Structural parameters governing the dynamic combinatorial synthesis of catenanes in water

Article abstract of DOI:10.1021/ja3075727

We report the first dynamic combinatorial synthesis in water of an all-acceptor [2]catenane and of different types of donor-acceptor [2] and [3]catenanes. Linking two electron-deficient motifs within one building block using a series of homologous alkyl chains provides efficient and selective access to a variety of catenanes and offers an unprecedented opportunity to explore the parameters that govern their synthesis in water. In this series, catenane assembly is controlled by a fine balance between kinetics and thermodynamics and subtle variations in the building block structure, such as the linker length and building block chirality. A remarkable and unexpected odd-even effect with respect to the number of atoms in the alkyl linker is reported.

Full text of DOI:10.1021/ja3075727

Article
pubs.acs.org/JACS
Structural Parameters Governing the Dynamic Combinatorial
Synthesis of Catenanes in Water
,†,‡
Fabien B. L. Cougnon,†,§ Nandhini Ponnuswamy,†,§ Nicholas A. Jenkins,† G. Dan Pantos,
̧*
and Jeremy K. M. Sanders*,†
†University Chemical Laboratory, University of Cambridge, Lensfield Road, CB2 1EW, Cambridge, United Kingdom
‡Department of Chemistry, University of Bath, BA 7AY, Bath, United Kingdom
S
* Supporting Information
ABSTRACT: We report the first dynamic combinatorial
synthesis in water of an all-acceptor [2]catenane and of
different types of donor−acceptor [2] and [3]catenanes.
Linking two electron-deficient motifs within one building
block using a series of homologous alkyl chains provides
efficient and selective access to a variety of catenanes and offers
an unprecedented opportunity to explore the parameters that
govern their synthesis in water. In this series, catenane
assembly is controlled by a fine balance between kinetics and
thermodynamics and subtle variations in the building block structure, such as the linker length and building block chirality. A
remarkable and unexpected odd−even effect with respect to the number of atoms in the alkyl linker is reported.
rendering the formation of the all-acceptor (AAAA) and all-
donor (DDDD) catenanes elusive. Building on these results, we
showed in an earlier communication5 that the entropic penalty
associated with the formation of larger structures such as
[3]catenanes could be overcome by linking two electron-
deficient units within one building block, using a butyl linker.
In the current study, we generalize our strategy by using a
series of building blocks composed of two electron-deficient
units linked with homologous alkyl chains ((−CH2−)n where n
= 2−9). First, using libraries composed of only electron-
deficient building blocks, we demonstrate that the hydrophobic
effect alone can be used to direct the synthesis of all-acceptor
[2]catenanes. The formation of these all-acceptor [2]catenanes
is driven by the intercalation of the hydrophobic alkyl chain
between the electron-deficient units and therefore does not
contradict our earlier study which showed that, under our
dynamic combinatorial conditions (water pH 8.0, 1 M
NaNO3), the formation of all-acceptor catenanes with four
electron-deficient aromatic units stacking in a conventional
fashion was improbable.
In the second part of this work, we show that addition of an
electron-rich building block to the above-mentioned electron-
deficient building blocks shifts the library distribution toward
the formation of the smallest accessible donor−acceptor
catenane, which may be either a [2] or a [3]catenane
depending on the length of the aliphatic linker. These results
reveal the operation of an odd−even effect with respect to the
number of CH2 units, n, of the linking alkyl chain. This trend is
clearly noticeable for short alkyl linkers and vanishes when the
INTRODUCTION
■
We report here the synthesis of a family of [2] and
[3]catenanes, assembled from homologous linear building
blocks in water. Until now, the low solubility of organic
molecules in water and the ability of aqueous media to suppress
weak supramolecular interactions have challenged the synthesis
of interlocked molecules in water. Catenanes, for example, have
mainly been assembled and studied in organic solvents,1 an
approach that has restricted their utility in biological contexts.
