Catalytic deoxygenation of C18 fatty acid over supported metal Ni catalysts promoted by the basic sites of ZnAl2O4 spinel phase

Article abstract of DOI:10.1039/c8cy02027b

Highly active Zn-Al composite oxides were synthesized via a hydrothermal process followed by thermal treatment and were used as supports to prepare Ni-based hydrogenation catalysts for catalytic deoxygenation of oleic acid, stearic acid, and 1-octadecanol. The results showed that increasing the temperature of hydrothermal synthesis changed the morphology of the Zn-Al composite oxides from sheet-like structures to spheroidal structures. High hydrothermal synthesis temperatures enhanced the interaction between Zn and Al atoms, resulting in more ZnAl₂O₄ spinel phase. This phase not only improved the chemical stability of the support but also supplied strong basic sites which efficiently inhibited the formation of by-products and increased the yield of heptadecane in the catalytic deoxygenation of oleic acid. Stearic acid and 1-octadecanol could be readily transformed to alkanes in the presence of metallic Ni and ZnAl₂O₄ phase. Decarbonylation of the octadecanal intermediate and dehydrogenation of 1-octadecanol were key reaction pathways to produce heptadecane, in which decarbonylation was catalyzed by metallic Ni, while the dehydrogenation was attributed to synergistic catalysis between metallic Ni and the strong basic sites of the support. Individual metallic Ni only catalyzed the cleavage of C-H bonds but did not affect the O-H bond of 1-octadecanol.

Full text of DOI:10.1039/c8cy02027b

Journal of
Materials Chemistry A
View Article Online
View Journal
PAPER
Hyperbranched poly(ionic liquid) functionalized
poly(ether sulfone) membranes as healable
antifouling coatings for osmotic power generation†
Cite this: DOI: 10.1039/c8ta10484k
ab
ab
Jia Le Li,ab Yu Zhang,c Sui Zhang,c Minqiu Liu,c Xue Li
and Tao Cai
*
*
Membrane fouling and membrane deterioration are two major concerns since they greatly worsen
membrane performance in pressure retarded osmosis (PRO) and shorten the membrane lifetime. Herein,
inspired by biological systems where microdamage induces an autonomous repair process, intrinsically
healable poly(ionic liquid) (PIL) coating layers with sufficiently high mobility from short-distance
electrostatic interactions have been fabricated by sequential immobilization of sulfonated hyperbranched
polyglycerol (hbSPG) and quaternized polyethylenimine (QPEI) on polydopamine (PDA) pretreated
poly(ether sulfone) (PES) membranes. XPS, SEM, and AFM results confirmed the successful incorporation
of a PIL polymer pair onto the PES supports. Zeta potential analysis validated the charge alteration by
electrostatic conjugation of the QPEI polymer onto the hbSPG modified surface. The antifouling and
self-healing characteristics of the resultant PES-g-hbSPG-QPEI hollow fiber membranes were
demonstrated by the excellent anti-protein adsorption behaviors and improved antibacterial
performances on both non-aged and aged samples that were soaked in real municipal wastewater
(WWRe) for two months. In PRO tests, the pristine PES membranes were contaminated seriously under
high pressure operation, leading to a significant flux drop of 60%. In comparison, a flux reduction of 30%
and a recovery rate of 98% after backflushing and hydraulic pressure impulsion on the non-aged
membrane and a flux reduction of 40% on the aged membranes were observed. The intrinsically
healable and antifouling PIL coating layers exhibit great potential of the developed strategy for the
fabrication of high-performance PRO membranes toward osmotic power generation.
Received 31st October 2018
Accepted 21st December 2018
DOI: 10.1039/c8ta10484k
rsc.li/materials-a
The impurities would readily be stuck in the porous membrane
matrix rather than the dense selective layer, and thereby
aggravate the fouling behavior.17–22 In an effort to boost the
power density of the PRO process, the establishment of multi-
layer coating structures with superior fouling resistance, self-
healing characteristics and high osmotic pressure toleration
on the membranes is of great signicance.
Most coating layers are so and susceptible to mechanical
damage especially for membranes under great pressure in
practical usage. Continuous attacks and corrosion by inorganic
substances, organic compounds and microorganisms in feed
solutions result in occasional scratches or partial exfoliation
on the coating layers and thereby disclose the underlying
membrane matrix. With regard to the antifouling coating layers,
myriad metabolic substances including corrosive ammonia and
organic acids excreted by the attached bacteria during their
reproductive processes further destroy the already defective
coating layers and eventually deprive them of their protective
functions.23–25 Inspired by biological systems where micro-
damage triggered an autonomous repair process with
minimum external intervention, self-healing materials with
diverse dynamic bond connections have been developed.26–31
Current self-healing behavior mainly focuses on bulk materials
1. Introduction
Pressure retarded osmosis (PRO) has provided intensive
impetus to the exploration and development of high perfor-
mance membranes for converting clean and sustainable
osmotic power into electricity.1–11 Preliminary experiments
demonstrated that the replacement of deionized (DI) water with
wastewater retentates resulted in the sharp decline of power
density from a maximum of ꢀ27 W mꢁ2 to 0 W mꢁ2 aer a short
operation period.12–16 Although pretreatment processes on feed
solutions brought about some improvements, the accumulation
of contaminants within the porous support and vulnerable
coating inevitably caused severe internal concentration polari-
zation and bacterial biolm formation in long term operations.
aKey Laboratory of Biomedical Polymers of Ministry of Education, College of Chemistry
and Molecular Science, Wuhan University, Wuhan, Hubei 430072, P. R. China. E-mail:
chemcaitao@whu.edu.cn; li.x@whu.edu.cn
bWuhan University Shenzhen Research Institute, Shenzhen, Guangdong 518057, P. R.
