Host–Guest Chemistry Within Cellulose Nanocrystal Gel Receptors

Article abstract of DOI:10.1002/anie.201913030

Cellulose nanocrystals (CNCs) spontaneously assemble into gels when mixed with a polyionic organic or inorganic salt. Here, we have used this ion-induced gelation strategy to create functional CNC gels with a rigid tetracationic macrocycle, cyclobis(paraquat-p-phenylene) (CBPQT⁴⁺). Addition of [CBPQT]Cl₄ to CNCs causes gelation and embeds an active host inside the material. The fabricated CNC gels can reversibly absorb guest molecules from solution then undergo molecular recognition processes that create colorful host–guest complexes. These materials have been implemented in gel chromatography (for guest exchange and separation), and as elements to encode 2- and 3-dimensional patterns. We anticipate that this concept might be extended to design a set of responsive and selective gel-like materials functioning as, for instance, water-pollutant scavengers, substrates for chiral separations, or molecular flasks.

Full text of DOI:10.1002/anie.201913030

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Host–Guest Chemistry Within Cellulose Nanocrystal Gel
Receptors
Authors: Dongjie Zhang, Miguel Soto, Lev Lewis, Wadood Hamad, and
Mark MacLachlan
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201913030
Angew. Chem. 10.1002/ange.201913030
Link to VoR: http://dx.doi.org/10.1002/anie.201913030
http://dx.doi.org/10.1002/ange.201913030

10.1002/anie.201913030
Angewandte Chemie International Edition
Host–Guest Chemistry Within Cellulose Nanocrystal Gel Receptors
+
†
∥
†
†
§
Dongjie Zhang,†‡ Miguel A. Soto, Lev Lewis, Wadood Y. Hamad, and Mark J. MacLachlan*
⊥
⊥
† Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, British Columbia V6T 1Z1, Canada
‡ MIIT Key Laboratory of Critical Materials Technology for New Energy Conversion and Storage, School of Chemistry and
Chemical Engineering, Harbin Institute of Technology, Harbin 150001, P. R. China
§ Transformation and Interfaces Group, Bioproducts Innovation Centre of Excellence, FPInnovations, 2665 East Mall, Vancouver,
British Columbia V6T 1Z4, Canada
∥ Quantum Matter Institute, University of British Columbia, 2355 East Mall, Vancouver, British Columbia, V6T 1Z4 Canada
+
WPI Nano Life Science Institute, Kanazawa University, Kanazawa, 920-1192 Japan
⊥
These authors contributed equally
Abstract
Cellulose nanocrystals (CNCs) spontaneously assemble into gels when mixed with a
polyionic organic or inorganic salt. Here, we have used this ion-induced gelation
strategy to create functional CNC gels with a rigid tetracationic macrocycle,
cyclobis(paraquat-p-phenylene) (CBPQT4+). Addition of [CBPQT]Cl4 to CNCs
causes gelation and embeds an active host inside the material. The fabricated CNC gels
can reversibly absorb guest molecules from solution then undergo molecular
recognition processes that create colorful host–guest complexes. These materials have
been implemented in gel chromatography (for guest exchange and separation), and as
elements to encode 2- and 3-dimensional patterns. We anticipate that this concept might
be extended to design a set of responsive and selective gel-like materials functioning
as, for instance, water-pollutant scavengers, substrates for chiral separations, or
molecular flasks.
