Wei Jiang1,2, Dong Peng1, Wei-Rong Cui1, Ru-Ping Liang1, Jian-Ding Qiu1,3. 1. College of Chemistry, Nanchang University, Nanchang 330031, China. 2. Nanchang Institute for Food and Drug Control, Nanchang 330038, China. 3. College of Materials and Chemical Engineering, Pingxiang University, Pingxiang 337055, China.
Abstract
The effective removal of organic pollutants in wastewater is a key environmental challenge. In this work, an anionic covalent organic framework (named TpPa-SO3Na) was synthesized through a green two-in-one synthesis strategy with autocatalytic imine formation. The slowly generated acetic acid as a catalyst is favorable to sustain the reversibility of the covalent organic framework (COF) formation reaction and improve the crystallinity of TpPa-SO3Na. TpPa-SO3Na consists of a homogeneous distribution of sulfonate groups to produce negatively charged regular channels. The strong electrostatic and hydrogen-bonding interactions between the sulfonate groups anchored in the nanochannels and the amine groups in organic pollutants improve the adsorption selectivity and capacity. These structures allow a high degree of control over adsorption processes to boost the adsorption kinetics and improve selective separation. TpPa-SO3Na exhibits ultrafast adsorption (<1 min) of cationic antibiotics and dyes (average over 95%). Furthermore, TpPa-SO3Na exhibits high selectivity for the uptake of dye molecules on the basis of the differences in charge and molecular size. This work explored functional designs and green manufacturing of anionic COFs for removal of hydrophilic organic pollutants.
The effective removal of organic pollutants in wastewater is a key environmental challenge. In this work, an anioniccovalent organic framework (named TpPa-SO3Na) was synthesized through a green two-in-one synthesis strategy with autocatalyticimine formation. The slowly generated acetic acid as a catalyst is favorable to sustain the reversibility of the covalent organic framework (COF) formation reaction and improve the crystallinity of TpPa-SO3Na. TpPa-SO3Naconsists of a homogeneous distribution of sulfonate groups to produce negatively charged regular channels. The strong electrostatic and hydrogen-bonding interactions between the sulfonate groups anchored in the nanochannels and the amine groups in organic pollutants improve the adsorption selectivity and capacity. These structures allow a high degree of control over adsorption processes to boost the adsorption kinetics and improve selective separation. TpPa-SO3Na exhibits ultrafast adsorption (<1 min) of cationic antibiotics and dyes (average over 95%). Furthermore, TpPa-SO3Na exhibits high selectivity for the uptake of dye molecules on the basis of the differences in charge and molecular size. This work explored functional designs and green manufacturing of anionicCOFs for removal of hydrophilic organic pollutants.
With
the development of the chemical industry, water resources
are seriously contaminated by organic pollutants such as antibiotics
and synthetic dyes.[1,2] Illegal discharges of antibiotics
increased bioaccumulation,[3,4] and dyes might even
transform into carcinogenic and mutagenic agents.[5,6] Adsorption
is an effective technique to remove organic pollutants from water
owing to its excellent separation efficiency and reusability.[7]To achieve high adsorption capacity, rapid
kinetics, and outstanding
selectivity, many functional materials have been designed to combat
highly water-soluble antibiotic and artificial dye pollution.[7] A series of conventional materials have been
developed as adsorbents, such as silica and clay.[8] Nonetheless, only organic pollutants interact mainly with
the surface of adsorbents due to their lack of pore structures, limiting
adsorption capacity.[7] Because of their
high surface area, porous organicpolymers offer enhanced adsorption
efficiency compared to conventional materials.[9] Though the correlation between surface area and adsorption capacity
is often documented, the inner surfaces of open-ended polymers were
found to be partially ineffective due to the irregular pore structures
and lack of functional groups, compromising binding capacity.[7]Metal–organic frameworks (MOFs)
were used to mitigate this
problem due to added functional groups and uniform pore structures,
which facilitate the interaction between adsorbents and targeted organic
pollutants.[10] However, MOFs with a relative
density of functional groups and interpenetrating channels greatly
affected site accessibility, thus resulting in a low adsorption affinity
and a relatively long adsorption equilibrium time.[7,11] Therefore,
it is highly desirable to develop new types of materials to overcome
these limitations.Covalent organic frameworks (COFs) are a
new type of porous crystalline
material that integrate organic building blocks into predetermined
nanometric networks.[12] The functionalization
of COFs with specific functional groups on the organic building blocks