China
cDepartment of Chemical
& Biomolecular Engineering, National University of
Singapore, 4 Engineering Drive 4, Kent Ridge 117585, Singapore
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c8ta10484k
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
like hydrogels that are capable of restoring their mechanical
and structural integrity. It has been gradually applied to surface
coatings to sustain preliminarily incorporated functions, such
as antifouling and anticorrosive performances. Differing from
conventional coatings, self-healing coatings capable of detect-
ing internal defects within the coatings and autonomously
repairing themselves to recover their original functions are
acknowledged as a promising strategy to address the increasing
sophisticated fouling problems and are of crucial necessity to
reduce maintenance costs, extend service lifespan and enhance
reliability of the underlying membranes.
Poly(ionic liquid)s (PILs) are an emerging and intriguing
subclass of functional polymers with repeating units that bear
ionic liquid (IL) moieties.32,33 The structural synergy between the
IL component and the macromolecular architecture broadens
the window of physicochemical properties of common poly-
electrolytes and thereby enriches their function and application
spectrum by combining the benecial characteristics from ILs,
such as high ionic conductivity, remarkable chemical, electro-
chemical and thermal stability, structural variability and strong
polarizability, with the intrinsic polymer features including
superior processability and spatial tunability.34–37 Although
there are many self-healing paradigms in nature, it is still
greatly challenging to transform them into engineering mate-
rials. Moreover, in self-healing coatings, the conned mobility
of the polymer brushes within the nanoscale thickness has
2.2 Preparation of the PES-g-hbSPG-QPEI or PES-g-hbSPG-
PDDA hollow ber membranes
As shown in Scheme 1, the construction of sandwiched multi-
layer coatings on the PES-TFC dual-layer hollow ber substrates
was carried out in three successive steps. A Tris buffer solution
(0.05 mol Lꢁ1, pH 8.5) containing DA (2.0 mg mLꢁ1), CuSO4
(0.85 mg mLꢁ1) and H2O2 (0.9 mg mLꢁ1) was preliminarily
dosed to the shell channel of the PES-TFC hollow ber
membranes at ambient temperature for 3 h to provide adequate
reaction sites for the subsequent polymer graing. The pre-
treated membranes, denoted as PES–PDA membranes, were
washed with deionized (DI) water for 0.5 h. Aerward, the poly-
dopamine coating solution was replaced by a hbSPG solution
(10 g Lꢁ1) containing TEA (0.1 mol Lꢁ1) and the Michael addition
reaction was allowed to proceed for 12 h. Aer that, the inter-
mediate membranes, denoted as PES-g-hbSPG membranes, were
rinsed with DI water for an additional 0.5 h. For the second
polyelectrolyte layer, a diluted QPEI or PDDA solution (10 g Lꢁ1
)
was dosed to the shell channel of the membranes for another
12 h. The resultant membranes, denoted as PES-g-hbSPG-QPEI or
PES-g-hbSPG-PDDA membranes, were then extracted thoroughly
and stored in DI water to eliminate any leover polymers before
further tests.
2.3 Characterization
hitherto restricted the self-recovery behavior upon mechanical X-ray photoelectron spectroscopy (XPS) analysis was performed
damage because the reversible interaction can be rebuilt only on a Kratos AXIS Ultra DLD spectrometer with an Al Ka X-ray
when the polymers can migrate across the damaged area and source to investigate the surface chemistry evolution of the
come into intimate contact with each other. In this contribu- membranes. A eld emission scanning electronic microscope
tion, upon establishing the role of PILs in membrane modi- (FESEM, JEOL JSM-6700F) was employed to observe the surface
cation, intrinsically healable PIL coating layers with sufficiently morphologies of the resultant membranes. The membrane
high mobility from short-distance electrostatic interactions, specimens were initially cut in liquid nitrogen for sharp cross-
have been constructed by the sequential conjugation of sections and then platinum-coated by a JEOL JFC-1100e Ion
sulfonated hyperbranched polyglycerol (hbSPG) and quater- Sputtering device prior to FESEM observation. The surface
nized polyethylenimine (QPEI) on polydopamine (PDA) pre- charge proles of the modied membranes were evaluated via
treated membrane supports. To the best of our knowledge, the a SurPASS electrokinetic analyzer (Anton Paar GmbH, Austria).
current work is the rst example on the preparation of PIL- PES at sheet membranes were prepared using a similar
based PRO membranes with self-healing characteristics and method to simulate the outer surface charge properties of the
fouling resistant features for osmotic power generation.
hollow ber membranes. The at sheet membrane sample was
rst placed in a 0.01 mol Lꢁ1 NaCl solution of pH 2.5. During
measurement, the NaCl solution was slowly auto-titrated with
0.05 mol Lꢁ1 NaOH to pH 10. The z-potential of the membrane
surface was then measured against the elevated pH values.
Fluorescein isothiocyanate conjugated bovine serum
2. Experimental section
2.1 Materials
Dopamine hydrochloride (DA, 99%), tris(hydroxymethyl)ami- albumin (BSA-FITC) and g-globulin (g-globulin-FITC) were
nomethane (Tris, $99.8%), triethylamine (TEA, $99%), cop- adopted in the protein adsorption experiments. The membrane
per(II) sulfate (CuSO4, $99%), hydrogen peroxide (H2O2, specimens were initially cut to 2 cm in length and positioned in
30 wt%) and poly(diallyldimethylammonium chloride) solu- isotonic phosphate buffered saline (PBS, 0.01 mol Lꢁ1, pH 7.4)
tion (PDDA, Mw < 100 000, 35 wt%) were purchased from containing BSA-FITC or g-globulin-FITC (1.0 mg mLꢁ1) at 37 ꢂC
Sigma-Aldrich and used as received except where mentioned for 2 h. The relative uorescence intensity of the surface-bound
otherwise. Sulfonated hyperbranched polyglycerol (hbSPG, protein amount aer being thoroughly rinsed with PBS was
Mn,NMR ¼ 12 200 g molꢁ1, Mn,GPC ¼ 9800 g molꢁ1, Đ ¼ quantied using a Leica DMLM uorescence microscope (Leica
1.22),38–41 quaternized poly(ethylenimine) (QPEI, linear, Microsystems, Wetzlar, Germany).
