FRACTURE MECHANICS AND STATISTICAL MODELING OF TERNARY BLENDS OF POLYLACTIDE/ETHYLENE-ACRYLATE COPOLYMER /WOOD-FLOUR COMPOSITES By Kojo Agyapong Afrifah A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Forestry 2012 ABSTRACT FRACTURE MECHANICS AND STATISTICAL MODELING OF TERNARY BLENDS OF POLYLACTIDE/ETHYLENE-ACRYLATE COPOLYMER /WOOD-FLOUR COMPOSITES By Kojo Agyapong Afrifah This study examined the mechanisms of toughening the brittle bio-based poly(lactic acid) (PLA) with a biodegradable rubbery impact modifier to develop biodegradable and cost effective PLA/wood-flour composites with improved impact strength, toughness, high ductility, and flexibility. Semicrystalline and amorphous PLA grades were impact modified by melt blending with an ethylene-acrylate copolymer (EAC) impact modifier. EAC content was varied to study the effectiveness and efficiency of the impact modifier in toughening the semicrystalline and amorphous grades of the PLA. Impact strength was used to assess the effectiveness and efficiency of the EAC in toughening the blends, whereas the toughening mechanisms were determined with the phase morphologies and the miscibilities of the blends. Subsequent tensile property analyses were performed on the most efficiently toughened PLA grade. Composites were made from PLA, wood flour of various particle sizes, and EAC. Using two-level factorial design the interaction between wood flour content, wood flour particle size, and EAC content and its effect on the mechanical properties of the PLA/wood-flour composites was statistically studied. Numerical optimization was also performed to statistically model and optimize material compositions to attain mechanical properties for the PLA/wood-flour composites equivalent to at least those of unfilled PLA. The J-integral method of fracture mechanics was applied to assess the crack initiation (Jin) and complete fracture (Jf) energies of the composites to account for imperfections in the composites and generate data useful for engineering designs. Morphologies of the fractured surfaces of the composites were analyzed to elucidate the failure and toughening mechanisms of the composites. The EAC impact modifier effectively improved the impact strength of the PLA/EAC blends, regardless of the PLA type. However, the EAC was more efficient in the semicrystalline grades of PLA compared to the amorphous grade. The semicrystalline blends showed decreased tensile strength and modulus with increased impact modifier content. In contrast, the ductility, elongation at break, and energy to break increased significantly. Mechanisms of toughening of PLA with EAC included impact modifier debonding, fibrillization, crack bridging and matrix shear yielding resulting in a ductile behavior. Increasing the EAC content in PLA/wood-flour composites enhanced the impact strength and elongation at break, but reduced the tensile modulus and strength of the composites. Composites with fine wood particles showed greater improvement in elongation at break than those with coarse particles; an opposite trend was observed for impact strength, tensile modulus and tensile strength. Numerical optimization produced two scenarios based on materials compositions to produce composites with similar mechanical properties as unfilled PLA. These optimization solutions were successfully validated experimentally. The crack initiation (Jin) and complete fracture (Jf) energies of unmodified PLA/wood-flour composites showed the deleterious effect of wood fiber incorporation into the plastic matrix by significantly decreasing the fracture toughness of PLA as the wood flour content increased. By contrast, impact modification of wood plastic composites with EAC significantly increased both the resistance to crack initiation (Jin) and complete fracture (Jf). Microscopic morphological studies revealed that the major mechanisms of toughening was through the EAC existing as separate domains in the bulk matrix of the composites which tended to act as stress concentrators that initiated local yielding of the matrix around crack tips and enhanced the toughness of the composites. ACKNOWLEDGEMENTS I would like to thank my advisor Dr. Laurent Matuana for his guidance and support throughout my Ph.D. program. He helped me develop my research, scientific writing, and communication skills. I have no doubt that these skills will be an asset to me in my career. I am also indebted to my committee members, Dr. Bix, Dr. Chhin and Dr. Nzokou for their invaluable suggestions and advice. They always availed themselves whenever I contacted them. My gratitude goes to my colleagues Dr. Faruk, Dr. Acosta-Diaz, Dr. Jin and the undergraduate students (Deja Torrence, Thomas McKay, Anthony Weatherspoon and Ross Hickok) who worked with me. They were a great team to work with. Their willingness to help and share knowledge was unparalleled. Many thanks go to my wife, children, parents, and siblings for their unconditional love, support, and encouragement during my studies. I especially appreciate my wife and children for their understanding and the sacrifices they made during this entire program. I also want to recognize all my friends who have helped me throughout this process. The help of my friends at MSU has been invaluable in getting me through my Ph.D. program. Finally I want to acknowledge the USDA-CSREES Grant-Advanced Technology Applications to Eastern Hardwood Utilization (Grant No. 2008-34158-19510) for providing the financial support for my research. iv TABLE OF CONTENTS LIST OF TABLES ....................................................................................................................... viii LIST OF FIGURES .........................................................................................................................x LIST OF ABBREVIATIONS ...................................................................................................... xiii CHAPTER 1 Introduction ......................................................................................................................................1 1.1 Introduction ....................................................................................................................1 1.2 Objectives ......................................................................................................................7 1.3 References ......................................................................................................................9 CHAPTER 2 Background and Literature Review ...............................................................................................12 2.1 Introduction ..................................................................................................................12 2.2 Wood ............................................................................................................................12 2.3 Anatomy of Wood........................................................................................................14 2.4 Chemical Composition of Wood .................................................................................16 2.4.1 Cellulose .......................................................................................................16 2.4.2 Hemicelluloses ..............................................................................................18 2.4.3 Lignin ............................................................................................................19 2.4.4 Cell wall ultrastructure ..................................................................................21 2.4.5 Extractives.....................................................................................................23 2.5 Wood Moisture Content ...............................................................................................23 2.6 Wood Flour ..................................................................................................................24 2.7 Poly(lactic acid) (PLA) ................................................................................................26 2.7.1 Synthesis of poly(lactic acid) ........................................................................26 2.7.2 Applications and challenges of poly(lactic acid) ..........................................30 2.7.3 Toughening mechanisms of PLA..................................................................30 2.7.3.1 Copolymerization of PLA ..............................................................31 2.7.3.2 Plasticization of PLA .....................................................................32 2.7.3.3 Blending of PLA with tough polymers ..........................................34 2.7.3.4 Microcellular foaming of PLA.......................................................35 2.8 Wood Plastic Composites (WPCs) ..............................................................................36 2.8.1 Toughening mechanisms of wood plastic composites ..................................38 2.8.2 Effect of wood particle size on the properties of wood plastic composites (WPCs)..........................................................................................................38 2.8.3 Surface modification and compatibilization of the wood fillers and the matrix polymer ..............................................................................................39 2.9 Mechanisms of Failure of Polymers and Wood Plastic Composites ...........................46 2.10 Fracture Mechanics ....................................................................................................47 2.10.1 Load cases in fracture mechanics ...............................................................48 2.10.2 Fracture toughness measurement ................................................................50 2.10.2.1 Linear elastic fracture mechanics (LEFM) ..................................50 v 2.10.2.2 Elastic-plastic fracture mechanics (EPFM) .................................51 2.11 References ..................................................................................................................57 CHAPTER 3 IMPACT MODIFICATION OF POLYLACTIDE WITH A BIODEGRADABLE ETHYLENE/ACRYLATE COPOLYMER...................................................................................63 3.1 Abstract ........................................................................................................................64 3.2 Introduction ..................................................................................................................65 3.3 Experimental Part.........................................................................................................68 3.3.1 Materials .......................................................................................................68 3.3.2 Blending and molding of PLA/ethylene-acrylate copolymer .......................70 3.3.3 Differential scanning calorimetry (DSC .......................................................71 3.3.4 Scanning electron microscopy (SEM) ..........................................................71 3.3.5 Mechanical property evaluation ....................................................................72 3.4 Results and Discussion ................................................................................................73 3.4.1 Miscibility and Crystallinity of PLA/ethylene-acrylate copolymer blends ..73 3.4.2 Effectiveness and efficiency of impact modifier (ethylene-acrylate copolymer) in PLA .......................................................................................79 3.4.3 Tensile properties ..........................................................................................84 3.4.4 Morphology of the impact fractured surface and mechanisms of toughening.....................................................................................................89 3.4.5 Temperature effect on impact strength of impact modified PLA .................95 3.5 Conclusion ...................................................................................................................99 3.6 References ..................................................................................................................101 CHAPTER 4 STATISTICAL OPTIMIZATION OF TERNARY BLENDS OF POLY(LACTIC ACID)/ETHYLENE-ACRYLATE COPOLYMER/WOOD-FLOUR COMPOSITES ..............104 4.1 Abstract ......................................................................................................................105 4.2 Introduction ................................................................................................................106 4.3 Experimental Section .................................................................................................108 3.3.1 Materials .....................................................................................................108 3.3.2 Compounding and injection molding .........................................................108 3.3.3 Mechanical property evaluation ..................................................................111 3.3.4 Scanning electron microscopy (SEM) ........................................................112 4.4 Results and Discussion ..............................................................................................113 4.4.1 Statistical analysis of the impact strength of PLA/wood-flour composites...................................................................................................113 4.4.2 Statistical analysis of the tensile properties of PLA/wood-flour composites...................................................................................................124 4.4.3 Numerical optimization of the mechanical properties of PLA/wood-flour composites...................................................................................................129 4.5 Conclusion .................................................................................................................136 4.6 References ..................................................................................................................138 vi CHAPTER 5 FRACTURE TOUGHNESS OF POLY(LACTIC ACID)/ETHYLENE-ACRYLATE COPOLYMER/WOOD-FLOUR COMPOSITES TERNARY BLENDS...................................140 5.1 Abstract ......................................................................................................................141 5.2 Introduction ................................................................................................................142 5.3 Experimental ..............................................................................................................145 5.3.1 Materials .....................................................................................................145 5.3.2 Compounding and compression molding ...................................................145 5.3.3 Fracture energy determination ....................................................................146 5.3.4 Scanning electron microscopy (SEM) ........................................................151 5.3.5 Statistical analysis .......................................................................................151 5.4 Results and Discussion ..............................................................................................152 5.4.1 Geometric factor (η) calibration .................................................................152 5.4.2 Effects of wood flour and ethylene-acrylate copolymer (EAC) concentrations on fracture toughness of PLA/wood-flour composites .......157 5.5 Conclusions ................................................................................................................165 5.6 References ..................................................................................................................167 CHAPTER 6 Conclusions and Future Work .....................................................................................................170 6.1 Conclusions ................................................................................................................170 6.2 Future Work ...............................................................................................................175 6.3 References ..................................................................................................................179 APPENDICES Appendix A ..................................................................................................................................181 Appendix B ..................................................................................................................................197 vii LIST OF TABLES Table 3.1. Characteristics of different PLA resins used in study (2) ......................................69 Table 3.2. Thermal properties of neat PLA (3001D), pure ethylene-acrylate copolymer (EAC), and PLA/EAC blends ................................................................................77 Table 4.1. Experimental design matrix in terms of actual and coded factor levels generated ® by Design-Expert software ................................................................................110 Table 4.2. Analysis of variance (ANOVA) for two-level factorial model ...........................114 Table 4.3. Numerical optimization settings ..........................................................................131 Table 4.4. Numerical optimization solutions for impact strength and tensile properties .....133 Table 4.5. Results of validation test for the two scenarios of the optimization solutions for the targeted mechanical properties of the PLA/wood–flour composites .............135 Table A.1. Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (3001D)/EAC Blends ..................................................................................182 Table A.2. Pairwise Comparison of the Impact Strength of PLA (3001D)/EAC Blends ......183 Table A.3. Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (2002D)/EAC Blends ..................................................................................184 Table A.4. Pairwise Comparison of the Impact Strength of PLA (2002D)/EAC Blends ......185 Table A.5. Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (8302D)/EAC Blends ..................................................................................186 Table A.6. Pairwise Comparison of the Impact Strength of PLA (8302D)/EAC Blends ......187 Table A.7. Analysis of Variance for the Effect of EAC Content on the Energy to Break of PLA (3001D)/EAC Blends ..................................................................................189 Table A.8. Pairwise Comparison of the Energy to Break of PLA (3001D)/EAC Blends .....189 Table A.9. Analysis of Variance for the Effect of EAC Content on the Tensile Strength of PLA (3001D)/EAC Blends ..................................................................................190 Table A.10. Pairwise Comparison of the Tensile Strength of PLA (3001D)/EAC Blends .....190 viii Table A.11. Analysis of Variance for the Effect of EAC Content on the Tensile Modulus of PLA (3001D)/EAC Blends ..................................................................................191 Table A.12. Pairwise Comparison of the Tensile Modulus of PLA (3001D)/EAC Blends ....191 Table A.13. Analysis of Variance for the Effect of Temperature on the Impact Strength of Neat PLA 3001D..................................................................................................193 Table A.14. Pairwise Comparison of the Impact Strength of Neat PLA 3001D at Different Temperatures........................................................................................................194 Table A.15. Analysis of Variance for the Effect of Temperature on the Impact Strength of PLA 3001D/15 % EAC Blend .............................................................................195 Table A.16. Pairwise Comparison of the Impact Strength of PLA 3001D/15 % EAC Blends at Different Temperatures ........................................................................................196 Table B.1. Analysis of Variance for the Effect of Wood Flour Content on the J-Integral (Jin) of the Composites ................................................................................................198 Table B.2. Pairwise Comparison of the J-Integral (Jin) of the PLA/Wood-Flour Composites198 Table B.3. Analysis of Variance for the Effect of Wood Flour Content on the Fracture Energy (Jf) of the Composites .............................................................................199 Table B.4. Pairwise Comparison of the Fracture Energy (Jf) of the PLA/Wood-Flour Composites...........................................................................................................199 Table B.5. Analysis of Variance for the Effect of EAC Content on the J-Integral (Jin) of the Composites with 40 % Wood Flour Content .......................................................200 Table B.6. Pairwise Comparison of the J-Integral (Jin) of the PLA/Wood-Flour/EAC Composites...........................................................................................................200 Table B.7. Analysis of Variance for the Effect of EAC Content on the Fracture Energy (Jf) of the Composites with 40 % Wood Flour Content .................................................201 Table B.8. Pairwise Comparison of the Fracture Energy (Jf) of the PLA/Wood-Flour/EAC Composites...........................................................................................................201 ix LIST OF FIGURES Figure 2.1. Cross section of oak tree trunk: (A) pith (B) sapwood (C) heartwood (D) cambium and (E) bark. ...........................................................................................13 Figure 2.2. Schematic anatomical structures of (a) softwood and (b) hardwood. ....................15 Figure 2.3. Chemical structure of cellulose (the carbon atoms in the ring vertexes have been omitted to simplify the formula). ...........................................................................17 Figure 2.4. Chemical structures of (a) monolignol monomers and (b) lignin ..........................20 Figure 2.5. Sketch of the layered cell wall structure of a softwood longitudinal trachied showing the orientation of the cellulose microfibrils. The wall consists of primary (P) and secondary (S1, S2, and S3) walls. Adjacent tracheids are joined together by the middle lamella (m.l.) ...................................................................................22 Figure 2.6. Scanning electron micrograph of pine wood flour. ...............................................25 Figure 2.7. Chemical structures of L-, D-, and meso-lactide ...................................................27 Figure 2.8. Schematic of the synthesis of poly(lactic acid)......................................................29 Figure 2.9. Acetylation mechanism ..........................................................................................41 Figure 2.10. Mechanism of the graft-copolymerisation process ................................................43 Figure 2.11. Reaction of vinyltrimethoxysilane with cellulose fibers .......................................45 Figure 2.12. Modes of fracture for engineering materials ..........................................................49 Figure 2.13. Elastic-plastic behavior; (a) separation of the elastic and plastic contributions and (b) decrease of potential energy due to crack growth. ...........................................53 Figure 3.1. DSC thermograms of semicrystalline PLA (3001D) and its blends with various concentrations of ethylene-acrylate copolymer: (a) 0, (b) 5, (c) 15, and (d) 20 wt.%. The curve labeled (e) represents the thermogram of pure ethylene-acrylate copolymer. .............................................................................................................75 Figure 3.2. DSC curve of pure ethylene-acrylate copolymer ...................................................76 Figure 3.3. Effect of ethylene-acrylate copolymer (EAC) content on the notched Izod impact strengths of both amorphous (8302D) and semicrystalline (2002D and 3001D) PLA. Error bars are not shown in the graphs to prevent overlapping of the curves. x However, the values of standard deviation for the data shown in this figure were below 10 % of the mean values of the notched Izod impact strength....................81 Figure 3.4. Influence of ethylene-acrylate copolymer (EAC) content on the tensile stress strain curves of semicrystalline PLA (3001D).......................................................85 Figure 3.5. Effect of ethylene-acrylate copolymer (EAC) content on the energy to break and ductility of semicrystalline PLA (3001D). Some error bars did not appear in the graph due to the smaller values of standard deviation. ..........................................87 Figure 3.6. Effect of ethylene-acrylate copolymer (EAC) content on the tensile strength and modulus of semicrystalline PLA (3001D). Some error bars did not appear in the graph due to the smaller values of standard deviation ...........................................88 Figure 3.7. Scanning electron micrographs of impact fractured surfaces of semicrystalline PLA (3001D) and its blends with various concentrations of ethylene-acrylate copolymer: (a) 0, (b) 3, (c) 5, (d) 10, (e) 15, (f) 20, (g) 30, and (h) 40 wt.-% .......90 Figure 3.8. Photographs of notched Izod impact tested samples showing fracture modes: complete breakage for both (a) neat PLA and (b) PLA blended with 5 wt.-% ethylene-acrylate copolymer, and (c) partial breakage for PLA blended with 20 wt.-% ethylene-acrylate copolymer. (sw) in the figure stands for stress whitening. ‘For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation.’. ........93 Figure 3.9. Influence of testing temperature on the notched Izod impact strength of semicrystalline PLA (3001D) and its blend with 15 wt.-% ethylene-acrylate copolymer (EAC). Some error bars did not appear in the graph due to the smaller values of standard deviation...................................................................................96 Figure 4.1. Perturbation plots of impact strength of the composites against wood flour content (factor A), particle size (factor B), and impact modifier (EAC) content (factor C). .............................................................................................................116 Figure 4.2. Three-dimensional graphs of the variation of the impact strength of the composites as a function of the interaction between wood flour content (factor A) and impact modifier (EAC) content (factor C) with wood flour particle sizes (factor B) of (a) 20 and (b) 100 mesh sizes .........................................................118 Figure 4.3. Effects of wood flour particle sizes [(a) 20 mesh, (b) 40 mesh, (c) 60 mesh, and (d) 100 mesh] and contents on the impact strength of PLA/wood-flour composites............................................................................................................122 Figure 4.4. Scanning electron micrographs of impact fractured surfaces of composites with 100 mesh particle size and wood flour contents of (a) 5 % and (b) 40 % ...........123 xi Figure 4.5. Cube graphs of the relationships between tensile properties and wood flour content (factor A), particle size (factor B), and impact modifier (EAC) content (factor C) for (a) tensile strength, (b) tensile modulus, and (c) elongation at break of PLA/wood-flour composites............................................................................125 Figure 5.1. Schematic drawing of fracture toughness test specimen .....................................147 Figure 5.2. Scheme for the evaluation of energy U at different displacements ‘q’................153 Figure 5.3. Charts for the evaluation of the geometry factor (η) for neat PLA. (a) Energy input during fracture plotted as a function of crack length ‘a’ at different displacements ‘q’. (b) Geometry factors at different displacements ‘q’ plotted as a function of a/W ratios........................................................................................154 Figure 5.4. The overall geometry factors (η) as a function of a/W ratios for neat PLA, unmodified and EAC-modified PLA/wood-flour composites with 40 % wood flour (WF) content ...............................................................................................156 Figure 5.5. Effect of wood flour content on fracture toughness of PLA/wood-flour composites: (a) J-integral (Jin) and (b) fracture energy (Jf) .................................159 Figure 5.6. SEM of PLA/wood-flour composites with 40 % wood flour content and EAC contents of (a) 0 % and (b) 10 % .......................................................................161 Figure 5.7. Effect of EAC impact modifier content on fracture toughness of PLA/wood-flour composites containing 40 % wood flour content: (a) J-integral (Jin) and (b) fracture energy (Jf) ...............................................................................................163 xii LIST OF ABBREVIATIONS ANOVA -- analysis of variance APS -- 3-aminopropyltriethoxysilane ASTM -- American Society for Testing and Materials BDT -- brittle-to-ductile transition COO --` carboxylate CTOD -- crack tip opening displacement DLLA -- D-L-lactide DSC -- differential scanning calorimetry EAC -- ethylene-acrylate copolymer EB -- elongation at break EPDM -- ethylene/propylene/diene terpolymers EPFM -- elastic-plastic fracture mechanics FDA -- Food and Drug Administration IS -- impact strength ISO -- international standard organization LEFM -- linear elastic fracture mechanics LLA -- L-lactide MOE -- modulus of elasticity (tensile modulus) PCL -- polycaprolactone PDLA -- poly(D-lactide) PDLLA -- poly(D-L-lactic acid) xiii PE -- polyethylene PEG -- polyethylene-glycol PGA -- polyglycolic acid PHB -- polyhydroxy butyrate PLA -- polylactic acid PLLA -- poly(L-lactide) PP -- polypropylene PS -- polystyrene PVAc -- poly(vinyl acetate) PVC -- poly(vinyl chloride) SCF -- super critical fluid SEM -- scanning electron microscopy SEN -- single-edge-notched TBC -- tributyl citrate TEC -- triethyl citrate WPCs -- wood plastic composites xiv CHAPTER 1 Introduction 1.1 Introduction Composite materials made from bio-based plastics and natural and renewable fibers have become important as alternatives to the traditional wood plastic composites (WPCs) in recent years. The increased interest in bio-composites in applications such as automotive, leisure, building and furniture industries is attributable to environmental concerns on the use of petroleum-based resins, instability in oil producing regions and increasing oil prices (1). The most common bio-based plastics are aliphatic polyesters such as polylactic acid (PLA), polyglycolic acids (PGA), polycaprolactone (PCL) and polyhydroxybutyrate (PHB) (2). Among these polymers, PLA made from starch of annual crops is the front-runner because of its relatively high modulus, reasonable strength, excellent flavor and aroma barrier, good heat sealability and ease of processing using existing equipments (2-4). Compared to a conventional plastic such as polypropylene (PP) for instance, PLA has superior tensile strength and modulus (5). In spite of these good attributes, PLA is more expensive and has several drawbacks including brittleness and lower impact resistance when contrasted with most conventional plastics such as polyethylene (PE) and polypropylene (PP) (1,3,6). Historically, the high cost of PLA production limited its use to the specialty biomedical niches such as sutures and drug delivery devices due to its biodegradability and biocompatibility (7). Presently, technological innovations have decreased PLA production cost while the recent increasing crude oil and natural gas prices has conversely increased the cost of traditional petroleum derived plastics, enhancing PLA economic competitiveness (7). More recently, PLA 1 has been used as packaging materials for food and consumer