Enhanced properties of oil palm fronds (OPF) lignin fractions produced via tangential ultrafiltration technique

Page 1

Ep

MDa

b

F

a

ARR1A

KOLUPA

1

er(ubgtt2bcwct

tto

h0

Industrial Crops and Products 66 (2015) 1–10

Contents lists available at ScienceDirect

Industrial Crops and Products

jo ur nal home p age: www.elsev ier .com/ locate / indcrop

nhanced properties of oil palm fronds (OPF) lignin fractionsroduced via tangential ultrafiltration technique

. Hazwan Hussin a,b, Afidah Abdul Rahim a, Mohamad Nasir Mohamad Ibrahim a,ominique Perrin b, Nicolas Brosse b,∗

Lignocellulosic Research Group, School of Chemical Sciences, Universiti Sains Malaysia, 11800 Minden, Penang, MalaysiaLaboratoire d’Etude et de Recherche sur le MAteriau Bois (LERMAB), Faculte des Sciences et Techniques, Universite de Lorraine, Bld des Aiguillettes,-54500 Vandoeuvre-les-Nancy, France

r t i c l e i n f o

rticle history:eceived 7 October 2014eceived in revised form4 November 2014ccepted 13 December 2014

a b s t r a c t

The present study sheds light on the structural characteristic and antioxidant properties of different oilpalm fronds lignin fractions (kraft, soda and organosolv) separated via 5 kDa ultrafiltration membranes.The different permeates obtained have improved the physicochemical properties of lignin in terms oftheir specific molecular weight fragments, phenolic OH content (especially the phenolic S units) andsolubility. It was found that the antioxidant activity of lignin permeates was better than the crude ones

eywords:il palm frondsigninltrafiltrationhenolic groups

(highest oxygen uptake inhibition, OUI up to 83%) and was closely related to its average molecular weightand phenolic OH content. The ultrafiltration process allows the production of lignins with differentchemical structures and antioxidant capacity, enhanced solubility and subsequently making it moreadequate for commercial applications.

© 2014 Elsevier B.V. All rights reserved.

ntioxidant

. Introduction

One of the most critical issues faced by the world today is tonsure the sustainability of consumption for energy and naturalesources. As the petroleum fuels is creating problematic issuessuch as global warming, increase in price and running out), these of renewable resources to shift the oil-based economy intoio-based economy leads to alternative new discoveries. With aoal of reducing net greenhouse gas emission, this marks an impor-ant turning point in effort to promote the use of renewable energyo fulfill the commitments of the Kyoto Protocol (Ragauskas et al.,006; Sarkar et al., 2012). Lignocellulosic biomass was said to beest-suited for energy and chemical applications due to its suffi-ient availability, is inexpensive and environmentally safe. Recentork in this area has mainly focused on the delignification of ligno-

ellulosic biomass separating lignin, cellulose and hemicelluloseso be used in both physical and chemical applications.

Lignin is an aromatic amorphous biopolymer which is expected

o play an important role in the near future as raw materials forhe production of bioproducts. It is built up by oxidative couplingf three major C6–C3 (phenylpropanoid) units; p-coumaryl alco-

∗ Corresponding author. Tel.: +333 83 68 48 62; fax: +333 83 68 44 98.E-mail address: [email protected] (N. Brosse).

ttp://dx.doi.org/10.1016/j.indcrop.2014.12.027926-6690/© 2014 Elsevier B.V. All rights reserved.

hol, coniferyl alcohol and sinapyl alcohol (Ammalahti et al., 1998;Garcia et al., 2009; She et al., 2010). Lignins as well as other polyphe-nols are potent free radical scavengers and considered to be avaluable source of antioxidant phenolic compounds (due to highcontent of diverse functional groups such as phenolic and aliphatic

OH, carbonyls, carboxyls, etc.). The applicability of lignins fromdifferent sources as potential antioxidants has been also success-fully tested (Urgatondo et al., 2009; El Hage et al., 2012; Hussinet al., 2014a). Consequently, the antioxidant properties exhibitedby lignin can give broader applications as anti-microbial, anti-agingagents and corrosion inhibitors.

Nevertheless, the high non-homogeneity complex structure oflignin with high molecular weight distributions has limited itsusage in the industrial sectors. Although this phenolic moiety rep-resents a low and variable fraction of the total lignin, it can stronglyaffect the reactivity of the polymer (Garcia et al., 2009). In addition,high hydrophobicity of lignin can limit its capability to be employedin other possible applications. Therefore, the modulation of suitablelignin structures (by considering its solubility, molecular weight,phenolic OH content) is important so that it can overcome suchimplications. Obviously, the properties of lignin can be improvedby modifying the structure into a more suitable structure type.

Therefore, fractionation of lignin has become a promising methodto obtain a more specific molecular weight fraction with differentchemical compositions and functionalities.

Page 2

2 Crops

m1M2tmuTmaeurta

awgme(fObfoaw

2

2

OmTcla0OpaaSuSatar

2

rlpfiaap

M.H. Hussin et al. / Industrial

Various methods have been proposed to fractionate the ligninolecules. These comprises of solvent extraction (Thring et al.,

996; Yuan et al., 2009), differential precipitation (Sun et al., 1999;ussatto et al., 2007) and membrane technology (Wallberg et al.,

003; Toledano et al., 2010a). Membrane technology via ultrafiltra-ion was said to be a better method than differential precipitation

ethod to obtain different fractions of lignin with specific molec-lar weights and low contamination (Toledano et al., 2010b).he effectiveness of the ultrafiltration technology to separate theacromolecular solution is indeed beneficial for the purification

nd fractionation of lignin. In addition, a previous study by Garciat al. (2010) has revealed that the permeated lignin produced by theltrafiltration system will possess a good antioxidant activity thanough lignin. Thus, ultrafiltration technology would also improvehe properties of lignin polymer, making it suitable for any potentialpplications.

The oil palm (Elaeis guineensis) fronds, OPF have been identifieds the major contributor of biomass waste in Malaysia. Recently,e have demonstrated the potential of lignin obtained from OPF for

reen material applications (Hussin et al., 2013) and their possibleodification to improve the physicochemical properties (Hussin

t al., 2014a,b) and applicability especially as corrosion inhibitorHussin et al., 2014c). In the present work, the lignin was obtainedrom different pulping (kraft, soda and organosolv) processes ofPF. The resulting lignins were then re-dissolved and fractionatedy ultrafiltration membrane technique to obtain different ligninractions with smaller molecular weight. The improved physic-chemical characteristics, phenolic OH content, solubility andntioxidant activity were studied in order to establish differencesith the crude ones.

