Ball Milling Pretreatment of Oil Palm Biomassfor Enhancing Enzymatic Hydrolysis

Mohd Rafein Zakaria & Shinji Fujimoto & Satoshi Hirata &

Mohd Ali Hassan

Received: 11 March 2014 /Accepted: 15 May 2014# Springer Science+Business Media New York 2014

Abstract Oil palm biomass, namely empty fruit bunch and frond fiber, were pretreated usinga planetary ball mill. Particle sizes and crystallinity index values of the oil palm biomass weresignificantly reduced with extended ball mill processing time. The treatment efficiency wasevaluated by the generation of glucose, xylose, and total sugar conversion yields from thepretreatment process compared to the amount of sugars from raw materials. Glucose andxylose contents were determined using high-performance liquid chromatography. An increas-ing trend in glucose and xylose yield as well as total sugar conversion yield was observed withdecreasing particle size and crystallinity index. Oil palm frond fiber exhibited the best materialyields using ball milling pretreatment with generated glucose, xylose, and total sugar conver-sion yields of 87.0, 81.6, and 85.4 %, respectively. In contrast, oil palm empty fruit bunchafforded glucose and xylose of 70.0 and 82.3 %, respectively. The results obtained in this studyshowed that ball mill-treated oil palm biomass is a suitable pretreatment method for highconversion of glucose and xylose.

Keywords Ball milling . Pretreatment . Oil palm empty fruit bunch . Oil palm frond fiber .

Enzymatic hydrolysis

Introduction

Malaysia is the second largest palm oil producer and exporter after Indonesia (2012 season),with 18.7 million tonnes of crude palm oil production over a cultivated area of 4.917million ha [1]. Oil palm processing leads to the generation of several waste by-products,namely oil palm empty fruit bunch (OPEFB), oil palm mesocarp fiber (OPMF), and oil palmfrond fiber (OPFF) at both plantations and oil processing mills. Approximately 1–2 fronds are

Appl Biochem BiotechnolDOI 10.1007/s12010-014-0964-5

M. R. Zakaria (*) : S. Fujimoto : S. HirataBiomass Refinery Research Center, National Institute of Advanced Industrial Science and Technology(AIST), 3-11-32 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-0046, Japane-mail: rafein-zakaria@aist.go.jp

M. R. Zakaria :M. A. HassanDepartment of Bioprocess Technology, Faculty of Biotechnology and Biomolecular Sciences, UniversitiPutra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

removed during the pruning of one fresh fruit bunch (FFB) and left to decompose on theground as a natural carbon and nitrogen source for the oil palm plants. Meanwhile, OPEFBand OPMF are employed as a source of solid fuel in boiler systems where produced steampasses through a steam turbine to generate electricity for the mill’s internal usage [2].

Over the past decade, many studies have focused on the efficient use of oil palm biomass infeedstock chemical production and as potential biofuels, as these materials are abundant,inexpensive, and readily available [3]. Oil palm biomass is a lignocellulosic organic materialcomprising mainly three components: cellulose, hemicellulose, and lignin. These organiccomponents are incorporated to form a strong cellulose-hemicellulose-lignin complex withinthe plant material. In order to get maximum access to cellulose, this cellulose-hemicellulose-lignin network must be destroyed by physical, chemical, or biological treatments. Such pre-treatment is a crucial step in successful enzymatic saccharification, which involves destructionof the lignin seal, removal of hemicelluloses, increase surface area and pore volume, disruptionof the crystalline structure of cellulose, and an increase in the amorphization of cellulose [4, 5].

In Malaysia, OPEFB is the most studied lignocellulosic waste material that has the potentialto be converted to glucose monosaccharides via enzymatic saccharification after pretreatmentprocesses; however, a limited number of studies have been reported on similar conversions ofOPFF. Acid-alkaline treatment and other chemical pretreatments have been reported to resultin high removal of hemicellulose and lignin as well as a higher recovery of cellulose [6, 7]compared to untreated oil palm biomass. Nevertheless, these pretreatments lead to a high loadof chemicals, material loss, and formation of fermentation inhibitors [8, 9]. Pretreatmentsinvolving compressed hot water, steam, and superheated steam are gaining interest recently, asthese treatments offer an environmentally friendly process without the addition of chemicals,are time-saving (operated at shorter time), and limit the production of inhibitors [4, 10–12]. Tothe best of our knowledge, there are no reports thus far on the pretreatment of oil palm biomass(OPEFB and OPFF) using physical/mechanochemical pretreatment for increasing surface areaby the reduction of raw materials to nanoscale particle sizes.

