1

Fast-responsive monolithic hydrogels as draw agent for

forward osmosis membrane process

Kha L Tu 1, George P Simon 2, and Huanting Wang 1*

1 Department of Chemical Engineering, Monash University, Victoria 3800, Australia

2 Department of Material Science and Engineering, Monash University, Victoria 3800,

Australia

* Corresponding author: Huanting Wang, Email: huanting.wang@monash.edu

Abstract

A monolithic copolymer hydrogel was synthesised by polymerising n-isopropylacrylamide

and sodium acrylate in the presence of poly(sodium 4-styrenesulfonate) (PSSS) chains, and

studied as potential draw agent for forward osmosis (FO) process. The monolithic hydrogel

synthesised from 19% NIPAM, 1% SA, and 15% PSSS at 60 °C generated higher FO water

flux and higher recovery rate than other hydrogels, due to improved membrane-hydrogel

contact and water transport. In particular, the synthesised hydrogel draw agent achieved twice

water flux compared with the particles. The monolithic hydrogel containing a smaller amount

of water produced a higher water flux due to its high swelling pressure.

Keywords: Forward osmosis, hydrogel, membrane process, water treatment.

1. Introduction

In the last decade, forward osmosis (FO) has been emerging as a promising water treatment

process and has attracted a large number of studies worldwide [1, 2]. Unlike other membrane

processes such as nanofiltration (NF) and reverse osmosis (RO), which require a hydraulic

pressure to be applied to drive pure water through the semi-permeable membrane, the driving

2

force in FO is a chemical potential difference between the feed and osmotic draw agent.

Because hydraulic pressure is not required in the FO step, FO is considered as an energy-

efficient process when compared with RO process. However, the diluted draw agent, in most

of FO applications, needs to be brought to the next treatment stage where it is reconcentrated

and pure water is collected as the final product. This draw agent recovery stage plays the key

role in the overall energy consumption of the whole FO process. Recent thermodynamic

analysis indicates that this two-step FO process theoretically requires more energy inputs than

RO [3, 4]. As a result, FO process is favoured only in cases where a cost-effective energy

source is used in the draw agent recovery stage, such as industrial waste heat or sunlight.

The development of temperature-responsive polymer hydrogels as FO draw agent was first

reported by our group in 2011 [5]. Polymer hydrogels have a good potential for use as a FO

draw agent because: (1) the hydrogel can be functionalised to obtain high osmotic (or

swelling) pressure [6]; (2) there is negligible reverse solute flux; and (3) the permeate water

can advantageously be recovered by the application of a range of chemical or physical stimuli.

N-isopropylacrylamide (NIPAM)-based hydrogels have been the most common hydrogels

tested for FO draw agent so far, due to their high sensitivity to the temperature.

Much effort has been dedicated to improve the performance of hydrogel draw agent, but its

FO flux is significantly lower compared with inorganic electrolyte solutions. Also there still

exists a trade-off between the water flux and recovery rate in the polymer hydrogel driven FO

process. For instance, some studies achieved water flux improvement by incorporating

hydrophilic components, such as sodium acrylate (SA) [7, 8] or reduced graphene oxide [9],

into a NIPAM hydrogel. Since these ionic components are unresponsive to typical stimuli

such as light and temperature, increasing their proportion would reduce the recovery

capability of the hydrogels [10]. To improve the recovery of the NIPAM-based hydrogels,

magnetic nanoparticle inclusion coupled with magnetic field-induced heating was suggested

3

[11]. Cai et al. [12] employed a hydrogel draw agent having a semi-interpenetrating polymer

network. The hydrogel composed of crosslinked PNIPAM network interpenetrated by free

poly(sodium acrylate) (PSA) chains achieved good recovery capability [12]. A later study [13]

demonstrated that a high FO water flux and higher recovery ratio could be obtained by using

NIPAM-SA copolymer microgels as the draw agent. The sub-micrometer size of the hydrogel

particles improved their contact with membrane surface and also facilitated a fast diffusion of

water out of the particle during phase-transition [13]. Electrically-responsive hydrogels have

recently been tested as FO draw agent [14]; the water flux generated by these hydrogels was

of modest level, and the hydrogel recovery was not investigated in the study.

To date, hydrogels studies in the FO process are in form of small particles, which have

several limitations. Firstly, small particles tend to aggregate after each swelling-deswelling

cycle, resulting in decreased contact area of the FO membrane and hydrogel particles [15].