. Materials and methods

.1. Materials

The oil palm fronds (OPF) were obtained from the Valdor Palmil Mill near Sungai Bakap plantation (Seberang Prai, Malaysia) inid 2012. The composition (% w/w) of the OPF according to TAPPI

203 cm-09 and laboratory analytical procedure (LAP) method isellulose 35.73 ± 1.34%, hemicelluloses 28.39 ± 1.34% and Klasonignin 24.62 ± 1.17% on a dry weight basis. It also contains sug-rs such as glucans 56.30 ± 3.20%, xylans 16.80 ± 0.60%, arabinans.90 ± 0.10%, mannans 0.90 ± 0.00% and galactans 0.40 ± 0.00%. ThePF leaves were removed and the strands were chipped into smallieces. After sun dried for 3 days, the chips were then ground to

1–3 mm size using Wiley mill and the fiber was further dried inn oven at 50 ◦C for 24 h. The OPF biomass was first subjected tooxhlet extraction with ethanol/toluene (2:1, v/v) for 6 h beforese. All chemical reagents used in this study were purchased fromigma–Aldrich, Merck, QRec (Malaysia) and VWR (France) and useds received. Dried matter contents were determined using a mois-ure balance, KERN MRS 120-3 Infra-red moisture analyzer (dryingt 105 ◦C to constant weight). The effective dry matter content ofaw OPF biomass was ∼89%.

.2. Alkaline lignin extraction

Both kraft and soda pulping processes were carried out in a 4 Lotary digester. All pulping conditions followed the method out-ined by Wanrosli et al. (2007), with slight modifications. For kraftulping, a 20% of active alkali and 30% of sulfidity with water to

ber ratio of 8 were used. The time of maximum cooking temper-ture (170 ◦C) was set for 3 h. For soda pulping, 30% of active alkalilone was applied at the same condition as described above. Theressure of both kraft and soda pulping was around 12–15 bar. The

and Products 66 (2015) 1–10

pulp was washed and separated by screening through a sieve andthe black liquor was collected.

Kraft and soda lignins were precipitated from the concentratedblack liquor by acidifying it with 20% (v/v) sulfuric acid until pH2 (Mohamad Ibrahim et al., 2011). The precipitated lignins werefiltered and washed with pH 2 water. Both lignins were then driedin an oven at 50 ◦C for 48 h. The purification of lignin was conductedby extracting it in the Soxhlet apparatus for 6 h with n-pentaneto remove lipophilic non-lignin matters such as wax and lipids.The precipitate was filtered and washed twice with pH 2 water toremove the excess n-pentane and non-lignin phenolic compoundswhich may remain after the pulping process. The purified kraft andsoda lignins were then dried in an oven for another 48 h.

2.3. Organosolv lignin extraction

The oil palm fronds (OPF) were treated with 65% (v/v) of aque-ous ethanol with addition of 0.5% (w/w) sulfuric acid as a catalyst at190 ◦C for 60 min (El Hage et al., 2009). The material to liquid ratioused was 1:8. After the reactor was loaded with OPF and cookingliquor, it was heated to the operating temperature which was thenmaintained throughout the experiment (pressure around 25 bar).After the treatment, the pre-treated OPF was filtered and washedwith warm aqueous ethanol (8:1, 3 × 50 mL). The liquor was dilutedwith 3 volumes of water to precipitate the ethanol organosolvlignin and the precipitate was then collected by centrifugation at3500 rpm for 10 min. The resulting ethanol organosolv lignin wasdried in an oven at 50 ◦C for 48 h.

2.4. Fractionation of lignin by ultrafiltration unit

The tangential ultrafiltration module used to fractionate thelignins was a Pall Omega Centramate pilot unit equipped with a1 L reservoir feed with a recirculation pump and a set of Omegacassette equipped with polyethersulfone (PES) membrane cut-offin the interval of 5 kDa manufactured by PALL Corporation, USA.The retentate and permeate were recirculated to the feed tank for1 h before samples of permeate and retentate were withdrawn. Thepressure was measured at the inlet and at the outlet of the mem-brane tube. As the pressure on the permeate side was atmospheric,the trans membrane pressure, TMP is the average of the inlet andoutlet pressure on the feed side. The experiments were done atthe following experimental conditions: TMP: 300 kPa; cross-flowvelocity: 5.6 m s−1 and temperature: 25 ± 2 ◦C.

Approximately, around 2.00 g of lignins were weighed and dis-solved in 300 mL of mobile phase solvents (0.1 M NaOH for kraft andsoda; 65% v/v aqueous ethanol for organosolv lignin). The solutionswere sonicated for 5 min before gradually filtered. The ultrafiltra-tion system was rinsed 2–3 times with water and 0.1 M NaOH toremove any residues. Blank solution (0.1 M NaOH or 65% aq. EtOH)was introduced into the ultrafiltration system and the flux ratewas measured by controlling pump flow and retentate valve untila stable flux was achieved. Lignin solution was introduced into theultrafiltration system and the circulation was maintained for 1 hbefore withdrawal. The permeate solutions were collected until allsolution has passed the ultrafiltration membrane. The flow flux foreach sample was maintained at around 9 mL min−1 for kraft andsoda lignin and 0.30 mL min−1 for organosolv lignin. Next, the blanksolution was introduced again for flushing purposes for 30 min.All retentate solutions were collected by opening the retentatevalve. The permeate and retentate solution of kraft and soda ligninwere next acidified with 20% v/v H2SO4 and then centrifuged for

20 min (4000 rpm) to collect the precipitated lignin. The perme-ate and retentate solution of organosolv lignin were introduced tothe rotary evaporator to evaporate all ethanol. Then the remain-ing solution was freeze-dried to remove any water impurities. The

Page 3

Crops

pf

L

2

mTai(d

mQdE2pioc(sasMo5lpw

ectsfreefcfw4ldcdcw

tmwTPs1

2a

M.H. Hussin et al. / Industrial

ercentage of lignin fraction recovery (LPR %) was calculated asollows:

PR% = (Total dry weight of lignin after ultrafiltration)(Initial weight of lignin before ultrafication)

× 100 (1)

.5. Characterization of lignin

The FTIR spectrophotometry was carried out in a direct trans-ittance mode using PerkinElmer model System 2000 instrument.

he region between 4000 and 400 cm−1 with a resolution of 4 cm−1

nd 20 scans was recorded. The samples were prepared accord-ng to the potassium bromide technique, in a proportion of 1:100200 mg of KBr approximately). Interpretation of the IR spectra wasone using PerkinElmer software.

