Polyvinyl chloride and layered double hydroxide composite as a novel

Desalination 421 (2017) 149–159

Contents lists available at ScienceDirect

Desalination journal homepage: www.elsevier.com/locate/desal

Polyvinyl chloride and layered double hydroxide composite as a novel substrate material for the forward osmosis membrane Pankaj M. Pardeshi, Arvind K. Mungray, Alka A. Mungray ⁎ Chemical Engineering Department, Sardar Vallabhbhai National Institute of Technology, Ichchhanath, Surat 395007, Gujarat, India

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Novel PVC/LDH composite was used for FO substrate preparation. • Effect of LDH on substrate morphology was studied by thermodynamic aspect. • Effect of LDH on FO membrane performances was investigated. • Value of S was able to decrease from 699 μm to 303 μm.

a r t i c l e

i n f o

Article history: Received 13 July 2016 Received in revised form 23 January 2017 Accepted 31 January 2017 Available online 24 March 2017 Keywords: Forward osmosis membrane Polyvinyl chloride Layered double hydroxide Thin film composite

a b s t r a c t In this study, a composite of the polyvinyl chloride (PVC) and layered double hydroxide (LDH) has been used as substrate material for the forward osmosis (FO) membrane preparation. Different concentrations of LDH (0, 0.5, 1, 1.5, 2, 2.5 and 3 wt%) were incorporated into PVC to fabricate FO substrates. The substrate morphology was evaluated by scanning electron microscope (SEM) and thermodynamic aspect. Furthermore, the substrate transport properties and osmotic performance of TFC-FO membranes was investigated. It was observed that the increasing concentration of LDH in PVC matrix results an increase in hydrophilicity, surface free energy, pore size and porosity of the substrates. Among all seven TFC-FO membranes, 2 wt% LDH substrate's TFC membrane (denoted as TFC-SUB 2) was unveiled a notable decrease in structural parameter (S). In osmotic flux evaluation, the TFC-SUB 2 membrane exhibited the high water flux of 37.46 Lm−2 h−1 (LMH) and 50.89 Lm−2 h−1 (LMH) using DI water as a feed solution and 1 M NaCl draw solution under an active layer facing feed solution (AL-FS) and an active layer facing draw solution (AL-DS) respectively. This study shows that PVC and LDH can be good candidates for the FO substrate preparation. © 2017 Published by Elsevier B.V.

1. Introduction In recent years, numbers of large scale seawater desalination plants have been built in water stressed areas to supplement existing water ⁎ Corresponding author. E-mail address: [email protected] (A.A. Mungray).

http://dx.doi.org/10.1016/j.desal.2017.01.041 0011-9164/© 2017 Published by Elsevier B.V.

resources, and it is likely to upturn in the near future. Regardless of major development in desalination technology, seawater desalination is still more energy demanding [1]. Forward osmosis (FO) is a state-ofthe-art technology which uses natural osmosis to address comprehensive issues associated to water and energy [2]. In this process, osmotic pressure difference propels the water across a semi-permeable membrane from a low concentrated feed solution to a high concentrated

150

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

draw solution. The applications of FO have been growing in many areas such as fresh water production at low energy consumption [3,4], high value solid recovery [5,6] and trace organic compounds removal [7]. Nevertheless, the concept of FO has stuck in lab scale studies due to limitations in membrane design, appropriate draw solution development and its recovery, process integration, scale up, and the feasibility of a process economy [8]. This study emphasize on membrane design, as it has been pondering a primary hindrance to FO development. The polyamide (PA) thin film composite (TFC) membrane made-up from interfacial polymerization upon a suitable substrate is the most common commercial form of reverse osmosis (RO) and nanofiltration (NF) membrane. The main advantage of the thin film composite membrane is that interaction of both active layer and porous support layer can be individually studied to optimize overall membrane performance [9]. This concept has also been using for the FO membrane fabrication since the last two decades [9]. A TFC-FO membrane is usually composed of (i) ultra-thin polyamide active layer for high salt rejection, and (ii) highly porous support layer for water transport. A TFC membrane fabricated for FO must yield a high water flux with low salt permeability in feed and restrain the diffusion of draw solute. It must also be chemically, thermally stable and easy to manufacture at large scale. Most importantly, the FO membrane must be designed to have low structural parameter (S) to mitigate the osmotic inefficiency. In the consequence of the internal concentration polarization (ICP), the importance of design of substrate in the FO membrane has conceded by many FO membrane scientists [10–12]. It does not only provide mechanical support to the polyamide active layer but also act as a suitable surface for the formation of defect free polyamide active layer [13]. To enhance the influence of the osmotic driving force more effectively, the improvement of substrate plays a crucial part in optimizing TFCFO membrane performance. Despite of using different polymers to tailor FO substrate, the properties like hydrophilicity, thermal, chemical, biological and mechanical stability are still big questions to achieve. The cellulose acetate, cellulose triacetate, polysulfone (PSf) and polyether sulfone (PES) are most widely used polymers for FO substrate preparation. The hydrophilic nature of cellulose acetate and cellulose triacetate can reduce ICP and increase water flux. But poor resistance to hydrolysis and biological attack makes cellulose acetate and cellulose triacetate deprived for a wide range of FO application [14]. The PSf and PES are synthetic polymers known for good chemical resistance, good mechanical properties, thermal oxidative resistance, and resistance to hydrolysis and industrial solvent. However, hydrophobic nature and a high cost of PSf and PES make them unfavorable for the FO process. In order to enhance the desirable properties of FO substrate, the incorporation of inorganic particles in polymer matrix has been attracting more attentions. It has unlocked the possibilities of making new material by combining inorganic particles into virgin polymer and also validated its performance in terms of permeability, thermal stability, selectivity, mechanical strength, and hydrophilicity [15]. Whereas, this approach has been accepted for all pressure driven membrane making process [16–18]. A varied range of polymer nanocomposites have been used to design FO porous support i.e., polysulfone (PSf)/Zeolite [19], polyether sulfone (PES)/carbon nanotube [20], polysulfone (PSf)/TiO2 [21], polysulfone (PSf)/graphene oxide [22], polysulfone (PSf)/silica [23], polyether sulfone (PES)/reduced graphene oxide (rGO) modified graphitic carbon nitride (gC3N4) [24] etc. Regardless of increased efforts to tailor highly stable FO substrate, most of the fabricated polymer nanocomposite suffers with particle agglomeration, and non-uniform dispersion of particles. One of the possible approaches to improve the dispersion of particles in polymer matrix is applying specific functional group by surface modification of particles. The modified particles usually possess specific functional group to restrain the agglomeration in the non-polar polymer solution. Nevertheless, the complex functionalization process and deficiency of thorough understanding of the effect of functionalize particles on FO substrate

