Biomacromolecules 2001, 2, 581-587
581
On the Relationship between Crystalline Structure and Amorphous Phase Dynamics during Isothermal Crystallization of Bacterial Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) Copolymers I. S ˇ ics,† T. A. Ezquerra,* A. Nogales,‡ and F. J. Balta´ -Calleja Instituto de Estructura de la Materia, CSIC Serrano 119, Madrid 28006, Spain
M. Kalnin¸ sˇ and V. Tupureina Riga Technical University, Institute of Polymer Materials, 14 Azenes str., Riga LV 1048, Latvia Received February 26, 2001; Revised Manuscript Received April 24, 2001
The isothermal crystallization process of a poly(3-hydroxybutyrate-co-3-hydroxyvalerate) copolymer, P(HBco-HV) with a HB/HV ratio 78/22 was investigated by simultaneous small-angle X-ray scattering (SAXS), wide-angle X-ray scattering (WAXS), and dielectric spectroscopy (DS). By use of this experimental setup (SWD), we have obtained simultaneous information about changes occurring in both the crystalline and the amorphous phases during crystallization. By using the Havriliak-Negami formalism to analyze the dielectric relaxation data, a strong dependence of the relaxation curve shape with the development of the crystalline phase was found. However, in this particular copolymer, the developing crystalline domains do not affect significantly the average segmental mobility in the amorphous phase. This effect is discussed in the light of the enrichment of amorphous phase by HV comonomer units during primary crystallization, hindering the secondary crystallization processes. Results support the hypothesis that the decrease of the physical-aginglike behavior, observed in P(HB-co-HV) copolymers as the amount of HV increases, can be attributed to the progressive inhibition of secondary crystallization mechanisms. Introduction Poly(3-hydroxybutyrate) (PHB) is a stereoregular homopolyester produced naturally by a great variety of bacterial strains.1,2 There has been growing interest in this type of polymers over the last 20 years due mainly to two aspects: On one hand, the biodegradability of this polymer makes it potentially suitable for the fabrication of biodegradable plastic items.1,3 On the other hand, PHB is obtained with very high purity due to the absence of catalyst remnants. Therefore, PHB is a polymer system which is very appropriate for basic crystallization and nucleation studies.2 The absence of impurities in this polymer induces the crystallization kinetics to be slower as compared with other commercial polymers like the polyolefins.4 In addition, the absence of impurities seems to be responsible, in this polymer, for severe physical-aging-like effects observed at room temperature, which negatively affect its mechanical properties enhancing embrittlement.5,6 In an attempt to compensate for these technical drawbacks, PHB has been copolymerized with poly(3-hydroxyvalerate) (PHV), giving rise to P(HB-co-HV) copolymers.2 These materials, which * To whom correspondence should be addressed. E-mail: imte155@ iem.cfmac.csic.es. † Permanent address: Riga Technical University, Institute of Polymer Materials, 4 Azenes str., Riga LV 1048, Latvia. ‡ Present address: JJ Thomson Laboratory, The University of Reading, Whiteknights, RG6 6AF, UK.
are commercialized under the trade name of “Biopol” (Zeneca Bio Products), belong to the family of the polyhydroxyalkanoates and can be obtained by a bacterial method.1,7 In contrast to PHB, P(HB-co-HV) copolymers present an improvement in mechanical properties, partially due to a reduction of the physical-aging-like effects.6 Although it is assumed that secondary crystallization plays an important role on the above-mentioned physical-aginglike phenomena, it was shown that microstructure development in these copolymers also affects selectively the segmental dynamics above the glass transition temperature (Tg).8 While copolymers with low HV content (HV < 22 mol %) present evidence of constrained amorphous phase dynamics after crystallization, those with higher HV content do not present such an effect.8 As it is known, at temperatures above Tg mechanical energy can be dissipated to some extent by the primary relaxation, R, through molecular motions. Therefore, an accurate description of the relationships between microstructure and dynamics is needed in these copolymers in order to understand the reason for the influence of copolymerization on the mechanical properties. To obtain precise information about the changes occurring in a polymeric system during crystallization, a real time experimental setup is of crucial importance.9 As far as microstructure development is concerned, high intensity synchrotron radiation offers the possibility to perform simultaneous, real time small- and wide-angle X-ray scat-
10.1021/bm0155266 CCC: $20.00 © 2001 American Chemical Society Published on Web 05/23/2001
582
S ˇ ics et al.
