A Strategy to Synthesize Molecular Knots and Links Using the Hydrophobic Effect

Article abstract of DOI:10.1021/jacs.8b05220

Conventional approaches to the synthesis of molecular knots and links mostly rely on metal templation. We present here an alternative strategy that uses the hydrophobic effect to drive the formation of complex interlocked structures in water. We designed an aqueous dynamic combinatorial system that can generate knots and links. In this system, the self-assembly of a topologically complex macrocycle is thermodynamically favored only if an optimum packing of all its components minimizes the hydrophobic surface area in contact with water. Therefore, the size, geometry, and rigidity of the initial building blocks can be exploited to control the formation of a specific topology. We illustrate the validity of this concept with the syntheses of a Hopf link, a Solomon link, and a trefoil knot. This latter molecule, whose self-assembly is templated by halides, binds iodide with high affinity in water. Overall, this work brings a fresh perspective on the synthesis of topologically complex molecules: Solvophobic effects can be intentionally harnessed to direct the efficient and selective self-assembly of knots and links.

Full text of DOI:10.1021/jacs.8b05220

Subscriber access provided by Kaohsiung Medical University
Article
A Strategy to Synthesize Molecular Knots
and Links Using the Hydrophobic Effect
Fabien B. L. Cougnon, Kenji Caprice, Marion Pupier, Antonio Bauza, and Antonio Frontera
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 28 Aug 2018
Downloaded from http://pubs.acs.org on August 29, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
A Strategy to Synthesize Molecular Knots and Links Using the Hydrophobic
Effect
a
a
a
b
Fabien B. L. Cougnon, * Kenji Caprice, Marion Pupier, Antonio Bauzá and Antonio
b
Frontera
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Department of Organic Chemistry, University of Geneva, 30 Quai Ernest-Ansermet
1
211 Geneva 4, Switzerland
b
Department de Química, Universitat de les Illes Balears, Carretera de Valldemossa km 7.5,
07122 Palma de Mallorca, Baleares, Spain
E-mail: fabien.cougnon@unige.ch
Abstract
Conventional approaches to the synthesis of molecular knots and links mostly rely on
metal-templation. We present here an alternative strategy that uses the hydrophobic effect
to drive the formation of complex interlocked structures in water. We designed an aqueous
dynamic combinatorial system that can generate knots and links. In this system, the self-
assembly of a topologically complex macrocycle is thermodynamically favored only if an
optimum packing of all its components minimizes the hydrophobic surface area in contact
with water. Therefore, the size, geometry and rigidity of the initial building blocks can be
exploited to control the formation of a specific topology. We illustrate the validity of this
concept with the syntheses of a Hopf link, a Solomon link and a trefoil knot. This latter
molecule, whose self-assembly is templated by halides, binds iodide with high affinity in
water. Overall, this work brings a fresh perspective on the synthesis of topologically complex
molecules: solvophobic effects can be intentionally harnessed to direct the efficient and
selective self-assembly of knots and links.
Introduction
1
Nearly thirty years after the synthesis of the first molecular trefoil knot, topologically
complex molecules remain extremely difficult to produce and their properties are nearly
2
unexplored. In traditional synthetic approaches, ligands are pre-organized using metal
coordination and linked together to generate molecular knots and links. This approach,
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
1
4
originally devised by Sauvage, who notably synthesized a trefoil knot and a Solomon link
2
b, 5-7
from linear helicates, has been more recently revisited by Leigh.
The Leigh group
considerably expanded the complexity of available interlocked architectures by replacing
linear helicates with circular helicates. Their most impressive example is probably the recent
7
,8
synthesis of a 819 knot with eight crossings. Surprisingly, supramolecular interactions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
other than metal coordination have rarely been employed to promote the synthesis of
9-12
knots.
Using donor-acceptor π-π interactions to build structures more complex than
2
,8h
singly-interlocked catenanes has proved challenging.
Recent studies by Sanders et al.
demonstrated that macrocycles could adopt non-trivial topologies in water to minimize
11,12
11-15
solvent exposed hydrophobic surfaces.
The idea that the hydrophobic effect,
and
perhaps more generally solvophobic effects, could provide an alternative driving force to
promote the formation of knots and links in the absence of metals is extremely appealing.
However, translating these observations into a reliable synthetic strategy represents a
critical step that has not been overcome until today.
We present now a strategy that uses the hydrophobic effect to direct the synthesis of
1
6-
knots and links. We considered the virtual hydrazone-based dynamic combinatorial library
1
8
(
Figure 1) engendered by dialdehyde A (blue cartoon) and dihydrazide H (transparent
cartoon). Dialdehyde A is a hydrophobic building block constituted of two quinolinium-5-
carboxaldehyde units connected with a p-xylyl unit. Dihydrazide H is a fictive neutral
hydrophobic building block that could be ideally stretched and bent to enable the closure of
any macrocycle. Folding the quinolinium-based units (A) in a loop-like conformation around
the neutral hydrophobic moiety of H represents the optimum way to decrease the number
of hydrophobic surfaces exposed to the environment. The permanent positive charges of
the quinolinium units remains on the outer surface, while the neutral moiety H and a large
part of the aromatic surfaces of A are sheltered from water.
