ArticlePDF Available

Abstract and Figures

Forward osmosis (FO) has emerged as a promising membrane technology to yield high-quality reusable water from various water sources. A key challenge to be solved is the bidirectional solute flux (BSF), including reverse solute flux (RSF) and forward solute flux (FSF). Herein, zwitterion functionalized carbon nanotubes (Z-CNTs) have been coated onto a commercial thin film composite (TFC) membrane, resulting in BSF mitigation via both electrostatic repulsion forces induced by zwitterionic functional groups and steric interactions with CNTs. At a coating density of 0.97 g m −2 , a significantly reduced specific RSF was observed for multiple draw solutes, including NaCl (55.5% reduction), NH 4 H 2 PO 4 (83.8%), (NH 4) 2 HPO 4 (74.5%), NH 4 Cl (70.8%), and NH 4 HCO 3 (61.9%). When a synthetic wastewater was applied as the feed to investigate membrane rejection, FSF was notably reduced by using the coated membrane with fewer pollutants leaked to the draw solution, including NH 4 +-N (46.3% reduction), NO 2 −-N (37.0%), NO 3 −-N (30.3%), K + (56.1%), PO 4 3−-P (100%), and Mg 2+ (100%). When fed with real wastewater, a consistent water flux was achieved during semi-continuous operation with enhanced fouling resistance. This study is among the earliest efforts to address BSF control via membrane modification, and the results will encourage further exploration of effective strategies to reduce BSF.
Content may be subject to copyright.
Contents lists available at ScienceDirect
Environment International
journal homepage: www.elsevier.com/locate/envint
Mitigation of bidirectional solute ux in forward osmosis via membrane
surface coating of zwitterion functionalized carbon nanotubes
Shiqiang Zou
a,1
, Ethan D. Smith
b,1
, Shihong Lin
c
, Stephen M. Martin
b,
, Zhen He
a,
a
Department of Civil and Environmental Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA
b
Department of Chemical Engineering & Macromolecules Innovation Institute, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA
c
Department of Civil and Environmental Engineering, Vanderbilt University, Nashville, TN 37235, USA
ARTICLE INFO
Handling Editor: Hefa Cheng
Keywords:
Forward osmosis
Reverse solute ux
Forward solute ux
Zwitterion
Membrane modication
ABSTRACT
Forward osmosis (FO) has emerged as a promising membrane technology to yield high-quality reusable water
from various water sources. A key challenge to be solved is the bidirectional solute ux (BSF), including reverse
solute ux (RSF) and forward solute ux (FSF). Herein, zwitterion functionalized carbon nanotubes (Z-CNTs)
have been coated onto a commercial thin lm composite (TFC) membrane, resulting in BSF mitigation via both
electrostatic repulsion forces induced by zwitterionic functional groups and steric interactions with CNTs. At a
coating density of 0.97 g m
2
, a signicantly reduced specic RSF was observed for multiple draw solutes,
including NaCl (55.5% reduction), NH
4
H
2
PO
4
(83.8%), (NH
4
)
2
HPO
4
(74.5%), NH
4
Cl (70.8%), and NH
4
HCO
3
(61.9%). When a synthetic wastewater was applied as the feed to investigate membrane rejection, FSF was
notably reduced by using the coated membrane with fewer pollutants leaked to the draw solution, including
NH
4+
-N (46.3% reduction), NO
2
-N (37.0%), NO
3
-N (30.3%), K
+
(56.1%), PO
43
-P (100%), and Mg
2+
(100%). When fed with real wastewater, a consistent water ux was achieved during semi-continuous operation
with enhanced fouling resistance. This study is among the earliest eorts to address BSF control via membrane
modication, and the results will encourage further exploration of eective strategies to reduce BSF.
1. Introduction
Alternative water resources, e.g., via water reuse and desalination,
are of signicant interest for addressing mounting global water de-
mand. Membrane-based treatment processes have been extensively
developed to produce reusable water due to their reliable performance
and compact footprint (Shannon et al., 2008). As an emerging mem-
brane technology, forward osmosis (FO) is able to supply high-quality
water by utilizing a natural osmotic pressure gradient, and can oer
unique merits such as reduced pressure operation, low fouling pro-
pensity, excellent solute rejection, and relatively low energy con-
sumption if proper regeneration/separation of draw solutes (DS) could
be achieved (Zou et al., 2016). Advancement of FO-based technologies
will need to address several technical bottlenecks, especially further
enhancement of fouling resistance and mitigation of reverse solute ux
(RSF). RSF is dened as the cross-membrane DS diusion to the feed
(She et al., 2012), and has thus far received less attention than en-
hancing fouling resistance in the FO eld. RSF has been studied via
mathematical modeling (Lu et al., 2014), and can be aected by factors
such as intrinsic membrane parameters (e.g., thickness, tortuosity, and
porosity) (McCutcheon and Elimelech, 2006) and solute characteristics
(e.g., hydrated ion radius, aqueous diusivity, solution viscosity, and
ion charge) (Zhao and Zou, 2011). Gradual DS leakage via RSF can lead
to reduced osmotic driven force, salinity buildup on the feed side, po-
tential feed contamination, aggravated membrane fouling, and elevated
operating costs due to the need to continuously replenish DS (Lu and
He, 2015).
Increasing awareness of RSF's detrimental eects has led to the
development of indirect and direct control strategies. Indirect RSF
control focuses on addressing the consequence of RSF, i.e. salinity
buildup on the feed side, and major approaches include solute removal
by using microorganisms to eliminate/assimilate biodegradable DS
(e.g., NH
4
HCO
3
)(Li et al., 2016) and integration of parallel desalina-
tion/separation process (e.g., microltration or electrodialysis) (Qiu
et al., 2015;Zou and He, 2017a). Nonetheless, alleviation of salinity
accumulation in the feed solution does not mitigate RSF itself. The
leaching of DS via RSF still leads to economic loss in operation.
Therefore, direct RSF control needs to be prioritized via smart DS
https://doi.org/10.1016/j.envint.2019.104970
Received 17 April 2019; Received in revised form 27 May 2019; Accepted 26 June 2019
Corresponding authors.
E-mail addresses: martinsm@vt.edu (S.M. Martin), zhenhe@vt.edu (Z. He).
1
These authors contribute equally to this paper.
Environment International 131 (2019) 104970
0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T
selection, operational strategies, and advanced membrane development
(Zou et al., 2019). Novel DS, such as stimuli-responsive polymers, can
theoretically produce low RSF, due to their large hydrodynamic dia-
meter (> 100 nm), while achieving energy-ecient phase separation
(Hartanto et al., 2015). Operational strategies have also been in-
vestigated, including pressure- (Blandin et al., 2013), electrolysis- (Zou
and He, 2017b), and ultrasonic-assisted osmosis (Kowalski et al., 2015).
However, additional energy is required to manipulate ion transport
across the FO membrane, and inconsistent RSF reduction was reported
due to potential membrane damage under a certain operating condi-
tions (Lutchmiah et al., 2015).
Advanced membrane development aims at creating a better DS
barrier via membrane fabrication or surface modication (Akther et al.,
2015). New membrane fabrication techniques have yielded promising
results in RSF reduction, including optimized substrate composition
(Wang et al., 2015b), advanced substrate fabrication (Yasukawa et al.,
2015), selection of novel supporting materials (Huang et al., 2013), and
modied active layer composition (Kwon et al., 2015). Compared to
membrane fabrication, surface modication via alterations of surface
functional groups and hydrophilicity can be a straightforward method
to enhance membrane performance (Guo et al., 2018). Zwitterionic
monomers have emerged as an attractive surface modier to enhance
water ux and prevent adsorption of foulants in NF (Mi et al., 2017)
and RO systems (Wang et al., 2015a), owing to their remarkable hy-
dration capacity and electroneutrality (Zhou et al., 2014). Zwitterionic
modication of FO membranes can lead to notable enhancement of
fouling resistance (Wang et al., 2018), for example, a zwitterionic poly
amino acid 3-(3,4-Dihydroxyphenyl)-
L
-alanine (
L
-DOPA) coated FO
membrane exhibited up to 30% less fouling compared to an uncoated
membrane (Nguyen et al., 2013). However, to the best of our knowl-
edge, no prior studies have used zwitterionic materials for simultaneous
RSF and forward solute ux (FSF) reduction (i.e. bidirectional ux re-
duction, BSF) in FO systems.
Herein, zwitterion functionalized carbon nanotubes (Z-CNT) have
been coated onto a commercial Aquaporin Inside®FO membrane for
performance enhancement (Holt et al., 2006) and BSF reduction. The
specic objectives of this study are to (1) investigate the feasibility and
consistency of RSF reduction by a Z-CNT coated FO membrane; (2)
evaluate the eects of key operation parameters on RSF reduction, in-
cluding Z-CNT coating density, DS concentration, and various DS spe-
cies; (3) analyze membrane rejection (i.e., FSF) of the Z-CNT coated
membrane; and (4) assess water ux consistency and membrane fouling
resistance under semi-continuous operation.
