Levulinic acid (4-oxopentanoic acid) is an important biomass derivative that can be obtained by acid hydrolysis of lignocellulosic wastes, such as paper mill sludge, urban waste paper, and agricultural residues, through the Biofine process. [123] Levulinic acid has been recently selected by the US Department of Energy (DOE) as one of the top 15 carbohydrate — derived chemicals in view of its potential to serve as a building block for the development of biorefinery processes. [91] Thus, taking advantage of its dual functionality (i. e., a ketone and an acid group), a number of use­ful chemicals can be synthesized from levulinic acid including methyl — tetrahydrofuran (MTHF) (a gasoline additive) and d-aminolevulinic acid (DALA) (a biodegradable pesticide). [124]

Recently, our group has developed a series of catalytic approaches to convert aqueous solutions of levulinic acid into liquid hydrocarbon trans­portation fuels of different classes (Fig. 9). The catalytic pathways involve oxygen removal by dehydration/hydrogenation (in the form of water) and decarboxylation (in the form of CO2) reactions, combined with C-C cou­pling processes such as ketonization, isomerization, and oligomerization that are required to increase the molecular weight and to adjust the struc­ture of the final hydrocarbon product. As a first step, aqueous levulinic acid is hydrogenated to water-soluble GVL, which is the key intermedi­ate for the production of hydrocarbon fuels. This hydrogenation step can be achieved with high yields by operating at low temperatures (e. g., 423 K) over non-acidic catalysts (e. g., Ru/C) to avoid formation of angelica lactone, a known coke precursor [125] which is produced by dehydration over acidic sites at higher temperatures (e. g., 573-623 K). [119] Interest­ingly, because equimolar amounts of formic acid (a hydrogen donor) are coproduced along with levulinic acid in the C6-sugars dehydration pro­cess, this hydrogenation step could be potentially carried out without uti­lizing hydrogen from an external source, and several groups have already explored this possibility. [126,127] This route is promising in that GVL has applications as a gasoline additive, [128] and as a precursor to poly­mers [129] and fine chemicals. [130]

Aqueous solutions of GVL can be upgraded to liquid hydrocarbon fu­els by following two main pathways: the C9 route and the C4 route (Fig. 9). In the former route, GVL is converted to 5-nonanone over a water-stable multifunctional Pd/Nb2O5 catalyst. In this process, GVL is first transformed into hydrophobic pentanoic acid by means of ring-opening (on acid sites) and hydrogenation reactions (on metal sites) at moderate temperatures and pressures. Pentanoic acid is subsequently ketonized to 5- nonanone, and reaction conditions can be adjusted to allow this transformation to take place on the same Pd/Nb2O5 reactor with a maximum of 70% carbon yield. [119] Nonanone yield can be increased to almost 90% by using a dual-cat­alyst approach with Pd/Nb2O5 + Ce0 5Zr0 5O2 in a reactor with two different temperature zones, which allows for optimum control of reactivity. [131] 5-Nonanone, which is obtained in a high purity organic stream that spon­taneously separates from water, is subsequently transformed into its cor­responding alcohol that serves as a platform molecule for the production of hydrocarbon fuels for gasoline and diesel applications. For example, the C9 alcohol can be processed (through hydrogenation/dehydration cycles) over a bifunctional metal-acid catalyst such as Pt/Nb2O5 [110] into linear n-nonane, with excellent cetane number and lubricity to be used as a diesel blender agent. Alternatively, the functionality of 5-nonanol can be utilized to upgrade the alcohol to gasoline and diesel components. In particular, 5-nonanol can be dehydrated and isomerized in a single step over an USY

FIGURE 9: Catalytic routes for the conversion of levulinic acid (LA) and g-valerolactone (GVL) into liquid hydrocarbon transportation fuels. Blue colour indicates water-soluble compounds, yellow symbolizes hydrophobic compounds, and green refers to liquid hydrocarbon fuels.

