Xylene Vapor Permeation in MFI-Zeolite ... - ACS Publications

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26 ... permeation for p-x/o-x have been achieved at varied temperatu...
1 downloads 0 Views 14MB Size
Subscriber access provided by the Henry Madden Library | California State University, Fresno

Separations

Xylene Vapor Permeation in MFI-Zeolite Membranes Made by Templated and Template-free Secondary Growth of Randomly Oriented Seeds: Effects of Xylene Activity and Microstructure Fateme Banihashemi, Lie Meng, Ali A. Babaluo, and Jerry Y.S. Lin Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b01373 • Publication Date (Web): 04 Jun 2018 Downloaded from http://pubs.acs.org on June 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Xylene Vapor Permeation in MFI-Zeolite Membranes Made by Templated and Template-free Secondary Growth of Randomly Oriented Seeds: Effects of Xylene Activity and Microstructure Fateme Banihashemia,b, Lie Menga, Ali A. Babaluob, Y.S. Lina,* a. School for Engineering of Matter, Transport and Energy, Arizona State University, Tempe, AZ 85287, USA b. Nanostructure Material Research Center, Chemical Engineering Department, Sahand University of Technology, P.O. Box 51335/1996, Tabriz, I.R. Iran

ACS Paragon Plus Environment

1

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 49

ABSTRACT Many studies have reported high p-xylene/o-xylene perm-selectivities with MFI zeolite membranes, however, these results were obtained at low xylene activity ( 400 and a p-

ACS Paragon Plus Environment

3

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 49

x flux of ~ 3.5 µmol m−2 s−1 at 200oC on HZSM-5/alumina composite membranes synthesized by the pore-plugging method. All above results were obtained by vapor permeation at low xylene partial pressures (< 2 kPa) or activity (ratio of partial pressure to the saturated vapor pressure, < 0.01). Falconer and co-workers studied p-x/o-x vapor separation through ZSM-5 zeolite membranes at xylene partial pressures from 0.4 to 2.5 kPa 13, and found that with increasing xylene partial pressure, p-x/o-x separation factor decreases from 25 to 7.5. Gu et al.

14

also observed a decrease in p-x/o-x

separation factor from 17.8 to 8.8 with an increase in xylene partial pressure from 0.91 to 87.2 kPa at 250oC. While the selectivity drops, p-x flux increases from 0.49 to 5.9 × 10−5 mol m−2 s−1. Xomeritakis et al. 15 investigated vapor permeation of p-x and o-x in single- and binary-gas as a function of xylene pressure at 100oC for MFI zeolite membranes. The permeation flux of o-x in binary-gas approaches that in single-gas only at very low p-x partial pressure in feed and surpasses the flux of pure-gas permeation with increasing p-x partial pressure. These results indicate that the separation performance of MFI zeolite membranes is affected by the presence of p-x in the feed. As the xylene partial pressure decreases, the isomers exhibit less interaction, which in turn results in an enhanced p-x/o-x selectivity. Moreover, structural deformation of MFI zeolite may occur in existence of p-x in the feed, which could cause a reduction in separation factor of p-x over other xylene isomers. 16, 17 Besides xylene partial pressures, the separation performance of zeolite membrane is also influenced by the zeolitic microstructure, which could be changed in the template removal step during membrane synthesis.

6

We recently applied positron annihilation spectroscopy (PAS) to

characterize the pore structure of MFI zeolite membranes and found that template-synthesized MFI zeolite membranes contain a moderate amount of inter-crystalline micropores while

ACS Paragon Plus Environment

4

Page 5 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

template-free synthesized membranes have a nearly no defect or very small amount of intercrystalline micropores.

4

We reported that template-free synthesized MFI zeolite membranes

exhibited better pervaporation separation factors of p-x over o-x than those template-synthesized MFI zeolite membranes.

3, 18, 19

However, no vapor permeation xylene separation data were

reported for these randomly-oriented MFI zeolite membranes of the different microstructure. Random-oriented zeolite membranes have received much more attention for industrial application (and have been commercialized for solvent dehydration) because of simplicity in membrane synthesis allowing easier scaling up as compared to oriented zeolite membranes. For industrial application of randomly-oriented MFI zeolite membranes for vapor permeation xylene separation, it is important to understand the effects of the membrane microstructure and permeation operation conditions (especially the activity of xylene) on separation performance of the membrane, which are not clear. The objective of this investigation is to understand how xylene activity affects the permeation and separation properties of the randomly-oriented MFI zeolite membranes of different microstructure, prepared with/without an organic template.