goods (transparent bottles, meat trays, bags, films, etc.), with the polymer typically discarded after use. These applications benefit from their ability to decompose relatively quickly in a compost environment. Unfortunately, PLAs commercial expansion has been limited only to these areas due to its brittleness and lower impact resistance at room temperature (8,9), resulting in splitting and other handling problems during sheet manufacture (10-17). Therefore, the necessity for improvement in PLAs impact strength and toughness, which would eliminate the processing and handling deficiencies, precedes the possibility of widespread utilization and substitution for commodity plastics. In this regard, this research project will first address the toughening of PLA and subsequently manufacture PLA/wood-flour composites with improved impact strength and toughness. The toughening of PLA includes blending with other polymers (7,11-13,18-20), an exceptional method that is industrially relevant (21). Blending of PLA with rubbery polymers, has predominantly emphasized biomedical applications, resulting in the use of biodegradable and biocompatible polymers such as poly(vinyl alcohol), poly(e-caprolactone), poly(ethylene glycol), polyhydroxyalkanoate, and poly(butylene succinate) as second phase polymers (13,19). These impact modifying additives are relatively exorbitant rendering their PLA blends expensive. Improving the impact strength of PLA by direct mechanical blending with inexpensive nondegradable polymers such as poly(ethylene oxide), poly(vinyl acetate), polyisoprene, acrylonitrile/butadiene/styrene copolymer, and polyethylene has successfully reduced cost and expanded its commercial applications (11,19). However, the majority of these blends require compatibilizers to improve the miscibility between the impact modifier and the PLA. Therefore, using a commercially cost-effective impact modifier which is compatible with PLA would 2 produce greater advantages. In response to this challenge, DuPont Packaging and Industrial Polymers have introduced Biomax strong 100. This is a petroleum-based ethylene/acrylate copolymer (EAC) for commercial applications, noted as compatible with PLA and maintains its biodegradability at low concentrations (22). The feasibility of using this EAC as an impact modifier has been explored in a few studies and reported in literature (23,24). However, these studies failed to extensively account for the more subtle considerations of this impact modifier on different grades of PLA in terms of crystallinity (effectiveness) and concentrations of impact modifier (efficiency) (25) as well as providing detailed fracture mechanisms of impact modified PLA samples tested under various environmental conditions. Consequently, this study will investigate the mechanisms involved in the toughening of PLA with EAC. An in depth understanding of the toughening mechanisms of PLA is a crucial step in developing PLA with improved impact strength which will expand the use of PLA significantly in other applications. Both semicrystalline and amorphous PLA grades will be considered in view of the significant role crystallinity of PLA plays in the mechanical and durability performance in rigid molded applications (10). The modification of PLA with EAC is expected to yield improved flexibility, impact resistance and toughness of the PLA/EAC blends. These enhanced mechanical properties will be achieved at the expense of additional cost due to the inclusion of the EAC additive. Cost reduction is therefore a prerequisite to ensure the economic competitiveness of the PLA/EAC blends to commodity plastics. A well accepted approach to mitigate cost and provide specific properties such as low density, high specific stiffness and biodegradability is to compound wood 3 fibers with the polymer matrix (2,3,26). The wood fibers possess additional advantages of renewability, low hardness resulting in less wear on machinery, carbon dioxide sequestration and recyclability (3,27,28). A PLA/wood-flour composite would therefore be a convenient approach to control the material cost, engineer the composites mechanical properties and enhance environmental friendliness. Few scientific literatures have reported studies on PLA/wood-flour composites. Such studies have shown that PLA/wood-flour composites exhibit improved flexural and tensile moduli with wood flour content (2,3,5). However, tensile strength, flexural strength, impact strength, toughness and elongation at yield and break decreased with wood flour content (3,5,6,29). The decreased tensile strength and impact strength have been attributed to poor adhesion between the PLA matrix and wood fibers, short length of wood fibers and poor distribution of the wood fibers in the composite (6). Several studies have shown that toughness and impact strength of WPCs can be improved by impact modification and/or plasticization of the matrix polymer of the composites (27,28,30). However, the research conducted so far on PLA/wood-flour composites has mostly focused on the effects of wood flour content, coupling agent and plasticizer types and addition levels on the mechanical properties (2-6,31,32) without due consideration of the effects of impact modifier content and the particle size of the fiber. In this regard, for the full exploitation of the use of wood fillers in PLA/wood-flour composites, the effects of wood flour particle size, content, and impact modifier addition level on the mechanical properties must be fully determined. In view of the multiple variables involved, the traditional technique of varying one-factor at a time while holding other variables constant 4 would not be a good option to assess the interactions between factors as it is time consuming and often easily misses the interaction effects between factors. A statistical technique capable of developing a mathematical model that describes the relationships between the responses of interest and independent variables in which the significance of individual factors and multifactor interactions can be determined should be used (9,24,33,34). Therefore, this study used a twolevel factorial design to evaluate the effects of wood flour particle size (mesh size), wood flour content, and impact modifier concentration on the Izod impact strength and tensile properties of PLA/wood-flour composites. Even though, the Izod impact test is a simple and convenient testing procedure for toughness, it only measures the energy to break the sample without consideration of the crack initiation and propagation energies. The results from this test is useful to quickly assess the toughening potential of additives and characterize experimental formulations but do not represent the true material constants as they are size and geometry dependent (14,27,35). It is also difficult to interpret the results and compare with other test results since they have poor reproducibility (27). Moreover, impact test does not take into consideration cracks present in the tested sample, which can affect the testing results. Indeed, weak or poor interfacial adhesion occurs in the complex structure of ternary blends such as the composites of this study. Poor interfacial adhesion between components in the composites leads to cracks in the composites. Fibers also act as discontinuities in the composites, capable of initiating cracks (14). Due to the possibility of flaws or cracks in the composites, a method that takes those imperfections into consideration should be used to effectively evaluate the toughness of the composites. Fracture mechanics approach is one of the methods suitable for testing the fracture toughness of the material by taking its imperfections into consideration and produces parameters 5 that are true material constants, independent of both its size and geometry (36). In addition, fracture mechanics concepts have been proven to establish morphology-property correlations for thermoplastic materials, hence its use as an assessment tool for the PLA/EAC/wood-flour ternary composites is preferred (37). The J-integral method of fracture mechanics, was therefore used to assess the fracture toughness of both unmodified and ethylene-acrylate copolymer (EAC)toughened PLA/wood-flour composites. Particular emphasis was placed on evaluating the effects of wood flour concentration and EAC content on the fracture resistance of PLA/woodflour composites to gain an in-depth understanding of the mechanisms of crack initiation and propagation in the composites. In this study, we hypothesized that PLA toughened with ethylene/acrylate copolymer (EAC) will produce PLA/wood-flour composites with improved toughness. The concentration of ethylene/acrylate copolymer as well as crystallinity of PLA would dictate the extent of PLA toughening. The impact strength of the PLA/EAC blends and composites is expected to also be affected by the environmental temperature under which they will be tested. 6 1.2 Objectives The main aim of this proposed research project is to gain an in-depth understanding of the toughening mechanisms of PLA in order to develop PLA/wood-flour composites with improved impact strength, toughness, high ductility and flexibility using biodegradable ethylene/acrylate copolymer while reducing cost and maintaining biodegradability. To achieve this aim the following specific objectives must be accomplished: 1. Assess the efficiency and effectiveness of the EAC in toughening amorphous and semicrystalline grades of PLA with the ultimate goal of gaining an in depth understanding their toughening mechanisms. 2. Study how the testing temperature affects the notched Izod impact strength of toughened PLA in order to determine appropriate service temperature conditions for PLA/EAC blends. 3. Evaluate the effects of wood flour particle size (mesh size), wood flour content and impact modifier concentration on the mechanical properties of PLA/EAC/wood-flour composites. This is aimed at understanding the interactions between the composites component materials and developing model equations establishing the relationships between them. 4. Examine the fracture toughness of the composites using fracture mechanics in order to determine the energy consumed at each phase of the fracture process and obtain data that would be useful in engineering designs applying PLA/EAC/wood-flour composites. 7 REFERENCES 8 1.3 References 1. Haq, M., Burgueno, R., Mohanty, A.K. and Misra, M., “Hybrid bio-based composites from blends of unsaturated polyester and soybean oil reinforced with nanoclay and natural fibers,” Composites Science and Technology, 68: 3344-51 (2008). 2. Lee, S.Y., Kang, I.A., Doh, G.H., Yoon, H.G., Park, B.D. and Wu, Q., “Thermal and mechanical properties of wood flour/talc-filled polylactic acid composites: effect of filler content and coupling treatment.” Journal of Thermoplastic Composite Materials, 21 (3): 209–23 (2008). 3. Pilla, S., Gong, S., O’Neill, E., Rowell, R.M., and Krzysik, A.M., “Polylactide-pine wood flour composites,” Polymer Engineering and Science, 48: 578-87 (2008). 4. Oksman, K., Skrifvars, M. and Selin, J.F., “Natural fibres as reinforcement in polylactic acid (PLA) composites,” Composites Science and Technology, 63: 1317–24 (2003). 5. Huda, M.S., Drzal, L.T., Misra, M. and Mohanty, A.K., “Wood-fiber-reinforced poly(lactic acid) composites: evaluation of the physicomechanical and morphological properties,” Journal of Applied Polymer Science, 102: 4856-69 (2006). 6. Sykacek, E., Schlager, W. and Mundigler, N., “Compatibility of softwood flour and commercial biopolymers in injection molding,” Polymer Composites, 31: 443-51 (2010). 7. Schreck, K.M. and Hillmyer, M.A., “Block copolymers and melt blends of polylactide with nodax™ microbial polyesters: preparation and mechanical properties,” Journal of Biotechnology, 132: 287-95 (2007). 8. Matuana, L.M., “Solid State Microcellular foamed PLA: morphology and property characterization,” Bioresource Technology, 99 (9): 3643-50 (2008). 9. Matuana, L.M., Faruk, O. and Diaz, C.A., “Cell morphology of extrusion foamed poly(lactic acid) using endothermic chemical foaming agent,” Bioresource Technology, 100 (23): 594754 (2009). 10. Harris, A.M. and Lee, E.C., “Improving mechanical performance of injection molded PLA by controlling crystallinity,” Journal of Applied Polymer Science, 107 (4): 2246-55 (2008). 11. Li, Y.J. and Shimizu, H., “Toughening of polylactide by melt blending with a biodegradable poly(ether) urethane elastomer”, Macromolecular Bioscience, 7: 921-28 (2007). 12. Baird, J.C., Christiano, J.P. and Morris, B.A., “An extrusion study: examination of the improved processing characteristics of a PLA impact modified blend,” ANTEC Technical Papers, 155-60 (2009). 13. Ishida, S., Nagasaki, R., Chino, K., Dong, T. and Inoue, Y., “Toughening of poly(L-lactide) by melt blending with rubbers”, Journal of Applied Polymer Science, 113: 558-66 (2009). 9 14. Afrifah, K.A., Hickok, R.A. and Matuana, L.M., “Polybutene as a matrix for wood plastic composites,” Composites Science and Technology, 70: 167-72 (2010). 15. Matuana, L.M., Park, C. B. and Balatinecz, J. J., “Processing and cell morphology relationships for microcellular foamed PVC/cellulosic-fiber composites,” Polymer Engineering & Science, 37 (7): 1137–47 (1997). 16. Matuana, L.M., Park, C.B. and Balatinecz, J.J., “The effect of low levels of plasticizer on the rheological and mechanical properties of polyvinyl chloride/newsprint-fiber composites,” Journal of Vinyl & Additive Technology, 3 (4): 265-273 (1997). 17. Matuana, L.M., Woodhams, R.T., Balatinecz, J.J. and Park, C.B., “Influence of interfacial interactions on the properties of PVC/cellulosic fiber composites,” Polymer Composites, 19 (4): 446-55 (1998). 18. Kobayashi, S. and Sugimoto, S., “Biodegradation and mechanical properties of poly(lactic acid)/poly(butylene succinate) blends,” Journal of Solid Mechanics and Materials Engineering, 2(1): 15-24 (2008). 19. Anderson, K.S., Shawn, H.L. and Hillmyer, M.A., “Toughening of polylactide by melt blending with linear low-density polyethylene,” Journal of Applied Polymer Science, 89: 3757-68 (2003). 20. Murariu, M., Ferreira, A.D.S., Pluta, M., Bonnaud, L., Alexandre, M. and Dubois, P., “Polylactide (PLA)–CaSO4 composites toughened with low molecular weight and polymeric ester-like plasticizers and related performances,” European Polymer Journal, 44: 3842-52 (2008). 21. Gao, X., Qu, C. and Fu, Q., “Toughening mechanism in polyoxymethylene/thermoplastic polyurethane blends,” Polymer International, 53: 1666-71 (2004). 22. Dupont, Biomax Resins, Material Safety Data Sheet, 11/28/2006. 23. Zhu, S., Rasal, R. and Hirt, D., “Polylactide toughening using impact modifiers,” ANTEC Technical Papers, 1616-20 (2009). 24. Murariu, M., Ferreira, A.D.S., Duquesne, E., Bonnaud, L. and Dubois, Ph., “Polylactide (PLA) and highly filled PLA - calcium sulfate composites with improved impact properties,” Macromolecular Symposia, 272: 1-12 (2008). 25. Lutz, J.T. and Dunkelberger, D.L., Impact modifiers for PVC; the history and practice, John Wiley and Sons: Canada (1992). 26. Stark N.M. and Berger M.J., “Effect of particle size on properties of wood-flour reinforced polypropylene composites,” Proceedings of Functional Fillers for Thermoplastics and Thermosets Conference.1997: San Diego, CA. 27. Park, B.D. and Balatinecz, J.J., “Mechanical properties of wood fiber/toughened isotactic polypropylene composites,” Polymer Composites, 18: 79-89 (1997). 10 28. Oksman, K. and Clemons, C., “Mechanical properties and morphology of impact modified polypropylene-wood flour composites,” Journal of Applied Polymer Science, 67: 1503-13 (1998). 29. Bledzki, A.K. and Gassan, J., “Composites reinforced with cellulose based fibers,” Progress in Polymer Science, 24(2): 221–74 (1999). 30. Mengeloglu, F., Matuana, L.M. and King, J.A., “Effects of impact modifiers on the properties of rigid PVC/wood-fiber composites,” Journal of Vinyl and Additive Technology, 6(3): 153-57 (2000). 31. Li, Y., Venkateshan, K. and Sun, X.S., “Mechanical and thermal properties, morphology and relaxation characteristics of poly(lactic acid) and soy flour/wood flour blends,” Polymer International, 59: 1099-1109 (2010). 32. Sykacek, E., Hrabalova, M., Frech, H. and Mundigler, N., “Extrusion of five biopolymers reinforced with increasing wood flour concentration on a production machine, injection moulding and mechanical performance,” Composites: Part A, 40: 1272-82 (2009). 33. Sivaraman, P., Chandrasekhar, L., Mishra, V.S., Chakraborty, B.C. and Varghese, T.O., “Fracture toughness of thermoplastic co-poly(ether ester) elastomer – acrylonitrile butadiene styrene terpolymer blends,” Polymer Testing, 25: 562-67 (2006). 34. Rice, J., “A path independent integral and approximate analysis of strain concentration by notches and cracks,” Journal of Applied Mechanics, 2: 379-86 (1968). 35. Hughes, M., Hill, C.A.S. and Hague, J.R.B., “The fracture toughness of bast fibre reinforced polyester composites,” Journal of Materials Science, 37: 4669-76 (2002). 36. Wu, J., Mai, Y.W. and Cotterell, B., “Fracture toughness and fracture mechanisms of PBT/PC/IM blend,” Journal of Materials Science, 28: 3373-84 (1993). 37. Reincke, K., Grellmann, W., Lach, R. and Heinrich, G., “Toughness optimization of SBR elastomers – use of fracture mechanics methods for characterizationa,” Macromolecular Materials and Engineering, 288: 181-89 (2003). 11 CHAPTER 2 Background and Literature Review 2.1 Introduction Consistent with the scope of this study, a background and literature review on current studies of wood plastic composites (WPCs) and fracture mechanics is presented in this section. The review focuses on how the component materials of WPCs used in this study including wood, poly(lactic acid) (PLA) and additives affect the properties of the composites. Discussion on fracture mechanics methods emphasized developments in that field and how fracture toughness tests are performed. 2.2 Wood Wood is the xylem of the vascular tissue lying beneath the bark (Phloem) of woody plants (1). The wood in most species can be subdivided into sapwood, heartwood and pith. Figure 2.1 presents the structure of the cross-section of an oak tree showing the sapwood, heartwood, pith and bark. Living and dead cells are found in the sapwood which primarily stores food and transports water and sap in the plant. Heartwood on the other hand is inactive and made up of dead cells derived from the gradual change in sapwood (2,3). The transition from sapwood to heartwood is accompanied by increase in extractive content which gives the wood its characteristic dark color and odor (3). The pith constitutes a small core of tissue at the center of tree stems about which initial wood growth takes place (Figure 2.1) (3). 12 D E A C C B Figure 2.1. Cross section of oak tree trunk: (A) pith, (B) sapwood, (C) heartwood, (D) cambium, and (E) bark (4). 13 Wood has been an important construction material since humans began building. It has widely been used in applications such as furniture, weapons, musical instruments, domestic tools and many others. Commercially, important woody plants are classified as softwoods (conifers or evergreens) and hardwoods (deciduous or broad leaved). This classification does not reflect the hardness of their wood, as some softwood species are harder than some hardwood species. Anatomical characteristics are the more accurate means of distinguishing between these two classes (1). 2.3 Anatomy of Wood Wood consists of different types of cells that are mainly placed vertically to the three axes of the tree (5). Figures 2.2a and b, presents the schematic anatomical structures of softwoods and hardwoods, respectively. The main functions of wood cells in a living tree includes: providing support to the plant structure, transportation of sap from the roots to the leaves, storing substances such as sugars, starches, and fats, and metabolizing particular substances used by the tree for different functions (5). There are significant differences in the cell structure of hardwoods and softwoods (Figure 2.2a and b) (5). The xylem of softwoods is made up of mainly tracheids (Figure 2.2a). These are hollow, elongated and spindle shaped conduction and support cells of about 1.4 - 4.4 mm in length. They are usually present with a uniform appearance free from large pores (5,6). Hardwoods conversely have more complex anatomy due to the presence of additional and more specialized cell types (Figure 2.2b) (5). Besides tracheids, hardwoods have vessels and fibers which are for conduction and support, respectively (Figure 2.2b) (5). The lumen of the fibers 14 (a) (b) Figure 2.2. Schematic anatomical structures of (a) softwood and (b) hardwood (6). 15 can be completely or partially filled with tyloses which are growth from neighboring cells or resins or gums secreted from epithelial cells (6). Parenchyma cells are present in both softwoods and hardwoods. They are specialized for storage and are numerous in hardwoods than in softwoods (5). 2.4 Chemical Composition of Wood Chemically, wood is a complex composite material, composed of several organic compounds produced in living cells of a tree near its cambium (Figure 2.1). The main chemical components of wood are: cellulose (45-50 % by weight), and hemicelluloses (20-25 %), held together by lignin (20-30 %) (4). Other polymeric substances called extractives are also present in small amounts (3-30 %) (5,6). 2.4.1 Cellulose Cellulose is the main component of the cell walls providing stability and strength to wood (2,5,6). It is a highly crystalline, linear polymer consisting of 7000 to 15000 D-glucose units. The glucose units are produced during photosynthesis and are linked by β 1,4 glycosidic bonds in the cellulose chains (Figure 2.3) (4). The longest cellulose molecules are about 10 microns (µm) in length. 16 Figure 2.3. Chemical structure of cellulose (the carbon atoms in the ring vertexes have been omitted to simplify the formula) (5). 17 Three hydroxyl groups are present on every glucose unit in a cellulose molecule. Figure 2.3 presents the chemical structure of cellulose showing hydroxyl groups on each structural unit. These large numbers of hydroxyl groups on the cellulose chains accounts for its polarity and crystalline nature. The hydroxyl groups allow the chains of the cellulose to interact to form hydrogen bonds with adjacent chains and develop crystalline domains. Typically cellulose is about 60-90 % crystalline by weight and the portions that are not ordered into crystalline structure are amorphous. The hydroxyl groups in the amorphous regions are free and interact with water molecules through hydrogen bonding. This impart to the cellulose a high affinity for water (2,5,6). However, air humidity of the environment determines the amount of water the cellulose absorbs or releases (5). Due to the great dimensions of the cellulose chains, it absorbs water without dissolving in it (5). Additionally, the weak hydrogen bonds of cellulose and water renders the interaction between them reversible. 2.4.2 Hemicelluloses Apart from glucose, photosynthesis in plants produces other six carbon sugars such as galactose and mannose and five carbon sugars such as xylose and arabinose. These sugars together with other sugar derivatives like galacturonic acid, glucoronic acid and even glucose are used to synthesize hemicelluloses. Most hemicelluloses are branched chain polymers with degree of polymerization in the hundreds (2,5,6). Hemicelluloses act as cement between cellulose and lignin contributing to strength and stiffness of the wood (5,6). Examples of hemicelluloses are xylan, galactoglucomannan, glucomannan, and arabinogalactans. 18 The proportions of these hemicelluloses vary in softwoods and hardwoods. For instance in softwoods, the hemicellulose galactoglucomannan is about 16 % of the weight of wood whereas in hardwoods it is as high as 22–24 % (7). 2.4.3 Lignin Lignin is a 3-dimensional polymer made up of an array of variously bonded hydroxyland methoxy-substituted phenyl propane units (Figures 2.4a and b) (5,6). Primarily lignin is derived from three monolignol monomers methoxylated to various degrees. They are P- coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol (Figure 2.4a) (7). These lignols are incorporated in the lignin as phenyl propanoids such as p-hydroxyphenyl, guaiacyl, and syringyl, respectively. The general structure of lignin is illustrated in Figure 2.4b. There is a variation in lignin’s chemical structure depending on the source (6). Softwoods have lignin that is mainly made up of guaiacyl and small quantities of p-hydroxyphenyl units. By contrast hardwoods are a mixture of guaiacyl and syringyl units with very small quantities of p-hydroxyphenyl units (7). The polarity of lignin is less than that of cellulose due to its fewer hydroxyl groups. Lignin acts as a chemical adhesive within and between the cellulose fibers (6) giving rigidity to the cell and wood (2). It also limits penetration of water, reduces dimensional change with moisture content fluctuation and adds to wood toxicity making it resistant to decay and insect attack (2,5). 19 (a) CH2OH CH2OH CH2OH OMe MeO OH OH OH P-coumaryl alcohol Coniferyl alcohol Sinapyl alcohol CH2CHCH2OH (b) H3CO HOH2CHCH2C H3CO H H OCH3 CH2CHCH2OH CH2CHOH H H O H H H3CO CH2CHCH2OH CH2CHCH2OH H OCH3 O H H OCH3 OCH3 OH Figure 2.4. OMe O Chemical structures of (a) monolignol monomers (7) and (b) lignin (5). 20 2.4.4 Cell wall ultrastructure The linear nature of cellulose molecules allows them to form strong inter- and intramolecular bonds aggregating into bundles of molecules called microfibrils. Microfibrils form the elementary units of the cell wall of tracheids and fibers and have dominant influence on the overall behavior of wood as a material (4). In the cell wall, microfibrils are arranged in distinct layers in the primary and secondary subdivisions. Figure 2.5 shows the layered cell wall structure of a softwood longitudinal tracheid. The primary wall (i.e. outermost cell wall layer) is composed of randomly oriented microfibrils. The secondary wall makes up most of the cell wall volume and it is divided into three distinct layers of S1, S2, and S3. S1 is a thin layer of about 0.1 µm and have a microfibril o angle of 50–70 with respect to the cell’s longitudinal axis. The S2 is the thickest layer compared to the S1 and S3. It is about six times thicker than the S1. Consequently, it exerts a dominant influence on the overall behavior of the cell wall. S2 has microfibrillar angle of 10– o 30 and influences the anisotropic and orthotropic nature of wood. The S3 layer is the innermost o layer with typical thickness of 0.1µm and microfibrillar angle of 60–90 (4). 21 S3 S2 S1 m.l. P Figure 2.5. Sketch of the layered cell wall structure of a softwood longitudinal tracheid showing the orientation of the cellulose microfibrils. The wall consists of primary (P) and secondary (S1, S2, and S3) walls. Adjacent tracheids are joined together by the middle lamella (m.l.) (4). 22 2.4.5 Extractives These are inorganic (i.e. oxides, salts) and organic extraneous substances found in wood. They vary in quantity among the different tree species. The organic compounds can be extracted using appropriate solvents, thus are called extractives (5). Extractives which include substances such as fats, waxes, resins, proteins, gums, terpenes and simple sugars make up about 3-10 % of temperate woods, but higher quantities are found in tropical woods. They function in tree metabolism, act as energy reserves, defend against microbial attack, and impart color and odor to wood (6). The inorganic matter termed ash is also present in small quantities (~1 %) (6). 2.5 Wood Moisture Content Wood is hygroscopic or hydrophilic attracting moisture through hydrogen bonding because of the presence of hydroxyl and other oxygen containing groups on the main chemical constituents of the cell wall. This property is a drawback in composites fabrication and the performance of their end products (5,6). Commercially, moisture is removed from wood flour before processing or during processing into composites (5,6). However, once dried wood flour can still absorb moisture quickly because of its hygroscopic nature. Absorbed moisture interferes with and reduces hydrogen bonding between cell wall polymers, altering the mechanical performance of the product (6). Consequently, wood flour compounded with plastic matrices for wood plastic composites (WPCs) often need to be dried prior to further processing. Water uptake in composites depends on wood flour content, wood flour particle size, matrix type, processing method and additives used (6). 23 2.6 Wood Flour Wood flour refers to wood reduced to finely divided particles (6,7). The wood flour particles are usually small enough to pass through a screen with 850 micron openings (i.e. 20 US standard mesh) (6). The wood particles are also composed of bundles of wood fibers rather than individual fibers and have low aspect ratios (i.e. length to diameter ratios) which allow them to be metered and fed easily than individual wood fibers which tend to bridge (6). A scanning electron micrograph of pine wood flour is shown in Figure 2.6. Wood flour has been used in wood plastic composites, as extenders for glues, absorbents for explosives and in soil amendments. Bakelite made of phenyl formaldehyde and wood flour constitutes its earliest application in plastics (6). In the last few decades, wood flour has been used in thermoplastics for the manufacture of wood plastic composites for exterior building products, such as: railings, window and door profiles and decking. Normally, the wood flour used for these composites has a mesh size of 40 which is about 400 µm (7). Preferred plastics o for the composites are those that can be processed at temperatures lower than 200 C such as polypropylene (PP), polyethylene (PE), polyvinyl chloride (PVC), etc. due to the low thermal stability of the wood flour (6). 24 Figure 2.6. Scanning electron micrograph of pine wood flour (6). 25 2.7 Poly(lactic acid) (PLA) Poly(lactic acid) (PLA) is the first commodity polymer produced from annually renewable resources (8) such as starch of corn or sugar beets (9). Due to its source and ability to rapidly degrade in the environment, PLA is considered environmentally friendly. The by- products of its degradation have very low toxicity and are eventually converted to carbon dioxide and water (10-12). Additionally, it is biocompatible, has relatively high modulus, high strength, excellent flavor and aroma barrier, good heat sealability, and easy fabrication using existing technology and techniques (13-15). 2.7.1 Synthesis of poly(lactic acid) Nearly all agricultural materials such as sugar beets, corn and even waste can be used as basic materials for PLA production (16). Starch and sugar from agricultural materials are used directly as carbon sources for bacterial fermentation to produce the monomer lactic acid, the basic chemical needed for PLA production (17). Lactic acid is one of the simplest chiral molecules and exists as 2 stereo isomers namely L- and D-lactic acid (Figure 2.7). L-lactic acid (2-hydroxypropionic acid) is the natural and most common form of the acid. D-lactic acid can be produced by microorganisms or through racemization (8). Due to the stereo isometric nature of lactic acid (L- and D- lactic acid), three different types of lactides, which are the cyclic dimers of the lactic acids, can be generated by means of a combined process of oligomerization and cyclicization. These cyclic dimers are Llactide, D-lactide, and meso-lactide (Figure 2.7) (16). 26 O O O O O O O O O O O L-lactide Figure 2.7. O D-lactide Meso-lactide Chemical structures of L-, D-, and meso-lactide (10). 