w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: www.elsevier.com/locate/watres
Evaluation of poly (aspartic acid sodium salt) as a
draw solute for forward osmosis
Gimun Gwak a,1, Bokyung Jung b,1, Sungsoo Han b, Seungkwan Hong a,*
a
School of Civil, Environmental & Architectural Engineering, Korea University, 1-5 Ga, Anam-Dong, Sungbuk-Gu,
Seoul 136-713, Republic of Korea
b
Energy Laboratory, SAIT, Samsung Electronics, 130 Samsung-ro, Suwon-si, Gyeonggi-do 443-803, Republic of
Korea
article info
abstract
Article history:
Poly (aspartic acid sodium salt) (PAspNa) was evaluated for its potential as a novel draw
Received 8 January 2015
solute in forward osmosis (FO). The inherent advantages of PAspNa, such as good water
Received in revised form
solubility, high osmotic pressure, and nontoxicity, were first examined through a series of
28 April 2015
physicochemical analyses and atomic-scale molecular dynamics simulations. Then, lab-
Accepted 29 April 2015
scale FO tests were performed to evaluate its suitability in practical processes. Compared
Available online 14 May 2015
to other conventional inorganic solutes, PAspNa showed comparable water flux but
significantly lower reverse solute flux, demonstrating its suitability as a draw solute.
Keywords:
Moreover, fouling experiments using synthetic wastewater as a feed solution demon-
Forward osmosis (FO)
strated that PAspNa reversely flowed to the feed side reduced inorganic scaling on the
Draw solution
membrane active layer. The recyclability of PAspNa was studied using both nanofiltration
Antiscaling
(NF) and membrane distillation (MD) processes, and the results exhibited its ease of re-
Wastewater reclamation
covery. This research reported the feasibility and applicability of FO-NF or FO-MD pro-
Desalination
cesses using PAspNa for wastewater reclamation and brackish water desalination.
© 2015 Elsevier Ltd. All rights reserved.
1.
Introduction
Forward osmosis (FO) has gained recognition over the last
decade as an emerging technology for water treatment and
desalination (Cath et al., 2006; Boo et al., 2013; Shannon et al.,
2008; Valladares Linares et al., 2013). In contrast to pressuredriven membrane processes, such as reverse osmosis (RO),
this process utilizes an osmotic gradient between two streams
separated by a semi-permeable membrane as the driving
force for moving water into the higher osmotic potential draw
solution. Consequently, the performance of FO mainly
* Corresponding author. Tel.: þ82 2 3290 3322; fax: þ82 2 928 7656.
E-mail address: skhong21@korea.ac.kr (S. Hong).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.watres.2015.04.041
0043-1354/© 2015 Elsevier Ltd. All rights reserved.
depends on the osmotic pressure exerted by the draw solution
among many other factors.
Many researchers have demonstrated the feasibility of FO
technology based on its special characteristics, i.e., low energy
consumption, high rejection of a wide range of contaminants
and easy fouling control (Holloway et al., 2007; Kim et al.,
2012). However, FO is still a relatively immature technology
and some remaining limitations, such as low membrane
performance, hinder its industrial application. One of the key
challenges to the further development of FO is identifying a
suitable draw solute. The development of an optimized draw
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
solute would be a powerful breakthrough for the improvement of FO performance.
The ideal draw solute should be able to provide significantly improved FO performance, and separated easily and
completely by means of a low cost recovery method. In this
sense, three major criteria can be applied to evaluate the
suitability of a draw solute: (1) high water flux; (2) low reverse
draw solute flux; and (3) easy and low-cost recovery from the
diluted draw solution. Additional points, such as water solubility, nontoxicity, and the potential for commercialization,
should be also considered.
Various materials, including conventional inorganic salts,
organic solutes, and newly developed materials, have been
studied and evaluated in the search for an optimized draw
solute. In previous work, traditional draw solutes based on
inorganic salts, such as NaCl and MgCl2, have been extensively studied in various FO fields (Holloway et al., 2007; Achilli
et al., 2010). The draw solutions made from inorganic solutes
produce reasonably high water flux and can be readily separated by RO. However, both energy requirement of RO recovery and corresponding solute leakage in FO have been found
to be high, leading to concerns about operational costs.
To overcome these disadvantages, many novel draw solutes have been proposed and evaluated. Especially, some
types of single molecular organic salts and polyelectrolytes,
such as EDTA sodium salts, sodium polyacrylate (PAANa), and
sodium lignin sulfonate (NaLS) have been investigated as potential draw solutes, taking advantage of their large molecular
size (Hau et al., 2014; Duan et al., 2014; Ge et al., 2012; Yen
et al., 2010; Stone et al., 2013). These types of solutes not
only generated good water flux but also reduced reverse solute
flux significantly. The regenerations of these salts, especially
EDTA sodium salts and PAANa, were also successfully achieved by low-pressure nanofiltration (NF) (Hau et al., 2014) or
ultrafiltration (UF) (Ge et al., 2012), respectively. Despite good
performance in the FO process and recovery system, these
solutes were ultimately found to be impractical due to drawbacks such as commercial availability.
Water-soluble magnetic nanoparticles (MNPs) have also
been investigated as a novel draw solute based on their
superparamagnetic properties. Various hydrophilic single
molecule- or oligomer-coated MNPs such as 2-pyrrolidoneMNP, triethylene glycol-MNP, poly(acrylic acid)-MNP, and
poly(ethylene glycol)diacid-MNP were synthesized to incorporate both large molecular size and magnetic properties and
evaluated for their potential as draw solutes (Ling et al., 2010;
Ge et al., 2011; Na et al., 2014). The large molecular size of these
MNPs is led to better prevention of reverse solute flux but the
coating materials on the MNPs adversely affected the magnetic properties. Therefore, this resulted in a low separation
rate and the separated MNPs showed irreversible aggregation.
The combination of these factors caused difficulty in the recovery and sustainability for these materials. Moreover, the
risk assessment for human health and the environmental
hazards of synthesized nanomaterials are an ongoing concern
in various research fields.
Draw solutes based on thermosensitive materials with
temperature dependent phase separation behavior have
received increasing attention (Ling et al., 2011; Noh et al., 2012;
Zhao et al., 2014; Cai et al., 2013). Solutes such as N-acylated
295
polyethyleneimine (PEI) derivatives and poly (sodium styrene4-sulfonate-co-n-isopropylacrylamide) (PSSS-PNIPAM) are
soluble in water under their lower critical solution temperature (LCST), but agglomeration of polymer chains occurs
above the LCST (Noh et al., 2012; Zhao et al., 2014). Based on
this thermoresponsivity, the diluted draw solutes can be
easily concentrated by thermal recovery methods such as
membrane distillation (MD). In spite of all these pioneering
work, a practical solution to the low performance of FO due to
their low osmotic pressure has remained elusive. Despite the
enormous research efforts undertaken in terms of these
various solutes, each has had serious drawbacks, and therefore, the search for a practical and optimized draw solute
must continue if the successful application of FO is to be
achieved.
