- Email: [email protected]
Preparation of porous diffusion dialysis membranes by functionalization of polysulfone for acid recovery
Journal of Membrane Science 524 (2017) 557–564
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Preparation of porous diffusion dialysis membranes by functionalization of polysulfone for acid recovery
MARK
Xiaocheng Lina, Seungju Kima, De Ming Zhub, Ezzatollah Shamsaeia, Tongwen Xuc, Xiya Fangd, ⁎ Huanting Wanga, a
Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia c CAS Key Laboratory of Soft Matter Chemistry, Laboratory of Functional Membrane, School of Chemistry and Material Science, University of Science and Technology of China, Hefei 230026, China d Monash Center for Electron Microscopy, Monash University, Clayton, Victoria 3800, Australia b
A R T I C L E I N F O
A BS T RAC T
Keywords: Polysulfone Anion exchange membrane Diffusion dialysis Acid recovery Ultrafiltration membrane
Asymmetrically porous anion exchange membranes for acid recovery by diffusion dialysis are prepared by onestep functionalization of chloromethylated polysulfone (PSF) ultrafiltration membrane via reaction with N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA) to achieve simultaneous crosslinking and quaternization. The performance of the resulting TMPDA modified PSF (TPSF) membranes can be easily tailored by controlling the reaction time in TMPDA solution and thus the charged density and crosslinking degree. The TPSF membranes exhibit excellent diffusion dialysis performance, especially high acid dialysis coefficient, stemming from the porous matrix and thin selective layer ( < 0.6 µm). The optimized TPSF membrane has an acid dialysis coefficient of 0.065 m h−1 and a separation factor of 34.0, which are 6.6 and 0.84 times greater than the corresponding values of the commercial DF-120 membranes at the same testing condition using a mixture of HCl and FeCl2 solution as feed solution. The membranes reported in this work are promising for scale-up for industrial acid recovery applications.
1. Introduction Anion exchange membrane (AEM) is a positively charged membrane for anion conduction. AEMs have so far attracted growing attention for many applications such as in alkaline fuel cells (AFCs) [1–3], all vanadium flow battery (VRB) [4–6] and electrodialysis (ED) [7–9]. In addition, diffusion dialysis employing an AEM is a promising process for the treatment of acidic wastewater produced in industrial processes [10–14], because of its significant advantages over many other common methods for acid recovery in terms of low cost and environmental friendliness [15–17]. AEM is the core component in determining the process efficiency of diffusion dialysis and the purity of the recovered acid. It is noted that diffusion dialysis is a spontaneous process without external power and its driving force comes from the concentration gradient of the two solutions separated by AEM [16]. Moreover, the cost of acidic wastewater treatment has increased in recent years [18]. So the development of an AEM with high acid permeability to enhance the diffusion dialysis process capacity is of great interest. The dense AEMs, including the homogeneous [16] and pore-filled ⁎
AEMs [19–22], have been the main focus for many years. Although many strategies had been developed to modify dense AEMs for diffusion dialysis, their performance remained unsatisfied given their dense morphology and large thickness [23–33]. Obviously, the thickness of AEMs should be reduced to improve acid dialysis coefficient [22,34]. However, the decrease in the dense membrane thickness results in reduced mechanical strength and acid-salt selectively [34]. Until recently, ultrafiltration membranes have been considered as the membrane matrix to design mechanically robust AEMs with asymmetric microstructure composing of a thin (~1 µm) nanoporous selective layer and a thick (50–150 µm) macroporous supporting layer [35]. The large-pore supporting layer provides sufficient mechanical strength but presents minimal ion transport resistance [36]. Very recently, our group demonstrated that it is an effective way to use porous ultrafiltration membrane as substrate to prepare high-performance AEMs for diffusion dialysis [37–39]. Specifically, brominated poly(phenylene oxide) (BPPO) polymer was used as the starting material to produce ultrafiltration membrane, which was then crosslinked and quaternized to obtain the porous AEMs, via the nucleophilic substitution between reactive –CH2Br groups and amine agents. The
Corresponding author. E-mail address: [email protected] (H. Wang).
http://dx.doi.org/10.1016/j.memsci.2016.11.059 Received 27 August 2016; Received in revised form 9 November 2016; Accepted 15 November 2016 Available online 25 November 2016 0376-7388/ © 2016 Elsevier B.V. All rights reserved.
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
ion transport rate through the membrane was tremendously enhanced because the asymmetrically porous structure with a thin selective layer resulted in low ion permeation resistance; hence, simultaneous enhancements in both high acid permeability and high acid-salt selectivity were realized. It is noted that only BPPO was investigated as the starting material for the membrane fabrication using direct conversion of ultrafiltration membrane. Therefore, it should be interesting to use other polymers to prove the applicability of this strategy, and it is also valuable to further investigate the polymer type to optimize the resulting membrane performance for diffusion dialysis application. Note that diffusion dialysis process has been widely used in many countries [16], but research is still needed to improve the process efficiency and operating cost. As the core component, the AEM mainly is the main contributor to the cost of the diffusion dialyzer. Therefore, the reduction of the cost of membrane fabrication is important for the development of low-cost diffusion dialysis process. BPPO is a less favored polymer for ultrafiltration membrane fabrication, and its price is high because of the limited large-scale production. By contract, polysulfone (PSF) is a commercial polymer with a much lower price, and it has been widely used for the fabrication of ultrafiltration membranes [40–42]. The chloromethylation of PSF is very mature and of low cost for production of chloromethylated PSF (CMPSF) [43,44]. Until now, PSF polymer has not been used for the fabrication of the porous AEMs. Therefore, the scientific and economic values can be expected when BPPO is replaced with PSF as the starting material for the membrane preparation. In our previous work, we found that PSF ultrafiltration membrane [45] had smaller pore size at the top surface than BPPO ultrafiltration membrane [46] at the same condition. This indicates that the higher acid-salt selectivity of the resultant AEMs can be achieved when BPPO is replaced with PSF for the fabrication of porous AEMs. In the present work, chloromethylated PSF ultrafiltration membranes were prepared by simple non-solvent phase immersion, followed by one-step functionalization of simultaneous crosslinking and quaternization by immersing the membranes in the aqueous solution of the bifunctional agent of N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA). The membrane structure and morphology as well as diffusion dialysis related performance were investigated in detail, and finally the comparison between our new membranes and other reported membranes was made.
Scheme 1. Schematic illustration of CMPSF synthesis process.
2.3. Preparation of CMPSF ultrafiltration membrane CMPSF ultrafiltration membrane was prepared via the non-solvent phase inversion method [46]. A 25 wt% CMPSF/NMP solution was firstly formed by dissolving CMPSF polymer in NMP. After ultrasonication to remove the bubbles, the CMPSF/NMP solution was cast onto a glass plate using a Gardco® adjustable micrometer film applicator with a stainless steel blade (Paul N. Gardner Company, Inc. USA), whose gap was set as 250 µm. After immersing the glass plate into a water bath for non-solvent phase inversion, the CMPSF ultrafiltration membrane was obtained. 2.4. Preparation of porous TPSF AEMs As shown in Scheme 2, porous AEMs were prepared similar to our previous work [39]. The pre-formed CMPSF ultrafiltration membrane was simply immersed in 1 mol L−1 TMPDA solution at 60 °C for different times to tailor the membrane microstructure and diffusion dialysis related performance. The final AEMs prepared from CMPSF ultrafiltration membrane treated by TMPDA were named TPSF-X h AEMs, where X h is the immersion time of CMPSF ultrafiltration membrane in TMPDA solution. 2.5. Membrane characterization 2.5.1. NMR, XPS, SEM and TGA The degree of chloromethylation (DC) of CMPSF was determined using a Bruker Avance 400 (9.4 T magnet) nuclear magnetic resonance (NMR) spectrometer operating at 400 MHz. X-ray photoelectron spectroscopy (XPS) analysis was performed with an AXIS Nova spectrometer (Kratos Analytical Inc., Manchester, UK) with a monochromated Al Kα source operating at a power of 150 W (10 kV, 15 mA). The membrane morphologies were investigated using a scanning electron microscope (FEI Magellan Nano SEM 450 FEG SEM, USA). Thermogravimetric analysis (TGA) was conducted for thermal stability investigation using a TG-DTA/DSC, NETZSCH 449 F3 thermal gravity analyzer. The TGA measurement was performed by heating the sample from 30 to 800 °C in flowing argon at a heating rate of 10 °C min−1.
2. Experimental 2.1. Materials Polysulfone (PSF, Mw~35,000), anhydrous ferrous chloride (FeCl2, 98%), chloroform (≥99%), paraformaldehyde (95%), trimethylchlorosilane (≥97%), stannic chloride (SnCl4, 99%), 1-methyl-2-pyrrolidone (NMP, 99.5%) and N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA, ≥99%) were purchased from Sigma-Aldrich (Australia). Hydrochloric acid was purchased from Ajax Finechem Pty Ltd, Australia. Deionized water was used throughout the experiments.
2.5.2. Ion exchange capacity (IEC) The membrane sample was firstly immersed in 0.2 mol L−1 NaOH aqueous solution at 25 °C for 12 h to ensure that all Cl− ions within the membrane were ion-exchanged with OH−. After thoroughly washing
2.2. Synthesis of CMPSF polymer As shown in Scheme 1, chloromethylated polysulfone (CMPSF) was fabricated according to the reported method [47]: 10 g of PSf was added into 500 mL of chloroform in a flask equipped with a reflux condenser to form a homogenous solution under stirring, 6.78 g of paraformaldehyde and 24.6 g of trimethylchlorosilane were added into the PSF solution; afterwards, 1.178 g stannic chloride was added dropwise, the resulting solution was heated at 50 °C for 48 h. The final CMPSF was obtained by pouring the solution into ethanol bath, followed by drying at 60 °C in an oven for 12 h.
Scheme 2. Schematic of the fabrication of porous TPSF AEMs.
