Effects of acyl chloride monomer functionality on the properties of polyamide reverse osmosis (RO) membrane

Journal of Membrane Science 440 (2013) 48–57

Contents lists available at SciVerse ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Effects of acyl chloride monomer functionality on the properties of polyamide reverse osmosis (RO) membrane Tunyu Wang a,b, Lei Dai a,b, Qifeng Zhang a, Ang Li a, Suobo Zhang a,n a b

Key Laboratory of Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China Graduate School of Chinese Academy of Sciences, Beijing 100039, China

art ic l e i nf o

a b s t r a c t

Article history: Received 20 October 2012 Received in revised form 28 March 2013 Accepted 30 March 2013 Available online 6 April 2013

In this work, three novel polyacyl chloride monomers: 2,4,4′,6-biphenyl tetraacyl chloride (BTAC), 2,3′,4,5′,6-biphenyl pentaacyl chloride (BPAC) and 2,2′,4,4′,6,6′-biphenyl hexaacyl chloride (BHAC) have been successfully synthesized. For the purpose of investigating the effects of the polyacyl chloride functionality on the reverse osmosis (RO) membrane properties, the thin film composite (TFC) RO membranes were prepared through interfacial polymerization of trimesoyl chloride (TMC), BTAC, BPAC, and BHAC with m-phenylenediamine (MPDA) respectively. The membrane properties including physicochemical properties and separation performances were evaluated by a combination of attenuated total reflection Fourier transform infrared (ATR FT-IR) spectroscopy, X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), atomic force microscopy (AFM), contact angle (CA), streaming potential measurements as well as cross-flow reverse osmosis tests. The results reveal that the functionality of the acid chloride monomer strongly influences the properties of the RO membrane. As the functionality of the acid chloride monomer increased, the resulting membrane skin layer became more negatively charged, thinner and smoother. However, the change of hydrophilicity did not seem to follow any rule or trend, which could be ascribed to the cooperative effects of surface roughness and the carboxylic acid group content. In addition, all the four membranes exhibited close salt rejection rates according to the RO separation performance tests. However, with the increase of acid chloride functionality the permeate flux of the resulting RO membrane became lower, due to a combination of the increase in the carboxylic acid groups on the membrane surface, lower surface roughness, and lower mobility of the crosslinked polyamide chains. & 2013 Elsevier B.V. All rights reserved.

Keywords: Thin-film composite membrane Polyacyl chloride Reverse osmosis Membrane property

1. Introduction Reverse osmosis (RO) separation is an effective technology, and has been used extensively in many fields such as for the desalination of seawater and brackish water, ultrapure water production and wastewater treatment [1–3]. Nowadays, the RO membrane market is mainly dominated by thin-film-composite (TFC) polyamide membranes containing three layers: a polyester web serving as the structural support (120–150 μm thick), a microporous polysulfone film acting as the supporting mid-layer (about 40 μm), and a selective ultra-thin barrier layer on the upper surface (about 0.2 μm) [4]. The topmost ultra-thin polyamide layer is generally fabricated through interfacial polymerization (IP) of polyfunctional amine and acid chloride monomers at the interface of two immiscible solvents.

n

Corresponding author. Tel.: +86 431 85262118; fax: +86 431 85685653. E-mail address: [email protected] (S. Zhang).

0376-7388/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.memsci.2013.03.066

The most important properties of the TFC membranes are permeability and selectivity, which are basically determined by the physicochemical properties of the upper polyamide layer such as surface roughness, hydrophilicity, charge performance as well as skin layer thickness. Factors affecting these physicochemical properties include: support membrane structure and chemistry [5,6], monomer structures and concentration [7–16], catalysts and other additives in the aqueous solution and/or in the organic solution during the interfacial polymerization [17–22], reaction and curing conditions [23–25], and other post-treatments [26–28]. Among all these factors, the inherent chemistry of the monomers employed in the polymerization has been proven to play a major role as confirmed from various studies over the past decades. Roh and Khare [12] for example, found that the network polyamide synthesized from MPDA and TMC shows higher water flux as well as salt rejection compared to the linear polyamide made from MPDA and isophthloyl chloride. The higher water flux was attributed mainly to the increased hydrophilicity of the network polyamide due to the presence of the pendant carboxylic acid groups. In addition, the influence of the isomeric diamine

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

monomers on membrane performances were also studied [14]. It was revealed that the meta-positioned polyamide (TMC/MPDA) had higher hydrophilicity and greater polymer chain mobility compared with the para-positioned polyamide (TMC/PPDA), hence resulting in higher water flux. Moreover, Zhou et al. found that the incorporation of m-phenylenediamine-5-sulfonic acid (SMPD) into the amine solution of MPDA during the IP process could increase the membrane flux at the same time reducing the surface roughness, due to the lower reactivity of SMPD with TMC arising from the electron withdrawing nature of the sulfonic acid group [15]. Since the interfacial polymerization process is diffusion controlled in the organic layer, developing a new type of organic phase monomer can be seen as an effective strategy towards improving the RO performance [29]. However, as is widely known, a comprehensive and profound understanding of the structure– property relationship of the RO membrane is definitely necessary which will enable the rational design of new membrane materials at the molecular level. In our previous report, RO membranes based on a series of isomeric tetra-functional biphenyl acid chloride monomers (mm-Biphenyl tetraacyl chloride, om-Biphenyl tetraacyl chloride, and op-Biphenyl tetraacyl chloride) were prepared for the first time [8]. Our research showed the obvious differences in separation performance, chemical composition, hydrophilicity and surface morphology among the three RO membranes, which resulted from the different reactivity of the three isomeric acid chlorides with MPDA during the interfacial reaction. Driven by our continuing interests towards the investigation of structure–property relationship between the acid chloride monomer structure and RO membrane properties [7–10], three novel polyacyl chloride monomers: 2,4,4′,6-biphenyl tetraacyl chloride (BTAC), 2,3′,4,5′,6-biphenyl pentaacyl chloride (BPAC) and 2,2′,4,4′,6,6′-biphenyl hexaacyl chloride (BHAC) were designed and successfully synthesized in this work (Scheme 1). The significant distinction in the functionality of the active acid chloride group among the four monomers of TMC, BTAC, BPAC and BHAC could be observed, which increases in the order of TMC oBTAC oBPAC oBHAC, from trifunctional to hexafunctional. Nevertheless, unlike our previous research in which we mainly focused on addressing the correlation between the position of substitution of the acyl chloride group and the corresponding membrane performance [8], the objective of the current study was to investigate the impacts of the polyacyl chloride functionality on the separation performance and physicochemical properties of the resulting RO membranes. Our consequential insights of the specific role of monomer functionality on the RO membrane properties would deepen the understanding of the structure–property relationship between the monomer structure and membrane properties. Furthermore, to the best of our knowledge, there have been no reports of the RO membranes derived from the penta and hexafunctional acid chloride monomers. However, as a complementary data towards the full understanding of the properties of interfacially polymerized TFC membranes based on a series of organic phase reactants, the information about such highly-functional acid chloride based RO membranes can greatly enrich the knowledge for designing new membrane materials with improved performance [30]. Herein, we report the preparation of the RO membranes based on these new polyacyl chlorides. The membrane

