POLYMERIZATION OF POLYESTERS IN CARBON DIOXIDE: STUDY OF DEGRADATION Mzrro, R., Sánchez, L. A., de Lucs, A., Grci, I.*, Rodríguez, J. F. Deprtment of Chemicl Engineering. University of Cstill-L Mnch. Fcultd de C. C. Químics. Avd. Cmilo José Cel 12, 13071 Ciudd Rel, Spin. igncio.grci@uclm.es. Fx: +34926295256 INTRODUCTION Crbon dioxide hs been recently used for pplictions in polymer processing. There re severl exmples in the literture regrding polylctide (PLA) nd polyglycolide (PGA), most of them re relted to polymer composites, microcellulr foming nd prticle production [1-3]. The min dvntge of polymer processing in CO 2 is the considerble reduction in the viscosity of molten polymer [3]. The temperture of the polymer processing process depends on severl prmeters, such s crystllinity nd therml properties of the polymer, but it is usully bove glss trnsition temperture (Tg) or melting point (Tm), for morphous or semicrystlline polymers, respectively. However, in CO 2 the temperture of the process cn be lower thn in the conventionl processing processes becuse of the plsticiztion of the polymers [4,5]. An dditionl dvntge in polymer processing in crbon dioxide could be the in situ polymeriztion due to the elimintion of using orgnic solvents for the polymer production. Nevertheless, up to now, there re only few exmples in the literture, nd even less bout the bovementioned polymers [6,7]. On the other hnd, degrdtion of poly(l,llctide), PLLA, ws observed t tempertures close to its melting point. According to previous studies the therml degrdtion of polyesters follows complex process which is the result of the contributions of different pthwys, such s (1) hydrolysis by trce mounts of wter; (2) zipper-like depolymeriztion; (3) oxidtive, rndom min-chin scission; (4) intermoleculr nd intrmoleculr trnsesterifictions [8]. However, the experimentl results showed tht the min contribution to therml degrdtion of PLA goes through the trnsesterifiction rections, especilly the intrmoleculr ones which produce cyclic oligomers [9]; moreover, the presence of residul trces of ctlyst ccelertes the degrdtion by the intrmoleculr trnsesterifiction even t tempertures s low s 160ºC [8]. More recently, two kinds of intrmoleculr trnsesterifiction rections hve been pointed out s effective for the degrdtion process, the first one, clled rndom chin scission is opertive t low tempertures, wheres the second one (specific chin end scission) tkes plce t higher tempertures [10]. Similr results hve lso been observed for the therml degrdtion of PGA [11]. Aprt from differences in the composition of the degrdtion products, nd in the ctivtion energy of the processes, the rndom chin scission nd specific chin end scission rections cn be distinguished for their opposite effect on both moleculr weight nd mss loss. Wheres the rndom chin scission produces high reduction in the moleculr weight of the smple without prcticlly ny mss loss [10,12]; the specific chin end scission does not ffect the number verge moleculr weight (Mn), i.e. the mole number of polymer chins, but it highly increses the mss loss of the smple becuse of the production of low moleculr weight species [10]. Due to the different therml properties of poly(d,l-lctide) (PDLLA), poly(d,d-lctide) (PDLA), PLLA nd PGA, the tempertures of degrdtion re lso different. So, Gogolewski et l. [12] reported significnt reduction in moleculr weight fter injection molding for PDLLA, PDLA nd PLLA t 148, 174 nd 188ºC, respectively. Jmshidi et l. [8] observed this reduction in Mn for PLLA bove tempertures of 190ºC, but mss loss of purified polymer, i.e. without ctlyst, ws not observed below 220ºC. On the other hnd, for PGA significnt mss loss ws observed bove 270ºC [11]. In generl, the results relted to mss loss gree quite well with dt given by Sivlingm nd Mdrs [10]