Considering the aforementioned requirements for a draw
solute, poly (aspartic acid sodium salt) (PAspNa) is a more
suitable candidate than the previously studied solutes.
PAspNa, a kind of polyelectrolyte commonly used as an
antiscalant, has many properties which suggest it may have
potential as an alternative draw solute of FO. It has significantly large molecular size and good water solubility.
Furthermore, it is non-toxic and commercially available with
reasonable unit price. As these various merits of PAspNa had
been verified through our autonomous evaluation procedure
for screening draw solutes, in order to show the potential of
PAspNa as a new draw solute systematically, this paper is
organized with detailed scrutiny for physicochemical analyses and practical performance measurements. The evaluation flow for selecting new draw solute was composed of
four stages and all the criteria of each level were categorized
in Fig. 1.
To assess the suitability of PAspNa, we first analyzed the
physicochemical properties of PAspNa as a draw solute, specifically its molecular characteristics, water solubility, osmotic pressure, viscosity, and toxicity. The basic performance
aspects of PAspNa as a draw solute in FO, such as its water flux
and reverse solute flux, were examined using deionized (DI)
water as the feed solution. The effect of reverse PAspNa flux
on inorganic fouling was also assessed through a scaling
experiment using inorganic synthetic wastewater as the feed
solution. Finally, recyclability studies were explored using
nanofiltration (NF) and membrane distillation (MD) processes.
2.
Materials and methods
2.1.
Preparation of PAspNa
Commercially available PAspNa (CAS-No. 181828-06-8) in
powder form was obtained from Lanxess AG (Cologne, Germany). For the characterization analyses, including molecular
structure and toxicity, purified PAspNa was used, and the
molecular structural data obtained was used to inform the
molecular dynamics study of its solvation behavior in water.
Unwanted NaOH was removed from the PAspNa through
rigorous washing with MeOH/distilled water (1:9 vol%) followed by drying in vacuo at 80 C. For all the FO and recovery
tests, unrefined PAspNa was used as supplied, in order to
directly evaluate its real-world applicability.
296
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
monomer units. The syndiotactic configuration was found to
have the lowest energy, and was verified by ab initio calculations using a short polymer strand. The molecular weight
used for the commercially available PAspNa was informed by
the GPC results, which corresponded to approximately 10
monomer units. For simplification of chemical structure,
PAspNa was consisted of only a-form of L-aspartic acid units
(See Fig. S1). For more details of the simulation methodology is
described in the supplementary information.
2.4.
Cell culture and cellular toxicity test
A colorimetric assay using 3-(4, 5-Dimethylthiazol-2-yl)-2, 5diphenyltetrazolium bromide (MTT) was performed to
examine the cellular toxicity of PAspNa. Normal human fibroblasts from neonatal foreskin, obtained from the dermatology unit of Ajou University Hospital in the Republic of
Korea, were used as target cells. The cells were grown as
adherent culture and maintained in Dulbecco's modified eagle's medium (DMEM) supplemented with 10% fetal bovine
serum, 100 U/mL penicillin (Gibco BRL), and 100 mg/mL streptomycin (Gibco BRL) at 37 C in a humidified atmosphere with
5% CO2. For testing, cells at a concentration of 2 104 cells per
well were seeded in a 24-well plate. After 24 h, PAspNa dissolved in serum-free DMEM at different concentrations was
introduced to the cells, which were then incubated at 37 C in
a humidified atmosphere with 5% CO2 for 72 h. The control
cells were treated with the same amount of serum-free
DMEM. Then, MTT solution (1 mg/mL, PBS) with one tenth of
the total volume of the cell media was added. The cells were
then incubated under the same conditions for a further 4 h.
After removing the media, 500 mL of DMSO was added to
dissolve the precipitated formazan crystals formed in the
well. The absorbance of the sample proportional to cellular
viability was measured spectrophotometrically at 540 nm
with a background absorbance at 650 nm by a microplate
reader (PowerWave X, Bio-Tek Inc., VT, USA).
Fig. 1 e Evaluation flow for selecting new draw solute.
2.2.
Physicochemical properties
The molecular weight of PAspNa was measured by aqueous
gel permeation chromatography (GPC) (Breeze System, Waters, USA) using 0.02 N NaNO3 as the eluent at a flow rate of
0.8 mL/min. The amount of sodium in PAspNa was quantified
with an inductively coupled plasma-atomic emission spectrometer (ICP-AES) (ICPS-8100, Shimadzu, Japan). The viscosity and osmolality of PAspNa were measured using an SV-10
Vibro Viscometer (A&D Company LTd., Japan) and Osmomat
030 (Gonotec GmbH, Germany), respectively.
2.3.
Atomistic model of draw solute for solvation
behavior study
In the simulations, the chemical structure of the PAspNa was
represented by a single chain polymer strand in the syndiotactic configuration, with a degree of polymerization (n) of 10
2.5.
FO process
All of the FO experiments were conducted with a lab-scale FO
unit as described elsewhere (Boo et al., 2013; Kim et al., 2012;
Boo et al., 2012). A custom-built cross-flow FO cell with symmetric rectangular channels (7.7 cm in length, 2.6 cm in width
and 0.3 cm in height) on both feed and draw sides was used for
the co-current flows of the feed and draw solutions. The crossflow velocity of both solutions was fixed at 8.5 cm/s and the
temperature was maintained at 25 ± 0.5 C. Two types of flat
sheet membranes made with different materials were used,
namely, asymmetric cellulose triacetate (CTA) and thin film
composite (TFC) membranes (Hydration Technology Inc.,
Albany, OR, USA). The membrane coupons were inserted in
the membrane cell in two different orientations; active layer
facing the feed solution (AL-FS) mode, and active layer facing
the draw solution (AL-DS) mode. PAspNa draw solutions of
different concentrations in DI water were prepared at an
initial volume of 1 L. Two different types of solution, i.e., DI
water and inorganic synthetic wastewater, were used as the
feed solution. DI water was used to evaluate basic FO performance such as water flux and reverse solute flux, and
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
inorganic synthetic wastewater was employed for the fouling
experiments. The chemical composition of the synthetic
wastewater shown in Table 1 is based on the secondary
effluent quality from selected wastewater treatment plants in
California, USA (Boo et al., 2013; Huertas et al., 2008). All feed
solutions were prepared at the same initial volume of the
draw solution, i.e., 1 L. The water flux, Jw, given in units of liters per square meter per hour (LMH), was calculated by
recording the mass change in the draw solution reservoir
versus time, thus:
Jw ¼
DV
ADt
(1)
where DV (L) is the volume change over the time interval Dt
(h), and A (m2) is the effective membrane surface area.