Scheme 1. Structures of the novel polyacyl chloride monomers.

49

properties including physicochemical properties and separation performances were characterized by attenuated total reflection Fourier transform infrared (ATR FT-IR) spectroscopy, X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), atomic force microscopy (AFM), contact angle (CA), streaming potential measurements and cross-flow reverse osmosis tests. The influences of the acid chloride monomer functionality on the RO membrane properties were evaluated from the comparative studies of their separation property, chemical composition of the skin layer, surface hydrophilicity, morphology, charge performance and skin layer thickness of the membranes based on the above mentioned four monomers.

2. experimental 2.1. Synthesis of monomers 2.1.1. Materials N,N-dimethylformamide (DMF) was stirred with calcium hydride for 24 h, then distilled under reduced pressure, and stored over 4 Å molecular sieves. Copper bronze was dried at 100 1C under vacuum for 3 h before use. Isopa G (Isoparaffin type hydrocarbon oil) was purchased from Guangdong Jesan Chemical Ltd. TMC (purity 498%) was purchased from Qingdao Ocean Chemical Company. MPDA (purity 499%) was purchased from TianJin Guangfu Fine Chemical Research Institute. Other reagents and solvents were obtained commercially and used without further purification. 2.1.2. Synthesis of 2,4,4′,6-tetramethylbiphenyl A 500 ml three-necked round-bottomed flask equipped with mechanical stirring, nitrogen inlet and outlet was placed with 2-bromo-1,3,5-trimethylbenzene (7.6 ml, 50 mmol), Pd(PPh3)4 (2.9 g, 2.5 mmol), Cs2CO3 (18 g, 55 mmol), dimethoxyethane (200 ml) and water (0.1 ml). The flask was evacuated and filled with nitrogen for three times. The mixture was stirred and gradually warmed to 80 1C. Then, a solution of 4-methylphenylboronic acid (6.8 g, 50.3 mmol) in 50 ml ethanol was added dropwise and the resulting reaction mixture was refluxed for 5 h. After cooling to room temperature, the solution was diluted with ethyl ether, washed with H2O and brine, dried over anhydrous MgSO4 and concentrated under reduced pressure. The concentrate was purified by silica gel column chromatography to give 2,4,4′,6-tetramethylbiphenyl as a waxy white solid. Yield: 65%. 1H NMR (300 MHz, CDCl3): 7.23 (d, 2 H, J¼ 7.9 Hz), 7.03 (d, 2 H, J¼7.9 Hz), 6.91 (s, 2 H), 2.40 (s, 3 H), 2.33 (s, 3 H), 2.01 (s, 6 H). 2.1.3. Synthesis of 2,4,4′,6-biphenyl tetracarboxylic acid To a stirred mixture of 2,4,4′,6-tetramethylbiphenyl (6.5 g, 31 mmol), aqueous NaOH (10 g, 248 mmol, 300 ml H2O) and 100 ml pyridine under reflux solid KMnO4 (78 g, 500 mmol) was added portionswise during 4 h. After additional 4 h heating, the excess of KMnO4 was removed by several drops of formaline, the mixture was filtered and the separated MnO2 was washed with hot water. The filtrate and washings were combined, concentrated to 100 ml and acidified to pH 1 with concentrated HCl. The deposited tetra-acid was isolated by filtration, washed with cold water and dried. Yield: 85%. 1H NMR (300 MHz, DMSO-d6): 8.35 (s, 2 H), 7.93 (d, 2 H, J ¼9.0 Hz), 7.32 (d, 2 H, J ¼9.0 Hz). 2.1.4. Synthesis of 2,4,4′,6-biphenyl tetraacyl chloride (BTAC) The method for the preparation of BTAC was similar to our previous report [7]. Yield 495%. 1H NMR (300 MHz, CDCl3): 8.89 (s, 2 H), 8.23 (d, 2 H, J¼ 8 Hz), 7.40 (d, 2 H, J¼ 8 Hz). 13C NMR (75 MHz, CDCl3): 167.69, 165.80, 165.70, 144.45, 141.34, 137.85,

