Bioresource Technology 152 (2014) 526–529
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Short Communication
Production and purification of xylooligosaccharides from oil palm empty fruit bunch fibre by a non-isothermal process Ai Ling Ho a, Florbela Carvalheiro b, Luís C. Duarte b, Luísa B. Roseiro b, Dimitris Charalampopoulos a, Robert A. Rastall a,⇑ a b
Department of Food and Nutritional Sciences, University of Reading, Whiteknights, P.O. Box 226, Reading RG6 6AP, United Kingdom LNEG, Unidade de Bioenergia, Estrada do Paço do Lumiar 22, 1649-038 Lisboa, Portugal
h i g h l i g h t s Information on oligosaccharides production from oil palm biomass is very scarce. Oligosaccharides production from OPEFB fibre via autohydrolysis is demonstrated. Substantial yield of XOS is achieved using non-isothermal conditions. High purity of refined XOS with wide ranges of DP were obtained.
a r t i c l e
i n f o
Article history: Received 28 August 2013 Received in revised form 28 October 2013 Accepted 30 October 2013 Available online 8 November 2013 Keywords: Autohydrolysis Gel filtration chromatography Oil palm empty fruit bunch Xylooligosaccharides
a b s t r a c t Oil palm empty fruit bunches (OPEFB) fibre, a by-product generated from non-woody, tropical perennial oil palm crop was evaluated for xylooligosaccharides (XOS) production. Samples of OPEFB fibre were subjected to non-isothermal autohydrolysis treatment using a temperature range from 150 to 220 °C. The highest XOS concentration, 17.6 g/L which relayed from solubilisation of 63 g/100 g xylan was achieved at 210 °C and there was a minimum amount of xylose and furfural being produced. The chromatographic purification which was undertaken to purify the oligosaccharide-rich liquor resulted in a product with 74–78% purity, of which 83–85% was XOS with degree of polymerisation (DP) between 5 and 40. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction A new wave of novel of non-digestible oligosaccharide (NDO) prebiotic candidates is emerging such as pectic oligosaccharides and xylooligosaccharides (XOS), which are obtained from plant cell wall polysaccharides of agro-industrial or agricultural wastes (Manderson et al., 2005; Parajó et al., 2004). The potential of these NDOs lies not only in their potential nutraceutical properties, but also in their economic benefits as they constitute an opportunity for agro-food industries to re-capture value from wastes and indirectly improves the environment through reduction of agro-food wastes and the costs associated with their management. Hydrothermal treatments such as autohydrolysis have been investigated as the practical processes for obtaining oligosaccharides, as these treatments enable the hemicelluloses to be hydrolysed selectively in a relatively short time (Garrote and Parajó, 2002). This hydrolysis mechanism has been well described in the ⇑ Corresponding author. Tel.: +44 118 378 6726; fax: +44 118 931 0080. E-mail address: [email protected] (R.A. Rastall). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.10.114
literature (Belkacemi et al., 1991; Carvalheiro et al., 2008; Garrote et al., 2002; Moure et al., 2006). Since only water is used in the process as compared to acid hydrolysis, equipment corrosion and impact on the environment are reduced (Carvalheiro et al., 2008). Nevertheless, optimisation of operational conditions is required to minimise reaction by-products such as monosaccharides and sugar degradation compounds generated from oligosaccharides thus decreasing yield. Acetic acid (a structural component of hemicelluloses) is typically also present in autohydrolysis liquors. These need to be removed to produce XOS with high purity (Moure et al., 2006). Most of the feedstock sources for XOS production studied to date have been wastes from crops of temperate climates. An alternative tropical crop source, largely from South East Asia, is oil palm (Elaeis guineensis Jacq.) wastes. Lignocellulosic residues arise from oil palm biomass either as mill by-products (oil palm empty fruit bunches – OPEFB) or field by-products (oil palm frond and trunks). The upgrade of these biomass sources, especially from the mill derived by-products such as OPEFB, is now considered an important research topic due to the potential economic impact that could be
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
derived. On average, approximately 20% of the empty fruit bunches processed for oil extraction is discarded (Astimar et al., 2002). In Malaysia, one of the world largest oil palm producers, an average of 83.8 million tonnes of EFB was produced yearly from 2007 to 2011 (MPOB, 2012). Since EFB biomass is rich in hemicelluloses, this may constitute a practical source of XOS for food applications. The present study was undertaken to determine the potential of producing XOS from EFB fibre by autohydrolysis using a stirred, high pressure reactor. A non-isothermal process approach was used in this study as it is a simple but yet effective method to determine the extent of the treatment severity on hemicellulose degradation. The hemicellulosic hydrolysates obtained were purified by gel filtration chromatography. 2. Methods 2.1. Feedstock material The dried oil palm empty fruit bunches (OPEFB) fibre was supplied by the Malaysian Oil Palm Board and was milled to pass a 1.5 mm sieve cassette. The homogenised lot was stored in a screwed cap plastic container at room temperature. 2.2. Autohydrolysis of feedstock material A 2 L stainless steel reactor (4532, Parr Instruments Co., Illinois, USA) with PID temperature controller (4842, Parr Instruments Co., Illinois, USA) was used for autohydrolysis. Feedstock was mixed with water at a liquid to solid ratio (LSR) of 8 (w/w) and continuously agitated at 150 rpm. The reactor was heated to final temperatures between 150 and 220 °C in a non-isothermal operation with a typical average heating rate (100 °C to the desired temperature) at 3.8 °C/min. This was followed by rapid cooling and the hydrolysate was separated from the solids using a manual hydraulic press (Sotel, Portugal). The effect of the treatment conditions is interpreted based on the severity factor, log Ro (Overend and Chornet, 1987). The condition resulting in the highest XOS yield was selected for production of XOS-rich liquors and subsequent purification. 