Following the synthesis of the first water-soluble interlocked
molecules,2 notably pioneered by Fujita et al.,3 we4,5 and
others6 have recently begun to develop strategies to access
catenanes in water.
Our strategy, based on a dynamic combinatorial approach,7,8
presents many attractions, especially to explore the potential
synthesis of unprecedented structures. We use simple and
readily available dithiol building blocks containing either an
electron-donor (D) or an electron-deficient (A) aromatic unit
functionalized with amino acid residues. The presence of polar
amino acid residues not only renders the catenanes water
soluble but also permits the introduction of new functionalities,
such as carboxylates or amides, which can be used in the
context of molecular recognition5 and could also be utilized for
catalysis or post-synthetic modifications.
More importantly, we have shown that the hydrophobic
effect could be exploited to control the arrangement of the
stacked donor and acceptor aromatic groups,4 allowing us to
synthesize [2]catenanes with the conventional alternating
DADA stacking sequence as well as with the previously
unknown DADD, AADA, and DAAD motifs. The assembly of
these catenanes is not driven purely by the hydrophobic effect:
donor−acceptor interactions still play an important role,
Received: August 1, 2012
Published: November 1, 2012
© 2012 American Chemical Society
19129
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
Figure 1. Different types of [2] and [3]catenanes formed in dynamic combinatorial libraries from electron-rich 1 and electron-deficient 2−9 building
blocks. The cartoon yellow dots represent the sulfur atoms of the building blocks.
chain length increases. Odd−even effects are often observed in
the solid state9 but appear to be less common in solution
state.10 We, therefore, discuss the possible origins of this effect,
which seems to be associated with kinetically controlled
processes and tends to disappear when the library progresses
toward thermodynamic equilibrium. Finally, we also demon-
strate that the catenane synthesis can be influenced by
differences in the building block chirality.
building blocks 2−9 are all based on the same design: two
large hydrophobic electron-deficient π-systems (1,4,5,8-naph-
thalenetetracarboxylic diimide) are connected via a flexible
aliphatic chain and terminated by L-cysteine hydrophilic side
chains.11 Eight consecutive diamines, from 1,2-diaminoethane
to 1,9-diaminononane, were used to produce flexible aliphatic
linkers of systematically increasing length. The notation 2−9
refers both to the acceptor building block and to the
corresponding number n of CH2 of the alkyl chain linking
the two aromatic moieties. Some of the results related to
building block 4 have been previously published in preliminary
form;5 they are presented here as a part of a more general
picture.
Exploiting the Hydrophobic Effect to Assemble All-
Acceptor Catenanes. The first set of libraries was prepared
by dissolving each acceptor 2−9 at a concentration of 5 mM in
water at pH 8 in the presence of 1 M NaNO3 and was analyzed
by reverse-phase HPLC. Absorbance was recorded at 383 nm,
the optimum wavelength for the naphthalenediimide, and the
library members were unambiguously identified by tandem ESI-
MS and MS/MS.
Overall, these catenanes constitute a model from which
synthetic rules can be extracted. Our results provide an insight
into the kinetics and thermodynamics governing the library
distribution and highlight the role of various parameters
directing catenane synthesis, particularly hydrophobicity,
flexibility, and chirality. Representative samples of each type
of catenanes were isolated by preparative HPLC and fully
1
characterized as individual compounds by H NMR and UV−
vis in addition to the LCMS analyses. This work, which is
summarized in Figure 1, therefore constitutes a benchmark for
the easy and versatile synthesis of different types of water-
soluble [2] and [3]catenanes.