limits the applications of the membrane. Hence there is need to explore a different polymer-inorganic particle composite which could improve the FO substrate quality with low cost. Among all polymers for membrane preparation, polyvinyl chloride (PVC) has been drawing more attention as capable membrane material for its applicable mechanical, chemical resistance, film forming property, excellent thermal stability and easy to modify [25]. The PVC is very low cost material compared with other FO membrane material such as polysulfone (PSf), polyether sulfone (PES), polybenzimidazole, polyvinylidene fluoride (PVDF), polyacrylonitrile (PAN). The PVC is easily dissolved in a many organic solvent such as N,N dimethylacetamide (DMAc), dimethylformamide (DMF), N-methyl-pyrrolidinone (NMP), and tetrahydrofuran (THF), which sorts it suitable for industrial membrane separation application. In recent years, numerous studies have been focused on the synthesis of ultrafiltration membrane with PVC [26,27]. However, there is need to add hydrophilic filler in to PVC matrix to increase porosity and wettability of the membrane. Layered double hydroxides (LDHs) are highly functionalize and crystalline minerals with a positively charged brucite layer of mixed metal hydroxide [28]. LDH have many applications such as polymer additive, elimination of environmental hazardous and medical use etc. [28]. Layered double hydroxides (LDHs) comprise of two types of metallic cations such as Mg2+ and Al3+ surrounded by a close packed configuration of OH– groups in a positively charged brucite like layer, with charge compensated by anions and water located in the interlayer space [28]. This exceptional type of structure and high surface energy of LDHs provide a prodigious potential for making polymer composite material [29]. The LDH has been used as filler for a fabrication of ultrafiltration and nanofiltration composite membrane to enhanced hydrophilicity, water flux, rejection of solute and antifouling properties [30]. Furthermore, LDH can be used in a wide range of health applications and also suitable excipients in the pharmaceutical formulation [31]. Whereas, carbon base nonmaterial and graphene base nonmaterial can be a toxic for human and environment [32,33]. To the best of our knowledge, PVC/LDH composite has never been explored earlier as TFC-FO substrate material for FO. In this study, we present a PVC/LDH composite as new material for FO substrate preparation. Our approach is to employ the properties of PVC and LDH composite for obtaining stable material to fabricate hydrophilic PVC/LDH substrate. The objectives of this study are to (i) explore the effect of LDH particles concentration in the PVC matrix on morphology by SEM and thermodynamic aspect; (ii) investigate the transport properties of the flat-sheet FO substrate; (iii) evaluate the mass transport properties and osmotic flux performance of the prepared TFC-FO membrane. 2. Materials and methods 2.1. Materials If else specified, all chemicals and reagents were analytical grade and used without further purification. For the synthesis of LDH, magnesium chloride hexahydrate (MgCl2·6H2O), aluminum chloride hexahydrate (AlCl3·6H2O), NaOH and Na2CO3 were purchased from Fisher Scientific. Poly (vinyl chloride) PVC (Mw: ~ 48,000) (CAS No. 9002-86-2), N,N Dimethylformamide (DMF), Trimesoyl chloride (TMC), mphenylenediamine (MPD), and n-hexane were purchased from SigmaAldrich for thin film composite membrane preparation. Polyester woven fabric sheet was supplied by Suneeta Garments (Mumbai, India). Sodium chloride (NaCl) purchased from FINAR reagent. For all experiments, deionized water (DI) (pH 5.9 ± 0.2 and conductivity 1.0 (μS/cm)) was used (MilliQ, Millipore, India). 2.2. Preparation of LDH The co-precipitation method was used to prepare LDH [34]. A 100 mL solution of 0.75 M MgCl2·6H2O and 0.25 M AlCl3. 6H2O (molar

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

ratio of Mg/Al = 3:1) was added drop wise at 1 mL/min into 100 mL solution of 1 M NaOH and 1 M Na2CO3. The 2 M NaOH solution was continuously added to maintain pH at 11, and the temperature of solution was maintained at 65 ± 5 °C under a magnetic stirring at 600 rpm. The mixed solution was kept at 65 ± 5 °C for 24 h. The precipitated solution separated by the centrifuge at 6000 rpm and the separated slurry was washed with DI water to make it neutral. The precipitates were dry in an oven at 65 ± 5 °C for 24 h and then crushed with a mortar and pestle to make powder. The prepared LDH was kept at 200 °C for 24 h in the muffle furnace. The calcined LDH powder was screened by 75 μm mesh and a small quantity of LDH was further utilized for XRD and FT-IR analysis.

151

30 mA and 40 kV over 2θ = 0 to 70°. The zeta average diameter of LDH dispersed in DMF was measured by using Zetasizer laser light scattering equipment (Malvern Zetasizer, Nano ZS-90, UK). Fourier transform infrared (FTIR) spectra for prepared LDH was verified for wave numbers 400–4000 cm−1 using a Brucker IFS 66v/s FTIR spectrometer. The mixer of sample and KBr powdered were converted in the pellet for analysis. The cross-sectional and surface morphologies of prepared substrates were observed using field-emission scanning electron microscopy (Thermal FEG-SEM, JSM-7600F, IIT Mumbai, India). For cross-section morphologies, samples were dipped into the liquid nitrogen for 5 min in a closed chamber and carefully fractured.

2.3. PVC/LDH substrate and TFC-FO membrane preparation The PVC/LDH substrates were tailored by the non-solvent phase inversion process. The dope solutions with polymer concentration of 15 wt% in DMF were prepared from PVC and different wt% of LDH (0, 0.5, 1, 1.5, 2, 2.5 and 3) at 60 °C for 8 h under vigorous stirring. The mixtures were cast on polyester woven fabric supported by glass plate using adjusted 100 μm gap size of automatic film applicator (Sheen Instrument Ltd., Vacuum based, Model number: 1133 N) at a speed of 50 mms−1. After casting, immediately glass plate was immersed in DI water coagulation bath at room temperature and kept in a bath for 12 h. The prepared substrates were symbolized as SUB 0, SUB 0.5, SUB 1, SUB 1.5, SUB 2, SUB 2.5 and SUB 3. Subsequently, the thin film composite (TFC) FO membranes were prepared by interfacial polymerization [22]. The prepared substrates were enclosed in order to expose the only top surface for interfacial polymerization. Every framed substrate was immersed in 4 wt% MPD aqueous solution for 5 min, followed by the excess MPD solution removal. After that, the 0.15 wt% TMC/n-hexane solution was softly poured on the MPD-adsorbed substrate and kept it for 1 min. The excess solution on the substrate was drained off and dried in room temperature for 5 min. The prepared TFC membranes were stored in DI water for further evaluation. The fabricated TFC FO membranes were also named as TFCSUB 0, TFC-SUB 0.5, TFC-SUB 1, TFC-SUB 1.5, TFC-SUB 2, TFC-SUB 2.5 and TFC-SUB 3. 2.4. Casting solution viscosity and thermodynamic study The kinematic viscosity of all casting solution was measured by Redwood viscometer at room temperature (29 ± 2 °C). The phase separation of polymer solution can be done by inducing thermodynamic instability between a polymer solution and anti-solvent. Therefore, the constituent's thermodynamic stability is an important factor to study the membrane formation process. It has been well recognized that, the ternary phase diagram of the polymer/solvent/non-solvent system can be used as a tool to study the thermodynamic aspect of the membrane formation process. In most of the study, the cloud point measurement was used to construct a ternary phase diagram [35]. Cloud point measurements were carried out by the titration of the polymer solutions with a concentration of 15 wt% PVC in DMF, to which 0, 0.5, 1, 1.5, 2, 2.5 and 3 wt% LDH was added. All solutions were maintained at 30 °C. The 10 g of homogeneous sample was taken into a sealed container with a magnetic needle for titration and DI water slowly added drop wise by micro-syringe. Due to a high viscous polymer solution, mixing of water and the solution was carried out by heating up to 60 °C and then cooled to 30 °C. When a solution appeared turbid, cloud point was measured by weighing the container to obtain fraction of water added. Subsequently, solvent was added to transfer the turbid solution into a clear and same procedure was followed eight times for each sample.