2.3. Oligosaccharide purification The filtered XOS-rich liquors obtained by autohydrolysis were purified with preparative gel filtration chromatography (GFC) using ÄKTA Basic 100 with a column (BPG 100 950 mm) of Superdex 30TM (Amersham Pharmacia Biotech, Uppsala, Sweden) and with a bed volume of 4.2 L. The amount of liquor injected into the column was 400 mL at a flow rate of 15 mL/min followed by elution with deionised water at 25 mL/min. The eluate was detected by refractive index (K-2401 Knauer, Berlin, Germany) and fractions (125 mL) collected using a fraction collector. All the fractions were then freeze-dried. 2.4. Analytical methods 2.4.1. Compositional analysis of feedstock and processed solids Samples were milled to less than 0.5 mm particle size. Moisture content was determined after incubation at 105 °C ± 1 °C for at least 18 h. Ash was determined at 550 ± 5 °C for at least 5 h. Protein was measured by Kjeldahl analysis (AOAC method with N 6.25 as the conversion factor). Total extractives were measured sequentially with deionised water and ethanol 96% using a Soxhlet apparatus. Glucan, xylan, arabinan and acetyl groups were determined by a procedure provided by the National Renewable Energy Laboratory,
527
U.S. (NREL/TP-510-42618) followed by HPLC analysis (Section 2.4.2). Acid insoluble residue was recovered by filtration and considered as Klason lignin after correction for ash. 2.4.2. Compositional analysis of autohydrolysis liquor HPLC was used to quantify free glucose, xylose, arabinose, acetic acid, formic acid, levulinic acid and the furan derivatives furfural and 5-hydroxymethylfurfural, HMF. An Aminex HPX-87H column (BioRad) was used at 50 °C with 5 mM H2SO4 as mobile phase. Monosaccharides and aliphatic acids were detected by refractive index and furfural and HMF were detected by absorbance at 280 nm. Oligosaccharides were quantified indirectly after an aliquot of liquor was subjected to quantitative acid hydrolysis according to NREL/TP-510-42623. Xylooligosaccharides (XOS) were estimated based on both xylose and arabinose. The basis of calculations for the material conversion was illustrated in Carvalheiro et al. (2004). The total phenolic content was assayed spectrophotometrically by Folin Ciocalteu method (Singleton and Rossi, 1965) using gallic acid as standard. 2.4.3. Characterisation of purified oligosaccharide fractions Freeze dried fractions were dissolved in deionised water at a concentration of 10 g/L. The apparent molar mass was determined by HPLC using a BIOSEP-SEC S2000 column (Phenomenex) at 30 °C and with 50 mM NaNO3 as mobile phase at 0.7 mL/min. Xylose, maltose, raffinose, stachyose and dextrans (1–71 kDa) were used for calibration. 3. Results and discussion 3.1. Composition of the feedstock material Oil palm empty fruit bunches (OPEFB) fibre had the following composition: 36.8% cellulose (as glucan), 28.2% hemicellulose (as the sum of xylan (21.5%), arabinan 2.2%) and acetyl groups (4.6%)), 20.3% Klason lignin, 3.0% protein, 2.9% ash and 9.8% total extractives. Xylan was the main hemicellulose component while arabinose, is only present in minor amounts. The glucan and xylan content were close to those reported by Lau et al. (2010), and Timilsena et al. (2013). Compared to other agricultural by-products, OPEFB fibre has similar cellulose and Klason lignin content to corn cobs, rice husks, rye straw and wheat straw while the hemicellulose is slightly higher except when compared to corn cobs (Garrote et al., 2007; Gullón et al., 2010; Carvalheiro et al., 2009a,b). 3.2. Effect of autohydrolysis on the liquor composition Table 1 shows the OPEFB autohydrolysis liquor compositions at different temperatures. The main products were XOS and their maximum concentration was obtained at 210 °C that corresponds to production of 63 g XOS/100 g feedstock arabinoxylan; a yield that is comparable to the highest reported for corn cobs at 66% (Garrote et al., 2002). Sabiha-Hanim et al. (2011), using oil palm frond as source of XOS, showed that hydrothermal operating conditions have a significant influence on hemicellulose solubilisation from recalcitrant biomass. Using an autoclave (121 °C, 20–80 min) a maximum of 48% of the hemicellulose was hydrolysed, which indicate that the treatment conditions using the conventional autoclave were not efficient (Sabiha-Hanim et al., 2011). The gluco-oligosaccharides (GlcOS) and glucose, presumably derived from cellulose, were present at less than 1 g/L. This finding is consistent with previous studies on wood and other agricultural residues, which indicate the selectivity of autohydrolysis on
528
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
Table 1 Composition (g/L) of the liquors obtained from autohydrolysis of OPEFB fibre. Severity factor, log Ro
pH XOS GlcOS AcOS Xylose Arabinose Glucose Acetic acid Formic acid Furfural HMF Total phenolics
2.19 (150)
2.73 (170)
3.33 (190)
3.60 (200)
3.91 (210)
4.03 (215)
4.22 (220)
4.83 0.70 0.36 0.10 0.63 0.43 0.69 0.21 0.06 0.01 0.02 0.47
4.60 1.52 0.26 0.19 0.79 0.45 0.84 0.37 0.15 0.00 0.01 0.55
4.12 11.56 0.33 1.34 1.08 0.98 0.81 1.30 0.37 0.04 0.03 1.10
4.12 13.62 0.29 2.15 1.10 1.06 0.50 1.82 0.47 0.10 0.04 1.28
3.72 17.64 0.51 2.83 2.19 1.21 0.58 3.42 0.87 0.66 0.09 2.05
3.47 14.54 0.60 1.90 3.66 1.17 0.40 5.15 2.32 1.35 0.13 2.96
3.39 9.49 0.57 1.37 4.60 0.75 0.42 6.50 2.54 2.58 0.19 3.42
Value in parentheses indicate the final reaction temperature (°C). XOS – arabinose substituted xylooligosaccharides; GlcOS – glucooligosaccharides; AcOS – O-acetyl groups linked to oligosaccharides; HMF – hydroxymethylfurfural.