2 2.1

EXPERIMENTAL Synthesis and Characterization of MFI Zeolite Membranes MFI zeolite membranes were synthesized via secondary growth method onto α-alumina

substrates coated with a yttrium stabilized zirconia (YSZ) intermediate layer. The polished surface of a homemade macroporous α-alumina supports (2 mm thick, 20 mm in diameter, average pore diameter, 200 nm; porosity, 45%) were dip-coated in a stable YSZ suspension for preparation of the intermediate layer according to a previously reported procedures 18.

ACS Paragon Plus Environment

5

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 49

To prepare silicalite seeds, fumed silica (particle size: 7 nm, surface area: 350-440 m2 g−1, Sigma–Aldrich) and sodium hydroxide (NaOH, 98%, Sigma–Aldrich) were dissolved in tetrapropylammonium hydroxide (TPAOH, 1M in water, Sigma–Aldrich) at 80oC under stirring. The prepared precursor solution has a molar composition of 10SiO2: 2.4TPAOH: 1NaOH: 110H2O. The followed hydrothermal synthesis was performed at 125oC for 8 h. The obtained silicalite seeds were washed repeatedly with distilled water and a centrifugation was used to collect the solids. The seed layer was coated by dip-coating the support with the silicalite seed suspension with a composition of 1 g silicalite seed-0.14 g HPC solution-94 ml H2O, followed by drying at 40oC and calcination at 550oC for 8 h (heating/cooling rate: 18oC h−1). The HPC solution was prepared by adding 0.5 wt.% hydroxy propyl cellulose (HPC, molecule weight = 100,000 g mol−1, Aldrich) as a binder and adjust the pH of the solution from 11 to 3 by adding a few drops of 1M HNO3. The silicalite seed layer was calcined after each dip coating; the dipcoating procedure was repeated three times to ensure adequate coverage of the seed layer. Silicalite seed layer was grown to a continuous layer by a templated secondary growth solution with a molar composition of 0.9 NaOH: 0.9 TPABr: 4 SiO2: 1000 H2O: 16 EtOH under a hydrothermal condition at 175◦C for 8 h. The synthesized membrane was washed and dried, then calcined at 500◦C for 8 h in air to remove the template. These membranes are referred to as templated membranes. Template-free secondary growth solution was prepared by adding a given amount of fumed silica powder to NaOH solution at 80◦C with vigorous stirring. The resulting solution with a molar composition of 1SiO2: 0.16NaOH: 10.5H2O was aged for 2h and then used for secondary growth of the seeded silicalite layer under hydrothermal treatment conditions at 180◦C for 4h. After cooling to the room temperature, the membranes were removed from the autoclave, washed with distilled water, and dried, according to the procedure reported

ACS Paragon Plus Environment

6

Page 7 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

previously.

20, 21

These membranes did not go through high temperature calcination as no

template was used in the secondary growth step, and are referred to as template-free membranes. Table S1 outlines the secondary growth conditions for synthesized random membranes with a different structure. The morphology of surface and cross-section of the substrate and synthesized zeolite membranes was characterized using scanning electron microscope (SEM, Amray 1910). The crystallographic properties of the α-alumina supported zeolite membranes were examined by Xray diffraction (XRD, AXS-D8, Bruker) with a scan step of 0.015o. Heat treatment condition of MFI zeolite membranes was studied by thermogravimetric analysis (TGA, SDT-Q600, TA Instruments Inc.) with a heating rate of 5 ºC min−1 from room temperature to 400oC.

2.2

Gas Permeation, Pervaporation and Vapor Permeation/Separation Experiments He and SF6 single-gas permeation data were measured on the apparatus shown in our

previous work.

22

The permeation properties of the support and MFI zeolite membranes were

studied at room temperature and a transmembrane pressure range of 60-280 kPa. The permeate flow rate was measured with a bubble flow meter. The gas permeance is defined as

 =

  ∙ ( −  )

(1)

where  is the permeance of species through the membrane in mol m−2 s−1 Pa−1;  is the gas flow rate at permeate side of the membrane in mol s−1,  and  are the feed and permeate measured pressures in Pa;  is the active membrane area in m2. In this study, the active permeation area was 2 × 10−4 m2 for all membranes. Pervaporation experiments of xylene isomers through prepared MFI zeolite membranes were performed on the setup shown in detail in our previous publication.

