Preparation and characterization of sulfated carboxymethyl cellulose nanofiltration membranes with improved water permeability

Desalination 338 (2014) 74–83

Contents lists available at ScienceDirect

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

Preparation and characterization of sulfated carboxymethyl cellulose nanofiltration membranes with improved water permeability Ling-Ling Shao a, Quan-Fu An a,⁎, Yan-Li Ji a, Qiang Zhao a, Xue-San Wang a, Bao-Ku Zhu a, Cong-Jie Gao b,c a b c

MOE Key Laboratory of Macromolecular Synthesis and Functionalization, Department of Polymer Science & Engineering, Zhejiang University, Hangzhou 310027,China Department of Chemical Engineering and Bioengineering, Zhejiang University, Hangzhou 310027, China The Development Center of Water Treatment Technology, Hangzhou 310012, China

H I G H L I G H T S • Hydrophilic SCMC was synthesized and utilized to prepare NF membranes. • SNFMs exhibited higher water permeability while retaining their salt rejection. • SNFMs had good separation performance in treating acidic dye solution.

a r t i c l e

i n f o

Article history: Received 7 November 2013 Received in revised form 11 January 2014 Accepted 29 January 2014 Available online 19 February 2014 Keywords: Nanofiltration membrane Sulfated carboxymethyl cellulose Water permeability Anionic dye

a b s t r a c t Sulfated carboxymethyl cellulose (SCMC) with tailored amount of sulfate groups was synthesized, and their composite nanofiltration membranes (SNFMs) were prepared by solution casting onto polysulfone (PSF) supporting membrane and crosslinking by glutaradehyde (GA). Chemical structure and composition of SCMCs and SNFMs were characterized by Fourier transform infrared spectroscopy and X-ray photoelectron spectroscopy. SNFM surfaces were examined by field emission scanning electron microscopy, water contact angle and streaming potential measurement. The effects of the chemical composition on the hydrophilicity, the surface charge and the nanofiltration performance of SNFMs were determined. It is found that the incorporation of sulfate groups improves the water permeability of SNFMs while retaining their salt rejection. For the feed mixture of aqueous Na2SO4 (1.0 g L−1, pH = 6.5), the optimum water flux is 39.6 L m−2 h−1, which is 2.2 times of the pristine SNFM-0. For the xylenol orange (XO) dye solution (0.1 g L−1, pH = 4.0), the flux of SNFM-3 (30.0 L m−2 h−1) is also higher than that of SNFM-0 (11.1 L m−2 h−1). In addition, SNFM-3 exhibits good separation performance and chemical stability in treating saline anionic dye aqueous solution (MYB/NaCl) in acidic conditions. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Nanofiltration (NF) is a pressure-driven membrane separation technique with the separation performance between reverse osmosis (RO) and ultrafiltration (UF). Compared with RO membrane, NF membrane exhibits lower retention and higher permeability to monovalent ions at lower operating pressure. Compared with UF membrane, NF membrane processes smaller pore size and molecules with molecular weight ranges from 200 to 1000 Da can be retained [1]. Most NF membranes are charged, the rejection of organic and inorganic species is thought to be controlled not only by the stereo-hindrance effect but also by the electrostatic repulsion effect [2]. NF membranes are usually in the form of composite membranes, whereas a plethora of polymers such as polyamide [3,4], Poly (piperazine amide) [5,6], sulfonated polysulfone [7], sulfonated polyether sulfone [8] and cellulose acetate [9,10] have been exploited as selectivity layers on the substrate ⁎ Corresponding author. Tel./fax: +86 571 87953780. E-mail address: [email protected] (Q.-F. An). 0011-9164/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.desal.2014.01.025

membrane. Very recently, polyelectrolyte complexes has emerged as promising membrane materials for NF membranes [11,12]. Due to the above-mentioned characteristic, NF processes have been widely used in water softening, drinking water purification as well as industrial process fluids treatment [13–18]. In some textile and dye manufacturing processes, synthetic dyes are considered as the most difficult to treat due to the complex aromatic molecular structures, which make them more stable and difficult to be biodegraded [19]. NF technique has been utilized for dye bath wastewater since 1990, and in recent years, more and more researchers consider NF as an effective membrane process to treat the dye wastewater stream [20–22]. Cellulose and its derivatives, being biopolymers with abundant and sustainable sources, represent a family of well-established membrane materials [23–25]. When it comes to NF membrane, cellulose derivatives with good hydrophilicity and charge have been actively pursued recently, such as carboxymethyl cellulose (CMC) that possess a combination of crosslink-able hydroxyl groups and carboxylic acid groups. Yu et al. fabricated CMC/polypropylene thin film composite hollow fiber membranes (cross-linked with AlCl3) for NF, allowing for efficient

L.-L. Shao et al. / Desalination 338 (2014) 74–83

removal of anionic dyes from saline aqueous solution as a result of the hydrophilic and the charge properties of CMC [26]. Miao et al. prepared CMC/PVDF composite NF membranes through the coating and crosslinking method (cross-linked with ECH) with good rejection achieved by modulating the preparing conditions [27]. For instance, the rejections to Na2SO4 and NaCl salts were 88.5 and 33.3%, respectively, while the permeation fluxes of those salts were 10.6 and 11.2 L m−2 h−1, respectively. Although CMC highlights a good rejection, it suffers from the relatively low flux. So it is of importance to improve the permeation flux considerably while retaining the selectivity. On this occasion, we consider that the incorporation of strong acid groups such as sulfate or sulfonate groups is a feasible route, because these groups possess permanent charge and improved hydrophilicity, both of which facilitate the water transport through the membrane [28,29]. Moreover, compared with the carboxylate groups that are pH-dependent, sulfate groups are pH-independent [30], making CMC membranes survive acidic conditions that are encountered in practical applications such as the treatment of acid dye wastewater [31]. In this study, a series of the sulfated carboxymethyl cellulose (SCMC) with modulated composition was synthesized and their membranes (SNFMs) were prepared by means of surface coating and chemical cross-linking. The properties of SNFMs and SCMCs were characterized by ATR-FTIR, XPS, SEM, water contact angle and streaming potential measurements. NF performance confirmed that the water flux to salt and dye molecules of SNFMs was highly improved by the incorporating sulfate groups, while the rejection was maintained stable. 2. Experimental

75

Technology, Hangzhou, China. Deionized water (pH ≈ 7) with a resistance of 18 MΩ was used in all experiments.

