Highly water selective silicotungstic acid (H4SiW12O40) incorporated novel sodium alginate hybrid composite membranes for pervaporation dehydration of acetic acid

Highly water selective silicotungstic acid (H4SiW12O40) incorporated novel sodium alginate hybrid composite membranes for pervaporation dehydration of acetic acid

Separation and Purification Technology 54 (2007) 178–186 Highly water selective silicotungstic acid (H4SiW12O40) incorporated novel sodium alginate h...

674KB Sizes 3 Downloads 242 Views

Separation and Purification Technology 54 (2007) 178–186

Highly water selective silicotungstic acid (H4SiW12O40) incorporated novel sodium alginate hybrid composite membranes for pervaporation dehydration of acetic acid夽 Shivanand B. Teli a , Gavisiddappa S. Gokavi a , Malladi Sairam b , Tejraj M. Aminabhavi b,∗ a


Kinetics and Catalysis Laboratory, Department of Chemistry, Shivaji University, Kolhapur 416 008, India Membrane Separations Division, Center of Excellence in Polymer Science, Karnatak University, Dharwad 580 003, India Received 14 June 2006; received in revised form 3 September 2006; accepted 3 September 2006

Abstract Silicotungstic acid incorporated sodium alginate (STA-NaAlg) hybrid composite membranes were prepared by incorporating 1, 2, 3 and 5 wt.% of silicotungstic acid (STA) into sodium alginate (NaAlg) and crosslinked with glutaraldehyde. Scanning electron microscopy and universal testing machine were used to investigate the morphology and the mechanical strength properties of the membranes. These membranes were tested for pervaporation (PV) dehydration of acetic acid at lower water concentrations of 10–25 wt.% in the feed. Addition of STA into NaAlg could result in a dramatic increase of water selectivity over that of plain NaAlg membrane dehydrating acetic acid for lowest STA containing NaAlg membrane. Thus, infinite selectivity was observed for STA-NaAlg membrane containing 1 wt.% of STA for all feed mixture compositions ranging from 10 to 25 wt.% water. With increasing STA content from 2, 3 and 5 wt.% in NaAlg membranes, selectivities ranged from 22,491 to 288, 12,848 to 192 and 6914 to 108, respectively for the studied composition range of the feed mixtures. PV performances of STA-NaAlg hybrid composite membranes were also tested at 40, 50, 60 and 70 ◦ C typically in case of 10 wt.% water-containing feed mixture, which indicated a decrease in selectivity and increase in flux for all hybrid composite membranes. Arrhenius plots of flux data versus reciprocal of temperature exhibited linear trends. The hydrophilic STA is responsible to offer increased selectivity and flux to water as compared to plain NaAlg membrane for the feed mixtures studied. Of all the membranes tested, the 1 wt.% STA containing NaAlg membrane exhibited the best PV performance characteristics. Temperature did not influence much the selectivity data of STA-NaAlg membrane containing 1 wt.% of STA up to 50 ◦ C, but flux increased with increasing temperature for all membranes. The results of this study are far superior to the previously reported data on acetic acid dehydration by the PV technique. Published by Elsevier B.V. Keywords: Pervaporation; Silicotungstic acid; Water–acetic acid mixture; Sodium alginate; Membrane; Activation energy

1. Introduction Acetic acid (HAc) is one of the most basic organic intermediates used in chemical and other allied industries, which finds applications in synthesizing vinyl acetate, terephthalic acid, phthalic anhydride, acetic anhydride, cellulose esters, etc. Acetic acid is also used as a solvent in plastics industries in the production of resins as well as in pharmaceutical and biochemical industries. Currently used processes for acetic acid production include carbonylation of methanol, liquid-phase oxidation of hydrocarbons and oxidation of acetaldehyde [1–3]. 夽 ∗

This article is CEPS communication #131. Corresponding author. Tel.: +91 836 2215372; fax: +91 836 2771275. E-mail address: [email protected] (T.M. Aminabhavi).

1383-5866/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.seppur.2006.09.002

It is however, difficult to separate acetic acid from industrial waste streams because acetic acid–water mixture does not form azeotrope, since the relative volatility of water and acetic acid is close to unity. Thus, their separation by conventional distillation becomes energy-intensive. Therefore, separation of acetic acid from water has been a challenging problem in chemical industries. In general, the more dilute the acetic acid solution is, the more attractive the solvent extraction becomes. In solvent extraction, acetic acid to water feed ratio is generally high, ranging from 2:1 to 5:1 [4]. Compared to other general types of solvent extraction and separation methods, from the energy-savings standpoint, pervaporation (PV) technique is a useful alternative method, since it is energy-intensive and environmentally clean. In recent years, PV has been the widely accepted technique for the separation of aqueous–organic mixtures. It is also a

