Polysulfone porous hollow fiber membranes for ethylene-ethane separation in gas-liquid membrane contactor

Accepted Manuscript Polysulfone porous hollow fiber membranes for ethylene-ethane separation in gas-liquid membrane contactor Anna Ovcharova, Vladimir Vasilevsky, Ilya Borisov, Stepan Bazhenov, Alexey Volkov, Alexandr Bildyukevich, Vladimir Volkov PII: DOI: Reference:

S1383-5866(16)32898-2 http://dx.doi.org/10.1016/j.seppur.2017.03.023 SEPPUR 13609

To appear in:

Separation and Purification Technology

Received Date: Revised Date: Accepted Date:

29 December 2016 15 March 2017 15 March 2017

Please cite this article as: A. Ovcharova, V. Vasilevsky, I. Borisov, S. Bazhenov, A. Volkov, A. Bildyukevich, V. Volkov, Polysulfone porous hollow fiber membranes for ethylene-ethane separation in gas-liquid membrane contactor, Separation and Purification Technology (2017), doi: http://dx.doi.org/10.1016/j.seppur.2017.03.023

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Polysulfone porous hollow fiber membranes for ethylene-ethane separation in gas-liquid membrane contactor Anna Ovcharova 1,*, Vladimir Vasilevsky 1, Ilya Borisov 1, Stepan Bazhenov 1, Alexey Volkov 1, Alexandr Bildyukevich 2, and Vladimir Volkov 1 1

A.V. Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences; [email protected] 2 Institute of Physical Organic Chemistry, National Academy of Sciences of Belarus; [email protected] * Correspondence: [email protected]; Tel.: +7-495-955-4893 Abstract Separation of olefin from paraffin having same carbon number is one of the most energy intensive separations in petrochemical processes. Employing gas-liquid membrane contactors for olefin/paraffin separation is very attractive. In the present work, the membrane contactor ethylene/ethane separation was investigated based on porous asymmetric polysulfone hollow fiber membranes having mesoporous (dav ~ 2 nm) skin layer structure in contrast to conventional ultrafiltration membranes (dav ~ 50-100 nm). Also, the membranes selective layer surface was modified to increase its hydrophobic properties expecting that pore size and surface properties tailoring prevents liquid absorbent penetration into membrane pores. The membrane contactor performance was studied under various operating parameters such as absorbent (aqueous AgNO3) solution concentration and gas and liquid flow rates. Ethylene permeance value was 185 l/(m2·h·bar), which is at the level of the best results available in literature for porous membranes, at ethylene recovery rate up to 39%, which is three times higher than that in literature. Modified membranes were characterized after membrane contactor experiments. SEM, EDX and XRD analysis as well as gas permeance measurements showed that silver nitrate crystals deposition in the membranes pore space occurs. However, no noticeable change of membrane contactor performance was observed during two months of experiments on ethylene/ethane separation. Keywords: olefin/paraffin separation; membrane contactor; polysulfone; porous hollow fiber membrane; surface modification.

1. Introduction Olefin/paraffin separation is one of the most important processes in the petrochemical industry [1]. Light olefins such as ethylene and propylene are produced in volume: in 2011, ethylene worldwide overall production was 141 MT, propylene – 70 MT [2]. About 80% of ethylene produced in Europe and USA is used for ethylene oxide, ethylene dichloride and polyethylene production. Ethylene oxide is a key precursor for detergents and antifreeze compounds production. The main difficulty in separation of olefin from paraffin having same carbon number is low difference in components boiling temperatures, e.g. for ethylene and ethane mixture, ΔTB = 14.7° [3]. During the last 60 years, traditional olefin/paraffin separation technology was low temperature distillation [4]. In the process, distillation columns having trays number 100 and higher are used. The technology disadvantages are high capital costs and large metal consumption due to application of low temperatures and high pressures. Using low temperature distillation is only economically viable when flows containing large olefin quantities are to be separated, e.g. large-scale fluid catalytic crackers outflows [1,5] An alternative approach is using non-porous gas separation membranes but industrial employment of modules based on such membranes is limited because of very close olefin and paraffin molecule sizes [6,7]. In the literature [3,8,9-10] employment of supported liquid membranes is also studied. Supported liquid membrane system represents membrane phase (in most cases, Ag+ or Cu+ salts solution) is immobilized in the polymeric substrate pores. From the high pressure side, specific olefin absorption takes place, while desorption occurs on the low pressure side. The main disadvantage of such systems is their low separation performance stability [11] because of the membrane drying-out. An attractive option for separation of olefin from paraffin is using gas-liquid membrane contactor where gas and liquid phases are divided by membrane and liquid absorbent is able to complex with olefin [12-14]. The membrane contactor benefit is high area per unit volume (up to 3000 m2/m3 [15,16]). Furthermore, it provides an opportunity for independent gas and liquid flow rate control which is impossible in conventional columns. Generally, for the olefins chemical absorption transition metals salts solutions are used [17]. Bonding force in the transition metal ion and olefin π-complex is mainly defined by the metal electronegativity value and the lattice energy of its salt [8]. The most studied liquid absorbent for olefin/paraffin separation is aqueous silver nitrate solution [12-14,18-20]. Using of Cu+ salts is mentioned [21] too, but it is less versatile because in the presence of water vapors or oxidants copper salts are susceptible to oxidizing or disproportionating [22]; moreover, olefin desorption in Cu-containing systems occurs at higher temperatures. Olefin/paraffin separation using gas-liquid membrane contactor is studied in a number of works [5,9,12-14,23,24]. Employing of porous membranes in membrane contactor is attractive, as porous structure provides an opportunity to obtain high mass-exchange characteristics. But the main challenge is that for the membrane contactor based on porous membranes, one of the most important operating conditions is non-wetting mode in which membrane pores are gas-filled; when liquid penetrates into membranes pores, mass-exchange properties drastically decrease [27,28]. As shown in [9,13,17], the major disadvantage of UF grade membranes having average pore size value dav ~ 300 nm is liquid absorbent penetration into pores which significantly