2.2. Synthesis of sulfated CMC and their composite membranes SCMC was synthesized according to the literature [32] and the reaction scheme was shown in Fig. 1a. Typically, 0.98 g anhydrous p-TsOH, 1.82 g CMC and 45 mL DMA were added into a 100 mL flask and stirred at 60 °C for 50 min. The highly swollen gel-suspension formed was cooled to room temperature and kept stirring for 8 h. A certain amount of SO3/pyridine complex was added into the activated gel-suspension under stirring at room temperature for 1 h. The product was precipitated in acetone and washed three times with acetone. The obtained SCMC was then dissolved in 0.5 M NaOH aqueous solution and precipitated in ethanol and washed three times with ethanol/water solution (volume ratio of 80 to 20) to get rid of the residual salts. Finally, SCMC was dried at 50 °C in vacuum oven to a constant weight. In the next step, the composite NF membranes SNFMs were prepared through surface coating and chemical cross-linking methods (as shown in Fig. 1b), which were similar to that reported in our previous work [33]. In detail, a mixed aqueous solution of SCMC, cross-linking agent GA and H2SO4 was cast on a PSF ultrafiltration supporting membrane and then the excess casting solution was drained off the membrane surface after being steeped for 3 min. The SNFMs were obtained by cross-linking the \OH in SCMC and the \CHO in GA at 50 °C for 3 h. The obtained SNFMs were thoroughly washed with deionized water and stored in deionized water before NF tests.

2.1. Materials 2.3. Characterization Carboxymethyl cellulose (CMC) (Mw = 700,000 g mol−1) with degree of substitution (D.S.) values 0.90 was purchased from Aldrich. Sulfur trioxide pyridine complex (SO3/pyridine) and p-toluenesulfonic acid (p-TsOH) were obtained from Aladdin. Glutaraldehyde (GA), 25 wt.% aqueous solution, was purchased from Sinopharm Chemical Reagent Co. Ltd., Shanghai, China. All organic solvents (analytical grade) including ethanol, acetone and N,N-dimethylacetamide (DMA) were obtained from Sinopharm Chemical Reagent Co. Ltd., Shanghai, China, and were used as obtained. Inorganic salts including K2SO4, Na2SO4, MgSO4, NaCl, MgCl2 and NaOH, HCl (33.6–38.6 wt.%), and H2SO4 (95.0–98.0 wt.%) were all analytical reagents and obtained from Sinopharm Chemical Reagent Co. Ltd., Shanghai, China. Xylenol orange (XO) and methyl blue (MYB) were purchased from Aladdin. Polysulfone ultrafiltration (PSF-UF) supporting membranes were provided by the Development Center of Water Treatment

The elemental contents of SCMC were measured by X-ray photoelectron spectra (XPS, PerkinElmer PHI 5300 ESCA), with Mg/Al Dual Anode Hel/Hell ultraviolet source (400 W, 15 kV, 1253.6 eV). Chemical structure of SNFMs was characterized with a BRUKER VECTOR 22 attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR, Germany). Membrane morphologies of SNFMs were observed with a field emission scanning electron microscope (FESEM, SIRION-100, USA). SNFMs were fractured in liquid nitrogen to get their crosssection structure. Both surface and cross-section membranes were coated with a thin golden layer before all SEM measurements. Dynamic water contact angles of SNFMs were measured by the sessile drop method using a contact angle meter (OCA 20, Data physics Instruments GmbH, Germany). All the membrane samples were vacuum dried at 30 °C for 24 h prior to characterizations.

(a)

(b)

Fig. 1. Schematic diagram for the synthesis of SCMC (a) and the preparation of their NF membranes (b).

L.-L. Shao et al. / Desalination 338 (2014) 74–83

2.4. Zeta potential measurements

Table 1 Composition of SCMC determined by XPS.

Zeta potential was determined by streaming potential measurements with a self-made setup [34]. The zeta potential ζ is calculated from the Helmholtz–Smoluchowski equation: ζ¼

ΔEηκ ΔPε0 εr

ð1Þ

where ΔE is the potential difference, η is the viscosity, κ is the conductivity, ΔP is the pressure difference across channel, ε0 is the permittivity of vacuum, and εr is the relative electric constant of the solvent. The streaming potential of SNFMs was measured at 25 °C and applied pressure difference ranging from 0.1 to 0.5 MPa in 0.01 M KCl aqueous solution. The pH of the KCl aqueous solution was in the range of 6.5 to 4.0, which was adjusted with 0.1 M HCl. A digital multimeter (Model MY60, Hangzhou Huayi Electric Industrial Co. Ltd., China) was used to determine the ΔE value. Detailed description of the instrument and the measurement procedure can be found in the literatures [35]. The data are average values from three samples of each membrane type. 2.5. Nanofiltration performance Membrane permeation tests were conducted by a laboratory scale cross-flow flat membrane module consisting of three small cells connected in parallel [36]. Prior to testing, membranes were prepressurized under 0.7 MPa for 1 h to make sure the membranes were in the steady state. The permeation tests were carried out at 25 °C and 0.6 MPa, and the results of permeation tests were recorded until the flux and conductivity of permeation reached equilibrium values. The concentration of K2SO4, Na2SO4, NaCl, MgCl2 and MgSO4 in the feed solution was 1.0 g L−1, the XO and MYB feed solution was 0.1 g L−1. The feed solution was adjusted by 0.1 M HCl and 0.1 M NaOH. Permeation flux (J) and rejection (R) were calculated according to the following equations:
 V −2 −1 J Lm h ¼ At

Rð%Þ ¼

Cp 1− Cf

ð2Þ

!  100:

ð3Þ

Above, V is the volume of the solution that penetrates through the membrane; A is the effective membrane area (22.4 cm2); t is the operation time. Cp and Cf are the concentrations of permeate and feed solutions, respectively. In this work, inorganic salts concentrations of permeate and feed solutions were determined via a conductivity meter (DDS-11A, Hangzhou Dongxing Instrument Works, China). Dye concentrations of permeate and feed solution were measured using ultraviolet– visible spectrophotometer (UV722, Shanghai) at 433 nm of XO and 603 nm of MYB. The feed pH was determined with a pH meter (PHS-3C, Hangzhou Dongxing Instrument Works, China). All experiments were repeated three times, and the data presented in this study was the average value of these results. 3. Results and discussion

Sample

Molar ratio (sulfate agent: CMC)

S (At. %)

O (At. %)

