water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles

water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles

Accepted Manuscript Pervaporative dehydration of ethanol/water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles Gabriela ...

1MB Sizes 0 Downloads 82 Views

Accepted Manuscript Pervaporative dehydration of ethanol/water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles Gabriela Dudek, Roman Turczyn, Małgorzata Gnus, Krystyna Konieczny PII: DOI: Reference:

S1383-5866(17)31808-7 https://doi.org/10.1016/j.seppur.2017.09.023 SEPPUR 14032

To appear in:

Separation and Purification Technology

Received Date: Revised Date: Accepted Date:

6 June 2017 7 September 2017 7 September 2017

Please cite this article as: G. Dudek, R. Turczyn, M. Gnus, K. Konieczny, Pervaporative dehydration of ethanol/ water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles, Separation and Purification Technology (2017), doi: https://doi.org/10.1016/j.seppur.2017.09.023

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.

Pervaporative dehydration of ethanol/water mixture through hybrid alginate membranes with ferroferic oxide nanoparticles Gabriela Dudek1, Roman Turczyn1, Małgorzata Gnus1, Krystyna Konieczny2 1

Silesian University of Technology, Faculty of Chemistry, Strzody 9, 44-100 Gliwice, Poland, email: [email protected] 2 Silesian University of Technology, Institute of Water and Wastewater Engineering, Konarskiego 18, 44-100 Gliwice, Poland

Highlights Study of hybrid alginate/magnetite Alg/Fe3O4 membranes in pervaporative dehydration of ethanol. Membranes are crosslinked with four crosslinking agents: glutaraldehyde, phosphoric(V) acid, calcium chloride or citric acid, respectively. Presence of Fe3O4 in polymer matrix further improves all separation parameters. The highest PSI is obtained for hybrid alginate membrane loaded with 15 wt% of ferroferric oxide.

ABSTRACT Composite alginate membranes filled with various amount of iron ferroferric oxides (Fe3O4) and cross-linked by four different agents, i.e. calcium chloride (AlgCa), phosphoric(V) acid (AlgP), glutaraldehyde (AlgGA) and citric acid (AlgC) were prepared. Alteration of the separation performance related to the membrane’s composition was examined by the pervaporative dehydration process of water/ethanol mixture. The observed influence of the nature of chosen crosslinking agent and the contents of Fe 3O4 on the transport parameters was discussed in detail. Among all the tested membranes, the alginate membrane loaded with 15 wt% of magnetite and crosslinked with phosphoric(V) acid (AlgP), showed the highest PSI of 33.86 kg·m-2·h-1·μm-1. KEYWORDS Sodium alginate, hybrid membrane, pervaporation, ethanol/water separation, dehydration, ferroferric oxide 1. Introduction The application of biopolymers has increased significantly recently, and these compounds are now regularly used in nanotechnology, environmental and medical sciences [1-5]. The use of biopolymers in place of synthetic polymers has been rising due to their excellent biocompatibility, biodegradability, high hydrophilicity, relative non-toxicity and low costs. The most often studied biopolymers are polysaccharides and proteins, with their constituent sugars and amino acids, their sequence, and contribution to hydrodynamic three-dimensional structure in the solvent influencing the overall polymer functionality [6-7]. Proteins applied in industry are usually derived from a variety of sources, including animals, plants and fungi. Polysaccharides can be obtained from botanical and animal sources and from bacterial fermentation. Among the polysaccharides, these are cellulose, chitosan and alginate that are mostly investigated [7]. Alginate is a material of interest due to its unique and useful

properties. Being extracted from marine brown algae, alginates are non-toxic and edible polysaccharides. This compound is a copolymer of 1,4-linked β-D-mannuronic acid and α-Lguluronic acid, in varying proportions. These two uronic acids have only minor differences in structure, and they adopt different chair conformations, such that the bulky carboxyl group is in the energetically favoured equatorial position [8-9]. Alginate that exhibits many intrinsic properties, is considered as a prospective membrane material for membrane separation technology [10-11], adsorption of dyes [12,13] and metal ions [14-15]. During the past few decades, membrane separation processes have become one of the emerging technologies that underwent a rapid growth [16-18]. It has drawn the attention of researchers to the separation technology field, with its better performance when compared to the conventional separation technology. Most of the membrane separation technologies are well developed and established. Among these technologies, pervaporation (PV) is still a rapidly developing membrane separation technique. Especially, PV is a promising technique for separation of organic liquid mixtures such as azeotropic mixtures [19-21]. The application of alginate membranes in pervaporation processes has played an important role, and therefore these materials have been intensively studied and developed in numerous research groups, and widely used in a pristine or modified forms [10,11,22]. Nigiz at al. [10] applied zeolite filled sodium alginate membrane (NaAlg) in the pervaporation of ethanol/water mixtures. The study showed the availability of sodium alginate membranes in PV process. Addition of zeolite into the NaAlg matrix improved the mechanical properties of alginate membranes and the magnitudes of flux. Moulik et al. [11] applied sodium alginate/polyaniline (SA/PAni) composite

membranes,

synthesized

on

ultraporous

polyacrylonitrile

(PAN)

and

polyethersulfone (PES) supports and subjected to pervaporation dehydration of acetic acid. In this work, the effect of the feed water concentration, permeate pressure and membrane thickness on the process flux and selectivity were evaluated. The results showed that such