The reverse solute flux, Js, given in units of grams per
square meter per hour (gMH), was determined by converting
conductivity and total organic carbon of the feed DI water
using Equation (2).
Js ¼
DðCt Vt Þ
ADt
NF process
A lab-scale NF system similar to that reported in our earlier
publications (Ju and Hong, 2014) was used for the recyclability
experiments. TFC flat sheet membranes (model NF270-4040D)
with a 200e300 Da molecular weight cut-off (MWCO) from
Dow Chemical Company were used. For stable operation, a
porous stainless steel spacer 1.75 mm thick was placed in the
feed channel. The membrane cell specifications and flow
conditions were the same as those of the FO experiments. The
feed solutions were PAspNa solutions ranging from 0.01 to
0.03 g/mL. Hydraulic pressure was applied to keep a constant
pressure of 20 bar, and the temperature of the feed PAspNa
solution was maintained at 20 ± 0.5 C. The water flux was
calculated with Equation (1), but the mass change in the
Table 1 e Chemical composition of the synthetic
wastewater solution used in the inorganic fouling
experiments (Boo et al., 2013).
Compound
Sodium citrate
Ammonium chloride
Potassium
phosphate
Calcium chloride
Sodium bicarbonate
Sodium chloride
Magnesium sulfate
permeated water was measured instead of the draw solution.
The salt rejection was computed with Equation (3).
R¼
Cp
1
100
CF
(3)
where R (%) is the salt rejection, Cp (g/L) is the solute concentration of the permeate water, and CF (g/L) is the solute concentration of the feed solution.
2.7.
MD process
MD experiments were carried out using the same lab-scale
set-up as for the FO tests during 3 h. The effective membrane area, flow rate, and direction were identical to those of
the FO experiments. Polytetrafluoroethylene (PTFE) flat sheet
membrane (EMD Millipore, Germany) of 0.2 mm pore size was
used. The PAspNa feed solution was prepared and maintained
at 60 ± 0.5 C, while the permeate solutions were cold DI water
maintained at 20 ± 0.5 C. The water flux was calculated with
Equation (1).
(2)
where Ct (g/L) and Vt (L) are the reverse solute concentration
and the volume of feed solution, respectively, at an arbitrary
time t. The conductivity was measured using a conductivity
meter (Hach-Lange, UK), and the amount of total organic
carbon was analyzed with a TIV-V CPH (Shimadzu, Japan).
2.6.
297
Molecular
weight
(g/mol)
Concentration
(mM)
294.09
53.49
136.09
1.16
0.94
0.45
147.01
84.01
58.44
246.47
0.5
0.5
2.0
0.6
3.
Results and discussion
3.1.
Physicochemical properties of PAspNa
In order to investigate the suitability of the new draw solute,
the physicochemical characteristics of PAspNa were explored
on the basis of our evaluation procedure. The basic molecular
characteristics of PAspNa were first analyzed to identify its
molecular size and composition. Its water solubility, osmotic
pressure, and toxicity were also assessed.
3.1.1.
Molecular characterizations
Commercially available PAspNa is supplied in oligomeric
form. To define the composition of PAspNa as a polyelectrolyte, we first analyzed its molecular weight and total
sodium content by aqueous GPC and ICP-AES, respectively.
The average molecular weight of PAspNa, represented by Mn,
was found to be 1313 g/mol, and the polydispersity index (PDI)
was found to be 1.14. The number of sodium ions per carboxylic group was shown to be almost unity, with the result
for Na content being 93.7 mol% (Table S1). These characteristics of the PAspNa was used the molecular dynamics simulation study of the next section. In addition, 1H NMR
spectroscopy was used to explore the structural properties of
PAspNa related to biodegradability. As shown in the 1H NMR
spectra (Fig. S2), both unpurified and purified PAspNa clearly
show characteristic peaks of well-defined aspartic acid units
with broad peaks for a/b-amide methine protons at d ¼ 4.4 and
4.6 ppm, and also another broad peak from methylene protons
at 2.7 ppm. Moreover, regardless of purification method, both
1
H NMR spectra of PAspNa showed almost no peaks in the
8e10 ppm region, where branched and/or opened amide
groups would resonate. Thus, it is indicated that the content
of irregular forms in the PAspNa is very small. In general, the
irregular structural forms display a much slower biodegradation rate than the linear form (Tomida and Nakato, 1997;
Rowenton et al., 1997). As the PAspNa sample consists
almost entirely of linear type molecules with a negligible
298
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
(a)
2.0
1.5
1.0
0.5
0.0
0.0
Water solubility
To verify the high water solubility of PAspNa through its solvation behavior at the atomic-scale, we examined how water
molecules exist around the draw solute molecules by investigating the pair correlation function, g(r), between the two
oxygen atoms of the PAspNa carboxylate ion (COO) and
water molecules as a function of distance, r. In order to
demonstrate the suitability of PAspNa as draw solute through
comparison study with the criteria of material selection, we
also investigated PAANa with a similar molecular weight
(nominal molecular weight ~ 1200 g/mol), which is a previously reported polyelectrolyte draw solute. To simplify our
approach, we assumed that PAspNa and PAANa are (1) a
syndiotactic linear molecule with short side chains, and (2)
composed of 10 repeat units. Since the number of solvating
water molecules would simply increase proportionally to the
number of repeating units in the model molecule, we assumed
that PAANa has the same number of repeating units as
PAspNa. In addition, the oxygen and hydrogen atoms of the
water molecules surrounding the carboxylic groups were
taken into account as solvating molecules.
As plotted in Fig. 2(a), the g(r) curves between COO of both
PAANa and PAspNa, and the oxygen of water (OW) showed
similar values of unity for r < 2 nm. The sharp peaks below
r ¼ 0.35 nm, representing possible hydrogen bond formation
between the fully ionized carboxylate groups and the water
molecules, suggest that approximately equal numbers of
water molecules are present for every COO in both PAANa
and PAspNa. There are 2.41 and 2.21 water molecules present,
respectively. The function g(r) between the COO and two
hydrogens of water (HW) is also shown in Fig. 2(b). The trends
of both curves are very similar to Fig. 2(a) and every COO
group of PAANa and PAspNa had 2.45 and 2.26 water molecules, respectively. As the number of water molecules existing
around the draw solute could be interpreted to the water
solubility directly, these two pair correlation functions assure
that PAspNa has comparable water solubility to PAANa.