can efficiently tune their adsorption properties. In two-dimensional
(2D) COFs, the layered structures stack via π–π
interactions and form regular channels running along the stacking
direction. Thus, the functional groups can be uniformly distributed
on the surface, and the density of functional groups can be precisely
controlled.[13] AnionicCOFs, which consist
of negatively charged active sites and well-ordered one-dimensional
(1D) nanochannels, are a subtype of COF with extraordinary functions.[14] However, the introduction of ionic sites on
the pore walls needs ionic building blocks, which prevents π–π
stacks because of strong charge repulsion, usually leading to low
porosity and crystallinity.[15] Homogeneous
negatively charged groups in the channels would generate a strong
interaction for cationic organic pollutants, improving the binding
capacity.[16,17] Compared with polymers and MOFs, well-defined
hydrophilic 1D channels of COFs are favorable for the diffusion of
highly water-soluble antibiotics and artificial dyes, increasing the
access of pollutants to the surface of channels.[7,13] Negatively
charged 1D nanochannels will generate ionic interfaces to enhance
the accessibility of the anionic sites, allowing high adsorption kinetics
and selective adsorption.[17] These traits
suggest that anionicCOFs can be pertinently designed for highly efficient
removal of hydrophilic antibiotics and artificial dyes.In this
study, a two-in-one strategy was developed for the synthesis
of sulfonate-based covalent organic frameworks (denoted as TpPa-SO3Na) for trapping antibiotics and synthetic dyes. Sulfonates
were homogeneously introduced onto the surface of 2D COFs. The negatively
charged site can provide strong electrostatic and hydrogen-bonding
affinity, which can promote the interaction between cationic pollutants
and the anionic sites. Hydrophilicchannels are favorable for the
diffusion of water-soluble pollutants. Meanwhile, the ionic interface
on the 1D channel walls helps to expose the active sulfonate sites
and promotes adsorption efficiency and selectivity. TpPa-SO3Na shows ultrafast adsorption kinetics with high removal efficiency
(99.0%) in 1 min toward antibiotics such as norfloxacin (NOR). Owing
to its charged nanochannels, TpPa-SO3Na exhibits high selectivity
to separate cationic dye molecules owing to the differences in molecular
size and charge. These results reveal the broad application of the
anionicCOF to the treatment of organic pollutants.
Results and Discussion
Synthesis and Characterization
The
anionicCOF was prepared by condensing aldehydes and amines at an
elevated temperature in a mixture of organic solvents and aqueous
acetic acid. Acetic acid was used to catalyze imine formation.[13] The obtained COF was reacted with an organic
salt to exchange its protons with metal ions.[18] A tedious preparation process, long reaction time, and strong acid
solution are necessary in the entire COF formation reaction. A relatively
high concentration of acid in the reaction can damage the reversible
reaction of diamines on the linkers, resulting in irregular nanochannels.[19] However, the networks transform into crystalline
COFs, and this transformation would be hindered if acetic acid is
not present.[20] As mentioned earlier, the
COF was controlled using different amounts of AcOH.[21] The slowly generated acetic acid was favorable to sustain
the COF formation reaction and improve crystallinity.[13] In our work, a two-in-one synthesis strategy was used for
the synthesis of TpPa-SO3Na as illustrated in Figure . First, the sulfone
acid group of Pa-SO3H reacted with NaOAc to exchange its
proton with Na+. The slowly generated acetic acidcatalyzed
imine formation without additional catalysts,[22] which ensured a low concentration of acid catalyst in the synthetic
process. The low concentration of acetic acid pushed the equilibrium
toward COF formation and slowed down the entire reaction, improving
self-healing and error correction during the crystallization[19] Meanwhile, the two-in-one synthesis strategy
was used for the synthesis of anionicTpPa-SO3Na, avoiding
the traditional synthesis procedures that are time- and solvent-consuming.
Thus, the dual role of NaOAc is very intriguing as both an ion exchanger
and a catalyst in the reaction. The cation exchange reaction was quantitatively
confirmed by the inductively coupled plasma optical emission spectroscopy
(ICP-OES) measurement of the Na content (7.32 wt % (calcd) and 7.25
wt % (found), Table S1). The sulfonate
groups in TpPa-SO3Na work as anionic sites for efficient
adsorption of cationic antibiotics and artificial dyes, which also
provide sites for charge-assisted hydrogen bonds to further improve
the affinity.
Figure 1
(a) Preparation scheme of TpPa-SO3Na and TpPa;
(b) graphic
view of the eclipsed AA stacking structure of TpPa-SO3Na
(top view); and (c) stacking model showing individual layers (side
view) (blue, N; red, O; yellow, S; purple, Na; gray, C; white, H).
