Effect of adding wood flour to the physical properties of a biodegradable polymer

Effect of adding wood flour to the physical properties of a biodegradable polymer

Available online at www.sciencedirect.com Composites: Part A 39 (2008) 503–513 www.elsevier.com/locate/compositesa Effect of adding wood flour to the ...

1MB Sizes 3 Downloads 38 Views

Available online at www.sciencedirect.com

Composites: Part A 39 (2008) 503–513 www.elsevier.com/locate/compositesa

Effect of adding wood flour to the physical properties of a biodegradable polymer M. Morreale, R. Scaffaro *, A. Maio, F.P. La Mantia Universita` di Palermo, Dipartimento di Ingegneria Chimica dei Processi e dei Materiali, Viale delle Scienze, 90128 Palermo, Italy Received 23 July 2007; received in revised form 13 September 2007; accepted 12 December 2007

Abstract Wood flour/polymer composites (WPC) gained a significant interest during the last decades, due to several advantages related to the use of a natural-organic filler rather than an inorganic-mineral one. However, most of the studies have been performed on composites based on polyolefin matrices. A further step is the use of biodegradable polymers instead of traditional ones. In this work, wood flour (WF), under the form of short fibers, with two different sizes (coarse and fine) was added to a corn starch based biodegradable polymer of the Mater-BiÒ family. The effect of WF size, WF content, thermal treatment on the mechanical properties was investigated. The tensile mechanical tests showed an increase of rigidity of the composites upon increasing the WF content, together with a sharp decrease of the elongation at break. With regard to the tensile strength, no remarkable differences were observed upon changing wood flour size or type. In both cases, on average, there were often slightly better results with the samples which had underwent a drying pre-treatment. The increased rigidity was confirmed also by the impact tests, even though it decreased upon increasing the WF content. The heat deflection temperature followed the same trend as the elastic modulus. The immersion tests suggested that these materials are not suitable to prolonged contact with water. Humidity absorption tests revealed that the matrix plays a fundamental role in humidity absorption. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: A. Wood; A. Polymer–matrix composites (PMCs); B. Mechanical properties; D. Electron microscopy

1. Introduction Wood flour filled polymer composites attracted interest both in academia and in industry in the 1980s, because of several advantages they can grant if compared with the then dominating mineral filler-polymer composites. These advantages regard the low cost of wood based fillers (which usually come from sawmill wastes), the lower specific weight, the reduced hazards for production workers in case of inhalation, the special aesthetic features (these composites can look like some kinds of wood) and several environmental issues. In particular, all the natural-organic fillers are biodegradable, and allow reducing the use of non-biodegradable plastic materials. All these characteristics have *

Corresponding author. Tel.: +39 0 916567223; fax: +39 0 916567280. E-mail address: scaff[email protected] (R. Scaffaro).

1359-835X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.compositesa.2007.12.002

an overall positive impact regarding the reduction in the use of non-renewable sources throughout the whole lifecycle of the composite [1–8]. Scientific literature reports studies regarding polymer matrices (polyolefins in prevalence) in combination with several different natural-organic fillers (wood fibers and flour, maize, tapioca, sago, rice starch, sisal fibers, kenaf fibers, olive stone flour, etc.) [3,9–18]. However, a limit of these composites is that there is not a full biodegradability: this, in fact, regards only the filler, therefore the environmental performance is lower than expected. A possible solution to overcome this limit is to replace the traditional, non-biodegradable polymer matrices (typically polyolefins, as already highlighted) with biodegradable ones. The biodegradable polymers can be synthetic (polyesters, polyester amides, polyvinyl alcohol) or coming from renewable sources (i.e., polymers derived from vegetable sources or

