Journal of Membrane Science 455 (2014) 305–311

Contents lists available at ScienceDirect

Journal of Membrane Science

journal homepage: www.elsevier.com/locate/memsci

Exploration of EDTA sodium salt as novel draw solution in forward

osmosis process for dewatering of high nutrient sludge
Nguyen Thi Hau a, Shiao-Shing Chen a,n, Nguyen Cong Nguyen a, Kuo Zue Huang a,
Huu Hao Ngo b, Wenshan Guo b
a
Institute of Environmental Engineering and Management, National Taipei University of Technology, No. 1, Sec. 3, Chung-Hsiao E. Road,
Taipei 106, Taiwan, ROC
b
Centre for Technology in Water and Wastewater, School of Civil and Environmental Engineering, University of Technology, Sydney,
Broadway, NSW 2007, Australia

art ic l e i nf o a b s t r a c t

Article history: In this study, a hybrid forward osmosis–nanofiltration (FO/NF) process was designed for dewatering high
Received 26 September 2013 nutrient containing sludge and recovering draw solution with minimum energy as well as low fouling.
Received in revised form A novel draw solution – EDTA sodium salt – was also systematically studied for dewatering process.
23 December 2013
Results show that using EDTA sodium salt produced higher water flux and lower reverse salt flux when
Accepted 26 December 2013
Available online 3 January 2014
compared to conventional inorganic salt (NaCl) at pH 8. The final sludge concentration reached
32,000 mg/L after 16 h of operation. Moreover, nutrient compounds in sludge were successfully removed
Keywords: by the FO membrane with a removal efficiency of approximately 97% of NH4þ –N, 90% of NO3 –N, 97% of
Forward osmosis NO2 –N and 99% of PO34  –P, which was attributed to the multi-barrier layers of sludge forming on
Sludge dewatering
membrane surface and the steric effect of the FO membrane. The NF recovery of EDTA sodium salt
Osmotic pressure
indicated that all NF membranes performed well and TS-80 was the best among the tested membranes.
Complexion
& 2014 Elsevier B.V. All rights reserved.

1. Introduction
Forward osmosis (FO) has been considered as a promising
technology to solve environmental challenges with excellent
advantages such as high rejection of many pollutants, without
aid pressure and low membrane fouling. It not only reduces
the stresses of wastewater treatment and seawater desalination
but also promotes sludge dewatering meaningfully [1–6]. Unlike
reverse osmosis (RO), FO is an engineering process utilizing
natural osmosis whereby the water from the feed solution passes
through a semi-permeable membrane to minimize operation cost
and increase the lifetime of membrane [7]. Theoretically, the water
flux (Jw) across the FO membrane is the product of the membrane
permeability coefficient (A), the reflection coefficient (s) and the
osmotic pressure driving force (Δπ) [8,9]:
J w ¼ As Δπ
with
π ¼ iCRT
where i is the Van't Hoff factor, C is the concentration of solute
(mol/L), R is the universal gas constant (R=0.0831 L bar/(mol/1C))
and T is the absolute temperature (1C). However, the actual water
n
Corresponding author. Tel.: + 886 2 27712171x4142; fax: +886 2 27214142.
E-mail address: f10919@ntut.edu.tw (S.-S. Chen).
flux is much lower than the theoretical water flux due to both
internal concentration polarization and external concentration
polarization effects [10–13].
Recently, new FO membranes have been synthesized to
improve membrane permeability [14–16]. Chung and co-worker
(2012) showed that the modified membranes could achieve 160%
higher water flux compared with the traditional FO membranes,
which was a good evidence to apply the FO technology in both
academia and industry [10]. However, so far, the development of
FO technology is still hindered by the unavailability of suitable
draw solutions and the exploration of new draw solutes is vital for
future FO applications. Basically, an ideal draw solution must
simultaneously have the following characteristics: (i) high osmotic
pressure; (ii) minimal reverse draw solute flux; and (iii) easy
recovery. Previously, Holloway et al. [17]) used traditional inor-
ganic salt (NaCl) as the draw solution in a continuous FO/RO
ð1Þ
process for concentration of anaerobic digester [17]. Although
promising results have been recorded with high water flux
ð2Þ (Jw ¼ 14 L/m2 h at 23 71 1C with 70 g/L NaCl draw solution) and
75% water recovery, the most critical drawback was high recovery
energy of RO (4 kWh/m3). As presented in our previous study [3],
utilizing available seawater as the draw solution to sludge dewater-
ing process without draw solution recovery requirement proved
economically feasible. However, the high salt leakage from the
seawater side into the feed solution side was the major obstacle to
reuse sludge for agricultural purposes.