composite membrane was suitable for removal of small concentrations of preferentially sorbed water from acetic acid. Sajjan et al. [22] investigated chitosan-wrapped multiwalled carbon nanotubes (CS-wrapped MWCNTs) incorporated sodium alginate membranes. These membranes were employed for the separation of water–isopropanol mixtures. The study indicated that the increase of the content of CS-wrapped MWCNTs in the membrane matrix results in a simultaneous increase of both permeation flux and selectivity. This was explained on the bases of selective adsorption, molecular sieving action and significant enhancement of hydrophilicity of the membrane matrix. The highest separation selectivity of 6419 with a flux of 0.218 kg/m2h at 300C was obtained for the membrane containing 2 wt% of CS-wrapped MWCNT. In our previous research [23-27] we investigated the performance of chitosan membranes with different content of iron oxide nanoparticles in the process of ethanol dehydration. The results showed that the addition of magnetite particles to chitosan matrix created extra free volumes in polymer, and in consequence, offered space for water molecules to permeate easier through the membranes. In this paper we apply sodium alginate membranes with various amount of ferroferric oxides (Fe3O4), cross-linked by four different agents, i.e. calcium ions (AlgCa), phosphoric acid (AlgP), glutaraldehyde (AlgGA) and citric acid (AlgC). We determined how cross-linking agents and iron oxide powder content can change the transport properties of alginate membranes in dehydration of ethanol by pervaporation. 2. Experimental Sodium alginate was obtained from Across. FeCl6 ∙6H2O, sodium acetate, ethylene glycol and 2,2’-(ethylenedioxy)bis(ethylamine) (EDBE),that were used in the preparation of iron oxide nanoparticles, were purchased from Sigma-Aldrich.2.5 wt% glutaraldehyde solution, Ca2+ions as CaCl2, phosphoric acid and citric acid, applied as crosslinkers, were purchased from Avantor Performance Materials.

2.1 Preparation of iron oxide nanoparticles In a 100 ml beaker, a mixture of FeCl6∙6H2O (10 mmol), sodium acetate (10 mmol), and ethylene glycol (30 ml) was stirred vigorously at 50˚C to give a transparent solution. 30 mmol of EDBE was added to the resultant solution and the temperature of the reaction mixture was raised to 120˚C. Stirring and heating was continued for additional 2 hours. Then, the mixture was poured into 800 ml beaker and ethylene glycol was evaporated until the brown suspension transformed into black colloid. The resulting suspension was cooled to room temperature and, finally, the obtained particles were subjected to magnetic decantation and repeatedly washed with distilled water, ethanol and acetone. The size distribution of the prepared magnetite nanopowder was determined using image analysis acquired by Tecnai F20 Twin TEM microscope.

2.2 Membrane preparation The scheme of the preparation of sodium alginate membranes is shown in Fig.1. Briefly, 1.5% sodium alginate solution was prepared by dissolving an appropriate amount of sodium alginate powder in deionised water. This solution was mixed with a suitable portion of iron oxide nanoparticles (5; 10; 15; 20; 25 wt%). The sodium alginate solution was then casted onto a levelled glass plate and evaporated to dryness at 40 oC. After 24 h, the membrane was crosslinked using an appropriate crosslinkers, i.e. 2.5 wt% calcium chloride in water, 1.25 wt% glutaraldehyde solution in water, 3.5 vol.% of phosphoric acid or 3.5 wt% citric acid in isopropanol/water mixture. In case of crosslinking with phosphoric acid (AlgP), sodium alginate membranes were immersed in isopropanol–water bath (90/10 vol.%) containing 3.5 vol.% of phosphoric acid for 180 min at room temperature. Crosslinked glutaraldehyde membranes (AlgGA) were prepared by immersing the dry membrane in the 50 cm3 of

glutaraldehyde solution for 24 h and subsequent washing with deionised water. In case of crosslinking with Ca2+ ions (AlgCa), sodium alginate membranes were immersed in calcium chloride solution for 120 min at room temperature. Crosslinked citric membranes (AlgC) were prepared by immersing of the dry membrane in 50 cm3 3.5 wt% citric acid solution in isopropanol/water mixture (7:3 v/v) for 3 hours, then washing with distilled water until neutral pH. The pristine NaAlg membrane was prepared in the same manner as above except for the addition of iron oxide nanoparticles. The membrane thickness was measured by the micrometer screw gauge and was equal to 28.0 ± 2.0 μm.

FIGURE 1

2.3 Membrane characterization Membranes were characterized using scanning electron microscope (SEM) Phenom ProX equipped with 3D Roughness Reconstruction software and energy dispersive X-ray (EDS) analyzer,

differential scanning calorimeter (DSC), and FTIR spectroscopy. Basing on the results of the sorption tests, the degree of swelling was calculated.. The SEM images were used to determine the size distribution of magnetite embedded in the hybrid membranes. DSC thermograms of pristine Alg, sodium alginate membrane and membranes crosslinked with four different agents, i.e. Ca2+ ions, phosphoric acid, glutaraldehyde and citric acid were recorded on Mettler-Toledo DSC 822e differential scanning calorimeter. The thermal measurements were performed over the temperature range of -60–350 oC with the 10K/min heating rate, and under N2 atmosphere (flow rate 250 cm3/min). Static contact angles of the dry membranes were measured as an average value of 15 independent measurements using Portable Type Phoenix - I goniometer Surface Electro Optics (SEO). FTIR measurements were performed for the membrane films of 20–30 μm thickness using Perkin Elmer Spectrum Two FTIR spectrometer for scanning the films, at ambient

temperatures, in the spectral range of 650-4000 cm-1, and with the spectral resolution of 2 cm1

. The degree of swelling (DS) of the pristine and hybrid Alg membranes with different iron

oxide loadings was determined using sorption test. In this case, membrane samples were immersed in water or ethanol and mass changes of analysed samples were determined during one week using analytical balance. The degree of swelling (DS) was calculated using following equation: (1) where Wwet is the weight of a wet membrane and Wdry is the weight of a dry membrane.

Tensile strength measurements were conducted using MTS Insight® Universal Test System equipped with a 100 N sensor. The applied tension speed was 0.05 mm/s. The Young modulus was calculated from the following equation: (2) where E is the Young modulus (Pa), F is force exerted on the specimen under tension (N), A o is initial area of cross-section (m2), ΔL is the change in the length of the specimen (m) and Lo is the initial length of the specimen (m).

