ECTFE membranes produced by non-toxic diluents for organic solvent filtration separation

Abstract

A new grade of ethylene-chlorotrifluoroethylene, low melting point HALAR® ECTFE (LMP ECTFE), was studied and used as a polymer for the preparation of solvent-resistant flat-sheet membranes. Among the different types of non-toxic solvents tested, di-ethyl adipate (DEA) was selected for preparing flat sheet membranes via thermally induced phase separation (TIPS). The morphology of the membranes has been analyzed by scanning electron microscopy (SEM) and atomic force microscopy (AFM). Dense and porous membranes have been obtained and characterized by contact angle, pore size and porosity tests. Porous membranes showed an asymmetric structure made of a denser top-side and a spherulitic porous structure on the bottom side. Membrane resistance was studied using the dense membrane in contact with most aggressive organic solvents, such as polar protic, polar aprotic and non-polar solvents. The results suggest that the newly developed LMP ECTFE membranes are very promising candidates for organic solvent separation. Ultrafiltration (UF) and nanofiltration (NF) tests with alcohols and di-methyl formamide (DMF) demonstrated their solvent separation potential.

---

1. Introduction

Organic solvents are usually employed in different production processes, such as chemical production, the pharmaceutical industry, the petrochemical sector, cosmetics, the purification and processing of food, nutraceuticals and natural products. Therefore, solvent recycling is one of the main issues facing the chemical and pharmaceutical industries; in fact, industrial waste may be toxic, corrosive or reactive, which can lead to environmental and human health consequences. The traditional practices of solvent recycling rely on pre-treatment (addition of additives), evaporation and distillation. However, these processes are costly, require high temperatures or use of other types of chemicals. With respect to these techniques, membranes allow the facile, safe and low-cost recovery, concentration or purification of the target molecules (non-thermal separation).¹ In particular, ultrafiltration (UF) and nanofiltration (NF) pressure-driven membrane processes are of particular interest for the organic solvent separation. The first publications concerning their application were reported by Nguyen² for UF and Eriksson³ for NF, respectively. Main examples of UF process include the fractionation and purification of peptide or impurities from protein solution⁴ and the extraction polyphenols from seeds.⁵ Organic solvent nanofiltration (OSN) has gained popularity as membrane process for different applications, such as purification of pharmaceutically active ingredients,⁶ specific recognition of genotoxins⁷ recovery of catalyst in chemical synthesis,⁸ separation of ionic liquids,⁹ and solvent exchange.¹⁰ Both UF and NF processes were used in pharmaceutical and biotechnological applications to extract, isolated and concentrated compounds of interest^11–15 from organic solvent media. For both types of membranes it is of primary importance the membranes stability in the presence of harsh solvents.

New generation of NF membranes is more stable towards organic solvents, but full-scale applications are still limited, because of the low number of available commercial solvent-resistant membranes. Nowadays, the typical polymers used for preparing NF membranes are polyimide (PI) including co-polyimides (co-PIs), polydimethylsiloxane (PDMS), polyacrilonitrile (PAN),¹⁶ polyamide (PA),¹⁷ polysulfone (PS).¹⁸ Table 1 reports examples of typical polymers used in UF and NF membranes preparation and applications.

Table 1 Examples of typical polymers used in organic solvent UF and NF membranes preparation and applications^a

Authors	Polymer type	Solvents tested	Permeability (L m^−2 h^−1 bar^−1)	Application of organic solvent resistant membranes
a PDMS: polydimethylsiloxane; PEEKWC: polyetheretherketone; PI: polyimide, TFC: thin-film composite membrane; DMF: di-methyl formamide; THF: tetrahydrofuran; NMP: N-methyl pyrrolidone; PAN: polyacrylonitrile; PPS: polyphenylsulfone, PES: polyethersulfone, PS: polysulfone, PVP: poly(vinylpyrrolidone), PASS: polyarylene sulfide sulfone, PA: polyamide.
S. Darvishmanesha et al.^19	PPS	Methyl ethyl ketone, diethyl ether, ethyl acetate, methanol, ethanol, 2-propanol, n-hexane, n-heptane, acetone and toluene	0.02–3.21	OSN membranes permeation performance in pure solvents at lab scale
S. Darvishmanesha et al.^20	PPS	Methanol	0.8–9	OSN membranes permeation performance in pure solvents at lab scale
M. G. Buonomenna et al.^21	PEEKWC	Water, methanol, ethanol, 2-propanol and n-butanol	0.27–3.63	OSN membrane performance in pure solvents at lab scale
I. Soroko et al.^22	PI	DMF	0.23–11	OSN membrane performance in pure solvents at lab scale
M. Peyravi et al.^23	Thin film composite (TFC) membrane of PA and PS	Methanol, ethyl acetate and n-hexane	1.4–7.4	OSN membrane performance in pure solvents at lab scale
D. Fritscha et al.^24	TFC membrane of polymers of intrinsic microporosity (PIMs) and PAN	n-Heptane, toluene, chloroform, tetrahydrofuran, methanol and ethanol	0.1–7.3	OSN membrane performance in pure solvents at lab scale
Jansen et al.^25	PPS and PI	Methanol, ethanol, 2-propanol, pentane, hexane, heptane, acetone, methyl ethyl ketone, methylacetate, ethylacetate, isopropyl acetate	0.5 (NF)–2000 (MF)	OSN membrane performance in pure solvents at lab scale
K. Hendrix et al.^26	TBPEEK	Alcohol, alkanes, alkylacetate and ketone	0.09–0.77	OSN membrane performance in pure solvents at lab scale
M. F. J. Solomon et al.^27	PI TFC	THF, toluene and ethyl acetate	0. 3–3.83	OSN membrane performance in pure solvents at lab scale
L. Liu et al.^28	PASS	Water, ethanol, methanol, n-butanol	0.5–1.24	OSN membrane performance in pure solvents at lab scale
F. M. Penha et al.^29	PES and hydrophilic PES	Water, ethanol, 2-propanol and n-hexane	0.3 (NF)–250.3 (UF)	UF–NF membrane performance in pure solvents at lab scale
F. M. Penha et al.^30	PES and hydrophilic PES	Oil and hexane	0.1–2.5	NF permeation of oil/hexane mixture at lab scale
M. Saxena et al.^31	PS	Hexane	79–364	UF membrane performance in pure solvents at lab scale
M. V. Tres et al.^32	PES/PVP	Oil and hexane	2 (NF)–27.5 (UF)	UF–NF separation of soybean oil/n-butane at lab scale
M. E. Tsui et al.^11	PAN-based	Aqueous solutions of ethanol	0.45–2.34	NF treatment of ethanol extracts of corn at lab scale
H. Nawaz et al.^12	Cellulose acetate/cellulose nitrate mixed esters	Aqueous solutions of ethanol	—	Extraction and concentration of polyphenols at lab scale

