Elsevier

Polymer

Volume 41, Issue 25, 1 December 2000, Pages 8909-8919
Polymer

Epitaxial crystallization and crystalline polymorphism of polylactides

https://doi.org/10.1016/S0032-3861(00)00234-2Get rights and content

Abstract

Two crystal phases of poly(l-lactide) and that of the racemate of poly(l-lactide) and poly(d-lactide) can be grown epitaxially on one and the same crystalline substrate, hexamethylbenzene (HMB), which had been shown by Zwiers et al. [Polymer 1983;24:167] to form a eutectic with these polymers. The stable α-crystal modification of the optically active polymer, based on a 103 helix conformation (for PDLA; 107 for PLLA), is obtained for Tc near 155°C. A new crystal modification is produced by epitaxial crystallization at slightly lower Tc (≈140°C). The crystal structure of this new form is established by electron diffraction and packing energy analysis. Two antiparallel helices are packed in an orthorhombic unit-cell of parameters a=9.95Å,b=6.25Å and c=8.8Å. The racemate of poly(l-lactide) and poly(d-lactide) also crystallize epitaxially (at ≈165°C) on HMB, which appears to be a very versatile substrate.

Introduction

Polylactide, of formula (O–CO–C·H(CH3))n, is the polyester equivalent of the polypeptide, polyalanine (NH–CO–CH(CH3))n. Since a heteroatom is present in the main chain (which has a chemical polarity), it is a genuine chiral polymer, the two enantiomers of which have been synthesized. It is of particular interest as a biocompatible material since it is metabolized to non-toxic compounds.

The optically active form is known to exist in two crystal structures with different helix conformations and cell symmetries. The conformation of the chain in the α-phase was determined by De Santis and Kovacs [1] to be a left-handed 107 helix for the l-isomer (PLLA), and a right handed 103 helix for the d-isomer (PDLA). Two chains are included in an orthorhombic unit-cell of parameters a=1.06nm,b=0.610nm and c=2.88nm. The ratio of a and b parameters, 1.737, is nearly equal to √3, indicating an almost hexagonal packing of helices. Several more recent investigations report slightly different parameters for the unit cell, but the detailed crystal structure has not been determined; Hoogsten et al. [2] observed extra 00l reflections which suggest some deviation from a “pure” 107 or 103 helix conformation. The second phase, hereafter the β-phase, was first observed in 1982 by Eling et al. [3] when investigating stretched fibers of PLLA. Its crystal structure has not yet been solved, although several investigations indicate that it is based on a three-fold helical conformation with a c-axis parameter of 0.88 nm4 or 0.9 nm2. Hoogsten et al. [2] further suggest an orthorhombic unit cell (a=1.031nm,b=1.821nm) which houses six helices with, again, a near-hexagonal packing (the b/a ratio is 1.76, i.e. ≈√3); however, given the large size of the cell and number of helices, the detailed arrangement of the latter could not be worked out. More recently, Brizzolara et al. [5] made an extensive molecular modeling of the above structures and suggested, on the basis of data from Hoogsten et al., an orthorhombic unit cell with two parallel chains. As will be shown in the present and companion paper, the situation is more complex: there exist actually two distinct (but related) phases based on the three-fold helix conformation. The β-phase just considered is actually a frustrated structure, which has a trigonal cell with three chains in the cell (rather than six or two); the other, new phase introduced in the present paper (the γ-phase) turns out to be very close to the model proposed by Brizzolara et al. for the β-phase: it has an orthorhombic unit-cell that houses two antiparallel helices.

The racemic blend of polylactides forms a crystal structure which is known [6] to have a melting temperature (220°C) significantly higher than those of the enantiomeric α- or β-forms (185 and 175°C, respectively), in spite of very similar or (for the latter structure) even identical helix conformations. Following the initial work of Ikada et al. [6] a crystal structure has been proposed for the racemate by Okihara et al. [4]. The unit-cell is triclinic (space group P1̄) with parameters a=b=0.916nm,c=0.89nm,α=β=109°2, γ=109°8 and houses two enantiomorphous helices. In several recent works (Okihara et al. [4], [7], Brizzolara et al. [5] and Cartier et al. [8]), the racemate was observed to produce highly unconventional single crystals with triangular morphology. As shown by Cartier et al. [8] this morphology results from differences in growth rates on opposite sides of the (110) growth planes, which reflect differences in molecular characteristics and/or concentrations of the co-crystallizing enantiomeric species: for identical Mw and concentrations, hexagonal single crystals are formed [8]. These authors also pointed out that the unit cell considered by Okihara et al. [5] is a sub-cell of a larger trigonal cell that includes six three-fold helices, with parameters a=b=1.498nm,c=0.87nm and symmetry R3c or R3̄c. This unit cell is more familiar for chiral but racemic homopolymers, in particular isotactic polyolefins (poly(1-butene) [9], polystyrene [10], etc.) which can adopt either right- or left-handed helical conformations. By contrast, the unit cell of the polylactide racemate houses of course two different (enantiomeric) molecular species.