2.4 Pervaporation experiments Pervaporation experiments were carried out at room temperature and the pressure on the permeate side was equal to 300 Pa. As the feed solution, an aqueous solution of 97 wt% ethanol was used. The membrane (effective area of 112·10-4 m2) was placed in the cell where the feed solution was loaded. The permeate was condensed in a vacuum trap immersed in liquid nitrogen (-196 0C). Flux was calculated from the measured weight of liquid collected in the cold traps during a certain time intervals at steady-state condition. The feed, permeate and retentate composition was analysed by a gas chromatography on PerkinElmer Clarus 500 GC

equipped with 30 m elite-WAX ETR column and a flame ionization detector (FID). For each of the membrane used, the experiment was repeated three times. The results, presented an a mean ± standard deviation, showed the repeatability of measurements and the errors of the order of a few percent only. The permeation flux of component i was calculated using the following equation [28-29]: (3) where mi – weight of component i in permeate [g], A – effective membrane area [m2], t – permeation time [s]. Knowing the flux and the difference of vapour pressure at both sides of the membrane it is possible to estimate the permeability coefficient of component i [28-29]: (4)

where P – permeation coefficient [g·m/N·s],l – membrane thickness [m],Δp – difference of vapour pressure at both sides of the membrane [Pa], Js –flux [g/m2·s]. For describing the separation properties of the membrane, three parameters are used: separation factor (αAB),selectivity coefficient (ScAB) and enrichment factor (βAB). Separation factor was calculated by [28-29]: (5) where xA , xB – weight fraction of component A, B in the feed [wt.%], yA, yB – weight fraction of components A, B in permeate [wt.%]. Selectivity coefficient was equal to the ratio of permeability of separated components [28-29]: (6) In order to compare the separation efficiency of different investigated membranes, pervaporation separation index expressed by following equation [28-29] was used: (7)

where J – total permeate flux [g/m2·s], αAB – separation factor.

3. Results and discussion 3.1 Membrane characterization 3.1.1. FTIR studies The investigation of surface functional groups was carried out using FTIR analysis for noncrosslinked (Alg), calcium chloride (AlgCa), glutaraldehyde (AlgGA), phosphoric acid (AlgP) and citric acid crosslinked alginate membranes; the corresponding spectra were shown in Fig. 2. The unmodified alginate showed characteristic bands at 1598, 1410, and 1084 cm-1 corresponding to COO, C-O(H), and C-O-C (ring) vibrational modes, respectively. Calcium chloride used as crosslinker does not affect on the structural properties of alginate as a consequence of the ionic interaction of the chains of polymer which form electronegative cavities with calcium cations. After phosphorylation, new bands are observed at 1681, 1235, 1072, 977, and 928 cm-1. These are assigned to water, P=O antisymmetric stretching, P-O-C (aliphatic), and two P-O(H) vibrational modes, respectively. These bands are similar to those observed in [30]. Crosslinking reactions by glutaraldehyde and citric acid is confirmed by IR spectra which show two sets of significant changes: an increase in the absorbance of peaks at 1730 cm-1, which is attributable to the characteristics of the aldehyde or carboxylic acid, and 1240 cm-1, which could be assigned to the formation of an acetal ring and ether linkage as a result of the reaction between the hydroxyl groups of SA and the aldehydes of GA or carboxylic of CA. The significant aldehyde peaks are also found as observed in the reaction between Alg and GA [31].

FIGURE 2

3.1.2. DSC studies The differential scanning calorimetry is a very useful tool for the determination of glass transition and melting behaviour of polymer materials. Initially, a study of the thermal behaviour of crosslinked Alg was performed under N2 atmospheres. Since the nature of alginate polymer containing hydrophilic groups, usually presence of strong interaction with water molecules is expected. Since the humidity may influences on the operative properties of polymer matrix, DSC analysis allow to probe the states of water contained in membranes. Under heating, crosslinked alginates present initially a dehydration process followed usually by the complex multi-step decomposition of crosslinker and/or polymer backbone. The DSC thermograms of crosslinked pristine Alg membranes are presented in Fig. 3. The first endothermic peak below 100°C is connected with the dehydration process of loosely bounded water. In case of crosslinked alginate matrix, a shift of dehydration peak to higher temperatures is observed [32]. Depends on the type of crosslinker used in our study sodium alginate loses water between 71-93.5°C. Gonzales-Rodriguez et al. reported that observed changes depend mainly on the degree of crosslinking and the nature of the multivalent cation [32]. The lowest dehydration temperature, i.e. 71 and 77°C, was found for the two acid applied as crosslinker – tricarboxylic citric acid and inorganic phosphoric acid, respectively. Similar to the results of Gonzales-Rodriguez the highest dehydration temperature is observed in case of AlgCa membrane, indicating the strongest bounding of water molecules by the Ca 2+ ions. The presence of metal oxides additive in the hybrid membranes further influences on the dehydration process [33]. Magnetite (Fe3O4) is known for the easy oxidation on the surface and formation of hydrophilic layer of corresponding hydroxide. For all studied membranes shifts of the dehydration peak toward higher temperatures is noticed. This phenomenon is the most pronounced in case of the hydrophobic AlgGA membrane crosslinked with

glutaraldehyde. For this matrix very broad dehydration endotherm with maximum about 130°C is observed. The sodium alginate decomposes in two steps after the dehydration process. It is postulated, that the decomposition of the polymer leads to the formation of carbonate Na2CO3 and carbonized residue and is represented by an broad exothermic peak between 200-270°C. Finally, at the higher temperature subsequent decomposition of the carbonaceous residue occurs due to the pyrolysis reaction [34]. The decomposition of crosslinked matrix have somewhat more complex mechanism involving several endo- and exothermic steps, depending on the nature and decomposition mechanism of crosslinker. Usually an exothermic peak followed by an endothermic peak has appeared. The endothermic peak connected with the polymer melting and breaking of crosslinks appears just below 200°C (186-198) and is followed by the broad exothermic peak of decomposition. Only in case of AlgCa membrane the exothermic decomposition process occurring in the range of 200-315°C overlaps an endothermic process with maximum about 230°C, connected with the polymer melting. Such shape of the thermogram curve indicates strong effects between calcium ions and polymer matrix, and is typically attributed to the decomposition resulting in the carbonized residue. Furthermore, it was also observed for other AlgCa membranes independent on the type of oxide , inter alia, Ag2O and ZnO [33].

