Hydrodeoxygenation of Lignin-Derived Phenols: from Fundamental Studies Towards Industrial Applications

catalysts

Review Hydrodeoxygenation of Lignin-Derived Phenols: From Fundamental Studies towards Industrial Applications

Päivi Mäki-Arvela and Dmitry Yu. Murzin *

Johan Gadolin Process Chemistry Centre, Åbo Akademi University, 20500 Turku/Åbo, Finland; pmakiarv@abo.ﬁ * Correspondence: dmurzin@abo.ﬁ

Received: 7 August 2017; Accepted: 22 August 2017; Published: 7 September 2017

Abstract: Hydrodeoxygenation (HDO) of bio-oils, lignin and their model compounds is summarized in this review. The main emphasis is put on elucidating the reaction network, catalyst stability and time-on-stream behavior, in order to better understand the prerequisite for industrial utilization of biomass in HDO to produce fuels and chemicals. The results have shown that more oxygenated feedstock, selection of temperature and pressure as well as presence of certain catalyst poisons or co-feed have a prominent role in the HDO of real biomass. Theoretical considerations, such as density function theory (DFT) calculations, were also considered, giving scientiﬁc background for the further development of HDO of real biomass.

Keywords: hydrodeoxygenation; bio-oil; guaiacol; lignin; heterogeneous catalyst; fuel

1. Introduction Lignocellulosic biomass, which is abundant, and does not compete with the food chain, has mainly been used for production of pulp and paper via the Kraft process, in which lignin and hemicellulose have been conventionally burnt to generate energy. New transformation processes should be developed in order to utilize the above-mentioned fractions more efficiently. The research in biomass transformation to produce bio-oil via flash pyrolysis [1] and utilize different lignin fractions to produce chemicals and fuels [2,3] has been very intensive in the recent years in part due to the depleting fossil based feedstock and strive towards carbon neutral fuels and chemicals. Further transformation of these fractions, can, however, be demanding, since bio-oil is unstable, exhibiting low pH, whereas lignin is a complex polyphenolic polymer difficult to transform selectively. One of the most important research area in transformation of bio-oil [4] and lignin [2,3] is catalytic hydrodeoxygenation (HDO) during which fuels or suitable blending agents are produced including benzene, toluene as well as cycloalkane and–ketones. Several reviews have already been published on HDO of lignin based phenols [5–15]. In these works, the main emphasis has been on different catalysts [5–8,11], deactivation [6,8,9], mechanism [6,8,10], kinetics [6]. Some specific topics, such as the use of carbide catalysts [12] and production of jet fuel from lignin [7] have been recently reviewed. Bio-oil HDO including the patent literature was reviewed by Zacher et al. in 2014 [13]. The recent trends in HDO of bio-oils have been summarized since this research area is highly important and very many new studies are constantly appearing [16–64]. Although HDO of phenolic compounds has already been investigated during several decades [65,66] and was shown to be a promising method to upgrade bio-oil, it is still challenging. The main challenges are catalyst deactivation, polymerization of phenolic compounds at elevated temperatures and the use of real bio-oils which contain various compounds and catalyst poisons.

Catalysts 2017, 7, 265; doi:10.3390/catal7090265 www.mdpi.com/journal/catalysts Catalysts 2017, 7, 265 2 of 40

In this review, the main emphasis is in depth elucidation of the reaction network in hydrodeoxygenation of phenolic model compounds, especially focusing on which types of model molecules have been used for HDO and comparison of their reactivity. The experimental data will be correlated with the theoretical results using density functional theory and bond dissociation energies for different functional groups. In addition, to bring the HDO process closer to industry, several relevant aspects, such as catalyst deactivation, recycling and reuse have to be properly addressed. Typically, in continuous HDO of phenolic compounds deactivation of catalyst is signiﬁcant, therefore relevant literature has been described. It is also of utmost importance for industrial processes to know exactly the mass balances and in particular formation of various undesired side products including gaseous ones and coke. The aim in this review is to report the main trends in the HDO of lignin derived model molecules, bio-oil and lignin per se. Special attention was put on the effect of temperature, pressure, metal and support selection. In addition, a reaction network during the HDO of model molecules will be discussed and reactivity of different model molecules in a single feed studies over the same catalyst will be compared. Continuous operation, catalyst reuse and deactivation will also be discussed in depth in order to relate the current state-of-art more closely to industrial applicability.

2. Bio-Oil and Lignin Properties and Need for Upgrading

2.1. Bio-Oil Properties Bio-oil is a product, which is most commonly produced via fast pyrolysis, in which the biomass is treated in the temperature range of 300–600 ◦C and short residence time from 1 s to 2 s [15]. In this process biomass is deconstructed giving char as well as liquid, gases. The liquid fraction called bio-oil contains typically a large amount of water, varying in the range of 10–30 wt %, and different chemicals including aldehydes, acids, carbohydrates, phenolics, ketones, alcohols, etc. [15,67]. The composition of bio-oil varies signiﬁcantly depending on the raw material. Bio-oils being typically non-stable, have a high viscosity, acidity (pH = 2.8–3.8) and low energy content or heating value due to high oxygen content [47,64]. Typically, bio-oils can contain up to 40 wt % oxygen [15], requiring upgrading during which different reactions occur, such as cracking, decarbonylation decarboxylation, hydrogenation, hydrocracking and hydrodeoxygenation. The main aim in hydrodeoxygenation is to reduce the oxygen content in bio-oil, especially present in aromatic hydroxyl and methoxy containing molecules. The HDO reactions of these molecules (e.g., guaiacol), are, however, challenging due to their low stability which causes catalyst deactivation [48].

2.2. Structure of Different Lignin Types

Properties of Organosolv, Kraft and Soda Lignin Organosolv lignin can be prepared under mild extraction conditions using different solvents, such as an ethanol/water mixture at 200 ◦C[68]. Sulfur-free organosolv lignin suitable for catalytic transformation [69] has a lower average molecular weight compared to other lignin [1]. The organosolv lignin utilized by Kasakov et al. exhibited seven interconnected monolignols in its structure [2]. On the other hand, kraft and soda lignin is more condensed [1]. This can be explained by the preparation method involving alkaline reactions, including degradation of aryl-ethyl bonds and condensation because of carbanions addition to quinine methide during precipitation [1].

3. HDO of Lignin-Derived Model Compounds, Simulated and Real Bio-Oil and Lignin In the HDO of bio-oil the main target is to decrease the oxygen content of the product and to improve the product heating value. It has been reported that the heating value of, for example, wheat based bio-oil is 21.2 MJ/kg on the dry basis [35] while the corresponding value for yellow poplar based bio-oil is 17.3 MJ/kg dry basis [4]. In the latter work the heating value of the processed Catalysts 2017, 7, 265 3 of 40

◦ Catalystsbio-oil 2017 was, 7 increased, 265 maximally to 22.8 MJ/kg over Ni/SBA-15 catalyst during HDO at 300 C3 under of 42 30 bar hydrogen being 24% better than in the feed. Furthermore, van Krevelan plot showing the O/CH/O molar and O/Cratio molarof different ratio of products different gives products an idea gives of an how idea HDO of how decreases HDO decreases oxygen content oxygen during content theduring process the (Figure process 1). (Figure For example,1). For when example, the light when oil thefraction light in oil wheat fraction based in pyrolysis wheat based oil containing pyrolysis 57%oil containingwater was 57%hydrodeoxygenated water was hydrodeoxygenated with different catalysts with different at 250 catalysts°C under at 80 250 bar◦C hydrogen, under 80 its bar molarhydrogen, O/C ratio its molar decreased O/C ratio to 0.22 decreased using Ni/ZrO to 0.222 using as a catalyst. Ni/ZrO 2Inas another a catalyst. study In anotherwith bio study‐oil from with yellowbio-oil poplar from yellow the O/C poplar ratio the was O/C decreased ratio was from decreased 0.8 to from 0.2 using 0.8 to 0.2Pt/C using as a Pt/C catalyst as a [4]. catalyst When [4 ]. consideringWhen considering the fuel the properties fuel properties after HDO, after HDO, it should it should be kept be kept in inmind mind that that even even cyclohexanol, cyclohexanol, cyclohexanonecyclohexanone and and cyclohexane cyclohexane are are valuable valuable products products which which can can be be blended blended with with petroleum petroleum [70]. [70 ]. Furthermore,Furthermore, for for example, example, cyclohexane, cyclohexane, toluene toluene and and xylene xylene exhibit exhibit suitable suitable vapor vapor pressures pressures and and carboncarbon number number to to be be used used in in gasoline gasoline [62]. [62].

FigureFigure 1. 1. VanVan Krevelan Krevelan plot plot of of the the light light phase phase of of pyrolysis pyrolysis oil oil prepared prepared from from wheat wheat straw straw and and its its ◦ upgradedupgraded products products using using different different catalysts catalysts at at 250 250 °CC under under 80 80 bar bar hydrogen hydrogen [35]. [35].

InIn the the recent recent review review [7] [7 ]results results from from HDO HDO of of model model molecules, molecules, simulated simulated and and real real bio bio-oils‐oils and and ligninlignin were were summarized. summarized. Model Model molecules molecules have have been been selected selected among among the the three three phenols phenols derived derived from from lignin,lignin, namely namely p-coumarylp‐coumaryl alcohol, alcohol, coniferyl coniferyl alcohol alcohol and and sinapyl sinapyl alcohol alcohol (Figure (Figure2). The main2). The emphasis main emphasishas been puthas onbeen reactivity put on of reactivity different molecules,of different the molecules, proposed reactionthe proposed pathways reaction and thermodynamics pathways and thermodynamicsof different reactions. of different reactions. Catalysts 2017, 7, 265 4 of 40 Catalysts 2017, 7, 265 4 of 42

Catalysts 2017, 7, 265HO HO HO 4 of 42

HO HO HO

CH3H3C CH3 O O O

CH3H3C CH3 OH OH O O OH O Figure 2. The main phenols from lignin adapted from reference [7]. Figure 2.OHThe main phenolsOH from lignin adapted fromOH reference [7]. 3.1. Conversion and MainFigure Products 2. The main in the phenols HDO offrom Lignin lignin‐Derived adapted Model from Moleculesreference [7]. 3.1. Conversion and Main Products in the HDO of Lignin-Derived Model Molecules 3.1. ConversionConversion and values Main and Products the main in the products HDO of inLignin HDO‐Derived of lignin Model derived Molecules model molecules are shown Conversionin Tables 1 and values 2 and and Figure the main 3. For products guaiacol, inwhich HDO is of one lignin OF the derived most studied model moleculesmodel molecule, are shown in Tablescontaining1Conversion and 2both and methoxyvalues Figure and3 and. the For hydroxyl main guaiacol, products group, which in it HDO is ispossible oneof lignin OF to derived theobtain most completemodel studied molecules conversion model are molecule, shownunder containingeitherin Tables high both 1 andtemperatures methoxy 2 and Figure and [50] hydroxyl 3.or For at guaiacol,slightly group, lowerwhich it is temperatures possibleis one OF to the obtain under most completestudiedhigh hydrogen model conversion molecule,pressure under either(Tablecontaining high temperatures1, entry both 1) methoxy [46]. [ 50However,] and or at hydroxyl slightly in most group, lower of the temperatures itcases is possible lower conversionto under obtain high complete levels hydrogen have conversion been pressure obtained under (Table 1, either high temperatures [50] or at slightly lower temperatures under high hydrogen pressure entry[26,33,38,58]. 1) [46]. However, Complete in mostdeoxygenation of the cases for lowerguaiacol conversion is also difficult, levels since have these been yieldsobtained vary between [26,33,38 ,58]. (Table 1, entry 1) [46]. However, in most of the cases lower conversion levels have been obtained Complete10–75%, deoxygenation except for a high for HDO guaiacol yield with is also cyclohexane difﬁcult, as sincethe main these product yields obtained vary betweenover Ni/SiO 10–75%,2‐ ZrO[26,33,38,58].2 at 300 °C Complete under 50 deoxygenation bar (Figure 3). forWhen, guaiacol however, is also the difficult, methoxy since group these is in yields position vary 3, between HDO is except10–75%, for a highexcept HDO for a high yield HDO with yield cyclohexane with cyclohexane as the as main the main product product obtained obtained over over Ni/SiO Ni/SiO22-ZrO‐ 2 easier,◦ most probably due to less steric hindrance (Table 1, entry 2). The reasons for this difficulty to at 300deoxygenateZrOC2 at under 300 °C guaiacol 50 under bar 50 (Figureand bar other (Figure3). model When, 3). moleculesWhen, however, however, are discussed the the methoxy methoxy in Section groupgroup 3.3. is is in in position position 3, HDO 3, HDO is is easier,easier, most most probably probably due due to to less less steric steric hindrance hindrance (Table 11,, entry 2). 2). The The reasons reasons for for this this difficulty difﬁculty to to deoxygenatedeoxygenate guaiacol guaiacol and and other other model model molecules molecules are discussed in in Section Section guaiacol 3.3. 3.3 . phenol anisole guaiacol 100 15. 2. 1. phenol anisole 100 15.6. 7. 80 1. 11. 2. 18. 13.6. 7.14. 6080 12.11. 19. 4. 18.20. 3. 13. 14. 60 12. 16. 19. 40 4. 8. 20. Y HDO (%)Y HDO 9. 5. 3. 16. 2040 8.

Y HDO (%)Y HDO 9. 5. 10.

20 10. 0 17.

200 250 300 350 400 450 0 17. Temperature (°C) 200 250 300 350 400 450 Temperature (°C) Figure 3. Hydrodeoxygenation (HDO) yield in phenol, anisole and guaiacol HDO over different

Figurecatalysts 3. Hydrodeoxygenation as a function of temperature. (HDO) yieldNotation: in phenol, For phenol anisole 1. CoNx, and 20 guaiacol bar, cyclohexane HDO over [63], different 2. MoOFigure3/ZrO 3. Hydrodeoxygenation2, 60 bar, benzene [39], (HDO) 3. MoO yield3/ZrO in phenol,2, 60 bar, anisole benzene and [39],guaiacol 4. ReOx/CNFox, HDO over different 50 bar, catalysts as a function of temperature. Notation: For phenol 1. CoNx, 20 bar, cyclohexane [63], cyclohexanecatalysts as a[37], function 5. MoO of3 ,temperature. 1 bar, benzene Notation: [36], for Foranisole phenol 6. ReOx/CNFox, 1. CoNx, 20 bar,50 bar, cyclohexane cyclohexane [63], [37], 2. 2. MoO3/ZrO2, 60 bar, benzene [39], 3. MoO3/ZrO2, 60 bar, benzene [39], 4. ReOx/CNFox, 50 bar, 7.MoO MoO3/ZrO3, 1 bar,2, 60 benzene bar, benzene [36], 8. [39],MoO 33./ZrO MoO2, 603/ZrO bar,2 benzene, 60 bar, [39], benzene 9. Pt‐ Sn/CNF/Inconel,[39], 4. ReOx/CNFox, 1 bar, benzene50 bar, cyclohexane[48],cyclohexane 10. Ni [37‐catalyst,], [37], 5. MoO 5. MoO1.73 ,bar, 13, bar,1 benzenebar, benzene benzene [50], [ 36[36],for], guaiacol: forfor anisole 11. 6. 6.Pt/Alumina ReOx/CNFox, ReOx/CNFox, silicate, 50 bar, 50 30 bar, cyclohexanebar, cyclohexane cyclohexane [37], [37], 7. MoO[49],7. MoO3 ,12. 13, 1Ni/ZrO bar, bar, benzene benzene2, 100 [36],bar, [36 8.main], MoO 8. product MoO3/ZrO23, /ZrO 60cyclohexanol bar,2 ,benzene 60 bar, S [39], = benzene50%, 9. Pt ‐cyclohexaneSn/CNF/Inconel, [39], 9. S Pt-Sn/CNF/Inconel, = 125% bar, [47],benzene 13.

1 bar,Rh/ZrO[48], benzene 10.2 +aluminosilicate,Ni‐catalyst, [48], 10. 1.7 Ni-catalyst, bar, 40 benzenebar, cyclohexane 1.7 [50], bar, for benzene guaiacol:[45], 14. MoO11. [50 ],Pt/Alumina3, for1 bar, guaiacol: cyclohexane, silicate, 11. 30 Pt/Alumina[36],bar, cyclohexane15. Ni/SiO silicate,2‐ 30 bar,ZrO[49], cyclohexane2 , 12.50 bar,Ni/ZrO cyclohexane2, [ 49100], 12.bar, [62], Ni/ZrO main 16. productCoMoS,2, 100 40 bar,cyclohexanol bar, main benzene product S main= 50%, cyclohexanolproduct cyclohexane [46], 17. SS Mo == 50%,225%N, 50 [47], cyclohexane bar, 13.no Rh/ZrO2+aluminosilicate, 40 bar, cyclohexane [45], 14. MoO3, 1 bar, cyclohexane, [36], 15. Ni/SiO2‐ S = 25%HDO [47 [58],], 13. 18. Rh/ZrO Fe/SiO2,2 +aluminosilicate,1 bar, benzene and 40toluene bar, cyclohexane[33], 19. Pt‐Sn/CNF/Inconel, [45], 14. MoO 31, bar, 1 bar, toluene cyclohexane, [48], 20. [36], ZrO2, 50 bar, cyclohexane [62], 16. CoMoS, 40 bar, benzene main product [46], 17. Mo2N, 50 bar, no 15. Ni/SiONi‐catalyst,2-ZrO 1.72, 50bar, bar, benzene cyclohexane and toluene [62 ],[50]. 16. Notation: CoMoS, Inside 40 bar, the benzene area surrounded main product by the [ 46curve,], 17. the Mo 2N, mainHDO products[58], 18. Fe/SiO are cycloalkanes.2, 1 bar, benzene and toluene [33], 19. Pt‐Sn/CNF/Inconel, 1 bar, toluene [48], 20. 50 bar, no HDO [58], 18. Fe/SiO2, 1 bar, benzene and toluene [33], 19. Pt-Sn/CNF/Inconel, 1 bar, Ni‐catalyst, 1.7 bar, benzene and toluene [50]. Notation: Inside the area surrounded by the curve, the toluene [48], 20. Ni-catalyst, 1.7 bar, benzene and toluene [50]. Notation: Inside the area surrounded by main products are cycloalkanes. the curve, the main products are cycloalkanes. Catalysts 2017, 7, 265 5 of 40

Catalysts 2017, 7, x FOR PEER REVIEWTable 1. Single model molecule feed HDO results. Notation: Y denotes yield and S selectivity. X is conversion. 5 of 42

Entry ReactantTable 1. Catalyst Single model molecule T (◦C) feed P (bar) HDO results. Solvent Notation: Y Conversion denotes yield % and (Time, S selectivity. h) X is HDOconversion. Selectivity Note Reference Rh/MCM-36 250 40 decaneConversion % X = 69%, 80 min 29% hydrocarbons [26] Entry Reactant Catalyst T (°C) P (bar) Solvent HDO Selectivity Note Reference (Time, h) Ybenzene+toluene = 71% Co-feed with H2, CO2, Rh/MCMFe/SiO‐36 2 250 40040 1decane -X = 69%, 80 min X = 92%29% hydrocarbons [26] [33] after 21 min time-on-stream CO, H2O Ybenzene+toluene = 71% after 21 Fe/SiO2 400 1 ‐ X = 92% Co‐feed with H2, CO2, CO, H2O [33] X = 11%min time‐on‐stream Yphenol = 8% Mainly phenol and Mo2N 300 50 decalin [38] X = 11% 14,500 sYphenol = 8% Ycatechol = 2.5% catechol Mo2N 300 50 decalin Mainly phenol and catechol [38] 14,500 s Ycatechol = 2.5% Yphenol = 22% Mo2N 300 50 decalin X = 40% Yphenol = 22% [58] Mo2N 300 50 decalin X = 40% Yheavy = 12% [58] Yheavy = 12%

