Bioresource Technology 214 (2016) 496503

Contents lists available at ScienceDirect

Bioresource Technology
j o u r n a l h o m e p a g e : w w w.e l s e v i e r . c o m / l o c a t e / b i o r t e c h

Multi-level dissolution and hydrolysis of lignocellulosic waste with a

semi-flow hydrothermal system
Yan Zhao , Haobo Tan, Yingjie Xu, Lei Zou
School of Environment, Beijing Normal University, Beijing 100875, China

h i g h l i g h t s
Multi-level hydrothermal process was studied to optimize lignocellulosic conversion.
Semi-flow system offers separate and long residence times to undissolved materials.
Dissolution of amorphous compounds at 195200 C without affecting cellulose.
Cellulose dissolution ratios higher than 75% are achieved at 245250 C.
Soluble sugars can be further obtained after the hydrolysis of dissolved cellulose.

a r t i c l e

i n f o

Article history:
Received 25 March 2016
Received in revised form 28 April 2016
Accepted 29 April 2016
Available online 2 May 2016
Keywords:
Cellulose hydrolysis
Hydrothermal process
Lignocellulosic waste
Multi-level dissolution
Semi-flow reaction system

a b s t r a c t
The hydrothermal process is efficient in lignocellulosic conversion and is beneficial to potential
bioethanol production. In batch- and flow-type processes, concurrent dissolution and hydrolysis of
lignocellulose result in product loss and inhibitory intermediates. Therefore, multi-level hydrothermal
conversion of corn stalks was implemented with a semi-flow system to provide different residence times
to undissolved compounds and facilitate dissolution or hydrolysis at respective optimal conditions. Firststage dissolution dissolved amorphous hemicellulose and lignin at 195200 C. Xylan, acid soluble lignin,
and part of Klason lignin were dissolved without affecting glucan. In second-stage dissolution, the
crystallinity of the undissolved materials suddenly decreased at 245250 C. The cellulose dissolution
ratio was higher than 75%. Soluble sugars were obtained after the hydrolysis of dissolved cellulose at
280 C. The results provide significant information on the multi-level hydrothermal process and its
potential applications for recovering valuable chemicals from lignocellulosic waste.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Biofuel production from biomass conversion has received
increasing attention as a promising renewable energy that can
overcome the disadvantages of fossil fuel, such as climate change
and limited availability. As the most commonly produced biofuel
in the world, bio-ethanol is traditionally obtained from foodbased sources, such as agricultural crops with high contents of
carbohydrates, which are facing the problem of competition for
food requirement (Bommarius et al., 2008). As a result, interest
in second-generation processes utilizing low-cost feedstock, such
as lignocellulosic materials, has emerged (Sittijunda et al., 2013).
Cellulose, one of the main components in lignocellulose, can be
converted into fermentable saccharides and fermented into

Corresponding author.
E-mail address: yanzhao@bnu.edu.cn (Y. Zhao).

http://dx.doi.org/10.1016/j.biortech.2016.04.135
0960-8524/ 2016 Elsevier Ltd. All rights reserved.

ethanol by applying chemical or biological processes (Abdullah

et al., 2014).
Given the complex structure of cellulose combined with hemicellulose and the fact that lignin restricts its contact with reaction
media or enzymes, converting lignocellulosic biomass into saccharides or other chemicals requires a pretreatment or hydrolysis process to overcome its high resistance to dissolution and hydrolysis
(Li et al., 2014; Xiao et al., 2014). Among many options, hydrothermal treatment by using supercritical or subcritical water has
gained momentum as a promising process (Kruse, 2008). Benefitting from the high dissolving ability of hot compressed water
+
and catalysis of ionized H , hydrothermal treatment offers the
major advantages of rapid reaction, no chemical reagents, and efficient cost in lignocellulosic conversion (Xiao et al., 2014).
Hydrothermal processes have been widely studied for lignocellulosic pretreatment (Zhao et al., 2009a), cellulose hydrolysis
(Tolonen et al., 2013), and fractionation for biorefineries from