0376-7388/$ - see front matter & 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.memsci.2013.12.068
306 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311
To overcome the disadvantage mentioned above, Yong et al. [18]
investigated the reverse salt flux of neutral draw solute glucose
across an asymmetric FO membrane. Low reverse salt flux was
obtained and diluted glucose could be recovered by the NF
membrane. Nevertheless, the measured water flux for the glucose
system (around 4.7 L/m2 h with 1 M glucose) is one-half of the
water flux measured for the NaCl system. Similarly, McGinnis and
Elimelech [19] investigated a novel ammonia carbon dioxide
forward osmosis desalination process and solute recovery was
conducted by a single vacuum distillation column at low tempera-
ture (40 1C). The cost of the system was expected to be very
competitive with an electrical power requirement of typically less
than 0.25 kWh/m3; however, the reverse salt flux of monovalent
ions (NH4þ ) was a problem. In addition, synthesis of magnetic
nanoparticles (MNPs) could provide a new way of designing draw
solutions for the FO process [20–22]. It is noteworthy that the large
particle size of the MNPs had a superior advantage in preventing
salt leakage, whereas the observed particle aggregation was trou-
blesome during recycling via a magnetic separator. Switchable
polarity solvents (SPS) were also used as viable FO draw solutions
[23]. With the presence of carbon dioxide, the tertiary amines were
soluble in water and the removal of carbon dioxide from the SPS
draw solution resulted in separation of the tertiary amine from the
product water due to increased hydrophobicity of the SPS draw
solution. However, polar to nonpolar switch was a complex process
that depends on reaction chamber surface area, temperature, purge
gas, and pressure. Chung and co-workers explored several new
types of draw solutions for use in FO such as polyelectrolytes and
2-methylimidazole-based organic compounds where product water
is recovered by application of pressure-driven UF process and
membrane distillation process, respectively [24,25]. More recently,
using copper sulfate as the draw solute in FO desalination has been
considered as a good idea as diluted draw solution was recovered by
metathesis precipitation without thermal or electrical energy
[26,27]. However, copper sulfate, barium hydroxide, and sulfuric
acid are all consumables and barium sulfate is produced as a waste.
Hence, recovery of copper sulfate draw solution is costly and causes
secondary environmental pollution. Stone et al. [28] used hexava-
lent phosphazene salts as draw solutes in forward osmosis. These
salts created high osmotic pressure and might be suitable for
seawater desalination.
Although many research efforts have been tested with various
chemicals as solutes for draw solutions in the FO process, no work
has been found using EDTA sodium salt as the draw solution. Thus,
the objective of this study was to explore a novel draw solute –
EDTA sodium salt – that could generate a high water flux with a
minimal reverse salt diffusion. Fundamentally, EDTA exhibits high
charged compounds in high pH conditions according to the
following equations:
H4EDTA2H3EDTA  þH þ pKa1 ¼1.99 (3)
H3EDTA  2H2EDTA2  þ H þ pKa2 ¼2.67 (4)
H2EDTA2  2HEDTA3  þ H þ pKa3 ¼6.16 (5)
Table 1
Characteristics of nanofiltration membranes.
Membrane name Manufacturer Material
NF-XN45 TriSep Polypiperazine amide
NF-DL GE Osmonics Thin film composite
NF-TS80 TriSep Polyamide
NF-DK GE Osmonics Thin film composite
HEDTA3  2EDTA4  þ H þ pKa4 ¼10.26 (6)
Clearly, at high pH, trivalent and tetravalent ions are much
easier to remain at the draw solution side, which limits the
diffusion of the counterion. More specifically, high charged com-
pounds in draw solution will be easily regenerated with the aid of
nanofiltration (NF). Furthermore, the safety of EDTA sodium salt
for the environment and human is one of the advantages for its
application in water treatment. In this study, the performance of
EDTA sodium salt as draw solution in the FO process for dewater-
ing high nutrient sludge was symmetrically investigated. Firstly,
the effects of pH, cross-flow rate and draw solution concentration
on the FO process were evaluated with deionized (DI) water as the
feed solution with the active-layer-facing-draw-solution orienta-
tion. Next, simultaneous sludge dewatering and nutrient removal
of FO were examined with activated sludge as the feed solution.
Finally, the recovery of the diluted EDTA sodium draw solution
after FO tests was also conducted with four kinds of NF mem-
branes to investigate the separation mechanisms and determine
the appropriate membrane.
2. Materials and methods
2.1. Materials
Commercial asymmetric cellulose triacetate (CTA) cartridge-
type membranes (Hydration Technology Inc., Albany, OR, USA)
were used for all FO experiments. The FO membrane thickness is
approximately 50 μm with measured contact angle of 611 and the
membrane has been recorded to be negatively charged in typical
feed waters [29]. Characteristics of four NF membranes used in
EDTA sodium recovery are shown in Table 1. Laboratory-grade
EDTA–2Na (purity of 99.0%) was purchased from Sigma-Aldrich Co,
Germany. D-glucose, NH4Cl and K2HPO4 (Merck Co. Ltd., Germany)
were used as organic and nutrient concentrations of activated
sludge. Samples of activated sludge were first collected at the
secondary sedimentation tank of new Taipei wastewater treat-
ment plant (WWTP) in Taiwan with mixed liquor suspended solid
(MLSS) concentration of 8000 mg/L. And then glucose, NH4Cl and
K2HPO4 solution was added to the raw sludge to produce synthetic
sludge with dissolved organic carbon (DOC) of 550 mg/L, NH4þ –N
of 150 mg/L and PO34  –P of 150 mg/L (Table 2). The calculation of
concentration units for FO studies should be molality (mol/kg)
instead of molarity (M, mol/L) [9]. However, in diluted solution,
the difference of these two is small and thus molarity is used
throughout the manuscript.
2.2. The production of draw solutions
Since the formation of high charged species of EDTA strongly
depends on pH, the key point is the pH control of the EDTA draw
solution. Hence, 4 M concentration of NaOH solution was prepared
first, and for example, the pH of 0.3 M EDTA sodium salt draw
Molecular weight Rejection 25 1C pH range
cut-off (MWCO)
500 (95.0% MgSO4) 2–11
1–11
150 (99.0% MgSO4) 2–11
150–300 (96% MgSO4) 2–10
 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311 307

Table 2
Characteristics of synthetic sludge from new Taipei WWTP as feed solution.