6

The physical

ACS Paragon Plus Environment

7

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 49

properties of p-xylene and o-xylene are summarized in Table S2. Pervaporation experiments were first conducted with pure components p-x, o-x and then equimolar binary mixture at room temperature. The membrane was sealed in the vertical stainless-steel cell, and the liquid feed was contained in a reservoir above, with the MFI zeolite layer facing the organic liquid. The permeate vapors were collected in the cold trap while the vacuum was applied to the downstream side and weight of the cold trap before and after each run was measured for flux calculation. In the case of binary mixtures, the feed and the permeate composition were determined by gas chromatography (GC, 7890A, Agilent Technologies, Inc.) with a 30 m and 0.320 mm capillary column of HP-5 and flame ionizing detector (FID). The permeate samples were diluted with hexane in the 5:1 hexane/xylene weight ratio before injecting to the GC. Ideal separation factor was determined by taking the ratio of the flux of pure p-x to o-x. For binary mixture experiments, membrane pervaporation separation factor (  ) and total flux () are defined as:

  = =

  /   /

(2)

 ∙

(3)

where   and  , and,   and  are the molar fractions of p-x and o-x in the feed and permeate streams, respectively.  is the weight of the collected permeate (kg),  is the active membrane area (m2), and  is the permeation time interval for the pervaporation (h). The membrane permeance ( ) and selectivity ( ) were calculated by.

 =

 ∙   ,



− ,

(4)

ACS Paragon Plus Environment

8

Page 9 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

 =

  

(5)

  where  is the membrane thickness, , is the equilibrium vapor pressure of at the

pervaporation temperature, and , is the partial pressure for each component at the permeate   23 side. Since , is very low in the vacuum stream, it can be neglected compared to , .

Figure 1. Schematic of the setup for the vapor permeation measurements. Vapor permeation/separation of xylene isomers was performed on the setup schematically depicted in Figure 1. The membrane was sealed with Viton O-ring in a custommade stainless-steel permeation cell, which was placed in a high-temperature oven to control membrane temperature. The membrane in the cell was first activated at 150°C under helium flow for 2 h before measurements. The total partial pressure of xylene mixture in the feed stream was adjusted by regulating the flow rate of carrier gas through the saturator, in a range of 30–60 ml min-1 and temperature of saturator, which was filled with 50:50 p-x and o-x liquids, was kept at 25-60°C. The actual partial pressure of each xylene in the feed for each experiment was measured from the amounts of xylene collected in the cold-trap, and the total amount of feed gas

ACS Paragon Plus Environment

9

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 49

passed. The flow rate of the sweep gas was fixed at 20 ml min−1. The total pressure on either side of the membrane was maintained at the atmospheric pressure. To avoid any condensation and ensure proper partial pressure throughout the setup, all the tubing lines were heated and maintained at 180°C using heating tapes. The gas composition of retentate and permeate streams was analyzed by GC. The permeance of xylene vapor was calculated from the flux normalized by the transmembrane partial pressure difference. The p-x/o-x selectivity (!) was defined as the ratio of the vapor permeance of p-x over o-x.

!=

  

(6)

The accuracy of pervaporation and vapor permeation results was reported with ±3.0% for the permeation of pure components, and ±5.0% for binary mixtures separation data.

3 3.1

RESULTS AND DISCUSSION Characteristics of MFI Zeolite Membranes The SEM image of the top-view of the YSZ intermediate layer in Figure 2a confirms the

alumina support surface is well covered with aggregated YSZ particles with no cracks and defects. The average pore diameter for the supported YSZ intermediate layer is less than 100 nm, which is consistent with our previous study18. The YSZ-coated alumina support was then covered with MFI zeolite seeds. As can be seen in Figure 2b, The MFI zeolite seeds are roughly spherical with a uniform particle size less than 200 nm, and form a continuous layer that provides a comparatively flat surface covered with nucleation sites for the zeolite crystallization. Because of employing a well-dispersed and stable seeds suspension as the dip-coating sol, the aggregation of MFI particles is not observed.

ACS Paragon Plus Environment

10

Page 11 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 2. SEM images of top-view of (a) YSZ intermediate layer coated on an α-alumina support and (b) seeded support.

