Combined supercritical and subcritical conversion of cellulose for fermentable hexose production in a flow reaction system

Abstract

Using research on a batch system as basis, a flow reactor was designed and applied in the combined supercritical and subcritical hydrolysis of cellulose for fermentable hexose production. The results show that when the supercritical parameters were maintained, the hexose yield first increased with the rise in subcritical temperature, and then decreased after the maximum yield was obtained. This maximum yield of fermentable hexoses from cellulose was 31.5% ± 1.4%, which was obtained under the following conditions: cellulose concentration of 3.53 ± 0.24 g L−1, supercritical temperature of 380 °C, supercritical reaction time of 9.70 ± 0.66 s, subcritical temperature of 240 °C, and subcritical reaction time of 48.49 ± 3.31 s. The appropriate ranges of cellulose concentration (around 3.5 g L−1) and reaction time (9–10 s for supercritical process and 45–50 s for subcritical process), which depended on the flows of water and material sludge, were also crucial in obtaining a high hexose yield. Compared with the batch system, the flow reaction system can yield a reasonable amount of hexose from cellulose hydrolysis and proved to be considerably promising for practical applications, especially for combined supercritical and subcritical technology on lignocellulosic resources.

Full text

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright

Author's personal copy
Chemical Engineering Journal 166 (2011) 868–872
Contents lists available at ScienceDirect
Chemical Engineering Journal
journal homepage: www.elsevier.com/locate/cej
Combined supercritical and subcritical conversion of cellulose for fermentable
hexose production in a flow reaction system
Yan Zhaoa, Hong-Tao Wanga,∗, Wen-Jing Lua, Hao Wangb
aDepartment of Environmental Science and Engineering, Tsinghua University, Beijing 100084, China
bCollege of Environmental Science and Engineering, Beijing Forestry University, Beijing 100083, China
article info
Article history:
Received 25 September 2010
Received in revised form
13 November 2010
Accepted 16 November 2010
Keywords:
Combined supercritical and subcritical
hydrolysis
Flow reaction system
Cellulose
Fermentable hexose production
Hydrothermal technology
abstract
Using research on a batch system as basis, a flow reactor was designed and applied in the combined
supercritical and subcritical hydrolysis of cellulose for fermentable hexose production. The results show
that when the supercritical parameters were maintained, the hexose yield first increased with the rise
in subcritical temperature, and then decreased after the maximum yield was obtained. This maximum
yield of fermentable hexoses from cellulose was 31.5% ±1.4%, which was obtained under the following
conditions: cellulose concentration of 3.53 ±0.24 g L−1, supercritical temperature of 380 ◦C, supercrit-
ical reaction time of 9.70 ±0.66 s, subcritical temperature of 240 ◦C, and subcritical reaction time of
48.49 ±3.31 s. The appropriate ranges of cellulose concentration (around 3.5 g L−1) and reaction time
(9–10 s for supercritical process and 45–50 s for subcritical process), which depended on the flows of
water and material sludge, were also crucial in obtaining a high hexose yield. Compared with the batch
system, the flow reaction system can yield a reasonable amount of hexose from cellulose hydrolysis and
proved to be considerably promising for practical applications, especially for combined supercritical and
subcritical technology on lignocellulosic resources.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
The current pressure to adopt more efficient energy consump-
tion methods stems from the high price of fossil fuel, energy
security, and environmental concerns. Renewable energy is an
acknowledged solution to this problem [1]. Biomass has elicited
increasing attention because it is renewable, inexpensive, and read-
ily available worldwide, thereby guaranteeing a high level of energy
security. It is also a carbon-neutral resource and does not cause a
net increase in green-house gases [2]. Therefore, because of the
polymeric and crystalline structure of lignocellulose, many pre-
treatment, hydrolysis, and fermentation technologies have been
investigated and developed to convert lignocellulose into energy
or fuel, including bioethanol [3–6].
Hydrothermal technologies have proven promising in ligno-
cellulose conversion because of their high efficiency in dissolving
and hydrolyzing cellulose [7–9]. Supercritical water technology has
obvious advantages in lignin separation and cellulose hydrolysis,
which is attributed to its high dissolution and catalyzing capac-
ity [10,11]. However, considering the high decomposition rate of
hydrolyzates in supercritical water, subcritical water was intro-
duced for the hydrolysis of dissolved cellulose [12]. Combined
∗Corresponding author. Tel.: +86 10 6277 3438.
