BAHAGIAN A – Pengesahan Kerjasama* Adalah disahkan bahawa projek penyelidikan tesis ini telah dilaksanakan melalui kerjasama antara _____________________ dengan ______________________ Disahkan oleh: Tandatangan : ………………………………….. Tarikh: ………………. Nama : ………………………………….. Jawatan (Cop rasmi) : ………………………………….. * Jika penyelidikan tesis/projek melibatkan kerjasama. BAHAGIAN B – Untuk Kegunaan Pejabat Sekolah Pengajian Siswazah Tesis ini telah diperiksa dan diakui oleh: Nama dan Alamat Pemeriksa Luar : Nama dan Alamat Pemeriksa Dalam I : Pemeriksa Dalam II : Prof. Dr. Sahrim Bin Hj. Ahmad School of Applied Physics Faculty of Science & Technology Universiti Kebangsaan Malaysia 43600 UKM Bangi, Selangor Assoc. Prof. Dr. Wan Aizan Bt. Wan Abd. Rahman Department of Polymer Engineering Faculty of Chemical and Natural Resources Engineering, Universiti Teknologi Malaysia 81310 UTM Skudai, Johor Assoc. Prof. Dr. Shahrir Bin Hashim Department of Polymer Engineering Faculty of Chemical and Natural Resources Engineering, Universiti Teknologi Malaysia 81310 UTM Skudai, Johor Nama Penyelia Lain : (jika ada) Disahkan oleh Timbalan Pendaftar di SPS: Tandatangan : …………………………………. Nama ………………………………… : Tarikh: ……………… MECHANICAL, THERMAL AND PROCESSING PROPERTIES OF OIL PALM EMPTY FRUIT BUNCH-FILLED IMPACT MODIFIED UNPLASTICISED POLY (VINYL CHLORIDE) COMPOSITES AZNIZAM BIN ABU BAKAR A thesis submitted in fulfilment of the requirements for the award of the degree of Doctor of Philosophy Faculty of Chemical and Natural Resources Engineering Universiti Teknologi Malaysia JANUARY 2006 ii iii To my beloved wife, parents, son and daughter… iv ACKNOWLEDGEMENT I would like to express my sincere gratefulness and appreciation to my supervisor, Associate Professor Dr. Azman Hassan for his tremendous encouragement and endless support during my Ph.D. study. I learned the art of science throughout your supervision. You taught me what is not taught in the course. I am very grateful to Dr. Ahmad Fuad Mohd Yusof, my co-supervisor, from SIRIM. My special thanks to my colleagues, all from Universiti Teknologi Malaysia, who provided constructive and scientific advice for the completion of this thesis. This include: Dr. Afendi, Munirah Mokhtar, Zainab Salleh, Sukor Ishak, Suhee Tan Hassan, Nordin Ahmad, Ambiga Gopal, Jaafar, Jamil Tahir, Ahmad Norani Sadiron, Yahya Khalid, Arshad Abu Hassan, Mazatusziha and Toh Soh Chong. I would also like to thank IRM, especially Dato Karim, for providing PVC resins and additives, and allowing me to use the Brabender Torque Rheometer equipment for pocessability studies. My special thanks also go to Wan Ali from MINT for his assistance in extrusion processing. This study will not to be accomplished without finance and therefore I would like to express my appreciation to the Malaysia Government and Universiti Teknologi Malaysia for the financial support. Finally, I wish to thank to my beloved wife, son and daughter, Sukina, Ahmad Afzanizam and Afza Firzana for their love, support, and understanding and most importantly for their patience. v ABSTRACT The main objective of this study is to investigate the effects of EFB filler and two types of impact modifiers on mechanical and thermal properties, and processability of PVC-U composites. Two methods were used to prepare the samples, i.e., compression moulding and single screw extrusion. A torque rheometer was used to determine the fusion characteristics of PVC. EFB increased the flexural modulus of 22 % but decreased the impact strength and flexural strength of 20 % and 18 % of PVC, respectively, as the filler contents increased. Increasing the impact modifiers contents, on the other hand, improved the impact strength but slightly decreased the flexural modulus and flexural strength. Acrylic impact modifier performed better than CPE in producing composites with balanced properties of toughness and stiffness. The use of Prosil 9234 coupling agent for filler surface treatment proved to be more reliable compared to NZ 44 in providing better filler-matrix interaction. The TGA study showed that both the EFB filler and acrylic impact modifier decreased the PVC thermal stability. Oil residues in the EFB did not influence the mechanical properties of PVC-U composites. The fusion time also decreased significantly with the addition of acrylic. Both the temperature and screw speed significantly influenced the fusion level of PVC, hence the mechanical properties of the extrudates. Overall, the study showed that EFB has the potential to be used as filler for PVC-U composites. vi ABSTRAK Objektif utama kajian ini ialah untuk menyelidik kesan pengisi EFB dan pengubahsuai hentaman terhadap sifat mekanikal, terma dan kebolehanproses komposit PVC-U. Dua kaedah telah digunakan untuk penyediaan sampel, iaitu pengacuan mampatan dan penyemperitan berskru tunggal. Reometer tork telah digunakan untuk menentukan ciri-ciri pelakuran PVC. EFB telah meningkatkan modulus lenturan sebanyak 22 % tetapi telah mengurangkan kekuatan hentaman PVC sebanyak 20 % dan kekuatan lenturan sebanyak 18 % apabila meningkatnya kandungan pengisi. Peningkatan kandungan pengubahsuai hentaman sebaliknya telah meningkatkan kekuatan hentaman tetapi mengurangkan sedikit modulus lenturan dan kekuatan lenturan. Pengubahsuai hentaman akrilik telah menunjukkan prestasi yang lebih baik dalam menghasilkan komposit bersifat kekukuhan dan kekakuan yang seimbang berbanding CPE. Penggunaan agen gandingan Prosil 9234 untuk rawatan permukaan pengisi telah menghasilkan interaksi pengisi-matrik yang lebih baik berbanding dengan NZ 44. Kajian TGA menunjukkan bahawa pengisi EFB dan pengubahsuai hentaman akrilik telah mengurangkan kestabilan terma PVC. Sisa minyak pada EFB didapati kurang mempengaruhi sifat-sifat mekanikal komposit PVC-U. Masa pelakuran juga telah menurun dengan penambahan kandungan akrilik. Suhu dan laju skru telah mempengaruhi tahap lakuran PVC begitu juga sifat-sifat mekanikal ekstrudat. Keseluruhannya kajian menunjukkan bahawa EFB berpotensi digunakan sebagai pengisi di dalam komposit PVC. vii TABLE OF CONTENTS CHAPTER 1 2 TITLE PAGE TITLE PAGE i DECLARATION ii DEDICATION iii ACKNOWLEDGEMENT iv ABSTRACT v ABSTRAK vi TABLE OF CONTENTS vii LIST OF TABLES xiii LIST OF FIGURES xviii LIST OF SYMBOLS xxv INTRODUCTION 1 1.1 Overview 1 1.2 Significance of Study 4 1.3 Objectives of Study 6 1.4 Outline of Thesis 7 LITERATURE REVIEW 9 2.1 Poly (vinyl chloride) 9 2.1.1 Definitions and General Feature Properties 9 2.1.2 PVC Polymerisation 11 2.1.3 Morphology of PVC 13 viii 2.1.4 PVC Fusion 16 2.1.4.1 Mechanisms of Fusion 17 2.1.4.2 Fusion Level Assessment: Differential Scanning Calorimetry 19 2.1.4.3 Interrelation between Fusion and Properties 2.2 2.3 2.4 21 PVC and Additives 24 2.2.1 Stabilizer 24 2.2.2 Lubricants 26 2.2.3 Processing Aid 32 2.2.4 Pigment 34 2.2.5 Impact Modifier 35 Impact Modification 38 2.3.1 Theory of Impact Modification 38 2.3.2 Impact Modified PVC 40 Filled PVC Composites 42 2.4.1 Composites: Definitions and Categories 42 2.4.2 Filler and Reinforcement 44 2.4.2.1 Mineral Fillers 46 2.4.2.2 Organic Fillers 47 Filler-Matrix Interface 50 2.4.3 2.4.3.1 Coupling Agent and Surface Modification 2.5 Oil Palm Empty Fruit Bunch-Polymer Composites 2.5.1 56 Theory of Mechanical Properties of Filled Composites 2.7 55 Utilization of EFB Fibre in Polymer Composites 2.6 54 Oil Palm Empty Fruit Bunch and its Chemical Composition 2.5.2 51 59 Recent Studies on Impact Modified PVC-U by UTM researchers 63 ix 3 MECHANICAL PROPERTIES OF COMPRESSION MOULDED PVC-U COMPOSITES 65 3.1 Introduction 65 3.2 Experimental 67 3.2.1 Materials 67 3.2.2 Filler Preparations 69 3.2.3 Coupling Agent Treatment of Filler 71 3.2.4 Blend Formulations 72 3.2.5 Dry Blending 82 3.2.6 Two Roll Milling 82 3.2.7 Compression Moulding 84 3.2.8 Density Determination 85 3.2.9 Impact Testing 87 3.3 3.2.10 Flexural Testing 87 3.2.11 Accelerated Weathering Testing 88 3.2.12 Water Absorption Testing 89 3.2.13 Scanning Electron Microscopy Study 89 3.2.14 Fourier Transform Infra-Red Analysis 90 Results and Discussion 90 3.3.1 3.3.2 Density of the Filler and EFB-filled PVC-U Composites 90 Mechanical Properties 92 3.3.2.1 Notched Izod Impact Strength 92 3.3.2.2 Flexural Properties 108 3.3.2.3 EFB-Filled Acrylic-Impact Modified PVC-U Composites Treated with Coupling Agents 120 3.3.2.4 Effect of Accelerated Weathering on the Impact and Flexural Properties 137 3.3.2.5 Effect of NPCC on the Impact and Flexural Properties 153 3.3.2.6 Effect of Extracted Filler on the Impact and Flexural Properties 159 x 3.3.2.7 Water Absorption 3.4 4 Conclusions 167 THERMAL PROPERTIES OF COMPRESSION MOULDED PVC-U COMPOSITES 169 4.1 Introduction 169 4.2 Experimental 171 4.2.1 4.3 4.4 5 162 Materials, Filler Preparation and Blend Formulations 172 4.2.2 Sample Preparation 172 4.2.3 Heat Deflection Temperature Testing 172 4.2.4 Differential Scanning Calorimetry Study 173 4.2.5 Thermogravimetric Analysis 173 Results and Discussion 174 4.3.1 Heat Deflection Temperature 174 4.3.2 Glass Transition Temperature 177 4.3.3 Thermal Degradation Temperature 178 Conclusions 190 PROCESSABILITY STUDIES OF PVC-U COMPOUNDS 192 5.1 Introduction 192 5.2 Experimental 195 5.2.1 Materials and Blend Formulations 195 5.2.2 Filler Preparation 195 5.2.2.1 EFB Oil Extraction 195 5.2.3 Dry Blending 197 5.2.4 Processability Study 198 5.2.5 Fourier Transform Infra-Red Analysis 198 5.2.6 Scanning Electron Microscopy Study 199 5.3 Results and Discussion 199 5.3.1 Fusion Characteristics 199 5.3.1.1 Effect of EFB Filler 201 5.3.1.2 Effect of Acrylic Impact Modifier 203 xi 5.3.1.3 Effect of CPE Impact Modifier 207 5.3.1.4 Effect of Temperature 212 5.3.2 Postulated Lubrication Mechanisms 214 5.3.3 Effect of Extracted EFB Filler on the Processability of PVC-U Compound 219 5.3.3.1 Extractives Removal 219 5.3.3.2 FTIR Spectra 219 5.3.3.3 Extracted Filler Surface 222 5.3.3.4 Fusion Characteristics of Extracted EFB-Filled PVC-U Compound 5.4 6 Conclusions 223 225 EXTRUSION PROCESS AND MECHANICAL PROPERTIES OF EXTRUDATES 227 6.1 Introduction 227 6.2 Experimental 229 6.2.1 Materials 229 6.2.2 Dry Blending 230 6.2.3 Extrusion Process 230 6.2.4 Izod Impact and Flexural Testing 231 6.2.5 Tensile Testing 232 6.2.6 Differential Scanning Calorimetry Study 232 6.3 Results and Discussion 232 6.3.1 Extrudates Characteristics 233 6.3.2 Differential Scanning Calorimetry 238 6.3.3 Impact Strength 240 6.3.3.1 Effect of Processing Temperature 240 6.3.3.2 Effect of Screw Speed 244 Flexural Properties 246 6.3.4.1 Effect of Processing Temperature 246 6.3.4.2 Effect of Screw Speed 248 6.3.5 Tensile Properties 249 6.3.6 Yield Stress Analysis 255 6.3.7 General Discussion 258 6.3.4 xii 6.4 7 Conclusions 260 CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK 262 7.1 Conclusions 262 7.2 Suggestions for Future Work 264 REFERENCES 266-282 xiii LIST OF TABLES TABLE NO. 2.1 TITLE PAGE Some physical properties of moulding PVC products and compounds 11 2.2 PVC nomenclatures 15 2.3 Chemical analysis of EFB 55 3.1 Specifications of PVC suspension resin MH-66 68 3.2 Types of additives 68 3.3 Blend formulations 73 3.4 Effect of impact modifiers content on the impact strength of PVC-U compounds 3.5 94 Percentage of impact strength reduction of EFB-filled impact modified composites as filler content increased from 10 to 40 phr 3.6 101 Percentage of impact strength increment for EFB-filled impact modified composites as the impact modifier content increased from 0 to 12 phr 3.7 107 Percentage of flexural modulus increment for EFB-filled impact modified composites as the filler content increased from 0 to 40 phr 3.8 114 Percentage of flexural strength reduction for EFB-filled impact modified composites as the filler content increased from 0 to 40 phr 3.9 Percentage of flexural modulus reduction for EFB-filled impact modified composites as the impact modifier 114 xiv increased from 0 to 12 phr 3.10 118 Percentage of flexural strength reduction for EFB-filled impact modified composites as the impact modifier increased from 0 to 12 phr 3.11 Impact strength of EFB–filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 3.12 130 Comparison between the predicted and the experimental of relative modulus of composites 3.15 130 Flexural strength of EFB–filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 3.14 126 Flexural modulus of EFB–filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 3.13 118 131 Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 3 phr of acrylic 3.16 132 Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 6 phr of acrylic 3.17 132 Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 9 phr of acrylic 3.18 132 Comparison between the predicted and the experimental relative flexural modulus of filled acrylic-impact modified composites at 12 phr of acrylic 3.19 Impact strength reduction for impact-modified compounds after accelerated weathering for 504 hours 3.20 139 Percentage of impact strength reduction for filled impact modified composites after weathered for 504 hours 3.21 133 142 Percentage of flexural modulus reduction and percentage of flexural strength increment of the impact-modified compounds after 504 hours exposure to accelerated weathering 3.22 Percentage of flexural modulus reduction and percentage of flexural strength increment of the filled composites and filled impact-modified composites after 504 hours exposure 146 xv to accelerated weathering 3.23 Percentage of water absorption of impact modified compounds and EFB-filled composites 3.24 163 Percentage of water absorption of impact modified compounds and impact modified treated composites 4.1 146 166 HDT of acrylic-impact modified and CPE-impact modified PVC-U compounds 175 4.2 HDT of EFB-filled PVC-U composites 176 4.3 HDT of EFB-filled acrylic-impact modified and EFB-filled CPE-impact modified PVC-U composites 176 4.4 Tg of EFB-filled PVC-U composites 177 4.5 Tg of EFB-filled acrylic-impact modified and EFB-filled CPE-impact modified PVC-U composites 4.6 Degradation temperature of neat components obtained from TG and DTG curves 4.7 185 Degradation temperature of EFB-filled CPE-impact modified PVC-U composites 4.9 180 Degradation temperature of EFB-filled PVC-U composites obtained from TG and DTG curves 4.8 178 187 Degradation temperature of EFB-filled acrylic-impact modified PVC-U composites obtained from TG and DTG curves 5.1 Fusion characteristics of EFB-filled compounds at 50 rpm and 80 oC 5.2 202 Fusion characteristics of acrylic-impact modified compounds at 50 rpm and 180 oC 5.3 189 204 Fusion characteristics of EFB-filled acrylic-impact modified compounds added with 9 phr of acrylic at 50 rpm and 180 oC 5.4 205 Fusion characteristics of EFB-filled acrylic-impact modified compounds filled with 20 phr of filler at 50 rpm and 180 oC 5.5 206 Fusion characteristics of CPE-impact modified compounds at 50 rpm and 180 oC 208 xvi 5.6 Fusion characteristics of EFB-filled CPE-impact modified compounds added with 9 phr CPE at 50 rpm and 180 oC 5.7 Fusion characteristics of EFB-filled CPE-impact modified compounds filled with 20 phr of EFB at 50 rpm and 180 oC 5.8 211 Fusion time of PVC-U compounds at different processing temperatures and constant rotor speed of 50 rpm 5.9 210 212 End torque of PVC–U compounds and composites at different processing temperatures and constant rotor speed of 50 rpm 5.10 214 Fusion characteristics of extracted EFB-filled PVC-U compounds at 50 rpm and 180 oC. 224 6.1 The conditions of extrusion processing 231 6.2 Residence time of extrudates at constant die temperature 237 6.3 Residence time of extrudates at constant screw speed of 60 rpm 238 6.4 Heat of fusion of extrudates at screw speed of 60 rpm 240 6.5 Effect of processing temperatures on the fracture properties of unfilled and filled extrudates 6.6 Effect of processing temperatures on the heat of fusion and impact strength of unfilled extrudates 6.7 247 Effect of screw speeds on the flexural modulus of unfilled and filled extrudates 6.14 247 Effect of processing temperatures on the flexural strength of unfilled and filled extrudates 6.13 246 Effect of processing temperatures on the flexural modulus of filled extrudates 6.12 244 Effect of processing temperatures on the flexural modulus of unfilled extrudates 6.11 244 Effect of screw speeds on the impact strength of unfilled and filled extrudates 6.10 243 Effect of screw speeds on the fracture properties of unfilled and filled extrudates 6.9 241 Effect of processing temperatures on the heat of fusion and impact strength of filled extrudates 6.8 241 Effect of screw speeds on the flexural strength of unfilled 249 xvii and filled extrudates 6.15 Effect of processing temperatures on the yield stress of unfilled extrudates at screw speed of 60 rpm 6.16 258 Predicted yield stress of extrudates at 3 m/s, die temperature of 190 oC and screw speed of 60 rpm 6.24 255 Predicted yield stress of extrudates at 3 m/s, die temperature of 180 oC and screw speed of 60 rpm 6.23 255 Effect of strain rate on the yield stress of unfilled and filled extrudates 6.22 254 Effect of screw speed on the percentage strain at break at die temperature of 180 oC 6.21 254 Effect of processing temperature on the percentage strain at break of filled extrudates at screw speed of 60 rpm 6.20 251 Effect of processing temperature on the percentage strain at break of unfilled extrudates at screw speed of 60 rpm 6.19 251 Effect of screw speeds on the yield stress of extrudates at die temperature of 180 oC 6.18 250 Effect of processing temperature on the yield stress of filled extrudates at screw speed of 60 rpm 6.17 249 258 Impact strength, yield stress and strain at break at die temperature 180 oC and 190 oC 259 xviii LIST OF FIGURES FIGURE NO. TITLE PAGE 2.1 The chemistry of vinyl chloride monomer production 12 2.2 PVC particle morphology 15 2.3 Mechanism of fusion during processing extruder 18 2.4 Schematic presentation of fusion process for 19 2.5 DuPont thermal analysis of PVC-U compound compression moulded at different temperatures 2.6 20 Zip-elimination of HCl from PVC and simultaneous formation of sequences containing conjugated double bonds (polyenes) in the chain during degradation 2.7 25 A model of PVC lubrication (not to scale) showing metal lubrication and lubrication between PVC primary particle flow units for (a) no lubricant; (b) calcium stearate ; (c) for paraffin ; and (d) calcium stearate and paraffin 2.8 29 Lubrication mechanism, showing metal lubrication and lubrication between PVC micro particles flow units of PVC/CPE, PVC/OPE and PVC/CPE/OPE compounds 2.9 31 Stress distribution in polymer matrix surrounding a rubbery impact modifier particle 37 2.10 Aspect ratio calculation 46 2.11 Cellulose molecular structure 48 2.12 Bonding mechanism of silane coupling agent to filler’s surface 52 3.1 Chemical structure of (a) Prosil 9234 (b) NZ 44 69 3.2 EFB fillers distribution 70 xix 3.3 Morphology of EFB fillers (Magnification of 600x) 70 3.4 Two roll milling of PVC-U dry blend at 165 oC 83 3.5 Two-roll milled composite sheet 83 3.6 A mould with five cavities placed in between the hot platens before compression 3.7 The theoretical and experimental densities of EFB-filled composites (S9 - S12) with respect to EFB filler contents 3.8 84 91 SEM micrograph of impact-fractured surface of S11 (Magnification 2000x) 92 3.9 Hinge breaks of impact-modified PVC-U samples (S3) 95 3.10 SEM micrograph of impact-fractured surface of S0 (Magnification 200x) 3.11 SEM micrograph of impact-fractured surface of S3 (Magnification 200x) 3.12 96 SEM micrograph of impact-fractured surface of S7 (Magnification 200x) 3.13 96 97 Effect of filler content on the impact strength of EFB-filled composites (S9-S12) and EFB-filled acrylic-impact modified composites (S13-S28) 3.14 98 Effect of filler content on the impact strength of EFB-filled composites (S9-S12) and EFB-filled CPE-impact modified composite (S29- S44) 3.15 SEM micrograph of the impact-fracture surface of composite incorporated with 10 phr filler (Magnification 500x) 3.16 100 Polar-polar interaction between PVC molecule and cellulose molecule. 3.20 100 SEM micrograph of the impact-fractured surface of composite incorporated with 40 phr filler (Magnification 500x) 3.19 99 SEM micrograph of the impact-fractured surface of composite incorporated with 30 phr filler (Magnification 500x) 3.18 99 SEM micrograph of the impact-fracture surface of composite incorporated with 20 phr filler(Magnification 500x) 3.17 98 SEM micrograph of filler debonding on the impact-fractured 102 xx surface of composite incorporated with 30 phr filler (Magnification 900x) 3.21 103 SEM micrograph of filler bundles pulled-out area on the impactfractured surface composite incorporated with 30 phr filler of S11 (Magnification 1000x) 3.22 Effect of acrylic content on the impact strength of EFB-filled acrylic-impact modified composites (S0-S4 and S9-S28). 3.23 104 105 Effect of CPE impact modifier content on the impact strength of EFB-filled CPE-impact modified composites (S5 -S8 and S29-S44) 3.24 Effect of acrylic and CPE impact modifier on the impact strength of EFB-filled impact-modified composites 3.25 116 Effect of filler and CPE content on the flexural strength of EFB-filled CPE-impact modified composites (S29 - S44) 3.35 116 Effect of filler and CPE content on the flexural modulus of EFB-filled CPE-impact modified composites (S29 -S44) 3.34 115 Effect of filler and acrylic content on the flexural strength of EFB-filled acrylic-impact modified composites (S13-S28) 3.33 113 Effect of filler and acrylic content on the flexural modulus of EFB-filled acrylic-impact modified composites (S13-S28) 3.32 112 Effect of filler and CPE content on the flexural strength of EFB-filled CPE-impact modified composites (S29-S44) 3.31 112 Effect of filler and CPE content on the flexural modulus of EFB-filled CPE-impact modified composites (S29-S44) 3.30 111 Effect of filler and acrylic content on the flexural strength of EFB-filled acrylic-impact modified composites (S9-S28) 3.29 109 Effect of filler and acrylic content on the flexural modulus of EFB-filled acrylic-impact modified composites (S9-S28) 3.28 109 Effect of impact modifiers content on the flexural strength of impact modified compounds (S0-S8). 3.27 107 Effect of impact modifiers content on the flexural modulus of impact modified compounds (S0-S8) 3.26 105 Effect of acrylic and CPE impact modifier on the flexural 117 xxi modulus of EFB-filled impact modified composites 3.36 Effect of acrylic and CPE impact modifier on the flexural strength of EFB-filled impact modified composites 3.37 119 119 Effect of EFB filler and acrylic content on the impact strength of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S49-S68) 3.38 121 Effect of EFB filler and acrylic content on the impact strength of EFB-filled acrylic-impact modified composites treated with NZ 44 (S69-S88) 3.39 Bonding mechanism of NZ 44 coupling agent to the filler’s surface 3.40 121 122 SEM micrographs of the impact fractured-surface of composites (a) Prosil 9234 treatment (S51) (Magnification 100x) (b) S51 (Magnification 2000x) (c) NZ 44 treatment (S71) (Magnification 500x) and (d) S71 (Magnification 2000x) 3.41 125 Effect of EFB filler and acrylic contents on the flexural modulus of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S9-S28 and S49-S68) 3. 42 127 Effect of EFB filler and acrylic contents on the flexural strength of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S9-S28 and S49-S68) 3.43 128 Effect of EFB filler and acrylic contents on the flexural modulus of EFB-filled acrylic-impact modified composites treated with NZ 44 (S9-S28 and S69-S88) 3.44 128 Effect of EFB filler and acrylic content on the flexural strength of EFB-filled acrylic-impact modified composites treated with NZ 44 (S9-S28 and S69-S88) 129 3.45 FTIR spectra of (a) PVC-U compound (S0) (b) composite (S11) 135 3.46 FTIR spectra of (a) filled composite (S11) (b) Prosil 9234-treated filled composite (S51) (c) NZ 44-treated filled PVC-U composite (S71). 3.47 Impact strength of unweathered and weathered impact modified compounds 3.48 137 Impact strength of unweathered and weathered EFB filled 138 xxii composites 3.49 Suggested reaction scheme for the accelerated photo-oxidation degradation 3.50 155 Effects of EFB (S9-S12) and NPCC (S89-S91) on the flexural modulus of composites 3.65 154 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the impact strength of EFB-filled composites 3.64 153 Effects of EFB filler and NPCC on the impact strength of composites 3.63 152 FTIR spectra of unweathered and weathered-acrylic-impact modified composites 3.62 152 FTIR spectra of (a) unweathered and (b) weathered-acrylic impact modified composites 3.61 151 FTIR spectra of (a) unweathered and (b) weathered-EFB-filled composites 3.60 149 FTIR spectra of (a) unweathered and (b) weathered PVC-U compounds 3.59 148 Appearance of some samples surface before and after exposed to UV irradiation for 504 hours 3.58 148 Flexural strength of unweathered and weathered EFB-filled impact modified composites 3.57 147 Flexural modulus of unweathered and weathered EFB-filled impact modified composites 3.56 147 Flexural strength of unweathered and weathered EFB-filled composites 3.55 145 Flexural modulus of unweathered and weathered EFB-filled PVC-U composites 3.54 145 Flexural strength of unweathered and weathered impact modified compounds 3.53 143 Flexural modulus of unweathered and weathered impact compounds 3.52 141 Impact strength of unweathered and weathered EFB-filled impact modified PVC-U composites 3.51 140 Effects of EFB (S9-S12) and NPCC (S89-S91) on the flexural 156 xxiii modulus of composites 3.66 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the flexural modulus of the EFB-filled composites 3.67 161 Effect of unextracted and extracted EFB filler on the flexural strength of composites at filler content of 30 phr 3.71 160 Effect of unextracted and extracted EFB filler on the flexural modulus of composites at filler content of 30 phr 3.70 159 Effect of unextracted and extracted EFB filler on the impact strength of composites at filler content of 30 phr 3.69 158 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the flexural strength of the EFB-filled composites 3.68 157 162 Effects of impact modifier content on water absorption of acrylic-impact modified (S1-S4) and CPE-impact modified compounds (S5-S8) 3.72 Effect of EFB filler content on the water absorption of unmodified and modified composites 3.73 164 Effect of coupling agents on the water absorption of treated unmodified composites 3.74 163 165 Effect of coupling agents on the water absorption of treated acrylic-impact modified composites 166 4.1 (a) TG and (b) DTG curves for neat components 179 4.2 DSC trace of EFB filler 181 4.3 (a) TG and (b) DTG curves of unmodified filled PVC-U composites 184 4.4 Polar-polar interaction between H20 and PVC molecules 185 4.5 (a) TG and (b) DTG curves for EFB-filled CPE-impact modified PVC-U composites 187 4.6 (a) TG and (b) DTG curves for EFB filled acrylic-impact PVC-U 188 4.7 Polar interaction between MMA and HCl 189 5.1 Rigid PVC fusion mechanism in the Torque Rheometer 193 5.2 Soxhlet extraction apparatus 197 5.3 A typical temperature and torque curves of PVC-U compound (S0) blended in the Brabender Torque Rheometer 200 xxiv 5.4 A typical temperature and torque curves of EFB-filled PVC-U composite (S0) blended in the Brabender Torque Rheometer 5.5 201 Postulated lubrication mechanism of among PVC micro-particles flow units, EFB fillers and acrylic impact modifiers 5.6 217 Postulated lubrication mechanism of among PVC micro-particles flow units, EFB fillers and CPE impact modifiers 218 5.6 FTIR spectra: a) EFB filler b) extracted EFB filler 220 5.7 FTIR spectra (a) solvents mixture condensate (b) concentrated extract 221 5.8 Palm oil molecular structure 222 5.9 SEM Micrograph for EFB Filler (Magnification 3000x) 223 5.10 SEM micrograph for extracted EFB filler (Magnification 1000x) 6.1 223 o Extrudates of S0, S3, S7, S9 and S21 at die temperature of 185 C, and extrudate of S37 at die temperature of 190 oC 233 6.2 Shape difference between the die and extrudate 234 6.3 Filled extrudates of S37 at die temperature of 180 oC and 185 oC and at constant screw speed of 60 rpm 6.4 235 Filled extrudate of S9 with air bubbles on surface 236 o 6.5 Filled extrudates of S10, S22 and S38 at die temperature of 195 C 6.6 A typical DSC thermogram of S0 at different processing temperatures 6.7 6.8 237 239 A typical stress-strain graph of tensile for extrudates at 60 rpm and 180 oC 253 Predicted yield stress values at 3 m/s for S0 257 xxv LIST OF SYMBOLS b - width d - thickness CPE - chlorinated polyethylene EFB - oil palm empty fruit bunch Ec - flexural modulus of composite Ep - flexural modulus of polymer Gf - shear modulus of filler Gp - shear modulus of polymer HDT - heat deflection temperature k - Boltzmann’s constant L - span Mt - water absorption at time t NPCC - nano-precipitated calcium carbonate NZ 44 - zirconate based coupling agent Prosil 9234 - silane based coupling agent PVC-U - unplasticised poly (vinyl chloride) S - sample Tg - glass transition temperature TG - thermogravimetric curve DTG - differential thermogravimetric curve v - velocity Vf - volume of filler Vp - volume pyknometer Vw - volume of water W - failure load xxvi Wadditives - weight fraction of additives Wd - dry weight Wf - weight fraction of filler Wpvc - weight fraction of poly (vinyl chloride) Ww - weight after immersed into water ∆H - activation energy ∆S - deflection increment ∆W - load increment ε - strain rate εc - elongation at break of composite εp - elongation at break of polymer φf - volume fraction of filler φm - maximum packing fraction γ - stress concentration factor ν - activation volume νp - Poisson’s ratio of polymer ρadditives - density of additives ρc - density of composite ρf - density of filler ρw - density of water σc - strength of composite σp - strength of polymer σy - yield stress ψ - reduced concentration term CHAPTER 1 INTRODUCTION 1.1 Overview Poly (vinyl chloride) has grown into one of the major thermoplastics material since it was first produced in the 1930’s. More than one-half of the PVC produced is used in the building industry (Mengeloglu, 2001). Because of its blending versatility (ability to be modified through the introduction of various additives) and costeffectiveness, PVC is one of the most worldwide used thermoplastics and it plays an important role in the plastic industry as a major component of composite manufacturing (Wenguang and La Mantia, 1996). Cost reduction is achieved when fillers are incorporated into the PVC matrix. In addition to reducing cost, other typical reasons for using fillers are stiffness enhancement and thermal properties alteration of vinyl products. Currently, the most important fillers for PVC are inorganic materials such as glass fibre, calcium carbonate, agalmatolite, talc (Saad et al., 1999; Wee, 2001; Pedro et al., 2001; Xie et al., 2001). These fillers are high-density materials (Mengeloglu, 2001). While they offer wide property changes in the composites, their use is not cost effective on a volumetric basis (Titow and Lanham, 1975). In addition, these inorganic fillers cause 2 wear of equipment during processing. Therefore the use of natural fillers such as wood has been suggested as an alternative to mineral materials (Nabi Saheb and Jog, 1999). Interest in the use of natural fillers has grown during the last decade and most of the natural fibres used in thermoplastics composites are derived from wood. The use of natural fillers in USA plastic industry in 2000 was estimated 0.4 billion pounds or 7 % (Gurram et al., 2002). In fact, industry experts believe that the demand for natural fillers for plastics composite applications will grow at least six-fold in the next 5 to 7 years (Stokke, 2002). The rationale behind this interest is that the use of natural fibres offers several benefits, including lower cost, less abrasiveness to equipment, high specific properties such as higher specific modulus (modulus/specific gravity), renewable nature and biodegradability (Nabi Saheb and Jog, 1999). Even though natural fibres offer several advantages, certain drawbacks such as incompatibility of fibre with the thermoplastics matrix, the tendency to form aggregates, and poor resistance to moisture greatly reduce the potential of these fibres to be used as reinforcement. All these drawbacks, especially incompatibility, lead to low fibre-matrix interfacial bonding strength and poor wetting of the fibres by the matrix resin, which can cause reduction in mechanical performance of the composites. Therefore, surface modification by the use of coupling agents is commonly suggested as solution to overcome the drawbacks of lowered strength property. These chemical reagents convert the hydrophilic surface of cellulosic fibres to hydrophobic one. As a result, the fibre-fibre hydrogen bonds is reduced and the adhesion is also improved via the mechanism such as formation of chemical bonds between treated fibres and polymer matrix (Xanthos, 1983; Raj et al., 1989, 1989, 1990; Maldas and Kokta, 1990,1990,1991; Myres et al., 1991; Maiti and Subbarao, 1991; Ahmad Fuad et al., 1994,1997; Sivaneswaran, 2002; Baiardo et al., 2002; Fabbri et al., 2004). One of the natural fillers that have been tried was oil palm empty fruit bunch (EFB). EFB is one of the natural fillers obtained from oil palm. The incorporation of 3 EFB as a reinforcing component in the polymer composites has received much attention (Sareekala et al., 1996/97; Rozman et al., 1998, 2001, 2001; Sarani et al., 2001; Abdul Khalil et al., 2001). Most of the research works on EFB so far dealt with polymers such as polypropylene, polyester, polystyrene and phenol-formaldehyde resin. The researchers focused on the effect of filler loading, filler particle size distributions and filler treatment on the mechanical properties of the composites. Like other natural fillers, many issues have been identified in the processing of the composites. These include low compatibility between EFB fibre and polymer matrix, higher moisture intake, poor distribution of the fibre in the matrix and effect of oil residue on mechanical properties. Although EFB has been useful as fillers in various polymers mentioned above, the use of EFB filler in PVC has not been studied extensively. Most of the researchers have used wood flour, wood sawdust and rice husk ash as fillers in PVC matrix (Matuana et al., 1998, 2001; Sivaneswaran, 2002; Sombatsompop et al., 2003). The use of these fibres in the PVC matrix produces products that are more brittle than neat PVC. The incorporation of the natural fibres alters the ductile mode of failure of the matrix. The brittleness of PVC/wood fibre composite compared to the unfilled PVC may prevent this emerging class of materials from capturing their full market potential in applications such as door and window frames (Mengelollu, 2001). One of the most important aspects in the development of PVC compound by using fillers is to achieve a good combination of properties and processability at a moderate cost. As far as mechanical properties are concerned, the main target is to strike a balance of stiffness, strength and toughness (Mohd Ishak et al., 2000). The incorporation of filler in thermoplastic such as PVC is associated with an enhancement in stiffness and strength but a decrease in toughness. The incorporation of impact modifiers in PVC, on the other hand will result in a significant improvement in toughness at the expense of stiffness and strength. 4 The enhancement of impact properties of PVC-U by introduction of a rubbery phase is a well-known phenomenon which has been widely exploited commercially. Among those which have extensively used are chlorinated polyethylene (CPE) and various modified acrylic. Hiltner et al. (1984) reported that CPE modifier formed a network enclosing PVC primary particles in order to develop the desired impact strength whereas Robinovic (1983) found that acrylic modifier formed a distinct dispersed phase in the continuous PVC phase. The dispersed rubber particles of acrylic were independent of the processing time and temperature, whereas the CPE network structure was dependent of those factors. The studies have concluded that the function of the impact modifiers (whether in the form of discrete phase or network structure) was to increase the toughness of PVC by facilitating the ability of PVC matrix to shear yield and/or craze. This means that the PVC would deform rather than fracture when subjected to a sudden load (Petrich, 1973; Lutz, 1993). Meanwhile, Stevenson (1995) has pointed out that the impact modifiers can also be used to correct the detrimental effects of other additives such as fillers by boosting the impact performance of the material. 1.2 Significance of Study Based on the previous findings mentioned above, no study has been reported on the use of EFB fibre in the PVC matrix. Most studies involved investigation of the mechanical properties of natural fibre polymer composites in terms of natural fibre content, various treatments of the fibre, whereas very few have investigated processability and thermal stability of the composites. According to Sombatsompop et al. (2003), the area for PVC/wood fibre composites are still actively pursued, especially in areas of processability and thermal stability-related mechanical properties of the composites. 5 With the rising environmental issues in the limelight and the need for new applications of agricultural by-products such as EFB to be more useful in order to minimize industrial waste, therefore agricultural by-products seem attractive enough to be use as reinforcement filler alternative for filled thermoplastics composites. Since oil palm has now become a major cash crop and is cultivated commercially in Malaysia, many million tons of EFB on dry weight basis are produced annually throughout the world. EFB is an agricultural waste, which is left unutilized after the removal of the oil seeds for oil extraction. In order to minimize the abundance of this waste and environmental problems, therefore new applications are urgently required for EFB to be more useful (Sreekala et al., 1996/97). The utilization of this EFB as a reinforcement in polymers has some economical as well as ecological importance. PVC is basically a rigid thermoplastic. It is inherently ductile because of the polarity and low-temperature molecular relaxation. However, it is brittle at high deformation rate and under concentrated stress, and PVC is known as a notchsensitive polymer (Wimolmala et al., 2001). Previous studies have also shown that the impact strength reduces drastically with addition of natural fillers (Nabi Saheb and Jog, 1999; Mengeloglu, 2001). Impact modifier is therefore added to increase the ductility and toughness of PVC. Although ductility increases, the impact modifier used is costly and tends to decrease the PVC-U stiffness. In this study, the strategies to use EFB fillers together with impact modifier in PVC-U are deliberately done to develop PVC-U composites with good stiffness without sacrificing both desirable toughness and cost effectiveness. Therefore, in order to understand the properties of the newly developed of PVC-U composites, the effects of filler and impact modifier contents, and filler treatment on the mechanical and thermal properties, processability and extrusion process on the PVC-U composites are investigated. 6 1.3 Objectives of Study The main objective of the study is to produce EFB-filled PVC-U composites. To achieve the objectives, the following stages have been carried out: (a) To determine the effect of EFB filler, acrylic and CPE content on the impact strength and flexural properties of the compression moulded composites. The influence of EFB filler, acrylic and CPE content on the thermal properties such as heat deflection temperature, thermal degradation temperature and glass transition temperature are also investigated. In addition, composites accelerated weathering and water absorption properties rare also determined. (b) To study the effect of silane (Prosil 9234) and zirconate (NZ 44) coupling agents on the impact and flexural properties and water absorption of the compression moulded composites. From this study, the most suitable coupling agent is determined. (c) To determine the effect of EFB filler, acrylic and CPE content on the fusion characteristics of PVC-U compounds. The effect of different processing temperatures on the composites is also studied. (d) To determine the effect of processing temperatures and screw speeds on the impact, tensile and flexural properties of extrudates. The influence of processing temperatures on fusion level of extrudates is also determined. 7 1.4 Outline of Thesis Chapter 1 presents the overview of the previous research works that has been conducted by other researchers who used either EFB fibres as reinforcement or PVC as matrix in their studies. Both the advantages and disadvantages of natural fibres are also reported as well as the role of impact modifier in the PVC. The significance of the study, the objectives of study and the layout of research are presented. Chapter 2 begins with some definitions and general feature properties of PVC. Several theories on PVC such as impact modification and fusion mechanism, and the function of additives are described. The definitions and categories of composites and the properties of EFB filler are also described. Literature survey on PVC filled composites is also reported, followed by reviews on the application of various coupling agents and the application of EFB as a reinforcement filler or fibre in the thermoplastics composites systems. Several theories that have been used to predict the relative modulus of filled composites are also reviewed. Chapter 3 mainly describes the effect of filler content, impact modifier content and coupling agents on the impact strength, flexural properties and water absorption of composites. The effect of filler content on the density of composites, the application of models to predict the relative modulus, and the impact strength and flexural properties of weathered composites are also described. The morphological studies on the impact-fractured surface and the presence of functional groups on the samples investigated using Scanning Electron Microscope (SEM) and Fourier Transform Infra-Red (FTIR), respectively, are also reported. Chapter 4 reports the thermal properties of the composites. A wide array of thermal analysis techniques such as thermogravimetric analyser (TGA), heat deflection temperature analyser (HDT) and differential scanning calorimetry (DSC) are used to 8 investigate the thermal degradation temperature, heat deflection temperature and glass transition temperature, respectively. Chapter 5 reports on the processability of PVC-U compounds. The effects of EFB filler content, extracted EFB filler, CPE and acrylic impact modifier content and mixing temperature are reported. The effect of extracted EFB filler content on the processability is also investigated as a minor part of this study. FTIR is used to identify the related functional groups present in the extractives and the changes in EFB filler surface, followed by morphological study on the extracted EFB filler surface using SEM. Chapter 6 reports the effect of screw speed and processing temperature of extrusion on the impact strength, flexural properties, tensile properties and fusion level of the composites. The effect of crosshead speed on the tensile properties is studied to predict the yield stress of extrudates. The fusion level of extrudates at different temperatures is also investigated using DSC. Chapter 7 presents the overall conclusions and suggestions for further work. CHAPTER 2 LITERATURE REVIEW 2.1 Poly (vinyl chloride) Poly (vinyl chloride) is the second largest of thermoplastics (polypropylene being the first on the list) being used as raw material in the plastic industry. The world market reported that the growth rate of the world production and the growth rate of the world consumption of PVC thermoplastic from 1999 to 2005 would be 3.30 % and 3.83 %, respectively (Murphy, 2001). It shows that although poly (vinyl chloride) has been dubbed as a toxic material but its excellent properties and versatility cannot be denied. 2.1.1 Definitions and General Feature Properties Poly (vinyl chloride), commonly known as PVC, is a predominantly amorphous polymer with a basic repeat unit [-CH2-CHCl-]. Based on its chemical structure, PVC is structurally very similar to polyethylene (PE), with a chlorine atom 10 substituted for a hydrogen atom on alternating carbons in the chain. The number of chlorine molecules attached to these components governs the overall characteristics of PVC polymer or compound to a substantial extent (Penn, 1971). The chlorine atom is many times larger than the hydrogen atom and greatly reduces the ability of the molecules to rotate and flex. The base polymer is therefore quite stiff and has low creep properties (long term deflection under stress) at room temperature. Being highly electronegative, chlorine atoms create polarity and are electrically repulsive to each other. This feature results in further immobility, inflexibility and low level of crystallinity. Therefore, the impact resistance of the basic vinyl chloride is quite poor but is relatively resistant to organic solvents and flame. The polarity of PVC is responsible for its remarkable ability to form compatible compositions with many other polymers and plasticizers as well as other formulating ingredients (Coaker and Wypart, 1993). These advantageous properties offer rigid PVC the opportunity to be used as base polymer matrix for the reinforced version. This can be explained by their characteristics as listed in Table 2.1. Its main limitations are its low softening temperature (82 oC) and glass transition (80-84 oC) and its tendency to degrade at elevated temperature. Under heat or ultraviolet light, the chlorine atoms become unstable and break out from the rest of the molecules. This heat sensitivity makes the base polymer extremely difficult to be processed by conventional melt processes. However, due to its affinity for additives, this undesirable property of PVC could be overcome by the development of suitable stabilizer in the early 1930’s (Butters, 1982). Despite its poor thermal stability, when correctly formulated, PVC is still widely used since it is relatively cheap, inherently flame retardant, strong and stiff, and good chemical and weathering resistance (Whelan and Goff, 1989). 