MLSS (mg/L) Ca (mg/L) Na (mg/L) Zn (mg/L) Cu (mg/L) Ni (mg/L) Cr (mg/L)

8000 745 68 7 1 1207 2 3.177 0.12 0.747 0.08 0.217 0.05 0

DOC (mg/L) pH SVI (mL/g) NO3 –N (mg/L) NO2 –N (mg/L) NH4þ –N (mg/L) PO34  –P (mg/L)

550 7 6 7.2 70.2 68 7 3 287 2 30 72 1507 4 150 73

Fig. 1. Schematic illustration of the high nutrient sludge dewatering FO–NF system (membrane orientation: active layer facing to feed side as using sludge as feed solution).

solution was adjusted from 4.5 to 5.5, 6.0, 6.5, 7.0, 7.5 and 8 by
adding 0, 14, 38, 46, 58, 70 and 81 mL NaOH solution, respectively.
2.3. Experimental set-up
The experimental set-up used in this work is shown in Fig. 1 and
the system was operated in batch experiments. The FO test cell (FO
Sterlitech) was designed with symmetric channels on both sides for
the feed and draw solutions, and each channel was 4.5 cm in width,
9.2 cm in length and 0.2 cm in height. The total active FO membrane
area for mass transfer was 41.40 cm2 and the initial volumes of the
feed solution and draw solutions in FO process were 500 and
1000 mL, respectively. Two peristaltic pumps (Master Flux L/S Drive,
Model 7518-00) were used to continuously circulate the feed and
draw solutions on both sides of the FO membrane, where the cross-
flow rate of the feed solution was kept the same as that of the draw
solution. The fluctuation of temperature during the experiment was
in the range of 25–28 1C. Conductivity and pH sensors were installed
in the containers for the feed solution and the draw solution to
monitor the changes. Moreover, mass transport across the mem-
brane was determined by measuring the weight change of the draw
solution container using a scale to calculate the water flux. The NF
tests were carried out using a laboratory-scale cross-flow NF
membrane cell (CF042 Delrin Acetal Crossflow Cell, USA). The feed
water (with initial volume and concentration of 2000 mL and 0.07 M,
respectively) was circulated into the membrane cell by a diaphragm
pump (Triwin, Taiwan) under operating hydraulic pressure of 5.5 bar.
All experimental data were collected after 2 h to avoid the influence
of adsorption of ions on the membrane surface to the rejection.
2.4. Measurement of water flux and reverse solute flux
The experimental water flux (L/m2 h) was calculated by mea-
suring the mass change in the draw solution container versus time
as follows:
ΔV
Jw ¼ ð7Þ
A Δt
where ΔV is total increase volume of permeate water (L) collected
over a predetermined time Δt (h), and A is the effective FO
membrane area (m2).
The reverse solute flux, Js (g/m2 h), of EDTA sodium salt was
determined by the conversion of its electrical conductivity mea-
sured by a conductivity meter when EDTA sodium salt dissociates
in its aqueous solution as follows:
V t C t V 0 C 0
Js ¼ ð8Þ
A t100
where Ct and Vt are the concentration and volume of feed solution
measured at time t, respectively, and C0 and V0 are the initial
concentration and volume of feed solution, respectively.
2.5. Analytical methods
The concentration of EDTA was measured by high performance
liquid chromatography (HPLC). The HPLC includes a 20 μL sample
loop, a thermo BDS Hypersil C18 column (Thermo Fisher Scientific
Inc., USA), 150 mm in length, 4.6 mm in diameter, 5 μm packing
diameter with the corresponding guard column, and a photo-
diode array detector operated at a wavelength of 254 nm. The
concentrations of Na þ , PO34  –P, NO3 –N, NO2 –N and NH4þ –N
were analyzed using Ion Chromatography (a Dionex ICS-90) and a
UV–vis spectrophotometer (HACH Model DR-4000, Japan). DOC
samples were prepared using 0.45 μm filter paper, and then
measured using Aurora 1010C TOC Analyzer purchased from O.I.
Analytical Corporation in USA. The concentration of MLSS was
analyzed using methods 2540 D described in Standard Methods for
the Examination of Water and Wastewater [30]. The conductivity
308 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311
was monitored by a conductivity meter (Sension156, Hach, China).
Chromate, zinc, calcium, copper and nickel concentrations were
analyzed using the flame atomic adsorption spectroscopy (GBC
932, GBC scientific equipment, Australia). The fouled membranes
were observed by Scanning Electron Microscopy & Energy Dis-
persive X-ray Spectroscopy (SEM-EDS) with a Philips XL30 system
operating with an accelerating voltage of 5 kV. Viscosity was
measured by Vibro Viscometer, AD Company, Japan.
3. Results and discussion
3.1. Effect of cross-flow rate and pH on water flux and reverse
salt flux
To check whether the cross-flow rate of EDTA sodium draw
solution and feed solution could improve the water flux, a wide
range of velocities in the FO process was carried out. The result in