increases mass transfer resistance. Ghasem et al. [13] obtained olefin permeance value of 200 l/(m2·h·bar), but the membranes long-term stability was not investigated. To increase the process stability and avoid membrane wetting, numerous studies were carried out focused on development of composite hollow fiber membranes, constituting thin dense polymer layer on a porous support. Nymeijer et al. [14] used composite hollow fiber membranes for separation of ethylene from ethane in a membrane contactor unit with aqueous silver nitrate solution as liquid sorbent. Depending on silver nitrate solution concentration and liquid flow rate values, ethylene permeance of 5.7 – 16.5 l/(m2·h·bar) was obtained with selectivity of 72.5 – 14.7. It should be mentioned that under relatively low absorbent flow rates, separation selectivity is defined by silver nitrate solution selectivity and the overall performance is quite low; and conversely, when liquid flow rates are high, large separation performance is observed whereas the separation factor is defined by the membrane itself. In the present work, home-made asymmetric porous hollow fiber polysulfone (PSf) membranes having mesoporous (dav = 2 nm) skin layer structure and modified surface are used for ethylene/ethane separation in the gas-liquid membrane contactor. Such porous structure was chosen based on the expectation that decreased average pore size value compared to that of the UF-grade membranes will provide an opportunity to escape liquid sorbent penetration into the porous structure still keeping high olefin permeance values (being orders of magnitude higher than those of composite membranes). The latter benefit of such membranes structure is to be achieved not only by pore size tailoring, but also by increasing the skin layer surface hydrophobic properties. PSf is widely used as membrane material due to its mechanical strength, thermal and chemical resistance and relative hydrophobicity. Furthermore, it is well suited for producing hollow fiber membranes with controlled pore size [25,26]. In the literature, various approaches to membranes hydrophobic/hydrophilic properties modification are described including bulk modification [29,30] as well as surface modification [31-33]. Much attention is paid to modification focused on surface hydrophilization [34,35], including plasma treatment [36] and dopes introduction [37], but the aim of the modification used in the present work was to increase membranes skin layer surface hydrophobic properties, as aqueous absorbent was used. To this end, membranes used in the present work were modified by industrial modifier Protect Guard® application. The modifier represents a perfluorinated acrylic copolymer (PAC) and hydrophobizates the surface. Modified PSf porous membranes were employed in gas-liquid membrane contactor using aqueous AgNO3 as liquid absorbent. 2. Materials and methods 2.1. Materials The materials used to prepare spinning solutions were PSf pellets, Ultrason® S 6010 (from BASF) and N-methylpyrrolidone (NMP 99% extra pure) supplied from Acros Organics, used as the base polymer and solvent, respectively, with no supplementary purification. The poreforming additive used in the polymer solution was polyethylene glycol having average molecular weight 400 g/mol (PEG-400) supplied from Acros Organics. For membranes surface modification, aqueous phase perfluorinated acrylic copolymer (PAC) ProtectGuard® Pro was used purchased from Guard Industrie, France. AgNO3 was supplied by Chimmed, Russia. Distilled water was used to dissolve silver nitrate. The gas mixture used for membrane contactor experiments (ethylene (20%)/ethane

(80%)) was obtained from JSC Moscow’s Gas Refinery Plant. All chemicals were used without further purification. 2.2. Membranes fabrication 2.2.1. Preparation of spinning solutions For spinning solutions preparation, mass ratio PSf to PEG-400 was 1:1.25. PSf concentration was 23.9 wt %. PSf and PEG-400 were placed into temperature-controlled reservoir and stirred under 70 °C temperature and mixing rate 150 rpm; after that, NMP solvent was added into the solution and stirring continued under 120 °С and 500 rpm for 4-5 h. After preparation, spinning solution was cooled to 23 ± 0.1 °С and its dynamic viscosity was measured using Brookfield viscometer DV2T-RV and RV-07 spindle (rotation speed 100 rpm). For spinning solution with given composition average dynamic viscosity value was 32000 ± 400 cPs. Hollow fiber membranes spinning was preceded by dope solution heating to 120 °С in order to reduce its viscosity. Then, polymer solution was filtered through metal mesh (cutoff rating 4-5 μm) under gas pressure 0.18 – 0.20 MPa. After filtration, polymer solution was cooled to room temperature and degassed under vacuum. Filtration and degasation steps are extremely important for hollow fiber spinning because nonsoluble particles or gas bubbles presence in the spinning dope solution leads to fiber breakage. 2.2.2. Fabrication of hollow fiber membranes Hollow fiber membranes were prepared via a dry-wet phase inversion technique in the free spinning mode in air when bore fluid was brought into liquid polymer solution orifice, resulting in the selective layer appearance on the fibers lumen side. Under this mode, the spun fiber gets into coagulation bath by gravity and coils of its own accord. No external coagulation bath was used in the process allowing instrumental design facilitation. The spinneret ring sectional area was 1.77 mm2. Hollow fibers prepared upon contact with an aqueous medium (internal non-solvent and coagulating bath) were subsequently treated and dried in order to remove residual water from pore volume. For this purpose, membranes were washed in polar solvent (ethanol), then in nonpolar solvent (hexane) and, finally, dried under ambient conditions. The method was used to prevent capillary mesopores contraction [37] which may be present when water is drastically removed by conventional drying. 2.3. Membranes modification In order to enhance membranes hydrophobic properties, modification was carried out. The hollow fibers inner surface was modified by application of aqueous phase perfluorinated acrylic copolymer (PAC) Protect Guard® Pro, containing acrylic acid fluorine derivative monomers. The modification was carried out using tailor-made syringe enabling to fix hollow fiber from one side while other side was open to modifying liquid injection. After injecting PAC into the fibers lumen, residual modifier was removed by air flush after which fibers were dried under ambient conditions for 4 days to stabilize the modifier layer properties.