558
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
with water, the sample was ion-exchanged again by immersing in 1 mol L−1 NaCl aqueous solution at 25 °C for 12 h. The amount of the released OH− was measured by titration using a freshly prepared HCl solution as titrant and methyl orange as indicator. The IEC with units of mmol of OH- per gram of dry membrane can be calculated as follow:
IEC =
C. V W
(1)
where C and V are the concentration and the consumed volume of HCl solution, respectively; W the dry weight of the membrane. 2.5.3. Water uptake (WU) The membrane samples were firstly immersed in the sealed containers filling with water, the containers were then kept in an oven at 25 °C or 65 °C for 48 h. After that, the membranes were taken out of the container and water on the membrane surface was wiped and then weighted to get the wet weight of the membrane (Wwet). The wet membranes were then dried in an oven at 100 °C overnight to get the dry weight of the membrane (Wdry). The water uptake (WU) can be determined by the following equation:
WU =
Wwet − Wdry Wdry
×100%
Fig. 1. 1H NMR spectra of PSF and CMPSF.
following equation:
RW =
(2)
M At ∆C
3. Results and discussions 3.1. Chloromethylation degree of CMPSF Chloromethylated polysulfone (CMPSF) was prepared by chloromethylation of polysulfone (PSF). This step is important for preparation of TPSF AEMs and determines the amount of active site of CMPSF for quaternization and the ion transport ability of TPSF AEMs. Fig. 1 shows the 1H NMR spectra of PSF and CMPSF. For PSF, the signal with the chemical shift of 6.90–7.86 ppm can be assigned to the multi hydrogens on phenyl groups (HA) while the peak with the chemical shift of 1.72 ppm assigned to the hydrogens on methyl (HB, -CH3) [49]. By contrast, for CMPSF spectrum, a new signal appears at the chemical shift of 4.56 ppm is assigned to the hydrogens of benzyl chloride (HC, – CH2Cl) [49]. This confirms the formation of the –CH2Cl groups in CMPSF. The degree of chloromethylation (DC) of CMPSF can be calculated to be 129% by the relative area ratio of hydrogens with the chemical shift of 4.56–1.72 ppm according to the formula: DC=3×SHC/ SHB, where S is the area of the proton peak.
(3)
where M is the amount of component transported in moles, A the effective area in square meters, t the time in hours, and ΔC the logarithm average concentration between the two chambers in moles per cubic meters and defined by the following equation:
∆C =
C0f − (C tf − Cdt) ln [(C0f − Cdt)/C tf ]
whereC0f
(5)
The diffusion dialysis performance of TPSF AEM after abovementioned treatment for 7 d was also tested for comparison.
2.5.4. Diffusion dialysis (DD) One home-made cell with two chambers was used for diffusion dialysis tests [23]. Prior to each test, AEM sample was immersed in a mixture of HCl (1 mol L−1) and FeCl2 (0.2 mol L−1) aqueous solution for 12 h and then washed thoroughly with water. After that, an AEM sample was fixed to separate two chambers with a coincident hole, whose area was 5.73 cm2. 0.14 L HCl/FeCl2 mixture solution was added into the feed side and 0.14 L water into the permeate side. The solutions in both sides were stirred at the same rate at 25 °C for minimizing the well-known concentration polarization. The diffusion process ran for 45 min under stirring. A certain amount of solution was taken out from the both sides for concentration test. Specifically, a Na2CO3 aqueous solution as titrant and a methyl orange aqueous solution as indicator were used to test the HCl concentration while KMnO4 solution was used solely to analyze the FeCl2 concentration. Dialysis coefficients (U) of the species in the solution can be calculated as follows [48]:
U=
Wi − Wt ×100% Wi
3.2. Solvent resistance The solubility of CMPSF ultrafiltration membrane and TPSF AEMs in the different solvents was evaluated to investigate the solvent resistance as well as the crosslinking effect of TMPDA on the TPSF membranes. TPSF-5 h was selected for the investigation. As shown in Table 1, CMPSF exhibited outstanding solubility in
(4)
C tf
and are the feed concentrations at time 0 and t, respectively, and Cdt the dialysate concentration at time t. The separation factor (S) with respect to one species over another one can be calculated from the ratio of dialysis coefficients (U) of the two species [48].
Table 1 Solubility
2.5.5. Chemical stability Acidic resistance was tested to investigate the chemical stability of the prepared TPSF AEM according to our previous work [37]. It was determined by the difference of dry weight between the initial sample without treatment (Wi) and the samples after immersing in the mixture of HCl/FeCl2 solution (Wt) at 65 °C for different times (1-7 d). The chemical stability of the membrane was evaluated by investigating the residual weight percentage (RW) of the sample according to the
H2O CHCl3 NMP DMF DMAc
a
of CMPSF and TPSF membranes. CMPSF
TPSF-5h
I S S S S
I I I I I
a S: soluble; I: insoluble; CHCl3: chloroform; NMP: N-methylpyrrolidone; DMF: N,Ndimethylformamide; DMAc: N, N-dimethylacetamide.
559
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
observed, which is assigned to the positively charged QA groups [50,51]. In our preliminary experiment, acid permeation of CMPSF ultrafiltration membrane was too low to be measured, this phenomenon was consistent with XPS results, revealing the absence of QA groups in CMPSF membrane for ion transport. 3.4. Membrane morphology The membrane morphologies of CMPSF and TPSF membranes were investigated by SEM, and the images are shown in Fig. 4. All the membranes formed by immersion precipitation in water have an asymmetric structure. Numerous nanopores can be observed on the top surface of CMPSF membrane. By immersion of CMPSF membrane in TMPDA solution, the pore size and density on the membrane top surface simultaneously decrease (such as TPSF-1 h membrane) because of the swelling after quaternization; finally the pores become unobservable for TPS-4 h and TPSF-5 h membraens, due to the full crosslinking. It is noted that CMPSF and TPSF membranes show similar cross-sectional morphology. Specifically, the CMPSF membrane presents an asymmetric structure composed of a thin active layer and a thick supporting layer. No obvious difference in the crosssectional microstructure between TPSF and CMPSF membranes can be observed even after immersed in TMPDA solution for 5 h; in other words, the treatment of TMPDA does not affect the membrane formation notably. It is noted that the thickness of the active layer of TPSF membrane is only about 0.5–0.6 µm. Compared with the reported dense membranes [23–33], whose thickness were tens to hundreds μm, the actual thickness of the membrane was significantly reduced. Therefore the improved acid permeability can be expected due to the greatly reduced resistance. For the membrane bottom surface, the change between the CMPSF membrane and TPSF membranes is similar to that for their top surface. With increasing immersing time in TMPDA, the pores on the bottom surface become smaller and finally become unobservable as a result of the enhanced swelling of TPSF membranes. Also, the surface becomes wrinkled for TPSF-4 h and TPS5 h membrane at the higher immersion time in TMPDA because of the enhanced crosslinking degree and hydrophilicity.
Fig. 2. The XPS survey spectra of CMPSF and TPSF membranes.
polar solvents including chloroform (CHCl3), N-methylpyrrolidone (NMP), N, N-dimethylformamide (DMF) and N, N-dimethylacetamide (DMAc). By contrast, TPSF-5 h was insoluble in the above-mentioned solvents, resulting from the crosslinking effect of TMPDA. Moreover, the good solvent resistance of TPSF-5 h membrane also indicates the chemical stability of the membrane in the practical application. 3.3. Membrane composition The membrane chemical composition was investigated by XPS. The survey spectra of CMPSF and TPSF membrane (top surface) are shown in Fig. 2. Generally, the two main emission peaks at 532 eV and 285 eV in both CMPSF and TPSF survey spectra can be attributed to O1s and C1s, respectively. For CMPSF, the emission peaks are found as follows (eV): 168 (S2p), 200 (Cl2p) and 399 (N1s) [50]. Notably, the peaks assigned to Cl2p can prove the formation of -CH2Cl groups in CMPSF membrane and is consistent with NMR results. For TPSF membrane prepared from CMPSF with an immersing time in TMPDA solution of 5 h, the peak intensity for Cl2p decreases while the peak intensity for N1s increases; this is due to the growth of TMPDA onto the membrane matrix from the conversion of –CH2Cl groups to quaternary ammonium (QA) groups after the reaction between –CH2Cl and TMPDA. The high resolution XPS spectra with respect to the N1s peak are shown in Fig. 3. From PSF to TPSF, a new emission peak at 402.1 eV can be
3.5. Ion exchange capacity The ion exchange capacity (IEC) is important for AEMs because it determines the amount of ion transport site of AEMs. Because TPSF AEMs are prepared by direct immersion of CMPSF ultrafiltration membrane in TMPDA solution, IECs of TPSF membranes can be simply controlled by the immersing time in TMPDA solution. The bare CMPSF membrane has no IEC, as confirmed by XPS results. For TPSF membranes, the IEC values depend on the immersing time; the longer immersing time results in higher IEC. As shown in Fig. 5, IEC values of TPSF-1 h to TPSF-4 h increases from 0.72 to 1.18 mmol g−1. Generally, the IEC values of TPSF membrane are comparable with those of the reported dense AEMs [23–33], and similar to those of our previous porous membranes [37–39]. This is anticipated as the amount of QA groups should increase upon increasing immersion time that directly determines the extent of reaction between –CH2Cl and TMPDA. When the immersing time increases to 5 h, there is no change in IEC for TPSF-5 h, because of the full conversion of –CH2Cl groups to QA groups. 3.6. Water uptake The presence of water molecules in AEM is considered to be essential for facilitated ion transport, and water uptake with respect to the membrane hydrophilicity thus becomes an important parameter of AEM. As shown in Fig. 6, the CMPSF membrane has a water uptake of 172% at 25 °C. By contract, the water uptake value increases from TPSF-1 h to TPSF-4 h with increasing IEC at a given temperature.
Fig. 3. The XPS high-resolution spectra of N1s region of CMPSF and TPSF membranes.
560
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
Fig. 4. SEM images of CMPSF and TPSF membranes.
Fig. 5. Ion exchange capacity (IEC) values of CMPSF and TPSF membranes.
Fig. 6. Water uptake (WU) values of CMPSF and TPSF membranes.
Specifically, the water uptake values of TPSF-1 h and TPSF-4 h membranes are 260% and 310% at 25 °C, respectively. This is due to the enhanced hydrophilicity of QA groups. The water uptake value of TPSF-5 h is similar to that of TPSF-4 h because of the similar IEC. Normally, the membrane swells severely at the higher temperature because of more water adsorption, leading to an increase in free space in the membrane. However, the water uptake values of TPSF membranes at 65 °C are similar to those at 25 °C because the crosslinking effect of TMPDA leads to the improved swelling stability of the final TPSF membranes. It is noted that the mechanical properties of the conventional dense AEMs in the wet state are poor at elevated temperature due to the membrane swelling, thus limiting their
practical application. By contrast, TPSF membranes with the good swelling stability show the good mechanical properties at high temperature.
