all complexes. This observation showed that the
pseudotetrahedral metals represent stereogenic centers that
are configurationally stable on the NMR timescale.
To establish the structure of the complexes, we tried to
obtain single crystals suited for a crystallographic analysis.
For the Ru(p-cymene) complex, this was possible by adding
an excess of K2HPO4. The addition of the basic phosphate
salt lead to a deprotonation of the dimethylamino groups
and consequently to a reduced solubility in water. Alterna-
tively, the complex can be precipitated by using CsOH as
the base. The molecular structure of one enantiomer of the
neutral macrocycle 2 is shown in Figure 1.
Complex 2 shows the expected[19] trigonal structure with a
(crystallographic) C3 symmetry. The metal centers are bridg-
ed by the two adjacent oxygen atoms and the nitrogen atom
of the pyridonate ligand. The planes of the heterocyclic li-
gands are nearly perpendicular (V=85.978) to the plane de-
Chem. Eur. J. 2006, 12, 1058 – 1066
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1059

K. Severin et al.
Figure 1. ORTEP[34] representation of the molecular structure of complex
2 in the crystal. The co-crystallized water solvent molecules and the hy-
drogen atoms are not shown for clarity. Selected bond lengths (Š) and
Scheme 4. a) When solutions of the macrocycles (AX)3 and (CX)3 are
combined, an equilibrium with the mixed macrocycles AX(CX)2 and
CX(AX)2 is established b) A mixture of identical composition is obtained
by self-assembly of the building blocks A, C, and X. c) A dynamic library
of the macrocycles (BX)3, (DX)3, BX(DX)2 and DX(BX)2 is obtained by
the self-assembly of the building blocks B, D, and X.
À
À
À
angles (8): Ru1 O1=2.067(3), Ru1 O2=2.080(3), Ru1 N1’=2.174(3);
O1-Ru1-O2=79.34(12),
O1-Ru1-N1’=84.81(12),
O2-Ru1-N1’=
82.13(13).
fined by the three ruthenium atoms. The Ru atoms are
7.277(1) Š apart from each other.
[{Ru(p-cymene)Cl2}2] and [{Ir(Cp*)Cl2}2] lead to the forma-
tion of the four complexes (BX)3, BX(DX)2, DX(BX)2 and
(DX)3 (Scheme 4c). For the last reaction, a mixture of D2O
and CD3OD (7:3) was employed due to the low solubility of
Ir(Cp*)-containing complexes in plain D2O.
Having established that ligand 1 can be used for the as-
sembly of trinuclear metallamacrocycles in water, scram-
bling reactions were investigated. It turned out that mixed-
metal complexes are rapidly formed upon combination of
two symmetrical macrocycles. For the discussion of these re-
sults, we will use the notation depicted below, in which the
metal fragments are denoted by A–D and the bridging li-
gands by X and Y.
1
The H NMR spectrum of the equilibrated mixture gener-
ated from (AX)3 and (CX)3 showed eight signals for the aro-
matic CH group of the ligand X in the region from d=7.1
to 7.4 ppm (Figure 2). One signal is expected for each of the
1
Figure 2. Part of the H NMR spectrum of a mixture of (AX)3 and (CX)3
after equilibration (D2O, phosphate buffer, pD 8.0). The signals corre-
spond to the aromatic CH proton of the ligand X (7.1–7.4 ppm) and to
the methyl group of the p ligand of A (1.9–2.0 ppm).
When aqueous solutions (5.0mm, D2O, phosphate buffer,
pD 8.0) of the complexes (AX)3 and (CX)3 were mixed, the
asymmetric complexes AX(CX)2 and CX(AX)2 could be ob-
symmetrical macrocycles (AX)3 and (CX)3 and three signals
for each of the mixed complexes AX(CX)2 and CX(AX)2.
In the aliphatic region of the spectrum, four signals corre-
sponding to the methyl group of the C6H5Me p-ligand were
observed. Again, this was in agreement with a scrambling
reaction since one signal was expected for the symmetric
macrocycle (AX)3, one signal for the mixed complex
AX(CX)2, and two signals for the macrocycle CX(AX)2.
The relative intensities of the respective signals were ap-
proximately equal, which suggested a nearly statistical ratio
of (AX)3:AX(CX)2:CX(AX)2:(CX)3 ~1:3:3:1.
1
served by H NMR spectroscopy within a few minutes. The
final equilibrium was reached after 2 h (Scheme 4a). A mix-
ture with an identical composition (1H NMR spectroscopy)
was obtained when the complexes were prepared in situ, by
adding buffered D2O to a mixture of ligand 1 and the two
complexes [{Ru(C6H5Me)Cl2}2] and [{Rh(Cp)Cl2}2] (Sche-
me 4b). Similarly, addition of a buffered D2O/CD3OD (7:3)
solution to a mixture of ligand 1 and the two complexes
1060
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1058 – 1066

Combinatorial Chemistry
FULL PAPER
Very similar results were obtained for a mixture generat-
ed from (BX)3 and (DX)3: eight 1H NMR signals were
found for the aromatic CH group of the ligand X and the
peak intensity was in agreement with a nearly statistical
1:3:3.1 ratio of the four different complexes.
pD 8.0 as evidenced by NMR spectroscopy and ESI mass
spectrometry. Thus, a combination of (AY )3 and (CY)3 gave
rapidly the mixed complexes AY (CY)2 and CY(AY )2 (Sche-
me 5a). As it was observed for the corresponding mixtures
1
with ligand X, four equally intense H NMR signals for the
The formation of four-component DCLs by scrambling
was furthermore confirmed by ESI mass spectrometry. Iso-
tope-resolved peaks were observed for the four different
macrocycles, with the peaks of the mixed complexes being
the dominant ones. Part of the spectrum obtained for a mix-
ture of (AX)3 and (CX)3 is shown in Figure 3.
methyl group of the p ligand of A were observed at d=1.8–
2.0 ppm in agreement with a nearly statistical distribution.
For the self-assembly of the building blocks B, D, and Y, the
1H NMR data supported a statistical distribution as well
(Scheme 5b).
