Sie sind auf Seite 1von 12

Applied Catalysis A: General 397 (2011) 112

Contents lists available at ScienceDirect

Applied Catalysis A: General


journal homepage: www.elsevier.com/locate/apcata

Review article

Renewable fuels via catalytic hydrodeoxygenation


T.V. Choudhary , C.B. Phillips
ConocoPhillips Company, Bartlesville Technology Center, Bartlesville, OK 74004, USA

a r t i c l e

i n f o

Article history:
Received 1 November 2010
Received in revised form 14 February 2011
Accepted 20 February 2011
Available online 24 February 2011
Keywords:
Bio-fuels
Deoxygenation
Renewables
Pyrolysis oils
Triglycerides

a b s t r a c t
There is considerable interest in investigating the deoxygenation process, due to the high oxygen content
of the feed-stocks used for the production of renewable fuels. This review addresses studies related to
the catalytic hydrodeoxygenation of two feed-stocks (a) oils with high content of triglycerides and (b)
oils derived from high pressure liquefaction or pyrolysis of biomass. Future research directions that could
potentially bridge the existing gaps in these areas are provided.
2011 Elsevier B.V. All rights reserved.

Contents
1.
2.

3.

4.
5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HDO of triglyceride-based feeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Hydroprocessing in stand-alone mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.1.
Inuence of process conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.2.
Inuence of feed and catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Hydroprocessing in co-processing mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.
Potential future research areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HDO of bio-oil feeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
HDO of bio-oils derived from high pressure liquefaction of biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1.
Inuence of nature of feed-stock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.2.
Inuence of catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.3.
Inuence of operating conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
HDO of bio-oils derived from pyrolysis of biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1.
Stabilization studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.2.
Dual stage studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.3.
Combined dual stage studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.
Potential future research areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Life cycle assessment (LCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction
Renewable transportation fuels may be generally dened as
those derived from the processing and upgrading of various forms

Corresponding author.
E-mail address: tvchoud@yahoo.com (T.V. Choudhary).
0926-860X/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.02.025

1
3
3
4
5
6
6
7
7
7
7
7
7
8
9
9
10
11
11
11

of biomass and degradable municipal waste feedstocks. Typical products are hydrogen, methane, propane, ethanol, butanol,
gasoline and diesel. Another denition used by regulatory and governing bodies around the world describes a renewable fuel as any
fuel derived from renewable sources of biomass designed to reduce
the amount of fossil fuel within the transportation fuel pool of a
region [1]. Renewable fuels are often classied into three generations [2]; those produced from (i) the conventional processing

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

of edible feedstocks (e.g., cane-based ethanol via fermentation,


biodiesel via esterication); (ii) the advanced processing (e.g., gasication, hydroprocessing, pyrolysis) of non-edible feedstocks (e.g.,
waste greases, lignocelluloses, refuse) and (iii) the harvesting and
advanced processing of ultra-high yield biomass (e.g., algae).
A simplistic schematic showing renewable fuels production
from biomass is provided in Fig. 1. The biomass itself may be
broken down into three basic categories, carbohydrates, lignin
and fats/oils. Carbohydrates primarily include cellulose and hemicellulose fractions. Fats are mainly comprised of triglycerides and
fatty acids. Corma [3], Huber [4], Kamm [5] and Centi and van Santen [6] have adequately outlined the various bio-rening strategies
for upgrading these fractions into chemicals and fuels using block
process ow diagrams. The intent of this paper is not to review
the entire research area but rather to focus exclusively on conventional hydroprocessing studies (with real feed-stocks) as related to
production of renewable fuels from catalytic hydrodeoxygenation
(HDO) of fats/vegetable oils and biomass derived oils.
HDO is an example of hydroprocessing application for the
production of both renewable (green) gasoline and renewable
(green) diesel as discussed in detail by Marker et al. [7]. HDO
reactions also occur during the hydroprocessing of petroleum
fractions along with hydrodesulfurization (HDS), hydrodenitrogenation (HDN), hydrodemetallization (HDM) and the saturation
of olens/aromatics [8].
(thiophene)C4 H4 S + 3H2 C4 H8 + H2 S
C4 H8 + H2 C4 H10

HDS

(1)

Saturation

(pyrrole)C4 H5 N + 3H2 C4 H8 + NH3


(furan)C4 H4 O + 3H2 C4 H8 + H2 O

(2)
(3)

HDN
HDO

(4)

These hydroprocessing reactions use high pressure hydrogen


to remove S, O and N heteroatoms out of petroleum feedstocks
through a series of hydrogenolysis and hydrogenation steps. Since,
the oxygen in petroleum is typically less than 3000 ppmw (Table 1),
less attention has been paid to HDO as compared to HDS as far
as petroleum upgrading research is concerned. However, in the
hydroprocessing of biomass feedstocks, HDO is a critical reaction step since a neat biomass feedstock may contain up to
500,000 ppmw oxygen with minimal amounts of sulfur. The oxygen
speciation within these feedstocks varies signicantly with the type
of biomass and upgrading methods employed [911]. For instance,
a renewable HDO feedstock produced from a gasication-to-liquids
facility may contain a signicant amount of alcohols and ethers.
However, a bio-oil HDO feedstock derived from the pyrolysis of
lignocellulose may contain a relatively disproportionate amount
of phenols, carboxylic acids and ketones. According to the work
by Bridgwater and Peacocke [12], a bio-oil produced from many
fast pyrolysis operations yields mostly water (2030 wt%), insoluble lignin (1520 wt%), aldehydes (1020 wt%), carboxylic acids
(1015 wt%) and carbohydrates (510 wt%).
Esterication and HDO are considered important commercial
routes for producing distillate range bio-fuels from triglyceridebased feed-stocks such as vegetable oils, animal fat and edible
grease. In the liquid-phase, commercial esterication reaction, a
fatty acid methyl ester (FAME) biofuel formed by reacting methanol
with triglycerides in the presence of caustic or acid catalyst is commonly referred to as bio-diesel.
CH2 (R 1 COO)CH(R 2 COO)CH2 (R 3 COO) + 3MeOH C3 H5 (OH)3
+ 3R  COOMe

(5)
R

= R1 ,

R2

R3

or
in
Here, the fatty hydrocarbon tail may vary
carbon chain length, as well as, number of unsaturated regions.
The FAME molecule (R COOMe) contains two oxygen atoms thereby

making it a diesel oxygenate. Additional information on this topic


can be found elsewhere [13].
When the distillate fuel is produced by catalytically removing
oxygen (e.g., HDO) completely from the triglyceride (assuming saturated acyl groups and no oxides of carbon production),
CH2 (R 1 COO)CH(R 2 COO)CH2 (R 3 COO) + 12H2 C3 H8 + 3R  H3
+ 6H2 O

(6)

the resulting straight-chain, renewable hydrocarbon fuel product (R H3 ) is referred to as renewable diesel in the marketplace.
Renewable diesel molecules are indistinguishable in molecular
structure from conventional petroleum-derived diesel molecules.
A comparison of the fuel properties was provided in the reports by
Kalnes et al. [14], Kuronen et al. [15] and Rantanen et al. [16]. From
an economics viewpoint according to Holmgren and co-workers
[17], one of the most important advantages of the HDO route is
that the bio-based feedstocks can be processed at reneries using
existing equipment and thereby minimizing capital cost.
A signicant amount of work has been dedicated to using
model compounds/surrogates for understanding the HDO reaction
chemistry. Model compounds/surrogates are selected based on the
fundamental structural classes describing the upgraded biomassderived feed-stocks namely; (i) lignin precursors to represent bio
oil, (ii) sugars and sugar alcohols to represent cellulose/hemicellulose and (iii) triglycerides and fatty acid derivatives to
represent lipids or fats/oils. A review by Furimsky [18] mentions
many of the oxygen containing model compounds present in both
petroleum oils and bio-oils derived from pyrolysis or liquefaction
processes. Bridgwater and co-workers [19] recently identied over
120 of these type of compounds in the fast GCMS pyrolysis of
rhizome from cassava plants. Liquefaction and pyrolysis oils are
often represented in the literature by the model compounds such
as, 2-methoxyphenol or guaiacol, ethylphenol, diethylsebecate and
methoxybenzene or anisole. However, one should note that these
compounds represent only the 25 wt% phenolic portion of the biooil and are more representative of lignin precursor molecules. For
cellulosic HDO feeds, C5 and C6 sugar alcohols such as, xylitol and
sorbitol, respectively have served as model species [4,20]. Fats and
lipid-oils are often represented by model triglycerides and fatty
acids such as, tristearate and stearic acid, respectively or as fatty
alkyl esters (e.g., methyl stearate). Krause and co-workers [2127]
have contributed signicantly to the understanding of the HDO
reaction for fats and their corresponding derivatives using model
compounds over the last decade. The model feedstock studies are
useful for determining the relative activities of HDO catalysts and
understanding HDO reaction kinetics. While some relevant model
compound studies will be mentioned herein, the main purpose of
this paper is to report the recent progress in the catalytic HDO of real
feed-stocks (e.g., vegetable oils, pyrolysis oils, etc.) that contain
varying degrees of oxygen.
Unlike model feed-stock studies which focus on the chemistry of a single representative compound or a simple mixture
of few representative compounds [28,29], real feed-stocks studies involve investigation of bulk effects arising from intertwining
interactions of complex mixtures of molecules [30,31]. From previous studies, it has been established that individual molecules can
lose their identity and respective inuence within complex mixtures such as those found in real feed-stocks [32]. This is related
to the strong adsorption of certain molecules on the active catalyst sites resulting in inhibition of other molecules that exhibit
weaker interaction with the catalyst sites [33]. Real feed-stock
studies thus offer practically relevant information that cannot be
easily obtained from model compound studies. These types of studies also offer information related to various contaminants that may
strongly inuence the overall feed-stock processability and impact

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

Fig. 1. Hydrodeoxygenation of the basic building blocks of biomass to renewable hydrocarbon fuels.

the catalyst stability. Real feedstock studies related to catalytic


HDO can be broadly classied into two main categories; (i) the
hydroprocessing of triglyceride-based feeds and (ii) the hydroprocessing of bio-oil feeds derived from the liquefaction or pyrolysis
of biomass. Herein, triglyceride-based feeds are addressed rst followed by a discussion on studies related to bio-oils derived from
biomass. Sugar-based feed-stocks have purposely been avoided in
this review since the production of higher hydrocarbons involves
condensation steps [34,35] outside of the traditional hydroprocessing operations.