(a) Preparation scheme of TpPa-SO3Na and TpPa;
(b) graphic
view of the eclipsed AA stacking structure of TpPa-SO3Na
(top view); and (c) stacking model showing individual layers (side
view) (blue, N; red, O; yellow, S; purple, Na; gray, C; white, H).The structure of TpPa-SO3Na was analyzed
based on the
powder X-ray diffraction (PXRD) pattern (Figure a). The highly prominent PXRD peak at 4.8°
can be attributed to the {100} facet of a regularly ordered lattice,
which is arranged in a hexagonal orientation.[23,24] The sharp diffraction peaks indicate the successful formation of
the imine-linked COF, and the acetic acid formed in the two-in-one
step reaction is an effective catalyst for the COF. Noticeably, the
slowly generated acetic acid is favorable to sustain the reversibility
of the COF formation reaction and improve the crystallinity of TpPa-SO3Na (Figure S1). Additionally, the
broad peak at 27.0° assigned to the {001} facet further confirms
the formation of 2D COFs in a crystalline and π–π
stacking form.[24,25] Intramolecular hydrogen bonds
between the sulfonate units and imine group on the frame can avoid
disturbing the π–π stacking between the sheets,
forming a highly crystalline structure.[26] The PXRD profile obtained from the Pawley refinement of the AA stacking
mode (Figure a, red
curve) agrees well with the experimental result. The average interplanar
distance was estimated to be 3.4 Å. The AA-stacking mode provides
1D channel, allowing adjacent alignment of sulfonate. This spatial
structure efficiently boosted the accessibility of the anionic sites
by cationic pollutants.[17,18] Based on the facts
presented above, superior anionicCOFs have been synthesized using
the two-in-one synthesis strategy. The method also has the advantages
of simplified operation and green environmental protection.
Figure 2
(a) Comparison
of PXRD profiles of TpPa-SO3Na between
the experimental sample (black), the simulated AA eclipsed stacking
modes (red), and the simulated AB staggered stacking modes; (b) Fourier-transform
infrared spectroscopy (FTIR) spectra of Tp (black), Pa-SO3H (red), and TpPa-SO3Na (blue); (c) pore size distribution
of TpPa-SO3Na; and (d) stability tests of TpPa-SO3Na in 6 M NaOH, 6 M HCl, and boiling water.
(a) Comparison
of PXRD profiles of TpPa-SO3Na between
the experimental sample (black), the simulated AA eclipsed stacking
modes (red), and the simulated AB staggered stacking modes; (b) Fourier-transform
infrared spectroscopy (FTIR) spectra of Tp (black), Pa-SO3H (red), and TpPa-SO3Na (blue); (c) pore size distribution
of TpPa-SO3Na; and (d) stability tests of TpPa-SO3Na in 6 M NaOH, 6 M HCl, and boiling water.The typical FTIR spectra of TpPa-SO3Na are shown in Figure b. The keto tautomer
is confirmed by the presence of a new C=C peak at 1576 cm–1 and the absence of C=N stretching peaks (1620
cm–1).[27] The aromaticC=C peak appears at 1438 cm–1, and the newly
formed C–N bonds of the keto tautomer in the COF appear at
1231 cm–1.[24,27] The formation of the
−SO3– unit is further verified
by the peak at 1005 cm–1.[22] The Brunauer–Emmett–Teller (BET) surface area for
TpPa-SO3Na is 236 m2 g–1 (Figure S2). Because of the high surface-to-volume
ratio of the COF and internal sulfonate groups, the surface charges
may generate abundant adsorption sites for charged organiccompounds
and enhance the adsorption capacity and affinity. The pore diameter
is concentrated at ∼15.2 Å (Figure c), and the narrow pore size reflects the
successful preparation of the crystalline network.[28] Owing to the presence of the narrow nanochannels, the internal
sulfonate groups can easily interact with amine groups of antibiotics
and artificial dyes.[17] SEM images of TpPa-SO3Na are shown in Figure S3. The
material exhibits a nanofiber morphology. The SEM image with a 2 μm
scale plate shows a randomly shaped aggregate composed of many macroscopic
fibers.
Stability Test
To investigate the
stability of TpPa-SO3Na, samples were dispersed in different
solutions, including NaOH (6 M) and aqueous HCl (6 M) and boiling
water solutions for 7 days. Surprisingly, all of the TpPa-SO3Na samples exhibited intense PXRD patterns without a change in the
peak position and intensity, indicating that the high stability is
retained under these extreme conditions (Figure d). Compared with traditional imine-linked
COFs,[13] irreversible enol-to-keto tautomerization
and the hydrogen-bonding interaction can further enhance the chemical
stability.[29,30] The thermogravimetric analysis
(TGA) curves exhibit two weight losses in the studied temperature
range. The initial weight loss was caused by the removal of the occluded
solvent molecules at 50–120 °C, and the second weight
loss was derived from the decomposition of the sulfonate groups at
300 °C (Figure S4). This suggested
that the obtained COF exhibits superior stability,[23] which is important for its practical applications.
