Chinese Journal of Chemical Engineering, 16(6) 890ü894 (2008)

Kinetics of Non-catalyzed Decomposition of Glucose in

High-temperature Liquid Water*

JING Qi (࠲ត) and LÜ Xiuyang (ে༎ཽ)**

Department of Chemical and Biochemical Engineering, Zhejiang University, Hangzhou 310027, China

Abstract The decomposition kinetics of glucose was studied in high-temperature liquid water (HTLW) from 180
to 220qC under a pressure of 10 MPa. It was found the main products from glucose decomposition were
5-hydroxymethylfurfural (5-HMF) and levulinic acid (LA). The decomposition kinetics of 5-HMF and stability of
LA in HTLW were further investigated. A kinetic model for glucose decomposition was proposed accordingly. In
the model, a series of first-order reactions with the consideration of parallel by-reactions were used to illustrate the
decomposition of glucose. The decomposition activation energies of glucose, 5-HMF, and LA were evaluated as
118.85, 95.40, and 31.29 kJ·molˉ1, respectively.
Keywords high-temperature hydrothermal decomposition, glucose, 5-hydroxymethylfurfural, levulinic acid, reac-
tion kinetics

1 INTRODUCTION furfural] were obtained. Under a pressure of 25 MPa

and temperature from 448 to 673 K, Matsumura et al.
[15] found that the reaction order was less than unity
Glucose is a very common monosaccharide, a
monomer of many polysaccharides such as starch and and the reaction mechanism was changed with tem-
cellulose, and has been often employed as a model perature.
compound for the study of lignocellulosic biomass 5-HMF and LA are valuable chemical feedstocks
decomposition. with numerous industrial applications [16, 17]. The
Typical approaches for biomass decomposition generic decomposition pathway from glucose to
are hydrolysis with acid, alkaline, or enzyme as cata- 5-HMF and LA is shown in Fig. 1 [1820]. However,
lyst [13]. The acid or alkaline hydrolysis process is there is still lack of knowledge regarding formation
associated with production of significant amount of rates of 5-HMF and LA from glucose and degradation
acidic/alkalinous waste. The enzymatic hydrolysis rates of 5-HMF and LA in HTLW. In order to establish
offers an environmentally benign alternative, but the a complete kinetics equation for glucose decomposi-
reaction rate is often too low. tion in HTLW, the kinetics data of glucose decompo-
As an environmentally benign medium, high- sition from 180 to 220°C, 5-HMF and LA degradation
temperature liquid water (HTLW) [4] has drawn in- from 180 to 280°C will be determined in this paper.
creasing attention for organic chemical reactions [5, 6] The work will provide important basic data for proc-
and biomass conversion [79]. Non-catalytic decom- ess development of biomass conversion in HTLW.
position of biomass is an environmentally-benign
process with high efficiency, and is anticipated suit- 2 EXPERIMENTAL
able for large-scale treatment of biomass. In our pre-
vious work, we studied the degradation of xylose in 2.1 Experimental procedures
HTLW from 180 to 220°C under a pressures of 10
MPa. We found that the main products of xylose de- All the reagents were commercially available and
composition were furfural and formic acid [10]. used without further purification. A 500 ml stirred
Kabyemela et al. [1114] investigated D-glucose de- autoclave reactor made of 316L stainless steel was
composition in HTLW and discovered the existence of used for experiment. Prior to the decomposition reac-
dehydration, retro-aldol, hydration and isomerization tion, the reactor was filled with 300 ml deionized wa-
reactions. Various linear [such as lactic acid (LA), ter, deoxidized with high purity nitrogen for 30 min,
glyceraldehyde, and glycolaldehyde], and ring com- and then heated to a set temperature. The agitation
speed was fixed at 300 r·min 1 and the temperature
ˉ
pounds [such as 5-hydroxymethylfurfural (5-HMF),

Figure 1 Decomposition pathway from glucose to 5-HMF and LA

Received 2007-12-05, accepted 2008-07-31.

* Supported by the National Natural Science Foundation of China (20674068) and the Natural Science Foundation of Zhejiang
Province (Y405157).
** To whom correspondence should be addressed. E-mail: [email protected]
 Chin. J. Chem. Eng., Vol. 16, No. 6, December 2008 891

was controlled within f1qC of the set value. The re- 3.1.2 Effect of reaction temperature on glucose con-
action started with charging a 50 ml of aqueous reac- version
tant solution into the reactor using a pump. The charge Decomposition of glucose was investigated at
pipeline was washed with additional 20 ml of deion- 180220qC under 10 MPa. Glucose conversions at
ized water. The end of the loading procedure was set various temperatures are shown in Fig. 3. It was ob-
as the initial time (t 0) for the reaction. The reaction served that high reaction rate was achieved at elevated
pressure was maintained at 10 MPa with high pressure temperature. The glucose conversion was 35.2% in
nitrogen from a cylinder. During the reaction, 23 ml 180 min at 180qC, while the conversion reached
of samples were collected after discarding first 23 ml 95.8% in 90 min at 220qC.
of solution from the sampling line at certain time in-
stants. The samples were filtrated with a 0.25 ȝm filter
and analyzed with HPLC. The initial concentrations of
glucose, 5-HMF, and LA solution were 0.060 mol·L 1,
ˉ
ˉ1 ˉ1
0.060 mol·L , and 0.046 mol·L , respectively.