11 Table 2.1: Some physical properties of moulding PVC products and compounds (Titow, 1991) Typical values or values range for Rigid PVC 1.30-1.45 Properties Density (gcm-3) Tensile strength (MPa) 31-60 Elongation at break (%) 2-40 Flexural strength (MPa) 62-100 Young’s modulus (GPa) a. in tension, E 2.5-3.5 b. in shear, G 1.0-1.8 c. in flexural, EB 2.0-3.5 d. in compression, E 2.2-3.5 Thermal conductivity (Wm-1K-1) 0.14-0.28 2.1.2 PVC Polymerisation Commercial grade PVC is produced by two main routes as shown in Figure 2.1. The vinyl chloride monomer used in production of PVC is produced either by direct hydrochlorination of acetylene or by chlorination of ethylene. Process 1. CH ≡ CH Acetylene + HCl → CH2=CHCl Hydrochloric acid Vinyl chloride monomer 12 Process 2 CH2=CH2 + Ethylene CH2Cl-CH2Cl 1,2-dichloroethane Figure 2.1 Cl2 Chlorine → CH2Cl-CH2Cl 1,2-dichloroethane ⎯⎯ ⎯ ⎯→ pyrolisis CH2=CHCl Vinyl chloride monomer + HCl Hydrochloric acid The chemistry of vinyl chloride monomer production PVC resins are normally produced by polymerisation procedure (Penn, 1971). Three of these essential polymerisation procedures are, in terms of commercial importance: (1) suspension polymerisation (about 80 % of the commercial production) (2) emulsion (about 10-15 %) and (3) bulk or mass and solution polymerisation (about 10 %). The majority of PVC is produced via the suspension technique due to the fact that PVC produced by this technique forms the basic compound intended for extrusion or moulding. For suspension resins, vinyl chlorides are dispersed in the aqueous phase to form a dispersion of droplets. The polymerisation takes place within the droplet until the required conversion of monomer to polymer is achieved. The PVC resins normally have a compact, relatively non-porous closed structure. In the emulsion polymerisation, the vinyl monomer is emulsified in water, which contains water-soluble catalyst and is characterized by high polymerisation rates leading to high-molecular weight resins. This type of PVC is normally less expensive than other forms of PVC, but requires higher amount of lubricants to assure the glossy surface of the finished commodity products. Bulk polymerization resins are snowball-like particles having very porous structure with a large surface area. Because of this, bulk PVC resins are much easier to fuse into homogeneous melt. 13 The suspension and mass polymerisation processes both produce a particulate with a mean particle size in the range 100-150 µm. Meanwhile, emulsion polymerisation produces a product in the latex form with very small and narrow particle size distribution of the order 0.1 to 2 µm. However, all PVC resins have a series of particulate-sub structures. Commercial PVC materials are often classified according to the polymerisation method and the molecular weight. The molecular weight of PVC is often quoted in terms of a K-value instead of the actual molecular weight. It is the measurement of the viscosity of dilute solution. The higher molecular weight in general is indicated by the increase difficulty in processing. In commercial resins, the number average molecular weight lies in the range of 50,000 to 100,000 (Sarvetnick, 1969). 2.1.3 Morphology of PVC The PVC resin particle is unusual in structure and to understand the properties and processing of this unique polymer requires an understanding of its structure or morphology. Suspension PVC resin particles usually posses a pericellular “skin” or “membrane” which extends almost continuously over the entire outer surface of the grain. This skin is formed during the early stages of polymerisation. Suspension-polymerised PVC in a powder form possesses various level of morphology. Powder particles, which are visible to the naked eyes, are known as grains. Each grain is an assembly of primary particles, which are loosely packed 14 together. Each primary particle appears to be made up of smaller structure called domains. These are illustrated in Figure 2.2 (Hattori et al., 1972). Domain 10-30 nm Colloidal Skin Primary Particle 0.2-1.5 µm Grain 65-150 µm Figure 2.2 PVC particle morphology (Gilbert, 1985) There are arguments among researchers in defining the terminology to describe PVC particle morphology. Thus, to avoid confusion and to maintain consistency of the PVC terminology, a table with a standard terminology was summarized and published by Geil (1977). This standard PVC nomenclature is listed in Table 2.2 and it will be used for further discussion on the morphology. Although PVC is an amorphous polymer, it has been reported that it contains about 5-10 % of crystallinity (Soni et al., 1982 and Gilbert, 1985). Based on small angle X-ray scattering (SAXS) data, the crystalline structure has been proved by the existence of crystalline noduli with the size of 2.7 nm and d-spacing of 10 nm for virgin PVC. This crystalline nucleus which was surrounded by constrained amorphous has been associated with the basic micro-domain structure of PVC. 15 Table 2.2: PVC nomenclatures (Geil, 1977) Term Grain Range (µm) 50-250 Average (µm) 130 Description Visible constituents of free flowing powders, made up of more than 1 monomer droplet Sub-grain Agglomerate 10-150 40 Polymerised monomer droplet 2-10 5 Formed during early stage of polymerisation by coalescence of primary particles (1-2 µm). Grows with conversion to size shown. Primary particle 0.6-0.8 0.7 Grows from domain. Formed at low conversion (less than 2 %) by coalescence of micro-domain: grows with conversion to size shown. Domain 0.1-0.2 0.2 Nucleus of primary particle. Only observed at low conversion (less than 2 %) or after mechanical working. It becomes primary particle as soon as growth starts. Micro-domain 0.01-0.02 ~0.02 Smallest species so far identified. Aggregate of polymer chains. Summers (1981) studied the nature of crystallinity of PVC and found that the crystallites in the micro-domain structures were connected by tie molecules with a • 100 A spacing. These crystallites acted as crosslinks in a three-dimensional network which prevented primary particles interaction at low temperature. However, at high temperature (177 oC) some of the crystallite melted and recrystallization after cooling led to strong ties between the primary particles. It was also revealed that PVC contained two types of crystallinity: primary crystallinity, which was formed during polymerisation and secondary crystallinity formed during recrystallization (Lacatus and Rogers, 1986). 16 2.1.4 PVC Fusion PVC must undergo fusion in every fabrication process in order to achieve its best mechanical properties. It is generally agreed by researchers that the level of fusion has significant influence on the mechanical, physical and chemical properties of the PVC. To obtain optimum properties an appropriate level of fusion is needed (Covas et al., 1988) The fusion process for PVC is the subject of great deal of discussion, and there are many views about the exact courses of this process. Although there are several ideas suggested by the researchers, but one conclusion to describe the fusion process of PVC can be made, namely the fusion process is the breakdown process of the PVC grains and is affected by a combination of such external parameters such as temperature, pressure and shear. The methods used to break down the PVC grains, whether via Brabender torque rheometer, two roll mill, single screw extruder or twin screw extruder, results in the different breakdown mechanisms (Benjamin,1980; Parey and Manges, 1981; Portingell, 1982). Benjamin (1980) defined that the fusion of PVC as the process of removing the boundaries between base particles whereby a continuous mass that resulted have a homogeneous molecular structure. The grains have broken down to the so-called primary particles. Further broken down of the primary particles to the sub-primary particles took place if sufficient energy was applied. The structure of these subprimary particles was said to be a continuous network of molecular chains, which to some extent, was bound together by a crystalline nucleus. The sub-primary particles contain approximately 10 % crystalline material. Parey and Manges (1981) concluded that the fusion of PVC proceeds in two stages. In the first stage, the PVC dry-blend was changed into a material capable of fusing under the influence of temperature and particularly of shear. In the second 17 stage, mainly through the effect of temperature, this material becomes fused. The shear here, only served to homogenize the material. Meanwhile, Portingell (1982) has defined fusion process is a destruction process of grains and primary particles and the establishment of an entanglement network locked by a network of crystallites. This destruction process is accomplished by the combined action of heat, pressure and shear during processing. The extent of destruction of particulate of PVC is known as the degree of fusion. 2.1.4.1 Mechanisms of Fusion There are two mechanisms of PVC fusion as suggested by Allsopp (1982), and these are illustrated in Figure 2.3. If the condition such that shear is applied before densification is complete, such as Banbury mixer and Brabender mixing, grain comminution results, but if grain densification is complete before shear force is applied, such as extrusion, two roll milling, and compression moulding, then compaction-densification-fusion-elongation (CDFE) mechanism results. The route CDFE mechanism occurs at less shear-intensive processing conditions. The mechanism has been proposed by Allsopp (1982). The route of CDFE is illustrated in Figure 2.3. As the material progresses in the screw, the grains are compacted together and simultaneously eliminate the internal porosity of the grains. Further along the screw the PVC grains start to show a high level of densification, with the additives deposited in pockets around the grain surface. Midway along the screw, grain densification is complete, with almost complete fusion of the grains, but the additives still remain concentrated at the grain surface. Between the end of the screw and the form side of the breaker plate, the grains become highly elongated and homogenisation of PVC and additives becomes very good. The mechanism is generally similar with the works reported by other researchers which confirmed that in extrusion, the main route for fusion was through the deformation of grains by low shear (Gilbert et al., 1983; Covas et al., 1988). They also suggested the overall view of fusion by postulating a melting 18 and recrystallizing process leading to network formation at the latter stages of the process. Figure 2.3 Mechanism of fusion during processing extruder (Allsopp, 1982) The route of fusion by comminution mechanism occurs in the shear intensive processing conditions. Allsopp (1982) suggested that a PVC fusion mechanism for a torque rheometer as shown in Figure 2.3 where the original PVC grains underwent comminution and formation of agglomerates, which consisted of bound fragments of comminuted grains. When densification of primary particles occurred, the additives were well dispersed throughout the matrix. With further densification, the primary particles structure could still be observed and finally the boundaries between primary particles disappeared, the particles lost their identity, and a three-dimensional entanglement network of polymer chains was formed and the additives were uniformly distributed. There were crystalline regions inside the primary particles, which were connected together by molecules forming a network structure (Krzewki and Collins, 1981). Similar fusion mechanism was also described by Benjamin (1980) who divided the mechanism into two distinct phases as shown in Figure 2.4. 19 Phase 1: The PVC grain broke down into primary particles. The molecular network remained within the boundaries of the particles although the primary particles underwent the compaction. Phase 2: With the continuation of the process, further breakdown of the primary particles occurred and followed by the formation of a continuous molecular network of entanglement. Figure 2.4 Schematic presentation of fusion process for PVC (Benjamin, 1980) 2.1.4.2 Fusion Level Assessment: Differential Scanning Calorimetry Assessment of the state of fusion is of utmost importance in order to achieve satisfactory product properties. Methods presently employed include acetone immersion, zero length die extrusion, and calorimetry (Butters, 1982). Recent 20 investigations (Gilbert and Vyvoda, 1981; Teh et al., 1989) indicated that thermal analysis being the simplest and most effective quantitative method for obtaining a considerable amount of information about the fusion of processed PVC compounds. Differential Scanning Calorimetry (DSC) approach to quantify fusion as defined in the previous studies involves detecting two endotherm peaks, A and B as shown in Figure 2.5 (Gilbert and Vyvoda, 1981; Teh et al., 1989). The first an endothermic baseline shift before two endotherms, corresponded to the glass transition temperature (Tg) of the rigid PVC. Gilbert (1985) reported that Tg of PVC was largely unchanged (except for plasticized compounds) with the heat treatment or processing but not for the two endotherm peaks. ENDO ∆T EXO Moulding Temperature Temperature oC Figure 2.5 DuPont thermal analysis of PVC-U compound compression moulded at different temperatures Peak B represents the melting of crystallites, which decreased in size and shifted to higher temperature as the processing temperature increased. The size decrease was attributed to the melting of less perfect or smaller crystallites, whilst the 21 temperature shift was caused by annealing of unmelted crystallites. Other additional useful information is available from position of peak B where the B onset temperature is equal to the maximum processing temperature reached by the polymer sample. Peak A increased in size as the processing temperature increased. The area of A, however, is not yet fully understood but is probably be related to the degree of fusion. The nature of endotherm of A probably corresponds to the formation of imperfect ordered regions which is produced as a continuous PVC network develops, and which subsequently melt upon reheating. The heat of fusion of endotherms was used by researchers to analyse the degree of fusion of PVC compounds. The heat of fusion of the first endotherm, A has been employed by Gilbert and Vyvoda (1981) to asses the degree of fusion but Teh et al. (1989), reported that a more convenient normalized form where the ratio of the heats of fusion of endotherm A to the sum of the endotherms A and B was used. The similar method was also used by Cora et al. (1999) to determine the relationship between impact strength and morphology in PVC window profiles with fusion level. 2.1.4.3 Interrelation between Fusion and Properties Benjamin (1980) extruded a series of pipes under controlled conditions to provide four fusion levels. Tensile results (stress at yield) suggested a constant maximum value at a higher fusion level of 68 %, while ductility tests, such as tensile impact, tensile strain at break, resistance to crack formation, produced an optimum value at approximately 44-68 % fusion level. Benjamin concluded that a fusion level of approximately 60-70 % would provide the best balance of properties for particular pipe and composition chosen. This corresponds to some primary particle still remaining within the network morphology. 22 Covas et al. (1988, 1992) provide a correlation between processing, structure and properties. This processing study considered both a single and a twin screw extruders, and a mechanism of fusion was conjectured based upon the Allsopp (1982). Tensile properties were evaluated by producing families of stress-strain curves with different fusion level at different test rates and test temperatures. Yield properties were, again, seen to be unaffected by fusion level and were attributed to molecular relaxation mechanism, however elongation at break and stress at break were dependent upon fusion level. A maximum value of impact strength occurred between fusion levels 40-65 % as observed by Benjamin (1980). This fusion range corresponds to a processing temperature of 193-203 oC and the optimum network morphology with possible residual primary particles. Similar reason was related to the impact performance, where the maximum was achieved for the maximum levels in the range 69-70 %, which was also reported by other workers (Marshall et al., 1983; Gilbert, 1985). The reason for the decrease in impact strength after reaching the maximum value was not fully understood, but it seemed to occur when primary particles have disappeared to produce a reasonably homogeneous matrix. Meanwhile the maximum value was possibly due to the presence of a new network structure and some primary particles were beneficial in providing a mechanism for stress distribution. Uitenham and Geil (1981) and Terselius and Jansson (1984) have studied the relationship between tensile properties of the extrudates and level of fusion for rigid PVC and found that the yield parameters were not sensitive to variation in the degree of fusion. Information on this relationship has shown that yield stress at low strain rate was independent of the relative coherence of the entanglement network and number of crystalline junctions, and it was probably related to molecular relaxation processes. Terselius et al. (1981,1984,1985) also, published a number of papers, which discussed the effect of fusion levels on pipe properties. The effect of mass 23 temperature of 176 oC-205 oC at extrusion on the impact strength of specimens from pipe was studied. Falling weight tests have clearly demonstrated the existence of a maximum in impact strength at a moderate mass temperature (190 oC) of extruded pipes (unnotched). It was explained by increased post-yield deformation, presumably due to voiding and local yielding in the interparticular weak regions of specimens of reduced entanglement density. Charpy test on notched specimens showed a marked increase of impact strength at moderate mass temperature (fusion level). This was attributed to improved strength and coherence of the entanglement network, and thus increased the resistance to crack propagation. Summers et al. (1982) who studied the dropped dart impact strength of extruded specimens found that toughness went through a maximum then decreased with increased melt temperature and increased degree of fusion. This is similar to the observations made by Benjamin (1980) and Terselius et al. (1985). In contrary with the previous findings, Summers et al. (1982) explained that the loss of impact strength at high extrusion temperatures was due to lubricant failure and melt fracture (as discussed in lubricant section) Cora et al. (1999) provide a recent correlation between impact strength and morphology in poly (vinyl chloride) window profiles with fusion level. In this study, the toughness of extruded PVC window profiles not only depended on intrinsic variables such as type and level of impact modifier, filler and lubricants but for a given impact test (i.e. given stress and strain rate), toughness was also greatly affected by the free volume within the matrix, an extrinsic variable. Free volume for a given formulation is determined by the degree of fusion the matrix (crystallinity). As melt temperature increases the three-dimensional crystalline network fuses, leading to glassy state with lower free volume. This increased the stress at yield and subsequently reduced toughness of the materials. This study contradicted with the results obtained by other researchers (Uitenham and Geil, 1981; Terselius et al., 1984, 1985; Covas et al., 1988, 1992). 24 2.2 PVC and Additives Even though PVC polymer resins have been reported to possess many attractive advantages with vast commercial demands, some imperfections in virgin PVC such as susceptibility to heat degradation, relatively high melt viscosity, comparably low heat deflection temperature, limited strength and stiffness. To counteract these imperfections and to extend its range of applications some modifications are required within original PVC materials such as adding small amounts of additives such as heat or thermal stabilizer, lubricants, processing aids, pigment, impact modifier, and so forth. 2.2.1 Heat Stabilizer PVC is thermally unstable and degrades rapidly at normal processing temperatures unless stabilized. Thermal stability of PVC has been reported in detail by Wypych (1985). It has been well documented that PVC undergoes dehydrochlorination at its normal processing temperatures (180-200 oC) or upon long term exposure to ultraviolet light. As shown in Figure 2.6, the dehydrochlorination reaction proceeds like a chain reaction, producing long sections of conjugated polyenes. The process of heat degradation liberates hydrogen chloride (HCl), which has an autocatalytic effect on the degradation. Many stabilizers act primarily by reaction with hydrogen chloride to inhibit the hydrogen chloride from accelerating the process of thermal degradation. Therefore, the heat or thermal stabiliser such as metal carboxylates and organotin compounds added to the PVC compound, usually posses two major roles: scavenging HCl and reacting with the polymer to prevent HCl loss and simultaneous discoloration. 25 HC CH CH2 CH Cl CH CH2 Cl CH Cl -H C l HC CH H C C H CH CH2 Cl CH Cl -H C l HC CH C H H C C H CH CH Cl Figure 2.6 Zip-elimination of HCl from PVC and simultaneous formation of sequences containing conjugated double bonds (polyenes) in the chain during degradation (Wypych, 1985) The effectiveness of stabilisation depends on the capability of the stabiliser to mitigate or alter the degradation and to disrupt the chain reactions already started. The principal phenomena symptomatic involved in the thermal degradation of PVC those react with stabiliser are listed below (Wypych, 1985): • Dehydrochlorination of the chains through the elimination of HCl. The presence of HCl is to promote further dehydrochlorination, and consequent degradation, sometimes referred to as an ‘unzipping mechanism.’ • Formation of double bonds, which are conjugated in the polyene sequences. They are responsible for the development of colour. The polyenes are active in secondary reactions leading to crosslinking and the formation of aromatic pyrolysates. 26 • The presence of tertiary chlorine atoms (chlorine on branch carbons) and end groups in the PVC chain. The tertiary chlorine atoms are particularly heatlabile and they are responsible for the initiation of evolution of HCl and the formation of double bonds. Decomposition can also be initiated at the end groups. 2.2.2 Lubricants According to Marshall et al. (1969), lubricant, in general, can be defined as a chemical compound necessary for minimisation of shearing forces during processing. During processing, lubricants firstly coat the PVC granules and then, with continued heating soften the PVC as they become soluble in the polymer. The granules are forced together by processing and coalesce to give a state of fusion, such that, the more soluble the lubricant, the more readily fusion takes place. In addition the metal/polymer adhesion is reduced and surface slip increases. For that reason, processing lubricants are widely used to assist the passage of the polymer melts through the processing machinery. Depending on the lubricating action and effects, lubricants are usually classified into two categories, internal and external lubricants. Internal lubricant is highly compatible with PVC to promote a breakdown of the PVC resin granules to the primary particles at processing temperature of about 160 oC and influences flow through reduction in the bulk viscosity of the melt or reduction of the chain-chain intermolecular forces. The fatty esters, metal stearates (calcium stearate) or alcohol are typical types of this lubricant (Bower, 1986). Meanwhile, external lubricant is incompatible or less compatible (compared to internal lubricant) at all temperatures with the PVC, and therefore during processing it migrates to the melt-metal interface, promoting some effective slippage of the melt by reducing interfacial layer viscosity. It also prevents the sticking of PVC to the processing equipment and improves the 27 surface finish of the product. The examples of this type of lubricant are metal soaps, amide waxes and hydrocarbon waxes. Mostly, external lubricants delay the fusion of the PVC compound, and internal lubricants speed up the fusion time or have very little effect on it. In actual applications, however, lubrications do not always abide by the rules (Chen and Lo, 1999). The effect of lubricants on the fusion process or processability of PVC has been studied extensively and particular attention has been given to the combination of calcium stearate with a paraffin wax. Based on this combination, Rabinovitch et al. (1984) has proposed a model of PVC lubrication of rigid PVC, as shown in Figure 2.7. As suggested by Figure 2.7, fusion between the primary particles of PVC without lubricants took place when three-dimensional crystallite networks within the primary particles started melting, thus releasing polymer chain ends and loops for partial interdiffusion across the original primary particle boundaries. Because of the polarity, PVC attracted and stuck to the polar metal surface of processing equipment. When calcium stearate was added to PVC (Figure 2.7b), the polar ends of its molecules (CaCOOC-group) adhered not only to the polar PVC but also to the polar metal surface, leaving the non polar (CH2)16CH3 to do the lubricating to displace PVC melt from the metal surface, preventing it from sticking and providing limited slip between the metal and PVC melt surfaces. When paraffin wax was added to PVC (Figure 2.7c), it did not wet the metal surface because it was a non-polar hydrocarbon, and therefore did not displace the more polar PVC. Thus, in the presence of only paraffin wax, PVC stuck to the metal surface. 28 When calcium stearate and paraffin were used together (Figure 2.7d), interaction between the non-polar adjacent calcium stearate molecules is hindered by the intervening presence of mobile non-polar alkane paraffin wax molecules which were not attracted to the polar groups of calcium stearate, nor to PVC, nor to the metal surface, but were weakly attracted to themselves and the non-polar hydrocarbons of calcium sterate molecules. This mechanism led to the formation of a mobile layer, which increased the separation between the thin calcium stearate films on the metal surface, and on the PVC melt surface. It also provided a more mobile interface between the calcium sterate molecules adhered to adjacent PVC primary particles. Based on this mechanism and synergistic effect, it was concluded that the paraffin wax delayed the fusion, while calcium stearate mildly promoted fusion. The similar effect has also been found by Bower (1986). Summers et al. (1982), have reported that at very high processing temperatures (above 200 oC) where the PVC was substantially fused, the calcium sterate/paraffin wax failed to perform well. Melt fracture took place, leading to a brittle product. It is believed that, above 200 oC, the crystalline structure almost disappeared and PVC became a continuous phase with the insoluble calcium stearate/paraffin wax lubricants coalescing to form droplets rather than surrounding individual primary particles flow units. The droplets of calcium stearate/paraffin wax lubricant initiated melt fracture mostly at the surface, which led to surface roughness and brittleness. 29 Figure 2.7 A model of PVC lubrication (not to scale) showing metal lubrication and lubrication between PVC primary particle flow units for (a) no lubricant (b) calcium stearate (c) for paraffin and (d) calcium stearate and paraffin (Rabinovitch et al., 1984) According to Krzewki and Collins (1981), calcium stearate has shown to facilitate or delay the fusion process in the compound. Increased calcium sterate levels in compounds with wax enhanced the fusion process whereas increased calcium stearate levels without wax decreased fusion process at relatively low temperatures but accelerate fusion at higher temperatures. With increasing 30 temperature, calcium stearate began to play a main role due to its increased solubility in the polymer matrix and was capable to facilitate interdifussion of polymer chains across the microparticle and submicroparticle boundaries to increase the rate of fusion. They also found that the paraffinic wax delayed the resin particle breakdown. Currently, Chen et al. (1995, 1999) investigated the fusion characteristics and morphology of rigid PVC compounds using Haake torque rheometer. They used oxidized polyethylene (OPE), calcium stearate (CaSt) and paraffin wax as lubricants and CPE as impact modifier in the PVC compounds. Similar to Rabinovitch’s theory of lubrication, the postulated mechanism of this PVC blend is given in Figure 2.8. The fusion time of PVC/OPE compounds was the longest among other PVC blends. This was attributed to a mobile layer, resulting from synergistic combined lubrication between OPE and CaSt, which increased the separation and slipping among particles (Figure 2.8e). Meanwhile, the fusion time of PVC/CPE/OPE compounds was the shortest among other PVC blends. The interaction between OPE and CPE formed a powerful,viscous material and appeared to act as a glue, which enabled the PVC particles to fuse together easily. 31 Figure 2.8 Lubrication mechanism, showing metal lubrication and lubrication between PVC microparticles flow units of PVC/CPE, PVC/OPE and PVC/CPE/OPE compounds (Chen and Lo, 1999) 32 2.2.3 Processing Aid Processing aids are used in PVC formulations for many reasons, but they can all be grouped into two general categories of effects (Dunkelberger et al., 1993). Any proposed mechanism for the action of processing aids must consider these two different categories. • They accelerate and control the fusion process in PVC compound. • They strongly affect the rheological characteristic of the fully fused PVC melt. The melt rheology can be further divided into four groups, i.e. promotes homogeneous melt, increase melt strength, increases melt viscosity and controls die swell. The processing aid mechanism for promoting the fusion is still not well understood. Bohme (1975) proposed that the processing aids promoted friction between hot processing surfaces and the PVC compound during the fusion process. Menges and Berndtsen (1976) argued that the processing aids became part of the interstitial putty that initially bound the particles together and thus promoted the fusion process. Krzewki and Collins (1981) proposed a combination of particleparticle and particle-metal friction processes as the role of the processing aids. These previous findings, in general, elucidated that processing aids generated frictional heats and improved the heat transfer throughout the system during processing. Casey and Okano (1986) used acrylic processing aid to investigate the rheological characteristic of PVC. They suggested a generally accepted theory to explain the function of processing aids. The performance of processing aids is due to their high affinity for PVC, and their very high molecular weights. These characteristics form a rubber-like elasticity by the entanglement of processing aids molecules with PVC. The processing aid is thought to melt first and causes the PVC to become a tacky mass. The long chain of the processing aid increases the viscosity 33 of this mass (melt viscosity), causing a heat build up from the processing shear. The resulting PVC is fully melted and homogeneous before it exits the processing equipment. A homogeneous melt is one where there is minimum amount of primary particles present to minimize melt fracture (increase melt strength) and shark skin effects. Pfening and Dunkelberger (1986) studied the effect of processing aids on PVC die swell. The die swell behaviour of PVC melts is a manifestation of elastic properties of the polymer melt. They reported that the addition of high molecular weight acrylic processing aids to PVC provided better die swell control. The die swell, including melt strength and melt elasticity increased with increasing molecular weight of the processing aid (Casey and Okano, 1986). Hence, the knowledge of molecular weight variables of processing aids is important for controlling the die swell. Sombatsompop and Phromchirasuk (2004) studied the effect of acrylic-based processing aids on processability, rheology, thermal and structural stability and mechanical properties of PVC/wood sawdust composites. They found that the addition of acrylic-based processing aids into composites improved the thermal and structural stability composites, which were evidenced by an increase in glass transition and decomposition temperature and a decrease in polyene sequences, respectively. The changes in the mechanical properties of the composites involved a composite homogeneity, which was varied by degree of PVC entanglement and the presence of wood. There are several types of acrylic-based processing aids and they are highly compatible with PVC. They are methyl methacrylate-ethylacrylate (MMA-EA), methacrylate-butyl acrylate (MMA-BA), methylmethacrylates-butyl acrylate-styrene (MMA-BA-ST) (Dunkelberger et al., 1993). 34 2.2.4 Pigment Photo-oxidation is an oxidation process of the polymers initiated by ultraviolet (UV) radiation in the presence of oxygen and/or water. The oxidation of the polymer chain occurs preferably in the polyene sequences and reduces their length, resulting in a reduction or even elimination of colour. Besides that, oxidation process leads to scission of the polymer chain and intermolecular crosslinking. Intermolecular crosslinking causes material embrittlement (Gardiner et al., 1997). In order to achieve long-term weatherability of PVC products, degradation process must be avoided. A number of different pigments may be added to PVC-U but the most commonly used pigment for outdoors applications is inorganic pigment, namely titanium dioxide (TiO2). This pigment is extremely efficient in scattering the visible light and in absorbing some ultraviolet light down to 300 nm. Sunlight in the 300-320 nm regions is reported to be most effective in initiating the degradation of PVC (Mathur and Vanderheiden, 1982). According to Mathew (1993), TiO2 pigment is readily dispersible and has good chemical stability and hiding power. Besides that, the most important property of TiO2 is its light scattering power for visible light. Hence, its use as a white pigment, and its ability to absorb UV radiation and protect underlying layers, depends on the amount of TiO2 present. Matuana et al. (2001) investigated the photo-degradation and stabilization of rigid PVC/wood fibre composites exposed to the accelerated UV weathering tests. The experimental results indicated that wood fibres were effective sensitizers and that their incorporation into a rigid PVC matrix had a deleterious effect on the ability of the matrix to resist degradation caused by UV irradiation. The light stability of these composites has improved quite efficiently with the addition of rutile (less photoactive pigment) TiO2 during formulation. 35 2.2.5 Impact Modifier Impact modifier is a semi-compatible rubbery polymer, which is dispersed as small discrete domains in the rigid matrix to increase ductility and impact strength (Lutz, 1993). Impact resistance is defined as the ability of a material to withstand the application of the sudden load without failure. The polymer with good impact strength is the polymer that has ability to yield or craze before failure. The enhancement of impact strength of PVC through the introduction of impact modifier has been recognized, mainly to compensate for the detrimental effects of other additives on impact resistance. The detrimental effects occur when the additives used to replace the volume fraction of polymer cannot dissipate stress through the mechanisms known as shear yielding or crazing. Furthermore, certain additives are miscible in the polymer matrix and restricting the mobility of polymer molecules, which allow them to shear yield. Impact modifier is capable of improving the impact resistance of the plastic by facilitating the ability of polymer matrix to shear yield and/or craze (Stevenson, 1995). Lutz (1993) has classified impact modifiers into three classifications: • Predefined elastomers (PDE) impact modifiers, which have constant ultimate particle size and shape modifiers, are sometimes called discrete particle modifiers (DPM). Usually, PDE modifiers have a rubbery core surrounded by a shell of higher glass transitions temperature (Stevenson, 1995). While the rubber is responsible for the impact-modifying performance of the additive, the shell has also important functions: it enables the additive to be isolated as fine free-flowing powders, it facilitates dispersion of the additive into the PVC matrix, and it interacts with the matrix to couple the additive to the matrix. Currently, the major modifiers of this type are methacrylate butadiene styrene (MBS), all-acrylic (ACR) and modified acrylic (MACR). 36 • Non-predefined (NPDE) particles size and shape modifiers. They can also be classified as network polymers (NM). It means that these modifiers must form the network structures enclosing PVC primary particles at the temperature typically in range of 150 – 200 oC in order to develop the desired impact strength (McGill and Wittstock, 1983). The major modifiers in type II are chlorinated polyethylene (CPE) and ethylene vinyl acetate (EVA). Increasing the chlorine content of CPE or the vinyl acetate content of EVA can attain higher degree of compatibility of these modifiers with PVC, and in this case they act as plasticisers rather than impact modifiers. Besides that, because of their inherent solubility in PVC, type II modifiers assume varying morphologies in PVC melt depending on the conditions of temperature, time and shear rates encountered in processing. • Transitional modifiers, due to their properties between the PDE and NPDE. These modifiers are crosslinked rubber/elastomer moiety (compound) and are sensitive to processing conditions. Acrylonitrile butadiene styrene (ABS) terpolymer is categorized in this type of modifier. By varying the ratio of ABS, a wide variety of impact, modulus and clarity properties in PVC can be obtained. Stevenson (1995) reported that one of the major functions of impact modifier is to act as stress concentrators located throughout of the polymer matrix. The stress concentrations occur in the polymer matrix around an impact modifier particle upon application of a tensile stress to the material. The tensile stress is amplified in the region surrounding the equator and is decayed rapidly to the level of the applied stress as the distance from the modifier particle increases. The stress also changes quickly as one moves from equator toward the poles of the modifier particle and is greatly reduced from the applied stress near the poles. This stress distribution is illustrated in Figure 2.9. 37 The localized concentrations of the stress when dispersed through the polymer matrix provide multiple sites at which shear yielding and/or crazing of the polymer can initiate simultaneously upon impact. The result is a structure with a large number of small crazes and/or shears bands instead of a structure with a small number of large crazes or shear bands that is more prone to failure. Figure 2.9 Stress distribution in polymer matrix surrounding a rubbery impact modifier particle (Stevenson, 1995) In the case of additives where the modulus is higher than the polymer matrix, the tensile stress is amplified at the poles rather than equator of the particle and much reduced stress is found at the equator. Even though the stress distribution is different; the stress concentrations are still able to induce shear yielding and crazing and are therefore able to increase the impact resistance of the material. While providing sites of stress concentration in the polymer matrix, rubbery impact modifier is also able to elongate significantly without completely debonding from the matrix or completely fracturing, which can help keep crazes from becoming cracks. It means that impact modifier can restrain craze growth by bridging the crazes. 38 2.3 Impact Modification The modification of PVC by the incorporation of impact modifier to improve the toughness is a common industrial practice. The morphology and subsequently the properties of the PVC itself are very sensitive to processing conditions. Benjamin (1980) and Bystedt et al. (1988) have reported that the optimum ductility occurred when some of the primary particles remained as a dispersed phase in the PVC continuum. Thus, the effectiveness of impact modifiers to modify the toughness of PVC not only depends on their degree of dispersion and distribution and good adhesion to the PVC matrix but also on the presence of primary particles of PVC. 2.3.1 Theory of Impact Modification Mertz et al. (1956) have proposed the earliest mechanism of impact modification in which high impact polystyrene (HIPS) was used in their study. They suggested that the increased energy absorption in the heterogeneous blend represented the energy required to fracture the rubber particles as cracks passed through them. This model is inadequate since the energy absorbed by composites far exceeded than that required breaking the rubber particles (Newman and Strella, 1965). Newman and Strella (1965) proposed that increased energy absorption in the rubber-modified plastics was caused by the large energy involved in the local deformation or cold drawing of glassy the matrix polymer. They associated this increased cold drawing with an increase in free volume of the glassy matrix near the rubber particles. Such was because the rubber particle, with Poisson ratio near 0.5 exhibited no volume change on deformation, then the lack of volume change in the rubber must be due to an accentuated expansion of the matrix immediately surrounding the rubber particle. This effectively lowered the glass transitions 39 temperature of the surrounding matrix, making it deforms more easily, and absorbed more energy in the viscous flow process. A craze theory developed by Bucknall and Smith (1965) proposed that the rubber particles increased toughness by inducing greater crazing prior to fracture. The increased degree of crazing was accompanied by greater energy absorption. Besides, rubber particles lowered the craze initiation stress by providing sources of local stress concentration entirely in the matrix and further strengthened the craze by bearing (relevant) a portion of the triaxial stresses present at the particle-matrix interface (equators). Thus, crazes initiated and terminated at rubber particles. It can be concluded that the function of rubber particles was to promote and control deformation in the matrix. Bucknall’s crazing mechanism of impact reinforcement is the generally accepted one until now because it has been demonstrated conclusively by using HIPS system (Petrich, 1973). Currently, Lu et al. (2000) studied the toughening mechanisms in commercial thermoplastic polyolefin blends and found that the crazes started at the rubber-matrix interface, propagated through the matrix, and terminated at the surrounding rubber particles. This indicates that the rubber phase acted not only as a stress concentrator but also as a craze stabilizer to help prevent crazes from transforming into premature cracks that could lead to catastrophic failure. Lutz (1993) reported that the impact strength increased monotonically with impact modifier particle size until an optimum was reached. For PS, large particles were needed to yield by multiple crazing whereas in PVC, small particles were needed to yield predominantly by shear band formation. Liang and Li (2000) divided thermoplastic polymer matrices into two types, brittle matrix and ductile matrix. For rubber-toughening brittle polymer system, such as PS and PMMA, the external energy is mainly dissipated by the formation of crazing in the matrix. On the contrary, for rubber toughening ductile polymer such as PVC (PVC is intrinsically more ductile 40 than other glassy plastics such PS) the dissipation of the external impact energy depends on the shear yielding of the matrix. 2.3.2 Impact Modified PVC PVC systems modified with EVA and CPE have shown that the EVA or CPE with minimum impact modifier concentration envelops the PVC primary particles to form a network structure (Robinovic, 1983). This structure was able to dissipate the impact forces without cracking and resulted in high impact strength. However, at higher processing temperature (180 oC) the network structure was destroyed, and the decrease in impact strength observed was attributed to a phase inversion in which the modifier became distributed as finer droplets within a continuous PVC phase (McGill and Wittstock, 1987). The mechanism of PVC system modified with acrylic was completely different than that in EVA or CPE systems. Instead of a network structure, the acrylic impact modifier formed a distinct dispersed phase (acrylic particle) and continuous phase (PVC) in order to achieve high impact strength. The impact strength obtained from this system was independent of the processing temperature. This was because the morphology of dispersed rubber particles in the continuous PVC phase was maintained at higher temperature (Robinovic, 1983). Meanwhile, CPE system was dependent on processing temperature, shear and time. Around 150 oC, the network structure has not developed and beyond 200 oC, melting of primary particles resulted in the CPE converting into spherical globs dispersed in the PVC matrix resulting in poor impact strength (Robinovic, 1983; Siegman and Hiltner, 1984). Siegmann and Hiltner (1984) studied the impact modification PVC with CPE. The PVC/CPE blends were roll milled at temperature of 178 oC for 5 min and then 41 moulded to the desired thickness in the preheated press at 193 oC. The morphology of the two phases system changed with increasing CPE content from a dispersion of discrete CPE particles to a continuous CPE network, which surrounded the primary particles. The discrete particles of CPE had very little effect compared to that achieved by continuous network of impact modifier in improving the impact strength of PVC. The network structure was destroyed when the blend milled for longer times or above the fusion temperature of the primary particles. McGill and Wittstock (1987) found that both systems, acrylic (KM330)-PVC and EVA- or CPE – PVC required a minimum melt processing temperature of about 170-175 oC before high impact properties were attained. This temperature provided high energy to cause disintegration of PVC grains and dispersion of primary particles. It seems that at least partial fusion of primary particles was necessary for high impact strength. ABS terpolymers have been widely used as impact modifiers in rigid PVC, because the polybutadiene rubber phase dispersed in the PVC matrix to form small discrete domains, and the graft terpolymer provided interfacial bonding between the phases to permit efficient stress transfer and thus increased the impact strength but lowered the rigidity of PVC-U. Deanin and Chuang (1986) reported that ABS had less softening effect and gave the best overall balance of properties than CPE. The reasons behind this effect were SAN had high modulus and was quite miscible in the PVC matrix and a polybutadiene formed separate domains and did not soften the PVC matrix, whereas CPE was both soft and flexible, and also fairly miscible in PVC matrix, thus softening it more aggressively. 42 2.4 Filled PVC Composites PVC resin plays important role in plastic industry as a major component of composites manufacturing. During the last few years, researchers have concentrated their efforts to develop natural fibre-filled PVC composites (Wenguang and La Mantia, 1996). Various types of natural fibres such as wood fibres, wood sawdust, rich husk ash, have been extensively investigated and reviewed (Matuana et al., 1998; Sivaneswaren, 2002; Sombatsompop et al., 2003). Other common fillers for filled PVC composites include calcium carbonate, talc, kaolin, carbon black, mica (Wee, 2001; Saad et al., 1999). Characterised by high stiffness and strength, natural fibre as fillers offer a number of advantages over the currently used reinforced inorganic materials in terms of cost on a unit-volume basis, flexibility during processing and stiffness (Matuana et al., 2001) Although natural fillers have received increased attention but the inorganic fillers are still widely used in PVC composites. Hence, this section briefly discusses the use of inorganic fillers in the PVC. 2.4.1 Composite: Definitions and Categories The word ‘composite’ means ‘consisting of two or more distinct parts’. A composite material is formed when two or more materials are combined so that the properties of the composite are different from, and usually better than, those of the individual constituents. Composites are made up of continuous and discontinuous mediums. The continuous medium is called matrix and the discontinuous medium, which is usually the harder and the stronger one, is called reinforcement. The properties of a composite are dependent on the properties of the constituent materials, and their distribution and interaction (fibre-matrix interface)(Mahmood Husein Datoo, 1992). 