ACS Paragon Plus Environment

Page 3 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. A virtual dynamic combinatorial library of knots and links generated from
dialdehyde A (blue cartoon) and a fictive dihydrazide H (transparent cartoon). All these
structures were constructed from a basic motif (top right) multiplied n times using an axis of
n n
symmetry C or S (represented by the red dot).
n n
This motif was simply multiplied with an n-fold rotation axis (C or S ) to construct
macrocycles with minimum surface area exposed to water. Each time, we obtained a unique
architecture. These architectures consistently alternated blue and transparent components
and were characterised by a topological complexity of n crossings (Figure 1). A C
generated a Hopf link or singly interlocked [2]catenane (two crossings); a C axis generated
a trefoil knot (three crossings); a C axis generated a Solomon link or doubly interlocked
[2]catenane (four crossings); a S axis generated a figure-eight knot (four crossings), etc. The
2
axis
3
4
4
list presented here, intentionally limited to the simplest topologies tabulated by
18
mathematicians, is non-exhaustive. Many other topologies could be similarly constructed
through the same process of multiplication. Building blocks A and H could therefore ideally
produce a complex library of knots and links. Macrocycles with trivial topologies may also
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
form, but their formation should not be thermodynamically favored because their
hydrophobic surfaces would remain exposed to water. Even from the simplified cartoon
representation showed in Figure 1, it is noticeable that building block A (blue cartoon) can
only generate all these topologies if H is indefinitely stretchable and bendable. Indeed, each
topology enforces specific geometrical constraints on their backbone. This consideration
was important as it suggested that the geometry and the flexibility of building block H could
be used to bias the outcome of this virtual library in favor of a given topology.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. HPLC traces of the libraries generated from dialdehyde A·2X (2.5 mM) and
dihydrazide a) H1, b) H2 and c) H3 (2.5 mM). The yields were evaluated from the integration
of the peak areas using catechol as an internal standard. The HPLC traces of the libraries
prepared from A·2X and H1, or from A·2X and H2, were identical for any of the counter-ions
-
X tested (Figures S51).
Here, we demonstrate that the choice of the dihydrazide (H1-3) is indeed critical and
allows for the selective amplification of three topologies from the virtual pool described
above: a Hopf link, a Solomon link and a trefoil knot. The syntheses described here are the
result of a rapid screening of readily-available dihydrazides. In the following sections, we
describe these syntheses as three independent stories. In each case, we also investigated
-
the role of the quinolinium counter-ion (X ) on the outcome of the libraries. We first
ACS Paragon Plus Environment

Page 5 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
prepared analytical-scale hydrazone-based dynamic combinatorial libraries following a
1
5
20
protocol optimized by Sessler and Li. Equimolar amounts of dialdehyde A·2X (2.5 mM)
and dihydrazide H1-3 (2.5 mM) were dissolved in D O (700 L, 0.01% TFA). The libraries
2
ꢀ
were stirred overnight at 70
°C. The structures formed were all purified by semi-preparative
-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HPLC. During the process of purification, the counter-ion X was replaced by the
trifluoroacetate anion present in the HPLC eluent. All the products were thus isolated as
trifluoroacetate salts.
1
We identified knots and links using H NMR and ion-spray mass spectrometry (ESI-MS).
The interlocked structures represented in Figure 1 feature n equivalent arms, each
comprised an outer quinolinium-based loop (originating from dialdehyde A) surrounding an
inner thread (originating from dihydrazide H). In all these structures, the moiety H is buried
inside and inevitably experiences a symptomatic NMR shielding.
ESI-MS provided additional structural information and notably allowed for the
characterization of the type of topology formed: a Hopf link is a [2+2] species and fragments
into two [1+1] species; a trefoil knot is a [3+3] species and fragments as a single macrocycle;
a Solomon link is a [4+4] species and fragments into two [2+2] species; a figure-eight knot is
a [4+4] species and fragments as a single macrocycle; etc. If the identification of the
topologies formed becomes more difficult as the number of crossing increases, this
analytical method allows for the differentiation of the simplest knots and links shown in
Figure 1.
Finally, theoretical calculations of the knots and links were carried out using the DFT-D3
methodology, providing useful information regarding the adequate conformation of each
building block within the interlocked architectures. The optimized geometries compared
well with the NMR data and provided further evidence of the structural viability of the knots
and links synthesized.
A Hopf link, or singly interlocked [2]catenane
High-performance liquid chromatography (HPLC, Figure 2a) revealed that A·2X and
dihydrazide H1 generated a rather simple library, mostly dominated by one product, 1
-
(~70%). This outcome was independent of the choice of the counter-ion X (Figure S51). ESI-
MS analysis showed that 1 was a [2+2] species. Three major peaks at mass-to-charge ratio
4+
3+
2+
m/z 299.15, 398.54 and 597.29, corresponded to [1] , [1-H] and [1-2H] , respectively
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(Figure S15). At a collision voltage of 45 V, dissociation of 1 yielded a fragmentation pattern
characteristic of a [2]catenane composed of two interlocked [1+1] macrocycles (Figure
21
S17). The [2+2] species 1 is relatively simple and cannot be interlocked more than once,
implying that 1 was a singly interlocked [2]catenane, or Hopf link.
1
The Hopf link 1 was isolated by HPLC and further characterized by H NMR in D O. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
spectrum of 1·4CF CO at 298 K (Figure 3a) confirmed that it was composed of two identical
3
2
arms, each constituted of a quinolinium-based loop (A) threaded by the aliphatic chain (H1).