Adsorption Kinetics
Taking into account
the porosity and high density of sulfonate sites within the ordered
pore channels in TpPa-SO3Na, its adsorption performance
toward the positively charged amine group of antibiotics and artificial
dyes was further investigated. The effects of adsorbent dosage on
the removal efficiency were investigated. As shown in Figure S5, it can be seen that the removal efficiency
increases with the increase of TpPa-SO3Na dose as a result
of increased available adsorption active sites and surface area. Taking
the adsorption efficiency and cost into consideration, the optimum
dosage of TpPa-SO3Na was used in the subsequent investigations.
Typically, TpPa-SO3Na was mixed with pollutant solutions
at 10 mg L–1 under pH 7.0. As shown in Figure a, TpPa-SO3Na showed ultrafast adsorption kinetics, and it efficiently removed
most of the pollutants such as NOR (∼99%), ciprofloxacin (CIP)
(∼98%), enrofloxacin (ENR) (∼98%), tetracycline hydrochloride
(TC) (∼91%), methylene blue (MB) (∼96%), and crystal
violet (CV) (∼89%) within 1 min. It is worth noting that almost
90% of its maximum adsorption capacity was reached within 30 s. This
suggests that TpPa-SO3Na has excellent affinity toward
the antibiotics and dyes. The adsorption kinetics studies show that
the adsorption process fits well with the pseudo-second-order model
(Figure b), confirming
that the interactions between TpPa-SO3Na and pollutants
are mainly controlled by chemical processes. The pseudo-second-order
rate constants (kobs) of NOR, CIP, ENR,
TC, MB, and CV adsorption are 4.78, 3.53, 3.37, 1.64, 2.18, and 1.08
g mg–1 min–1, respectively (Figure S6). NOR was selected as one of the representative
pollutants. The removal rate constant is higher than those of most
reported porous pollutant removers (Table S2). Compared to inorganic porous materials such as clays and zeolites
with low inherent porosity[8] and amorphous
organic porous polymers with an irregular pore structure, COFs are
desirable materials that provide a specific pore structure and interactions
according to the contaminant of choice.[13] The excellent removal rate constant should be attributed to TpPa-SO3Nacontaining nanochannels and compact sulfonate sites, significantly
improving the accessibility to these cationic pollutants rather than
the amorphous activated carbon with an uneven pore distribution and
undefined functional groups.[31] The 1D nanochannel
with ionic interfaces in TpPa-SO3Na is favorable for diffusion
of the antibiotics and dyes compared with the self-interpenetrated
nanochannel of MOFs.[7] The ultrafast adsorption
kinetics is the key to remarkably reduce the exposure time, especially
under extreme conditions, improving the configurable preservation
during sewage disposal.
Figure 3
(a) Adsorption kinetics of TpPa-SO3Na toward different
organic pollutants (initial concentration was 10 mg L–1; the dosage of TpPa-SO3Na was 0.6 g L–1 (for NOR, CIP, and ENR), 0.8 g L–1 (for MB and
CV), and 1.0 g L–1 (for TC); solution volume 100
mL; and pH 7.0). (b) Pseudo-second-order kinetics of adsorption of
different organic pollutants. (c) Adsorption isotherms for different
organic pollutants on TpPa-SO3Na (initial concentration
was 10, 20, 30, 50, 100, 200, 300, and 500 mg L–1; the dosage of TpPa-SO3Na was 0.6 g L–1 (for NOR, CIP, and ENR), 0.8 g L–1 (for MB and
CV), and 1.0 g L–1 (for TC); solution volume 40
mL; and pH 7.0). (d) Removal efficiency and ζ-potential of TpPa-SO3Na toward NOR at different pH conditions.
(a) Adsorption kinetics of TpPa-SO3Na toward different
organic pollutants (initial concentration was 10 mg L–1; the dosage of TpPa-SO3Na was 0.6 g L–1 (for NOR, CIP, and ENR), 0.8 g L–1 (for MB and
CV), and 1.0 g L–1 (for TC); solution volume 100
mL; and pH 7.0). (b) Pseudo-second-order kinetics of adsorption of
different organic pollutants. (c) Adsorption isotherms for different
organic pollutants on TpPa-SO3Na (initial concentration
was 10, 20, 30, 50, 100, 200, 300, and 500 mg L–1; the dosage of TpPa-SO3Na was 0.6 g L–1 (for NOR, CIP, and ENR), 0.8 g L–1 (for MB and
CV), and 1.0 g L–1 (for TC); solution volume 40
mL; and pH 7.0). (d) Removal efficiency and ζ-potential of TpPa-SO3Na toward NOR at different pH conditions.