27 Synthesis of PLA can be by either direct condensation of the lactic acid or by the ring opening polymerization of the cyclic lactide dimer. The direct condensation route has the drawback of generating relatively low molecular weight PLA polymers due to difficulties in removing trace amounts of water generated at the late stages of polymerization. Consequently, most of PLA production is by the ring-opening polymerization of lactide. This approach produces high molecular weight polymers and allows the control of PLA properties by adjusting the ratios and the sequence of L-and D- lactic acid units (8). Figure 2.8 shows the schematic of the synthesis of poly(lactic acid) (PLA) using the ring opening polymerization method. Many important properties of PLA are controlled by the type of lactide used and their sequence of arrangement in the polymers. For instance, the isotactic homopolymers poly(Llactide) (PLLA) and poly(D-lactide) (PDLA) in which the monomers in the chain are of the same optical composition are semicrystalline (18). However, random copolymers of L-, D-, and mesolactide result in amorphous PLAs (18). Commercial PLAs are mostly a blend of PLLA and PDLA or copolymer PDLLA obtained by the polymerization of LLA and DLLA (19). PLA with 90 % or more of PLLA content tends to be crystalline while that with lower optical purity is amorphous (19). Thermal properties such as melting temperature (Tm), glass transition temperature (Tg) and percent crystallinity of PLA decreases with decreasing amounts of the PLLA (18,19). 28 O OH HO O HO O O Lactic acid OH O O n High molecular weight PLA Ring-opening polymerization Condensation -H2O O O O HO O O OH O n O Depolymerization O O Prepolymer Mn ~ 5000 Figure 2.8. Schematic of the synthesis of poly(lactic acid) (8). 29 2.7.2 Applications and challenges of poly(lactic acid) In spite of the unique combination of characteristics, PLA has several drawbacks. The main concerns are that, it is currently more expensive, has inferior moisture barrier properties, poor thermal resistance, inherent brittleness and lower impact resistance compared to most conventional plastics such as PE and PP. The high price of the polymer and the brittleness however has been the major bottlenecks for its large scale commercial application (20). At room temperature, PLA typically fractures through crazing mechanism (21). Approaches that have been adopted to improve the toughness and cost are discussed in the following sections. Until recently, PLA was used mainly in specialty biomedical applications such as sutures and drug delivery devices due to its biodegradability and biocompatibility. Presently, PLA has also been used as panels in automobiles and packaging materials for food and consumer goods (transparent bottles, meat trays, bags, films, etc.) (22). However, extensive application of PLA as replacement for petroleum based plastics in commodity products can only be achieved after further improvement in its cost and impact resistance. 2.7.3 Toughening mechanisms of PLA The ability of a material to undergo massive yielding at impact speeds (i.e. impact strength or toughness) is very important in many engineering applications such as in automobiles, electronic devices, buildings, etc, where they will be subjected to impact stresses. Gordon, (1968), emphasized the importance of toughness of materials in engineering applications and surmised it as “The worst sin in an engineering material is not lack of strength or stiffness but lack of toughness” (23). 30 The impact strength and toughness of PLA have been enhanced by approaches such as plasticization, copolymerization, blending with tough polymers, and rubber toughening (20). However, the improved toughness of PLA is at the expense of other strength properties such as tensile strength and modulus (24). Consequently, PLA toughness modification strategies should also aim at having a good balance between toughness and other strength properties. 2.7.3.1 Copolymerization of PLA Copolymerization which allows two or more monomers to be combined into one polymeric chain has been investigated as a means to obtain polymer materials with properties unattainable by homopolymers (17). A wide range of mechanical properties can be attained in a versatile way by manipulating the architecture of the copolymer molecule, monomers sequence and composition (17). Random, block, and graft copolymers of PLA with other monomers has been produced to modify its properties (16,17,24). Although copolymerization of PLA can be conducted through polycondensation of lactic acid with other monomers, it has mainly been done using the ring opening copolymerization of lactide with other cyclic monomers. This is because of the precise control of the chemistry and the higher molecular weight of the copolymers produced by using the ring opening copolymerization (17). Some monomers that have been copolymerized with lactic acid include glycolide derivatives, lactones, cyclic amide esters, cyclic ether esters, and cyclic carbonates (24). Poly(lactic-glycolic acid) a copolymer approved by the food and drug administration (FDA) for clinical uses is composed of lactic acid and glycolic acid. A variant having lactic acid and glycolic acid in 2:23 ratio has been used in controlled drug release systems (20). 31 Copolymerization of lactic acid has generally been successful with several monomers and produced copolymers with promising properties. Unfortunately, none of these copolymerization processes are so far economically viable (17). 2.7.3.2 Plasticization of PLA Plasticizers are low molecular weight organic substances that are added to rigid polymers to improve processability, impart softness and enhance their low temperature flexibility and ductility (17). They also decrease the concentration of intermolecular forces between the host polymer chains and consequently the glass transition temperature (Tg) of the polymers (25). Selection of a plasticizer for PLA should be influenced by bio-degradability, non-volatility, nontoxicity, efficiency at lowering Tg of PLA, and minimal leach or migration during aging (17,26). For a plasticizer to be effective it must first be thoroughly mixed and incorporated between the polymer matrix/chains and interact with the functional groups to reduce the polymer’s chain to chain interaction. Effective miscibility of a plasticizer in the matrix polymer is achieved by heating and mixing until either the resin dissolves in the plasticizer or the plasticizer dissolves in the resin (25). The possibility of plasticizers and host polymers mixing is determined by the closeness of their solubility parameters (δ) and the magnitude of their interaction parameters (χT) (27). High interaction parameters and close solubility parameter values between the plasticizer and polymer are required for good miscibility. It must be emphasized that, the plasticizing efficiency of a plasticizer depends not only on its miscibility with host polymers but also its molecular weight and loading level. The 32 plasticizing efficiency can be evaluated by the depression in the Tg of the blend and improvement in toughness (17). There are several theoretical models to predict the single Tg of the plasticized blends, however Fox equation is the most frequently used (Equation 1). W W 1 = 1 + 2 Tg Tg1 Tg 2 (1) with Tg1 and Tg2 as the Tgs of the pure components in Kelvin andW1 andW2 as their respective weight fractions in the mixture (28). The extent of Tg depression in PLA is directly related to the amount of plasticizer used (27). The higher the plasticizer content however, the higher is the chain mobility of PLA leading to faster cold crystallization (27). Generally, low molecular weight plasticizers are more efficient than larger ones (17). Nevertheless, low molecular weight plasticizers have the problem of migrating to the surface in the long term, which would cause embrittlement. Attempts have been made to prevent the above problem with relatively high molecular weight polymers but the expected improvements in properties were not obtained due to their thermodynamically immiscibility and phase segregation (29). Several monomers and oligomers have been tried to plasticize PLA. The commonly investigated plasticizers are polyethylene-glycol (PEG) and citrate esters (17). Low molecular 33 weight plasticizers such as tributyl citrate (TBC), triethyl citrate (TEC), lactide molecules, diethyl bis(hydroxymethyl) and triacetin are effective for PLA below 25 % concentration (26). 2.7.3.3 Blending of PLA with tough polymers The toughening of PLA by blending with other polymers (9,12,30-34) is an exceptional method that is industrially relevant (35). This is because it is a much more economic and convenient methodology than synthesizing new polymers to achieve the properties unattainable with existing polymers (17). Blending of PLA with rubbery polymers, has predominantly emphasized biomedical applications, resulting in the use of biodegradable and biocompatible polymers such as poly(vinyl alcohol), poly(ε-caprolactone), poly(ethylene glycol), polyhydroxyalkanoate, and poly(butylene succinate) as second phase polymers (31,34). These toughening additives are relatively exorbitant rendering their PLA blends expensive. Compatibilizers sometimes have to be used to improve interaction between the PLA and the other polymer in order to improve mechanical properties of the blend. For instance, in blends of PLA/polycaprolactone (PCL), compatibilized with dicumyl peroxide and lysine triisocyanate, the impact strength and impact fracture toughness respectively were markedly increased (17). Improving the impact strength of PLA by direct mechanical blending with inexpensive non-degradable polymers such as poly(ethylene oxide), poly(vinyl acetate), polyisoprene, acrylonitrile-butadiene-styrene copolymer and polyethylene has successfully reduced cost and expanded its commercial applications (9,31). However, the majority of these modifiers are also thermodynamically immiscible with PLA and lack favorable interactions due to differences in their chemical structures which results in weak interfacial adhesion (indicated by poor 34 dispersion, very broad size distribution and distinct particle interfaces) and poor mechanical properties. Compatibilizers are therefore required to improve the interaction between the impact modifier and the PLA. Therefore using a commercially cost-effective impact modifier which is biodegradable and compatible with PLA would be preferred and produce greater advantages. 2.7.3.4 Microcellular foaming of PLA Microcellular thermoplastic foam is based on the theory that the creation of a very large number of micro bubbles, smaller than the pre-existing natural flaws in a polymer, can reduce the material cost and consumption in mass-produced plastic parts without compromising mechanical properties. This concept has attracted a lot of attention in the research community in recent years. In particular, Matuana (36) used this approach to address the brittleness and utilization cost of PLA. Using a batch foaming process he produced microcellular foams in the PLA which resulted in significant improvements in the; volume expansion ratio (a two times increase in expansion over unfoamed PLA), impact resistance (about four times increase over unfoamed PLA), strain at break (about two times increase over unfoamed PLA), and toughness (about four times increase over unfoamed PLA). These improvements were due to the presence of a very high cell population density of minute bubble cells. The small bubbles inhibited crack propagation by blunting the crack tip and increasing the amount of energy needed to propagate the crack (36). However, the batch foaming process is time consuming due to the multiple steps involved, making it unattractive for industrial application (37,38). Due to the limitations of the batch-microcellular foaming process, manufacturing of PLA foams through the continuous 35 extrusion and injection molding processes using super critical blowing agents has attracted attention in recent years. Additional cost is introduced as these processes require modification of the existing equipment by adding a super critical fluid (SCF) delivery system that provides high pressure and accurately metered mass flow of SCF to the processing equipment. Screw and die modification may also be required to achieve a single phase solution which is a critical step in the gas foaming process for homogeneous cell nucleation (39-43). 2.8 Wood Plastic Composites (WPCs) A composite is a hybrid material. It is a combination of two or more components to achieve desired properties, by taking advantage of the beneficial characteristics of each component material (44,45). The term wood plastic composites (WPCs) refer to any composite that contain wood of any form and thermosets or thermoplastics (46,47). Thermosets include plastics such as epoxies and phenolics which once cured cannot be melted by reheating. Contrastingly, thermoplastics are plastics that can be repeatedly melted (47). This review concentrates on wood thermoplastic composites, which are often referred to as wood plastic composites (WPCs) with the understanding that the plastic is a thermoplastic. Thermoplastic composites have become attractive due to their unique properties, such as good acoustical properties, thermal insulating properties, high strength, and elasticity modulus (48). The interest in the use of wood and cellulosic fibers for reinforcement in thermoplastic composites is due to their high aspect ratio, high specific stiffness and high strength. In addition to the above properties, they have relatively low density and low hardness which minimizes 36 abrasion of the equipment during processing. They are also environmentally friendly and economical by being renewable, biodegradable and low cost (49,50). WPCs combine the properties of both wood and plastic. Consequently, in addition to the improved physical and mechanical properties, WPCs are more durable than wood, while being more cost-effective than plastics. Because of the plastic, WPCs absorb much less moisture than solid wood. This helps to reduce swelling, as well as fungi and insect attacks. WPCs can also be sanded, painted, stained, and finished like solid wood (49,51,52). Wood plastic composites have a variety of uses. Some of the commercial applications include automotive interior substrates, packaging, furniture and housing (53). The majority of WPCs currently being produced are using polyethylene (PE), polypropylene (PP), polystyrene (PS), and polyvinyl chloride (PVC) as matrices. These plastics are chosen because of their lower processing temperatures (150-220 °C) which prevent the deterioration of the added cellulosic materials (51). Despite the advantages of WPCs, the less desirable characteristic of these composites is the drastic reduction in some of their mechanical properties such as toughness due to the poor interfacial adhesion between the hydrophobic matrix and hydrophilic filler (51). Toughness of composites reinforced with wood flour can be improved by methods such as; i) increasing the matrix toughness, ii) optimizing the interphase between the filler and the matrix through the use of coupling agents and compatibilizers, and iii) optimizing the filler-related properties such as filler content, particle size, aspect ratio, orientation and dispersion (53). Discussions on how the above approaches are used to improve the mechanical properties of WPCs are presented elsewhere in this review. 37 2.8.1 Toughening mechanisms of wood plastic composites Studies have shown that toughness and impact strength of WPCs can be improved by impact modifiers (45,51,52). Some reported studies used impact modifiers such as ethylene/propylene/diene terpolymers (EPDM), styrene-ethylene/butylene-styrene and their maleated derivatives for improvement of impact strength and toughness of polypropylene/woodflour composites (45,51). The addition of these impact modifiers increased impact strength and toughness while it reduced stiffness and strength of the composites (45,51). Mengeloglu et al. (52) examined the effects of impact modifier types and addition levels on the mechanical properties of rigid PVC/wood-fiber composites. They observed a strong dependence of impact resistance of PVC/wood-fiber composites on the type and content of impact modifier. Of the impact modifiers studied, methacrylate-butadiene-styrene and all acrylic modifiers were observed as more effective and efficient in improving the impact resistance of the PVC/woodfiber composites compared to the chlorinated polyethylene modifier. 2.8.2 Effect of wood particle size on the properties of wood plastic composites (WPCs) In the plastic industry, fillers are classified as either particulate or fibrous. Particulates can have any shape and normally have dimensions that are approximately equal in all directions. They are classified mainly as fillers but also considered reinforcing fillers when their interphase adhesion with the matrix polymer is high leading to effective stress transfer between the wood particles and the polymer matrix. Fibers on the other hand are considered reinforcing because they bear a majority of the load applied and their length to cross-sectional dimensions are not equal (54). 38 One of the factors to consider when using wood flour as filler in wood plastic composites is particle size. This property, which tends to have significant influence on some mechanical properties, has not been extensively studied. Stark and coworkers (55) studied the effect of species and particle size on properties of wood flour filled polypropylene composites. They observed that at wood flour loading of 40 %, particle size did not affect specific gravity, but it did affect other properties. As particle size increased from 120 to 20 mesh, the melt flow index, heat deflection temperature, notched impact energy, flexural modulus, tensile modulus and strength increased. These observations not withstanding, small fibers are preferred for manufacturing WPCs because they are easily processed by plastic processing equipments (54). 2.8.3 Surface modification and compatibilization of the wood filler and the matrix polymer The surface properties of the wood fibers and the polymer matrix influence the nature of the interfacial interactions between them. Good surface properties are required to obtain composites with high tensile and flexural strengths. The polar or hydrophilic nature of wood fiber due to the existence of many hydroxyl groups in cellulose, hemicelluloses and lignin is a major concern for their use in wood plastic composites. This is because it results in low compatibility with the non-polar or hydrophobic polymer matrices (56) leading to poor wetting of the fibers by the polymer, inhomogeneous dispersion of the fibers in the matrix polymer, and poor adhesion between the fibers and the polymer. Consequently, this incompatibility leads to composites with lower strength properties (e.g. tensile and flexural strengths) than the unfilled polymer (57,58). Hence it is important to enhance the surface properties of the wood fibers 39 and/or the polymer matrix to improve adhesion between them and derive optimum tensile and flexural strengths from the composites. The surface modification of wood fibers by coupling agents is an effective approach that is generally used to attain improved interfacial adhesion and subsequently satisfactory tensile and flexural strengths of the composites (57). The coupling agents compatibilize or promote the affinity between the wood fibers and the plastics and induce the bond formation between the fibers and plastic matrix. When strong bonds are formed, the stress transfer from the polymer to the load bearing fibers takes place, resulting in the enhancement of the tensile and flexural strength properties of the composites (57). Some examples of coupling agents that have been used in WPCs include acetic anhydride (acetylation) and maleated polyolefins (compatibilizing agents). Acetylation or esterification processes in which the hydroxyl group of wood fibers react with acetic anhydride to form esters on the wood surface has been explored (Figure 2.9) (56,59). The acetic anhydride substitutes the hydroxyl groups on the wood fiber cell wall with acetyl groups. This modifies the surface of wood fibers making them more hydrophobic than the untreated counterparts (60). The hydrophobic fibers obtained after this process are more compatible with the non-polar and hydrophobic polymers used as matrix for WPCs (56). Enhanced compatibility between the wood fibers and the polymer matrix due to the acetylation process generally results in improved adhesion between the interfaces of the composite leading to dimensional stability of the composites, increased tensile and flexural strength, and improved moisture resistance (59). 40 Cellulose-OH + CH3-CO-O-CO-CH3 Figure 2.9. Cellulose-O-CO-CH3 + CH3COOH Acetylation mechanism (56). 41 Compatibilizing agents have also been used to provide good adhesion between polymers such as PLA and natural fibers such as wood flour (58). These additives react with both the polymer and fibers to create chemical bridge at the interface. Maleated compounds and silanes are the generally used compatibilizers in WPCs (56). Maleic anhydride functionalized polyolefin is commonly used as a compatibilizing agent for polyolefin/natural fiber composites due to its strong effect in altering the surface energy of the fibers to a value close to that of neat polyolefin matrix (57,61). For instance the introduction of bi-functional polypropylene grafted with maleic anhydride in a polypropylene/wood-flour composite, improves compatibility by chemically reacting with the fibers on one end and physically forming entanglements with the polymer matrix on the other end (56). The improved adhesion between the two phases results in reduced moisture absorption and enhanced tensile and flexural strength properties of the composite (56). The mechanism underlying this process of compatibilization is shown in Figure 2.10a and b: 42 (a) O O H2 HO C C HO C C H O Δ C O H C C2 C C H O C H 2O + PP chain PP chain o Activation of the copolymer by heating (t=170 C) (b) O H2 O C C O C C C H O Cellulose fiber O OH + OH O H C C2 C C H O C O O PP chain O Cellulose fiber H Cellulose fiber Esterification of cellulose Figure 2.10. PP chain Mechanism of the graft-copolymerisation process (56). 43 H2 C C O C C C H H O PP chain The effectiveness of maleated polyolefins in improving the adhesion between the wood fiber and polymer matrix is attributed to two mechanisms. Firstly the carboxylic acid portions of the maleated compound react with hydroxyl groups on the wood surface to form monoester or diester linkages. This leaves the remaining polyolefin portions of the maleated compound as free pendant chains. The second mechanism involves the diffusion of the free pendant polyolefin chain into the polymer matrix which forms a physical bond upon cooling (60,62,63). Successfully achieving these two mechanisms, results in the formation of adhesive bridge between the wood fiber and the matrix polymer. Polyisocyanates, triazines and organosilanes have also been used as coupling agents. Silane coupling agents consist of organo-functional group in one side of the chain and an alkoxy group in the other (56). The organofunctional group causes reaction with the polymer while the alkoxy group undergoes hydrolysis to form silanol groups. The silanol groups formed are able to react with the hydroxyl groups of cellulose, hemicellulose and lignin by etherification process to form strong covalent bonds between the wood fiber and the matrix polymer (Figure 2.11) (56). The chemical reactions in Figure 2.11 illustrate the processes. The strong covalent bonds formed between the fibers and the matrix polymer ensures effective stress transfer from the polymer to the fiber resulting in improved strength properties, especially tensile and flexural strengths (58,59). Sometimes adhesion promoters are also used to increase the interfacial interactions between cellulosic fibers and the matrix polymer. Lignin has been used as adhesion promoter in PLA/cotton fiber reinforced composites with observed improvements in tensile strength and modulus but decreased impact strength (58). 44 O CH3 H2C CH Si O CH3 O CH3 CH2 CH CH2 CH + OH H2C CH Si OH OH 3 H 2O OH CH2 CH OH HO + OH CH2 CH Si O Si O Si OH O OH CH2 CH O O Cellulose fiber OH Cellulose fiber Figure 2.11. 3H3C OH CH2 CH HO Si O Si O Si OH OH + Reaction of vinyltrimethoxysilane with cellulose fibers (56). 45 2.9 Mechanisms of Failure of Polymers and Wood Plastic Composites Several studies have reported occurrence of various energy-dissipative mechanisms locally in a stressed plastic specimen before catastrophic crack development (9,64-67). Mechanisms such as crazing, bond rupture, cavitation, crack growth, plastic and viscoelastic deformations, etc. relieve stresses and consequently reduce the stored elastic energy. High strength and toughness result primarily from special combination of these mechanisms that retard or arrest the growth of cracks (66). Using fractured surfaces of notched Izod impact tested samples as a means of differentiating brittle to ductile fractures, previous studies reported noticeable whitening only occurs at the origin of the notched tip for brittle fracture, while ductile fracture involves all of the material around the fractured surface in stress whitening and forms a yielding zone (65). In rubber-toughened plastic systems, two types of cavitations induced by impact or tensile tests are discerned, which includes internal cavitations in the rubber domains for the blends with strong interfacial adhesion and debonding cavitations between the interfaces with insufficient interfacial adhesion (9). Generally, the mechanisms causing damage during brittle fracture have been prescribed as mainly crazing or microcracks (leading to stress whitening) and cavitation (67). Crazing, cavitation, shear banding, crack bridging and shear yielding reportedly occur as important energy dissipation processes involved in the impact fracture of toughened polymer systems (64). In filled polymers or polymer blends such as wood plastic composites, the increased impact strength may be explained by the role of impact modifier in toughening the matrix. For instance, in wood flour filled PP, impact modifiers have been reported to affect its morphology by either existing as separate phase in the matrix, partially or completely encapsulating the filler or a mixed condition of the above (51). Impact modifier particles existing as a separate phase in 46 the matrix of the composites generally act as stress concentrators that initiate local yielding of the matrix avoiding brittle failure of the material (51). The encapsulation of the fibers by the polymer on the other hand reduces the stress concentrations at the fiber-polymer interface, thus enhancing the resistance of the composites to fracture (51). 2.10 Fracture Mechanics Fracture mechanics deals with predicting the level of loads at which cracks may grow. Data obtained by this technique are important for the safe design of structures (68). For the characterization of a materials fracture parameters, the originator of fracture mechanics Griffith made assumptions that fracture begins with the initiation and growth of cracks from pre-existing defects or cracks in the structure of the materials (69). He indicated that the potential energy of the system varied with the size of the crack and consists of the internal stored elastic energy and the external potential energy of the applied loads. The elastic strain energy model (Equation 2) is used in fracture mechanics to demonstrate the propagation of cracks in materials (70): σ= 2 Eγ s (2) πa where E is the modulus of elasticity, γs is the specific surface energy and a is one half the length of an internal crack. 47 The above relation (Equations 2) only holds for brittle materials and cannot be applied in materials that experience some plastic deformation during fracture (70). This plasticity deficiency in the model was remedied by Irwin and Orowan (71). They suggested that in a ductile material a great deal of the released strain energy was absorbed by energy dissipation due to plastic flow in the material near the crack tip and not for creating new surfaces (71). Additionally, they concluded that catastrophic fracture occurs when the strain energy is released at a rate sufficient to satisfy all the energy requirements for fracture. The critical strain energy release rate required for fracture was denoted by the parameter Gc, hence the Griffith equation was rewritten as (71); σ= EGc πa (3) 2.10.1 Load cases in fracture mechanics A loaded body can be classified as one of three characteristic load cases based on the orientation of the stress fields near the crack tips which affect the local mode of failure. The load cases are Mode I, Mode II, and Mode III (Figure 2.12) (70). Mode I represents opening in a purely tensile field (Figure 2.12). Mode II fracture causes the cracked walls or surfaces to slide relative to each other in a direction normal to the crack front (Figure 2.12). In Mode III the crack surfaces slide (shear) with respect to one another in a direction parallel to the crack front (Figure 2.12) (70,72). Commonly found failures are due to cracks propagating predominately in Mode I making it the most critical mode (70,72). Discussions in this paper therefore focused only on Mode I. 48 Mode I Tensile Figure 2.12. Mode II In-plane Modes of fracture for engineering materials (70). 49 Mode III Anti-plane 2.10.2 Fracture toughness measurement Two common fracture toughness measurement approaches based on the ductility of the material are used to analyze cracked solids in fracture mechanics. In dealing with brittle materials the stress intensity factor, K, is determined using the linear elastic fracture mechanics method. For ductile materials elastic-plastic fracture mechanics (EPFM) methods, such as the Jintegral and crack tip opening displacement (CTOD) methods are used. The following sections describe these fracture toughness measurement approaches. 2.10.2.1 Linear elastic fracture mechanics (LEFM) Linear elastic fracture mechanics (LEFM) is one of the most frequently used methods to evaluate polymer fracture (73). It is effective with materials showing only linear elastic deformation or brittle failure (74). In the application of LEFM, the fracture toughness may be represented by the stress intensity factor, K or the energy release rate, G. The stress intensity factor is based on stresses around a cracked tip and failure occurs when the stress reaches a critical value Kc (Equation 4) (70). K c = σ c πaY (4) where σc is the critical stress which will lead to crack propagation, Y is a geometry factor which takes different geometries into account and a is the crack length. Theoretically, Kc is the inherent ability of a material to withstand a given stress-field intensity at the tip of a crack and 50 prevents tensile crack extension under plane-strain condition (70). Measurement of Kc is done at crack initiation for materials subjected to tensile or bending stresses (73,75). The strain energy release rate is the energy needed to extend a crack of a unit area. It causes fracture when it also reaches a critical value Gc (Equation 5). Gc = (1 − v 2 ) K c 2 E (5) where v is Poisson’s ratio and E is the modulus of elasticity. The stress intensity factor K and strain energy release rate Gc represents true material constants only when limited plastic deformation occurs at the crack tip or when size criteria of the testing specimen that leads to plane-stain fracture have been satisfied (73). Plane stress or strain conditions are achieved in mode I fracture, if the thickness, B, of the sample satisfy Equation 6. B ≥ 2.5 (Kc )2 (6) σ2 y where Kc is the critical stress intensity factor and σy is the tensile yield stress (70). 