-
[show abstract]
[hide abstract]
ABSTRACT: Since their first synthesis in the 1940s, zeolites have found wide applications in catalysis, ion-exchange, and adsorption. Although the uniform, molecular-size pores of zeolites and their excellent thermal and chemical stability suggest that zeolites could be an ideal membrane material, continuous polycrystalline zeolite layers for separations were first prepared in the 1990s. Initial attempts to grow continuous zeolite layers on porous supports by in situ hydrothermal synthesis have resulted in membranes with the potential to separate molecules based on differences in molecular size and adsorption strength. Since then, further synthesis efforts have led to the preparation of many types of zeolite membranes and better quality membranes. However, the microstructure features of these membranes, such as defect size, number, and distribution as well as structure flexibility were poorly understood, and the fundamental mechanisms of permeation (adsorption and diffusion), especially for mixtures, were not clear. These gaps in understanding have hindered the design and control of separation processes using zeolite membranes. In this Account, we describe our efforts to characterize microstructures of zeolite membranes and to understand the fundamental adsorption and diffusion behavior of permeating solutes. This Account will focus on the MFI membranes which have been the most widely used but will also present results on other types of zeolite membranes. Using permeation, x-ray diffraction, and optical measurements, we found that the zeolite membrane structures are flexible. The size of defects changed due to adsorption and with variations in temperature. These changes in defect sizes can significantly affect the permeation properties of the membranes. We designed methods to measure mixture adsorption in zeolite crystals from the liquid phase, pure component adsorption in zeolite membranes, and diffusion through zeolite membranes. We hope that better understanding can lead to improved zeolite membranes and eventually facilitate the large-scale application of zeolite membranes to industrial separations.
Accounts of Chemical Research 08/2011; 44(11):1196-206. · 21.64 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Defect-free, microporous Al(2)O(3)/SAPO-34 zeolite composite membranes were prepared by coating hydrothermally grown zeolite membranes with microporous alumina using molecular layer deposition. These inorganic composite membranes are highly efficient for H(2) separation: their highest H(2)/N(2) mixture selectivity was 1040, in contrast with selectivities of 8 for SAPO-34 membranes. The composite membranes were selective for H(2) for temperatures up to at least 473 K and feed pressures up to at least 1.5 MPa; at 473 K and 1.5 MPa, the H(2)/N(2) separation selectivity was 750. The H(2)/CO(2) separation selectivity was lower than the H(2)/N(2) selectivity and decreased slightly with increasing pressure; the selectivity was 20 at 473 K and 1.5 MPa. The high H(2) selectivity resulted either because most of the pores in the Al(2)O(3) layer were slightly smaller than 0.36 nm (the kinetic diameter of N(2)) or because the Al(2)O(3) layer slightly narrowed the SAPO-34 pore entrance. These composite membranes may represent a new class of inorganic membranes for gas separation.
Journal of the American Chemical Society 01/2011; 133(6):1748-50. · 9.91 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Gated ion diffusion is found widely in hydrophobic biological nanopores, upon changes in ligand binding, temperature, transmembrane voltage, and mechanical stress. Because water is the main media for ion diffusion in these hydrophobic biological pores, ion diffusion behavior through these nanochannels is expected to be influenced significantly when water wettability in hydrophobic biological nanopores is sensitive and changes upon small external changes. Here, we report for the first time that ion diffusion through highly hydrophobic nanopores (approximately 3 nm) showed a gated behavior due to change of water wettability on hydrophobic surface upon small temperature change or ultrasound. Dense carbon nanotube (CNT) membranes with both 3-nm CNTs and 3-nm interstitial pores were prepared by a solvent evaporation process and used as a model system to investigate ion diffusion behavior. Ion diffusion through these membranes exhibited a gated behavior. The ion flux was turned on and off, apparently because the water wettability of CNTs changed. At 298 K, ion diffusion through dense CNT membranes stopped after a few hours, but it dramatically increased when the temperature was increased 20 K or the membrane was subjected to ultrasound. Likewise, water adsorption on dense CNT membranes increased dramatically at a water activity of 0.53 when the temperature increased from 293 to 306 K, indicating capillary condensation. Water adsorption isotherms of dense CNT membranes suggest that the adsorbed water forms a discontinuous phase at 293 K, but it probably forms a continuous layer, probably in the interstitial CNT regions, at higher temperatures. When the ion diffusion channel was opened by a temperature increase or ultrasound, ions diffused through the CNT membranes at a rate similar to bulk diffusion in water. This finding may have implications for using CNT membrane for desalination and water treatment.