Adsorption Isotherms
The adsorption
isotherms of TpPa-SO3Na fitted well with the Langmuir model
(Figure c and Table S3) with a correlation coefficient R2 of greater than 0.99, suggesting that the uniform and monolayer
adsorption isotherm model is more propitious to the adsorption of
NOR. It is worth noting that 2D COFs tend to form uniform and controllable
1D nanochannels by the suitable stacking of monolayers.[17] The monolayer adsorption model is consistent
with the structural characteristics of stacking of monolayers. The
distribution coefficient value (Kd) can
be applied to further evaluate the interaction between adsorbents
and pollutants. Generally, a material with a Kd value of more than 104 mL g–1 can be regarded as an excellent adsorbent.[32] The Kd of TpPa-SO3Na toward
NOR is 1.7 × 107 mL g–1 in 10 min
(Table S4). The saturation sorption capacity
toward NOR was 653 mg g–1 at room temperature (Table S3), higher than those of other reported
adsorbents (Table S2). The self-adsorption
behavior was also tested under dark conditions. Spectroscopic monitoring
of NOR solution at different intervals indicated that the concentration
of NOR did not exhibit any noticeable change under dark conditions.
It is suggested that the self-adsorption behaviors did not significantly
impact the adsorption process of NOR on TpPa-SO3Na (Figure S7). Its excellent adsorption capacity
is mainly attributed to the strong affinity interaction between antibiotics
and the sulfonate groups of TpPa-SO3Na. In addition, the
sufficiently long 1D nanochannels provide more active sites.[33]
Effect of pH and Reusability
The
capture of antibiotics and dyes over a wide pH range is greatly needed
for actual sewage treatments. The effect of pH on the adsorption capacity
toward NOR was researched from pH 1 to 12 (Figure d). The adsorption efficiencies of TpPa-SO3Na toward NOR were over 90% within a wide pH range of 3–10.
Meanwhile, the ζ-potentials of TpPa-SO3Na at different
pH values were also measured (Figure d). As shown in Figure d, when pH <6.30, the adsorption capacity of NOR
increases with the increase of pH value. The cationicNOR is mainly
formed in the solution in this pH range (Figure S8a,b); therefore, the increase in adsorption capacity is attributed
to the sharp increase in the negative potential of TpPa-SO3Na. In the range of 6.30 < pH < 8.38,
the increase in the negative potential of TpPa-SO3Na slows
down, while the content of anionicNOR increases gradually with pH.
The adsorption capacity reaches the highest value at pH about 7.0.
When pH > 8.38, anionicNOR as the main form exists
in the solution (Figure S8a,b). Compared
with the expected result,[34] increased adsorption
of NOR was obtained in spite of an unfavorable electrostatic interaction
in the range. Hydrogen bonding may be one of the potential mechanisms
for enhanced adsorption of NOR. The N–H group of NOR is considered
to be able to serve as a very strong hydrogen donor for hydrogen bonds.
Therefore, NORcan interact efficiently via hydrogen bonds with adsorbents
having a hydrogen acceptor.[26] As TpPa-SO3Na has ultrafast adsorption kinetics, a long-time adsorption
and desorption experiment was implemented to determine the stability
of TpPa-SO3Na. After being treated with NOR solutions for
12 h, the desorption of NOR from TpPa-SO3Na was studied
using an HCl solution (0.1 M). No obvious decrease in the adsorption
capacity of TpPa-SO3Na for NOR (100 mg L–1) was observed even after five adsorption–desorption cycles
(Figure S9). The PXRD result also showed
that there was no change in the peak pattern after five cycles of
reuse compared with the fresh one (Figure S10). Furthermore, all of the characteristic FTIR peaks remained the
same after recycling, and no extra peaks were observed (Figure S11). These excellent properties endow
TpPa-SO3Na with great capability in the removal of antibiotics
and artificial dyes.
Adsorption Mechanism
The adsorption
property of 2D COFs is actuated by “horror vacui” and
guided by the diffusion and accessibility of the pores.[35] The nanochannel diameter of TpPa-SO3Na is concentrated at 15.2 Å, which is large enough to allow
entry of NOR inside the nanopore and its interaction with the sulfonate
groups.[36,37] To further verify the possibility of entering
the pore during adsorption, N2 sorption measurements at
77 K after NOR adsorption were made. The experiment results showed
a decrease in surface area to 49 m2 g–1, corresponding to a drop of 80%, which could indicate that NOR is
adsorbed within the pores of TpPa-SO3Na (Figure S12). The ζ-potential of TpPa-SO3Na
is negative at pH 1–12 owing to the abundant sulfonate groups.[11,38] The negatively charged surface of TpPa-SO3Na helped in
capturing the positively charged antibiotics and artificial dyes.
