Redox-Responsive Covalent Organic Nanosheets from Viologens and Calix[4]arene for Iodine and Toxic Dye Capture

Article abstract of DOI:10.1002/chem.201800623

Owing to their chemical and thermal stabilities, high uptake capacities, and easy recyclability, covalent organic polymers (COPs) have shown promise as pollutant sponges. Herein, we describe the use of diazo coupling to synthesize two cationic COPs, COP1<

Full text of DOI:10.1002/chem.201800623

A Journal of
Accepted Article
Title: Redox-responsive Covalent Organic Nanosheets from Viologens
and Calix[4]arene for Iodine and Toxic Dye Capture
Authors: Tina Skorjanc, Dinesh Shetty, Sudhir Sharma, Jesus Raya,
Hassan Traboulsi, Dong-Suk Han, Jayesh Lalla, Ryan
Newlon, Ramesh Jagannathan, Serdal Kirmizialtin, John-Carl
Olsen, and Ali Trabolsi
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: Chem. Eur. J. 10.1002/chem.201800623
Link to VoR: http://dx.doi.org/10.1002/chem.201800623
Supported by

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
Redox-responsive Covalent Organic Nanosheets from Viologens
and Calix[4]arene for Iodine and Toxic Dye Capture
Tina Skorjanc,[a] Dinesh Shetty,[a] Sudhir Kumar Sharma,[b] Jesus Raya,[c] Hassan Mohamed
Traboulsi,[d] Dong Suk Han,[e] Jayesh Lalla,[f] Ryan Newlon,[f] Ramesh Jagannathan,[b] Serdal
[
a]
[f]
[a]
Kirmizialtin, John-Carl Olsen, and Ali Trabolsi*
Abstract: Owing to their chemical and thermal stabilities, high Introduction
uptake capacities, and easy recyclability, covalent organic polymers
(
COPs) have shown promise as pollutant sponges. Here, we
Molecular switches are molecules that can reversibly switch
between two or more stable states upon exposure to external
stimuli such as light, temperature, specific pH, or a reducing
environment.[1] Derivatives of 4,4’-bipyridine known as viologens,
are a class of redox-active molecular switches that can exist in
dicationic, radical cationic, or neutral states,[2] with each state
characterized by unique physical and chemical properties. For
example, dicationic viologens are aromatic and hydrophilic,
whereas neutral ones are antiaromatic, hydrophobic, and Lewis
basic. Because of their diverse properties that can be accessed
in a single molecule by means of simple reduction or oxidation
describe the use of diazo coupling to synthesize two cationic COPs,
++
++
COP1 and COP2 , that incorporate a viologen-based molecular
switch and an organic macrocycle, calix[4]arene. Both COPs form
nanosheets with height profiles of 6.00 and 8.00 nm, respectively,
based on AFM measurements. The sheets remain morphologically
intact upon one- or two-electron reductions of their viologen
subunits. MD simulations of the dicationic COPs indicate that
calix[4]arene adopts a partial cone conformation and that, in height,
++
the individual 2D polymer layers are 5.48 Å in COP1 and 5.65 Å in
++
COP2 , which, together with the AFM measurements, suggests that
the nanosheets are composed of 11 and 14 layers, respectively. The
COPs, in either dicationic, radical cationic, or neutral form exhibit
high affinity for iodine, reaching up to 200% mass increase when
exposed to iodine vapor at 70 °C, which makes the materials among
the best-performing nanosheets for iodine capture reported in the
literature. In addition, the COPs effectively remove Congo red from
solution in the pH range of 2 – 10, reaching nearly 100% removal
within 15 minutes at acidic pH.
reactions, viologens have been incorporated into covalent
organic polymers (COPs)[
3–7]
and covalent organic frameworks
(COFs)[8] for the construction of robust stimuli-responsive
materials. We recently showed, for instance, that changing the
redox state of a viologen in a COP affected the material's
selectivity for specific dyes.[ Here, we describe the combination
of a viologen with an organic macrocycle, calix[4]arene, for the
synthesis of two redox-switchable, azo-bridged COPs.
9]
Calix[n]arenes (n = 4, 6, 8) are well-known organic macrocycles
with a hydrophobic cavity and easily functionalizable upper and
lower rims for guest recognition.[10] Because of these structural
characteristics, calix[n]arenes have been incorporated into
polymers, but mostly as side-chain moieties.[11–13] Only a few
polymers having calixarenes in their main chains have been
reported in the literature.[14] We postulated that incorporating
both positively charged viologens and calix[4]arenes within the
backbone of COPs would produce multifunctional materials
useful for pollutant adsorption. For instance, it is known that
viologens[15] and other π electron-rich organic groups[16,17] have
high affinities for iodine and iodides. For example, cationic
viologens are known to bind negatively-charged polyiodides
[
[
a]
b]
T. Skorjanc, Dr. D. Shetty, Prof. S. Kirmizialtin, Prof. A. Trabolsi
Science Division
New York University Abu Dhabi
Experimental Research Building (C1), Saadiyat Island, UAE
E-mail: ali.trabolsi@nyu.edu
Dr. S. K. Sharma, Prof. R. Jagannathan
Engineering Division
New York University Abu Dhabi
Experimental Research Building (C1), Saadiyat Island, UAE
Dr. J. Raya
[
[
[
c]
d]
e]
CNRS/Université de Strasbourg
–
–
–
[15]
such as I
3
, I
5
, I
7
etc. through electrostatic interactions.
1, Rue Blaise Pascal, Strasbourg 67000, France
Dr. H. M. Traboulsi
King Faisal University - Ahsaa
Furthermore, it has been reported that calix[n]arenes have the
[14,18]
ability to bind toxic anionic azo dyes such as Congo red,
31982 Ahsaa, Kingdom of Saudi Arabia
Reactive Black[19] and Direct Blue,[20] among others.[21] This
Dr. D. S. Han
property is the result of noncovalent interactions including (i)
hydrogen bonding involving the OH-rich lower rim of
calix[4]arene; (ii) π-π interactions between the phenyl rings of
the macrocycles and the dyes; and (iii) in some cases, ionic
interactions between positively-charged calix[n]arenes and
negatively-charged dyes.
