Synthesis of polylactic acid by direct polycondensation under vacuum without catalysts聚乳酸直接缩聚

直接缩聚

ChemicalEngineeringJournal151(2009)342–350

ContentslistsavailableatScienceDirect

ChemicalEngineering

Journal

journalhomepage:http://wendang.chazidian.com/locate/ce

j

Synthesisofpolylacticacidbydirectpolycondensationundervacuumwithoutcatalysts,solventsandinitiators

FeerzetAchmad,KenjiYamane,ShiQuan,TakaoKokugan?

DepartmentofChemicalEngineering,TokyoUniversityofAgricultureandTechnology,2-24-16Naka-cho,Koganei-shi,Tokyo184-8588,Japan

articleinfoabstract

Inordertoreducetheproductioncostofpolylacticacid(PLA),l-lacticacidwaspolymerizedbydirectpolycondensation(DP)undervacuumwithoutcatalysts,solventsandinitiators.Experimentswerecon-ductedatpolymerizationtemperatures(Tp)of150–250?C.ThemaximumPLAmolecularweightobtainedwas90kDaat200?Cafter89hundervacuum.Above200?C,PLAisthermallydegradedbyspeci?cscis-sion.TheDPactivationenergy(Ea)wasalsoinvestigatedandwasfoundtobelargerthanthatrequiredbycatalyzedringopeningpolymerization(ROP).Furthermore,theEawashigherthanthedeactivationenergy(Ed)forbothDPandROP.PLAyieldswerehigheratlowerTp,whiletheyieldoflactideincreasedwithTp.ThetotalyieldofPLAandlactidewasapproximately52–75wt.%.TheuncatalyzedDPmethodmayhaveapplicationsinon-sitecellplantproduction,wherecompactnessisrequired,aswellassafeandsimpleoperatingprocedures.

©2009ElsevierB.V.Allrightsreserved.

Articlehistory:

Received27November2008

Receivedinrevisedform2April2009Accepted3April2009Keywords:

Polylacticacid

DirectpolycondensationOn-sitecellplant

1.Introduction

Recently,thedemandforplasticsmadefrombiomass(bio-plastics)hasincreasedbecausetheycanreduceenvironmentalcarbondioxideemissions,whichinducesglobalwarming,andtheirproductiondoesnotrequirefossilfuelresources,thusreducingtheenvironmentalloadrequiredtoproducetheseplastics.Ifbio-plasticsaregoingtobeabletoaddresstheissuesofglobalwarmingorfossilfueluse,theywillneedtobeusedinlargequantities.However,theuseofbio-plasticsremainslimitedbecausetheirpro-ductioncostsare3–10timeshigherthanthoseofconventionalplastics[1].

Polylacticacid(PLA)derivedfrombiomassisatypicalbio-plasticthatcanbemadeinathree-stepprocessoffermentation,separationandpolymerization.NatureWorksCo.(USA)hasproducedPLAfromrecombinantcornbymeansofanintensivelarge-scaleprocessthatreducesproductioncosts[2].

InanattempttofurtherreducethecostofproducingPLA,weproposedanon-sitecellplantmethodforallthreeprocesses.Forthefermentationprocess,l-lacticacidwasfermentedusingfreshcassavaroot(FCR)asthesubstrateandconcentrateofmagurowaste(CMW)andtofuliquidwaste(TLW)asthemedium,withoutsac-chari?cationandStreptococcusbovisasthebacterialstrain[1,3,4].ThepolymerizationprocessusedtosynthesizePLAwasdirectpoly-condensation(DP)undervacuumwithoutcatalysts,solventsand

initiators.Thisprocessful?llstheconditionsofcompactness,aswellassafeandsimpleoperationforanon-sitecellplantconcept.PLAistypicallysynthesizedbyringopeningpolymerization(ROP)oflactide,aprocessthatusescatalystssuchastin,zinc,alu-minumandlead,initiatorssuchasn-,sec-andtert-butyllithium,andsolventssuchasdiphenylether,tolueneandchloroform.Manyofthesecomponentsaretoxicor?ammable.Asaresult,thispro-cessrequirespuri?cationtoremovetheunwantedmaterialsfromtheproduct,whichrequirescomplexfacilities[5,6].