FIGURE 3

3.1.3. Swelling and contact angle measurements The results of swelling experiments for pristine and hybrid crosslinked sodium alginate membranes were shown in Fig.4. As a crosslinkers calcium chloride, glutaraldehyde, phosphoric acid and citric acid were used. Three of them (AlgCa, AlgP and AlgC) have

hydrophilic character and crosslinking with these agents cause bigger membrane swelling. The degree of swelling (DS) reached 180% after 1.5 h. Different result is obtained for glutaraldehyde alginate membranes. GA has hydrophobic properties and crosslinking with this agent introduce more hydrophobic fragments to sodium alginate membranes. In this case, DS is about 80%. Addition of iron oxide nanoparticles into polymer matrix changes the hydrophilic-hydrophobic balance resulting in continuous decreasing of water uptake. Since ferroferric oxide particles have hydrophobic properties, the most influence is observed for alginate membranes with hydrophilic character of crosslinkers. For the highest loaded membranes the DS was equal to 96, 105 and 119 % for AlgC, AlgP and AlgCa, respectively. Loading AlgGA membranes with iron oxide nanoparticles also caused the reduction of degree of swelling but in this case this change was not significant.

FIGURE 4

The consistent results was obtained in case of contact angle measurement. Contact angle measurements describe the surface hydrophilicity of the membrane for the prediction of the membrane performance. Each experiment was performed eight times to guarantee a statistical integrity. The contact angle values of pristine membranes are equaled 31.5 ± 0.2 0, 34.0 ± 0.20, 38.9 ± 0.20 and 55.6 ± 0.20 for AlgCa, AlgP, AlgC and AlgGA, respectively (SI Table 1). All investigated membranes are characterized as hydrophilic (q < 90 0). The hydrophilicity depends on the crosslink moiety and increase in following order: AlgGA < AlgC < AlgP < AlgCa (Fig.5A). Like in the case of swelling, the three types of membrane show similar contact angles and are more hydrophilic that AlgGA. The higher contact angle of AlgGA membrane results from the hydrophobic nature of the glutaric aldehyde used as a crosslinking agent. In all cases, the addition of iron oxide particles increases slightly the contact angle. For

AlgCa, AlgP, AlgC membranes, contact angle reaches the values in the range 42-480. In case of more hydrophobic AlgGA membrane increase in the contact angle is less pronounced and is equal only 2-30. (Fig. 5B).

FIGURE 5 A, B

3.1.4. Structural properties The surface morphology of sodium alginate membranes observed by SEM revealed that both pristine and hybrid cross-linked sodium alginate membranes have a homogeneous, dense surface. The size distribution of the iron oxide particles as found from a TEM measurements is presented in Fig. 6. The determined D50 of magnetite particles equals to 13.62 nm and is typical for the in situ hydrolysis and precipitation method of Fe3O4 preparation [35]. The smallest magnetite particles have 3.18 nm and the biggest 42.13 nm (SI Fig.1). As shown in the SI Fig. 2 prepared nanoparticles are finely dispersed in the cross-section of hybrid membranes. In polymer matrix aggregation phenomena of magnetite is observed, because of the existing magnetic attraction forces allowing to overcame the viscosity of polymer solution. The measured size of iron oxide particles and agglomerates ranges from 19.8 to 2870 nm and their distribution is shown in Fig. 6. The biggest aggregates with diameter in order of several micrometers appear in membrane rarely. In case of magnetite in membranes the determined D50 equals to 53.15 nm and is clearly bigger than found from a TEM measurement.

FIGURE 6

3.1.5. Mechanical properties

The mechanical properties of pristine and hybrid AlgCa, AlgGA, AlgP andAlgC membranes with 15 wt% of magnetite content were presented in Fig. 7. Generally, the pristine membranes had the highest value of Young modulus. In all cases, the addition of iron oxide particles into polymer matrix caused the decrease in the mechanical strength of the membranes. The most significant difference of mechanical properties between pristine and hybrid membrane was shown for alginate membranes crosslinked with calcium chloride. This membrane was also the least mechanically stable. Nearly the same values of Young's modulus were observed for membranes crosslinked with citric and phosphoric acid. For these membranes, the addition of magnetite only slightly influenced their mechanical properties. The best mechanical strength was reached by the pristine AlgGA membrane. These good mechanical properties were reduced with the presence of iron oxide particles in the alginate matrix and reached the same value as for AlgP and AlgC membranes.

FIGURE 7

3.2 Pervaporation performance of alginate-based membranes 3.2.1 The calcium chloride crosslinked hybrid sodium alginate membranes The evaluated parameters describing transport properties i.e., normalized flux, permeation coefficients and selectivity coefficients of ethanol and water in pervaporation process through pristine and hybrid sodium alginate membranes crosslinked with calcium chloride, with different amount of iron oxide nanoparticles were collected in SI Table 2. According to literature [36] Ca2+ cross-linking is a common method to improve alginate membrane performance. Huang et al. [37] investigated alginate membranes crosslinked with six different metal ions in pervaporative ethanol and isopropanol dehydration. It was proved, that the highest separation performance for both alcohol mixtures was achieved by the crosslinking

with Ca2+. The difference between the values of fluxes and separation efficiency after crosslinking by various metal ions was explained by another closely packed membrane structure and the thermodynamic interaction difference between the permeating component and the counter ions. For pristine AlgCa membrane, normalized flux and water permeation coefficient were equal to 0.17 kg·m-2·h-1·μm-1 and 4.83·10-5 Barrer, respectively. After the addition of magnetite, both parameters increased, reaching the values of 0.48 kg·m-2·h-1·μm-1 and 24.42·10-5 Barrer for 25 wt% of iron oxide loaded. The increase in total flux and permeation coefficients was probably due to the structure of the obtained membranes, which become more sponge-like. The another relation was noticed in case of selectivity coefficient. The highest value of Sc was equal to 48.26 and was achieved for AlgCa membrane with 15wt% of iron oxide nanoparticles. For larger amounts of magnetite, the selectivity coefficient decreased to 23.31 for 25 wt% of ferroferric oxide nanoparticles but was still greater than for pristine alginate membrane. This phenomenon could be explain by the difference in the diameters of the molecules of water and ethanol . The water diameter equals 0.28 nm and is smaller than the diameter of ethanol (0.44 nm). In consequence, water molecules are able to pass through the small pores (when less magnetite is present in the alginate matrix), whereas the transport of ethanol molecules is limited/restricted. In case of 20 and 25 wt% of ferroferric oxide nanoparticles, the larger free space is formed, through which both water and ethanol molecules can easily permeate [38-39]. Regarding the parameters describing effectiveness of membranes, i.e. separation factor, enrichment factor and PSI, the best membranes for ethanol/water separation by pervaporation process appeared to be AlgCa membrane with 15 wt% of magnetite loaded. For this type of membrane, the obtained value of αH2O/EtOH, βH2O and PSI were equal to 58.76, 46.04 and 21.07 kg·m-2·h-1·μm-1, respectively. In this case pervaporation separation index was about five times greater than for pristine membrane.