Fig. 2 shows that when the cross-flow rates of draw and feed
solutions were simultaneously raised from 62 to 385 cm/min, the
water flux was gradually increased from 6.13 to 7.12 L/m2 h,
respectively. An explanation for this phenomenon would be that
the increasing velocity of draw solution reduced the external
concentration polarization effect, and thus increased the water
flux. In addition, the higher cross-flow rate of draw solution and
feed solution kept more effective osmotic pressure at the mem-
brane support layer due to rapid dilution of permeate flow.
Fig. 3(a) depicts the effect of pH on the water flux and salt flux
using DI water as feed solution and 0.3 M EDTA sodium salt as
draw solution on FO performance process. The experiments
recorded that average water flux increased from 6.93 to 7.68 L/
m2 h, while pH value of draw solution increased from 4.5 (original
Fig. 2. Various water fluxes versus cross-flow rates (feed solution: DI water; draw
solution: 0.3 M EDTA sodium salt; pH: 6 and active layer facing the draw solution).
Fig. 3. (a) Effect of pH values on water flux and reverse salt flux and (b) comparison
of EDTA sodium and NaCl as draw solution on FO process (feed solution: DI water;
draw solution: 0.3 M EDTA sodium and 0.3 M NaCl; membrane orientation: active
layer facing the draw solution and cross-flow rate: 384 cm/min).
pH of EDTA–2Na draw solution) to 8. The reason is that adding 14
and 81 mL of 4 M NaOH solution to adjust pH 5.5–8 of draw
solution, respectively, increased free Na þ and osmotic pressure as
well. However, the reverse salt flux had a slightly increasing trend
(from 0.18 to 0.29 g/m2 h) in the pH range of 4.5–7 and then
decreased at pH 8. Obviously, at the pH value of 4.5–7, free Na þ
ions increase in draw solution was the main reason for increasing
salt concentration in feed solution. Although Xie et al. [31]
reported an FO membrane with a mean pore radius of 0.37 nm
could remove most of the high valent ions and organic contami-
nants, Israelachvili [32] concluded that some monovalent ions as
Na þ with a hydrated radius of 0.36 nm was able to pass through
the FO membranes into feed solution container.
In contrast, as presented in Fig. 4, the salt flux was lower than
that of pH 7 due to 14.4% of complexion between EDTA4  and Na þ
at pH 8 (Na[EDTA]3  ) (complex formation is determined by
Mineqlþ based on the chemical equilibrium model from the
thermodynamic database [33,34]), which reduced free Na þ ions
significantly. Moreover, the higher charged compounds of EDTA
were recorded according to Eqs. (3)–(6) and these charged
compounds could not pass through the FO membrane because of
the higher pH condition. Hence, the Donnan equilibrium effect
was enhanced (Fig. S1 in Supplementary material) whereby the
diffusion of HEDTA3  or H2EDTA2  anions and Na þ cations must
be equal on a molar-equivalent basis (ion-pairing), leading to the
salt leakage decline. In addition, the draw solution pH also affected
membrane properties and membrane performance in the FO
process. According to Valladares Linares et al. [35], the higher
negative charge of the FO membrane was recorded at the higher
pH value of 8, resulting in increasing water flux. When the FO
membrane contacted with the EDTA sodium salt, the cross-linked
membrane polymer structure and the FO membrane hydrophobi-
city were changed as a function of solution pH. When the pH of
EDTA sodium draw solution increased, the electrostatic repulsion
between ionizable functional groups of the membrane polymeric
matrix increased, causing an increased pore size and higher water
flux [36]. This clearly demonstrated that pH 8 was the desirable
value for EDTA sodium salt as draw solution of which the
composition is predominantly H[EDTA]3  .
Furthermore, in order to evaluate the feasibility of EDTA
sodium draw solution in FO process, NaCl draw solution at the
same concentration was selected for comparison study. As shown
in Fig. 3(b), using 0.3 M EDTA sodium salt at pH 8 resulted in a
higher water flig (7.68 L/m2 h) compared to 0.3 M NaCl draw
solution (6.68 L/m2 h) due to the Van't Hoff factor effect according
to Eq. (2) (i¼2 with NaCl and i¼4 with EDTA sodium salt). Since
HEDTA3  is the predominant composition (around 85.6%) of EDTA
sodium salt at pH 8, an ideal Van't Hoff index of 4 could be gained,
Fig. 4. Complex formation between EDTA4  and Na þ from Mineqlþ for EDTA4 
concentration of 0.3 M and NaOH concentration of 0.35 M (reproduced from the
data of Mineqlþ ), where 14.4% of complexion (Na[EDTA]3  ) between EDTA4  and
Na þ was observed at pH 8.
 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311 309