49

Y. Zhao
Y. Zhao
et al.
et/ al.
Bioresource
/ Bioresource
Technology
Technology
214 214
(2016)
(2016)
496503
496503

lignocellulose (Gullon et al., 2012). Several studies have also presented hydrothermal technologies for lignocellulosic liquefaction,
gasification, and carbonization (Resende et al., 2007; Funke and
Ziegler, 2010; Tymchyshyn and Xu, 2010; Hoekman et al., 2011).
Several researchers previously reported the hydrothermal
dissolution and hydrolysis of lignocelluloses with relatively high
temperatures, especially with the involvement of supercritical
water (above 374
C and 22.1 MPa) and subcritical water
(normally above 300 C and 10 MPa) (L and Saka, 2012; Zhao et
al., 2012). However, many of them found that monosaccharides
decompose rapidly into unfermentable products during cellulose
dissolution and hydrolysis, resulting
in
the loss
of
fermentable sugars (Kumar and Gupta, 2008; Zhao et al., 2015).
This result is mainly attributed to the synchronous dissolution
and hydrolysis of cellu- lose, especially in the conventional batchor flow-type hydrother- mal
process for
lignocellulosic
conversion
(Zhao
et al.,
2012). Moreover, synchronous
dissolution and hydrolysis of
lignocellu- lose
at
high
temperatures result in the formation of various intermediates,
such as
phenolic and furfural, which are
derived from
hemicellulose and lignin degradation and inhibit cellulose
hydrolysis or subsequent yeast fermentation (Qing et al., 2010;
Kont et al., 2013).
To avoid the disadvantages caused by product degradation,
considerable attention has been devoted to the reaction mechanisms of lignocellulosic dissolution and hydrolysis during
hydrothermal treatment. According to previous studies, the
decomposition of lignin, hemicellulose, and cellulose can be
characterized by temperature as follows: hemicellulose mainly
decomposes between 200 C and 400 C, cellulose decomposes
at high temperatures of 300400 C, and lignin gradually decomposes between 180 C and 600 C (Libra et al., 2011). Several
researchers reported that all hemicellulose, 3560% of lignin,
and 422% of cellulose can be dissolved with flow-type
hydrothermal treatment at 200230 C, 34.5 MPa, and 15 min
(Mosier et al., 2005). Therefore, different treatment temperatures
can be utilized for hemicellulose and cellulose hydrolysis separately, given that cellulose is more resistant to hydrolysis than
hemicellulose (L et al., 2009). The multi-step hydrothermal process has been applied to convert biomass waste, such as rapeseed straw, and depolymerization of hemicellulose, cellulose,
and lignin has been implemented with a batch system at
210 C, 270 C, and 330 C, respectively (Pinkowska and Wolak,
2013). L and Saka also reported that the semi-flow hydrothermal system, which is superior to the batch system in the respect
of total saccharide production, provides the maximum saccharides from hemicellulose and cellulose in Japanese beech at
230 C and 270 C, respectively (L and Saka, 2010). However,
clear fundamental understanding of the process, especially
multi-level dissolution and hydrolysis with the semi-flow process, remains lacking (Kumar et al., 2010). In addition, no standard process conditions exist; as a result, each raw material
should be individually studied and optimized (Rostagno et al.,
2015).
To overcome the problem in the synchronous dissolution and
hydrolysis of lignocellulosic compounds, a semi-flow reaction system was
applied in this study to implement a multi-level
hydrothermal process for lignocellulosic conversion. The semiflow system keeps the undissolved materials in the dissolution
unit as stationary phase and discharges the dissolved compounds
from the system as mobile phase. This system structure offers
different residence times to undissolved and dissolved compounds
and thus avoids their insufficient dissolution and excessive
degradation, respectively. Furthermore, the multi-level operation
benefits the dissolution or hydrolysis of the major compounds in
their respec- tive
optimal
conditions. To
examine the
mechanisms of asyn- chronous dissolution and hydrolysis of
lignin, hemicellulose, and

49

cellulose, multi-level hydrothermal conversion of corn stalks was

investigated with the semi-flow system in this study from the perspectives of chemical structure and component analysis. The variation in
crystallinities, chemical bonds, compositions, and
hydrolysates of lignocellulose was
revealed with different
hydrothermal conditions to
enhance the understanding of
hydrothermal lignocellulosic conversion and reveal the promising
application of the multi-level semi-flow system to the recovery
of valuables from lignocellulosic waste.

2. Materials and methods

2.1. Raw materials and chemicals
Corn stalks, which were obtained from farms in the suburbs of
Beijing, China, were utilized as lignocellulosic substrate. The corn
stalks were mechanically crushed into powder and sieved with a
0.63 mm sieve prior to the experiments.
The standard substances for high-performance liquid chromatography (HPLC) analysis, including cellopentaose, cellotetraose, cellotriose, and erythrose, were purchased from Toronto
Research Chemicals Inc. (Canada). The other standard substances,
such as cellobiose, glucose, xylose, fructose, glyceraldehyde, 1,6anhydroglucose, and 5-HMF, were provided by SigmaAldrich
Inc. (MO, USA).

2.2. Experimental apparatus and designs

A semi-flow hydrothermal reaction system was specially
designed and manufactured for the sequential dissolution and
hydrolysis of lignocellulosic materials. A diagram of the apparatus is shown in Fig. 1. The entire system is made of 316L stainless steel, and its different units are connected by pipelines with
a diameter of 32 mm. Initially, a certain amount of raw materials
in powder was loaded as solid phase in the dissolution unit,
which was then sealed with the upstream and downstream pipelines. Subsequently, distilled water as liquid phase was pumped
into the preheating unit and heated to a set temperature and
then sent to the dissolution unit to dissolve the lignocellulosic
materials. For first-stage dissolution, the dissolved materials flowed with the liquid phase into the termination unit to stop the
reaction by water cooling. Meanwhile, the undissolved materials
remained in the dissolution unit for further reaction until a set
retention time. For the purpose of combining dissolution and
hydrolysis experiments, after first-stage dissolution, the dissolution unit was shifted to a high temperature for second-stage dissolution. The liquid phase carrying dissolved materials flowed
into the hydrolysis unit to hydrolyze the substances into small
fractions, such as monosaccharides. After the entire process,
the solid residues and liquid hydrolysates were collected and
further analyzed.
Before each round of experiments, 7.0 0.7 g of raw materials
were loaded in the dissolution unit. The pressures of the dissolution unit were maintained at 4.5 0.5 MPa for all the experimental
rounds. The solid retention times for undissolved materials in the
first- and second-stage dissolution were both 30 min, and the
hydraulic retention times for liquid phase in both stages were
6 min. The temperatures for the different experimental strategies
are listed in Table 1. The temperature of the hydrolysis unit was
maintained at 280 C according to our previous work on oligosaccharide hydrolysis (Zhao et al., 2009a). Two parallel experiments
were performed for all the experimental conditions, and the residual samples collected from the parallel experiments were mixed
for analysis purpose.