Cai et al. [12] pressed hydrogel particles into a disc, showing an interesting concept of

designing hydrogel draw agent.

The objective of the present study is to investigate the potential of using a hydrogel monolith

directly prepared from monomers as FO draw agent. In particular, a NIPAM-SA copolymer

monolith is synthesised at 60 °C in the presence of poly(sodium 4-styrenesulfonate) (PSSS)

chains. It is known that NIPAM-based hydrogels synthesised at a temperature higher than its

lower critical solution temperature (LCST) would have a fast-phase-transition property due to

the formation of a heterogeneously porous structure which contained numerous free PNIPAM

domains [16, 17]. In addition, a semi-interpenetrating polymer network would increase the

hydrophilicity and porosity of the synthesised hydrogel [18, 19], therefore an improved FO

water flux is expected. Also this is the first time that the hydrogel monolith is used for FO

draw agent application. The chemical composition of this hydrogel is systematically tailored

and optimized for obtaining a highest water flux and recovery rate.

4

2. Materials and methods

2.1. Materials

NIPAM (96%), SA (99%), and PSSS (Mw = 200 kDa) were used for synthesizing the semi-

interpenetrating hydrogels. N,N’-methylenebisacrylamide (MBA, 99%), ammonium

persulfate (APS, ≥98%), and N,N,N’,N’-tetramethylethylenediamine (TEMED, 99%) were

used as crosslinker, initiator, and accelerator, respectively. Sodium chloride (ACS reagent)

was used for preparing the FO feedwater. All the chemicals were purchased from Sigma

Aldrich (St. Louis, MO). A commercial cellulose triacetate flat-sheet FO membrane with

embedded polyester support (CTA-ES, Hydration Technology Inc., Albany, OR) was used in

this study. The morphology and performance of this membrane has been well documented in

the literature [20, 21]. A SPEX 6870 Freezer/Mill (SPEX SamplePrep, Metuchen, NJ) was

used to produce small hydrogel particles. A field-emission scanning electron microscope (FEI

Nova NanoSEM 450) was used to obtain the morphological structure of the hydrogels. De-

ionised (DI) water was used to prepare the hydrogels.

2.2. Preparation of the hydrogels

In this study, copolymer hydrogels of NIPAM and SA interpenetrated by linear PSSS chains

were prepared. The concentration of monomer (NIPAM and SA), MBA as crosslinker, and

APS as initiator was 20, 2, and 1% wt, respectively. The hydrogels with different NIPAM to

SA ratios, such as 19-1% and 18-2%, were synthesised. To prepare semi-interpenetrating

network hydrogels, given amounts of monomers, crosslinker and initiator were dissolved in

aqueous solutions containing different amounts of PSSS (from 0 to 15% wt). Our

experiments indicated that the hydrogel synthesised with a higher PSSS content (i.e. 20%

PSSS) had a very low mechanical strength and did not maintain its form, so it was not

included in this study. The hydrogels were prepared at two different polymerisation

temperatures, 20 °C and 60 °C. The polymerisation temperature was maintained by putting

5

the glass tube filled with the monomer solution in a water bath at the set temperature. The

polymerisation at 20 °C was done using APS and TEMED as couple redox initiators, whereas

only APS was used as the initiator for 60 °C polymerisation. The polymerisation was

conducted in a glass tube with 28 mm diameter for two hours. Afterwards, the formed

monolithic hydrogels were submerged in large amounts of DI water for 7 days to wash away

any detachable species. The hydrogels synthesised in this study are labelled according to the

concentration of NIPAM-SA-PSSS used in its pre-gel solutions. For example, the hydrogel

synthesised from 19% NIPAM, 1% SA, and 15% PSSS is labelled as 19-1-15.

2.3. Hydrogel swelling and deswelling characterisation

The washed monolithic hydrogels were dried in a convection oven at 50 °C until reaching a

constant weight (usually within 3 days), which was considered as the dry weight in the

calculation of swelling ratio. To study the swelling kinetic of the hydrogel, the dry monolithic

hydrogels were submerged in a copious amount of DI water, and its weight was recorded as a

function of time. With the wet weight was the weight of swollen hydrogel, the swelling ratio

was calculated as:

(g/g) t][Dry weigh

t][Dry weight][Wet weigh]ratio Sweling[

−= Eq.1

To study the deswelling property of the hydrogels, equilibrium swollen monolithic hydrogels

were placed in a 50 °C convection oven. The weight of the hydrogels was recorded as a

function of drying time.