All nuclear magnetic resonance (NMR) spectroscopy experi-ents were performed on a Bruker Avance-400 spectrometer.uantitative NMR spectra were acquired using an inverse-gatedecoupling (Waltz-16) pulse sequence to avoid Nuclear Overhauserffect (NOE). The phosphitilating reagent employed here was-chloro-4,4,5,5-tetramethyl-1,1,3,2-dioxaphospholane (TMDP) asreviously described by Granata and Argyropoulos (1995). Approx-

mately, 25 mg of lignin was added into 400 �L solvent mixturef pyridine:CDCl3 (1.6/1, v/v) (Aldrich). Then each 150 �L ofhromium(III) acetylacetonate (3.6 mg mL−1) and cyclohexanol4.0 mg mL−1) solution in pyridine/CDCl3 was added to the ligninolution. The chromium(III) acetylacetonate served as a relax-tion reagent and cyclohexanol as the internal standard. The ligninolution was then vigorously stirred until completely dissolved.inutes before starting the NMR experiment, approximately 50 �L

f TMDP was added to the vial and the solution transferred into a mm NMR tube. The 31P NMR spectra were recorded with the fol-

owing acquisition parameters; inverse-gated pulse sequence, 25 sulse delay, 200 acquisitions, 61.7 ppm sweep width and 30◦ pulseidth (p1 = 6.5 �s, pl 1 = −2.0 db).

Lignin samples were subjected to acetylation in order tonhance their solubility in organic solvents, used in gel permeationhromatography (GPC). The acetylation reaction will substitute allhe hydroxylic functions into acetyl groups. Lignin (20 mg) was dis-olved in a 1:1 acetic anhydride/pyridine mixture (1 mL) and stirredor 24 h at room temperature. Ethanol (25 mL) was added to theeaction mixture, left for 30 min, and then removed with a rotaryvaporator. The addition and removal of ethanol were repeated sev-ral times to ensure complete removal of acetic acid and pyridinerom the sample. Afterwards, the acetylated lignin was dissolved inhloroform (2 mL) and added drop-wise to diethyl ether (100 mL)ollowed by centrifugation. The precipitate was washed three times

ith diethyl ether and dried under vacuum (762 mm of Hg) at0 ◦C for 24 h (Hussin et al., 2013). The number average molecu-

ar weights (Mn) and weight (Mw) of lignin were determined aftererivatization by GPC with a Dionex Ultimate-3000HPLC systemonsisting of an autosampler and a UV detector and using tetrahy-rofuran as eluent. Standard polystyrene samples were used toonstruct a calibration curve. Data were collected and analyzedith Chromeleon software Version 6.8 (Dionex Corp., USA).

The thermal behaviors of lignin samples were studied byhermal gravimetric analysis (TGA) using a PerkinElmer TGA 7 ther-

ogravimetric analyzer. Scans were recorded from 30 to 900 ◦Cith a heating rate of 10 ◦C min−1 under nitrogen atmosphere.

he glass transition temperatures (Tg) were obtained using aerkinElmer Pyris 1 differential scanning calorimeter (DSC). Theamples were heated from −50 to 200 ◦C with a heating rate of

0 ◦C min−1.

The dissolution test was carried out as follows. Approximately,00 mg sample was added to 500 mL distilled water at 25 ◦C andgitated constantly at 100 rpm for 6 h. 2 mL solution was with-

and Products 66 (2015) 1–10 3

drawn at predetermined intervals and filtered through a 0.45 �msyringe filter. An equal volume of distilled water was replacedafter each withdrawal (Francesco et al., 2009; Lu et al., 2012). Afterdiluted with distilled water, the solution absorbance was measuredat 280 nm (Shimadzu UV-2550 Japan). According to the UV lin-ear regression equation, lignin concentration and the percentageof dissolution (D %) are calculated as follows (Hussin et al., 2014b):

A − A0 = K × C × L (2)

D% = (Cmax × Vtotal)minitial

× 100 (3)

where A0 is the absorption of distilled water, A is the absorption ofsample, K is the absorption constant (21 L g−1 cm−1) (Lu et al., 2012),C is the concentration of lignin, Cmax is the maximum concentrationat 600 min, L is the thickness of quartz cell (cm), Vtotal is the totalvolume and minitial is the initial mass of lignin.

2.6. Measurement of lignin antioxidant activity

2.6.1. Oxygen uptake inhibitionAntioxidant properties of organosolv lignins were investi-

gated by evaluating oxygen uptake inhibition during oxidationof methyl linoleate. The induced oxidation by molecular oxy-gen was performed in a gas-tight borosilicate glass apparatus.Butan-1-ol was used as solvent for lignin dissolution. Tem-perature was set to 60 ◦C, initial conditions inside the vesselwere as follows; methyl linoleate (Fluka, 99%) concentration:0.32 mol L−1; 2,2′-azobisisobutyronitrile (AIBN) (Fluka, 98%) con-centration: 7.2 × 10−3 mol L−1; lignin concentration: 0.2 g L−1;oxygen pressure: 150 Torr. Oxygen uptake was monitored con-tinuously by a pressure transducer (Viatron model 104). Withoutany additive, oxygen uptake is roughly linear and constitutes thecontrol. In the presence of an antioxidant, oxygen consumption isslower, and the antioxidative capacity (OUI) of organosolv ligninwas estimated by comparing oxygen uptake at a chosen time (3 h),in the presence of this compound (pressure variation �Psample) andin the absence of the compound (�Pcontrol) according to:

OUI(%) = (�Pcontrol − �Psample)�Pcontrol

× 100 (4)

This ratio defines antioxidative capacity as an oxygen uptake inhi-bition index (OUI); it should spread from 0 to 100%, for poor andstrong antioxidants, respectively, and may be negative for proxi-dants.

2.6.2. Reducing power assayThe reducing power of samples was determined by the method

proposed by Gulcin et al. (2003) after a slight modification. Standardsyringaldehyde and vanillin (Aldrich) solution of concentrations2.0 × 10−2 mg mL−1 – 0.1 mg mL−1 were prepared. To 1.0 mL of thestandard solution, 2.5 mL of 0.2 M phosphate buffer at pH 6.6 (pre-pared from the addition of 0.2 M Na2HPO4 and 0.2 M NaH2PO4(QRec) and 2.5 mL of 1% (w/v) potassium ferricynide, K3Fe(CN)6(Aldrich) solution were added. The mixture was incubated at 50 ◦Cfor 20 min after which 2.5 mL 10% (w/v) trichloroacetic acid (Merck)was added. The resultant mixture was centrifuged for 20 min at2500 rpm. The upper layer (2.5 mL) was dispensed and 2.5 mL dis-tilled water and 0.5 mL of 0.1% (w/v) ferric chloride hexahydrate,FeCl36H2O (Merck) solution were added. The absorbance was mea-sured at 700 nm. The procedure was repeated for lignin samples.

