Pervaporation of dichlorinated organic compounds through silicalite-1 zeolite membrane

Pervaporation of dichlorinated organic compounds through silicalite-1 zeolite membrane

Journal of Membrane Science 279 (2006) 459–465 Pervaporation of dichlorinated organic compounds through silicalite-1 zeolite membrane Hyoseong Ahn, Y...

436KB Sizes 0 Downloads 50 Views

Journal of Membrane Science 279 (2006) 459–465

Pervaporation of dichlorinated organic compounds through silicalite-1 zeolite membrane Hyoseong Ahn, Yongtaek Lee ∗ Department of Chemical Engineering, Chungnam National University, 220 Gungdong, Daejeon, Republic of Korea Received 25 March 2005; received in revised form 17 December 2005; accepted 25 December 2005 Available online 2 February 2006

Abstract Pervaporation utilized with a zeolite membrane is one of the economic separation technologies for liquid mixtures including organic/water mixtures. The silicalite-1 membrane was selected for pervaporation of dichloromethane, 1,2-dichloroethane and trans-1,2-dichloroethylene from their aqueous solutions since it shows very hydrophobic properties. Silicalite-1 crystals were hydrothermally grown and deposited on the inside of a porous sintered stainless steel tube by the secondary growth method. Fluxes of dichlorinated organic compounds were observed to be 0.2–71 g/(m2 h) while separation factors were 12–332 depending on a mole fraction of a dichlorinated organic compound in the feed solution and the operation temperature. The silicalite-1 membrane was found to be applicable to the selective removal of dichlorinated organic compounds from aqueous solutions composed of a low concentration at a relatively low temperature. © 2006 Elsevier B.V. All rights reserved. Keywords: Organic/water separation; Dichlorinated compounds; Pervaporation; Silicalite-1; Zeolite membrane

1. Introduction Due to their unique pore structure, mechanical, chemical and biological stabilities, zeolite membranes have been widely studied on various applications: gas permeation, vapor permeation, pervaporation, etc. [1–3]. Pervaporation is not only an economic separation technology since it needs electric powers to maintain the permeate side in vacuum but also an environmentally clean technology in which potential pollution sources such as entrainers for azeotropic distillation are not needed [4,5]. A hydrophobic membrane could be used to separate organic compounds from their aqueous solution. Separation of organic compounds from aqueous solutions is very important in a point of view from either prevention of water pollution or recycle of valuable materials. Even if hydrophobic polymer membranes might show a high selectivity, their operation conditions such as a concentration and a temperature may limit their application. Also thermal, chemical and mechanical stabilities of polymer membrane are not good enough for a certain usage. On the other hand, a zeolite membrane might be used for organic separation

Corresponding author. Tel.: +82 428 215 686; fax: +82 428 228 995. E-mail address: [email protected] (Y. Lee).

0376-7388/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2005.12.060

with pervaporation technique since it shows not only molecular sieve effects but also good physicochemical stabilities [6]. Chlorinated organic compounds are useful materials which have been used for various applications such as solvents, coating agents, extractants, cleaning agents, dry cleaning and so on [7]. Chlorinated organic compounds are strongly regulated on discharges so that air and water pollutions could be prevented. Many research results were published about the adsorption relation between zeolites and chlorinated organic compounds. For example, adsorption of chlorinated organic compounds including methyl chloride on NaY zeolite [8], 1,2dichloroethane on ZSM-5 zeolite [9], dichlorinated and trichlorinated organic compounds on synthetic sorbents [10]. Also, the decomposition or the catalytic oxidation of chlorinated organic compounds were reported using ion-exchanged zeolites such as Co Y, Cr Y, Mn Y [11], H Y, Na Y [12] and H Y, H-ZSM-5 [13]. Silicalite-1 zeolite classified as a MFI structure has been developed for organic material separation from aqueous solutions since it is known to be hydrophobic [14]. Olson et al. [15] reported about hydrophobicity of H-ZSM-5 zeolites with various aluminum oxide contents in the framework. Their research showed that water adsorption amount was strongly related with contents of Al2 O3 and increased with increase of Al2 O3


