Constitutional Dynamic Selection at Low Reynolds Number in a Triple Dynamic System: Covalent Dynamic Adaptation Driven by Double Supramolecular Self-Assembly

Article abstract of DOI:10.1021/jacs.1c04446

A triple dynamic complex system has been designed, implementing a dynamic covalent process coupled to two supramolecular self-assembly steps. To this end, two dynamic covalent libraries (DCLs), DCL-1 and DCL-2, have been established on the basis of dynamic covalent C-C/C-N organo-metathesis between two Knoevenagel derivatives and two imines. Each DCL contains a barbituric acid-based Knoevenagel constituent that may undergo a sequential double self-organization process involving first the formation of hydrogen-bonded hexameric supramolecular macrocycles that subsequently undergo stacking to generate a supramolecular polymer SP yielding a viscous gel state. Both DCLs display selective self-organization-driven amplification of the constituent that leads to the SP. Dissociation of the SP on heating causes reversible randomization of the constituent distributions of the DCLs as a function of temperature. Furthermore, diverse distribution patterns of DCL-2 were induced by modulation of temperature and solvent composition. The present dynamic systems display remarkable self-organization-driven constitutional adaption and tunable composition by coupling between dynamic covalent component selection and two-stage supramolecular organization. In more general terms, they reveal dynamic adaptation by component selection in low Reynolds number conditions of living systems where frictional effects dominate inertial behavior.

Full text of DOI:10.1021/jacs.1c04446

pubs.acs.org/JACS
Article
Constitutional Dynamic Selection at Low Reynolds Number in a
Triple Dynamic System: Covalent Dynamic Adaptation Driven by
Double Supramolecular Self-Assembly
Ruirui Gu and Jean-Marie Lehn*
Cite This: https://doi.org/10.1021/jacs.1c04446
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: A triple dynamic complex system has been
designed, implementing a dynamic covalent process coupled to
two supramolecular self-assembly steps. To this end, two dynamic
covalent libraries (DCLs), DCL-1 and DCL-2, have been
established on the basis of dynamic covalent CC/CN
organo-metathesis between two Knoevenagel derivatives and two
imines. Each DCL contains a barbituric acid-based Knoevenagel
constituent that may undergo a sequential double self-organization
process involving first the formation of hydrogen-bonded
hexameric supramolecular macrocycles that subsequently undergo
stacking to generate a supramolecular polymer SP yielding a
viscous gel state. Both DCLs display selective self-organization-
driven amplification of the constituent that leads to the SP.
Dissociation of the SP on heating causes reversible randomization of the constituent distributions of the DCLs as a function of
temperature. Furthermore, diverse distribution patterns of DCL-2 were induced by modulation of temperature and solvent
composition. The present dynamic systems display remarkable self-organization-driven constitutional adaption and tunable
composition by coupling between dynamic covalent component selection and two-stage supramolecular organization. In more
general terms, they reveal dynamic adaptation by component selection in low Reynolds number conditions of living systems where
frictional effects dominate inertial behavior.
revealing the organizational power of a soft matter, complex
fluid medium5 as characterized by bulk properties, viscosity
and therefore low Reynolds number conditions. Constitutional
dynamic chemistry (CDC)6 offers such perspectives. It is for
these relationships that we wish to dedicate the present study
to the memory of Edward Purcell.1
INTRODUCTION
■
In the famous paper “Life at Low Reynolds Number”, Edward
Purcell pointed out that living systems operate in conditions
where frictional effects strongly dominate inertial features.1
Purcell had a major role in the development of NMR,
particularly in its nuclear relaxation aspect,2 which bears direct
dependence on molecular motions and thus on frictional
effects. These relationships establish a connection between
nuclear relaxation and supramolecular organization as em-
bodied in medium viscosity. As a consequence, chemical
systems might be able to change/adapt their constitution
under the driving force of an increase in organization.
Considering that frictional effects result from intermolecular
interactions and determine bulk properties such as viscosity, an
increase in viscosity corresponds to higher supramolecular
(passive) self-organization.3 It would thus be of great interest
to devise chemical systems whose behavior would reveal the
driving force of self-organization4 by inducing the generation
of just the very chemical entity that leads to higher
organization in a sort of organizational autocatalytic process.
Such would be the case if the increase in organization
occurring on transition from a solution to a gel state were to
enforce the amplification of the gel forming entity, thus
RATIONALE
■
CDC rests on the generation of dynamic libraries of molecular
(covalent) or supramolecular (noncovalent) constituents by
recombination of their components linked, respectively, by
reversible chemical bonds or intermolecular interactions or by
both simultaneously. In response to the application of physical
stimuli or chemical effectors, such complex dynamic systems
may undergo constitutional changes through selection of the
Received: May 4, 2021
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
© XXXX American Chemical Society
A

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
appropriate components and display adaptation to the acting
agent.
response to the application of diverse external (environmental)
agents. Hydrogen bonding, electrostatic, and van der Waals
interactions play a major role in the buildup of biological
architectures, for instance, of the double helix structure of
nucleic acids or the assembling of polypeptide chains in
proteins. To implement such elaborate processes, to expand on
them, and to create functional materials with diverse chemical
or physical properties, numerous chemical structures of various
types have been designed and synthesized, involving in
particular multiple hydrogen-bonding motifs.13−15 Among the
latter, derivatives of barbituric acid or analogues have been
demonstrated to be able to form sextuple hydrogen-bonding
patterns with a complementary receptor moiety and undergo
supramolecular polymerization,13h−j as well as form hydrogen-
bonded macrocyclic hexamers (rosettes) that may assemble
into supramolecular polymeric stacks.13d−g,15
The implementation of CDC at the molecular level, that is,
dynamic covalent chemistry (DCC), involves reversible
reactions and generates dynamic covalent libraries (DCLs) in
which the relationships between constituents are described by
an underlying constitutional dynamic network (CDN). Such
reactions include amine/carbonyl condensations,6 disulfide
exchange,6b,c,7 peptide exchange,8 olefin metathesis,9 and
Diels−Alder condensation.10 Furthermore, the CC/CN
organo-metathesis taking place between Knoevenagel (Kn)
compounds derived from 1,3-dimethylbarbituric acid and
imines (aldimines or Schiff bases) in low polar organic
solvents was found to be an efficient way to establish DCLs
(Scheme 1a).11 One may note that the well-balanced
Scheme 1. (a) Component Exchange between Knoevenagel
Barbituric Derivatives to Undergo Component Exchange
with Imines (or Schiff Bases) by CC/CN Organo-
metathesis; (b) Hexameric Hydrogen-Bonded
In this Article, we describe constitutional dynamic systems
that involve a triple dynamic process, a dynamic covalent
component exchange based on CC/CN organo-meta-
thesis followed by two sequential supramolecular assembly
steps (formation of rosettes and their self-assembly into
polymeric stacks). We demonstrate the direct correlation
between the degree of molecular organization (as displayed by
Supramolecular Macrocycle (“Rosette”) Formation (Left to
Center) Followed by a Second Self-Assembly to Generate a
Supramolecular Polymer SP by Stacking of the Rosettes
(Center to Right) in a Double Supramolecular Process;
(Right) Simplified “Artist” Representation of SP; and (c)
Steric Hindrance to Stacking Polymerization of the
1
viscosity and measured by corresponding H NMR signal line
width) and amplification by component selection of the
constituent that yields higher order in a self-organization
driven mode.16 It represents a direct demonstration of the
spontaneous trend of “passive” self-organization to drive
selection and thus, in Purcell’s terms, low Reynolds number
driving chemical selection, both being characteristic features of
living systems.
a
“Rosette” Assembly
We have shown earlier in systems involving supramolecular
organization linked to a dynamic covalent process that indeed
the possibility of the formation of a gel causes the selection of
just those components that generate the very constitution that
yields the gel.6a,16a,d Thus, a high friction/low Reynolds
number marks an increase in medium organization, which
drives component selection and constituent adaptation.
