Semi-aromatic polyamides containing methylene and thioether units: synthesis and membrane properties

Abstract

A series of semi-aromatic diamines containing a different content of methylene and thioether units were synthesized via reaction with 4-aminobenzenethiol (4-ABT) and a semi-aromatic difluorobenzamide. They were then reacted with 4,4′-thiodibenzoyl chloride (TDC) to yield a series of semi-aromatic polyamides. Thermal property investigation confirmed a high glass transition temperature (T_g) of 163–203 °C and a good thermal stability with initial degradation temperatures (T_d) of 409 to 419 °C for the synthesized semi-aromatic polyamides, as determined by differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) characterization. A water flux and retention rate for the resultant semi-aromatic polyamide membranes in the range of 89–371 L (m² h)⁻¹ was observed when measured by ultrafiltration cup experiments, which suggests that they are suitable for separation processes. The resultant semi-aromatic polyamides showed good corrosion resistance especially in a strong alkaline environment. It was found that they can withstand a solution of NaOH (2 mol L⁻¹) for nearly 6 months without loss of mechanical strength.

---

Introduction

In recent years, ultrafiltration, as a novel, green and powerful technique, has received tremendous attention in fractionating protein solutions and in the wastewater treatment field.^1–6 In order to meet the requirement of the application of these membranes in harsh environments, several kinds of high performance polymeric membrane material have been developed.^7–9 These approaches comprised the following methods such as: introduction of an aromatic heterocyclic group into the main polymeric chain^10–12 and incorporation of polycyclic units into the polymer backbone.^13,14 All these membrane materials were found to have excellent thermal performance. And these newly developed polymeric membrane materials with their excellent mechanical strength and thermal stability should have a potential application in high temperature ultrafiltration and organic solvent nanofiltration. In order to improve the comprehensive properties of the membranes, to allow their use in more rigorous environments, our interest has focused on improving the corrosion resistance whilst at the same time maintaining the membrane materials’ thermal properties.

At present there are some reports concerning membrane materials used in organic solvents, but studies about membranes applied in strongly alkaline and polar solvent conditions except for a few resins, such as polyphenylene sulfide¹⁵ and so on, are scant. In addition, we found that some of the commercially available membrane materials containing full aromatic units such as polyether sulfone (PES),^16–20 polyphenylene sulfone containing diphenyl units (PPSU)^21,22 and aromatic polyimides (PI)²³ could be used at high temperature, but they were soluble or became brittle when applied in strongly polar organic solvents or alkaline conditions.

During the experimental process, we found another interesting phenomenon: the mechanical properties and exterior of the buckets made from polyolefins such as polypropylene (PP) or PE changed little after immersion in strongly alkaline solutions (NaOH 2 mol L⁻¹) for about 2 years. This observed phenomenon, generated a question for us: why does the bucket display such excellent alkaline corrosion resistance for so long a time? Then we found that the flexible aliphatic units such as methylene played an important role in the materials’ alkaline corrosion resistance. The methylene group is un-reactive to alkaline reagent. In addition, the methylene unit is a flexible bond, it is beneficial for improving the polymer’s degree of crystallinity. Thus, the solvent is hard to permeate into the polymers. Additionally, as we know, poly(aryl thioether)s and poly(phenylene sulfide sulfone)^24–29 are a type of high performance engineering thermoplastic. They all have excellent corrosion resistance properties because of the introduction of a high content of thioether units (–S–) and aromatic benzene rings.

In order to further investigate how the methylene and thioether units affect the polymer’s corrosion resistance, we designed a series of diamines containing thioether and a different content of methylene units to react with 4,4′-thiodibenzoyl chloride (TDC) containing a thioether group (the flexible thioether is beneficial to yield high molecular weight semi-aromatic polyamides and maintain the polymers’ corrosion resistance) to yield semi-aromatic polyamides. The effects of introduction of thioethers and a different content of methylene units in the polymeric main chain on the thermal properties and corrosion resistance of these semi-aromatic polyamides are discussed. The membrane structure and the separation properties were also studied.

Experimental

Materials

4-Fluorobenzoic chloride (4-FBC) (99.5%, Lanning Chemical Company Limited), 4-aminobenzenethiol (4-ABT) (99.0%, ShouErFu Chemical Company Limited), semi-aromatic difluorobenzamides [1,2-N,N′-bis(4-fluorobenzamide)ethane (BFBE), 1,4-N,N′-bis(4-fluorobenzamide)butane (BFBB), 1,6-N,N′-bis(4-fluorobenzamide)hexane (BFBH), 1,8-N,N′-bis(4-fluorobenzamide)octane (BFBO), 1,10-N,N′-bis(4-fluorobenzamide)decane (BFBD)] and 4,4′-thiodibenzoyl chloride (TDC) were prepared according to the method reported earlier by our group,^30,31 N-methyl-2-pyrrolidone (NMP) (JiangSu NanJing JinLong Chemical Industry Company), polyether sulfone (PES, BASF, E6020P), polyvinylidene fluoride (PVDF, Solvay 6010), sodium hydroxide (NaOH) and other reagents were obtained commercially.

