A catalyst composed primarily of magnetite prepared from red mud via H2 reduction at 300 °C, simultaneously reduced acidity, allowed recovery of carbon, and generated upgradable intermediates from the aqueous fraction of fast pyrolysis oil in a “continuous” process.

No abstract

A mixture of fast pyrolysis oil (FPO) and methanol (1/1 v/v) was continuously converted to methyl levulinate (ML), methyl acetate (MA), and C3 or greater methyl esters using metal-acid functionalized zeolites (Ni and Ru/HZSM-5) and an iron oxide catalyst with both acid and base sites (250 °C, 600 psig). Fractional conversion of FPO components was 60% or greater using the iron oxide catalyst, and space time yields approached 150 and 30–50 g/L cat/h for MA and C3 methyl esters, respectively, at 250 °C (W/F = 0.4 h, liquid hourly space velocity = 5–11.2 h–1). Product yield and concentration using the iron oxide catalyst were comparable to those of the Ni and Ru/HZM-5 catalysts and achieved performance levels higher than those of SiO2‑Al2O3 and HZSM-5. Two potential pathways for acetic acid conversion (ketonization and esterification) and ML formation from levoglucosan were observed. Using the bifunctional catalysts in the presence of hydrogen resulted in significant coke reduction (60–80%) and the production of esters of carboxylic acids C3 or greater (e.g., pentanoic and hexanoic acid methyl esters) and MA from the mixture. More interestingly, contrary to the other catalysts, an increase in phenolic levels (e.g., 2-methoxy phenol) was observed using the iron oxide catalyst with H2 and isopropanol (replacing H2), indicating the presence of undetected lignin oligomers in the feed and their subsequent hydrogenolysis. Simultaneous esterification and hydrogenation resulted in percent reduction in total acid numbers ranging from 66 to 76%.

A two-stage catalytic process for converting model bio-oil oxygenates (acetic acid, formic acid, acetol, and levoglucosan) to higher value compounds was conducted. Ketonization using iron oxide catalysts prepared from red mud was implemented in a packed bed reactor system using fast pyrolysis oil model compounds (stage one). High levels of acetone (15− 25 g/L), 2-butanone (∼5 g/L), and cyclic ketones (9−13 g/L) were observed. Time-on-stream studies (7 h) indicated no measurable decline in conversion of acetol, formic acid, and levoglucosan and only a 4% decline in acetic acid conversion. Subsequently, a second stage continuous hydrogenation of the ketonization products was conducted using Pd on activated carbon monolith (ACM) catalysts generated from wood and Pd−C granules. The best results were achieved using the Pd-ACM catalyst at 180 °C and 300 psi (H 2 ), which converted ketones to alcohols at ∼60−80%. The Pd-ACM catalyst achieved higher space time yields, selectivity, and conversions compared to Pd−C granules.

A mixture of levoglucosan (LG) and acetic acid (AA), representing water extracted fast pyrolysis oil, was continuously converted to ethyl levulinate (EL) and ethyl acetate (EA) using H-ZSM5 [120–230 °C, 600 psig, 80% ethanol (v/v)]. Fractional conversion of both reactants was 65% or greater at temperatures above 120 °C, and space time yields (STY) approached 140 and 15 g/L-cat/h for EA and EL, respectively, at 180 °C (LHSV = 4.9 h–1). Two potential pathways for EL formation from levoglucosan were apparent, one with glucose and ethyl α-d-glucopyranoside as intermediates and the other with furfural. Adding metal functionality (Ru/H-ZSM5) resulted in the production of valerate biofuels (esters of carboxylic acids C3 or greater; e.g., pentanoic and hexanoic acid ethyl esters) and EA from the mixture in the presence of hydrogen. Conversions for LG and AA using Ru/H-ZSM5 were similar to H-ZSM5, but ethyl levulinate space time yield declined (∼5 g/L-cat/h) as valerate biofuel STY increased (∼10 g/L-cat/h) at an optimum temperature of 180 °C. Our results indicate that valerate biofuels can be produced from levoglucosan (and possibly other sugars) in a continuous single stage, integrated process. However, due to low yields and coke formation, it is clear that ethanol/water ratios, pore size, and acid site type and density must be optimized when coupled with metal functionality for industrial application.

No abstract