Chemical Engineering Program
Texas A&M University at Qatar
Education City, Doha, Qatar
J. Lalla, R. Newlon, Dr. J.-C. Olsen
School of Sciences
[f]
Indiana University Kokomo
2300 S. Washington Street
Kokomo, Indiana, 46902, USA
The azo-bridged viologen- and calix[4]arene-containing COPs
reported here, COP1++ and COP2++, serve as efficient Congo
red sorbents in aqueous solution from pH 2 to 10. They are also
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
potent iodine sorbents capable of adsorbing up to 200% of their
weight in iodine vapor. Interestingly, these materials exhibit
nanosheet morphology regardless of the redox state of their
constituent viologens. According to AFM measurements, the
dicationic sheets have height profiles of 6.00 and 8.00 nm, and
are consistent with other aromatic carbons of the linkers, and
peaks at 138 and 140 ppm correspond to aromatic carbons of
+
+
++
calixarenes in COP1 and COP2 , respectively. A peak at 63
+
+
ppm in the spectrum of COP2 corresponds to the methyl
bridge in the linker. FTIR and NMR analyses both confirm the
formation of COP1++ and COP2++ as shown in Scheme 1.
molecular dynamics (MD) simulations indicate that single layers
thick in COP1++ and COP2 ,
++
The powder X-ray diffraction (PXRD) patterns of COP1 and
++
are 5.48
Å
and 5.65
Å
respectively, which implies that the nanosheets observed by
COP2++ exhibit broad signals corresponding to the amorphous
nature of the materials (Figure S4). Thermogravimetric analysis
(TGA) demonstrates that the polymers have enhanced thermal
stability up to 400 °C (Figure S5). Gas adsorption experiments
demonstrate that COP1++ and COP2++ have modest Brunauer-
Emmett-Teller (BET) surface areas of 17.9 and 51.8 m2 g-1
+
+
AFM are composed of 11 layers in COP1 and 14 layers in
+
+
COP2 .
Results and Discussion
(
Figure S6). Average pore diameters as determined by the
Synthesis and characterization of azo-bridged COP1++ and
Barrett-Joyner-Halenda (BJH) analysis are 192.7 Å and 148.7 Å,
respectively (Figure S6).
COP2++
+
+
The synthesis of COP1++ and COP2++ was accomplished by
Taking advantage of the viologen subunits in COP1 and
+
+
diazo bond formation[22] between
COP2 , we performed one- or two-electron reductions to obtain
a
nitro derivative of
+
•
+•
calix[4]arene (pTNC4A)[23] and two different viologen-based
radical cationic polymers COP1 and COP2 , and neutral
polymers COP10 and COP20, respectively. One-electron
diamines (1 and 2) in presence of a strong base (Scheme 1,
synthetic procedure in Experimental Section). The reaction was
carried out in a Schlenk tube using dimethylformamide (DMF) as
a solvent for 24 h at 150 °C. The resulting polymers were
purified by Soxhlet extraction with DMF, tetrahydrofuran (THF),
water and ethanol. Both materials were obtained as dark brown
powders, which were stable at highly acidic and highly basic pH
reductions were achieved with sodium dithionate (Na
2 2 4
S O ) in
degassed water under atmosphere. Although no color
N
2
change was observed for either COP, due to their dark color in
the dicationic state, electron paramagnetic resonance (EPR)
spectra show a peak at 338 mT, confirming the formation of
radical species in both COP1+• and COP2+• (Figure S7). We also
analyzed solid state UV-Vis spectra of the as-synthesized and
reduced polymers (Figure S8). The spectrum of COP1+• exhibits
a strong peak at ~420 nm, analogous to that seen in spectra of
monomeric viologen radicals.[24] A peak at ~600 nm also
confirms the formation of viologen radicals.[ On the contrary,
the spectrum of COP2+• does not contain a strong reflectance
band at 420 nm, which suggests that the radicals do not exist as
monomers but rather as dimers. In addition, the Raman
(
Figure S1), and insoluble in DMF, THF, ethanol, acetone,
diethyl ether and DMSO.
Molecular-level characterization of COP1++ and COP2++ was
accomplished with Fourier transform infrared (FTIR)
spectroscopy and cross polarization magic angle spinning (CP-
15]
MAS) 1 C solid state nuclear magnetic resonance (NMR). Peaks
3
groups, which appear at 1336 cm–1 and
corresponding to NO
516 cm–1 in the FTIR spectrum of pTNC4A, are absent in
2
1
+
•
-1
spectrum of COP1 shows a peak at 1560 cm , which is
spectra of the polymers (Figure S2), indicating successful
coupling. A weak peak at 1550 cm–1 that corresponds to N=N
[
25]
characteristic of radical monomers (Figure S9).
This peak is
+
•
absent in COP2 , which further confirms radical dimerization in
COP2+• that is likely due to flexibility of the viologen linker. EPR,
solid state UV-Vis and Raman spectroscopy confirmed that both
COP1+• and COP2+• radical cations were successfully formed.
Upon prolonged exposure to air for several days, the radical
character of the materials disappeared.
bond vibration is also present in the spectra of both COPs. CP-
MAS spectra show peaks in the range of 140 – 180 ppm that are
consistent with aromatic carbons as well as peaks near 37 ppm
that correspond to the aliphatic methylene bridges of
calix[4]arene (Figure S3). Peaks at 163 and 162 ppm are due to
the aromatic carbons directly bonded to the N of viologen in
COP1++ and COP2 , respectively. Peaks at 127 and 129 ppm
++
Scheme 1. Schematic illustration of the synthetic route to COP1++ and COP2++
.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
Figure 1. Characterization of the morphology of COP1++ by SEM (A), TEM (B) and AFM (C, D); and that of COP2 by SEM (E), TEM (F) and AFM (G,
++
H). The measured height profiles are 6.00 and 8.00 nm for COP1++ and COP2 , respectively.