MoMo2C/CNF2C/CNF 300 30040 40dodecane dodecane X = 98% (4 h) X = 98% (4 h)Sphenol = 48% Sphenol = 48% [56] [56] Yphenol = 38% Yphenol = 38% MoMo2C/CNFC/CNF 350 35055 55dodecane dodecane X = 100% X = 100% [54] [54] 2 Ycreo = 13%, 6 h Ycreo = 13%, 6 h WHSV = 4 h−1 WHSV = 4 h−1 Xgua = 100% DODGua = 62% DODGua = 62% OH Ni/ZrONi/ZrO2 250 250100 1001‐octanol 1-octanol Xgua = 100% Trickle bed, poisonsTrickle bed,[47] poisons [47] 2 Xoctanol = 30% after 80 h TOS after 80 h TOS Xoctanol = 30% after 78 h TOS O after 78 h TOS H C YHDO = 50% YHDO = 50% 1 1 3 CoMoSCoMoS 300 30040 ‐ 40 -X = 100% X = 100% H2S in the feed, fixedH bedS in the feed,[46] ﬁxed bed [46] Yphenol = 36% Yphenol = 36% 2 Stoluene = 32% Stoluene = 32% Ni‐catalyst 400 1.7 ‐ - 100% [50] Ni-catalyst 400 1.7 100% Sbenzene = 30% [50] guaiacolguaiacol Sbenzene = 30% 8 h Ni‐SiO2‐ZrO2 300 50 Ycyclohexane = 97 [62] X = 100 8 h Ni-SiO2-ZrO2 300 50 Ycyclohexane = 97 [62] X = 100 SPhenol = 40% Scyclopentanone = 18% Most stable among Pt, Ru, Rh Pt/C 300 1 ‐ X = 87% SPhenol = 40% [31] Scatechol = 5% and Pd/C catalysts, MB = 70% - W/F = 0.30 g cat h−1/gguaScyclopentanone = 18% Most stable among Pt, Pt/C 58– 300 1 X = 87% Temperature programmedRu, Rh and Pd/C [31] Rh/ZrO2 + SiAl 40 decane X = 100% Ycyclohexane = 64% Scatechol = 5% [45] 310 experiment catalysts, MB = 70% −1 Pt/alumina W/F = 0.30 g cat h /ggua 250 30 hexadecane 76 (1 h) Ycyclohexan = 75% [49] silicate Temperature Rh/ZrO2 + SiAl 58–310 40 decaneW/F = 0.067 h−1 X = 100% Ycyclohexane = 64% [45] Ru/C 400 40 ‐ Ybenzene = 69% programmed[16] experiment X = 34% Pt/alumina 250 30 hexadecane 76 (1 h) Ycyclohexan = 75% [49] silicate W/F = 0.067 h−1 Ru/C 400 40 - Ybenzene = 69% [16] X = 34%

Catalysts 2017, 7, 265; doi:10.3390/catal7090265 www.mdpi.com/journal/catalysts Catalysts 2017, 7, 265 6 of 40

Table 1. Cont. Catalysts 2017, 7, 265 6 of 42 CatalystsCatalysts 2017 2017, ,7 7, ,265 265 66 of of 42 42 EntryCatalysts 2017, 7, 265 Reactant Catalyst T (◦C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note6 of 42 Reference OH OHOH OH

2 2 CoNxCoNx 200 20020 20dodecane dodecane X = 100% (2 h) X = 100%Ycyclohexane (2 h) = 95% Ycyclohexane =No 95% sintering No sintering[63] [63] 22 CoNxCoNx 200200 2020 dodecanedodecane XX = = 100% 100% (2 (2 h) h) YcyclohexaneYcyclohexane = = 95% 95% NoNo sintering sintering [63][63] 2 CH3 CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexane = 95% No sintering [63] CHCH33 O CH OO 3 4-methoxyphenol4‐methoxyphenolO 44‐‐methoxyphenolmethoxyphenol 4‐methoxyphenol 25 wt25 % wt % S = 41%, aromatic 2525 wt wt % % 320 3201 1X = 41% XS = = 41% 41%, aromatic compounds [51] [51] MoO25MoO wt3/ZrO %/ZrO 2 320320 11 XX = = 41% 41% SS = = 41%, 41%, aromatic aromatic compounds compounds compounds [51][51] MoOMoO33/ZrO/ZrO3 22 2 320 1 X = 41% S = 41%, aromatic compounds [51] O MoO3/ZrO2 Time‐on‐stream OO TimeTime‐‐onon‐‐streamstreamTime-on-stream Scyclohexane 60,000 s = 20% Scyclohexane = 20% H C Mo2MoC C150 1501.1 ‐ 1.1 - 60,000 s ScyclohexaneScyclohexane = = 20% 20% CO co‐feed CO co-feed[44] [44] HH3 CC O MoMo22CC 2 150150 1.11.1 ‐ ‐ Time60,00060,000‐on‐stream s s Sbenzene = 80% COCO co co‐‐feedfeed [44][44] 33 X = 3.5% X = 3.5%ScyclohexaneSbenzeneSbenzene = = 80% =80% 20% Sbenzene = 80% H C Mo2C 150 1.1 ‐ 60,000 s CO co‐feed [44] 3 3 XX = = 3.5% 3.5% Sbenzene = 80% 3 33 X = 3.5% Yphenol = 38% Yphenol = 38% 3 WHSV = 0.54 hWHSV = 0.54YphenolYphenol h = = 38% 38% Ni catalystNi catalyst 350 3501.7 1.7 WHSVWHSV = = 0.54 0.54 h h YbenzeneYphenol = 38%22% Ybenzene = 22% [50] [50] NiNi catalyst catalyst 350350 1.71.7 WHSVX = 100% = 0.54 h X = 100%YbenzeneYbenzene = = 22% 22% [50][50] anisole Ni catalyst 350 1.7 XX = = 100% 100% YbenzeneYcresol = 27%= 22% Ycresol = 27% [50] anisoleanisole X = 100% YcresolYcresol = = 27% 27% anisole X = 98%, X = 98%, Ycresol = 27% Pt/HPt/H-Beta‐Beta 400 4001 ‐ 1 - XX = = 98%, 98%, Ybenzene = 48% Ybenzene = 48%[24] [24] Pt/HPt/H‐‐BetaBeta 400400 11 ‐ ‐ W/FX = 98%,= 0.2 W/F = 0.2YbenzeneYbenzene = = 48% 48% [24][24] Pt/H‐Beta 400 1 ‐ W/FW/F = = 0.2 0.2 Ybenzene = 48% [24] CH3 W/F = 0.2 CHCH33 OCH3 O OO OO CH3 O O CHCH33 CH 4 3 Rh/MCM‐36 250 40 decane X = 7%, 80 min SHDO = 62% [26] 4 44 Rh/MCMRh/MCMRh/MCM-36‐‐3636 250250 2504040 40decanedecane decane XX = = 7%, 7%, 80 80 min min X = 7%, 80 minSHDOSHDO = = 62% 62% SHDO = 62%[26][26] [26] 4 Rh/MCM‐36 250 40 decane X = 7%, 80 min SHDO = 62% [26] O H C OO HH3 CC O 33 1,3,5H3C‐trimethoxybenzene 1,3,5-trimethoxybenzene1,3,51,3,5‐‐trimethoxybenzenetrimethoxybenzene 1,3,5‐trimethoxybenzeneOH OHOH Ycyclohexane = 9% OOOH YcyclohexaneYcyclohexane = = 9% 9% OOOO H3C CH3 Ycyclohexane = 9% 4 HH3CC OOCHCH3 CoNx 200 20 dodecane X = 44% (2 h) Ycyclohexane = 9% [63] 44 3 3 CoNxCoNx 200200 2020 dodecanedodecane XX = = 44% 44% (2 (2 h) h) [63][63] H3C CH3 4 4 CoNxCoNx 200 20020 20 dodecane X = 44% (2 h) X = 44% (2Yguaiacol h) = 10% [63] dodecane YguaiacolYguaiacol = = 10% 10% [63] syringol Yguaiacol = 10% syringolsyringol Yguaiacol = 10% Catalysts5 2017, 7, 265syringol CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexane = 91% [63]7 of 42 55 syringol CoNxCoNx 200200 2020 dodecanedodecane XX = = 100% 100% (2 (2 h) h) YcyclohexaneYcyclohexane = = 91% 91% [63][63] 5 CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexane = 91% [63] OH CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexane = 91% [63]

5 MoO3/ZrO2, 350 60 n‐decane Ybenzene = 97 [39]

MoO3/ZrO2, 350 60 n-decane Ybenzene = 97 [39]

phenol phenol OH Ycyclohexane = 52% HO 6 CoNx 200 20 dodecane X = 100% (2 h) [63]

Ycyclohexanol = 62%

catechol OH Ycyclohexane = 45%

7 CoNx 200 20 dodecane X = 100% (2 h) [63]

Ycyclohexanol = 43% OH 1,3‐dihydroxybenzene CO co‐feedTemperature OH Mo2C 150 1.2 ‐ toluene [59] programmed reaction CH3 8 h Ni/SiO2‐ZrO2 300 50 Stoluene = 93% [62] 8 X = 100%

8 h Ni/SiO2‐ZrO2 300 50 Stoluene = 93% [62] o‐cresol X = 100% Catalysts 2017, 7, 265 7 of 42 Catalysts 2017, 7, 265 7 of 42 CatalystsCatalysts2017, 20177, 265, 7, 265 7 of 42 7 of 40 OH OH OH

MoO3/ZrO2, 350 60 n‐decane Table 1. Cont. Ybenzene = 97 [39] MoO3/ZrO2, 350 60 n‐decane Ybenzene = 97 [39] MoO3/ZrO2, 350 60 n‐decane Ybenzene = 97 [39] Entry Reactant Catalyst T (◦C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note Reference phenol phenolOH phenol OH Ycyclohexane = 52% Ycyclohexane = 52% HO OH Ycyclohexane = 52% 6 HO CoNx 200 20 dodecane X = 100%Ycyclohexane (2 h) = 52% 6 HO CoNx 200 20 dodecane X = 100% (2 h) [63] [63] 6 CoNx 200 20 dodecane X = 100% (2 h) [63] 6 CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexanol = 62% [63] Ycyclohexanol = 62% Ycyclohexanol = 62% catechol catechol Ycyclohexanol = 62% catecholOH catechol OH Ycyclohexane = 45% OH Ycyclohexane = 45% Ycyclohexane = 45% 7 CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexane = 45% [63] 7 7 CoNxCoNx 200 20020 20 dodecanedodecane X = 100% (2 h) X = 100% (2 h) [63] [63] 7 CoNx 200 20 dodecane X = 100% (2 h) Ycyclohexanol = 43% [63] OH Ycyclohexanol = 43% 1,3‐dihydroxybenzeneOH Ycyclohexanol = 43% OH Ycyclohexanol = 43% 1,3-dihydroxybenzene1,3‐dihydroxybenzene CO co‐feedTemperature OH Mo2C 150 1.2 ‐ toluene [59] 1,3‐dihydroxybenzene COprogrammed co‐feedTemperature reaction OH Mo2C 150 1.2 ‐ toluene CO co-feedTemperature[59] OH CH Mo2C 150 1.2 -8 h tolueneCOprogrammed co‐feedTemperature reaction [59] 3 Mo2C 150 1.2 ‐ toluene programmed[59] reaction CH Ni/SiO2‐ZrO2 300 50 8 h Stoluene = 93% programmed reaction [62] 8 3 X = 100% CH Ni/SiO2‐ZrO2 300 50 8 h 8 h Stoluene = 93% [62] 8 3 X = 100% 8 Ni/SiONi/SiO2‐ZrO2-ZrO2 2 300 30050 50 Stoluene = 93% Stoluene = 93% [62] [62] 8 X = 100%8 h X = 100% Ni/SiO2‐ZrO2 300 50 8 h Stoluene = 93% [62] X = 100% Catalysts 2017, 7, 265 o‐cresol Ni/SiO2‐ZrO2 300 50 8 h 8 h Stoluene = 93% [62]8 of 42 X = 100% o‐cresol Ni/SiONi/SiO2‐ZrO2-ZrO2 2 300 30050 50 Stoluene = 93% Stoluene = 93% [62] [62] o-cresol X = 100% o‐cresol X = 100% OH

9 9 CoNxCoNx 200 20020 20dodecane dodecane X = 100% (2 h) X = 100%Yisopropylcyclohexane (2 h) Yisopropylcyclohexane= 99.2% = 99.2%[63] [63]

H C CH 3 3 4-isopropylphenol4‐isopropylphenol OH

10 CoNx 200 20 dodecane X = 100% (2 h) Yisopropylcyclohexane = 99.9% [63] CH3

CH 3 3‐isopropylphenol OH

Hydrogen 40 1 h 11 Pt/AC(N) 280 water Ypropylcyclo‐hexane = 97% [57] Loade at X = 100% RT

CH 3 4‐propylphenol Hydrogen 40 6 h Ypropylcyclo‐hexane = 65 12 Pt/AC(N) 280 water [57] Loade at X = 100% Ypropylcyclopentane = 5 RT Catalysts 2017,, 7,, 265 8 of 42

Catalysts 2017, 7, 265 8 of 40

9 CoNx 200 20 dodecane X = 100% (2 h) Yisopropylcyclohexane = 99.2% [63]

Table 1. Cont. H C CH 3 3 ◦ Entry4‐isopropylphenolisopropylphenol Reactant Catalyst T ( C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note Reference OH

Yisopropylcyclohexane = 1010 CoNxCoNx 200 20020 20dodecane dodecaneX = 100% (2 h) Yisopropylcyclohexane X = 100% (2 h) = 99.9% [63] [63] CH3 99.9%

CH 3 3-isopropylphenol3‐isopropylphenolisopropylphenol OH

Hydrogen 40 Hydrogen 40 1 h 1 h 1111 Pt/AC(N)Pt/AC(N) 280 280 water water Ypropylcyclo‐hexane = 97% Ypropylcyclo-hexane = 97%[57] [57] Loade atLoade at RT X = 100% X = 100% RT

CH3 Catalysts 2017, 7, 265 9 of 42 4-propylphenol4‐propylphenol Hydrogen OH CH3 40 6 h Ypropylcyclo‐hexane = 65 12 Pt/AC(N) 280 water [57] Loade atHydrogen 40 X = 100% Ypropylcyclopentane = 5 O RT 6 h Ypropylcyclo-hexane = 65 Pt/AC(N) 280 water [57] X = 100% Ypropylcyclopentane = 5

12 Pd/C + HSM‐5 240 50 Loade at RT X = 86%, 4 h Ypropenylcyclohexane = 75% [27]

CH 3 Pd/C + HSM-5 240 50 X = 86%, 4 h Ypropenylcyclohexane = 75% [27] 4-propylguaiacol4‐propylguaiacol

OH CH3 O

Hydrogen 40 6 h Ypropylcyclo‐hexane = 65 13 Pt/AC(N) 280 water [57] Loade at X = 100% Ypropylcyclo‐pentane = 6 O RT

CH 3 4‐(2‐propenal)‐guaiacol

CH3 OH CH3 O O

Hydrogen 40 6 h Ypropylcyclo‐hexane = 58 14 Pt/AC(N) 280 water [57] Loade at X = 97% Ypropylcyclopentane = 10 RT

CH 2 4‐propenalsyringol 15 250 30 hexadecane (1 h) X = 100% Ypropylcyclo‐hexane = 17% [49] CatalystsCatalysts 2017 2017, ,7 7, ,265 265 99 of of 42 42

OHOH CHCH33 OO

Catalysts 2017, 7, 265 9 of 40 Pd/CPd/C + + HSM HSM‐‐55 240240 5050 XX = = 86%, 86%, 4 4 h h YpropenylcyclohexaneYpropenylcyclohexane = = 75% 75% [27][27]

Table 1. Cont. CHCH 33 ◦ Entry44‐‐propylguaiacolpropylguaiacol Reactant Catalyst T ( C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note Reference

OHOH CHCH33 OO

HydrogenHydrogen 4040 Hydrogen 40 66 h h YpropylcycloYpropylcyclo6 h ‐‐hexanehexane = = 65 65 Ypropylcyclo-hexane = 65 131313 Pt/AC(N)Pt/AC(N)Pt/AC(N) 280280 280 waterwater water [57][57] [57] LoadeLoade at atLoade at RT XX = = 100% 100% YpropylcycloYpropylcycloX = 100%‐‐pentanepentane = = 6 6 Ypropylcyclo-pentane = 6 OO RTRT

CHCH 33 4-(2-propenal)-guaiacol44‐‐(2(2‐‐propenal)propenal)‐‐guaiacolguaiacol

CHCH33 OHOH CHCH33 OO OO

HydrogenHydrogen 4040 Hydrogen 40 66 h h YpropylcycloYpropylcyclo6 h ‐‐hexanehexane = = 58 58 Ypropylcyclo-hexane = 58 141414 Pt/AC(N)Pt/AC(N)Pt/AC(N) 280280 280 waterwater water [57][57] [57] LoadeLoade at atLoade at RT XX = = 97% 97% YpropylcyclopentaneYpropylcyclopentaneX = 97% = = 10 10 Ypropylcyclopentane = 10 RTRT

CHCH 22 Catalysts 2017,4-propenalsyringol 474‐,‐propenalsyringol propenalsyringol265 10 of 42 1515 250250 3030 hexadecanehexadecane (1(1 h) h) X X = = 100% 100% YpropylcycloYpropylcyclo‐‐hexanehexane = = 17% 17% [49][49]

Pt/Silica Ypropylcyclo-hexane = 17% OH Ydihydroeugenol = 75% alumina O Pt/Silica alumina 250 30 hexadecane (1 h) X = 100% [49] CH 3 Ydihydroeugenol = 75% 8 h 15

Ni/SiO2‐ZrO2 300 50 Spropylcyclohexane8 h = 98% [62]

Ni/SiO -ZrO 300 50 Spropylcyclohexane = 98% 2 2 X = 100% [62] CH2 X = 100% eugenol O

(5) Smethylcyclo‐hexane = 65% 16 Ni/SiO2‐ZrO2 300 50 dodecane [62] Xvan = 83% Scyclohexane = 31 CH3 O

OH Vanillin Y 2‐methoxy‐4‐methylphenol = Pd/TiO2‐N‐C 150 10 water X = 91% (2 h) [29] 78% Pd‐ carbonaceous 90 5.5 water X = 99% (3 h) Sp‐creosol = 50% [60] microspheres Mo2C/AC 100 10 water X = 100% (3 h) Sp‐creosol = 60% [25] HO

X ≥ 99%

17 ZnCl2 MeOH Y1,4‐dimethoxy‐phenol = 60% [21] CH3 O 8 h OH Vanillyl alcohol Catalysts 2017,, 7,, 265 10 of 42

OH Pt/Silica OH Ydihydroeugenol = 75% alumina O CH3 Catalysts 2017, 7, 265 8 h 10 of 40

Ni/SiO22‐ZrO22 300 50 Spropylcyclohexane = 98% [62] Table 1. Cont. X = 100% CH X = 100% 2 ◦ Entry Reactanteugenol Catalyst T ( C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note Reference O

(5) (5)Smethylcyclo‐hexane = 65%Smethylcyclo-hexane = 65% 16 16 Ni/SiONi/SiO22‐ZrO2-ZrO22 2 300 30050 50dodecane dodecane [62] [62] Xvan = 83% Xvan = 83%Scyclohexane = 31 Scyclohexane = 31 CH3 O OH OH VanillinVanillin

Pd/TiO2-N-C 150 10 water X = 91%Y 2‐methoxy (2 h)‐4‐methylphenol Y 2-methoxy-4-methylphenol = = 78% [29] Pd/TiO22‐N‐C 150 10 water X = 91% (2 h) [29] 78% Pd-carbonaceous 78% 90 5.5 water X = 99% (3 h) Sp-creosol = 50% [60] microspheresPd‐ carbonaceous 90 5.5 water X = 99% (3 h) Sp‐creosol = 50% [60] microspheresMo2C/AC 100 10 water X = 100% (3 h) Sp-creosol = 60% [25] Mo22C/AC 100 10 water X = 100% (3 h) Sp‐creosol = 60% [25] HO X ≥ 99% X ≥ ≥ 99%