O:S

Xa

Membrane

SCMC-0 SCMC-1 SCMC-2 SCMC-3 SCMC-3.5

0 1 2 3 3.5

– 0.96 2.38 3.92 4.51

55.24 56.14 57.48 58.77 59.67

– 58.48 24.15 14.99 13.23

– 0.12 0.33 0.58 0.68

SNFM-0 SNFM-1 SNFM-2 SNFM-3 SNFM-3.5

a X was defined as the molar ratio of sulfate to carboxylate groups, and the specific calculation was given in the supplementary information.

different contents of sulfate groups. On the basis of curves (a) and (b) in Fig. 2, the absorption intensity at 1115 cm−1 is highly enhanced in the SNFM-0, which is attributed to the formation of ether bond between CMC and GA [37], and this absorption peak exists in all the prepared SNFMs. The results indicate that the SNFMs are successfully crosslinked by the addition of GA. Absorption bands at 814 cm−1 reveal for all SNFMs, which is attributed to stretching vibration of SO3 [32] and the absorption intensity increases with increasing molar ratio of the sulfated agent to CMC (as shown in curves (c)–(f) in Fig. 2). This indicates that the sulfate groups were indeed introduced into SNFMs in a tunable manner. The absorption intensity at 1732 cm−1 (COOH groups) for all SNFMs is much larger than that at 1635 cm−1 (COONa groups) [38], because the casting solutions are tuned at acidic conditions (pH = 3.0) to promote the cross-linking reactions between GA and SCMC. Moreover, by calculating the absorption intensity ratio of \C\O\C\ and \OH (I1115/I3432), the relative crosslinking degree (RCD) [39] is obtained (Fig. 3), revealing that RCD reduces with increasing molar ratio of the sulfated agent to CMC in the feed; understandably, it is because a part of crosslinkable \OH groups is replaced by newly incorporated SO3 groups. Collectively, these results all support that SCMCs with tailored chemical structure and composition were synthesized, along with the preparation of their composite membrane SNFMs. Fig. 4 provides the surface and cross-section morphologies of SNFMs determined by FESEM. In contrast to the substrate membrane that exhibits a porous morphology (Fig. 4a), the surface of SNFM-3 is dense and smooth, with no appreciable defections being visualized (Fig. 4b). Detailed cross-sectional examination (Fig. 4c and d) shows that the thickness of both SNFM-3 and SNFM-0 ranges from 550 to 600 nm, which is a reasonable thickness of active skin layer of the common NF

C-O-C

-OH

OSO3

-

(a) (b) Transmittance (%)

76

(c) (d) (e) (f)

3.1. Characterizations of SCMC and SNFMs The chemical structure and composition of SCMCs and their nanofiltration membranes (SNFMs) were determined by XPS and ATR-FTIR. From the XPS characterization of SCMCs (Table 1), it is quantified that the sulfate content of SCMC depends on the molar ratio of sulfate agent to CMC, namely, the sulfate groups in SCMC increase with increasing sulfate agent. Fig. 2 gives the ATR-FTIR spectra of uncrosslinked CMC, crosslinked CMC, and crosslinked SNFMs with

3500

3000

2500

2000

1500

1000

Wavenumbers (cm-1) Fig. 2. Transmittance IR spectra of the un-crosslinked CMC membranes and SNFMs. (a) un-crosslinked CMC membranes, (b) SNFM-0, (c) SNFM-1, (d) SNFM-2, (e) SNFM-3 and (f) SNFM-3.5.

L.-L. Shao et al. / Desalination 338 (2014) 74–83

77

100

0.8

75 80

Flux (L m-2 h-1)

RCD (I1115/I3432)

0.4

60

45

40

30

Rejection (%)

60

0.6

0.2 20

15

0.0

SNFM-0

SNFM-1

SNFM-2

SNFM-3

SNFM-3.5

Fig. 3. Relative crosslinking degree (RCD, I1115/I3432) of SNFMs with different contents of sulfate groups.

membranes prepared via the surface coating and chemical cross-linking method [40]. 3.2. NF performance of SNFMs with different contents of sulfate groups NF performance of SNFMs with different contents of sulfate groups were all prepared under the same conditions (SCMC concentration: 1.0 wt.%; GA concentration: 0.3 wt.%; pH = 3.0; curing at 50 °C for 3 h); these membranes were tested with 1.0 g L−1 Na2SO4 aqueous solution (pH = 6.5) at 25 °C and 0.6 MPa (Fig. 5). As shown in Fig. 5, with increasing the content of sulfate groups, i.e., from SNFM0 to SNFM3.5, the flux increases from 17.7 to 53.1 L m−2 h−1; meanwhile, the rejection retains stable above 90% till SNFM-3 and drops thereafter. As such, we consider that SNFM-3 shows the optimum NF performance of Na2SO4 (J = 39.6 L m−2 h−1, R = 92.6%), with the flux which is 2.2 times of the pristine SNFM-0 (J = 17.7 L m−2 h−1, R = 92.6%). The improved flux of the SNFMs is likely attributed to the improved

0

0 SNFM-0

SNFM-1

SNFM-2

SNFM-3

SNFM-3.5

Fig. 5. Effect of the content of sulfate groups on the nanofiltration performance of SNFMs testing with 1.0 g L−1 Na2SO4 aqueous solution (pH = 6.5) at 25 °C and 0.6 MPa.

hydrophilicity and the decreased cross-linking degree, both of which stem from the incorporated sulfate groups. Because the sulfate groups are strong ion groups that possess high affinity with water, SNFMs with higher content of sulfate groups are more hydrophilic (Fig. 6a) [41,42], and bear higher surface charge density (Fig. 6b). In this case, not surprisingly the flux of SNFMs increases with increasing sulfate group contents, whereby the rejection drops at a certain stage owing to the excessive swelling, because the crosslinking degree of SNFMs decreases in the meantime (Fig. 3).

3.3. Effect of the preparing conditions on the NF performance of SNFM-3 In this part, by choosing SNFM-3 as the model membrane, the influences of preparing conditions of SNFMs were studied, including casting solution concentration, cross-linking reagent concentration and casting solution pH value. All the NF testing were carried out with the feed

Fig. 4. FESEM micrographs of (a) PSF surface, (b) SNFM-3 surface, (c) SNFM-0 cross-section, and (d) SNFM-3 cross-section.

