This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
A looped flow process was designed for the synthesis of well-defined multiblock copolymers using Reversible Addition-FragmentationchainTransfer(RAFT)polymerization.Thereactionconditions were optimized to reach high conversions whilstmaintainingahighend-group fidelity. The loop set-upproved tobe a flexible, robust and time-efficient process for scaling-upmutliblockcopolymers.
Industrial screening of block copolymer libraries has recentlyreceived increasing interest,1 therefore simpler processes forthe synthesis of highly-definedpolymer-libraries are requiredas demonstrated with the success of the automatedsynthesizerfortheone-potsynthesisofquasiblockcopolymersbyHavenandco-workers.2ReversibleAddition-Fragmentationchain Transfer (RAFT) polymerization is an extremely robustand versatile Reversible-Deactivation Radical Polymerization(RDRP)techniqueandmaybeusedtoreadilyobtainpolymersof defined molecular weight with narrow molecular weightdistributions3,4,5 for a range of monomer families.Furthermore, developments in RDRPmethods now allow forgreater control over the monomer distribution along thepolymer backbone6 which can have an interesting influenceover the physico-chemical properties of the resultingpolymers.7 Multiblock copolymers may now be prepared viaRAFT using a one-pot iterative addition process throughcareful optimization of polymerisation conditions to achievequantitativemonomerconversionwhilstmaintaininghighend-group fidelity.6,8 With such time and resource-efficiency,complex multiblock copolymer architectures are becomingincreasingly accessible and offer enormous potential forindustrialapplications.However,thescale-upofsolutionpolymerizationsisgenerallyundermined with issues concerning temperature control,9
hence affecting end-group fidelity and the properties of thepolymersobtained.10Emulsion polymerization generally allow for lower viscositiesand possess superior temperature control compared tosolution polymerizations, offering potential for scalability.11RAFTemulsionpolymerizationwasemployedbyEngelisetal.12to prepare well-defined polymethacrylate multiblockcopolymersonmulti-gramscales.13Polymerization in continuous flow is another alternative totraditional solution polymerization in batch. Tubular reactorsareadvantageousduetotheirexcellentheattransferandtheability to readilymaintain homogeneity, possessing potentialforup-scalingwithreproducibility.14,15Therefore, Russum16 and co-workers compared a mini-emulsion RAFT polymerization performed via a continuousflow process with a batch equivalent. However, higherpolydispersities were observed with reactions in a tubularreactor as compared to the batch process.17 This differencewas explained by the distribution of residence time of theparticles of the emulsion. Therefore, homogeneouspolymerizations might be more adapted than emulsionpolymerizations for continuous flow processes. Additionally,homogeneouspolymerizationsare far less limited in termsofpolymerizable monomers compared to heterogeneoussystems.12The first homogeneous continuous flow RAFT polymerizationwas performed by Diehl and co-workers18 to obtainpolyNIPAm.Theincreasedkineticsofflowreactionscomparedtothebatchprocesswasexplainedbyamoreuniformheating.Hornung et al.19 also used a continuous flow process tosynthesise diblock copolymers20 with solution-phase RAFTpolymerization, adapting the reaction conditions in order tokeep a low viscosity throughout the reactions. By connectingtworeactorsinseries,diblockcopolymerswereobtainedona2 gram scale without purification between blocks. However,there appeared to be reduced control of the polymerizationwith a lowmolecularweight tail observedon theGPC tracesafterchainextension.