3.7. Diffusion dialysis performance Fig. 7 shows the DD performance including acid dialysis coefficient (UH+) and separation factor (S) of TPSF membranes at 25 °C. With increasing the immersing time in TMPDA solution, UH+ and separation factor from TPSF-1 h to TPSF-5 h generally increase from 0.042 m h−1 and 24.5–0.065 m h−1 and 34.0, respectively, resulting from the simultaneously increased IEC and crosslinking degree. 561
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
Fig. 7. Acid dialysis coefficients (UH+) and separation factors (S) of TPSF AEMs at 25 °C. Fig. 8. TGA curves of PSF and TPSF membranes.
The diffusion dialysis related performances including IEC, UH+ and separation factor of the commercial DF-120 membrane and other reported AEMs for acid recovery from the mixture of HCl/FeCl2 solution are summarized in Table 2. The commercial DF-120 membrane has a UH+ value of 0.0085 m h−1 and S value of 18.5. Therefore, TPSF-5 h membrane shows 6.6 times increase in UH+ and 84% increase in S. This is a great improvement in diffusion dialysis performance when comparing the commercial DF-120 membrane with TPSF-5 h membrane. In other words, there are 6.6 times increase in acid recovery rate and 84% increase in the final acid purity if our membrane is used to replace the commercial DF-120 membrane. Obviously, TPSF5 h hold promise for further development for commercial acid recovery. Moreover, TPSF-5 h membrane also shows relatively higher UH+ at a given separation factor as compared with other reported dense anion exchange membranes [23–33]. The improvement in the diffusion dialysis performance should be attributed to the asymmetric structure of TPSF membranes composed of a thin selective layer and a porous supporting layer. It is worth mentioning that the supporting layer with abundant macropores would present minimal ion transport resistance due to the extremely high free space [37–39]. Therefore, the ion transport resistance should be mainly from the top layer. Different from the conventional dense membranes with thicknesses of tens to hundreds μm, TPSF membranes have a selective layer thickness of 0.5– 0.6 µm. The greatly reduced thickness will significantly facilitate the ion transport. In addition to the improvement in diffusion dialysis performance, TPSF membranes are also of low cost as compared with the conventional dense membrane, due to the simple fabrication method involving the industry-standard non-solvent phase inversion process and one-step functionalization. In our previous research, similar AEMs were fabricated using BPPO
ultrafiltration membrane as a substrate, their diffusion dialysis performances were also summarized in Table 2. When polyetherimide (PEI) and trimethylamine (TMA) were used as crosslinker and quaternizing agent respectively, the optimal membrane showed UH+ of 0.063 m h−1 and S of 20.0 [37]. When the crosslinker was replaced with butanediamine (BTDA), the optimal membrane showed the corresponding values of 0.063 m h−1 and 30.4 [38], respectively. Note that TPSF-5 h membrane shows higher separation factor at the similar UH+. Moreover, when tetramethylethylenediamine (TEMED) was used as simultaneous crosslinker and quaternization agent, the optimal membranes showed UH+ of only 0.043 mol h−1 [39], TPSF-5 h membrane shows about 50% increase in UH+ due to the decreased thickness of the active layer (from 0.8 to 0.9 µm to 0.5–0.6 µm). The improvement in diffusion dialysis performance indicates the PSF is more suited than BPPO to be a starting material for fabrication of porous AEM membranes for acid recovery.
3.8. Thermal stability The thermal stability of the resultant porous AEMs was investigated by TGA. Fig. 8 shows TGA curves of the selected TPSF-5 h and PSF membrane. For PSF membrane, it was thermally stable even when the temperature is up to 500 °C. For TPSF membrane, the initial weight loss ( < 5%) at 50–120 °C was caused by the water evaporation of the residual water and thus can be ignored. The second weight loss from 150 °C was caused by the degradation of QA groups [52]. It means that no chemical degradation occurs at temperatures lower than 150 °C, indicating the good thermal stability of TPSF membranes.
Table 2 Ion exchange capacity (IEC), acid dialysis coefficient (UH+) and separation factor (S) of the reported membranes at 25 °C using HCl/FeCl2 solution as model acidic waste solution. Membrane
Compactness
IEC (m molg−1)
UH+ (10−3 m h−1)
S
Ref.
TPSF membranes DF-120 commercial membrane Quaternized PPO based hybrid membranes PVA and glycidyltrimethyl ammonium chloride (EPTAC) blending membranes PVA and multi-alkoxy silicon copolymer blending membranes PVA treated with alkoxysilanes membranes Quaternized poly(VBC-co-γ-MPS) membranes Quaternized Bionic multisilicon copolymers Quaternized blending of PPO and PVA membranes Quaternized aromatic amine based hybrid PVA membranes Quaternized PPO blending with PVA and silanol Quaternized blending of PVC and P(DMAM-co-DVB) Imidazolium functionalized hybrid membrane blending with PVA BPPO ultrafiltration membranes crosslinked by PEI and quaternized by TMA BPPO ultrafiltration membranes crosslinked by BTDA and quaternized by TMA BPPO ultrafiltration membranes simultaneously and crosslinked by TEMED
Porous Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Porous Porous Porous
0.82–1.18 1.96 1.7–2.2 0.58–1.15 0.34–0.76 0.52–1.01 0.6–0.9 0.46–1.25 1.22–1.92 0.65–1.12 1.04–1.08 0.34–1.32 1.34–1.86 2.13 1.45–1.01 0.51–1.43
42–65 8.5 5–11 11–18 10–17 8–10 24–43 7.2–74.8 21–49 17.2–25.2 9.5–14.5 12.0–40.0 18.7–48.3 63 62–41 9–43
24.5–34.0 18.5 17.0–32.0 18.5–21.0 24.0–30.1 15.9–21.0 22–26 25.9–42.8 26–39 14–21 45.0–67.5 36–61 12.72–52.5 20.0 30.4–80.4 46.6–73.8
This work – [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [37] [38] [39]
562
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
References [1] G. Merle, M. Wessling, K. Nijmeijer, Anion exchange membranes for alkaline fuel cells: a review, J. Membr. Sci. 377 (2011) 1–35. [2] J.R. Varcoe, R.C. Slade, Prospects for alkaline anion‐exchange membranes in low temperature fuel cells, Fuel cells 5 (2005) 187–200. [3] G. Couture, A. Alaaeddine, F. Boschet, B. Ameduri, Polymeric materials as anionexchange membranes for alkaline fuel cells, Prog. Polym. Sci. 36 (2011) 1521–1557. [4] X. Li, H. Zhang, Z. Mai, H. Zhang, I. Vankelecom, Ion exchange membranes for vanadium redox flow battery (VRB) applications, Environ. Sci. Technol. 4 (2011) 1147–1160. [5] S. Zhang, C. Yin, D. Xing, D. Yang, X. Jian, Preparation of chloromethylated/ quaternized poly (phthalazinone ether ketone) anion exchange membrane materials for vanadium redox flow battery applications, J. Membr. Sci. 363 (2010) 243–249. [6] T. Mohammadi, M.S. Kazacos, Modification of anion-exchange membranes for vanadium redox flow battery applications, J. Power Sources 63 (1996) 179–186. [7] T. Sata, K. Teshima, T. Yamaguchi, Permselectivity between two anions in anion exchange membranes crosslinked with various diamines in electrodialysis, J. Polym. Sci. A: Polym. Chem. 34 (1996) 1475–1482. [8] M. Kariduraganavar, R. Nagarale, A. Kittur, S. Kulkarni, Ion-exchange membranes: preparative methods for electrodialysis and fuel cell applications, Desalination 197 (2006) 225–246. [9] T. Xu, Ion exchange membranes: state of their development and perspective, J. Membr. Sci. 263 (2005) 1–29. [10] L. Rocchetti, F. Vegliò, B. Kopacek, F. Beolchini, Environmental impact assessment of hydrometallurgical processes for metal recovery from WEEE residues using a portable prototype plant, Environ. Sci. Technol. 47 (2013) 1581–1588. [11] S. Sahu, K. Sahu, B. Pandey, Leaching of zinc sulfide concentrate from the ganeshhimal deposit of nepal, Metall. Mater. Trans. B 37 (2006) 541–549. [12] A.Kumari, M.K.Sinha, S.K.Sahu, B.D.Pandey, Investigation of a novel ionic liquid, Cyphos IL 104 for the solvent extraction of mineral acids, Hydrometallurgy.http:// dx.doi.org/10.1016/j.hydromet.2015.09.016 [13] J.G. Dean, F.L. Bosqui, K.H. Lanouette, Removing heavy metals from waste water, Environ. Sci. Technol. 6 (1972) 518–522. [14] H.C. Bramer, Pollution control in the steel industry, Environ. Sci. Technol. 5 (1971) 1004–1008. [15] M. Regel-Rosocka, A review on methods of regeneration of spent pickling solutions from steel processing, J. Hazard. Mater. 177 (2010) 57–69. [16] J. Luo, C. Wu, T. Xu, Y. Wu, Diffusion dialysis-concept, principle and applications, J. Membr. Sci. 366 (2011) 1–16. [17] S.J. Oh, S.-H. Moon, T. Davis, Effects of metal ions on diffusion dialysis of inorganic acids, J. Membr. Sci. 169 (2000) 95–105. [18] M.-S. Kang, K.-S. Yoo, S.-J. Oh, S.-H. Moon, A lumped parameter model to predict hydrochloric acid recovery in diffusion dialysis, J. Membr. Sci. 188 (2001) 61–70. [19] D.M. Stachera, R.F. Childs, Tuning the acid recovery performance of poly(4vinylpyridine)-filled membranes by the introduction of hydrophobic groups, J. Membr. Sci. 187 (2001) 213–225. [20] D.M. Stachera, R.F. Childs, A.M. Mika, J.M. Dickson, Acid recovery using diffusion dialysis with poly(4-vinylpyridine)-filled microporous membranes, J. Membr. Sci. 148 (1998) 119–127. [21] A.K. Pandey, R.F. Childs, M. West, J.N. Lott, B.E. McCarry, J.M. Dickson, Formation of pore‐filled ion‐exchange membranes with in situ crosslinking: Poly (vinylbenzyl ammonium salt)‐filled membranes, J. Polym. Sci. A: Polym. Chem. 39 (2001) 807–820. [22] D.-H. Kim, J.-H. Park, S.-J. Seo, J.-S. Park, S. Jung, Y.S. Kang, J.-H. Choi, M.S. Kang, Development of thin anion-exchange pore-filled membranes for high diffusion dialysis performance, J. Membr. Sci. 447 (2013) 80–86. [23] J. Luo, C. Wu, Y. Wu, T. Xu, Diffusion dialysis of hydrochloride acid at different temperatures using PPO–SiO2 hybrid anion exchange membranes, J. Membr. Sci. 347 (2010) 240–249. [24] C. Cheng, Z. Yang, J. Pan, B. Tong, T. Xu, Facile and cost effective PVA based hybrid membrane fabrication for acid recovery, Sep. Purif. Technol. 136 (2014) 250–257. [25] C. Wu, Y. Wu, J. Luo, T. Xu, Y. Fu, Anion exchange hybrid membranes from PVA and multi-alkoxy silicon copolymer tailored for diffusion dialysis process, J. Membr. Sci. 356 (2010) 96–104. [26] Y. Wu, C. Wu, Y. Li, T. Xu, Y. Fu, PVA–silica anion-exchange hybrid membranes prepared through a copolymer crosslinking agent, J. Membr. Sci. 350 (2010) 322–332. [27] Y. Wu, J. Luo, L. Yao, C. Wu, F. Mao, T. Xu, PVA/SiO2 anion exchange hybrid membranes from multisilicon copolymers with two types of molecular weights, J. Membr. Sci. 399–400 (2012) 16–27. [28] Y. Wu, J. Luo, C. Wu, T. Xu, Y. Fu, Bionic multisilicon copolymers used as novel cross-linking agents for preparing anion exchange hybrid membranes, J. Phys. Chem. B 115 (2011) 6474–6483. [29] Y. Wu, J. Luo, L. Zhao, G. Zhang, C. Wu, T. Xu, QPPO/PVA anion exchange hybrid membranes from double crosslinking agents for acid recovery, J. Membr. Sci. 428 (2013) 95–103. [30] A.N. Mondal, C. Cheng, Z. Yao, J. Pan, M.M. Hossain, M.I. Khan, Z. Yang, L. Wu, T. Xu, Novel quaternized aromatic amine based hybrid PVA membranes for acid recovery, J. Membr. Sci. 490 (2015) 29–37. [31] Y. Wu, M. Jiang, J. Cao, T. Xu, F. Mao, Combination of OH– ions and –OH groups within QPPO/PVA hybrid membranes for acid recovery, Desalination Water Treat. (2015) 1–11.