2.4. Membranes characterization 2.4.1. Scanning electron microscopy and energy dispersive X-ray analysis Cross-sections of the hollow fiber membranes were analyzed by scanning electron microscopy (SEM) on a Hitachi Table top Microscope TM 3030 Plus with proprietary highly sensitive low-vacuum secondary electron detector (Hitachi High Technologies America Inc., USA), accelerating voltage was equal to 15 kV. For dispersive X-Ray (EDS) measurements, Bruker Quantax 70 EDS system was used. The cross-section area of the samples was studied as follows: the samples were immersed in isopropyl alcohol to fill up their pores; then, the samples were fractured in liquid nitrogen and sputtered with a thin gold layer (5 nm) under vacuum. The SEM images were processed via the Quantax 70 Microanalysis software. 2.4.2. X-ray diffraction analysis X-ray diffraction (XRD) patterns were captured using X-ray diffractometer Rigaku D/maxRC, fitted with powerful 12 kW X-ray generator having rotating anode, CuKα-emission in Bragg-Brentano geometry (at the reflection). Generator operating conditions were as follows: U = 50 kV, I = 160 mA, scanning pitch 0.02°. For XRD patterns processing, MDI Jade 9 software was used. 2.4.3. Gas permeance measurements Membranes gas permeance is one of the key properties and depends on porous skin layer structure. Unmodified as well as modified membranes specimens were characterized by means of gas permeance measurement using helium and carbon dioxide, as molecular mass difference of these gases allows to determine Knudsen flow mode based on ideal selectivity value (individual gases permeance coefficients ratio). The membranes gas permeance was carried out via a volumetric technique using the setup depicted at Fig. 1. Fiber specimen having length of 16 cm was fixed in the measurement cell C. Through the cross valve V1, one of the two gases (helium or carbon dioxide) got into system. Gas pressure in the tank was controlled by the M1 manometer and gas pressure in the system – by M2 manometer. Gas inlet into system was carried out by opening the V2 valve. The volumetric flux passed through the membrane was measured using gas meter and timer. The fibers gas permeance was measured under transmembrane pressure from 0.5 to 2 bar while permeate gas pressure was kept constant at 1 bar.

Figure 1. Functional diagram of the laboratory setup for membranes gas permeance measurement.

Based on gas permeance data, average membranes pore size value was calculated by the Dusty Gas Model (DGM) [38]. When structural parameters are directly associated with membrane permeance properties, pores shape and their sizes are to be distinguished. Simple descriptive models usually do not take into account morphological deviations while these deviations may be critical. Gas permeance values help to understand which gas transport mechanism is implemented: Knudsen (free molecular flow) or Poiselle (viscous flow). Knudsen mechanism takes place usually under low pressures and gas molecule free path length exceeds pore diameter value. Viscous flow, in turn, occurs under high pressures and in this case molecule free path length is small compared to the pore size value. Under transition pressure values both transport mechanisms should be taken into account. DGM considers both mechanisms influence. General equation for gas flowing through the porous media by mixed Knudsen-Poiselle mechanism can be expressed as follows: (1), where N is gas mass flow [kg/m2·s]; R is molar gas constant [J/mole·K]; T is absolute temperature [K]; M is gas molecular weight [kg/mole]; ∆p is transmembrane pressure [Pa]; μ is gas dynamic viscosity [Pa·s]; δ is membrane thickness [m]; v is average arithmetic gas velocity [m/s]; pavg is mean pressure [Pa] and B0 is a constant depending exclusively on porous media parameters. The К0 constant is related to model particles geometric characteristics. It also takes into account angular dispersion of colliding gas particles. Assuming that pores are uniform straight cylinders, K0 and B0 can be expressed as follows:

where ε is membrane porosity coefficient; τ is pores tortuosity coefficient and r is mean pore size. Equation (1) may be rewritten as:

and expressed as follows: where . So, K0 and B0 values can be obtained by plotting. Mean pore size value may be expressed as follows:

2.4.4. Contact angle measurement Contact angle values were measured by via the conventional sessile drop technique using the LK-1 goniometer. Analysis was performed for the membranes inner surface, as selective layer was on the fibers lumen side. For image capture and digital processing of the drops images,

DropShape software was used providing Laplace-Young contact angle calculation. Measurement error was 2°. Experiments were carried out at room temperature (23±2°). Membrane surface energy value was determined according to the Owens-Wendt interfacial interaction model [40] The relation between surface energy and equilibrium contact angle of the liquid phase placed onto solid phase is derived from the Fowkes equation [41]:

where «d» and «p» subscripts relate to the dispersive and polar components of the liquid surface energy and the membrane surface energy . Water and ethylene glycol were used as test liquids, as surface energy components of both liquids are well known and widely described in the literature [42-44].

2.5. Membrane contactor experiments Membrane contactor setup functional diagram is depicted at Fig. 2.

Figure 2. Membrane contactor functional diagram: MC – membrane contactor; P – peristaltic pump; ST – silver nitrate solution tank; HR1 and HR2 – hydraulic reverse tanks 20 ml each; MFC – mass flow controller; S – sampler; V – equalizing valve; T – terminal sealing.

The membrane absorber module represented a stainless steel tube containing three hollow fiber membranes fixed by epoxy resin. Input gas mixture (ethylene (20%)/ethane (80%)) flow was brought into fibers shell side without excessive pressure (gas mixture pressure was approx. 1 bar) while silver nitrate solution was brought into fibers lumen by peristaltic pump LS-301, LOIP, Russia. Liquid sorbent was flowing in recycling mode. Gas mixture flow rate was controlled by mass flow controller (MFC), Alicat Scientific, USA. To study the membrane contactor operating conditions effect on the separation performance, various AgNO3 solution concentration values were used (1, 3 and 4M). Gas and liquid flow rates were also varied: absorbent linear velocity value was from 5 to 40 cm/s (estimated Reynolds

numbers 160-1300) and gas linear velocity value was 1, 2 and 3 m/s. Reynolds numbers were calculated as described in [45]. In the membrane contactor experiments, ethylene was separated from ethane in several stages (cycles): within each cycle, gas mixture was got into contact with liquid under preset gas and liquid flow rates. Gas mixture composition was analyzed by Crystallux-4000M gas chromatograph (GC) with a TCD detector. The GC parameters were as follows: injection port temperature 230 °C, column temperature 60 °C, and detector temperature was 230 °C. The mixtures were analyzed on a Porapak N packed column. Sample volume was 0.2 ml. All experiments were carried out under room temperature. Membrane contactor characteristics are given in Table 1. Table 1. Membrane contactor characteristics Reference values Membrane contactor effective length, cm Membrane contactor tube inner diameter, cm Hollow fiber outer diameter, cm Hollow fiber inner diameter, cm Number of fibers, pcs 1M AgNO3 solution ethylene absorptivity, l/l solution Design areas Contactor tube inner sectional area, cm2 Hollow fiber outer sectional area, cm2 Overall outer sectional area per 3 fibers, cm2 Contactor gas flow sectional area, cm2 Hollow fiber inner sectional area, cm2 Overall inner sectional area per 3 fibers, cm2 Outer surface sectional area per 3 fibers, cm2 Reference values at effective contactor length 50 cm Contactor tube inner volume, cm3 Volume per 3 fibers, cm3 Contactor gas space volume, cm3 Inner volume of 3 fibers, cm3 Membrane area per unit volume, cm2/cm3 gas Equivalent gas layer thickness over the membrane, mm

50 0.3 0.17 0.08 3 5.6 0.072 0.023 0.069 0.003 0.005 0.015 80 3.6 3.45 0.15 0.75 533 0.02

Ethylene permeance P/l, l/(m2·h·bar), was calculated as gas volume V, l, ratio to the product of membranes area S, m2, contact time τ, h, and transmembrane gas pressure difference Δp, bar. Δ

(2)

Ethylene recovery rate ηi, %, per one cycle was calculated as follows: (3) where Vi and Vi+1 – ethylene volume in mixture for preceding and following steps, respectively. The ethylene volume was calculated as overall gas volume in the system multiplied by ethylene concentration.

