Sie sind auf Seite 1von 19

Review pubs.acs.

org/IECR

Hydroprocessing in Aqueous Phase


Edward Furimsky*
IMAF Group 184 Marlborough Avenue Ottawa, Ontario, Canada K1N 8G4 ABSTRACT: A large consumption of H2 aects the overall economy of conventional hydroprocessing. The costs can be decreased by using water as the source of active hydrogen. This can be achieved under subcritical and supercritical water conditions providing that an active and stable catalyst is developed. Hydroprocessing in aqueous phase has been studied for potential applications in upgrading of high oxygen content feeds and heavy petroleum feeds to liquid hydrocarbons. The feeds were tested at temperatures ranging from less than 200 to 500 C and total pressure from 1 to 30 MPa. These conditions cover subcritical and supercritical regions of water. Water takes part in hydroprocessing reactions as a free radical scavenger and a hydrogen donor. Hydrogen generated in situ via partial reforming and watergas shift reactions is more reactive than external hydrogen. Catalyst development for hydroprocessing in aqueous phase has been receiving much attention. High performance was observed over the catalysts containing noble metals (Pt, Pd, Ru, and Rh) supported on various supports; however, the information on a long-term stability of these catalysts is limited.

1. INTRODUCTION Hydroprocessing (HPR) is the most important route for upgrading petroleum and nonpetroleum feeds to commercial fuels. It involves conversion of compounds containing contaminants such as sulfur, nitrogen, and metals to hydrocarbons via reactions with hydrogen. In some liquids, a nal polishing step is required to attain specication of transportation fuels and lubricants. For example, aromatics must be removed by hydrogenation while straight chain hydrocarbons by hydroisomerization. The cost of conventional HPR is aected by a large consumption of hydrogen. Water has been identied as an alternative source of hydrogen. Both subcritical and supercritical water (SCW) conditions have been attracting attention. Water is an important constituent of the feeds produced via hydrothermal liquefaction and pyrolysis of biomass. In this case, operating conditions and/or type of biomass dictate that liquid products are obtained in an aqueous medium. Separation of the water-soluble components from the aqueous phase may be dicult and inecient. Therefore, conversion of polar compounds to hydrocarbons directly in the same environment may be more advantageous.1 In such applications, HPR is the method of interest. Once polar components in the feed are converted to hydrocarbons, the separation of hydrophobic phase from the aqueous phase is simple. A similar approach may be applied for upgrading of the aqueous phase separated from the primary products obtained during FischerTropsch synthesis (FTS). This byproduct may contain up to 10 wt % of dissolved oxygenates. Direct removal of these oxygenates from aqueous phase via catalytic route has also been attracting attention.3,4 Depending on the method of production, petroleum crudes may be obtained in the mixture with large quantities of water. This may be the case of heavy crudes produced during the enhanced-oil-recovery using steam ooding method and via hot water separation process employed during the bitumen production from tar sands.5 A direct conversion of such crudes (without dewatering) via HPR may be a potential route for
2013 American Chemical Society

primary upgrading. A unique case of an aqueous phase may be slurry bed hydrocracking (HCR) of heavy feeds. In this case, a catalyst dissolved in water is coslurried with feed before entering the reactor. The information on conventional HPR methods has been extensively reviewed elsewhere.6 All reactions occurring in parallel during the HPR of conventional feeds, that is, hydrodesulfurization (HDS), hydrodenitrogenation (HDN), hydrodeoxygenation (HDO), hydrocracking (HCR), hydrogenation (HYD), hydroisomerization (HIS), hydrodemetallization (HDM), and hydrodeasphaltization (HDAs), have been discussed in details. In addition, the most important HPR reactions were reviewed separately (i.e., HDS,68 HDN,911 HDO,12,13 HYD,14 HIS,2 HCR,2,15 HDM15,16 and HDAs15,16). Some similarities in the mechanisms of these reactions in the presence of water may be anticipated. However, rather than repeat this information here, the main focus of this review is on potential role of water in modifying the mechanism of HPR. Also, the eect of water on operating parameters under aqueous conditions requires attention. Of particular signicance are the eects of water on catalyst activity and stability. In this regard, the advances in catalyst development for applications in aqueous phase are one of the objectives of this review.

2. PROPERTIES OF WATER During the HPR in aqueous phase, water plays an important role as both solvent and reactant. Chemical and physical properties of water as well as their change with temperature and pressure were described in details elsewhere.17 This included the properties in subcritical and supercritical regions. Thus, under mild conditions, a direct involvement of water in HPR reactions may be much less evident. For HPR, the miscibility and/or solubility of various feeds (including H2) in water as
Received: Revised: Accepted: Published:
17695

October 15, 2013 October 30, 2013 November 21, 2013 November 21, 2013
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research well as diusivity and reactivity of water are of prime interests. For the purpose of this review, only a brief account of these properties is given in the following text. Thus, rather extensive information on these and other aspects of water may be readily accessed in several books published elsewhere.1719 The original structure of liquid water, dominated by hydrogen bonds, is changing with increasing temperature. While critical temperature is being approached, an almost complete collapse of the hydrogen bond network occurs. As the result of this change, polarity of water is signicantly diminished. This was conrmed by a dramatic decrease in dielectric constant.1719 Above the critical point, water behaves as a nonpolar medium, capable of dissolving organic substrates. In this regard, water is approaching properties of solvents such as acetone, methanol, ethanol, etc. Then, the solubility of various feeds (e.g., bio oils, petroleum residues, coal derived liquids, etc.) in SCW is signicantly enhanced.19 From the HPR point of view, it is important that under super critical conditions, gaseous H2 is completely miscible with SCW. A high homogeneity of reaction streams attained in SCW is favorable for the ecient transfer of hydrogen to reactant molecules. Above the critical point, water behaves as a dense gas while still retaining some characteristics (e.g., density) of liquid water. This behavior is the reason for a high diusivity and unique transportation properties of SCW. While increasing temperature from subcritical region toward critical temperature at 22 MPa, the density of water abruptly decreases, for example, from about 0.6 g/mL at 350 C to less than 0.2 g/mL at 374 C. Although to a lesser extent, the SCW density further decreased with temperature increase above critical point temperature. This density decrease may be oset by increasing pressure.19 Some eect of density of subcritical water and SCW on HPR reactions may be anticipated. Then, if necessary, an optimal combination of density with temperature and pressure may be established. It is obvious that the reactivity of water may change dramatically as the consequence of hydrogen bonds network collapse. For example, while approaching 374 C, the pKw of the water dissociation equilibrium almost doubled.1719 A much higher concentration of H3O+ and HO ions in SCW than that in liquid water increases the chances for the involvement of these ions during HPR. For example, H3O+ ions tend to add readily to heteroatoms such as S, N, and O.6,9 An interaction of HO ions with carbons, particularly those attached to heteroatoms, may be anticipated.20 These facts increase the probability of an ionic mechanism as part of the overall mechanism during the HPR in aqueous phase. Ionic reactions are favored by high density of water. This may be achieved under subcritical conditions; however, under supercritical conditions, high pressures are needed to get densities suitable for ionic chemistry.

Review

The objective of research in this eld was a direct upgrading of such complex mixtures without pretreatment. The most typical source of high water content feeds is the conversion of biomass (both by pyrolysis and hydrothermal treatment) always to a high water content biocrude. Detailed accounts of the conversion of an aquatic biomass via hydrothermal liquefaction and gasication, both in the presence and absence of catalysts was given by Yeh et al.21 and Savage.22 An example of the biocrude (primary liquids) from pyrolysis and liquefaction of lignocellulosic biomass is shown in Table 1.23 At least two stages may be needed to upgrade such feeds to Table 1. Property Ranges of Bio-crude Obtained by Liquefaction and Pyrolysis of Biomass23
liquefaction carbon, wt % sulfur + nitrogen, wt % oxygen,a wt % water in crude, wt % density, g/cm3
a

pyrolysis 5666 0.1 2738 2452 1.111.23

6881 0.1 925 625 1.101.14

Dry basis.

hydrocarbons via HPR route.24 Chemical compositions of these biocrudes and corresponding products were discussed in details elsewhere.13 Unless an extensive dewatering was conducted, a high water content in the biocrude obtained from algae biomass and municipal solid wastes using similar methods may be anticipated.12 Interests in the catalytic conversion of sorbitol to a great variety of products have been noted. The HPR of sorbitol in an aqueous phase to produce hydrocarbons has been one of the evaluated routes.25 A unique case of the feed for potential HPR in the presence of water may be the reaction water produced in FTS process. Such aqueous phase contains a mixture of water-soluble oxygenates, among which alcohols, ketones, aldehydes, and carboxylic acids are far predominant structures. The total amount of the oxygenates in the aqueous phase may approach 10 wt %. The most abundant oxygenates dissolved in the aqueous phase from FTS are shown in Table 2.24 Table 2. Most Abundant Oxygenates in Aqueous Product from High Temperature FTS2
rank 1 2 3 4 5
a

oxygenate ethanol propanone butanone 1-propanol acetic acid

yield, mass %a 3.4 2.5 1.2 1.0 0.9

On total syncrude basis.

3. PROPERTIES OF FEEDS In this section, attention is being paid to those feeds that have been included in the studies on HPR in aqueous phase. In this regard, model compounds alone and/or mixtures of various model compounds used to study HPR under conventional conditions have been also studied under aqueous phase conditions. This is illustrated on several examples presented in this review. In the case of real feeds, the focus is on those feeds which are being produced in an aqueous environment.
17696

Besides the high oxygen containing feeds, heavy petroleum feeds have been focus of attention for potential upgrading via HPR in an aqueous phase. Table 35 indicates a signicant dierence between the properties of the latter feeds and the high oxygen content feeds shown in Tables 1 and 2. A strong emulsifying potential of some asphaltenic and resinous components suggests that heavy feeds produced via enhanced oil recovery using steam ooding method may contain a large amount of water in the form of stable emulsions.5 Similarly, relatively high water content bitumen may be produced from tar sands using a hot water separation process.15 If necessary,
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research Table 3. Properties of Heavy Feedsa for Hydroprocessing in Aqueous Phase5
heavy feed Maya density, kg/L sulfur, wt % nitrogen, wt % vanadium, ppm nickle, ppm CCRb, wt %
a

Review

Cold Lake 1.0 4.9 0.6 160 80 19

Arab heavy 0.89 2.9 0.2 50 16 7

0.93 3.8 0.3 273 50 15

Vacuum residues. bConradson carbon residue.

stable emulsions involving heavy feeds and water may be prepared with the aid of surface agents using mechanical means. An atomization of heavy feeds is achieved by water evaporation during rapid temperature increase, that is, during the introduction of water-heavy feed emulsion into HPR reactor. At temperatures employed during HCR of heavy feeds (e.g., >400 C), water molecules can stabilize free radicals and transfer hydrogen to products and, as such, oset high hydrogen consumption. A water-soluble catalyst may be added before the emulsion preparation. A catalytic process referred to as the slurry bed HCR has been approaching commercial stage.5 Properties of the heavy feeds, which can be upgraded by this method as well as the operating parameters of slurry bed reactor, were discussed elsewhere.15 In practical situation, a number of other high water content feeds can be identied. For example, such feeds may be disposed from various industrial operations and municipalities (e.g., renery sludge, waste from pulp and paper industry, municipal solid waste, etc.). Little information on the upgrading of such materials under aqueous conditions could be found in the literature, so far. Apparently, HPR in an aqueous medium may be an attractive option, although the homogeneity of these feeds for processing may require attention. Other feeds that can be upgraded under aqueous phase conditions include coal derived pitch, waste plastics, etc. With respect to HPR, the miscibility and/or solubility of organic phase in water phase is of a primary importance. High oxygen content feeds are usually in the form of a homogeneous mixture of water with an organic phase. However, signicant problems may be encountered during the preparation of the waterheavy feed mixtures for HPR under aqueous conditions. For such feeds, a temperature exceeding that of SCW is required to achieve desirable conversion. Also, an optimal combination of pressure and temperature has to be identied to ensure homogeneity of the system. The phase diagram in Figure 126 developed for heavy feed containing 37 wt % of asphaltenes shows three miscibility regions, that is, nonmiscible two phase region, partially miscible two phase region, and pseudo-single phase region. It was evident that the regions are inuenced by the water/feed ratio. The dashed region, which is suitable for upgrading of the heavy feed, is, in fact, the SCW region. Mechanistic aspects of the conversion of various feeds under subcritical and supercritical conditions are discussed in more details later in the review.

Figure 1. Phase structure of petroleum residue in subcritical and supercritical water; water/residue (1) 1/4; (2) 1/2.

200 to almost 500 C and 440 MPa, respectively.2636 Processing under such conditions may have some energetic advantages. Thus, avoiding the liquid water-steam phase transformation results in substantial energy savings the extent of which is inuenced by severity. For the purpose of this review, the severity of conditions is referred to temperature ranges employed, i.e., mild, subcritical and supercritical (e.g., below 300 C, 300374 C, and above 374 C, respectively). Under subcritical conditions, a high pressure is required to ensure that most of the water in the system is in a liquid phase. Besides energy savings, this improves the interaction of water molecules with reactants thus ensuring a higher conversion of the latter. Rather unique properties are exhibited by water in the supercritical region (above 374 C and 22 MPa).26 The products from upgrading in an aqueous phase include H2, synthesis gas (H2+CO), gaseous and liquid hydrocarbons. The conditions employed during the upgrading may be optimized to maximize the yield of products of interest.36 The production of liquid hydrocarbons via HPR in an aqueous phase is the primary focus of this review. The presence of large quantities of water in the system indicates on the occurrence of reactions (Figure 2), which

Figure 2. Tentative reactions during biomass reforming.