504

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

wastes). This latter class includes, in particular, starch and starch derivatives, polylactic acid (PLA), polyhydroxyalcanoates (PHA), cellulose, etc. [19–21]. Possible applications for these materials can be panels for automotive interior, indoor furnishing, packaging and, in general, to replace polyolefins [22–24]. Starch, in particular, is an especially interesting biopolymer. It is a polysaccharide extracted from several natural sources (corn in prevalence). Its biodegradability comes mainly from the presence of oxygen in its consecutive-ring structure. Starch is a thermoplastic material which can be processed by the typical polymer-processing techniques. It reacts with water, degrading via hydrolysis. However, it is usually modified for the purpose: Novamont MaterBiÒ, for instance, comes from corn starch which undergoes a ‘‘complexation” process, by means of biodegradable complexing agents which aim is to create several molecular superstructures, depending on the desired final properties of the product. The scientific literature does not report many studies on biocomposites based on Mater-Bi. The ‘‘Y” grade was investigated in combination with sisal [25,26], flax [27] and myscanthus [28,29], while the ‘‘Z” grade was studied in conjunction with flax-derived cellulose pulp and sisal [21,30]. With regard to the Mater-Bi Y/sisal biocomposites, biodegradability tests [26], mechanical [25] and rheological [31] characterization studies have been performed. On the whole, it was observed that elastic modulus and impact strength increase, and that the sisal fibers play a substantially marginal role with concern to degradation and biodegradation phenomena, since the starch polymer matrix is more susceptible to them. In addition, water sorption was reduced too. Accurate analyses were also carried out on the impact behavior of Mater-Bi Y/myscanthus composites. In this case, the most important processing variable is the temperature, followed by the processing speed and the filler size [28,29]. The behavior of Mater-Bi Y and Z with flax cellulose pulp [27] reflected, as predictable, a dramatic increase of elastic modulus, upon increasing the filler content. Rather surprisingly, tensile strength increased as well, while it was observed decreasing for other similar systems [17,32]. In this work, the mechanical, thermomechanical and hygroscopic behavior of Mater-Bi–wood flour biocomposites were investigated. The Mater-Bi used for this purpose is a commercial N grade, and it is particularly interesting since few literature studies are available on it. The influence of several processing parameters (WF size, WF content, humidity content of the components) was assessed.

starch-based fraction and a synthetic biodegradable polyester. More details on the properties of the matrix are reported in Table 1. The wood flour used in this work was kindly supplied by LA.SO.LE. (Italy) in two different types: the ‘‘35” (average particle diameter 350–500 lm, average L/D = 3.9) and ‘‘150/200” (average particle diameter 150–200 lm, L/ D = 2.8). Throughout the work, these are indicated respectively as ‘‘SDC” and ‘‘SDF”. 2.2. Preparation and processing The investigated materials were mixed in a Brabender PLE330 batch mixer, thus preparing several blends with WF content ranging from 15% to 60% (by weight). The materials underwent several pre-treatment procedures before compounding. In particular, the wood flour was always thermally treated before mixing, by drying it in an oven at 70 °C overnight. Some samples were prepared with the polymer without any treatment (‘‘humid”), while other (‘‘dry”) were characterized by a thermal treatment on the polymer prior to processing, in order to reduce its amount of water. This treatment was chosen in order to investigate the effects of humidity on the properties of the materials, since water can typically cause hydrolytic chain scission reactions of polyesters during processing. The treatment was performed in a vacuum oven at 70 °C overnight, plus one hour at 90 °C. The mixing conditions were: temperature 160 °C, rotating speed 60 rpm, mixing time 4 min. After mixing, the blends were compression-molded in a Carver (USA) laboratory press at 180 °C for 2–3 min in order to obtain the specimens for tensile, impact, and heat deflection temperature (HDT) tests. 2.3. Characterization Tensile tests according to ASTM D882 were performed by an Instron (USA) mod. 3365 universal machine on specimens (thickness 0.5–1.4 mm, width = 10 mm) cut off from compression-molded sheets, with a crosshead speed of 5 mm/min and 30 mm initial length. At least 20 replicates were tested. Impact tests according to ASTM D256 were carried out by a Ceast (Italy) mod. 6545 universal apparatus on notched samples in the Izod mode. At least 15 replicates were tested. Heat Deflection Temperature (HDT) tests (at least five replicates) according to ASTM D2990 were performed by means of a Ceast (Italy)

2. Experimental Table 1 Properties of the raw polymer

2.1. Materials Ò

The Mater-Bi used in this work is a commercial ‘‘N” grade, produced by Novamont. The chemical composition is unknown but, according to the producer, it contains a

Property

Value

Melting temperature [°C] Melt flow index [g/10 min] (at T = 150 °C, load = 5 kg) Density [g/cm3]

