Bioresource Technology 98 (2007) 554–559

Optimization studies on acid hydrolysis of oil palm empty fruit

bunch fiber for production of xylose
S.H.A. Rahman, J.P. Choudhury *, A.L. Ahmad, A.H. Kamaruddin
School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, 14300 Nibong Tebal, SPS Penang, Malaysia

Received 25 February 2005; received in revised form 9 January 2006; accepted 6 February 2006
Available online 2 May 2006

Abstract

Oil palm empty fruit bunch fiber is a lignocellulosic waste from palm oil mills. It is a potential source of xylose which can be used as a
raw material for production of xylitol, a high value product. The increasing interest on use of lignocellulosic waste for bioconversion to
fuels and chemicals is justifiable as these materials are low cost, renewable and widespread sources of sugars. The objective of the present
study was to determine the effect of H2SO4 concentration, reaction temperature and reaction time for production of xylose. Batch reac-
tions were carried out under various reaction temperature, reaction time and acid concentrations and Response Surface Methodology
(RSM) was followed to optimize the hydrolysis process in order to obtain high xylose yield. The optimum reaction temperature, reaction
time and acid concentration found were 119 C, 60 min and 2%, respectively. Under these conditions xylose yield and selectivity were
found to be 91.27% and 17.97 g/g, respectively.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Lignocellulose; Xylose; Oil palm empty fruit bunch (OPEFB); Acid hydrolysis; Optimization

1. Introduction OPEFB biomass is comprised of 24% xylan, a sugar poly-

Bioconversion of lignocellulosic waste materials to
chemicals and fuels are receiving interest as they are low
cost, renewable and widespread in nature (Kuhad and
Singh, 1993). Malaysia is well known for its potential in
renewable resources such as oil palm waste, sugar cane
bagasse and rice straw. In the process of extraction of palm
oil from oil palm fruit, a lignocellulosic material oil palm
empty fruit bunch (OPEFB) is generated as a waste prod-
uct. Approximately fifteen million tons of OPEFB biomass
waste is generated annually throughout Malaysia by palm
oil mills. In practice this biomass is burned in incinerators
by palm oil mills which creates environmental pollution
problems in nearby localities. The OPEFB biomass con-
tains cellulose, hemicellulose and lignin. It is estimated that
*
Corresponding author. Tel.: +60 4 5937788x6424; fax: +60 4 5941013.
E-mail address: jyoti_choudhury@yahoo.com (J.P. Choudhury).
mer made of pentose sugar xylose. This xylose can be used
as substrate for production of a wide variety of compounds
by chemical and biochemical processes (Wyman, 1994;
Almeida e Silva et al., 1995). One such compound is xylitol,
which is extensively used in food, pharmaceutical and thin
coating applications (Parajo et al., 1995; Torget et al.,
1991). The most important application of xylitol is its use
as an alternative sweetener in foods for diabetic patients
(Pepper and Olinger, 1988). Other important uses of xylitol
are: as an anticariogenic agent in tooth paste formulations,
as thin coatings on chewing vitamin tablets, in mouth
washes, in beverages and in bakery products (Makinen,
1976; Emodi, 1978; Hyvonen and Koivstoinen, 1982).
Research investigations on dilute acid hydrolysis of various
raw materials such as sugar cane bagasse, sorghum straw,
corn cobs and eucalyptus wood have been carried out by
several workers (Roberto et al., 1995; Herrera et al.,
2003). From the research studies it was revealed that under
controlled treatment conditions, acid hydrolysis of ligno-