43 Composite materials can be broadly classified into three categories: particulate composites, fibrous composites and laminated composites (Mahmood Husein Datoo, 1992). The definition of fibrous composites is referred to the fibre which has the particle size longer than 100 µm with a length to diameter ratio lesser (aspect ratio) than 10, whereas the opposed particle size as stated above is referred to the particulate composites. (Katz and Brandmaier, 1987) The main roles of the matrix are to bind and transmit and distribute stresses among the individual fibres, and to prevent the fibres separation and to give them desired orientation. The matrix also provides protection against both fibre abrasion and fibre exposure to moisture or other environmental conditions, and causes the fibres to act as a team in resisting failure and deformation under load. The maximum service temperature of the composites is limited by the matrix ((Katz and Brandmaier, 1987; Miller, 1996) Although thermoset polymers have some great virtues when used as matrices in plastic composites such as low viscosity to facilitate thorough wetting of reinforcing particles, economical for production large components and high softening points, however recent years have seen rapid growth in the used of reinforced thermoplastic polymers. The major advantage of thermoplastic matrix is that forming is possible by injection moulding or extrusion techniques. These are the most economical processes when cheap and precise manufacture of very large quantities of components is required (McCrum et al., 1997). However, some special consideration must be made for thermoplastics due to their poor high-temperature capability (Collyer and Clegg, 1986). Most of the PVC is used as composite matrix in the mineral filler-filled PVC composites. Calcium carbonate (CaCO3) has been extensively used in PVC composites to enhance mechanical and thermal properties of PVC, such as hardness, stiffness and heat resistance. However, the use of fillers at high loading leads to a 44 decrease in yield stress, tensile strength and ductility compared to the original polymer (Saad et al., 1999). Recently, PVC matrix is used as the matrix for nano CaCO3 particles. The impact strength, flexural modulus and softening point temperature of PVC are significantly enhanced after addition of 0-15 phr nano CaCO3 but the yield strength and elongation at break of PVC are decreased (Chen et al., 2004). PMMA treated talc is an other inorganic filler used as filler in the PVC matrix. The dispersion of talc and the mechanical properties of PVC composite especially the impact strength are found to be improved. The enhancement of interfacial adhesion is attributed to the impact strength improvement (Xie et al., 2001). Agalmatolite mineral filler is also used in the PVC matrix. The composite made with agalmatolite shows slightly higher values in Young’s modulus and elongation at break than those using CaCO3. Thus the resistance to elastic deformation and plasticity is enhanced when agalmatolite is used in PVC composites (Pedro et al., 2001) The previous findings mentioned above show that PVC thermoplastic is also competent to be a good matrix in the composite system. Although unlike thermosetting resins, if more ductile and versatile PVC thermoplastic is modified with some consideration on heat processing, then the potential to use PVC as matrix in the thermoplastic composite is possible. 2.4.2 Filler and Reinforcement The basic purpose of fillers is to ‘fill’ a plastic that is to increase the bulk at low cost and to improve the economics. With good adhesion between filler and polymer, the filler can play a role in improving the mechanical properties of the polymers (Murphy, 2001). Therefore, currently the primary purpose for using fillers 45 in thermoplastics is to modify properties such as stiffness, hardness, thermal stability and shrinkage, instead of reducing the material cost. In this case, the fillers function as the reinforcing fillers in the composites. Reinforcing fillers are additives, which improve the tensile strength and tear resistance of the composites whereas extenders reduce both of these properties. In general, the less fibrous (more particulate) fillers, which have the lower aspect ratio, provide less reinforcement (Schwartz, 1987). Whether extender or reinforcing filler, the effect of fillers on the composite properties are largely dependent on the mean particle size, particle shape, particle size distribution and the bond strength between filler particles and polymer matrix (Wiebking, 1986). Most filler do not consist of particles, which are all the same size and the same shape. The shape of the particles affects the modulus and other mechanical properties significantly. Spherical particles are less likely to initiate cracking compared to more slender particles. However, flat platelet-type particles tend to produce stronger bond between the filler particle and matrix, due to increased surface area available for bonding. One measurement of shape is the aspect ratio. The greater aspect ratio, the greater the increase in bonding area available, which usually improves mechanical properties. (Miller, 1996). Example of aspect ratio calculation is illustrated in Figure 2.10. According to Wiebking (1986), the flexural modulus or stiffness of PVC composite was directly proportional to the aspect ratio of the filler being used. A high aspect ratio can also increase the tensile properties of PVC composite by providing more surfaces to contact the polymer. As a material is made stiffer, it also becomes more brittle. Miller (1996) reported that the tensile strength depended on the interfacial bond strength, which determined the distribution of stress between filler and matrix. He concluded that, in general, the smaller the average size of the filler particles, the greater their surface areas and the greater the bonding, resulting in composite with good tensile strength. 46 Figure 2.10 Aspect ratio calculation (Wiebking, 1986) 2.4.2.1 Mineral Fillers As mentioned earlier, mineral fillers have been widely used in the PVC formulations. Among the mineral fillers used in the PVC are CaCO3, clay, talc, mica and so forth. CaCO3 is the predominant filler used in PVC composites because of the cost advantages and other appealing properties, which includes strong resistance to thermal degradation during processing. Clays are useful for PVC in some specific application such as electric insulator products. Meanwhile talc, mica, and other silica are mainly employed with the aim of lowering material cost (Schlumpf, 1993). Another category of filler materials used in PVC is termed “functional filler” and this includes glass fibre, glass sphere, carbon fibres, aramid fibres, nano –clay and nano calcium carbonates (Schlumpf, 1993; Wan et al., 2003; Chen et al., 2004) Their use is mainly for consideration on the performance of the composites rather than of cost. Although these organic fillers are low-cost on weight basis, their densities (specific gravities about 2.5) can increase the cost on volume basis and make processing more difficult. 47 2.4.2.2 Organic Fillers Organic fillers such as wood flour, pulp fibre, rice husk are also used in PVC compounds. They possess low specific gravities (about 1.4) and when used to extend the resins they offer a significant cost saving. However, the use of cellulosic fillers in the PVC materials is not as common as in polyolefins (polypropylene and polyethylene). Organic filler or natural fillers are biological structures composed mostly of holocellulose (combination of cellulose and the hemicellulose) and lignin, and other traces of plant materials, namely ash content and extractives (a group of cell wall chemicals mainly consisting of fats, fatty acids, fatty alcohol and so forth). The inorganic content can be quite high in plants containing large amounts of silica. The extractives can be removed by one or several extraction procedures (Han and Rowell, 1997). Natural fibres or fillers are rich in hydroxyl groups, which are responsible for moisture sorption via hydrogen bonding, bio-degradation and thermal degradation of the fibre whereas lignin is thermally stable but is responsible for the UV degradation. Generally, without non-wood fibres, the fibres contain 60-80 % cellulose, 5-20 % lignin and up to 20 % moisture. The hydrogen bonds and other linkages provide the stiffness to the fibres (Saheb and Jog, 1999). Figure 2.11 shows the structure of cellulose molecules. They have a tendency to form intra and intermolecular hydrogen bonds. Low thermal stability of natural fillers is the primary drawback. The thermal degradation of natural fillers is a two-stage process, namely degradation of hemicellulose and cellulose at low temperatures (220-280 oC) and degradation of lignin at high temperatures (280-300 oC). The suggested processing temperatures are thus limited to about 200 oC because of the possibility of natural fillers degradation 48 and/or the volatiles emissions that could affect the composites properties if processed at higher temperatures. This factor will limit the type of the thermoplastics that can be used with agro-fibres to only commodity thermoplastics. O H OH OH H H H C H 2O H H OH OH H O O H H O H O OH H H H H H OH H H O O H OH H H OH H O O C H 2O H C H 2O H C H 2O H n n= repeat unit Figure 2.11 Cellulose molecular structure (Han and Rowell, 1997) Another primary drawback is the high moisture absorption of the natural fillers. The moisture absorption can result in swelling of the fibres, dimensional variation in composites and affects the mechanical properties of the composites. The moisture evaporation, during processing of thermoplastics composites can lead to poor processability and porous products. (Sanadi et al., 1997). Uniform distribution and interaction of natural fillers is also important, so that as many polymer chains as possible can be bound to the free filler surface. However, natural fillers have poor interactions with the polymer matrix and tend to aggregate among themselves due to strong hydrogen bonding, which holds the fillers together, thus the fillers do not disperse well in a hydrophobic polymer matrix. Due to the poor adhesion at the fibre-matrix interface, the fillers do not function as an effective reinforcement system. Modification or surface treatment of the fibres with suitable chemical reagents prior to processing can lead to good dispersion and adhesion, significantly improves mechanical properties and also reduces water absorption of the composites. 49 Thus, the important factors, which need to be considered in the application of natural fibres as fillers or reinforcements in the composite system, are; • thermal stability of the fillers • surface adhesion characteristics of the fillers • dispersion of the fillers The incorporation of wood fibre derived from cellulose into PVC matrix has been studied extensively by Matuana et al. (1997, 1998, and 2001). The main reason for using wood fibre or fibrous cellulose materials lies in the fact that these fibres offer a number of advantageous over currently used reinforcing inorganic materials in terms of cost on unit volume basis, flexibility during processing and stiffness. Matuana et al. (2001) studied the photo-aging and stabilisation of rigid PVC/wood-fibre composites. The experimental results showed that wood fibres were effective sensitisers and that their incorporation into a rigid PVC matrix has accelerated the photo-degradation of the PVC matrix when exposed to UV radiation. They also reported that the light stability of these composites could be improved quite efficiently with the addition of rutile titanium dioxide photoactive pigment during formulation. In their previous investigation, the incorporation of wood fibre increased stiffness and melt viscosity of the composites and limited the potential of producing a foamed composite with high void fraction (greater than 20 %). However with good treatment of fibre (silane), the foamed PVC composites with void fraction larger than 20 % were successfully achieved (Matuana et al.,1997). Mengeloglu (2001) has studied the effects of impact modifiers such as methacrylate-butadiene-styrene (MBS), all acrylic (ACR) and CPE on the properties of rigid PVC/wood fibre composites. The wood fibre from hardwood maple was used. Based on the mechanical property testing, the author concluded that the MBS and ACR were more efficient and effective in improving the impact resistance of the 50 composites than the CPE modifier. Meanwhile, the tensile strength and modulus of composites were significantly decreased by the addition of all of the impact modifiers. The effect of rice husk ash (RHA) in the PVC-U has been studied by Sivaneswaran (2002). The RHA-filled PVC has been impact modified with ABS and the coupling agent with Lica 12. The results showed that the sample containing 8 phr ABS and 40 phr RHA fillers treated with Lica 12 was the most effective formulation in terms of flexural modulus, impact strength and cost. Sombatsompop et al. (2003) have used sawdust as filler in the PVC to examine the effect of sawdust content on structural and thermal changes, and rheological and mechanical properties of PVC composite. The results revealed that the extrudate swell monotonically decreased, up to 33.3 wt % sawdust content, and became independent for sawdust content at greater than 33.3 wt %. Meanwhile, tensile, impact, flexural and hardness properties considerably decreased with up to 16.7 wt % sawdust content before levelling off at higher sawdust loading. The decrease in mechanical properties was associated with the presence of moisture, interfacial defects and fibre dispersion in PVC matrix. Thermal degradation of composite also decreased with increasing sawdust content, which was caused by chlorine atoms cleavage due to strong hydrogen bonds of wood fibre and PVC molecules. However the glass transition temperature was found to increase after adding 16.7 wt % sawdust due to strong hydrogen bonds and polar-polar interaction between PVC and sawdust molecules. 2.4.3 Filler-Matrix Interface One of the factors controlling the performance of the composite materials is the bond between the filler and the matrix. The localized stresses are usually highest at or near the interface, which may be the point of premature failure of the composite. 51 As a result, a high interfacial strength is desirable to maximize the overall composite strength. The filler-matrix interface bond acts as a bridge to allow the dispersed phase (filler) to ‘communicate’ with matrix to transmit the stress or strain due to a mechanical load from one phase to the other (Shackelford, 1996). The interface must have appropriate chemical and physical features to provide the necessary load transfer from the matrix to the filler. The use of a coupling agent can provide improved interfacial condition (Katz and Brandmaier, 1986). 2.4.3.1 Coupling Agents and Surface Modification The strength performance of the cellulosic fibre/polymer composites is generally lower compared to unfilled polymer. The decreased strength is likely a result of poor adhesion and/or dispersion of fibres due to the natural incompatibility of phases during mixing of the hydrophilic cellulosic fibres with the hydrophobic polymer matrix. To improve the adhesion between fibres and polymer, coupling agents are usually used. Coupling agents are able to react chemically on both ends during processing, with the fibres on the one side and the polymer on the other, thus creating a chemical bridge at the interface (Xanthos,1983; Raj et al., 1990; Ahmad Fuad et al., 1997; Rozman et al.,1998; Sivaneswaren, 2002). The scheme for the reaction of silanes coupling agent (for instance) with cellulosic materials is shown in Figure 2.12 (Ahmad Fuad et al., 1997). In the first step, silanes are hydrolyzed to form alkoxysilanol groups with water molecules on the surface of cellulosic fillers. In the second step, the absorbed alkoxysilanols condense with the surface hydroxyl groups of cellulosic materials to form ether bonds. The alkoxysilanols attached to hydroxyl groups of cellulose and/or lignin at the surface of cellulosic materials were accomplished through the ether linkage (-O-) via structure 1, as illustrated in Figure 2.12. This structure assumed the condensation of alkoxysilanols to the filler surface (covalent deposition) has involved only one alkoxy 52 group of alkoxysilanols. However, silanols are also known to condensate amongst themselves (intermolecular silane formation) during hydrolysis (structure II). Step 1: X − Si(OR ) 3 ⎯hydrolysis ⎯⎯ ⎯→ X − Si(OH) 3 Step 2: Reaction with filler surface + OH + X Si OH OH O H H OH X Si O Filler Surface X Si OH OH OH H OH Filler Surface X Si OH OH OH O O H OH Condensation O X Si O OH Structure II Figure 2.12 + OH X Si OH O OH X Si O Filler Surface O Filler Surface OH X Si O OH Structure I Bonding mechanism of silane coupling agent to filler’s surface (Ahmad Fuad et al., 1997) According to Xanthos (1983) and Raj et al. (1990) the wood flour or fibres has the potential to be used as reinforcing fillers in the thermoplastics when they are treated with chemical reagents. Therefore, various chemical reagents have been 53 employed to enhance the compatibility between the constituent materials. In order to overcome the incompatibility between cellulosic fibre and PVC, Kokta et al. (1990) have assessed the effects of various types of isocyanates such as poly(methylene polyphenyl isocyanate)(PMPPIC), toulene,4-diisocyanate (TDIC), hexamethylene diisocyanate (HMDIC) and ethyl isocyanate (EIC) as coupling agents. The mechanical properties of hardwood aspen in the form of chemithermomechanical pulp (CTMP) and sawdust were studied (Maldas and Kokta, 1991). The results showed that the isocyanates acted either as a promoter or inhibitor, depending on the concentration of the isocyanates used. For example, when used with moderate concentration (0.5 to 5 % by weight), isocyanates promoted maximum mechanical properties (tensile and flexural properties) whereas with greater concentration, the mechanical properties deteriorated. PMPPIC had the best reactivity among the four types of isocyanates studied and offered better performance of composites. A stronger interface formed due to the delocalized electrons of benzene ring in PMPPIC chemical structure provided little polarity, which interacted with the opposite polarity of PVC developed by the presence of electronegative chlorine and electropositive carbon and hydrogen. Besides that, the nature of pulp also played important role in the mechanical properties of composites. The CTMP fibres performed better in stress and modulus properties of composites than sawdust fibres. It was explained by the dissimilar preparation techniques. CTMP fibres were separated by chemical and mechanical means whereas sawdust fibres were separated by cutting in mechanical way only. Therefore CTMP fibres possessed more specific area leading to better polymer linkage at interface. The effectiveness of other coupling agents has been studied. Matuana et al. (1998) have studied the effects of various types of coupling agents such as aminopropyltriethoxy-silane (A-1100), dichlorodiethylsilane (DCS), phthalic anhydride (PA) and maleated polypropylene (E-43) on the compatibility between wood veneer laminates and PVC. The effectiveness of surface treatment on the wood –fibres in the PVC/wood fibre composites was investigated. The results showed that the adhesion between PVC and wood veneer laminates was significantly improved 54 when wood veneers were treated with A-1100, while no improvement was observed for the other coupling agents. The improvement was attributed to the fact that A-1100 treated wood veneer chemically reacted and formed an ionic bond with the matrix because of the highly electronegative nature of chlorine atoms. Recently, the rice husk ash (RHA) filled PVC-U matrix was investigated (Azman Hassan and Kannan, 2001). RHA was modified with silane (Prosil 9234), titanate (Lica 12 and Lica 38) and zirconate (NZ 44) coupling agents. The results showed that the impact strength and flexural strength of impact-modified PVC decreased with increasing RHA contents, however flexural modulus was increased. The coupling agents used improved the mechanical properties but in overall application, Lica 12 provided the optimum balance of properties for impact modified PVC-U/RHA composites. These recent studies revealed that although the strength property of PVC/cellulosic fibre composites increased slightly after fibre treatments, the strength property of the composites was still less than the base polymer, i.e., PVC. Therefore PVC/cellulosic fibre composites with similar or greater strength to unfilled PVC have not yet been reported in the literature. This is because an appropriate coupling agent or its suitable form of application for PVC/cellulosic fibre composites has not yet been identified. 2.5 Oil Palm Empty Fruit Bunch -Polymer Composites As mentioned before most of the published papers have reported the use of EFB fibre or filler as reinforcement in the thermoplastics polymers such as polyethylene (Mohd Ishak et al., 1998), polypropylene (Rozman et al., 2001) and polystyrene (Sarani et al., 2001) and in thermosets such as polyester (Hill and Abdul 55 Khalil, 2000 ; Abdul Khalil et al., 2001) and phenol-formaldehyde (Sreekala et al., 1996; Sreekala and Thomas, 2003). However, areas for using of EFB fibres or fillers in the PVC composites are still wide, especially in terms of the processability, thermal stability-related to mechanical properties of the composites (Sombatsompop et al., 2003). 2.5.1 Oil Palm Empty Fruit Bunch Fibre and its Chemical Composition Oil palm empty fruit bunch (EFB), one of the by-products of oil palm milling process, can be found in abundance in palm oil mills in Malaysia. Compared to other by-products such as oil palm trunk (OPT) and oil palm fronds (OPF), the EFB has the economical advantage of being collected at the oil palm mills. Currently, Sabutek Sdn. Bhd, Ipoh, Perak, Malaysia has commercially produced the pulverized EFB fibres and fillers. The EFB fibres are derived from oil palm tree (Elais guineensis) component. The chemical composition of untreated EFB fibre strands has been reported by Peng (2002) as shown in Table 2.3. Table 2.3: Chemical analysis of EFB (Peng, 2002) Chemical Composition Percentage (%) Extractive 2.3 Holocellulose 82.5 α-cellulose 60.6 Lignin 17.2 Hemicellulose 32.5 Ash content 5.4 56 2.5.2 Utilization of EFB Fibre in Polymer Composites Utilization of short EFB as a reinforcement in phenol-formaldehyde (PF) resins has been studied by Sreekala et al. (1996). Four different fibre lengths of 20, 30, 40 and 50 mm with various fibre loadings were used for composite fabrication by means of hand lay-up. The tensile, flexural and impact properties showed improvement upon reinforcing with the fibre. The elongation, brittle nature and buckling characteristics of PF resin were considerably improved by incorporating the fibre. The fibre length of 40 mm and at 38 % fibre loading has produced the maximum tensile and flexural strength. Fibre entanglement at higher fibre length was the reason for the decrease in properties beyond 40 mm fibre length. Meanwhile the impact properties showed linear enhancement with increase in fibre length up to 50 mm. However, for the best balance of mechanical properties, the optimum fibre length and fibre loading were considered to be 40 mm and 30 %, respectively. EFB fibre is also utilized in thermoplastic composites. Rozman et al. (1998) have studied the effect of compounding techniques, the effect of oil palm EFB filler loading and the effect of coupling agents on the mechanical properties of EFB/PP composites. The study revealed that the incorporation of EFB into PP matrix resulted in the improvement in the tensile modulus. However, the tensile strength, elongation at break and impact strength decreased with increasing filler loading. Composites produced by internal mixer have displayed higher mechanical properties than those produced by extrusion. The better performance has been attributed to the effectiveness of the internal mixer in improving the wetting of the filler surface. Meanwhile, the incorporation of compatibilizer and coupling agent, that is, maleic anhydridemodified PP (Epolene E-43) and 3-Aminopropyl triethoxysilane (3-APE), respectively, have produced composites with improved tensile and impact strength. The presence of coupling agent and compatibilizer has also affected the amount of water absorbed. The water absorption decreased as the amount of coupling agents increased. 57 The effect of chemical modified oil palm EFB, fibre-matrix interaction of composites, the effect of silane-based coupling agents treated EFB, and acrylic acid based compatibilizers on the mechanical properties of high density polyethylene (HDPE) also has been investigated by Mohd Ishak et al. (1998). They have found that tensile modulus of EFB-HDPE without chemical treatment showed an increase, whereas tensile strength, elongation at break and impact strength decreased with the increasing filler loading. The strong tendency of EFB to exist in the form of fibre bundles, poor filler-matrix interaction was believed to be responsible for the poor strength displayed by the composites. Meanwhile, almost all chemical treatments increased the stiffness of the composites but limited improvement has been observed in the case of tensile strength. This has been attributed to the presence of fibre bundles that remained intact even after several types of chemical treatment have been carried out. Scanning electron microscopy (SEM) micrograph revealed that the main energyabsorbing mechanisms contributing towards toughness enhancement was through the fibre bundle pullout process. Rozman et al. (2001) in their recent paper reported the effect of oil extraction of the oil palm EFB on the mechanical properties of PP/EFB/glass fibre hybrid composites. The results showed that composites with oil-extracted EFB produced significantly higher flexural and tensile strength and toughness than those without extraction. The study showed that the efficiency of coupling agent was enhanced with the removal of oil from the EFB surface. The main reason being the oil extraction has resulted in the formation of continuous interfacial region i.e. good adhesion between EFB and PP matrix. Abdul Khalil et al. (2001) have used other methods to improve the compatibility between oil palm EFB and the polyester polymer matrix, where EFB was chemically modified using non-catalysed reaction with acetic, propionic and succinic anhydrides. The effects of various anhydride modifications on mechanical properties and water absorption of oil palm reinforced polyester composites were investigated. The properties were improved for modified fibres whereas unmodified EFB fibres exhibited poor mechanical properties and higher water absorption. It was 58 concluded that the chemical modification resulted in hydrophobic fibre and hence improved the fibre-matrix bonding. Another research, oil palm EFB was used in the polystyrene matrix as reported by Sarani Zakaria et al. (2001). The benzylation process treatment was used in order to decrease the hydrophilic property of EFB fibres where the hydroxyl groups were replaced by aromatic rings from benzyl groups to produce benzylated EFB fibres. The results showed that the modified EFB/polystyrene produced better mechanical and physical properties than unmodified EFB/polystyrene composites, due to the enhancement of compatibility between EFB fibres and polystyrene. Recently, Sreekala and Thomas (2003) studied the effect of fibre surface modification on water-sorption characteristics of EFB fibres. Surface modification such as mercerization, latex coating, gamma irradiation, silane treatment, isocyanate treatment, acetylation and peroxide treatment were used to reduce the hydrophilicity of fibres. Treatment reduced overall water uptake at all temperatures (30, 50, 70 and 90 oC). Such reduction was due to the physical and chemical changes in the fibre upon modifications. The effect of sorption on the mechanical performance of the treated and untreated fibres was also investigated and it was found that the properties of the fibres decrease upon sorption and regains on desorption. Based on the literatures reviews, the following conclusions are derived. The addition of EFB fibres or other natural fillers as reinforcement generally has increased the modulus of the composites while decreased the strength, elongation and water absorption. An alternative way to improve the composite properties is by modifying the fibre surface by chemical reagents. However, some of the treated fillers have not shown any improvement and they even destroyed the composites because they could not react with a hydroxyl groups. Rowell (1997) has suggested that the basic principles that must be considered in selecting a reagent and a reaction system are as follows: (1) the chemical must contain functional groups to react with the hydroxyl groups on the fibres, (2) the moisture content must be reduced, due to OH group in 59 water being more reactive (rate of hydrolysis) than the OH group available in the fibre. The favourable condition is a reaction, which requires a trace of moisture content and (3) the chemical bond formed between the reagent and the fibres components. In the order of stability, the types of covalent chemical bonds that may be formed are ethers, acetyls and esters. The covalent carbon-oxygen bond i.e. the ether bond is the most desirable. 2.6 Theory of Mechanical Properties of Filled Composites Both the matrix and fibre properties are important in improving mechanical properties of the composites. Modulus is a bulk property, which depends primarily on the filler properties such as geometry (aspect ratio), particle size distribution and concentration of the filler, whereas tensile strength of a composite depends strongly on local polymer-filler interaction as well as the above factors. The filler particle geometry has significant influence on the strength of a filled polymer. Impact strength of the composite depends on the degree of fibre-matrix adhesion and is very complex. The microscale morphological changes in the matrix caused by the fibre affect the impact strength of composites. The development of the most of the theories is generally based on the assumption that perfect adhesion takes place between matrix and dispersed phases. By perfect adhesion, means that the displacement continuity is maintained at the boundary of the two phases when stress is applied to the system. In many practical cases, however, it cannot be assumed that there is perfect adhesion between matrix and dispersed phases, and in extreme cases, there is no adhesion at the boundary. In this case, the modulus equation is considerably different from the one for perfect adhesion (Kunori and Geil, 1980). 60 One of the first equations developed was that of Einstein’s. It is valid only for composites filled with low concentrations of non-interactive spheres and assumes that the filler is very much more rigid than the matrix (Kunori and Geil, 1980; Maiti and Subbarao, 1991; Zaini et al., 1994). For cases when there is perfect adhesion between phases, the result is: ( E c = 1 + 2.5φ f E p ) (2.1) and for the case when there is no adhesion between the two phases : E E ( c = 1+ φ f p ) (2.2) where Ec, Ep and φf are the modulus of composite, modulus of polymer matrix and volume fraction of filler, respectively. To account for interparticle interactions at higher filler concentrations, the following equation from Guth-Smallwood is more relevant (Zaini et al., 1994). This equation is an extension of Einstein’s theory and is valid for the perfect adhesion system. E c = ⎛⎜1 + 2.5φ + 14.1φ 2 ⎞⎟ f f⎠ ⎝ E p (2.3) Kerner (1969) afterward derived a versatile equation to describe the reinforcing action of the spherical fillers by assuming perfect bonding of the spherical grains to each other: ⎡ ⎤ ⎡ ⎤ φ G / ⎢⎛⎜ 7 − 5ν ⎞⎟G + ⎛⎜ 8 − 10ν ⎞⎟G ⎥ + φ / ⎢15⎛⎜1 − ν ⎞⎟⎥ f f p p p f p p ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎣ ⎦ ⎣ ⎦ c = E ⎡⎛ ⎤ ⎡ ⎤ p φ G / ⎢⎜ 7 − 5ν ⎞⎟G + ⎛⎜ 8 − 10ν ⎞⎟G ⎥ + φ / ⎢15⎛⎜1 − ν ⎞⎟⎥ p f ⎣⎝ p⎠ p ⎝ p⎠ f ⎦ p ⎣ ⎝ p ⎠⎦ E (2.4) 61 where φf and φp are the volume fractions of filler and polymer respectively; Gf and Gp are the shear moduli of filler and polymer, respectively. Since many types of filler are much more rigid than the matrices, thus Equation (2.4) may be simplified to: Ec Ep = 1+ 15(1− ν p ) ⎛ φ ⎜ f 8 − ν p ⎜ 1 − φf ⎝ ⎞ ⎟ ⎟ ⎠ (2.5) The most frequently used model is the Lewis-Nielsen model (also known as modified Kerner’s model)(Nielsen, 1974). Lewis and Nielsen have put forward an equation which is more applicable for fillers with lower aspect ratio. This equation is for the system having perfect adhesion at the boundary. By modifying Equation (2.4), they came to the following equation: ( ( 1 + ABφ c = f E 1 − Bψφ p f E ) ) (2.6) where ⎛ 7 − 5ν ⎞ ⎜ p ⎟⎠ ⎝ ; A= ⎛ 8 − 10ν ⎞ ⎜ p ⎟⎠ ⎝ ⎡⎛ G ⎞ ⎤ ⎢⎜ f ⎟ − 1⎥ ⎢⎜⎜ G ⎟⎟ ⎥ ⎢⎝ p ⎠ ⎦⎥ B= ⎣ ; ⎤ ⎡⎛ G ⎞ ⎢⎜ f ⎟ + A ⎥ ⎥ ⎢⎜⎜ G ⎟⎟ ⎦⎥ ⎣⎢⎝ p ⎠ ( ⎡ 1− φ m ψ = 1− ⎢ ⎢ 2 φ m ⎣⎢ )⎤⎥φ ⎥ f ⎦⎥ A is a constant related to the Einstein coefficient and is a function of the filler’s geometry, B is related to the relative shear moduli of filler, Gf and polymer, Gp, and ψ is a reduced concentration term dependent on the maximum packing fraction, φm of the filler in the polymer while υp is Poisson’s ratio of the polymer. 62 Maiti and Subbarao (1991) found that the data for both untreated and treated wood flour-filled polypropylene composites fit Equation 2.1, while Equations 2.2 and 2.6 exhibited higher values of modulus. The higher values obtained from Equations 2.2 and 2.6 were due to the surface treatment of wood flour with the coupling agent (Lica 38), which improved the interaction only marginally. The increase in modulus was due to the restriction in the mobility of the polymer molecules in the presence of wood flour. However, the increase was not very high because of the absence of chemical interaction between wood flour and polypropylene. In order to describe the dependence of mechanical properties (i.e., tensile strength and tensile yield) of filled composites on the volume fraction of filler, the equation from Nielsen (1966) and Nicolais and Narkis (1971), which expressed by two-thirds power law relationship are given by Equation (2.7) and (2.8), respectively, are commonly used: σ σ σ σ c = ⎛⎜1 − φ 2 / 3 ⎞⎟S f ⎠ ⎝ p (2.7) c = ⎛⎜1 − Kφ 2 / 3 ⎞⎟ f ⎠ ⎝ p (2.8) where σ c and σ p denote the given mechanical properties of the composite and the matrix, respectively. According to Nielsen’s definition, parameter S in Equation 2.7 describes the stress concentration at the filler-polymer interphase arising from the discontinuity in stress transfer. The maximum value of S is unity for “no stress concentration effect,” which is valid for unfilled polymer. The lower the value of S the greater the stress concentration effect (or poorer the adhesion), which is valid for filled polymer composites. The weightage factor K in Equation 2.8 accounts for the adhesion quality between the matrix and the inclusion or the filler. The lower K value indicates better adhesion. The extreme case with K= 0 represents unfilled polymer. 63 For spherical inclusions in the extreme case of poor adhesion, K assumes a value of 1.21. Nielsen (1966) has also proposed a simple model to predict the relative elongation at break with perfect adhesion: εc εp = 1 − φ f 1/3 (2.9) where εp and εc are the elongation at break of the polymer and the elongation of the composites, respectively. Maiti and Subbarao (1991) and Maiti and Lopez (1992) found that the prediction values of relative elongation of untreated wood flour-filled polypropylene composites and kaolin-filled polypropylene composites were slightly higher than the experimental results. It was apparent that the presence of kaolin particles or wood flour interferes with the stress transfer process, in particular in the absence of any specific adhesion between polypropylene and fillers. 2.7 Recent Studies on Impact Modified PVC-U by UTM Researchers Impact modified PVC-U is an area which is being actively pursued by the researchers from Universiti Teknologi Malaysia (UTM). Azman Hassan (1996) in his study on the effect of acrylic impact modifier particle size on unfilled PVC-U has found that the effectiveness in decreasing the yield stress of unfilled PVC-U became more pronounced with decreasing particle size. The smaller particle size impact modifiers were also shown to be more effective in enhancing the impact strength of PVC-U. 64 Wee, (2001) studied the effect of calcium carbonate fillers on the mechanical properties of modified PVC-U. The different sizes and types of calcium carbonate fillers were used to determine the most optimum formulation in terms of mechanical properties of modified PVC-U and cost. He concluded that the impact-modified PVCU filled with 20 phr of 0.8 µm precipitated calcium carbonate was the most optimum formulation based on the criteria. Besides the use of mineral fillers, the application of RHA, which is an organic material as a filler in impact modified PVC-U composites has also been studied (Sivaneswaran, 2002). The sample containing 8 phr ABS and 40 phr RHA filler treated with Lica 12 was found to be the most effective formulation in terms of flexural modulus, impact strength and cost. This study showed that the strategy to develop materials which possessed a significant improvement in toughness without sacrificing the desirable stiffness properties at a moderate cost was achieved. CHAPTER 3 MECHANICAL PROPERTIES OF COMPRESSION MOULDED PVC-U COMPOSITES 3.1 Introduction Unmodified poly (vinyl chloride) (PVC-U) is intrinsically more ductile than other thermoplastics such as polystyrene, and when compared with polyethylene and polypropylene, it is more rigid and stronger. However, it is extremely sensitive to stress rate, temperature, and notching. Increasing the impact rate, reducing the test temperature or providing notch as a stress concentrator, can dramatically reduce PVC’s ductility (Grohman, 1994). PVC has to be modified in order to satisfy end-use requirements for rigid applications. Thus, therefore impact modifiers are incorporated to PVC to improve its ductility. The enhancement of impact properties of PVC by introduction of a rubbery phase is a well-known phenomenon, which has been widely exploited commercially. Basically, the effect of impact modifier is to shift the ductilebrittle transition of the PVC to the higher impact rate, lower temperature, or sharper notch. While PVC’s ductility is improved, there are other effects, which must be considered such as the tensile and flexural properties of rigid PVC. Lutz (1993) 66 reported that these properties of PVC have adversely affected with increasing impact modifier concentration. The incorporation of soft rubbery impact modifier has lowered the rigidity of the rigid PVC. Modifiers with lower glass transition temperature and greater compatibility have a greater effect on the impact strength. Among the families of plastic tougheners, acrylic core-shell impact modifier (acrylic) and chlorinated polyethylene (CPE) are commonly used in formulation of PVC, particularly for window profiles or outdoor applications. Rigid PVC is used in many applications where cost is a major factor. In order to reduce the cost of products, as a result, fillers are often added in the rigid PVC formulations. However, the incorporation of fillers into thermoplastics results in stiffness enhancement but reduces the impact strength or toughness. Thus, the rubbery impact modifier is added in the rigid PVC in order to produce a significant improvement in toughness but at the expense of stiffness and strength. One of the most important aspects in the modification properties of thermoplastics by using fillers is to achieve a good combination of properties and processability at a moderate cost. As far as mechanical properties are concerned, the main target is to strike a balance of stiffness, strength, and toughness (Mohd Ishak et al., 2000) The several advantages offered by natural organic fillers have gradually stimulated the researchers to use these kinds of filler for plastics modification (Raj et al., 1990). For instance, the application of EFB fibre as filler or reinforcement for production of polymer filled composites has been reported by several researchers (Zaini et al., 1996; Rozman et al., 1998; Abdul Khalil et al., 2001). The incorporation of EFB into PP matrix has resulted in the improvement in tensile modulus, but the tensile strength, elongation at break, and impact strength decreased with increasing filler content. However, the application of compatibiliser and coupling agent has improved the tensile properties and impact strength of the composites (Rozman et al., 1998). The modification of EFB with anhydrides has improved the impact and flexural properties of polyester reinforced with EFB (Abdul Khalil et al., 2001). In terms of filler length, oil palm wood flour for all filler lengths has decreased the 67 mechanical properties of PP/Oil Palm wood flour composites with increasing filler contents (Zaini et al., 1996). Although EFB has been useful as a filler in the various polymers, the use of EFB in PVC has not been studied extensively. Most of the researchers using PVC have used wood flour, wood sawdust, and rice husk ash (Mautana et al., 2001; Sivaneswaran, 2002; Sombatsompop et al., 2003). Therefore, the objectives of this chapter are to discuss the impact and flexural properties of the PVC-U composites. 3.2 Experimental This section describes in detail the materials, filler preparations, samples preparation, density determination, impact and flexural testing, accelerated weathering testing and water absorption property. 3.2.1 Materials The suspension PVC resin used in this study, with solution viscosity K-value, and trade name MH-66 was supplied by Industrial Resin Malaysia (IRM) Sdn.Bhd. Tampoi, Johor, Malaysia. This homopolymer PVC is a medium molecular weight resin designed for general purpose, rigid and flexible applications. Its specifications are summarized in Table 3.1. 68 Table 3.1: Specifications of PVC suspension resin MH-66 Appearance White Powder Degree of polymerisation 1000 + 50 K-value 66 Specific gravity 1.4 Bulk density (g/cm3) 0.50+0.05 Volatile Matter (max) (%) 0.5 Foreign Matter (grain/100g) 15 Particle size (retained on 250 µ), (max) (%) 0.3 Table 3.2: Types of additives Additives Types Manufacturers Internal lubricant Calcium Stearate (CaSt) External lubricant Stearic Acid (HSt) Processing aid (PA-Acrylic) Pigment Titanium Dioxide (TiO2) Heat Stabilizer Tin (Sn) Elf Atochem Impact Modifier Acrylic Elf Atochem Chlorinated Polyethylene (CPE) Filler Oil Palm Empty Fruit Bunch (EFB) Sun Ace Koh Palm star Kaneka (M) Sdn. Bhd. Kromos Dupont Dow Elastomers Sabutek Sdn. Bhd. The additives used in the PVC formulations are shown in Table 3.2. IRM Sdn. Bhd also supplied most of the additives used in the formulations except for the impact modifiers and filler. Kureha Chemicals supplied core and shell impact modifier with a trade name of Paraloid KM355P. It consists of poly (methyl methacrylate) (PMMA) as a shell while poly (butyl acrylate) (PBA) acts as a core. The chlorinated polyethylene (CPE) for thermoplastic applications with trade name Tyrin 702P and containing 36 % chlorine content was supplied by Dupont Dow Elastomers. The EFB in the powder form was purchased from Sabutek Sdn. Bhd., Teluk Intan Perak. The 69 properties of EFB are shown in Table 2.3 (Chapter 2). The coupling agents used were Prosil 9234 and NZ 44 supplied by Kenrich Petrochemicals, Inc. The Prosil 9234 is silane-based with chemical name is n-octyltriethoxysilane, while NZ 44 is zirconatebased with chemical name is tris (2-ethylene-diamino) ethanolate. The chemical structures of Prosil 9234 and NZ 44 are shown in Figure 3.1. OC2H5 CH3CH2CH2CH2CH2CH2CH2CH2 Si OC2H5 OC2H5 (a) n-octyltriethoxysilane OC2H4-NH-C2H4-NH2 RO-Zr OC2H4-NH-C2H4-NH2 OC2H4-NH-C2H4-NH2 (b) tris(2-ethylene-diamino)ethanolate Figure 3.1 Chemical structure of (a) Prosil 9234 (b) NZ 44 3.2.2 Filler Preparations The EFB fillers were supplied in bulk with unknown sizes by the supplier. Thus, in this study a Restsch shaker AS200 was used to separate the EFB fillers into different sizes. The shaking time was set at 10 minutes. The filler size used in this study was less than 75 µm with filler distribution and filler morphology is shown in Figures 3.2 and 3.3, respectively. EFB filler distribution was determined by using 70 Malvern Instruments Mastersizer. EFB fillers were dried in an oven at 105 oC for about 24 hours to constant weight, the weight loss then being calculated. The original moisture content of the EFB filler was found to be around 5 %. Figure 3.2 EFB fillers distribution Figure 3.3 Morphology of EFB fillers (Magnification of 600x) 71 The temperature and time duration used for drying process in this study were referred from previous researchers who used natural fibre in their studies (Zaini et al., 1996; Hill and Abdul Khalil, 2000). The dried EFB fillers were placed in desiccators for cooling and were left in desiccators before dry blending with PVC and other additives to produce the dry blends of PVC-U compounds. Nano-precipitated calcium carbonate (NPCC) with filler size of less than 40 nm and extracted EFB filler size of less than 75 µm were also used as filler for the composites. NPCC filler was supplied by NanoMaterials Technology, Singapore. Extracted EFB fillers were also dried in an oven at similar conditions as unextracted EFB fillers before dry blending. The preparation of extracted EFB fillers was reported in detail in Chapter 5 (Section 5.2.2). The effect of NPCC and extracted EFB fillers on the mechanical properties of the composites was investigated to compare with unextracted EFB. This is a minor part of this study. 3.2.3 Coupling Agent Treatment of Filler Filler treatment techniques and dosages of the coupling agent were applied as advised by the manufacturers of the respective coupling agents. For zirconate coupling agent (NZ 44), the PVC resin itself was treated rather than the filler. This is contrary to silane coupling agent (Prosil 9234) where the EFB filler itself underwent coupling agent treatment prior to dry blending process. The dosage used for the Prosil 9234 coupling agent was 0.5 % by weight of the filler. The Prosil 9234 was diluted in ethanol to make up 10 % solution for better dispersion. The EFB filler was charged into a bench top tumbler mixer and coupling agent was added slowly over the period of 15 minutes to ensure uniform distribution 72 of the coupling agent. The filler was continuously mixed for another 30 minutes. The Prosil 9234-treated fillers were dried at 100 oC for about 2 hours. The dosage of the NZ 44 coupling agent is a slight different. The manufacturer recommended the dosage should be either 0.2 % or 0.5 % by weight of filler. This dosage has also been used by the previous researchers and they found that this dosage was able to improve the mechanical properties of filled composites (Ahmad Fuad et al., 1997 ; Sivaneswaran, 2002). Since with increasing filler content, the PVC fraction decreases correspondingly, the applied dosage of NZ 44 has to be adjusted. Thus, at lower filler contents (at which level, the PVC predominates), 0.2 % by weight of PVC was used. At higher filler contents (30 and 40 phr filler content), 0.5 % by weight of EFB was applied. The coupling agent was diluted in propanol to assist distribution of the coupling agent in the PVC resin prior to dry blending process. 