Polylactide (PLA) - a new way of production.
Polylactide (PLA) is a polymer well known from applications in the biomedical field. It has been used for more than 20 years for surgical devices such as sutures or clips. In recent times these biomedical applications have been extended to controlled drug delivery systems and larger parts, such as screws for fracture fixation. In the biomedical field PLA is highly accepted because of its good mechanical properties combined with its biocompatibility and its ability to degrade both in vivo and in vitro (1-3).
This paper aims to introduce this aliphatic polyester as a polymer able to find its place among traditional commodity polymers. For PLA many applications in the agricultural sector as well as in the packaging field (the food and non-food sector) are open (4-6). These new applications for PLA are only thinkable because of recent cost reductions in the production of lactide as the basic monomer and because of the change in the polymerization technique from batch polymerization to continuous polymerization, using reactive extrusion. For this reactive extrusion process an innovative processing concept is presented, using a closely intermeshing co-rotating twin screw extruder.
In Fig. 1 the life cycle of polylactide (PLA) is shown. As basic materials for PLA production, nearly all agriculture raw materials, such as sugar or corn, but also waste materials from the agricultural production, can be used. These basic materials will be transferred by means of a bacterial fermentation process into lactic acid, the basic chemical needed for PLA production. This lactic acid can then be transferred into lactide, the cyclic dimer of lactic acid, by means of a combined process of oligomerization and cyclization. Due to the stereoisometric nature of lactic acid (L- and D-lactic acid), three different types of lactide can be generated (L, L-, D, D-, and meso-lactide). For the actual polymerization of lactide into polylactide this generated crude lactide has to be purified with regard to residual water content and free acidity. A water content smaller than 50 ppm and a free acidity smaller than 0.1% meq/kg is advisable. This purified lactide can be polymerized in a ring-opening polymerization into polylactide, an aliphatic polyester. Depending on the type of lactide used, polymers with different properties will be generated. A polymer made from pure L, L- or D, D- lactide will be of a semi-crystalline nature, while mixtures of both or a small content of meso-lactide together with L, L- or D, D-lactide as a main component will result in amorphous polymers. These polymers can be processed like all other thermoplastic polymers with extrusion, injection molding, blow molding, or fiber spinning processes into various products. The products can be recycled after use either by remelting and processing the material a second time, or they can be hydrolyzed into lactic acid, the basic chemical. This lactic acid can be re-introduced into the production process of PLA. The last possibility is to compost the polylactide to introduce it into the natural life cycle of all biomass, where it degrades into C[O.sub.2] and water. Here the diversity of PLA becomes obvious. It can be recycled following the traditional ways, composted like all other organic matter, and it will do no harm if burned in an incineration plant or introduced into a classical waste management system.
Comparison of Polylactide and Poly(Lactic Acid)
For both polylactide and poly(lactic acid) the abbreviation PLA is used. In Fig. 2 the difference of these two polymers is shown. Indeed, polylactide and poly (lactic acid) are the same chemical products, though they differ from each other in means of production. Poly(lactic acid) is generated using a polycondensation reaction starting from lactic acid. The main drawback of this polycondensation reaction is that it results in polymers of low molecular weight (7). Water produced during the polymerization has to be removed during the polymerization process, and soon an equilibrium between polymerization and the depolymerization reaction is reached (8). Furthermore, this polymerization requires long reaction times in combination with high temperatures. In contrast, polylactide is produced in a ring-opening polymerization, starting from the cyclic dimer of lactic acid. For this process the additional step of dimerization of the lactic acid is necessary which, as a disadvantage, increases the overall cost of the polymerization. On the other hand, the ring-opening polymerization is easier to control and polymers with a higher molecular weight are generated.
The ring-opening polymerization of lactide is usually catalyzed by organo-metallic compounds [ILLUSTRATION FOR FIGURE 3 OMITTED], which have been first classified into anionic or cationic initiators (9). Nevertheless, various studies have shown that most metal derivatives initiate the chain reaction through covalent bonds (10). Polymerization is usually assumed to proceed through a coordination-insertion mechanism, the detail of which depends on the metal compound. From these organometallic compounds two major groups are of special interest.