L.-L. Shao et al. / Desalination 338 (2014) 74–83

90

SNFM-0 SNFM-1 SNFM-2 SNFM-3 SNFM-3.5

(a) Water Contact Angle (o)

80 70 60 50

75

(a)

Na2SO4 NaCl

60

Flux (L m-2 h-1)

78

40

45

30

30 20 0

15 10

20

30

40

50

60

70

0.4

Time (s) 100

-5.0

(b) -4.5

1.2

1.6

-4.0

-3.5

2.0

(b) Na2SO4

80

Rejection (%)

Zeta potential (mV)

0.8

SCMC-3 concentration (wt.%)

NaCl 60

40 -3.0 20 0.4

-2.5 SNFM-0

SNFM-1

SNFM-2

SNFM-3

SNFM-3.5

Fig. 6. (a) Dynamic water contact angle of SNFMs and (b) zeta potential of SNFMs at pH = 6.5.

solution of 1.0 g · L−1 Na2SO4 aqueous solution (pH = 6.5) and 1.0 g L−1 NaCl aqueous solution (pH = 6.5) at a pressure of 0.6 MPa and temperature of 25 °C. Fig. 7 shows the effect of the SCMC-3 concentration on the NF performance of SNFM-3. With increasing SCMC-3 concentration from 0.5 to 2.0 wt.%, the flux of Na2SO4 decreases from 68.1 to 23.5 L m−2 h−1, while the rejection increases from 74.4% to 96.8%. For NaCl feed solution, the flux and rejection of SNFM-3 follow the same trend. The phenomena can be explained by the negatively charged density of the membrane surface and the active layer cross-linking density. The sulfate groups increase with increasing the casting solution concentration from 0.5 to 2.0 wt.%; meanwhile, the hydroxyl groups also provide more crosslinking points and enhance the cross-linking density [43]. The same trend was also reported in the literatures [26,27]. Considering both the salt rejection and water flux of SNFM-3, the 1.0 wt.% concentration of SCMC-3 in the casting solution was selected for following studies. Moreover, for SNFMs, apparently their separation performance also depends on the effective crosslinking between SCMC and GA, which is influenced by both the GA content (Fig. 8) and acid catalysts for the reaction, i.e., the solution pH (Fig. 9). As seen from Fig. 8, the rejection of both Na2SO4 and NaCl increases and the water flux decreases with increasing GA concentration from 0.05 to 0.3 wt.%. An optimum NF performance of SNFM-3 is achieved at the GA concentration of 0.3 wt.%, where the flux of Na2SO4 is 38.5 L m− 2 h−1, and the rejection of Na2SO4 is 92.4%. With increasing GA concentration in the casting solution, the cross-linking reaction between SCMC-3 and GA is enhanced,

0.8

1.2

1.6

2.0

SCMC-3 concentration (wt.%) Fig. 7. Effects of SCMC-3 concentration on the flux (a) and rejection (b) of SNFM-3 testing with 1.0 g L−1 Na2SO4 and NaCl aqueous solution at 0.6 MPa and 25 °C (membrane preparation conditions: GA = 0.3 wt.%; pH = 3.0; curing temperature = 50 °C; curing time = 3 h).

leading to denser top layer, accompanied by the increase in rejection and decrease in flux. Interestingly, with further increasing GA concentration above 0.3 wt.%, the rejection of SNFM-3 decreases and the water flux increases. This observation correlates well with the previous literature [44], which is interpreted by the membrane pore forming effect caused by excessive GA molecules. Furthermore, with adding more acid into the casting solution, i.e., decreasing solution pH from 5.0 to 2.0, for both Na2SO4 and NaCl feed solutions, the rejection of SNFM-3 increases and flux decreases with the feed pH reduction from 5.0 to 2.0 (Fig. 9). It is because the stronger acidic conditions can accelerate the chemical crosslinking reaction between SCMC-3 and GA [45], which induces a higher crosslinking degree of SNFM-3 and leads to higher salt rejection coupled with lower flux. In view of both the rejection and water flux of SNFM-3, pH = 3.0 is designed as an optimum preparation condition. 3.4. Effect of operating conditions on the NF performance of SNFMs Fig. 10 shows NF performance of SNFM-3 and SNFM-0 for different inorganic salts and dye aqueous solutions. It is seen that the permeate flux of different salts is approximately similar, with the exception of NaCl solution that exhibits a higher flux (Fig. 10a), which might result from a lower osmotic pressure [46]. The rejection of SNFM-3 and SNFM-0 increases in the same sequence: XO N K2SO4 ≈

L.-L. Shao et al. / Desalination 338 (2014) 74–83

150

(a)

Na2SO4

90

79

(a)

Na2SO4

NaCl

NaCl 75

Flux (L m-2 h-1)

Flux (L m-2 h-1)

120

90

60

60

45

30

15

30 0.0

0.1

0.2

0.3

5.0

0.4

4.5

100

(b)

100

3.5

3.0

2.5

Na2SO4 NaCl

60

2.0

(b) Na2SO4

80

40

Rejection (%)

Rejection (%)

80

NaCl 60

40

20 0.0

4.0

Casting solution pH

GA concentration (wt.%)

20 0.1

0.2

0.3

0.4

GA concentration (wt.%)

5.0

4.5

4.0

3.5

3.0

2.5

2.0

Casting solution pH

Fig. 8. Effects of GA concentration on the flux (a) and rejection (b) of SNFM-3 testing with 1.0 g L−1 Na2SO4 and NaCl aqueous solution at 0.6 MPa and 25 °C (membrane preparation conditions: SCMC = 1.0 wt.%; pH = 3.0; curing temperature = 50 °C; curing time = 3 h).

Fig. 9. Effects of casting solution pH on the flux (a) and rejection (b) of SNFM-3 testing with 1.0 g L−1 Na2SO4 and NaCl aqueous solution at 0.6 MPa and 25 °C (membrane preparation conditions: SCMC = 1.0 wt.%; GA = 0.3 wt.%; curing temperature = 50 °C; curing time = 3 h).