regrding PDLLA nd PGA. The im of this work is to nlyze the therml degrdtion of PGA t low temperture nd to study the effect of CO 2 on this process. MATERIALS AND METHODS Mterils. The mterils used were glycolide (1,4-dioxne-2,5-dione; Purc Biochem bv, The Netherlnds) with purity higher thn 99.5%. Zinc octote (ZnOct 2, zinc (II) 2-ethylhexnote; Nus, Spin) with metl content of 12%. %. Methnol nhydre (MeOH) (SDS S.A., Spin) with purity higher thn 99.85%. And crbon dioxide (Crburos Metálicos, S.A., Spin) with purity of 99.8%. All products were used s received. Procedures. Experiments in supercriticl medi were crried out in the lb-scle unit described in previous pper [13]. The procedure followed ws: the mounts of monomer, ctlyst nd inititor were directly chrged in the rector t room temperture. The recipe followed ws: 5 g of monomer; the molr rtio of monomer to ctlyst ws 100; the molr rtio of inititor to ctlyst ws 0.5; nd the rection temperture ws fixed t 160ºC. Rector ws seled nd filled with 100 g of CO 2. After, the stirring (1000 rpm) nd the heting strted. The zero time ws considered when temperture reched 80ºC. Then, the rection ws run for ech rection time. To quench the rection, the rector ws cooled ner to room temperture nd lter it ws depressurized nd open, nd the smple ws tken for chrcteriztion. Bulk rections were crried out in stirred-tnk rector with controlled temperture nd constnt high enough nitrogen flow in order to remove ir nd humidity from the system. The stirring ws initilly set t 100 rpm but due to the increse in the viscosity of the system, it ws necessry to stop it. Firstly, 100 g of monomer ws dded nd heted until 160ºC. After complete therml equilibrtion, the sme molr rtios of monomer to ctlyst nd inititor to ctlyst s in supercriticl conditions were injected into the rector. Severl smples were tking during the rection. TGA (Therml Grvimetric Anlysis). In order to determine the conversion, the thermogrvimetric nlysis ws used. A TGA (TA Instruments, Q600) with refrigerted cooling system nd purged with nitrogen gs. DSC (Differentil Scnning Clorimetry). A DSC (TA Instruments Q1000) equipped with refrigerted cooling system nd utosmpler ws used for the mesurement of the fusion enthlpy, glss trnsition temperture nd percent of crystllinity. The smple (4-10 mg) ws heted until 250ºC followed by cooling until - 50ºC nd, finlly, it ws heted until 250ºC gin. Heting nd cooling were crried out with heting rte of 10ºC min -1. MALDI-TOF mss spectrometry (Mtrix-ssisted lser desorption/ioniztion time of flight mss spectrometry). MALDI-TOF mss spectr were mesured by n Autoflex II TOF/TOF Bruker spectrometer (Bremen, Germny) using IAA s mtrix mteril. Smples cocrystllized with mtrix on the probe were ionized by nitrogen lser pulse (337 nm) nd ccelerted under 20 kv with time-delyed extrction before entering the time-of-flight mss spectrometer. Mtrix (10 mg ml -1 ) nd smple (1 mg ml -1 ) were crefully mixed s powder. Typiclly, 5 L mixture of mtrix nd smple ws pplied to MALDI-TOF MS probe nd ir- dried. All spectr were preformed in positive reflection mode. Externl clibrtion ws performed by using Peptide Clibrtion Stndrd II (covered mss rnge: ~700 D-3200 D), Protein Clibrtion Stndrd I (covered mss rnge: ~5000 D-17500 D) nd Protein Clibrtion Stndrd II (covered mss rnge: ~20000 D- 70000 D) from Cre (Bruker).
P / br RESULTS In order to study the influence of crbon dioxide on the degrdtion of PGA, two sets of experiments were crried out t 160 C in bulk nd in CO 2. Due to the impossibility of tking smples during the rection in CO 2, the sme rection ws crried out t different rection times. To check the reproducibility of the experiments in crbon dioxide, the profile of pressure versus time is plotted in Figure 1. A good greement mong ll the dt is observed wht mens tht the reproducibility of the experiments is fine. 250 200 150 100 50 0 Interpoltion 10 min 15 min 20 min 30 min 40 min 0 10 20 30 40 50 60 t / min Figure 1. Reproducibility study of homopolymeriztion glycolide in CO 2. Ctlyst: ZnOct 2. Inititor: methnol. T: 160ºC. N: 1000 rpm. m monomer : 5 g. Molr rtio monomer to ctlyst: 100. Molr rtio inititor to ctlyst: 0.5. m CO2 : 100 g. On the other hnd, the smples were nlysed by MALDI-TOF to determine if the kinetic mechnism of the rection is the sme in both systems. Figure 2 shows the MALDI-TOF spectr of two smples, one from the bulk polymeriztion nd the other from the polymeriztion in crbon dioxide. The sme two min fmilies re observed in both cses, which correspond to the species produced by initition of wter nd methnol, nd b, respectively, following the kinetic scheme previously described in the literture [14,15]. Nevertheless, the proportion between them is little bit different, being less importnt the contribution of species produced by methnol in the bulk polymeriztion. This could indicte loss of methnol during the bulk polymeriztion since the system is open. Moreover, prcticlly ny cycle oligomer is observed even for longer rection times.