For the simulation of PAspNa and PAANa configurations in
water, we illustrated the first shell of solvation with the end
carboxylate group of each draw solute molecule using spatial
distribution functions (SDF) which gives the probability of
finding an atom in the three-dimensional space around the
central molecule. The SDF is shown for water molecules at a
distance of less than 0.35 nm, where the first solvation shell is
found, from the COO of PAspNa and PAANa. As depicted in
Fig. 3, the water molecules are almost spherically distributed
around the carboxylate groups.
PAANa_10mer
PAspNa_10mer
2.5
0.5
1.0
1.5
2.0
r (nm)
3.0
2.5
g(r) [C(COO-), HW]
3.1.2.
3.0
g(r) [C(COO-), OW]
amount of branching, its structure is more prone to biodegradation. Considering an increasing environmental concern
and discharge limitations, the biodegradability of PAspNa
could make it more suitable as draw solute due to its ecofriendliness. According to previous studies, however, biodegradable polymer, which was used as an antiscalant, could
accelerate biofouling on a membrane surface (Vrouwenvelder
et al., 2000). This gives a deducible fact that an inevitable
leakage of PAspNa to feed solution may cause an aggravated
membrane fouling in FO. In order to confirm the impact of
biodegradable draw solute on the FO performance, further
research should be needed.
(b)
PAANa_10mer
PAspNa_10mer
2.0
1.5
1.0
0.5
0.0
0.0
0.5
1.0
1.5
2.0
r (nm)
Fig. 2 e Pair correlation function between the two
carboxylate oxygens (COO¡) of the draw solutes and (a) the
oxygen of water (OW) and (b) both the hydrogens of water
(HW). The solid and dashed curves indicate PAspNa and
PAANa, respectively.
Our previous study also compared the hydrogen bonding
dynamics between charged carboxylic group and water molecules of both PAANa and PAspNa (Ramachandran et al.,
2013). Quantitatively, the average number of H-bonds for
PAANa and PAspNa is 48 and 56, respectively. This result indicates that the PAspNa has a more favorable interaction with
water, which results in relatively open conformations of the
solute. Indeed, the radius of gyration (Rg) of PAANa was
calculated to have a much smaller size of 0.7 nm with a narrow distribution, while PAspNa showed a very broad distribution of Rg. These demonstrated facts, i.e., the large
configuration and molecular size of solvated PAspNa, through
the analytical scrutiny lead us to expect much less reverse
solute flux, which makes it the more suitable draw solute,
than PAANa.
3.1.3.
Osmotic pressure and viscosity
To examine the driving force potential for FO, the osmotic
pressure of PAspNa was measured in terms of osmolality. As
shown in Fig. 4(a), the osmolality of PAspNa solution has an
299
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
2.5
O smo lality (O s mo l/kg )
(a)
2.0
1.5
1.0
0.5
0.0
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.30
0.35
Concentration (g/mL)
5.0
4.5
(b)
Viscosity(cP)
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.00
0.05
0.10
0.15
0.20
0.25
Concentration (g/mL)
Fig. 4 e (a) Osmolality of the PAspNa solution as a function
of concentration. (b) Viscosity of the PAspNa solution at
different concentrations.
Fig. 3 e Image of spatial density isosurface of water
molecules within 0.35 nm of the carboxylate oxygens
(COO¡) of the first residue of (a) PAANa and (b) PAspNa.
almost linear correlation with concentration from 0.05 to
0.3 g/mL. The osmolality values ranges from 0.21 to 2.15
Osmol/kg corresponding to osmotic pressure from 5.08 to
51.5 atm. According to previous studies, simulated seawater
(0.6 M NaCl) and 0.72 g/mL PAANa (1200) produced an osmotic
pressure of approximately 27.6 and 44.5 atm, respectively
(Achilli et al., 2010; Ge et al., 2012). In comparison with these
osmotic pressure values, 0.3 g/mL PAspNa generated a much
higher osmotic pressure, indicating its great potential as an
alternative draw solute. Especially, in the case of polyelectrolytes, it has been reported that the osmotic pressure
can be implied from the water activity, which is influenced by
freely mobile counterions in proximity to charged polyions
(Vlachy, 2008; Manning, 1972). Lower ion pairing between
counterions and polyions leads to more freely mobile counterions, resulting in reduced water activity around them. This
can be correlated to the results of earlier publication, in which
PAspNa has lower ion paring between its charged carboxylic
groups and Na counterions than in PAANa (Ramachandran
et al., 2013), resulting in a higher osmotic pressure. The
possible ion pairing was represented by the value for average
number of ion pairs <Nip>, which was calculated to be 0.9 and
1.8 for PAspNa and PAANa, respectively.
Viscosity may affect the efficiency of FO performance by
causing a serious concentration polarization (CP) effect.
Therefore, the viscosity was measured and used as another
criterion for draw solute evaluation. As illustrated in Fig. 4(b),
the viscosity of the solution increased exponentially as the
PAspNa concentration increased from 0.05 to 0.3 g/mL. At all
concentrations, the viscosity was lower than 4.4. This phenomenon is explained by the fact that high concentration not
only causes high friction between the neighboring PAspNa
molecules, but also reduces the water activity. When converting the unit to relative viscosity, a maximum value of 5.0
300
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
was recorded at 0.3 g/mL PAspNa. This viscosity is significantly higher than that of simulated seawater, i.e., 1.08, but
much lower than the measured value of 0.3 g/mL PAANa
corresponded to 13.0.
3.1.4.
Cellular toxicity
To assess the toxicity of PAspNa, we performed an in vitro
toxicology test by means of an MTT assay with primary
cultured human fibroblast cells. Control cells were treated
with the same medium as in the PAspNa-treated experimental cell groups, but without PAspNa, in order to exclude
the effect of the medium on the cells. We examined the
cellular toxicity of PAspNa at 3000 ppm total dissolved solids
(TDS), which is ten times the concentration of the maximum
accepted TDS value for drinking water, i.e., 300 ppm (WHO,
1996). As shown in Fig. 5, the PAspNa-treated cells showed
almost the same or better cellular viability than the control,
which was treated with empty media. This indicates that even
under the tested maximum concentration of 3000 ppm for
72 h, PAspNa does not show any significant toxicity to human
skin cells. Although this in vitro toxicology test is admittedly
quite different from the test for determining the standard TDS
value for drinking water, this is meaningful in that it shows
exposure to PAspNa does not harm human skin cells.
3.2.