Na2SO4 N NaCl N MgSO4 N MgCl2 (Fig. 10b), which is a typical negatively charged NF performance and is in good agreement with previously reported CMCNa/PVDF composite membranes [26]. This rejection sequence can be explained by electrostatic effect and steric effect [47]. Compared with inorganic salts, the XO has a larger molecular size and its rejection is up to 98.9% for SNFM-3, which is much higher than the rejection of K2SO4 or Na2SO4. Another important feature in Fig. 10, the flux of SNFM-3, is higher than that of the pristine SNFM-0 for all the inorganic salts and dye molecules, and the rejection to multivalent ions and dye molecule is maintained at a good level. As for NaCl, the rejection of SNFM-3 is only 25% while SNFM-0 is about 40%, indicating that the SNFM-3 exhibits better ability for separating monovalent ions and multivalent ions. Next, the effect of operating pressure and feed temperature on NF performance of SNFM-0 and SNFM-3 was studied. Water flux of both membranes increases almost linearly with increasing operating pressure, while the salt rejection also increases slightly (Fig. 11). This result is commonly observed in the NF performance [48,49], and in accordance with the Spiegler–Kedem model [50]. Moreover, the water flux of SNMF-3 increases more rapidly than the SNFM-0 with increasing operating pressure. For instance, the increase ratio of water flux to operating pressure for SNFM-3 is 76.7 L m− 2 h− 1 MPa− 1, whereas that for SNFM-0 is only 31.0 L m− 2 h− 1 MPa− 1. This result further confirms that the water permeability of SNFMs is significantly

improved by the introducing sulfate groups, which is attributed to the higher permeability of the sulfate groups than the carboxylate groups [30]. With increasing feed temperature, the water flux of SNFM-3 and SNFM-0 increases dramatically, while the salt rejection is stable (Fig. 12). This phenomenon has been observed in other NF processes, and is likely due to the increase in water molecular mobility with the increasing feed temperature [51]. Moreover, while both the SNFM-3 and SNFM-0 reject more than 95% Na2SO4, SNFM-3 have a higher flux (61.0 L m− 2 h− 1) than SNFM-0 (25.8 L m− 2 h− 1) at 40 °C, thus affording good separation performances in a wide range of feed temperatures. In addition to the operation pressure and temperature, the pH of feed solution is an important factor to NF performance [52]. For SNFM-0 and SNFM-3, both the flux and rejection of XO decrease slightly with decreasing feed pH from 6.5 to 4.0 (Fig. 13), which is in accordance with literature report [53]. This result is due to the fact that the COONa groups on SCMC chains are pH sensitive [20]. Indeed, the water contact angles of both SNFM-0 and SNFM-3 both increase with decreasing the feed pH (Fig. S1a, supporting information), as a result of the protonation of COONa groups. In the same vein, zeta potential of SNFMs also decreases with the decreasing feed pH (Fig. S1b, supporting information), resulting in the concomitant decrease in rejection. In the feed pH ranges from 6.5 to 4.0, SNFM-3 exhibits higher flux than SNFM-0, while the two

80

L.-L. Shao et al. / Desalination 338 (2014) 74–83

60

70

SNFM-0 SNFM-3

(a)

100

60

80

30

50 60 40

SNFM-0 SNFM-3

30

40

Rejection (%)

Flux (L m-2 h-1)

Flux (L m-2 h-1)

45

15 20

20

0

XO

K

2

SO

Na

2

4

100

SO

Na

Cl

10

M

M

O

(b)

20

25

30

35

40

0

Feed temperature (oC)

l

Fig. 12. Effects of feed temperature on NF performance of SNFM-3 and SNFM-0 testing with 1.0 g L−1 Na2SO4 aqueous solution at 25 °C.

SNFM-0 SNFM-3

membranes show similar rejection to XO molecules (N 90%) due to the incorporation of sulfate groups, which make SNFM-3 higher hydrophilicity and surface charge as shown in Fig. S1. Collectively, the high flux coupled with rejection (N90%) of SNFM-3 renders it applicable in the

80

Rejection (%)

15

2

4

4

10

gC

gS

60 35 40

(a)

SNFM-0 SNFM-3

30

0

XO

K

2

SO 4

Na

2

SO

Na

Cl

M

M

gC

gS

O

l 2

4

4

Flux (L m-2 h-1)

20

Fig. 10. NF performance (a) flux and (b) rejection of different organic and inorganic electrolytes by SNFM-0 and SNFM-3 testing with 1.0 g L−1 salt aqueous solution or 0.1 g L−1 dye aqueous solution at 25 °C and 0.6 MPa.

25 20 15 10 5 0

50

pH 6.5

pH 5.0

pH 4.5

(b)

SNFM-0 SNFM-3

80

30

60

20

40

10

SNFM-0 SNFM-3 0.2

0.3

0.4

0.5

0.6

20

0

Operating pressure (MPa) Fig. 11. Effects of operating pressure on NF performance of SNFM-3 and SNFM-0 testing with 1.0 g L−1 Na2SO4 aqueous solution at 25 °C.

80

Rejection (%)

40

Rejection (%)

Flux (L m-2 h-1)

100

0

pH 4.0

100

60

40

20 pH 6.5

pH 5.0

pH 4.5

pH 4.0

Fig. 13. Effects of feed pH on (a) the flux and (b) the rejection of SNFM-0 and SNFM-3 testing with XO aqueous solution (0.1 g L−1) at 25 °C and 0.6 MPa.

L.-L. Shao et al. / Desalination 338 (2014) 74–83

acidic conditions, as exemplified by the treatment of acidic dye solutions.

40

3.5. Further evaluation of the separation performance of SNFMs

32

(a)

SNFM-0 SNFM-3

Flux (L m-2 h-1)

25 20 15 10 5 0 0

10

20

30

24

80

60

16

40

8

20

0

0

20

40

60

80

100

120

140

160

Rejection (%)

Flux Salt rejection Dye rejection

0 180

Operating time (h) Fig. 15. Variations of flux and rejection of salt and dye with filtration time of SNFM-3 in treating aqueous solution containing 0.1 g L−1 MYB and 1.0 g L−1 NaCl at 0.6 MPa and 25 °C at pH 2.5 ± 0.5.

pH = 2.5 ± 0.5 for 7 days and the result was described in Fig. 15. As shown in Fig. 15, SNFM-3 exhibited good operation stability, which was attributed to the chemical cross-linking structure of SNFM-3. After operation of 168 h for 0.1 g L−1 MYB and 1 g L−1 NaCl at pH 2.5, the flux of SNFM-3 was maintained at 16.0 L m−2 h−1. Moreover the rejection of dye and salt were kept at 99.0% and 40.2%, respectively. It should be mentioned that the firstly decreasing flux is a common phenomenon for treating dye wastewater which was attributed to the accumulation of dye molecules on the membrane surface for membranes' processes [52]. Furthermore, Table 2 gives the comparison of SNFM with some commercial NF membranes and cellulose-based membranes. When compared with some commercial NF membranes, SNFM-3 exhibited a moderate water permeability and considerable salt rejection. Moreover, SNFM-3 showed a relatively higher flux without sacrificing the rejection to multivalent ions in comparison with cellulose acetate, regenerate cellulose and carboxymethyl cellulose membranes. It can be confirmed that the incorporation of sulfate groups is a facile method to fabricate NF membranes with improved water permeability.