H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465

contents. Compared with their research, silicalite-1 could be regarded as more hydrophobic than H-ZSM-5 since silicalite1 is composed only silicon and oxygen without Al2 O3 in their framework. According to the hydrophobicity of silicalite-1 and advantages from pervaporation, organic compounds might be separated economically from their aqueous solution by pervaporation using a silicalite-1 membrane. Silicalite-1 zeolite membrane has been investigated for several purposes such as gas permeation including single gas and gas mixture to determine membrane characteristics and separation of gas mixtures [16–23], vapor permeation [24,25] and pervaporation for organic/water mixtures. Sano et al. [26] synthesized a silicalite-1 zeolite membrane on a sintered stainless steel tube as a porous support. Their membrane was used for pervaporation of acetic acid from its water mixture. It was shown that the constant total flux was obtained at the concentration of above 20 vol.% of acetic acid in the feed solution and the permeate concentration of acetic acid was fairly higher than the feed. Nomura et al. [27] reported about pervaporation characteristics of ethanol/water mixture by a silicalite-1 zeolite membrane synthesized on a sintered stainless steel tube. Concentrations of ethanol in the permeate is higher than those obtained by the vapor–liquid equilibrium at each feed concentration of ethanol. It was shown that the ethanol flux was increased as increase of a feed concentration of ethanol and the highest separation factor was observed at 4.65 wt.% of ethanol in the feed. Noble and co-workers [6] separated organic compounds from their aqueous methanol solutions using a silicalite-1 zeolite membrane by pervaporation. The total flux increased as the feed concentration increased and the highest separation factor could be obtained at 16.5 wt.% of the feed concentration of methanol in its aqueous solution. Their membrane showed more outstanding separation behavior on acetone/water pervaporation than on methanol/water. They obtained the highest separation factor of 255 at the acetone feed concentration of 0.8 wt.% and the highest acetone flux of 0.95 kg/(m2 h) at the acetone concentration of 43 wt.%. Falconer and co-workers [28] reported on a relationship between the pervaporation characteristics and the fugacity of feed components for various C1 –C4 organic/water mixtures with Ge-ZSM-5 zeolite membranes. Among pervaporation characteristics, the separation factor and the organic coverage on the feed side were correlated strongly with the organic feed fugacity. However, a molecular size and a heat of adsorption did not significantly affect the separation factor while they influenced the flux and the coverage. The organic flux also depends strongly on competitive adsorption resulting from the feed fugacity difference of feed components. Fugacity was not expected to affect diffusion coefficients. In this study, a thin film of silicalite-1 zeolite was prepared by the secondary growth method where the seed crystals were spread on the inside of a porous support as crystal growing nuclei; it was used to separate dichlorinated organic compounds (DCOs) from an aqueous feed stream. The effect of a temperature and a concentration are presented on the

