• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Poly(L-lactic acid) composites with flax fibers modified by plasticizer absorption

Posted on: Saturday, 11 October 2003, 06:00 CDT

The effect of modified flax fibers by plasticizer absorption in poly(L-lactic acid) composites was investigated. The plasticizers chosen were triethyl citrate (TEC), tributyl citrate (TBC) and glycerol triacetate (GTA), which were derived from natural sources. Characterization was performed by differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA). The morphology was examined from scanning electron microscopy (SEM) and optical microscopy (OM). The results showed that the plasticizer caused a marked increase in the storage modulus of the composites, which could be due to an improvement in the morphology of the matrix and a smoother surface coverage of the fibers by the matrix. The thermal properties were also affected, in which the glass transition temperature (T^sub g^), the crystallization temperature (T^sub c^) and the crystallinity (X^sub c^) were reduced depending on the plasticizer. The citrate esters revealed to be the most effective plasticizers of those tested.

INTRODUCTION

Fiber reinforced polymer composites are extensively used in many applications, ranging from aerospace technology to the automobile industry (1). These composites offer high performance such as high strength and stiffness, great versatility, and processing advantages at favorable cost (2). By altering the fiber-polymer combination, the composite properties can easily be modified to the requirements specific for an application (3). The properties of the polymer and the fibers are important for their contribution to the properties of the composite, but the key factor controlling composite properties is the polymer-fiber interfacial bond strength.

Many existing fiber composites are based on petroleum-derived polymers and synthetic fibers such as carbon or glass. However, the components of these traditional composites are difficult to dispose of or recycle after use. With increasing environmental concerns over waste disposal, a promising means of reducing waste is to use biodegradable products. There has been a renewed interest in composites composed of natural fibers and biodegradable polymers (4- 6).

A large number of natural fibers are successfully used in many composites in recent publications. These include jute, hemp, kenaf, ramie, sisal and flax. These natural fibers are of low density, have high strength and stiffness relative to their density, low cost, good thermal and acoustic insulating properties, and are friendly to processing equipment (7). Also, they are fully recyclable and combustible and emit no noxious gases (8). However, when used as reinforcement in polymers, a major problem is their incompatibility with hydrophobic polymers, such as polypropylene, because of their hydrophilic nature. Insufficient wetting of the fibers by the polymer matrix is proven to lower the tensile strength (9) and stiffness of a composite, as a poor interface cannot effectively transfer the stress from the polymer matrix to the fibers. Moisture absorption is another problem, as the moisture present is known to cause voids, reducing the strength of the composite (10). The moisture content will vary, depending on relative humidity or wetting of the composite. Moisture will interfere with melt dispersion and processing of the composites, since processing temperatures of the order of 200[degrees]C are necessary. When moisture is removed from natural fibers, such as flax, they become brittle, thereby losing their effectiveness as reinforcements.

Of the many available biodegradable polymers such as thermoplastic starch and soy plastics, poly(lactic acid) (PLLA) is one that exhibits good mechanical properties and processability (10). PLLA is derived from 100% renewable resources such as corn and sugar beets (11), and it degrades to non-toxic compounds in landfills. It can be produced from either direct condensation of lactic acid and by ring-opening polymerization of the cyclic lactide dimer (11). Lactic acid is chiral, but L-lactic acid can be produced exclusively when biosynthesis is used. Poly(L-lactic acid) is an isotactic polymer that is crystallizable, unlike the racemic form that remains amorphous. The L-enantiomorph is the form referred to in this research. Poly (L-lactic acid) is slow to crystallize and can form large spherulites. However, one limitation of PLLA that restricts its range of applications is its brittleness despite its high modulus and tensile strength.

To overcome the brittleness of PLLA, plasticizers such as citrate esters (12, 13), 1, 2-propylene glycol, glycerol (14), poly (ethylene glycol) (15), glucose monoesters, and partial fatty acids (16) have been used with some success. The role of plasticizers is to decrease the glass transition temperature (T^sub g^) and the crystallinity of a polymer (X^sub c^). Of these plasticizers, citrate esters have been preferred because of their effectiveness as a plasticizer, but also because they are derived from naturally occurring citric acid, which is non-toxic (12). Depending on the citrate ester used as well as the composition, depressions in the T^sub g^ and melting temperature (T^sub m^) as large as 45[degrees]C and 15[degrees]C (respectively) can be obtained (12).