ThispaperreportsthesynthesisofPLAbyDPwithoutcatalysts,solventsandinitiatorsbyvaryingtheTpfrom150to250?Candthepressurefromatmospherepressuretovacuumfor96h,inaccor-dancewiththespeci?cationsforon-sitecellplants.ThepropertiesofthePLAsynthesizedbyDPwerecomparedwiththoseofPLAsynthesizedbyROP.2.Materialsandmethods2.1.Materials

l-Lacticacid(l-LA:specialgrade)wassuppliedbyWakoPureChemicalIndustries,Ltd.(Tokyo,Japan)withamonomerconcentra-tionof60wt.%,atotalacidityof90wt.%andadensityof1.224kg/L.Chloroform(WakoPureChemicalIndustries,Ltd.(Tokyo,Japan))wasusedastheeluentingelpermeationchromatography(GPC).ThepolystyrenestandardforGPCcalibrationandthelactidestan-dardforFTIRanalysesweresuppliedbySigma–AldrichChemicalCo.(USA).WeusedPLAproducedbyLeceaMitsuiChemicalsInc.(Japan)asastandard.

?Correspondingauthor.Tel.:+81423887062;fax:+81423887062.E-mailaddress:kokugant@cc.tuat.ac.jp(T.Kokugan).1385-8947/$–seefrontmatter©2009ElsevierB.V.Allrightsreserved.doi:10.1016/j.cej.2009.04.014

直接缩聚

F.Achmadetal./ChemicalEngineeringJournal151(2009)342–350343

Nomenclature[A]0initialconcentrationofthe–COOHor–OHgroup

[molL?1]

[A]concentrationofthe–COOHgroup[molL?1]

[B]concentrationofcyclicmonomerB,e.g.,lactide[molL?1]

[B]0initialconcentrationofcyclicmonomerB[molL?1][Cat]catalystconcentration[molL?1]

Eaactivationenergyofpolymerization[kJmol?1]

Eddeactivationenergyofthermaldegradation[kJmol?1]

k1?rstorderrateconstantforROPpolymerizationsoflactide[min?1]

k?1

?rstorderrateconstantforthermaldegradationinROPpolymerizationsoflactide[min?1]

kRappapparentrateconstantofpolymerizationforROP[Lmol?1min?1]

kR?app

apparentrateconstantofthermaldegradationforROP[min?1]

k2

secondorderrateconstantofpolymerizationforDP[Lmol?1min?1]

kD2

app

apparentpolymerizationrateconstantforDP[min?1]

kD2

?app

apparentrateconstantofthermaldegradationforDP[min?1]

k3thirdorderrateconstantofpolymerizationforDP[L2mol?2min?2]

Mnnumberaveragemolecularweight[g/mol]Mwweightaveragemolecularweight[g/mol]

ndegreeofpolymerizationinstepwisepolymeriza-tion

Pdegreepolymerizationinringopeningpolymeriza-tion

tTime[min]

TTemperature[?C],[K]

ωinitialmonomerconcentrationofl-LA[wt.%]??

Density[kg/L]

2.2.Apparatusandprocedures

Fig.1showstheexperimentalapparatususedforpolymer-izations.Polymerizationswereconductedina1Lfour-neckedseparable?askequippedwithamechanicalstirrerandare?uxcondenserconnectedtoaninlinecoldtrapandavacuumpump.Thermocoupleswereusedtomeasuretheinternalandexteriortemperaturesofthesolutioninthereactorandtheexteriorofthe?ask.Tocontroltheinternalandexteriortemperaturesofthereac-tor,mainandsub-heaterswereinstalledsurroundingthereactor.Thereactorpressurewasmeasuredbyatransducerandmonitoredbyacomputer.Thepressureinthereactorwasloweredincremen-tallyusingavacuumpump.

Twohundredgramsofl-LAwasaddedtothe?ask,andPLAwassynthesizedbydirectpolycondensation(DP)throughthreeoperations:distillation,oligomerizationandpolymerization.Thetemperatureandpressureduringthepolymerizationwerepre-ciselycontrolled.