S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186

promising technique for the fractionation of azeotropic mixtures, closely-boiling mixtures and mixtures consisting heat-sensitive compounds as well as low volatile aqueous solutions. In the past, hydrophilic and hydrophobic membranes have been used in PV dehydration of water–acetic acid mixtures [5–8]. Hydrophilic polymer membranes are preferred in PV dehydration studies of organics due to their strong affinity to water molecules. For PV dehydration of organic acids, new membranes with hydrophilic groups are preferred, which will absorb water molecules preferentially, leading to increased flux and selectivity. But, incorporation of hydrophilic groups in polymeric matrices will lead to polymer plasticization effect due to excessive membrane swelling. Swelling can be minimized by either crosslinking the polymer or by adding filler particles to improve the membrane performance. In recent years, research efforts in membrane-based separations have been devoted to the separation of aqueous alcohol mixtures using NaAlg membranes [9–11], but acetic acid–water separation has received relatively a lesser attention. Many membranes used in water–acetic acid separation have met with a limited success to achieve the desired values of flux and selectivity. NaAlg and its modified forms are among the various types of membranes used in PV dehydration of acetic acid. Wang [12] used the modified NaAlg composite membrane for the PV dehydration of acetic acid. Toti and Aminabhavi [13] used different viscosity grade NaAlg and modified NaAlg membranes in PV separation of water–acetic acid mixtures. Several reports have been published previously on the use of mixed matrix membranes by incorporating inorganic super acids (HPAs) into polymer matrix [14–16]. Such organic and inorganic hybrid compounds, when combined at the molecular level, have the proven advantages of both the organic materials of being light weight with a greater flexibility and good moldability, while inorganic components have high strength, thermal, and chemical stability [17]. HPAs have both acidic and redox catalytic properties [18], which show the characteristic adsorption behaviors, depending upon the properties of the adsorbates [19]. One of the distinctive features of HPAs is that they are highly soluble in polar solvents like water, alcohols and amines, but have generally poor solubility in non-polar solvents like benzene and olefins [20]. Taking advantages of their solubility properties, HPAs were blended with polymeric materials to form a membrane that has been used in catalytic applications [21,22]. In continuation of our ongoing efforts to develop novel types of membranes based on particulate filled NaAlg membrane with improved selectivity and flux [23–25], we now present some new experimental data on PV performance of STA filled-NaAlg membrane, for the first time used in acetic acid dehydration. The literature survey indicated that hitherto, no reports are available on the use of HPA-filled hybrid membranes of NaAlg for PV dehydration of water–acetic acid mixtures. The main objective of this study is to prepare of novel membrane of STA filled NaAlg hybrid composite membranes by incorporating different amounts of STA (1, 2, 3 and 5 wt.% based on the weight of NaAlg) into NaAlg matrix, and then crosslinking the membranes to achieve an effective dehydration of acetic acid. PV perfor-


mance of the hybrid membranes of this study is much better than plain NaAlg membrane giving an improved flux and selectivity to water. Pervaporation performance of these membranes was studied at 30, 40, 50, 60 and 70 ◦ C to evaluate the Arrhenius activation energy parameters using the PV results. Results are explained in terms of membrane–solvent interactions. 2. Experimental 2.1. Materials Sodium alginate, glutaraldehyde (GA), acetic acid (HAc), acetone and conc. hydrochloric acid (HCl), all of analytical reagent grade chemicals, were purchased from s.d. fine Chemicals, Mumbai, India. Silicotungstic acid (H4 SiW12 O40 ) was purchased from Loba Chemie, Mumbai, India. Deionized water, having a conductivity of 20 ␮S/cm, was produced in the laboratory itself from the Permeonics pilot plant (Vadodara, India) using the Nanofiltration membrane module. 2.2. Preparation of silicotungstic acid-filled NaAlg membranes Silicotungstic acid-filled NaAlg membranes were prepared by solution casting. Required amounts of silicotungstic acid (1, 2, 3 and 5 wt.% based on the weight of NaAlg) were dissolved in 100 mL of water, to which 4 g of NaAlg was added and stirred thoroughly for 24 h at the ambient temperature. The resulting homogeneous solution was cast uniformly on a glass plate with the help of a casting knife and dried at ambient temperature. Membranes were immersed in a crosslinking bath containing 180 mL of water–acetone mixture (1:2, v/v) with 2 mL of glutaraldehyde and 2 mL of conc. HCl for about 10 h. The crosslinked membranes were washed with deionized water to remove the excess amount of GA and HCl. Membranes were dried at ambient temperature. The resulting 1, 2, 3 and 5 wt.% STA-loaded NaAlg membranes were designated, respectively as: STA-NaAlg-1, STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5. The crosslinked pristine NaAlg membrane was prepared in the same manner as described before in the absence of STA. 2.3. Mechanical properties The equipment used for carrying out mechanical strength properties of the membranes was that of universal testing machine (UTM) (Model H25 KS Hounsfield, Surrey, United Kingdom). The test specimens were prepared in the form of dumbbell shapes as per the ASTM D-638 standards. Films of gauge length of 50 mm and width of 10 mm were stretched at a crosshead speed of 10 mm/min. Cross-sectional area of the sample of known width and thickness was calculated. Tensile strength was calculated using the equation:   maximum load tensile strength = (1) (N/mm2 ) cross-sectional area