Fig. 1. (A) Schematic of the membrane coating and testing process, (B) comparison of water recovery volume among the pristine, AL-coated, or SL-coated membranes
over three successive batch tests, and (C) quantication of the corresponding water ux, RSF and SRSF. In panel A, Z-CNT coating is conrmed by SEM image under
50,000 times magnication. In panels B and C, new pristine or coated membrane (0.97 g m
2
) was used for each batch test. In panel C, error bars represent triplicates
with dierent membrane samples. Testing conditions: 100-mL 1-M NaCl as the draw and 500-mL DI water as the feed.
S. Zou, et al. Environment International 131 (2019) 104970
2
2. Materials and methods
2.1. Z-CNT preparation, membrane coating, and characterization
The functionalization process and detailed characterization were
conducted according to a previous study (Appendix A. Supplementary
data) (Chan et al., 2016). The nal zwitterion functional group attached
to CNTs (25.2 wt%, equal to ~2 zwitterion groups per 100 carbon
atoms, Supplementary data) had a positive charge at the tertiary amine
group and a negative charge at the carboxylate group (Fig. S1, Sup-
plementary data). Before membrane coating, a Z-CNT stock solution
(1.0 mg mL
1
) was prepared by adding 20 mg Z-CNTs into 20 mL of
deionized (DI) water, followed by probe sonication for 5 min (Fisher
EB705 Probe Sonicator). After being cooled to room temperature,
10 mL of the Z-CNT solution (unless otherwise stated) was uniformly
distributed onto the surface of a commercial TFC FO membrane
(Aquaporin A/S, Denmark), which was placed under a Teon frame and
clamped onto a glass plate with either its active layer (i.e., AL-coated)
or supportive layer facing up (i.e., SL-coated). Multiple batches of
commercial membranes were coated (0.97 g CNT m
2
, unless other-
wise stated) using the same protocol. The Z-CNT coated membranes
were then placed inside a fume hood overnight to allow water to eva-
porate (Fig. 1A), and van der Waals force was expected to be the main
binding force due to charged groups (e.g., hydroxyl and amine groups)
between FO membrane surface and the functionalized Z-CNTs (Khalid
et al., 2018;Park et al., 2016). The coated membrane was stored in DI
water before use.
Both the pristine and coated FO membranes were characterized for
zeta potential, membrane hydrophilicity, and surface morphology and
composition. Membrane zeta potential was measured over a range of
pH from 3.8 to 8.7 using a streaming potential analyzer (SurPASS,
Anton Paar) with an adjustable gap cell and 1-mM KCl as the electro-
lyte. Hydrochloric acid (HCl) and potassium hydroxide (KOH) were
used for pH adjustment. The membrane hydrophilicity was determined
by measuring the contact angle (CA) of a sessile drop (DI water,
~25 μL) using an optical goniometer (KSV Instruments CAM 200). For
each sample, at least three CA measurements were performed using
dierent spots of the membrane surface. Membrane surface mor-
phology was examined using eld emission scanning electron micro-
scopy (FESEM) with a 7-nm Pt/Pd coating (80:20 wt%, Leica ACE600
Sputter), with composition analysis performed using elemental map-
ping via energy dispersive spectroscopy (EDS).
2.2. Setup of FO testing cells
Two identical cross-ow FO testing cells were constructed, with one
equipped with a pristine membrane (control system) and the other
equipped with a coated membrane (AL-coated or SL-coated experi-
mental system). Each solution chamber (5 × 5 × 1 cm) was connected
to an external 600-mL reservoir. One piece of FO membrane (25 cm
2
eective area) was installed inside each testing module with its active
layer facing the feed solution (i.e. AL-Feed, FO mode) to minimize
potential membrane fouling. Upon membrane installation, the FO
testing modules were rinsed with DI water at a owrate of
120 mL min
1
for 3 h to remove unattached Z-CNT or other residue
chemicals from the membrane surface. Under a default setting, 100-mL
of 1-M NaCl (conductivity 86.887.1 mS cm
1
) and 500-mL of DI water
were recirculated at a owrate of 60 mL min
1
as the draw and feed
solutions, respectively. The FO testing systems were operated under
either a batch or a semi-continuous mode. During the batch operation
(triplicate runs), both FO testing modules were operated for 12 h with
water sampled at the beginning and the end of the experiment.
Sucient physical crossow ushing (3 h with DI water, 60 mL min
1
)
was performed between batch tests for membrane cleaning. During the
semi-continuous operation, samples were taken daily with no mem-
brane cleaning performed. All experiments were conducted in a
temperature-controlled lab (20 ± 2 °C).
2.3. Experimental procedure
The Feasibility of RSF Reduction via Z-CNT surface coating was rst
investigated in three successive batch tests (under default settings) by
comparing the performance between the pristine and coated mem-
branes (AL-coated or SL-coated, 0.97 g m
2
). The Eects of Operating
Parameters on FO performance (i.e. water ux and RSF) were then
evaluated by using the AL-coated membranes with dierent Z-CNT
coating densities (0, 0.10, 0.48, 0.97, or 1.45 g m
2
, equivalent to a
stock solution volume of 0, 1, 5, 10, or 15 mL, respectively, Fig. S2), DS
concentrations (0.25, 0.50, 0.75, or 1.00 mol L
1
NaCl), and DS com-
positions (0.25 mol L
1
NaCl, (NH
4
)
2
HPO
4
,NH
4
H
2
PO
4
,NH
4
Cl, or
NH
4
HCO
3
). Various coating densities were achieved by uniformly dis-
tributing 1, 5, 10, and 20 mL of the Z-CNT solution onto FO membrane
surface. Coating-Enhanced FSF Reduction was subsequently analyzed
using two dierent synthetic contaminated solutions as the feed
(500 mL). The ions of signicant interest include NH
4+
-N
(111.5 ± 3.5 mg L
1
), NO
3
-N (110.1 ± 0.2 mg L
1
), NO
2
-N
(119.6 ± 1.5 mg L
1
), K
+
(107.1 ± 1.4 mg L
1
), Mg
2+
(100.1 ± 0.9 mg L
1
), and Ca
2+
(99.8 ± 0.3 mg L
1
) in synthetic
solution 1, or SO
42
-S (130.4 ± 2.2 mg L
1
) and PO
43
-P
(131.8 ± 3.9 mg L
1
) in synthetic solution 2. This experiment design
was to prevent potential precipitation within the synthetic solutions.
Finally, Water Flux Consistency was explored with a focus on membrane
fouling resistance via a 12-day semi-continuous operation, in which
100-mL 1-M NaCl (i.e. the draw) was replaced on a daily-basis, whereas
the secondary euent (500 mL constant, Table S1) from a local was-
tewater treatment plant (Christiansburg, VA) was supplied as the feed
to achieve enhanced water reuse (Zou and He, 2016).
2.4. Measurement and analysis
Detailed water quality analysis is provided in Supplementary data.
Quantication of water ux and RSF (J
s
, gMH) was based on Eqs.
(S1)(S3) and (S4)(S6), respectively (Zou and He, 2017a, 2017b). FSF
(J
f
, gMH) and membrane rejection (%) were determined to quantify the
diusion of feed solutes (FS, usually pollutants) to the draw solution
(Eqs. (S7)(S8)). Specic RSF (SRSF,gL
1
) was calculated to rule out
the inuence of both membrane structural parameters and bulk draw
solution concentration (Phillip et al., 2010). SRSF quanties the leakage
of a DS or an ion that can leak across FO membrane to the feed side per
unit of recovered water.
=SRSF J
J
s
w(1)
RSF mitigation ratio (MR, %) was quantied by the ratio between
the mitigated SRSF with coated membrane (SRSF
m
) and that with the
pristine membrane (SRSF
o
).
⎜⎟
⎜⎟
=
×=
×
M
RSRSF
SRSF
JJ
JJ
1 100% 1 /
/100%
m
o
sm wm
so wo
,,
,, (2)
where J
s,o
and J
w,o
are the original RSF and water ux of the pristine
membrane, respectively; J
s,m
and J
w,m
are the mitigated RSF and water
ux of the coated FO membrane, respectively.
3. Results and discussion
3.1. RSF reduction via Z-CNT surface coating
The RSF reduction and water recovery capability of the Z-CNT
coated membrane were investigated by comparing its performance to a
pristine membrane. The FO system with the AL-coated membrane
(0.97 g m
2
) could recover 264.5 ± 3.8 mL of water within 12 h,
S. Zou, et al. Environment International 131 (2019) 104970
3
comparable to that of the pristine membrane (261.3 ± 2.2 mL,
p> 0.18) but higher than that of the SL-coated membrane
(246.6 ± 3.7 mL, p< 0.05, Fig. 1B). This result indicated that coating
Z-CNTs on the dense and smooth active layer side had a negligible ef-
fect on water transport, whereas coating on the porous support layer
resulted in a higher possibility of inner clogging, pore blockage, and a
more severe internal concentration polarization, leading to reduced
water extraction. It should be noted that the SL-coated membrane did
exhibit a comparable maximum water ux (11.96 ± 0.94 LMH,
Fig. 1C) to that of the pristine or AL-coated membranes (~12 LMH,
p> 0.46), as this parameter was predominantly determined by the
initial osmotic pressure gradient. In terms of RSF reduction, a notable
decrease of both RSF and SRSF was observed with the coated mem-
branes, compared to the pristine membrane (Fig. 1C), yielding an MR of
56.3% (AL-coated) and 28.1% (SL-coated), respectively.