3. Results and discussion

The resulted permeates and retentates of three different typesof lignins (kraft, soda and organosolv) after passing through 5 kDa

Page 4

4 Crops

uttwl1mtbsMka1t

tmcsmeabprci

3

s

M.H. Hussin et al. / Industrial

ltrafiltration membrane were collected and characterized in ordero determine their physicochemical properties. It was found thathe highest recovery of lignin permeate fraction (LFR %, w/w)as obtained by soda lignin followed by kraft and organosolv

ignin (LFRsoda: 12.29 ± 0.54% > LFRkraft: 5.41 ± 2.04% > LFRorganosolv:.48 ± 0.15%). Since the fractionation of lignin via ultrafiltrationembrane gave very low percentage of recovery, the implementa-

ion of ultrafiltration technique especially in industrial sector maye less effective. Unexpectedly, the LFR value of all lignin sampleshows unclear trend with their crude average molecular weight,

w as reported in our previous study (Hussin et al., 2013). Indeed,raft lignin shows low permeate yield due to the higher crude aver-ge molecular weight (Mw: 2063 g mol−1) than soda lignin (Mw:660 g mol−1). Higher Mw of kraft lignin makes it difficult to passhrough smaller membrane cut-off (5 kDa).

In contrast, organosolv lignin gives lower permeate yield evenhough it has a low crude Mw value (Mw: 1215 g mol−1). Lower per-

eate yield for organosolv lignin could be explained by its lowhemical affinity with the relatively high hydrophilic polyether-ulfone membrane and also by the use of aqueous ethanol asobile phase. In addition, during organosolv pulping the lignin was

xtracted from the raw material, which produced different sizesnd shapes of lignin fragments (Toledano, 2012). Besides, it shoulde understood that the ultrafiltration method produces not onlyroduct fractionation but also product concentration. Hence, alletentates presented high lignin concentration. Since all permeatesould potentially possesses low molecular weight (LMW) of lignin,t was further characterized.

.1. FTIR analysis

The FTIR spectra of the obtained ultrafiltration fractions arehown in Fig. 1. In this study, the following structure signals

Fig. 1. FTIR spectra of crude and ultrafiltrat

and Products 66 (2015) 1–10

were detected; aromatic phenylpropane vibrations (1600, 1515and 1427 cm−1), aromatic and aliphatic OH groups stretching(3400 and 1033 cm−1), C H in methyl and methylene bonds (2923,2850 and 1462 cm−1). The signals for C O stretching in uncon-jugated ketone and conjugated carboxylic groups (1706 cm−1),C O C bridges (1217 cm−1) and aromatic C H out-of-plane defor-mation (833 cm−1) were also observed. This result is in goodagreement with previous studies with different types and frac-tions of lignin after ultrafiltration (Alriols et al., 2010; Garciaet al., 2010; Toledano et al., 2010a,b). A small band at 620 cm−1

which was detected for both kraft and soda lignin fractionswas attributed to hemicelluloses and silicates. This suggestedthat lignin fractions obtained through acid precipitation containhemicelluloses contamination. The contamination was perhapsdue to high chemical affinity between hemicellulose moleculeswith the polyethersulfone membrane (both considered have goodhydrophilic properties) facilitated by their mobile phase (0.1 MNaOH). It was also believed that suspended lignin drags thehemicellulose and lignin-carbohydrate complex (LCC) when pre-cipitation occurred (Toledano et al., 2010b). Organosolv ligninfraction showed higher intensities in aliphatic/aromatic OHgroups and C H bands as compared to organosolv lignin beforethe ultrafiltration process. Meanwhile soda lignin fraction showedintensified peak at 1700 cm−1 which is attributed to C O con-jugated aromatic bonds (Hibbert ketones) presented in quinonicstructures.

Besides these typical lignin structure bands, signals that areassociated with syringyl (S) and guaiacyl (G) were also detected.The S-type aromatic C H in-plane and out-of-plane deformationswere observed at 1116 and 833 cm−1 while breathing S ring with

C O stretching gave a signal at 1327 cm−1. In addition, breath-ing G ring with C O stretching (1265 cm−1), G-type aromaticC H in-plane and out-of-plane bending (1033 cm−1) were found,

ed kraft, soda and organosolv lignins.

Page 5

M.H. Hussin et al. / Industrial Crops and Products 66 (2015) 1–10 5

Table 1Quantitative analysis of OPF lignins via 31P NMR.

ppm Assignments (31P NMR) mmol g−1

Krafta Sodaa Organosolva Permeated kraft Permeated soda Permeated organosolv

150–145 Aliphatic OH 0.45 0.42 1.49 0.77 0.69 0.89144.8 Cyclohexanol (internal standard)144–140 Syringyl OH 1.42 1.07 0.69 1.10 1.21 0.99140–138 Guaiacyl OH 0.65 0.48 0.45 0.58 0.47 0.44

6

7

rf(cfbhlfte

chains produced after organosolv treatment) as compared to kraftand soda lignin fractions. The sharp peak observed at 147 ppm for

138–136 p-Hydroxyphenyl OH 0.13 0.11 0.2135–133 Carboxylic acid – – 0.0

a (Hussin et al., 2014c).

espectively. It was noticed that the S-bands intensity of all ligninractions were higher than the G-bands. According to Alriols et al.2010), the intensity of signal was dependant on the lignin fractionut off. Lignin with higher percentages of G units will produce ligninraction with high molecular weight through the formation of C5onds. In the case of S units, it is impossible to form C5 bonds as theyave C5 position substituted by the methoxy groups. Therefore, the

ow membrane cut-off used in this study (5 kDa) will produce ligninraction with higher S-band intensity (low molecular weight) and

his observation is in good agreement with study done the by Alriolst al. (2010).

Fig. 2. 31P NMR spectra of OPF lignin fractions afte

0.07 0.16 0.88– – –

3.2. 31P NMR

The quantitative 31P NMR values after derivatization with TMDPand their corresponding spectra of the resulted lignin fraction sam-ples are shown in Table 1 and Fig. 2. From this data, it can beseen that the organosolv lignin fraction produce highest aliphatic

OH content (due to a lower dehydration/degradation of lateral

both kraft and soda lignin fractions could corresponds to secondary(erythro and threo) aliphatic OH, while the peak at 145.8 ppm

r ultrafiltration (5 kDa) in expanded region.