++
0
0
++
classical molecular mechanics approaches,[31–34] MD simulations
To synthesize the fully reduced COP1 and COP2 , COP1 and
COP2++
were reacted with an excess of
bis(cyclopentadienyl)cobalt(II) (cobaltocene) in degassed
[35]
of covalent calixarene-based polymeric systems are still rare.
Initial structures of the repeating units of COP1 and COP2++
were constructed in primitive unit cells such that parallel layers
of sheets were stacked on top of each other in the first set
(Figure S12). Chloride ions were added to neutralize the
simulation box. These models were denoted as AA stacking
sheets. In the second set of simulations, we varied the unit cell
geometry in the initial models such that the interlayers are slid
by half of the repeating unit in one of the directions. These
constructs were denoted as AB stacking (Figure S13 and Table
S1). A rapid transformation from the planar structure to a partial
cone was observed for calix[4]arenes in both COP1++ and
COP2++ in all stacking conditions, suggesting that the partial
cone structure is dictated by the molecular topology. The partial
cone[36] conformation has three OH groups pointing downward
and one OH group pointing upward (Figure S14). The flexibility
of the calix[4]arene rings resulted in rapid exchanges of the
linker positions (up and down), which led to nanoscale
undulations of the sheets during thermal motion (Supplementary
movie SM1). However, despite the unit cell shape changes that
were allowed in simulations, we did not observe any intra-sheet
structural transitions (AA AB) in the time scale of 100 ns,
suggesting that such transitions are not accessible at room
temperature due to high energy barriers. As our objective was to
find the stable intra-sheet structure and not the transitions, we
compared the stabilities of each intra-sheet state energetically.
We estimated the free energies in the basin of attraction of each
stacking condition by computing enthalpy and free energy (Table
1).
++
acetonitrile under N
2
atmosphere. As expected, solid state EPR
0
0
spectra of COP1 and COP2 show no signal. FTIR spectra of
the COPs after one- or two-electron reductions (Figure S10) are
similar to those of the corresponding COP1++ and COP2 with
++
minor differences, including a small additional peak near 860 cm-
1
in the spectrum of COP10 and a change in the shape of the
peak at 1265 cm-1 in the spectra of both COP1+• and COP10,
which can be attributed to previously described changes in the
geometry of the viologen subunit upon reduction.[26] No peak
appears in the EPR spectrum of the materials even after a week
0
0
of exposure to air. This suggests that neutral COP1 and COP2
are stable species which do not get spontaneously oxidized into
radical cationic and afterwards to dicationic materials when
exposed to air.
Experimental and theoretical investigation of nanosheet
morphology
The morphology of COP1++ and COP2++ was investigated by
scanning electron microscopy (SEM), transmission electron
microscopy (TEM), and atomic force microscopy (AFM) (Figure
1). SEM images hint at the nanosheets which are rolled on top
of one another (Figures 1A and 1E). TEM images clearly show
thin individual nanosheets in both COPs (Figures 1B and 1F).
Neither of the reductions (one and two electron) changed the
nanosheet morphology, as demonstrated by SEM and TEM
images recorded after reductions (Figure S11). Observations by
TEM were further confirmed by AFM (Figures 1C and 1G), which
shows individual stacks of nanosheets. A typical stack is about 1
Despite the structural similarity of their components, the two
polymers differ in their preference for stacking structure. COP1++
is more stable in the AB stacking arrangement, whereas COP2++
exhibits greater stability in the AA orientation (Table 1). These
observations can be explained in terms of non-bonded
interactions, in particular, the average van der Waals and
Coulomb interactions. COP1++ is greatly stabilized by van der
–
3 μm in width and length and has a height profile well under 10
nm, as illustrated by Figures 1D and 1H.
To further investigate the formation of the nanosheets, we
performed MD simulations of COP1++ and COP2 . Although
monomeric and dimeric calixarene derivatives have been
studied extensively using both quantum mechanical[27–30] and
++
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
Figure 2. Structures of calixarene-based polymers deduced from Molecular Dynamics simulations. (A) AB stacking orientation in COP1, with π-π interactions
between the linker moieties spaced 3.89 Å apart and 2.57 Å hydrogen bonds between calixarene OH groups. (B) AA stacking orientation in COP2, with π-π
interactions between linkers spaced 5.58 Å apart and 2.94 Å hydrogen bonding between calixarene OH groups.
Waals interactions in the AB form (55.89 kcal mol-1 compared to
13.73 kcal mol-1 in AA stacking). The reason for this difference
measured by AFM are 6.00 and 8.00 nm, respectively. To relate
simulation results to experiments, we defined Δz as the average
distance between the highest and lowest atoms in a unit cell with
respect to the z coordinate in a unit cell. Height profiles from
AFM (Δh) were divided by Δz to calculate the number of 2D
polymer layers per nanosheet: n = Δh/Δz. Remarkably, we found
n = 10.97 for COP1++ and n = 13.94 for COP2 , values very
close to integers, which demonstrates that the computational
models are consistent with our physical observations. In
3
can be attributed to the formation of π-π stacking in AB, and its
absence in AA. In AA conformation, the distance between the
two adjacent aromatic rings is 3.89 Å, allowing strong π-π
stacking interactions between the viologen linkers (Figure 2A). In
contrast, parallel viologen linkers are about 8.06 Å away from
each other in the AA conformation, which prevents
intramolecular attraction between the linkers (Figure S15A). In
addition, the conformation of calix[4]arene appears to be
somewhat distorted in the AA stacking with distances between
OH groups of adjacent calixarenes being too large for hydrogen
bond formation (4.24 Å). The same distances in the AB stacking
mode are only 2.57 Å, allowing hydrogen bond formation.