[3638]. In fact, co-processed renewable fuel is virtually indistinguishable in the blend with petroleum fuel. Fatty-acid methyl esters
(FAME) derived from esterication processes on the other hand
contain signicant amounts of oxygen (up to 11 wt%) and differ signicantly in cold ow, volatility and combustion properties from
petroleum-based diesel fuel (Fig. 2). Compared to petroleum fuel,
FAME also suffers from other disadvantages in that it has storage
stability issues that can damage the fuel system, degrade the motor
oil more quickly and cause corrosion problems. The triglyceridebased feeds can be hydroprocessed to renewable fuels in presence
of (co-processing) or absence of (stand-alone) petroleum feedstocks. Related studies will be discussed in subsequent sections.

2. HDO of triglyceride-based feeds


One of the main advantages of the HDO route relative to other
methods for making biomass-derived diesel (e.g., FAME synthesis)
is that the corresponding renewable-fuel product is a high quality,
oxygen free, hydrocarbon fuel completely fungible with conventional petroleum-based renery fuel blendstocks and components

2.1. Hydroprocessing in stand-alone mode


Studies related to saturation of triglyceride-based feeds have
relevance in the food industry as well as the fuel industry. Efforts
in the food industry are often focused towards selective hydrogena-

Table 1
Atomic composition of various feed stocks and fuels adapted from Ref. [3,18,87].
Substance

C (mol%)

H (mol%)

O (mol%)

Pine (Hard)
Corn stover
Rice husk
Cottonseed hull
Cellulose
Hemi-cellulose
Lignin
Crude oil
Liquefaction bio-oil
Pyrolysis bio-oil
Gasoline
Diesel
Ethanol
DME
MTBE
Methane

32
33
29
22
29
33
40
3846
4044
3035
3337
3337
22
22
28
20

48
46
48
48
48
48
46
5472
4953
4550
6367
6367
67
67
66
80

20
21
23
30
23
19
14
<2
7
20
<0.01
<0.01
11
11
6
0

H/C
1.5
1.4
1.7
2.1
1.7
1.4
1.2
1.52.0
1.11.3
1.31.6
1.72.0
1.72.0
3.0
3.0
2.4
4.0

O/C
0.6
0.7
0.8
1.3
0.8
0.6
0.3
<0.03
0.10.3
0.60.7
<0.01
<0.01
0.5
0.5
0.2
0

Avg. formula
C1.6 H2.4 O
C1.6 H2.2 O
C1.3 H2.1 O
C0.7 H2.2 O
C1.3 H2.1 O
C1.7 H2.5 O
C2.9 H3.3 O
CH1.5
C6 H7.3 O
C1.6 H2.4 O
CH1.9
CH1.9
C2 H6 O
C2 H6 O
C5 H12 O
CH4

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

Fig. 2. Key property differences between different types of available diesel renewable fuels.
Adapted from [2].
Fig. 3. Effect of reaction temperature on Buriti oil.
Adapted from [53].

tion (Eq. (7)) of vegetable oils [39,40].


(triolein)C57 H104 O6 + 3H2 C57 H110 O6 (tristearate)

(7)

However, the hydroprocessing studies related to the renewable


fuel production involve complete conversion of the vegetable oils
to alkanes [4143]. As shown below, a single mole of a triglyceride
molecule with an acyl group (Ri ) containing xi number of carbon
atoms and ni CC double bonds can be reduced completely via
CO bond hydrogenolysis and hydrogenation to produce 3 mole of
straight-chain parafn diesel, 1 mol of propane and 6 mole of water.

carbon monoxide, respectively. The release of carbon monoxide is


accompanied by the simultaneous rejection of water.
(stearicacid)C17 H35 COOH C17 H36 + CO2

According to Vonghia et al. [52], triglyceride thermal decomposition occurs through a concerted -elimination pathway
producing unsaturated glycol di-fatty acid esters (UGDEs) and fatty
acids;

R1
O

R1 = (Cx1:n1)
R2

R3

R1

(12+3n) H2

R2
R3

H
H
H

H3C

CH3

R1 = (Cx1:n1)

2.1.1. Inuence of process conditions


Several sequential and concurrent reactions occur during the
hydroprocessing of triglyceride-based feeds. These reactions can be
simplistically classied as saturation of olenic bonds [44], degradation of triglycerides [45] and hydrogenation of intermediates
produced from the degradation of triglycerides [4648]. A complete
reaction network was illustrated by Huber and co-workers using
sunower oil [49] and more recently in the work by Kubicka and
Kaluza [50] using rapeseed oil. In the presence of hydrogen at elevated temperatures, the fatty acyl chains become saturated and/or
decomposed into fatty acids, mono- and diglyceride fractions
followed by decarboxylation, decarbonylation and hydrogenation
reactions to produce renewable diesel. In hydrogenolysis, fatty
alcohols are reduced to diesel and fatty esters [51] are reduced to
fatty alcohols directly or via dehydrationhydrogenation steps
(octadecanol)C18 H37 OH + H2 C18 H38 + H2 O

(8)

(methyloctadecanoate)C17 H33 COOCH3 + 3H2


C18 H37 OH + CH3 OH

R1
H

6H2O

In recent years, there has been considerable interest in


optimizing the production of renewable hydrocarbon fuels via
the hydroprocessing route. As described below, some studies
have focused on optimizing the process conditions while others
have investigated different catalyst options for renewable fuels
production.

(9)

The decarboxylation and decarbonylation reactions produce


long-chain alkanes and alkenes by releasing carbon dioxide and

(10)

R1
H

O
H

O
H

O
R2

R3

O
R2

O
R3

or by -hydrogen transfer resulting in CC bond cleavage within


the acyl group also producing straight-chain, olen diesel.
R3

R1
O

H2C

O
O

HO

O
R2
O

R1
O

O
O

O
R2

H
R'3

R'3 = (R3-1:n3+1)

In the absence of hydrogen, both decomposition routes can


occur simultaneously even within the same triglyceride molecule
allowing only one -elimination. However, in the presence of
hydrogen and conventional hydroprocessing catalysts, alkenes are
quickly saturated, thus allowing additional -eliminations to occur
off the glycerol back-bone.
Expectedly, the reaction temperature plays an important role
in dening the nal product composition. The inuence of reaction temperature on the yields from Buriti oil over a sulded NiMo
catalyst at a hydrogen pressure of 14 MPa is shown in Fig. 3 [53].
The long-chain alkane yield was found to decrease with increasing reaction temperature (360430 C), while the gas, cycloalkane
and alkylbenzenes yields were found to increase. Simacek and coworkers investigated the effect of reaction temperature on the
hydroprocessing of rapeseed oil over a commercial hydrotreating
NiMo catalyst [54]. The small content of un-reacted feed and oxygenated intermediates that was observed at 310 C was absent at
360 C. Formation of n-heptadecane and iso-alkanes was found to

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

increase with increasing temperature, whereas n-octadecane content decreased. These studies indicate that the decarboxylation
reaction, which results in formation of an alkane with one less
carbon atom than the corresponding carboxylic acid, is favored at
higher temperatures. The cold ow properties such as cloud point
and pour point were found to improve with increasing temperature, as expected from the increase in iso-alkanes content. The
qualitative trend for the temperature effect was similar at both
hydrogen pressures studied (7 MPa and 15 MPa). Also, a similar
temperature effect on the product yields (as described above) was
observed for different NiMo catalysts [55]. Investigation of CoMobased catalysts on different supports showed that an increase in
temperature decreased the selectivity of oxygenated products irrespective of the support used [56].
Model compounds as well as vegetable oil hydroprocessing
studies have shown that higher pressures are favorable for the
process [57,58]. In the HDO of tricaprylin and caprylic acid over
supported NiMo oxide catalysts, Boda et al. [59] discussed how
qualitative trends in total hydrogen pressure (1022 atm) could be
explained by LangmuirHinshelwoodHougenWatson (LHHW)
kinetic expressions;
r=

kKH2 pH2 Kacid Cacid


(1 + KH2 pH2 + Kacid Cacid )2

(11)