[21] 17 17 ZnClZnCl22 2 MeOHMeOH Y1,4‐dimethoxy‐phenol = Y1,4-dimethoxy-phenol60% = 60% [21] CH3 O 8 h 8 h OH Catalysts 2017, 7VanillylVanillyl, 265 alcohol 11 of 42

CH3 8 h 8 h

18 O Ni/SiO2‐ZrO2 300 50 Spropylcyclohexane = 99% [62] [62] 18 Ni/SiO2-ZrO2 300 50 Spropylcyclohexane = 99% CH 3 X = 100% transantholetransanthole X = 100%

OCH3 X = 75% Sphenol = 87% 19 Pt/Al2O3 325 5 W/F = 1 kg H2/O = 23 [61] O Sbenzene = 20% cat.s/mol phenylacetate O

20 CoNx 200 20 dodecane X = 100% Ycyclohexane = 92% [63]

Biphenyl ether O

21 HOHO CoNx 200 20 dodecane C = 100% Ycyclohexane = 99.9% [63] 4,4‐dihydroxydiphenyl ether H C OH 3 O O Y2‐methoxy‐4‐propylphenol = 2 h 22 Zn/Pd/C 150 21 MeOH 85% [21] HO OH X = 100% O Yguaiacol = 85% CH 3 dimer

CatalystsCatalysts 2017 20172017,, , 7 77,, , 265 265265 111111 of ofof 42 4242 Catalysts 2017, 7, 265 11 of 40 CH CHCH333 888 h hh

18 O Ni/SiO2‐ZrO2 300 50 Spropylcyclohexane = 99% [62] 181818 OO Ni/SiONi/SiO2‐2‐‐ZrOZrO2 2 300300300 505050 Table 1. Cont. SpropylcyclohexaneSpropylcyclohexaneSpropylcyclohexane = == 99% 99%99% [62][62][62]

CHCH3 X = 100% 333 XX = = 100% 100% ◦ Entrytransantholetransantholetransanthole Reactant Catalyst T ( C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Note Reference OCH OCHOCH333 X == 75%75% X = 75% Sphenol = 87% 2 3 X = 75% SphenolSphenol = = 87% 87% Sphenol = 87% 2 19 1919 O Pt/AlPt/AlPt/Al222O333 O 325325 32555 5 W/F == 11 kgkg H222/O/O == 2323 H /O[61][61] = 23 [61] 19 OO Pt/Al O 2 3 325 5 W/F = 1 kg Sbenzene = 20% H /O = 23 2 [61] cat.s/molW/F = 1 kg cat.s/molSbenzeneSbenzene = = 20% 20% Sbenzene = 20% cat.s/molcat.s/molcat.s/mol phenylacetatephenylacetatephenylacetatephenylacetate OO

20202020 CoNxCoNxCoNx 200200200 200202020 20dodecanedodecanedodecane dodecane XX = == 100% 100%100% X = 100%YcyclohexaneYcyclohexane = == 92% 92%92% Ycyclohexane = 92%[63][63][63] [63]

Biphenyl ether BiphenylBiphenylBiphenyl ether etherether OOO

HOHO 21212121 HOHHOHHOHOOO CoNxCoNxCoNx 200200200 200202020 20dodecanedodecanedodecane dodecane CC = == 100% 100%100% C = 100%YcyclohexaneYcyclohexane = == 99.9% 99.9%99.9% Ycyclohexane = 99.9%[63][63][63] [63] 4,4-dihydroxydiphenyl4,44,44,4‐‐‐dihydroxydiphenyldihydroxydiphenyldihydroxydiphenyl etheretherether H C HHHCCC OH 3 O OHOHOH 33 OOO O OOO Y2‐‐methoxy‐‐44‐‐propylphenolpropylphenol == 2 h Y2‐methoxy‐4‐propylphenolY2-methoxy-4-propylphenol = 22 Zn/Pd/C 150 21 MeOH 22 h h 2 h 85% [21] 22222222 HO OH Zn/Pd/CZn/Pd/CZn/Pd/C 150150150 150212121 21MeOHMeOH MeOH 85%85%85% = 85% [21][21][21] [21] HHOO OHOH XX = == 100% 100%100% X = 100% O Yguaiacol = 85% OOO YguaiacolYguaiacol = = 85% 85% Yguaiacol = 85% CH 3 CHCH33 dimerdimerdimer

Catalysts 2017, 7, 265 12 of 40

Table 2. Comparative single feed HDO results using one catalyst. Notation: Yi, j is yield in which i is the reactant and j product, X denotes conversion and S selectivity.

Comparison of the Conversion % Catalyst T (◦C) P (bar) Solvent Products Notation Reference Reactivity (Time, h) Xphe = 48% Yphe, benzene = 46% Phenol, anisole, MoO /ZrO 350 60 Xanisole = 55% Yani, benzene = 40% [39] guaiacol 3 2 Trickle bed Xgua = 51% Ygua, phe = 23% Xphe = 29% Yphe, benzene = 44% Phenol, cresol, Xani = 79% Yani, benzene = 31% With phenol the anisole, guaiacol, MoO3 320 1 - Xgua = 74% Ygua, phenol = 31% solvent is [36] diphenyl ether Xcre = 49% Ycre, toluene = 49% mesitylene Xdip = 83% Ydip, benzene = 72% (4 h) Yphe, cyclohexane = 51% Xphe = 90% Ygua, cyclohexane = 65% Guaiacol, anisole, 10ReO /CNFox 300 50 Xanisole = 100% Yani, cyclohexane = 79% [37] phenol, o-cresol x dodecane Xgua = 100% Ycre, methylcyclohexane = 18% Xcre = 44% Ycre, toluene = 16% 50 h Spheac, benzylalcohol = 83% Anisole, time-on-stream Spheac, benzene = 16% Pt/Al O 325 5 H /O = 23.2 [61] phenylacetate 2 3 Xani = 50% Sani, benzylalcohol = 55% 2 Xpheac = 58% Sani, benzene = 12 125 min TOS W/F = 2 g cat/g Y gua, toluene = 63% Anisole, guaiacol Pt-Sn/CNF/Inconel 400 1 react h−1 [48] Y ani, benzene = 38% Xgua = 97% Xani = 67% Catalysts 2017, 7, 265 13 of 40

CatalystsCatalysts 2017 2017, 7, ,7 x, xFOR FOR PEER PEER REVIEW REVIEW 1313 of of 42 42 For methoxylated benzenes, such as anisole, and trimethoxybenzene (Table1, entries 3 and 4), it is easierForFor to methoxylated achievemethoxylated HDO benzenes, benzenes, compared such such to as as guaiacol. anisole, anisole, and and For trimethoxybenzene trimethoxybenzene example, selectivity (Table (Table to 1, anisole1, entries entries 3 HDO 3and and 4), products4), it it isis easier easier to to achieve achieve HDO HDO compared compared to to guaiacol. guaiacol. ◦For For example, example, selectivity selectivity to to anisole anisole HDO HDO products products was 100% at relatively mild temperature of 150 C under atmospheric pressure over Mo2C catalyst was 100% at relatively mild temperature of 150 °C under atmospheric pressure over Mo2C catalyst (Tablewas3, entry 100% 3) at [ relatively44]. mild temperature of 150 °C under atmospheric pressure over Mo2C catalyst (Table(Table 3, 3, entry entry 3) 3) [44]. [44]. Table 3. Homolytic bond dissociation energies for lignin derived model molecules calculated by TableTable 3. 3. Homolytic Homolytic bond bond dissociation dissociation energies energies for for lignin lignin derived derived model model molecules molecules calculated calculated by by B3LYP/6–311 G(d,p) level theory at 320 ◦C in the gas phase adapted from reference [36] and comparison B3LYP/6–311B3LYP/6–311 G(d,p) G(d,p) level level theory theory at at 320 320 °C °C in in the the gas gas phase phase adapted adapted from from reference reference [36] [36] and and withcomparisoncomparison other bond with with dissociation other other bond bond dissociation energies dissociation from energies energies the from literature. from the the literature. literature. Bond Bond dissociationBond dissociation dissociation energies energies energies are are are given in kJ/mol.givengiven in in kJ/mol. kJ/mol.

DiphenylDiphenyl MonolignolMonolignolMonolignol Diphenyl EtherEther EtherO BondBond GuaiacolGuaiacol PhenolPhenol MM‐ M-CresolCresol‐Cresol AnisoleAnisole Anisole O

Ph-OHPh‐OH 472472 481 469 469 469 469 ‐ ‐ ‐ - - - Ph-OCHPh‐OH 384472 397481 469 - 469 - ‐ ‐ ‐ 384 - - Ph‐OCH33 384 397 ‐ ‐ 384 ‐ ‐ Ph‐OCH3 384 397 ‐ ‐ 384 ‐ ‐ Ph-O—CH3 218 205 - - 238 - - Ph‐O—CH3 218 205 ‐ ‐ 238 ‐ ‐ PhPh-OPh‐O—CH3 -218 205 - ‐ ‐ - -238 ‐ ‐ 280 - aliphaticPhPh‐OPh‐OPh C-O ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ - - - -280280 ‐ ‐ 339 [43] aliphaticaliphatic C C‐O‐O ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ 339339 [43] [43]

PhenolPhenolPhenol HDO HDO HDO is is alsois also also giving giving giving highhigh high conversionsconversions conversions and and high high high yields yields yields of of ofcyclohexane cyclohexane cyclohexane at at moderate moderate at moderate temperaturestemperaturestemperatures under under under relatively relatively relatively high high high hydrogenhydrogen hydrogen pressures pressurespressures (Table (Table (Table 1, 1,1 entry ,entry entry 5) 5) 5)[63]. [63]. [ 63On On]. the Onthe other theother other hand, hand, hand, dihydroxylateddihydroxylateddihydroxylated catechol catechol catechol was was was more more more difﬁcult difficult difficult to to deoxygenate deoxygenate deoxygenate compared compared compared to to phenol, to phenol, phenol, since since since this this molecule thismolecule molecule exhibitedexhibitedexhibited sterical sterical sterical restrictions restrictions restrictions and and and strong strong strong adsorption adsorption containing containing two two two hydroxyl hydroxyl hydroxyl groups groups groups in in the the in vicinity vicinity the vicinity of each other (see Section 3.5.5). It was also observed that when the other hydroxyl was in the meta of eachof each other other (see (see Section Section 3.5.5 3.5.5).). It It was was also also observed observed that when when the the other other hydroxyl hydroxyl was was in the in meta the meta position, degree of HDO was much higher (Table 1, entry 7). The products in the HDO of guaiacol, position,position, degree degree of HDO of HDO was was much much higher higher (Table (Table1 1,, entry entry 7). The The products products in inthe the HDO HDO of guaiacol, of guaiacol, phenolphenol andand anisoleanisole areare mainlymainly aromaticaromatic compounds,compounds, suchsuch asas benzenebenzene andand toluenetoluene atat highhigh phenoltemperatures, and anisole whereas are mainly cyclohexane aromatic compounds,can be formed such below as benzene 340 °C and over toluene catalysts at high exhibiting temperatures, temperatures, whereas cyclohexane can be ◦formed below 340 °C over catalysts exhibiting whereas cyclohexane can be formed below 340 C over catalysts exhibiting3 hydrogenation activity, hydrogenationhydrogenation activity, activity, such such as as supported supported Ni, Ni, Rh, Rh, ReOx, ReOx, CoNx CoNx and and MoO MoO 3[36,37,43,62,63]. [36,37,43,62,63]. such as supportedMonooxygenatedMonooxygenated Ni, Rh, o o‐ ReOx,cresol‐cresol with CoNx with a amethyl andmethyl MoO group group3 [36 in in, 37ortho ortho,43, 62position position,63]. is is easy easy to to deoxygenate deoxygenate (Table (Table 1,Monooxygenated1, entry entry 8) 8) [62]. [62]. For For molecules o-cresolmolecules withcontaining containing a methyl three three group oxygenated oxygenated in ortho groups, groups, position such such is as as easy vanillin, vanillin, to deoxygenate maximally maximally only only (Table 1, 2 2 entry65%65% 8) [selectivity 62selectivity]. For moleculestowards towards methylcyclohexane methylcyclohexane containing three was was oxygenated obtained obtained at at groups, 300 300 °C °C under under such 50 as50 bar vanillin,bar over over Ni/SiO Ni/SiO maximally‐2ZrO‐ZrO 2 only catalyst [62]. ◦ 65% selectivitycatalyst [62]. towards methylcyclohexane was obtained at 300 C under 50 bar over Ni/SiO2-ZrO2 catalyst [The62The]. propyl propyl substituted substituted phenols, phenols, including including 4 ‐4isopropyl‐isopropyl‐,‐ ,and and 4 ‐4propylphenol‐propylphenol easily easily underwent underwent hydrogenation and deoxygenation (Table 1, entries 1 and 9–11), whereas 4‐propylguaiacol and 4‐(2‐ Thehydrogenation propyl substituted and deoxygenation phenols, (Table including 1, entries 4-isopropyl-, 1 and 9–11), and whereas 4-propylphenol 4‐propylguaiacol easily and underwent 4‐(2‐ propenyl)‐guaiacol gave slightly lower HDO degree due to presence of three oxygenated substituents hydrogenationpropenyl)‐guaiacol and deoxygenation gave slightly lower (Table HDO1, degree entries due 1 andto presence 9–11), of whereas three oxygenated 4-propylguaiacol substituents and (Table(Table 1, 1, entries entries 12 12 and and 13). 13). Noteworthy Noteworthy is is also also that that 4 ‐4propensyringol‐propensyringol was was transformed transformed not not only only to to 4 ‐4‐ 4-(2-propenyl)-guaiacol gave slightly lower HDO degree due to presence of three oxygenated propylcyclohexane,propylcyclohexane, but but also also to to 4 ‐4propylcyclopentane‐propylcyclopentane indicating indicating that that in in addition addition to to deoxygenation deoxygenation substituents (Table1, entries 12 and 13). Noteworthy is also that 4-propensyringol was transformed andand ring ring hydrogenation hydrogenation in in this this case case ring ring contraction contraction occurred occurred (Table (Table 1, 1, entry entry 14). 14). not onlyFor toFor 4-propylcyclohexane,eugenol eugenol HDO HDO under under 30 30 butbar bar at also at 250 250 to °C °C 4-propylcyclopentane the the main main product product was, was, indicatinghowever, however, dihydroeugenol dihydroeugenol that in addition to deoxygenationshowingshowing difficulties difficulties and ring to hydrogenationto hydrogenate hydrogenate the the in phenyl thisphenyl case ring ring ring at at contractionthese these conditions conditions occurred (Table (Table (Table 1, 1, entry entry1, entry 15) 15) [49], 14).[49], ◦ whereasForwhereas eugenol at at elevated elevated HDO temperature temperature under 30 and bar and pressure at pressure 250 Cthe the the main main main product product product was was 4 was,‐4propylcyclohexane‐propylcyclohexane however, dihydroeugenol [62]. [62]. The The showingsamesame difﬁculties waswas validvalid to forfor hydrogenate HDOHDO ofof vanillin thevanillin phenyl whichwhich ring maximallymaximally at these conditionsgavegave 65%65% (Tableselectivityselectivity1, entry towardstowards 15) [ 49], methylcyclohexane at 300 °C under 50 bar over Ni/SiO2‐ZrO2 catalyst (Table 1, entry 16). The main whereasmethylcyclohexane at elevated temperature at 300 °C under and 50 pressure bar over Ni/SiO the main2‐ZrO product2 catalyst was(Table 4-propylcyclohexane 1, entry 16). The main [62]. product from vanillyl alcohol HDO is 1,4‐dimethoxyphenol using ZnCl2 as a catalyst in methanol The sameproduct was from valid vanillyl for alcohol HDO HDO of is vanillin 1,4‐dimethoxyphenol which maximally using ZnCl gave2 as 65%a catalyst selectivity in methanol towards (Table(Table 1, 1, entry entry 17) 17) [21]. [21]. Transanthole Transanthole◦ contains contains both both propyl propyl and and methoxy methoxy substituents substituents and and it it is is easy easy to to methylcyclohexane at 300 C under 50 bar over Ni/SiO2-ZrO2 catalyst (Table1, entry 16). The main deoxygenatedeoxygenate (Table (Table 1, 1, entry entry 18). 18). Phenylacetate Phenylacetate formed formed in in bio bio‐oil‐oil via via a areaction reaction between between acetic acetic acid acid product from vanillyl alcohol HDO is 1,4-dimethoxyphenol using ZnCl2 3 2 as a catalyst in methanol andand phenol phenol underwent underwent HDO HDO at at 325 325 °C °C under under 5 5bar bar hydrogen hydrogen over over Pt/Al Pt/AlO2O 3catalyst catalyst [61]. [61]. At At elevated elevated (Table1, entry 17) [ 21]. Transanthole contains both propyl and methoxy substituents and it is easy to temperaturetemperature due due to to thermodynamic thermodynamic limitations limitations the the main main product product was was phenol phenol (Figure (Figure 4, 4, Table Table 1, 1, deoxygenateentryentry 19) 19) [61]. (Table[61]. 1, entry 18). Phenylacetate formed in bio-oil via a reaction between acetic acid ◦ and phenol underwent HDO at 325 C under 5 bar hydrogen over Pt/Al2O3 catalyst [61]. At elevated temperature due to thermodynamic limitations the main product was phenol (Figure4, Table1, entry 19) [61].

CatalystsCatalysts 2017 2017, 7, ,7 265;, 265; doi:10.3390/catal7090265 doi:10.3390/catal7090265 www.mdpi.com/jouwww.mdpi.com/journal/catalystsrnal/catalysts Catalysts 2017, 7, 265 14 of 40

Table 4. Comparison of the HDO using mixtures or bio-oil. Notation: X is conversion, Y yield and S selectivity.

Comparison of Catalyst T (◦C) P (bar) Solvent Conversion % (Time, h) HDO Selectivity Reference the Reactivity 25 wt % Guaiacol, (4 h) 25 wt % o-cresol, Xphe = 50% Ycycloalkane = 22% NiFe/H-Beta 300 16 water [40] 50 wt % phenol Xcre = 68% Yarom. HC = 29% mixture Xgua = 88% 78% HDO, reuse, Van Bio-oil Pt/C 350 Supercritical EtOH [30] Krevelan Xphe = 100% Phenol, vanillin, Xtransanethole = 100% Scyclohexane = 98% cresol, eugenol, Ni/SiO -ZrO 300 50 dodecane [62] 2 2 Xeug = 100% Smethylcyclohexane = 93%e trans-anethole Xo-cre = 100% (8 h) Yphenolic = 8 wt % 86% of identiﬁed Bio-oil Ni/SiO -ZrO 300 50 Dodecane Yarom. HC = 8 wt % [62] 2 2 compounds transformed Ycyclic alkanes = 55 wt % to cyclic hydrocarbons (1 h) Ydihydroeugenol = 80% Guaiacol, eugenol Pt/ 250 30 hexadecane Xgua = 88% [49] Ycyclohexane = 85% Xeug = 100% Sbenzene = 66% m-cresol, anisole, 1,2-dimethoxybenzene, Mo2C 280 1.14 - Xmixture = 94% Stoluene = 27% [42] guaiacol Scyclohexane = 5% Yield 13 g/100 g oil in water Bio-oil Ni/SiO 250 80 at 25 ◦C Water 57% [35] 2 and organic phase Bio-oil Pd/C 350 30 ethanol Y = 22 wt % [30] Bio-oil Ni/SBA-15 300 30 ethanol Deoxygenation degree 54.9% [4] CuMgAlOx + NiO/SiO Ylignin monomer = 51% Soda-lignin 2 340 20 ethanol [3] containing Ni2P HDO degree = 84% Organosolv lignin Ni/H-BEA 20 320 hexadecane X = 100%, 6 h Shydrocarbon = 70% [2] CatalystsCatalysts 20172017,, 77,, 265x FOR PEER REVIEW 1515 ofof 4042

Figure 4. Results from kinetic results of the gas-phase HDO of phenyl acetate over Pt/Al2O3 catalyst Figure◦ 4. Results from kinetic results of the gas‐phase HDO of phenyl acetate over Pt/Al2O3 catalyst at 325 C under 5 bar with H2/O ratio of 23 [61]. at 325 °C under 5 bar with H2/O ratio of 23 [61].