0960-8524/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2006.02.016
 S.H.A. Rahman et al. / Bioresource Technology 98 (2007) 554–559
cellulosic biomass mainly produced xylose from xylan with
the cellulosic and lignin fractions remaining unaltered.
Xylan is more susceptible to hydrolysis by mild acid treat-
ment due its amorphous structure compared to cellulose
which needs severe treatment conditions for its crystalline
nature. Although xylose was the main sugar obtained from
hemicellulose, other byproducts such as glucose, acetic
acid, furfural etc., were also produced in low amounts dur-
ing the hydrolysis process (Domı́nguez et al., 1997; Silva
et al., 1998). It was also reported that amount of sugar
released during hydrolysis, depended on type of raw mate-
rial and operating conditions of the experiment. Acid con-
centration is an important parameter for release of sugars
whereas temperature is mainly responsible for degradation
of sugars to various byproducts (Neureiter et al., 2002).
Acetic acid and furfural are potential inhibitors to yeast
metabolism. When these compounds are present in the
hydrolysate, they inhibit the fermentation process by caus-
ing cell morphological change or ultimate death of the
organism (Van Zyl et al., 1991). To keep the concentration
of byproducts (glucose, acetic acid and furfural) in the
hydrolysate at a low level it is necessary to run the hydro-
lysis reaction at less severe conditions.
To our knowledge, there are no reports on acid hydro-
lysis of OPEFB biomass. The aim of the present investiga-
tion was to determine the effect of acid concentration,
reaction temperature and reaction time period on release
of sugars (xylose and glucose) and formation of byproducts
(acetic acid and furfural) from OPEFB biomass. To maxi-
mize the formation of xylose in hydrolysate optimization
process was followed using response surface methodology
(RSM) with central composite as statistical design in order
to attain high xylose selectivity.
2. Methods
2.1. Raw material
Oil palm empty fruit bunch (OPEFB) fiber was collected
from local palm oil mill (United Oil Palm Industries Sdn
Bhd, Malaysia), sun-dried and ground to a particle size
<1 mm. The homogenized OPEFB biomass was then
oven-dried at 105 C for overnight and was analyzed fol-
lowing standard method for determination of its main
composition (Garrote et al., 1999).
2.2. Acid hydrolysis
Acid hydrolysis of OPEFB biomass were carried out in
125 ml Erlenmeyer flasks. The media consisted of 2–6 g
H2SO4/100 g liquor using a charge of 1 g OPEFB fiber/
8 g liquor on dry basis. Operating temperature of hydroly-
sis were varied between 100 and 130 C and samples were
collected at various time intervals in the range of 30–
90 min. After reaction was completed, solids were sepa-
rated from aqueous solution by filtration. The filtrate was
analyzed for xylose, glucose, acetic acid and furfural.
555
2.3. Analytical methods
Xylose and glucose in the acid hydrolysate were ana-
lyzed by High Performance Liquid Chromatograph
(HPLC, SHIMADZU) using SUPELCOSIL LC-NH2 col-
umn and RI detector. Aqueous acetonitrile (75%) was used
as mobile phase with flow rate of 1.5 ml/min and oven tem-
perature was maintained at 50 C. Acetic acid and furfural
were analyzed by Gas Liquid Chromatograph (GLC, Per-
kin Elmer) using 80/120 carbopack column with nitrogen
as carrier gas and detection was done by FID. The column
and injector temperature was maintained at 225 and
250 C, respectively while oven was operated at 175 C.
2.4. Experimental design and RSM
In the experimental plan, response surface methodology
(RSM) was utilized to optimize the hydrolysis process and
a 23 rotatable central composite design (CCD) was adopted
in order to fit a second order model and the design
consisted of 20 sets of experiments. The basic theoretical
aspects, the fundamental assumptions and the experimental
implications of RSM have been discussed elsewhere (Mont-
gomery, 2001). The second order model was selected for
predicting the optimal point and is expressed as
Y ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 3 þ b11 X 21 þ b22 X 22
þ b33 X 23 þ b12 X 1 X 2 þ b13 X 1 X 3 þ b23 X 2 X 3 ð1Þ
where Y represents response variable (xylose yield and
selectivity), b0 is interception coefficient, b1 and b2 are lin-
ear terms, b11, b22 and b33 are quadratic terms and X1, X2
and X3 are independent variables studied (temperature,
reaction time and acid concentration). Regression analysis
was performed by Design Expert v.6.0.7 (Stat-Ease Inc.
Minneapolis). Fischer’s test was used for determination
of type of model equation, while the student’s t-test was
performed for determination of statistical significance of
regression coefficients.
3. Results and discussion
3.1. Composition of empty fruit bunch
Analysis of OPEFB was carried out for determination of
principal components using quantitative acid hydrolysis.
The composition is shown in Table 1, expressed on an
Table 1
Main components of oil palm empty fruit bunch fiber (on an oven-dry
basis)
Main fraction Composition (%)
Glucan 42.85
Xylan 24.01
Lignin (acid insoluble) 11.70
Ash 0.52
Others 20.92
556 S.H.A. Rahman et al. / Bioresource Technology 98 (2007) 554–559

35
Xylose Glucose Acetic acid Furfural

30

25

20
g/L

15

10

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

RUN

Fig. 1. Formation of sugars and sugar degradation products under various hydrolysis conditions.