2.10.2.2 Elastic-plastic fracture mechanics (EPFM) In practice, many engineering materials such as toughened polymer blends are not strictly elastic and develop large plastic deformations ahead of the crack tip (75,76). These materials 51 have low yield stress and high toughness, requiring large/unrealistic specimen thickness sizes to validate the use of LEFM (78). Such toughened materials are better assessed with post yield fracture mechanics approaches such as the crack-tip-opening displacement (CTOD) and the Jintegral concepts (74). However, the J-integral concept is the most frequently used technique due to its energetic interpretation of the fracture process (77). Therefore, subsequent discussions in this review will focus on the J-integral concept as a criterion for fracture toughness. J-integral concept proposed by Rice (77) is a path independent line integral that encloses the plastically deformed area and describes the stress-strain field around the crack-tip (68,70,79). It is measured at crack initiation point in the fracture of elastic plastic materials (79). The Jintegral can be divided into an elastic component compatible to Kc and a plastic component derived from the plastic area under the load displacement curve. The separation of the elastic and plastic contributions of J-integral is shown in Figure 2.13a. J-integral concept can be demonstrated by two identical specimens equally loaded, but having different crack lengths and potential energies. Assuming initial crack lengths a and (a + da) and crack propagations of S and S’, respectively, after load application for the two identical specimens (Figure 2.13) (70). The difference between their potential energies would be the shaded zone between the two load-displacement curves shown in Figure 2.13b and equal to the energy required to produce a crack surface (70,76). This represents the energy intake per unit area to create new fracture surfaces in a loaded body containing a crack at constant displacement. 52 (a) Force (b) Force Crack length, a S S’ Up Ue Crack length, a + da Displacement Figure 2.13. Displacement Elastic-plastic behavior; (a) separation of the elastic and plastic contributions and (b) decrease of potential energy due to crack growth (70). 53 It can be expressed as: J in = − 1 dU B da (7) where B is the thickness of the specimen, U is the total potential energy (i.e. the area under the load-displacement curve) and a is the crack length. For ease of application of this method to experimental determination and computation of fracture toughness of tough materials, the above equation has been rewritten as (80): J in = ηU in (8) B(W − a ) where η is a geometry factor, Uin the potential energy at crack initiation and W the width of the specimens. Sumpter and Turner (81) expanded Equation (8) and expressed it as: J in = J e + J p = ηeU e B(W − a ) + η pU p B(W − a ) (9) where Je and Jp are the elastic and plastic components of the whole J value. Correspondingly, ηe and ηp are the elastic and plastic work factors. Ue and Up are the elastic and plastic contributions of the total energy (80). The values of the geometry factors ηe and ηp are equal to 54 2 when the crack’s length-to-width ratio a/W of notched specimens satisfies the criterion 0.4 < a/w and < 0.6 (81), simplifying Equation 9 to: J= 2U B(W − a ) (10) 55 REFERENCES 56 2.11 References 1. Bodig, J., and Jayne, B.A., Mechanics of Wood and Wood Composites, Van Nostrand Reinhold Company Inc.: New York. p. 712 (1982). 2. Bowyer, J.L., Shmulsky, R., and Haygreen, J.G., Forest Products and Wood Science: An Introduction, Fifth ed. Blackwell Publishing: Ames, Iowa. p. 558 (2007). 3. Miller, R.B., Wood Handbook: Wood as an Engineering Material, Structure of Wood, Forest Products Society, United States. pp. 2.1-2.4 (1999). 4. Stokke, D. and Gardner, D., “Fundamental aspects of wood as a component of thermoplastic composites,”Journal of Vinyl Additive Technology, 9 (2): 96 (2003). 5. Bulian, F. and Graystone, J., Wood Coatings: Theory and Practice, Elsevier Science, San Diego. p. 320 (2009). 6. Clemons, C.M., Functional Fillers for Plastics, Wiley VCH, Weinhem. pp. 213-23 (2010). 7. Kim, J.K. and Pal, K., Recent Advances in the Processing of Wood-Plastic Composites, Springer, London. p. 176 (2010). 8. Henton, D.E., Gruber, P., Lunt, J. and Randall, J., Polylactic Acid Technology. Natural Fibers, Biopolymers, and Biocomposites, Taylor & Francis, Boca Raton, FL: pp. 528-69 (2005). 9. Li, Y.J. and Shimizu, H., “Toughening of polylactide by melt blending with a biodegradable poly(ether) urethane elastomer”, Macromolecular Bioscience, 7: 921-28 (2007). 10. Garlotta, D., “A literature review of poly(lactic acid),” Journal of Polymers and the Environment, 9 (2): 63-84 (2001). 11. Lee, S.T., Kareko, L., and Jun, J., “Study of thermoplastic PLA foam extrusion,” Journal of Cellular Plastics, 44 (4): 293-305 (2008). 12. Kobayashi, S., and Sugimoto, S., “Biodegradation and mechanical properties of poly(lactic acid)/poly(butylenes succinate) blends,” Journal of Solid Mechanics and Materials Engineering, 2 (1): 15-24 (2008). 13. Lee, S.Y., Kang, I.A., Doh, G.H., Yoon, H.G., Park, B.D. and Wu, Q., “Thermal and mechanical properties of wood flour/talc-filled polylactic acid composites: effect of filler content and coupling treatment.” Journal of Thermoplastic Composite Materials, 21 (3): 209–23 (2008). 14. Pilla, S., Gong, S., O’Neill, E., Rowell, R.M., and Krzysik, A.M., “Polylactide-pine wood flour composites,” Polymer Engineering and Science, 48: 578-87 (2008). 57 15. Oksman, K., Skrifvars, M. and Selin, J.F., “Natural fibers as reinforcement in polylactic acid (PLA) composites,” Composites Science and Technology, 63: 1317–24 (2003). 16. Jacobsen, S. and Fritz, H. G., “Plasticizing polylactide - The effect of different plasticizers on the mechanical properties,” Polymer Engineering and Science, 39: 1303-10 (1999) 17. Liu, H., and Zhang, J., “Research progress in toughening modification of poly(lactic acid),” Polymer Physics, 49: 1051-83 (2011). 18. Auras, R., Harte, B. and Selke, S., “An overview of polylactides as packaging materials,” Macromolecular Bioscience, 4: 835-64 (2004). 19. Jiang, X., Luo, Y., Tian, X., Huang, D., Reddy, N. and YANG, Y., Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons, Inc.: Hoboken, N.J. pp.69-82 (2010). 20. Nampoothiri, K.M., Nair, N. R., and John, R. P., “An overview of the recent developments in polylactide (PLA) research,” Bioresource Technology, 101: 8493-8501 (2010). 21. Perego, G., Cella, G. D. and Bastioli, C., “Effect of molecular weight and crystallinity on poly (lactic acid) mechanical properties,” Journal of Applied Polymer Science, 59: 37-43 (1996). 22. Harris, A.M., and Lee, E.C., “Improving mechanical performance of injection molded PLA by controlling crystallinity,” Journal of Applied Polymer Science, 107 (4): 2246-55 (2007). 23. Lutz, J.T., and Dunkelberger, D.L., Impact Modifiers for PVC; The History and Practice, John Wiley and Sons, Inc.,: Canada. p. 1-87 (1992). 24. Dorgan, J.R., Braun, B., Wegner, J.R. and Knauss, D.M., “Poly(lactic acids): a brief review. in degradable polymers and materials,” American Chemical Society, 102-25, (2006). 25. Bouchareb, B., and Benaniba, T., “Effects of epoxidized sunflower oil on the mechanical and dynamical analysis of the plasticized poly(vinyl chloride),” Journal of Applied Polymer Science, 107: 3442-50, (2008). 26. Detyothin, S., Kathuria, A., Jaruwattanayon, W., Selke, S.E.M., and Auras, R., Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons, Inc.: Hoboken, N.J. pp.227-71 (2010). 27. Painter, P.C., and Coleman, M.M., Fundamentals of Polymer Science: An Introductory Text, CRC Press: Washington, D.C. p. 478 (1997). 28. Pillin, I., Montrelay, N., and Grohens, Y., "Thermo-mechanical charaterization of plasticized PLA: Is the miscibility the only significant factor?” Polymer, 47: 4676-82, (2006). 58 29. Lin, Y., Zhang, K.Y., Dong, Z.M., Dong, L.S., and Li, Y.S., “Study of hydrogen-bonded blend of polylactide with biodegradable hyperbranched poly (ester amide),” Macromolecules, 40 (17): 6257-67 (2007). 30. Schreck, K.M., and Hillmyer, M.A., “Block copolymers and melt blends of polylactide with Nodax™ microbial polyesters: preparation and mechanical properties,” Journal of Biotechnology, 132 (3): 287–95 (2007). 31. Anderson, K.S., Shawn, H.L., Hillmyer, M.A., “Toughening of polylactide by melt blending with linear low-density polyethylene”, Journal of Applied Polymer Science, 89: 3757-68 (2003). 32. Murariu M., Ferreira A.D., Pluta M., Bonnaud L., Alexandre M., Dubois, P., “Polylactide (PLA)-CaSO4 composites toughened with low molecular weight and polymeric ester-like plasticizers and related performances,” European Polymer Journal, 44: 3842-52 (2008). 33. Baird, J.C., Christiano, J.P., Morris, B.A., “An extrusion study: examination of the improved processing characteristics of a PLA impact modified blend,” ANTEC, 155-160 (2009). 34. Ishida, S., Nagasaki, R., Chino, K., Dong, T., and Inoue, Y., “Toughening of poly (L-lactide) by melt blending with rubbers,” Journal of Applied Polymer Science, 113 (1): 558-66 (2009). 35. Gao, X., Qu, C., and Fu, Q., “Toughening mechanism in polyoxymethylene/thermoplastic polyurethane blends,” Polymer International, 53 (11): 1666-71 (2004). 36. Matuana, L.M., “Solid state microcellular foamed poly (lactic acid): morphology and property characterization,” Bioresource Technology, 99 (9): 3643-50 (2008). 37. Matuana, L.M., Faruk, O. and Diaz, C.A., “Cell morphology of extrusion foamed poly (lactic acid) using endothermic chemical foaming agent,” Bioresource Technology, 100: 5947-54 (2009). 38. Matuana, L.M. and Diaz, C.A., “Study of cell nucleation in microcellular poly (lactic acid) foamed with supercritical CO2 through a continuous-extrusion process,” Industrial & Engineering Chemistry Research, 49 (5): 2186-93 (2010). 39. Lee, S.T., Kareko, L., Jun, J., “Study of thermoplastic PLA foam extrusion,” Journal of Cellular Plastics, 44: 293–305 (2008). 40. Mihai, M., Huneault, M.A., Favis, B.D., and Li, H., “Extrusion foaming of semicrystalline PLA and PLA/thermoplastic starch blends,” Macromolecular. Bioscience, 7 (7): 907–20 (2007). 41. Kramschuster, A., Pilla, S., Gong, S., Chandra, A., and Turng, L., “Injection molded solid and microcellular polylactide compounded with recycled paper shopping bag fibers,” International Polymer Processing, 12 (5): 436–445 (2007). 59 42. Reignier, J., Gendron, R., and Champagne, M.F., “Extrusion foaming of poly (lactic acid) blown with CO2: toward 100 % green material,” Cellular Polymers 26 (2): 83–115 (2007). 43. Vanvuchelen, J., Perugini, C., Deweerdt, M., Chen, L., and Burnham, T., “Microcellular PVC foam for thin wall profile,” Journal of Cellular Plastics, 36 (2): 148–57 (2000). 44. Matuana, L.M., Park, C.B. and Balatinecz, J.J., “The effect of low levels of plasticizer on the rheological and mechanical properties of polyvinyl chloride/newsprint-fiber composites,” Journal of Vinyl & Additive Technology, 3 (4): 265-73 (1997). 45. Park, B.D. and Balatinecz, J.J., “Mechanical properties of wood-fiber/toughened isotactic polypropylene composites,” Polymer Composites, 18: 79-89 (1997). 46. Nourbakhsh, A. and Ashori, A., “Fundamental studies on wood–plastic composites: Effects of fiber concentration and mixing temperature on the mechanical properties of poplar/PP composite,” Polymer Composites, 29: 569-73 (2008). 47. Clemons, C., “Wood-Plastic composites in the United States: The interfacing of two industries”, Forest Products Journal, 52: 10 (2002). 48. Lin, Q., Zhou, X. and Dai, G., “Effect of hydrothermal environment on moisture absorption and mechanical properties of wood flour–filled polypropylene composites,” Journal of Applied Polymer Science, 85 (14): 2824-32 (2002). 49. Matuana, L.M., Cam, S., Yuhasz, K.B. and Armstrong, Q.J., “Composites of acrylonitrilebutadiene-styrene filled with wood-flour,” Polymers & Polymer Composites, 15 (5): 343-8 (2007). 50. English, B.W., and Falk, R.H., Proceedings of the woodfiber-plastic conference, Forest Products Society, May 1-3, 1995, Madison, Wisconsin. 51. Oksman, K. and Clemons, C., “Mechanical properties and morphology of impact modified polypropylene-wood flour composites,” Journal of Applied Polymer Science, 67: 1503-13 (1998). 52. Mengeloglu, F., Matuana, L.M., and King, J.A., “Effects of impact modifiers on the properties of rigid PVC/wood-fiber composites,” Journal of Vinyl and Additive Technology, 6 (3):153-7 (2000). 53. Afrifah, K.A., Hickok, R.A., and Matuana, L.M., “Polybutene as a matrix for wood plastic composites,” Composites Science Technology, 70:167-72 (2010). 54. Wolcott, M. P., and Englund, K., “A Technology Review of Wood-Plastic Composites,” 33rd International Particleboard/Composite Materials Symposium, Lecture conducted from Washington State University, Pullman, WA. April 13-15, 1999. 60 55. Stark, N., “Effect of species and particle size on properties of wood-flour-filled polypropylene composites,” Functional Fillers for Thermoplastics & Thermosets. Intertech Conference, San Diego, CA. December 8-10, 1997. 56. Espert, A., “Natural fibres/polypropylene composites from residual and recylced materials: Surface modification of cellulose fibers, properties, and environmental degradation,” KTH Fiber - och polymerteknologi, 100 (44), 1-36 (2003). 57. Zazyczny, J.M., and Matuana, L.M., Fillers and Reinforcing Agents, PVC handbook Hanser: Munich. pp. 235-72 (2005). 58. Ghosh, B., Bandyopadhyay-Ghosh, S., and Sain, M., Composites, Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons Inc.: Hoboken. pp. 293-307 (2010). 59. Shibata, M., Biodegradable Polymer Blends and Composites from Renewable Resources, John Wiley & Sons, Inc.: Hoboken, N.J. pp.287-99 (2009). 60. Carlborn, K., and Matuana, L.M., “Influence of processing conditions and material compositions on the performance of formaldehyde-free wood-based composites,” Polymer Composites, 27 (6): 599-607 (2006). 61. Carlborn, K., and Matuana, L.M., “Modeling and optimization of formaldehyde-free woodbased composites using a Box-Behnken Design,” Polymer Composites, 27(5): 497-503 (2006). 62. Kazayawoko, M., Balatinecz, J.J., and Matuana, L.M., “Surface modification and adhesion mechanisms of woodfiber-polypropylene composites,” Journal of Materials Science, 34: 6189-99 (1999). 63. Carlborn, K., and Matuana, L.M., “Functionalization of wood particles through a reactive extrusion process,” Journal of Applied Polymer Science, 101(5): 3131-42 (2006). 64. Jiang, L., Wolcott, M.P., and Zhang, J., “Study of biodegradable polylactide/poly(butylene adipate-co-terephthalate) blends,” Biomacromolecules, 7 (1): 199-207 (2006). 65. Yin, L., Shi, D., Liu, Y., and Yin, J., “Toughening effects of poly (butylene terephthalate) with blocked isocyanate-functionalized poly (ethylene octene),” Polymer International, 58: 919-26 (2009). 66. Smith, T.L., “Fracture of polymers in biaxial and triaxial tension,” Polymer Science Symposium, 32: 269-82 (1971). 67. Liang, Z., and Li, R.K.Y., “Rubber toughening in polypropylene: A review,” Journal of Applied Polymer Science, 77: 409-17 (2000). 68. Rosler, J., Harders, H., and Baker, M., Mechanical Behaviour of Engineering Materials: Metals, Ceramics, Polymers, and Composites, Springer: New York. p. 534 (2007). 61 69. Griffith, A.A., “The phenomena of rupture and flow in solids,” Philosophical Transactions of the Royal Society, London, 221: 163-98 (1920). 70. Arencon, D. and Velasco, J.I., “Fracture toughness of polypropylene-based particulate composites,” Materials, 2: 2046-94 (2009). 71. Kumar, S. and Barai, S.V., Concrete Fracture Models and Applications, Springer-Verlag: Berlin Heidelberg. p 262 (2011). 72. Dagang, L., Xiaojun, X., Jiansheng, D., Qing, S., Zhengyuan, W. and Feng, W., “Fracture behavior of wood-plastic composites,” presented at 9th International Conference on Wood & Biofiber Plastic Composites, Madison, WI (2007). 73. Wu, J. and Mai, Y.W., “The essential fracture work concept for toughness measurement of ductile polymers,” Polymer Engineering and Science, 36 (18): 2275- 88 (1996). 74. Reincke, K., Grellmann, W., Lach, R. and Heinrich, G., “Toughness optimization of SBR elastomers – use of fracture mechanics methods for characterizationa,” Macromolecular Materials and Engineering, 288: 181-89 (2003). 75. Williams, J.G. and Cawood, M.J., “European group on fracture: Kc and Gc methods for polymers,” Polymer Testing, 9: 15-26 (1990). 76. Wu, J., Mai, Y.W. and Cotterell, B., “Fracture toughness and fracture mechanisms of PBT/PC/IM blend,” Journal of Materials Science, 28: 3373-84 (1993). 77. Rice, J., “A path independent integral and approximate analysis of strain concentration by notches and cracks,” Journal of Applied Mechanics, 2: 379-86 (1968). 78. Dongming, L., Wenge, Z. and Zongneng, Q., “The J-integral fracture toughness of PP/CaCO3 composites,” Journal of Materials Science, 29: 3754-58 (1994). 79. Ha, C.S., Kim, Y. and Cho, W.J., “Fracture mechanics investigation on the PP/EPDM/Ionomer ternary blends using J-integral by locus method,” Journal of Applied Polymer Science, 51: 1381-88 (1994). 80. Frassine, R., Rink, M. and Pavan, A., “Size effects in the fracture of a pipe-grade high density polyethylene,” Fatigue and Fracture of Engineering Materials and Structures, 20: 1217-23 (1997). 81. Sumpter, J.D.G., and Turner, C.E., “Application of J to elastic-plastic materials,” International Journal of Fracture, 9: 320-21 (1973). 62 CHAPTER 3 IMPACT MODIFICATION OF POLYLACTIDE WITH A BIODEGRADABLE ETHYLENE-ACRYLATE COPOLYMER This chapter is slightly modified from Macromolecular Materials and Engineering, published in September 14, 2010. 295 (9): 802-811. It is co-authored by K.A. Afrifah and L.M. Matuana. 63 3.1 Abstract The effectiveness and efficiency of an ethylene/acrylate copolymer in toughening semicrystalline and amorphous PLA through melt blending is studied. The mechanical properties, phase morphologies, miscibilities, and toughening mechanisms of the blends are assessed. The ethylene/acrylate impact modifier effectively improved the impact strength of the blends, regardless of the PLA type. The semicrystalline blends showed decreased tensile strength and modulus with increased impact modifier content. In contrast, the ductility, elongation at break, and energy to break increased significantly. The relatively low BDT temperature obtained for the PLA blends renders the ethylene/acrylate copolymer impact modifier a desirable additive to toughen PLA for use in cold temperatures. 64 3.2 Introduction The continued increase in oil prices, U.S. dependency on foreign oil, and environmental concerns about the use of common petroleum-based plastics have recently led to a growing interest in biobased plastics (1,2). Polylactic acid (PLA) an aliphatic and compostable polyester, comprising completely of renewable resources has attracted much attention as a popular alternative to traditional petroleum-based plastics. Various factors contribute to the success of PLA as alternative to traditional petroleum-based plastics, including physical properties, favorable compostable and degradation characteristics as well as its ability to maintain carbon dioxide balance after its decomposition (3-5). Currently, annual crops such as corn and sugar beets predominate as feedstock in PLA’s commercial production. The presence of two stereogenic centers in the lactide monomer allows for the formation of both amorphous and semicrystalline forms of PLA (3,6-9). Historically, the high cost of PLA production limited its use to the specialty biomedical niches such as sutures and drug delivery devices due to its biodegradability and biocompatibility (6). Presently technological innovations have slightly decreased PLA production cost while the recent increasing crude oil and natural gas prices has conversely increased the cost of traditional petroleum derived plastics, enhancing PLA economic competitiveness (6). More recently, PLA has been used as packaging materials for food and consumer goods (transparent bottles, meat trays, bags, films, etc.), with the polymer typically discarded after use. These applications benefit from their ability to decompose relatively quickly in landfill or compost environments. Unfortunately, PLA has not been used extensively beyond these areas due to its brittleness and lower impact resistance at room temperature (1,2), resulting in splitting and other handling problems during sheet manufacture (8,10-12). The necessity for improvement in its impact 65 strength and toughness, which would eliminate the processing and handling deficiencies, precedes the possibility for widespread utilization and substitution for commodity plastics. The toughening of PLA includes blending with other polymers (5-7,9-12), an exceptional method that is industrially relevant (13). Blending of PLA with rubbery polymers, as previously mentioned, has predominantly emphasized biomedical applications, resulting in the use of biodegradable and biocompatible polymers such as poly(vinyl alcohol), poly(ε-caprolactone), poly(ethylene glycol), polyhydroxyalkanoate, and poly(butylene succinate) as second phase polymers (7,12). These impact modifying additives are relatively exorbitant rendering their PLA blends expensive. Improving the impact strength of PLA by direct mechanical blending with inexpensive non-degradable polymers such as poly(ethylene oxide), poly(vinyl acetate), polyisoprene, acrylonitrile-butadiene-styrene copolymer and polyethylene has successfully reduced cost and expanded its commercial applications (7,10). However, the majority of these blends require compatibilizers to improve the miscibility between the impact modifier and the PLA. Therefore, using a commercially cost-effective impact modifier which is compatible with PLA would produce greater advantages. In response to this challenge, DuPont Packaging and ® Industrial Polymers has introduced Biomax strong 100 . This is a petroleum-based ethyleneacrylate copolymer for commercial applications, noted as compatible with PLA and maintains its biodegradability at low concentrations (14). The feasibility of using this ethylene-acrylate copolymer as an impact modifier has been explored in a few studies and reported in literature. One study produced and analyzed not only PLA blend films with either Sukano PLA im S550 or ethylene-acrylate copolymer as additives for tensile properties and clarity but also correlated these properties to their micro-structures. 66 Ethylene-acrylate copolymer at a loading rate of 12 % resulted in elongation at break of 255 %, a significant improvement over that of neat PLA with about 90 % (15). Another study reported the use of ethylene-acrylate copolymer as an impact modifier for neat PLA and highly filled PLA–calcium sulfate composites (16). Addition of 5 to 10 % ethylene-acrylate copolymer into highly filled composites (30 to 40 % of filler) led to a threefold increase of impact strength with respect to the compositions without the modifier (16). However, these studies failed to extensively account for the more subtle considerations of this impact modifier on different grades of PLA (effectiveness) and concentrations of impact modifier (efficiency) (17) as well as providing detailed fracture mechanisms of impact modified PLA samples tested under various environmental conditions. Consequently, this study investigated the mechanisms involved in the toughening of PLA with this ethylene-acrylate copolymer. The basis for considering both semicrystalline and amorphous PLA grades lay in the significant role crystallinity of PLA plays in the mechanical and durability performance in rigid molded applications (8). The specific objectives of this research included assessing the efficiency, effectiveness and ductility of blends of ethyleneacrylate copolymer with semicrystalline and amorphous grades of PLA using notched Izod impact and tensile tests. Evaluation also included the miscibility, phase morphology, interfacial adhesion between PLA matrix and ethylene-acrylate copolymer and temperature effect on notched Izod impact strength for a semicrystalline PLA grade. 67 3.3 Experimental Part 3.3.1 Materials The PLA resins from NatureWorks (USA) consisted of two grades of semicrystalline (PLA 2002D and PLA 3001D) and one amorphous grade (PLA 8302D). Table 3.1 lists the properties of these resins as measured by Matuana and coworkers (2). Dupont Packaging and Industrial Polymers supplied the impact modifier, ethylene-acrylate copolymer or Biomax strong ® 100 , used in experimentation. Documented typical characteristics of this ethylene-acrylate o o copolymer includes: melting point: 72 C; glass transition temperature: -55 C; MFI (190 o C/2.16kg): 12g/10 min and elongation at break: 950 % (16). 68 Table 3.1. Characteristics of different PLA resins used in study (2). PLA types Density of solid Melt Properties Thermal properties 3 (g/cm ) MFR (g/10 min) χc Tm (g/cm ) (%) ( C) ρm 3 o 2002D – semicrystalline 1.256 3.4 1.142 15 149 3001D – semicrystalline 1.254 12.1 1.144 15 167 8302D – amorphous 1.257 6.5 1.149 1.2 - 69 3.3.2 Blending and molding of PLA/ethylene-acrylate copolymer o All of the polymers were dried in an oven at 55 C for at least 24 hours before processing. The PLA and impact modifier were first mechanically mixed in a tray at room temperature. With the exception of PLA 3001D which had 8 compositions of the blends, 7 compositions of the blends that varied by the proportion of impact modifier at 0, 3, 5, 10, 15, 20, and 30 wt.-% based on the total weight of the blends were prepared for all three PLA grades. The additional composition of PLA 3001D blend had 40 wt.-% impact modifier. After mechanical mixing, the materials were melt blended in a 32 mm conical counter-rotating twinscrew extruder (C.W. Brabender Instruments Inc.) with a length to diameter ratio of 13:1 fitted with a rectangular profile die with dimensions of 1 cm by 2.5 cm. A 5.6 kilowatt (7.5 hp) IntelliTorque Plasti-Corder Torque Rheometer® (C.W. Brabender Instruments Inc.) powered the extruder. Starting from the hopper to the die, the melt blending temperature profile was: 180o 180-175-170 C. The melt blended materials were cooled for about an hour at room temperature and granulated in a Conair Wortex granulator (Model JC-5). The granulated samples were then dried o in an oven at 55 C for about 12 hours and injection molded using a BOY 30T2 equipment to produce 3 mm thick specimens for tensile (Type I, ASTM D638) and notched Izod impact testing. The temperature profile, from the hopper to the nozzle used for the injection process, o o was set: 180-180-175-170 C. The mold temperature of the injection equipment set at 50 C was controlled by a water temperature controller (Advantage Sentra, LE Series Model SK-1035LE). 70 3.3.3 Differential scanning calorimetry (DSC) Thermal analysis of the blends was carried out on a Q200 differential scanning calorimeter from TA instruments. Powdered samples of about 10 mg were heated in a o o temperature range of -70 to 200 C at a heating and cooling rate of 10 C/min. The glass transitions (Tg), melting temperatures (Tm) and crystallinities recorded during the second heating scan ensured consistent thermal histories. The percent crystallinity (χc) of the blends and unfilled PLA was determined using the following equation (2): χ c (% ) = ∆H m − ∆H c 0 ∆H m (1) .100 0 with, ΔHm as the melt enthalpy, ΔHc as the enthalpy for the cold crystallization and ∆H m = 93 J/g as the enthalpy of fusion of a PLA crystal of infinite size (3). 3.3.4 Scanning electron microscopy (SEM) Studying the fracture mechanisms of the blends used an examination of the morphology of notched Izod impact-fractured surfaces with a JOEL JSM-6400 scanning electron microscope with an acceleration voltage of 10 kV at a magnification factor of 3630x. SEM micrographs were taken after coating the surfaces with a thin layer of gold. 71 3.3.5 Mechanical property evaluation o A conditioning room at 23 ± 2 C and 65 ± 4 % relative humidity housed the samples for mechanical properties evaluation for at least 48 hours prior to testing. The conditioning room also served as the location for the tensile and notched Izod impact tests. Tensile properties were measured with Instron 5585H testing machine using the Instron Bluehill 2, version 2.14 software and in accordance with the procedures outlined in ASTM standard D638. Each composition included ten tested replicates to obtain a reliable mean and standard deviation using a crosshead speed of 5 mm/min. From the tensile results, the ductility or the materials ability to undergo plastic deformation without fracture was determined, using the following relation (18): Ductility = failure strain yielding strain (2) Notched Izod impact tests were performed at room and sub-ambient temperatures in conformance to ASTM standard D256 on Tinius Olsen Izod impact tester (Model 892). The ten o specimens tested for each composition and temperature were V-notched at 45 angle using Tinius Olsen specimen notcher (Model 899). The specimens for the study of temperature effect on notched Izod impact strength contained 15 wt.-% ethylene-acrylate copolymer impact 3 3 modifier. They were cooled in a freezer (General Electric 0.38 m or 13.4 ft chest freezer) set o at various sub-ambient temperatures (-27, -21, -15, -3.2, 1.2, 6.9, and 23 C) for at least 8 hours and tested before any significant warming could occur in the specimens. 72 3.4 Results and Discussion 3.4.1 Miscibility and crystallinity of PLA/ethylene-acrylate copolymer blends The glass transition temperatures (Tg) of the blends were of particular interest to determine whether the two components were miscible or immiscible. While the Tg of each polymeric component does not change in immiscible mixtures, generally perfect miscible blends exhibit a single broad Tg located between the Tgs of the individual components of the blend (1923). The final Tg of miscible binary blends can be predicted by the following well known Fox equation as a function of the weight fraction of the two polymeric components and their Tgs, respectively (24): 1 W W = 1 + 2 Tg Tg1 Tg 2 (3) with Tg1 and Tg2 as the Tg’s of the pure components in Kelvin, and W1 and W2 as their respective weight fractions present in the mixture (24). Figure 3.1 shows the DSC thermograms of neat PLA, neat ethylene-acrylate copolymer, and PLA blended with different impact modifier contents after crystallizing from melt. Figure 3.2 illustrates the DSC curve of pure ethylene-acrylate copolymer. As seen in these figures, the DSC curves of neat PLA and its blends with ethylene-acrylate copolymer show three apparent 73 transitions upon heating, which includes PLA glass transition temperature, cold crystallization exotherm peak and melting endotherm peak. The thermal properties of pure components and blends calculated from these thermograms are summarized in Table 3.2. 74 0.2 PLA Tg Exo -0.2 Heat flow (W/g) a -0.6 b c d -1.0 e -1.4 -1.8 -100 -50 0 50 100 150 200 o Temperature ( C) Figure 3.1. DSC thermograms of semicrystalline PLA (3001D) and its blends with various concentrations of ethylene-acrylate copolymer: a) 0, b) 5, c) 15, and d) 20 wt.-%. The curve labeled e) represents the thermogram of pure ethylene-acrylate copolymer. 75 Heat flow (W/g) Exo 0.26 0.16 0.06 Tg -0.04 -0.14 -100 -50 0 50 100 o Temperature ( C) Figure 3.2. DSC curve of pure ethylene-acrylate copolymer. 76 150 200 Table 3.2. Thermal properties of neat PLA (3001D), pure ethylene-acrylate copolymer (EAC), and PLA/EAC blends. o Tm Materials ΔHm ΔHc χc o Tg ( C) ( C) (J/g) (J/g) (%) Measured Theoretical Pure EAC -48.7 -55 72.1 - - - Neat PLA 60.1 60.1 167.8 25.4 11.5 15 PLA/5 wt.-% EAC 60.2 52.1 168.1 29.0 17.1 13 PLA/15 wt.-% EAC 59.7 37.5 168.1 31.9 23.0 10 PLA/20 wt.