3.2.4 Blend Formulations The PVC-U dry blend formulations are shown in Tables 3.3a-3.3i, which were based upon typical commercial PVC window formulations with some modifications. The basis of this formulation was also used by previous researchers (Wee, 2001; Sivaneswaran, 2002). The additives content were unchanged in the formulations except for EFB filler, CPE, and acrylic impact modifiers. Based on the preliminary study, the two-roll milled composite sheet became brittle as the filler content incorporated up to 50 phr. Therefore, in this study the EFB filler content was only varied from 10 to 40 phr. 73 74 75 76 77 78 79 80 81 82 3.2.5 Dry Blending The purpose of dry blending is to obtain good homogeneity of the blends. Before dry blending took place, the exact proportion of the resin and the additives were weighed and mixed appropriately. PVC dry blending with additives was done using a heavy-duty laboratory mixer. Each of PVC-U formulation as shown in Tables 3.3a-3.3i, except for formulations S45, S46, S48, was mixed at low speed mixing for 10 minutes before being collected. Low speed mixing was used to minimize the less dense of additives from floating in the air and sticking on the wall inside the column mixer. As a result, the percentage of homogeneity of dry blends was able to be increased. The PVC-U dry blends were poured into the sealed plastic bags and stored in the desiccators before used for two-roll milling process. The similar dry blending process was applied to the PVC-U dry blends filled with coupling agent-treated EFB, oil-extracted EFB and NPCC-200 fillers. 3.2.6 Two-Roll Milling The dry blended PVC-U were milled and sheeted using a two-roll mill as shown in Figure 3.4 with 10 minutes of milling time. The temperature used for the front and back roller was 165 oC. The milled sheet as shown in Figure 3.5 was then compression moulded. 83 Back Roller Dry Blended PVC-U Front Roller Figure 3.4 Figure: 3.5 Two roll milling of PVC-U dry blend at 165 oC Two-roll milled composite sheet 84 3.2.7 Compression Moulding A mould was designed and fabricated for compression moulding as shown in Figure 3.6. The mould was designed to have five cavities and the size of each sample was according to ASTM D256 (A) with thickness 3 mm. The two roll milled sheets were placed into a mould with five cavities at 180 oC and 120 kg/m2, 3 minutes for preheating, 5 minutes for compression moulding and 5 minutes for cooling before being removed. The water circulation was used for the cooling process. All the moulded specimens were then conditioned for 48 hours at 23 oC prior to testing. Top Hot Platen Mould Cavities Bottom Hot Platen Figure 3.6 compression A mould with five cavities placed in between the hot platens before 85 3.2.8 Density Determinations The density of EFB fillers was determined according to ISO 8962 using a glass pyknometer with some modification as reported by Ahmad Fuad et al. (1993). According to them, the standard describes only a method for determining the density of polymer dispersion and not the density of solid powder-like particles. The volume of the pyknometer was initially determined from the following equations; volume = Vp = mass density (M 2 − M1 ) ρw (3.1) (3.2) where M1 is the mass of cleaned and dried pyknometer, M2 is the mass of pyknometer filled with distilled water, in which both of them were weighed accurately to four decimal places on a Mettler Toledo Analytical Balance Model AX 201. Vp is the volume of the pyknometer and ρw is the density of water at the test temperature. Since the test was carried out at a constant laboratory temperature of 24 oC, the density of water was taken to be 0.9973 g/cm3. The pyknometer was dried thoroughly again and approximately 2 g of the dried EFB filler was placed into the bottle. The bottle was reweighed. It was then filled with distilled water, stoppered, and shaken slightly to disperse the particles. Care was taken to ensure absence of air bubbles. Excess water outside the bottle was wiped clean with absorbent paper. The volume of distilled water, Vw was determined as follows; Vw = (M 4 − M 3 ) ρw (3.3) 86 where, M3 is the mass of pyknometer containing the EFB, M4 is the mass of pyknometer containing the EFB with distilled water. Knowing the Vp and Vw, the volume of EFB filler, Vr was calculated as; Vr = VP − VW (3.4) Thus, the density of the EFB, ρf would be: ρf = (M 3 − M 1 ) Vr = ρw (M 3 − M 1 ) (M 3 − M 1 + M 2 − M 4 ) (3.5) (3.6) The experimental density of the PVC-U and composites was determined according to ASTM D792, Method A using a Mettler Toledo Model AX201 density balance. The theoretical density of PVC-U and composite, ρ c was calculated using the following equation; ρc = 1 ⎛ Wf ⎞ ⎛ WPVC ⎞ ⎛ Wadditives⎞ ⎟⎟ ⎟⎟ + ⎜⎜ ⎜⎜ ⎟⎟ + ⎜⎜ ⎝ ρf ⎠ ⎝ ρPVC ⎠ ⎝ ρadditives ⎠ (3.7) where, Wf ,WPVC and Wadditives are weight fractions of filler, PVC resin and additives in composite, respectively. ρ f , ρ m and ρ additivess refer to the density of filler, PVC resin and additives, respectively. 87 3.2.9 Impact Testing Specimens test bars for impact testing were moulded from the compressed sheet according to the size specified in ASTM D256 (A). The specimens were notched with a Davenport notch cutting apparatus. The notch depth was fixed at 2.6 + 0.02 mm with angle 45o.The Izod impact strength test was carried out on the notched specimens using the IMPats pendulum impact tester at impact velocity of 3.0 m/s and 90o swing angle using a 7.5 J hammer at room temperature. The Izod impact strength was calculated by dividing the indicator reading of impact energy by the cross sectional area of the specimen. All the reported values for the impact testing were the average of seven specimens. 3.2.10 Flexural Testing Specimen test bars for flexural testing were also moulded from the compressed sheet. The flexural test was conducted on the Instron Machine Model 5567 (static load cell of 30 kN) according to ASTM D790. The specimens, with dimensions 125 x 13x 3 mm3, were tested at crosshead speed of 3 mm/min at room temperature. The support span for the test was 51 mm. All the reported values for the flexural testing were the average of seven specimens The slope of the initial linear portion of the stress-strain curve represented flexural modulus. Flexural modulus was represented the resistance to elastic deformation under applied stress. The resistance to plastic deformation and failure under applied stress was represented by flexural strength. They were calculated according to the following equations; 88 Flexural modulus = L3 ∆ W 4bd 3 ∆ S (3.8) Flexural strength = 3WL 2 bd 2 (3.9) where W is the ultimate failure load (N), L is the span between the centres of support, b is the mean width of the specimens (m), d is the mean thickness of the specimens of the sample, ∆ W is the increment in load (N) and ∆ S is the increment in deflection. 3.2.11 Accelerated Weathering Testing The specimen test bars were tested using ATLAS UV accelerated weathering tester according to ASTM G53. The tester simulated the deterioration caused by sunlight and rain by means of ultraviolet light and condensation. The ATLAS UV accelerated weathering test apparatus consists of a corrosion resistant material test chamber enclosing eight fluorescent UV lamps, a heated water pan, test specimen racks and means for controlling and indicating operating times and temperatures. The specimens were mounted in the specimen racks with the test surfaces facing the lamps. The specimens were alternatively exposed to ultraviolet only (4 hours at 70 oC) and to condensation only (4 hours at 50 oC) in repetitive cycle without interruption, 24 hours a day. Condensation was produced by exposing the test surface to a heated, saturated mixture of air and water, while the reverse side of the test specimen was exposed to the cooling effect of ambient air. The total exposure time was 504 hours (3 weeks). The experimental was stopped at this total exposure time as the PVC-U specimens (as an indicator) suffered a change in colour (from white to 89 yellowish). The durability of each specimen was evaluated by comparing the mechanical properties of unweathered and weathered specimens. 3.2.12 Water Absorption Testing The specimen test bars were immersed into the distillate water for 60 days at room temperature according to ASTM D570. Water absorption was determined by weighing the sample before and after immersion at specific time intervals. The water absorption at any time, Mt was calculated from the following equation; Mt = Ww − Wd Wd (3.10) where Wd and Ww are the initial dry weight and the weight after immersion in water, respectively. The purpose of this test was to observe the influence of EFB filler and impact modifiers content on the degree of water absorption of the specimens. 3.2.13 Scanning Electron Microscopy Study Studies of the specimen impact-fractured surfaces were performed with a JEOL model JSM-6301F scanning electron microscope (SEM). A small portion of the fractured surface of impact-tested specimen was mounted on the copper stub and sputter-coated with a thin layer of gold to avoid electrostatic charging during examination. 90 3.2.14 Fourier Transform Infra-Red Analysis Fourier Transform Infra-Red (FTIR) was used to identify the related functional groups present in specimens. A small quantity of specimen was thoroughly mixed with dry potassium bromide powder (KBr) in a mortar and then pressed at pressure of 6 bars for 2 minutes to form a moisture-free KBr thin disc. The thin disc was placed in a sample cup of a diffuse reflectance accessory and was then scanned from 4000 cm-1- 370 cm-1 for 16 times to reduce the noise to signal ratio. All the spectra of specimens were obtained from the KBr technique, using a Perkin Elmer 2000 Infrared Spectrometer. 3.3 Results and Discussion This section discusses the density, impact strength, flexural properties, accelerated weathering properties and water absorption of the composites. The functional groups and impact-fractured surface morphologies of the composites obtained from FTIR and SEM, respectively, will also be discussed. 3.3.1 Density of the Filler and EFB-Filled PVC-U Composites The density of the EFB filler and PVC-U (So) found were 1.444 + 0.041 g/cm3 and 1.411 + 0.001 g/cm3, respectively. These values were within the known range of densities reported by other researchers (Titow, 1991; Sreekala and Thomas, 2003). The results showed that the density of EFB filler was slightly higher than that of PVC-U matrix. Consequently, the density of the EFB-filled PVC-U composites 91 increased as the filler content increased from 0 to 40 phr, as shown in Figure 3.7. All experimental densities obtained were lower than the theoretical densities calculated using Equation 3.7. The lowering of experimental densities was attributed to the presence of voids within the filled composites. The voids probably occurred in the matrix, at the fillermatrix interface, or within the filler bundles. The presence of voids in natural fibrefilled polymer composites has also been noted by other researchers who found that the void content increased as the fibre content increased (Hill et al., 2000; Klaric et al., 2000). The SEM micrograph in Figure 3.8 shows the presence of voids at the fibrematrix interface in the composites. The hydrophilic nature of the EFB fillers might have caused them to pick up moisture, which evaporated from the filler-matrix interface during processing. This resulted in the formation of voids, which increased with increasing EFB filler content. The density of the composites, however, still increased with increasing filler content because the effect of voids was less significant than the effect of an increase in filler content. Theoretical 1.428 1.427 1.426 Density (g/cm3) 1.424 1.422 1.425 1.423 1.422 1.420 1.426 1.421 1.419 1.418 1.417 1.415 1.416 Experimental 1.414 1.412 1.411 1.410 1.408 0 10 20 30 40 Filler content (phr) Figure 3.7 The theoretical and experimental densities of EFB-filled composites (S9 - S12) with respect to EFB filler contents 50 92 Void at fibre-matrix interface Figure 3.8 SEM micrograph of impact-fractured surface of S11 (Magnification 2000x) 3.3.2 Mechanical Properties This section discusses the impact strength and flexural properties of the composites with respect to filler content, impact modifiers concentration, and coupling agent-treated filler content. The size of EFB filler incorporated into PVC-U matrix was less than 75 µm. Two types of impact modifier used were acrylic and CPE. The effect of other fillers such as NPCC and extracted filler on the impact strength and flexural properties was also investigated as a minor part of this study. Tensile properties are important parameters of composites, however, in structural applications frequently involve flexural and impact stress modes. Therefore, only flexural and impact properties were investigated. 3.3.2.1 Notched Izod Impact Strength The impact strength represents the amount of energy that can be absorbed by a material when subjected to sudden load before failure occurs. The influence of EFB 93 filler, acrylic, and CPE impact modifier contents on the impact strength is discussed in this section. A. Impact Modified PVC-U Compounds Table 3.4 shows the impact strength of impact-modified PVC-U compounds at various contents of acrylic and CPE impact modifier. The result of unmodified PVCU (S0) was included as a comparison. The impact strength of impact-modified compounds rapidly increased when 3 phr impact modifier added in PVC-U and increased continuously as the impact modifier content increased from 3 to 12 phr. This is typical of toughened plastics containing impact modifier. It can be seen that at 3 phr, acrylic was able to produce approximately 59 % increment in impact strength of PVC-U, while the CPE gave 57 % increment. It was observed that all samples containing 3 phr and lower of impact modifier broke completely into two halves during impact testing. There were mixtures in the failure mode for the samples containing 6 phr of both impact modifiers. However, at 9 phr, the samples underwent hinge breaks (Figure 3.9) i.e. the samples did not break completely. This result suggested that the brittle-ductile transition occurred at around 6 phr for both the acrylic and CPE impactmodified PVC-U samples. This transition also resulted in the experimental errors at 6 phr for both impact modifiers as shown in Table 3.4 were higher than other impact modifier contents. From the number of unbroken samples at 6 phr and impact strength values at 3 phr, results showed that acrylic was a more effective impact modifier than CPE at low impact modifier content. Table 3.4 also shows that the optimum impact strength of acrylic-impact modified compound occurred at 9 phr. This is because the impact strength was relatively constant with further increased to 12 phr. However, the impact strength of CPE-impact modified compound kept increasing with increasing CPE content, even 94 though at 9 phr all the samples underwent hinge breaks. The difference in the effectiveness (optimum proportion) showed that the toughening mechanism of both impact modifiers was definitely different (Robinovic, 1983; Hiltner et al., 1984; Lutz, 1993). The difference in the effectiveness of both impact modifiers can be observed at the impact-fractured surface of the samples. SEM micrographs in Figure 3.10 shows that the failure modes of impact-fractured surface of S0 which is brittle in nature. However, as shown in Figures 3.11 and 3.12, the addition of impact modifier transformed the brittle mode of fracture surface to the ductile mode, where the impact-fractured surface of impact-modified compound was smoother than unmodified ones. For impact-modified compounds, the impact fractured surface of acrylic-impact modified (Figure 3.11) was smoother than CPE-impact-fractured surface (Figure 3.12). It indicates that the acrylic is more effective in enhancing impact strength of compound compared to CPE. Table 3.4: Effect of impact modifiers content on the impact strength of PVC-U compounds S0 Acrylic-Impact Modified Complete Impact Broken Strength (no. of 2 (kJ/m ) sample) 8.04 + 0.85 7 S0 CPE-Impact Modified Complete Impact Broken Strength (no. of 2 (kJ/m ) sample) 8.04 + 0.85 7 S1 19.65 + 1.81 7 S5 18.55 + 0.70 7 S2 93.13 + 30.30 3 S6 48.60 + 41.44 5 S3 109.53 + 5.99 0 S7 99.60 + 11.10 0 S4 109.03 + 4.23 0 S8 119.58 + 3.80 0 Sample Sample 95 Figure 3.9 Hinge breaks of impact-modified PVC-U samples (S3) According Hiltner et al. (1984), the most effective effect in impact modification of rigid polymers was when the rubbery impact modifier dispersed in the polymeric matrix as discrete small particles. The ability of the discrete small particles of acrylic to well disperse in the continuous matrix was contributed to the better effect on the impact strength of acrylic-impact modified compound. This is clearly not the case for CPE-impact modified compound where the discrete particles of CPE have only a very small effect on the impact strength compared to that achieved by a continuous network of CPE (Hiltner et al., 1984). 96 Figure 3.10 SEM micrograph of impact-fractured surface of S0 (Magnification 200x) Figure 3.11 200x) SEM micrograph of impact-fractured surface of S3 (Magnification 97 Figure 3.12 SEM micrograph of impact-fractured surface of S7 (Magnification 200x) B. EFB-Filled Impact-Modified PVC-U Composites Figures 3.13 and 3.14 show that with increasing EFB filler contents, the impact strength of both unmodified and impact-modified EFB-filled composites decreased gradually. This observation was expected for filled composites and has also been reported by other researchers (Raj et al., 1990; Zaini et al., 1996; Ismail and Mega, 2001). Furthermore, as shown by the impact-fractured surface of composites in Figures 3.15 through 3.18, the incorporation of EFB filler into PVC-U matrix reduced the ductile behaviour of the matrix by making the composites more brittle. 98 14 2 Impact strength (kJ/m ) 13 0 phr 3 phr 12 11 6 phr 9 phr 12 phr 10 9 8 7 6 5 0 10 20 30 40 50 Filler content (phr) Figure 3.13 Effect of filler content on the impact strength of EFB-filled composites (S9-S12) and EFB-filled acrylic-impact modified composites (S13-S28) 0 phr 3 phr 13 2 Impact strength (kJ/m ) 12 6 phr 9 phr 12 phr 11 10 9 8 7 6 5 0 10 20 30 40 50 Filler content (phr) Figure 3.14 Effect of filler content on the impact strength of EFB-filled composites (S9-S12) and EFB-filled CPE-impact modified composite (S29- S44) 99 Figure 3.15 SEM micrograph of the impact-fracture surface of composite incorporated with 10 phr filler (Magnification 500x) Figure 3.16 SEM micrograph of the impact-fracture surface of composite incorporated with 20 phr filler (Magnification 500x) 100 Figure 3.17 SEM micrograph of the impact-fractured surface of composite incorporated with 30 phr filler (Magnification 500x) Filler bundle Figure 3.18 SEM micrograph of the impact-fractured surface of composite incorporated with 40 phr filler (Magnification 500x) Table 3.5 shows the percentage of impact strength reduction of composites increased as the filler content increased from 10 to 40 phr. The percentage of reduction of the EFB-filled acrylic-impact modified composites (S13-S28) was higher 101 than EFB-filled CPE-impact modified composites (S29-S44). The results indicated that the impact strength of S13-S28 was higher at 10 phr but lower at 40 phr compared to S29-S44. This means that the acrylic was efficient in enhancing the impact strength of composites at low filler content while CPE was efficient at high filler content. Table 3.5 also shows that EFB-filled composites (S9-S12) have the lowest percentage of reduction (20 %) due to their lowest impact strength values at all filler contents compared to S13-S28 and S29-S44. Table 3.5: Percentage of impact strength reduction of EFB-filled impact-modified composites as filler content increased from 10 to 40 phr Impact modifier content (phr) 0 Impact strength reduction (%) EFB-filled acrylic-impact EFB-filled CPE-impact modified composites modified composites 20 20 3 36 17 6 45 37 9 42 39 12 49 36 There are several reasons why the impact performance of composites decreased as the filler content increased. Firstly, there is the detrimental effect of the fillers caused by the volume they take up. Fillers, unlike the matrix, are incapable of dissipating stress through the mechanism known as shear yielding prior to fracture. They may also hinder the local chain motions of the polymer molecules to shear yield (Stevenson, 1995). As a result, the ability of the composites to absorb energy during fracture propagation decreased. Figures 3.10 and 3.15-3.18 clearly showed that the impact fractured surface of composites changed from smooth to rough as the filler content increased from 0 to 40 phr. In other words, the ductile portion contributed by PVC-U matrix reduced, and the failure mode became more brittle as the filler content increased. 102 Secondly, composites materials with satisfactory mechanical properties can only be achieved if there is good dispersion and wetting of the fillers in the matrix, giving rise to strong interfacial adhesion. However, this is not the case when EFB fillers are used in PVC. Although the polarity of EFB fillers made them capable of forming a physical interaction with polar PVC (Figure 3.19), it was relatively weak interaction compared to the strong filler-filler interaction caused by hydrogen bonding. Therefore, EFB fillers had a greater tendency to agglomerate among themselves into filler bundles, which consequently, lowered the area of contact with the matrix. This can be explained with the aid of the SEM micrographs shown in Figures 3.17-3.18. The EFB fillers tended to cling together in bundles and to resist dispersion as the filler content increased. The filler bundles can be observed in the SEM micrograph distributing unevenly throughout the matrix. H H C C δ+H O - δ O - H H Cl δ + Hδ C C OH O HOH2C δ + δ Cl δO + Hδ δ+ H OH Cellulose O HOH2C PVC C - OH O O Figure 3.19 H H C C - δ Cl δ H δδ+ O H + H H O O HOH2C Polar-polar interactions between PVC molecule and cellulose molecule The presence of moisture as mentioned in Section 3.3.1 might interfere with the physical bonding between fillers and PVC, and potentially acted as a lubricant at the filler-matrix interface (Mengeloglu, 2001; Sombatsompop et al., 2004). The presence of moisture in the composites was also detected in the DTG curves for all composites as mentioned in Chapter 5. Other researchers had also detected the moisture content in the composites, even though careful care had been taken prior to drying of cellulose before mixing (Sapieha et al., 1989; Thiebaud et al., 1997; Peters and Bruch, 2001). This moisture was also able to vaporize during composite 103 preparation to form a number of small voids in the PVC matrix. The presence of voids in the matrix might also become the potential sites for crack initiation. These factors increased the weakness of interfacial adhesion between the filler and the matrix, and became the potential sites for crack growth as the inability of fillers to support the stress transfer from the polymer matrix decreased (Rozman et al., 1998). As shown in Figures 3.20 and 3.21, the weakness of interfacial adhesion resulted in debonding and pulled of filler bundles in the composites. Filler debonding Figure 3.20 SEM micrograph of filler debonding on the impact-fractured surface of composite incorporated with 30 phr filler of S11 (Magnification 900x) 104 Fillers pulled-out areas Figure 3.21 SEM micrograph of filler bundles pulled-out area on the impact- fractured surface composite incorporated with 30 phr filler of S11 (Magnification 1000x) The toughening of composites by introducing impact modifiers was also investigated. Figures 3.22 and 3.23 show the impact strength of the impact-modified composites improved with increasing impact modifier content. Without impact modification, the deleterious effect of incorporating EFB fillers to the PVC-U matrix was apparent because unmodified composites exhibited a much lower capability for absorbing impact energy. For instance, the impact strength of composite filled with 10 phr filler (7.1 kJ/m2) was inferior compared to that of PVC-U compound (8.0 kJ/m2). 105 2 Imapct Strength (kJ/m ) 14 10 phr 20 phr 12 30 phr 40 phr 10 8 6 4 0 Figure 3.22 3 6 9 Acrylic content (phr) 12 15 Effect of acrylic content on the impact strength of EFB-filled acrylic- impact modified composites (S0-S4 and S9-S28) 13 2 Impact strength (kJ/m ) 12 11 10 phr 10 20 phr 30 phr 9 40 phr 8 7 6 5 4 0 Figure 3.23 3 6 9 CPE content (phr) 12 15 Effect of CPE impact modifier content on the impact strength of EFB- filled CPE-impact modified composites (S5 -S8 and S29-S44) 106 The lowering of impact strength of composites due to incorporation of EFB filler was well recovered by the introduction of impact modifiers into the composites. Impact modifiers have the capability to compensate the detrimental effect caused by the filler that is lowering the yield stress of PVC-U matrix. Therefore, PVC-U matrix was able to undergo shear yielding rather than fracture when subjected to sudden load. Through this mechanism, the composites were able to suppress catastrophic failure (brittle failure). As a result, the enhancement of impact strength with increasing impact modifier content was expected. Mengeloglu (2001) has also reported a similar result, in which the addition of impact modifiers efficiently reduced brittle crack propagation in the wood flour-filled PVC composite. It is interesting to observe the effectiveness of the impact modifier in enhancing the impact strength of EFB-filled composites. Table 3.6 shows that the percentage of impact strength increment for the composites decreased as the filler content increased, as the presence of EFB fillers in composites dominated. The results showed that the ability of both the impact modifiers to compensate the detrimental effect of filler reduced as the filler content increased. As reported in Table 3.5, the difference in the effectiveness of both the impact modifiers can also be observed in Table 3.6. Acrylic is believed to be more effective at low filler content, whereas CPE at high filler content. This is due to the discrete acrylic particles were able to disperse evenly throughout the matrix at low filler contents and effectively improved the impact strength of composites. However, the dispersion of acrylic particles was restricted by EFB at high filler contents and thus acrylic unable to improve the impact strength of composites effectively. Figure 3.24 illustrates that in general, CPE is more efficient and effective as impact modifier than acrylic at filler content of 20 phr and higher. Gerlach (1998) reported the similar advantage of CPE at higher filler content. The plausible reason for this observation is the adhesion between filler and matrix might be improved by the addition of CPE. It may be elucidated in terms of polarity between hydroxyl (OH) groups on the EFB filler surface and chlorine (Cl) atoms of PVC resin. The addition of CPE impact modifier, which contains 36 % of Cl atoms, into the PVC resin may 107 increase the Cl content and polarity of CPE-impact modified PVC-U. Because of that, the physical interaction between OH groups and Cl atoms took place and consequently produced better wetting of filler surface by PVC-U matrix. Therefore, the filler-matrix adhesion of composite was improved. Table 3.6: Percentage of impact strength increment for EFB-filled impact-modified composites as the impact modifier content increased from 0 to 12 phr EFB filler content (phr) 10 Impact strength increment (%) EFB-filled acrylic-impact EFB-filled CPE-impact modified composites modified composites 78 69 20 37 38 30 31 51 40 14 35 16.0 10 phr 14.0 20 phr 30 phr 40 phr 2 Impact Strength (kJ/m ) 12.0 10.0 8.0 6.0 4.0 2.0 0.0 Acrylic- CPE3 phr 3 phr Figure 3.24 Acrylic- CPE6 phr 6 phr Acrylic9 phr CPE9 phr Acrylic- CPE12 phr 12 phr Effects of acrylic and CPE impact modifier on the impact strength of EFB-filled impact-modified composites 108 3.3.2.2 Flexural Properties Flexural strength is the ability of a material to withstand bending forces applied perpendicular to its longitudinal axis. Flexural modulus is a measure of stiffness during the first or initial part of the bending process. Similar to the section on impact strength, the influences of EFB filler, acrylic and CPE impact modifier contents on the flexural properties of the composites will be discussed in this section. A. Impact Modified PVC-U Compounds Figures 3.25 and 3.26 show the effect of acrylic and CPE impact modifiers on the flexural properties of impact-modified compounds. While the impact strength was improved, the flexural properties were adversely affected by the addition of acrylic and CPE. As the impact modifier contents increased from 0 to 12 phr, acrylic generated 28 % and 22 % decrement in flexural modulus and flexural strength of PVC-U, respectively, while CPE caused 29 % and 31 % decrement in flexural modulus and flexural strength, respectively. The rubbery phase introduced by the impact modifiers decreased both the stiffness and flexural strength of the compounds. This was probably due to the softening effect generated by the impact modifier, which lowered the yield stress, thereby increasing the energy absorption and consequently lowered the modulus and strength of the PVC-U compound. In other words, impact modifier increased the local chain motion of the PVC molecules that enabled them to shear yield. Several researchers also reported that the flexural properties of PVC modified with impact modifiers decreased as the impact modifier contents increased (Deani et al., 1986; Lutz, 1993; Wimolmala et al., 2001). 109 Flexural modulus (MPa) 3800 Acrylic 3600 CPE 3400 3200 3000 2800 2600 2400 0 3 6 9 12 Impact modifier content (phr) Figure 3.25 Effect of impact modifiers content on the flexural modulus of impact modified compounds (S0-S8) Flexural strength (MPa) 90.0 Acrylic 85.0 80.0 CPE 75.0 70.0 65.0 60.0 55.0 50.0 45.0 40.0 0 Figure 3.26 3 6 9 Impact modifier content (phr) 12 Effect of impact modifiers content on the flexural strength of impact modified compounds (S0-S8) 110 Although the flexural properties decreased linearly with increasing impact modifier contents, the flexural properties of acrylic-impact modified PVC-U compound were found higher than CPE-impact modified PVC-U compound. This result indicates that acrylic which comprised a PBA core and a PMMA shell has less softening effect than CPE for two possible reasons. The first reason being that the PMMA shell had higher stiffness and was more compatible with PVC matrix. The compatibility between impact modifier and matrix has resulted in the reduction in free volume of matrix, which increased the modulus of compound (Wimolmala et al., 2001). Besides that, compatibility also provided good interfacial bonding between the phases to allow efficient stress transfer and thus enhanced the flexural properties of acrylic-impact modified compounds. Secondly, PBA rubber core formed separate domains and do not soften the PVC matrix whereas CPE was soft and flexible and was also less compatible in the PVC matrix, thus softening it more aggressively (Deanin and Chuang, 1986). B. EFB-Filled Impact-Modified PVC-U Composites The flexural modulus increased whereas the flexural strength decreased for all composites as the filler content increased. The influence of filler content on the flexural properties of the composites is shown in Figures 3.27-3.30. Modulus or the material stiffness increased significantly with increasing filler content for all composites. As highlighted by other researchers, the enhancement of modulus depends on a number of factors, such as filler aspect ratio, filler modulus, and filler content. This result indicates that even though EFB fillers have a low aspect ratio (5:1), they are able to impart a significant improvement in stiffness by hindering the movement of PVC molecules (Woodhams et al., 1984). The fact that the modulus increased while the strength decreased with increasing filler content was in agreement with the trend observed in other fibre-filled thermoplastics (Rozman et al., 2003). As discussed in the impact strength section, agglomeration of fillers in the matrix, moisture pick-up in the fillers and interfacial defects could be the causes of 111 the reduction in flexural strength (Figures 3.28 and 3.30). Furthermore, the irregular shape and a relatively low aspect ratio of EFB filler increased the inability of fillers to support the stress transmitted from the polymer matrix. This effect is amplified when the proportion of EFB filler increased (Ichazo et al., 2000; Rozman et al., 2000). A similar result has been reported for works on the cellulose-filled polymer composites (Raj et al., 1990; Rozman et al., 1998; Azman Hassan and Sivaneswaran, 2001). Flexural modulus (MPa) 5000 0 phr 3 phr 6 phr 9 phr 12 phr 4500 4000 3500 3000 2500 2000 0 10 20 30 40 Filler content (phr) Figure 3.27 Effects of filler and acrylic content on the flexural modulus of EFB- filled acrylic-impact modified composites (S9-S28) 112 Flexural strength (MPa) 100.0 0 phr 3 phr 6 phr 9 phr 12 phr 90.0 80.0 70.0 60.0 50.0 40.0 0 10 20 30 40 Filler content (phr) Figure 3.28 Effects of filler and acrylic content on the flexural strength of EFB- filled acrylic-impact modified composites (S9-S28) 5000 0 phr 3 phr 6 phr 9 phr 12 phr Flexural modulus (MPa) 4500 4000 3500 3000 2500 2000 0 10 20 30 40 Filler content (phr) Figure 3.29 Effects of filler and CPE content on the flexural modulus of EFB- filled CPE-impact modified composites (S29-S44) 113 100.0 0 phr 3 phr 6 phr 9 phr 12 phr Flexural strength (MPa) 90.0 80.0 70.0 60.0 50.0 40.0 0 10 20 30 40 Filler content (phr) Figure 3.30 Effects of filler and CPE content on the flexural strength of EFB-filled CPE-impact modified composites (S29-S44) Tables 3.7 and 3.8 show the percentage of modulus increment and the percentage of flexural strength reduction, respectively, for the unmodified and impact modified composites as the filler contents increased from 0 to 40 phr. It can be seen that the percentage of modulus increment for EFB-filled composites (S9-S12), EFBfilled acrylic-impact modified composites (S13-S28), and EFB-filled CPE-impact modified composites (S29-S44) was relatively similar. The result reveals that all composites underwent a similar degree of increment as the filler contents increased. On the other hand, the percentage of flexural strength reduction of S13-S28 was higher than S29-S44. This is due to the flexural strength values of S13-S28 were higher than S29-S44 at 0 phr and relatively similar at 40 phr of filler. As mentioned earlier in Section 3.3.2.2 A, the softening effect aggressively generated by CPE compared to that by acrylic resulted in the flexural strength of S29-S44 was lower than S13-S28. 114 Table 3.7: Percentage of flexural modulus increment for EFB-filled impact modified composites as the filler content increased from 0 to 40 phr Impact modifier content (phr) 0 Flexural modulus increment (%) EFB-filled acrylic-impact modified PVC-U composites EFB-filled CPE-impact modified PVC-U composites 22 22 3 24 22 6 26 21 9 24 27 12 25 24 Table 3.8: Percentage of flexural strength reduction for EFB-filled impact modified composites as the filler content increased from 0 to 40 phr Impact modifier content (phr) 0 Flexural strength reduction (%) EFB-filled acrylic-impact modified PVC-U composites EFB-filled CPE-impact modified PVC-U composites 18 18 3 26 15 6 16 16 9 21 12 12 20 11 Figures 3.31-3.34 show that the flexural properties of composites reduced significantly as the impact modifiers contents increased from 0 to 12 phr. As mentioned previously, the softening effect introduced by the impact modifier was the main factor, which resulted in the reduction of flexural properties. Figures 3.31 and 3.33 show that this effect on the flexural modulus became less significant as the filler content in the impact modified composites increased. This was due to the separation among the impact modifier particles increased and thus reduced the effectiveness of impact modifier to soften the matrix. As a result, composites containing 40 phr filler 115 produced the highest values of modulus as the impact modifier content increased form 0 to 12 phr. Meanwhile, in case of flexural strength, the adverse effect was observed. The flexural strength of composites at 40 phr with the increase of impact modifier content from 0 to 12 phr has the lowest series of value (Figures 3.32 and 3.34). The possible factors severely affected the flexural strength of the composites were the weak filler-matrix interface, and agglomeration and randomly dispersed fillers. 5000 0 phr 10 phr 20 phr 30 phr 40 phr Flexural Modulus (MPa) 4500 4000 3500 3000 2500 2000 0 3 6 9 12 Acrylic content (phr) Figure 3.31 Effects of filler and acrylic content on the flexural modulus of EFB- filled acrylic-impact modified composites (S13-S28) 116 Flexural strength (MPa) 100.0 0 phr 10 phr 20 phr 30 phr 40 phr 90.0 80.0 70.0 60.0 50.0 40.0 0 3 6 9 12 Acrylic content (phr) Figure 3.32 Effects of filler and acrylic content on the flexural strength of EFB- filled acrylic-impact modified composites (S13-S28) Flexural modulus (MPa) 5000 0 phr 4500 10 phr 20 phr 30 phr 40 phr 4000 3500 3000 2500 2000 0 Figure 3.33 3 6 CPE content (phr) 9 12 Effects of filler and CPE content on the flexural modulus of EFB-filled CPE-impact modified composites (S29 -S44) 117 100.0 0 phr 10 phr 20 phr 30 phr 40 phr Flexural strength (MPa) 90.0 80.0 70.0 60.0 50.0 40.0 0 6 3 9 12 CPE content (phr) Figure 3.34 Effects of filler and CPE content on the flexural strength of EFB-filled CPE-impact modified composites (S29 - S44) Tables 3.9 and 3.10 show the percentage of flexural modulus and flexural strength reduction, respectively. Table 3.9 shows that the degree of modulus reduction for both acrylic-impact modified and CPE-impact modified filled composites was relatively similar as the impact modifier contents increased from 0 to 12 phr. Meanwhile, Table 3.10 shows that the percentage of flexural strength reduction of EFB-filled CPE-impact modified composites was higher than EFB-filled acrylic-impact modified composites. This is due to the softening effect aggressively generated by CPE compared to that by acrylic. 118 Table 3.9: Percentage of flexural modulus reduction for EFB-filled impact modified composites as the impact modifier increased from 0 to 12 phr EFB filler content (phr) 0 Flexural modulus reduction (%) EFB-filled acrylic-impact EFB-filled CPE-impact modified composites modified composites 28 28 10 29 28 20 27 27 30 24 26 40 26 27 Table 3.10: Percentage of flexural strength reduction for EFB-filled impact modified composites as the impact modifier increased from 0 to 12 phr EFB filler content (phr) 0 Flexural strength reduction (%) EFB-filled acrylic-impact EFB-filled CPE-impact modified composites modified composites 22 31 10 34 35 20 30 32 30 29 36 40 24 25 Based on Figures 3.35 and 3.36, the effect of both impact modifiers on the flexural properties of the composites can be compared. Due to the softening effect aggressively generated by CPE, thus EFB-filled CPE-impact modified PVC-U composites have lower values of flexural properties compared to the EFB-filled acrylic-impact modified PVC-U composites. 119 5000 0 phr Flexural modulus (MPa) 4500 10 phr 20 phr 30 phr 40 phr 4000 3500 3000 2500 2000 1500 1000 500 0 Acrylic-3 CPE- 3 Acrylic-6 CPE- 6 Acrylic-9 CPE- 9 Acrylic- CPEphr phr phr phr phr phr 12 phr 12 phr Effects of acrylic and CPE impact modifier on the flexural modulus of Figure 3.35 EFB-filled impact modified composites 100 Flexural strength (MPa) 90 0 phr 10 phr 20 phr 30 phr 40 phr 80 70 60 50 40 30 20 10 0 Acrylic- CPE- Acrylic- CPE- Acrylic- CPE- Acrylic- CPE3 phr 3 phr 6 phr 6 phr 9phr 9 phr 12 phr 12 phr Figure 3.36 Effects of acrylic and CPE impact modifier on the flexural strength of EFB-filled impact modified composites 120 3.3.2.3 EFB-Filled Acrylic-Impact Modified PVC-U Composites Treated With Coupling Agents The purpose of adding coupling agents to the composites is to improve the interfacial adhesion bond strength and wetting of the filler by the matrix. The purpose of this study is to investigate whether coupling agents would lead to any improvement in the impact properties of composites or vice versa. The coupling agents were only applied to the EFB-filled PVC-U composites and EFB-filled acrylic-impact modified PVC-U composites. As mentioned earlier in Section 3.2.4, the blend formulations used in this study are based on the PVC window profile formulation, which is suitable for outdoor applications, mainly for blend formulations added with acrylic impact modifier. This is because the acrylic is reported more weather resistance compared to CPE (Robinovic, 1983). Therefore, the acrylic-impact modified composites are only used for coupling agents treatment. A. Impact Strength Figures 3.37 and 3.38 show that the impact strength of EFB-filled and EFB– filled acrylic-impact modified composites, respectively, was generally higher compared to those that were not treated with Prosil 9234 and NZ 44 coupling agent. Both coupling agents were able to enhance the impact strength of the composites by means of improving the filler-matrix adhesion. 121 20.0 18.0 10 phr 16.0 20 phr 30 phr 40 phr Impact strength (kJ/m2) 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0 Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil0 phr treated Figure 3.37 3 phr treated 6 phr treated 9 phr treated 12 phr treated Effects of EFB filler and acrylic content on the impact strength of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S49-S68) 16.0 14.0 10 phr 20 phr 30 phr 40 phr 2 Impact strength (kJ/m ) 12.0 10.0 8.0 6.0 4.0 2.0 0.0 Acrylic-NZ 44- Acrylic-NZ 44- Acrylic-NZ 44- Acrylic-NZ 44-Acrylic-NZ 440 phr treated 3 phr treated Figure 3.38 6 phr treated 9 phr treated 12 phr treated Effects of EFB filler and acrylic content on the impact strength of EFB-filled acrylic-impact modified composites treated with NZ 44 (S69-S88). 122 As mentioned in Section 2.4.3.1 (Chapter 2), the reaction of the tetrafunctional organo-metallic compounds based on silicone and zirconium was believed to have occurred in two steps (Ahmad Fuad et al., 1997). First, the alkoxy groups in the coupling agent underwent hydrolysis process. Water from the hydrolysis might come from the surface humidity of filler (in case of the Prosil 9234 treatment), and from the surface or in the PVC resin (in the case of NZ44 treatment) (Ahmad Fuad et al., 1997). The hydroxyl groups were supposed to link to EFB filler through the formation of hydrogen bonds with the hydroxyl groups at the EFB filler surface. Then Si-O or Zr-O cross-links formed between the filler surface and the adjacent functional groups via the condensation reaction with the elimination of water. The scheme reaction in Figure 2.12 (Chapter 2) can be used to describe the reaction of Prosil 9234 with fillers, while that of NZ 44 is shown in Figure 3.39. Step 1 hydrolis X-Zr(OR)3 X-Zr(OH)3 Step 2 OH OH OH X-Zr + OH + OH O O OH H OH H X-Zr O OH O Filler surface X-Zr OH OH OH Filler surface X-Zr H OH H Condensation OH X-Zr OH OH Figure 3.39 O + O OH OH X-Zr OH O Filler surface O Filler surface O X-Zr X-Zr O Bonding mechanism of NZ 44 coupling agent to the filler’s surface 123 SEM micrographs of the impact-fractured surface of Prosil 9234 treated-filled composites (Figures 3.40a-3.40b) and NZ 44 treated-filled composites (Figures 3.40c3.40d) also show that the formation of interaction between treated-filler and matrix took place in the composites. This is because the treated-fillers pulled out from the matrix but a fair amount of polymeric residues remained on the fillers. It means that the matrix wetted the treated fillers in composites extensively. Besides that the treated fillers in composites in Figures 3.40a and 3.40c show only slight fillers were pulled out and formed a more topographically even surface compared to fracture surface of the untreated filler-filled composite in Figure 3.17. The formation of ether linkage (-O-) or interaction between treated fillers and coupling agents also shown by the FTIR spectra of both treated EFB-filled composites in the Figure 3.46 (discuss in detail at part D in this section). The remaining chain of coupling agents (Figure 3.39) adhered to the PVC with the help of a Van der Waals type of weak interaction. By this means, coupling agent made a link between EFB filler and PVC. As a result, the impact strength of treated-filled composites improved as compared to the untreated ones. Polymeric residues remained on the fillers (a) 124 Polymeric residues remained on the filler (b) (c) 125 Polymeric residues remained on the fillers (d) Figure 3.40 SEM micrographs of the impact fractured-surface of composites (a) Prosil 9234 treatment (S51) (Magnification 100x) (b) S51 (Magnification 2000x) (c) NZ 44 treatment (S71) (Magnification 500x) and (d) S71 (Magnification 2000x) Table 3.11 shows the comparison on the effect of Prosil 9234 and NZ 44 on the impact strength of composites. It may be worth noted that the performance of both coupling agents in enhancing the impact strength of composites is difficult to be distinguished. This is due to the impact strength values obtained in Table 3.11 was relatively similar except for composites containing 12 phr acrylic at filler contents of 10 phr and 20 phr, whereby in general, the impact strength of composites treated with Prosil 9234 was higher than composites treated with NZ 44 coupling agent. This may be due to the techniques used for treating the composites with both coupling agents in this study were different. As reported in Section 3.2.3, Prosil 9234 was used to treat the EFB fillers surfaces whereas NZ 44 was used to treat the PVC matrix rather than the fillers. Both techniques, the filler surface treatment and matrix resin treatment, were reported to improve the interaction between fibres and matrix at the interface regions in the composites (Maldas et al., 1991). However, the results show that the treatment of fillers, which alters the chemical nature of fillers surfaces, has been considered to be more promising method in enhancing the filler-matrix interaction compared to the matrix treatment (Maldas et al., 1990). As a result, Prosil 9234 has brought a slight higher improvement in the impact strength compared to the 126 composites treated with NZ44. The effectiveness of Prosil 9234 has also been reported by Ahmad Fuad et al. (1997) who used these two types of coupling agents in the rice husk ash-polypropylene composites. Table 3.11: Impact strength of EFB –filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 Filler 0 0 Content Prosil NZ44 (phr) 9234 7.7 7.7 10 + + 0.6 0.6 7.0 6.9 20 + + 0.5 0.4 6.3 6.3 30 + + 0.4 0.5 5.7 5.7 40 + + 0.6 0.6 B. Acrylic Impact Modifier Content (phr) 3 3 6 6 9 9 12 12 Prosil Prosil Prosil Prosil NZ44 NZ44 NZ44 NZ44 9234 9234 9234 9234 9.4 9.4 9.1 11.3 11.4 11.4 15.9 11.9 + + + + + + + + 0.9 0.9 0.4 1.3 0.9 1.4 2.5 0.3 7.1 7.7 7.9 8.0 8.6 9.0 9.7 8.8 + + + + + + + + 0.5 0.5 0.3 0.4 0.4 0.6 0.2 0.9 6.6 6.8 7.2 7.0 7.5 7.4 7.6 7.5 + + + + + + + + 0.3 0.8 0.3 0.4 0.6 0.3 0.5 0.3 6.6 6.4 6.7 6.5 6.9 6.6 7.0 6.9 + + + + + + + + 0.6 0.4 0.2 0.3 0.4 0.4 0.3 0.3 Flexural Properties Figures 3.41-3.44 show that the addition of both coupling agents improved the flexural properties, particularly for the flexural modulus of composites. This is because the coupling agents were able to enhance the interfacial adhesion between filler and matrix. It could be also that the initially strong filler-filler interaction caused by the intermolecular hydrogen bonding has also been weakened, which leads to the better dispersion of fillers in the matrix. The flexural modulus of composites treated with coupling agents improved significantly as shown by Figures 3.41 and 3.43. Although these two coupling agents 127 were able to enhance the stiffness of the composites, the flexural modulus (Table 3.12) indicates that Prosil 9234 has brought a greater stiffness to composites than composites treated with NZ44. Similar results on the stiffness improvement of sisal fibre reinforced polystyrene composites upon treatment with benzoyl chloride have been reported (Manikandan Nair et al., 1996). 