Monomer synthesis

Synthesis of 1,6-N,N′-bis(4-aminobenzenethiolbenzamide)hexane (BATH) (as shown in Scheme 1). BFBH (36 g, 0.1 mol), 4-ABT (50 g, 0.2 mol), NaOH (8 g, 0.2 mol) and N,N-dimethylacetamide (DMAC, 100 mL) were added to a three-necked round bottomed flask, equipped with a stirring machine, thermometer and a nitrogen inlet and outlet. The mixture was heated to 120–130 °C to remove the byproducts of water for about 1–2 h. The color of the mixture changed from yellow to brown with the evaporation of water. Then the temperature was gradually increased to the boiling point of the solvents (DMAC) and maintained another 8 h to complete the reaction. When the temperature of the mixture cooled down to room temperature, it was poured into 800 mL of water to precipitate the crude product. Then, the mixture was filtered and the solid sample collected, the solid product was further purified with water and ethanol washing. Then it was vacuum dried at 80 °C for 12 h to yield the pure product 52.2 g (91.6%).

FT-IR (KBr, cm⁻¹): 3353, 3238 (–NH₂), 3441 (–CONH–), 3024 (C–H, aromatic ring), 2931, 2860 (–CH₂–), 1617, 1533 (–CONH–), 1594, 1496 (C=C aromatic ring), 1084 (–S–), 826 (para-substituted benzene ring); ¹H-NMR (400 MHz, DMSO, ppm): 1.296 (s, 4H, H₁), 1.466–1.497 (m, 4H, H₂), 3.178–3.226 (m, 4H, H₃), 5.599 (s, 4H, H₄), 6.622–6.657 (m, 4H, H₅), 6.999–7.021 (d, 4H, H₆), 7.173–7.209 (m, 4H, H₇), 7.673–7.695 (m, 4H, H₈), 8.317–8.345 (m, 2H, H₉).

BATE, BATB, BATO and BATD were synthesized and purified with a similar procedure as BATH.

BATE: yield, 47.6 g, 91.3%, FT-IR (KBr, cm⁻¹): 3358, 3215 (–NH₂), 3434 (–CONH–), 3049 (C–H, aromatic ring), 2944, 2861 (–CH₂–), 1631, 1534 (–CONH–), 1594, 1493 (C=C aromatic ring), 1086 (–S–), 825 (para-substituted benzene ring); ¹H-NMR (400 MHz, DMSO, ppm): 3.378 (s, 4H, H₁), 5.596 (s, 4H, H₂), 6.622–6.657 (d, 4H, H₃), 7.005–7.027 (d, 4H, H₄), 7.139–7.172 (d, 4H, H₅), 7.682–7.703 (d, 4H, H₆), 8.485 (s, 2H, H₇).

BATB: yield, 50.2 g, 92.0%, FT-IR (KBr, cm⁻¹): 3357, 3314 (–NH₂), 3458 (–CONH–), 3050 (C–H, aromatic ring), 2932, 2861 (–CH₂–), 1615, 1527 (–CONH–), 1593, 1496 (C=C aromatic ring), 1085 (–S–), 829 (para-substituted benzene ring); ¹H-NMR (400 MHz, DMSO, ppm): 1.513 (s, 4H, H₁), 3.229–3.243 (d, 4H, H₂), 5.599 (s, 4H, H₃), 6.635–6.656 (m, 4H, H₄), 7.003–7.024 (d, 4H, H₅), 7.184–7.205 (m, 4H, H₆), 7.678–7.700 (d, 4H, H₇), 8.365 (s, 2H, H₈).

BATO: yield, 53.8 g, 90.5%, FT-IR (KBr, cm⁻¹): 3349, 3232 (–NH₂), 3456 (–CONH–), 3029 (C–H, aromatic ring), 2929, 2857 (–CH₂–), 1629, 1531 (–CONH–), 1594, 1495 (C=C aromatic ring), 1084 (–S–), 824 (para-substituted benzene ring); ¹H-NMR (400 MHz, DMSO, ppm): 1.266 (s, 8H, H₁–H₂), 1.458–1.488 (m, 4H, H₃), 3.169–3.219 (m, 4H, H₄), 5.599 (s, 4H, H₅), 6.621–6.656 (m, 4H, H₆), 7.000–7.021 (d, 4H, H₇), 7.172–7.207 (m, 4H, H₈), 7.673–7.694 (d, 4H, H₉), 8.311–8.339 (m, 2H, H₁₀).