The reduced RSF can be attributed primarily to the electrostatic
repulsion force induced by the charged functional groups from the Z-
CNTs on the membrane surface. For the pristine membrane, the diu-
sion of DS across the active layer (J
sA
) under the AL-Feed orientation
(i.e. FO mode) can be described using Eq. (3) (Fig. 2A) (Phillip et al.,
2010).
=− =− − =− −
J
DdC
dt
D
tCC D
tC() (0)
sAAA
Aa
if
A
Aa
i
(3)
where D
A
is the diusion coecient of DS in active layer, t
A
is the
thickness of active layer, C
ai
is the concentration of DS on the active
layer side of the supportive layer-active layer interface, and C
f
is the
concentration of draw solute in the feed (assuming C
fi
C
f
in an FO
system with negligible external concentration polarization) (Cath et al.,
2006). Initially, C
f
is zero (DI water as the feed). Once the membrane
surface is coated with Z-CNT, the negatively and positively charged
functional groups provide an electrostatic repulsion against anions and
cations that attempt to diuse across the coating layer, resulting in a
decrease in RSF (J
s,1A
<J
sA
and J
s,2A
<J
sA
, where subscripts 1 and 2
represent the AL- and SL-coated orientations, respectively). To accu-
rately measure the extent of net electrostatic repulsion, the Debye
length (κ
1
, nm) is introduced (Eq. (4), for a symmetric monovalent
electrolyte) (Russel et al., 1991). With decreasing Debye length, charges
are increasingly electrically screened, providing reduced electrostatic
repulsion force.
=×
κ
εεRT
FC210
r
10
320(4)
where ε
0
is the permittivity of free space, ε
r
is the dielectric constant, R
is the gas constant, Tis the temperature, Fis the Faraday constant, and
C
0
is the electrolyte concentration in molar units. Eq. (4) can be sim-
plied as the following equation in the water solution at room tem-
perature (~25 °C) (Israelachvili, 2011).
=
κ
I
(nm) 0.304
(M)
1
(5)
where Iis the ionic strength (molar unit) in the surrounding environ-
ment. This simplied Eq. (5) can further explain the SRSF reduction
dierence between the two coating orientations (i.e. J
s,1A
<J
s,2A
).
Once coated on the active layer, Z-CNTs are exposed to a less con-
centrated environment with a relatively low ionic strength, due to the
dilution eect of the clean water ux as well as rejection by the active
layer (C
c,1i
C
d
,C
c,1i
represents DS concentration at the interface of the
coating layer for an AL-coated membrane). Thus, a relatively large
Fig. 2. The proposed DS ion diusion pattern within (A) the pristine FO membrane; (B) the AL-coated membrane; and (C) the SL-coated membrane. The black line in
each gure represents the original DS ion diusion pattern, while the red and blue lines represent the DS ion diusion aected by electrostatic repulsion induced by
Z-CNT coating materials. Subscript 1 and 2 indicate DS diusion pattern in AL-coated and SL-coated membranes, respectively. (For interpretation of the references to
color in this gure legend, the reader is referred to the web version of this article.)
S. Zou, et al. Environment International 131 (2019) 104970
4
Debye length and an extensive electrostatic repulsion leads to enhanced
RSF reduction (Fig. 2B). Coating on the supportive layer, nonetheless,
allows Z-CNTs to be directly exposed to the highly concentrated draw
solution (C
c,2i
C
d
,C
c,2i
represents the DS concentration at the inter-
face of the coating layer in a SL-coated membrane), resulting in a
smaller Debye length, a weakened electrostatic repulsion force due to
the strong screening eect, and a smaller reduction in RSF (Fig. 2C).
Considering the overall system performance, the AL-coated membrane
was chosen for subsequent experiments.
3.2. Eects of Z-CNT coating density
Surface coating density (0, 0.10, 0.48, 0.97, and 1.45 g m
2
), as a
key operating parameter, was investigated to assess its eects on zeta
potential, membrane hydrophilicity, water ux, and RSF. In terms of
zeta potential, the pristine FO membrane exhibited a negative charge
due to abundant carboxylic groups (-COOH) on the TFC membrane
(Fig. 3A) (Tiraferri and Elimelech, 2012). A more negatively charged
surface (from 39.1 to 67.4 mV) was observed with higher pH due to
stronger deprotonation of the carboxylic groups (Jin et al., 2012).
Surface charge neutralization was observed with Z-CNT coating, con-
sistent with the results of a prior zwitterionic coating study (Wang
et al., 2018). With an increase of coating density, the zeta potential
became less negative. For example, at a pH of 5.7, the surface charge
increased from 60.9 mV (pristine membrane) to 46.2 mV
(1.45 g m
2
). Such a positive shift towards neutral surface charge could
hinder Donnan-facilitated cation transport (Sarkar et al., 2010) while
minimizing the diusion of counter anions to maintain solution elec-
troneutrality (Epsztein et al., 2018). For membrane hydrophilicity, the
dominance of the hydrophobic CNT backbone (74.8 wt%) over the
hydrophilic zwitterionic functional groups (25.2 wt%) resulted in
higher contact angles compared to that of the pristine membrane, even
at a low coating density of 0.10 g m
2
(Fig. 3B) (Tijing et al., 2016;
Wang et al., 2018). Increasing the coating density did not signicantly
aect membrane hydrophilicity, and the largest contact angle was ob-
served under 1.45 g m
2
(an average of 28.5 ± 4.1°).
The eect of coating density on water recovery was subsequently
investigated by using DI water as the feed (Fig. 3C). At lower coating
densities (0.100.97 g m
2
), the volumes of recovered water
(~247 mL) and water uxes (1013 LMH) were comparable to those of
the pristine membrane. This could result from a dynamic balance be-
tween a negligible mass transport resistance added owing to the Z-CNT
coating and enhanced water permeation via zwitterionic groups. A
further density increase to 1.45 g m
2
yielded in an increased re-
covered water volume (293.2 ± 1.5 mL, p< 0.01) and water ux
(13.8 ± 0.2 LMH, p< 0.01), suggesting that a higher density of sur-
face zwitterionic groups could facilitate water transport. The perfor-
mance of RSF reduction was closely linked to coating densities (Fig. 3D)
and eective surface electrostatic repulsion force. A lower coating
density provided limited surface electrostatic repulsion, rendering a
comparable RSF. With a higher coating density, a gradual decrease in
both RSF and SRSF was observed. The lowest RSF and SRSF were
3.41 ± 0.04 gMH and 0.281 ± 0.013 g L
1
, respectively, with a
coating density of 0.97 g m
2
, resulting in an MR of 55.5% compared to
Fig. 3. Comparison of FO performance and membrane characterization among various coating densities regarding (A) membrane zeta potential; (B) membrane
contact angle; (C) normalized water recover volume and maximum water ux to that of the pristine membrane; and (D) RSF and SRSF values. Membrane coating
densities include 0 (pristine membrane), 0.48, 0.97, and 1.45 g m
2
. In panels C and D, triplicate tests were performed for coated membrane at each coating density.
Error bars were calculated from triplicate test results by using the same membrane sample under each coating density. Testing conditions: 100-mL 1-M NaCl as the
draw and 500-mL DI water as the feed.
S. Zou, et al. Environment International 131 (2019) 104970
5
that of the pristine membrane (SRSF
o
of 0.631 ± 0.008 g L
1
). A fur-
ther increase of density to 1.45 g m
2
led to an increase in RSF
(4.77 ± 0.07 gMH) due to a higher water recovery volume (Fig. 3C)
and potentially diminished electrostatic repulsion via surface charge
neutralization (Fig. 3A). However, a similar SRSF
(0.299 ± 0.019 g L
1
,p> 0.45) was obtained to that of 0.97 g m
2
after normalizing RSF to water ux.
3.3. Eects of DS concentration and composition
To further investigate the relationship between ion concentration
and the eectiveness of the electrostatic repulsion force (Eq. (5)), a
series of DS concentrations (0.25, 0.50, 0.75, and 1.00 M NaCl) were
tested as a key operating parameter. A gradual decrease in the draw
concentration and osmotic driven force resulted in a reduced water ux
for both the pristine (J
w,o
) and the AL-coated membranes (J
w,m
,Fig. 4A).
Meanwhile, smaller amounts of NaCl leaked into the feed side with the
lower draw concentration (0.25-M NaCl), leading to a RSF of
1.98 ± 0.10 gMH for the AL-coated (J
s,m
) and 4.17 ± 0.16 gMH for
the pristine membranes (J
s,o
). A comparable SRSF was determined
under 0.501.00 mol L
1
NaCl for the AL-coated membrane (0.301,
0.309, and 0.308 g L
1
, respectively), indicating a similar C
c,1i
(DS level
at the coating-active layer interface, Fig. 2B) and comparable Debye
length (Eq. (5)). However, SRSF decreased to 0.279 ± 0.001 g L
1
(SRSF
m
) when 0.25-M NaCl was tested with an AL-coated membrane.
Hence, the Z-CNT coated FO membranes are more eective to elec-
trostatically repel ions at lower DS concentrations (Eq. (5)). It is worth
noting that a consistent SRSF was obtained for the pristine membrane
under all DS concentrations (0.500 ± 0.011 g L
1
,SRSF
o
), conrming
that the SRSF of a certain DS in a selected pristine membrane was only
aected by membrane intrinsic parameters (Phillip et al., 2010).