35 30

100

Flux (L m-2 h-1)

In an effort to further assess the ability of the membrane in removing acid dye bath wastewaters (acidic pH 2.5–7), MYB was chosen as a model molecule as it was deleterious and extensively used for dyeing cotton, cotton based fibers and leather [54]. In Fig. 14, both the SNFM-3 and SNFM-0 were filtrated continuously with 0.1 g L−1 MYB aqueous solutions containing 1.0 g L−1NaCl at pH = 4.0 ± 0.3 with 0.6 MPa and 25 °C operating condition. Within the first five hours, the flux declines rapidly then levels off (Fig. 14a); meanwhile, the salt rejection also levels off in 2 h while the dye rejection remains constant around 99.0% during the whole filtration time (Fig. 14b). Noteworthy, after 40 h continuous filtration, the flux of SNFM-3 (19.2 L m−2 h−1) is much higher than that of SNFM-0 (4.5 L m−2 h−1), indicating that SNFM-3 possesses better long-term performance stability and improved water flux, in particular for treating saline anionic dye aqueous solutions at acidic conditions; this beneficial attribute originates from the incorporation of sulfate groups. To further investigate the chemical stability of SNFM-3 in the treatment of acidic dye wastewater, the SNFM-3 was filtrated with 0.1 g L− 1 MYB aqueous solution containing 1.0 g L− 1 NaCl at

81

40

Operation time (h)

100

100

(b)

Table 2 Comparison of NF performance of SNFM-3 with commercial NF membranes and cellulosebased composite membranes.

80

SNFM-0 SNFM-3

Membrane

60 60 40

40

Salt rejection (%)

Dye rejection (%)

80

20

20

0

0 0

10

20

30

40

Operation time (h) Fig. 14. Variations of (a) flux and (b) rejection of salt and dye with filtration time for SNFM-0 and SNFM-3 in treating aqueous solution containing 0.1 g L−1 methyl blue and 1.0 g L−1 NaCl at 0.6 MPa, 25 °C and pH = 4.0 ± 0.3.

SCMC-3 SCMC-3 NF270 NTR7450 OPMN-P70 Desal-5DL RC0 CA + LSMM CA CMC/PVDF a b c

Flux at 0.6 MPa (L m−2 h−1) 50.4a a

41.77 45.54 b 78 b 30b 45.6b 0.57b 10.53b 6.15b 16.8c

RNa2SO4 (%)

RNaCl (%)

Feed solution

Ref.

92.6

24.9

This work

87.38 96.8

– 55.2

– –

– N50

– – – – 88.5

– – 72.4 – 33.3

1 g L−1 Na2SO4, 1 g L−1 NaCl, 2 g L−1 Na2SO4 2 g L−1 Na2SO4, 2 g L−1 NaCl Pure water Pure water, 1.5 g L−1 NaCl Pure water Pure water 0.2 g L−1 NaCl Pure water 1 g L−1 Na2SO4, 1 g L−1 NaCl

Flux measured with NaCl solution in this study. Calculated by water permeability. Estimate with the Spiegler–Kedem model with NaCl feed solution.

This work [16] [52] [52] [52] [55] [10] [9] [27]

82

L.-L. Shao et al. / Desalination 338 (2014) 74–83

4. Conclusion SCMC was synthesized by sulfating CMC with SO3/pyridine complex, along with the preparation of composite membranes (SNFMs) through a combination of solution casting and GA crosslinking. FTIR and XPS analysis confirmed that the sulfate groups were successfully introduced into the SNFMs in a tunable manner, leading to a beneficial increase in flux without sacrificing rejection of their membranes. In detail, SNFM-3 exhibited the water flux of 39.8 L m−2 h−1, totally 2.2 times of the pristine SNFM-0 (17.7 L m−2 h−1) in the separation of Na2SO4 aqueous solution (1.0 g L−1). When used for the separation of MYB and NaCl solution in acidic conditions (pH = 4.0), the flux of SNFM-3 (19.2 L m−2 h−1) was also higher than that of SNFM-0 (4.5 L m−2 h−1). Moreover, the rejection of SNFM-3 was kept above 90% in separating salt aqueous solutions and nearly 99.0% of MYB, coupled with a long-term stability of both the water flux and rejection. These features of SNFMs were attributed to the incorporated sulfate groups, which improved electrostatic repulsion between membranes and negative ions and enhanced the affinity with water, resulting in the increase of the charge density and hydrophilicity of SNFMs. Nomenclature CMC carboxymethyl cellulose GA glutaraldehyde MYB methyl blue NF nanofiltration J flux R rejection RO reverse osmosis RCD relative cross-linking degree SCMC sulfated carboxymethyl cellulose SNFM-0 membrane prepared from CMC SNFM-1 membrane prepared from SCMC-1 SNFM-2 membrane prepared from SCMC-2 SNFM-3 membrane prepared from SCMC-3 SNFM-3.5 membrane prepared from SCMC-3.5 UF ultrafiltration X molar ratio of sulfate group to carboxylate groups XO xylenol orange

Acknowledgments This research was financially supported by NNSFC (No. 21106126, 51173160), the National High Technology Research and Development Program of China (No. 2012AA03A602), the Fundamental Research Funds for the Central Universities (No. 2013QNA4049), and the National Basic Research Program of China (No. 2009CB623402). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.desal.2014.01.025. References [1] N. Hilal, H. Al-Zoubi, N.A. Darwish, A.W. Mohammad, M. Abu Arabi, A comprehensive review of nanofiltration membranes: treatment, pretreatment, modelling, and atomic force microscopy, Desalination 170 (2004) 281–308. [2] B. Van der Bruggen, J. Schaep, D. Wilms, C. Vandecasteele, Influence of molecular size, polarity and charge on the retention of organic molecules by nanofiltration, J. Membr. Sci. 156 (1999) 29–41. [3] S.P. Sun, K.Y. Wang, N. Peng, T.A. Hatton, T.S. Chung, Novel polyamide–imide/ cellulose acetate dual-layer hollow fiber membranes for nanofiltration, J. Membr. Sci. 363 (2010) 232–242. [4] Y. Zhang, Y.L. Su, J.M. Peng, X.T. Zhao, J.Z. Liu, J.J. Zhao, Z.Y. Jiang, Composite nanofiltration membranes prepared by interfacial polymerization with natural material tannic acid and trimesoyl chloride, J. Membr. Sci. 429 (2013) 235–242.