BATD: yield, 56.3 g, 91.1%, FT-IR (KBr, cm⁻¹): 3336, 3221 (–NH₂), 3424 (–CONH–), 3028 (C–H, aromatic ring), 2921, 2850 (–CH₂–), 1629, 1535 (–CONH–), 1594, 1495 (C=C aromatic ring), 1085 (–S–), 824 (para-substituted benzene ring); ¹H-NMR (400 MHz, DMSO, ppm): 1.243 (s, 12H, H₁–H₃), 1.453–1.485 (m, 4H, H₄), 3.171–3.220 (m, 4H, H₅), 5.600 (s, 4H, H₆), 6.637–6.653 (m, 4H, H₇), 7.003–7.024 (d, 4H, H₈), 7.182–7.203 (m, 4H, H₉), 7.677–7.698 (d, 4H, H₁₀), 8.312–8.340 (m, 2H, H₁₁).

Polymer synthesis (as shown in Scheme 2)

A typical polymerization was conducted as following: BATH (5.7 g, 0.01 mol), Et₃N (2.02 g, 0.02 mol) and NMP (50 mL) were added to a round bottomed single-necked flask, and the resulting mixture was stirred with a magnetic stirring bar. Equimolar TDC (3.1 g, 0.01 mol) was added to the mixture when BATH had dissolved completely. The reaction was maintained for about 8–10 h at ambient temperature. A viscous brown solution was obtained. The solution was poured into water to precipitate a fibrous light-yellow solid. It was crushed into a powder and washed with deionized water and ethanol 3 times, respectively. Then the crude product was vacuum dried at 100 °C for 12 h to yield TDC-BATH 7.6 g (94.5%).

FT-IR (KBr, cm⁻¹): 3426 (–CONH–), 3092 (C–H, aromatic ring), 2927, 2855 (–CH₂–), 1637, 1518 (–CONH–), 1590, 1481 (C=C aromatic ring), 1082 (–S–), 828 (para-substituted benzene ring).

TDC-BATE, TDC-BATB, TDC-BATO and TDC-BATD were synthesized and purified by a similar procedure to TDC-BATH. But the molecular weight of TDC-BATE is so small that it can not form a membrane. So we no data were obtained for TDC-BATE.

TDC-BATB: yield, 7.3 g, 93.3%, FT-IR (KBr, cm⁻¹): 3427 (–CONH–), 3056 (C–H, aromatic ring), 2928, 2854 (–CH₂–), 1636, 1520 (–CONH–), 1590, 1481 (C=C aromatic ring), 1081 (–S–), 827 (para-substituted benzene ring).

TDC-BATO: yield, 7.8 g, 93.8%, FT-IR (KBr, cm⁻¹): 3434 (–CONH–), 3076 (C–H, aromatic ring), 2919, 2848 (–CH₂–), 1638, 1516 (–CONH–), 1589, 1481 (C=C aromatic ring), 1080 (–S–), 831 (para-substituted benzene ring).

TDC-BATD: yield, 8.0 g, 93.1%, FT-IR (KBr, cm⁻¹): 3433 (–CONH–), 3052 (C–H, aromatic ring), 2919, 2848 (–CH₂–), 1637, 1516 (–CONH–), 1590, 1482 (C=C aromatic ring), 1079 (–S–), 834 (para-substituted benzene ring).

Characterization

Intrinsic viscosity analysis: the intrinsic viscosity (η_int) of these semi-aromatic polyamides was measured by a Cannon–Ubbelodhe viscometer at 30 ± 0.1 °C. Polymer solutions were prepared by dissolving 0.500 g polymer into 100 mL H₂SO₄. The results were obtained by the one-point method (or Solomon–Ciuta equation) as follows:where η_r = t/t₀, η_sp = t/t₀ − 1, where t is the solution flow time (s), t₀ is the solvents (H₂SO₄) flow time (s) and C is the solution concentration (g dL⁻¹).

Chemical structure analysis

FT-IR spectroscopic measurements were performed on a NEXUS670 FT-IR instrument. ¹H-NMR was obtained on a Bruker-400 NMR spectrometer in deuterated DMSO. The X-ray diffraction (XRD) was performed with Philips X’pert Pro MPD to study the polymer’s aggregation structure.

Thermal analysis

Thermal analyses of the samples were performed by differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA). DSC measurements were determined using a NETZSCH DSC 200 PC thermal analysis instrument. The heating rate for the DSC measurements was 10 °C min⁻¹ under a nitrogen atmosphere. TGA measurements were determined using a TGA Q500 V6.4 Build 193 thermal analysis instrument. The heating rate for the TGA measurements was 10 °C min⁻¹ under a nitrogen atmosphere.

Preparation of separation membranes (8 wt%)

To a round bottomed flask, 5 g of semi-aromatic polyamides [PES or PVDF (16 wt%): 5 g, NMP: 26.25 g] and 57.5 g of H₂SO₄ were added, the mixture was dissolved under ultrasonication at 60 °C to form an 8 wt% polymer solution. The solution was coated onto a clean glass plate; then the glass plate was immersed into deionized water to form the separation membranes.