50

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

135.52, 134.15, 134.12, 131.56, 129.07. Anal. Calcd for C16O4Cl4H6: C, 47.56%, H, 1.50%, Cl, 35.10%.. Found: C, 47.54%, H, 1.55%, Cl, 35.13%. 2.1.5. Synthesis of 2,3′,4,5′,6-pentamethylbiphenyl The 2,3′,4,5′6-pentamethylbiphenyl was prepared by following the literature procedure [31]. Yield: 35%. 1H NMR (300 MHz, CDCl3): 6.98 (s, 1 H), 6.94 (s, 2 H), 6.77 (s, 2 H), 2.36 (s, 6 H), 2.35 (s, 3 H), 2.03 (s, 6 H). 2.1.6. Synthesis of 2,3′,4,5′,6-biphenyl pentacarboxylic acid The method for the preparation of 2,3′,4,5′6- biphenyl pentaboxylic acid was similar to the synthesis of 2,4,4′,6-biphenyl tetracarboxylic acid mentioned above. Yield: 80%. 1H NMR (300 MHz, DMSO-d6): 8.45 (s, 1 H), 8.41 (s, 2 H), 7.93 (s, 2 H). 2.1.7. Synthesis of 2,3′,4,5′,6-biphenyl pentaacyl chloride (BPAC) The method for the preparation of BPAC was similar to our previous report [7]. Yield 495%. 1H NMR (300 MHz, CDCl3): 9.05 (s, 2 H), 8.95 (s, 1 H), 8.22 (s, 2 H). 13C NMR (75 MHz, CDCl3): 166.43, 165.56, 165.52, 142.05, 137.38, 136.58, 135.85, 134.85, 134.71, 133.98. Anal. Calcd for C17O5Cl5H5: C, 43.77%, H, 1.08%, Cl, 38.00%. Found: C, 43.72%, H, 1.05%, Cl, 38.08%. 2.1.8. Synthesis of 2,2′,4,4′,6,6′-biphenyl hexacarboxylic acid The 2,2′,4,4′,6,6′-biphenyl hexaboxylic acid was prepared by following the literature procedure [32]. 1H NMR (300 MHz, DMSOd6): 8.55 (s, 4 H). 2.1.9. Synthesis of 2,2′,4,4′,6,6′-biphenyl hexaacyl chloride (BHAC) The method for the preparation of BHAC was similar to our previous report [7]. Yield 495%. 1H NMR (300 MHz, CDCl3): 9.23 (s, 4 H). 13C NMR (100 MHz, CDCl3): 166.44, 165.68, 144.21, 139.12, 134.97, 134.82. Anal. Calcd for C18O6Cl6H4: C, 40.87%, H, 0.76%, Cl, 40.22%. Found: C, 40.93%, H, 0.72%, Cl, 40.18%. 2.2. Preparation of the composite membrane 2.2.1. Preparation of microporous polysulfone support membrane To fabricate the TFC membrane, a support substrate composed of microporous polysulfone was first prepared according to a procedure reported in our previous paper. In summary, a solution of 16.5% (w/v) polysulfone (Udel P-3500, U.S. Amono Comp), 13.5% (w/v) ethyleneglycol monomethyl ether (EGM), 0.03% (w/v) sodium dodecyl sulfate (SDS) and DMF 69.97% (w/v) was cast onto a glass plate with a polyester non-woven fabric using a 0.29 mm knife gape. The plate was immediately immersed in a water bath at room temperature with a smooth motion. Within 30 s, the polysulfone gelled into a white microporous sheet, and the top face was used as a support surface for the TFC membranes. 2.2.2. Fabrication of thin-film composite membrane The polyamide skin layer of the composite RO membranes was prepared by interfacial polymerization technique in an assembly clean room. First, the aqueous amine solution was poured on top of the polysulfone support membrane and allowed to soak for 3 min and then the excess amine solution was removed. Subsequently, the impregnated polysulfone substrate was covered with an organic phase solution for 20 s. After removal of the excess organic solution, the membrane was heated in an oven at 90 1C to induce further polymerization. Finally, the membrane was rinsed with DI water and stored in a 1% (w/v) NaHSO3 solution. The aqueous solution containing MPDA (2%, w/v), triethyl amine (TEA) (1%, w/v) and SDS (0.05%, w/v) was prepared with pH 10 adjusted by camphor sulfonic acid. It is generally believed that the triethylamine salt of camphor sulfonic acid (CSA) as an

additive in the amine solution could protect the microporous structure of the polysulfone support membrane from shrinkage during the cuing process [1,23]. In addition, SDS was used to improve the absorption of MPDA in the microporous polysulfone support membrane, hence decreasing the possibility that pinhole defects would form during the IP process [18]. For preparing the organic phase solution, owing to the relatively weak solubility of BPAC and BHAC in Isopa G (0.039% (w/v) and 0.042% (w/v), respectively), toluene was introduced as the co-solvent, a method also used by Arthur [33]. The approach of organic phase preparation was as follows: dissolved 0.1 g of BPAC or BHAC in 5 ml toluene, afterword the solution was diluted to 100 ml with Isopa G. Thus the Isopa solution containing 0.1% (w/v) polyacyl chloride and 5% (v/v) toluene as the co-solvent was obtained. In order to eliminate the influences from the organic solvent, the organic phase of TMC and BTAC monomer was also prepared using the same method mentioned above.

2.3. Performance testing The membrane samples were checked carefully under a fluorescent lamp to avoid some obvious defects before test. All tests for RO performance were characterized at 1.55 MPa using 2000 ppm NaCl solution in cross-flow cells. The feed flow rate in the flow cell was fixed at 100 L/h. The temperature of the feed tank, with a 40 L capacity, was kept constant at 2571 1C. The circular membrane samples were placed in the test cell with the active skin layer facing the incoming feed. The effective membrane area (for each cell) was around 19 cm2. The membranes were initially subjected to pure water with pressure of 1.55 MPa for 5 h prior to performing the RO test experiments. The water flux was determined by direct measurement of the permeate flow (l/m2 h). The salt rejection rate was calculated using the following equation: rejection (%) ¼100  [1−(Cp/Cf)], where Cf and Cp were salt concentrations in the feed and permeate, respectively. The salt concentration in permeate and feed solutions was determined by measuring the electrical conductivity of the salt solution. For the minimization of experimental error, the membrane separation properties were measured from at least three samples and the results were averaged.