Fig. 9. The weight percentage of the TPSF-5 h membrane after immersion in aqueous HCl/FeCl2 solution at 65 °C for different periods of times.
3.9. Chemical stability As mentioned above, the weight loss of TPSF membrane in the mixture of HCl/FeCl2 solution after different times was evaluated to investigate the chemical stability of TPSF membrane according to our previous work [38]. As shown in Fig. 9, it can be seen that TPSF-5 h shows good chemical stability as no weight loss occurs even after immersing the TPSF-5 h in the mixture of HCl/FeCl2 aqueous solution at 65 °C for 7 days. Moreover, the diffusion dialysis performance of TPSF-5 h after immersing in the above hot acidic solution for 7 days was also tested, it was found that the treated TPSF-5 h AEM finally showed UH+ of 0.064 m h−1 and S of 36.9, both of which are similar to the initial values. The chemical stability of the membrane in the hot acidic solution demonstrates that our membranes are suitable for practical diffusion dialysis applications.
4. Conclusions The common chloromethylated PSF rather than BPPO was selected as starting material to fabricate the porous AEMs at a lower cost for practical diffusion dialysis application to further explore the recently developed method based on ultrafiltration membrane for the fabrication of the porous AEMs. Specifically, porous TPSF AEMs were prepared by immersing chloromethylated PSF ultrafiltration membrane in TMPDA solution to realize simultaneous crosslinking and quaternization. The resulting membranes show good swelling, thermal and chemical stabilities. The optimal TPSF-5 h membrane shows an acid dialysis coefficient of 0.065 m h−1 and a separation factor of 34.0, which are superior to the common dense AEMs and our previous porous membranes based on BPPO ultrafiltration membranes for diffusion dialysis at the same test condition. Specifically, the optimal TPSF-5 h membrane shows 6.6 times higher acid dialysis coefficient 0.84 time higher separation factor than DF-120 membranes. This work also suggests that polysulfone is a better polymer than BPPO for the fabrication of porous membrane for acid recovery. Given the simple process for the membrane fabrication, TPSF membranes with high performance but low cost are promising for the large-scale, practical application of high-efficiency acid recovery.
Acknowledgement This work was supported by the Australian Research Council (Project no. DP140101591). The authors thank technical support from staff at Monash Centre for Electron Microscopy. 563
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
J. Membr. Sci. 120 (1996) 239–259. [43] R.P. Pinell, G.D. Khune, N.A. Khatri, S.L. Manatt, Concerning the chloromethylation of alkylbenzenes and polystyrenes by chloromethyl methyl ether, Tetrahedron Lett. 25 (1984) 3511–3514. [44] R.C. Fuson, C. McKeever, Chloromethylation of aromatic compounds, Org. React., 1942. [45] E. Shamsaei, Z.-X. Low, X. Lin, Z. Liu, H. Wang, Polysulfone and its quaternary phosphonium derivative composite membranes with high water flux, Ind. Eng. Chem. Res. 54 (2015) 3333–3340. [46] X. Lin, K. Wang, Y. Feng, J.Z. Liu, X. Fang, T. Xu, H. Wang, Composite ultrafiltration membranes from polymer and its quaternary phosphonium-functionalized derivative with enhanced water flux, J. Membr. Sci. 482 (2015) 67–75. [47] S. Gu, R. Cai, T. Luo, Z. Chen, M. Sun, Y. Liu, G. He, Y. Yan, A Soluble and Highly Conductive Ionomer for High-Performance Hydroxide Exchange Membrane Fuel Cells, Angew. Chem. Int. Ed. 48 (2009) 6499–6502. [48] T. Xu, W. Yang, Sulfuric acid recovery from titanium white (pigment) waste liquor using diffusion dialysis with a new series of anion exchange membranes — static runs, J. Membr. Sci. 183 (2001) 193–200. [49] C.G. Arges, V. Ramani, Two-dimensional NMR spectroscopy reveals cationtriggered backbone degradation in polysulfone-based anion exchange membranes, Proc. Natl. Acad. Sci. USA 110 (2013) 2490–2495. [50] J. Chastain, R.C. King, J. Moulder, Handbook of X-ray Photoelectron Spectroscopy: A Reference Book of Standard Spectra for Identification and Interpretation of XPS Data, Physical Electronics Eden Prairie, MN, 1995. [51] M. Maniruzzaman, D.J. Morgan, A.P. Mendham, J. Pang, M.J. Snowden, D. Douroumis, Drug–polymer intermolecular interactions in hot-melt extruded solid dispersions, Int. J. Pharm. 443 (2013) 199–208. [52] W. Xie, Z. Gao, W.-P. Pan, D. Hunter, A. Singh, R. Vaia, Thermal degradation chemistry of alkyl quaternary ammonium montmorillonite, Chem. Mater. 13 (2001) 2979–2990.
[32] C. Cheng, Z. Yang, Y. He, A.N. Mondal, E. Bakangura, T. Xu, Diffusion dialysis membranes with semi-interpenetrating network for acid recovery, J. Membr. Sci. 493 (2015) 645–653. [33] K. Emmanuel, C. Cheng, B. Erigene, A.N. Mondal, M.M. Hossain, M.I. Khan, N.U. Afsar, G. Liang, L. Wu, T. Xu, Imidazolium functionalized anion exchange membrane blended with PVA for acid recovery via diffusion dialysis process, J. Membr. Sci. 497 (2016) 209–215. [34] T.Asawa, T.Gunjima, Acid diffusion-dialysis process utilizing anion exchange membrane of 4–55 μm thickness, in, Google Patents, 1976. [35] B. Van der Bruggen, C. Vandecasteele, T. Van Gestel, W. Doyen, R. Leysen, A review of pressure‐driven membrane processes in wastewater treatment and drinking water production, Environ. Prog. 22 (2003) 46–56. [36] X.-L. Wang, T. Tsuru, S.-i Nakao, S. Kimura, The electrostatic and steric-hindrance model for the transport of charged solutes through nanofiltration membranes, J. Membr. Sci. 135 (1997) 19–32. [37] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, Y. Hu, T. Xu, H. Wang, Porous diffusion dialysis membranes for rapid acid recovery, J. Membr. Sci. 502 (2016) 76–83. [38] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, D. Zhao, T. Xu, Z. Xie, C.D. Easton, H. Wang, Asymmetrically porous anion exchange membranes with an ultrathin selective layer for rapid acid recovery, J. Membr. Sci. 510 (2016) 437–446. [39] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, T. Xu, H. Wang, Fabrication of asymmetrical diffusion dialysis membranes for rapid acid recovery with high purity, J. Mater. Chem. A 3 (2015) 24000–24007. [40] K. Zodrow, L. Brunet, S. Mahendra, D. Li, A. Zhang, Q. Li, P.J. Alvarez, Polysulfone ultrafiltration membranes impregnated with silver nanoparticles show improved biofouling resistance and virus removal, Water Res. 43 (2009) 715–723. [41] T.A. Tweddle, O. Kutowy, W.L. Thayer, S. Sourirajan, Polysulfone ultrafiltration membranes, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983) 320–326. [42] M. Ulbricht, M. Riedel, U. Marx, Novel photochemical surface functionalization of polysulfone ultrafiltration membranes for covalent immobilization of biomolecules,
564
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
Preparation of porous diffusion dialysis membranes by functionalization of polysulfone for acid recovery
MARK
Xiaocheng Lina, Seungju Kima, De Ming Zhub, Ezzatollah Shamsaeia, Tongwen Xuc, Xiya Fangd, ⁎ Huanting Wanga, a
Department of Chemical Engineering, Monash University, Clayton, Victoria 3800, Australia Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia c CAS Key Laboratory of Soft Matter Chemistry, Laboratory of Functional Membrane, School of Chemistry and Material Science, University of Science and Technology of China, Hefei 230026, China d Monash Center for Electron Microscopy, Monash University, Clayton, Victoria 3800, Australia b
A R T I C L E I N F O
A BS T RAC T
Keywords: Polysulfone Anion exchange membrane Diffusion dialysis Acid recovery Ultrafiltration membrane
Asymmetrically porous anion exchange membranes for acid recovery by diffusion dialysis are prepared by onestep functionalization of chloromethylated polysulfone (PSF) ultrafiltration membrane via reaction with N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA) to achieve simultaneous crosslinking and quaternization. The performance of the resulting TMPDA modified PSF (TPSF) membranes can be easily tailored by controlling the reaction time in TMPDA solution and thus the charged density and crosslinking degree. The TPSF membranes exhibit excellent diffusion dialysis performance, especially high acid dialysis coefficient, stemming from the porous matrix and thin selective layer ( < 0.6 µm). The optimized TPSF membrane has an acid dialysis coefficient of 0.065 m h−1 and a separation factor of 34.0, which are 6.6 and 0.84 times greater than the corresponding values of the commercial DF-120 membranes at the same testing condition using a mixture of HCl and FeCl2 solution as feed solution. The membranes reported in this work are promising for scale-up for industrial acid recovery applications.