Mechanical testing of the separation membranes

The stress–strain behavior of the resultant polymer membranes was measured with an Instron Corporation 4302 instrument at room temperature. The specification of the membrane specimens was 50 mm × 3.5 mm and five replicas should be prepared for every sample. An average value of five replicas was used.

SEM analysis of the resultant separation membranes

Scanning electron microscopy (SEM) was employed to observe the cross section morphology of the obtained semi-aromatic polyamides separation membranes. The membranes were frozen in liquid nitrogen to break into pieces. The membrane pieces were gold-sputtered for use as a conductive material. Then the treated membranes were measured on a J Corporation JSM-7500F instrument.

Flux testing of the separation membranes

Fluxes of the separation membranes were performed on a Mosu Corporation MSC50 Ultrafiltration cup with a volume of 50 mL. The membranes were fixed in the cell of the ultrafiltration cup and then the deionized water was filled in the cell. Nitrogen was used as a pressure gas. The membranes were compacted by nitrogen gas at 0.2 MPa for 10 min; then kept at a pressure of 0.1 MPa to measure the pure water fluxes of the membranes. The effective area of the membranes was 1075.21 mm². The pure water flux was obtained by the following equation:where J is the pure water flux (L (m² h)⁻¹), Q is the volume of permeated water (L), A is the effective area (m²), Δt is the sampling time (h). The solute (bovine serum albumin) retention rate (R) was defined by the following equation:where C_p and C_f (mg mL⁻¹) were the concentration of permeate and feed solutions, respectively. The concentrations of the feed and permeate solutions were examined by UV-vis spectroscopy.

Porosity testing of the separation membranes

The porosity of the separation membranes was tested by the following method:

The water of the membrane surface was carefully wiped with filter paper after the membrane was taken out from the water. Then the membrane was weighed to obtain the wet membrane weight. After that, the membrane was dried at 100 °C for 10 h and then weighed again to obtain the dry membrane weight. The membrane porosity is given by the following equation:

where P is the porosity (%), W₁ is the wet membrane weight (g), W₂ is the dry membrane weight (g), A is the membrane area (cm²), d is the thickness of the membrane (cm), ρ is the density of water (g cm⁻³). Five replicas were prepared for every sample. An average value of five replicas was used.

Alkaline and strongly polar organic solvent corrosion resistance experiments:

The separation membrane was immersed in a solution of NaOH (2 mol L⁻¹), concentrated aqueous ammonia and NMP for about 1 week, 1 month, 3 months and 6 months, respectively. Then the tensile strength of each sample was tested.

Results and discussion

Synthesis and chemical structure of monomers (BATE, BATB, BATH, BATO and BATD) and polymers (TDC-BATE, TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD)

The monomers were prepared according to a nucleophilic substitution reaction with the difluoro-substituted semi-aromatic amide and 4-ABT as reagent. The yield of these monomers reached up to 90%. The main reason for this being the high reactivity of the difluoro-substituted precursor caused by the strong electron-withdrawing effect of the amide unit and F atom. The FT-IR spectra of the monomers are shown in Fig. 1. From Fig. 1, the characteristic absorption of amine and thioether units near 3360 and 3230 cm⁻¹, 1080 cm⁻¹ was observed. This suggests a nucleophilic substitution reaction between the difluoro-substituted semi-aromatic amide and 4-ABT has occurred. These monomers were also characterized with ¹H-NMR. As shown in Fig. 2 and the ESI Fig. S1,† BATE, BATB, BATH, BATO and BATD was found to have 7, 8, 9, 9 and 9 groups of proton signals, respectively. According to the molecular chemical structure, BATO and BATD should have 10 and 11 groups of peaks, but the protons’ chemical environment which is far from the amide group were so similar that they proved difficult to split, then the proton signals (H₁–H₂ in BATO and H₁–H₃ in BATD) were found to be a single and wide peak. But the integral ratio of the whole monomers is completely consistent with the molecular formula. The above results suggest the monomers are synthesized successfully as described in Scheme 1.

Fig. 1 The FT-IR spectra of monomers (BATE, BATB, BATH, BATO and BATD).

Fig. 2 The ¹H-NMR spectra of monomers (BATE, BATH and BATD).

Scheme 1 The synthesis route of monomers (BATE, BATB, BATH, BATO and BATD).

Scheme 2 The synthesis route of polymers (TDC-BATE, TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD).