---

The influence of the polymeric materials, in terms of separation performance, was well summarized in the paper of Cheng et al.³³ Other authors showed that, beside the polymer type, also the selected solvent and eventual additives influence the performance of the solvent resistant nanofiltration (SRNF) membranes.³⁴ A suitable solvent-resistant materials, as Halar® ECTFE (ethylene-chlorotrifluoroethylene), a perfectly alternating copolymer of ethylene and chlorotrifluoroethylene, could be used in the chemical process industry due to its properties such as excellent chemical resistance and mechanical properties.^35,36 However, due to its chemical–physical stability, ECTFE is difficult to be processed with the conventional membrane manufacturing techniques. In fact, ECTFE membranes are usually prepared via thermally induced phase separation (TIPS) technique and the polymer is solubilized in organic solvents at high temperature. The patent of Mutoh and Miura was the first one reporting successful TIPS casting of ECTFE.³⁷ Several studies were performed to identify high-boiling organic solvents able to dissolve ECTFE.³⁸ Among them, tri-ethyl citrate (CTF), glycerol triacetate (GTA) and di-octyl adipate (DOA) have been used in the ECTFE membrane formation thanks to their low toxicity in comparison than phthalates, such as di-butyl phthalate (DBP) or di-ethyl phthalate (DEP).^39–43 However, in all these cases, the ECTFE polymer was solubilised at temperature over 180 °C. Solvay Specialty Polymers has recently developed a lower melting point grade of Halar® ECTFE, named here LMP ECTFE, which still offers the above quoted outstanding chemical resistance in caustic environments. Moreover, it shows comparable properties with standard Halar® (hydrophobicity and mechanical properties), but lower crystallinity and lower melting point⁴⁴. ECTFE Halar® 901 and LMP ECTFE properties are summarized in Table 2. Molecular weight (MW) of ECTFE polymers cannot be directly determined by GPC. However, since the Melt Flow Index (MFI) can be provided and it is proportional to the molecular weight, information on their MW could be indirectly obtained.⁴⁵ For Halar 901, the average Melt Flow Index (MFI) at 275 °C (527 °F), under a load of 2.16 kg, is 0.1 g min⁻¹; for LMP ECTFE, the MFI, at 225 °C following ASTM D1238, is 0.15 g min⁻¹. On the basis of this parameter, it can be concluded that LMP ECTFE has lower MW and lower viscosity with respect to Halar 901. The same issue is related to the ethylene content, which is also not given by the supplier but it can be identified by other polymer properties. In fact, it is possible to obtain information on the melting point and the heat of fusion. For Halar 901 the melting point is 242 °C, while the heat of fusion is 42 J g⁻¹. For LMP ECTFE the melting point is 175–185 °C, while the heat of fusion is 18 J g⁻¹. These parameters are connected to the ethylene content. Since they are lower with respect to the standard Halar, in which the ethylene/chloro-trifluoroethylene ratio is 1 : 1; it could be concluded that, for LMP ECTFE, the ethylene content is less than 50 molar%.

Table 2 ECTFE Halar® 901 and LMP ECTFE properties

	Melting point Tm (°C)	Heat of fusion (J g^−1)	Tensile modulus (MPa)	Contact angle (°)	Melt flow index (g/10 min, @2.16 kg)
a At 275 °C following ASTM D 1238 “Standard Test Method for Melt Flow Rates of thermoplastics by Extrusion Plastometer”.b At 225 °C following ASTM D1238.
ECTFE Halar® 901	242	42	1650	90–95	1^a
LMP ECTFE	175–185	18	1100	90–95	1.5^b

---

In this work, LMP ECTFE flat-sheet membranes were prepared by means of TIPS. In this perspective, solvents with low toxicity were employed. In particular, the first part of this investigation focused on the study of different non-toxic solvents as possible diluents selected on the basis of their environmental impact, high boiling point and solubility towards the polymer used. Then, LMP ECTFE flat membranes were prepared and characterized in terms of morphology, contact angle, mechanical test, porosity and pore size. In particular, swelling tests in pure organic solvents as methanol, ethanol, 2-propanol, hexane, cyclohexane, tetrahydrofuran, toluene, N-methyl pyrrolidone, di-methyl acetamide, di-methyl formamide were carried out for evaluating the LMP ECTFE dense film resistance. Finally, organic solvent permeation tests on selected solvents were performed using the novel LMP ECTFE membranes produced.

2. Experimental

2.1 Materials

The ECTFE based polymers (experimental ECTFE Halar® 901 and LMP ECTFE) were kindly supplied by Solvay Specialty Polymers and used without any further purification. Di-ethyl adipate (DEA), ethanol (EtOH), 2-propanol (IPA), methanol (MetOH), acetone, tetrahydrofuran (THF), toluene (Tol), N-methyl pyrrolidone (NMP), di-ethylene glycol (DEG), di-butyl itaconate (DBI), glycerol, chloroform, di-methyl acetamide (DMA), di-methyl formamide (DMF), cyclohexane (C₆H₁₂), hexane, Fluorinert® FC-40, kerosene oil, were all purchased from Sigma-Aldrich and used without any further purification. Liquid nitrogen was purchased from Pirossigeno (Cosenza, Italy).

