(811b) Enhancement of Membrane Pervaporation By Manipulation of Nanoparticles Within the Concentration Boundary Layer | AIChE

(811b) Enhancement of Membrane Pervaporation By Manipulation of Nanoparticles Within the Concentration Boundary Layer

Authors 

Smart, J. L. - Presenter, University of Kentucky



Dilute concentrations of volatile organic compounds (VOCs) can be effectively pervaporated from aqueous systems with use of silicone rubber hollow fibers [1,2].  A source of vacuum or sweep gas on the downstream side of a membrane can be effectively used to provide the driving force for pervaporation of these VOCs.  
    With use of pervaporation technology, the effect of mixing perturbations in the boundary layer adjacent to the upstream side of a silicone rubber hollow fiber was studied.  With use of a sweep gas imposed on the downstream side of the hollow fiber, the downstream side resistance is removed, and overall resistance to component permeation is reduced to only two resistances.  These are resistances offered by the membrane wall thickness (easily designed and controlled) and resistance of the boundary layer.  Resistance between the bulk feed solution and interface of the membrane is an interplay of several competing phenomena, including effects of component diffusion, component sorption, and hydrodynamics.  
    Resistance of the concentration boundary layer cannot be entirely eliminated, but its effects can be reduced.  Varying concentrations of ferromagnetic nanoparticles (size < 100 nm) were introduced into a dilute feed mixture of toluene/water.  As the feed flowed through the inside of the hollow fiber, external magnets were used to manipulate movement of the particles.  Perturbation effects upon the boundary layer were evaluated.  Micromixing action provided by the tiny magnetized nanoparticles act to increase turbulence and promote component diffusion across the effectively-thinned boundary layer.  Besides precise control of mixing action, another advantage of the use of magnetized particles is the ability to easily separate and collect them from the feed retentate for recycle.
    The work as outlined in this study is the first known application of use of magnetic fields to enhance component flux through a membrane.  Other investigators have used sonication, vibration, mechanical turbulence promoters, moving boundaries, gas sparging, and pulsating flow to enhance component flux.  In many cases, these earlier studies have shown marked improvements in flux, but it is believed manipulation of magnetic particles within the boundary layer offers added advantages of better process control through manipulation of magnetic fields and the reduction of fouling layers that accumulate on the inside wall of a hollow fiber membrane.
    The initial experimental apparatus consists of a single silicone rubber hollow fiber that is surrounded by a variable rotating geometric arrangement of permanent magnets.  The toluene/water feed containing a low concentration of ferrofluid is pumped through the inside of the hollow fiber.  A nitrogen sweep gas is fed past the outside surface of the membrane to collect the permeated species.  The sweep gas is fed directly into a gas chromatograph to quantify toluene flux.  A side stream of the sweep gas is also condensed into a cold trap for later quantification to evaluate membrane separation.  Early efforts in this study using magnetic nanoparticles have demonstrated greater than double the flux rate over conventional hollow fiber membranes.  Continuing efforts are underway to optimize flux rates by improved location and manipulation of magnetic fields surrounding the nanoparticles.

1.  J. Smart, R.C. Schucker, and D.R. Lloyd, Pervaporative extraction of volatile organic compounds from aqueous systems with use of a tubular transverse flow module.  Part 1.  Composite membrane study, J. Mem. Sci., 143, (1998), pp. 137 - 157.
2.  J. Smart, V.M. Starov, R.C. Schucker, and D.R. Lloyd, Pervaporative extraction of volatile organic compounds from aqueous systems with use of a tubular transverse flow module.  Part 2.  Experimental results, J. Mem. Sci., 143, (1998), pp. 159 - 179.

Topics