2.4. Characterization 2.4.1. Monomer characterization The monomers were identified by elemental analysis of C, H, and Cl. 1H NMR spectra were measured at 300 MHz on an AV300 spectrometer and the 13C NMR spectra were measured at 300 MHz on an AV300 spectrometer as well as 400 MHz on an AV400 spectrometer.

2.4.2. Characterization of the membranes The membranes used for the chemical structure and morphology analysis of the skin layer were rinsed with DI water several times. Subsequently, the membranes were dried under vacuum at 40 1C for 24 h. Attenuated total reflectance infrared (ATR-IR) characterization of the TFC membrane surface was carried out with a Bio-Rad digilab Division FST-80 spectrometer. For ATR-IR analysis of the membrane samples, an Irtran crystal at a 451 angle of incidence was employed. Surface chemical characterizations were performed by X-ray photoelectron spectroscopy (XPS) on a Thermo ESCALAB 280 system with Al/Kα (hν ¼1486.6 eV) anode mono X-ray source.

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

The zeta potential of membrane surface was calculated from the measured streaming potential using the Fairbrother–Mastin approach. The contact angles of the active layer were estimated by a Drop Shape Analysis DSA10 (Krüss Gmbh, Germany) at room temperature. At least fifteen water droplets (of approx. 5 mg) were carefully dropped onto each sample. The minimum and maximum equilibrium angles were dropped from consideration. Average membrane contact angles and standard deviations were determined from the remaining data. The solid–liquid interfacial free energy was determined from −ΔGSL ¼γL[1+cos θ/Δ], where θ is the average contact angle and γL (¼72.8 mJ/m2 for pure water at 25 1C) is the liquid surface tension, Δ is the relative surface area. Scanning electron microscopy (SEM) was performed with an XL 30 ESEM FEG from the FEI Company, and magnifications of up to 10,000 were obtained. Atomic force microscopy (AFM) was carried out on SPA300HV with an SPI 3800 controller (Seiko Instruments Industry, Co. Ltd.).

3. Results and discussion 3.1. Synthesis and characterization of monomers BTAC, BPAC and BHAC The synthetic routes for the three novel polyacyl chlorides: BTAC, BPAC and BHAC are outlined in Scheme 2. Monomer BTAC was synthesized through the Suzuki–Miyaura coupling reaction of 2-bromomesitylene and 4-methylphenylboronic acid, oxidation, and final chlorination reaction. Monomer BPAC was synthesized from 3,5-dimethylbenzenediazonium tetrafluoroborate salt and

51

mesitylene via the Gomberg–Bachmann coupling reaction, oxidation, and subsequent chlorination. The BHAC monomer was synthesized through the Ullmann coupling reaction of 2-bromobenzene-1,3,5-tricarboxylate, hydrolysis, followed by a chlorination reaction. The chemical composition and structure of the three monomers were confirmed by elemental analysis; 1H NMR as well as 13C NMR (Fig. 1). 3.2. Characterization of the active layer 3.2.1. ATR-IR spectra and X-ray photoelectron spectroscopy (XPS) analysis The depth of penetration of the reflected IR beam in the ATRFTIR technique used herein was around 1 μm; as a result the radiation could also penetrate into the polysulfone region of the composite membrane. Hence, the ATR-FTIR spectra of the TFC membranes comprise of bands from both the polyamide layer and the polysulfone support, which are displayed in Fig. 2. It can be confirmed from the figure that the interfacial polymerization had occurred since the acyl chloride at 1770 cm−1 was absent and a strong band at 1664 cm−1 (amide I, C¼ O stretching vibrations of amide) was present. Additionally, peaks at 1541 cm−1 (amide II) and 1610 cm−1 could be assigned to the N–H in-plane bending and the polyamide aromatic ring breathing, respectively. The above two peaks also prove the existence of amide functionalities. The band at 1587 cm−1 is one of the characteristic bands for polysulfone. Furthermore, since the intensity of the absorbance bands is proportional to the total amount of the polyamide layer that the IR beam penetrated [20,34], the areas of the four peaks (A1, A3, and A4 for polyamide, A2 for polysulfone) of all the four membranes were calculated to evaluate the polyamide film thickness.

Scheme 2. Synthetic routes for BTAC, BPAC and BHAC.

52

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

Fig. 1. 1H NMR spectra of monomers: BTAC (a), BPAC (b) and BHAC (c).

Fig. 2. ATR-FTIR spectra of the RO membranes: (a) TMC-MPDA, (b) BTAC-MPDA, (c) BPAC-MPDA and (d) BHAC-MPDA.