Here, we implement two four-membered DCLs that take
advantage of the dynamic covalent organo-metathesis reactivity
of Kn compounds with imines to generate library constituents
capable of undergoing hierarchical two-level supramolecular
self-assembly, the initial formation of discrete hydrogen-
bonded hexameric rosettes and their subsequent stacking to
give a cylindrical supramolecular polymer (Scheme 1b). Driven
by supramolecular self-organization under temperature mod-
ulation, the distributions of the DCLs form a constitutional
dynamic network that displays variation from a four-member-
statistically averaged state to constitutionally selected states
(Scheme 2b). Such selection could be reversibly switched by
heating/cooling of the reaction mixture. The distributions of
the libraries were investigated by 1H NMR and UV−vis
spectroscopy, while the supramolecular polymeric entities were
characterized by DLS, AFM, and cryo-TEM.
a
Kn1, Kn2, and Kn3 form all three of the “rosettes”, but subsequent
polymerization to SP is hindered for Kn1 by the bulky adamantane
groups in the R1 side chains (see Scheme 2a).
Knoevenagel/imine reversible exchange also represents a very
sensitive touchstone for the quantitative evaluation of the
forces acting on the equilibrating system and shifting the
equilibrium position.
The lability of noncovalent interactions allows for the
establishment of constitutional dynamic systems at the
supramolecular level. Such interactions may drive the
formation of more or less highly organized supramolecular
entities by recognition between their complementary molec-
ular components.12−15 The resulting self-organization ability
(passive, equilibrium) leads to the generation of complex
architectures with component selection and adaptation in
RESULTS AND DISCUSSION
■
Two DCLs were investigated: DCL-1 consisting of Kn1, A1,
Kn2, and A2, and DCL-2 consisting of Kn1, A1, Kn3, and A3
(Scheme 2a). DCL-1′ and DCL-2′, involving N-methylated
barbituric groups, which cannot undergo association by H-
bonding, were studied for comparison with the libraries DCL-1
and DCL-2, respectively. Detailed synthetic procedures and
B
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Compound Kn1, present in both DCLs, was chosen as a 3,4-
disubstituted structure to ensure better solubility in nonpolar
solvents. It was also equipped with two adamantane groups
expecting that their steric bulk would hinder the formation of
supramolecular polymeric assemblies by stacking of Kn1-based
hexameric rosettes (Scheme 1c) and thus reduce their
thermodynamic stability, so as to create a difference in free
energy between reactants and exchange products. The four-
member DCLs were obtained by mixing two diagonally related
agonistic constituents or the orthogonal ones in the chosen
solvent. Thus, for DCL-1, an equimolar mixture of Kn1+A2
gave the same distribution of constituents as Kn2+A1 after
equilibration. One may expect that at a temperature high
enough to dissociate the hydrogen-bonded aggregates, the four
constituents would be present as monomers with a statistical
distribution of about 25% each. Upon cooling of the reaction
mixture, the monomeric Knoevenagel derivatives Kn1 and
Kn2 (or Kn3) would start to assemble and coassemble into
hexameric rosette supramolecular macrocycles, followed by
self-organization and self-sorting of the coaggregated rosettes
into Kn2 (or Kn3)-dominated supramolecular polymers as a
consequence of the lower thermodynamic stability of Kn1
assemblies. Meanwhile, the difference in free energy would be
expected to drive the CC/CN metathesis reaction to
favor the formation of Kn2 (or Kn3) and its agonist A1,
resulting in a constitutionally selected mixture. The process is
expected to be reversible with randomization on temperature
increase.
Scheme 2. (a) The Two Dynamic Combinatorial Libraries
DCL-1 and DCL-2 Composed Each of Four Constituents;
and (b) (Left) Component Exchange between the Agonistic
Constituents Kn1/Ax or Knx/A1 by Organo-metathesis
(Scheme 1a), Generating the Same Equilibrated [2 × 2]
Square Constitutional Dynamic Network (CDN) with
(Center) No Constitutional Selection; (Right) The CDN
Undergoes Reversible Amplification of Knx and of Its
Agonist A1 as a Function of Temperature, Driven by Self-
Organization-Induced Constitutional Selection on
a
Formation of SP
Medium. Chloroform/cyclohexane (CHCl3/CH) or
1,1,2,2-tetrachloroethane/methylcyclohexane (C2H2Cl4/
MCH; for studies at higher temperatures) mixtures were
chosen as the reaction medium for the present DCLs. The
Knoevenagel/imine CC/CN exchange reaction is known
to be fast in chloroform.11b Regarding the effect of solvent on
H-bonding-based self-assembly, one expects that the addition
of a low polarity solvent increases the degree of aggregation
and strengthens the thermodynamic stability of the supra-
molecular polymers.14f The ratio of the two solvents was tuned
according to the specific situation as it is an important
parameter of the DCLs affecting not only the equilibration rate
but also the constitutional selectivity.
a
For Ax and Knx, x = 2 or 3.
characterization data of all of the compounds are presented in
the Supporting Information.
Design of the Constituents and Generation of the
DCLs. It has been reported that Knoevenagel condensation
products of barbituric acid and aromatic aldehyde components
form six-member rosettes connected by double hydrogen
bonding and then undergo polymerization via stacking of the
rosette layers in low-polarity solvents (for example, hexane and
methylcyclohexane).15 It has also been shown that Knoevena-
gel derivatives perform dynamic covalent component exchange
with imines catalyzed by L-proline.11a However, due to their
low solubility, these exchange processes had to be run in
dimethyl sulfoxide. For the present purposes, the Knoevenagel
compounds were thus equipped with aliphatic chains to reach
sufficient solubility in solvents of low polarity. Indeed, it has
been confirmed in this work that they do undergo the CC/
CN exchange in chloroform without a catalyst (Scheme 1a).
On the basis of these studies, compounds Kn1 and Kn2
were used to generate the four constituents of DCL-1 by
reaction with their imine agonists A2 and A1, respectively
(Scheme 2b, left). Another library DCL-2 was designed to
induce better constitutional selectivity. It was envisaged that
the extra −C3H6CONH− fragment present in the side chains
of compound Kn3 would lead to the formation of
intermolecular hydrogen bonds between the amide units and
thus generate supramolecular architectures of higher thermo-
dynamic stability, as has been reported.14d,e Both DCL-1 and
DCL-2 form a square CDN linking their four constituents.
Dynamic Covalent Library-1 (DCL-1). A solution of 2
mM Kn1 + A2 in 20% CDCl3 + 80% CH-d12 was allowed to
1
equilibrate at 343 K and monitored by H NMR. After the
equilibrium was reached, all four constituents could be
observed in the spectrum, the distribution being essentially
statistical with about 25% of each constituent Kn1, A2, Kn2,
1
and A1 (A1:A2 = 1:1.04; see Figure S1a for the H NMR
spectrum). As the mixture was gradually cooled to 293 K at a
rate 0.5 K/min, the 1H NMR peaks of the two Kn compounds
became broad and unobservable, which indicated the
formation of supramolecular self-assemblies. The equilibrium
state at 293 K was reached in about 33 h. The ratio of A1:A2
became 1:0.91 (Figure S1b). As compared to the distribution
at 343 K, the equilibrium of the reaction had undergone a
small shift. It should be noted that, in this case and most of the
other cases presented below, the NMR signals of the Kn
compounds (Kn1, Kn2, and Kn3) were rarely observable
because of the supramolecular assembly and coassembly.