RESULTS AND DISCUSSION
Libraries prepared from the short-linker building blocks (n <
7) are dominated by the cyclic dimer, while the proportion of
cyclic monomer increases with longer linkers (7 ≤ n ≤ 9, see
SI). For n = 8, we observed trace amount of a new species that
corresponded to an all-acceptor [2]catenane composed of two
■
The synthesis of catenanes was achieved in water in one pot
from simple linear building blocks: an electron-rich dialkox-
ynaphthalene building block (1) and a series of homologous
acceptor building blocks (2−9, Figure 1). The acceptor
19130
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
Figure 2. (a) Reverse-phase HPLC trace of a library prepared by dissolving 9 in water 1 M NaNO3, pH 8, and structure of the all-acceptor
1
[2]catenane. (b) H NMR (500 MHz, 298 K) spectrum of the all-acceptor [2]catenane in D2O.
interlocked monomeric rings. The yield of this all-acceptor
catenane is significantly higher for the linker length n = 9 (60%
of the library composition, Figure 2.a). The catenane was first
characterized by MS (see SI). The MS displays a singly charged
molecular ion (m/z 1723.7) corresponding to the mass of a
dimer, while the MS/MS fragmentation pattern is characteristic
of a interlocked structure, the loss of one ring resulting in the
formation of the cyclic monomer (m/z 861.7) and smaller
fragments.
wavelength for each building block (donor units at 260 nm, and
acceptor units at 383 nm), allowing for easy estimation of the
quantity of catenane formed. The presence of the donor
building block 1 results in a shift in the equilibrium toward the
formation of donor−acceptor [2] and [3]catenanes. Results are
shown in Figures 3 and 4.
Short Linkers (n ≤ 6): Synthesis of Donor−Acceptor
[3]Catenanes and Manifestation of an Unexpected Odd−
Even Effect. For short aliphatic linkers (n ≤ 6), an unexpected
odd−even effect dictates the library behavior, leading to the
formation of either simple or extremely complex libraries.
Acceptor building blocks with an even-numbered linker (n = 2,
4, and 6) produced simple libraries containing significant
quantities of [3]catenane (Figures 3a,c,e). Kinetic experiments
showed the formation of the [3]catenane to be a stepwise
process in which the acceptor dimer is formed first and then
successively threaded by the donor units.4 No other macro-
cycles than the [3]catenane and the intermediate species (the
noninterlocked dimers and the intermediate [2]catenane) were
observed. The [3]catenane is likely to be the thermodynamic
product of the libraries, and its yield is therefore relatively good;
however, the residual presence of intermediates suggests that
the libraries are fully oxidized and that disulfide exchange stops
before the library reaches thermodynamic equilibrium.
In stark contrast, libraries consisting of acceptor building
blocks with an odd-numbered linker (n = 3 and 5) displayed
little selectivity for any particular library member. The n = 3
library is the most striking example, while the effect is less
pronounced with the more flexible (n = 5) before disappearing
when the length of the linker is further increased. In the odd-
numbered libraries, the [3]catenane is clearly not the dominant
species, and we observed the formation of many giant
macrocycles.
The odd−even effect is even more pronounced when the
libraries are prepared in a 1:1 molar ratio of the two building
blocks. Figure 4a illustrates the complexity of the n = 5 library
in a 1:1 molar ratio. The library is characterized by the
formation of a variety of surprisingly large macrocycles and
catenanes, such as 10 (m/z 1555.0, triply charged, Figure 5).
The fragmentation pattern generated by MS/MS is consistent
with that of a giant [2]catenane. The soft fragmentation at 0.4
V shows fragmentation into the separated rings at m/z 1852.0
and 958.8. Smaller fragments were also detected, but no
evidence of linkage between these two rings could be identified,
suggesting that they were not covalently bonded to each other.