Ball milling is one of the most widely used mechanical activation processes to increasesurface area of lignocellulosic biomass. Despite the high energy required, the ball millingpretreatment of lignocellulosic residues has a great impact on the reduction of particle sizes viachanges in cellulose crystalline structures and chemical bonding distortions due to imposedstress, as well as decreases in crystallinity and degrees of polymerization of cellulose.Furthermore, this process is simple yet no additional chemicals are required and no otherfunctional groups are generated [8, 9, 13, 14]. Thus, the aim of this study was to determine theeffect of ball milling pretreatment on physical properties of oil palm biomass (OPEFB andOPFF) by means of particle size reduction, cellulose crystallinity index, and surface morphol-ogy changes over milling time. Efficient pretreatment by this process was demonstrated by thehigh recovery of glucose, xylose, and total conversion yields from enzymatic hydrolysis oftreated oil palm biomass.

Materials and Methods

Raw Materials

The OPEFB was collected from Seri Ulu Langat Palm Oil Mill, Dengkil, Selangor, Malaysia,and OPFF was obtained from the oil palm plantation at Universiti Putra Malaysia, Serdang,Selangor, Malaysia. All biomass samples were sun-dried and then cut by milling cutters.Particles that pass through a sieve with a 2-mm size were used for further studies.

Appl Biochem Biotechnol

Compositional Analysis of Oil Palm Biomass

The oil palm biomass samples were ground using a Pulverisette 15 cutting mill(Fritsch, Germany) to a particle size of 0.25 mm and dried in vacuo at 40 °C priorto use. The oil palm biomass components were determined using a method recom-mended by Sluiter et al. [15].

Ball Milling Pretreatment

The ball milling (BM) pretreatment was performed according to the method reported by Silvaet al. [16]. Raw materials (2-mm particle size, dried in vacuo at 40 °C) were treated using thePulverisette 5 planetary ball mill (Fritsch, Germany). The sample (20 g) was milled at 250 rpmin a 500-mL milling cup with 25 spheres (ψ=20 mm). Milling was carried out for a total timeof 3–120 min (with a cycle of 10-min run and 10-min pause) at room temperature. The millingtime chosen was dependent on the types of biomass applied. All experiments were performedin duplicate. The BM time indicated in this study refers to the actual milling time, excludingthe paused time. Samples were stored in vacuo at room temperature prior to enzymaticsaccharification.

Enzymatic Hydrolysis

Enzymatic hydrolysis was performed using an enzyme cocktail comprising 40 FPU/mLAcremonium cellulase (Meiji Seika Co, Japan) and 10 % OPTIMASH BG (Genencor Inter-national, CA, USA). In a standard assay, 2.5 mL of 1.0 M acetate buffer (pH 5.0), 0.75 mL(10 FPU/g substrate) of Acremonium cellulase, and 0.6 mL of 10 % OPTIMASH BG wereadded to 3 g of BM samples (dry weight) in a 65-mL tube with cap (NEG, Japan). Themixtures were diluted with distilled water to a total volume of 50 mL. Enzymatic hydrolysiswas performed at 50 °C for 72 h with stirring; these hydrolysis reactions were performed intriplicate. The enzymatic digestibility was expressed by the obtained sugars (mg sugars/gmaterials) or sugar yield as calculated using Eq. (1):

Sugar yield %ð Þ ¼ ½weight of monomeric sugars after enzymatic hydrolysis=

weight of potential total monomeric sugars after hydrolysis using H2SO4� � 100

ð1Þ

Wide-Angle X-ray Diffraction

Wide-angle X-ray diffraction (WAXD) patterns were obtained using a Rigaku RINT-TTR IIIX-ray diffractometer (Tokyo, Japan) equipped with nickel-filtered Cu Kα radiation (λ=0.1542 nm) at 50 kV and 300 mA. The disk pellets were prepared by compacting oven-dried samples at 2-t pressures using a KBr disk apparatus. The diffractograms were obtained inthe range 2θ=2−60° at a scan rate of 2°/min. The crystallinity index (CrI) was calculated usingEq. (2) based on the method of Segal et al. [17]:

CrI %ð Þ ¼ I002−Iam.I002

h i� 100 ð2Þ

where I002 is the intensity at about 2θ=22.5° and Iam is the intensity at 2θ=18.7°.