2.6. Porosity (ε), mean pore radius (rm) and interfacial free energy analysis of the substrates The porosity (ε) of the substrates was determined by the gravimetric analysis method. The small piece of substrates was vacuum-dried for 0.5 h prior to soak in DI water for 24 h. After 24 h, piece of substrate weighted without excess water on the surface (m1 , g). Subsequently, wet substrate was dried for overnight in the vacuum dryer and weighted (m2, g) to determine the amount of absorbed water into the membrane pores and the polymer weight. The porosity of the membrane substrates ε was then calculated as follow [22]: ε¼

ðm1 −m2 Þ=ρw ðm1 −m2 Þ=ρw þ m2 =ρp

ð1Þ

where, ρw is the density of water and ρp is the density of polymer. The mean pore radius (rm) of the substrates was also measured by the Guerout–Elford–Ferry equation, as follows [36]: rm ¼

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð2:9−1:75εÞ  8μtQ ε  A  ΔP

ð2Þ

where Q is the water flux (m3/s), μ is the water viscosity (8.9 × 10−4 Pas) and ΔP is the pressure difference (100 kPa). The surface wettability of the substrates was evaluated by the contact angle measurement of DI water using the sessile drop method. Before contact angle measurement substrates were vacuum dried. Minimum five contact angle was measured for every substrate at different location using OCA 15EC, Dataphysics Instruments, Germany. The relative solid liquid interfacial free energy or “wettability” was quantified using Young–Dupré equation as follow, −ΔGML ¼ γL ð1 þ cosθÞ

ð3Þ

where, γL is the total liquid surface tension of DI water (72.8 mJ m−2 at 25 °C), θ the measured contact angle, M and L denoted for membrane and liquid respectively. The larger value of −ΔGML indicates more wettable surface [37]. Additionally, Lifshitz–van der Waals acid–base approach was used to determine surface tension components of solids and interfacial free energies per unit area between two phases. By measuring contact angles of one apolar liquid (diiodomethane) and two polar liquid (water and glycerol), the Lifshitz-van der Waals (γLW), electron donor (γ−), and electron acceptor (γ+) components of the membrane surface tension were calculated using the extended Young–Dupré equation as follow, ð1 þ cosθÞγ L ¼ 2

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffi þ LW − γ LW γ− γþ M γL þ M γL M γL þ

ð4Þ

2.5. Characterization of LDH and PVC/LDH substrates X-ray diffraction (XRD) pattern for synthesized LDH was obtained using an X-ray diffractometer with Cu–Kα radiation generated at

Eq. (4) was solved simultaneously using the contact angle of − diiodomethane and water and glycerol with known values of γLW L γL and γ+ . The total surface energy of the membrane surfaces was L

152

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

calculated by a summation of the surface tension due to Lifshitz–van der Waals and Lewis acid-base components as follow, γ TOT ¼ γLW þ γ AB

ð5Þ

pffiffiffiffiffiffiffiffiffiffiffiffiffi where, γAB ¼ 2 γþ γ− [37]. The total free energy of cohesion of membrane interfaces immersed in water or “hydrophilicity” (ΔGTOT MLM) as also calculated by using the surface tension components of the membrane and the water. Higher the value of free energy, more the membrane is hydrophilic or noncohesive when immersed in water. 2.7. Mass transfer characteristics of TFC-FO membranes The mass transfer properties of TFC-FO membranes, water permeability coefficient (A, L m−2 h−1 bar−1), salt rejection (Rs, %), and salt permeability coefficient (B, L m−2 h− 1) were determined by using a lab-scale RO filtration setup. The total effective membrane area of the RO module was 19.6 cm2. The experiments were performed using the feed solution containing 500 ppm NaCl at 100 KPa. The water permeability coefficient (A) of the membranes was calculated as follows: ΔV A¼ Am  Δt  ΔP

ð6Þ

where Am is the effective membrane area, and ΔV is the permeate volume, Δt is time and ΔP is pressure difference. To find membrane salt rejection Rs, the following equation was employed.   Cp Rs ¼ 1−  100% Cf

ð7Þ

where Cf and Cp are the salt concentrations in the feed and permeate solution, respectively. The salt permeability coefficient (B) was calculated based on the solution–diffusion theory by [38]: 1−R B ¼ R AðΔP−ΔπÞ

2.9. Evaluation of forward osmosis performance Fig. 1 shows the schematic diagram of a lab-scale FO test setup. The rectangular stainless steel module was designed with two symmetric parts having the size of 7 cm × 5 cm × 0.5 cm, creating an effective area of 35.5 cm2 and a total volume of 17.5 cm3. The FO membrane was sandwiched between two stainless steel plate using a rubber gasket and six stainless steel screws. Feed solution and draw solution were individually pumped using peristaltic pumps. The draw solution tank was located on a digital weighing machine attached with AD-1688 Weighing Data Logger - A&D Company Ltd. to determine permeates flux. The conductivity meter (HACH, USA) was used to determine the conductivity of feed solution at the end of each cycle to measure reverse salt flux. All the TFC-FO membranes were tested for active layer facing feed solution (AL–FS) and active layer facing draw solution (AL–DS) mode. Each experiment was performed for 180 min and repeated three times. For performance evaluation, DI water as the feed solution and 1 M NaCl solutions as a draw solution was used. The FO water flux Jw was determined by measuring the weight change of draw solution as follow; Jw ¼

ΔV Am Δt

ð11Þ

where ΔV is the volume change of feed solution, Am is an effective membrane area, Δt is the measuring time interval. The reverse salt flux Js, was determined by the change of the salt concentration in feed by the following equation; Js ¼

ΔðC t V t Þ Am Δt

ð12Þ

where Am is an effective membrane area, Δt is a time interval, Ct and Vt are the salt concentration and the volume of the feed measured at the beginning and the end of the time interval, respectively. 3. Results and discussion 3.1. XRD analysis of LDH and prepared substrates

ð8Þ

where A is water permeability coefficient, ΔP is pressure difference and Δπ is the osmotic pressure difference across the membrane. 2.8. Determination of membrane structural parameter (S)

Fig. 2 shows XRD patterns of LDH, SUB 0, SUB 0.5, SUB 1, SUB 1.5, SUB 2, SUB 2.5 and SUB 3. The XRD peaks of LDH (Fig. 2(a) are attributed to hydrotalcite, a hydroxycarbonate of magnesium (JCPDS card 22–700) and aluminum represented as Mg6Al2(OH)16CO3.4H2O. The XRD patterns of LDH were indexed based on a hexagonal unit cell. The strong peak at approximately 11.58° 2 theta (d-value = 7.64 a.u.) was

The structural parameter (S) examines the measure of internal concentration polarization in the porous substrate layer. It is recognized as resistance to the water flux and expressed as the multiplication of tortuosity and thickness, divided by porosity of the membrane's substrate. The S value was calculated using the internal concentration polarization (ICP) model as expressed in Eq. (6) (FO mode) [21]. S¼

D Jv

 ln

 A:π draw þ B A:πfeed þ J v þ B

ð9Þ

where D is the solute diffusion coefficient in water; πdraw and πfeed is the osmotic pressures of the draw solution and the feed solution, respectively. While the thickness t of the membranes was measured by an analog micrometer, the ratio of tortuosity and porosity (τ/ε) was determined as follow:

τ S ¼ ε t

ð10Þ

Fig. 1. Schematic diagram of FO system used in this study; (1) membrane module (2) FO membrane (3) Feed solution reservoir (4) draw solution reservoir (5) weighing data logger.