3. Results and discussion 3.1. Scanning electron microscopy Cross section images were captured for both unmodified fibers and PAC modified ones. The results are given in Fig. 3. As can be seen from Fig. 3, membranes have asymmetric structure consisting of thick drainage layer with finger-like macrovoids, transition layer having spongelike structure and thin (20-30 µm) skin layer from the lumen side. Calculated outer hollow fiber diameter value was 1.55 mm, inner diameter value 1.16 mm. The morphology of unmodified and PAC modified membranes do not differ significantly from each other but comparing Fig. 3 (b) and (d) one can see that transition layer of PAC modified membrane looks more dense than that of the unmodified one. It appears to be related to the PAC layer presence on the membrane skin layer surface.

a

b

c

d

Figure 3. Hollow fibers cross section images: a – enlarged fragment of unmodified fiber top view, magnification 500х; b – unmodified fiber transition and skin layers image, magnification 3000х; c – enlarged fragment of PAC modified fiber top view, magnification 500х; d – PAC modified fiber transition and skin layers image, magnification 3000х.

In order to recognize change in the membrane chemical structure after PAC treatment, EDX analysis was performed (see Fig. 4).

b a Figure 4. EDX analysis results for the PAC modified fiber: a – magnification 500x; b – magnification 2000x. Green corresponds to fluorine.

As can be seen from Fig. 4, hollow fiber selective layer surface is totally covered by PAC which indicates successful modification. Fluorine can also be seen in the macroporous layer, as shown at Fig. 3 (d). It should be noted that modification was focused on the skin layer surface hydrophobization, so, PAC penetration into the drainage layer can be regarded to as epiphenomenon. 3.2. Gas permeance and average pore size of the membranes Gas permeance measurements were carried out for the unmodified and PAC modified membranes. Table 2 gives the gas permeance values for carbon dioxide and helium. The values are equal for all transmembrane pressure intervals studied (0.5-2 bar). Ideal selectivity α (He/CO2) values as well as calculated by DGM pore sizes are presented. Table 2. Gas permeance data and calculated pore size values. Membrane

P/l (CO2), m3/(m2·h·bar)

P/l (He), m3/(m2·h·bar)

Ideal selectivity α (Не/СО2)

Average pore size dav, nm

Unmodified

560

1300

2.3

24

PAC modified

5

10.5

2.1

2

As can be seen from Table 2, unmodified membranes have high gas permeance values for both carbon dioxide and helium. Ideal selectivity value indicates that mixed Knudsen-Poiselle flow regime is present (for helium and carbon dioxide, ideal Knudsen selectivity is 3.3). Calculated by DGM average selective layer pore size was 24 nm which corresponds to mesoporous structure type. For PAC modified membrane, significant gas permeance decrease is observed. This is due to the fact that modifying polymer application leads to the formation of a thin PAC film on the membrane surface which, in turn, decreases average pore size in comparison with initial one. PAC layer deposition on the hollow fibers skin layer surface is confirmed by the EDX data. According to DGM astimations, the calculated average selective layer pore size for the PAC modified membanes was 2 nm which is one order smaller than that of the unmodified membranes.

3.3. Contact angle and surface energy of the membranes Table 3 gives water and ethylene glycol contact angle values for the inner surface of the unmodified and PAC modified hollow fiber membranes. Free surface energy values for the membranes inner surface as well as its polar and dispersive components values are presented. For the unmodified membranes, water contact angle value is 82° and ethylene glycol contact angle value is 70°. The θ (H2O) value is close to that of the polysulfone given by the manufacturer (78° as given in [46]). Slight increase is, probably, related to the membranes inner surface porosity and roughness. θ (ethylene glycol) is lower than that for water because ethylene glycol surface tension is smaller. Table 3. Contact angle and free surface energy values for the membranes inner surface. Membrane

θ (H2O), °

θ (ethylene glycol), °

Unmodified

82

PAC modified

90

,

,

,

mJ/m2

mJ/m2

mJ/m2

70

23.5

7.8

15.7

79

18.1

6.6

11.5

As can be seen from Table 3, membranes modification by PAC application leads to water contact angle value increase up to 90° which indicates the fibers inner surface hydrophobicity increase. Furthermore, free surface energy decrease is observed which is due to the presence of the perfluorinated groups on the membrane surface (the same explanation can be given to the surface energy polar component decrease: polarity decreases when fluorinated compounds are present). As for the dispersive component , it can be seen that after modification its value also slightly decreased. 3.4. Membrane contactor experiment 3.4.1. Aqueous silver nitrate solution concentration effect on the ethylene permeance

P/l (C2 H 4 ), l/(m2·h·bar)

Ethylene permeance values, P/l (C2H4), were obtained under three different absorbent concentration values: 1, 3 and 4M. Fig. 5 depicts the effect of the silver nitrate solution concentration on the maximum ethylene permeance values. 200

185

160

163

133 120 80 40 0

0

1

2

3

4

C (AgNO3 ), mol/L

Figure 5. Maximum ethylene permeance dependence on the silver nitrate concentration.