Next, we investigated a more complex mixture containing
the metal building blocks A and C as well as both bridging
ligands X and Y. Analysis of the resulting dynamic combina-
torial library of macrocycles by 1H NMR spectroscopy re-
vealed a surprisingly low complexity. In fact, the spectrum
showed the species identified in mixtures of (AX)3 and
(CX)3 and of (AY )3 and (CY)3, but no other complexes. This
was confirmed by adding the equilibrated mixture depicted
in Scheme 4a to the mixture shown in Scheme 5a. No fur-
1
ther re-equilibration was observed by H NMR spectroscopy
(Figure 4). These data suggested that the assembly process
of the building blocks A, C, X, and Y was strongly self-sort-
Figure 3. Part of the ESI mass spectrum of an equilibrated mixture of the
complexes (AX)3, (CX)3, AX(CX)2, and CX(AX)2.
When macrocycles with the bridging ligand Y were em-
ployed, dynamic mixtures of homo- and heterometallic com-
plexes were also formed in aqueous buffered solution at
1
Figure 4. Part of the H NMR spectra (D2O, phosphate buffer, pD 8.0) of
a mixture of (AX)3 and (CX)3 after equilibration (bottom), a mixture of
(AY )3 and (CY)3 after equilibration (middle), and a mixture of (AX)3,
(CX)3, (AY )3, and (CY)3 after equilibration (top). The signals correspond
to the aromatic CH proton of the ligand X (d=7.1–7.4 ppm), to one of
the aromatic CH protons of the ligand Y (d=6.8–7.0 ppm) and to the
methyl group of the p-ligand of A (d=1.8–2.0 ppm).
ing: out of the 24 possible trinuclear macrocycles, only eight
were formed (Scheme 6). These eight complexes contained
either ligand X or ligand Y, but no combination of both.
The self-sorting of X- and Y-containing macrocycles is
likely the result of geometric restraints. Whereas complexes
based on ligand X have a trigonal prismatic structure, com-
plexes based on ligand Y have a concave, domelike struc-
ture (Figure 5). Apparently, there is no low-energy geometry
for hypothetical complexes containing a mixture of the li-
gands X and Y.
Scheme 5. a) When solutions of the macrocycles (AY )3 and (CY)3 are
combined, an equilibrium with the mixed macrocycles AY (CY)2 and
CY(AY )2 is established. b) Similar results are obtained for the self-assem-
bly of the building blocks B, D, and Y.
To investigate the mechanism of the scrambling reactions,
we have examined the reaction of the two symmetrical mac-
Chem. Eur. J. 2006, 12, 1058 – 1066
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1061

K. Severin et al.
1
This corresponds to what was observed by H NMR spec-
troscopy (Figure 6). Before equilibration, a single peak was
observed in the region around d=7 ppm, which can be at-
Scheme 6. Self-assembly of the building blocks A, C, X, and Y leads to
the formation of eight different macrocycles. Complexes containing both
the ligand X and the ligand Y are not observed.
Figure 6. Part of the 1H NMR spectra (D2O/CD3OD, 7:3, phosphate
buffer, pD 8.0) of a mixture of (DY)3 and (BX)3 before (top) and after
equilibration (bottom). The signals correspond to the aromatic CH pro-
tons of the ligand X.
tributed to the aromatic proton of the X ligand in (BX)3.
After equilibration, eight signals were found that correspond
to (BX)3 (one signal), (DX)3 (one signal), DX(BX)2 (three
signals), and BX(DX)2 (three signals). The remaining four
complexes containing the Y ligand showed less separated
1
signals in the H NMR spectrum, but could still be identi-
Figure 5. Ball and stick representation of the molecular structure of com-
plex 2 containing the bridging ligand X (left) and of a macrocycle com-
prised of Ru(p-cymene) fragments and a bridging ligand of type Y
(right).[35] To highlight the differences in geometry, the alkyl groups of
the aromatic p ligand, the aminomethyl groups and the hydrogen atoms
have been omitted.
fied. The formation of (DX)3 from (BX)3 was clear evidence
that both the metal–nitrogen and the metal–oxygen bonds
are readily exchanged.
From the experiments described above, it can be conclud-
ed that upon mixing of two different metal fragments with
the ligands X and Y, a DCL of eight macrocycles is formed.
Each of these complexes contains exclusively one type of
ligand; the 16 hypothetical macrocycles with mixed ligands
are not formed, since the self-assembly process is strictly
self-sorting. The eight members of the library can exchange
metal fragments, but an exchange of ligands is only possible
within the sub-library of complexes containing the same
rocycles (DY)3 and (BX)3. The weakest connection in these
macrocycles is assumed to be the metal–nitrogen bond. For
exchange reactions proceeding exclusively through a break-
age of these bonds, no scrambling would be expected for the
reaction of (DY)3 with (BX)3, given that mixed macrocycles
containing both ligand X and Y are excluded due to self-
sorting. However, if the metal–oxygen bonds are likewise
labile, eight different macrocycles should form (Scheme 7).
Scheme 8. The assembly of two different metal fragments with the li-
gands X and Y leads to the formation of a dynamic library of eight mac-
rocycles. The library can be dived into two sub-libraries, which are con-
nected by exchange of metal fragments but not of ligands.
Scheme 7. Scrambling of the complexes (DY)3 and (BX)3 leads to the for-
mation eight different macrocycles.
1062
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1058 – 1066

Combinatorial Chemistry
FULL PAPER
ligand (Scheme 8). This type of network topology leads to a
distinct behavior in selection experiments as outlined below.
Recent theoretical[21] and experimental[17,22] studies have
shown that the adaptation process of a DCL upon addition
of a target is strongly dependent on the boundary condi-
tions. Importantly, it is not necessarily the library member
with the highest affinity to the target that is amplified the
most. One of the parameters that was found to have a
strong effect on the outcome of a selection experiment is
the target concentration. This was demonstrated with a
model DCL composed of the macrocyclic receptors (DY)3,
(BY)3, BY(DY)2, and DY(BY)2 and lithium ions as the
target.[17] All four complexes were able to act as a receptor
for Li+ and the binding affinity increases in the order
(DY)3 <BY(DY)2 <DY(BY)2 <(BY)3.[23] Upon addition of
small amounts of Li+ to the four-component DCL, the dom-
inant Li+-containing complex was the best receptor (BY)3.
