The Mechanism of the Ring-opening Polymerization of Lactide and Glycolide

-To obtain more insight in the ring-opening polymerization of lactones, complex formation of the initiators tetraphenyltin, stannous octoate, tin tetrachloride, aluminium bromide and triisobutylaluminium (TIBA), and the monomers L(-)-lactide, D,e-lactide and glycolide was studied by i.r.-spectroscopy. When equimolar benzene or toluene solutions of initiators and monomers were combined, only complexes of aluminium bromide and D,L-lactide or glycolide, and of TIBA and D,L-lactide or glycolide were observed. The complex formation was studied in detail by varying the initiator and monomer concentrations. From these results and theoretical considerations, it is concluded that complexes are formed by the coordination of a carbonyl oxygen of monomer and the aluminium atom of initiator. The corresponding polymers were formed when TIBA was used as an initiator. When AIBr 3 was used, polymers were obtained only when traces of water were added. It is concluded that in the case of AIBr~ the actual initiating species is HBr and that the polymerization initiated with TIBA proceeds very probably through a coordinated insertion of the lactone monomer into the aluminium-carbon bond. INTRODUCTION molecular 1:1 compound of both enantiomers Over the past 15 years there has been considerable L( -)-lactide and D( + )-lactide [8]. Lactide and glyinterest in the application ofpoly(glycolic acid) PGA, colide can be considered as modified lactones and poly(Land n,L-lactic acid) P L L A and PDLA and of their polymerization behaviour has been reported to co-polymers of glycolic and L-lactic acid GA/LLA as be rather similar to that of 3-valerolactone [9]. biodegradable polymers in medicine and surgery [l]. In several publications, mechanisms have been Absorbable sutures prepared from PGA [2] and from proposed for the initiation or, in a broader sense, for GA/LLA co-polymers (molar ratio 90:10) [3] are the polymerization of lactones. These mechanisms commercially available. In our laboratories we have can be subdivided into (i) cationic [5, 10-15); (ii) recently applied PDLA as temporary eardrum grafts anionic [16--20]; and (iii) coordination-insertion types in rats [4] and dogs. [21 27]. The latter type is considered by Young et al. Nowadays the preferred method for the prepara[28] as the intermediate case between the two other tion of high molecular weight PGA, P L L A and modes of initiation. In spite of the many data pubPDLA or corresponding co-polymers is the ringlished on the polymerization oflactones, especially of opening polymerization of the six-membered cyclic fl-propiolactone and E-caprolactone [29-31], the diesters glycolide (m.p. 87°), L-lactide (m.p. 95 °) and question whether a lactone ring is opened by D,L-lactide (m.p. 127°), respectively, initiated with a acyl-oxygen bond cleavage or by alkyl oxygen bond suitable initiator, e.g. tin tetrachloride [5], stannous cleavage is still disputed. Another as yet insufficiently octoate or tetraphenyltin [6]. The ring-opening polyanswered question is related to the true nature of the merization is schematically depicted in Scheme 1. The initiating species. polymerization is preferably performed in the melt With the exception of the results obta,ned by at higher temperatures, although lactide has also Kogan et al. [15], there are virtually no data on the been polymerized successfully in solution under mild interaction between the initiator and the lactone conditions [7]. monomer prior to the actual ring opening. InPurified D,L-lactide, derived from the racemic mixvestigation of this interaction is the subject of this ture of L( + )-lactic acid and n( -)-lactic acid, is the study. For that purpose the interaction between initiators and glycolide, L-lactide and D,L-lactide, *Present address: Firet by, P.O. Box 45, 3900 AA respectively, was studied by i.r.-spectroscopy. In the Veenendaal, The Netherlands. study by Kogan et al., only precipitates which could