As can be seen from Fig. 5, ethylene permeance vs silver nitrate concentration curve shows a distinct maximum at the C (AgNO3) value of 3 mol/L and ethylene permeance under 4M silver concentration is slightly lower than that for 3M. Initial ethylene permeance increase with the AgNO3 concentration rise from 1 to 3 mol/L is accounted for by the fact that, in contrast to ethane, ethylene is absorbed both physically and chemically, and ethylene absorption flux increases when using the more concentrated salt solution. As shown in [19], solution molar absorptivity (moles of olefin absorbed by 1 mole of Ag+) decreases when salt concentration increases. However, total absorbed olefin amount per solution volume unit rises. Molar absorptivity decrease is due to the enhanced Ag+/NO3interactions in concentrated solutions. These interactions restrict silver-olefin complexation. Total olefin absorptivity for aqueous AgNO3 solutions initially increases with salt concentration increase, as shown in the present work. But, as shown in the literature [4], solutions having 4-10 M concentrations show absorptivity decrease. 3.4.2. Liquid absorbent and gas mixture linear velocities effect on the ethylene permeance and recovery rate In order to define the most effective operating conditions of the gas-liquid membrane contactor based on porous hollow fiber membranes, ethylene/ethane separation performance was studied under different gas and liquid linear flow rates. The aqueous silver nitrate linear flow rate values were 5, 10, 20 and 40 cm/s and gas mixture flow rates were 1, 2 and 3 m/s. Table 4 presents results for all the experiments carried out.

Table 4. Membrane contactor tests results on ethylene/ethane separation at different process parameters. C (AgNO3), mol/L

v (absorbent), cm/s 5

10 1 20

40

5

10 3 20

40

5

10 4 20

40

v (gas mixture), cm/s

P/l (C2H4), l/(m2*h*bar)

η (per 1 cycle), %

η (total), %

1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3

70 113 133 63 73 91 58 68 87 68 74 84 119 160 185 102 145 165 85 105 146 75 97 120 80 114 163 73 120 149 68 89 131 72 105 117

22 20 19 29 14 12 25 14 13 28 16 15 39 30 24 32 24 18 33 22 17 28 21 16 31 22 18 27 25 20 25 20 19 28 21 18

73 66 64 72 53 47 66 54 52 61 59 58 92 83 75 87 74 64 86 72 61 81 69 59 85 72 63 80 77 68 80 67 65 76 69 63

Table 4 gives maximum ethylene permeance values obtained in each experiment. It can be seen that the highest P/l (C2H4) value corresponds to the following operating parameters: C (AgNO3) = 3M; absorbent linear velocity in the fibers lumen is 5 cm/s (estimated Reynolds number Re = 160); gas mixture linear velocity in the membranes shell side is 3 m/s. It can be seen that in all experiments inverse relation between ethylene permeance and absorbent flow rate was observed. As shown by Rajabzadeh et al. [23], absorbent velocity increase leads to the absorption flux rise due to the mass transfer coefficient increase. However, such relation takes place only in case when membrane pores are gas-filled. In the present work, as it is shown in the next section, salt deposition into membranes pores occurred which,

P/l (C2 H4 ), l/(m2 ·h·bar)

probably, was accounted for by partial liquid intrusion into the pores, and this is why ethylene permeance decreased with sorbent velocity increase. As shown in Table 4, maximum ethylene from ethylene-ethane mixture recovery rate was also obtained under minimal liquid velocity studied, but gas mixture flow rate increase led to the recovery rate value decrease. This is due to the fact that when liquid velocity increases, period of contact during which complexation reaction between reacting compounds (ethylene and silver nitrate) takes place, shortens, and separation efficiency also decreases. The same phenomena was observed by Ghasem et al. [13]. Ethylene concentration in the feed flow and gas mixture velocity influence on the P/l (C2H4) value at the optimum (3M) absorbent concentration and liquid linear velocity value of 5 cm/s is depicted at Fig. 6. 200

160

120 gas velocity 1 m/s gas velocity 2 m/s

80

gas velocity 3 m/s 40

0 0,05

0,1

0,15

0,2 C (C2 H4 )

Figure 6. Ethylene concentration in the feed flow influence on the ethylene permeance at different gas flow rates.

As can be seen from Fig. 6, as the ethylene concentration in the feed flow reduces, P/l (C2H4) slightly decreases. Maximum ethylene permeance is observed under gas flow velocity of 3 m/s which is accounted for by the process driving force increase, which, in turn, is due to the transmembrane pressure difference rise when gas flow rate is increased. The same relation was described by Tsou et al. [9]. Table 5 gives a comparison of the results obtained in the present work compared to those described in literature focused on separation performance parameters study in equivalent system (ethylene and ethane gas mixture, aqueous silver nitrate solution as an absorbent). It can be seen that in most of studies, such system employs composite membranes (having dense skin layer). In that case, high selectivity is a benefit, but overall ethylene performance is not high. All presented in Table 5 values of ethylene permeance and its recovery rate are the maximum obtained.

Table 5. Results obtained in the present work compared to those described in literature. Membranes type

[14]

[13]

[12]

[20]

[9]

This work

Composite hollow fiber membranes: Accurel (polypropylene) support and EPDM selective layer Fabricated via TIPS technique fibers: 28% PVDF, 72% triacetin. Pore size ~ 300 nm, porosity 56%, θ (H2O) ~ 108°, CEPw ~ 1.6 bar Composite hollow fiber membranes: Accurel (polypropylene) support and PEO/PBT selective layer Composite hollow fiber membranes: Accurel (polypropylene) support and SPEEK selective layer Porous PSf hollow fibers pore size not given, wall thickness varied. Liquid membrane system. Porous PAC modified PSf hollow fibers, pore size ~ 2 nm, θ(H2O)=90°

Gas mixture

Absorbent

Absorbent flow rate

Inlet pressure

η (C2H4), %

P/l (C2H4), l/(m2·h·bar)

C2H4 (80%) + C2H6 (20%)

Aqueous AgNO3, 1.8М or 3.5М

30 – 350 ml/min

1 or 3 bar

N/A

~ 20

C2H4 + C2H6 (various concentrati ons)

Aqueous AgNO3, 1, 3 and 4M

10 – 50 ml/min

1 bar

~ 14%

~ 200

C2H4 (20%) + C2H6 (80%)

Aqueous AgNO3 3.5М

50 – 350 ml/min

3 bar

N/A

~ 30

C2H4 (80%) + C2H6 (20%)

Aqueous AgNO3 3.5М

50 – 600 ml/min

1 or 3 bar

N/A

~3

C2H4 (74%) + C2H6 (26%)

Aqueous AgNO3 0.5-5М

0 – 10 ml/min

25 – 125 psig

N/A

~ 30

C2H6 (80%) + C2H4 (20%)