Table 1 Analysis of peak areas for the RO membranes. Membrane

A1 (1541 cm−1)

A2 (1587 cm−1)

A3 (1610 cm−1)

A4 (1664 cm−1)

TMC-MPDA BTAC-MPDA BPAC-MPDA BHAC-MPDA

5.098 4.079 2.498 0.916

3.130 3.239 3.396 4.411

1.918 1.904 0.836 0.708

10.558 8.562 8.339 4.458

As shown in Table 1, the areas of the three peaks (1664, 1610, and 1541 cm−1) for polyamide decreases in the following order of TMCMPDA 4 BTAC-MPDA 4 BPAC-MPDA 4BHAC-MPDA, whereas the peak area of the 1587 cm−1 band for polysulfone follows the opposite order of TMC-MPDA oBTAC-MPDA oBPAC-MPDA o BHAC-MPDA. The results imply that a thinner polyamide skin layer is formed over the polysulfone support when an acyl chloride with higher functionality is employed to interfacially react with MPDA during the IP process. Thus the penetration depth of the IR beam into the polysulfone increased, resulting in a larger peak area for the polysulfone interlayer. The decrease in the thickness of the resultant polyamide skin layer based on the highly-functional acid chloride monomer can be demonstrated from SEM results. The chemical compositions of the ultra-thin polyamide layer of the four RO membranes were measured using X-ray photoelectron spectroscopy (XPS). The atomic composition information is presented in Fig. 3 and Table 2. According to the results, the O/N atomic ratio for the surface of the polyamide barrier layer decreases in the order of BHAC-MPDA 4BPAC-MPDA 4BTACMPDA 4 TMC-MPDA. Since oxygen atoms exist theoretically in the –COOH (hydrolyzed from the unreacted acyl chloride groups during IP) and the –NHCO– bond functionality whereas nitrogen atoms are present in –NHCO– bond and the –NH2 end groups, the higher O/N ratio of the membrane skin layer derived from BHAC indicates the existence of more free carboxylic acid groups at the BHAC-MPDA membrane surface.

3.2.2. Morphological studies Representative surface morphology and cross-section structures of the four RO membranes are shown in Fig. 4. Obviously, the TMC-MPDA membrane displays a characteristic surface

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

Fig. 3. XPS spectra of the RO membranes: (a) TMC-MPDA, (b) BTAC-MPDA, (c) BPAC-MPDA and (d) BHAC-MPDA.

Table 2 Surface atomic composition of the skin layer and their concentration ratios. Sample

C (%)

N (%)

O (%)

O/C

N/C

O/N

TMC-MPDA BTAC-MPDA BPAC-MPDA BHAC-MPDA

73.47 70.64 70.45 69.92

9.77 9.04 8.04 7.13

16.75 20.32 21.51 22.95

0.228 0.288 0.305 0.328

0.133 0.128 0.114 0.102

1.714 2.248 2.675 3.219

morphology known as the “ridge-and-valley” structure, in which the white parts represent ridges and the black areas correspond to valleys, identical to that in previous reports [35]. However, the surface morphology of BTAC-MPDA, BPAC-MPDA, and BHAC-MPDA are visibly different from that of TMC-MPDA. Especially for the BPAC-MPDA and BHAC-MPDA membranes, tightly packed globules are distinctly observed and the ear-shaped polyamide ridges (white parts) become almost inconspicuous. Furthermore, it seems that with the increased functionality of the acid chloride monomer, the globules at the resultant membrane surface become smaller and pack much more tightly, indicating a relatively smoother skin layer. In addition, according to a visual inspection of the cross-section images, the film thickness decreases in the order of TMC-MPDA4BTAC-MPDA4BPAC-MPDA4BHAC-MPDA. The estimated film thickness based on image analysis is on the order of 261714 nm, 20576 nm, 179710 nm and 14478 nm, for TMC-MPDA, BTACMPDA, BPAC-MPDA and BHAC-MPDA, respectively. As a complement to SEM, atomic force microscopy (AFM) was also used to characterize the surface morphology of the polyamide skin layer. Quantitative information about the surface roughness can be expressed in terms of various roughness parameters such as root means square (RMS) roughness, average roughness (Ra) and relative surface area (Δ) [36]. The RMS roughness is defined as the mean of the root for the deviation from the standard surface to the indicated surface, the average roughness is the average deviation of peaks and valleys from the mean plane, and the relative surface area Δ is the actual surface area divided by the projection area. The high values of the RMS, Ra and low values of Δ correspond to a significant surface roughness. The 10 mm  10 mm three-dimensional AFM images of the four RO membranes are presented in Fig. 5 and the data of roughness analyses is listed in Table 3. According to the AFM results, the order of surface roughness for the four membranes follows the order: TMC-MPDA 4BTAC-MPDA 4BPAC-MPDA 4BHAC-MPDA, this result is consistent with the surface morphology from SEM measurements.

53

3.2.3. Hydrophilicity The surface hydrophilicity of the TFC membranes is measured by the contact angle (CA) between the membrane surface and the air–water interface. The lower contact angle suggests the greater tendency for water to wet the membrane surface hence. For every membrane measurement, at least fifteen contact angles at different locations of the membrane samples were averaged to make the CA information more reliable. Additionally, the surface solid– liquid interfacial free energy −ΔGSL is also employed to suggest the extent of surface hydrophilicity, which is a modified form of the Young–Dupre equation and widely used in the previous reports. A larger value of −ΔGSL suggests a more hydrophilic surface [11,23]. As shown in Table 4, the surface hydrophilicity of the four RO membranes is in the following order: BTAC-MPDA≈BPACMPDA 4 TMC-MPDA 4BHAC-MPDA. Since the wettability of a solid surface is determined not only by the chemical composition but also by the geometrical structure of the surface, the value of the contact angle on a rough surface where the contact angle is below 901 is less than that on a smooth surface with the same material [23]. Hence, the difference in hydrophilicity among the four RO membranes can be attributed to the cooperative effect of surface roughness and the carboxyl group content on the membrane surface. Taking the BHAC-based membrane as an example, on one hand, there is more carboxylic acid groups existing at the surface of the BHAC-MPDA membrane which is advantageous for the hydrogen bonding interaction due to higher affinity between water molecules and the polymer membrane; on the other hand, its relatively smoother surface as compared to the other three membranes increases the resistance towards water wettability due to the lower surface energy. Ultimately, the greater extent of reduction of the surface roughness acts as the leading factor making the BHAC-MPDA membrane skin layer relatively more hydrophobic compared with the other three RO membranes.

