sugarcane bagasse pretreatment using three imidazolium-based

Karatzos et al Biotechnology for Biofuels 2012 562httpwwwbiotechnologyforbiofuelscomcontent5162

RESEARCH Open Access

Sugarcane bagasse pretreatment using threeimidazolium-based ionic liquids mass balancesand enzyme kineticsSergios Kimon Karatzos1 Leslie Alan Edye1 and William Orlando Sinclair Doherty2

Abstract

Background Effective pretreatment is key to achieving high enzymatic saccharification efficiency in processinglignocellulosic biomass to fermentable sugars biofuels and value-added products Ionic liquids (ILs) still relativelynew class of solvents are attractive for biomass pretreatment because some demonstrate the rare ability to dissolveall components of lignocellulosic biomass including highly ordered (crystalline) cellulose In the present study threeILs 1-butyl-3-methylimidazolium chloride ([C4mim]Cl) 1-ethyl-3-methylimidazolium chloride ([C2mim]Cl) 1-ethyl-3-methylimidazolium acetate ([C2mim]OAc) are used to dissolvepretreat and fractionate sugarcane bagasse In theseIL-based pretreatments the biomass is completely or partially dissolved in ILs at temperatures greater than 130degCand then precipitated by the addition of an antisolvent to the IL biomass mixture For the first time mass balancesof IL-based pretreatments are reported Such mass balances along with kinetics data can be used in processmodelling and design

Results Lignin removals of 10 mass of lignin in bagasse with [C4mim]Cl 50 mass with [C2mim]Cl and 60mass with [C2mim]OAc are achieved by limiting the amount of water added as antisolvent to 05 waterIL massratio thus minimising lignin precipitation Enzyme saccharification (24 h 15FPU) yields ( cellulose mass in startingbagasse) from the recovered solids rank as [C2mim]OAc(83)gtgt[C2mim]Cl(53) = [C4mim]Cl(53) Composition of[C2mim]OAc-treated solids such as low lignin low acetyl group content and preservation of arabinosyl groups arecharacteristic of aqueous alkali pretreatments while those of chloride IL-treated solids resemble aqueous acidpretreatments All ILs are fully recovered after use (100 mass as determined by ion chromatography)

Conclusions In all three ILs regulated addition of water as an antisolvent effected a polysaccharide enrichedprecipitate since some of the lignin remained dissolved in the aqueous IL solution Of the three IL studied [C2mim]OAc gave the best saccharification yield material recovery and delignification The effects of [C2mim]OAcpretreatment resemble those of aqueous alkali pretreatments while those of [C2mim]Cl and [C4mim]Cl resembleaqueous acid pretreatments The use of imidazolium IL solvents with shorter alkyl chains results in accelerateddissolution pretreatment and degradation

Keywords Ionic liquids Pretreatment Sugarcane bagasse Enzyme hydrolysis

Correspondence ledyequteduau1School of Chemistry Queensland University of Technology GPO Box 2434Brisbane QLD 4001 AustraliaFull list of author information is available at the end of the article

copy 2012 Karatzos et al licensee BioMed Central Ltd This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (httpcreativecommonsorglicensesby20) which permits unrestricted use distribution andreproduction in any medium provided the original work is properly cited

Karatzos et al Biotechnology for Biofuels 2012 562 Page 2 of 12httpwwwbiotechnologyforbiofuelscomcontent5162

BackgroundLignocellulosics whether in the form of dedicated en-ergy crops such as switchgrass agricultural residuessuch as sugarcane bagasse or from forestry residuespresent an abundant renewable energy resource Withworld energy demand predicted to increase in the nearfuture [1] and fossil fuel reserves being depleted andnon-renewable biomass resources have drawn much at-tention as renewable and sustainable feedstocks for al-ternative fuels and chemicalsThe conversion of the lignocellulosic biomass poly-

saccharides to ethanol fuel andor other products offermentation (eg butanol) involves hydrolysis fermen-tation and product separation However these polysac-charide molecules are not readily hydrolysed since theyare contained in the chemically recalcitrant and struc-turally complex lignocellulosic matrix This matrix isdesigned by nature to protect the plant organism fromchemical and biological attack Therefore a pretreatmentstep is added to the process prior to hydrolysis in orderto improve saccharification of the polysaccharidesMost pretreatment technologies are either physical

(eg size comminution steam explosion and hydro-thermolysis) or chemical utilizing organic solvents acidsor alkalis [2-4] These chemical pretreatment technol-ogies occur by either acid or alkali mechanisms at hightemperatures and pressures Acid pretreatments areknown to dissolve hemicellulose and cleave arabinosylglycosidic linkages whereas alkali pretreatments dissolvelignin and cleave acetyl bonds [5]Recently ionic liquids (ILs) have drawn a great deal of

attention as solvents for pretreatment of lignocellulosicsILs are a class of organic salts that are liquid at temper-atures below 100degC Many ILs are non-volatile non-explosive stable at a wide range of temperatures andreaction-condition severities and compatible with awide array of organic and inorganic functional chemicalsand solvents [6] Pretreatment of the lignocellulosicmatrix with ionic liquids occurs by the dissolution-then-precipitation of solids which exhibit reduced crystallin-ity and improved enzyme saccharification of cellulose[7-11] For a thorough review which focuses on dissol-ution of biomass lists 74 ILs that have been tested forthe ability to dissolve cellulose wood pulp hemicellulose

Table 1 Effect of ionic liquid choice on bagasse dissolution

IL mg starting mass

Startingmass

Undissolvedsolids

Dissolved-and-resolids

[C4mim]Cl 224 37 43

[C2mim]Cl 224 15 26

[C2mim]OAc 224 4 56

Masses expressed on dry basis

and lignin and also includes other solvents of relevanceto woody biomass processing viz choline urea mixture - adeep eutectic salt and N-methylmorpholine-N-oxidemonohydrate - a conventional cellulose solvent thereader is referred to Pinkert et al [12] While it isexpected that ionic liquids impart compositional changesand cleave covalent bonds of the original biomass in orderto dissolve its components little is known about the na-ture of these changes and how they vary with varyingionic liquids In addition there are few accounts of fullmass closures of ionic liquid pretreatments For exampleArora et al [13] reported mass balances of [C2mim]OActreated switchgrass (160degC 3 h 3 loading) and Sunet al [14] reported mass balances for [C2mim]OAc treat-ment of southern yellow pine (110degC for 16 h 5 loading)These few accounts do not provide direct comparisonsbetween ILs while they do not account for IL massesand do not differentiate between hemicellulose compo-nents in the recovered solids Three most cited ionicliquids for biomass pretreatment are the imidazoliums1-butyl-3-methylimidazolium chloride or [C4mim]Cl 1-ethyl-3-methylimidazolium chloride or [C2mim]Cl and1-ethyl-3-methylimidazolium acetate or [C2mim]OAcThis study a) investigates the structural and composi-tional characteristics of sugarcane bagasse pretreatedwith each of the aforementioned ionic liquids and b) pro-vides full mass balances for these pretreatment processes

Results and discussionDissolution of bagasse in ILsThe dissolution of bagasse in the three ILs under identi-cal conditions (150degC 90 min and 5 mass bagasse inIL conditions as in previous work [10]) was investigatedand the results are shown in Table 1 It has to be notedthat the [C2mim]OAc dissolution at this loading (5mass) was very viscous and hard to stirThe effect of cation size is exhibited when comparing

[C4mim]Cl with [C2mim]Cl In agreement with the lit-erature [12] the smaller [C2mim] cation imparts higherdissolution (85 cf 63) and this may be due to theenhanced penetration of the smaller solvent moleculeresulting in higher dissolution The effect of anion isstudied by comparing [C2mim]Cl to [C2mim]OAc Theacetate anion seems to favour dissolution as opposed to