hemicellulose depolymerisation (Garrote and Parajó, 2002; Parajó et al., 2004). Xylose and arabinose increased gradually with treatment severity factor due to XOS hydrolysis, reaching a maximum of 4.3% (w/w feedstock). pH decreased (4.8–3.4) with increasing severity factor due to cleavage of acetyl groups to free acetic acid, in turn increasing XOS production (Carvalheiro et al., 2009a,b; Moure et al., 2006). Furfural and HMF are degradation products of pentose and hexose sugar, respectively. Under the highest XOS-yielding conditions, furfural concentration was at 0.66 g/L and increases to 2.58 g/L at the highest temperature tested. HMF was always present at a lower concentration, indicating minimal impact on hexose-based polysaccharides. Other reaction by-products that may arise from autohydrolysis treatment are formic acid and levulinic acid from degradation of furfural and HMF (Carvalheiro et al., 2008). No levulinic acid was detected in the liquor, formic acid concentration reached 0.87 g/L under highest XOS-yielding conditions. Phenolic compounds resulting from both from the extractives and the acid soluble lignin fraction were present in the liquor. Their concentration increased with treatment severity up to 3.42 g/L.
Fig. 1. Gel filtration chromatography (GFC) elution profile of OPEFB fibre liquor from autohydrolysis at 210 °C. Sample was fractionated on a BPG column (100 950 mm) eluted at 25 mL/min with refractive index detection.
purification batch was 6.41 g ± 0.48 freeze-dried hydrolysates/ 100 mL. This was lower than purified fractions obtained from olive tree pruning (7.3 g hydrolysates/100 mL; Cara et al., 2012). The overall purity from OPEFB was more than 74% and at least 83% consisted of XOS. Cara et al. (2012) obtained a purity of 71% for olive tree prunings with average DP of 5 and 82–90% purity with average DP 7–25. Compared to commercial XOS, which consists mainly of DP 2–3 with a purity of 70–95% (Moure et al., 2006), the XOS mixtures obtained in this work show an adequate purity with the advantage that a wide range of DP values were obtained. These could be used separated, or combined as required, to evaluate their prebiotic potential as well as to optimise their functional properties in food. 3.4. Oligosaccharide characterisation The recovered oligosaccharides had an estimated apparent molar mass of 700–10,500 Da corresponding to average degree of polymerisation (avDP) of 5–70. The characterisation of fractions with avDP 5–40 is presented in Table 2. As expected, the purified samples consisted mainly of xylose oligomers and purification has reduced the monosaccharides to 0.11 g/L and the acetic acid to 0.12 g/L in the avDP 5–10 fractions, while fractions with avDP 21–40 contained 0.02 g/L monosaccharides. GFC was also used in the purification of hydrolysates from olive tree prunings and corn
3.3. Oligosaccharide purification At the best conditions for oligosaccharide production, the impurities in the reaction liquor were mainly monosaccharides at 11.4%, which is at the lower end of values described for other materials using similar treatments (Carvalheiro et al., 2009a,b; Garrote et al., 2007). Conversely, for acetic acid, the value is higher (10.9%) compared with most materials. Other impurities were furfural (2.9%), HMF (0.36%), formic acid (2.8%) and phenolic compounds (6.6%). With the overall relative impurities found in reaction liquor at 35%, purification of XOS from OPEFB is essential. As monosaccharides are normally present in commercial oligosaccharides, from 5% in FOS (Raftilose P95, Orafti, Tienen, Belgium) to 15% in soy oligosaccharides (Calpis, Tokyo, Japan) and 38% in GOS (Oligomate55, Yakult, Tokyo, Japan), the aliphatic acids and furans are of most concern. As the by-products are mostly low molecular weight compounds, gel filtration chromatography can be employed to separate them from the oligosaccharides and also to fractionate the oligosaccharides with respect to degree of polymerisation (Palm and Zacchi, 2004). Fig. 1 shows the GFC elution profile of OPEFB fibre liquor under optimum XOS-yielding conditions. The average yield per
Table 2 Composition of freeze dried purified liquor with different average DP range. Composition
XOS GlcOS AcOS Ara/Xylc Acetyl/Xylc Xylose Arabinose Glucose Acetic acid Yieldb,d
Concentrationa,b (g/L) avDP 5–10 (16–18)
avDP 11–20 (14–15)
avDP 21–30 (13)
avDP 31–40 (12)
6.15 ± 0.12 0.17 ± 0.02 0.85 ± 0.08 0.02 ± 0.00 0.35 ± 0.04 0.07 ± 0.01 0.02 ± 0.00 0.02 ± 0.00 0.12 ± 0.01 6.63 ± 0.41
6.19 ± 0.24 0.16 ± 0.02 0.99 ± 0.06 0.01 ± 0.00 0.41 ± 0.03 0.04 ± 0.02 0.03 ± 0.02 0.02 ± 0.00 0.08 ± 0.06 3.62 ± 0.11
6.54 ± 0.55 0.15 ± 0.02 1.04 ± 0.01 0.01 ± 0.00 0.41 ± 0.03 0.02 ± 0.01 n.d. n.d. 0.04 ± 0.01 1.55 ± 0.15
6.29 ± 0.06 0.18 ± 0.02 1.10 ± 0.03 0.01 ± 0.00 0.44 ± 0.02 0.02 ± 0.00 n.d. n.d. 0.03 ± 0.01 1.38 ± 0.22
Numbers in parentheses indicate the corresponding sample purification fraction number to each avDP range. XOS – arabinose substituted xylooligosaccharides, GlcOS – glucooligosaccharides, AcOS – O-acetyl groups linked oligosaccharides, n.d. – not detected. a Solution of 10 g/L freeze dried sample. b Mean ± standard error based on average of three purification batches. c Ratio in mol/mol. d As in g/100 g dry wt. OPEFB fibre.