DCOs/water separation. Since the solubility of DCOs in water is quite low such as an order of 10−3 in a mole fraction, the feed concentrations were selected below 0.001 mole fraction of DCOs. 2. Experimental 2.1. Preparation and analysis of membrane Silicalite-1 zeolite membranes were synthesized from liquid mixtures in which the chemical compositions were 1 TPABr:21 SiO2 :3 NaOH:788 H2 O. Ludox As40 (Dupont, U.S.A.) was used as a source material for Si; sodium hydroxide (Daejung, Korea) was used as a source chemical for Na; tetrapropylammonium bromide (Aldrich, U.S.A.) was used as a template. After preparing a Si solution and a template solution, the template solution was added to the Si solution. The synthesized silicalite1 zeolite powders (mean particle size 6.15 ␮m, analyzed with HELOS particle analyzer, Sympatec) were used as seed crystals; they were rubbed on the inside surface of a support tube (sintered stainless steel tube, Mott. Co., o.d. 9.5 mm, i.d. 6.4 mm, mean pore size 0.5 ␮m) using a sponge brush. Silicalie-1 zeolite membranes were synthesized under the reaction temperature of 150–180 ◦ C and the reaction time of 8–16 h. After synthesis, inside of the support tube was brushed carefully with distilled water as a cleaning procedure. After two to three times of synthesis, washing and drying, silicalite-1 membranes were calcined at upto 480 ◦ C. The crystal structure of the zeolite membrane was confirmed with a XRD (D/Max-2200 Ultima/PC, Rigaku Co., Japan, 30 kV, 15 mA) and the morphology was analyzed with a SEM (JSM6300, Jeol Ltd., Japan). The synthesized silicalite-1 powders were prepared as follows: the powders were collected during a cleaning procedure of zeolite membranes and they were calcined at the same temperature as the silicalite-1 membranes were prepared. 2.2. Pervaporation The pervaporation experiments were carried out using a pervaporation apparatus designed and set up as schematically shown in Fig. 1. The membrane was installed in a membrane test cell. The aqueous feed solutions were fed and circulated with a diaphragm pump (DMA-05, Daekyung, Korea). The feed concentration was regulated upto the solubility of each DCOs in water as shown in Table 1. The permeate was collected in a liquid N2 trap in which the pressure was controlled to be about 6.66 × 102 Pa (5 mmHg). The experiments were performed at the temperature ranged from 25 to 35 ◦ C for dichloromethane, 25 to 45 ◦ C for dichloroethane and 25 to 45 ◦ C for dichloroethylene for pervaporation experiments to carry out liquid feed mixture separation. Both the feed and the permeate concentrations were analyzed with a Gas Chromatograph (M600D, Younglin Co., Korea) where Porapak Q (Supelco, U.S.A.) column was equipped. A flame ionization detector (FID) was used to determine the concentration of organic compounds.

H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465


Fig. 1. A schematic diagram of pervaporation apparatus. Fig. 2. XRD pattern of silicalite-1 zeolite membrane.

2.3. Pervaporation characteristics Pervaporation characteristics might be presented in terms of a flux and a separation factor defined as follows: Flux =

P , A×t

Separtion factor =

yDCOs /yH2 O xDCOs /xH2 O

where P represents a weight of permeate (g), A the membrane area (m2 ) and t is a permeation time (h). yDCOs and yH2 O refer to mole fractions of DCOs and water at the permeate side, respectively. xDCOs and xH2 O represent mole fractions of DCOs and water at the feed side, respectively. 3. Results and discussion 3.1. Membrane characterization Fig. 2 shows XRD patterns of the synthesized silicalite-1 and the seed silicalite-1 as a standard which was confirmed by Ref. [32].

The synthesized silicalite-1 was confirmed as a MFI structure since the XRD pattern was same as that of the standard silicalite1 zeolite. SEM images of the synthesized silicalite-1membrane were shown in Fig. 3. As shown in Fig. 3, the synthesized silicalite1 zeolite crystals were randomly grown to the size of about 10–20 ␮m and the thickness of the zeolite layer was found to be about 15 ␮m. 3.2. Organic and water fluxes Fluxes of DCOs and water through the silicalite-1 membrane are shown in Figs. 4 and 5. Fluxes of DCOs increased as a mole fraction of DCOs increased and also they increased as a temperature increased while water fluxes decreased with increase of a mole fraction of DCOs. This may be due to the increment of the fugacity of DCOs in the feed side shown in Table 2. The increment of either the concentration or the temperature results

Table 1 Physical properties of DCOs Name




Chemical structure Molecular weight Boiling point (◦ C) Solubility (g/100 g H2 O, 20 ◦ C)

CH2 Cl2 84.92 39.8 1.32

C2 H4 Cl2 98.95 83.7 0.869

C2 H2 Cl2 96.94 48 0.63 (25 ◦ C)

Vapor pressure (kPa)a 25 (◦ C) 30 (◦ C) 35 (◦ C) 45 (◦ C)

57.72 70.17 84.67 120.69

5.23 13.90 17.41 26.64

45.28 55.13 66.61 95.19


Vapor pressure was calculated from the KDB correlation equation provided by Ref. [29].