Coupled with the pore structure, the adsorption kinetics and capacity
are also influenced by surface charge.[39] TpPa-SO3Na shows a high negative ζ-potential (−22.2
mV) before NOR adsorption, suggesting that it prefers to adsorb cationicNOR. The ζ-potential of the NOR-loaded adsorbents was also tested
(Figure S13). The results indicate an improvement
in the ζ-potential. The adsorption of cationicNOR was promoted
by electrostatic attraction. Negatively charged groups may act in
concert with pore size to facilitate adsorption of cationicNOR into
the channels of TpPa-SO3Na. The typical FTIR spectra of
TpPa-SO3Na before and after adsorption of NOR are shown
in Figure S14. The characteristic peaks
for N–O at 1101 cm–1 and the S–O stretching
vibration at 1024 cm–1 shifted to 1026 cm–1 are attributed to the occurrence of the strong reaction between
sulfonate and amine.[24] This interaction
was further confirmed by X-ray photoelectron spectroscopy (XPS). The
spectra of N 1s XPS analysis are consistent with the presence of the
electrostatic attraction process as the new peak of N–O (401.9
eV) appeared in TpPa-SO3Na-NOR (Figure a,b). Moreover, the O 1s core level of S–O
(533.1 eV) shifted to a higher binding energy (533.5 eV) (Figure c,d), corroborating
the results obtained from the XPS of N 1s. The sulfonate groups are
aligned on the nanochannel’s internal surface, and the open
nanochannel effectively exposes the active sites, increasing the accessibility
to NOR and enhancing the adsorption efficiency toward NOR.[17,18] Integration of anionic sites into the nanochannel has generated
anionic interfaces to accelerate the diffusion of charged NOR in the
direction of the channel.[40,41]
Figure 4
N 1s of TpPa-SO3Na (a) and TpPa-SO3Na-NOR
(b) and O 1s of TpPa-SO3Na (c) and TpPa-SO3Na-NOR
(d).
N 1s of TpPa-SO3Na (a) and TpPa-SO3Na-NOR
(b) and O 1s of TpPa-SO3Na (c) and TpPa-SO3Na-NOR
(d).Nanochannels of 2D COFscan offer
efficient pathways for the transport
of ions and molecules. With decreasing size, the interfacial wettability
within nanochannels plays a significant role in the efficient transportation
of various molecules.[42,43] Modulating hydrogen bonding is
a powerful strategy to increase the interior wettability of nanochannels.[44,45]Figure S14 shows the FTIR spectra of
TpPa-SO3Na before and after adsorption of NOR. The characteristic
peak at 1237 cm–1 may be attributed to the symmetric
stretching vibration of O=S. Following the adsorption of NOR, the
peak at 1237 cm–1 shifted up to 1232 cm–1, which signified that a hydrogen-bonding interaction occurred between
N-H and O=S. This interaction was further confirmed by XPS. The O
1s core level of S=O...H...N (532.6 eV) shifted
to a lower binding energy (532.3 eV),[26] verifying that the sulfonate sites were the binding sites toward
the amine group in NOR by charge-assisted hydrogen-bonding attraction
(Figure c,d).[23,24] To further confirm the hydrogen-bonding interaction, NOR was added
to TpPa-SO3Na in the presence of 6 M carbamide (Figure a).[46] The adsorption capacity toward NOR significantly declined
with the addition of carbamides in the pollutant solution. The disruptive
effect of carbamide further confirms the hydrogen-bonding mechanism.
This shows that negative charge-assisted and enhanced H-bonding attraction
makes the hydrophilic nanochannels robust. With the hydrophilic modification
of nanochannels, the wettability of the material was effectively improved,
facilitating diffusion in the direction of the channel.[38,42]
Figure 5
(a)
Competitive adsorption of NOR and carbamide on TpPa-SO3Na (initial concentration was 50 mg L–1;
carbamide 6M; TpPa-SO3Na 0.6 g L–1; and
pH 7.0). (b) adsorption kinetics of TpPa-SO3Na and TpPa
toward NOR (initial concentration was 10 mg L–1;
COF 0.6 g L–1; and pH 7.0).
(a)
Competitive adsorption of NOR and carbamide on TpPa-SO3Na (initial concentration was 50 mg L–1;
carbamide 6M; TpPa-SO3Na 0.6 g L–1; and
pH 7.0). (b) adsorption kinetics of TpPa-SO3Na and TpPa
toward NOR (initial concentration was 10 mg L–1;
COF 0.6 g L–1; and pH 7.0).In addition, to investigate the role of charges in the adsorption,
4-nitrophenol (4-NP) containing only a hydrogen donor (hydroxyl) was
tested. As demonstrated in Figure S15,
4-NP was hardly adsorbed by TpPa-SO3Na, which showed as
low as 2% of adsorption efficiency even after 2 h, implying that the
individual hydrogen-bonding effect was not as dominant as expected
in the adsorption process because of the lack of charged binding sites.