Introduction
Natural and synthetic gels have been known and investigated for centuries. Today, these
materials are ubiquitous and are used in diverse applications, spanning from simple
food additives[1] and drug carriers,[2–4] to intricate smart materials[5,6] and soft robotics.[7]
Gels are entangled networks where a fluid is physically confined, yielding a
material with a unique soft-solid behavior.[8] Like other porous materials,[9–11] gels can
1
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
serve as molecular receptors, absorbing chemical species from solution or air.[12–14]
When installing functional units inside a gel structure (e.g. a ligand or a molecular host),
either chemically or physically, reversible, responsive, and selective absorption
processes can occur.[15–18] Supramolecular approaches to embed functional moieties
within gels remain rare.[19–22]
Cellulose nanocrystals (CNCs) are spindle-shaped particles obtained by the acid
(H2SO4) hydrolysis of bulk cellulose. This chemical treatment produces nanoparticles
-
that contain sulfate half-ester groups (R–OSO3 ) on their surface.[23] When dispersed in
water, CNCs form colloidal suspensions that remain stable because of sulfate–sulfate
interparticle repulsions, which are poorly screened by the counterions.[24,25] These
colloids can be disturbed by the addition of inorganic or organic salts, which modifies
the ionic strength of the medium and causes gel formation (Figure 1a). This process is
spontaneous and entropically-driven,[26] and has been used by several groups to
assemble cellulose-based hydrogels.[27–31] In fact, there is considerable interest in CNC-
based gels as they are attractive materials for separation/purification purposes,[32] drug-
delivery,[33] sensing,[34,35] thermal insulation,[36] tissue engineering,[37] and 3D
printing.[38]
Here, we present a new methodology to create CNC-based materials that contain
functional ions. Specifically, we envisioned ionic macrocycles driving gelation of
CNCs to generate materials with embedded host molecules. The gels could function as
receptors to absorb guest species from solution to subsequently undergo reversible
assembly of host–guest complexes in their interior. A visible macroscale change during
the sequential process of guest absorption/self-assembly would also allow for the use
of these materials in gel chromatography and guest-conditioned patterning.
Results and Discussion
To pursue this concept, we selected the macrocycle cyclobis(paraquat-p-phenylene)
(CBPQT4+) as the functional ion.[39] This host was chosen based on the following four
characteristics: (1) its tetracationic character will facilitate CNC gelation; (2) the salt
2
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
solubility can be tuned by appropriate counterion selection; (3) this host has a shape-
persistent cavity that will not collapse inside the gels; and (4) CBPQT4+ reversibly
binds -electron-rich guests, such as G1 and G2 (Figure 1b), to produce colorful host–
guest complexes.[40–42]
Figure 1. (a) Schematic representation of the ion-induced gelation of CNCs. (b) Chemical structures of the
selected host, CBPQT4+, and guest molecules, G1 and G2. Protons assigned in the NMR discussion are labeled
in this figure.
We started by mixing an aqueous suspension of sodium-neutralized cellulose
nanocrystals (CNC-Na+, 6 wt%) with [CBPQT]Cl4 in solution. The host concentration
was varied from 0.4 to 3.8 mM to generate five independent systems. At a low salt
concentration (0.4 and 0.8 mM), the combination of both components, CNC-Na+ and
CBPQT4+, did not produce a visible change; the samples remained semitransparent and
liquid. At higher concentrations (CBPQT4+ at 1.3, 2.5, and 3.8 mM), however, we
obtained colorless hydrogels (CNC-CBPQT4+) that were either semitransparent or
3
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
opaque (Fig. S1).
From the prepared series of hydrogels, we noticed that low amounts of host (1.3
mM) produced materials with liquid-like behavior, displaying an injectable consistency.
Higher concentrations of CBPQT4+ (2.5 and 3.8 mM) generated more robust hydrogels
that could be molded and physically manipulated while preserving their shape (Fig. 2a).
These observations qualitatively suggest that more robust hydrogels can be produced
with increasing the proportion of CBPQT4+.
The enhancement of the gel stiffness with increasing host concentration was
confirmed through rheological studies of the CNC-CBPQT4+ systems (see ESI for
details). The complex modulus (G*) of an aqueous suspension of CNCs was 5.1 Pa at
an oscillatory frequency of 8.9 rad s-1; addition of a small quantity of [CBPQT]Cl4 (0.4
mM) caused a ca. 70% increase of G*, reaching 8.6 Pa. Continuous addition of host led
to a concomitant increase of G*, attaining 17.4 kPa at 3.8 mM of CBPQT4+. This
represents a 200-fold increase with respect to our weakest hydrogel, prepared at 1.3
mM of CBPQT4+ (G* = 63.2 Pa, see Fig. S2 and Table S1).
Figure 2. (a) Photographs of CNC-CBPQT4+ hydrogels at varied concentrations of [CBPQT]Cl4: 1.3 mM
(injectable viscous liquid), 2.5 mM (inverted vial), and 3.8 mM (disc-shaped gel). (b) Changes of the complex
viscoelastic modulus (G*) and -potential plotted as functions of [CBPQT]Cl4 concentration. Solid curves
depict the observed trends. The discontinuous line at 1.0 mM differentiates between the liquid and gel states
of the CNC-CBPQT4+ system.
