Biodegradable Composite Foams of PLA and PHBV Using Subcritical CO₂

Abstract

One key strategy for increasing the application potential for biodegradable plastics lies in improving the physical and mechanical characteristics, which can be attained by inducing a cellular morphology in the pure polymer with the aid of a blowing agent, as well as by blending two or more polymers with the desirable properties. This paper examines the effect that blending two biodegradable polymers has on the thermal properties and morphology of the resultant foams blown with carbon dioxide (CO₂). Polylactic acid (PLA), polyhydroxybutyrate-co-valerate (PHBV) and blends of both were foamed and characterized in terms of thermal characteristics, relative density, cell size, and foam morphology. The results indicate that although PLA and PHBV are immiscible, the presence of small quantities of PHBV (25 wt%) could lead to low density foams with finer, more uniform cells. Furthermore, the crystallinity of PHBV was found to be unaffected by the presence of PLA in the composite, which supports the immiscibility of PLA and PHBV.

Introduction

Biodegradable polymers are believed to be a suitable, environmentally benign replacement to current petrochemical-based polymers. Recent interest in these polymers is spurred on by the increasing costs and eventual depletion of petrochemicals feedstocks, concern for the global environment and green house gases, and a general paradigm shift towards sustainable manufacturing. To date, the applications for biodegradable polymers have grown to include mainly agricultural, biomedical, and food packaging applications; however, there remains a large unrealized potential for many other applications such as landfill cover film, fishing lines and nets, consumer packaging, and various sanitary products. A range of renewable biodegradable polymers are currently available on the market including polylactic acid (PLA) and polyhydroxyalkanoates.

PLA is one of the more widely used biodegradable polymers. Produced from renewable resources (usually starch-rich products like corn, wheat), it may be used to fabricate porous structures for biomedical, and food packaging applications. PLA possesses properties that lie between those of polystyrene (PS) and polyethylene terephthalate (PET) [1], and as a biodegradable alternative it is expected to replace these commodity polymers. Its production cost, estimated at $1–2/kg [1] is expected to decrease further as the markets for the material expand.

Polyhydroxyalkanoates (PHAs) refer to a family of polyesters produced by microorganisms. Philip et al. [2] summarizes the history of PHAs and report that Lemoigne was perhaps the first to extract a polymer of 3-hydroxybutyric acid (P(3HB)). Commercial exploration of the production of P(3HB) was pioneered in subsequent decades by companies like ICI Zeneca, Monsanto, Metabolix and others and eventually led to the commercial availability of present-day polyhydroxybutyrate (PHB) and polyhydroxyvalerate (PHV) [2]. PHB is 100% renewable and biodegradable, and exhibits high stiffness and crystallinity. In order to increase its processing capabilities, PHB is often copolymerized with PHV to form polyhydroxybutyrate-co-valerate (PHBV). Increasing the PHV content in PHB reduces stiffness, melting point, and crystallinity of PHB. Producing PHBV microbially has been estimated at over four times the cost of PLA [1], however, PHBV does perform better in terms of degradation and requires a shorter time to degrade than PLA under the same conditions, as shown in one report [3].

As with conventional petrochemical plastics, key strategies for increasing the application potential for biodegradable plastics lie in reducing the production costs as well as improving the physical and mechanical characteristics, such as strength-to-weight ratio, insulating properties, impact-resistant characteristics, and buoyancy. Most of these advances can usually be attained by inducing a cellular morphology in the pure polymer with the aid of a blowing agent, as well as by blending two or more polymers with the desirable properties. Inducing a cellular structure in plastics can dramatically alter their physical properties, where the degree of change is influenced by the quality of the cell structure. This can result in light weight materials with higher strength- and stiffness-to-weight ratios, which contribute to material savings and ultimately to less waste at the end-of-life stage. Thus cellular structures have the ability to impart biodegradable polymers with a desired range of physical properties, and can also affect the rate of biodegradation.

This paper examines the effect of blending two biodegradable polymers on thermal properties and morphology of resultant foams blown with carbon dioxide (CO₂). This is done as a first step towards identifying a suitable composition which may be used to generate desirable characteristics and thereby improve application potential and increase market penetration for biodegradable polymers.

A temperature soak batch foaming technique was employed to fabricate foams using CO₂ as the foaming agent. This offers advantages over other techniques such as solvent casting, particulate leaching, fiber weaving, and phase inversion [4–6], which are used to produce foams especially for biomedical applications. CO₂ is a relatively benign solvent compared to other organic solvents and residual CO₂ in the cellular structures generally has no detrimental effect for various applications.

