Upgrading of lignin-derived bio-oils by catalytic hydrodeoxygenation

Abstract

The incentive for use of renewable resources to replace fossil sources is motivating extensive research on new and alternative fuels derived from biomass. Bio-oils derived from cellulosic biomass offer the prospect of becoming a major feedstock for production of fuels and chemicals, and lignin is a plentiful, underutilized component of cellulosic biomass. Lignin conversion requires depolymerization and removal of oxygen. Likely processes for lignin conversion involve depolymerization (e.g., by pyrolysis) and catalytic upgrading of the resultant bio-oils. A major goal of the upgrading is catalytic hydrodeoxygenation (HDO), which involves reactions with hydrogen that produce hydrocarbons and water. The aim of this review is to present a critical introduction to HDO chemistry focused on compounds derived from lignin, including a summary of HDO reactions and those that accompany them, with a comparison of catalysts addressing their activities, selectivities, and stabilities. The reactions are evaluated in terms of reaction pathways of compounds representative of lignin-derived bio-oils, including anisole, guaiacol, and phenol. The review includes recommendations for further research and an attempt to place HDO in a context of options for renewable fuels and chemicals, but it does not provide an economic assessment.

---

Majid Saidi

Majid Saidi is working toward his PhD at Shiraz University in the research group of Professor Mohammad Reza Rahimpour. He holds a Bachelor of Science in HSE Engineering (2008) from the Petroleum University of Technology, Iran, and a Masters of Science in Chemical engineering (2011) from the University of Tehran. His undergraduate thesis was about the effects of heavy metals on biological wastewater treatment, and his graduate thesis is about optimization of continuous catalytic regeneration using a bee colony algorithm. His research interests include process modeling and optimization, chemical reaction engineering, catalytic upgrading of bio-oils, and environmental science. He wrote the book Heavy Metals Effects on Biological Wastewater Treatment.

Fereshteh Samimi

Fereshteh Samimi is a PhD student of the Department of Chemical Engineering, School of Chemical and Petroleum Engineering, Shiraz University. She received her Bachelor of Science in Chemical Engineering in 2011, working on the crystallization of calcium oxalate under the influence of potassium citrate. Her current research focuses on the development of ultraclean fuels, renewable energy, and process engineering; she is a member of Professor Mohammad Reza Rahimpour's research group.

Dornaz Karimipourfard

Dornaz Karimipourfard is a PhD student of the Department of Chemical Engineering, School of Chemical and Petroleum Engineering, Shiraz University. She completed her Bachelor of Science in Chemical Engineering in 2011, working on an investigation of various methods of soil decontamination. She now works on synthesis gas production via reforming reactions in Professor Mohammad Reza Rahimpour's research group.

Tarit Nimmanwudipong

Tarit Nimmanwudipong is a postdoctoral associate in the Department of Chemical Engineering at the Massachusetts Institute of Technology, working in the research group of Professor Yuriy Román-Leshkov. Working under the mentorship of Bruce C. Gates and David E. Block, he obtained his M.S. and Ph.D. degrees in chemical engineering from the University of California, Davis (2012), focusing on heterogeneous catalysis with a designated emphasis in biotechnology. He investigated reaction networks of compounds derived from lignocellulosic biomass and bio-oils upgrading processes. He obtained his B. Eng. (2006) from Chulalongkorn University, Thailand, and received a Fulbright Open Competition Scholarship (2008) to pursue his graduate studies. His research interests include renewable energy, biomass conversion, catalytic materials synthesis, and chemical reaction engineering.

Bruce C. Gates

Bruce C. Gates teaches in the Department of Chemical Engineering and Materials Science at the University of California, Davis. His research group is active in catalysis, with a focus on supported metal complex and metal cluster catalysts that are essentially molecular and lend themselves to design and in-depth characterization by spectroscopy and atomic-resolution microscopy. The group also works on catalysts for biomass conversion. Gates is a co-author of the textbook Chemistry of Catalytic Processes and author of Catalytic Chemistry, and he edits Advances in Catalysis. He serves on DOE's Basic Energy Sciences Advisory Committee and the North American Catalysis Society's Board of Directors.

Mohammad Reza Rahimpour

Mohammad Reza Rahimpour is a professor in the Department of Chemical Engineering at Shiraz University. He has spent sabbatical leaves at the University of Sydney and the University of Newcastle (Australia). His research group is active in fuel processing technology, with a focus on catalytic conversion of natural gas, petroleum, and bio-oils to fuels. His group also works on catalysts for chemical looping. He has authored more than 200 peer-reviewed papers and more than 100 conference presentations. He is a visiting Professor in the Department of Chemical Engineering and Materials Science at the University of California, Davis, working in the group of Professor Bruce C. Gates, where they are engaged in research on the catalytic conversion of compounds found in bio-oils derived from lignin.

---

Broader context

The goal of replacing fossil fuels has focused attention on lignocellulosic biomass as a renewable source—and especially on lignin, a major, underutilized component of biomass. Lignin is highly aromatic, with a polymeric structure characterized by ether linkages and hydroxy and methoxy groups. Prospective processes for converting lignin to fuels involve pyrolysis to produce bio-oils and subsequent refining to give fuels. The bio-oils themselves are poor fuels; their high oxygen contents make them poorly compatible with today's infrastructure for hydrocarbon fuels. Consequently, extensive effort has been devoted recently to methods for converting lignin-derived bio-oils to remove the oxygen, typically as water, methanol, CO, or CO₂. The best conversions in prospect are catalytic, with H₂ as a co-reactant. Catalysts such as noble metals (e.g., platinum) and metal sulfides (e.g., molybdenum sulfide) catalyze the reactions of compounds in bio-oils (typified by phenol, anisole, and guaiacol) to give water and hydrocarbons. However, numerous reactions occur, not all of them leading to oxygen removal, and much work remains to identify optimal catalysts that are active, stable, and selective for producing good fuel components without consuming too much expensive H₂. The available data provide a good starting point for future research by indicating what catalyst components and what reaction conditions favor the production of desired products, such as hydrocarbons, but extensive work is needed.

Introduction

To begin to meet the needs for renewable fuels, many researchers have turned their attention recently to a massive resource that today finds few applications: lignin.¹ Fast pyrolysis of biomass, a process that is relatively well developed, gives products called bio-oils.² Bio-oils derived from lignin are poor fuels, being unstable and having undesirable physical properties, and so they require upgrading. Fig. 1 illustrates a candidate biomass-to-biofuel conversion cycle. A central goal of the upgrading is to convert the oxygen-rich, high-molecular-weight lignin into hydrocarbons that are compatible with today's petroleum-derived fuels. Thus, a potentially valuable processing goal is to convert lignin to bio-oils and to subject the bio-oils to hydrodeoxygenation (HDO), a chemical conversion that takes place at high H₂ partial pressures (as high as 100–200 bar) and high temperatures (570–670 K) to remove oxygen primarily in the form of water.

Fig. 1 Renewable fuels from biomass: simplified representation of conversion cycle.

HDO of bio-oils gives hydrocarbons and oxygen-containing organic compounds in addition to water, and the classes of reactions include decarboxylation, hydrogenation, hydrogenolysis, hydrocracking, and dehydration (Fig. 2). Oxygen-containing products, in addition to water, include CO₂, formed in decarboxylation reactions, CO, and methanol. Like bio-oils themselves, the organic products of HDO of bio-oils still include oxygen-containing compounds unlike the hydrocarbons in fossil fuels, and researchers are working both on producing hydrocarbon products from bio-oil conversion and on opportunities for blending bio-oils that contain oxygen-containing compounds with fossil fuels for use in an infrastructure like today's.

Fig. 2 Reactions that may take place in a hydrodeoxygenation process.

Here we review recent work on the catalytic HDO of bio-oils derived from lignin, first presenting short summaries of the biomass feedstocks and their conversion into bio-oils. Interest in this subject has been growing rapidly, as shown (Fig. 3) by the frequency of publications and the number of reviews that have appeared. Consequently, we summarize the earlier work only briefly and focus instead on recent work. In the following sections, we summarize information about biomass, lignin-derived bio-oils, and processes for bio-oils upgrading, then focusing in more detail on the reactions and reaction networks characterizing HDO of compounds in bio-oils and on potential bio-oils HDO processes. We conclude with a statement of research opportunities.

Fig. 3 Number of papers published on the subject of bio-oils upgrading by year.

Biomass and lignin

Biomass offers the prospect of playing a key role as a renewable energy resource, because it can be converted into products to replace fossil fuels directly, for example, biogas and synthetic gas instead of natural gas; liquid biofuels such as biodiesel and bioethanol instead of diesel and gasoline (petrol); and solids such as pellets and briquettes instead of coal.³ Lignocellulosic biomass is the most abundant source for candidate biofuels. It consists (Fig. 4) of hemicellulose (typically 25–35% of the total), cellulose (40–50%), and lignin (15–20%), with the composition depending on the plant species.^4–7 Cellulose is a linear homopolymer of glucose units, and hemicellulose is a mixture of various polymerized monosaccharides including glucose, mannose, galactose, xylose, arabinose, 4-O-methyl glucuronic acid, and galacturonic acid residues.⁵ Lignin consists of aromatic units linked in large polymeric structures. In nature, lignin acts as an essential resin to strengthen lignocellulose matrices by filling the spaces between cellulose and hemi-cellulose, lending structural integrity to plants.^7,8

Fig. 4 Typical composition of lignocelluloses.⁵

Our focus here is on lignin because it is an underutilized resource available in massive quantities. Lignin residue materials are available from biomass-to-ethanol processes and other biorefineries. Other sources of lignin are agricultural products, municipal wastes, and byproducts of pulp and paper processing.^9–11 The enormous quantities and potential economic benefit of lignin processing provide a compelling motivation for further work on its conversion. However, the history of research on lignin conversion is marked with little success in terms of technology, and today lignin is used only in industries such as those producing low-grade fuels and those generating electricity via direct combustion.¹² If efficient, economical methods for lignin conversion could be developed, lignin could become a major source of renewable energy.^8,13

Lignin is markedly different in structure and composition from hemicellulose and cellulose, being highly aromatic and containing less oxygen. Of the biomass components, lignin is the one most similar to petroleum. Two important criteria for evaluation of biomass in our context are the hydrogen-to-carbon (H/C) and oxygen-to-carbon (O/C) atomic ratios. For comparison, such H/C ratios of crude oil are typically between about 1.60 and 2.10 and the O/C ratios between 0 and 0.03. In contrast, wood typically has O/C and H/C ratios greater than 0.61 and 1.4, respectively. In comparison with wood, lignin has lower O/C and H/C ratios, typically about 0.32 < O/C < 0.46 and 1.1 < H/C < 1.3. The low oxygen content relative to that of wood makes lignin a potentially attractive source of fuels.^14,15

The structure of lignin is complex and not fully resolved; it is known to be amorphous and polyaromatic and to incorporate numerous ether linkages, OH, and methoxy groups. Lignin is regarded as a cross-linked macromolecule composed of three types of monolignol, including p-coumaryl alcohol, coniferyl alcohol, and synapyl alcohol, with the proportions depending on the source. Lignin from softwoods is mostly made up of coniferyl alcohol-derived components, but lignin from hardwoods consists of mixtures of coniferyl- and syringyl-derived structures.

In fast pyrolysis, lignin is converted substantially into compounds such as phenol, anisole, guaiacol, cresol, syringol, etc., as summarized below.^1,2,16,17

Biomass conversion processes

Most of the reported work on biomass conversion to fuels has focused on feedstocks other than lignin; much effort has been devoted to triglycerides and to cellulose and hemicellulose and the sugars formed from them by hydrolysis.¹ Crops such as corn, sugar cane, soybeans and palm oil and rapeseed oil are among the most widely investigated feedstocks. Processes in operation today produce bioethanol and biodiesel from food-grade biomass.^16,17 But there are limitations of this technology related to the increasing need for food and concerns about the relatively low overall energy efficiencies of the processes.^14,15,18,19 Therefore, in recent years, researchers have focused on developing new methods of producing bio-oils from sources such as agricultural wastes, wood, and plants such as switchgrass and miscanthus that do not compete much with food production and do not consume too much water.

The potential processes that may emerge from research on lignocellulosic feedstocks are at a relatively early stage of development and not yet economical, but they offer tantalizing potential based on the expectation that research will lead to increasing efficiencies for the production of feedstocks that can be converted into fuels that are economically competitive and compatible with the existing infrastructure.^16,17 It is highly likely that any economic biomass conversion processes will involve catalysis—biocatalysis and/or catalysis by solids for conversion of fluid products formed from biomass conversion processes such as pyrolysis, liquefaction, catalytic pyrolysis, or catalytic liquefaction. The products of such processes include light gases and solids such as chars in addition to the liquids (bio-oils) that are the focus here.

Catalytic upgrading of bio-oils

The bio-oils that can be formed from lignocellulosic biomass require massive chemical transformations to be useful as fuels in today's petroleum-based infrastructure. Bio-oils derived from lignin have such high oxygen contents that oxygen removal is an essential processing goal.²⁰ Thus there is a rapidly growing literature related to the catalytic removal of oxygen from bio-oils, and especially from compounds in bio-oils. Removal of oxygen from bio-oils by reactions with H₂ that form water and deoxygenated compounds are called HDO reactions. There are numerous reports of catalytic HDO of lignin-derived bio-oils and compounds found in bio-oils, including numerous reports of the conversions of such compounds with catalysts that are not selective for HDO.^21–23 A premise underlying this review is that HDO may provide economical routes for the production of fuels from bio-oils, and we emphasize that this viewpoint reflects a prospect rather than a technological reality.

Catalytic HDO is superficially similar to the well-established catalytic hydroprocessing carried out in petroleum refining—hydrodesulfurization (HDS) and hydrodenitrogenation (HDN)—but in petroleum refining, these give liquid products which are single phases, largely hydrocarbons, and are well suited to further processing or direct application as fuels.

There is already a substantial literature of catalytic HDO and related topics, much of it motivated by early work in coal conversion to liquid fuels. Little of this work has been focused on petroleum, because there is only little oxygen in petroleum (typically less than a few tenths of a weight percent), and the oxygen is readily removed in processes that have the goals of HDS and HSN.

In 2000, the literature of kinetics and reaction networks of HDO was reviewed by Furimsky.²⁴ In 2004, Czernik and Bridgwater⁹ reviewed developments in the applications of bio-oils, and in 2006 Mohan et al.²⁵ discussed pyrolysis of wood/biomass for bio-oils. A 2007 review by Elliott²⁶ gives a historical perspective on developments in catalytic hydroprocessing of bio-oils. In a 2007 review, Zhang et al.²⁷ focused on the properties of compounds and compositions of biomass pyrolysis oil. In 2010, Venderbosch and Prins²⁸ and Oasmaa et al.²⁹ reviewed processes for fast pyrolysis, and in 2011 Mortensen et al.³⁰ considered catalytic upgrading of bio-oils to produce engine fuels, and Choudhary and Phillips³¹ reviewed applications of catalytic HDO from an industrial perspective. In 2012, Bu et al.³² considered catalytic HDO of lignin-derived phenols. In 2013, He and Wang³³ considered recent advances in HDO of pyrolysis bio-oils with various catalysts, and Laskar et al.³⁴ reviewed catalytic processing of lignin derivatives to give hydrocarbon fuels. Furimsky³⁵ has just contributed a review of hydroprocessing catalysis with a focus on biomass conversion.