Mn,GPC ¼ 18 000 g molꢁ1, Đ ¼ 1.24)42,43 and PES-TFC dual-layer
Two bacterial strains including E. coli (ATCC 25922) and
hollow ber supports were prepared according to the protocols S. aureus (ATCC 25923) were cultivated to investigate the anti-
reported in the literature.
microbial behavior of the poly(ionic liquid) conjugated
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
Scheme 1 Schematic of the procedure for the fabrication of PES-g-hbSPG-QPEI membranes.
membranes. The bacterial suspension was xed at 1 ꢃ 108 testing run, where DI water was dosed to the feed solution. In
colony forming unit (CFU) per mL by diluting overnight between each fouling test run, DI water was fed into the lumen
cultured bacteria with sterile PBS. The membrane samples of side of the membranes to backush the fouled membrane at
2 cm length were positioned in a sterile plastic tube with 10 mL a transmembrane pressure (Dp) to facilitate the removal of the
bacterial suspension. The tube was incubated in a shaking bed contaminants. The water permeation ux in this process, Jw
at 100 rpm at 37 ꢂC for 6 h. The samples were then removed and (L mꢁ2 hꢁ1), can be deduced from the following relation:
gently rinsed thrice with PBS to eliminate the unattached or
Jw ¼ Dm/(ADtr)
dead bacterial cells. Qualitative analyses of the contaminated
membranes were directly performed by FESEM observation
while quantitative measurements were obtained by the spread
plate method with each result reported from the mean value
where Dm (g) denotes the mass of water transported across the
effective membrane area A (m2) at predetermined time intervals
Dt (h) and r denotes the density of water (g Lꢁ1). The normalized
over at least ve concurrent samples.
water ux is calculated via Jw/Ji, where Ji is the initial water
permeation ux during the rst 10 min. In parallel assays, the
modied membranes were aged by immersion in WWRe for two
months to further assess the antifouling and self-healing
characteristics of the PIL coating layers. For each membrane
module, the effective membrane area was 11.2 cm2.
2.4 Membrane fouling assays in PRO processes
The fouling tests against WWRe were performed in a lab-scale
PRO setup as demonstrated in Scheme S1,† feeding 3 L
synthetic SWBr (0.81 mol Lꢁ1 NaCl) in the draw side and 1 L
WWRe in the feed side. The WWRe has total dissolved solids
(TDS) in the range of 1000–1350 ppm, of which the major
cations are Na2+ꢁ(13–17%) and Ca2+ (8–10%) and the major
anions are SO4 (22–25%) and Clꢁ (14–17%). Its biochemical
oxygen demand (BOD) is in the range of 3–10 ppm, and its
3. Results and discussion
3.1 Preparation of the poly(ionic liquid) functionalized PES-
g-hbSPG-QPEI or PES-g-hbSPG-PDDA hollow ber membranes
chemical oxygen demand (COD) is in the range of 150–220 ppm. Mussel-inspired polydopamine (PDA) coatings have been widely
Touati and Tadeo reported that reverse salt diffusion and implemented in various membrane separation techniques as
produced power are affected by the operating conditions.44 The they enable a reversible or irreversible bridge between the inert
operating conditions have been optimized in our previous inorganic or organic surface and diverse functional polymers
studies45,46 and therefore the operating pressure was main- via Schiff base or Michael additions.47–49 Ingole et al. investi-
tained at 15 ꢄ 0.2 bar at the draw side during each 10 h testing gated the effect of the PDA coating on PRO membranes and
cycle. To provide high velocity, equal ow rates of 0.2 L minꢁ1 achieved an increase in the power density for the PDA-TFC
were maintained for both the feed and draw sides without any membranes.50 Oxidative self-polymerization of dopamine by
concentration/temperature adjustments to both the solutions using air as an exclusive oxidant, which usually proceeds
throughout the whole test period. The membranes were sub- randomly on the membrane surface, enabled the generation of
jected to pressure-conditioning at 15 bar for 0.5 h before each
a
decient amount of irregular PDA nanoparticles and
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
incomplete coverage of the surface in a sluggish growth fashion. hydrolysis of commercial poly(2-ethyl-2-oxazoline) (Mn,GPC
¼
In consideration of the reduced selectivity and permeability 25 000 g molꢁ1, Đ < 1.4), followed by quaternization with methyl
arising from the inappropriate coating strategies especially in iodide. Detailed information about the synthesis and character-
the increasingly sophisticated membrane separation systems, ization of hbSPG and QPEI is presented in our previous publi-
a stronger and stable CuSO4/H2O2 oxidant stood out as cation and in the ESI (Fig. S2 and S3†).38–43 For comparison
a promising candidate, leading to the continuous deposition of purpose, commercial poly(diallyldimethylammonium chloride)
small and uniform PDA nanoparticles on the membrane surface (PDDA, Mw < 100 000, 35 wt%) was also employed for the fabri-
with negligible structural defects, tunable thickness and favor- cation of PIL coatings.