Similarly, we assessed the change in surface potential of dispersed CNCs when
combined with the tetracationic host. By -potential analyses, we found a value of -49.4
4
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
mV for a pristine CNC-Na+ dispersion, indicating that sodium ions poorly screen the
negative surface charge of the nanocrystals, yielding a stable aqueous suspension. The
addition of CBPQT4+ (1.25 mM) to generate CNC-CBPQT4+ increased the -potential,
reaching a value of -27.2 mV (Fig. 2b). With an increase of CBPQT4+ (3.8 mM) the
potential further increased to -11.8 mV (Table S2). These results show that the
tetracation effectively shields the CNCs surface charge, even under dilute conditions
(0.4 mM, -potential = -41.0 mV), driving the aggregation of nanocrystals and the
subsequent formation of hydrogels with different mechanical strengths.
We anticipated that the prepared hydrogels could function as guest receptors in
aqueous media. Despite their apparent sturdiness, however, we observed that all gels
lost cohesion when immersed in water, and eventually collapsed (see Fig. S3). This
could be ascribed to the leaching of [CBPQT]Cl4 that leads to the redispersion of CNCs.
To address this, we employed a less polar medium composed of ethanol:water (3:1), in
which [CBPQT]Cl4 is poorly soluble. In this medium, our materials did not collapse
even after one week of storage. The ability of the CNC-CBPQT4+ gels (CBPQT4+ at
3.8 mM) to act as guest receptors was thus evaluated in an ethanol/water solvent system.
Host CBPQT4+ can accommodate guests G1 or G2 in its cavity, as thoroughly
documented by Stoddart’s group.[43,44] The assembly process has been reported in non-
competitive organic media, such as acetonitrile or acetone, but rarely explored in
water.[45] We found that host CBPQT4+ rapidly self-assembles with G1 and G2 in
ethanol/water solution to generate pseudorotaxane complexes with a 1:1 stoichiometry
4+
⸦
(see Fig. S4–S7). Both complexes, named hereafter as G1 CBPQT
and
4+
4+
⸦
⸦
G2 CBPQT , differ in two aspects. First, they have different colors: G1 CBPQT
4+
⸦
is orange (max = 460 nm) while G2 CBPQT is green in solution (max = 706 nm).
Second, they have different stability (Fig. S8), as measured by the equilibrium constant
3
-1
4
-1
4+
⸦
(Ka) and estimated as (1.3 ± 0.1) × 10 M and (4.3 ± 0.2) × 10 M for G1 CBPQT
4+
⸦
and G2 CBPQT , respectively.
We rationalized that both factors, color and affinity, would prevail when employing
the CNC-CBPQT4+ gels as receptors in a dual-phase (gel/solution) system. To prove
this, we fabricated a ~500 μL disc-shaped gel (CBPQT4+ at 3.8 mM), followed by its
5
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
solvent exchange from water to ethanol/water (see ESI). The obtained CNC-CBPQT4+
gel was immersed in a colorless solution of G1 (25 mM, 5 mL) and kept at 25 ºC. After
a few minutes, the surface of the material turned into orange while the interior remained
white; the colored portion gradually extended along the disc until, after ca. one hour,
the whole gel was uniformly orange (Fig. 3a). This macroscopic observation suggests
that as the diffusion process occurs, G1 permeates into the gel structure where it
encounters the embedded CBPQT4+ units. Then both species, being close in proximity,
4+
⸦
undergo molecular recognition to yield pseudorotaxane G1 CBPQT inside the gel,
which is unveiled by its intense orange color.
4+
⸦
Figure 3. (a) Assembly of in-gel pseudorotaxanes. From left to right, photographs of CNC-G1 CBPQT ,
6
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
4+
4+
4+
⸦
⸦
CNC-CBPQT , and CNC-G2 CBPQT gels. (b) UV-vis spectra (3:1 ethanol/water) of G1 CBPQT
1
complex within gel (dashed line) and in solution (bold line).[46] (c) H NMR spectra (400 MHz, D2O) of
4+
⸦
pseudorotaxane G1 CBPQT (top) gel (orange) and solution (black). G1 guest (bottom). (d)
Absorption/desorption cycles, using heat (red) and acid (green) stimuli, measured by 1H NMR spectroscopy.