[5] L. Li, S.B. Zhang, X.S. Zhang, Preparation and characterization of poly (piperazine amide) composite nanofiltration membrane by interfacial polymerization of 3,3′,5,5′-biphenyl tetraacyl chloride and piperazine, J. Membr. Sci. 335 (2009) 133–139. [6] R.L. Han, S.H. Zhang, L.J. Hu, S.S. Guan, X.G. Jian, Preparation and characterization of thermally stable poly (piperazine amide)/PPBES composite nanofiltration membrane, J. Membr. Sci. 370 (2011) 91–96. [7] A.R. Roudman, F.A. DiGiano, Surface energy of experimental and commercial nanofiltration membranes: effects of wetting and natural organic matter fouling, J. Membr. Sci. 175 (2000) 61–73. [8] S. Butylina, S. Luque, M. Nyström, Fractionation of whey-derived peptides using a combination of ultrafiltration and nanofiltration, J. Membr. Sci. 280 (2006) 418–426. [9] N. Ghaemi, S.S. Madaeni, A. Alizadeh, H. Rajabi, P. Daraei, M. Falsafi, Effect of fatty acids on the structure and performance of cellulose acetate nanofiltration membranes in retention of nitroaromatic pesticides, Desalination 301 (2012) 26–41. [10] D. Rana, B. Scheier, R.M. Narbaitz, T. Matsuura, S. Tabe, S.Y. Jasim, K.C. Khulbe, Comparison of cellulose acetate (CA) membrane and novel CA membranes containing surface modifying macromolecules to remove pharmaceutical and personal care product micropollutants from drinking water, J. Membr. Sci. 409–410 (2012) 346–354. [11] Q. Zhao, Q.F. An, Y.L. Ji, J.W. Qian, C.J. Gao, Polyelectrolyte complex membranes for pervaporation, nanofiltration and fuel cell applications, J. Membr. Sci. 379 (2011) 19–45. [12] J.F. Shi, W.Y. Zhang, Y.L. Su, Z.Y. Jiang, Composite polyelectrolyte multilayer membranes for oligosaccharides nanofiltration separation, Carbohydr. Polym. 94 (2013) 106–113. [13] W.X. Feng, L. Shi, R. Wang, Interfacially polymerized composite nanofiltration hollow fiber membranes for low-pressure water softening, J. Membr. Sci. 430 (2013) 129–139. [14] N. Hilal, A. Al-Zoubi, N.A. Darwish, A.W. Mohammad, M. Abu Arabi, A comprehensive review of nanofiltration membranes: treatment, pretreatment, modeling, and atomic force microscopy, Desalination 170 (2004) 281–308. [15] B. Van der Bruggen, Integrated membrane separation progresses for recycling of valuable wastewater streams: nanofiltration, membrane distillation, and membrane crystallizers revisited, Ind. Eng. Chem. Res. 52 (2013) 10335–10341. [16] A. Giacobbo, A.M. Bernardes, M.N. de Pinho, Nanofiltration for the recovery of low molecular weight polysaccharides and polyphenols from winery effluents, Sep. Purif. Technol. 48 (2013) 2524–2530. [17] K.Y. Wang, Y.C. Xiao, T.S. Chung, Chemically modified polybenzimidazole nanofiltration membrane for the separation of electrolytes and cephalexin, Chem. Eng. Sci. 61 (2006) 5807–5817. [18] M.V. Galiana-Aleixandre, J.A. Mendoza-Roca, A. Bes-Piá, Reducing sulfates concentration in the tannery effluent by applying pollution prevention techniques and nanofiltration, J. Clean. Prod. 19 (2011) 91–98. [19] M.F. Abid, M.A. Zablouk, A.M. Abid-Alameer, Experimental study of dye removal from industrial wastewater by membrane technologies of reverse osmosis and nanofiltration, Iran. J. Environ. 9 (2012) 2–9. [20] W.J. Lau, A.F. Ismail, Polymeric nanofiltration membranes for textile dye wastewater treatment: preparation, performance evaluation, transport modeling and fouling control - a review, Desalination 245 (2009) 321–348. [21] E. Kurt, D.Y. Koseoglu-Imer, N. Dizge, S. Chellam, I. Koyuncu, Pilot-scale evaluation of nanofiltration and reverse osmosis for process reuse of segregated textile dyewash wastewater, Desalination 302 (2012) 24–32. [22] L. Shao, X.Q. Cheng, Y. Liu, S. Quan, J. Ma, S.Z. Zhao, K.Y. Wang, Newly developed nanofiltration (NF) composite membranes by interfacial polymerization for Safranin O and Aniline blue removal, J. Membr. Sci. 430 (2013) 96–105. [23] S. Khan, A.K. Ghosh, V. Ramachandhran, J. Bellare, M.S. Hanra, M.K. Trivedi, B.M. Misra, Synthesis and characterization of low molecular weight cut off ultrafiltration membranes from cellulose propionate polymer, Desalination 128 (2000) 57–66. [24] G. Jonsson, C.E. Boesen, Water and solute transport through cellulose acetate reverse osmosis membranes, Desalination 17 (1975) 145–165. [25] J.C. Su, Q. Yang, J.F. Teo, T.S. Chung, Cellulose acetate nanofiltration hollow fiber membranes for forward osmosis processes, J. Membr. Sci. 355 (2010) 36–44. [26] S.C. Yu, Y.P. Zheng, Q. Zhou, S. Shuai, Z.H. Lü, C.J. Gao, Facile modification of polypropylene hollow fiber microfiltration membranes for nanofiltration, Desalination 298 (2012) 49–58. [27] J. Miao, G.H. Chen, L.L. Li, S.X. Dong, Formation and characterization of carboxymethyl cellulose sodium (CMC-Na)/poly (vinylidene fluoride) (PVDF) composite nanofiltration membranes, Sep. Purif. Technol. 42 (2007) 3085–3099. [28] G. Chen, S.H. Li, X.S. Zhang, S.B. Zhang, Novel thin-film composite membranes with improved water flux from sulfonated cardo poly (arylene ether sulfone) bearing pendant amino groups, J. Membr. Sci. 310 (2008) 102–109. [29] S.S. Guan, S.H. Zhang, R.L. Han, B.G. Zhang, X.G. Jian, Preparation and properties of novel sulfonated copoly (phthalazinone biphenyl ether sulfone) composite nanofiltration membrane, Desalination 318 (2013) 56–63. [30] J.E. Hensley, J.D. Way, Synthesis and characterization of perfluorinated carboxylate/sulfonate ionomer membranes for separation and solid electrolyte applications, Chem. Mater. 19 (2007) 4576–4584. [31] G. Capar, L. Yilmaz, U. Yetis, Reclamation of acid dye bath wastewater: effect of pH on nanofiltration performance, J. Membr. Sci. 281 (2006) 560–569. [32] S. Vogt, D. Klemm, T. Heinze, Effective esterification of carboxymethyl cellulose in a new non-aqueous swelling system, Polym. Bull. 36 (1996) 549–555. [33] Y.L. Ji, Q.F. An, Q. Zhao, H.L. Chen, C.J. Gao, Preparation of novel positively charged copolymer membranes for nanofiltration, J. Membr. Sci. 376 (2011) 254–265. [34] Q.F. An, W.D. Sun, Q. Zhao, Y.L. Ji, C.J. Gao, Study on a novel nanofiltration membrane prepared by interfacial polymerization with zwitterionic amine monomers, J. Membr. Sci. 431 (2013) 171–179.