-% EAC 60.1 30.6 168.2 30.4 21.7 9 77 All blends showed a single Tg (Figure 3.1), which may indicate miscible binary system. However, the appearance of the single Tg in the blends did not meet the most frequently used criterion for miscibility (19-23) since the Tg of the blends did not occur at a temperature intermediate between those of pure components (Table 3.2). The single Tgs of the blends did not change as predicted by the Fox equation (Table 3.2), instead not only did they remain similar to the Tg of the neat PLA but also did not change with increasing impact modifier content. These results imply that the two components were immiscible in the blends with a two-phase morphology lacking significant molecular interactions (25). The SEM micrographs of the PLA and its blends (presented later in this paper) support this conclusion, as they show a clear phaseseparated morphology with impact modifier dispersed in the PLA matrix. It should be pointed out that, the two distinct separate Tgs, typical of immiscible binary blends were not clearly identified in the thermograms of the blends illustrated in Figure 1 mainly due to the insensitivity of DSC technique to monitor the Tg of pure ethylene-acrylate copolymer. Diluting this impact modifier into PLA reduced its heat flow in DSC thermograms (Figure 3.2), making its Tg identification difficult. Similar trends were reported by other investigators (20,23,26). Several researchers (19-23) have reported the limitations of DSC in studying the miscibility/immiscibility of a two phase system containing a small amount of a second component. Studying immiscibility in polymer blends, Jorda and Wilkes (20) showed that observation of the second component’s Tg or the effect of this minor component on the Tg of the 78 more dominant component may not be always easily measured. Another limitation of DSC technique lies with the difficulty to easily distinguish the two separate Tgs for a binary blend o where the Tg of each component differs only slightly (~10-20 C) (22). Since the close proximity of Tg between pure PLA and pure ethylene-acrylate copolymer was not the case in this study (Table 3.2), it is believed that the no real distinct appearance of two Tgs in the DSC scans of the blends shown in Figure 3.1 could be attributed to the lack of DSC sensitivity in monitoring the Tg of the impact modifier in the blends. The heat of cold-crystallization (ΔHc) generally increased with increasing impact modifier concentration in the blends (Table 3.2), which indicates the decreased crystalline ability of PLA. This implies that the introduction of ethylene-acrylate copolymer limited the formation of crystals in the PLA, leading to lower crystallinities for the blends as shown in Table 3.2. The melting temperature of the neat PLA was unaffected by the addition of the ethylene-acrylate copolymer into the matrix. 3.4.2 Effectiveness and efficiency of impact modifier (ethylene-acrylate copolymer) in PLA Figure 3.3 illustrates the effect of impact modifier content on the notched Izod impact strength of the semicrystalline and amorphous PLA grades. Because the impact strength of the PLA increased with ethylene-acrylate copolymer content, irrespective of the PLA type, ethyleneacrylate copolymer effectively acts as an impact modifier. Additives that contribute to damping and improve impact resistance, which include rubbers, acrylic impact modifiers, plasticizers, and 79 so on accounts for these expected results. Generally, reports suggest the overall performance and damping properties (i.e., the ability of a material to dissipate impact energy by converting it into heat) of the material subjected to impact loading is primarily dependent on its composition (27). Consequently, introduction of 30 wt.-% of the ethylene-acrylate copolymer into the semicrystalline and amorphous PLA resulted in significant changes in Izod impact strength of about 594 % and 372 %, respectively. Detailed statistical analyses of variance are presented in Appendices A.1 to A.6 A supertough behavior, where there was a dramatic increase in impact strength of the blend was observed with PLA 3001D when the ethylene-acrylate copolymer content was increased to 40 wt.-% (Figure 3.3). Notched Izod impact strength of semicrystalline PLA increased in two steps with increase in ethylene-acrylate copolymer content in the range of 0 to 40 wt.-%. A continuous increase in impact strength was observed by increasing the impact modifier concentrations from 0 wt.-% (16.9 J/m) to 20 wt.-% (88.0 J/m) after which it increased at a lower rate to 30 wt.-% (118.2 J/m) and then increased sharply at 40 wt.-% (348.4 J/m) representing about 195 % increase over the impact strength of the blend with 30 wt.-%. The immiscibility of the ethylene-acrylate copolymer with the matrix polymer, which allows for a rubbery polymer to impart toughness to PLA or any other polymer, accounts for this positive characteristics of the blend. This consequently allows the rubber to induce energy dissipation mechanisms into PLA, which retard crack initiation and propagation, and ultimately result in a material with improved toughness (28). Additional explanation will be provided in Section on the morphology of the impact fractured surface and mechanisms of toughening. 80 Figure 3.3. Effect of ethylene-acrylate copolymer (EAC) content on the notched Izod impact strengths of both amorphous (8302D) and semicrystalline (2002D and 3001D) PLA. Error bars are not shown in the graphs to prevent overlapping of the curves. However, the values of standard deviation for the data shown in this figure were below 10 % of the mean values of the notched Izod impact strength. 81 Figure 3.3 also shows an obvious brittle-to-ductile transition (BDT) of the notched Izod impact strength for the PLA with an increase in impact modifier content. The impact strengths of the blends remained nearly unchanged up to 5 wt.-% ethylene-acrylate copolymer content, regardless of the PLA grade, and then increased remarkably with increasing ethylene-acrylate copolymer content (Figure 3.3). Experimentation found the threshold impact modifier content at which the transition from brittle to ductile behavior in PLA blends occurred at 8 wt.-% for semicrystalline PLA (2002D and 3001D) and 12 wt.-% for amorphous PLA (8302D). The efficiency of ethylene-acrylate copolymer as an impact modifier varied within the three grades of PLA. Impact modifier efficiency, understood as the contribution to impact resistance of the blend per unit of modifier, indicates the impact resistance obtained per part of impact modifier used. The classic output over input equation determines this as follows (17,29): Modifier efficiency = impact strength impact modifier content (4) Above 5 wt.-%, superior efficiency of the impact modifier resulted in the semicrystalline PLAs to the amorphous due to the minute amount of the ethylene-acrylate copolymer needed to initiate a significant increase in impact strength of the semicrystalline PLA than the amorphous. For example, by drawing a horizontal line parallel to the X-axis (ethylene-acrylate copolymer content) at an arbitrary impact strength value (e.g., 50 J/m in Figure 3.3), the data can produce two intersection points. The first intersection point detected at lower impact modifier content for semicrystalline PLA (2002D and 3001D) and the next seen at higher impact modifier content for 82 amorphous PLA (8302D), which clearly indicates the higher efficiency of the ethylene-acrylate copolymer in semicrystalline PLA. Since a reduction in impact resistance accompanies greater crystallinity within a family of materials, the experimental results came as a surprise. (27). However, our results agree with those reported by Baird (11), which shows a higher spencer impact of PLA sheets modified with ethylene-acrylate copolymer in the semicrystalline as opposed to the amorphous. For instance, at a 2 % impact modifier, the spencer impact of 3,500 J and 2,600 J were attained for the semicrystalline and amorphous PLA, respectively (27). Other investigators (7) also observed a similar trend for PLA toughened with 20 % linear low-density polyethylene (LLDPE), reporting the impact strengths of 34 J/m and 350 J/m for the amorphous and semicrystalline, respectively. The increased toughness in the semicrystalline PLA blends could potentially be attributed to the presence of overlapping crystalline layers. Investigations suggest that in semicrystalline blends, the rubber/matrix interface nucleates crystallization of the matrix producing crystallographically oriented material at the interface region (7,30). At ligament thickness below the critical value for the matrix in question, the oriented layers merge into a percolating material component with reduced plastic resistance throughout the matrix resulting in tough behavior (31). Therefore, with all other conditions for impact modification fulfilled, the semicrystalline blend should surpass the amorphous one in toughness because of the presence of oriented layer around the rubber particles. The low critical brittle-to-ductile transition (BDT) concentrations of 8 wt.-% for semicrystalline PLA blends compared to the 12 wt.-% of the amorphous blends confirms the greater efficiency of the impact modifier in the semicrystalline PLAs to the amorphous. Consequently, cost/performance advantages contribute to the desirability of the semicrystalline grades for commercial applications with the ethylene-acrylate copolymer as impact modifier. 83 Although the effectiveness and efficiency of this impact modifier were relatively similar in both semicrystalline PLA grades, the injection molding grade (3001D) was selected for further studies, instead of the extrusion one (2002D). 3.4.3 Tensile properties Figure 3.4 illustrates the stress-strain curves of PLA (3001D) and its blends with 10 and 15 wt.-% ethylene-acrylate copolymer. Tensile property data including tensile modulus, energy to break and ductility were evaluated from these stress-strain curves. Observations indicate that the fracture behavior of the specimens fluctuated from brittle fracture in the neat PLA (0 wt.-% impact modifier) to ductile fracture in the blends with 10 and 15 wt.-% impact modifier (Figure 3.4). Neat PLA includes characteristics such as extreme rigidity and brittleness and has a tensile strength of 63 MPa and elongation at break of 4.4 %. It showed a distinct yield point (maximum load) with subsequent failure by neck instability. By contrast, all the blends with ethyleneacrylate copolymer content above 5 wt.-% showed distinct yielding and stable neck growth through cold drawing. The samples finally broke at a significantly increased elongation, compared to that of the neat PLA (Figure 3.4). Analyses of variance of the significance of the effect of EAC on the tensile properties of the PLA/EAC blends are presented in Appendices A.7 to A.12. 84 Figure 3.4. Influence of ethylene-acrylate copolymer (EAC) content on the tensile stress strain curves of semicrystalline PLA (3001D). 85 Figure 3.5 shows the energy to break and ductility as a function of impact modifier content and Figure 3.6 presents the effect of impact modifier concentration on tensile strength and modulus of PLA and its blends. The general increase in energy to break and ductility [measured from Equation (2)] of the blends with increase in ethylene-acrylate copolymer content confirmed the toughening capacity of the impact modifier (Figure 3.5). As shown in Figure 3.5, the addition of up to 5 wt.-% ethylene-acrylate copolymer into PLA matrix affected neither the ductility, nor the energy to break of the samples. Whereas, above this concentration, a significant increase in both ductility and energy to break occurred, which dropped slightly in the blend with 20 wt.-% impact modifier. Alternatively, the tensile strength and modulus of the PLA decreased almost linearly with increasing ethylene-acrylate copolymer content (Figure 3.6). Tensile strength and modulus decreased by 38 % and 31 %, respectively by adding 20 wt.-% ethylene-acrylate copolymer in the blends. The lower tensile strength and modulus of neat ethylene-acrylate copolymer compared to those of neat PLA produced this anticipated consequence. As the concentration of the ethylene-acrylate copolymer increases, it becomes the dominant component of the blend, and the compounds become softer and flexible, with a lower tensile strength and modulus (32). For this reason, in rubber-toughened blends, observations denote the modulus and impact strength as competing properties (33). 86 Figure 3.5. Effect of ethylene-acrylate copolymer (EAC) content on the energy to break and ductility of semicrystalline PLA (3001D). Some error bars did not appear in the graph due to the smaller values of standard deviation. 87 Figure 3.6. Effect of ethylene-acrylate copolymer (EAC) content on the tensile strength and modulus of semicrystalline PLA (3001D). Some error bars did not appear in the graph due to the smaller values of standard deviation. 88 3.4.4 Morphology of the impact fractured surface and mechanisms of toughening Scanning electron micrographs taken on the surface of impact fractured samples assisted in examining the morphology and understanding the fracture mechanisms in PLA blends. Figure 3.7 shows the SEM micrographs of semicrystalline neat PLA (PLA 3001D) and its blends with various concentrations of ethylene-acrylate copolymer (3 to 40 wt.-%). The blends have a twophase morphology, a necessary condition for toughening (34-35). The SEM images show a uniform distribution and dispersion of the ethylene-acrylate copolymer in the PLA matrix (Figure 3.7b-h). A gradual increase in domain size, change in shape from spherical to ellipsoid and broadening of particle size distribution with increasing ethylene-acrylate copolymer content also prevails. Particle size distribution broadens at higher concentrations (Figure 3.7d-h) because breakup and coalescence occur simultaneously during blending (36). Investigation has also shown that the elongated/ellipsoid domain particles act as path arresters during impact loadings particularly at higher impact modifier concentrations, rendering good withholding of impact properties even at low temperatures (35). Particle coalescence could account for the increase in the modifier’s particle domain size at high concentrations. 89 a b 10 µm 10 µm c d 10 µm 10 µm e f 10 µm 10 µm h g 10 µm Figure 3.7. 10 µm Scanning electron micrographs of impact fractured surfaces of semicrystalline PLA (3001D) and its blends with various concentrations of ethylene-acrylate copolymer: a) 0, b) 3, c) 5, d) 10, e) 15, f) 20, g) 30, and h) 40 wt.-%. 90 Several studies have reported occurrence of various energy-dissipative mechanisms locally in a stressed specimen before catastrophic crack development (10,25,33,37,38). Mechanisms such as crazing, bond rupture, cavitation, crack growth, plastic and viscoelastic deformations, etc. relieve stresses and consequently reduce the stored elastic energy. High strength and toughness therefore, result primarily from special combination of these mechanisms that retard or arrest the growth of cracks (37). Using fractured surfaces of notched Izod impact tested samples as a means of differentiating brittle to ductile fractures, previous studies reported noticeable whitening only occurs at the origin of the notched tip for brittle fracture, while ductile fracture involves all of the material around the fractured surface in stress whitening and forms a yielding zone (33). In rubber-toughened plastic systems two types of cavitations induced by impact or tensile tests are discerned, which includes internal cavitations in the rubber domains for the blends with strong interfacial adhesion and debonding cavitations between the interfaces with insufficient interfacial adhesion (10). Generally, the mechanisms causing damage during brittle fracture have been prescribed as mainly crazing or microcracks (leading to stress whitening) and cavitation (38). Crazing, cavitation, shear banding, crack bridging and shear yielding reportedly occur as important energy dissipation processes involved in the impact fracture of toughened polymer systems (25). The surfaces of fractured neat PLA samples (Figure 3.7a) and its blends with up to 5 wt.% ethylene-acrylate copolymer (Figures 3.7b and 3.7c) remained relatively flat and smooth without any signs of plastic deformation, which are typical characteristics of brittle failure. The images of neat PLA (Figure 3.8a) and its blend with 5 wt.-% ethylene-acrylate copolymer (Figure 3.8b) clearly verify the brittleness effect since complete breakage occurred in all impact tested samples. Moreover, evident cavitation caused by debonding and stress whitening near the 91 notched tip due to crazing/microcracks resulted. Although debonding absorbs a considerable amount of fracture energy, a limited effect resulted in these samples due to a lower concentration of the ethylene-acrylate copolymer particles used in the blends (up to 5 wt.-%), which caused mainly the matrix to bear most of the stress (38). Lower impact modifier content implies a higher distance between particles in the blends compared to the desired critical inter-particle distance achieved at the brittle-ductile transition (13,39). This high inter-particle distance accounts for the brittle fracture observed in PLA blends containing up to 5 wt.-% ethyleneacrylate copolymer. These results corroborate the notched Izod impact strength data illustrated in Figure 3.3, where the addition of up to 5 wt.-% ethylene-acrylate copolymer into PLA did not toughen the matrix. By contrast, with 10 wt.-% or more ethylene-acrylate copolymer into PLA, observations included evidences of ductile fracture in the SEM images (Figure 3.7d-h). Blends with 10 wt.-% or more ethylene-acrylate copolymer experienced partial breakage when impact tested (Figure 3.8c), a sign of improved toughness and ductility, which can be attributed to several fracture mechanisms as described below. 92 sw sw (a) Figure 3.8. (b) (c) Photographs of notched Izod impact tested samples showing fracture modes: Complete breakage for both a) neat PLA and b) PLA blended with 5 wt.-% ethylene-acrylate copolymer, and c) partial breakage for PLA blended with 20 wt.-% ethylene-acrylate copolymer. sw in the figure stands for stress whitening. ‘For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation.’ 93 Visible factors illustrated stress whitening of the entire fractured surface due to crazing (Figure 3.8c), forming of yielding zone, debonding of the particle/matrix interface and deformation of the impact modifier particles (Figure 3.7d-h). The observed plastic deformation of the PLA matrix at the fractured region implies that shear yielding of the PLA matrix had occurred (Figure 3.8c), i.e., whitening at all around the fracture surface. Additionally, evidence showed the apparent attraction of the impact modifier to fibrillar morphology (Figure 3.7d-h), a process which absorbs a substantial amount of energy and tends to bridge cracks (40). The fibrils increased in number and length with increase in impact modifier content, hence increasing its capacity to bridge and retard the propagation of cracks. The dispersed domains of the impact modifier act as stress concentrators under impact stress due to the difference in their elastic properties compared to the PLA matrix. This promotes disruptive processes; however, the evenly dispersed impact modifier’s domains play the role of preventing the growth of cracks to a critical size by initiating the disruptive processes in a uniform manner throughout the specimen instead of in a few isolated regions. The stress concentration leads to the development of triaxial stress in the impact modifier particles. Due to insufficient interfacial adhesion between the impact modifier domains and the PLA matrix, interfacial debonding occurred instead of cavitation within the core of the impact modifier particles under triaxial stress. Once debonding occurred, the stress state of the PLA matrix surrounding the voids altered, triaxial tension released locally in the surrounding voids and the yield strength lowered. With debonding progression, PLA matrix strands between impact modifier particles deformed more easily to achieve shear yielding (10,25). The intense debonding in the blends at all concentrations of the ethylene-acrylate copolymer evidently portrays weak interfacial bonding between the PLA matrix and the impact modifier domain particles (Figure 3.7b-h). 94 In summary, up to 5 wt.-% impact modifier (ethylene-acrylate copolymer), the fracture of the blends occurred through crazing/microcracking and debonding of modifier’s particles with the matrix mainly bearing all the stress due to the fewer number of modifier’s particles in the blends resulting in brittle failure. Whereas in the blends with 10 wt.-% or more impact modifier, their fracture arose through crazing, impact modifier fibrillization, crack bridging, debonding cavitations, and matrix shear yielding resulting in a ductile behavior. 3.4.5 Temperature effect on impact strength of impact modified PLA The measured notched Izod impact strength of semicrystalline neat PLA (3001D) and its blend with 15 wt.-% ethylene-acrylate copolymer tested at sub-ambient and room temperatures also determined the effectiveness of the impact modifier at cold temperatures. The choice of the 15 wt.-% impact modifier lays in its high impact strength and also the correspondingly highest elongation at break, energy to break and ductility determined at room temperature compared to the other concentrations (Figures 3.3-3.5). Figure 3.9 depicts the response of the PLA blend to Izod impact strength as a function of test temperature. Analysis of variance of the effect of temperature on the impact strength of the neat PLA and the PLA/EAC blend can be found in Appendices A.13 to A.16. 95 o ( C) Figure 3.9. Influence of testing temperature on the notched Izod impact strength of semicrystalline PLA (3001D) and its blend with 15 wt.-% ethylene-acrylate copolymer (EAC). Some error bars did not appear in the graph due to the smaller values of standard deviation. 96 Neat PLA displayed brittle fracture having almost the same impact strength values at all tested temperatures (Figure 3.9). Suppositions assumed the actual BDT temperature for the neat PLA as situated near its Tg. The findings of Hassan and Haworth (41) supports this contention in their work on the influence of temperature on impact properties of acrylate rubber-modified PVC. Their findings denoted the unmodified PVC as the first to undergo the BDT when the o temperature decreased in the range between 60 and 80 C, just below the glass transition temperature of PVC. Thereafter, the impact strength remained constant at lower temperatures. Jansen (42) reported that implementation of most semicrystalline polymers predominates at temperatures between their Tg and melting point (Tm). Above the Tm, the resin becomes liquid and below the Tg, the resin loses the kinetic energy required for the amorphous tie molecules to move in response to applied stress. The polymer then lacks the capacity to undergo substantial deformation and yielding, thus it undergoes molecular disentanglement and exhibits brittle properties. Therefore, the glass transition temperature of a neat semicrystalline polymer (e.g., o 60.1 C for PLA in Table 3.2) represents its BDT. Consequently, the BDT of neat PLA was not observed in this study because the testing temperature range used was below its Tg (Figure 3.9). By contrast, PLA blended with 15 wt.-% ethylene-acrylate copolymer significantly surpassed the unmodified PLA in notched Izod impact strength, irrespective of the testing temperature (Figure 3.9). In addition, the BDT of this blend occurred at lower testing o temperature. In fact, a clear BDT transition was observed at around -8 C (Figure 3.9). This BDT represents, a transition in major deformation mechanisms from shear yielding to crazing, or vice versa, accompanied by a sudden change in crack resistance (35). Therefore, the addition of 97 ethylene-acrylate copolymer reduces the BDT temperature of PLA well below end use requirements due to its low Tg (-55 o C) compared to the neat PLA (60.1 98 o C). 3.5 Conclusion Two grades of semicrystalline PLA and a grade of amorphous PLA were each melt blended with various concentrations of ethylene-acrylate copolymer impact modifier using twin screw extruder. The extrudates were pelletized and injection molded into samples tested for tensile and impact properties. The ethylene-acrylate copolymer toughened all three grades of PLA. The efficiency of the ethylene-acrylate copolymer in the semicrystalline grades superseded the amorphous grade. Thus the semicrystalline grades recorded lower impact modifier content for the brittle-to-ductile transition compared to the amorphous grade. A detailed analysis of the injection molding grade of the semicrystalline PLA revealed increasing ductility, elongation at break, and energy to break with ethylene-acrylate copolymer content compared to neat PLA. Contrarily, the tensile strength and modulus decreased with ethyleneacrylate copolymer content. SEM and DSC analyses indicated the blends as an immiscible two phase system. The fracture of PLA and its blends with up to 5 wt.-% impact modifier occurred through crazing or microcracking and debonding of modifier’s particles with the matrix mainly bearing all the stress due to the low content of modifier’s particles resulting in brittle failure. For the blends with 10 wt.-% or more impact modifier, the fracture mechanisms included impact modifier debonding, fibrillization, crack bridging and matrix shear yielding resulting in a ductile behavior. A relatively low brittle-to-ductile transition temperature obtained for the PLA blended with 15 wt.-% ethylene-acrylate copolymer confirms that this impact modifier is a good additive for toughening PLA for use in sub-ambient temperatures. 99 REFERENCES 100 3.6 References 1. Matuana, L.M., “Solid state microcellular foamed poly(lactic acid): morphology and property characterization,” Bioresource Technology, 99 (9): 3643-50 (2008). 2. Matuana, L.M., Faruk, O., and Diaz, C.A., “Cell morphology of extrusion foamed poly(lactic acid) using endothermic chemical foaming agent,” Bioresource Technology, 100 (23): 594754 (2009). 3. Garlotta, D., “A literature review of poly(lactic acid),” Journal of Polymers and the Environment, 9 (2): 63-84 (2001). 4. Lee, S.T., Kareko, L., and Jun, J., “Study of thermoplastic PLA foam extrusion,” Journal of Cellular Plastics, 44 (4): 293-305 (2008). 5. Kobayashi, S., and Sugimoto, S., “Biodegradation and mechanical properties of poly(lactic acid)/poly(butylenes succinate) blends,” Journal of Solid Mechanics and Materials Engineering, 2 (1): 15-24 (2008). 6. Schreck, K.M., and Hillmyer, M.A., “Block copolymers and melt blends of polylactide with Nodax™ microbial polyesters: preparation and mechanical properties,” Journal of Biotechnology, 132 (3): 287–95 (2007). 7. Anderson, K.S., Shawn, H.L., Hillmyer, M.A., “Toughening of polylactide by melt blending with linear low-density polyethylene”, Journal of Applied Polymer Science, 89: 3757-68 (2003). 8. Harris, A.M., and Lee, E.C., “ Improving mechanical performance of injection molded PLA by controlling crystallinity,” Journal of Applied Polymer Science, 107 (4): 2246-55 (2007). 9. Murariu M., Ferreira A.D., Pluta M., Bonnaud L., Alexandre M., Dubois, P., "Polylactide (PLA)-CaSO4 composites toughened with low molecular weight and polymeric ester-like plasticizers and related performances,” European Polymer Journal, 44: 3842-52 (2008). 10. Li, Y., and Shimizu, H., “Toughening of polylactide by melt blending with a biodegradable poly(ether) urethane elastomer,” Macromolecular Bioscience, 7: 921-28 (2007). 11. Baird, J.C., Christiano, J.P., Morris, B.A., “An extrusion study: examination of the improved processing characteristics of a PLA impact modified blend,” ANTEC, 155-160 (2009). 12. Ishida, S., Nagasaki, R., Chino, K., Dong, T., and Inoue, Y., “Toughening of poly(L-lactide) by melt blending with rubbers,” Journal of Applied Polymer Science, 113 (1): 558-66 (2009). 13. Gao, X., Qu, C., and Fu, Q., “Toughening mechanism in polyoxymethylene/thermoplastic polyurethane blends,” Polymer Internationl, 53 (11): 1666-71 (2004). 14. Dupont, Biomax Resins, Material Safety Data Sheet, 11/28/2006. 15. Zhu, S., Rasal, R., and Hirt, D., “Polylactide toughening using impact modifiers,” ANTEC, 1616-20 (2009). 101 16. Murariu, M., Ferreira, A.D.S., Duquesne, E., Bonnaud, L., and Dubois, P., “(Bio)degradable polymers from renewable resources,” Macromolecular Symposia, 272 (1): 1-12 (2008). 17. Lutz, J.T., and Dunkelberger, D.L., Impact Modifiers for PVC; The History and Practice, John Wiley and Sons, Inc.,: Canada. p. 1-87 (1992). 18. Shackelford, J.F., Introduction to Materials Science for Engineers, Macmillan Publishing Company: New York. p. 605 (1985). 19. Thirtha, V.M., Lehman, R.L., and Nosker, T.J., In Proceedings of the ANTEC 2005, Boston, MA, 2005; p 2380 20. Jorda, R., and Wilkes, G.L., “A novel use of physical aging to distinguish immiscibikity in polymer blends,” Polymer Bulletin, 20 (5): 479-85 (1988). 21. Icoz, D.Z., and Kokini, J.L., “Probing the boundaries of miscibility in model carbohydrates consisting of chemically derivatized dextrans using DSC and FTIR spectroscopy,” Carbohydrate Polymers, 68 (1): 68-76 (2007). 22. Cameron, N., Cowie, J.M.G., Ferguson, R., Gomez Ribelles, J.L., and Mas Estelles, J., “Transition from miscibility to immiscibility in blends of poly(methyl methacrylate) and styrene–acrylonitrile copolymers with varying copolymer composition: a DSC study,” European Polymer Journal, 38 (3): 597-605 (2002). 23. Horrion, J., Jerome, R., Teyssie, Ph., Vanderschueren, J., and Corapci, M., “Thermally stimulated currents studies in blends of immiscible telechelic polymers,” Polymer Bulletin, 21 (6): 627 (1989). 24. Painter, P.C., and Coleman, M.M., Fundamentals of Polymer Science: An Introductory Text, CRC Press: Washington, D.C. p. 478 (1997). 25. Jiang, L., Wolcott, M.P., and Zhang, J., “Study of biodegradable polylactide/poly(butylene adipate-co-terephthalate) blends,” Biomacromolecules, 7 (1): 199-207 (2006). 26. Lin, Y., Zhang, K.Y., Dong, Z.M., Dong, L.S., and Li, Y.S., “Study of hydrogen-bonded blend of polylactide with biodegradable hyperbranched poly(ester amide),” Macromolecules, 40 (17): 6257-67 (2007). 27. Jansen, J.A., “Ductile-to-brittle transition of plastic materials,” Advanced Materials and Processes, 25-27 (February 2007). 28. NatureWorks, Technology Focus Report; Toughened PLA, Ver.3/1/2007. 29. McMurrer, M.C., “Update: impact modifiers for rigid PVC,” Plastic Compounding, 77-86 (1983). 30. Muratoglu, O.K., Argon, A.S., Cohen, R.E., and Wienberg, M., “ Toughening mechanism of rubber-modified polyamides,” Polymer, 36: 921-30 (1995). 102 31. Bartczak, Z., Argon, A.S., Cohen, R.E., and Wienberg, M., “Toughness mechanism in semicrystalline polymer blends: I. high-density polyethylene toughened with rubbers,” Polymer, 40: 2331-46 (1999). 32. Mengeloglu, F., Matuana, L.M., and King, J.A., “Effect of impact modifiers on the properties of rigid PVC/wood-fiber composites,” Journal of Vinyl and Additive Technology, 6 (3):15357 (2000). 33. Yin, L., Shi, D., Liu, Y., and Yin, J., “Toughening effects of poly(butylene terephthalate) with blocked isocyanate-functionalized poly(ethylene octene),” Polymer International, 58: 919-26 (2009). 34. Ratna, D., and Banthia, A.K., “Toughening of epoxy resin by modification with 2-ethylhexyl acrylate-acrylic acid copolymers,” Polymer International, 49: 309-15 (2000). 35. Das, V., Pandey, A.K., and Krishna, B., “Low temperature izod impact studies of blends based on impact grade polypropylene and ethylene-α-octene copolymer,” Journal of Reinforced Plastics and Composites, 28: 2879-88 (2009). 36. Sundararaj, U., and Macosko, C.W., “Drop breakup and coalescence in polymer blends: the effects of concentration and compatibilization,” Macromolecules, 28: 2647-57 (1995). 37. Smith, T.L., “Fracture of polymers in biaxial and triaxial tension,” Polymer Science Symposium, 32: 269-82 (1971). 38. Liang, Z., and Li, R.K.Y., “Rubber toughening in polypropylene: A review,” Journal of Applied Polymer Science, 77: 409-17 (2000). 39. Zhao, C., Wu, G., Zhou, C., Yang, H., and Zhang, H., “Independence of the brittle-ductile transition from the rubber particle size for impact-modified poly(vinyl chloride),” Journal of Polymer Science, 44: 696-702 (2006). 40. Zhou, C., Wang, K., and Fu, Q., “Toughening of polyamide 11 via addition of crystallizable polyethylene derivatives,” Polymer International, 58:538-44 (2009). 41. Hassan, A., and Haworth, B., “Impact properties of acrylate rubber-modified PVC: influence of temperature,” Journal of Materials Processing Technology, 172:341-45 (2006). 42. Jansen, J.A., “Ductile-to-brittle transition of plastic materials,” Material Science Forum, 3942 (February 2006). 