Materials and Methods

Materials

Two biodegradable polymers were used in this research. Semi-crystalline PLA was obtained from NatureWorks LLC (Grade 3051D). It was reported to have a density of 1.25 g/cm³, melting temperature of 145–155 °C, glass transition temperature of 55–58 °C, and crystallinity of 37%. Although semi-crystalline, the PLA was also reported to have slow nucleation and crystallization rates such that the sample would be 100% amorphous after normal quenching.

PHBV was acquired from GoodFellow Corporation (Oakdale, PA), and the co-polymer consisted of 88% PHB and 12% PHV. It had a reported density of 1.25 g/cm³, melting temperature of 160 °C, and glass transition temperature of 0 °C, with a degree of crystallinity of 60%.

The blowing agent used to generate foams was commercial purity (>99%) CO₂ at 800 psi, obtained from BOC Canada, Ltd.

Sample Preparation

Compounded samples of PLA and PHBV were made using a 50 gm compounder in mixtures containing 25/75, 50/50 and 75/25 wt% PLA/PHBV. Compounding was done at 180 °C for approximately 10 min or until the blends appeared to be well mixed, at which point they were removed, allowed to air cool and subsequently cut into pellet-sized pieces. Disc-shaped samples were then produced using the procedure outlined below.

PLA, PHBV and compounded PLA/PHBV were formed into disc-shaped samples having an average diameter of 12.5 mm and average thickness of 1.5 mm utilizing a Carver Hydraulic Press and a steel mold. The steel mold basically comprised of a stainless steel sheet of 1.5 mm thickness with holes of the required diameter drilled through. Pellets of the required material or blend were placed in the mold, heated above their melting temperatures to 180 °C for 5 min and subsequently compressed at 1,000 psi for a further 5 min, after which they were removed and quenched in ice-cold water.

Sorption

In order to determine the optimum amount of time to allow for saturation of the polymer samples at various gas pressures, absorption studies were performed where the mass uptake of CO₂ was monitored over various time periods (0–48 h) using the experimental setup shown in Fig. 1. Samples were saturated at the desired pressure (2.76, 4.14 or 5.52 MPa) and the mass change measured at various time intervals. The percentage of CO₂ uptake by mass was measured as a function of time. From these studies, a time period of 24 h was determined to be more than sufficient for saturation of the samples and was chosen as it simplified the logistics involved in successive experiments.

Fig. 1

Schematic of experimental setup

Foaming

The samples produced were subjected to various conditions to induce cellular structures using the temperature soak method on the batch foaming system depicted in Fig. 1. Samples were sealed in high pressure chambers and saturated with CO₂ at room temperature and at three different pressures (2.76, 4.14, and 5.52 MPa) for 24 h. The high pressures required were generated using a Teledyne ISCO syringe pump (model 260D). After 24 h, the gas pressure was released and the samples transferred to a thermostatted water bath and ‘soaked’ at three different temperatures of 65, 75 and 95 °C for three distinct time periods (10, 20 and 40 s). This increase in temperature reduces the solubility of the gas in the polymer, and at the same time softens the polymer to allow the nucleation and growth of cells. Samples were subsequently quenched in cold water after the designated foaming time to freeze the cell-morphology and prevent further changes. Furthermore, the quenched samples were quickly and carefully dried to remove moisture and minimize premature hydrolytic degradation of the samples.

Characterization

The thermal, physical, and morphological characteristics of the foams produced using the temperature soak method were analyzed. The thermal characteristics were determined using differential scanning calorimetry (DSC 2000, TA Instruments) using a heat–cool–heat cycle between 0 and 200 °C with a heating and cooling rate of 10 °C/min. Glass transition points, melting temperatures and degree of crystallinity were all determined from the thermograms.

The physical property of interest was the foam density and this was measured using the density measurement kit for a Denver Instruments mass balance. The density measurement involves the mass of the sample in air and in water and calculating the specific gravity while correcting for the effects of the water temperature. The relative density was calculated by Eq. 1 and the values reported were the average of at least three samples.

$$ {\text{Relative density}} = \frac{\text{Density\,of\,foamed\,sample}}{\text{Density\,of\,unfoamed\,sample}} $$

(1)

The cellular structures of the foams were characterized using images obtained from a JEOL JSM 6060 scanning electron microscope (SEM). Samples were frozen in liquid nitrogen and fractured to obtain a clean cross-section, sputter coated with platinum or gold, and micrographs taken in the SEM. Micrographs were subsequently analyzed using Image-Pro Plus (Media Cybernetics) to determine the average cell diameter at each foaming condition.