The synergetic electrostatic interaction and charge-assisted hydrogen
bonds strongly enhance the binding ability between TpPa-SO3Na and NOR.To get a better idea of the effect of 1D negatively
charged nanochannels
in NOR removal, the TpPa COF without the sulfonate group was prepared
for comparison (Figure S16).[29] Similarly, 10 mg of TpPa COF was immersed in
20 mL of NOR solution (10 mg L–1), and we could
find that only 1% of NOR were absorbed, suggesting that the absence
of sulfonate sites in the network led to the poor NOR uptake efficiency
(Figure b). TpPa-SO3Na with uniform 1D negatively charged channels exhibited a
higher adsorption efficiency than the TpPa COF. The above results
show that the negatively charged nanochannels are indispensable for
fast and efficient positively charged organic pollutant adsorption
from water.[17,42]On the basis of these FTIR,
XPS, and confirmatory adsorption experiments,
a possible adsorption mechanism for the removal of NOR is proposed
and described in Figure . The synergetic electrostatic interaction and charge-assisted hydrogen
bonds strongly enhance the binding ability between TpPa-SO3Na and NOR.
Figure 6
Proposed mechanism for the removal of NOR by TpPa-SO3Na.
Proposed mechanism for the removal of NOR by TpPa-SO3Na.
Selective
Adsorption and Separation
The adsorption and separation of
hydrophilic organic pollutants are
critically important due to serious environmental problems. The adsorption
and separation are highly dependent on the porous structures, that
is, the pore size and pore environment.[47−51] Compared with other separation materials, due to
the capacity to regulate their pore size and chemical functionalities,
MOFs and COFs have received increasing attention in adsorption and
separation. MOFs have limitations such as poor chemical stability
and interpenetrating channels in the separation of dyes.[47] COFs overcame the limitations of chemical instability
in aqueous and organic and strongly acidic or alkaline media compared
with MOFs.[13] Meanwhile, in 2D COFs, the
1D pores are open and free of interpenetration.[37,50] Owing to the availability of functional groups and design of pore
size, a pore-surface engineering and skeleton design strategy offers
a powerful tool to broaden the diversity of porous structures of COFs
in adsorption and separation fields, thus enabling the screening of
impact of porous structures on specific organic pollutants.[37,47,50] Effective selective separation
of specific molecules can be achieved based on different interactions
between molecules and functional nanochannels. To demonstrate the
charge- selective separation, we treated TpPa-SO3Na with
various dye molecules, including positively charged MB, positively
and negatively charged methyl red (MR), neutral p-benzoquinone (p-BQ), and negatively charged calcein
(CAL) (Figure S17). As shown in Figure a, MB was completely
adsorbed in 10 min by TpPa-SO3Na, MR was removed by 85%
(Figure b), while
CAL was hardly adsorbed at all even after 2 h (Figure c). Neutral p-BQ was also
hardly adsorbed by TpPa-SO3Na, which showed as low as 2%
of separation efficiency even after 2 h (Figure S18). Although p-BQ molecules were small enough
for entry, the uptake capacity of neutral p-BQ was
low because of the absence of electrostatic interaction between TpPa-SO3Na and p-BQ. Therefore, electrostatic interactions
appeared to dominate the adsorption, which was conducive to specific
target pollutant removal.
Figure 7
UV absorption spectra of aqueous MB (a), MR
(b), and CAL (c) treated
with TpPa-SO3Na at different intervals. UV absorption spectra
of the dye mixture solution (d) before and after passing through a
TpPa-SO3Na-packed column. The inset shows the photograph
of the tested column.
UV absorption spectra of aqueous MB (a), MR
(b), and CAL (c) treated
with TpPa-SO3Na at different intervals. UV absorption spectra
of the dye mixture solution (d) before and after passing through a
TpPa-SO3Na-packed column. The inset shows the photograph
of the tested column.To further test the charge
separation effect, TpPa-SO3Na was utilized in a column
experiment. The decrease of the UV spectra
after the column treatment suggested that the dye mixture was completely
separated with a sharp decrease of MBconcentration (Figure d). The concentration change
was 2 and 100% for CAL and MB (Figure d), respectively, suggesting the effective separation
of the two molecules. However, when treated with TpPa, a poor separation
selectivity toward CAL and MB was observed under the same conditions
(Figure S19), suggesting that the negatively
charged nanochannels play a critical role. Experiments show that the
pore environment is essential to cause interactions between pores
and guest molecules because the surfaces form the interface that interacts
with guests.We then treated TpPa-SO3Na with dye
molecules including
CV and Alcian blue 8GX (AB8GX) (Figure S17), which have different molecular dimensions and are positively charged,
to further confirm the confinement effect of charged nanochannels
in selective separation. MB was completely removed in 10 min with
TpPa-SO3Na, while CV and AB8GX did not exhibit a noticeable
change after 95% and 16% adsorption, respectively, for the same time
period (Figure a,b).