DimericDimeric compounds,compounds, such such as as biphenyl biphenyl ether ether and and the correspondingthe corresponding 4,4-dihydroxydiphenyl 4,4‐dihydroxydiphenyl ether canether be can also be easilyalso easily deoxygenated deoxygenated to cycloalkanes to cycloalkanes (Table (Table1, entries 1, entries 20 20 and and 21), 21), whereas whereas a a synthetic synthetic β-‐OO-4‐4 ligninlignin dimer,dimer, guaiacylglycerol-guaiacylglycerol‐β-4-‐4‐OO-guaiacyl‐guaiacyl couldcould onlyonly bebe transformedtransformed toto 2-methoxyphenol2‐methoxyphenol andand guaiacol via via breaking breaking the the ether ether linkage linkage in its in HDO its HDO reaction reaction performed performed with Zn/Pd/C with Zn/Pd/C system ◦ systemunder 21 under bar 21hydrogen bar hydrogen at 150 at °C 150 in Cmethanol in methanol [21]. [ 21The]. The minor minor product product was was 3‐ 3-(3-methoxy-(3‐methoxy‐4‐ 4-hydroxyphenyl)-1-propanol.hydroxyphenyl)‐1‐propanol. Synergy Synergy between between Pd Pd and and Zn Zn promoted promoted the the bond cleavage and a further reactionreaction ofof thethe monomericmonomeric unitsunits onon thethe catalystcatalyst surfacesurface with with spillover spillover hydrogen. hydrogen. 3.2. Comparative Studies of a Single Feed Reactant over the Same Catalysts 3.2. Comparative Studies of a Single Feed Reactant over the Same Catalysts From HDO of different model compounds, comprising guaiacol, anisole, phenol and o-cresol From HDO of different model compounds, comprising guaiacol, anisole, phenol and o‐cresol with a single reactant over the same catalyst it is possible to compare their relative reactivity, with a single reactant over the same catalyst it is possible to compare their relative reactivity, which which is of high interest for industrial applications (Table2)[ 36,37,39,48]. Results over ReOx/CNF is of high interest for industrial applications (Table 2) [36,37,39,48]. Results over ReOx/CNF catalyst catalyst at elevated temperature, 300 ◦C and pressure of 50 bar showed that conversion of four at elevated temperature, 300 °C and pressure of 50 bar showed that conversion of four different model different model molecules decreased in the following order: guaiacol = anisole > phenol > cresol molecules decreased in the following order: guaiacol = anisole > phenol > cresol (Figure 5) [37]. In the (Figure5)[ 37]. In the work of Prasomsri et al. [36] the corresponding reactivity was the following: work of Prasomsri et al. [36] the corresponding reactivity was the following: diphenyl > anisole > diphenyl > anisole > guaiacol > phenol > o-cresol [36]. Thus, dehydroxylation of the phenyl ring is guaiacol > phenol > o‐cresol [36]. Thus, dehydroxylation of the phenyl ring is difficult [37,39,48,61]. difﬁcult [37,39,48,61]. Reactivity of anisole higher than guaiacol reported in [39] similarly to the work Reactivity of anisole higher than guaiacol reported in [39] similarly to the work of Prasomsri et al. of Prasomsri et al. [36], can be explained by a higher bond dissociation energy required for the aromatic [36], can be explained by a higher bond dissociation energy required for the aromatic hydroxyl hydroxyl compared to methoxyl (Table3). For o-cresol the preferred route is demethoxylation followed compared to methoxyl (Table 3). For o‐cresol the preferred route is demethoxylation followed by ring by ring hydrogenation [37]. Only when HDO was performed at 400 ◦C over Pt-Sn/CNF/Inconel hydrogenation [37]. Only when HDO was performed at 400 °C over Pt‐Sn/CNF/Inconel nickel‐ nickel-(Inconel is chromium-based superalloy), guaiacol reactivity was higher than for anisole probably (Inconel is chromium‐based superalloy), guaiacol reactivity was higher than for anisole probably due due to a very high temperature [48]. Biphenyl with an ether bond is easy to break, as experimentally to a very high temperature [48]. Biphenyl with an ether bond is easy to break, as experimentally and and theoretically shown [36]. theoretically shown [36].

CatalystsCatalysts 20172017,, 77,, x 265 FOR PEER REVIEW 1616 of of 42 40

100

Conversion/ Yield (mol-%) 20

0 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min)

(a)

100

Conversion/ Yield (mol-%) 20

0 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min) (b)

Figure 5. Cont.

Catalysts 2017,, 7,, x 265 FOR PEER REVIEW 1717 of of 42 40

100

Conversion/(mol-%) Yield 20

0 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min)

Conversion/ Yield (mol-%) 20

0 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min)

(d)

Figure 5. ComparativeComparative HDO of different model compounds over 10 wt % ReOx/CNFReOx/CNF in in dodecane dodecane at 300300 °C◦C under under 5 5 MPa MPa hydrogen hydrogen adapted adapted from from reference reference [37], [37], HDO of ( (aa)) guaiacol; guaiacol; ( (bb)) phenol; phenol; ( (cc)) anisole anisole and (d) o‐cresol. Notation: (■) conversion, (+) phenol, (x) benzene, (◊) toluene, (o) cyclohexane, (▲) and (d) o-cresol. Notation: () conversion, (+) phenol, (x) benzene, (♦) toluene, (o) cyclohexane, methoxycyclohexane, () cyclohexanol, () cyclohexene, (●) lights. (N) methoxycyclohexane, (∆) cyclohexanol, (∇) cyclohexene, ( ) lights.

In a mixture of guaiacol and eugenol a slightly higher conversion levels were obtained for HDO In a mixture of guaiacol and eugenol a slightly higher conversion levels were obtained with the latter molecule at 250 °C under 30 MPa over Pt/C and Pt/aluminosilicate catalysts [49]. The for HDO with the latter molecule at 250 ◦C under 30 MPa over Pt/C and Pt/aluminosilicate main product with Pt/SA was 2‐methoxy‐4‐propylphenol, whereas 2‐methoxy‐4‐propylcyclohexanol catalysts [49]. The main product with Pt/SA was 2-methoxy-4-propylphenol, whereas 2-methoxy-4- was mainly generated over Pt/C [49]. It is also interesting that when dehydroxylation occurred over propylcyclohexanol was mainly generated over Pt/C [49]. It is also interesting that when an acidic Pt supported on silica alumina, the reaction did not proceed to ring hydrogenation contrary dehydroxylation occurred over an acidic Pt supported on silica alumina, the reaction did not proceed to Pt/C. to ring hydrogenation contrary to Pt/C. Temperature is one of the main parameters affecting product distribution in the HDO of phenolic mixtures. This was visible, for example, in the HDO of mixtures at high temperature, above 350 °C, when aromatic compounds were the main products [39,48], whereas below this temperature

Catalysts 2017, 7, 265 18 of 40

Temperature is one of the main parameters affecting product distribution in the HDO of phenolic mixtures. This was visible, for example, in the HDO of mixtures at high temperature, above 350 ◦C, when aromatic compounds were the main products [39,48], whereas below this temperature non-aromatic cyclic hydrocarbons are also formed [36,37]. For example, in guaiacol HDO, phenol was the main product in accordance with a higher bond dissociation energy for Ar-OH compared to that of Ar-OCH3 [39] (Table3). In case of o-cresol HDO the main products were methylcyclohexane and toluene requiring dehydration with a high bond dissociation energy for Ar–OH probably also explaining a low conversion of o-cresol. In the gas-phase HDO of a mixture of model compounds, m-cresol, anisole, dimethoxybenzene ◦ and guaiacol at 280 C under 114 kPa over Mo2C catalyst, phenol was seen to decrease with an increase of conversion. Two other main products were benzene and toluene. These results indicated that aromatic –OCH3 bond was preferentially cleaved from guaiacol prior to the cleavage of aromatic –OH group [42]. These results were in accordance with the computational chemistry calculations of the bond dissociation energies being for Ar-OCH3 and Ar-OH in guaiacol equal to 376 kJ/mol and 456 kJ/mol [36].

3.3. Reaction Pathways and the Thermodynamic Feasibility of Different Reactions during HDO of Model Compounds Since several different reactions occur during HDO of model compounds one of the main aims in using them is to elucidate the reaction pathways during HDO [32,37,48–50]. In addition to kinetic studies, also adsorption studies [48], theoretical DFT (density function theory) calculations [22,71–75] and bond dissociation energy calculations [36] have been performed.

3.3.1. Kinetic Studies Elucidating the Reaction Pathways during HDO of Model Compounds Guaiacol hydrodeoxygenation can proceed via either demethoxylation pathway forming in the first step phenol at 300 ◦C under 50 bar hydrogen over ReOx/CNF (Figures5 and6)[ 37] and mainly phenol over Ni-catalysts at 350 ◦C (Figure7)[ 50] or alternatively via dehydroxylation pathway forming first phenol and subsequently cyclohexanol and finally cyclohexane at moderate temperature, when hydrogenation is thermodynamically feasible, e.g., over Pt/silica aluminate at 250 ◦C under 30 bar hydrogen [49] and above 250 ◦C over an Rh catalyst (Figure8)[ 45]. Phenol was also the main product in guaiacol HDO at 400 ◦C under atmospheric conditions over Co/Al-MCM-41 [34]. These results clearly show that both the metal type and reaction conditions, i.e., pressure and temperature, are decisive for the product distribution, for example, towards ring hydrogenation or HDO products. On acidic catalysts catechol is formed as the major product from guaiacol and anisole having low reactivity for further transformations [32]. Substantial amounts of catechol were also formed over Pt/C catalyst in HDO of guaiacol at 300 ◦C under atmospheric pressure (Figure9)[ 22]. Further reaction is very slow due to strong adsorption on the catalyst surface, in line with DFT calculations (see Section 3.3.2)[22]. It is also interesting to note that catechol formation was more prominent in HDO of guaiacol over Ru/C catalyst at 350 ◦C than at 400 ◦C, which was probably also related to temperature dependence of adsorption, namely low desorption of catechol at lower temperature [16]. According to a DFT study for Ru(0001) surface it was concluded that catecholate is formed prior to phenolate, finally yielding phenol as the main product [74]. Catalysts 2017, 7, 265 19 of 40

Catalysts 2017, 7, x FOR PEER REVIEW 19 of 42

CH3

CH 12 3

+CH4 ME

OH CH3 CH3 CH ME 3 +H2 +3 H2

‐ H2O HYD +CH 4 ‐ H2 9 10 11

OH OH OH CH3 DMO +5 H O +H2 2 + H2 ‐ H2O ‐ CH3OH HYD 4 5 1 2 3 +H2 ‐ CH3OH +3 H2 DMO

7 +H +H2 ‐ H2O 2 CH CH3 3 +H2 O O +3 H2 HYD DDO 6 8

Figure◦ 6. Reaction scheme for guaiacol HDO at 300 °C under 50 bar hydrogen over ReOx/CNF catalyst Figure 6. Reaction scheme for guaiacol HDO at 300 adaptedC under from reference 50 bar [37]. hydrogen Notation: 1. guaiacol, over ReOx/CNF 2, phenol, 3. cyclohexanol, catalyst 5. cyclohexene, adapted 5, from reference [37]. Notation: 1. guaiacol, 2, phenol, 3. cyclohexanol, 5. cyclohexene, 5, cyclohexane, 6. methoxybenzene,cyclohexane, 6. methoxybenzene, 7. benzene, 7. benzene, 8.8. methoxycyclohexanol, methoxycyclohexanol, 9. O‐cresol, 10. 9. toluene, O-cresol, 10. toluene, methylcyclohexne and 11. xylene. methylcyclohexne and 11. xylene. DMO denotes demethoxylation, DDO direct deoxygenation, ME DMO denotes demethoxylation, DDO direct deoxygenation,methanation and ME HYD methanation hydrogenation. and HYD hydrogenation. Catalysts 2017, 7, x FOR PEER REVIEW 15 of 42

Figure 7. Proposed reactionFigure 7. network Proposed reaction for network guaiacol for guaiacol hydrodeoxygenation hydrodeoxygenation over Ni‐based catalysts over at 350 °C Ni-based [50]. catalysts at 350 ◦C[50].

Catalysts 2017, 7, x FOR PEER REVIEW 15 of 42 Catalysts 2017, 7, 265 20 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 15 of 42

Figure 8. Guaiacol HDO over bifunctional metal supported acidic catalysts [45]. FigureFigure 8.8. GuaiacolGuaiacol HDOHDO overover bifunctionalbifunctional metalmetal supportedsupported acidicacidic catalystscatalysts [ [45].45].

◦ Figure 9. Kinetic for guaiacol HDO over Pt/C catalyst at 300 C under atmospheric pressure [22]. Figure 9. Kinetic for guaiacol HDO over Pt/C catalyst at 300 °C under atmospheric pressure [22]. InFigure case of 9. anisole Kinetic for the guaiacol reaction HDO rate over is very Pt/C high catalyst due at to300 demethoxylation °C under atmospheric occurring pressure more [22]. easily In case of anisole the reaction rate is very high due to demethoxylation occurring more easily for for anisole compared to guaiacol, most probably due to a strong adsorption of the hydroxyl group anisole compared to guaiacol, most probably due to a strong adsorption of the hydroxyl group loweringIn case demethoxylation of anisole the reaction rate of guaiacol rate is very [37 ].high Phenol due dehydroxylationto demethoxylation is rather occurring slow more due toeasily a high for lowering demethoxylation rate of guaiacol [37]. Phenol dehydroxylation is rather slow due to a high bondanisole dissociation compared energyto guaiacol, required most for probably phenolic due hydroxyl to a strong group asadsorption already mentioned. of the hydroxyl In addition, group bond dissociation energy required for phenolic hydroxyl group as already mentioned. In addition, dehydroxylationlowering demethoxylation rate is diminished rate of guaiacol for methyl [37]. substituted Phenol dehydroxylation phenols due to is steric rather hindrance. slow due to a high bonddehydroxylation dissociation rateenergy is diminished required for for phenolic methyl substitutedhydroxyl group phenols as alreadydue to steric mentioned. hindrance. In addition, In eugenol HDO the following reaction network was proposed over Pd supported on Al2O3, dehydroxylationIn eugenol rateHDO is thediminished following for reaction methyl◦ substitutednetwork was phenols proposed due overto steric Pd hindrance.supported on Al2O3, SiO2, C and alumina silicate catalysts at 250 C under 30 bar hydrogen in hexadecane (Figure 10): SiO2, C and alumina silicate catalysts at 250 °C under 30 bar hydrogen in hexadecane (Figure 10): hydrogenationIn eugenol forming HDO the 2-methoxy-4-propylcyclohexanol following reaction network was followedproposed byover its Pd demethoxylation supported on Al and2O3, hydrogenation forming 2‐methoxy‐4‐propylcyclohexanol followed by its demethoxylation and hydrolysisSiO2, C and forming alumina 2-hydroxy-4-propylcyclohexanol silicate catalysts at 250 °C under and 30 methanol bar hydrogen [49]. Over in hexadecane acidic sites dehydration (Figure 10): hydrolysis forming 2‐hydroxy‐4‐propylcyclohexanol and methanol [49]. Over acidic sites ofhydrogenation hydroxypropylcyclohexanol forming 2‐methoxy can happen‐4‐propylcyclohexanol with further hydrogenation followed toby the its corresponding demethoxylation saturated and dehydration of hydroxypropylcyclohexanol can happen with further hydrogenation to the compound.hydrolysis Informing addition, 2‐hydroxy dealkylation‐4‐propylcyclohexanol can occur on the acidicand sitesmethanol causing [49]. cleavage Over ofacidic the propyl sites corresponding saturated compound. In addition, dealkylation can occur on the acidic sites causing chain,dehydration ﬁnally formingof hydroxypropylcyclohexanol cyclohexanol. These results can show happen also difﬁculties with further of deoxygenation hydrogenation over to Pd/C. the cleavage of the propyl chain, finally forming cyclohexanol. These results show also difficulties of Incorresponding addition, HDO saturated of 2-methyl-4-propylphenol compound. In addition, was testeddealkylation using a can mixture occur of on Pd/C–H-ZSM-5 the acidic sites giving causing a deoxygenation over Pd/C. In addition, HDO of 2‐methyl‐4‐propylphenol was tested using a mixture relativelycleavage of high the conversion propyl chain, and yieldfinally for forming mainly cyclohexanol. propylcyclohexene These [ 27results] showing show that also some difficulties acidity isof of Pd/C–H‐ZSM‐5 giving a relatively high conversion and yield for mainly propylcyclohexene [27] neededdeoxygenation for HDO. over In eugenol Pd/C. In HDO addition, the best HDO selectivity of 2‐methyl towards‐4‐propylphenol the desired product was tested was using achieved a mixture using showing that some acidity is needed for HDO. In eugenol HDO the best selectivity towards the Pd/Cof Pd/C–H and H-ZSM-5‐ZSM‐5 giving (Figure a relatively11). HDO high activity conversion increased and with yield acidity for mainly increase. propylcyclohexene The kinetic diameter [27] showing that some acidity is needed for HDO. In eugenol HDO the best selectivity towards the

Catalysts 2017, 7, x FOR PEER REVIEW 16 of 42 Catalysts 2017, 7, 265 21 of 40 desired product was achieved using Pd/C and H‐ZSM‐5 (Figure 11). HDO activity increased with of eugenolacidity is, however, increase. smallerThe kinetic than diameter the pore of size eugenol of H-ZSM-5. is, however, Alkali smaller leached than H-ZSM-5 the pore exhibiting size of H‐ZSM‐5. mesoporesAlkali was leached used in H eugenol‐ZSM‐5 HDOexhibiting giving mesopores 60% selectivity was used to hydrocarbons in eugenol HDO at 240 giving◦C under 60% 50 selectivity bar to hydrogenhydrocarbons [27]. It was at stated 240 °C by under Zhang 50 et bar al. [hydrogen27] that the [27]. key It factorswas stated inhibiting by Zhang transformations et al. [27] that of the key polysubstitutedfactors inhibiting phenols aretransformations sterical hindrance, of polysubstituted for example, phenols the bulkiness are sterical of the hindrance, alkyl substituent for example, in the the aromaticbulkiness ring. of the alkyl substituent in the aromatic ring.

Figure 10. Reaction scheme for eugenol HDO over various Pd catalysts supported on Al2O3, Figure 10. Reaction scheme for eugenol HDO over various◦ Pd catalysts supported on Al2O3, SiO2, SiO2, carbon and aluminosilicate under 30 bar hydrogen at 250 C in hexadecane as a solvent adaptedcarbon from reference and aluminosilicate [49]. Notation: under 1. 30 Eugenol, bar hydrogen 2. Isoeugenol, at 250 °C in 3. hexadecane 2-methoxy-4-propylphenol, as a solvent adapted from 4. 2-methoxy-4-propylcyclohexanol,reference [49]. Notation: 1. 5.Eugenol, 3-propylmethoxycyclohexane, 2. Isoeugenol, 3. 2‐methoxy 6. 4-methyl-2-methoxyphenol,‐4‐propylphenol, 4. 2‐methoxy‐4‐ 7. 4-propylphenol,propylcyclohexanol, 8. 4-propylcyclohexanol 5. 3‐propylmethoxycyclohexane, and 9. propylcyclohexane. 6. 4‐methyl‐2‐methoxyphenol, 7. 4‐ propylphenol, 8. 4‐propylcyclohexanol and 9. propylcyclohexane. In vanillin HDO full deoxygenation was not achieved at 150 ◦C under 10 bar hydrogen in In vanillin HDO full deoxygenation was not achieved at 150 °C under 10 bar hydrogen in water water over mesoporous Pd/TiO2-CNx giving the main product 2-methoxy-4-methylphenol with 78% over mesoporous Pd/TiO2‐CNx giving the main product 2‐methoxy‐4‐methylphenol with 78% selectivity at 91% conversion [29]. The reaction proceeded via hydrogenation of the aldehyde group selectivity at 91% conversion [29]. The reaction proceeded via hydrogenation of the aldehyde group followed by dehydroxylation of the formed alcohol (Figure 12). The pressure applied was not sufﬁcient followed by dehydroxylation of the formed alcohol (Figure 12). The pressure applied was not to hydrogenate the phenyl ring while acidity was not enough to catalyze deoxygenation. sufficient to hydrogenate the phenyl ring while acidity was not enough to catalyze deoxygenation. Cleavage of lignin β-O-4-link model molecules has been experimentally investigated over Cleavage of lignin◦ ‐O‐4‐link model molecules has been experimentally investigated over CuO/γ-Al2O3 catalyst at 150 C under 25 bar hydrogen. The experimental results have been conﬁrmed CuO/‐Al2O3 catalyst at 150 °C under 25 bar hydrogen. The experimental results have been confirmed by calculating the bond dissociation energies of the corresponding compounds [53]. by calculating the bond dissociation energies of the corresponding compounds [53].