Upon increasing the Li+ concentration, however, the lithi-
um adduct of the second best receptor DY(BY)2 outcompet-
ed the best one.
they are coupled to the sub-library of Y complexes through
metal-exchange reactions (Scheme 8).
Increasing amounts of Li2SO4 were added to solutions of
the respective DCL (100mm phosphate buffer, pD 8.0) in
D2O/CD3OD (7:3) . The mixtures were tempered at 408C
for 40 h and the concentration of the Li+ adducts of the
high-affinity receptors (BY)3 and DY(BY)2 was approximat-
7
ed by integration of the corresponding Li NMR signals.[24]
The results are summarized in Table 1. For low Li+ concen-
Table 1. Ratio of the Li+ complexes [DY(BY)2·Li+] and [(BY)3·Li+]
upon addition of increasing amounts of Li2SO4 to DCLs comprised of the
building blocks B, D, Y or B, D, X, Y, respectively, as determined by 7 Li
NMR spectroscopy.[a]
[Li+] [mm]
4-component DCL
8-component DCL
[DY(BY)2·Li+]:[(BY)3·Li+] [DY(BY)2·Li+]:[(BY)3·Li+]
1
2
3
4
1
5
25
50
0.23
0.41
1.30
3.91
0.19
0.28
0.60
1.50
[a] The solutions (D2O/CD3OD, 7:3, 100mm phosphate buffer, pD 8.0)
were equilibrated for 40 h at 408C. The total concentration of the four
complexes containing Y ligands was in both cases 5mm.
It appeared interesting to compare the behavior of the
more complex eight-component DCL obtained from the
building blocks B, D, X, and Y (Scheme 9b) with the four-
component DCL described above (Scheme 9a). For the
trations with respect to the total macrocycle concentration,
the dominant host–guest complex was found to be in both
cases [(BY)3·Li+] with a [(BY)3·Li+]:[(DY)(BY)2·Li+] ratio
of 4:1 or 5:1, respectively (entry 1). Interesting differences,
however, were observed at higher lithium concentrations. At
[Li+]=25mm, the dominant lithium complex in the four-
component DCL was the adduct [DY(BY)2·Li+], whereas in
the eight-component DCL it was still the complex of the re-
ceptor with the highest affinity for Li+, (BY)3 (Table 1,
entry 3 and Figure 7). The pref-
erential
formation
of
[(BY)3·Li+] was also observed
for other lithium concentra-
tions.
The different result obtained
for the eight-component DCL
can be explained by the fact
that the sub-library of X com-
plexes can act as a reservoir for
the metal fragment B, which is
required for the formation of
the high affinity receptors
(BY)3 and DY(BY)2. The com-
petition between (BY)3 and
DY(BY)2 is therefore “buf-
fered” by the four additional
DCL members. A related situa-
tion is found for virtual combi-
natorial libraries, that is, DCLs
in which the monomeric build-
ing blocks are the dominating
species. Here, the competition
between the aggregates is buf-
Scheme 9. Formation of host-guest complexes upon addition of lithium
ions to a four (a) or eight-component DCL (b) comprised of the building
blocks B, D, Y or B, D, X, Y, respectively.
Figure 7. Part of the 7 Li NMR
spectrum (D2O/CD3OD, 7:3,
100 mm
pD 8.0) of DCLs comprised of
the building blocks B, D,
phosphate
buffer,
Y
former case, only the complexes containing the Y ligands
were expected to bind to lithium ions, because the geometry
of the X complexes does not permit a strong interaction
with the alkali metal ion. Nevertheless, the four complexes
with X ligands can affect the adaptation process, because
(top) or B, D, X, Y (bottom)
highlighting the signals of the
complexes [DY(BY)2·Li+] and
[(BY)3·Li+]. Concentrations:
[Y]total =15mm, [Li+]=25mm.
Chem. Eur. J. 2006, 12, 1058 – 1066
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1063

K. Severin et al.
3) (2.0 mL, 100mm phosphate buffer, pD 8.0) was added to the respective
pyridone ligand (30 mmol) and the [M(L)nCl2]2 complex (15 mmol). The
mixture was stirred for 2 h until a clear solution was obtained.
fered by the excess of monomeric building blocks. As a
result, an improved correlation between binding affinity and
amplification factor is found in selection experiments.[21a]
Complex (AX)3: 1H NMR (400 MHz, D2O): d=1.94 (s, 9H; C6H5Me),
2.41 (s, 9H; CH3), 2.79 (s, 18H; N(CH3)2), 3.60 (d, 2J=13 Hz, 3H;
2
3
NCH2), 4.08 (d, J=13 Hz, 3H; NCH2), 5.29 (d, J=6 Hz, 3H; C6H5Me),
3
3
Conclusion
5.34 (d, J=6 Hz, 3H; C6H5Me), 5.47 (t, J=5 Hz, 3H; C6H5Me), 5.78 (t,
3J=6 Hz, 3H; C6H5Me), 5.86 (t, 3J=6 Hz, 3H; C6H5Me), 7.14 ppm (s,
3H; CH, pyridone).
Over the last years, selection experiments with DCLs have
been used to discover new receptors for a given target as
well as new guest molecules for a given host. The power of
the approach is highlighted by a recent report of Sanders
et al., in which an acetylcholine target was shown to amplify
a high-affinity catenane receptor from a DCL of macrocy-
cles with hydrazone linkages.[25] It would have been very dif-
ficult to discover this receptor with a more traditional
“design approach”.[26] So far, research in the field of dynam-
ic combinatorial chemistry has focused on relatively simple
DCLs, but libraries with more complex network topologies
are increasingly being investigated.[11–13] In the present work,
we have demonstrated that self-assembled metallamacrocy-
cles with two types of pyridone ligands form dynamic libra-
ries with a unique network topology. Since the assembly
process is strictly self-sorting with respect to the bridging
ligand, only eight out of 24 possible macrocycles are formed.