Figure 3 shows surface and cross-sectional view of MFI zeolite membranes after secondary growth with and without a template. From the top-view SEM images, both templated (Figure 3a) and templated-free (Figure 3c) MFI zeolite membranes are continuous, defect-free, and do not have well-defined grain boundaries. However, MFI zeolite layer obtained without template

ACS Paragon Plus Environment

11

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 49

shows better inter-grown with much higher integrity than that synthesized with the template. From the cross-section of the MFI zeolite membranes (Figure 3b and 3d), the zeolite layer, the YSZ intermediate layer, and the alumina substrate can be distinguished. The thickness of YSZ intermediate layer is about 4 µm, which effectively eliminates the influence of the uneven surface of porous alumina support. The thickness of the templated and template-free MFI zeolite membranes used in this study is respectively 9.7 and 5.5 µm, due to the difference in synthesis conditions. The XRD patterns of MFI zeolite seeded α-alumina support, and as-prepared templated and template-free MFI zeolite membranes are shown in Figure 4. It can be clearly observed that the MFI zeolite seeded support obtained from 3 sequential dip-coating-drying-calcination processes, and as-prepared templated and template-free MFI zeolite membranes all contain diffraction peaks for MFI-type zeolite crystal and alumina, confirming the formation of thin MFI zeolite layer on the α-alumina supports. The diffraction pattern of zeolite top-layer are in agreement with the standard MFI crystal structure

24

, and confirms these polycrystalline zeolite

membranes formed have a random orientation. The peak intensity ratio of the MFI zeolite layer to alumina substrate increases in the order: seeded support < template-free MFI zeolite membrane < templated MFI zeolite membrane, due to an increase in crystallinity and the thickness of MFI zeolite layer.

ACS Paragon Plus Environment

12

Page 13 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

ACS Paragon Plus Environment

13

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 49

Figure 3. SEM images of top-view and cross-section of (a, b) templated and (c, d) template-free MFI zeolite membrane.

Figure 4. XRD patterns of (a) MFI zeolite seeded α-alumina support, (b) synthesized templatefree MFI zeolite membrane, and (c) synthesized templated MFI zeolite membrane.

ACS Paragon Plus Environment

14

Page 15 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

In this study, multiple measurements were conducted to characterize the performance of MFI zeolite membranes in pervaporation and vapor permeation. A heat treatment in which the membrane is exposed to an elevated temperature at 300oC for an extended period (4 h) was established based on TGA results (Figure 5) and was applied to remove the adsorbed xylene molecules from zeolite layer after the xylene permeation evaluation. Prior to TGA, MFI zeolite powder soaked in pure p-x and o-x solution, respectively, for 48 h. Figure S1 shows the weight loss of MFI zeolite as a function of temperature. Both TGA curves for MFI zeolites contain p-x and o-x show a near 9% weight loss with temperature increasing from 25 to 300oC. The zeolite weights decreased less than 0.5% in the first 2 h of isothermal step at 300oC, and it was almost constant during the rest of isothermal step at 300oC, and no significant change is observed under 400oC. The weight loss for p-x is about 1% higher than that for o-x, due to the higher adsorption of p-x than o-x in MFI zeolitic pores.

Figure 5. TGA results of MFI zeolite powders soaked in p-xylene and o-xylene solutions.

ACS Paragon Plus Environment

15

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3.2

Page 16 of 49

Gas Permeation and Xylene Pervaporation Properties

The He and SF6 gas permeances for as-prepared membranes are independent of the mean pressure (Figure 6), indicating the absence of macroscopic defects in the zeolite membranes. The ideal He/SF6 selectivities for templated and template-free MFI zeolite membranes are, respectively, 21 and 42, higher than Knudsen selectivity. This indicates the microporous MFI zeolite layers formed onto the α-alumina substrates are high-quality and almost free of intercrystalline gaps. It can be seen that membranes synthesized with a template that require a hightemperature calcination (~500oC) for template removal, shows higher gas permeance but lower selectivity than those fabricated without the use of the template. Since the thickness of the templated membrane is larger than template-free membrane, these results suggest template-free MFI zeolite membranes contain fewer transport pathways within the inter-crystalline regions, or non-zeolitic pores. The high-performance of as-prepared membranes could be attributed to the MFI zeolite crystals are highly inter-grown in the zeolite layer, which limits the formation of non-selective permeation paths through the layer. Also, the removal of the organic template at high temperature can either create or enlarge inter-crystalline gaps, leading to an improvement in ideal He/SF6 selectivity for template-free MFI zeolite membranes. Therefore, these achieved high gas selectivities indicate the high quality of the membranes prepared in this work. Table S3 in supporting information compares the He permeance, and He/SF6 permselectivity of MFI zeolite membranes obtained from previously studies and the present work. To qualitatively analyze the contribution of zeolitic pores and inter-crystalline gaps in gas diffusion, the permeation results of He and SF6 for templated and template-free MFI zeolite membranes are discussed using a transport model in Appendix 4 in supporting information.

