Você está na página 1de 8

Article

pubs.acs.org/cm

Polystyrenesulfonate Threaded in MIL-101Cr(III): A Cationic


Polyelectrolyte Synthesized Directly into a MetalOrganic
Framework
Liang Gao, Chi-Ying Vanessa Li,* and Kwong-Yu Chan*
Department of Chemistry, The University of Hong Kong, Pokfulam, Hong Kong
S Supporting Information
*

ABSTRACT: Incorporation of an ion-exchange polymer in a metal


organic framework (MOF) is an attractive strategy to achieve fast ion
exchange by increasing surface area and porosity of the material.
Synthesis of a cationic polyelectrolyte in a MOF is reported here for
the rst time. Sodium poly(4-styrenesulfonate) threaded in MIL-101
(NaPSSMIL-101) is synthesized directly with polymerization in situ
of the MOF. NaPSSMIL-101 exhibits superior exchange kinetics,
high selectivity with co-ion rejection, reversibility, and durability. The
polyelectrolyte threaded in MOF has a larger specic volume
compared to its bulk state and possesses advantageous properties.
The xed charges of the polyelectrolyte are exposed for full interaction
with solvated ions and solvent, without the need of swelling or
restructuring the porous framework.
materials.20 They have only recently been explored for ionexchange applications. Ion-exchange functionality can be
incorporated into MOFs in three ways: (1) charge-balancing
species attached to the metal sites of MOF. These species are
almost exclusively anionic and can be formed by introducing
charged secondary building units21,22 or anion stripping
postsynthesis modications;23 (2) counterions of charged
groups covalently bonded to the ligands of MOFs. These
charged organic ligands are generated by self-assembly24 or
postsynthesis modications;25 and (3) ions of guest inorganic
salts residing in pores of MOFs and introduced by
impregnation26 or one-pot synthesis.27 Although these
approaches yield MOFs incorporated with ion exchange
functionality, there are still limitations in their applications.
For instance, the cationic MOFs typically have slow ionexchange rate and MOF encapsulating inorganic Keggin acid27
or H2SO426 are inevitably vulnerable to leaching.
Incorporating an ion-exchange polymer into a MOF has been
demonstrated recently with the polymer threading through the
cavities of the framework.28 Enhanced ion exchange kinetics
and durability was displayed and attributed to the full exposure
and dispersion of non-cross-linked polymers in the highly
porous MOF matrix. The synthesis was based on the
pioneering works of Kitagawa, Uemura, and co-workers
approach of polymerization inside MOFs2935 but with a
further amination step to introduce ion-exchange function. The
open porous structure of MOF allows the physically trapped

1. INTRODUCTION
Separation of ions and charged molecules is an important
process in a wide range of applications, such as water
purication, protein separation, precious metals recovery, and
remediation of radioactive contaminations.13 An eective
means is solid-phase extraction by ion exchange.1 The
conventional ion-exchange polymeric resins, however, have
very low surface area, and charge sites are hidden due to strong
binding and interaction of polymer chains.46 In most cases,
time-consuming prewetting or swelling is needed to facilitate
ion exchange. In addition, structural changes due to swelling are
macroscopic and nonuniform, leading to osmotic shock and
polymer leaching from the beads. Eorts to increase the contact
eciency between ion-exchange sites and guest species has
been demonstrated to be an eective strategy to improve the
overall performance of conventional ion-exchange beads.
Economy and Domingurez5 proposed a class of ion-exchange
nanobers, which yields orders of magnitude increase in ionexchange rate because of higher surface-to-volume ratio
compared to the bead-type. Harmer et al.6 prepared a highsurface area silica-Naon composite, whose catalytic activities
were hundreds of times higher than low-surface area pure
Naon sphere. Similarly, Choi et al.7 also reported that
extending ion-exchange polymer within high-surface mesoporous silica can remarkably improve its catalytic activity.
Metalorganic frameworks (MOFs), alternatively named
porous coordination polymers (PCPs),816 is a new class of
nanoporous materials generally possessing high surface area and
high porosity. MOFs have been applied in a wide range of
elds, such as catalysis,17 gas storage and separation,18
conducting materials,19 and sensing and delivery of bioactive
XXXX American Chemical Society

Received: December 17, 2014


Revised: April 2, 2015

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials
polymers to contact solvent and exchange ions eciency. The
MOF framework provides durable regenerative ion exchange.
Furthermore, the non-cross-linked linearity of polymers, a
feature of synthesis in situ of the MOF framework, enhances
contact eciency and full utilization of ion-exchange sites. In
principle, this strategy of forming an ion exchange polymer
within a MOF is applicable to a variety of MOFs and ion
exchange polymers. ZIF-8 was chosen in the previous work for
its high stability in alkaline condition and threaded with
polyvinyl benzyl trimethylammonium hydroxide (PVBTAH)
for a compatible synthesis and its anion exchange function.28
The hydrophobic ZIF-8 can only be impregnated with
hydrophobic species. Hence, vinylbenzyl chloride monomers
were impregnated into ZIF-8 MOF and polymerized in situ.
Charges were introduced by animation of the PVBCZIF-8
matrix which subsequently turned hydrophilic.
It is important to extend the strategy of threading an ionexchange polymer in MOF with an example of a cation
exchange polymer. Furthermore, the high temperature and high
pH requirements of the amination step precludes many MOFs
from being selected, thus limiting the range of structural
properties of the nal polyelectrolyteMOF composite. It is
desired to simplify the multistep synthesis of polyelectrolyte in
MOF with a one-step in situ polymerization process. We report
here for the rst time a direct synthesis of polyelectrolyte
threaded in MOF with a one-step in situ polymerization.
Report here is the rst cation exchange polymer in MOF, viz.
sodium poly(4-styrenesulfonate) (NaPSS) synthesized directly
in MIL-101. In addition to being hydrophilic, the inertness of
Cr(III) prevents structure modication of MIL-101 upon
electrolyte contact.36 Although the redox active Cr(III) of MIL101 can potentially aect radical polymerizations due to
quenching of propagation radicals, there are recent reports37
of successful radical polymerization within MIL-101Cr(III).
This suggests that the quenching process did not hinder
polymerization. Most likely, radicals generation far exceeds
quenching and facilitates polymerization. X-ray photoelectron
spectroscopy (XPS) analysis of the synthesized NaPSSMIL101Cr(III) conrms the absence of Cr(II) which would have
appeared if signicant quenching has taken place (see Figure S1
in the SI).
Successful exchange with mono- and multivalent metal
cations, as well as exclusion of anion, are demonstrated.
Compared to the neat MIL-101(Cr) and commercialized cation
exchange resin Amberlite IR-120, a gel type resin of sulfonated
styrene-divinylbenzene, NaPSSMIL-101 exhibits signicantly
improved adsorption capacity, charge selectivity, cationic dye
removal, and release kinetics, making it a promising candidate
for the practical applications of solid-phase extraction.