Results

Effect of PLA/PHBV Composition on Thermal Behavior

The DSC heating and cooling thermograms for PLA/PHBV combinations are shown in Fig. 2a and b, respectively. The observed values of the glass transition and the enthalpy of melting for the different polymer combinations are summarized in Table 1. From the heating curves (Fig. 2a), samples containing 50% or more of PHBV exhibit two endothermic peaks at approximately 156 and 145 °C, which correspond to common crystal perfection behavior seen in many semi-crystalline polymers. The difference between these two melting points becomes progressively less distinguishable until 25% PHBV where it becomes indistinguishable. For pure PLA, the glass transition was observed at 60 °C and there was no melting or crystallization peak observed, which indicates the sample to be fully amorphous under the DSC experimental conditions. As the PHBV content increases the observed glass transition temperature of the composite decreases closer to the glass transition point of pure PHBV (0 °C) in a non-linear fashion. The observed heat of fusion, ΔH _f, of the composite increases as the amount of PHBV is increased, since PHBV is significantly more crystalline than PLA because of much faster crystallization kinetics under the chosen DSC conditions.

Fig. 2

a Heating endotherms for PLA/PHBV blends; b cooling exotherms for PLA/PHBV blends

Table 1 Observed thermal characteristics for PLA/PHBV blends

The composition dependence of the glass transition temperature in a polymer blend can be modeled using the Gordon–Taylor (G–T) equation [7, 8]:

$$ T_{\text{g}} = \frac{{W_{1} T_{{{\text{g}}1}} + kW_{2} T_{{{\text{g}}2}} }}{{W_{1} + kW_{2} }} $$

(2)

where W ₁ and W ₂ are the weight fractions of the two blend components, and T _g1 and T _g2 are the respective glass transition temperatures of the pure components. The parameter k is formally represented by Δα₂/Δα₁, where Δα_i is the change between the liquid and glassy thermal expansion coefficient at T _gi. More usually, k is used as a fitting parameter; a value close to unity indicates very good miscibility between blend components, whereas too high or low a value for k indicates limited miscibility [8]. In this paper, the T _g of PLA was observed to be depressed to 55, 48 and 41 °C with the addition of 25%, 50% and 75% PHBV respectively. This dependence of T _g on the PHBV content is depicted graphically in Fig. 3, and is compared with the G–T model. It must be noted that no T _g was observed for neat PHBV during the DSC runs, consequently, the supplier’s T _g of 0 °C was used to determine the G–T relationship. As Fig. 3 shows, there is fairly good agreement between the observed T _g and the G–T model for k = 0.18. However, this low k value indicates low miscibility which is borne out in the SEM images presented in a following section.

Fig. 3

a Composition dependence of T _g; b heat of fusion vs. PHBV content

It is worth mentioning that for the PLA sample, relaxation phenomenon was observed as indicated by the peak immediately following the glass transition (see Fig. 2b). This phenomenon which has been reported by other researchers [9–11] is indicative of the gradual migration (with age) in the relaxation of the properties of an amorphous polymer while in the thermodynamically unstable glassy state. The relaxation enthalpy (ΔH _relax) can be obtained by integrating the area under the peak in the region of the glass transition, and is generally reported to increase with aging time and temperature up to T _g. Since the first DSC run removed any previous thermal history, the enthalpy relaxation observed and presented in Fig. 2 (0.365 J/g) for pure PLA, reflects the aging of the polymer between the previous cooling cycle below the observed T _g of 60 °C and the subsequent reheating to T _g (approximately 12 min).

The degree of crystallinity of PHBV (X _c,PHBV) in each of the blends was also calculated using Eq. 3:

$$ {X}_{\text{c, PHBV}} = \left( {\frac{{\Updelta H_{\text{f}} m_{\text{sample}} }}{{\phi_{\text{PHBV}} m_{\text{sample}} }}} \right)\frac{1}{{\Updelta H_{\text{f, PHBV}}^{0} }} $$

(3)

where ΔH _f is the heat of fusion (J/g) calculated by integrating the area under the melting peak of each thermogram (see Table 1), m _sample is the mass of each sample (g) studied, ϕ _PHBV is the mass fraction of PHBV in the PLA/PHBV composite, and $\Delta H_{\text{f, PHBV}}^{0}$is the heat of fusion for 100% crystalline PHBV taken as 146 J/g [12]. The calculated results are plotted in Fig. 3b. It can be seen from this figure that the degree of crystallinity of PHBV in the PLA/PHBV composite remains fairly constant regardless of the amount of PLA in the composite, indicating that PLA has neither an inhibiting nor catalyzing effect on the crystallization of PHBV; this provides further support for the molecular immiscibility of PLA and PHBV.