Mass fraction

covered Dissolvedunrecovered

Totaldissolved

Unrecoveredtotal dissolved

21 63 033

60 85 070

40 96 042


losses Moreover losses may be exacerbated due to thefact that dissolution (96) is well into the last recalci-trant bagasse fraction It is probable that by reducing theseverity of the treatment conditions the ratio of dissol-ution to losses will be improved Acetate has a higherhydrogen bond basicity than chloride [15] and thus itsability to disrupt hydrogen bonds and dissolve celluloseis higher This positive correlation between the hydrogenbond basicity (β a Kamler-Taft solvation parameter) ofthe IL anion and the ILrsquos ability to dissolve cellulose orlignocellulose is reported in the literature [15] The su-perior dissolving capacity of [C2mim]OAc has also beenreported by Sun et al [14] who measured 935 dissol-ution extent of southern yellow pine in [C2mim]OAcand only 26 in [C4mim]Cl under the same conditions(particle size 025 ndash 050 mm 5 mass loading 110degCfor 16 h)

Fractional precipitation by stepwise addition of waterThe potential of using incremental amounts of water tofractionally precipitate IL-dissolved bagasse was testedon three ILs and the mass balances of each processdetermined The fractional precipitation process wasdesigned to yield a polysaccharide-rich and a lignin-richfraction using two incremental additions of water (andacidification to pH le10 and a third addition of water toprecipitate remaining dissolved material)The first water addition used for partial precipitation

of dissolved solids resulted to a water IL mass ratio of05 Preliminary qualitative studies based on visualobservations of fractional precipitation with water(where bagasse soda lignin -prepared in house- and Avi-cel were used) indicated that this water amount shouldprecipitate all cellulose and keep lignin in solution forwater-[C4mim]Cl mixtures and possibly also for water-[C2mim]OAc mixtures It was also indicated that ligninprecipitation was complete at a water[C4mim]Cl massratio of 20 However native bagasse lignin dissolved inIL is likely to have different properties to lignin extractedwith aqueous NaOH and then dissolved in IL Maximumlignin recovery was ensured by acidification to a pH le10and the addition of a further 15 IL mass equivalents of

Table 2 Compositional analysis of SF1 solids from pretreatme

Sample dry mass

Mass recovery Ash AIL ASL Total lignin

plusmn04 plusmn04 plusmn01

Untreated (extracted) 100 31 208 54 262

[C4mim]Cl 90 35 208 54 261

[C2mim]Cl 48 61 250 38 287

[C2mim]OAc 66 57 99 60 158

AIL Acid insoluble lignin ASL Acid Soluble Lignin

water Water-[C2mim]Cl solutions are assumed to be-have in a similar manner to water-[C4mim]Cl solutionsBagasse pretreatments with [C4mim]Cl [C2mim]Cl

and [C2mim]OAc under identical conditions (35 min at150degC 5 bagasse in IL (25 for [C2mim]OAc))imparted partial dissolution and the polysaccharide richsolid fractions (SF1) were recovered using water addition(water IL mass ratio of 05) These conditions are delib-erately less severe than in the first dissolution experi-ments so as to lessen the degradation reactions andallow for the recovery of sufficient SF1 solid mass formass balance and further analysis The SF1 solids werewashed and freeze-dried prior to analysis and enzymesaccharification Bagasse was extracted with water andethanol prior to treatment since better mass balance clo-sures are obtained by removing non-structural moleculeswhich can interfere with characterisation of solid andliquid fractions

Compositional analysisThe composition of SF1 fractions from each IL pretreat-ment (at 150degC for 35 min with 25 min temperature ramp)are shown in Table 2 These solids contain both theundissolved bagasse and the dissolved-then-precipitatedbagasse and the proportion of each has not been mea-sured However they have been measured (at 150degC for90 min see Table 1) Of course the extents of biomassdissolution and degradation to material not recoveredin the solid fraction (ldquolossesrdquo to liquid fraction) wouldboth be expected to increase with temperature and treat-ment timeIn the mass balance experiments reported in Table 2

the extent of dissolution has been estimated indirectlyfrom the solids ldquolossesrdquo (and ratios of dissolved mass tolosses derived from Table 1) For the [C4mim]Cl dissol-ution a ratio of unrecovered to total dissolved mass of13 was observed (Table 1) This ratio was found to be123 for extracted bagasse The ldquolossesrdquo incurred for[C4mim]Cl are ca 10 mass (Table 2) thus the dissol-ution extent in these experiments for [C4mim]Cl is esti-mated to be around 23 mass The other two ILs areknown to effect different ratios of dissolution extents tolosses (Table 1) and the losses incurred in these ILs

nt of ethanol-extracted bagasse with three different ILs

Ratios

Glucan Xylan Arabinan Acetyl Arabinan Xylan Acetyl Xylan

plusmn02 plusmn05 plusmn008 plusmn005

449 222 150 311 007 014

476 205 106 296 005 014

534 110 060 175 005 016

682 133 158 142 012 011


(34 - 52 Table 2) suggest that the dissolutions areclose to complete These estimates were confirmed visu-ally it was observed that [C4mim]Cl contained a verylarge amount of undissolved fibre [C2mim]OAc only asmall amount and [C2mim]Cl almost noneFor [C4mim]Cl-treated solids the compositional

changes are small with a slight cellulose enrichmentThis is primarily because only ca 23 of the bagassedissolved For [C2mim]Cl the SF1 is rich in celluloseand lignin and low in hemicellulose saccharides Giventhe fact that [C2mim]Cl dissolves and degrades bagassecomponents faster than [C4mim]Cl it is not surprisingthat the [C2mim]Cl solids are low in hemicellulose com-ponents The [C2mim]OAc SF1 is rich in cellulose andarabinose and it is low in lignin xylan and acetyl con-tent The [C2mim]Cl SF1 is also rich in cellulose butalso rich in lignin Hemicellulose appears to have beensubstantially removed by [C2mim]Cl treatmentIt is tempting to speculate that the acetate IL and

chloride ILs cleave different bonds It appears that theacetate IL effects preferential delignification and deacety-lation whereas the chloride ILs effect preferential removalof arabinose and xylan In terms of solids composition

Table 3 Mass balance of bulk biomass and of biomass compo

Dry mass (mg) Bulk biomass Ash Lignin (AIL + ASL)

[C4mim]Cl

Untreated 1461 46 383

SF1 1318 47 344

LF1 92 nd 04

LF1 oligo

Total mass recovered 1411 47 344

[C2mim]Cl


SF1 643 39 185

LF1 446 nd 3

LF1 oligo


[C2mim]OAc


SF1 463 26 73

LF1 88 nd 17

LF1 oligo


Masses (average of duplicates)Masses expressed on dry basis Glucan is considered equal to cellulose HMF (Hydroequivalents SF1 solid fraction 1 recovered from the first water addition (waterIL m(waterIL mass ratio = 05) LF1 oligo oligosaccharides in LF1 nd not determined nThe estimates of standard deviation (absolute based on duplicate IL pretreatmentscomponent) in SF1 are 2 for glucan 2 for xylan 3 for arabinan 1 for acetylstandard deviation for the oligosaccharides are 1 for glucan 2 for xylan 18 foglucan 02 for xylan 15 for arabinan and 07 for acetyl The unacceptably highthe liquid fractions This mass does not represent all the lignin mass in the LF1 bu

the acetate IL appears to effect a similar outcome toaqueous alkali treatment while the chloride ILs producea similar outcome to dilute acid treatments This is a keyfinding Note that the aqueous acid and alkali treatmentsinvolve biomass dissolution and decomposition while ILtreatments involve dissolution decomposition and pre-cipitation Thus the chemical processes involved are dif-ferent but the gross compositional changes are similarThe acetate may be expected to be basic and be more

reactive than chloride because acetate is a stronger basethan chloride (acetic acid pKa = 475 cf HCl pKa = minus70)but these chemical behaviours do not necessarily holdtrue for non-aqueous IL solutions

Mass recovery of bagasse components after pretreatmentMass balances for bagasse components after the firstwater addition (waterIL mass ratio of 05 yielding SF1and LF1) of the three IL pretreatments ([C2mim]OAc[C4mim]Cl [C2mim]Cl) were determined and areshown in Table 3The composition of the SF1 solids and the LF1 liquids