110 3 1.3

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

mod.6505 apparatus, with a temperature ramp of 120 °C/h. All the results were processed by calculating the average values and taking into account the data scattering and then reporting the related error bars. The extent to which the material tends to undergo dimensional and weight variations when exposed to environmental humidity were analyzed by water and humidity uptake tests. Water uptake test were performed by putting the tensile specimens into a water bath at a fixed temperature of 30 °C (by means of a ventilated oven), while humidity uptake tests were carried out in a Binder (USA) climatic chamber set at 30 °C and RH = 80% on tensile specimens. Weight and thickness increase due to absorbed water were monitored through periodical measuring. The samples were taken out from the water bath or the controlled humidity chamber, dried by blotting paper (if necessary), weighed (precision 104 g), the thickness (always in the same, previously marked point) measured, and the samples then put again in the test environment. The water/humidity uptake (at the generic time t) was thus monitored by measuring the weight increase and the thickness swelling through the following Eqs. (1) and (2): DW ¼ ðW t  W 0 Þ=W 0  100 DTH ¼ ðTHt  TH0 Þ=TH0  100

ð1Þ ð2Þ

where W0 is the weight of dry specimen, TH0 its thickness, Wt and THt the weight and thickness of the wet sample at t time. These quantities where then normalized, dividing them, respectively, by W0 and TH0. The possible solubilization of starch was checked by a titration carried out at the end of the immersion tests, with a procedure commonly used in food science. In the immersion test bath, a few drops of tincture of iodine (red) were poured. A colour change of the bath to dark purple would, in turn, reveal the presence of starch. Finally, in order to get a deeper understanding of some phenomena emerged during the characterization, SEM analysis of specimens was carried out by a Philips (The Netherlands) ESEM XL30 equipment. 2.4. Calculation of the mixing index In order to evaluate the change in the dispersion of the WF before (i.e., directly coming out of the batch mixer) and after compression molding, chunks of some blends were first solubilized in an appropriate solvent and therefore filtered to recover the WF fraction. Several measurements allowed calculating the related weight fraction of WF for each observation. These results have been therefore treated to calculate the mixing index (MI) according to Eq. (3) [33,34]: MI ¼ 1 

s s0

where s is the square root of the variance

ð3Þ

s2 ¼

505

N X 1 2 ðci  cÞ ðN  1Þ i¼1

ð4Þ

where N is the number of measurements (7 in our case), ci is the concentration of the ith observation and c is the mean concentration calculated following Eq. (5): c ¼

N 1 X ci N i¼1

ð5Þ

3. Results and discussion In order to study how the processing influenced the aspect ratio of the WF particles, an analysis was carried out on the WF extracted from some samples using an appropriate solvent. In Table 2, the values of length-todiameter ratio are reported for the two neat WF samples and for 30% and 60% WF containing composites (humid). As expected, after processing there is a reduction of the aspect ratio due to the breakup of the WF fibers during processing. The phenomenon is more intense when a higher content of wood flour is used even if the major reduction occurs at low concentration of fibers. This can be explained hypothesizing that, at lower concentration, the fibers are likely well wetted and the stresses during processing are easily transmitted to the fiber causing their breakup. Upon increasing the concentration, physical networks of fiber agglomerates, as observed by solubility tests, are present in the blend and the wetting of the fibers is consequently drastically reduced. In this case, the stress transmission is less efficient and the breakup levels off. This hypothesis is confirmed by the mixing torque values (not reported here). At 60% of wood flour the mixing torque is lower than that observed for the 30% composite: a worse dispersion and the segregation of the fibers can be invoked to explain this behavior as commented further below. Fig. 1 shows the elastic modulus as a function of the WF content for all the investigated materials. The experimental results show that the WF imparted a significant improvement of the elastic modulus, increasing upon increasing the WF content, while the treatment, on average, did not significantly affect this property. On the other hand, SDF filled samples showed to be, on average, significantly more rigid than the SDC filled ones, especially at high WF content. This may seem in contrast with the

Table 2 Average L/D ratios for neat and extracted WF Sample

L/D ratio (average)

SDC SDF 30SDC 30SDF 60SDC 60SDF

3.9 2.8 3.5 2.4 2.5 2.3

506

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

Fig. 1. Elastic modulus as a function of WF content.