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
cob and small amounts of monosaccharides were present in the oligosaccharide fraction at average DP of 3–6 (Moura et al., 2007; Cara et al., 2012). 4. Conclusion OPEFB fibre contained xylan-rich hemicelluloses, with substantial acetyl substitution, and low ash content suggesting the choice of autohydrolysis to produce XOS. The challenge of utilising autohydrolysis hydrolysates in food applications is the presence of reaction by-products. Gel filtration chromatography however, successfully fractionated the oligosaccharide-rich liquor into fractions with different molar mass with removal of most of the undesirable low molecular weight compounds. The XOS fractions will be evaluated for their biological properties in future studies. Acknowledgement The authors are grateful to Malaysian Oil Palm Board for providing the OPEFB fibre and to Malaysian Ministry of Higher Education for scholarship funding to Ai Ling Ho. References Astimar, A.A., Kumudeswar, D., Mohamad, H., Anis, M., 2002. Effects of physical and chemical pre-treatments on xylose and glucose production from oil palm press fibre. J. Oil Palm Res. 14, 10–17. Belkacemi, K., Abatzoglou, N., Overand, R.P., Chornet, E., 1991. Phenomenological kinetics of complex systems: mechanistic considerations in the solubilization of hemicelluloses following aqueous/steam treatments. Ind. Eng. Chem. Res. 30, 2416–2425. Cara, C., Ruiz, E., Carvalheiro, F., Moura, P., Ballesteros, I., Castro, E., Gírio, F., 2012. Production, purification and characterisation of oligosaccharides from olive tree pruning autohydrolysis. Ind. Crops Prod. 40, 225–231. Carvalheiro, F., Duarte, L.C., Girio, F.M., 2008. Hemicellulose biorefineries: a review on biomass pretreatments. J. Sci. Ind. Res. 67, 849–864. Carvalheiro, F., Duarte, L.C., Silva-Fernandes, T., Lopes, S., Moura, P., Pereira, H., Girio, F.M. 2009a. Hydrothermal processing of hardwoods and agro-industrial residues: evaluation of xylooligosaccharides production. Nordic Wood Biorefinery Conference (NWBC), 2–4 September, Helsinki, pp. 96–102. Carvalheiro, F., Esteves, M.P., Parajo, J.C., Pereira, H., Girio, F.M., 2004. Production of oligosaccharides by autohydrolysis of brewery’s spent grain. Bioresour. Technol. 91, 93–100.
529
Carvalheiro, F., Silva-Fernandes, T., Duarte, L.C., Girio, F.M., 2009b. Wheat straw autohydrolysis: process optimisation and products characterisation. Appl. Biochem. Biotechnol. 153, 84–93. Garrote, G., Domínguez, H., Parajó, J.C., 2002. Autohydrolysis of corncob: study of non-isothermal operation for xylooligosaccharide production. J. Food Eng. 52, 211–218. Garrote, G., Falque, E., Domı´nguez, H., Parajó, J.C., 2007. Autohydrolysis of agricultural residues: study of reaction byproducts. Bioresour. Technol. 98, 1951–1957. Garrote, G., Parajó, J.C., 2002. Non-isothermal autohydrolysis of Eucalyptus wood. Wood Sci. Technol. 36, 111–123. Gullón, B., Yáñez, R., Alonso, J.L., Parajó, J.C., 2010. Production of oligosaccharides and sugars from rye straw: a kinetic approach. Bioresour. Technol. 101, 6676– 6684. Lau, M.J., Lau, M.W., Gunawan, C., Dale, B.E., 2010. Ammonia fiber expansion (AFEX) pretreatment, enzymatic hydrolysis, and fermentation on empty palm fruit bunch fiber (EPFBF) for cellulosic ethanol production. Appl. Biochem. Biotechnol. 162, 1847–1857. Manderson, K., Pinart, M., Tuohy, K.M., Grace, W.E., Hotchkiss, A.T., Widmer, W., Yadhav, M.P., Gibson, G.R., Rastall, R.A., 2005. In vitro determination of prebiotic properties of oligosaccharides derived from an orange juice manufacturing byproduct stream. Appl. Environ. Microbiol. 71, 8383–8389. Moura, P., Barata, R., Carvalheiro, F., Gírio, F., Loureiro-Dias, M.C., Esteves, M.P., 2007. In vitro fermentation of xylo-oligosaccharides from corn cobs autohydrolysis by Bifidobacterium and Lactobacillus strains. LWT – Food Sci. Technol. 40, 963–972. Moure, A., Gullón, P., Domínguez, H., Parajó, J.C., 2006. Advances in the manufacture, purification and applications of xylo-oligosaccharides as food additives and nutraceuticals. Process Biochem. 41, 1913–1923. MPOB, 2012. Sectoral status: Fresh fruit bunch (FFB) processed by mill [Online]. Selangor: Malaysian Oil Palm Board (MPOB). Available: (accessed 23. 05.12.). Overend, R.P., Chornet, E., 1987. Fractionation of lignocellulosics by steam–aqueous pretreatments. Philos. Trans. R. Soc. London, Ser. A 321, 523–536. Palm, M., Zacchi, G., 2004. Separation of hemicellulosic oligomers from steamtreated spruce wood using gel filtration. Sep. Purif. Technol. 36, 191–201. Parajó, J.C., Garrote, G., Cruz, J.M., Domínguez, H., 2004. Production of xylooligosaccharides by autohydrolysis of lignocellulosic materials. Trends Food Sci. Technol. 15, 115–120. Sabiha-Hanim, S., Mohd Azemi, M.N., Ahmad, R., 2011. Effect of autohydrolysis and enzymatic treatment on oil palm (Elaeis guineensis Jacq.) frond fibres for xylose and xylooligosaccharides production. Bioresour. Technol. 102, 1234–1239. Singleton, V.L., Rossi, J.A.J., 1965. Colorimetry of total phenolics with phosphomolybdic–phosphotungstic acid reagents. Am. J. Enol. Vitic. 16, 144– 158. Timilsena, Y.P., Abeywickrama, C.J., Rakshit, S.K., Brosse, N., 2013. Effect of different pretreatments on delignification pattern and enzymatic hydrolysability of miscanthus, oil palm biomass and typha grass. Bioresour. Technol. 135, 82–88.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Short Communication
Production and purification of xylooligosaccharides from oil palm empty fruit bunch fibre by a non-isothermal process Ai Ling Ho a, Florbela Carvalheiro b, Luís C. Duarte b, Luísa B. Roseiro b, Dimitris Charalampopoulos a, Robert A. Rastall a,⇑ a b
Department of Food and Nutritional Sciences, University of Reading, Whiteknights, P.O. Box 226, Reading RG6 6AP, United Kingdom LNEG, Unidade de Bioenergia, Estrada do Paço do Lumiar 22, 1649-038 Lisboa, Portugal
h i g h l i g h t s Information on oligosaccharides production from oil palm biomass is very scarce. Oligosaccharides production from OPEFB fibre via autohydrolysis is demonstrated. Substantial yield of XOS is achieved using non-isothermal conditions. High purity of refined XOS with wide ranges of DP were obtained.