Intensity /.u. : HO - [C 2 H 2 O 2 ] n - H b: CH 3 O - [C 2 H 2 O 2 ] n - H b b CO 2 10 min Bulk 6 min 800 900 1000 1100 1200 1300 1400 1500 1600 m / z Figure 2. MALDI-TOF spectr of two smples of PGA. Up from the polymeriztion in crbon dioxide nd down from the bulk polymeriztion process. In second step, the evolution of the rection is nlysed by chrcteriztion of the products with DSC. The conversion ws estimted from TGA nlyses. The results for both sets of experiments re shown in Tbles 1 nd 2. The moleculr weight of ech smple hs been clculted from the Flory-Fox expression (eq. (1)) whose prmeters for PGA hve recently published by Gutier et l. [16]: T g T g K M n (1) Where T g is the glss trnsition temperture for polymer with n infinitive moleculr weight, nd K is constnt, the vlues for PGA re 44.8ºC nd 1.10 10 5, respectively. Tble 1. Homopolymeriztion glycolide in CO 2. Ctlyst: ZnOct 2. Inititor: methnol. T: 160ºC. N: 1000 rpm. m monomer : 5 g. Molr rtio monomer to ctlyst: 100. Molr rtio inititor to ctlyst: 0.5. m CO2 : 100 g. t (min) X (-) T g (ºC) M n (D) ΔH f (J g -1 ) Xc (-) 10 0.84 40.47 24390 57.20 0.27 15 0.86 36.58 13382 50.91 0.24 20 0.87 10.05 3165 41.41 0.20 30 0.86 1.50 2540 57.16 0.27 40 0.85-17.90 1754 44.99 0.21
On the other hnd, the cristllinity percentge cn be clculted using the fusion enthlpy s following: Hf Xc (2) Hc PGA Where Hc PGA is the melting enthlpy of pure crystl of polyglycolide, 206,6 J/g [17]. Tble 2. Homopolymeriztion glycolide in bulk. Ctlyst: ZnOct 2. Inititor: methnol. T: 160ºC. m monomer : 100 g. Molr rtio monomer to ctlyst: 100. Molr rtio inititor to ctlyst: 0.5. t (min) X (-) T g (ºC) M n (D) ΔH f (J g -1 ) Xc (-) 4.5 0.91 21.73 4768 118.3 0.57 6 0.97 39.36 20221 108.9 0.53 16 0.96 40.8 27574 101.3 0.49 26 0.97 11.8 3333 120.1 0.58 36 0.95 19.4 4324 99.83 0.48 66 0.96-12.7 1914 125.7 0.61 In both cses it is observed tht fter reching the mximum conversion, the moleculr weight decreses long with the time increse, wht indictes tht the degrdtion process tkes plce even t such low temperture compred to the melting point of PGA (round 220 C) nd it is minly through rndom chin scission process, wht it supports by the fct tht the mss loss observed fter the experiment completion cn be considered neglected in both cses [10,12]. However, the Mn decrese is fster in crbon dioxide thn in bulk. As result, to chieve reduction of 90% in the moleculr weight, only 40 minutes re required in CO 2 ; while to get the sme percentge of reduction in the bulk polymeriztion more thn one hour is required. Consequently, it seems tht crbon dioxide enhnces the degrdtion process, most probbly becuse the plsticizing effect of CO 2 on the polymer which cn enhnce the mobility of the chins nd ccelertes the trnsesterifiction process. On the other hnd, the intrmoleculr trnsesterifiction seems not to be very importnt since prcticlly ny cycle ws observed in MALDI-TOF spectr for longer rection times. It is lso observed tht while the vlues of crystlinity of the smples from the bulk rection re quite similr to the typicl vlue reported in literture [17], round 50%, they re considerbly lower in the cse of supercriticl rections. Wht mens tht the cooling in supercriticl rection is fster thn in the bulk process, which is probbly due to the Joule-Thomson effect during the depressuriztion process. In this wy, the cooling time is shorter thn the crystllition one nd the grde of crystllinity chieved is lower thn the one expected. Finlly, pointing out tht the conversion in bulk polymeriztion could be overestimted becuse of the glycolide loss during the rection since the sublimtion of glycolide is fct.