Much of the early work on HDO of whole bio-oils emerged from the group of Elliott,³⁶ who worked primarily with hydroprocessing catalysts of the types used to remove sulfur and nitrogen from petroleum. Elliott's work demonstrated the need for severe reaction conditions for bio-oils hydroprocessing relative to those for petroleum hydroprocessing; these researchers encountered severe reactor plugging resulting from carbonaceous deposits on the catalysts. Their experiments indicated that a stabilization step would be of value preceding catalytic hydroprocessing.^37–40 Gagnon and Kaliaguine⁴¹ investigated the effect of a hydrogenation pretreatment, and Sheu et al.⁴² suggested models based on an empirical function of temperature and pressure for overall removal of oxygen and compositional changes during hydroprocessing. In the early 1990s, Oasmaa and Boocock⁴³ reported the reduction of the oxygen content of peat pyrolysate oils from 22 to 3 wt% resulting from catalytic HDO.

The early work demonstrated that hydroprocessing of bio-oils was feasible, although not economical, requiring severe reaction conditions and leaving open the question of whether economical processes were likely to emerge. The early work largely ignored issues of fundamental understanding of the chemistry of bio-oils conversion, identification of active, selective, and stable catalysts, and a foundation for estimation of process economics.

The recent emphasis on renewable fuels has focused interest on these issues, and the field is moving forward much faster than before, although extensive work remains before economical bio-oils processing is likely to be achieved. The review presented here addresses this topic. It is not comprehensive, and the reader is referred to the literature cited in the preceding paragraphs in this section for more foundational information. We consider here only reactions catalyzed by solids, although there are early results suggesting the possible benefits of soluble transition metal complex catalysts for biomass conversion.⁴⁴

Lignin- and other biomass-derived bio-oils

Increasing attention is being paid to catalytic upgrading of lignin-derived bio-oils for production of aromatic compounds such as phenolics as well as for the production of hydrocarbons that could be used as fuels. The high concentrations of oxygen in lignin-derived bio-oils (typically, 35–40 wt%) make their conversion into fuels a processing challenge that is markedly different from those of petroleum refining.^8,9,22,45

Processes for conversion of lignin to phenolic compounds and biofuels are based on depolymerization reactions, which may be pyrolitic and (in part) catalytic, and several investigations have provided information about potential commercial processes.⁴⁶ Pyrolysis involves anaerobic thermal decomposition for degrading of biomass into mixtures of small molecules, which can be used to form chemicals and fuels. At the high temperatures of pyrolysis (650–800 K), vapor-phase products react and condense to form liquid mixtures. To obtain relatively high liquid product yields and to decrease char formation, high temperatures and short reactor residence times are favorable. Lignin degradation processes such as fast pyrolysis produce oxygen-containing compounds including guaiacol (typically, 39%), syringyl (16%), and hydroxyphenyl derivatives (45%) (Fig. 5). Pyrolysis of softwood lignin mainly yields guaiacol derivatives, whereas pyrolysis of hardwood lignin produces both guaiacol and syringyl derivatives.^47–60

Fig. 5 Compounds in lignin-derived bio-oils.

Lignin can alternatively be depolymerized in an aqueous catalyst solution. Water is the solvent, but methanol, ethanol, or combinations thereof can also be used. Catalysts used in aqueous lignin depolymerization include NaOH, KOH, Ca(OH)₂, and Mg(OH)₂. Often, for preventing precipitation and reactor plugging, continuous stirred reactors are used. The reactions are typically carried out in the ranges of 500–625 K and 70–160 bar with liquid hourly space velocities in the range of 0.5 to 6 h⁻¹.

Methods of catalytic upgrading of bio-oils

Bio-oils obtained from fast pyrolysis have high oxygen contents—typically more than 10 wt% and even as much as 50 wt%. The oxygen-containing compounds are responsible for chemical and thermal instability of the liquids and a strong tendency to polymerize. Low heating values, high viscosities, the acidic nature of the oils (which causes corrosion of vessels and pipes), and the high water contents are all major disadvantages of bio-oils as potential fuels.^{48,58,61–63} These limitations make bio-oils essentially unusable for almost all of today's fuel applications.

Therefore, the (partially) depolymerized products formed by pyrolysis must be refined. The major prospective refining processes have the goal of forming stabilized deoxygenated products. Lignin-derived compounds can in prospect be converted to components that are good liquid fuels, such as high-octane-number alkylbenzenes.

Thus, in the processing step following pyrolysis, the partially depolymerized product may be subjected to a catalytic process for removal of oxygen. This step can in prospect convert compounds such as methoxyphenols and benzenediols to hydrocarbons that are valuable fuel components. The desired reactions involve formation of hydrocarbons and water in addition to other compounds that can be removed from the hydrocarbons; the reactions typically involve H₂ as a reactant and are classified as HDO reactions.

Catalysts that are active for these conversions include noble metals and metal sulfides supported on metal oxides or carbon. The reactions typically take place at temperatures exceeding 475 K and at H₂ partial pressures from 35–140 bar.^64–66 For example, in the conversion of lignin to naphthenic kerosene, the HDO reaction temperature is about 475 to 675 K, and suitable catalysts are alumina-supported sulfides of molybdenum or tungsten, among others.^67–70

Much of the research on catalytic HDO is based on the premise that the products should be usable in today's infrastructure built for hydrocarbon fuels. The lignin-derived products might be blended with petroleum-derived products, and they might also be refined as co-feeds with them. Another practical goal is to minimize the yields of light gases, char, and coke formed in the processing.⁷¹

The most commonly investigated methods for bio-oils upgrading (Fig. 6) involve physical, chemical, and catalytic processing. Physical upgrading of bio-oils involves filtration, emulsification, and solvent extraction. In emulsification and solvent extraction, the bio-oils are mixed with diesel oil and solvents, respectively, to extract the compounds with lower oxygen contents. Chemical upgrading includes aqueous-phase processing such as reforming and gasification, mild cracking, esterification, etc.

Fig. 6 Simplified schematic flow diagram for candidate process for conversion of biomass to bio-fuels.

Catalytic upgrading routes that have been investigated include steam reforming, zeolite-catalyzed cracking, and HDO.^63,72 Catalytic steam reforming of pyrolysis oils is a recently explored method to produce H₂. Cracking is related to the “fluid catalytic cracking” of petroleum fractions (which takes place with zeolite-containing catalysts in fluidized-bed reactors) and leads to the removal of oxygen; the process is operated at atmospheric pressure and high temperatures (575–875 K); no H₂ is required, which is advantageous for processing costs.^58,73 However, a major disadvantage of cracking is the high yields of carbonaceous deposits (coke) formed on the catalysts, which deactivate them within seconds and require regeneration by burning of the deposits.⁷⁴ Zeolite-catalyzed cracking is not nearly as well developed as HDO in our context.⁴⁸ HDO is considered in detail in the following sections.

Yan et al.^75,76 investigated the degradation of lignin catalyzed by noble metals in a two-step process. The first step included the catalytic cleavage of C–O–C bonds without breaking of C–C bonds under relatively mild conditions, and the second step was hydrogenation of species to remove the aromaticity followed by hydrogenolysis of the C–O bonds to produce compounds classified as C₉ and C_14–18. The transformation of the phenolic compounds into desirable fuels in the second step is challenging.

HDO with conventional metal sulfide catalysts is problematic because of the rapid catalyst deactivation by coke, water, and other components. Catalytic reactions in the presence of aqueous-phase reactants offer prospects for the conversion of phenolic compounds that may help to overcome some of the limitations of more conventional catalytic hydroprocessing, but high operating temperatures are needed. Recently ionic liquids (ILs) have been proposed as promising reaction media (and catalysts) for biofuel production, and these offer some of the advantages of the water-containing reactants. Yan et al.⁷⁶ examined the HDO of lignin-derived phenols into alkanes using nanoparticle catalysts dispersed in Brønsted acidic ILs. Their results indicate that in the HDO of phenol, high conversions with excellent selectivity to cyclohexane could be obtained with [bmim][BF₄] or [bmim][TF₂N]. They concluded that in comparison with metal sulfide or mineral acid/supported metal catalysts in water, upgrading of the lignin derivatives to alkanes catalyzed by nanoparticles combined with Brønsted acidic ILs is an efficient and less energy consuming process.

Jongerius et al.⁷⁷ considered a process in which lignin was depolymerized to form monomers, dimers, and oligomers by liquid-phase reforming catalyzed by Pt/γ-Al₂O₃ followed by HDO of the extracted lignin-derived bio-oils catalyzed by CoMo/Al₂O₃ and Mo₂C/CNF. The authors carried out the HDO at 573–648 K and a high H₂ partial pressure (55 bar), observing that the latter catalyst outperformed the former and yielded more oxygen-free products such as benzene and toluene.

HDO reactions and those accompanying them

Because most of the reported work on catalytic HDO has been carried out with solid catalysts and under conditions resembling those of conventional catalytic hydroprocessing, our review is focused on such work. Thus, in the following sections, the reactions, catalysts, and potential processes for bio-oils HDO are considered, and results obtained with prototypical catalysts and individual compounds that are representative of lignin-derived bio-oils provide a foundation for the discussion. For example, phenol itself has been used often as a representative of phenolic compounds. Experiments with phenol and related compounds have been carried out to determine the reactions taking place and to elucidate what catalyst functions are engaged in these conversions. The approach mirrors that applied in years past to help elucidate the hydroprocessing reactions involved in the conversion of petroleum and of coal.⁷⁴ We believe that this approach offers the prospect of being similarly valuable for improving bio-oils upgrading process.

Reaction networks

Several reaction networks have been proposed for HDO and the reactions that accompany HDO. Understanding of these reaction networks has emerged from investigations of various individual compounds as reactants—one can refer to them as model compounds—with various catalysts.

More than one pathway has been inferred for oxygen removal by reactions with H₂ from compounds such as guaiacol, a compound that a number of researchers have considered to be a prototype. One pathway is direct hydrodeoxygenation (direct deoxygenation, DDO), which leads to the formation of aromatics and water. Another is hydrogenation of the phenolic ring followed by dehydration to form cyclohexene derivatives and rehydrogenation to produce cyclohexane derivatives.^73,78–80 Furthermore, removal of methoxy groups from phenolic compounds can proceed by demethylation (DME) and by demethoxylation (DMO), which involve C–O bond breaking and respectively produce CH₄ and methanol.⁷⁸

Details follow in this section, showing that the catalyst type and reaction conditions are important in determining the relative importance of the competing reaction pathways.⁷³ We illustrate with results determined in investigations of three compounds, phenol, anisole (methoxybenzene), and guaiacol (2-methoxyphenol) reacting with H₂. These compounds were chosen to illustrate the main characteristics of the reaction networks because each of them is abundant in lignin-derived bio-oils, each is regarded as a good model compound representing lignin-derived bio-oils, and the data characterizing these compounds are more fundamental and complete than those characterizing the reactions of other candidate compounds to represent the bio-oils.

Because understanding of the mechanisms (details of the elementary steps of bond breaking and bond forming) of catalytic HDO reactions is still not well developed, we do not address the mechanisms here.

Conversion of phenol

Phenol is the simplest of the widely investigated compounds representative of bio-oils, incorporating the aromatic ring and hydroxyl group that are typical of lignin and lignin-derived compounds. Here we summarize the reactions characterizing HDO of phenol, with an emphasis on reactions catalyzed by metal sulfides (which have been used commonly) and noble metals. Regardless of which of these catalyst types is used, the major routes for HDO of phenol are hydrogenolysis and hydrogenation.^81–84

Fig. 7 illustrates the conversion of gas-phase phenol in the presence of a metal sulfide catalyst taking place by two pathways.^{49,79,85–87} Pathway (1) involves direct hydrogenolysis to give benzene and water by removal of a hydroxyl group. Subsequent hydrogenation gives cyclohexene, which can be hydrogenated to cyclohexane. Alternatively, in Pathway (2), phenol is converted to cyclohexanone, which is hydrogenated to form cyclohexanol. In most reported experiments characterizing these reactions, conversions have been high and the intermediate cyclohexanol has not been observed. Pathway (1) is typically dominant and is markedly favored at higher temperatures. Mass balance data indicate the hydrogen consumptions.^{62,72,79,80,82,87–89}

Fig. 7 Reaction network characterizing phenol conversion catalyzed by metal sulfides and noble metals.^{50,79,85–88}

When the metal sulfide catalysts are replaced by noble metals on supports, similar chemistry is observed, but the hydrogenation activity is typically higher. For example, Senol et al.⁶² concluded that as a consequence of the high hydrogenation activity of a NiMo catalyst, the formation of cyclohexane from phenol was predominant. In contrast, in HDO of phenol catalyzed by a La–Ni–W–B catalyst, reaction Pathway (1) was not observed, and hydrogenation of phenol by the second pathway was dominant.^62,85,86

Conversion of anisole

Anisole is representative of bio-oil compounds that incorporate C–O–C moieties—which are prevalent in lignin as well as lignin-derived bio-oils. In the HDO of anisole, four reaction pathways—demethylation, hydrogenation, hydrogenolysis, and methyl group transfer (transalkylation)—occur, not all of them leading to the removal of oxygen. Because the C_methyl–O bond is weaker than the C_aromatic–O bond, demethylation via the rupture of the C_methyl–O bond is typically favored kinetically.⁹⁰

Experiments with various catalysts (details are given below) have led to the linking of the reaction pathways with specific catalyst functions. A noble metal such as platinum catalyzes demethylation, HDO, and hydrogenation in sequence, giving phenol. Phenol, consistent with results summarized above, is converted to benzene by breaking of the C_aromatic–O bond and to cyclohexane by hydrogenation of the aromatic ring, which is followed by hydrogenolysis of the C_alicylic–O bond.^62,73,90 When the solid catalyst incorporates an acidic function (such as γ-Al₂O₃ as the support), methyl group transfer reaction are kinetically significant.⁷³

Fig. 8 is a scheme representing the reactions of anisole conversion that is consistent with the results of numerous researchers.^{52,58,73,90,91} Pathway (1) is demethylation, which takes place on the metal surface. Another reaction (Pathway 2) involves direct deoxygenation, leading to methanol and benzene.⁵⁸ On catalysts that incorporate both metal and acid functions, in addition to Pathway (1), methyl group transfer and HDO are catalyzed by acidic sites via Pathway (3). Thus, the combination of a noble metal and an acidic support gives a higher catalytic activity than the metal on a nonacidic support, but not all the reactions involve oxygen removal.⁷³

Fig. 8 Reaction network characterizing anisole conversion catalyzed by acidic and non-acidic supported metal catalysts.^{52,58,73,90,91}

