PLA Synthesis. From the Monomer to the Polymer

KAZUNARI MASUTANI AND YOSHIHARU KIMURA

Kyoto Institute of Technology, Japan

1.1 Introduction

It is recorded that Théophile-Jules Pelouze first synthesized poly(lactic acid) (PLA) by polycondensation of lactic acid in 1845. In 1932, Wallace Hume Carothers et al. developed a method to polymerize lactide into PLA. This method was later patented by Du Pont in 1954. Until the late 1970s, PLA and its copolymers were developed as biomedical materials based on their bioabsorbable and biocompatible nature and have been utilized in many therapeutic and pharmaceutical applications such as drug delivery systems (DDS), protein encapsulation and delivery, development of microspheres and hydrogels etc. Recently, the biomedical application of PLA has been extended to tissue engineering including scaffold materials as well as to biocompatible materials for sutures and prostheses in which high- and low-molecular-weight PLAs are used, respectively. In the early 1990s, a breakthrough occurred in the production of PLA. Cargill Inc. succeeded in polymerizing high-molecular-weight poly(L-lactic acid) (PLLA) by ring-opening polymerization (ROP) of L-lactide in industrial scale and commercialized the PLLA polymer in the mid 1990s. Showing high mechanical properties in addition to a biodegradable nature, PLLA was thought to provide large opportunities to replace non-degradable oil-based polymers, such as poly(ethylene terephthalate) (PET) and polystyrene (PS). Since then, PLA has been utilized as biodegradable plastics for short-term use, such as rigid packaging containers, flexible packaging films, cold drink cups, cutlery, apparel and staple fibres, bottles, injection- and extrusion-moulds, coatings, and so on. All of them can be degraded under industrial compositing conditions. In the late 1990s, the bio-based nature of PLA was highlighted and its production as a bio-based polymer started. In this case, the newly developed polymers ought to have high-performances and long-life utilities that can compete with those of the ordinary engineering plastics. Various types of bio-based polymers are now under development, and several PLA types are also developed as promising alternatives to commercial commodities. In particular, PLLA polymers comprising high L-contents and stereo-complex PLA polymers showing high melting temperatures are now expected to be candidates for high-performance materials. The above historical view reveals the three specific features of PLA in terms of application, i.e. bio-absorbable, bio-degradable and bio-based.

Now, the synthesis of PLA polymers can be performed by direct polycondensation of lactic acid as well as by ring-opening polymerization of lactide (LA), a cyclic dimer of lactic acid. While the former method needs severe conditions to obtain a high-molecular-weight polymer (high temperature of 180–200 ?C, low pressure as low as 5 mmHg and long reaction times), the latter method can afford a high-molecular-weight PLA with narrow molecular weight distribution at relatively mild reaction conditions (low temperature of 130 ?C and short reaction times). Consequently, ROP of L-lactide is adopted in the ordinary industrial production of PLLA. On the other hand, since Ikada discovered the formation of stereo-complexes of PLLA and its enantiomer poly(D-lactic acid) (PDLA) in 1987, many trials have been done for its industrial production. Manufacturing of D-lactic acid and improvement of the stereo-complexibility of the enantiomeric segments have been the big challenges in the trials thus far. Synthesis of stereo-block polymers consisting of PLLA and PDLA macromolecular chains is a promising method for the preferential formation of stereo-complexes. This chapter deals with the whole synthetic aspects of these PLA polymers and their starting monomers.

1.2 Synthesis of Lactic Acids

1.2.1 Stereoisomers of Lactic Acid

Lactic acid (2-hydroxypropanoic acid) is the simplest 2-hydroxycarboxylic acid with a chiral carbon atom and exists in two optically active stereo-isomers, namely L and D enantiomers (S and R in absolute configuration, respectively), as shown in Scheme 1.1. These L- and D-lactic acids are generally synthesized by fermentation using suitable micro-organisms. Racemic DL-lactic acid (RS configuration) consisting of the equimolar mixture of D- and L-lactic acids shows characteristics different from those of the optically active ones. DL-lactic acid is conveniently synthesized by chemical method rather than fermentation.

1.2.2 Fermentation with Lactic Acid Bacteria

Lactic acid fermentation is one of the bacterial reactions long utilized by mankind along with alcoholic fermentation. The lactic acid bacteria are generally divided into several classes in terms of cell morphology, i.e. Lactobacillus, Streptococcus, Pediococcus, Aerococcus, Leuconostoc and Coryne species. They are also divided into various genera. Most of them produce L-lactic acid while some produce D- or DL-lactic acids. Table 1.1 compares which of D- or L-lactic acid is produced by different bacteria. The species belonging to the same Lactobacillus genus produce either L- or D-lactic acid preferentially. Lactobacillus helvetics and Sporolactobacillus produce DL- and D-lactic acids, respectively. In the lactic acid formation, therefore, stereoselectivity is much lower than in the amino acid formation where the absolute L-selectivity is shown. Table 1.2 shows the mono- and di-saccharides assimilated by the lactic acid bacteria. Each bacterium assimilates most mono-saccharides, but shows its own assimilation ability for di-saccharides. This difference in assimilation ability is important in the selection of bacteria. Since the breakdown of cellulose and starch often produces disaccharides, the species that can assimilate these di-saccharides must be used in the fermentation. In the ordinary lactic acid fermentation, the yields of L- and D-lactic acids reach 85–90% and 70–80% based on carbon usage, respectively.