Appl Biochem Biotechnol

Scanning Electron Microscopy Analysis

The unmilled and BM-treated lignocellulosic biomass samples (OPEFB and OPFF) weresputtered with Pt-Pd for 100 s (ion sputter; Hitachi, Japan). The coated samples were examinedby field emission scanning electron microscopy (FE-SEM; S-3400 N, Hitachi, Japan) at 1 kV.

Particle Size Analysis

The sample particle size distributions and mean values were determined using a LA-920HORIBA laser scattering particle size distribution analyzer. The particle size was calculatedbased on the average value of the particle’s geometrical lengths measured from the differentorientations of incident scattering light.

High-Performance Liquid Chromatography Analysis

Detection of sugars before and after enzymatic hydrolysis was performed using high-performance liquid chromatography (HPLC) equipped with a refractive index detector (RID-10A, Shimadzu, Japan) using an Aminex HPX-87P column (7.8×30 cm I.D., BioRad, USA)with a Carbo-P micro-guard cartridge. The column oven was set at 80 °C, and samples wereeluted at 0.60 mL/min with water. Acetic acid, furfural, 5-hydroxymethylfurfural (5-HMF),and other chemical compounds were analyzed as reported earlier [18].

Results and Discussion

Chemical Compositions of Oil Palm Biomass

Chemical compositions of untreated oil palm biomass are shown in Table 1. The cellulose,hemicellulose, and lignin contents of OPEFB and OPFF found in this study varied among thebiomass types. The major component found in OPEFB and OPFF was cellulose followed byhemicelluloses and lignin. The results obtained were expected, as diverse compositions ofcellulose, hemicellulose, and lignin are present in different plant tissues. Furthermore, theratios between various constituents within a single plant vary among the stages of growth, age,and other conditions [19]. Because different compositions of hemicellulose and lignin weremeasured, it is crucial that appropriate methods are used to break down the strong bonds in thecellulose-hemicellulose-lignin matrix to make it more accessible and amenable to enzymatic

Table 1 Compositions of oil palmbiomass

The experiments were performedin triplicatesOPEFB oil palm empty fruit bunch,OPFF oil palm frond fiber

Content (wt%)

Components OPEFB OPFF

Cellulose 40.4±2.4 32.7±4.9

Hemicellulose 20.2±2.3 22.5±4.9

Lignin 23.1±0.5 15.2±0.6

Acetone extractives 2.5±1.9 6.4±1.4

Moisture 0.2±1.5 5.0±1.9

Ash 5.9±0.25 3.4±0.2

Appl Biochem Biotechnol

hydrolysis by cellulase. The effects of mechanical activation by BM pretreatment of oil palmbiomass were next studied and are discussed in a later section.

Effect of BM on Size Reduction

The main purpose of this work was to carry out a preliminary study of BM pretreatment of oilpalm biomass over an appropriate milling period to reduce the particle size of lignocellulosicbiomass, thus enhancing its enzymatic digestibility and biosugar production. After BMtreatment over an appropriate period, a significant reduction in oil palm biomass particle sizecould be clearly distinguished compared to its original size (<2 mm), as shown in Table 2. Thetwo oil palm biomass types responded differently over the examined milling times to producea fine powder. OPEFB was ball milled from 0-120 min, achieving a size reduction of about61.1 % when milled for 40 min; the highest reduction in size was obtained (88.9 %) for a BMtreatment time of 120 min. OPFF exhibited a significant size reduction over a milling time asshort as 3 min, with 71.1 % size reduction obtained; size reduction was further improved whenBM time was extended to 10–60 min with a reduction of 89.9–92.6 %.