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

153

713 nm to 1990 nm and Zeta average diameter of LDH particles were 1180 nm. At a maximum volume percentage of 12.8%, 18.7%, 21.5%, 19.3% and 12.5%, the LDH particles size of 1106 nm, 1281 nm, 1484 nm, 1718 nm and 1990 nm respectively were detected. This result was good agreement with Zhao et al. [41]. The functional groups and interaction between intercalated anion and inorganic compounds of LDH were investigated by FT-IR shown in Fig. 3(b). In accordance with previous FT-IR studies of LDH [42], all peaks have been allotted to corresponding functional groups. The peak at 3440.21 cm− 1 was due to the stretching of OH group attached to Mg and Al in the layer. The peak at 2946.1 cm−1 was attributed to the hydrogen bonding of interlayer water with CO2– anions. The 3 1615.59 cm−1 peak was observed due to the bending vibration of interlayer water. Due to the vibration of CO2– 3 group in the LDH, 1343.91, 960.6 and 564.15 cm−1 were detected. The wavenumbers at 960.60, 807.93 and 564.15 cm−1 were attributed to the stretching of Al\\O bond.

Fig. 2. XRD patterns of LDH (a), SUB 3 (b), SUB 2.5 (c), SUB 2 (d), SUB 1.5 (e), SUB 1 (f), SUB 0.5 (g), and SUB 0 (h).

attributed to diffraction from the (003) family of a crystallographic plane [39]. According to Carlino [39], these planes are spaced one third of unit cell distance apart and correspond to the interlayer repeat distance. The interlayer spacing between metal hydroxide sheet (approximately 4.80 a.u.) and the intercalated anions (approximately 2.79 a.u.) can be termed as d spacing d300 or as the gallery height. These d spacing values were found good agreement with reported data by Phuong [40]. The strong evidence of a high crystalline order of the LDH attributed to the presence of reflection (003), (006), (009), (015) and (018). Hence, it was confirmed that well structure LDH synthesized by given procedure. In case of the SUB 0.5, 1, 1.5, 2, 2.5, and 3 substrates XRD pattern indicate the dispersion of LDH particles in PVC matrix. The XRD peaks of SUB 3, 2.5, 2, 1.5, 1, 0.5 and PVC (SUB 0) substrates are shown in Fig. 2(b–h). The peaks are disappearing in the range of 2 theta = 10– 30° when concentration of LDH decreased as shown in Fig. 2(b–h). These disappearances of LDH peaks in the PVC/LDH composite might be due to the partially exfoliated structure of substrates. 3.2. Particle size distribution and FTIR analysis of LDH The particle size distribution by volume is shown in Fig. 3(a). It shows that the maximum particle size of LDH was found between

3.3. Pore size, porosity, pure water permeability (PWP) and interfacial free energy analysis of the substrates Pore size and porosity are not just a part of the structural parameter but also plays an important role in formation of the defect free polyamide selective layer. The effect of pore size on polyamide layer formation has been investigated by Singh et al. [43]. According to the Singh et al. [43], the small pore size of the substrate membrane would restrain the components of polyamide to penetrate into the pores. Therefore, the polyamide layer formed on the surface would be thick dense skin and less bonded on substrate. In other case, the larger pores would allow components to form polyamide inside the pores with the thin layer of polyamide. The porosity, average pore diameter, and pure water permeability (PWP) after the phase inversion are presented in the Table 1. The average pore diameter and porosity of substrate were increased with increasing LDH concentration. It has already been investigated that the incorporation of hydrophilic components into the membrane casting solution could speed up the mass exchange rate of solvent and nonsolvent, causing rise in pore size and porosity [21]. This might be the reason for the increasing pore size and porosity of substrate as LDH concentration increases. This increasing pore size and porosity of substrate was a further consequence that PWP of the substrate also increases as LDH content increased. The initial PWP for pure PVC (SUB 0) substrate was about 145.1 LMH bar−1 and rate began to increase with the addition of LDH (Table 1). These results could be supported by the HagenPoiseuille pore flow model [44], according to the model pore size and porosity is directly proportional to the permeability.

Fig. 3. (a) Particle size distribution, and (b) FTIR spectra of LDH.

154

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

Table 1 Porosity, mean pore diameter, pure water permeability (PWP) and kinematic viscosity as a function of LDH concentration. Substrates

Porosity (%)

Mean pore diameter (nm)

PWP (LMH bar−1)

Kinematic viscosity (mm2/s)

SUB 0 SUB 0.5 SUB 1 SUB 1.5 SUB 2 SUB 2.5 SUB 3

62.87 67.68 72.98 73.15 77.53 79.59 79.88

56.44 ± 1.2 65.36 ± 0.9 68.52 ± 2.1 77.86 ± 1.8 83.76 ± 3.8 105.44 ± 3.6 117.46 ± 4.3

145.1 ± 8.9 239.4 ± 16.2 281 ± 10.5 398 ± 13.8 490 ± 9.7 609 ± 20.1 958 ± 14.9

5.42 8.01 10.40 15.16 19.04 24.55 32.86

± ± ± ± ± ± ±

0.8 1.3 1.5 0.7 1.2 1.8 1.4

These all quantitative analysis was further confirmed with contact angle measurement and SEM morphological analysis. This is may be due to the fact that PWP is not only depends on pore size and porosity but also on hydrophilicity of substrate. It has been suggested that the hydrophilic surface of the membrane could easily allow water molecule to penetrate the membrane. Hence, the hydrophilic membrane could be more favorable for the enhancing water flux [21]. The average water contact angle of pure PVC (SUB 0) substrate was about 75.2° ± 1.6 and moved towards the hydrophilicity by reducing an average contact angle as low as 38.3° ± 0.9 as concentration of LDH increased (Table 2). This large decreasing contact angle indicate that incorporation of LDH particle enhanced the surface free energy of substrate which helped the water drop to spread more over the substrate. The presence of the abundant amount of hydroxyl (OH−) group around the Mg/Al molecule and water molecule located in the interlayer space of a LDH molecule might have pushed the substrate towards hydrophilicity. Table 2 listed the surface energy parameters and interfacial free energy of cohesion calculated from one apolar liquid and two polar liquid contact angle measurements. The value of ΔGTOT MLM is used to quantitatively define hydrophilicity/hydrophobicity of the membrane. A negative values of ΔGTOT MLM indicates the thermodynamic instability (hydrophobicity/attraction) and positive values of ΔGTOT MLM indicates the thermodynamic stability (hydrophilicity/repulsion) [37]. From these values, a quantitative approach in terms of hydrophobicity/hydrophilicity of membrane can be obtained. Except the ΔGTOT MLM value of pure PVC (SUB 0), all ΔGTOT MLM values were positive. The wettability (−ΔGML) of SUB 0 was obtained 91.4 mJ m−2 and the interfacial free en−2 ergy of cohesion (ΔGTOT . This might MLM) of SUB 0 was found −19.4 mJ m TOT −2 be due to the low value of surface energy γ (=2mJm ) of SUB 0 substrate. Moreover, wettability (−ΔGML) increased from 91.4 to 130 mJ m−2 with increasing LDH concentration. The total free energy of −2 cohesion (ΔGTOT and MLM) also increased from −19.4 to 13.6 mJ m showed the presence of a LDH particle in PVC matrix was responsible for increasing hydrophilicity. This increasing trend was attributed to the presence of a LDH particle in PVC matrix which was caused of enhancing the Lifshitz–van der Waals and Lewis acid base components. Fathizadeh et al. [45] suggested that not only pore size but hydrophilicity of substrate also plays a vital role in formation of the polyamide layer. Fathizadeh et al. [45] reported that due to the presence of hydrophilic groups of additive increases the adsorption of MPD on the surface and within the pore. More hydrophilic substrate could form a thin and less rough polyamide layer which can act as an antifouling surface.