The first is the group of Lewis-acid-catalysts, including metal-halides, oxides, and carboxylates. One typical catalyst of this group traditionally used in polylactide synthesis is tin octoate, i.e., Sn(II)di(ethyl-2-hexanoate). Even though the mechanism is not yet clearly established, it is widely accepted that the ring-opening polymerization is actually initiated from compounds containing hydroxyl groups such as water and alcohols, which are either present in the lactide feed or can be added by demand [ILLUSTRATION FOR FIGURE 4 OMITTED]. Though several organometallic compounds are known to allow for the synthesis of polyesters with high molecular weight, the control of the actual ring-opening polymerization remains a problem. The Lewis acid catalysts are not chemically bonded to the growing chain and can therefore activate more than one chain. Thus, the degree of polymerization is not directly controlled by the monomer-to-catalyst molar ratio. In addition, intermolecular and intramolecular transesterification reactions perturb the chain propagation, broaden the molecular weight distribution, and yield cyclic oligomers.
The second group of catalytic systems known to effectively promote the ring-opening polymerization of polylactide is the group of metal alkoxides, and here especially the Mg, Sn, Ti, Zr, Zn, and Al-alkoxides. These metal-alkoxides are, in contrast to the above-mentioned Lewis-acid catalysts, true initiators of the ring-opening polymerization [ILLUSTRATION FOR FIGURE 5 OMITTED]. This ring-opening polymerization occurs within two steps. At first a complex between initiator and monomer is formed, followed by a rearrangement of the covalent bonds. The monomer is included in between the metal-oxygen bond of the initiator, cleaving the acyloxygen bond of the cyclic monomer, so that the metal is incorporated with an alkoxide bond into the growing chain.
POLYLACTIDE SYNTHESIS - A NEW APPROACH
Comparison of Traditional and Innovative Polymerization Concepts
In Fig. 6 the traditional way of polylactide synthesis is compared with a new approach used for the reactive extrusion polymerization of polylactide. Conventionally, polylactide is mixed, as can be seen on the left side of Fig. 6, with tin octoate as a catalytic system and introduced into a batch reactor, where the polymerization occurs at reaction temperatures between 130 [degrees] C and 180 [degrees] C. The actual polymerization time (depending on the amount of catalyst added and the temperature conditions used) varies between two hours and up to two days. It has to be noted that the limitation in finalizing the polymerization is the time needed for the remaining monomer to diffuse through the already formed high viscous polymer in order to reach the reactive sites. The polymer obtained with such a process often has a low thermal stability in melt processing. No stabilizing agent is added and the catalytic system is still present in the polylactide. It not only catalyzes the polymerization reaction but also the corresponding depolymerization reaction. This PLA of low thermal stability is therefore dissolved in a suitable solvent (e.g., CH[Cl.sub.3]), purified, a stabilizing system is added, and it is recrystallized to receive the final PLA product. PLA produced with this concept is expensive because of the solvent recrystallization step, slow because of the low reactivity of the catalytic system used, and complicated because of its two-step process. Overall, it is discontinuous, with all the disadvantages known to be apparent in batch processes. The only known continuous process for PLA production is the process developed for Cargill Inc.
As an alternative, an innovative continuous polymerization process is presented. The lactide is mixed together with a tin octoate-based catalytic system and a suitable stabilizer system that does not prevent or hinder the polymerization is added. This mixture is introduced directly into a co-rotating, closely intermeshing twin screw extruder and is polymerized during a reactive extrusion process directly into the final PLA product within a residence time of about five minutes. This process is not only cheaper and faster compared with the ones originally used, but it is also a continuous process.
Continuous Polymerization by Reactive Extrusion
ln Fig. 7 the processing concept used for the continuous polymerization by reactive extrusion is presented. A co-rotating, closely intermeshing twin screw extruder with a screw diameter of 25 mm (Berstorff, type ZE 25) is used. The screw as well as the barrel can be built up in a modular way. The processing concept as well as the screw concept can be adapted to any processing problem, especially to reactive processes where totally and partially melt-filled sections have to be realized following each other, where energy has to be introduced at specific stages of the process into the material. The extruder used has an overall length of 48 D. With this, it is one of the longer extruders, traditionally used in the reactive modification of polymers. Therefore it can be adopted for ring-opening polymerization.