2.2 LMP ECTFE solubility tests

The solubility tests were carried out using different types of solvents: DBI and DEA. Their chemical structure is reported in Fig. 1.

Fig. 1 Chemical structure of solvents of interest.

Solvents were selected on the basis of their high boiling point, solubility parameters, lower toxicity and environmental impact, compared with the phthalates, 1,3,5-trichlorobenzene (TCB), di-butyl phthalate (DBP) and di-octyl phthalate (DOP),³⁷ which are the solvents usually used for the preparation of ECTFE-based membrane by TIPS (Table 3). Similar solubility parameters indicate good affinity between solvent and polymer. In this case would be expected the completely dissolution of the polymer, whilst those with dissimilar values would not.

Table 3 Solvents properties

Solvent	Molecular formula	Molar mass (g mol^−1)	Density (g cm^−3)	Boiling point (°C)	δd (MPa)^1/2	δp (MPa)^1/2	δ (MPa)^1/2
DEA	C10H18O4	202	1.01	251	16.4	6.2	7.5
DBI	C13H22O4	242	0.98	284	16.9	10	22
ECTFE Halar® 901					16.8	8.4	7.8

---

Solubility tests were carried out heating and magnetically stirring (50 rpm) the polymeric solution (15 wt% LMP ECTFE–85 wt% solvent) in an oil bath. The polymer solubilisation was evaluated by increasing the temperature of 10 °C each 30 min, from room to a maximum temperature value, close to the boiling point of the solvent employed.

In particular, an homogeneous solution (15 wt% LMP ECTFE/85 wt% solvent) was observed at 140 °C for DEA and 170 °C for DBI. The possibility of decreasing drastically the temperature of polymer solution makes easier the polymer processability. Based on these results and considering also the low toxicity, DEA was selected as solvent for the preparation of LMP ECTFE flat sheet membranes.

2.3 Polymeric dope solution preparation

Polymeric dopes were prepared by dissolving the polymer in DEA at different concentrations (15–20–25% w/w). Each solution was stirred for 1 h at temperature of 193 °C until complete dissolution of the components was achieved. The polymeric dope was allowed to degas, keeping the temperature, for 6 h before casting.

2.4 Preparation of LMP ECTFE membranes and dense films

Membranes were prepared by casting the polymeric solution over a suitable smooth glass support by means of an automatized casting knife (DeltaE srl, Italy) Fig. 2.

Fig. 2 Automatized casting knife.

The dope solution, having a polymer concentration in the range of 15–25 wt%, was cast by keeping both the casting knife and the support to the temperature of 193 °C, to prevent premature precipitation of the polymer.

After casting and evaporation, polymeric membranes were cooled down by immersing them in the coagulation bath of pure di-ethylene glycol (DEG) at 5 °C. This value of temperature, among the others used (25 °C and 60 °C), has been found optimal for the preparation of reproducible polymeric porous membranes. After coagulation, the membranes were washed overnight in 2-propanol. In case of preparation of dense film, specifically made for evaluating the solvent resistant of the LMP ECTFE, no coagulation bath was used and the cast dope solutions were cool down overnight slowly. In the dense film (D), DEA was extract by washing in ethanol (typical step in a TIPS process), three times and drying in an oven for 6 h.

All the membrane conditions are resumed in Table 4.

Table 4 Summary of the main preparation conditions for LMP ECTFE membranes

Membrane type	Polymer conc. (wt%)	Casting knife (μm)	Air exposure time (s)	Coagulation bath type	Coagulation bath temperature (°C)	Drying procedure
L2	15	300	0	DEG	5	Isopropanol over night and after drying at air
M2	20	300	0
N2	25	300	0
Dense film (D)	20	250	Over night	—	—	Direct drying

---

2.5 Determination of the binary phase diagram

ECTFE Halar® 901 and LMP ECTFE solubility in DEA was evaluated by monitoring the cloud point (CP) of solutions containing different weight percentages of polymer. In each case, the polymer was added to the solvent at room temperature and solubilised by magnetically stirring the solution (50 rpm) increasing the temperature by using the oil bath. Once the polymeric solution was completely solubilised, the dope solution was cooled (rate of cooling was 0.1 °C min⁻¹) until the cloud point (solid–liquid demixing) occurred. The tests were performed varying the polymer concentration from 5 wt% to 35 wt%.

2.6 Membrane characterization

2.6.1 Scanning electron microscopy (SEM). Membranes morphology (cross section, top and bottom side) was observed by using a scanning electron microscope (Zeiss EVO MA 100, Assing, Italy). The sample for the evaluation of the membrane cross-section was fractured in liquid nitrogen. Samples were sputter-coated with a thin gold film prior to SEM observation.

2.6.2 Atomic force microscopy (AFM). Atomic force microscopy was used to study the top and bottom surface morphology and roughness of the prepared membranes. The AFM device was a Bruker Multimode 8 with Nanoscope V controller. Data were acquired in tapping mode, using silicon cantilevers (model TAP150, Bruker). The membrane surfaces were imaged in a scan size of 10 μm × 10 μm.

2.6.3 Contact angle measurements. Contact angle measurements were performed with ultrapure water using the sessile drop method by a CAM100 instrument. For all membranes, at least five measurements were taken both on the air and the glass sides; the average values and the corresponding standard deviation were then calculated.

2.6.4 Swelling tests. According to Standard Practices for Evaluating the Resistance of Plastics to Chemical Reagents,⁴⁶ to measure the swelling degree, dense film samples (A = 4 cm²) were weighted and placed in suitable solvent resistant containers. The quantity of reagent shall be approximately 12.5 mL cm⁻² of specimen surface area. Samples were kept totally immersed, for 72 h–120 h–172 h at standard laboratory atmosphere (25 °C), in one of the following pure organic solvents: EtOH, MetOH, acetone, THF, Tol, NMP, EtOAc, DMA and DMF. The swelling degree (S_w) was calculated as follows:
where W_W is the weight of the dense film after 72 h–96 h–172 h of immersion and W_D is the initial weight of the dry dense film.