S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186

2.4. Scanning electron microscopic (SEM) studies Surface morphology as well as cross-sectional view of the pristine NaAlg and STA (5 wt.%) incorporated NaAlg membranes were obtained using SEM, Hitachi, Model S-2150, Tokyo, Japan, equipped with Phoenix energy dispersive analysis of X-rays (EDAX). Since these films were non-conductive, gold coating (15 nm thickness) was done on the samples. 2.5. Degree of swelling Degree of swelling (DS) was performed gravimetrically in pure HAc and water at 30 ◦ C and in 10, 15, 20, 25 and 30 wt.% water containing acetic acid feed mixtures at 30, 40 and 50 ◦ C. Initial mass of the circularly cut (dia. = 2.5 cm) pristine NaAlg and STA-NaAlg hybrid composite membrane was measured on a single-pan digital microbalance (Model AE 240, Mettler, Greifensee, Switzerland) sensitive to ±0.01 mg. Samples were placed inside the specially designed airtight test bottles containing 30 cm3 of the test solvent. Test bottles were transferred to the oven maintained at constant desired temperature. Mass of the samples was measured at the selected time intervals by removing the test samples from the bottles and wiping out the surface-adhered liquid droplets by gently pressing them in between filter paper wraps; samples were then placed back into the oven. In order to minimize solvent evaporation losses, this step was completed within 15–20 s. All the samples reached equilibrium swelling within 15 min, which remained constant over an extended period up to 48 h. The increments in weight were equal to the total weight of acetic acid and water sorbed by the membrane. The %DS of membranes was calculated using Eq. (2):   W∞ − W0 %DS = × 100 (2) W0

The % error values in computing different mixture compositions were less than 3%, since all the weight measurements were done within an accuracy of ±0.01 mg. Mass of permeate in the trap was collected and its composition was determined by measuring refractive index using Abbe refractometer (Atago, Model 3T, Tokyo, Japan). Alternatively, the condensed permeate as well as feeds were warmed up to ambient temperature, weighed and analyzed by Nucon Gas chromatograph (GC model, 5765, Mumbai, India) containing thermal conductivity detector (TCD) equipped with a DEGS or a Tenax packed column of 1/8 in. ID having 2 m length. The oven temperature was maintained at 70 ◦ C (isothermal), while the injector and detector temperatures were maintained at 150 ◦ C. Pure hydrogen was used as a carrier gas at a pressure of 0.75 kg/cm2 . The sample injection volume was 1 ␮L and GC response was calibrated for column and for known compositions of water + acetic acid mixtures. Calibration factors were fed into GC software to obtain the analysis for unknown sample. Selectivity (α) of the membrane was determined as: α=

Yacetic acid /Ywater Xacetic acid /Xwater


where X and Y are weight fractions in the feed and permeate, respectively. Permeation rate, also known as pervaporation flux, J (kg/m2 h) was determined as: w (4) J= At where w is the weight of permeate collected (in g), t the duration of the experiment (h), and A is the effective area of the membrane (in m2 ). At least three independent readings were taken under similar set of conditions of temperature and feed compositions to confirm the steady-state permeation. Standard errors in these results were 3%, but average values are considered for data display.

where W∞ is the weight of the swollen membrane and W0 is the initial weight of the membrane.

3. Results and discussion

2.6. Pervaporation experiments

3.1. Mechanical stability of the NaAlg and STA-NaAlg membrane

Pervaporation experiments were performed in an apparatus indigenously designed; detailed protocols have been described elsewhere [26]. Effective area of the membrane in the PV cell was 28.27 cm2 with a required solution volume capacity of 200 cm3 . PV apparatus consists of a stirred stainless steel cell in which the feed stock solution was maintained at the required temperature by a thermostatically controlled water jacket. PV cell has an efficient three-blade stirrer powered by a dc motor in the feed compartment and feed mixture was stirred at 200-rpm rotation speed. Downstream side of the pressure was maintained below 13.332 × 102 Pa (10 Torr) using a vacuum pump (Toshniwal, Mumbai, India). Before starting the PV experiment, test membrane was equilibrated for 1 h with the feed mixture and after establishment of steady state, the liquid permeate was collected in traps immersed in liquid nitrogen and condensed in traps using liquid nitrogen. Experiments were repeated in triplicate with the feed mixtures ranging from 10 to 25 wt.% water.