Figure 1. Schematic of NaPSSMIL-101 synthesis and structures of


MIL-101 and NaPSSMIL-101.

previous synthesis of an ion exchange polymer in MOF, viz.


PVBTAHZIF-8.28
Figure 2(a) shows the scanning electron microscopy (SEM)
image of the synthesized NaPSSMIL-101 powder which has
the same octahedral shape as a neat MIL-101 powder. Powder
X-ray diraction (XRD) conrms that NaPSSMIL-101 has
the same XRD pattern as that of MIL-101 (Figure 2(b)). The
decrease in relative diraction intensity between 2 = 5 to 6
can be explained by the change in electron density of MIL-101
after lling with NaPSS. The corresponding transmission
electron microscopy (TEM) image in Figure 2(c) conrms
the highly porous structure of NaPSSMIL-101 with cavities
ordered in the same pattern as MIL-101.
Chemical conrmation of NaPSS conned in MIL-101 was
given by Fourier transform infrared spectroscopy (FTIR), 1H

2. RESULTS AND DISCUSSION


2.1. Synthesis and Characterization of NaPSSMIL101. The MIL-101Cr(III) framework was synthesized
following a procedure developed by Ferey et al.38 Its structure
and the synthesis of NaPSSMIL-101 is shown schematically
in Figure 1. Sodium-4-styrenesulfonate monomer was dissolved
in a mixed DMF/H2O solvent and impregnated into MIL-101.
In situ radical polymerization was initiated by AIBN and
allowed to proceed at 80 C for 5 days. The product was
washed thoroughly to remove polymers formed exterior of the
MOF and unreacted monomers. The synthesized NaPSS is
interlocked in the porous structure of MIL-101Cr(III). The
synthesis procedure is simple and has fewer steps than the

Figure 2. (a) SEM image of NaPSSMIL-101 particles (scale bar = 2


m); (b) XRD patterns for NaPSSMIL-101 and MIL-101; (c) TEM
images for NaPSSMIL-101 from dierent directions (scale bar = 50
nm).
B

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials
nuclear magnetic resonance (NMR), and energy-dispersive Xray uorescence spectroscopy attached to a transmission
electron microscope (TEM- EDX). Figure S2(a) shows the
FTIR spectra of NaPSSMIL-101 and neat MIL-101. Two
additional peaks appeared at 1215 and 1044 cm1 in the
NaPSSMIL-101 spectrum, corresponding to the symmetric
OSO vibration and SO stretching.39 Since both MIL101 and NaPSS polymer have an aromatic ring, the
corresponding FTIR peaks cannot be used to conrm the
presence of a polystyrene backbone in NaPSSMIL-101. To
provide additional chemical characterization, the NaPSS
polymer within NaPSSMIL-101 was isolated from its MIL101 host (procedure detailed in the SI) for 1H NMR
spectroscopy. As shown in Figure S2(b), locations of the 1H
NMR peaks of the NaPSS isolated from NaPSSMIL-101
match well with those of a bulk NaPSS separately prepared with
a standard procedure in solution. The peak at 9 ppm is assigned
to the H of terephthalic acid in MIL-101 fragments that remain
after digestion. The peaks between 0 and 3 ppm are assigned to
aliphatic H, while two peaks between 6 and 8 ppm belong to H
of the aromatic ring. All the peaks of NaPSSMIL-101 are
broader than those of bulk NaPSS. This broadening is likely to
be caused by the contamination with paramagnetic Cr(III)
species which is present in the isolated NaPSS as conrmed by
EDX (Figure S2(c)).
The EDX mapping via TEM in Figure S3 indicates uniform
distributions of the elements S, Na, and Cr. A uniform
distribution of NaPSS into the MIL-101 framework is therefore
suggested. The loading of NaPSS is estimated to be ca. 15.4 wt
% by XPS (Figure S4) and ca. 14 wt % by CHNS element
analysis. By digesting MIL-101, the NaPSS polymer in the
cavities was released and characterized by gel permeation
chromatography (GPC) to have a molecular weight of around
Mn = 11300 (Figure S5). This molecular weight is equivalent to
ca. 55 sodium-4-sytrene sulfonate monomers.
2.2. Porous Structure and Stability of NaPSSMISL101. The porosity and structure of NaPSSMIL-101 was
analyzed by N2 sorption experiments and compared to identical
characterizations performed for neat MIL-101. Figure 3 shows
the adsorption isotherms and corresponding pore size
distributions of MIL-101 and NaPSSMIL-101. As expected,
a decrease in BrunauerEmmettTeller (BET) surface area
from 3024 m2/g for neat MIL-101 to 1850 m2/g for
NaPSS101 was observed. This large decrease is signicantly
more than what is caused by an increase in mass when MIL-101
is loaded with 15%(m/m) NaPSS. Similarly, a decrease of pore
volume from 1.42 mL/g for MIL-101 to 0.85 mL/g for
NaPSSMIL-101 is signicantly more than the factor 1/(1.15)
due to mass increase. Therefore, NaPSS must be present inside
the cavities of MIL-101, leading to decrease in pore volume and
surface area accessible by nitrogen. From the increase of mass
and decease of pore volume, the specic volume of NaPSS in
MIL-101 is estimated to be 2.17 mL/g and the theoretical
maximum loading of NaPSS in MIL-101 is 61% mNaPSS/
mMIL-101. Furthermore, the pore size distribution calculated
by the BarrettJoynerHalenda (BJH) method suggests
narrowing of pore from 2.4 nm for MIL-101 to 2.1 nm upon
lling of NaPSS (Figure 3b). In addition to decrease of mean
pore size of NaPSSMIL-101, the broadening of the 2.1 nm
indicates perturbation in the cavity and window dimensions due
to varying NaPSS occupation of MIL-101 cavities. These N2
sorption results support the representation of NaPSSMIL-101

