(315c) Carboxylic Acid Effects on Ethanol Recovery from Aqueous Mixtures Using Pervaporation through Mfi Zeolite-Filled Polydimethylsiloxane Membranes

Most bioethanol is produced by fermenting sugars released from biomass and using distillation to recover the ethanol. Recovering ethanol from the fermentation broths using pervaporation through hydrophobic membranes is potentially economically competitive with distillation for small- to medium-scale applications (~ 2 ? 20 million liters per year).[1] In addition, pervaporation can be used during fermentation to maintain ethanol concentrations at levels where the microbes produce ethanol efficiently.

Pure polydimethylsiloxane (PDMS) membranes typically deliver ethanol/water pervaporation separation factors around 8, whereas mixed-matrix membranes with high-silica ZSM-5 (MFI structure) zeolite dispersed in the PDMS matrix have delivered much higher pervaporation separation factors as high as 59.[1] These separation factors are usually measured using aqueous ethanol binary feed mixtures, but many compounds are present in fermentation broths. These additional compounds can include carboxylic acids, such as acetic acid, which may be produced by the microbes as byproducts or generated during hydrolysis of biomass.[2]

The effects of acetic acid on long-term pervaporation of ethanol/water mixtures using high-silica ZSM-5 zeolite-filled PDMS membranes were investigated. Pervaporation was carried out continuously for extended periods of time for each membrane. Initially, the feed contained 5 wt% ethanol in water. After several days, acetic acid was added to the feed at 1 wt%. The ethanol and water pervaporation fluxes through a pure PDMS membrane were not measurably affected by the presence of acetic acid. When mixed-matrix membranes containing 50 ? 65 wt% ZSM-5 zeolite particles dispersed in PDMS were used, however, both ethanol and water fluxes decreased when acetic acid was present. One day after adding acetic acid, the ethanol and water fluxes were around 80 % of the pre-acetic acid values. This initial flux decrease was due to acetic acid molecules competing with ethanol and water for adsorption sites on the zeolite particles. The flux decrease was reversible if the feed was changed back to 5 wt% ethanol in water after a short time (~2 days). Longer-term exposure to acetic acid resulted in a steady decline in ethanol flux that was not reversible upon changing the feed back to 5 wt% ethanol in water. In contrast, when the feed pH was increased above the pKa value of acetic acid, the ethanol and water fluxes did not change measurably in the short or long term after acetic acid addition to the feed. At increased pH, most of the acetic acid molecules are dissociated into the corresponding ions, which do not adsorb easily into the PDMS or the zeolite particles. Acetic acid-aluminum complex formation in the ZSM-5 framework was postulated as a potential reason for the irreversible decline in ethanol flux after long term exposure to acetic acid with no pH adjustment. Thus, membranes with decreased aluminum content in the zeolite were tested. Adsorption of carboxylic acids on zeolite powder was also investigated to further clarify possible reasons for the observed pervaporation behaviors.

[1] L.M. Vane, A review of pervaporation for product recovery from biomass fermentation processes, J. Chem. Technol. Biotechnol., 80 (2005) 603-629.

[2] C. Tengborg, M. Galbe, G. Zacchi, Reduced inhibition of enzymatic hydrolysis of steam-pretreated softwood, Enzyme Microb. Technol., 28 (2001) 835.

*This is an abstract of a proposed presentation and does not necessarily reflect U.S. EPA policy.