ACS Paragon Plus Environment

16

Page 17 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 6. Effect of mean pressure on (a) permeance and (b) ideal selectivity of He and SF6 for templated and template-free MFI zeolite membranes at 25oC. Pervaporation is another practical approach to evaluate the quality of MFI zeolite membranes. Table 1 lists the flux, permeance and separation factor obtained in pervaporation of single and binary xylene isomers through templated and template-free MFI zeolite membranes. The xylene fluxes and permeances recorded in pervaporation of single-component are higher than those of xylene mixture because a higher concentration of p-x or o-x in feed side leads to an enhancement of concentration gradient, which increases the driving force in pervaporation

ACS Paragon Plus Environment

17

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 49

process with a single-component feed. Compared with o-x, p-x suffers more reduction of flux and permeance in binary-component separation, lowering the separation factor as compared to the pure component pervaporation data. This could be explained by the fact that the transport of p-x in binary mixtures is restricted in the presence of the slower permeating component o-x. Moreover, compared with templated MFI zeolite membranes, those prepared by the templatefree method exhibit a lower flux and permeance but much higher separation factor in the pervaporation for both single and binary xylene isomers. Pure and binary components pervaporation results of templated and template-free membranes are in agreement with previously published data in our group.

3, 6

These findings suggest that fewer or no inter-

crystalline gaps exist in the template-free zeolite layer, which offers permselectivity mainly based on the differences in the molecule size-sensitive intra-crystalline (zeolitic) diffusivity.

Table 1. Pervaporation performance of xylene isomers for MFI zeolite membranes at 25oC. Flux (kg m−2 h−1) Membrane Templated Templatefree

3.3

Single

Permeance (10-6 mol m−2 h−1 Pa−1)

Binary

Single

Binary

Separation Factor (SF) Ideal

Binary

p-x

o-x

p-x

o-x

p-x

o-x

p-x

o-x

p-x/o-x

p-x/o-x

0.71

0.28

0.27

0.11

1.43

0.74

1.09

0.58

2.2

1.9

0.17

0.007

0.08

6 ?@A9B

6 89:; 2
89:; L:

]

(14)

Thus, the measured selectivity decreases with increasing activity due to the effect of activity on two terms, as shown in eq. (14). The first term on the right represents the decrease in the zeolitic perm-selectivity due to relative changes in the solubility and diffusivity for p-x or o-x in zeolitic pores with increasing activity. The measured activity decrease with increasing activity is magnified by the second term in eq. (14), which is related to the presence of the intercrystalline gaps. It is known that increasing xylene loading in MFI zeolite framework can result in a slight change in the unit cell of the zeolite. For example, as xylene loading increases from 0 to 2 molecules per unit cell (with increasing activity), the unit cell volume decreases from 5.211 nm3 to 5.209 nm3

28

corresponding to less than 0.01% change in the crystallite size (assuming cubic

crystals). Such small change in the crystal size would cause a negligible change in crystalline gap size and porosity. Thus, the permeance through intercrystalline gaps depends minimally on the activity of xylene, and - 45 can be considered constant for a given membrane within the range of activity studied. As a result, increasing activity would cause a decrease in only -01 , and thus a decrease in the second term of eq. (14). At the given activity, a membrane with larger - 45 would have a smaller second term, and hence lower measured selectivity. Comparing the

ACS Paragon Plus Environment

26

Page 27 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

templated and template-free membranes, the former has - 45 larger than the later. From eq. (14), the decreasing -01 due to increase in activity will have a more pronounced effect on selectivity. This explains more decrease in the separation factor as the activity increases for templated membranes than template-free membranes. The strong activity dependence of the xylene separation performance indicates importance to compare xylene separation/permeation data for zeolite membranes from different studies in the same activity range. Figure 9 compares the xylene separation performance for randomly oriented MFI zeolite membranes obtained in this study at a partial pressure of 1.05 and 1.65 kPa with data reported in the literature and summarized by Daramola et al.1 via vapor permeation at low xylene partial pressures (