Page 6

6 Crops

f

twacclthst(trMp2cfc

3

tf(lmdnlw

omSttpsekt

3

3

tfidaers

TG

M.H. Hussin et al. / Industrial

or organosolv lignin fraction was assigned to primary aliphaticOH groups. This is in good agreement to our previous report

hat organosolv lignin gave higher primary aliphatic OH content,hereas alkaline lignins (kraft and soda) gave higher secondary

liphatic OH content (Hussin et al., 2013). On the other hand, itan be observed that all the resulted lignin fractions gave loweroncentration of aliphatic OH groups compared to the crudeignins before fractionation process. Besides, some increment ofhe total HGS phenolic OH (syringyl OH + guaiacyl OH + p-ydroxyphenyl OH) content than the crude was observed foroda (1.84 mmol g−1) and organosolv (1.52 mmol g−1) lignin frac-ions. However, the total HGS phenolic OH content for kraft1.75 mmol g−1) lignin fraction shows some reduction. Accordingo Garcia et al. (2010), the molecular weight of crude lignin mayeflect the total phenolic content in the lignin fractions. Higher

w value of crude lignin will give lignin fractions with lower totalhenolic OH content. Perhaps that is why crude kraft lignin (Mw:063 g mol−1) produces lignin fraction with low total phenolic OHontent. Hence, the amount of phenolic OH content of ligninractions through the utilization of low ultrafiltration membraneut-off will strongly rely on their crude molecular weight.

.3. GPC analysis

Gel permeation chromatography (GPC) was carried out to obtainhe distributions of molecular weight of different acetylated ligninractions obtained by ultrafiltration. The average molecular weightMw), number average (Mn) and polydispersity (PD = Mw/Mn) ofignin fractions are shown in Table 2. Clearly, it can be seen that the

olecular weight distributions of lignin fractions were significantlyecreased after the ultrafiltration process. The ultrafiltration tech-ique also produced lignin fractions with low polydispersity. The

ow PD values indicated high fraction of lignin with low moleculareight (LMW) (Toledano et al., 2010b).

The number of C C bonds between units is connected to the Mw

f lignin, mainly to the structures involving aromatic C5 ring. Theost abundant aromatic rings in lignin are G-type (guaiacyl) and

-type (syringyl) units. It was believed that G-type units are ableo form this kind of bonds and is not possible for S-type units dueo the presence of methoxy functional groups at both C3 and C5ositions. Thus, lignins that are mostly composed of G units tend tohow higher Mw (Glasser et al., 1993; Toledano et al., 2010a; Garciat al., 2010). This is in agreement with the 31P NMR study whereraft lignin fraction gave higher guaiacyl OH content comparedo soda and organosolv lignin fractions.

.4. Thermal behavior

.4.1. Thermal gravimetric analysisIn order to study the decomposition of organic polymers, the

hermogravimetric analysis (TGA) was carried out for the ultra-ltrated fractions. The TG curves (Fig. 3) show three major stepsuring sample degradation and as similarly reported by several

uthors (Alriols et al., 2010; Garcia et al., 2010; Mohamad Ibrahimt al., 2011). Similar to other lignin thermograms, the first one cor-esponded to water evaporation (observed around 90 ◦C) and theecond one was related to the degradation of hemicelluloses (at

able 2PC results of weight-average (Mw), number-average (Mn) and polydispersity (PD: Mw/M

Kraftb Sodab Organosolvb Pe

Mw(g mol-1) 2063 1660 1215 13Mn(g mol-1) 1063 713 1036 10Mw/Mn(PD) 1.89 2.33 1.17 1.3

b (Hussin et al., 2013).

and Products 66 (2015) 1–10

around 200–300 ◦C), which are present in the lignin samples. Thesedegradation peaks were more pronounced for the alkaline ligninfractions (kraft and soda permeates) which entail small hemicellu-lose impurity content (as observed in the FTIR spectra). The thirdstep was related to the degradation of lignin, which occurred pro-gressively from 200 up to 400 ◦C.

In some cases, wide degradation peaks suggested the pres-ence of hemicellulose–lignin complexes. This phenomenon wasobservable to that of organosolv lignin fraction. The lower DTG tem-perature obtained for organosolv lignin fraction (DTGmax: 340 ◦C)than kraft and soda lignin fractions (DTGmax: 380 ◦C) was due tothe less recondensed structure (less C C bonds with more �-O-4linkages) of organosolv lignin. Besides, this could also due to thedecomposition temperature of hemicellulose–lignin intermediate(Toledano et al., 2010b). It was believed that the maximum weightloss rate depended strongly on the lignin structure, particularlytheir functional groups and linkages formed.

Additionally, the residue obtained after the thermal degradationwas also related to the lignin molecule structural complexity andtheir grade of linkages. High percentage of residue will imply highthermal stability of the sample (Dominguez et al., 2008). However,high residue percentage could also indicate the presence of inor-ganic matter in the lignin samples (Garcia et al., 2010) due to poorwashing step. In this analysis, it was found that kraft lignin frac-tion shows high residue value (50–60%) due to the use of sodiumsalts (sodium hydroxide and sodium sulfide) during crude kraftdelignification or ultrafiltration process.

3.4.2. Differential scanning calorimetryThe thermal behavior of lignin fractions was further analyzed

with differential scanning calorimetry (DSC) to study the ther-mal effect associated with physical and/or chemical changes. Theobtained organosolv lignin fraction presented a lower glass tran-sition (Tg: 49.39 ◦C) value than alkaline lignin fractions (kraft Tg:51.58 ◦C; Soda Tg: 55.28 ◦C). It was reported previously by otherauthors (Feldman and Banu, 1997; Lora and Glasser, 2002; Alriolset al., 2010) that the Tg values of alkaline lignins (kraft and soda)were higher than for organosolv ones. Besides, it can be seenthat the ultrafiltration lignin fractions Tg increases with the ligninmolecular weight. This confirms that the fractionation process hasa great influence on the obtained lignin molecular weight (Randhiret al., 2004; Lora, 2008).

This observation is also in concordance with the results obtainedby other techniques (FTIR, 31P NMR and GPC) where the relationbetween lignin molecular weight and the S and G groups con-tent was established. Fractionation of lignin using low membranecut-off (5 kDa) will produce small molecular weight fractions withlower G content (units that contribute to the formation of C Cbonds) and low Tg value.