Threading of the aliphatic chain was clearly manifested by a significant upfield shift of the
aliphatic protons k, l and m (below 0 ppm) compared to the same protons in the parent
building block H1 (~1.16-1.78 ppm). Moreover, relatively intense nuclear Overhauser effect
(
NOE) cross-peaks between the aliphatic protons and the surrounding quinolinium and xylyl
protons supported the proposed structure (Figure S10). In spite of its interlocked nature, 1
remained a flexible structure. The ability of the hydrazone bond to rotate freely was notably
highlighted by the presence of NOE correlations between the hydrazone proton a and both
protons b and g of the neighboring quinolinium (Figure S9). Furthermore, the presence of
several minor conformations in dynamic exchange was noticeable at 298 K. Their proportion
considerably increased at lower temperatures (below 298 K, Figure S7).
The minimum energy structure of the Hopf link is represented in Figure 3c. The H-atoms
of the flexible aliphatic chain were located between the quinolinium π-systems, in
agreement with the upfield shift observed by NMR. The analysis of bond lengths and angles
indicated that this link did not present any strained functional group, bond or ring, thus
explaining its dominant formation (~70%).
ACS Paragon Plus Environment

Page 7 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Figure 3. Comparison between the H NMR spectra (500 MHz, 298 K, D
2
O) of a) the Hopf
3 2
link 1·4CF CO and b) the parent dihydrazide building block H1. c) BP86-D3/def2-TZVP
optimized geometry of Hopf link 1 in sticks (left) and space filling (right). H-atoms were
omitted for clarity. The hydrazone bonds were represented in yellow, moiety A in blue and
moiety H1 in red.
The same library also contained a small amount of the [1+1] macrocycle 1´ (~25%, Figure
2
3 2 3 2
a). We isolated 1´·2CF CO to compare its NMR features to that of the Hopf link 1·4CF CO .
Macrocycle 1´ can only be topologically trivial and did not feature any major change of
chemical environment compared to the original building blocks (Figure S6). The only
noteworthy exception was that of the aldehyde proton a (10.42 ppm in A) transformed
during the reaction into a hydrazone proton (8.65 ppm in 1´).
The self-assembly of a Hopf link (1), driven by the hydrophobicity of the aliphatic moiety,
1
4,20
is in full agreement with findings reported in other studies.
of the length of the aliphatic chains (Figure S57). The formation of the Hopf link was always
favored if the length of the aliphatic chain was superior or equal to (-CH -) . In all these
We also explored the effect
2
6
cases, the simplest non-trivial topology is formed because the aliphatic chain is highly
flexible. Introducing more rigidity in the backbone of the dihydrazide building block should
prevent the formation of a Hopf link, thereby opening up the possibility to build other more
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
complex topologies. It is precisely what was observed in the two following libraries,
obtained from the rigid isophthalic (H2) and 2,7-naphthalene (H3) building blocks.
A Solomon Link, or doubly interlocked [2]catenane
The HPLC trace in Figure 2b showed that reaction between A·2X and the isophthalic
dihydrazide H2 almost exclusively generated a single product, 2 (>90%). Here again, this
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-
outcome was independent of the counter-ion X used (Figure S51). To our surprise, ESI-MS
4
+
3+
2+
indicated that 2 was a [4+4] species (m/z [2-4H] = 575.18, [2-5H] = 766.91, [2-6H] =
1
150.36, Figure S30). The exclusive self-assembly of such a large macrocycle was remarkable
and immediately caught our attention. Collision-induced dissociation of 2 produced a
fragmentation pattern corresponding to a [2]catenane composed of two interlocked [2+2]
1
2
macrocycles (Figure S31). This fragmentation pattern immediately ruled out the possibility
that 2 could be a trivial or a knotted macrocycle, such as a figure-eight knot. On the other
hand, two [2+2] macrocycles may be interlocked more than once, raising our hope that 2
would exhibit a topology more complex than that of a Hopf link.
1
Figure 4. Comparison between the H NMR spectra (500 MHz, 298 K, D
2
O) of a) the
3 2
Solomon link 2·8CF CO and b) the parent dihydrazide building block H2. c) Several views of
ACS Paragon Plus Environment

Page 9 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
the BP86-D3/def2-TZVP optimized geometry of Solomon link 2. H-atoms were omitted for
clarity. The hydrazone bonds were represented in yellow, moiety A in blue and moiety H2 in
pink.
1
The H NMR spectrum of the isolated [4+4] species 2·8CF CO in D O was significantly
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
2
complex (Figure 4a), with aromatic protons spanning an astonishingly wide range of
chemical shifts from 10.91 to 3.88 ppm. The spectrum displayed sharp signals corresponding
to a molecule composed of four identical arms, each constituted of a quinolinium-based
loop (A) threaded by an isophthalic moiety (H2). The unexpected complexity of the
spectrum arose from the inequivalence of every proton within an individual arm. The most
significant upfield shifts, compared to the parent building blocks, were observed for the
isophthalic protons k (7.94 ppm in H2 to 3.88 ppm in 2) and l (7.66 ppm in H2 to 6.04 ppm in
2
). These shifts not only indicated that the isophtahlic moiety was stacked between the
quinolinium units, but also that protons k and l were pointing directly towards the
perpendicular xylyl unit. Proton j, located on the opposite side of the isophathlic unit, was
therefore oriented towards the central cavity of the molecule. As a consequence, it did not
experience any strong shielding effect from nearby aromatic units (8.07 ppm in H2 versus
7
.98 ppm in 2).
Each arm of the structure possessed two hydrazone bonds, with one on each side of the
2 2
isophthalic moiety. NOE correlations obtained in a 95:5 H O/D O mixture (Figure S25)
disclosed that these two hydrazone bonds adopted different conformations, with one NH
pointing toward the interior (observed correlations: j↔NH↔a↔b) and another one
pointing toward the exterior (observed correlations: k´↔NH´↔a´↔g´) of the structure.