Biomacromolecules, Vol. 2, No. 2, 2001
Figure 1. Schematic presentation of sample capacitor (sandwich): (1) metal electrodes, (2) Kapton spacer with cutoffs, and (3) polymer sample.
tering, SAXS and WAXS, respectively.9 An improvement in understanding of the correlation between microstructure and dynamics is obtained when both SAXS and WAXS experiments are accompanied by dielectric spectroscopy (DS) performed simultaneously (SWD).10 In a SWD experiment, one may monitor simultaneously, in real time, both the microstructure development, through SAXS and WAXS, and the dynamic changes occurring in the amorphous phase, by DS. Consequently, a more complete picture of the crystallization process can be obtained. Previous attempts of combining X-ray scattering measurements, over particular spatial ranges, and dielectric spectroscopy were shown to be useful to characterize changes occurring in the induction time for crystallization.11 In the present study, we report simultaneous measurements of small and wide-angle X-ray scattering and dielectric relaxation spectroscopy (SWD) during isothermal crystallization of a poly(3-hydroxybutyrate-co-3-hydroxyvalerate) copolymer with 22 mol % of HV units. The selection of the sample was done based on the fact the P(HB-co-HV) copolymers with HV > 22 mol % do not present evidence of constrained amorphous phase dynamics after crystallization. Because of the fact that the incorporation of HV units tends to reduce the physical-aging-like effects, it is a question wether these two facts may be related. The aim of the work is to shed some light on the physical-aging-like mechanism “in situ” by simultaneous monitorization of the evolution of both crystalline and amorphous phases in this copolymer. Experimental Section Materials. A poly(3-hydroxybutyrate-co-3-hydroxyvalerate) sample with a HV comonomer units content of 22 mol % was obtained from ZENECA Bio Products, U.K. The sample was characterized by gel permeation chromatography using polystyrene standards and chloroform as a solvent (Mn ) 234 000 g/mol, Mw ) 428 000 g/mol, Mw/Mn ) 1.8). Samples for dielectric measurements were prepared in a fivestep procedure (schematized in Figure 1) in order to ensure that the polymer sample fills in completely the space between electrodes. This procedure includes the following: (1) Polymer films were prepared by melt compression. (2) A disk-shaped sample, with a diameter slightly smaller than
that of the electrode to be used, was cut from the polymer film. (3) The disk-shaped sample was sandwiched between two thin aluminum electrodes provided with a circular Kapton spacer. The inner diameter of the spacer was chosen to fit exactly the sample disk and the thickness was slightly smaller than that of sample. (4) Amorphous films were prepared by heating the complete set (sample, electrodes and spacers) above the melting temperature at T ) 180 °C. Narrow cuts were made in the spacer to allow removal of excess of molten material. (5) A slight compression was then produced following quenching between iron blocks held at 0 °C temperature. Precaution was taken to hold the sample in molten state for periods of time, as short as possible (t < 60 s), to avoid thermal degradation. Thereafter, the sandwich was introduced into the experimental cell and heated to the crystallization temperature. The latter, Tc ) 35 °C, was chosen in order to ensure that Tc >Tg and high enough for cold crystallization to occur at rates suitable for experimental observation. Additionally, at the chosen Tc, the R-relaxation process appears to be centered in the measured frequency window. The contribution of spacer dielectric constant was automatically subtracted from the dielectric measurements. Techniques. Simultaneous SAXS, WAXS, and DS experiments were performed in the polymer beam-line at HASYLAB (DESY) synchrotron facility in Hamburg, Germany. A detailed description of experimental setup for these simultaneous measurements has been reported elsewhere.10 A novel