Levulinic acid (4-oxopentanoic acid) is a high-boiling point, water-soluble biomass-derived acid that crystallizes at room temperature. Levulinic acid contains two reactive functional groups (single bondCdouble bond; length as m-dashO and single bondCOOH) that provides, as in the case of lactic acid, a rich chemistry to this compound [37]. Levulinic acid occupies a prominent place in the selected list of biomass platform molecules [11] since it is simply and inexpensively produced from lignocellulose wastes (paper mill sludge, urban waste paper, agricultural residues) by acid dehy­dration of C6 sugars [38]. Interestingly, equimolar amounts of formic acid

are co-produced along with levulinic during the dehydration process. As will be described below, this formic acid can be used as a renewable inter­nal source of hydrogen in the conversion of levulinic acid into advanced biofuels. Dehydration of sugars to levulinic acid is typically accompanied by unwanted polymerization reactions that produce intractable black in­soluble materials denoted as humins. These humins are typically burnt in the industrial process to generate heat and electricity. Interestingly, C5 sugars such as xylose (the main component of hemicellulose which typi­cally accounts for 20-30% of lignocellulose) can also serve as a source of levulinic acid. The process is not as straightforward as in the case of hexoses since it involves previous dehydration of xylose to furfural, sub­sequent hydrogenation to furfuryl alcohol and final hydrolysis of the latter to levulinic acid [39]. However, this route would benefit from an easier de­construction of amorphous hemicellulose compared to highly recalcitrant crystalline cellulose (the main component of lignocellulose and the source of C6 sugars) leading to potential operation at milder acidic conditions at which formation of humins is more controlled.

Among the different processes that have been developed for the large — scale continuous production of levulinic acid, the most promising tech­nology has been patented by the Biofine Corporation [40] and [41]. This approach utilizes a double-reactor system that minimizes the formation of by-products and the resulting separation problems. Lignocellulose is fed to the reactor along with a solution of sulfuric acid. The by-products of the process (furfural and formic acid) are condensed and separately col­lected, while solid humins are removed from the levulinic acid solution and collected as combustible wastes. This technology allows production of levulinic acid at low cost (0.06-0.18 €/kg) [42], making this compound suitable for use as a platform molecule. There are currently several plants operating with tons of waste cellulosic materials per day, both in the U. S. [42] and in Europe [43].

As shown in Fig. 3, levulinic acid possesses great potential for the pro­duction of advanced biofuels of diverse classes. Most of the routes involve intermediate formation of y-valerolactone (GVL), a stable and water-sol­uble biomass derivative with potential to be blended with gasoline as well as to serve as a precursor of polymers and fine chemicals [44] and [45]. The reduction of levulinic acid to GVL is thus a process with interest, and several routes involving different catalysts and hydrogen sources for this reduction have been explored in recent years. The most simple route in­volves utilization of carbon-supported noble metal catalysts under hydro­gen pressure [46], [47] and [48] which achieves near quantitative yields of GVL at mild temperatures (e. g. 150 °C) with slight deactivations with time on stream. The utilization of mild temperature conditions and non-acidic supports such as carbon is crucial to direct synthesis through 4-hydroxy — pentanoic acid which subsequently undergoes highly favorable internal esterification to the five member ring GVL compound [49]. The utilization of higher temperatures and/or acidic catalysts promotes dehydration of levulinic acid to a-angelica lactone. This compound polymerizes readily over acidic surfaces leading to loss of carbon as coke and severe catalyst deactivation.

As remarked above, the formic acid co-produced in the sugar dehydra­tion process can serve as a internal source of hydrogen for the reduction of levulinic acid to GVL [13], [50], [51], [52], [53], [54] and [55]. These technologies take advantage of the easy decomposition of formic acid into CO2 and H2 to generate an in situ source of this gas within the reaction system. Interestingly, the same materials used for reduction of levulinic acid to GVL are also able to catalyze decomposition of formic acid. In this sense, excellent GVL yields (96%) have been reported by Deng and co-workers in two different works involving homogenous [13] and het­erogeneous [51] Ru-based catalysts. This ability of Ru to quantitatively convert levulinic acid into GVL via formic acid decomposition has been utilized to expand this route to hexoses thereby allowing production of GVL from biomass sugars in a one-pot process and without hydrogen re­quirements [52]. The process, however, requires additional utilization of an acidic medium which serves as a catalyst for sugar dehydration, and GVL yields obtained, limited by the sugar dehydration process, are typi­cally modest (50%).