FO performance
In the previous section, we demonstrated that the PAspNa is a
good candidate for a new draw solute for FO in terms of material characteristics, including its large molecular size, high
osmotic pressure, and non-toxicity. To evaluate the practical
suitability of PAspNa in real membrane processes, basic FO
performance tests were carried out using both CTA and TFC
membranes. A simple economic feasibility was also assessed
based on the results of the FO performances.
3.2.1.
Basic performance: CTA membrane
As most of draw solute researches have been conducted with
CTA membranes, this study also used CTA membrane to
measure the water flux and reverse solute flux of PAspNa in
FO. Fig. 6 shows that, in both AL-FS and AL-DS modes, as the
concentration of PAspNa increased from 0.1 to 0.3 g/mL, the
water flux also increased. Its increment was not proportional
to that of the concentration, reflecting the correlation between
osmotic pressure and CP of the PAspNa solution. More specifically, higher concentrations of PAspNa not only generate
greater osmotic driving force, which lead to higher water flux,
but also simultaneously reduce its diffusivity. The enhanced
hindrance of diffusion at high concentration of PAspNa exacerbates the CP effect, which results in a non-linear relationship between the concentration and water flux.
Furthermore, the water flux in AL-DS mode always surpassed
that observed in AL-FS mode. This means that the internal CP,
which occurs in AL-FS mode, has a greater effect on the flux
decline than the external CP in AL-DS mode (Yen et al., 2010).
The reverse solute flux in both operational modes increases
following the same trend as that seen for water flux. The
maximum value of reverse solute flux was recorded as 2.4
gMH at the concentration of 0.3 g/mL under AL-DS mode.
The specific reverse solute flux, which refers to the ratio of
reverse solute flux to water flux, i.e., Js/Jw, is usually used as an
indicator of draw solute loss per volume of water permeation
in the FO operation. So, we first calculated the Js/Jw of 0.2 g/mL
PAspNa, of which an osmotic pressure of 26.6 atm, based on
the results of the FO experiments. As NaCl and MgCl2 are
extensively used as draw solutes, we also referred to their Js/Jw
values as comparative benchmarks for a better evaluation of
PAspNa: 0.035 g/mL of NaCl and 0.034 g/mL MgCl2 with an
osmotic pressure of 27.6 atm were adopted as model concentrations. As shown in Fig. 7(a), PAspNa has the lowest
value of Js/Jw, 0.19 g/L, while those of NaCl and MgCl2 are
recorded as 0.75 g/L and 0.57 g/L, respectively (Achilli et al.,
2010). This result indicates that PAspNa has a lower loss to
the feed solution than the inorganic salts for each liter of
water produced during the FO operation.
For a more specific comparison in an economic perspective, we also calculated the FO operating cost of those solutes.
As an operating cost of draw solution in FO usually regarded
as a cost for replenishing draw solutes, the costs were
computed by multiplying the Js/Jw values and unit solute
120
Water flux (LMH)
Cellular Viability [%]
20
80
60
40
20
0
Control
10
100
1000
3000
Concentration of PAspNa [ppm]
Fig. 5 e Cell viability of human fibroblast cells after a 72 h
incubation with PAspNa at predetermined concentrations.
25
Water flux (AL-FS, LMH)
Water flux (AL-DS, LMH)
Reverse solute flux (AL-FS, gMH)
Reverse solute flux (AL-DS, gMH)
20
15
15
10
10
5
5
0
0.05
0.10
0.15
0.20
0.25
0.30
Reverse solute flux (gMH)
25
100
0
0.35
Concentration(g/mL)
Fig. 6 e Basic FO performance results, including water flux
and reverse solute flux, with the CTA membrane.
301
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
3.2.2.
0.8
(a)
Js/Jw
0.6
0.4
0.2
0.0
NaCl
MgCl2
PAspNa
20
FO operating cost, 10-3$/L
(b)
15
10
5
0
NaCl
MgCl2
PAspNa
Fig. 7 e (a) Specific reverse solutes flux, Js/Jw, of PAspNa
and (b) FO operating cost compared to two conventional
inorganic draw solutes, namely NaCl and MgCl2.
prices. The specific conditions for this assessment were
summarized in Table 2. As illustrated in Fig. 7(b), the operating
cost of PAspNa is recorded as the lowest value among three
draw solutes. Considering those of NaCl and MgCl2 are 0.011
and 0.016 $/L, respectively, the replenishment cost of PAspNa,
0.002 $/L, is a remarkably low value. Although, much more
weight of PAspNa is needed to express a comparable osmotic
pressure with NaCl or MgCl2, which results in high initial investment cost (specific cost, 5th column in Table 2), in the long
run, a more important cost associated with the draw solution
is the operating cost. Therefore, it is logical to judge that
PAspNa has a potential as an alternative draw solute from the
economic perspective.
Basic FO performance: TFC membrane
PAspNa was also evaluated using a TFC membrane with a pH
tolerance of 2e11 in order to prevent the possibility of membrane damage caused by its basicity, which was pH~10.3 at all
the concentrations tested. As shown in Fig. 8(a), the FO performance with the TFC membrane, including water flux and
reverse solute flux, shared very similar characteristics with
the performance seen with the CTA membrane. More specifically, both water flux and reverse solute flux increased under
both operational modes as the concentration of PAspNa
increased. The water flux under AL-DS mode always clearly
surpassed that observed under AL-FS mode, and the
maximum water flux was recorded as 31.8 LMH under AL-DS
mode at a PAspNa concentration of 0.3 g/mL.
Because intensely high water flux is observed in Fig. 8(a),
PAspNa solutions with concentrations under 0.1 g/mL, which is
the minimum concentration employed in the preceding FO
experiments with the TFC membrane, were additionally used as
the draw solution in AL-DS mode. As shown in Fig. 8(b), the
water flux of the PAspNa solution increased linearly with concentration at low concentrations ranging from 0.01 to 0.1 g/mL.
However, this trend of increasing water flux changes suddenly
at 0.1 g/mL. Beyond 0.1 g/mL, the rate of water flux increase with
concentration is much lower. This result is attributed to the fact
that the increase in water flux is inversely proportional to that of
solution viscosity. In addition, a reasonable water flux of 4.81
LMH was achieved at the lowest concentration, 0.01 g/mL. This
meaningful result gives us a potential that the NF process,
which requires a feed solution of low osmotic pressure, could be
the optimal recovery method for use with PAspNa as a feed
solution. Thus, the hybrid FO-NF process could be practical at
low PAspNa concentrations under 0.1 g/mL, attaining a
competitive FO performance and ease of recovery through low
pressure-driven NF. In addition, to compare the FO performance of PAspNa draw solution, same FO experiments were
conducted differing only in their use of 0.3 g/mL PAANa draw
solution. In these tests, as in the PAspNa draw solution tests,
unpurified PAANa solution was used as draw solution. As
shown in Fig. 9, the water flux from the PAspNa draw solution
test was higher than that from the PAANa draw solution test.