2.6.5 Mechanical tests. The Young's or elastic modulus (E_mod), the tensile stress at break (R_m) and breaking elongation or stress at break (e_Break) were measured by means of a ZWICK/ROELL Z 2.5 test unit. Each sample was stretched unidirectionally at a constant rate of 5 mm min⁻¹; the initial distance between the clamps was of 50 mm. Five specimens were tested for each sample.

2.6.6 Porosity. Membrane porosity (ε_m) was determined according to the gravimetric method, described in literature.⁴⁷ Porosity was defined as the ratio between the volume of the membrane and the volume of voids present within it. Dry membrane pieces were weighted and impregnated in kerosene for 24 h; after this time, the excess of liquid was removed with tissue paper, and membranes weight was measured again. Finally, porosity was calculated applying the following formula:
where W_W is the weight of the wet membrane, W_D is the weight of the dry membrane, ρ_K is the kerosene density (0.82 g cm⁻³) and ρ_P is the polymer density (1.71 g cm⁻³). For all membranes types, three measurements were performed; then, the average values and corresponding standard deviation were calculated.

2.6.7 Bubble point and pore size distribution. Membrane bubble points, largest pore size and pore size distribution were determined using a PMI Capillary Flow Porometer (CFP1500 AEXL, Porous Materials Inc., USA). For each test, membranes samples were initially fully wetted using Fluorinert FC-40 (16 dyne per cm), for 24 h and placed in the sample holder. Bubble point, gas pressure and flow rates through the dry membranes were measured. This operating mode, named wet-up/dry-up, was selected using the software Capwin. The measurement of bubble point, largest pore size and pore size distribution is based on the Laplace's equation:
where d_P is the pore diameter, τ is the surface tension of the liquid, θ is the contact angle of the liquid (assumed to be 0 in case of full wetting, which means cos θ = 1) and P is the external pressure. The results of each test were exported as an excel file using the software Caprep for further processing.

2.6.8 Solvent filtration experiments. Filtration tests were performed in a high pressure crossflow filtration cell (model HP4750) supplied from Sterlitech corporation (USA). The volume was 300 mL and the diameter 5.1 cm. The effective membrane area was 20.4 cm².

Experiments were performed using the solvents at room temperature, as indicated in Table 5.

Table 5 Characteristics of the organic solvents used in the filtration experiments⁴⁸

Solvent	Molecular weight (g mol^−1)	Density (g mL^−1)	Surface tension γ (mN m^−1)	Viscosity (cP)	Polarity
Methanol	32.04	0.791	22.1	0.60	Protic polar
Ethanol	46.10	0.789	21.9	1.20	Protic polar
DMF	73.09	0.944	37.1	0.82	Aprotic polar

---

The cell was filled with one of the following solvents: methanol, ethanol, and DMF. Before the tests, each membrane was conditioned by immersing in the target pure solvent for 24 h, and then placed in the cell. Experiments were carried out by applying different N₂ gas pressures (trans-membrane pressure (TMP)) from 2 to 10 bar. The permeate was collected at atmospheric pressure. Each membrane filtration test was conducted three times. Solvent flux (J) through each membrane, at a given pressure, was defined as the volume permeated per unit area and per unit time. J was calculated by the following equation:

where V (L) is the volume of permeate, A (m²) is the membrane area, and Δt (h) is the operation time. The average and relative standard deviation were calculated.

From the slope of the J vs. P linear relationship, the membrane permeability was calculated according to a least-square fitting method.

3. Results and discussion

3.1 Determination of the binary phase diagram

The cloud point (CP) of LMP ECTFE/DEA and ECTFE Halar® 901 system, as a function of polymer concentration, was determined. The initial thermodynamically stable homogeneous solution, made of polymer and solvent, separates into two phases decreasing the temperature. The polymer-rich phase forms the membrane structure, and the polymer-lean phase forms the pores.⁴⁹ The binary phase diagram indicates the miscibility gap of the solution, at different polymer concentration. In general, the CP of a polymer/solvent system depends on its stability, which in turn is influenced by the solubility of the polymer in the same solvent. This depends on the Hansen's solubility parameters, but also on the polymer degree of crystallinity. Since, usually, a crystalline polymer is more stable, the interactions between the chains are stronger and, therefore, it is more difficult to dissolve. As reported in Fig. 3, LMP ECTFE CP was lower that the ECTFE Halar® 901. In this case, the two polymers' solubility parameters are very close. However, LMP ECTFE is easier to dissolve because of its lower crystallinity (Table 2) and therefore, the compatibility of polymer/solvent is higher for LMP ECTFE/DEA. Moreover, it was observed that the CP increased at higher polymer concentration. Phase separation mechanism usually influences the membrane morphology. In fact, the liquid–liquid (L/L) demixing is favoured at low temperature and low polymer concentration, leading to cellular morphologies, while solid–liquid (S/L) demixing generally occurred at high polymer concentration and high temperature and brings to the formation of spherulites and axialites structures.⁴⁹ In particular, both our systems, LMP ECTFE/DEA and ECTFE Halar 901/DEA, did not become cloudy, until they began to form gels at the sol–gel transition temperature. Similar results were observed for PVDF/Citroflex system reported by Sawada et al.⁵⁰ This result is in agreement with TIPS processes, where the higher temperature is necessary to keep the polymer/solvent system homogeneous (latent solvent). According to literature,^38,41 and SEM observation, S/L phase separation, typical in a TIPS process, is observed in our experiments. The obtained membranes morphology is discussed more in details in Section 3.2.

Fig. 3 Sol–gel transition temperature of ECTFE Halar® 901 and LMP ECTFE solutions as a function of polymer concentration.

3.2 SEM, AFM and contact angle analyses

The SEM pictures of the produced LMP ECTFE membranes are shown in Fig. 4. Membrane D shows a fully dense structure, whilst L2, M2, and N2 membranes are asymmetric. In the latter, the top-side (airside during casting) is dense while the bottom side (glass side) is porous. This is in agreement with previous literature work. Moreover, it can be clearly seen that spherulitic structures are well defined in all the prepared membranes. This morphology is due to S/L demixing, which takes place during film cooling, as reported in Section 3.1. As described in literature,⁴⁹ when membranes are prepared via TIPS, S–L phase separation can take place only if, during solution cooling, the crystallization temperature of the polymer is reached. Semi-crystalline and crystalline polymers can give, then, rise to chain folded lamellae and supramolecular architectures as axialites and spherulites.