3.4. Morphological study SEM analysis was used to depict the effect of LDH on the crosssectional and surface morphology of substrate structure as shown in Fig. 4. All substrate membranes had appeared as an asymmetric structure comprising of a top dense layer, a thick porous sub layer with finger like macrovoids as shown in Fig. 4. Furthermore, not every substrate membrane had exposed pores at the dense top layer. The interaction rate of solvent in casting solution and non-solvent in the coagulation bath is a dominating factor for the phase inversion process. The existence of macrovoids were attributed to the good miscibility between solvent (DMF) and non-solvent (water), which leads to generate pores by exchanging water in the coagulation bath with solvent in casting solution. Incorporation of LDH particles even at low concentration (0.5 wt%) had affected both thermodynamic and mass transfer properties of PVC/DMF/water ternary system, by turning the sponge like structure of pristine PVC (SUB 0) substrate to more open pore structure. As we increased the concentration of LDH up to 2 wt%, increasing connectivity between up layer and sub layer, enhancement in pore size and porosity, and decrement of sponge like structure were observed. This is a good agreement with the result of pore size and porosity of the substrate in the Table 1. The presence of a large amount of OH– group around the Mg/Al molecule and water molecule located in the interlayer space made the casting solution thermodynamically unstable when it came to contact with non-solvent. This situation of casting solution might have caused to decrease in free energy by forming polymer-poor phase (nascent pores) and polymer-rich phase [46]. It can be seen from Fig. 4, surface pores size had increased as LDH concentration increased. It is attributed to the great affinity of a LDH particle towards water which allowed more water diffusion in to casting solution [46]. From cross-section images of all substrate, the LDH loading at 0.5 to 2 wt% resulted in substrate with long finger-like voids with insignificant presence of sponge like structure. It was due to the increased concentration of LDH in casting solution affect the mass transfer rates of nonsolvent influx and solvent escape. The presence of LDH particles in casting solution made solution more affinitive towards non-solvent, and might have responsible for the rapid and uninterruptedly demixing process [46]. This favorable macrostructure was expected to decrease structural parameter by providing more passage for water in the FO separation process. But as the concentration of LDH further increased for 2.5 and 3 wt%, disturbed finger-like structure seem to be appeared. This was attributed

Table 2 Contact angle, surface tension parameters and interfacial free energy of cohesion of substrate. Substrates

Contact angle (°)

γLW S (mJ/m2)

γ+ S (mJ/m2)

γ− S (mJ/m2)

γAB S (mJ/m2)

−ΔGML (mJ/m2)

γTOT (mJ/m2)

ΔGTOT MLM (mJ/m2)

SUB 0 SUB 0.5 SUB 1 SUB 1.5 SUB 2 SUB 2.5 SUB 3

75.2 ± 1.6 66.9 ± 1.5 58 ± 0.9 51.9 ± 1 46.6 ± 1.3 42.5 ± 0.6 38.3 ± 0.9

34.10 35.86 36.90 38.65 40.90 41.50 43.42

0.960 0.367 0.521 0.096 0.036 0.096 0.309

16.23 31.18 36.20 38.62 41.54 39.19 38.62

7.90 6.80 8.70 3.85 2.45 3.90 6.90

91.4 101.4 111.4 117.7 122.8 126.5 129.9

42.0 42.6 45.6 42.5 43.3 45.3 50.3

−19.4 6.1 12.8 17.3 21.2 16.7 13.6

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

to increasing viscosity of casting solution causes delayed in the phase inversion process. These results were corroborated with thermodynamic study by cloud point measurement.

155

The effects of LDH on the morphology of the membrane were substantiated by thermodynamic phase equilibria. Whereas, the morphology of the final prepared membrane is intensely depends on the casting

Fig. 4. Surface and cross-section morphology of prepared substrates (a to n for SUB 0 to SUB 3).

156

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

Fig. 4 (continued).

Fig. 5. Ternary phase diagram of PVC/LDH/DMF/water system.

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

157

Table 3 Thickness, Pure water permeability coefficient A, Salt permeability coefficient B, salt rejection Rs, Structural parameter S and τ/ε of prepared TFC-FO membranes based on different substrate. Membrane

Thickness (×10−6 m)

A (LMH bar−1)

B (LMH)

Rs (%)

TFC-SUB 0 TFC-SUB 0.5 TFC-SUB 1 TFC-SUB 1.5 TFC-SUB 2 TFC-SUB 2.5 TFC-SUB 3

97 89 95 87 91 94 88

1.96 2.33 2.63 3.00 3.61 3.73 4.04

0.085 ± 0.0055 0.1084 ± 0.023 0.1238 ± 0.019 0.1489 ± 0.041 0.1816 ± 0.033 0.2276 ± 0.011 0.2834 ± 0.029

83.8 81.6 80.8 79.6 79.2 73.0 68.4

± ± ± ± ± ± ±

15.9 08.2 10.8 07.3 05.8 11.2 13.7

± ± ± ± ± ± ±

0.019 0.092 0.084 0.021 0.019 0.017 0.045

solution composition, binodal curve position and precipitation path [47]. The thermodynamic study involves the ternary phase diagram which helps to demonstrate the main two path of the phase inversion process. These are: i) liquid-liquid phase separation, in which binodal boundary separates the triangle into two region: homogeneous solution region, and two phase region, ii) solidification, in which movement of polymer precipitation restrained by increasing viscosity of polymer solution to fix the membrane structure. The phase inversion process of seven membrane casting solution (i.e., 15/0, 15/0.5, 15/1, 15/1.5, 15/2, 15/2.5, and 15/3 wt% PVC/LDH composite) were investigated by thermodynamic study (Fig. 5). By increasing the LDH concentration into the casting solution, the binodal line shifted nearer to the polymer/water axis. Shifting of the binodal line towards the polymer/water axis indicated that more water required for precipitation of PVC in the quaternary system. Therefore, the presence of LDH particle made casting solution more tolerant to water without causing phase separation. This was related to the increasing concentration of LDH particles results in increasing surface free energy of casting solution. Consequently, this free energy difference between casting solution and water was acted as driving force for more water diffusion, more and large pores formation [48]. These experimental results could be in agreement with the pore size and porosity of substrate (Table 1), and top surface SEM images. Furthermore, as a concentration of LDH increased, viscosity of casting solution increased (as shown in Table 1) and might have affected the rate of precipitation of casting solution. It exposed that the overall mass transfer was thermodynamically dominant with increased concentration of LDH. Up to 2 wt% concentration of LDH, the increasing regularity of finger like structure could be attributed to the consecutive instantaneous demixing process of casting solution. This indicated that the thermodynamic factor overcomes the rheological hindrance related with viscosity increased. In case of 2.5 and 3 wt% LDH concentration, further increased in viscosity of casting solution might have caused an inappropriate way of water permeation (delayed demixing) into casting solution resulted in a disturbing finger like structure (Fig. 4). The delayed demixing process slows the liquid-liquid phase separation and restrains the macrovoids formation [48].