1. Introduction Anion exchange membrane (AEM) is a positively charged membrane for anion conduction. AEMs have so far attracted growing attention for many applications such as in alkaline fuel cells (AFCs) [1–3], all vanadium flow battery (VRB) [4–6] and electrodialysis (ED) [7–9]. In addition, diffusion dialysis employing an AEM is a promising process for the treatment of acidic wastewater produced in industrial processes [10–14], because of its significant advantages over many other common methods for acid recovery in terms of low cost and environmental friendliness [15–17]. AEM is the core component in determining the process efficiency of diffusion dialysis and the purity of the recovered acid. It is noted that diffusion dialysis is a spontaneous process without external power and its driving force comes from the concentration gradient of the two solutions separated by AEM [16]. Moreover, the cost of acidic wastewater treatment has increased in recent years [18]. So the development of an AEM with high acid permeability to enhance the diffusion dialysis process capacity is of great interest. The dense AEMs, including the homogeneous [16] and pore-filled ⁎
AEMs [19–22], have been the main focus for many years. Although many strategies had been developed to modify dense AEMs for diffusion dialysis, their performance remained unsatisfied given their dense morphology and large thickness [23–33]. Obviously, the thickness of AEMs should be reduced to improve acid dialysis coefficient [22,34]. However, the decrease in the dense membrane thickness results in reduced mechanical strength and acid-salt selectively [34]. Until recently, ultrafiltration membranes have been considered as the membrane matrix to design mechanically robust AEMs with asymmetric microstructure composing of a thin (~1 µm) nanoporous selective layer and a thick (50–150 µm) macroporous supporting layer [35]. The large-pore supporting layer provides sufficient mechanical strength but presents minimal ion transport resistance [36]. Very recently, our group demonstrated that it is an effective way to use porous ultrafiltration membrane as substrate to prepare high-performance AEMs for diffusion dialysis [37–39]. Specifically, brominated poly(phenylene oxide) (BPPO) polymer was used as the starting material to produce ultrafiltration membrane, which was then crosslinked and quaternized to obtain the porous AEMs, via the nucleophilic substitution between reactive –CH2Br groups and amine agents. The
Corresponding author. E-mail address: [email protected] (H. Wang).
http://dx.doi.org/10.1016/j.memsci.2016.11.059 Received 27 August 2016; Received in revised form 9 November 2016; Accepted 15 November 2016 Available online 25 November 2016 0376-7388/ © 2016 Elsevier B.V. All rights reserved.
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
ion transport rate through the membrane was tremendously enhanced because the asymmetrically porous structure with a thin selective layer resulted in low ion permeation resistance; hence, simultaneous enhancements in both high acid permeability and high acid-salt selectivity were realized. It is noted that only BPPO was investigated as the starting material for the membrane fabrication using direct conversion of ultrafiltration membrane. Therefore, it should be interesting to use other polymers to prove the applicability of this strategy, and it is also valuable to further investigate the polymer type to optimize the resulting membrane performance for diffusion dialysis application. Note that diffusion dialysis process has been widely used in many countries [16], but research is still needed to improve the process efficiency and operating cost. As the core component, the AEM mainly is the main contributor to the cost of the diffusion dialyzer. Therefore, the reduction of the cost of membrane fabrication is important for the development of low-cost diffusion dialysis process. BPPO is a less favored polymer for ultrafiltration membrane fabrication, and its price is high because of the limited large-scale production. By contract, polysulfone (PSF) is a commercial polymer with a much lower price, and it has been widely used for the fabrication of ultrafiltration membranes [40–42]. The chloromethylation of PSF is very mature and of low cost for production of chloromethylated PSF (CMPSF) [43,44]. Until now, PSF polymer has not been used for the fabrication of the porous AEMs. Therefore, the scientific and economic values can be expected when BPPO is replaced with PSF as the starting material for the membrane preparation. In our previous work, we found that PSF ultrafiltration membrane [45] had smaller pore size at the top surface than BPPO ultrafiltration membrane [46] at the same condition. This indicates that the higher acid-salt selectivity of the resultant AEMs can be achieved when BPPO is replaced with PSF for the fabrication of porous AEMs. In the present work, chloromethylated PSF ultrafiltration membranes were prepared by simple non-solvent phase immersion, followed by one-step functionalization of simultaneous crosslinking and quaternization by immersing the membranes in the aqueous solution of the bifunctional agent of N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA). The membrane structure and morphology as well as diffusion dialysis related performance were investigated in detail, and finally the comparison between our new membranes and other reported membranes was made.
Scheme 1. Schematic illustration of CMPSF synthesis process.
2.3. Preparation of CMPSF ultrafiltration membrane CMPSF ultrafiltration membrane was prepared via the non-solvent phase inversion method [46]. A 25 wt% CMPSF/NMP solution was firstly formed by dissolving CMPSF polymer in NMP. After ultrasonication to remove the bubbles, the CMPSF/NMP solution was cast onto a glass plate using a Gardco® adjustable micrometer film applicator with a stainless steel blade (Paul N. Gardner Company, Inc. USA), whose gap was set as 250 µm. After immersing the glass plate into a water bath for non-solvent phase inversion, the CMPSF ultrafiltration membrane was obtained. 2.4. Preparation of porous TPSF AEMs As shown in Scheme 2, porous AEMs were prepared similar to our previous work [39]. The pre-formed CMPSF ultrafiltration membrane was simply immersed in 1 mol L−1 TMPDA solution at 60 °C for different times to tailor the membrane microstructure and diffusion dialysis related performance. The final AEMs prepared from CMPSF ultrafiltration membrane treated by TMPDA were named TPSF-X h AEMs, where X h is the immersion time of CMPSF ultrafiltration membrane in TMPDA solution. 2.5. Membrane characterization 2.5.1. NMR, XPS, SEM and TGA The degree of chloromethylation (DC) of CMPSF was determined using a Bruker Avance 400 (9.4 T magnet) nuclear magnetic resonance (NMR) spectrometer operating at 400 MHz. X-ray photoelectron spectroscopy (XPS) analysis was performed with an AXIS Nova spectrometer (Kratos Analytical Inc., Manchester, UK) with a monochromated Al Kα source operating at a power of 150 W (10 kV, 15 mA). The membrane morphologies were investigated using a scanning electron microscope (FEI Magellan Nano SEM 450 FEG SEM, USA). Thermogravimetric analysis (TGA) was conducted for thermal stability investigation using a TG-DTA/DSC, NETZSCH 449 F3 thermal gravity analyzer. The TGA measurement was performed by heating the sample from 30 to 800 °C in flowing argon at a heating rate of 10 °C min−1.
2. Experimental 2.1. Materials Polysulfone (PSF, Mw~35,000), anhydrous ferrous chloride (FeCl2, 98%), chloroform (≥99%), paraformaldehyde (95%), trimethylchlorosilane (≥97%), stannic chloride (SnCl4, 99%), 1-methyl-2-pyrrolidone (NMP, 99.5%) and N,N,N′,N′-tetramethyl-1,3-propanediamine (TMPDA, ≥99%) were purchased from Sigma-Aldrich (Australia). Hydrochloric acid was purchased from Ajax Finechem Pty Ltd, Australia. Deionized water was used throughout the experiments.
2.5.2. Ion exchange capacity (IEC) The membrane sample was firstly immersed in 0.2 mol L−1 NaOH aqueous solution at 25 °C for 12 h to ensure that all Cl− ions within the membrane were ion-exchanged with OH−. After thoroughly washing
2.2. Synthesis of CMPSF polymer As shown in Scheme 1, chloromethylated polysulfone (CMPSF) was fabricated according to the reported method [47]: 10 g of PSf was added into 500 mL of chloroform in a flask equipped with a reflux condenser to form a homogenous solution under stirring, 6.78 g of paraformaldehyde and 24.6 g of trimethylchlorosilane were added into the PSF solution; afterwards, 1.178 g stannic chloride was added dropwise, the resulting solution was heated at 50 °C for 48 h. The final CMPSF was obtained by pouring the solution into ethanol bath, followed by drying at 60 °C in an oven for 12 h.
Scheme 2. Schematic of the fabrication of porous TPSF AEMs.