Fig. 4 SEM pictures of membranes D, L2, M2 and N2.

The roughness of the top and bottom surfaces was investigated using AFM and only for the membranes having the intermediate polymer concentration and the best drying procedure. AFM, with respect to SEM imaging, provides quantitative information on the sample topography. Fig. 5a and b show respectively the top and bottom surfaces topography of the membrane M2, while Fig. 5c and d show the correspondent 3D views.

Fig. 5 AFM pictures of membranes M2 where (a) and (b) show respectively the top and bottom surfaces topography of the membrane, while (c) and (d) show the correspondent 3D views.

Topography was measured on five different areas on the sample surface and the RMS roughness (with its standard deviation) was calculated. The RMS roughness for the top-side of the membrane is found to be 2.79 ± 1.01 nm while the RMS roughness for the bottom part is 590 ± 160 nm. The obtained results indicate that the bottom surface has a much higher roughness with respect to the top surface in agreement with SEM analysis. In fact, the bottom side of the membrane presents a porous structure that contributes to the higher roughness.

These values are useful for the interpretation of the contact angle measurements which are performed to quantify the membranes hydrophobic properties, on both membrane surfaces. The average values are reported in Table 7. The relative standard error is less than 5% in all cases. Contact angle measurements also confirmed that the produced membranes are, in general, asymmetric with top-side more hydrophilic than bottom side; due to surface smoothness (denser skin layer) as shown in AFM and SEM results. In fact, the apparent contact angle of a sessile droplet changes not only with the chemical texture, determined by the composition of the polymeric solution, but also with the roughness. Wenzel et al.⁵¹ suggested a phenomenological model for understanding how roughness affects wetting:

cos θ* = cos θ × r

where θ is the Young's angle and r the surface roughness.

Table 6 Hansen solubility parameters and surface tension of the solvents used for swelling test^a

Solvent	Solubility parameters (MPa)^1/2				Surface tension@20 °C in mN m^−1
	δh (MPa)^1/2	δd (MPa)^1/2	δp (MPa)^1/2	δ (MPa)^1/2
a δh, δd and δp are the solubility parameters related to hydrogen bonding, dispersion parameters, and polar forces, respectively.
Ethanol	9.5	7.73	4.3	12.92	22.1
Methanol	10.9	7.42	6	14.28	22.7
2-Propanol	8.5	7.75	3.3	11.97	23
Hexane	0	7.24	0	7.24	18.43
Cyclohexane	0	8.18	0	8.18	24.95
Acetone	6.9	15.5	10.4	19.9	25.2
DMA	11.8	17.8	14.1	22.7	36.7
DMF	11.3	17.4	13.7	24.8	37.1
THF	3.7	19.0	10.2	22.5	26.4
NMP	7.2	18.4	12.3	22.9	40.79
Toluene	2.0	18.0	1.4	18.2	28.4

---

Table 7 Properties of the LMP ECTFE membranes^a

Membrane	Contact angle (°)		Mechanical tests			Porosity (%)	Pore size measurements
	Top side	Bottom side	Emod (N mm^−2)	Tensile-strength (N mm^−2)	eBreak (%)		Bubble point (bar)	Largest pore size (μm)	Mean flow pore diameter (μm)	Diameter at maximum pore size distribution (μm)
a The relative standard error is less than 5% in all cases.
L2	98	137	40.2	1.3	35.1	69.5	2.08	0.22	0.03	0.03
M2	98	108	277.5	4.6	141.7	56	2.11	0.21	0.01	0.02
N2	105	114	370	17.1	155.4	42.3	2.43	0.18	0.01	0.01

---

Smooth surfaces reduce the absolute value of cos θ and for this reason the corresponding contact angle is lower.

3.3 Membrane properties

Swelling tests on the dense film (membrane D) were carried out in order to evaluate the membrane resistance to the most used solvents in chemical and pharmaceutical industries. According to the procedure, reported in the materials and method Section 2.6.4, membrane stability plays a crucial role in organic solvent separation, and the swelling test could be an important criteria to understand, predict, and describe the membrane performance. Swelling is a thermodynamic phenomenon which takes place in three different steps: solvent absorption on membrane surface, penetration/diffusion into the polymer free volume and, finally, polymer expansion.

Membrane swelling depends on the affinity between solvent and polymer. Indeed, if mutual affinity between them is higher, swelling will be enhanced. This may lead to an increase in the free volume (polymer expansion), which will affect membrane's morphology. In case of porous membranes, the solvent uptake is much more pronounced than swelling in dense membranes. This is because the solvent also filled their porous structure. Membrane swelling could reduce selectivity, increase solvent flux and reduce membrane cut-off. Fig. 6 indicates that swelling takes 120 hours to reach equilibrium.

Fig. 6 Swelling tests for membrane D (dense membrane).

Moreover, depending on the type of solvent employed (Table 6) a different behaviour is observed:

• polar protic solvents, the swelling increases in function of the surface tension (ethanol > methanol > propanol).

• non-polar solvents, swelling increases with the Hansen solubility parameter (THF > toluene > cyclohexane > hexane).

• polar aprotic solvents, the results indicate that the total degree of swelling is influenced by both superficial tension NMP > DMA > DMF and hydrogen bonding parameters acetone > NMP > DMA > DMF.