The mixture of lactide, catalytic system, and stabilizer is added using a gravimetric feeding system at the beginning of the screw (left side in [ILLUSTRATION FOR FIGURE 7 OMITTED]). Additional additives, stabilizer systems, colors, or plasticizers can be added at any stage of the process on purpose. The screw concept shown in the lower part of Fig. 7 is divided into three zones. The first zone of a length of around three barrel sections is used to melt the lactide and to heat the mixture to the reaction temperature of around 185 [degrees] C. For this purpose only right-handed screw elements are used. In this first zone no significant polymerization occurs.
The second zone, having a length of around 4.5 barrel sections, is divided into three major functional segments, all built up in a similar structure. Each of the segments is built with kneading elements used for dispersive mixing to introduce mechanical energy into the melt, followed by mixing elements that initiate distributive mixing. Only with these functional elements can it be assured that the reactive components reach each other within a short time. These segments are closed against the flow with either left-handed elements or blisters. Such elements guarantee fully filled channels within the kneading and mixing elements. The closing element is chosen the stronger the further down the screw the segment is located. First a left-handed kneading block, then a left-handed screw element combined with a left-handed kneading element, and last, a blister element are used. This increasing strength of the barrier elements is necessary as the viscosity increases, together with the reaction reaching higher conversion rates. As a consequence, the ability of the monomer to move and diffuse within the already generated high viscous polymer melt becomes smaller. Therefore a proper mixing of the material becomes more important to enable the reactive components to reach each other.
The whole system is concluded by a third zone of about 2.5 barrel sections, using right-handed screw elements only. At the beginning of this zone it is possible to install a devolatilization unit to remove the remaining low molecular weight components such as leftover monomer or side reaction products at a reduced pressure level. The main part of this zone is used to build up the pressure to overcome the die resistance. In this pumping zone the final part of polymerization will be achieved at increased pressure levels of up to 100 bars, until the thermodynamic equilibrium of the polymerization reaction is reached at around 96-99% monomer conversion, depending on the processing conditions used.
TYPICAL EXPERIMENTAL CONDITIONS AND MOLECULAR PARAMETERS OF PLLA PRODUCED BY REACTIVE EXTRUSION TECHNOLOGY
Brussels Biotech provided (L,L)-lactide, having a water content of [less than] 40 ppm and containing 0.2% remaining toluene. The remaining free acidity was determined to be 6.5 mequ/kg. The 2-ethylhexanoic acid tin(II) salt [Sn(Oct).sub.2] was purchased from Th. Goldschmidt and used without purification. A 0.15 molar solution of the tin octoate-based catalytic system was prepared by dilution in freshly dried toluene. Toluene was dried by refluxing over Ca[H.sub.2]. Ultranox 626, used as a stabilizing agent, was provided by GE Specialty Chemicals and used without further purification.
The monomer conversion was determined by FTIR, and tin residues were extracted by washing successively the organic layer once with an aqueous HCl solution (0.1M) and twice with deionized water. Part of this solution was evaporated to dryness and the solid residue (monomer + polymer) was analyzed by size exclusion chromatography (SEC). The second part was precipitated in cold methanol and the polymer was filtered off and dried under vacuum to a constant weight.
The L-LA conversion was calculated from the FTIR spectrum of cast film on NaCl. A calibration plot of the PL-LA to L-LA molar ratio versus the [A.sub.1383] to [A.sub.935] ratio was established, where [A.sub.1383] and [A.sub.935] were the absorption of the bands at 1383 and 935 [cm.sup.-1] respectively. The absorption band at 1383 [cm.sup.-1] corresponded to a vibration mode shared by the polymer and the monomer, while the absorption hand at 935 [cm.sup.-1] was characteristic of the monomer. Practically, monomer conversion (c) was calculated on the following equation:
c = 100/([L - LA]/[PL - LA] + 1) where
[PL - LA]/[L - LA] = - 12.27 + 31.56
([A.sub.1383]/[A.sub.935]) - 26.76 [([A.sub.1383]/[A.sub.935]).sup.2] + 10.43
[([A.sup.1383]/[A.sup.935).sup.3] - 1.81 ([A.sub.1383]/[A.sub.935]).sup.4] + 0.12 [([A.sub.1383]/[A.sub.935]).sup.5] (1)
Occasionally conversion was also calculated by 1H NMR from the relative intensity of the methine group of the monomer and the polymer ([Delta][CH.sub.PL-LA] = 5.16 ppm, [Delta][CH.sub.L-LA] = 5.02 ppm). The 1H and 13C NMR spectra were recorded in [CDCl.sub.3] with a Bruker AM400 apparatus at 25 [degrees] C. The solution concentration was 5 wt/v%. Quantitative analysis of the polylactide microstructure by 13C NMR required using the "INVGATE" sequence with a pulse width of 30 [degrees], an acquisition time of 0.7 s, and a delay of 3 s between pulses.