558
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
with water, the sample was ion-exchanged again by immersing in 1 mol L−1 NaCl aqueous solution at 25 °C for 12 h. The amount of the released OH− was measured by titration using a freshly prepared HCl solution as titrant and methyl orange as indicator. The IEC with units of mmol of OH- per gram of dry membrane can be calculated as follow:
IEC =
C. V W
(1)
where C and V are the concentration and the consumed volume of HCl solution, respectively; W the dry weight of the membrane. 2.5.3. Water uptake (WU) The membrane samples were firstly immersed in the sealed containers filling with water, the containers were then kept in an oven at 25 °C or 65 °C for 48 h. After that, the membranes were taken out of the container and water on the membrane surface was wiped and then weighted to get the wet weight of the membrane (Wwet). The wet membranes were then dried in an oven at 100 °C overnight to get the dry weight of the membrane (Wdry). The water uptake (WU) can be determined by the following equation:
WU =
Wwet − Wdry Wdry
×100%
Fig. 1. 1H NMR spectra of PSF and CMPSF.
following equation:
RW =
(2)
M At ∆C
3. Results and discussions 3.1. Chloromethylation degree of CMPSF Chloromethylated polysulfone (CMPSF) was prepared by chloromethylation of polysulfone (PSF). This step is important for preparation of TPSF AEMs and determines the amount of active site of CMPSF for quaternization and the ion transport ability of TPSF AEMs. Fig. 1 shows the 1H NMR spectra of PSF and CMPSF. For PSF, the signal with the chemical shift of 6.90–7.86 ppm can be assigned to the multi hydrogens on phenyl groups (HA) while the peak with the chemical shift of 1.72 ppm assigned to the hydrogens on methyl (HB, -CH3) [49]. By contrast, for CMPSF spectrum, a new signal appears at the chemical shift of 4.56 ppm is assigned to the hydrogens of benzyl chloride (HC, – CH2Cl) [49]. This confirms the formation of the –CH2Cl groups in CMPSF. The degree of chloromethylation (DC) of CMPSF can be calculated to be 129% by the relative area ratio of hydrogens with the chemical shift of 4.56–1.72 ppm according to the formula: DC=3×SHC/ SHB, where S is the area of the proton peak.
(3)
where M is the amount of component transported in moles, A the effective area in square meters, t the time in hours, and ΔC the logarithm average concentration between the two chambers in moles per cubic meters and defined by the following equation:
∆C =
C0f − (C tf − Cdt) ln [(C0f − Cdt)/C tf ]
whereC0f
(5)
The diffusion dialysis performance of TPSF AEM after abovementioned treatment for 7 d was also tested for comparison.
2.5.4. Diffusion dialysis (DD) One home-made cell with two chambers was used for diffusion dialysis tests [23]. Prior to each test, AEM sample was immersed in a mixture of HCl (1 mol L−1) and FeCl2 (0.2 mol L−1) aqueous solution for 12 h and then washed thoroughly with water. After that, an AEM sample was fixed to separate two chambers with a coincident hole, whose area was 5.73 cm2. 0.14 L HCl/FeCl2 mixture solution was added into the feed side and 0.14 L water into the permeate side. The solutions in both sides were stirred at the same rate at 25 °C for minimizing the well-known concentration polarization. The diffusion process ran for 45 min under stirring. A certain amount of solution was taken out from the both sides for concentration test. Specifically, a Na2CO3 aqueous solution as titrant and a methyl orange aqueous solution as indicator were used to test the HCl concentration while KMnO4 solution was used solely to analyze the FeCl2 concentration. Dialysis coefficients (U) of the species in the solution can be calculated as follows [48]:
U=
Wi − Wt ×100% Wi
3.2. Solvent resistance The solubility of CMPSF ultrafiltration membrane and TPSF AEMs in the different solvents was evaluated to investigate the solvent resistance as well as the crosslinking effect of TMPDA on the TPSF membranes. TPSF-5 h was selected for the investigation. As shown in Table 1, CMPSF exhibited outstanding solubility in
(4)
C tf
and are the feed concentrations at time 0 and t, respectively, and Cdt the dialysate concentration at time t. The separation factor (S) with respect to one species over another one can be calculated from the ratio of dialysis coefficients (U) of the two species [48].
Table 1 Solubility
2.5.5. Chemical stability Acidic resistance was tested to investigate the chemical stability of the prepared TPSF AEM according to our previous work [37]. It was determined by the difference of dry weight between the initial sample without treatment (Wi) and the samples after immersing in the mixture of HCl/FeCl2 solution (Wt) at 65 °C for different times (1-7 d). The chemical stability of the membrane was evaluated by investigating the residual weight percentage (RW) of the sample according to the
H2O CHCl3 NMP DMF DMAc
a
of CMPSF and TPSF membranes. CMPSF
TPSF-5h
I S S S S
I I I I I
a S: soluble; I: insoluble; CHCl3: chloroform; NMP: N-methylpyrrolidone; DMF: N,Ndimethylformamide; DMAc: N, N-dimethylacetamide.
559
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
observed, which is assigned to the positively charged QA groups [50,51]. In our preliminary experiment, acid permeation of CMPSF ultrafiltration membrane was too low to be measured, this phenomenon was consistent with XPS results, revealing the absence of QA groups in CMPSF membrane for ion transport. 3.4. Membrane morphology The membrane morphologies of CMPSF and TPSF membranes were investigated by SEM, and the images are shown in Fig. 4. All the membranes formed by immersion precipitation in water have an asymmetric structure. Numerous nanopores can be observed on the top surface of CMPSF membrane. By immersion of CMPSF membrane in TMPDA solution, the pore size and density on the membrane top surface simultaneously decrease (such as TPSF-1 h membrane) because of the swelling after quaternization; finally the pores become unobservable for TPS-4 h and TPSF-5 h membraens, due to the full crosslinking. It is noted that CMPSF and TPSF membranes show similar cross-sectional morphology. Specifically, the CMPSF membrane presents an asymmetric structure composed of a thin active layer and a thick supporting layer. No obvious difference in the crosssectional microstructure between TPSF and CMPSF membranes can be observed even after immersed in TMPDA solution for 5 h; in other words, the treatment of TMPDA does not affect the membrane formation notably. It is noted that the thickness of the active layer of TPSF membrane is only about 0.5–0.6 µm. Compared with the reported dense membranes [23–33], whose thickness were tens to hundreds μm, the actual thickness of the membrane was significantly reduced. Therefore the improved acid permeability can be expected due to the greatly reduced resistance. For the membrane bottom surface, the change between the CMPSF membrane and TPSF membranes is similar to that for their top surface. With increasing immersing time in TMPDA, the pores on the bottom surface become smaller and finally become unobservable as a result of the enhanced swelling of TPSF membranes. Also, the surface becomes wrinkled for TPSF-4 h and TPS5 h membrane at the higher immersion time in TMPDA because of the enhanced crosslinking degree and hydrophilicity.
Fig. 2. The XPS survey spectra of CMPSF and TPSF membranes.
polar solvents including chloroform (CHCl3), N-methylpyrrolidone (NMP), N, N-dimethylformamide (DMF) and N, N-dimethylacetamide (DMAc). By contrast, TPSF-5 h was insoluble in the above-mentioned solvents, resulting from the crosslinking effect of TMPDA. Moreover, the good solvent resistance of TPSF-5 h membrane also indicates the chemical stability of the membrane in the practical application. 3.3. Membrane composition The membrane chemical composition was investigated by XPS. The survey spectra of CMPSF and TPSF membrane (top surface) are shown in Fig. 2. Generally, the two main emission peaks at 532 eV and 285 eV in both CMPSF and TPSF survey spectra can be attributed to O1s and C1s, respectively. For CMPSF, the emission peaks are found as follows (eV): 168 (S2p), 200 (Cl2p) and 399 (N1s) [50]. Notably, the peaks assigned to Cl2p can prove the formation of -CH2Cl groups in CMPSF membrane and is consistent with NMR results. For TPSF membrane prepared from CMPSF with an immersing time in TMPDA solution of 5 h, the peak intensity for Cl2p decreases while the peak intensity for N1s increases; this is due to the growth of TMPDA onto the membrane matrix from the conversion of –CH2Cl groups to quaternary ammonium (QA) groups after the reaction between –CH2Cl and TMPDA. The high resolution XPS spectra with respect to the N1s peak are shown in Fig. 3. From PSF to TPSF, a new emission peak at 402.1 eV can be
3.5. Ion exchange capacity The ion exchange capacity (IEC) is important for AEMs because it determines the amount of ion transport site of AEMs. Because TPSF AEMs are prepared by direct immersion of CMPSF ultrafiltration membrane in TMPDA solution, IECs of TPSF membranes can be simply controlled by the immersing time in TMPDA solution. The bare CMPSF membrane has no IEC, as confirmed by XPS results. For TPSF membranes, the IEC values depend on the immersing time; the longer immersing time results in higher IEC. As shown in Fig. 5, IEC values of TPSF-1 h to TPSF-4 h increases from 0.72 to 1.18 mmol g−1. Generally, the IEC values of TPSF membrane are comparable with those of the reported dense AEMs [23–33], and similar to those of our previous porous membranes [37–39]. This is anticipated as the amount of QA groups should increase upon increasing immersion time that directly determines the extent of reaction between –CH2Cl and TMPDA. When the immersing time increases to 5 h, there is no change in IEC for TPSF-5 h, because of the full conversion of –CH2Cl groups to QA groups. 3.6. Water uptake The presence of water molecules in AEM is considered to be essential for facilitated ion transport, and water uptake with respect to the membrane hydrophilicity thus becomes an important parameter of AEM. As shown in Fig. 6, the CMPSF membrane has a water uptake of 172% at 25 °C. By contract, the water uptake value increases from TPSF-1 h to TPSF-4 h with increasing IEC at a given temperature.
Fig. 3. The XPS high-resolution spectra of N1s region of CMPSF and TPSF membranes.
560
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
Fig. 4. SEM images of CMPSF and TPSF membranes.
Fig. 5. Ion exchange capacity (IEC) values of CMPSF and TPSF membranes.
Fig. 6. Water uptake (WU) values of CMPSF and TPSF membranes.
Specifically, the water uptake values of TPSF-1 h and TPSF-4 h membranes are 260% and 310% at 25 °C, respectively. This is due to the enhanced hydrophilicity of QA groups. The water uptake value of TPSF-5 h is similar to that of TPSF-4 h because of the similar IEC. Normally, the membrane swells severely at the higher temperature because of more water adsorption, leading to an increase in free space in the membrane. However, the water uptake values of TPSF membranes at 65 °C are similar to those at 25 °C because the crosslinking effect of TMPDA leads to the improved swelling stability of the final TPSF membranes. It is noted that the mechanical properties of the conventional dense AEMs in the wet state are poor at elevated temperature due to the membrane swelling, thus limiting their
practical application. By contrast, TPSF membranes with the good swelling stability show the good mechanical properties at high temperature.