4. UPGRADING UNDER AQUEOUS CONDITIONS The information in literature suggests that, in most cases, temperature and pressure employed during the upgrading under aqueous conditions of various feeds range from less than
17697

under conditions of conventional HPR are either absent or play a minor role. The reforming of hydrocarbons with the aid of steam (reaction {1}) may be one of the hydrogen sources generated in situ. The high yield of CO2, combined with the low yield of CO conrmed the involvement of the watergas shift (WGS) reaction {2}, which may always be present as the next step of reforming reactions. A higher reactivity of the in
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research

Review

Figure 3. Structure of asphaltenes derived from (a) Athabasca bitumen; (b) Maya crude. Reprinted with permision from ref 50. Copyright 2005, Elsevier.

situ generated hydrogen than that of the external gaseous H2 may be anticipated. As it was pointed out, production of liquid hydrocarbons from various feeds under aqueous conditions is the primary objective of this review. In the case of glucose, this may be indicated by reaction {3}. In addition, reaction {4} (methanation) is also involved. The hydrogen required for the reactions {3} and {4} would be supplied in situ via reactions {1} and {2} as well as using an external source (gaseous H2). During the HPR in an aqueous phase, the reactions {1} to {4} occur in parallel. The extent of these reactions depends on the experimental conditions (e.g., temperature, total pressure, type of experimental system, origin of feed, etc.) and type of catalyst. The success of HPR in aqueous phase for the production of liquid fuels depends on the optimization of experimental conditions to ensure the maximization of reaction {3} and minimization of reaction {4}. A large hydrogen consumption in reaction {3} should be noted. This may be oset by the in situ hydrogen production via reactions {1} and {2}. Therefore, an ecient process for conversion of various feeds to liquid fuels using a concept based on the HPR in aqueous phase may require a delicate balance involving a number of operating parameters. These issues were discussed extensively in the study published by Davda et al.37 In this regard, signicant advancements in the understanding of reactions occurring under aqueous phase conditions were made by Dumesic and co-workers.3743 For example, at 483 and 498 K, Pt and Pd supported on silica were selective for
17698

generation of H2, while Rh, Ru, and Ni supported on silica exhibited a low selectivity for H2 and a high selectivity for alkane production.42 The selectivity for H2 production was also inuenced by the structure of substrate; for example, it increased from sugars toward ethylene glycol and methanol.43 For ethanol, ethylene glycol, and sorbitol (below 500 K, 3 MPa, Pt/Al2O3 and a tin-promoted Raney-Ni catalysts), almost complete suppression of reaction {4} in favor or reactions {1} and {2} could be achieved.38,39 On the other hand, methanation of ethanol (7.5 wt % in SCW) was dominant reaction over Ru/C catalyst at 400 C and 24.5 MPa.44,45 Lercher and co-workers29,3336 made an important contribution to the understanding of the HPR under aqueous conditions using model compounds typical of those present in biofeeds. In this case, hydrocarbons were the targeted products. Therefore, the conditions favoring reaction {3} were the focus of their attention. More detailed accounts of these studies are given in the latter sections of this review. Huber et al.40 conducted extensive evaluations of various catalysts in a wide range of experimental conditions. This study is introduced here to illustrate the attempt for identifying optimal conditions ensuring high yields of liquid hydrocarbons (reaction {3}). Using Pt(4%)/SiO2Al2O3 catalyst (498 K; 4 MPa; continuous system) and 5 wt % sorbitol in water, the combined yield of pentane and hexane approached 60 wt % without external H2 being present. The yield was further increased to about 80% in the presence of external H2.
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research

Review

Figure 4. Tentative structure of soft wood lignin. Reprinted with permission from ref 49. Copyright 2012, Elsevier.

For most part, the discussions on aqueous phase upgrading was focusing on high oxygen feeds. To various extents, reactions {1} to {4} are also present during the upgrading of petroleum residues, coal tar pitch, waste plastics, etc. For such feeds, thermal cracking of large molecules to light fractions may play a dominant role during the overall conversion to liquid hydrocarbons. Apparently, HPR is not the only option for liquid fuels production under aqueous conditions. For example, the concept based on the conversion of polysacharides via dehydration, aldol-condensation, and hydrogenation can produce liquid alkanes under mild conditions.46 In addition, a high yield of synthesis gas (CO+H2) may be generated if reaction {1} is carried out under controlled conditions. Liquid fuels can then be produced via FTS using synthesis gas as the feed.47 It should be noted that, for this alternative, all stages of the FTS process and subsequent upgrading of products have been used on a commercial scale.24 There might be another potential non-HPR routes for production of liquid hydrocarbons under aqueous conditions.48 These routes are not in the scope of the present review. However, they should be always considered as an alternative to HPR while evaluating the viability of upgrading under aqueous phase conditions. In some specic cases (e.g., for renery applications), the production of H2 (reaction {1}) may be attractive. In other cases, the production of synthetic natural gas via methanation (reaction {4}) may also be of an interest.

5. HYDROPROCESSING MECHANISM IN AQUEOUS PHASE For the purpose of this review, HPR reactions are those in which both an in situ produced hydrogen and an external
17699

gaseous H2 are involved. Participation of the former was indicated by the reactions {1} to {4}. In addition, H2O molecules may be directly involved in supplying hydrogen. During the HPR under aqueous conditions, both noncatalytic and catalytic reactions occur simultaneously. For the overall mechanism of HPR, decoupling the noncatalytic reactions from catalytic reactions may be of an interest. The extent of the former reactions increases during the temperature increase from mild conditions to subcritical and nally to supercritical region. Because of the unique physical state of SCW, the occurrence of entirely new chemical reactions (e.g., ionic reactions) between SCW and substrate in the latter region may be anticipated. Additional noncatalytic reactions occur under SCW + H2 conditions. It is believed that, under mild conditions, rather low conversion of feed involving water should be observed unless an active catalyst is present. Once a catalyst is present, a new set of reactions becomes evident even under mild conditions. Although a primary focus is on catalytic HPR, a brief account of noncatalytic reactions occurring in parallel with catalytic reactions may be useful for better comprehending the overall mechanism of HPR under aqueous conditions. As it was pointed out earlier, nonpetroleum high oxygen content feeds and petroleum residues are two main types of the feeds that have been used for the HPR in an aqueous phase. Figures 349 and 450 respectively show tentative structures of lignin and asphaltenes. Lignin containing feeds may be used for the HPR in aqueous phase either directly or as the source of a bio-oil after upgrading (e.g., via pyrolysis and liquefaction). A high oxygen content ensures a good miscibility/solubility in the aqueous phase. On the other hand, vigorous mixing of heavy feeds of petroleum origin with water may be required to obtain
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research a homogeneous mixture suitable for HPR. The structures in Figures 3 and 4 clearly indicate a signicant dierence between the overall mechanism of HPR of these feeds in aqueous phase. 5.1. Noncatalytic Reactions. Because of higher temperatures involved, thermal cleavage of chemical bonds in various reactants in SCW (above 374 C) is more extensive than under subcritical conditions.51 The thermal cleavage of the weak CC bonds begins at bout 300 C. It was reported that free radicals produced by the cleavage of organic bonds were rapidly stabilized in the presence of water via the set of tentative reactions shown in Figure 5.52,53 Based on the models in

Review

Figure 5. Tentative reactions during hydroprocessing in aqueous phase.

Figures 3 and 4, the radical (R) may involve complex structures comprising aromatic and aliphatic entities as well as heteroatoms. Once generated, such radicals may decompose to lighter products. This suggests that in the presence of a free radicals generating agent, the conversion of large reactant molecules could be enhanced. This was indeed conrmed in the study of Zhu et al.54 who added ditert-butyl peroxide to a heavy feed for pyrolysis in SCW (653 K; water density 0.30 kg/ L). Consequently, the overall conversion of aspahltenes and resins was increased. However, decomposition of large radicals involves a parallel formation of lighter products and smaller radicals. The latter may lead to coke formation unless they are stabilized. This was conrmed by a signicant decrease in coke formation during copyrolysis of the heavy feed with polyethylene conducted at 683 K under otherwise similar conditions.54 Polyethylene is the source of paranic hydrogen, an excellent stabilizer of free radicals. The reactions {5} and {8} (Figure 5) suggest that hydrogen from H2O may be transferred to the feed and corresponding products. This was indeed conrmed by Dutta et al.57 using the mixture of H2O + D2O during thermal cracking of bitumen between 350 to 530 C. In this case, deuterium was transferred both to liquid products and coke. In the absence of water, the conversion of radicals would proceed via reaction {9} which represents the formation of coke. According to this mechanism, water acts as both radical scavenger and hydrogen donor. In recent study, Xu et al.58 concluded that SCW cannot donate hydrogen to reactants via the dissociation of H2O to HC and HOC radicals, followed by the reaction of radicals with a reactant. This reaction is energetically unfavorable and should be distinguished from the reaction in which hydrogen is abstracted from H2O by radical via reaction {5}. Ability of the radicals generating reactants to transfer hydrogen from H2O observed in their study may be attributed to reactions {7} and {8} (Figure 5). Similar observations were made during the decomposition of several S-containing compounds in SCW.59
17700

In this case, the experimental results could be interpreted in terms of free radicals mechanism. At suciently high temperatures, the reforming of hydrocarbons initiated by free radicals, according to the tentative reaction {1} may take place. For example, at about 600 C, the transfer of oxygen from water to carbon of a highly disintegrated organic molecules was a dominant reaction producing high yields of CO2 and H2.51 At such temperature of SCW, the total pressure may be in the range of 30 MPa. The high yields of CO2 and H2 combined with low yields of CO suggested that H2O reacted with CO via WGS reaction (reaction {2}). Because of the equilibrium eects, the excess of H2O in the system was favorable for WGS reaction while the presence of H2 had an opposing eect. Therefore, the consumption of H2 in HPR reactions, that is, its removal from equilibrium mixture, drives WGS reaction forward. It is believed that the extent of reforming and WGS reactions (reactions {1} and {2}) is being gradually diminished by decreasing temperature from about 600 to 374 C and below. However, a high yield of CO2 obtained between 300 to 350 C during hydrothermal liquefaction of lignocellulosic biomass in water as reported by Goudriaan and Peferoen60 conrmed that in this temperature range the presence of these reactions was still evident. Between 100 to 200 C, little conversion of relatively reactive compounds such as alky-aryl ethers and arylether in water even in the presence of H2 was observed.36 A low conversion of the most reactive oxygenates such as alcohols under mild conditions may be anticipated as well. Even the most probable reaction such as dehydration would be inhibited considerably because of the excess of water in the system favoring the shift of the equilibrium in reaction {10} to the left where R represents corresponding olen. Of course, the shift of this equilibrium to the right would be maintained by rapidly removing R from the system via HYD, that is, in the presence of catalyst with a high HYD activity (reaction {11}). In the study of Liu et al.26 residual feeds were converted in the aqueous phase (sub- and supercritical) in an autoclave under typical cracking conditions, that is, absence of H2. It was proposed that thermal conversion of asphaltenes proceeded via radical mechanism rather than via an ionic hydrolysis mechanism. In the presence of water, the formation of unwanted coke was signicantly suppressed. It is believed that water molecules behaved as free radicals scavenger and hydrogen donor involving reactions {5} to {8}, thus slowing down reaction {9}. In the case of a heavy feed, the radical (R) may involve complex structures comprising aromatic and aliphatic entities as well as heteroatoms. For example, according to the model of asphaltenes shown in Figure 3,50,61 the CS C entity (site 1) in model A represents the weakest bonds in the molecule which after rupture yield two large free radicals. Such radicals must be stabilized (e.g., via reaction {5} to {8}) before being converted to coke (reaction {9}). An optimal temperature and pressure, ensuring the highest level of upgrading in SCW (determined by high yield of liquid products and a low yield of coke) may be identied. Such temperature depends on the origin of feed. For example, for a vacuum residue, Cheng et al.62 reported maximum of the yield of liquids and the lowest coke at 420 C while for coal derived asphaltenes Han et al.52 observed an optimal temperature of 460 C. In the latter study, the yield of maltenes in SCW was signicantly greater compared with the experiment conducted in N2. The H/C ratio of the former was higher as well. In the study conducted by Zhao et al.,63 a VR was upgraded in SCW
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research as indicated by decrease in the content of asphaltenes, resins, and aromatics and a signicant increase in the content of saturates. The experiments were carried out in an autoclave from 380 to 460 C at 25.0 MPa. In addition, a signicant reduction in viscosity of products, as well as the content of sulfur, nitrogen, and metals was observed. Bitumen upgrading in SCW alone was compared with that in the presence of 10% HCOOH (semibatch reactor from 633 to 693 K) by Sato et al.64,65 The decomposition of HCOOH yielded the mixture of SCW + H2 + CO2. In the case of the mixture, the involvement of H2 was conrmed by the higher H/ C ratio of the unconverted asphaltenes compared with the asphaltenes in bitumen and those obtained in SCW alone. The coke formation in the SCW + H2 + CO2 mixture was suppressed as well. These observations suggest that H2 successfully competed with water (reactions {5} to {8}) in suppressing coke formation (reaction {9}). In this case, reaction {12} was involved. Hydrogen donor mechanism involving saturated hydrocarbons, particularly naphthenic structures, may also be involved. Thus, it has been generally observed that naphthenic structures can readily donate hydrogen.14 In the presence of hydrogen, this may be depicted by the hydrogen transfer cycle shown in Figure 6. The occurrence of such reactions may be

Review

Figure 6. Free radicals scavenging cycle using naphthenic hydrogen.