able polymerization kinetics.51 It also enabled large-scale
The construction procedure for the hierarchically structured
membrane functionalization with consistent performance and PIL layer consisted of the successive conjugation of anionic and
feasibility at low industrial cost. In the present study, compact cationic polyelectrolytes from the PDA gutter layer under mild
coverage of the PDA gutter layer with thickness over 10 nm was conditions. hbSPG possesses many outstanding characteristics
formed by dosing the Tris buffer solution with dopamine and of dendrimers including tunable molecular weight, narrow
CuSO4/H2O2 oxidant to the shell channel of the PES-TFC dispersity, abundant end-functionalities, high water solubility,
membranes for 3 h, as elucidated by the signal screening of thermal stability, chemical resistance and biocompatibility. In
the underlying S element and a [N]/[C] ratio of approximately addition, it has been widely reported that a branched polymer
0.11 in the X-ray photoelectron spectroscopy (XPS) wide-scan provides a larger shielding area than its linear counterpart at an
spectrum (Fig. S1a and S1d†). The deconvoluted C 1s core- equivalent graing density. XPS wide-scans and C 1s and S 2p
level spectrum split into ve peaks, with the respective core-level spectra of the exterior surface of PES-g-hbSPG
binding energies (BEs) at 284.6 eV for the –CH–/–C–C species, membranes were collected and the results are displayed in
285.6 eV for the –C–N species, 286.2 eV for the –C–O species, Fig. 1a, d and e. The signal shielding of the underlying N
287.4 eV for the –C]O species and 288.8 eV for the O–C]O element and simultaneous signal renascence of S and Na
species, which is in good correlation with the chemical struc- elements in the XPS wide-scan spectrum of the PES-g-hbSPG
ture of neat PDA (Fig. S1e†).40,52,53
membrane surface validated the fact that the PDA layer is fully
As macromolecular precursors for the fabrication of poly- buried under the thick hbSPG branches (Fig. 1a).39,52,53 The
(ionic liquid) (PIL) coatings, sulfonated hyperbranched poly- deconvoluted C 1s core-level spectrum comprised four distinc-
glycerol (hbSPG, Mn,NMR ¼ 12 200 g molꢁ1, Mn,GPC ¼ 9800 g tive peaks with BEs at 284.6, 285.6, 286.5 and 288.7 eV, which
molꢁ1, Đ ¼ 1.22, Scheme S2†) was preliminarily produced by the can be associated with the neutral –CH–/–C–C, –C–S, –C–O/–
ꢁ
anionic ring-opening multi-branching polymerization of glycidol, C–SO3 and O–C]O species (Fig. 1d), respectively. The decon-
followed by sulfonation of the hydroxyl groups with sulfamic acid voluted S 2p core-level spectrum is composed of two spin–orbit
whereas quaternized poly(ethylenimine) (QPEI, linear, Mn,GPC
¼
doublets, including the predominant part with BEꢁs at 168.6 and
ꢁ
18 000 g molꢁ1, Đ ¼ 1.24) was synthesized by the partial 169.8 eV for the –C–SO3 2p3/2 and –C–SO3 2p1/2 peak
Fig. 1 XPS wide-scan, C 1s, N 1s and S 2p core-level spectra of the (a, d, and e) PES-g-hbSPG, (b, f, and f0) PES-g-hbSPG-QPEI and (c and g) PES-
g-hbSPG-PDDA hollow fiber membranes.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 2 Membrane morphologies of (a1–a4) pristine PES-TFC, (b1–b4) PES-g-hbSPG and (c1–c4) PES-g-hbSPG-QPEI hollow fiber membranes.
components in the hbSPG polymer edges and the rest with BEs correlation with the decreased density of quaternary ammo-
at 163.1 and 164.3 eV for the –C–S 2p3/2 and –C–S 2p1/2 peak nium groups in PDDA.
components formed by Michael addition (Fig. 1e).40,52,53
The combination of a Schiff base complex with residual
imine and the PDA gutter layer, and short-distance electrostatic
attraction between anionic and cationic polyelectrolytes
accounts for the rm binding of QPEI polymers to the
3.2 Surface chemistry and morphology characterization of
membranes
Fig. 2 illustrates the morphologies of the pristine, hbSPG
functionalized and PIL coated hollow ber membranes. All
membrane surface anchors with high chain mobility and
three membranes demonstrate a typical ridge-and-valley poly-
reversible ion-pair interactions.42,43 In self-healing coatings, the
amide structure on the interior surface. The inner surface
(Fig. 2(a2)–(c2)) and cross-section (Fig. 2(a1)–(c1)) of the poly-
amide layer maintain the same structure with rough ridges,
demonstrating that these modication processes cause no
conned mobility of the polymer brushes within the nanoscale
thickness has hitherto restricted the self-recovery behavior
upon mechanical damage because the reversible interaction
can be rebuilt only when the polymers can migrate across the
damaged area and come into intimate contact with each other.
In other words, linear polymers would possess higher mobility
than branched polymers and therefore are able to migrate to the
edge of the damaged interfaces in close proximity. The degree of
neutralization of sulfonate groups in hbSPG was regulated by
the feed molar ratio of the corresponding ammonium groups in
QPEI in combination with the fast kinetics of the neutralization
reaction. The C 1s core-level spectrum can be curve-tted into
two peak components having BEs at 284.6 and 286.5 eV
assigned to the –CH–/–C–C and combined –C–O/–C–SO3ꢁ/–
C–N+ species, respectively (Fig. 1f). The same result can also be
deduced from the complete substitution of the Na element with
the N element in the XPS wide-scan spectrum of the PES-g-
hbSPG-QPEI membranes (Fig. 1b). The N 1s core-level line is
occupied by an exclusive peak with a BE at around 402 eV,
attributable to the quaternary nitrogen (–N+) species from the
coupled QPEI (Fig. 1f0). With regard to PES-g-hbSPG-PDDA
membranes, similar trends were observed (Fig. 1c and g),
Fig.
3 z-Potential of the PES-g-hbSPG and PES-g-hbSPG-QPEI
except for the increased [–CH–/–C–C]/[–C–O/–C–SO3ꢁ/–C–N+] in membranes as a function of pH.