Guest concentration was determined from solution before and after immersing a gel receptor. Uptake is
normalized by the concentration of embedded host (3.8 mM = 100%). G1 and G2 shown in red and green,
respectively.
4+
⸦
A UV-vis spectrum of the gel CNC-G1 CBPQT resembles the spectrum of
4+
⸦
G1 CBPQT in solution, with an absorption band at 460 nm (Fig. 3b). This result is
in good agreement with the formation of a charge-transfer complex between the
electron-poor paraquat moieties of the host and the electron-rich phenylene core of G1.
4+
1
⸦
We also confirmed the self-assembly of the in-gel G1 CBPQT complex by H
1
NMR spectroscopy. Figure 3c shows the H NMR spectrum of the resulting
4+
⸦
CNC-G1 CBPQT gel (see ESI for sample preparation). Interestingly, all resonances
4+
⸦
compare well with the signals assigned to G1 CBPQT in solution, although with
significant broadening. All the aromatic host resonances (, , and ) show a slight shift
4+
-
with respect to CNC CBPQT (Fig. S11); for instance,  exhibits a  = -0.2 ppm.
More noticeably, the phenylene protons (a) of G1 present an upfield shift of 3.2 ppm
compared to a pure solution of the guest. This dramatic shift can be explained by the
shielding of the G1 core while it is inside the host cavity, as in a pseudorotaxane
complex. Together, the UV-vis and NMR data verify formation of a pseudorotaxane
4+
-
host–guest complex inside the gel CNC CBPQT upon introducing guest G1.
The second guest selected, G2, yielded analogous results with only one difference:
the apparent color. Following the diffusion-driven self-assembly of
4+
⸦
CNC-G2 CBPQT , the disc-shaped gels became green (Fig. 3a), showing an
absorbance band at 706 nm in the UV-vis spectrum (Fig. S10). 1H NMR spectroscopy
confirmed the formation of pseudorotaxane within the gel by comparison with the self-
4+
⸦
assembly of G2 CBPQT in solution (Fig. S11). Of note, other properties of the gels,
e.g. stiffness and -potential, were unresponsive to the presence of G1 or G2 (Fig. S12
and Tables S3/S4). For instance, the difference in -potential between CNC-CBPQT4+
4+
⸦
and CNC-G1 CBPQT is only 0.6 mV (Table S4). We rationalize that the neutral
7
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
charge of G1 and G2, and their relatively small sizes, do not generate a significant effect
on the ion–ion interactions that hold the CNC–CBPQT4+ gel components together.
In principle, the formation of both in-gel pseudorotaxanes could be reversed by
disrupting the noncovalent interactions that hold host and guest together. We therefore
explored their controlled dissociation by employing physical and chemical stimuli.
4+
⸦
A CNC-G1 CBPQT gel was immersed in an ethanol/water mixture and
maintained at room temperature for one hour. Within this period, no visible changes
were observed. Heating the system at 60 ºC caused the orange color to fade and, after
ca. one hour, the disc-shaped material turned homogenously white. This color change
indicates that the thermal stimulus promotes the dissociation of G1 from the host cavity.
At this point, the exclusion of G1 from the gel was not ensured, so heating was
continued for four hours in fresh ethanol/water. Both phases, gel and solution, were
separated. Noteworthy, the isolated gel did not recuperate its original orange color,
1
whereas the recovered solution, analyzed by H NMR spectroscopy (Fig. S13),
contained only compound G1.
4+
⸦
Disassembly of CNC-G2 CBPQT was also accomplished. By immersing the
4+
⸦
CNC-G2 CBPQT gel in an acidic solution (3 M HCl, 2 mL), the gel visibly changed
color from green to white in ca. one hour. After keeping the gels submerged in the acidic
solution for about four hours, we recovered [G2‧H2]2+ in solution (Fig. S14 for 1H NMR
data), and CNC-CBPQT4+ as a white gel. We ascribe this dissociation to the protonation
of G2 within the material, which yields [G2‧H2]2+. This species formed inside the gel
is expected to repel the tetracationic host and escape from its cavity.