3.2.2 The glutaraldehyde crosslinked hybrid sodium alginate membranes The evaluated parameters i.e. normalized flux, permeation coefficient and selectivity coefficient, describing transport properties, and separation factor, enrichment factor and pervaporation separation index, describing effectiveness of membranes, of ethanol and water in pervaporation process through glutaraldehyde crosslinked alginate membranes with magnetite loaded were collected in SI Table 3. The results showed that the evaluated values for pristine glutaraldehyde crosslinked alginate membrane were similar to obtained in [40]. The total normalized flux was equal to 0.26 kg·m-2·h-1·μm-1 and was larger by 0.09 kg·m-2·h1

·μm-1 from the pristine alginate membrane cross-linked with calcium chloride. Wang et al.

[41] found that when a hydrophilic membrane was crosslinked with glutaraldehyde, its degree of crystallinity decreased, resulting in an improved flux. According to the character of hybrid polymer membranes, after adding magnetite to polymer matrix, the total flux increased reaching a value of 0.53 kg·m-2·h-1·μm-1 for the membrane containing 25 wt% of iron oxide. The same behavior was observed in case of the permeation coefficients. The water permeation coefficient increased from 8.06 Barrer for pristine membrane to 31.73 Barrer for 25 wt% of magnetite loaded. At the same time, the permeation coefficient of ethanol slightly increased with the increase in iron oxide particle content in alginate matrix. This difference between water and ethanol permeation coefficients influenced the value of selectivity coefficient. For pristine alginate membrane, selectivity coefficient was equal to 15.20. The addition of iron oxide nanoparticles caused the decrease in the value of this parameter. Just above 20 wt% of magnetite content, the selectivity coefficient increased and reached the value of 16.71 for 25 wt%. This phenomenon is explained by the fact that glutaraldehyde is miscible in both water and ethanol, indicating that aldehyde group has affinity towards both of them. However, the acetal group has a good affinity toward alcohol rather than water. From these facts, it could be

postulated that the glutaraldehyde crosslinked alginate membrane would have less affinity to water (i.e., less preferential sorption of water) [38]. The same relation was observed in case of separation

factor

and

pervaporation

separation

index.

For

pristine

alginate

membranes,αH2O/EtOH and PSI were equal to 23.62 and 5.88 kg·m-2·h-1·μm-1, respectively. Addition of the magnetite to alginate matrix reduced these parameters up to 15 wt% of iron oxide particles, to the values of 9.78 and 3.42 kg·m-2·h-1·μm-1, respectively. This fact showed that membranes cross-linked by glutaraldehyde were less hydrophilic due to the consumption of OH groups, and these membranes were more rigid, leading to greater obstruction to diffusion [42]. Further increase in the amount of magnetite in the membrane resulted in the increase in αH2O/EtOH and PSI, that for 25 wt% of magnetite loaded reach the values of 24.65 and 5.93 kg·m-2·h-1·μm-1, respectively. In consequence, only this membrane had the better parameters comparing with pristine glutaraldehyde crosslinked alginate membranes and could be used in pervaporation process.

3.2.3 The phosphoric acid crosslinked hybrid sodium alginate membranes The evaluated parameters (i.e. normalized flux, permeation coefficient, selectivity coefficient, separation factor, enrichment factor and pervaporation separation index) describing the transport of water and ethanol from the 96% ethanol solution through phosphoric acid crosslinked alginate membrane were presented in SI Table 4. The results showed that phosphorylated forms of alginate are promising for breaking the ethanol azeotropic barrier and in consequence, the pervaporation with the use of these membranes can be an advantageous technique for ethanol dehydration. The normalized flux for pristine phosphoric acid alginate membrane was equal to 0.27 kg·m-2·h-1·μm-1and was higher than for pristine AlgCa membrane. This is due to the fact that the AlgP membrane has a high affinity for water and is very hydrophilic in nature. The ionically cross-linked membrane structure effectively

excludes the permeation of ethanol, while the total and water permeabilities increase. This phenomenon can be explained by the fact that ionized groups hydrate easily and exclude organic solvents. Moreover, residual-OH groups of alginate are also available for interaction with water molecules through hydrogen bonding [43]. The obtained selectivity coefficient, separation factor and PSI were equal to 33.07, 55.83 and 14.96 kg·m-2·h-1·μm-1, respectively, and they corresponded with the results obtained in [27]. After adding iron oxide nanoparticles, the values of evaluated parameters were changed. In range from 5 to 15 wt% of magnetite content, the rise in all received parameters was observed. For 15 wt% of ferroferric oxide particles in AlgP matrix the obtained selectivity coefficient, separation factor and PSI reached the values of 60.42, 90.46 and 33.86 kg·m-2·h-1·μm-1, respectively, and they were about 2 times greater than for pristine AlgP membrane. Further addition of magnetite influenced the decrease in the values of received parameters. The normalized total flux was reduced to 0.13 kg·m-2·h-1·μm-1 for 25 wt% iron oxide nanoparticles. In consequence, all evaluated parameters i.e., Sc, αH2O/EtOH, βH2O and PSI also decreased, reaching smaller values than for pristine AlgP membrane. This phenomenon was associated with the nature of hybrid membranes. In hybrid membranes, inorganic components are dispersed in a polymer matrix. With low inorganic particles loading, the contribution of passes through membranes molecules firmly immobilized on the inorganic particles surface was inconsiderable low, and improvement effect contributed to the free volume formation was dominated in the separation process, and increasing of the permeation flux was observed. With excess inorganic particles loading, molecules stayed immobilized on the inorganic particles surface, and the permeation flux decreased [44].