consisting of one H[EDTA]3  and three Na þ . Besides, the salt
leakage of EDTA sodium as the draw solution was 3 times lower
than that of NaCl, because the size of HEDTA3  anions in the EDTA
sodium draw solution is much larger than that of Cl  anions in the
NaCl draw solution.
3.2. Effect of draw solution concentrations on water flux
and reverse salt flux
The characteristic of good solubility in water at high pH made
EDTA sodium salt an excellent candidate for being a draw solution,
even to be used in high concentration. Fig. 5(a) presents the effect
of draw solution concentrations on the water flux and salt flux
using DI water as feed solution and EDTA sodium salt as draw
solution on the performance of the FO process. Water flux had an
increased trend from 4.02 to 13.08 L/m2 h when EDTA sodium
concentration increased from 0.1 to 1.0 M, respectively. Thus,
higher draw solution concentration could generate greater osmo-
tic driving force and lead to a higher water flux across the
membrane. However, the increase in water flux was not directly
proportional to the EDTA sodium concentration and Fig. 5(a)
illustrates the flux attenuation at high concentration, which could
be explained by concentration polarization. In addition, between
0.7 M (Jw ¼12.60 L/m2 h) and 1 M (Jw ¼13.08 L/m2 h) EDTA sodium
salt concentration, the difference of water fluxes was negligible, as
the sharp increase in viscosity and decrease in dissociation of
EDTA sodium salt at a concentration of 1 M prevented permeable
flux from feed into draw solution (Fig. 5(b)). Similarly, the reverse
salt flux also increased when EDTA sodium concentration
increased from 0.1 to 0.5 M because of increased free Na þ to
adjust pH to 8, whereas the reverse salt flux of EDTA sodium was
stable at the draw solution concentration between 0.5 and 1 M.
The results could be explained by the following two reasons:
(i) increasing complex formation of Na[EDTA]3  (from 17.8% to
26.7% corresponding to 0.5 to 1 M EDTA sodium) as recorded in
Table 3 and (ii) increasing viscosity which quickly (from 1.42 to
2.19 cp, respectively) decreased the salt leakage significantly. The
results also suggested that 0.7 M EDTA sodium salt was the
preferable concentration for this study.

3.3. Sludge dewatering

Fig. 6(a) demonstrates the variation in the water flux versus

Fig. 5. (a) Effect of draw solution concentrations on water flux and reverse salt flux
and (b) effect of viscosity on water flux (feed solution: DI water; draw solution:
EDTA sodium salt at pH 8; membrane orientation: active layer facing the draw
solution and cross-flow rate: 384 cm/min).
dewatering time using 0.7 M EDTA sodium salt as draw solution and
the initial concentration of sludge feed solution was 8000 mg/L.
It was observed that water flux declined slowly after a sharp
decrease within the first 8 h (average water flux decreased from
8.45 to 4.22 L/m2 h). As the experiment was conducted using batch
mode, the dilution of the draw solution caused decreasing water
flux with the extension of running time.

Table 3
Components of EDTA sodium at pH 8 were determined by Mineql þ based on the chemical equilibrium model.

EDTA sodium [EDTA]4  (%) H[EDTA]3  (%) H2[EDTA]2  (%) H4[EDTA] (%) H3[EDTA]  (%) Na[EDTA]3  (%)
concentration (M)

0.1 0.0 98.1 1.8 0.0 0.0 0.0

0.3 0.0 83.9 1.6 0.0 0.0 14.4
0.5 0 80.6 1.5 0 0 17.8
0.7 0 76.8 1.4 0 0 21.6
1.0 0 71.9 1.3 0 0 26.7