The 1H NMR spectrum (500 MHz, 298 K, Figure 2b) of the
all-acceptor [2]catenane in D2O exhibits resonances corre-
sponding to a single highly symmetric conformation: four
doublets were assigned to the four inequivalent aromatic
protons of each naphthalenediimide unit. The chemical shifts of
these protons (δ = 8.7 to 8.5 ppm) are consistent with the
proposed geometry of the all-acceptor [2]catenane, where the
aromatic units are not stacked and therefore do not experience
shielding. On the other hand, the diastereotopic aliphatic
protons sandwiched between two aromatic units experience a
strong shielding environment and thus appear at very low
chemical shift values (from 0.9 to −2.0 ppm). We identified
nuclear Overhauser effect (NOE) correlations between the
naphthalenediimide protons and aliphatic protons H1 to H4
confirming that the aliphatic protons are in close proximity to
the aromatic units. Only the most shielded aliphatic proton H5,
deeply buried within the cavity of the catenane, does not
display any NOE correlations with the aromatic protons.
Moreover, NOE cross peaks were only observed between
alternating protons of the alkyl chain, (H1, H3, and H5) or (H2
and H4), indicating that the alkyl chain is confined to a
staggered zigzag geometry within the core of the catenane.
The formation of this all-acceptor catenane appears to be
purely driven by the need to minimize solvent exposed surface
area. The hydrophobic surfaces are efficiently buried, and the
hydrophilic cysteine side chains are exposed to the solvent. This
structure highlights the importance of hydrophobic effects in
the synthesis of interlocked and folded structures in water.
Although we did not investigate the possibility of synthesizing
an all-donor catenane, this result, along with previous ones,4
suggests that such catenanes might be accessible.
Combining the Hydrophobic Effect with Donor−
Acceptor Interactions: Synthesis of [2] and [3]-
Catenanes. The second set of libraries was prepared by
dissolving each acceptor 2−9 in the presence of donor 1 to give
a total concentration of 5 mM (1:2 molar ratio) in water at pH
8 (1 M NaNO3). Absorbance was recorded at the optimum
19131
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
Figure 4. Reverse-phase HPLC analysis of aqueous libraries composed
of 1 and 5 in the presence of 1 M of NaNO3 in a: (a) 1:1 molar ratio,
(b) 2:1 molar ratio prepared in one pot, and (c) 2:1 molar ratio after
stepwise addition of 1. The total concentration of building block is 5
mM in every case. Absorbance was recorded at 383 nm.
increase of the catenane yield, which is thermodynamically
more stable. Stepwise addition to the n = 5 library is a good
example of this behavior (Figure 4b,c). Adding another
equivalent of fresh thiol 1 to a library prepared in a 1:1
molar ratio dramatically simplified the library in favor of the
[3]catenane. While the final ratio between the building blocks
in the ‘stepwise’ library is equal to the ratio in the ‘one pot’
library, the stepwise procedure led to an increase of the
[3]catenane yield from 10% to 25%. On the other hand, the
giant catenanes, such as 10, disappear in the ‘stepwise’ library,
implying that their formation is mainly kinetically controlled.
The origin of the odd−even effect remains unclear. From our
results, it seems that the [3]catenane is the thermodynamic
product in all the libraries, regardless of the aliphatic linker
length. Although the odd−even effect might influence the
stability of the [3]catenane itself, it seems that it mainly
controls the stability of the intermediates and competing
species met along the kinetic pathways leading to the catenane
formation.
Long Linkers (n > 6): Synthesis of Donor−Acceptor
[2]Catenanes. When the aliphatic linker is long enough (n >
6), templating by the electron-rich building block favors
intramolecular closure of the acceptor building block and
results in the formation of donor−acceptor [2]catenanes that
are entropically favored over the larger [3]catenane. Our
previous studies suggest that the formation of these [2]-
catenanes is possible when the aliphatic linker is longer than a
cysteine−cysteine linkage, which roughly corresponds to a
hexyl chain (n > 6).4 In the n = 7 library, the presence of the
two catenanes (Figure 3f) in the same library suggests that the
[2]catenane might be relatively strained and that each catenane,
being a kinetic trap, is persistent once it is formed. For n = 8
and 9, the [2]catenane is generated in yields of roughly 70% in
otherwise simple libraries. The formation of these [2]catenanes
Figure 3. Reverse-phase HPLC analysis of aqueous libraries composed
of 1 and (a) 2, (b) 3, (c) 4, (d) 5, (e) 6, (f) 7, (g) 8, and (h) 9. These
libraries were prepared in a 1:2 molar ratio (5 mM total) in water at
pH 8 in the presence of 1 M of NaNO3. Absorbance was recorded at
383 nm.