6000 10 phr 20 phr 30 phr 40 phr 50003.32: Effect of EFB filler and acrylic content on the flexural modulus of EFBFigure Flexural modulus (MPa) filled acrylic-impact modified PVC-U composites treated with Prosil 9234 coupling 4000 agent (S9 to S28 and S49 to S68). 3000 2000 1000 0 Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil0 phr treated Figure 3.41 3 phr treated 6 phr treated 6 phr treated 12 phr treated Effect of EFB filler and acrylic contents on the flexural modulus of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S9-S28 and S49-S68) 128 100.0 10 phr 90.0 20 phr 30 phr 40 phr Flexural strength (MPa) 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic- Prosil- Acrylic Prosil0 phr treated 3 phr treated 6 phr treated 9 phr treated 12 phr treated Figure 3. 42 Effect of EFB filler and acrylic contents on the flexural strength of EFB-filled acrylic-impact modified composites treated with Prosil 9234 (S9-S28 and S49-S68) 6000 10 phr 20 phr 30 phr 40 phr Flexural modulus (MPa) 5000 4000 3000 2000 1000 0 Acrylic-NZ44- Acrylic- NZ44- Acrylic-NZ44- Acrylic-NZ44- Acrylic- NZ440 phr treated Figure 3.43 3 phr treated 6 phr treated 9 phr treated 12 phr treated Effect of EFB filler and acrylic contents on the flexural modulus of EFB-filled acrylic-impact modified composites treated with NZ44 (S9-S28 and S69-S88) 129 100.0 10 phr 90.0 20 phr 30 phr 40 phr Flexural strength (MPa) 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Acrylic- NZ44- Acrylic- NZ44- Acrylic- NZ44- Acrylic- NZ44- Acrylic- NZ440 phr treated Figure 3.44 3 phr treated 6 phr treated 9 phr treated 12 phr treated Effect of EFB filler and acrylic content on the flexural strength of EFB-filled acrylic-impact modified composites treated with NZ44 (S9-S28 and S69-S88) Based on the flexural strength values of composites as shown in Table 3.13, both coupling agents have a similar ability in enhancing the flexural strength of composites. However, in general, Prosil 9234 showed slightly better ability in enhancing the flexural strength of composites compared to NZ44. Although NZ 44 has a similar purpose as Prosil 9234 to impart a better wetting, dispersion and interaction, the flexural modulus (Table 3.12) and flexural strength (Table 3.13) showed that Prosil 9234 was more effective than NZ 44. It indicates that NZ 44 insufficient to form a strong bridge between filler and matrix at the interface regions in the composites. On the other hand, Prosil 9234 capable to generate a strong bonds with fillers to give a better performance of the composites (Wu et al., 2000). 130 Table 3.12: Flexural modulus of EFB–filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 EFB Filler 0 Content Prosil (phr) 9234 4196 + 10 80 4208 20 + 72 4475 30 + 252 4636 40 + 150 Acrylic Impact Modifier Content (phr) 3 3 6 6 9 9 12 12 Prosil Prosil Prosil Prosil NZ44 NZ44 NZ44 NZ44 NZ44 9234 9234 9234 9234 3859 3583 3481 3573 3302 3145 3007 2910 2766 + + + + + + + + + 91 69 122 218 90 84 41 81 31 4272 4041 3823 3605 3574 3566 3172 3196 3058 + + + + + + + + + 147 67 97 61 51 182 99 74 38 4505 4252 4111 3774 3823 3632 3451 3306 3287 + + + + + + + + + 161 68 32 59 139 110 91 94 89 4698 4426 4362 4103 3965 3804 3729 3571 3392 + + + + + + + + + 196 181 63 93 92 79 83 168 42 0 Table 3.13: Flexural strength of EFB–filled acrylic-impact modified composites treated with Prosil 9234 and NZ 44 EFB Filler 0 Content Prosil (phr) 9234 84.9 10 + 3.0 82.2 20 + 4.3 72.9 30 + 3.1 67.2 40 + 7.5 Acrylic Impact Modifier Content (phr) 3 3 6 6 9 9 12 12 Prosil Prosil Prosil Prosil NZ44 NZ44 NZ44 NZ44 NZ44 9234 9234 9234 9234 85.0 78.6 73.9 74.0 68.2 65.0 64.4 62.8 59.8 + + + + + + + + + 2.4 2.1 1.9 1.7 1.9 1.6 2.9 1.7 1.7 77.7 79.9 67.8 70.9 70.3 65.3 58.5 59.3 53.7 + + + + + + + + + 6.4 1.9 3.5 3.0 1.6 6.1 1.6 3.1 3.2 76.5 69.6 64.2 60.0 63.6 55.3 55.8 55.3 52.0 + + + + + + + + + 4.8 5.0 1.3 3.4 2.3 3.2 1.0 4.4 1.2 69.8 62.3 60.7 57.9 63.4 51.4 51.8 51.0 50.3 + + + + + + + + + 7.3 2.7 4.7 5.0 2.6 2.0 1.6 4.0 3.6 0 131 C. Relative Modulus of Composites A number of models have been used to calculate the relative modulus of composites. The most commonly used was discussed in the Chapter 2. It was found that the relative elastic modulus (Ec/Ep) of the composites varied with volume fraction, φf as shown in Tables 3.14-3.18. The relative elastic modulus of untreated and treated composites increased with increasing volume fractions. The modulus values of treated composites were higher than modulus of untreated ones. These tables also show the comparison between the experimental relative modulus values and the predicted relative modulus values obtained from the predictive models, which take into account the shape, packing fraction, and adhesion between the filler and matrix polymer. Table 3.14: Comparison between the predicted and the experimental of relative modulus of composites EFB-Filled Composites Experimental Values EFB (phr) φ f 10 Predicted Values Einstein Kerner (perfect adhesion Untreated (S9 - S12) Prosil 9234 (S49 -S52) NZ44 (S69 - S72) Einstein (without adhesion) 0.08 1.07 1.20 1.08 1.08 1.20 1.19 1.07 20 0.15 1.15 1.20 1.18 1.15 1.38 1.39 1.14 30 0.21 1.18 1.28 1.23 1.21 1.53 1.59 1.19 40 0.27 1.29 1.32 1.26 1.27 1.68 1.83 1.24 LewisNielsen 132 Table 3.15: Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 3 phr of acrylic EFB-Filled Acrylic-Impact Modified Composites Experimental Values Untreated NZ44 Prosil 9234 (S13 to (S73 to (S53 to S56) S16) S76) Predicted Values Einstein Kerner (perfect adhesion) EFB (phr) φ f 10 0.08 1.08 1.11 1.08 1.08 1.20 1.19 1.07 20 0.15 1.18 1.26 1.19 1.15 1.38 1.39 1.14 30 0.21 1.23 1.32 1.28 1.21 1.53 1.59 1.19 40 0.26 1.26 1.38 1.36 1.26 1.65 1.78 1.23 Einstein (without adhesion) LewisNielsen Table 3.16: Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 6 phr of acrylic EFB (phr) EFB-Filled Acrylic-Impact Modified Composites Experimental Values Predicted Values Einstein Einstein Untreated Prosil 9234 NZ44 (without (perfect Kerner (S17 - S20) (S57-S60) (S77 - S80) adhesion) adhesion) φ f LewisNielsen 10 0.08 1.14 1.21 1.12 1.08 1.20 1.19 1.07 20 0.15 1.20 1.22 1.21 1.15 1.38 1.39 1.14 30 0.20 1.26 1.27 1.29 1.20 1.50 1.56 1.18 40 0.25 1.35 1.39 1.34 1.25 1.63 1.74 1.23 Table 3.17: Comparison between the predicted and the experimental of relative modulus of filled acrylic-impact modified composites at 9 phr of acrylic EFB-Filled Acrylic-Impact Modified Composites EFB (phr) φ f 10 Experimental Values Predicted Values Untreated (S21 - S24) Prosil 9234 (S61 - S64) NZ44 (S81 - S84) Einstein (without adhesion) Einstein (perfect adhesion) Kerner LewisNielsen 0.08 1.08 1.16 1.11 1.08 1.20 1.19 1.07 20 0.14 1.21 1.32 1.17 1.14 1.35 1.36 1.13 30 0.20 1.23 1.34 1.27 1.20 1.50 1.56 1.18 40 0.25 1.31 1.40 1.38 1.25 1.63 1.74 1.23 133 Table 3.18: Comparison between the predicted and the experimental relative flexural modulus of filled acrylic-impact modified composites at 12 phr of acrylic EFB-Filled Acrylic-Impact Modified PVC-U Composites Experimental Values Prosil Untreated NZ44 9234 (S25 -S28) (S85 - S88) (S65 - S68) Predicted Values Einstein Kerner (perfect adhesion) EFB (phr) φ f 10 0.07 1.05 1.16 1.10 1.07 1.18 1.17 1.06 20 0.14 1.16 1.27 1.21 1.14 1.35 1.36 1.13 30 0.19 1.24 1.31 1.30 1.19 1.48 1.52 1.17 40 0.24 1.33 1.42 1.35 1.24 1.60 1.70 1.22 Einstein (without adhesion) LewisNielsen A few assumptions have been made in order to apply the predictive models. Assumptions such as perfect adhesion, spherical shape of EFB, well-dispersed and random packing have used for Lewis-Nielsen model (Equation 2.6). The maximum packing volume fraction of filler, φm used was 0.64 (for random packing), which was suggested by Lewis-Nielsen to calculate ψ (a reduced concentration term) in their studies. The Poisson’s ratio of PVC was assumed 0.39 (Fried, 2003) and it was substituted in Kerner’s model (Equation 2.5) to calculate the value of relative modulus of composites whereas in the Lewis-Nielsen’s model this Poisson’s ratio was used to calculate the value of A (Einstein coefficient). The value of B (relative shear modulus of filler and polymer) in Lewis-Nielsen’s model was calculated by assuming that the modulus of EFB was 9000 MPa, which was also used by Sreekala and Thomas (2003) in their studies. In this study, the value of A and B obtained by calculation was 1.23 and 0.413, respectively. Based on the Equation 3.10, the volume fraction of filler in the composites for each formulation in this study was calculated. Mρ φ = f f f Mρ c c (3.10) where Mf and Mc are mass of EFB filler and composite in the formulation, respectively. The volume fraction of filler, φf is a volume occupied by filler in the 134 composites, which varies with EFB filler content and impact modifier content in the formulation. Experimental values for untreated composites in Tables 3.14-3.18 were in good agreement with predicted values of the Einstein model without adhesion. For treated composites, experimental values were in between predicted values of the Einstein’s model without adhesion and predicted values of Einstein with perfect adhesion and with Kerner’s model. These results indicate that the interaction between fillers and matrix of treated composites has been improved marginally by the addition of Prosil 9234 and NZ 44. Because the experimental values of relative modulus of composites treated with Prosil 9234 higher than composites treated with NZ 44, it proves that Prosil 9234 has a good ability than NZ 44 to enhance the flexural modulus of composites. Tables 3.14-3.18 also show that predicted values of the Lewis-Nielsen’s model which is modified from Kerner’s model and applicable for lower aspect ratio exhibited slightly lower than experimental values of untreated composites. This is because the value of B was calculated based on the EFB modulus obtained from the literature (Sreekala and Thomas, 2003), and not from the experimental work. This means that the EFB modulus used to calculate B might not represent the actual value of EFB modulus in this study. In actual case, there is a larger variation of EFB modulus due to the non-uniformity of the fibres geometry (Sreekala and Thomas, 2003). D. FTIR Spectra of Untreated and Treated EFB-Filled PVC-U Composites A typical FTIR spectrum of PVC-U compound (Figure 3.45) shows two characteristics bands; a very broad one at 617 and 687 cm-1, probably due to C-Cl stretching, and a sharp band at 1425 cm-1 due to C-H groups (Yagoubi et al., 1994). Some oxygenated structures were also detected in the PVC chain or/and in the 135 additives chain, as evidenced by the absorption band at 1727 cm-1, which corresponds to the carbonyl (C=O) groups (Matuana et al., 2001). The FTIR spectra of untreated EFB-filled composite (S11) (Figure 3.45b) showed the increase in the absorption band intensity at 1727 cm-1. Besides that, two new bands were also observed at 1597 cm-1 and 1163 cm-1. All these bands are characteristic of the main components of wood (EFB filler). The presence of cellulose was obvious from the intense and sharp band appearing at 1727 cm-1. This band was due to the stretching vibration of carbonyl groups (C=O). This stretching might also come from carbonyl (C=O) linkage in acetyl ester groups in hemicelluloses, aldehyde in lignin, extractives, and PVC’s additives (Matuana et al., 2001). The two absorption bands at 1597 cm-1 and 1425 cm-1 of untreated composite (S11) were indicative of the presence of lignin, an aromatic compound, and attributed to the C=C vibrations of the benzene ring. The absorption bands at 1163 cm-1 and 1045 cm-1 were assigned to the ether linkages (C-O-C) from lignin or hemicellulose. Figure 3.45 FTIR spectra of (a) PVC-U compound (S0) (b) composite (S11) 136 The FTIR spectra of untreated and treated EFB-filled PVC-U composites are shown in Figure 3.46. The untreated composite (S11), Prosil 9234-treated composite (S51), and NZ 44-treated composite (S71) samples were used in this analysis. The absorption band regions at 3200-3600 cm-1 of untreated and treated composites were probably due to absorption vibration of hydroxyl groups (-OH) and of secondary amine groups (N-H). The FTIR spectra band of S51 (Figure 3.46) at 1045 cm-1 was more intense compared to S11 (Figure 3.46) and it may be assigned to the Si-O stretching vibration. The presence of this functional group was supported by the bands at 3418 cm-1, and 772 cm-1- 835 cm-1, due to the OH stretching and OH bending vibration, respectively, from the functional group of Si-OH.. The stretching vibration of aryl alkyl ether or ether linkage (C-O-C) was also found not to differ greatly from that of the Si-O system. Similar to Si-O, the absorption band at 1045 cm-1 was assigned to the C-O-C symmetric stretching vibration from the filler. This band was supported by C-O-C asymmetric stretching at 1247 cm-1 (Sarani et al., 2001). The presence of these bands indicated that the Prosil 9234 coupling agent probably interacted with EFB fillers. NZ 44 coupling agent also interacted with EFB fillers due to the similarity of its FTIR spectra (Figure 3.46) with that of S51. The Si-O and Si-OH functional groups of S51 might be replaced and assigned to the Zr-O and Zr-OH. The formation of ether linkage also found in this composite. The new bands at 1513 cm-1 and 798-850 cm-1 also emerged on the both spectra of treated composites. The new band at 1513 cm-1 was probably due to NH (amine) symmetric bending which suggested the formation of amine salt (Matuana et al., 1997) while band at 798-850 cm-1 was due to N-H out of plane bending. These bands were also supported by the secondary amine groups (N-H) stretching vibration, which observed at band 3418 cm-1. There is a possibility that the amino groups, particularly from NZ 44 reacted with PVC. This has also been reported by Matuana et al. (1997). 137 Figure 3.46 FTIR spectra of (a) filled composite (S11) (b) Prosil 9234-treated filled composite (S51) (c) NZ 44-treated filled PVC-U composite (S71) 3.3.2.4 Effect of Accelerated Weathering on the Impact and Flexural Properties This part of study was carried out to observe the effect of accelerated weathering on the impact strength and flexural properties of PVC-U compound (S0), impact modified PVC-U compound (S1-S8), EFB-filled PVC-U composites (S9-S12) and EFB-filled impact modified PVC-U composites (S21-S24 and S37-S40). The impact modifier content was varied from 3 to 12 phr for S1 to S8 and was fixed at 9 phr for S21-S24 and S37-S40. The EFB filler content was varied from 10 to 40 phr for all samples, except samples S1-S8. All samples were weathered for 504 hours. 138 A. Impact Strength Figure 3.47 shows the impact strength of PVC-U compound (S0), acrylicimpact modified (S1-S4) and CPE-impact modified (S5-S8) compounds insignificantly decreased after being exposed to the accelerated weathering. This is due to the values of impact strength of weathered samples fell within the experimental errors. Unweathered CPE-impact modified compounds (S0 and S5- S8) 180.0 Impact strength (kJ/m2) 160.0 Weathered acrylic-impact modified compounds 140.0 120.0 100.0 Unweathered acrylic-impact modified compounds (S0 and S1-S4) 80.0 60.0 Weathered CPEimpact modified 40.0 20.0 0.0 0 3 6 9 12 Impact modifier content (phr) Figure 3.47 Impact strength of unweathered and weathered impact modified compounds Although the impact strength decreased insignificantly, Table 3.19 shows that the impact strength of CPE-impact modified compounds greatly decreased compared to that of acrylic-impact modified compounds due to the effect of photo-oxidation degradation, which was initiated by UV irradiation in the presence of oxygen and/or water (Gardiner et al., 1997). The possible reason is the amount of labile chlorine atom, which releases from the molecular chains during photo-oxidation degradations, was increased in these compounds. Consequently, the number of active site for 139 propagation reaction of degradation increased (Figure 3.49). The decrease in impact strength was only marginal for acrylic-impact modified compounds (S1-S4). This is due to the acrylic was able to tolerate the devastating effects of accelerated weathering and therefore these compounds were hardly susceptible to the effects. The excellent weathering performance of acrylic-impact modified compound compared to CPE has also been reported by Robinovic (1983). Table 3.19 : Impact strength reduction for impact-modified compounds after accelerated weathering for 504 hours Impact modifier content (phr) 0 Impact strength reduction (%) Acrylic-impact modified CPE-impact modified compounds compounds 3 3 3 14 25 6 2 59 9 -1 17 12 -1 6 Notation: negative sign (-) shows the value of weathered samples was higher than unweathered. Figure 3.48 also shows the decrease in impact strength of EFB-filled composites (S9-S12) after 504 hours exposure to the accelerated weathering. The effect of photo-oxidative degradation process, which was initiated by UV irradiation in the presence of oxygen and moisture in the amorphous regions of PVC, was the main cause for the decrease in impact strength of weathered filled composites (Gardiner et al.,1997). The rate of permeability of oxygen and moisture is greater in amorphous region because the molecular packing is less dense that the degradation process is able to take place. Some photo-oxidation processes lead to scission of the polymer chain, while others to intermolecular crosslinking. Intermolecular crosslinking defined as the joining of two chains or more, (Sarmoria and Enrique, 2004) causes material embrittlement (Gardiner et al., 1997). Most researchers agreed that crosslinking and chain scission contribute to the deterioration of the impact strength and other mechanical properties (Raghi et al., 2000; Matuana et al., 2001; Leong et al., 2003). 140 Unweathered EFB-filled composites (S9-S12) 9.0 Impact strength (kJ/m2) 8.0 7.0 6.0 5.0 Weathered EFB-filled composites 4.0 3.0 2.0 1.0 0.0 0 10 20 30 40 Filler content (phr) Figure 3.48 Impact strength of unweathered and weathered EFB filled composites The suggested reaction scheme for photo-oxidative degradation (similar to photo degradation) is shown in Figure 3.49 (Matuana et al., 2002; Owen, 1989; Torikai and Hirose, 1999). The action of UV radiation on PVC produces the singletexcited state of polyene sequence (Figure 3.49), which resulted from the absorption of conjugated double bonds and carbonyl groups. The most likely bond to be broken in the excited singlet polyene sequence is the allylic C-Cl linkage, due to its low dissociation energy. This result in the chlorine radical being released from the excited singlet polyene sequence and produces a polyenyl radical. The radical then reacts with the oxygen to produce peroxy radicals on the chain and leads to chain scission (degradation) or/ and crosslinking of PVC molecular chains. 141 Cl CH 2 CH CH=CH CH 2 n hv Cl CH 2 CH CH=CH CH 2 n * CH 2 C H CH=CH n CH 2 + Cl + O2 O O H2C H and/or Two in chain radicals may recombine to form crosslink chain * CH 2 C H CH=CH + n CH 2 * * CH 2 CH CH=CH O C * C O n CH 2 CH 2 C CH=CH CH 2 n H H CH 2 C CH=CH CH 2 C CH=CH CH=CH n CH 2 Chain Scission n n CH 2 CH 2 H Crosslinking Figure 3.49 Suggested reaction scheme for the accelerated photo-oxidation degradation The lignin of the incorporated EFB filler primarily broke down under UV irradiation leading to the generation of the chromophoric functional groups (impurities) such as carbonyl and hydroperoxy radicals, which accelerated the degradation of PVC. The increase in the intensity of the C=O stretching bands at around 1735- 1731 cm-1 as shown in FTIR spectra (Figure 3.59 in part D) provided good evidence of the incorporation of carbonyl groups into the PVC-U matrix during processing. The presence of these groups explained the pronounced degradation observed in the filled composites as compared to unfilled composites (Matuana et al., 2001). As a result, the brittleness of the filled composites was increased. Table 3.20 shows that the brittleness of the filled composites increased as the filler content increased. 142 Table 3.20: Percentage of impact strength reduction for filled impact-modified composites after weathered for 504 hours EFB filler content (phr) 0 EFB-filled composites 3 10 4 1 13 20 4 2 4 30 6 2 15 40 6 3 9 Impact strength reduction (%) EFB-filled acrylic-impact EFB-filled CPE-impact modified compounds modified compounds -1 17 Notation: Negative sign (-) shows the value of weathered samples was higher than unweathered. Figure 3.50 shows that the addition of 9 phr of acrylic or CPE into EFB filled PVC-U composites decreased the impact strength of weathered filled impact-modified composites. The reduction in impact strength of filled CPE-impact modified composites (S37-S40) was greater than filled acrylic-impact modified composites (S21S24). Similar to the performance of acrylic-impact modified PVC-U compound (S1-S4) in Figure 3.38, the results in Figure 3.50 and Table 3.20 indicated that the filled composites added with acrylic were also more stable upon photo-oxidation than filled CPE-impact modified composites because they were less susceptible UV irradiation. The reduction of impact strength of filled CPE-impact modified composites in Table 3.20 obviously shows that the combination EFB filler with CPE increased the propagation reaction of photo-oxidation degradation in the weathered composites. The synergistic combination of chlorine radicals from PVC and CPE with carbonyl, hydroperoxy and other radicals from fillers might attribute to the greater reduction in impact strength of filled CPE-impact modified composites. 143 12.0 Unweathered EFB-filled acrylic-impact modified composites (S21- S24) Impact strength (kJ/m2) 11.0 10.0 Unweathered EFB-filled CPE-impact modified composites (S37- S40) Weathered EFB-Filled acrylic-impact modified composites 9.0 8.0 Weathered EFB-filled CPE-impact modified composites 7.0 6.0 5.0 0 10 20 30 40 Filler content (phr) Figure 3.50 Impact strength of unweathered and weathered EFB-filled impact modified PVC-U composites B. Flexural Properties Similar to impact strength, the reduction in flexural modulus of acrylic-impact modified compounds (S1-S4) (Figure 3.51), CPE-impact modified compounds (S5-S8) (Figure 3.51), EFB-filled composites (S9-S12)(Figure 3.53), filled acrylic-impact modified composites (S21-S24) (Figure 3.56) and filled CPE-impact modified composites (S37 to S40) (Figure 3.56) were also observed. The reasons for the reduction of flexural modulus are similar to the reasons as discussed on the reduction of impact strength. However, some of weathered samples of CPE-impact modified compounds at 3 phr to 9 phr CPE (Figure 3.51 and Table 3.21) and filled CPE-impact modified composites at 10phr EFB filler (Figure 3.55 and Table 3.22) slightly increased in the flexural modulus. The results indicate that samples containing CPE might be more susceptible to UV irradiation, which resulted in PVC molecular chains undergoing crosslinking process during photo-oxidation degradation. The influence of 144 crosslinked chains in samples might cause a slight increase in flexural modulus (Leong et al., 2003). The flexural strength for all weathered samples is shown in Figures 3.52, 3.54, and 3.56. Unlike impact strength and flexural modulus, the flexural strength of weathered samples was higher than that of unweathered samples. The increment of flexural strength of weathered samples is shown in Tables 3.21 and 3.22. The compounds and filled composites containing CPE impact modifier exhibited higher series of values in the increment of flexural strength after 504 hours exposure in accelerated weathering. These results show that CPE probably has an ability to accelerate the crosslinking of PVC molecular chains to occur predominantly during photo-oxidation degradation due to the increase of chlorine atom concentration in the samples. The enhancement in the properties of weathered samples is similar with the findings of other researchers (Radghi et al., 2000) who reported that the tensile strength of PVC/lignin blend increased after undergoing 129 hours and 480 hours of accelerated weathering. Similar to the increase of flexural modulus of some weathered samples as discussed earlier, the increase of flexural strength might be caused by the crosslinked chains in weathered samples. However, this evidence was not conclusive because of the fluctuation and small magnitude of increase in the flexural strength and flexural modulus as well, which fell within the experimental error. Furthermore, other factors such as filler agglomerations and filler-matrix interaction bonding as mentioned earlier in Section 3.3.2.2 also influenced the flexural properties results. 145 Unweathered acrylic-impact modified compounds (S1 -S4) 3600 Flexural modulus (MPa) 3400 Unweathered CPE-impact modified compounds (S5 - S8) 3200 3000 2800 Weathered acrylic-impact modified compounds 2600 Weathered CPE-impact modified compounds 2400 2200 2000 6 3 0 9 12 Impact modifier content (phr) Figure 3.51 Flexural modulus of unweathered and weathered impact compounds Weathered acrylic-impact modified Weathered CPE-impact modified compounds compounds Flexural strength (MPa) 100 80 60 Unweathered acrylic-impact modified compounds (S1-S4) 40 Unweathered CPE-impact modified compounds (S5 - S8) 20 0 0 3 6 9 Impact modifier content (phr) Figure 3.52 compounds Flexural strength of unweathered and weathered impact modified 12 146 Table 3.21: Percentage of flexural modulus reduction and percentage of flexural strength increment of the impact-modified compounds after 504 hours exposure to accelerated weathering Impact modifier content (phr) 0 Flexural modulus reduction (%) Acrylic-impact CPE-impact modified modified compounds compounds 13 13 Flexural strength increment (%) Acrylic-impact CPE-impact modified modified compounds compounds 13 13 3 7 0 4 18 6 5 -2 10 22 9 1 -1 7 13 12 3 7 8 0 Notation: Negative sign (-) shows the value of weathered samples was higher than unweathered. Table 3.22: Percentage of flexural modulus reduction and percentage of flexural strength increment of the filled composites and filled impact-modified composites after 504 hours exposure to accelerated weathering. EFB filler content (phr) 0 Flexural modulus reduction (%) EFB-filled EFB-filled CPEacrylicEFB-filled impact impact composites modified modified composites composites 13 2 -1 Flexural strength increment (%) EFB-filled EFB-filled CPEacrylicEFBimpact impact filled modified modified composites composites composites 13 7 13 10 9 4 -2 4 13 16 20 10 10 3 0 5 10 30 0 10 11 11 2 7 40 15 8 23 0 4 15 Notation: Negative sign (-) shows the value of weathered samples was higher than unweathered. 147 Flexural modulus (MPa) 5000 Unweathered EFB-filled composites (S9 -S12) 4000 3000 2000 Weathered EFB-filled composites 1000 0 0 Figure 3.53 10 20 Filler content (phr) 30 40 Flexural modulus of unweathered and weathered EFB-filled PVC-U composites Flexural strength (MPa) 100.0 Weathered EFB-filled composites 80.0 60.0 Unweathered EFB-filled composites (S9- S12) 40.0 20.0 0.0 0 Figure 3.54 composites 10 20 Filler content (phr) 30 Flexural strength of unweathered and weathered EFB-filled PVC-U 40 148 Flexural modulus (MPa) 4000 Unweathered EFB-filled acrylicimpact modified composites (S21- S24) 3500 Unweathered EFB-filled CPE-impact modified composites (S37- S40) 3000 2500 Weathered EFB-Filled CPEimpact modified composites Weathered EFB-filled acrylicimpact modified composites. 2000 1500 0 10 20 30 40 Filler content (phr) Flexural modulus of unweathered and weathered EFB-filled impact Figure 3.55 modified composites Weathered EFB-filled acrylic-impact modified composites Flexural strength (MPa) 80.0 70.0 Weathered EFB-filled CPEimpact modified composites 60.0 Unweathered EFB-filled acrylic impact modified composites (S21-S24) 50.0 Unweathered EFB-filled CPE-impact modified composites (S37-S40) 40.0 30.0 0 10 20 30 40 Filler content (phr) Figure 3.56 Flexural strength of unweathered and weathered EFB-filled impact modified composites 149 C. Visual Inspection The effect of accelerated weathering on the appearance of all the samples surfaces was also studied by visual inspection of the exposed sides of samples (Figure 3.57). Samples of PVC-U compound (S0) and impact modified PVC-U compounds (S1-S4 and S5-S8) suffered discolouration (from white to yellowish). This was due to the presence of chromophoric groups, particularly carbonyl functional groups, which formed on PVC under the effect of the UV irradiation. Meanwhile, for unmodified and impact modified filled composites (S9-S12, S21-S24 and S37-S40), the initial colour was dark brown, which later changed to light brown after exposure to UV irradiation. In terms of discolouration level, the composites samples exhibited greater discolouration than that of unfilled samples. This showed that the addition of carbonyl functional groups contributed by EFB fillers within the PVC-U matrix caused the discolouration process to take place aggressively. The discolouration also provided an additional evidence of photo-oxidation of the samples. No cracks have been noticed after this period of weathering for all samples. Figure 3.57 Appearance of some samples surface before and after exposed to UV irradiation for 504 hours 150 D. FTIR Spectra The FTIR spectroscopy has been used to study changes in the surface chemistry of polymer after weathering (Matuana et al., 2001; Stark and Matuana, 2004). FTIR spectroscopy was used as a complementary technique to monitor functional groups of weathered compounds and composites samples. The FTIR spectra of unweathered and weathered samples for PVC-U compound (S0), acrylicimpact modified compound (S3), CPE-impact modified compound (S7) and EFB-filled composite (S11) are shown in Figures 3.58-3.61, respectively. It can be seen that the formation of conjugated carbonyl groups (C=O stretching) was easily recognized at 1735 cm-1 for weathered PVC-U sample (Figure 3.58). This absorption band (intensity) may be assigned to functional groups containing carbon-carbon double bonds (R2-C=C) conjugated to carbonyl groups (Matuana et al., 2001). Similar to the weathered PVC-U compound, the absorption bands at 1735 cm-1 of weathered composite (Figure 3.59), acrylic-impact modified (Figure 3.60) and CPE-impact modified sample (Figure 3.61) were also due to carbonyl groups. Although the carbonyl groups absorption band also emerged at 1735 cm-1 for unweathered composite and unweathered impact-modified samples but the peak of this band was blunted. The photo-oxidation degradation was also reflected by the formation of unsaturated vinyl groups (RCH=CH2) in the absorption of 1653-1648 cm-1 (C=C stretching) in the FTIR spectra of weathered samples as shown in Figures 3.58-3.61 (Matuana et al., 2001; Wan et al., 2004). The vinyl group formation was also supported by the new region of 970 cm-1 for all weathered samples (Stark and Matuana, 2004). Based on this peak, similar to the carbonyl formation, the increase in vinyl group concentrations suggested that the chain scission took place not only at the polymer chains but also at the EFB filler. Furthermore, the absorption bands at the regions of 1430 cm-1 in the FTIR spectra of weathered samples, which were associated with =C-H out of plan bending were also 151 indicative of photo degradation of samples upon exposure to UV irradiation (Pavia et al., 1979). (b) Figure 3.58 compounds FTIR spectra of (a) unweathered and (b) weathered PVC-U 152 (a) (b) Figure 3.59 FTIR spectra of (a) unweathered and (b) weathered-EFB-filled composites (a) (b) Figure 3.60 FTIR spectra of (a) unweathered and (b) weathered-acrylic-impact modified composites 153 (a) (b) Figure 3.61 FTIR spectra of (a) unweathered and (b) weathered-CPE-impact modified composites 3.3.2.5 Effect of NPCC on the Impact Strength and Flexural Properties The study was conducted to observe the possibility of using NPCC as impact modifier for the EFB-filled PVC-U composites. A preliminary study on the effect of NPCC on the PVC-U compound was first carried out. A. Impact Strength Figure 3.62 illustrates that the impact strength has significantly enhanced upon the addition of 10 phr NPCC. No further increase was observed with further incorporation of NPCC fillers. Recall from Section 3.3.2.1 (B), that the incorporation of EFB fillers decreased the impact strength of the composites. This preliminary result 154 indicates that NPCC has the potential to be used as an impact modifier in the PVC. The ability of nano calcium carbonate (NCC) in enhancing the impact strength of the composites has also been reported by Chen et al.(2004). They found that the impact strength of PVC/NCC compound was significantly enhanced after the incorporation of 10 phr NCC. 16.0 NPCC-filled composites (S89-S91) Impact strength (kJ/m2) 14.0 12.0 10.0 EFB-filled composites (S9-S12) 8.0 6.0 4.0 2.0 0.0 0 Figure 3.62 10 20 30 Filler content (phr) 40 50 Effects of EFB filler and NPCC on the impact strength of composites The second part of the study the investigation was to observe whether the NPCC fillers can also significantly enhance the impact strength of the composite when used together with EFB fillers. Besides that, the previous results of impact strength of EFB-filled composites (S9-S12) and EFB-filled impact modified composites containing 3 phr acrylic (S13-S16) and 3 phr CPE (S29-S32) were also replotted in Figure 3.63 in order to compare the ability of NPCC filler with acrylic and CPE impact modifier in enhancing the impact properties of filled composites. The results shows that the impact strength of EFB-filled composites containing 10 phr NPCC (S92-S95) and 20 phr NPCC (S96-S99), respectively, decreased drastically as the EFB filler contents increased from 10 to 40 phr. Although all the filled composites show a similar trend on the impact strength with increasing EFB 155 filler contents, the rate of decrease of filled composites incorporated with NPCC fillers was larger than that of the acrylic and CPE-impact modified filled composites. The results also show that when hybrid fillers are used, the impact strength was lower than when only EFB filler was used. The results also revealed that NPCC filler unable to perform as impact modifier in EFB-filled composites. The inability of stress to distribute uniformly upon impact and inability of matrix to shear yield before fracture due to the increase of filler proportion in the matrix are the plausible reasons for the decreasing in the impact strength of hybrid filled composites (Marthur and Vanderheiden,1986). EFB filled composites (S9-S12) EFB-filled composites added with 3 phr acrylic (S13-S16) EFB-filled composites added with 3 phr CPE (S29-S32) EFB filled composites added with 10 phr NPCC (S92-S95) EFB-filled composites added with 20 phr NPCC (S96-S99) Impact strength (kJ/m2) 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 0 10 20 30 40 50 EFB filler content (phr) Figure 3.63 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the impact strength of EFB-filled composites 156 B. Flexural Properties Figure 3.64 shows the effect of NPCC filler on the flexural modulus of filled composites. The result on the effect of EFB filler (from Section 3.3.2.2 B) is also replotted as a comparison. The results revealed that EFB filler was more effective in increasing the stiffness of composites at all filler contents. The ability of NPCC and EFB fillers to restrict the mobility of PVC molecular chains by reducing the free volume of the PVC molecules may be attributed to the increasing filled composites stiffness. However, the result revealed that EFB was more effective in increasing the stiffness of composites compared to NPCC at all filler contents. The lower aspect ratio of NPCC could not restrict effectively the mobility of PVC molecular chains. Similar to flexural modulus, the flexural strength of NPCC-filled composites was lower than EFB-filled composites as shown in Figure 3.65. The inability to support stresses transferred from the matrix due to the lower aspect ratio is a reason for the Flexural modulus (MPa) decreasing in flexural strength of NPCC-filled composites. 5000 4500 4000 3500 EFB-filled composites (S9-S12) 3000 2500 2000 1500 1000 NPCC-filled composites (S89-S91) 500 0 0 10 20 30 40 50 Filler content (phr) Figure 3.64 composites Effects of EFB (S9-S12) and NPCC (S89-S91) on the flexural modulus of Flexural strength (MPa) 157 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 EFB-filled composites (S9-S12) NPCC-filled composites (S89-S91) 0 10 20 30 40 50 Filler content (phr) Figure 3.65 Effects of EFB (S9-S12) and NPCC (S89-S91) on the flexural modulus of composites Figure 3.66 depicts that the EFB-filled composites containing 20 phr NPCC (S96-S99) has the greatest stiffness compared to other filled composites. The flexural modulus of the composites with 20 phr NPCC was significantly higher than the EFBfilled composites with only 10 phr NPCC (S92-S95). The increase of filler proportion in the composites is attributed to the increase of flexural modulus. The results also show that the filled impact modified composites (S13-S16 and S29-S32) samples have lower stiffness compared to the EFB-filled composites containing 10 phr NPCC. The softening effect generated by 3 phr acrylic (S13-S16) and 3 phr CPE (S29-S32) is a cause for the lower stiffness of the filled impact modified composites. Figure 3.67 shows that the flexural strength of EFB-filled composites containing 10 phr or 20 phr NPCC was lower than the EFB-filled composites without NPCC filler. The increase of detrimental effect due to the increase of filler proportion in the composites was the main reason. The result shows that the composites containing 10 phr and 20 phr NPCC had a lower flexural strength compared to other filled composites. As mentioned earlier, the presence of NPCC with low aspect ratio 158 and incapable of EFB filler to facilitate a better transfer of stress, thus increased the deleterious effect of filler on the filled composites. Therefore the flexural strength of EFB-filled composites containing NPCC filler decreased as the EFB filler content increased from 0 to 40 phr. EFB-filled composites (S9-S12) EFB-filled composites added with 3 phr acrylic (S13-S16) EFB-filled composites added with 3 phr CPE (S29-S32) EFB-filled composites added with 10 phr (S92-S95) EFB-filled composites added with 20 phr NPCC (S96-S99) 5000 Flexural modulus (MPa) 4500 4000 3500 3000 2500 0 10 20 30 40 50 EFB filler content (phr) Figure 3.66 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the flexural modulus of the EFB-filled composites 159 EFB-filled composites (S9-S12) EFB-filled composites added with 3 phr acrylic (S13-S16) EFB-filled composites added with 3 phr CPE (S29-S32) EFB-filled composites added with 10 phr NPCC (S92-S95) EFB-filled composites added with 20 phr NPCC (S96-S99) Flexural strength (MPa) 90.0 85.0 80.0 75.0 70.0 65.0 60.0 55.0 50.0 0 10 20 30 40 50 EFB filler content (phr) Figure 3.67 Effects of 10 phr NPCC, 20 phr NPCC, 3 phr acrylic and 3 phr CPE on the flexural strength of the EFB-filled composites 3.3.2.6 Effect of Extracted EFB filler on the Impact Strength and Flexural Properties This study was carried out in order to observe the effect of residues oil in the EFB fillers on the mechanical properties of composites. Comparison will be done to observe whether the extracted EFB filler has a better ability than the unextracted EFB filler in enhancing the mechanical properties of composites. Therefore, the previous mechanical properties of filled composite containing 30 phr unextracted EFB filler was also used as comparison. 160 A. Impact Strength Figure 3.68 shows the effect of extracted fibre on the impact strength of filled composites. It shows that the impact strength of the composite filled with 30 phr extracted filler is marginally higher than the composites with unextracted filler. The degree of enhancement is only about 0.3 %. This indicates that the amount of the oil present in the EFB filler is not a significant factor to influence the impact strength of the filled PVC-U composites. The impact strength of oil extraction filler in the present study is less significant compared to previous study on polypropylene composite (Rozman et al., 2001), which may be due to the polarity of PVC molecules, which enable the polymer to have a good interaction even with oil residue present in the fillers. 8.00 30 phr Impact strength (kJ/m2) 6.11 6.13 6.00 4.00 2.00 0.00 Unextracted-filled composite (S11) Figure 3.68 Extracted-filled composite (S47) Effect of unextracted and extracted EFB filler on the impact strength of composites at filler content of 30 phr 161 B. Flexural Properties Figure 3.69 shows that the flexural modulus of extracted-filled composite is slightly lower compared to unextracted filler. The plausible reason is that the extraction does not only remove the residues of oil but also other extractives in the fillers. Thus, it causes the extracted filler to decrease in compaction and to have many cavities on the surface (Figure 5.10). The decrease of extracted filler rigidity caused a slight decrease in the flexural modulus of the filled composite. Figure 3.70 depicts the results of flexural strength of filled composites. The result shows that there is no significant improvement (only about 3 %) in flexural strength of the composite filled with 30 phr extracted filler compared to the unextracted filler. The relatively similar result in the interaction of extracted or unextracted filler with the matrix is a plausible reason to the insignificant changed in the flexural strength. 4400 30 phr 4300 Flexural modulus(MPa) 4200 4142 4100 4000 3900 3900 3800 3700 3600 3500 3400 Unextracted-filled composites (S11) Figure 3.69 Extracted-filed composites (S47) Effect of unextracted and extracted EFB filler on the flexural modulus of composites at filler content of 30 phr 162 80.0 30 phr Flexural strength (MPa) 78.0 76.0 74.5 74.0 72.4 72.0 70.0 68.0 66.0 64.0 Unextracted-filled composites (S11) Figure 3.70 Extracted-filled composites (S47) Effect of unextracted and extracted EFB filler on the flexural strength of composites at filler content of 30 phr 3.3.2.7 Water Absorption In this study, the effect of acrylic, CPE, untreated and coupling agent-treated filler content on the water absorption of filled composites was investigated. The results in Figure 3.71 and Table 3.23 show that the incorporation of impact modifier increased the percentage of water absorption PVC for 60 immersion days. The water absorption increased with increasing impact modifier content. The increase in the polarity, which was generated by the presence CPE and acrylic in compounds is the main reason for the increasing of water absorption. The polarity of molecular chains of CPE and acrylic is due to chlorine atoms and oxygen atom, respectively. Table 3.23 also shows that acrylic-impact modified compounds has a slight higher values of water absorption compared to CPE-impact modified compounds. 163 0.4 12 phr CPE (S8) Water absorption (%) 0.3 12 phr acrylic (S4) 9 phr acrylic (S3) 6 phr acrylic (S2) 9 phr CPE (S7) 6 phr CPE (S6) 0.2 3 phr acrylic (S1) 3 phr CPE (S5) 0.1 PVC-U (S0) 0.0 0 10 20 30 40 50 60 70 Immersion Day Figure 3.71 Effects of impact modifier content on water absorption of acrylic- impact modified (S1-S4) and CPE-impact modified compounds (S5-S8) Table 3.23: Percentage of water absorption of impact modified compounds and EFB- filled composites Impact modifier content (phr) Water absorption of Impact-modified compounds(%) 0 Acrylic 0.10 CPE 0.10 3 0.16 6 EFB filler content (phr) Water absorption of EFB-filled composites (%) 0 Untreated 0.10 Prosil 9234 0.10 NZ44 0.10 0.13 10 0.74 0.60 0.72 0.22 0.18 20 1.69 1.33 1.53 9 0.23 0.21 30 2.89 2.21 2.32 12 0.27 0.30 40 4.08 3.95 3.84 164 In order to compare the effect of filler on the absorption clearly, only the results of unmodified and modified composites filled with 20 phr and 40 phr were plotted. The result given in Figure 3.72 shows that the water absorption of unmodified and impact modified composites was significantly increased as the filler content increased. This was expected due to the hydrophilic nature of the cellulosic filler. The polar hydroxyl groups in their molecular structures are able to form hydrogen bonds with water molecules. The higher the proportion of EFB filler in the composites, the higher possibility of water is being absorbed through the formation of hydrogen bonds. Thus, the water absorption increased proportionally with the filler content, as shown in Table 3.23. Similar observations have been reported by other researchers who used EFB filler in different polymer matrices (Rozman et al., 2000; Abdul Khalil et al., 2000). Figure 3.72 shows that the water absorption of impact-modified composites was higher than unmodified composites. Similar to the impact modified compounds, the increase in polarity, which was contributed by impact modifier, is the main cause for the higher water absorption of impact-modified composites. 5.0 40 phr EFB,9 phr acrylic (S24) 4.5 40 phr EFB,9 phr CPE (S40) Water absorption(%) 4.0 40 phr EFB (S12) 3.5 3.0 2.5 20 phr EFB,9 phr CPE (S38) 2.0 20 phr EFB,9 phr acrylic (S22) 1.5 20 phr EFB (S10) 1.0 0.5 0.0 0 Figure 3.72 10 20 30 40 Immersion day 50 60 70 Effect of EFB filler content on the water absorption of unmodified and modified composites 165 Figures 3.73 and 3.74 show the effect of coupling agents on the water absorption of unmodified and impact-modified composites, respectively. The results show that Prosil 9234 and NZ44 significantly reduced the water absorption of composites. The ability of coupling agents to modify the cell wall of filler made the EFB filler to be less hydrophilic. Coupling agent has an ability to interact with hydroxyl groups of EFB, thus blocking the latter from hydrogen bonding with water. It was discovered that unmodified (Table 3.23) and impact-modified composites (Table 3.24) treated with Prosil 9234 exhibited slightly less hydrophilic than NZ 44 as filler content increased from 10 to 30 phr. Figure 3.73 also shows that at 40 phr the water absorption of filler treated with both coupling agents was almost similar. Meanwhile, Figure 3.74 and Table 3.24 show that at 40 phr the water absorption of filler treated with Prosil 9234 was slightly higher than untreated. This result shows that the ability of Prosil 9234 to reduce the hydrophilic of filler was ineffective at higher filler content (40 phr), due to the domination of filler content. It shows that the amount of Prosil 9234 used to treat the 40 phr fillers was insufficient to modify the fillers completely. NZ 44 was able to modify the hydrophilic of fillers at higher filler content. 