Conversion of guaiacol

Guaiacol appears to be the consensus preferred choice as a prototype compound to represent lignin-derived bio-oils—because it incorporates two types of C–O bond that represent both lignin and many of its derivatives. The moieties are hydroxy (C_sp2OH) and methoxy (C_sp2OCH₃). Reactions of guaiacol have been used to resolve the roles of catalyst functions.⁹² Early experiments indicated consecutive transformations to produce catechol, methylated products, and deoxygenated compounds such as benzene and cyclohexane.⁷⁸ Demethylation (or methyl group transfer) occurs by two pathways. One is hydrogenolysis of the O–CH₃ bond, which leads to the formation of catechol and methane, and the other is demethoxylation, which occurs by cleavage of the C_aromatic–OCH₃ bond, leading to phenol and methanol. In the conversion of the phenolic OH group, one pathway includes hydrogenolysis of the C_aromatic–OH bond and one includes hydrogenation of the aromatic ring.⁶¹

The reaction pathways depend substantially on the catalyst composition. For example, the HDO of guaiacol catalyzed by sulfided CoMo and NiMo begins with demethylation/demethoxylation and dehydroxylation as the primary and kinetically significant reactions, followed by hydrogenation of the benzene ring. But when the catalysts are supported rhodium or supported palladium, hydrogenation of the benzene ring is followed by demethylation/demethoxylation.^{62,74,81,85,87,93–99}

Because the acidic sites on catalyst surfaces are active for deoxygenation reactions, the deoxygenation activity is enhanced when more acidic sites are introduced. For example, when the conversion is catalyzed by noble metals supported on nitric-acid-treated carbon black, 2-methoxycyclohexanol is formed selectively by hydrogenation of the benzene ring, but when the support is SiO₂–Al₂O₃, which incorporates acidic sites, the major product is cyclohexane.⁹⁵

To elucidate more precisely and quantitatively which catalyst functions catalyze the various reactions, Nimmanwudipong et al.⁵⁴ used a supported platinum catalyst, Pt/γ-Al₂O₃, under mild enough conditions to provide a detailed characterization of intermediate products. They showed that deoxygenation by removal of the methoxy group is more rapid than hydroxy group removal under their conditions. The main products of the reactions were phenol, catechol, and 3-methylcatechol. When the support was changed from the acidic γ-Al₂O₃ to the basic MgO, the products formed via transalkylation reactions (on acidic sites) were not formed in kinetically significant amounts.¹⁰⁰ Phenol and catechol were then the major products, along with the unexpected product cyclopentanone.¹⁰⁰

Results of other authors confirm that the acidic sites catalyze transalkylation and metal and metal sulfide sites catalyze the other classes of reactions.⁷³

Summary reaction network synthesizing data from a number of investigations

A figure summarizing the reaction pathways occurring in the conversion catalyzed by platinum supported on γ-Al₂O₃ (Fig. 9) is representative of the results presented above for guaiacol, anisole, phenol, and other reaction intermediates that were mentioned (such as cyclohexanone and methylanisole), and it is based on data obtained in experiments in which guaiacol, anisole, and phenol each was used as a feed (with H₂); data were also obtained when some of the intermediates were used as feeds. At the risk of overgeneralizing, we suggest that this reaction network is qualitatively representative of observations with most supported metal catalysts^{54,55,95,97,101,102} and even most supported metal sulfide catalysts; details of the performance of various catalyst types are given in a following section. The summary shown in Fig. 9 includes the several documented reaction classes, each of which is coded with a different type of arrow in the figure.¹⁰³

Fig. 9 Reaction network for the conversion of lignin-derived compounds (each compound shown in red was used as a reactant) with H₂ catalyzed by Pt/γ-Al₂O₃ at 573 K and 1.40 bar. HDO, hydrogenolysis, and hydrogenation (or dehydrogenation) reactions are represented by dashed green, blue, and black arrows, respectively. Transalkylation reactions are represented by solid black arrows.¹⁰³

Comparison of catalysts

Research has been done with various HDO catalysts, both in attempts to elucidate the roles of various catalyst functions and to seek improved catalysts—those with good combinations of activity, selectivity, and resistance to deactivation.

The catalytic species that accelerate the HDO reactions include metals, metal sulfides, metal phosphides, metal carbides, and metal nitrides. These have been used on various supports, and in general the acidic supports catalyze methyl group transfers and also lead to catalyst deactivation by carbonaceous deposits (coke) that form on their surfaces. Noble metals such as platinum offer high activities for HDO reactions, but they are expensive.^{48,97,104–108} Furthermore, these metals are highly active for hydrogenation of aromatic rings, and consequently catalysts incorporating them are characterized by high H₂ consumptions,¹⁰⁹ which is an economic disadvantage and raises the question of the availability and cost of the H₂. Consequently, many investigators have worked with supported non-noble transition metals which are less expensive than noble metals.^90,101,110

Various supports have been investigated, notably those that offer prospects of being resistant to coke formation, including carbon, ZrO₂, SiO₂, and MgO,^{62,68,111–113} and these are discussed in a following section.

In the next paragraphs, the performance of various catalysts for conversion of lignin-derived compounds is summarized; details are given in Tables 1–4.

Table 1 HDO catalyzed by non-noble transition metals (T is temperature; P is pressure; t is time; WHSV is weight hourly space velocity)

Catalyst	Support	Reactor configuration	Operating conditions				Lignin-derived compound	Major products	Ref.
			T (K)	P (bar)	t (h)	WHSV (h^−1)
Fe	SiO2	Fixed bed	623–723	1	0.5	0.67–9.09	GUA	Ben, Tol	114 and 115
Co	Kieselguhr	Fixed bed	623–723	1	—	1.22–2	GUA	—	114
CoMo	γ-Al2O3	Batch	613 or 628	70	—	—	4MP	Tol, MCYHA	89
CoMo	γ-Al2O3	Flow reactor	523	15	—	—	Ph	Ben	62
CoMo	γ-Al2O3	Batch	523	75	2	—	Ph	—	62
CoMo	Al2O3	Fixed bed	613	70	—	—	2EP	—	141
CoMo	Al2O3	Batch	573	50	4.17	—	4MP	Tol	96
CoMo	Al2O3	Batch	598	69	1.68	—	4MG	Tol, Cr, MC	152
CoMo	Al2O3	Batch	573	69	5.74	—	4MC	Tol, Cr, MCYHA	152
CoMo	Al2O3	Batch	573	69	4	—	EUG	PCYHA, PP, PG, PC	152
CoMo	Al2O3	Batch	573	69	4.24	—	Van	MCYHA, Cr, MC	152
CoMo	Al2O3	Batch	523–598	34	6.67–10	—	ANI	Ph, Ben, CYHA	142
CoMo	Al2O3	Batch	523	34	20	—	GUA	CAT, Ph, Ben, CYHA	142
CoMo	Al2O3	Flow reactor	548–623	50	—	—	o-MEP	ph, dioxygen compounds	153
CoMo	Al2O3	Flow reactor	548–623	50	—	—	m-MEP	ph, dioxygen compounds	153
CoMo	Al2O3	Batch	553	70	2.5	—	GUA	Ph. CAT	67
CoMo	Al2O3	Fixed bed	573	50	—	—	ANI	ph, 2MP, Ben	154
CoMo	Al2O3	Fixed bed	573	50	—	—	ANI	ph, 2MP, Ben	98
CoMo	Al2O3	Flow reactor	523	15	—	—	Ph	Ben	98
CoMo	Al2O3	Flow reactor	573	15	—	—	Ph	Ben, CYHA	98
CoMo	Al2O3	Flow reactor	523	15	—	—	ANI	2MP, Xol, ph, Ben	98
CoMo	Al2O3	Flow reactor	573	15	—	—	ANI	Tol, ph, Ben, 2MP	98
CoMo	Al2O3	Batch	573	55	4	—	GUA, Ph	Ben, Tol, Xy	77
CoMo	C	Batch	553	70	—	—	GUA	Ph, Ben, CYHA	155
Co–Mo–B	—	Batch	548	40	10	—	Ph	CYHA	156
Mo	Al2O3	Fixed bed	613	70	—	—	2EP	—	141
Mo	γ-Al2O3	Fixed bed	573	50	—	2	ANI	Ph, Cr, 2,6-DMP	117
MoO2	—	Batch	623	44	5	—	4MP	Tol	109
MoO3	—	Batch	623	44	5	—	4MP	Tol	109
MoO3	—	Packed-bed	673	1	1	—	CYHAONE, ANI	Ben	157
Mo2N	—	Batch	573	50	—	—	GUA	Ph	102
Mo2N	C	Batch	573	50	6	—	GUA	Ph	94
Mo2C	CNF	Batch	573–648	55	4	—	GUA	Ph	119
NiMo	Al2O3	Fixed bed	613	70	—	—	2EP	—	141
NiMo	Al2O3	Fixed bed	573	50	—	—	ANI	Ph, 2MP	154
NiMo	Al2O3	Fixed bed	573	50	—	—	ANI	Ph, 2MP, CYHA	154
NiMo	Al2O3	Batch	573	70	2.5	—	GUA	Ph, CAT	67
NiMo	γ-Al2O3	Flow reactor	523	15	—	—	Ph	CYHA	62
NiMo	γ-Al2O3	Batch	523	75	2	—	Ph	—	62
NiMo	SiO2–Al2O3	Fixed bed	573	50	—	—	Ph	C6 hydrocarbons	146
NiMo	SiO2–Al2O3	Fixed bed	598	50	—	—	Ph	C6 hydrocarbons	146
NiMo	SiO2–Al2O3	Fixed bed	598	50	—	—	2MP	Ph, C7 hydrocarbons	146
Ni	SiO2	Batch	593	170	1	—	GUA	CYHA, CYHAONE	61
Ni	SiO2	Batch	593	170	—	—	GUA	CYHA, CYHAONE	132
Ni	SiO2	Fixed bed	423	1	0–4	—	Ph	CYHAONE, CYHAOL, Ben	130
Ni	SiO2	Fixed bed	423	1	0–4	—	CYHAONE	CYHAOL, Ph, Ben	130
Ni	SiO2	Fixed bed	573	50	—	2	ANI	Ph, Ben	117
Ni	γ-Al2O3	Fixed bed	573	50	—	2	ANI	Ph, Ben	117
Ni	SBA-15	Flow reactor	563–583	3	—	20.4 and 81.6	ANI	Ben	135
	Al-SBA-15
	γ-Al2O3
	Microporous carbon
	TiO2
	CeO2
Ni	SiO2–ZrO2	Batch	573–613	50	5	—	GUA	CYHA, MCYHA, PCYHA	133
Ni	ZrO2–SiO2	Batch	523–613	50	—	—	Ph, GUA	CYHA	158
Ni–W	AC	Fixed bed	423–573	15	0–6	0.5	Ph	—	85
Ni–W–B	—	Batch	498 or 523	40	4	—	Ph	CYHAOL, CYHE, CYHA	86
Ni–W–B	—	Batch	498	40	7	—	Ph	—	122
Ni–W–B	—	Batch	498	40	—	—	Ph	—	123
Ni–Cu	Al2O3	Batch	593	170	1	—	GUA	CYHA, CYHAONE	61
Ni–Cu	SiO2	Batch	593	170	1	—	GUA	CYHA, CYHAONE	61
Ni–Cu	SiO2	Batch	593	170	—	—	GUA	CYHA, CYHAONE	132
Ni–Cu	SiO2–ZrO2	Batch	593	170	—	—	GUA	CYHA, CYHAONE	132
Ni–Cu	CeO2–ZrO2	Batch	593	170	1	—	GUA	CYHA, CYHAONE	61
Ni–Cu	CeO2–ZrO2	Batch	593	170	—	—	GUA	CYHA, CYHAONE	132
Ni–Cu	CeO2	Fixed bed	593	10	1	—	ANI	CYHA	156
Ni–Cu	ZrO2–SiO2–La2O3	Batch	593	170	1	—	GUA	CYHA, CYHAONE	61
Ni–Cu	ZrO2–SiO2–La2O3	Batch	593	170	—	—	GUA	CYHA, CYHAONE	132
Ni–Cu	δ-Al2O3	Packed-bed	573	10	4	3–6	ANI	Ben, Tol, CYHA, MCYHA	118
Ni and NiCu	ZrO2–SiO2	Batch	523–613	50	1–8	—	GUA	CYHA, MCYHA, Ben	134
Sn	Inconel	Quartz tube	673	1	3	—	GUA	Ph	58
Ga	H-Beta zeolite	Packed-bed	673–823	1	2–22	—	3MP	Tol, Ben, Xy	159
V	Al2O3	Flow reactor	623	—	—	—	GUA	Methylated phenol	160
W2C	CNF	Batch	573–648	55	4	—	GUA	Ph	119
Pd, Pt, Ru, Fe, Cu, Pd–Fe	C	Fixed bed	350 and 450	1	2	—	GUA	Ph	161
CuCr2O4·CuO	—	Batch	423–548	50	0–20	—	Ph	CYHAOL	116
							ANI	Ben, CYHA, CYHAOL, MCYHA
							Cr	MCYHA
							GUA	CYHAOL, MCYHDIOL, CYHA
							VA	2ME4MEPH

---

Table 2 HDO catalyzed by noble metals (T is temperature; P is pressure; t is time; WHSV is weight hourly space velocity)

Catalyst	Support	Reactor configuration	Operating conditions				Lignin-derived compound	Major products	Ref.
			T (K)	P (bar)	t (h)	WHSV (h^−1)
Pt	Al2O3	Fixed bed	473	40	—	—	GUA	CYHA	162
Pt	Al2O3	Fixed bed	473	40	—	—	Cr	—	162
Pt	Al2O3	Packed bed	573	1.4	—	17.6–437	ANI	Ph, 2MP, Ben	52 and 53
Pt	Al2O3	Packed bed	573	1.4	0–6.67	0–4	4MA	4MP, Tol	147
Pt	γ-Al2O3	Tubular	573	1.4	0–3	20	GUA	Ph, CAT	55
Pt	γ-Al2O3	Tubular	573	1.4	0–4.17	7.9	EUG	4-PG	56
Pt	γ-Al2O3	Packed bed	573	1.4	0–6.67	0–4	4MA	2,4,6-TMP	147
Pt	γ-Al2O3	Packed bed	523–623	1.4	0–6	56	CYHAONE	Ph	51
Pt	MZ-5	Fixed bed	473	40	—	—	GUA	—	163
Pt	MZ-5	Fixed bed	473	40	—	—	Cr	MCYHA, alkylated cycloalkanes	163
Pt	SiO2–Al2O3	Packed bed	573	1.4	—	1–4	4MA	4MP, Tol	147
Pt	SiO2	Fixed bed	673	1	0–5	2	ANI	Ben	73
Pt	AC(N)	Batch	553	40	2	—	4PP	—	163
Pt	H Beta zeolite	Fixed bed	673	1	0–5	12	ANI	Ben, Tol, Xy	73
Pt	HY zeolite	Fixed bed	473–523	40	—	5–20	Ph	CYHA	49
Pt	Inconel	Quartz tube	673	1	3	—	GUA	Ph	58
Pt–Sn	Inconel	Quartz tube	673	1	0–3	—	GUA	Ph, Ben	58
Pt–Sn	Inconel	Quartz tube	673	1	0.75, 2.08	0.32	ANI	Ph, Ben	58
Pt–Rh	ZrO2	Batch	573	80	3	—	GUA	—	47 and 164
Rh	—	Batch	400–700	50	1	—	GUA	2MECYHAOL, 2EXYHAONE	97
Rh	ZrO2	Batch	573	80	3	—	GUA	Ben	47 and 164
Rh	SiO2	Fixed bed	573	10	—	—	ANI	—	156
Ru	C	Batch	473	69	—	—	GUA	2MECYHAOL, CYHAOL	45
Ru, Mo	C	Packed bed	623–673	40	0.067	—	GUA	Ben, Ph	165
Pd	C	Fixed bed	353	5	7	—	Ph	—	50
Pd	C, HZSM-5	Batch	473	50	2	—	GUA	Cycloalkanes	166
Pd	C, HZSM-5	Batch	433	50	2	—	Ph	Cycloalkanes	166
Pd	C, H3PO4	Batch	423	50	0–3.34	—	ANI	—	87
Pd	C, H3PO4	Batch	473	50	0–3.34	—	CAT	—	87
Pd	C, H3PO4	Batch	473	50	0.5		Ph	CYHA	87
Pt, Rh, Pd, Ru	Al2O3, SiO2–Al2O3, nitric-acid-treated carbon black	Batch	523	—	2	—	GUA	—	95