T
T

Fig. 1. Apparatus diagram of the semi-flow hydrothermal reaction system. 1. water tank; 2. metering pump; 3. preheating unit; 4. dissolution unit (equipped with filters);
5. hydrolysis unit; 6. termination unit (equipped with a water cooling jacket); 7. temperature controller; 8. check valve; 9. thermocouple; 10. counterbalance valve; 11.
pressure meter.

Table 1
Reaction temperatures for the different experimental strategies.
Experimental strategy

Unit

Tested temperature range

Temperature interval

First-stage dissolution

First-stage dissolution
Second-stage dissolution and hydrolysis

180215 C
Null

5 C
Null

Combining dissolution and hydrolysis

First-stage dissolution
Second-stage dissolution
Hydrolysis

205 C
230250 C
280 C

Null
5 C
Null

2.3. Analysis methods

74.68%

80
70.58%
70
CI (%)

The water content of the corn stalks was calculated by measuring weight loss after oven drying at 105 C for 24 h, and the ash
content was further calculated by measuring weight loss after furnace ignition at 600 C for 2 h. The compositions, including cellulose, hemicellulose, and lignin, of the raw materials and solid
residues after treatment were determined to be polymeric carbohydrates and acid insoluble and soluble lignin through highperformance anion exchange chromatography (ICS3000, Dionex,
USA) and UVvis spectroscopy following two-step acid hydrolysis

62.83%

73.04%

72.82%

64.63%
61.80% 61.93%

60

50

45.66%

40

according to the standard method recommended by the National

Renewable Energy Laboratory of the United States (Sluiter et al.,
2008).
The crystallinity of the raw materials and residues was studied
with an X-ray diffractometer (XPert PRO, PANalytical Co., Netherlands) operated at 40 kV and 40 mA with scan ranges of 2h = 550
and step size of 0.017 . The crystallinity index (CI) for cellulose I
was calculated according to Eq. (1) (Kumar et al., 2010).

CI I002 - I 18:5 =I002

Raw materials180 C

185 C

190 C

195 C

200 C

205 C

210 C

215 C

Fig. 2. Crystallinity indexes of raw corn stalks and undissolved materials in the
first-stage dissolution.

yacetone, and 5-HMF, were quantified through HPLC (Agilent Technologies 1260 Infinity, USA) with a special sugar column (Shodex,
Sugar KS-801, Japan) at 50 C and 1.0 mL min 1.

1
3. Results and discussion

where I0 0 2 is the peak intensity corresponding to the (0 0 2) lattice

plane of the cellulose molecule (2h = 22.7 ) and I18.5 is the peak
intensity observed at 2h = 18.5 that corresponds to the amorphous
portion.
Chemical bond changes were studied through Fourier
transform-infrared spectroscopy (FT-IR, Nicolet Nexus 670,
Thermo Electron, USA) using the KBr pellet method with a resolu1
tion ratio of 4 cm . The characteristic peaks of cellulose, lignin,
and hemicellulose were summarized from previous work (Zhao
et al., 2013; Jiang et al., 2014; Cao et al., 2015; Yoo et al., 2016),
and these values are listed in Table A.1.
The compounds in the liquid hydrolysate, including cellopentaose, cellotetraose, cellotriose, cellobiose, glucose, xylose,
fructose, erythrose, glyceraldehyde, 1,6-anhydroglucose, dihydrox-

3.1. Chemical reactions of lignocellulose in first-stage dissolution

3.1.1. Chemical structure changes of the undissolved lignocelluloses
According to the chemical structures of natural lignocelluloses,
such as corn stalks, the cellulose in the structure is mainly crystalline and relatively stable; hemicellulose is mainly amorphous
and is thus easier to be dissolved and hydrolyzed; while most lignin is considered difficult to dissolve. The CI of the raw corn stalks
utilized in this study was determined to be approximately 46%;
that of microcrystalline cellulose is around 73%. In the first-stage
dissolution, all the CIs of the undissolved materials increased to
above 60% after 30 min treatment. They increased at different
extents under different tested conditions (Fig. 2). In the
temperature

range of 180190 C, the variation in the CIs was unobvious,

although the CIs were significantly higher than that of raw corn
stalks. This result can be attributed to the partial dissolution of
the amorphous hemicellulose and lignin. Furthermore, a sudden
increase was observed between the data of 195 C and 200 C, suggesting that most structures of the hemicellulose and lignin were
destroyed in the semi-flow reaction system. Particularly, the CIs
of the undissolved materials above 200 C ( 73%) were close to
that of microcrystalline cellulose ( 74%), the highest level that
can be achieved in the first-stage dissolution. This increasing trend
of CIs has also been observed in other related studies with
hydrothermal technologies. However, because of the difference in
material characteristics, compositions and the treatment parameters, the results varied significantly. For example, a study on
hydrothermal pretreatment of bamboo reported a CI of 50.2%,
which increased from 45.5% after treatment at 200 C for
120 min; the increase was attributed to the removal of hemicellulose and the preservation of cellulose at such hydrothermal
condi- tions (Xiao et al., 2014).
Analysis of the chemical bonds through FT-IR was further performed to expound the structure variation. The characteristic
1
peaks of cellulose, such as 1372, and 897 cm , were completely
reserved in all the test conditions (Fig. B.1). These peaks represented CAH stretching and CAH deformation in crystalline and
amorphous cellulose. Moreover, the peaks at 1165 and 559 cm 1
presenting the structure of glycosidic bonds and pyranose were
also reserved. This finding indicates that cellulose structures were
hardly affected in the first-stage dissolution. By contrast, the char1
acteristic peaks of hemicellulose (C@O at 1736 cm
and CAO
1
stretching at 1040 cm ) changed significantly and even disappeared at temperatures higher than 195 C. This result indicates
the destructive effect on hemicellulose in the first-stage dissolution. With regard to lignin, the characteristic peaks of aromatic
skeletal vibration were basically reserved, although they
presented a reduced trend with the increase in temperature.
This phe- nomenon is different from that observed in a previous
study on bamboo with similar treatment; in the study, the
intensity of the same peaks increased slightly as
the
temperature increased because of the increment in lignin in
the residues (Xiao et al.,
2014). In the current study, the contents of lignin decreased with
the increase in temperature, as verified by the results of chemical
composition analysis (Table 2). Thus, the reserved peaks of lignin
were probably caused by retention of the stable structure of benzene rings in the undissolved residues after the first-stage
dissolu- tion. Hydrothermal conversion is also considered a
promising method for the exploitation of lignin by recovering
phenolic com- pounds as valuable chemical intermediates (Kang
et al., 2013). It can be achieved with a stepwise process by
separate collection.
3.1.2. Chemical composition changes in the undissolved lignocelluloses
Chemical composition analysis of the raw corn stalks and undissolved materials after treatment with the semi-flow system can

