(27a) Condensed Phase Ketonization of Organic Acids Produced By the Hydrothermal Liquefaction of Lignocellulosic Biomass
AIChE Spring Meeting and Global Congress on Process Safety
2015 AIChE Spring Meeting and 11th Global Congress on Process Safety
Fuels and Petrochemicals Division
Developments in Bio-Petrochemicals I
Monday, April 27, 2015 - 1:30pm to 2:00pm
The direct liquefaction of lignocellulosic biomass (e.g. fast pyrolysis, catalytic fast pyrolysis, or hydrothermal liquefaction) and subsequent bio-oil upgrading is one strategy for producing hydrocarbon fuels from biomass. The aqueous stream generated during the hydrothermal liquefaction (HTL) of terrestrial biomass contains 35-40 percent of the total carbon fed to the HTL reactor. Short chain organic acids (e.g. acetic acid) are one of the primary organic constituents of the aqueous phase. The ketonization of these short organic acids whereby two organic acid equivalents react to form a single ketone in the condensed aqueous phase was investigated. Ketonization is desirable because 1) a C-C bond is formed between two shorter chain compounds; 2) ¾ of the O present in the original reactants is rejected as one equivalent of CO2 and H2O each without the need for additional hydrogen; 3) the ketone products have higher vapor pressure than the parent acids and can be readily concentrated via distillation. Furthermore, performing the reaction in the condensed phase is advantageous because the energy intensive step of vaporizing a high volume stream rich in water is avoided. High throughput catalyst screening via small (~4 mL) batch reactors at 275°C and ~1000 psig (N2) was performed to identify active catalysts. Selected catalysts were tested in a 3 mL flooded bed continuous reactor at 300°C and 1350-1400 psig. The effect of WHSV and acetic acid concentration on the rate of converting acetic acid to acetone was investigated. Model compounds as well as real feeds were tested. Under similar conditions, the rate of ketonization using the real feed was only about one-half that observed with model compounds. The lowered rate when utilizing real feeds may be due to a high concentration of solubilized Na (4000-6000 ppm) present from the HTL process. Fresh and spent catalyst characterization will be reported to suggest catalyst structure/activity relationships. Downstream, it is envisioned the ketones could undergo further processing such as reduction and dehydration steps to produce olefins. An overall flow diagram of the conceptualized process for producing olefins as well as preliminary techno-economic analysis incorporating an HTL bio-refinery will be presented.