(sum of monosaccharides and oligosaccharides) weredetermined The total mass accounted for by the mass

nents from three treatments with different ILs

Glucan Xylan Arabinan Acetyl HMF Furfural

656 324 22 46 na na

627 270 14 39 na na

16 60 10 7 06 12

13 58 2 6

636 317 24 46

605 299 20 42 na na

343 71 4 11 na na

188 214 14 28 12 14

175 204 3 27

516 267 20 39

314 155 10 22 na na

316 62 7 7 na na

1 85 5 nd 11 18

bdl 83 5 nd

316 125 12 7

xymethyl furfural) is expressed as glucan equivalents and furfural as xylanass ratio = 05) LF1 liquid fraction 1 recovered from the first water additiona not applicable bdl barely detectable 3 df) of the recovery (and analysis) of these components (asmass startingand 2 for lignin (acid soluble + acid insoluble) In LF1 these estimates ofr arabinan and 3 for acetyl and for the monosaccharides they are 02 forstandard deviation for arabinan is attributed to its very low concentrations in

t only the recoverable lignin mass in the sum of SF2 and SF3 precipitates


balance determination protocol differs greatly betweenILs For [C4mim]Cl the original bagasse mass accountedfor is 97 mass (1411 mg out of 1461 mg) and this maybe attributed to the low dissolution extent achieved bythis treatment For [C2mim]Cl the mass recovery is 81mass (1089 mg out of 1346 mg) and for [C2mim]OAc79 mass (551 mg out of 699 mg) For all 3 IL treat-ments the remaining bagasse mass (not accounted forby mass balance determinations) is mainly (more thanhalf ) lignin that was not recoverable from the liquidfraction followed by polysaccharides (glucan followed byxylan for the chloride ILs and xylan only for [C2mim]OAc) Note that in the liquid fraction of the [C2mim]OAc pretreatment the biomass-derived acetyl content(representing ca 3 of starting bagasse mass) is not de-tectable since the IL counter ion is acetateThe percent mass of each starting bagasse component

in each pretreatment fraction (viz solid fraction liquidfraction oligosaccharides and liquid fraction monosac-charides) is listed in Table 3 Note that the degradationproducts HMF and furfural represent a small fraction ofthe mass in the liquid fraction and are included in themass balances listed in Table 3 as glucan and xylanequivalentsFor [C4mim]Cl 95 mass of the starting cellulose

83 of xylan and 90 of lignin are recovered in the solidfraction Out of the 10 lignin in the liquid fractiononly 03 was recoverable (see Table 4) indicating thatthe vast majority of the lignin mass remaining in the li-quid fraction after addition of 05 IL mass equivalents ofwater is in a form that cannot be recovered by furtheradditions of water or acidification (ie it is soluble andlikely low molecular weight material) Hemicellulosecomponents are depolymerised preferentially while a bigpart of the arabinose (36) is removed all of which is in

Table 4 Mass recovery and lignin content of solidsrecovered from the liquid fraction after treatment withthree ILs

mg mass mg ( startingmass of lignin)

Sample Recoveredmass

Lignincontent

Ligninrecovery

pH

[C4mim]Cl SF2 12 31 04 (03) 61

[C4mim]Cl SF3 03 nd nd 10

[C4mim]Cl TOTAL 15 04 (03)

[C2mim]Cl SF2 108 30 32 (15) 36



[C2mim]OAc SF2 582 23 134 (99) 70

[C2mim]OAc SF3 122 26 32 (17) 10

[C2mim])OAc TOTAL 704 166 (117)


the monomeric form The arabinofuranosyl glycosidiclinkages are acid labile and consequently arabinose lossis a characteristic of acid treatments [5] The fact thatarabinose is found in the monomeric form indicates ei-ther that the arabinosyl groups removed are terminal orif they are not that the ester bonding of arabinose to lig-nin is concomitantly cleavedMass balance analysis for [C2mim]Cl treatment reveals

the same trends as for [C4mim]Cl but since dissolutionis near complete in this IL the trends are more pro-nounced Lignin and cellulose are the predominant com-ponents of the recovered solids reflecting the relativethermal and chemical stability of these polymers in solu-tion when compared to hemicelluloses Out of 78 massarabinan in the liquid fraction 55 mass is in mono-meric form Out of 50 lignin expected in the liquidfraction only 15 is recoverablePreferential hemicellulose removal resulting in lignin

and cellulose enrichment of the treated solids is also acharacteristic of dilute acid treatments The chlorideimidazolium ILs appear to remove arabinose preferen-tially Note that in previous work [10] where the undis-solved fraction of bagasse after [C4mim]Cl dissolutionwas analysed it was shown that cellulose dissolved to agreater extent than hemicelluloses (the effect of dissol-ution only) The analysis of SF1 here indicates thathemicellulose is preferentially removed (the net effect ofdissolution and reprecipitation) Cellulose is preferen-tially dissolved but at a high DP (degree of polymerisa-tion) and mostly recovered by precipitation with theaddition of water whereas the little hemicellulose thatdissolves depolymerises and becomes soluble in thewater-IL mixture (ie does not reprecipitate)The mass distributions of bagasse components after

[C2mim]OAc pretreatment are also shown in Table 3While 100 mass of the original cellulose is recoveredin the solid fraction the lignin content is reduced to 40mass Cellulose appears to get solubilised in a high DPform since it is all precipitated with the addition of waterand none of it is found in the aqueous liquid fractionThese features indicate that [C2mim]OAc affords excel-lent cellulose preservation and substantial delignifica-tion However out of the 60 mass lignin extracted inthe liquid fraction only 117 was recoverable Sincearabinose content is comparatively high in the solid frac-tion and found in the oligomeric liquid fraction only itcan be concluded that arabinose is preserved and thatno terminal arabinosyl groups have been removed (iethe arabinosyl glycosidic linkages in hemicellulose arestable in [C2mim]OAc) Finally the acetate content ofthe solid fraction is reduced by 70 mass indicating sub-stantial deacetylationThe arabinan recovery is inflated (115 mass) and

this is attributed to the large standard deviation of

Table 5 Assignments of FTIR-ATR absorption bands forbagasse

Band position(cm-1)

Assignment

1730 C = O stretching vibration in acetyl groups ofhemicelluloses

1600 C = C stretching vibration in aromatic ring of lignin


1421 CH2 scissoring at C(6) in cellulose

1368 Symmetric CndashH bending in cellulose

1236 C-O stretching vibration in lignin xylan and estergroups

1100 O-H association band in cellulose and hemicelluloses(associated with crystalline cellulose)

1030 C-O stretching vibration in cellulose and hemicelluloses

974 C-O stretching vibration in arabinosyl side chains inhemicellulose

895 Glucose ring stretch C1-H deformation

[17-20]


arabinose measurements in the liquid fraction as dis-cussed earlierOverall the [C2mim]OAc pretreatment affords dis-

tinctly different compositional changes to the chlorideIL pretreatments These distinct differences are delignifi-cation deacetylation preservation of cellulose glycosidicbonds (as deduced from the absence of cellulose mass inthe liquid fraction) and preservation of arabinosyl groupsin hemicellulose These differences are similar to the dif-ferences between acid and alkali aqueous pretreatmentsNote that previous work by Fu et al [16] where triticalestraw was treated with six different ILs is in agreementwith the present study in that [C2mim]OAc affords ex-tensive delignification while preserving the cellulosefraction of the treated biomass

ATR-FTIRThe SF1 from each IL was also analysed using ATR-FTIR and the spectra are shown in Figure 1 Table 5 liststhe assignments of the absorption bands of interest Ingeneral these spectra reflect the compositional charac-teristics discussed above For example it is clearly dis-cernible that the band absorbances which arecharacteristic of lignin are relatively low in the spectrumof SF1 treated with [C2mim]OAc when compared toother spectraInfrared spectra were also used to estimate crystallinity

of each SF1 The absorbance at 1421 cm-1 is viewed astypical of crystalline regions of cellulose and the absorp-tion band at 893 cm-1 typical of amorphous regions theratio of these two bands represents a crystallinity index