Catalysts 2017, 7, 265 22 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 17 of 42

Catalysts 2017, 7, x FOR PEER REVIEW 17 of 42

Figure 11. Conversion and product selectivity in HDO of eugenol at 240 ◦C under 50 bar hydrogen Figure 11. Conversion and product selectivity in HDO of eugenol at 240 °C under 50 bar hydrogen afterFigure 4 h over 11. Pd/C Conversion catalyst and together product with selectivity HZSM-5 in HDO with of different eugenol acidityat 240 °C [27 under]. 50 bar hydrogen afterafter 4 h 4over h over Pd/C Pd/C catalyst catalyst together together with HZSM HZSM‐5‐ 5with with different different acidity acidity [27]. [27].

Figure 12. Reaction scheme for HDO of vanillin under 10 bar hydrogen in water at 70–150 ◦C over Figure 12. Reaction scheme for HDO of vanillin under 10 bar hydrogen in water at 70–150 °C over Pd/TiOPd/TiO2 adapted2 adapted from from reference reference [29 [29].]. Notation Notation ((a) vanillin; ( (bb) )4 4-hydroxymethyl-2-methoxyphenol‐hydroxymethyl‐2‐methoxyphenol Figure 12. Reaction scheme for HDO of vanillin under 10 bar hydrogen in water at 70–150 °C over and (cand) 2-methoxy-4-methylphenol. (c) 2‐methoxy‐4‐methylphenol. Pd/TiO2 adapted from reference [29]. Notation (a) vanillin; (b) 4‐hydroxymethyl‐2‐methoxyphenol and (c) 2‐methoxy‐4‐methylphenol. 3.3.2.3.3.2. Different Different Reactions Reactions and and Their Their Thermodynamic Thermodynamic Feasibility in in HDO HDO of ofModel Model Compounds Compounds 3.3.2.Thermodynamic DifferentThermodynamic Reactions feasibility feasibility and Their of of different differentThermodynamic reactions reactions occurring occurringFeasibility during during in HDO HDO HDO of of Model lignin of lignin Compoundsbased based model model molecules together with the specific catalyst properties are summarized below. During HDO several molecules together with the speciﬁc catalyst properties are summarized below. During HDO several reactions,Thermodynamic such as demethoxylation, feasibility of different hydrogenation, reactions transalkylation, occurring during hydrogenolysis, HDO of demethylation,lignin based model reactions, such as demethoxylation, hydrogenation, transalkylation, hydrogenolysis, demethylation, moleculesring opening, together decarbonylation, with the specific oligomerisation catalyst properties can occur. are summarized below. During HDO several ring opening, decarbonylation, oligomerisation can occur. reactions,Demethoxylation such as demethoxylation, is one of the mainhydrogenation, desired reactions transalkylation, for HDO occurring hydrogenolysis, typically at demethylation,relatively Demethoxylation is one of the main desired reactions for HDO occurring typically at relatively ringhigh opening, temperature decarbonylation, prior to dehydroxylation oligomerisation during can guaiacoloccur. HDO [50]. Demethoxylation of 1,3,5‐ high temperature prior to dehydroxylation during guaiacol HDO [50]. Demethoxylation of trimethoxybenzeneDemethoxylation followed is one of by the ring main hydrogenation desired reactions occurred for over HDO Rh/MCM occurring‐36 at typically250 °C under at relatively 40 1,3,5-trimethoxybenzenebar hydrogen in decane followed [26]. It was by ring concluded hydrogenation in [34] that occurred demethoxylation over Rh/MCM-36 was catalyzed at 250by metal◦C under high temperature prior to dehydroxylation during guaiacol HDO [50]. Demethoxylation of 1,3,5‐ 40 barsites hydrogen using Co in‐ decaneand Ni‐catalysts [26]. It was supported concluded with inAl [‐34MCM] that‐41 demethoxylation at 400 °C under atmospheric was catalyzed pressure by metal trimethoxybenzene followed by ring hydrogenation occurred over Rh/MCM‐36 at 250 °C under 40 sites using[34]. Cyclopentanone Co- and Ni-catalysts formation supported occurred with during Al-MCM-41 HDO of guaiacol at 400 over◦C under Pt/C catalyst atmospheric at 300 °C pressure under [34]. bar hydrogen in decane [26]. It was concluded in [34] that demethoxylation was catalyzed by metal Cyclopentanone formation occurred during HDO of guaiacol over Pt/C catalyst at 300 ◦C under sites using Co‐ and Ni‐catalysts supported with Al‐MCM‐41 at 400 °C under atmospheric pressure [34]. Cyclopentanone formation occurred during HDO of guaiacol over Pt/C catalyst at 300 °C under

Catalysts 2017, 7, 265 23 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 18 of 42 atmosphericatmospheric pressure pressure [31 ].[31]. The The proposed proposed reaction reaction network network for for guaiacol guaiacol HDO HDO proposed proposed over over Pt/C Pt/C is is shownshown in Figure in Figure 13a. 13a. Cyclopentanone Cyclopentanone was was formed formed viavia hydroxycycloalkene intermediate intermediate and andits ring its ring openingopening [41]. [41]. Slow Slow demethoxylation demethoxylation of of guaiacol guaiacol observed observed experimentally experimentally was was in agreement in agreement with with DFT DFT calculationscalculations for HDOfor HDO on on Pt(111) Pt(111) surface surface (Figure (Figure 1313b)b) [22] [22] leading leading to to formation formation of large of large amounts amounts of of catechol. Formation of cyclopentanone was confirmed by DFT calculations made for guaiacol HDO catechol. Formation of cyclopentanone was conﬁrmed by DFT calculations made for guaiacol HDO over Pt(111). This product can be formed by partial hydrogenation of the aromatic ring in guaiacol over Pt(111). This product can be formed by partial hydrogenation of the aromatic ring in guaiacol followed by formation of alpha diketone which undergoes decarbonylation and ring closing [22]. followedAnalogous by formation results were of alpha reported diketone by Lee which et al. [68] undergoes who performed decarbonylation DFT calculations and ring for closing Pt(111) [22]. Analogoussurface. results were reported by Lee et al. [68] who performed DFT calculations for Pt(111) surface.

(a)

(b)

Figure 13. (a) Reaction scheme for HDO of guaiacol over Pt/C catalyst in the temperature range of Figure 13. (a) Reaction scheme for HDO of guaiacol over Pt/C catalyst in the temperature range 275–325 °C under atmospheric pressure over Pt/C and (b) energy levels for reaction coordinate of 275–325 ◦C under atmospheric pressure over Pt/C and (b) energy levels for reaction coordinate calculated with DFT for Pt(111) surface [22]. calculated with DFT for Pt(111) surface [22]. Demethoxylation of dihydroeugenol resulting in formation of 4‐propylphenol requires acidity Demethoxylation(Figure 10) [49]. In temperature of dihydroeugenol programmed resulting HDO of in guaiacol formation at 280 of °C 4-propylphenol over Rh/ZrO2 together requires with acidity ◦ (Figurealuminosilicate 10)[49]. In temperatureunder 40 bar programmedhydrogen in decane HDO ofcomplete guaiacol demethoxylation at 280 C over of Rh/ZrO guaiacol2 togetherwas withaluminosilicate under 40 bar hydrogen in decane complete demethoxylation of guaiacol was achieved at 280 ◦C showing clearly its easiness for breaking this bond (Figure 14) (Table3)[ 45]. It was also stated in [37] that ReOx/CNF is a promising catalyst favoring C-O bond cleavage and forming Catalysts 2017, 7, x FOR PEER REVIEW 19 of 42

achieved at 280 °C showing clearly its easiness for breaking this bond (Figure 14) (Table 3) [45]. It was Catalysts 2017, 7, 265 24 of 40 also stated in [37] that ReOx/CNF is a promising catalyst favoring C‐O bond cleavage and forming phenol at 300 °C under 50 bar hydrogen (Figures 5 and 6) without breaking C‐C bond being thus phenolhydrogen at 300efficient◦C under and carbon 50 bar hydrogenatom economical (Figures catalyst.5 and6) Demethoxylation without breaking was C-C faster bond compared being thus to hydrogentransalkylation efﬁcient in andcontinuous carbon atomHDO economicalof a phenylacetate catalyst. and Demethoxylation anisole mixture was at 325 faster °C compared under 5 bar to transalkylationhydrogen over in Pt/C continuous catalyst HDO [61] ofand a phenylacetatethus it was proposed and anisole that mixture anisole at 325is converted◦C under 5more bar hydrogenpreferentially over to Pt/C phenol catalyst than [61 to] m and‐cresol. thus it was proposed that anisole is converted more preferentially to phenol than to m-cresol. a)

Figure 14. (a) Temperature and pressure proﬁles during HDO of guaiacol; (b) yields of different Figure 14. (a) Temperature and pressure profiles during HDO of guaiacol; (b) yields of different products in HDO of guaiacol as a function of temperature, solvent decane [45]. products in HDO of guaiacol as a function of temperature, solvent decane [45].

DehydroxylationDehydroxylation of theof phenylthe phenyl ring is thermodynamicallyring is thermodynamically more demanding more than demanding demethoxylation than (Table3)[ 22,36]. Substantial amounts of catechol were formed from guaiacol prior to phenol formation demethoxylation (Table 3) [22,36]. Substantial amounts of catechol were formed from◦ guaiacol prior showingto phenol apparent formation difficulties showing apparent of direct difficulties phenol formation of direct fromphenol guaiacol formation at 300fromC guaiacol over Pt/C at 300 [22 °C]. Typically demethoxylation of bulky 1,3,5-trimethoxybenzene was rather slow even over mesoporous over Pt/C [22]. Typically◦ demethoxylation of bulky 1,3,5‐trimethoxybenzene was rather slow even Rh/MCM-36over mesoporous at 250 Rh/MCMC under‐36 40 at bar 250 hydrogen °C under due40 bar to thehydrogen sterical due hindrance, to the sterical despite hindrance, the fact that despite it is easythe fact to break that it the is methoxyeasy to break bond the compared methoxy to bond the hydroxyl compared bond to the connected hydroxyl to thebond phenyl connected ring [ 26to]. the phenylSubstantial ring [26]. demethylation has been observed over Ni-, Fe-based catalysts [50] as well as over Mo2S and CoMoS catalysts [46] in HDO of guaiacol at relatively high temperatures. Methanation was ◦ also observed in HDO of guaiacol over CoMo/ZrO2 catalyst at 300 C under 50 bar hydrogen [46] and Co/Al-MCM-41 at 400 ◦C under 1 bar [34]. In the latter work it was concluded that demethylation Catalysts 2017, 7, 265 25 of 40 occurs only on the acid sites especially with high catalyst weight to flow rate [34]. Methanation could however, be inhibited in guaiacol HDO via transalkylation increasing the carbon efficiency [34,50]. On the other hand, a very low methanation activity occurred in HDO of anisole at 320 ◦C under 1 bar using MoO3/ZrO2 catalyst [51]. It is also noteworthy that substantial methanation occurred in HDO of guaiacol and anisole over Pt-Sn/CNF [48]. In addition, methanation is especially promoted by Rh and Ru [76]. Methanation in anisole HDO increased with increasing temperature over Mo2C[43]. Phenol and methanol were the primary products in anisole HDO over Mo2C, whereas methanol deoxygenated to methane at higher temperatures [43]. It was thus concluded that Mo2C is capable for breaking phenolic C-O bond in anisole at low temperature. Demethylation of guaiacol involves C(sp3)-Oar and is from the thermodynamic point of view the easiest reaction compared to, e.g., phenol, via C(sp2)-OMe cleavage [46]. Non-promoted MoS2 catalyst was beneficial for demethylation [46]. 2-methylphenol is formed over CoMoS/Al2O3 via adsorption of the methoxy oxygen group followed by methylation of the phenyl ring. The deoxygenation of methoxyphenol is more difficult than that of phenol [77]. Ring hydrogenation is thermodynamically favored during HDO of phenolic compounds under high pressure using relatively low temperatures [27,37,45,47–49,62,63]. Such ring hydrogenation makes the overall process more expensive because of hydrogen consumption. Typically hydrogenation during HDO is favored over supported Ni catalysts [47,62], CoNx [63], Rh/ZrO2 with the mixture of aluminosilicates [43], Pt/alumina [49] and even with MoO3 [36]. If the target is to perform HDO to produce BTX products, lower hydrogen consumption can be achieved under high temperatures [16,34]. For example, no ring hydrogenation occurred in HDO of guaiacol and anisole at 400 ◦C and 1 bar over Pt-Sn/CNF [48]. At the same temperature nickel and iron catalysts were also inefficient for hydrogenation [33,50]. Even when using 60 bar at 350 ◦C in anisole and phenol HDO, the main product was benzene over MoO3/ZrO2 catalyst showing clearly thermodynamic limitations at this temperature [39]. Ring hydrogenation in HDO of 4-propylphenol occurred with very high selectivity over Pt/AC at 280 ◦C under 40 bar hydrogen giving propylcyclohexane as the main product [57]. It was also pointed out that under these conditions using water as a solvent, protons can be formed increasing acidity of the reaction media [57]. Slightly lower yields of propylcyclohexane were achieved in HDO of 4-propylguaiacol, since Pt/AC catalyst promoted also formation of propylcyclopentane [57]. In eugenol HDO, ring hydrogenation occurred after demethoxylation over supported Pd catalysts under 30 bar hydrogen at 250 ◦C in hexadecane as a solvent [49]. It should, however, be noticed that ring hydrogenation could occur already prior to dehydroxylation [49]. Noteworthy is that 10 bar hydrogen was not sufficient to facilitate ring hydrogenation during vanillin HDO over Pd/CNF catalyst in the temperature range of 70–150 ◦C in water [29]. Acidic catalysts are active in HDO facilitating often complete conversion of phenolic compounds, but not hydrogenation [25,55], for example, vanillin transformation to p-creosol [25]. Transalkylation occurring after demethylation is beneficial for carbon efficiency [50] and it is favored by acidic catalysts [32]. Several transalkylated C7–C12 products were formed in guaiacol HDO ◦ at 400 C under atmospheric pressure due to acidity of Co-Ni-Al-MCM-41 [34]. MoO3/ZrO2 promoted also transalkylation in anisole HDO at 350 ◦C under 60 bar hydrogen [39]. It was also proposed based on the product distribution that transalkylation occurred parallel with hydrogenolysis [39]. On the other hand, no transalkylation occurred with Mo2C in HDO of a mixture of anisole, m-cresol, dimethoxybenzene and guaiacol at 280 ◦C under 114 kPa [42]. These results suggest that temperature is one of the decisive factors determining the transalkylation activity. A low alkylation ability was, however, observed in HDO of guaiacol at 400 ◦C over Pt-Sn/CNF catalyst exhibiting low acidity (Figure 15)[48]. Catalysts 2017, 7, 265 26 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 21 of 42

CH3 O OH +H2

‐H2O +

CH3 CH3 O O CH3 CH3 CH +H OH 3 2 +H2

‐CH4 ‐H2O

Figure 15. Reaction scheme for HDO and transalkylation of anisole at 400 ◦C under 1 bar over Figure 15. Reaction scheme for HDO and transalkylation of anisole at 400 °C under 1 bar over Pt-Sn/CNF adapted from ref. [48]. Pt‐Sn/CNF adapted from ref. [48].

Oligomerisation [25[25]] and and formation formation of gaseousof gaseous products products are alsoare typicalalso typical as evidenced as evidenced from several from severalHDO studies HDO studies regarding regarding lignin derived lignin derived phenolics. phenolics. As a result As a massresult balance mass balance is very is often very belowoften below 100%. ◦ 100%.For example, For example, in vanillin in vanillin HDO, performedHDO, performed over Mo over2C/AC Mo2 catalystC/AC catalyst under 20under bar hydrogen20 bar hydrogen at 100 atC 100oligomerisation °C oligomerisation of vanillin of vanillin occurred occurred causing causing incomplete incomplete mass balance mass [balance25]. Analogously [25]. Analogously in guaiacol in guaiacolHDO the HDO mass balancethe mass was balance not complete was not when complete performing when performing HDO over Mo HDO2C/CNF over inMo the2C/CNF temperature in the temperaturerange of 300–375 range◦C of under 300–375 55 bar °C under hydrogen. 55 bar The hydrogen. highest mass The highest balance mass closure balance of ca. 96%closure was of obtained ca. 96% wasat 375 obtained◦C[54]. at It was375 °C stated [54]. [35 It] was that stated in order [35] to that avoid in polymerisation, order to avoid stirringpolymerisation, of the bio-oil stirring should of the be biostarted‐oil should during itsbe heating.started during In bio-oil its HDO heating. CO2 Inwas bio the‐oil main HDO product CO2 was due the to themain presence product of due carboxylic to the presenceacids in the of carboxylic feed [35]. When acids in comparing the feed [35]. gaseous When composition comparing gaseous after bio-oil composition HDO it could after bio be‐ seenoil HDO that itRu/C could produced be seen morethat Ru/C methane produced from CO more2 than methane supported from Ni-catalysts, CO2 than insupported line with Ni 80%‐catalysts, more hydrogen in line withconsumption 80% more [35 hydrogen]. consumption [35].

3.4. HDO of Simulated and Real Bio Bio-Oils‐Oils and Lignin 3.4.1. HDO of Simulated Bio-Oils 3.4.1. HDO of Simulated Bio‐Oils Mixtures of different phenolic compounds to simulate bio-oil have been used in HDO (Figure 16, Mixtures of different phenolic compounds to simulate bio‐oil have been used in HDO Table4)[ 40,62]. The main products in the HDO of simulated bio-oil containing phenol, o-cresol (Figure 16, Table 4) [40,62]. The main products in the HDO of simulated bio‐oil containing phenol, o‐ and guaiacol over Ni- and Ni-Fe/H-Beta catalysts at 300 ◦C during 4 h under 16 bar hydrogen cresol and guaiacol over Ni‐ and Ni‐Fe/H‐Beta catalysts at 300 °C during 4 h under 16 bar hydrogen were oxygenated aromatic compounds [40]. The lowest amount of oxygenates was produced over were oxygenated aromatic compounds [40]. The lowest amount of oxygenates was produced over NiFe-H-Beta catalysts along with substantial amounts of cycloalkanes. It was also stated that water as NiFe‐H‐Beta catalysts along with substantial amounts of cycloalkanes. It was also stated that water a solvent is beneﬁcial for HDO, whereas methanol diminished selectivity towards hydrocarbons [40]. as a solvent is beneficial for HDO, whereas methanol diminished selectivity towards hydrocarbons The proposed reaction network for HDO of this simulated bio-oil is shown in Figure 17. It contains [40]. The proposed reaction network for HDO of this simulated bio‐oil is shown in Figure 17. It in addition to hydrogenation and deoxygenation also ring contraction of 1,2-cyclohexanediol to contains in addition to hydrogenation and deoxygenation also ring contraction of 1,2‐ cyclopentanol over acidic Ni-Fe/H-Beta catalyst. In HDO of a mixture of phenol, cresol, eugenol and cyclohexanediol to cyclopentanol over acidic◦ Ni‐Fe/H‐Beta catalyst. In HDO of a mixture of phenol, trans-anethole over Ni/SiO2-ZrO2 at 300 C under 50 bar during 8 h it was observed that this catalyst cresol, eugenol and trans‐anethole over Ni/SiO2‐ZrO2 at 300 °C under 50 bar during 8 h it was was rather efﬁcient towards hydrocarbons with 62% selectivity to cyclohexane at 91% conversion [62]. observed that this catalyst was rather efficient towards hydrocarbons with 62% selectivity to Gas-phase pyrolytic products contain also hydrogen, CO2, CO and water vapours. Guaiacol HDO was cyclohexane at 91% conversion [62]. Gas‐phase pyrolytic products contain also hydrogen, CO2, CO investigated using a co-feed of the simulated pyrolysis vapors [33]. The effect of gaseous co-feed is and water vapours. Guaiacol HDO was investigated using a co‐feed of the simulated pyrolysis discussed, however, in Section 3.5.6. vapors [33]. The effect of gaseous co‐feed is discussed, however, in Section 3.5.6.

