Documente Academic
Documente Profesional
Documente Cultură
Received 10 December 2001; received in revised form 17 May 2002; accepted 22 May 2002
Abstract
Experimental trials of the dilute acid hydrolysis of bagasse hemicellulose to produce xylose, arabinose, glucose, acid-soluble
lignin (ASL) and furfural were conducted using a temperature-controlled digester. The reaction conditions varied were;
temperature (80200 C), mass ratio of solid to liquid (1:5 1:20), type of bagasse material (i.e. bagasse or bagacillo),
concentration of acid (0.25 8 wt% of liquid), type of acid (hydrochloric or sulphuric) and reaction time (10 2000 min).
Kinetic modelling of the global rates of formation of products was performed. The most accurate kinetic model of the global
reaction for the decomposition of xylan was a simple series hydrolysis of xylan to xylose followed by xylose decomposition.
Similar schemes were used to model the production of arabinose, glucose and furfural from the hemicellulose. The production
of ASL was modelled by a 9rst-order decomposition of lignin to ASL followed by a reversible decomposition of ASL.
Yields of up to 220 mg xylose=g solid were achieved, i.e. about 80% of the theoretical xylose available from the bagasse.
The bagasse particle size was found to negligibly a<ect the rate of hydrolysis. Hydrochloric acid was found to be less active
for the degradation of xylose compared to sulphuric acid. ? 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Bagasse; Hemicellulose; Acid hydrolysis; Xylose; Arabinose; Glucose; Lignin; Acid-soluble lignin; Furfural
1. Introduction
Sugarcane bagasse constitutes the 9brous residue of
sugar cane after undergoing conventional milling. An
absolute minimum of about 50% of this 9bre is needed
to generate heat and power to run the sugar milling
process [1] and the remainder can be stockpiled. The
stockpiled bagasse is of low economic value and constitutes an environmental problem to sugar mills and
surrounding districts, especially if stockpiled for extended periods, due to the risk of spontaneous combustion occurring within the pile [2]. As a means of
minimising this hazard, many mills have highly inef9cient furnaces that burn large portions of the excess
bagasse. It has been estimated that bagasse consumption within sugar mill boilers could be reduced by up
to 36% by ceasing such practices and adopting more
energy e(cient modes of operation [3,4]. If a suitable
use could be found for the excess bagasse there is,
0961-9534/02/$ - see front matter ? 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 9 6 1 - 9 5 3 4 ( 0 2 ) 0 0 0 6 6 - 1
368
369
Table 1
Chemical content of dried bagasse fractions
Bagasse
fraction
Pentosan content
(mg g1 material) (dry basis)
Ratio of arabinose=xylose
Bagacillo
Bagasse
277 18
268 18
315 20
305 20
0:11 0:01
0:11 0:01
0.68
0.58
370
Table 2
HPLC analytical conditions
Column
Packing
Mobile phase
Flow rate
Pre-column 1
Pre-column 2
Temperature
Injection volume
Run time
Detector
xylan(s)
k1
k2
xylose(aq)
Furfural
300 6:5 mm
Interaction CHO-620
0:05 mol=dm3 Na2 Ca(EDTA)
0:5 ml=min
Keystone Scienti9c, Inc. C-18
guard column, 11 6 mm
Interaction CHO-620 guard
column, 23 6 mm
90 C
20 l
22 min
GBC 1240 RI detector
23 6 mm (Guard column)
Interaction GC-coregel 87H
0:05 mol=dm3 sulphuric acid
0:5 ml=min
40 C
20 l
5 min
GBC 1240 RI detector
xylaneasy-to-hydrolyse (s)
k 1,e
decomposition products
k2
xylose(aq)
Scheme 1.
furfural
k 1,h
xylanhard-to-hydrolyse (s)
k3
decomposition products
Scheme 3.
k2
xylose(aq)
k 1,h
xylanhard-to-hydrolyse (s)
Scheme 2.
decomposition products
k1
bound arabinose(s)
371
k2
arabinose(aq) decomposition products
Scheme 4.
k1
k2
bound glucuronic acid(s) glucose(aq) decomposition products
Scheme 5.
k1; e
(ek1; e t ek2 t )
k2 k1; e
(1 )k1; h
(ek1; h t ek2 t ) ;
(2)
+
k2 k1; h
k1; e
L
s
(ek1; e t e(k2 +k3 )t )
Cx = Cx0
(k2 + k3 ) k1; e
(1)k1; h
(ek1; h t e(k2 +k3 )t ) ;
(3)
+
(k2 + k3 )k1; h
s
CxL = Cx0
(4)
k1
k2
pentosans(s) furfural(aq) decomposition products
Scheme 6.