H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465

Table 2 Fugacity of the feed mixture Dichloromethane 0.0001a



Fugacity of organics (kPa)b 25 (◦ C) 1.56 5.45 30 (◦ C) 1.93 6.76 35 (◦ C) 2.33 8.16 – – 45 (◦ C)










10.90 13.52 16.31 –

15.58 19.31 23.30 –

0.17 – 0.56 0.83

0.34 – 1.11 1.67

1.20 – 3.90 5.83

1.71 – 5.57 8.33

0.60 – 0.94 1.44

2.10 – 3.30 5.03

3.60 – 5.66 8.63

5.99 – 9.44 14.38







Dichloromethane 0.9999c Fugacity of water (kPa)b 25 (◦ C) 30 (◦ C) 35 (◦ C) 45 (◦ C)






3.16 4.24 5.62 –



3.16 – 5.62 9.59


0.999965c 3.16 – 5.62 9.59


Mole fraction of organic compounds. Activity coefficient was obtained from Refs. [30,31]. Fugacity was calculated by the equation, fi = xi γi Pisat , xi : mole fraction, γ i : activity coefficient, Pisat : saturated vapor pressure (kPa). c Mole fraction of water. b

in the increment of the fugacity of DCOs since the fugacity is a function of the mole fraction and the saturated vapor pressure given at a temperature. The coverages of DCOs on silicalite-1 zeolite increased with fugacity which might result in a greater transportation of DCOs. However, the change in fluxes is not proportional to the change in fugacity. It is expected that the fluxes of DCOs depend strongly on the competitive adsorption between DCOs and water because silicalite-1 zeolite is hydrophobic. The water coverages could be decreased according to the increase of DCOs coverage. A decrease of water coverage would result in a significant decrease of the driving force for water transportation and accordingly the water flux decreases as shown in Fig. 5. The water flux is higher than the DCOs flux at the same experimental conditions because water is the dominant component

Fig. 3. SEM images of the silicalite-1 membrane.

Fig. 4. DCOs fluxes through silicalite-1 membrane.

H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465

Fig. 5. Water fluxes through silicalite-1 membrane.

in a feed mixture and it can be adsorbed and transported. Olson et al. [15] reported that water was adsorbed upto 4% of the adsorption capacity of a H-ZSM-5 which is regarded very hydrophobic (SiO2 /Al2 O3 = 8660). According to their research, it is expected that water could be adsorbed and transported even though silicalite-1 is highly hydrophobic. Moreover, water adsorption occurs more easily than DCOs adsorption under our experimental situation where water is the dominant component of mole fraction over 0.999 even though DCOs have a higher fugacity than water. Under these conditions, water is preferably transported over DCOs as shown in Fig. 5. However, the water fluxes dramatically decreased with increase of a mole fraction of DCOs in a feed solution. This suggests that the fugacity does not seriously affect the coverages of DCOs at a very low concentration of DCOs. Kita and co-workers [33] reported that their tubular NaA zeolite module showed the water flux of 1100 g/(m2 h) at the feed concentration of 5 wt.% (0.12 mole fraction) and 75 ◦ C. The flux of DCOs in our study was ranged between 14 and 71 g/(m2 h) which was obtained below 0.001 mole fraction of DCOs in a feed solution. The flux of DCOs seems reasonable because the experiments were carried out under very low feed mole fractions and low temperatures compared with Kita’s experimental conditions.