Journal of the American Chemical Society 06/2010; 132(24):8285-90. · 9.91 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: A method is presented to prepare high-density, vertically aligned carbon nanotube (VA-CNT) membranes. The CNT arrays were prepared by chemical vapor deposition (CVD), and the arrays were collapsed into dense membranes by capillary-forces due to solvent evaporation. The average space between the CNTs after shrinkage was approximately 3 nm, which is comparable to the pore size of the CNTs. Thus, the interstitial pores between CNTs were not sealed, and gas permeated through both CNTs and interstitial pores. Nanofiltration of gold nanoparticles and N(2) adsorption indicated the pore diameters were approximately 3 nm. Gas permeances, based on total membrane area, were 1-4 orders of magnitude higher than VA-CNT membranes in the literature, and gas permeabilities were 4-7 orders of magnitude higher than literature values. Gas permeances were approximately 450 times those predicted for Knudsen diffusion, and ideal selectivities were similar to or higher than Knudsen selectivities. These membranes separated a larger molecule (triisopropyl orthoformate (TIPO)) from a smaller molecule (n-hexane) during pervaporation, possibly due to the preferential adsorption, which indicates separation potential for liquid mixtures.
Nano Letters 01/2009; 9(1):225-9. · 13.20 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Methods that use n-hexane (n-hexane permporosimetry and n-hexane/2,2-dimethylybutane (DMB) separation) are shown to not be effective for characterizing MFI zeolite membranes because n-hexane adsorption swells MFI crystals and shrinks the size of nonzeolitic pores. Measurements on a membrane in which 30% of its helium flux at 300 K was through nonzeolitic pores demonstrate that benzene permporosimetry and isooctane vapor permeation as a function of feed activity provide better characterizations. Isooctane condensed in nonzeolitic pores at high activities, and this was used to estimate the sizes of those pores. The average nonzeolitic pore size in this membrane decreased from approximately 3.0 to 1.5 nm as the temperature increased from 300 to 348 K, apparently due to thermal expansion of MFI crystals. Benzene permporosimetry yielded dramatically different results from n-hexane permporosimetry because benzene does not swell the MFI crystals significantly. Single-component pervaporation fluxes as a function of molecular kinetic diameter verified the results from benzene permporosimetry. Larger molecules had higher fluxes than n-hexane because they diffused through nonzeolitic pores that were shrunk by n-hexane adsorption. Nonzeolitic pores were estimated to account for only 0.5% of the membrane permeation area, but 30% of the helium flux, because these pores were significantly larger than MFI pores.
04/2008;
-
[show abstract]
[hide abstract]
ABSTRACT: Isotopic transient permeation of methanol/ethanol, methanol/2-propanol, and methanol/acetone mixtures through a MFI zeolite membrane was investigated experimentally. For the methanol/2-propanol and methanol/ethanol mixtures, the more mobile species (methanol) was slowed down in the mixture and the tardy species (2-propanol or ethanol) was speeded up. The extent of slowing down and speeding up depended on the mixture composition. The Maxwell–Stefan (M–S) diffusion equations reproduced the observed mixture permeation results, at least qualitatively, provided the self-exchange coefficient Ð ii for each species was taken to be a tenth of the pure component M–S diffusivity Ð i . For the methanol/acetone mixtures, both species slowed down in the mixture. Adjusting the value of the self-exchange coefficient Ð ii in the M–S equations did not provide an explanation of the observed experimental results; it appeared that the component diffusivities, Ð 1 and Ð 2 , in the mixture were both lower than the values of the pure components, an effect that has not earlier been reported in the literature.