3.5. Dissolution test of lignin

Fig. 4 shows that the dissolution of all lignin fractions

(after ultrafiltration) was greater compared to the crude lignins(Dkraft: 0.69%; Dsoda: 1.57% and Dorganosolv: 2.07%). Kraft ligninfraction presented the highest percentage of dissolution and fol-lowed by organosolv and soda lignin fractions (Dpermeated kraft:

n) of OPF lignins.

rmeated kraft Permeated soda Permeated organosolv

94 955 84864 837 8231 1.14 1.03

Page 7

M.H. Hussin et al. / Industrial Crops and Products 66 (2015) 1–10 7

Fig. 3. (A) TG and (B) DTG curves for OPF lign

Fig. 4. Dissolution profiles of ultrafiltrated and crude lignins.

in fractions after ultrafiltration (5 kDa).

10.27% > Dpermeated organosolv: 10.11% > Dpermeated soda: 9.49%). Betterdissolution properties of all lignin fractions can be rationalized bytwo major factors; (1) low molecular weight/polydispersity (due tothe uniformity of the overall packing structures) (Alriols et al., 2009)and (2) higher phenolic content (more polar structures), which pro-motes better hydrogen bonding with water molecules. Therefore,it can be affirmed that fractionation of lignin could substantiallyimprove their hydrophilicity and applicability.

3.6. Lignin antioxidant activity

3.6.1. Oxygen uptake inhibitionThe results of the antioxidant activity of different lignin frac-

tions obtained by the oxygen uptake inhibition (OUI) method areshown in Fig. 5A . In this study, it was found that fractionation oflignin samples through the ultrafiltration system improved their

antioxidant activity compared to the crude lignins (Supplementarymaterials). Soda lignin fraction shows higher OUI value followedby kraft and organosolv lignin fractions (OUIsoda: 83% > OUIkraft:79% > OUIorganosolv: 77%). The performance of the lignin antioxi-

Page 8

8 Crops

dIh(gOf

a2tol(otfiBa

M.H. Hussin et al. / Industrial

ant activity could be dependent on their phenolic OH content.t can be noticed from 31P NMR analysis that soda lignin fractionas higher HGS phenolic OH (1.84 mmol g−1) content than kraft1.75 mmol g−1) and organosolv (1.51 mmol g−1) fractions. Thus, aood correlation between lignin phenolic OH content with theirUI value affirmed that membrane technology is a suitable process

or the production of compounds with high antioxidant capacity.Another factor that strongly affected the radical scavenging

ctivity of lignin was its heterogeneity and purity (Dizhbite et al.,004). The presence of hemicellulose components could diminishhe antioxidant capacity of lignin samples; due to the formationf hydrogen bondings between carbohydrates and lignin pheno-

ic OH groups that later interfere with the antioxidant propertiesGarcia et al., 2010). It was proposed from the TGA analysis thatrganosolv lignin fraction could contain higher carbohydrate con-amination (hemicellulose–lignin complexes) than other lignin

ractions. Thus, this could corroborate the low antioxidant capac-ty of organosolv lignin fraction as observed from the OUI value.esides, high S-type units in both soda and kraft lignin fractions canlso contribute to a better antioxidant activity. The ortho-methoxyl

Fig. 5. Antioxidant profile of ultrafiltrated lignins by; (A) oxy

and Products 66 (2015) 1–10

substitution exhibited in S might provide resonance stabilizationto the incipient phenoxyl radical that can increase the antioxi-dant activity (Barclay et al., 1997; Nadji et al., 2009). Therefore, theantioxidant activity of lignin fraction should increase with the freephenolic OH and ortho-methoxyl content, through the stability ofthe radical formed.

3.6.2. Reducing power assayFig. 5B shows the reducing power of ultrafiltrated alkaline

lignins (kraft and soda) and organosolv lignin as a function of theirconcentration. In this assay, the yellow colour of the test solu-tion changes to various shades of green and blue, depending onthe reducing power of each lignin. The presence of reducers (i.e.,antioxidants) causes the reduction of the Fe3+ (ferricyanide) com-plex to the ferrous complex that has an absorption maximum at700 nm (Ferreira et al., 2007). As a comparison, reducing power of

syringaldehyde and vanillin standard was employed to mimic thereduction ability of syringyl (S) and guaiacyl (G) units. In this study,the reducing power of lignin increased with concentration to give aplateau at a concentration of 0.08 mg mL−1. Increased absorbance of

gen uptake inhibition and (B) reducing power method.

Page 9

Crops

trtilou

pttitrssaltf

3

tpaamr

oio

mharGtcffsbwrbeytwacg

4

lIpl(

M.H. Hussin et al. / Industrial

he reaction mixture indicates increase in reducing power. Highereducing power of syringaldehyde than vanillin standard revealedhat the S unit could facilitate better reduction ability. The reduc-ng power of all ultrafiltrated lignins was higher than the crudeignin ones (Supplementary materials). In addition, similar trendsf antioxidant capacity through lignin reducing power and oxygenptake inhibition method were observed.

Soda lignin fraction gives better reducing power ability com-ared to kraft and organosolv lignin fractions. It could be assumedhat soda lignin fraction might contain higher amounts of reduc-ants, which are able to reduce Fe3+ ion into Fe2+ ions. Themprovement of lignin reducing power ability was again relatedo their phenolic OH and S unit content. As seen in Fig. 5B, lignineducing power ability increased and behaved almost similar totandard S (syringaldehyde) unit. Thus, this confirms that ligninamples with high S unit content would probably improve thentioxidant ability. It could also be affirmed that fractionation ofignin by ultrafiltration method produced high phenolic OH con-ent and better antioxidant activity. All these properties are usefulor various applications especially as a corrosion inhibitor.

.7. Possible applications of ultrafiltrated lignin

Structural fractionation of lignin via ultrafiltration membraneechnology has tremendously altered both physical and chemicalroperties of the lignin. Higher S-type unit (C3 and C5 positionsre linked to methoxy groups) of ultrafiltrated lignins offers a goodntioxidant property due to its ortho-methoxyl substitution whichight provide resonance stabilization to the incipient phenoxyl

adical that can increase the antioxidant activity. Higher phenolicOH and ortho-methoxyl groups content with smaller fragmentsf lignin structures (low average molecular weight) was produced

n which later gave a good antioxidant activity as observed fromxygen uptake inhibition and reducing power assay.