Junkersandco-workersdemonstratedtheuseofhomogenousRAFT in continuous flow using a micro-reactor to prepare apoly(acrylate) multiblock copolymer (up to five blocks).21However, a final yield of 100 mg with a molecular weightdistributionof1.46andtheneedforawork-upbetweeneachchainextensionmakethismethodlessamenabletoscale-up.Loopedprocesseshavepreviouslybeenusedandoptimisedatan industrial scale for the production of polymers by freeradicalpolymerizationinaslurry-liquidphase22,23aswellasinemulsion.24 The set-up was chosen as it allows a higherpolymerconcentrationcompared toabatchprocess,25henceincreasingtheproductivity.However,thisprocesshasnotyetbeenappliedtoacontrolledradicalpolymerizationtothebestofourknowledge.Thepresentworkaimstoshowthattheloopset-upallowsthesynthesis of highly defined multiblock copolymers usingmaterials which are commercially available and of relativelylowcostsinordertocomplywithindustrialrequirements.Acrylamide monomers were chosen for the design of themultiblock copolymers since thesemonomers are associatedwithahighpropagationratecoefficient(kp)whichallowshortpolymerization times, andmean that low [I]0 can be used toachievefullconversion.26Theycanalsobepolymerizedathightemperatureastheyarenotsensitivetosidereactionsofchaintransfer. In addition many useful acrylamide monomers arewater soluble allowing for aqueous solution polymerization,whichenhancestheirkp,
27andimportantlypermitstheuseofthewater soluble azoinitiatorVA-044whichpossesses ahighdecomposition rate coefficient (kd)witha10hourhalf-lifeof44°C(inwater).Theuseofanazo-initiatorwithsuchahighkdresults in rapid generation of radicals at temperatures above60 °C allowing for short polymerization times. However, theoverallnumberofradicalsgeneratediskeptlowbylimitingtheconcentration of azo-initiator in solution in order to retain ahigh fraction of ω-functional chains throughout thepolymerization.26The choice of Chain Transfer Agent (CTA) is also a keyparameter for the design ofmultiblock copolymers since thecontrolofthe
polymerization is strongly dependent on the structure of theCTA. Trithiocarbonate RAFT agentswith a secondary R grouphave exhibit a good control over the polymerization ofacrylamide and acrylate monomers.6,8,26 In addition, theinduction period of the polymerization can be reduced bychoosing a secondary R group. Furthermore, 3-((((1-carboxyethyl)thio)carbonothioyl)thio)propanoicacid(BM1429)was selected because it is water soluble, hence compatiblewiththemostlyaqueoussystemwhichwasdesigned.Dioxanewas added (20% in volume) to themixture in order to keepthepolymersinsolution.Thesynthesisinbothflowandbatchprocesses have been optimised to achieve high monomerconversions, circumventing the need for purification stepsbetween each chain extension, limiting the work-up andsimplifyingtheprocess.Theset-upof theautomated loopreactorwasdesigned,asaproofofconcept,basedonastandardcontinuousflowset-upwithasinglereactorcoilof10mL.However,thereactionsmaybefurtherscaled-upbyincreasingthevolumeofthecoil.Themain feature of the system is the loop pump (which will bereferredtoaspumpB)whichrecirculatesthereactionmixtureinto the coil. As the volume contained in the reactor coil (10mL) ismuch higher as compared to the rest of the loop (1.5mL) and due to the high flow rate used (5 mL min-1), theresidencetimeoutsidethereactorcoilisnegligibleintermsofvariations in the temperature. All tubing used is made ofstainless steel to limit the permeation of air into the systemwhich would interfere with the polymerization. Another keyelementistheuseofthree-wayvalvestoisolatedefinedpartsof the circuit depending on the stage of the reaction. Thevarious options are illustrated in Fig. 1 with the relevantsectionsoftheset-uphighlightedthroughoutthestages.The operational procedure for the synthesis of multiblockcopolymers isdivided into threesteps.After filling theset-upwith solventprior toanyoperation, themonomer solution isloaded using pump A, with pump B being switched off asshown