Under these circumstances, the distributions of the constitu-
ents of the libraries can be inferred from the ratio of the two
imines A1 and A2, which do not participate in the
supramolecular aggregates.
C
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 1. 400 MHz 1H NMR spectra of the equilibrated solutions of 2 mM DCL-1 in 10% CDCl3 + 90% CH-d12 (a) at 343 K and (b) at 293 K; 2
mM DCL-1′ in 10% CDCl3 + 90% CH-d12 (c) at 343 K and (d) at 293 K; 2 mM DCL-2 in 30% CDCl3 + 70% CH-d12 (e) at 343 K and (f) at 293
K; and 2 mM DCL-2′ in 10% CDCl3 + 90% CH-d12 (g) at 343 K and (h) at 293 K. SP: supramolecular polymer. Benzyl −CH2− signals of the two
imines in each DCL are highlighted in the dashed gray boxes, and their ratio indicates the “ON” or “OFF” state of the DCLs. The error in signal
integration amounts to about 5%.
DCL-1 in 10% CDCl3 + 90% CH-d12 was submitted to the
same procedure, the less polar solvent being expected to
strengthen the driving force of self-assembly. At 343 K, the 1H
NMR peaks of both Kn1 and Kn2 were broad and difficult to
integrate. In the equilibrated solution, the A1:A2 ratio was
1:1.03 (Figure 1a). Cooling to 293 K caused a progressive
amplification of A1 and a reduction of A2. After 106 h
(indicating a much slower rate due to the lower solvent
polarity), the equilibrium was reached, and the ratio of A1:A2
became 1:0.6 (Figure 1b). Meanwhile, 22% of aldehyde
hydrolysis products were detected as indicated by the CHO 1H
NMR proton signal. Comparing these two cases, the change in
reaction medium from 20% CDCl3 to 10% affected the DCL
by slowing the reaction rate and inducing a larger shift in the
equilibrium toward the formation of Kn2 and A1.
After the equilibration of a 2 mM Kn1′ + A2 solution at 343 K,
a mixture of 24% Kn1′, 23% A2, 27% Kn2′, and 26% A1 was
obtained (Figure 1c). On cooling to 293 K, all four
constituents remained clearly observable, while the thermody-
namic equilibrium reached 12 h later indicated only a slight
variation in distribution (23% Kn1′, 22% A2, 28% Kn2′, and
27% A1, with less than 5% hydrolysis; see the spectrum in
Figure 1d). In this DCL, the introduction of two methyl
groups on the barbituric rings of the Kn compounds prevented
them from self-assembling, resulting in the disruption of the
constitutional selectivity in this library.
Dynamic Covalent Library-2 (DCL-2). The two reactants
Kn1 and A3 (2 mM each) were mixed in 10% CDCl3 + 90%
CH-d12 and allowed to equilibrate at 343 K to give the four-
member mixture, DCL-2. The 1H NMR spectrum at
equilibrium gave A1:A3 as 1:0.15, while Kn1 and Kn3 peaks
were not observable (Figure S2a). In contrast to DCL-1, the
exchange products Kn3 and A1 were already selected as the
main species. The equilibrium mixture obtained 20 h after
mixing was cooled to 293 K to give A1:A3 as 1:0.13 (Figure
DCL-1′, containing the bis-N-methylated Knoevenagel
compounds Kn1′ and Kn2′ in place of Kn1 and Kn2 (see
Figure 1c), was investigated in 10% CDCl3 + 90% CH-d12 as a
case comparative to DCL-1. The structures of the four
constituents Kn1′, A2, Kn2′, and A1 are shown in Figure 1c.
D
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
S2b). The considerable thermodynamic stability of Kn3
indicated by these distributions at the two equilibrium states
is much stronger than that of Kn2 and may be attributed to the
extra intermolecular hydrogen bonding between the amide
units.
10%:90%), taking about 3 h at 343 K and 25 h on cooling
to 293 K with less than 5% hydrolysis. When their NMR
signals could not be integrated, the amounts of Kn1 and Kn3
were, respectively, calculated from those of their agonists A3
and A1. There was almost no variation of composition when
the solvent polarity was either too high (100% CDCl3) or too
low (10% CDCl3 + 90% CH-d12), indicating that no
aggregation was occurring at high CDCl3 fraction and that
heating to 343 K was not sufficient to destroy the assemblies at
the lowest CDCl3 fraction. Only small shifts of the equilibrium
were found below a 50%:50% solvent composition, which thus
corresponds to the composition for which heating to 343 K is
sufficient for melting of the supramolecular assemblies. In line
with the titration results, above a composition of 30% CDCl3 +
70% CH-d12, the selectivity was highest and did not change
anymore on increasing cyclohexane content (see Figure 1e,f for
However, to optimize the aggregation-induced variation of
distribution, the effect of the composition of the reaction
medium (the ratio of CDCl3 and CH-d12) on the behavior of
Kn3 was explored by titration of CDCl3 into a CH-d12 solution
1
of Kn3 at 293 K (see Figure S3 for the H NMR spectra). At
the beginning (15% CDCl3 + 85% CH-d12), only aliphatic
1
chains in the low ppm region of the H NMR spectrum could
be detected. On increasing the fraction of CDCl3, the signals at
high ppm values started to show up (30% CDCl3 + 70% CH-
d12). After the CDCl3 proportion reached 50%, all of the peaks
became narrow and clear. During the titration process, the Kn3
solution always remained transparent. This titration experi-
ment indicated, as expected, the negative influence of CDCl3
on the self-assembly of compound Kn3 and suggested that a
composition of 30% CDCl3 + 70% CH-d12 could be a suitable
medium for DCL-2.
To confirm the effect of solvent polarity on the dynamic
library, a series of DCL-2 reactions were conducted in media
of five different CDCl3:CH-d12 compositions. The correspond-
ing distributions obtained at 343 and 293 K are illustrated in
Figure 2. The equilibration was slowest for the composition
with the lowest fraction of chloroform (CDCl3:CH-d12
1
the H NMR spectra). Taken together, the largest variation in
distribution at equilibrium was from 24% Kn1, 24% A3, 26%
Kn3, 26% A1 to 6% Kn1, 6% A3, 44% Kn3, and 44% A1.
Thus, both the constituent Kn3 and consequently its agonistic
constituent A1 were subject to strong amplification driven by
the sequential double self-organization first of Kn3 into
hydrogen-bonded hexameric rosettes, which set the stage for
the subsequent assembly by stacking to generate the
supramolecular polymer SP. The corresponding driving forces
for DCL-1 and DCL-2 are calculated to be, respectively, 0.63
and 2.21 kcal/mol (see Table S1 for the free energies of all of
the present DCLs, calculated following eq S1). The difference
of about 1.6 kcal/mol between these two values indicates a
markedly stronger stabilization of the SP aggregates, which
may be attributed to the significantly longer spacer and the
additional interactions introduced by the amide group in the
side-chain of Kn3; it thus may well be considered as a third
organizational driving force.