The libraries are fully oxidized after a few hours and may not
reach thermodynamic equilibrium. We showed, however, that
the library distribution can be shifted toward thermodynamic
equilibrium if the reversible process of disulfide exchange is
reinitiated by adding fresh thiol to an oxidized library.4
Therefore, addition of the 2 equiv of building block 1 in one
pot or in successive fractions may lead to a different library
distribution. In all the libraries, stepwise addition of 1 (two
batches of 1 equivalent added at an interval of 2 h) led to an
19132
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
Figure 5. Tandem (a) MS and (b) MS/MS fragmentation of catenane 10.
Figure 6. Comparison of the yield of: (a) homo and heterochiral catenanes and (b) homo and heterochiral catenane following the one pot and the
stepwise procedures (in the case of n = 4).
is highly favored by the tightness of the acceptor cyclic
monomer derived from the building blocks of this study (7 ≤ n
≤ 9), in full agreement with our previous observations and
mechanistic proposal.
observable (see SI). Although the effect of chirality on the
behavior of complex dynamic systems is difficult to ration-
alize,12 comparative kinetic studies suggest that chirality
influences the whole energy landscape of the libraries. The
formation of heterochiral catenanes seems to be recurrently
preferred compared to the formation of catenanes with
homochirality. In the most impressive example, the assembly
of the [2]catenane becomes nearly quantitative for n = 9
(homochiral [2]catenane: 70%, heterochiral [2]catenane:
90%). The formation of the [3]catenanes is also improved
and reaches 50% for n = 4 (homochiral [3]catenane: 33%).
Once again, this yield can be further improved to 70% when 1′
is added in a stepwise fashion. To date, these are the best yields
obtained for the synthesis of these dynamic combinatorial
catenanes.
Influence of Chirality on Catenane Formation. The
aliphatic linker length of the building block is not the only
feature that can be used to control catenane synthesis. In
compact structures, such as catenanes, the negatively charged
carboxylates accumulate around the disulfide bridges, poten-
tially having an adverse effect on their formation. The geometry
of the carboxylates is dictated by the cysteine chiral center in
the building blocks. Therefore, a change in chirality can be used
to influence the energy landscape of the library. Toward this
end, we analyzed libraries composed of 2−9 in the presence of
the enantiomeric building block 1′, derived from D-cysteine
(Figure 6). Indeed, the libraries distribution is dramatically
affected by the change in chirality. Despite the formation of
overall more complex libraries, the odd−even effect is still
1H NMR Characterization of a Family of Catenanes (n =
7). The n = 7 library provides a unique opportunity to isolate
19133
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
1
Figure 7. H NMR spectrum (D2O, 298 K, 500 MHz) of the catenanes formed from 1 and 7. (a) Partial spectrum of the [2]catenane. (b) Partial
spectra of the [3]catenane and the macrocyclic intermediates on the pathway to its formation. An impurity is labeled by *.
1
and compare by H NMR (500 MHz, 298 K, D2O) a [2] and
dimer displays the expected sharp signals of a rapidly rotating
ring. Upon threading a second donor ring, the NMR signals
sharpen remarkably due to the high level of organization of the
aromatic units within the [3]catenane.
[3]catenane produced from the same building blocks, 1 and 7.