2.4. Morphological characterisation of the hydrogels

The microstructure of the hydrogels was observed using a scanning electron microscope (FEI

Nova NanoSEM 450 FEGSEM). To preserve the internal microstructure, the equilibrium-

swollen monolithic hydrogels were frozen in liquid nitrogen, and then freeze-dried for 48 h.

The freeze-dried samples were fractured using a scalpel, and then sputter-coated with iridium

before SEM observation.

6

2.5. FO experiments

A home-made FO cell consisting of a 27-mm diameter tube, whose one end was covered with

a piece of HTI FO membrane, was used in this study. The membrane was carefully glued on

the tube with the active layer facing out of the tube. A 2,000 ppm NaCl aqueous solution was

used as feedwater. The feedwater was approximately 300 mL in volume, and mildly agitated

during the FO experiments to minimize the external concentration polarization. Similar setup

has been utilized in the previous studies [8, 22]. The conductivity of the feedwater was

monitored and maintained constant during FO experiments. All FO experiments were

conducted at 20 °C and under FO-mode (i.e. the membrane active layer facing the feedwater).

A monolithic hydrogel used for FO draw agent should have a flat surface, which maximizes

the contact area between the hydrogel and the FO membrane. In this study, a swollen

monolithic hydrogel was attached between two glass Petri dishes, and then was dried at 50 °C

oven until the hydrogel obtained a predetermined swelling ratio. A flat and smooth surface of

hydrogel was achieved using this method. Before each FO experiment, the FO membrane

was saturated with DI water, and the membrane tube was carefully examined for any water

leakage. The monolithic or particle hydrogel was placed on the membrane inside the tube to

maintain the best hydrogel-membrane contact during FO experiment. The original hydrogel

had a thickness between 1-2 mm and weight between 0.34-1.8 g depending on its swelling

ratio. Gravimetric method was used to calculate the FO water flux using the following

equation:

LMH)or /h,L/m(At

W]fluxWater [ 2

= Eq.2

where W (g) is the change in the weight of the FO cell during the t (h) (the density of

permeate water was considered as 1,000 g/L); A (m2) is the hydrogel-membrane contact area

which is varied in different experiments depending on the hydrogel swelling ratio.

3. Results and discussion

7

3.1. Hydrogel formulation

To select the most suitable hydrogel formulations for FO draw agent application, different

hydrogel formulations were tested to determine swelling and deswelling rates. The hydrogels

with high swelling and deswelling rates would result in a high FO water flux and recovery.

Changes in the swelling ratio of the monolithic hydrogels in the first 1 h and 72 h

(equilibrium state) swelling are presented in Figure 1. The concentration of free PSSS chains

used in the pregel solution appeared to have a significant impact on the swelling behavior of

the hydrogels. When the PSSS concentration in the pregel solution increased from 5 to 15%,

the hydrogels swelled faster (higher swelling ratio in 1 h), but obtained a lower equilibrium

swelling ratio (Figure 1). This phenomenon was observed for both hydrogels having 19-1 and

18-2% of NIPAM-SA. The fast swelling property of the hydrogel having more free chains in

their pregel solution has been reported in the literature [23-25]. The free hydrophilic PSSS

chains trapped in the crosslinked NIPAM-SA network would increase the hydrophilicity of

the whole gel structure, therefore resulting in a higher swelling rate of the hydrogels (Figure

2).

On the other hand, the decrease in the equilibrium swelling ratio of the hydrogels having

higher PSSS portions in the pregel solution (Figure 1) can be attributed to the decrease in the

crosslinking density of these hydrogels [26]. The presence of PSSS chains in the pregel

solution reduced the overall contact between the monomer and crosslinker molecules,

subsequently resulted in a lower crosslinking density of the whole hydrogel structure. In fact,

it was observed in this study that the hydrogels having more PSSS in pregel solution were

softer and more fragile, which indicated a lower crosslinking density, than the ones having

less PSSS chains in their pregel solutions. There were marginal differences in the swelling

behaviour of the hydrogels having 19-1 and 18-2% NIPAM-SA in their crosslinked network.