The eight different complexes can be divided into two par-
tially orthogonal sub-libraries. Within these sub-libraries, an
exchange of metal fragments and ligands is possible, but
communication between the sub-libraries is restricted to an
exchange of metal fragments. This partial orthogonality is
reflected in selection experiments as demonstrated in reac-
tions with Li+ ions as targets.
Complex (BX)3: 1H NMR (400 MHz, D2O): d=1.09 (d, 3J=7 Hz, 9H;
3
CH(CH3)2), 1.14 (d, J=7 Hz, 9H; CH(CH3)2), 1.79 (s, 9H; CH3), 2.41 (s,
3
9H; CH3), 2.65 (sept, J=7 Hz, 3H; CH(CH3)2), 2.73 (s, 18H; N(CH3)2),
3.58 (d, 2J=13 Hz, 3H; NCH2), 4.01 (d, 2J=13 Hz, 3H; NCH2), 5.15 (d,
3J=5 Hz, 3H; MeC6H4iPr), 5.38 (d, 3J=5 Hz, 3H; MeC6H4iPr), 5.73 (d,
3
3J=5 Hz, 3H; MeC6H4iPr), 5.76 (d, J=5 Hz, 3H; MeC6H4iPr), 7.06 ppm
(s, 3H; CH, pyridone).
Complex (CX)3: 1H NMR (400 MHz, D2O): d=2.43 (s, 9H; CH3), 2.76
(s, 18H; N(CH3)2), 3.68 (d, 2J=13 Hz, 3H; NCH2), 4.08 (d, 2J=13 Hz,
3H; NCH2), 5.72 (s, 15H; CH, Cp), 7.33 ppm (s, 3H; CH, pyridone).
Complex (DX)3: 1H NMR (400 MHz, D2O/CD3OD: 7/3): d=1.49 (s,
2
45H; CH3, Cp*), 2.38 (s, 9H; CH3), 2.66 (s, 18H; N(CH3)2), 3.55 (d, J=
2
13 Hz, 3H; NCH2), 4.13 (d, J=13 Hz, 3H; NCH2), 6.98 ppm (s, 3H; CH,
pyridone).
Complex (AY)3: 1H NMR (400 MHz, D2O): d=2.00 (s, 9H; C6H5Me),
2.66 (s, 18H; N(CH3)2), 3.97 (d, 2J=13 Hz, 3H; NCH2), 4.03 (d, 2J=
13 Hz, 3H; NCH2), 5.24 (d, 3J=6 Hz, 3H; C6H5Me), 5.52 (t, 3J=6 Hz,
3H; C6H5Me), 5.56 (d, 3J=5 Hz, 3H; C6H5Me), 5.84 (d, 3J=7 Hz, 3H;
CH, pyridone), 5.88 (t, 3J=6 Hz, 3H; C6H5Me), 6.10 (t, 3J=6 Hz, 3H;
3
C6H5Me), 6.80 ppm (d, J=7 Hz, 3H; CH, pyridone).
Complex (BY)3: 1H NMR (400 MHz, D2O): d=1.25 (d, 3J=7 Hz, 9H;
3
CH(CH3)2), 1.28 (d, J=7 Hz, 9H; CH(CH3)2), 1.82 (s, 9H; CH3), 2.66 (s,
3
2
18H; N(CH3)2), 2.82 (sept, J=7 Hz, 3H; CH(CH3)2), 3.86 (d, J=13 Hz,
3H; NCH2), 4.07 (d, 2J=13 Hz, 3H; NCH2), 5.25 (d, 3J=5 Hz, 3H; Me-
3
3
C6H4iPr), 5.50 (d, J=5 Hz, 3H; MeC6H4iPr), 5.81 (d, J=6 Hz, 3H; CH,
pyridone), 5.81 (d, 3J=6 Hz, 3H; MeC6H4iPr), 6.04 (d, 3J=5 Hz, 3H;
MeC6H4iPr), 6.71 ppm (d, J=7 Hz, 3H; CH, pyridone).
3
Complex (CY)3: 1H NMR (400 MHz, D2O): d=2.69 (s, 18H; N(CH3)2),
4.02 (d, 2J=13 Hz, 3H; NCH2), 4.08 (d, 2J=13 Hz, 3H; NCH2), 5.78 (s,
15H; CH, Cp), 5.99 (d, 3J=7 Hz, 3H; CH, pyridone), 6.98 ppm (d, 3J=
7 Hz, 3H; CH, pyridone).
Experimental Section
Complex (DY)3: 1H NMR (400 MHz, D2O): d=1.65 (s, 45H; CH3, Cp*),
2.71 (s, 18H; N(CH3)2), 4.02 (d, 2J=13 Hz, 3H; NCH2), 4.11 (d, 2J=
13 Hz, 3H; NCH2), 5.80 (d, 3J=7 Hz, 3H; CH, pyridone), 6.79 ppm (d,
3J=7 Hz, 3H; CH, pyridone).
General: The complexes [{Ru(C6H5Me)Cl2}2],[27] [{Ru(p-cymene)Cl2}2],[28]
[{Rh(Cp)Cl2}2],[29] and [{Ir(Cp*)Cl2}2][30] and the ligand 4-dimethylamino-
methyl-3-hydroxy-2-(1H)-pyridone[31] were prepared according to litera-
ture procedures. The synthesis of all complexes was performed under an
Synthesis of complex 2:
A suspension of the [{Ru(p-cymene)Cl2}2]
1
atmosphere of dry dinitrogen, using standard Schlenk techniques. The H
(17.8 mg, 29.1 mmol) and ligand 1 (10.6 mg, 58.2 mmol) in water (3.5 mL)
was stirred for 2 h until a red solution was obtained. CsOH (116 mmol)
was added and complex 2 precipitated as an orange powder which was
isolated by centrifugation and dried under vacuum (yield: 21.3 mg, 87%).