a r t i c l e
i n f o
Article history: Received 28 August 2013 Received in revised form 28 October 2013 Accepted 30 October 2013 Available online 8 November 2013 Keywords: Autohydrolysis Gel filtration chromatography Oil palm empty fruit bunch Xylooligosaccharides
a b s t r a c t Oil palm empty fruit bunches (OPEFB) fibre, a by-product generated from non-woody, tropical perennial oil palm crop was evaluated for xylooligosaccharides (XOS) production. Samples of OPEFB fibre were subjected to non-isothermal autohydrolysis treatment using a temperature range from 150 to 220 °C. The highest XOS concentration, 17.6 g/L which relayed from solubilisation of 63 g/100 g xylan was achieved at 210 °C and there was a minimum amount of xylose and furfural being produced. The chromatographic purification which was undertaken to purify the oligosaccharide-rich liquor resulted in a product with 74–78% purity, of which 83–85% was XOS with degree of polymerisation (DP) between 5 and 40. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction A new wave of novel of non-digestible oligosaccharide (NDO) prebiotic candidates is emerging such as pectic oligosaccharides and xylooligosaccharides (XOS), which are obtained from plant cell wall polysaccharides of agro-industrial or agricultural wastes (Manderson et al., 2005; Parajó et al., 2004). The potential of these NDOs lies not only in their potential nutraceutical properties, but also in their economic benefits as they constitute an opportunity for agro-food industries to re-capture value from wastes and indirectly improves the environment through reduction of agro-food wastes and the costs associated with their management. Hydrothermal treatments such as autohydrolysis have been investigated as the practical processes for obtaining oligosaccharides, as these treatments enable the hemicelluloses to be hydrolysed selectively in a relatively short time (Garrote and Parajó, 2002). This hydrolysis mechanism has been well described in the ⇑ Corresponding author. Tel.: +44 118 378 6726; fax: +44 118 931 0080. E-mail address: [email protected] (R.A. Rastall). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.10.114
literature (Belkacemi et al., 1991; Carvalheiro et al., 2008; Garrote et al., 2002; Moure et al., 2006). Since only water is used in the process as compared to acid hydrolysis, equipment corrosion and impact on the environment are reduced (Carvalheiro et al., 2008). Nevertheless, optimisation of operational conditions is required to minimise reaction by-products such as monosaccharides and sugar degradation compounds generated from oligosaccharides thus decreasing yield. Acetic acid (a structural component of hemicelluloses) is typically also present in autohydrolysis liquors. These need to be removed to produce XOS with high purity (Moure et al., 2006). Most of the feedstock sources for XOS production studied to date have been wastes from crops of temperate climates. An alternative tropical crop source, largely from South East Asia, is oil palm (Elaeis guineensis Jacq.) wastes. Lignocellulosic residues arise from oil palm biomass either as mill by-products (oil palm empty fruit bunches – OPEFB) or field by-products (oil palm frond and trunks). The upgrade of these biomass sources, especially from the mill derived by-products such as OPEFB, is now considered an important research topic due to the potential economic impact that could be
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
derived. On average, approximately 20% of the empty fruit bunches processed for oil extraction is discarded (Astimar et al., 2002). In Malaysia, one of the world largest oil palm producers, an average of 83.8 million tonnes of EFB was produced yearly from 2007 to 2011 (MPOB, 2012). Since EFB biomass is rich in hemicelluloses, this may constitute a practical source of XOS for food applications. The present study was undertaken to determine the potential of producing XOS from EFB fibre by autohydrolysis using a stirred, high pressure reactor. A non-isothermal process approach was used in this study as it is a simple but yet effective method to determine the extent of the treatment severity on hemicellulose degradation. The hemicellulosic hydrolysates obtained were purified by gel filtration chromatography. 2. Methods 2.1. Feedstock material The dried oil palm empty fruit bunches (OPEFB) fibre was supplied by the Malaysian Oil Palm Board and was milled to pass a 1.5 mm sieve cassette. The homogenised lot was stored in a screwed cap plastic container at room temperature. 2.2. Autohydrolysis of feedstock material A 2 L stainless steel reactor (4532, Parr Instruments Co., Illinois, USA) with PID temperature controller (4842, Parr Instruments Co., Illinois, USA) was used for autohydrolysis. Feedstock was mixed with water at a liquid to solid ratio (LSR) of 8 (w/w) and continuously agitated at 150 rpm. The reactor was heated to final temperatures between 150 and 220 °C in a non-isothermal operation with a typical average heating rate (100 °C to the desired temperature) at 3.8 °C/min. This was followed by rapid cooling and the hydrolysate was separated from the solids using a manual hydraulic press (Sotel, Portugal). The effect of the treatment conditions is interpreted based on the severity factor, log Ro (Overend and Chornet, 1987). The condition resulting in the highest XOS yield was selected for production of XOS-rich liquors and subsequent purification. 