Fig. 6. Separation factors of DCOs through silicalit-1 membrane.

fluxes in Fig. 4 is greater than the increasing rate of the denominator (xDCOs /xH2 O ) from the increment of the feed mole fraction. 3.4. Comparison of separation behavior of DCOs Organic fluxes at the same mole fraction of 0.0001 in the feed solution are shown in Fig. 7 to compare the pervaporation behavior among DCOs as a function of the temperature. While dichloromethane showed the highest fluxes, 1,2-dichloroethane showed the lowest fluxes. This may be due to the fugacity difference of each compound according to the difference of their activity coefficient. The highest flux of dichloromethane can be said that dichloromethane has the highest coverage because it has the highest fugacity among DCOs and it can diffuse eas˚ [34] is small enough to ily since its kinetic diameter of 4.898 A pass the pore of silicalite-1 zeolite which is known to be 5.1– ˚ 5.6 A. Even though trans-1,2-dichloroethylene and 1,2-dichloroethane might have a similar kinetic diameter, 1,2-dichloroethane

3.3. Separation factor In Fig. 6, the separation factors of DCO were plotted as a function of the feed mole fraction of DCOs through the silicaite-1 membrane. It was observed that both the mole fraction of DCOs and the temperature strongly affected the separation factor. It should be noted that the separation factor increases almost linearly with the feed fugacity of DCOs while the flux of DCOs increases slowly with the feed fugacity of DCOs as shown in Fig. 4. Increase of the separation factor could be explained by the definition of the separation factor equation. As shown in the definition of the separation factor, the increasing rate of the numerator (yDCOs /yH2 O ) according to the increment of DCO


Fig. 7. DCOs fluxes at the same feed mole fraction.


H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465

4. Conclusions Silicalite-1 membranes were successfully synthesized on the inside of a porous sintered stainless steel tube by the secondary growth method; they showed reasonable fluxes and separation factors for dichloromethane/water, 1,2-dichloroethane/ water and trans-1,2-dichloroethylene/water mixtures. The flux and the separation factor through silicalite-1 zeolite membranes were strongly affected by both the feed concentration and the operating temperature. The separation factors of dichloromethane/water mixture and trans-1,2-dichloroethylene/ water were found to be two times and 1.5 times higher than that obtained from 1,2-dichloethane/water mixture, respectively. Silicalite-1 membrane might be used as a hydrophobic pervaporation membrane to separate DCOs from their aqueous solutions. Fig. 8. Separation factors of DCOs at the same feed mole fraction.

Acknowledgement fluxes are lower than trans-1,2-dichloroethylene since its fugacity is two to five times lower than that of trans-1,2dichloroethylene as shown in Table 2. Separation factors of DCOs at the same mole fraction of 0.0001 in the feed mixture are plotted in Fig. 8. The separation factor of dichloromethane shows the highest value and 1,2-dichloroethane shows the lowest value. The separation factor difference among DCOs might be explained by the difference of the permeation flux according to Fig. 7. The separation factor increased as the temperature increased which leads to the increment of the fugacity. At the same operation conditions, the separation factor of dichloromethane was observed to be two times larger than that of 1,2-dichloroethane and the separation factor of trans-1,2-dichloroethylene was shown to be 1.5 times higher than that of 1,2-dichloroethane. From Table 2, it is expected that the separation factor of trans1,2-dichloroethylene is higher than dichloromethane since its fugacity is higher than dichloromethane fugacity [28]. Regarding the relationship between the fugacity and the separation factor, Falconer and co-workers [28] used 5 wt.% of feed organic concentration. At this concentration, high and enough coverages of organics on zeolite membrane could be obtained because their membrane is hydrophobic and the organic fluxes is higher than water flux even though the fugacity of organic is lower than fugacity of water. On the other hand, when the organic has a higher fugacity than water, the higher separation factor was also obtained in their case. In our study, the feed concentration is very low such as below 0.001 mole fraction. In this case, it is expected that the coverages among DCOs are not significantly different even if their fugacity is different. This suggests that the difference of the diffusivity seems to do a major role than the fugacity difference, which affects fluxes and separation factors among DCOs at a very low feed concentration. In this situation, the flux and the separation factor of dichloromethane is higher than trans-1,2dichloroethylene since dichloromethane has a smaller kinetic diameter than trans-1,2-dichloroethylene which means a higher diffusivity in the membrane.