The hydrogels having higher SA portion (2% SA) obtained a slightly higher swelling ratio in

8

1 h and equilibrium swelling ratios (Figure 1). It was reported that the content of hydrophilic

molecules in NIPAM-based copolymer hydrogels had a significant impact on the swelling

rate of the hydrogels [24]. In this study, the impact of the difference in NIPAM-SA ratios on

the swelling behaviour of the hydrogels might have been overwhelmed by the impact of the

porous structure as described above.

18-2-15

18-2-10

18-2-5

18-2-0

19-1-15

19-1-10

19-1-5

19-1-0

0 1 2 3 4 5 6 7 8 30 35 40

Swelling ratio (g/g)

Hyd

rog

el (N

IPA

M-S

A-P

SS

S)

1h 72h (Equilibrium)

Figure 1. Changes in swelling ratios of the 60 °C-synthesised hydrogels after 1 h and 3 d

(equilibrium) submerged in 20 °C DI water. The error bars indicate standard deviation of two

replicate experiments.

The hydrogels 19-1-10, 19-1-15, 18-2-5, 18-2-10, and 18-2-15, which showed fastest

swelling capability (Figure 1), were selected for the deswelling assessment. Equilibrium

swollen monolithic hydrogels were deswollen in a 50 °C convection oven, and changes in its

swelling ratio as a function of time were recorded and presented in Figure 2. The hydrogels

containing 19-1% NIPAM-SA showed a significantly faster deswelling property than the

hydrogels containing 18-2% NIPAM-SA regardless their PSSS content (Figure 2). Similar

phenomenon has been reported in the literature [24]. The portion of hydrophilic and

9

hydrophobic contents in the crosslinked structure and their distribution (i.e. random or block

copolymer) play a crucial role in governing the LCST of PNIPAM-based hydrogel [27].

According to Zarzyka et al. [27], the hydrogel containing 19-1% NIPAM-SA would have a

LCST of approximately 36 °C, whereas it was 41 °C for the hydrogel containing 18-2%

NIPAM-SA. Such difference in the LCST of the hydrogels explained the superior deswelling

property of the 19-1-10 and 19-1-15 hydrogels.

0 2 4 6 8 10 12 14 16 18 20 22 24 26

5

10

15

20

25

30

35

40

45

Sw

elli

ng r

atio

(g

/g)

Time (h)

Hydrogel

18-2-5

18-2-10

18-2-15

19-1-10

19-1-15

Equilibrium swelling ratio

Figure 2. Deswelling kinetic of the 60 °C-synthesised hydrogels in 50 °C oven. The error

bars indicate standard deviation of two replicate experiments.

Interestingly, the deswelling property of the hydrogels used in this study was also affected by

the content of semi-interpenetrating PSSS chains. The hydrogels containing higher portion of

free PSSS chains obtained a faster deswelling capability (Figure 2). This phenomenon was

observed for both the hydrogels containing 19-1 and 18-2% NIPAM-SA in their crosslinked

network. Similar results were reported [12], and can be explained by the porosity-induced

phenomenon as previously described in this study. The formed pores played the role of water

10

release channels which reduced the hinder effect of the crosslinked network on the water

diffusing out of the hydrogel, therefore facilitated a phase-transition of the whole hydrogel

structure [28, 29]. These pores would be more interconnected when a higher PSSS

concentration was used in the pregel solution [28]. The fast phase-transition of the hydrogels

containing higher PSSS portion in their pregel solutions could also be attributed to the more

flexible structure of these hydrogels as observed during the hydrogel synthesis. Since the 19-

1-15 hydrogel indicated a superior swelling and deswelling capability, this hydrogel

formulation was selected for further investigations, such as the hydrogel morphology and FO

draw agent tests.

3.2. Impact of the synthesis temperature on the morphology and swelling/deswelling

properties of the hydrogels

The polymerisation temperature had a significant impact on the morphology of the

synthesised hydrogels. The monolithic hydrogel synthesised at 20 °C appeared to be more

homogeneous and mechanically tougher than the monolithic hydrogel synthesised at 60 °C

(Figure 3a). The lower mechanical strength of the 60 °C-synthesised hydrogel implied that

this hydrogel contained a lower crosslinking density. The 20 °C-synthesised hydrogel had a

homogeneously micro-porous structure, whereas the 60 °C-synthesised hydrogel had a

heterogeneously macro-porous structure (Figure 3b and 3c). The impact of polymerisation

temperature on the structure and property of the hydrogels was previously reported, and was

explained by the phase-transition of NIPAM according to the synthesis temperature [17, 26].