In this investigation, composites consisting of PLLA and modified flax fibers with plasticizers were prepared to achieve improved properties. Modification of the fibers was achieved by absorption of the plasticizers into or onto the fibers that were previously dried. The plasticizer was intended to serve two functions: 1) to overcome weakening of the fibers as a result of the moisture lost; 2) to increase the ductility of PLLA, thereby producing composites with enhanced mechanical properties. The properties of the composites were characterized by differential scanning calorimetry, dynamic mechanical flexural analysis, and hot stage optical and scanning electron microscopy.

EXPERIMENTAL

Materials

A commercial sample of poly(L-lactic acid) (PLLA) (Resomer L206; Mw = 110,000 g/mol ) obtained from Boehringer Ingelheim (Ingelheim, Germany) was used for this study. Flax fibers (Durafiber Grade One of 95% purity) were obtained from Durafiber Inc. (Cargill) and were dried in a vacuum oven at 100[degrees]C for 2 hours before preparation of composites. The plasticizers used were tributyl citrate (TBC) (Merck), triethyl citrate (TEC) (Merck), and glycerol triacetate (GTA) (Hopkins & Williams Ltd). Their physical properties are given in Table 1.

Table 1. Physical Properties of Plasticizers.

Fiber Composite Preparation

Treatment of flax with plasticizers was performed after removal of moisture to a constant weight in a vacuum oven. Plasticizers were used as supplied. A weighed amount of dried flax fibers was added into the neat plasticizer and the mixture heated at 120[degrees]C for 8 hours. The long heating period was to ensure that air was expelled from the interstitial spaces in the flax, allowing the plasticizer to absorb into or onto the flax. The high temperature decreased the viscosity of the plasticizer, further assisting with absorption into the flax. After treatment, it was expected that at room temperature the viscous plasticizer would be strongly absorbed into or onto the flax. After cooling to room temperature, the treated flax was washed with acetone repeatedly to remove any excess plasticizer. The treated flax was then again dried in a vacuum, at room temperature, to remove any residual acetone. The amount of plasticizer present in the flax was determined by the difference in weight before and after treatment. This was typically about 10 vol%, which was similar to the 8-10 vol% of water contained in undried flax.

PLLA-flax composites of 1:1 ratio by volume were prepared by dissolving PLLA in a minimum amount of chloroform under reflux. After cooling, an appropriate amount of flax was added to the PLLA solution and they were mixed together thoroughly, casted onto a glass plate, and allowed to dry. This was the most convenient method to mix small quantities of polymer and flax to ensure homogeneous dispersion of the matrix onto the fibers. Larger amounts could be melt-mixed in an internal mixer or via an extruder without the need for the chloroform. The dried PLLA and flax were cut into small pieces that were dried further at 5O0C for 2 hours to remove any residual solvent. The small pieces were then heat pressed at 175[degrees]C between polytetrafiuoroethylene sheets into thin sheets. They were then stacked to obtain sample bars of 15 mm x 20 mm x 1.5 mm. The temperature used to press the samples was chosen after careful consideration of the decomposition temperatures of the polymer, in order to minimize degradation. The short residence time in the press of about 3 minutes was to ensure that some uncontrollable evaporation was reduced. The sample bars were then annealed under nitrogen at 100[degrees]C for 3 hours to allow the PLLA to crystallize and equilibrate with the fibers and plasticizer. Samples of PLLA with no libers that contained about the same amount of plasticizers were prepared and subjected to the same treatment as the PLLA-flax composites. These were used to observe the effect of the plasticizer on the matrix without the influence from the fibers.

Measurement of Properties

The dynamic mechanical properties were evaluated in a 3-point bend \mode using a Perkin-Elmer DMA7e dynamic mechanical analyzer (DMA). The sample dimensions was 18 x 10 x 1.5 mm^sup 3^. A static and dynamic force of 500.0 and 400.0 mN respectively were used and the frequency was 10.0 Hz. These parameters were determined to be within the linear viscoelastic range. Temperature scans were performed from -50[degrees]C to 100[degrees]C at a rate of 2[degrees]C/min.

Differential scanning calorimetry (DSC) (Perkin-Elmer Pyris 1 equipped with a Perkin Elmer Intracooler 2P) was used to determine the melting temperatures, the crystallization temperatures and the heat of fusion of PLLA and PLLA-flax systems. Approximately 5 mg samples were used and the temperature scans were performed from - 40[degrees]C to 200[degrees]C at a rate of 10 K/min. At least two replicates were performed for each sample and the results are averaged.

Scanning electron microscopy (SEM) (Philips XL-30) was used to observe the effect of the plasticizer on composite morphology and the matrix adhesion to the fibers. Small sections of the composite were cut and attached to aluminum stubs, and then coated with gold. Electron energy of 20 kV was used for all micrographs.