Since industrial manufacturing of levulinic acid involves treatment of biomass with aqueous sulfuric acid, it would be interesting to find a cata­lyst that can efficiently transform aqueous sulfuric acid streams of levu — linic acid and formic acid into GVL without the need for previous and waste-producing neutralization steps. In this sense, aqueous sulfuric acid solutions of levulinic and formic acids, obtained after acid hydrolysis of

solid cellulose, can be transformed into GVL with acceptable yields over Ru/C and Ru-Re/C catalysts [53] and [56]. Importantly, formation of GVL allows the design of strategies for the recycling of most of the sulfuric acid utilized for biomass depolymerization and sugar dehydration processes. Thus, alkylphenol solvents, with superior abilities to selectively extract GVL from aqueous sulfuric acid solutions, have been recently proposed for this task [57].

As summarized in Fig. 3, GVL presents high versatility to synthesize advanced transportation biofuels of diverse classes. The most direct route involves conversion into methyltetrahydrofuran (MTHF) via hydrogena­tion to 1,4-pentanediol over metal catalysts at moderate temperatures (250 °C) and subsequent dehydration of the diol to yield the cyclic ether [42]. The process takes advantage of the natural tendency of 1,4 diols to under­go dehydration/cyclation with temperature (AG = -73 kJ/mol for 1,4-pen — tanediol at 250 °C) to afford MTHF from levulinic acid with high yields (83%). MTHF is a hydrophobic molecule which, unlike ethanol, can be blended with gasoline up to 60% (v/v) without adverse effects on engine performances or gas mileage and can be distributed by existing pipeline for hydrocarbons without water contamination. MTHF is one of the com­ponents of the so-called P-series fuels which are approved by the US DOE for use in gasoline vehicles.

One route that is gaining interest in recent years involves transforma­tion of GVL into pentanoic acid. The process involves acid-catalyzed ring opening of GVL to pentenoic acid and subsequent hydrogenation of the latter over bifunctional (metal and acid) catalysts at moderate tempera­tures and hydrogen pressures [48] and [58]. The formation of pentanoic acid achieves reduction of the oxygen content of levulinic acid thereby producing a less-reactive intermediate which is more appropriate for new upgrading strategies to larger compounds. For example, Lange and co­workers [59] have used this route to produce the so-called valeric biofuels (i. e. alkyl valerates). Valeric biofuels can be used in conventional engines without any modification since they present similar energy-density, polar­ity and volatility-ignition properties than hydrocarbon fuels. The process is flexible in that by varying the alkyl chain length the fuels can be adapted to fit in both gasoline and diesel engines. The main drawback of this tech­nology lies in the need for external alcohol source for esterification.

Alternatively, liquid hydrocarbon fuels appropriate for gasoline and diesel applications can be produced via 5-nonanone, the ketonization product of pentanoic acid (Fig. 3). Interestingly, 5-nonanone can be pro­duced in high yields (70%) from aqueous GVL over a single bed of Pd/ Nb2O5 catalyst in which niobic support catalyzes GVL ring opening and pentanoic ketonization reactions [48]. Nonanone yield can be increased to almost 90% by using a double-bed reactor configuration with Pd/Nb2O5 + Ce05Zr05O2 operating at two different temperature zones (325 and 425 °C) which allows for optimum control of reactivity [53]. As shown in Fig. 4, the C9 ketone, which is obtained in high yields stored in an organic layer that spontaneous separates from water, can be upgraded to liquid hydrocarbon fuels by means of well-known petroleum-based chemistry. For example, by consecutive cycles of hydrogenation/dehydration over a bifunctional metal-acid catalyst such as Pt/NbPO4, 5-nonanone can be converted into n-nonane which possesses excellent cetane number, lubric­ity and cloud point properties to be used as a blender agent for winter die­sel applications. Alternatively, 5-nonanol, obtained by the hydrogenation of the C9-ketone, can be dehydrated and isomerized in a single step over an USY 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 [53]. Larger hydrocarbons such as those required for diesel vehicles can be produced from the nonanone stream by means of oligomerization reactions of the previously formed C9-alkenes over an acid catalyst such as Amberlyst 70 [60]. This process allows conversion of approximately half of the mass of GVL into C18 alkenes which retain more than 90% of its energy content.