anticipated during the upgrading of distillation residues derived from naphthenic crude. The recent study published by Zachariah et al.66 provides a direct experimental evidence for the occurrence of such reactions. There are contradictory reports on the role of water during the bitumen upgrading in SCW. For example, Morimoto et al.67 observed little dierence between the yield and composition of gaseous products obtained during the treatment of bitumen (autoclave, at 420450 C and 2030 MPa for up to 120 min) in SCW and nitrogen. This conrmed that only very small amount of water was involved in upgrading. Moreover, the residue produced in SCW had lower molecular weight distribution, lower H/C ratio, and higher aromaticity. In other studies, benecial role of SCW was observed by an increased yields of liquid products and decreased yield of coke.52,53,57,6063,68,69 It should be noted that all these results were obtained in autoclave. This suggests that the distribution of products was changing with time.70 For example, little participation of water may be anticipated during the early stages because of the naphthenic structures present eectively trapped free radicals formed (Figure 6). However, once this source of hydrogen donors was exhausted, the involvement of water as radical scavenger appears to be plausible. Therefore, experimental system and conditions used for bitumen upgrading in SCW are another factor to be considered in designing the overall mechanism under noncatalytic conditions. For heavy petroleum feeds, the involvement of noncatalytic reactions during overall conversion was reported by Mara et al.71 under typical HPR conditions. A higher probability for such and additional reactions in an aqueous environment may
17701

be anticipated. For example, an interaction of H2O with metals (V and Ni) leading to disintegration of the porphyrin skeleton may be the initial step of HDM during an aqueous phase HPR, as it was indicated by Kokubo et al.69 The eect of sub- and supercritical conditions on conversion of coal tar was also investigated.72,73 It was observed that at the same temperature, liquefaction of tar (autoclave, 623 and 673 K, 2540 MPa) increased with increasing water density at the same reaction temperature. It was proposed that under sub- and supercritical conditions, hydrolysis was involved in the conversion of macromolecular structure of tar to lighter products such as phenol, biphenyl, diphenylether and diphenylmethane. 5.2. Ionic Reactions. The involvement of an ionic mechanism as part of the overall conversion of dierent feeds in aqueous phase may be anticipated, although this issue has been receiving little attention. Yet, it was indicated earlier that, compared with liquid water, the dissociation constant of H2O increased with increasing temperature from 100 C toward subcritical and supercritical conditions.1719 Therefore, in an aqueous medium, the involvement of H+ and HO ions in parallel with the free radicals as well as conventional HPR reactions, as part of the overall conversion in the presence of catalysts and H2, is highly probable. By adding to a heteroatom (e.g., S, O and N), H+ may enhance the rate of hydrogenolysis of the corresponding heterobonds. Autocatalysis by fatty acid products formed during the hydrolysis of triglycerides (soybean oil) in subcritical water (250 300 C) supports the involvement of an ionic mechanism as well.74 The H+ ions required for such mechanism were generated by the partial dissociation of the fatty acids produced by triglycerides hydrolysis. The nitrogen content of biomass of an algae origin may exceed 10 wt %. A strong tendency of N-heterorings to combine with H+ ions has been well documented.9,12 This suggests that the ionic mechanism plays key role during the conversion of algae biomass to hydrocarbons and ammonia under sub- and supercritical water conditions without a catalyst being present. The HPR of biocrude obtained from an algae biomass via hydrothermal route may be aected unless most of nitrogen ends up in aqueous phase rather than in biocrude. This issue was the focus of attention of the study published by Valdez et al.75 Additional eorts may be needed to clarify mechanistic aspects of the conversion of N-heterorings under hydrothermal conditions. During the noncatalytic hydrothermal conversion of lignin in subcritical (300370 C) and supercritical (390450 C) regions, Yong and Y. Matsumura7678 obtained results which support both radical and ionic mechanisms occurring in parallel. A rapid change in pKw on approaching supercritical region from subcritical region enhanced the involvement of ionic reactions. However, increased temperature required for this change favored radicals formation as supported by the increased yield of coke in supercritical region. The kinetic network for the noncatalytic conversion of guaiacol in sub- and supercritical regions proposed by Yong and Matsumura78 considered both ionic and radical reactions. Thus, the rate constants for the overall conversion of guaiacol obeyed Arrhenius law in subcritical region, but they deviated in supercritical region unless both radical and ionic reactions were considered. Under typical HPR conditions (e.g., with H2 and catalyst being present), the involvement of H+ ions during the
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research elimination of H2O and NH3 from alcohols and amins, as the nal intermediates of HDO and HDN reactions, respectively, has been well documented.9,12,13 This may be illustrated using reactions {13} and {14} in Figure 7. In these reactions,

Review

Figure 7. Dehydration {13} and ammonia elimination {14} with aid of H+ ions.

carbenium ion intermediate may be converted to either an olen (e.g., RCHCH2) or isomerized to dierent iso-olens before it is hydrogenated to the nal saturated hydrocarbons. Rates of the H2O and NH3 elimination reactions may be further enhanced over acidic catalysts, which can be the additional source of protons.2 Montgomery et al.79 reported that a disintegration of asphaltenes in water to light fractions began already at 250 C and 4 MPa total pressure. This was signicantly enhanced by increasing temperature and pressure to 350 C and 11 MPa, respectively. These observations are consistent with the participation of H+ and HO ions in the overall conversion. Thus, thermal eects could not account for all reactions observed. The results published by Li and Egiebor80 may be interpreted in terms of ionic reactions being present during the extraction of oxygen from coal in SCW (360 to 400 C) and their absence during the extraction in supercritical toluene. Thus, a signicantly greater amount of oxygen removed under the former conditions can be attributed to the enhanced hydrogenolysis aided by H+ ions. The involvement of H+ ions during the dehydration of alcohols is indicated by reaction {12}. The observations made during the HDO of phenol by Massoth et al.81 could be attributed to the participation of H+ during the overall HDO as well. Thus, even the hydrogenolysis of phenol to benzene could be aided by H+ ions. The promotional eect of acetic acid on the aqueous phase HDO of p-cresol over Ru/C (300 C; 4.8 MPa) observed by Wan et al.82 could only be attributed to the participation of H+ ions. Methyl-cyclohexanol was the main product in the absence of acetic acid. The appearance of methyl-cyclohexane as the main product in the presence of acetic acid can be attributed to the increased rate of dehydration of methyl-cyclohexanol (reaction {12}) caused by the presence of H+ ions. It should

be noted that the content of acetic acid in some bio-oils from pyrolysis of biomass may exceed 10 wt %.12,13 This suggests that ionic reactions play an important role during the HPR of these feeds in aqueous phase. The HO ions formed via partial dissociation of H2O may also participate in some reactions. The involvement of HO ions in the overall conversion of anisole to phenol and methanol was proposed by Wu et al.83 For heavy feeds, HO ions may interact with Me-C bonds present in porphyrins, where Me = Ni and/or V metals.50 In the absence of any experimental data on the ionic mechanism, only a speculative argument on the involvement oh HO ions may be forwarded.57 In the presence of solid catalysts, HO ions may behave as Lewis bases suggesting that they may adsorb on Lewis acids of solid supports. For the (-Al2O3 support, this may be the beginning of a gradual transformation toward boehmite. Some interaction of HO ions with carbon supports may be anticipated as well. It is believed that such an interaction will increase with the increasing irregularities of carbon surface. 5.3. Catalytic Reactions. The above discussions indicated a distinct dierence between the mechanisms in an aqueous phase involving high oxygen content feeds (biofeeds, FTS aqueous phase, etc.) and that involving distillation residues derived from petroleum, although some common reactions may be evident. This is illustrated on several examples from literature which are based on experimental observations. In this regard, a signicant dierence in the structure of these feeds and their miscibility/solubility in water may play an important role. Of course, in the presence of catalyst, the role of external H2 during the overall HPR mechanism can be dominant. In this case, hydrogen activation, that is, the conversion of dihydrogen to an active surface hydrogen with the aid of catalyst, must precede HPR reactions. The hydrogen generated in situ must be activated as well. The tentative scheme shown in Figure 8 accounts for the potential sources of hydrogen. In an aqueous phase under HPR conditions, sources 1 and 2 represent steam reforming of hydrocarbons and WGS (i.e., reactions {1} and {2} in Figure 1, respectively). Potential of H2O as free radicals scavenger in noncatalytic reactions (reactions {5} to {8} in Figure 5) was indicated earlier. The overwhelming evidence for the presence of reforming and WGS reactions was clearly conrmed in several studies.8493 Apparently, active surface hydrogen formed on catalyst surface according to Figure 8 can be abstracted by free radicals.94 Consequently, the involvement of noncatalytic reactions (e.g., {5} to {8} and {11}) in radicals stabilization is diminished. At the same time, radical stabilization via hydrogen transfer cycle in Figure 6 may be sustained in the presence of active hydrogen. It is believed that, in the presence of HPR catalysts, the reactions shown in Figures 6 and 8 may

Figure 8. Active hydrogen sources during hydroprocessing in aqueous phase.


17702
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research play much more important role in the overall mechanism of HPR of petroleum feeds than that of a high oxygen content feeds. 5.3.1. HPR Mechanism for High Oxygen Feeds. For high oxygen content feeds (e.g., biofeeds and aqueous phase from FTS), the dehydration of alcohols (ROH) to alkenes (R) may be the rate determining step during the overall HDO, although a direct hydrogenolysis (e.g., phenol to benzene) may be involved. The corresponding dehydration equilibrium reaction (reaction {10} in Figure 5) is aected by the excess of water in the system which tends to shift the equilibrium from alkenes to alcohol. An ecient removal of oxygen from the system can still be accomplished providing that a rapid removal of alkenes (R) from the system can be ensured. This may be achieved by a rapid HYD of alkenes to alkanes (reaction {11}. Therefore, catalysts with a high HYD activity and at the same time stable in aqueous environment may be suitable for these applications. The HDO mechanism applicable to conventional HPR conditions was discussed extensively elsewhere.2,12,13 The rate of dehydration of alcohols, as the nal stage of the HDO of many oxygenates, may be inuenced by reaction conditions. For example, over Pd/C catalyst at 200 C, phenol was converted predominantly to cyclohexanol while hydrogenolysis to benzene was not observed.34,35 However, after acidifying the solution, cyclohexanol was quantitatively dehydrated to cyclohexene followed by HYD to cyclohexane.95 This conrmed that hydronium ions aided dehydration of cyclohexanol (e.g., reaction {13} in Figure 7). At the same time, noble metals facilitated the HYD function. This concept of bifunctional catalysis was conrmed using more complex reactants (e.g., guaiacols and syringols). Zhang et al.96 reported results on the HDO of phenol over Ni catalysts supported on HZSM-5 (Si/Al = 38 and 50) and on (-Al2O3 (autoclave; 160240 C; 2.0 g phenol; 1.5 g catalyst; 40 mL water; 4.0 MPa of H2). For all three catalysts, the hydrogenolysis to benzene and the HYD to cyclohexanol were the main routes. Dehydration of the latter to cyclohexene was enhanced over the catalysts supported on HZSM-5 compared with that on Al2O3. Moreover, over the former catalysts, small amounts of methyl-cyclopentane were observed. This clearly conrmed the involvement of H+ in dehydration of alcohols to alkenes (reaction {13}).2 Isomerization of the carbocation is another potential reaction to occur as indicated by the presence of methyl-cyclopentane among the products over Ni/HZSM-5 catalysts. For Ni/Al2O3 catalyst, an insucient strength of Bronsted sites may be the reason for the absence of isomerized product. The study published by Peng et al.29 contributes to the fundamental understanding of the HPR of oxygenates in an aqueous environment under mild conditions. The reactants such as 1-propanol, 2-propanol, 1,2-propanediol, 1,3-propanediol, and glycerol (10 wt % in water) were studied in a batch reactor at 473 K, 4 MPa of H2, and 0.3 g of Pt(3 wt %)/Al2O3. Under these conditions, the direct cleavage of CC and CO bonds was not observed. For 2-propanol and 1,2-propanol, the deHYD to ketone was the main reaction while for 1,3-propanol and glycerol the CO bond was cleaved by dehydration. For alcohols with the terminal hydroxyl group, the CC bond cleavage occurred in steps via deHYD to aldehyde followed by either disproportionation and subsequent decarboxylation or decarbonylation. Reactivity of the alcohols increased with increasing number of hydroxyl groups. Thus, over Pt/Al2O3, the following overall reactivity order was established: glycerol
17703

Review

1,3-propanol > 1,2-propanol > 1-propanol 2-propanol. However, a signicant eect of catalyst type on the HDO mechanism was indicated in the study conducted by Chen et al.30 Diphenyl ether, 2-phenylethyl phenyl ether and benzylphenyl ether, as representatives of lignin, were studied over Ni/SiO2 catalyst at 120 C and 0.6 MPa of H2 in an autoclave in an excess of water.36 For diphenyl ether, the initial reactions (Figure 9) were dominated by HYD (65% cyclohexyl phenyl

Figure 9. Mechanism of conversion of diphenyl ether in water (Ni/ SiO2; 120 C; 0.6 MPa).