The tensile strength at break of the 1 and 5 wt.% STAincorporated NaAlg hybrid composite membrane as well as pristine NaAlg in dry state are given in Table 1. From the results, it can be observed that the composite membrane containing 1 wt.% STA exhibits a higher tensile strength of 60.32 N/mm2 than the pristine NaAlg membrane (20.01 N/mm2 ). This enhancement may be attributed to the interaction of STA with the polymer Table 1 Tensile strength and % elongation of pristine NaAlg and 1 and 5 wt.% STAincorporated NaAlg hybrid composite membranes Membrane type

Tensile strength (N/mm2 )

Elongation at break (%)

NaAlg 5 wt.% STA + NaAlg 1 wt.% STA + NaAlg

20.01 53.10 60.32

7.6 6.5 5.4

S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186


Fig. 1. SEM (a) surface of plain NaAlg and (b) cross-sectional scanning electron micrograph of STA (5 wt.%) incorporated NaAlg membrane.

matrix making the membrane stronger than the pristine NaAlg. Polymer chains having interactions with STA particles could experience a restriction in mobility, resulting in an increase of rigidity or tensile strength of the membrane, thereby reducing the elongation at break. 3.2. Scanning electron microscopy Fig. 1(a) shows the SEM surface photograph of pristine NaAlg and (b) the cross-sectional morphologies of 5 wt.% STAincorporated NaAlg membrane. Notice that molecular level distribution of STA particles is observed in the NaAlg matrix. This type of morphology would facilitate higher amount of water molecules to transport through the membrane by restricting the transport of acetic acid. 3.3. Effect of feed concentration and STA content on membrane swelling Pervaporation results are influenced by the extent of membrane swelling. Results of % degree of swelling of pristine crosslinked NaAlg and STA-NaAlg hybrid composite membranes at 30, 40 and 50 ◦ C in 10, 15, 20, 25 and 30 wt.% watercontaining feed mixtures are shown in Fig. 2. The efficiency of

Fig. 2. % Degree of swelling vs. wt.% of water in the feed mixture at 30, 40 and 50 ◦ C for () pristine NaAlg, () STA-NaAlg-1, (♦) STA-NaAlg-2, () STA-NaAlg-3 and () STA-NaAlg-5 membranes.

a membrane depends upon its selectivity to the preferred liquid component viz., water from the mixture, which depends upon the extent of membrane swelling and temperature. STA-NaAlg hybrid membranes have shown higher % degree of swelling than the pristine NaAlg membrane at all the temperatures due to their high hydrophilic interactions with water molecules. The fact that % degree of swelling increased with increasing STA content of the membrane is indicative of the increased hydrophilicity of the STA-NaAlg hybrid membranes in the presence of water. For instance, with the membrane containing a higher amount of STA (i.e., 5 wt.%), the degrees of swelling and permeation flux are quite high. Notice that DS increases with increasing temperature, which follows the conventional wisdom that at increased


S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186 Table 2 Flux and selectivity of pristine NaAlg and STA-NaAlg hybrid membranes at 30 ◦ C wt.% of water in feed Pristine NaAlg 10 15 20 25

Fig. 3. % Degree of swelling of pristine NaAlg, STA-NaAlg-1, STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5 membranes in pure acetic acid at 30 ◦ C.

temperature; the polymer chain segments move quite rapidly allowing the higher amount of liquids to penetrate into the free volume pores of the membranes. Swelling experiments were also performed in pure water and pure acetic acid to study degree swelling of the pristine and hybrid composite NaAlg membranes. Pristine NaAlg membrane showed a DS of 162% in water, whereas STA-NaAlg membranes disintegrated after 5 h due to excessive swelling, since they could absorb higher amount of water. As shown in Fig. 3, the %DS is quite small for pristine NaAlg and STANaAlg hybrid membranes in pure acetic acid. Pristine NaAlg membrane has a DS value of 9.25%, whereas for STA-NaAlg1, STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5 membranes, DS values were 11.1 and 13.5, 15.7 and 18.6%, respectively for 48 h of immersion. Swelling characteristics of pristine NaAlg and hybrid composite membranes for each of the feed mixtures as a function of STA loading in NaAlg clearly indicate the water selective nature of all the STA-NaAlg hybrid composite membranes. Notice that in general, swelling due to water increased more than double by increasing the STA content of NaAlg matrix membrane.

STA-NaAlg-1 10 15 20 25

wt.% of water in permeate

Flux, J (kg/m2 h)

Selectivity, α

71.13 70.65 68.65 67.91

0.081 0.115 0.168 0.234

22 14 9 6

0.152 0.246 0.345 0.457

∞ ∞ ∞ ∞

100 100 100 100

STA-NaAlg-2 10 15 20 25

99.96 99.51 99.32 98.97

0.190 0.318 0.420 0.511

22,491 1151 584 288

STA-NaAlg-3 10 15 20 25

99.93 99.43 98.97 98.46

0.232 0.380 0.483 0.590

12,848 1024 384 192

STA-NaAlg-5 10 15 20 25

99.87 99.30 98.80 97.81

0.263 0.423 0.535 0.679

6914 803 329 108

3.4. Effect of silicotungstic acid content on pervaporation

that flux values of all the STA-NaAlg hybrid composite membrane increased systematically in the range 0.152–0.263 kg/m2 h for 10 wt.% water containing feed; these values are comparatively higher than that observed for pristine NaAlg membrane. Notice that flux and selectivity results are in accordance with the swelling data. Heteropolyacids have the hydrated structures in which the number of water molecules are as high as 30 per one formula unit. Silicotungstic acid, being hydrophilic, will absorb