3.2.4. Surface charge studies The surface charge properties of the four RO membranes were characterized in terms of the zeta potential calculated from the streaming potential measurements. All streaming potential measurements were conducted in a background electrolyte solution containing 10 mM KCl. Hydrochloric acid and potassium hydroxide were used to adjust the pH by means of manual titration. Fig. 6 illustrates the streaming potential of the composite membranes measured at three different pH values of 3.0, 7.0 and 10.0. Obviously, all the four membranes displayed positive zeta potentials at pH 3.0, due to the protonation of the amine functional groups. In contrast, the deprotonation of the carboxylic acid and amine groups at pH 7.0 makes the membrane surfaces negatively charged. The more negative zeta potential at pH 10.0 can be due to the dissociation of the more carboxylic acid groups on the membrane surface. Additionally, it can be seen clearly from Fig. 6 that the zeta potential decreases in the order of TMCMPDA 4 BTAC-MPDA 4BPAC-MPDA 4BHAC-MPDA. This trend indicates that the surface of the membrane based on acid chloride monomer possessing higher functionality contains more carboxylic acid groups, a result which is in agreement with the XPS analysis. According to Morgan [29], owing to the negligible solubility of acid chlorides in water and the fairly good solubility of amines in the organic phase, amines spontaneously diffuse from the aqueous phase into the organic phase to react with the acid chlorides, thus it is commonly believed that the polycondensation takes place at the organic side of the interface. In the initial stage of the IP process, since the nascent polyamide film is quite thin, the amine monomers can constantly diffuse through the dense film into the organic phase where it reacts with the acid chloride molecules

54

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

Fig. 4. SEM images of the RO membranes: TMC-MPDA (a, b), BTAC-MPDA (c, d), BPAC-MPDA (e, f) and BHAC-MPDA (g, h).

thereby increasing the film thickness. The growing thin polyamide film behaves as a barrier for the diffusion of the amines and when the mass transfer resistance becomes great enough to block the amine transport into the organic phase, the growth of the polyamide layer finally stops, which reflects the self-restraining character of the IP process. However, when an acid chloride monomer with higher functionality is employed to interfacially react with MPDA in the IP process, the degree of cross-linking of the nascent polyamide film

formed on the interface should be much higher, as a result the denser nascent film increases the diffusion resistance of the MPDA into the organic phase, hence leading to the hydrolysis of more acyl chloride groups in the reaction zone, which consequently results in more carboxylic acid groups existing at the membrane surface. Also, since the roughness and thickness of the membrane skin layer are induced by the overlap-add of the hard polyamide chains [11], the lesser diffused diamine in the reaction zone will directly lead to lower polyamide content on the polysulfone

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

55

Fig. 5. AFM images of the RO membranes: (a) TMC-MPDA, (b) BTAC-MPDA, (c) BPAC-MPDA and (d) BHAC-MPDA.

Table 3 Data of surface roughness of the RO membranes obtained from AFM. Membrane

RMS (nm)

Ra (nm)

Δ

TMC-MPDA BTAC-MPDA BPAC-MPDA BHAC-MPDA

88.2 84.1 74.6 49.1

70.7 68.2 48.7 33.4

1.61 1.58 1.40 1.31

Table 4 Hydrophilicity of the RO membranes. Membrane

Contact angle (1)

−ΔGSL (mJ/m2)

TMC-MPDA BTAC-MPDA BPAC-MPDA BHAC-MPDA

60.9 7 4.58 56.8 7 6.69 57.2 7 4.16 69.0 7 5.71

94.8 98.0 100.9 92.7 Fig. 6. Zeta potential of the RO membranes at various pH values.

support membrane, thereby resulting in the lower thickness and surface roughness of the polyamide active layer derived from acid chloride monomer with higher functionality. 3.3. Permeation properties The separation performances of the four RO membranes are shown in Table 5. It can be seen that there are no significant differences in the salt rejection among the four membranes, which exhibit high average salt rejections (98.8–99.1%) when tested with 2000 mg/L NaCl solution at 1.55 MPa and 25 71 1C. It has been widely accepted that the thickness of the polyamide skin layer in TFC membranes plays an important role in determining the water flux; the thicker film increases the water transfer resistance, hence

Table 5 Separation performances of the RO membranes. Membrane

Water flux (L/(m2h))

NaCl rejection (%)

TMC-MPDA BTAC-MPDA BPAC-MPDA BHAC-MPDA

54.1 74.8 43.3 78.1 31.2 70.8 22.1 72.1

99.17 0.3 98.8 7 0.4 99.0 7 0.2 99.17 0.2

Test conditions: 1.55 MPa, 25 71 1C, 2000 mg/L NaCl solution.

resulting in lower water permeability. However, in this study lower permeate flux with the order of TMC-MPDA 4BTACMPDA 4 BPAC-MPDA 4BHAC-MPDA is observed, suggesting that