directly reveal the changes in cellulose, hemicellulose, and lignin

during the reactions to determine the dissolution laws of lignocellulosic waste at hydrothermal conditions. Table 2 shows the compositions of the undissolved materials after the first-stage
dissolution, including cellulose content reflected by glucan, hemicellulose content reflected by arabinose, galactose, and xylan, and
lignin content reflected by Klason lignin and acid soluble lignin.
The sums of all the contents varied from 73% to 88% and were
not equal to 100% because of the existence of protein, ash, and several other decomposed products that cannot be determined by the
applied analysis method.
In the raw corn stalks, hemicellulose and lignin mainly consisted of xylan (86.6%) and Klason lignin (87.2%), respectively. With
the increase in temperature in the first-stage dissolution, arabinose
and galactose decomposed rapidly, and the relative contents of
xylan decreased gradually until they almost completely disappeared. Different from the nearly eliminated acid soluble lignin,
the contents of Klason lignin decreased significantly with the
increase in temperature; a relative content (in terms of treated
materials) of approximately 10% remained in the undissolved residues. This result supports the FT-IR analysis. However, the relative
contents of glucan increased significantly after the first-stage dissolution and changed insignificantly as the temperature increased.
This result indicates that in the first-stage dissolution with the
semi-flow hydrothermal system, the major reactions are the complete dissolution of hemicellulose and partial dissolution of lignin.
The reaction of glucan dissolution can hardly occur.
Because the chemical composition analysis only provided relative contents, the real dissolved amounts and the dissolution ratios
of the components can hardly be obtained from the data in Table 2.
To present the absolute contents of the components and the variation in the compositions intuitively, the data in Table 2 were normalized under the assumption that the dissolution of cellulose or
glucan is negligible, and its contents remain unchanged in the
first-stage dissolution. The assumption is reasonable given the
results on both chemical structures and compositions. In addition,
it is supported by related reports indicating that cellulose is not
dissolved or hydrolyzed at temperatures below 230 C, except for
trace amounts of paracrystalline cellulose (L et al., 2009; Xiao
et al., 2014). Therefore, all the absolute contents (in terms of raw
materials) of cellulose (glucan) in the undissolved materials were
defined as 35.74% according to that in the raw corn stalks. On this
basis, the absolute contents of hemicellulose and lignin in the
undissolved materials (shown in Table 3) were calculated according to the relative contents listed in Table 2. The variation in the
compositions and the corresponding dissolution ratios of the components can further be calculated by comparing with the data of
raw materials.
Compared with those in the raw corn stalks, hemicellulose and
lignin were heavily dissolved or decomposed at temperatures
ranging from 180 C to 200 C. Specifically, in the undissolved
materials after the first-stage dissolution, the lignin content

Table 2
Chemical compositions of raw corn stalks and undissolved materials in the first-stage dissolution (%).
Temperature

Glucan

Arabinose

Galactose

Xylan

Klason lignin

Acid soluble lignin

Sum

Raw materials
180 C
185 C
190 C
195 C
200 C
205 C
210 C
215 C

35.74
63.94
63.71
66.29
67.95
65.77
64.91
74.30
65.88

1.98
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.

0.63
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.

16.91
4.24
5.49
6.84
2.67
2.76
1.17
1.06
1.00

15.75
16.67
14.02
13.90
12.16
12.99
10.61
8.26
7.67

2.31
0.66
0.65
0.87
N.D.
0.67
N.D.
N.D.
N.D.

73.32
85.52
83.87
87.90
82.77
82.19
76.69
83.62
74.55

N.D.: not detected.

Table 3

temperatures lower than 245 C. In contrast, when the tempera-

undissolved materials in the first-stage dissolution (% of raw materials).