E-mail address: htwang@mail.tsinghua.edu.cn (H.-T. Wang).
supercritical and subcritical technology has been suggested and
proven efficient for hexose production from lignocelluloses. In this
combined approach, cellulose in biomass is first dissolved and
hydrolyzed in supercritical water to produce oligosaccharides, to
which subcritical water is then applied for hydrolysis into fer-
mentable hexoses [13,14].
In our previous work on a batch reaction system, the feasibility
and reaction mechanism of the combined supercritical and subcrit-
ical hydrolysis of cellulose and lignocellulosic waste were studied
and demonstrated [13,15,16]. The batch system cannot be used for
practical purposes considering its non-continuity and high energy
costs. Therefore, in this study, a flow reaction system was designed
and investigated. The combined supercritical and subcritical
hydrolysis of cellulose using the flow reaction system was exam-
ined, along with the effects of subcritical temperature, cellulose
concentration, and reaction time on the final hexose production.
The relatively optimal parameters obtained can be valuable for the
conversion of lignocellulosic waste, such as in the conversion of
corn stalks into fermentable hexoses, using the flow system.
2. Materials and methods
2.1. Reagents and analysis methods
Microcrystalline cellulose powder, the substrate used for the
combined supercritical and subcritical hydrolysis, was obtained
1385-8947/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2010.11.058

Author's personal copy
Y. Zhao et al. / Chemical Engineering Journal 166 (2011) 868–872 869
3
P
4
T
T
T
5
68
9
P
P
T
7
12
10
11
12 12
12
13
13 13 13
14
14
15
Preheating subsystem Supercritical reaction subsystem
Feed-in subsystem
Subcritical reaction subsystem
Temp. control subsystem
Fig. 1. Flow reaction system for the combined supercritical and subcritical process. (1) Water tank, (2) material sludge tank, (3) water pump, (4) material sludge pump,
(5) preheater, (6) supercritical reactor, (7) primary water cooler, (8) subcritical reactor, (9) final water cooler, (10) product collector, (11) temperature control system, (12)
manometer, (13) thermoelement, (14) reducing valve, and (15) safety valve.
from Beijing Fengli Jingqiu Commerce and Trade Co., Ltd.
(Beijing, China). The liquid hydrolysis products were analyzed
by high performance liquid chromatography (HPLC, Shimadzu,
LC-10ADvp, RID-10A, Japan) using a sugar column (Shodex, Sugar
KS-801, Japan). The products were analyzed under the following
conditions: 50 ◦C, 1.0 mL min−1, and 3.0 MPa. The standard sub-
stances for HPLC analysis, such as cellopentaose, cellotetraose,
cellotriose, cellobiose, glucose, xylose, fructose, erythrose, glyc-
eraldehyde, 1,6-anhydroglucose, dihydroxyacetone, and 5-HMF,
were from Sigma–Aldrich Inc. (Missouri, USA).
2.2. Flow reaction system for the combined supercritical and
subcritical hydrolysis
The structure of the flow reaction system is shown in Fig. 1. The
main body of the flow system is made of stainless steel 316 and
has five subsystems, namely, the feed-in, preheating, supercritical
reaction, subcritical reaction, and temperature control subsystems.
The feed-in subsystem comprises a water tank and pump, as well
as a material sludge tank and pump for storing and feeding water
and material sludge through the flow pipes into the respective
preheating and supercritical reaction subsystems. The preheating
subsystem is composed of a preheater with a coil pipe inside and a
salt bath outside. The salt bath is filled with NaNO3and KNO3(1:1,
w/w), and it can heat the water in the coil pipe to a temperature
ranging from 260 to 500 ◦C. The supercritical reaction subsystem
includes a supercritical reactor (= 4 mm, l= 800 mm, V=10mL)
with a salt bath outside similar to that of the preheating subsys-
tem, and a primary water cooler for cooling the products from
the supercritical reaction. The subcritical reaction system includes
a subcritical reactor (= 7 mm, l= 1300 mm, V=50 mL) with an
electric heater outside (providing a temperature of 100–350 ◦C),
and a final water cooler for cooling the products from subcritical
reaction. The temperature controlling subsystem can monitor and
control the temperatures in the preheater, supercritical reactor,
water cooler, and subcritical reactor through four thermoelements.
Moreover, the supercritical and subcritical reactors can be modi-
fied into reactors of different lengths to provide different volumes.