Figure 3. (a) N2 isothermal (77 K) and (b) BJH pore size distribution
calculated based on adsorption branch for MIL-101 and NaPSSMIL101.

in Figure 1 with the NaPSS polymer threading through cavities


of MIL-101
The 2.4 nm peak in the pore size distribution of MIL101Cr(III) is the average representation of several characteristic
dimensions of the MIL-101Cr(III) porous framework. The
MOF has two types of cavities interconnected by three types of
windows.37 The largest internal diameters of the cavities are 2.9
and 3.4 nm, correspond to internal volumes 12,700 3 and
20,600 3, respectively. The number of small cavities are twice
the amount of the large cavities. Inscribed diameters are 1.6 and
1.2 nm for the hexagonal and pentagonal, respectively. Eective
volume of a sodium styrenesulfonate molecule in MIL-101 can
be estimated from the specic volume of NaPSS in the MOF,
which is 2.17 mL/g. Therefore, small and large cavities can
contain a maximum of 17 and 28 monomer units, respectively.
The molecular weight of extracted NaPSS is 55 monomer
units. Therefore, NaPSS must interpenetrate through at least
three large cages or four small cages, or a combination of them
in MIL-101 as shown in the 3D structure of Figure 1.
Similar to PVBTAH threaded in ZIF-8,28 the NaPSS in MIL101 has a larger free volume compared to bulk NaPSS. This can
be quantied by estimating the mass density of NaPSS threaded
in MIL-101, much lower than that of bulk NaPSS. The linear
chain structure and non-cross-linked tacticity also suggest the
larger free volume. This free volume of NaPSS allows full
contact with solvent and solvated ions leading to fast and
eective ion exchange.
The residing NaPSS is not chemically bonded to the host
MIL-101, nor strongly adsorbed onto the pore wall of MIL-101.
Undulating forces of the NaPSS with 55 monomer units is
suggested and contributed to entropy that is localized within
the MIL-101 framework. This localized entropy is an important
feature for eective ion-exchange. Without the MOF framework, the NaPSS may entangle into a bulk material, losing its
C

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials

As the actual ion exchange sites in IR-120 is much more than


NaPSSMIL-101, it is reasonable to speculate that the highly
porous and open network structure of NaPSSMIL-101
signicantly enhance the accessibility to the internal active
sites. Access to hidden sites of IR-120 is a much slower process,
likely due to the limited swelling in DMF. While neat MIL-101
also adsorbs cobalt ion, as shown in Figure 4a and Figure S8.
This is due to physical adsorption rather than columbic
interaction. The cobalt adsorbed on neat MIL-101 can be
completely removed by DMF washing, with corresponding
peaks in the EDX spectra disappearing (Figure S9(b)) after
washing. On the other hand, as shown in EDX of Figure S9(a),
the Co(II) ions in NaPSSMIL-101 cannot be washed away by
DMF due to the electrostatic interaction between Co(II) and
the sulfonate group on polystyrene within MIL-101. DMF was
used as a solvent since low concentration cobalt ions has a
more pronounced UVvis absorption peak in DMF compared
to that of cobalt ions in water, as shown in Figure S10(a). Thus,
better accuracy and sensitivity can be obtained by using DMF
when UVvis was used to monitor the progress of cobalt ion
exchange. Co2+ ion exchange in water is studied using ICP-MS
instead to avoid baseline drifting and the poor UVvis signal/
noise ratio of Co2+. Since Co2+ is much better solvated by water
than DMF, adsorption(removal) by NaPSSMIL-101 is not as
high as removal from DMF. However, as shown in Figure
S10(b), there is still high cobalt ion removal of 60% by
NaPSSMIL-101 compared to only 28% by IR-120 for the
cobalt ion removal in water. The observed dierence of Co
adsorption between DMF and water is shown Figure S10(c).
There is enhanced ion exchange than that of DMF, perhaps due
to better swelling of IR-120 resin in water.
Although cobalt(II) ion exchange is clearly indicated, it is
necessary to investigate whether this ion-exchange occurs inside
the cavities of NaPSSMIL-101 or just on the external surface.
To clarify this issue, cobalt form of NaPSSMIL-101 is cut by
ultramicrotome to expose the internal surface of the nanoparticles and subject to spatial elemental characterization. The
TEM-EDX mapping of the cross section of cobalt exchanged
NaPSSMIL-101 (Figure 4b) clearly shows the highly uniform
distribution of both Co and S elements in the MIL-101 matrix.
This conrms ion exchange occurred within the cavities of
NaPSSMIL-101. Since the ion exchange capacity of
NaPSSMIL-101 is higher than the total amount of Co2+ in
the solutions, there are sodium form of PSS remaining in the
material.
2.4. Anion Rejection. In addition to fast and high capacity
exchange, charge selectivity is a critical performance indicator of
a good cation exchange material. According to Donnans
theory,40 the presence of xed negative charge inside
NaPSSMIL-101 can counteract concentration gradient of
anions, and thus reject anions entering into its pores, provided
that the charge concentration is suciently high. For
NaPSSMIL-101, in each nanocage, the xed negative charge
density is estimated to be ca. 1.4 108 C/m3 (compares to 9.6
107 C/m3 at 1 M concentration). However, it is also noted
that neat MIL-101 displayed preferential adsorption of anions,
likely due to its metal anchored groups and inherent charges.
Furthermore, van der Waals forces may compete with
electrostatic repulsion and provide adsorption insensitive to
charges. It is therefore important to demonstrate the anion
exclusion property of NaPSSMIL-101.
We observed that NaPSS has a dramatic inuence on the
anion adsorption property of MIL-101. The anion exclusion