56

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57

the polyamide film thickness seems not to behave as a leading factor in deciding the membrane flux. The above change in the permeate flux can be ascribed to three aspects. Firstly, based on the description of the water diffusion process in RO membranes by Reid and Breton [37], the water molecules originally hydrogen bond with both the carboxylic acid groups and amide functionality at the membrane surface. The bound water passes through the membrane from one hydrogen-bonding site to another site in the membrane matrix, and the passage of a solute component that does not form a hydrogen bond is resisted. However, carboxylic acid groups could form a relatively more stable and strong hydrogen-bond with water than hydrophilic amide groups [19]. The higher stability of the hydrogen-bond for carboxylic acid groups with water should increase the resistance for transport of the bound water from one site to another during the water absorption–desorption process [38]. In our case, the lower water flux for membranes based on acid chloride monomer with higher functionality could be attributed to the higher abundance of carboxylic acid groups on the membrane surface, which retards the diffusion of water from the film surface into the polyamide matrix. Secondly, it is well known that the mobility of the polyamide chains play an important role in controlling the RO permeability [39]. Recognizing that the crosslinks in the polymer severely restrict the segmental mobility of the chains and the polymers with higher extent of crosslinking are inherently less flexible [14], the chain mobility of polyamide in the dense inner barrier layer of the four RO membranes is expected to be in the order: TMC-MPDA4BTAC-MPDA4BPAC-MPDA4BHAC-MPDA. Lower chain mobility should increase the resistance towards the passage of the diffused-in water molecule in the polyamide matrix, accordingly resulting in lower water flux. Finally, Hirose and his colleagues have investigated the relationship between the skin layer surface structure of the polyamide RO membranes and their performances. They found that the membranes having rougher skin layer exhibited higher flux, owing to the enlargement of the effective contact area [17]. Nevertheless, some subsequent reports concluded that the surface morphology of the RO membrane was not intrinsically related to water permeability [23,36]. However, in this study we found that the water flux is positively correlated with the surface roughness, which is consistent with Hirose's observation.

4. Conclusions Three novel polyacyl chloride monomers: 2,4,4′,6-biphenyl tetraacyl chloride (BTAC), 2,3′,4,5′,6-biphenyl pentaacyl chloride (BPAC) and 2,2′,4,4′,6,6′-biphenyl hexaacyl chloride (BHAC) were successfully synthesized. TFC reverse osmosis membranes were prepared by using TMC, BTAC, BPAC as well as BHAC to interfacially react with MPDA on the polysulfone support through interfacial polymerization respectively, for the purpose of investigating the effects of the polyacyl chloride functionality on the RO membrane properties. Characterizations of the membrane physicochemical properties including chemical composition, surface roughness, hydrophilicity, charge performance and skin layer thickness demonstrated that the RO membrane skin layer based on acid chloride monomer with higher functionality became more negatively charged, thinner and smoother, due to the higher crosslinking degree of the nascent polyamide film with heavier resistance of amine diffusion into organic phase during the IP process. However, the change in hydrophilicity did not seem to follow any rule or trend, which could be ascribed to the cooperative effect of the surface roughness and the carboxylic acid group content. Moreover, it is expected that the low surface roughness and negative charge of the membranes based on penta- and hexa-

functional acid chloride monomers would be beneficial as it can increase the membrane's resistance to fouling, although additional work is needed to verify this point [40]. Finally, despite all of the four membranes exhibiting close salt rejection rates according to the RO separation performance tests, the permeate flux of the more highly functional acid chloride based membrane was fairly lower, owing to the greater extent of carboxylic acid groups on the membrane surface, lower surface roughness, and lower mobility of the crosslinked polyamide chains.

Acknowledgments We thank the National Basic Research Program of China (No. 2009CB623401), the National Science Foundation of China (Nos. 51021003 and 51133008), and the National High Technology Research and Development Program of China (No. 2012AA03A601) for the financial support.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.memsci.2013.03. 066.

References [1] R.J. Petersen, Composite reverse osmosis and nanofiltration membranes, J. Membr. Sci. 83 (1993) 81–150. [2] M.E. Williams, D. Bhattacharya, R.J. Ray, S.B. McCray, Selected applications reverse osmosis, in: W.S. Winston Ho, K.K. Sirkar (Eds.), Membrane Handbook, Van Nonstrand Reinhold, New York, 1992. [3] J.E. Cadotte, R.J. Petersen, R.E. Larson, E.E. Erickson, A new thin film composite sea water reverse osmosis membrane, Desalination 32 (1980) 25–31. [4] K.P. Lee, T.C. Arnot, D. Mattia, A review of reverse osmosis membrane materials for desalination-development to date and future potential, J. Membr. Sci. 370 (2011) 1–22. [5] A.K. Ghosh, B.H. Jeong, X.F. Huang, E.M.V. Hoek, Impacts of support membrane structure and chemistry on polyamide–polysulfone interfacial composite membranes, J. Membr. Sci. 336 (1993) 140–148. [6] P.S. Singh, S.V. Joshi, J.J. Trivedi, C.V. Devmurari, A.P. Rao, P.K. Ghosh, Probing the structure 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. [7] L. Li, S.B. Zhang, X.S. Zhang, G.D. Zheng, Polyamide thin film composite membranes prepared from 3,4′,5-biphenyl triacylchloride, 3,3′,5,5′-biphenyl tetraacyl chloride and m-phenylenediamine, J. Membr. Sci. 289 (2007) 258–267. [8] L. Li, S.D. Zhang, X.S. Zhang, G.D. Zheng, Polyamide thin film composite membranes prepared from isomeric biphenyl tetraacyl chloride and mphenylenediamine, J. Membr. Sci. 315 (2008) 20–27. [9] 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. [10] H.F. Wang, L. Lei, X.S. Zhang, S.B. Zhang, Polyamide thin-film composite membranesprepared from a novel triamine: 3,5-diamino-N-(4-aminophenyl)-benzamide monomer and m-phenylenediamine, J. Membr. Sci. 353 (2010) 78–84. [11] M.H. Liu, D.H. Wu, S.C. Yu, C.J. Gao, Influence of the polyacyl chloride structure on the reverse osmosis performance, surface properties and chlorine stability of the thin-film composite polyamide membranes, J. Membr. Sci. 326 (2009) 205–214. [12] I.J. Roh, V.P. Khare, Investigation of the specific role of chemical structure on the material and permeation properties of ultrathin aromatic polyamides, J. Mater. Chem. 12 (2002) 2334–2338. [13] C.K. Kim, J.H Kim, I.J. Roh, I.J. Kim, The changes of membrane performance with polyamide molecular structure in the reverse osmosis process, J. Membr. Sci. 165 (2000) 189–199. [14] I.J. Roh, Effect of the physicochemical properties on the permeation performance in fully aromatic crosslinked polyamide thin films, J. Appl. Polym. Sci. 87 (2003) 569–576. [15] Y. Zhou, S.H. Yu, M.H. Liu, C.J. Gao, Polyamide thin film composite membrane prepared from m-phenylenediamine and m-phenylenediamine-5-sulfonic acid, J. Membr. Sci. 270 (2006) 162–168.