Optical microscopy with cross polarizers (Nikon Labophot2 optical microscope) was used to observe the morphology of the PLLA-flax composites. All samples were melted using a Mettler FP82 HT hot stage at 175[degrees]C and crystallization was observed upon cooling at a rate of 10[degrees]C/min to 25[degrees]C. Photographs were captured with the aid of IPLab Spectrum Scientific Image Processing Software 3.1a on a Macintosh computer.

RESULTS AND DISCUSSION

Thermal Properties

The thermal properties of PLLA with each of the plasticizers at the concentrations are shown in Table 2. The glass transition temperature (T^sub g^) was taken from the maximum of the loss modulus curve (G'') in DMA, since a T^sub g^ inflection was not observed in DSC (because of the high crystallinity of PLLA and its slow crystallization process). It can be seen that the T^sub g^ of all plasticized PLLA were lower than that of the pure PLLA. This was expected, as the primary function of a plasticizer is to disrupt crystallization and lower the T^sub g^. However, the T^sub g^ of PLLA was mostly affected in the presence of the citrate esters, TEC and TBC. These plasticizers showed about the same efficiency at these low concentrations. But from previous work done by Labrecque et al. (12), TBC proved to be one of the most efficient citrate esters when used at higher concentrations in PLLA of lower molecular weight than the one used in this study.

Table 2. Thermal Properties of PLLA and Plasticizers.

The crystallization and melting temperatures (T^sub c^ and T^sub m^, respectively) are also shown in Table 2. It can be seen that both the T^sub m^ and T^sub c^ of the plasticized PLLA were not significantly different from those of the pure PLLA. This could be because a higher quantity of plasticizer may be required to have a noticeable effect on these properties, as the crystallinity of the PLLA was high (0.62). The crystallinity (X^sub c^) of the PLLA samples (calculated from the enthalpy of melting) was slightly reduced in the presence of all the plasticizers.

Table 3 shows the thermal properties of PLLA-flax composites, with each of the plasticizers at the concentrations shown. The overall plasticizer concentrations were almost the same as those studied without fibers (approximately 10% v/v). In a composite, the T^sub g^ of the pure PLLA showed an increase from 71[degrees]C to 77[degrees]C. This could be attributed to a filler effect, where a decrease in molecular mobility of the matrix was observed when the matrix was adsorbed onto the filler. This suggested that good interfacial bonding between the fibers and the matrix exists. The T^sub g^s of the plasticized PLLA composites were decreased more than their counterparts without fibers. This was because the PLLA was expected to be less strongly adsorbed onto the flax when the plasticizer was located at the surface of the fiber.

In a composite, the T^sub c^ of the pure PLLA was increased from 106[degrees]C to 110[degrees]C. This indicated that the fibers acted as a nucleating agent for the PLLA as crystallization proceeded earlier upon cooling from melt. However, the nucleating effect was not observed for the plasticized PLLA composites. This was because the plasticizer was situated at the surface of the fibers, which inhibited the PLLA from nucleating on the fibers. The T^sub m^ of the PLLA was also slightly increased in the composite, which was expected from an increase in nucleation. The T^sub m^, of the plasticized composites was the same as those without fibers.

Table 3. Thermal Properties of Reinforced PLLA-Plasticizer Composites.

The X^sub c^ of pure PLLA was slightly increased in the composite compared with the PLLA in the absence of the fibers (0.66 and 0.62 respectively). This was expected, as the T^sub c^ of the PLLA was increased in the composite so that a greater degree of crystallization took place. The X^sub c^ of the plasticized composites was observed to be lower than that of the pure PLLA composite and the lowest X^sub c^ exhibited by the TBC composite (0.44). From the thermal properties of the composites, one could say that the most efficient plasticizer was TBC, as the X^sub c^, T^sub g^, and T^sub c^ showed the greatest effect compared with the unplasticized PLLA composite.

Dynamic Flexural Properties

The DMA curves for PLLA and with each of the plasticizers are shown in Fig. 1a, b. The storage modulus (G') of PLLA is shown in Fig. 1a. It can be seen that all the plasticized PLLA exhibited lower G' values than the pure PLLA within the temperature range investigated, indicating that the polymer is more flexible upon the addition of the plasticizers. This was consistent with the thermal properties determined by DSC as the T^sub g^ and the X^sub c^ were reduced. The plasticized PLLA with TBC showed the lowest G' as its T^sub g^ and the X^sub c^ were the lowest, while the plasticized PLLA with GTA showed the highest G' with the highest T^sub g^ and X^sub c^, of the plasticized PLLA. This trend suggested that the G' was highly dependent on their T^sub g^ and X^sub c^, and that the most efficient plasticizer for PLLA was TBC. After the glass transition region, the curves all merged towards a similar G^sub '^ of 1 GPa regardless of the type of plasticizer, as the rigidity of PLLA was lost upon heating above its T^sub g^.