In the polymerization reaction, TDC was selected as one of the reagents. It not only produced highly viscous polymers, but also kept the resins’ good corrosion resistance. Actually, we tried a commercially available diacid chloride such as terephthalic acid chloride (TPC) or isophthaloyl chloride (IPC) as the monomer, but we found that the resultant polymer molecular weight was too small when used with TPC and IPC as the monomer because of their poor solubility in solvents. The intrinsic viscosities of these semi-aromatic polyamides were obtained using H₂SO₄ as the solvent. They were in the range of 0.8–1.2 dL g⁻¹ (as shown in Table 1). During the reaction procedure, we observed that TDC-BATE always dissolved in NMP. But the high molecular weight TDC-BATE could almost not be obtained. That may be due to the high rigidity of the diamine monomer BATE, but the exact reason for this needs further study. In addition, we observed that TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD gradually precipitate after reacting for about 3 h, 2.5 h, 1 h and 0.5 h. The solubility of the polymers became increasingly small with the addition of the number of methylene units in the polymeric main chain. With the increase of methylene units in the polymer chain, the flexibility of the resins became increasingly larger, then the corresponding polymer was found to tend to arrange regularly and crystallize. So the sample of TDC-BATD was the first one that precipitated from the polymerization solution.

Table 1 Intrinsic viscosity (η_int), thermal and separation membrane mechanical properties of TDC-BATB, TDC-BATH, TDC-BATO, TDC-BATD, PES and PVDF^a

Polymers	ηint (dL g^−1)	Tg (°C)	T5% (°C)	Char yield (%)	Tensile strength (separation membrane, MPa)	Elongation at break (%)
a —: not detected.
TDC-BATB	0.81	203	409	12.7	4.1	12.1
TDC-BATH	1.02	192	419	21.1	4.2	17.2
TDC-BATO	1.16	176	417	16.0	3.8	20.7
TDC-BATD	1.23	163	419	16.8	3.3	21.8
PVDF	—	—	—	—	0.4	52.7
PES	—	—	—	—	2.1	38.9

---

The FT-IR spectra of the polymers are shown in Fig. 3. Comparing with the FT-IR spectra of monomers, the characteristic absorption of amine near 3360 and 3230 cm⁻¹ disappeared, and appeared at a new absorption near 3430 and 1640 cm⁻¹ attributed to the amide group. This indicates that condensation polymerization between the amine group and acid chloride (TDC) has occurred. The resultant polymers can not be dissolved in common solvents except for concentrated H₂SO₄, so the NMR characterization of the polymers can not be supplied.

Fig. 3 The FT-IR spectra of polymers (TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD).

Aggregate structure of TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD

X-ray diffraction patterns of these resultant semi-aromatic polyamides which had been annealed at 230 °C for 6 h are shown in Fig. 4. As shown in Fig. 4, these polymers exhibited crystalline peaks at about 19–22°. This indicates these resultant resins were crystalline. From Fig. 4 it can be observed that the crystalline peaks of TDC-BATB and TDC-BATH were weaker than those of TDC-BATO and TDC-BATD. This result can be explained by the phenomenon that TDC-BATO and TDC-BATD precipitated more quickly than that of TDC-BATB and TDC-BATH during the synthetic process. And these results suggest that longer aliphatic chains in the polymer main chain are beneficial to the crystallinity of these semi-aromatic polyamides.

Fig. 4 XRD profiles of TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD.

Thermal properties of polymers (TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD)

The thermal properties of these polymers were examined by DSC and TGA (as shown in the ESI, Fig. S2 and S3†). The T_g values of the resultant polymers were in the range of 163–203 °C (as shown in Table 1). It is about 30–60 °C higher than that of traditional semi-aromatic polyamides such as PA6T. It was expected that these resins should have melting points as the XRD results indicated that they were crystalline resins. But from the DSC curves, the melting endothermic peaks of these polymers could not be observed. The main reason is that a high density of hydrogen bonding results in a high melting temperature (may well surpass 350 °C) of these semi-aromatic polyamides, then their melting temperature is much higher than their degradation temperature. As shown in Fig. S3,† the initial degradation temperatures (T_5%) of TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD in nitrogen were in the range of 409–419 °C, the initial degradation temperatures of these resultant polyamides were also similar. We also found that these samples provided about 12.7–21.1% char yield at 800 °C in nitrogen (it was much higher than that of PA6T, PA9T). In this series of polymers, TDC-BATB contains a larger content of aromatic units than that of TDC-BATH, TDC-BATO and TDC-BATD. In theory, it should be the most stable resin, while TDC-BATB was found to have a lower char yield than other samples. The main reason was that TDC-BATB has a smaller molecular weight than the other samples. Combining the results of DSC, it suggests that these resultant semi-aromatic polyamides maintain good thermal properties whilst incorporating thioether linkages and aliphatic units (methylene group) into the polymeric backbone.