To conclude, the full analysis of this spectrum highlighted that 2 was a rigid, twisted
molecule locked into a well-defined conformation. Importantly, two pairs of diastereotopic
protons (h and h´) suggested that the topology of 2 induced a chiral environment. The only
topology consistent with such an NMR spectrum, with the mass of a [4+4] species and with
the MS/MS fragmentation pattern observed is the topology of a Solomon link, or doubly
interlocked [2]catenane.
At higher temperatures, exchange cross-peaks appeared between pairs of inequivalent
protons (a↔a
´
, b↔b
´
, etc.), corresponding to a global inversion of the conformation of all
) of the molecule. The activation parameters were
the eight hydrazone bonds (NH↔NH
´
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
calculated from the cross-peaks intensity at variable temperatures (Figures S22-S23). An
−1
unfavourable enthalpy of activation (
ꢁ
H⧧ = +115 kJ
—
mol ) and a favourable entropy of
−1
−1
⧧
activation (
ꢁ
S⧧ = +106 J
—K
—
mol ) contributed to a rather large energy barrier (
ꢁ
G = +838
−
1
kJ
structure.
The C -symmetric Solomon link was easy to model using an antiparallel conformation of
—mol at 298 K) reflecting the penalty associated to the re-organisation of the entire
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
the quinolinium moieties in each arm and an opposite conformation of the two hydrazone
bonds located at both sides of the isophthalic moiety (see Figure 4c, pink fragments). The
minimum energy structure agrees well with the NOE correlations (Figures S24-S25),
providing further support to the formation of the Solomon link. In addition, the DFT-D3
study on this system strongly supported the idea that the formation of a figure-eight knot
from building blocks A and H2 was not possible: all attempts to model the figure-eight knot
topology were unsuccessful due to its highly strained geometry.
The near-quantitative self-assembly of Solomon link 2 demonstrated the importance of
introducing rigidity in the design of the dihydrazide building block to favor the formation of
complex topologies. The yield of the synthesis was particularly impressive, presumably
because the geometry of each unit allowed an optimum packing of all the hydrophobic
aromatic surfaces. Indeed, when the same reaction was performed with the closely related
terephthalic dihydrazide, we only observed the formation of a complex mixture of products
with undecipherable NMR features (Figure S56). The angle between the hydrazide bonds
thus seemed to be of crucial importance. Therefore, when we designed building block H3,
we maintained a similar rigidity and a similar angle between the hydrazide bonds and we
chose to increase the
π surface.
A Trefoil knot
The 2,7-naphthalene-based dihydrazide building block H3 (Figure 2c) also generated a
3
+
relatively simple library composed of two main species, a [3+3] species (3, m/z [3-3H] =
2
+
6+
6
3
25.23, [3-4H] = 937.84, Figure S45) and a [4+4] species (3´, m/z [3´-2H] = 417.14, [3´-
5+
4+
H] = 500.57, [3´-4H] = 625.20, Figure S46). The formation of two such large macrocycles
was already intriguing. More importantly, the HPLC traces depicted in Figure 2c showed that
-
the outcome of the library was highly dependent on the choice of the counter-ion X . Only
-
-
trace amounts of 3 were found in the absence of halides (X = CF
3
CO
2
). In the presence of
ACS Paragon Plus Environment

Page 11 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
-
-
-
-
halides (X = Cl , Br or I ), 3 was formed in non-negligible quantities. The yield of 3 increased
-
regularly with the size of the anion, from ~45% for the smallest halide used (Cl ) to ~72% for
-
23
the largest one (I ). This amplification phenomenon strongly suggested that 3 was binding
halides in a size-dependent manner. Selective anion recognition is challenging in pure
2
4
water, so we focused our attention on the characterization of 3.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
After HPLC purification, the [3+3] macrocycle 3·6CF CO displayed broad, featureless,
3
2
1
uninterpretable H MNR signals in D O (Figure 5a), indicating its ability to adopt many
2
conformations in slow exchange. We were delighted to observe a progressive signal
sharpening upon addition of either NaCl, NaBr or NaI (Figure S34). Figure 5b shows, as a
typical example, the spectrum of 3·6CF CO after addition of ca. 10 equivalents of NaBr.
3
2
Minor conformations remained visible in the spectrum baseline after addition of NaBr
(Figures S34 and S39). However, the spectrum of the bromide-bound macrocycle mainly
showed a single conformation that echoed a highly symmetrical structure, consistent with a
molecule composed of three identical arms, each constituted of a quinolinium-based loop
(A) threaded by the naphthalene moiety (H3). Compared to the parent building block H3,
the symptomatic upfield shift of the naphthalene protons l (7.91 to 4.14 ppm) and k (8.10 to
6
.98 ppm) suggested that: 1) each naphthalene moiety was stacked between two
quinoliniums; 2) protons l and k were pointing towards the xylyl unit.