CONCLUSION The therml degrdtion of polyglycolide t low temperture hs been studied; it hs been demonstrted tht rndom chin scission process hs the lrger influence on the process. Moreover, the presence of the residul ctlyst cn contribute to ccelerte the process in polymeriztion in bulk nd in crbon dioxide; however, the process in CO 2 is even fster becuse of the plsticising effect of it on the polymer, which probbly enhnces the mss trnsfer nd the mobility of the chins. REFERENCES : [1] URBANCZYK, L.; NGOUNDJO, F.; ALEXANDRE, M.; JEROME, C.; DETREMBLEUR, C.; CALBERG, C. Europen Polymer Journl 45, (3), 2009, 643. [2] TSIVINTZELIS, I.; ANGELOPOULOU, A. G.; PANAYIOTOU, C. Polymer 48, (20), 2007, 5928. [3] HAO, J. Y.; WHITAKER, M. J.; WONG, B.; SERHATKULU, G.; SHAKESHEFF, K. M.; HOWDLE, S. M. Journl of Phrmceuticl Sciences 93, (4), 2004, 1083. [4] PINI, R.; STORTI, G.; MAZZOTTI, M.; TAI, H. Y.; SHAKESHEFF, K. M.; HOWDLE, S. M. Journl of Polymer Science Prt B-Polymer Physics 46, (5), 2008, 483. [5] LIU, D. H.; TOMASKO, D. L. Journl of Supercriticl Fluids 39, (3), 2007, 416. [6] BRATTON, D.; BROWN, M.; HOWDLE, S. M. Chem. Comm. 2004, 808. [7] GANAPATHY, H. S.; HWANG, H. S.; JEONG, Y. T.; LEE, W.-K.; LIM, K. T. Eur. Polym. J. 43, 2007, 119. [8] JAMSHIDI, K., HYON S. H., IKADA, H. Polymer 29, 1988, 2229. [9] KOPINKE, F. D., REMMLER, M., MACKENZIE, K., MILDER, M., WACHSEN, O. Polymer Degrdtion nd Stbility 43, 1996, 329. [10] SIVALINGAM, G., MADRAS, G. Polymer Degrdtion nd Stbility 84, 2004, 393. [11] MCNEILL, I. C., LEIPER, H. A. Polymer Degrdtion nd Stbility 12, 1985, 373. [12] GOGOLEWSKI, S., JOVANOVIC, M., PERREN, S. M., DILLON, J. G., HUGHES, M. K. Polymer Degrdtion nd Stbility 40, 1993, 313. [13] MAZARRO, R., DE LUCAS, A., GRACIA, I., RODRIGUEZ, J. F. J. Biomed. Mter. Res. Prt B: Appl. Biomter., 85B, 2008, 196. [14] KOWALSKI, A.; DUDA, A.; PENCZEK, S. Mcromolecules 33, 2000, 7359. [15] YU, Y., STORTI, S., MORBIDELLI, M. Mcromolecules 42, 2009, 8187. [16] GAUTIER, E.; FUERTES, P.; CASSAGNAU, P.; PASCAULT, J. P.; FLEURY, E. J. Polym. Sci. Prt A: Polym. Chem. 47, 2009, 1440. [17] WU, H. C.; SHEN, F.W.; HONG, X.; CHANG W. V.; WINET H. Biomterils 24, 2003, 3871.