ether), followed by hydrolysis (25% cyclohexanol) and hydrogenolysis (10% benzene). With progress of reaction, the yield of cyclohexyl phenyl ether reached a maximum and then decline to zero while that of cyclohexanol and benzene increased. The importance of hydrolysis was conrmed by much higher yield of cyclohexanol compared with that of benzene. It is postulated that hydrolysis was result of the electron deciency on carbon of the CO bond which was oset by the interaction with unpaired electrons on oxygen of H2O molecule. In this case, catalyst surface could facilitate a at adsorption of diphenyl ether which improved the access of H2O molecules to the CO bond. The HYD of phenyl ring, as the initial step was only minor reaction. Initially, hydrogenolysis was a dominant route for the other two reactants. Under the same conditions, no reactions took place over SiO2, thus conrming the role of Ni in the overall conversion. The study of Duan and Savage 86 showed how the observations made during the HPR of a real feed in aqueous phase can be explained in terms of mechanistic aspects discussed above. In this case, a microalgae paste and a series of catalysts were mixed with deionized water in batch reactor and tested under subcritical conditions (350 C). The reactor was pressurized with either He or 3.5 MPa of H2. In the absence of catalyst, the external H2 had benecial eect on the yield of bio-oil. However, except for Pt/C catalyst, the yield of bio-oil in H2 over other catalysts (e.g., Pd/C, Ru/C, CoMo/ Al2O3, Ni/SiO2Al2O3, and zeolite) was lower than that in the inert environment. The analysis of gaseous products conrmed CO2 to be dominant product. This indicated the involvement of WGS reaction. Thus, it was shown that decarbonylation generating CO was an important route during decomposition of the algae biomass.12 Therefore, in the presence of a catalyst,
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research the reaction pathways shown in Figure 8 may be part of the overall HPR mechanism even under subcritical conditions. 5.3.2. HPR Mechanism for Petroleum Residues. In the case of residual feeds, the temperature of at least 400 C is required to achieve desirable conversion. This results from the presence of high molecular weight components (e.g., resins, asphaltenes, and porphyrins) in the feed. Such temperatures dictate that for most part, the HPR in aqueous phase must be carried out under supercritical conditions. This suggests that thermal cracking of organic bonds, leading to the formation of free radicals is an important part of the overall mechanism. Several weak bonds as potential cracking sites can be identied in the models of asphaltenes shown in Figure 3. The reactions occurring under the conditions of conventional HPR are also occurring during the HPR in an aqueous phase. The extent of the HPR in the latter phase depends on the origin of residues and type of catalyst. Generally, catalysts comprising noble metals (Pt, Pd, Ru, Rh, etc.) exhibit a high HYD activity. This facilitates a high rate of the HYD of heterorings accompanied by the change of the CAR-S(N,O) bonds to corresponding CAL-S(N,O) bonds. Consequently, strength of the latter bonds is signicantly decreased. The stabilization of free radicals involving the cycle in Figure 6 is expected to be much more important for residual feeds than that for a high oxygen content feeds because of a much higher content of naphthenic structures in the former feeds.1416 The involvement of ionic reactions, particularly over an acidic catalyst should be anticipated as well. Model compounds used to study the mechanism of HPR reactions under conditions of conventional HPR have been also evaluated under aqueous conditions. This may simulate secondary upgrading of reactants produced initially during the conversion of heavy components (e.g., asphaltenes). For example, the conversion of DBT over CoMo/Al2O3 catalyst in SCW was observed in the presence of CO suggesting that the required hydrogen was produced via WGS reaction. Similar observation was made by Arai et al.53,87,88 over sulded NiMo/ Al2O3 catalyst, while little conversion was observed over the corresponding oxidic catalysts. In fact, in CO+SCW system, the reaction rate was higher than in H2+SCW suggesting that the hydrogen which originated from the source 1 and 2 (Figure 8) was more reactive than from the source 3. In the study conducted by Ng and Milad89 on the HDS of BT, the hydrogen produced via WGS reaction was about seven times more reactive than the external H2. The transfer of hydrogen from SCW to reactants was also conrmed using D2O.90 No WGS reaction was observed without catalyst. Even in the CO2 + H2 + SCW system, the rate of reaction was higher than in H2 + SCW. When the feed was partially oxidized in situ in SCW to generate CO, the conversion of several reactants (e.g., DBT, carbazole, quinolin and naphthalene) involving the HYD route was higher than in H2 + SCW.8389,9193 The involvement of H+ ions during the HPR of residues in SCW, as part of ionic reactions (e.g., reactions {13} and {14}) was anticipated above. Under the conditions of conventional HPR, the nal stages of HDO12,13 and HDN,911 that is, elimination of oxygen and nitrogen from the last intermediates, respectively were interpreted in terms of a proton transfer from catalyst to the intermediate. Such reactions are more likely to take place in SCW than in a liquid water. Heavy Arabian heavy crude containing 3 wt % sulfur was treated in an autoclave (without external H2) in SCW.97 Under these conditions, only about 7% of sulfur in the feed were removed in the absence of catalyst. It is believed that most of
17704

Review

this sulfur was removed thermally. In the presence of MoS2 the removal of sulfur approached 12%. Again, thermal reactions accounted for most of the sulfur removed. Rather low sulfur removal can be attributed to the absence of reforming and WGS reactions, which are the source of active hydrogen, as it was demonstrated by Ng and Milad89 who carried out similar experiments in the presence of CO. In the absence of any experimental data, only a speculative mechanism of hydrodemetallization (HDM) in SCW may be proposed. A high reactivity of H2O molecules in SCW would favor a direct interaction with the porphyrinic form of V and Ni metals in the residue. Consequently, the rate of disintegration of porphyrin skeleton as the nal stage of HDM, would be enhanced. The presence of V in a vanadyl form (O = V) suggests a more direct interaction with Ni than that with the V = O group. The HO ions can interact with the metals as well. However, this would have to be conrmed by experimental data on the HDM of model porphyrin compounds obtained in SCW over a catalyst. Based on the results obtained during the pyrolysis of bitumen in SCW (at 723 K) in the presence of cubic (8 nm) and octahedral (50 nm) CeO2 nanoparticles, Dejhosseini et al.98 observed a higher activity of the former than that of hexagonal CeO2. This was attributed to a much higher oxygen storage capacity of the cubic CeO2. It is speculated that the HO ions generated from H2O were involved in red-ox reactions producing surface hydrogen and carbon oxide. 5.3.3. HPR Mechanism for Other Feeds. The HPR in an aqueous phase involving other feeds (e.g., coal derived liquids, oil shale liquids, etc.) has been receiving little attention. However, the database of experimental results established for biofeeds and petroleum feeds may serve as a basis for proposing the mechanism of the HPR of other feeds. Thus, after a detailed characterization of any feed and corresponding products, a set of reactions, particularly those involving H2O may be proposed. Obviously, experimental results are always needed to strengthen arguments one way or the other.

6. CATALYST DEVELOPMENT FOR HYDROPROCESSING IN AQUEOUS PHASE Attempts have been made to develop catalysts with desirable activity and selectivity as well as a high stability in the presence of large quantities of water. In this regard, mild, subcritical and supercritical conditions have been receiving attention. Apparently, the conventional HPR catalysts consisting of the Co(Ni)-Mo(W)-S active phases supported on -Al2O3 may not be suitable for a direct use unless they are modied to enhance stability.69,1116,94 As it was indicated above, both the supports and active metals have to be carefully selected for catalyst design for the applications under aqueous phase conditions. In this regard, besides conventional materials, combinations of various nonconventional metals and supports have been explored. However, a limited number of articles published in scientic literature indicates that the development of HPR catalysts for applications in aqueous phase is still in an early stage and may be rather challenging. The adverse eects of H2O on performance of the conventional catalysts were discussed in detail elsewhere.12,13 It is believed that in the presence of large amounts of H2O as in the case of sub- and supercritical conditions employed during the HPR in an aqueous phase, a detrimental eect of H2O would be much more evident than that under the conditions of conventional HPR. The requirement of a high HYD activity to
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research Table 4. Summary of Catalysts Tested in Aqueous Phase under Mild, Subcritical, and Supercritical Conditions
catalyst feed conditions batch.; 200300 C; 4 MPa cont.; 225 and 265 C; co-fed H2 batch; 200 C; 4 MPa; H2 batch; 150 and 300 C; 4.8 MPa of H2 batch; 150300 C; 6.9 MPa total cont.; 150230 C; 6.4 MPa total batch; 130279 C; 611 MPa total batch.; 200 and 250 C; 4 MPa H2 batch; 180 and 200 C 280 C; 4 MPa; H2 batch; 200300 C; 412 MPa batch; 150190 C batch; 245 C; 6 MPa; H2 cont.; 160240 C; ow of H2 batch; 160 C; 0.6 MPa; H2 batch; 16240 C; 4 MPa; H2 cont.; 200 C; 4 MPa; H2 batch; 150 C; 5 MPa batch; 350 C; 3.5 MPa H2 batch; 350 C batch; 330 C batch reactor; 350 C; 0.55 h; without H2 batch; 330370 C; 4 h; either CO or H2 batch; 380 C; 629 MPa batch; 380420 C; 06.9 MPa H2 batch; 380 C; 28 MPa; no H2 batch; 400 C; 3.4 MPa; H2 batch; 400 C; 30 MPa total batch; 380420 C; 6.9 MPa; H2 batch, 400 C; 2.54.0 MPa; H2 batch; 723 K

Review

ref. 35 39 119 82, 110 111 30 31 32 33, 34 112 114 115 120 116 36 101 113 118 86 24 122 123 124

Mild Conditions Pd/C, RANEY Ni, Ni/SiO2, Ni/ASA, Naon suspensions, Naon/SiO2, propyl-phenol zeolites Pt/SiO2Al2O3, Pd/SiO2Al2O3 sorbitol Ni-MgO sorbitol Ru/C, Ru/Al2O3, Pt/Al2O3, Pt/C, Pd/Al2O3, Pd/C, sulded CoMo/Al2O3, acetic acid, p-cresol NiMo/Al2O3, and NiW/Al2O3 Ru/C and Pd/C furfural, guaiacol, acetic acid Ru/ZrO2 and RuMo/ZrO2 Pt (35 nm) protected by polyethyleneimine Pt on HY, H$, HZSM-5, (-Al2O3 and SiO2 Pd/C Pt on AC, MWCN and CB; Pt on ZrO2, TiO2, and CeO2 Ru(5%)/H- Pt(1%)/H- zeolite NiW/SiO2 Pt/ZrO2 with H4SiW12O40, H3PW12O40, H3PMo12O40 Ni/SiO2 Ni/HZM-5, Ni/Al2O3 Cu/ZrO2 varying Cu/Zr ratio Ru + transition metal (e.g., Zn, Cr, Mn, Co, Fe, and Ni) Pd/C, Pt/C, Ru/C, Ni/SiO2Al2O3, zeolite, sulf. CoMo/Al2O3 sulf. NiMo/Al2O3 Pt(5%)/C Pt, Pd and Ni all supported on carbon sulf. Pt/C propanoic acid glucose and fructose phenol phenol 4-propylphenol lignin cellulose cellulose glycerol aryl-ethers phenol glycerol benzene Subcritical Conditions microalga pyrolysis oil fatty acids (stearic, palmitic, lauric, oleic, and linoleic) jathropa biofeed hydrothermal liquef. biofeed Supercritical Conditions benzofuran pyridine palmitic acid microalgae, algae bio-oil DBT algae biomass DBT, naphthalene, tetraline, carbazole bitumen

Pt(5%)/C Pt/C, Pd/C, Ru/C, Rh/C, Pt/Al2O3, Mo2C, MoS2, PtO2, Al2O3, sulf. CoMo/Al2O3 Pt/C and Pd/C Pd(5%)/C, Pt(5%)/C NiMo/Al2O3, CoMo/Al2O3

125 126 127, 128 127, 128 53, 87, 88 128 92, 93 93

Pt/C, Pd/C, Ru/C, Rh/C, Pt/Al2O3, Mo2C, MoS2, PtO2, Al2O3, CoMo/ Al2O3, carbon NiMo/Al2O3 cubic and octahedral CeO2 nanoparticles

minimize catalyst deactivation during the HPR in aqueous phase was indicated above. It appears that for these applications, the catalysts comprising noble metals (e.g., Pt, Pd, Ru, and Rh) supported on various supports may be more suitable than conventional HPR catalysts. The summary of studies on catalyst development is given in Table 4. 6.1. Methods for Catalysts Testing. There is little dierence in the methodology for catalysts preparation for the applications under aqueous conditions and those used for the conventional HPR. An extensive information on preparation of the latter catalysts can be found in the scientic literature.2,612 The validity of experimental data on catalyst performance is inuenced by the testing protocol employed.99,100 The catalyst evaluations in both batch and continuous systems have been generally observed. Dierent sizes of batch reactors, starting
17705

with micro autoclave up to several liters volume size systems, have been available for testing. They are useful screening tools, however catalyst activity may be aected by some products (e.g., NH3, H2S, etc.), which accumulate in the system rather than being carried out with reaction streams as it is in the case of continuous systems. Moreover, some inherent limitations of batch systems (i.e., an unsteady-state operation, continuously varying catalyst/oil ratio, a lengthy heat-up periods between preheat and reaction, etc.) should not be overlooked. On the other hand, in batch systems, the mixture/slurry (water + catalyst + feed) for testing in an aqueous phase can be readily prepared and tested. High-throughput techniques have been used for rapidly prescreening a large number of catalysts in batch reactors. However, to obtain a more comprehensive information, more
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research

Review

Figure 10. Continuous reactor system for hydrothermal conversion of biomass. Reprinted with permission from ref 102. Copyright 2010, American Chemical Society.

detailed studies may be required on selected catalysts. Lee et al.101 used this method for determining the initial activity of the supported monometallic catalysts such as Pd, Pt, Ru, Rh, Ni, and Co for the HYD of various carbonyl group containing reactants in an aqueous phase. They used reactor which consisted of 24 wells machined into a cylindrical stainless steel high pressure chamber. Using this system, the database was generated much faster than in a typical batch reactor. Patwardhan et al.59 described the operation of a continuous stirring tank reactor (CSTR) rated for more than 30 MPa and up to 650 C during the conversion of S-containing reactants in SCW. In this case, water and model reactants were fed separately to the top, without premixing. It might be the only study, in which CSTR system was used to study conversion in SCW. Figure 10 shows the continuous system,102 which may be adapted for catalyst testing in the presence of water. The list of major components and modes of operation are evident from the schematics. The continuous system is one of the few found in literature, being used for catalyst testing under SCW conditions. Of a particular importance is the separate unit required for the generation of SCW and subsequent mixing with reaction streams. Absence of the source of external H2 in the schematic of continuous system should be noted. However, because of the catalytic upgrading of an algae feed being studied,102 hydrogen required for HPR reactions was generated in situ via WGS reaction involving CO produced during decarbonylation of the feed. Obviously, for the HPR of liquid feeds (e.g., biofeeds, FTS reaction water, etc.) in an aqueous phase, the continuous system in Figure 10102 requires modications. First of all, the preheaterreactor unit would have to be replaced by a catalytic reactor. It is believed that such a change could be made without any diculties. For example, in the study conducted by Zohrer et al.,103 biofeed (glycerol) was injected directly into SCW entering catalytic reactor. Little problems during the feeding into continuous reactors are anticipated for high water content liquid feeds such as biofeeds and reaction water from FTS
17706

contrary to that for heavy petroleum feeds. For the latter feeds, a special feeding system would have to be designed. Nevertheless, it is believed that a special continuous system needs to be designed to bring catalyst testing for the applications in an aqueous phase to the next level. 6.2. Selection of Supports. The choice of support for preparation of the catalysts to be used in aqueous media is crucial because of a potential reaction with water. Thus, in the case of conventional catalysts, the most frequently used -Al2O3 tends to undergo the H 2 O aided transformation to boehmite,29,104106 as it is shown by reaction {15} in Figure 11.

Figure 11. Potential reactions of H2O with -Al2O3 and carbon.