porosity and ion-exchange sites. On the other hand, the


absence of binding between MIL-101 and NaPSS may suggest a
freely mobile NaPSS which can eventually leave the MIL-101
framework with the solvent. This concern of leaching is
dispelled since there is no observable changes in NaPSSMIL101 soaked in water over a month (700 h). The S/Cr atom
ratio remains constant (Figure S6) and there is no observable
NaPSS absorbance peak of the ltrate of NaPSSMIL-101 in
ultravioletvisible (UVvis) spectrum.
2.3. Metal Cations Exchange. The ion-exchange property
of NaPSSMIL-101 is evaluated for metal cations. Exchange of
Na+ with K+, Cu2+, Fe3+, and Co2+ is performed by soaking
NaPSSMIL-101 in separate DMF solutions containing
individual metal nitrates. After equilibration, the metal ion
exchanged NaPSSMIL-101 material is rinsed with DMF and
characterized by EDX with spectra of Na+, K+, Cu2+, Fe3+
shown in Figure S7. Successful metal cation exchange is
demonstrated by disappearance of Na peaks and appearance of
the peaks of exchanged metal species. In the control
experiments of metal nitrates interacted with neat MIL-101,
no changes were observed in EDX metal peaks. These results
conrm that cation exchange functionality is only contributed
by the NaPSS polymer.
To investigate the kinetics of ion exchange and compared
quantitatively with a commercial material, Co(II) ion exchange
is conducted for it can be monitored by UVvis (Figure S8).
Concentration proles of Co(II) after contact with NaPSSMIL-101, neat MIL-101, and a commercial cation-exchange
resin IR-120, are shown in Figure 4a. NaPSSMIL-101

Figure 4. (a) Cobalt(II) ion removal by 0.1 g each of NaPSSMIL101, MIL-101, and a commercial cation exchange resin, IR-120. The
volume of cobalt nitrate DMF solution is 2 mL, and the initial
concentration of Co2+ is 6.25 mM. The UVvis absorbance spectra of
the respective ion exchange material can be found in Figure S8. Inset
photography: the light red cobalt nitrate solution (at time = 0) turned
colorless after contacting with NaPSSMIL-101 for 40 min. (b) TEMEDX mapping measured on the ultramicrotomed slice of Co(II)
exchanged NaPSSMIL-101. Cross sections of cobalt stained
NaPSSMIL-101 was exposed after the particles were cut into slices
of 5080 nm thick. The pink square in the top STEM image is where
TEM-EDX mapping was performed.

demonstrated signicantly higher Co(II) adsorption and at a


much faster rate than that of IR-120 and neat MIL-101. In the
rst 10 min, over 70% cobalt in solution was adsorbed by
NaPSSMIL-101, while only 15% and 25% Co(II) were
adsorbed by IR-120 and MIL-101, respectively. After 40 min,
the total adsorption amount for NaPSSMIL-101 is more than
four times higher than that of IR-120.
D

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials

is 24 nm (section S-2 in the SI). This value is about ten times


higher than that of the pore width of NaPSSMIL-101.
Therefore, the anionic PSS can eectively repel the
ferricyanide anions within a MIL-101 cavity. While the PSS
repulsion may not extended directly to a neighboring cavity, it
may perturb the charge distribution of the neighboring
framework, hence indirectly excluding ferricyanide ions in a
neighboring cavity should it be empty without PSS. The term
rejection coecient (R) can be used to estimate the degree of
such kind of depletion of co-ions. The rejection coecient (R)
is estimated to be larger than 99% for 6.25 mM ferricyanide
anions (section S-2 in the SI). We emphasize that after hot
water, EtOH and DMF purication process, additional NH4F
treatment procedure41 can eectively suppressed the adsorption of ferricyanide on MIL-101.
2.5. Selective and Regenerative Adsorption of
Organic Dye. Lastly, we demonstrate NaPSSMIL-101
provides fast, selective, and regenerative removal of a cationic
dye Rhodamine B (RhB), with properties listed in Table S2, is a
common pollutant in the euents of textile plants.42 RhB is a
more bulky cation (MW = 479) than the inorganic ions tested
above. After contacting 10 mL 105 M RhB with 15 mg
NaPSSMIL-101 for 3 min, almost 100% RhB was removed.
This compares favorably to 40% and 10% removal by MIL-101
and IR-120, respectively (Figure 6). The maximum adsorption

property of NaPSSMIL-101 is demonstrated by monitoring


the uptake of ferricyanide (Fe(CN)63) ions by UVvis
spectroscopy. The Fe(CN)63 ion is suciently small (hydrated
diameter: 0.94 nm) with steric hindrance to its transport or
adsorption. Shown in Figure 5, NaPSSMIL-101 exhibits 100%

Figure 5. UVvis spectra of 0.4 mM Fe(CN)63 aqueous solution as a


function of contacting time with (a) MIL-101 and (b) NaPSSMIL101. Given the concentration of ferricyanide is relatively low, Donnan
exclusion eect can be pronounced, and thus ferricyanide was strictly
rejected by NaPSSMIL-101 as we observed for the constant
ferricyanide concentration UVvis spectra in (b). (c) Ferricyanide
adsorption kinetic comparison between MIL-101 and NaPSSMIL101.