First,duringthedistillationperiod,l-LAwasheatedtothedistil-lationtemperature(Td)of150?C,afterwhichthetemperaturewasmaintainedattheTdfor150minatatmospherepressure.Thiswasfollowedbytheoligomerizationperiod,whichlastedbetween150and300min.Thetemperatureandpressurewerethenbroughttothepolymerizationtemperature(Tp)andvacuum,respectively.The

pressurewasreducedincrementallyto10mmHgover?ve30minperiods.Thedistillatewasremovedfromthecondensertrapandmeasuredevery30min.Then,duringthepolymerizationperiod,thel-LAwaspolymerizedatvariouspolymerizationtemperatures(Tp)between150and250?Cat10mmHgfor96h.Samplesweretakenfromthereactorandthetrapat12hintervals.Attheendofthereaction,theproductswerepouredintoaPetridishandallowedtocooltoroomtemperature.2.3.Analyticalmethod

Theconcentrationsofthel-LAmonomerinthereactorandtrapwereanalyzedbyabiosensorequippedwithanautosam-pler(BF-5:OjiScienti?cInstruments,Osaka).Thetotalacidityinthel-LAsolutionwasanalyzedtodeterminetheoligomerconcen-tration,ascalculatedbyanacid–basetitrationmethod(JISK8726).Theamountofoligomerpresentwascalculatedfromthediffer-encebetweentheamountofmonomerandthetotalacidity.ThePLA’sweightaveragemolecularweight(Mw),numberaveragemolecularweight(Mn)andMw/Mnratioweredeterminedusingagelpermeationchromatography(GPC)systemequippedwithtwochromatographycolumns(ShodexK803andShodexK804)andaRIdetector.Chloroformwasusedastheeluentata?owrateof1mL/min,andthemolecularweightswerecalibratedtoapolystyrenestandardat40?C.ThefunctionalgroupsofPLA,lactideandl-LAwereanalyzedbyFourierTransformInfraredSpectroscopy(FTIR:JASCOFT/IR-4100ST).3.Basicconsiderations

3.1.Therateofstepwisepolymerization

a.Self-catalyzedpolymerization

Intheabsenceofanexogenousstrongacid,theacidmonomeractsasitsowncatalystintheesteri?cationreaction.Thecon-densationcanbeexpectedtobethirdorderoverallintheconcentrationofthe–OHand–COOH(orA)functionalgroups[7,8]:d[A]

dt

=?k3·[?OH]·[A]2,(1)

wherek3istherateconstantofthethirdorderreaction,and[–OH]and[A]aretheconcentrationsof–OHandAgroups,respectively.

Sincethereisone–OHgroupforeach–COOHgroup,theequa-tioncanbewrittenas:d[A]

dt

=?k3·[A]3.(2)

IntegrationofEq.(2)yields:[A]20[A]

2

?1=2[A]20·k3·t.

(3)

However,thedegreeofpolymerization(n)inastepwisepoly-merizationcanbeexpressedbyEq.(4)(assumingnomassislost):n=

[A]0

[A]

.(4)

Thus,Eq.(3)canbesimpli?edto:n2?1=2[A]20·k3·t.

(5)

Eq.(5)makesclearthattherelationshipbetween(n2?1)andtshouldbelinearifthepolymerizationfollowsthismodelofself-catalyzedpolymerization.

直接缩聚

344F.Achmadetal./ChemicalEngineeringJournal151(2009)

342–350

Fig.1.TheexperimentalapparatusforPLApolymerization.(1)Four-neckedseparable?ask,(2)main-heater,(3)thermocouple,(4)sub-heater,(5)condensatetrap,(6)condenser,(7)coldtrap,(8)controlvalve,(9)vacuumpump,(10)stirrer,(11)pressuregauge,(12)transducer,(13)ampli?er,(14)A/Dconverter,(15)temperatureindicatingcontroller(TIC),and(16)personalcomputer.

b.Externalcatalysisofpolymerization

Thestepwisepolymerizationinvolvinganexternalcatalystfol-lowsthesecondorderreaction[7,8]:d[A]

=?k2·[?OH]·[A],dt

(6)

Ontheotherhand,thecoef?cientk1isproportionaltothecon-centrationofthecatalystinROP,thatis:

R

k1=kapp·[Cat],

(14)

wherek2istherateconstantofthesecondorderreaction.Setting

[–OH]=[A]gives:d[A]

=?k2·[A]2.dt

IntegrationofEq.(7)gives:[A]0

?1=[A]0·k2·t.[A]

CombiningEq.(4)andEq.(8)yields:n?1=[A]0·k2·t.