Comparing the swelling tests results to what observed by S. Simone et al.⁴¹ it can be noticed that: (a) the swelling percentages are, in general, lower, since the pure ECTFE Halar® 901 membranes prepared in that work were asymmetric dense, and, hence, solvent uptake by the porous polymer matrix was observed, (b) the pure ECTFE Halar® 901 membranes solvent uptake percentages increased in the following order: DMF < DMA < acetone < toluene < NMP < EtOH < MetOH < THF. In the case of LMP ECTFE, DMF presented the lowest swelling degree; while, acetone, THF and toluene produced the highest one swelling, (c) the swelling degree observed of DMF using LMP ECTFE dense film, was lower than that observed with DMA, and this is in agreement with the ECTFE Halar® 901 solvent-uptake reported in literature.⁴¹ A possible explanation could be that despite these two solvents are similar, some differences in their chemical structure and physical properties may lead to a different behaviour. In fact, DMA is polar and possess an electron lone pair, which makes it more reactive than DMF. Moreover, DMF readily interacts by hydrogen bonding with the polymer backbone enhancing the decreasing of swelling, while DMA forms H-bonds only at high temperature, as observed in literature for sulfonated polymers.⁵²

3.4 Mechanical tests, porosity and pore size characterisation

The tensile strength of the prepared membranes was measured in order to study how polymer concentration affected the mechanical properties and the results are summarised in Table 7. The elongation at break (e_Break) increases from 35.15% for membrane L2 (15 wt%), to 155.45% for membrane N2 (25 wt%). In addition, the elastic modulus and the tensile stress at break both increased when raising the polymer concentration. These results were in agreement with the porosity measurements. In fact, porosity decreased whilst all tensile properties (strength, modulus and extension at break) increased at higher concentration of LMP ECTFE as also reported in Table 7.

The mean pore size of the membranes is in the range 0.01–0.03 μm and it decreased at higher polymer concentration. Moreover, the bubble point was increased from 2.08 bar, for the membrane with 15 wt% of polymer (L2), to 2.43 bar, for the membrane with 25 wt% of polymer (N2) as shown in Table 5. This is in agreement with the corresponding largest pore size of 0.22 and 0.18 μm of the membranes L2 and N2, respectively.

3.5 Membrane filtration performance in organic solvents

The low degree of swelling for dense membranes in different solvents, as MetOH, EtOH, and DMF, suggested the possible use of the novel LMP ECTFE membranes for organic solvent filtration separation. The permeability of the membranes L2, M2 and N2 prepared using 15-20–25 wt% of polymer, respectively, was measured using the dead-end filtration cell described in Section 2.6.8. The trans-membrane flux was measured at different pressure values and the permeability was calculated as described in the materials and methods section. The results are shown in Fig. 7.

Fig. 7 Solvent permeability trends with three different organic solvents and the membranes L2 (a), M2 (b) and N2 (c).

In all the cases, the permeate flux increased linearly when increasing the pressure. In particular, ethanol flux was lower with respect to methanol flux. In agreement with literature,⁵³ this result can be justified considering the molecular weight and the viscosity of the two solvents (Table 5). In fact, the viscosity of the ethanol is twice of that one of methanol. Moreover, the molecular weight of ethanol is 46.10 g mol⁻¹ while the one of methanol is 32.09 g mol⁻¹.

A lower DMF flux was observed and it is due by the combination of different factors such as molecular weight (73.09 g mol⁻¹), the density (0.94 g mL⁻¹) and the superficial tension (37.1 mN m⁻¹). In fact, the higher molecular weight of DMF is responsible of an increase of the steric hindrance, while the higher superficial tension (37.1 mN m⁻¹), is responsible of the decrease of the organic solvent diffusion through the membranes. Moreover, the increase of the polymer content (from 15 wt% to 25 wt%) determines a decrease of the solvent permeability, as expected. This allows tailoring the produced membranes depending on the type of application. In fact, the permeability values were between UF and NF range for the membranes L2 (MetOH: 20 L m⁻² h⁻¹ bar⁻¹; EtOH: 13 L m⁻² h⁻¹ bar⁻¹; DMF: 10 L m⁻² h⁻¹ bar⁻¹), and M2 (MetOH: 15.5 L m⁻² h⁻¹ bar⁻¹; EtOH: 8.81 L m⁻² h⁻¹ bar⁻¹; DMF: 7.66 L m⁻² h⁻¹ bar⁻¹), and in the NF range for the membrane N2 (MetOH: 3.6 L m⁻² h⁻¹ bar⁻¹; EtOH: 3 L m⁻² h⁻¹ bar⁻¹; DMF: 2.9 L m⁻² h⁻¹ bar⁻¹), as reported in literature.⁵⁴

4. Conclusions

Novel LMP ECTFE dense and porous asymmetric flat membranes were prepared using a suitable non-toxic solvent as the di-ethyl adipate via TIPS technique. Polymeric membranes with different concentrations (from 15 wt% to 25 wt%) were successfully cast via TIPS technique. All the produced porous asymmetric membranes showed a spherulitic structure, with a top-side more hydrophilic than the bottom side, this result was confirmed by SEM and AFM analysis. Dense membrane shows good resistance to most aggressive organic solvents, as confirmed by the swelling tests carried out for 192 h. Porous asymmetric membranes were tested for organic solvent filtration using pure methanol, ethanol and DMF. Varying polymer concentration in dope solution, it is possible to tailor the produced membranes in the NF and UF range. The results obtained show that LMP ECTFE membranes are promising candidates to be use in separation filtration processes under harsh conditions. Moreover, this polymer can be used as alternative of Halar® 901, thanks to its property, which makes it easy processable at lower temperature keeping the material properties unchanged.

Acknowledgements

The authors gratefully acknowledge Solvay Specialty Polymers, Bollate (Italy) for financial support of the research performed.