Ravenelle et al.107 observed that the conversion of -Al2O3 to boehmite was complete within 10 h at 200 C. However, for the Ni/-Al2O3 and Pt/-Al2O3, the transformation was signicantly slowed down. On the other hand, -Al2O3 exhibited a high stability at 350 C during dehydration of the 1:1 mixture of heavy alcohols in water during more than 30 days.3 Under similar conditions, SiO2Al2O3 exhibited a high stability as part of the Ni/SiO2Al2O3 catalyst used for the partial HYD of the aqueous phase obtained during FTS.3 It is believed that other metal oxides (e.g., zeolites, TiO2, ZrO2, mixed oxides, etc.) may be more suitable supports than traditionally used -Al2O3. It should be emphasized that for the catalysts to be used for the aqueous phase HPR of biofeeds and reaction water from FTS, a proper acidity of supports may be required to maintain a high rate of alcohol dehydration as the last step of HDO (reaction {13} in Figure 7). In this case, some zeolites may exhibit a desirable acidity. Therefore, the interests in zeolites as
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research the supports for catalysts have been steadily growing. Ravenelle et al.108 investigated the hydrothermal stability of zeolites Y and ZSM-5 with the varying Si/Al ratios in liquid water at 150 and 200 C. Under these conditions, ZSM-5 zeolite was not modied. However, the zeolite Y with the Si/Al ratio of 14 or higher was transformed into an amorphous solid. The zeolite degradation was caused by hydrolysis of SiOSi bonds rather than by dealumination. This resulted in the loss of micropore volume and that of accessible acidic sites. The stability of several ZrO2 and TiO2 samples in SCW was evaluated by Zohrer et al.103 with the aim to select the most stable support for Ru catalysts. The stability of supports and corresponding Ru (2 wt %) catalysts were tested in an autoclave at 400 C, 28.5 MPa, for 20 h. The Ru/ZrO2 catalyst supported on tetragonal ZrO2 exhibited the highest activity and stability during the conversion of glycerol in a continuous system operating under SCW conditions. Carbons are neutral and hydrophobic supports; therefore, in an aqueous medium, they exhibit high stability. Various forms of carbons (i.e., activated carbons, carbon blacks, carbon composites, carbon nanotubes, etc.) may be suitable support.20 For reactive carbon solids, the potential reaction of carbon with H2O (reaction {16}) becomes evident at above 700 C. A low reactivity for this reaction is ensured by a high severity used for carbon supports preparation.20,109 For example, activated carbon is prepared by steaming and/or partial oxidation (diluted air) of various carbonaceous solids at about 850 C. Oxygen containing entities (carbon centered peroxides and peroxy radicals, etheric groups, hydroxyl groups, etc.) on the surface of activated carbons left behind may play an important role during the impregnation with active metals. Therefore, a desirable stability of the carbon supported catalysts in an aqueous environment is anticipated, although some experimental data obtained under subcritical and particularly under supercritical conditions are still needed. 6.3. Catalyst Testing. As Table 4 shows, the catalysts testing was conducted under mild conditions (less than 300 C) as well as in subcritical (300370 C) water and SCW (above 374 C) using both the batch and continuous systems. Most of the testing has been carried out under mild conditions. Studies were dominated by model feeds while to a lesser extent real feeds were used as well. 6.3.1. Mild Conditions. The catalyst development for HPR in aqueous media under mild conditions (less than 300 C) has been focusing on two groups of metals, that is, noble metals (e.g., Pt, Pd, Ru, and Rh) and other metals (Ni, Co, Cu, etc). A wide range of supports were tested; however, carbons were the supports of choice. -Al2O3 has also been used, although some stability problems during a long-term performance of catalysts may be anticipated, as it was indicated above. 6.3.1.1. Noble Metals Containing Catalysts. As the most abundant product in biomass pyrolysis liquids, acetic acid alone and in the mixture with p-cresol was investigated at 150 and 300 C over a series of catalysts (i.e., Ru/C, Ru/Al2O3, Pt/ Al2O3, Pt/C, Pd/Al2O3 and Pd/C).82 The experiments were conducted in a batch reactor at 4.8 MPa of H2. For experiments, 0.05 mol of substrate and 0.2 g of catalyst were added to 40 mL of water. With respect to the conversion of acetic acid, the following activity order was established at 300 C: Ru/C > Ru/Al2O3 > Pt/C > Pt/Al2O3 > Pd/Al2O3 > Pd/C. However, ethane as the minor product was formed only over Pt/Al2O3 and trace amounts of ethane over Ru/C and Pt/C catalysts. The Ru/C was the only catalyst used for the HDO of
17707

Review

p-cresol and the mixture of p-cresol with acetic acid. The inhibiting eect of p-cresol on conversion of acetic acid and a benecial eect of acetic acid on the HDO of p-cresol was discussed earlier.82 Thus, the benecial eect resulted from the contribution of ionic mechanism to the overall conversion. The conventional sulded CoMo/Al2O3, NiMo/Al2O3, and NiW/Al2O3 catalysts were compared with the Pt, Pd, and Ru catalysts supported either on carbon or on -Al2O3 by Wan et al.110 under identical conditions as above82 at 300 C. Under these conditions, conventional catalysts were inactive. The highest activity for hydrocarbons formation was exhibited by the Pt catalysts. Elliott and Hart111 used the mixture containing 200 g water and 5 wt % of substrate (furfural, guaiacol, and acetic acid each) in a batch reactor to study reactivity of the substrates between 150 to 300 C and 6.9 total pressure of H2 over Ru/C and Pd/ C. With respect to HYD, the former catalyst was more active than Pd/C catalyst; however, above 250 C, reforming and methanation reactions (i.e., reactions {1}, {2} and{4} in Figure 1) became quite evident over Ru/C catalyst. A lower activity of Pd/C could be increased by increasing temperature. At the same time, the production of methane and CO2 was kept at a minimum. The aqueous phase HDO of propanoic acid was used by Chen et al.30 to compare Ru/ZrO2 and RuMo/ZrO2 catalysts in a trickle-bed reactor after the catalysts (6 g of 2040 mesh) were activated in situ in the ow of H2 at 300 EC for 3 h. In the temperature range 150230 C, the total pressure of 6.4 MPa ensured an aqueous-phase system of 0.83 mol/L of the reactant. For the RuMo/ZrO2 catalyst, the eect of Mo/Ru ratio (01.5) on the activity and selectivity was investigated. The catalyst with the Mo/Ru ratio of 0.2 exhibited the highest activity for the conversion of propanoic acid. The activity decreased with further increase in the Mo/Ru ratio. The products included propanol, methane, ethane, propane, and trace amounts of ethanol and acetic acid. Over Ru/ZrO2 catalyst, CC bond cleavage to methane and ethane was dominant reaction compared with the HYD of CO bond, while the RuMo/ZrO2 catalyst favored the HYD of CO bond in propanoic acid. The Pt nanoparticles (3 to 5 nm) protected by polyethyleneimine (Pt-PEI) were used as catalyst for the conversion of glucose and fructose at 403543 K in subcritical water and H2 in a batch reactor.31 For the experiments, H2 was introduced into reactor at ambient temperature until the pressure reached 5.0 MPa. In the temperature range used, the total pressure increased from 6 to 11 MPa. Typically, 0.48 g of glucose, 1 g of aqueous dispersion Pt-PEI ( 5 mg Pt) and 60 g of ion-exchanged water, were used. Compared with inert atmosphere (Ar), conversion was signicantly enhanced under H2. At 403 K, glucose could be readily isomerized to fructose. In the temperature range 483543 K, glucose produced 1,2propanediol, 1,2-hexanediol, and ethylene glycol, while fructose yielded 1,2-propanediol, 1,2-hexanediol, and glycerol. Other catalysts tested included the Pt protected by polyvinylpyrrolidone, Pt/SiO2, and Pt/Al2O3; however, Pt-PEI exhibited superior activity. Hong et al.32 studied the eect of support on the activity of Pt (1 wt %) catalysts during the conversion of phenol. In this case, zeolites such as HY, H$, and HZSM-5 as well as -Al2O3 and SiO2, were compared (473 and 523 K; H2 pressure of 4 MPa; WHSV of 20 h1; 10 wt % H2O). For all catalysts, the overall phenol conversion reached almost 100%. However,
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research signicant dierence in product distribution, was observed. Thus, cyclohexane accounted for more than 90% of all products over Pt/zeolite catalysts compared with less than 3 wt % over Pt/Al2O3 and Pt/SiO2 catalysts. For the latter catalysts, cyclohexanol accounted for almost 95% of the converted phenol. This may be attributed to a lack of surface acidity which would enhance dehydration of cyclohexanol (e.g., via reaction {10} and {13}). Small amounts of bicyclics and tricyclic products were observed over the Pt/zeolite catalyst. During the HDO of phenol in aqueous medium (below 453 K, Pd/C catalyst, batch reactor), Zhao et al.33,34 observed an increased yield of cyclohexcanol and decreased yield cyclohexanone with time on stream. This suggests that the latter was an intermediate for the formation of cyclohexenol. At 453 K, small amount of cyclohexane was formed if the solution was acidied with H3PO4. However, temperature increase to 473 K resulted in a complete dehydration of cyclohexanol to cyclohexene (reaction {10}) followed by the HYD of latter (reaction {11) to cyclohexane. There was little evidence of the hydrogenolysis of phenol to benzene. Similarly, a high selectivity to cyclohexanol in a neutral aqueous solution was reported over Pt-, Ru-, and Rh-based catalysts.74 However, with temperature increase to 473 K of the acidied solution, cyclohexanol was quantitatively dehydrated to cyclohexene followed by HYD to cyclohexane. Similarly, 4-n-propylguaiacol, 4-allylguaiacol, and 4-acetonylguaiacol were converted to cycloalkanes (80%), methanol (78%) 1218% intermediate cycloalcohols or cycloketones (1218%). In this case, the reaction mixture comprised 5 wt % Pd/C (0.040 g), reactant (0.0106 mol), 0.5 wt % H3PO4 in 80 mL of water. Ohta et al.109 prepared series of Pt catalysts (2 wt % Pt) supported on activated carbon (Norit and Wako), mesoporous carbon, multiwalled carbon nanotube, and carbon black via impregnation with aqueous solution of H2PtCl6. The catalysts were used for the HDO of 4-propylphenol in water at 280 EC under 4 MPa H2. The Pt/ACN exhibited high activity with 97% yield of propylcyclohexane, similarly to the Pt catalyst supported on mesoporous carbon and carbon nanotube. The Pt supported on carbon black was less active. The Pt catalysts supported on ZrO2, TiO2, and CeO2 were moderately active, but besides propylcyclohexane, propylbenzene in 310% yield was also present. These results are consistent with the HDO mechanism observed during conventional HPR. Contrary to these observations, Pt/Al2O3 catalyst was inactive because of the structural transformation of (-Al2O3 into boehmite. The activity of the Rh, Ru, and Pd catalysts supported on activated carbon was much lower than that of Pt/C. The mixture of 0.5 g lignin in 10 mL of water was used to study conversion at 300 C and 5 MPa of H2 in an autoclave in the presence of the Ru(5%)/H- catalyst.113 A dozen of oxygenates were identied in liquid products. However, rather low overall conversion of lignin (less than 20%) should be noted. Patil et al.114 used Ru(5%)/H- catalyst for the conversion of lignin to hydrocarbons in an autoclave from 200 to 300 C and total pressure of 4 to 12 MPa. Under these conditions, less than 20% conversion of lignin was observed. After adding a basic solution (1 M NaOH) to the mixture, the conversion increased to almost 33%. This supports the involvement of HO ions during the overall lignin conversion. The Pt(1%)/H- zeolite catalyst tested by Kato and Sekine115 exhibited high selectivity for C3 and C4 hydrocarbons during the conversion of cellulose in distilled water under mild conditions (423463 K) in batch reactor. Under the same
17708

Review

conditions, high yields of gaseous products were formed over the catalysts supported on H- and H-Y zeolites with threedimensional structure and large pores. Apparently, among the catalysts tested, a catalyst which can maximize the yield of the product of interest can be selected. Pt/ZrO2 catalysts modied with heteropolyacids such as H4SiW12O40, H3PW12O40, and H3PMo12O40 were used for the hydrogenolysis of aqueous glycerol (10% glycerol) to 1,3propanediol.116 The unmodied Pt/ZrO2 was used for comparison. The modied Pt/ZrO2 catalysts were more active because of a higher acidity. Thus, the catalyst modied with H4SiW12O40 was the most active because of suitable Brnsted acid sites. At the same time, 1,2-propanediol yield increased with increasing concentration of Lewis acid sites. In this study, experiments were conducted in a continuous system in the ow of H2 between 160 to 240 C. The catalyst consisting of Ru nanoparticles having average diameter of 3 nm dispersed on carbon spheres (500 nm) was used by Yang et al.117 for the HYD of ethyl lactate to 1, 2propanediol in water in an autoclave (423 K; 5 MPa H2; 8 h). The catalyst exhibited a high activity and selectivity even after being recycled six times. The bimetallic catalysts containing Ru and a transition metal (e.g., Zn, Cr, Mn, Co, Fe, and Ni) were used by Sun et al.118 for the HYD of benzene at 150 C and 5 MPa in an autoclave. Under these conditions, transition metals were present in an oxidic form, while Ru in a reduced form. The reaction mixture comprised 49.2 g catalyst, 280 mL H2O, and 140 mL benzene. The catalyst preparation involved adding NaOH solution to the mixture of RuCl3H2O and sulfate of the corresponding transition metal under continuous stirring at 353 K. The black precipitate obtained was dispersed in distilled water and transferred to autoclave to be treated at 423 K under 5 MPa of H2. The isolated powder was washed and vacuum-dried before being used for activity tests and extensive spectroscopic evaluations. The bimetallic catalysts such as RuMn(0.23), RuFe(0.47), and RuZn(0.27) exhibited a high selectivity for cyclohexene. 6.3.1.2. Other Metals Containing Catalysts. Series of the Ni/HZSM-5 catalysts containing variable amount of Ni (e.g., 6, 10, 14, and 17 wt %) and the Ni(10%)/Al2O3 catalyst were compared during the HDO of phenol between 160 to 240 C and 4 MPa of H2 in an autoclave (2 g phenol, 40 mL water).101 The highest activity was exhibited by the Ni/HZSM-5 (Si/Al = 38) containing 10 wt % of Ni. Benzene and cyclohexane were dominant products. Small amounts of methyl-cyclopentane were formed as well. The Ni/SiO2 catalyst exhibited a high activity for the HDO of aryl-ethers in water at 160 C and 0.6 MPa of H2 in an autoclave.36 Zhao et al.35 expanded their study with the aim to develop a solid acids as the source of hydroniom ions, which are stable in an aqueous medium at high temperatures. For example, the Naon polymer used as one of the solid acids was hardly ionized as conrmed by little change in pH with time on stream. To test the concept, a series of liquid and solid acids in the presence of Pd/C and RANEY Ni catalysts were used for the aqueous HDO of 4-propylphenol. The results are summarized in Table 5.35 A combination of aqueous solutions of either H3PO4 or CH3COOH with Pd/C yielded 84 and 74% of propylcyclohexane, respectively, while 98% yield of propylcyclohexane was obtained with both Naon suspension in water and the Naon supported on SiO2 (13 wt % of Naon). A combination of either Naon water suspensions or
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research Table 5. Aqueous-Phase HDO of 4-n-propylphenol over Pd and Ni Based Catalysts and Acids (473 K, 4 MPa H2, and 0.5 h)35
catalyst Pd/C Pd/C RANEY Ni RANEY Ni Pd/C Pd/C Pd/C Pd/C RANEY Ni RANEY Ni RANEY Ni 2400 RANEY Ni 4200 Ni/SiO2 Ni/ASA acid H3PO4 CH3COOH H3PO4 CH3COOH zeolite (H-) zeolite (H-Y) Naon solution Naon/SiO2 Naon/SiO2 Naon solution Naon/SiO2 Naon/SiO2 Naon/SiO2 Naon/SiO2 conv., % 100 100 0 0 100 100 100 100 100 100 51 96 9 37 cycloalkane select. % 84 74