Membrane preparation and SEM analysis of these separation membranes

The prepared semi-aromatic polyamides were insoluble at room temperature in any solvent except for concentrated sulfuric acid. So in this study concentrated sulfuric acid was taken to as the solvent for the separation membranes. The solutions were prepared by dissolving the resins into concentrated sulfuric acid at concentrations of 8% (wt%). We prepared the membranes by using a knife on the resin’s solutions cast onto the glass plates, then the glass plates were immersed into a water bath at room temperature. It was found that the membranes which had formed in the water bath were peeled off from the glass plates gradually. We immersed the membranes into another water bath. The water in the water bath should be changed regularly until the pH value of the water bath is nearly 7. A series of separation membranes were prepared. The thickness of the membranes was in the range of 130 to 200 μm. The conference sample membranes (PES and PVDF) were prepared by a similar method. While the solvent was NMP and the solution concentration is 16 wt% (because the obtained membrane (solution concentration: 8 wt%) was too weak, we prepared the comparative sample solution at 16 wt%). The cross section morphology of the resultant separation membranes were examined for visualization by SEM. As shown in Fig. 5, the cross section morphology of the separation membrane displayed the asymmetric structure with a dense top layer and a porous sub-layer. It was found that the porous sub-layer consisted of fingerlike structures which acted as a support layer to withstand the external stress. The dense top layer acted as selective permeation layer. From Fig. 5, we can observe that with increasing length of the aliphatic chain in the polymer backbone, the thickness of the top layer increased gradually. In general, instantaneous demixing and delayed demixing would take place during the phase inversion process. With increasing the length of the aliphatic chain in the polymer backbone, the diffusion rates of both non-solvent to the sub-layer and solvent to non-solvent would be reduced to increase delayed demixing. The thickness of the top layer would increase and the macroporous structure near the bottom layer would increase in terms of number and size.

Fig. 5 SEM images of the resultant separation membranes (cross-section).

Fluxes and permeation of the resultant separation membranes

The fluxes of the resultant separation membranes, PES and PVDF were tested and are depicted in Fig. 6. As shown in Fig. 6, with increasing the length of the aliphatic chain in the polymer backbone, the fluxes would reduce gradually. It reduced from 371 L (m² h)⁻¹ to 89 L (m² h)⁻¹. With increasing the number of methylene units, the top layer of the membranes was denser and the thickness of the top layer increased gradually to result in a decrease of flux, which agreed with the results of the SEM images. Compared with the conference samples, the fluxes of the resultant membranes were larger than PVDF while smaller than PES. While the changes of the resultant membranes’ retention rate (bovine serum albumin) was found to show a converse trend to that of the water flux (as shown in Fig. 7). The retention increased with increase in the number of methylene units in the polymer main chain.

Fig. 6 Effect of the aliphatic chain length in the polymer backbone on fluxes of the membranes and the fluxes of PES and PVDF.

Fig. 7 Effect of the aliphatic chain length in the polymer main chain on retention rate (bovine serum albumin) of the membranes.

Porosity of the resultant separation membranes

The porosity of the prepared separation membranes was measured and shown in Fig. 8. From Fig. 8, with increasing length of the aliphatic chain in the polymer backbone, the porosity of the membranes would reduce gradually. Much more flexible aliphatic units in the polymer main chain enhanced the entanglement of the polymer molecular chain, and that would result in a decrease inn the membrane porosity. In addition, the introducing more aliphatic units may lead to a decrease in the hydrophilic properties of the resin. Combined with both aspects, the membrane solution tended to occur delay demixing with increasing length of aliphatic chain in the polymer backbone, and then the membranes would form a much more compacted structure consistent with the SEM and water flux results.

Fig. 8 Effect of the aliphatic chain length on the polymer backbone on the membrane porosity.

Mechanic properties of the resultant resin separation membranes

Because the resultant semi-aromatic polyamides’ melting temperature surpasses their degradation temperature, we can not obtain the tensile strength of the pure resin. But we tested the average tensile strength of the separation membranes (solution concentration: 8 wt%) of the resultant semi-aromatic polymers (TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD) and the reference sample PES and PVDF, these are summarized in Table 1. As shown in Table 1, the average tensile strengths of these samples were 4.1, 4.2, 3.8, 3.3, 2.1 and 0.4 MPa, respectively. The elongation at break was in the range of 12.1–21.8%, 38.9% and 52.7%, respectively. From the results, we find that as the aliphatic chain becomes increasingly long, the resultant membrane has a much smaller tensile strength and a larger elongation at break. This indicates that a long aliphatic chain is beneficial for membrane flexibility. Also we tested the tensile strength of the comparative sample membranes of PES and PVDF (solution concentration: 16 wt%); they were 2.1 and 0.4 MPa, respectively. Combining the results, it can be concluded that the prepared membranes have much better mechanical properties compared to those of the commercial products.