Temperature

Cellulose

Hemicellulose

Lignin

Sum

35.74
35.74
35.74
35.74
35.74
35.74
35.74
35.74

19.52
2.37
3.08
3.69
1.40
1.50
0.64
0.51

18.06
9.69
8.23
7.96
6.40
7.42
5.84
3.97

73.32
47.80
47.05
47.39
43.54
44.66
42.23
40.22
4.16

Raw materials
180 C
185 C
190 C
195 C
200 C
205 C
210 C
215 C
40.44

35.74

0.54

decreased significantly with the increase in temperature, but

approximately 2530% of lignin remained at temperatures above
205 C. Meanwhile, the hemicellulose contents decreased sharply
from 19.52% to 2.37% even at the lowest tested temperature and
were close to zero at temperatures above 205 C. These quantitative results verify the analysis of chemical structures with CI and
FT-IR. Furthermore, the compositions of the undissolved materials
varied insignificantly when the temperature was above 205 C,
suggesting that 205 C is sufficient for the first-stage dissolution
of corn stalks. Therefore, 205 C was applied as the fixed temperature in the first stage in the following experiments combining dissolution and hydrolysis with the semi-flow hydrothermal reaction
system.
3.2. Chemical reactions of lignocellulosic residues in second-stage
dissolution
3.2.1. Chemical structure changes in the undissolved lignocelluloses
With the temperature of the first-stage dissolution fixed at
205 C, experiments on combining dissolution and hydrolysis were
conducted according to the strategy shown in Table 1. The CIs from
XRD analysis present the chemical structure changes in the
undis- solved materials after combining dissolution and
hydrolysis (Fig. 3). According to the data in Fig. 2, the CI of
undissolved mate- rials after the first-stage dissolution at 205 C
was approximately
73%, which is close to that of pure microcrystalline cellulose. By
contrast, the CIs of the undissolved materials after combining dissolution and hydrolysis decreased from 73% when the temperature
increased from 230 C to 250 C, suggesting the gradually
enhanced dissolution of crystalline cellulose. However, the CIs
obtained at temperatures below 245 C were still approximately
70%. Considering that the hydrothermal treatment was performed
for 30 min at 205 C and another 30 min at a temperature in the
range of 230245 C, this result indicates that the crystalline
struc- ture of the undissolved materials was
not
affected
significantly by

80
73.84%

72.90%
69.79%

CI (%)

70

68.92%

the undissolved materials decreased to 36.33%, which was even

lower than that of the raw corn stalks by approximately 10%. This
result is attributed to the intensive dissolution of crystalline cellulose in the lignocellulose and indicates that the chemical structure
of the undissolved materials was also changed significantly.
The chemical bond variation in the undissolved materials after
combining dissolution and hydrolysis is shown in Fig. B.2 through
FT-IR spectrograms. Similar to those in Fig. B.1, the characteristic
1
peak of hemicellulose (1736 cm ) did not exist after the secondstage dissolution in all the tested conditions because almost all
the hemicellulose had been already destroyed in the first-stage dissolution. Although the characteristic peaks of aromatic skeletal
vibration in lignin (1602 cm 1) after the second-stage dissolution
attenuated with the increase in temperature, they can still be
observed in all the tested conditions because of the recalcitrant
structure of the benzene ring that attached to the undissolved
materials. Considering the solid retention times of over 60 min in
total, high temperatures are probably required to remove lignin
completely. However, the characteristic peaks of cellulose, including pyranose, CAH deformation, CAOAC vibration, and intermolecular H-bonds in cellulose, which attenuated after the second-stage
dissolution at temperatures below 245 C compared with those in
raw corn stalks, were hardly observed in the spectrogram obtained
at 250 C with a relatively small fluctuation. Together with the
results on CIs, these results verify that saltation on the chemical
structures of lignocellulose was achieved after the second-stage
dissolution at 250 C.
3.2.2. Chemical composition changes in the undissolved lignocelluloses
In the compositions of the undissolved materials after the
second-stage dissolution, arabinose, galactose, and acid soluble lignin were not detected in all the tested conditions because they had
already been removed in the first-stage dissolution at 205 C. The
relative contents of xylan were also close to zero after the
second-stage dissolution, as shown in Table 4. By contrast, the relative contents of Klason lignin increased from 11.66% to 33.13%
when the temperature of the second-stage dissolution increased
from 230 C to 240 C. As the data of chemical compositions in
Table 4 were relative contents, rather than absolute contents
which could not increase during the experiments, the increase in
Klason lignin was mainly attributed to the decrease of glucan
and other undetermined components. According to the analysis
data, the rates of glucan dissolution at these conditions were speculated to be higher than the rates of lignin dissolution. For the
same reason, the further decrease in the relative contents of
Klason lignin when the temperatures increased to 245 C and 250
C indi- cates the destruction of Klason lignin during the
second-stage
dissolution.
The relative contents of glucan in Table 4 can hardly present the
reaction laws of cellulose in the tested conditions. Therefore, the
changes in cellulose contents were further estimated in terms of
the ratios of glucan and Klason lignin (shown in Table 4). The ratios

60

50

Table 4
Chemical compositions of raw corn stalks and undissolved materials in the secondstage dissolution.

45.66%
36.33%

Temperature

Glucan
(%)

Xylan
(%)

Klason
lignin (%)

Sum
(%)

230 C

63.12

0.16

11.66

74.94

40

30
Raw materials 230 C

235 C

240 C

245 C

250 C

Fig. 3. Crystallinity indexes of raw corn stalks and undissolved materials in the
second-stage dissolution.