The H NMR spectrum of the [2]catenane is characteristic
1
of a fully asymmetric catenane in which each of the aromatic
protons is inequivalent (Figure 7a). In the acceptor region
(7.2−8.2 ppm), eight doublets were assigned by correlation
spectroscopy (COSY) to the eight protons of the naphthale-
nediimide units: four upfield-shifted (inner acceptor) and four
downfield-shifted (outer acceptor). A similar pattern was
observed in the donor region (5.5−7.1 ppm).
Figure 7b shows the NMR spectra of the corresponding
[3]catenane and of each of the macrocyclic intermediates
leading to this [3]catenane. These spectra clearly illustrate how,
during the process of the [3]catenane formation, increasing
rigidity arising from the successive threading of donor dimers
around the acceptor dimer progressively sharpens the NMR
signals. The acceptor dimer displays broad resonances
characteristic of large, conformationally flexible macrocycles.3
Threading of one donor ring does not significantly constrain
the flexibility of the acceptor dimer: its signals are consequently
broad and cannot be assigned. On the other hand, the donor
The complexity of the [3]catenane structure, with two
possible conformations, renders its full characterization difficult
by NMR. However, it is important to point out that the
exchange between the two conformations can be restricted by
both the length of the aliphatic linker and the building block
chirality. Indeed, for n = 4, the homochiral [3]catenane exhibits
only one conformation, while the heterochiral [3]catenane
exhibits two conformations in a molar ratio of 2:1 (see SI). Of
course, the effect of ring size on the energy barrier between the
two conformations is more pronounced, and for n = 7, the two
conformations are present in equal proportions.
CONCLUSIONS
■
We have shown here that the length of the aliphatic linker, the
chirality of the building blocks, and kinetic and thermodynamic
parameters can be efficiently manipulated to synthesize a
variety of [2] and [3]catenanes in high yields.
19134
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135

Journal of the American Chemical Society
Article
(c) Anderson, S.; Claridge, T. D. W.; Anderson, H. L. Angew.
Chem., Int. Ed. 1997, 36, 1310.
In libraries consisting only of the acceptor building blocks
connected via long linkers (n ≥ 8), the hydrophobic effect
drives the formation of all-acceptor [2]catenanes. In the
libraries composed of acceptor and donor building blocks,
changing the length of the linker leads to dramatic variations in
library distributions. For short-linkers (n ≤ 6) we can
distinguish two characteristic types of energy landscapes
highlighting an odd−even effect. In the first situation (even
libraries, n = 2, 4, 6), the formation of the [3]catenane is largely
favored over the other possible macrocycles. In the second
situation (odd libraries, n = 3 and 5), many other macrocycles
compete with the formation of the [3]catenane: the yield of the
[3]catenane is low, and the libraries are dominated by the
presence of many large macrocycles, some of which are also
catenated. Reinitiation of the reversible process through
stepwise addition of 1 leads to the production of the
thermodynamically more stable [3]catenane at the expense of
the unusually large catenanes. Furthermore, changing the
building block chirality influences the library distribution in
favor of catenane formation.
(3) (a) Fujita, M.; Ibukuro, F.; Hagihara, H.; Ogura, K. Nature 1994,
367, 720. (b) Fujita, M.; Ibukuro, F.; Ogura, K. J. Am. Chem. Soc. 1995,
117, 4175. (c) Hori, A.; Kumazawa, K.; Kusukawa, T.; Chand, D. K.;
Fujita, M.; Sakamoto, S.; Yamaguchi, K. Chem.Eur. J. 2001, 7, 4142.
(4) (a) Cougnon., F. B. L.; Au-Yeung, H. Y.; Pantos,̧ G. D.; Sanders,
J. K. M. J. Am. Chem. Soc. 2011, 133, 3198. (b) Au-Yeung, H. Y.;
Pantos, G. D.; Sanders, J. K. M. J. Org. Chem. 2011, 76, 1257. (c) Au-
Yeung, H. Y.; Pantos, G. D.; Sanders, J. K. M. Angew. Chem., Int. Ed.