Aqueous AgNO3, 1, 3 and 4M

Linear velocity 5 - 40 cm/s

1 bar

~ 40%

~ 190

It is clear that porous membranes enable to obtain gas permeance values order of magnitude higher than those of composite membranes. To compare our results with ones given by Ghasem et al. [13] would be more appropriate, as system parameters given in that work were similar to ours. As can be seen from Table 5, ethylene permeance values are comparable for both studies while in the present work ethylene recovery rate per 1 cycle value is three times higher than that in [13]. 3.5. Membranes characterization after the membrane contactor experiments Unfortunately, the most of studies do not consider long-term stability issue for the gas-liquid membrane contactor based on porous membranes and commonly, membrane characterization after contactor experiments is not carried out. Faiz et al. [47] investigated properties of PVDF and PTFE hollow fiber membranes after membrane contactor experiments using aqueous silver solution as liquid sorbent and discovered that noticeable silver deposition was found on the membranes surface but it was not severe enough to cause noticeable membrane contactor performance decline.

To examine changes occuring in the PAC modified PSf membranes after 2 months of contactor experiments, characterization was carried out including SEM, EDX and X-ray diffraction analysis, as well as gas permeance and contact angle measurement. Fig. 7 depicts SEM and EDX images of the membranes.

a

b

с

d

Fig. 7. SEM and EDX images of the PAC modified PSf membranes after membrane contactor experiment. a – SEM image of the membrane top view (enlarged fragment), magnification 500x; b – SEM image of the membrane transition and skin layers, magnification 3000x; с – the membrane EDX image, magnification 500x; d – the membrane EDX image, magnification 2000x.

It can be seen from both SEM and EDX images that after 2 months of membrane contactor experiments (when silver nitrate solution was supplied into hollow fibers lumen), PAC layer on the skin layer surface is still present but pores space is partially clogged by silver compound depositions (penetration depth ~ 750 μm). Nevertheless, silver traces amount from the hollow fiber shell side is small to negligible which indicates that no absorbent leakage into the shell side occurred during all membrane contactor operating period. Probably, salt deposition into the membranes pore space is due to the sorbent pressure rise when its velocity in the fibers lumen is increased and also to the fact that deposition of silver compounds increases membrane wettability. In order to determine chemical composition of silver compounds in membranes pore space, comparative X-ray diffraction analysis was carried out. Fig. 8 gives XRD patterns for both PAC modified hollow fiber PSf membranes and the same membranes after contactor studies.

a

b

Fig. 8. XRD patterns of hollow fiber membranes: a – PAC modified hollow fiber PSf membrane; b – PAC modified hollow fiber PSf memnbrane after 2 months of membrane contactor experiments.

As can be seen from Fig. 8 (a), wide diffusional scattering halo was observed for PAC modified PSf membranes, so, it can be concluded that hollow fiber materials, namely, PSf and PAC are X-ray amorphous. However, hollow fibers XRD pattern after membrane contactor experiment drastically differs from that presented in Fig. 8 (a). As is seen from Fig. 8 (b), in the 5-100° range of angle 2θBragg values, crystalline phase reflexes were observed as well as amorphous halo. XRD analysis results indicate AgNO3 phases of two structural modifications which are distinct in symmetry group and elementary cell parameters. The first phase has rhombohedral elementary cell with following lattice parameters: a = b = 5.16 Å, с = 16.58 Å. The second phase has orthorhombic elementary cell with following lattice parameters: a = 6.995 Å, b = 7.335 Å, c = 10.122 Å. Average crystallites size was determined via Williamson-Hall procedure, considering diffraction reflexes widening; its value is 69 nm. It is important that ethylene permeance in the membrane contactor decreased only slightly with time and no drastic membrane contactor mass exchange parameters decrease was observed which would be inevitable in case when membrane pores are filled with liquid instead of gas. The fact may indicate that only a fraction of membrane pores is blocked by absorbent and pores enabling effective gas transport are present as well. Table 6 gives gas permeance and contact angle values of the membranes after membrane contactor experiment. Table 6. Gas permeance and contact angle values for the membranes after membrane contactor experiment. P/l (CO2), Ideal , P/l (He), θ (ethylene , , m3/(m2·h·bar selectivity α θ (H2O), ° mJ/m 3 2 2 m /(m ·h·bar) glycol), ° mJ/m mJ/m2 2 ) (Не/СО2) 0.4 0.9 2.1 89 84 19.5 2.6 16.9

It can be seen that gas permeance of the membranes after membrane contactor experiments decreased by an order of magnitude compared to initial values (see Table 2). Such decrease is probably due to the silver nitrate crystals deposition on the membranes pore walls which leads to the blocking of a transport pores fraction. However, it should be noted that after membrane contactor experiments, the membranes gas permeance values for both helium and carbon dioxide remained significantly higher than maximum (185 l/(m2·h·bar)) ethylene permeance in the contactor, which is accounted for by extra liquid phase resistance in the membrane module.

After 2 months of membrane contactor experiments, the membranes free surface energy is slightly higher than that of the PAC modified PSf membranes but still lower than surface energy of the unmodified membranes (see Table 3), which indicates membranes surface hydrophilization after contact with aqueous AgNO3 solution. However, as can be concluded from the EDX data, layer containing perfluorinated groups remains on the hollow fibers inner surface after membrane contactor tests. 4. Conclusions In the present work, home-made porous polysulfone hollow fiber membranes were fabricated and further modified by perfluorinated acrylic copolymer (PAC) application in order to increase their inner surface hydrophobic properties. The gas-liquid membrane contactor containing PAC modified PSf hollow fiber membranes was used for ethylene-ethane separation with aqueous silver nitrate solution as liquid absorbent. Membranes characterization via gas permeance and contact angle measurement, as well as SEM and EDX analysis was carried out. The gas-liquid membrane contactor separation performance was examined under different AgNO3 solution concentrations and various gas and liquid linear velocities. It was shown that maximum ethylene permeance corresponds to the following operating conditions: C (AgNO3) = 3М, liquid absorbent velocity value 5 cm/s (estimated Reynolds number Re = 160), gas mixture linear velocity in the fibers shell side 3 m/s. Under given operating conditions, maximum P/l (C2H4) value was 185 l/(m2·h·bar) which is order of magnitude higher than that available in literature for composite membranes. Ethylene recovery rate values per 1 cycle were 12-39% which is higher than those given in literature for porous membranes. Analysis based on gas permeance data as well as SEM and EDX techniques showed that after membrane contactor experiments, the membranes pore space is partially clogged by silver nitrate crystals which may be because of the liquid sorbent penetration in the pores. However, no noticeable change of membrane contactor performance was observed. Acknowledgements This work was performed at the A.V. Topchiev Institute of Petrochemical Synthesis and supported by the Russian Science Foundation, project no. 14-49-00101. A.A. Ovcharova, V.P. Vasilevsky, I.L. Borisov, S.D. Bazhenov, A.V. Bildyukevich and V.V. Volkov acknowledge the financial support of RSF. Authors thank K.A. Kutuzov for carrying out the membrane contactor experiments, S.P. Molchanov and S.N. Polyakov for providing X-ray diffraction analysis data and D.S. Bakhtin for providing SEM and EDX images. References 1. 2. 3.