Single crystals of complex 2 were obtained from a concentrated aqueous
solution of (BX)3 after addition of an excess of K2HPO4. 1H NMR
(400 MHz, CDCl3): d=1.26 (d, 3J=7 Hz, 9H; CH(CH3)2), 1.32 (d, 3J=
7 Hz, 9H; CH(CH3)2), 2.09 (s, 9H; CH3), 2.13 (s, 18H; N(CH3)2), 2.38 (s,
and 13C spectra were recorded on a Bruker Avance DPX400 spectrome-
ter by using the residual protonated solvents (1H, 13C) as internal stan-
dards or LiCl in water as an external standard. All spectra were recorded
at room temperature. The ESI MS studies were performed with solutions
of the respective macrocycles (total concentration of macrocycles:
2.5mm) in water/acetonitrile (3:1) containing a 50mm phosphate buffer
(pH~8).
5-Dimethylaminomethyl-3-hydroxy-2-methyl-4-(1H)-pyridone (1): Palla-
dium on charcoal (43 mg) was added to a solution of 3-benzyloxy-5-dime-
thylaminomethyl-2-methyl-4-(1H)-pyridone (504 mg, 1.85 mmol) in etha-
nol (45 mL). The mixture was stirred for 20 h under an atmosphere of di-
hydrogen. The catalyst was then eliminated by filtration. Evaporation of
the ethanol under vacuum gave ligand 1 as a white solid (yield: 223 mg,
63%). 1H NMR (400 MHz, D2O): d=2.34 (s, 3H; CH3), 2.62 (s, 6H; N-
(CH3)2), 3.90 (s, 2H; NCH2), 7.54 ppm (s, 1H; CH); 13C NMR (400 MHz,
D2O): d=16.71 (CH3), 45.52 (N(CH3)2), 58.37 (NCH2), 118.7, 134.7,
136.3, 152.1, 174.1 (pyridone); elemental analysis calcd (%) for
C9H14N2O2·0.5H2O: C 56.53, H 7.91, N 14.65; found: C 56.28, H 7.84, N
15.18.
2
3
9H; CH3), 2.64 (d, J=13 Hz, 3H; NCH2), 2.73 (sept, J=7 Hz, 3H; CH-
(CH3)2), 3.41 (d, 2J=13 Hz, 3H; NCH2), 4.96 (d, 3J=5 Hz, 3H; Me-
3
3
C6H4iPr), 4.97 (d, J=5 Hz, 3H; MeC6H4iPr), 5.09 (d, J=5 Hz, 3H; Me-
3
C6H4iPr), 5.30 (d, J=5 Hz, 3H; MeC6H4iPr), 6.79 ppm (s, 3H; CH, pyri-
done); 13C NMR (101 MHz, CDCl3): d=17.60, 18.02, 22.60, 23.15 (CH3),
31.42 (CH(CH3)2), 44.98 (N(CH3)2), 55.08 (NCH2), 79.42, 79.57, 80.14,
80.63 (CH, MeC6H4iPr), 94.10, 99.17 (C, MeC6H4iPr), 115.77, 140.81,
141.81, 157.13, 167.62 ppm (pyridone); elemental analysis calcd (%) for
C57H78Ru3N6O6·H2O: C 54.14, H 6.38, N 6.65; found: C 53.86, H 6.38, N
6.78.
Preparation of a four-component DCL
Method A: A D2O solution (2.0 mL, 100mm phosphate buffer, pD 8.0)
General procedure for the synthesis of the complexes (AX)3, (BX)3,
(CX)3, (DX)3, (AY)3, (BY)3, (CY)3, and (DY)3: D2O or D2O/CD3OD (7/
was added to a mixture of ligand L (L=X, Y, 30 mmol), [{Ru-
1064
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1058 – 1066

Combinatorial Chemistry
FULL PAPER
(C6H5Me)Cl2}2] (3.9 mg, 7.5 mmol), and [{Rh(Cp)Cl2}2] (3.6 mg, 7.5 mmol).
The mixture was equilibrated for 2 h at RT.
(DY)3 ([Y]total =15mm, D2O/CD3OD (7:3), 100 mm phosphate buffer,
pD 8.0), the reaction mixture was tempered at 408C for 40 h.
Reaction of an eight-component DCL with Li+: After addition of various
amounts of a Li2SO4 stock solution (D2O, 50mm, 0.5m or 2.0m) to the
equilibrated mixture of the macrocycles (BY)3, DY(BY)2, BY(DY)2,
Method B: A solution of the macrocycle (AL)3 (5.0mm, 100mm phos-
phate buffer, pD 8.0) in D2O (1.0 mL) was mixed with a solution of the
macrocycle (CL)3 (5.0mm, 100mm phosphate buffer, pD 8.0) in D2O
(1.0 mL) and equilibrated for 2 h at RT (L=X, Y).
(DY)3, (BX)3, DX(BX)2, BX(DX)2, and (DX)3 ([X]total =15 mm, [Y]total
=
15mm, D2O/CD3OD (7:3), 100mm phosphate buffer, pD 8.0), the reac-
tion mixture was tempered at 408C for 40 h.
DCLs containing the metal fragment B and D were prepared following
the method A with D2O/CD3OD (7:3) instead of D2O as the solvent. The
composition of the library was analyzed using 1H NMR spectroscopy. In
order to verify that the final equilibrium was reached, NMR spectra were
recorded at later times.
X-ray crystallography: Details about the crystal and the structure refine-
ment of 2 are listed in Table 2, whereas some relevant geometrical pa-
rameters are included into the picture captions. Data collection was per-
formed at 140(2) K on a marresearch mar345 IPDS diffractometer. Data
reduction was carried out with CrysAlis RED, release 1.7.0.[32] Absorp-
tion correction was applied. Structure solution and refinement were per-
formed with the SHELXTL software package, release 5.1.[33] The struc-
ture was refined by using the full-matrix least-squares on F2 with all non-
H atoms anisotropically defined. H atoms were placed in calculated posi-
tions using the “riding model” (except those belonging to the water mol-
ecules that were not included into the final model). The asymmetric unit
contains 19/3 water molecules of which four (O3, O4, O7, O8) do not
show any sign of disorder, two look disordered (O5 and O6) and have
been treated with the split model and then their occupancy factor fixed,
the remaining 1/3H2O (O9) lies on a threefold symmetry axis. CCDC
271907 contains the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge Crystal-
lographic Data Centre via www.ccdc.cam.ac.uk/data request/cif.