---

Table 3 HDO catalyzed by metal sulfides (T is temperature; P is pressure; t is time; WHSV is weight hourly space velocity)

Catalyst	Support	Reactor configuration	Operating conditions				Lignin-derived compounds	Major products	Ref.
			T (K)	P (bar)	t (h)	WHSV (h^−1)
CoMoWS	SBA-15	Fixed bed	583	30	0.84–4.17	24.5	ANI		91
CoMoWS	SBA-16	Fixed bed	583	30	0.84–4.17	24.5	ANI	Ph, Cr, Xol	91
CoMoS	—	Fixed bed	573	40	—	—	GUA	Ben	92
CoMoS	—	Batch	673	50	1	—	GUA	Ben, Ph, CYHE, CYHA	97
CoMoS	—	Batch	553	70	—	—	GUA	—	68
CoMoS	—	Batch	573	50	3.34	—	4MP	Tol, CYHA	96
CoMoS	γ-Al2O3	Fixed bed	573	40	—	—	GUA	Ben	92
CoMoS	γ-Al2O3	Batch	523	75	0.08–5	—	Ph	Ben	88
CoMoS	γ-Al2O3	Flow reactor	523	15	—	—	Ph	Ben	62
CoMoS	γ-Al2O3	Batch	523	75	2	—	Ph	—	62
CoMoS	Al2O3	Fixed bed	573	40	—	—	GUA	—	113
CoMoS	Al2O3	Trickle bed	573	40	—	—	GUA	—	78
CoMoS	Al2O3	Batch	573	80	1–5	—	GUA	Ben	48
CoMoS	Al2O3	Batch	573	50	—	—	GUA	Ph, CAT	120
CoMoS	Al2O3	Batch	573	50	—	—	ANI	Ph	120
CoMoS	Al2O3	Fixed bed	573	28.5	—	—	Ph	Aromatics, CYHA	79
CoMoS	Al2O3	Batch	573	50	4	—	Ph	Ben	120
CoMoS	TiO2	Fixed bed	573	40	—	—	GUA	—	113
CoMoS	ZrO2	Fixed bed	573	40	—	—	GUA	—	105
CoMoS	C	Batch	553	70	2	—	GUA	—	113
CoMoS	Al2O3	Fixed bed	623		1	—	2EP, GUA	—	145
NiMoS	—	Batch	400–700	50	1	—	GUA	Ben, Ph, CYHE, CYHA	97
NiMoS	—	Parr	623	28	1	—	Ph	—	79
NiMoS	γ-Al2O3	Batch	523	75	0.08–5	—	Ph	CYHA	88
NiMoS	γ-Al2O3	Flow reactor	523	15	—	—	Ph	CYHA	69
NiMoS	γ-Al2O3	Batch	523	75	2	—	Ph	—	69
NiMoS	γ-Al2O3	Batch	723	28	1	—	CAT	Ph, Ben, CYHA	167
NiMoS	γ-Al2O3	Batch	723	28	1	—	GUA	Ben, Tol	167
NiMoS	γ-Al2O3	Batch	723	28	1	—	Syr	Ben, Tol	167
MoS2	—	Fixed bed	573	40	—	—	GUA	CYHE	92
MoS2	—	Fixed bed	573	50	—	1	ANI	Ph, Ben	117
MoS2	—	Batch	593–643	28	—	—	4MP	Tol, Ph, 2,4-Xol	81
MoS2	—	Batch	623	44	5	—	4MP	Tol	101
MoS2	—	Parr	623	28	1	—	Ph	Ben	79
MoS2	—	Batch	593–643	28	6	—	4MEP	Ben, Ph, ANI, MPh	82
MoS2	C	Batch	573	50	4.17		GUA	Ph, CAT	121
MoS2	γ-Al2O3	Fixed bed	573	40	—	—	GUA	CYHE, MCYHE	92
MoS2	γ-Al2O3	Fixed bed	573	50	—	2	ANI	Ph, Cr, 2,6-DMP	117
ReS2	SiO2	Batch	573	50	4.4	—	GUA	PH, Ben, CYHE, CYHA	168
	γ-Al2O3							PH, CAT

---

Table 4 HDO catalyzed by metal phosphides (T is temperature; P is pressure; t is time. WHSV is weight hourly space velocity)

Catalyst	Support	Reactor configuration	Operating conditions				Reactant	Major products	Ref.
			T (K)	P (bar)	t (h)	WHSV (h^−1)
Ni2P	SiO2	Packed bed	573	1	0.33	—	GUA	Ben	101
Ni2P	SiO2	Fixed bed	573	15	0–10	10	ANI	Ph, Ben, CYHA	90
NiMoP	SiO2	Fixed bed	573	15	0–10	10	ANI	Ph, Ben, CYHA	90
MoP	SiO2	Packed bed	573	1	0.33	—	GUA	Ben	101
MoP	SiO2	Fixed bed	573	15	0–10	10	ANI	Ph, Ben, CYHA	90
MoP	none	Batch	573	44	5	—	4MP	Tol, MCYHA	127
Co2P	SiO2	Packed bed	573	1	0.33	—	GUA	Ben	101
Fe2P	SiO2	Packed bed	573	1	0.33	—	GUA	Ph	101
WP	SiO2	Packed bed	573	1	0.33	—	GUA	Ph	101

---

Metals, including bimetallics

Both mono- and bi-metallic catalysts have been tested for upgrading of lignin-derived bio-oils, and for appropriate accessibility of their surfaces (and corresponding high activity), the metals are dispersed on porous supports (Tables 1 and 2).

In the upgrading reactions, there is generally a competition between HDO and hydrogenation of aromatic rings. Metals including iron, cobalt, nickel, ruthenium, palladium, rhodium, platinum, and iridium show activity. For example, Olcese et al.¹¹⁴ used iron supported on silica (Fe/SiO₂) for HDO of guaiacol, finding it to have good selectivity for formation of benzene, toluene, and xylene without much aromatic ring hydrogenation. Fe/SiO₂ was found to have a lower activity but higher selectivity for HDO than cobalt catalysts. Moreover, Fe/SiO₂ produced less methane than cobalt-containing catalysts even at high conversions.¹¹⁴

An investigation¹¹⁵ of the effects of individual gases present in a pyrolysis mixture (H₂, CO, CO₂, H₂O, CH₄) on the selectivity of a Fe/SiO₂ catalyst led to the conclusion that although CO has a strong positive effect on guaiacol HDO by reducing the rate of catalyst deactivation, CO₂ has a strong negative effect on the initial reaction rate because of the resultant carburization of iron and coke formation. Water hinders the overall catalytic conversion and enhances the production of phenols rather than benzene as a consequence of the hydrolysis of adsorbed precursors. CH₄ was found not to have a significant effect on the guaiacol conversion.¹¹⁵

In an investigation of the HDO of substituted phenols (including anisole, guaiacol, o-cresol, benzyl alcohol, catechol, and vanillyl alcohol) catalyzed by copper chromite at 423–548 K and 50 bar H₂ partial pressure, Deutsch and Shanks¹¹⁶ concluded that copper catalyzed hydrogenation and hydrogenolysis reactions and Cr₂O₄, as a weak acid, catalyzed dehydration and transalkylation reactions. The authors concluded that in the conversion of anisole, demethoxylation to give benzene was the dominant reaction pathway, in contrast to the demethylation and transalkylation observed with conventional hydroprocessing catalysts. Thus, the authors inferred that their catalyst is characterized by unique selectivity and reaction networks for the removal of hydroxyl and methoxy groups relative to the widely investigated HDO catalysts.

Gutierrez et al.⁴⁸ reported that the activity for guaiacol HDO decreased in the following order:

Rh/ZrO₂ > Co-MoS₂/Al₂O₃ > Pd/ZrO₂ > Pt/ZrO₂ (1)

A number of investigators have reported experiments with bimetallic catalysts. For example, in the conversion of guaiacol, the combinations of rhodium with platinum and with palladium led to much better results than were observed with mono-metallic platinum and palladium catalysts.⁴⁸ Huuska¹¹⁷ considered the effect of catalyst composition on the hydrogenolysis of anisole with catalysts containing molybdenum or nickel (Tables 5 and 6). When the catalysts were supported on an acidic oxide, the main products were phenol and methyl-substituted phenols, consistent with the rapid methyl group transfer catalyzed by the acidic sites. When the support was not acidic, only limited methyl group transfer occurred, and the main products were phenol and cyclic hydrocarbons. In related work, Ardiyanti et al.¹¹⁸ investigated catalytic hydroprocessing of anisole using Ni–Cu catalysts with various Ni/Cu ratios supported on δ-Al₂O₃. They observed lower conversions with the monometallic nickel than with the bimetallic catalyst. The higher activity of the bimetallic catalysts may be attributed to the incorporation of less nickel in the bulk Al₂O₃ when the copper is present and as a result less formation of nickel spinel than in the monometallic nickel catalyst.¹¹⁸

Table 5 Effect of catalyst composition and properties on the hydrogenolysis of anisole¹¹⁷

Catalyst	Surface area (m^2 g−^l)	Bulk density (g cm^−3)	Catalyst age (h)	Anisole conversion (%)
γ-Al2O3	100	0.8	4	22.5
SiO2–Al2O3	100	0.6	4	10.2
MoS2 (60% Mo)	6.9	1.7	5	50.0
Ni(11%)/SiO2	280	0.6	5	19.4
Mo(6.7%)/γ-Al2O3	160	0.9	6	74.5
Ni(11%)/γ-Al2O3	65	1.0	6	42.1
Ni(0.5%)/γ-Al2O3	100	0.9	7	90.1

---

Table 6 Effect of catalyst composition on the product distribution in the conversion of anisole; see the footnotes for details^a,117

Catalyst	Product	Selectivity, %
a Reaction conditions: experiments were carried out in a fixed-bed flow reactor fed with H2 and neat anisole containing CS2 to maintain catalysts in a sulfided state; catalysts D–G were presulfided; catalyst E was not. The temperature was 573 K; the H2 partial pressure 5 mPa; and the LHSV 2 h^−1, except when the catalyst was C, for which the LHSV was 1 h^−1. The data were collected after the each catalyst had been onstream for several hours. See the original reference for details regarding the mass balance closures and the calculation of the selectivities.
A, γ-Al2O3	Phenol	46.8
	o-Cresol	27.9
	2,6-Dimethylphenol	9.7
	Other oxygen-containing compounds	13.1
	Total oxygen-containing compounds	97.5
	Benzene	1.9
	Other hydrocarbons	0.6
	Total hydrocarbons	2.5
B, SiO2–Al2O3	Phenol	51.1
	o-Cresol	12.9
	2,6-Dimethylphenol	1.8
	Other oxygen-containing compounds	28.7
	Total oxygen-containing compounds	94.5
	Benzene	1.3
	Other hydrocarbons	4.1
	Total hydrocarbons	5.4
C, MoS2 (60% Mo)	Phenol	63.1
	o-Cresol	1.9
	2,6-Dimethylphenol	—
	Other oxygen-containing compounds	0.4
	Total oxygen-containing compounds	65.4
	Benzene	27.8
	Other hydrocarbons	6.8
	Total hydrocarbons	34.6
D, Ni(11%)/SiO2	Phenol	53.8
	o-Cresol	0.8
	2,6-Dimethylphenol	—
	Other oxygen-containing compounds	5.1
	Total oxygen-containing compounds	59.7
	Benzene	31.4
	Other hydrocarbons	9.0
	Total hydrocarbons	40.4
E, Mo(6.7%)/γ-Al2O3	Phenol	47.0
	o-Cresol	29.5
	2,6-Dimethylphenol	14.8
	Other oxygen-containing compounds	1.8
	Total oxygen-containing compounds	93.1
	Benzene	3.0
	Other hydrocarbons	3.8
	Total hydrocarbons	6.8
F, Ni(11%)/γ-Al2O3	Phenol	77.1
	o-Cresol	8.6
	2,6-Dimethylphenol	—
	Other oxygen-containing compounds	0.6
	Total oxygen-containing compounds	86.3
	Benzene	8.3
	Other hydrocarbons	5.4
	Total hydrocarbons	13.7
G, Ni(0.5%)/γ-Al2O3	Phenol	40.4
	o-Cresol	35.0
	2,6-Dimethylphenol	18.6
	Other oxygen-containing compounds	5.0
	Total oxygen-containing compounds	99.0
	Benzene	—
	Other hydrocarbons	1.0
	Total hydrocarbons	1.0

---

Jongerius et al.¹¹⁹ investigated the HDO of guaiacol catalyzed by carbon nanofiber-supported transition metal carbides (W₂C/CNF and Mo₂C/CNF) and obtained higher selectivity and conversion to phenolic compounds than with CoMo/Al₂O₃. According to their results, Mo₂C/CNF has higher activity, selectivity, and stability than W₂C/CNF and yields more completely deoxygenated aromatic products. The authors observed that higher temperatures and longer reaction times gave higher yields of the fully deoxygenated products benzene and toluene.

Metal sulfides

Metal sulfide catalysts are widely used in hydroprocessing technology to remove sulfur and nitrogen from petroleum. Many such catalysts have been tested for HDO. These include MoS₂ and supported catalysts that incorporate cobalt and molybdenum or nickel and molybdenum (abbreviated as CoMoS and NiMoS).^{48,68,78–81,88,91,92,96,97,111,113,117,120,121} The supported catalysts typically incorporate clusters of MoS₂ that include cobalt or nickel ions. These structures are maintained in the sulfide form in petroleum refining operations because the reaction environments contain H₂ and H₂S, which is formed in HDS reactions from components in petroleum such as thiols and thiophenic compounds. Because sulfur is almost absent from biomass, the supported metal sulfides would not be stable in the presence of the oxygen-containing compounds, and the catalysts would be converted into oxide forms—unless a sulfur source such as H₂S or a thiol were added to the feed,^89,122,123 with the detriment (a) that some sulfur would then be present in the product, which upon further processing with a noble metal catalyst would be poisoned by the sulfur-containing compounds or (b) that the product, upon combustion, would form sulfur oxides that pollute the atmosphere.