To better characterize the composition of the SP formed in
1
DCL-2, the H NMR data were analyzed in view of obtaining
information about the amount of coassembly that may occur
between Kn1 and Kn3 in the actual DCL-2. Considering the
NMR spectrum shown in Figure 1f, as expected the signals of
Kn3 could not be observed because it was entirely contained in
the supramolecular gel assembly. However, it was still possible
to obtain the ratio Kn1:Kn3 as 1:7.33 by comparing the
integrals of the N−CH2− peaks of their agonists A3 and A1.
Moreover, the −OCH2− peak of Kn1 at 4.21 ppm remained
observable, indicating that a certain amount of Kn1 molecules
were present as free molecules instead of participating in the
coassembly with Kn3. From the amount of this free Kn1 (0.7
with respect to A3), the ratio of free Kn1 to total Kn1 was
found to be 0.7:1, giving a coassembly ratio Kn1:Kn3 of
0.3:7.33. Thus, the SP formed in DCL-2 consisted of a
supramolecular coassembly composed of 96% Kn1 and 4%
Kn3. The ability of Kn1 and Kn3 themselves to coassemble
was further investigated by mixing stoichiometric amounts of
Kn1 and Kn3 in 30% CDCl3 + 70% CH-d12. As expected, the
1
400 MHz H NMR spectrum (Figure S4a) did not show any
Figure 2. Representation of the variation in constitutional selectivity
of DCL-2 between 343 and 293 K as a function of solvent
composition/polarity. The change from the “ON” state to “OFF”
state represents the variation in constituent distribution driven by
formation of the self-organized supramolecular aggregates. The time
to reach equilibrium depends on the solvent composition and
increases from bottom (30 min at 343 K and 90 min at 293 K) to top
(3 h at 343 K and 25 h at 293 K). The error in % determination
amounts to about 5%.
Kn3 signals, and only broad peaks of Kn1 were observable,
yielding an amount of free Kn1 of 0.39 with respect to an
internal standard (mesitylene). When the solvent composition
of the sample was then tuned to 70% CDCl3 + 30% CH-d12 by
adding CDCl3 (Figure S4b), all of the signals of Kn1 and Kn3
showed up as regular peaks in equal amounts of 0.59 with
respect to the internal standard, yielding a composition of the
coassembly of 25% Kn1 + 75% Kn3 (in Figure S4a).
E
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
For comparison, the library DCL-2′ (also in 30% CDCl3 +
70% CH-d12), which cannot lead to hydrogen bonding
between the barbituric groups, was also examined. When a
mixture of 2 mM Kn1′ + A3 was equilibrated at 70 °C, a
statistical distribution of 25% was obtained for all four
constituents (Figure 1g). After being cooled to 20 °C, the
distribution underwent only a very small variation (22% Kn1′,
22% A3, 28% Kn3′, and 28% A1, Figure 1h). These results
confirm the crucial role of the hydrogen-bonding-based self-
assembly of the barbituric units into the hexameric rosettes.
The present DCL-2 displays a larger distribution bias (6%,
6%, 44%, 44%) than the previously established two CC/
CN DCLs: the DCL driven by addition of metal ions where
a distribution of 8%, 9%, 42%, 40% was obtained as the
chemical effector selected state,16f and the DCL driven by
gelation of a specific constituent where a distribution of 10%,
9%, 40%, 40% was obtained as the physical stimulus selected
state.16d
(1 mM) are shown in Figure 4a. DCL-2 was obtained by
mixing Kn1 and A3 (1 mM each) in 30% CHCl3 + 70% CH.
To investigate the reversibility of the dynamic system, DCL-
2 in 30% CDCl3 + 70% CH-d12 was cyclically modulated
between 343 and 293 K, and the evolution of the constituent
distributions was monitored by 1H NMR. The resulting
1
changes in the ratio of (A1)/(A3) (obtained by H NMR
signal integration) as a function of time are shown in Figure 3
Figure 4. (a) UV−vis spectra of the four constituents of DCL-2,
measured separately in 30% CHCl3 + 70% CH (c = 1 mM, T = 293
K). UV−vis spectra of DCL-2 starting from Kn1+A1 (1 mM each):
(b) spectra at 343 K, measured every 15 min; (c) spectra after cooling
(0.5 K/min) to 293 K, measured every 1 h. Corresponding kinetic
plots of absorptions at 271, 323, 410, and 450 nm of DCL-2: (d) at
343 K; (e) after cooling to 293 K. Before the UV−vis measurements,
all of the samples were placed in 1 mm quartz cuvettes, used for the
concentrated solutions, and sealed by parafilm.
Figure 3. Reversible switching of the distribution of the constituents
of DCL-2 and adaptation as a function of temperature modulation in
30% CDCl3 + 70% CH-d12. (A1)/(A3) (left Y axis, black curve) and
amount of hydrolysis products (%) versus time (right Y axis, red
curve). Temperature and time scale of each pattern are annotated.
(A1) and (A3) represent the amount of compound as obtained by
integration of their CHN and CH2 signals in the 1H NMR
spectrum of the DCL. The error in % determination amounts to about
5%.
The change in the absorption spectra of the reaction at 343 K
is shown in Figure 4b. The absorptions at 323 and 410 nm,
respectively, representing A3 and Kn1 decreased as the
absorption at 271 and 450 nm, respectively, due to A1 and
Kn3 increased (see Figure 4d for the changes observed at the
four wavelengths versus time). From 65 to 110 min, no change
was observed, which means the thermodynamic equilibrium
was reached. After being cooled to 293 K (0.5 K/min), the
DCL displayed continuous amplification of Kn3 and A1 in the
first 12 h, as indicated by Figure 4c and e.
and indicate a good reversibility of the system. When the
reaction solution was heated to 343 K, the equilibrium always
displayed the selection “OFF” state ((A1)/(A3) ≈ 1) after 1 h,
while after the solution was cooled to 293 K, the system
underwent a slow (in about 20 h) variation in constituent
distribution to the selection “ON” state (Scheme 2, center and
right, Figure 2). The (A1)/(A3) selectivity in the “ON” states
was gradually reduced from 7.4 (first cycle) to 5 (third cycle),
and even to 3.4 (fourth cycle, after 100 h), a loss in reversibility
that may be a consequence of the increasing hydrolysis of the
constituents as indicated by the red curve in Figure 3 (see also
Figure S5).
Structural Investigations of the Supramolecular
Polymer SP by DLS, AFM, and Cryo-TEM. Dynamic light
scattering (DLS) measurements of equilibrated DCL-2 in 30%
CHCl3 + 70% CH at 293 K showed entities of an average
hydrodynamic diameter (DH) of 31.52 nm with a poly-
dispersity index (PDI) of 0.436 (see black curve in Figure 5a
for the DLS of DCL-2). The single constituent solutions of
Kn1 and Kn3 were also measured in the same conditions. The
average DH of Kn3 (see red curve in Figure 5a) was 49.54 nm
with a PDI of 0.240. The smaller object sizes and broader
diameter region of DCL-2, as compared to Kn3 alone, can be
attributed to an adverse effect on the formation of supra-
molecular self-assemblies due to the participation/coassembly
of the hindering bulky constituent Kn1. On the other hand, the
correlogram (correlation coefficient versus time; see Figure S8
Electronic Spectroscopy Investigations of DCL-2. The
dynamic behavior of DCL-2 was also monitored by UV−vis
spectrometry. The absorption spectra of the four constituents
F
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
for the correlograms of all of the present samples) for Kn1 did
not show a well-defined curve, pointing to the absence of
supramolecular polymers. These DLS data not only supported
the suggestion that Kn1 itself does not form nanoscale
supramolecular self-assemblies, but also indicated that the
formation of supramolecular copolymers in DCL-2 was driven
by the self-assembly of Kn3.