Preparation of an eight-component DCL containing A and C: Method
A: A D2O solution (2.0 mL, 100 mm phosphate buffer, pD 8.0) was added
to a mixture of ligand X (5.5 mg, 30 mmol), ligand Y (5.1 mg, 30 mmol),
[{Ru(C6H5Me)Cl2}2] (7.9 mg, 15 mmol) and [{Rh(Cp)Cl2}2] (7.2 mg,
15 mmol). The mixture was equilibrated for 2 h at RT.
Method B: A solution of an equilibrated mixture of the macrocycles
(AX)3, AX(CX)2, CX(AX)2, and (CX)3 ([X]total =30mm, 100mm phos-
phate buffer, pD 8.0) in D2O (1.0 mL) was mixed with a solution of an
equilibrated mixture of the macrocycles (AY )3, AY (CY)2, CY(AY )2, and
(CY)3 ([Y]total =30mm, 100mm phosphate buffer, pD 8.0) in D2O
(1.0 mL) and stirred for 2 h at RT.
Preparation of an eight-component DCL containing B and D: A solution
of the macrocycle (BX)3 (10mm, 100mm phosphate buffer, pD 8.0) in
D2O/CD3OD (7:3; 1.0 mL) was mixed with a solution of the macrocycle
(DY)3 (10mm, 100mm phosphate buffer, pD 8.0) in D2O/CD3OD (7:3;
1.0 mL) and equilibrated for 2 h at RT. The composition of the library
was analyzed using 1H NMR spectroscopy. In order to verify that the
final equilibrium was reached, NMR spectra were recorded at later
times.
Acknowledgements
Reaction of a four-component DCL with Li+: After addition of various
amounts of a Li2SO4 stock solution (D2O, 50mm, 0.5m or 2.0m) to the
equilibrated mixture of the macrocycles (BY)3, DY(BY)2, BY(DY)2, and
The work was supported by the Swiss National Science Foundation and
by the EPFL.
[1] a) P. A. Brady, R. P. Bonar-Law, S. J. Rowan, C. J. Suckling, J. K. M.
Sanders, Chem. Commun. 1996, 319–320; b) S. J. Rowan, P. A.
Brady, J. K. M. Sanders, Angew. Chem. 1996, 108, 2283–2285;
Angew. Chem. Int. Ed. Engl. 1996, 35, 2143–2145; c) S. J. Rowan,
D. G. Hamilton, P. A. Brady, J. K. M. Sanders, J. Am. Chem. Soc.
1997, 119, 2578–2579; d) S. J. Rowan, J. K. M. Sanders, Chem.
Commun. 1997, 1407–1408.
[2] a) B. Hasenknopf, J.-M. Lehn, B. O. Kneisel, G. Baum, D. Fenske,
Angew. Chem. 1996, 108, 1987–1990; Angew. Chem. Int. Ed. Engl.
1996, 35, 1838–1840; b) B. Hasenknopf, J.-M. Lehn, N. Boumediene,
A. Dupont-Gervais, A. Van Dorsselaer, B. Kneisel, D. Fenske, J.
Am. Chem. Soc. 1997, 119, 10956–10962; c) I. Huc, J.-M. Lehn,
Proc. Natl. Acad. Sci. USA 1997, 94, 2106–2110.
Table 2. Crystallographic data for complex 2.
2”19H2O
formula
C57H116N6O25Ru3
1588.77
Mr [gmolÀ1
crystal size
]
0.33”0.20”0.20
rhombohedral
crystal system
space group
a [Š]
¯
R3
19.673(4)
b [Š]
19.673(4)
c [Š]
33.224(8)
a [8]
90
b [8]
90
[3] a) T. Hayashi, T. Asai, H. Hokazono, H. Ogoshi, J. Am. Chem. Soc.
1993, 115, 12210–12211; b) B. A. Katz, J. Finer-Moore, R. Morte-
zaei, D. H. Rich, R. M. Stroud, Biochemistry 1995, 34, 8264–8280;
c) P. G. Swann, R. A. Casanova, A. Desai, M. M. Frauenhoff, M. Ur-
bancic, U. Slomcznska, A. J. Hopfinger, G. C. LeBreton, D. L.
Venton, Biopolymers 1996, 40, 617–625; d) S. Sakai, Y. Shigemase,
T. Sasaki, Tetrahedron Lett. 1997, 38, 8145–8148; e) A. V. Eliseev,
M. I. Nelen, J. Am. Chem. Soc. 1997, 119, 1147–1148; f) B. Klekota,
M. H. Hammond, B. L. Miller, Tetrahedron Lett. 1997, 38, 8639–
8642; g) A. V. Eliseev, M. I. Nelen, J. Am. Chem. Soc. 1997, 119,
1147–1148.
g [8]
120
11135(4)
6
1.422
140(2)
0.678
3.22 to 25.03
À22ꢀhꢀ23
À23ꢀkꢀ23
À39ꢀlꢀ39
23827
V [Š3]
Z
1calcd [gcmÀ3
T [K]
]
m [mmÀ1
]
q range [8]
index ranges
reflections collected
[4] For reviews see: a) J. D. Cheeseman, A. D. Corbett, J. L. Gleason,
R. J. Kazlauskas, Chem. Eur. J. 2005, 11, 1709–1716; b) J.-L. Rey-
mond, Angew. Chem. 2004, 116, 5695–5697; Angew. Chem. Int. Ed.
2004, 43, 5577–5579; c) O. Ramstrçm, T. Bunyapaiboonsri, S. Loh-
mann, J.-M. Lehn, Biochim. Biophys. Acta 2002, 1572, 178–186;
d) S. J. Rowan, Cantrill, G. R. L. Cousins, J. K. M. Sanders, J. F.
Stoddart, Angew. Chem. 2002, 114, 938–993; Angew. Chem. Int. Ed.