To add weight on our finding, SEM images illustrating changes in surface morphology ofunmilled raw materials (OPEFB and OPFF) and BM-treated oil palm biomass over varyingBM processing times, as shown in Fig. 1. Rigid surfaces and intact particles were clearly seenfrom unmilled biomass samples (OPEFB and OPFF), as shown in Fig. 1a, d, indicating astrong cellulose-hemicellulose-lignin network. After BM treatment for a certain time, themorphologies of BM-treated OPEFB and OPFF were obviously changed, suggesting destruc-tion of the lignin seal between the cellulose-hemicellulose matrixes. The sizes of the particlesdecreased sharply; this increased the specific surface area by destroying lignocellulose micro-fibrils and provided greater access to enzymatic reaction. The effect of BM treatment on otherorganic residues has been previously discussed and yielded a reduction in particle size andincrease in specific surface area that has increased the rate of enzymatic hydrolysis. Aftermilling for 60 min, rice straw consisting of particle sizes less than 30 μm yielded glucose

Table 2 Effects of BM-treated oilpalm biomass on particle size andcrystallinity index

aCrystallinity index of BM-treated samples determined bywide-angle X-ray diffraction andcalculated as described in materi-al and methods section

Oil palm biomass/ballmilling time

Geometric meandiameter (μm)

Size reduction(%)

CrIa (%)

OFEFB

Unmilled (<2 mm) 523.6 0 56.1

BM- 30 min 312.0 40.4 53.1

BM- 40 min 203.7 61.1 51.7

BM- 50 min 61.9 88.2 46.6

BM- 60 min 77.4 85.2 44.0

BM- 120 min 58.1 88.9 9.3

OPFF

Unmilled (<2 mm) 600.6 0 57.6

BM- 3 min 173.5 71.1 50.9

BM- 5 min 119.4 80.1 47.3

BM- 10 min 61.3 89.8 39.0

BM- 20 min 44.7 92.6 24.1

BM- 30 min 46.1 92.3 9.0

BM- 60 min 62.7 89.6 4.5

Appl Biochem Biotechnol

amounts 3.8 times higher than unmilled materials (500 μm–2 mm) [20]. Khullar and co-workers [21] showed that without any pretreatment, Miscanthus ground to a particle size of56 μm yielded a glucose recovery double that compared to larger particles (695 μm). Thus,this study indicates that a reduction in particle size provides larger surface areas of BM-treatedOPEFB and OPFF samples. There have been diverging reports regarding correlation betweendirect and indirect conversion of lignocellulose into glucose from solely particle size reduction[22]. Besides reduction of particle size, greater enzymatic hydrolysis efficiency could beachieved by increasing pore size and volume by removing a larger percentage of hemicelluloseand lignin [22, 23].

Effect of BM on Crystallinity

A reduction in cellulose crystallinity is necessary in order to maximize the conversion ofglucose since the presence of high crystallinity is one of the limiting factors in the enzymatichydrolysis of cellulose [24]. The effects of BM on the CrI of oil palm biomass are shown inTable 2, and WAXD patterns are shown in Fig. 2. The WAXD diffractograms had lowintensity peaks at 2θ=18.7° and 22.3° that contributed to the crystallinity of cellulose.Reductions in CrI values were clearly seen for OPEFB and OPFF samples, where CrI valueswere lowest (9.3 and 4.5 %) when the materials were ball milled for 120 and 60 min,respectively. In general, the WAXD spectra and CrI values confirmed the disruption of thecrystalline structure because of the increased milling time. The decrease in peak intensity of(I002) at 2θ=22.3° accompanied by a broadening of the diffraction peak indicated a decrease incrystalline phase and crystallite size. The decreasing trend in CrI values revealed the effec-tiveness of BM as a pretreatment for OPEFB and OPFF. The cause of decreasing CrI valuescan be explained by the fragmentation of crystalline grains, the deformation of crystalline

a b c

d e f

Fig. 1 Scanning electron micrographs of ball mill-treated oil palm biomass. a OPEFB cutter milled to less than2 mm (×1,000), b OPEFB ball milled for 60 min (×1,000), c OPEFB ball milled for 120 min (×1,000), d OPFFcutter milled to less than 2 mm (×1,000), e OPFF ball milled for 20 min (×1,000), f OPFF ball milled for 60 min(×1,000)