4.5 40 phr EFB(S12) Prosil 9234-40 phr EFB(S52) Water absorption (%) 4.0 3.5 NZ 44-40 phr EFB(S72) 3.0 2.5 20 phr EFB(S10) 2.0 1.5 NZ 44-20 ph rEFB(S70) 1.0 Prosil 9234-20 phr EFB(S50) 0.5 0.0 0 10 20 30 40 50 60 70 Immersion day Figure 3.73 Effect of coupling agents on the water absorption of treated unmodified composites Water absorption (%) 166 Prosil 9234-40 phr EFB, 9 phr acrylic (S64) 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 40 phr EFB, 9 phr acrylic (S24) NZ 44-40 phr EFB,9 phr acrylic(S84) 20 phr EFB,9 phr acrylic(S22) NZ 44-20 phr EFB,9 phr acrylic(S82) Prosil 9234-20 phr EFB,9 phr acrylic(S62) 0 10 20 40 30 60 50 70 Immersion day Figure 3.74 Effect of coupling agents on the water absorption of treated acrylic- impact modified composites Table 3.24: Percentage of water absorption of impact modified compounds and impact modified treated composites EFB filler content (phr) 0 Water absorption EFB-filled impact modified composites (%) 9 phr CPE 9 phr acrylic Untreated Untreated Prosil 9234-treated NZ44-treated 0.21 0.23 0.23 0.23 10 1.03 0.95 0.65 0.76 20 2.08 2.10 1.45 1.66 30 3.15 3.14 2.35 2.54 40 4.46 4.61 4.81 4.14 167 3.4 Conclusions The main purpose in this study is to determine the impact and the flexural properties of EFB filled PVC-U composites. The composites were impact modified with acrylic and CPE, and treated with Prosil 9234 and NZ 44 coupling agents. Based on the results obtained, the following conclusions can be made: (a) The flexural modulus of EFB-filled PVC-U composites increased with EFB filler contents, whereas the impact and flexural strength decreased. The increase of filler contents contributed to the increase composites stiffness. Factors such as filler agglomeration and low aspect ratio of filler and the weakness of filler-matrix interface contributed to the impact and flexural strength decrement. (b) The addition of acrylic and CPE impact modifier improved the impact strength of the composite. Impact modifier allowed PVC-U matrix to yield before fracture. However, the rubbery phase of impact modifier also softened the composites. As a result, the flexural properties of impact-modified composites were decreased. CPE has a greater softening effect on the composite than acrylic impact modifier. The polarity of both impact modifiers increased water absorption of the composites. (c) The Prosil 9234 and NZ 44 improved the impact and flexural properties of the composites. The coupling agents have ability to improve the filler-matrix interface of the composite by introducing ether linkage between the filler and matrix. The FTIR spectra obtained in this study showed that there was interaction between filler and PVC matrix. The presence of this interaction was supported by a reduction in water absorption of treated composites. The coupling agents interacted with hydroxyl groups of EFB, thus blocking the latter from hydrogen bonding with water The interaction, however, 168 insufficient to bring about better enhancement of composite impact strength and flexural properties. Therefore, the flexural modulus obtained from Einstein Model without adhesion in good agreement with the modulus obtained in this study. In other words, the interaction between treated filler and PVC was only improved slightly. Prosil 9234 was found more effective than NZ 44 in improving the mechanical properties of composite. (d) The flexural strength of compounds and composites increased after exposed for 504 hours in the accelerated weathering environment, but the flexural modulus and impact strength were decreased. Overall, acrylic impact modified compounds and EFB-filled acrylic-impact modified composites were more stable than others. Incorporation of EFB accelerated the photo-oxidation degradation, particularly for EFB-filled composites containing CPE impact modifier. (e) The impact and flexural strength of composites insignificantly improved but the flexural modulus decreased marginally when the extracted EFB filler incorporated in composites. NPCC filler was able to increase the impact strength but not for flexural properties of composites compared to EFB filler. The NPCC filler cannot be used as impact modifier in the EFB-filled composites due to the inability of NPCC to enhance the impact strength of composites. CHAPTER 4 THERMAL PROPERTIES OF COMPRESSION MOULDED PVC-U COMPOSITES 4.1 Introduction PVC is thermally unstable, especially at elevated temperature, and in presence of oxygen and humidity. The degradation of PVC has been shown by most researchers to proceed in two distinct stages (Marcilla and Beltran, 1995; McNeill et al., 1995). The first stage corresponds to the dehydrochlorination to form polyene with subsequent formation of conjugated polyene sequences of doubles bonds (-HC=CH-) in the backbone of the polymer after splitting of the first hydrogen chloride (HCl) molecules (initiation reaction), accompanied by the formation of small amounts of hydrocarbons, mainly aromatics such as benzene, napthelene and others. The second stage corresponds to the scission of the polyene sequences, which formed during the first stage (Marcilla and Beltran, 1995; Mc Neill et al., 1995). Accordingly, one might expect that in the thermogravimetric (TG) curve of PVC, the first weight loss would be due to HCl evolution, while the second, would be due to the evolution of hydrocarbons (Montaudo and Puglisi, 1991). At the last stage, a black carbonaceous residue known as char was obtained. 170 Klaric et al. (2000) have studied the thermal stability of PVC blended with chlorinated polyethylene (CPE) impact modifier. The blend has also produced the two stages of degradation in which the dehydrochlorination took place in the first step and followed by the degradation of dehydrochlorinated sequences, in the second step. The addition of CPE have improved the thermal stability of PVC for all compositions investigated. The thermal stability of PVC blended with acrylic impact modifier has not been reported in literature. Pavlinec et al. (1997) have, however, studied the thermal stability of polymer bead consists of poly (methyl methacrylate) (PMMA) core and poly (butyl acrylate) (PBAC) copolymer shell. They found that the PBAC, which is an impact modifier, was more stable than PMMA. The PBAC degradation temperature was observed to occur at 380-390 oC while PMMA at 340-350 oC. The effect of impact modifiers on the heat deflection temperature of PVC has been studied by previous researchers (Deanin and Chuang, 1986; Lutz, 1993). HDT is often related to the mechanical behaviours of the materials. Deanin and Chuang (1986) reported that addition of impact modifiers (ABS and CPE) have lower the HDT of rigid vinyl polymers progressively. It means that impact modifiers were able to reduce the modulus of rigid vinyl polymers. However, according to Lutz (1993), HDT has appeared to be more affected by the molecular weight (K-value) of the PVC than by the impact modifier content. Pedro et al. (2001) characterized the agalmatolite-filled PVC composites using a differential scanning calorimetry (DSC). DSC results showed that there were a slight decrease in the glass transition temperature (Tg) as the agalmatolite filler content increased from 10 to 40 phr. However, the similar trend was not observed in the composites filled with calcium carbonate filler, which practically did not show any difference in Tg, as the filler content increased (Pedro et al., 2001). Meanwhile, Sombatsompop et al. (2003) has reported that the Tg of PVC increased with increasing wood sawdust content. 171 Since most polymers are processed at high temperature, thermal stability of natural fibres is also one of the factors to be considered in developing natural fibrefilled composite. Thermal stability of lignin was considered to be greater than that of hemicellulose and less than that of cellulose (Beall and Eickner, 1970). Based on thermogravimetric analysis (TGA) conducted by Thiebaud et al. (1997), they found that the untreated oak wood started to lose weight at the temperature range of 22-200 o C due to the evaporation of water in the wood. The significant weight loss occurred at temperature range of 200-330 o C, which corresponded to the degradation of hemicellulose while at temperature range of 330-490 oC corresponded to cellulose. Nguyen et al. (1981) found that the pyrolisis of hemicelluloses, lignin and cellulose of wood fibre occurred at 200, 220 and 250 oC, respectively. In recent years, most studies conducted involved investigating the mechanical properties of natural fibre-filled polymer composites as a function of fibre loading, using various treatments for the fibres. Very few workers have investigated the thermal stability of the EFB-filled PVC-U composites. Therefore, the aim of the study in this chapter is to investigate the thermal behaviours such as heat deflection temperature, glass transition temperature, and thermal degradation temperature of the composites. 4.2 Experimental In this section, the materials, filler preparations, and blend formulation used are similar to Chapter 3. The HDT, DSC, and TGA analysis carried out in this chapter are described. 172 4.2.1 Materials, Filler Preparation and Blend Formulations The blend formulations as summarized in Table 3.3 were selected and used in this study. The blend formulations used for HDT were PVC-U compound (S0), impactmodified PVC-U (S1-S8), EFB-filled PVC-U composites (S9-S12), and EFB-filled impact-modified PVC-U composites (S21-S24 and S37-S40). The blend formulations used for DSC were S0, S3, S7, S9-S12, S21-S24 and S37-S40. For TGA, blend formulation used were S0, S9 to S12, S21-S24 and S37-S40. The neat PVC resin, EFB filler, acrylic, and CPE impact modifier were also analysed using TGA. 4.2.2 Sample Preparations As described in detail in Sections 3.2.4-3.2.6 (Chapter 3), the dry blending of each formulation was first done using a laboratory high-speed mixer. The dry blended compounds were then melt-blended and sheeted using a two-roll mill. The milled sheets were then placed into a mould and hot pressed. The moulded test specimens were used as samples for thermal analysis. 4.2.3 Heat Deflection Temperature Testing The heat deflection temperature test was conducted using Toyo-Seiki HDT S-3 tester, following ASTM D 648, with flexural load of 1.2 kg and a heating rate of 2 + 0.2 oC/min. The samples with standard dimensions of 125 x 13 x 3 mm3 were immersed under flexural load into a heated bath of silicone oil. The temperature of the oil was measured when the test bar deflected by 0.25 mm. This temperature was recorded as 173 the deflection temperature under flexural load of the test sample. All reported values for the test were the average of two specimens. 4.2.4 Differential Scanning Calorimetry Study Differential scanning calorimetry (DSC) was carried out according to ASTM D3418-82 at a heating rate of 10 0C/min to ensure high resolution of the DSC curve, and at temperature ranging from 30 to 250 0C. A Mettler Toledo DSC model 821 with STAR system was used. The liquid nitrogen rate was maintained at 50 ml/min. The calorimeter was calibrated with an indium reference. Meanwhile, samples in the size range 4.5-5.5 mg were placed in a sealed aluminium pan and an empty aluminium pan was used as the reference. The tangent method was used to determine the glass transition temperature (Tg) of the samples. The mid-point of the first endothermic baseline in the DSC heating curve was taken as the Tg. 4.2.5 Thermogravimetric Analysis The thermogravimetric analyser used for the thermal degradation study was a Perkin Elmer-TGA 7. Approximately 9.5-10.0 mg of samples was scanned from 30 to 800oC, at heating rate of 10 oC /min. Each sample was placed in an open platinum sample pan. The purge gas used was nitrogen at a flow rate 20 cm3/min. The onset temperature of a 5% (T5%) weight loss deviation from the baseline of the thermogravimetric (TG) curve was used as the indicator of the composite’s thermal stability. The onset temperature for degradation was found to be tremendously dependent on the run conditions (Klaric et al., 2000). The differential thermogravimetric (DTG) curve for weight loss was also recorded. 174 4.3 Results and Discussion In this section, results on heat deflection temperature testing, differential calorimetry study and themogravimetric analysis of the unmodified and impact modified of composites will be presented. 4.3.1 Heat Deflection Temperature Table 4.1 shows the effect of impact modifier content on the HDT of the impact-modified PVC-U. The HDT of the PVC-U compound (S0) obtained was 66 oC. This value was within the range of HDT of PVC commercial grades reported by Fried (2003). In general, the decrease of HDT due to the addition of both impact modifiers was not significant. The decreased in HDT was only marginal as the impact modifier content increased from 3 to 12 phr. A similar trend has also been reported by Deanin and Chuang (1986) for the acrylonitrile butadiene styrene (ABS) and CPE impact modified PVC. Both impact modifiers have slightly decreased the HDT of PVC. The results shown in this study were also in an agreement with the results obtained by Lutz (1993), who reported that the HDT was more affected by the molecular weight of PVC compared to the impact modifier content. Although the results in Table 4.1 do not show significant difference in HDT as the impact modifier content increased, the information can be used to distinguish the softening effect contributed by acrylic and CPE impact modifier. The softening effect was able to reduce the rigidity of PVC-U compounds by means of reducing the strength of intermolecular interaction (Deanin and Chuang, 1986). The HDT of acrylic-impact modified PVC-U (S1-S4) was found slightly higher than CPE (S5-S8). It shows that the CPE impact modifier had greater softening effect than acrylic for the similar reasons enumerated in Section 3.3.2.2 (Chapter 3). As a result, the flexural modulus and 175 strength of the CPE-impact modified PVC-U compounds were also lower than acrylicimpact modified PVC-U as the impact modifier content increased. Table 4.1: HDT of acrylic-impact modified and CPE-impact modified PVC-U compounds S0 Acrylic content (phr) 0 HDT (oC) 66.0 + 0.3 S1 3 S2 S0 CPE content (phr) 0 HDT (oC) 66.0 + 0.3 65.8 + 0.1 S5 3 65.0 + 0.1 6 65.6 + 0.1 S6 6 64.9 + 0.1 S3 9 64.9 + 0.6 S7 9 64.7 + 0.4 S4 12 65.0 + 0.1 S8 12 64.3 + 0.1 Sample Sample Tables 4.2 and 4.3 show the HDT for the unmodified filled composites (S9-S10) and impact-modified filled PVC-U composites (S21-S24 and S37-S40), respectively. The results showed that the HDT of S0 was slight higher than that of the S9 composites. A similar result was also found in the work of Chiu and Chiu (1996, 1998), who investigated the relationship between HDT and carbon black content in copolymer polypropylene. They found that that the HDT of the blend decreased as the carbon black content increased. The decrease in HDT upon incorporation of 10phr EFB filler was probably caused by the fact that the silicone oil was able to penetrate into the composites and acted as a lubricant, affecting the filler-matrix interface. It was found that the filled unmodified composites slightly increased in weight by 0.2 % on immersion in silicone oil for the HDT measurement. The HDT of composites increased insignificantly with a further increase of fillers from 10 to 40 phr. It has been reported that the incorporation of fillers into highly crystalline polymers generally contributed to an increase in HDT (Schwartz, 1987). However, PVC polymer, with only about 5-10 % crystallinity was incapable of undergoing a significant improvement in HDT as the filler content increased. The amorphous PVC and the lubrication effect of silicon oil were attributed to the insignificant improvement in HDT. This is due to the mobility of PVC chains was less restricted. 176 Similar results (Table 4.3) were also obtained for S21-S24 and S37-S40 where the HDT was not significantly increased with increasing filler content. The softening effect produced by the acrylic and CPE was also experienced by the S21-S24 and S37-S40. Therefore, the HDT of these impact-modified filled composites were slightly lower than the unmodified filled composites. As mentioned earlier, CPE had a greater softening effect than acrylic, thus the HDT values of S37-S40 were slightly lower than S21-S24. Table 4.2: HDT of EFB-filled PVC-U composites S0 EFB filler content (phr) 0 HDT (oC) 66.0 + 0.3 S9 10 64.7 + 0.1 S10 20 65.1 + 0.6 S11 30 66.3 + 1.6 S12 40 66.0 + 1.0 Sample Table 4.3: HDT of EFB-filled acrylic-impact modified and EFB-filled CPE-impact modified PVC-U composites EFB HDT filler (oC) content (phr) 0 64.9 + 0.6 Sample CPE content (phr S7 9 EFB filler content (phr) 0 65.0 + 0.1 S37 9 10 63.9 + 0.2 20 64.9 + 0.8 S38 9 20 64.5 + 0.3 9 30 65.5 + 0.2 S39 9 30 64.9 + 1.3 9 40 65.6 + 0.7 S40 9 40 64.8 + 0.5 Sample Acrylic content (phr) S3 9 S21 9 10 S22 9 S23 S24 HDT (oC) 64.7 + 0.4 177 4.3.2 Glass Transition Temperature The glass transition temperature (Tg) is an important physical parameter characterizing amorphous polymers in terms of their chain rigidity and intermolecular forces. The Tg of S0 showed in Table 4.4 was comparable with the Tg of PVC reported in the literature (Pedro et al., 2001). Similar to S0, the Tg values of the unmodified filled composites (S9-S12) and impact-modified filled composites (S21-S24 and S37-S40) in the Tables 4.4 and 4.5, respectively, recorded as a function of filler content, were also obtained from the DSC traces. They show a slight increase in Tg with increasing filler content from 0 to 40 phr. This was predictable since the incorporation of EFB fillers in PVC-U matrix increased the flexural modulus. Increased in Tg can be explained on the basis of reduced mobility of chain molecular segments in the vicinity of the filler (Ash et al., 2002). The physical interactions, due to dipole-dipole interactions, between PVCU matrix and filler provided another reason. However, these interactions were not strong enough to cause a severe restriction in the molecular segmental motion of PVC. As discussed in HDT section, the greater softening effect produced by CPE compared to acrylic can also be observed in the Table 4.5. The Tg values of S37-S40 were slightly lower than S21-S24. Table 4.4: Tg of EFB-filled PVC-U composites S0 EFB filler content (phr) 0 Tg (oC) 80.4 S9 10 81.0 S10 20 81.4 S11 30 81.9 S12 40 82.1 Sample 178 Table 4.5: Tg of EFB-filled acrylic-impact modified and EFB-filled CPE-impact modified PVC-U composites S7 CPE content (phr 9 EFB filler content (phr) 0 80.4 81.1 S37 9 10 81.1 20 81.4 S38 9 20 81.1 9 30 82.0 S39 9 30 81.6 9 40 83.1 S40 9 40 82.5 EFB filler content (phr) 0 Tg (oC) Sample S3 Acrylic content (phr) 9 80.5 S21 9 10 S22 9 S23 S24 Sample Tg (oC) 4.3.3 Thermal Degradation Temperature Figure 4.1 shows the thermogravimetric (TG) and differential thermogravimetric (DTG) curves of neat components degradation, respectively. The first stage degradation of the neat PVC resin (uncompounded PVC) as shown in Figure 4.1 occurred in the temperature range of 200-400 oC with a peak at 300 oC while the second stage occurred in the range of 400 –530 oC with a peak at 470 oC. Meanwhile, for S0 (compounded PVC), its first stage was shifted to the slightly higher temperature in range of 200-410 oC with a peak at 336 oC while the second stage was in the temperature range of 410-600 oC with a peak at 500 oC. The TG and DTG curves of neat PVC resin recorded in this study were also comparable to the previous findings (Marcilla and Beltran, 1995; Mc Neill et al., 1995). As mentioned earlier, the researchers have been suggested that the first stage corresponded to the dehydrochlorination of the PVC with the subsequent formation of conjugated double bonds. This was considered to be the main reaction and HCl was assumed to be the major volatile product of degradation. The second stage was corresponded to the scission of polyene structure sequences formed during the first 179 stage that contributed to the measured weight loss of the PVC resin. A char was obtained at the end of degradation. Figure 4.1 (a) TG and (b) DTG curves for neat components Table 4.6 shows the T5% of the neat components obtained from TG and DTG curve (Figure 4.1). The onset temperature (T5%) of the S0 was higher than that of PVC resin while the weight loss of S0 (59 wt %) was lower than the weight loss of PVC resin 180 (64 wt %) as shown in Figure 4.1(a). Similar findings were also reported by other researchers (Miranda et al., 1999). They found that the weight loss of the suspension PVC resin at the first stage (200-375 oC) was 64 wt % while Marcilla and Beltran (1995) found that the weight loss of emulsion PVC resin was 65 wt % in the temperature range of 227-377 oC. The slightly higher weight loss than the theoretical value was due to the elimination of low molecular hydrocarbon molecules, which took place simultaneously with the volatilisation of HCl during this stage (Okieimen and Eromonsele, 2000). The degradation of S0 also occurred in two stages but its onset temperature was increased by 14 oC, indicating an improved thermal stability of S0.The total weight loss at the first stage was similar with stoichiometric amount of HCl of 58.7 wt %. The similarity was corresponded to HCl volatilisation totally without involving the small amount of hydrocarbons. A tin heat stabilizer, which was added into the S0, suppressed the small amount hydrocarbons evolved during pyrolisis as reported by Montaudo and Puglisi (1991). They reported that the production of benzene and naphthalene was drastically suppressed by the presence of tin dioxide (SnO2) in the S0, whereas other aromatics were suppressed to a lesser extent. It means that the crosslinking of the linear polyene chains formed and transformed directly to char without being able to produce benzene and other hydrocarbons during the dehydrochlorination process. The suppression was resulted in the S0 in this study producing a higher residual char (18 wt %) than PVC resin (8 wt %) after the pyrolisis was completed (Figure 4.1a). Table 4.6: Degradation temperature of neat components obtained from TG and DTG curves Neat Components Temperature (oC) Peak 1 300 PVC resin T5% 281 PVC-U compound (S0) 295 335 500 EFB filler 263 340 none Acrylic 359 421 none CPE 287 335 484 Peak 2 470 181 The results of EFB filler degradation are shown in Table 4.6 and Figure 4.1. Table 4.6 shows that the T5% of 263 oC of EFB filler was the lowest temperature compared to other neat components. The EFB fillers degraded at this temperature, which corresponds to the weight loss caused by production and volatilisation of volatile products. However, based on the DTG curve of EFB filler (Figure 4.1b), there was a minor peak in the temperature range of 100-200 oC. At this temperature, EFB fillers were heated above the boiling temperature of water, and thus the moisture content of EFB fillers evaporated as vapours due to the drying process. The EFB fillers lost their weight due to the devolatilisation of moisture, approximately 1.2 wt % (Figure 4.1b). The presence of moisture in EFB filler was also observed in the DSC thermogram as shown in Figure 4.2. The minor endothermic peak at about 110 oC corresponding to the heat required for evaporation of moisture in the EFB filler (Thiebaud et al., 1997; Peters and Bruch, 2001). Figure 4.2 DSC trace of EFB filler A more intense peak of DTG implies the maximum rate of weight loss of EFB filler occurred at 340 oC (Figure 4.1b). Meanwhile, TG curve (Figure 4.1a) shows that 182 only one main step of weight loss of 78 % occurred in the temperature range of 200550 oC and produced 22 % of solid char at 550oC. The thermal degradation of EFB filler was due to the decomposition of hemicellulose, lignin and cellulose to give off volatiles. Rowell (1997) reported that the hemicellulose and cellulose degraded by heat much before the lignin. The lignin component contributes to char formation. Based on the literature (Nguyen et al., 1981; Thiebaud et al., 1997), as well as from the DSC studies, (Figure 4.2) the exothermic peaks at 207, 223 and 234 oC can be attributed to the pyrolisis of hemicelluloses, lignin and cellulose, respectively. The degradation of EFB fillers started at 250 oC, followed by exothermic peak at 267 oC and 329 oC that were related to volatilisation/degradation and oxidative combustion of the charred residues, respectively. The slightly dissimilar results obtained from literature may be due to the different type of fibre used. According to Saheb and Jog (1999) the chemical composition of natural fibres varies depending upon the type of the fibre. Similar to S0, the dehydrochlorination was also the dominant reaction in the CPE impact modifier degradation. Klaric et al. (2000) reported that the saturated backbone and partly crystalline structure of CPE were the reasons responsible for the good thermal stability of CPE before HCl started to evolve. The TG and DTG curve of CPE in Figure 4.1 also showed two stages of degradation. The first stage occurred in the temperature range of 200-400 oC with a peak at 335 oC while the second stage occurred in the temperature range of 400-540 oC with a peak at 484 oC. CPE has better thermal stability compared to PVC resin because of its maximum weight loss of 35 wt % in the first stage (Figure 4.1a) whereas PVC resin weight loss was 64 wt %. The CPE weight loss was almost 36 %, similar to the chlorine content in CPE. Thus, the weight loss in TG curve could be used to estimate the chlorine contents of CPE (Varma et al., 1999). As shown in Figure 4.1a, the TG curve of acrylic impact modifier, which consists of PMMA shell and PBAC core, revealed that only one stage of degradation was observed in the temperature range of 250-500 oC with a peak at 420 oC and weight loss of 97 wt %. However, from the corresponding DTG curve in Figure 4.1b, it indicated that the one stage of degradation consisted of two stages; the one with the 183 temperature between 250-380 oC corresponded to PMMA. As a shell, PMMA should be initially subjected to heat compared to the PBAC core during pyrolisis. The other with the temperature between 380-500 oC corresponded to PBAC (Pavlinec et al., 1997). The TG and DTG curves of unmodified filled PVC-U composites (S9-S12) in Figure 4.3 clearly shows that devolatilisation of moisture of EFB fillers occurred in temperature range of 100-200 oC for all EFB filler contents even after the EFB fillers incorporated into the PVC-U matrix. Moisture content in cellulose-containing composites has also been detected by other researchers (Sapieha et al., 1989). Such moisture content might play a significant role in degradation processes due to OH group in water was more reactive the OH group available in the EFB filler (Rowell, 1997). Table 4.7 shows that T5% decreased progressively with increasing fibre content while the DTG curves (Figure 4.3b) also changed from two peaks for the S0 to the three peaks for the S9-S12. Besides, the size of the first peak (maximum rate of weight loss) also proportionally increased with EFB filler content. Table 4.7 also shows that the first peaks were largely associated with the degradation of the fillers and were only marginally related to the PVC-U matrix. Therefore, the temperature range of 200-300 o C corresponded to the thermal degradation of the EFB filler in the composites, while the remaining peaks (i.e. the second and third) at about 333-343 oC and 496-505 oC corresponded to the degradation of PVC-U matrix. 184 (b) Figure 4.3 (a) TG and (b) DTG curves of unmodified filled PVC-U composites 185 Table 4.7: Degradation temperature of EFB-filled PVC-U composites obtained from TG and DTG curves Temperature (oC) Sample S0 Filler content (phr) 0 S9 10 277 271 339 493 S10 20 280 281 342 505 S11 30 276 278 343 496 S12 40 267 278 333 496 T5% Peak 1 Peak 2 Peak 3 295 336 500 none As mentioned earlier, the hydroxyl groups in the fillers facilitated extensive hydrogen bonding between the macromolecules of the filler. Since the moisture in the unmodified filled composites evaporated during the thermogravimetric test before the degradation began, the hydroxyl groups of moisture did not influence the degradation. However, the water molecules released by the fillers during the thermal degradation significantly affected thermal degradation of unmodified filled composites. The amount of water released would increase as the filler content increased. Because of their polarity, the released water molecules would be able of inducing the chlorine atoms of PVC to cleave from the main chains and undergo dehydrochlorination, as shown in Figure 4.4. (Naqvi and Sen, 1991; Sombatsompop et al., 2003, 2004). H2C PVC molecular chain CH2 2HC C H2C δ+ δO + Hδ Figure 4.4 Cl H δδ+ water molecule Polar-polar interaction between H20 and PVC molecules 186 Figure 4.5 shows the TG and DTG curves for the CPE-impact modified PVC-U compound (S7) and EFB filled CPE-impact modified PVC-U composites (S37-S40). As shown in Table 4.8, the addition of 9 phr CPE into PVC-U increased the thermal stability of S7 as indicated by theT5% and first peak. The first peak of S7 was shifted to the higher temperature. The first peak of S7 occurred at temperature of 349 oC, while the first peak of S0 occurred at 336 oC. This indicates that the interaction between CPE and PVC in S7, resulting in improvement in the thermal stability of the compound. The S7 also showed two degradation steps, the first peak was attributed to dehydrochlorination of PVC and the second peak was attributed to decomposition of polyene structure sequences formed during the first stage. The degradation step of S37-S40 also changed from two to three stages when EFB filler was incorporated. The devolatilisation of moisture content of EFB fillers in the S37-S40 also occurred in temperature range of 100-200 oC. The first peaks of S37-S40 were also due to the degradation of EFB fillers. As shown in Table 4.8, this first temperature peak of S37-S40 was relatively similar to the first temperature peak of S9-S12 in Table 4.7. 187 Figure 4.5 TG and (b) DTG curves for EFB-filled CPE-impact modified PVC-U composites Table 4.8: Degradation temperature of EFB-filled CPE-impact modified PVC-U composites EFB filler CPE Temperature (oC) Sample content content T5% Peak 1 Peak 2 Peak 3 (phr) (phr) S0 0 0 295 336 500 none S7 9 0 300 349 507 none S37 9 10 285 273 352 506 S38 9 20 283 285 351 500 S39 9 30 275 278 339 502 S40 9 40 249 254 310 464 Figure 4.6 shows that the thermal degradation of acrylic-impact modified PVCcomposites (S3) was shifted to the lower temperature instead of higher temperature, even though acrylic impact modifier has good thermal stability (Figure 4.1). The T5% of S3 was also lower than the T5% of S0. This indicated that the addition of 9 phr of acrylic decreased the thermal stability of PVC-U matrix. Even though the degradation of neat 188 acrylic impact modifier began at higher temperature compared with PVC-U but it believed that small methyl methacrylates (MMA) molecules produced from the depropagation of PMMA reacted with the macromolecules and the radicals formed during degradation of PVC (Kovacic et al., 1993). The introduction of highly polar MMA might intensify the undesirable polar tensions, resulting in heat destabilisation. This polar interaction (similar to Figure 4.4) accelerated by the dehydrochlorination of PVC, which was activated by the polarisation as illustrated in Figure 4.7. Figure 4.6 composites (a) TG and (b) DTG curves for EFB filled acrylic-impact PVC-U 189 δ− O C δ+ δ− Cl O δ+ H Figure 4.7 CH3 Polar interaction between MMA and HCl. Figure 4.6 also shows the volatilization of moisture in the temperature range of 100-200 oC. Similar to DTG of S9-S12 (Figure 4.3), the DTG of S21-S24 also showed three peaks but all these peaks were shifted to the temperatures lower than the S9-S12 and S37-S40 composites. The first peak at temperature range of 251-258 oC (Table 4.9) corresponded to the degradation of incorporated EFB filler in the composites while the major second peak and third peak at 308-320 oC and 460-507 oC, respectively, corresponded to the dehydrochlorination of PVC-U matrix. Table 4.9: Degradation temperature of EFB-filled acrylic-impact modified PVC-U composites obtained from TG and DTG curves Temperature (oC) EFB filler content (phr) 0 T5% Peak 1 Peak 2 Peak 3 S0 Acrylic content (phr) 0 295 336 500 none S3 9 0 275 311 469 none S21 9 10 257 254 320 475 S22 9 20 248 251 314 464 S23 9 30 249 252 315 474 S24 9 40 247 258 308 460 Sample The T5% of S21-S24 as given in Table 4.9 is lower compared to S9-S12 composites (Table 4.7). It shows that the combination EFB fillers and 9 phr of acrylic impact 190 modifier actively accelerated the thermal degradation of PVC-U matrix. This may be attributed to the presence of EFB filler (which contained polar groups) in the composite had enhanced the concentrations of like-poles in the PVC-U matrix. Such concentrations, randomly distributed in the PVC-U matrix, may be considered to constitute weak or high-energy spots and possible sites of initiation of thermal dehydrochlorination (Naqvi and Sen, 1991). 4.4 Conclusions The results obtained from the HDT, DSC and TGA analysis in this study, the following conclusions can be made. (a) CPE and acrylic impact modifier slightly decreased the HDT of PVC. CPE has a greater softening effect than acrylic impact modifier. Thus, the HDT of filled CPE-impact modified composites was greater than filled acrylic-impactmodified as the impact modifier content increased from 3 to 12 phr. The HDT of unfilled and filled impact-modified PVC-U was not significantly affected by the increase of impact modifier contents. (b) The HDT of unmodified and impact modified filled PVC-U composites was not significantly affected by the increase of EFB filler content. The amorphous structure of PVC and the lubrication effect of silicon oil were attributed to the insignificant improvement in the HDT of filled composites. (c) The Tg of unmodified and impact-modified filled PVC-U composites slightly increased as the EFB filler contents increased. The restriction of segmental 191 mobility of the polymer chains in the vicinity of the fillers was contributed to the increment of Tg. (d) The incorporation of EFB filler decreased the thermal stability of unmodified and impact-modified filled PVC-U composites, where the degradation process has changed from two to three stages. The first stage corresponded to the degradation of EFB filler, while the second and third stage corresponded to PVC-U matrix. The maximum rate of weight loss increased with the increase of EFB filler content. (e) The CPE-impact modified PVC–U compound have a good thermal stability compared to the acrylic impact modified PVC-U compound. The presence of highly polar of MMA produced from depropagation of PMMA in the compound intensified the undesirable polar tensions, which accelerated the dehydrochlorination of PVC. This dehydrochlorination actively accelerated with the presence of OH groups of EFB filler, and thus EFB-filled acrylic-impact modified PVC-U composites have the lowest thermal degradation temperature compared to other composites. CHAPTER 5 PROCESSABILITY STUDIES OF PVC-U COMPOUNDS 5.1 Introduction Torque Rheometer is a useful and widely accepted tool for assessing the state of PVC fusion. During mixing in a torque rheometer, PVC compound is subjected to temperature and shear and it passes through all stages of fusion. These stages are studied by measuring the torque and PVC temperature as a function of time for constant chamber temperature and rotor speed (Faulkner, 1975). The fusion process involves a series of particulate changes from larger to smaller particles followed by fusion of PVC resin primary particles. The fusion process is affected by a combination of external parameters such as temperature and shear. Figure 5.1 shows the mechanism of PVC fusion in the torque rheometer (Allsopp, 1982). PVC resins undergo a compaction and densification process yielding individual primary particles and primary particles agglomerates, which eventually fuse together. This route of fusion is a comminution mechanism where PVC grains are broken down instead of elongated into the primary particles. As mixing continues, the particulate nature of the melt is gradually eliminated, provided a sufficient temperature is attained (Allsopp, 1982). 193 Resin grains compacted Cold material compacted in chamber Figure 5.1 Sub-grains and agglomerates slide over one another at minimum torque High torques cause some pulverization Fusion of agglomerates and primary particles Torque increases as interparticle fusion initiates and propagates “Fused” melt of primary particles Torque reaches maximum Temperature increases, torque decreases Rigid PVC fusion mechanism in the torque rheometer (Allsopp, 1982). Since the properties of PVC products depend on the morphology and the degree of fusion, additives that control the rate of fusion within the processing equipment become important ingredients in a successful formulation. Many studies have shown that external lubricants delay the fusion time of PVC compound and internal lubricants promote the fusion time. But the synergistic reaction between them has promoted the fusion time of PVC compound significantly (Hawkins, 1982; Pedersen; 1984; Rabinovitch et al., 1984; Chen et al., 1995). Processing aid, which used to improve the processing properties of PVC has been recognized as a fusion promoter, although the mechanism by which this processing aid accelerates the fusion of PVC has not yet been fully elucidated. According to other researchers, processing aid has increased the rate of fusion by means of increasing the friction between the particles (Cogswell et al., 1980; Krzewki and Collins, 1984). Impact modifier has also been extensively used as an additive in the commercial PVC formulation with the purpose to improve the toughness properties of PVC products. Chlorinated polyethylene (CPE) is commonly used as an impact modifier of PVC. At low content of CPE (less than 10 phr), the fusion time of CPEimpact modified PVC compound has been delayed, whereas the fusion process has 194 been promoted as the CPE content increased. Thus, CPE has dual functions, i.e. as an external lubricant or processing aid depending on the content used in the formulation (Chen and Lo, 1999). The acrylic impact modifier, which consists of core and shell, is another type of impact modifier commonly used in PVC formulations. It promotes the fusion process of PVC. McGill and Wittstock (1987) found that the primary particles are broken-down easily when the PVC blended with this impact modifier. During processing, poly (methyl methacrylate) shell of the impact modifier is capable of having good interaction with the PVC melts, which promotes the fusion process. Many PVC compounds are incorporated with filler either to reduce overall cost or to improve the certain properties of the finished product. Schramm (1965) has studied the effect of varying the percentage of clay filler in a PVC-U compound. The fusion rate of the compound decreased with increasing clay filler content. It is also more difficult to obtain good homogeneity at higher filler loading. The clay filler occupying the voids between the resin particles restricted the movement of the resin particles and reduced the flow ability of the compound. A similar trend was also found by other researchers who studied the effect the rice husk filler loading on the processability of PVC-U compounds (Azman and Sivaneswaran, 2002). The processability of PVC compounds, which is determined by the fusion characteristics, depends on its composition and processing conditions. Although PVC fusion characteristics have been widely studied by other researchers, no study has been reported on fusion characteristics of EFB-filled PVC-U and EFB-filled-impact modified PVC-U compounds. Thus, the main objectives of the study reported in this chapter are to investigate the influences of EFB filler, and impact modifiers contents and mixing temperatures on the fusion time and the end torque of PVC-U compounds. 195 5.2 Experimental In this section, the materials, filler preparation and blend formulation are similar to Chapter 3. Filler extraction, processability test, and FTIR analysis are carried out in this study. 5.2.1 Materials and Blend Formulations The PVC and additives used are shown in Tables 3.1 and 3.2 (Chapter 3). The blend formulations, S0-S12, S14, S18, S21-S24, S26, S30, S34, S37-S40, S42 and S45 -S48 were used in this study. 5.2.2 Filler Preparation The size of EFB fillers and extracted EFB fillers used in this study were less than 75 µm. EFB fillers were dried in an oven at similar conditions as reported in Chapter 3, while the preparation for the extracted EFB fillers is described in detail in the following section. 5.2.2.1 EFB Oil Extraction According to Han and Rowell (1997), most of plant materials consist of extractives, holocellulose, lignin, and inorganics (ash). A Soxhlet apparatus (Figure 196 5.2) described by Han and Rowell (1997) was used in this study for EFB fillers extraction. The chemical reagent used consists of three types of solvents, which were ethanol, toluene, and acetone with purity of 99.7, 99.5 and 99.8 %, respectively. They were mixed in proportion to ratio of four volumes of toluene, one volume of acetone and one of volume ethanol (Han and Rowell, 1997). These solvents were employed to extract the oil residues of EFB fillers. Approximately 31 g of EFB fillers were weighed and placed into preweighed extraction thimble. The thimble was placed in vacuum oven at 45 oC for 24 hours. The thimble was cooled in a desiccator for one hour and weighed. Then the thimble was placed in the Soxhlet extraction unit. The 300 ml of the toluene-acetone-ethanol mixture was placed in 500 ml round bottom flask with several boiling chips to prevent bumping. The extraction was then carried out in a well-ventilated chemical fume hood for 24 hours. The heating temperature of 120 oC was used, which was slightly higher than the boiling point of toluene. The boiling point of toluene, ethanol and acetone is 111 oC, 79 oC, and 56 oC, respectively. After extraction with solvents mixture, the thimble was taken out of the extractors, and the excess solvent was drained. The extracted EFB fillers were washed with ethanol and placed in the vacuum oven overnight at 45 oC for 24 hours to allow complete evaporation of the ethanol. When dried, it was removed to desiccators for one hour and weighed. The collected extract that consisted of extractives and solvents mixture in the 500 ml round bottom flask was condensed using a condensation method in order to separate the extractives from solvents. The extractives were left in the flask while the solvents were vaporized and then condensed as a condensate. 197 5.2.3 Dry Blending Similar to the dry blending process as reported in Chapter 3, the dry blending process for each blend formulations in this study was done using a heavy-duty laboratory mixer. All the dry-blended samples were left in desiccators before using for processability study. Figure 5.2 Soxhlet extraction apparatus. 198 5.2.4 Processability Study The fusion characteristics of the PVC-U compounds were determined by using Brabender Torque Rheometer model PL2200. A 54 gram dry-blended sample was placed into the mixing chamber through a loading chute. Immediately after powder has been loaded, the piston with 5 kg weight was inserted in place, and it was pressed gently to force all the dry blend completely went into the mixing chamber as quickly as possible (within 20 seconds) for best reproducibility and comparability of test result. All the dry blends in this study were run at mixing temperature of 180 oC and rotor speed of 50 rpm. The dry blends, S0, S3, S7, S10, S22 and S38 were also run at 175, 180 and 185 oC at constant rotor speed of 50 rpm. The fusion characteristics of compounds were studied by observing the changes of torque curve. 5.2.5 Fourier Transform Infra-Red Analysis The method procedure used for Fourier Transform Infra-Red (FTIR) analysis in this study was similar as reported in Chapter 3 (Section 3.2.14). The concentrated extract and condensate for FTIR analysis were injected in between sodium chloride plates and fitted in the cell unit, while filler and extracted filler were mixed with KBr to form a thin disc. All these samples were also scanned from 4000 - 370 cm-1 for 16 times. 199 5.2.6 Scanning Electron Microscopy Study Studies of the filler and extracted filler surfaces were also performed with a JEOL model JSM-6301F scanning electron microscope (SEM) as reported in Chapter 3 (Section 3.2.13). The purpose of this study is to investigate the surface of filler after extraction. 5.3 Results and Discussion The discussion will focus on the influence of EFB filler, acrylic and CPE impact modifier content, and processing temperature on the processability of compounds. The effect of extracted EFB filler on the processability will also be discussed. The lubrication mechanism will be suggested and explained at the end of this chapter. 5.3.1 Fusion Characteristics Fusion characteristics such as fusion time, torque at equilibrium stage (end torque) and maximum torque at fusion, which were obtained from the fusion curve, were used to elucidate the processability of the studied compounds. Figures 5.3 and 5.4 show a typical fusion curve of PVC-U compound (S0) and EFB-filled PVC-U composite (S10) melted in the Brabender Torque Rheometer at starting temperature of 180 oC and screw rotor speed of 50 rpm. The compound was discharged when the stabilized torque reached at 7 minutes. The fusion curve obtained in Figures 5.3 and 5.4 illustrated the changes of viscosity of compound related to torque and temperature 200 versus time. The viscosity-related torque curve showed two different peaks. The first peak, A, in the torque curve was due to the loading of sample, and the second peak, X, (fusion peak) was due to the compaction and onset of melting. When the PVC dry blend was loaded into the system, the initial peak was generated, then the torque started to decrease sharply because of the free-material flow before it started to compact; the torque then started to increase and generated the second peak. At this peak, X, the material has reached an effectively void-free state and started to melt at the interface between the compacted material and the hot metal surface. The temperature and torque with respect to this point, X was defined as fusion temperature and fusion torque, respectively. The time between the loading point A and the fusion point X was defined as fusion time, t. The compound was in equilibrium state (end torque) as the torque curve and stock temperature curve being smooth and stable with time. At this state, the melt was in homogenous. Stock Temperature 100 o Torque (Nm) 180 Temprature ( C) 200 80 60 A X 160 Torque 40 140 20 120 End torque t 0 -1 100 1 3 5 7 9 11 13 Time (min) Figure 5.3 A typical temperature and torque curves of PVC-U compound (S0) blended in the Brabender Torque Rheometer Stock temperature curve in Figures 5.3 and 5.4 show the generation of frictional heat during mixing process. Due to the shearing forces of the mixing action, and the influence of the surface temperature in the mixing chamber and on the rotors, 201 the PVC resin particles of the compound started to soften at the surface. Thus, the coefficient of friction between the particles when passing each other in the mixing processing was increased (Schramm, 1965). As a result, the frictional heat in the system was increased. At the beginning of the test, the stock temperature was increased due to heat conduction from the heated mixer, afterwards temperature increased due to transforming work into heat input. 