Fig. 6. (a) Variations in water flux and sludge concentration via time (feed solution: activated sludge; draw solution: 0.7 M EDTA sodium salt at pH 8). (b) SEM micrograph of
the active layer of used FO membrane with activated sludge as feed solution and EDTA sodium salt as draw solution. All orientated membranes: active layer facing the feed
solution.
310 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311
Moreover, both increase in salt concentration of feed solution
and deposition of sludge cake layer on the FO membrane surface
decreased water flux during sludge dewatering period. Further-
more, when activated sludge was used as feed solution in FO
process, the reversed salt flux was less than 0.2 g/m2 h, indicating
the positive effect of sludge cake layer on salt leakage. Deposition
of these sludge cakes on the FO membrane could block the mem-
brane pore and enhance solute rejection by steric hindrance [37],
which alleviated reverse salt transport through the membrane as
shown in Fig. 6(b). Meanwhile, as can be seen from Fig. 6(c), using DI
water as feed solution achieved higher water flux than that of
activated sludge, and water flux reduced gradually versus time due
to the only effect of increasing salt concentration in feed solution
and diluted draw solution without membrane fouling. Based on
the experimental results, sludge concentrations of feed solution
depended on the permeable water volume of the FO membrane.
From Fig. 6(a), it is clear that the MLSS concentration increased
quickly with dewatering time. After 16 h of operation, the MLSS
concentration reached approximately 32,000 mg/L, which could
replace simultaneous secondary settler and thickening tanks of the
conventional wastewater treatment plant.
3.4. Nutrient removal on sludge dewatering of FO membrane
Notably, an excellent advantage of FO membrane is the high
rejection of nutrient and organic compounds to achieve superior
effluent water quality. Fig. 7 illustrates the nutrient removal
efficiency of the FO process for sludge dewatering. The initial
concentrations of MLSS, NH4þ –N, PO34  –P and DOC in activated
sludge feed solution were 8000, 150, 150 and 550 mg/L, respec-
tively. Activated sludge often contains colloids and other func-
tional groups (e.g., carboxyl, phenolic and hydroxyl groups) that
can adsorb NH4þ , NO3 , NO2 and PO34  [38,39] to form multi-
barrier layers of sludge on membrane surface. As a result, nutrient
compounds can be rejected almost completely by this steric effect
of the FO membrane. The results exhibited high removal efficiency
of nutrient compounds (approximately 97% of NH4þ –N, 90% of
NO3 –N, 97% of NO2 –N and 99% PO34  –P) during sludge dewater-
ing process.
Apparently, according to Eqs. (9) and (10), nitrification process
occurred slightly on this system during 16 h of sludge dewatering
because of dissolved oxygen brought by sludge recirculation:
NH4þ þO2-NO2 þH2O þ2H þ (9)
2NO2 þO2-2NO3 (10)
Consequently, a small amount of NH4þ
converted to and NO3 ,
finally to NO2 under aerobic condition. This was one possible
reason to explain why the removal of NO2 –N (97%) was higher
than that of NO3 –N (90%). In addition, the difference in rejection
Fig. 7. Rejection of NH4þ –N, NO3 –N, NO2 –N and PO34  –P on the high nutrient
sludge dewatering FO system.
between PO34  –P (99%) and NH4þ –N (97%) was due to both effects
of hydrated radius and charged repulse. According to Kiriukhin
and Collins [40] and Lide [41], the hydrated radius of ammonium
was 0.104 nm with a diffusion coefficient in water of 1.96 
10  9 m2/s whereas the hydrated radius of phosphate was
0.339 nm with a diffusion coefficient in water of 0.44  10  9 m2/s.
Obviously, larger hydrated radius and lower diffusion coefficient
could have a higher removal efficiency. Moreover, as pH of activated
sludge always remained at about 7–8 and the CTA membrane is
negatively charged at pH47 [29], NH4þ ions were easy to attach on
the surface of membrane and into water flux by electrostatic
attractions, whereas the CTA membrane rejected PO34  ions by
electrostatic repulsions.
3.5. Recovery of EDTA sodium draw solution
After FO tests, EDTA sodium recovery was studied with
pressure-driven NF process to test the effectiveness of the four
selective membranes. Each NF concentration experiment was
repeated three times with three independent membranes. Fig. 8
shows the variation in the EDTA sodium removal efficiency and
permeate flow versus different kinds of NF membranes using
0.07 M EDTA sodium salt as diluted draw solution (pressure of
5.5 bar and pH of 8). The reason for using 0.07 M EDTA sodium salt
was (1) according to Fig. 5, concentration as low as 0.1 M still has a
water flux of 4 L/m2 h, and (2) the diluted draw solution can be
concentrated at low pressure which also provides energy saving
for this process. Under the basic condition, the permeate concen-
trations of all NF experiments were low due to the fact that EDTA
sodium was converted to 83.9% HEDTA3  , 1.6% H2EDTA2  and
14.4% NaEDTA3  . Generally, NF membranes are negatively charged
at pH 8 [45,46] and easy to repulse the negatively charged
HEDTA3  or H2EDTA2  on the membrane surface due to electro-
static repulsion and increased removal efficiency. Hence, these
were multivalent ions removed by steric-hindrance effect of NF