PO

LI

Wavenumber

[C2mim]Cl [C2mim]OAc [C4mim]Cl Untreated

Figure 1 FTIR spectra of bagasse treated with different ILs (absorbanc

(CrI) also known as lateral order index [17] The crystal-linity indices of the solids recovered from all three ILtreatments (019 for [C2mim]OAc 021 for [C4mim]Cland 037 for [C2mim]Cl) were significantly lower that ofthe untreated bagasse solids (088) Decrystalisation ofcellulose is a known and unique effect of IL pretreat-ments [10] The estimate of the standard deviation (ab-solute) for this crystallinity index-measurement is 002(based on duplicate IL pretreatments 3 df) These indi-ces combined with the observed shift of the 1034 cm-1

band to lower wavenumber represent a significant loss

LYSACCHARIDES

GNIN

s (cm-1)

e ndash common scale)


of crystallinity of cellulose after treatment in all threeILs Note that despite only ca 23 dissolution of ba-gasse in [C4mim]Cl the SF1 CrI is significantly lowerthan starting bagasse As shown in previous work by theauthors the cellulose in the IL-undissolved fraction canalso undergo decrystallisation [10]

Enzyme saccharificationThe progress of saccharification resulting from each ILtreatment for cellulose and hemicellulose (as xylan) wasmonitored and is plotted in Figure 2 Initial rates of cel-lulose saccharification are very fast for all IL treatmentsand this is the effect of cellulose decrystallisation whichis common to all three treatments The extent of sac-charification is higher for the [C2mim]+ ILs than for the

Figure 2 Glucan and xylan saccharification of extractedbagasse treated with 3 ILs Glucan refers to glucan in pretreatedsolids and includes glucan equivalents of cellobiose released Xylanrefers to xylan in pretreated solids

[C4mim]+ IL and of the [C2mim]+ salts the chlorideanion gives a higher saccharification yield Given that[C2mim]OAc delignifies as well as decrystallises bagasseit is surprising that [C2mim]Cl yielded a higher final sac-charification This is possibly due to the different lignin-hemicellulose bonds (as identified by compositional andinfrared analysis and discussed above) that survive in thesolids after these two pretreatments The fast initial sac-charification rates resulting from [C4mim]Cl treatmentreflect the presence of structural changes while the lowfinal saccharification yields are a consequence of a lowerextent of dissolution and indicate the absence of com-positional and covalent bonding perturbation in the un-dissolved solidsInterestingly the hemicellulose saccharification of

[C2mim]OAc treated bagasse proceeds faster and closerto completion than for the other two ILs This showsthat the delignification achieved by [C2mim]OAc resultsin improved hemicellulose saccharification results whilefor the chloride IL treated solids the hemicellulose is notas accessible to enzymes due to the persistence of ligninIt is likely that covalent linkages between lignin andhemicelluloses survive chloride IL treatment and limitthe extent of saccharification of hemicellulosesThe cellulose and hemicellulose saccharification

extents achieved by each IL at 24 h as mass theoreticalin the starting bagasse (prior to pretreatment) areshown in Figure 3 Factoring in both the saccharificationextent and the recovery of the starting polysaccharidemass in the pretreated solids [C2mim]OAc affords thehighest cellulose conversion due to a combination ofrapid saccharification and no degradation of celluloseupon pretreatment [C2mim]Cl and [C4mim]Cl have

Figure 3 Fraction of starting bagasse polysaccharidessaccharified in 24 h (15 FPU g-1 glucan) after pretreatment inthree ILs Glucan includes glucan equivalents of cellobiose released


similar theoretical cellulose conversions The muchhigher dissolution extent in [C2mim]Cl did not benefitthe overall performance of this pretreatment since lossesin [C2mim]Cl were much higher than in [C4mim]Cl Interms of hemicellulose saccharification yield [C4mim]Clperforms best mainly because more hemicellulose is pre-served in polymeric and recoverable formIn general and in agreement with the literature for the

cations it appears that the shorter alkyl chain of[C2mim]Cl (cf [C4mim]Cl) imparted faster dissolutionand greater extent of saccharification However higherdissolution rates were accompanied by higher degrad-ation rates The anion effect is greater since it impartsentirely distinct dissolution patterns In the case of acet-ate compared to chloride the acetate ion appears to im-part a more alkali-resembling effect while the chlorideones a more acid-resembling effect

Lignin recoveryThe second addition of water to the liquid fractions ofthe three IL pretreatments to a waterIL mass ratio of20 yielded solid fractions (SF2) Acidification of the li-quid fractions and a third water addition yielded someadditional precipitate (SF3) For all SF2 and SF3 sam-ples lignin content was measured by the acetyl bromidemethod and FTIR spectra were obtained SF2 and SF3lignin contents and recovered masses are reported inTable 4 The weight of all liquid fraction precipitates andtheir lignin content was used to determine the totalamount of lignin recoverable from the liquid fraction ofeach IL pretreatment The lignin recovery from the li-quid fraction of all three IL pretreatments was low (03to 117 mass of starting lignin)

Precipitate from [C2mim]Cl liquid fraction Precipitate from [C4mim]Cl liquid fraction Precipitate from [C2mim]OAc liquid fraction Untreated extracted bagasse

Wavenu

Figure 4 FTIR spectra of precipitate recovered after precipitation in 3(absorbance ndash common scale)

The ATR-FTIR spectra for the solids precipitated fromeach IL pretreatment from the 35 mass ratio wateraddition (SF3) are shown in Figure 4 Spectra of SF2 preci-pitates were also obtained but were not different to thoseof the SF 3 precipitates These spectra have intense absor-bances at the hemicellulose characteristic bands between1175 cm-1 and 1000 cm-1 This indicates that the majorityof the non-lignin component of these fractions (ca 70mass) is comprised of hemicellulose This is in agreementwith previous discussion demonstrating the preservationof lignin-hemicellulose bonding upon biomass dissolutionin ILs The [C2mim]OAc precipitate in particular has astrong characteristic band at 974 cm-1 indicative of arabi-nosyl groups This again is in agreement with previousdiscussion supporting preservation of covalent bonds toarabinosyl moieties during [C2mim]OAc dissolutionwhilst they are labile in chloride IL dissolutionsUndoubtedly a large proportion of the dissolved lignin

remains in the water-IL mixture even after addition ofmore water (35 water IL mass ratio) and lowering ofthe pH to le10 In this fractional precipitation approachthere are difficulties in quantitatively recovering ligninsince only a small fraction of it precipitates Furthermorethe precipitate is far from pure lignin as it can containup to 70 mass hemicellulose While the cellulose frac-tion can be obtained in an enriched and extensivelydecrystallised form the poor lignin recovery and co-precipitation of hemicelluloses argues against the use ofthis approach in an industrial setting

Ionic liquid recoveryThe recovery of the ionic liquid solvent forms part ofthe mass balance closure and it was measured using ion

mbers (cm-1)

5 water IL mass ratio (acidified to pH le1) in three ILs


chromatography of the liquid fraction of each pretreat-ment The recovery of all ions appears to be full (100mass) within experimental error The standard deviationof repeat analyses was 2 At least under laboratoryconditions and scale little or no degradation of ILoccurs and the IL could be completely recovered Cova-lent bonding of acetate ions from [C2mim]OAc to bio-mass [21] and the reaction of imidazole cations withsaccharides to form imidazole glycosides [22] may occurbut at the low biomass concentrations used here themagnitudes of these reactions are small by comparisonto the experimental error in the measurement of ILrecovery

ConclusionsMass balances are reported for pretreatment of bagassewith [C4mim]Cl [C2mim]Cl and [C2mim]OAc Theacetate IL preferentially removes lignin and acetyl whileit preserves arabinosyl groups resembling the effects ofaqueous alkali pretreatments Chloride ILs impart anopposite effect and resemble dilute acid pretreatmentsThe shorter IL alkyl chain accelerated the dissolutionpretreatment and degradation Regulated addition ofwater-antisolvent affected a polysaccharide rich precipi-tate by maintaining dissolved lignin in all three ILs Interms of saccharification yields material recovery anddelignification [C2mim]OAc ranks as the most suitableIL for biomass pretreatment The mass of all ILs usedwas fully recovered Future work should focus on eluci-dating the mechanisms of lignocellulose dissolution anddepolymerisation in acetate and chloride-based ILs Inaddition the distinct chemistries of these IL solventsmight be used to development an understanding of thereactions involved in other solvent based biomasspretreatments