3.7. Diffusion dialysis performance Fig. 7 shows the DD performance including acid dialysis coefficient (UH+) and separation factor (S) of TPSF membranes at 25 °C. With increasing the immersing time in TMPDA solution, UH+ and separation factor from TPSF-1 h to TPSF-5 h generally increase from 0.042 m h−1 and 24.5–0.065 m h−1 and 34.0, respectively, resulting from the simultaneously increased IEC and crosslinking degree. 561
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
Fig. 7. Acid dialysis coefficients (UH+) and separation factors (S) of TPSF AEMs at 25 °C. Fig. 8. TGA curves of PSF and TPSF membranes.
The diffusion dialysis related performances including IEC, UH+ and separation factor of the commercial DF-120 membrane and other reported AEMs for acid recovery from the mixture of HCl/FeCl2 solution are summarized in Table 2. The commercial DF-120 membrane has a UH+ value of 0.0085 m h−1 and S value of 18.5. Therefore, TPSF-5 h membrane shows 6.6 times increase in UH+ and 84% increase in S. This is a great improvement in diffusion dialysis performance when comparing the commercial DF-120 membrane with TPSF-5 h membrane. In other words, there are 6.6 times increase in acid recovery rate and 84% increase in the final acid purity if our membrane is used to replace the commercial DF-120 membrane. Obviously, TPSF5 h hold promise for further development for commercial acid recovery. Moreover, TPSF-5 h membrane also shows relatively higher UH+ at a given separation factor as compared with other reported dense anion exchange membranes [23–33]. The improvement in the diffusion dialysis performance should be attributed to the asymmetric structure of TPSF membranes composed of a thin selective layer and a porous supporting layer. It is worth mentioning that the supporting layer with abundant macropores would present minimal ion transport resistance due to the extremely high free space [37–39]. Therefore, the ion transport resistance should be mainly from the top layer. Different from the conventional dense membranes with thicknesses of tens to hundreds μm, TPSF membranes have a selective layer thickness of 0.5– 0.6 µm. The greatly reduced thickness will significantly facilitate the ion transport. In addition to the improvement in diffusion dialysis performance, TPSF membranes are also of low cost as compared with the conventional dense membrane, due to the simple fabrication method involving the industry-standard non-solvent phase inversion process and one-step functionalization. In our previous research, similar AEMs were fabricated using BPPO
ultrafiltration membrane as a substrate, their diffusion dialysis performances were also summarized in Table 2. When polyetherimide (PEI) and trimethylamine (TMA) were used as crosslinker and quaternizing agent respectively, the optimal membrane showed UH+ of 0.063 m h−1 and S of 20.0 [37]. When the crosslinker was replaced with butanediamine (BTDA), the optimal membrane showed the corresponding values of 0.063 m h−1 and 30.4 [38], respectively. Note that TPSF-5 h membrane shows higher separation factor at the similar UH+. Moreover, when tetramethylethylenediamine (TEMED) was used as simultaneous crosslinker and quaternization agent, the optimal membranes showed UH+ of only 0.043 mol h−1 [39], TPSF-5 h membrane shows about 50% increase in UH+ due to the decreased thickness of the active layer (from 0.8 to 0.9 µm to 0.5–0.6 µm). The improvement in diffusion dialysis performance indicates the PSF is more suited than BPPO to be a starting material for fabrication of porous AEM membranes for acid recovery.
3.8. Thermal stability The thermal stability of the resultant porous AEMs was investigated by TGA. Fig. 8 shows TGA curves of the selected TPSF-5 h and PSF membrane. For PSF membrane, it was thermally stable even when the temperature is up to 500 °C. For TPSF membrane, the initial weight loss ( < 5%) at 50–120 °C was caused by the water evaporation of the residual water and thus can be ignored. The second weight loss from 150 °C was caused by the degradation of QA groups [52]. It means that no chemical degradation occurs at temperatures lower than 150 °C, indicating the good thermal stability of TPSF membranes.
Table 2 Ion exchange capacity (IEC), acid dialysis coefficient (UH+) and separation factor (S) of the reported membranes at 25 °C using HCl/FeCl2 solution as model acidic waste solution. Membrane
Compactness
IEC (m molg−1)
UH+ (10−3 m h−1)
S
Ref.
TPSF membranes DF-120 commercial membrane Quaternized PPO based hybrid membranes PVA and glycidyltrimethyl ammonium chloride (EPTAC) blending membranes PVA and multi-alkoxy silicon copolymer blending membranes PVA treated with alkoxysilanes membranes Quaternized poly(VBC-co-γ-MPS) membranes Quaternized Bionic multisilicon copolymers Quaternized blending of PPO and PVA membranes Quaternized aromatic amine based hybrid PVA membranes Quaternized PPO blending with PVA and silanol Quaternized blending of PVC and P(DMAM-co-DVB) Imidazolium functionalized hybrid membrane blending with PVA BPPO ultrafiltration membranes crosslinked by PEI and quaternized by TMA BPPO ultrafiltration membranes crosslinked by BTDA and quaternized by TMA BPPO ultrafiltration membranes simultaneously and crosslinked by TEMED
Porous Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Dense Porous Porous Porous
0.82–1.18 1.96 1.7–2.2 0.58–1.15 0.34–0.76 0.52–1.01 0.6–0.9 0.46–1.25 1.22–1.92 0.65–1.12 1.04–1.08 0.34–1.32 1.34–1.86 2.13 1.45–1.01 0.51–1.43
42–65 8.5 5–11 11–18 10–17 8–10 24–43 7.2–74.8 21–49 17.2–25.2 9.5–14.5 12.0–40.0 18.7–48.3 63 62–41 9–43
24.5–34.0 18.5 17.0–32.0 18.5–21.0 24.0–30.1 15.9–21.0 22–26 25.9–42.8 26–39 14–21 45.0–67.5 36–61 12.72–52.5 20.0 30.4–80.4 46.6–73.8
This work – [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [37] [38] [39]
562
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
References [1] G. Merle, M. Wessling, K. Nijmeijer, Anion exchange membranes for alkaline fuel cells: a review, J. Membr. Sci. 377 (2011) 1–35. [2] J.R. Varcoe, R.C. Slade, Prospects for alkaline anion‐exchange membranes in low temperature fuel cells, Fuel cells 5 (2005) 187–200. [3] G. Couture, A. Alaaeddine, F. Boschet, B. Ameduri, Polymeric materials as anionexchange membranes for alkaline fuel cells, Prog. Polym. Sci. 36 (2011) 1521–1557. [4] X. Li, H. Zhang, Z. Mai, H. Zhang, I. Vankelecom, Ion exchange membranes for vanadium redox flow battery (VRB) applications, Environ. Sci. Technol. 4 (2011) 1147–1160. [5] S. Zhang, C. Yin, D. Xing, D. Yang, X. Jian, Preparation of chloromethylated/ quaternized poly (phthalazinone ether ketone) anion exchange membrane materials for vanadium redox flow battery applications, J. Membr. Sci. 363 (2010) 243–249. [6] T. Mohammadi, M.S. Kazacos, Modification of anion-exchange membranes for vanadium redox flow battery applications, J. Power Sources 63 (1996) 179–186. [7] T. Sata, K. Teshima, T. Yamaguchi, Permselectivity between two anions in anion exchange membranes crosslinked with various diamines in electrodialysis, J. Polym. Sci. A: Polym. Chem. 34 (1996) 1475–1482. [8] M. Kariduraganavar, R. Nagarale, A. Kittur, S. Kulkarni, Ion-exchange membranes: preparative methods for electrodialysis and fuel cell applications, Desalination 197 (2006) 225–246. [9] T. Xu, Ion exchange membranes: state of their development and perspective, J. Membr. Sci. 263 (2005) 1–29. [10] L. Rocchetti, F. Vegliò, B. Kopacek, F. Beolchini, Environmental impact assessment of hydrometallurgical processes for metal recovery from WEEE residues using a portable prototype plant, Environ. Sci. Technol. 47 (2013) 1581–1588. [11] S. Sahu, K. Sahu, B. Pandey, Leaching of zinc sulfide concentrate from the ganeshhimal deposit of nepal, Metall. Mater. Trans. B 37 (2006) 541–549. [12] A.Kumari, M.K.Sinha, S.K.Sahu, B.D.Pandey, Investigation of a novel ionic liquid, Cyphos IL 104 for the solvent extraction of mineral acids, Hydrometallurgy.http:// dx.doi.org/10.1016/j.hydromet.2015.09.016 [13] J.G. Dean, F.L. Bosqui, K.H. Lanouette, Removing heavy metals from waste water, Environ. Sci. Technol. 6 (1972) 518–522. [14] H.C. Bramer, Pollution control in the steel industry, Environ. Sci. Technol. 5 (1971) 1004–1008. [15] M. Regel-Rosocka, A review on methods of regeneration of spent pickling solutions from steel processing, J. Hazard. Mater. 177 (2010) 57–69. [16] J. Luo, C. Wu, T. Xu, Y. Wu, Diffusion dialysis-concept, principle and applications, J. Membr. Sci. 366 (2011) 1–16. [17] S.J. Oh, S.-H. Moon, T. Davis, Effects of metal ions on diffusion dialysis of inorganic acids, J. Membr. Sci. 169 (2000) 95–105. [18] M.-S. Kang, K.-S. Yoo, S.-J. Oh, S.-H. Moon, A lumped parameter model to predict hydrochloric acid recovery in diffusion dialysis, J. Membr. Sci. 188 (2001) 61–70. [19] D.M. Stachera, R.F. Childs, Tuning the acid recovery performance of poly(4vinylpyridine)-filled membranes by the introduction of hydrophobic groups, J. Membr. Sci. 187 (2001) 213–225. [20] D.M. Stachera, R.F. Childs, A.M. Mika, J.M. Dickson, Acid recovery using diffusion dialysis with poly(4-vinylpyridine)-filled microporous membranes, J. Membr. Sci. 148 (1998) 119–127. [21] A.K. Pandey, R.F. Childs, M. West, J.N. Lott, B.E. McCarry, J.M. Dickson, Formation of pore‐filled ion‐exchange membranes with in situ crosslinking: Poly (vinylbenzyl ammonium salt)‐filled membranes, J. Polym. Sci. A: Polym. Chem. 39 (2001) 807–820. [22] D.-H. Kim, J.-H. Park, S.-J. Seo, J.-S. Park, S. Jung, Y.S. Kang, J.-H. Choi, M.S. Kang, Development of thin anion-exchange pore-filled membranes for high diffusion dialysis performance, J. Membr. Sci. 447 (2013) 80–86. [23] J. Luo, C. Wu, Y. Wu, T. Xu, Diffusion dialysis of hydrochloride acid at different temperatures using PPO–SiO2 hybrid anion exchange membranes, J. Membr. Sci. 347 (2010) 240–249. [24] C. Cheng, Z. Yang, J. Pan, B. Tong, T. Xu, Facile and cost effective PVA based hybrid membrane fabrication for acid recovery, Sep. Purif. Technol. 136 (2014) 250–257. [25] C. Wu, Y. Wu, J. Luo, T. Xu, Y. Fu, Anion exchange hybrid membranes from PVA and multi-alkoxy silicon copolymer tailored for diffusion dialysis process, J. Membr. Sci. 356 (2010) 96–104. [26] Y. Wu, C. Wu, Y. Li, T. Xu, Y. Fu, PVA–silica anion-exchange hybrid membranes prepared through a copolymer crosslinking agent, J. Membr. Sci. 350 (2010) 322–332. [27] Y. Wu, J. Luo, L. Yao, C. Wu, F. Mao, T. Xu, PVA/SiO2 anion exchange hybrid membranes from multisilicon copolymers with two types of molecular weights, J. Membr. Sci. 399–400 (2012) 16–27. [28] Y. Wu, J. Luo, C. Wu, T. Xu, Y. Fu, Bionic multisilicon copolymers used as novel cross-linking agents for preparing anion exchange hybrid membranes, J. Phys. Chem. B 115 (2011) 6474–6483. [29] Y. Wu, J. Luo, L. Zhao, G. Zhang, C. Wu, T. Xu, QPPO/PVA anion exchange hybrid membranes from double crosslinking agents for acid recovery, J. Membr. Sci. 428 (2013) 95–103. [30] A.N. Mondal, C. Cheng, Z. Yao, J. Pan, M.M. Hossain, M.I. Khan, Z. Yang, L. Wu, T. Xu, Novel quaternized aromatic amine based hybrid PVA membranes for acid recovery, J. Membr. Sci. 490 (2015) 29–37. [31] Y. Wu, M. Jiang, J. Cao, T. Xu, F. Mao, Combination of OH– ions and –OH groups within QPPO/PVA hybrid membranes for acid recovery, Desalination Water Treat. (2015) 1–11.