Various DSs were subsequently tested at 0.25 mol L
1
for SRSF re-
duction to evaluate the broad applicability of Z-CNTs. All the selected
DSs were ammonium-based fertilizers and have been studied in ferti-
lizer-driven FO (FDFO) to bypass energy-intensive DS regeneration/
separation, including NH
4
H
2
PO
4
(monoammonium phosphate, MAP),
(NH
4
)
2
HPO
4
(diammonium phosphate, DAP), NH
4
Cl, and thermolytic
NH
4
HCO
3
. Ammonium-based DSs were found to exhibit relatively large
SRSF in commercial FO membranes (Achilli et al., 2010), compared to
that of the benchmark NaCl (Fig. 4B). DAP, however, exhibited a much
lower SRSF due to the higher mole amount of phosphate ions and their
larger hydrated radii (Zou and He, 2017a). A signicantly reduced SRSF
was observed with the coated membrane (0.97 g m
2
) for all the DSs
(SRSF
m
,Fig. 4B), including MAP (0.110 ± 0.004 g L
1
, 83.8% MR),
DAP (0.060 ± 0.002 g L
1
, 74.5% MR), NH
4
Cl (0.394 ± 0.013 g L
1
,
70.8% MR), and NH
4
HCO
3
(0.279 ± 0.009 g L
1
, 61.9% MR). The
high degree of repulsion for multiple DS ions, especially nutrient ions
(e.g., N and P) could eectively prevent contamination of the feed so-
lution (e.g., brine water, groundwater, or wastewater) and reduce
subsequent polishing costs before nal discharge.
3.4. Membrane rejection and forward solute ux
FO has been recognized for its superior rejection of a variety of
contaminants (over 95%) in the feed stream, such as inorganic pollu-
tants (nutrients and heavy metals) (Xie et al., 2016), trace organic
compounds (pharmaceuticals and pesticides) (Madsen et al., 2015), and
microorganisms (Liu et al., 2013). However, minor permeation of exotic
contaminants via FSF could lead to enhanced system costs for down-
stream processes (e.g., fouling and/or scaling during pressure- or
thermal-separation) and potential environmental concerns if being di-
rectly reused (e.g., agriculture irrigation in FDFO). To quantify the
magnitude of contaminant permeation via FSF, a synthetic solution
containing common pollutant ions (each with a concentration of
100130 mg L
1
) was fed into an FO cell equipped with the pristine
membrane. The results indicated that multivalent ions (e.g., Mg
2+
and
PO
43
-P) exhibited signicantly less permeation than monovalent ions
(NO
2
-N and K
+
,Fig. 5A and Table S2), due to their relatively larger
hydrated ion radii and better size exclusion eect. It should be noted
that Ca
2+
was completed rejected (i.e., 100% rejection rate) by the
pristine membrane. Among the monovalent ions, cations (NH
4+
-N and
K
+
) demonstrated a higher propensity to diuse towards the draw side
than anions (NO
2
-N and NO
3
-N), owing to the negative membrane
charge and Donnan facilitated cation transport (similar to RSF). NH
4+
-
N had the lowest rejection rate (12.2 ± 1.2%) and the highest FSF
(1.68 ± 0.02 gMH). Hence, membrane rejection must be properly
addressed to promote water reuse.
Enhanced membrane rejection was realized by using the Z-CNT
coated FO membrane. The AL-coated membrane (0.97 g m
2
) exhibited
zero or negligible permeation of multivalent ions, with an FSF of 0 gMH
(i.e., 100% rejection, Fig. 5A) for Mg
2+
,Ca
2+
, and PO
43
-P, and an
FSF < 0.06 gMH for SO
42
-S (95.7 ± 0.2% rejection). Signicantly
improved rejection was also obtained for monovalent ions (Fig. 5A and
B), including NH
4+
-N (49.8% rejection, FSF of 0.90 gMH, or 46.3% FSF
reduction compared to pristine membrane), NO
2
-N (83.6% rejection,
FSF of 0.30 gMH, or 37.0% FSF reduction), NO
3
-N (72.5% rejection,
FSF of 0.54 gMH, or 30.3% FSF reduction), and K
+
(73.6% rejection,
FSF of 0.43 gMH, or 56.1% FSF reduction). The enhanced rejection and
reduced FSF could be attributed primarily to electrostatic repulsion as
the Z-CNT was in direct contact with the feed. It should also be noted
Fig. 4. Comparison of FO performance between pristine and coating-AL
membranes in terms of (A) various draw solution concentration (NaCl) and (B)
various draw solutes. The subscripts o and m in panel A represent pristine
membrane and AL-coated membrane (0.97 g m
2
), respectively. In panel B, the
concentration for all the draw solutions is 0.25 M, and the MR of each DS is
labelled beside the column (i.e., (1 SRSF
m
/ SRSF
o
) × 100%).
S. Zou, et al. Environment International 131 (2019) 104970
6
that domestic wastewater tends to have lower ion concentrations
(< 50 mg L
1
) than the synthetic solution used in this study
(100130 mg L
1
), resulting in elevated electrostatic repulsion forces
based on Eq. (5). Together with a desirable RSF reduction, Z-CNT
surface coatings have been proved to be eective in signicantly de-
creasing BSF for multiple draw or feed solutes, allowing FO to poten-
tially treat a variety of source waters (Lu et al., 2014).
3.5. Water ux consistency under semi-continuous operation
The lowest SRSF should be obtained by both minimizing DS pene-
tration (i.e., low RSF) and more importantly, obtaining a consistent
water ux under long-term operation (Eq. (1)). However, when fed
with real wastewater, gradual fouling on the membrane surface could
hinder ecient water transport through FO membrane while accel-
erating DS leakage via fouling-intensied concentration polarization
(She et al., 2012). Zwitterionic materials, either coated on membrane
surface or embedded inside the membrane, have been utilized to
maintain water ux consistency in previous studies (Lee et al., 2018;
Zhao et al., 2016). In this study, the Z-CNT coated FO membranes were
further examined under semi-continuous operation focusing on water
ux consistency and membrane fouling resistance (Fig. 6A). On day 1,
less water (184.4 mL and 3.76 LMH) was recovered in the FO equipped
with the AL-coated membrane compared to that of the pristine mem-
brane (230.7 mL and 4.92 LMH). The reduced water extraction via
coated membrane was likely due to the initial attraction of charged
substances in the wastewater and potential accumulation of charged
substances in the feed-membrane boundary layer, leading to additional
water transport resistance and reduced water ux. As the operation
continued, relatively consistent daily water recovery volumes and
water uxes were obtained with the coated membrane, while the
pristine membrane exhibited a dramatic decrease. By the end of day 12,
the AL-coated membrane exhibited only a 25.2% and 14.9% decrease in
water volume and ux, respectively, much lower than the pristine
membrane (50.4% and 54.5%, respectively). This desirable enhance-
ment of fouling resistance is attributed to several possible mechanisms
(He et al., 2016): (1) formation of a hydration shell assisted by zwit-
terionic materials on the membrane surface as a barrier to prevent di-
rect contact of foulants (Chen et al., 2005); (2) steric hindrance eects
induced by the chain of zwitterionic functional groups (Chen et al.,
2010); and (3) potential antimicrobial properties of the CNT (Tiraferri
et al., 2011), leading to an apparent visual dierence of fouling cov-
erage area on the membrane surface (Fig. 6B). Simple physical ushing
appears to be very eective in removing most foulants on both mem-
branes (day 13, Fig. 6B). On day 13 (after membrane cleaning), the
recovery eciency for the pristine membrane could reach 99.2%.
However, attened biofouling residues could still be observed under
SEM (50 K magnication). For the AL-coated membrane, a higher than
100% recovery was observed after membrane cleaning due to detach-
ment of Z-CNTs (Fig. 6B, SEM image after membrane cleaning). The
elemental mapping via EDS indicated that the remaining foulants on
the AL-coated membrane were marginally scattered inorganic scaling
(e.g., Ca
2+
,Fig. 6C), rather than biofouling.
3.6. Perspectives
Our results have collectively demonstrated eective control of bi-
directional solute ux and consistent water ux by coating Z-CNTs on
commercial FO membranes. To fully understand the contribution of
each coating element to BSF reduction and antifouling performance, a
direct membrane surface functionalization with zwitterionic groups and
a pure CNT coating (without functionalization process) should be per-
formed separately on commercial membranes and such a comparison
may better reveal the underlying mechanisms. Several key challenges
also need to be addressed to fully realize the potential of zwitterionic
materials for the elimination of bidirectional solute ux. First, more
robust approaches need to be employed to incorporate Z-CNTs onto the
TFC membrane. Natural adhesion via physical adsorption in this study
is one of the easiest, economically viable, and most scalable coating
methods. However, Z-CNTs bond with FO active layer only through van
der Waals force and may detach from the membrane during long-term
operation (especially under higher hydraulic sheer force). The detached
Z-CNTs can leave membrane surface together with accumulated fou-
lants, rendering temporarily enhanced fouling resistance. Other stra-
tegies have been exploited in previous studies to strengthen the bond
between zwitterionic materials and membrane surface, including
grafting zwitterionic polymers via polymerization, surface segregation,
and biomimetic adhesion (He et al., 2016). However, most of these
strategies were only assessed for their performance in anti-fouling re-
sistance rather than BSF mitigation. Alternatively, embedding zwitter-
ionic materials inside the TFC membrane can oer a more permanent
solution for possible detachment (Chan et al., 2016). Nonetheless, such
an approach requires high compatibility between the zwitterionic ma-
terial and the intrinsic membrane framework to maintain membrane
integrity. Second, other zwitterionic materials than Z-CNTs should be
explored for potential control of bidirectional solute ux in FO. The
current Z-CNTs do provide a good electrostatic repulsion force to repel
Fig. 5. Comparison of FO performance between pristine and AL-coated mem-
branes in terms of (A) membrane rejection and (B) FSF of various pollutant ions
in the feed. Synthetic solution, instead of DI water, was selected as the feed
solution. The subscripts o and m in panel B legend represent pristine membrane
and AL-coated membrane (0.97 g m
2
), respectively. The percentage represents
for FSF reduction eciency (i.e., (1 FSF
m
/ FSF
o
) × 100%). Error bars were
calculated from triplicate test results with the pristine and AL-coated mem-
branes.