2010, 49, 5331. (d) Au-Yeung, H. Y.; Pantos, G. D.; Sanders, J. K. M. J.
Am. Chem. Soc. 2009, 131, 16030. (e) Au-Yeung, H. Y.; Pantos, G. D.;
Sanders, J. K. M. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 10466.
(5) Cougnon, F. B. L.; Jenkins, N. A.; Pantos, G. D.; Sanders, J. K. M.
Angew. Chem., Int. Ed. 2012, 124, 1472.
̧
̧
̧
̧
̧
(6) (a) Fang, L.; Wang, C.; Fahrenbach, A. C.; Trabolsi, A.; Botros,
Y. Y.; Stoddart, J. F. Angew. Chem., Int. Ed. 2011, 50, 1805. (b) Fang,
L.; Basu, S.; Sue, C.−H.; Fahrenbach, A. C.; Stoddart, J. F. J. Am.
Chem. Soc. 2011, 133, 396.
(7) For reviews of dynamic combinatorial chemistry see:
(a) Cougnon, F. B. L.; Sanders, J. K. M. Acc. Chem. Res. DOI:
10.1021/ar200240m; (b) Herrmann, A. Org. Biomol. Chem. 2009, 7,
3195. (c) Ladame, S. Org. Biomol. Chem. 2008, 6, 219. (d) Lehn, J.-M.
Chem. Soc. Rev. 2007, 36, 151. (e) Rozenman, M. M.; McNaughton, B.
R.; Liu, D. R. Curr. Opin. Chem. Biol. 2007, 11, 259. (f) Corbett, P. T.;
Leclaire, J.; Vial, L.; West, K. R.; Wietor, J.-L.; Sanders, J. K. M.; Otto,
S. Chem. Rev. 2006, 106, 3652. (g) de Bruin, B.; Hauwert, P.; Reek, J.
N. H. Angew. Chem., Int. Ed. 2006, 45, 2660.
While the origins of the observed odd−even effect are still
unclear, it seems to affect both the shape and the relative energy
of the macrocycles formed in solution and the kinetic pathways
leading to the formation of [3]catenanes. The right balance
between flexibility and rigidity, partially controlled by the odd−
even effect, constitutes an important component of the self-
assembly of complex structures in dynamic systems.
(8) For other examples of [2]catenane discovery from dynamic
combinatorial chemistry: (a) Chung, M.-K.; White, P. S.; Lee, S. J.;
Waters, M. L.; Gagne,
(b) Chung, M.-K.; Lee, S. J.; Waters, M. L.; Gagne,
Soc. 2012, 134, 11430. (c) Chung, M.-K.; White, P. S.; Lee, S. J.;
Gagne, M. R. Angew. Chem., Int. Ed. 2009, 48, 8683. (d) West, K. R.;
Ludlow, R. F.; Corbett, P. T.; Besenius, P.; Mansfeld, F. M.; Cormack,
P. A. G.; Sherrington, D. G.; Goodman, J. M.; Stuart, M. C. A.; Otto, S.
J. Am. Chem. Soc. 2008, 130, 12218. (e) Lam, R. T. S.; Belenguer, A.;
Roberts, S. L.; Naumann, C.; Jarrosson, T.; Otto, S.; Sanders, J. K. M.
Science 2005, 308, 667.
́
M. R. J. Am. Chem. Soc. 2012, 134, 11415.
ASSOCIATED CONTENT
* Supporting Information
Detailed procedure for building block synthesis, library
preparations, HPLC/LC-MS methods and data, UV−vis and
NMR spectra of isolated catenanes and macrocycles. This
material is available free of charge via the Internet at http://
pubs.acs.org.
́
M. R. J. Am. Chem.
■
S
́
(9) For examples of odd−even effect in materials, see: (a) Thuo, M.
M.; Reus, W. F.; Nijhuis, C. A.; Barber, J. R.; Kim, C.; Schulz, M. D.;
Whitesides, G. M. J. Am. Chem. Soc. 2011, 133, 2962. (b) de Jeu, W.