oven-dry basis. The potential concentration of xylose Table 3

produced in hydrolysate was 34.1 g/L. Experimental design and results obtained with OPEFB biomass hydrolysis
Runs Variables Responses
3.2. Sugar and byproduct formation X1 X2 X3 Y1 (%) Y2 (g/g)
1 1.000 1.000 1.000 13.16 0.59
It was observed that release of xylose and glucose in the 2 0.000 0.000 0.000 70.88 11.13
hydrolysate was dependent on experimental operating con- 3 1.000 1.000 1.000 11.64 7.99
ditions, which is shown in Fig. 1. Maximum concentration 4 1.000 1.000 1.000 86.91 17.57
5 0.000 0.000 0.000 90.03 19.24
of xylose, 30.81 g/L was achieved when reaction was car-
6 1.000 1.000 1.000 54.72 25.07
ried out at 115 C for time period of 60 min with acid con- 7 0.000 0.000 0.000 90.00 15.16
centration maintained at 4%. On the other hand release of 8 1.000 1.000 1.000 28.95 7.61
glucose was highest (7.61 g/L) when operating temperature 9 1.682 0.000 0.000 27.95 10.83
and time period were 130 C and 90 min, respectively while 10 1.682 0.000 0.000 41.70 2.42
11 0.000 0.000 1.682 74.55 10.44
acid concentration was maintained at 6%. The maximum
12 1.000 1.000 1.000 45.04 3.67
concentration of acetic acid and furfural obtained were 13 1.000 1.000 1.000 62.99 6.78
6.54 and 3.99 g/L, respectively when reaction temperature, 14 0.000 0.000 0.000 89.74 17.54
acid concentration and reaction time period were 130 C, 15 0.000 0.000 0.000 89.94 15.98
6% and 90 min, respectively. 16 0.000 1.682 0.000 81.55 14.84
17 0.000 0.000 0.000 90.35 18.30
18 0.000 0.000 1.682 32.90 14.46
3.3. Statistical modelling 19 1.000 1.000 1.000 70.81 12.93
20 0.000 1.682 0.000 56.30 37.57
Experimental range and levels of independent process
variables, reaction temperature (X1), reaction time (X2)
Table 4
Analysis of variance (ANOVA) for xylose yield
Table 2 Source Sum of Degree of Mean F-value P-value
Experimental range and levels of independent process variables square freedom square
Independent variable Symbol Range and levels Model 13026.27 9 1447.36 14.84 0.0001
Residual 975.49 10 97.55
a 1 0 +1 +a
Lack of fit 670.27 5 134.05 2.20 0.2041
Temperature (C) X1 90 100 115 130 140 Pure error 305.22 5 61.04
Reaction time (min) X2 10 30 60 90 110 Total 14001.76 19
Acid concentration (%) X3 1 2 4 6 7 R2 0.93
 S.H.A. Rahman et al. / Bioresource Technology 98 (2007) 554–559 557

and acid concentration (X3) studied for hydrolysis of empty liquor/maximum xylose concentration) · 100] and Y2
fruit bunch fiber are shown in Table 2. The value of a was (selectivity) was defined by (xylose/glucose ratio).
calculated as ±1.682. The quadratic models in terms of coded variables are
The design of the present experiment including depen- shown in Eqs. (2) and (3), where (Y1) represents xylose
dent variables or responses Y1 and Y2 are given in Table yield and (Y2) represents selectivity, as a function of reac-
3, where Y1 (% yield) was defined by [(xylose obtained in tion temperature (X1), reaction time (X2) and acid concen-
tration (X3).
Y 1 ¼ 86:94 þ 2:41X 1 þ 3:83X 2 þ 6:99X 3
Table 5  19:15X 21  7:09X 22  12:47X 23
Analysis of variance (ANOVA) for selectivity
 11:15X 1 X 2  22:08X 1 X 3  1:15X 2 X 3 ð2Þ
Source Sum of Degree of Mean F-value P-value
square freedom square Y 2 ¼ 16:38  3:55X 1  4:05X 2 þ 0:35X 3
Model 1100.77 9 122.31 528 0.0079  4:40X 21 þ 2:53X 22  2:34X 23  0:17X 1 X 2
Residual 231.67 10 23.17
Lack of fit 189.39 5 37.88 4.48 0.0627  5:31X 1 X 3  0:51X 2 X 3 ð3Þ
Pure error 42.28 5 8.46
To fit the response function and experimental data, regres-
Total 1332.43 19
R2 0.83 sion analysis was performed and the second order model

95
95
73.75
xylose yield
xylose yield

73.75
52.5
52.5
31.25

31.25 10

10

110.00
7.00
110.00 85.00
140.00 5.50
85.00 60.00
127.50 Ti 4.00 .
m
e cen
60.00 35.00 2.50 c on
Tim 115.00 re cid
e atu A
35.00 102.50 m per 10.00 1.00
Te
10.00 90.00
xylose yield
110.00
xylose yield
110.00 65.3797

85.00
85.00

58.0229
50.6661
Time

65.3797
60.00 6
Time

60.00
50.6661 58.0229
58.0229 50.6661
65.3797
80.0934
80.0934
35.00
35.00
65.3797
72.7366
72.7366
58.0229

10.00 10.00
90.00 102.50 115.00 127.50 140.00 1.00 2.50 4.00 5.50 7.00
Temperature Acid concn.