S. Zou, et al. Environment International 131 (2019) 104970
7
feed and draw solutes, yet it decreases the membrane hydrophilicity
with larger surface contact angles. A hydrophilic polymer-based back-
bone, instead of a hydrophobic CNT, could be selected to carry the
zwitterionic functional groups. Finally, in the case when zwitterionic
materials (e.g., Z-CNT) detach from the membrane surface, their pre-
sence in the feed stream could lead to potential environmental con-
cerns. The toxicity and ecological eects of such substances need to be
comprehensively studied with further research.
4. Conclusions
In this study, zwitterion functionalized carbon nanotubes (Z-CNTs)
have been coated onto a commercial thin lm composite (TFC) mem-
brane to achieve BSF mitigation via electrostatic repulsion forces in-
duced by zwitterionic functional groups and steric interactions with
CNTs. The results have important implications for promoting high-
quality water recovery via FO and will inspire further development of
eective strategies for BSF and fouling control. The following conclu-
sions are reached:
Better mitigation of RSF was achieved (under the AL-Feed mode)
when coating Z-CNT on the active layer, rather than on the support
layer, likely due to more extended electrostatic repulsion in the
presence of lower ionic strength solutions.
With an optimal coating density of 0.97 g m
2
, a signicantly re-
duced specic RSF was observed for multiple draw solutes,
including NaCl (55.5% reduction), NH
4
H
2
PO
4
(83.8%), (NH
4
)
2
HPO
4
(74.5%), NH
4
Cl (70.8%), and NH
4
HCO
3
(61.9%).
FSF was notably reduced with fewer pollutants leaked to the draw
solution, including NH
4+
-N (46.3% reduction), NO
2
-N (37.0%),
NO
3
-N (30.3%), K
+
(56.1%), PO
43
-P (100%), and Mg
2+
(100%).
Successful BSF mitigation (both RSF and FSF) could allow FO to
potentially treat a variety of source waters with a wide spectrum of
DS.
When fed with real wastewater, a much more stable water ux (only
14.9% decrease) was achieved with the Z-CNT coated membrane
during a 12-day semi-continuous operation, compared to that of the
pristine membrane (54.5% ux decline). Nearly all membrane fou-
lants could be removed via simple physical ushing, rendering a
more robust and cost-eective FO operation.
This study is among the earliest eorts to address BSF control via FO
membrane modication, and the results warrants further eort to
explore alternative zwitterionic materials while strengthening the
bond between coating materials and membrane surface.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inu-
ence the work reported in this paper.
Fig. 6. Comparison of FO performance between pristine and AL-coated membranes under a semi-continuous operation in terms of (A) daily maximum water ux and
recovery volume; (B) visual images of membrane on day 12 (before membrane cleaning) and day 13 (after membrane cleaning with SEM characterization); and (C)
elemental mapping of AL-coated membrane (day 13) under EDS. In panel A, error bars were calculated by using the maximum water ux quantied at 20 min,
30 min, and 40 min of each cycle (23h). In panel C, the green and red color indicate sulfur from membrane framework and calcium from inorganic scaling. Raw
secondary euent from a local WWTP was collected as the feed solution. (For interpretation of the references to color in this gure legend, the reader is referred to
the web version of this article.)
S. Zou, et al. Environment International 131 (2019) 104970
8
Acknowledgements
This research was nancially supported by Institute for Critical
Technology and Applied Science, Virginia Tech. Shiqiang Zou was
partially supported by a Fellowship from Water INTERface IGEP at
Virginia Tech. We sincerely thank Mr. Li Wang (Vanderbilt University)
for his help with the analysis of membrane zeta potential and Virginia
Tech Open Access Subvention Fund for covering publication expense.
Appendix A. Supplementary data
The functionalization process and characterization of Z-CNT, sche-
matic of Z-CNT chemical structure (Fig. S1), surface coating density
(Fig. S2), composition of secondary euent from local WWTP (Table
S1), hydrated ion radii (Table S2), detailed water quality analysis
methods, and calculation equations for water ux, RSF, and FSF are
provided in Appendix A. Supplementary data. Supplementary data to
this article can be found online at doi:https://doi.org/10.1016/j.envint.
2019.104970.
References
Achilli, A., Cath, T.Y., Childress, A.E., 2010. Selection of inorganic-based draw solutions
for forward osmosis applications. J. Membr. Sci. 364 (1), 233241.
Akther, N., Sodiq, A., Giwa, A., Daer, S., Arafat, H.A., Hasan, S.W., 2015. Recent ad-
vancements in forward osmosis desalination: a review. Chem. Eng. J. 281, 502522.
Blandin, G., Verliefde, A.R., Tang, C.Y., Childress, A.E., Le-Clech, P., 2013. Validation of
assisted forward osmosis (AFO) process: impact of hydraulic pressure. J. Membr. Sci.
447, 111.
Cath, T.Y., Childress, A.E., Elimelech, M., 2006. Forward osmosis: principles, applica-
tions, and recent developments. J. Membr. Sci. 281 (1), 7087.
Chan, W.-F., Marand, E., Martin, S.M., 2016. Novel zwitterion functionalized carbon
nanotube nanocomposite membranes for improved RO performance and surface anti-
biofouling resistance. J. Membr. Sci. 509, 125137.
Chen, S., Zheng, J., Li, L., Jiang, S., 2005. Strong resistance of phosphorylcholine self-
assembled monolayers to protein adsorption:insights into nonfouling properties of
zwitterionic materials. J. Am. Chem. Soc. 127 (41), 1447314478.
Chen, S., Li, L., Zhao, C., Zheng, J., 2010. Surface hydration: principles and applications
toward low-fouling/nonfouling biomaterials. Polymer 51 (23), 52835293.
Epsztein, R., Shaulsky, E., Dizge, N., Warsinger, D.M., Elimelech, M.J.E.S., Technology,
2018. Role of ionic charge density in Donnan exclusion of monovalent anions by
nanoltration. Environ. Sci. Technol. 52 (7), 41084116.
Guo, H., Yao, Z., Wang, J., Yang, Z., Ma, X., Tang, C.Y., 2018. Polydopamine coating on a
thin lm composite forward osmosis membrane for enhanced mass transport and
antifouling performance. J. Membr. Sci. 551, 234242.
Hartanto, Y., Yun, S., Jin, B., Dai, S., 2015. Functionalized thermo-responsive microgels
for high performance forward osmosis desalination. Water Res. 70, 385393.
He, M., Gao, K., Zhou, L., Jiao, Z., Wu, M., Cao, J., You, X., Cai, Z., Su, Y., Jiang, Z., 2016.
Zwitterionic materials for antifouling membrane surface construction. Acta Biomater.
40, 142152.
Holt, J.K., Park, H.G., Wang, Y., Stadermann, M., Artyukhin, A.B., Grigoropoulos, C.P.,
Noy, A., Bakajin, O., 2006. Fast mass transport through sub-2-nanometer carbon
nanotubes. Science 312 (5776), 10341037.
Huang, L., Bui, N.-N., Meyering, M.T., Hamlin, T.J., McCutcheon, J.R., 2013. Novel hy-
drophilic nylon 6, 6 microltration membrane supported thin lm composite mem-
branes for engineered osmosis. J. Membr. Sci. 437, 141149.
Israelachvili, J.N., 2011. Intermolecular and Surface Forces. Academic Press.
Jin, X., Shan, J., Wang, C., Wei, J., Tang, C.Y., 2012. Rejection of pharmaceuticals by
forward osmosis membranes. J. Hazard. Mater. 227228, 5561.
Khalid, A., Abdel-Karim, A., Ali Atieh, M., Javed, S., McKay, G., 2018. PEG-CNTs nano-
composite PSU membranes for wastewater treatment by membrane bioreactor. Sep.
Purif. Technol. 190, 165176.
Kowalski, S.J., Szadzińska, J., Pawłowski, A., 2015. Ultrasonic-assisted osmotic dehy-
dration of carrot followed by convective drying with continuous and intermittent
heating. Dry. Technol. 33 (13), 15701580.
Kwon, S.-B., Lee, J.S., Kwon, S.J., Yun, S.-T., Lee, S., Lee, J.-H., 2015. Molecular layer-by-
layer assembled forward osmosis membranes. J. Membr. Sci. 488, 111120.