onFig.1-1.Uponclosureof thevalvesCandD,pumpAandthewastecollectionsectionareisolatedfromtheloopsothereaction mixture can then be homogenised by letting thesolutioncirculatethroughtheset-upatahighflowrateusingpumpB,withpumpAturnedoff(Fig.1-2).Thepolymerizationwill accelerateas it reaches the targeted temperature.Whenthereactionreachescompletion,bothvalvesareopenedandthenextmonomerisintroducedasshownonFig.1-3,whilstasample is collected for analysis. The loop is closed after theloadingiscompletetoreturntotheset-updescribedbyFig.1-2 and the chain extension can take place. The last two stepsare repeated as many times as targeted number of blocksrequires. At the end of the reaction of the final block, bothvalvesareopenedandtheentireset-upisflushedwithsolventto collect the product through valveD. PumpAwas cleanedbetween each chain extension by purging with solvent toremoveanyresidualmonomerfromthepreviousblock.Theentireset-upwasmonitoredusingpressuresensorsandapressure relief valve was introduced as an outlet in casesystempressureexceededsafelevelsduetoanincreaseintheviscosityofthereactionmixtureortoablockage.Eachpolymerizationwasperformedat70°Cinordertogiveahigh rate of radical generation with VA-044, allowing thepolymerizationtimeforeachblocktobereducedto2h(≈95% VA-044 consumed under these conditions). A triblockcopolymerwithadegreeofpolymerization(DP)of20foreachblockandtwohexablockcopolymerswithblocksofDP10eachwere synthesizedusing twodifferent processes. The reactionconditions were firstly optimized for the loop set-up, thenadaptedforabatchprocess.Limitations in flow reactions are mainly associated withpotential blockages in the coil due to precipitation of thepolymer or as a result of excessive viscosity of the reaction
mixture, leading to a pressure build-up. Becausepolyacrylamides generally have a high glass transitiontemperature, characterized by a rapid increase in viscosityduring polymerization, the solvent system chosen as well asthe concentration and the temperature of the reactionmixture, is of particular importance in order to maintain arelatively low viscosity. The reaction conditions were thusoptimisedintheloopset-upinordertoachievehighpolymerconcentrationswhilstavoidingablockage.The loop set-up was optimized after several runs. The backpressureregulator(referredtoasFinFigure1)playedamajorrole in maintaining a pressure difference between the inletandtheoutletofpumpBwhentheset-upfunctionedasaloop(stepdescribedinFig.1-2).PumpBrequiresabackpressureinorder to operate and circulate the polymer solution throughthe systemwhilst the newmonomer solution is addedwhenthe viscosity is relatively low. Nevertheless, as thepolymerization(s) proceed, the viscosity of the solutionincreases and eventually the pressure applied by the backpressure regulator becomes toohigh for pumpB to operate.Whentheviscosityofthesolutionreachesthispoint,typicallyobserved from the third chain extension, the regulator isremovedsincethepressureofthesolutionishighenoughforpumpBtofunctionwithoutit.The loading of the monomer solution for each block was ofparticular interest as it could affect the control on the chainextension. Indeed, the final molecular weight distribution iscloselyrelatedtothehomogeneityofthecontentintheloop.For this reason, themonomer stock solutionwas loadedataflowrateof0.5mLmin-1whilstsimultaneouslycirculatingthepreviousblockat1.15mLmin-1.Indeed,themonomersolutionwouldnothavemixedwiththesolutionofmacroCTAifthe
latter had not been circulated during the injection of themonomer solution. The circulation within the loop afterinjection would have allowed the two solutions to mixeventually but free radical polymerization of the monomer
couldhaveoccurredpriortoanymixing.Potentialfreeradicalpolymerizationoftheadditionalmonomercanalsobelimitedby maintaining the temperature of the reactor coil at roomtemperature until the homogeneity of the mixture was
ensured by circulation of the polymer solution for 5minutesaftermonomerinjection.Firstly, an ABC triblock copolymer was synthesized in thelooped reactor, composedofpNAM20-b-pDMAm20-b-pDEAm20
(Fig. 2-1).According to the 1HNMRanalysis (Fig. S1-1), near-quantitative monomer conversion (>98 %) was achieved foreach block while SEC shows a clear shift towards highermolecular weight (maintaining narrow molecular weightdistributions) with each successful extensions (Fig. S2-1).However, the viscosity of the solutionbecame toohigh afterthe third block and further chain extensions could not becarried out in the tubular reactor. Therefore, shorter blockswere synthesized to attempt additional chain extensions.Subsequently, we chose to target a lower DP for each block(DP of 10) in order to conduct a higher number of chainextensions. By targeting blocks of DP 10 each, a hexablock(ABCABC) of pNAM10-b-pDMAm10-b-pDEAm10-b-pNAM10-b-pDMAm10-b-pDEAm10(Fig.2-2)wassuccessfullyobtained(Fig.3-1).Again, fullconversionwasobtainedforeachblock (>98%) as determined by 1H NMR (Fig. S3-1). This synthesis wascompleted in 12 hours, spread over 2 days, by producing 3blocks per day. The polymer solution was left at roomtemperature in the loop overnight without having the pumpoperating.Anotherhexablock(ABCABC)wassynthesizedusingdifferent monomers pNAM10-b-pHEAm10-b-NIPAm10-b-pNAM10-b-pHEAm10-b-pNIPAm10 (Fig. 2-3). The process wascompleted over 12 hours as well, within 3 days with thesynthesisof2blocksaday.Near-quantitativeconversionwasobtained with each polymerization (with the exception ofblock 3, 94 % by 1H NMR, Fig. S4-1). However, this may beconsidered acceptable given the low DP targeted. Chain