Finally, stabilization in COP1++ is also supported by electrostatic
interactions. AB stacking mode shows an average Coulomb
energy of –115.76 kcal mol-1 whereas AA stacking results in a
++
+
+
contrast, n = 7.42 for COP1 in its disfavored AA stacking, and
n = 13.79 for COP2++ in its disfavored AB stacking. Given that n
+
+
must be an integer, AA stacking in COP1 and AB stacking in
COP2++ are inconsistent with our experimental results.
Table 1. Thermodynamic parameters obtained in MD simulations of COP1++
and COP2++ in both AA and AB stacking configurations.
COP1++
AA
COP1++
AB
COP2++ COP2++
-1
higher electrostatic energy of –72.80 kcal mol . Interestingly,
AA
AB
+
+
++
Enthalpy [kcal mol-1]
Free energy [kcal mol-1]
unlike COP1 , AA stacking is more stable in COP2 . Between
577.96
579.91
313.73
189.37
191.26
55.89
79.42
81.33
30.23
345.94
347.79
212.69
the two non-bonded energy terms the electrostatic energy of AA
+
+
and AB stacking modes are similar in COP2 . van der Waals
interactions on the other hand, show a drastic difference that
amounts to 30.23 kcal mol-1 for AA and to 212.69 kcal mol-1 in
AB. A major stabilization of the AA stacking arises from
favorable orientation, small enough distances between linkers in
adjacent layers (5.58 Å), and hydrogen bond formation between
adjacent calix[4]arene moieties (2.94 Å; Figure 2B). Linker
moieties in AB stacking are not well aligned across viologen
rings. Also, they are too far apart (> 6.32 Å) to form π-π
interactions (Figure S15B).
van der Waals int.
kcal mol-1]
[
Coulomb int.
-72.80
-115.76
-96.99
-102.49
kcal mol-1]
[
Iodine capture with COP1 and COP2
Knowing that viologen species,
rings[16] interact strongly with iodine, we decided to test our
[15]
azo bonds[37,38] and phenyl
polymers as iodine sorbents. Significant amounts of iodine
radionuclides (I129 and I131) with half-lives of up to 15.7 million
Our results show how interplay between hydrogen bonding and
π-π stacking governs the overall topology of nanosheet
formation in calix[4]arene-based COPs. These two forces favor
the AB super-structure in COP1++ and disfavor the AB
[39]
years are generated in nuclear power plants, and these pose
a great danger to the environment and human health if released.
Although inorganic materials[40] and metal organic frameworks
+
+
configuration in COP2 . Our molecular modeling results for
MOFs)[41] have been used for iodine capture, these classes of
(
+
+
++
COP1 and COP2 are in good agreement with our AFM
material suffer from low uptake capacity and poor thermal and
water stability, respectively. COPs, in contrast, display excellent
measurements. The height profiles of COP1 and COP2++
+
+
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
thermal and chemical stability, can be synthesized on large
scale and are recyclable.[42]
monitored by UV-Vis spectrophotometry (Figure S21).
Regardless of redox state, COP2 performed better than COP1
Considering the importance of developing materials for I
2
2
(Figure 3C): COP2 removed > 97% of I in all redox states, while
removal, we utilized our viologen-based materials for iodine
capture, and investigated the effect of viologen redox state on
the uptake capacities. In vapor phase experiments, COPs were
exposed to iodine vapors in a closed chamber at 70 °C and
ambient pressure, and iodine uptake was measured
gravimetrically after 24 h (Figure 3A). The particularly high
the performance of COP1 varied between 57% and 95%
depending on the redox state. The flexibility of COP2 likely
allows the material to adjust its structure in solution so as to
maximize interactions with the small molecule. Furthermore,
both radical-containing polymers, COP1+• and COP2+•,
performed worse (57% and 96%, respectively) than the
+
•
++
uptake capacity observed for COP1 , up to 200 wt %, can be
corresponding dicationic polymers (95% and 100% for COP1
+
+
explained by a mechanism that involves an interaction between
and COP2 , respectively) and neutral polymers (91% and 98%
COP1+• and the polymerized I –
species and that ultimately
for COP1 and COP2 , respectively). Although COP1 , COP1 ,
COP2++ and COP20 all removed nearly 100% of iodine in 210
0
0
++
0
3
yields oxidized COP1++ and I– ions (Figure S16), as supported
by the EPR-silent spectrum of the polymer after iodine
adsorption (Figure S17). Radical stabilization through
0
min, COP2 showed a particularly quick uptake, as 100% was
removed after only 15 min (Figure S22). Upon two-electron
reduction of viologen, the counter ion is no longer bound to the
COP, so some additional surface area is available for interaction
with iodine. In addition, viologen gains the character of a Lewis
+
•
conjugation is less effective in COP2 because of the methylene
bridges present in the linker molecule and the presence of
radical dimers rather than monomers. The effects of Lewis
0
basicity in COP2 are more prominent, so the neutral polymer is
2
base, which further enhances interactions with I .
the most effective iodine capturer among the three redox states
in COP2, reaching up to a 180% weight increase. In most forms,
our COPs performed better than previously reported organic or
inorganic nanosheet materials used for iodine capture (Table
S2). X-ray photoelectron spectroscopy (XPS) of the COPs
pH dependence of dye adsorption by COP1++ and COP2++
Congo red, an anionic azo-based dye known to cause allergic
reactions as well as cytotoxicity, genotoxicity, hematotoxicity and
neurotoxicity,[45] interacts with the hydroxyl portal of
calix[n]arenes.[18] Knowledge of this favorable interaction
prompted us to test our COPs as Congo red adsorbents. Congo
red-containing effluents from textile, printing, dyeing, paper, and
other industries[46] are often acidic[47,48] but generally range in
pH depending on the source of pollution. We therefore
investigated the abilities of COP1 and COP2 in their various
redox states to remove Congo red from aqueous solutions of
different pH (Figure S23).