Lee, W., Goh, P., Lau, W., Ong, C., Ismail, A.J.S., Technology, P., 2018. Antifouling
zwitterion embedded forward osmosis thin lm composite membrane for highly
concentrated oily wastewater treatment. Sep. Purif. Technol. https://doi.org/10.
1016/j.seppur.2018.07.009.
Li, X., Sun, S., Badgley, B.D., He, Z., 2016. Long-term performance and microbial com-
munity characterization of an osmotic anammox system for removing reverse-uxed
ammonium. Bioresour. Technol. 211, 628635.
Liu, X., Qi, S., Li, Y., Yang, L., Cao, B., Tang, C.Y., 2013. Synthesis and characterization of
novel antibacterial silver nanocomposite nanoltration and forward osmosis mem-
branes based on layer-by-layer assembly. Water Res. 47 (9), 30813092.
Lu, Y., He, Z., 2015. Mitigation of salinity buildup and recovery of wasted salts in a hybrid
osmotic membrane bioreactorelectrodialysis system. Environ. Sci. Technol. 49 (17),
1052910535.
Lu, X., Boo, C., Ma, J., Elimelech, M., 2014. Bidirectional diusion of ammonium and
sodium cations in forward osmosis: role of membrane active layer surface chemistry
and charge. Environ. Sci. Technol. 48 (24), 1436914376.
Lutchmiah, K., Harmsen, D.J.H., Wols, B.A., Rietveld, L.C., Jianjun, Q., Cornelissen, E.R.,
2015. Continuous and discontinuous pressure assisted osmosis (PAO). J. Membr. Sci.
476, 182193.
Madsen, H.T., Bajraktari, N., Helix-Nielsen, C., Van der Bruggen, B., Søgaard, E.G., 2015.
Use of biomimetic forward osmosis membrane for trace organics removal. J. Membr.
Sci. 476, 469474.
McCutcheon, J.R., Elimelech, M., 2006. Inuence of concentrative and dilutive internal
concentration polarization on ux behavior in forward osmosis. J. Membr. Sci. 284
(1), 237247.
Mi, Y.-F., Zhao, F.-Y., Guo, Y.-S., Weng, X.-D., Ye, C.-C., An, Q.-F., 2017. Constructing
zwitterionic surface of nanoltration membrane for high ux and antifouling per-
formance. J. Membr. Sci. 541, 2938.
Nguyen, A., Azari, S., Zou, L., 2013. Coating zwitterionic amino acid l-DOPA to increase
fouling resistance of forward osmosis membrane. Desalination 312, 8287.
Park, C.H., Tocci, E., Fontananova, E., Bahattab, M.A., Aljlil, S.A., Drioli, E., 2016. Mixed
matrix membranes containing functionalized multiwalled carbon nanotubes: me-
soscale simulation and experimental approach for optimizing dispersion. J. Membr.
Sci. 514, 195209.
Phillip, W.A., Yong, J.S., Elimelech, M., 2010. Reverse draw solute permeation in forward
osmosis: modeling and experiments. Environ. Sci. Technol. 44 (13), 51705176.
Qiu, G., Law, Y.-M., Das, S., Ting, Y.-P., 2015. Direct and complete phosphorus recovery
from municipal wastewater using a hybrid microltration-forward osmosis mem-
brane bioreactor process with seawater brine as draw solution. Environ. Sci. Technol.
49 (10), 61566163.
Russel, W.B., Russel, W., Saville, D.A., Schowalter, W.R., 1991. Colloidal Dispersions.
Cambridge University Press.
Sarkar, S., SenGupta, A.K., Prakash, P., 2010. The Donnan membrane principle: oppor-
tunities for sustainable engineered processes and materials. Environ. Sci. Technol. 44
(4), 11611166.
Shannon, M.A., Bohn, P.W., Elimelech, M., Georgiadis, J.G., Marinas, B.J., Mayes, A.M.,
2008. Science and technology for water purication in the coming decades. Nature
452 (7185), 301311.
She, Q., Jin, X., Li, Q., Tang, C.Y., 2012. Relating reverse and forward solute diusion to
membrane fouling in osmotically driven membrane processes. Water Res. 46 (7),
24782486.
Tijing, L.D., Woo, Y.C., Shim, W.-G., He, T., Choi, J.-S., Kim, S.-H., Shon, H.K., 2016.
Superhydrophobic nanober membrane containing carbon nanotubes for high-per-
formance direct contact membrane distillation. J. Membr. Sci. 502, 158170.
Tiraferri, A., Elimelech, M., 2012. Direct quantication of negatively charged functional
groups on membrane surfaces. J. Membr. Sci. 389, 499508.
Tiraferri, A., Vecitis, C.D., Elimelech, M., 2011. Covalent binding of single-walled carbon
nanotubes to polyamide membranes for antimicrobial surface properties. ACS Appl.
Mater. Interfaces 3 (8), 28692877.
Wang, J., Wang, Z., Wang, J., Wang, S., 2015a. Improving the water ux and bio-fouling
resistance of reverse osmosis (RO) membrane through surface modication by
zwitterionic polymer. J. Membr. Sci. 493, 188199.
Wang, Y., Ou, R., Wang, H., Xu, T., 2015b. Graphene oxide modied graphitic carbon
nitride as a modier for thin lm composite forward osmosis membrane. J. Membr.
Sci. 475, 281289.
Wang, J., Xiao, T., Bao, R., Li, T., Wang, Y., Li, D., Li, X., He, T., 2018. Zwitterionic surface
modication of forward osmosis membranes using N-aminoethyl piperazine propane
sulfonate for grey water treatment. Process. Saf. Environ. Prot. 116, 632639.
Xie, M., Shon, H.K., Gray, S.R., Elimelech, M., 2016. Membrane-based processes for
wastewater nutrient recovery: Technology, challenges, and future direction. Water
Res. 89, 210221.
Yasukawa, M., Mishima, S., Shibuya, M., Saeki, D., Takahashi, T., Miyoshi, T.,
Matsuyama, H., 2015. Preparation of a forward osmosis membrane using a highly
porous polyketone microltration membrane as a novel support. J. Membr. Sci. 487,
5159.
Zhao, S., Zou, L., 2011. Relating solution physicochemical properties to internal con-
centration polarization in forward osmosis. J. Membr. Sci. 379 (1), 459467.
Zhao, D., Qiu, G., Li, X., Wan, C., Lu, K., Chung, T.-S., 2016. Zwitterions coated hollow
ber membranes with enhanced antifouling properties for osmotic power generation
from municipal wastewater. Water Res. 104, 389396.
Zhou, Q., Lei, X.-P., Li, J.-H., Yan, B.-F., Zhang, Q.-Q., 2014. Antifouling, adsorption and
reversible ux properties of zwitterionic grafted PVDF membrane prepared via
physisorbed free radical polymerization. Desalination 337, 615.
Zou, S., He, Z., 2016. Enhancing wastewater reuse by forward osmosis with self-diluted
commercial fertilizers as draw solutes. Water Res. 99, 235243.
Zou, S., He, Z., 2017a. Electrodialysis recovery of reverse-uxed fertilizer draw solute
during forward osmosis water treatment. Chem. Eng. J. 330, 550558.
Zou, S., He, Z., 2017b. Electrolysis-assisted mitigation of reverse solute ux in a three-
chamber forward osmosis system. Water Res. 115, 111119.
Zou, S., Yuan, H., Childress, A., He, Z., 2016. Energy consumption by recirculation: a
missing parameter when evaluating forward osmosis. Environ. Sci. Technol. 50 (13),
68276829.
Zou, S., Qin, M., He, Z., 2019. Tackle reverse solute ux in forward osmosis towards
sustainable water recovery: reduction and perspectives. Water Res. 149, 362374.
S. Zou, et al. Environment International 131 (2019) 104970
9
... Surface modification of FO membranes by polyethylenimine (PEI) effectively improved the ammonium rejection rate from 75.5% to 89.3% when treating return activated sludge (RAS), due to the enhanced surface hydrophilicity and positively charged PEI-modified FO membranes [52]. In another study, Zou et al. [53] reported a facile approach to fabricate a surface-modified TFC-FO membrane with zwitterion functionalised carbon nanotubes. The modified TFC-FO membrane showed reduced reverse solute flux due to more extended electrostatic repulsion in the presence of lower ionic strength solutions, improved fouling resistance, and stable water flux during the filtration of the secondary effluent from a local wastewater treatment plant. ...
Article
Full-text available
The application of membrane technologies for wastewater treatment to recover water and nutrients from different types of wastewater can be an effective strategy to mitigate the water shortage and provide resource recovery for sustainable development of industrialisation and urbanisation. Forward osmosis (FO), driven by the osmotic pressure difference between solutions divided by a semi-permeable membrane, has been recognised as a potential energy-efficient filtration process with a low tendency for fouling and a strong ability to filtrate highly polluted wastewater. The application of FO for wastewater treatment has received significant attention in research and attracted technological effort in recent years. In this review, we review the state-of-the-art application of FO technology for sewage concentration and wastewater treatment both as an independent treatment process and in combination with other treatment processes. We also provide an outlook of the future prospects and recommendations for the improvement of membrane performance, fouling control and system optimisation from the perspectives of membrane materials, operating condition optimisation, draw solution selection, and multiple technologies combination.