± ± ± ± ± ± ±

0.12 0.09 0.16 0.27 0.18 0.29 0.11

S (μm)

τ/ε

699 480 357 312 303 387 538

7.20 5.40 3.76 3.59 3.33 4.12 6.12

could be attributed to increased hydrophilicity and porosity of substrate which have a tendency to form a thin polyamide layer [13]. Nevertheless, salt permeability coefficient B increased from 0.085 ± 0.0055 LMH to 0.29 ± 0.029 LMH as LDH concentration increased. The salt rejection Rs of TFC-SUB 0 membrane slightly decreases from 83.8 ± 0.12% to 79.2 ± 0.18% of the TFC-SUB 2 membrane. Further increased in concentration of LDH to 2.5 and 3 wt%, drastic decreased in salt rejection was observed for TFC-SUB 2.5 (73 ± 0.29%) and TFC-SUB 3 (68.4 ± 0.11%) membranes. Both increasing B and decreasing Rs could also be attributed to the increasing hydrophilicity and large porosity of substrate which caused thin polyamide layer formation, and low diffusion resistance to the salt. Though, the value of A is high for TFC membrane, it does not expose the applicable performance of FO. Because the use of hydraulic pressure to calculate A value makes it insufficient to evaluation of FO performance. To further recognize the effect of LDH on the TFC membrane, osmosis experiments were performed. It has been recognized and understands that support layer must have low structural parameter (S) to mitigate the effect of ICP. Osmosis experiments were carried out to calculate the S parameter and depicted in the Table 3. The additions of LDH up to 2 wt% concentration in PVC had led to decrease S parameter from 699 μm to 303 μm. The further addition of LDH to 2.5 and 3 wt%, slightly increasing S parameters were observed. The low structural parameter of TFC-SUB 2 membrane could be attributed to many several factors: (i) the finger like pore structure (less tortuous) (ii) the high porosity and (iii) thinner sponge like layer close to the top of substrate. These results concluded that PVC substrate with 2 wt% concentration of LDH would be more promising for higher water flux performance. Whereas, the more addition of LDH could be unfavorable for FO performance. Teraferri et al. [49], suggested that the ratio of

3.5. Characterization TFC-FO membrane separation properties The characterized substrates were employed to fabricate TFC-FO membrane by applying polyamide layer. The effect of different substrate on the pure water permeability coefficient (A), salt permeability coefficient (B) and salt rejection (Rs) of the synthesized TFC-FO membranes were compared in the Table 3. The TFC-SUB 0 membrane had shown pure water permeability A of 1.96 ± 0.019 LMH bar−1 with applied pressure of 1 bar in RO testing mode. Whereas, the TFC-SUB 0 membrane could reject 83.8 ± 0.12% of salt with 500 ppm NaCl solution as feed at 1 bar applied pressure. As concentration of LDH increases, water permeability increased up to 4.04 ± 0.045 LMH bar− 1 for the TFC-SUB 3 membrane with the same RO testing mode conditions which was nearly 106% higher than the TFC-SUB 0 membrane. This

Fig. 6. The water flux and reverse salt flux of TFC membranes in AL-FS and AL-DS orientation.

158

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

Fig. 7. Water flux to reverse salt flux (Jw/Js) for TFC membranes in AL-FS and AL-DS.

τ/ε can be an appropriate sign of the intrinsic diffusional resistance employed by the membrane structure. The ratio of τ/ε can directly provide information about the effect of the membrane structure on the FO performance. From Table 3, it has also been seen that the structure of TFC-SUB 2 was optimal for the FO process in relations of osmotic flux.

The concentrative ICP, dilutive ICP, ECP, structural parameter and salt diffusion are the main factors affecting the water flux of the FO membrane. Due to use of DI water as feed, the concentrated ICP was considered as negligible. Increasing hydrophilicity and decreasing structural parameters S might have responsible to suppress dilutive ICP and results in a high water flux for 0 to 2 wt% LDH concentration. The decrement in the water flux for TFC-SUB 2.5 and the TFC-SUB 3 wt% membrane could be attributed to the drastic reverse salt flux (Js) increment caused an enhancement of osmotic pressure in feed solution. The smaller S value of TFC-SUB 2 (303 μm) had produced lower dilutive ICP and resulted in a higher water flux membrane. The increment in Js for TFC-SUB 2.5 and TFC-SUB 3 might be due to the defective PA active layer formation. The PA active layer supported by SUB 2.5 and SUB 3 substrates were incapable to generate a necessary driving force in terms of osmotic pressure difference. Fig. 7 shows the reverse solute flux selectivity (RSFS), i.e. the ratio of water flux/reverse salt flux (Jw/Js) for TFC membranes. High Jw and low Js is the main criteria for an ideal FO membrane, hence, the higher Jw/Js ratio would be preferable for the FO membrane. For the prepared membrane with 0 to 2 wt% LDH contents substrate, Jw/Js ratios exhibited increasing trends (Fig. 7) and for TFC-SUB 2.5 to TFC-SUB 3 wt% membranes decreasing tends were observed. This can be hypothesized that even though incorporation of the LDH particle was reason to make the substrate membrane more hydrophilic and porous, with higher concentration of LDH might have made substrate adverse for PA active layer formation. The high performance of TFC-SUB 2 might have attributed to the perfect substrate membrane structure resulted in formation of a defect-free PA active layer. In Table 4 compares the Jw/Js ratio of prepared TFC-SUB 2 membrane with the reported flat sheet TFC-FO membrane. It can be seen that, the first time fabricated lab-scale TFC-SUB 2 membrane was comparable with the recently reported flat sheet lab-scale the FO membrane.

3.6. Osmotic flux performance of TFC-FO membrane 4. Conclusions The comparisons between the AL-FS and AL-DS performance of prepared TFC membranes using DI water as feed and 1 M NaCl as a draw solution portrayed in Fig. 6. The AL-DS orientation for all TFC membrane revealed higher water flux (Jw) and higher reverse salt flux (Js) than AL-FS orientation. This fact attributed to the strong dependency of net osmotic pressure difference on a different degree of dilutive/concentrative internal concentration polarization (ICP) in AL-FS and AL-DS orientations [50]. Nevertheless, water flux (Jw) in both AL-FS and AL-DS increased with an increase in LDH content in the supports from 0 to 2 wt%. Whereas, water flux started decreasing with a further increase in LDH concentration from 2.5 to 3 wt% for both AL-FS and AL-DS. In the osmotic flux experiment, the TFC-SUB 2 membrane had been emerged as the best performing membrane with an average Jw of 37.46 ± 0.85 LMH and Js of 3.57 ± 0.2 gMH in the AL-FS, and average Jw of 50.89 ± 1.13 LMH and Js of 13.284 ± 0.67 gMH in the AL-DS. Nevertheless, in comparison with the TFC-SUB 2 membrane, TFC-SUB 2.5 and TFC-SUB 3 membranes could have achieved Jw of 31.92 ± 1.8 LMH and 25.78 ± 0.8 LMH in AL-FS respectively. Similarly, for ALDS orientation, TFC-SUB 2.5 and TFC-SUB 3 membranes could achieve Jw of 48.37 ± 0.61 LMH and 47.77 ± 1.1 LMH in AL-DS respectively.