2.3. Oligosaccharide purification The filtered XOS-rich liquors obtained by autohydrolysis were purified with preparative gel filtration chromatography (GFC) using ÄKTA Basic 100 with a column (BPG 100 950 mm) of Superdex 30TM (Amersham Pharmacia Biotech, Uppsala, Sweden) and with a bed volume of 4.2 L. The amount of liquor injected into the column was 400 mL at a flow rate of 15 mL/min followed by elution with deionised water at 25 mL/min. The eluate was detected by refractive index (K-2401 Knauer, Berlin, Germany) and fractions (125 mL) collected using a fraction collector. All the fractions were then freeze-dried. 2.4. Analytical methods 2.4.1. Compositional analysis of feedstock and processed solids Samples were milled to less than 0.5 mm particle size. Moisture content was determined after incubation at 105 °C ± 1 °C for at least 18 h. Ash was determined at 550 ± 5 °C for at least 5 h. Protein was measured by Kjeldahl analysis (AOAC method with N 6.25 as the conversion factor). Total extractives were measured sequentially with deionised water and ethanol 96% using a Soxhlet apparatus. Glucan, xylan, arabinan and acetyl groups were determined by a procedure provided by the National Renewable Energy Laboratory,
527
U.S. (NREL/TP-510-42618) followed by HPLC analysis (Section 2.4.2). Acid insoluble residue was recovered by filtration and considered as Klason lignin after correction for ash. 2.4.2. Compositional analysis of autohydrolysis liquor HPLC was used to quantify free glucose, xylose, arabinose, acetic acid, formic acid, levulinic acid and the furan derivatives furfural and 5-hydroxymethylfurfural, HMF. An Aminex HPX-87H column (BioRad) was used at 50 °C with 5 mM H2SO4 as mobile phase. Monosaccharides and aliphatic acids were detected by refractive index and furfural and HMF were detected by absorbance at 280 nm. Oligosaccharides were quantified indirectly after an aliquot of liquor was subjected to quantitative acid hydrolysis according to NREL/TP-510-42623. Xylooligosaccharides (XOS) were estimated based on both xylose and arabinose. The basis of calculations for the material conversion was illustrated in Carvalheiro et al. (2004). The total phenolic content was assayed spectrophotometrically by Folin Ciocalteu method (Singleton and Rossi, 1965) using gallic acid as standard. 2.4.3. Characterisation of purified oligosaccharide fractions Freeze dried fractions were dissolved in deionised water at a concentration of 10 g/L. The apparent molar mass was determined by HPLC using a BIOSEP-SEC S2000 column (Phenomenex) at 30 °C and with 50 mM NaNO3 as mobile phase at 0.7 mL/min. Xylose, maltose, raffinose, stachyose and dextrans (1–71 kDa) were used for calibration. 3. Results and discussion 3.1. Composition of the feedstock material Oil palm empty fruit bunches (OPEFB) fibre had the following composition: 36.8% cellulose (as glucan), 28.2% hemicellulose (as the sum of xylan (21.5%), arabinan 2.2%) and acetyl groups (4.6%)), 20.3% Klason lignin, 3.0% protein, 2.9% ash and 9.8% total extractives. Xylan was the main hemicellulose component while arabinose, is only present in minor amounts. The glucan and xylan content were close to those reported by Lau et al. (2010), and Timilsena et al. (2013). Compared to other agricultural by-products, OPEFB fibre has similar cellulose and Klason lignin content to corn cobs, rice husks, rye straw and wheat straw while the hemicellulose is slightly higher except when compared to corn cobs (Garrote et al., 2007; Gullón et al., 2010; Carvalheiro et al., 2009a,b). 3.2. Effect of autohydrolysis on the liquor composition Table 1 shows the OPEFB autohydrolysis liquor compositions at different temperatures. The main products were XOS and their maximum concentration was obtained at 210 °C that corresponds to production of 63 g XOS/100 g feedstock arabinoxylan; a yield that is comparable to the highest reported for corn cobs at 66% (Garrote et al., 2002). Sabiha-Hanim et al. (2011), using oil palm frond as source of XOS, showed that hydrothermal operating conditions have a significant influence on hemicellulose solubilisation from recalcitrant biomass. Using an autoclave (121 °C, 20–80 min) a maximum of 48% of the hemicellulose was hydrolysed, which indicate that the treatment conditions using the conventional autoclave were not efficient (Sabiha-Hanim et al., 2011). The gluco-oligosaccharides (GlcOS) and glucose, presumably derived from cellulose, were present at less than 1 g/L. This finding is consistent with previous studies on wood and other agricultural residues, which indicate the selectivity of autohydrolysis on
528
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
Table 1 Composition (g/L) of the liquors obtained from autohydrolysis of OPEFB fibre. Severity factor, log Ro
pH XOS GlcOS AcOS Xylose Arabinose Glucose Acetic acid Formic acid Furfural HMF Total phenolics
2.19 (150)
2.73 (170)
3.33 (190)
3.60 (200)
3.91 (210)
4.03 (215)
4.22 (220)
4.83 0.70 0.36 0.10 0.63 0.43 0.69 0.21 0.06 0.01 0.02 0.47
4.60 1.52 0.26 0.19 0.79 0.45 0.84 0.37 0.15 0.00 0.01 0.55
4.12 11.56 0.33 1.34 1.08 0.98 0.81 1.30 0.37 0.04 0.03 1.10
4.12 13.62 0.29 2.15 1.10 1.06 0.50 1.82 0.47 0.10 0.04 1.28
3.72 17.64 0.51 2.83 2.19 1.21 0.58 3.42 0.87 0.66 0.09 2.05
3.47 14.54 0.60 1.90 3.66 1.17 0.40 5.15 2.32 1.35 0.13 2.96
3.39 9.49 0.57 1.37 4.60 0.75 0.42 6.50 2.54 2.58 0.19 3.42
Value in parentheses indicate the final reaction temperature (°C). XOS – arabinose substituted xylooligosaccharides; GlcOS – glucooligosaccharides; AcOS – O-acetyl groups linked to oligosaccharides; HMF – hydroxymethylfurfural.