Furthermore, we demonstrated that using the appropriate stimulus (heat for G1 or
acid for G2), CNC-CBPQT4+ gels can reversibly uptake G1 and G2 up to five times
without significant decay, as shown in Figure 3d. Interestingly, the ratio of absorbed
guest varies between G1 and G2 by ca. 25%. This might be explained by the difference
in affinity between the host and both guests; in solution, CBPQT4+ binds stronger to
G2 than to G1. We also proved that a CNC-CBPQT4+ gel absorbs non-active guests,
such as tetraethylene glycol monotosylate (M1) (Fig. S19). However, the amount of
M1 absorbed was just 45% with respect to the G2 uptake. (All experiments were
8
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
performed at same concentrations of G1, G2 and M1.)
These observations confirm that the amount of absorbed (and complexed) guests
depends on the intercomponent affinity. Nevertheless, and if needed, the absorption
capacity of the receptors can be tuned by modifying the ratio of embedded host. For
example, a material prepared with CBPQT4+ at 7.5 mM can uptake about twice the
amount of G1 absorbed by a gel assembled with 3.8 mM of host (Fig. S15).
Following the analysis of the in-gel host–guest complexes, we sought to take
advantage of the contrasting affinity and apparent color of the pseudorotaxanes by
implementing the CNC-CBPQT4+ materials in gel chromatography for exchange and
separation purposes. To investigate this, a solvent-exchanged CNC-CBPQT4+ gel was
transferred into a glass pipette, then a solution of G1 (25 mM) was passed through the
packed column, which gradually turned orange accordingly with the diffusion of G1.
Although the stationary phase appeared intensely orange (Fig 4a), the collected solution
remained colorless, indicating that CBPQT4+ did not leach out along with the guest. A
1H NMR spectrum of the collected solution showed exclusively the expected
resonances for G1 (Fig 4b).
Eluting a solution of G2 (10 mM) through the same column produced guest-
exchange, so that the stationary phase slowly changed color, from orange to green, as
G2 displaced G1 from the host cavity. During the exchange process, G1 was identified
by 1H NMR spectroscopy of the collected eluent and, when completed, G2 was solely
observed (Fig. 4b). To regenerate the stationary phase, we treated the column with an
acidic solution (50 mM HCl). This caused protonation of G2 and its exclusion from the
host (and from the stationary phase), accompanied by a color change (Fig 4a). The
generated cation [G2‧H2]2+ was recovered from the column and unequivocally
identified by 1H NMR spectroscopy (Fig. 4b). Further treatment of the stationary phase
with ethanol/water allowed for its pH neutralization. No other components were noticed
in the collected solvent.
Analogously, we attempted the purification of a G1/G2 mixture through gel
column chromatography. For this, a stationary phase was made with silica gel (SiO2)
and CNC-CBPQT4+ in a 1:1 volume ratio. SiO2 provided structural support to the
9
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
CNC-CBPQT4+ gels so that a packed column could operate under pressurized
conditions. The recurrent ethanol/water solvent system was employed as eluent. After
loading the G1/G2 mixture (1:2 molar ratio),[47] a light-green color quickly developed
at the column baseline (Fig. 4c). As elution progressed, the colored baseline slowly
separated into two components that appeared orange and green in color. The orange
4+
⸦
component, assigned to the in-gel pseudorotaxane G1 CBPQT , moved faster within
4+
⸦
the stationary phase, while the green component (in-gel G2 CBPQT ) was left
behind (Fig. 4c). This difference relies, like in regular chromatographic separations, on
4+
4+
analyte-stationary phase affinity; for our system Ka(
⸦
> Ka(
⸦
). Once
G2 CBPQT
G1 CBPQT
)
the stationary phase became colorless, we assumed that separation was completed (Fig.
S20). All collected fractions, recovered as colorless solutions, were evaporated and
1
analyzed by H NMR spectroscopy. In the resulting spectra (Fig. 4d), resonances for
G1 and G2 were identified in separate fractions, whereas CBPQT4+ was not detected.