Catalysts 2017, 7, x FOR PEER REVIEW 22 of 42

Catalysts 2017, 7, 265 27 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 22 of 42

Figure 16. Conversion and selectivity to different products as a function of time on stream in HDO of FigureFigure 16. Conversion 16. Conversion and and selectivity selectivity to to different different productsproducts as as a a function function of oftime time on onstream stream in HDO in HDO of of a mixture of m‐cresol, anisole, dimethoxybenzene and guaiacol at 280 ◦°C under 144 kPa over Mo2C a mixturea mixture of m-cresol, of m‐cresol, anisole, anisole, dimethoxybenzene dimethoxybenzene and guaiacol at at 280 280 °C Cunder under 144 144 kPa kPa over over Mo2C Mo 2C catalystcatalyst [[42].42]. [42].

Figure 17. Proposed reaction scheme for the HDO of simulated bio‐oil containing guaiacol, phenol and o‐cresol as feedstock over Ni‐H‐Beta catalyst. Notations: DDO: direct deoxygenation, HYD:

hydrogenation, DMO: demethoxylation, DME: demethylation, AL: alkylation and DHY: dehydration Figure[40]. 17. ProposedProposed reaction reaction scheme scheme for the for HDO the HDO of simulated of simulated bio‐oil bio-oilcontaining containing guaiacol, guaiacol, phenol phenoland o‐cresol and o-cresol as feedstock as feedstock over Ni over‐H‐Beta Ni-H-Beta catalyst. catalyst. Notations: Notations: DDO: direct DDO: deoxygenation, direct deoxygenation, HYD: HYD:hydrogenation, hydrogenation, DMO: demethoxylation, DMO: demethoxylation, DME: demethylation, DME: demethylation, AL: alkylation AL: and alkylation DHY: dehydration and DHY: dehydration[40]. [40].

Catalysts 2017, 7, 265 28 of 40

3.4.2. HDO of Real Bio-Oil HDO of real bio-oils [4,30,35,62] have been investigated in several publications. Bio-oil derived from yellow poplar via fast pyrolysis was hydrodeoxygenatied at 300 ◦C in ethanol at 30 bar hydrogen [4]. The lowest O/C content was obtained with Pt/C catalyst [4]. HDO of the bio-oil prepared via fast pyrolysis of Miscanthus sinensis at 350 ◦C and 30 bar over noble metals supported on carbon, in supercritical ethanol [30] showed deoxygenation of the heavy oil fraction of the bio-oil to 78% with the heating value of the product 27.8 MJ/kg. It is also important to note that such unstable compounds, as acids, furfural, vanillin and levoglucosan were converted to esters, ketones and saturated phenols [30]. Macromolecular products in that work were also characterized [28]. The results showed that the carbon content in the polymer increased and at the same time oxygen content decreased indicating lower amounts of phenolic and hydroxyl and methoxy groups. It was also observed that after prolonged reaction times the amount of methoxy groups decreased while phenol groups remained constant indicating that dehydroxylation of the phenol groups was not feasible under these reaction conditions [28]. On the other hand, when the bio-oil was hydrodeoxygenated over Pd/C catalyst at different temperatures varying from 250 ◦C to 350 ◦C phenolics started to polymerize [28]. In addition to Pt/C also Ni/SBA-15 was rather active in HDO of bio-oil giving ca. 55% deoxygenation in 1 h [4]. The main reactions occurring during HDO of bio-oil over the latter catalyst were guaiacol and 4-propylsyringol. In this case stability of the deoxygenated bio-oil was improved, since acetic acid, levoglucosan and furfural were totally converted. Gas phase products contained methane, CO, ethane and propane. The produced heavy oil fraction contained only 9.3% water with the heating value of 22.8 MJ/kg [4]. HDO of a real lignin derived bio-oil was successfully carried out over Ni/SiO2-ZrO2 catalyst at 300 ◦C under 50 bar hydrogen resulting in formation of 63% of cyclic hydrocarbons, out of which 8% were aromatics. In addition it was stated that most of the identiﬁed products, were cyclic hydrocarbons, such as cyclohexane, toluene and xylene exhibiting suitable vapor pressures and carbon number to be used in gasoline [62].

3.4.3. HDO of Lignin

Soda lignin HDO was investigated over various Ni/SiO2 catalysts impregnated with Ni2P and also with CuMgAlOx [3]. The best results were obtained using a mixture of catalysts comprising ◦ Ni/SiO2 containing Ni2P together with CuMgAlOx. At 340 C under 20 bar of hydrogen for HDO in ethanol of oxygen removal 84% was achieved with the aromatic lignin monomer yield of 51% and relatively low ring hydrogenation [3]. It was stated that the role of CuMgAlOx was to generate hydrogen during lignin ethanolysis [19] which in turn activates the reduced and passivated Ni2P/SiO2 catalyst [78]. Complete conversion of organosolv lignin was obtained during its HDO with Ni-BEA at 320 ◦C under 20 bar hydrogen [2]. The hydrocarbon selectivity of 70% was reported, while the amount of water and the solid residue were 21% and 4%, respectively. The reaction network includes disruption of phenolic compounds of lignin from ether linkages. Thereafter ring hydrogenation followed by deoxygenation occur over Ni-catalyst with slow hydrogenolysis. It was also concluded that at lower temperatures oligomers are formed via recombination of phenolic compounds.

3.5. Catalyst Deactivation, Regeneration and Reuse during HDO of Phenolic Compounds Catalyst deactivation and continuous HDO have been intensively investigated for phenolic compounds [31,33,42,49,50,57]. Catalyst stability, activity as a function of time-on- stream [31,33,36,39,42,43,47,48,58,59,61], reuse [30,54,56–58,63], regeneration [34,36], origin of the catalyst deactivation [30], chemical composition of coke [31], thermogravimetrical analysis of the spent catalysts [62], modelling of deactivation [33] are also very important from the industrial point of view. Since the real bio-oils contain in addition to lignin derived compounds also water [59], acetic acid Catalysts 2017, 7, 265 29 of 40 or furfural [56] as well several catalyst poisons, such as chlorine, potassium and sulphur containing compounds [47], it is very important not only to investigate HDO of model compounds but also the real feed [30] using different approaches, such as a single component feed, co-feed studies, real bio-oil [30], simulated bio-oil [40], adsorption studies [48,79] as well as continuous operation [33,36,39,42,43,48].

3.5.1. Catalyst Reuse in HDO of Model Compounds and Bio-Oils Reuse of the catalysts in HDO of phenolic compounds has been intensively ◦ investigated [49,54,56,57,62]. Mo2C/CNF catalyst was reused in guaiacol HDO at 350 C under 55 bar after rinsing it with diethyl ether [54]. The results were very promising showing that in three consecutive experiments more than 99% conversion was achieved. On the other hand, W2C/CNF gave lower activity exhibiting only 48% conversion after three consecutive experiments. Mo2C/CNF catalyst suffered from irreversible coking during three consecutive HDO experiments of guaiacol at ◦ ◦ 300 C under 40 bar [56]. The reuse of Ni/SiO2-ZrO2 was investigated in guaiacol HDO at 300 C under 50 bar hydrogen [62]. The results showed that in four consecutive experiments the conversion decreased as follows: 100%, 91%, 90% and 87% indicating minor catalyst deactivation. The use of Pt/C is favorable in HDO of phenolic compounds in order to avoid coke formation [57]. Successful recycling was demonstrated in 4-propylphenol HDO over Pt on active carbon giving more than 99% yield of propylcyclohexane in three consecutive experiments at 280 ◦C under 40 bar hydrogen [57]. The results revealed also that no sintering occurred. Recycling of noble metal supported catalysts in guaiacol HDO was done in three consecutive experiments, with catalyst washing with acetone and reduction at 200 ◦C in between. The results showed minor deactivation of Ru/C and Pd/silica alumina [49]. It was conﬁrmed that especially in HDO of eugenol the catalyst was poisoned by strongly adsorbed products causing also a brownish colour of the product. In bio-oil HDO over noble metal catalysts supported on active carbon at 300 ◦C under 30 bar hydrogen, the catalyst deactivation was intensive during reuse [30]. It was also concluded that Ru/C deactivated mainly via fouling and blocking of active sites, whereas activity loss of Pt/C was attributed to extensive char deposition on the catalyst surface.

3.5.2. Time-on Stream Behavior in HDO of Model Compounds ◦ Time-on-stream behavior during HDO of cresol over MoO3 catalyst at 320 C at atmospheric pressure showed catalyst deactivation. When the catalyst was regenerated at 400 ◦C via calcination, its original activity was restored, however, after each 24 h a repeated regeneration was required [36]. ◦ Promising results have been obtained in HDO of phenyl acetate over Pt/Al2O3 at 325 C under 5 bar hydrogen. After 25 h time-on-stream a stable period of 40 h with constant activity followed [61]. Phenol formation was rapid from phenyl acetate, whereas the consecutive benzene formation was rather slow. Only minor catalyst deactivation was also observed in HDO of m-cresol, anisole, ◦ dimethoxybenzene and guaiacol mixture at 280 C under 1.5 bar pressure over Mo2C catalyst giving mostly toluene and benzene as products (Figure 18)[42]. On the other hand, the most rapid deactivation ◦ of Mo2C/ZrO2 as a function of time on stream was observed in HDO of guaiacol at 350 C under 60 bar hydrogen, followed by anisole and phenol showing clearly that compounds bearing two oxygen atoms caused the most intensive deactivation (Figure 18)[39]. Prominent catalyst deactivation observed in guaiacol HDO can partially be explained by a strongly bound phenate mode of guaiacol by double anchoring to the catalyst surface covering a large part of active sites [80]. Catalyst deactivation occurred also in HDO of anisole and guaiacol over Pt-Sn/CNF/Inconel as a funbction of time on stream (Figure 19)[48]. In addition, in the HDO of anisole, less water is formed than, for example, in phenol HDO. At higher temperature catalyst stability is improved in the presence of smaller amounts of adsorbed water. Catalyst deactivation in guaiacol HDO was also more intensive than in HDO of anisole over Pt-Sn/CNF catalyst at 400 ◦C under atmospheric pressure (Figure 20) i.e., conditions favoring deactivation [48]. Guaiacol is overall more reactive than anisole at 400 ◦C, thus more prominent catalyst Catalysts 2017, 7, 265 30 of 40 deactivation in HDO of guaiacol compared to anisole is mainly seen from the product distribution. ForCatalysts anisole, 2017, the 7, x mainFOR PEER product REVIEW was benzene, whereas it was anisole for HDO of guaiacol. 25 of 42

50 "OH" 40

30 "-OCH3"

20 Conversion (%) "-OH + -OCH3"

0 02468 Time on stream (h)

◦ FigureFigure 18. 18.Conversion Conversion of of guaiacol, guaiacol, anisole anisole and and phenol phenol in in their their single single feed feed HDO HDO at at 350 350 C°C under under 60 60 bar bar inin a downa down ﬂow flow trickle trickle bed reactorbed reactor adapted adapted from [ 39from]. Notation: [39]. Notation: ( ) phenol, (●) (phenol,) anisole (■ and) anisole (o) guaiacol. and (o) guaiacol.

Figure 19. Continuous hydrodeoxygenation of (a) anisole and (b) guaiacol at 400 ◦C under atmospheric pressure over Pt-Sn/CNF/Inconel catalyst [48].

Figure 19. Continuous hydrodeoxygenation of (a) anisole and (b) guaiacol at 400 °C under atmospheric pressure over Pt‐Sn/CNF/Inconel catalyst [48].

Catalysts 2017, 7, 265 31 of 40 Catalysts 2017, 7, x FOR PEER REVIEW 26 of 42

80 calcination

60 regenerated calcination

Conversion (C-mol%) 20

0 0 5 10 15 20 25 30 35 40 45 50 TOS (h)

◦ Figure 20. FigureConversion 20. Conversion of m-cresol of m‐cresol as a as function a function of of timetime on on stream stream at 320 at 320°C underC under 1 bar over 1 bar MoO over3 MoO3 catalyst. Regenerationcatalyst. Regeneration of the of catalyst the catalyst was was performed performed at 400 400 °C◦ Cin oxygen in oxygen for 3 h for adapted 3 h adapted from [36]. from [36].

Catalyst regeneration has also been investigated in continuous operation [36]. Typically, the Catalystcatalyst regeneration is calcined in has air also at 400 been °C in investigated order to avoid insintering continuous of metal operation‐like particles, [36 as]. it Typically, was the case the catalyst is calcinedin in HDO air at of 400 m‐cresol◦C in at order 320 °C to under avoid atmospheric sintering pressure of metal-like over MoO particles,3 (Figure 21) as [36]. it was The the results case in HDO showed clearly that already after 24 h TOS the catalyst activity declined from 50% conversion to 28% of m-cresol at 320 ◦C under atmospheric pressure over MoO (Figure 21)[36]. The results showed conversion and the intermittent calcination was required in order3 to keep the catalyst activity high clearly thatenough. already after 24 h TOS the catalyst activity declined from 50% conversion to 28% conversion and theCatalysts intermittent 2017, 7, x FOR calcination PEER REVIEW was required in order to keep the catalyst activity high enough.27 of 42

100

50 Conversion/Selectivity (%)

0 0246

Reaction time (h)

Figure 21. Results from HDO of guaiacol over Mo2C/CNF catalyst at 350 °C under◦ 55 bar hydrogen. Figure 21. Results from HDO of guaiacol over Mo2C/CNF catalyst at 350 C under 55 bar hydrogen. Symbols: mass balance (■), conversion (●), phenol (o) and cresol (□) adapted from reference [54]. Symbols: mass balance (), conversion ( ), phenol (o) and cresol () adapted from reference [54]. 3.5.3. Catalyst Selection in HDO of Phenolic Compounds Precious metal catalysts have some advantages during HDO of phenolic compounds being highly active at low pressures and temperatures and enabling use of smaller reactors and other economic benefits [61]. Metal‐like carbides have also shown activity in HDO of phenolic compounds [56]. For nitride and carbide catalysts, neutral carbon and zirconia are preferred as supports in direct deoxygenation of guaiacol [46]. Hydrophobic carbons are also efficient supports for bio‐oil HDO [78] deactivating only slightly in the presence of water during bio‐oil HDO [55]. Hydrothermal stability of the catalysts is important during HDO of lignin derived phenolic compounds. It is known that ZrO2 exhibits hydrothermal stability, whereas alumina is unstable in the presence of water under hydrothermal conditions [81]. Nitrides have been more stable against sulfides in HDO of guaiacol in a continuous operation [58]. Transition metal carbides, nitrides and phosphides require demanding preparation conditions and exhibit poor stability in HDO of phenolic compounds [37]. Performing HDO at higher temperature enhances stability of MoO3/ZrO2 due to lower amounts of water adsorbed on the catalyst surface [39].

Catalysts 2017, 7, 265 32 of 40

3.5.3. Catalyst Selection in HDO of Phenolic Compounds Precious metal catalysts have some advantages during HDO of phenolic compounds being highly active at low pressures and temperatures and enabling use of smaller reactors and other economic benefits [61]. Metal-like carbides have also shown activity in HDO of phenolic compounds [56]. For nitride and carbide catalysts, neutral carbon and zirconia are preferred as supports in direct deoxygenation of guaiacol [46]. Hydrophobic carbons are also efficient supports for bio-oil HDO [78] deactivating only slightly in the presence of water during bio-oil HDO [55]. Hydrothermal stability of the catalysts is important during HDO of lignin derived phenolic compounds. It is known that ZrO2 exhibits hydrothermal stability, whereas alumina is unstable in the presence of water under hydrothermal conditions [81]. Nitrides have been more stable against sulfides in HDO of guaiacol in a continuous operation [58]. Transition metal carbides, nitrides and phosphides require demanding preparation conditions and exhibit poor stability in HDO of phenolic compounds [37]. Performing HDO at higher temperature enhances stability of MoO3/ZrO2 due to lower amounts of water adsorbed on the catalyst surface [39].

3.5.4. Coking during HDO of Model Compounds Possible reasons for catalyst deactivation during HDO of phenolic compounds are coking [34] and poisoning [47], in addition to a loss of active sites. In some cases minor sintering has been observed [31]. In guaiacol HDO over Pt/C slight catalyst deactivation occurred at 325 ◦C[31]. Coking can be explained by ring condensation favored by low pressure and high temperature [82], whereas aromatic hydrogenation is favored at lower temperature using high hydrogen pressure. Condensed ring products are more strongly adsorbed on the catalyst surface and can be desorbed only at higher temperatures [83]. Co-MCM-41 was used for HDO of guaiacol resulting in a quite rapid catalyst deactivation occurred at 400 ◦C under atmospheric conditions [34], as these conditions promoted extensive coking. Coking during HDO is typically favored over catalysts having acidic supports, while for less acidic supports, such as zirconia and activated carbon, the catalysts retain their activity longer [58]. More intensive coking during HDO of guaiacol at 300 ◦C was observed for Ru/C compared to Pt/C [31]. Catalyst coking has been also investigated in HDO of guaiacol and anisole [50]. Conﬁrming that reactant nature has an effect on coking in particular more coke has been formed from guaiacol than anisole on Ni-based catalysts at relatively high temperatures, between 350–400 ◦C[50]. Coke formation was conﬁrmed in guaiacol HDO at 400 ◦C under atmospheric pressure over Co/Al-MCM-41 catalyst by TGA [34]. Furthermore, it was stated that the catalyst regeneration at 450 ◦C was not able to release all coke. From TEM images it was found that coke is located on both on acidic sites and close to Co-particles [34] leading to conclusion that coke formation depends not only on acidity, but also on the presence of metal sites [34]. Analysis of coke deposited on the spent catalyst used in HDO of guaiacol at 300 ◦C under atmospheric pressure was performed using thermogravimetry and GC-MS [31]. The TGA results revealed that more coke was desorbed from Ru/C compared to Pt/C. Furthermore, linked ring structures such as biphenyl, aromatic compounds, e.g., benzene, and condensed rings including naphthalene and 1-methylnaphtalene were found after extraction of adsorbed species by dichloromethane not being present in the liquid product. It was additionally concluded that there was no sintering of Ru/C under HDO conditions, while minor sintering of Pt/C was observed.

3.5.5. Deactivation Due to Strong Adsorption on the Catalyst Surface Deactivation studies of different supports, such as alumina and silica have been performed by adsorption of different model molecules (phenol, anisole, guaiacol) on these supports and desorbing them at different temperatures [79]. Based on IR-studies interaction mechanisms of these molecules at higher temperatures have been proposed. The results revealed that anisole is demethylated at Catalysts 2017, 7, 265 33 of 40 higher temperature, whereas guaiacol is bound via two oxygen atoms on Al-sites. The amounts of adsorbed species on SiO2 were much lower than those recorded for Al2O3 in the case of phenol and anisole. Even if for guaiacol adsorption quantitative data were not achieved, it was concluded that its interaction with supports was clearly stronger than for phenol and anisole. Strong adsorption of reactants or other molecules during eugenol hydrodeoxygenation has been proposed to cause catalyst deactivation [49]. In eugenol HDO incomplete carbon balance at 250 ◦C under 30 bar hydrogen in hexadecane over noble metal catalysts supported was ascribed strong adsorption of the products on the catalyst surface [49]. Adsorption strength of guaiacol and anisole was investigated using pulse experiments at 400 ◦C by adsorbing these molecules on Pt-Sn/CNF catalyst [48]. It was conﬁrmed that guaiacol is more strongly adsorbed than anisole. Furthermore, it was also stated that guaiacol and its HDO-product, catechol with two oxygen atoms, are more strongly adsorbed on the catalyst surface than their single functionalized counterparts, e.g., anisole and phenol [48].