MethodsBagasseSugarcane bagasse (Rocky Point sugar mill PimpamaQueensland) comprising long cuticle fibres and core pithparticles (mostly ca 5 mm to 50 mm) was air dried for aweek on metal trays then size reduced (to lt10 mm) witha knife mill before being mixed and subsampled by thecone and split method and stored at 4degC Before use thebagasse was ground (to lt2 mm) using an electric labmill (Retsch SM100 Haan Germany) The material wasground for ca 1 min per batch to avoid excess heatingplaced on top of two brass sieves (05 mm and025 mm) and in a sieve shaker for 20 min and the frac-tion collected between the two sieves was used as thestarting material Moisture content was measured gravi-metrically (convection oven 105degC overnight) beforeevery use and was 10 plusmn 1 mass except where otherwiseindicated

ChemicalsThe ionic liquids (1-butyl-3-methylimidazolium chloride[C4mim]Cl (ge 95) melting point as per MSDS (mp)73degC 1-ethyl-3-methylimidazolium chloride [C2mim]Cl(ge 95) mp 80degC and 1-ethyl-3-methylimidazoliumacetate [C2mim]OAc (ge 90) mp ndash20degC Sigma-Aldrich NSW) were all dried in a vacuum oven (at 80degCndash 90degC ca 4 mm Hg gt 12 h) prior to use Initial mois-ture content (at the time of weighing the IL for eachuse) was typically ca 2 of total mass for [C4mmim]Cland 1 for [C2mim]Cl and [C2mmim]OAc as measuredby Karl Fischer titration At this point it is worth notingthat although the mp of neat [C4mim]Cl is 73degC (as perthe MSDS provided by the manufacturer) its 2 mois-ture content was sufficient to maintain it in liquid phaseat room temperature Cellulose (Avicel PH-101) di-methyl sulphoxide (DMSO) (999) and Karl FischerHYDRANAL titrant 2E and solvent E were purchasedfrom Sigma-Aldrich (Sydney NSW) Cellulase β-glucosidase mixture (Accelerase 1000) was purchasedfrom Genencor (Danisco AS Denmark) Water wasMillipore-filtered and deionised (Milli-Q-plus) to a spe-cific resistivity of 182 μS at 25degC All other solvents andchemicals were analytical grade

Bagasse soda lignin preparationBagasse soda lignin was prepared by soda pulping of ba-gasse (175degC 2 h bagasse 10 mass NaOH 10 mass)and precipitating the resulting black liquor with acid(2 M H2SO4) add to reduce the pH to 30 The precipi-tate was then redissolved in aqueous NaOH (10 mass)and reprecipitated by addition of acid to reduce the pHto 30 The recovered lignin solids were washed anddried (40degC vacuum oven)

Karl Fischer titrationA Karl Fischer automated titrator (Radiometer Copen-hagen TIM 900) with ethanol based HYDRANALreagents was used to measure moisture content of ILsafter drying and prior to use

Preliminary dissolution experimentsNon-extracted bagasse was treated in [C4mim]Cl[C2mim]Cl [C2mim]OAc under conditions used in pre-viously reported work [10] (150degC 90 min and 5 massbagasse in IL) The treated mixture (partially dissolvedbagasse in IL) was then diluted with DMSO and the un-dissolved and the dissolved-then-precipitated (withwater) solid fractions were recovered dried and weighedas described in previous work by the authors [10] Theestimates of standard deviation (absolute) for this tech-nique (based on duplicate dissolution experiments 5degrees of freedom (df )) are 6 starting mass for

Karatzos et al Biotechnology for Biofuels 2012 562 0 of 12httpwwwbiotechnologyforbiofuelscomcontent5162

undissolved solids and 3 starting mass for dissolved-then-precipitated solids

Mass balance determinations for three IL treatmentsBagasse (025 mm ndash 05 mm) was extracted with ethanoland water using a Sohxlet device according to the NRELprotocol for biomass extractives [23] ILs (ca 30 g of ei-ther [C4mim]Cl or [C2mim]Cl or [C2mim]OAc in du-plicate) were weighed in sealable pressure glass tubes(ACE glass 50 mL) At this point IL (ca 05 g) wasweighed and set aside for IL recovery analysis (using ionchromatography) Extracted bagasse (35 moisture) (ca15 g for [C4mim]Cl and [C2mim]Cl and 075 g for[C2mim]OAc) was added to each pressure tube sealedwith Teflon stoppers and placed in an oil bath whichwas stabilised at 150degC with magnetic stirring at200 rpm Sealing the tubes prevented volatile losses suchas acetic acid (bp 1181degC) or furfural (bp 1617degC)from the degradation of xylose The tubes were left inthe oil bath for 60 min (25 min of which at temperatureramp and 35 min at 150degC) and upon removal placedin an ice bath with magnetic stirring After 2 min thetubes were removed from the ice bath and water wasadded equal to 05 mass fraction of the originally addedIL The tube was sealed again and agitated vigorouslyuntil a homogenous solution between water and ILappeared to form The contents of each tube were quan-titatively transferred into a preweighed polypropylenecentrifuge tube and centrifuged at 10000 times g for 20 minThe liquid contents of the centrifuge tube were decantedto a new preweighed polypropylene centrifuge tube andweighed (liquid fraction 1 or LF1) The pellet (SF1) wascentrifuge washed with distilled water (5 times 30 mL at10000 times g and 5 min - 10 min cycles) freeze dried over-night (minus85degC 80 mT) and weighed LF1 1 was precipi-tated with additional water resulting to a water IL massratio of 20 Precipitation and coagulation of solids wasaided by storing at 4degC overnight followed by shaker in-cubating at 55degC - 70degC for 60 min The resulting pre-cipitate (solid fraction 2 or SF2) was centrifuge washedfreeze dried and weighed The resulting liquid was acid-ified to pH le1 and mixed with a further 15 IL massequivalents of water to maximise precipitation of ligninin solution This final precipitate (solid fraction 3 orSF3) was centrifuge washed freeze dried and weighedLosses of liquid components to washings of pellets

were accounted for by weighing pellets prior to washingand after drying (it is assumed that the composition ofthese lost liquid components is the same as the bulk li-quid) Similarly subsampling for analysis was accountedfor by careful attention to mass changesSF1 and the starting biomass were characterised using

the NREL acid hydrolysis protocol [23] Solid fractions 2and 3 were characterised for lignin content with the

acetyl bromide protocol described by Iiyama and Wallis[24] The sample of LF1 was directly injected onto theHPLC and the Ion Chromatograph (IC) for quantifica-tion of monosaccharides and IL ions respectively whilethe soluble oligosaccharides were determined by acid hy-drolysis All methods are described in detail in the follow-ing sections The distribution of cellulose hemicelluloseand lignin between solid fractions liquid fraction mono-saccharides and liquid fraction oligosaccharides was finallyreported as mass of the components in the starting ma-terial (Table 3) The estimates of standard deviation (abso-lute based on duplicate IL pretreatments 3 df) of therecovery (and analysis) of these components (asmassstarting component) in SF1 are 2 for glucan 2 forxylan 3 for arabinan 1 for acetyl and 2 for lignin(acid soluble + acid insoluble) In LF1 these estimates ofstandard deviation for the oligosaccharides are 1 for glu-can 2 for xylan 18 for arabinan and 3 for acetyl andfor the monosaccharides they are 02 for glucan 02 forxylan 15 for arabinan and 07 for acetyl The unaccept-ably high standard deviation for arabinan is attributed toits very low concentrations in the liquid fractions