3.2.4 The citric acid crosslinked hybrid sodium alginate membranes

The evaluated parameters i.e. normalized flux, permeation coefficient and selectivity coefficient, describing transport properties, and separation factor, enrichment factor and pervaporation separation index, describing effectiveness of membranes, of ethanol and water in pervaporation process through citric acid crosslinked alginate membranes with iron oxide nanoparticles were collected in SI Table 5. Citric acid has three carboxylic functional groups that are responsible for the high ability to be a good crosslinker. For this reason, the researchers frequently use citric acid in the formation of crosslinked chitosan and poly(vinyl alcohol) membranes [45-46]. The investigation on such membranes showed that chitosan and poly(vinyl alcohol) citrate formed hydrocolloidal matrices and stable membranes. Our results showed that the total normalized flux of pristine citric acid crosslinked alginate membrane was the lowest among all investigated membranes and was equal to 0.15 kg·m-2·h1

·μm-1. This phenomenon was explained by Burshe et al. [46]. They applied poly(vinyl

alcohol) membranes crosslinked with citric acid, adipic acid, maleic acid, glutaraldehyde and glyoxal in pervaporative separation of acetone/water and isopropanol/water systems. They observed a low flux values for citric acid crosslinked membranes is due to the trifunctional carboxylic acid groups of citric acid which gave a larger number of sites for crosslinking between different polymer molecules. It caused the formation of a more compact structure of the membrane causes by the reduction of chain mobility. After adding magnetite into alginate matrix, the value of total normalized flux increased but this growth was very small. Higher flux appeared only at 25 wt% of iron oxide particles. The formation of the ester groups as a result of crosslinking between alginate and citric acid influenced the water and ethanol permeability. According to the nature of the ester groups, that have a hydrophilic character, therefore AlgC membranes possess good affinity to water. As a result, the high values of selectivity coefficients, separation factor, enrichment factor and

PSI were observed, which further increased after the addition of magnetite. The largest values of αH2O/EtOH, βH2O and PSI that equal to 133.24, 99.59 and 24.21 kg·m-2·h-1·μm-1, respectively, were obtained for alginate membrane with 20 wt % of magnetite. For further addition of iron oxide particles, the values of considered parameters decreased due to the larger free volumes in the polymer matrix.

3.3 Comparison of alginate membranes crosslinked with different crosslinking agents The comparison of investigated Alg membranes according to separation factor, flux and pervaporation separation index (PSI) was shown in Fig. 8. Fig. 8A presented the relation between separation factor and magnetite content in polymer matrix for different crosslinking alginate membranes. For pristine membranes, the values of α H2O/EtOH were distributed in two groups. Membranes crosslinked with citric and phosphoric acid were characterized by higher values of separation factor equal to 64.31 and 55.83, respectively. Crosslinking of alginate membranes with glutaraldehyde and calcium chloride caused the less separation of water/ethanol mixture. In this case,αH2O/EtOH for AlgGA and AlgCa membranes reached the values of 23.62 and 26.48, respectively. Addition of iron oxide particles into polymer matrix affected the increase in separation factor for membranes crosslinked with calcium chloride, citric and phosphoric acid. Due to the fact that glutaraldehyde has less hydrophilic properties, magnetite does not affect the improvement of separation but on the deterioration of this coefficient. For each crosslinked membranes, it can be distinguished that the extreme (minimum or maximum) value of separation factor occurs at 15 or 20 wt% of iron powder. Further addition of magnetite reduces αH2O/EtOHfor AlgC, AlgP and AlgCa membranes, and increases for AlgGA membrane. The highest values of separation factor that equal to 109.91 and 133.24 reach the membranes crosslinked with citric acid with 15 and 20 wt% of ferroferric oxide content, respectively. The another relation is observed in case of fluxes (Fig.

8B). For pristine membranes, the highest values of this parameter are characterized by the membranes crosslinked with glutaraldehyde and phosphoric acid. The addition of magnetite change this tendency, showing that also for AlgCa membrane is possible to achieve high fluxes. As it was aforementioned, the trifunctional carboxylic acid groups of citric acid influence on a low flux values for AlgC membranes. The highest values of fluxes are reached by AlgGA and AlgCa membranes with 25 wt% of magnetite, that equal to 0.53 and 0.48 kg·m-2·h-1·μm-1, respectively. FIGURE 8 A,B,C

Pervaporation separation index (PSI), which is a relative measure of the separation ability of a membrane, has been defined as the product of total permeation and separation factor. This index can be used as a relative guideline index for the design of pervaporation membrane separation processes. Fig. 8C shows the variation of PSI as a function of magnetite content in hybrid AlgCa, AlgP, AlgGA and AlgC membranes. It is found that for pristine alginate membranes the highest value of PSI has AlgP membrane. Addition of iron oxide particles influences the increase in PSI values. The tendency of the greatest values of PSI for AlgP membranes remains the same. Despite the fact that the glutaraldehyde crosslinked membranes show the highest flux values, PSI are the lowest because of the small value of the separation coefficients. Among all investigated membranes, membrane with an optimal combination of flux and selectivity show AlgP membrane with 15 wt% of magnetite. For this membrane, PSI equals to 33.86 kg·m-2·h-1·μm-1 and is about ten times higher than for AlgGA membrane with the same amount of iron oxide particles. AlgP membranes also show good mechanical properties that only slightly deteriorate with the addition of magnetite.

4. Conclusion

The paper presents the results of water/ethanol mixture pervaporation studies through alginate membranes loaded with ferroferric oxide, crosslinked with four different agents i.e. calcium ions (AlgCa), phosphoric acid (AlgP), glutaraldehyde (AlgGA) and citric acid (AlgC). The investigated membranes were characterized by the swelling behavior, contact angle, FTIR, DSC, SEM, TEM and Tensile strength measurements. The experimental study led to the following significant conclusions: 

In the context of the determination of parameters describing transport properties, the highest values of fluxes are shown for the membranes crosslinked with glutaraldehyde, phosphoric acid and calcium chloride.