Results of the effect of STA content on flux and selectivity of the hybrid composite membranes obtained for different feed mixtures are given in Table 2. Pristine NaAlg membrane has a flux of 0.081 kg/m2 h and a selectivity of 22 removing 71.13 wt.% of water on the permeate side for 10 wt.% water containing feed mixture. The highest selectivity value of infinity was observed at 30 ◦ C for all the feed compositions with NaAlg composite membrane containing 1 wt.% STA. Using GC, we have analyzed the amount of water removed in the permeate. The chromatogram is shown in Fig. 4 (water peak is seen at the retention time of 2.08 min with the injection volume of 1 ␮L). Selectivity data declined with increasing amount of STA in NaAlg matrix as well as with an increasing concentration of water in the feed mixture. For instance, STA-NaAlg-5 membrane has the lowest selectivity of 6194 at 10 wt.% water containing feed, while that containing 2 wt.% STA exhibited a selectivity of 22,491. An intermediate value of selectivity of 12,848 was observed for STA-NaAlg-3. It is interesting to note

Fig. 4. Plot of peak intensity vs. retention time (typical gas chromatogram of water (25 wt.%) + acetic acid mixture). Plot taken from the ASC-II data of WinAcds GC software installed in our lab. Injection volume of the mixture: 1 ␮L.

S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186


brane that is hydrophilic, but inhibits the transport of acetic acid. Swelling of STA-NaAlg membranes in pure acetic acid is smaller than pristine NaAlg membrane, which supports the observed flux data. The flux increases due to increase in driving force of transport of water as well as a faster desorption of water molecules on the permeate side. This effect is more beneficial for water transport, which further justifies the observed marked increase in selectivity with a recovery of almost 100% of water on the permeate side with a concomitant increase in flux as observed for STA-NaAlg-1 at all feed compositions. In case of other hybrid membranes, containing higher amounts of STA, % recovery of water on the permeate side is >99% at all feed compositions. 3.5. Effect of feed composition on pervaporation Fig. 5. Flux and selectivity vs. wt.% of water in the feed mixture at 30 ◦ C. () Pristine NaAlg, () STA-NaAlg-1, (♦) STA-NaAlg-2, () STA-NaAlg-3 and () STA-NaAlg-5 membranes.

more of water molecules selectively when it is incorporated in the hydrophilic NaAlg membrane. An increase in flux and a decrease in selectivity observed for STA-NaAlg-2, STA-NaAlg3 and STA-NaAlg-5 composite membranes compared to STANaAlg-1 membrane is due to higher levels of absorption of water molecules onto STA-loaded NaAlg even at lower content of STA. Thus, the hydrophilic nature of NaAlg membrane is enhanced further with the addition of 5 wt.% STA, resulting in higher membrane swelling with a consequent increase in flux by sacrificing in selectivity. Fluxes of all the membranes versus wt.% of water in feed at 30 ◦ C are displayed in Fig. 5. The decrease in selectivity at higher amount of STA is due to excess water uptake capacity of the membrane, thereby causing an excessive membrane swelling, resulting in increased free volume of the composite membrane, which has resulted in the selective transport of large number of feed water molecules on the permeate side with a decrease in selectivity, due to restriction of acetic acid molecules. Since both NaAlg polymer and STA filler particles are hydrophilic, the overall membrane performance can be explained on the basis of solution-diffusion principles [27,28] as well as adsorption–diffusion–desorption concepts [29]. It is visualized that for the pristine NaAlg membrane, permeating water molecules first get adsorbed in the micro-voids of the membrane and then diffuse out on the permeate side as a result of the concentration gradient between the barrier surface and the liquid environment. In case of hybrid composite membranes, the selective separation of water is due to the combined effect of hydrophilic–hydrophilic interactions between STA and NaAlg along with water molecules, which would induce the plasticization effect. Most water molecules get adsorbed by the hydrophilic STA, but lesser amount will be absorbed by the hydrophilic regions of NaAlg, thus promoting for an easy diffusion of water through the composite membranes. Flux of all the STA-NaAlg membranes also increased with increasing concentration of water in the feed (see Table 2). At any rate, it may be noted that separation occurs mainly due to the selective adsorption of water molecules onto STA particles, facilitating an easy transport of water molecules through the NaAlg mem-