Journal of Membrane Science 01/2007; 293:167-173. · 3.85 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: A technique that measures the effective density of a zeolite after adsorption from the liquid phase was developed to measure the absolute amounts of liquid mixtures adsorbed on zeolites without using a nonadsorbing solvent. Since the fugacities of the adsorbing components in solution can be dramatically different with or without the addition of a nonadsorbing solvent, this technique measures mixture isotherms that can be used for analyzing pervaporation through zeolite membranes. A nonideal solution, methanol/acetone, was used as an example to show that its adsorption isotherms on silicalite-1 zeolite at 294 K differ dramatically from those measured with the nonadsorbing solvent method. The methanol/acetone fugacity ratio is different for the two methods because of different concentrations in the liquid phase. Methanol preferentially adsorbs on silicalite-1 at low methanol concentrations and acetone preferentially adsorbs at high methanol concentrations. The density bottle method was used to show that n-hexane preferentially adsorbs from n-hexane/3-methylpentane liquid mixtures, and at high n-hexane concentrations, essentially no 3-methylpentane adsorbs, as has been predicted previously by simulations. A larger molecule, 2,2-dimethylbutane, adsorbed so slowly at 294 K that silicalite had only 16% of saturation coverage after 370 h, but it was saturated after 1650 h; at 423 K, saturation was obtained in less than 24 h.
Langmuir 09/2005; 21(16):7390-7. · 4.19 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: A SAPO-34 zeolite membrane was made essentially impermeable to high-pressure hydrogen at room temperature by adsorbing methanol in the SAPO-34 layer. Hydrogen permeance decreased three orders of magnitude when the methanol feed activity was ∼0.1, and it decreased more than six orders of magnitude when the methanol feed activity was higher than 0.85 at 293 K. The hydrogen permeance at 293 K was below ∼10−14 mol/m2 s Pa for at least five days for a H2 feed pressure of 6.6 MPa. At higher temperatures, methanol desorbed and the H2 flux increased. The hydrogen permeance could be controlled by the activity of the methanol on the feed side. These results demonstrate that the SAPO-34 membrane had low fluxes through defects, and H2 flow through these defects was blocked by capillary condensation of methanol at high methanol activities.
Microporous and Mesoporous Materials 110:579-582. · 3.29 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Pervaporation and vapor permeation of hydrocarbons were used to characterize three MFI zeolite membranes that had significant fluxes through non-zeolitic pores. The single-component pervaporation fluxes of n-hexane were only 5–55% of the fluxes of larger molecules (2,2-dimethylbutane (DMB), isooctane, and 1,3,5-trimethylbenzene (TMB)) at ∼300 K, all of which did not adsorb in the MFI pores under the conditions used. In spite of this behavior, these membranes were selective for C6 isomer separation, and the n-hexane/DMB vapor-phase separation factor was 3200 for one membrane at low hydrocarbon partial pressures. The pervaporation behavior was consistent with the explanation that MFI membrane structures are flexible; MFI crystals increase in size when n-hexane adsorbs, and this expansion dramatically decreases the non-zeolitic pore sizes. Thus, less than 0.8 mol% n-hexane reduced DMB and TMB pervaporation fluxes to ∼5% of their single-component fluxes for one membrane. At high vapor concentrations, molecules condensed in the non-zeolitic pores, which had almost no selectivity, and this capillary condensation decreased the n-hexane/DMB separation factor. Non-zeolitic pores were shown to increase in size after a membrane was calcined at 1000 K.
Journal of Membrane Science.