Review

1.5 5.2 98 98 99 98 36 64 43 50

the Naon/SiO2 composite with freshly prepared Ni catalysts (RANEY Ni) resulted in 100% n-propylcyclohexane yield compared with 51 and 96% yields for commercial RANEY Ni2400 and RANEY Ni4200, respectively. Compared with Naon, zeolites were poor source of hydronium ions as indicated by rather low yields of propylcyclohexane. The combination of RANEY Ni with Naon/SiO2 was used to study the HDO of 2-methoxy-4-n-propylphenol. The conversion of glycerol to propanediols has been attracting attention as the source of the latter. In the presence of a catalyst and H2, this may be achieved under mild conditions. This is illustrated using two studies published recently. In one study, an aqueous solution of glycerol (40% glycerol) was used as the feed to produce 1,2-propanediol over a series of Cu/ZrO2 catalysts with dierent Cu contents.114 Experiments were carried out in batch reactor at 200 C and 4.0 MPa H2. The selectivity increased with the increasing Cu/Zr ratio while the overall glycerol conversion exhibited little change. Chen et al.119 used an aqueous solution of sorbitol (20%) to study its hydrogenolysis to glycols and glycerol (473 K; 4 MPa H2; batch) over a series of Ni-MgO catalysts with varying Ni/ Mg ratio. In terms of conversion and selectivity, the best performance was exhibited by the catalyst with the Ni/Mg ratio of 3/7. However, catalyst deactivation was noted under more severe conditions. This was indicated by the more complex mixture of products, although the sorbitol conversion increased with increasing severity. The bimetallic Ni/W/SiO2 catalysts containing 5 wt % Ni and 25 wt % W was used for conversion of cellulose into low molecular weight polyols (sorbitol, mannitol, erythritol, ethylene glycol, and 1,2-propanediol) in an aqueous solution.120 For the experiments, 500 mg cellulose and 50 mg catalysts were coslurried with 30 mL deionized water in an autoclave which was subsequently pressurized with H2 to 6 MPa at ambient temperature. The experiments were conducted at 518 K. The bimetallic catalyst was much more active than either Ni/SiO2 + W or Ni/SiO2 + WO3 mixtures. The reduced form of the bimetallic catalyst was more active than the oxidic form. The nano-metallic carbides of the NiWMoC formulations were prepared by mechanical alloying and used as catalysts for upgrading a residual feed in an autoclave at 200 C, 3 MPa, and 24 h.121 The size of catalyst particle decreased with increasing
17709

length of milling, that is, from 126 to 10 nm from 0 to 240 h milling period. For the experiments, 50 g of residue were mixed with 50 g of seawater and 1 g of catalyst. A signicant viscosity reduction was achieved with catalyst prepared with milling time exceeding 200 h. This coincided with a marked decrease in the content of resins and asphaltenes in the feed. Rather superior activity of the was evident considering the extent of viscosity reduction under rather mild conditions. 6.3.2. Subcritical Conditions. Fu et al.122 observed a high activity of the Pt(5%)/C catalyst for the conversion of fatty acids (stearic, palmitic, lauric, oleic, and linoleic). Oleic and linoleic acids had one and two double bonds, respectively, while the remaining acids were saturated. The experiments were conducted under subcritical water conditions at 330 C in an autoclave. For the unsaturated acids, the conversion involved the HYD of double bond rst, followed by decarboxylation. The latter reaction dominated the overall conversion of the acids to saturated n-alkanes. In line with the mechanism discussed earlier, the HYD of double bonds involved hydrogen generated in situ. Thus, no external H2 was present. Hayashi et al.123 studied the conversion of Jathropa biofeed in subcritical water (batch reactor; 350 C; 0.55 h; without H2) over the carbon supported Pt, Pd, and Ni catalysts. The results in Table 6 conrmed a high activity of the Pt and Ni Table 6. Yields of Products (% of C) from Conversion of Fatty Acids at 350 C (1 h) in Subcritical Water123
catalysts no catalyst Pt/C Pd/C NiC Pt/NiC PtC-Pa PdC-Pa
a

C14 0 1.2 0.3 54.5 19.0 0.2 0

C714,16,18 0.2 1.4 0.4 2.0 3.6 0.3 <0.1

C15 <0.1 5.5 0.7 0.4 1.6 0.4 <0.1

C17 <0.1 40.8 6.2 1.0 5.6 3.8 <0.1

fatty acids 80.9 7.6 66.7 0.7 21.6 48.3 83.3

During preparation, strongly basic anion-exchange resin was used after treatment with aqueous NaOH.

catalysts as indicated by low content of free fatty acids in the product mixture. However, while the former catalyst had a high selectivity for C15 and C17 hydrocarbons, methane was a dominant product over the Ni catalyst. At the end of experiments, the spent Pt and Ni catalysts were isolated from the reaction mixture for reuse. A high activity was maintained during the three subsequent test cycles conducted under identical conditions. The results support the involvement of the active hydrogen generated in situ via reactions {1} and {2} (Figure 1), as well as the tentative mechanism in Figure 8. The pyrolysis oil containing 33 wt % of water was upgraded in two stages. After the rst stage carried out under mild conditions, partially upgraded feed was further upgraded over the sulded NiMo/Al2O3 catalyst at 350 C in a batch reactor.24 Under such conditions, stability of the catalyst was aected due to hydrolysis of the (-Al2O3 support. Duan et al.124 studied the biofeed from hydrothernal liquefaction of the duckweed biomass under subcritical conditions (330370 C; 2 and 4 h; autoclave) in either H2 or CO using the sulded Pt/C catalyst. The yield and quality of liquids (e.g., lower viscosity, higher hydrogen content, etc.) were better under CO than under H2. This conrmed a higher reactivity of the in situ generated hydrogen as discussed earlier (Figure 8). Rather viscous and tarry product was obtained
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research during the parallel experiments conducted in the absence of catalyst. The catalysts such as Pd/C, Pt/C, Ru/C, Ni/SiO2Al2O3, sulded CoMo/Al2O3 and zeolite with the variable SiO2/Al2O3 ratio were used for the HPR of a microalga Nanochloropsis feed.86 The experiments in a batch reactor (350 C; 60 min) were conducted with the mixture containing 0.384 g of catalyst, 4.27 g of microalgae paste and 13.5 mL of deionized water. All catalysts produced higher yields in N2 than in H2. The higher yields under N2 can be attributed to a highly reactive hydrogen generated in situ via WGS of CO released via decarbonylation of microalgae (i.e., reactions {1} and {2}).12 With respect to the yield, the most active catalysts under N2 and H2 (3.5 MPa) were Pd/C and Pt/C catalysts, respectively. 6.3.3. Supercritical Conditions. Dickenson et al.125 studied the HDO of benzofuran in di-ionized water over the Pt(5%)/C catalyst in a batch reactor at 380 C. The eects of the water density, amount of water, and corresponding total pressure (6 to 29 MPa) as well as the H2/reactant ratio on the conversion and products distribution were investigated. With respect to hydrocarbons production, the Pt/C catalyst exhibited a good activity and stability. Duan and Savage126 studied hydrothermal transformations of pyridine in an aqueous medium with the aim to simulate the HPR of an algae biofeed under SCW conditions. Nitrogen removal from the algae biofeed above critical temperature of water may have some advantages. For example, a large portion of ammonia produced from nitrogen compounds ends up in aqueous phase rather than in oil phase. The eciency of upgrading can be enhanced in the presence of catalyst. In this regard, a series of commercially available catalysts (i.e., 5% Pt/ C, 5% Pd/C, 5% Ru/C, 5% Rh/C, 5% Pt/C-sulded, 5% Pt/ Al2O3, Mo2C, MoS2, PtO2, Al2O3, CoMo/Al2O3-sulded and activated carbon) were evaluated for potential applications (380, 400, 420 C, 10150 min, catalyst loading of 50200 wt %, H2 pressure of 06.9 MPa). The 5% Pt/Al2O3 catalyst exhibited the highest activity and stability. The conversion of palmitic acid was studied in batch reactor using Pt (5%)/C and Pd(5%)/C catalysts under a near critical and SCW (380 C; 28 MPa) conditions without H2 being present.127 The Pt/C catalyst was more active than Pd/C catalyst giving more than 90% selectivity to pentadecane. At the end of run, the catalysts retained most of their activity for reuse. Under similar conditions, activated carbon alone was used for conversion palmitic and oleic acids. In this case, n-C8C15 and n-C12C17 were the major products.128 The absence of alkenes among the products conrmed an in situ formation of active hydrogen which was consumed in HYD reactions. Both water and fatty acid molecules could be the source of active hydrogen. The SCW hydrothermal process, operating in the presence of Pt/C catalyst under a high H2 pressure was developed by Duan and Savage127,128 for upgrading biofeed from liquefaction of a microalgae. The upgraded oil was a freely owing liquid compared with a tarry consistency biofeed. The characterization of upgraded bio oil identied 72 compounds which accounted for almost 70% of the product mixture A series of n-alkanes starting at C9 were dominant species in the products. This conrmed a high level of the HYD of alkenes present in the feed. For example, phytenes present in the biofeed were converted to phytane (2,6,10,14-tetramethylhexadecane). A signicant amount of alkyl substituted benzenes was formed as well. The derivatives of piperidine, indole and O-methyloxime, which were present in the crude material, were not detected in
17710

Review

the products after the supercritical upgrading process. This suggests that the catalyst and reaction conditions used caused an extensive denitrogenation. The overall content of cholesterol, cholestane, and cholestene decreased in the treated oils. The only sulfur-containing compound, such as 1-methyl-2-piperidinethione, detected in the biofeed, was removed during upgrading. The experimental conditions had a pronounced eect on the products distribution.128 Without Pt/C catalyst, the content of fatty acid in the produced liquids was rather high. Basic conditions favored the conversion of fatty acids. At the same time, the relative amount of pentadecane in the product oil was much higher than that in the crude feed. In another study, the biofeed from the hydrothermal liquefaction of a microalgae was upgraded over the Pd(5 wt.%)/carbon catalyst in a SCW at 400 EC and a partial pressure of 3.4 MPa of H2.127 The longer reaction times and a lower feed/catalyst ratio increased the amount of gas and coke. At the same time, the quality of liquid products improved, although their yield decreased. The cubic and octahedral CeO2 nanoparticles of 8 and 50 nm size, respectively, were used for the conversion of bitumen in SCW at 723 K.97 In this case, the solution of 10 wt % of bitumen in 1-methylnaphthalene was used to reduce viscosity of the former. The batch reactor was lled with the solution (1 g) and water (1 g) and slurried with 1020 mg of catalyst. Signicant reduction in the asphaltenes content and an increase in the yield of maltenes (toluene solubles-pentane insolubles) was achieved, particularly in the presence of the cubic CeO2. The temperature employed during these experiments indicates complex mechanism of the overall conversion involving thermal decomposition of asphaltenes initially, followed by stabilization of radicals via reactions {5} to {8} and scavenging cycle in Figure 6.