The effect of feed composition on pervaporation performance of both pristine and STA-loaded NaAlg hybrid composite membranes was investigated and results are shown in Table 2. STAloaded NaAlg membranes swell to a larger extent in the presence of water as compared to pristine NaAlg membrane. The flux data of STA-loaded NaAlg membranes are higher than those of the pristine NaAlg membrane. For STA-loaded NaAlg hybrid composite membranes, flux increased with increasing water composition from 10 to 25 wt.% of the feed mixture. At 10 wt.% water in the feed mixture, a strong adsorptive effect of water molecules onto hybrid composite membranes was obtained, which would facilitate to increase the transport of water from the feed side [30]. For STA-NaAlg-1 membrane, the values of infinite selectivity are observed for all compositions of water in the feed. For STA-NaAlg-2 membrane, the observed selectivity is 22,491 at 10 wt.% of water in the feed, which decreased with increasing concentration of water in the feed (see Fig. 5). At higher concentration of water in the feed mixture, STANaAlg-5 membrane could absorb more amounts of water molecules when compared to STA-NaAlg-1 membrane due to plasticization effect of the polymer. However, selectivity has decreased, but flux increased considerably at 5 wt.% STA containing NaAlg membrane. For 25 wt.% water containing feed mixture, selectivity decreased to 108, but flux was enhanced to 0.679 kg/m2 h for the STA-NaAlg-5 membrane. In case of pristine NaAlg membrane, the respective values are much lower, i.e., 0.081 kg/m2 h and 0.269 kg/m2 h. In all cases, flux and selectivity of STA-NaAlg membranes are higher than that of plain NaAlg membrane. In any ease, the present study demonstrates the positive role played by silicotungstic acid upon incorporation into NaAlg to enhance the membrane performance over that of pristine crosslinked NaAlg membrane. 3.6. Effect of temperature One of the most important methods to increase the permeation rates through the membrane is to increase the temperature of permeating mixtures. In general, flux increases with increasing temperature due to increased segmental mobility of the polymeric chain along with diffusion rates of the feed liquids. Permeation, sorption and diffusion are the activated processes


S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186

Table 3 Flux and selectivity of pristine NaAlg and STA-NaAlg hybrid membranes at higher temperatures for 10 wt.% water in feed mixture Temperature (◦ C)

wt.% of water in permeate

Flux, J (kg/m2 h)

Selectivity, α

Pristine NaAlg 40 50 60 70

60.31 57.32 54.29 50.34

0.154 0.187 0.222 0.269

15 13 10 9

STA-NaAlg-1 40 50 60 70

100 100 99.80 98.91

0.165 0.189 0.258 0.349

∞ ∞ 4491 817

STA-NaAlg-2 40 50 60 70

99.89 99.83 99.77 98.89

0.178 0.195 0.263 0.423

8172 5285 3904 802

STA-NaAlg-3 40 50 60 70

99.67 99.05 97.57 95.71

0.235 0.269 0.325 0.486

2718 938 361 201

STA-NaAlg-5 40 50 60 70

99.58 98.68 97.11 93.23

0.265 0.288 0.393 0.660

2134 673 302 124

and temperature effect on these parameters can be described by the Arrhenius relationship. In the present research, membrane performance was studied at 40, 50, 60 and 70 ◦ C typically with 10 wt.% water containing acetic acid feed mixture by keeping the pressure constant. In all the membrane-based PV processes, usually an increase in temperature will result in the linear increase of permeation flux, with a decrease in selectivity. The results of flux (J) and selectivity (α) for 10 wt.% water containing feed mixture at 40, 50, 60 and 70 ◦ C are presented in Table 3, while the plots are displayed in Fig. 6. It is observed that flux increased with increasing temperature for all the membranes, but selectivity was infinity for STA-NaAlg-1 membrane up to 50 ◦ C, but later, it was decreased to 4491 and 817 at 60 and 70 ◦ C. STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5 membranes have the moderate selectivity and flux values at higher temperature. This type of unusual behavior by STA-incorporated NaAlg membranes is due to the presence of STA, which can form the protonated clusters with water molecules at higher temperatures. Membrane containing 1 wt.% STA offered the infinite value of selectivity to water along with higher values of flux, but selectivity values dropped from infinity to 22,491, 12,848 and 6914, respectively for STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5 membranes at 10 wt.% water containing feed at 30 ◦ C. These values dropped further with increasing concentrations of water in the feed; this is due to the fact that as the STA content of the membrane increased, the membrane, resulting in the excess swelling of the membrane, absorbs more of water molecules.

Fig. 6. Flux vs. wt.% of water in the feed at different temperatures: () pristine NaAlg, () STA-NaAlg-1, (♦) STA-NaAlg-2, () STA-NaAlg-3 and () STANaAlg-5 membranes.

The temperature dependency of flux was analyzed by Arrhenius equation of the type:   Ep Jp = Jpo exp − (5) RT where Jp is the permeation flux of water, Jpo the permeation rate constant, Ep the activation energy of permeation, R the molar gas constant and T is the temperature in Kelvin. If activation energy is positive, then permeation flux increases with increase in temperature, which is generally observed in majority of PV experiments [31,32]. The driving force is concentration gradient, resulting from a difference in the partial vapor pressure of the permeants between feed and permeate sides. As the feed temperature increases, vapor pressure of the feed compartment also increases, but vapor pressure on the permeate side will not be affected, resulting in an increase of driving force with increasing temperature. Arrhenius plots of ln Jp versus 1/T displayed in Fig. 7 for all the membranes for water–acetic acid mixtures display the linear trends. The Ep values computed for water in water–acetic acid mixture for pristine NaAlg, STA-NaAlg1, STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5 membranes are, respectively 21, 25, 29, 30 and 32 kJ/mol. This trend suggests that membranes with lesser selectivity have lower Ep values and the same trend was observed for all the feed mixtures of this study. As expected, the flux has increased with increasing temperature, but selectivity decreased. Membranes were stable up to operating temperature of 70 ◦ C and the same membrane withstands multiple cycles of PV experiments without any loss in selectivity. 3.7. Comparison of present membranes with literature data A comparison of the present PV results with the published results is important to assess the superiority of the membranes over the existing data. The present data are compared in Table 4 with the available literature data. To compare the data, we have