The DMA curves of PLLA composites and the plasticized composites are shown in Fig. 2a, b. For all plasticized composites, the G' throughout the temperature range investigated were significantly increased compared with the pure PLLA composite (Fig. 2a). The G' values were at least double that of the pure PLLA composite up to temperatures before the glass transition. This was in contrast to the behavior observed earlier for the PLLA with plasticizers, where the G' decreased in the presence of the plasticizers as their T^sub g^ and X^sub c^ decreased. This could be because the matrix became less brittle as the X^sub c^ was decreased, which created a better surface coverage of the fibers by the matrix and consequently a more homogeneous morphology of the matrix in general. This was evident in the scanning electron microscopy photos (Fig. 3a-d). This was an important factor in controlling the mechanical properties of composites because if the matrix does not adequately cover the fibers, efficient stress transfer from the matrix to the fibers is not obtained, causing a weakening effect rather than a reinforcing one in the presence of the fibers. And if the matrix was brittle, cracks could propagate under stress and result in failure of the matrix.

After the glass transition region, the G' of all plasticized composites were still significantly higher than that of the pure PLLA composite, which was not observed with the unreinforced samples. At higher temperatures than T^sub g^, the reinforcing effect of the fibers was more evident such that the stiffness was still retained to a certain degree, unlike that observed without fibers where the G' all merged to a similar value.

Figures 1b and 2b shows the loss modulus (G'') for PLLA samples and the PLLA composites (with and without plasticizers) respectively. The most noticeable observation was the breadth of the G'' peaks for the composites with plasticizers compared with their counterparts without fibers. The G'' peaks were much broader and this peak broadening of G'' or tan [delta] has also been observed in many studies with fiber composites (10, 17-19). Verghese et al. (19) have credited the G'' or tan [delta] peak broadening to an increase in the breadth of the distribution of relaxation times. The relaxation time was highly affected by the degree of coupling of polymer chains and the level of intermolecular interactions. In our case, the broad peaks indicated that the PLLA molecules existed in different environments. The different environments could arise from some of the PLLA having been adsorbed onto the fibers and therefore exhibiting a higher T^sub g^. Other molecules farther away from the fibers may have more plasticizer, resulting in lower crystallinity and therefore a lower T^sub g^. There are many environments in which the PLLA molecules may reside in the plasticized composites, without suggesting that the cause be due to sample preparation problems. Repeated measurements with similar thermal histories provided the same results.

Scanning Electron Microscopy (SEM)

Figure 3a-d shows scanning electron microscope (SEM) pictures of composite surfaces. In Fig. 3a, the surface of the PLLA composite with no plasticizer shows a very rough topography, with a high incidence of cracks and voids throughout the matrix. This is indicative of brittle behavior, which could result in a composite of lower mechanical properties, as discuss\ed earlier.

Fig. 1. Storage modulus (G') (a) and loss modulus (G'') (b) of PLLA and plasticizers.

Figures 3b-d show the PLLA composites with plasticizers. Generally, these photos show a more ductile matrix as indicated by the smooth topography. The occurrence of cracks and voids was also observed to be much less than in the unplasticized PLLA composite. Of all the plasticized composites, the samples with GTA and TEC showed the smoothest surfaces with excellent dispersion of the fibers within the matrix. This could account for the higher G' observed in DMA compared with pure PLLA and with TBC. The composite with TBC showed an improvement compared with the pure PLLA composite, but some uneven texture and cracks still exist, which indicated some degree of brittleness. This explained the lower G' observed in DMA compared with the other plasticized composites. Optical Microscopy (OM)

Fig. 2. Storage modulus (G') (a) and loss modulus (G'') (b) of PLLA-plasticizers composites.

Figure 4a-c shows some hot stage optical microscopy pictures for the PLLA-plasticizer-flax system. The specimen was prepared as to show one fiber so that the crystallization of the PLLA could be observed adjacent to and farther away from the fiber. The pictures were taken with crossed polarizers after cooling on the hot stage from above the melting temperature.