1.2.3 Isolation and Purification of Lactic Acids

The fermenting liquor finally obtained in the above fermentation contains lactic acid together with various impurities such as un-reacted raw materials, cells and culture media-derived saccharides, amino acids, carboxylic acids, proteins and inorganic salts. Therefore, the isolation and purification steps are needed for obtaining a highly pure product needed in the polymer's synthesis. In the usual fermentation process, the generated lactic acid is neutralized in situ with calcium oxide or ammonia. When calcium oxide is used for the neutralization, calcium lactate is precipitated out.

This salt is isolated by filtration in the final step, washed with water and acidified with sulfuric acid to liberate free lactic acid with formation of calcium sulfate as solids. When ammonia is used for the neutralization, the ammonium lactate is formed and directly converted into butyl lactate by esterification with n-butanol, as shown in Scheme 1.2. Here, the ammonia is recovered and recycled. The following distillation and hydrolysis of butyl lactate gives an aqueous lactic acid with high efficiency. The lactic acid obtained by this method has higher purity than that obtained by the calcium salt method. The technologies for the above lactic acid fermentation and purification have well been established, and the production of both D- and L-lactic acids is conducted industrially in a plant scale of 100000 ton year-1.

1.2.4 Chemical Synthesis of Lactic Acids

Racemic DL-lactic acid can be synthesized by fermentation using appropriate bacteria (Lactobacillus helvetics in Table 1.1), but it is more easily synthesized by following the chemical process shown in Scheme 1.3. Here, the DL-lactic acid is produced by hydrolysis of lactonitrile that is generally formed by the addition reaction of acetaldehyde and hydrogen cyanide. Industrially, the lactonitrile is obtained as a by-product of acrylonitrile production (Sohio process). The lactic acid thus prepared is purified by distillation of its ester as described above.

1.3 Synthesis of Lactide Monomers

1.3.1 Stereoisomers of Lactides

Scheme 1.4 shows three lactides consisting of different stereoisomeric lactic acid units. L- and D-lactides consist of two L- and D-lactic acids, respectively, while meso-lactide consists of both D- and L-lactic acids. Racemic lactide (rac-lactide) is an equimolar mixture of D- and L-lactides. The melting points (Tm) of these lactides are compared in Table 1.3. Note that the Tm is higher in rac-lactide and is lower in meso-lactide.

1.3.2 Synthesis and Purification of Lactides

Each of the aforementioned lactides is usually synthesized by depolymerization of the corresponding oligo(lactic acid) (OLLA) obtained by polycondensation of relevant lactic acid, as shown in Scheme 1.5. Because of the ring-chain equilibrium between lactide and OLLA, unzipping depolymerization generates lactide through the back-biting mechanism involving the -OH terminals of OLLA as the active site as shown in Scheme 1.6. This reaction is well catalyzed by metal compounds involving Sn, Zn, Al and Sb ions, etc. The crude lactide can be purified by melt crystallization or ordinary recrystallization from solution.

1.4 Polymerization of Lactide Monomers

1.4.1 Structural Diversities of the Polylactides

As shown in Scheme 1.7, there are two major synthetic routes to PLA polymers: direct polycondensation of lactic acid and ring-opening polymerization (ROP) of lactide. Industrial production of PLA mostly depends on the latter route. The polymerization of optically pure L- and D-lactides gives isotactic homopolymers of PLLA and PDLA, respectively. Both PLLA and PDLA are crystalline, showing a Tm around 180 1C. Their crystallinity and Tm usually decrease with decreasing optical purity (OP) of the lactate units. Optically inactive poly(DL-lactide) (PDLLA), prepared from rac- and mesolactides, is an amorphous polymer, having an atactic sequence of D and L units. However, crystalline polymers can be obtained when the sequence of both D and L units are stereo-regularly controlled. The most interesting issue comes from the fact that mixing of isotactic PLLA and PDLA in 1 : 1 ratio affords stereo-complex crystals (sc-PLA) whose Tm is 50 ?C higher than that of PLLA or PDLA. This sc-PLA is formed by co-crystallization of the helical macromolecular chains having opposite senses. Stereo-block copolymers (sb-PLA) consisting of isotactic PLLA and PDLA sequences are also synthesized by stereo-regular polymerization techniques involving block copolymerization. These structural diversities of PLA polymers provide a broad range of physicochemical properties for PLA materials when processed.