Effect of PLA/PHBV Composition on Cell-Morphology

Figures 4 and 5 illustrate the effects of the PLA/PHBV composition on the relative density and average cell size of the foams obtained respectively. Figure 4a–c present the variation of the relative density with the PHBV content at foaming temperatures of 65, 75 and 95 °C respectively. In these graphs, the first bar in each cluster of three, represents a foaming time of 10 s, while the second and third bar represent 20 and 40 s foaming time, respectively. It can be seen that a significant reduction in density was achieved at low PHBV contents with the lowest being achieved for pure PLA. At PHBV contents of 50% or greater, no cellular structure could be achieved under the foaming conditions investigated, and the relative density remained virtually unchanged—any deviation from a relative density of 1.0 is attributed to the formation of a few large crevices within the sample possibly due to the accumulation and expansion of gas in pre-existing cavities during foaming.

Fig. 4

Effect of PHBV content on relative density at: a 65 °C; b 75 °C and c 95 °C

Fig. 5

Effect of PHBV content on cell size at: a 65 °C; b 75 °C and c 95 °C

For the pure PLA foamed at 75 and 95 °C, the lowest foam densities were achieved at 2.76 and 4.14 MPa, followed by a significant jump in relative density at 5.52 MPa. This increase was as much as four times the value at 4.14 MPa when the foaming temperature was 95 °C. At 65 °C a similar trend is observed except that there was very little change from 4.14 to 5.52 MPa. The observation of increased relative density at 5.52 MPa is most likely due to the CO₂ induced crystallinity of PLA which occurs under high pressures as reported by Hu et al. [13]; the resulting increase in crystalline regions at higher pressures effectively reduces the foamability of the PLA material.

At a PHBV content of 25%, the lowest density foams were generally achieved at the higher temperature of 95 °C. Also at this temperature, the increase in density was observed on moving from 4.14 to 5.52 MPa, which further suggests the theory of CO₂ induced crystallinity. However at 75 °C, the relative density was fairly constant at 2.76 and 4.14 MPa and then was reduced at 5.52 MPa; furthermore the density reduction from 4.14 to 5.52 MPa was more significant at 65 °C. It may be that at the lower foaming temperature, the gas dissolution is sufficiently slow, so that the reversal of the CO₂-induced crystallinity (caused by the dissolution of the gas) permits cell nucleation and growth. Conversely at the higher temperatures, the gas diffusivity is higher and the gas leaves the polymer at too fast a rate to take advantage of the reversal in the CO₂-induced crystallinity.

No significant trends were seen in relation to the effect of foaming time on the relative densities and its effects were considerably smaller than other parameters studied. According to the classical nucleation theory as time goes by during the foaming process, gas bubbles nucleate, decreasing the CO₂ concentration in the polymer matrix. This nucleation process slows down with time therefore further decrease of CO₂ concentration in the polymer matrix is now facilitated by growing the cell size. However the differences in the cell size and relative density during the transition from nucleation phase to growth phase (increasing foaming time) are quite small compared to differences in other parameters such as saturation pressure and foaming temperature. It should be added that caution must be taken when making conclusions about the effects of foaming time as it is a wholly kinetic parameter and will vary from laboratory to laboratory.

Figure 5a–c demonstrates the variation in cell size with the PHBV content at foaming temperatures of 65, 75 and 95 °C, respectively. These graphs only show the cell sizes for PHBV content up to 25% as there was no foam structure (hence no cells) seen above 25% PHBV. The largest cells were obtained for PLA, and varied from 20 to 1,500 μm depending on the pressure and temperature. At 2.76 MPa, the cell size was observed to decrease with increase in temperature from 65 to 75 °C, followed by a significantly large increase at 95 °C. This decrease in cell size for pure PLA at 2.76 MPa as the foaming temperature is raised from 65 to 75 °C is attributed to the reduced need for phase separation and cell nucleation because of the low CO₂ concentration in the polymer matrix at 2.76 MPa coupled with the reduced thermal energy at 65 °C. The remaining CO₂ in the polymer matrix diffuses to the few gas bubbles that are present and grows them, resulting in a large cell size. At 75 °C there is greater nucleation density than at 65 °C, resulting in more bubbles to grow with more or less the same gas concentration therefore the cell size is lower than the cell size at 65 °C. The increase in cell size at 2.76 MPa as foaming temperature is raised from 75 to 95 °C is due to cell coalescence precipitated by the reduced melt strength of the polymer and the high diffusivity of the gas contributing to unregulated cell growth. At the higher pressure of 4.14 MPa, there was an approximate fivefold increase in the cell size from 75 to 95 °C. Again, coalescence played a dominant role in determining the cell size at the higher temperature.