For the first time, polyvinyl chloride (PVC) and layered double hydroxide (LDH) composite substrates were successfully fabricated for FO membrane preparation via interfacial polymerization. Seven different concentration of LDH composite substrate were demonstrated. The effect of LDH loading on the morphology and transport properties of substrates and TFC membrane were examined. The substrates with LDH were more porous, hydrophilic and low value of S than substrate without LDH. The 2 wt% LDH supported TFC membrane was emerged as best performed membrane with high water fluxes of 37.46 ± 0.85 LMH and 50.89 ± 1.13 LMH using 1 M NaCl as the draw solution and DI water as feed in AL-FS and AL-DS orientation respectively. Reverse solute flux selectivity (RSFS) (Jw/Js) of TFC-SUB 2 was found high among all prepared membrane. Thus, PVC/LDH support for a fabrication of the FO TFC membrane can be used as new material with discovered performance of low S value. Acknowledgement Authors acknowledge the financial support from the Institute (SVNIT, Surat) Research Grant {Dean (R&C)/1503/2013-14} from the

Table 4 Comparison of Jw/Js and other experimental data of prepared TFC-SUB 2 membrane with reported flat sheet TFC FO membrane at draw and feed solution conditions. Membrane

Feed solution

Draw solution

Jw/Js AL-FS/AL-DS (L/g)

Water permeability coefficient A (LMH bar−1)

Salt permeability coefficient B (LMH)

Salt rejection Rs (%)

Reference

TFC-SUB 2 No pre-wetted HTI-TFC Pre-wetted HTI-TFC Sulphonated polyethersulfone (SPES) 40% PTA-POD TFC PSf/0.5 wt% halloysite nanotubes TFC PSf and 0.25 wt% GO TFC

DI water DI water DI water DI water DI water 10 mM NaCl DI water

1 M NaCl 1 M NaCl 1 M NaCl 2 M NaCl 1 M NaCl 2 M NaCl 0.5 M NaCl

10.49/3.83 7.58/2.35 3.57/1.86 3.54/3.8 6.81/6.37 1.9/1.57 5.75/6.42

3.61 1.75 2.4 2.9 1.893 2 1.76

0.18 1.25 1.75 0.184 0.207 0.33 0.19

79.2 91.5 90.5 91.1 96.8 93.7 98.71

This work [51] [51] [52] [53] [54] [22]

P.M. Pardeshi et al. / Desalination 421 (2017) 149–159

project entitled “A revolutionary approach for regeneration of graphene based membrane and draw agent in UASB-EFO (Up flow anaerobic sludge blanket-Electric field attached, Forward Osmosis) for the production of pure water from municipal wastewater” for carrying out research work. We thank Sophisticated Analytical Instrument Facility (SAIF), IIT Bombay for analytical facilities. References [1] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology, and the environment, Science 333 (2011) 712–717. [2] D.L. Shaffer, J.R. Werber, H. Jaramillo, S. Lin, M. Elimelech, Forward osmosis: where are we now? Desalination 356 (2015) 271–284. [3] J.O. Kessler, C.D. Moody, Drinking water from sea water by forward osmosis, Desalination 18 (1976) 297–306. [4] J.R. McCutcheon, R.L. McGinnis, M.A. Elimelech, Novel ammonia-carbon dioxide forward (direct) osmosis desalination process, Desalination 174 (2005) 1–11. [5] J. Zhang, Q. She, V.W.C. Chang, C.Y. Tang, R.D. Webster, Mining nutrients (N, K, P) from urban source-separated urine by forward osmosis dewatering, Environ. Sci. Technol. 48 (2014) 3386–3394. [6] F.X. Kong, H. Yang, Y. Wu, X. Wang, Y.F. Xie, Rejection of pharmaceuticals during forward osmosis and prediction by using the solution–diffusion model, J. Membr. Sci. 476 (2015) 410–420. [7] B.D. Coday, B.G.M. Yaffe, P. Xu, T.Y. Cath, Rejection of trace organic compounds by forward osmosis membranes: a literature review, Environ. Sci. Technol. 48 (2014) 3612–3624. [8] N. Akther, A. Sodiq, A. Giwa, S. Daer, H.A. Arafat, S.W. Hasan, Recent advancements in forward osmosis desalination: a review, Chem. Eng. J. 281 (2015) 502–522. [9] W.J. Lau, S. Gray, T. Matsuura, D. Emadzadeh, J.P. Chen, A.F. Ismail, A review on polyamide thin film nanocomposite (TFN) membranes: history, applications, challenges and approaches, Water Res. 80 (2015) 306–324. [10] N.N. Bui, J.R. McCutcheon, Hydrophilic nanofibers as new supports for thin film composite membranes for engineered osmosis, Environ. Sci. Technol. 47 (2013) 1761–1769. [11] Y. Wang, R. Ou, Q. Ge, H. Wang, T. Xu, Preparation of polyethersulfone/carbon nanotube substrate for high-performance forward osmosis membrane, Desalination 330 (2013) 70–78. [12] M. Yasukawa, S. Mishima, M. Shibuya, D. Saeki, T. Takahashi, T. Miyoshi, H. Matsuyama, Preparation of a forward osmosis membrane using a highly porous polyketone microfiltration membrane as a novel support, J. Membr. Sci. 487 (2015) 51–59. [13] A.K. Ghosh, E.M.V. Hoek, Impacts of support membrane structure and chemistry on polyamide–polysulfone interfacial composite membranes, J. Membr. Sci. 336 (2009) 140–148. [14] G.M. Geise, H.S. Lee, D.J. Miller, B.D. Freeman, J.E. McGrath, D.R. Paul, Water purification by membranes: the role of polymer science, J. Polym. Sci. Polym. Phys. 48 (2010) 1685–1718. [15] J. Yin, B. Deng, Polymer-matrix nanocomposite membranes for water treatment, J. Membr. Sci. 479 (2015) 256–275. [16] J. Shen, H. Ruan, L. Wu, C. Gao, Preparation and characterization of PES–SiO2 organic–inorganic composite ultrafiltration membrane for raw water pretreatment, Chem. Eng. J. 168 (2011) 1272–1278. [17] H. Rabiee, M.H.D.A. Farahani, V. Vatanpour, Preparation and characterization of emulsion poly(vinyl chloride) (EPVC)/TiO2 nanocomposite ultrafiltration membrane, J. Membr. Sci. 472 (2014) 185–193. [18] S. Zinadini, A.A. Zinatizadeh, M. Rahimi, V. Vatanpour, H. Zangeneh, Preparation of a novel antifouling mixed matrix PES membrane by embedding graphene oxide nanoplates, J. Membr. Sci. 453 (2014) 292–301. [19] N. Ma, J. Wei, S. Qi, Y. Zhao, Y. Gao, C.Y. Tang, Nanocomposite substrates for controlling internal concentration polarization in forward osmosis membranes, J. Membr. Sci. 441 (2013) 54–62. [20] Y. Wang, R. Ou, Q. Ge, H. Wang, T. Xu, Preparation of polyethersulfone/carbon nanotube substrate for high-performance forward osmosis membrane, Desalination 330 (2013) 70–78. [21] D. Emadzadeh, W.J. Lau, T. Matsuura, M. Rahbari-Sisakht, A.F. Ismail, A novel thin film composite forward osmosis membrane prepared from PSf–TiO2 nanocomposite substrate for water desalination, Chem. Eng. J. 237 (2014) 70–80. [22] M.J. Park, S. Phuntsho, T. He, G.M. Nisola, L.D. Tijing, X.M. Li, G. Chen, W.J. Chung, H.K. Shon, Graphene oxide incorporated polysulfone substrate for the fabrication of flatsheet thin-film composite forward osmosis membranes, J. Membr. Sci. 493 (2015) 496–507. [23] X. Liu, H.Y. Ng, Fabrication of layered silica–polysulfone mixed matrix substrate membrane for enhancing performance of thin-film composite forward osmosis membrane, J. Membr. Sci. 481 (2015) 148–163. [24] Y. Wang, R. Ou, H. Wang, T. Xu, Graphene oxide modified graphitic carbon nitride as a modifier for thin film composite forward osmosis membrane, J. Membr. Sci. 475 (2015) 281–289. [25] N.N. Guo, Q.G. Zhang, H.M. Li, X.M. Wu, Q.L. Liu, A.M. Zhu, Facile fabrication, structure, and applications of polyvinyl chloride mesoporous membranes, Ind. Eng. Chem. Res. 53 (2014) 20068–20073.