2
suggest that the adsorbed iodine exists as I , with signals at 619
(
3d 5/2) and 630 (3d 3/2) eV appearing in the spectra of all
+
•
iodine-exposed COPs (Figure S18) except COP1 , where peaks
appear at 617 eV (I
3
–
2
3d 5/2), 618 eV (I
3
3d 5/2), 628 eV (I
2
3d
–
/2) and 631 eV (I
3
3d 3/2). This can be explained by the
electron transfer from COP1+• to iodine (Figures S16).
Adsorption of iodine in all COPs results in an additional weight
loss in a TGA profile between 150 and 300 °C (Figure S19),
which can be attributed to iodine evaporation. COP1++ and
COP2++ were also tested for their recyclability (Figure S20). A
loss of iodine uptake capacity was observed in the 2nd cycle.
This phenomenon, which we have observed previously during
the testing of similar COPs, is likely caused by residual iodine
that remains in the material due to strong intermolecular
attractions.[
Since our materials were found to be stable from pH 2 to pH 10
(Figure S1), Congo red solutions at pH 2.0, 4.0, 6.0, 8.0 and
10.0 were prepared, and a fixed amount of the polymer was
added to each. Both COP1++ and COP2++ were effective at
removing the dye over the whole pH range after 375 min
(Figures 4A and S24). At all tested pH, COP1++ adsorbed the
15]
++
same or a lower percentage of the dye than COP2 (pH 2: 97%
Because the counter ion of viologen has been shown to
significantly influence the gas adsorption capacity of COPs,[43]
we studied the effects that changing the counter ion had on the
uptake capacities of COP1 and COP2 for iodine vapor. By
each; pH 4: 80% and 91%; pH 6: 87% and 97%; pH 8: 73% and
97%; pH 10: 84% and 97%, respectively), likely, again, because
of flexible structure of COP2++ which allows for maximum
+
+
++
interaction with the guest molecule. Both COP1 and COP2
soaking the materials in saturated aqueous solutions of KI or
performed best at pH 2.0 (97% removal in both cases). At higher
pH, the carboxylic acid groups of the dye molecule are
deprotonated, making the dye anionic. At basic pH, the phenolic
(OH) groups of calix[4]arene become deprotonated, as well, to
–
NH
4
PF
6
followed by several washes with water, Cl counter ions
–
in as-synthesized polymers were successfully replaced by I or
–
PF
6
, as demonstrated by energy dispersive X-ray (EDAX)
+
+
–
++
–
͞
spectroscopy (Figure S21). COP1 (I ) and COP2 (I ) adsorbed
about 10 and 20% more of vapor iodine than COP1 (Cl ) and
form phenoxides (Ph-O ), which makes interaction between the
+
+
–
dye and the polymer less favorable due to electrostatic
+
+
–
[18]
COP2 (Cl ), respectively (Figure 3B). This can be attributed to
repulsion. Adsorption by the as-synthesized COPs at pH 2.0
–
++
I counter ions being able to catalyze the formation of polyiodide
was investigated in greater detail in terms of kinetics. COP1
species on the surface of the polymer.[44] In contrast, substituting
and COP2++ adsorb Congo red with a pseudo second-order rate
constants of 0.0116 g/mg·min and 0.0239 g/mg·min,
respectively (Figure S24). Maximum quantity of Congo red
Cl– with bulky PF
–
reduced iodine uptake capacity by about
6
+
+
++
–
30% in both COP1 and COP2 . Due to its size, PF
6
occupies
+
+
a larger area of the polymer’s surface, so fewer intermolecular
forces can develop with vapor iodine and the uptake capacity
drops significantly.
adsorbed by COP1 as determined from a Langmuir isotherm
was 928 mg g-1 (Figure S25).
Also, adsorption experiments involving COP1 , COP1 , COP2
+
•
0
+•
In a typical solution-phase iodine adsorption experiment, 5.0 mg
of polymer was added to 5 mL of cyclohexane having an iodine
concentration of 1 mM. The progress of iodine uptake was
and COP20 were conducted at pH 2.0, the optimal pH for
adsorption. In these, 17%, 95%, 86% and 96% of the dye was
absorbed after 15 minutes, respectively. COP1+• and COP2+•
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
Figure 3. Histograms illustrating iodine adsorption by COP1 and COP2. (A) Vapor phase iodine adsorption by the three redox states of COP1 and COP2; (B)
Effect of counter ion in COP1++ and COP2++ on iodine sorption, and (C) solution phase iodine adsorption by the three redox states of COP1 and COP2.
thus performed significantly worse than the corresponding
dicationic and neutral COPs (Figure 4B). However, it is known
Cationic dye Rhodamine B and neutral dye Nile red were also
removed from aqueous solution by COP1++ and COP2++,
demonstrating that our materials serve as versatile dye
adsorbers (Figure S27).
[49,50]
that viologens are oxidized under acidic conditions,
so after
longer incubation times, COP1 and COP2+• are oxidized to
+
•
+
+
++
+•
COP1 and COP2 , respectively. Also, COP2 is oxidized
much more rapidly because the polymer is less conjugated and Conclusions
therefore stabilizes the radical form of its viologens
lesseffectively. Due to oxidation, the uptake capacity of both
radical cationic polymers becomes comparable to that of
dicationic and neutral COPs after longer incubation times
In summary, we used diazo coupling to synthesize two viologen-
based polymers that incorporate calix[4]arene in their
backbones. Interestingly, both polymers were isolated as
nanosheets, and the sheets remained intact upon chemical
reduction of their viologen subunits. The materials adsorbed
iodine vapor and iodine dissolved in cyclohexane, and removed
Congo red from aqueous solutions of different pH. The flexible
neutral polymer removed iodine from solution most efficiently,
whereas the rigid radical polymer was the most efficient at
removing iodine from the vapor phase. The cationic polymers
effectively removed Congo red from water in the pH range of 2 –
(
Figure S26).
10. These studies have provided insight into the effects that
backbone rigidity and redox state have on the uptake capacities
of viologen-based polymers and will aid the design of materials
that are even more effective at pollutant adsorption.