100 220 90 Stock Temperature Torque (Nm) 70 180 60 50 160 40 Torque 140 30 20 Temprature (oC) 200 80 120 10 0 -1 100 1 3 5 7 9 11 13 Time (min) Figure 5.4 A typical temperature and torque curves of EFB filled PVC-U composite (S10) blended in the Brabender Torque Rheometer 5.3.1.1 Effect of EFB Filler Table 5.1 shows that the incorporation of EFB filler in the compound decreased the fusion time of PVC-U (approximately 38 % decrement) as the filler content increased from 0 to 10 phr. The fusion time then remained relatively constant as the filler content increased from 10 to 40 phr. These results were contradicted with the results obtained by other researchers. They found that the PVC-U fusion time 202 increased with increasing filler content of clay or rice husk ash. (Schramm,1965; Azman and Sivaneswaran, 2002). Unlike clay and rice husk ash fillers, oil residues are present in EFB fillers (as reported in Section 5.3.3). The presence of oil residue on the EFB fillers has also been reported by Rozman et al. (2001). Oil residues in the fillers may be able to migrate onto the filler surface due to the shearing actions of the mixer blades and the frictional heat during mixing process. This oil residue forms a layer on the outer surface of EFB filler and tends to be an internal lubricant in the filled compounds. As an internal lubricant, calcium stearate interacts with the ester components of oil residue to form a better adhesion that binds the PVC resin particles together easily. As a result, the PVC resin particles of filled compounds fused at the shorter time. Furthermore, the oil layer on the EFB fillers surface may also obstruct the thermal heat to be absorbed by the EFB fillers. Consequently, most of heat was transferred to the PVC resin particles during mixing process. Another reason for the fusion times of filled compounds were shorter than PVC-U compound is the frictional heat generated by the friction of fillerfiller and filler-PVC resin particles. The generated frictional heat increased the thermal heat of the compounds and allowed PVC resin particles to fuse easily. Table 5.1: Fusion characteristics of EFB-filled compounds at 50 rpm and 80 oC S0 Filler content (phr) 0 Fusion time (s) 96 End torque (Nm) 31.1 Fusion temperature (oC) 177 S9 10 60 28.2 175 S10 20 68 27.3 178 S11 30 60 27.5 177 S12 40 58 26.1 175 Sample As mentioned earlier, the fusion time of filled compounds remained relatively constant as the filler content increased. This is due to the increase of frictional heat of 203 filled compounds and adhesive property of calcium stearate were probably nullified by the increase of separation and slipping among PVC resin particles as the filler content increased. This reason can also be cited to explain the decrease of end torque and fusion temperature of filled compound (Table 5.1) at 10 phr and remained relatively constant as the filler content increased from 10 to 40 phr. In general, the melt viscosity of these compounds decreased with increasing EFB filler content. 5.3.1.2 Effect of Acrylic Impact Modifier Table 5.2 shows that the fusion time of acrylic-impact modified compounds (S1-S4) decreased linearly with increasing acrylic impact modifier content. The acrylic impact modifier, which has PMMA shell, increased the adhesive property of calcium stearate. The formation of adhesive property in the impact-modified compounds was more powerful than filled compounds due to the acrylic shell having a greater compatibility with PVC resin particles. The combined lubrication effect between acrylic shell and calcium stearate facilitated the PVC resin particles to interact closely. Consequently, the local friction between the particles, which generated the frictional heat in the impact modified compounds, was highly enhanced. Therefore, McGill and Wittstock (1987) reported that a greater degree of interaction between the PVC primary particles and acrylic promoted the fusion process. The acrylic shell was also thought to melt first and became a viscous adhesive in the compound. This viscous adhesive increased the viscosity of the compound, causing a build up in heat from the shearing process. Besides, it also appeared to perform as a powerful adhesive that binds the PVC resin particles to fuse together easily. Kosior and Stachurski, (1986), also reported that the PVC resin particles underwent further shear process as the acrylic impact modifier content increased which resulted in the increase of resin particles breakdown and followed by the increase rate of cohesion of primary particles agglomerates. Based on these reasons, therefore, the fusion time of acrylic-impact modified compounds in Table 5.2 was decreased approximately 73 % when the impact modifier content increased from 0 to 12 phr. 204 The end torque values of acrylic-impact modified compounds are interesting to observe, because the end torque remained relatively constant as the acrylic content increased. It means that the melt viscosity of these compounds was unaffected with the increasing acrylic content from 3 to 12 phr. The addition of acrylic in the compounds was able to promote the PVC resin particles to reach a homogeneous (equilibrium) melt at the shorter time. In other words, through the process of mixing, heat transfer from the mixer wall and frictional heat from the inter-particles interaction, the PVC resin particles increases in temperature at the shorter time, and eventually the torque increases to a maximum value as the fusion peak since a coherent melt is observed at this stage. As the mixing continues the temperature gradually increases due to the fractional heat, causing a faster reduction in torque as the melt viscosity is decreased (Kosior and Stachurski, 1986). The increase of frictional heat in the compound at the shorter time resulted in the torque curve capable to stabilize rapidly and uniformly. Table 5.2: Fusion characteristics of acrylic-impact modified compounds at 50 rpm and 180 oC. Sample Acrylic content (phr) Fusion time (s) End torque (Nm) So 0 96 31.1 Fusion temperature (oC) 177 S1 3 58 30.9 174 S2 6 40 30.3 173 S3 9 30 30.8 171 S4 12 26 31.7 171 Table 5.3 shows the influence of EFB filler contents on the processability of the filled acrylic-impact modified PVC-U compounds. The result shows that the fusion time of filled impact-modified compound was not significantly changed at filler content of 10 phr. It indicates that the filler less inhibited the function of acrylic as a fusion promoter. This is due to the agglomeration of filler in the filled impact- 205 modified compound were still not predominant. Besides, the frictional heat, which was generated by the fillers, might also contribute to increase the thermal energy in the compounds. As a result, the powerful adhesive property of acrylic and calcium stearate and the generated frictional heat were able to overcome the detrimental effect of filler and able to bind PVC resin particles together easily and allowed them to fuse at the shorter time. Table 5.3 also shows that the fusion time of filled impact-modified compound was linearly increased as the filler content increased from 10 to 40 phr. The increase of filler in the compound probably inhibited the ability of acrylic to perform as a fusion promoter effectively. However, the fusion times of filled impact-modified compounds in the Table 5.3 were still shorter compared to the fusion times of filled unmodified compounds as shown in Table 5.1. It indicates that the influence of acrylic impact modifier in reducing the fusion time of PVC-U compound was more significant than the oil residue of filler. Table 5.3: Fusion characteristics of EFB-filled acrylic-impact modified compounds added with 9 phr of acrylic at 50 rpm and 180 oC S3 Acrylic content (phr) 9 EFB filler content (phr) 0 Fusion Time (s) End Torque (Nm) 30 30.8 Fusion Temperature (oC) 171 S21 9 10 28 30.7 171 S22 9 20 36 28.4 174 S23 9 30 42 28.6 175 S24 9 40 52 27.9 178 Sample The acrylic became less effective in increasing the adhesive property of calcium stearate due to the tendency of EFB fillers to agglomerate among them than to disperse throughout the compound as the filler content increased. As a result, the acrylic took a longer time to bind the PVC resin particles together. Furthermore, the 206 increase of filler agglomerations in the compound with increasing filler content was also increased the separation and slipping among PVC resin particles. As a result, the transferring of heat to PVC resin particles in the compounds was also delayed. Thus, PVC resin particles took a longer time to fuse together. Similar to the fusion time, the fusion temperature of filled impact-modified compounds in Table 5.3 was increased as the filler content increased. Since the fusion time of these compounds was increased, it could be explained that more thermal energy are required to be absorbed by these compounds in order for PVC resin particles to fuse together. Thus, the fusion temperature was increased and resulted in a decrease in end torque of these compounds. It shows that the melt viscosity of these compounds decreased with increasing EFB filler content. Table 5.4 shows the effect of acrylic impact modifier content on the processability of the filled acrylic-impact modified PVC-U compounds. The fusion characteristics of these compounds in Table 5.4 show a similar trend with the results obtained in Table 5.2 (acrylic-impact modified compounds). The fusion time and fusion temperature of filled impact-modified compounds decreased and the end torque remained relatively constant as the acrylic impact modifier content increased from 3 to 12 phr. Table 5.4: Fusion characteristics of EFB-filled acrylic-impact modified compounds filled with 20 phr of filler at 50 rpm and 180 oC 27.3 Fusion Temperature (oC) 178 60 28.6 177 20 50 28.4 176 9 20 36 28.4 174 12 20 26 29.8 171 S10 Acrylic content (phr) 0 EFB filler content (phr) 0 Fusion Time (s) 68 S14 3 20 S18 6 S22 S26 Sample End Torque (Nm) 207 Table 5.4 also shows that the fusion time of filled impact-modified compounds was progressively decreased to reach the fusion time of the acrylic-impact modified compounds (Table 5.2) as the impact modifier increased from 0 to 12 phr. The difference of fusion time between both the filled impact-modified (Table 5.4) and acrylic-impact modified compounds (Table 5.2) get smaller with increasing impact modifier content. The fusion time of both compounds was similar at impact modifier content of 12 phr. It indicates that the presence of 20 phr filler content did not affect the fusion time of the filled compounds containing 12 phr acrylic impact modifier. The contribution of acrylic in increasing the particles interaction and frictional heat in the filled impact-modified compounds allowed the PVC resin particles to fuse at the shorter fusion time when the impact modifier content increased. The results also highlight that the influence of acrylic was more predominant than fillers in controlling the fusion rate. According to Chen and Lo (1999), the decrease of fusion time resulted in decreasing the fusion temperature of the PVC compound, due to the less thermal energy was needed to be absorbed by the compound in order for the PVC resin particles to fuse together. Because of this, the melt viscosity of PVC compound was increased. In other words, the melt took a longer time to homogenize and stable when the torque decreased. Thus, the similar reason can be used to explain a slight increase in melt viscosity of filled acrylic-impact modified compounds in Table 5.4. Since the fusion temperature of these compounds was slightly higher, thus these compounds showed a slight lower in melt viscosity compared to the acrylic-impact modified compounds (Table 5.2). Similar to the acrylic-impact modified compounds, the homogeneous melt of these compounds was achieved rapidly. 5.3.1.3 Effect of CPE Impact Modifier Table 5.5 shows that the fusion time of CPE-impact modified compounds with increasing CPE content from 0 to 12 phr. Interestingly, in general, the fusion time of these compounds decreased as the CPE content increased. The fusion time was 208 significantly decreased when the CPE content reached 12 phr. Similar trend has been reported by other researchers (Chen and Lo, 1999). Based on the trend of fusion time shown in Table 5.5, CPE was not very effective to decrease the fusion time at lower content in range of 0-9 phr. CPE can only perform as a processing aid at 12 phr, promoting the fusion process of PVC resin particles to occur faster and uniformly. This also agreed by Chen et al. (2001) in their studies where they found that the CPE impact modifier did not perform as an effective processing aid at a concentration of lower than 10 phr. Table 5.5 also shows that CPE was not effective as the acrylic impact modifier (Table 5.2) in promoting the fusion time of PVC resin particles. The reduction of fusion time of CPE-impact modified compound was about only 42 % (compared to acrylic-impact modified compound which was 73 %) as the impact modifiers content increased from 0 to 12 phr. The reason is that the CPE is less compatible with the PVC resin particles, that resulted in less inter-particles interactions to generate the frictional heat in the compounds. Besides that the combined lubricant effect between calcium stearate and CPE was probably insufficient to the bind the PVC resin particles together easily. Table 5.5: Fusion characteristics of CPE-impact modified compounds at 50 rpm and 180 oC Sample CPE content (phr) Fusion Time (s) End Torque (Nm) S0 0 96 31.1 Fusion Temperature (oC) 177 S5 3 88 29.4 177 S6 6 90 28.6 178 S7 9 88 28.2 179 S8 12 56 28.4 178 209 As mentioned earlier, the fusion time has a proportional relationship to the fusion temperature of compound. However, this relationship was found to be not proportional for the fusion characteristics of the CPE-impact modified compound. The fusion temperatures remained relatively constant while the fusion time of compounds decreased as the CPE content increased. This may be that the frictional heat generated by CPE particles in the compounds did not give enough contribution to assist PVC resin particles to fuse. A more significant amount of thermal energy was supplied by the heaters during processing. Since the fusion temperature was not significantly changed, the end torque of the compounds decreased marginally as the CPE content increased. This shows that the homogeneous melt of compounds was also rapidly achieved before seven minutes of time taken. The end torque values of CPE-impact modified compounds were slightly lower than acrylic-impact modified compounds (Table 5.2) due to the slightly longer fusion times and higher fusion temperatures of CPE-impact modified compounds. In general, the melt viscosity of these compounds decreased with increasing CPE content. Table 5.6 shows the fusion time of EFB-filled CPE-impact modified PVC-U compounds added with 9 phr of CPE. The fusion time of these compounds increased as the filler content increased from 0 to 40 phr. The fusion times of these compounds were higher compared to the fusion times of EFB-filled PVC-U compounds shown in Table 5.1. The decrease in the adhesive property of calcium stearate due to the addition of CPE caused the fusion time to increase. The ability of CPE in reducing the adhesive property of calcium stearate has also been reported by Chen et al. (1995). The predominant effect of EFB filler in increasing the fusion time of the compounds occurred at 30 and 40 phr filler content. At these filler loadings, EFB fillers decreased the inter-particles friction among the PVC resin particles in the compounds by increasing the separation and slipping particles in the compounds. In this case, EFB filler performed as a fusion delayer for EFB-filled PVC-U compounds 210 with 9 phr of CPE. Since the fusion time was higher, more thermal energy was needed to be absorbed for PVC resin particles to fuse. This thermal energy resulted in the fusion temperature becoming higher, since less frictional heat was contributed by the CPE in the compounds. Table 5.6: Fusion characteristics of EFB-filled CPE-impact modified compounds added with 9 phr CPE at 50 rpm and 180 oC S7 CPE content (phr) 9 EFB filler content (phr) 0 Fusion Time (s) End Torque (Nm) 88 28.2 Fusion Temperature (oC) 179 S37 9 10 80 30.3 180 S38 9 20 82 30.8 180 S39 9 30 150 32.3 186 S40 9 40 160 32.9 189 Sample In contrast to the previous compounds, the end torque of these compounds increased as the filler content and fusion temperature increased. It shows that the homogeneous melt of these compounds could not be achieved within seven minutes. The delaying of fusion time affected the time taken of PVC resin particles to break down and stabilize completely in order to achieve the homogeneous melt. Because of this, the end torque of these compounds increased and was higher than other compounds. It indicates that the melt viscosity of these compounds increased with increasing EFB filler content. Table 5.7 shows the effect of 20 phr of EFB filler on the fusion characteristics of CPE-impact modified PVC-U. The fusion time increased as the compound added with 3 phr CPE, but remained relatively constant as the CPE content increased from 3 to 12 phr. It shows that the CPE delayed the fusion time of compounds. In this case, it possible that CPE performed as an external lubricant. Similar trend has been reported 211 by other researchers (Chen and Lo, 1999 and Chen et al.2001). This may be due to the two factors. The first factor is that the less compatibility between the PVC resin particles and CPE particles, and secondly the CPE particles decreased the adhesive property of the calcium stearate. The fusion temperature of these compounds increased marginally as the CPE content increased. This shows that CPE did not contribute significantly to the generation of frictional heat, which can accelerate the fusion process of the PVC resin particles. The end torque increased as the CPE content increased from 0 to 6 phr and then remained relatively constant with further increase of CPE content from 6 to 12 phr. The delaying of the fusion process resulted in the end torque of the compounds to increase slightly. The compounds took slightly longer to completely stabilize and achieve the homogeneity. The melt viscosity of these compounds also increased with the addition of CPE. Table 5.7: Fusion characteristics of EFB-filled CPE-impact modified compounds filled with 20 phr of EFB at 50 rpm and 180 oC 0 EFB filler content (phr) 20 Fusion Time (s) 68 End Torque (Nm) 27.3 Fusion Temperature (oC) 178 S30 3 20 84 28.7 179 S34 6 20 84 30.8 180 S38 9 20 82 30.8 180 S42 12 20 86 30.4 181 Sample CPE content (phr) S10 212 5.3.1.4 Effect of Temperature Table 5.8 shows the effect of processing temperatures on the fusion time of compounds. It shows that the fusion time of the compounds decreased as the temperature increased from 175 0C to 185 0C. The increase of thermal energy due to the increase of temperatures was attributed to the decrease in fusion time for the compounds. This thermal energy contributed to the enhancement of PVC chains mobility that resulted in the rate of primary particles cohesion increased. Consequently, PVC resin particles fused at a shorter time. Another reason is the adhesive property of calcium stearate was enhanced significantly as the temperature increased. The mobile layer, which increased the separation and slipping among PVC resin particles, caused by the non-polar hydrocarbon chain of calcium stearate and stearic acid, was also destroyed (Chen et al., 1995; Chen and Lo, 1999; Chen et al., 2001). Therefore the fusion process of the compounds took place at shorter time at higher temperature. Table 5.8: Fusion time of PVC-U compounds at different processing temperatures and constant rotor speed of 50 rpm Fusion Time (s) Sample S0 S10 S3 S22 S7 S38 EFB filler content (phr) 0 20 0 20 0 20 Acrylic content (phr) 0 0 9 9 0 0 CPE acrylic content (phr) 0 0 0 0 9 9 175 104 88 42 54 94 102 180 96 68 30 36 88 82 185 50 52 18 42 52 78 Temperature (oC) 213 PVC-U compound (S0) took a longest time to fuse at 175 0C compared to other compounds due to the calcium stearate and stearic acid in the PVC-U compound were able to develop a mobile layer. This mobile layer resulted in the increase of fusion time since less mechanical energy during mixing process was converted to thermal energy (Chen et al.,1995; Chen and Lo,1999). This reason can also be used to explain the longer fusion time of the EFB-filled CPE-impact modified PVC-U composite (S38). The fusion time of PVC-U compound at 185 0C was almost similar to the fusion time of EFB-filled PVC-U composite (S10) and CPE-impact modified PVC-U compound (S7). In this case, calcium stearate in these compounds performed as a processing aid due to its adhesive property was increased at this temperature. Table 5.8 also shows that the fusion time of acrylic-impact modified PVC-U compounds (S3) at 185 oC was the shortest compared to other compounds. The synergistic combined lubricant between calcium stearate and acrylic impact modifier was able to form a powerful viscous material, which appeared to act as an adhesive that allowed the PVC resin particles to fuse together easily. Since the fusion time of acrylic-impact modified compounds was the shortest compared to other compounds at all temperatures, it can be deduced that the acrylic impact modifier is the most efficient additive in reducing the fusion time of PVC-U compounds and followed by the EFB filler and then CPE impact modifier. Table 5.9 shows the end torque of PVC-U compounds with varying processing temperatures. In general, the end torque decreased as the temperature increased. The higher temperature did not only reduce the fusion time but also decreased the melt viscosity of the compounds. The reduction of viscosity was attributed to the greater molecular mobility due to the higher thermal energy supplied for the PVC particles resin during mixing process. Table 5.9 also shows that in general the end torque of EFB-filled CPE-impact modified compounds was the highest at all processing temperature. It means that the combination between EFB and CPE increased the melt viscosity of compound. This is 214 may be due to the combined lubricant effect between EFB and CPE had increased the time taken of the compound to fuse and stabilize completely and formed a homogeneous melt. The end torque of unfilled CPE-impact modified compound at 185 0C was the lowest. It shows that the CPE was effective in reducing the melt viscosity of compound at this temperature. Table 5.9: End torque of PVC–U compounds and composites at different temperatures and constant rotor speed of 50 rpm End Torque (Nm) Sample S0 S10 S3 S22 S7 S38 EFB filler content (phr) 0 20 0 20 0 20 Acrylic content (phr) 0 0 9 9 0 0 CPE acrylic content (phr) 0 0 0 0 9 9 175 29.7 30.5 31.6 32.3 29.0 32.0 180 31.1 27.3 30.3 28.4 28.2 30.8 185 27.7 26.2 29.5 28.0 25.1 30.0 o Temperature ( C) 5.3.2 Postulated Lubrication Mechanisms The concepts of lubricant mechanisms, similar to Robinovitch et al. (1984) and Chen et al. (1995) as shown in Figure 2.8 (Chapter 2), are proposed based on the present study. Previous studies have used the paraffin wax or polyethylene oxide and CPE as an external lubricant and impact modifier, respectively. In this study, the stearic acid, acrylic, and EFB were used as the internal lubricant, impact modifier, and filler, respectively. Thus, a slight addition and modification on the previous lubrication mechanisms was done. 215 PVC is known as a polar thermoplastic polymer due to the presence of C-Cl groups in its molecular chains. Because of this polarity, the molten PVC is attracted and stuck to the polar metal surface of processing equipment (Figure 5.5a). As mentioned in Chapter 2 (Section 2.2.2), the strong polar ends of calcium stearate molecules adhere to the polar PVC and also the polar metal surface, and leaving the non-polar tails to perform the lubrication. The non-polar tails of calcium stearate displace the PVC melt from sticking on the metal surface and providing limited slip between the metal and PVC melt surfaces (Figure 5.5b). The polar groups in calcium stearate are also strongly attracted to each other and form a viscous layer structure, which is similar to a liquid crystal (Figure 5.5b)(Chen et al., 1995). The stearic acid with HCOOC groups, which was used as an external lubricant in this study, is also polar material. Therefore, in the presence of stearic acid, the polar groups of stearic acid will also attach to the metal surface, to the surface of PVC resin particles, and to each other, while the long-chain aliphatic acid provides the lubrication effect. The combined interaction between calcium stearate and stearic acid in the compound is also illustrated in Figure 5.5b. The external lubricants, however, have a lower polarity than the internal lubricants (Marshall, 1969). Thus, the ability of calcium stearate to hold PVC resin particles together in the compounds is better than stearic acid. The acrylic impact modifier is known to be highly compatible with PVC. The polarity of PMMA in the acrylic shell is higher, thus a better interfacial adhesion (compatibility) between PVC particles and acrylic shell is formed. The interaction between acrylic shell and PVC is also well distributed (Figure 5.5c) due to the ability of acrylic to disperse in PVC matrix (Lutz, 1993). Therefore, acrylic shell acts evenly as a bridge and appears to increase the number of ties between the PVC particles. The acrylic core remains in between PVC particles. It acts as a grit to increase the frictional heat in the compound (Figure 5.5c). The acrylic shell is also attracted and stuck to the metal surface. Due to the polarity, the acrylic shell attracted to the calcium stearate to form a powerful adhesive property to bind the PVC particles together (Figure 5.5f). 216 In contrast with acrylic, the interaction between CPE and PVC resin particles is insufficient to hold the PVC particles together due to it less compatible to PVC (Figure 5.6c). Due to the C-Cl group in the molecular chains, CPE is also attracted and attached to the polar metal surface. The shorter fusion time and higher melt viscosity for acrylic-impact modified compounds (Table 5.2) compared to CPEimpact modified compounds (Table 5.5) may be traced to two possible factors: (a) good compatibility and dispersion and (b) combined lubricant effect with calcium stearate. The EFB is a highly polar filler. Although the polarity of EFB filler makes them capable of forming a physical interaction with polar PVC (Woodham et al., 1995), it is a relatively weak interaction compared to the strong filler-filler interaction caused by hydrogen bondings. Therefore, EFB fillers have a greater tendency to attract among themselves to form the filler bundles and leave in between the PVC particles (Figure 5.5d). Thus, there is a slight attraction between EFB fillers surface to the PVC particles. In this case, there are two significant factors that can be contributed by EFB filler. The first factor is it increases the frictional heat and the second factor is it increases the separation among PVC resin particles especially at higher filler loading (Figure 5.6g). Interestingly, the two factors can give opposing effect to the fusion time of PVC-U compound. Due to the polarity, the EFB fillers are attracted to calcium stearate, acrylic, and CPE (Figure 5.5g and Figure 5.6g). The ability of EFB filler to increase the frictional heat and interact (due to the presence of ester component of oil residue) with calcium stearate to form a fairly good adhesion between PVC resin particles were the factors responsible for the shorter fusion of EFB-filled compound (Table 5.1) compared to CPE-impact modified compounds (Table 5.5). The increase of separation among PVC resin particles contributed by agglomeration of fillers, resulted in the melt viscosity of EFB-filled compounds to be lower than acrylic (Table 5.2) and CPE-impact modified compounds (Table 5.5). 217 (a) METAL 1µm PVC microparticle flow unit PVC PVC calcium stearate and stearic acid (b) METAL acrylic core-shell impact modifier (c) METAL EFB filler bundles (agglomerates) (d) METAL (e) METAL (f) METAL (g) METAL Figure 5.5 Postulated lubrication mechanism of among PVC micro-particles flow units, EFB fillers and acrylic impact modifiers 218 (a) METAL 1µm PVC microparticle flow unit PVC PVC calcium stearate and stearic acid (b) METAL CPE impact modifier (c) METAL EFB filler bundles (agglomerates) (d) METAL (e) METAL (f) METAL (g) METAL Figure 5.6 Postulated lubrication mechanism of among PVC micro-particles flow units, EFB fillers and CPE impact modifiers 219 5.3.3 Effect of Extracted EFB Filler on the Processability of PVC-U Compound In this section, the processability of extracted EFB filler-filled compounds with difference filler content was investigated. The analysis of extracted filler, concentrated extract and solvents condensate using FTIR and SEM are reported. 5.3.3.1 Extractives Removal The percentage of extractives removed from the EFB filler was approximately 4.5 %. The effect of extraction on the appearance of solvent mixture was studied by the visual inspection. The solvents mixture changed in colour from plain to dark brown, while the extracted fillers were found to be brighter in colour than the unextracted fillers. This visual inspection indicates that the extraction has effectively extracted the extractives, which include oil residues of fillers. The analysis on the extractives and extracted fillers was carried out to verify that the extraction was successfully achieved. 5.3.3.2 FTIR Spectra The cellulose molecular structure as shown in Figure 2.11(Chapter 2) is useful for functional groups identification in the EFB filler. FTIR analysis was done on the fillers, concentrated extract, and condensate. Figure 5.7 shows the FTIR spectrum of extracted and unextracted EFB filler. Both spectrums were found to be very similar. It shows that the functional groups of the extracted and unextracted filler cannot be distinguished using this analysis. Both spectrums show a broad peak at 3393 and 3391 cm-1 due to absorption vibration of the hydroxyl (OH) group (Sarani et al., 2001). The peaks at 2919 and 220 2916 cm-1 are due to the stretching of C-H groups. The presence of cellulose is obviously seen from the sharp peak appearing at 1730 and 1731 cm-1. This is due to the stretching vibration of carbonyl group (C=O). Peaks at 1629 cm-1 and 1630 cm-1 are an indicative of the presence of lignin, an aromatic compound, and due to the C=C vibration of benzene ring.The broad peaks (both spectra) at 1050 and 1050 cm-1 are assigned to ether linkage (C-O-C) from lignin or hemicelloluse (Matuana et al., 2001). Figure 5.7 FTIR spectra: a) EFB filler b) extracted EFB filler Figure 5.8 shows the FTIR spectrum of concentrated extract and solvent condensate. By using the palm oil molecular structure as a reference (Figure 5.9), the existance of ester component in the concentrated extract was clearly detected. The sharp absorption band appearing at 1712 and 1739 cm-1 are attributed to the stretching of carbonyl groups (C=O). The sharp absorption band at 1261 cm-1 is due to 221 asymmetric stretching of C-O-C from the oil chemical structure (Conley, 1972). Since the oil chemical structure contains a long chain of hydrocarbon, thus the sharp and strong band at 2925 and 1458.87 cm-1 due to the stretching of C-H groups and bending of CH2 groups, respectively, are clearly observed (Streitwieser and Heathcook, 1981). This spectrum confirmed the existence of oil residue in the concentrated extract. Figure 5.8 FTIR spectra (a) solvents mixture condensate (b) concentrated extract 222 H O C O C16H31 Palmitic H C O C O C17H35 Stearic H C O C C17H35 Stearic H C O H Figure 5.9 Palm oil molecular structure 5.3.3.3 Extracted Filler Surface By visual observation, the extracted EFB fillers were brighter in colour compared to the EFB fillers. Both fillers were also analyzed using SEM to see the effects at higher magnification. Figures 5.10 and 5.11 show the SEM micrographs of unextracted and extracted filler surface, respectively. The surface of the extracted filler was coarser than the unextracted. It was also observed that the extracted filler surface contained a lot of pits. The difference is believed due to the removal oil and other extractives from the extracted filler surface. 223 Smoother Surface Figure 5.10 SEM Micrograph for EFB filler (Magnification 3000x) Coarser Surface Figure 5.11 SEM micrograph for extracted EFB filler (Magnification 1000 x) 5.3.3.4 Fusion Characteristics of Extracted EFB Filled PVC-U Compound Table 5.10 shows the effect of extracted EFB filler content on the fusion characteristics of extracted filler-filled PVC-U compound. Similar to the unextracted fillers, the fusion time of extracted fillers of filled compounds is shortened than the unfilled compounds. 224 Table 5.10 also shows that the fusion time of extracted filler filled-compound decreased drastically upon the incorporation at 10 phr and increased linearly with further increase of filler content from 10 to 40 phr. It shows that the extracted filler was also able to reduce the fusion time of filled compound. Based on the fusion time of unextracted filler-filled compounds in Table 5.1, the extracted filler-filled compound took a shorter time to fuse compared to the unextracted filler as the filler content increased from 10 to 20 phr. It showed that the coarser surface of fillers (Figure 5.11) was more effective to enhance the frictional heat of filled compounds compared to unextracted fillers. As the filler content increased from 30 to 40 phr, the separation and slipping among the PVC resin were also increased. The lack of oil residues resulted in the adhesive property of calcium stearate additive became weaker as filler content increased. Therefore, calcium stearate took a longer time to bind PVC resin particles together in extracted filler-filled compounds. Consequently, the fusion time of extracted filler-filled compounds increased and slightly higher than the unextracted filler-filled compound as filler content increased from 30 to 40 phr. Since the fusion time increased, the fusion temperature also increased and resulted in the end torque of the compound to decrease. The slightly higher end torque of these compounds compared to unextracted filler-filled compounds in Table 5.1 is due to extracted fillers not being coated with oil residues. It elucidates why the melt viscosity of the extracted filler-filled compounds was lower than unextracted filler. Table 5.10: Fusion characteristics of extracted EFB-filled PVC-U compounds at 50 rpm and 180 oC S0 Extracted filler content (phr) 0 Fusion time (s) 96 End torque (Nm) 31.1 Fusion temperature (oC) 177 S45 10 44 30.8 176 S46 20 58 27.5 177 S47 30 66 27.7 178 S48 40 78 27.0 180 Sample 225 5.4 Conclusions The main objective of this chapter is to investigate the effect of EFB filler, acrylic and CPE impact modifier, and temperatures on the fusion characteristics of PVC-U compounds. Based on the results obtained, several conclusions can be drawn: (a) The fusion time of PVC decreased as both the EFB filler and extracted EFB filler were incorporated. However, the fusion time of EFB-filled PVC-U composite remained relatively constant as the filler content increased from 10 to 40 phr. The frictional heat generated by the filler and followed by the increase of the adhesive property due to the synergistic combined lubricant between of calcium stearate and oil residue probably nullified by the increase of separation among PVC resin particles. The fusion time of extracted EFB filler-filled compound increased with increasing filler content from 10 to 40 phr. The increase of separation and slipping among the PVC resin particles and the lack of oil residues weakened the adhesive property of calcium stearate. The weakened calcium stearate was attributed to the increase of fusion time. (b) The acrylic impact modifier performed as a fusion promoter for all compounds whereas CPE changed from a fusion promoter to a fusion delayer in the formulations contents of 9 phr CPE and 30 phr or 40 phr EFB filler. The fusion time of PVC compound added with 9 phr acrylic at 185 oC is the shortest while the longest fusion time is produced by the PVC compound added with 9 phr CPE and 40 phr EFB filler at 180 oC. (c) The incorporation of EFB filler decreased the melt viscosity of the compounds. The melt viscosity of EFB-filled compound is the lowest and followed by the extracted filler-filled compound. The PVC-U compound 226 containing 9 phr CPE and 40 phr EFB at 180 oC produced the highest melt viscosity. (d) The increase of processing temperature decreased significantly the fusion time and melt viscosity of the compounds due to the increasing of thermal energy. (e) The lubrication mechanisms of the studied compounds have been postulated. Due to the polarity of molecules, calcium stearate was able to interact effectively with PVC. This result in a strong adhesion, which binds the PVC resin particles together and cause them to fuse at a shorter time. On the other hand, the increase in fusion time with increasing filler content is due to the increase of separation and slipping among the PVC resin particles. CHAPTER 6 EXTRUSION PROCESS AND MECHANICAL PROPERTIES OF EXTRUDATES 6.1 Introduction As an amorphous thermoplastic polymer, PVC does not exhibit a distinct melting point but gradually soften above its glass transition temperature. The extrusion process consists of softening the PVC by heat, fusing it together and forcing it through a die of the required dimensions, followed by cooling of the melt in order to retain the shape. The mechanical properties of PVC-U extrudates produced by single screw extruder have been studied by Covas and Gilbert (1992). In their study, the effect of fusion on the mechanical properties of extrudates has been investigated. Owing to the significant level of shear present in single screw extrusion, the route to fusion seems to be different from that observed in twin screw extrusion. Therefore, the hypothesis for the fusion of PVC-U compounds in single screw extruder has been developed from that originally derived by Allsopp (1982). The mechanism of fusion is illustrated in Figure 2.4 (Chapter 2). The melting of solid materials occurred either by comminution of grains with internal structure substituting or by ruptures of internally 228 fused grains. In other words, the comminution, compaction and internal fusion of grains could be observed simultaneously in the extruder machine. Progressive fusion and gradual increase in melt homogeneity were postulated. Network structures of entanglements, primary and secondary crystalline were also present upon cooling. It is well known in the PVC profile extrusion industries that the mechanical properties of the product are related to the shear and temperature during extrusion. The fusion process of PVC is strongly influenced by these two factors. Increasing shear in the extruder can be done by increasing the screw speed. However, the residence time decreased counteracting the effect of the shear increased. In the previous study, Summers (1982) found that the morphology and toughness of a PVC compound in single screw extrusion was only slightly affected by shear rate but dominated by melt temperature. The toughness of PVC compound went through a maximum then decreased with further increasing of melt temperature. The decrease of this toughness was due to the failure of lubrication system, which resulted in the melt fracture on the extrudates. The similar observation has also been reported by Lacatus and Rogers (1986). They found that the optimum toughness of extrudate was obtained at melt temperature of 177 oC (at lower fusion level). At 204 oC, the melt fracture of the extrudates which was the reason for the toughness decreasing. Other researchers also studied the effect of fusion on the tensile and impact properties of extrudates. Covas and Gilbert (1992) reported that the yield parameters were insensitive to the level of fusion whereas the impact strength increased with increasing of the level of fusion and exhibited a maximum value. Similar results have also been extensively reported by other researchers (Benjamin, 1980; Uitenham and Geil, 1981; Terselius and Jansson, 1981, 1984, 1985). They concluded that in order to obtain the optimum mechanical properties, an appropriate level of fusion was needed. It means that although during processing the original PVC grains particles was progressively changed into a network of entanglements, primary and secondary crystallinity, but the presence of some remaining primary particle was required in order to achieve the optimum properties. 229 Despite the known effect of fusion level on the mechanical properties of PVC, a limited number of studies have been carried out on the influence of screw speed and processing temperature on the fusion level and mechanical properties of the EFBfilled PVC-U extrudates. With this in mind, the main objective is to study the mechanical properties of unfilled and filled extrudates after being extruded using a single screw extruder. 6.2 Experimental In this section, the materials and blend formulations used are similar to Chapter 3 (Section 3.1). The dry blending, extrusion processing, mechanical testing, and DSC analysis are described. 6.2.1 Materials Six blends formulations were prepared for this study. The blend formulations, namely PVC-U compound (S0), acrylic-impact modified PVC-U (S3), CPE-impact modified PVC-U (S7), EFB-filled PVC-U (S9 and S10) and impact modified EFBfilled PVC-U composites (S21, S22, S37 and S38) as shown in Table 3.3 (Chapter 3) were used. 12 kg of each formulation was prepared to meet the requirement of extrusion process. 230 6.2.2 Dry Blending The formulations were dry blended using a Lab-Tech 5L high-speed mixer of the impeller type and water-jacketed. The formulations were dry blended in 4 kg batches, using the following sequence: 1 min at low speed and followed by another 1 minute at high speed. The temperature was built up during mixing process, due to the generation of frictional heat, thus the blend was discharged from the mixer chamber after 2 minutes blending in order to avoid the temperature exceeding 60 oC. Above this temperature, the blend became sticky and stuck on the mixer chamber wall and impeller. 6.2.3 Extrusion Process Each PVC-U dry blend was oven dried at 105 oC for 24 hours before they were compounded on a single screw extruder at two conditions as shown in Table 6.1. For the first condition, the extrudates were extruded at constant screw speed with difference processing temperatures. The extrudates were also extruded at constant temperature with difference screw speeds as a second condition. The ratio of the length of the flighted portion of the screw to the outside diameter of the screw flights (L/D) was 27:1. The slit die with 1.5 mm in aperture and 100 mm in width was attached to the extruder. The extrudate in the stripe form with approximate dimension of 2.4 mm in thickness and 100 mm in width was cooled in the ambient temperature after leaving the die. The extrudates were firstly cut into bars using a band saw machine before they were shaped according to the test specimens using the templates and cutting machine. The cutting sides of specimens test were polished with a sand paper before carrying out the mechanical testing. 231 Table 6.1: The conditions of extrusion processing Screw Speed (rpm) 60 Feed Zone 160 Constant Screw Speed Processing Temperature (oC) Compression Zone Metering Zone 170 180 Die 180 60 165 175 185 185 60 170 180 190 190 60 175 185 195 195 Constant Processing Temperature 50 160 170 180 180 60 160 170 180 180 70 160 170 180 180 6.2.4 Izod Impact and Flexural Testing The dimensions of the test specimens for the impact and flexural testing were similar to the ones reported in Chapter 3 (Sections 3.2.9-3.2.10) except for the specimen thickness. The average thickness of the specimen used for both testing was 2.4+0.2 mm. The notched specimens were impact tested in the Izod mode using a Toyo-Seiki Pendulum Impact Tester with impact velocity of 3.0 m/s and 150o swing angle using a 31.87 N hammer at room temperature. The machine and method used for the flexural testing were also similar to Chapter 3 (Section 3.2.10). The reported values for the testing were average of seven specimens. 232 6.2.5 Tensile Testing The tensile properties, in accordance with the ASTM D638-91(Type 1) method, were measured using a Lloyd Instrument model EZ20 (cell load of 20 kN) with four variations of crosshead speed, namely of 5, 20, 80 and 100 mm/min. The specimen dimensions used were 2.4+0.2 mm in thickness, while the gauge length used was 60 mm. The reported values for the test were also average of seven specimens. 6.2.6 Differential Scanning Calorimetry Study A Mettler Toledo Differential Scanning Calorimetry (DSC) instrument model 821 as reported in Chapter 4 (Section 4.2.4) was also used in this study. The similar experimental procedure was also utilized for this study. The extrudates that were processed at difference processing temperatures were analysed to determine the heat of fusion. 6.3 Results and Discussion In this section the influence of extruder screw speeds and processing temperatures on the impact, flexural and tensile properties of extrudates will be discussed. This section will also highlight the relationships between the processing temperatures and heat of fusion. The die temperature as already shown in Table 6.1 will represent the processing temperatures. 233 6.3.1 Extrudates Characteristics The blend formulations of S0, S3, S7, S9 and S21 were extruded using a single screw at the processing conditions as shown in Table 6.1. The extrudates with smooth surfaces were produced in this study are shown in Figure 6.1. Figure 6.1 Extrudates of S0, S3, S7, S9 and S21 at die temperature of 185 oC, and extrudate of S37 at die temperature of 190 oC Although the dimension of slit die used was 1.5 mm x 100 mm, the dimensions of extrudates produced were slightly larger due to the swelling of extrudates as it exited the slit die. This phenomenon is called die swell and illustrated in Figure 6.2. The die swell was caused by the viscoelastic nature of the polymer melt that dissipates forces slowly. The compressive forces that were needed to push the polymer melt through the land and small die were not completely relieved by the time the polymer exits the die. Therefore, the polymer expanded when it exited the die in response to the residual compressive forces. The die swell resulted in the extrudate surface being slightly curved at the top and flat at the bottom. The auxiliary 234 equipment attached with 1.85 kg rotating roller was used to haul off the melt from the extruder in order to minimize the die swell. Extrudate (top surface) Slit Die Figure 6.2 Shape difference between the die and extrudate Unlike other extrudates (S0,S3,S7,S9,S21), the extrudate of S37 which contained 10 phr EFB and 9 phr CPE were severely distorted shortly after exiting the slit die at die temperature of 180 0C and 185 0C (Table 6.1). The distorted extrudate is known as a bambooing defect (a type of melt fracture), which can be seen in Figure 6.3. It might be that the fusion time of S37 which contains CPE and EFB fillers are relative long during extrusion and some of the PVC remained in particulate form, resulting in an increase of melt viscosity. As the melt was flowing, it developed a flow profile with the centre of the melt flowing faster than the edges, because of the friction of the melt against the edges of the die. The melt in the centre of the flow slid against itself easier than it slid against the die walls. Because of the difference in velocity between edges and centre, as the melt lef the die, it was distorted somewhat like a bamboo. This also generated tensile stress and if the stress exceeds the melt strength, the edges of extrudate rupture causing the defect. The bambooing defect and the ruptured edges of S37 are shown in Figure 6.3. The defects were eliminated as the die temperature increased to 190 oC (Figure 6.1) and 195 oC, whereby the extrudates with smoother surface were produced. It indicates that at these temperatures the melt viscosity of S37 was reduced, as a result of the sufficient fusion process taking place (Strong, 2000). 