Table 3 is a summary of results characterizing HDO catalyzed by metal sulfides. HDO of phenolic compounds catalyzed by sulfided CoMo/Al₂O₃ was investigated by Furimsky et al.¹²⁴ Presulfiding of the catalyst had a pronounced effect on the conversion of phenols to benzene and cyclohexane. Vuori et al.¹²⁵ observed a benefit of sulfiding of such catalysts in the conversion of guaiacol. Lin et al.⁹⁷ inferred that demethylation, demethoxylation, and deoxygenation reactions following saturation of the aromatic ring in guaiacol accounted for the product distributions observed with sulfided CoMo/Al₂O₃ and NiMo/Al₂O₃ catalysts.

Ferrari et al.¹²⁶ investigated the effect of H₂S partial pressure on HDO of a mixture of compounds including guaiacol. They observed that increasing the H₂S partial pressure led to inhibition of the hydrogenolysis and hydrogenation pathways. Gutierrez et al.⁴⁸ reported that in guaiacol HDO with a conventional sulfided CoMo/Al₂O₃ catalyst, the products were contaminated with sulfur and the catalyst underwent deactivation associated with carbonaceous deposits.

Metal phosphides

Transition metal phosphides have been investigated widely as catalysts for petroleum hydroprocessing and drawn interest as candidate catalysts for bio-oils upgrading.^{90,101,109,127} Results are summarized in Table 4. These catalysts have relatively high activities in our context, with the reactions characterized by relatively low activation energies. Zhao et al.¹⁰¹ indicated that at high conversions of guaiacol, the major products were phenol, benzene, and methoxybenzene, whereas at low conversions, the main product was catechol. These authors observed higher HDO conversions with metal phosphide catalysts than with sulfided CoMo/Al₂O₃ catalysts, which underwent rapid deactivation. However, in the conversion of guaiacol, more coke formed on the phosphide than on the sulfide catalyst, consistent with the tendency of the Lewis acid sites (or Brønsted acid sites formed from them) of phosphide catalysts to favor the demethylation pathway. The authors reported the following decreasing order of activities for guaiacol HDO:

Ni₂P > Co₂P > Fe₂P > WP > MoP (2)

Whiffen et al.^109,127 considered HDO of 4-methylphenol on a MoP catalyst and compared it with that on other molybdenum-containing low-surface-area catalysts. They found the following decreasing order of activity measured by the conversion of 4-methylphenol:

MoP > MoS₂ > MoO₂ > MoO₃ (3)

The activation energy of the reaction increased in the reverse order.

Li et al.⁹⁰ investigated the conversion of anisole catalyzed by Ni₂P/SiO₂, MoP/SiO₂, and NiMoP/SiO₂ with various Ni/Mo ratios. The reaction sequence was demethylation of anisole, hydrogenolysis of phenol, and hydrogenation of benzene. As the Ni/Mo ratio increased, the activity and stability of the phosphide catalyst increased. The nickel phosphide-containing catalysts, especially Ni₂P/SiO₂, were found to have higher activities than conventional NiMo/γ-Al₂O₃ catalysts.⁹⁰ Yang et al.⁷² investigated the effect of phosphorus as an additive in a CoMo/MgO catalyst, finding that its addition gave a much more active catalyst for phenol conversion because of reduced coke formation.

Metal nitrides

Metal nitrides have also drawn attention as hydroprocessing catalysts and been investigated for HDO. Because of electronegativity differences between the metal and nitrogen atoms, nitrides can offer both acidic and basic sites. The potential advantages of metal nitride catalysts include their resistance to oxidation, ease of preparation, and low cost. Also, because of unique catalytic reaction pathways, hydrogen consumption may be minimized.

In HDO reactions on nitride catalysts, nitrogen may be released from the bulk lattice and create surface sites for adsorption and catalysis.¹²⁸ There is only little reported work on metal nitrides such as Mo₂N as HDO catalysts.^102,129 Ghampson et al.^93,94 used Mo₂N to catalyze guaiacol HDO, observing a high activity and a significant conversion of guaiacol to phenol when the catalyst was γ-Mo₂N but not when it was β-Mo₂N_0.78 or other molybdenum-containing compounds such as MoO₂ or molybdenum metal. Furthermore, their experiments showed that the bimetallic nitride catalyst CoMoN gave higher yields of deoxygenated products than the monometallic nitride catalyst, but the overall activity of the monometallic nitride catalyst was higher than that of the bimetallic.

Supports

A major role of a support is to stabilize the active catalytic species in a highly dispersed form, and one of the most commonly applied supports in HDO catalysts is alumina. The acidic character of a transition alumina influences the performance of the catalyst, being active for reactions such as transalkylation. Data characterizing the conversion of guaiacol are illustrative.^68,92 A major disadvantage of the acidic support is the high yields of coke that are associated with the acidic surface sites. Weak Lewis acid sites have been inferred to lead to coke formation during HDO.⁶⁸

A number of recent investigations have focused on developing new supports for decreasing H₂ consumption, increasing selectivity for direct oxygen removal, and minimizing coke formation. Echeandia et al.⁸⁵ used Ni–W catalysts supported on active carbon for HDO of phenol, observing that much less coke formed on the carbon than on the alumina. Silica and carbon supports give catalysts with relatively high selectivities in HDO, but alumina-supported catalysts have higher activities.¹³⁰

Lee et al.⁹⁵ investigated various supports (Al₂O₃, SiO₂–Al₂O₃, and nitric-acid-treated carbon black) for noble metal catalysts. Fig. 10 shows that the main product was cyclohexane when the support was SiO₂–Al₂O₃, but when the support was Al₂O₃ or nitric-acid-treated carbon black, the yield of cyclohexanol and 2-methoxycyclohexanol exceeded that of cyclohexane. This comparison indicates that deoxygenation activity of the SiO₂–Al₂O₃ support is higher than those of the other supports, and the SiO₂–Al₂O₃ might have catalyzed dehydration of the alcohols.

Fig. 10 Product distributions in the HDO of guaiacol with platinum, rhodium, palladium, and ruthenium supported on alumina, silica-alumina, and nitric acid-treated carbon black.⁹⁵

The acidic Al₂O₃ favors the formation of catechol via a demethylation route and subsequently the formation of methylphenols, but the more weakly acidic SiO₂ favors the formation of phenol via a demethoxylation route. Sepúlveda et al.¹³¹ investigated the conversion of guaiacol catalyzed by ReS₂ supported on Al₂O₃ and on SiO₂, concluding that the ReS₂/Al₂O₃ catalyst gives a higher initial rate of guaiacol conversion because of the direct formation of phenol via demethylation but that ReS₂/SiO₂ is more active for HDO because of the higher rate of formation of deoxygenated compounds. The main products observed with the ReS₂/Al₂O₃ catalyst were phenol, catechol, and deoxygenated compounds such as benzene, cyclohexene, and cyclohexane, whereas with the ReS₂/SiO₂ catalyst, phenol and deoxygenated compounds were the major products.¹³¹

Loricera et al.⁹¹ investigated the HDO of anisole catalyzed by mesoporous silica (SBA-15 and SBA-16)-supported CoMoW catalysts, pointing out the potential advantages of these supports, including high surface areas and the opportunities to vary the framework compositions. Ghampson et al.⁹³ compared the activities of Al₂O₃- and SBA-15-supported molybdenum nitride catalysts for HDO of guaiacol on the basis of reaction rates and phenol/catechol yields. They observed that the SBA-15-supported catalysts gave higher phenol/catechol yields than the Al₂O₃-supported catalysts because of the direct transformation of guaiacol to phenol. The lower production of catechol with the SBA-15 support is a potential advantage in minimizing coke formation and pore blocking in upgrading of mixtures of compounds in lignin-derived bio-oils.⁹³

Bui et al.^92,113 compared various supports for MoS₂ and CoMoS catalysts in guaiacol HDO (Fig. 11); the supports were γ-Al₂O₃, ZrO₂, and TiO₂. ZrO₂ gave a catalyst with a high activity and a higher selectivity for hydrogenolysis than the others. As expected, methyl group transfer reactions were kinetically significant when the support was γ-Al₂O₃, and the activity for formation of deoxygenated compounds was relatively low.^48,113,132 An investigation of the conversion of lignin-derived phenolic compounds to hydrocarbons catalyzed by Ni/SiO₂–ZrO₂^133,134 demonstrated high conversions and selectivities to hydrocarbons with high octane numbers. According to the authors, Ni/SiO₂–ZrO₂ catalysts have appropriate recyclability and anti-coking performance in the HDO reaction of model phenolic compounds because of the amphoteric character of the SiO₂–ZrO₂ support.

Fig. 11 (A) Conversion of guaiacol catalyzed by cobalt-promoted MoS/Al₂O₃.¹¹³ (B) Conversion of guaiacol catalyzed by cobalt-promoted MoS/TiO₂.¹¹³ (C) Conversion of guaiacol catalyzed by cobalt-promoted MoS/ZrO₂.¹¹³

Yang et al.¹³⁵ investigated the influence of supports on the selectivity of anisole HDO for aromatic products with a series of nickel-containing catalysts on supports including SBA-15, Al-SBA-15, γ-Al₂O₃, microporous carbon, TiO₂, and cerium oxide. The data show that Ni/C catalysts not only give high conversions similar to those observed with Ni/SBA-15, Ni–Al-SBA-15, and Ni/γ-Al₂O₃, but that they lead to relatively low yields of oxygen-containing compounds in comparison with the other catalysts. The results indicate that, like Ni/C, the nickel catalysts supported on reducible oxides such as TiO₂ and CeO₂ are characterized by significant selectivities for formation of aromatic compounds. The data also demonstrate a trend of increasing selectivity for benzene at the expenses of cyclohexane for the series Ni/C, Ni/Al-SBA-15, and Ni/γ-Al₂O₃; the authors inferred that carbon deposition on nickel sites led to a moderate loss of hydrogenation activity, as a result enhancing aromatic production.¹³⁵

Nimmanwudipong et al.^54,55 investigated the conversion of guaiacol catalyzed by Pt/Al₂O₃ in the presence of H₂, comparing the performance of this catalyst with that of the acidic HY zeolite. They observed that the two catalysts were similarly selective for HDO and transalkylation reactions but that the oxygen removal reactions were limited when the catalyst was the zeolite.

A few researchers have worked on catalysts supported on the basic MgO.^{100,136–140} Yang et al.⁷² used MgO-supported catalysts for HDO of phenolic compounds, concluding that they were highly active for conversion of phenol to benzene and cyclohexyl-aromatics. Nimmanwudipong et al.¹⁰⁰ used Pt/MgO to catalyze the reactions of guaiacol, finding that the basic support gave a catalyst that was more stable than the acidic γ-Al₂O₃, presumably because the MgO led to less coke formation on the surface; the two catalysts were similar in activity.

Promoters

The metal sulfide catalysts used in petroleum hydroprocessing are much improved by the presence of promoters, typically cobalt or nickel in the presence of MoS₂ supported on Al₂O₃; the promoter is present as ions in the layer structure of MoS₂.⁸⁰ Romero et al.¹⁴¹ investigated the effect of cobalt and nickel promoters on the performance of sulfided molybdenum catalysts used for HDO of 2-ethylphenol; each promoter increased the activity by about 1.7 times under their conditions. Yoosuk et al.⁸⁰ showed that Ni–Mo sulfides were more active than nickel or molybdenum sulfides alone, and the maximum activity was obtained at a Ni/(Ni + Mo) atomic ratio of 0.3. Bui et al.^92,113 considered the effect of cobalt as a promoter of MoS₂ catalysts for HDO of guaiacol (Fig. 12), finding that it improved performance by increasing the rate of direct deoxygenation. These authors observed that the major deoxygenated products formed in the presence of the promoted catalyst were methyl-substituted aromatic compounds such as toluene, whereas methylcyclohexene was a major product when the catalyst was unpromoted.

Fig. 12 (A) Conversion of guaiacol at 573 K catalyzed by MoS₂ (ref. 92) these data represent performance of the unpromoted catalyst. (B) Conversion of guaiacol at 573 K catalyzed by cobalt-promoted MoS.⁹² A comparison with (A) shows the influence of the promoter.

One might tentatively suggest that the promotion of sulfided catalysts may affect HDO more or less as it affects HDS, although work is needed to test the suggestion. The roles of promoters in non-sulfide HDO catalysts remain to be elucidated.

Catalyst deactivation

Catalyst deactivation in HDO processes is caused by deposition of carbonaceous material (coke), sintering and loss of surface area of catalytic components such as metals, and poisoning of catalytic sites by compounds such as those containing nitrogen and phosphorus, which are present in biomass.¹¹² Oxide-containing catalysts are not stable in liquid water at high temperatures, and water produced in HDO can destroy catalysts.

Catalyst deactivation typically influences the rates of HDO and accompanying reactions such as transalkylation differently because they often take place on different components of the catalyst (e.g., metal particles and acidic supports).⁷³

The major cause of deactivation of HDO catalysts as stated in published reports is evidently associated with deposits of coke or related materials; besides covering catalytic sites, the deposits may block pores, especially those of small-pored materials such as zeolites. Coke formation occurs through polymerization and polycondensation reactions and takes place at rates that depend strongly on the nature of catalyst, the nature of the bio-oil compounds, and the operating conditions. For example, Nimmanwudipong et al.⁵⁵ in an investigation of guaiacol conversion found that the conversion declined by about 50% within only an hour on stream when the catalyst support was HY zeolite, whereas the deactivation was much slower with Pt/γ-Al₂O₃ under similar conditions (Fig. 13), presumably because the blocking of zeolite pores was much more pronounced than the blocking of the larger γ-Al₂O₃ pores.

Fig. 13 Conversion of guaiacol catalyzed by (■) Pt/γ-Al₂O₃ and (△) HY zeolite at 573 K.⁵⁵

Because of their high concentrations of unsaturated hydrocarbon moieties, lignin-derived bio-oils are strongly prone to coke formation. These reactants typically interact strongly with surface catalytic sites, likely facilitating coke formation. During hydrotreating reactions, it is observed that adsorption of phenolic components on the support hinders the access of other compounds to the active sites.¹⁴ Oxygen-containing compounds with more than one oxygen atom, such as catechol, have been inferred to be more prone to coke formation than phenol.^67,101,142 As mentioned above, although the acidity (Lewis and Brønsted) increases HDO activity, it also fosters catalyst deactivation by coke formation.³⁰ Thus, one might expect that an optimum in strength and number of acidic sites may maximize catalyst performance in HDO, but this suggestion is still preliminary. One also expects that coke formation can be mitigated by lowering the temperature and reactant-catalyst contact time and by increasing the H₂ partial pressure, especially when metals are present to catalyze dissociation of H₂ and thereby facilitate hydrogenation of coke precursors.^73,91,92 Because higher temperatures increase the rates of HDO reactions, it may be useful to use staged reactors operating with different catalysts and different temperatures to find the best trade-offs between rates of desired catalytic reactions and rates of catalyst deactivation.¹⁴³ Catalyst staging is common in petroleum hydroprocessing. More data are needed to test these hypotheses and provide more thorough quantitative results.