values of torque registered during the processing (not reported here) showing that the SDF containing blends display lower mixing torques with respect to the SDC containing ones. This occurrence would suggest, in fact, a worse WF dispersion when SDF is used instead of SDC. This apparent discrepancy could be interpreted considering the presence of flow-induced dispersion phenomena during the compression molding step, which improve the dispersion of the WF within the thermoplastic matrix. To have a direct confirmation of this statement, measurements of the mixing index, according to the procedure reported in the experimental section have been reported in Table 3. The results indicate that the mixing index increases for SDF filled materials when they are compression molded confirming that this further operation contributes to improve the dispersion of the WF. This can be also verified by observing Fig. 2a–b, which show the SEM micrographs of the same SDF filled sample (60 wt%) before and after compression molding, respectively. The surface is clearly smoother and more compact in the second case (compression-molded material), and this can therefore explain the higher rigidity. Differently from SDF containing blends, and according to the mechanical tests results, the mixing index for the SDC containing blends, see Table 3, remains almost

Table 3 Mixing index of selected samples Sample

MI

30SDF mixed 30SDF compression-molded 30SDC mixed 30SDC compression-molded 60SDF mixed 60SDF compression-molded 60SDC mixed 60SDC compression-molded

0.958 0.994 0.995 0.983 0.915 0.983 0.923 0.927

unchanged. This results eventually confirm the above considerations about the relationships between flow-induced phenomena during the compression molding operation. Fig. 3 reports the tensile strength as a function of the WF content for all the materials. It can be observed that the tensile strength keeps practically constant upon varying all the operative conditions. Even if the two treatments give similar results in terms of ultimate resistance level, it is worth noting that, beyond the deviation, the dried sample showed higher values of this property with respect to the humid samples. This occurrence can be explained considering that polycondensation polymers endure, during processing, hydrolytic chain scission reactions if water is present and, as a result, the molecular weight rapidly decreases. The addition of wood flour did not cause, in this case, a drop of tensile strength, in contrast to the observations of other studies regarding wood flour filled thermoplastic polymers [17,32,35,36]. On the other hand, the Mater-Bi/wood flour system showed better results probably because of a better interfacial adhesion between the WF and the matrix, both of them being bearing polar groups on their molecules. The SEM micrographs of materials prepared with dry and humid Mater-Bi are reported, respectively, in Fig. 4a–b. In Fig. 4b, it is possible to see isolated fibers, badly embedded, coming out of the fracture surface of the material as highlighted by the white ellipse put on the micrograph. On the contrary, the material prepared with the dry treated polymer, Fig. 4a, shows a smoother and more uniform fracture surface. The elongation at break as a function of WF content is plotted in Fig. 5. The dramatic drop of the elongation at break is related to the stiffening effect of wood flour. The behavior is practically the same for all the investigated cases, in agreement with previous studies [35]. Figs. 6 and 7 report the values of impact strength tests for humid and dry samples, respectively. The unfilled samples did not break under the previously specified test conditions.

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

Fig. 2. SEM micrographs of a 60 wt% SDF filled sample before (a) and after (b) compression molding.

Fig. 3. Tensile strength as a function of WF content.

507

508

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

Fig. 4. SEM micrographs of fracture surfaces of a dry (a) and a humid (b) sample.

The Mater-Bi is ductile and contains hydrophilic groups, which should grant a good matrix–WF interfacial adhesion and therefore a relatively high value of impact strength even in the composites. However, a reduction of the impact strength in the composite was somehow expected because of the increased stiffness and because of the particle agglomeration, which cause easier crack propagation. The ductility of the material rapidly decreases upon increasing the WF content over 30 wt%. The effect of treatment is significant, and indeed is detrimental. This is probably related to the loss of plasticizing effect given by water, while the humid samples are more ductile. This would be in agreement also with the previous observations regarding the elongation at break. Furthermore, it was noted that the tests had a greater data scattering with WF contents above the 30% (as highlighted by the error bars). It was necessary to test at least

fifteen specimens for each material in order obtaining sufficiently reliable data. This can be attributed to the intrinsic variations (lignin, cellulose and hemi cellulose variable contents) of lignocellulosic fibers [28]. Singleton et al. [37] investigated a HDPE/flax composite and observed, in the impact strength data, fluctuations up to 175%. A study [38] on a traditional composite (epoxy/carbon fiber) showed a fluctuation of only 27.3%. Fig. 8 reports the values of HDT as a function of WF content for humid and dry samples. HDT increases upon increasing the WF content. With regard to WF size, it can be observed that the SDF filled composites have a higher HDT than the SDC filled ones, however the differences are small and, at higher WF concentrations, these are negligible, in agreement with the trends already observed for the elastic modulus. It is likely that the enhanced rigidity of SDF composites arises from the fact

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

509

Fig. 5. Elongation at break as a function of WF content.