(9)(8)(7)

RwherekappistheapparentrateconstantofpolymerizationforROP

and[Cat]isthecatalystconcentration.

Insertingthecoef?cientk1intoEq.(12)yields:

d[B]R

=?kapp·[B][Cat].dt

(15)

3.3.Therateofthepolymer’sthermaldegradationbythecatalystTherateofthepolymer’sthermaldegradationinthepresenceofacatalystisgivenbyEq.(16)[12]:dP

=?k?1·P,dt

(16)

Eq.(9)showsthattherelationshipbetween(n?1)andtshouldbelinearifthepolymerizationfollowsthismodelofexternalcatalysis.Wede?netheapparentpolymerizationrateconstantDk2asshowninEq.(10):

app

wherek?1istherateconstantforthermaldegradationandPisthe

degreeofpolymerization.

Thecoef?cientk?1isproportionaltothecatalystconcentration,thatis:

R

k?1=k?app·[Cat],

(17)

D

k2=k2·[A]0.app

(10)

SubstitutingEq.(10)intoEq.(9)yields:

Dn?1=k2·t.app

(11)

Rwherek?appistheapparentrateconstantofthermaldegradation

bythecatalyst.

Insertingthecoef?cientk?1intoEq.(16)yieldsEq.(18):

3.2.Therateofringopeningpolymerization

Theringopeningpolymerization(ROP)reactionisa?rstorderreaction[9–11],thus:d[B]

=?k1·[B],dt

(12)

dPR

=?k?app·P·[Cat].dt

(18)

Thus,therateofthethermaldegradationreactioninthepres-enceofacatalystisthesameastherateofROPpolymerization.4.Resultsanddiscussion

4.1.Impactoftemperatureandpressureontotalcondensateduringthedistillationandoligomerizationperiods

Fig.2showsthetemperatureandpressureinthereactor(a)andthetotalcondensateremoved(b)atvariouspolymerizationtemperatures(Tp)andpressuresover300mininthedistillationandoligomerizationperiods.Atabout45min,asecondorderphasetransitionoccurredat128±2?Cinthereactant(pointPinFig.2a),

where[B]istheconcentrationofcyclicmonomerandk1isthe?rstorderrateconstantforthereaction.WeobtainEq.(13)byintegrat-ingEq.(12):ln

[B]0

=k1·t,[B]

(13)

where[B]0istheinitialconcentrationofcyclicmonomerB.

直接缩聚

F.Achmadetal./ChemicalEngineeringJournal151(2009)342–350

345

Thecondensateonthewallsofthe?askandinthecondenserwaspreliminarilymeasuredtoweigh6.0g.Duringtheoligomerizationperiod(150min<t<300min),thetotalcondensateincreasedasaresultoftheincrementalreductioninpressure.Duringthepolymer-izationperiod,thetemperatureandpressurewerekeptconstantatTpand10mmHgfrom300minto96h.

4.2.Therateofpolymerizationandthermaldegradation

Fig.3showstheeffectofpolymerizationtime(t)onthedegreeofpolymerizationofthepolylacticacid(n)atvariouspolymerizationtemperatures(Tp)between150and250?C.Fortheself-catalyzedpolymerizationwiththirdorderkinetics,therelationshipbetween(n2?1)andtisshowninFig.3a,whiletheplotofthesec-ondorderreactionusinganexternalcatalystisshowninFig.3b.Theexperimentalresultssupportthemodelofanexternallycat-alyzedpolymerization,since(n?1)andthavealinearrelationship(Fig.3b)fromEq.(11).Thatis,thereactionmechanismforl-LApolymerizationbyuncatalyzedDP,isoverallsecondorder,becausel-LAisastrongacidthatactsasacatalystduringpolymerization,inspiteofthelargeamountofmasslostundertheseconditions.