... Another important future application of f-CNTs is (suggested to be) water purification and forward osmosis [195,196], as charged functional groups on the CNTs' tips are shown to enhance ionic selectivity and salt rejection due to electrostatic repulsion [144]. Moreover, f-CNTs are able to purify water from organic toxins and heavy metals [27]. ...
Article
Full-text available
Confined water inside carbon nanotubes (CNTs) has attracted a lot of attention in recent years, amassing as a result a very large number of dedicated studies, both theoretical and experimental. This exceptional scientific interest can be understood in terms of the exotic properties of nanoconfined water, as well as the vast array of possible applications of CNTs in a wide range of fields stretching from geology to medicine and biology. This review presents an overreaching narrative of the properties of water in CNTs, based mostly on results from systematic nuclear magnetic resonance (NMR) and molecular dynamics (MD) studies, which together allow the untangling and explanation of many seemingly contradictory results present in the literature. Further, we identify still-debatable issues and open problems, as well as avenues for future studies, both theoretical and experimental.
... For example, physical separation via microfiltration, chemical precipitation, and biological methods was used for removing the solute transferred to the feed due to the reverse solute flux (Ferby et al., 2020). Other researchers suggested methods for alleviating reverse solute flux such as membrane surface functionalization (Zou et al., 2019) or the addition of surfactants to the OA (Chekli et al., 2018). ...
Article
Concentration of milk in the dairy industry is typically achieved by thermal evaporation or reverse osmosis (RO). Heat concentration is energy intensive and leads to cooked flavor and color changes in the final product, and RO is affected by fouling, which limits the final achievable concentration of the product. The main objective of this work was to evaluate forward osmosis (FO) as an alternative method for concentrating milk. The effects of fat content and temperature on the process were evaluated, and the physicochemical properties and sensory qualities of the final product were assessed. Commercially pasteurized skim and whole milk samples were concentrated at 4, 15, and 25°C using a benchtop FO unit. The FO process was assessed by monitoring water flux and product concentration. The color of the milk concentrates was also evaluated. A sensory panel compared the FO concentrated and thermally concentrated milks, diluted to single strength, with high temperature, short time pasteurized milk. The FO experimental runs were conducted in triplicate, and data were analyzed by single-factor ANOVA. Water flux during FO decreased with time under all processing conditions. Higher temperatures led to faster concentration and higher concentration factors for both skim and whole milk. After 5.75 h of FO processing, the concentration factors achieved for skim milk were 2.68 ± 0.08 at 25°C, 2.68 ± 0.09 at 15°C, and 2.36 ± 0.08 at 4°C. For whole milk, after 5.75 h of FO processing, concentration factors of 2.32 ± 0.12 at 25°C, 2.12 ± 0.36 at 15°C, and 1.91 ± 0.15 at 4°C were obtained. Overall, maximum concentration levels of 40.15% total solids for skim milk and 40.94% total solids for whole milk were achieved. Additionally, a triangle sensory test showed no significant differences between regular milk and FO concentrated milk diluted to single strength. This work shows that FO is a viable nonthermal processing method for concentrating milk, but some technical challenges need to be overcome to facilitate commercial utilization.
... Surface charge of the FO membrane can be modified by surface modification or developed from the beginning with membrane fabrication. Although, surface modification by alteration of its functional group can result in a more direct effect on the membrane performance compared to membrane fabrication [55,56]. The strategy formulated in this study was therefore carried out by simulating the surface charge modification employed under the fixed pH and cross-flow velocity of 16.7 cm/s, both membranes appeared to have a reduction of the J s /J w ratio as their negative charge was enhanced using our current model (Fig. 11). ...
Article
Perm-selectivity consisting of water flux Jw and solute flux Js or in form of Js/Jw ratio is an important parameter of designing Forward Osmosis (FO) membrane as it indicates the membrane performance and how much solute replenishment over the extracted pure water from the feed solution. Parameter Js/Jw ratio is dependent on hydrodynamic condition i.e cross-flow velocity (CFV), solute type i.e. diffusivity, trans-membrane surface potential. This study employed Cellulose Triacetate (CTA) membrane for representing low-charge membrane and Polyamide-Thin Film Composite (PA-TFC) for the membrane of highly negative surface charge. Six (6) different models were used to quantify the effect of external mass transfer, the ideal solution, non-ideal solution, trans-membrane surface potential on the transport of solute across the membrane. Our new model was proven to improve the prediction of perm-selectivity. The improvement was attributed to the application of trans-membrane potential-dependent solute partitioning for solute permeability correction and the application of experimentally obtained mass transfer coefficient. By the incorporation of the Donnan effect in determining the transmembrane potential, it was found that in the low charge membrane namely CTA, the dominant transport mechanism was diffusion, while in a highly negative surface charged membrane namely TFC, the partition of solute to be the dominant mechanism. Operation at a low CFV posed less impact of the membrane charge. The newly developed model provided a good foundation for FO process design under different CFVs and modified surface charges.
... Zwitterion functionalized carbon nanotubes (Z-CNTs) coated onto a commercial Aquaporin Inside ® FO thin film composite (TFC) membrane have been reported by Zou et al. [77]. They obtained BSF mitigation through zwitterion-induced repulsive electrostatic interaction along with the carbon nanotube-induced steric interaction. ...
Article
Full-text available
Forward osmosis (FO) is an important desalination method to produce potable water. It was also used to treat different wastewater streams, including industrial as well as municipal wastewater. Though FO is environmentally benign, energy intensive, and highly efficient; it still suffers from four types of fouling namely: organic fouling, inorganic scaling, biofouling and colloidal fouling or a combination of these types of fouling. Membrane fouling may require simple shear force and physical cleaning for sufficient recovery of membrane performance. Severe fouling may need chemical cleaning, especially when a slimy biofilm or severe microbial colony is formed. Modification of FO membrane through introducing zwitterionic moieties on the membrane surface has been proven to enhance antifouling property. In addition, it could also significantly improve the separation efficiency and longevity of the membrane. Zwitterion moieties can also incorporate in draw solution as electrolytes in FO process. It could be in a form of a monomer or a polymer. Hence, this review comprehensively discussed several methods of inclusion of zwitterionic moieties in FO membrane. These methods include atom transfer radical polymerization (ATRP); second interfacial polymerization (SIP); coating and in situ formation. Furthermore, an attempt was made to understand the mechanism of improvement in FO performance by zwitterionic moieties. Finally, the future prospective of the application of zwitterions in FO has been discussed.
Article
Full-text available
Membrane technology has rapidly gained popularity in wastewater treatment due to its cost-effectiveness, environmentally friendly tools, and elevated productivity. Although membrane performance in wastewater treatment has been reviewed in several past studies, the key techniques for improving membrane performance, as well as their challenges, and solutions associated with the membrane process, were not sufficiently highlighted in those studies. Also, very few studies have addressed hybrid techniques to improve membrane performance. The present review aims to fill those gaps and achieve public health benefits through safe water processing. Despite its higher cost, membrane performance can result in a 36% reduction in flux degradation. The issue with fouling has been identified as one of the key challenges of membrane technology. Chemical cleaning is quite effective in removing accumulated foulant. Fouling mitigation techniques have also been shown to have a positive effect on membrane photobioreactors that handle wastewater effluent, resulting in a 50% and 60% reduction in fouling rates for backwash and nitrogen bubble scouring techniques. Membrane hybrid approaches such as hybrid forward-reverse osmosis show promise in removing high concentrations of phosphorus, ammonium, and salt from wastewater. The incorporation of the forward osmosis process can reject 99% of phosphorus and 97% of ammonium, and the reverse osmosis approach can achieve a 99% salt rejection rate. The control strategies for membrane fouling have not been successfully optimized yet and more research is needed to achieve a realistic, long-term direct membrane filtering operation.
Article
Reverse solute flux (RSF) is a key issue for operating forward osmosis (FO) systems and can cause the loss of draw solute (DS) and salt accumulation in the feed. Herein, an electrolysis-assisted FO (e-FO) system was developed for simultaneous RSF reduction and recovery of the reverse-fluxed DS. Applying a voltage of 1.5 V led to RSF of 3.34 ± 0.01 mmol m⁻² h⁻¹ (0.47 g m⁻² h⁻¹) in the e-FO system, 67.5 ± 0.5% lower than that of the control system; in addition, the e-FO system recovered ~0.32 g L⁻¹ of the reverse-fluxed DS and this could not be realized in the control. When the e-FO system was examined with three types of the mimicked fertilizer, RSF reduction and DS recovery were largely affected by the individual components of a fertilizer DS. The energy consumption of the e-FO system was reduced by ~90%, from 0.38 ± 0.01 to 0.04 ± 0.01 kWh m⁻³ when the recirculation rate decreased from 60 to 15 mL min⁻¹. Those results have demonstrated the technical feasibility of an e-FO system that is capable of reducing RSF and recovering the lost DS and will encourage further investigation by addressing several identified challenges.