372
k1
k2
lignin(s) ASL(aq) decomposition products
Scheme 7.
k2 + k 3
k3 + k2 k1
exp ((k2 + k3 )t) +
k3
;
k2 + k 3
(6)
lignin(s)
determine these fractions, and the total amount of xylose present in the bagasse, samples were hydrolysed
ASL(aq)
decomposition products
k3
Scheme 8.
373
250
200
Bagasse - xylose
Bagacillo - xylose
Bagasse - arabinose
Bagacillo - arabinose
150
100
H2SO4 concentration = 4%
= 0.05
Temperature = 90oC
50
0
0
200
400
600
800
1000
1200
1400
1600
1800
2000
1600
1800
2000
time (min)
Figure 1. Hydrolysis of bagasse=bagacillo.
e0
1 - (CLx / Csx0 )
Easy-to-hydrolyse
Hard-to-hydrolyse
e1
Bagasse
Bagacillo
H2SO4 concentration = 4%
= 0.05
Temperature = 90 C
e2
0
200
400
600
800
1000
1200
1400
time (min)
Figure 2. Determination of easy-to-hydrolyse xylan fraction in bagasse material.
374
9ner material (bagacillo) showed no measurable difference in yield over the coarser bagasse material.
Table 4 shows the predicted parameters, regression
coe(cients and standard error for Scheme 1 using
the data collected for xylose yield with hydrochloric
acid. As results using sulphuric acid had indicated
that particle size was not a strong inQuence on the
degradation rates, only bagacillo was tested using hydrochloric acid. Fig. 5 shows a typical set of hydrolysis results with the experimental data (data points)
and 9tted model (solid and dashed lines). Included in
the 9gure are the predicted results using the kinetic
parameters derived from the sulphuric acid trials. To
compare the results of the two acids, an equivalent sulphuric acid concentration was calculated by multiplying the hydrochloric acid concentration by 98=35.5 and
using this concentration in the sulphuric acid model.
As the pH of all hydrolysis solutions was well below
1.9 (the approximate pKa of the bisulphate ion) it was
assumed the sulphuric acid acted as a monoprotic acid.
It can be seen that the rate of xylose production for
the 9rst 100 min is very similar for both models, indicating that the type of acid used does not inQuence the
degradation of the bagasse. However, after this 9rst
100 min, the sulphuric acid model shows a much more
rapid degradation of the xylose and lower maximum
yield. Comparison between the xylose kinetic parameters derived for bagasse hydrolysis using sulphuric
acid (Scheme 1, Bagasse, Table 3) and hydrochloric
acid (Xylose, Table 4) show the main di<erence lies
with the pre-exponential function for xylose degradation. This is an order of magnitude lower when HCl
is used as the catalyst. This indicates that the type of
acid used does inQuence xylose degradation kinetics.
Fig. 5 also shows the a<ect of changing the solid
to liquid ratio, . As expected, decreasing the solid
to liquid ratio reduced the rate at which the xylose
decomposed as the xylose produced is more dilute
within the hydrolysate.
4.3. Arabinose, glucose and furfural kinetics
results
Fig. 6 shows a typical set of data for the yield of
arabinose, glucose and furfural from the hydrolysis of
bagasse. The kinetic models of these data are shown
as lines. To determine the amount of glucuronic acid
in the bagasse, experiments were conducted under
375
Table 3
Kinetic model results for xylose yield using sulphuric acid
Scheme 1
Feed
Bagasse
s1
Bagacillo
106
A1 ;
E1 , kJ mol1
A1; e , s1
E1; e , kJ mol1
A1; h , s1
E1; h , kJ mol1
A2 , s1
E2 , kJ mol1
A3 , s1
E3 , kJ mol1
n
Correlation coe(cient, R
Std error in xylose yield, mg g1
21:6
82.8
2:66 1012
118.9
0.82
0.88
32.5
250
Scheme 2
109
30:3
105.9
2:66 1012
117.6
0.97
0.96
22.6
Scheme 3
Bagasse
Bagacillo
Bagasse
Bagacillo
97:6 106
88.1
1:46 106
73.5
0:294 1012
111.2
0.79
0.88
32.5
1346:0 106
98.6
11:9 106
76.5
0:353 1012
113.8
0.27
0.83
44.2
28:5 106
83.8
28:5 106
83.8
4:72 109
140.5
87:0 1010
115.0
0.81
0.88
32.5
46370 106
107.2
46370 106
107.2
4:72 109
140.5
70:6 1010
113.0
0.97
0.96
23.2
= 0.05
200
150
100
Model
Exp.