At 25% PHBV content, there was a marked decrease in the cell size compared to 100% PLA; cell sizes were in the range of 3–14 μm. No significant variations or trends could be noted between different pressures or temperatures at 25% PHBV.

Effect of Foaming Parameters on Cell Morphology

The SEM images of the cellular structures of PLA foams processed at various pressures and temperatures and a 20 s foaming time are shown in Fig. 6. Similar to observations in Fig. 5 for PLA saturated at 2.76 MPa, the cell size decreases and then increases as the foaming temperature is increased from 65 to 75 to 95 °C. This is further evidence to our understanding that at low CO₂ saturation pressure there is not enough thermodynamic need for phase separation and cell nucleation at 65 °C. Therefore cell growth rather than nucleation is preferred. At the higher temperature, 75 °C, a greater amount of cell nucleation is present resulting in less gas concentration to grow the cells further; therefore the cell sizes remain small. At an even higher temperature, 95 °C, the greater need for phase separation coupled with faster diffusion kinetics results in an almost instantaneous foaming process as indicated by a constant cell size with change in foaming time from 10 to 40 s.

Fig. 6

PLA foams at 20 s foaming time (all images at 70× unless otherwise stated)

At higher saturation pressures the effects of CO₂ induced crystallinity are evident on the cellular morphologies of the foams. PLA specimens saturated at 4.14 MPa have a lower induced crystalline content than specimens saturated at 5.52 MPa. PLA specimens saturated with CO₂ at 5.52 MPa lack the uniform cellular structures obtained at lower saturation pressures of 2.76 and 4.14 MPa, with only surface foaming evident at higher temperatures (75 and 95 °C). The surface foaming is explained by the higher temperature at the surface resulting in softening of the semi-crystalline polymer allowing for cells to nucleate and grow.

The low magnification SEM images of the cellular structures of 25% PHBV/PLA foams processed at various pressures and temperatures and a foaming time of 20 s are shown in Fig. 7. In general, it can be observed that the addition of 25% PHBV to the mixture resulted in finer cells, when compared to 100% PLA. The high magnification SEM images of the cellular structures of 25% PHBV/PLA foams processed at a saturation pressure of 4.14 MPa, foaming temperature of 95 °C and a foaming time of 20 s are shown in Fig. 8. Figure 8 illustrates the immiscibility of the two polymer components. Spherical particles assumed to be PHBV, can be seen evenly dispersed throughout the PLA matrix. Where cells are formed, the particles are still visible and they may have acted as sites for heterogeneous nucleation. In addition, given the even distribution and micron-sized dimensions of these particles, they may also account for the reduced cell size (by impeding cell growth) and increased cell density observed when compared to 100% PLA.

Fig. 7

75/25 PLA–PHBV composite foams at 20 s foaming time (all images at 70× unless otherwise stated)

Fig. 8

SEM micrographs for 25% PHBV/PLA foams at 4.14 MPa saturation pressure, 95 °C foaming temperature, and 20 s foaming time: a 900×; b 2000×

Conclusions

This study investigated the thermal properties and morphology of foams manufactured from blends of PLA and PHBV with CO₂ as the blowing agent. The DSC results indicate that PLA and PHBV are immiscible, and blending achieves only a physical dispersion of one component into the other. The composition dependence of the glass transition temperature was found to obey the Gordon–Taylor model, with a modeling parameter value indicative of immiscibility. The finest cellular morphologies were obtained at a PHBV content of 25%. No foams could be obtained using the temperature soak batch foaming technique with PHBV contents greater than 25%, presumably due to the high crystallinity of the blend. At this ideal PHBV content of 25%, lower density foams were achieved at the higher temperature of 95 °C. Furthermore, the CO₂ induced crystallization of PLA was found to most likely occur as the cellular morphology deteriorated at higher saturation pressures for both PLA and the blend. In general cell sizes in the range of 20–1,500 μm and 3–14 μm were obtained for PLA and 75/25 PLA/PHBV blend, with cells at the larger end of the scale being primarily due to cell coalescence.