4.

Eldridge R. B. Olefin/paraffin separation technology: a review //Industrial & engineering chemistry research. – 1993. – Т. 32. – №. 10. – С. 2208-2212. True W. R. Global ethylene capacity continues advance in 2011 //Oil & gas journal. – 2012. – Т. 110. – №. 7. – С. 78-84. Krull F. F., Medved M., Melin T. Novel supported ionic liquid membranes for simultaneous homogeneously catalyzed reaction and vapor separation //Chemical Engineering Science. – 2007. – Т. 62. – №. 18. – С. 5579-5585. Safarik D. J., Eldridge R. B. Olefin/paraffin separations by reactive absorption: a review //Industrial & engineering chemistry research. – 1998. – Т. 37. – №. 7. – С. 2571-2581.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14.

15. 16. 17.

18.

19. 20.

21.

22.

Bessarabov D. G., Sanderson R. D., Jacobs E. P., Beckman I. N. High-efficiency separation of an ethylene/ethane mixture by a large-scale liquid-membrane contactor containing flatsheet nonporous polymeric gas-separation membranes and a selective flowing-liquid absorbent //Industrial & engineering chemistry research. – 1995. – Т. 34. – №. 5. – С. 17691778. Staudt-Bickel C., Koros W. J. Olefin/paraffin gas separations with 6FDA-based polyimide membranes //Journal of Membrane Science. – 2000. – Т. 170. – №. 2. – С. 205-214. Krol J. J., Boerrigter M., Koops G. H. Polyimide hollow fiber gas separation membranes: preparation and the suppression of plasticization in propane/propylene environments //Journal of Membrane Science. – 2001. – Т. 184. – №. 2. – С. 275-286. Ravanchi M. T., Kaghazchi T., Kargari A. Supported liquid membrane separation of propylene–propane mixtures using a metal ion carrier //Desalination. – 2010. – Т. 250. – №. 1. – С. 130-135. Tsou D. T., Blachman M. W., Davis J. C. Silver-facilitated olefin/paraffin separation in a liquid membrane contactor system //Industrial & engineering chemistry research. – 1994. – Т. 33. – №. 12. – С. 3209-3216. Azizi S., Kaghazchi T., Kargari A. Propylene/propane separation using N-methyl pyrrolidone/AgNO 3 supported liquid membrane //Journal of the Taiwan Institute of Chemical Engineers. – 2015. – Т. 57. – С. 1-8. Van Zyl A. J., Kerres J. A., Cui W., Junginger M. Application of new sulfonated ionomer membranes in the separation of pentene and pentane by facilitated transport //Journal of membrane science. – 1997. – Т. 137. – №. 1. – С. 173-185. Nymeijer K., Visser T., Assen R., Wessling M. Super selective membranes in gas–liquid membrane contactors for olefin/paraffin separation //Journal of membrane science. – 2004. – Т. 232. – №. 1. – С. 107-114. Ghasem N., Al-Marzouqi M., Ismail Z. Gas–liquid membrane contactor for ethylene/ethane separation by aqueous silver nitrate solution //Separation and Purification Technology. – 2014. – Т. 127. – С. 140-148. Nymeijer K., Visser T., Assen R., Wessling M. Composite hollow fiber gas–liquid membrane contactors for olefin/paraffin separation //Separation and purification technology. – 2004. – Т. 37. – №. 3. – С. 209-220. Mansourizadeh A., Ismail A. F. Hollow fiber gas–liquid membrane contactors for acid gas capture: a review //Journal of hazardous materials. – 2009. – Т. 171. – №. 1. – С. 38-53. Bazhenov S.D., Lyubimova E.S. Gas-Liquid Membrane Contactors for Carbon Dioxide Capture from Gaseous Streams //Petroleum Chemistry. – 2016. T. 56. – № 10. – С.893-919. Faiz R., Li K. Olefin/paraffin separation using membrane based facilitated transport/chemical absorption techniques // Chemical engineering science. – 2012. – Т. 73. – С. 261-284. Nymeijer K., Visser T., Brilman W., Wessling M. Analysis of the complexation reaction between Ag+ and ethylene //Industrial & engineering chemistry research. – 2004. – Т. 43. – №. 11. – С. 2627-2635.. Cho I. H., Yasuda H. K., Marrero T. R. Solubility of ethylene in aqueous silver nitrate //Journal of Chemical and Engineering Data. – 1995. – Т. 40. – №. 1. – С. 107-111. Nymeijer K., Visser T., Assen R., Wessling M. Olefin-selective membranes in gas-liquid membrane contactors for olefin/paraffin separation //Industrial & engineering chemistry research. – 2004. – Т. 43. – №. 3. – С. 720-727. Yu G., Deng L., Abdeltawab A.A., Al-Deyab S.S., Chen S., Zhang J. Functional solution composed of Cu (I) salt and ionic liquids to separate propylene from propane //Industrial & Engineering Chemistry Research. – 2014. – Т. 53. – №. 34. – С. 13430-13435. Agel F., Pitsch F., Krull F.F., Schulz P., Wessling M., Melin T., Wassercheid P. Ionic liquid silver salt complexes for propene/propane separation //Physical Chemistry Chemical Physics. – 2011. – Т. 13. – №. 2. – С. 725-731.