Experimental Section
Synthesis of COP1++ and COP2++. The polymers were synthesized
under solvothermal conditions. A magnetic stir bar, pNC4A (0.200 g,
0.33 mmol), two equivalents of the appropriate viologen diamine (1 or 2),
and KOH (0.186 g, 3.31 mmol) were placed into a Schlenk tube, and 15
mL of DMF was added under Ar atmosphere. The contents were
degassed by three freeze-pump-thaw cycles. The tube was then placed
in an oil bath and heated at 150 °C for 24 h. The reaction mixture was
centrifuged and the solid was collected and partially dried. Impurities
were removed by Soxhlet extraction with DMF, THF, water, and ethanol
as solvents. Finally, the washed polymer was dried in a vacuum oven at
55 °C for 16 hours.
MD simulations of COP1++ and COP2++ nanosheets. The repeating
units of COP1 and COP2 were drawn using GaussView[51] as shown in
Figure S10. Initial geometries were assumed to be planar. These
repeating units were imported into Materials and Process Simulation
Figure 4. pH-dependence of Congo red adsorption by COP1++ and COP2++
after 375 min of incubation (A); the effect of redox state on Congo red
adsorption at pH 2.0 after 15 minutes of incubation (B).
MAPS)[52] to create unit cells of periodic systems in the Crystal Builder
(
module as well as to prepare the input files and forcefield parameters for
the molecular simulation set-ups. Two types of unit cells were created for
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
each polymer by adjusting the angle α and the height c (Figure S12) so
that layers of sheets directly on top of one another (AA) or offset by half
of the repeating unit (AB) were created. Specific parameters used in
each unit cell can be found in Table S1 and unit cells visualized in MAPS
can be found in Figure S11. The net charge of COP1 and COP2 unit cell
was +4. To neutralize the polymer, we added 4 Cl– counter ions in
random positions into the simulation box. Geometry optimization and MD
simulations were conducted using LAMMPS[53] program with
DreidingX6[54] force field. Equations of states were integrated using a
velocity-Verlet[55] integrator with a timestep of 2 fs in the isothermal-
isobaric (NPT) ensemble for each model. A temperature of 298.15 K and
a pressure of 101.325 kPa (1 atm) were maintained using a Nose-Hoover
thermostat[56] and barostat,[57] respectively. Coulomb interactions were
Congo red. The data were fitted to Ho and McKay’s pseudo second-order
adsorption model t/q
t
= t/q
e
+ 1/(kobs·q
e
2), where q
t
is the quantity of
adsorbed Congo red at time t (min), q
e
is the adsorbed amount at
equilibrium, and kobs is the second-order rate constant (g mg-1 min-1).[59]
To determine maximum adsorption capacity, adsorption isotherms were
obtained by varying the initial concentration of the dye from 0.1 mM to
.5 mM. The adsorption isotherms were fitted (correlation coefficients, R2
1
>
0.96) by using the Langmuir adsorption model: q
e
= q
(mg L-
(mg g ) is the amount of dye adsorbed at equilibrium, C
e
m e e
·b·C /(1+b·C ),
-1
where q
1)
e
is the equilibrium solute concentration remaining in solution when q
e
is
achieved, q
m
is the maximum adsorption capacity corresponding to
-1 [14]
complete monolayer coverage, and b is a constant (L mg ).
[
58]
evaluated using an implementation of the Ewald summation algorithm
-
4
with a relative force accuracy of 1.0 × 10 . Lennard-Jones and Coulomb
interactions were cut off at a distance of 12 Å. Tail corrections were
employed for accurate computation of the long-range interactions. Each
structure was left to sample conformations for 100 ns to ascertain its
stability and to visualize and analyze the equilibrium populations. Data
were recorded with 1 ps intervals with the first 10 ns considered
equilibration and discarded. All observables reported here were obtained
by time series averaging of the snapshots.
Acknowledgements
This work was supported by New York University Abu Dhabi.
We thank NYUAD and the Core Technology Platforms for their
generous support for the research program. This research was
carried out on the High Performance Computing resources at
New York University Abu Dhabi. The authors also thank Dr.
Sajeewa Walimuni Degawe for his help with the MAPS software
and Dr. Thirumurugan Prakasam for his help with EPR
spectroscopy.
To compare the stability of alternative sheet configurations we computed
the average enthalpy ∆H =(1/푡) ∑ tt =1 ∆H , where ∆Hi is the instantaneous
i
enthalpy of snapshot i, and t is the total number of snapshots after
equilibrium is reached. The entropy of inter-layer state is estimated from
the trajectory using Gibbs Entropy ∆S = - k ∫ P(E)ln(P(E))dE , where kB
B
is the Boltzmann constant, and P(E) is the probability density of total
energy, E sampled during the simulations. Using these two entities we
compute the Gibbs free energy as ∆G = ∆H - T∆S.
Keywords: viologen • calix[4]arene • nanosheet • iodine capture
•
Congo red
Iodine capture experiments. For vapor phase experiments, ~10 mg of
the polymer was placed on a small crystallization dish, which was placed
[
1]
H. Li, D.-H. Qu, Sci. China Chem. 2015, 58, 916–921.
in
a
closed chamber containing solid I
2
and incubated at 70 °C.
[2]
[3]
C. L. Bird, A. T. Kuhn, Chem. Soc. Rev. 1981, 10, 49–82.
C. Hua, B. Chan, A. Rawal, F. Tuna, D. Collison, J. M. Hook, D. M.
D’Alessandro, J. Mater. Chem. C 2016, 4, 2535–2544.
O. Buyukcakir, S. H. Je, S. N. Talapaneni, D. Kim, A. Coskun, ACS
Appl. Mater. Interfaces 2017, 9, 7209–7216.