zeolite catalyst to produce a mixture of branched C9 alkenes with the ap­propriate molecular weight and structure for use in gasoline after hydro­genation to the corresponding alkanes. [131] Additionally, 5-nonanol can be converted into a C9-alkene stream (by means of dehydration reactions) which can be subsequently oligomerized over an acid catalyst such as Am — berlyst 70 to achieve good yields of C18 alkanes (after hydrogenation) for diesel applications. [132]

Recently, a promising route to upgrade aqueous solutions of GVL into jet fuels through the formation of C4 alkenes has been developed by Bond et al. [133] (Fig. 9). The process is based on a dual reactor system. In the first catalytic reactor the GVL feed undergoes decarboxylation at elevated pres­sures (e. g., 36 bars) over a silica/alumina catalyst, producing a gas stream composed of butene isomers and CO2. In a second reactor connected in se­ries, the gaseous butene stream is passed over an acidic catalyst (H-ZSM5, Amberlyst 70) that achieves oligomerization of butene monomers, yielding

TABLE 1: Summary of the different technologies for the conversion of biomass into liquid hydrocarbon transportation fuels

TABLE 1: Cont.

a [] indicates that hydrolysis and dehydration can be carried out in the same reactor, b 3 reactors if WGS syngas conditioning is required. c Depending on the upgrading process required. d According to ref. 134. e Calculated as: [0.50-0.70 yield of bio-oil from lignocellulose in pyrolysis135] <?> [0.65yield of LHF in HDO,136 or 0.30yield of aromaticsLHF in zeolite upgrading79].f Calculated as: [0.20 content of oil in soybeans137] <?> [0.1 glycerol in biodiesel process] <?> [0.552yield of alkanes104]. g Calculated as: [0.80-0.60 sugar content of lignocellulose] <?> [0.96yield of hydrolysis enzymatic] <?> [0.53-0.31 yield isomerization glucose to fructose138,139] <?> [0.69-0.58 yield of LHF from fructoseU0]. h Calculated as: [0.80-0.60 sugar content of lignocellulose] <?> [0.96 yield of hydrolysis enzymatic] <?> [0.52 yield of organic C116] <?> [0.57 yield of C7+ ketones116]. i Yield to n-nonane (diesel blender). Calculated as: [0.80-0.60 sugar content of lignocellulose] <?> [0.96 yield of hydrolysis enzymatic] <?> [0.45 yield to levulinic acid of biofine process] <?> [0.96 yield of levulinic acid to GVL119] <?> [0.80 yield of 5-nonanone from GVL119] <?> [1.00] hydrogenation of 5-nonanone to n-nonane. j Calculated as: [0.80-0.60 sugar content of lignocellulose] <?> [0.96 yield of hydrolysis enzymatic] <?> [0.45 yield to levulinic acid of biofine process] <?> [0.96 yield of levulinic acid to GVL119] <?> [0.78 yield of C8+ alkenes133] <?> [1.00] hydrogenation to final alkane product.

a distribution of alkenes centred at C12. While CO2 does not affect the oligo­merization process other than by dilution, water inhibits the acidic oligo­merization catalyst (especially Amberlyst), and it has to be removed prior to the second reactor by using a gas-liquid separator operating at 36 bars and 373-398 K. The final optimized yield to C8+ alkenes reaches 75% when silica/alumina and Amberlyst 70 are used.

Since GVL can be potentially produced from levulinic acid with no external hydrogen requirements, this technology allows the production of liquid hydrocarbon alkanes from lignocellulose with minimal utilization of hydrogen (i. e., hydrogen is only used during the final alkene hydro­genation step). Furthermore, the process is potentially cost-competitive with petroleum-derived technologies, since only two reactors are required, operating in series and using non-precious metalcatalysts. Finally, a CO2 gas stream is produced with high purity and at high pressures, thereby permitting effective utilization of sequestration or capture technologies to mitigate greenhouse gas emissions.