3.5.6. Poisons in Bio-Oil Causing Catalyst Deactivation Hydrodeoxygenation (HDO) is an important treatment method for bio-oils which have low stability and heating value as described above. Due to a complex structure of bio-oil and presence of possible catalyst poisons [47] and water, its HDO is, however, demanding. Bio-oils stemming from hemicellulose and cellulose contain also acetic acid and furfural, which have been used as co-feed in HDO of guaiacol [56]. Co-feed including also H2O, CO2 and O2 has been used in HDO of guaiacol [33,59]. The effect of several poisons, such as chlorine, sulfur and potassium was investigated in HDO ◦ of guaiacol and 1-octanol over Ni/ZrO2 catalyst in a trickle bed reactor at 250 C under 100 bar hydrogen [47]. In these experiments, the catalyst was deactivated with 1-octanethiol, 1-chlorooctane, KCl or KNO3. When using potassium salts a stoichiometric amount of salt compared to nickel was used. The results revealed that with sulfur present in the catalyst, the latter retained its activity during more than 4 h which was followed by a dramatic activity drop almost to zero. Effect of chlorine poisoning was more rapid decreasing the catalyst activity immediately after the start of the chlorine feed. On the other hand, a potassium poisoned catalyst exhibited high activity in guaiacol conversion, even if its deoxygenation activities towards guaiacol were 3% and 18% for respectively KCl and KNO3 in comparison with the not poisoned catalyst displaying 66% deoxygenation degree of guaiacol. It was also noteworthy that chlorine and potassium poisoning was reversible and chloride was found in the liquid product. Sulfur poisoning inhibited both hydrogenation of the aromatic ring in ◦ guaiacol as well as subsequent deoxygenation reactions over Ni/ZrO2 catalyst at 250 C under 100 bar hydrogen, whereas potassium poisoned mainly deoxygenation sites. Chlorine poisoning decreased both hydrogenation and deoxygenation activity of Ni/ZrO2 catalyst. The presence of acetic acid and furfural was investigated in HDO of guaiacol over Mo2C/CNF catalyst in dodecane at 300 ◦C under 40 bar hydrogen [56]. The results showed that both acetic acid and furfural were converted to a large extent, whereas guaiacol conversion was only 6% after 4 h. The main products were phenol and catechol, while without furfural and acetic acid catechol formation was much lower. Catalyst stability was investigated during HDO of m-cresol with Mo2C using co-feeds of H2O, CO2 and O2 [59]. The results revealed that especially toluene formation rate was retarded with adsorption of oxygen on metal like sites catalyzing HDO reaction. In the gas phase HDO of m-cresol-anisole, dimethoxybenzene and guaiacol mixture the experimental observations could be correlated with the bond dissociation energies showing that the most prominent products were toluene ◦ and benzene at 280 C over Mo2C catalysts. This catalyst was completely inactive for hydrogenation of the products. It was stated [59] that CO and water, which are typically present in pyrolysis vapor, inhibit hydrogenation of the aromatic ring. Co-feeding CO and CO2 has been also investigated in HDO of guaiacol [33]. Direct hydrodeoxygenation of phenolic compounds was achieved over Mo2C catalyst without ring Catalysts 2017, 7, 265 34 of 40 hydrogenation which was also conﬁrmed by co-feed of water and CO. In guaiacol HDO in the ◦ presence of CO, CO2 and water over Fe/SiO2 catalyst at 400 C under atmospheric pressure inhibition of guaiacol transformations to cresol was observed (Figure 22)[33]. In the vapor-phase anisole HDO, benzene was formed with high selectivity over metal-like carbides, Mo2C and W2C[43,44]. It was, however, observed that with co-fed CO formation of benzene during anisole HDO was suppressed ◦ over Mo2C catalyst at 133 C under atmospheric pressure [43]. Analogously in the presence of Catalysts 2017, 7, x FOR PEER REVIEW ◦ 31 of 42 CO-co-feeding during anisole HDO at 150 C over Mo2C under atmospheric pressure benzene formation was suppressed and completely recovered after CO-co-feed was stopped [44]. Pt/C is active for water gas shift reaction forming carbon dioxide during HDO [84].

2.0x10-6 1

1.5x10-6 ) cat kg . 1.0x10-6 k (kmol/s

5.0x10-7 2 3 0.0 4 0 50 100 150 200 Time on stream (min)

Figure 22. Decrease of the kinetic constants as a function of time on stream in the HDO ◦ of guaiacolFigure at 22. 400 DecreaseC under of the 1 kinetic bar over constants Fe/SiO as2 aadapted function from of time ref. on [ 33stream]. The in reaction the HDO mixture of guaiacol at contained400 °C also under 50% 1 H bar2, over 5% CO Fe/SiO2, 2.5%2 adapted CO and from 2.5% ref. [33]. H2O. The Notation: reaction (1) mixture guaiacol contained to products, also 50% H2, −6 −4.61·10−3 x −7 −7.29·10−3 x line fitting:5% CO= 2, 2.852.5%·10 COe and 2.5%, (2) H guaiacol2O. Notation: to cresol, (1) line guaiacol fitting :to y products,= 4.93 ·10line efitting: , 2.85 ∙ − − · −3 (3) phenol10 to.∙ benzene,, line 2 guaiacol to cresol, fitting: y = 1.37 line fitting: ·10 7e 4.85 10 4.93x, and ∙ 10 (4) cresol.∙ to toluene,, line fitting: − − · −3 y = 2.813 phenol to benzene,·10 8e 5.03 10 x. line fitting: 1.37 ∙ 10.∙ and 4 cresol to toluene, line fitting: 2.81 ∙ 10.∙ . 3.5.7. Carbon Balance during HDO of Model Compounds 4. Conclusions Mass balance in HDO of phenolic model compounds has often been reported to be below 100% [25,31Hydrodeoxygenation,54]. One example isof guaiacolbio‐oils, lignin HDO and over their Mo 2modelC/CNF compounds and W2C/CNF has been catalysts intensively (Figureinvestigated 12)[54], where in the the recent authors years. also The stated main that findings the mass in balance this area increased are summarized with increasing in the reactioncurrent work. temperature.Several Amodel relatively molecules high mass have balance been closureused in above HDO 90% of bio was‐oil achieved including at 375 single,◦C under double 55 baror triple hydrogenoxygenated [54]. Analogously phenolic tocompounds. the work of Several Jongerius reaction et al., massnetworks balances based vary on in kinetic the range results of 70–98% have been ◦ in guaiacolproposed HDO especially at 300 C underfor guaiacol 40 bar usingHDO. Mo The2C/CNF main catalystsparameters (Figure affecting 21). It alsoactivity pointed and out product that thedistribution mass balances are temperature contained only and thepressure, liquid type phase of and a catalytically coke from theactive catalyst metal, surface nature omittingof the support the gaseousand its products acidity. [ 56Two]. Clearly main cokingpathways [56] andduring release HDO of productsare either into ring the hydrogenation gas phase decrease followed the by mass balancedeoxygenation calculated at high on the pressure basis ofand the temperature liquid phase below products. 300 °C, In as HDO well ofas guaiacoldeoxygenation the highest at elevated carbontemperatures, recovery in the when liquid hydrogenation phase over is noble thermodynamically metals supported suppressed. on carbon was ca. 70% obtained at 300 ◦C[31The]. Especiallydegree of deoxygenation the ring opening is reactionimportant at highin order temperatures to increase leads the energy to formation density of and more heating value of the products. Typically, guaiacol is deoxygenated at elevated temperatures or under very high pressures at temperatures below 350 °C affording HDO yields between 60–80%. One exception in guaiacol HDO is Ni/SiO2‐ZrO2 giving 97% cyclohexane at 300 °C under 50 bar. In addition to cyclohexane, also cyclopentanone is formed, especially over Pt. These results can be explained theoretically at a DFT level, confirming the reactivity of phenolic compounds decreases with an increasing number of oxygen atoms. Moreover, thermodynamically the hydroxyl group connected to benzene is more difficult to cleave compared to a methoxy group. HDO of real bio‐oils was also successfully demonstrated over various Ni‐and Pd‐supported catalysts. Unstable compounds, such as furfural, acetic acid and levoglucosan were also transformed to more stable ketones and esters. In addition, high HDO activity was obtained for lignin itself, especially over Ni/H‐BEA catalyst.

Catalysts 2017, 7, 265 35 of 40 gaseous products [18]. On the other hand, total carbon yields varied in the range of 95–98% in the HDO ◦ of different phenolic compounds at 300–350 C over MoO3 catalyst under atmospheric pressure [36]. Solid products are also formed during HDO of phenolic compounds and lignin [2,25,62]. Polymerization, occurring especially for vanillin, is intensive, decreasing the carbon balance [25], ◦ for example, in vanillin HDO at 100 C under 30 bar hydrogen over Mo2C/AC catalyst. As a result mass balances in the range of 65–70% were reported. It was concluded by Zhang et al. [62] that the mass losses are caused by formation of coke and polymers as well as formation of volatile gaseous products. In addition large amounts of char were formed being in the range of 9–19% in the HDO of pyrolysis oil over Ni-supported catalysts at 300 ◦C under 30 bar hydrogen [4]. Noteworthy is that formation of a solid residue in lignin HDO decreased with increasing reaction temperature indicating that at higher temperature unwanted reactions of phenolic products are retarded [2]. Formation of gaseous products is typically enhanced at higher temperatures [31,34], while at lower temperatures less gases were formed [43]. For example, in the temperature range of 147–247 ◦C no gaseous C2–C5 products were observed in vapor phase HDO of anisole on Mo2C and the main product was benzene formed by C-O bond cleavage [43]. In addition, low gas generation during HDO was observed at 250 ◦C under 80 bar hydrogen over various Ni catalysts [35]. At elevated temperature, 350 ◦C, both char and gas formation from HDO of miscanthus bio-oil was enhanced when HDO was performed over Pd/C catalyst under 30 bar of hydrogen [28]. Under these conditions char and gas yields vary in the ranges of 16% and 19–36%, respectively depending on the reaction time in the range of 30–60 min [28]. The liquid yields, composed of light and heavy oil fractions were the highest, above 80%, when HDO of bio-oil was performed over Pd/C catalyst between 250–300 ◦C[28]. Formation of gaseous products depends also on the type of active metal in the catalyst [31]. For noble metal catalysts the amounts of gaseous products formed during guaiacol HDO increased in the following order: Pd/C < Ru/C < Rh/C < Pt/C [31]. A substantial amount of gaseous products, 50%, was formed in HDO of guaiacol at 400 ◦C under atmospheric pressure over NiCo-Al-MCM-41 catalysts [34]. Methane formation was very extensive during HDO of guaiacol over Co/Al-MCM-41 at 400 ◦C under atmospheric pressure and it increased with increasing ratio of catalyst weight to feed rate [34]. The gas phase product amount was was the highest with Ni/Al-MCM-41, being 78%, whereas it was the lowest (30%) for Co-MCM-41 [34]. The main gaseous product yield in the HDO of bio-oil over supported Ni-catalysts at 300 ◦C under 30 bar hydrogen was CO being in the range of 88–95% from the gaseous products, followed by methane varying in the range of 2–7.1% [4]. It was also possible to obtain a few % of propane, ethyne and ethane [4]. Only minor amounts of methane were formed over supported Ni-catalysts in the HDO of guaiacol under 7.9 bar hydrogen in the temperature range of 350–450 ◦C, whereas coking was rather high under these conditions over Ni-catalysts being in the range of 3–8 wt % from the total products [50]. Analogously, in the HDO of pyrolysis bio-oil from yellow polar over Ni supported catalysts at 300 ◦C under 30 bar of hydrogen, formation of gaseous products was signiﬁcant over Ni supported catalysts varying in the range of 6.6–27% depending on the support type [4]. On the other hand, low methanation degree was achieved over MoO3/ZrO2 at 320 ◦C under atmospheric pressure in anisole HDO showing clearly its carbon efﬁciency [51].

3.5.8. Modelling of Deactivation for HDO of Model Compounds Modelling of deactivation along with kinetic modelling was performed [33] for HDO of guaiacol over Fe/SiO2 catalyst [33]. When the rate constants were plotted against time on stream it was noticed that the decrease of rate constants occurred similarly for HDO of phenol to benzene and cresol to toluene, whereas in guaiacol transalkylation to cresol the decline of the rate constant was more rapid (Figure 22)[33]. Thus, it was concluded that inhibition of guaiacol to cresol transformations may happen on different sites compared to HDO reactions. Catalysts 2017, 7, 265 36 of 40

4. Conclusions Hydrodeoxygenation of bio-oils, lignin and their model compounds has been intensively investigated in the recent years. The main findings in this area are summarized in the current work. Several model molecules have been used in HDO of bio-oil including single, double or triple oxygenated phenolic compounds. Several reaction networks based on kinetic results have been proposed especially for guaiacol HDO. The main parameters affecting activity and product distribution are temperature and pressure, type of a catalytically active metal, nature of the support and its acidity. Two main pathways during HDO are either ring hydrogenation followed by deoxygenation at high pressure and temperature below 300 ◦C, as well as deoxygenation at elevated temperatures, when hydrogenation is thermodynamically suppressed. The degree of deoxygenation is important in order to increase the energy density and heating value of the products. Typically, guaiacol is deoxygenated at elevated temperatures or under very high pressures at temperatures below 350 ◦C affording HDO yields between 60–80%. One exception ◦ in guaiacol HDO is Ni/SiO2-ZrO2 giving 97% cyclohexane at 300 C under 50 bar. In addition to cyclohexane, also cyclopentanone is formed, especially over Pt. These results can be explained theoretically at a DFT level, confirming the reactivity of phenolic compounds decreases with an increasing number of oxygen atoms. Moreover, thermodynamically the hydroxyl group connected to benzene is more difficult to cleave compared to a methoxy group. HDO of real bio-oils was also successfully demonstrated over various Ni-and Pd-supported catalysts. Unstable compounds, such as furfural, acetic acid and levoglucosan were also transformed to more stable ketones and esters. In addition, high HDO activity was obtained for lignin itself, especially over Ni/H-BEA catalyst. Catalyst deactivation studies revealed that main reason for catalytic activity decline during HDO of phenolic compounds is coking in addition to reactant polymerization. Interestingly higher mass balance closures exceeding 90% for guaiacol at 375 ◦C could be achieved at very high temperatures, due to suppression of oxygen compounds, such as phenol over Mo2C/CNF. Catalyst poisoning was systematically investigated showing irreversible bulk poisoning of the catalyst with sulphur, whereas chlorine poisoning was reversible. Hydrothermal stability of the catalyst is of crucial importance and suitable candidates are active carbon and ZrO2, whereas alumina and metal carbides exhibit lower stability under catalytically relevant conditions. The HDO of phenolic compounds is challenging and requires optimization of different parameters, such as temperature, pressure and the catalyst per se in order to maximize the liquid product yield and produce desired compounds with high enough heating value and RON number. In addition, further research efforts are still needed to diminish hydrogen consumption and provide high carbon efficiency.

Author Contributions: P.M-A designed the review, collected references and wrote the first draft. D.Yu.M. revised the text and contributed to the final paper. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Hussin, M.H.; Rahim, A.A.; Ibrahim, M.N.M.; Brosse, N. Physicochemical characterization of alkaline and ethanol organosolv lignins from palm oil (Elaeis guineensis) fronds as phenol substitutes for green material applications. Ind. Crops Prod. 2013, 49, 23–32. [CrossRef] 2. Kasakov, S.; Shi, H.; Camaioni, D.M.; Zhao, C.; Baráth, E.; Jentys, A.; Lercher, J.A. Reductive deconstruction of organosolv lignin catalyzed by zeolite supported nickel nanoparticles. Green Chem. 2015, 17, 5079–5090. [CrossRef]

3. Koranyi, T.I.; Hensen, E.J.M. Preparative aspects of supported Ni2P catalysts for reductive upgrading of technical lignin to aromatics. Catal. Lett. 2017, 147, 1722–1731. [CrossRef] 4. Oh, S.; Choi, H.S.; Choi, I.G.; Choi, J.W. Evaluation of hydrodeoxygenation reactivity of pyrolysis bio-oil with various Ni-based catalysts for improvement of fuel properties. RSC Adv. 2017, 7, 15116–15126. [CrossRef] Catalysts 2017, 7, 265 37 of 40

5. Wang, H.; Male, J.; Wang, Y. Recent advances in hydrotreating of pyrolysis bio-oil and its oxygen-containing model compounds. ACS Catal. 2013, 3, 1047–1070. [CrossRef] 6. Bu, Q.; Lei, H.; Zacher, A.H.; Wang, L.; Ren, S.; Liang, S.; Wei, Y.; Liu, Y.; Tang, J.; Zhang, Q.; et al. A review of catalytic hydrodeoxygenation of lignin-derived phenols from biomass pyrolysis. Bioresour. Technol. 2012, 124, 470–477. [CrossRef][PubMed] 7. Cheng, F.; Brewer, C.E. Producing jet fuel from biomass lignin: Potential pathways to alkylbenzenes and cycloalkanes. Renew. Sustain. Energy Rev. 2017, 72, 673–722. [CrossRef] 8. He, Z.; Wang, X. Hydrodeoxygenation of model compounds and catalytic systems for pyrolysis bio- oils upgrading. Catal. Sustain. Energy 2013, 28–52. [CrossRef] 9. Lee, H.; Kim, Y.-M.; Lee, I.-G.; Jeom, J.-K.; Jung, S.-C.; Chung, J.D.; Choi, W.G.; Park, Y.-K. Recent advances in the catalytic hydrodeoxygenation of bio-oil. Korean J. Chem. Eng. 2016, 33, 3299–3315. [CrossRef] 10. Saidi, M.; Samimi, F.; Karimipourfard, D.; Nimmanwudipong, T.; Gates, B.C.; Rahimpour, M.R. Upgrading of lignin-derived bio-oils by catalytic hydrodeoxygenation. Energy Environ. Sci. 2014, 7, 103–129. [CrossRef] 11. Shafaghat, H.; Rezaei, P.S.; Mohd, W.; Wan Daud, A. Effective parameters on selective catalytic hydrodeoxygenation of phenolic compounds of pyrolysis bio-oil to high-value hydrocarbons. RSC Adv. 2015, 5, 103999–104042. [CrossRef] 12. Sullivan, M.M.; Chen, C.-J.; Bhan, A. Catalytic deoxygenation on transition metal carbide catalysts. Catal. Sci. Technol. 2016, 6, 602–616. [CrossRef] 13. Zacher, A.H.; Olarte, M.V.; Santosa, D.M.; Elliott, D.C.; Jones, S.B. A review and perspectives of recent bio-oil hydrotreating research. Green Chem. 2014, 16, 491–515. [CrossRef] 14. Wang, Y.; De, S.; Yan, N. Rational control of nano-scale metal-catalysts for biomass conversion. Chem. Commun. 2016, 52, 6210–6224. [CrossRef][PubMed] 15. Mortensen, P.M.; Grundwaldt, J.-D.; Jensen, P.A.; Knudsen, K.G.; Jensen, A.D. A review of catalytic upgrading of bio-oil to engine fuels. Appl. Catal. A Gen. 2011, 407, 1–19. [CrossRef] 16. Chang, J.; Danuthai, T.; Dewiyanti, S.; Wang, C.; Borgna, A. Hydrodeoxygenation of guaiacol over carbon-supported metal catalysts. ChemCatChem 2013, 5, 3041–3049. [CrossRef] 17. Nie, L.; de Souza, P.M.; Noronha, F.B.; An, W.; Sooknoi, T.; Resasco, D.E. Selective conversion of m-cresol to toluene over bimetallic Ni–Fe catalysts. J. Mol. Catal. A Chem. 2014, 388–389, 47–55. [CrossRef] 18. Sun, J.; Karim, A.M.; Zhang, H.; Kovarik, L.; Li, X.S.; Hensley, A.J.; McEwen, J.-S.; Wang, Y. Carbon-supported bimetallic Pd–Fe catalysts for vapor-phase hydrodeoxygenation of guaiacol. J. Catal. 2013, 306, 47–57. [CrossRef] 19. Huang, X.; Atay, C.; Koranyi, T.I.; Boot, M.D.; Hensen, E.J.M. Role of Cu–Mg–Al mixed oxide catalysts in lignin depolymerization in supercritical ethanol. ACS Catal. 2015, 5, 7359–7370. [CrossRef] 20. Do, P.T.M.; Foster, A.J.; Chen, J.; Lobo, R.F. Bimetallic effects in the hydrodeoxygenation of meta-cresol on