L.-L. Shao et al. / Desalination 338 (2014) 74–83 [35] C. Lettmann, D. Möckel, E. Staude, Permeation and tangential flow streaming potential measurements for electrokinetic characterization of track-etched microfiltration membranes, J. Membr. Sci. 159 (1999) 243–251. [36] Q.F. An, Y.L. Ji, W.S. Hung, K.R. Lee, C.J. Gao, AMOC positron annihilation study of zwitterionic nanofiltration membranes: correlation between fine structure and ultrahigh permeability, Macromolecules 46 (2013) 2228–2234. [37] K.J. Kim, S.B. Lee, N.W. Han, Kinetics of crosslinking reaction of PVA membrane with glutaraldehyde, Korean J. Chem. Eng. 11 (1994) 41–47. [38] L.T. Cuba-Chiem, L. Huynh, J. Ralston, D.A. Beattie, In situ particle film ATR FTIR spectroscopy of carboxymethyl cellulose adsorption on talc: binding mechanism, pH effects, and adsorption kinetics, Langmuir 24 (2008) 8036–8044. [39] Y.X. Liu, M.H. Zhu, Q. Zhao, Q.F. An, J.W. Qian, K.R. Lee, J.Y. Lai, The chemical crosslinking of polyelectrolyte complex colloidal particles and the pervaporation performance of their membranes, J. Membr. Sci. 385–386 (2011) 132–140. [40] R.W. Baker, Membrane Technology and Applications, second ed. John Wiley &Sons Ltd., Chichester, 2004. [41] X.S. Wang, Q.F. An, Q. Zhao, K.R. Lee, J.W. Qian, C.J. Gao, Homogenous polyelectrolyte complex membranes incorporated with strong ion-pairs with high pervaporation performance for dehydration of ethanol, J. Membr. Sci. 435 (2013) 71–79. [42] J.S. Liu, T.W. Xu, X.Z. Han, Y.X. Fu, Synthesis and characterizations of novel zwitterionic hybrid copolymer containing both sulfonic and carboxylic groups via sulfation and zwitterionic process, Eur. Polym. J. 42 (2006) 2755–2764. [43] C.K. Yeom, K.H. Lee, Pervaporation separation of water-acetic acid mixtures through poly (vinyl alcohol) membranes crosslinked with glutaraldehyde, J. Membr. Sci. 109 (1996) 257–265. [44] Y.L. Ji, Q.F. An, Q. Zhao, H.L. Chen, J.W. Qian, C.J. Gao, Fabrication and performance of a new type of charged nanofiltration membrane based on polyelectrolyte complex, J. Membr. Sci. 357 (2010) 80–89. [45] Y.L. Ji, Q. Zhao, Q.F. An, L.L. Shao, K.R. Lee, Z.K. Xu, C.J. Gao, Novel separation membranes based on zwitterionic colloid particles: tunable selectivity and enhanced antifouling property, J. Mater. Chem. A 1 (2013) 12213–12220.

83

[46] X.L. Wang, J.F. Wei, Z. Dai, K.Y. Zhao, H. Zhang, Preparation and characterization of negatively charged hollow fiber nanofiltration membrane by plasma-induced graft polymerization, Desalination 286 (2012) 138–144. [47] D.X. Wang, M. Su, Z.Y. Yu, X.L. Wang, M. Ando, T. Shintani, Separation performance of a nanofiltration membrane influenced by species and concentration of ions, Desalination 175 (2005) 219–225. [48] Y.L. Ji, Q.F. An, Q. Zhao, W.D. Sun, K.R. Lee, H.L. Chen, C.J. Gao, Novel composite nanofiltration membranes containing zwitterions with high permeate flux and improved anti-fouling performance, J. Membr. Sci. 390–391 (2012) 243–253. [49] X.L. Wang, W.N. Wang, D.X. Wang, Experimental investigation on separation performance of nanofiltration membranes for inorganic electrolyte solutions, Desalination 145 (2002) 115–122. [50] K.S. Spiegler, O. Kedem, Thermodynamics of hyperfiltration (reverse osmosis): criteria for efficient membranes, Desalination 1 (1966) 311–326. [51] M.F.A. Goosen, S.S. Sablani, S.S. Al-Maskari, R.H. Al-Belushi, M. Wilf, Effect of feed temperature on permeate flux and mass transfer coefficient in spiral-wound reverse osmosis systems, Desalination 144 (2002) 367–372. [52] M. Manttari, A. Pihlajamaki, M. Nystrom, Effect of pH on hydrophilicity and charge and their effect on the filtration efficiency of NF membranes at different pH, J. Membr. Sci. 280 (2006) 311–320. [53] A. Braghetta, F.A. Digiano, W.P. Ball, Nanofiltration of natural organic matter: pH and ionic strength effects, J. Environ. Eng. 123 (1997) 628–641. [54] C. Sahoo, A.K. Gupta, Optimization of photocatalytic degradation of methyl blue using silver ion doped titanium dioxide by combination of experimental design and response surface approach, J. Hazard. Mater. 215–216 (2012) 302–310. [55] X.P. Xiong, J.J. Duan, W.W. Zou, X.M. He, W. Zheng, A pH-sensitive regenerated cellulose membrane, J. Membr. Sci. 363 (2010) 96–102.