membrane [42–44]. Nevertheless, a small portion of divalent
H2EDTA2  and free monovalent Na þ were not removed by the
NF and passed through the membrane during recovery. Moreover,
the results indicated that NF-TS80 could achieve the highest EDTA
sodium rejection of 93%, followed by NF-DK (89%), NF-XN45 (86%)
and then NF-DL (81%) from the conducted experiments. Therefore,
the permeation of 4 NF membranes could be written as follows:
NF-DL4NFXN45 4NF-DK 4NF-TS80, which elucidated that
NF-TS80 had the smallest pore size (MWCO of 150 Da), and could
retain the EDTA–2Na in water of the concentrate side instead of
passing through the membrane. On the contrary, NF-DL generated
the highest permeate water flux of 5.5 L/m2 h, suggesting that
NF-DL was the loosest membrane and the production of permeate
across the NF membrane was easier. Consequently, NF-TS80 is the
Fig. 8. The recovery of EDTA sodium salt with different kinds of NF membranes
(pressure: 80 psi; diluted EDTA sodium concentration: 0.07 M; temperature: 25 oC
and pH: 8). Error bars were based on the standard deviations of three replicate tests
of three independent membranes.
 N.T. Hau et al. / Journal of Membrane Science 455 (2014) 305–311
most preferable membrane for EDTA sodium recovery if high
recovery of draw solution is considered, whereby draw solution
was concentrated approximately three times (from 0.07 to 0.2 M).
4. Conclusion
The successful application of EDTA sodium draw solution for
sludge dewatering in an FO–NF hybrid system has been demon-
strated. The good solubility and high charged species of EDTA at
pH 8 not only provided effective osmotic pressure to draw water
from wasted sludge, but also reduced salt leakage significantly
compared to traditional inorganic salts. In addition, FO tests
showed that using 0.7 M EDTA sodium salt as draw solution for
sludge dewatering achieved a high water flux of 8.45 L/m2 h
within the first hour and then water flux decreased sharply with
the extension of experiment time due to both increase in deposi-
tion of sludge cake layer and diluted draw solution. After 16 h of
sludge dewatering, MLSS concentration reached 32,000 mg/L and
nutrient compounds were removed with high efficiency (97% of
NH4þ –N and 99% of PO34  –P). Furthermore, the result of NF
recovery demonstrated that NF-TS80 was the best membrane
among those tested for the recovery of diluted EDTA sodium draw
solution under the operating pressure of 5.5 bar.
Acknowledgment
The authors gratefully acknowledge the support of Hydration
Technology Innovations (HTI, Albany, OR), USA, for supplying the
FO membranes.
Appendix A. Supplementary material
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.memsci.2013.12.068.
References
[1] R. Valladares Linares, Z. Li, M. Abu-Ghdaib, C.H. Wei, G. Amy, J.S. Vrouwenvelder,
Water harvesting from municipal wastewater via osmotic gradient: an evaluation
of process performance, J. Membr. Sci. 447 (2013) 50–56.
[2] C. Boo, M. Elimelech, S. Hong, Fouling control in a forward osmosis process
integrating seawater desalination and wastewater reclamation, J. Membr. Sci.
444 (2013) 148–156.
[3] N.C. Nguyen, S.S. Chen, H.Y. Yang, N.T. Hau, Application of forward osmosis on
dewatering of high nutrient sludge, Bioresour. Technol. 132 (2013) 224–229.
[4] Q. Li, Application of forward osmosis membrane in a sequential batch reactor
for water reuse (Master thesis), King Abdullah University of Science and
Technology, 2012.
[5] H. Zhu, L. Zhang, X. Wen, X. Huang, Feasibility of applying forward osmosis to
the simultaneous thickening, digestion, and direct dewatering of waste
activated sludge, Bioresour. Technol. 113 (2012) 207–213.
[6] A. D0 Haese, P. Le-Clech, S. Van Nevel, K. Verbeken, E.R. Cornelissen, S.J. Khan,
A.R.D. Verliefde, Trace organic solutes in closed-loop forward osmosis applica-
tions: influence of membrane fouling and modeling of solute build-up, Water.
Res. 47 (2013) 5232–5244.
[7] Q. Ge, M. Ling, T.-S. Chung, Draw solutions for forward osmosis processes:
developments, challenges, and prospects for the future, J. Membr. Sci. 442
(2013) 225–237.
[8] W. Pusch, R. Riley, Relation between salt rejection r and reflection coefficient σ
of asymmetric cellulose acetate membranes, Desalination 14 (1974) 389–393.
[9] A.D. Wilson, F.F. Stewart, Deriving osmotic pressures of draw solutes used in
osmotically driven membrane processes, J. Membr. Sci. 431 (2013) 205–211.
[10] J. Su, T.-S. Chung, Sublayer structure and reflection coefficient and their effects
on concentration polarization and membrane performance in FO processes,
J. Membr. Sci. 376 (2011) 214–224.
[11] C. Suh, S. Lee, Modeling reverse draw solute flux in forward osmosis with
external concentration polarization in both sides of the draw and feed
solution, J. Membr. Sci. 427 (2013) 365–374.
[12] C.H. Tan, H.Y. Ng, Revised external and internal concentration polarization
models to improve flux prediction in forward osmosis process, Desalination
309 (2013) 125–140.
311
[13] Y. Xu, X. Peng, C.Y. Tang, Q.S. Fu, S. Nie, Effect of draw solution concentration
and operating conditions on forward osmosis and pressure retarded osmosis
performance in a spiral wound module, J. Membr. Sci. 348 (2010) 298–309.
[14] G. Han, S. Zhang, X. Li, N. Widjojo, T.S. Chung, Thin film composite forward