exclusion of ferricyanide ion which is present at 0.4 mM


concentration. There is no change in the Fe(CN)63 peak in
the UVvis adsorption spectra for 120 min, as shown in Figure
5(b). On the other hand, neat MIL-101 adsorbs ferricyanide
steadily, as observed from the UVvis spectra in Figure 5(a).
After 120 min, 95% of the 0.4 mM Fe(CN)63 is adsorbed. This
anion adsorption of MIL-101 is likely due to anion exchange
between ferricyanide and OH/F ions attached on the Cr
metal site of MIL-101. These results clearly conrm the
eectiveness of sulfonate group on MIL-101 for excluding coions in the pores of MIL-101.
From the decrease of pore volume shown in the sorption
isotherms of Figure 3a, NaPSS only partially lled the pores of
MIL-101. There is, however, only one single peak centering at
2.1 nm in the pore size distribution. The absence of any peaks
of shoulders around 2.4 nm which is the pore size of neat MIL101, suggests there are very few empty cavities in MIL-101.
Should there be many unoccupied cavities, there should be two
peaks in the pore size distribution, at 2.1 and 2.4 nm,
corresponding to lled and unlled cavities. This is also
supported by the TEM-EDX mapping analysis, which shows
the uniform distribution of NaPSS within MIL-101 in
microscopic scale (Figure S3). Even though each pore is only
partially lled, electrostatic repulsion will exclude ferricyanide
anions eectively. In the negatively charged environment of a
NaPSS lled MIL-101 cavity, there should be exclusion of
ferricynaide anions. In pristine MIL-101, van der Waals (vdw)
attraction plays a major role in adsorption of ferricyanide ions,
in addition to possible electrostatic interaction. However, the
presence of NaPSS in partially lled MIL-101 pore has
sucient electrostatic repulsion to dominate over the vdw
interaction. The Debye length of a 6.25 mM ferricyanide anions

Figure 6. (a) Dye adsorbed vs time and (b) Rhodamine B removal


normalized to xed initial [RhB] but with increasing amounts of
absorbent of NaPSSMIL-101, MIL-101, and IR-120. The corresponding Langmuir tting and determination of adsorption rates are
shown in Figure S11 in the SI.

capacity Q0 can be estimated by tting the adsorption


isotherms to a Langmuir model. The Q0 of NaPSSMIL-101
is 35 mg/g, which is an order of magnitude higher than those
of MIL-101 (1.3 mg/g) and IR-120 (1.7 mg/g) (Figure
S11).
The adsorption rate for NaPSSMIL-101 can be estimated
from the initial linear region of the Langmuir plot to be 0.179
g/(mg s). This value is 2 orders of magnitude higher than
that of IR-120 (0.001 g/(mg s)), and 2.5 times higher than
MIL-101 (0.068 g/(mg s)) (Figure S11).
The economic feasibility of using NaPSSMIL-101 for dye
removal in wastewater relies on its reuse and ease of
regeneration. RhB dye desorption from the dye loaded
NaPSSMIL-101 was tested in a batch system. We observed
E

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials

solution. The change in UV peak intensity is converted to


concentration change in Figure 7(a), showing high selectivity of
NaPSSMIL-101 toward charge of the dyes. In the control
experiments of adsorption by neat MIL-101, both UVvis
peaks at 550 and 630 nm, decreased, indicating indiscriminative
adsorption, as shown in Figure S17. However, the decrease in
AB9 peak intensity is larger, indicating a preferential adsorption
of the anionic dye. The color of the control experiment remains
purple, but slightly pinkish. The adsorption of anions by neat
MIL-101 is similarly indicated in Figure 5(a) for ferricyanide
ion.
2.6. Superior Ion Exchange Oered by the Unique
Properties of Polymers in MOF. The outstanding ion
exchange performance of NaPSSMIL-101 is in many ways
parallel to that of PVBTAHZIF-8,28 the anionic polymer in
MOF. In NaPSSMIL-101 there is a linear polymer threaded
through a lattice of open cells of the MOF. For a polymer chain
in the bulk form of dry state, van der Waals forces bind the
chains tightly together and any pores created due to random
binding are macroscopic, nonuniform, and uncoordinated. The
apparent specic volume of NaPSS in MIL-101, 2.17 mL/g,
estimated by a decrease of MOF pore volume and added mass
of the polymer, is signicantly higher than the bulk equivalent
of 1.23 mL/g. The high free volume can also oer high
diusion coecients to enhance the molecular trac.43
Since the charge sites are hidden due to strong binding and
interaction of polymer chains in conventional resins, signicant
swelling is needed to facilitate ion exchange with the increase in
pore volume. Swelling takes time as the solvent slowly
impregnates into the resin which is initially nonporous. The
structural changes due to swelling are macroscopic, nonuniform, and not reversible. This limits the eciency in
reutilization and repeated use of ion-exchange resins.
On the other hand, polymers are uniformly organized in the
porous network of MOF whose structure does not change
upon contact with solvent. As illustrated in Figure 8(a), the

that 20% of the loaded RhB dye can be released in a 50 v/v%


H2O/EtOH solvent. On the other hand, pure water cannot
elude loaded RhB. The result suggests that there is some
hydrophobic interaction between RhB and NaPSSMIL-101.
Since ion-exchange plays a role in RhB capture, the RhB elusion
process is highly sensitive to ionic strength of the eluding agent.
Indeed, variations in the NaCl concentration of the elusion
solution produce large dierences in the desorption eciency.
Optimized elusion is given by 1 M NaCl in 50% EtOH
solution. (Figure S12) Beneting from the high porosity of
NaPSSMIL- 101, the RhB dye desorption is fast and 80%
RhB desorbed in 2 h. IR-120 exhibits slower and ineective
RhB dye release with only 15% RhB desorbed in 2 h (Figure
S13). Reuse and regeneration of NaPSSMIL-101 is
demonstrated in four repeated cycles of adsorption and
desorption without signicant loss in performance, according
to Figure S14. No change in the MOF structure is observed
from the XRD pattern of NaPSSMIL-101 subjected to
repeated cycles of RhB adsorption and desorption (Figure
S15).
Selectivity of the NaPSSMIL-101 is further tested with an
anionic dye of Acid Blue 9 (AB9), which can be partially
adsorbed due to signicant van der Waals attraction due to its
size, despite facing charge repulsion. The properties of AB9 are
listed in Table S1 and compared to those of RhB. UVvis
spectroscopy monitored adsorption experiments are performed
in three separate solutions of pure RhB, pure AB9, and an equal
molar mixture of RhB and AB9. In Figure S16(a), excellent
RhB cation exchange can be observed with a rapid drop in the
RhB intensity @ 550 nm in the rst 20 s. The RhB
concentration normalized to initial concentration is shown in
Figure.S16(c) indicating a 90% removal in 50 s. On the other
hand, the AB9 intensity @ 630 nm, as shown in Figure.S16(b)
remains the same after 30 min, indicating 100% rejection by
NaPSSMIL-101for the negatively charged AB9 dye.
In the mixed solution with equal molar RhB and AB9, the
UVVis spectrum (Figure 7) displays two peaks at 550 and