Catalyst poisons such as compounds of nitrogen and even oxygen-containing compounds (e.g., water, CO) may compete with the reactants for catalytic sites in HDO. Adsorption of poisons on Lewis and Brønsted acidic sites depends on reaction conditions.¹¹² Water and other oxygen-containing compounds have an inhibiting effect on conventional metal sulfide and metal phosphide catalysts (e.g., NiMo/Al₂O₃ and CoMo/Al₂O₃).^90,144 Sulfide catalysts may be deactivated as the sulfide structures are converted to oxides. Addition of sulfur-containing compound such H₂S to the feed can minimize such catalyst deactivation, but at the potential cost of formation of undesirable sulfur-containing products.^48,85,90,144 Water also causes the oxidation of phosphide catalysts, leading to the formation of metal oxides and metal phosphates.⁹⁰ Odebunmi and Ollis⁸³ observed a reduction of catalytic activity caused by water during HDO of m-cresol, and they regenerated the catalyst by resulfiding it. Laurent and Delmon^67,89 observed that water caused only weak inhibition in HDO of phenols relative to that observed with compounds such as H₂S and NH₃, but when a high enough concentration of H₂O–H₂S was present in the feed, the detrimental effect of water was more significant. Nevertheless, water may have a beneficial effect by mitigating the temperature increases resulting from the strong exothermicity of hydrotreating reactions and thereby minimize catalyst deactivation resulting from sintering.

In attempts to understand catalyst deactivation in HDO, Popov et al.¹⁴⁵ investigated the interactions of guaiacol, phenol, and ethylphenols with a sulfided CoMo/Al₂O₃ catalyst. They concluded that the surface properties of the catalyst support and the basicity and nature of the oxygen-containing compound affect the interactions and the resultant poisoning. Phenate-type species on the Al₂O₃ support obstruct the accessibility to the metal sulfide crystal edge sites where catalysis occurs. The metal sulfide sites are also poisoned by oxygen-containing compounds.¹⁴⁵

The data summarized in Table 7 provide information about carbon formation on various catalysts after catalytic reactions of guaiacol. The data indicate that mono-metallic catalysts are less resistant to coke formation than bimetallics, but one should be cautious about generalizing this statement. A generalization that is more strongly supported by data is that catalysts with more strongly acidic supports typically undergo deactivation faster than those with less strongly acidic supports, as exemplified by data characterizing NiMo supported on SiO₂–Al₂O₃vs. γ-Al₂O₃.¹⁴⁶ H-Beta zeolite was found to be characterized by a higher rate of deactivation than Pt/H-Beta zeolite (or Pt/SiO₂), because dissociation of H₂ on the metal regenerated acidic sites by hydrogenation of coke precursors and/or coke.⁷³ The hydrophobic character of carbon surfaces corresponds to less rapid catalyst deactivation by water than occurs on acidic supports.^85,92 But recall that acidic groups on supports, although they are responsible for coke formation, also may be desired because they help catalyze hydrogenolysis and hydrocracking.⁹¹

Table 7 Carbon formation on various catalysts in batch reactor experiments performed with guaiacol as the reactant (T is temperature; P is pressure; t is time)

Catalyst	Operating conditions			Carbon deposition (wt%)	Ref.
	T (K)	P (bar)	t (h)
γ-Al2O3	553	70	3	10.3	68
K–CoMoS–Al2O3	553	70	3	9.5	68
Pt–CoMoS–Al2O3	553	70	3	9.1	68
CoMoS–Al2O3	553	70	3	8.9	68
CoMoS	553	70	3	2.8	68
CoMoS/SiO2	553	70	3	2.7	68
Ni/SiO2	593	170	1	15.8	48
NiCu/SiO2	593	170	1	2.3	48
CoMoS–Al2O3	373	80	5	6.7	61
Rh/ZrO2	373	80	5	1.8	61
Pd/ZrO2	373	80	5	0.7	61
Pt/ZrO2	373	80	5	0.6	61
ZrO2	373	80	5	0.5	61

---

Kinetics

Because of the complexity and variability of structures of compounds in bio-oils and variations in molecular weights and concentration, little quantitative information characterizing the kinetics of bio-oils hydroprocessing is available. In HDO process, oxygen can be removed as CO₂, CO, H₂O, and other compounds, by a variety of reactions. Little information is available to resolve the reactions and characterize their rates separately.

Most of the available kinetics data characterize reactions of individual oxygen-containing compounds in the presence of H₂. A number of researchers have represented their data in terms of reactions that are first-order in the oxygen-containing compound.^{51–53,56,68,79,82,85,101,109,127,147,148} Summaries of some pseudo-first-order rate constants and apparent activation energies for conversion of lignin-derived bio-oil compounds are summarized in Tables 8–10.

Table 8 Pseudo-first-order rate constants characterizing HDO reactions of compounds derived from lignin

Reactant compound	Catalyst	Pseudo-first-order rate constant	Dimensions of rate constant	Ref.
a At 598, 623, 648 K. b At 598, 623, 648 K. c At 598, 623, 648 K. d At 598, 623, 648 K.
ANI	Pt/Al2O3	19 (at 573 K)	L per g of catalyst per h	52 and 53
GUA	Pt/Al2O3	16.2 (at 573 K)	L per g of catalyst per h	55
GUA	γAl2O3	0.02 (at 553 K)	L per g of catalyst per h	68
GUA	CoMo/Al2O3	0.08 (at 553 K)	L per g of catalyst per h	68
GUA	CoMo/Al2O3 modified with Pt	0.04 (at 553 K)	L per g of catalyst per h	68
GUA	CoMo/SiO2	0.02 (at 553 K)	L per g of catalyst per h	68
GUA	CoMo/C	0.01 (at 553 K)	L per g of catalyst per h	68
GUA	CoMoS	0.02 (at 553 K)	L per g of catalyst per h	68
4MP	MoO3	4.54, 21.7, 29.3^a	L h^−1 molMo^−1)	109
4MP	MoO2	1.04, 3.17, 6.16^b	L h^−1 molMo^−1	109
4MP	MoS2	1.04, 2.70, 6.19^c	L h^−1 molMo^−1	109
4MP	MoP	2.74, 9.43, 13.25^d	L h^−1 molMo^−1	109
CYHAONE	Pt/γ-Al2O3	90 (at 573 K)	L per g of catalyst per h	51
4MA	Pt/γ-Al2O3	11 (at 573 K)	L per g of catalyst per h	147
4MA	Pt/SiO2–Al2O3	31 (at 573 K)	L per g of catalyst per h	147
EUG	Pt/γ-Al2O3	12.4 (at 573 K)	L per g of catalyst per h	56
4MP	CoMoS	0.14, 0.03 (at 573 K)	L per g of catalyst per h	96
2MP	CoMoS	0.08, 0.01 (at 573 K)	L per g of catalyst per h	96

---

Table 9 Apparent activation energies characterizing HDO reactions of compounds derived from lignin

Reactant compound	Catalyst	Activation energy (kJ mol^−1)	Ref.
GUA	Ni2P/SiO2	40	101
GUA	MoP/SiO2	63	101
GUA	Co2P/SiO2	52	101
GUA	WP/SiO2	23	101
Ph	MoS2	102	148
4MP	MoS2	135	148
4MEP	MoS2	104	148
4MP	MoO3	122	109
4MP	MoO2	113	109
4MP	MoS2	110	109
4MP	MoP	98	109

---

Table 10 Pseudo-first-order rate constants characterizing reactions in the network of Fig. 14 (ref. 103)

Reaction number (keyed to Fig. 14)	Reactant	Rate constant^a	Reaction class
a Rate constants in L per g catalyst per h were determined for the conversion of reactant to a particular product; for example, phenol is formed in more than one reaction and the rate constant accounts for the formation of phenol in both of these reactions.
1	Guaiacol	0.11	Hydrodeoxygenation
2	Guaiacol	4.4	Hydrodeoxygenation
3	Guaiacol	6.5	Hydrogenolysis
4	Guaiacol	1.8	Transalkylation
5	Guaiacol	0.50	Bimolecular transalkylation
6	Guaiacol	0.21	Bimolecular transalkylation
7	Guaiacol	0.26	Bimolecular transalkylation
8	Anisole	12	Hydrogenolysis
9	Anisole	0.86	Hydrodeoxygenation
10	Anisole	2.8	Transalkylation
11	Anisole	0.039	Transalkylation
12	Anisole	0.14	Bimolecular transalkylation
13	4-Methylanisole	0.76	Hydrodeoxygenation
14	4-Methylanisole	4.2	Hydrogenolysis
15	4-Methylanisole	2.2	Transalkylation
16	Cyclohexanone	77	Dehydrogenation
17	Cyclohexanone	5.5	Hydrogenation
18	Cyclohexanone	0.05	Hydrodeoxygenation

---

Runnebaum et al.⁵³ considered the reaction network of anisole conversion catalyzed by Pt/Al₂O₃ and presented pseudo-first-order rate constants for the removal of anisole, using selectivity-conversion data to determine the rate constants for formation of individual products. The rate constants for formation of phenol, 2-methylphenol, 2-methylanisole, and 4-methylphenol as primary products were found to be 12, 2.8, 0.14, and 0.039 L per g of catalyst per h, respectively, at 573 K and atmospheric pressure. Nimmanwudipong et al.⁵⁵ developed an extensive reaction network for the conversion of guaiacol catalyzed by Pt/γ-Al₂O₃, reporting a set of pseudo-first-order rate constants for guaiacol conversion at 573 K (Table 10; this table is keyed to Fig. 14).

Fig. 14 Reaction network accounting for formation of primary products determined from analysis of selectivity-conversion plots for the conversion of individual reactants (shown in red), guaiacol, anisole, 4-methylanisole, and cyclohexanone, catalyzed by Pt/Al₂O₃ in the presence of H₂ at 573 K and atmospheric pressure. The reactant H₂ is omitted for simplicity. Pseudo-first-order rate constants for the individual primary products formed in the conversion of the individual reactants catalyzed by Pt/Al₂O₃ are shown in Table 10; the numbers next to the arrows in this figure are keyed to the list in Table 10.¹⁰³

Zhang et al.¹⁴⁹ investigated the HDO of bio-oils in the presence of a Co–MoS₂/Al₂O₃ catalyst in a batch reactor and temperatures between 633 and 663 K. They inferred an overall reaction order of 2.3 for HDO (and an order such as this may correspond to simultaneous first-order reactions of each of a group of oxygen-containing compounds). A weak dependence of rate on H₂ partial pressure was observed in the range of 15 to 30 bar. In an investigation of guaiacol HDO on transition metal phosphide catalysts,¹⁰¹ activation energies in the range of 23–63 kJ mol⁻¹ were observed.

Massoth et al.⁷⁹ used Langmuir–Hinshelwood expressions to represent the kinetics of HDO of methyl-substituted phenols catalyzed by a sulfided CoMo/Al₂O₃ in a flow microreactor at 573 K and 2.85 MPa H₂ partial pressure. The equations account for adsorption of the organic reactants on catalytic sites and for competitive direct deoxygenation and hydrogenation reactions; they do not account for adsorption of hydrogen or products and do not distinguish between adsorption of reactants on different catalytic sites for direct deoxygenation and hydrogenation. One may expect that, as more complete data become available to characterize HDO reactions, Langmuir–Hinshelwood expressions will be found to apply with some generality, but now there are too few data to justify any generalizations.

Bui et al.^92,113 investigated the kinetics of HDO of guaiacol in the presence of CoMo catalysts in a fixed-bed reactor, obtaining data that they represented by first-order kinetics. Such an approximation was also made by Runnebaum et al.,¹⁰³ as shown in the comparison of reactivities of various compounds shown in Table 10 for reactions catalyzed by Pt/Al₂O₃.

In sum, there is a lack of detailed, quantitative determinations of kinetics of HDO and a need for further research in this direction, with the expected benefits being quantitative indications of the relative strengths of adsorption of the various reactants and products on various catalysts and comparisons of the reactivities of various compounds with various catalysts. Such results would help elucidate patterns of competitive adsorption and reaction in catalytic HDO and lead to equations corresponding to Langmuir–Hinshelwood forms.

Reactor configurations and catalyst staging

Data indicating the types of reactors and operating conditions in investigations of HDO of lignin-derived bio-oils are summarized in Tables 1–4. Testing of catalysts for HDO of bio-oils has been carried out under a wide range of conditions, and, roughly speaking, one refers to mild HDO taking place at about 523 K and 100 bar and to severe HDO taking place at about 623 K and 200 bar. We emphasize that most of the experiments with individual compounds as reactants were carried out under milder conditions than these.

When the reactors are staged, operation of the upstream reactor operating under the milder conditions may facilitate hydrogenation, and operation of the downstream reactor under more severe conditions may facilitate deoxygenation. This approach allows selection of the optimum catalyst for each of the stages with respect to the oil yield and degree of deoxygenation. Elliot et al.^36,37 proposed such a configuration of reactors, with the upstream reactor operated at a temperature below 553 K, a pressure of about 140 bar, and a liquid hourly space velocity (LHSV) of 0.62. A deep deoxygenation step was performed in the downstream reactor in the presence of H₂ at a total pressure of 142 bar, a temperature higher than 623 K, and a very low LHSV of 0.11. Oil yields between 30 and 55% and deoxygenation conversions up to 99% were achieved.

In early bio-oils HDO research, batch reactors were typically used, and two-step procedures were applied; in the first step (mild HDO), thermal decomposition reactions were significant, and overall conversions for oxygen removal were typically about 50%. In the second step, conversions for oxygen removal of more than 90% at about 623 K and 200 bar were typically achieved in reaction times of 2 h.

Most recent experiments characterizing bio-oils HDO have been carried out in continuous packed-bed flow reactors. Downflow has given better performance and less tendency to plug the reactors than upflow, and residence time distributions have favored higher conversions in downflow operation because of the optimized contacting of reactant and catalyst.^27,150

Research opportunities

Investigations of HDO of lignin-derived bio-oils and bio-oil components for fuel production are still at an early stage, and much more information is needed to improve candidate catalysts and processes if commercialization is to become economical. Commercial applications do not seem to be close at hand absent public subsidies for research and development or for avoidance of non-renewable fuels. It is not yet evident what the most appropriate catalysts will be, as insufficient work has been done to identify the optimum combinations of catalyst functions, let alone the specific catalyst components.