Corrosion resistance properties of the prepared semi-aromatic polyamides and resultant separation membranes

The corrosion resistance of TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD is summarized in Table 2. Comparing with traditional membrane materials such as PES and PVDF, the resultant polymers show a relatively excellent corrosion resistance in polar organic solvents. PES and PVDF dissolved in some strongly polar solvents such as DMF, DMAC, NMP, concentrated sulfuric acid and so on, while TDC-BATB, TDC-BATH, TDC-BATO and TDC-BATD had almost no change in the above solvents. Interestingly, we found that the resultant semi-aromatic polyamides showed an excellent stability in strongly alkaline solution such as a solution of NaOH (2 mol L⁻¹) and concentrated aqueous ammonia. From Fig. 9, it can be observed that the prepared membranes exhibited almost no change after being kept for about 6 months in a solution of NaOH, while the membrane of the PASS turned into small pieces. We also investigated the tensile strength of these samples. As shown in Fig. 10, the untreated membrane samples’ tensile strength is in the range of 3.3–4.2 MPa, indicating that these samples have a similar tensile strength to the commercial membrane materials. We also observed that the resultant polyamide membranes’ tensile strength decreased gradually on extending the treatment time in a solution of NaOH (2 mol L⁻¹). But the decreasing trend was mild, and the total change was about 30%. The tensile strength remained at 2.0–3.0 MPa even in those samples treated in concentrated base solution for 6 months whilst the membrane of PASS turned into small pieces and had almost no tensile strength when kept in the solution of NaOH (2 mol L⁻¹) for one week only. Combined with the above experimental results, we can confidently conclude that the resultant semi-aromatic polyamides have a better solvent resistance than commercial membrane materials such as PVDF and PES (in organic solvents) and PASS (in alkaline solution).

Table 2 Corrosion resistance behavior of the resultant semi-aromatic polyamides^a

Solvents	Polymers
	TDC-BATB	TDC-BATH	TDC-BATO	TDC-BATD	PES	PVDF
a ++: soluble at room temperature; +: soluble at solvents boiling point; +−: swelling with heating; −: insoluble with heating.
Concentrated sulfuric acid	++	++	++	++	++	++
Formic acid	−	−	+−	+−	−	−
NMP	−	−	−	−	++	++
DMF	−	−	−	−	++	++
DMAC	−	−	−	−	++	++
Pyridine	−	−	−	−	−	−
Acetone	−	−	−	−	−	−
Chloroform	−	−	−	−	+−	−
DMSO	−	−	−	−	++	++
1,4-Dioxane	−	−	−	−	−	−
Toluene	−	−	−	−	−	−
NaOH (2 mol L^−1)	−	−	−	−	−	−
Concentrated aqueous ammonia	−	−	−	−	−	−
Phenol + tetrachloroethane	−	−	−	−	++	−

---

Fig. 9 The images of the resultant semi-aromatic polyamides’ and PASS membranes (the left sample: TDC-BATB, the second sample: TDC-BATH, the third sample: TDC-BATO, the right sample: TDC-BATD) kept for different times in NaOH (2 mol L⁻¹) and concentrated aqueous ammonia.

Fig. 10 The tensile strength of the resultant semi-aromatic polyamides’ membranes kept for different times (7 d, 30 d, 90 d, 180 d) in a solution of NaOH (2 mol L⁻¹).

Conclusions

Four kinds of semi-aromatic diamine (BATB, BATH, BATO and BATD) monomers containing flexible thioether units were synthesized via a nucleophilic substitution reaction. They were reacted with TDC by electrophilic polycondensation to prepare a series of semi-aromatic polyamides with a high glass and thermal degradation temperature. The resultant polyamides were dissolved in concentrated sulfuric acid yielding separation membranes with a flux of 89–371 L m⁻² h⁻¹ and a retention rate of 87–94.2%. Interestingly, the introduction of flexible thioether units was beneficial for improving the polymers’ molecular weight.

The solvent resistance of these resultant polyamide membranes is much better than that of commercial products such as PVDF and PES. The samples are insoluble in strongly polar organic solvents such as NMP, DMF, etc. and exhibit almost no change in concentrated alkaline solution. This approach provides a convenient method for preparation of chemically and thermally stable polymer membranes with excellent corrosion resistance. Thus, they should be good potential candidates for membrane materials for industrial filtration especially in harsh environments (high operating temperature and strongly alkaline conditions), although the pure water flux is in need of improvement.

Acknowledgements

This work was supported by research grants from Outstanding Young Scholars Fund of Sichuan University (Grant No. 2015SCU04A25) and the Youth Fund Natural Science Foundation of China (21304060).