Figure 8. (a) In the dry state, an ion-exchange polymer is attached to


the pore wall of the MOF and counterions are tightly paired. (b) The
MOF structure does not alter when immersed in a solvent. The
polymer partially detaches from the pore wall and becomes more
extended after contacting the solvent.
Figure 7. Charge selectivity and ion-exchange kinetics of NaPSSMIL-101: (a) percentage of cationic (RhB) and anionic dyes
remaining in solution monitored with time and (b) UVvis spectra
in the presence of NaPSSMIL-101 monitored with time.

polymers in the dry state are attached to the pore wall and
conform to the outline of the MOF. There is strong interaction
between the framework and NaPSS, contributed separately by
hydrophobic interaction and electrostatic interaction between
SO3 group and the Cr(III) site. Upon immersion into a
good (aqueous or polar) solvent, the MOF structure remains
unchanged. But it is reasonable to expect the solvent to
signicantly reduce the interaction between the SO3 group
and the framework. Hydrophobic interaction may not be
aected by the solvent, hence the NaPSS only partially
detached from the framework and it recongures with the

630 nm, corresponding to the two respective species. When


immersed with NaPSSMIL-101, the RhB peak at 550 nm
steadily decreases in intensity, indicating eective cation
exchange (Figure 7). The AB9 peak at 630 nm, however,
remains unchanged, indicating eective anion exclusion by
NaPSSMIL-101. The initial purple color of the mixed
solution gradually changes to blue, the color of pure AB9
F

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Chemistry of Materials
SO3 groups solvated o the framework, as shown in Figure
8(b). Thus, the ion exchange function can be enhanced by the
more exposed SO3 sites.
In addition to fast and high capacity exchange of counterions,
the polymers in MOF demonstrate excellent co-ions exclusion.
This high selectivity is dicult to achieve in conventional resin
type ion exchange materials where channelling can occur
through macro-pores and through regions of lower charge
density. It should be noted that the loadings of the polymer
achieved are below 50% of the maximum. Higher IEC can be
achieved by increased loading. At the same time, increased
loading can also increase co-ion exclusion, though there may be
a trade-o in exchange kinetics due to the decease in porosity
with a higher packing of polymers in the MOF.
While NaPSSMIL-101 and PVBTAHZIF-828 have similar
characteristics of polymersMOF and advantages over conventional resins, they show a few dierences that can be further
exploited and developed for other specic applications.
Synthesis of NaPSSMIL-101 has fewer steps since it is
possible to impregnate the hydrophilic MIL-101 directly with
an ionic monomer. On the other hand, the alkaline resistance
required for anion exchange in high pH favors the choice of
ZIF-8 which is hydrophobic. Therefore, a nonionic monomer,
VBC, was impregnated, and amination was carried out only
after in situ polymerization. Fortunately, ZIF-8 has good
thermal and chemical stability and can resist the high pH and
high temperature conditions. The variety of ion exchange
applications and operation conditions calls for dierent choices
of MOFs, polymers, and synthesis paths. The success of
NaPSSMIL-101 and PVBTAHZIF-8 is encouraging for
more polymer/MOF combinations to be developed.

Article

EXPERIMENTAL SECTION

ASSOCIATED CONTENT

Synthesis. For NaPSSMIL-101 synthesis, typically, 0.3 g of


sodium-4-styrenesulfonate (SS) was dissolved by 1.5 mL of DMF and
water mixture (DMF:water = 10:2 v/v). Then, 3 wt % AIBN (based
on SS) was introduced as initiator. The obtained SS-AIBN in DMF/
H2O solution was mixed with 0.7 g of MIL-101 in a mortar. The
polymerization was conducted at 80 C for 5 days or longer to increase
the polymerization yield. The obtained powder was soaked in HPLC
grade DMF for 3 days to completely remove the unreacted monomer
and the loosely attached polymer between MIL-101 particles. DMF
was then removed by acetone soaking. The nal product
NaPSSMIL-101 was obtained after drying at 80 C overnight.
Perhaps due to the charge rejection eects, the second relling of the
SS monomer into NaPSSMIL-101 cannot further increase the
loading percentage of the NaPSS polymer.
Metal Ions Exchange and Ferricyanide Rejection. For the
metal ions exchange test, 0.05 g of NaPSSMIL-101 was soaked in 0.1
M metal nitrate (KNO3, Cu(NO3)2, Fe(NO3)3) DMF solution for 12
h. The solid NaPSSMIL-101 was ltrated out and soaked in DMF
for another 12 h to completely remove physically trapped nitrates.
Finally, the metal loaded NaPSSMIL-101 was repeatedly washed
with acetone to remove DMF and dried at 80 C overnight. For
ferricyanide rejection test, 0.03 g of NaPSSMIL-101 was soaked in
0.4 mM potassium ferricyanide DMF solution, and UVvis was used
to monitor the concentration. To conrm the ion exchange and
rejection performance were related to the cation exchange polymer
NaPSS, rather than the physical trapping of MIL-101, neat MIL-101 is
used as a control experiment under identical treatments.
Dye Adsorption and Desorption. For kinetics test, 15 mg of
NaPSSMIL-101 was mixed with 10 mL of 105 M Rhodamine B
(RhB) aqueous solution. The suspensions were then ltered by PTFE
lter (pore size: 450 nm) and analyzed for the concentration of
residual RhB at 5 s, 10 s, 20 s, 40 s, 1 min, 2 min, 5 min, and 10 min.
The syringe lter had no eect on the dye concentration. The
maximum absorbance was chosen to determine the dye content, and
the absorbance value for the original solution was normalized as 100%.
The absorbance value at 750 nm was used as a baseline value to
correct the baseline drift. For comparison, the dye adsorption kinetics
of 200 mg of IR-120 and 15 mg of MIL-101 were also tested under
identical conditions. The batch dye desorption was performed by
soaking dye-loaded NaPSSMIL-101 in 1 M NaCl EtOH/H2O
solution. After ltering, the RhB concentration in ltrate was
monitored by UVvis. The complete cycle process can be divided
into 4 steps and is described in section S-3.