We reemphasize the usefulness of the approach of identifying catalyst functions for particular classes of reactions (determined with individual compounds) and quantification of the catalyst performance data as a basis for identifying improved catalysts. We also recommend investigations of the conversions of individual compounds in whole bio-oils by application of analysis methods like those used for investigations of petroleum conversion; the methods include high-resolution mass spectrometry and gas chromatography. These methods are expensive and time consuming but can yield a wealth of data that would not be characterized by the limitations of those determined with individual compounds and would correctly reflect competitive adsorption and reaction effects that cannot be determined in experiments with individual compounds as feeds.

The reported work reflects a strong emphasis on catalysts that have been well investigated for hydroprocessing of reactants that do not contain oxygen, and the implicit assumption that HDS, HDN, and hydrogenation catalysts may be good choices for HDO is not well tested. The oxygen-containing compounds present in biomass-derived feedstocks have their own reactivities that seemingly demand their own catalysts, and we suggest that these catalysts are yet to be discovered and may be much different from those used in hydroprocessing today. A specific goal of catalyst development might be to identify catalysts that function with minimal H₂ consumption by allowing C–O bond breaking without substantial aromatic ring saturation. Rapid throughput catalyst testing is expected to have a place in future research on HDO.

The optimum catalysts for any application will depend on the processing goals, which could range from production of partially deoxygenated mixtures that could be co-fed with petroleum in petroleum refineries for production of fuels or fuel blending components—or production of chemicals, of which there are many that could be formed from bio-oils. It is still an open question which candidate processing goals may turn out to be practically important, and one realizes that projections of such goals will depend on the discovery of new catalysts. Process conception and catalyst discovery are intertwined (and one should not neglect the truism that catalysts are discovered for needed processes, and processes are almost never found for intriguing new catalysts).

Much remains to be learned about how candidate catalysts function over long times and how their activities, selectivities, and stabilities depend on composition, temperature, pressure, the ratio of H₂ to feedstock, and the presence of minor components such as those containing phosphorus, metals, and other heteroatoms in the bio-oils. The effects of these components vary widely from one catalyst class to another. Systematic variations of these processing variables and long-term testing to determine catalyst lifetimes and performance are needed, but such testing is expensive and time consuming. It might be valuable to apply modified rapid-throughput catalyst testing methods to screen not just for initial activity and selectivity but to test for long-term catalyst performance.

Development of improved catalysts and processes would also benefit from fundamental investigations of reaction networks and kinetics expanded beyond compounds such as phenol, anisole, and guaiacol to larger compounds with more functional groups (such as those reported by Sergeev and Hartwig^44,151 (Fig. 15)) and to reactions of mixtures of compounds that adsorb competitively on catalytic sites—one might begin with pairs of compounds and extend the work to increasingly complex mixtures. It would be valuable in all these experiments to use catalysts with prototypical functions (e.g., noble metals) and to explore a wide range of reaction conditions, especially pressure, as little is known about how H₂ partial pressure affects catalyst performance. Investigations of reaction mechanism would also help in the identification of catalyst functions. Kinetics of HDO would also be helpful in providing a quantitative framework for interpreting data and elucidating reaction mechanisms.

Fig. 15 Compounds investigated by Sergeev and Hartwig with: (A) 4-O–5 linkages, (B) α-O–4 linkages, and (C) β-O–4 linkages in lignin.^44,151

It might be also fruitful to explore reactions that produce H₂ taking place simultaneously with or in concert with HDO to minimize costs of H₂. And research with reactions in ionic liquids and supercritical fluids and even in plasmas is worthy of attention.

Conclusions

Lignin is a major component of biomass that is chemically similar to petroleum but differs in having high oxygen content. Lignin offers the prospect of being an important source of renewable fuels, but production of fuels from lignin requires the removal of much of the oxygen. Removal of the oxygen-containing compounds results from catalytic conversion with H₂ of the compounds formed from lignin, for example, by pyrolysis. The reactions are called hydrodeoxygenation (HDO), and extensive recent research has been done on HDO of various biomass-derived compounds, including those derived from lignin and typified by phenol, anisole, and guaiacol. Depending on the catalyst, the reactions include direct removal of water, hydrogenation of aromatic rings, and migration of methyl groups in the bio-oils compounds. Catalysts with various functions, such as noble metals and metal sulfides typified by MoS₂, catalyze HDO, but side reactions also occur, and the catalysts undergo deactivation by deposition of carbonaceous residues. Much work is needed to determine good catalysts and elucidate how various catalyst functions (such as metals and acidic supports) affect the conversions and product distributions. Recent work with individual compounds representative of bio-oils is helping to elucidate the chemistry, and one may anticipate systematic progress in the identification of catalysts and process conditions for HDO, although it is likely to be slow to determine optimum catalysts and processing conditions for whole bio-oils. Economical processes for conversion of lignin to fuels that are compatible with today's infrastructure are not close at hand, and much research is needed, as recommended in this review.

Abbreviations

ANI	Anisole
Ben	Benzene
CAT	Catechol
CNF	Carbon nanofiber
Cr	Cresol
CYHA	Cyclohexane
CYHAOL	Cyclohexanol
CYHAONE	Cyclohexanone
CYHE	Cyclohexene
DDO	Direct deoxygenation
DME	Demethylation
DMO	Demethoxylation
2,6-DMP	2,6-Dimethylphenol
2EP	2-Ethylphenol
EUG	Eugenol
2EXYHAONE	2-Ethoxycyclohexanone
GUA	Guaiacol
HDO	Hydrodeoxygenation
HYD	Hydrogenation
LHSV	Liquid hourly space velocity
MA	Methylanisole
4MA	4-Methylanisole
MCYHA	Methylcyclohexane
MCYHE	Methylcyclohexene
2MECYHAOL	2-Methoxycyclohexanol
2ME4MEPH	2-Methoxy-4-methylphenol
4MEP	4-Methoxyphenol
MCYHDIOL	1-Methyl-1,2-cyclohexanediol
MP	Methylphenol
4MP	4-Methylphenol
P	Pressure
Ph	Phenol
PCYHA	Propylcyclohexane
4PG	4-Propylguaiacol
4PP	4-Propylphenol
Syr	Syringol
t	Time
T	Temperature
2,4,6-TMP	2,4,6-Trimethylphenol
Tol	Toluene
VA	Vanillyl
WHSV	Weight hourly space velocity
Xol	Xylenol
2,4-Xol	2,4-Xylenol
Xy	Xylene.