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
The z-potential of the PES-g-hbSPG and PES-g-hbSPG-QPEI
membranes as a function of pH is exhibited in Fig. 3. In the
pH range from 3 to 10, the hbSPG modied membrane is
negatively charged, which is mainly attributable to the depro-
ꢁ
tonation of the ionizable –C–SO3 groups in the hbSPG poly-
electrolyte branches. However, aer the electrostatic
conjugation of the QPEI polymer onto the hbSPG modied
surface, the corresponding z-potential reverses and becomes
positive. The overall membrane charge proles could be asso-
ciated with the difference between the positively charged
quaternary ammonium groups and negatively charged sulfo-
nate groups. With the increase in the charge density of cationic
polymers, a more positive deviation from charge neutrality in
the PIL networks was found. The hybrid chelating polymers may
provide extra adsorption and repulsion effects for both nega-
tively and positively charged foulants, preventing the foulants
from entering into the porous substrates.
Fig. 4 Relative fluorescence intensities of (a) pristine PES, (b) PES-g-
hbSPG, (c) PES-g-hbSPG-QPEI and (d) PES-g-hbSPG-PDDA hollow
fiber membranes after exposure to 1.0 mg mLꢁ1 BSA-FITC and g-
globulin-FITC solution for 2 h. Inset pictures are the respective fluo-
rescence microscopy images of protein adsorption on the aged
samples.
3.3 Anti-protein and antimicrobial performances of the PES-
TFC, PES-g-hbSPG, PES-g-hbSPG-QPEI and PES-g-hbSPG-
PDDA hollow ber membranes
structural damage to the polyamide selective layer. On the other
hand, the outer surface of the three membranes exhibits
a porous structure with a relatively smoother surface compared
to the polyamide layer. There are many small nodules observed
on the outer surface of both PES-g-hbSPG and PES-g-hbSPG-
QPEI, indicating the deposition of coating polymers. The AFM
results shown in Fig. S4† indicate that the pristine PES has the
lowest roughness and the polymer graing amplies the
surface roughness to a small extent, where Rms increases from
6.9 nm to 9.1 nm. The porous exterior surface of the pristine
PES support allows impurities to pass through and accumulate.
For PES-g-hbSPG and PES-g-hbSPG-QPEI hollow ber
membranes, dense and rough structures could be well-
preserved and the original surface porous structures are fully
covered, elucidating the successful incorporation of the PIL
polymer pair onto the PES supports.
The amount of protein adsorbed on the membrane surfaces is
signicantly inuenced by the ionic strength of the poly-
electrolyte pair and the specic features of the ions. In general,
anionic polymers tend to bind positively charged proteins (g-
globulin, pl ¼ 7.3–8.2) whereas cationic polymers tend to bind
negatively charged proteins (BSA, pl ¼ 4.7) via electrostatic
attraction to afford different protein build-up on the membrane
surfaces.54 To elucidate the effect of surface chemistry evolution
on protein adsorption, the involved membranes were placed in
two uorescein labeled protein solutions (g-globulin-FITC and
BSA-FITC). Fluorescence intensities and uorescence micros-
copy measurements were recorded for the original PES-TFC and
PIL decorated membranes aer soaking in 1.0 mg mLꢁ1 BSA-
FITC or g-globulin-FITC solution for 2 h. The hydrophobic–
hydrophobic interactions among PES and proteins enable the
surface to bind a large quantity of proteins. In contrast, the
Fig. 5 FESEM images of (a and e) pristine PES, (b and f) PES-g-hbSPG, (c and g) PES-g-hbSPG-QPEI and (d and h) PES-g-hbSPG-PDDA hollow
fiber supports after exposure to PBS suspensions of S. aureus (top) and E. coli (below) at an initial cell concentration of 1 ꢃ 108 CFU mLꢁ1 for 6 h at
37 ꢂC.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
as one of the great challenges in membrane separation tech-
niques in water treatment. In the present study, water based
assays against two strains, Gram-negative E. coli and Gram-
positive S. aureus, were adopted to examine the antibacterial
behavior of the PIL-coated membranes. A myriad of live bacterial
clusters with smooth and unperturbed features were observed on
the domestic PES-TFC membranes when contaminated by PBS
suspensions of S. aureus and E. coli for a prolonged incubation
time of 6 h, with the former exhibiting stronger biolm
formation capability (Fig. 5).40 Aer surface tethering of hbSPG
branches with abundant sulfonate groups that could resist the
fouling of hydrophobic bacterial cells, the PES-g-hbSPG hollow
ber membranes became less prone to bacterial attachment,
with the respective viabilities of E. coli and S. aureus
decreasing to 8.1% and 10.2%, as determined by the spread
plate method (Fig. 6). As expected, only sparse bacterial cells
with lysed and wrinkled shapes were found on the PIL func-
tionalized membrane surfaces. The synergistic effect of the
bacteria-repelling features in the PIL hierarchical architecture
and the bactericidal characteristics of the free quaternary
ammonium base from the extended QPEI polymer in the
aqueous environment enabled the killing of approaching
bacterial cells and release of loosely bound bacterial debris by
gentle rinsing, leaving behind a much cleaner membrane
surface. Consequently, the PES-g-hbSPG-QPEI and PES-g-
hbSPG-PDDA membrane containing different concentrations
of the remaining quaternary ammonium salts further
decreased S. aureus attachment to 2.8% and 1.6%, and E. coli
adhesion to 1.8% and 0.8%, respectively. These results, when
combined, indicate that the PIL-based polymer coatings are
highly efficient in suppressing bacterial attachment, prolifer-
ation and thus biolm formation.