Fig. 9. The weight percentage of the TPSF-5 h membrane after immersion in aqueous HCl/FeCl2 solution at 65 °C for different periods of times.
3.9. Chemical stability As mentioned above, the weight loss of TPSF membrane in the mixture of HCl/FeCl2 solution after different times was evaluated to investigate the chemical stability of TPSF membrane according to our previous work [38]. As shown in Fig. 9, it can be seen that TPSF-5 h shows good chemical stability as no weight loss occurs even after immersing the TPSF-5 h in the mixture of HCl/FeCl2 aqueous solution at 65 °C for 7 days. Moreover, the diffusion dialysis performance of TPSF-5 h after immersing in the above hot acidic solution for 7 days was also tested, it was found that the treated TPSF-5 h AEM finally showed UH+ of 0.064 m h−1 and S of 36.9, both of which are similar to the initial values. The chemical stability of the membrane in the hot acidic solution demonstrates that our membranes are suitable for practical diffusion dialysis applications.
4. Conclusions The common chloromethylated PSF rather than BPPO was selected as starting material to fabricate the porous AEMs at a lower cost for practical diffusion dialysis application to further explore the recently developed method based on ultrafiltration membrane for the fabrication of the porous AEMs. Specifically, porous TPSF AEMs were prepared by immersing chloromethylated PSF ultrafiltration membrane in TMPDA solution to realize simultaneous crosslinking and quaternization. The resulting membranes show good swelling, thermal and chemical stabilities. The optimal TPSF-5 h membrane shows an acid dialysis coefficient of 0.065 m h−1 and a separation factor of 34.0, which are superior to the common dense AEMs and our previous porous membranes based on BPPO ultrafiltration membranes for diffusion dialysis at the same test condition. Specifically, the optimal TPSF-5 h membrane shows 6.6 times higher acid dialysis coefficient 0.84 time higher separation factor than DF-120 membranes. This work also suggests that polysulfone is a better polymer than BPPO for the fabrication of porous membrane for acid recovery. Given the simple process for the membrane fabrication, TPSF membranes with high performance but low cost are promising for the large-scale, practical application of high-efficiency acid recovery.
Acknowledgement This work was supported by the Australian Research Council (Project no. DP140101591). The authors thank technical support from staff at Monash Centre for Electron Microscopy. 563
Journal of Membrane Science 524 (2017) 557–564
X. Lin et al.
J. Membr. Sci. 120 (1996) 239–259. [43] R.P. Pinell, G.D. Khune, N.A. Khatri, S.L. Manatt, Concerning the chloromethylation of alkylbenzenes and polystyrenes by chloromethyl methyl ether, Tetrahedron Lett. 25 (1984) 3511–3514. [44] R.C. Fuson, C. McKeever, Chloromethylation of aromatic compounds, Org. React., 1942. [45] E. Shamsaei, Z.-X. Low, X. Lin, Z. Liu, H. Wang, Polysulfone and its quaternary phosphonium derivative composite membranes with high water flux, Ind. Eng. Chem. Res. 54 (2015) 3333–3340. [46] X. Lin, K. Wang, Y. Feng, J.Z. Liu, X. Fang, T. Xu, H. Wang, Composite ultrafiltration membranes from polymer and its quaternary phosphonium-functionalized derivative with enhanced water flux, J. Membr. Sci. 482 (2015) 67–75. [47] S. Gu, R. Cai, T. Luo, Z. Chen, M. Sun, Y. Liu, G. He, Y. Yan, A Soluble and Highly Conductive Ionomer for High-Performance Hydroxide Exchange Membrane Fuel Cells, Angew. Chem. Int. Ed. 48 (2009) 6499–6502. [48] T. Xu, W. Yang, Sulfuric acid recovery from titanium white (pigment) waste liquor using diffusion dialysis with a new series of anion exchange membranes — static runs, J. Membr. Sci. 183 (2001) 193–200. [49] C.G. Arges, V. Ramani, Two-dimensional NMR spectroscopy reveals cationtriggered backbone degradation in polysulfone-based anion exchange membranes, Proc. Natl. Acad. Sci. USA 110 (2013) 2490–2495. [50] J. Chastain, R.C. King, J. Moulder, Handbook of X-ray Photoelectron Spectroscopy: A Reference Book of Standard Spectra for Identification and Interpretation of XPS Data, Physical Electronics Eden Prairie, MN, 1995. [51] M. Maniruzzaman, D.J. Morgan, A.P. Mendham, J. Pang, M.J. Snowden, D. Douroumis, Drug–polymer intermolecular interactions in hot-melt extruded solid dispersions, Int. J. Pharm. 443 (2013) 199–208. [52] W. Xie, Z. Gao, W.-P. Pan, D. Hunter, A. Singh, R. Vaia, Thermal degradation chemistry of alkyl quaternary ammonium montmorillonite, Chem. Mater. 13 (2001) 2979–2990.
[32] C. Cheng, Z. Yang, Y. He, A.N. Mondal, E. Bakangura, T. Xu, Diffusion dialysis membranes with semi-interpenetrating network for acid recovery, J. Membr. Sci. 493 (2015) 645–653. [33] K. Emmanuel, C. Cheng, B. Erigene, A.N. Mondal, M.M. Hossain, M.I. Khan, N.U. Afsar, G. Liang, L. Wu, T. Xu, Imidazolium functionalized anion exchange membrane blended with PVA for acid recovery via diffusion dialysis process, J. Membr. Sci. 497 (2016) 209–215. [34] T.Asawa, T.Gunjima, Acid diffusion-dialysis process utilizing anion exchange membrane of 4–55 μm thickness, in, Google Patents, 1976. [35] B. Van der Bruggen, C. Vandecasteele, T. Van Gestel, W. Doyen, R. Leysen, A review of pressure‐driven membrane processes in wastewater treatment and drinking water production, Environ. Prog. 22 (2003) 46–56. [36] X.-L. Wang, T. Tsuru, S.-i Nakao, S. Kimura, The electrostatic and steric-hindrance model for the transport of charged solutes through nanofiltration membranes, J. Membr. Sci. 135 (1997) 19–32. [37] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, Y. Hu, T. Xu, H. Wang, Porous diffusion dialysis membranes for rapid acid recovery, J. Membr. Sci. 502 (2016) 76–83. [38] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, D. Zhao, T. Xu, Z. Xie, C.D. Easton, H. Wang, Asymmetrically porous anion exchange membranes with an ultrathin selective layer for rapid acid recovery, J. Membr. Sci. 510 (2016) 437–446. [39] X. Lin, E. Shamsaei, B. Kong, J.Z. Liu, T. Xu, H. Wang, Fabrication of asymmetrical diffusion dialysis membranes for rapid acid recovery with high purity, J. Mater. Chem. A 3 (2015) 24000–24007. [40] K. Zodrow, L. Brunet, S. Mahendra, D. Li, A. Zhang, Q. Li, P.J. Alvarez, Polysulfone ultrafiltration membranes impregnated with silver nanoparticles show improved biofouling resistance and virus removal, Water Res. 43 (2009) 715–723. [41] T.A. Tweddle, O. Kutowy, W.L. Thayer, S. Sourirajan, Polysulfone ultrafiltration membranes, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983) 320–326. [42] M. Ulbricht, M. Riedel, U. Marx, Novel photochemical surface functionalization of polysulfone ultrafiltration membranes for covalent immobilization of biomolecules,
564