Ratio of glucan and

Klason lignin
5.41

235 C
240 C

60.50
52.57

0.31
0.39

18.56
33.13

79.37
86.09

3.26
1.59

245 C
250 C

61.59
74.33

0.79
N.D.

24.53
6.79

86.91
2.51
81.12
10.95

of glucan and Klason lignin presented a trend of initial decrease in

the range of 230240 C. Between 245 C and 250 C, the ratio of
glucan and Klason lignin increased significantly, which was attributed to the higher rate of Klason lignin decomposition than that
of glucan dissolution. Therefore, the absolute contents of Klason
lig- nin was decreased and thus led to the decrease in relative
content of Klason lignin and the increase in that of glucan. Given
that the rates of glucan dissolution at 230 C, 235 C, and 240
C were higher than those of lignin dissolution, it was assumed
that the Klason lignin remained unchanged (minimum
dissolution rate) in the second-dissolution to estimate the
possible dissolution ratios of cellulose. According to the absolute
contents of lignin
(5.84%) obtained after the first-stage
dissolution at 205 C (see Table 3), the highest absolute contents
of cellulose were estimated from their ratios, and the lowest
dissolution ratios were further calcu- lated by using the cellulose
content of 35.74% in the raw corn stalks (Fig. 4). Considering the
significant destruction of Klason lignin at temperatures above 240
C, the assumption mentioned above is invalid under these
conditions. However, the dissolution ratios of cellulose at 245 C
and 250 C were not lower than those at
240 C. Therefore, the actual dissolution ratios of cellulose, which
were higher than 11.5%, 46.7%, 74.1% at 230 C, 235 C, and temperatures above 240 C, respectively, were in accordance with
the trend of the shadow area shown in Fig. 4. The dissolution
ratios of cellulose were supposed to increase rapidly with the
increase in temperature from 230 C to 240 C and gradually at
above 245 C. We thus conclude that temperatures of 240245 C
are sufficient for the second-stage dissolution of cellulose and
Klason lignin.
3.3. Hydrolysis of dissolved cellulose after second-stage dissolution
After the second-stage dissolution, the liquid phase carrying
the dissolved materials flowed into the hydrolysis unit to
hydrolyze the substances at 280 C according to our previous
research (Zhao et al., 2009a). Considering that the effective
volume of the hydrolysis unit was 90 mL and could not be
adjusted, the hydraulic retention time was
2.25
min as
calculated from the flow rate of

100

Dissolution ratio (%)

80

60

40
Possible dissolution ratio

20

Lowest dissolution ratio

0
230 C

235 C

240 C

245 C

250 C

40 mL/min. The hydrolysates were collected for 30 min with a total

volume of 1.2 L, given that the solid retention time in the secondstage dissolution was 30 min. According to previous research, the
hydrolysates from cellulose mainly include oligosaccharides (e.g.,
cellopentaose, cellotetraose, cellotriose, and cellobiose), monosaccharides (e.g., glucose and fructose), and degradation compounds
from monosaccharides (e.g., glyceraldehyde, dioxyacetone, and 5hydroxymethyl-2-furaldehyde (HMF)). Table 5 shows the substance concentrations of the hydrolysates from different experiments on the second-stage dissolution.
The substance concentrations in the final hydrolysate varied
significantly with the increase in temperature in the secondstage dissolution. In particular, the concentrations of oligosaccharides, such as cellopentaose and cellobiose, doubled and even
tripled at 245 C and 250 C compared with those obtained at
tem- peratures below 240 C. The concentrations of hexoses,
including glucose and fructose, obtained at 245 C and 250 C
were also much higher than those from the other conditions,
although the degradation compounds (e.g., glyceraldehyde and
dioxyacetone) showed similar trends at the same time. HMF,
which can originate from either glucose or fructose and behaves as
a typical inhibitor in possible microbial fermentation for further
ethanol production (Rasmussen et al., 2014), was found at
relatively low
concentra- tions in the tested hydrolysates.
Furfural, which is another inhibi- tor degraded from xylose, was
not detected in the hydrolysates probably because it had flown
out of the system together with xylose during the first- and
second-stage dissolutions before the hydrolysate collection. The
stepwise collection in the integrated process can avoid high
concentrations of inhibitors, and further refining and mixing can
be added to recover xylose if required.
Given that oligosaccharides are oligomers of glucose and can be
further hydrolyzed into glucose at much gentler conditions,
oligosaccharides together with glucose and fructose, the isomeride
of glucose, are all considered soluble sugars. Therefore, the collected yields of the soluble sugars were calculated in terms of
the contents of cellulose in the raw materials, as shown in Table 6.
The data of the collected yields show that 245 C and 250 C in
the second-stage dissolution provided the highest recovery efficiencies of soluble sugars in all the tested conditions. Although
the values are comparable with previously reported maximum
concentrations of oligosaccharides (78%) and hexoses (45%)
obtained from cellulose I hydrolyzed at 270 C for 20 min with a
semi-flow reaction system (Abdullah et al., 2014), they are relatively low when compared with the sugar yields obtained with
batch reaction systems in our previous research (Zhao et al.,
2009b). One of the reasons is that the hydrolysis unit was operated
at a fixed condition and could not be adjusted because of the manufacturing limitation of the reaction system. This scenario led to a
sub-optimal condition for the hydrolysis process, which did not
match the prior dissolution processes. Another reason is that the
collection times were unified as 30 min according to the solid
retention time in the second-stage dissolution. This operation
resulted in low concentrations and significant loss of the

Fig. 4. Lowest and corresponding possible dissolution ratios of cellulose in the

undissolved materials in the second-stage dissolution.