Fig. 6 Viable adherent fractions of (a) pristine PES, (b) PES-g-hbSPG,
(c) PES-g-hbSPG-QPEI and (d) PES-g-hbSPG-PDDA hollow fiber
supports after exposure to PBS suspensions of S. aureus and E. coli at
an initial cell concentration of 1 ꢃ 108 CFU mLꢁ1 for 6 h at 37 ꢂC. The
cell number was determined by the spread plate method. Error bars
represent the standard deviation of five measurements.
relative uorescence intensities of BSA and g-globulin adsorp-
tions signicantly reduced to less than 10% on the PIL coated
membrane surfaces. The combined results indicated that PIL
functionalized membranes can bind various amounts of water
and thereby establish a compact surface hydration network that
generally exhibits reduced fouling propensity and provides
a steric and energetic barrier against the adsorption of hydro-
phobic impurities (Fig. 4).
A stubborn biolm forms on the membrane surfaces once the
accumulation of bacterial colonies exceeds a certain amount,
leading to the decline of membrane permeability, selectivity and
lifespan, which is identied as the quorum sensing effect.25,26 3.4 Antifouling behavior in the PRO process
Antimicrobial surface coatings, which could effectively eliminate
Frequent backwash, chemical treatment and even membrane
or neutralize bacteria before they become destructive, are pref-
erable in combating membrane fouling which is acknowledged
replacement are oen prerequisite steps towards water reme-
diation in modern membrane separation systems prior to the
Fig. 7 Normalized water fluxes of the PES-TFC and PES-g-hbSPG-QPEI membranes as a function of normalized accumulative permeate volume
in the PRO process (blank marker: PES-TFC; solid marker: PES-g-hbSPG-QPEI). The hydraulic pressure difference between the feed and draw
solution was 15.0 ꢄ 0.2 bar. No concentration adjustments were made to either the feed or draw solutions throughout the entire test.
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 8 FESEM images of the outer surfaces of the PES-TFC and PES-g-hbSPG-QPEI hollow fiber membranes. “Pristine membrane” indicates the
fresh membrane as prepared and “cleaned membrane” indicates the fouled membrane after the 3rd backwash cleaning. Backwash cleaning was
carried out using DI water under 15.0 ꢄ 0.2 bar for 30 min.
unintended impurities reaching and accumulating on the ux while the water ux of the PES-g-hbSPG-QPEI remains at
membrane module.55 To investigate the comprehensive anti- 60% of the initial value, indicating the lower fouling tendency of
fouling performance of the pristine and PIL modied the dual-layer modied PRO membrane.
membranes, WWRe taken from a local municipal water treat-
Fig. 8 shows the FESEM observations of the exterior surfaces
ment plant was employed as the feed solution during PRO tests of PES-TFC and PES-g-hbSPG-QPEI before PRO fouling tests and
under 15 bar. Fig. 7 shows the plots of the normalized water aer the 3rd backwash. The pristine PES-TFC exterior surface is
uxes of PES-TFC and PES-g-hbSPG-QPEI membranes as fully porous, which makes it vulnerable to the accumulation of
a function of the normalized cumulative volume of the foulants. Even aer the 3rd backwash cleaning, the exterior
permeate. When WWRe is dosed to the feed side of PES-TFC surface of PES-TFC is still fully sheltered with a thick coverage of
membranes, the water ux of PES-TFC rst drops to 90% of salt crystals, nodules and particles, which can be explained by
the initial value and nally decreases to about 60% of the initial the sticky bonding of these inorganic foulants to the porous and
water ux when the normalized accumulative permeate volume hydrophobic PES-TFC membrane surface. On the other hand,
reaches 100 L mꢁ2 in run 1. This phenomenon can be explained
by the fast and severe gathering of the contaminants inside the
porous structure of the original PES-TFC membrane. In
contrast, the PES-g-hbSPG-QPEI membrane exhibits a relatively
slower initial water ux decline. Over the subsequent testing
period, the water ux slowly decreases and stabilizes at 80% of
the initial water ux. The smaller drop in water ux validates
that the dual-layer chelating polymer modications enable the
PRO membrane to have a minimum fouling tendency due to the
generation of a hydration layer caused by the polyelectrolyte
brushes and combined adsorption/exclusion mechanisms.
Aer each 10 h operation, backwashing by feeding DI water
into both sides of the membranes at a Dp of 15 bar at the
lumen side for 1 h was carried out. For the subsequent runs,
the PES-TFC membrane still retains a higher attraction to
various contaminants. In comparison, the PES-g-hbSPG-QPEI
membranes become more sluggish and exhibit a limited
decline in terms of water ux over the whole testing period. It is
worth noting that the nal water ux aer each subsequent run
PES-g-hbSPG-QPEI membranes as a function of normalized accu-
Fig. 9 Normalized water fluxes of the PES-TFC, PES-g-hbSPG and
was lower compared with that of the previous run, which may be
attributed to the incomplete removal of foulants during the
simple backwashing for both membranes. Eventually, aer run
3, the water ux of the PES-TFC drops to 35% of the initial water
mulative permeate volume in the PRO process (blank marker: PES-
TFC; triangle maker: PES-g-hbSPG; solid marker: PES-g-hbSPG-
QPEI). The hydraulic pressure difference between the feed and draw
solution was 15.0 ꢄ 0.2 bar. No concentration adjustments were made
to either the feed or draw solutions throughout the entire test.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
Journal of Materials Chemistry A
Scheme 2 Proposed mechanism of the self-healing property of the poly(ionic liquid) hierarchical structure.
the PIL modied membrane could regain its original both the non-aged and WWRe aged samples. In PRO tests, slow
morphology aer backwash cleaning and only sparse crystals ux reductions and high ux recoveries aer backushing and
and particles could be observed on the exterior surface of the hydraulic pressure impulsion on the non-aged and aged
PES-g-hbSPG-QPEI membrane. Clearly, the PIL functionalized membranes were also found. The higher chain mobility of the
membrane surfaces prove effective in inhibiting foulant adhe- PIL network endows the underlying PES-TFC membranes with
sion and reproduction.
self-healing properties in aqueous environments, in which the
One intriguing advantage of PIL coating layers lies in the edges of damaged interfaces in close proximity would be readily
long-term stability and environmentally friendly behavior in compensated via reversible ion-pair exchange, leading to
aqueous environments. These robust traits of the PES-g-hbSPG- repeated healing processes, maintained membrane perfor-
QPEI membranes encouraged us to evaluate the performance of mance and prolonged membrane lifespan upon usage.
modied membranes aer aging in WWRe for two months.