This phenomenon could be simply explained by a viscosity
difference between those two solutions. Although same solution concentration was used, the viscosity of 0.3 g/mL PAANa
solution was recorded much higher value than that of PAspNa
solution. It means that diffusion of PAANa solutes in the
membrane support layer is much slower than that of PAspNa
solutes, resulting in much severe ICP effect. In contrast to the
water flux, the reverse solute flux from the PAspNa test was
Table 2 e Osmotic pressure (pDS ), unit cost ($/kg), specific cost ($/L), specific reverse solute flux (g/L) and FO operating cost
($/L) for each draw solution. The specific cost of each draw solution was obtained by calculating the cost of solute needed to
produce 1 L of draw solution.
Draw solutes
NaCl
MgCl2
PAspNa
a
CDS, g/L
pDS , atm
Cost, $/kg
Specific cost, $/L
Js/Jw, g/L
FO operating
cost, $/L
35.2a
33.8a
200
27.6a
27.6a
26.6
15a
28a
8.8
0.53a
0.95a
1.76
0.74a
0.57a
0.19
0.011a
0.016a
0.002
Reported values from previous study (Achilli et al., 2010).
302
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
W a t e r f lu x ( L M H )
40
50
Water flux (AL-FS, LMH)
Water flux (AL-DS, LMH)
Reverse solute flux (AL-FS, gMH)
Reverse solute flux (AL-DS, gMH)
40
30
30
20
20
10
10
0
0.05
0.10
0.15
0.20
0.25
0.30
Reverse solute flux (gMH)
50
0
0.35
Concentration (g/mL)
(a)
Wa t e r f l u x ( L M H)
40
50
Water flux (LMH)
Reverse solute flux (gMH)
40
30
30
20
20
10
10
0
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Reverse solute flux (gMH)
50
0
0.35
Concentration (g/mL)
(b)
Fig. 8 e Behavior of water and reverse solute fluxes under
different PAspNa concentrations: (a) comparison of
operation modes (AL-FS and AL-DS) and (b) variations at
low PAspNa concentration conditions in AL-DS mode of
TFC FO membrane.
14
12
10
Water flux (LMH)
Reverse solute flux (gMH)
Viscosity (cP)
8
6
4
2
0
PAspNa
PAANa
Fig. 9 e Water flux and reverse solute flux of 0.3 g/mL
PAspNa and 0.3 g/mL PAANa draw solutions with TFC
membrane in AL-FS mode.
lower than that from the PAANa test. This result is in complete
accord with our prediction in the previous section, which was
that PAspNa may show much lower solute leakage to the feed
solution due to the larger configuration in water compared with
PAANa. The overall comparison test results proved that the
PAspNa had a better FO performance than that of PAANa.
3.3.
Antiscaling effect of reverse PAspNa flux in the FO
process
In various membrane processes, from pressure-driven RO to
thermal-driven MD, inorganic scaling is a serious problem
which causes increased energy consumption, cleaning frequency and membrane replacement (Antony et al., 2011; He
et al., 2009). Particularly, calcium phosphate scaling is regarded as a major difficulty for the treatment of wastewater
effluent by NF or RO (Greenberg et al., 2005). A commonly
applied method to control this inorganic fouling is dosing the
feed solution with a very small amount of effective antiscalant
(He et al., 2009; Rahardianto et al., 2006). The antiscalant reduces crystal formation and deposition on the membrane
surface by impeding crystal nucleation and growth.
As briefly mentioned above, PAspNa is usually used as a
component in antiscalants. Therefore, scaling experiments in
FO were carried out to verify that the reversely flowed PAspNa
could act as an antiscalant in an FO feed solution. To simulate
wastewater reclamation, synthetic wastewater containing
only inorganic ionic species was used as the feed solution, and
0.3 g/mL PAspNa was used as the draw solution. The resulting
flux decline curve was compared to experiments differing
only in their use of a 0.3 M NaCl draw solution which generated a similar initial water flux with 0.3 g/mL PAspNa draw
solution. The initial water fluxes from the 0.3 g/mL PAspNa
draw solution test and 0.3 M NaCl draw solution test were 8.22
LMH and 8.37 LMH, respectively. As shown in Fig. 10(a), a
gradual and constant decline in flux was observed with the
PAspNa draw solution. Although the permeate flux with a
NaCl draw solution initially showed a similar gentle decline, it
dropped abruptly in the middle of the fouling run. To investigate this difference between the two flux decline curves, the
membrane was removed from the membrane cell and
analyzed by scanning electron microscopy (SEM) after each
experiment. As shown in Fig. 10(c), a dense deposition of
inorganic precipitates (specifically, calcium phosphate salts,
see Fig. S3) was observed on the membrane surface after the
NaCl draw solution test. On the contrary, a distinct salt
deposition was not detected in the SEM image of FO membrane obtained after the PAspNa draw solution test (Fig. 10(b)).
To confirm this densely deposited layer originated from
synthetic wastewater, the qualitative and quantitative
chemical characterizations of the fouled FO membranes were
performed with SEM-EDS (Fig. S3) and ICP-AES (Table S2) analyses, respectively. The FO membrane when using NaCl draw
solution shows predominantly the characteristic peak of calcium as well as phosphate organic atoms from supporting
layer of TFC membrane. While FO membrane using PAspNa
draw solution as well as virgin FO membrane do not show the
characteristic peaks from inorganic atoms of wastewater, due
to the relatively low detection limit for inorganic metals of
small atomic mass with SEM-EDS, we could not observe other
303
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
Fig. 10 e Antiscaling effect of PAspNa draw solution: (a) Flux decline curves obtained during the inorganic scaling
experiments. (b) SEM image of an FO membrane after the scaling experiment using a 0.3 g/mL PAspNa as draw solution. (c)
SEM image of an FO membrane after the scaling experiment using a 0.3 M NaCl as draw solution.
PAspNa recovery process
To verify the ease and efficiency of PAspNa recovery, a recyclability study was conducted on the two membrane processes, i.e., pressure-driven NF and thermal-driven MD. First,
the NF process was applied to demonstrate the effectiveness
of size exclusion by large molecular size of PAspNa with low
energy consumption. In addition, in order to reduce the viscosity of the PAspNa solution and prove the viability of a
hybrid FO-MD system, the MD process was also adopted as
another recovery method.