The reaction time is considered the residence time in which the
mixture flows through the supercritical or subcritical reactor.
However, due to the heating time (less than 1 s) of the mixture
from the temperature after preheating or primary cooling to the
chosen supercritical or subcritical temperature, the real reaction
time is actually slightly shorter than the residence time. The reac-
tion pressures inside the reactors are measured using manometers
and adjusted by the reducing valves.
2.3. Experimental design
In the combined supercritical and subcritical hydrolysis exper-
iments on cellulose, deionized water was pumped from the water
tank into the preheater maintained at 370 ◦C. Prepared material
sludge with a cellulose concentration of 10 gL−1was pumped from
the material sludge tank into the supercritical reactor immediately
after being mixed with the preheated water. The products from the
supercritical reactor were cooled by the primary water cooler to
stop supercritical reaction and then transferred into the subcritical
reactor to undergo further hydrolysis under subcritical conditions.
Finally, the products were collected by a product collector. Con-
sidering the interaction of the operational parameters including
temperature, pressure, and flow, the reaction system was deemed
to achieve steady state only when the measured temperatures for
the preheating, supercritical and subcritical reactions reached the
chosen temperatures (with an accuracy of ±1◦C), and at the same
time, the measured pressures for supercritical and subcritical reac-
tions reached the chosen pressures (with an accuracy of ±1 MPa).
The stable experimental period was defined as the period during
which the system maintained steady state, and all the operational
parameters and samples mentioned in this paper were obtained
during the stable experimental period. The flows of the two pumps
were measured based on the water and material sludge consump-
tion during the stable experimental period, and adjusted to provide
different levels of cellulose concentration. In accordance with our
previous work on the supercritical hydrolysis of cellulose [15],a
temperature of 380 ◦C was applied for supercritical reaction, and
the cellulose concentration of the mixture was adjusted in the range
of 3.5–4.5 g L−1. Four subcritical reaction temperatures, i.e., 210,
240, 270, and 300 ◦C, were investigated. The supercritical and sub-
critical reaction times were calculated according to Eq. (1). Three
parallel samples were collected for HPLC analysis at equal intervals

Author's personal copy
870 Y. Zhao et al. / Chemical Engineering Journal 166 (2011) 868–872
Table 1
Oligosaccharide yields from the supercritical hydrolysis of cellulose at 380 ◦C.
No. Cellulose
concentration
(g L−1)
Pressure
(MPa)
Reaction
time (s)
Oligosaccharide
yield (%)
1 3.29 25 ±1 8.45 35.9
2 3.64 24 ±1 8.44 31.5
3 3.66 25 ±1 7.34 28.9
4 5.47 25 ±1 6.56 24.3
5 5.32 25 ±1 6.44 17.9
6 6.34 25 ±1 9.73 16.5
7 4.62 25 ±1 4.99 16.2
8 1.72 25 ±1 15.64 –
(e.g., 5, 10, and 15 min) during each stable experimental period,
which lasted for at least 15 min. The yield of each component was
the average value of those in the parallel samples with an error bar.
t=V
Q(1)
where Qstands for the flow of the mixture (mL s−1), and Vstands
for the volumes of the supercritical or subcritical reactor, which are
10 and 50 mL, respectively.
3. Results and discussion
3.1. Oligosaccharide production from the supercritical hydrolysis
of cellulose at 380 ◦C
The hermetic structure of the flow reaction system makes
the collection of intermediate products from the supercritical
reactor difficult when performing combined experiments. There-
fore, considering that a temperature marginally higher than the
critical point is more suitable for accumulating oligosaccharides
and enhancing the stability of oligosaccharide production in super-
critical reactions (as determined in our previous work on the
supercritical hydrolysis of cellulose) [15], 380 ◦C was chosen as the
optimal and fixed supercritical temperature. Furthermore, when
the cellulose concentrations were in the range of 3–4.5 g L−1and
the reaction times in the range of 7–10 s, the oligosaccharide yields
in the supercritical reaction can normally reach over 30% to around
40%, which are relatively high compared with those obtained under
other conditions. The results are shown in Table 1. A yield of
35.9 ±1.5% oligosaccharide was accumulated when the cellulose
concentration was at 3.29 g L−1and the reaction time was 8.45 s.
These results served as an important guide for supercritical param-
eter control in the combined experiments conducted in this study,
although the intermediate products obtained after the supercritical
reaction were not analyzed because of the flow-type structure.