Figure 4a shows the PLLA-flax system. The crystals were of different sizes and of mixed perfections. No transcrystallinity was observed along the fiber although intimate contact between the fiber and the matrix was observed. Figure 4b shows the PLLA-flax system with GTA. The crystals were smaller than those in Fig. 4a, which may account for the decrease in X^sup c^. The large number of smaller crystals also suggests that more nuclei were formed in the presence of the GTA. The contact between the fibers and the matrix was also good as no gaps between the fibers and the matrix were observed. Figure 4c shows the PLLA-flax system with TBC. The crystals were large, typical spherulites, which were very different from those observed with no plasticizer and with GTA. They were of similar size and had similar growth rates, as indicated by the straight interfaces between the spherulites. The nuclei were fewer in number than observed with GTA. On close inspection, the contact between the fibers and the matrix was limited, as gaps along the fiber and the crystal face were observed in the lower right of the picture. This may be because some of the plasticizer was at the surface of the fiber, which inhibited the nucleation at the fiber surface.

Fig. 3a. SEM photograph of PLLA composite.

Fig. 3b. SEM reinforced PLLA-TBC composite.

CONCLUSIONS

Flax fibers have been modified by removal of moisture and absorption of selected plasticizers onto or into the fibers. The presence of the plasticizer enhanced the storage modulus (G') of the composites, which indicated an increase in stiffness. The improvement in G' was at least double, depending on the plasticizer. This was because the matrix was less brittle, thereby creating a better surface coverage of the matrix on the fibers. The thermal properties of these composites were also greatly affected when the plasticizers were present. The citrate esters showed the most influence on the PLLA, but TBC proved to be the most efficient plasticizer for the matrix. The T^sub g^, T^sub c^, and X^sub c^ were all reduced in the presence of the plasticizers either with or without the fibers.

Fig. 3c. SEM reinforced PLLA-TEC composite.

Fig. 3d. SEM reinforced PLLA-GTA composite.

ACKNOWLEDGMENTS

The assistance of Cooperative Research Centre for Polymers (CRC- P) in providing a postgraduate scholarship for Susan Wong is acknowledged.

Fig. 4a. OM photograph of PLLA composite.

Fig. 4b. OM photograph of PLLA-GTA composite.

Fig. 4c. OM photograph of PLLA-TBC composite.

REFERENCES

1. J. Jiang, K. Okubo, and T. Fujii, 31st International SAMPE Technical Conference, 31, 355-366 (1999).

2. A. Amash and P. Zugenmaiser, Polymer, 41, 1589-1596 (2000).

3. S. Mishra, M. Misra, S. S. Tripathy, S. K. Nayak, and A. K. Mohanty, Macromolecular Materials and Engineering, 286, 107-113 (2001).

4. A. K. Mohanty, M. Misra, and G. Hinrichsen, Macromolecular Materials and Engineering. 276/277, 1-24 (2000).

5. D. N. Saheb and J. P. Jog, Advances in Polymer Technology, 18, 351-363 (1999).

6. U. Riedel and J. Nickel, Die Angewandte Makromolekulare Chemie, 272, 34-40 (1999).

7. W. D. Brouwer, SAMPE Journal 36, 19 (2000).

8. K. Van de Velde and E. Baetens, Macromolecular Materials and Engineering, 286, 342 (2001).

9. S. C. Tjong, Y. Xu, and Y. Z. Meng, Journal of Applied Polumer Science, 72, 1647 (1999).

10. A. K. Saha, S. Das, D. Bhatla, and B. C. Mitra, Journal of Applied Polymer Science, 71, 1509.

11. R. E. Drumright, P. R. Gruber, and D. E. Henton, Advanced Materials, 12, 1841 (2000), 1510 (1999).

12. L. V. Labrecque, R. A. Kumar, V. Dave, R. A. Gross, and S. P. McCarthy, Journal of Applied Polymer Science, 66, 1507 (1997).

13. O. Martin and L. Averous, Polymer, 42, 6209-6219 (2001).

14. A. G. Andreopoulos, Clinical Materials, 15, 89 (1994).

15. M. Sheth, R. A. Kumar, V. Dave, R. A. Gross, and S. P. McCarthy, Journai of Applied Polymer Science, 66, 1495-1505 (1997).

16. S. Jacobsen and H. G. Fritz, Polymer Engineering and Science, 39, 1303 (1999).

17. A. K. Rana, B. C. Mitra, and A. N. Banerjee, Journal of Applied Polymer Science, 71, 537 (1999).

18. M. A. Lopez-Manchado and M. Arroy, Polymer, 41, 7761- 7767(2000).

19. K. N. E. Verghese, R. E. Jensen, J. J. Lesko, and T. C. Ward, Polymer, 42, 1634 (2001).

SUSAN WONG, ROBERT A. SHANKS*, and ALMA HODZIC

CRC for Polymers, Applied Chemistry, RMIT University GPO Box 2476V, Melbourne, 3001, Australia

*Corresponding author. E-mail address: polymer@rmit.edu.au

Copyright Society of Plastics Engineers Sep 2003

More News in this Category