Chapter
With industrial and technological growth, the generation of waste has increased significantly in the last decades. The need arises for new techniques and methodologies more effective in treating waste generated, especially from contaminated water. Nanomaterials have been developed to treat contaminated water in various parts of the world to meet the demand for pure water. Carbon nanotubes (CNTs) have aroused researchers' interest in developing new technologies and devices that replace or enhance existing materials. They are applied in filtration devices because they have good water permeability and are selective for different substances. Their functionalization can enhance the ability to remove salts and other membrane properties based on CNT. In this work, studies focused on using functionalized CNT in the treatment and purification of water will be presented.
Article
The membrane based separation technology has been widely applied in water treatment, due to its high efficiency and lower energy consumption. Compared with the traditional polyamide membrane, the novel non-polyamide membrane possesses unique physicochemical structures and surface properties. The introduction of carbon based nanomaterials further endows the membrane with excellent separation performances and resistance to the various feed water and long utilization duration. This paper intends to review the state-of-the-art application of carbon based nanomaterials on the non-polyamide desalination membrane, including carbon nanotubes (CNTs), graphene, carbon quantum dots (CQDs), porous carbon and activated carbon. The exclusive properties of carbon based nanomaterials are discussed in the initial section of the review. The novel carbon based free-standing membranes without supporting layer for desalination are comprehensively addressed. The focus is placed on reporting the utilization of carbon based nanomaterials in the fabrication of the different non-polyamide membranes, including reverse osmosis (RO) membrane, forward osmosis (FO) membrane, nanofiltration (NF) membrane, membrane distillation (MD) membrane and pervaporation (PV) membrane. The effects on the non-polyamide membrane performances including water flux, salt rejection, anti-fouling, mechanical properties, and chlorine resistance by incorporating the carbon nanomaterials are detailedly summarized and analyzed. We also discuss the relationship between the membrane microstructure such as chemical modification, porosity and hydrophilicity and key performance mentioned above. Furthermore, the development challenges and research opportunities in this field are also proposed from the aspects of technological challenges, commercialization challenges and environmental impacts/sustainability.
Article
Forward osmosis (FO) has emerged as a potentially energy-efficient membrane treatment technology to yield high-quality reusable water from various wastewater/saline water sources. A key challenge remained to be solved for FO is reverse solute flux (RSF), which can cause issues like reduced concentration gradient and loss of draw solutes. Yet no universal parameters have been developed to compare RSF control performance among various studies, making it difficult to position us in this “battle” against RSF. In this paper, we have conducted a concise review of existing RSF reduction approaches, including operational strategies (e.g., pressure-, electrolysis-, and ultrasound-assisted osmosis) and advanced membrane development (e.g., new membrane fabrication and existing membrane modification). We have also analyzed the literature data to reveal the current status of RSF reduction. A new parameter, mitigation ratio (MR), was proposed and used together with specific RSF (SRSF) to evaluate RSF reduction performance. Potential research directions have been discussed to help with future RSF control. This review intends to shed more light on how to effectively tackle solute leakage towards a more cost-effective and environmental-friendly FO treatment process.
Article
The raising oil consumption in oil and gas industries has exacerbated the disposal of oil waste into various water streams. This phenomenon has called for treatments to prevent threats to the human and environment. With some great advantages such as lower membrane fouling rate, lower energy requirement and higher water recovery rate compared to the conventional pressure-driven membrane processes, forward osmosis (FO) has been recognized as a potential candidate for oily wastewater treatment. In this study, a poly[3-(N-2-methacryloylxyethyl-N,N-dimethyl)-ammonatopropanesulfonate] (PMAPS) incorporated thin film composite (TFC) membrane with excellent anti-fouling properties was fabricated for oily wastewater through forward osmosis process. PMAPS was blended with polyethersulfone (PES) dope solution and cast into PES support layer. Interfacial polymerization (IP) technique was applied to form a thin polyamide (PA) layer atop of the PES support layer. The PMAPS incorporated TFC membranes were characterized for their morphology and surface hydrophilicity. The resultant 1% PMAPS-TFC membrane exhibited high water flux of 15.79 ± 0.3 L/m².h and oil flux of 12.54 ± 0.8 L/m².h when tested in FO mode for oil removal from oily wastewater using 1000 ppm emulsified oily solution as feed solution and 2 M NaCl as draw solution. The oil rejection up to 99% was also obtained. Most significantly, PMAPS incorporated TFC membrane outperformed neat TFC membrane with lower fouling propensity for oily waste treatment. When treating 10000 ppm oil emulsion, PMAPS-TFC was able to achieve average flux recovery rate of 97% while neat TFC only able to achieve 70.8% of average flux recovery rate.
Article
Zwitterionic amide monomer (N-aminoethyl piperazine propane sulfonate, AEPPS) was used to modify the active layer of the thin-film composite (TFC) Forward Osmosis Membranes (FOMs) by either adding into the water phase before the interfacial polymerization (route 1) or grafting to the initial active separation layer after the interfacial polymerization (route 2). Their separation performance and anti-fouling property were investigated. Different from the literature results, the synthesized AEPPS was a mixture of isomers which is too complicated to be purified. Results showed that both fourier transform infrared spectroscopy and X-ray photo-electron spectroscopy confirmed that the zwitterionic materials were successfully incorporated into the FOMs. However, in a long-term FO fouling test by using grey wastewater as the feed solution, surface grafted zwitterionic FO membrane prepared after the interfacial polymerization (route 2) showed superior stable FO performance than the membrane prepared by imbedding the AEPPS inside the active layer (route 1). Imbedded AEPPS chemical inside the active layer could not prevent the adsorption of foulant, thus no relief in the fouling was observed. The results demonstrated that the zwitterionic material surface grafting after the interfacial polymerization way is a more efficient approach to prepare fouling resistant FOMs for treating greywater.
Article
The main objective of this study is to examine how the charge density of four monovalent anions  fluoride (F-), chloride (Cl-), bromide (Br-), and nitrate (NO3-)  influences their Donnan (charge) exclusion by a charged nanofiltration (NF) membrane. We systematically studied the rejection behavior of ternary ion solutions containing sodium cation (Na+) and two of the monovalent anions as a function of pH with a polyamide NF membrane. In the solutions containing F- and Cl- or F- and Br-, F- rejection was higher than Cl- or Br- rejection only when the solution pH was higher than 5.5, suggesting that F- (which has higher charge density) was repelled more strongly by the negatively charged membrane. The order of change in the activation energy for the transport of the four anions through the polyamide membrane as a response to the increase of the membrane negative charge was the following: F- > Cl- > NO3- > Br-. This order corroborates our main hypothesis that an anion with a smaller ionic radius, and hence a higher charge density, is more affected by the Donnan-exclusion mechanism in NF. We conclude with a proposed mechanism for the role of ionic charge density in the rejection of monovalent anions in NF.
Article
Forward osmosis (FO) has been widely studied for desalination or water recovery from wastewater, and one of its key challenges for practical applications is reverse solute flux (RSF). RSF can cause loss of draw solutes, salinity build-up and undesired contamination at the feed side. In this study, in-situ electrolysis was employed to mitigate RSF in a three-chamber FO system (“e-FO”) with Na2SO4 as a draw solute and deionized (DI) water as a feed. Operation parameters including applied voltage, membrane orientation and initial draw concentrations were systematically investigated to optimize the e-FO performance and reduce RSF. Applying a voltage of 1.5 V achieved a RSF of 6.78 ± 0.55 mmol m−2 h−1 and a specific RSF of 0.138 ± 0.011 g L−1 in the FO mode and with 1 M Na2SO4 as the draw, rendering ∼57% reduction of solute leakage compared to the control without the applied voltage. The reduced RSF should be attributed to constrained ion migration induced by the coactions of electric dragging force (≥1.5 V) and high solute rejection of the FO membrane. Reducing the intensity of the solution recirculation from 60 to 10 mL min−1 significantly reduced specific energy consumption of the e-FO system from 0.693 ± 0.127 to 0.022 ± 0.004 kWh m−3 extracted water or from 1.103 ± 0.059 to 0.044 ± 0.002 kWh kg−1 reduced reversed solute. These results have demonstrated that the electrolysis-assisted RSF mitigation could be an energy-efficient method for controlling RSF towards sustainable FO applications.
Article
Fouling on pressure-retarded osmosis (PRO) membranes leads to severe declines in water flux and power density because their porous substrates are facing the wastewater feed. Thus, inorganics, organics and microorganisms in the wastewater are prone to depositing on the substrate surface and even in its pores. In order to reduce the fouling propensity, coating the substrate surface of PRO membranes with zwitterionic materials proves to be an effective way. In this work, 2-methacryloyloxyethylphosphorylcholine (MPC), is modified and grafted onto the polydopamine (PDA) coated poly (ether sulfone) (PES) hollow fiber substrate. Both the synthesis and surface coating of MPC are easy and facile to be scaled up. Compared with the pristine PES and PES-PDA substrates, the MPC modified substrate (PES-PDA-MPC) exhibits high resistance to protein adsorption as well as bacteria adhesion. By using a state-of-the-art thin-film composite poly (ether sulfone) (TFC-PES) hollow fiber membrane as the control for power generation, the power density of the TFC-PES-PDA-MPC membrane can achieve as high as 7.7 W/m(2) while the unmodified one has only 6.0 W/m(2) after 3 h's PRO tests. In conclusion, the osmotic power generation of PRO membranes can be significantly sustained by modifying the membrane surface with zwitterions.