3.2. Effect of temperature on subcritical reaction and hexose
production
Four temperature levels (210, 240, 270, and 300 ◦C) were inves-
tigated to reveal the effect of temperature on the subcritical
reaction for the combined hydrolysis of cellulose. Under the same
parameters as those in the supercritical reaction, subcritical reac-
tion, especially hexose production, can be influenced to a great
extent by the reaction temperature. Fig. 2 presents the oligosaccha-
ride and hexose yields at different subcritical temperatures under
the following operational parameters: cellulose concentration of
3.53 ±0.24 g L−1, supercritical temperature of 379 ±1◦C, supercrit-
ical reaction time of 9.70 ±0.66 s, and subcritical reaction time of
48.49 ±3.31 s.
Fig. 2 shows that after a subcritical reaction lasting around 48 s
at 210 ◦C, approximately 13.1% of the oligosaccharides produced
Fig. 2. Effect of subcritical temperature on the oligosaccharide, glucose, and fructose
yields.
in the supercritical reaction remained in the final liquid prod-
uct, whereas the glucose and fructose yields reached 21.6% and
2.2%, respectively. With the increment of subcritical temperature,
the hydrolysis rate of the oligosaccharides increased accordingly,
inducing the higher consumption of oligosaccharides during sub-
critical reactions. However, the yields of hexoses, including glucose
and fructose, increased at first and then decreased after a maxi-
mum yield was obtained. This is because the hexoses were further
decomposed as they were being produced from oligosaccharides.
Moreover, the hexoses decomposed more rapidly at higher temper-
atures, a finding that has been proven and analyzed in our previous
work on the batch system. The hexose yields were 23.8%, 31.5%,
26.1%, and 22.6% at 210, 240, 270, and 300 ◦C, respectively. There-
fore, 240 ◦C was determined as the optimal subcritical temperature
of the combined process for cellulose conversion.
3.3. Influence of cellulose concentration and reaction time on
hexose yields
Due to the structural integrity of the flow reaction system, the
variety of cellulose concentrations and reaction times influenced
both supercritical and subcritical reactions in the combined experi-
ments. In fact, the reaction time depended on the flows of water and
material sludge when the reactor volumes were fixed. Therefore,
four levels of cellulose concentrations in the range of 3.5–4.5 g L−1
were adjusted by water and material sludge flows and investigated
to reveal their effects on hexose production. The corresponding four
levels of operational parameters (groups A–D in terms of cellulose
concentration), including information on water and material sludge
flow, are shown in Table 2.
Fig. 3 represents the final yields of hexoses at different sub-
critical temperatures for each experimental group. The initial
increase and subsequent decrease in the yields of hexoses after a
maximum yield was reached, along with an increase in the sub-
critical temperature, can be observed in each experimental group.
Furthermore, the cellulose concentration and reaction time con-
siderably influenced hexose production. For example, when the
subcritical temperature was 240 ◦C, the hexose yield obtained
at a cellulose concentration of 3.53 ±0.24 g L−1and supercritical
reaction time of 9.70 ±0.66 s (group A) was 31.5% ±1.4%, which
was larger than those obtained at higher cellulose concentrations,
including 25.1% ±1.1% obtained at a cellulose concentration of

Author's personal copy
Y. Zhao et al. / Chemical Engineering Journal 166 (2011) 868–872 871
Table 2
Operational parameters for the investigation of the effects of cellulose concentration and reaction time on combined hydrolysis.
Experimental group A B C D
Water flow (mL s−1) 0.67 ±0.06 0.90 ±0.04 0.75 ±0.04 0.56 ±0.04
Material sludge flow (mL s−1) 0.36 ±0.03 0.60 ±0.05 0.55 ±0.03 0.43 ±0.05
Cellulose concentration in mixture (g L−1) 3.53 ±0.24 4.00 ±0.24 4.22 ±0.17 4.38 ±0.35
Supercritical temperature (◦C) 379 ±1 380 ±0 379 ±1 379 ±1
Supercritical pressure (MPa) 24.0 ±0.5 23.5 ±0.5 23.5 ±0.5 23.5 ±0.5
Supercritical reaction time (s) 9.70 ±0.66 6.65 ±0.33 7.72 ±0.31 10.11 ±0.59
Subcritical pressure (MPa) 9.0 ±0.0 9.0 ±1.0 9.0 ±0.0 8.5 ±0.5
Subcritical reaction time (s) 48.49 ±3.31 33.24 ±1.64 38.60 ±1.54 50.55 ±2.94
Fig. 3. Hexose yields produced from cellulose at different subcritical temperatures
for each experimental group.