Figure 4. (a) Exchange chromatography. Photographs showing the observed changes as elution progresses
(top). Structures of the expected species and complexes present at each stage (bottom). (b) 1H NMR spectra
(400 MHz, D2O) of selected fractions collected throughout the exchange process. (c) Chromatographic
separation of a G1/G2 mixture. Images showing loaded (left) and partially eluted (right) stages; schematic
1
representations of the complexes are included. (d) H NMR spectra (400 MHz, D2O) of the G1/G2 mixture
(top) and the components G1 (middle) and G2 (bottom) separated by column chromatography.
Furthermore, we envisaged that color changes yielded by in-gel pseudorotaxanes
10
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
could be used to encode patterns into 2D and 3D structures (Fig. S22). For this, we used
a CNC-CBPQT4+ gel as the active patterning component while an analogous
CNC-PQT2+ gel was used as background (PQT2+ = methylviologen cation).
Beforehand, we conofirmed that PQT2+ does not self-assemble with G1 or G2 (Fig.
S16), and therefore when CNC-PQT2+ absorbs G1 or G2 from solution, it does not
form a colorful complex.
To encode a pattern, CNC-CBPQT4+ was cast onto a glass substrate and shaped
into a design; this was coated with the background gel to produce a homogeneous
surface (Fig. 5a). The generated material was then immersed in a solution of G2. As
the guest permeated into the material, a green color gradually appeared in localized
regions, until the pattern was fully intelligible. After submerging the material into an
acidic solution (3M HCl), the pattern was completely erased, recovering the unreadable
2D surface. This demonstrates the opportunity to use the molecular recognition
capability of CNC-CBPQT4+ for covert encryption.
Figure 5. (a) 2D patterning exposure (by G2) and erasing (through G2 protonation). (b) 3D-patterned objects
revealed by exposure to G1 (left) and G2 (right). The observed air bubbles were unintentionally trapped within
the materials during gelation.
11
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
Similarly, we fabricated a 3D-patterned object. For this, a cube-shaped CNC-
CBPQT4+ gel was located within a cylindrical CNC-PQT2+ gel. After exposing the
system to solutions of G1 or G2, only the embedded cube component became colored
(Fig. 5b). This implies that the guests travel through the CNC-PQT2+ gel without
undergoing any molecular recognition, but as they encounter CNC-CBPQT4+ the in-
gel pseudorotaxanes are formed, turning on a color and revealing the patterned shape.
We also proved that the 3D patterns can be successfully erased by applying the
appropriate stimulus: heat for G1 and acid for G2.
Conclusions
In summary, we have employed a supramolecular ion-induced strategy to produce
CNC-CBPQT4+ gels with a cationic host embedded in a cellulose nanocrystal matrix.
These materials can absorb species from solution, especially electron-rich guests such
as G1 and G2, to produce color changes. Interestingly, the guest molecules could be
expelled from the host cavity (and from the gel) by applying stimuli, heat for G1 and
acid for G2, allowing for the recovery of vacant CNC-CBPQT4+ receptors. The
reversible assembly of in-gel host–guest complexes allowed us to utilize CNC-
CBPQT4+ in gel chromatography for the exchange and separation of G1 and G2.
Additionally, the controlled emergence/fading of color was employed to encode 2D and
3D patterns, which are conditioned to the presence of a guest molecule.
We believe that this simple, but robust, gel fabrication strategy might be useful in
the future to create new functional materials. The use of polyionic species that trigger
gelation and perform a secondary function might be relevant for water remediation,
chiral separations, drug-transport/delivery, and chemical reactivity.
Acknowledgments
DZ thanks the Chinese Science Council (CSC) for a scholarship and Professors Yuyan
12
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
Liu and Zhongjun Cheng for their support. MAS thanks CONACYT for a postdoctoral
fellowship. MJM thanks NSERC for funding (Discovery and CREATE NanoMat
grants). The authors acknowledge Prof. Savvas Hatzikiriakos for access to his
rheometer.
References
[1] A. Nazir, A. Asghar, A. Aslam Maan, Adv. Food Rheol. Its Appl. 2016, 8398,
335–353.
[2] S. W. Kim, B. Jeong, Y. H. Bae, D. S. Lee, Nature 1997, 388, 860–862.