Generally, the reactivity of lignins with high fractions of lowolecular weight, LMW molecules are higher than the one with

igh molecular weight, HMW (Pizzi, 1994). High fractions of LMWre suitable to be used as a component of phenol-formaldehydeesins or an extender due to their high reactivity (Benar et al., 1999;oncalves and Benar, 2001; El Mansouri and Salvado, 2006). Addi-

ionally, based from their Mw and PD values, the ultrafiltrated ligninould also be used to improve the properties of viscous media usedor inks and paints and for favoring blend capability in polymerormulations. An improvement of lignin properties toward theirolubility right after being fractionated with ultrafilration mem-rane was observed. The hydrophobicity of lignin was then reducedith this technique; and therefore, this could widen the area of

esearch for lignin to be applied in any aqueous systems (e.g., adsor-ent of waste water, corrosion inhibitors, drilling fluid, additives,tc.). Even though this technique produces lignin with low recoveryield (less effective to be applied in industrial sector) but throughhis study it was evidence that properties of the lignin fractionere enhanced, in particular to their solubility and antioxidant

ctivity (which increase its applicability). Our current research isoncerning the utilization of modified lignin to be employed asreen corrosion inhibitors.

. Conclusion

In this work, it was successfully proven that the modulation ofignin structure can be achieved by using a fractionation technique.

n fact, fractionation of lignin via ultrafiltration technique is one of aroper way to obtain smaller fractions of lignin (with low molecu-

ar weight). Improved physicochemical properties of ultrafiltratedkraft, soda and organosolv) lignins were observed where smaller

and Products 66 (2015) 1–10 9

fragments of lignin with higher phenolic OH content (especiallythe phenolic S units) with better solubility were produced from thismethod. In addition, the second major finding was that the oxy-gen uptake inhibition of ultrafiltrated lignins seems be dependenton the increased free phenolic OH and ortho-methoxyl content,through the stabilization of the radical formed. Besides, the reduc-ing power assay has also revealed that lignin samples with high Sunit content could probably improve the antioxidant ability. Theenhanced properties of the new lignin structure are indeed benefi-cial for its subsequent applications.

Acknowledgments

The authors are grateful for the financial support of this researchfrom CPER 2007-2013 “Structuration du Pôle de CompétitivitéFibres Grand’Est” and from Universiti Sains Malaysia throughUSM Research University Grant–1001/PKIMIA/854002. M. HazwanHussin would like to express his gratitude to the Ministry of HighEducation (MOHE) of Malaysia and Ministere Affaires Etrangeresde France for the MyPhD scholarship and Boursier du Governe-ment Francais (CampusFrance). The EA 4370 LERMAB is supportedby the French National Research Agency through the Laboratory ofExcellence ARBRE (ANR-12- LABXARBRE-01).

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2014.12.027.

References

Alriols, M.G., Tejado, A., Blanco, M., Mondragon, I., Labidi, J., 2009. Agriculturalpalm oil tree residues as raw material for cellulose: lignin and hemicellulosesproduction by ethylene glycol pulping process. Chem. Eng. J. 148, 106–114.

Alriols, M.G., Garcia, A., Llano-ponte, R., Labidi, J., 2010. Combined organosolv andultrafiltration lignocellulosic biorefinery process. Chem. Eng. J. 157, 113–120.

Ammalahti, E., Brunow, G., Bardet, M., Robert, D., Kipelainen, I.J., 1998.Identification of side-chain structures in a poplar lignin usingthree-dimensional HMQC-HOHAHA NMR spectroscopy. J. Agric. Food Chem.46, 5113–5117.

Barclay, L.R.C., Xi, F., Norris, J.Q., 1997. Antioxidant properties of phenolic ligninmodel compounds. J. Wood Chem. Technol. 17 (1–2), 73–90.

Benar, P., Goncalves, A.R., Mandelli, D., Schuchardt, U., 1999. Eucalyptus organosolvlignins: study of the hydroxymethylation and use in resols. Bioresour. Technol.68, 11–16.

Dizhbite, T., Telysheva, G., Jurkjane, V., Viesturs, U., 2004. Characterization ofradical scavenging activity of lignin-natural antioxidants. Bioresour. Technol.95 (3), 309–317.

Dominguez, J.C., Oliet, M., Alonso, M.V., Gilarranz, M.A., Rodriguez, F., 2008.Thermal stability and pyrolysis kinetics of organosolv lignins obtained fromEucalyptus globulus. Ind. Crop Prod. 27, 150–156.

El Hage, R., Brosse, N., Chrusciel, L., Sanchez, C., Sannigrahi, P., Ragauskas, A., 2009.Characterization of milled wood lignin and ethanol organosolv lignin frommiscanthus. Polym. Degrad. Stabil. 94, 1632–1638.

El Hage, R., Perrin, D., Brosse, N., 2012. Effect of pre-treatement severity on theantioxidant properties of ethanol organosolv Miscanthus x giganteus lignin.Nat. Resource 3, 29–34.

El Mansouri, N.E., Salvado, J., 2006. Structural characterization of technical ligninsfor the production of adhesives: application to lignosulfonate, kraft,soda-anthraquinone, organosolv and ethanol process lignins. Ind. Crop Prod.24, 8–16.

Feldman, D., Banu, D., 1997. Contribution to the study of rigid PVC polyblends withdifferent lignins. J. Appl. Polym. Sci. 66 (9), 1731–1744.

Ferreira, I.C.F.R., Baptista, P., Vilas-Boas, M., Barros, L., 2007. Free-radical scavengingcapacity and reducing power of wild edible mushrooms from northeastPortugal Individual cap and stipe activity. Food Chem. 100, 1511–1516.

Francesco, L., Chiara, S., Guido, E., Francesca, M., Giaime, M., Anna Maria, F., 2009.Diclorfenec nanosuspension: influence of preparation procedure and crystalform on drug dissolution behavior. Int. J. Pharm. 373, 124–132.

Garcia, A., Toledano, A., Serrano, L., Egues, I., Gonzales, M., Marinn, F., Labidi, J.,2009. Characterization of lignins obtained by selective precipitation. Sep. Purif.Technol. 68, 193–198.

Garcia, A., Toledano, A., Andres, M.A., Labidi, J., 2010. Study of the antioxidantcapacity of Miscanthus sinesis lignins. Process Biochem. 45, 935–940.