Size exclusion chromatography (SEC) was carried out in CH[Cl.sub.3] at 35 [degrees] C using a Waters 610 liquid chromatograph equipped with a Waters 410 refractometer index detector and two STYRAGEL columns (HR1, HR5E). The molecular weight and molecular weight distribution of polylactides were calculated in reference to a polystyrene calibration and corrected to an absolute basis using a universal calibration curve ([K.sub.PS) = 1.67 [degrees] [10.sup.-4], [a.sub.PS] = 0.692, [K.sub.PLA] = 1.05 [center dot] [10.sup.-3], [a.sub.PLA] = 0.563 in the [[Eta]] = K [center dot] [M.sup.a] Mark-Houwink equation).
POLYMERIZATION EXPERIMENT AND RESULTING PARAMETERS
Following the principal technology described before, the Lactide was mixed with the catalytic system using a monomer-tin molar ratio of 5000. The amount 0.5 wt% of Ultranox 626 was added as a stabilizing system to this mixture. The toluene was removed under reduced pressure and the monomer catalyst stabilizer mixture was added using a gravimetric feeding system into the previously described extruder system. A throughput rate of I kg/h was used. The screw speed was fixed to 100 rpm, which resulted in a melt temperature of 192 [degrees] C measured inside the polymer shortly before exiting through the die.The pressure in front of the die was determined to be 45 bars.
The resulting polymer has been analyzed by means of molecular weight and degree of conversion as described above. The conversion reached was 0.97 and the molecular weight distribution can be seen in Fig. 8. A PLA with a monomodal, very narrow molecular weight distribution was achieved. The number average molecular weight was [M.sub.n] = 91,000 with a polydispersity of 1.58.
POLYLACTIDE-MECHANICAL PROPERTIES TO REACH SUITABLE APPLICATIONS
Polylactide (PLA) is at room temperature a stiff and brittle polymer with a glass transition temperature of 55 [degrees] C and a melting temperature of 180 [degrees] C, depending on the amount of L- and D- lactide used in the original monomer composition (11-13). In Fig. 9 typical values of mechanical properties of such a PLA polymer are given, together with the stress-strain interrelationship. These properties include a modulus of elasticity of more than 3.500 MPa and a tensile strength of around 60 MPa, with only 3% of elongation at break. In Fig. 10 the mechanical properties of polylactide are compared with those of typical petrochemically based commodity polymers. As already said, PLA is a stiff and brittle polymer with comparable properties to polystyrene.
But polylactide can be modified by adding suitable plasticizers such as glucosemonoesters, polyethyleneglycols, or citrate esters to change the mechanical properties, similar to the way this is done in the traditional PVC industry (14-18). Using such plasticizer additions, it is possible to generate properties similar to those of a typical polypropylene. These results are shown for the two properties of elasticity modulus and tensile strength in Fig. 10, but they apply in a similar way for the other mechanical parameters. In the example a plasticized PLA is obtained, having an elasticity modulus of 1200 MPa and a tensile strength of 28 MPa.
Other Aliphatic Polyesters end Copolyesters
But lactide is only one of several cyclic monomers to be polymerized into aliphatic polyesters. Fig. 11 shows a list of them, including glycolide, caprolactone, and valerolactone. In principle all of these cyclic monomers or dimers can be polymerized in a similar process, adapting the processing conditions and the catalytic systems to generate a multitude of polyesters and copolyesters with the same multitude of mechanical and thermal properties (19, 20). As an example, the change in the mechanical properties using copolymerization of L-lactide with [Epsilon]-caprolactone is shown in Fig. 12. The tensile strength at break and the elongation at break of several compositions are shown, reaching from a stiff and brittle material with 0% caprolactone to a rubbery polymer with around 30 wt% of caprolactone.