Such a spectrum and the mass of a [3+3] species were indicative of a trefoil knot
2
5
topology. In agreement with this interpretation, the presence of diastereotopic protons
h1 and h2) reflected a topologically chiral environment. NOE correlations (b↔a↔NH↔j,
Figure S37) in a 95:5 H O/D O mixture revealed that all the six hydrazone bonds of the
(
2
2
bromide-bond macrocycle adopted an identical conformation. Protons NH and j were both
oriented towards the center of the molecule and a total of six NH hydrazone pointing
toward the central cavity provided an ideal binding site for halides. The presence of bromide
at the center of the molecule was further confirmed by a downfield shift of the nearby
proton j (from 8.14 in H3 to 9.32 ppm in 3).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
Figure 5. Comparison between the H NMR spectra (500 MHz, 298 K, D
2
O) of a) the
3 2 3 2
trefoil knot 3·6CF CO , b) the trefoil knot 3·6CF CO in the presence of ca. 10 equivalents of
NaBr, and c) the parent dihydrazide building block H3. d) Several views (sticks and CPK) of
the BP86-D3/def2-TZVP optimized geometry of the trefoil knot 3 complexed to two bromide
anions (3 could also accommodate two iodide ions). H-atoms were omitted for clarity. The
hydrazone bonds were represented in yellow, moiety A in blue and moiety H3 in gold.
We also succeeded in constructing and optimizing the trefoil knot using DFT-D3
calculations. The optimized geometry in water showed that the central cavity should be able
-
-
-
to incorporate two spherical anions (Cl , Br or I ) since six NH groups were pointing to the
main symmetry axis in two different layers, thus generating a ditopic cavity (Figure S58). In
Figure 5d, we show the optimized geometry of the trefoil knot interacting with two bromide
3
anions in water. The minimum energy structure presents C symmetry where each Br atom
ACS Paragon Plus Environment

Page 13 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
interacts with three NH and three CH groups, thus establishing six hydrogen bonds (Figure
S58B). We concluded that the amplification originally observed in the libraries (Figure 2c)
could be explained by the selective binding of halides to the central cavity of the [3+3]
trefoil knot 3·6CF CO rather than to the [4+4] macrocycle 3´·8CF CO . Indeed, no
3
2
3
2
interaction between the [4+4] macrocycle 3´·8CF CO and halides could be evidenced by
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
NMR (Figure S35), microcalorimetry (Figure S50) and ESI-MS (Figure S47). Unfortunately,
´·8CF CO displayed broad NMR signals in D O, even at high temperatures and after
3
3
2
2
addition of NaBr (Figure S35). Consequently, the structural basis underlying its lack of
interaction with halides remained unknown.
We explored the interaction between the [3+3] macrocycle 3·6CF CO and each halide by
3
2
2
6
microcalorimetry (Figure S49). As expected, the strongest binding was observed with
iodide (NaI). The data confirmed that two equivalents of iodide could bind with high affinity
4
-1
−1
(
(
K = 1.6 H = -20.2 kJ
—
10 M ). Binding was driven both by enthalpy (
ꢁ
—mol ) and by entropy
−1
−1
ꢁ
S = +12.8 J—K —mol ). Early saturation during the titration with NaBr indicated that
binding with bromide was about an order of magnitude lower but the binding constant
could not be measured. Nevertheless, the stoichiometry of both the iodide and the bromide
4
+
4+
complexes were confirmed by ESI-MS, as both the unbound ([3-2H] ) and bound ([3+2X] )
species were identified (Figure S47) in the MS spectra of the crude libraries. We compared
theoretically the affinity of 3 for bromide and iodide by evaluating the following substitution
–
–
–
–
2 6 2 6
reaction: 3·2Br + 2[I(H O) ] → 3·2I + 2[Br(H O) ] (eq. 1). We used explicit water
molecules in the first coordination sphere and a dielectric continuum model (COSMO) to
account for solvent effects. This procedure was more convenient than the direct
comparison of interaction energies because the energy of the free knot at room
temperature was not needed. The reaction energy of the substitution of bromide by iodide
(
eq. 1 and Figure S59) at the BP86-D3/def2-TZVP level of theory was energetically favorable
−
1
(ΔG = –10.03 kJ—mol ), which was good agreement with the experimental data that
suggested that binding of bromide was an order of magnitude lower compared to iodide.
Binding with chloride, presumably even weaker, could not be measured by microcalorimetry
4
+
either. The adduct [3+2Cl] also remained elusive in ESI-MS. Overall, the affinity of
3·6CF CO with halides appeared to be weaker than we had expected from the magnitude
3
2
of the amplifications observed during the synthesis. It is noteworthy to mention that the
-
CF
3
CO
2
anion is likely to interact with the NHs of the trefoil knot and probably competes
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with the binding of halides during the microcalorimetric titrations performed on the isolated
trifluoroacetate salt 3·6CF CO . In contrast, only a catalytic amount of trifluoroacetic acid
3
2
was present during the syntheses carried out using the halide salts A·2X.
The self-assembly of trefoil knot 3 is driven by the hydrophobicity of the naphthalene
moiety and is templated by halides, which presumably direct all the hydrazone bonds in the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
conformation required to close the structure. Several knots and links binding halides in their
5,27-29
central cavities have already been reported.
Here, this last example highlights that a
complex interplay of subtle interactions can participate in the selection of a specific
topology. The broad NMR features of 3·6CF CO in D O in the absence of halides (Figure 5a)
3
2
2
remained puzzling. To complete this study, we performed additional experiments to
ascertain the knotted topology of 3·6CF CO . The ability of a molecule to adopt many
3
2
conformations in slow exchange is highly dependent on the solvent used. On the other
hand, a knot cannot unfold once it has been closed and maintains its topology
independently of the solvent used. Illustrating these two statements, we observed that the
1
H NMR signals of 3·6CF CO in CD CN were sharp and retained all the symptomatic features
3
2
3
characteristic of a trefoil knot (e.g. diasteretopic protons and significantly upfield-shifted
naphthalene protons, Figure S43). Furthermore, splitting of the signals upon addition of the
anionic chiral shift reagent (R)-2-phenyl-propioniate corroborated that 3 was produced as a
racemic mixture of two enantiomers, the right- and left-handed knots (Figure S44).