hemicellulose depolymerisation (Garrote and Parajó, 2002; Parajó et al., 2004). Xylose and arabinose increased gradually with treatment severity factor due to XOS hydrolysis, reaching a maximum of 4.3% (w/w feedstock). pH decreased (4.8–3.4) with increasing severity factor due to cleavage of acetyl groups to free acetic acid, in turn increasing XOS production (Carvalheiro et al., 2009a,b; Moure et al., 2006). Furfural and HMF are degradation products of pentose and hexose sugar, respectively. Under the highest XOS-yielding conditions, furfural concentration was at 0.66 g/L and increases to 2.58 g/L at the highest temperature tested. HMF was always present at a lower concentration, indicating minimal impact on hexose-based polysaccharides. Other reaction by-products that may arise from autohydrolysis treatment are formic acid and levulinic acid from degradation of furfural and HMF (Carvalheiro et al., 2008). No levulinic acid was detected in the liquor, formic acid concentration reached 0.87 g/L under highest XOS-yielding conditions. Phenolic compounds resulting from both from the extractives and the acid soluble lignin fraction were present in the liquor. Their concentration increased with treatment severity up to 3.42 g/L.
Fig. 1. Gel filtration chromatography (GFC) elution profile of OPEFB fibre liquor from autohydrolysis at 210 °C. Sample was fractionated on a BPG column (100 950 mm) eluted at 25 mL/min with refractive index detection.
purification batch was 6.41 g ± 0.48 freeze-dried hydrolysates/ 100 mL. This was lower than purified fractions obtained from olive tree pruning (7.3 g hydrolysates/100 mL; Cara et al., 2012). The overall purity from OPEFB was more than 74% and at least 83% consisted of XOS. Cara et al. (2012) obtained a purity of 71% for olive tree prunings with average DP of 5 and 82–90% purity with average DP 7–25. Compared to commercial XOS, which consists mainly of DP 2–3 with a purity of 70–95% (Moure et al., 2006), the XOS mixtures obtained in this work show an adequate purity with the advantage that a wide range of DP values were obtained. These could be used separated, or combined as required, to evaluate their prebiotic potential as well as to optimise their functional properties in food. 3.4. Oligosaccharide characterisation The recovered oligosaccharides had an estimated apparent molar mass of 700–10,500 Da corresponding to average degree of polymerisation (avDP) of 5–70. The characterisation of fractions with avDP 5–40 is presented in Table 2. As expected, the purified samples consisted mainly of xylose oligomers and purification has reduced the monosaccharides to 0.11 g/L and the acetic acid to 0.12 g/L in the avDP 5–10 fractions, while fractions with avDP 21–40 contained 0.02 g/L monosaccharides. GFC was also used in the purification of hydrolysates from olive tree prunings and corn
3.3. Oligosaccharide purification At the best conditions for oligosaccharide production, the impurities in the reaction liquor were mainly monosaccharides at 11.4%, which is at the lower end of values described for other materials using similar treatments (Carvalheiro et al., 2009a,b; Garrote et al., 2007). Conversely, for acetic acid, the value is higher (10.9%) compared with most materials. Other impurities were furfural (2.9%), HMF (0.36%), formic acid (2.8%) and phenolic compounds (6.6%). With the overall relative impurities found in reaction liquor at 35%, purification of XOS from OPEFB is essential. As monosaccharides are normally present in commercial oligosaccharides, from 5% in FOS (Raftilose P95, Orafti, Tienen, Belgium) to 15% in soy oligosaccharides (Calpis, Tokyo, Japan) and 38% in GOS (Oligomate55, Yakult, Tokyo, Japan), the aliphatic acids and furans are of most concern. As the by-products are mostly low molecular weight compounds, gel filtration chromatography can be employed to separate them from the oligosaccharides and also to fractionate the oligosaccharides with respect to degree of polymerisation (Palm and Zacchi, 2004). Fig. 1 shows the GFC elution profile of OPEFB fibre liquor under optimum XOS-yielding conditions. The average yield per
Table 2 Composition of freeze dried purified liquor with different average DP range. Composition
XOS GlcOS AcOS Ara/Xylc Acetyl/Xylc Xylose Arabinose Glucose Acetic acid Yieldb,d
Concentrationa,b (g/L) avDP 5–10 (16–18)
avDP 11–20 (14–15)
avDP 21–30 (13)
avDP 31–40 (12)
6.15 ± 0.12 0.17 ± 0.02 0.85 ± 0.08 0.02 ± 0.00 0.35 ± 0.04 0.07 ± 0.01 0.02 ± 0.00 0.02 ± 0.00 0.12 ± 0.01 6.63 ± 0.41
6.19 ± 0.24 0.16 ± 0.02 0.99 ± 0.06 0.01 ± 0.00 0.41 ± 0.03 0.04 ± 0.02 0.03 ± 0.02 0.02 ± 0.00 0.08 ± 0.06 3.62 ± 0.11
6.54 ± 0.55 0.15 ± 0.02 1.04 ± 0.01 0.01 ± 0.00 0.41 ± 0.03 0.02 ± 0.01 n.d. n.d. 0.04 ± 0.01 1.55 ± 0.15
6.29 ± 0.06 0.18 ± 0.02 1.10 ± 0.03 0.01 ± 0.00 0.44 ± 0.02 0.02 ± 0.00 n.d. n.d. 0.03 ± 0.01 1.38 ± 0.22
Numbers in parentheses indicate the corresponding sample purification fraction number to each avDP range. XOS – arabinose substituted xylooligosaccharides, GlcOS – glucooligosaccharides, AcOS – O-acetyl groups linked oligosaccharides, n.d. – not detected. a Solution of 10 g/L freeze dried sample. b Mean ± standard error based on average of three purification batches. c Ratio in mol/mol. d As in g/100 g dry wt. OPEFB fibre.