In a different context, Zwiers et al. [11] have reported that blends of polylactides and hexamethylbenzene [12] (HMB) form eutectics, with the eutectic composition at 65% concentration of PLLA. Scanning electron micrographs obtained after removal of the HMB solvent–substrate indicate oriented growth of the polymer upon solidification of HMB-rich mixtures. In line with several earlier observations made on binary polymer–crystallizable solvent systems [13], [14], these results suggest that upon solidification of the eutectic the polymer crystallizes epitaxially on the freshly formed substrate crystals.

We report in this paper on the epitaxial crystallization of polylactides (both enantiomers and racemates), and on its use in a further investigation of their crystal structures and polymorphism. Rather surprisingly, three crystal phases of polylactides have been epitaxially crystallized on HMB, which turns out to be a highly versatile substrate. These are: (1) the racemate; (2) the “standard” α-phase of the enantiomer (based on 103 or 107 helices); and (3) an original modification, referred to hereafter as the γ-phase. This phase has been obtained so far only by epitaxial crystallization; its structure is actually very close to a model proposed by Brizzolara et al. [5] for the “stretched” β-form. In a companion paper [15], we deal specifically with this “stretched” β-phase of PLLA, which is not obtained by epitaxial crystallization, and which rests on a recently uncovered frustrated packing scheme[16], [17] that applies frequently (but not exclusively [18]) for polymers with three-fold helical symmetry.

Section snippets

Experimental

The synthesis of the poly(l-lactide) and poly(d-lactide) has been described in detail [19]. The samples have been used in numerous investigations on the structure and morphology of triangular single crystals, and on the properties of enantiomeric and racemic polylactides [4], [7], [20], [21], [22], [23], [24], [25]. Several samples were used during this study, but mainly a couple of enantiomers with identical, 7000 molecular weight. Since, however, the structural features discussed in the

Epitaxy of three crystal phases of polylactides on HMB

A surprising outcome of the present investigation is that two crystalline modifications of the polylactides (plus the racemate) could be produced by epitaxial crystallization on a single substrate, namely HMB. The relevant variable is the crystallization temperature. As already indicated, the crystal phases of enantiomeric and racemic polylactides have significantly different melting temperatures, ranging from 175 and 185°C for the enantiomeric β- and α-phases to 230°C for the racemate.

Conclusion

Polylactides in either enantiomeric or racemic forms crystallize epitaxially on HMB, a low molecular weight organic solvent that had been shown by Zwiers et al. [11] to form eutectics with the polymer. Epitaxial crystallization of polymers associated with eutectic formation in binary low molecular weight organic solvent/polymer systems has already been documented in a number of cases [13], [14]. However, the original feature revealed by the present study is that three different crystal phases

Acknowledgements

Special thanks are due to the Spanish Secretaria de Estado de Universidades, Investigación y Desarrollo for supporting the stay of G.P. in Strasbourg.

References (28)

  • B. Eling et al.

    Polymer

    (1982)
  • R.J.M. Zwiers et al.

    Polymer

    (1983)
  • J. Puiggali et al.

    Polymer

    (2000)
  • J.C. Wittmann et al.

    Prog Polym Sci

    (1990)
  • P. De Santis et al.

    J Biopolym

    (1968)
  • W. Hoogsten et al.

    Macromolecules

    (1990)
  • T. Okihara et al.

    J Macromol Sci Phys B

    (1991)
  • D. Brizzolara et al.

    Macromolecules

    (1996)
  • Y. Ikada et al.

    Macromolecules

    (1987)
  • T. Okihara et al.

    Bull Inst Chem Res Kyoto Univ

    (1988)
  • L. Cartier et al.

    Macromolecules

    (1997)
  • G. Natta et al.

    Nuovo Cimento

    (1960)
  • G. Natta et al.

    Nuovo Cimento

    (1960)
  • L.O. Brockway et al.

    J Chem Soc, Part II

    (1939)
  • Cited by (490)

    View all citing articles on Scopus
    1

    Present address. ENSCL, Laboratoire GEPIFREM, Cité Scientifique, Bat C7 Av. Mendeleiev, BP108, 59652 Villeneuve d'Ascq, France.

    2

    Permanent address. Department of Applied Chemistry, Faculty of Engineering, Okayama University, Okayama 700, Japan.

    3

    Permanent address. ETSEIB, Department d'Enginyera Quimica, Diagonal, 647 E-08028 Barcelona, Spain.

    View full text