The morphologies of the supramolecular self-assemblies
were investigated by atomic force microscopy (AFM; Figure
5c,d) and cryogenic transmission electron microscopy (cryo-
TEM; Figure 5g,h). The 1 mM 30% CHCl3 + 70% CH
solutions of Kn3 and DCL-2, equilibrated at 293 K, were
separately spin-coated onto highly oriented pyrolytic graphite
(HOPG) for AFM imaging and drop-casted onto graphene
oxide grids for cryo-TEM studies (see details in the Supporting
Information).
The AFM images of Kn3 (Figure 5c) showed bundled,
parallel nanofibers with an average width of about 7 nm, in line
with the diameter of 7.8 nm indicated by a computed
geometry-optimized rosette structure of Kn3 (Figure 5b; see
the Supporting Information for details). In the same way, the
cryo-TEM image of Kn3 displayed overlapping nanosheets of
crossed bundles of nanofibers (width 6−7 nm) in parallel
arrangement (Figure 5e). Both AFM and cryo-TEM images of
Kn3 demonstrated the formation of supramolecular nano-
fibers, corresponding to our expectation of double supra-
molecular self-assembly as represented in Scheme 1b. For
DCL-2, its AFM height image (Figure S9a) was blurry,
probably caused by interference of nonassembled molecules of
A1 and A2. In comparison, its corresponding phase image
(Figure 5d) clearly showed a mixture of disordered nanofibers
with a wide range of lengths (5−340 nm, also in line with the
DLS curve of DCL-2 in Figure 5a), together with amorphous
aggregates, probably due to the inclusion of molecules of the
adamantane-bearing Kn1 by some amount of copolymerization
and resulting consequently in a perturbation in the stacking
structure of the global supramolecular assemblies. The cryo-
TEM image of DCL-2 (Figure 5f) displayed bundled
nanofibers (width 6−8 nm) similar to that of Kn3 (Figure
5e), confirming that the constituent Kn3 was selected as the
dominating chemical species among the library. On the other
hand, the cryo-TEM also recorded DCL-2 (see Figure S9b) as
less well-defined crossed nanofibers of larger complexity as one
might expect in view of the possible incorporation of Kn1 and
the presence of the two other constituents A1 and A2 in the
medium.
When the solution of DCL-2 in 30% C2H2Cl4 + 70% MCH
was equilibrated at 363 K before being spin-coated onto
HOPG, the corresponding AFM image (Figure 5g) shows only
amorphous nanoparticles due to the full randomization of the
constituents at high temperature. In comparison, when the
same solution was equilibrated at 303 K before spin coating,
the image obtained (Figure 5h) shows irregularly arranged
nanofibers assembled and amorphous aggregates indicating
both the major role of Kn3 and the presence of the other three
constituents.
Figure 5. (a) DLS profiles of DCL-2 and of Kn3, both in 30% CHCl3
+ 70% CH (c = 2 mM, T = 293 K). (b) Optimized computed
structure for the rosette of Kn3, obtained using a semiempirical
quantum mechanical GFN2-xTB method.17 AFM images of (c) 1 mM
Kn3 in 30% CHCl3 + 70% CH at 293 K (left) with (right)
enlargement of the dashed box and cross-sectional analysis of the
white solid line (inset); (d) 1 mM DCL-2 in 30% CHCl3 + 70% CH
at 293 K; and 1 mM DCL-2 in 30% C2H2Cl4 + 70% MCH at (g) 363
K and (h) 303 K. Samples were spin-coated from equilibrated
solutions onto HOPG. The AFM samples were dried under high
vacuum for 1 week to ensure that no residual solvent remained on the
plate. (e,f) Cryo-TEM micrographs (Titan-Krios) of (e) 0.1 mM Kn3
at 293 K showing superimposed sheets of bundles of nanofibers of 6−
7 nm diameter and (f) 0.1 mM DCL-2 at 293 K showing a
morphology similar to that in (e). Samples were drop-casted from
30% CHCl3 + 70% CH equilibrated solutions onto graphene oxide
grids.
Constituent Amplification and Component Selection
at Low Reynolds Number. To establish the correlation
between, on one hand, the level of medium organization
reflected in frictional effects and bulk medium viscosity and, on
the other hand, constituent amplification (component
1
selection), H NMR line widths (at half height LW1/2) were
determined as line broadening results from motional restriction
G
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
in viscous media.2 Thus, correlating signal line width with
component amplification would establish selection at low
Reynolds number, as an increase in viscosity corresponds to
increasing supramolecular (passive) self-organization.
It is seen in Figure 1e that, when 30% CDCl3 + 70% CH-d12
is used as the solvent of DCL-2, the supramolecular assemblies
are not fully melted even at 343 K, as indicated by the broad
Kn3 peaks. Thus, to reach the gel to sol transition at a higher
temperature, a mixed solvent of C2D2Cl4 (deuterated 1,1,2,2-
tetrachloroethane) and MCH-d14 (deuterated methylcyclohex-
ane) was utilized. The double self-assembly behavior of the Kn
compounds and the DCLs was then investigated under fine
temperature control.
First, the solutions of 4 mM Kn1 and Kn3 in 30% C2D2Cl4
+ 70% MCH-d14 were, respectively, heated to 363 K in the
NMR probe and then stepwise cooled to 303 K with
stabilization at 11 intermediate temperatures (15 min each)
1
for the recording of the 400 MHz H NMR spectra (Figure
6a). As is seen (red spectra in Figure 6a), the Ha proton NMR
signal of Kn1 underwent a large shift from 7.9 to 9.1 ppm,
indicating the gradual formation of the hexameric supra-
molecular rosettes by hydrogen bonding. There was almost no
line-broadening during the cooling process, as indicated by the
LW1/2 of the Hb signal as a function of temperature shown in
Figure 6c (LW1/2 values obtained by the MNova software
suite; see Figure S6 for details). Such a behavior is in line with
our expectation that stacking of the rosettes of Kn1 to form
larger supramolecular assemblies would be sterically hindered.
In comparison, the green spectra of Kn3 (Figure 6a) show not
only a large shift of the H1 signal but also the broadening of all
of the peaks. The LW1/2 of H2 (Figure 6c) underwent a sharp
increase when the temperature was below 348 K, suggesting a
critical temperature of supramolecular aggregation at about
348 K.
Figure 6. Partial 400 MHz 1H NMR spectra of (a) single constituent
solution of Kn1 (left), Kn3 (right) and (b) DCL-2 under fine
temperature control from 363 to 303 K (concentration, 4 mM;
solvent, 30% C2D2Cl4 + 70% MCH-d14). (c) LW1/2 (line width at half
height in Hz) of Hb (Kn1) and H2 (Kn3) in the single constituent Kn
solution as a function of temperature data obtained from the spectra
in (a). (d) LW1/2 of Hb and H2 in DCL-2 as a function of temperature
(left Y axis, data obtained from the spectra in (b)), (A1)/(A3) ratio
versus temperature (black curve, right Y axis), and (e) (A1)/(A3)
ratio as a function of LW1/2. All of the LW1/2 data were obtained using
the MNova software suite, and the detailed procedure is described in
Figure S6. (f) Logarithm of the specific viscosities of 4 mM DCL-2
(black curve), 4 mM Kn1 (red curve), and 4 mM Kn3 (green curve)
in 30% C2D2Cl4 + 70% MCH-d14 as a function of temperature (see
the Supporting Information for the detailed description of viscosity
measurements). The green and red shaded areas in (a) and (b)
The solution of 4 mM DCL-2 in 30% C2D2Cl4 + 70%
MCH-d14 was submitted to a similar procedure. At each
temperature, the library was allowed to equilibrate until a
1
stable distribution was obtained (see Figure 6b for the H
NMR spectra). The LW1/2 values of the Hb and H2 signals
versus temperature are shown in Figure 6d. The curve of H2
(Kn3) shows a sharp variation of the slope at about 348 K, in
line with the green curve in Figure 6c. However, as compared
to the Hb curve for Kn1 alone where the line width barely
changes, the curve of Hb (Kn1) in the DCL-2 displays a
different behavior with a growth in line width down from 348
K, indicating the participation of Kn1 in stacking of the
rosettes, by supramolecular copolymerization, which is also
suggested by the less defined appearance of the assembled
entities (Figure 5d). Furthermore, the variation of the
equilibria of DCL-2 under temperature cooling as illustrated
by the black curve in Figure 6d for the ratio of (A1)/(A3)
versus temperature (right Y axis) shows a synchronous trend
with the line width variation of the Hb and H2 signals as
confirmed by the linear correlation between the (A1)/(A3)
ratio and the Hb and H2 signal line widths shown in Figure 6e.