Gravimetric measurements of the polymer were recorded at regular time
intervals up to 24 h, and the % mass increase was calculated as follows:
%
mass increase at [t=t ]= (mass at [t=t ]-mass at [t=0])⁄mass at [t=0].
i i
[
4]
2
For solution phase experiments, 5 mL of 1 mM stock solution of I was
prepared in cyclohexane. 5.0 mg of polymer was added to the stock
solution, and the mixture was continuously stirred. The progress of iodine
capture was monitored by UV-Vis spectrophotometry at regular time
intervals up to 210 min. % iodine uptake was calculated by dividing the
absorbance at λmax = 523 nm at a given time by the absorbance at λmax of
[5]
G. Das, T. Skorjanc, S. K. Sharma, F. Gándara, M. Lusi, D. S.
Shankar Rao, S. Vimala, S. Krishna Prasad, J. Raya, D. S. Han, et
al., J. Am. Chem. Soc. 2017, 139, 9558–9565.
[
[
6]
7]
H. Yen, C. Tsai, S. Chen, G. Liou, Macromol. Rapid Commun. 2017,
2
the starting I solution.
38, 1–7.
M. Burgess, E. Che, R. S. Assary, J. Hui, S. Moore, Chem. Mater.
2016, 28, 7362–7374.
Congo red adsorption experiments. To determine the effect of pH on
Congo red uptake, 25 μM aqueous dye solutions of various pH were
prepared. Solution pH was adjusted with 1 M NaOH or 1 M HCl solutions.
UV-Vis spectra of the starting solutions were measured. Then, 5.0 mg of
polymer was added to 5 mL of each solution, and the mixture was
continuously stirred. The progress of dye uptake was monitored by UV-
Vis spectrophotometry at regular time intervals up to 375 min. Percent
Congo red uptake was calculated by dividing the absorbance of the dye
solution at λmax after 375 min by the absorbance at λmax before polymer
addition and multiplying by 100. Uptake was calculated for each polymer,
and for each of the three redox states of its constituent viologens. To
compare the different redox states of COPs in their ability to adsorb
Congo red, the % uptake after 15 min was considered.
[8]
H. Ma, B. Liu, B. Li, L. Zhang, Y.-G. Li, H.-Q. Tan, H.-Y. Zang, G.
Zhu, J. Am. Chem. Soc. 2016, 138, 5897–5903.
[
9]
G. Das, T. Skorjanc, T. Prakasam, S. Nuryyeva, J.-C. Olsen, A.
Trabolsi, RSC Adv. 2017, 7, 3594–3598.
[10]
T. Skorjanc, F. Benyettou, J.-C. Olsen, A. Trabolsi, Chem. – A Eur.
J. 2017, 23, 8333–8347.
[
[
11]
12]
L. Wang, X. Shi, P. Jia, Y. Yang, J. Polym. Sci. Part A Polym.
Chem. 2004, 42, 6259–6266.
S. Memon, O. Oguz, A. Yilmaz, M. Tabakci, M. Yilmaz, Ş. Ertul, J.
Polym. Environ. 2001, 9, 97–101.
For investigation of the kinetics, the amount of pollutant adsorbed at each
[13]
[14]
Y. Yang, T. M. Swager, Macromolecules 2007, 40, 7437–7440.
D. Shetty, I. Jahovic, J. Raya, F. Ravaux, M. Jouiad, J.-C. Olsen, A.
Trabolsi, J. Mater. Chem. A 2017, 5, 62–66.
time point was determined using the equation q
where q (mg/g) is the quantity of adsorbed Congo red per 1 g of
adsorbent at a given time t (min), C (mmol/L) is the initial concentration
before polymer treatment, C (mmol/L) is the residual concentration of the
dye after treatment with polymer, V is the volume of dye solution, m (g) is
the amount of adsorbent used, and M (g/mol) is the molecular weight of
t 0 t w
= [(C – C )·M ·V]/m,
t
0
[
15]
G. Das, T. Prakasam, S. Nuryyeva, D. S. Han, A. Abdel-Wahab, J.-
C. Olsen, K. Polychronopoulou, C. Platas-Iglesias, F. Ravaux, M.
Jouiad, et al., J. Mater. Chem. A 2016, 4, 15361–15369.
t
w
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
[
[
[
[
[
[
[
[
[
[
16]
17]
18]
19]
20]
21]
22]
23]
24]
25]
Z. Yan, Y. Yuan, Y. Tian, D. Zhang, G. Zhu, Angew. Chemie Int. Ed.
015, 54, 12733–12737.
[44]
[45]
P. H. Svensson, L. Kloo, Chem. Rev. 2003, 103, 1649–1684.
Z. L. Yaneva, N. V Georgieva, Int. Rev. Chem. Eng. 2012, 4, 127–
146.
2
D. Shetty, J. Raya, D. S. Han, Z. Asfari, J.-C. Olsen, A. Trabolsi,
Chem. Mater. 2017, 29, 8968–8972.
[46]
[47]
[48]
V. Vimonses, S. Lei, B. Jin, C. W. K. Chow, C. Saint, Chem. Eng. J.
2009, 148, 354–364.
M. A. Kamboh, I. B. Solangi, S. T. H. Sherazi, S. Memon, J. Hazard.
Mater. 2009, 172, 234–239.
L. Kos, K. Michalska, J. A. N. Perkowski, Environ. Prot. Eng. 2012,
38, 29–39.
M. A. Kamboh, I. B. Solangi, S. T. H. Sherazi, S. Memon, J. Hazard.
Mater. 2011, 186, 651–658.
M. Joshi, R. Bansal, R. Purwar, Indian J. Fibre Text. Res. 2004, 29,
239–259.
E. Akceylan, M. Bahadir, M. Yılmaz, J. Hazard. Mater. 2009, 162,
9
60–966.
M. Chen, T. Shang, W. Fang, G. Diao, J. Hazard. Mater. 2011, 185,
14–921.
R. Zhao, C. Tan, Y. Xie, C. Gao, H. Liu, Y. Jiang, Tetrahedron Lett.
011, 52, 3805–3809.