This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOEHRD) (R052003-000-10119-0). References [1] K. Aoki, K. Kusakabe, S. Morooka, Separation of gases with an A-type zeolite membrane, Ind. Eng. Chem. Res. 39 (2000) 2245. [2] K. Weh, M. Noack, I. Sieber, J. Caro, Permeation of single gases and gas mixtures through faujasite-type molecular sieve membranes, Microporous Mesoporous Mater. 54 (2002) 27. [3] H. Kita, K. Fuchida, T. Horita, H. Asamura, K. Okamoto, Preparation of faujasite membranes and their permeation properties, Sep. Purif. Technol. 25 (2001) 261. [4] C.C. Pereira, A.C. Habert, R. Nobrega, C.P. Borges, New insights in the removal of diluted volatile organic compounds from dilute aqueous solution by pervaporation process, J. Membr. Sci. 138 (1998) 227. [5] D. Hofmann, L. Fritz, D. Paul, Molecular modelling of pervaporation separation of binary mixtures with polymeric membranes, J. Membr. Sci. 144 (1998) 145. [6] Q. Liu, R.D. Noble, J.L. Falconer, H.H. Funke, Organics/water separation by pervaporation with a zeolite membrane, J. Membr. Sci. 117 (1996) 163. [7] R.E. Kirk, D.F. Othmer, Encyclopedia of chemical technology, fourth ed., John Wiley & Sons, New York, 1991. [8] J. Konya, I. Hannus, I. Kiricsi, Infrared spectroscopic study of adsorption and reactions of methyl chloride on acidic, neutral and basic zeolites, Appl. Catal. B Environ. 8 (1996) 391. [9] B. Garrot, G. Couderc, M.-H. Simonot-Grange, F. Stoeckli, Coadsorption of 1,2-dichloroethane and 1-bromo,2-chloroethane on zeolite ZEM-5 from the liquid and vapor phases, using the Myers–Prausnitz–Dubinin model, Microporous Mesoporous Mater. 52 (2002) 199. [10] G. Rexwinkel, J.T.A.M. Berkhout, A.B.M. Heesink, Adsorption of chlorinated hydrocarbons from aqueous solutions by wetted and non-wetted synthetic sorbents: dynamics, Chem. Eng. Sci. 58 (2003) 1301. [11] S. Chatterjee, H.L. Greene, Y.J. Park, Comparison of modified transition metal-exchanged zeolite catalysts for oxidation of chlorinated hydrocarbons, J. Catal. 138 (1992) 179. [12] B. Ramachandran, H.L. Greene, S. Chatterjee, Decomposition characteristics and reaction mechanisms of methylene chloride and carbon tetrachloride using metal-loaded zeolite catalysts, Appl. Catal. B Environ. 8 (1996) 157. [13] J.R. Gonz´alez-Velasco, R. L´opez-Fonseca, A. Aranzabal, J.J. Guti´errezOrtiz, P. Steltenpohl, Evaluation of H-type zeolites in the destructive