On the other hand, bulk viscosity measurements (see the
Supporting Information for details) showed that the logarithm
of the specific viscosity of DCL-2 (black curve in Figure 6f)
also exhibits a change in slope at 343−348 K and a nonlinear
growth under further cooling caused by both the increase in
self-organization and the amplification of the concentration of
Kn3. In comparison, the specific viscosity of Kn1 solution (red
curve in Figure 6f) was almost left unchanged upon cooling,
1
represent the domains of line broadening. The full H NMR spectra
are shown in Figure S7.
while that of Kn3 solution (green curve in Figure 6f)
underwent a sharp linear growth after being cooled below
338 K as a result of self-organization.
Further information on the relation between macroscopic
properties and microscopic behavior was obtained by
rheological investigations of the mechanical properties of 4
mM DCL-2 in 30% C2H2Cl4 + 70% MCH (equilibrated at 298
K for 40 h) at 298 K. As shown in Figure 7a, it was found that
the storage modulus (G′) was larger than the loss modulus
(G′′) with small modulus values (∼10 Pa) and that G′ was
independent of the angular frequency ω, indicating the
formation of a soft organogel. Furthermore, the strain
H
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
amino acid exchange and/or reshuffling in peptides and
proteins. In general terms, self-organization-driven selection of
those components that reinforce organization amounts to a
self-amplification of organization toward soft matter systems of
increasing complexity.4 Overall, component selection, con-
stituent amplification, and adaptation take place in a medium
dominated by frictional effects and are thus enforced by low
Reynolds number conditions.1
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c04446.
Details of instruments and measurements, synthetic
procedures and characterization of the involved
compounds, and supplementary data and figures
(PDF)
Figure 7. (a) Frequency dependence of G′ (storage modulus) and G′′
(loss molulus) of 4 mM DCL-2 (equilibrated at 298 K for 40 h) with
oscillation strain at 1%. (b) Dynamic strain sweep measurement with
angular frequency at 6.28 rad/s of DCL-2. (c) Temperature
dependence (293−374 K) of G′ and G′′ of DCL-2.
AUTHOR INFORMATION
■
Corresponding Author
Jean-Marie Lehn − Lehn Institute of Functional Materials,
School of Chemistry, Sun Yat-Sen University, Guangzhou
510275, China; Laboratoire de Chimie Supramoléculaire,
Institut de Science et d’Ingénierie Supramoléculaires (ISIS),
Université de Strasbourg, Strasbourg 67000, France;
orcid.org/0000-0001-8981-4593; Email: lehn@
unistra.fr
amplitude sweep in Figure 7b showed that there was a
crossover of G′ and G′′ at 27% strain, demonstrating a
disruption of the gel state. When the gel was gradually heated
to 374 K (Figure 7c), the storage modulus underwent two
sharp declines at 333 and 353 K, indicating a two-step
decomposition of the supramolecular assemblies, which may
be attributed, respectively, to the disruption of π stacking and
hydrogen bonding. The crossover of G′ and G′′ at 373 K
demonstrated a full phase transition of DCL-2 from gel to
solution.
Author
Ruirui Gu − Lehn Institute of Functional Materials, School of
Chemistry, Sun Yat-Sen University, Guangzhou 510275,
China; Laboratoire de Chimie Supramoléculaire, Institut de
Science et d’Ingénierie Supramoléculaires (ISIS), Université de
Strasbourg, Strasbourg 67000, France; orcid.org/0000-
0002-1831-7212
CONCLUSION
■
We have here demonstrated the occurrence of self-
organization-driven component selection and constituent
amplification in the two DCLs, DCL-1 and DCL-2, based
on a dynamic covalent CC/CN organo-metathesis
reaction. Constitutional adaptation in both DCLs is driven
by a two-step sequential and conditional process involving first
the formation of supramolecular rosettes, which conditions
their subsequent stacking into a supramolecular polymeric
entity SP, leading to the formation of a gel. It involves
amplification of the thermodynamically favored constituent,
that is, Kn2 for DCL-1 and Kn3 for DCL-2, together with
their constitutional agonist A1. The process is reversed with
randomization of the constituents at elevated temperature on
the gel to sol transition. Furthermore, the susceptibility of
DCL-2 to medium and to temperature results in diverse
distribution patterns, showing a remarkable (fast) adaptivity to
environmental information. Thus, DCLs based on CC/C
N DCC may serve as a sensitive touchstone for the detection
of perturbations by physical stimuli (temperature, pressure,
fields) or chemical effectors (pH, ions, medium, etc.) through
the shift in equilibrium distributions they cause.
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c04446
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This study is dedicated to the memory of Edward M. Purcell in
recognition of his inspiring paper.1 We thank the ERC
(Advanced Research Grant SUPRADAPT 290585), ANR
DYNAFUN grant no. ANR-15-CE29-0009-01, and the
University of Strasbourg Institute for Advanced Study
(USIAS) for financial support. We thank Dr. Jean-Louis
Schmitt and Dr. Cyril Antheaume for discussions of the NMR
and mass spectral data and Dr. Artem Osypenko for extensive
discussions. This work was also supported by the Integrated
Structural Biology Platform at the Centre for Integrative
Biology/IGBMC, by the French Infrastructure for Integrated
Structural Biology (FRISBI) ANR-10-INSB-05-01, and by
Instruct-ERIC, and we thank Dr. Brice Beinsteiner and Dr.
Bruno Klaholtz for the cryo-TEM analysis. We are very grateful
to Dr. Li Gong (SYSU) and Ye Wang (ISIS) for the AFM
imaging and Dr. Wende Hu (Shanghai) for the computational
results. R.G. gratefully acknowledges a postdoctoral fellowship
from Sun-Yat Sen University and financial support from the
China Postdoctoral Science Foundation.
Most significantly, the above studies involving NMR signal
line widths, bulk medium features (viscosity, rheology), and
dynamic equilibration in this complex system demonstrate the
correlation between (passive) self-organization and constituent
amplification under component selection. They thus unequiv-
ocally establish the phenomenon of self-organization-driven
adaptation of a constitutional dynamic system. It may also
operate in biological processes such as enzyme-catalyzed
I
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Cages. J. Am. Chem. Soc. 2020, 142, 7749−7753. (h) Wang, G. T.;
Lin, J. B.; Jiang, X. K.; Li, Z. T. Cholesterol-Appended Aromatic
Imine Organogelators: A Case Study of Gelation-Driven Component
Selection. Langmuir 2009, 25, 8414−8418. (i) Zhang, Q.; Deng, Y.