H.; van der Veen, J. Mol. Cryst. Liq. Cryst. 1977, 40, 1. (c) Amabilino,
D. B.; Serrano, J.-L.; Sierra, T.; Veciana, J. J. Pol. Sci. Part A: Polym.
Chem. 2006, 44, 3161. (d) Amabilino, D. B.; Ramos, E.; Serrano, J.-L.;
Sierra, T.; Veciana, J. J. Am. Chem. Soc. 1998, 120, 9126.
(e) Nakashima, H; Koe, J. R.; Torimitsu, K.; Fujiki, M. J. Am. Chem.
Soc. 2001, 123, 4847. (f) Zhi, J.; Zhu, Z.; Liu, A.; Cui, J.; Wan, X.;
Zhou, Q. Macromolecules 2008, 41, 1594. (g) Dauselt, J.; Zhao, J.;
Kind, M.; Binder, R.; Bashir, A.; Terfort, A.; Zharnikov, M. J. Phys.
Chem. C 2011, 115, 2841.
AUTHOR INFORMATION
Corresponding Author
g.d.pantos@bath.ac.uk; jkms@cam.ac.uk
■
Author Contributions
§These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank EPSRC (F.B.L.C., J.K.M.S.), Gates Cambridge
(N.P.), and the University of Bath (G.D.P.) for financial
support and Dr. Ana Belenguer for maintaining the HPLC
facility.
(10) (a) Stals, P. J. M.; Smulders, M. M. J.; Martín-Rapun, R.;
́
Palmans, A. R. A.; Meijer, E. W. Chem.Eur. J. 2009, 9, 2071.
(b) Lermo, E. R.; Langeveld-Voss, B.; Janssen, R. A. J.; Meijer, E. W.
Chem. Commun. 1999, 791. (c) de Silva, A. P.; Sandanayake, K. R. A. S.
Angew. Chem., Int. Ed. 1990, 29, 1173.
(11) Tambara, K.; Ponnuswamy, N.; Hennrich, G.; Pantos,
Org. Chem. 2011, 76, 3338.
(12) (a) Bulos, F.; Roberts, S. L.; Furlan, R. L. E.; Sanders, J. K. M.
Chem. Commun. 2007, 3092. (b) Liu, J.; West, K. R.; Bondy, C. R.;
Sanders, J. K. M. Org. Biomol. Chem. 2007, 5, 778.
̧ G. D. J.
REFERENCES
■
(1) For some recent examples, see: (a) Coskun, A.; Spruell, J. M.;
Barin, G.; Fahrenbach, A. C.; Forgan, R. S.; Colvin, M. T.; Carmieli,
R.; Benítez, D.; Tkatchouk, E.; Friedman, D. C.; Sarjeant, A. A.;
Wasielewski, M. R.; Goddart, W. A., III; Stoddart, J. F. J. Am. Chem.
Soc. 2011, 133, 4538. (b) Spruell, J. M.; et al. Nat. Chem. 2010, 2, 870.
(c) Blanco, V.; Chas, M.; Abella, D.; Peinador, C.; Quintela, J. M. J.
Am. Chem. Soc. 2007, 129, 13978. (d) Chas, M.; Blanco, V.; Peinador,
C.; Quintela, J. M. Org, Lett. 2007, 9, 675. (e) Hubbard, A. L.;
Davidson, G. J. E.; Patel, R. H.; Wisner, J. A.; Loeb, S. J. Chem.
Commun. 2004, 138.
(2) (a) Wenz, G.; Han, B. H.; Muller, A. Chem. Rev. 2006, 106, 782.
(b) Harada, A.; Li, J.; Kamachi, M. Nature 1994, 370, 126.
19135
dx.doi.org/10.1021/ja3075727 | J. Am. Chem. Soc. 2012, 134, 19129−19135