Aviation requires fuels with high energy density (to allow storage of large amounts of fuels in tanks with a size determined by aircraft design) and with extremely low cloud points (to ensure operational use at high altitude temperatures). Branched hydrocarbons in the C9-C16 range meet those requirements and, consequently, routes for the production of these compounds from biomass sources are highly valuable. 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. [23] (Fig. 3). In this process, GVL undergoes decarboxylation at elevated pressures (e. g. 36 bar) over a inexpensive silica/alumina catalysts, producing a clean gas

stream composed of butenes isomers and CO2. This gaseous stream is then passed over an acidic catalyst (H-ZSM5, Amberlyst) that achieves oligo­merization of butenes yielding a distribution of branched alkenes centered at C12 suitable to be used as jet fuels after hydrogenation. This technology presents important economic and environmental advantages: (a) minimum amounts of external hydrogen are required in the process, (b) precious metal catalysts are not required, (c) a gas stream of pure CO2 is produced at the elevated pressures, thereby permitting effective utilization of se­questration or capture technologies to mitigate greenhouse gas emissions.

H2 is the typical reagent employed for levulinic acid deoxygenation to advanced biofuels, and efforts are currently aimed to obtain this gas from biomass sources. Electricity, based on renewable wind or solar pow­er, could be alternatively used for electrochemical deoxygenation of bio­mass to fuels. Nilges and co-workers have recently developed an approach for the electrochemical conversion of levulinic acid into octane [61]. As shown in Fig. 3, the process involves the combination of electroreduction of aqueous levulinic acid into pentanoic acid and subsequent oxidative Csingle bondC coupling of the latter (e. g. Kolbe reaction) to octane. Apart from the spontaneous separation of octane from the water medium, the use of electricity gives this process a number of green characteristics includ­ing the use of mild conditions (room temperature and water solutions), the replacement of any reducing chemical agent by immaterial electrons, and the minimization of waste generation. Although this preliminary work still has to solve some issues such as the effect of potential impurities in levu — linic acid (from biomass depolymerization steps) on the electrochemical reactions and the design of more efficient electrodes for large-scale ap­plications, it represents an interesting approach for converting the unused electricity generated during overproduction cycles (typical of fluctuating electricity production profiles of wind and photovoltaics) into a storable biofuel form.

Deoxygenation of levulinic acid to advanced biofuels can be carried out without utilization of hydrogen by means of a pure thermal pyrolysis treatment. The process is denoted as thermal deoxygenation (TDO) and involves previous formation of calcium levulinate and heating of the latter to temperatures ranging 350-450 °C under inert atmosphere [62] (Fig. 3). At these conditions, calcium levulinate simultaneously con­denses (by ketonization) and deoxygenates (by internal cyclation and dehydration) leading to the production of a broad product distribution of cyclics and aromatics with very low oxygen content. Production of aromatics, which are valuable components of gasoline and jet fuels, can be maximized by operating at higher temperatures. The process has been recently improved by addition of equimolar amounts of calcium formate which serves as an in situ hydrogen source allowing a deeper TDO pro­cess leading to the formation of a petroleum-like oil which could be further processed in existing refinery facilities [63].

7.4 CONCLUSIONS

Our society is highly dependent on fossil fuels, which are non-renewable and contribute to global warming. The conversion of biomass into fuels for the transportation sector can help to partially alleviate this reliance. Biodiesel and bioethanol, the main biofuels used today, present serious compatibility issues which can be overcome by the production of ad­vanced biofuels such as higher alcohols and green hydrocarbons which are fully compatible with our existing hydrocarbons-based transportation infrastructure.

However, working with complex biomass feedstocks is difficult and ap­proaches based on the formation of simpler and more stable intermediate derivatives, denoted as platform molecules, have been shown to be effec­tive for efficient biomass conversion to fuels and chemicals. Lactic acid and levulinic acid are two of these relevant biomass derivatives that can be trans­formed into advanced biofuels by a number of catalytic routes involving deoxygenation reactions combined with Csingle bondC coupling processes. The present paper offers a state of the art overview of the most relevant cata­lytic strategies available today for this paradigmatic conversion.