Compositional analysis of ldquosolid fraction 1rdquoCompositional analysis of SF1 samples was carried outusing the standard NREL procedure for determination ofstructural carbohydrates and lignin in biomass [25] Allsamples were freeze dried overnight prior to analysisEach sample (250 mg) was treated with H2SO4 (72mass) at 30degC for 1 h These samples and a sugar recov-ery standard (SRS containing known concentrations ofglucose xylose and arabinose) were then exposed to di-lute H2SO4 (4) at 121degC for 1 h The hydrolysis pro-ducts were determined by HPLC (Waters) equippedwith a RI detector (Waters 410) and a Bio-Rad HPX-87 H column operated at 85degC The mobile phase con-sisted of 5 mM H2SO4 with a flow rate of 06 mL min-1The glucose xylose and arabinose results were correctedfor acid decomposition using the mass recovery fromthe SRS The polysaccharide and acetyl mass contentwere calculated by conversion of the monosaccharideand acetic acid results with appropriate multiplicationfactors (090 for glucose 088 for xylose and arabinose0683 for acetic acid) In bagasse glucan content is con-sidered equal to cellulose content since there is no glu-cose in the hemicelluloses of sugarcane and sucrose hasbeen removed previously at the sugar mill The acid- in-soluble lignin (AIL) after acid hydrolysis was measuredas the mass loss of insoluble residue at 575degC The acid-soluble lignin (ASL) was measured by UVndashvis spectro-photometer (Cintra 40) at 240 nm with an extinctioncoefficient value of 25 L g-1 cm-1 [25] Ash was deter-mined by placing separate sample fractions at 575degC


Compositional analysis of monosaccharides in liquidfraction 1Each sample of LF1 (05 mL) was weighed in 15 mLEppendorf tubes and diluted with water (05 mL) Thecontents were vortexed thoroughly filtered through a045 μm nylon filter and injected to a Waters HPLC asdescribed earlier Glucose xylose arabinose and aceticacid masses were converted to glucan xylan arabinanand acetate masses using appropriate multiplication fac-tors (as listed earlier) In addition it was assumed thatthe detected hydroxymethylfurfural (HMF) and furfuralwere products of cellulose and xylan degradation re-spectively Therefore HMF and furfural masses wereconverted to cellulose and xylan mass equivalents usingmultiplication factors of 128 and 138 respectively [23]

Compositional analysis of oligosaccharides in liquidfraction 1Each sample of LF1 (05 mL) and SRS solution (05 mL)were weighed in 2 mL twist-top Eppendorf tubes dilutedwith water (1 mL) and acidified (with 72 mass H2SO4)to a pH of 03 The contents were vortexed thoroughlyand autoclaved (121degC for 60 min autoclaving did notaffect mass) After cooling to room temperature theautoclaved tube contents were filtered through a045 μm nylon filter and injected onto the HPLC systemdescribed earlier After SRS correction for acid decom-position of sugars and subtraction of the monosacchar-ide composition results the difference was converted topolysaccharide mass equivalents (using appropriatemultiplication factors as listed earlier) in order to arriveat the composition of the soluble oligosaccharides inLF1

ATR-FTIRA small amount of freeze dried fibre enough to coverthe surface of the probe was placed on the diamondprobe of a Thermo Nicolet 870 FTIR (software OMNIC73) The sample was pressed with an anvil to increasethe surface contacting the probe Sixty-four scans wereacquired for each spectrum and the two replicate spectrafor each sample were overlayed No differences in thereplicate spectra of this study were observed and thusonly the first spectrum of each sample was used foranalysis

Acetyl bromide for lignin quantification in solid fractions2 and 3The acetyl bromide method as described by Iiyama andWallis [24] was used to determine the mass lignin con-tent of solid fractions 2 and 3 Freeze dried solids (ca10 mg) were weighed in glass tubes and acetyl bromidein acetic acid (25 mass 10 mL) and then perchloricacid (70 mass 01 mL) were added The tubes were

sealed with Teflon screw caps and placed in temperaturecontrolled rotary shaker (70degC and 100 rpm for 30 min)After cooling to room temperature the tubes wereopened and 2 M NaOH (10 mL) and then glacial aceticacid (25 mL) were added After agitation absorbance(280 nm quartz cuvettes Cintra UV spectrometer) wasmeasured against glacial acetic acid The resulting solu-tion was analysed with a Cintra-40 UV spectrometer andthe absorbance was referenced to a cuvette with glacialacetic acid Dilutions with glacial acetic acid were neces-sary for some samples so that the absorbance was lt10The absorbance was converted to percent mass con-centration of lignin using an extinction coefficient of25 Lmiddotg-1middotcm-1 The extinction coefficient was determinedwith the use of a calibration curve based on bagasse ofknown lignin content The estimate of standard devi-ation (absolute) of this technique (duplicate samples ofuntreated bagasse and soda lignin 4 df) (as dry massof solid analysed) is 3

Enzymatic saccharification of solids from 3 IL treatmentsEnzymatic hydrolysis experiments (SF1 fraction) wereperformed in 20 mL scintillation vials on a rotary shaker(150 rpm 50degC) in volumes of 5 mL with a biomass loadof 50 mg cellulose equivalent and Accelerase 1000 (Gen-encor) activity of 15 FPU g-1 (25 μL of Accelerase asreceived) in 50 mM citrate buffer (pH 47) Samples(02 mL) were periodically removed placed in ice thenin boiling water (2 min) and centrifuged Cellobiose glu-cose and xylose concentrations were measured by HPLC(HPLC system as described earlier except a ShodexSPO-810 HPLC column at 85degC with a mobile phase ofultrapure water at 06 L min-1 were used) The glucoseand cellobiose results were converted to glucan massequivalents and xylose was converted to xylan massequivalents using appropriate multiplication factorsThe estimates of standard deviation (absolute) of this

analysis (based on duplicate IL pretreatmentsrsquo saccharifi-cation extents at different time points 18 df ) are 2mass of glucan and 09 mass of xylan

Recovery of ILIL set aside at the start of mass balance experiments(starting IL) was brought to a volume of 50 mL withdeionised water Similarly known masses of LF1 werediluted with deionised water and injected onto the ionchromatograph (Metrohm 761 with a conductivity de-tector) For cation analysis samples were injected onto aMetrosep C 2 150 (150 mm times 4 mm) column with anaqueous mobile phase (25 volume acetone 6 mM tar-taric acid and 075 mM dipicolinic acid) at 1 mL min-1For anion analysis samples were injected onto a Metro-sep ASupp5 (150 mm times 4 mm) column with an aqueousmobile phase (1 mM NaHCO3 and 32 mM Na2CO3) at


07 mL min-1 and suppressed by post-column additionof H2SO4 (50 mM) IL mass balance was determinedfrom the results of these analyses The estimate ofstandard deviation (absolute) of this technique (asmass of ions in starting IL) for both cations andanions and for all 3 ILs is 2 (based on duplicate ILpretreatments 6 df)

AbbreviationsIL Ionic Liquid [C4mim]Cl 1-butyl-3-methylimidazolium chloride [C2mim]Cl 1-ethyl-3-methylimidazolium chloride [C2mim]OAc 1-ethyl-3-methylimidazolium acetate

Competing interestsThe authors declare that they have no competing interests

Authorsrsquo contributionsAll work has been carried out by SK under the supervision of LE and WD Allauthors read and approved the final manuscript

AcknowledgementsFunding was generously provided by the Greek State ScholarshipFoundation (IKY) the Queensland Government and Queensland University ofTechnology Bagasse soda lignin was prepared by Dylan Cronin

Author details1School of Chemistry Queensland University of Technology GPO Box 2434Brisbane QLD 4001 Australia 2Sugar Research and Innovation QueenslandUniversity of Technology GPO Box 2434 Brisbane QLD 4001 Australia

Received 16 February 2012 Accepted 23 July 2012Published 24 August 2012

References1 EIA (US Energy Information Association) International Energy Outlook 2010

httpwwweiagovforecastsarchiveieo10indexhtml2 Kumar P Barrett DM Delwiche MJ Stroeve P Methods for pretreatment of

lignocellulosic biomass for efficient hydrolysis and biofuel productionInd Eng Chem Res 2009 48(8)3713ndash3729