where the adsorption and reactions of fatty acids with surface


hydrogen determine the kinetic rates. For the cases where the relative amount of surface hydrogen is low due to the nature of the
catalyst (Kacid Cacid  KH2 CH2 ), the hydrogen pressure will increase
rate of reaction. Product yields obtained from Buriti oil processed
at 310 C and different hydrogen pressures over a sulded NiMo
catalyst in a batch reactor further illustrates the effect of pressure
[53]. The concentration of un-reacted fatty acid was high at the lowest pressure (7 MPa) investigated, however, at the highest pressure
investigated (14 MPa), the amount of un-reacted fatty acids was
negligible due to complete hydrogenation. Although the concentration of alkanes increased signicantly when the pressure was
increased from 7 MPa to 10 MPa, it remained almost constant when
the pressure was increased from 10 MPa to 14 MPa. The gas make on
the other hand increased almost linearly with increasing pressure.
Simacek and co-workers also investigated the effect of hydroprocessing pressure on the upgrading of rapeseed oil over a commercial
hydrotreating NiMo catalyst [54]. Interestingly, different trends
with pressure were observed in this study depending on the temperature investigated. At 310 C, the selectivity for n-heptadecane
decreased signicantly on increasing the pressure from 7 MPa to
15 MPa, however it was insensitive to pressure at 360 C. The selectivity for n-octadecane also showed a different trends depending
on the reaction temperature. At the lower reaction temperature
the selectivity for n-octadecane increased with increasing pressure, while the pressure effect on n-octadecane selectivity was
exactly opposite at the higher reaction temperature. The opposing
effects were also observed from the cold ow (cloud point and pour
point) properties of the products obtained under different pressure
and temperature conditions. Unfortunately no explanation was forwarded for these unexpected results.
The effect of pressure has also been investigated over organized mesoporous alumina (OMA) supported CoMo catalysts [56].
At 280 C, the conversion of triglycerides (in rapeseed oil) dropped
from 100% (pressure = 7 MPa) to 80% (pressure = 0.7 MPa). The
selectivity for oxygenated products and the nC18 /nC17 ratio
(280 C) was also found to be very sensitive to pressure. At the lower
pressure the selectivity for the oxygenated product was 45% while
it was about 10% at 7 MPa. The nC18 /nC17 ratio was considerably
lower at 0.7 MPa, indicating that there was a higher contribution
from the hydrodecarboxylation reaction at the lower pressure.

Fig. 4. Fatty acid composition of vegetable oils as determined by analysis of methyl


esters.
Adapted from [53].

OMA catalyst supports are synthesized with surfactants similar


to fatty acid intermediates (e.g., stearic and lauric acids) [60], so
the pore sizes (24 nm) allow space for reasonable diffusion rates
and reaction. A typical oleoyl or stearyl group with a characteristic chain length of roughly 21 A [61] can access the active sites on
the internal surfaces of these materials. The increased hydrodecarboxylation selectivity on OMA may be due to the small population
(26%) of penta-coordinated aluminum that provides intermediate
Lewis acidity [62] between octahedral (6570%) and tetrahedral
sites (2533%).
2.1.2. Inuence of feed and catalyst
The fatty acid composition of different vegetable oils is shown
in Fig. 4 [53]. Hydroprocessing studies have shown that the product
alkanes contain equivalent or one less carbon than the parent fatty
acids. The renewable diesel product composition of the vegetable
oil hydroprocessing reaction can thereby be roughly estimated
based on the fatty acid composition of the feed. However, depending on the chain length and degree of saturation of the vegetable
oils, some differences have been observed in the product yields [53].
From Fig. 4, the soybean, buriti and maracuja oils contain fatty acids
that have a longer chain length and are more unsaturated compared
to tucuma and babassu oils. The product gas yields were found to
increase with decrease in chain length and the degree of unsaturation of the vegetable oils. Further, the alkane selectivity was found
to increase while the cycloalkane selectivity was found to decrease
with an increase in the degree of saturation.
Studies have been undertaken to understand the effect of both
the active component as well as the support on the stand-alone
hydroprocessing of triglyceride-based feed stocks [6365]. The efcacy of NiMo/Al2 O3 , Ni/Al2 O3 and Mo/Al2 O3 sulded catalysts has
recently been compared in hydrotreating rapeseed oil [63]. At a
similar content of active sites (3.4 atoms/nm2 ), the NiMo/Al2 O3
was found to be much more active than the mono-metallic catalysts. The selectivity to hydrocarbons (at iso-conversion levels)
over the bimetallic NiMo catalyst was also signicantly higher as
compared to the Mo/Al2 O3 and Ni/Al2 O3 catalysts. The enhanced
performance of the NiMo catalyst was attributed to the synergy
between NiMo for the bimetallic catalysts. From a reaction network viewpoint, the Ni/Al2 O3 catalyst resulted in products from
the decarboxylation pathway whereas the NiMo/Al2 O3 catalyst
predominantly formed products by the hydrogenation pathway.
Nevertheless, the NiMo/Al2 O3 catalyst yielded signicant products
from both pathways. In a separate study, Yakovlev et al. found
that the bimetallic NiCu catalysts were superior than Ni cata-

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

lysts for the upgrading of bio-diesel [66]. Their study indicated


that the deoxygenation over Ni/CeO2 proceeded via CO bond
hydrogenolysis, while decarboxylation was the major route over
the NiCu/CeO2 catalyst. Synergy for bimetallic systems (rhenium
modied Pt/H-ZSM-5) has also been observed in a recent study
involving hydrotreating of jatropha oils [65].
Although much of the research undertaken thus far employed
conventional alumina as the active component support or carrier, some studies have also considered other supports [6569].
In a recent study, Kubicka and co-workers investigated rapeseed
oil deoxygenation over Co and Mo suldes supported on MCM41 with varying Si/Al ratios [70]. Although the incorporation of
Al into the MCM-41 framework enhanced the selectivity towards
hydrocarbons, the performance of the catalysts with MCM-41 supports was inferior to a catalyst using conventional alumina support.
CoMo catalysts supported on MCM-41 were also found to have
signicantly inferior performance to CoMo catalysts supported on
organized mesoporous alumina (OMA) [56]. From a deoxygenation mechanism viewpoint the CO hydrogenolysis (HDO) pathway
was more dominant for the CoMo/OMA catalysts as compared
to the CoMo/MCM-41 catalyst. Nava et al. recently investigated
the effect of mesoporous silicate supports (SBA-15, HMS, SBA16, DMS-1) on the hydroprocessing of an olive oil by-product
[71]. The selectivity for parafn production was found to decrease
in the following order: CoMo/SBA-15 > CoMo/HMS > CoMo/SBA16  CoMo/DMS-1. The SBA-15, SBA-16 and DMS-1 based catalysts
were found to be signicantly more effective in deoxygenation than
the HMS-supported catalyst.
2.2. Hydroprocessing in co-processing mode
From an economics viewpoint, co-processing triglyceride-based
feed-stocks with petroleum fractions in existing renery units is
a relatively low cost method to produce renewable fuels. While
some studies have considered co-processing of vegetable oils with
vacuum gas oil (VGO) boiling range petroleum, others have investigated co-processing with light gas oil (LGO) fractions.
In the studies involving blends of VGO and sunower oil over a
commercial sulded NiMo/Al2 O3 catalyst, Corma and co-workers
found that the kinetics of oxygen removal from sunower oil was
much faster than that of sulfur removal from VGO [49]. The conversion of sunower oil at 350 C was 100% while the corresponding
sulfur conversion was only 41%. Similar to stand-alone vegetable oil
processing studies [54] even in the sunower oil-VGO blend studies
the selectivity for decarboxylation products was found to increase
with increasing temperature and increasing sunower oil content.
The studies also revealed that sunower oil in the concentration
investigated (up to 50%) did not decrease the rate of desulfurization
over the sulded NiMo/Al2 O3 catalyst.
Additional agreement was provided in the studies by Knudsen
and co-workers that also suggested rapeseed oil (up to 25%) did not
inhibit sulfur removal from LGO over a sulded NiMo/Al2 O3 catalyst [72]. However in contrast to the VGO blend work described
earlier [49], the LGO-rapeseed oil studies revealed a decrease in
the selectivity of decarboxylation products when the rapeseed oil
content was increased from 15 to 25%. Although the total hydrogen consumption was highest for the 25% rapeseed oil feed, the
aromatic conversion was the highest for the LGO alone feed. The
decrease in aromatic conversion for rapeseed oil containing feeds
was attributed to the inhibition of the hydrogenation of mono-ring
aromatic compounds by CO formed from the rapeseed oil conversion. Due to the increase in parafnic content of the product, the API
gravity and cetane of the product increased with increasing rapeseed oil content. In general, it has been observed that an increase
in vegetable oil content is detrimental to the cold ow properties
of the diesel product. An increase in the rapeseed oil content in

the feed from 0% to 25% resulted in an increase in the cloud point


product from 4.4 C to 10.5 C [72]. In a separate study [73], the cold
lter plugging point was found to be lower than 10 C for a 10%
rapeseed oil content feed, while it was higher than 0 C for the 20%
rapeseed containing feed.
Along with conventional hydrotreating catalysts, some studies
have also considered the use of hydrocracking catalysts (bifunctional catalysts containing metals and acidic component) for
upgrading blends of vegetable oils and petroleum feed-stocks
[7476]. Light cycle oil (LCO) from the uidized catalytic cracking
(FCC) unit is a very poor quality diesel range material. Yunqi investigated the mild hydrocracking of soybean oil in LCO blend (030%)
over a NiMoP-HUSY/Al2 O3 catalyst at 370 C and 4 MPa pressure
[74]. The product cetane number was found to increase from 32.7
to 39.2, when the soybean oil content was increased from 0% to
30%. Although the corresponding sulfur conversion was not signicantly affected by the presence of additional triglycerides, the
nitrogen conversion dropped from 86.1% to 83.1%. Bezergianni et al.
investigated the hydroprocessing of sunower oil and VGO blends
over different hydrocracking (differing acidity) catalysts [75]. The
diesel to gasoline ratio in the product could be manipulated based
on the hydrocracking catalyst used. As expected, higher reaction
temperatures were found to be favorable for maximizing naphtha
production from the sunower oil-VGO blends.
2.3. Potential future research areas
Although previous studies on the hydroprocessing of
triglyceride-based feed-stocks have forwarded considerable
information, from a commercial viewpoint there are still some
research areas that would benet from more attention. The
conventional hydroprocessing of triglycerides, irrespective of
the mechanisms, dominantly results in the formation of linear
saturated hydrocarbons. While linear saturated hydrocarbons
(C15+ n-alkanes) are extremely attractive from the viewpoint of
desirable diesel combustion properties such as cetane, they are
also unfortunately responsible for inferior cold-ow properties.
In order to maximize diesel volume without signicantly sacricing the desirable properties, it would be benecial for the
research community to further increase their research focus on
optimizing catalytic isomerization or dewaxing of the long-chain
parafnic product formed from the triglyceride hydroprocessing
process [77]. The research in this area could be classied in the
following broad sections: (a) improved isomerization catalyst
and (b) process optimization. The desirable dewaxing catalysts
should have sufcient acidity to perform isomerization activity
while minimizing cracking activity for maintaining high distillate
yields. It would be benecial to have a series of isomerization
catalysts which have a range of activity for providing optimal
amount of cold-ow property upgrading (depending on the
regional cold-ow product property constraints). From a process
economics viewpoint, it would be most benecial to locate the
isomerization/dewaxing catalyst in the same reactor after the
hydroprocessing catalyst.
From an economics and environmental viewpoint, minimizing
unnecessary hydrogen consumption is important. The amount of
hydrogen consumed for triglyceride conversion by the HDO route
(breaking the CO linkage with no release of CO/CO2 ) is considerably higher than that by the decarboxylation route (breaking of a
CC bond with release of CO or CO2 ). However, the total hydrogen consumed by the decarboxylation route could theoretically
exceed the HDO route through secondary reactions such as water
gas shift and methanation [72]. Moreover, there is a loss in volume
of the desired product associated with the decarboxylation route.
The extent of hydrogen consumption via the different mechanisms
is expected to be sensitive to the process conditions and catalysts.

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

More studies need to be undertaken to develop further insights into


minimizing unnecessary hydrogen consumption.
To ensure optimal co-processing of triglycerides with petroleum
fractions in existing renery units, it is critical to develop
an enhanced understanding of the effect of triglycerides (feed
and conversion products) on the processing of the petroleum
fraction. Some of the studies described earlier have revealed
that there is no inhibition effect for the HDS reaction (VGO
and diesel) over a NiMo catalyst [49,72]. However, inhibition over the NiMo catalysts has been observed for aromatic
saturation [72] and nitrogen removal [74]. Studies related to
triglyceride co-processing over CoMo catalysts are relatively
scarce. Since CoMo catalysts inherently consume less hydrogen during LGO hydrotreating, a large number of ULSD units
worldwide use CoMo catalysts. It would therefore be benecial
to conduct focused studies to understand the effect of triglycerides on the processing of the petroleum fractions over CoMo
catalysts.
3. HDO of bio-oil feeds
Conversion of biomass to bio-oil feedstocks has been mainly
studied via two routes: (a) high pressure hydrothermal liquefaction
and (b) pyrolysis. The differences in physical properties between
bio-oil produced by both processes can vary signicantly due to the
type of biomass feed, conditions and reactor design (Table 2). The
catalytic HDO studies of the bio-oils obtained from the high pressure liquefaction and pyrolysis processes are discussed in separate
sections below.
3.1. HDO of bio-oils derived from high pressure liquefaction of
biomass
3.1.1. Inuence of nature of feed-stock
Researchers at Pacic Northwest National Laboratory (PNNL)
compared the processability of a bio-oil with that of a lighter oil
fraction obtained from the (same) bio-oil [78,79]. The bio-oil (TR7)
was produced from high pressure liquefaction of wood. Hydroprocessing studies were conducted using a continuous ow xed bed
catalyst bed system using an up-ow conguration. Compared to
the light oil fraction, the complete oil required more severe operating conditions such as a higher residence time (lower space
velocity) and operating pressure. Also, the complete oil consumed
a signicantly larger amount of hydrogen. This observation is similar to that for petroleum feed-stocks wherein lighter fractions
consume less hydrogen and are easier to hydrotreat (HDS, HDN)
compared to a crude oil.
In a separate study the same research group compared the processability of two bio-oils, which exhibited different properties due
to operational differences employed by the liquefaction process
[80]. GCMS characterization studies revealed the difference in the
composition of the two bio-oils. While cyclic ketones and single
ring phenolics were predominantly observed in bio-oil 1, a larger
concentration of multi-ring phenolics was observed in bio-oil 2.
Interestingly the two bio-oils also showed signicant differences in
their ability to be upgraded via the hydroprocessing process. Bio-oil
1 was a superior feed-stock in terms of ease of oxygen removal as
well as production of a desirable lighter hydrocarbon product. The
poor HDO performance was exacerbated even further with Biooil 2 due to a signicant alkali content, which was considered to
be the main reason for catalyst deactivation via a pore plugging
mechanism.
3.1.2. Inuence of catalyst
Conventional hydrotreating catalysts were found to have a superior HDO performance as compared to other catalysts investigated

such as copperchromite and supported Ni catalyst [79]. While the


copperchromite catalyst suffered from low activity, the Ni catalyst
produced more undesirable gaseous products, consumed considerably more hydrogen and was less stable. Also, the sulded form
of the CoMo hydrotreating catalyst considerably outperformed the
oxidic form of the same catalyst. This implies that if a hydrotreating
catalyst is utilized for completing the HDO of liquefaction derived
bio-oil, it must remain in a sulded form (using appropriate source
of H2 S) during the bio-oil upgrading process. A study involving
screening of different hydrotreating catalysts revealed that there
were no dramatic differences between the performances within
the same family of catalysts [80]. In general, all the hydrotreating
catalysts showed a similar trend for residual oxygen content and
hydrogen to carbon ratio with varying liquid hourly space velocities.
Studies have also been undertaken to investigate the effect
of hydrotreating catalyst pore size on the upgrading of liqueed
biomass [81]. Three catalysts were used in this study; two catalysts with large average pore diameters and one catalyst with
much lower pore diameter. Although the smaller pore size catalyst showed the best HDO performance, it was unable to maintain
its pore volume during the course of the run (6 h). Since the two
catalysts with larger pore diameters were better able to maintain
their pore volumes, it was speculated that the higher pore diameter
catalysts would show a better long term performance compared to
the narrow pore catalyst due to higher stability.
3.1.3. Inuence of operating conditions
Gevert et al. investigated solvent extraction as a process for
pretreating bio-oils (separating light oil from heavier fraction and
residual sodium salts) before hydroprocessing (HDO) [82,83]. The
extraction yields were found to be very sensitive to the solvent
varying from 3 to 74 vol% pentane to acetone, respectively. The solvent extracted bio-oil was subsequently investigated as feed-stock
for hydrotreating studies over a sulded CoMo/Al2 O3 catalyst in
a magnetically stirred autoclave system. Studies were undertaken
over a wide range of operating pressure (515 MPa) and a temperature range of 300390 C. The hydrotreating process resulted in
the conversion of some heavier oxygen-containing components of
the feed-stocks into lighter hydrocarbon components. The yields
of naphtha were small without a catalyst and were found to be
more sensitive to temperature than pressure. While the naphtha
and atmospheric gas oil yields increased gradually with temperature, coking was found to be minimal in the medium temperature
range.
Increased gasoline yields can be obtained by using a dual catalyst system: hydrotreating catalyst for decreasing the oxygen
content and a cracking component for enhancing heavy component
conversion. Baker et al. proposed a two step process modication to increase the aromatic gasoline yields while minimizing
hydrogen consumption [84]. Their process involved separation
of the lighter components from the products obtained after the
hydrotreating step and thereby only feeding the heavier components to the hydrocracking section. According to the authors the
proposed modication maximized aromatic gasoline production by
avoiding the unnecessary saturation/loss of aromatic compounds
(over the hydrocracking catalyst) of the gasoline fraction produced
in the hydrotreating step.
3.2. HDO of bio-oils derived from pyrolysis of biomass
Bio-oil production from the pyrolysis of biomass is considered
to have superior economics than that from the high pressure liquefaction process [85]. Consequently more studies have targeted the
upgrading of pyrolysis bio-oils. The properties of different pyrolysis oils are compared with high pressure liquefaction oils in Table 2

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

Table 2
Comparison of properties of pyrolysis oils with high pressure liquefaction oils [86].
Pyrolysis oils

Carbon (wt%)
Hydrogen (wt%)
Oxygen (wt%)
Moisture (wt%)
Density@55 C (g/ml)
Viscosity@65 C cps

High pressure liquefaction oils

Georgia tech

Waterloo

SERI

Laval

VTT peat

TR7

39.5
7.5
52.6
29.0
1.23
10

45.3
7.5
46.9
24.5
1.20
59

48.6
7.2
44.2
NA
1.23
NA

49.9
7.0
43.0
18.4
1.23
NA

51.0
7.8
40.3
NA
1.15
NA

74.8
8.0
16.6
3.5
1.10
3000

[86]. The pyrolysis oils have considerably higher oxygen content


than the high pressure liquefaction oils [80]. Correspondingly the
pyrolysis oils, due to their more polar nature, contain signicant
amounts of dissolved water (%). Although the pyrolysis oils have a
higher density, they have a much lower viscosity due to the high
water content and low molecular weight organics, such as formic
acid and acetic acid [87]. A comparison of properties between real
pyrolysis oils and model compounds is shown for reference in
Table 3.
As a result of the high oxygen content, and presence of some
highly reactive species (e.g., guaiacol, alkoxyphenols), the pyrolysis oils are signicantly more difcult to upgrade. The pyrolysis
oils have a high propensity for coke formation even under mild
hydrotreating conditions. An additional low-temperature, stabilization step is therefore recommended before upgrading pyrolysis
oils [86]. For the purpose of this review, the pyrolysis oil upgrading
work has been classied as follows: (a) stabilization studies dealing
with low temperature hydrotreating (b) dual stage studies dealing
with low temperature and high temperature hydrotreating in separate steps (c) combined dual stage studies that couple the low
temperature and high temperature steps in one system (product
from rst stage not collected).
3.2.1. Stabilization studies
Early studies by Elliott et al. revealed that hydrotreating of pyrolysis oils at temperatures above 350 C resulted in plugging of the
reactor system and catalyst encapsulation by coke-like material
[86]. Based on these studies the authors proposed a low temperature step for stabilization (to prevent rapid coke formation) of the
pyrolysis oils. Subsequently the authors demonstrated that the low
temperature stabilization step required a catalyst with hydrogenation activity, such as Pd, Pt, Ru or Ni [88]. Experiments conducted in
a continuous ow reactor with inert alumina at 260 C using hardwood derived pyrolysis oil showed very rapid coking followed by
reactor plugging. In contrast a stable operation was observed under
similar operating conditions with a sulded CoMo catalyst. ComTable 3
Comparison of properties between real and corresponding model HDO feeds derived
from bio-oil.
Real
Property
Density (kg/L)
HHV (MJ/kg)
KV (cSt)@50 C
Water (%)
C (wt%)
H (wt%)
N (wt%)
O (wt%)
S (wt%)
Cl (ppm)
Ash (ppm)
pH

Fast-py
1.111.30
1619
1080
1530
3249
6.98.6
00.4
4460
0.010.05
375
100200
2.03.7

1.11

0
68
6
0
36
0
0
0

Acetic acid
1.05
32.9
1.2
0
40
7
0
53
0
0
0
2.4 (1 M)

Guaiacol (C7 H8 O2 ), acetic acid (C2 H4 O2 ), phenol (C6 H6 O).

Phenol
1.07

0
77
6
0
16
0
0
0

72.6
8.0
16.3
5.0
1.09
17, 000

Table 4
Comparison of feed and product properties after low temperature hydrotreating of
wood-derived pyrolysis oil [89].
Property

Acetone insolubles
pH
Moisture (%)
Density (g/cm3 )@20 C
HHV (MJ/kg m.f.)
H/C ratio

Ensyn pyrolysis oil


Feed

Product

2.98
3.17
24.8
1.21
19.4
0.117

0.21
6.5
3.6
1.07
30.2
0.111

pared to the feed the stabilized pyrolysis oil product had a lower
oxygen content and higher viscosity [86]. The properties of the stabilized pyrolysis oil were found to be closer to that of the bio-oil
obtained from the high pressure liquefaction process.
Conti et al. investigated the low temperature hydrotreating of
wood-derived pyrolysis oil using a sulded NiMo catalyst in a continuous ow reactor system [89] (Table 4). The hydrogen stream
was fed directly into the reactor (no premixing with the liquid
feed) to avoid plugging of feed lines. Instead of operating under
isothermal conditions the reactor was operated using a temperature prole (140 C at the inlet and 280 C at the outlet) and H2
partial pressure of 15 MPa. The authors believed that the specic
temperature prole was important from the viewpoint of operation stability. The reactor was operated for 120 h and was shut
down voluntarily. While there was some initial decrease in catalyst
activity, the catalyst remained fairly stable after 60 h on stream. A
stabilized oil yield of 72 wt% with respect to the dry feed bio-oil was
obtained. The product properties are compared to the feed properties in Table 5. The upgrading resulted in a 60% reduction in
oxygen content and corresponding hydrogen uptake of 264 L/kg of
the feed.
Instead of using a sulded NiMo catalyst, Xu et al. investigated reduced NiMo based catalyst for mild upgrading of pine

Table 5
Single stage upgrading of pyrolysis oil over CoMo/Al2 O3 [86].
Pyrolysis oils

Model
Guaiacol

TR12

Processing conditions
Temperature-inlet ( C)
Temperature-outlet ( C)
Pressure (psig)
LHSV (h1 )
Product yields
Total oil (L/L feed oil)
Aqueous phase (L/L feed oil)
C5 -225 C frac. (L/L feed oil)
Carbon conversion to gas (wt%)
Carbon in aqueous phase (wt%)
Product properties
Oxygen (wt%)
H/C ratio (mol/mol)
Specic gravity (kg/L)

Laval

SERI

VTI peat

258
400
2020
0.13

259
376
1980
0.10

302
391
2000
0.19

0.42
0.57
0.37
35.5
1.0

0.37
0.51
0.27
25.0
0.6

0.44
0.34
0.36
30.0
1.0

0.8
1.7
0.83

1.3
1.68
0.85

1.5
1.8
0.86

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

sawdust-derived pyrolysis oil [90]. The hydrotreating reaction was


investigated in a batch reactor system under the following process conditions: temperature = 273 C, pressure = 3 MPa, reaction
time = 2 h. During the upgrading process the pH value of the bio-oil
increased from 2.16 to 2.84 indicating some conversion of organic
acids. The H content of the bio-oil increased from 6.61 wt% to
6.93 wt%. Aging studies (storage at 5 C for 8 months) of the raw and
upgraded bio-oils indicated a phase separation for the raw bio-oils,
which was not observed for the upgraded bio-oil.
Although conventional hydrotreating catalysts have been the
focus of most stabilization studies on bio-oil upgrading [78,91],
some studies have also considered the use of noble metal catalysts [9296]. Elliott et al. investigated the low temperature
hydrotreating of pyrolysis oils over Ru-based catalysts under the
following process conditions [96]: temperature = 180240 C; pressure = 1315 MPa; LHSV = 0.220.67 h1 . The pyrolysis oils were
derived from whitewood and bagasse (sugarcane). Under the process conditions investigated a deoxygenation of 3170% and a
product yield of 0.540.79 g/g feed (dry) was reported for whitewood oil while a deoxygenation of 3246% and product yield of
0.640.81 g/g feed (dry) was reported for bagasse oil. The product
oils were formed in two separate phases consisting of an aqueous
phase with more hydrophilic components and a dense tar phase
with more hydrophobic components. The two phase formation was
attributed to specic chemical transformations that made the product oil less hydrophilic and not from polymerization as suggested
previously. A large loss in catalyst activity was observed during the
course of the experiments. An important conclusion of the study
was that even low levels of contaminants (sulfur and iron) could
act as very strong poisons for the Ru catalyst. de Miguel Mercader
et al. [97] also investigated the HDO of pyrolysis oil (forest-residue
derived) over Ru-based catalysts in a batch reactor. A typical reaction proceeded for 4 h at a temperature range between 230 and
340 C, 30 MPa and 5 wt% catalyst. Above 300 C, three distinct
phases were observed; a heavy bottom oil/catalyst layer, a top-oil
product and an aqueous layer. The amount of top-oil micro-carbon
residue decreased from 4.7 wt% to 2.2 wt% over the temperature
range. Calculated oxygen content in the top-oil also reduced by
6 wt% under the same conditions suggesting an increase in HDO
activity. The overall dry product yields were insensitive to temperature and remained relatively constant at 50%. HDO conversion
based on oxygen removal occurred between 57 and 67%.
3.2.2. Dual stage studies
While the earlier section dealt with work that was primarily
focused on the stabilization stage, herein studies dealing with both
rst and the second stage will be considered. Such studies are
important from a process optimization viewpoint as the second
stage upgrading is strongly inuenced by the quality of the product
obtained after the stabilization (low temperature hydrotreating)
stage. Researchers at PNNL investigated the dual stage hydroprocessing of hardwood-derived pyrolysis oil in a continuous ow
reactor over a sulded CoMo catalyst [86]. The stabilization (stage
I) step was undertaken at 274 C and a space velocity of 0.62 h1 ,
whereas signicantly more severe operating conditions (temperature = 353 C and LHSV = 0.11 h1 ) were used for stage II. Stage I
process conditions resulted in a decrease in oxygen content (feed
as fed) from 52.6 wt% to 32.7 wt%. Under the high severity conditions (stage II) the oxygen content dropped further to 2.3 wt%. The
combined yield (stage I + stage II) for gasoline and oil products was
found to be 0.31 and 0.43 L/L feed, respectively and the combined
conversion of carbon to gas (C1 C4 ) was 17%. On a combined basis
about 8.8 wt% carbon was present in the aqueous phase.
Gagnon and Kaliaguine investigated a Ru-based catalyst for the
stabilization stage and a NiOWO3 /Al2 O3 catalyst for the second
stage [98]. The studies were carried out in a batch reactor with

focused efforts to optimize the rst stage with the second stage. The
rst stage studies spanned a temperature (80140 C) and pressure
range (410 MPa), whereas the second stage studies were undertaken at 350 C and 17 MPa. The dual stage studies showed that
the optimal rst stage operating conditions for the Ru-based catalyst were a temperature of 80 C and a pressure of 4 MPa. The
authors attributed this to a better control of the polymerization
reactions under the less severe rst stage operating conditions. The
information related to the extent of polymerization was obtained
by estimating the average molecular weight of the liquid samples
following the reaction. Based on the molecular weight data, the
authors further concluded that along with aldehyde hydrogenation and polymerization, hydrogenolysis reactions also occur over
the Ru catalyst.
Vispute and Huber also studied Ru-based catalysts for the low
temperature hydrogenation step, however they used a Pt-based
catalyst for producing alkanes in the second stage [99]. Studies were
conducted only on the aqueous phase of the oak wood-derived
pyrolysis oil, which was obtained by extraction with water. The
authors suggest that separation of the pyrolysis oil into two phases
prior to upgrading allows for better control on catalyst design and
optimization. The low temperature aqueous phase hydrogenation
step was conducted in a Parr batch reactor at a pressure of 6.89 MPa
and a temperature range of 25175 C. Although low temperatures
were found to be favorable for minimizing undesirable methane
formation, the higher temperatures favored faster hydrogenation
of the undesired components (e.g., sugars and levoglucosan). The
studies indicated the need for time and temperature optimization.
In a second step a bifunctional (4 wt% Pt/SiO2 Al2 O3 ) catalyst was
investigated for production of alkanes from the product formed
after the rst stage (temperature = 260 C and pressure = 5.17 MPa).
Alkane yields up to 48% of the theoretical yields were obtained in
this study in absence of externally added hydrogen.
Pd-based catalyst was investigated by Elliott et al. for stage
I hydrogenation of the different pyrolysis oils [100]. The studies were conducted in a continuous ow reactor at temperatures
ranging from 310 C to 375 C and space velocities ranging from
0.18 to 1.12 h1 . From a long term durability (e.g., stability, activity) viewpoint, the temperatures investigated in this study were
higher than the optimal temperatures reported previously. Due to
the higher temperatures used in this study, reactor plugging was
frequently observed. However, adequately stable operations were
maintained to obtain sufcient product for the next hydrocracking stage. Depending on the feed-stock the oil yields varied from
0.45 to 0.78 g/g dry feed. Higher oil yields were obtained from the
heavy phase and whole pyrolysis oil. An increase in temperature
decreased the oil yields and increased the gas yields, whereas the
oxygen conversion passed though a maximum at 340 C. Increase
in space velocity did not affect the oil yields but signicantly
decreased the oxygen conversion. The rst stage products were
biphasic in nature and the carbon content of the aqueous fraction
was found to correlate well with the oxygen content of the product
oil. Hydrocracking studies (stage II) were performed with the oil
phase products obtained from stage I using a sulded conventional
hydrocracking catalyst. The hydrocracking studies were conducted
at 405 C, 10 MPa and 0.2 LHSV and stable operations were observed
without any reactor plugging. Depending on the pyrolysis oil used
the oil yields ranged from 0.61 to 0.82 g/g/dry feed (shown in Fig. 5).
Interestingly, elemental analysis showed that the nal products had
a very similar composition irrespective of the source of bio-mass.
3.2.3. Combined dual stage studies
From an efciency viewpoint it is better to perform the dual
stage studies in a combined reactor system wherein the product
from the rst stage (low severity) is directly fed into the second
stage (high severity). Moreover it is difcult to pump the inter-

10

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

0.9
0.8

Oil yield (g/g) dry basis

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Mixed wood

Corn stover light Corn stover heavy


phase
phase

Corn stover 2

Poplar (hotfiltered)

Pyrolysis oil
Fig. 5. Product oil yields from different pyrolysis oils.
Adapted from [100].

mediate product (on account of its high viscosity) into the second
stage. In order to simplify the overall process, researchers at PNNL
investigated the pyrolysis oil hydroprocessing in a non-isothermal
continuous ow reactor such that reactor inlet was maintained at
250280 C while the reactor outlet was maintained at a higher
(370400 C) temperature [86]. The nal oxygen content was less
than 2 wt% and the yield of gasoline range material ranged from
0.27 to 0.37 L/L feed oil (Table 5). The product oil yields and quality were found to be similar to the separate dual stage. However, as
compared to the separate dual stage studies, the carbon conversion
to gas was higher, while the carbon lost in the aqueous phase was
much lower in the combined dual stage process. This was attributed
to the fact that the organics in the aqueous phase are converted to
gas in the combined stage studies. However, during the separate
stage studies, the aqueous phase was removed after stage I and not
processed through stage II. Recently Elliott and co-workers [100]
used non-isothermal processing conditions to study the combined
dual stage hydroprocessing over a Pd-based hydrotreating catalyst
(at lower temperature) and conventional hydrocracking catalyst
(at higher temperature). Studies over mixed wood derived pyrolysis oil provided a product oil yield of 0.5 g/g dry feed at a pressure of
14 MPa and space velocity of 0.15. The oxygen content of the product oil was less than 1 wt% and correspondingly the carbon content
of the aqueous fraction was less than 0.5 wt%. In agreement with
separate dual stage (hydrotreating/hydrocracking) studies over the
same catalysts [100], the chemical composition of the nal product
was found to be independent of the source biomass.
Combined dual stage low severity hydrotreating has also been
investigated for upgrading pyrolysis oils. The study involved a continuous ow system with two reactors connected in series. Using
a space velocity of 0.28 h1 , rst stage temperature of 148 C, second stage temperature of 355 C, a product yield of 0.53 g/g feed
was obtained over a sulded NiMo catalyst [101]. The corresponding deoxygenation was 95.8% and carbon conversion to gas was
29%. Use of a weak hydrogenation catalyst in the rst stage (spinel
supported CoMo) resulted in exothermic pyrolysis condensation
reactions which in turn caused rapid coke build up and plugging
of efuent lines. Upgrading of pyrolysis oil obtained from poplar
was found to be signicantly more efcient than that obtained
from eucalyptus. Even though eucalyptus-derived pyrolysis oil was

processed under more severe conditions, the product yields and


deoxygenation was lower compared to the poplar-derived pyrolysis oil. This was a surprising result as previous studies by the same
authors had not indicated major differences based on the source of
pyrolysis oils. Another important conclusion from this study was
that the oxygen content of product oil had to be dropped below
10 wt% to facilitate easy separation of water indicating the presence
of organic acids and highly polar intermediate species after the rst
stage. Thus, it is evident that hydrogenation over hydrogenolysis is
critical in the rst stage.
3.3. Potential future research areas
Since the bio-oils for the upgrading studies are not easily available there are several constraints for undertaking these studies.
However, to make adequate progress towards commercialization
it is important that a number of additional studies are undertaken
in this area. Previous studies have provided some interesting information about the upgrading of these bio-oils. However, a more
systematic study is desirable in the following areas (a) effect of
bio-oil quality (b) effect of process parameters and (c) optimal catalyst. Identication of an optimal catalyst system with adequate
activity and stability is critical. Although poor catalyst stability has
been identied as a major problem, very few studies have actually
addressed the problem. Due to the poor quality of the bio-oils the
conventional hydrotreating catalysts are expected to have a considerably lower catalyst life in bio-oil upgrading operations than that
observed with petroleum feed-stocks. While the current generation
commercial catalysts are excellent hydroprocessing catalysts they
are optimized for petroleum feedstocks. Since the bio-oils have signicantly different properties than petroleum feed-stocks, it would
be worthwhile to dedicate efforts towards developing catalysts
specically for upgrading bio-oils. From a wide spread commercial
applications viewpoint an ideal bio-oil upgrading catalyst should
have the attributes described below.
a) High activity for deoxygenation: this is important from the viewpoint of minimizing the reactor size and obtaining desired yields.
b) Ability to withstand large quantities of coke and/or minimize
coke formation: bio-oils have a much higher propensity to form

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

c)

d)
e)

f)

coke than typical petroleum fractions. In order to elongate the


time between regeneration cycles and to achieve stable operations, the catalyst should be able to hold large quantities of coke
and/or minimize coke formation.
High tolerance to water: previous studies have indicated that
presence of water can have a detrimental effect on the deoxygenation catalyst performance [102,103]. Since the bio-oil
upgrading catalyst is expected to be exposed to large amounts
of water, it is important for the catalyst to exhibit high tolerance
for water. Based on the work done by Resasco and co-workers
[104], it may also be worthwhile to consider catalyst systems
that can stabilize water-oil emulsions and catalyze reactions at
the oilwater interface.
Ability to regenerate using a simple process (e.g., hot air burn):
this will simplify the process and minimize capital expenditure.
High tolerance to poisons: the activity lost due to coke deposition related deactivation can be regained during hot air burn
regeneration; however activity loss from poisons is typically irreversible. High tolerance for poisons is critical from the viewpoint
of increasing the overall life of the catalyst load.
Availability should not be an issue: if this process is commercially practiced on a wide-scale, a huge amount of catalyst will
be required. This will preclude the use of materials/components
that are scarce.

The above list can be used to narrow down the possible materials and approaches that could be potentially considered to develop
superior bio oil upgrading catalysts. Catalyst development efforts
are expected to be beneted by using a rationalized design of catalyst approach. However in order to use this approach it is important
to develop a realistic understanding of the bio-oil catalyst systems.
Development of a realistic understanding will require the following: (i) analytical advances for rapid and quantitative identication
of oxygenated compounds in complex mixtures and (ii) systematic
studies that are designed to obtain comprehensive understanding of interaction between representative molecules present in
bio-oils (as individual compounds and mixtures) and different catalyst systems (active metals, supports etc.) and (iii) understanding
the catalyst deactivation mechanisms for bio-oils. Since it seems
unlikely that it would be possible to use the relatively simple reactors used in conventional hydroprocessing of petroleum fractions,
it is also worthwhile to identify/develop appropriate reactor technology to optimize the reaction-regeneration process.
4. Life cycle assessment (LCA)
The LCA on an energy and Green House Gases (GHGs) basis for
renewable fuels is required to ascertain the degree of GHG reduction and thus its categorization as a renewable fuel. The LCA is also
an essential part of renewable fuel and low carbon regulations like
the Energy Independence and Security Act of 2007 (EISA 2007), the
California Low Carbon Fuels Standard (CA LCFS), and the European
Unions Renewable Energy Directive (RED). A number of papers have
addressed the LCA of renewable fuels for HDO, most of these studies deal with production of renewable diesel [105108]. Elliot et al.
[109] have discussed the process efciency for two liquefaction
processes for cellulosic biomass: atmospheric fast pyrolysis, and
liquefaction in pressurized solvents. The LCA results are sensitive to
the assumptions made in the models and the allocation techniques
used [107]. Thus there is a need to develop LCA standards for clearly
dening system boundaries, setting up feedstock life cycle inventories, and allocation methods. The renewable diesel LCA studies
indicate that GHG reductions of around 5090% can be realized
compared to petroleum diesel [105,108]. Tallow being a more saturated lipid feedstock can produce a GHG reduction of about 88%

11

[105]. For cellulose liquefaction processes atmospheric fast pyrolysis can provide process energy efciencies of around 5060%, while
pressurized liquefaction processes can be 5055% efcient [109].
5. Concluding remarks
While some interesting information has been forwarded by previous HDO studies, several gaps remain in this very important
research area. These gaps can be divided into two types (optimization and technology gaps) depending on the feed-stock used
for producing renewable fuels. The gap in case of HDO of triglycerides is an optimization gap and can be eliminated via some
additional research on dewaxing catalysts, process optimization
and developing a better understanding of inhibition effects. The
technology gap in case of HDO of biomass derived bio-oils, on
the other hand, is exceedingly challenging due to the poor quality of the bio-oils (high oxygen content, impurity levels, molecular
complexity and coking propensity). From a technical viewpoint
extensive research on nature of feeds, catalyst systems as well as
reactor technology is needed in this area. The challenge of commercial implementation of renewable fuels production is not just
limited to technical issues. The non-technical issues (such as infrastructure, logistics and economics) are expected to be at-least as
challenging as the technical aspects of renewable fuel production.
Although there are several uncertainties associated with renewable fuels, one thing is certain the economic and environmental
importance coupled with the enormous challenges of this topic,
will continue to maintain this as one of the most vibrant areas of
future research.
References
[1] Energy Independence and Security Act of 2007, 110th Congress of the United
States, Public Law 110-140.
[2] S.N. Naik, P.K. Rout, A.K. Dalai, Renew. Sustain. Energy Rev. 14 (2010) 578597.
[3] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 24112502.
[4] G. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 40444098.
[5] B. Kamm, P.R. Gruber, M. Kamm (Eds.), Bioreneries Industrial Processes
and Products, Wiley/VCH, Weinheim, 2006.
[6] G. Centi, R.A. van Santen (Eds.), Catalysis for Renewables: From Feedstock to
Energy Production, Wiley/VCH, Weinheim, 2007.
[7] T. Marker, J. Petri, T. Kalnes, M. McCall, D. Mackowiak, B. Jerosky, B. Reagan,
L. Nemeth, M. Krawczyk, S. Czernik, D. Elliott, and D. Shonnard, in: DOE (Ed.),
DEFG36-05GO15085, 2005.
[8] R.G. Leliveld, S.E. Eijsbouts, Catal. Today 130 (2008) 183189.
[9] D.C. Elliott, D. Beckman, A.V. Bridgwater, J.P. Diebold, S.B. Gevert, Y. Solantausta, Energy Fuels 5 (1991) 399.
[10] A.V. Bridgwater, G.V.C. Peacocke, Renew. Sustain. Energy Rev. 4 (2000) 1.
[11] P.L. Thigpen, W.L. Berry, in: D.L. Klass (Ed.), Energy from Biomass and Wastes
VI, IGT, Chicago, 1982, p. 1057.
[12] A.V. Bridgwater, G.V.C. Peacocke, Renew. Sustain. Energy Rev. 4 (2000)
173.
[13] L.C. Meher, D.V. Sagar, S.N. Naik, Renew. Sustain. Energy Rev. 10 (2006)
248268.
[14] T. Kalnes, T. Marker, D.R. Shonnard, Int. J. Chem. Reactor Eng. 5 (2007).
[15] M. Kuronen, S. Mikkonen, P. Aakko, T. Murtonen, SAE Int. (2007), 2007-014031.
[16] L. Rantanen, R. Linnaila, P. Aakko, T. Harju, SAE Int. (2005), 2005-01-3771.
[17] J. Holmgren, C. Gosling, R. Marinangeli, T. Marker, G. Faraci, C. Perego, Hydrocarbon Process. (2007) 6772.
[18] E. Furimsky, Appl. Catal. A: Gen. 199 (2000) 147190.
[19] A. Pattiya, J.O. Titiloyse, A.V. Bridgwatger, Fuel 89 (2010) 244253.
[20] R.R. Davda, G.W. Huber, R.D. Cortright, J.A. Dumesic, Appl. Catal. B: Environ.
56 (2005) 171186.
[21] O.I. Senol, E.-M. Ryymin, A.O.I. Krause, J. Mol. Catal. A: Chem. 277 (2007)
107112.
[22] O.I. Senol, T.-R. Viljava, A.O.I. Krause, Appl. Catal. A: Gen. 326 (2007) 236244.
[23] O.I. Senol, E.-M. Ryymin, T.-R. Viljava, A.O.I. Krause, J. Mol. Catal. A: Chem. 268
(2007) 18.
[24] O.I. Senol, T.-R. Viljava, A.O.I. Krause, Catal. Today 106 (2005) 186189.
[25] O.I. Senol, T.-R. Viljava, A.O.I. Krause, Catal. Today 100 (2005) 331335.
[26] T.-R. Viljava, E.R.M. Saari, A.O.I. Krause, Appl. Catal. A: Gen. 209 (2001)
3343.
[27] T.-R. Viljava, R.S. Komulainen, A.O.I. Krause, Catal. Today 60 (2000) 8392.
[28] M.J. Girgis, B.C. Gates, Ind. Chem. Eng. Res. 30 (1991) 2021.

12

T.V. Choudhary, C.B. Phillips / Applied Catalysis A: General 397 (2011) 112

[29] B. Andrea, D. Resasco, S. Crossley, W. Alvarez, T.V. Choudhary, Catal. Lett. 123
(2008) 181.
[30] T.V. Choudhary, Ind. Chem. Eng. Res. 46 (2007) 8363.
[31] T.V. Choudhary, J. Malandra, J. Green, S. Parrott, B. Johnson, Angew. Chem. Int.
Ed. 45 (2006) 3299.
[32] T.V. Choudhary, S. Parrott, B. Johnson, Environ. Sci. Technol. 42 (2008)
1944.
[33] T.V. Choudhary, S. Parrott, B. Johnson, Catal. Commun. 9 (2008) 1853.
[34] G.W. Huber, N. Chheda, C.J. Barrett, J.A. Dumesic, Science 308 (2005)
1446.
[35] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007) 7164.
[36] M. Stumborg, A. Wong, E. Hogan, Bioresour. Technol. 56 (1996) 13.
[37] I. Sebos, A. Matsoukas, V. Apostolpoulos, N. Papayannakas, Fuel 88 (2009) 145.
[38] S. Bezergianni, S. Voutetakis, A. Kalogianni, Ind. Eng. Chem. Res. 48 (2009)
8402.
[39] J.W. Veldsink, M.J. Bouma, N.-H. Schoon, A.A.C.M. Beenackers, Catal. Rev. Sci.
Eng. 39 (1997) 253.
[40] D. Singh, M.E. Rezac, P.H. Pfromm, J. Membr. Sci. 348 (2010) 99.
[41] W.K. Craig, D.W. Soveran, US patent 4992605 (1991).
[42] J. Yao, E.L. Sughrue, J.B. Cross, J.B. Kimble, H. Hsing, M.M. Johnson, US Patent
7550634 (2009).
[43] D.B. Ghonasgi, E.L. Sughrue, J. Yao, US Patent 7626063 (2009).
[44] J.H. Benedict, B.F. Daubert, J. Am. Chem. Soc. 72 (1950) 43564359.
[45] E. Vonghia, D.G.B. Boocock, S. Konar, A. Leung, Energy Fuels 9 (1995)
10901096.
[46] A. Leung, D.G.B. Boocock, S.K. Konar, Energy Fuels 9 (1995) 913920.
[47] M. Snare, I. Kubickova, P. Maki-Arvela, D. Chichova, D.Y. Eranen, D.Y. Murzin,
Fuel 87 (2008) 933945.
[48] M. Krar, S. Kovacs, J. Hancsok, Bioresour. Technol. 101 (2010) 9287.
[49] G.W. Huber, A. Corma, Appl. Catal. A: Gen. 329 (2007) 120129.
[50] I. Kubickova, L. Kaluza, Appl. Catal. A: Gen. 372 (2010) 199208.
[51] V.A. Yakovlev, O.V. Sherstyuk, V.O. Dundich, D.Yu. Ermakov, V.M.
Novopashina, M.Yu. Lebedev, O. Bulavchenko a, V.N. Parmon, Catal. Today
144 (2009) 362366.
[52] E. Vonghia, D.G.B. Boocock, S.K. Konar, A. Leung, Energy Fuels 9 (1995)
1090.
[53] G.N. da Rocha Filho, D. Brodzki, G. Djega-Mariadassou, Fuel 72 (1993)
543.
[54] P. Simacek, D. Kubicka, G. Sebor, M. Pospisil, Fuel 89 (2010) 611.
[55] P. Simacek, D. Kubicka, G. Sebor, M. Pospisil, Fuel 88 (2009) 456.
[56] D. Kubicka, P. Simacek, N. Zilkova, Top. Catal. 52 (2009) 161.
[57] J. Gusmao, D. Brodzki, G. Djega-Mariadassou, R. Frety, Catal. Today 5 (1989)
533.
[58] A. Guzman, L.P. Prada, M.L. Nunez, ACS Fall Meeting Preprints, 2009.
[59] L. Boda, G. Onyestyak, H. Solt, F. Lonyi, J. Valyon, A. Thernesz, Appl. Catal. A:
Gen. 374 (2010) 158169.
[60] J. Cejka, Appl. Catal. A: Gen. 254 (2003) 327338.
[61] T.J. Benson, R. Hernandez, W.T. French, E.G. Alley, W.E. Holmes, J. Mol. Catal.
(2009).
[62] J.A. van Bokhoven, D.C. Koningsberger, P. Kunkeler, H. van Bekkum, J. Catal.
211 (2002) 540547.
[63] D Kubicka, L. Kaluza, Appl. Catal. A: Gen. 372 (2010) 199.
[64] J. Monnier, H. Sulimma, A. Dalai, G. Caravaggio, Appl. Catal. A: Gen. 382 (2010)
176.
[65] K. Murata, Y. Liu, M. Inaba, I. Takahara, Energy Fuels 24 (2010) 2404.
[66] V.A. Yakovlev, S.A. Khromova, O.V. Sherstyuk, V.O. Dundich, D.Yu. Ermakov,
V.M. Novopashina, M.Yu. Lebedev, O. Bulavchenko, V.N. Parmon, Catal. Today
144 (2009) 362.
[67] J.G. Immer, M.J. Kelly, H.H. Lamb, Appl. Catal. A: Gen. 375 (2010) 134139.
[68] K.D. Maher, K.M. Kirkwood, M.R. Gray, D.C. Bressler, Ind. Eng. Chem. Res. 47
(2008) 53285336.
[69] P. Maki-Arvela, I. Kubickova, M. Snare, K. Eranen, D.Y. Murzin, Energy Fuels
21 (2007) 3041.
[70] D. Kubicka, M. Bejblova, J. Vlk, Top. Catal. 53 (2010) 168.
[71] R. Nava, B. Pawelec, P. Castano, M.C. Alvarez-Galvan, C.V. Loricera, J.L.G. Fierro,
Appl. Catal. B: Environ. 92 (2009) 154.

[72] B. Donnis, R.G. Egeberg, P. Blom, K.G. Knudsen, Top. Catal. 52 (2009)
229.
[73] J. Walendziewski, M. Stolarski, R. Luzny, B. Klimek, Fuel Process. Technol. 90
(2009) 686.
[74] L. Yunqui, ACS Fall Meeting Preprints, 2009.
[75] S. Bezergianni, A. Kalogianni, I.A. Vasalos, Bioresour. Technol. 100 (2009) 3036.
[76] A.A. Lappas, S. Bezergianni, I.A. Vasalos, Catal. Today 145 (2009) 55.
[77] D.E. Resasco, S. Crossley, AICHE J. 55 (2009) 1082.
[78] D.C. Elliott, Energy Fuels 21 (2007) 1792.
[79] D.C. Elliott, E.G. Baker, Biotechnol. Bioeng. Symp. 14 (1984) 159.
[80] E.G. Baker, D.C. Elliott, Pyrolysis Oils from Biomass Producing, Analyzing
and Upgrading, in: E.J. Soltes, T.A. Milne (Eds.), ACS Symposium Series, vol.
376, 1988, p. 228, Washington, DC.
[81] S.B. Gevert, P.B.W. Andersson, S.P. Sandqvist, S.G. Jaras, M.T. Tokarz, Energy
Fuels 4 (1990) 78.
[82] S.B. Gevert, J.E. Otterstedt, Energy from Biomass and Wastes X, IGT, Chicago,
1986, p. 105.
[83] S.B. Gevert, J.E. Otterstedt, Biomass 13 (1987) 105.
[84] E.G. Baker, D.C. Elliott, US Patent 5180868.
[85] D.C. Elliott, E.G. Baker, D. Beckman, Y. Solantausta, V. Tolenhiemo, S.B. Gevert,
C. Hornell, A. Ostman, B. Kjellstrom, Biomass 22 (1990) 251.
[86] E.G. Baker, D.C. Elliott, in: A.V. Bridgwater, J.L. Kuester (Eds.), Research in Thermochemical Biomass Conversion, Elseiver Appl. Sci., Barking, UK, 1988, p.
883.
[87] G. Huber (Ed.), B. NSF Chemical, Environmental and Transport Systems Division, 2008.
[88] D.C. Elliott, E.G. Baker, US Patent 4795841.
[89] L. Conti, G. Scano, J. Boufala, S. Mascia, in: A.V. Bridgwater, D.G.B. Boocock
(Eds.), Developments in Thermochemical Biomass Conversion, vol. 1, Chapman & Hall, London, UK, 1997, p. 622.
[90] Y. Xu, T. Wang, L. Ma, Q. Zhang, L. Wang, Biomass Bioenergy 33 (2009) 1030.
[91] R.J. French, J. Hrdlicka, R. Baldwin, Environ. Prog. Sustain. Energy 29 (2010)
142.
[92] C.A. Fisk, T. Morgan, Y. Ji, M. Crocker, C. Crofcheck, S.A. Lewis, Appl. Catal. A:
Gen. 358 (2009) 150.
[93] P. De Wild, R. Van der Laan, A. Kloekhorst, E. Heeres, Environ. Prog. Sustain.
Energy 28 (2009) 461.
[94] J. Wildschut, I. Melian-Cabrera, H.J. Heeres, Appl. Catal. B: Environ. 99 (2010)
298.
[95] J. Wildschut, M. Iqbal, F.H. Mahfud, I. Melian-Cabrera, R.H. Venderbosch, H.J.
Heeres, Energy Environ. Sci. 3 (2010) 962.
[96] D.C. Elliott, G.G. Neuenschwander, T.R. Hart, J. Hu, A.E. Solana, C. Cao, in: A.V.
Bridgwater, D.G.B. Boocock (Eds.), Science in Thermal and Chemical Biomass
Conversion, CPL Press, Newbury berks, UK, 2006, p. 1536.
[97] F. de Miguel Mercadera, S.R.A. Kerstena, N.W.J. Wayb, C.J. Schaverienb, J.A.
Hogendoorna, Appl. Catal. B Environ. (2010).
[98] J. Gagnon, S. Kaliaguine, Ind. Eng. Chem. Res. 27 (1988) 1783.
[99] G.W. Huber, T.P. Vispute, Green Chem. 11 (2009) 1433.
[100] D.C. Elliott, T.R. Hart, G.G. Neuenschwander, L.J. Rotness, A.H. Zacher, Environ.
Prog. Sustain. Energy 28 (2009) 441.
[101] D.C. Elliott, G.G. Neuenschwander, in: A.V. Bridgwater, D.G.B. Boocock (Eds.),
Developments in Thermochemical Biomass Conversion, vol. 1, Chapman &
Hall, London, UK, 1997, p. 611.
[102] L. Laurent, B. Delmon, J. Catal. 146 (1994) 281.
[103] O.I. Senol, T.-R. Vijaya, A.O.I. Krause, Catal. Today 106 (2005) 186.
[104] S. Crossley, J. Faria, M. Shen, D.E. Resasco, Science 327 (2010) 68.
[105] H. Jin, R. Nielsen, The 2008 Annual AIChE Meeting, November 19th, Philadelphia, PA, 2008.
[106] T.N. Kalnes, K.P. Koers, T. Marker, D.R. Shonnard, Environ. Prog. Sustain. Energy
28 (2009) 111120, No. 1.
[107] H. Huo, M. Wang, C. Bloyd, V. Putsche, Environ. Sci. Technol. 43 (2009)
750756.
[108] T. Kalnes, T. Markery, D.R. Shonnard, Int. J. Chem. Reactor Eng. 5 (2007), Article
A48.
[109] D.C. Elliott, E.G. Baker, D. Beckman, Y. Solantausta, V. Tolenhiemo, S.B. Gevert,
C. Hrnell, A. stmanf, B. Kjellstrm, Biomass 22 (1990) 251269.

Das könnte Ihnen auch gefallen