The size-dependent separation of dye molecules can be attributed to
the effect of the nanochannel structure. In TpPa-SO3Na,
the 1D pores are open and accessible only from the top and bottom
sides of the 2D layers, and they are free of interpenetration. TpPa-SO3Na has a uniform and persistent hexagonal shape that enables
accurate size recognition.[37] As displayed
in Table S5, the size of AB8GX is larger
than the accessible nanochannel diameter of TpPa-SO3Na,
but the size of MB fits the nanochannel diameter of TpPa-SO3Na, leading to size-exclusive behaviors. The molecular size of CV
is close to the nanochannel width of TpPa-SO3Na. The positive
charge of CV assists considerably in the selective adsorption when
the molecular size approaches the nanochannel diameter of the charged
COF. Results show that selective adsorption can be specifically tuned
by introducing functional sites and modifying their pore size. This
would generate a high degree of control over host–guest interactions,
improving the adsorptive selectivity and promoting the separation
processes.
Figure 8
UV absorption spectra of aqueous CV (a) and AB8GX (b) treated with
TpPa-SO3Na at different intervals.
UV absorption spectra of aqueous CV (a) and AB8GX (b) treated with
TpPa-SO3Na at different intervals.
Conclusions
In summary, an anioniccovalent
organic framework was synthesized
through a green two-in-one synthesis strategy with autocatalyticimine
formation. The as-prepared TpPa-SO3Na has ultrafast adsorption
kinetics and allows molecular separation on the basis of sulfonate
functionalization of nanochannels. Our work can greatly promote the
green manufacturing and investigation of anionicCOFs with different
kinds of binding sites to remove hydrophilic pollutants, further broadening
the application of COFs as flexible and tunable materials for wastewater
treatment and real-time separation, to cope with challenges of sustainability
in the field of chemical and environmental protection technology.
Experimental Section
Synthesis of TpPa-SO3Na
Briefly, 0.3 mmol of 1,3,5-triformylphloroglucinol
(Tp) and 0.45
mmol of 2,5-diaminobenzenesulfonic acid (Pa-SO3H) were
added to 5 mL of mesitylene and 1,4-dioxane mixture (4:1, v/v). Subsequently,
the mixture was ultrasonicated for 20 min to form a homogeneous solution.
Afterward, 1 mL of 5 M NaOAc was added to the mixture dropwise at
room temperature, and then it was flash-frozen under liquid N2.The reaction tube was sealed and then heated under 120 °C
for 72 h. Precipitates with red color were successfully obtained and
then washed successively with water and tetrahydrofuran five times.
Finally, the product was dried at 120 °C under vacuum for 12
h to yield TpPa-SO3Na as a red-colored powder (90% yields).
Synthesis of TpPa followed the same process, except that 2,5-diaminobenzenesulfonic
acid was replaced by p-phenylenediamine.
Adsorption
Experiments
Norfloxacin
(NOR), ciprofloxacin (CIP), enrofloxacin (ENR), tetracycline hydrochloride
(TC), methylene blue (MB), and crystal violet (CV) were dissolved
in deionized water to obtain the stock solutions (0.5 g L–1). To study the adsorption kinetics, 60 mg of TpPa-SO3Na was added to 100 mL of the target pollutant solution (10 mg L–1) at pH 7.0 for 30 min. After being shaken for different
time intervals, 0.5 mL of the solution sample was collected and immediately
filtered by a 0.45 μm filter. The concentrations of NOR, CIP,
ENR, and TC were determined by high-performance liquid chromatography
(HPLC) with an ultraviolet (UV) detector. The concentrations of MB
and CV were measured by UV–vis absorption spectroscopy. During
the adsorption processes, the pH value of the solution was adjusted
using NaOH or HCl solutions (0.1 M). For the regeneration experiments,
the used adsorbents were added into the HCl solution (0.1 M). After
desorbing for 10 min, the adsorbent was centrifuged with a speed of
8000 rpm for 5 min. Finally, the adsorbent was washed with deionized
water and dried for reuse.
Authors: René P Schwarzenbach; Beate I Escher; Kathrin Fenner; Thomas B Hofstetter; C Annette Johnson; Urs von Gunten; Bernhard Wehrli Journal: Science Date: 2006-08-25 Impact factor: 47.728
Authors: Dusan Licina; Glenn C Morrison; Gabriel Bekö; Charles J Weschler; William W Nazaroff Journal: Environ Sci Technol Date: 2019-05-09 Impact factor: 9.028