Keynote Presentation
Pervaporation Membranes from Addition Polynorbornenes for
Effective Butanol Recovery
Dr. Richard A. Register
Department of Chemical and Biological Engineering
Princeton University, Princeton, NJ
While ethanol is the best-known fermentation product, there is growing interest (both research
and commercial) in the production of higher alcohols, such n-butanol and isobutanol. Currently,
such molecules are already utilized as chemical building blocks; in the longer term, they could
potentially serve as biofuels more compatible than ethanol with our fuel distribution and
utilization infrastructure, and with higher energy density. While there have been significant
advances in the biosynthesis of such heavier alcohols, an unmet need is for their economical
recovery from aqueous solutions (fermentation broth). Since the broth must be maintained dilute
in these higher alcohols (which are toxic to the producing organisms), isolation of the alcohol by
distillation is energetically prohibitive, and a lower-energy separation process is sought;
membrane pervaporation (liquid feed, vapor permeate) is the leading candidate. Currently, the
state-of-the-art membranes for such separations are based on crosslinked polydimethylsiloxane
(PDMS), a soft rubbery polymer which must be applied as a relatively thick layer to maintain
mechanical integrity.
Polymers derived by vinyl addition polymerization of substituted norbornenes—where the
backbone of every repeat unit contains a bicyclic norbornene unit—have extremely high glass
transition temperatures (T g ), and thus retain stiffness even in very thin, uncrosslinked films. Since
transmembrane flux is inversely proportional to thickness, such high-T g polymers are potentially
attractive for separation membranes (high flux), if they can also exhibit high component
selectivity—which is potentially tunable through the chemical identity of sidechains attached to
the norbornene monomer. However, until recently, initiators which could enchain norbornenes
via vinyl addition polymerization showed very poor tolerance of even relatively unreactive
sidegroups. In collaboration with Promerus LLC, we have shown that certain Pd-
trialkylphosphine complexes can initiate the living homopolymerization of a rather broad range
of substituted norbornenes, bearing alkyl, aryl, perfluoroaryl, and even hexafluoroisopropanol
(HFA) substituents. Moreover, these initiators can also yield random, gradient, and block
copolymers of narrow molecular weight distribution and targeted block molecular weights.
Optimized polymers containing the HFA substituent outperform PDMS in both selectivity and
flux for recovery of n-butanol from dilute aqueous solution, providing a potential solution to this
challenging separation problem.
2