The best separation properties are exhibited by AlgC and AlgP membranes, that for AlgC with 20 wt% of magnetite and AlgP with 15 wt% of iron oxide loaded, α H2O/EtOH reaches the values of 133.24 and 90.46, respectively.



The addition of iron oxide particles in polymer matrix influences the ethanol/water separation, giving the higher values of fluxes and separation factors.



Combination of flux and separation coefficient, indicating the separation ability, show that among all investigated membranes, this is alginate membrane with 15 wt% of magnetite loaded, crosslinked with phosphoric acid, that has the highest value of PSI equal to 33.86 kg·m-2·h-1·μm-1.

References [1]A. Abe, K. Dušek, S. Kobayashi, Biopolymers, Springer-Verlag Berlin Heidelberg, 2010. [2] M. Niaounakis, Biopolymers: Processing and Products, William Andrew, Elsevier, UK, 2015. [3] M. Niaounakis, Biopolymers: Applications and Trends, William Andrew, Elsevier, UK, 2015. [4] M.U. Adikwu, Biopolymers in Drug Delivery: Recent Advances and Challenges, Benthan Science Publishers, 2009. [5] S. Kalia, L. Avérous, Biopolymers: Biomedical and Environmental Applications, 1st ed., Wiley-Scrivener, Salem, Massachusetts, 2011.

[6] T. Heinze, Polysaccharides I: Structure, Characterization and Use, Springer-Verlag Berlin Heidelberg, 2005. [7] P.A. Williams, Renewable Resources for Functional Polymers and Biomaterials, RCS Publishing, UK, 2011. [8] M.E. Molina, A.J. Quiroga, Alginates: Production, Types and Applications, Nova Science Publishers, New York, 2012. [9] Z.A. Butt, Alginate: Biosynthesis and Properties, LAP LAMBERT Academic Publishing, 2014. [10] F.U.Nigiz, N.D.Hilmioglu, Pervaporation of ethanol/water mixtures by zeolite filled sodium alginate membrane, Desalination Water Treat. 51 (2013) 637-643. [11] S. Moulik, S. Nazia, B. Vani, S. Sridhar, Pervaporation separation of acetic acid/water mixtures through sodium alginate/polyaniline polyion complex membrane, Sep. Purif. Technol.170 (2016) 30–39. [12] S. Thakur, S. Pandey, O.A. Arotiba, Development of a sodium alginate-based organic/inorganic superabsorbent composite hydrogel for adsorption of methylene blue, Carbohydr. Polym.153 (2016) 34-46. [13] C. Wu, Y. Wang, B. Gao, Y. Zhao, Q. Yue, Coagulation performance and floc characteristics of aluminum sulfate using sodium alginate as coagulant aid for synthetic dying wastewater treatment, Sep. Purif. Technol. 95 (2012) 180-187. [14] H.Ch. Vu, A.D. Dwivedi, T.T. Le, S.-H. Seo, E.-J. Kim, Y.-S. Chang, Magnetite graphene oxide encapsulated in alginate beads for enhanced adsorption of Cr(VI) and As(V) from aqueous solutions: Role of crosslinking metal cations in pH control, Chem. Eng. J. 307 (2017) 220-229. [15] J.E. Varg, J. Dussán, Encapsulation and immobilization of the S-layer protein of Lysini bacillus sphaericus in an alginate matrix for chromium adsorption, Int. Biodeterior. Biodegrad. 116 (2017) 141-146. [16] R.W. Baker, Membrane Technology and Applications, 3st ed., Wiley&Sons, New York, 2012. [17] N. Hilal, M. Khayet, Ch.J. Wright, Membrane Modification: Technology and Applications, CRC Press, Boca Raton, London, New York, 2012. [18] A.Basile, A. Figoli, M. Khayet, Pervaporation, Vapour Permeation and Membrane Distillation, 1st ed.; Woodhead Publishing: Elsevier, 2015. [19] Q. Wang, N. Li, B. Bolto, M. Hoang, Z. Xie, Desalination by pervaporation: A review. Desalination 387 (2016) 46-60. [20] N. Wang, G. Zhang, L. Wang, J. Li, Q. An, S. Ji, Pervaporation dehydration of acetic acid using NH2-UiO-66/PEI mixed matrix membranes, Sep. Purif. Technol. 186 (2017) 20-27. [21] S. Xu, L. Liu, Y. Wang, Network cross-linking of polyimide membranes for pervaporation dehydration, Sep. Purif. Technol. 185 (2017) 215-226. [22] A.M. Sajjan, B.K.J. Kumar, A.A. Kittur, M.Y. Kariduraganavar, Novel approach for the development of pervaporation membranes using sodium alginate and chitosan-wrapped multiwalled carbon nanotubes for the dehydration of isopropanol, J. Membr. Sci. 425– 426 (2013) 77–88. [23] G. Dudek, M. Krasowska, R. Turczyn, M. Gnus, A. Strzelewicz, Structure, morphology and separation efficiency of hybrid Alg/Fe3O4 membranes in pervaporative dehydration of ethanol. Sep. Purif. Technol., 182 (2017) 101-109. [24] G. Dudek, M. Gnus, R. Turczyn, A. Strzelewicz, M. Krasowska, Pervaporation with chitosan membranes containing iron oxide nanoparticles, Sep. Purif. Technol. 133 (2014) 8-15.