23. Rajabzadeh S., Teramoto M., Al-Marzouqi M., Kamio E., Yoshikage O., Maruyama T. Experimental and theoretical study on propylene absorption by using PVDF hollow fiber membrane contactors with various membrane structures //Journal of Membrane Science. – 2010. – Т. 346. – №. 1. – С. 86-97. 24. Faiz R., Fallanza M., Ortiz I., Li K. Separation of olefin/paraffin gas mixtures using ceramic hollow fiber membrane contactors //Industrial & Engineering Chemistry Research. – 2013. – Т. 52. – №. 23. – С. 7918-7929. 25. Plisko T.V., Bildyukevich A.V., Usosky V.V., Volkov V.V. Influence of the concentration and molecular weight of polyethylene glycol on the structure and permeance of polysulfone hollow fiber membranes. //Petroleum Chemistry. – 2016. T. 56. – № 10. – С.321-329. 26. Ismail A. F., Mansourizadeh A. A comparative study on the structure and performance of porous polyvinylidene fluoride and polysulfone hollow fiber membranes for CO 2 absorption //Journal of Membrane Science. – 2010. – Т. 365. – №. 1. – С. 319-328. 27. Kirsch V. A. et al. Simulation of convective-diffusional processes in hollow fiber membrane contactors //Separation and Purification Technology. – 2016. – Т. 167. – С. 63-69. 28. Fallanza M., Ortiz A., Gorri D., Ortiz I. Effect of liquid flow on the separation of propylene/propane mixtures with a gas/liquid membrane contactor using Ag+-RTIL solutions //Desalination and Water Treatment. – 2011. – Т. 27. – №. 1-3. – С. 123-129. 29. Khayet M., Matsuura T. Application of surface modifying macromolecules for the preparation of membranes for membrane distillation //Desalination. – 2003. – Т. 158. – №. 1. – С. 51-56. 30. Plisko T. V., Bildyukevich A.V., Volkov V.V., Osipov N.N. Formation of hollow fiber membranes doped with multiwalled carbon nanotube dispersions //Petroleum Chemistry. – 2015. – Т. 55. – №. 4. – С. 318-332. 31. Korminouri F., Rahbari-Sisakht M., Matsuura T., Ismail A.F. Surface modification of polysulfone hollow fiber membrane spun under different air-gap lengths for carbon dioxide absorption in membrane contactor system //Chemical Engineering Journal. – 2015. – Т. 264. – С. 453-461. 32. He T., Mulder M.H.V., Strathmann H., Wessling M. Preparation of composite hollow fiber membranes: co-extrusion of hydrophilic coatings onto porous hydrophobic support structures //Journal of membrane science. – 2002. – Т. 207. – №. 2. – С. 143-156. 33. Khulbe K. C., Feng C., Matsuura T. The art of surface modification of synthetic polymeric membranes //Journal of Applied Polymer Science. – 2010. – Т. 115. – №. 2. – С. 855-895. 34. Qin J. J., Wong F. S. Hypochlorite treatment of hydrophilic hollow fiber ultrafiltration membranes for high fluxes //Desalination. – 2002. – Т. 146. – №. 1. – С. 307-309. 35. Bildyukevich A.V., Plisko T.V., Liubimova A.S., Volkov V.V., Usosky V.V. Hydrophilization of polysulfone hollow fiber membranes via addition of polyvinylpyrrolidone to the bore fluid. //Journal of Membrane Science – 2017. In Press. 36. Volkov V.V., Ibragimov R.G., Abdullin I.Sh., Gallyamov R.T., Ovcharova A.A., Bildyukevich A.V. Modification of polysulfone porous hollow fiber membranes by air plasma treatment. //Journal of Physics: Conference Series 751 (2016) 012028. 37. Chakrabarty B., Ghoshal A. K., Purkait M. K. Preparation, characterization and performance studies of polysulfone membranes using PVP as an additive //Journal of Membrane Science. – 2008. – Т. 315. – №. 1. – С. 36-47. 38. Bildyukevich A.V., Usosky V.V. Prevention of the capillary contraction of polysulfone based hollow fiber membranes. // Petroleum Chemistry. – 2014. – Т. 54. – №. 8. – С. 652658. 39. Mason E. A., Malinauskas A. P. Gas transport in porous media: the dusty-gas model. – Elsevier Science Ltd, 1983. – Т. 17. 40. Owens D. K., Wendt R. C. Estimation of the surface free energy of polymers //Journal of applied polymer science. – 1969. – Т. 13. – №. 8. – С. 1741-1747.

41. Fowkes F. M. Calculation of work of adhesion by pair potential suummation //Journal of colloid and interface science. – 1968. – Т. 28. – №. 3-4. – С. 493-505. 42. Decker E. L., Frank B., Suo Y., Garoff S. Physics of contact angle measurement //Colloids and Surfaces A: Physicochemical and Engineering Aspects. – 1999. – Т. 156. – №. 1. – С. 177-189. 43. Šikalo Š., Tropea C., Ganić E. N. Dynamic wetting angle of a spreading droplet //Experimental Thermal and Fluid Science. – 2005. – Т. 29. – №. 7. – С. 795-802. 44. Volkov V. V., Lebedeva V.I., Petrova I.V., Bobyl A.V., Konnikov S.G., Roldughin V.I., Van Erkel J., Tereshchenko G.F. Adlayers of palladium particles and their aggregates on porous polypropylene hollow fiber membranes as hydrogenization contractors/reactors //Advances in colloid and interface science. – 2011. – Т. 164. – №. 1. – С. 144-155. 45. Kirsch V. A., Volkov V. V., Bildukevich A. V. Modeling the Transverse Shell-side Mass Transfer in Hollow Fiber Membrane Contactors at Low Reynolds Numbers //Physics Procedia. – 2015. – Т. 72. – С. 162-165. 46. http://iwww.plasticsportal.com/membranes/BASF_Membranes_Flyer.pdf 47. Faiz R., Fallanza M., Boributh S., Jiraratananon R., Ortiz I., Li K. Long term stability of PTFE and PVDF membrane contactors in the application of propylene/propane separation using AgNO 3 solution //Chemical Engineering Science. – 2013. – Т. 94. – С. 108-119.

    

Mesoporous PSf hollow fibers with hydrophobic polymer coating fabricated. Ethylene/ethane separation in membrane contactor with silver nitrate absorbent. Olefin permeance 185 l/(m2·h·bar) higher than in composite membranes; recovery 40%. Fibers after experiment characterized by SEM, EDX, gas permeance and contact angle. No MC mass-exchange properties decline observed despite silver compounds deposits.