3. CONCLUSIONS
In conclusion, we prepared a cation-exchange polymerMOF
composite, NaPSSMIL-101, by in situ polymerizing a sodium4-styrenesulfonate monomer within the nanocavities of MIL101Cr(III). The functional cation-exchange polymer threading
through the MIL-101 matrix shows robustness to water soaking
for a long period (>30 days). Comparing to a conventional ionexchange resin, the high surface area (>1800 m2/g) and high
pore volume (0.85 mL/g) oers ecient contacting of
NaPSSMIL-101 with guest species. This structural feature
translated into orders of magnitude increases in both ionexchange and regeneration rates and high ion adsorption
capacity. Due to the presence of the negatively charged
sulfonate group inside MIL-101, only the cationic guest can
enter NaPSSMIL-101 and access to the anionic guest is
denied. Based on these ion-exchange properties demonstrated
here, this composite can be potentially useful to several
applications ranging from separation, purication, and ionconduction and therefore present promising features for solidphase extraction.
With this example of cation exchange type and the earlier
anion exchange type, we have demonstrated eectively the
concept of a polymer in MOF and how excellence in
performance can be correlated to the unique structural
properties of the material. In particular, the excess free volume
and exposed charged sites of the polymer facilitate full
interaction with solvent and ions without structural changes
of the MOF framework during solvent entry. Further variations
in combinations of MOFs and polymers can ne-tune to a
specic application with the desired chemical compatibility,
stability, ion exchange capacity, selectivity, and kinetics.

S Supporting Information
*

General characterization, chemicals, data tting theory, Figures


S1S17, and Table S1. The Supporting Information is available
free of charge on the ACS Publications website at DOI:
10.1021/cm504623r.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: hrsccky@hku.hk (K.-Y.C.).


*E-mail: cyvli@hku.hk (C.-Y.V.L.).
Funding

The authors acknowledge nancial support from Strategic


Research Theme (SRT) on Clean Energy and University
Development Fund (UDF) for Initiative for Clean Energy and
Environment (ICEE).
Notes

The authors declare no competing nancial interest.


G

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Article

Chemistry of Materials

(24) Akiyama, G.; Matsuda, R.; Sato, H.; Takata, M.; Kitagawa, S.
Adv. Mater. 2011, 23, 3294.
(25) Goesten, M. G.; Juan-Alcaniz, J.; Ramos-Fernandez, E. V.; Sai
Sankar Gupta, K. B.; Stavitski, E.; van Bekkum, H.; Gascon, J.;
Kapteijn, F. J. Catal. 2011, 281, 177.
(26) Ponomareva, V. G.; Kovalenko, K. A.; Chupakhin, A. P.;
Dybtsev, D. N.; Shutova, E. S.; Fedin, V. P. J. Am. Chem. Soc. 2012,
134, 15640.
(27) Li, R.; Ren, X.; Zhao, J.; Feng, X.; Jiang, X.; Fan, X.; Lin, Z.; Li,
X.; Hu, C.; Wang, B. J. Mater. Chem. A 2014, 2, 2168.
(28) Gao, L.; Li, C. Y. V.; Chan, K. Y.; Chen, Z. N. J. Am. Chem. Soc.
2014, 136, 7209.
(29) Uemura, T.; Kitagawa, K.; Horike, S.; Kawamura, T.; Kitagawa,
S.; Mizuno, M.; Endo, K. Chem. Commun. 2005, 5968.
(30) Uemura, T.; Yanai, N.; Kitagawa, S. Chem. Soc. Rev. 2009, 38,
1228.
(31) Comotti, A.; Bracco, S.; Mauri, M.; Mottadelli, S.; Ben, T.; Qiu,
S.; Sozzani, P. Angew. Chem. Int. Ed. 2012, 51, 10136.
(32) Yanai, N.; Uemura, T.; Kitagawa, S. Chem. Mater. 2012, 24,
4744.
(33) Distefano, G.; Suzuki, H.; Tsujimoto, M.; Isoda, S.; Bracco, S.;
Comotti, A.; Sozzani, P.; Uemura, T.; Kitagawa, S. Nat. Chem. 2013, 5,
335.
(34) Uemura, T.; Kaseda, T.; Kitagawa, S. Chem. Mater. 2013, 25,
3772.
(35) Radhakrishnan, L.; Reboul, J.; Furukawa, S.; Srinivasu, P.;
Kitagawa, S.; Yamauchi, Y. Chem. Mater. 2011, 23, 1225.
(36) Kim, M.; Cahill, J. F.; Fei, H.; Prather, K. A.; Cohen, S. M. J. Am.
Chem. Soc. 2012, 134, 18082.
(37) (a) Bai, L.; Wang, P.; Bose, P.; Li, P.; Zou, R.; Zhao, Y. ACS
Appl. Mater. Interfaces 2015, 7, 5056. (b) Bromberg, L.; Su, X.; Hatton,
T. A. Chem. Mater. 2014, 26, 6257.
(38) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour,
J.; Surble, S.; Margiolaki, I. Science 2005, 309, 2040.
(39) Juan-Alcaniz, J.; Gielisse, R.; Lago, A. B.; Ramos-Fernandez, E.
V.; Serra-Crespo, P.; Devic, T.; Guillou, N.; Serre, C.; Kapteijn, F.;
Gascon. J. Catal. Sci. Technol. 2013, 3, 2311.
(40) Donnan, F. G. J. Membr. Sci. 1995, 100, 45.
(41) Hong, D. Y.; Hwang, Y. K.; Serre, C.; Ferey, G.; Chang, J. S.
Adv. Funct. Mater. 2009, 19, 1537.
(42) ONeill, C.; Hawkes, F. R.; Hawkes, D. L.; Lourenco, N. D.;
Pinheiro, H. M.; Delee, W. J. Chem. Technol. Biotechnol. 1999, 74,
1009.
(43) Budd, P. M.; McKeown, N. B.; Fritsch, D. J. Mater. Chem. 2005,
15, 1977.