Page 10

1 Crops

G

G

G

G

H

H

H

H

L

L

L

M

M

0 M.H. Hussin et al. / Industrial

lasser, W.G., Dave, V., Frazier, C.E., 1993. Molecular weight distribution ofsemi-commercial lignin derivatives. J. Wood Chem. Technol. 13 (4), 545–559.

oncalves, A.R., Benar, P., 2001. Hydroxymethylation and oxidation of organosolvlignins and utilization of the products. Bioresour. Technol. 79, 103–111.

ranata, A., Argyropoulos, D.S., 1995.2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane, a reagent for theaccurate determination of the uncondensed and condensed phenolic moietiesin lignins. J. Agric. Food Chem. 43, 1538–1544.

ulcin, I., Oktay, M., Kirecci, E., Kufrevioglu, O.I., 2003. Screening of antioxidant andantimicrobial activities of anise (Pimpinella anisum L.) seed extracts. FoodChem. 83, 371–382.

ussin, M.H., Rahim, A.A., Mohamad Ibrahim, M.N., Brosse, N., 2013.Physicochemical characterization of alkaline and ethanol organosolv ligninsfrom oil palm (Elaeis guineensis) fronds as phenol substitutes for green materialapplications. Ind. Crop Prod. 49, 23–32.

ussin, M.H., Rahim, A.A., Mohamad Ibrahim, M.N., Yemloul, M., Perrin, D., Brosse,N., 2014a. Investigation on the structure and antioxidant properties ofmodified lignin obtained by different combinative processes of oil palm fronds(OPF) biomass. Ind. Crop Prod. 52, 544–551.

ussin, M.H., Rahim, A.A., Mohamad Ibrahim, M.N., Perrin, D., Yemloul, M., Brosse,N., 2014b. Impact of catalytic oil palm fronds (OPF) pulping on organosolvlignin properties. Polym. Degrad. Stabil. 109, 33–39.

ussin, M.H., Shah, A.M., Rahim, A.A., Mohamad Ibrahim, M.N., Perrin, D., Brosse,N., 2014c. Antioxidant and anticorrosive properties of oil palm fronds ligninsextracted with different techniques. Ann. For. Sci.,http://dx.doi.org/10.1007/s13595-014-0405-1.

ora, J.H., Glasser, W.G.J., 2002. Recent industrial applications of lignin – asustainable alternative to non-renewable materials. J. Polym. Environ. 10,39–48.

ora, J.H., 2008. Industrial commercial lignins: sources, properties andapplications. In: Belgacem, M.N., Gandini, A. (Eds.), Monomers, Polymers andComposites from Renewable Resources. Elsevier, Oxford, pp. 225–241.

u, Q., Zhu, M., Zu, Y., Liu, W., Yang, L., Zhang, Y., Zhao, X., Zhang, X., Zhang, X., Li,W., 2012. Comparative antioxidant activity of nanoscale lignin prepared by asupercritical antisolvent (SAS) process with non-nanoscale lignin. Food Chem.135, 63–67.

ohamad Ibrahim, M.N., Zakaria, N., Sipaut, C.S., Sulaiman, O., Hashim, R., 2011.Chemical and thermal properties of lignin from oil palm biomass as a substitutefor phenol formaldehyde resin production. Carbohydr. Polym. 86, 112–119.

ussatto, S.I., Fernandes, M., Roberto, I.C., 2007. Lignin recovery from brewer’sspent grain black liquor. Carbohydr. Polym. 70, 218–223.

and Products 66 (2015) 1–10

Nadji, H., Diouf, P.N., Benaboura, A., Bedard, Y., Riedl, B., Stevanovic, T., 2009.Comparative study of lignins isolated from Alfa grass (Stipa tenacissma L.).Bioresour. Technol. 100, 3585–3592.

Pizzi, A., 1994. Advanced Wood Adhesives Technology. Marcel Dekker, New York,pp. 219–241.

Ragauskas, A., Williams, C., Davison, B., Britovsek, G., Cairney, J., Eckert, C.,Frederick, W., Hallett, J., Leak, D., Liotta, C., Mielenz, J., Murphy, R., Templer, R.,Tschaplinski, T., 2006. The path forward for biofuels and biomaterials. Science311, 484–489.

Randhir, R., Lin, Y.T., Shetty, K., 2004. Stimulation of phenolics: antioxidant andantimicrobial activities in dark germinated mung bean sprouts in response topeptide and phytochemical elicitors. Process Biochem. 39, 637–646.

Sarkar, N., Ghosh, S.K., Bannerjee, S., Aikat, K., 2012. Bioethanol production fromagricultural wastes: an overview. Renew. Energy 37, 19–27.

She, D., Xu, F., Geng, Z., Sun, R., Jones, G.L., Baird, M.S., 2010. Physicochemicalcharacterization of extracted lignin from sweet sorghum stem. Ind. Crop Prod.32, 21–28.

Sun, R.C., Tomkinson, J., Bolton, J., 1999. Effects of precipitation pH on thephysicochemical properties of the lignins isolated from the black liquor ofoil palm empty fruit bunch fibre pulping. Polym. Degrad. Stabil. 63 (2),195–200.

Thring, R.W., Vanderlaan, M.N., Griffin, S.L., 1996. Fractionation of Alcell lignin bysequential solvent extraction. J. Wood Chem. Technol. 16 (2), 139–154.

Toledano, A., Garcia, A., Mondragon, I., Labidi, J., 2010a. Lignin separation andfractionation by ultrafiltration. Sep. Purif. Technol. 71, 38–43.

Toledano, A., Serrano, L., Garcia, A., Mondragon, I., Labidi, J., 2010b. Comparativestudy of lignin fractionation by ultrafiltration and selective precipitation.Chem. Eng. J. 157, 93–99.

Toledano, A., 2012. Lignin Extraction, Purification and Depolymerization Study.PhD Thesis. Universidad del Pais Vasco, Pais Vasco, Spain.

Urgatondo, V., Mitjans, M., Vinardell, M.P., 2009. Applicability of lignin fromdifferent sources as antioxidants based on the protective effects onlipid peroxidation induced by oxygen radicals. Ind. Crop Prod. 30 (2),184–187.

Wallberg, O., Jonsson, A.S., Wimmerstedt, R., 2003. Ultrafiltration of kraft blackliquor with a ceramic membrane. Desalination 156, 145–153.

Wanrosli, W.D., Zainuddin, Z., Law, K.N., Asro, R., 2007. Pulp from oil palm frondsby chemical processes. Ind. Crop Prod. 25, 89–94.

Yuan, T.Q., He, J., Xu, F., Sun, R.C., 2009. Fractionation and physicochemical analysisof degraded lignins from the black liquor of Eucalytup pellita KP-AQ pulping.Polym. Degrad. Stabil. 94, 1142–1150.