2002, 41, 898–952; e) S. Otto, R. L. E. Furlan, J. K. M. Sanders,
Curr. Opin. Chem. Biol. 2002, 6, 321–327; f) S. Otto, R. L. E.
independent reflections
absorption correction
max/min transmission
data/restraints/parameters
goodness-of-fit on F2
final R indices [I>2s(I)]
R indices (all data)
4381 (Rint =0.0691)
semi-empirical
0.9209 and 0.8879
4381/0/293
1.024
R1 =0.0479, wR2 =0.1267
R1 =0.0671, wR2 =0.1395
1.221/À1.270
À3
largest diff. peak/hole [eŠ
]
Chem. Eur. J. 2006, 12, 1058 – 1066
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1065

K. Severin et al.
Furlan, J. K. M. Sanders, Drug Discovery Today 2002, 7, 117–125;
g) C. Karan, B. L. Miller, Drug Discovery Today 2000, 5, 67–75;
h) J.-M. Lehn, Chem. Eur. J. 1999, 5, 2455–2463; i) A. Ganesan,
Angew. Chem. 1998, 110, 2989–2992; Angew. Chem. Int. Ed. 1998,
37, 2828–2831.
[12] a) N. Giuseppone, J.-L. Schmitt, J.-M. Lehn, Angew. Chem. 2004,
116, 5010–5014; Angew. Chem. Int. Ed. 2004, 43, 4902–4906;
b) J. R. Nitschke, Angew. Chem. 2004, 116, 3135–3137; Angew.
Chem. Int. Ed. 2004, 43, 3073–3075; c) J. R. Nitschke, D. Schultz, G.
Bernardinelli, D. GØrard, J. Am. Chem. Soc. 2004, 126, 16538–
16543; d) J. R. Nitschke, J.-M. Lehn, Proc. Natl. Acad. Sci. USA
2003, 100, 11970–11974; e) V. Goral, M. I. Nelen, A. V. Eliseev, J.-
M. Lehn, Proc. Natl. Acad. Sci. USA 2001, 98, 1347–1352; f) D. M.
Epstein, S. Choudhary, M. R. Churchill, K. M. Keil, A. V. Eliseev,
J. R. Morrow, Inorg. Chem. 2001, 40, 1591–1596; g) B. Klekota,
B. L. Miller, Tetrahedron 1999, 55, 11687–11697.
[13] J. Leclaire, L. Vial, S. Otto, J. K. M. Sanders, Chem. Commun. 2005,
1959–1961.
[14] a) P. Mukhopadhyay, A. Wu, L. Isaacs, J. Org. Chem. 2004, 69,
6157–6164; b) A. Wu, L. Isaacs, J. Am. Chem. Soc. 2003, 125, 4831–
4835, and references therein.
[15] Z. Grote, M.-L. Lehaire, R. Scopelliti, K. Severin, J. Am. Chem.
Soc. 2003, 125, 13638–13639.
[16] Z. Grote, R. Scopelliti, K. Severin, J. Am. Chem. Soc. 2004, 126,
16959–16972.
[17] I. Saur, K. Severin, Chem. Commun. 2005, 1471–1473.
[18] M. K. Patel, R. Fox, P. D. Taylor, Tetrahedron 1996, 52, 1835–1840.
[19] T. Habereder, M. Warchhold, H. Nçth, K. Severin, Angew. Chem.
1999, 111 3422–3425; Angew. Chem. Int. Ed. 1999, 38, 3225–3228.
[20] The Ir(Cp*) macrocycle showed a limited solubility in pure water.
Addition of 30% CD3OD lead to an improved solubility and experi-
ments with this complex were thus performed in this solvent mix-
ture.
[5] The notion “assembly” includes reversible covalent bond formation
under thermodynamic control.
[6] For reviews about complex dynamic networks see: a) S. H. Strogatz,
Nature 2001, 410, 268–276; b) R. Albert, A.-L. Barbarµsi, Rev. Mod.
Phys. 2002, 74, 47–97.
[7] a) A. L. Kieran, S. I. Pascu, T. Jarrosson, J. K. M. Sanders, Chem.
Commun. 2005, 1276–1278; b) S. Sando, A. Narita, Y. Aoyama,
Bioorg. Med. Chem. Lett. 2004, 14, 2835–2838; c) A. M. Whitney, S.
Ladame, S. Balasubramanian, Angew. Chem. 2004, 116, 1163–1166;
Angew. Chem. Int. Ed. 2004, 43, 1143–1146; d) A. L. Kieran, A. D.
Bond, A. M. Belenguer, J. K. M. Sanders, Chem. Commun. 2003,
2674–2675; e) S. Otto, S. Kubik, J. Am. Chem. Soc. 2003, 125, 7804–
7805; f) B. Brisig, J. K. M. Sanders, S. Otto, Angew. Chem. 2003, 115,
1308–1311; Angew. Chem. Int. Ed. 2003, 42, 1270–1273; g) Y.
Krishnan-Ghosh, S. Balasubramanian, Angew. Chem. 2003, 115,
2221–2227; Angew. Chem. Int. Ed. 2003, 42, 2171–2173; h) S. Otto,
R. L. E. Furlan, J. K. M. Sanders, Science 2002, 297, 590–593; i) S.
Otto, R. L. E. Furlan, J. K. M. Sanders, J. Am. Chem. Soc. 2000, 122,
12063–12064; j) O. Ramstrçm, J.-M. Lehn, ChemBioChem 2000, 1,
41–48.
[8] K. C. Nicolaou, R. Hughes, S. Y. Cho, N. Winssinger, C. Smethurst,
H. Labischinski, R. Endermann, Angew. Chem. 2000, 112, 3981–
3986; Angew. Chem. Int. Ed. 2000, 39, 3823–3828.
[9] a) S. Zameo, B. Vauzeilles, J.-M. Beau, Angew. Chem. 2005, 117,
987–991; Angew. Chem. Int. Ed. 2005, 44, 965–969; b) A. Bugaut,
J.-J. ToulmØ, B. Rayner, Angew. Chem. 2004, 116, 3206–3209;
Angew. Chem. Int. Ed. 2004, 43, 3144–3147; c) N. Guiseppone, J.-M.