With this trefoil knot, we complete a first catalogue of several structures with complex
topologies. The hydrodynamic radii R
H
, evaluated from the NMR diffusion coefficients
(
Figures S11, S26 and S38), mirrored the difference in size between all these species: 0.66
nm for the [1+1] species 1´; 0.72 nm for the [2+2] species 1; 0.87 nm for the [3+3] species 3;
and 1.06 nm for the [4+4] species 2. These values are in reasonable agreement with the
theoretical values (0.57 nm for 1’, 0.64 nm for 1, 0.92 nm for 3 and 1.12 nm for 2).
Conclusions
To conclude, we proposed an original strategy to synthesize molecules with complex
topologies and we illustrated our approach with the synthesis of three structures: a Hopf
link (1), a Solomon link (2) and a trefoil knot (3). We are working on obtaining crystals
suitable for diffraction to prove these topologies unambiguously. Identifying topologies
before obtaining an X-ray crystal structure is a challenging intellectual puzzle. We applied a
ACS Paragon Plus Environment

Page 15 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
solution-based analytic method to characterize the type of topology formed and the results
of our interpretation were all confronted with success to DFT calculations. In some cases (3
1
and 3’), the lack of H NMR signals in D O increased the difficulty to assign the topology of
2
the isolated macrocycles. If crystal structures ultimately confirm the topology of all three
molecules 1-3, this work will probably represent the simplest strategy to date to access
molecular knots and links. Our approach is simple, cheap, particularly high-yielding and
requires minimum effort in terms of synthesis and purification.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The formation of these interlocked architectures was relatively fast (< 3 hours, Figure
S55) and mostly driven by the hydrophobic effect. Indeed, the yield of the three
macrocycles 1-3 decreased when the libraries were prepared in water/DMSO mixtures
(
Figure S52). However, the structures survived a relatively large ratio of DMSO (>20% of
DMSO in the case of 2 and 3), suggesting that several supramolecular interactions such as
intramolecular hydrogen bonds, hydrogen halide bonds and
in the process of self-assembly.
π-π interactions also intervene
The interlocked structures 1-3 are stable in the aqueous medium for several weeks. The
hydrazone bond is reversible but each intertwined molecule represents a kinetic trap from
which it is difficult to escape (Figure S53). In contrast, non-interlocked macrocycles such as
the trivial [1+1] macrocycle 1
isolated sample of 1 ·4CF CO
´
continue to evolve over time. The NMR spectrum of the
left for two months in the fridge (278 K) in D O in neutral
´
3
2
2
conditions showed a slow reorganization into the more stable Hopf link 1 (Figure S54).
From a more fundamental perspective, the hydrophobicity, the size and the geometry of
the building blocks are certainly determining factors in the process of selection and
amplification of a specific topology from a virtual dynamic pool of knots and links. The exact
nature of the relationship between the type of dihydrazide building block employed and the
topology naturally selected is not yet understood. It is currently impossible to predict the
formation of a knot or a link in our system. Due to the simplicity of implementation, the
system presented here should nevertheless allow us to understand better how the
hydrophobic effect can be exploited to construct specific topologies. The hydrophobicity of
the building blocks is undoubtedly the main feature that drives the formation of metal-free
molecular knots and links in water. Therefore, we expect that our strategy will not be
limited to quinolinium-based building blocks and may be generalizable to other hydrophobic
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
chemical structures. A synergetic combination of the hydrophobic effect and donor-
acceptor π-π interactions or metal-templation may prove particularly powerful in the future.
Supporting Information
1
13
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The synthesis and characterization ( H and C NMR, HR-MS and MS/MS) of the building
blocks and of the interlocked structures 1-3 are available in the Supporting Information, as
well as computational methods and cartesian coordinates of all optimized complexes. We
also provided additional figures with the optimized structures and details on the binding
energy calculations of the halide complexes.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work is dedicated to Prof. Jeremy K. M. Sanders. We would like to take the
opportunity to thank him for his kind support and for allowing us to use his LC-MS. The
Swiss National Science Foundation (PZ00P2_161270), the Department of Organic Chemistry
at the University of Geneva and the MINECO/AEI of Spain (projects CTQ2014-57393-C2-1-P
and CTQ2017-85821-R FEDER funds) are gratefully acknowledged for financial support. We
thank Dr. Ana Belenguer for help with the LC-MS. We also thank Dr. Daniel Palomino and
Irina Dorin (Malvern Panalytical Ltd) for help with ITC data processing. Finally, we
acknowledge Emmanuel Varesio, Harry Théraulaz and Eliane Sandmeier at the Mass
Spectrometry Core Facility of the University of Geneva for performing thorough HR-MS
characterizations.
References and footnotes
(
(
1) Dietrich-Buchecker, C. O.; Sauvage, J.-P. J. Am. Chem. Soc. 1989, 28, 189.
2) (a) Forgan, R. S.; Sauvage, J.-P.; Stoddart, J. F. Chem. Rev. 2011, 111, 5434; (b) Ayme, J.-F.;
Beves, J. E.; Campbell, C. J.; Leigh, D. A. Chem. Soc. Rev. 2013, 42, 1700; (c) Fielden, S. D. P.;
Leigh, D. A.; Woltering, S. L. Angew. Chem. Int. Ed. 2017, 56, 11166.
(3) Sauvage, J.-P. Acc. Chem. Res. 1990, 23, 319.