The 19-1-15 hydrogel used in this study had a LCST of approximately 36 °C [27]. When

prepared in 20 °C which was below the LCST, the NIPAM-SA polymer chains was in a

relaxed and hydrophilic state, and therefore the hydrogel obtained a homogeneous structure.

On the other hand, when prepared in 60 °C which was well higher than the LCST, the

polymerisation system became phase-separated with some rich- and poor- polymer domains

11

[16]. Consequently, a heterogeneous hydrogel structure was formed, as seen on Figure 3c.

The highly porous structure of the hydrogels seen in Figure 3 may be attributed to the

diffusion of free PSSS chains as previously described in this study.

Figure 3. (a): Optical image of the 19-1-15 hydrogels prepared at 20 and 60 °C; (b) and (c):

SEM images of the hydrogels prepared at 20 and 60 °C, respectively.

The swelling and deswelling of the hydrogels were significantly affected by its

polymerisation temperature (Figure 3). The 60 °C-synthesised hydrogel adapted to changes in

the environmental temperature faster than the 20 °C-synthesised hydrogel (Figure 4a and b).

For instance, the 60 °C-synthesised hydrogel released 65% of its absorbed water after 5 h

placed in 50 °C oven, whereas it was only 45% for the 20 °C-synthesised hydrogel. This

phenomenon can be attributed to the highly and heterogeneously porous structure of the

60 °C-synthesised hydrogel as observed in Figure 3. According to Ju et al. [16], the microgel-

like clusters within the 60 °C-synthesised hydrogel had numerous free ends which overcame

phase-transition with minimum restriction, therefore the hydrogel could be fast-responsive to

the surrounding temperature. On the other hand, having a homogeneously porous

microstructure, the phase-transition of the 20 °C-synthesised hydrogel was restricted by the

three-dimensional stereo pulling, and consequently obtained a slower response to the ambient

temperature [16]. The fast response rate (swelling and deswelling) of the 60 °C-synthesised

hydrogel may indicate its advantages in the FO draw agent application.

12

0 5 10 15 20 25

0

5

10

15

20

25

30

35

0 5 10 15 20 25

0

5

10

15

20

25

30

35

Sw

elli

ng

ra

io (

g/g

)

Time (h)

20 oC synthesised hydrogel

60 oC synthesised hydrogel

(a) In 20 oC water

(b) In 50 oC oven

Sw

elli

ng

ra

io (

g/g

)

Time (h)

Figure 4. Swelling (a) and deswelling (b) kinetic of the monolithic hydrogels synthesised at

20 °C and 60 °C.

3.3. FO performance of the hydrogels in the form of monolith and particles

To study the impact of hydrogel conformation on the FO water flux, the 19-1-15 monolithic

hydrogel synthesised at 60 °C was ground into small particles (approximately 20 µm

diameter), and then tested for FO performance. It is noteworthy that this form of hydrogel

(small ground particles) was employed by the previous studies on hydrogel-driven FO

process [9, 13, 22]. By using the hydrogel draw agent in form of a monolith, this study found

that double amounts FO water flux can be obtained compared to the small particles (Figure 5).

There are two possible explanations for this flux improvement. First of all, the flattened

surface of the monolithic hydrogel had a more complete contact with the FO membrane,

therefore would result in a higher FO water flux. In fact, when the hydrogel was used in the

form of small particles, the membrane-hydrogel contact area was limited due to the irregular

shapes of the particles, resulting in a low FO water flux [15]. Secondly, the interconnected

porous structure of the monolith facilitated internal water diffusion more efficiently than the

inter-particle transport in the particle hydrogel conformation [30, 31]. This explanation is

supported by visual observation during the experiment, which indicated that the top surface

13

of the hydrogel particles was still dry, whereas that of the monolithic hydrogel was partially

hydrated after 90 min FO process.

0 10 20 30 40 50 60 70 80 90 100

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Wate

r flux (

LM

H)

Time (min)

60 oC-monolith

20 oC-monolith

60 oC-particle

Figure 5. FO water flux obtained by the 19-1-15 hydrogels in the forms of monolith and

particles. The hydrogel particles have an average diameter of 20 µm; 2,000 ppm NaCl

solution at room temperature was used as feedwater. Approximately 0.6 g hydrogel at a

swelling ratio of 1 g/g was used. The active area ranges from 3.85 to 5.72 cm2 depending on

the actual area of the hydrogel used. The error bars indicate the standard deviation of three

replicate measurements.