Appl Biochem Biotechnol

5 10 15 20 25 30 35 40 45 50

Diffraction angle 2θ (degree)

BM 30 min

BM 60 min

BM 120 min

5 10 15 20 25 30 35 40 45 50

Diffraction angle 2θ (degree)

(a)

(b)

BM 10 min

BM 30 min

BM 60 min

UM

UM

Fig. 2 Wide-angle X-ray diffraction profiles of a ball milled OPEFB and b ball milled OPFF, at various millingtime. UM unmilled biomass (<2 mm) BM ball milling

Appl Biochem Biotechnol

structure, and the increase of amorphization during BM pretreatment [14, 24]. Kim andcoworkers [9] have reported that planetary milling was more effective at reducing CrI thanattrition milling, down to 19.5 % from unmilled rice straw (48 %) after a 2-h milling process.Our findings are in agreement with other studies [16, 20], which reported that BM samplesexhibit greatly reduced cellulose CrI values, thus granting greater access to enzyme digestion.

Effect of BM on Glucose, Xylose, and Total Conversion Yield

Table 3 shows the by-products generation, yield of glucose and xylose from BM-treated oilpalm biomass at varying milling times. It was observed that low production of acetic acid wasrecorded from BM-treated OPEFB samples at milling time 30–60 min with concentration 1.3–2.0 mg/g substrate, respectively. The generation of acetic acid was probably due todeactylation process that took place from hemicellulose degradation [4, 5], and the valuesobtained from this study were far below from inhibitory level. There was no acetic acidgeneration recorded from OPFF samples. Other inhibitory compounds such as furfural and 5-HMF were not recorded from this experiment indicating BM as suitable method for pretreat-ment of OPEFB and OPFF. To support our hypothesis, BM-treated oil palm biomass sampleswere subjected for enzymatic hydrolysis. From the results obtained, OPFF required theshortest milling time (60 min), followed by OPEFB (120 min), with glucose yields 80.3 and67.5 %, respectively. Meanwhile, the highest yield of xylose was obtained from OPEFB(80.1 %), followed by OPFF (78.6 %). Total conversion yield of sugars from OPEFB and

Table 3 Effect of milling time on by-products generation, glucose, xylose, and total conversion yield of oil palmbiomass after BM treatment

Milling time (min) By-products produced (mg/g biomass)a Yields of hydrolyzed sugarsb Total yield (%)c

Acetic acid Furfural 5-HMF Glucose (%) Xylose (%)

OPEFB

0 - - - 15.9±0.7 5.4±0.3 11.4±0.2

30 1.8±0.8 - - 16.5±3.2 28.6±1.7 20.7±2.2

40 1.7±1.1 - - 31.4±2.3 44.7±1.2 36.1±2.2

50 1.3±0.5 - - 36.1±5.3 56.4±3.4 43.2±4.5

60 2.0±0.8 - - 33.1±1.9 46.8±2.4 37.9±3.1

120 - - - 67.5±4.4 80.1±4.1 71.9±4.3

OPFF

0 - - - 23.2±1.5 17.7±0.3 21.7±1.2

3 - - - 35.7±5.7 23.9±0.8 32.3±4.2

5 - - - 44.2±5.4 28.8±0.7 39.8±4.1

10 - - 53.2±6.8 39.5±0.5 49.3±4.9

20 - - - 64.9±9.4 55.8±3.2 62.3±7.6

30 - 68.0±10.8 65.4±5.1 67.2±9.2

60 - - - 80.3±11.6 78.6±5.4 79.8±9.8

Ball milled treated oil palm biomass (6 %) was hydrolyzed with Acremonium cellulase at 10 FPU/g substrate,incubated for 72 h at 50 °C. Values presented herewith are mean of triplicate samplesa Untreated and ball milled treated samples (3 g) were suspended in 45 ml of distilled water prior buffer andenzymes addition. The by-products produced were determined by HPLCbGlucose and xylose were determined by HPLC

“-” not detected

Appl Biochem Biotechnol

OPFF were 71.9, and 79.8 %, respectively. Both OPEFB and OPFF exhibited suitablelignocellulosic biomass to be treated with the BM process. It is worth mentioning that BM-treated OPEFB and OPFF yielded 4-fold of glucose and 4–16-fold of xylose, respectively,compared to untreated oil palm biomass. Furthermore, there was no other functional group thatgenerated via BM treatment, and the co-substrates produced were suitable for ethanol produc-tion since the availability of microorganism to convert pentose to ethanol [25].