(
3
4) Nierengarten, J.-F.; Dietrich-Buchecker, C. O.; Sauvage, J.-P. J. Am. Chem. Soc. 1994, 116,
75.
ACS Paragon Plus Environment

Page 17 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(5) Ayme, J.-F.; Beves, J. E.; Leigh, D. A.; McBurney, R. T.; Rissanen, K.; Schultz, D. Nat. Chem.
2
011, 4, 15.
(
(
6) Leigh, D. A.; Pritchard, R. G.; Stephens, A. J. Nat. Chem. 2014, 6, 978.
7) Danon, J. J.; Krüger, A.; Leigh, D. A.; Lemonnier, J.F.; Stephens, A. J.; Vitorica-Yrezabal, I.
J.; Woltering, S. L. Science 2017, 355, 159.
8) For other examples of successful syntheses of knots and links, see: (a) Chichak, K. S.;
(
Cantrill., S. J.; Pease, A. R.; Chiu, S.-H.; Cave, G. W. V.; Atwood, J. L.; Stoddart, J. F. Science
2004, 304, 1308; (b) Prakasam, T.; Lusi, M.; Elhabiri, M.; Platas-Iglesias, C.; Olsen, J.-C.; Asfari,
Z.; Cianférani-Sanglier, S.; Debaene, F.; Charbonnière, L. J.; Trabolsi, A. Angew. Chem. Int. Ed.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
013, 52, 9956; (c) Carina, R. F.; Dietrich-Buchecker, C.; Sauvage, J.-P. J. Am. Chem. Soc. 1996,
18, 9110; (d) Sawada, T.; Yamagami, M.; Ohara, K.; Yamaguchi, K.; Fujita, M. Angew. Chem.
Int. Ed. 2016, 55, 4519. (e) Guo, J.; Mayers, P. C.; Breault, G. A.; Hunter, C. A. Nat. Chem.
010, 2, 218; (f) Wood, C. S.; Ronson, T. K.; Belenguer, A. M.; Holstein, J. J.; Nitschke J. R. Nat.
2
Chem. 2015, 7, 354; (g) Liu, Y.; Ma, Y.; Zhao, Y.; Sun, X.; Gándara, F.; Furukawa, H.; Liu, Z.; Zhu,
H.; Zhu, C.; Suenaga, K.; Oleynikov, P.; Alshammari, A. S.; Zhang, X.; Terasaki, O.; Yaghi, O. M.
Science 2016, 351, 365; (h) Peinador, C.; Blanco, V.; Quintela, J. M. J. Am. Chem. Soc. 2009,
1
31, 920.
(
(
(
9) Safarowsky, O.; Nieger, M.; Frölich, R.; Vögtle, F. Angew. Chem. Int. Ed. 2000, 39, 1616.
10) Elbaz, J.; Wang, Z.-G.; Wang, F.; Willner, I. Angew. Chem. Int. Ed. 2012, 51, 2349.
11) Ponnuswamy, N.; Cougnon, F. B. L.; Clough, J. M.; Pantoş, G. D.; Sanders, J. K. M. Science
2012, 338, 783.
(12) Ponnuswamy, N.; Cougnon, F. B. L.; Pantoş, G. D.; Sanders, J. K. M. J. Am. Chem. Soc.
2
014, 136, 8243.
(
(
13) Fujita, M.; Ibukuro, F.; Hagihara, H.; Ogura, K. Nature 1994, 367, 720.
14) Caprice, K.; Pupier, M.; Kruve, A.; Schalley, C. A.; Cougnon, F. B. L. Chem. Sci. 2018, 9,
1
317.
(
15) Li, H.; Zhang, H.; Lammer, A. D.; Wang, M.; Li, X.; Lynch, V. M.; Sessler, J. L. Nat. Chem.
2015, 7, 1003.
(16) For reviews on dynamic combinatorial chemistry, see: (a) Corbett, P. T.; Leclaire, J.; Vial,
L.; West, K. R.; Wietor, J.-L.; Sanders, J. K. M.; Otto, S. Chem. Rev. 2006, 106, 3652. (b) Lehn,
J.-M. Chem. Soc. Rev. 2007, 36, 151. (c) Li, J.; Nowack, P.; Otto, S. J. Am. Chem. Soc. 2013,
1
35, 9222. (d) Jin, Y.; Yu, C.; Denman, R. J.; Zhang, W. Chem. Soc. Rev. 2013, 42, 6634.
(
17) (a) Otto, S.; Furlan, R. L. E.; Sanders, J. K. M. J. Am. Chem. Soc. 2000, 122, 12063; (b)
Otto, S.; Furlan, R. L. E.; Sanders, J. K. M. Science 2002, 297, 590; (c) Hasenknopf, B.; Lehn, J-
M.; Kneisel, B. O.; Baum, G.; Fenske, D. Angew. Chem. Int. Ed. 1996, 35, 1838; (d)
Hasenknopf, B.; Lehn, J-M.; Boumediene, N.; Dupont-Gervais, A.; Van Dorsselaer, A.; Kneisel,
B. O.; Fenske, D. J. Am. Chem. Soc. 1997, 119, 1095; (e) Huc, I.; Lehn, J.-M. Proc. Natl. Acad.
Sci. USA 1997, 94, 2106; (f) Carnall, J. M. A.; Waudby, C. A.; Belenguer, A. M.; Stuart, M. C.