1.4.2 Thermodynamics for the Polymerization of D- and L-Lactides

The heat capacities and enthalpies of combustion were measured to analyze the thermodynamics of polymerization of D- and L-lactides into their polymers. The enthalpies and entropies of the lactide polymerization determined from these data are as follows: ΔHp = -27.0 kJ mol-1 and ΔSp - -13.0 J mol-1 K-1 at 400 K, indicating an exothermic reaction. The kinetics of ROP of lactide have also been studied with various catalysts, showing that the polymerization rate is in first-order of each of the monomer and catalyst concentrations.

Witzke proposed a reversible kinetic model for the melt polymerization of L-lactide in the presence of tin(II) octoate as the catalyst and determined the following parameters: Ea = 70.9 [+ or -] 1.5 kJ mol-1, ΔHp = -23.3 [+ or -] 1.5 kJ mol-1 and ΔSp = -22.0 [+ or -] 3.2 J mol-1 K-1, and ceiling temperature (Tc) = 786 [+ or -] 87 ?C. Model equations for monomer concentration and conversion as a function of time were derived as follows:

Mt = Meq + (M0 - Meq) exp(-KpIt) (1)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

Kp = 86.0exp{(-Ea/R) (1/T - 0.00223)} (3)

where

Meq = (ΔHp/RT - ΔS/R)

Mt = monomer concentration at time

Meq = equilibrium monomer concentration

M0 = initial monomer concentration

Kp = propagation rate constant in (1/cat.mole %-hr)

I = catalyst concentration in (mole %)

t = time in hours

Xt = monomer to polymer conversion at time t

Ea = activation energy

R = gas constant

T = polymerization temperature in Kelvin

ΔH = enthalpy of polymerization

ITLΔS = entropy of polymerization

Three reaction mechanisms have been proposed thus far for ROP of lactide: anionic, cationic and coordination mechanisms. In the anionic polymerization, undesirable reactions such as racemization, back-biting reaction and other side reactions are often caused by the highly active anionic reactants that hinder the chain propagation. In the cationic polymerization, undesirable side reactions and racemization likely occur because of the nucleophilic attacks on the activated monomers and the propagating species. The decreases in molecular weight and optical purity lower the crystallinity and mechanical properties of the obtained polymers. On the contrary, coordination polymerization with metal catalysts (mostly alkoxides) can give a large molecular weight with the high optical purity maintained. Therefore, a variety of catalysts have been studied. The following sections deal with these polymerization aspects in detail.

1.4.3 Metal Catalysts

Metal complexes of Al, Mg, Zn, Ca, Sn, Fe, Y, Sm, Lu, Ti and Zr have been widely used as the catalysts in the ROP of various lactone monomers involving lactides. The standard catalyst system utilized for lactide polymerization is tin(II) octoate (stannous bis(2-ethylhexanoate): Sn(Oct)2), to which lauryl alcohol (1-dodecanol) is usually added as a real initiator. This catalyst system has many advantages over the other systems in that it is highly soluble in organic solvents and molten lactide in bulk state and very stable on storage. It also shows excellent catalytic activity to give high molecular weight of PLLA. The most important characteristic is that this catalyst is biologically safe and approved by the FDA (the US Food and Drug Administration) for use in medical and food applications, although the approval is dependent on empirical safety data. With these characters, Sn(Oct)2 has been used as the catalyst in the industrial production of PLAs.

The mechanism of this tin-catalyzed polymerization of lactide has been disputed for a long time, i.e. discussing whether the polymerization is induced by "insertion-coordination mechanism" or "monomer activation mechanism. Duda and Penzek proposed a comprehensive polymerization scheme based on the insertion-coordination mechanism. In the ordinary lactide polymerization catalyzed by Sn(Oct)2, a hydroxyl compound (alcohol) is added as the real initiator. The alcohol initiator first reacts with Sn(Oct)2 to generate a tin alkoxide bond by ligand exchange. In the next stage, one of the exocyclic carbonyl oxygen atoms of the lactide temporarily coordinates with the tin atom of the catalyst having the alkoxide form. This coordination enhances the nucleophilicity of the alkoxide part of the initiator as well as the electrophilicity of the lactide carbonyl group. In the next step the acyl-oxygen bond (between the carbonyl group and the endocyclic oxygen) of the lactide is broken, making the lactide chain opened to insert into the tin-oxygen bond (alkoxide) of the catalyst. The following propagation is induced by identical mechanism and continues as additional lactide molecules are inserted into the tin-oxygen bond (Scheme 1.8).

This mechanism was strongly supported by the MALDI-TOF mass spectrum showing molecular peaks that correspond to the oligomeric PLLA chains connecting with the tin residue, which are propagating species formed with the Sn(Oct)2/lauryl alcohol system. Since the polymerization is pseudo-living, the molecular weight can be relatively well controlled. However, in the last stage of propagation where the monomer concentration becomes significantly lower, the reverse depolymeriztion by back-biting mechanism as well as intermolecular trans-esterification that is referred to chain transfer or polymer interchange reaction becomes evident to broaden the molecular weight distribution. Despite the presence of this mechanism, the degrees of racemization and chain scrambling are much lower than those with anionic or cationic catalysis.