19-01-2013, 12:46 PM
A CONTINUOUS PERVAPORATION MEMBRANE REACTOR FOR THE STUDY OF ESTERIFICATION REACTIONS USING A COMPOSITE POLYMERIC/CERAMIC MEMBRANE
A CONTINUOUS.pdf (Size: 807.78 KB / Downloads: 52)
Abstract
The esterification reaction between acetic acid and ethanol was studied in a continuous flow
pervaporation membrane reactor utilizing a polymeric/ceramic composite membrane. For a range of
experimental conditions reactor conversions were observed which are higher than the corresponding
calculated equilibrium values. This is due to the ability of the membrane to remove water, a product of the
reaction. A theoretical model has been developed which gives a reasonable fit of the experimental results.
Copyright '~(~ 1996 Elsevier Science Ltd
INTRODUCTION
Polymeric membranes have found uses in gas and
liquid separations. Polymers, like PVA, polyimide,
and Nation have a high permselectivity to a number
of chemical compounds. They are commonly used as
membranes in industrial separations. A review paper
on the use of polymeric membranes for gas separations
has recently been published (Koros and
Fleming, 1993) discussing materials selection, membrane
formation, and trends in module and system
design.
Polymeric/inorganic composite type membranes
have received increased attention in the last few years
(Kita et al., 1987; Rezac and Koros, 1994, 1995). Such
membranes are formed by the deposition of a thin
dense polymeric film on an underlying inorganic
(metal or ceramic) macroporous support. The inorganic
layer support is believed to endow such membranes
with a higher mechanical strength while maintaining
the permselectivity of the dense polymeric film
and increasing the overall permeance. An appropriate
coating technique which deposits a crack-free polymeric
film on the underlying inorganic porous support
is important in the preparation of such composite
membranes.
EXPERIMENTAL APPARATUS AND PROCEDURES
Membrane preparation
The membranes utilized in the experiments reported
here were prepared by the dip-coating method
as described by Kita et al. (1987). A polyetherimide
(UltemR-1000) kindly provided to us by G.E. was
dissolved in dichloroethane to form a 3 wt.% polymer
solution, while stirring on the top of a heating plate.
A ceramic support tube (see below) was then dipped in
the polymeric solution and allowed to stay for 1 h.
The tube was then withdrawn from the solution, excess
solution was removed and it was allowed to dry
overnight. The dip-coating procedure was repeated
several times until the desired permeability is attained,
as measured by conventional techniques which are
described in prior publications (Champagnie et al.,
1992). It was not attempted to make the membrane
completely impermeable to Ar (i.e. defect-free). In our
case a composite membrane is considered defect-free
as long as its Ar permeability is less than 40 GPU
[ 1 GPU = 10- 6 cm 3 (STP)/(cm 2 s cmHg)].
Experimental procedures~data analysis methods
The permeabilities of the liquid reactants (ethanol
and acetic acid) and products (ethyl acetate and water)
through the membrane were measured in situ in the
reactor. A typical experiment consisted of feeding
a mixture of known composition in the tubeside at
a predetermined flow rate. Samples on the tubeside
and shellside were withdrawn and analyzed by gas
chromatography. Phase separation has not been observed
during the experiments for any of the mixtures
applied in agreement with the theoretical calculations
of Furzer (Furzer, 1994). For small space times, the
changes in the tubeside composition are negligible
and one can calculate the permeabilities at the average
composition. For larger space times, the permeability
must be calculated iteratively by using eq. (8)
below, after zeroing the reaction term.
RESULTS AND DISCUSSION
For the series of experiments from which data are
reported in this paper a total of three membranes were
used, designated as membranes # 1, 22 and 23.
They were all prepared by a similar procedure. Their
liquid permeances were not exactly the same, but
generally close to each other as can be seen in Table 1.
This table lists the water permeance and water to
ethanol separation factor ~H20/EtOH (the ratio of the
molar fraction of water to the molar fraction of
ethanol in the downstream over the same ratio in the
upstream) measured under similar experimental conditions
(T = 60°C, a composition of 37% each of
water and ethyl acetate, and 13% each of ethanol and
acetic acid for membrane # 1; and a composition of
37.5% each of water and ethyl acetate, and 12.5%
each of ethanol and acetic acid for membranes # 2
and # 3). Based on these data, it is our belief, that the
conclusions drawn on concentration polarization
effects and the concentration dependence of permeance
based on experiments with individual membranes
(see discussion to follow) also apply generally
to all the other membranes.
CONCLUSIONS
A model of a pervaporation membrane reactor
has been presented. It gives a reasonable fit to a
set of experimental data which were measured with
the esterification reaction between ethanol and acetic
acid. In the experiments a new type of composite
polymeric membrane has been utilized which was
prepared in our laboratories. The membranes show
reasonable fluxes and separation efficiencies towards
water, a product of the esterification reaction, as
can be seen in Table 2 which summarizes prior
data with other water permeable membranes. During
membrane reactor experiments removal of the
water from the membrane tubeside results in
reactor conversions which exceed the calculated
equilibrium conversions for a region of experimental
conditions,