X.; Luo, H. X.; Shi, C. Y.; Geise, G. M.; Feringa, B. L.; Tian, H.; Qu,
D. H. Assembling a Natural Small Molecule into a Supramolecular
Network with High Structural Order and Dynamic Functions. J. Am.
Chem. Soc. 2019, 141, 12804−12814.
REFERENCES
■
(1) Purcell, E. M. Life at Low Reynolds Number. Am. J. Phys. 1977,
45, 3−11.
(2) Bloembergen, N.; Purcell, E. M.; Pound, R. V. Relaxation Effects
in Nuclear Magnetic Resonance Absorption. Phys. Rev. 1948, 73,
679−712.
(3) Lehn, J.-M. Perspectives in Chemistry−Steps towards Complex
Matter. Angew. Chem., Int. Ed. 2013, 52, 2836−2850.
(13) (a) Fouquey, C.; Lehn, J.-M.; Levelut, A.-M. Molecular
Recognition Directed Self-Assembly of Supramolecular Liquid
Crystalline Polymers from Complementary Chiral Components.
Adv. Mater. 1990, 2, 254−257. (b) Lehn, J.-M.; Mascal, M.; Decian,
A.; Fischer, J. Molecular Recognition Directed Self-Assembly of
Ordered Supramolecular Strands by Cocrystallization of Comple-
mentary Molecular Components. J. Chem. Soc., Chem. Commun. 1990,
479−481. (c) Ghadiri, M. R.; Granja, J. R.; Milligan, R. A.; McRee, D.
E.; Khazanovich, N. Self-Assembling Organic Nanotubes Based on A
Cyclic Peptide Architecture. Nature 1993, 366, 324−327. (d) Marsh,
A.; Silvestri, M.; Lehn, J.-M. Self-Complementary Hydrogen Bonding
Heterocycles Designed for the Enforced Self-Assembly into Supra-
molecular Macrocycles. Chem. Commun. 1996, 13, 1527−1528.
(e) Fenniri, H.; Mathivanan, P.; Vidale, K. L.; Sherman, D. M.;
Hallenga, K.; Wood, K. V.; Stowell, J. G. Helical Rosette Nanotubes:
Design, Self-Assembly, and Characterization. J. Am. Chem. Soc. 2001,
123, 3854−3855. (f) Bielejewska, A. G.; Marjo, C. E.; Prins, L. J.;
Timmerman, P.; De Jong, F.; Reinhoudt, D. N. Thermodynamic
Stabilities of Linear and Crinkled Tapes and Cyclic Rosettes in
Melamine-Cyanurate Assemblies: A Model Description. J. Am. Chem.
Soc. 2001, 123, 7518−7533. (g) Prins, L. J.; De Jong, F.; Timmerman,
P.; Reinhoudt, D. N. An Enantiomerically Pure Hydrogen-Bonded
Assembly. Nature 2000, 408, 181−184. (h) Chang, S. K.; Hamilton,
A. D. Molecular Recognition of Biologically Interesting Substrates:
Synthesis of An Artificial Receptor for Barbiturates Employing Six
Hydrogen Bonds. J. Am. Chem. Soc. 1988, 110, 1318−1319.
(i) Chang, S. K.; Van Engen, D.; Fan, E.; Hamilton, A. D. Hydrogen
Bonding and Molecular Recognition: Synthetic, Complexation, and
Structural Studies on Barbiturate Binding to An Artificial Receptor. J.
Am. Chem. Soc. 1991, 113, 7640−7645. (j) Berl, V.; Schmutz, M.;
Krische, M. J.; Khoury, R. G.; Lehn, J.-M. Supramolecular Polymers
Generated from Heterocomplementary Monomers Linked through
Multiple Hydrogen-Bonding Arrays− Formation, Characterization,
and Properties. Chem. - Eur. J. 2002, 8, 1227−1244.
(14) (a) Aida, T.; Meijer, E. W. Supramolecular Polymers − we’ve
Come Full Circle. Isr. J. Chem. 2020, 60, 33−47. (b) Korevaar, P. A.;
George, S. J.; Markvoort, A. J.; Smulders, M. M. J.; Hilbers, P. A. J.;
Schenning, A. P. H. J.; De Greef, T. F. A.; Meijer, E. W. Pathway
Complexity in Supramolecular Polymerization. Nature 2012, 481,
492−496. (c) Kang, J.; Miyajima, D.; Mori, T.; Inoue, Y.; Itoh, Y.;
Aida, T. A Rational Strategy for the Realization of Chain-Growth
Supramolecular Polymerization. Science 2015, 347, 646−651.
(d) Armao, J. J.; Maaloum, M.; Ellis, T.; Fuks, G.; Rawiso, M.;
Moulin, E.; Giuseppone, N. Healable Supramolecular Polymers as
Organic Metals. J. Am. Chem. Soc. 2014, 136, 11382−11388.
(e) Adelizzi, B.; Aloi, A.; Markvoort, A. J.; Ten Eikelder, H. M. M.;
Voets, I. K.; Palmans, A. R. A.; Meijer, E. W. Supramolecular Block
Copolymers under Thermodynamic Control. J. Am. Chem. Soc. 2018,
140, 7168−7175. (f) Korevaar, P. A.; Schaefer, C.; De Greef, T. F. A.;
Meijer, E. W. Controlling Chemical Self-Assembly by Solvent-
Dependent Dynamics. J. Am. Chem. Soc. 2012, 134, 13482−13491.
(4) Lehn, J.-M. Toward Complex Matter: Supramolecular Chemistry
and Self-Organization. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 4763−
4768.
(5) De Gennes, P.-G. Soft Matter (Nobel Lecture). Angew. Chem.,
Int. Ed. Engl. 1992, 31, 842−845.
(6) (a) Lehn, J.-M. From Supramolecular Chemistry towards
Constitutional Dynamic Chemistry and Adaptive Chemistry. Chem.
Soc. Rev. 2007, 36, 151−160. (b) Corbett, P. T.; Leclaire, J.; Vial, L.;
West, K. R.; Wietor, J.-L.; Sanders, J. K. M.; Otto, S. Dynamic
Combinatorial Chemistry. Chem. Rev. 2006, 106, 3652−3711.
(c) Ladame, S. Dynamic Combinatorial Chemistry: on the Road to
Fulfilling the Promise. Org. Biomol. Chem. 2008, 6, 219−226.
(d) Lehn, J.-M. Perspectives in Chemistry−Aspects of Adaptive
Chemistry and Materials. Angew. Chem., Int. Ed. 2015, 54, 3276−
3289. (e) Belowich, M. E.; Stoddart, J. F. Dynamic Imine Chemistry.
Chem. Soc. Rev. 2012, 41, 2003−2024.
(7) (a) Black, S. P.; Sanders, J. K. M.; Stefankiewicz, A. R. Disulfide
Exchange: Exposing Supramolecular Reactivity through Dynamic
Covalent Chemistry. Chem. Soc. Rev. 2014, 43, 1861−1872. (b) Liu,
B.; Pappas, C. G.; Zangrando, E.; Demitri, N.; Chmielewski, P. J.;
Otto, S. Complex Molecules That Fold Like Proteins Can Emerge
Spontaneously. J. Am. Chem. Soc. 2019, 141, 1685−1689.
(8) Ruff, Y.; Garavini, V.; Giuseppone, N. Reversible Native
Chemical Ligation: A Facile Access to Dynamic Covalent Peptides.