DTheratecoef?cientk2inEq.(11)wasobtainedfromtheslope

app

Fig.2.Thetemperatureandpressureinthereactor(a)andtotalcondensateremoved

(b)duringthedistillationandoligomerizationperiods.Atfollowingpolymerizationtemperature150?C(??),170?C(??),180?C(??),190?C(?),200?C(??),210?C(?),220?C(??),230?C(??),250?C(*).

whereheatwasconstantlysupplieduntil150?Cwasreachedandthetotalacidityofthel-LAincreasedtoabout100wt.%from90wt.%bydistillation.Theamountsofwater,monomerandoligomerinthereactorandcondensatewereinagreementwiththediagrambyVickroy[13].Atabout90min,thereactortemperaturereachedthespeci?eddistillationtemperatureTdof150?C,whichwasmain-tainedthroughoutthedistillationperioduntil150minhadelapsed.Thetemperatureandpressureconditionsthroughoutthedistilla-tionperiodwerethesameforalltrials.

Thetotalcondensatewasmeasuredasthecondensateinthetrapandthecondensateadheringtothewallsofthe?askandcondenser.

ofthegraphinFig.3b,whilethecoef?cientk2wascalculatedfromEq.(10).

WealsocomparedthepolymerizationrateconstantforDPwiththatofringopeningpolymerization(ROP).Thepolymerizationrate

RisanforROPisrepresentedbyEq.(15),andthecoef?cientkapp

apparentsecondorderreactionrateconstant,inwhichthetermsthatappearintheratelawaretheconcentrationsoflactideand

Rcatalyst.Thepolymerizationrateconstantsk2andkapphavethe

?1?1samedimension[Lmolmin].

AtTp=200?C,thedegreeofPLApolymerization(n)reached1260at89handdecreasedafterthatpoint.Athightemperatures,PLAisinequilibriumwithradicallactidebyradicalhomolysispathways[14].Sincethemeltingpointoflactideis94–96?C,lactideisvapor-izedathightemperaturesandlowpressures,andisthentrappedbythecondenser.Asthermaldegradationtakesplaceandlactideisdistilled,PLAshiftstoalowermolecularweight.PLA’sequilibriumwithradicallactidedependsonthetemperature.Thermaldegrada-tionwasfoundtooccurat200?Cwhennocatalystispresent.Highertemperaturesleadtofasterandmoreprevalentthermaldegrada-tion.Forexample,atTp=250?C,thedegreeofpolymerizationisabout509after10h.Theboundarywherepolymerizationand

ther-

Fig.3.Theeffectofpolymerizationtime(t)onthedegreeofPLApolymerization(n).Symbols:150?C(??),170?C(??),180?C(??),190?C(?),200?C(??),210?C(?),220?C(??),230?C(??),250?C(*).

直接缩聚

346F.Achmadetal./ChemicalEngineeringJournal151(2009)

342–350

Fig.4.TheeffectsofpolymerizationtemperatureonthePLA’smolecularweightandMw/Mn.Symbols:Mw(??),Mn(??),ratioMw/Mn(??);atTp≤200?Cfort=89h,atTp>200?Cfort=maximummolecularweight.

maldegradationoccursimultaneouslyisshownbythedashedlineinFig.3b.

Thisstudyevaluatedtherateofthermaldegradationbyrep-resentingtheexperimentalreactionrateconstant(kD2?app

)fortheuncatalyzedthermaldegradationas:

n?1=kD

2

?app

·t+b.(19)

However,OkamotoandMatsuo[12]studiedthecatalyzedther-maldegradationofPLAandproposedtheratelawshowninEq.(16).Thecoef?cientk?1isthe?rstorderrateconstantforthermaldegradationinthepresenceofacatalyst.Thethermaldegradation

rateconstantskD2?app

andk?1havethesamedimensions[min?1].Fig.4showstheeffectsofpolymerizationtemperature(Tp)on

theweightaveragemolecularweight(Mw),thenumberaveragemolecularweight(Mn)andthePLA’sMw/Mnratio.ForTp<200?C,MwandMnarethemolecularweightsatt=89h,thetimewhenthePLAreacheditsmaximummolecularweight(90kDa)atTp=200?C(inFig.3b)andforTp>200?C,theMwandMnparametersarethemaximumvaluesallowedbythermaldegradation.