2.2 Method of analysis

A quantitative analysis of the samples was con-

ducted with Agilent 1100 HPLC. The column was a
Sugar SH1011 (Shodex, 8 mm ID×300 mm). The
HPLC was operated at 60qC and eluted with 0.5
ml·min 1 of sulfuric acid water solution (0.5 mmol·L 1
ˉ ˉ
Figure 3 Effect of temperature and reaction time on glu-
in water). The temperature of Refractive Index Detec- cose conversion
tor was set at 40qC. The identification was made by T/qC: Ɣ 180; ż 190; Ÿ 200; Ŷ 210; Ƒ 220
comparison of the retention times of the samples with üü kinetic model
those of pure compounds. The concentrations of each
component were determined using calibration curves Figures 4 and 5 show the effect of reaction tem-
of the standards. perature on the yields of 5-HMF and LA. The yield of
5-HMF increased with reaction time at the temperature
3 RESULTS AND DISCUSSION
3.1 Decomposition of glucose in HTLW
3.1.1 Decomposition products
Figure 2 (a) shows a typical HPLC chromato-
graph of glucose decomposition product. In general,
the colorless solution turned to brown after decompo-
sition in HTLW. In addition to LA, 5-HMF, and for-
mic acid, there existed two unknown components, a
and b. Also existed are insoluble dark-brown sub-
stances, maybe humic matter.

Figure 4 Effect of temperature and reaction time on the

yields of 5-HMF
T/qC:ƽ180;ƻ190;Ʒ200;Ƶ210;ƶ220
üü kinetic model

(a) Glucose decomposition at 200qC and 120 min

Figure 5 Effect of temperature and reaction time on the

(b) HMF decomposition at 220qC and 180 min yields of LA
Figure 2 Typical HPLC chromatograph of glucose and T/qC:ƻ190;Ʒ200;Ƶ210;ƶ220
HMF decomposition product üü kinetic model
892 Chin. J. Chem. Eng., Vol. 16, No. 6, December 2008

of 180 to 200qC. When the temperatures were ele- experimental conditions. The maximum decomposi-
vated to 210qC and 220qC, the yield of 5-HMF in- tion of LA was 7.08% at 280qC for 32 h.
creased at first and then decreased. The maximum
yield for 5-HMF was 32.0% in 30 min at 220qC. 3.4 Kinetic modeling

3.2 Decomposition of 5-HMF in HTLW During the decomposition of glucose, 5-HMF, two
3.2.1 Decomposition products unidentified compounds a and b, and insoluble humic
Figure 2 (b) shows a typical HPLC chromato- matter were found. It was also observed that 5-HMF
graph of 5-HMF decomposition product. It appeared further degraded to formic acid, levulinic acid, and
that two unknown components in the 5-HMF decom- humic matter. The following first order kinetic model
position product were identical to components a and b was used for the glucose decomposition in HTLW:
found in the glucose decomposition product, indicat-
ing the 5-HMF decomposition reaction mainly con-
tributed to the formation of unknown components a
and b. In addition, humic matter appeared in all runs.
3.2.2 Effect of reaction temperature on 5-HMF con-
version which is a modified version of what was reported in a
In order to obtain actual formation rates of previous paper [21].
5-HMF during the glucose decomposition, it is neces- The following differential equations were derived:
sary to understand degradation rates of 5-HMF in dCGLU / dt  k1  k2  CGLU (1)
HTLW. 5-HMF degradation experiments were carried
out at various temperatures from 180 to 260qC (Fig. 6). dC5-HMF / dt k1CGLU   k3  k4  C5-HMF (2)
The decomposition rates of 5-HMF were substantially
lower than those of glucose in HTLW. dCLA / dt k3C5-HMF  k5CLA (3)
kG  k1  k2  (4)
kH  k3  k 4  (5)
with the analytical solution as
CGLU CGLU,0 e kG t (6)
k1CGLU,0
CHMF  e k t  e k t 
G H (7)
k H  kG
k1k3CGLU,0
CLA ª¬ k5  kH  e  kG t 
 H G  5 G  5 H 
k  k k  k k  k
Figure 6 Effect of temperature on decomposition conver-
sion of 5-HMF in HTLW  k5  kG  e k t   kH  kG  e k t º¼
H 5