Table 5
Substance concentrations in liquid hydrolysates collected after combining dissolution and hydrolysis with different temperatures of the second-stage dissolution (mg/L).
Substance

230 C

235 C

240 C

245 C

250 C

Cellopentaose
Cellotriose
Cellobiose
Glucose
Fructose
Glyceraldehyde
Dioxyacetone
HMF

35.07 6.68
28.72 3.14
25.50 2.02
5.36 0.35
28.38 6.99
26.08 2.86
15.37 1.53
1.75 0.11

36.90 7.45
23.09 4.85
23.30 3.41
4.71 1.25
47.15 24.16
33.05 10.86
13.45 2.88
1.63 0.26

56.94 3.49
36.82 2.19
27.29 0.13
9.22 0.67
45.08 10.21
38.61 4.86
16.67 2.37
3.07 0.32

88.70 9.20
32.42 2.67
54.42 5.25
13.83 2.73
111.43 5.32
49.87 3.12
17.76 2.83
2.41 0.06

135.00 3.26
30.21 1.06
51.28 12.35
14.68 3.27
88.98 23.34
44.87 6.33
28.02 0.12
3.84 1.30

Table 6
Collected yields of soluble sugars in the hydrolysates after combining dissolution and hydrolysis with different temperatures of the second-stage dissolution.
Item

230 C

235 C

240 C

245 C

250 C

Mass of cellulose in raw materials (g)

Concentration of soluble sugars (mg/L)
Collected mass of soluble sugars (mg)
Collected yields of soluble sugars (%)

2.71 0.03
123.03 0.46
147.64 0.55
5.45

2.75 0.07
135.15 31.42
168.18 37.70
6.12

2.70 0.01
175.35 16.70
210.42 20.04
7.79

2.71 0.01
300.79 9.18
360.95 11.01
13.32

2.71 0.02
320.14 34.64
384.17 41.57
14.18

hydrolysates during the heating and cooling periods of the reaction

units as well as the residual hydrolysates in the reaction system.
In addition, the absolute value of the collected yields is not the
focus of this study. By contrast, the hydrolysis units were added to
com- plete the integrated processes and to help reveal the
variation trend of the different dissolution processes. Thus, the
relative com- parison of the data from different tested conditions
is more impor- tant than the absolute values. Furthermore,
previous research indicated that the degraded products from
glucose are
generated at an almost similar time as the
hydrolyzed products from cellu- lose (Abdullah et al., 2014). To
prevent the degradation, a much higher temperature and a very
short reaction time are probably required (Tolonen et al., 2013).
Thus, further study on the opti- mization of the hydrolysis unit
and its
connection with
prior first- and second-stage
dissolution units remains desirable; such study should be
conducted before the application of the entire combined process
for lignocellulosic conversion.
4. Conclusions
Multi-level dissolution and hydrolysis of the major compounds
in corn stalks were studied with a semi-flow hydrothermal
process. Amorphous hemicellulose and lignin were dissolved in
first-stage dissolution above 195 C, as evidenced by the high crystallinity of undissolved materials and the dissolution of all xylan,
acid soluble lignin, and part of Klason lignin. In second-stage dissolution above 245 C, over 75% of glucan was dissolved, with the
crystallinity decreasing significantly. Soluble sugars were further
obtained from glucan hydrolysis. The mechanisms of the multilevel process were discussed to provide significant knowledge on
its promising application in the conversion of lignocellulosic waste.
Acknowledgements
This work was supported by the National Natural Science Foundation of China (Grant Nos. 51578071 and 51208039). We would
like to thank the College of Materials Science and Technology, Beijing Forestry University, for helping in the lignocellulosic composition analysis.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in
the
online
version,
at
http://dx.doi.org/10.1016/j.biortech.2016.04.
135.
References
Abdullah, R., Ueda, K., Saka, S., 2014. Hydrothermal decomposition of various
crystalline celluloses as treated by semi-flow hot-compressed water. J. Wood
Sci. 60, 278286.
Bommarius, A.S., Katona, A., Cheben, S.E., Patel, A.S., Ragauskas, A.J., Knudson, K., Pu,
Y., 2008. Cellulase kinetics as a function of cellulose pretreatment. Metab. Eng.
10, 370381.
Cao, C., Yang, Z.L., Han, L.J., Jiang, X.P., Ji, G.Y., 2015. Study on in situ analysis of
cellulose, hemicelluloses and lignin distribution linked to tissue structure of
crop stalk internodal transverse section based on FTIR microspectroscopic
imaging. Cellulose 22, 139149.