7. CONCLUSIONS For high oxygen and water contents feeds, the HPR in aqueous phase for production of liquid hydrocarbons is an alternative to conventional HPR. Potential for generating in situ hydrogen via partial reforming and WGS during the HPR in aqueous phase may be an answer to large hydrogen consumption observed during upgrading high oxygen content feeds using conventional HPR. Advancements toward commercialization of the aqueous phase HPR of high oxygen content feeds may be accelerated if an active and selective catalysts exhibiting desired performance in the presence of large quantities of water, can be developed. Depending on feed composition, a multifunctional catalyst possessing high activities for HYD, HDO, HCR, hydrolysis, hydrogenolysis, WGS, and dehydration may be required. Published information suggests that this stage of catalysis may be approached by an optimal selection of active metals and support as well as by performing overall HPR in stages. Good performance was exhibited by noble metals (e.g., Pt, Pd, Ru, and Rh) catalysts supported on supports with suitable surface acidity. It should be, however, noted that most of the experimental results were obtained in batch reactors. A parallel testing in continuous xed bed reactors is needed to obtain data for scale-up consideration. In this case, the kinetics of HPR reactions in aqueous phase require attention. Thus, traditional expressions used to study kinetics of HPR under conventional conditions must be modied to account for potential involvement of water.
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research Among other factors, high consumption of hydrogen during the conventional HPR of heavy petroleum feeds may be the reason for interests in HPR in aqueous phase. A higher reactivity of the in situ generated hydrogen (via WGS and partial reforming) than that of the external H2 was clearly demonstrated. Consequently, formation of coke was diminished and the yields of liquid products increased. In this regard, both active hydrogen and water molecules played a benecial role. Moreover, the transfer of hydrogen from water to products was conrmed. In spite of the potential benets, the interests in aqueous phase HPR of heavy feeds have faded. Apparently, catalysts and reactor design for such applications are much more challenging than those for liquid feeds. The results obtained predominantly in batch reactors need to be compared with those from continuous systems. Among reactors type, xed bed, ebullated bed, and slurry bed reactors have to be carefully evaluated for aqueous HPR. It is believed that ebullated and slurry bed reactors may be more advantageous compared with xed bed reactors. For these purposes, an extensive database of experimental parameters has still to be established. In published studies on the HPR in aqueous phase, material problems caused by enhanced corrosivity by water are rarely addressed. As expected, corrosivity increases with increasing temperature. Therefore, it may require special attention during the design of catalytic reactors as well as upstream and downstream units, particularly for SCW.

Review

AUTHOR INFORMATION

Corresponding Author

*E-mail: edfurimsky@hotmail.com.
Notes

The authors declare no competing nancial interest.

REFERENCES

(1) Amin, S. Review on biofuels and gas production processes from microalgae. Energy Convers. Manage. 2009, 50, 1834. (2) De Klerk; A. and Furimsky, E. Catalysis in the Rening Fischer Tropsch Syncrude; RSC Publishing: Cambridge, U.K., 2010. (3) Nel, R. J. J.; de Klerk, A. FischerTropsch aqueous phase refining by catalytic alcohol dehydration. Ind. Eng. Chem. Res. 2007, 46, 3558. (4) Nel, R. J. J.; de Klerk, A. Oxygenates in FischerTropsch products. Am. Chem. Soc. Div. Fuel Chem. Prepr. 2009, 54, 118. (5) Furimsky, E. Lower emissions schemes for upgrading extra heavy petroleum feeds. Ind. Eng. Chem. Res. 2009, 48, 2752. (6) Topsoe, H.; Clausen, B. S.; Massoth, F. E. Hydrotreating catalysts. In Catalysis, Science and Technology; Anderson, J., Boudart, M. Eds.; Springer: New York, 1996; Vol. 11, p 1. (7) Massoth, F. E. Characterization of molybdena catalysts. Adv. Catal. 1978, 27, 265. (8) Ratnasamy, P.; Sivashanker, S. Structural chemistry of Co-Mo alumina catalysts. Catal. Rev.: Sci. Eng. 1980, 22, 401. (9) Furimsky, E.; Massoth, F. E. Hydrodenitrogenation of petroleum. Catal. Rev.: Sci. Eng. 2005, 47, 297. (10) Prins, R. Catalytic hydrodenitrogenation. Adv. Catal. 2001, 46, 399. (11) Ho, T. Hydrodenitrogenation catalysts. Catal. Rev.: Sci. Eng. 1988, 30, 117. (12) Furimsky, E. Hydroprocessing challenges in biofuels production. Catal. Today 2013, 217, 13. (13) Furimsky, E. Catalytic hydrodeoxygenation. Appl. Catal. 2000, 199, 147. (14) Stanislaus, A.; Cooper, B. H. Aromatic hydrogenation catalysts. Catal. Rev.: Sci. Eng. 1994, 36, 75.
17711

(15) Furimsky, E. Catalysts for upgrading heavy petroleum feeds. Stud. Surf. Sci. Catal. 2007, 169, 1. (16) Marafi, A.; Stanislaus, A.; Furimsky, E. Kinetics of hydroprocessing of heavy feeds. Catal. Rev.: Sci. Eng. 2013, 52, 204. (17) Eisenberg, D. and Kauzmann, W. The Structure and Properties of Water; Oxford University Press: Oxford, U.K., 2005. (18) Markus, M. Supercritical Water, Properties and Uses; Wiley: New York, 2012. (19) Palmer, D. A.; Fernandez-Prini, R.; Harvey, A. H. Aqueous Systems at Elevated Temperatures and Pressures; Elsevier: Amsterdam, 2004. (20) Furimsky, E. Carbons and Carbon Supported Catalysts in Hydroprocessing; RSC Publishing: Cambridge, U.K., 2008. (21) Yeh, T. M.; Dickinson, J. G.; Franck, A.; Linic, S.; Thompson, L. T., Jr.; P.E. Savage, P. E. Hydrothermal catalytic production of fuels and chemicals from aquatic biomass. J. Chem. Technol. Biotechnol. 2013, 88, 13. (22) Savage, P. E. A perspective on catalysis in supercritical water. J. Supercrit. Fluids 2009, 47, 407. (23) Elliott, D. C.; Baker, E. G. Energy Biomass Wastes 1986, 765. (24) French, R. J.; Stunkel, J.; Baldwin, R. M. Mild hydrotreating of bio-oil. Energy Fuels 2011, 25, 3266. (25) Zhang, J.; Li, J.-B.; Wu, S.-B.; Liu, Y. Advances in catalytic production and utilization of sorbitol. Ind. Eng. Chem. Res. 2013, 52, 11799. (26) Liu, Y.; Bai, F.; Zhu, C.-C.; Yuan, P.-Q.; Cheng, Z.-M.; Yuan, W.-K. Upgrading of residual oil in sub- and supercritical water. Fuel Proc. Technol. 2013, 106, 281. (27) Peterson, A. A.; Vogel, F.; Lachance, R. P.; Froling, M.; Antal, M. J., Jr.; J.W. Tester, J. W. Thermochemical biofuel production in hydrothermal media. Energy Environ. Sci. 2008, 1, 32. (28) Huber, G. W.; Dumesic, J. A. An overview of catalytic processes for production of hydrogen and alkanes in biorefinery. Catal. Today 2006, 111, 119. (29) B. Peng, B.; Zhao, C.; Meja-Centeno, I.; Fuentes, G. A.; Jentys, A.; Lercher, J. A. Comparison of kinetic and reaction pathways for hydrodeoxygenation of C3 alcohols on Pt/Al2O3. Catal. Today 2012, 183, 3. (30) Chen, L.; Zhu, Y.; Zheng, H.; Zhang, C.; Y. Li, Y. Aqueous phase HDO of propanoic acid over Ru/ZrO2 and Ru-Mo/ZrO2. Appl. Catal. 2011, 411/412, 95. (31) Kanie, Y.; Akiyama, K.; Iwamoto, M. Reaction pathways of glucose and fructose on Pt nano particles in supercritical water. Catal. Today 2011, 178, 58. (32) Hong, D. Y.; Miller, S. J.; Agrawal, P. K.; Jones, C. W. Hydrodeoxygenation and coupling of aqueous phenolics over bifunctional zeolite supported metal catalysts. Chem. Commun. 2010, 46, 1038. (33) Zhao, C.; Kou, Y.; Lemonidou, A. A.; Li, X. B.; Lercher, J. A. Highly selective catalytic conversion of phenoliv bio-oil to alkanes. Angew. Chem., Int. Ed. 2009, 48, 3987. (34) Zhao, C.; He, J.; Lemonidou, A. A.; Li, X. B.; Lercher, J. A. Aqueous phase hydrodeoxygenation of bio-derived phenols to cycloalkanes. J. Catal. 2011, 280, 8. (35) Zhao, C.; Kou, Y.; Lemonidou, A. A.; Li, X. B.; Lercher, J. A. Hydrodeoxygenation of bio-derived phenols to hydrocarbons using Raney Ni and and Nafion-SiO2 catalysts. Chem. Commun. 2010, 46, 412. (36) He, J.; Zhao, C.; Lercher, J. A. Ni-catalyzed cleavage of aryl ethers. J. Am. Chem. Soc. 2012, 134, 20768. (37) Davda, R. R.; Shabaker, J. W.; Huber, G. W.; Cortright, R. D.; Dumesic, J. A. Review of catalytic issues and process conditions for renewable hydrogen and alkanes by aqueous phase reforming of oxygenated hydrocarbons over supported catalysts. Appl. Catal., B 2005, 56, 171. (38) Shabaker, J. W.; Davda, R. R.; Huber, G. W.; Cortright, R. D.; Dumesic, J. A. Aqueous phase reforming of of methanol and ethylene glycol over alumina supported Pt catalysts. J. Catal. 2003, 215, 344.
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research


(39) Huber, G. W.; Shabaker, J. W.; Dumesic, J. A. Raney Ni-Sn catalyst for H2 production from biomass derived hydrocarbons. Science 2003, 300, 2075. (40) Huber, G. W.; Cortright, R. D.; Dumesic, J. A. Renewable alkanes by aqueous phase reforming of biomass-derived oxygenates. Angew. Chem., Int. Ed. 2004, 43, 1549. (41) Huber, G. W.; Chheda, J. N.; Barret, C. J.; Dumesic, J. W. Production of liquid alkanes by aqueous phase processing of biomassderived carbohydrates. Science 2005, 308, 1446. (42) Davda, R. R.; Shabaker, J. W.; Huber, G. W.; Cortright, R. D.; Dumesic, J. A. Aqueous phase reforming of ethylene glycol on silica supported catalysts. Appl. Catal. Eviron. 2003, 43, 13. (43) Cortright, R. D.; Davda, R. R.; Dumesic, J. A. Hydrogen from catalytic reforming of biomass-derived hydrocarbons in liquid water. Nature 2002, 418, 946. (44) Dreher, M.; Johnson, B.; Peterson, A. A.; Nachtegaal, M.; Wambach, J.; Vogel, F. Catalysis in supercritical water. Pathways of the methanation reaction and sulfur poisoning over Ru/C catalyst during reforming of biomolecules. J. Catal. 2013, 301, 38. (45) Rabe, S.; Nachtegaal, M.; Ulrich; Vogel, F. Towards understanding the catalytic reforming of biomass in supercritical water. Angew. Chem., Int. Ed. 2010, 49, 6434. (46) Chheda, J. N.; Dumesic, J. A. An overview of dehydration, aldol condensation and hydrogenation processes for production of liquid alkanes from biomass derived carbohydrates. Catal. Today 2007, 123, 59. (47) Leibbrandt, N. H.; Aboyade, A. O.; Knoetze, J. H.; Gorgens, J. F. Process efficiency of biofuels production via gasification and Fischer Tropsch synthesis. Fuel 2013, 109, 484. (48) Pham, T. N.; Shi, D.; Resasco, D. E. Evaluating strategies for catalytic upgrading of pyrolysis oil in liquid phase. Appl. Catal. Environ. 2013, 145, 10. (49) Jongerius, A. L.; Jastrzebski, R.; Bruijnincx, P. C. A.; Weckhuysen, B. M. Co-Mo sulfide catalyzed HDO of lignin model compounds. J. Catal. 2012, 285, 315. (50) Ancheyta, J.; Rana, M. S.; Furimsky, E. Hydroprocessing of heavy petroleum feed. Tutorial. Catal. Today 2005, 109, 3. (51) Kozhevnikov, I. V.; Nuzhdin, A. L.; Martyanov, O. N. Transformation of asphaltenes in supercritical water. J. Supercrit. Fluids 2010, 55, 217. (52) Han, L. N.; Zhang, R.; Bi, J. C. Reactivity of coal tar pitch in supercritical water. Fuel Proc. Technol. 2009, 90, 292. (53) Sato, T.; Adschiri, T.; Arai, K.; Rempel, G. R.; Ng, F. T. T. Upgrading of asphalt with and without supercritical water. Fuel 2003, 82, 1231. (54) Zhu, C.-C.; Ren, C.; Tan, X.-C; Chen, G.; Yuan, P.-Q.; Cheng, Z.-M.; Yuan, W.-K. Initiated pyrolysis of heavy oil in presence of near critical water. Fuel Process. Technol. 2013, 111, 111. (55) Tan, X.-C.; Zhu, C.-C.; Liu, Q.-K.; Ma, T.-Y.; Yuan, P.-Q.; Cheng, Z.-M.; Yuan, W.-K. Co-pyrolysis of heavy oil and low density polyethylene in presence of supercritical water. Fuel Process. Technol. 2014, 108, 49. (56) Bai, F.; Zhu, C.-C.; Liu, Y.; Yuan, P.-Q.; Cheng, Z.-M.; Yuan, W.-K. Co-pyrolysis of residual oil and polyethylene in sub- and supercritical water. Fuel Process. Technol. 2013, 106, 267. (57) Dutta, W. R. P.; McCaffrey, W. C.; Gray, M. R.; Muehlenbachs, K. Thermal cracking of Athabasca bitumen. Influence of steam on reaction chemistry. Energy Fuels 2013, 14, 671. (58) Xu, T.; Liu, Q.; Liu, Z.; J. Wu, J. The role of supercritical water in pyrolysis of carbonaceous compounds. Energy Fuels 2013, 27, 3148. (59) Patwardhan, P. R.; Timko, M. T.; Class, C. A.; Bonomi, R. E.; Kida, Y.; Hernandez, H. H.; Tester, J. W.; Green, W. H. Supercritical water desulfurization of organic sulfides. Energy Fuels 2013, 27, 6108. (60) Goudriaan, F.; Peferoen, D. G. R. Liquid fuels from biomass via hydrothermal process. Chem. Eng. Sci. 1990, 45, 2729. (61) Suzuki, T.; Itoh, M.; Takegami, Y.; Y. Watanabe, Y. Chemical structure of tar sands bitumen by 1H and 13C NMR spectroscopy. Fuel 1982, 61, 402.
17712