T/qC:ƽ180;ƻ200;Ʒ220;Ƶ240;ƶ260
(8)
The reaction rate constants were estimated as
3.3 Stability of LA in HTLW follows: (1) the overall glucose decomposition rate
constants (kG) were calculated from the plots of
Stability of LA in HTLW was tested. The LA  ln(1  X ) versus reaction time using the data in Fig. 3;
conversions from 220 to 280qC are shown in Fig. 7. It (2) the decomposition rate constants of 5-HMF (kH)
can be seen that LA was moderately stable under the were obtained from the plots of  ln(1  X ) versus
time with the data in Fig. 6. We got the decomposition
rate constants of 5-HMF of 0.0021 min 1 at 190qC
ˉ
ˉ1
and 0.0059 min at 210qC, respectively by interpola-
tion; (3) as listed in Table 1, the decomposition rate

Table 1 Rate constants for LA decomposition at

different temperatures
k5×105/min
ˉ1
T/ć
180 
190 0.903
200 1.072
210 1.264
Figure 7 Effect of temperature on LA conversion
220 1.480
T/qC:ƻ220;ƽ240;Ƶ260; ƶ280
 Chin. J. Chem. Eng., Vol. 16, No. 6, December 2008 893

Table 2 Rate constants for glucose decomposition at different temperatures

ˉ1 ˉ1 ˉ1 ˉ1 ˉ1 ˉ1
T/ć k1/min k2/min kG/min kH/min k3/min k4/min

180 0.00178 0.00082 0.0026 0.0010  0.001

190 0.00355 0.00225 0.0058  0.00023 0.00187
200 0.00562 0.00618 0.0118 0.0039 0.00032 0.00358
210 0.00982 0.00808 0.0179  0.00059 0.00527
220 0.0196 0.0165 0.0361 0.012 0.00091 0.0112

constants of LA (k5) were obtained from the plots of tion time are essential, but for higher yield of LA, we
 ln(1  X ) versus time in Fig. 7; (4) k1 was correlated need longer reaction time and lower temperature.
with Eq. (7), k2 was obtained from the difference be-
tween kG and k1, and k3 was fitted using Eq. (8). NOMENCLATURE
The correlated kinetic parameters are listed in
Table 2. As shown in Figs. 35, a good fit between the
pre-exponential factor, min 1
ˉ
experimental data and kinetic modeling was achieved. A
concentration of glucose, mol·L 1
ˉ
The apparent activation energies tabulated in Table 3 CGLU
initial concentration of glucose, mol·L 1
ˉ
CGLU,0
were evaluated with the Arrhenius equation. ˉ1
C5-HMF concentration of 5-HMF, mol·L
concentration of LA, mol·L 1
ˉ
CLA
ˉ1
Table 3 Apparent activation energies for E activation energy, kJ·mol
activation energy of overall glucose decomposition, kJ·mol 1
ˉ
glucose decomposition EG
ˉ1
EH activation energy of overall 5-HMF decomposition, kJ·mol
ˉ1 ˉ1
E/kJ· mol A/min R2 k reaction rate constants, min 1
ˉ