Funke, A., Ziegler, F., 2010. Hydrothermal carbonization of biomass: a summary and
discussion of chemical mechanisms for process engineering. Biofuel Bioprod.
Biorefin. 4, 160177.
Gullon, P., Romani, A., Vila, C., Garrote, G., Parajo, J.C., 2012. Potential of
hydrothermal treatments in lignocellulose biorefineries. Biofuel Bioprod.
Biorefin. 6, 219232.
Hoekman, S.K., Broch, A., Robbins, C., 2011. Hydrothermal carbonization (HTC) of
lignocellulosic biomass. Energy Fuel 25, 18021810.
Jiang, W., Peng, H.D., Li, H.Q., Xu, J., 2014. Effect of acetylation/deacetylation on
enzymatic hydrolysis of corn stalk. Biomass Bioenergy 71, 294298.
Kang, S.M., Li, X.L., Fan, J., Chang, J., 2013. Hydrothermal conversion of lignin: a
review. Renewable Sustainable Energy Rev. 27, 546558.
Kont, R., Kurasin, M., Teugjas, H., Valjamae, P., 2013. Strong cellulase inhibitors from
the hydrothermal pretreatment of wheat straw. Biotechnol. Biofuels 6, 135
148.
Kruse, A., 2008. Supercritical water gasification. Biofuel Bioprod. Biorefin. 2, 415
437.
Kumar, S., Gupta, R.B., 2008. Hydrolysis of microcrystalline cellulose in subcritical
and supercritical water in a continuous flow reactor. Ind. Eng. Chem. Res. 47,
93219329.
Kumar, S., Gupta, R., Lee, Y.Y., Gupta, R.B., 2010. Cellulose pretreatment in
subcritical water: effect of temperature on molecular structure and enzymatic
reactivity. Bioresour. Technol. 101, 13371347.
Li, H.J., Pu, Y.Q., Kumar, R., Ragauskas, A.J., Wyman, C.E., 2014. Investigation of lignin
deposition on cellulose during hydrothermal pretreatment, its effect on
cellulose hydrolysis, and underlying mechanisms. Biotechnol. Bioeng. 111,
485492.
Libra, J.A., Ro, K.S., Kammann, C., Funke, A., Berge, N.D., Neubauer, Y., Titirici, M.M.,
Fuhner, C., Bens, O., Kern, J., Emmerich, K.H., 2011. Hydrothermal carbonization
of biomass residuals: a comparative review of the chemistry, processes and
applications of wet and dry pyrolysis. Biofuels 2, 71106.
L, X., Saka, S., 2010. Hydrolysis of Japanese beech by batch and semi-flow water
under subcritical temperatures and pressures. Biomass Bioenergy 34, 1089
1097.
L, X., Saka, S., 2012. New insights on monosaccharides isomerization, dehydration
and fragmentation in hot-compressed water. J. Supercrit. Fluid 61, 146156.
L, X., Yamauchi, K., Phaiboonsilpa, N., Saka, S., 2009. Two-step hydrolysis of
Japanese beech as treated by semi-flow hot-compressed water. J. Wood Sci. 55,
367375.
Mosier, N., Wyman, C., Dale, B., Elander, R., Lee, Y.Y., Holtzapple, M., Ladisch, M.,
2005. Features of promising technologies for pretreatment of lignocellulosic
biomass. Bioresour. Technol. 96, 673686.
Pinkowska, H., Wolak, P., 2013. Hydrothermal decomposition of rapeseed straw in
subcritical water. Proposal of three-step treatment. Fuel 113, 340346.
Qing, Q., Yang, B., Wyman, C.E., 2010. Xylooligomers are strong inhibitors of
cellulose hydrolysis by enzymes. Bioresour. Technol. 101, 96249630.
Rasmussen, H., Sorensen, H.R., Meyer, A.S., 2014. Formation of degradation
compounds from lignocellulosic biomass in the biorefinery: sugar reaction
mechanisms. Carbohydr. Res. 385, 4557.
Resende, F.L.P., Neff, M.E., Savage, P.E., 2007. Noncatalytic gasification of cellulose in
supercritical water. Energy Fuel 21, 36373643.
Rostagno, M.A., Prado, J.M., Mudhoo, A., Santos, D.T., Forster-Carneiro, T., Meireles,
M.A.A., 2015. Subcritical and supercritical technology for the production of
second generation bioethanol. Crit. Rev. Biotechnol. 35, 302312.
Sittijunda, S., Tomas, A.F., Reungsang, A., O-Thong, S., Angelidaki, I., 2013. Ethanol
production from glucose and xylose by immobilized Thermoanaerobacter
pentosaceus at 70 C in an up-flow anaerobic sludge blanket (UASB) reactor.
Bioresour. Technol. 143, 598607.
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008.
Determination of Structural Carbohydrates and Lignin in Biomass. Laboratory
Analytical Procedure. National Renewable Energy Laboratory, NREL/TP-51042618.
Tolonen, L.K., Penttila, P.A., Serimaa, R., Kruse, A., Sixta, H., 2013. The swelling and
dissolution of cellulose crystallites in subcritical and supercritical water.
Cellulose 20, 27312744.
Tymchyshyn, M., Xu, C.C., 2010. Liquefaction of bio-mass in hot-compressed water
for the production of phenolic compounds. Bioresour. Technol. 101, 24832490.
Xiao, X., Bian, J., Li, M.F., Xu, H., Xiao, B., Sun, R.C., 2014. Enhanced enzymatic
hydrolysis of bamboo (Dendrocalamus giganteus Munro) culm by hydrothermal
pretreatment. Bioresour. Technol. 159, 4147.
Yoo, C.G., Kim, H., Lu, F.C., Azarpira, A., Pan, X.J., Oh, K.K., Kim, J.S., Ralph, J., Kim, T.H.,
2016. Understanding the physicochemical characteristics and the improved
enzymatic saccharification of corn stover pretreated with aqueous and gaseous
ammonia. Bioenergy Res. 9, 6776.

Zhao, X.Y., Chen, J., Chen, F.L., Wang, X.C., Zhu, Q.J., Ao, Q., 2013. Surface
characterization of corn stalk superfine powder studied by FTIR and XRD.
Colloid Surf. B 104, 207212.
Zhao, Y., Lu, W.J., Wang, H.T., Li, D., 2009a. Combined supercritical and subcritical
process for cellulose hydrolysis to fermentable hexoses. Environ. Sci. Technol.
43, 15651570.
Zhao, Y., Lu, W.J., Wang, H.T., Yang, J.L., 2009b. Fermentable hexose production from
corn stalks and wheat straw with combined supercritical and subcritical
hydrothermal technology. Bioresour. Technol. 100, 58845889.

Zhao, Y., Lu, W.J., Wu, H.Y., Liu, J.W., Wang, H.T., 2012. Optimization of supercritical
phase and combined supercritical/subcritical conversion of lignocellulose for
hexose production by using a flow reaction system. Bioresour. Technol. 126,
391396.
Zhao, Y., Zhang, S.Y., Chen, J.J., 2015. Mechanisms of sequential dissolution and
hydrolysis for lignocellulosic waste using a multilevel hydrothermal process.
Chem. Eng. J. 273, 3745.