[3] Y. Qiu, K. Park, Adv. Drug Deliv. Rev. 2001, 53, 321–339.
[4] C. Karavasili, D. G. Fatouros, Drug Discov. Today 2016, 21, 157–166.
[5] Z. Wei, J. H. Yang, J. Zhou, F. Xu, M. Zrínyi, P. H. Dussault, Y. Osada, Y. M.
Chen, Chem. Soc. Rev. 2014, 43, 8114–8131.
[6] Z. Qi, C. A. Schalley, Acc. Chem. Res. 2014, 47, 2222–2233.
[7] L. Hines, K. Petersen, G. Z. Lum, M. Sitti, Adv. Mater. 2017, 29, 1603483.
[8] J. Zhang, Y. Hu, Y. Li, Gel Chemistry: Interactions, Structures and Properties,
Springer Singapore, 2018.
[9] J.-R. Li, R. J. Kuppler, H.-C. Zhou, Chem. Soc. Rev. 2009, 38, 1477–1504.
[10] N. Giri, M. G. Del Pópolo, G. Melaugh, R. L. Greenaway, K. Rätzke, T. Koschine,
L. Pison, M. F. C. Gomes, A. I. Cooper, S. L. James, Nature 2015, 527, 216.
[11] K. Jie, Y. Zhou, E. Li, Z. Li, R. Zhao, F. Huang, J. Am. Chem. Soc. 2017, 139,
15320–15323.
[12] T. Wu, M. Chen, L. Zhang, X. Xu, Y. Liu, J. Yan, W. Wang, J. Gao, J. Mater.
Chem. A 2013, 1, 7612–7621.
[13] F. Jiang, Y. Lo Hsieh, J. Mater. Chem. A 2014, 2, 6337–6342.
[14] D. M. Alzate-Sánchez, B. J. Smith, A. Alsbaiee, J. P. Hinestroza, W. R. Dichtel,
Chem. Mater. 2016, 28, 8340–8346.
[15] D. Li, Q. Li, N. Bai, H. Dong, D. Mao, ACS Sustain. Chem. Eng. 2017, 5, 5598–
5607.
[16] J. Wang, X. Wang, X. Zhang, J. Mater. Chem. A 2017, 5, 4308–4313.
[17] S. Zhao, W. J. Malfait, N. Guerrero-Alburquerque, M. M. Koebel, G. Nyström,
Angew. Chem. Int. Ed. 2018, 57, 7580–7608.
[18] D. Zhao, Y. Tian, X. Jing, Y. Lu, G. Zhu, J. Mater. Chem. A 2019, 7, 157–164.
[19] T. N. Plank, L. P. Skala, J. T. Davis, Chem. Commun. 2017, 53, 6235–6238.
[20] G.-F. Gong, Y.-Y. Chen, Y.-M. Zhang, Y.-Q. Fan, Q. Zhou, H.-L. Yang, Q.-P.
Zhang, H. Yao, T.-B. Wei, Q. Lin, Soft Matter 2019, DOI 10.1039/C9SM01035A.
[21] B. Li, B. Wang, X. Huang, L. Dai, L. Cui, J. Li, X. Jia, C. Li, Angew. Chem. Int.
Ed. 2019, 58, 3885–3889.
[22] M. Kieffer, A. M. Garcia, C. J. E. Haynes, S. Kralj, D. Iglesias, J. R. Nitschke,
S. Marchesan, Angew. Chem. Int. Ed. 2019, 58, 7982–7986.
13
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
[23] Y. Habibi, Chem. Soc. Rev. 2014, 43, 1519–1542.
[24] S. Shafeiei-Sabet, W. Y. Hamad, S. G. Hatzikiriakos, Rheol. Acta 2013, 52, 741–
751.
[25] X. M. Dong, J. F. Revol, D. G. Gray, Cellulose 1998, 5, 19–32.
[26] S. Lombardo, A. Gençer, C. Schütz, J. Van Rie, S. Eyley, W. Thielemans,
Biomacromolecules 2019, DOI 10.1021/acs.biomac.9b00755.
[27] H. Dong, J. F. Snyder, K. S. Williams, J. W. Andzelm, Biomacromolecules 2013,
14, 3338–3345.