References

1. Z. Ademovic, D. Klee, P. Kingshott, R. Kaufmann and H. Hocker, Biomol. Eng., 2002, 19, 177–182.
2. K. Katsoufidou, S. G. Yiantsios and A. J. Karabela, J. Membr. Sci., 2005, 266, 40–50.
3. W. Zhao, Y. L. Su, C. Li, Q. Shi, X. Ning and Z. Y. Jiang, J. Membr. Sci., 2008, 318, 405–412.
4. M. E. Rogers, and T. E. Long, Synthetic Methods in Step-Growth Polymers, Wiley-Interscience, New York, 2003, pp. 327–375.
5. G. M. Geise, D. R. Paul and B. D. Freeman, Prog. Polym. Sci., 2014, 39, 1–42.
6. G. M. Geise, H. S. Lee, D. J. Miller, B. D. Freeman, J. E. McGrath and D. R. Paul, J. Polym. Sci., Part B: Polym. Phys., 2010, 48, 1685–1718.
7. D. M. Tigelaar, E. Palkera, C. M. Jackson, K. M. Anderson, J. S. Wainrigh and F. Robert, Macromolecules, 2009, 42, 1888–1896.
8. Q. F. Zhang, S. B. Zhang, L. Dai and X. S. Chen, J. Membr. Sci., 2010, 349, 217–224.
9. G. P. Yu, C. Liu, H. X. Zhou, J. Y. Wang, E. C. Lin and X. G. Jian, Polymer, 2009, 50, 4520–4528.
10. Y. Song, J. Y. Wang, G. H. Li, Q. M. Sun, X. G. Jian, J. Teng and H. B. Zhang, Polymer, 2008, 49, 724–731.
11. G. P. Yu, C. Liu, B. Li, L. W. Wang, J. Y. Wang and X. G. Jian, J. Polym. Res., 2012, 19, 9829,  DOI:10.1007/s10965-012-9829-1.
12. C. Yan, S. H. Zhang, C. Liu, D. L. Yang, F. J. Yang and X. G. Jian, J. Appl. Polym. Sci., 2009, 113, 1389–1397.
13. D. W. Shin, S. Y. Lee, N. R. Kang, K. H. Lee, M. D. Guiver and Y. M. Lee, Macromolecules, 2013, 46, 3452–3460.
14. F. A. Bottino, A. R. Cinquegrani, G. D. Pasquale, L. Lucrezia and A. Pollicino, Eur. Polym. J., 2003, 39, 2203–2208.
15. E. Montoneri, L. Giuffré, G. Modica and M. Gennuso, J. Mater. Sci., 1983, 18, 3311–3322.
16. H. Qin, C. C. Sun, C. He, D. S. Wang, C. Cheng, S. Q. Nie, S. D. Sun and C. S. Zhao, J. Membr. Sci., 2014, 468, 172–183.
17. W. F. Zhao, Q. B. Mou, X. X. Zhang, J. Y. Shi, S. D. Sun and C. S. Zhao, Eur. Polym. J., 2013, 49, 738–751.
18. C. S. Zhao, J. M. Xue, F. Ran and S. D. Sun, Prog. Mater. Sci., 2013, 58, 76–150.
19. Z. M. Wei, Q. C. Zhang, L. H. Wang, M. L. Peng, X. J. Wang, S. R. Long and J. Yang, Colloid Polym. Sci., 2013, 291, 1293–1296.
20. Z. M. Wei, M. L. Peng, F. Qiu, X. J. Wang, S. R. Long and J. Yang, RSC Adv., 2015, 5, 42305–42310.
21. A. Mamo, A. Aureliano, S. Battiato, G. Cicala, F. Samperi, A. Scamporrino and A. Recca, Polymer, 2010, 51, 2972–2983.
22. Z. Wu, Process for preparing copolymer containing biphenyl polyether ether-sulfone and poly(ether sulfone), China Patent CN1844196A, 2006.
23. L. Q. Feng and O. J. Iroh, Prog. Org. Coat., 2014, 77, 590–599.
24. G. Zhang, S. S. Yuan, Z. M. Li, S. R. Long and J. Yang, RSC Adv., 2014, 4, 23191–23201.
25. S. S. Yuan, J. Wang, X. J. Wang, S. R. Long, G. Zhang and J. Yang, Polym. Eng. Sci., 2015, 55, 2829–2837.
26. S. X. Xu, J. Yang, S. R. Long, Y. R. Chen and G. X. Li, Polym. Bull., 2005, 54, 251–261.
27. Y. Liu, A. Bhatnagar, Q. Ji, J. S. Riffile, J. E. McGrath, J. F. Geibei and T. Kashiwagi, Polymer, 2000, 41, 5137–5146.
28. H. D. Wang, J. Yang, S. R. Long, X. J. Wang, Z. Yang and G. X. Li, Polym. Degrad. Stab., 2004, 83, 229–235.
29. J. Yang, H. D. Wang, S. X. Xu, G. X. Li and Y. J. Huang, J. Polym. Res., 2005, 12, 317–323.
30. G. Zhang, G. S. Huang, D. S. Li, X. J. Wang, S. R. Long and J. Yang, Ind. Eng. Chem. Res., 2011, 50, 7056–7064.
31. G. Zhang, D. S. Li, G. S. Huang, X. J. Wang, S. R. Long and J. Yang, React. Funct. Polym., 2011, 71, 775–781.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20717k

---