Harseetal.[15]studiedthesynthesisofPLAfroml-LAbyDPandfoundthatthePLAreachedamaximumMwof7kDausingSnCl2asthecatalystandaMwof2.35kDawithoutacatalystat180?Cafter24hundernitrogenatmosphere.Kaitianetal.[16]reportedamaximumMwof6kDaford,l-PLAsynthesizedbyuncatalyzedandsolvent-freeDPat200?Cfor24hundernitrogenatmosphere.LowMwpolymersresultedduetodif?cultiesinremovingwaterfromthebulkpolymerization.Ajiokaetal.[17]havereportedthepolymerizationofl-LAbyDPusingaSncatalystanddiphenyletherasthesolvent.HighMWPLA(240kDa)wasproducedafter24hat130?C.Inthepresenceofacatalystandsolvent,thesmallamountofwaterremaininginthereactionmixturecanberemovedazeotrop-icallyatarelativelylowtemperatureunderhighvacuum.However,suchaprocessrequiresmultiplereactors,complexfacilitiesandapuri?cationsteptoremovethecatalystandthesolventfromtheproduct[6].

CommerciallyavailablePLA,suchasthatproducedbyNatureWorks,issynthesizedfromlactidebyROPusingatinoctoatecatalystat180–210?Cfor5h.TheMwofthePLAproducedis111–145kDa[http://wendang.chazidian.comingthesamemethod,Ajiokaetal.[17]reportedthepolymerizationofl-LactidebyROPusingtinoctoateasthecatalystand1-dodecanolasthesolvent.ThisprocedureproducedhighMWPLA(430kDa)after5hat200?C.WhiletheMwofPLAobtainedbyourpresentconditionsislowerthanthat

obtained

Fig.5.Arrheniusplotforthepolymerizationandthermaldegradationofpolylacticacid.Symbols:Polymerization(??),andthermaldegradation(?)bypresentwork,Kaitian(),Eguiburu(),Tang(??),polymerization()andthermaldegradation(??)byOkamoto,HarsheusedTi(BuO)4(??)andSnCl2()ascatalysts,Ra?er().

byROP,ourpolymerizationconditionsaresuitableforon-siteproduction.

TheuncatalyzedpolycondensationprocessdidnotgiveahigherPLAmolecularweight.Becausethedehydrativepolycondensationofl-LAisdrivenbyself-catalysisfroml-LA’scarboxylgroup,theprogressofthecondensationdecreasesasthetotalamountofcar-boxylgroupinsituisreducedbythepolycondensationreaction[7,19].AnotherreasonforthegenerationoflowMwpolymersisthatthepolycondensationprocessiscarriedoutatmoderatetohightemperaturesandunderastrongvacuum,whichnotonlyallowsforafastreactionrate,butalsoaidsintheremovalofasmallamountofthereactants[7].

TheMw/MnratioforPLAproducedinuncatalyzedDPreac-tionsvariesfrom2.0to3.2.Ontheotherhand,thisratiois1.5–3.77forPLAproducedbyROP[10,20,21],1.4–5.5forcatalyzedDP[6,15,16,19,22,23]and2.16–2.65foruncatalyzedDPusingd,l-lacticacid[15,16].TheratioforDPwaslargerthanthatobtainedbyROP,aprocesswithahigherpolymerizationrateandalowerpolymerizationtemperaturethanthoseinthepresentstudy.ThebroadrangeintheMw/Mnratiooccurredbecauseofintramolec-ulartrans-esteri?cationofthepolymerchainsthatresultedintheformationoflactideandlowMwpolymers[16,22,24].Thesephe-nomenawerecausedbyhigherpolymerizationtemperaturesandlongerpolymerizationtimes.

4.3.Activationanddeactivationenergiesforthepolymerizationofpolylacticacid

Wecomparedtheactivationenergy(Ea)ofpolymerizationandthedeactivationenergy(Ed)ofthermaldegradationfortheDPmethodwiththoseoftheROPmethod.Fig.5showstheArrheniusplotforpolymerizationsandthermaldegradationsunderdirectpolycondensationandringopeningpolymerizationconditions.The

polymerizationrateconstantsk2andkRappwerecalculatedfromEq.

(10)fortheDPmethod(presentstudy)andfromEq.(14)fortheROPmethod,respectively.

[A]0inEq.(4)iscalculatedusingEq.(20):[A]ω×??0=

(20)

whereωisthemonomerconcentration,??isthedensity,andMistheinitialmolecularweightofl-LA.Propagationconstantsareplottedagainstthelefthandy-axisassolidlinesinFig.5.However,

thethermaldegradationrateconstantskD2?app

andk?1werecalcu-