Acknowledgements

T. Nimmanwudipong was supported at the University of California, Davis, by a fellowship from Chevron.

References

1. F. K. Forson, E. K. Oduro and E. Hammond-Donkoh, Renewable Energy, 2004, 29, 1135–1145.
2. J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem., Int. Ed., 2007, 46, 7164–7183.
3. A. Németh, M.Sc. thesis, Wageningen University Rabobank, 2009.
4. J. Akhtar and N. A. S. Amin, Renewable Sustainable Energy Rev., 2011, 15, 1615–1624.
5. D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem., 2010, 12, 1493–1513.
6. M. V. Olarte, Ph.D. thesis, Georgia Institute of Technology, Atlanta, 2011.
7. L. Ma, T. Wang, Q. Liu, X. Zhang, W. Ma and Q. Zhang, Biotechnol. Adv., 2012, 30, 859–873.
8. J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552–3599.
9. S. Czernik and A. V. Bridgwater, Energy Fuels, 2004, 18, 590–598.
10. C. E. Wyman, B. E. Dale, R. T. Elander, M. Holtzapple, M. R. Ladisch and Y. Y. Lee, Bioresour. Technol., 2005, 96, 1959–1966.
11. H. L. Chum and R. P. Overend, Prepr. Symp. – Am. Chem. Soc., Div. Fuel Chem., 2004, 49, 798–799.
12. D. Ferdous, A. K. Dalai, S. K. Bej and R. W. Thring, Can. J. Chem. Eng., 2010, 79, 913–922.
13. J. B. Binder, M. J. Gray, J. F. White and Z. C. Zhang, Biomass Bioenergy, 2009, 33, 1122–1130.
14. J. Holmgren, R. Marinageli, P. Nair, D. C. Elliott and R. Bain, Hydrocarbon Process., 2008, 95–103.
15. R. H. Venderbosch, A. R. Ardiyanti, J. Wildschut, A. Oasmaa and H. J. Heeres, J. Chem. Technol. Biotechnol., 2010, 85, 674–686.
16. L. Lin, Z. Cunshan, S. Vittayapadung, S. Xiangqian and D. Mingdong, Appl. Energy, 2011, 88, 1020–1031.
17. M. F. Demirbas, Appl. Energy, 2011, 88, 17–28.
18. H. Wenzel, Breaking the biomass bottleneck of the fossil free society, Technical report, Concito, 2010.
19. P. McKendry, Bioresour. Technol., 2002, 83, 37–46.
20. S. Crossley, J. Faria, M. Shen and D. E. Resasco, Science, 2010, 327, 68–72.
21. P. Grange, E. Laurent, R. Maggi, A. Centeno and B. Delmon, Catal. Today, 1996, 29, 297–301.
22. C. Perego and A. Bosetti, Microporous Mesoporous Mater., 2011, 144, 28–39.
23. J. G. Rogers and J. G. Brammer, Biomass Bioenergy, 2009, 33, 1367–1375.
24. E. Furimsky, Appl. Catal., A, 2000, 199, 144–190.
25. D. Mohan, C. U. Pittman and P. H. Steele, Energy Fuels, 2006, 20, 848–889.
26. D. C. Elliott, Energy Fuels, 2007, 21, 1792–1815.
27. Q. Zhang, J. Chang, T. Wang and Y. Xu, Energy Convers. Manage., 2007, 48, 87–92.
28. R. H. Venderbosch and W. Prins, Biofuels, Bioprod. Biorefin., 2010, 4, 178–208.
29. A. Oasmaa, Y. Solantausta, V. Arpiainen, E. Kuoppala and K. Sipil, Energy Fuels, 2010, 24, 1380–1388.
30. P. M. Mortensen, J. D. Grunwaldt, P. A. Jensen, K. G. Knudsen and A. D. Jensen, Appl. Catal., A, 2011, 407, 1–19.
31. T. V. Choudhary and C. B. Phillips, Appl. Catal., A, 2011, 397, 1–12.
32. Q. Bu, H. Lei, A. H. Zacher, L. Wang, S. Ren, J. Liang, Y. Wei, Y. Liu, J. Tang, Q. Zhang and R. Ruan, Bioresour. Technol., 2012, 124, 470–477.
33. Z. He and X. Wang, Catal. Sustainable Energy Prod., 2013, 28–52.
34. D. D. Laskar, B. Yang, H. Wang and J. Lee, Biofuels, Bioprod. Biorefin., 2013, 7, 602–626.
35. E. Furimsky, Catal. Today, 2013, 217, 13–56.
36. D. C. Elliott and E. G. Baker, Biotechnol. Bioeng. Symp., 1984, 14, 159–174.
37. D. C. Elliott and G. F. Schiefelbein, Abstr. Pap. Am. Chem. Soc., 1989, 34, 1160–1166.
38. D. C. Elliott and G. G. Neuenschwander, Liquid fuels by low-severity hydrotreating of biocrude, in Developments in thermochemical biomass conversion, ed. A. V. Bridgwater and D. G. Boocock, Blackie, London, 1996, vol. 1, pp. 611–621.
39. D. C. Elliott, K. L. Peterson, D. S. Muzatko, E. V. Alderson, T. R. Hart and G. G. Neuenschwander, Appl. Biochem. Biotechnol., 2004, 113–116, 807–825.
40. A. H. Zacher, D. Elliott and D. Santosa, AIChE Annual Meeting, Lake City, 2010.
41. J. Gagnon and S. Kaliaguine, Ind. Eng. Chem. Res., 1988, 27, 1783–1788.
42. Y. H. E. Sheu, R. G. Anthony and E. J. Soltes, Fuel Process. Technol., 1988, 19, 31–50.
43. A. Oasmaa and D. G. B. Boocock, Can. J. Chem. Eng., 1992, 70, 294–300.
44. A. G. Sergeev and J. F. Hartwig, Science, 2011, 332, 439–443.
45. D. C. Elliott and T. R. Hart, Energy Fuels, 2009, 23, 631–637.
46. D. Meier, J. Berns, O. Faix, U. Balfanz and W. Baldauf, Biomass Bioenergy, 1994, 7, 99–105.
47. G. Telysheva, G. Dobele, D. Meier, T. Dizhbite, G. Rossinska and V. Jurkjane, J. Anal. Appl. Pyrolysis, 2007, 79, 52–60.
48. A. Gutierrez, R. K. Kaila, M. L. Honkela, R. Siloor and A. O. I. Krause, Catal. Today, 2009, 147, 239–246.
49. D. Y. Hong, S. J. Miller, P. K. Agrawal and C. W. Jones, Chem. Commun., 2010, 46, 1038–1040.
50. C. Zhao, Y. Kou, A. A. Lemonidou, X. B. Li and J. A. Lercher, Angew. Chem., Int. Ed., 2009, 48, 3987–3990.
51. T. Nimmanwudipong, R. C. Runnebaum, D. E. Block and B. C. Gates, Catal. Lett., 2011, 141, 1072–1078.
52. R. C. Runnebaum, T. Nimmanwudipong, D. E. Block and B. C. Gates, Catal. Lett., 2011, 141, 817–820.
53. R. C. Runnebaum, R. J. Lobo-Lapidus, T. Nimmanwudipong, D. E. Block and B. C. Gates, Energy Fuels, 2011, 25, 4776–4785.
54. T. Nimmanwudipong, R. C. Runnebaum, D. E. Block and B. C. Gates, Catal. Lett., 2011, 141, 779–783.
55. T. Nimmanwudipong, R. C. Runnebaum, D. E. Block and B. C. Gates, Energy Fuels, 2011, 25, 3417–3427.
56. T. Nimmanwudipong, R. C. Runnebaum, S. E. Ebeler, D. E. Block and B. C. Gates, Catal. Lett., 2012, 142, 151–160.
57. X. Zhu, R. G. Mallison and D. E. Resasco, Appl. Catal., A, 2010, 379, 172–181.
58. M. A. Gonzalez-Borja and D. E. Resasco, Energy Fuels, 2011, 25, 4155–4162.
59. E. Dorrestijn, L. J. J. Laarhoven, I. W. C. E. Arends and P. Mulder, J. Anal. Appl. Pyrolysis, 2000, 54, 153–192.
60. A. V. Bridgwater and M. L. Cottam, Energy Fuels, 1992, 6, 113–120.
61. M. V. Bykova, D. Y. Ermakov, V. V. Kaichev, O. A. Bulavchenko, A. A. Saraev, M. Y. Lebedev and V. A. Yakovlev, Appl. Catal., B, 2012, 113–114, 296–307.
62. O. I. Senol, E.-M. Ryymin, T.-R. Viljava and A. O. I. Krause, J. Mol. Catal. A: Chem., 2007, 277, 107–112.
63. A. V. Bridgwater, Biomass Bioenergy, 2012, 38, 68–94.
64. R. W. Thring, S. P. Katikaneni and N. N. Bakhshi, Fuel Process. Technol., 2000, 62, 17–30.
65. S. M. Kadangode, Ph.D. thesis, University of Utah, Salt Lake City, 2001.
66. D. K. Johnson, E. Chornet, W. Zmierczak and J. Shabtai, Prepr. Symp. – Am. Chem. Soc., Div. Fuel Chem., 2002, 47, 380–381.
67. E. Laurent and B. Delmon, Appl. Catal., A, 1994, 109, 77–96.
68. A. Centeno, E. Laurent and B. Delmon, J. Catal., 1995, 154, 288–298.
69. C. E. Burgess, D. J. Clifford and J. R. Horvath, Prepr. Symp. – Am. Chem. Soc., Div. Fuel Chem., 2002, 47, 376–379.
70. R. W. Thring, E. Chornet and R. P. Overend, Can. J. Chem. Eng., 1993, 71, 107–115.
71. L. Liguori and T. Barth, J. Anal. Appl. Pyrolysis, 2011, 92, 477–484.
72. Y. Yang, A. Gilbert and C. Xu, Appl. Catal., A, 2009, 360, 242–249.
73. X. Zhu, L. L. Lobban, R. G. Mallinson and D. E. Resasco, J. Catal., 2011, 281, 21–29.
74. M. J. Girgis and B. C. Gates, Ind. Eng. Chem. Res., 1991, 30, 2021–2058.
75. N. Yan, C. Zhao, P. J. Dyson, C. Wang, L. Liu and Y. Kou, ChemSusChem, 2008, 1, 626–629.
76. N. Yan, Y. Yuan, R. Dykeman, Y. Kou and P. J. Dyson, Angew. Chem., Int. Ed., 2010, 49, 5549–5553.
77. A. L. Jongerius, P. C. A. Bruijnincx and B. M. Weckhuysen, Green Chem., 2013, 15, 3049–3056.
78. V. N. Bui, G. Toussaint, D. Laurenti, C. Mirodatos and C. Geantet, Catal. Today, 2009, 143, 172–178.
79. F. E. Massoth, P. Politzer, M. C. Concha, J. S. Murray, J. Jakowski and J. Simons, J. Phys. Chem. B, 2006, 110, 14283–14291.
80. B. Yoosuk, D. Tumnantong and P. Prasassarakich, Fuel, 2012, 91, 246–252.
81. T. R. Viljava and A. O. I. Krause, Stud. Surf. Sci. Catal., 1997, 106, 343–352.
82. Y. Yunquan, L. He'an, T. Gangsheng, K. J. Smith and T. C. Thian, Chin. J. Chem. Eng., 2008, 16, 733–739.
83. E. O. Odebunmi and D. F. Ollis, J. Catal., 1983, 80, 76–89.
84. E. Furimsky and F. E. Massoth, Catal. Rev. Sci. Eng., 2005, 47, 297–489.
85. S. Echeandia, P. L. Arias, V. L. Barrio, B. Pawelec and J. L. G. Fierro, Appl. Catal., B, 2010, 101, 1–12.
86. W. Wang, Y. Yang, H. Luo, H. Peng and F. Wang, Ind. Eng. Chem. Res., 2011, 50, 10936–10942.
87. C. Zhao, J. He, A. A. Lemonidou, X. Li and J. A. Lercher, J. Catal., 2011, 280, 8–16.
88. E. M. Turpeinen, Ph.D. thesis, Aalto University, 2011.
89. E. Laurent and B. Delmon, Ind. Eng. Chem. Res., 1993, 32, 2516–2524.
90. K. Li, R. Wang and J. Chen, Energy Fuels, 2011, 25, 854–863.
91. C. V. Loricera, B. Pawelec, A. Infantes-Molina, M. C. Alvarez-Galvan, R. Huirache-Acuña, R. Nava and J. L. G. Fierro, Catal. Today, 2011, 172, 103–110.
92. V. N. Bui, D. Laurenti, P. Afanasiev and C. Geantet, Appl. Catal., B, 2011, 101, 239–245.
93. I. T. Ghampson, C. Sepúlveda, R. Garcia, B. G. Frederick, M. C. Wheeler, N. Escalona and W. J. DeSisto, Appl. Catal., A, 2012, 413–441, 478–484.
94. I. T. Ghampson, C. Sepúlveda, R. Garcia, L. R. Radovic, J. L. García Fierro, W. J. DeSisto and N. Escalona, Appl. Catal., A, 2012, 439–440, 111–124.
95. C. R. Lee, J. S. Yoon, Y. W. Suh, J. W. Choi, J. M. Ha, D. J. Suh and Y. K. Park, Catal. Commun., 2012, 17, 54–58.
96. B. S. Gevert, J. E. Otterstedt and F. E. Massoth, Appl. Catal., 1987, 31, 119–131.
97. Y. C. Lin, C. L. Li, H. P. Wan, H. T. Lee and C. F. Liu, Energy Fuels, 2011, 25, 890–896.
98. T. R. Viljava, R. S. Komulainen and A. O. I. Krause, Catal. Today, 2000, 60, 83–92.
99. E. M. Ryymin, M. L. Honkela, T. R. Viljava and A. O. I. Krause, Appl. Catal., A, 2010, 389, 114–121.
100. T. Nimmanwudipong, C. Aydin, J. Lu, R. C. Runnebaum, K. C. Brodwater, N. D. Browning, D. E. Block and B. C. Gates, Catal. Lett., 2012, 142, 1190–1196.
101. H. Y. Zhao, D. Li, P. Bui and S. T. Oyama, Appl. Catal., A, 2011, 391, 305–310.
102. E. Furimsky, J. A. Mikhlin, D. Q. Jones, T. Adley and H. Baikowitz, Can. J. Chem. Eng., 1986, 64, 982–985.
103. R. C. Runnebaum, T. Nimmanwudipong, D. E. Block and B. C. Gates, Catal. Sci. Technol., 2012, 2, 113–118.
104. D. Prochazkova, P. Zamostny, M. Bejblova, L. Cerveny and J. Cejka, Appl. Catal., A, 2007, 332, 56–64.
105. D. C. Elliott and T. R. Hart, Energy Fuels, 2008, 23, 631–637.
106. J. Wildschut, F. H. Mahfud, R. H. Venderbosch and H. J. Heeres, Ind. Eng. Chem. Res., 2009, 48, 10324–10334.
107. C. Zhao, Y. Kou, A. A. Lemonidou, X. Li and J. A. Lercher, Angew. Chem., Int. Ed., 2009, 121, 4047–4050.
108. T. T. Pham, L. L. Lobban, D. E. Resasco and R. G. Mallinson, J. Catal., 2009, 266, 9–14.
109. V. M. L. Whiffen and K. J. Smith, Energy Fuels, 2010, 24, 4728–4737.
110. S. T. Oyama, X. Wang, Y. K. Lee and W. J. Chun, J. Catal., 2004, 221, 263–273.
111. G. de la Puente, A. Gil, J. J. Pis and P. Grange, Langmuir, 1999, 15, 5800–5806.
112. E. Furimsky and F. E. Massoth, Catal. Today, 1999, 52, 381–495.
113. V. N. Bui, D. Laurenti, P. Delichere and C. Geantet, Appl. Catal., B, 2011, 101, 246–255.
114. R. N. Olcese, M. Bettahar, D. Petitjean, B. Malaman, F. Giovanella and A. Dufour, Appl. Catal., B, 2012, 115–116, 63–73.
115. R. Olcesea, M. M. Bettahar, B. Malamanc, J. Ghanbajac, L. Tibavizco, D. Petitjean and A. Dufour, Appl. Catal., B, 2013, 129, 528–538.
116. K. L. Deutsch and B. H. Shanks, Appl. Catal., A, 2012, 447–448, 144–150.
117. M. K. Huuska, Polyhedron, 1986, 5, 233–236.
118. A. R. Ardiyanti, S. A. Khromova, R. H. Venderbosch, V. A. Yakovlev and H. J. Heeres, Appl. Catal., B, 2012, 117–118, 105–117.
119. A. L. Jongerius, R. W. Gosselink, J. Dijkstra, J. H. Bitter, P. C. A. Bruijnincx and B. M. Weckhuysen, ChemCatChem, 2013, 5, 2964–2972.
120. A. L. Jongerius, R. Jastrzebski, P. C. A. Bruijnincx and B. M. Weckhuysen, J. Catal., 2012, 285, 315–323.
121. P. E. Ruiz, B. G. Frederick, W. J. De Sisto, R. N. Austin, L. R. Radovic, K. Leiva, R. García, N. Escalona and M. C. Wheeler, Catal. Commun., 2012, 27, 44–48.
122. W. Wang, Y. Yang, J. Bao and Z. Chen, J. Fuel Chem. Technol., 2009, 37, 701–706.
123. W. Wang, Y. Yang, J. Bao and H. Luo, Catal. Commun., 2009, 11, 100–105.
124. E. Furimsky, J. A. Mikhlin, D. Q. Jones, T. Adley and H. Baikowitz, Can. J. Chem. Eng., 1986, 64, 982–985.
125. A. Vuori, A. Helenius and J. B. Son Bredenberg, Appl. Catal., 1989, 52, 41–56.
126. M. Ferrari, R. Maggi, B. Delmon and P. Grange, J. Catal., 2001, 198, 47–55.
127. V. M. L. Whiffen, K. J. Smith and S. K. Straus, Appl. Catal., A, 2012, 419–420, 111–125.
128. S. T. Oyama, Catal. Today, 1992, 15, 179–200.
129. D. Sajkowski and S. T. Oyama, in ACS Div. Petr. Prepr. 199th ACS Meeting, Boston, 1990.
130. E. J. Shin and M. A. Keane, Ind. Eng. Chem. Res., 2000, 39, 883–892.
131. C. Sepúlveda, N. Escalona, R. García, D. Laurenti and M. Vrinat, Catal. Today, 2012, 195, 101–105.
132. M. V. Bykova, O. A. Bulavchenko, D. Y. Ermakov, M. Y. Lebedev, V. A. Yakovlev and V. N. Parmon, Catal. Ind., 2011, 3, 15–22.
133. X. Zhang, Q. Zhang, T. Wang, L. Ma, Y. Yu and L. Chen, Bioresour. Technol., 2013, 134, 73–80.
134. X. Zhang, T. Wang, L. Ma, Q. Zhang, Y. Yu and Q. Liu, Catal. Commun., 2013, 33, 15–19.
135. Y. Yang, C. Ochoa-Hernández, V. A. O'Shea, P. Pizarro, J. M. Coronado and D. P. Serrano, Appl. Catal., B: Environ., 2014, 145, 91–100.
136. M. Zdražil, Catal. Today, 2003, 86, 151–171.
137. H. Shimada, T. Sato, Y. Yoshimura, J. Haraishi and A. Nishijima, J. Catal., 1988, 110, 275–284.
138. M. J. Ledoux, A. Peter, E. A. Blekkan and F. Luck, Appl. Catal., A, 1995, 133, 321–333.
139. T. Klicpera and M. Zdražil, J. Mater. Chem., 2000, 10, 1603–1608.
140. T. Klicpera and M. Zdražil, Appl. Catal., A, 2001, 216, 41–50.
141. Y. Romero, F. Richard and S. Brunet, Appl. Catal., B, 2010, 98, 213–223.
142. S. J. Hurff and M. T. Klein, Ind. Eng. Chem. Fundam., 1983, 22, 426–430.
143. T. P. Vispute, H. Zhang, A. Sanna, R. Xiao and G. W. Huber, Science, 2010, 330, 1222–1227.
144. M. Badawi, J. Paul, S. Cristol and E. Payen, Catal. Commun., 2011, 12, 901–905.
145. A. Popov, E. Kondratieva, L. Mariey, J. M. Goupil, J. E. Fallah, J. P. Gilson, A. Travert and F. Maugé, J. Catal., 2013, 297, 176–186.
146. J. B. Son Bredenberg, M. Huuska, J. Räty and M. Korpio, J. Catal., 1982, 77, 242–247.
147. R. C. Runnebaum, T. Nimmanwudipong, R. R. Limbo, D. E. Block and B. C. Gates, Catal. Lett., 2012, 142, 7–15.
148. Y. Q. Yang, C. T. Tye and K. J. Smith, Catal. Commun., 2008, 9, 1364–1368.
149. S. Zhang, Y. Yan, R. Zhengwei and L. Tingchen, Energy Sources, 2003, 25, 57–65.
150. W. Baldauf, U. Balfanz and M. Rupp, Biomass Bioenergy, 1994, 7, 237–244.
151. J. C. Hicks and J. Phys, Chem. Lett., 2011, 2, 2280–2287.
152. F. P. Petrocelli and M. T. Klein, Ind. Eng. Chem. Prod. Res. Dev., 1985, 24, 635–641.
153. J. B. S. Bredenberg, M. Huuska and P. Toropainen, J. Catal., 1989, 120, 401–408.
154. M. Huuska and J. Rintala, J. Catal., 1985, 94, 230–238.
155. M. Ferrari, B. Delmon and P. Grange, Carbon, 2002, 40, 497–511.
156. V. A. Yakovlev, S. A. Khromova, O. V. Sherstyuk, V. O. Dundich, D. Y. Ermakov, V. M. Novopashina, M. Y. Lebedev, O. Bulavchenko and V. N. Parmon, Catal. Today, 2009, 144, 362–366.
157. T. Prasomsri, T. Nimmanwudipong and Y. R. Leshkov, Energy Environ. Sci., 2013, 6, 1732–1738.
158. X. Zhang, T. Wang, L. Ma, Q. Zhang, X. Huang and Y. Yu, Appl. Energy, 2013, 112, 533–538.
159. A. Ausavasukhi, Y. Huang, A. T. To, T. Sooknoi and D. E. Resasco, J. Catal., 2012, 290, 90–100.
160. J. Filley and C. Roth, J. Mol. Catal. A: Chem., 1999, 139, 245–252.
161. J. Sun, A. M. Karim, H. Zhang, L. Kovarik, X. S. Li, A. J. Hensley, J. S. McEwen and Y. Wang, J. Catal., 2013, 306, 47–57.
162. Y. Wang, T. He, K. Liu, J. Wu and Y. Fang, Bioresour. Technol., 2012, 108, 280–284.
163. H. Ohta, H. Kobayashi, K. Hara and A. Fukuoka, Chem. Commun., 2011, 47, 12209–12211.
164. A. Gutierrez, R. K. Kaila and A. O. I. Krause, Proc. 14th Int. Congr. Catal., Seoul, Korea, 2008.
165. J. Chang, T. Danuthai, S. Dewiyanti, C. Wang and A. Borgna, ChemCatChem, 2013, 5, 3041–3049.
166. C. Zhao and J. A. Lercher, ChemCatChem, 2012, 4, 64–68.
167. R. K. M. R. Kallury, W. M. Restivo, T. T. Tidwell, D. G. Boocock, A. Crimi and J. Douglas, J. Catal., 1985, 96, 535–543.
168. C. Sepúlveda, N. Escalona, R. García, D. Laurenti and M. Vrinat, Catal. Today, 2012, 195, 101–105.

---