3 Mosier N Wyman C Dale B Elander R Lee YY Holtzapple M Ladisch MFeatures of promising technologies for pretreatment of lignocellulosicbiomass Bioresour Technol 2005 96(6)673ndash686

4 Wyman CE Dale BE Elander RT Holtzapple M Ladisch MR Lee YYCoordinated development of leading biomass pretreatmenttechnologies Bioresour Technol 2005 96(18)1959ndash1966

5 Hamilton JK The behaviour of wood carbohydrates in technical pulpingprocesses Pure Appl Chem 1962 5(1ndash2)197ndash218

6 El Seoud OA Koschella A Fidale LC Dorn S Heinze T Applications of ionicliquids in carbohydrate chemistrya window of opportunitiesBiomacromolecules 2007 8(9)2629ndash2647

7 Edye LA Doherty WOS Fractionation of a lignocellulosic material 2007AU2007900603

8 Edye LA Doherty WOS Blinco JA Bullock GE The sugarcane biorefineryEnergy crops and processes for the production of liquid fuels andrenewable commodity chemicals International Sugar Journal 2006108(1285)19ndash20 22ndash27

9 Fort DA Remsing RC Swatloski RP Moyna P Moyna G Rogers RD Canionic liquids dissolve wood Processing and analysis of lignocellulosicmaterials with 1-n-butyl-3-methylimidazolium chloride Green Chem 20079(1)63ndash69

10 Karatzos S Edye LA Doherty WOS Enhanced saccharification kinetics ofsugarcane bagasse pretreated in 1-butyl-3-methylimidazolium chlorideat high temperature and without complete dissolution Bioresour Technol2011 102(19)9325ndash9329

11 Kilpelainen I Xie H King A Granstrom M Heikkinen S Argyropoulos DSDissolution of wood in ionic liquids J Agric Food Chem 2007 55(22)9142ndash9148

12 Pinkert A Marsh KN Pang SS Staiger MP Ionic liquids and theirinteraction with cellulose Chem Rev 2009 109(12)6712ndash6728

13 Arora R Manisseri C Li C Ong M Scheller H Vogel K Simmons B Singh SMonitoring and analyzing process streams towards understanding ionicliquid pretreatment of switchgrass (Panicum virgatum L) Bioenergy Res2010 3(2)134ndash145

14 Sun N Mustafizur R Qin Y Maxim ML Rodriacuteguez H Rogers RD Completedissolution and partial delignification of wood in the ionic liquid1-ethyl-3-methylimidazolium acetate Green Chem 2009 11(5)646ndash655

15 Brandt A Hallett JP Leak DJ Murphy RJ Welton T The effect of the ionicliquid anion in the pretreatment of pine wood chips Green Chemistry2010 12(4)672ndash679

16 Fu D Mazza G Tamaki Y Lignin extraction from straw by ionic liquidsand enzymatic hydrolysis of the cellulosic residues J Agric Food Chem2010 582915ndash2922

17 Širokyacute J Blackburn R Bechtold T Taylor J White P Attenuated totalreflectance Fourier-transform infrared spectroscopy analysis ofcrystallinity changes in lyocell following continuous treatment withsodium hydroxide Cellulose 2010 17(1)103ndash115

18 Labbeacute N Rials TG Kelley SS Cheng Z-M Kim J-Y Li Y FT-IR imaging andpyrolysis-molecular beam mass spectrometry new tools to investigatewood tissues Wood Sci Technol 2005 39(1)61ndash76

19 Pandey KK A study of chemical structure of soft and hardwood andwood polymers by FTIR spectroscopy J Appl Polymer Sci 199971(12)1969ndash1975

20 Sun JX Sun XF Sun RC Su YQ Fractional extraction and structuralcharacterization of sugarcane bagasse hemicelluloses Carbohydr Polym2004 56(2)195ndash204

21 Karatzos S Edye LA Wellard RM The undesirable acetylation of celluloseby the acetate ion of 1-ethyl-3-methylimidazolium acetate Cellulose 201119(1)307ndash312

22 Ebner G Schiehser S Potthast A Rosenau T Side reaction of cellulose withcommon 1-alkyl-3-methylimidazolium-based ionic liquids TetrahedronLett 2008 497322ndash7324

23 Sluiter A Ruiz R Scarlata C Sluiter J Templeton D Determination ofextractives in biomass National Renewable Energy Laboratory 2008Available online httpwwwnrelgovbiomasspdfs42619pdf

24 Iiyama K Wallis AFA An improved acetyl bromide procedure fordetermining lignin in woods and wood pulps Wood Sci Technol 1988 22(3)271ndash280

25 Sluiter A Hames B Ruiz R Scarlata C Sluiter J Templeton D Crocker DDetermination of structural carbohydrates and lignin in biomass NationalRenewable Energy Laboratory 2008 Available online httpwwwnrelgovbiomasspdfs42618pdf

doi1011861754-6834-5-62Cite this article as Karatzos et al Sugarcane bagasse pretreatmentusing three imidazolium-based ionic liquids mass balances and enzymekinetics Biotechnology for Biofuels 2012 562

Submit your next manuscript to BioMed Centraland take full advantage of

bull Convenient online submission

bull Thorough peer review

bull No space constraints or color figure charges

bull Immediate publication on acceptance

bull Inclusion in PubMed CAS Scopus and Google Scholar

bull Research which is freely available for redistribution

Submit your manuscript at wwwbiomedcentralcomsubmit

Abstract

Background

Results

Conclusions

Background

Results and discussion

Dissolution of bagasse in ILs

Fractional precipitation by stepwise addition of water

Compositional analysis

Mass recovery of bagasse components after pretreatment

ATR-FTIR

Enzyme saccharification

Lignin recovery

Ionic liquid recovery

Conclusions

Methods

Bagasse

Chemicals

Bagasse soda lignin preparation

Karl Fischer titration

Preliminary dissolution experiments

Mass balance determinations for three IL treatments

Compositional analysis of ldquosolid fraction 1rdquo

Compositional analysis of monosaccharides in liquid fraction 1

Compositional analysis of oligosaccharides in liquid fraction 1

ATR-FTIR

Acetyl bromide for lignin quantification in solid fractions 2 and 3

Enzymatic saccharification of solids from 3 IL treatments

Recovery of IL

Abbreviation

Competing interests

Authorsrsquo contributions

Acknowledgements

Author details

References

Karatzos et al Biotechnology for Biofuels 2012 562 of 12httpwwwbiotechnologyforbiofuelscomcontent5162

BackgroundLignocellulosics whether in the form of dedicated en-ergy crops such as switchgrass agricultural residuessuch as sugarcane bagasse or from forestry residuespresent an abundant renewable energy resource Withworld energy demand predicted to increase in the nearfuture [1] and fossil fuel reserves being depleted andnon-renewable biomass resources have drawn much at-tention as renewable and sustainable feedstocks for al-ternative fuels and chemicalsThe conversion of the lignocellulosic biomass poly-

saccharides to ethanol fuel andor other products offermentation (eg butanol) involves hydrolysis fermen-tation and product separation However these polysac-charide molecules are not readily hydrolysed since theyare contained in the chemically recalcitrant and struc-turally complex lignocellulosic matrix This matrix isdesigned by nature to protect the plant organism fromchemical and biological attack Therefore a pretreatmentstep is added to the process prior to hydrolysis in orderto improve saccharification of the polysaccharidesMost pretreatment technologies are either physical

(eg size comminution steam explosion and hydro-thermolysis) or chemical utilizing organic solvents acidsor alkalis [2-4] These chemical pretreatment technol-ogies occur by either acid or alkali mechanisms at hightemperatures and pressures Acid pretreatments areknown to dissolve hemicellulose and cleave arabinosylglycosidic linkages whereas alkali pretreatments dissolvelignin and cleave acetyl bonds [5]Recently ionic liquids (ILs) have drawn a great deal of

attention as solvents for pretreatment of lignocellulosicsILs are a class of organic salts that are liquid at temper-atures below 100degC Many ILs are non-volatile non-explosive stable at a wide range of temperatures andreaction-condition severities and compatible with awide array of organic and inorganic functional chemicalsand solvents [6] Pretreatment of the lignocellulosicmatrix with ionic liquids occurs by the dissolution-then-precipitation of solids which exhibit reduced crystallin-ity and improved enzyme saccharification of cellulose[7-11] For a thorough review which focuses on dissol-ution of biomass lists 74 ILs that have been tested forthe ability to dissolve cellulose wood pulp hemicellulose