S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186


Table 4 Comparison of PV performance of STA-NaAlg hybrid membranes with literature data at different temperatures for the separation of water–acetic acid mixtures Selectivity, α


0.152 0.180 0.214 0.258 0.350

∞ ∞ ∞ 4.491 817

Present work

30 40 50 60 70

0.194 0.231 0.269 0.325 0.489

22,491 2134 673 302 124

Present work






15 90

70 50

0.262 0.094

161 18





NaAlg + 5% PVA + 10% PEG NaAlg + 10% PVA + 10% PEG NaAlg + 15% PVA + 10% PEG



0.0239 0.425 0.739

40 21 11



4.16 50

90 90

950 5400

450 125


Polyimide (BDTA/ODA) Polyaniline (salt) Polyaniline (base) Poly(vinyl chloride) Poly(sodiumacrylate)/poly(divinyl benzene) TPX-g-PGMAS with 13.7% degree of grafting Polyacrylic acid/nylon 6 blend Crosslinked/poly(vinyl alcohol) Poly(4-vinylpyridine-co-acrylonitrile)

20 47.62 47.62 50 50 16 50 50 40

65 25 25 40 30 35 25 30 21

0.048 401 13 6.0 11 0.073 170 1400–1800 45

417 1370 2.4 45 1320 125 32 20 100

[33] [7]

Poly(4-methyl-1-pentane)/ethylene-vinyl acetate copolymer TPX/P4-VP











Feed (wt.% of water)

Temperature (◦ C)



30 40 50 60 70



Composite membrane of NaAlg and PAN crosslinked with HDM Composite membrane of NaAlg and PAN crosslinked with PVA

Flux, J (kg/m2 h)

[34] [35] [36] [37] [38] [39]

NaAlg: sodium alginate; HPA: heteropolyacid; HMD: 1,6-hexanediamine; PVA: poly(vinylalcohol); PEG: polyethyleneglycol; BDTA: 3,3 ,4,4-benzophenonetetra carboxylic dianhydride; ODA: 4,4 -oxydianiline; PAN: poly(acrylonitrile); TPX: poly(4-methyl-1-pentene); PGMAS: poly(glycidyl methacrylate sulfonic acid); PAAm: poly(acrylamide).

converted the flux data of all the membranes for 50-␮m thickness of the membranes. There is a greater improvement in water selectivity at all compositions studied with STA-NaAlg membranes as compared to literature data. However, flux data are slightly moderate. 4. Conclusions

Fig. 7. Variation of ln Jp vs. 1/T for plain NaAlg, STA-NaAlg-1, STA-NaAlg-2, STA-NaAlg-3 and STA-NaAlg-5. Symbols () pristine NaAlg, () STA-NaAlg1, (♦) STA-NaAlg-2, () STA-NaAlg-3 and () STA-NaAlg-5 membranes.

The present study demonstrates that by incorporating different amounts of STA into NaAlg matrix and crosslinking with glutaraldehyde, it is possible to develop highly water selective STA-NaAlg hybrid membranes. Of all the membranes developed, hybrid composite membrane prepared with 1 wt.% of STA, i.e., STA-NaAlg-1 has exhibited an infinite selectivity value over the entire feed composition range from 10 to 25 wt.% water of the feed mixture at 30 ◦ C. Infinite selectivity to water was obtained at 40 and 50 ◦ C for only 10 wt.% feed containing water. The hybrid composite membranes containing increasing amount of


S.B. Teli et al. / Separation and Purification Technology 54 (2007) 178–186

STA increased the flux rates with the reasonable selectivity data to water for 10 wt.% water of the feed mixture. However, selectivity values decreased when higher feed water content is present in the mixture as well as at higher temperatures. The STA particles (hydrophilic) are quite compatible with the NaAlg matrix and these can be dispersed uniformly into hydrophilic NaAlg matrix. Such types of hydrophilic–hydrophilic interactions are responsible to offer increased membrane selectivity to water. It is observed that both flux and selectivity have improved tremendously with the hybrid composite membranes. These data are quite higher than pristine NaAlg membrane at all compositions of water in the feed mixture over the temperature range of the PV runs.