A.; Peyralans, J. J.-P.; Otto, S. Science 2010, 327, 1502 (g) Sadownick, J. W.; Mattia, E.;
Nowak, P.; Otto, S. Nat. Chem. 2016, 8, 264; (h) Wang, Q.; Zhang, C.; Noll, B. C.; Long, H.; Jin,
Y.; Zhang, W. Angew. Chem. Int. Ed. 2014, 126, 10663; (i) Wang, Q.; Yu, C.; Long, H.; Du, Y.;
Jin, Y.; Zhang, W. Angew. Chem. Int. Ed. 2015, 54, 7550; (j) Brachvogel, R.; Maid, H.; von
Delius, M. Nat. Commun. 2015, 6, 20641; (k) Seoane, A.; Brea, R. J.; Fuertes, A.; Podolsky, K.
A.; Devaraj, N. K. J. Am. Chem. Soc. 2018, DOI: 10.1021/jacs.8b04557.
(
18) For other uses of the acyl hydrazone bond in dynamic combinatorial chemistry, see: (a)
Furlan, R. L. E.; Ng, Y.-F.; Otto, S.; Sanders, J. K. M. J. Am. Chem. Soc. 2001, 123, 8876; (b)
Lam, R. T.; Belenguer, A.; Roberts, S. L.; Naumann, C.; Jarrosson, T.; Otto, S.; Sanders, J. K. M.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 19
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Science 2005, 308, 667; (c) Beeren, S. R.; Pittelkow, M.; Sanders, J. K. M. Chem. Commun.
2
011, 47, 7359; (d) Chung, M.-K.; White, P. S.; Lee, S. J.; Waters, M. L.; Gagné, M. R. J. Am.
Chem. Soc. 2012, 134, 11415; (e) Chung, M.-K.; Lee, S. J.; Waters, M. L.; Gagné, M. R. J. Am.
Chem. Soc. 2012, 134, 11430; (f) Cvrtila, I.; Fanlo-Virgós, H.; Schaeffer, G.; Santiago, G. M.,
Otto, S. J. Am. Chem. Soc. 2017, 56, 14658; (g) Mondal, M.; Radeva, N.; Fanlo-Virfós, H.;
Otto, S.; Klebe, G.; Hirsch, A. K. H. Angew. Chem. Int. Ed. 2016, 55, 9422.
(
19) Handbook of Knot Theory (Eds.: W. Menasco, M. Thistlethwaite) Elsevier Science
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Amsterdam, 2005, 209.
(
(
(
(
20) Wu, G.; Wang, C.-Y.; Jiao, T.; Zhu, H.; Huang, F.; Li, H. J. Am. Chem. Soc. 2018, 140, 5955.
21) Schill, G. Catenanes, Rotaxanes and Knots; Academic Press: 1971.
22) Pons, M.; Millet, O. Prog. Nucl. Magn. Reson. Spectrosc. 2001, 38, 267.
23) For examples of anion recognition in dynamic combinatorial systems, see: (a) Otto, S.;
Kubik, S. J. Am. Chem. Soc. 2003, 125, 7804; (b) Beeren, S. R.; Sanders, J. K. M. Chem. Sci.
2011, 2, 1560; (c) Beeren, S. R.; Sanders, J. K. M. J. Am. Chem. Soc. 2011, 133, 3804; (d)
Katayev, E. A.; Pantoş, G. D.; Reshetova, M. D.; Khrustalev, V. N.; Lynch, V. M.; Ustynyuk, Y.
A.; Sessler, J. L. Angew. Chem. Int. Ed. 2005, 44, 7386; (e) Riddell, I. A.; Smulders, M. M. J.;
Clegg, J. K.; Hristova, Y. R.; Breiner, B.; Thoburn, J. D.; Nitschke, J. R. Nat. Chem. 2012, 4, 751.
(
8
24) (a) Busschaert, N.; Caltagirone, C.; Rossom W. V.; Gale, P. A. Chem. Rev. 2015, 115,
038; (b) Kubik, S. Chem. Soc. Rev. 2010, 39, 3648-3663; (c) Kim, Y.; Gabbaï, F. P. J. Am.
Chem. Soc. 2009, 131, 3363.
(25) When the library was prepared in DMSO from 10 mM of A·2Br and 10 mM of H3 (Figure
S33), we could identify by LC-MS another [3+3] macrocycle with a different retention time
(
(
20 min). This second [3+3] macrocycle is likely to be the topologically trivial macrocycle.
26) Due to the lack of feature of the unbound form of 3, it was not possible to determine
the binding constants by NMR. Similarly, the UV-Vis spectrum of 3 did not change upon
addition of halides, and the binding constants could not be determined by UV titrations.
(
27) Bilbeisi, R. A.; Prakasam, T.; Lusi, M.; El-Khoury, R.; Platas-Iglesias, C.; Charbonnière, L. J.;
Olsen, J.-C.; Elhabiri, M.; Trabolsi, A. Chem. Sci. 2016, 7, 2524.
(28) Ayme, J.-F.; Beves, J. E.; Campbell, C. J.; Gil-Ramírez, G.; Leigh, D. A.; Stephens, A. J. J.
Am. Chem. Soc. 2015, 137, 9812.
(
29) Other types of interlocked structures can bind anions: (a) Langton, M. J.; Robinson, S.
W.; Marques, I.; Félix, V.; Beer, P. D. Nat. Chem. 2014, 6, 1039; (b) Caballero, A.; Zapata, F.;
White, N. G.; Costa, P. J.; Félix, V.; Beer, P. D. Angew. Chem. Int. Ed. 2012, 51, 1876.
ACS Paragon Plus Environment

Page 19 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table of content
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment