Direct liquefaction of paulownia in hot compressed water: Inuence of catalysts

Peiqin Sun, Mingxing Heng, Shaohui Sun

*
, Junwu Chen
School of Chemical and Energy Engineering, Zhengzhou University, No.100 of Science Road, 450001 Zhengzhou Henan, China
a r t i c l e i n f o
Article history:
Received 23 January 2010
Received in revised form
2 July 2010
Accepted 7 July 2010
Available online 17 August 2010
Keywords:
Biomass
Direct liquefaction
Paulownia
Hot compressed water
Iron
a b s t r a c t
The liquefaction of paulownia in hot compressed water with and without the presence of catalysts such
as Fe and Na
2
CO
3
was studied to investigate the catalysts effects on the biomass liquefaction process.
Experiments were conducted in a 1000 ml stainless steel autoclave in the conditions of temperature
range of 280e360

C for 10 min under vacuum, with 60 g biomass and 360 g water, respectively. All the
catalysts tested, i.e., Na
2
CO
3
and Fe, were found effective for enhancing the formation of heavy oil
products, while they signicantly promoted the formation of gas. The experimental results showed that
the maximum heavy oil yield rate was 36.34% with Fe catalyst and the minimum solid residue yield was
obtained when Na
2
CO
3
was employed as catalyst. The heavy oils from with and without catalyst were
characterized by GCeMS and FTIR. The results showed that the oils mainly contained phenol derivatives,
ketones, carboxylic acid/ester, benzene derivatives and long-chain alkanes, aldehydes and its derivatives
and that employing different catalysts could change the distribution and relative abundance of produced
compounds.
2010 Elsevier Ltd. All rights reserved.
1. Introduction
Nowadays, biomass is drawing increasing attention as a renew-
able energy source due to the advantages that it is renewable and
abundant. In addition, among all the renewable sources, biomass
derived energy sources can be (and are already) converted to liquid
fuels and valuable chemicals (e.g., derived crops, sugar cane.) so
far [1,2]. Since biomass can x CO
2
in the atmosphere through
photosynthesis, the CO
2
emitted by energy using biomass can be
offset by the CO
2
xed by photosynthesis; resulting in a net zero
contribution of carbon to the atmosphere. In addition, woody
biomass has less content of nitrogen and sulfur, therefore, its
burning causes less NO
x
and SO
x
emission [3] and it is much more
environmentally friendly compared with fossil fuels. Hydrocarbons
for transport fuel can be made from biomass liquefaction tech-
nology and subsequent upgrading process including hydro-
deoxygenation, catalytic cracking of bio-oil, emulsication, steam
reforming, chemicals extracted fromthe bio-oils and phenolic resin
synthesis et al. (e.g., hydrodeoxygenation, catalytic cracking of bio-
oil, emulsication, steam reforming, chemicals extracted from the
bio-oils and phenolic resin synthesis) [4e7]. Biomass liquefaction
technologies mainly include thermal conversion (include lique-
faction, pyrolysis and indirect liquefaction) and biochemical,
among which the direct liquefaction process in hot compressed
water has attracted much more attention due to its versatility to
utilize mixed biomass feedstock without any pretreatment or
drying. Not only is water environmental benign solvent, cheap, not
polluting the environment, but also hot compressed water has
a more enhanced solubility for organic compounds than ambient
liquid water as its lower dielectric constant [8]. More importantly,
employing water as a reactant and as the reaction medium in
a biomass conversion process will eliminate the costly drying
process of wet feedstocks, which make it a promising reaction
medium for biomass liquefaction.
Extensive research work has been conducted on direct lique-
faction of biomass in hot compressed water. Appell et al. [9] carried
out that a variety of lignocellulosic materials were efciently con-
verted to oily products in water at around 350

C in the presence of
CO and Na
2
CO
3
as catalyst. Minowa et al. obtained heavy oil at
a yield of 21e36 wt% from a variety of biomass feedstocks in water
at 300

C and around 10 MPa with Na
2
CO
3
as catalyst. Qu et al. [10]
obtained heavy oil at a yield of 7.38e23.78% by direct liquefaction of
Cunninghamia lanceolata in water at 280e360

C for 10e30 min
under vacuum. Mazaheri et al. [11] carried out the decomposition
of oil palm fruit press ber in subcritical water with and without
the presence of catalyst and found that the solid conversion
increased with the increasing of Na
2
CO
3
and NaOH. Xu and Lad [12]
reported that woody biomass direct liquefaction in a sub/near-
critical water with and without catalysts at 280e380

C and the
yield of heavy oil in the operation at 300

C for 30 min were
* Corresponding author. Tel.: 86 371 67781754; fax: 86 371 67781755.
E-mail address: hengjim@hotmail.com (S. Sun).
Contents lists available at ScienceDirect
Energy
j ournal homepage: www. el sevi er. com/ l ocat e/ energy
0360-5442/$ e see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2010.07.005
Energy 35 (2010) 5421e5429
improved signicantly from around 30% without catalyst to more
than 45% by Ba(OH)
2
. Bestue et al. [13] reported that liquefaction of
poplar wood in water at 340

C for 30 min with several iron addi-
tives as catalyst and revealed that iron powder gave good oil yields
of w40%. Zhang et al. [14] studied that co-liquefaction of secondary
pulp/paper-mill sludge and waste newspaper in sub/near-critical
water with and without catalysts (such as HCO
2
H, FeS, KOH) at
250e380

C and the yield of heavy oil was found to improve
signicantly at 300

C with catalysts.
Nowadays, various forestry and agricultural residues were liq-
ueed to fuel or chemicals in hot compressed water, and paulownia
was the only biomass that was investigated in this paper. To the
best of our knowledge, a systematic experimental study on the
paulownia wood in a wide range of reaction temperatures with and
without catalyst had not been performed so far. Paulownia was
native to China, which growed on marginal land or intercrop. It was
fast growing, widely distributed, low requirement of chemical
inputs material and the use of marginal land did not compete with
the food production. As paulownia was a very important renewable
resource in China, therefore it was worthwhile to further investi-
gate its applications.
The aim of this work was to study the inuence of the catalysts
such as iron powder and sodium carbonate at 280e360

C in
a batch reactor by direct liquefaction of paulownia in hot
compressed water. The emphasis was put on the formation of
heavy oil which can be easily separated (by decanting) from the
liquefaction products and contained much greater heating values
than WSOs (water-soluble oils), as evidenced by later and previous
work [12]. The extensive GCeMS analysis of the heavy oil to
observe the formation of key compounds from with and without
catalyst was also reported.
2. Materials and methods
2.1. Materials
The raw material was colleted in Zhengzhou, Henan province in
China. The decorticated samples were soaked in water to remove
adhering dirt and air-dried in ambient atmosphere for 72 h. Then
they were milled and screened to fractions with particle diameters
smaller than 0.30 mm. Only the fractions smaller than 0.20 mm
were selected in this work. The powders were dried in an oven at
105

C for 24 h so as to completely remove the inner water and then

kept in a desiccator at room temperature before use. The chemical
and elemental compositions of paulownia were analyzed. The
chemical composition was measured according to the TAPPI test
methods whereas elements analysis (e.g., CHNS/O) was determined
by a Flash EA 1112 CHNS/O analyzer (USA Thermo Electron). The
paulownia consisted of 42.35 wt% cellulose, 25.22 wt% hemi-
cllulose, 23.44 wt% lignin, 8.19 wt% extractives. The elemental
analysis for paulownia was 45.52 wt% carbon, 6.31 wt% hydrogen,
48.17 wt% oxygen.
The solvent used in this work was self-made de-ionized water
while the analytical grade acetone, sodium carbonate (Na
2
CO
3
) and
KBr were purchased from Tianjin Kemiou Chemical Reagent Ltd.
and used as received. The analytical grade iron powder (Fe)
(150 mm) was purchased from Tianjin Kaitong Chemical Reagent
Ltd. and used as received.
2.2. Liquefaction and separation
The reaction was carried out in a 1000 ml stainless steel auto-
clave with a magnetic stirrer. The autoclave was heated by an
external electrical furnace, and the temperature was measured by
a thermocouple and controlled within 3

C. In a typical lique-
faction experiment, 60 g (dry basis) of raw material powder and
360 ml de-ionized water were fed into the autoclave with catalyst
(if needed, with 6 g iron powder (10%) or 3 g sodium carbonate
(5%)). The autoclave was vaccumized and then sealed. Reactants
were agitated by magnetic stirring at 300 rpm. With heating rate of
5

C/min, reactions proceeded at specied temperature ranging
from 280 to 360

Cfor a desired residence time (varying from 5 min

up to 50 min). Heating was then stopped and the autoclave
temperature was dropped to room temperature by an internal
water cooling system.
After each run, once the reactor was cooled to room temper-
ature, the gas phase was covered in a gas-sampling bag and the
total volume was measured. The solid/liquid products were
carefully rinsed thoroughly with weighed accurately de-ionized
water and poured into a beaker, then weighed. The liquid phase
was ltered under reduced pressure through a pre-weighed lter
paper to separate the water-insoluble fraction. The ltered water
solution was evaporated at 60

C under reduced pressure in
a rotary evaporator, and the residues obtained were the so-called
WSO (water-soluble oil). The autoclave was then completely
rinsed three times with acetone, and the acetone solution was
collected in a beaker, into which the previous water-insoluble
fraction was added. In addition, the mixture in the beaker was
kept for 12 h, the two fractions were successfully ltered, and the
residue was rinsed with acetone until the elution acetone solvent
became colorless. The acetone solution was dried at 30

C under
reduced pressure in a rotary evaporator, and the residue was
designated as heavy oil (HO). The acetone-insoluble fraction
was eluted with acid to remove the catalyst, then the contents
were separated by ltration and the residue was washed with de-
ionized water, then dried at 105

C in oven for 24 h to get the
solid residue (SR). The separation method was performed in
a similar way described by Minowa et al. [3], Zhong and Wei [15],
Xu and Lancaster [16]. The yields of products, all expressed with
wt%, were calculated relative to the dry biomass feedstock. The
separation of liquefaction products can be found elsewhere
[3,15,16]. Two or three duplicate runs were performed for each
feedstock to ensure the repeatability of the results, and the
maximum error between the yield of products resulted from that
the duplicate runs was veried to be 3% of the average yield. In
alkali catalytic liquefaction, Zhong and Wei revealed that when
using potassium carbonate as the catalyst, almost all of K
260 280 300 320 340 360 380
10
20
30
WSO
HO
Gas
SR
%
t
w
/
d
l
e
i
Y
Reaction temperatures/
Fig. 1. Inuence of reaction temperature on product yields in the liquefaction of the
paulownia wood without catalyst. Conditions: raw material 60 g; de-ionized water
360 mL; reaction time 10 min.
P. Sun et al. / Energy 35 (2010) 5421e5429 5422
remained in WSO, while nearly no K remained in the HO and SR.
Bestue et al. studied that when using iron powder as the catalyst,
almost all of the total amount of iron was recovered as Fe
3
O
4
and
it remained in SR, while nearly no iron remained in the HO
and the WSO detected by atomic absorption spectrophotometer.
And when using iron as the catalyst, the measurement of iron
dissolved in the WSO, HO was performed with a UV spectro-
photometer and we had the similar results that only trace of iron
remained in the HO and the WSO. As such, for simplication of
the calculation, in the present work, the authors assumed that
when using sodium carbonate as the catalyst, the metal
compounds were primarily present in WSO; when using iron
powder as the catalyst, the iron was primarily present in the solid
residues like Fe
3
O
4
and it was removed thoroughly by acid. For
iron powder, iron was not modied and no trace amounts of
hydrogen were detected in the absence of wood in the reaction
conditions (340

C, 10 min, 360 g water, under vacuum). From all
the iron catalytic liquefaction runs studied in this paper, iron was
recovered as Fe
3
O
4
with only trace amounts of iron when reac-
tion temperature was higher than 280

C.
2.3. Analysis of products
Elemental analysis of the raw material and products was per-
formed on a Flash EA 1112 CHNS/Oanalyzer (USAThermo Electron).
The heating value was obtained from calculation by Dulongs
formula. The compositions of gaseous products collected in a gas-
sampling bag were analyzed by a gas chromatograph equipped
with a thermal conductivity detector (Shimadzu GC-14B). For
characterization of heavy oil, the FTIR spectra of the HO were
recorded using a Shimadzu Instrument IR-8000 spectrophotometer
within the range of 4000e400 cm
1
by the KBr tablet technique.
The components with lower boiling point in the HO products were
analyzed by a gas chromatograph equipped with a mass selective
detector [Agilent GC6890-MS5973(EI), Agilent GC6890, Agilent HP-
5MS (5% phenyl methyl siloxane, 30 m 250 mm 0.25 mm),
temperature program, 100

C(hold 2 min)2250

C (20

C/min,
hold 30 min), inlet temperature 250

C, MS5973 inlet temperature
280

C]. After a solvent delay of 1.4 min, full scan mass spectra were
acquired from 50 to 550 m/z. The identication of the peaks was
based on computer matching of the mass spectra with the NIST
(National Institute of Standards and Technology) 2005 library.
However it should be noted that the GC/MS analysis had a common
limitation in that only volatile compounds can pass through the GC
column, so high molecular heavy oil and non-volatile compounds
were unable to be detected by GC/MS. As such, the composition by
GC/MS represented only the volatile fraction of the heavy oil
products.
3. Results and discussion
3.1. Liquefaction of paulownia in hot compressed water without
a catalyst
Fig. 1 showed the products distribution by the liquefaction of the
paulownia wood in hot compressed water at various temperatures
ranging from 280

C to 360

C for 10 min and 14.29 wt% biomass
without the presence of catalyst. At all temperatures in the range of
280e360

C, generally, the formation of HO (18.59e27.01 wt%
yields) was much greater than that of WSO (10.09e17.35 wt%
270 300 330 360
8
16
24
32
%
t
w
/
d
l
e
i
Y
O
S
W
Reaction temperatures/
270 300 330 360
21
28
35
Fe
Na
2
CO
3
%
t
w
/
d
l
e
i
Y
O
H
270 300 330 360
7
14
21
%
t
w
/
d
l
e
i
Y
R
S
Reaction temperatures/
270 300 330 360
10
20
Reaction temperatures/
%
t
w
/
d
l
e
i
Y
s
a
G
Reaction temperatures/
Fig. 2. Inuence of reaction temperature on product yields in the liquefaction of the paulownia wood with Fe and Na
2
CO
3
catalyst. Conditions: raw material 60 g; de-ionized water
360 mL; reaction time 10 min.
P. Sun et al. / Energy 35 (2010) 5421e5429 5423
yields). As can be clearly shown in Fig. 2, the yield of WSO
decreased monotonously with the increasing reaction temperature,
which may mainly be caused by the dehydration, isomerization,
fragmentation, condensation and repolymerizaion of the small
compounds at high temperatures. And these small compounds
ultimately came into being heavy oil, gas and char. The yield of HO
increased with the increasing reaction temperature up to 300

C,
and it decreased with temperature increasing further. This was
caused by the competition between the two reactions (hydrolysis
and repolymerization) involved in the liquefaction. In fact, hot
compressed water accelerated biomass depolymerization by
hydrolysis. At the initial stage, biomass was swelled, dissoluted,
decomposed, pyrolyzed and depolymerized to small water-soluble
compounds, and these compounds may rearrange to water-insol-
uble compounds through dehydration, condensation, cyclization,
repolymerization and other spontaneous alterations and interac-
tions. The heavy oil tended to maximize at around 300

C, which
might be due to the enhanced gasication/dehydration/condensa-
tion reactions of the products at a higher temperature to yield more
gaseous, water and char products, as evidenced by the increase in
gas and SR yields above 300

C in Fig. 1. The gas product increased
with the increment of temperature from 280

C to 360

C and the
composition of gaseous products mainly contained CO
2
, CO, H
2
, CH
4
and trace of C
2
, C
3
hydrocarbon. The major gaseous product was
CO
2
for all runs. This nding indicated that the decomposition of
paulownia in hot compressed water was quite dependent on
temperature. It had been well demonstrated in previous studies
that when biomass was heated in hot compressed water, the
hemicellulose component was usually decomposed almost
completely at the temperature of 180

C, forming intermediates or
Table 1
Elemental compositions of the crude wood powder and the products obtained from
liquefaction of the paulownia inwater with/without catalyst for 10 min. (a) With 10%
Fe catalyst at 340

C. (b) With 5% Na
2
CO
3
catalyst at 300

C. (c) Without catalyst at
300

C.
Sample Elemental composition,
wt%(dry basis)
HHV
b
(MJ/kg)
C H O
a
N
Crude wood powder 45.52 6.31 48.17 0 15.81
HO, no catalyst 66.81 5.92 27.27 0 26.17
10% Fe 73.50 7.08 19.43 0 31.48
5% Na
2
CO
3
70.29 6.01 23.70 0 28.11
WSO, no catalyst 47.90 5.43 46.68 0 15.63
10% Fe 49.91 5.37 44.73 0 16.57
5% Na
2
CO
3
33.24 4.27
c
0
c
SR, no catalyst 64.41 5.17 30.42 0 23.73
10% Fe 57.44 5.12 37.44 0 20.06
5% Na
2
CO
3
61.52 4.93 33.55 0 21.86
a
By difference.
b
Calculated according to the Dulong Formula, i.e., HHV (MJ/
kg) 0.3383 C 1.422(H O/8).
c
The catalyst remains in the WSO in the Na
2
CO
3
catalytic liquefaction so as to
cause O which cant be calculated by difference and HHV cant be calculated.
Table 2
GCeMS Analysis results for the HO from the operations at 300

C for 10 min without catalyst.
No. RT (min) Name of the compounds Formula Mol. weight Area %
1 2.173 2-Cyclopenten-1-one,2-methyl- C
6
H
8
O 96.13 2.92
2 2.628 Phenol C
6
H
6
O 94.11 1.85
3 3.461 2-Cyclopenten-1-one,2,3-dimethyl- C
7
H
10
O 110.15 1.07
4 3.868 Phenol, 4-methyl- C
7
H
8
O 108.14 1.44
5 4.28 Phenol, 2-methoxy-(guaiacol) C
7
H
8
O
2
124.14 9.64
6 7.054 Phenol, 2-methoxy-4-methyl- C
8
H
10
O
2
138.16 2.87
7 8.672 1,2-Benzenediol,3-methoxy- C
7
H
8
O
3
140.14 1.56
8 8.745 Nonanoic acid C
9
H
18
O
2
158.24 1.98
9 8.992 Phenol, 4-ethyl-2-methoxy- C
9
H
12
O
2
152.19 3.58
10 10.011 Phenol, 2,6-dimethoxy-(syringol) C
8
H
10
O
3
154.16 13.20
11 10.094 Eugenol C
10
H
12
O
2
164.2 0.62
12 10.206 Phenol, 2-methoxy-4-propyl- C
10
H
14
O
2
166.22 2.37
13 10.51 Ethanone, 1-(3-hydroxyphenyl)- C
8
H
8
O
2
136.15 0.88
14 10.57 Vanillin C
8
H
8
O
3
152.15 1.03
15 11.03 Benzoic acid, 4-hydroxy-3-methoxy- C
8
H
8
O
4
168.15 3.23
16 11.065 Phenol, 2-methoxy-4-(1-propenyl)- C
10
H
12
O
2
164.2 0.72
17 11.407 Ethanone, 1-(4-hydroxy-3-methoxyphenyl)- C
9
H
10
O
3
166.17 1.44
18 11.507 1H-Inden-1-one, 2,3-dihydro-5-methoxy- C
10
H
10
O
2
162.19 0.83
19 11.568 Ethanone, 1-(2,5-dihydroxyphenyl)- C
8
H
8
O
3
152.15 0.64
20 11.719 Benzene, 1,2,3-trimethoxy-5-methyl- C
10
H
14
O
3
182.22 2.52
21 11.776 Benzeneacetic acid, 4-hydroxy-3-methoxy- C
9
H
10
O
4
182.17 1.26
22 12.344 2-Propanone, 1,1-diphenyl- C
15
H
14
O 210.27 2.90
23 12.569 Naphthalene, 1,2,3,4-tetrahydro-5-methoxy- C
11
H
14
O 162.23 0.72
24 12.625 2-Butanone, 4-(4-hydroxy-3-methoxyphenyl)- C
11
H
14
O
3
194.23 1.00
25 12.708 Benzaldehyde, 4-hydroxy-3,5-dimethoxy- C
9
H
10
O
4
182.17 0.77
26 12.781 2-Propenoic acid, 3-(4-hydroxy-3-methoxyphenyl)- C
10
H
10
O
4
194.18 1.56
27 12.833 5,6,7,8-Tetrahydro-2-methyl-1,4-naphthoquinone C
11
H
12
O
2
176.21 0.61
28 13.176 Ethanone,1-(4-hydroxy-3,5-dimethoxyphenyl)- C
10
H
12
O
4
196.2 0.71
29 13.228 [1,1
0
-Biphenyl]-2,2
0
-diol C
12
H
10
O
2
186.21 0.62
30 13.41 1-Butanone,1-(2,4,6-trihydroxy-3-methylphenyl)- C
11
H
14
O
4
210.23 1.46
31 13.484 Naphthalene, 2,3-dimethoxy- C
12
H
12
O
2
188.22 0.63
32 14.056 Dihydrocoumarin, 4,4,5,7-tetramethyl- C
13
H
16
O
2
204.26 1.33
33 14.689 Tridecanoic acid C
13
H
26
O
2
214.34 4.04
34 14.867 3-Methoxy-6,7,8,9-tetrahydro-dibenzofuran-2-ol C
13
H
14
O
3
218.25 0.73
35 14.953 3-Methoxy-6,7,8,9-tetrahydro-dibenzofuran-2-ol C
13
H
14
O
3
218.25 1.43
36 16.679 9-Octadecenoic acid, (E)- C
18
H
34
O
2
282.46 0.77
37 16.913 Octadecanoic acid C
18
H
36
O
2
284.48 0.81
38 20.351 Benzeneacetic acid, 4-hydroxy-3-methoxy-, methyl ester C
10
H
12
O
4
196.2 2.14
39 20.841 Benzeneacetic acid, 4-hydroxy-3-methoxy-,methyl ester C
10
H
12
O
4
196.2 0.77
Total area % 78.66
Only peak area more than 0.6% of total area was given.
P. Sun et al. / Energy 35 (2010) 5421e5429 5424
aqueous and oily products (carbohydrates, aldehydes, ketone,
alcohols, phenol and its derivatives), regardless of tree species[17]
and much of the lignin dissolved in the water at 230

C[18].
Hydrolysis and pyrolysis of the remaining lignocellulosic solids
took place at higher temperatures. Kobayashi et al. [19] revealed
that the cellulose crystal in the wood (Douglas r) began decom-
posing between 473 and 523 K in HCWtreatment, and the cellulose
structure was changed from cellulose I to cellulose II. The cellulose
crystal was decomposed completely at 573 K. Therefore decom-
position of woody biomass in hot compressed water treatment was
a sequence of complex reactions, which was swelling, dissolution
process, hydrolysis, dehydration and condensation. The hydrolysis
product was usually water-soluble, and the dehydration product
was usually water-insoluble. Therefore, the water-insoluble
product was the dehydration substance of cellulose and the con-
densationedehydration substance of lignin.
3.2. Inuence of catalyst on liquefaction of paulownia in hot
compressed water
Inuences of reaction temperature and catalysts on the
products distribution, particularly heavy oil products, which
possessed a high caloric value [12] were discussed below. These
experiments were conducted under conditions of temperature
range 280e360

C for 10 min and 14.29 wt% biomass in water
with Fe and Na
2
CO
3
catalyst.
3.2.1. Inuence of Fe catalyst
It was clear from Fig. 2, within the range of 280e360

C, the
yield of WSO decreased monotonously with the increasing of the
reaction temperature irrespective of the presence of a catalyst.
With iron as paulownia liquefaction catalyst, the yield of HO
increased from 21.88 wt% to 36.34 wt% with the increment of
temperature from 280

C to 340

C. The reason might be that when
temperature increased, the pyrolysis/isomerization/rearrangement
of the carbohydrates, alcohols, aldehydes produced by biomass
hydrolysis would occur due to the endothermic nature of the
pyrolysis reaction (being thermodynamically favorable as temper-
ature increased), and then ketone, aldehydes and acids compounds
were formed. With the Fe catalyst, the aldehyde was rst converted
to acid, followed by transformation into ketone [20]. The phenolic
derivatives were adsorbed dissociatively on the ferriferrous oxide
surface to form phenoxy species. Then the phenoxy-adsorbed
species were linked with eOH groups to form small molecular
weight phenolic compounds. This may account for that the yield of
HO increased as temperature increased from 280 to 340

C. Maybe
it was evidenced by the GCeMS analysis of heavy oil produced with
Fe catalyst that contained more ketone and phenolic derivatives.
With the temperature increasing further to 360

C, gasication/
dehydration reactions of the HO intermediates/products (to form
more gases and char) would be dominant over the phenoxy-
adsorbed reaction, hence leading to the declining yield of HO as
shown in Fig. 2. The dominance of gasication/dehydration of HO
may be evidenced by the apparently increased yields of gas and SR
from 340 to 360

C, as revealed in Fig. 2. In summary, in the tested
temperature range (280e360

C), the highest yield of total oil (at
about 53 wt%) was produced at the lowest temperature, while the
maximumyield of HO (at about 36.34 wt%) was obtained at 340

C
under the experimental conditions adopted in this work. The
reduction in the yields of liquid organic products accompanied by
the increases in the yield of gas and SR suggested the gasication,
pyrolysis and dehydration reactions of the liquid products form gas
and solid residue at higher temperatures. Similar conclusions were
made by Qu et al. and Xu et al. [10,12].
Table 3
GCeMS Analysis results for the HO from the operations at 340

C for 10 min with Fe catalyst.
No. RT (min) Name of the compounds Formula Mol. weight Area %
1 2.177 2-Cyclopenten-1-one, 2-methyl- C
6
H
8
O 96.13 6.87
2 2.385 2-Cyclopenten-1-one, 2,3-dimethyl- C
7
H
10
O 110.15 1.23
3 2.624 Phenol C
6
H
6
O 94.11 1.41
4 2.884 Ethanone, 1-(2-furanyl)- C
6
H
6
O
2
110.11 1.43
5 2.927 2-Cyclopenten-1-one, 2,3-dimethyl- C
7
H
10
O 110.15 2.90
6 3.239 2-Cyclopenten-1-one, 3,4,5-trimethyl- C
8
H
12
O 124.18 0.77
7 3.343 Furan, 2,4-dimethyl- C
6
H
8
O 96.13 1.01
8 3.456 2-Cyclopenten-1-one, 2,3-dimethyl- C
7
H
10
O 110.15 2.41
9 3.803 Phenol, 2-methoxy- C
7
H
8
O
2
124.14 0.81
10 3.864 Phenol, 4-methyl- C
7
H
8
O 108.14 0.73
11 4.275 Phenol, 2-methoxy-(guaiacol) C
7
H
8
O
2
124.14 11.15
12 4.484 Phenol, 2-methoxy- C
7
H
8
O
2
124.14 1.04
13 4.553 5-Ethyl-2-furaldehyde C
7
H
8
O
2
124.14 0.84
14 5.242 1,4-Benzenediol, 2-methyl- C
7
H
8
O
2
124.14 1.52
15 7.05 Phenol, 2-methoxy-4-methyl- C
8
H
10
O
2
138.16 3.72
16 7.921 Bicyclo[4.1.0]heptan-3-ol, 4,7,7-trimethyl-, [1R-(1.alpha.,3.beta.,4.beta.,6.alpha.)]- C
10
H
18
O 154.25 0.72
17 8.697 1,4-Benzenediol, 2,3,5-trimethyl- C
9
H
12
O
2
152.19 1.77
18 8.988 Phenol, 4-ethyl-2-methoxy- C
9
H
12
O
2
152.19 4.98
19 9.036 1H-Inden-1-one, 2,3-dihydro- C
9
H
8
O 132.16 0.65
20 10.007 Phenol, 2,6-dimethoxy-(syringol) C
8
H
10
O
3
154.16 9.11
21 10.089 Eugenol C
10
H
12
O
2
164.2 0.75
22 10.206 Phenol, 2-methoxy-4-propyl- C
10
H
14
O
2
166.22 4.73
23 10.765 1H-3a,7-Methanoazulene, 2,3,4,7,8,8a-hexahydro-3,6,8,8-tetramethyl-, [3R-(3.alpha.,3a.beta.,7.beta.,8a.alpha.)]- C
15
H
24
204.35 0.66
24 10.904 Ethanone, 1-(2,5-dimethoxyphenyl)- C
10
H
12
O
3
180.2 0.78
25 11.03 Benzoic acid, 4-hydroxy-3-methoxy- C
8
H
8
O
4
168.15 2.58
26 11.719 Benzene, 1,2,3-trimethoxy-5-methyl- C
10
H
14
O
3
182.22 2.31
27 11.78 Benzeneacetic acid, 4-hydroxy-3-methoxy- C
9
H
10
O
4
182.17 0.62
28 12.343 2-Propanone, 1,1-diphenyl- C
15
H
14
O 210.27 3.78
29 12.786 Phenol, 2,6-dimethoxy-4-(2-propenyl)- C
11
H
14
O
3
194.23 1.05
30 14.684 Tridecanoic acid C
13
H
26
O
2
214.34 1.84
31 14.953 Bicyclo[3.1.0]hexane, 1,6-diphenyl- C
18
H
18
234.34 1.06
32 20.368 Benzeneacetic acid, 4-hydroxy-3-methoxy-, methyl ester C
10
H
12
O
4
196.2 1.15
Total area% 76.37
Only peak area more than 0.6% of total area was given.
P. Sun et al. / Energy 35 (2010) 5421e5429 5425
Uddin et al. studied that Fe
3
O
4
possessed catalytic activity in
biomass tar decomposition [21]. Polychonopoulou et al. reported
the absorption-enhanced reforming of phenol (as a model
compound of biomass tar) by steam over supported Fe catalysts
[22] and suggested a mechanism for phenol reforming on sup-
ported iron oxide catalysts that phenol was adsorbed dissociatively
on the iron oxide surface to form surface phenoxy species. The
derived hydrocarbon fragments which mainly contained phenolic
derivatives and ketones from the phenoxy-adsorbed species
were oxidized by labile O and/or eOH species from H
2
O [20,23],
and by eOH groups residing on iron oxides surfaces to form H
2
, CO,
and CO
2
.
3.2.2. Inuence of Na
2
CO
3
catalyst
In order to investigate the effect of Na
2
CO
3
on the decomposi-
tion of paulownia in hot compressed water, the product distribu-
tion was analyzed by adding 5% Na
2
CO
3
as catalyst. The alkali
catalyst increased the yield of HO, WSO, conversion and inhibited
the formation of char. Fig. 2 showed the products yield (WSO, HO,
SR and Gas) from subjecting paulownia to hot compressed water
liquefaction with 5% Na
2
CO
3
as catalyst at different temperatures.
The yield of HO was found in the range of 23.58e31.91 wt% and
increased with the increasing reaction temperature up to 300

C,
while it decreased as temperature increased further. Subsequently,
the yield of WSO, SR and Gas products varied from 25.11% to
32.91 wt%, 7.23e14.71 wt% and 15.89e20.33 wt%, respectively.
Compared with the results obtained without catalyst, when cata-
lyst with 5% Na
2
CO
3
was used, the yield of solid residue was found
to drop from about 19.83 wt% to 8.12 wt% while the WSO, HO
products of paulownia increased signicantly. And the gaseous
products increased with the increasing temperature and when
using Na
2
CO
3
as catalyst, the products were more than that without
catalyst. Minowa et al. [24] studied the liquefaction of cellulose in
hot compressed water in the presence of Na
2
CO
3
. They reported
that the use of Na
2
CO
3
for liquefaction of cellulose decreased the
formation of char and favored the formation of oil. Our results were
in good agreement with this previous report. It seemed that the
alkaline catalyst such as Na
2
CO
3
, prevented the dehydration reac-
tion and hastened the hydrolysis and gasication reaction. They
weakened the CeC bond, thereby decreased the activation energy
for the complex reaction. It can be assumed that the alkali catalyst
caused swelling of paulownia and led to an increase in internal
surface area and separation of structure linkages between hol-
ocellulose and lignin. On the other hand, Na
2
CO
3
caused weakening
Table 4
GCeMS Analysis results for the HO from the operations at 300

C for 10 min with Na
2
CO
3
catalyst.
No. RT (min) Name of the compounds Formula Mol. weight Area %
1 2.177 2-Cyclopenten-1-one, 2-methyl- C
6
H
8
O 96.13 2.05
2 2.307 Benzene, (1-methylethyl)- C
9
H
12
120.19 4.26
3 2.627 Phenol C
6
H
6
O 94.11 2.47
4 2.931 2-Cyclopenten-1-one, 2,3-dimethyl- C
7
H
10
O 110.15 1.17
5 3.46 2-Cyclopenten-1-one, 2,3-dimethyl- C
7
H
10
O 110.15 0.90
6 4.279 Phenol, 2-methoxy-(guaiacol) C
7
H
8
O
2
124.14 8.31
7 7.05 Phenol, 2-methoxy-4-methyl- C
8
H
10
O
2
138.16 1.94
8 8.901 Nonanoic acid C
9
H
18
O
2
158.24 2.03
9 9.217 Phenol, 4-ethyl-2-methoxy- C
9
H
12
O
2
152.19 5.35
10 10.44 Phenol, 2,6-dimethoxy-(syringol) C
8
H
10
O
3
154.16 11.78
11 10.544 Eugenol C
10
H
12
O
2
164.2 0.84
12 10.687 Phenol, 2-methoxy-4-propyl- C
10
H
14
O
2
166.22 5.21
13 11.71 Benzoic acid, 4-hydroxy-3-methoxy- C
8
H
8
O
4
168.15 2.52
14 11.753 Phenol, 2-methoxy-4-(1-propenyl)- C
10
H
12
O
2
164.2 1.28
15 12.187 Ethanone, 1-(4-hydroxy-3-methoxyphenyl)- C
9
H
10
O
3
166.17 0.74
16 12.586 Benzene, 1,2,3-trimethoxy-5-methyl- C
10
H
14
O
3
182.22 4.39
17 12.655 2-Propanone, 1-(4-hydroxy-3-methoxyphenyl)- C
10
H
12
O
3
180.2 1.39
18 13.327 Phenol, 2,6-dimethoxy-4-(2-propenyl)- C
11
H
14
O
3
194.23 1.17
19 13.388 2-Propanone, 1,1-diphenyl- C
15
H
14
O 210.27 8.22
20 13.752 Benzeneacetic acid, 4-hydroxy-3-methoxy- C
9
H
10
O
4
182.17 0.81
21 13.947 Phenol, 2,6-dimethoxy-4-(2-propenyl)- C
11
H
14
O
3
194.23 0.67
22 14.745 2-Pentanone, 1-(2,4,6-trihydroxyphenyl) C
11
H
14
O
4
210.23 0.77
23 14.81 Naphthalene, 2,3-dimethoxy- C
12
H
12
O
2
188.22 0.86
24 15.434 1,4-Naphthalenedione, 2-hydroxy-3-methoxy- C
11
H
8
O
4
204.18 0.74
25 16.149 Tridecanoic acid C
13
H
26
O
2
214.34 2.79
26 16.297 3-Methoxy-6,7,8,9-tetrahydro-dibenzofuran-2-ol C
13
H
14
O
3
218.25 0.72
27 16.392 7-Isopropenyl-1,4a-dimethyl- C
15
H
22
O 218.33 2.11
4,4a,5,6,7,8-hexahydro-
3H-naphthalen-2-one
28 17.935 2,5-Cyclohexadien-1-one, C
16
H
24
O 232.36 0.82
2,6-bis(1,1-dimethylethyl)-4-ethylidene-
29 21.915 Benzeneacetic acid, 4-hydroxy-3-methoxy-, methyl ester C
10
H
12
O
4
196.2 1.19
Total area% 77.50
Only peak area more than 0.6% of total area was given.
0.0
20.0
40.0
60.0
80.0
100.0
500.0 750.0 1000.0 1250.0 1500.0 1750.0 2000.0 2500.0 3000.0 3500.0 4000.0
1/cm Testscan Shimadzu FTIR 8000 series
a
b
c
Fig. 3. FTIR spectra of HOs with and without catalyst. (a) With 10% Fe catalyst at
340

C. (b) With 5% Na
2
CO
3
catalyst at 300

C. (c) Without catalyst at 300

C.
P. Sun et al. / Energy 35 (2010) 5421e5429 5426
of the intermolecular interaction of the polymeric chains and at the
same time, catalyses not only interlinked dehydration but also
promoted the processes of retro-aldol cleavage and condensation of
the products produced [25].
3.3. Elemental analysis of the products and carbon balance
The elemental composition of liquefaction products and solid
products obtained with and without catalyst was analyzed. The
results and the HHV were presented in Table 1, compared with that
of the crude wood powder.
Compared with the data of the crude wood powder, the HO
obtained had a much higher content of carbon and a much lower
concentration of oxygen, leading to signicantly increased caloric
values, irrespective of whether the catalyst was present and
which catalyst was used. This might also account for the chem-
ical compositions and the originating sources of the HO. As evi-
denced by the GC/MS analysis results of the HO presented later in
Tables 2e4, the HO mainly consisted of phenolic compounds and
its derivatives, ketone, long-chain carboxylic acids/esters and
hydrocarbons, which were obviously the decomposition products
from lignin and cellulose. The greatly reduced oxygen content in
the HO may partially be due to the pyrolysis reactions (forming
CO
2
/CO) and the dehydration reactions (forming H
2
O) in the
liquefaction process. The presence of various catalysts only
slightly varied in the elemental compositions and HHVs of the
liquid products. The HO, WSO and SRs obtained at various
reaction temperatures had similar elemental compositions of
carbon, hydrogen, and oxygen and hence similar HHVs, respec-
tively, as similarly observed by Qu et al., Zhong and Wei [10,15].
As shown in Table 1, the HHVs of HOs were in the range of
26e31 MJ/kg, in comparison with the HHV of 15.8 MJ/kg for the
crude wood sample, the energy density increased almost by
100%, and the caloric value of the HOs was comparable to that
of high rank coal. The WSOs had similar HHVs with the HHV of
crude wood powder and it could be used as boiler fuel. The
residue was char-like material. It had high caloric value of
around 20.0 MJ/Kg, and it can be used as a solid fuel. If both the
Fig. 4. GCeMS spectrum of HOs with and without catalyst. (a) Without catalyst at 300

C for 10 min (b)With 10% Fe catalyst at 340

Cfor 10 min. (c) With 5% Na
2
CO
3
catalyst at
300

Cfor 10 min.
P. Sun et al. / Energy 35 (2010) 5421e5429 5427
HO and residue would be used as a solid fuel, no separation was
required. So liquefaction of woody biomass in water may be
a promising technique for upgrading biomass feedstocks with
low caloric value to liquid fuels with greatly increased caloric
value.
The overall material balance closed to within 1% and the carbon
balance within 10%. Reliable complete balances had been proved
difcult to obtain because of unavoidable losses during the
recovery and separations of the different products of the reaction,
especially for the sticky oil and the water adhering to the reactor
walls. And small carbon lost due to the loss of lowmolecular weight
organics when measuring the weight of water-soluble products in
the aqueous phase, since a drying process was needed to vaporize
the water present and it would also vaporize the light organic
fraction. This result showed good agreement with previous work by
Minowa et al. [3] and Bestue et al. [13].
3.4. FTIR analysis of HO
The highest yields of HO obtained with and without catalyst
were analyzed by FTIR spectrophotometer (Shimadzu IR-8000).
The oil samples were placed directly in the diffuse reectance
attachment sample holder (between two KBr salt plates in
a microcompression cell) and the FTIR spectra of HO were shown in
Fig. 3. As it can be seen from Fig. 3, the HO had similar functional
groups whether a catalyst was present or not. That the usual OeH
wavenumber drifted from 3396 cm
1
to 3363 cm
1
indicated that
there may be alcohols, phenols, carboxylic acids in the HO. The
strong absorbance at 2930 cm
1
and 1460 cm
1
were CeH vibra-
tion and CeH deformation vibration, respectively. The absorption
peak at 1700 cm
1
represented the C]O stretching vibration and
indicated the presence of ketone. A slight move of the C]O
stretching vibration to high wavenumber in HO without the pres-
ence of catalyst showed the presence of OeH electron withdrawing
group near the carbonyl group and revealed the presence of
carboxylic acids. The absorption peaks between 1680 and
1550 cm
1
suggested the possible presence of alkenes or aromatics.
The peaks at 1605 cm
1
, 1495 cm
1
, and 1456 cm
1
well repre-
sented aromatic nuclei and the existence of a medium intense
absorption peak in all HO, which indicated there could be aromatic
and its derivates. The strong peaks at 1361 cm
1
, 1271.9 cm
1
,
1217.9 cm
1
represented the syringyl ring, guaiacyl ring plus C]O
stretching in HO [26]. The existence of a CeO stretching within
1075e1020 cm
1
(1034 cm
1
) indicated the presence of primary,
secondary and tertiary alcohols, phenols, and esters. Absorption
peaks at 858.2 cm
1
, 815.8 cm
1
and 748.3 cm
1
indicated the
possible presence of substituted aromatic groups. Therefore, the HO
may mainly consist of phenolic compounds and its derivatives,
ketone, long-chain carboxylic acids/esters and hydrocarbons. The
heavy oils obtained with sodium carbonate catalyst were also
analyzed by FTIR by Qian et al. [27] and the analytical results are
consistent with their work.
3.5. Composition of the HOs
Theidenticationof all thecompounds inthese three HOobtained
withFe catalyst at 340

C, with Na
2
CO
3
catalyst at 300

C and without
the presence of catalyst at nal temperature of 300

C were qualita-
tively characterized by GCeMS as it had a comprehensive library of
the spectra of almost all the compounds. The compounds of the
identied compounds in the HOs with and without catalyst were
showed inFig. 4 and listedinTables 2e4 (only compounds withmore
than 0.6% of total area were shown). The area % for each compound
identied (dened by the percentage of the compounds
chromatographic area out of the total area) and the total area % for all
the identied compounds were shown in the Tables 2e4
It mainly consisted of phenolic compounds and its derivatives in
HO obtained by liquefaction of paulownia irrespective of whether
a catalyst was present or not. In the HO produced without the
presence of catalyst, the phenolic compounds and its derivatives
were the major compounds identied which comprised almost
37.54%, followed by carboxylic acids (12.07%) and ketones (10.79%).
Four major peaks appearing at about 4.28, 8.992, 10.011 and
14.689 min were identied as Phenol, 2-methoxy-(9.64%), Phenol,
4-ethyl-2-methoxy-(3.58%), Phenol, 2,6-dimethoxy-(13.2%) and
Tridecanoic acid(4.04%) respectively and these six compounds
accounted for about 36.61% of total peak area. Hydrothermal
treatment of pinewood was studied at 280

Cfor 15 min and

liquefaction of pinewood in water in an autoclave under a starting
pressure of 1 MPa argon gas and at 623 K for 20 min. And the heavy
oil compositions were analyzed by Karagz et al. [28], Liu and
Zhang [29], respectively. Compared with the compounds detected
in this previous report, we found that there was a mixture of
guaiacyl and syringyl type phenols presented in the heavy oil of
liquefaction of paulownia while only guaiacyl type phenols pre-
sented in the heavy oil of liquefaction of pinewood. Maybe it pre-
dicted that there was a mixture of guaiacyl propane and syringyl
propane in the lignin of paulownia but only guaiacyl propane in the
lignin of pinewood. The major compound for the HO produced
with the presence of Fe catalyst was Phenol, 2-methoxy- (11.15%)
with retention time 4.275 min, Phenol, 2,6-dimethoxy- (9.11%) with
retention time 10.007 min and 2-Cyclopenten-1-one, 2-methyl-
(6.87%) with retention time 2.177 min. In the HO produced with Fe
catalyst, the phenolic compounds and its derivatives contents were
42.77%, followed by ketones (14.83%) and carboxylic acids (5.04%).
The major compounds of the HO produced with the presence of
Na
2
CO
3
catalyst were Phenol, 2,6-dimethoxy- (11.78%) with reten-
tion time 10.44 min, Phenol, 2-methoxy- (8.31%) with retention
time 4.279 min and 2-Propanone, 1,1-diphenyl- (8.22%) with
retention time 13.388 min. The compound Benzene (1-methyl-
ethyl)- was observed at 2.307 min (4.26%) in HO with Na
2
CO
3
catalyst. However, The compound was not found with and without
Fe catalyst runs. In the HO produced with Na
2
CO
3
catalyst, the
phenolic compounds and its derivatives made up approximately
39.12% while in the heavy oil produced with carboxylic acids
(8.15%) and ketones (7.05%), they made up the total. Only three
compounds with more than 3% area, namely Phenol, 2-methoxy-,
Phenol, 2,6-dimethoxy- and Phenol, 4-ethyl-2-methoxy- were all
present in the HOs with and without the presence of tested catalyst.
Paulownia was mainly composed of cellulose, hemicellulose,
lignin and extractives. In the case of lignin degradation in near and
supercritical water without catalysts, ether bonds were easily
hydrolyzed and phenolic compounds were formed. In principle,
every compound could be obtained via different reaction pathways.
The most major ether bonds in lignin were that one propyl side
chain of a hydroxylphenylpropane unit and the hydroxyl group
attached to a benzene ring of another hydroxylphenylpropane unit.
So, disassembly of CeO ester bonds (a-o-4 and b-o-4) and car-
bonecarbon bonds were important factors in the degradation of
lignin. It was known that the degradation of lignin produced
predominantly 2-methoxy-phenol and then it was decomposed to
form 1,2-Benzenediol, which was then hydrolyzed into phenol. A
possible lignin mechanism via free radical species was proposed
where alkyl hydroxyl groups in a-position were rst eliminated
from the propane side chain along with a b-ether cleavage and
formed monolignols namely p-coumaryl alcohol, coniferyl alcohol,
sinapyl alcohol. Transfer of hydrogen from the hydroxyl group at g-
position to free radical species led to the formation of methoxy
phenols, dimers and formaldehyde [30]. In addition, condensation/
P. Sun et al. / Energy 35 (2010) 5421e5429 5428
cyclization of the cellulose/hemicellulose-derived carbohydrates in
the presence of alkali catalyst might also lead to the formation of
phenol and phenolic compounds. The ketones might be formed
from cellulose/hemicellulose-derived carbohydrates by isomeriza-
tion/retro-Aldol reactions. The carboxylic acids may be primarily
formed from wood extractives and the cellulose and hemicellulose
components of the biomass feedstock by complex hydrolysis and
hydration reactions. The signicant amount of the long-chain
alkanes were detected in the HO in this work, as shown in Tables 3
and 4, thus suggesting that they might be formed by condensation
and polymerization of the degradation products from cellulose/
hemicellulose during the liquefaction process. The GC/MS results
thus suggested that the catalysts used in this work can effectively
promote the degradation reactions of lignin and cellulose, yielding
more phenolic compounds and ketone. At the same time the
catalysts also inhibited the formation of acid, reducing the acid
content of heavy oil.
It should be noted that Tables 2e4 gave peak areas of identied
compounds in percents, which represented the relative concen-
trations rather than the actual concentrations of the detected
compounds. So the exact yields of the compounds, e.g., in mass % of
the initial wood, were not quantiable. However, with carefully
controlled GC/MS conditions, such as the concentration of the oil
sample (in solvent), the injected amount, the temperature program
and the relative concentrations derived from the peak areas may
still be useful for discussion on a qualitative basis.
4. Conclusions
An experimental study on direct liquefaction of paulownia in
hot compressed water with and without catalyst at 280e360

C
was carried out. The experimental results showed that with an
increase in the reaction temperature range, the yield of HO
increased and then decreased. In the present study, the maximum
heavy oil yield was found to be 36.34% with Fe catalyst and the
minimum solid residue yield was obtained when Na
2
CO
3
was
employed as catalyst. The heavy oil mainly consisted of phenol
derivatives, ketones, carboxylic acid/ester, benzene derivatives and
long-chain alkanes, aldehydes and its derivatives. This kind of oil
was unstable. Further research was needed to upgrade the heavy oil
products to make it more stable and value-added. Maybe catalytic
cracking as a cheaper route was a promising alternative for
upgrading this kind of oil and producing value added chemicals and
liquid fuels. The study had also shown that employing different
catalysts could change the distribution and relative abundance of
produced compounds. Further research was needed to select
proper catalyst to narrow the product distribution and produce
desired compounds to make the process more industrial.
Acknowledgements
The authors are grateful for the nancial support from the
science and technology innovative team (STIT) oil-for Comple-
mentary and Alternative Energy Research Team funded projects in
Henan Province, China. The authors would also like to thank Mr.
Zhang zhan at Dalian Institute of Chemical Physical, Chinese
Academy of Sciences for assistance on GCeMS measurements.
References
[1] Centi G, van Santen RA. Catalysis for renewables: from feedstock to energy
production. Weinheim: WILEY-VCH Verlag GmbH & Co. KGaA; 2007.
[2] Shuit SH, Tan KT, Lee KT, Kamaruddin AH. Oil palm biomass as a sustainable
energy source: a Malaysian case study. Energy 2009;34(9):1225e35.
[3] Minowa T, Kondo T, Sudirjo ST. Thermochemical liquefaction of Indonesia
biomass residues. Biomass and Bioenergy 1998;14:517e24.
[4] Zhang S, Yan Y, Li T, Ren Z. Upgrading of liquid fuel from the pyrolysis of
biomass. Bioresource Technology 2005;96:545e50.
[5] Hans L, Vilja P, Lea L, Kristjan K, Hindrek T, Rein V, et al. Trends in biomass thermo-
chemical liquefaction: global experience and recent studies in Estonia. Proceedings of
the Estonian Academy of Sciences Chemistry 2005;54(4):194e229.
[6] Zhang Q, Chang J, Wang T, Xu Y. Review of biomass pyrolysis oil properties
and upgrading research. Energy Conversion and Management 2007;48
(1):87e92.
[7] Park HJ, Heo HS, Jeon J, Kim J, Ryoo R, Jeong K, et al. Highly valuable chemicals
production from catalytic upgrading of radiata pine sawdust-derived pyrolytic
vapors over mesoporous MFI zeolites. Applied Catalysis B: Environmental
2010;95(3e4):365e73.
[8] Akiya N, Savage PE. Roles of water for chemical reactions in high-temperature
water. Chemical Reviews 2002;102(8):2725e50.
[9] Appell HR, Fu YC, Friedman S, Yavorsky PM, Wender I. Converting organic
wastes to oil. report no. 7560. Washington, D.C.: U.S. Bureau of Mines; 1971.
[10] Qu Y, Wei X, Zhong C. Experimental study on the direct liquefaction of Can-
ninghamia lanceolata in water. Energy 2003;28:597e606.
[11] Mazaheri H, Lee KT, Bhatia S, Mohamed AR. Subcritical water liquefaction of
oil palm fruit press ber for the production of bio-oil: effect of catalysts.
Bioresource Technology 2010;101:745e51.
[12] Xu C, Lad N. Production of heavy oils with high caloric values by direct
liquefaction of woody biomass in sub/near-critical water. Energy & Fuels
2008;22:635e42.
[13] Bestue-Labazuy C, Soyer N, Bruneau C, Brault A. Wood liquefaction with
hydrogen or helium in the presence of iron additives. The Canadian Journal of
Chemical Engineering 1985;63(4):634e8.
[14] Zhang L, Champagne P, Xu C. Bio-crude production from secondary pulp/
paper-mill sludge and waste newspaper via co-liquefaction in hot-
compressed water. Energy; 2010; doi:10.1016/j.energy.2010.05.029.
[15] Zhong C, Wei X. A comparative experimental study on the liquefaction of
wood. Energy 2004;29:1731e41.
[16] Xu C, Lancaster J. Conversion of secondary pulp/paper sludge powder to liquid
oil products for energy recovery by direct liquefaction in hot-compressed
water. Water Research 2008;42:1571e82.
[17] Hayashi N, Fujita S, Irie G, Sakaki T, Shibata M. Reaction kinetics of cellulose
decomposition in hot-compressed-water. Journal of the Japan Institute of
Energy 2004;83(10):805e14 [in Japanese].
[18] Hashaikeh R, Fang Z, Butler IS, Hawari J, Kozinski JA. Hydrothermal dissolution
of willow in hot compressed water as a model for biomass conversion. Fuel
2007;86:1614e22.
[19] Kobayashi N, Okada N, Hirakawa A, Sato T, Kobayashi J, Hatano S, et al.
Characteristics of solid residues obtained from hot-compressed-water treat-
ment of woody biomass. Industrial & Engineering Chemistry Research
2009;48(1):373e9.
[20] Masuda T, Kondo Y, Miwa M, Shimotori T, Mukai SR, Hashimoto K, et al.
Recovery of useful hydrocarbons from oil palm waste using ZrO
2
supporting
FeOOH catalyst. Chemical Engineering Science 2001;56:897e904.
[21] Uddin MA, Tsuda H, Wu S, Sasaoka E. Catalytic decomposition of biomass tars
with iron oxide catalysts. Fuel 2008;87:451e9.
[22] Polychronopoulou K, Bakandritsos A, Tzitzios V, Fierro JLG, Efstathiou AM.
Absorption-enhanced reforming of phenol by steam over supported Fe cata-
lysts. Journal of Catalysis 2006;241:132e48.
[23] Polychronopoulou K, Costa CN, Efstathiou AM. The role of oxygen and
hydroxyl support species on the mechanism of H
2
production in the steam
reforming of phenol over metal oxide-supported-Rh and -Fe catalysts.
Catalysis Today 2006;112:89e93.
[24] Minowa T, Zhen F, Ogi T. Cellulose decomposition in hot-compressed water
with alkali or nickel catalyst. The Journal of Supercritical Fluids 1998;
13:253e9.
[25] Ca glar A, Demirbas A. Conversion of cotton cocoon shell to liquid products by
supercritical uid extraction and low pressure pyrolysis in the present of
alkalis. Energy Conversion and Management 2001;42:1095e104.
[26] Meier D, Larimer DR, Faix O. Direct liquefaction of different lignocellulosics
and their constituents 2. Molecular weight determination, gas chromatog-
raphy, i.r. spectroscopy. Fuel 1986;65:916e21.
[27] Qian Y, Zuo C, Tan J, He J. Structural analysis of bio-oils from sub-and super-
critical water liquefaction of woody biomass. Energy 2007;32(3):196e202.
[28] Karagz S, Bhaskar T, Muto A, Sakata Y. Comparative studies of oil composi-
tions produced from sawdust, rice husk, lignin and cellulose by hydrothermal
treatment. Fuel 2005;84:875e84.
[29] Liu Z, Zhang F. Effects of various solvents on the liquefaction of biomass to
produce fuels and chemical feedstocks. Energy Conversion and Management
2008;49:3498e504.
[30] Amen-Chen C, Pakdel H, Roy C. Production of monomeric phenols by ther-
mochemical conversion of biomass: a review. Bioresource Technology
2001;79:277e99.
P. Sun et al. / Energy 35 (2010) 5421e5429 5429