4.00 ±0.24 g L−1, 22.0% ±1.1% at 4.22 ±0.17 g L−1, and 19.4% ±0.5%
at 4.38 ±0.35 g L−1. This reveals that in the 3.5–4.5 g L−1range, a
relatively low cellulose concentration can result in high yields of
hexoses when the reaction times are in reasonable ranges, such as
7–10 s for supercritical reaction and 40–50 s for subcritical reaction.
Conversely, an excessively short or long reaction time may cause
insufficient hydrolysis or immoderate decomposition, yielding rel-
atively low amounts of hexoses. In this research, the operational
conditions, including a cellulose concentration around 3.5 g L−1,
supercritical temperature of 380 ◦C, supercritical reaction time of
9–10 s, subcritical temperature of 240 ◦C, and subcritical reaction
time of 45–50 s, were thus determined as the optimal parameters
for the combined supercritical and subcritical process of cellulose
conversion.
3.4. Comparison of hexose production in the batch reaction and
flow reaction systems
The results obtained using the flow reaction system were com-
pared with those obtained with the batch reaction system carried
out in our previous work. The batch system was composed of two
parallel reactors (5 mL, stainless steel 316), two salt baths for the
supercritical and subcritical reactions respectively (providing tem-
peratures of 260–500 ◦C), an ice–water cooler, and a temperature
control subsystem. During the batch experiments, 60 mg cellulose
and 2.5 mL deionized water were mixed and dispensed into each
parallel reactor, and then hydrolyzed at 380 ◦C, 16 s and 280 ◦C,
44 s for the supercritical and subcritical reactions respectively,
which had been determined as the optimum conditions for cel-
lulose hydrolysis [13]. The corresponding operational parameters
and hexose yields are presented in Table 3. The maximum hexose
yield obtained with the flow reaction system is comparable to but
slightly lower than that obtained with the batch system. However,
the maximum oligosaccharide yield from the supercritical hydrol-
Table 3
Maximum yields of hexoses and the corresponding operational parameters in the
batch and flow systems.
Parameter Batch reaction
system
Flow reaction
system
Cellulose concentration (g·L−1) 24 3.53 ±0.24
Supercritical temperature (◦C) 380 379 ±1
Supercritical pressure (MPa) 25 24.0 ±0.5
Supercritical reaction time (s) 16 9.70 ±0.66
Subcritical temperature (◦C) 280 240
Subcritical pressure (MPa) 10 9.0 ±0.0
Subcritical reaction time (s) 44 48.49 ±3.31
Maximum hexose yield (%) 39.5 31.5% ±1.4%
ysis of cellulose in the flow system is almost equivalent to that in
the batch system. This probably stems from the fact that the oper-
ational parameters in the flow reaction system, such as reaction
time, cannot be adjusted separately for supercritical and subcriti-
cal reactions because of the fixed ratio of the reactor volumes. This
presents difficulties in the optimization of the combined process. In
contrast, supercritical and subcritical reactions can be performed
separately in the batch system, so that the reaction times can be
easily controlled as the combined process is optimized. Further-
more, the cellulose concentration in the flow system is much lower
than that in the batch system because of the problem of fluidity.
Despite the considerable contributions of the batch system to
theoretical research on the hydrothermal conversion of biomass
and the relatively higher amount of hexose it yields through the
combined hydrolysis of cellulose, it cannot be used in practical
applications because of the low efficiency of batch operation. The
flow reaction system is much more promising for hydrothermal
biomass conversion because of the high efficiency generated by
its continuous flow structure. It is especially suitable for com-
bined supercritical and subcritical technology on lignocellulosic
resources.
4. Conclusions
This paper examined the combined supercritical and subcritical
hydrolysis of cellulose in the flow reaction system. On the basis of
the optimal supercritical parameters obtained in a previous study,
the effects of subcritical temperature, cellulose concentration,
and reaction time on final hexose production were investigated.
When all other parameters were maintained, a maximum hexose
yield was obtained during the increase in subcritical tempera-
ture. Appropriate ranges of cellulose concentration and reaction
time, which depended on the flows of water and material sludge,
were also crucial in obtaining high hexose yields. The flow reac-
tion system can yield reasonable amounts of hexose from cellulose
hydrolysis compared with the batch system, and it has higher
potential for use in biomass conversion. Experiments on combined
supercritical and subcritical processes for the conversion of ligno-
cellulosic waste, such as corn stalks, are currently underway.