γ-Al2O3 supported Pt–Ni and Pt–Co catalysts. Green Chem. 2012, 14, 1388–1897. [CrossRef] 21. Parsell, T.H.; Owen, B.C.; Klein, I.; Jamell, T.M.; Marcum, C.L.; Haupert, L.J.; Amundson, L.M.; Kenttämaa, H.I.; Ribeiro, F.; Miller, J.T.; et al. Cleavage and hydrodeoxygenation (HDO) of C-O bonds relevant to lignin conversion using Pd/Zn synergistic catalysts. Chem. Sci. 2013, 4, 806–813. [CrossRef] 22. Gao, D.; Xiao, Y.; Varma, A. Guaiacol hydrodeoxygenation over platinum catalyst: Reaction pathways and kinetics. Ind. Eng. Chem. Res. 2015, 54, 10638–10644. [CrossRef] 23. Zanuttini, M.S.; Dalla Costa, B.O.; Querini, M.A.; Peralta, M.A. Hydrodeoxygenation of m-cresol with Pt supported over mild acid materials. Appl. Catal. A Gen. 2014, 482, 352–361. [CrossRef] 24. Zhu, X.; Lobban, L.L.; Mallinson, R.G.; Resasco, R.E. Bifunctional transalkylation and hydrodeoxygenation of anisole over a Pt/H-Beta catalyst. J. Catal. 2011, 281, 21–29. [CrossRef] 25. He, L.; Qin, Y.; Lou, H.; Chen, P. Highly dispersed molybdenum carbide nanoparticles supported on activated carbon as an efﬁcient catalyst for the hydrodeoxygenation of vanillin. RSC Adv. 2015, 5, 43141–43147. [CrossRef] 26. Yoon, J.S.; Lee, T.; Choi, J.-W.; Suh, D.J.; Lee, K.; Ha, J.-M.Y.; Choi, J. Layered MWW zeolite-supported Rh catalysts for the hydrodeoxygenation of lignin model compounds. Catal. Today 2017, 293–294, 142–150. [CrossRef] 27. Zhang, C.; Xing, J.; Song, L.; Xin, H.; Lin, S.; Xing, L.; Li, X. Aqueous-phase hydrodeoxygenation of lignin monomer eugenol: Inﬂuence of Si/Al ratio of HZSM-5 on catalytic performances. Catal. Today 2014, 234, 145–152. [CrossRef] Catalysts 2017, 7, 265 38 of 40

28. Oh, S.; Hwang, H.; Choi, H.S.; Choi, J.W. Investigation of chemical modifications of micro-and macromolecules in bio-oil during hydrodeoxygenation with Pd/C catalyst in supercritical ethanol. Chemosphere 2014, 117, 806–814. [CrossRef][PubMed] 29. Jiang, H.; Yu, X.; Peng, X.; Zhang, H.; Nie, R.; Lu, X.; Zhou, D.; Xia, Q. Efficient aqueous hydrodeoxygenation of vanillin over mesoporous carbon-nitride modified Pd nanoparticles. RSC Adv. 2016, 6, 69045–69051. [CrossRef] 30. Oh, S.; Hwang, H.; Choi, H.S.; Choi, J.W. The effects of noble metal catalysts of the bio-oil quality during the hydrodeoxygenative upgrading process. Fuel 2015, 153, 535–543. [CrossRef] 31. Gao, D.; Schweitzer, C.; Hwang, H.T.; Varma, A. Conversion of guaiacol on noble metal catalysts: Reaction performance and deactivation studies. Ind. Eng. Chem. Res. 2014, 53, 18658–18667. [CrossRef] 32. Zhu, X.; Mallinson, R.G.; Resasco, D.E. Role of transalkylation reactions in the conversion of anisole over HZSM-5. Appl. Catal. A 2010, 379, 172–181. [CrossRef] 33. Olcese, R.N.; Francois, J.; Bettahar, M.M.; Petitjean, D.; Dufour, A. Hydrodeoxygenation of guaiacol, a surrogate of lignin pyrolysis vapors, over iron based catalysts: Kinetic and modelling of the lignin to aromatic integrated process. Energy Fuels 2013, 27, 975–984. [CrossRef] 34. Tran, N.T.T.; Uemuyra, Y.; Chowdhury, S.; Ramli, A. Vapor-phase hydrodeoxygenation of guaiacol on Al-MCM-41 supported Ni and Co catalysts. Appl. Catal. A Gen. 2016, 512, 93–100. [CrossRef] 35. Boscagli, C.; Raffelt, K.; Zevaco, T.A.; Olbrich, W.; Otto, T.N.; Sauer, J.; Grunwealdt, J.D. Mild hydrotreatment of the light fraction of fast-pyrolysis oil produced from straw over nickel-based catalysts. Biomater. Bioeng. 2015, 83, 525–538. [CrossRef] 36. Prasomsri, T.; Shetty, M.; Murugappan, K.; Roman Leshkov, Y. Insights into the catalytic activity and surface

modiﬁcation of MoO3 during the hydrodeoxygenation of lignin-derived model compounds into aromatic hydrocarbons under low hydrogen pressures. Energy Environ. Sci. 2014, 7, 2660–2669. [CrossRef] 37. Ghampson, I.T.; Sepuldeva, C.; Garcia, R.; Fierro, J.L.G.; Escalona, N. Carbon nanoﬁber-supported ReOx catalysts for the hydrodeoxygenation of lignin-derived compounds. Catal. Sci. Technol. 2016, 6, 4356–4369. [CrossRef] 38. Sepuldeva, C.; Leiva, K.; Garcia, R.; Radovic, L.R.; Ghampson, I.T.; DeSisto, W.J.; Garcia Fierro, J.L.;

Escalona, N. Hydrodeoxygenation of 2-methoxyphenol over Mo2N catalysts supported on activated carbons. Catal. Today 2011, 172, 232–239. 39. Lodeng, R.; Ranga, C.; Rajkhowa, T.; Alexiadis, V.I.; Bjorkan, H.; Chytil, S.; Svenum, I.H.; Walmsley, J.;

Thybaut, J.W. Hydrodeoxygenation of phenolics on liquid phase over supported MoO3 and carburized analogous. Biomass Convers. Biorefin. 2017, 1–17. [CrossRef] 40. Shafaghat, H.; Rezaei, P.S.; Wan Daud, W.M.A. Catalytic hydrodeoxygenation of simulated phenolic bio-oil to cycloalkane and aromatic hydrocarbons over bifunctional metal/acid catalysts of Ni/Beta, Fe/H-Beta and NiFe/HBeta. J. Ind. Eng. Chem. 2016, 35, 268–276. [CrossRef] 41. Nimmanwudipong, T.; Aydin, C.; Lu, J.; Runnebaum, R.C.; Brodwater, K.C.; Browning, N.D.; Block, D.E.; Gates, B.C. Selective hydrodeoxygenation of guaiacol catalyzed by platinum supported on magnesium oxide. Catal. Lett. 2012, 142, 1190–1196. [CrossRef]

42. Chen, C.-J.; Lee, W.-S.; Bhan, A. Mo2C catalyzed vapor phase hydrodeoxygenation of lignin-derived phenolic compound mixtures to aromatics under ambient pressure. Appl. Catal. A Gen. 2016, 510, 42–48. [CrossRef] 43. Lee, W.-S.; Wang, Z.; Wu, R.J.; Bhan, A. Selective vapor-phase hydrodeoxygenation of anisole to benzene on molybdenum carbide catalysts. J. Catal. 2014, 319, 44–53. [CrossRef] 44. Lu, Q.; Chen, C.-J.; Luc, W.; Chen, J.G.; Bhan, A.; Jiao, F. Ordered mesoporous metal carbides with enhanced anisole hydrodeoxygenation selectivity. ACS Catal. 2016, 6, 3506–3514. [CrossRef] 45. Lee, C.R.; Yoon, J.S.; Suh, Y.W.; Choi, J.-W.; Ha, J.-M.; Suh, D.J.; Park, Y.-K. Catalytic roles of metals and supports on hydrodeoxygenation of lignin monomer guaiacol. Catal. Commun. 2012, 17, 54–58. [CrossRef] 46. Bui, V.N.; Laurenti, D.; Afanasiev, P.; Geantet, C. Hydrodeoxygenation of guaiacol with CoMo catalysts. Part I: Promoting effect of cobalt on HDO selectivity and activity. Appl. Catal. B Environ. 2011, 101, 239–245. [CrossRef] 47. Mortensen, P.M.; Grunwaldt, J.-D.; Jensen, P.A.; Knudsen, G.; Jensen, A.D. Stability and resistance of nickel catalysts for hydrodeoxygenation: Carbon deposition and effects of sulfur, potassium, and chlorine in the feed. Catal. Sci. Technol. 2014, 4, 3672–3686. [CrossRef] Catalysts 2017, 7, 265 39 of 40

48. Gonzalez-Borja, M.A.; Resasco, D.E. Anisole and guaiacol hydrodeoxygenation over monolithic Pt-Sn catalysts. Energy Fuels 2011, 45, 4155–4162. [CrossRef] 49. Deepa, A.K.; Dhepe, P.L.Function of metals and supports on the hydrodeoxygenation of phenolic compounds. ChemPlusChem 2014, 79, 1573–1583. [CrossRef] 50. Peters, J.E.; Carpenter, J.R.; Dayton, D.C. Anisole and guaiacol hydrodeoxygenation reaction pathway over selected catalysts. Energy Fuels 2015, 29, 909–916. [CrossRef] 51. Shetty, M.; Murugappan, K.; Green, W.H.; Roman-Leshkov, Y. Structural properties and reactivity trends of molybdenum oxide catalysts supported on zirconia for the hydrodeoxygenation of anisole. ASC Sustain. Chem. Eng. 2017, 5, 5293–5301. [CrossRef] 52. Yoon, J.S.; Lee, Y.; Yu, J.; Kim, Y.-A.; Park, E.D.; Choi, J.-W.; Ha, J.-M.; Suh, D.J.; Lee, H. Production of high carbon number hydrocarbon fuels from a lignin-derived-O-4 phenolic dimer, benzyl phenyl ether, via isomerization of ether to alcohols on high-surface-area silica-alumina aerogel catalyst. Appl. Catal. B Environ. 2013, 142–143, 668–676. [CrossRef] 53. Strassberger, Z.; Alberts, A.H.; Louwerse, M.J.; Tanase, S.; Rothenberg, G. Catalytic cleavage of lignin β-O-4 link mimics using copper on alumina and magnesia-alumina. Green Chem. 2013, 15, 768–774. [CrossRef] 54. Jongerius, A.L.; Gosselink, R.W.; Djikstra, J.; Bitter, J.H.; Bruijnincx, P.C.-A.; Weckhuysen, B.M. Carbon nanofiber supported transition-metal carbide catalysts for the hydrodeoxygenation of guaiacol. ChemCatChem 2013, 5, 2964–2972. [CrossRef] 55. Echeandia, S.; Arias, P.L.; Barrio, V.L.; Pawelec, B.; Fierro, J.L.G. Synergy effect in the HDO of phenol over Ni–W catalysts supported on active carbon: Effect of tungsten precursors. Appl. Catal. B 2010, 101, 1–12. [CrossRef] 56. Santillan-Jimenez, E.; Perdu, M.; Pace, R.; Morgan, T.; Crocker, M. Activated carbon, carbon nanoﬁber and carbon nanotube supported molybdenum carbide catalysts for the hydrodeoxygenation of guaiacol. Catalysts 2015, 5, 424–441. [CrossRef] 57. Ohta, H.; Kobayashi, H.; Hara, K.; Fukuoka, A. Hydrodeoxygenation of phenols as lignin models under acid-free conditions with carbon-supported platinum catalysts. Chem. Commun. 2011, 47, 12209–12211. [CrossRef][PubMed] 58. Ghampson, T.; Sepuldeva, C.; Garcia, R.; Radovic, L.R.; Garcia Fierro, J.L.; DeSisto, W.J.; Escalona, N. Hydrodeoxygenation of guaiacol over carbon-supported molybdenum nitride catalysts: Effects of nitriding and support properties. Appl. Catal. A Gen. 2012, 439–440, 111–124. [CrossRef]

59. Chen, C.-J.; Bhan, A. Mo2C modiﬁcation by CO2,H2O, and O2: Effect of oxygen content and oxygen source on rates and selectivity of m-cresol hydrodeoxygenation. ACS Catal. 2017, 7, 1113–1122. [CrossRef] 60. Zhu, Z.; Tan, H.; Wang, J.; Yu, S.; Zhou, K. Hydrodeoxygenation of vanillin as a bio-oil model over carbonaceous microspheres-supported Pd catalysts in the aqueous phase and Pickering emulsions. Green Chem. 2014, 16, 2636–2643. [CrossRef] 61. Gonzalez, C.; Marin, P.; Diez, F.V.; Ordonez, S. Gas-phase hydrodeoxygenation of benzaldehyde, benzyl alcohol, phenyl acetate, and anisole over precious metal catalysts. Ind. Eng. Chem. Res. 2016, 55, 2319–2327. [CrossRef] 62. Zhang, X.; Zhang, Q.; Wang, T.; Ma, T.; Yu, Y.; Chen, L. Hydrodeoxygenation of lignin-derived phenolic

compounds to hydrocarbons over Ni/SiO2-ZrO2 catalysts. Bioresour. Technol. 2013, 134, 73–80. [CrossRef] [PubMed] 63. Liu, X.; Xu, L.; Xu, G.; Jia, W.; Ma, Y.; Zhang, Y. Selective hydrodeoxygenaiton of lignin-derived phenols to cyclohexanols or cyclohexanes over magnetic CoNx@NC catalysts under mild conditions. ASC Catal. 2016, 6, 7611–7620. 64. Gutierrez, A.; Turpeinen, E.-M.; Viljava, T.-R.; Krause, O. Hydrodeoxygenation of model compounds on

sulﬁded CoMo/γ Al2O3 and NiMo/γ-Al2O3 catalysts: Role of sulfur-containing groups in reaction networks. Catal. Today 2017, 285, 125–134. [CrossRef]

65. Viljava, T.-R.; Komulainen, S.; Selvam, T.; Krause, A.O.I. Stability of CoMo/Al2O3 catalysts: Effect of HDO cycles on HDS. Stud. Surf. Sci. Catal. 1999, 127, 145–152.

66. Senol, O.I.; Viljava, T.-R.E.; Krause, A.O.I. Hydrodeoxygenation of methyl esters on sulphided NiMo/Al2O3 and CoMo/Al2O3 catalysts. Catal. Today 2005, 100, 331–335. [CrossRef] 67. Aho, A.; Kumar, N.; Eränen, K.; Salmi, T.; Hupa, M.; Murzin, D.Y. Catalytic pyrolysis of woody biomass in a ﬂuidized bed reactor: Inﬂuence of zeolite structure. Fuel 2008, 87, 2493–2501. [CrossRef] Catalysts 2017, 7, 265 40 of 40

68. Johansson, A.; Aaltonen, O.; Ylinen, P. Organosolv pulping—Methods and pulp properties. Biomass 1987, 13, 45–65. [CrossRef] 69. Mansouri, N.-E.E.; Salvado, J. Structural characterization of technical lignins for the production of adhesives: Application to lignosulfonate, kraft, soda-anthraquinone, organosolv and ethanol process lignins. Ind. Crops Prod. 2006, 24, 8–16. [CrossRef] 70. Boot, M.; Frijters, C.; Lujitern, C.; Somers, B.; Baert, R.; Donkerbroek, A.; Klein-Douwel, R.J.H.; Dam, N. Cyclic oxygenates: A new class of second-generation biofuels for diesel engines? Energy Fuels 2009, 23, 1808–1817. [CrossRef] 71. Lee, K.; Gu, G.H.; Mullen, C.A.; Boateng, A.A.; Vlachos, D.G. Guaiacol hydrodeoxygenation mechanism on Pt(111): Insights from density functional theory and linear free energy relations. ChemSusChem 2015, 8, 315–322. [CrossRef][PubMed] 72. Verma, A.M.; Kishore, N. DFT analyses of reaction pathways and temperature effects on various guaiacol conversion reactions in gas phase environment. Chem. Select 2016, 1, 6196–6205. [CrossRef] 73. Liu, C.; Zhang, Y.; Huang, Y.; Huang, X. Study of guaiacol pyrolysis mechanism based on density function theory. Fuel Process. Technol. 2014, 123, 159–165. [CrossRef] 74. Chiu, C.-C.; Genest, A.; Borgna, A.; Rösch, N. Hydrodeoxygenation of guaiacol over Ru (0001): A DFT study. ACS Catal. 2014, 4, 4178–4288. [CrossRef] 75. Lu, J.; Heyden, A. Theoretical study of the reaction mechanism of the hydrodeoxygenation of guaiacol over a Ru(0001) model surface. J. Catal. 2015, 321, 39–50. [CrossRef] 76. Wheeler, C.; Jhalani, A.; Klein, E.J.; Tummala, S.; Schmidt, L.D. The water–gas-shift reaction at short contact times. J. Catal. 2004, 223, 191–199. [CrossRef] 77. Gevert, B.S.; Otterstedt, J.-E.; Massoth, F.E. Kinetics of the HDO of methyl-substituted phenols. Appl. Catal. 1987, 31, 119–131. [CrossRef] 78. Prins, R.; Bussell, M.E. Metal phosphides: Preparation, characterization and catalytic reactivity. Catal. Lett. 2012, 142, 1413–1436. [CrossRef] 79. Popov, A.; Kondratieva, E.; Goupil, J.M.; Mariey, L.; Bazin, P.; Gilson, J.-P.; Travert, A.; Mauge, F. Bio-oils hydrodeoxygenation: Adsorption of phenolic molecules on oxidic catalyst supports. J. Phys. Chem. 2010, 114, 15661–15670. [CrossRef] 80. Badawi, M.; Paul, J.F.; Cristol, S.; Payen, E.; Romero, Y.; Richard, F.; Brunet, S.; Lambert, D.; Portier, X.; Popov, A.; et al. Effect of water on the stability of Mo and CoMo hydrodeoxygenation catalysts: A combined experimental and DFT study. J. Catal. 2011, 282, 155–164. [CrossRef]

81. Ravenelle, R.M.; Copeland, J.R.; Kim, W.G.; Crittenden, J.C.; Sievers, C. Structural changes of γ-Al2O3- supported catalysts in hot liquid water. ACS Catal. 2011, 1, 552–561. [CrossRef] 82. Kent, J.A. Handbook of Industrial Chemistry and Biotechnology; Springer: New York, NY, USA, 2012; Volume 18. 83. Bartholomew, C.H. Mechanisms of deactivation. Appl. Catal. A Gen. 2001, 212, 17–60. [CrossRef] 84. Snytnikov, P.V.; Sobyanin, V.A.; Belyaev, V.D.; Tsyrulnikov, P.G.; Shitova, D.A. Selective oxidation of carbon monoxide in excess hydrogen over Pt-, Ru-, and Pd-supported catalsyts. Appl. Catal. A Gen. 2003, 239, 149–156. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).