[25] G. Dudek, A. Strzelewicz, R. Turczyn, M. Krasowska, A. Rybak, The Study of Ethanol/Water Vapors Permeation through Sulfuric Acid Cross-Linked Chitosan Magnetic Membranes, Sep. Sci. Technol. 49(2014)1761-1767. [26] G. Dudek, R. Turczyn, A. Strzelewicz, M. Krasowska, A. Rybak, Z.J. Grzywna, Studies of separation of vapours and gases through composite membranes with ferroferric oxide magnetic nanoparticles, Sep. Purif. Technol.109 (2013) 55-63. [27] M. Gnus, G. Dudek, R. Turczyn, A. Tórz, D. Łącka, K. Konieczny, M. Łapkowski, Pervaporative investigation of ethyl alcohol dehydration. Prog. Chem. Appl. Chit. 20 (2015) 54-63. [28] J.G.Wijmans, R.W.Baker, The Solution Diffusion Model - A Review. J. Membr. Sci. 107 (1995) 1-21. [29] C. J. Geankoplis, Transport Processes and Separation Process Principles, 4th ed., Prentice Hall Press Upper Saddle River, NJ,, USA, 2013. [30] R.J. Coleman, G. Lawrie, L.K. Lambert, M. Whittaker, K.S. Jack, L. Grøndahl, Phosphorylation of Alginate: Synthesis, Characterization, and Evaluation of in Vitro Mineralization Capacity, Biomacromolecules 12 (2011) 889–897. [31] C.K. Yeom, K.H Lee, Characterization of sodium alginate membrane crosslinked with glutaraldehyde in pervaporation separation, J. Appl. Polym. Sci.67 (1998) 209–219. [32] M.L. Gonzalez-Rodriguez, M.A. Holgado, C.Sanchez-Lafuente, A.M. Rabasco, A. Fini, Alginate/chitosan particulate systems for sodium diclofenac release, Int. J. Pharmaceut. 232 (2002) 225-234. [33]M. Ali, M. Zafar, T. Jamil, M.T.Z. Butt, Influence of glycol additives on the structure and performance of cellulose acetate/zinc oxide blend membranes, Desalination270 (2011) 98-104. [34] J. Zhang, Q. Ji, , X. Shen, Y. Xia, L. Tan, Q. Kong, Pyrolysis products and thermal degradation mechanism of intrinsically flame-retardant calcium alginate fibre, Polym. Degrad. Stabil. 96 (2011) 936–942. [35] Ch. Cheng, F. Xu, H. Gu, Facile synthesis and morphology evolution of magnetic iron oxide nanoparticles in different polyol processes. New J. Chem. 35 (2011) 1072–1079. [36] Ch. Gao, M. Zhang, J. Ding, F. Pan, Z. Jiang, Y. Li, J. Zhao, Pervaporation dehydration of ethanol by hyaluronic acid/sodium alginate two-active-layer composite membranes, Carbohydr. Polym. 99 (2014) 158–165. [37] R.Y.M. Huang, R. Pal, G.Y. Moon, Characteristics of sodium alginate membranes for the pervaporation dehydration of ethanol-water and isopropanol-water mixtures, J. Membr. Sci. 160 (1999) 101-113. [38] M.J. Carmo, J.C. Gubulin, Ethanol-water adsorption on commercial 3A zeolites: kinetic and thermodynamic data, Braz. J. Chem. Eng. 14 (1997). [39] P. Schatzberg, Molecular diameter of water from solubility and diffusion measurements, J. Phys. Chem. 71 (1967) 4569–4570. [40] S.G. Adoor, V. Rajineekanth, M.N. Nadagouda, K.C. Rao,D.D. Dionysiou, T.M. Aminabhavi, Exploration of nanocomposite membranes composed of phosphotungstic acid in sodium alginate for separation of aqueous–organic mixtures by pervaporation, Sep. Purif. Technol. 113 (2013) 64–74. [41] L.Y. Wang, J.D. Li, Y.Z. Lin, C.X. Chen, Crosslinked poly(vinylalcohol) membranes for separation of dimethyl carbonate/methanol mixtures by pervaporation, Chem. Eng. J. 146 (2009) 71–78. [42] Z. Jia, G. Wu, Metal-organic frameworks based mixed matrix membranes for pervaporation. Microporous Mesoporous Mater.235 (2016) 151-159.

[43] P.S. Rao, S. Sridhar, M.Y. Wey, A. Krishnaiah, Pervaporative Separation of Ethylene Glycol/Water Mixtures by Using Cross-linked Chitosan Membranes, Ind. Eng. Chem. Res. 46 (2007) 2155-2163. [44] L. Lu, H. Sun, F. Peng, Z. Jiang, Novel graphite-filled PVA/CS hybrid membrane for pervaporation of benzene/cyclohexane mixtures, J. Membr. Sci. 281 (2006) 245–252. [45] T. Phaechamud, T. Koizumi, G.C. Ritthidej, Chitosan citrate as film former: compatibility with water-soluble anionic dyes and drug dissolution from coated tablet, Int. J. Pharm. 198 (2000) 97–111. [46] M.C. Burshe, S.A. Netke, S.B. Sawant, J.B. Joshi, V.G. Pangarkar, Pervaporative Dehydration of Organic Solvents, Sep. Sci. Technol. 32 (1997) 1335-1349.

FIGURE CAPTIONS FIGURE 1 General scheme of the pristine and hybrid AlgCa, AlgP, AlgGA and AlgC membranes preparation.

FIGURE 2 FTIR spectra of non-crosslinked (Alg), calcium chloride (AlgCa), glutaraldehyde (AlgGA), phosphoric acid (AlgP) and citric acid (AlgCA) crosslinked alginate membranes in the spectral range of 800–1800 cm-1.

FIGURE 3 DSC thermograms of calcium chloride (AlgCa), glutaraldehyde (AlgGA), phosphoric acid (AlgP) and citric acid (AlgCA) crosslinked alginate membranes under N 2.

FIGURE 4 Dependence of degree of swelling versus amount of magnetite particles in the AlgCa, AlgP, AlgC and AlgGA membranes.

FIGURE 5 Average contact angles C of (A) pristine alginate membranes versus crosslinking species and (B) hybrid alginate membranes versus magnetite content.

FIGURE 6 Size distribution of magnetite particles determined from TEM and in the membrane matrix (SEM)

FIGURE 7 The values of tensile strength for AlgCa, AlgP, AlgC and AlgGA membranes.

FIGURE 8 The plot of (A) separation factor (B) flux (C) max PSI versus corresponding magnetite content for hybrid AlgCa, AlgP, AlgGA and AlgC membranes.