References

1. J. P. Sheth, Y. Qin, K. K. Sirkar and B. C. Baltzis, Nanofiltration-based diafiltration process for solvent exchange in pharmaceutical manufacturing, J. Membr. Sci., 2003, 211, 251–261.
2. Q. T. Nguyen, P. Aptel and J. Neel, Characterization of ultrafiltration membrane. Part I. Water and organic-solvent permeabilities, J. Membr. Sci., 1979, 5, 235–251.
3. P. Eriksson, Nanofiltration extends the range of membrane filtration, Environ. Prog., 1988, 7, 58–62.
4. R. GHOSH, in Biopharmaceutical separations by ultrafiltration in Advanced Membrane Technology and Applications, John Wiley & Sons, Inc, 2008, ch. 5, p. 436.
5. H. Nawaz, J. Shi, G. S. Mittal and Y. Kakuda, Extraction of polyphenols from grape seeds and concentration, Sep. Purif. Technol., 2006, 48, 176–181.
6. D. Ormerod, B. Sledsens, G. Vercammen, D. Van Gool, T. Linsen, A. Buekenhoudt and B. Bongers, Demonstration of purification of a pharmaceutical intermediate via organic solvent nanofiltration in the presence of acid, Sep. Purif. Technol., 2013, 115, 158–162.
7. S. García Del Blanco, L. Donato and E. Drioli, Development of molecularly imprinted membranes for selective recognition of primary amines in organic medium, Sep. Purif. Technol., 2012, 87, 40–46.
8. D. Nair, S. S. Luthra, J. T. Scarpello, L. S. White, L. M. Freitas dos Santos and A. G. Livingston, Homogeneous catalyst separation and re-use through nanofiltration of organic solvents, Desalination, 2002, 147, 301–306.
9. S. Han, H. T. Wong and A. G. Livingston, Application of organic solvent to separation of ionic liquids and products from ionic liquid mediated reactions, Chem. Eng. Res. Des., 2005, 83, 309–316.
10. E. M. Rundquist, C. J. Pink and A. G. Livingston, Organic solvent nanofiltration: a potential alternative to distillation for solvent recovery from crystallisation mother liquors, Green Chem., 2012, 14, 2197–2205.
11. M. E. Tsui and M. Cheryan, Membrane processing of xanthophylls in ethanol extracts of corn, J. Food Eng., 2007, 83, 590–595.
12. H. Nawaz, J. Shi, G. S. Mittal and Y. Kakuda, Extraction of polyphenols from grape seeds and concentration by ultrafiltration, Sep. Purif. Technol., 2006, 48, 176–181.
13. M. Cisséa, F. Vaillant, D. Pallet and M. Dornier, Selecting ultrafiltration and nanofiltration membranes to concentrate anthocyanins from roselle extract (Hibiscus sabdariffa L.), Food Res. Int., 2011, 44(9), 2607–2614.
14. D. B. Reinoso, A. Moure, H. Domínguez and J. C. Parajó, Ultra- and nanofiltration of aqueous extracts from distilled fermented grape pomace, J. Food Eng., 2009, 91, 587–593.
15. Y. H. See Toh, X. X. Loh, K. Li, A. Bismarck and A. G. Livingston, In search of a standard method for the characterization of organic solvent nanofiltration membranes, J. Membr. Sci., 2007, 291, 120–125.
16. F. P. Cuperus and K. Ebert, in Nanofiltration – Principles and Applications, ed. A. I. Schafer, A. G. Fane and T. D. Waite, Elsevier, Amsterdam, 2005, pp. 522–536.
17. P. B. Kosaraju and K. K. Sirkar, Interfacially polymerized thin film composite membranes on microporous polypropylene supports for solvent-resistant nanofiltration, J. Membr. Sci., 2008, 321, 155–161.
18. A. K. Holda, B. Aernouts, W. Saeys and I. F. J. Vankelecom, Study of polymer concentrate on and evaporation time as phase inversion parameters for polysulfone-based SRNF membranes, J. Membr. Sci., 2013, 442, 196–205.
19. S. Darvishmanesha, F. Tasselli, J. C. Jansen, E. Tocci, F. Bazzarelli, P. Bernardo, P. Luis, J. Degrève, E. Drioli and B. Van der Bruggen, Preparation of solvent stable polyphenylsulfone hollow fiber nanofiltration Membranes, J. Membr. Sci., 2011, 384, 89–96.
20. S. Darvishmanesha, J. C. Jansen, F. Tasselli, E. Tocci, P. Luis, J. Degrève, E. Drioli and B. Van der Bruggen, Novel polyphenylsulfone membrane for potential use in solvent nanofiltration, J. Membr. Sci., 2011, 379, 60–68.
21. M. G. Buonomenna, G. Golemme, J. C. Jansen and S. H. Choi, Asymmetric PEEKWC membranes for treatment of organic solvent solutions, J. Membr. Sci., 2011, 368, 144–149.
22. I. Soroko, Y. Bhole and A. G. Livingston, Environmentally friendly route for the preparation of solvent resistant polyimide nanofiltration membranes, Green Chem., 2011, 13, 162.
23. M. Peyravi, A. Rahimpour and M. Jahanshahi, Thin film composite membranes with modified polysulfone supports for organic solvent nanofiltration, J. Membr. Sci., 2012, 423–424, 225–237.
24. D. Fritscha, P. Mertena, K. Heinricha, M. Lazar and M. Priske, High performance organic solvent nanofiltration membranes: Development and thorough testing of thin film composite membranes made of polymers of intrinsic microporosity (PIMs), J. Membr. Sci., 2012, 401–402, 222–231.
25. J. C. Jansen, S. Darvishmanesh, F. Tasselli, F. Bazzarelli, P. Bernardo, E. Tocci, K. Friess, A. Randova, E. Drioli and B. Van der Bruggen, Influence of the blend composition on the properties and separation performance of novel solvent resistant polyphenylsulfone/polyimide Nanofiltration membranes, J. Membr. Sci., 2013, 447, 107–118.
26. K. Hendrix, M. Vaneynde, G. Koeckelberghs and I. F. J. Vankelecom, Synthesis of modified poly(etheretherketone) polymer for the preparation of ultrafiltration and nanofiltration membranes via phase inversion, J. Membr. Sci., 2013, 447, 96–106.