Our findings are comparable in terms of glucose, xylose, and total conversion yields withthose reported by Hideno et al. [20], who reported that BM treatment of rice straw for 60 minsuccessfully provided glucose, xylose, and total sugar conversions in yields of around 89.4,54.3, and 78.0 %, respectively. Sugarcane bagasse and straw showed similar monosaccharidesugar yields, with total conversion yields of 82 and 74 %, respectively [16]. Recently, OPEFBwas pretreated in an environmentally friendly way using steam treatment. An improvementwas observed with an increase in temperature, residence time, and reaction pressure in thesteam chamber, with total conversion and glucose yields of about 30–37 and 66.3 %,respectively [10, 11, 26]. OPFF, on the other hand, showed high glucose recovery (83.7 %)when treated with hot compressed water at a liquid-solid ratio of 8.0 with a severity factor of3.31 [12]. Since no reports exist on the oil palm biomass treated by the BM method, ourfindings show that BM is a promising and efficient glucose/xylose recovery method forOPEFB and OPFF and is comparable with other methods mentioned above.

Effect of Cellulase Loadings

Figure 3 shows that the treated OPEFB (BM treatment for 120 min) and OPFF (BM treatmentfor 60 min) materials were hydrolyzed at different Acremonium cellulase loadings to observethe effect on monomeric sugars and total conversion yields. The enzymatic digestibility ofpretreated OPEFB and OPFF exhibited an increasing trend with higher cellulase loadings. Atcellulase loadings of 2–30 FPU/g substrate, glucose, xylose, and total conversion yields forOPEFB were determined to be in the ranges 54.0–62.4, 64.2–78.7, and 57.5–68.1 %, respec-tively. Meanwhile, at cellulase loadings of 2–40 FPU/g substrate, glucose, xylose, and totalconversion yields for OPFF were in the ranges 64.1–80.3, 67.1–84.2, and 65.0–81.5 %,respectively. In this study, further increase in cellulase loading from 20 to 30 FPU/g substrategave small increment in xylose, glucose, and total sugars yield from BM-treated OPEFBsamples. Meanwhile, no significant improvement in enzymatic digestibility obtained fromBM-treated OPFF samples when supplied with 20–40 FPU/g substrates. This might be due tothe efficient reduction in particle size and crystallinity of cellulose by BM treatment and was inagreement with the findings reported earlier [20, 27]. A higher cellulase loading (85.32 FPU/gsubstrate) was reported to achieve 37.76 % hydrolysis for OPEFB treated by high-pressuresteam at 210 °C [11]. Jung and coworkers [28] reported that enzymatic digestibility of OPFFtreated with a 7 % ammonia solution was 41.4 % at a 60 FPU/g glucan loading; this lowenzymatic digestibility obtained was due to the low degree of delignification of the treatedOPFF samples. Another report showed that a greater reduction in hemicellulose (90 %) andlignin content (70 %) from acid-alkali pretreated OPEFB improved enzymatic digestibility upto 83.9 % at cellulase loadings of 50 FPU/g substrate [29]. The digestibility of cellulose wasimproved because removal of hemicellulose and lignin alters and increases the accessiblesurface area for enzymatic absorption, thus increasing enzymatic hydrolysis [22, 23, 29]. Byremoving hemicellulose and lignin in the polymer matrix, lower cellulase loadings could beapplied, thus reducing the total ethanol production cost. Our present studies showed that higherxylose and glucose recovery were obtained from BM-treated OPEFB and OPFF samples withlower cellulase loadings (10 FPU/g substrate).