extensions could be carried out after leaving the macroCTAsolutionintheloopovernightinbothcases,demonstratingtherobustnessandtheimpermeabilitytoairoftheset-up,hencethepotentialtoconductmultiblockcopolymerssynthesisoverthecourseofseveraldaysifrequired.Similarly, the triblock and both hexablock copolymers weresynthesized in batch in order to compare the polymersobtained those from the loop system. The monomerconcentration could be increased as compared to equivalentreactioninflowsincetheviscositydoesnotaffecttheprocessas strongly. High conversionswere achieved as evidenced by1HNMR(Fig.S1-2,S3-2andS4-2),SECshowsclearshifttohighmolecularweight(Fig.3-2,S2-2andS5-2).Themaindifferencebetweentheloopandthebatchprocesseswas the scale atwhich the polymerizationswere performed:1.5 g of the triblock was yielded with the loop set-up,compared to 0.4 g obtained in batch. Similarly, the twohexablockswereobtainedatascaleofover3gramsusingflowandunderagramwithbatch.Interestingly,thescale-upofthesynthesisofmultiblockcopolymers in the loopset-updidnotaffectthedispersityofthepolymers.Indeed,accordingtotheSEC chromatograms (Fig. 3, S1andS2), themolecularweightdistribution of the three multiblock copolymers had a lowdispersityforbothprocesses(Đ≤1.12withflowandĐ≤1.18with batch, Table 1, S1 and S2). A high molecular weightshoulder was observed on the polymers obtained with thetubular reactor, as well as the hexablock with NAM, DMAmandDEAminbatch.Thiscouldbeattributedtosidereactionssuchaschaintransferorterminationbycombination.
Although thedispersitiesof thepolymersobtainedwithbothsystemsaresimilar,theaveragemolecularweightislowerforthosewhichweresynthesized inthe loopset-upaccordingtotheSECdata(Table1,S1andS2). Byimprovingtheprecision
of the flow rate atwhich themonomers are introduced, theaccuracyofthetargetedDPcouldbefurtheroptimized.Astheconcentrationsusedintheflowprocesswereoptimizedto be relatively high, a high livingness of the polymer chainscouldbemaintainedwithbothprocesses,whichiskeyforthe
synthesisofwell-definedmultiblockcopolymers(seeTableS3forpolymerconcentration).Usinga tubular reactorhelpedmaintaininggoodcontroloverthe temperature of the reaction mixture when scaling-upwhereas the temperature profile cannot be controlled asprecisely at large scale using round bottom flasks.9 The scalecan be further increased by choosing a reactor coil with alargervolumewhilstmaintainingthecharacteristicsofthefinalproduct. In addition, because the heat transfer is optimizedwithtubularreactors, theheatingandcoolingof thereactionmixture is more efficient than in batch,9 shortening thedurationoftheoverallpolymerizationprocess.Although tubular loop reactors have been explored for freeradical polymerization processes, their potential for thesynthesis of multiblock copolymers has been overlooked sofar.Thereareseveraladvantagesinusingaloopedsystemforthescale-upofmultiblockcopolymersoveracontinuousflowsystem. Primarily, there are fewer steps in the process sincethere is no loading andunloading of themacroCTA from thereactor coil needed between chain extensions. Equipmentcostsarealsokept lowsinceonlya single reactor is requiredinstead of several blocks in sequence. Furthermore, the
homogeneity of the reaction mixture was further improvedcompared to continuous flow polymerizations for which thehomogeneitystrictlyreliesonthesimultaneousloadingofthenew monomer solution with the solution containing thepreviousblock.All thechainextensionsweresuccessfulusingthe loop set-up, highlighting the difference in homogeneitywithpreviousblockcopolymersynthesis intubularreactors.20In addition, the looped system offers flexibility inpolymerization time, in contrast to continuous flowpolymerization, which is limited by a finite residence time(dictatedbycoil volumeand flow rate).As the stainless steeltubing is impermeabletoair, thereactionmixturecanbe leftovernight without affecting the polymerization, asdemonstratedbythesuccessfulchainextensions.Our work demonstrates a facile scale-up of the synthesis ofhighly defined multiblock copolymers using a loopedpolymerizationprocess.Theprocesswasshowntobeflexible,robustandtime-efficientwhilstmaintaininggoodcontroloverthe polymerization as attested by the low dispersity and thehighlivingnessoftheobtainedpolymers.