Polylactide - Fields of Application
Finally, it is necessary to find and define the future fields of application for those aliphatic polyesters and copolyesters. This starts in the field of medical applications, where they are already strong, taking over more and more tasks in fracture fixation and drug delivery, being used for ever bigger and more complex systems. Another very interesting field should be the agriculture industry, using PLA for mulch film production and for clips in vineyards. Polylactides can alternatively be used in a wide range for the production of threads, ranging from woven fabrics for clothes to yarn and nonwovens. An additional application will be for packaging, ranging from films to blow-molded bottles for drinks and injection-molded parts for all packaging applications, food as well as nonfood.
In this paper the complete life cycle of PLA polymers has been reviewed, presenting by chemical and processing means a completely innovative type of polylactide synthesis using reactive extrusion techniques as a polymerization step. With this technique it is possible to generate aliphatic polyesters and copolyesters based on lactide at a competitive price level so that polylactide will be able to compete in the medium-term future with traditional, petrochemically derived polymers.
The authors are grateful to the European Union, DG 12 SSMA, for funding work in this field in the frame of the AIR program.
1. R. G. Sinclair, ANTEC '87, 1214 (1987).
2. H. R. Kricheldorf and I. Kreiser-Saunders, Macromol. Symp., 103, 85 (1996).
3. R. G. Sinclair, Proceedings of the first annual corn utilisation conference (June 1987).
4. R. G. Sinclair, J. M. S. -Pure Appl Chem., A33(5) 585 (1996).
5. R. Datta, S.-P. Tsai, P. Bonsignore, S. H. Moon, and J. R. Frank, FEMS Microbiological Reviews, 16, 221 (1995).
6. H.-G. Fritz, T. Seidenstucker, U. Bolz, M. Juza, J. Schroeter, and H.-J. Endres, Study on Production of thermoplastics and fibres based mainly on biological materials, EUR 16102 EN, Study of the European Commission (1994).
7. R. Miyoshi, N. Hashimoto, K. Koyanagi, Y. Sumihiro, and T. Sakai, Intern. Polymer Processing XI, 4, 320 (1996).
8. M. Ajioka, K. Enomoto, K. Suzuki, and A. Yamaguchi, Bull. Chem. Soc. Jpn., 68, 2125 (1995).
9. R. D. Lundberg and E. F. Cox, in Ring-Opening Polymerisation, 6, 266, K. C. Frisch and S. L. Reegen, eds., Marcel Dekker, New York, London (1969).
10. H. R. Kricheldorf and I. Kreiser-Saunders, Macromol. Symp., 32, 285 (1990).
11. C. Marega, A. Marigo, V. Di Noto, and R. Zannetti, Makromol Chem, 193 (7), 1599 (1992).
12. W. Hoogsteen, A. R. Postema, A. J. Pennings, and G. ten Brinke, Macromolecules, 23, 634 (1990).
13. C. A. P. Joziasse, H. Veenstra, D. W. Grijpma, and A. J. Pennings, Macromol. Chem. Phys., 197, 2219 (1996).
14. H. Tsuji and Y. Ikada, J. Appl. Polym. Sci., 66, 2367 (1996).
15. M. Sheth, V. Dave, R. A. Gross, and S. P. McCarthy, ANTEC '95, 1829 (1995).
16. C. E. Rehberg, M. B. Dixon, T. J. Dietz, and C. H. Fisher, Industrial and Engineering Chemistry, July 1950, p. 1409.
17. S. Jacobsen and H. G. Fritz, Polym. Eng. Sci., 38, 2799 (1996).
18. L. V. Labreque, R. A. Kumar, V. Dave, R. A. Gross, and S. P. McCarthy, J. Appl Polym. Sci., 66, 1507 (1997).
19. D. W. Grijpma and A. J. Pennings, Macromol. Chem Phys., 195, 1633 (1994).
20. A. Lofgren, A.-C. Albertsson, Ph. Dubois, and R. Jerome, J. Macromol. Sci., Rev., Macromol. Chem. Phys., C35(3), 379 (1995).
21. H. G. Fritz, S. Jacobsen, R. Jerome, Ph. Dubois, and Ph. Degee, German patent application, 196 28 472.4 (1996).
|Printer friendly Cite/link Email Feedback|
|Author:||Jacobsen, S.; Degee, Ph.; Fritz, H.G.; Dubois, Ph.; Jerome, R.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 1999|
|Previous Article:||Plasticizing polylactide - the effect of different plasticizers on the mechanical properties.|
|Next Article:||Processing and properties of high-temperature creep-resistant PVC plastisols.|