osmosis membranes based on polydopamine modified polysulfone substrates
with enhancements in both water flux and salt rejection, Chem. Eng. Sci. 80
(2012) 219–231.
[15] Y.H. Cho, J. Han, S. Han, M.D. Guiver, H.B. Park, Polyamide thin-film composite
membranes based on carboxylated polysulfone microporous support mem-
branes for forward osmosis, J. Membr. Sci. 445 (2013) 220–227.
[16] S. You, C. Tang, C. Yu, X. Wang, J. Zhang, J. Han, Y. Gan, N. Ren, Forward osmosis
with a novel thin-film inorganic membrane, Environ. Sci. Technol. 47 (2013)
8733–8742.
[17] R.W. Holloway, A.E. Childress, K.E. Dennett, T.Y. Cath, Forward osmosis for
concentration of anaerobic digester centrate, Water Res. 41 (2007) 4005–4014.
[18] J.S. Yong, W.A. Phillip, M. Elimelech, Coupled reverse draw solute permeation
and water flux in forward osmosis with neutral draw solutes, J. Membr. Sci.
392–393 (2012) 9–17.
[19] R.L. McGinnis, M. Elimelech, Energy requirements of ammonia–carbon dioxide
forward osmosis desalination, Desalination 207 (2007) 370–382.
[20] H. Bai, Z. Liu, D.D. Sun, Highly water soluble and recovered dextran coated
Fe3O4 magnetic nanoparticles for brackish water desalination, Sep. Purif.
Technol. 81 (2011) 392–399.
[21] Q. Ge, J. Su, T.S. Chung, G. Amy, Hydrophilic superparamagnetic nanoparticles:
synthesis, characterization, and performance in forward osmosis processes,
Ind. Eng. Chem. Res. 50 (2011) 382–388.
[22] M.M. Ling, K.Y. Wang, T.S. Chung, Highly water-soluble magnetic nanoparticles
as novel draw solutes in forward osmosis for water reuse, Ind. Eng. Chem. Res.
49 (2010) 5869–5876.
[23] M.L. Stone, C. Rae, F.F. Stewart, A.D. Wilson, Switchable polarity solvents as
draw solutes for forward osmosis, Desalination 312 (2013) 124–129.
[24] Q. Ge, J. Su, G.L. Amy, T.-S. Chung, Exploration of polyelectrolytes as draw
solutes in forward osmosis processes, Water Res. 46 (2012) 1318–1326.
[25] S.K. Yen, F. Mehnas Haja N, M. Su, K.Y. Wang, T.-S. Chung, Study of draw solutes
using 2-methylimidazole-based compounds in forward osmosis, J. Membr. Sci.
364 (2010) 242–252.
[26] R. Alnaizy, A. Aidan, M. Qasim, Draw solute recovery by metathesis precipitation
in forward osmosis desalination, Desalin. Water Treat. 51 (2013) 5516–5525.
[27] R. Alnaizy, A. Aidan, M. Qasim, Copper sulfate as draw solute in forward
osmosis desalination, J. Environ. Chem. Eng. 1 (2013) 424–430.
[28] M.L. Stone, A.D. Wilson, M.K. Harrup, F.F. Stewart, An initial study of
hexavalent phosphazene salts as draw solutes in forward osmosis, Desalina-
tion 312 (2013) 130–136.
[29] J.L. Cartinella, T.Y. Cath, M.T. Flynn, G.C. Miller, K.W. Hunter, A.E. Childress,
Removal of natural steroid hormones from wastewater using membrane
contactor processes, Environ. Sci. Technol. 40 (2006) 7381–7386.
[30] A. Apha, WEF Standard Methods for the Examination of Waters and Waste-
waters, Washington, DC, 2005.
[31] M. Xie, L.D. Nghiem, W.E. Price, M. Elimelech, Comparison of the removal of
hydrophobic trace organic contaminants by forward osmosis and reverse
osmosis, Water Res. 46 (2012) 2683–2692.
[32] J.N. Israelachvili, Intermolecular and Surface Forces: With Applications to
Colloidal and Biological Systems, second ed., Academic Press, London, 2010.
[33] M.M. Benjamin, Water Chemistry, first ed., McGraw-Hill, New York, 2002.
[34] C.N. Sawyer, P.L. McCarty, G.F. Parkin, Chemistry for Environmental Engineer-
ing and Science, fifth ed., McGraw-Hill, New York, 2003.
[35] R. Valladares Linares, V. Yangali-Quintanilla, Z. Li, G. Amy, Rejection of
micropollutants by clean and fouled forward osmosis membrane, Water Res.
45 (2011) 6737–6744.
[36] M. Xie, W.E. Price, L.D. Nghiem, Rejection of pharmaceutically active com-
pounds by forward osmosis: role of solution pH and membrane orientation,
Sep. Purif. Technol. 93 (2012) 107–114.
[37] A. Achilli, T.Y. Cath, E.A. Marchand, A.E. Childress, The forward osmosis
membrane bioreactor: a low fouling alternative to MBR processes, Desalina-
tion 239 (2009) 10–21.
[38] P.H. Nielsen, A. Jahn, Extraction of EPS, in microbial extracellular substances:
characterization, Struct. Funct. (1999) 49–72.
[39] W. Guo, H.-H. Ngo, J. Li, A mini-review on membrane fouling, Bioresour.
Technol. 122 (2012) 27–34.
[40] M.Y. Kiriukhin, K.D. Collins, Dynamic hydration numbers for biologically
important ions, Biophys. Chem. 99 (2002) 155–168.
[41] D.R. Lide, CRC Handbook of Chemistry and Physics, CRC Press, New York, 1995.
[42] X.-L. Wang, T. Tsuru, S.-I. Nakao, S. Kimura, The electrostatic and steric-
hindrance model for the transport of charged solutes through nanofiltration
membranes, J. Membr. Sci. 135 (1997) 19–32.
[43] X.-L. Wang, W.-N. Wang, D.-X. Wang, Experimental investigation on separa-
tion performance of nanofiltration membranes for inorganic electrolyte
solutions, Desalination 145 (2002) 115–122.
[44] S.-S. Chen, B.-C. Hsu, C.-H. Ko, P.-C. Chuang, Recovery of chromate from spent
plating solutions by two-stage nanofiltration processes, Desalination 229
(2008) 147–155.
[45] J. Schaep, C. Vandecasteele, Evaluating the charge of nanofiltration mem-
branes, J. Membr. Sci. 188 (2001) 129–136.
[46] A.I.C. MorÃdo, A.M.B. Alves, M.D. Afonso, Concentration of clavulanic acid
broths: influence of the membrane surface charge density on NF operation,
J. Membr. Sci. 281 (2006) 417–428.