A.L. Ho et al. / Bioresource Technology 152 (2014) 526–529
cob and small amounts of monosaccharides were present in the oligosaccharide fraction at average DP of 3–6 (Moura et al., 2007; Cara et al., 2012). 4. Conclusion OPEFB fibre contained xylan-rich hemicelluloses, with substantial acetyl substitution, and low ash content suggesting the choice of autohydrolysis to produce XOS. The challenge of utilising autohydrolysis hydrolysates in food applications is the presence of reaction by-products. Gel filtration chromatography however, successfully fractionated the oligosaccharide-rich liquor into fractions with different molar mass with removal of most of the undesirable low molecular weight compounds. The XOS fractions will be evaluated for their biological properties in future studies. Acknowledgement The authors are grateful to Malaysian Oil Palm Board for providing the OPEFB fibre and to Malaysian Ministry of Higher Education for scholarship funding to Ai Ling Ho. References Astimar, A.A., Kumudeswar, D., Mohamad, H., Anis, M., 2002. Effects of physical and chemical pre-treatments on xylose and glucose production from oil palm press fibre. J. Oil Palm Res. 14, 10–17. Belkacemi, K., Abatzoglou, N., Overand, R.P., Chornet, E., 1991. Phenomenological kinetics of complex systems: mechanistic considerations in the solubilization of hemicelluloses following aqueous/steam treatments. Ind. Eng. Chem. Res. 30, 2416–2425. Cara, C., Ruiz, E., Carvalheiro, F., Moura, P., Ballesteros, I., Castro, E., Gírio, F., 2012. Production, purification and characterisation of oligosaccharides from olive tree pruning autohydrolysis. Ind. Crops Prod. 40, 225–231. Carvalheiro, F., Duarte, L.C., Girio, F.M., 2008. Hemicellulose biorefineries: a review on biomass pretreatments. J. Sci. Ind. Res. 67, 849–864. Carvalheiro, F., Duarte, L.C., Silva-Fernandes, T., Lopes, S., Moura, P., Pereira, H., Girio, F.M. 2009a. Hydrothermal processing of hardwoods and agro-industrial residues: evaluation of xylooligosaccharides production. Nordic Wood Biorefinery Conference (NWBC), 2–4 September, Helsinki, pp. 96–102. Carvalheiro, F., Esteves, M.P., Parajo, J.C., Pereira, H., Girio, F.M., 2004. Production of oligosaccharides by autohydrolysis of brewery’s spent grain. Bioresour. Technol. 91, 93–100.
529
Carvalheiro, F., Silva-Fernandes, T., Duarte, L.C., Girio, F.M., 2009b. Wheat straw autohydrolysis: process optimisation and products characterisation. Appl. Biochem. Biotechnol. 153, 84–93. Garrote, G., Domínguez, H., Parajó, J.C., 2002. Autohydrolysis of corncob: study of non-isothermal operation for xylooligosaccharide production. J. Food Eng. 52, 211–218. Garrote, G., Falque, E., Domı´nguez, H., Parajó, J.C., 2007. Autohydrolysis of agricultural residues: study of reaction byproducts. Bioresour. Technol. 98, 1951–1957. Garrote, G., Parajó, J.C., 2002. Non-isothermal autohydrolysis of Eucalyptus wood. Wood Sci. Technol. 36, 111–123. Gullón, B., Yáñez, R., Alonso, J.L., Parajó, J.C., 2010. Production of oligosaccharides and sugars from rye straw: a kinetic approach. Bioresour. Technol. 101, 6676– 6684. Lau, M.J., Lau, M.W., Gunawan, C., Dale, B.E., 2010. Ammonia fiber expansion (AFEX) pretreatment, enzymatic hydrolysis, and fermentation on empty palm fruit bunch fiber (EPFBF) for cellulosic ethanol production. Appl. Biochem. Biotechnol. 162, 1847–1857. Manderson, K., Pinart, M., Tuohy, K.M., Grace, W.E., Hotchkiss, A.T., Widmer, W., Yadhav, M.P., Gibson, G.R., Rastall, R.A., 2005. In vitro determination of prebiotic properties of oligosaccharides derived from an orange juice manufacturing byproduct stream. Appl. Environ. Microbiol. 71, 8383–8389. Moura, P., Barata, R., Carvalheiro, F., Gírio, F., Loureiro-Dias, M.C., Esteves, M.P., 2007. In vitro fermentation of xylo-oligosaccharides from corn cobs autohydrolysis by Bifidobacterium and Lactobacillus strains. LWT – Food Sci. Technol. 40, 963–972. Moure, A., Gullón, P., Domínguez, H., Parajó, J.C., 2006. Advances in the manufacture, purification and applications of xylo-oligosaccharides as food additives and nutraceuticals. Process Biochem. 41, 1913–1923. MPOB, 2012. Sectoral status: Fresh fruit bunch (FFB) processed by mill [Online]. Selangor: Malaysian Oil Palm Board (MPOB). Available: (accessed 23. 05.12.). Overend, R.P., Chornet, E., 1987. Fractionation of lignocellulosics by steam–aqueous pretreatments. Philos. Trans. R. Soc. London, Ser. A 321, 523–536. Palm, M., Zacchi, G., 2004. Separation of hemicellulosic oligomers from steamtreated spruce wood using gel filtration. Sep. Purif. Technol. 36, 191–201. Parajó, J.C., Garrote, G., Cruz, J.M., Domínguez, H., 2004. Production of xylooligosaccharides by autohydrolysis of lignocellulosic materials. Trends Food Sci. Technol. 15, 115–120. Sabiha-Hanim, S., Mohd Azemi, M.N., Ahmad, R., 2011. Effect of autohydrolysis and enzymatic treatment on oil palm (Elaeis guineensis Jacq.) frond fibres for xylose and xylooligosaccharides production. Bioresour. Technol. 102, 1234–1239. Singleton, V.L., Rossi, J.A.J., 1965. Colorimetry of total phenolics with phosphomolybdic–phosphotungstic acid reagents. Am. J. Enol. Vitic. 16, 144– 158. Timilsena, Y.P., Abeywickrama, C.J., Rakshit, S.K., Brosse, N., 2013. Effect of different pretreatments on delignification pattern and enzymatic hydrolysability of miscanthus, oil palm biomass and typha grass. Bioresour. Technol. 135, 82–88.