EG 118.85 1.40×10 11
0.993 t time, min
9
E1 108.03 5.06×10 0.994
E2 135.71 4.29×1012 0.970
REFERENCES
EH 95.40 1.21×108 0.983
1 Mosier, N.S., Sarikaya, A., Ladisch, C.M., Ladisch, M.R., “Charac-
E3 89.28 2.62×106 0.986
terization of dicarboxylic acids for cellulose hydrolysis”, Biotechnol.
E4 108.91 3.57×109 0.993 Prog., 17 (3), 474480 (2001).
E5 31.29 0.031 0.991 2 Karimi, K., Kheradmandinia, S., Taherzadeh, M.J., “Conversion of
rice straw to sugars by dilute-acid hydrolysis”, Biomass and Bio-
energy, 30 (3), 247253 (2006).
From the modeling, we can see: 3 Jin, S.W., Zhu, S.D., Wu, Y.X., Yu, Z.N., “Research progress on en-
(1) E2 is greater than E1, indicating that high zymatic hydrolysis of lignocellulosic materials”, Biomass Chem.
temperature enhances the formation of by-products. Eng. (China), 40 (3), 4853 (2006). (in Chinese)
(2) E3 is less than E4, indicating that lower tem- 4 Akiya, N., Savage, P.E., “Roles of water for chemical reactions in
perature is favorable for LA formation. high-temperature water”, Chem. Rev., 102 (8), 27252750 (2002).
(3) The formation rate of LA was substantially 5 Xu, X., Antal, M.J., Anderson, D.G.M., “Mechanism and tempera-
higher than its decomposition rate. ture-dependent kinetics of the dehydration of tert-butyl alcohol in
hot compressed liquid water”, Ind. Eng. Chem. Res., 36 (1), 2341
4 CONCLUSIONS (1997).
6 Lü, X.Y., Li, Z., Gao, F., “Base-catalyzed reactions in NH3-enriched
near-critical water”, Ind. Eng. Chem. Res., 45 (12), 41454149
An in-depth experimental study on the glucose (2006).
decomposition in HTLW from 180 to 220°C was 7 Ando, H., Sasaki, T., Kokusho, T., Shibata, M., Uemura, Y., Hatate,
conducted. At 220qC, approximately 100% of glucose Y., “Decomposition behavior of plant biomass in hot-compressed
was decomposed in 90 min without adding any cata- water”, Ind. Eng. Chem. Res., 39 (10), 36883693 (2000).
lyst. The main decomposition products were 5-HMF, 8 Moreschi, S.R.M., Petenate, A.J., Meireles, M.A.A., “Hydrolysis of
LA, humic matter, and two unidentified soluble com- ginger bagasse starch in subcritical water and carbon dioxide”, J.
pounds. The maximum yield of 5-HMF reached Agric. Food Chem., 52 (6), 17531758 (2004).
32.0% in 30 min. 9 Lü, X.Y., Sakoda, A., Suzuki, M., “Decomposition of cellulose by
A first-order approach was used for the kinetic continuous near-critical water reactions”, Chin. J. Chem. Eng., 8 (4),
modeling of the decomposition reaction. A good 321325 (2000).
agreement between the experimental data and model- 10 Jing, Q., Lü, X.Y., “Kinetics of non-catalyzed decomposition of
ing was reached. D-xylose in high temperature liquid water”, Chin. J. Chem. Eng., 15
From the kinetic analysis, we can see that glu- (5), 666669 (2007).
cose should be decomposed at lower reaction tem- 11 Kabyemela, B.M., Adschiri, T., Malaluan, R., Arai, K., “Degradation
perature to lessen by-products formation. For higher kinetics of dihydroxyacetone and glyceraldehyde in subcritical and
yield of 5-HMF, higher temperature and shorter reac- supercritical water”, Ind. Eng. Chem. Res., 36 (6), 20252030
894 Chin. J. Chem. Eng., Vol. 16, No. 6, December 2008

(1997). linlic acid and use as a platform chemical for derived products”, Re-
12 Kabyemela, B.M., Adschiri, T., Malaluan, R.M., Arai, K., “Kinetics sources Conserv. Recycling, 28 (3/4), 227239 (2000).
of glucose epimerization and decomposition in subcritical and su- 17 Huber, G.W., Chheda, J.N., Barrett, C.J., Dumesic, J.A., “Production
percritical water”, Ind. Eng. Chem. Res., 36 (5), 15521558 (1997). of liquid alkanes by aqueous-phase processing of biomass-derived
13 Kabyemela, B.M., Adschiri, T., Malaluan, R.M., Arai, K., Ohzeki, H., carbohydrates”, Science, 308 (3), 14461450(2005).
“Rapid and selective conversion of glucose to erythrose in super- 18 Qian, X.H., Nimlos, M.R., Johnson, D.K., Himmel, M.E., “Acidic
critical water”, Ind. Eng. Chem. Res., 36 (12), 50635067 (1997). sugar degradation pathways”, Appl. Biochem. Biotech., 121-124,
14 Kabyemela, B.M., Adschiri, T., Malaluan, R.M., Arai, K., “Glucose 989997 (2005).
and fructose decomposition in subcritical and supercritical water: 19 Asghari, F.S., Yoshida, H., “Acid-catalyzed production of
detailed reaction pathway, mechanisms, and kinetics”, Ind. Eng. 5-hydroxy-methylfurfural from D-fructose in subcritical water”, Ind.
Chem. Res., 38 (8), 28882895 (1999). Eng. Chem. Res., 45 (7), 21632173 (2006).
15 Matsumura, Y., Yanachi, S., Yoshida, T., “Glucose decomposition 20 Horvat, J., Klaic, B., Metelko, B., Sunjic, V., “Mechanism of levu-
kinetics in water at 25 MPa in the temperature range of 448673 K”, linic acid formation”, Tetrahedron Lett., 26 (17), 21112114 (1985).
Ind. Eng. Chem. Res., 45 (6), 18751879 (2006). 21 Cai, L., Lü, X.Y., “Decomposition kinetics of glucose in high tem-
16 Bozell, J.J., Moens, L., Elliott, D.C., Wang, Y., Neuenscwander, G.G., perature liquid water”, Chem. React. Eng. Tech. (China), 23 (1),
Fitzpatrick, S.W., Bilski, R.J., Jarnefeld, J.L., “Production of levu- 8891 (2007). (in Chinese)