[28] S. Shafiei-Sabet, W. Y. Hamad, S. G. Hatzikiriakos, Cellulose 2014, 21, 3347–
3359.
[29] M. Chau, S. E. Sriskandha, D. Pichugin, H. Thérien-Aubin, D. Nykypanchuk, G.
Chauve, M. Méthot, J. Bouchard, O. Gang, E. Kumacheva, Biomacromolecules
2015, 16, 2455–2462.
[30] P. Bertsch, S. Isabettini, P. Fischer, Biomacromolecules 2017, 18, 4060–4066.
[31] A. L. Oechsle, L. Lewis, W. Y. Hamad, S. G. Hatzikiriakos, M. J. MacLachlan,
Chem. Mater. 2018, 30, 376–385.
[32] F. Zhang, H. Ren, J. Dou, G. Tong, Y. Deng, Sci. Rep. 2017, 7, 40096.
[33] N. Lin, A. Gèze, D. Wouessidjewe, J. Huang, A. Dufresne, ACS Appl. Mater.
Interfaces 2016, 8, 6880–6889.
[34] J. T. Korhonen, P. Hiekkataipale, J. Malm, M. Karppinen, O. Ikkala, R. H. A.
Ras, ACS Nano 2011, 5, 1967–1974.
[35] A. E. Way, L. Hsu, K. Shanmuganathan, C. Weder, S. J. Rowan, ACS Macro Lett.
2012, 1, 1001–1006.
[36] Y. Kobayashi, T. Saito, A. Isogai, Angew. Chem. Int. Ed. 2014, 53, 10394–10397.
[37] K. J. De France, K. J. W. Chan, E. D. Cranston, T. Hoare, Biomacromolecules
2016, 17, 649–660.
[38] S. Sultan, A. P. Mathew, Nanoscale 2018, 10, 4421–4431.
[39] B. Odell, M. V Reddington, A. M. Z. Slawin, N. Spencer, J. F. Stoddart, D. J.
Williams, Angew. Chem. Int. Ed. 1988, 27, 1547–1550.
[40] E. Córdova, R. A. Bissell, N. Spencer, P. R. Ashton, J. F. Stoddart, A. E. Kaifer,
J. Org. Chem. 1993, 58, 6550–6552.
[41] R. Castro, K. R. Nixon, J. D. Evanseck, A. E. Kaifer, J. Org. Chem. 1996, 61,
7298–7303.
[42] S. Nygaard, K. C.-F. Leung, I. Aprahamian, T. Ikeda, S. Saha, B. W. Laursen, S.-
Y. Kim, S. W. Hansen, P. C. Stein, A. H. Flood, et al., J. Am. Chem. Soc. 2007,
129, 960–970.
[43] P. L. Anelli, P. R. Ashton, R. Ballardini, V. Balzani, M. Delgado, M. T. Gandolfi,
T. T. Goodnow, A. E. Kaifer, D. Philp, J. Am. Chem. Soc. 1992, 114, 193–218.
[44] C. H. Sue, S. Basu, A. C. Fahrenbach, A. K. Shveyd, S. K. Dey, Y. Y. Botros, J.
F. Stoddart, Chem. Sci. 2010, 1, 119–125.
[45] M. Bria, G. Cooke, A. Cooper, J. F. Garety, S. G. Hewage, M. Nutley, G. Rabani,
P. Woisel, Tetrahedron Lett. 2007, 48, 301–304.
+
[46] Complex G1 CBPQT was prepared at [host] = [guest] = 5 × 10-3 M.
4
⸦
+
4
⸦
CNC-G1 CBPQT gel (2 mL) was obtained and solvent-exchanged inside a
14
This article is protected by copyright. All rights reserved.

10.1002/anie.201913030
Angewandte Chemie International Edition
UV-vis cuvette (10 mm pathlength); [host] ≈ 1.9 × 10-3 M. The shown UV-vis
+
4
⸦
spectrum of CNC-G1 CBPQT was collected at t = 31 h after the addition of
G1. Full data set is shown in Fig. S9a.
[47] This molar ratio was chosen to facilitate the spectroscopic detection of both
guests by 1H NMR spectroscopy.
15
This article is protected by copyright. All rights reserved.