Fig. 6. Impact strength as a function of WF content, untreated samples.

that fine size flour has a better capability to immobilize the polymer chains, because of the higher filling degree it can grant. In general, there are not great differences between the dry and the humid samples, especially at low concentrations: the differences between the treatments are substantially small. This is not true at higher WF levels. Upon increasing WF concentration, also the data scattering increases and slightly higher values of HDT can be observed for humid samples. Some of these results could be therefore related (based also on the visual analysis of the samples and the trend observed during morphological analyses) to the formation of a ‘‘network” of poorly wetted, unified fibers, which stiffen the material and may alter the results. Finally, it cannot be excluded that some data

may have been affected by experimental errors due to specimen defects especially at higher filled loadings. Figs. 9 and 10 report the results of water uptake (immersion) tests. In particular, Fig. 9 shows the trends of specimens thickness variations, while Fig. 10 the changes in specimen weight, each curve representing a definite sample. The labels can be interpreted by referring to Table 4. By observing Figs. 9 and 10, it can be stated that, during the first 5 hours, the trend is roughly the same for all the samples, related to the absorption of water. Longer times highlight, on the other hand, different behaviors depending on the material: D, F, G, and H reach a typical saturation level, while the other samples show a weight decrease, due to the action of water, which clearly damaged those sam-

510

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

Fig. 7. Impact strength as a function of WF content, treated samples.

Fig. 8. Heat deflection temperature as a function of WF content.

20

(delta Th)/Tho, %

18

A

16

B C

14

D

12

E F

10

G H

8 6 4 2 0 0

5

10

15

20

Time, h Fig. 9. Thickness variation during immersion tests.

25

30

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

511

20 18 A

(delta W)/Wo, %

16

B

14

C D

12

E

10

F G

8

H

6 4 2 0 0

5

10

15

20

25

30

Time, h Fig. 10. Weight variation during immersion tests.

Table 4 Materials investigated by water and humidity uptake tests Label

WF (wt%)

WF size

Treatment

A B C D E F G H

15 30 15 30 15 30 15 30

SDF SDF SDC SDC SDF SDF SDC SDC

Humid Humid Humid Humid Dry Dry Dry Dry

ples and was testified by cracks and loss of fibers in the water bath. Nevertheless, it can be stated that this weight loss is not ascribable to a starch fraction solubilization, since the titration did not reveal starch traces in the water used for the immersion tests. However, the overall results suggest that the use of these materials in direct and prolonged contact with water is not recommended.

Figs. 11 and 12 show the trends of humidity uptake tests. Also in this case, it is possible to observe weight and thickness increases for all the samples during the first five hours, while at higher times saturation conditions are substantially achieved. The polymer itself proved to be hydrophilic, as shown by the absorption curves (‘‘NH” indicating the humid pure matrix, ‘‘ND” the dry one), in agreement with other works [39] where the Mater-Bi ‘‘Y” proved to absorb relatively high moisture contents. Mater-Bi ‘‘N” seems, however, to be less susceptible to moisture absorption. On average, it can be observed that the pre-treated blends (‘‘dry”) absorb more water than the non-treated (‘‘humid”) ones, and in particular there are noticeable absorption phenomena involving the 15 wt% filled samples. Furthermore, similar conclusions can be drawn by examining the absorption values of the humid samples. This seems to indicate, therefore, that the polymer matrix plays an important role in the absorption process. No clear correlation was found between WF size and humidity uptake. However, weight variations are more

Fig. 11. Thickness variation during absorption tests.

512

M. Morreale et al. / Composites: Part A 39 (2008) 503–513

Fig. 12. Weight variation during absorption tests.

reproducible than thickness variations due, of course, to lower experimental errors in the measurements, thus being more reliable.

for indoor applications. Humidity absorption tests revealed that the matrix plays a vital role in humidity absorption. Acknowledgement