103 CHAPTER 4 STATISTICAL OPTIMIZATION OF TERNARY BLENDS OF POLY(LACTIC ACID)/ETHYLENE-ACRYLATE COPOLYMER/WOOD-FLOUR COMPOSITES This chapter is slightly modified from Macromolecular Materials and Engineering, published online in August 25, 2011. 296 DOI:10.1002/mame.201100097. It is co-authored by K.A. Afrifah and L.M. Matuana.   104 4.1 Abstract The effects of wood flour content and particle size as well as ethylene-acrylate copolymer (EAC) impact modifier content on the mechanical properties of PLA/wood-flour composites were studied using a two-level factorial design. Increasing the EAC content enhance the impact strength and elongation at break, but reduce the tensile modulus and strength of the composites. Composites with fine wood particles show greater improvement in elongation at break than those with coarse particles; an opposite trend is observed for impact strength, tensile modulus and strength. Numerical optimization produced two scenarios based on materials compositions to produce composites with similar mechanical properties as unfilled PLA. These optimization solutions have been successfully validated experimentally.   105 4.2 Introduction Biodegradable plastics produced from renewable resources are gaining prominence due to their biomass origin and ability to decompose and maintain carbon dioxide balance (1,2). Polylactic acid (PLA), a plastic from starch of corn and sugar beets, has attracted much attention as replacement for traditional petroleum-based thermoplastics because it possess mechanical properties similar to some petroleum-based plastics and can be easily processed by existing plastic processing equipment (1-6). However, PLA is more expensive and has several drawbacks such as brittleness and lower impact resistance compared to conventional plastics such as PE and PP (1,3-6). Several investigators have recently studied the mechanical properties of PLA filled with cellulosic fibers (2-7) in an attempt to reduce the cost of PLA. The reduced cost due to addition of the cellulosic fibers is achieved at the expense of mechanical properties such as lowered impact strength, toughness, and elongation at break since the incorporated brittle cellulosic fibers alter the ductile mode of failure of the matrix making the composites more brittle than the neat polymer (8-13). Impact modification and/or plasticization are well known approaches to enhance the impact resistance and toughness of composites (4,14-17). However, the research conducted so far on PLA/wood-flour composites has mostly focused on the effects of wood flour content, coupling agent and plasticizer types and addition levels on the mechanical properties (2-7,18) without due consideration of the effects of impact modifier content and the particle size of the fiber.   106 For the full exploitation of the use of wood flour in PLA/wood-flour composites, the effects of wood flour particle size and content as well as impact modifier addition level on the mechanical properties must be determined. In view of the multiple variables involved, the traditional technique of varying one-factor at a time while holding other variables constant would not be a good option to assess the interactions between factors as it is time consuming and often easily misses the interaction effects between factors. A statistical technique capable of developing a mathematical model that describes the relationships between the responses of interest and independent variables in which the significance of individual factors and multifactor interactions can be determined should be used (19-22). Therefore, this study used a two-level factorial design to evaluate the effects of wood flour particle size (mesh size), wood flour content, and impact modifier concentration on the impact strength and tensile properties of PLA/wood-flour composites. The objective was to understand the interactions between the composites’ material composition variables and develop model equations establishing the relationships between these variables and the properties of the composites. Numerical optimization was also performed with the focus of attaining mechanical properties for the PLA/wood-flour composites equivalent to at least those of unfilled PLA.   107 4.3 Experimental Section 4.3.1 Materials The PLA resin used in this study (PLA 3001D) was purchased from NatureWorks, USA. -3 It had a density of 1.144 g.cm , 8 % of D-lactide enantiomer, crystallinity of 15 %, and melt flow index of 12.1 g.(10 min) -1 o at 190 C (12). An ethylene-acrylate (EAC) impact modifier (Biomax strong 100) supplied by DuPont Packaging and Industrial Polymers, was used as the o o impact modifier. It had a melting point of 72 C, glass transition temperature of -55 C and a -1 melt flow index of 12 g (10 min) o at 190 C (17). American Wood Fibers supplied the maple wood flours used in the composites. Their commercial grades were 2010, 4010, 6010, and 10010, which correspond to 20, 40, 60, and 100 nominal mesh sizes, respectively. 4.3.2 Compounding and injection molding ® A two level factorial design was formulated using Design Expert software versions V.6.0 and V.7.0 (Stat-Ease Corp. Minnesota). The three studied variables included wood flour content (5–40 % based on the total weight of the composite), wood flour particle size (mesh size 20–100), and impact modifier (EAC) content (0–30 % based on the total weight of PLA), whereas the obtained responses included the impact strength and tensile properties. The experimental design matrix is listed in Table 4.1. It shows the design in the two different methods of displaying the levels of factors that are: (i) the actual levels of factors or the actual   108 values in the experiment and (ii) the coded factor levels, which is presented as -1 for low levels, +1 for high levels, and 0 for center point. The coded factor levels are defined as (19-22): Coded factor levels  actual value  factor mean range of the factorial value/2 (1) The outputs generated after the inclusion of all responses in the design were first order response models. These models described the main effects and interactions on the response in terms of coded variables as follows (22): 3 3 Y  β0   βi Xi   i1 3  βijXi X j i1 j1 (2) where Y is the predicted response, Xi the main effect, XiXj the interaction and βo the intercept. βi and βij are the regression coefficients and are one-half of the corresponding factor effect estimates.   109 Table 4.1. ® Experimental design matrix in terms of actual and coded factor levels generated by Design-Expert software. Factors Experiment Serial Number Type 1 Responses Wood Flour content (%) Mesh Size EAC Content (%) Impact Strength -1 (J.m ) Tensile Properties Strength (MPa) Modulus (GPa) Elongation at break (%) Fact 5 (-1) 20 (-1) 0 (-1) 17.21 61.34 1.71 5.32 2 Fact 40 (+1) 20 (-1) 0 (-1) 15.60 60.24 2.87 3.29 3 Fact 5 (-1) 100 (+1) 0 (-1) 16.81 62.50 1.65 6.22 4 Fact 40 (+1) 100 (+1) 0 (-1) 6.49 64.26 2.73 3.64 5 Fact 5 (-1) 20 (-1) 30 (+1) 80.10 34.56 1.04 6.96 6 Fact 40 (+1) 20 (-1) 30 (+1) 32.76 25.97 1.55 4.03 7 Fact 5 (-1) 100 (+1) 30 (+1) 61.06 35.59 1.03 14.47 8 Fact 40 (+1) 100 (+1) 30 (+1) 14.54 25.85 1.34 5.39 9 Center 22.5 (0) 60 (0) 15 (0) 24.78 35.65 1.37 6.18 110   o o The PLA and wood flour were first dried at 55 C and 105 C, respectively, for at least 24 h before melt blending. The studied compositions were melt blended in a 32 mm conical counter-rotating twin-screw extruder with a length-to-diameter ratio of 13:1 (C.W. Brabender Instruments Inc. South Hackensack, NJ) fitted with a rectangular profile die. The extruder was powered with a 5.6 kW (7.5 hp) intelli-Torque Plasti-Corder Torque Rheometer (C.W. Brabender Instruments Inc. South Hackensack, NJ). After the melt blending, the extrudates were cooled at room temperature for about an hour and then granulated in a Conair Wortex granulator o (model JC-5). The granulated samples were dried for at least 24 h at 55 C and injection molded using a BOY 30T2 equipment to produce 3 mm thick test samples that conform to ASTM D638 (Type I) and ASTM D256 standards for tensile and notched Izod impact testing, respectively. The temperature profile used for both the extruder (from hopper to die) and injection molder o o (from hopper to the nozzle) was 180-180-175-170 C. The mold was heated at 50 C by a water temperature controller unit (Advantage Sentra, LE Series Model SK-1035LE) to prevent the injected samples from sticking in the injection equipment mold. 4.3.3 Mechanical property evaluation The evaluation of tensile properties (strength, modulus, and elongation at break) was carried out on an Instron 5585H testing machine using the Instron Bluehill 2, version 2.14 software. Tests were performed in accordance with ASTM standard D638 using a crosshead -1 speed of 5 mm.min . Notched Izod impact test was performed on a Tinius Olsen Izod impact tester (Model 892) in conformance with ASTM standard D256. The specimens were V-notched 111   o at a 45 angle using a Tinius Olsen specimen notcher (model 899). Prior to testing, the samples o were conditioned in a room at 23 ± 2 C and 65 ± 4 % relative humidity for at least 48 h and all tests were performed in this conditioning room. To obtain reliable means and standard deviations, at least 10 and 13 specimens were tested for the tensile properties and notched Izod impact strength, respectively. 4.3.4 Scanning electron microscopy (SEM) The morphology of the surface of impact fractured samples was studied to assess the mechanisms associated with impact failure in the various compositions of the PLA/wood-flour composites. The test was carried out using JOEL JSM-6400 SEM with an acceleration voltage of 12 kV at a magnification of 121x. The SEM micrographs were taken after coating the surfaces of the specimens with a thin layer of gold.   112 4.4 Results and Discussion 4.4.1 Statistical analysis of the impact strength of PLA/wood-flour composites Regression analysis on impact strength obtained the best fit model equation for the 2 2 experimental data with prob F < 0.05 and high R , adjusted R as well as adequate precision values (Table 4.2). Equation 3 describes the derived regression model predicting the impact strength in terms of coded factors: Impact strength  30.56  13.23A  5.86B  16.53C  10.25AC  3.48BC (3) where A is the wood flour content, B is the wood flour particle size, and C is the impact modifier content. These descriptions of the variables apply to all analyses in this paper. Except for the impact modifier content (factor C), wood flour content (factor A) and wood flour particle size (factor B) negatively affected the impact strength of the composites due to the negative algebraic signs of these factors (Equation 3). The results suggest that impact strength of the composites increased with impact modifier content; whereas opposite trends were observed by increasing wood flour content and decreasing its particle size.   113 Table 4.2. Analysis of variance (ANOVA) for two-level factorial model. Responses a) 2 2 b) F-Value Prob>F R Adjusted R Adequate Precision Impact strength (J.m ) 182.44 0.0001 0.9881 0.9827 49.8 Tensile strength (MPa) 3070.63 0.0001 0.9986 0.9983 156.8 Tensile modulus (GPa) 337.19 0.0001 0.9935 0.9906 66.3 Elongation at break (%) 74.54 0.0001 0.9830 0.9699 38.0 -1 a) Values of “Prob>F” less than 0.05 indicates significant model terms. b) Adequate precision measures the signal-to-noise ratio. A ratio greater than 4 is desirable for the model. 114   The perturbation plot (Figure 4.1) showed steeper plots for wood flour (factor A) and impact modifier (factor C) contents than for particle size (factor B); indicating a higher relative importance for factors A and C supporting their observed, relatively higher values of regression coefficients in Equation 3. All three main effect factors were involved in significant interactions (interactions AC and BC) implying that the effect of one factor depended on the level of the other. Hence these factors were investigated together (19,21,22). 115   Perturbation Perturbation Impact strength (J/m) Impact strength (J/m) 80.1 80 61.7 62 43 43.3 CC A A B B B B 24.9 25 A C A C 76.5 -1.0 -1.0 -0.5 -0.5 0.0 0.0 0.5 0.5 1.0 1.0 Deviation from Reference Point Deviation from reference point Figure 4.1.   Perturbation plots of impact strength of the composites against wood flour content (factor A), particle size (factor B), and impact modifier (EAC) content (factor C). 116 Three-dimensional plots illustrating the variation of the impact strength of the composites, with respect to the interaction between wood flour content and impact modifier content (interaction AC) are shown in Figure 4.2a and 4.2b. The impact strength of the composites increased significantly with impact modifier content, irrespective of wood flour content and particle size (Figure 4.2a and 4.2b). The increased impact strength may be explained by the role of impact modifier in toughening the matrix. In wood flour filled PP, impact modifiers have been reported to affect its morphology by either existing as separate phase in the matrix, partially or completely encapsulating the filler or a mixed condition of the above (15). Impact modifier particles existing as separate phase in composites, generally act as stress concentrators that initiate local yielding of the matrix avoiding brittle failure of the material (15). Our previous study showed that ethylene-acrylate copolymer (EAC) improves the impact strength of PLA (1). Existence of the EAC as separate phase in the matrix, which acted as stress concentrators, initiating the formation of a large number of energy dissipating microscopic crazes in the surrounding PLA matrix accounted for the improvement. Nevertheless, the extent of impact strength improvement was a strong function of EAC content (1). Up to 5 % EAC, brittle failure still occurred in the blends due to a fewer number of modifiers in the PLA/EAC blends, which left the matrix to solely bear the stress. By contrast, stress whitening, debonding and deformation shear yielding occurred in the blends with 10 % EAC or more, resulting in ductile failure (1). It is believed that these same mechanisms were responsible for the improvements witnessed in the impact modified PLA/wood-flour composites (Figure 4.2).   117 (a) Impact strength (J/m) 85 65 45 25 5 30 23 40 31 15 23 C: EAC content (%) 8 14 0   118 5 A: Wood flour content (%) (b) Impact strength (J/m) 85 65 45 25 5 30 23 40 31 15 23 C: EAC content (%) 8 14 0 Figure 4.2.   5 A: Wood flour content (%) Three-dimensional graphs of the variation of the impact strength of the composites as a function of the interaction between wood flour content (factor A) and impact modifier (EAC) content (factor C) with wood flour particle sizes (factor B) of (a) 20 and (b) 100 mesh sizes. 119 Both the wood flour content and particle size affected the efficiency of EAC. Lower impact strength values were obtained in composites with high wood flour content (Figure 4.2a and 4.2b), irrespective of wood particle size. The reduced impact strength is a typical trade-off of improving the stiffness of the matrix by incorporating wood flour into the PLA/impact modifier (EAC) matrix as will be discussed later on (8-13). The impact modifier was less efficient in the composites with fine particles (100 mesh) (Figure 4.2b) than in their counterparts with coarse particle size (20 mesh) (Figure 4.2a). This may be explained by the crack development and propagation mechanisms in the polymer, which depends on the strength of the interfacial regions between the components (11,16,23,24). Weak interfacial regions arise in composites due to poor interfacial adhesion between the hydrophilic wood flour and the hydrophobic polymer matrix. As a result, each fiber, acts as a discontinuity capable of initiating cracks in the composites (16,23). Since the finest particles (100 mesh) possess higher surface area per unit weight, their addition into PLA increases the number of weakened interfacial regions, thus making the composite more prone to crack propagation (25). Assessing composites without impact modifier depicted two distinct behaviors depending on wood flour particle size. Composites with coarse wood particles (20 mesh) did not show any significant changes in their impact strength with the increase in wood flour content (Figure 4.2a and Table 4.1). Whereas, impact strength of composites with fine wood flour (100 mesh) decreased slightly with wood flour content increase (Figure 4.2b and Table 4.1). This trend is not clearly seen in Figure 4.2a and 4.2b, although clearly shown in the data summarized in Table 4.1. Therefore, additional experiments were conducted to clarify these trends. The experiments assessed the effects of wood flour content and particle size on the impact strength of PLA/woodflour composites (Figure 4.3). The impact strengths of composites with coarse wood particles   120 (up to 40 mesh) were not significantly affected by the wood flour concentration (Figure 4.3a and 4.3b). Unlike fine particles, it is believed that coarse ones create fewer cracks in the composites due to their lower surface area per unit weight, reducing the propensity of brittle failure which decreases the impact strength. Conversely, the impact strengths of composites with finer wood particles (60 mesh and up) decreased drastically after adding 15 % or more wood flour into the matrix (Figure 4.3c and 4.3d). The observed trend is probably due to wood flour agglomeration in the matrix (Figure 4.4b), which resulted in stress concentration promoting crack propagation with low impact strength (18). It should be pointed out that, up to 5 %, the addition of wood flour did not affect the impact strength of the composites with fine particles (Figure 4.3c and 4.3d). Craze formation and the low amount of fiber (less cracks) (Figure 4.4a) may have accounted for the insensitivity of impact strength on wood flour content up to 5 %.   121 20 (a) Impact strength (J/m) Impact strength (J/m) 20 15 10 5 0 -10 0 10 20 30 40 50 60 (b) 15 10 5 0 -10 0 Wood flour content (% ) Wood flour content (% ) 20 (c) Impact strength (J/m) Impact strength (J/m) 20 15 10 5 0 -10 0 10 20 30 40 50 60 Wood flour content (% ) Figure 4.3.   10 20 30 40 50 60 (d) 15 10 5 0 -10 0 10 20 30 40 50 60 Wood flour content (% ) Effects of wood flour particle sizes [(a) 20 mesh, (b) 40 mesh, (c) 60 mesh, and (d) 100 mesh] and contents on the impact strength of PLA/wood-flour composites. 122 (a) Particle agglomeration (b) Craze 200 µm Figure 4.4.   200 µm Scanning electron micrographs of impact fractured surfaces of composites with 100 mesh particle size and wood flour contents of (a) 5 % and (b) 40 %. 123 4.4.2 Statistical analysis of the tensile properties of pla/wood-flour composites Regression analyses on tensile properties obtained best-fit model equations for the 2 experimental data with “Prob > F” values of less than 0.05, high R , and desirable adequate precision values (Table 4.2). The tensile strength data of the composites resulted in the power model, after dropping insignificant terms: Tensile strength 2.11  4.804  10004  1.225  10004 A  3.153  10004 C  1.232  10004 AC (4) The negative lambda value of the power model (Equation 4) indicates that the positive algebraic signs of the significant factors result in a decrease in tensile strength by increasing those factors. Based on the regression coefficients (Equation 4), factor A (wood flour content) had higher influence on the tensile strength than factor C (impact modifier content) due to its lower value of regression coefficient. These two significant factors were also involved in interactions and investigated together (Equation 4). Figure 4.5a show the cube graph of the variation of tensile strength of the composites as a function of wood flour content, wood flour particle size, and impact modifier content. In the absence of impact modifier, the tensile strength of the composites remained almost the same as the wood flour contents increased, insensitive of wood flour particle size. In general, the tensile strength of a polymeric matrix decreases with an increase in wood flour content due to the poor interfacial adhesion between the wood fiber and the matrix (3,8,11,18,24).   124 C: EAC content (%)   35.1 C+ 35.1 (a) 25.9 25.9 61.9 62.1 B+ B: Wood particle size C- A- 61.9 62.1 A+ B- A: Wood flour content (%)     1.36 B: Wood particle size 1.02   B+  (b) 2.71 1.66  1.53 1.05 C+ C: EAC content (%) B-  A- 1.70  2.89 A+ A: Wood flour content (%) 125 C- 14.5 5.4 (c) B: Wood particle size (c) Figure 4.5.   B+  6.2 3.6 7.0 4.0 C+ C: EAC content (%) B- A- 5.3 3.3 A+ A: Wood flour content (%) C  Cube graphs of the relationships between tensile properties and wood flour content (factor A), particle size (factor B), and impact modifier (EAC) content (factor C) for (a) tensile strength, (b) tensile modulus, and (c) elongation at break of PLA/wood-flour composites. 126 However, comparison between the tensile strength of the composites of this study and unfilled PLA (63 MPa) measured from our previous study (1) indicates that the addition of wood flour did not affect the tensile strength of PLA in the range of wood flour contents studied. These results are in agreement with those reported by other investigators (2,3). Pilla et al (3) obtained composites with tensile strength similar to that of unfilled PLA, irrespective of pine wood flour content (up to 40 %). This observation was attributed to the rough nature of the pine wood flour, which enhanced the interfacial adhesion between the particles and the PLA matrix. Regardless of wood particle size, the addition of impact modifier reduced the tensile strength of the composites especially at higher wood flour content (Figure 4.5a). This decreased trend was expected due to the plasticizing effect of the impact modifier, which makes the composites soft and flexible. (8,26). A linear equation (Equation 5) was obtained for the regression analysis on the tensile modulus of the composites: Tensile modulus  1.74  0.38A  0.053B  0.50C  0.035AB  0.18AC (5) The tensile modulus of composites increased with wood flour concentration (factor A), independent of both wood flour particle size (factor B) and impact modifier (EAC) content (factor C) (Figure 4.5b) as expected from the rule of mixtures (8,16). Similar results have been reported by other investigators (3,8,16). Nevertheless, reduction in tensile modulus occurred in impact modified composites due to the softening effect of the impact modifier, regardless of   127 wood flour content and particle size (15,16,26). The reduction in tensile modulus of impact modified composites was more pronounced at higher wood flour concentrations. The analysis of elongation at break, a measure to gauge the brittleness of a composite, resulted in a best-fit linear model for the three main factors: Elongation at break  6.17  2.08A  1.26B  1.55C  0.84AB  0.92AC  0.95BC  0.70ABC (6) The elongation at break of the composites was positively affected by wood flour particle size (factor B) and impact modifier (factor C) content and negatively by the wood flour content (factor A) (Equation 6). Wood flour content (factor A) had higher contribution to the elongation at break of the composites as shown by its higher regression coefficient, followed by factor C, with factor B having the least influence. All the three main effect factors in elongation at break were involved in interaction as shown in Equation 6. Generally, composites with fine wood flour particles (B+) exhibited superior elongation at break regardless of wood flour and impact modifier contents (Figure 4.5c). This is contrary to the impact strength data which is also a measure of toughness. A further investigation is therefore, required to explain this observation. An increase in wood flour content decreased the elongation at break of the composites regardless of the wood flour particle size and impact modifier content (Figure 4.5c). The higher degree of brittleness introduced by the stiffer wood particles, may have resulted in brittle composites (8-13,26). On the other hand, an increase in the impact modifier content increased the elongation at break of the composites irrespective of the   128 wood flour content. Nonetheless, the impact modifier was more efficient in the composite with fine particles. 4.4.3 Numerical optimization of the mechanical properties of PLA/wood-flour composites Numerical Optimization was carried out to determine the combinations of wood flour content, wood flour particle size, and impact modifier content that would result in targeted impact strength, tensile modulus and elongation at break (toughness). The targeted mechanical -1 properties for the composites were set at those of the unfilled PLA (impact strength of 17 J.m , tensile modulus of 1.96 GPa, and elongation at break of 4 %) as determined in our previous study (1). These targeted mechanical properties were selected to ensure that the composites will have at least impact strength, elongation at break, and stiffness similar to those of unfilled PLA. ® The optimization function in the Design Expert software was used for this analysis. This numerical optimization function is based on a desirability function, which transforms each response value to a desirability index (di). Three parameters (good, lower, and upper) define each desirability index. The program presents five desirability index options (minimum, maximum, target, in range, and equal to). After defining these settings, the desirability index varies between zero (worst case) and one (ideal case). Optimization results generated and ® presented according to the criteria settings made by Design Expert are a series of solutions that best maximize the desirability index (19,20).   129 Table 4.3 summarizes the optimization criteria settings used to optimize the tensile properties and the impact strength of the composites. The goals for the material compositions of the composites were set in range while the impact strength, tensile modulus, and elongation at break were set to targeted values as mentioned above (Table 4.3). The desirability functions of the impact strength (IS), tensile modulus or modulus of elasticity (MOE), and elongation at break (EB) were set as follows: (i) -1 if IS < 6.49 or IS > 80.1 J.m , MOE < 1.03 or MOE > 2.87 GPa, and EB < 3.29 or EB > 14.47 %, then di = 0 (worst case) (ii) if 6.49 ≤ IS ≤ 17, 1.03 ≤ MOE ≤ 1.95 and 3.29 ≤ EB ≤ 4, then di = 1 (ideal case). For all the optimization parameters, the weights of the upper and lower limits for the input factors were set at one and at three for the importance, which is a relative scale that weights each of the resulting values of dis in the desirability product. To obtain desirable results, ten cycles were run per optimization, with the epsilon value for the minimum difference in eliminating duplicate results set at its default value (19,20).   130 Table 4.3. Numerical optimization settings. Response Constraints Factors Goal Lower limit Upper limit Lower weight Upper weight Importance Wood flour content (%) In range 5 40 1 1 3 Mesh size In range 20 100 1 1 3 EAC content (%) In range 0 30 1 1 3 6.49 80.1 1 1 3 1.03 2.87 1 1 3 3.29 14.47 1 1 3 -1 a) Impact strength (J.m ) Target = 17 Tensile modulus (GPa) Target = 1.96 Elongation at break (%) Target = 4 a) a) a) Afrifah and Matuana (1). 131   Optimum solutions with desirability between 0.967 and 0.989 were produced for the tensile modulus, elongation at break, and impact strength (Table 4.4). The results show a variety of combinations of the wood flour content, particle size, and impact modifier (EAC) content for the composites that would result in at least similar impact strength, tensile modulus, and elongation at break to those of the unfilled PLA. The most preferable material compositions for the composites would be those providing the targeted mechanical properties and also reducing cost. Addition of wood fibers, apart from improving the stiffness of thermoplastic resins, also reduces their cost owing to its greater stiffness and lower price compared to the thermoplastic resins. As a result, high quantities of wood flour combined with low quantities of the matrix resin (PLA) and/or additives (EAC impact modifier) in the composites would be preferred. The optimization solutions resulted in the following two different scenarios in terms of material compositions for the composites to match the mechanical properties of unfilled PLA (Table 4.4): 1. The composite should contain a high concentration (40 %) of fine wood flour particles (70 mesh) and also include a high concentration of impact modifier (EAC) (18 % based on the total weight of PLA). 2. The composite should contain a low concentration (13 %) of coarse wood flour particles (30 – 36 mesh) without inclusion of impact modifier. The second scenario presents the best option in terms of cost reduction. Even though the wood flour content in this scenario (coarse particles) is lower than in scenario 1, its coarser particle size would reduce the grinding cost. Additionally, the exclusion of the more expensive impact modifier in the formulation could save cost of materials. 132   Table 4.4. Numerical optimization solutions for impact strength and tensile properties. Solutions Wood flour content (%) Mesh size EAC content (%) Impact strength -1 (J.m ) Tensile modulus (GPa) Elongation at break (%) Desirability 1 40.00 69.72 18.03 17.00 1.96 4.33 0.989 2 40.00 70.03 18.17 17.00 1.96 4.34 0.989 3 39.73 71.05 17.98 17.00 1.96 4.38 0.988 4 13.33 36.27 0.00 17.00 1.97 4.99 0.967 5 13.30 35.20 0.00 17.07 1.97 4.98 0.967 6 13.29 34.38 0.01 17.13 1.97 4.98 0.967 7 13.16 29.58 0.00 17.43 1.97 4.94 0.967 133   Since tensile strength, an important mechanical property of PLA/wood-flour composites was not considered in the above described numerical optimization, validation tests were carried out to verify the scenario that would provide tensile strength value equivalent to that of the unfilled PLA. Results of the validation test for the two scenarios of the optimization solutions are summarized in Table 4.5. The theoretical values shown in this table represent the predicted values obtained from the optimization solutions using the Design Expert ® software. The experimental data confirmed the numerical optimization solutions for the composites (Table 4.5). The results also show that, tensile strength of PLA/wood-flour composites similar to that of the unfilled PLA could be obtained using formulations with a low concentration of coarse wood flour particles without inclusion of the ethylene-acrylate impact modifier (scenario 2). 134   Table 4.5. Results of validation test for the two scenarios of the optimization solutions for the targeted mechanical properties of the PLA/wood–flour composites. Scenarios High concentration of fine wood a) flour with impact modifier Low concentration of coarse wood flour without impact b) modifier Theoretical Experimental Theoretical Experimental Impact strength -1 (J.m ) 18.6 22.4 ± 1.8 16.5 15.6 ± 2.4 Tensile strength (MPa) 31.5 34.6 ± 0.9 62.0 60.6 ± 0.6 Tensile modulus (GPa) 2.0 2.0 ± 0.04 2.0 2.4 ± 0.06 Elongation break (%) 4.2 3.8 ± 0.3 4.9 3.5 ± 0.1 Mechanical properties at a) Composite with 40 % of 60 mesh wood flour and 18 % EAC impact modifier content. b) Composite with 15 % of 40 mesh wood flour without impact modifier.   