159

[26] X. Fan, Y. Su, X. Zhao, Y. Li, R. Zhang, J. Zhao, Z. Jiang, J. Zhu, Y. Ma, Y. Liu, Fabrication of polyvinyl chloride ultrafiltration membranes with stable antifouling property by exploring the pore formation and surface modification capabilities of polyvinyl formal, J. Membr. Sci. 464 (2014) 100–109. [27] J. Zhu, Y. Su, X. Zhao, Y. Li, J. Zhao, X. Fan, Z. Jiang, Improved antifouling properties of poly (vinyl chloride) ultrafiltration membranes via surface zwitterionicalization, Ind. Eng. Chem. Res. 53 (2014) 14046–14055. [28] P. Nalawade, B. Aware, V.J. Kadam, R.S. Hirlekar, Layered double hydroxides: a review, J. Sci. Ind. Res. 68 (2009) 267–272. [29] Q. Wang, D. O'Hare, Recent advances in the synthesis and application of layered double hydroxide (LDH) nanosheets, Chem. Rev. 112 (2012) 4124–4155. [30] J. Liu, L. Yu, Y. Zhang, Fabrication and characterization of positively charged hybrid ultrafiltration and nanofiltration membranes via the in-situ exfoliation of Mg/Al hydrotalcite, Desalination 335 (2014) 78–86. [31] C. Del Hoyo, Layered double hydroxides and human health: an overview, Appl. Clay Sci. 36 (2007) 103–121. [32] V.S. Periasamy, J. Athinarayanan, M.A. Alfawaz, A.A. Alshatwi, Carbon nanoparticle induced cytotoxicity in human mesenchymal stem cells through upregulation of TNF3, NFKBIA and BCL2L1 genes, Chemosphere 144 (2016) 275–284. [33] Y. Ge, J.H. Priester, M. Mortimer, C.H. Chang, Z. Ji, J.P. Schimel, P.A. Holden, Long-term effects of multiwalled carbon nanotubes and graphene on microbial communities in dry soil, Environ. Sci. Technol. 50 (2016) 3965–3974. [34] Y.U. Han, W.S. Lee, C.G. Lee, S.J. Par, K.W. Kim, S.B. Kim, Entrapment of Mg−Al layered double hydroxide in calcium alginate beads for phosphate removal from aqueous solution, Desalin. Water Treat. 36 (2011) 178–186. [35] A.C. Sun, W. Kosar, Y. Zhang, X. Feng, A study of thermodynamics and kinetics pertinent to formation of PVDF membranes by phase inversion, Desalination 309 (2013) 156–164. [36] J.F. Li, Z.L. Xu, H. Yang, L.Y. Yu, M. Liu, Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane, Appl. Surf. Sci. 255 (2009) 4725–4732. [37] C.J. van Oss, Hydrophobicity of biosurfaces – origin, quantitative determination and interaction energies, Colloids Surf. B: Biointerfaces 5 (1995) 91–110. [38] X. Jin, C.Y. Tang, Y. Gu, Q. She, S. Qi, Boric acid permeation in forward osmosis membrane processes: modeling, experiments, and implications, Environ. Sci. Technol. 45 (2011) 2323–2330. [39] S. Carlino, Chemistry between the sheets, Chem. Br. 33 (1997) 59–62. [40] N.T.K. Phuong, Entrapment of Mg-Al layered double hydroxide into alginate/polyvinyl alcohol beads for water remediation, J. Environ. Chem. Eng. 2 (2014) 1082–1087. [41] Y. Zhao, F. Li, R. Zhang, D.G. Evans, X. Duan, Preparation of layered double-hydroxide nanomaterials with a uniform crystallite size using a new method involving separate nucleation and aging steps, Chem. Mater. 14 (2002) 4286–4291. [42] W. Yang, Y. Kim, P.K. Liu, M. Sahimi, T.T. Tsotsis, A study by in situ techniques of the thermal evolution of the structure of a Mg–Al–CO3 layered double hydroxide, Chem. Eng. Sci. 57 (2002) 2945–2953. [43] P.S. Singh, S.V. Joshi, J.J. Trivedi, C.V. Devmurari, A.P. Rao, P.K. Ghosh, Probing the structural variations of thin film composite RO membranes obtained by coating polyamide over polysulfone membranes of different pore dimensions, J. Membr. Sci. 278 (2006) 19–25. [44] M. Hayama, F. Kohori, K. Sakai, AFM observation of small surface pores of hollowfiber dialysis membrane using highly sharpened probe, J. Membr. Sci. 197 (2002) 243–249. [45] M. Fathizadeh, A. Aroujalian, A. Raisi, Effect of lag time in interfacial polymerization on polyamide composite membrane with different hydrophilic sub layers, Desalination 284 (2012) 32–41. [46] T.H. Young, L.W. Chen, Pore formation mechanism of membranes from phase inversion process, Desalination 103 (1995) 233–247. [47] M.J. Han, S.T. Nam, Thermodynamic and rheological variation in polysulfone solution by PVP and its effect in the preparation of phase inversion membrane, J. Membr. Sci. 202 (2002) 55–61. [48] Y.N. Yang, J. Wu, Q.Z. Zheng, X.S. Chen, H.X. Zhang, The research of rheology and thermodynamics of organic–inorganic hybrid membrane during the membrane formation, J. Membr. Sci. 311 (2008) 200–207. [49] A. Tiraferri, N.Y. Yip, W.A. Phillip, J.D. Schiffman, M. Elimelech, Relating performance of thin-film composite forward osmosis membranes to support layer formation and structure, J. Membr. Sci. 367 (2011) 340–352. [50] G.T. Gray, J.R. McCutcheon, M. Elimelech, Internal concentration polarization in forward osmosis: role of membrane orientation, Desalination 197 (2006) 1–8. [51] J. Ren, J.R. McCutcheon, A new commercial thin film composite membrane for forward osmosis, Desalination 343 (2014) 187–193. [52] S. Sahebi, S. Phuntsho, Y.C. Woo, M.J. Park, L.D. Tijing, S. Hong, H.K. Shon, Effect of sulphonated polyethersulfone substrate for thin film composite forward osmosis membrane, Desalination (2015)http://dx.doi.org/10.1016/j.desal.2015.11.028. [53] P.H.H. Duong, S. Chisca, P.Y. Hong, H. Cheng, S.P. Nunes, T.S. Chung, Hydroxyl functionalized Polytriazole-co-polyoxadiazole as substrates for forward osmosis membranes, ACS Appl. Mater. Interfaces 7 (2015) 3960–3973. [54] M. Ghanbari, D. Emadzadeh, W.J. Lau, H. Riazi, D. Almasi, A.F. Ismail, Minimizing structural parameter of thin film composite forward osmosis membranes using polysulfone/halloysite nanotubes as membrane substrates, Desalination 377 (2016) 152–162.