4. Conclusions In this work, wood flour with two different granulometries was added to a corn starch based biodegradable polymer. The tensile mechanical tests indicated that there is an increase of rigidity of the composites upon increasing the WF content. In particular, a significant increase of the elastic modulus was observed, together with a sharp decrease of the elongation at break. It is worth mentioning that slightly higher values of the modulus were observed for the SDF containing blends, with some more evident effect if a drying was performed before processing. The tensile stress seems not to be particularly affected both from the size and from the amount of wood flour. On the whole, slightly higher values were displayed by the pre-treated (dried) samples, beyond the scattering measured in the tests. The increased rigidity was confirmed by the impact tests. In all the cases, this property decreased on increasing the WF content. The highest values were observed in the not pre-treated samples and this may be explained considering that a higher amount of water can act as a plasticizer causing the increase of this property. The heat deflection temperature followed a trend similar to that of the elastic modulus. At high WF content, there was a dramatic increase of the data scattering suggesting that the bad dispersion of the WF is the key-factor that control the mechanical and the thermomechanical performance. The immersion tests did not reveal any solubilization of starch but, conversely, a progressive disgregation of several specimens due to the action of water. This latter result suggests that these materials are not appropriate for prolonged contact with water and, therefore, they are more suitable

The authors are grateful to Novamont for providing the Mater-Bi used in this work and for technical support. References [1] Rozman HD, Lai CY, Ismail H, Mohd Ishak ZA. Effect of coupling agents on the mechanical and physical properties of oil palm empty fruit bunch-polypropylene composites. Polym Int 2000;49:1273–8. [2] Gachter R, Muller H. Plastics additives. third ed. Munich: Hanser Publishers; 1990. [3] Joseph K, Thomas S, Pavithran C. Effect of chemical treatment on the tensile properties of short sisal fibre-reinforced polyethylene composites. Polymer 1996;37:5139–49. [4] Joseph PV, Joseph K, Thomas S. Effect of processing variables on the mechanical properties of sisal-fiber-reinforced polypropylene composites. Comp Sci Tech 1999;59:1625–40. [5] Canche`-Escamilla G, Rodriguez-Laviada J, Cauich-Cupul JI, Mendizabal E, Puig JE, Herrera-Franco PJ. Flexural, impact and compressive properties of a rigid-thermoplastic matrix/cellulose fiber reinforced composites. Comp Pt A 2002;33:539–49. [6] Nair KCM, Kumar RP, Thomas S, Schit SC, Ramamurthy K. Rheological behavior of short sisal fiber-reinforced polystyrene composites. Comp Pt A 2000;31:1231–40. [7] Joshi SV, Drzal LT, Mohanty AK, Arora S. Are natural fiber composites environmentally superior to glass fiber reinforced composites?. Comp Pt A 2004;35:371–6. [8] Coutinho FMB, Costa THS, Carvalho DL. Polypropylene–wood fiber composites: effect of treatment and mixing conditions on mechanical properties. J Appl Polym Sci 1997;65:1227–35. [9] Simpson RJ, Selke S. In: Andrews G, Subramamian P, editors. Emerging technologies in plastic recycling. Washington: ACS Symp Series 513, 1992 [Chapter 18]. [10] Balasuriya PW, Ye L, Mai YW, Wu J. Mechanical properties of wood flake–polyethylene composites II. Interface modification. J Appl Polym Sci 2002;83:2505–21. [11] Raj RB, Kokta VB, Maldas D, Daneault C. Use of wood fibers in thermoplastics. VII. The effect of coupling agents in polyethylene– wood fiber composites. J Appl Polym Sci 1989;37:1089–103.