[49]
[50]
R. Bauer, H. A. F. Werner, J. Mol. Catal. 1992, 72, 67–74.
M. Venturi, Q. G. Mulazzani, M. Z. Hoffman, Radiat. Phys. Chem.
1984, 23, 229–236.
9
[51]
D. J. Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E.,
Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V.,
Petersson, G. A., Nakatsuji, H., Li, X., Caricato, M., Marenich, A. V.,
Bloino, J., Janesko, B. G., Gomperts, R., Mennucci, B., Hratch,
2016.
2
S. Kumar, N. D. Kurur, H. M. Chawla, R. Varadarajan, Synth.
Commun. 2001, 31, 775–779.
J. W. Park, N. H. Choi, J. H. Kim, J. Phys. Chem. 1996, 100, 769–
774.
[52]
[53]
[54]
http://scienomics.com/, 2016.
M. Osawa, W. Suetaka, J. Electroanal. Chem. Interfacial
Electrochem. 1989, 270, 261–272.
S. Plimpton, J. Comput. Phys. 1995, 117, 1–19.
S. L. Mayo, B. D. Olafson, W. A. Goddard, J. Phys. Chem. 1990, 94,
8897–8909.
[
[
26]
27]
W. W. Porter, T. P. Vaid, J. Org. Chem. 2005, 70, 5028–5035.
G. Bifulco, L. Gomez-Paloma, R. Riccio, C. Gaeta, F. Troisi, P. Neri,
Org. Lett. 2005, 7, 5757–5760.
[55]
W. C. Swope, H. C. Andersen, P. H. Berens, K. R. Wilson, J. Chem.
Phys. 1982, 76, 637–649.
[
[
28]
29]
I. Mohammed-Ziegler, F. Billes, J. Incl. Phenom. Macrocycl. Chem.
[56]
[57]
W. G. Hoover, Phys. Rev. A 1985, 31, 1695–1697.
S. Melchionna, G. Ciccotti, B. Lee Holian, Mol. Phys. 1993, 78, 533–
544.
2
007, 58, 19–42.
J. Schatz, Collect. Czechoslov. Chem. Commun. 2004, 69, 1169–
194.
1
[58]
[59]
I.-C. Yeh, M. L. Berkowitz, J. Chem. Phys. 1999, 111, 3155–3162.
Y.-S. Ho, G. McKay, Process Biochem. 1999, 34, 451–465.
[
[
30]
31]
M. Athar, M. Y. Lone, P. C. Jha, J. Mol. Liq. 2017, 237, 473–483.
A. Ghoufi, J.-P. Morel, N. Morel-Desrosiers, P. Malfreyt, J. Phys.
Chem. B 2005, 109, 23579–23587.
[
[
[
32]
33]
34]
P. J. Costa, I. Marques, V. Félix, Biochim. Biophys. Acta (BBA)-
Biomembranes 2014, 1838, 890–901.
I. Thondorf, F. Broda, K. Rissanen, M. Vysotsky, V. Böhmer, J.
Chem. Soc. Perkin Trans. 2 2002, 0, 1796–1800.
M. Janke, Y. Rudzevich, O. Molokanova, T. Metzroth, I. Mey, G.
Diezemann, P. E. Marszalek, J. Gauss, V. Böhmer, A. Janshoff,
Nat. Nanotechnol. 2009, 4, 225–229.
[
[
35]
36]
J. Preat, F. Rodríguez‐Ropero, J. Torras, O. Bertran, D. Zanuy, C.
Alemán, J. Comput. Chem. 2010, 31, 1741–1751.
C. D. Gutsche, B. Dhawan, J. A. Levine, K. H. No, L. J. Bauer,
Tetrahedron 1983, 39, 409–426.
[
[
37]
38]
H. Li, X. Ding, B. Han, Chem. - A Eur. J. 2016, 22, 11863–11868.
Q.-Q. Dang, X.-M. Wang, Y.-F. Zhan, X.-M. Zhang, Polym. Chem.
2
016, 7, 643–647.
J. Whitby, R. Burgess, G. Turner, J. Gilmour, J. Bridges, Science.
000, 288, 1819–1821.
[
[
39]
40]
2
D. F. Sava, M. A. Rodriguez, K. W. Chapman, P. J. Chupas, J. A.
Greathouse, P. S. Crozier, T. M. Nenoff, J. Am. Chem. Soc. 2011,
1
33, 12398–12401.
W. Xu, Y. Li, H. Wang, J. Jiang, D. Fenske, C. Su, Chem. Asian J.
016, 11, 216–220.
B. L. Rivas, J. Sanchez, B. F. Urbano, Polym. Int. 2016, 65, 255–
67.
[
[
[
41]
42]
43]
2
2
O. Buyukcakir, S. H. Je, D. S. Choi, S. N. Talapaneni, Y. Seo, Y.
Jung, K. Polychronopoulou, A. Coskun, Chem. Commun. 2016, 52,
934–937.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201800623
FULL PAPER
FULL PAPER
Tina Skorjanc, Dinesh Shetty, Sudhir
Kumar Sharma, Jesus Raya, Hassan
Mohamed Traboulsi, Dong Suk Han,
Jayesh Lalla, Ryan Newlon, Ramesh
Jagannathan Serdal Kirmizialtin, John-
Carl Olsen, and Ali Trabolsi*[
Tunable sheets for the environment. Two cationic covalent organic polymers
Page No. – Page No.
(
COPs) are synthesized through diazo coupling of a calix[4]arene derivative and
Redox-responsive Covalent Organic
Nanosheets from Viologens and
Calix[4]arene for Iodine and Toxic
Dye Capture
viologen-based, redox-responsive linkers. Interestingly, the COPs exhibit
nanosheet morphology consistent with MD simulations. Because of their cationic
and aromatic nature, these materials serve as efficient sponges for iodine and for
organic dyes over a broad range of pH.
This article is protected by copyright. All rights reserved.