The monolithic hydrogel synthesised at 60 °C led to a higher FO water flux than the one

synthesised at 20 °C (Figure 5). This result can be attributed to the more flexible network of

the 60 °C-synthesised hydrogel as previously observed in this study. Such flexible network

reduced the structural resistance of the whole monolith on the swelling process of the

hydrogel, therefore facilitated a faster water diffusion into the hydrogel [31]. In fact, the

positive impact of a flexible network of the hydrogel draw agent on the FO water flux has

been highlighted in a previous study, in which reduced graphene oxide is added into the

14

hydrogel structure to decrease the stiffness of the hydrogel [9]. The macro-porous and

flexible structure of the 60 °C-synthesised monolithic hydrogel also helped to maintain a high

water flux in a longer FO time. After 90 min FO process, the water flux obtained by the

60 °C-synthesised monolithic hydrogel was approximately 74% of the original value,

whereas it was only 55% for the 20 °C-synthesised monolithic hydrogel (Figure 5).

The water flux obtained by the hydrogels used in this study (i.e. 1.4 – 1.9 LMH, Figure 5)

falls in the middle range of the previously reported studies using smart draw agents. For

instance, Razmjou et al. [15] reported a water flux between 0.8 and 1.3 LMH using particle

hydrogels. By utilising semi-interpenetrating polymer hydrogels, Cai et al. [12] achieved 0.1

– 0.25 LMH. Up to 2 LMH water flux has been reported in a latter study by Hartanto et al.

[13]. It is noted that a direct quantitative comparison between the fluxes reported in different

studies should be made with cautions because there are differences in the experimental

protocols, such as initial swelling ratio of the hydrogel draw agent or flux recording regime,

which can significantly affect the reported water flux.

3.4. Impact of the initial swelling ratio on water flux

Since the swelling rate of the hydrogel was significantly affected its water content (Figure 4a),

one can expect that a drier hydrogel draw agent would generate a higher FO water flux. This

was confirmed when the monolithic hydrogel draw agent was used at relatively high initial

swelling ratios, such as higher than 1 g/g (Figure 6). A lower water flux was obtained when a

hydrogel having a higher initial swelling ratio was used due to the decrease in its swelling

pressure, which was the driving force for the FO process. Interestingly, there appeared an

optimum swelling ratio of the hydrogel draw agent, which was approximately 1 g/g, where

the FO water flux reached the highest value (Figure 6). The water flux started to decrease

when the initial swelling ratio of the monolithic hydrogel went lower than 1 g/g (Figure 6).

This result can be attributed to the observable deformation of the dry monolithic hydrogel

15

which decreased the contact area between the hydrogel and the FO membrane, even though

the monolithic hydrogel was slightly pressed on the membrane surface during the FO process.

This result implied that in order to achieve a high FO water flux by using monolithic

hydrogel draw agent, a relatively dry hydrogel should be used, and its surface must be in

good contact with the FO membrane to produce a maximum contact area.

-1 0 1 2 3 4 5 6 7

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

FO

wa

ter

flu

x (

LM

H)

Initial swelling ratio (g/g)

Figure 6. FO water flux obtained in 1 h FO process as a function of the initial swelling ratio

of the 60 °C-synthesised monolithic hydrogel used. 2,000 ppm NaCl solution at room

temperature was used as feedwater. The error bars indicate the standard deviation of two

replicate measurements.

4. Conclusion

The results of this study demonstrated that temperature-responsive monolithic hydrogels

having a flexible and porous structure can be a suitable draw agent for hydrogel-driven FO

membrane processes. It was found that a hydrogel synthesised at 60 °C containing 19%

NIPAM, 1% SA, and 15% PSSS in the pregel solution can offer a high water flux and high

recovery rate in FO process. Such improvements were attributed to the improved hydrogel-

16

membrane contact when a monolithic hydrogel was used, and also due to the highly

hydrophilic and flexible network of the hydrogel. Furthermore, the hydrogel used in this

study requires relatively simple synthesis protocol and does not contain hazardous materials,

which result in the potential to be used in potable water production. This study also found a

significant correlation between the initial swelling ratio of the hydrogel used and the FO

water flux that it generated. A drier monolithic hydrogel (having lower swelling ratio) would

obtain a higher FO water flux due to its high osmotic potential. Nevertheless, the use of very

dry monolithic hydrogel may be compromised by an inadequate contact between the hydrogel

and membrane surface, which subsequently resulted in a decrease in water flux. Further

research to improve the contact between the membrane surface and dry monolithic hydrogel

is of necessity.