[10] [11] [12] [13] [14] [15]


[23] [24] [25]

Professor T.M. Aminabhavi and Dr. M. Sairam thanks the University Grants Commission (UGC), New Delhi, India for a major funding (F1-41/2001/CPP-II) to establish Center of Excellence in Polymer Science (CEPS) at Karnatak University, Dharwad. References [1] C.J. King, Acetic acid extraction, in: T.C. Lo, M.H.I. Baird, C. Hanson (Eds.), Handbook of Solvent Extraction, John Wiley & Sons, New York, 1983, pp. 567–573. [2] R.W. Helsel, Chem. Eng. Prog. 73 (1977) 55. [3] T.M. Aminabhavi, U.S. Toti, Design. Monom. Polym. 6 (2003) 211. [4] A. Aguilo, C.C. Hobbs, E.G. Zey, Acetic Acid, Ullmann’s Encyclopedia of Industrial Chemistry, vol. A1, VCH Verlagsgesellschaft GmbH, Weinheim, Germany, 1985, p. 45. [5] S. Kitao, M. Asaeda, J. Chem. Eng. Jpn. 23 (1990) 367. [6] R.Y.M. Huang, C.K. Yeom, J. Membr. Sci. 58 (1991) 33. [7] S.C. Huang, I.J. Ball, R.B. Kaner, Macromolecules 31 (1998) 5456. [8] S. Deng, S. Sourirajan, T. Matsuura, Sep. Sci. Technol. 29 (1994) 1209. [9] Y.M. Sun, T.L. Huang, J. Membr. Sci. 10 (1996) 211.

[16] [17] [18] [19] [20] [21] [22]

[26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40]

M. Yoshikawa, Y. Adachi, K. Sanui, N. Ogata, Polymer 17 (1995) 1281. R.Y.M. Huang, R. Pal, G.Y. Moon, J. Membr. Sci. 160 (1999) 101. X.P. Wang, J. Membr. Sci. 170 (2000) 71. U.S. Toti, T.M. Aminabhavi, J. Membr. Sci. 228 (2004) 199. Q. Liu, P. Jia, H. Chen, J. Membr. Sci. 159 (1999) 233. A. Bielanski, R. Dziembaj, A.M. Lubanska, J. Poziniczek, M. Hasik, M. Dorzdek, J. Catal. 185 (1999) 363. L.Y. Qu, R.Q. Lu, J. Peng, Y.G. Chen, Z.M. Dai, Synth. Met. 84 (1997) 135. K. Nakane, T. Yamashita, K. Iwakura, F. Suzuki, J. Appl. Polym. Sci. 74 (1999) 133. G.B. McGravey, J.B. Moffat, J. Catal. 128 (1991) 69. T. Baba, H. Watanabe, Y. Ono, J. Phys. Chem. 87 (1983) 2406. I.V. Kozhevnikov, Catal. Rev. Sci. Eng. 37 (1995) 311. T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41 (1996) 113. J. Gong, X.D. Li, C.L. Shao, B. Ding, D.R. Lee, H.Y. Kim, Mater. Chem. Phys. 79 (2003) 87. S.D. Bhat, T.M. Aminabhavi, Sep. Purif. Technol. 51 (2006) 85. S.D. Bhat, T.M. Aminabhavi, Micro. Meso. Mater. 91 (2006) 206. S.D. Bhat, B.V.K. Naidu, G.V. Shanbhag, S.B. Halligudi, M. Sairam, T.M. Aminabhavi, Sep. Purif. Technol. 49 (2006) 56. T.M. Aminabhavi, H.G. Naik, J. Appl. Polym. Sci. 83 (2002) 244. R.C. Binning, R.J. Lee, J.F. Jennings, E.C. Martin, Ind. Eng. Chem. 53 (1961) 45. J.G. Wijmans, R.W. Baker, J. Membr. Sci. 107 (1995) 1. X. Feng, R.Y.M. Huang, Liquid separation by membrane pervaporation: a review, Ind. Eng. Chem. Res. 36 (1997) 1048. M.D. Kurkuri, U.S. Toti, T.M. Aminabhavi, J. Appl. Polym. Sci. 86 (2002) 3642. M.C. Burshe, S.A. Netke, S.B. Sawant, J.B. Joshi, V.G. Pangarkar, Sep. Sci. Technol. 32 (1997) 1335. K. Satoshi, A. Masashi, J. Chem. Eng. Jpn. 23 (1990) 367. Y.D. Moon, B.K. Oh, Y.M. Lee, Polym. Bull. (Berlin) 29 (1992) 431. H. Okuno, H. Nishmito, T. Miyata, T. Urgami, Macromol. Chem. 194 (1993) 927. E. Ruckenstein, H.H. Chen, J. Membr. Sci. 66 (1992) 205. J.F. Lee, Y.C. Wang, Sep. Sci. Technol. 33 (1998) 187. R.Y.M. Huang, A. Moreira, R. Notarofonzo, Y.F. Xu, J. Appl. Polym. Sci. 35 (1998) 1191. R.Y.M. Huang, C.K. Yemon, J. Membr. Sci. 58 (1991) 33. Y.M. Lee, B.K. Oh, J. Membr. Sci. 85 (1993) 13. K.R. Lee, J.Y. Lai, J. Polym. Res. 1 (1994) 247.