3.4.1.
NF process
As PAspNa has a large molecular size and expanded configuration in water, it can be readily separated from water with the
aid of a relatively larger pore size membrane than that of RO.
Therefore, our first set of recyclability studies were designed as
an NF process to verify the recovery convenience of PAspNa.
Diluted PAspNa solutions ranging from 0.01 to 0.03 g/mL were
employed as feed solutions, which guaranteed a higher water
flux than 4.81 LMH under TFC AL-DS mode (Fig. 7(b)). As shown
100
1.2
1.0
95
0.8
90
0.6
0.4
85
0.2
0.0
0.00
Water flux (LMH)
Recovery rate (%)
0.01
0.02
0.03
80
0.04
Concentration (g/mL)
Fig. 11 e Specific water fluxes and recovery rates for
different concentrations of PAspNa solutions in NF
process.
Recovery rate (%)
3.4.
in Fig. 11, specific water flux, defined as the ratio of water flux to
hydraulic pressure, decreased from 1.07 to 0.44 LMH/bar as the
feed PAspNa concentration increased from 0.01 to 0.03 g/mL.
This behavior is attributed to a decrease in the driving force
caused by a reduced difference between the fixed hydraulic
pressure and the feed osmotic pressure at high concentration.
Simultaneously, there was an analogous decrease in rejection
rate, which resulted from increased occurrence of PAspNa
permeating through the membrane. Nevertheless, a high recovery rate of more than 98.9% was achieved at all the concentrations tested in this experiment.
To evaluate the NF recovery performance of PAspNa, EDTA
sodium salt was used as a benchmark, since its recyclability
by NF has already been established. According to Hau et al.
(Hau et al., 2014), 0.07 M EDTA sodium salt, which can
generate a water flux of 4 LMH in FO, was adopted as an NF
Specific water flux (LMH/bar)
inorganic atoms on the surface of the severely fouled FO
membrane when using NaCl draw solution. However, in ICPAES results, the additional inorganic atoms such as sodium,
magnesium and potassium of small atomic mass were also
detected with a higher magnitude of order on FO membrane
when using NaCl draw solution compared to the virgin FO
membrane and FO membrane when using PAspNa draw solution. Not only with the analysis of surface morphology and
chemical composition, we could deduce that severe scaling,
indicated by the rapid flux decline and dense fouling layer,
occurred when using a NaCl draw solution. These results
confirm that reversely flowed PAspNa in the feed solution can
inhibit the physical mechanisms of crystal formation, which
cause inorganic fouling.
304
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
feed solution, and its specific water flux and rejection rate in
NF ranged between 0.7 and 1.0 LMH/bar, and 80 and 93%,
respectively. Considering an NF performance of 0.01 g/mL
PAspNa had a specific water flux of 1.07 LMH/bar and a
rejection rate of 99.5%, PAspNa recovery under the NF process
was favorably comparable with the benchmark, assuring
satisfactory water production and solute rejection.
3.4.2.
MD process
In the continuous hybrid system combining FO and MD, the
heated FO draw solution is simply a feed solution for MD.
Considering the relationship between viscosity and temperature, a draw solution of high temperature can improve the
efficiency of FO by reducing its viscosity. Therefore, in order to
lower the viscosity of the PAspNa solution and verify the
applicability of the FO-MD hybrid system, MD experiments
were conducted as another recovery method. Fig. 12 shows
water flux as a function of solution concentration. The water
flux decreased from 24.6 to 15.0 LMH as the feed concentration
was increased from 0.1 to 0.3 g/mL. This reduction can be
ascribed to the fact that a higher concentration of PAspNa
solution not only reduces an effective vapor pressure of water
but also causes more severe temperature polarization. They
limit the water vaporization at the membrane surface and
thus, causes a decrease in vapor movement trough the
membrane which results in a decrement of water production
(He et al., 2011). The corresponding recovery rates for water
fluxes were recorded as more than 99.9% at all the concentrations used for this MD tests. Considering both reasonable
water flux and almost perfect regeneration of draw solute, the
successful operation results revealed the suitability of MD as a
recovery method for PAspNa, and the feasibility of the hybrid
FO-MD process for the improvement of FO performance.
4.
Conclusions
The suitability of PAspNa as a novel draw solute in the FO
process was demonstrated by a series of analyses comprising
Water flux (LMH)
30
20
10
0
0.0
0.1
0.2
0.3
Concentration (g/mL)
Fig. 12 e Water flux in the MD recovery process as a
function of concentration.
0.4
an atomistic molecular dynamics simulation and lab-scale
experiments. The inherent advantages of PAspNa, such as
good water solubility, high osmotic pressure, and non-toxicity,
were demonstrated systematically. The results of FO experiments proved the applicability of PAspNa as an alternative
draw solute compared to other previously studied materials,
particularly due to the much lower reverse solute fluxes
accompanying comparable water fluxes. Moreover, the reverse
PAspNa solutes showed a positive effect on the reduction of
inorganic scaling. The recyclability studies conducted by both
NF and MD processes also verified the ease of PAspNa recovery.
Although further studies are needed before PAspNa can be
applied in industry, this study has demonstrated its feasibility
and applicability to wastewater reclamation and/or brackish
water desalination by the FO-NF or FO-MD process.
Acknowledgments
This research was supported by a grant from the Fundamental
R&D Program for Technology of World Premier Materials
funded by the Ministry of Trade, Industry and Energy, Republic of Korea (10037794).
Appendix A. Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.watres.2015.04.041.
references
Achilli, A., Cath, T.Y., Childress, A.E., 2010. Selection of inorganicbased draw solutions for forward osmosis applications. J.
Membr. Sci. 364 (1e2), 233e241.
Antony, A., Low, J.H., Gray, S., Childress, A.E., Le-Clech, P., Leslie, G.,
2011. Scale formation and control in high pressure membrane
water treatment systems: a review. J. Membr. Sci. 383 (1e2), 1e16.
Boo, C., Elimelech, M., Hong, S.K., 2013. Fouling control in a
forward osmosis process integrating seawater desalination
and wastewater reclamation. J. Membr. Sci. 444, 148e156.
Boo, C., Lee, S., Elimelech, M., Meng, Z., Hong, S., 2012. Colloidal
fouling in forward osmosis: role of reverse salt diffusion. J.
Membr. Sci. 390e391, 277e284.
Cai, Y., Shen, W., Wang, R., Krantz, W.B., Fane, A.G., Hu, X., 2013. CO2
switchable dual responsive polymers as draw solutes for forward
osmosis desalination. Chem. Commun. 49 (75), 8377e8379.