Table 1 Effect of ionic liquid choice on bagasse dissolution

IL mg starting mass

Startingmass

Undissolvedsolids

Dissolved-and-resolids

[C4mim]Cl 224 37 43

[C2mim]Cl 224 15 26

[C2mim]OAc 224 4 56


and lignin and also includes other solvents of relevanceto woody biomass processing viz choline urea mixture - adeep eutectic salt and N-methylmorpholine-N-oxidemonohydrate - a conventional cellulose solvent thereader is referred to Pinkert et al [12] While it isexpected that ionic liquids impart compositional changesand cleave covalent bonds of the original biomass in orderto dissolve its components little is known about the na-ture of these changes and how they vary with varyingionic liquids In addition there are few accounts of fullmass closures of ionic liquid pretreatments For exampleArora et al [13] reported mass balances of [C2mim]OActreated switchgrass (160degC 3 h 3 loading) and Sunet al [14] reported mass balances for [C2mim]OAc treat-ment of southern yellow pine (110degC for 16 h 5 loading)These few accounts do not provide direct comparisonsbetween ILs while they do not account for IL massesand do not differentiate between hemicellulose compo-nents in the recovered solids Three most cited ionicliquids for biomass pretreatment are the imidazoliums1-butyl-3-methylimidazolium chloride or [C4mim]Cl 1-ethyl-3-methylimidazolium chloride or [C2mim]Cl and1-ethyl-3-methylimidazolium acetate or [C2mim]OAcThis study a) investigates the structural and composi-tional characteristics of sugarcane bagasse pretreatedwith each of the aforementioned ionic liquids and b) pro-vides full mass balances for these pretreatment processes

Results and discussionDissolution of bagasse in ILsThe dissolution of bagasse in the three ILs under identi-cal conditions (150degC 90 min and 5 mass bagasse inIL conditions as in previous work [10]) was investigatedand the results are shown in Table 1 It has to be notedthat the [C2mim]OAc dissolution at this loading (5mass) was very viscous and hard to stirThe effect of cation size is exhibited when comparing

[C4mim]Cl with [C2mim]Cl In agreement with the lit-erature [12] the smaller [C2mim] cation imparts higherdissolution (85 cf 63) and this may be due to theenhanced penetration of the smaller solvent moleculeresulting in higher dissolution The effect of anion isstudied by comparing [C2mim]Cl to [C2mim]OAc Theacetate anion seems to favour dissolution as opposed to

Mass fraction

covered Dissolvedunrecovered

Totaldissolved

Unrecoveredtotal dissolved

21 63 033

60 85 070

40 96 042






Sample dry mass




[C4mim]Cl 90 35 208 54 261

[C2mim]Cl 48 61 250 38 287

[C2mim]OAc 66 57 99 60 158







Ratios



449 222 150 311 007 014

476 205 106 296 005 014

534 110 060 175 005 016

682 133 158 142 012 011







[C4mim]Cl


SF1 1318 47 344

LF1 92 nd 04

LF1 oligo


[C2mim]Cl


SF1 643 39 185

LF1 446 nd 3

LF1 oligo


[C2mim]OAc


SF1 463 26 73

LF1 88 nd 17

LF1 oligo









656 324 22 46 na na

627 270 14 39 na na

16 60 10 7 06 12

13 58 2 6

636 317 24 46

605 299 20 42 na na

343 71 4 11 na na

188 214 14 28 12 14

175 204 3 27

516 267 20 39

314 155 10 22 na na

316 62 7 7 na na

1 85 5 nd 11 18

bdl 83 5 nd

316 125 12 7










Lignincontent

Ligninrecovery

pH

[C4mim]Cl SF2 12 31 04 (03) 61



[C2mim]Cl SF2 108 30 32 (15) 36



[C2mim]OAc SF2 582 23 134 (99) 70

[C2mim]OAc SF3 122 26 32 (17) 10

[C2mim])OAc TOTAL 704 166 (117)








Band position(cm-1)

Assignment











[17-20]






PO

LI

Wavenumber





LYSACCHARIDES

GNIN

s (cm-1)















Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References






Sample dry mass




[C4mim]Cl 90 35 208 54 261

[C2mim]Cl 48 61 250 38 287

[C2mim]OAc 66 57 99 60 158







Ratios



449 222 150 311 007 014

476 205 106 296 005 014

534 110 060 175 005 016

682 133 158 142 012 011







[C4mim]Cl


SF1 1318 47 344

LF1 92 nd 04

LF1 oligo


[C2mim]Cl


SF1 643 39 185

LF1 446 nd 3

LF1 oligo


[C2mim]OAc


SF1 463 26 73

LF1 88 nd 17

LF1 oligo









656 324 22 46 na na

627 270 14 39 na na

16 60 10 7 06 12

13 58 2 6

636 317 24 46

605 299 20 42 na na

343 71 4 11 na na

188 214 14 28 12 14

175 204 3 27

516 267 20 39

314 155 10 22 na na

316 62 7 7 na na

1 85 5 nd 11 18

bdl 83 5 nd

316 125 12 7










Lignincontent

Ligninrecovery

pH

[C4mim]Cl SF2 12 31 04 (03) 61



[C2mim]Cl SF2 108 30 32 (15) 36



[C2mim]OAc SF2 582 23 134 (99) 70

[C2mim]OAc SF3 122 26 32 (17) 10

[C2mim])OAc TOTAL 704 166 (117)








Band position(cm-1)

Assignment











[17-20]






PO

LI

Wavenumber





LYSACCHARIDES

GNIN

s (cm-1)















Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References







[C4mim]Cl


SF1 1318 47 344

LF1 92 nd 04

LF1 oligo


[C2mim]Cl


SF1 643 39 185

LF1 446 nd 3

LF1 oligo


[C2mim]OAc


SF1 463 26 73

LF1 88 nd 17

LF1 oligo









656 324 22 46 na na

627 270 14 39 na na

16 60 10 7 06 12

13 58 2 6

636 317 24 46

605 299 20 42 na na

343 71 4 11 na na

188 214 14 28 12 14

175 204 3 27

516 267 20 39

314 155 10 22 na na

316 62 7 7 na na

1 85 5 nd 11 18

bdl 83 5 nd

316 125 12 7










Lignincontent

Ligninrecovery

pH

[C4mim]Cl SF2 12 31 04 (03) 61



[C2mim]Cl SF2 108 30 32 (15) 36



[C2mim]OAc SF2 582 23 134 (99) 70

[C2mim]OAc SF3 122 26 32 (17) 10

[C2mim])OAc TOTAL 704 166 (117)








Band position(cm-1)

Assignment











[17-20]






PO

LI

Wavenumber





LYSACCHARIDES

GNIN

s (cm-1)















Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References








Lignincontent

Ligninrecovery

pH

[C4mim]Cl SF2 12 31 04 (03) 61



[C2mim]Cl SF2 108 30 32 (15) 36



[C2mim]OAc SF2 582 23 134 (99) 70

[C2mim]OAc SF3 122 26 32 (17) 10

[C2mim])OAc TOTAL 704 166 (117)








Band position(cm-1)

Assignment











[17-20]






PO

LI

Wavenumber





LYSACCHARIDES

GNIN

s (cm-1)















Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References


Band position(cm-1)

Assignment











[17-20]






PO

LI

Wavenumber





LYSACCHARIDES

GNIN

s (cm-1)















Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References














Wavenu





mbers (cm-1)





































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References




































































Abstract

Background

Results

Conclusions

Background






ATR-FTIR


Lignin recovery


Conclusions

Methods

Bagasse

Chemicals








ATR-FTIR



Recovery of IL

Abbreviation

Competing interests


Acknowledgements

Author details

References

sugarcane bagasse pretreatment using three imidazolium-based

Documents