Appl Biochem Biotechnol

54.0 57.6 58.5 58.6 62.4

64.2 68.2 71.9 77.3

78.7

0

10

20

30

40

50

60

70

0

20

40

60

80

100

120

140

160

2FPU 5FPU 10FPU 20FPU 30FPU

To

tal

conver

sion y

ield

(glu

cose

+xylo

se)

Yie

ld o

f su

gar

s, %

(g/g

-ra

w m

ater

ials

)

Cellulase (FPU/ g-substrate)

Xylose Glucose Total conversion(a)

64.1 70.3 73.1 78.1 80.3

67.174.5 78.0

83.5 84.2

0

10

20

30

40

50

60

70

80

90

0

20

40

60

80

100

120

140

160

180

2 FPU 5 FPU 10 FPU 20 FPU 40 FPU

To

tal

conver

sion y

ield

(glu

cos+

xylo

se)

Yie

ld o

f su

gar

s, %

(g

/ g

-ra

w m

ater

ials

)

Cellulase (FPU/ g- substrate)

Xylose Glucose Total conversion(b)

Fig. 3 Effect of cellulase loadings on glucose, xylose and total sugars yield. a Ball milled treated OPEFB for120 min and b ball mill-treated OPFF for 60 min. Values presented herewith are mean of triplicate samples. Theball mill-treated oil palm biomass (5 %) was hydrolyzed with Acremonium cellulose at varying FPU/g substrate,incubated for 72 h at 50 °C

Appl Biochem Biotechnol

Milling Energy and Integrated System

Pulverization is a process requiring high energy consumption, which results in significantfeedstock cost implications [22]. In this study, the BM pretreatment of OPFF and OPEFB wascalculated to require 234 and 468 MJ, respectively, per kilogram of biomass. This value wasmuch higher compared to rice straw as previously reported [9, 20] probably due to low loadingof biomass, different size, and specification of ball mill applied. Nevertheless, energy require-ments for the BM process could be reduced by expanding to pilot/industrial scales, wherehigher loadings of biomass and continuous processes could be implemented [9]. The problemof higher energy demands for the milling process could be solved by integration of thecomminution process in the oil palm industry, as the mills produce enough energy for theirown use. Surplus energy is generated from methane captured from anaerobic digestion of palmoil mill effluent (POME) under the Clean Development Mechanism (CDM). A study on theeconomic viability of integrated biogas energy and compost production for sustainable palmoil mill management was reported [2], suggesting that integrated technology is a moreattractive solution compared to a situation where the palm oil mill implements either biogasenergy or compost technology individually. They also found that even without implementingCDM, the integrated technology is still economically viable, which can be a good solution forsustainable palm oil industry management in the near future.

Conclusions

OPFF and OPEFB were successfully pretreated using planetary BM with significant reductionin particle sizes and cellulose crystallinity. At 30 and 40 FPU/g substrate cellulase loadings,OPEFB and OPFF showed high recovery of glucose (62.4 and 78.69 %) and xylose (80.35 and(84.23 %) production, respectively, suggesting that BM effectively increased surface area forenhancing enzymatic hydrolysis. Further investigation is necessary with regard to the efficientrecovery of sugars while enhancing energy efficiency, cost efficacy, and addressing environ-mental implications.

Acknowledgments This work was partly supported by the Science and Technology Research Partnership forSustainable Development (SATREPS), organized by Japan Science and Technology Agency (JST) and JapanInternational Cooperation Agency (JICA). We are grateful to Seri Ulu Langat Palm Oil Mill and Serting HilirPalm Oil Mill for providing the raw materials. Special thanks to Mr. Yoshihito Suwa for the technical assistance.

References

1. Malaysian Palm Oil Board. Available from: http://bepi.mpob.gov.my/images/area/2012/Area_summary.pdf2. Yoshizaki, T., Shirai, Y., Hassan, M. A., Baharuddin, A. S., Raja Abdullah, N. M., Sulaiman, A., & Busu, Z.

(2012). Environment, Development and Sustainability, 14, 1065–1079.3. Malaysian National Biomass Strategy. Available from: http://www.feldaglobal.com/site-content/National%

20Biomass%20Strategy%20Nov%202011%20FINAL.pdf.4. Mosier, N. S., & Wyman, C. (2005). Bioresource Technology, 96, 673–686.5. Kumar, P., Barrett, D. M., Delwiche, M. J., & Stroeve, P. (2009). Industrial & Engineering Chemistry

Research, 8, 3713–3729.6. Kim, S., Park, J. M., Seo, J. W., & Kim, C. H. (2012). Bioresource Technology, 109, 229–233.