Cath, T., Childress, A., Elimelech, M., 2006. Forward osmosis:
principles, applications, and recent developments. J. Membr.
Sci. 281 (1e2), 70e87.
Duan, J., Litwiller, E., Choi, S.H., Pinnau, I., 2014. Evaluation of
sodium lignin sulfonate as draw solute in forward osmosis for
desert restoration. J. Membr. Sci. 453, 463e470.
Ge, Q., Su, J., Amy, G.L., Chung, T.S., 2012. Exploration of
polyelectrolytes as draw solutes in forward osmosis
processes. Water Res. 46 (4), 1318e1326.
Ge, Q., Su, J., Chung, T.S., Amy, G., 2011. Hydrophilic
superparamagnetic nanoparticles: synthesis,
characterization, and performance in forward osmosis
processes. Industrial Eng. Chem. Res. 50 (1), 382e388.
w a t e r r e s e a r c h 8 0 ( 2 0 1 5 ) 2 9 4 e3 0 5
Greenberg, G., Hasson, D., Semiat, R., 2005. Limits of RO recovery
imposed by calcium phosphate precipitation. Desalination 183
(1e3), 273e288.
Hau, N.T., Chen, S.S., Nguyen, N.C., Huang, K.Z., Ngo, H.H.,
Guo, W., 2014. Exploration of EDTA sodium salt as novel draw
solution in forward osmosis process for dewatering of high
nutrient sludge. J. Membr. Sci. 455, 305e311.
He, F., Sirkar, K.K., Gilron, J., 2009. Effects of antiscalants to
mitigate membrane scaling by direct contact membrane
distillation. J. Membr. Sci. 345 (1e2), 53e58.
He, K., Hwang, H.J., Woo, M.W., Moon, I.S., 2011. Production of
drinking water from saline water by direct contact membrane
distillation (DCMD). J. Industrial Eng. Chem. 17 (1), 41e48.
Holloway, R.W., Childress, A.E., Dennett, K.E., Cath, T.Y., 2007.
Forward osmosis for concentration of anaerobic digester
centrate. Water Res. 41 (17), 4005e4014.
Huertas, E., Herzberg, M., Oron, G., Elimelech, M., 2008. Influence
of biofouling on boron removal by nanofiltration and reverse
osmosis membranes. J. Membr. Sci. 318 (1e2), 264e270.
Ju, Y., Hong, S., 2014. Nano-colloidal fouling mechanisms in
seawater reverse osmosis process evaluated by cake
resistance simulator-modified fouling index nanofiltration.
Desalination 343, 88e96.
Kim, C., Lee, S., Shon, H.K., Elimelech, M., Hong, S., 2012. Boron
transport in forward osmosis: measurements, mechanisms,
and comparison with reverse osmosis. J. Membr. Sci. 419e420,
42e48.
Ling, M.M., Chung, T.S., Lu, X., 2011. Facile synthesis of
thermosensitive magnetic nanoparticles as “smart” draw
solutes in forward osmosis. Chem. Commun. 47, 10788e10790.
Ling, M.M., Wang, K.Y., Chung, T.S., 2010. Highly water-soluble
magnetic nanoparticles as novel draw solutes in forward osmosis
for water reuse. Industrial Eng. Chem. Res. 49 (12), 5869e5876.
Manning, G.S., 1972. Polyelectrolytes. Annu. Rev. Phys. Chem. 23,
117e140.
Na, Y., Yang, S., Lee, S., 2014. Evaluation of citrate-coated
magnetic nanoparticles as draw solute for forward osmosis.
Desalination 347, 34e42.
Noh, M., Mok, Y., Lee, S., Kim, H., Lee, S.H., Jin, G.W., Seo, J.H.,
Koo, H., Park, T.H., Lee, Y., 2012. Novel lower critical solution
temperature phase transition materials effectively control
osmosis by mild temperature changes. Chem. Commun. 48,
3845e3847.
305
Rahardianto, A., Shih, W.-Y., Lee, R.W., Cohen, Y., 2006.
Diagnostic characterization of gypsum scale formation and
control in RO membrane desalination of brackish water. J.
Membr. Sci. 279 (1e2), 655e668.
Ramachandran, S., Katha, A.R., Kolake, S.M., Jung, B., Han, S.,
2013. Dynamics of dilute solutions of poly(aspartic acid) and
its sodium salt elucidated from atomistic molecular dynamics
simulations with explicit water. J. Phys. Chem. B 117 (44),
13906e13913.
Rowenton, S., Huang, S.J., Swift, G., 1997. Poly (aspartic acid):
synthesis, biodegradation, and current applications. J.
Environ. Polym. Degrad. 5 (3), 175e181.
Shannon, M.A., Bohn, P.W., Elimelech, M., Georgiadis, J.G.,
Marinas, B.J., Mayes, A.M., 2008. Science and technology for
water purification in the coming decades. Nature 452,
301e310.
Stone, M.L., Wilson, A.D., Harrup, M.K., Stewart, F.F., 2013. An
initial study of hexavalent phosphazene salts as draw solutes
in forward osmosis. Desalination 312, 130e136.
Tomida, M., Nakato, Takeshi, 1997. Convenient synthesis of high
molecular weight poly (succinimide) by acid-catalysed
polycondensation of L-aspartic acid. Polymer 38, 4733e4736.
Valladares Linares, R., Li, Z., Abu-Ghdaib, M., Wei, C.-H., Amy, G.,
Vrouwenvelder, J.S., 2013. Water harvesting from municipal
wastewater via osmotic gradient: an evaluation of process
performance. J. Membr. Sci. 447, 50e56.
Vlachy, V., 2008. Polyelectrolyte hydration: theory and
experiment. Pure Appl. Chem. 80 (6), 1253e1266.
Vrouwenvelder, J.S., Manolarakis, S.A., Veenendaal, H.R.,
Kooij, D.v.d, 2000. Biofouling potential of chemicals used for
scale control in RO and NF membranes. Desalination 132
(1e3), 1e10.
WHO, 1996. Health Criteria and Other Supporting Information.
Guideline for Drinking Water Quality, second ed., vol. 2. World
Health Organization, Geneva.
Yen, S.K., Mehnas Haja, N.F., Su, M., Wang, K.Y., Chung, T.S.,
2010. Study of draw solutes using 2-methylimidazole-based
compounds in forward osmosis. J. Membr. Sci. 364 (1e2),
242e252.
Zhao, D., Wang, P., Zhao, Q., Chen, N., Lu, X., 2014.
Thermoresponsive copolymer-based draw solution for
seawater desalination in a combined process of forward
osmosis and membrane distillation. Desalination 348, 26e32.