Author's personal copy
872 Y. Zhao et al. / Chemical Engineering Journal 166 (2011) 868–872
Acknowledgements
This work was supported by the National High-Tech Research
and Development Program of China (No. 2006AA10Z422), the
China Postdoctoral Science Foundation (No. 20100470018), and
the Science and Technology Innovation Program of Beijing Forestry
University (No. BLYX200911).
References
[1] J. Ruffell, B. Levie, S. Helle, S. Duff, Pretreatment and enzymatic hydroly-
sis of recovered fibre for ethanol production, Bioresour. Technol. 101 (2010)
2267–2272.
[2] M. Tymchyshyn, C.C. Xu, Liquefaction of bio-mass in hot-compressed water
for the production of phenolic compounds, Bioresour. Technol. 101 (2010)
2483–2490.
[3] B.A. Faga, M.R. Wilkins, I.M. Banat, Ethanol production through simultaneous
saccharification and fermentation of switchgrass using Saccharomyces cere-
visiae D5A and thermotolerant Kluyveromyces marxianus IMB strains, Bioresour.
Technol. 101 (2010) 2273–2279.
[4] T. Nakata, H. Miyafuji, S. Saka, Ethanol production with b-xylosidase,
xylose isomerase, and Saccharomyces cerevisiae from the hydrolysate of
Japanese beech after hot-compressed water treatment, J. Wood Sci. 55 (2009)
289–294.
[5] P. Kaparaju, M. Serrano, A.B. Thomsen, P. Kongjan, I. Angelidaki, Bioethanol,
biohydrogen and biogas production from wheat straw in a biorefinery concept,
Bioresour. Technol. 100 (9) (2009) 2562–2568.
[6] K. Ohgren, R. Bura, J. Saddler, G. Zacchi, Effect of hemicellulose and lignin
removal on enzymatic hydrolysis of steam pretreated corn stover, Bioresour.
Technol. 98 (13) (2007) 2503–2510.
[7] M.J. Díaz, C. Cara, E. Ruiz, I. Romero, M. Moya, E. Castro, Hydrothermal pre-
treatment of rapeseed straw, Bioresour. Technol. 101 (2010) 2428–2435.
[8] A. Sına˘
g, S. Gülbaya, B. Uskana, M. Güllü, Comparative studies of intermediates
produced from hydrothermal treatments of sawdust and cellulose, J. Supercrit.
Fluids 50 (2) (2009) 121–127.
[9] Y. Yu, X. Lou, H. Wu, Some recent advances in hydrolysis of biomass in hot-
compressed water and its comparisons with other hydrolysis methods, Energy
Fuels 22 (2008) 46–60.
[10] F.L.P. Resende, M.E. Neff, P.E. Savage, Noncatalytic gasification of cellulose in
supercritical water, Energy Fuels 21 (2007) 3637–3643.
[11] Y. Matsumura, S. Takami, M. Umetsu, T. Adschiri, M. Sasaki, S. Ohara, K. Okuda,
Supercritical water treatment of biomass for energy and material recovery,
Combust. Sci. Technol. 178 (2006) 509–536.
[12] K. Ehara, S. Saka, A comparative study on chemical conversion of cellulose
between the batch-type and flow-type systems in supercritical water, Cellulose
9 (2002) 301–311.
[13] Y. Zhao, W.J. Lu, H.T. Wang, D. Li, Combined supercritical and subcritical process
for cellulose hydrolysis to fermentable hexoses, Environ. Sci. Technol. 43 (5)
(2009) 1565–1570.
[14] K. Ehara, S. Saka, Decomposition behavior of cellulose in supercritical water,
subcritical water, and their combined treatments, J. Wood Sci. 51 (2005)
148–153.
[15] Y. Zhao, W.J. Lu, H.T. Wang, Supercritical hydrolysis of cellulose for oligosac-
charide production in combined technology, Chem. Eng. J. 150 (2–3) (2009)
411–417.
[16] Y. Zhao, W.J. Lu, H.T. Wang, J.L. Yang, Fermentable hexose production from corn
stalks and wheat straw with combined supercritical and subcritical hydrother-
mal technology, Bioresour. Technol. 100 (23) (2009) 5884–5889.