Review

(62) Cheng, Z.M..; Ding, Y.; Zhao, L. Q.; Yuan, P. Q.; Yuan, W. K. Effect of supercritical water in vacuum residue upgrading. Energy Fuels 2009, 23, 3178. (63) Zhao, L. Q.; Cheng, Z. M.; Ding, Y.; Yuan, P. Q.; Lu, S. X.; Yuan, W. K. Experimental studyon vacuum residue upgrading through pyrolysis in supercritical water. Energy Fuels 2006, 20, 2067. (64) Sato, T.; Tomita, T.; Trung, P. H.; Itoh, N.; Sato, S.; Takanohashi, T. Upgrading of bitumen in the presence of hydrogen and carbon dioxide in supercritical water. Energy Fuels 2012, 27, 646. (65) Sato, T.; Mori, S.; Watanabe, M.; Sasaki, M.; Itoh, N. Upgrading of bitumen with formic acid in supercritical water. J. Supercrit. Fluids 2010, 55, 232. (66) Zachariah, A.; Wang, L.; Yang, S.; Prasad, V.; de Klerk, A. Suppression of coke formation during bitumen pyrolysis. Energy Fuels 2013, 27, 3061. (67) Morimoto, M.; Sugimoto, Y.; Saotome, Y.; Sato, S.; Takanohashi, T. Effect of supercritical water on upgrading of oil sand bitumen. J. Supercrit. Fluids 2010, 55, 223. (68) Han, L.; Zhang, R.; Bi, J. Reactivity of coal tar pitch in supercritical water. J. Fuel Chem. Technol. 2008, 36, 1. (69) Kokubo, S.; Nishida, K.; Hayashi, A.; Takahashi, H.; Yokota, O.; Inage, S. I. Effective demetallization and suppression of coke formation in supercritical water. J. Japan Petr. Inst. 2008, 51, 309. (70) Dote, Y.; Sawayama, S.; Inoue, S.; Minowa, T.; Yokoyama, S. Recovery of liquid fuel from microalgae by thermochemical liquefaction. Fuel 1994, 73, 1855. (71) Marafi, A.; Kam, E.; Stanislaus, A. Kinetic study of non-catalytic reactions during hydroprocessing. Fuel 2008, 87, 2131. (72) Wahyudiono, M.; Sasaki, M.; Goto, M. Kinetic study for liquefaction of tar in sub- and super-critical water. Polymer Degrad. Stability 2008, 93, 1194. (73) Han, L.-N.; Zhang, R.; Bi, J.-C. High temperature coal tar upgrading in supercritical water. J. Fuel Chem. Technol. 2010, 41, 633. (74) Milliren, A. L.; Wissinger, J. C.; Gottumukala, V.; Schall, C. A. Kinetics of soybean oil hydrolysis in subcritical water. Fuel 2013, 108, 277. (75) Valdez, P. J.; Nelson, M. C.; Faeth, J. L.; Wang, H. Y.; Lin, X. L.; Savage, P. E. Hydrothermal liquefaction of bacteria and yeast monocultures. Energy Fuels 2013, DOI: 10.1021/ef401506u. (76) Yong, T.L.-K.; Matsumura, Y. Reaction kinetics if lignin conversion in supercritical water. Ind. Eng. Chem. Res. 2012, 51, 11975. (77) Yong, T.L.-K.; Matsumura, Y. Kinetic analysis of lignin hydrothermal conversion in sub- and supercritical water. Ind. Eng. Chem. Res. 2013, 52, 5626. (78) Yong, T.L.-K.; Matsumura, Y. Kinetic analysis of guaiacol conversion in sub- and supercritical water. Ind. Eng. Chem. Res. 2013, 52, 9048. (79) Montgomery, W.; Court, R. W.; Rees, A. C.; Sephton, M. A. High temperature reactions of water with heavy oil and bitumen. Fuel 2013, 113, 426. (80) Li, L.; Egiebor, N. O. Oxygen removal from coal during supercritical water and toluene extraction. Energy Fuels 1992, 6, 35. (81) Massoth, F. E.; Politzer, P.; Concha, M. C.; Murray, J. S.; Jakowski, J.; Simons, J. Catalytic HDO of methyl substituted phenols: Correlation of kinetic parameters with molecular properties. J. Phys. Chem. B 2006, 110, 14283. (82) Wan, H.; Chaudhari, R. V.; Subramanian, B. Aqueous phase hydrogenation of acetic acid and its promotional effect on HDO of pcresol. Energy Fuels 2013, 27, 487. (83) Wu, X.; Fu, J.; Lu, X. Kinetics and mechanism of hydrothermal decomposition of lignin model compounds. Ind. Eng. Chem. Res. 2013, 52, 5016. (84) Kumar, M.; Akgerman, A.; Anthony, R. G. Desulfurization by in situ hydrogen generation through watergas shift reaction. Ind. Eng. Chem. Res. 1984, 23, 88. (85) Hook, B.; Akgerman, A. Desulfurization of DBT by in situ hydrogen generation through watergas shift reaction. Ind. Eng. Chem. Res. 1986, 25, 278.
dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Industrial & Engineering Chemistry Research


(86) Duan, P.; Savage, P. E. Hydrothermal liquefaction of microalga with heterogeneous catalysts. Ind. Eng. Chem. Res. 2011, 50, 52. (87) Adschiri, T.; Shibata, R.; Sato, T.; Watanabe, M.; K. Arai, K. Catalytic HDS of DBT through partial oxidation and watergas shift reaction in supercritical water. Ind. Eng. Chem. Res. 1998, 25, 2634. (88) Arai, K.; Adschiri, T.; Watanabe, M. Hydrogenation of hydrocarbons through partial oxidation in supercritical water. Ind. Eng. Chem. Res. 2000, 39, 4697. (89) Ng, F. T. T.; Milad, I. K. Catalytic desulfurization of BT in emulsion via in situ hydrogen. Appl. Catal. 2000, 200, 243. (90) Liu, Y.; Shen, W.; Song, Y.; J. Cheng, J. The mechanism of diesel hydrogenation using supercritical water syngas. Pet. Sci. Technol. 2006, 24, 1283. (91) Yuan, P.-Q.; Cheng, Z.-M.; Zhang, X.-Y.; Yuan, W.-K. Catalytic denitrogenation through partial oxidation in supercritical water. Fuel 2006, 85, 367. (92) Lee, R. Z.; Ng, F. T. T. Effect of water on HDS of DBT over dispersed Mo catalysts using in-situ generated hydrogen. Catal. Today 2006, 116, 505. (93) Arai, K.; Adschiri, T.; Watanabe, M. Hydrogenation of hydrocarbons through partial oxidation in supercritical water. Ind. Eng. Chem. Res. 2000, 39, 4697. (94) Breysse, M.; Furimsky, E.; Kasztelan, S.; Lacroix, M.; Perot, G. Hydrogen activation by transition metal sulfides. Catal. Rev.: Sci. Eng. 2002, 44, 651. (95) Elliot, D. C. Historical developments in hydroprocessing biooils. Energy Fuels 2007, 21, 1792. (96) Zhang, X.; Wang, T.; Ma, L.; Zhang, Q.; Jiang, T. Hydrotreatment of bio-oil over Ni-based catalyst. Bioresour. Technol. 2013, 127, 306. (97) Ates, A.; Azimi, G.; Choi, K.-Y.; Green, W. H.; Timko, M. T. The role of catalyst in supercritical desulfurization. Appl. Catal. Envir. 2014, 147, 144. (98) Dejhosseini, M.; Aida, T.; Watanabe, M.; Takami, S.; Hojo, N.; Aoki, N.; Arita, T.; Kishita, A.; Adschiri, T. Catalytic cracking reaction of heavy oil in the presence of cerium oxide nanoparticles in supercritical water. Energy Fuels 2013, 27, 4624. (99) Bej, A. Performance and evaluation of hydroprocessing catalysts. Energy Fuels 2002, 16, 774. (100) Pitault, I.; Fongarland, P.; Mitrovic, M.; Ronze, D.; Forissier, M. Choice of laboratory scale reactors for HDT kinetic studies and catalyst testing. Catal. Today 2005, 98, 31. (101) Lee, J.; Xu, X.; Huber, G. W. High throughput screening of monometallic catalysts for aqueous phase hydrogenation of biomass derived oxygenates. Appl. Catal. Environ 2013, 140/141, 98. (102) Li, L.; Coppola, E.; Rine, J.; Miller, J. L.; D. Walker, D. Catalytic hydrothermal conversion of triglycerides to non-ester biofuels. Energy Fuels 2010, 24, 1305. (103) Zohrer, H.; Mayr, F.; F. Vogel, F. Stability and performance of ruthenium catalysts based on refractory oxide supports in supercritical water. Energy Fuels 2013, 27, 4739. (104) Liu, D.; Zhang, L.; Yang, B.; Li, J. Investigations on the loss mechanism and kinetics of MoO3 on alumina and silica. Appl. Catal. 1993, 105, 185. (105) Wittmann, Z.; Kantor, E.; Belafi, K.; Peterfi, L.; Farkas, P. L. Phase composition analysis of hydrous aluminium oxides by thermal analysis and infrared spectroscopy. Talanta 1992, 39, 1583. (106) Dabbagh, H. A.; Zamani, M. Catalytic conversion of alcohols over over alumina-zirconia mixed oxide. Appl. Catal. 2011, 404, 141. (107) Ravenelle, R. M.; Copeland, J. R.; Kim, W.-G.; Crittenden, J. C.; Sievers, C. Structural changes of -Al2O3 Supported Catalysts in hot water. ACS Catal. 2011, 1, 552. (108) Ravenelle, R. M.; Schussler, F.; DAmico, A.; Danilina, N.; van Bokhoven, J. A.; Lercher, J. A.; Jones, C. W.; Sievers, C. Stability of zeolites in hot water. J. Phys. Chem. C 2010, 114, 19582. (109) Orinakova, R.; Orinak, A. Recent applications of carbon nanotubes in hydrogen production and storage. Fuel 2011, 90, 3123.
17713

Review

(110) Wan, H.; Chaudhari, R. V.; Subramanian, B. Catalytic hydroprocessing of p-cresol. Metal, solvent, and mass transfer effects. Top. Catal. 2012, 55, 129. (111) Elliott, D. C.; Hart, T. S. Catalytic hydroprocessing of chemical models bio-fuels. Energy Fuels 2009, 23, 631. (112) Ohta, H.; Kobayashi, H.; Hara, K.; Fukuoka, A. Hydrodeoxyganation of phenols as lignin models under acid free conditions over carbon supported Pt catalysts. Chem. Commun. 2011, 47, 12209. (113) Duran -Martn, D.; Ojeda, M.; Lop ez Granados, M.; Fierro, J. L. G.; Mariscal, R. Stability and regeneration of Cu-ZrO2 catalyst used in glycerol hydrogenolysis. Catal. Today 2013, 210, 98. (114) Patil, P. T.; Armbruster, U.; Richter, M.; Martin, A. Heterogenously catalyzed hydroprocessing of organosolv lignin in sub- and super-critical solvents. Energy Fuels 2011, 23, 4713. (115) Kato, Y.; Sekine, Y. One pot direct catalytic conversion of cellulose to hydrocarbons by decarbonation using Pt/H-zeolite catalyst. Catal. Lett. 2013, 143, 418. (116) Zhu, S.; Qiu, Y.; Zhu, Y.; Hao, S.; Zheng, H.; Li, Y. Hydrogenolysis of glycerol to 1,3-propanediol over bifunctional catalyst containing Pt and heteropolyacids. Catal. Today 2013, 212, 120. (117) Yang, Q.; Zhang, J.; Zhang, L.; Fu, H.; Zheng, X.; Yuan, M.; Chen, H.; Li, R. Ruthenium nano particles on colloidal carbon spheres: Efficient catalyst for hydrogenation of ethyl lactate in aqueous phase. Catal. Commun. 2013, 40, 37. (118) Sun, H.-J.; Pan, Y.-J.; Jiang, H.-B.; Li, S.-H.; Zhang, Y.-X.; Liu, S.-C.; Liu, Z.-Y. Effect of transition metals on hydrogenation properties of benzene over Ru based catalysts. Appl. Catal. 2013, 464/465, 1. (119) Chen, X.; Wang, X.; Yao, S.; Mu, X. Hydrogenolysis of biomass derived sorbitol to glycols and glycerol over Ni-MgO catalyst. Catal. Commun. 2013, 39, 86. (120) You, S. J.; Baek, L. G.; Park, E. D. Hydrogenolysis of cellulose into polyols over NiW/SiO2 catalysts. Appl. Catal. 2013, 466, 161. (121) Rivera Olver, J. N.; Gutier rez, G.J..; Romero Serrano, J. A.; Medina Ovando, A.; Garibay Febles, V.; Barriga Arceo, L. D. Use of unsupported mechanically alloyed NiMoWC nano catalyst to reduce viscosity of aquathermolysis reaction of heavy oil. Catal. Commun. 2014, 43, 131. (122) Fu, J.; Lu, X.; Savage, P. E. Hydrothermal decarboxylation and hydrogenation of fatty acids over Pt/C. ChemSusChem 2011, 4, 481. (123) Idesh, S.; Kudo, S.; Norinaga, K.; Hayashi, J. Catalytic hydrothermal reforming of water soluble organics from pyrolysis of biomass using Ni carbon catalysts impregnated with Pt. Energy Fuels 2013, 27, 4796. (124) Duan, P.; Xu, Y.; Bai, X. Upgrading of crude duckweed bio-oil in sub- critical water. Energy Fuels 2013, 27, 4729. (125) Dickinson, J.; Poberezny, J. T.; Savage, P. E. Deoxygenation of benzofuran in supercritical water over Pt catalyst. Appl. Catal. Environ. 2012, 123/124, 357. (126) Duan, P.; Savage, P. E. Catalytic hydrothermal hydrodenitrogenation of pyridine. Appl. Catal. Environ. 2011, 108/109, 54. (127) Duan, P.; Savage, P. E. Catalytic hydrotreatment of crude algal bio-oil in supercritical water. Appl. Catal. Environ. 2011, 104, 136. (128) Duan, P.; Savage, P. E. Upgrading of crude algal bio-oil in supercritical water. Bioresour. Technol. 2011, 102, 1899.

dx.doi.org/10.1021/ie4034768 | Ind. Eng. Chem. Res. 2013, 52, 1769517713

Das könnte Ihnen auch gefallen