J. Am. Chem. Soc. 2014, 136, 6333−6339.
(9) Lu, Y.-X.; Guan, Z. Olefin Metathesis for Effective Polymer
Healing via Dynamic Exchange of Strong Carbon−Carbon Double
Bonds. J. Am. Chem. Soc. 2012, 134, 14226−14231.
(10) (a) Chen, X.; Dam, M. A.; Ono, K.; Mal, A.; Shen, H.; Nutt, S.
R.; Sheran, K.; Wudl, F. A Thermally Re-mendable Cross-Linked
Polymeric Material. Science 2002, 295, 1698−1702. (b) Boul, P.;
Reutenauer, P.; Lehn, J.-M. Reversible Diels−Alder Reactions for the
Generation of Dynamic Combinatorial Libraries. Org. Lett. 2005, 7,
15−18. (c) Sun, H.; Kabby, C. P.; Dai, Y.; Hill, M. R.; Ghiviriga, I.;
Bapat, A. P.; Sumerlin, B. S. Macromolecular Metamorphosis via
Stimulus-Induced Transformations of Polymer Architecture. Nat.
Chem. 2017, 9, 817−823.
(11) (a) Kulchat, S.; Meguellati, K.; Lehn, J.-M. Organocatalyzed
and Uncatalyzed C = C/C = C and C = C/C = N Exchange Processes
between Knoevenagel and Imine Compounds in Dynamic Covalent
Chemistry. Helv. Chim. Acta 2014, 97, 1219−1236. (b) Gu, R.;
Flidrova, K.; Lehn, J.-M. Dynamic Covalent Metathesis in the C = C/
C = N Exchange between Knoevenagel Compounds and Imines. J.
Am. Chem. Soc. 2018, 140, 5560−5568.
(12) (a) Aida, T.; Meijer, E. W.; Stupp, S. I. Functional
Supramolecular Polymers. Science 2012, 335, 813−817. (b) Mattia,
E.; Otto, S. Supramolecular Systems Chemistry. Nat. Nanotechnol.
2015, 10, 111−119. (c) Moulin, E.; Armao, J. J.; Giuseppone, N.
Triarylamine-Based Supramolecular Polymers: Structures, Dynamics,
and Functions. Acc. Chem. Res. 2019, 52, 975−983. (d) Sorrenti, A.;
Leira-Iglesias, J.; Markvoort, A. J.; De Greef, T. F. A.; Hermans, T. M.
Non-Equilibrium Supramolecular Polymerization. Chem. Soc. Rev.
2017, 46, 5476−5490. (e) Li, C.; Iscen, A.; Sai, H.; Sato, K.; Sather,
́
́
(g) Đorđevic, L.; Marangoni, T.; Miletic, T.; Rubio-Magnieto, J.;
Mohanraj, J.; Amenitsch, H.; Pasini, D.; Liaros, N.; Couris, S.;
Armaroli, N.; Surin, M.; Bonifazi, D. Solvent Molding of Organic
Morphologies Made of Supramolecular Chiral Polymers. J. Am. Chem.
Soc. 2015, 137, 8150−8160.
́
N. A.; Chin, S. M.; Alvarez, Z.; Palmer, L. C.; Schatz, G. C.; Stupp, S.
I. Supramolecular-Covalent Hybrid Polymers for Light-Activated
Mechanical Actuation. Nat. Mater. 2020, 19, 900−909. (f) Rao, S. J.;
Zhang, Q.; Mei, J.; Ye, X. H.; Gao, C.; Wang, Q. C.; Qu, D. H.; Tian,
H. One-Pot Synthesis of Hetero[6]rotaxane Bearing Three Different
Kinds of Macrocycle through A Self-Sorting Process. Chem. Sci. 2017,
8, 6777−6783. (g) Rizzuto, F. J.; Nitschke, J. R. Narcissistic,
Integrative, and Kinetic Self-Sorting within a System of Coordination
(15) (a) Yagai, S.; Goto, Y.; Lin, X.; Karatsu, T.; Kitamura, A.;
Kuzuhara, D.; Yamada, H.; Kikkawa, Y.; Saeki, A.; Seki, S. Self-
Organization of Hydrogen-Bonding Naphthalene Chromophores into
J-type Nanorings and H-type Nanorods: Impact of Regioisomerism.
Angew. Chem., Int. Ed. 2012, 51, 6643−6647. (b) Yagai, S.; Kitamoto,
J
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Y.; Datta, S.; Adhikari, B. Supramolecular Polymers Capable of
Controlling Their Topology. Acc. Chem. Res. 2019, 52, 1325−1335.
(c) Datta, S.; Kato, Y.; Higashiharaguchi, S.; Aratsu, K.; Isobe, A.;
Saito, T.; Prabhu, D. D.; Kitamoto, Y.; Hollamby, M. J.; Smith, A. J.;
Dagleish, R.; Mahmoudi, N.; Pesce, L.; Perego, C.; Pavan, G. M.;
Yagai, S. Self-Assembled Poly-Catenanes from Supramolecular
Toroidal Building Blocks. Nature 2020, 583, 400−405. (d) Aratsu,
K.; Takeya, R.; Pauw, B. R.; Hollamby, M. J.; Kitamoto, Y.; Shimizu,
N.; Takagi, H.; Haruki, R.; Adachi, S.; Yagai, S. Supramolecular
Copolymerization Driven by Integrative Self-sorting of Hydrogen-
Bonded Rosettes. Nat. Commun. 2020, 11, 1−12.
(16) (a) Sreenivasachary, N.; Lehn, J.-M. Gelation-Driven
Component Selection in the Generation of Constitutional Dynamic
Hydrogels Based on Guanine-Quartet Formation. Proc. Natl. Acad.
Sci. U. S. A. 2005, 102, 5938−5943. (b) Nguyen, R.; Allouche, L.;
Buhler, E.; Giuseppone, N. Dynamic Combinatorial Evolution within
Self-Replicating Supramolecular Assemblies. Angew. Chem., Int. Ed.
2009, 48, 1093−1096. (c) Minkenberg, C. B.; Li, F.; van Rijn, P.;
Florusse, L.; Boekhoven, J.; Stuart, M. C. A.; Koper, G. J. M.;
Eelkema, R.; van Esch, J. H. Responsive Vesicles from Dynamic
Covalent Surfactants. Angew. Chem., Int. Ed. 2011, 50, 3421−3424.
(d) Liang, C.; Kulchat, S.; Jiang, S.; Lehn, J.-M. Gelation-Driven
Selection in Dynamic Covalent C = C/C = N Exchange. Chem. Sci.
2017, 8, 6822−6828. (e) Seoane, A.; Brea, R. J.; Fuertes, A.;
Podolsky, K. A.; Devaraj, N. K. Biomimetic Generation and
Remodeling of Phospholipid Membranes by Dynamic Imine
Chemistry. J. Am. Chem. Soc. 2018, 140, 8388−8391. (f) Gu, R.;
Lehn, J.-M. Metal Ion-Driven Constitutional Adaptation in Dynamic
Covalent C = C/C = N Organo-Metathesis. Chem. - Asian J. 2021, 16,
44−48.
(17) Bannwarth, C.; Ehlert, S.; Grimme, S. GFN2-xTB−An Accurate
and Broadly Parametrized Self-Consistent Tight-Binding Quantum
Chemical Method with Multipole Electrostatics and Density-Depend-
ent Dispersion Contributions. J. Chem. Theory Comput. 2019, 15,
1652−1671.
K
https://doi.org/10.1021/jacs.1c04446
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX