DOE OSTI2019
Fast pyrolysis is an effective process to convert more than 60 wt. % of lignocellulosic materials into a bio-oil that can be further hydrotreated for transportation fuel production. However, bio-oil is acidic, viscous, and chemically unstable due to its high oxygen content. Therefore, increasing attention has being paid to deoxygenation of bio-oil through catalytic hydrotreatment into a “drop-in” fuel has been gaining attention from researchers. The first section of this review is devoted to summarize our current understanding of bio-oil composition. Bio-oil hydrotreatment is typically conducted in two-steps. In the first step, known as stabilization, the reactive carbonyl and carboxyl functional groups are transformed into alcohols at a temperature between 373 and 573 K. In the second step, between 623 and 673 K, cracking and hydro-deoxygenation (HDO) occurs. The most common catalysts used in the stabilization step are noble metals (Pt, Ru, Pd) supported on carbon and metal oxides (TiO2, Al2O3, SiO2, Fe2O3, ZrO2). The most common catalysts for deoxygenation and hydrocracking are sulfided CoMo or NiMo. More recently transition metal phosphides and carbides have been shown to be active. The effect of hydrotreatment parameters and the main global reaction mechanisms are discussed. In the first section, we update the progresses made on bio-oil composition, followed by a detailed discussion on the role of heterogeneous catalysts. Many of the bio-oil hydrotreament studies in the literature are based on model compound results. The reactions revealed by such approach are summarized and discussed. Although these studies provide valuable information for developing upgrading strategies, bio-oil chemistry is difficult to describe by model compounds. Thus, we also reviewed a smaller number of hydrotreatment studies of pyrolysis oil fractions. The presence of several oxygenated functional groups (phenols, carboxylic acids, carbonyls, alcohols) in monomeric and oligomeric molecules is a major challenge to study hydrotreatment reactions. Under realistic industrial operational conditions undesirable polymerization and crosslinking reactions leading to catalyst deactivation are very important. In spite of the extensive work on bio-oil hydrotreatment, the nature of the reactions happening in these complex oils is still poorly known. Pyrolytic lignin and sugars have been identified as important contributors to coke formation. Both fractions can then undergo successful hydrogenation or hydrogenolysis over the catalysts. The review ends with a discussion on future prospects and challenges to hydrotreat pyrolysis oils. Because bio-oil is a liquid formed by hundreds of molecules covering a wide range of molecular weights and functionalities, it is very unlikely that a single hydrotreatment technology that makes a good use of each of its molecules could be found. In fact, current hydrotreatment schemes sacrifice some fractions (convert them into low value products) to maximize the conversion of others into fuel. Separation of bio-oil in fractions and processing of each of them via hydrotreatment or by other suitable technology at optimized conditions with well suited catalysts to obtain high value products seems to be the most promising path to reduce hydrogen consumption, coke formation and enhance the economic viability of bio-oil refineries.