experimental setup, denominated SWD, has been recently developed to enable simultaneous measurements of small and wide-angle X-ray scattering and dielectric spectroscopy in real time.10 The SWD cell is incorporated into a vacuum chamber (10-2 Torr) specially designed to perform X-ray scattering measurements with synchrotron radiation. A wavelength λ ) 0.15 nm was employed for X-ray diffraction study. A semicrystalline PET standard sample was used for the WAXS and rat tail cornea for the SAXS calibration. Each frame was collected over 60 s by two linear detectors with no waiting time in between. Collected data were corrected for primary beam intensity fluctuations during experiment and background. Complex dielectric permittivity measurements, (* ) ′ - i′′ ) were performed in the frequency range 103 Hz < F < 106 Hz, using a HP 4192 impedance analyzer. Circular electrodes, 3 cm diameter, were employed to prepare a sandwich type capacitor as described above. Parallel isothermal crystallization experiments were performed in a microscope (Leitz) equipped with a hot stage (Mettler) under thermal conditions as close as possible to those used in the SWD experiment. Results 1. X-ray Diffraction. Figures 2 and 3 show the real time evolution of the WAXS and SAXS patterns with crystallization time during the isothermal crystallization of the investigated P(HB-co-HV) copolymer. The patterns are represented as a function of the scattering vector q ) 4π/λ(sin θ), 2θ being the scattering angle. The development of the crystalline phase is evidenced by the appearance in
Poly(3-hydroxybutyrate-co-3-hydroxyvalerate)
Biomacromolecules, Vol. 2, No. 2, 2001 583
Figure 3. Lorentz-corrected SAXS intensity profiles for the same crystallization times as in Figure 2. Presented curves are obtained by averaging over two frames; error ) (1 min. Each frame 60 s long.
Figure 2. Wide-angle X-ray diffraction patterns for selected crystallization times. The peak separation procedure for Bragg reflections is presented by dotted lines. Each frame is 60 s long.
the WAXS diffraction patterns of the characteristic Bragg peaks, corresponding to the 020 and 110 reflections of the orthorhombic PHB crystal lattice.12 This is in agreement with previous studies showing that poly(3-hydroxybutyrate-co3-hydroxyvalerate) copolymers crystallize in the PHB crystal lattice provided the HB molar concentration is smaller than 37%.13 To estimate the degree of crystallinity, a peak separation procedure was used.14 Gauss functions are assumed to describe the shape of the crystalline reflections. In addition, a function was generated describing the amorphous halo based on the WAXS pattern of the fully amorphous sample (tc ) 1 min). In the semicrystalline patterns, the generated halo function was allowed to diminish in intensity in order to consider the reduction of the amorphous material as crystallization proceed. The crystalline content of the sample after each scan was estimated as the ratio between the area under the Bragg reflections and the total scattered area. In doing so, an estimate of the mass fraction of crystallinity (Xcwax) is obtained (Figure 5a). The simultaneous time evolution of the SAXS patterns can be estimated from the Lorentz-corrected intensity (I*q2)
Figure 4. Isothermal plots of dielectric loss, ′′, (a) and dielectric constant, ′, (b) as a function of frequency for different crystallization times (tc ) 1, 15, 30, 40, 45, 50, 55, 60, 65, 70, 75, 80, 90, 100, 110, 120, 150). The continuos solid lines represent the results of fitting experimental points to the HN equation.
profiles presented in Figure 3. Despite the high noise-tosignal ratio, partly due to the X-ray absorption of the
584
S ˇ ics et al.
Biomacromolecules, Vol. 2, No. 2, 2001
Figure 6. Characteristic parameters, obtained from dielectric loss curves fittings to the Havriliak-Negami equation: (a) solid points, dielectric strength, ∆; (b) HN shape parameters; c ) asymmetry parameter and b ) symmetric broadening parameter.