Fig. 2. Effect of reaction temperature and reaction time on xylose yield Fig. 3. Effect of H2SO4 concentration and reaction time on xylose yield
when H2SO4 concentration was maintained at 4%. when reaction temperature was set at 115 C.
558 S.H.A. Rahman et al. / Bioresource Technology 98 (2007) 554–559

for both responses was evaluated by ANOVA which are

presented in Tables 4 and 5. The regression for both the re-
sponses was statistically significant at 95% of confidence le-
40
vel. For the first response (Y1), the model did not show any
lack of fit and the determination coefficient (R2) obtained 30.025

selectivity
20.05
was 0.93 which showed 93% of the variability in response.
The model for the second response also did not show any 10.075

lack of fit and the determination coefficient (R2) obtained 0.1

was 0.83, which explained 83% of the variability in

response.
Estimation of xylose yield and selectivity over indepen-
dent variables X1, X2 and X3 in terms of response surfaces 110.00
are shown in Figs. 2–5. The effect of reaction temperature 85.00
7.00

and reaction time on xylose yield when H2SO4 concentra- 5.50

60.00
tion was selected at 4% as the centre point, is shown in Ti 4.00 .
m en
e nc
35.00 co
Fig. 2. From the figure it can be interpreted that maximum 2.50
Ac
id
10.00 1.00

selectivity
110.00

15.1407

40
85.00
12.3196 12.3196
30.025
selectivity

15.1407
20.05
Time

10.075 60.00 6

0.1 17.9617

20.7828
35.00

110.00
140.00 10.00
85.00
127.50 1.00 2.50 4.00 5.50 7.00
60.00 Acid concen.
Ti 115.00 e
m tur
e 35.00 102.50 pera Fig. 5. Effect of H2SO4 concentration and reaction time on selectivity
m
Te
10.00 90.00 when reaction temperature was maintained at 115 C.

selectivity
110.00

and minimum xylose yield of 90.35% and 27.94% can be

obtained by conducting hydrolysis experiment for 60 min
reaction time and at 115 and 90 C reaction temperature,
85.00
12.3196
respectively. The effect of acid concentration and reaction
time on xylose yield is shown in Fig. 3, when reaction tem-
15.1407 perature was selected at 115 C as the centre point. From
the response surface figure it is evident that maximum
Time

9.49855
60.00 6
15.1401 72.3196 and minimum xylose yield of 90.35% and 32.9% can be
17.9617 obtained by working with 60 min reaction time and at
20.7828
4% and 1% acid concentration, respectively.
35.00
On the other hand the effect of variation of reaction tem-
perature and reaction time on selectivity is shown in Fig. 4,
when acid concentration was set at 4% as the centre point.
From the figure it can be interpreted that maximum and
10.00 minimum selectivity of 37.57 and 2.42 (g/g) can be achieved
90.00 102.50 115.00 127.50 140.00 by conducting experiment at temperature of 115 and
Temperature 140 C for reaction time period of 10 and 60 min, respec-
Fig. 4. Effect of reaction temperature and reaction time on selectivity tively. Analysis of the effect of variation of acid concentra-
when H2SO4 concentration was maintained at 4%. tion and reaction time on selectivity is shown in Fig. 5,
 S.H.A. Rahman et al. / Bioresource Technology 98 (2007) 554–559 559