135 4.5 Conclusion Using two level factorial design analyses, this study investigated the effects of wood flour content, wood flour particle size, and impact modifier content on the impact strength and tensile properties of PLA/wood-flour composites. Statistical models describing the relationships between the material composition variables of the composites and their resulting mechanical properties were derived. Additionally, numerical optimization was used to determine the best combination of the material composition variables for producing composites with mechanical properties similar to those of the unfilled PLA. The impact strength and elongation at break of the composites were positively affected by the addition of the impact modifier, which negatively affected the tensile strength and modulus, irrespective of the wood flour particle size and content. Increasing the wood flour content however, improved the tensile modulus but decreased the elongation at break of the composites. On the other hand, the tensile and impact strengths of the composites with coarse wood particles (up to 40 mesh) without impact modifier were not affected by the wood flour content. Numerical optimization led to several different combinations of the material compositions needed to manufacture PLA/wood-flour composites with mechanical properties similar to those of the unfilled PLA. Two scenarios were observed from these combinations depending on the particle size of the wood flour. High wood flour and impact modifier contents are required for composites made with fine wood flour particles. Whereas, the formulation requires low wood flour content and excludes impact modifier for composites with coarse wood flour particles. These optimization solutions were successfully validated experimentally.   136 REFERENCES   137 4.6 References 1. Afrifah, K.A. and Matuana, L.M., “Impact modification of polylactide with a biodegradable ethylene/acrylate copolymer,” Macromolecular Materials Engineering, 295: 802-11 (2010). 2. Lee, S.Y., Kang, I.A., Doh, G.H., Yoon, H.G., Park, B.D. and Wu, Q., “Thermal and mechanical properties of wood flour/talc-filled polylactic acid composites: effect of filler content and coupling treatment,” Journal of Thermoplastic Composites, 21: 209-23 (2008). 3. Pilla, S., Gong, S., O’Neill, E., Rowell, R.M. and Krzysik, A.M., “Polylactide-pine wood flour composites,” Polymer Engineering and Science, 48: 578-87 (2008). 4. Oksman, K., Skrifvars, M. and Selin, J.F., “Natural fibres as reinforcement in polylactic acid (PLA) composites,” Composites Science and Technology, 63: 1317-24 (2003). 5. Huda, M.S., Drzal, L.T., Misra, M. and Mohanty, A.K., “Wood-fiber-reinforced poly(lactic acid) composites: evaluation of the physicomechanical and morphological properties,” Journal of Applied Polymer Science, 102: 4856-69 (2006). 6. Sykacek, E., Schlager, W. and Mundigler, N., “Compatibility of softwood flour and commercial biopolymers in injection molding,” Polymer Composites, 31: 443-51 (2010). 7. Li,Y., Venkateshan, K. and Sun, X.S., “Mechanical and thermal properties, morphology and relaxation characteristics of poly(lactic acid) and soy flour/wood flour blends,” Polymer International, 59: 1099-109 (2010). 8. Afrifah, K.A., Hickok, R.A. and Matuana, L.M., “Polybutene as a matrix for wood plastic composites,” Composites Science and Technology, 70: 167-72 (2010). 9. Matuana, L.M., Park, C. B. and Balatinecz, J. J., “Processing and cell morphology relationships for microcellular foamed PVC/cellulosic-fiber composites,” Polymer Engineering & Science, 37 (7): 1137–47 (1997). 10. Matuana, L.M., Park, C.B. and Balatinecz, J.J., “The effect of low levels of plasticizer on the rheological and mechanical properties of polyvinyl chloride/newsprint-fiber composites,” Journal of Vinyl & Additive Technology, 3 (4): 265-273 (1997). 11. Matuana, L.M., Woodhams, R.T., Balatinecz, J.J. and Park, C.B., “Influence of interfacial interactions on the properties of PVC/cellulosic fiber composites,” Polymer Composites, 19 (4): 446-55 (1998). 12. Matuana, L.M., Faruk, O. and Diaz, C.A., “Cell morphology of extrusion foamed poly (lactic acid) using endothermic chemical foaming agent,” Bioresource Technology, 100: 5947-54 (2009). 13. Matuana, L.M. and Diaz, C.A., “Study of cell nucleation in microcellular poly(lactic acid) foamed with supercritical CO(2) through a continuous-extrusion process,” Industrial & Engineering Chemistry Research, 49 (5): 2186-93 (2010).   138 14. Park, B.D. and Balatinecz, J.J., “Mechanical properties of wood-fiber/toughened isotactic polypropylene composites,” Polymer Composites, 18: 79-89 (1997). 15. Oksman, K. and Clemons, C., “Mechanical properties and morphology of impact modified polypropylene-wood flour composites,” Journal of Applied Polymer Science, 67: 1503-13 (1998). 16. Mengeloglu, F., Matuana, L.M. and King, J.A., “Effects of impact modifiers on the properties of rigid PVC/wood-fiber composites,” Journal of Vinyl & Additive Technology, 6 (3):153-57 (2000). 17. Murariu, M., Ferreira, A.D.S., Duquesne, E., Bonnaud, L. and Dubois, Ph., “Polylactide (PLA) and highly filled PLA – calcium sulfate composites with improved impact properties,” Macromolecular Symposia, 272: 1-12 (2008). 18. Sykacek, E., Hrabalova, M., Frech, H. and Mundigler, N., “Extrusion of five biopolymers reinforced with increasing wood flour concentration on a production machine, injection moulding and mechanical performance,” Composites: Part A, 40: 1272-82 (2009). 19. Carlborn, K. and Matuana, L.M., “Modeling and optimization of formaldehyde-free wood composites using a box-behnken design,” Polymer Composites, 27 (5): 497-503 (2006). 20. Matuana, L.M. and Li, Q., “Statistical modeling and response surface optimization of extruded HDPE/wood-flour composite foams,” Journal of Thermoplastic Composites, 17: 185-99 (2004). 21. Shah, B.L. and Matuana, L.M., “Online measurement of rheological properties of PVC/wood-flour composites,” Journal of Vinyl & Additive Technology, 10 (3):121-8 (2004). 22. Jin, S. and Matuana, L.M., “Wood/plastic composites co-extruded with multi-walled carbon nanotube-filled rigid poly(vinyl chloride) cap layer,” Polymer International, 59: 648-57 (2010). 23. Zaini, M.J., Fuad, M.Y.A., Ismail, Z., Mansor, M.S. and Mustafah, J., “The effect of filler content and size on the mechanical properties of polypropylene/oil palm wood flour composites,” Polymer International, 40: 51-5 (1996). 24. Li, Q. and Matuana, L.M., “Effectiveness of maleated and acrylic acid-functionalized polyolefin coupling agents for HDPE-wood-flour composites,” Journal of Thermoplastic Composite Materials, 16: 551-564 (2003). 25. Stark, N.M. and Rowlands, R.E., “Effects of wood fiber characteristics on mechanical properties of wood/polypropylene composites,” Wood & Fiber Science, 35 (2): 167-74 (2003). 26. Matuana, L.M., Cam, S., Yuhasz, K.B. and Armstrong, Q.J., “Composites of acrylonitrilebutadiene-styrene filled with wood-flour,” Polymers & Polymer Composites, 15 (5): 343-8 (2007).   139 CHAPTER 5 FRACTURE TOUGHNESS OF POLY(LACTIC ACID)/ETHYLENE-ACRYLATE COPOLYMER/WOOD-FLOUR COMPOSITES TERNARY BLENDS This chapter has been submitted for publication in Polymer International (January 2012). It is co-authored by K. A. Afrifah and L. M. Matuana.   140 5.1 Abstract A fracture mechanics analysis based on J-integral method was adopted to determine the resistance of the composites with various concentrations of wood flour and ethylene-acrylate copolymer (EAC) to crack initiation (Jin) and complete fracture energy (Jf). The crack initiation (Jin) and complete fracture (Jf) energies of unmodified PLA/wood-flour composites showed the deleterious effect of incorporating wood fiber into the plastic matrix by significantly decreasing the fracture toughness of PLA as the wood flour content increases. The deteriorated fracture toughness of the matrix induced by adding brittle wood flour into PLA was well recovered by impact modification of the composites with ethylene-acrylate copolymer (EAC). Microscopic morphological studies revealed that the major mechanisms of toughening was through the EAC existing as separate domains in the bulk matrix of the composites which tended to act as stress concentrators that initiated local yielding of the matrix around crack tips and enhanced the toughness of the composites.   141 5.2 Introduction Poly(lactic acid) (PLA) is a bio-based and biodegradable polymer that has gained attention in recent years as potential alternative to petroleum-based thermoplastics due to its origin and environmental friendliness (1). However, its higher cost and brittleness as well as lower impact strength resistance (toughness) compared to most conventional plastics such as polyethylene (PE) and polypropylene (PP) are detrimental to its commercial competitiveness (15). Blending PLA with various additives such as lubricants, impact modifiers, plasticizers or a second polymer increases the toughness or impact strength of the PLA and broadens its applications (2-6). Unfortunately, the associated cost of the additives renders their PLA blends expensive (2,6). Previous efforts at reducing the cost of PLA and its blends included addition of cellulosic fibers (4-7). Nevertheless, the reduced cost is achieved at the expense of other mechanical properties such as impact strength, toughness, and elongation at break (6-8). This was expected since the ductile mode of failure of the matrix is altered by the incorporation of the brittle cellulosic fibers in the matrix, making the composites more brittle than the neat polymer (6-14). Attempts at enhancing the toughness of the PLA/wood-flour composites have included blending with rubbery polymers and plasticization (7,15). Such studies have often recorded improved impact strength and elongation at break for the composites due to the modifying effects of the additives (7,15). Previous studies of PLA/wood-flour composites, to the best of our knowledge have mainly assessed the toughness of unmodified and toughened-PLA/wood-flour composites with   142 notched Izod, unnotched Izod and charpy impact tests (4,5,7,15). Unfortunately, these impact tests measure the energy to break the sample without consideration of the crack initiation and propagation energies. The results from these tests do not represent the true material constants as they are size and geometry dependent (9,16). It is also difficult to interpret the results and compare with other test results since they have poor reproducibility (9). Moreover, impact test does not take into consideration cracks present in the tested sample, which can affect the testing results. Indeed, weak or poor interfacial adhesion occurs in the complex structure of ternary blends such as the composites of this study. Poor interfacial adhesion between components in the composites leads to cracks in the composites. Fibers also act as discontinuities in the composites, capable of initiating cracks (10). Due to the possibility of flaws or cracks in the composites, a method that takes those imperfections into consideration should be used to effectively evaluate the toughness of the composites. Fracture mechanics approach is one of the methods suitable for testing the fracture toughness of the material by taking its imperfections into consideration and produces parameters that are true material constants, independent of both its size and geometry (17). In addition, fracture mechanics concepts have been proven to establish morphology-property correlations for thermoplastic materials, hence its use as an assessment tool for the PLA/EAC/wood-flour ternary composites is preferred (18). Linear elastic fracture mechanics (LEFM) is one of the widely used fracture mechanics approaches to characterize the fracture toughness of polymeric materials (17). Fracture toughness measured by this method represents true material constants only when limited plastic deformation occurs at the crack tip or when certain restrictive size criteria of the testing specimen that leads to plane-strain fracture have been satisfied (17). For most brittle materials   143 such as polystyrene (PS) these requirements are easy to achieve. However, in toughened materials large scale plastic deformation may occur ahead of the crack tip during testing violating the validity requirements of LEFM (17,18). Such toughened materials require post yield fracture mechanics approaches such as J-integral method to quantitatively characterize their fracture toughness (17). Consequently, the J-integral method of fracture mechanics was used to assess the fracture toughness of both unmodified and ethylene-acrylate copolymer (EAC)-toughened PLA/woodflour composites. Particular emphasis was placed on evaluating the effects of wood flour concentration and EAC content on the fracture resistance of PLA/wood-flour composites to gain an in-depth understanding of the mechanisms of crack initiation and propagation in the composites. Toughening mechanisms of composites are more complicated than in single-phase systems, as stress concentrations, interactions between components and heterogeneity provide additional complications. Consequently, morphologies of fractured composites were analyzed through scanning electron microscopy to elucidate the failure and toughening mechanisms of the composites.   144 5.3 Experimental 5.3.1 Materials The PLA resin used in this study (PLA 3001D) was purchased from NatureWorks, USA. -3 It had a density of 1.144 g cm , 8 % of D-lactide enantiomer, crystallinity of 15 %, and melt -1 flow index of 12.1 g 10 min o at 190 C (19). An ethylene-acrylate copolymer (EAC) (Biomax strong 100) supplied by DuPont Packaging and Industrial Polymers, was used as the impact o o modifier. It had a melting point of 72 C, glass transition temperature of -55 C and a melt flow -1 index of 12 g 10 min o at 190 C (20). American Wood Fibers supplied the maple wood flours used in the composites. The commercial grade of wood flour was 6010, which corresponds to 60 nominal mesh size. 5.3.2 Compounding and compression molding The two studied variables included wood flour contents (0-40 % based on the total weight of the composites) and impact modifier (EAC) contents (0-20 % based on the total weight of PLA) in the composites. Impact modification was assessed using the composites containing a fixed amount of wood flour (40 %). o o Before melt blending, the PLA and wood flour were dried at 55 C and 105 C, respectively, for at least 24 h to remove moisture. A 32 mm conical counter-rotating twin-screw extruder with a length-to-diameter ratio of 13:1 (C.W. Brabender Instruments Inc. South Hackensack, NJ) fitted with a rectangular profile die was used for the melt blending of the   145 materials to ensure effective mixing of the composites’components. The extruder was powered with a 5.6 kW (7.5 hp) intelli-Torque Plasti-Corder Torque Rheometer (C.W. Brabender Instruments Inc. South Hackensack, NJ). Extrusion temperature from the hopper to the die for o all the processing was 180-180-175-170 C. Once blended, the extrudates were cooled at room temperature for about an hour and granulated in a Conair Wortex granulator (model JC-5). Reo extrusion of the granulated samples, which had been dried for at least 24 h at 55 C, produced 3 o mm thick sheet. Subsequently, the sheets were compression molded at 160 C for 150 sec in a Carver hot press to reduce the sheet thickness to 1.30 mm. The pressed composites were then sawn to a length and a width of 127 and 25.4 mm, respectively for fracture toughness tests. 5.3.3 Fracture energy determination Various approaches and specimen configurations are used for the determination of fracture toughness (21). This study used J-integral method with single-edge-notched (SEN) tensile specimens (Figure 5.1) to characterize the fracture behavior of the composites in terms of fracture resistance at crack initiation (Jin) and the work of fracture or fracture energy (Jf).   146 L=127 mm a=12.7 mm W=25.4 mm Figure 5.1.   Schematic drawing of fracture toughness test specimen. 147 J-integral (Jin) proposed by Rice as a criterion for fracture toughness is the most important fracture mechanics concept due to its energetic interpretation of the fracture process (22). It is a path independent line integral that encloses the plastically deformed area and describes the stress strain field around the crack-tip (23-25). J-integral is measured at crack initiation, and can be expressed as the energy intake per unit area to create new fracture surfaces in a loaded body containing a crack at constant displacement (23): J in   1 dU B da (1) with B as the thickness of the specimen, U the total potential energy (the area under the loaddisplacement curve up to crack initiation point) and a the crack length. For ease of application of this method to experimental determination and computation of fracture toughness of tough materials, the above equation has been rewritten as (26): J in  U in (2) B(W  a ) where η is a geometry factor, Uin the potential energy at crack initiation and W the width of the specimens.   148 The value of the geometry factor η is equal to 2 when the crack’s length-to-width ratio (a/W) of notched specimen satisfies 0.4 < a/W and < 0.6 criterion (27). However, in a recent study on the fracture resistance of vulcanized natural rubber/clay nanocomposites by J-integral method, Ramorino and coworkers obtained a value of 0.9 instead of 2 using notched specimens with a/W ratio of 0.5 (28). Implications are that the value of the geometry factor η is not always 2 for notched specimens and may be material dependent (28,29). Consequently, determination of the geometry factor characteristic of the specimens employed in J-integral (Jin) test is desirable to avoid any uncertainties and ensure accuracy of the measured fracture toughness. Detailed description of the derivation and results of the η factor used in this study is presented in the results and discussion section. The work of fracture or fracture energy (Jf), which quantifies the ability of a material to resist complete failure, is another approach to measure the fracture toughness of a material (9). This fracture parameter represents the energy per unit fracture surface dissipated during complete fracture of the specimen (Jf). It can be derived from the load-displacement curves at the point of complete failure by the following expression (30): Jf  Uf (3) B (W  a ) with Uf as the area under the load-displacement curve at the point of complete fracture and B, W, and a are the same as in Equation 2.   149 Tensile tests were performed on an Instron 5585H testing machine to determine the resistance to crack initiation (Jin) and fracture energy (Jf) of the composites. This equipment allowed the recording of the load-displacement curves using the Instron Bluehill 2, version 2.14 software. The length between the grips for the test was 100 mm and tensile tests were performed -1 at a crosshead displacement rate of 5 mm min until complete failure of the specimens. Unless otherwise mentioned, the specimens with initial crack length-to-width ratio (a/W) of 0.5 on one edge were used (Figure 5.1). Initial crack lengths were made by first sawing on a band saw with a 0.3 mm thick saw blade up to a point and the final 1.5–2 mm was done with a razor blade to create a sharp finish. Load versus displacement curves were recorded and the crack initiation points were marked on each curve during the tests to indicate the crack initiation point. Crack initiation points were easily observable since the crack opened widely prior to its propagation. The energies Uin and Uf in Equations 2 and 3 were obtained as the areas under the load-displacement curves at crack initiation and complete fracture, respectively. From these values, energies for crack initiation (Jin) and complete fracture (Jf) were determined. All tests o were conducted in a conditioned room at a temperature of 23 ± 2 C and 65 ± 4 % relative humidity. Prior to testing, the samples were conditioned in the conditioning room for at least 48 h. At least 15 samples for each formulation were tensile tested to obtain reliable means and standard deviation.   150 5.3.4 Scanning electron microscopy (SEM) JOEL JSM-6400 SEM was used to study the morphology of the surface of the fractured samples. This assisted in assessing the fracture mechanisms associated with the failure in the various compositions of the PLA/wood-flour composites and PLA/EAC/wood-flour ternary blends. The acceleration voltage and magnification for the test were 12 kV and 1200x, respectively. The SEM micrographs were taken after coating the surfaces of the specimens with a thin layer of gold. 5.3.5 Statistical analysis A one-way analysis of variance (ANOVA) was carried out with an α significance value of 0.05, comparing the effect of wood flour and EAC contents on the fracture toughness of the composites. All statistical analyses were performed using Design Expert software (v. 8) from Stat-Ease (Minneapolis, MN).   151 5.4 Results and Discussion 5.4.1 Geometric factor (η) calibration The determination of crack initiation energy (Jin) using Equation 2 and SEN specimens with ligaments satisfying the 0.4 < a/W < 0.6 condition, traditionally uses the value of 2 for the geometry factor (η). Nevertheless, as mentioned, to ensure the accuracy of the Jin determined in this study, the geometry factor η was calibrated for the specimens used. By combining Equations 1 and 2, the η factor was derived as:  (W  a) dU U da (4) Specimens with the same length, thickness and width as indicated in the experimental section (Figure 5.1) but having different crack lengths (a = 5.08, 7.62, 10.16, 12.70, and 15.24 mm) were tensile-tested. Crack initiation points were marked on their corresponding loaddisplacement curves (Figure 5.2). From the load-displacement curves, the energy U was determined for each specimen at various displacement values (q) of 0.10, 0.20, and 0.29 mm (Figure 5.2). By plotting the derived energies (U) versus initial crack lengths (a) for each displacement (q), the slopes dU/da were determined from the series of straight lines obtained by linearly interpolating the data as shown in Figure 5.3a. computed as a function of crack length using Equation 4.   152 The geometry factor η was then Load (N) 300 Crack initiation point 200 q3 q2 100 U1 q1 0 0 0.2 0.4 0.6 0.8 1 Extension (mm) Figure 5.2.   Scheme for the evaluation of energy U at different displacements ‘q’. 153 0.06 Energy U (J) (a) q=0.10 mm q=0.20 mm q=0.29 mm 0.04 0.02 0.00 0 0.005 0.01 0.015 0.02 a (m) 2.0 Geometry factor η (b) q=0.10 mm q=0.20 mm q=0.29 mm 1.5 1.0 0.5 0.0 0 0.2 0.4 0.6 0.8 a/w Figure 5.3.   Charts for the evaluation of the geometry factor (η) for neat PLA. (a) Energy input during fracture plotted as a function of crack length ‘a’ at different displacements ‘q’. (b) Geometry factors at different displacements ‘q’ plotted as a function of a/W ratios. 154 Results of geometry factor η as a function of crack length-to-width ratios (a/W) for each of the three displacements (q) examined are illustrated in Figure 5.3b for the neat PLA. It is clearly seen that the geometry factor η was less than 2, slightly varying from 1.0 to 0.6 as the crack length-to-width ratios (a/W) increased from 0.2–0.6. Additionally, this factor was not affected by the displacement values. Similar trends were obtained for all PLA/wood-flour composites examined in this study but not shown in this paper due to page limitation. Nevertheless, plots of geometry factors η versus crack length-to-width ratios (a/W) for neat PLA and some representative composites samples are shown in Figure 5.4. Once again the geometry factor η was independent of crack length within the range of a/W ratios studied. Therefore, averaging the values of η obtained at various crack lengths resulted in a mean value of 0.8 ± 0.1. The calibrated geometric factor of η = 0.8 was then used for the computation of the crack initiation energy (Jin) of the composites using Equation 2. It is worth noting that the experimental value of geometry factor η was less than 2, suggesting the material dependency for this factor, in agreement with the findings of Ramorino and coworkers (28).   155 Geometry factor η 2.0 Neat PLA 40 % WF 40 % WF, 15 % EAC 1.5 1.0 0.5 0.0 0 0.2 0.4 0.6 0.8 a/w Figure 5.4.   The overall geometry factors (η) as a function of a/W ratios for neat PLA, unmodified and EAC-modified PLA/wood-flour composites containing 40 % wood flour (WF) content. 156 5.4.2 Effect of wood flour and ethylene-acrylate copolymer (EAC) contents on fracture toughness of PLA/wood-flour composites Figure 5.5 illustrates the effect of wood flour content on the crack initiation energy (Jin) and fracture energy (i.e. ability of the composites to resist fracture) (Jf) of PLA/wood-flour composites. Both the crack initiation (Jin) and fracture (Jf) energies of PLA decreased as wood flour content increased in the composites, indicating the inability of the composites to resist crack initiation and propagation. This was expected since the ductility of plastic matrices reduces with the inclusion of wood fibers, making the composites more brittle (7,10-14). Up to 20 %, the addition of wood flour had no significant effect on both crack initiation (Figure 5.5a) and fracture (Figure 5.5b) energies of the matrix. The lower amount of added wood flours may have accounted for the insensitivity of the crack initiation (Jin) and complete fracture (Jf) energies on wood flour content up to 20 %. In contrast, the energies needed for crack initiation and sample fracture decreased significantly after adding more than 20 % wood flour into the matrix (Figures 5.5a and b). Detailed statistical analyses of the data are presented in Appendices B.1 to B.4. The reduced toughness may be explained by the crack development and propagation mechanisms in the polymer, which depends on the strength of the interfacial regions between the components (7). Generally, greater the intimate adhesion between components in composites, less microvoids or microflaws in the composite’s structure (12). Weak interfacial regions arise in composites due to poor interfacial adhesion between the hydrophilic wood flour and the hydrophobic polymer matrix. As a result, each fiber, acts as a discontinuity capable of initiating cracks in the composites (7,12). Increasing wood flour content   157 increases the interfacial regions between the PLA matrix and wood fibers. Since the fibers were not treated with coupling or other compatibilizing agents (12), it is believed that the number of weakened interfacial regions between PLA and wood particles increased with the wood flour contents, thus making the composite more prone to crack propagation.   158 Crack initiation energy (J/m2) 2000 (a) 1500 1000 R² = 0.9903 500 0 -10 0 10 20 30 40 50 Fracture energy Jf (J/m2) Wood flour content (% ) 4000 (b) 3000 2000 R² = 0.9927 1000 0 -10 0 10 20 30 40 50 Wood flour content (% ) Figure 5.5.   Effect of wood flour content on fracture toughness of PLA/wood-flour composites: (a) J-integral (Jin) and (b) fracture energy (Jf). 159 This fact was confirmed by SEM micrograph shown in Figure 5.6a. Insufficient or lack of intimate adhesion between components caused fiber pullout when the composites were fractured during tensile tests. The pulled out wood fibers were not coated with PLA resin (Figure 5.6a) indicative of poor interfacial adhesion between the matrix and wood fibers leading to brittle failure.   160 (a) Clean WF 20 µm (b) EAC domains 20 µm Figure 5.6.   SEM of PLA/wood-flour composites with 40 % wood flour content and EAC contents of (a) 0 % and (b) 10 %. 161 The deteriorated resistance to crack initiation and propagation of the matrix induced by incorporating brittle wood flour into PLA (Figures 5.5a and b) was well recovered by impact modification of the composites with ethylene-acrylate copolymer (EAC) (Figure 5.7). The effectiveness of the EAC impact modifier in toughening PLA/wood-flour composites is clearly illustrated in Figure 5.7 where one can observe that EAC-modified composites exhibited higher values of Jin and Jf than their unmodified counterparts (composites with 0 % EAC), implying that the addition of EAC enhanced the toughness of the composites (7). It must be pointed out that the extent of fracture toughness improvement was a strong function of EAC content (2,7). Both crack initiation energy (Jin) and fracture energy (Jf) of PLA/wood-flour composites increased significantly with EAC content up to 10 % and leveled off above this concentration (Figures 5.7a and b). Appendices B.5 to B.8 gives the detailed statistical analyses for the effect of the EAC on the crack initiation energy (Jin) and fracture energy (Jf) of the composites. The increased resistance to crack initiation (Jin) and complete fracture (Jf) of the composites is attributable to the modifying effect of the EAC, which allowed its rubbery nature to impart toughness to the matrix (7,30,31). Previous investigation showed immiscibility in EAC-modified PLA blends with a two-phase morphology lacking significant molecular interactions (7). Since this two-phase morphology in the blends was a necessary condition for toughening PLA and PLA/wood-flour composites (2,7,32,33), this immiscibility allowed the rubbery nature of EAC to induce energy dissipation mechanisms into PLA, which retarded crack initiation and propagation, and ultimately resulted in a material with improved fracture toughness (7).   162 Crack initiation energy (J/m2) 2000 (a) 1500 1000 R² = 0.9988 500 0 -5 0 5 10 15 20 25 Fracture energy Jf (J/m2) EAC content (% ) 4000 (b) 3000 2000 R² = 0.9680 1000 0 -5 0 5 10 15 20 25 EAC content (% ) Figure 5.7. Effect of EAC impact modifier content on fracture toughness of PLA/wood-flour composites containing 40 % wood flour content: (a) J-integral (Jin) and (b) fracture energy (Jf).   163 Moreover, morphological studies of the fractured surfaces of impact modified composites showed that EAC impact modifier existed as separate phase in the PLA matrix of the composite (Figure 5.6b). This separated phase of EAC modifier acted as stress concentrators by initiating local yielding of the matrix and the formation of a large number of energy dissipating microscopic crazes in the surrounding PLA matrix; thus improving the fracture toughness of the composites. Other investigators have reported similar results in talc-filled PP and PP/wood-flour composites as well as EAC-PLA and PLA/wood-flour composites (2,7,31).   164 5.5 Conclusions This study investigated the effects of wood flour concentration and EAC content on the toughness of PLA/wood-flour composites using fracture mechanics concepts. J-integral at crack initiation (Jin) and the fracture energy per unit surface area (Jf) were evaluated as the fracture properties. Assessment of the microscopic morphologies of the fracture surfaces was performed to characterize the fracture micro mechanisms. From the results, it can be concluded that the crack initiation (Jin) and complete fracture (Jf) energies of PLA/wood-flour composites significantly decreased with the incorporation of wood flour. This was expected since the ductility of plastic matrices reduces with the inclusion of wood fibers, making the composites more brittle. In contrast, impact modification of composites with EAC to form ternary blends of PLA/EAC/wood-flour composites increased the crack initiation (Jin) and complete fracture (Jf) energies. It was observed from the ductile morphologies of the fractured surfaces that the EAC existed in the toughened composites as separate domains in the bulk matrix which enhanced the toughness of the ternary composites. Subsequently, the main mechanisms of improvement of the fracture properties due to the EAC were deemed to be the impact modifier acting as stress concentrators that initiated local yielding of the matrix at the crack tip region due to the EAC domains in the bulk matrix.   165 REFERENCES   166 5.6 References 1. Park, S.D., Todo, M., Arakawa, K. and Koganemaru, M., “Effect of crystallinity and loadingrate on mode I fracture behavior of poly(lactic acid),” Polymer, 47: 1357-63 (2006). 2. Afrifah, K.A. and Matuana, L.M., “Impact modification of polylactide with a biodegradable ethylene/acrylate copolymer,” Macromolecular Materials Engineering, 295: 802-11 (2010). 3. Hu, R. and Lim, J.K., “Fabrication and mechanical properties of completely biodegradable hemp fiber reinforced polylactic acid composites,” Journal of Composite Materials, 41 (13): 1655-69 (2007). 4. Lee, S.Y., Kang, I.A., Doh, G.H., Yoon, H.G., Park, B.D. and Wu, Q., “Thermal and mechanical properties of wood flour/talc-filled polylactic acid composites: effect of filler content and coupling treatment,” Journal of Thermoplastic Composites, 21: 209-23 (2008). 5. Huda, M.S., Drzal, L.T., Misra, M. and Mohanty, A.K., “Wood-fiber-reinforced poly(lactic acid) composites: evaluation of the physicomechanical and morphological properties,” Journal of Applied Polymer Science, 102: 4856-69 (2006). 