T = 120oC, Cacid=1wt%
T = 160oC, Cacid=1wt%
T = 120oC, Cacid=4wt%
50
0
0
50
100
150
200
250
300
Time (min)
Figure 3. E<ect of hydrolysis temperature and sulphuric acid concentration on xylose yield.
376
250
= 0.05 T = 120oC
200
150
100
Model
Exp.
bagasse
Cacid = 1 wt%
bagacillo
50
bagasse
bagacillo
0
0
50
100
150
200
250
300
350
Time (min)
Figure 4. E<ect of bagasse material on xylose yield.
Table 4
Kinetic model results for xylose, arabinose, glucose and furfural yield using hydrochloric acid
A1 ; s1
E1 ; kJ mol1
A2 ; s1
E2 ; kJ mol1
n
Correlation coe(cient, R
Std error in yield, mg g1
Xylose
Arabinose
Glucose
Furfural
2:30 106
74.5
6:81 1011
114.8
0.93
0.79
32.5
0.604
21.3
4:41 1013
135.1
0.63
0.81
5.5
18:0 109
105.0
6:65 1011
117.6
1.24
0.77
7.3
87:2 1015
165.6
1:28 1014
131.1
1.87
0.87
7.4
250
T = 90oC
= 0.05
200
150
= 0.2
100
= 0.2
H2SO4 Model Prediction
50
0
0
200 400 600 800 1000 1200 1400 1600 1800 2000
Time (min)
20
Arabinose, glucose, furfural yield (mg/g bagasse)
18
= 0.05
T = 120oC
Arabinose
16
14
12
10
8
6
Glucose
4
Furfural
2
0
0
50
100
150
200
250
300
Time (min)
Figure 6. Yields of arabinose, glucose and furfural from acid hydrolysis.
377
378
Table 5
Kinetic model results for arabinose, glucose and furfural yield using sulphuric acid
Arabinose
Feed
Glucose
Bagasse
s1
Bagacillo
106
A1 ,
E1 , kJ mol1
A2 , s1
E2 , kJ mol1
n
Correlation coe(cient, R
Std error in yield, mg g1
Bagasse
106
17:1
84.1
1:51 108
87.9
0.49
0.733
5.7
Furfural
Bagacillo
109
375:0
90.3
2:12 107
79.2
0.82
0.542
8.2
18:0
107.3
2:66 1012
125.5
1.010
0.8751
4.6
20
109
18:0
107.4
2:66 1012
128.7
0.7747
0.9298
4.8
Bagasse
108
1:0
103.1
4:12 104
60.3
0.7217
0.8804
5.7
Bagacillo
3:25 1013
145.6
1:85 108
85.7
0.8425
0.9714
3.3
Scheme 7 Model
18
Scheme 8 Model
16
- experimental data points
14
12
= 0.05
T = 120oC
10
8
6
4
2
0
0
50
100
150
200
250
300
350
Time (min)
Figure 7. Yields of ASL from acid hydrolysis.
379
Table 6
Kinetic model results for ASL yield
H2 SO4
Feed
s1
A1 ,
E1 , kJ mol1
A2 , s1
E2 , kJ mol1
A3 , s1
E3 , kJ mol1
n
Correlation coe(cient, R
Std error in yield, mg g1
HCl
Bagasse
106
2:16
85.2
1:23 109
95.7
4:54 104
64.4
0.390
0.90
4.7
Bagacillo
103
115:5
77.2
0:574
20.57
2:41 105
73.4
0.843
0.85
5.8
Bagacillo
0:0305 106
73.0
1:11 105
65.1
2:82 105
78.7
0.650
0.21
4.9
380