Similar scenarios were observed on the aged membrane
samples in Fig. 5 for the anti-protein adsorption assays, Fig. 7
for the antimicrobial adhesion tests and Fig. 9 for the anti-
fouling performance in PRO tests. The higher chain mobility of
Conflicts of interest
There are no conicts to declare.
the PIL network endows the underlying PES-TFC membranes
with self-healing properties in aqueous environments, in which
the edges of damaged interfaces in close proximity would be
Acknowledgements
readily compensated via reversible ion-pair exchange, leading The nancial support from the Science and Technology
to repeated healing processes, maintained membrane Program
of
Shenzhen
(JCYJ20170303170403122
and
performance and prolonged membrane lifespan upon usage JCYJ20170303170340122), the National Natural Science Foun-
(Scheme 2).27–31 Self-healing coating layers are self-propellant dation of China (51503155, 51773156 and 51703167), the
systems that can revert to their original state aer micro- Natural Science Foundation of Hubei Province (2016CFB381
damage with a partial or complete recovery of their structural and 2017CFB127) and the Singapore National Research Foun-
integrity.
dation (NRF) under its Environment & Water Strategic Research
Programme, administered by the PUB under the project title
“Membrane Development for Osmotic Power Generation, Part
2. Module Fabrication and System Integration” (1102-IRIS-11-2)
is acknowledged. Special thanks are due to Prof. Tai-Shung
Chung and Dr Senlin Shao for their suggestions and support
for the experimental work.
4. Conclusions
In the present study, the antifouling and self-healing PIL
coating layers have been transformed into the engineered PES-
TFC hollow ber membranes by the sequential conjugation of
anionic hbSPG and cationic QPEI on the preliminarily
fabricated PDA gutter layer by CuSO4/H2O2 oxidation of DA. The
stronger and stable CuSO4/H2O2 oxidant resulted in the
continuous deposition of small and uniform PDA nanoparticles
on the membrane surface with negligible structural defects,
tunable thickness and favorable polymerization kinetics. It also
enabled large-scale membrane functionalization with consis-
tent performance and feasibility at low industrial cost. The
superior antifouling characteristics of the resultant PES-g-
hbSPG-QPEI membranes were veried by the excellent anti-
protein behavior and improved antibacterial performance on
References
1 G. Z. Ramon, B. J. Feinberg and E. M. V. Hoek, Energy
Environ. Sci., 2011, 4, 4423–4434.
2 M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis,
B. J. Marinas and A. M. Mayes, Nature, 2008, 452, 301–310.
3 X. B. Yang, Z. X. Wang and L. Shao, J. Membr. Sci., 2018, 549,
67–74.
4 Q. Chen, W. Xu and Q. C. Ge, Environ. Sci. Technol., 2018, 52,
4464–4471.
This journal is © The Royal Society of Chemistry 2019
J. Mater. Chem. A

View Article Online
Journal of Materials Chemistry A
Paper
5 K. Lee, R. Baker and H. Lonsdale, J. Membr. Sci., 1981, 8, 141– 30 X. Xing, L. Li, T. Wang, Y. Ding, G. Liu and G. Zhang, J.
147. Mater. Chem. A, 2014, 2, 11049–11053.
6 W. R. Thelin, E. Sivertsen, T. Holt and G. Brekke, J. Membr. 31 S. Owusu-Nkwantabisah, J. Gillmor, S. Switalski, M. R. Mis,
Sci., 2013, 438, 46–56.
G. Bennett, R. Moody, B. Antalek, R. Gutierrez and
G. Slater, Macromolecules, 2017, 50, 3671–3679.
32 T. L. Greaves and C. J. Drummond, Chem. Rev., 2015, 115,
11379–11448.
33 W. Qian, J. Texter and F. Yan, Chem. Soc. Rev., 2017, 46,
1124–1159.
7 P. G. Ingole, W. Choi, K. H. Kim, H. D. Jo, W. K. Choi,
J. S. Park and H. K. Lee, Desalination, 2014, 345, 136–145.
8 Z. Y. Li, R. V. Linares, M. Abu-Ghdaib, T. Zhan, V. Yangali-
Quintanilla and G. Amy, Water Res., 2014, 48, 200–209.
9 T. D. Lu, B. Z. Chen, J. Wang, T. Z. Jia, X. L. Cao, Y. Wang,
W. Xing, C. H. Lau and S. P. Sun, J. Mater. Chem. A, 2018, 34 V. F. Korolovych, A. Erwin, A. Stryutsky, H. Lee, W. T. Heller,
6, 15047–15056.
10 Q. C. Ge, C. H. Lau and M. H. Liu, Environ. Sci. Technol.,
2018, 52, 3812–3819.
11 Q. C. Xia, M. L. Liu, X. L. Cao, Y. Wang, W. Xing and
S. P. Sun, J. Membr. Sci., 2018, 562, 85–111.
12 M. M. Zhang, D. X. Hou, Q. H. She and C. Y. Tang, Water Res.,
2014, 48, 387–395.
V. V. Shevchenko, L. A. Bulavin and V. V. Tsukruk,
Macromolecules, 2018, 51, 4923–4937.