AcknowledgmentsThe authors wish to thank Michael Henry Steele for histechnical assistance. The CSIRO Julius Career Award (KL), theCSIROPhDscholarshipscheme(AK)andtheGermanResearchFoundation (DFG, GZ: HA 7725/1-1) (MH) are acknowledgedforfunding.
References
(1) A.-V.Ruzette;L.LeiblerNatMater2005,4,19. (2) J.J.Haven;C.Guerrero-Sanchez;D.J.Keddie;G.Moad;S.H.Thang;U.S.SchubertPolymerChemistry2014,5,5236. (3) J.Chiefari;Y.K.Chong;F.Ercole;J.Krstina;J.Jeffery;T.P.T.Le; R. T. A. Mayadunne; G. F. Meijs; C. L. Moad; G. Moad; E.Rizzardo;S.H.ThangMacromolecules1998,31,5559. (4) S.Perrier;P.Takolpuckdee JournalofPolymerSciencePartA:PolymerChemistry2005,43,5347. (5) G. Moad; E. Rizzardo; S. H. Thang Australian Journal ofChemistry2012,65,985. (6) G. Gody; T. Maschmeyer; P. B. Zetterlund; S. PerrierMacromolecules2014,47,3451.
(7) F. S. Bates;M. A. Hillmyer; T. P. Lodge; C.M. Bates; K. T.Delaney;G.H.FredricksonScience2012,336,434. (8) L. Martin; G. Gody; S. Perrier Polymer Chemistry 2015, 6,4875. (9) N. Micic; A. Young; J. Rosselgong; C. Hornung Processes2014,2,58. (10) T.Iwasaki;J.-i.YoshidaMacromolecules2005,38,1159. (11) S.W. Prescott;M. J. Ballard; E. Rizzardo; R. G. GilbertAustralianJournalofChemistry2002,55,415. (12) N. G. Engelis; A. Anastasaki; G. Nurumbetov; N. P.Truong;V.Nikolaou;A.Shegiwal;M.R.Whittaker;T.P.Davis;D.M.HaddletonNatChem2017,9,171. (13) G. Nurumbetov; N. Engelis; J. Godfrey; R. Hand; A.Anastasaki; A. Simula; V. Nikolaou; D. M. Haddleton PolymerChemistry2017,8,1084. (14) C. H. Hornung; A. Postma; S. Saubern; J. ChiefariMacromolecularReactionEngineering2012,6,246. (15) D.Wilms;J.Klos;H.FreyMacromolecularChemistryandPhysics2008,209,343. (16) J. P.Russum;C.W. Jones; F. J. SchorkMacromolecularRapidCommunications2004,25,1064. (17) J. P. Russum;C.W. Jones; F. J. Schork Ind. Eng. Chem.Res.2005,44,2484. (18) C. Diehl; P. Laurino; N. Azzouz; P. H. SeebergerMacromolecules2010,43,10311. (19) C. H. Hornung; C. Guerrero-Sanchez; M. Brasholz; S.Saubern; J. Chiefari; G. Moad; E. Rizzardo; S. H. Thang OrganicProcessResearch&Development2011,15,593. (20) C.H.Hornung;X.Nguyen;S.Kyi; J.Chiefari;S.SaubernAustralianJournalofChemistry2013,66,192. (21) J.Vandenbergh;T.deMoraesOgawa;T.JunkersJournalofPolymerSciencePartA:PolymerChemistry2013,51,2366. (22) M. A. Ferrero; M. G. Chiovetta Polymer-PlasticsTechnologyandEngineering1990,29,263. (23) J. D. Hottovy; F. C. Lawrence; B. W.Lowe; J. SFangmeier.;PhillipsPetroleumCompany:UnitedStatesofAmerica,1990. (24) C.Abad;J.C.DeLaCal;J.M.AsuaChemicalEngineeringScience1994,49,5025. (25) J.J.Zacca;W.H.RayChemicalEngineeringScience1993,48,3743. (26) G.Gody;T.Maschmeyer;P.B.Zetterlund;S.PerrierNatCommun2013,4. (27) A. Valdebenito; M. V. Encinas Polymer International2010,59,1246.