27. M. F. J. Solomon, Y. Bhole and A. G. Livingston, High flux hydrophobic membranes for organic solvent nanofiltration (OSN)—Interfacial polymerization, surface modification and solvent activation, J. Membr. Sci., 2013, 434, 193–203.
28. L. Liu, X. Wang, Y. Wang, L. Li, K. Pan, J. Yang and B. Cao, Preparation and characterization of asymmetric polyarylene sulfide sulfone (PASS) solvent-resistant nanofiltration membranes, Mater. Lett., 2014, 132, 11–14.
29. F. M. Penha, K. Rezzadoria, M. C. Pronera, V. Zanattaa, G. Zina, D. W. Tondob, J. Vladimir de Oliveiraa, J. C. C. Petrusa and M. Di Luccio, Influence of different solvent and time of pre-treatment on commercial polymeric ultrafiltration membranes applied to non-aqueous solvent permeation, Eur. Polym. J., 2015, 66, 492–501.
30. F. M. Penha, K. Rezzadori, M. C. Proner, G. Zin, L. A. Fogaça, J. C. C. Petrus, J. Vladimir de Oliveira and M. Di Luccio, Evaluation of permeation of macauba oil and n-hexane mixtures through polymeric commercial membranes subjected to different pre-treatments, J. Food Eng., 2015, 155, 79–86.
31. M. Saxena, P. Ray, P. S. Singh and A. Bhattacharya, Studies towards understanding the effect of hexane on polysulfone membranes, Polym. Bull., 2015, 72, 2157–2169.
32. M. V. Tres, J. C. Racoski, R. Nobrega, R. B. Carvalho, J. V. Oliveira and M. Di Luccio, Solvent recovery from soybean oil/n-butane mixtures using a hollow fiber ultrafiltration membrane, Braz. J. Chem. Eng., 2014, 31, 243–249.
33. X. Q. Cheng, Y. L. Zhang, Z. X. Wang, Z. H. Guo, Y. P. Bai and L. Shao, Recent Advances in Polymeric Solvent-Resistant Nanofiltration Membranes, Adv. Polym. Technol., 2014, 33 DOI:10.1002/adv.21455.
34. I. Soroko, M. P. Lopes and A. Livingston, The effect of membrane formation parameters on performance of polyimide membranes for organic solvent nanofiltration (OSN): Part A. Effect of polymer/solvent/non-solvent system choice, J. Membr. Sci., 2011, 381, 152–162.
35. J. P. Sibillia, J. P. Roldan and S. Chandrasekaran, Structure of Ethylene–Chlorotrifluoroethylene copolymers, J. Polym. Sci., Part A-1: Polym. Chem., 1972, 10, 249–563.
36. Halar® ECTFE ethylene-chlorotrifluoroethylene Design and Processing Guide, Solvay technical sheets, downloaded from http://www.solvaysolexis.com.
37. Y. Mutoh and M. Miura, Porous fluorine resin membrane and process for preparing the same, US Pat., N. 4.702.836, 1987.
38. S. Ramaswamy, A. R. Greenberg and W. B. Krantz, Fabrication of poly(ECTFE) membranes via thermally induced phase separation, J. Membr. Sci., 2002, 210, 175–180.
39. I. J. Roh, S. Ramaswamy, W. B. Krantz and A. R. Greenberg, Poly(ethylenechlorotrifluoroethylene) membrane formation via thermally induced phase separation (TIPS), J. Membr. Sci., 2010, 362, 211–220.
40. H. J. Müller, A new solvent resistant membrane based on ECTFE, Desalination, 2006, 199, 191–192.
41. S. Simone, A. Figoli, S. Santoro, F. Galiano, S. Alfadel, O. A. Al-Harbi and E. Drioli, Preparation and characterization of ECTFE solvent resistant membranes and their application in pervaporation of water/toluene mixtures, Sep. Purif. Technol., 2012, 27, 147–161.
42. E. Drioli, S. Santoro, S. Simone, G. Barbieri, A. Brunetti, F. Macedonio and A. Figoli, ECTFE membrane preparation for recovery of humidified gas streams using membrane condenser, React. Funct. Polym., 2014, 79, 1–7.
43. J. Pan, C. Xiao, Q. Huang, C. Wang and H. Liu, Fabrication and properties of poly(ethylene chlorotrifluoroethylene) membranes via thermally induced phase separation (TIPS), RSC Adv., 2015, 5, 45249–45257.
44. Solvay Specialty Polymers, HALAR®ECTFE XPH-793, Technical Presentation March 2012.
45. Standard Practices for Evaluating the Resistance of Plastics to Chemical Reagents Copyright by ASTM Int'l (all rights reserved) for licensee's use only, A00390972 19 Dec 06.
46. Rheology, principles, measurements and Applications, C. W. Macosko, Wiley VCH, 1994.
47. P. Sukitpaneenit and T. S. Chung, Molecular elucidation of morphology and mechanical properties of PVDF hollow fiber membranes from aspects of phase inversion, crystallization and rheology, J. Membr. Sci., 2009, 340, 192–205.
48. I. Smallwood, Handbook of Organic Solvent Properties, Arnold, London, 2000.
49. G. Ji, L. Zhu, B. Zhu, C. Zhang and Y. Xu, Structure formation and characterization of PVDF hollow fiber membrane prepared via TIPS with diluent mixture, J. Membr. Sci., 2008, 319, 264–270.
50. S. Sawada, C. Ursino, F. Galiano, S. Simone, E. Drioli and A. Figoli, Effect of citrate-based non-toxic solvents on poly(vinylidene fluoride) membrane preparation via thermally induced phase separation, J. Membr. Sci., 2015, 493, 232–242.
51. R. N. Wenzel, Resistance of solid surfaces to wetting by water, Ind. Eng. Chem., 1936, 28, 988–994.
52. G. P. Robertson, S. D. Mikhailenko, K. Wang, P. Xing, M. D. Guiver and S. Kaliaguine, Casting solvent interactions with sulfonated poly(ether ether ketone) during proton exchange membrane fabrication, J. Membr. Sci., 2003, 219, 113–121.
53. D. R. Machado, D. Hasson and R. Semiat, Effect of solvent properties on permeate flow through nanofiltration membranes. Part I: investigation of parameters affecting solvent flux, J. Membr. Sci., 1999, 163, 93–102.
54. M. Mulder, Basic Principles of Membrane technology, Kluwer Academic Publisher, 2003, ch. 1, p. 17.

---