Table 6 Acknowledgements
Solutions for optimum condition
Solution Temperature Time Acid conc. Desirability The authors are thankful to Palm Oil Board for tech-
number (C) (min) (%) nical assistance. The financial support by the Ministry of
1 118 44 3 1 Science, Technology and Environment, Malaysia Govern-
2 119 60 2 1 ment, under long-term IRPA grant (Project: 03-02-05-
3 110 44 6 1
4 123 86 5 1
2215 EA008) is gratefully acknowledged.
5 103 58 3 1
6 110 39 5 1 References
7 113 51 5 1
8 121 75 4 1 Almeida e Silva, J.B., de Mancilha, I.M., Vannetti, M.C.D., Teixeira,
9 117 32 3 1 M.A., 1995. Microbial protein production by Paecilomyces variotii
10 122 46 5 1 cultivated in Eucalyptus hemicellulosic hydrolysate. Biores. Technol.
52, 197–200.
Domı́nguez, J.M., Cao, N., Gong, C.S., Tsao, G.T., 1997. Dilute acid
when operating temperature was set at 115 C as the centre hemicellulose hydrolysates from corn cobs for xylitol production by
yeast. Biores. Technol. 61, 85–90.
point. From the response surface it is observed that maxi-
Emodi, A., 1978. Xylitol: its properties and food applications. Food
mum and minimum selectivity of 37.57 and 10.44 (g/g) can Technol., 28–32.
be attained by conducting hydrolysis at 4% and 6% acid Garrote, G., Domı́nguez, H., Parajo, J.C., 1999. Mild autohydrolysis:
concentration for reaction time period of 10 and 60 min, an environmentally friendly technology for xyloolignosaccharide
respectively. production from wood. J. Chem. Tech. Biotech. 74, 1101–1109.
Herrera, A., Tellez-Luist, S.J., Ramirez, J.A., Vazquez, M., 2003.
Based on the models, numerical optimization was car-
Production of xylose from sorghum straw using hydrochloric acid.
ried out with the help of Design Expert v.6.0.7 considering J. Cereal Sci. 37, 267–274.
each value of response and ten solutions were obtained Hyvonen, L., Koivstoinen, P., 1982. Food technological evaluation of
which are shown in Table 6. To confirm these results, xylitol. Adv. Food Res. 28, 373–403.
hydrolysis runs were carried out in triplicate under opti- Kuhad, R.C., Singh, A., 1993. Current and future prospects. Crit. Rev.
Biotechnol. 13, 151–172.
mized conditions and the best results of xylose yield and
Makinen, K.K., 1976. Xylitol: the sugar that prevents tooth decay. The
selectivity obtained were 91.27% and 17.97 (g/g), respec- Futurist, 135–139.
tively when hydrolyzing acid used was 2% under operating Montgomery, D.C., 2001. Design and Analysis of Experiments. John
temperature of 119 C and time period of reaction was Wiley and Sons, New York, pp. 427–450.
60 min. Neureiter, M., Danner, H., Thomasser, C., Saidi, B., Braun, R., 2002.
Dilute acid hydrolysis of sugar cane bagasse at varying conditions.
Appl. Biochem. Biotechnol. 98, 49–58.
4. Conclusion Parajo, J.C., Domı́nguez, H., Domı́nguez, J.M., 1995. Production of
xylitol from raw wood hydrolysate by Debaryomyces hansenii NRRL
Hydrolysis of OPEFB biomass was carried out with Y-7426. Bioprocess Eng. 13, 125–131.
dilute sulfuric acid under selected operating conditions to Pepper, T., Olinger, P.M., 1988. Xylitol in sugar-free confections. Food
Technol. 10, 98–106.
obtain high concentration of xylose in the resulting hydro-
Roberto, I.C., Felipe, M.G.A., Mancilha, I.M., Vitolo, M., Sato, S., Silva,
lysate. A 23 rotatable central composite design was adopted S.S., 1995. Xylitol production by Candida guilliermondii as an
in designing of experiments and response surface method- approach for the utilization of agro industrial residues. Biores.
ology was adopted in order to optimize the hydrolysis pro- Technol. 51, 255–257.
cess. It was found that under optimum conditions, xylose Silva, S.S., Felipe, M.G.A., Silva, J.B.A., Prata, A.M.R., 1998. Acid
hydrolysis of Eucalyptus grandis chips for microbial production of
yield and selectivity were 91.27% and 17.97 (g/g), respec-
xylitol. Process Biochem. 33, 63–67.
tively when operating temperature, reaction time and acid Torget, R., Walter, P., Himel, M., Grohmann, K., 1991. Dilute acid
concentration were 119 C, 60 min and 2%, respectively. pretreatment of corn residues and short-rotation woody crops. Appl.
Thus it is to be mentioned that under controlled treatment Biochem. Biotechnol. 28/29, 75–86.
conditions, oil palm empty fruit bunch fiber waste can be Van Zyl, C., Prior, B.A., Du Preez, J.C., 1991. Acetic acid inhibition
of D-xylose fermentation by Pichia stripitis. Enzyme Microb. Technol.
fruitfully utilized as a potential source of xylose, which
13, 82–86.
can be a starting raw material for production of various Wyman, C.E., 1994. Ethanol from lignocellulosic biomass: Technology,
chemicals, especially xylitol by microbial conversion economics and opportunities. Biores. Technol. 50, 3–16.
process.