Figure 5. Experimental data obtained from SWD measurement: (a) Xc, crystalline content from WAXS; (b and c) long spacing (L) and integrated intensity (Q, arbitrary units) respectively from SAXS; (d) frequency of maximum loss Fmax and maximum loss value ′′max, as a function of crystallization time tc.
aluminum electrodes which sandwich the polymer sample, the appearance of a well-defined SAXS maximum, centered around q ≈ 0.088 Å-1, is detected as crystallization time increases. The resulting long spacing, calculated as L ) 2π/ qmax , and its development with crystallization time are shown in Figure 5b. The measured value of L ) 70 Å is in agreement with previously reported data.7 Additionally, the integrated intensity in the measured q range, Q, was calculated and it is presented in Figure 5c in arbitrary units. This magnitude is related to the invariant.13 2. Dielectric Spectroscopy. The complex dielectric permittivity, * ) ′ - i′′, measured simultaneously with the WAXS and SAXS patterns, is illustrated in Figure 4 by selected curves of the dielectric loss, ′′ (Figure 4a), and of the dielectric constant, ′ (Figure 4b), as a function of the frequency. At the selected crystallization temperature of T ) 35 °C, the relaxation process which appears in the measured frequency window corresponds to the R-relaxation.8 This relaxation is characterized by a maximum of the dielectric loss and a steplike decrease in the dielectric constant. For the initial amorphous sample (tc ) 1 min), the frequency of maximum loss, Fmax, is centered around 4 × 104 Hz. Clearly, after a short period of relative constancy, a decrease in the intensity of the relaxation process, characterized by the value of the dielectric loss at the maximum ′′max, is observed. Although, at a first glance, no abrupt changes of Fmax with time are detectable as observed for other
polymers,8,10 at a closer look, a small continuous shift of Fmax toward lower frequencies is seen. These observations are visualized in Figure 5d which represents the variation of Fmax and ′′max with time. In addition, pronounced changes in the shape of the dielectric curves are observed. The dielectric loss curves become wider and apparently change from an asymmetric to a symmetric shape. A detailed evaluation of dielectric loss curves according to the fitting procedure of the Havriliak-Negami (HN) phenomenological equations15 has been performed in order to extract additional information about the changes in the dynamics of amorphous phase with crystallization time. The (HN) formalism proposes an analytical expression for the complex dielectric permittivity as *(ω) - ∞ )
o - ∞ [1 + (iωτo)b]c
(1)
where o - ∞ ) ∆ is the dielectric strength, το is the central relaxation time, and b and c are parameters which describe the shape of the relaxation time distribution function (symmetric and asymmetric broadening).15 The imaginary (′′) and real (′) parts of the complex dielectric constant from isothermal scans were simultaneously fitted to eq 1. Graphical results of this fitting procedure for ′′ are presented in Figure 4 by the continuous curves. Figure 6 shows the evolution with crystallization time of the characteristic parameters of the dielectric curves. The dielectric strength, ∆ (Figure 6a), decreases with a sigmoidal shape simiar to that followed by the experimentally measured ′′max values (Figure 5d). Discussion 1. Crystalline Phase Evolution during Isothermal Crystallization. As seen in Figure 5a, the crystallinity
Poly(3-hydroxybutyrate-co-3-hydroxyvalerate)
Biomacromolecules, Vol. 2, No. 2, 2001 585
Figure 7. Polarized optical micrographs taken during isothermal crystallization of PHB-22% HV sample at T ) 35 °C: (a) 30, (b) 45, (c) 60, (d) 75, (e) 90, and (f) 120 min. The white straight line at the bottom corresponds to 15 µm.
evolution in the copolymer, during isothermal crystallization, exhibits a typical sigmoidal shape. Here, after a slight induction time (0 < t