Lehn, J. Am. Chem. Soc. 2004, 126, 11448–11449; d) A. Gonzµlez-
ꢁlvarez, I. Alfonso, F. Lꢂpez-Ortiz, ꢁ. Aguirre, S. Garcꢃa-Granda,
V. Gotor, Eur. J. Org. Chem. 2004, 1117–1127; e) H. Li, P. Williams,
J. Micklefield, J. M. Gardiner, G. Stephens, Tetrahedron 2004, 60,
753–758; f) K. Oh, K.-S. Jeong, J. S. Moore, J. Org. Chem. 2003, 68,
8397–8403; g) M. Hochgꢄrtel, H. Kroth, D. Piecha, M. W. Hofmann,
K. C. Nicolau, S. Krause, O. Schaaf, G. Sonnenmoser, A. V. Eliseev,
Proc. Natl. Acad. Sci. USA 2002, 99, 3382–3387; h) S. Gerber-Lem-
aire, F. Popowycz, E. Rodrꢃguez-Garcꢃa, A. T. Carmona Asenjo, I.
Robina, P. Vogel, ChemBioChem 2002, 3, 466–470; i) O. Storm, U.
Lꢄnig, Chem. Eur. J. 2002, 8, 793–798.
[21] a) K. Severin, Chem. Eur. J. 2004, 10, 2565–2580; b) P. T. Corbett, S.
Otto, J. K. M. Sanders, Chem. Eur. J. 2004, 10, 3139–3143.
[22] a) Z. Grote, R. Scopelliti, K. Severin, Angew. Chem. 2003, 115,
3951–3955; Angew. Chem. Int. Ed. 2003, 42, 3821–3825; b) P. T.
Corbett, J. K. M. Sanders, S. Otto, J. Am. Chem. Soc. 2005, 127,
9390–9392.
[23] For an explanation of this trend see reference [16].
[24] The concentrations of the Li+ adducts of the low affinity receptors
(DY)3 and BY(DY)2 can be neglected under those conditions.
[25] R. T. S. Lam, A. Belenguer, S. L. Roberts, C. Naumann, T. Jarrosson,
S. Otto, J. K. M. Sanders, Science 2005, 308, 667–669.
[26] For a critical discussion of the limitations to design high-affinity re-
ceptors for a given target see: J. K. M. Sanders, Chem. Eur. J. 1998,
4, 1378–1383.
[27] M. A. Bennett, A. K. Smith, J. Chem. Soc. Dalton Trans. 1974, 2,
234–241.
[10] a) T. Ono, T. Nobori, J.-M. Lehn, Chem. Commun. 2005, 1522–
1524; b) W. G. Skene, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 2004,
101, 8270–8275; c) O. Ramstrçm, S. Lohmann, T. Bunyapaiboonsri,
J.-M. Lehn, Chem. Eur. J. 2004, 10, 1711–1715; d) S. L. Roberts,
R. L. E. Furlan, S. Otto, J. K. M. Sanders, Org. Biomol. Chem. 2003,
1, 1625–1633; e) S. L. Roberts, R. L. E. Furlan, G. R. L. Cousins,
J. K. M. Sanders, Chem. Commun. 2002, 938–939; f) R. L. E. Furlan,
Y.-F. Ng, G. R. L. Cousins, J. E. Redman, J. K. M. Sanders, Tetrahe-
dron 2002, 58, 771–778; g) R. L. E. Furlan, Y.-F. Ng, S. Otto,
J. K. M. Sanders, J. Am. Chem. Soc. 2001, 123, 8876–8877; h) T. Bu-
nyapaiboonsri, O. Ramstrçm, S. Lohmann, J.-M. Lehn, L. Peng, M.
Goeldner, ChemBioChem 2001, 2, 438–444; i) G. R. L. Cousins,
R. L. E. Furlan, Y.-F. Ng, J. E. Redman, J. K. M. Sanders, Angew.
Chem. 2001, 113, 437–443; Angew. Chem. Int. Ed. 2001, 40, 423–
428; j) M. G. Simpson, S. P. Watson, N. Feeder, J. E. Davis, J. K. M.
Sanders, Org. Lett. 2000, 2, 1435–1438; k) R. L. E. Furlan, G. R. L.
Cousins, J. K. M. Sanders, Chem. Commun. 2000, 1761–1762;
l) G. R. L. Cousins, S.-A. Poulsen, J. K. M. Sanders, Chem. Commun.
1999, 1575–1576.
[28] M. A. Bennett, T. -N. Huang, T. W. Matheson, A. K. Smith, Inorg.
Synth. 1982, 21, 74–78.
[29] J. W. Kang, K. Moseley, P. M. Maitlis, J. Am. Chem. Soc. 1969, 91,
5970–5977.
[30] C. White, A. Yates, P. M. Maitlis, Inorg. Synth. 1992, 29, 228–234.
[31] a) A. Nakamura, K. Shozo, Chem. Pharm. Bull. 1968, 16, 1466–
1471; b) A. G. Osborne, L. Jackson, Spectrochim. Acta Part A 1993,
49, 1703–1708.
[32] Oxford Diffraction Ltd., Abingdon, Oxfordshire, UK, 2003.
[33] G. M. Sheldrick, University of Gçttingen, Germany, 1997; Bruker
AXS, Inc., Madison, Wisconsin, 53719, USA, 1997.
[34] ORTEP 3 for Windows version 1.076: L. J. Farrugia, J. Appl. Crys-
tallogr. 1997, 30, 565.
[35] The graphic has been prepared using the crystallographic data of
complex [{Ru(cymene)(C8H12N2O3)}3]. For a more detailed descrip-
tion of this structure see reference [16].
[11] a) N. Sreenivasachary, J.-M. Lehn, Proc. Natl. Acad. Sci. USA 2005,
102, 5938–5943; b) E. Kolomiets, J.-M. Lehn, Chem. Commun.
2005, 1519–1521.
Received: June 1, 2005
Published online: August 15, 2005
1066
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1058 – 1066