M. Morreale et al. / Composites: Part A 39 (2008) 503–513 [12] Chen HL, Porter RS. Composite of polyethylene and kenaf, a natural cellulose fiber. J Appl Polym Sci 1994;54:1781–3. [13] Wool RP, Raghavan D, Wagner GC, Billieux S. Biodegradation dynamics of polymer–starch composites. J Appl Polym Sci 2000;77:1643–57. [14] Sharma N, Chang LP, Chu YL, Ismail H, Mohd Ishak ZA. A study on the effect of pro-oxidant on the thermo-oxidative degradation behaviour of sago starch filled polyethylene. Polym Degrad Stab 2001;71:381–93. [15] Danjaji ID, Nawang R, Ishiaku US, Ismail H, Mohd Ishak ZA. Sago starch-filled linear low-density polyethylene (LLDPE) films: Their mechanical properties and water absorption. J Appl Polym Sci 2000;79:29–37. [16] N. Tzankova Dintcheva. Recycling of post-consume plastics. PhD thesis. University of Palermo, Department of Chemical Engineering, 2000. [17] La Mantia FP, Tzankova Dintcheva N, Morreale M, Vaca-Garcia C. Green composites of organic materials and recycled post-consumer polyethylene. Polym Int 2004;53:1888–91. [18] Nitz H, Semke H, Mulhaupt R. Reactive extrusion of polycaprolactone compounds containing wood flour and lignin. J Appl Polym Sci 2001;81:1972–84. [19] Nishino T, Hirao K, Kotera M, Nakamae K, Inagaki H. Kenaf reinforced biodegradable composite. Comp Sci Tech 2003;63:1281–6. [20] Imam SH, Cinelli P, Gordon SH, Chiellini E. Characterization of biodegradable composite films prepared from blends of polyvinyl alcohol, cornstarch, and lignocellulosic fiber. J Polym Environ 2005;13:47–55. [21] Di Franco CR, Cyras VP, Busalmen JP, Ruseckaite RA, Vazquez A. Degradation of polycaprolactone/starch blends and composites with sisal fibre. Polym Degrad Stab 2004;86:95–103. [22] Netravali AN, Chabba S. Composites get greener. Mater Tod 2003;6:22–6. [23] Marsh G. Next step fort automotive materials. Mater Tod 2003;6:36–43. [24] Nickel J, Riedel U. Activities in biocomposites. Mater Tod 2003;6:44–8. [25] Alvarez VA, Kenny JM, Vazquez A. Creep behavior of biocomposites based on sisal fiber reinforced cellulose derivatives/starch blends. Polym Comp 2004;25:280–8.

513

[26] Alvarez VA, Ruseckaite RA, Vazquez A. Degradation of sisal fibre/ Mater Bi–Y biocomposites buried in soil. Polym Degrad Stab 2006;91:3156–62. [27] Puglia D, Tomassucci A, Kenny JM. Processing, properties and stability of biodegradable composites based on Mater-BiÒ and cellulose fibres. Polym Adv Tech 2003;14:749–56. [28] Johnson M, Tucker N, Barnes S, Kirwan K. Improvement of the impact performance of a starch based biopolymer via the incorporation of Miscanthus giganteus fibres. Ind Cro Prod 2005;22:175–86. [29] Johnson M, Tucker N, Barnes S. Impact performance of Miscanthus/ Novamont Mater-BiÒ biocomposites. Polym Test 2003;22:209–15. [30] Iannace S, Ali R, Nicolais L. Effect of processing conditions on dimensions of sisal fibers in thermoplastic biodegradable composites. J Appl Polym Sci 2001;79:1084–91. [31] Alvarez VA, Terenzi A, Kenny JM, Vazquez A. Melt rheological behavior of starch-based matrix composites reinforced with short sisal fibers. Polym Eng Sci 2004;44:1907–14. [32] La Mantia FP, Morreale M, Mohd Ishak ZA. Processing and mechanical properties of organic filler-polypropylene composites. J Appl Polym Sci 2005;96:1906–13. [33] Kalyon DM, Lawal A, Yazici R, Yaras P, Railkar S. Mathematical modelling and experimental studies of twin-screw extrusion of filled polymers. Polym Eng Sci 1999;39:1139–51. [34] Kalyon DM, Birinci E, Yazici R, Karuv B, Walsh S. Electrical properties of composites as affected by the degree of mixedness of the conductive filler in the polymer matrix. Polym Eng Sci 2002;42:1609–17. [35] La Mantia FP, Morreale M. Mechanical properties of recycled polyethylene ecocomposites filled with natural organic fillers. Polym Eng Sci 2006;46:1131–9. [36] Li XH, Meng YZ, Zhu Q, Tjong SC. Thermal decomposition characteristics of poly(propylene carbonate) using TG/IR and Py– GC/MS techniques. Polym Degrad Stab 2003;81:157–65. [37] Singleton ACN, Baillie CA, Beaumont PWR, Peijs T. On the mechanical properties, deformation and fracture of a natural fibre/ recycled polymer composite. Comp Pt B 2003;34:519–26. [38] Hallett SR. Three-point beam impact tests on T300/914 carbon–fibre composites. Comp Sci Tech 2000;60:115–24. [39] Alvarez VA, Fraga AN, Vazquez A. Effects of the moisture and fiber content on the mechanical properties of biodegradable polymer–sisal fiber biocomposites. J Appl Polym Sci 2004;91:4007–16.