Acknowledgements

This work is financially supported by the Baosteel-Australia Joint Research and Development

Centre (BAJC) and the Australian Research Council (Linkage Project No.: LP140100051).

Ms Ranwen Ou is thanked for SEM imaging.

References 1. Chung, T.S., S. Zhang, K.Y. Wang, J. Su, and M.M. Ling, Forward osmosis processes: Yesterday,

today and tomorrow. Desalination, 2012. 287: p. 78-81. 2. Ge, Q., M. Ling, and T.S. Chung, Draw solutions for forward osmosis processes: Developments,

challenges, and prospects for the future. Journal of Membrane Science, 2013. 442: p. 225-237. 3. McGovern, R.K. and J.H. Lienhard V, On the potential of forward osmosis to energetically

outperform reverse osmosis desalination. Journal of Membrane Science, 2014. 469(0): p. 245-250.

4. Shaffer, D.L., J.R. Werber, H. Jaramillo, S. Lin, and M. Elimelech, Forward osmosis: Where are we now? Desalination, 2015. 356(0): p. 271-284.

5. Li, D., X. Zhang, J. Yao, G.P. Simon, and H. Wang, Stimuli-responsive polymer hydrogels as a new class of draw agent for forward osmosis desalination. Chemical Communications, 2011. 47(6): p. 1710-1712.

6. Wack, H. and M. Ulbricht, Effect of synthesis composition on the swelling pressure of polymeric hydrogels. Polymer, 2009. 50(9): p. 2075-2080.

7. Li, D., X. Zhang, J. Yao, Y. Zeng, G.P. Simon, and H. Wang, Composite polymer hydrogels as draw agents in forward osmosis and solar dewatering. Soft Matter, 2011. 7(21): p. 10048-10056.

17

8. Li, D., X. Zhang, G.P. Simon, and H. Wang, Forward osmosis desalination using polymer hydrogels as a draw agent: Influence of draw agent, feed solution and membrane on process performance. Water Research, 2013. 47(1): p. 209-215.

9. Zeng, Y., L. Qiu, K. Wang, J. Yao, D. Li, G.P. Simon, R. Wang, and H. Wang, Significantly enhanced water flux in forward osmosis desalination with polymer-graphene composite hydrogels as a draw agent. RSC Advances, 2013. 3(3): p. 887-894.

10. Kipcak, A.S., O. Ismail, I. Doymaz, and S. Piskin, Modeling and investigation of the swelling kinetics of acrylamide-sodium acrylate hydrogel. Journal of Chemistry, 2014. 2014.

11. Razmjou, A., M.R. Barati, G.P. Simon, K. Suzuki, and H. Wang, Fast deswelling of nanocomposite polymer hydrogels via magnetic field-induced heating for emerging FO desalination. Environmental Science and Technology, 2013. 47(12): p. 6297-6305.

12. Cai, Y., W. Shen, S.L. Loo, W.B. Krantz, R. Wang, A.G. Fane, and X. Hu, Towards temperature driven forward osmosis desalination using Semi-IPN hydrogels as reversible draw agents. Water Research, 2013. 47(11): p. 3773-3781.

13. Hartanto, Y., S. Yun, B. Jin, and S. Dai, Functionalized thermo-responsive microgels for high performance forward osmosis desalination. Water Research, 2015. 70(0): p. 385-393.

14. Zhang, H., J. Li, H. Cui, H. Li, and F. Yang, Forward osmosis using electric-responsive polymer hydrogels as draw agents: Influence of freezing-thawing cycles, voltage, feed solutions on process performance. Chemical Engineering Journal, 2015. 259: p. 814-819.

15. Razmjou, A., G.P. Simon, and H. Wang, Effect of particle size on the performance of forward osmosis desalination by stimuli-responsive polymer hydrogels as a draw agent. Chemical Engineering Journal, 2013. 215-216: p. 913-920.

16. Ju, X.J., L.Y. Chu, X.L. Zhu, L. Hu, H. Song, and W.M. Chen, Effects of internal microstructures of poly(N-isopropylacrylamide) hydrogels on thermo-responsive volume phase-transition and controlled-release characteristics. Smart Materials and Structures, 2006. 15(6): p. 1767-1774.