T. Wang et al. / Journal of Membrane Science 440 (2013) 48–57 [16] A.L. Ahmad, B.S. Ooi, Properties-performance of thin film composites membrane: study on trimesoyl chloride content and polymerization time, J. Membr. Sci. 255 (2005) 67–77. [17] M. Hirose, H. Ito, Y. Kamiyama, Effect of skin layer surface structures on the flux behavior of RO membranes, J. Membr. Sci. 121 (1996) 209–215. [18] M.A. Kuehne, R.Q. Song, N.N. Li, R.J. Petersen, Flux enhancement in TFC RO membranes, Environ. Program 20 (2001) 23–26. [19] S.Y. Kwak, S.G. Jung, S.H. Kim, Structure–motion–performance of fluxenhanced reverse osmosis (RO) membranes composed of aromatic polyamide thin film, Environ. Sci. Technol. 35 (2001) 4334–4340. [20] M.R. Duan, Z. Wang, J. Xu, J.X. Wang, S.C. Wang, Influence of hexamethyl phosphoramide on polyamide composite reverse osmosis membrane performance, Sep. Purif. Technol. 75 (2010) 145–155. [21] C. Kong., M. Kanezashi, T. Yamomoto, T. Shintani, T. Tsuru, Controlled synthesis of high performance polyamide membrane with thin dense layer for water desalination, J. Membr. Sci. 362 (2010) 76–80. [22] C. Kong, T. Shintani, T. Kamada, V. Freger, T. Tsuru, Co-solvent-mediated synthesis of thin polyamide membranes, J. Membr. Sci. 384 (2011) 10–16. [23] A.k. Ghosh, B.H. Jeong, X.F. Huang, E.M.V. Hoek, Impacts of reaction and curing conditions on polyamide composite reverse osmosis membrane properties, J. Membr. Sci. 311 (2008) 34–45. [24] Y. Jin, Z.H. Su, Effects of polymerization conditions on hydrophilic groups in aromatic polyamide thin films, J. Membr. Sci. 330 (2009) 175–179. [25] W. Xie, G.M. Geise, B.D. Freeman, H.S. Lee, G. Byun, J.E. McGrath, Polyamide interfacial composite membranes prepared from m-phenylenediamine, trimesoyl chloride and a new disulfonated diamine, J. Membr. Sci. 403–404 (2012) 152–161. [26] D. Mukherjee, A. Kulkarni, W.N. Gill, Flux enhancement of reverse osmosis membranes by chemical surface modification, J. Membr. Sci. 97 (1994) 231–249. [27] A. Kulkarni, D. Mukherjee, W.N. Gill, Flux enhancement by hydrophilization of thin film composite reverse osmosis membranes, J. Membr. Sci. 114 (1996) 39–50.

57

[28] D. Mukherjee, A. Kulkarni, W.N. Gill, Chemical treatment for improved performance of reverse osmosis membranes, Desalination 104 (1996) 239–249. [29] P.W. Morgan, Condensation Polymers: By Interfacial and Solution Methods, Interscience, New York20–30. [30] 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.: Part B: Polym. Phys. 48 (2010) 1685–1718. [31] G.R. Beadle, S.H. Korzeniowski, D.E. Rosenberg, B.J. Garcia, G.W. Gokel, Phasetransfer-catalyzed Gomberg–Bachmann synthesis of unsymmetrical biarenes: a survey of catalysts and substrates, J. Org. Chem. 49 (1984) 1594–1603. [32] P. Holy, J. Zavada, J. Zezula, I. Cisarova, J. Podlaha, Formation of infinite grids with chiral square compartments by self-assembly of achiral 1,1′-biphenyl2,2′,6,6′-tetracarboxylic acids: effect of 4,4′-substitution on the grid stacking, Collect. Czech. Chem. Commun. 66 (2001) 820–833. [33] S.D. Arthur, Structure–property relationship in a thin film composite reverse osmosis membrane, J. Membr. Sci. 46 (1989) 243–260. [34] Y. Jin, Z.H. Su, Effects of polyerzation conditions on hydrophilic groups in aromatic polyamide thin films, J. Membr. Sci. 330 (2009) 175–179. [35] S.Y. Kwak, S.G. Jung, Y.S. Yoon, D.W. Ihm, Details of surface features in atomatic polyamide reverse osmosis membrane characterized by scanning electron and atomic force microscopy, J. Polym. Sci.: Part B 37 (1999) 1429–1440. [36] S.A. Jeshi, A. Neville, An investigation into the relationship between flux and roughness on RO membranes using scanning probe microscopy, Desalination 189 (2006) 221–228. [37] C.E. Reid, E.J. Breton, Water and ion floe across cellulosic membranes, J. Appl. Polym. Sci. 1 (1959) 133–143. [38] Y. Jin, W. Wang, Z.H. Su, Spectroscopic study on water diffusion in aromatic polyamide thin film, J. Membr. Sci. 379 (2011) 121–130. [39] S.Y. Kwak, Relationship of relaxation property to reverse osmosis permeability in aromatic polyamide thin-film-composite membranes, Polymer 40 (1999) 6361–6368. [40] G.D. Kang, Y.M. Cao, Development of antifouling reverse osmosis membranes for water treatment: a review, Water Res. 46 (2012) 584–600.