H. Ahn, Y. Lee / Journal of Membrane Science 279 (2006) 459–465










oxidation of chlorinated volatile organic compounds, Appl. Catal. B Environ. 24 (2000) 233. J.F. Smetana, J.L. Falconer, R.D. Noble, Separation of methyl ethyl ketone from water by pervaporation using a silicalite membrane, J. Membr. Sci. 114 (1996) 127. D.H. Olson, W.O. Haag, W.S. Borghard, Use of water as a probe of zeolite properties: interaction of water with HZSM-5, Microporous Mesoporous Mater. 35–36 (2000) 435. H. Takaba, R. Koshita, K. Mizukami, Y. Oumi, N. Ito, M. Kubo, A. Fahmi, A. Miyamoto, Molecular dynamics simulation of iso- and n-butane permeations through a ZSM-5 type silicalite membrane, J. Membr. Sci. 134 (1997) 127. W.J.W. Bakker, F. Kapteijn, J. Poppe, J.A. Moulijn, Permeation characteristics of a metal-supported silicalite-1 zeolite membrane, J. Membr. Sci. 117 (1997) 57. C. Bai, M. Jia, J.L. Falconer, R.D. Noble, Preparation and separation properties of silicalite composite membranes, J. Membr. Sci. 105 (1995) 79. J. Dong, K. Wegner, Y.S. Lin, Synthesis of submicron polycrystalline MFI zeolite films on porous ceramic supports, J. Membr. Sci. 148 (1998) 233. M. Yang, B.D. Crittenden, S.P. Perera, H. Moueddeb, J.-A. Dalmon, The hindering effect of adsorbed components on the permeation of a non-adsorbing component through a microporous silicalite membrane: the potential barrier theory, J. Membr. Sci. 156 (1999) 1. M. Nomura, T. Yamaguchi, S.-I. Nakao, Transport phenomena through intercrystalline and intracrystalline pathways of silicalite zeolite membranes, J. Membr. Sci. 187 (2001) 203. M. Noack, P. K¨olsch, R. Sch¨afer, P. Toussaint, I. Sieber, J. Caro, Preparation of MFI membranes enlarged area with high reproducibility, Microporous Mesoporous Mater. 49 (2001) 25.


[23] M. Arruebo, J. Coronas, M. Mem´endez, J. Santamaria, Separation of hydrocarbons from natural gas using silicalite membranes, Sep. Purif. Technol. 25 (2001) 275. [24] M. Noack, P. K¨olsch, J. Caro, M. Schneider, P. Toussaint, I. Sieber, MFI membranes of different Si/Al ratios for pervaporation and steam permeation, Microporous Mesoporous Mater. 35–36 (2000) 253. [25] W. Zhu, F. Kapteijn, J.A. Moulijn, Separation modeling of linear and branched C6 alkane permeation through silicalite-1 membrane, Sep. Purif. Technol. 32 (2003) 223. [26] T. Sano, S. Ejiri, K. Yamada, Y. Kawakami, H. Yanagishita, Separation of acetic acid–water mixtures by pervaporation through silicalite membrane, J. Membr. Sci. 123 (1997) 225. [27] M. Nomura, T. Yamaguchi, S. Nakao, Ethanol/water transport through silicalite membranes, J. Membr. Sci. 144 (1998) 161. [28] T.C. Bowen, S. Li, R.D. Noble, J.L. Falconer, Driving force for pervaporation through zeolite membranes, J. Membr. Sci. 225 (2003) 165. [29] [30] M.L. Brendel, S.I. Sandler, The effect of salt and temperature on the infinite dilution activity coefficients of volatile organic chemicals in water, Fluid Phase Equilib. 165 (1999) 87. [31] B. Khalfaoui, D.M.T. Newsham, Determination of infinite dilution activity coefficients and second virial coefficients using gas–liquid chromatography. I. The dilute mixtures of water and unsaturated chlorinated hydrocarbons and of water and benzene, J. Chromatogr. A 673 (1994) 85. [32] [33] Y. Morigami, M. Kondo, J. Abe, H. Kita, K. Okamoto, The first largescale pervaporation plant using tubular-type module with zeolite NaA membrane, Sep. Purif. Technol. 25 (2001) 251. [34] R.C. Reid, J.M. Prausnitz, B.E. Poling, Properties of Gases and Liquids, fourth ed., McGraw-Hill, New York, 1987.