Lignocellulosic biomass can be treated under inert atmosphere at tempera­tures of 648-800 K in a process called pyrolysis. At these conditions, solid biomass undergoes a number of processes including depolymerization, de­hydration and C-C bond breaking reactions which lead to the formation of reactive vapor species. [35] Upon subsequent cooling, the vapor products condense generating a dark viscous liquid referred to as bio-oil. This bio­oil is a complex mixture of more than 400 highly oxygenated compounds, including acids, alcohols, aldehydes, esters, ketones and aromatic species, along with some remnants of polymeric carbohydrates and lignin frag­ments. [63,64] Consequently, once separated into their components, bio­oil could serve as a source of chemicals. The final composition of the bio-oil depends on a large number of factors (e. g., feedstock type, reaction conditions, alkali content of the feedstock, storage conditions). Biooils typically contain about 25 wt% water (derived from the initial water of the feedstock and from the pyrolysis process itself), and retain up to 70% of the energy stored in the biomass feedstock, [52] thereby allowing for concentration of the energy of biomass in a dense liquid that is more easily transportable. The main advantage of pyrolysis over BTL is its simplic­ity, because it requires only a single reactor and low capital investments, thereby allowing the development of cost-effective processing units on small scale. Thus, small portable pyrolysis reactors (i. e., 50-100 tons of biomass per day) are currently commercially developed to produce liquid biofuels close to the biomass location in countries like the US, Canada and the Netherlands. [65,66]

Even though bio-oils can be used directly in simple boilers and tur­bines for heat and electricity production, their utilization as transporta­tion fuels has multiple shortcomings. The high oxygen content of bio-oils negatively affects the energy density (16-19 MJ kg-1 versus 46 MJ kg-1 of regular gasoline), and it leads to low volatility and poor stability proper­ties of the bio-oil liquid. Furthermore, the high corrosiveness (pH ~ 2.5) and viscosity of bio-oils discourage their utilization in internal combus­tion engines. Since the pyrolysis process does not involve a deep chemi­cal transformation in the feedstock, extensive oxygen removal is required for bio-oils to have hydrocarbon-like properties (e. g., high energy density, high volatility and high thermal stability), and several routes are available in this respect (Fig. 5).

Hydrodeoxygenation (i. e., treatment of the bio-oil at moderate temper­atures and high hydrogen pressures, HDO) is probably the most common method to achieve oxygen removal from biooils. [67,68] By means of this technology, bio-oil components are completely hydrogenated and oxygen is removed in the form of water, which appears in the reactor as a sepa­rate phase from the hydrocarbon layer. Hydrodeoxygenation is typically carried out over sulfided CoMo and NiMo based catalysts [68] (used in the petrochemical industry to achieve sulfur and nitrogen removal from crude oil). Precious metals such as Pt and Ru [69,70] show higher hydrogenation activities at the expense of low tolerance to sulfur impurities (typically pres­ent in bio-oils). The large amount of hydrogen required for bio-oil deoxy­genation represents the main drawback of this technology, [71] and strate­gies based on steam-reforming of the water-soluble fraction of bio-oils, [72] along with aqueous-phase reforming of biomass-derived sugars, [73,74] have been studied to avoid the need to supply hydrogen from external fossil fuel sources. Bio-oils typically contain significant amounts of lignin-derived phenols which, once transformed into aromatic hydrocarbons, are valuable gasoline components. [75] One of the challenges of the hydrodeoxygenation process is to achieve complete hydrogenation of aliphatic compounds while avoiding unnecessary hydrogen consumption in the reduction of the valu­able aromatic hydrocarbons. However, this control over the extent of the hy­drogenation process is difficult at the elevated hydrogen pressures required for hydrodeoxygenation (e. g., 100-200 bars). In addition, high pressures lead to increases in operational costs of the process.

Bio-oil deoxygenation can be alternatively carried out at milder condi­tions (e. g., 623-773 K, atmospheric pressure) and without external hydro­gen by processing the bio-liquid over acidic zeolites, in a route that resem­bles the catalytic cracking approach used in petroleum refining. [76-78] At these conditions, bio-oil components undergo a number of reactions in­volving dehydration, cracking and aromatization, and oxygen is removed in the form of CO, CO2 and water (Fig. 5). As a result, bio-oil is converted into a mixture of aliphatic and aromatic hydrocarbons, although a large fraction of the organic carbon reacts to form solid carbonaceous depos­its denoted as coke. Thus, hydrocarbon yields are relatively modest and regeneration cycles under air (to burn off the coke) are frequent. Irrevers­ible deactivation, caused by partial de-alumination of zeolite structures at the water contents typically found in bio-oils, is another drawback of this technology, and research is needed on new acidic catalytic materials with better resistance to water. [16] On the other hand, the conditions of pres­sure and temperature at which zeolite upgrading is carried out are similar to those used in pyrolysis, thereby allowing the integration of these two processes in a single reactor, as recently demonstrated by Huber et al. [79] A third route that could help to reduce oxygen content in biooils while leaving the bio-liquid more amenable for subsequent downstream processes is catalytic ketonic decarboxylation or ketonization [80] (Fig. 5). By means of this reaction, 2 molecules of carboxylic acids are condensed into a larger ketone (2n — 1 carbon atoms) with the release of stoichiometric amounts of CO2 and water. This reaction is typically catalyzed by inorganic oxides such as CeO2, TiO2, Al2 O3 and ZrO2 at moderate temperatures (573-698 K) and atmospheric pressure. [81-84] Interestingly, ketonization achieves oxygen removal (in the form of water and CO2) while consuming carboxylic acids, the latter of which represent an important fraction of bio-oils (up to 30 wt%). [85] Moreover, these acids are hydrogen-consuming compounds, and are responsible for unwanted properties of the bio-liquids such as corrosiveness and chemical instability. Consequently, as represented in Fig. 5, a pretreat­ment of the biooil over a ketonization bed would simultaneously reduce oxygen content and acidity, thereby reducing hydrogen consumption and leaving bio-oil more amenable for subsequent hydrodeoxygenation process­ing. Even though ketonization has not been used to process real lignocel — lulosic bio-oils so far, we believe that this route has potential to upgrade bio-liquids enriched in carboxylic acids. Furthermore, ketonization can also condense typical components of bio-oils like esters, [86-88] and, unlike zeolite upgrading, this reaction can be efficiently carried out under moderate amounts of water. [89]