ACKNOWLEDGMENTS
The authors thank Mr. Frankie Chan of Electron Microscopy
Unit at The University of Hong Kong for assistance in materials
characterizations. G. Liang gratefully acknowledges the Waters
Corporation in Shanghai for GPC testing. He also deeply
appreciates Dr. Xiaowu Dong at Zhejiang University and Prof.
Gerard Ferey at Institute Lavoisier for their knowledge sharing
and suggestions.

REFERENCES

(1) (a) Zagorodni, A. A. Ion Exchange Materials: Properties and


Applications; Elsevier: Amsterdam, 2006. (b) Simpson, N. J. K. SolidPhase Extraction: Principles, Techniques, and Applications; Marcel
Dekker: America, 2000.
(2) Elimelech, M.; Phillip, W. A. Science 2011, 333, 712.
(3) Shannon, M.; Bohn, P.; Elimelech, M.; Georgiadis, J.; Marinas, B.;
Mayes, A. Nature 2008, 452, 301.
(4) Lam, Y. L.; Yang, D.; Chan, C. Y.; Chan, K. Y.; Toy, P. H. Ind.
Eng. Chem. Res. 2009, 48, 4975.
(5) Economy, J.; Domingurez, L. Ind. Eng. Chem. Res. 2002, 41, 6436.
(6) Harmer, M. A.; Farneth, W. E.; Sun, Q. J. Am. Chem. Soc. 1996,
118, 7708.
(7) Choi, M.; Kleitz, F.; Liu, D.; Lee, H. Y.; Ahn, W.; Ryoo, R. J. Am.
Chem. Soc. 2005, 127, 1924.
(8) Kondo, M.; Yoshitomi, T.; Matsuzaka, H.; Kitagawa, S.; Seki, K.
Angew. Chem., Int. Ed. 1997, 36, 1725.
(9) Li, H.; Eddaoudi, M.; OKeeffe, M.; Yaghi, O. M. Nature 1999,
402, 276.
(10) Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Open, A. G.;
Williams, I. D. Science 1999, 283, 1148.
(11) Kitagawa, S.; Kitaura, R.; Noro, S. I. Angew. Chem., Int. Ed. 2004,
43, 2334.
(12) Rowsell, J. L. C.; Yaghi, O. M. Microporous Mesoporous Mater.
2004, 73, 3.
(13) Ferey, G. Chem. Soc. Rev. 2008, 37, 191.
(14) Furukawa, H.; Cordova, K. E.; OKeeffe, M.; Yaghi, O. M.
Science 2013, 341, 1230444.
(15) Foo, M. L.; Matsuda, R.; Kitagawa, S. Chem. Mater. 2014, 26,
310.
(16) Zhou, H.-C.; Long, J. R.; Yaghi, O. M. Chem. Rev. 2012, 112,
673.
(17) (a) Fujita, M.; Kwon, Y. J.; Washizu, S.; Ogura, K. J. Am. Chem.
Soc. 1994, 116, 1151. (b) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K.
A.; Nguyen, S. T.; Hupp, J. T. Chem. Soc. Rev. 2009, 38, 1450. (c) Ma,
L. Q.; Abney, C.; Lin, W. B. Chem. Soc. Rev. 2009, 38, 1248.
(d) Corma, A.; Garca, H.; Llabres i Xamena, F. X. Chem. Rev. 2010,
110, 4606. (e) Yoon, M.; Srirambalaji, R.; Kim, K. Chem. Rev. 2012,
112, 1196. (f) Zhu, Q. L.; Xu, Q. Chem. Soc. Rev. 2014, 43, 5468.
(g) Aijaz, A.; Xu, Q. J. Phys. Chem. Lett. 2014, 5, 1400.
(18) (a) Li, J. R.; Sculley, J.; Zhou, H. C. Chem. Rev. 2011, 112, 869.
(b) Dinca, M.; Long, J. R. Angew. Chem., Int. Ed. 2008, 47, 6766.
(c) Betard, A.; Fischer, R. A. Chem. Rev. 2011, 112, 1055. (d) Morris,
R. E.; Wheatley, P. S. Angew. Chem., Int. Ed. 2008, 47, 4966. (e) Rosi,
N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; OKeeffe, M.;
Yaghi, O. M. Science 2003, 300, 1127.
(19) (a) Yoon, M.; Suh, K.; Natarajan, S.; Kim, K. Angew. Chem., Int.
Ed. 2013, 52, 2688. (b) Horike, S.; Umeyama, D.; Kitagawa, S. Acc.
Chem. Res. 2013, 46, 2376. (c) Ramaswamy, P.; Wong, N. E.; Shimizu,
G. K. H. Chem. Soc. Rev. 2014, 43, 5913.
(20) (a) Rocca, J. D.; Liu, D.; Lin, W. Acc. Chem. Res. 2011, 44, 957.
(b) Horcajada, P.; Serre, C.; Vallet-Reg, M.; Sebban, M.; Taulelle, F.;
Ferey, G. Angew. Chem., Int. Ed. 2006, 45, 5974.
(21) Zhao, X.; Bu, X.; Wu, T.; Zheng, S. T.; Wang, L.; Feng, P. Nat.
Commun. 2013, 4 DOI: 10.1038/ncomms3344.
(22) Fei, H.; Rogow, D. L.; Oliver, S. R. J. J. Am. Chem. Soc. 2010,
132, 7202.
(23) Mao, C.; Kudla, R. A.; Zuo, F.; Zhao, X.; Mueller, L. J.; Bu, X.;
Feng, P. J. Am. Chem. Soc. 2014, 136, 7579.
H

DOI: 10.1021/cm504623r
Chem. Mater. XXXX, XXX, XXXXXX

Você também pode gostar