35 K. Zhang, X. Feng, C. Ye, M. A. Hempenius and G. J. Vancso,
J. Am. Chem. Soc., 2017, 139, 10029–10035.
36 J. K. Sun, Z. Kochovski, W. Y. Zhang, H. Hirmse, Y. Lu,
M. Antonietti and J. Yuan, J. Am. Chem. Soc., 2017, 139,
8971–8976.
13 Z. L. Cheng, X. Li, Y. Feng, C. F. Wan and T.-S. Chung, 37 J. Cui, F. M. Nie, J. X. Yang, L. Pan, Z. Ma and Y. S. Li,
J. Membr. Sci., 2017, 524, 97–107. J. Mater. Chem. A, 2017, 5, 25220–25229.
14 L. Shen, F. Wang, L. Tian, X. Zhang, C. Ding and Y. Wang, 38 X. Li, X. Hu and T. Cai, Langmuir, 2017, 33, 4477–4489.
J. Membr. Sci., 2018, 563, 284–297.
39 X. Li, J. L. Li, Y. Wang and T. Cai, ACS Appl. Nano Mater.,
2018, 1, 2559–2566.
40 Y. Zhang, J. L. Li, T. Cai, Z. L. Cheng, X. Li and T. S. Chung,
J. Membr. Sci., 2018, 563, 521–530.
41 J. Herzberger, K. Niederer, H. Pohlit, J. Seiwert, M. Worm,
F. R. Wurm and H. Frey, Chem. Rev., 2016, 116, 2170–2243.
42 V. R. de la Rosa, E. Bauwens, B. D. Monnery, B. G. De Geest
and R. Hoogenboom, Polym. Chem., 2014, 5, 4957–4964.
43 L. Tauhardt, K. Kempe, M. Gottschaldt and U. S. Schubert,
Chem. Soc. Rev., 2013, 42, 7998–8011.
15 D. Y. Zhang, Q. Hao, J. Liu, Y. S. Shi, J. Zhu, L. Su and
Y. Wang, Sep. Purif. Technol., 2018, 192, 230–239.
16 H. G. Sun, X. B. Yang, Y. Q. Zhang, X. Q. Cheng, Y. C. Xu,
Y. P. Bai and L. Shao, J. Membr. Sci., 2018, 563, 22–30.
17 Y. Li, Y. Su, X. Zhao, X. He, R. Zhang, J. Zhao, X. Fan and
Z. Jiang, ACS Appl. Mater. Interfaces, 2014, 6, 5548–5557.
18 W. Xu, Q. Chen and Q. C. Ge, Desalination, 2017, 419, 101–
116.
19 D. Zhao, G. Qiu, X. Li, C. F. Wan, K. J. Lu and T. S. Chung,
Water Res., 2016, 104, 389–396.
44 K. Touati and F. Tadeo, Desalination, 2016, 389, 171–186.
20 J. Li, L. Yin, G. Qiu, X. Li, Q. Liu and J. Xie, J. Mater. Chem. A, 45 X. Li and T.-S. Chung, Appl. Energy, 2014, 114, 600–610.
2015, 3, 6781–6786.
46 W. Gai, X. Li, J. Y. Xiong, C. F. Wan and T.-S. Chung,
21 D. Rana and T. Matsuura, Chem. Rev., 2010, 110, 2448–2471.
J. Membr. Sci., 2016, 516, 104–112.
22 Q. She, R. Wang, A. G. Fane and C. Y. Tang, J. Membr. Sci., 47 H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith,
2016, 499, 201–233. Science, 2007, 318, 1468–1492.
23 M. Lejars, A. Margaillan and C. Bressy, Chem. Rev., 2012, 112, 48 W. O. Yah, H. Xu, H. Soejima, W. Ma, Y. Lvov and
4347–4390. A. Takahara, J. Am. Chem. Soc., 2012, 134, 12134–12137.
24 X. Ding, X. Duan, X. Ding, R. Liu and F. J. Xu, Adv. Funct. 49 Y. Liu, K. Ai and L. Lu, Chem. Rev., 2014, 114, 5057–5115.
Mater., 2018, 1802140.
25 W. J. Yang, X. Tao, T. Zhao, L. Weng, E.-T. Kang and L. Wang,
Polym. Chem., 2015, 6, 7027–7035.
26 J. Liu, J. Liu, S. Wang, J. Huang, S. Wu, Z. Tang, B. Guo and
L. Zhang, J. Mater. Chem. A, 2017, 5, 25660–25671.
27 D. Mozhdehi, S. Ayala, O. R. Cromwell and Z. Guan, J. Am.
Chem. Soc., 2014, 136, 16128–16131.
50 P. G. Ingole, W. Choi, K. H. Kim, C. H. Park, W. K. Choi and
H. K. Lee, Chem. Eng. J., 2014, 243, 137–146.
51 C. Zhang, Y. Lv, W. Z. Qiu, A. He and Z. K. Xu, ACS Appl.
Mater. Interfaces, 2017, 9, 14437–14444.
52 X. Li, T. Cai and T. S. Chung, Environ. Sci. Technol., 2014, 48,
9898–9907.
53 X. Li, T. Cai, C. Chen and T. S. Chung, Water Res., 2016, 89,
50–58.
28 F. Luo, T. L. Sun, T. Nakajima, D. R. King, T. Kurokawa,
Y. Zhao, A. B. Ihsan, X. Li, H. Guo and J. P. Gong, 54 B. K. D. Ngo and M. A. Grunlan, ACS Macro Lett., 2017, 6,
Macromolecules, 2016, 49, 2750–2760. 992–1000.
29 T. J. Cuthbert, J. J. Jadischke, J. R. de Bruyn, P. J. Ragogna 55 T. Cai, X. Li, C. Wan and T. S. Chung, J. Membr. Sci., 2016,
and E. R. Gillies, Macromolecules, 2017, 50, 5253–5260.
497, 142–152.
J. Mater. Chem. A
This journal is © The Royal Society of Chemistry 2019