17. Rathjen, C.M., C.-H. Park, P.R. Goodrich, and D.D. Walgenbach, The effect of preparation temperature on some properties of a temperature-sensitive hydrogel. Polymer Gels and Networks, 1995. 3(2): p. 101-115.

18. Kayaman-Apohan, N. and B.M. Baysal, Semi-interpenetrating hydrogel networks of poly(2-hydroxyethyl methacrylate) with poly[(D,L-lactic acid)-co-(ε-caprolactam)]. Macromolecular Chemistry and Physics, 2001. 202(7): p. 1182-1188.

19. Zhang, Y. and L. Ye, Improvement of permeability of poly(vinyl alcohol) hydrogel by using poly(ethylene glycol) as porogen. Polymer - Plastics Technology and Engineering, 2011. 50(8): p. 776-782.

20. Gao, Y., W. Li, W.C.L. Lay, H.G.L. Coster, A.G. Fane, and C.Y. Tang, Characterization of forward osmosis membranes by electrochemical impedance spectroscopy. Desalination, 2013. 312: p. 45-51.

21. Wei, J., C. Qiu, C.Y. Tang, R. Wang, and A.G. Fane, Synthesis and characterization of flat-sheet thin film composite forward osmosis membranes. Journal of Membrane Science, 2011. 372(1-2): p. 292-302.

22. Razmjou, A., Q. Liu, G.P. Simon, and H. Wang, Bifunctional polymer hydrogel layers as forward osmosis draw agents for continuous production of fresh water using solar energy. Environmental Science and Technology, 2013. 47(22): p. 13160-13166.

23. Wang, J., X. Zhou, and H. Xiao, Structure and properties of cellulose/poly(N-isopropylacrylamide) hydrogels prepared by SIPN strategy. Carbohydrate Polymers, 2013. 94(2): p. 749-754.

24. Molina, M.A., C.R. Rivarola, and C.A. Barbero, Effect of copolymerization and semi-interpenetration with conducting polyanilines on the physicochemical properties of poly(N-isopropylacrylamide) based thermosensitive hydrogels. European Polymer Journal, 2011. 47(10): p. 1977-1984.

25. Mohan, Y.M., P.S.K. Murthy, J. Sreeramulu, and K.M. Raju, Swelling behavior of semi-interpenetrating polymer network hydrogels composed of poly(vinyl alcohol) and

18

poly(acrylamide-co-sodium methacrylate). Journal of Applied Polymer Science, 2005. 98(1): p. 302-314.

26. Lee, W.F. and S.H. Yen, Thermoreversible hydrogels. XII. Effect of the polymerization conditions on the swelling behavior of the N-isopropylacrylamide gel. Journal of Applied Polymer Science, 2000. 78(9): p. 1604-1611.

27. Zarzyka, I., M. Pyda, and M.L. Di Lorenzo, Influence of crosslinker and ionic comonomer concentration on glass transition and demixing/mixing transition of copolymers poly(N- isopropylacrylamide) and poly(sodium acrylate) hydrogels. Colloid and Polymer Science, 2014. 292(2): p. 485-492.

28. Harmon, M.E., W. Schrof, and C.W. Frank, Fast-responsive semi-interpenetrating hydrogel networks imaged with confocal fluorescence microscopy. Polymer, 2003. 44(22): p. 6927-6936.

29. Kaneko, Y., S. Nakamura, K. Sakai, T. Aoyagi, A. Kikuchi, Y. Sakurai, and T. Okano, Rapid deswelling response of poly(N-isopropylacrylamide) hydrogels by the formation of water release channels using poly(ethylene oxide) graft chains. Macromolecules, 1998. 31(18): p. 6099-6105.

30. Mohan, Y.M., J.P. Dickson, and K.E. Geckeler, Swelling and diffusion characteristics of novel semi-interpenetrating network hydrogels composed of poly[(acrylamide)-co-(sodium acrylate)] and poly[(vinylsulfonic acid), sodium salt]. Polymer International, 2007. 56(2): p. 175-185.

31. Ganji, F., S. Vasheghani-Farahani, and E. Vasheghani-Farahani, Theoretical description of hydrogel swelling: A review. Iranian Polymer Journal (English Edition), 2010. 19(5): p. 375-398.