6. Matuana, L.M and Diaz, C.A., “Study of cell nucleation in microcellular poly(lactic acid) foamed with supercritical CO2 through a continuous-extrusion process,” Industrial and Engineering Chemistry Research, 49: 2186-93 (2010). 7. Afrifah, K.A. and Matuana, L.M., “Statistical optimization of ternary blends of poly(lactic acid)/ethylene-acrylate copolymer/wood flour composites,” Macromolecular Materials and Engineering, 296, DOI: 10.1002/mame.201100097 (2011). 8. Pilla, S., Gong, S., O’Neill, E., Rowell, R.M. and Krzysik, A.M., “Polylactide-pine wood flour composites,” Polymer Engineering and Science, 48: 578-87 (2008). 9. Park, B.D. and Balatinecz, J.J, “Mechanical properties of wood-fiber/toughened isotactic polypropylene composites,” Polymer Composites, 18 (1): 79-89 (1997). 10. Afrifah, K.A., Hickok, R.A. and Matuana, L.M., “Polybutene as a matrix for wood plastic composites,” Composites Science and Technology, 70: 167-72 (2010). 11. Matuana, L.M., Park, C.B. and Balatinecz, J.J., “The effect of low levels of plasticizer on the rheological and mechanical properties of polyvinyl chloride/newsprint-fiber composites,” Journal of Vinyl & Additive Technology, 3 (4): 265-273 (1997). 12. Matuana, L.M., Woodhams, R.T., Balatinecz, J.J. and Park, C.B., “Influence of interfacial interactions on the properties of PVC/cellulosic fiber composites,” Polymer Composites, 19 (4): 446-55 (1998). 13. Matuana, L.M., Cam, S., Yuhasz, K.B. and Armstrong, Q.J., “Composites of acrylonitrilebutadiene-styrene filled with wood-flour,” Polymers & Polymer Composites, 15 (5): 343-8 (2007).   167 14. Mengeloglu, F., Matuana, L.M. and King, J.A., “Effects of impact modifiers on the properties of rigid PVC/wood-fiber composites,” Journal of Vinyl & Additive Technology, 6 (3):153-57 (2000). 15. Oksman, K., Skrifvars, M. and Selin, J.F., “Natural fibres as reinforcement in polylactic acid (PLA) composites,” Composites Science and Technology, 63: 1317-24 (2003). 16. Hughes, M., Hill, C.A.S. and Hague, J.R.B., “The fracture toughness of bast fibre reinforced polyester composites,” Journal of Materials Science, 37: 4669-76 (2002). 17. Wu, J., Mai, Y.W. and Cotterell, B., “Fracture toughness and fracture mechanisms of PBT/PC/IM blend,” Journal of Materials Science, 28: 3373-84 (1993). 18. Reincke, K., Grellmann, W., Lach, R. and Heinrich, G., “Toughness optimization of SBR elastomers – use of fracture mechanics methods for characterization,” Macromolecular Materials and Engineering, 288: 181-89 (2003). 19. Matuana, L.M., Faruk, O. and Diaz, C.A., “Cell morphology of extrusion foamed poly (lactic acid) using endothermic chemical foaming agent,” Bioresource Technology, 100: 5947-54 (2009). 20. Murariu, M., Ferreira, A.D.S., Duquesne, E., Bonnaud, L. and Dubois, Ph., “Polylactide (PLA) and highly filled PLA – calcium sulfate composites with improved impact properties,” Macromolecular Symposia, 272: 1-12 (2008). 21. Sivaraman, P., Chandrasekhar, L., Mishra, V.S., Chakraborty, B.C. and Varghese, T.O., “Fracture toughness of thermoplastic co-poly(ether ester) elastomer – acrylonitrile butadiene styrene terpolymer blends,” Polymer Testing, 25: 562-67 (2006). 22. Rice, J., “A path independent integral and approximate analysis of strain concentration by notches and cracks,” Journal of Applied Mechanics, 2: 379-86 (1968). 23. Arencon, D. and Velasco, J.I., “Fracture toughness of polypropylene-based particulate composites,” Materials, 2: 2046-94 (2009). 24. Rosler, J., Harders, H., and Baker, M., Mechanical Behaviour of Engineering Materials: Metals, Ceramics, Polymers, and Composites. Springer: New York (2007). 25. Ha, C.S., Kim, Y. and Cho, W.J., “Fracture mechanics investigation on the PP/EPDM/Ionomer ternary blends using J-integral by locus method,” Journal of Applied Polymer Science, 51: 1381-88 (1994). 26. Frassine, R., Rink, M. and Pavan, A., “Size effects in the fracture of a pipe-grade high density polyethylene,” Fatigue and Fracture of Engineering Materials and Structures, 20: 1217-23 (1997). 27. Sumpter, J.D.G., and Turner, C.E., “Application of J to elastic-plastic materials,” International Journal of Fracture, 9: 320-21 (1973).   168 28. Ramorino, G., Agnelli, S., De Santis, R., and Ricco, T., “Investigation of fracture resistance of natural rubber/clay nanocomposites by J-testing,” Engineering Fracture Mechanics, 77: 1527-36 (2010). 29. Anderson, T.L., McHenry, H. I., and Dawes, M.G., “Elastic-plastic fracture toughness tests with single-edge notched bend specimens,” Elastic-Plastic Fracture Test Methods: The User’s Experience, ASTM STP 856, American Society for Testing and Materials, 210-229 (1985). 30. Todo, M., Takayama, T., Tsuji, H. and Arakawa, K., “Effect of LTI blending on fracture properties of PLA/PCL polymer blend,” Journal of Solid Mechanics and Materials Engineering, 1 (9): 1157-64 (2007). 31. Oksman, K. and Clemons, C., “Mechanical properties and morphology of impact modified polypropylene-wood flour composites,” Journal of Applied Polymer Science, 67: 1503-13 (1998). 32. Ratna, D., and Banthia, A.K., “Toughening of epoxy resin by modification with 2-ethylhexyl acrylate-acrylic acid copolymers,” Polymer International, 49: 309-15 (2000). 33. Das, V., Pandey, A.K., and Krishna, B., “Low temperature izod impact studies of blends based on impact grade polypropylene and ethylene-α-octene copolymer,” Journal of Reinforced Plastics and Composites, 28: 2879-88 (2009).   169 CHAPTER 6 Conclusions and Future Work 6.1 Conclusions Environmental concerns about the use of petroleum-based resins in wood plastic composites (WPCs), instability in oil producing regions and increasing oil prices have helped to drive interest in composites made from renewable bio-based plastics and fibers. Polylactic acid (PLA) an aliphatic and compostable polyester, derived from renewable resources stands out as a popular alternative to petroleum-based plastics due to its favorable physical and degradation characteristics. In spite of its excellent properties, PLAs commercial application is limited because of its high cost and greater brittleness than most petroleum-based resins. To mitigate the material cost, PLA can be blended with cellulosic fibers. However, the lowered cost comes at the expense of flexibility and impact strength. Therefore, the main goal of this work was to gain an in-depth understanding of the toughening mechanisms of the PLA with biodegradable ethylene acrylate copolymer (EAC) in order to develop cost effective and biodegradable PLA/wood-flour composites with improved impact strength, toughness, high ductility and flexibility. The following specific objectives were achieved to accomplish the main goal of this project: 170 1. Assess the efficiency and effectiveness of the EAC in toughening amorphous and semicrystalline grades of PLA with the ultimate goal of gaining an in depth understanding of their toughening mechanisms. 2. Study how the testing temperature affects the notched Izod impact strength of toughened PLA in order to determine appropriate service temperature conditions for PLA/EAC blends. 3. Evaluate the effects of wood flour particle size (mesh size), wood flour content and impact modifier concentration on the mechanical properties of PLA/EAC/wood-flour composites. This was aimed at understanding the interactions between the composites component materials and developing model equations establishing the relationships between them. 4. Assess the fracture toughness of the composites using fracture mechanics in order to determine the energy consumed at each phase of the fracture process and obtain data that would be useful in engineering designs applying PLA/EAC/wood-flour composites. The following conclusions relating to the specific objectives were drawn from the experimental results. 171 Objectives 1 and 2: Results of impact strength tests indicated that the ethylene-acrylate copolymer toughened the semicrystalline and amorphous grades of PLA. The efficiency of the ethylene-acrylate copolymer in the semicrystalline grades superseded the amorphous grade. Thus the semicrystalline grades recorded lower impact modifier content for the brittle-to-ductile transition compared to the amorphous grade. A detailed analysis of the injection molding grade of the semicrystalline PLA revealed increasing ductility, elongation at break, and energy to break with ethylene-acrylate copolymer content compared to neat PLA. Contrarily, the tensile strength and modulus decreased with ethylene-acrylate copolymer content. SEM and DSC analyses indicated the blends as an immiscible two phase system. The fracture of PLA and its blends with up to 5 wt.% impact modifier occurred through crazing or microcracking and debonding of modifier particles with the matrix mainly bearing all the stress due to the low content of modifier particles resulting in brittle failure. For the blends with 10 wt.% or more impact modifier, the fracture mechanisms included impact modifier debonding, fibrillization, crack bridging and matrix shear yielding resulting in a ductile behavior. A relatively low brittle-to-ductile transition temperature obtained for the PLA blended with 15 wt.% ethylene-acrylate copolymer confirms that this impact modifier was a good additive for toughening PLA for use in sub-ambient temperatures. Objective 3: The impact strength and elongation at break of the PLA/wood-flour composites was positively affected by the addition of the impact modifier, which negatively affected the tensile strength and modulus, irrespective of the wood flour particle size and content. Increasing the 172 wood flour content however, improved the tensile modulus but decreased the elongation at break of the composites. On the other hand, the tensile and impact strengths of the composites with coarse wood particles (up to 40 mesh) without impact modifier were not affected by the wood flour content. Numerical optimization led to several different combinations of the material compositions needed to manufacture PLA/wood-flour composites with mechanical properties similar to those of the unfilled PLA. Two scenarios were observed from these combinations depending on the particle size of the wood flour. High wood flour and impact modifier contents are required for composites made with fine wood flour particles. Whereas, the formulation requires low wood flour content and excludes impact modifier for composites with coarse wood flour particles. These optimization solutions were successfully validated experimentally. Objective 4: Results from the fracture mechanics analysis concluded that the crack initiation (Jin) and complete fracture (Jf) energies of PLA/wood-flour composites significantly decreased with the incorporation of wood flour. This was expected since the ductility of plastic matrices reduces with the inclusion of wood fibers, making the composites more brittle. In contrast, impact modification of composites with EAC to form ternary blends of PLA/EAC/wood-flour composites increased the crack initiation (Jin) and complete fracture (Jf) energies. It was observed from the ductile morphologies of the fractured surfaces that the EAC existed in the toughened composites as separate domains in the bulk matrix which enhanced the toughness of 173 the ternary composites. The EAC domains improved the fracture properties of the composites by acting as stress concentrators that initiated local yielding of the matrix at the crack tip region 174 6.2 Future Work This study resulted in the development of a novel toughened bio-based wood plastic composite (WPC) made of wood flour, poly (lactic acid) (PLA) and ethylene acrylate copolymer (EAC) impact modifier. This composite is to serve as an alternative to traditional WPCs derived from petroleum based resins. The success of this composite, as replacement for the traditional WPCs depends on its renewability, biodegradability, physical and mechanical properties. In this study the mechanical properties of the newly developed composites have been duly determined and the mainly plant-based materials utilized to manufacture them supports their renewability claim. Biodegradability a major attribute of environmental importance is however outstanding and needs to be assessed in order to complete information on the properties of the new composites. The main components that make up biodegradable plastics or polymers are polysaccharides, polyesters, and polyamides, which can be hydrolyzed by enzymes such as glycosyl hydrolases, ester hydrolases, and peptide hydrolases, respectively (1). PLA is a biobased aliphatic polyester and it is susceptible to hydrolytic degradation (2). It degrades easily in compost environments in compliance with ISO and ASTM standards. Under conditions of high temperature and humidity in active compost, PLA degrades quickly and disintegrates within weeks to months (3, 4). Blending of biodegradable polymers with other polymers is one approach to modify them to achieve desired properties. Miscibility of the blends is an important factor that affects the final polymer properties (5). However, observations are that formation of either miscible or immiscible blends especially with non-biodegradable polymers can slow down or even inhibit 175 the degradation process of the biodegradable components of the blends (5). Reports on enzymatic degradation using PHB depolymerase from Alcalegenes feacalis of immiscible PHB/PCL and PHB/PBA blends and miscible PHB/PVAc blend showed that the weight loss of the blends decreased linearly with increase in the amounts of PCL, PBA and PVAc (6). This study indicated inhibition of degradation with increase in both miscible and immiscible additives. The EAC impact modifier used in the current study is claimed to be compostable at low concentrations in PLA/EAC blends (7). This implies that once the concentration of the EAC used in the composites exceeds a critical concentration, it may influence the degradation rate of the composites as has been observed in the above reported studies (6). Therefore, a study of the mode, extent and rate of the degradation of the PLA/EAC/wood-flour composites would be useful information not only for the commercialization of the composites, but also to understand their degradation process and determine their appropriate form of application. Investigation of the effect of ethylene acrylate copolymer (EAC) impact modifier incorporation into the PLA/wood-flour composites showed improved impact strength and toughness. However, the impact modifier negatively affected the tensile strength and modulus of the composites irrespective of the wood flour particle size and content. The mechanical properties (tensile and flexural strength) of wood plastic composites can be improved by enhancing the interfacial interactions between the wood fibers and the matrix polymer (8-10). Good surface properties are required to obtain composites with high tensile and flexural strengths (9). Consequently, a study to optimize the interaction at the interface between the wood fibers and the matrix with methods such as chemical surface modification of the wood fibers (e.g. mercerization, esterification, etc), compatibilization (e.g. using maleated compounds 176 and silanes), and adhesion promotion between wood fiber and matrix (e.g. using lignin) should be conducted with the aim of at least enhancing the properties (strength and modulus) reduced by the introduction of the impact modifier (8-12). Composites with improved mechanical properties would be commercially more attractive and useful for varied applications. 177 REFERENCES 178 6.3 References 1. Iwata, T., Abe, H., and Kikkawa, Y., Enzymatic Degradation. Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons, Inc.: Hoboken. pp. 383-398 (2010). 2. Tsuji, H., Hydrolytic Degradation. Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons, Inc.: Hoboken. pp. 345-376 (2010). 3. Calabia, B. P., Tokiwa, Y., Ugwu, C. U., and Aiba, S., Biodegradation. Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons, Inc.: Hoboken. pp. 423-8 (2010). 4. Nampoothiri, K. M., Nair, N. R., and John, R. P., “An overview of the recent developments in polylactide (PLA) research,” Bioresource Technology, 101: 8493-8501 (2010). 5. Tokiwa, Y., Calabia, B. P., Ugwu, C. U., and Aiba, S., “Biodegradability of Plastics,” International Journal of Molecular Sciences, 10: 3722-42 (2009). 6. Kumagai, Y.and Doi, Y., “Enzymatic degradation and morphologies of binary blends of microbial poly(3-hydroxybutyrate) with poly(ε-caprolactone), poly(1,4-butylene adipate and poly(vinyl acetate),” Polymer Degradation and Stability, 36: 241-8 (1992). 7. Dupont, Biomax Resins, Material Safety Data Sheet, 11/28/2006. 8. Espert, A., “Natural fibres/polypropylene composites from residual and recycled materials: Surface modification of cellulose fibers, properties, and environmental degradation,” KTH Fiber - och polymerteknologi, 100 (44), 1-36 (2003). 9. Zazyczny, J.M., and Matuana, L.M., Fillers and Reinforcing Agents, PVC Handbook, Hanser: Munich. pp. 235-72 (2005). 10. Ghosh, B., Bandyopadhyay-Ghosh, S., and Sain, M., Composites, Poly(lactic acid) Synthesis, Structures, Properties, Processing, and Applications, John Wiley & Sons Inc.: Hoboken. pp. 293-307 (2010). 11. Shibata, M., Biodegradable Polymer Blends and Composites from Renewable Resources, John Wiley & Sons, Inc.: Hoboken, N.J. pp.287-99 (2009). 12. Carlborn, K., and Matuana, L.M., “Influence of processing conditions and material compositions on the performance of formaldehyde-free wood-based composites,” Polymer Composites, 27 (6): 599-607 (2006). 179 APPENDICES 180 APPENDIX A THE FOLLOWING APPENDICES DEALS WITH CHAPTER 3 AND PRESENTS THE STATISTICAL ANALYSES OF THE EFFECT OF PLA CRYSTALLINITY AND IMPACT MODIFIER CONTENTS ON THE IMPACT STRENGTH OF PLA/EAC BLENDS 181 Table A.1. Source Model EAC Content Pure Error Cor Total Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (3001D)/EAC Blends. Sum of Squares 367078.29 367078.29 62053.27 429131.55 DF Mean Square F Value Prob > F 7 7 71 78 < 0.0001 < 0.0001 52439.76 52439.76 873.99 PLA 3001D: semicrystalline 182 60.00 60.00 significant Table A.2. Pairwise Comparison of the Impact Strength of PLA (3001D)/EAC Blends. EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 2.27 1 13.58 0.17 0.8679 1 vs 3 -0.56 1 13.58 -0.04 0.9675 1 vs 4 -17.12 1 13.58 -1.26 0.2117 1 vs 5 -42.97 1 13.58 -3.16 0.0023 1 vs 6 -71.16 1 13.58 -5.24 < 0.0001 1 vs 7 -101.29 1 13.58 -7.46 < 0.0001 1 vs 8 -209.15 1 13.58 -15.40 < 0.0001 2 vs 3 -2.82 1 13.22 -0.21 0.8316 2 vs 4 -19.38 1 13.22 -1.47 0.1470 2 vs 5 -45.24 1 13.22 -3.42 0.0010 2 vs 6 -73.43 1 13.22 -5.55 < 0.0001 2 vs 7 -103.55 1 13.22 -7.83 < 0.0001 2 vs 8 -211.42 1 13.22 -15.99 < 0.0001 3 vs 4 -16.56 1 13.22 -1.25 0.2144 3 vs 5 -42.41 1 13.22 -3.21 0.0020 3 vs 6 -70.60 1 13.22 -5.34 < 0.0001 3 vs 7 -100.73 1 13.22 -7.62 < 0.0001 3 vs 8 -208.59 1 13.22 -15.78 < 0.0001 4 vs 5 -25.85 1 13.22 -1.96 0.0545 4 vs 6 -54.04 1 13.22 -4.09 0.0001 4 vs 7 -84.17 1 13.22 -6.37 < 0.0001 4 vs 8 -192.03 1 13.22 -14.52 < 0.0001 5 vs 6 -28.19 1 13.22 -2.13 0.0364 5 vs 7 -58.32 1 13.22 -4.41 < 0.0001 5 vs 8 -166.18 1 13.22 -12.57 < 0.0001 6 vs 7 -30.13 1 13.22 -2.28 0.0257 6 vs 8 -137.99 1 13.22 -10.44 < 0.0001 7 vs 8 -107.86 1 13.22 -8.16 < 0.0001 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, 6 – 20 %, 7 – 30 %, and 8 – 35 % 183 Table A.3. Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (2002D)/EAC Blends. Source Sum of Squares Model 106483.06 7 15211.87 293.57 < 0.0001 EAC Content 106483.06 7 15211.87 293.57 < 0.0001 Pure Error 3730.77 72 51.82 Cor Total 110213.83 79 DF Mean Square F Value PLA 2002D: Semicrystalline 184 Prob > F significant Table A.4. Pairwise Comparison of the Impact Strength of PLA (2002D)/EAC Blends. EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 2.17 1 3.22 0.67 0.5025 1 vs 3 -2.73 1 3.22 -0.85 0.3989 1 vs 4 -18.73 1 3.22 -5.82 < 0.0001 1 vs 5 -29.41 1 3.22 -9.14 < 0.0001 1 vs 6 -63.25 1 3.22 -19.65 < 0.0001 1 vs 7 -72.23 1 3.22 -22.44 < 0.0001 1 vs 8 -102.28 1 3.22 -31.77 < 0.0001 2 vs 3 -4.90 1 3.22 -1.52 0.1323 2 vs 4 -20.90 1 3.22 -6.49 < 0.0001 2 vs 5 -31.58 1 3.22 -9.81 < 0.0001 2 vs 6 -65.42 1 3.22 -20.32 < 0.0001 2 vs 7 -74.40 1 3.22 -23.11 < 0.0001 2 vs 8 -104.45 1 3.22 -32.45 < 0.0001 3 vs 4 -16.00 1 3.22 -4.97 < 0.0001 3 vs 5 -26.68 1 3.22 -8.29 < 0.0001 3 vs 6 -60.51 1 3.22 -18.80 < 0.0001 3 vs 7 -69.49 1 3.22 -21.59 < 0.0001 3 vs 8 -99.55 1 3.22 -30.92 < 0.0001 4 vs 5 -10.68 1 3.22 -3.32 0.0014 4 vs 6 -44.51 1 3.22 -13.83 < 0.0001 4 vs 7 -53.49 1 3.22 -16.62 < 0.0001 4 vs 8 -83.54 1 3.22 -25.95 < 0.0001 5 vs 6 -33.83 1 3.22 -10.51 < 0.0001 5 vs 7 -42.81 1 3.22 -13.30 < 0.0001 5 vs 8 -72.86 1 3.22 -22.63 < 0.0001 6 vs 7 -8.98 1 3.22 -2.79 0.0067 6 vs 8 -39.03 1 3.22 -12.12 < 0.0001 7 vs 8 -30.05 1 3.22 -9.34 < 0.0001 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, 6 – 20 %, 7 – 25 %, and 8 – 30 % 185 Table A.5. Analysis of Variance for the Effect of EAC Content on the Impact Strength of PLA (8302D)/EAC Blends. Source Sum of Squares Model 18789.90 7 2684.27 185.05 < 0.0001 EAC Content 18789.90 7 2684.27 185.05 < 0.0001 Pure Error 1058.93 73 14.51 Cor Total 19848.84 80 DF Mean Square F Value PLA 8302D: Amorphous 186 Prob > F significant Table A.6. Pairwise Comparison of the Impact Strength of PLA (8302D)/EAC Blends. EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -0.34 1 1.75 -0.19 0.8461 1 vs 3 -1.55 1 1.75 -0.89 0.3776 1 vs 4 -5.24 1 1.75 -2.99 0.0038 1 vs 5 -18.06 1 1.80 -10.06 < 0.0001 1 vs 6 -25.62 1 1.71 -14.97 < 0.0001 1 vs 7 -32.41 1 1.71 -18.93 < 0.0001 1 vs 8 -40.82 1 1.71 -23.85 < 0.0001 2 vs 3 -1.21 1 1.70 -0.71 0.4788 2 vs 4 -4.90 1 1.70 -2.87 0.0053 2 vs 5 -17.72 1 1.75 -10.12 < 0.0001 2 vs 6 -25.28 1 1.66 -15.19 < 0.0001 2 vs 7 -32.07 1 1.66 -19.27 < 0.0001 2 vs 8 -40.48 1 1.66 -24.33 < 0.0001 3 vs 4 -3.68 1 1.70 -2.16 0.0339 3 vs 5 -16.51 1 1.75 -9.43 < 0.0001 3 vs 6 -24.07 1 1.66 -14.46 < 0.0001 3 vs 7 -30.85 1 1.66 -18.54 < 0.0001 3 vs 8 -39.27 1 1.66 -23.60 < 0.0001 4 vs 5 -12.82 1 1.75 -7.33 < 0.0001 4 vs 6 -20.39 1 1.66 -12.25 < 0.0001 4 vs 7 -27.17 1 1.66 -16.33 < 0.0001 4 vs 8 -35.59 1 1.66 -21.38 < 0.0001 5 vs 6 -7.56 1 1.71 -4.42 < 0.0001 5 vs 7 -14.35 1 1.71 -8.38 < 0.0001 5 vs 8 -22.76 1 1.71 -13.30 < 0.0001 6 vs 7 -6.79 1 1.62 -4.18 < 0.0001 6 vs 8 -15.20 1 1.62 -9.36 < 0.0001 7 vs 8 -8.41 1 1.62 -5.18 < 0.0001 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, 6 – 20 %, 7 – 25 %, and 8 – 30 % 187 THE FOLLOWING APPENDICES DEALS WITH THE STATISTICAL ANALYSES OF THE EFFECT OF IMPACT MODIFIER CONTENTS ON THE TENSILE PROPERTIES OF PLA (3001D)/EAC BLENDS 188 Table A.7. Analysis of Variance for the Effect of EAC Content on the Energy to Break of PLA (3001D)/EAC Blends. Source Sum of Squares DF Model 1761.65 5 352.33 43.45 < 0.0001 EAC Content 1761.65 5 352.33 43.45 < 0.0001 Pure Error 364.91 45 8.11 Cor Total 2126.56 50 Table A.8. Mean Square F Value Prob > F significant Pairwise Comparison of the Energy to Break of PLA (3001D)/EAC Blends. EAC Content Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -0.71 1 1.74 -0.41 0.6858 1 vs 3 -1.28 1 1.68 -0.76 0.4528 1 vs 4 -6.55 1 1.68 -3.89 0.0003 1 vs 5 -16.12 1 1.71 -9.42 < 0.0001 1 vs 6 -11.20 1 1.68 -6.65 < 0.0001 2 vs 3 -0.57 1 1.35 -0.42 0.6772 2 vs 4 -5.84 1 1.35 -4.33 < 0.0001 2 vs 5 -15.41 1 1.38 -11.13 < 0.0001 2 vs 6 -10.49 1 1.35 -7.77 < 0.0001 3 vs 4 -5.28 1 1.27 -4.14 0.0001 3 vs 5 -14.84 1 1.31 -11.34 < 0.0001 3 vs 6 -9.93 1 1.27 -7.79 < 0.0001 4 vs 5 -9.56 1 1.31 -7.31 < 0.0001 4 vs 6 -4.65 1 1.27 -3.65 0.0007 5 vs 6 4.91 1 1.31 3.76 0.0005 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, and 6 – 20 % 189 Table A.9. Analysis of Variance for the Effect of EAC Content on the Tensile Strength of PLA (3001D)/EAC Blends. Source Sum of Squares DF Model 4405.79 5 881.16 404.36 < 0.0001 EAC Content 4405.79 5 881.16 404.36 < 0.0001 Pure Error 98.06 45 2.18 Cor Total 4503.85 50 Table A.10. Mean Square F Value Prob > F significant Pairwise Comparison of the Tensile Strength of PLA (3001D)/EAC Blends. EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 0.19 1 0.90 0.21 0.8366 1 vs 3 1.30 1 0.87 1.49 0.1425 1 vs 4 11.96 1 0.87 13.69 < 0.0001 1 vs 5 17.10 1 0.89 19.27 < 0.0001 1 vs 6 24.12 1 0.87 27.62 < 0.0001 2 vs 3 1.12 1 0.70 1.59 0.1180 2 vs 4 11.77 1 0.70 16.81 < 0.0001 2 vs 5 16.91 1 0.72 23.57 < 0.0001 2 vs 6 23.94 1 0.70 34.18 < 0.0001 3 vs 4 10.66 1 0.66 16.14 < 0.0001 3 vs 5 15.79 1 0.68 23.28 < 0.0001 3 vs 6 22.82 1 0.66 34.57 < 0.0001 4 vs 5 5.14 1 0.68 7.57 < 0.0001 4 vs 6 12.17 1 0.66 18.43 < 0.0001 5 vs 6 7.03 1 0.68 10.36 < 0.0001 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, and 6 – 20 % 190 Table A.11. Analysis of Variance for the Effect of EAC Content on the Tensile Modulus of PLA (3001D)/EAC Blends. Source Sum of Squares DF Model 2061337.46 5 412267.49 1120.17 < 0.0001 EAC Content 2061337.46 5 412267.49 1120.17 < 0.0001 Pure Error 16561.81 45 368.04 Cor Total 2077899.27 50 Table A.12. Pairwise Comparison of the Tensile Modulus of PLA (3001D)/EAC Blends. Mean Square F Value Prob > F significant EAC Content Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 87.51 1 11.75 7.45 < 0.0001 1 vs 3 115.77 1 11.35 10.20 < 0.0001 1 vs 4 269.37 1 11.35 23.73 < 0.0001 1 vs 5 411.97 1 11.53 35.74 < 0.0001 1 vs 6 599.96 1 11.35 52.86 < 0.0001 2 vs 3 28.26 1 9.10 3.11 0.0033 2 vs 4 181.86 1 9.10 19.99 < 0.0001 2 vs 5 324.46 1 9.32 34.81 < 0.0001 2 vs 6 512.45 1 9.10 56.31 < 0.0001 3 vs 4 153.60 1 8.58 17.90 < 0.0001 3 vs 5 296.20 1 8.81 33.60 < 0.0001 3 vs 6 484.19 1 8.58 56.44 < 0.0001 4 vs 5 142.60 1 8.81 16.18 < 0.0001 4 vs 6 330.59 1 8.58 38.53 < 0.0001 5 vs 6 187.99 1 8.81 21.33 < 0.0001 EAC Contents: 1 – 0 %, 2 – 3 %, 3 – 5 %, 4 – 10 %, 5 – 15 %, and 6 – 20 % 191 THE FOLLOWING APPENDICES DEALS WITH THE STATISTICAL ANALYSES OF THE EFFECT OF TEMPERATURE ON THE IMPACT STRENGTH OF NEAT PLA AND PLA/15 % EAC BLEND 192 Table A.13. Analysis of Variance for the Effect of Temperature on the Impact Strength of Neat PLA 3001D. Source Sum of Squares DF Model 89.47 6 14.91 5.61 < 0.0001 Temperature 89.47 6 14.91 5.61 < 0.0001 Pure Error 199.39 75 2.66 Cor Total 288.87 81 Mean Square F Value 193 Prob > F significant Table A.14. Pairwise Comparison of the Impact Strength of Neat PLA 3001D at Different Temperatures. Temperature Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 1.87 1 0.72 2.61 0.0110 1 vs 3 -0.11 1 0.72 -0.15 0.8820 1 vs 4 2.43 1 0.71 3.44 0.0009 1 vs 5 2.93 1 0.72 4.08 0.0001 1 vs 6 1.71 1 0.71 2.41 0.0182 1 vs 7 1.41 1 0.73 1.92 0.0583 2 vs 3 -1.98 1 0.67 -2.98 0.0039 2 vs 4 0.56 1 0.65 0.86 0.3929 2 vs 5 1.06 1 0.67 1.59 0.1158 2 vs 6 -0.17 1 0.65 -0.26 0.7984 2 vs 7 -0.46 1 0.68 -0.68 0.4968 3 vs 4 2.54 1 0.65 3.89 0.0002 3 vs 5 3.04 1 0.67 4.57 < 0.0001 3 vs 6 1.81 1 0.65 2.78 0.0069 3 vs 7 1.52 1 0.68 2.23 0.0289 4 vs 5 0.50 1 0.65 0.76 0.4479 4 vs 6 -0.73 1 0.64 -1.14 0.2584 4 vs 7 -1.03 1 0.67 -1.54 0.1289 5 vs 6 -1.23 1 0.65 -1.88 0.0642 5 vs 7 -1.52 1 0.68 -2.24 0.0281 6 vs 7 -0.30 1 0.67 -0.45 0.6574 Temperatures: 1 – 23, 2 – 6.9, 3 – 1.2, 4 - -14.6, 5 - -17.3, 6 - -22, and 7 - -26 194 Table A.15. Analysis of Variance for the Effect of Temperature on the Impact Strength of PLA 3001D/15 % EAC Blend. Source Sum of Squares Model 4601.66 6 766.94 17.95 < 0.0001 Temperature 4601.66 6 766.94 17.95 < 0.0001 Pure Error 3846.01 90 42.73 Cor Total 8447.67 96 DF Mean Square F Value 195 Prob > F significant Table A.16. Pairwise Comparison of the Impact Strength of PLA 3001D/15 % EAC Blends at Different Temperatures. Temperatures Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 6.70 1 2.39 2.81 0.0061 1 vs 3 12.75 1 2.39 5.34 < 0.0001 1 vs 4 14.24 1 2.33 6.11 < 0.0001 1 vs 5 16.71 1 2.28 7.34 < 0.0001 1 vs 6 19.46 1 2.33 8.36 < 0.0001 1 vs 7 16.76 1 2.33 7.20 < 0.0001 2 vs 3 6.05 1 2.67 2.27 0.0258 2 vs 4 7.54 1 2.62 2.88 0.0050 2 vs 5 10.01 1 2.57 3.89 0.0002 2 vs 6 12.77 1 2.62 4.88 < 0.0001 2 vs 7 10.06 1 2.62 3.84 0.0002 3 vs 4 1.49 1 2.62 0.57 0.5706 3 vs 5 3.96 1 2.57 1.54 0.1267 3 vs 6 6.72 1 2.62 2.57 0.0119 3 vs 7 4.01 1 2.62 1.53 0.1289 4 vs 5 2.47 1 2.52 0.98 0.3284 4 vs 6 5.23 1 2.56 2.04 0.0444 4 vs 7 2.52 1 2.56 0.98 0.3282 5 vs 6 2.75 1 2.52 1.09 0.2772 5 vs 7 0.05 1 2.52 0.02 0.9854 6 vs 7 -2.71 1 2.56 -1.06 0.2939 Temperatures: 1 – 23, 2 – 6.9, 3 – 1.2, 4 - -3.2, 5 - -15, 6 - -21, and 7 - -27 196 APPENDIX B THE FOLLOWING APPENDICES DEALS WITH CHAPTER 5 AND PRESENTS THE STATISTICAL ANALYSES OF THE EFFECT OF WOOD FLOUR AND IMPACT MODIFIER CONTENTS ON THE FRACTURE TOUGHNESS OF PLA/WOOD-FLOUR COMPOSITES 197 Table B.1. Analysis of Variance for the Effect of Wood Flour Content on the J-Integral (Jin) of the Composites. Source Sum of Squares DF Mean Square F Value Prob > F Model 1451673.36 4 362918.34 11.27 < 0.0001 Wood Flour Content 1451673.36 4 362918.34 11.27 < 0.0001 Pure Error 1126878.34 35 32196.52 Cor Total 2578551.70 39 Table B.2. significant Pairwise Comparison of the J-Integral (Jin) of the PLA/Wood-Flour Composites. Wood Flour Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -112.61 1 105.07 -1.07 0.2911 1 vs 3 -9.76 1 105.07 -0.09 0.9265 1 vs 4 163.68 1 84.12 1.95 0.0597 1 vs 5 436.30 1 88.43 4.93 < 0.0001 2 vs 3 102.85 1 113.48 0.91 0.3710 2 vs 4 276.29 1 94.42 2.93 0.0060 2 vs 5 548.91 1 98.28 5.59 < 0.0001 3 vs 4 173.44 1 94.42 1.84 0.0747 3 vs 5 446.06 1 98.28 4.54 < 0.0001 4 vs 5 272.62 1 75.47 3.61 0.0009 Wood Flour Contents: 1 – 0 %, 2 – 10%, 3 – 20%, 4 – 30 %, and 5 – 40 % 198 Table B.3. Analysis of Variance for the Effect of Wood Flour Content on the Fracture Energy (Jf) of the Composites. Source Sum of Squares DF Mean Square F Value Prob > F Model 8989785.84 4 2247446.4 6 18.01 < 0.0001 Wood Flour Content 8989785.84 4 2247446.4 6 18.01 < 0.0001 Pure Error 4367980.69 35 124799.45 Cor Total 13357766.53 39 Table B.4. significant Pairwise Comparison of the Fracture Energy (Jf) of the PLA/Wood-Flour Composites. Wood Flour Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -53.71 1 206.85 -0.26 0.7967 1 vs 3 288.44 1 206.85 1.39 0.1720 1 vs 4 660.78 1 165.62 3.99 0.0003 1 vs 5 1224.41 1 174.09 7.03 < 0.0001 2 vs 3 342.15 1 223.43 1.53 0.1347 2 vs 4 714.49 1 185.90 3.84 0.0005 2 vs 5 1278.12 1 193.49 6.61 < 0.0001 3 vs 4 372.34 1 185.90 2.00 0.0530 3 vs 5 935.98 1 193.49 4.84 < 0.0001 4 vs 5 563.63 1 148.59 3.79 0.0006 Wood Flour Contents: 1 – 0 %, 2 – 10%, 3 – 20%, 4 – 30 %, and 5 – 40 % 199 Table B.5. Analysis of Variance for the Effect of EAC Content on the J-Integral (Jin) of the Composites with 40 % Wood Flour Content. Source Sum of Squares Model 950110.97 4 237527.74 9.27 < 0.0001 EAC Content 950110.97 4 237527.74 9.27 < 0.0001 Pure Error 2151546.39 84 25613.65 Cor Total 3101657.36 88 Table B.6. DF Mean Square F Value Prob > F significant Pairwise Comparison of the J-Integral (Jin) of the PLA/Wood-Flour/EAC Composites. EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -286.34 1 59.88 -4.78 < 0.0001 1 vs 3 -375.20 1 68.53 -5.48 < 0.0001 1 vs 4 -330.41 1 63.78 -5.18 < 0.0001 1 vs 5 -302.98 1 59.88 -5.06 < 0.0001 2 vs 3 -88.87 1 56.21 -1.58 0.1176 2 vs 4 -44.07 1 50.31 -0.88 0.3835 2 vs 5 -16.64 1 45.27 -0.37 0.7140 3 vs 4 44.79 1 60.34 0.74 0.4599 3 vs 5 72.22 1 56.21 1.29 0.2023 4 vs 5 27.43 1 50.31 0.55 0.5871 EAC Contents: 1 – 0 %, 2 – 5 %, 3 – 10 %, 4 – 15 %, and 5 – 20 % 200 Table B.7. Analysis of Variance for the Effect of EAC Content on the Fracture Energy (Jf) of the Composites with 40 % Wood Flour Content. Source Sum of Squares DF Model 11355656.98 4 2838914.25 31.81 < 0.0001 EAC Content 11355656.98 4 2838914.25 31.81 < 0.0001 Pure Error 7676143.87 86 89257.49 Cor Total 19031800.85 90 Table B.8. Pairwise Comparison of the Fracture Energy (Jf) of the PLA/Wood-Flour/EAC Composites. Mean Square F Value Prob > F significant EAC Contents Mean Difference DF Standard Error t for Ho Coeff = 0 Prob > |t| 1 vs 2 -865.65 1 111.79 -7.74 < 0.0001 1 vs 3 -942.59 1 123.70 -7.62 < 0.0001 1 vs 4 -1203.29 1 119.06 -10.11 < 0.0001 1 vs 5 -1160.85 1 111.79 -10.38 < 0.0001 2 vs 3 -76.94 1 99.73 -0.77 0.4426 2 vs 4 -337.64 1 93.92 -3.59 0.0005 2 vs 5 -295.20 1 84.50 -3.49 0.0008 3 vs 4 -260.70 1 107.82 -2.42 0.0177 3 vs 5 -218.27 1 99.73 -2.19 0.0313 4 vs 5 42.43 1 93.92 0.45 0.6525 EAC Contents: 1 – 0 %, 2 – 5 %, 3 – 10 %, 4 – 15 %, and 5 – 20 % 201