Appl Biochem Biotechnol

7. Iberahim, N. I., Jamaliah, M. J., Harun, S., Mohd Nor, M. T., & Hassan, O. (2013). International Journal ofChemical Engineering and Applications, 4, 101–105.

8. Taherzadeh, M. J., & Karimi, K. (2008). International Journal of Molecular Science, 9, 1621–1651.9. Kim, H. J., Lee, S., Kim, J., Mitchell, R. J., & Lee, J. H. (2013). Bioresource Technology, 144, 50–56.10. Bahrin, E. K., Baharuddin, A. S., Ibrahim, M. F., Abdul Razak, M. N., Sulaiman, A., Abd-Aziz, S., Hassan,

M. A., Shirai, Y., & Nishida, H. (2012). BioResources, 7, 1784–1801.11. Baharuddin, A. S., Md Yunos, N. S. H., Nik Mahmud, N. A., Zakaria, R., & Md Yunos, K. F. (2012).

BioResources, 7, 3525–3538.12. Goh, C. S., Tan, H. T., & Lee, K. T. (2012). Bioresource Technology, 110, 662–669.13. Lin, Z., Huang, H., Zhang, H., Zhang, L., Yan, L., & Chen, J. (2010). Applied Biochemistry and

Biotechnology, 162, 1872–1880.14. Liao, Z., Huang, Z., Hu, H., Zhang, Y., & Tan, Y. (2011). Bioresource Technology, 102, 7953–7958.15. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templenton, D., & Crocker, D. (2008). NREL/TP-

510-42618, National Renewable Laboratory, Golden, CO.16. Silva, A. S., Inoue, H., Endo, T., Yano, S., & Bon, E. P. S. (2010). Bioresource Technology, 101, 7402–7409.17. Segal, L., Creely, J. J., Martin, A. E., & Conrad, C. M. (1959). Textile Research Journal, 29, 786–794.18. Inoue, H., Yano, S., Endo, T., Sasaki, T., & Sawayama, S. (2008). Biotechnology for Biofuels, 1, 2.19. Perez, J., Dorado, J. M., Rubia, T. D., & Martinez, J. (2002). International Microbiology, 5, 53–63.20. Hideno, A., Inoue, H., Tsukahara, K., Fujimoto, S., Minowa, T., Inoue, S., Endo, T., & Sawayama, S. (2009).

Bioresource Technology, 100, 2706–2711.21. Khullar, E., Dien, B. S., Rausch, K. D., Tumbleson, M. E., & Singh, V. (2013). Industrial Crops and

Products, 44, 11–17.22. Vidal, B. C., Jr., Dien, B. S., Ting, K. C., & Singh, V. (2011). Applied Biochemistry and Biotechnology, 164,

1405–1421.23. Palonen, H., Thomasen, A. B., Tenkanen, M., Schmidt, A. S., & Viikari, L. (2004). Applied Biochemistry

and Biotechnology, 117, 1–17.24. Hendriks, A. T. W. M., & Zeeman, G. (2009). Bioresource Technology, 100, 10–18.25. Kuyper, M., Toirkens, M. J., Diderich, J. A., Winkler, A. A., Van Dijken, J. P., & Pronk, J. T. (2005). FEMS

Yeast Research, 5, 925–934.26. Shamsudin, S., Umi Kalsom, M. S., Zainudin, H., Abd-Aziz, S., Mustapa Kamal, S. M., Shirai, Y., &

Hassan, M. A. (2012). Biomass and Bioenergy, 36, 280–288.27. Chang, V. S., & Holtzapple, M. T. (2000). Applied Biochemistry and Biotechnology, 84–86, 5–37.28. Jung, Y. H., Kim, S., Yang, T. H., Lee, H. J., Seung, D., Park, Y. C., Seo, J. H., Choi, I. G., & Kim, K. H.

(2012). Bioprocess and Biosystem Engineering, 35, 1497–1503.29. Kim, S., & Kim, C. H. (2013). Renewable Energy, 54, 150–155.

Appl Biochem Biotechnol