235 Figure 6.3 Filled extrudates of S37 at die temperature of 180 oC and 185 oC and at constant screw speed of 60 rpm. Before the extrusion, the dry blends were oven dried and left to cool for a few minutes in the closed oven before being used for extrusion processing. If the drying process was insufficient, the air bubbles generated from the moisture of dry blends were formed during extrusion. Since the vent port was unavailable at the compression zone of the extruder, the air bubbles was unable to escape during the transition process of solid phase to melt phase. Shortly after the melt exited from the die, the air bubbles inside the melt started to move up to the top of the surfaces. Consequently the extrudates with air bubbles on the surface were produced. The air bubbles randomly distributed on the extrudates surfaces were observable as shown in Figure 6.4. 236 Figure 6.4 Filled extrudate of S9 with air bubbles on surface Rough surfaces were observed on the extrudates incorporated with 20 phr EFB (S10, S22 and S38). The roughness of these filled extrudates was greatly noticeable at the bottom surface. To eliminate this defect, the screw speed was increased up to 100 rpm at constant die temperature of 195 oC. However, no significant changes were observed on the quality of extrudates surfaces produced as shown in Figure 6.5. Based on the visual inspection on the filled extrudates surface, it was found that EFB fillers were poorly distributed and poorly dispersed throughout the matrix. The fillers precipitated at the bottom surface due to the EFB being slightly denser than PVC resin (as reported in Chapter 3). The small L/D ratio of the screw used in this study is also one of the reasons to explain the occurrence of the surface defect. The L/D is a measure of the ability of the screw to mix materials with high L/D ratio promoting good mixing and good melting capabilities (Strong, 2000). Thus, in PVC extrusion, high L/D ratio is required to ensure the dry blends have sufficient residence time for homogeneous mixing and melt before leaving the die. The results showed that the L/D ratio (27:1) of the screw used in this study was not sufficient to produce good quality filled extrudates containing 20 phr EFB filler. 237 Figure 6.5 Filled extrudates of S10, S22 and S38 at die temperature of 195 oC. Tables 6.2 and 6.3 show the residence time of extrudates at constant die temperature of 180 oC and at constant screw speed of 60 rpm, respectively. Table 6.2 shows that the residence time decreased as the screw speed increased from 50 to 70 rpm. On the other hand, the residence time of the extrudates was not significantly changed as the processing temperature increased at constant screw speed (Table 6.3). These results show that the residence time was strongly dependent on the screw speed used during extrusion as the drag force increased. Table 6.2: Residence time of extrudates at constant die temperature of 180 oC Residence Time (s) Speed (rpm) So S3 S7 S9 S21 50 60 62 63 61 63 60 55 53 54 53 50 70 41 43 42 43 44 238 Table 6.3: Residence time of extrudates at constant screw speed of 60 rpm Residence Time (s) Tdie (0C) So S3 S7 S9 S21 S27 180 55 53 54 53 53 - 185 53 57 56 54 50 - 190 54 58 57 55 53 60 195 53 58 58 53 53 55 6.3.2 Differential Scanning Calorimetry It has been reported that the differential thermal analysis can be used to evaluate the level of fusion of processed PVC samples (Gilbert and Vyvoda, 1981; Gilbert, et al. 1983). On reheating a sample of processed materials, melting of the secondary crystalline, which contributed to network development, might be detected. The detected endotherm in the DSC thermogram represented the melts of secondary crystallite. A typical thermogram for the unfilled PVC-U extrudate at difference processing temperatures is shown in Figure 6.6. It can be seen that the area of the endotherm, which represents a secondary crystallite population generated during cooling, increased with increasing processing temperature. This is in agreement with the results obtained by other researchers for the processed PVC compound (Gilbert and Vyvoda, 1981; Gilbert, et al. 1983; Cora et al.1999). The energy (heat of fusion) of the endotherm increased with processing temperatures, due to the increase of breakdown of secondary crystallites. The increase of endotherm also indicated the number of primary particles melts increased. Thus, the heat of fusion obtained from the endotherm was also related to the level of fusion. Similar results on the increase of 239 endotherm area with increasing processing temperature were shown by other extrudates. The other interesting and useful information available in DSC analysis is that the endotherms end-point temperatures of extrudates were equal to the maximum processing temperature used in the extrusion processing. This is a valuable internal measure of processing temperature (Gilbert, 1985). Interestingly, the results in Figure 6.6 confirm that maximum processing temperatures indicated by the DSC analysis were similar to the real die temperature during extrusion. End-point exo 195oC Heat Flow End-point 190oC 185oC End-point 180oC End-point Figure 6.6 A typical DSC thermogram of S0 at different processing temperatures Table 6.4 shows the heat of fusion of extrudates increased with increasing processing temperature, which indicates the degree of fusion increased with increasing processing temperature. A similar observation has been reported by other researchers (Benjamin, 1980; Gilbert et al., 1983; Axtell et al., 1994). It was found 240 that both EFB filler and acrylic acted synergistically in S21 to generate the highest heat of fusion at all processing temperatures. There are two reasons, namely the frictional heat generated by fillers and the ability of acrylic to accelerate the fusion process which were attributed to the increase of heat input of PVC to undergo fusion process. These reasons were strongly supported by the higher of heat of fusion S3 and S9 compared to S0. Table 6.4 also shows that the addition of CPE into S7 has a similar effect as acrylic in increasing the PVC heat of fusion, even though it was not as effective as acrylic. The effectiveness of acrylic to accelerate the fusion process of PVC compared to CPE has been discussed in previous chapter. Table 6.4: Heat of fusion of extrudates at screw speed of 60 rpm Heat of Fusion (J/g) S7 S9 2.68 1.91 Tdie (oC) 180 S0 2.23 S3 4.15 185 2.53 5.81 2.93 190 2.82 5.49 195 5.71 5.93 S21 4.52 S37 - 4.49 8.02 - 4.72 4.72 9.05 4.89 4.82 7.73 9.38 6.92 6.3.3 Impact Strength 6.3.3.1 Effect of Processing Temperature Table 6.5 shows the effect of processing temperatures on the impact strength of extrudates at screw speed of 60 rpm. The result shows that all extrudates underwent complete fracture at all temperatures except for S3. At 195 oC, the S3 extrudates which were added with acrylic impact modifier fractured completely giving a low impact strength. However, at 190 oC and below, S3 extrudates failed in a mix mode with some hinge break and some complete fractured. This shows that between 180 oC and 190 oC, 241 the extrudates were in ductile to brittle transition. Therefore, the experimental error for S3 was higher at these temperatures. The impact strength of S3 as shown in Table 6.6 is the highest compared to other extrudates due to the effectiveness of acrylic to enhance the impact strength. Table 6.5: Effect of processing temperatures on the fracture properties of unfilled and filled extrudates Tdie (oC) 180 Complete Fractured Samples (no. of samples) So S3 S7 S9 S21 S37 7 3 7 7 7 - 185 7 5 7 7 7 - 190 7 3 7 7 7 7 195 7 7 7 7 7 7 Table 6.6: Effect of processing temperatures on the heat of fusion and impact strength of unfilled extrudates Tdie. (oC) Heat of Fusion (J/g) Impact Strength (kJ/m2) So 4.7+0.7 Heat of Fusion (J/g) Impact Strength (kJ/m2) Impact Strength (kJ/m2) Heat of Fusion (J/g) S7 4.15 S3 82.2+40.1 2.68 15.5+1.6 180 2.23 185 2.53 5.4+0.4 5.81 48.2+43.5 2.93 7.7+3.9 190 2.82 5.3+0.5 5.49 63.1+37.3 4.72 9.3+1.2 195 5.71 4.8+0.6 5.93 15.8+2.1 4.82 9.7+2.4 Table 6.6 shows the effect of processing temperatures on the impact strength of extrudates. The impact strength of S0 remained relatively constant as the temperature increased from 180 oC to 195 oC, even though the heat of fusion at 195 o C significantly increased to 5.71 J/g. It shows that these processing temperatures did not affect the impact strength of this extrudate. 242 For S3, there were mixtures in the failure more for the samples at 180 oC, 185 o C and 190 oC. Therefore, the impact strength of S3 at these temperatures was fluctuated. Meanwhile, the impact strength of S3 dropped drastically as the temperature increased from 190 oC to 195 oC. It seems not to be in consistent with previous report. It was previously reported that the effectiveness of acrylic to enhance the impact strength of PVC was independent of processing temperatures when conducted at 175 oC, 180 oC, 185 oC and 190 oC (Robinovic, 1983). This is because the dispersed rubber particles maintained their particle morphology in the continuous PVC phase. Therefore, the high heat of fusion is the plausible factor that contributed to the reduction of impact strength at 195 oC. Summers (1982) reported that as the fusion level (heat of fusion) increased, the network structure related to the primary particles dispersed progressively, leading to the reduction of PVC free volume. It means that even though the crystallites formed from the recrystallization ties of primary particles to form a stronger network structure, which subsequently improves the toughness of PVC, the reduction of free volume nullify the contribution from the formation of crystallites. Therefore, the reduction in free volume probably was the overriding factor that resulted in the impact strength decreased as the processing temperature was increased from 190 oC to 195 oC. Similar observations have also been reported by other researchers (Cora et al., 1999). Based on the impact strength value, the optimum processing temperature of S3 was 180 oC. The impact strength of S7 dropped drastically as the processing temperature increased from 180 oC to 185 oC and remained relatively constant as the processing temperature increased from 185 oC to 195 oC. This indicates that the highest impact strength of S7 occurred at 180 oC with heat fusion of 2.68 J/g. CPE morphologies in the PVC melt was reported to be dependent on processing temperatures (Lutz, 1993). This was due to its inherent solubility in the PVC, in which CPE tends to act as a plasticizer rather than impact modifier as the temperature increased. Furthermore, at higher processing temperature, the rubbery networks of CPE started to collapse. Also, at higher processing temperature, primary particles of PVC fused completely and resulted in the toughness of PVC extrudates to decrease. 243 Table 6.7 shows the effect of processing temperatures on the impact strength and heat of fusion of filled PVC-U extrudates. The impact strength for S9, S21 and S37 was remained relatively constant as the temperature increased. However, the heat of fusions generated by S9, S21 and S27 were different where S9 has the highest heat fusion at all processing temperatures followed by S21 and S9. It indicates that the addition of 9 phr acrylic contributed to the higher heat of fusion which resulted in the higher impact strength compared to S9 and S37 at all processing temperatures. The dissimilar observation on the effect of impact modifier on the impact strength and heat of fusion can be seen at S37 extrudates. Even though heat of fusion of S37 (6.92 J/g) was slightly lower than S9 (7.73 J/g) at 195 oC, the impact strength of S37 was greater than S9. It shows that the primary particles left in S37 and the presence of CPE were able to maintain the good impact strength of this extrudate. On the other hand, the primary particles of S9 underwent more fusion process, which was indicated by its higher heat of fusion, and without impact modifier were unable to increase the impact strength of extrudate (S9). Table 6.7: Effect of processing temperatures on the heat of fusion and impact strength of filled extrudates Impact Strength (kJ/m2) Impact strength (kJ/m2) Impact strength (kJ/m2) Tdie. (oC) Heat of Fusion (J/g) 180 1.91 S9 4.4+0.3 4.52 S21 6.2+0.4 - - 185 4.49 4.7+0.5 8.02 6.6+1.1 - - 190 4.72 4.6+0.4 9.05 6.0+1.1 4.89 4.9+0.6 195 7.73 4.4+0.4 9.38 5.8+0.7 6.92 5.5+0.6 Heat of fusion (J/g) Heat of fusion (J/g) S37 Even though S9, S21, and S37 have higher values of heat of fusion compared to the unfilled PVC-U extrudates (S0 and S3), their impact strength was lower than S0 and S3. This is because of the influence of EFB fillers was significant in decreasing the impact strength of filled PVC-U extrudates than fusion level. 244 6.3.3.2 Effect of Screw Speed Table 6.8 shows the influence of screw speeds on the impact-fractured extrudates. The results show that all extrudates underwent complete fracture at all screw speeds except for S3. As reported in Chapter 3 (Section 3.3.2.1), the effectiveness of acrylic impact modifier in enhancing the impact strength of PVC compared to CPE was also shown in this study. The effect of screw speeds on the impact strength of extrudates is shown in the Tables 6.9. Table 6.8: Effect of screw speeds on the fracture properties of unfilled and filled extrudates Complete Fractured Samples (no. of sample) S7 S9 S3 5 7 7 Screw speed (rpm) 50 So 7 60 7 4 7 7 7 70 7 0 7 7 7 S21 7 Table 6.9: Effect of screw speeds on the impact strength of unfilled and filled extrudates S0 4.5+0.9 Impact Strength (kJ/m2) S3 S7 S9 80.8+30.3 15.9+1.1 4.9+0.3 S21 5.7+0.5 60 4.7+0.7 82.2+40.1 15.5+1.6 4.4+0.3 6.2+0.4 70 5.4+0.7 93.6+13.9 12.9+1.1 3.6+0.5 6.2+1.0 Screw speed (rpm) 50 Table 6.9 shows that the impact strength of all extrudates was relatively constant as the screw speed increased from 50 to 70 rpm, except for S7 and S9. It implies that the increase of shear rate (screw speed), which contributed to the increase 245 of fusion level, did not affect the impact strength of extrudates. Table 6.9 also shows that the impact strength of S3 was the highest compared to the rest of extrudates. It elucidates that S3 appeared to be more susceptible to fuse due to the presence of acrylic impact modifier when the rate of shear increased. It may be the fusion level of S3 progressively increased as the screw speed increased. Previous studies on PVC have shown that the impact strength increased with increasing fusion level. Marshall et al. (1983) reported that the increase of fusion level with the influence of screw speed was able to improve the impact strength of PVC. Covas and Gilbert (1992) also reported that although the temperature was the main effect, increasing screw speed also produced a significant jump in the level of fusion. The decreasing in impact strength of S7 may be due to the network structure of CPE to enclose primary particles that formed during the fusion process was destroyed as the screw speed increased from 60 to 70 rpm. CPE became distributed as droplets within a continuous PVC-U phase at higher shear rate (Mac Gill and Wittstock, 1987). In this case, CPE was unable to form strong ties between primary particles in order to improve the PVC toughness. Table 6.9 also shows that the impact strength of S9 decreased with increasing screw speeds. This was due to the EFB fillers poorly dispersed in the PVC-U matrix as the residence time decreased (Table 6.2). It indicates that the increase of shear rate to impart a good toughness through the increase of fusion level of S9 was inhibited by the agglomeration of fillers, which contributed to the reduction in impact strength of S9. It means that impact strength was more influenced by fusion level at lower screw speed, whereas fillers agglomeration was predominated at higher screw speed in decreasing the toughness. The addition of the acrylic in S21 was able to compensate the agglomeration of filler effect and improved the impact strength of filled extrudates. The impact strength of S21 remained constant as the screw speed increased from 50 to70 rpm. 246 6.3.4 Flexural Properties 6.3.4.1 Effect of Processing Temperature Tables 6.10 and 6.11 show the effects of processing temperatures and heat of fusion on the flexural modulus of extrudates. The flexural modulus of S0, S9 and S21 was relatively constant as the processing temperature increased from 180 oC to 195 oC. It shows that the stiffness of S0, S9 and S21 was insignificantly affected by the increase of processing temperatures and heat of fusion. Table 6.10 and 6.11 also show that the flexural modulus of S3, S7 and S37 was only significantly increased at 195 0C. The primary particles dispersion of these extrudates progressively increased at 195 oC due to the increase of heat of fusion was probably attributed to the increase of stiffness. Table 6.10: Effect of processing temperatures on the flexural modulus of unfilled extrudates Tdie. (oC) Heat of fusion (J/g) Flexural Modulus (MPa) So Heat of fusion (J/g) Flexural Modulus (MPa) Heat of fusion (J/g) Flexural Modulus (MPa) S7 S3 180 2.23 3486+23 4.15 2824+124 2.68 2792+126 185 2.53 3439+86 5.81 2892+115 2.93 2586+156 190 2.82 3363+160 5.49 2821+147 4.72 2781+164 195 5.71 3568+234 5.93 3088+81 4.82 2896+45 247 Table 6.11: Effect of processing temperatures on the flexural modulus of filled extrudates Tdie. (oC) Heat of fusion (J/g) Flexural Modulus (MPa) Flexural Modulus (MPa) Heat of fusion (J/g) S21 S9 Flexural Modulus (MPa) Heat of fusion (J/g) S37 180 1.91 3514+147 4.52 3051+63 - - 185 4.49 3609+133 8.02 2937+89 - - 190 4.72 3579+79 9.05 2926+121 4.89 2443+219 195 7.73 3613+16 9.38 2980+101 6.92 2797+157 The effect of extrusion processing temperatures on the flexural strength for all extrudates is shown in Table 6.12. The flexural strength for all extrudates was relatively constant as the temperature increased from 180 oC to 195 oC. It shows that the flexural strength of all extrudates was not affected by the processing temperatures and the progressive fusion of primary particles of the PVC. Table 6.12: Effect of processing temperatures on the flexural strength of unfilled and filled extrudates Tdie. (oC) Flexural Strength (MPa) S0 S3 S7 S9 S21 S27 180 90+2 69+1 68+2 76+3 61+2 - 185 86+4 67+2 64+4 76+2 60+1 - 190 82+4 66+1 65+2 75+1 60+1 55+4 195 88+4 67+2 65+3 73+2 59+2 57+2 248 As discussed in Chapter 3 (Section 3.3.2.2), Tables 6.10, 6.11 and 6.12 show that the addition of impact modifier in PVC-U reduced the flexural modulus and strength of S3 and S7. However, the results show that the modulus and strength of S3 were higher than S7 at all processing temperatures. This is due to CPE has more softening effect than acrylic in reducing the rigidity of PVC. The similar reason can also be used to elucidate the higher modulus and strength of filled extrudate, S21 compared to S37. The incorporation of EFB fillers in the filled extrudates resulted in the modulus of S21 was higher than unfilled extrudate (S3) at all temperatures, except at 195 oC. On the other hand, the flexural strength of S21 was lower than S3 for all temperatures. The combination effects such as poor distribution of EFB fillers and increment of heat of fusion were plausible factors for the slightly lower of modulus (at 195 oC) and strength of S21 (all processing temperatures) compared to S3. These factors can also be used as reasons to elucidate the lower modulus and strength of S37 compared to S7. 6.3.4.2 Effect of Screw Speed Tables 6.13 and 6.14 show the effect of screw speeds on the flexural properties of unfilled and filled extrudates. The flexural modulus and strength of S0 increased as the screw speed was increased from 50 to 60 rpm. As the screw speed further increased to 70 rpm, the modulus and strength remained relatively constant. The results indicate that in general the flexural modulus and strength of extrudates were not significantly affected by the screw speed. Extrudate, S0 was the most affected by the screw speed with 5 % and 8 % increment in modulus and strength, respectively, as the screw speed increased from 50 to 60 rpm. 249 Table 6.13: Effect of screw speeds on the flexural modulus of unfilled and filled extrudates Screw Flexural Modulus (MPa) Speed S0 S3 50 3309+133 2794+296 2685+93 3344+115 3066+97 60 3486+23 2824+124 2792+126 3514+147 3051+63 70 3426+97 2652+77 2621+191 3242+198 2918+127 (rpm) S7 S9 S21 Table 6.14: Effect of screw speeds on the flexural strength of unfilled and filled extrudates Flexural Strength (MPa) Screw Speed S0 S3 S7 S9 S21 50 83+5 68+4 66+4 74+2 61+2 60 90+2 69+1 68+2 76+3 61+2 70 88+2 65+1 65+3 75+2 59+2 (rpm) 6.3.5 Tensile Properties Tensile test was done to determine the yield stress and strain at break of extrudates at difference screw speeds and processing temperatures. The crosshead speed used was fixed at 20 mm/min. Tables 6.15 and 6.16 show that the yield stress of extrudates increased as the temperature increased from 180 to 185 oC and remained relatively constant as the 250 temperature increased to 195 oC. Table 6.17 shows that the yield stress values of the extrudates were also not significantly affected by the increase of screw speeds. The tensile test has indicated that yield stress was not sensitive to the variation of screw speeds whereas slightly influenced by the processing temperature especially below 185 oC. Similar observations have also been reported by Benjamin (1980) and Truss (1985) for unmodified PVC specimens from the wall of pipes extruded at difference temperatures. The tensile stress increased gradually with increasing fusion level to reach a maximum value and thereafter remained constant. Because of the yield stress was only marginally changed, they concluded that the yield stress was relatively independent to the processing temperature. Therefore, the yield stress obtained in this study can also be said to be relatively independent to the fusion level. The research works done by Terselius and Jasson (1984), and Patel and Gilbert (1985) also showed the similar results. The insensitivity of the yield stress on the fusion level is a fact, which is of considerable relevance to the discussion of molecular deformation mechanism at yielding. Since the increase of fusion level was associated with the increase of the strength and coherence of the entanglement network (Terselius and Jansson, 1985), yielding evidently involved molecular arrangements between entanglement and network junctions. The fact that the yield stress was independent of the molecular weight was a further support (Terselius and Jansson, 1984). Table 6.15: Effect of processing temperatures on the yield stress of unfilled extrudates at screw speed of 60 rpm Tdie. (oC) Yield Stress (MPa) Heat of Fusion (J/g) Yield Stress (MPa) Heat of Fusion (J/g) So Yield Stress (MPa) Heat of Fusion (J/g) S7 S3 180 2.23 43.8 + 2.0 4.15 42.1 + 1.4 2.68 41.0 + 1.7 185 2.53 55.4 + 2.1 5.81 44.1 + 0.5 2.93 42.7 + 3.9 190 2.82 57.6 + 1.6 5.49 47.3 + 1.8 4.72 42.2 + 2.7 195 5.71 55.7 + 3.4 5.93 47.3 + 1.5 4.82 44.5 + 1.0 251 Table 6.16 : Effect of processing temperature on the yield stress of filled extrudates at screw speed of 60 rpm Tdie. (oC) Heat of Fusion (J/g) Yield Stress (MPa) Heat of Fusion (J/g) Yield Stress (MPa) Heat of Fusion (J/g) S21 S9 Yield Stress (MPa) S37 180 1.91 40.9 + 1.0 4.52 34.3 + 1.4 - - 185 4.49 49.8 + 1.9 8.02 39.8 + 0.7 - - 190 4.72 48.9 + 1.0 9.05 38.9 + 0.4 4.89 32.9+2.5 195 7.73 47.7 + 1.1 9.38 38.9 + 0.9 6.92 33.5+3.5 Table 6.17: Effect of screw speeds on the yield stress of extrudates at die temperature of 180 oC Speed (rpm) 50 S0 43.0 + 3.7 Yield Stress (MPa) S9 S3 S7 38.9 + 1.4 43.2 + 1.2 41.6 + 2.5 S21 31.7 + 1.4 60 43.8 + 2.3 40.9 + 1.3 42.1 + 1.4 41.5 + 0.4 34.3 + 1.6 70 43.8 + 2.7 40.4 + 2.9 44.3 + 1.8 43.4 + 2.0 32.9 + 2.6 Tables 6.15-6.17 also show that the yield stress of S3, S7, S9 and S21 were lower than S0. It shows that the impact modifier and EFB filler were able to reduce the yield stress of PVC for yielding process to occur. The ability of impact modifier to reduce the yield stress and facilitate the cold drawing, which resulted in the increase of energy absorption of PVC, has also been reported by other researchers (Petrich, 1973; Marshall, 1982; Stevenson, 1995). The impact modifier provided sites of stress concentration that occurred in the PVC matrix around an impact modifier particle upon application of the tensile stress to the material. Since these localized concentrations were well dispersed throughout PVC-U matrix, the shear yielding of impact-modified extrudates was initiated easily upon tensile (Figure 6.7). 252 The reason EFB filler reduced the yield stress of PVC can be explained by the increase in numbers of voids, which acted as stress concentration sites for yielding to be initiated more easily (Nakamura et al., 1998; Liang et al., 2000). Unlike impact modifier (rubbery particles), fillers did not deform as the external stress was applied to specimen because the modulus of fillers are usually greater than that of the matrix. When the applied stress exceeded the interfacial adhesion strength between EFB and PVC, debonding at the interfaces would occur first, leading to the formation of microvoids. In this case, the deformation restraint of the matrix around the filler was released, resulting in the shear yielding (shear yielding was favored than crazing due to weak interfacial bonding) and absorbed strain energy of PVC without blocking the propagation of cracks (Liang et al., 2000). Although the voids were able to provide stress concentrations to facilitate the shear yielding, they were not as effective as the impact modifier (Figure 6.7). Besides that, the voids also act as a defect, which can be seen in the broken specimens. Because of the presence of voids, the reduction of yield stresses of filled extrudates was slightly higher than unfilled extrudates. The yield stresses of filled extrudates (S21 and S37) were further reduced when the impact modifier was added into the filled extrudates. The strain at break, which indicates ductility of the materials, is shown in Tables 6.18-6.20. It can be seen that in general the strain at break of extrudates was not significantly affected by screw speed and processing temperature except for S3. As mentioned earlier in the impact strength section, the decrease of the strain at break of S3 was due to the reduction of free volume of PVC. Tables 6.18-6.20 also show that the trend of the strain at break values was similar to the trend showed by the impact strength values in Tables 6.6, 6.7 and 6.9. This trend shows that the highest value of strain at break occurs at the highest value of impact strength. In other words, the results obtained in Tables 6.18-6.20 were well-correlated to the impact strength results as shown in Tables 6.6, 6.7 and 6.9. Thus, the fusion level determined the level of ductility of extrudates in terms of strain break. Therefore, the combination of new network structures and some remaining primary particles in the extrudates contributed to the strain at break. 253 S0 S3 S7 S9 S21 50 45 40 Stress (MPa) 35 30 S3 (117%) 25 20 15 10 5 0 0 2 4 6 8 10 12 14 16 18 20 Strain (%) Figure 6.7 A typical stress-strain graph of tensile for extrudates at 60 rpm and 180 oC Tables 6.19 and 6.20 show that the strain at break of filled extrudates was lower than unfilled extrudates. The strain break reduction was attributed to the detrimental effect of EFB fillers and the presence of voids in S9, S21 and S37. For unfilled impact-modified extrudates, as expected, the addition of impact modifier in the S3 ad S7 has promoted the PVC-U matrix to experience cold drawing after yielding and thus, increasing the energy absorption (Figure 6.7). The ability of the impact-modified PVC matrix to enhance the energy absorption by cold drawing (plastic deformation) has also been discussed by other researchers (Bucknall, 1977, Newman and Stella, 1965). A more applicable discussion to the behaviour of impactmodified PVC has been suggested by Newman and Stella (1965). They suggested that the cold drawing of the glassy matrix resulted in the increase of free volume of the matrix around rubbery particles, due to the resistance of the particles to change in volume. The expansion of matrix surrounding the particle generated heat that increase 22 254 the temperature of the matrix, so that it approached its glass transition temperature and deformed more easily and absorbed more energy in viscous flow process. Table 6.18: Effect of processing temperature on the strain at break of unfilled extrudates at screw speed of 60 rpm Tdie. (oC) Heat of Fusion (J/g) Strain at Break (%) Heat of Fusion (J/g) So Strain at Break (%) Heat of Fusion (J/g) Strain at Break (%) S7 S3 180 2.23 10.7 +5.1 4.15 132.1 + 65.4 2.68 38.4 + 3.0 185 2.53 19.3 + 5.0 5.81 103.4 + 77.9 2.93 21.1 + 9.6 190 2.82 15.1 + 1.7 5.49 62.5 + 57.3 4.72 19.4 + 5.4 195 5.71 6.9+ 0.7 5.93 29.2+10.5 4.82 20.7+2.0 Table 6.19: Effect of processing temperature on the strain at break of filled extrudates at screw speed of 60 rpm Tdie. (oC) Heat of Fusion (J/g) Strain at Break (%) Strain at Break (%) Heat of Fusion (J/g) S21 S9 Strain at Break (%) Heat of Fusion (J/g) S37 180 1.91 6.2 + 0.8 4.52 4.9 + 0.6 - - 185 4.49 7.6 + 0.9 8.02 11.1 + 1.8 - - 190 4.72 6.0 + 0.6 9.05 9.7 + 2.1 4.89 8.0 + 1.2 195 7.73 6.9+ 0.7 9.38 10.2+2.4 6.92 8.2 + 1.8 255 Table 6.20: Effect of screw speed on the strain at break at die temperature of 180 oC Speed (rpm) 50 S0 14.7 + 7.1 Strain at Break (%) S3 S7 S9 99.7 + 65.4 23.5 + 5.7 6.9 + 1.3 S21 5.0 + 1.0 60 10.7 + 5.2 132.1 + 77.9 38.4 + 3.0 6.2 + 0.8 4.9 + 0.6 70 16.7 + 6.2 172.3 + 57.3 36.9 + 11.3 6.0 + 0.5 4.6 + 0.9 6.3.6 Yield Stress Analysis In this section, the yield stress analysis was done to predict the yield stress of the extrudates at an impact velocity of 3 m/s. This impact velocity was used for pendulum impact testing. Extrudates produced at 180 oC and 190 oC with constant screw speed of 60 rpm were chosen for this analysis. A linear regression was performed on the yield stress data at lower strain rate (Table 6.21) using the Eyring equation of plastic yielding to obtain the yield stress at 3 m/s. Table 6.21: Effect of strain rate on the yield stress of unfilled and filled extrudates Crosshead Strain Strain speed rate rate (mm/min) -1 (ln) (s ) Yield stress ( MPa) S0 180 o S9 190 o 180 o S3 190 o 180 o S7 190 o 180 o S21 190 o 180 o 190 o S37 190 ( C) ( C) ( C) ( C) ( C) ( C) ( C) ( C) ( C) ( C) (oC) 5 0.001 -6.571 38.8 53.7 37.6 46.4 40.8 44.1 37.5 40.3 31.7 36.5 26.3 20 0.006 -5.185 43.8 57.6 40.9 48.9 42.1 47.3 41.0 42.2 34.3 38.9 26.9 80 0.022 -3.808 45.1 60.9 43.6 51.8 46.0 50.1 43.3 45.6 35.2 41.3 28.1 100 0.028 -3.583 45.2 61.7 44.1 52.2 46.6 49.3 42.8 46.2 37.2 41.7 29.0 The effect of strain rate on the yield stress of the extrudates is shown in Table 6.21. It shows that the yield stress increased as the strain rate increased. The yield 256 stress of S0 was higher than other extrudates at all strain rate. The increase of yield stress as the strain rate increased was also found by Wee (2001). It indicates that the yielding process being suppressed with the increase of strain rate, and thus the fracture tended to change from ductile to brittle behaviour. It can be explained in molecular aspect, where during yielding the molecular chains are uncoiled and slip past each others and gradually align closer in the reaction of the tensile stress. As the strain rate increased, the rate of extension molecular chains to align and stiff become faster and then the stress on the molecular chains increased. Eventually, the covalent bonds of the main chains break and fracture of the material occurred (Smith, 1996). The Eyring equation has been previously used by other researchers (Calvert et al.,1991; Azman Hassan,1996; Wee, 2001) in their studies to predict the yield stress values of impact-modified PVC-U at high strain rate. The Eyring theory has described that the yield behaviour of polymers to be dependent on the temperature and rate of deformation. In the high stress region where yield occurs, the Eyring equation gives the yield stress (σy) in terms of the strain rate at yield (ε) as : σy T = k ⎡ ln ε ∆H ⎤ + γν ⎢⎣ A kT ⎥⎦ (6.1) where, Α is a constant, ν is the activation volume, ∆H is a activation energy for the low process, k is Boltzmann’s constant, γ is a stress concentration factor and T is an absolute temperature. The plots of σy for all extrudates at temperature of 180 oC and 190 oC against natural log tensile strain rate were increased linearly as shown in Figure 6.8. The equation obtained by linear regression and extrapolating was used to predict the σy at constant testing temperature. The value of slope was represented for value of k/γνA while the intercept was represented for value of ∆H/k. Thus, to obtain σy, the strain rate at impact velocity was firstly determined. 257 100 y = 2.6301x + 71.064 2 Yield Stress (MPa) R = 0.9984 190 oC 80 60 y = 2.2871x + 54.497 2 R = 0.8988 40 180 oC 20 0 -8.000 -6.000 -4.000 -2.000 0.000 2.000 4.000 Strain Rate (ln) Figure 6.8 Predicted yield stress values at 3 m/s for S0 In order to determine the apparent strain rate of the impact test in three point bending, both equations from the flexural strength and modulus were used and rearranged using Hooke’s Law and resulting in the following equation: ε= 6vd L2 (6.2) where v is the impact velocity, d is the mean thickness of the specimens of the sample and L is the span between the centers of support. By substituting the relevant data, the strain rate for impact velocity of 3 m/s was found to be 20.76 s-1. It was substituted into the regression equation, which was obtained from the graph. The predicted σy of extrudates at 3 m/s are shown in Table 6.22 and Table 6.23. From the predicted yield stress, the incorporation of EFB filler and impact modifier decreased the yield stress of the filled and impact-modified extrudates. The ability of EFB filler and impact modifier in reducing the yield stress has previously discussed. 258 Table 6.22: Predicted yield stress of extrudates at 3 m/s, die temperature of 180 oC and screw speed of 60 rpm Extrudate Equation Predicted Yield Stress (MPa) S0 σy =2.2871(ln ε) + 54.497 61.4 S9 σy =2.1534(ln ε) + 51.853 58.4 S3 σy =2.0092(ln ε) + 53.492 59.6 S7 σy =1.8543(ln ε)+ 50.026 55.6 S21 σy =1.5461(ln ε)+ 41.976 46.7 Table 6.23: Predicted yield stress of extrudates at 3 m/s, die temperature of 190 oC and screw speed of 60 rpm Exrudate Equation Predicted Yield Stress (MPa) S0 σy =2.6301(ln ε) + 71.064 79.0 S9 σy =1.9601(ln ε) + 59.208 65.1 S3 σy =1.8882(ln ε) + 56.738 62.5 S7 σy =2.0056(ln ε)+ 53.175 59.3 S21 σy =1.7393(ln ε)+ 47.926 53.2 S37 σy =0.8430(ln ε)+ 31.619 34.2 6.3.7 General Discussion Table 6.24 shows the inter-relationship between impact strength, predicted yield stress and strain at break of extrudates at die temperature 180 oC and 190 oC. 259 The results show that the yield stress of all extrudates increased as the temperature increased. It is interesting to note the ability of S0 and S9 extrudates to increase in strain at break and yield stress with the increase of temperature. This could be attributed to the increase of fusion level thus increasing the number of effective entanglement and crystalline junctions favouring orientation hardening. Table 6.24: Impact strength, yield stress and strain at break at die temperature 180 oC and 190 oC 180 oC Sample 190 oC S0 Impact Strength (kJ/m2) 4.7 Yield Stress (MPa) 61.4 Strain at Break (%) 11 Impact Strength (kJ/m2) 5.3 Yield Stress (MPa) 79.0 Strain at Break (%) 15 S3 82.2 59.6 132 63.1 62.5 63 S7 15.5 55.6 38 9.3 59.3 19 S9 4.4 58.4 6 4.6 65.2 6 S21 6.2 46.7 5 6.0 53.2 10 Table 6.24 also shows the impact strength and strain at break of impactmodified extrudates (S3 and S7) decreased while the yield stress increased as the temperature increased. The increase of yield stress of S3 and S7 was attributed to the increase of fusion level of PVC, while the decrease of impact strength and strain at breaks was attributed to the increase of miscibility of acrylic and CPE in the PVC. This could be a reason why the strain at break and the impact strength decreased at higher temperature. Wimolmala et al. (2001) also reported that at higher temperature (195 oC) the PVC became more miscible with the acrylic rubber and resulted in the decrease of tensile toughness. Similar to the excessive increases of impact modifier concentration, the excessive increases of temperature can produce lower impact strength of PVC. 260 In terms of relationship between fusion level and mechanical properties, as reported previously, the highest impact strength and strain at break of S0, S3, S9 and S21 occurred at temperature of 185 oC with heat of fusion of 2.53 J/g, 4.15 J/g, 4.49 J/g, and 8.02 J/g, respectively. For S7 and S37, the highest impact strength and strain at break occurred at temperature 180 oCand 195 oC. The heat of fusion of S7 and S37 was 2.68 J/g and 6.92 J/g, respectively. The occurrence of the highest impact strength and strain at break at the moderate fusion level as found in this study was in agreement with the other researchers (Benjamin, 1980; Uitenham and Geil, 1981; Terselius and Jansson, 1981, 1984, 1985). The flexural properties and yield stress were found insensitive to the increase of fusion level, which contributed by the increase of temperature and screw speed. 6.4 Conclusions The mechanical properties and fusion analysis of extrudates were studied. Based on the results obtained several conclusions can be made: (a) The filled extrudates incorporated with 10 phr of filler can be extruded with a smooth surface. The rough surface of extrudates appeared after leaving the die for the extrudates incorporated with 20 phr filler, due to fillers poor distribution and dispersion in the matrix. (b) The filled compound (S37) containing 10 phr of EFB and 9 phr of CPE can only be extruded at die temperatures of 190 oC and 195 oC. This extrudate experienced melt fracture (melt bambooing) when extruded at below these temperatures. The increase of melt viscosity due to the delaying of PVC fusion resulted in the melt fracture of extrudates. 261 (c) The heat of fusion of extrudates increased with increasing processing temperature due to the increase of thermal energy. Similar to impact modifier, EFB filler has ability to increase the fusion level of filled extrudates, due to the increase of frictional heat, particularly when combined with acrylic impact modifier. (d) The impact strength of all extrudates was relatively constant as the screw speed increased from 50 to 70 rpm, except for S7 and S9. The influence of screw speed on the flexural properties and yield stress of extrudates is not significant. (e) The impact strength of S9, S21 and S37 relatively constant as the temperature increased from 180 to 195 oC while the impact strength of S7 relatively constant from 185 to 195 oC. The impact strength of S3 drastically decreased as the temperature increased to 195 oC. The increase of temperature from 180 to 185 oC slightly increased the yield stress of extrudates. However, yield stress remained relatively constant as the temperature increased from 185 to 195 oC. In most cases, the flexural properties and yield stress are independent of processing temperatures. (f) The yield stress of extrudates is directly proportional to the strain rate of tensile test. Similar to the experimental yield stress, the predicted yield stress also increased with processing temperatures. Although the incorporation of EFB reduced the yield stress of the PVC, the plastic deformation or strain hardening of the extrudates filled with EFB during tensile testing was limited (the inability of extrudates to elongate at higher value) . As a result, the impact modified and unmodified filled extrudates (S9 and S21) has lower values of strain at break and impact strength than the respective unfilled samples. CHAPTER 7 OVERALL CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK 7.1 Conclusions In this study, two processing methods were used to produce the PVC-U composites samples; compression moulding and extrusion. EFB-filled PVC-U composites with up to 40 phr filler content were produced using the compression moulding i.e. using a two-roll mill and hot press (Chapter 3). For the extrusion, only extrudates with 10 phr filler could be successfully produced (Chapter 6). The smooth surface of extrudates changed to be rough as the filler content increased to 20 phr. It shows that in single screw extrusion of EFB filled PVC-U composites, high L/D ratio is required to ensure the dry blends have a sufficient residence time for good melt homogeneity before leaving the die. The use of EFB fibre as a reinforcement filler in the compression moulded composites contributed to the enhancement of stiffness, but decreased the ductility, flexural strength of the PVC-U composites (Chapter 3). The results have shown that the addition of acrylic and CPE impact modifier improved the toughness but reduced marginally the flexural properties of EFB-filled composites. The study revealed that acrylic is a better impact modifier in improving toughness and cause a lesser decrease 263 in flexural properties of filled impact-modified composites compared to CPE. Besides that, the impact strength and flexural properties of filled composites added with acrylic impact modifier were also found to be more stable in the accelerated weathering than CPE. It was therefore decided to use only acrylic impact modifier for the studies of filler-surface treatment of EFB-filled acrylic-impact modified composites. Filler-surface treatment has fairly improved the toughness, stiffness, and strength of these composites when treated with coupling agents, and Prosil 9234 proved to be more effective compared to NZ 44 coupling agent. FTIR spectra and SEM micrograph showed that both coupling agents reacted with PVC to form chemical interactions and simultaneously reduced water absorption of the coupling agent-treated filled composites. The effect of EFB filler on the thermal properties is another important area which was investigated (Chapter 4). It was found that EFB filler marginally increased the Tg and insignificantly changed the HDT. EFB filler, however, decreased significantly the thermal stability of PVC. The addition of acrylic impact modifier has also contributed to the further declining of the thermal stability of PVC. Although the EFB-filled acrylic-impact modified composites have low thermal stability, these composites have good processability characteristics compared to other composites (Chapter 5). The study in Chapter 6 has shown that it is possible to extrude 10 phr EFB filled PVC-U composites at moderate temperatures (180-185 0C) and screw speeds (50-60 rpm) for the unmodified and acrylic impact modified formulations. In the extrusion, temperatures and screw speed governed the fusion level of PVC as indicated by DSC analysis. The correlation between mechanical properties and fusion level showed that the moderate fusion level contributed to the relatively better mechanical properties. Although the temperature and screw speed are important parameters governing extrusion process, the role of additives in the PVC-U compound is also important. Therefore, it is anticipated that filled-impact modified extrudates with 20 phr and higher can be extruded provided correlations between temperature, screw speed, and other additives are thoroughly studied. Overall, it can be concluded 264 that EFB, which is agricultural by-product from oil palm trees, has the potential to be used as filler in PVC composites as it enhanced the stiffness and reduced the cost of the composites. 7.2 Suggestions for Future Work From the present study, the following areas of research are needed for future investigations: (1) Microstructure of the interface between EFB fibre and matrix still needs to be investigated and the interfacial properties should be studied with more rigorous single fibre pullout tests. The reason for this suggestion is to study in depth the relationship between interface and untreated or between interface and treated fibre surface composite properties. This suggestion can also be used to evaluate the significance of using coupling agents in EFB fibre-filled PVC-U composites. (2) In the study of the fibre-surface treatment of composites, the results have shown that the improvement in mechanical properties is not encouraging. It would be interesting to find out a more suitable coupling agent for the studied composites. (3) In the FTIR analysis, it is very difficult to distinguish clearly the nature of new peaks observed for the treated composites. These peaks could be also the results of the interaction between other additives and fillers or PVC and fillers, since the formulation of PVC-U used in this study contained many additives. Further research is needed to identify the role of each additive on the adhesion between PVC and coupling agent-treated EFB filler. 265 (4) In extrusion, the amount of processing aid content in the blend formulations was kept constant at 1.5 phr. Thus, in order to overcome the difficulty to extrude the extrudates filled containing 20 phr and higher, a possible method can be investigated to increase the processing aid content in the blend formulations. This should be investigated. Besides that, the L/D of screw and type of extruder also play an important role in determining the efficiency of mixing of the blends, thus influencing the quality of filled extrudates. Therefore, the other possible methods need for further investigations, which can be used to overcome the problem of extrudate filled with 20 phr and higher, are to use a higher L/D ratio or twin screw extruder in order to increase the efficiency of mixing and avoid the thermal degradation. (5) New application should be found for EFB-fibre-based composites. Hybrid fibre composites with EFB fibre and other fibres rather than glass may open up new applications. For example, from the economics point of view, EFB fibres may be hybridised with carbon or aramid fibres to reduce the costs of these expensive fibres reinforced composites whilst maintaining their good mechanical performance. REFERENCES Abdul Khalil, H.P.S., Ismail, H., Ahmad, M.N.,Ariffin, A. and Hassan, K. (2001).The Effect of Various Anhydride Modifications on Mechanical Properties and Water Absorption of Oil Palm Empty Fruit Bunches Reinforced Polyester Composites.Polymer International. 50:395-402. Ahmad Fuad, M.Y., Yaakob, I., Mohd Ishak, Z.A. and Mohd Omar, K.(1993). Density Measurement of Rice Husk Ash Filler Particles in Polypropylene Composites. Polymer Testing.12:107-112. Ahmad Fuad, M.Y., Salleh, O., Mohd Ishak, Z.A. and Mohd Omar, A.K.(1994). Application of Rice Husk Ash as Fillers in Polypropylene: Effect of Silane Coupling Agents. ASEAN Journal of Science Technology Development.11 (1): 115-140. Ahmad Fuad, M.Y., Zawawi, M.T., Rahmad, S., Norazlan, R. and Ismail, Z.(1997). Effect of Silane, Titanate and Zirconate Coupling Agents in Polypropylene Composites. Journal of Industrial Technology. 6(2): 1-13. Allsopp, M.W.(1982). Morphology of PVC. In : Burgress, R.H. Manufacturing and Processing of PVC. New York: Applied Science Publishers.215. Axtell, F.H., Sophanowong, W. and Rujiranontapong, K.(1994). The Influence of Soft Filler Characteristics and Fusion Level on the Impact Strength of Poly(vinyl chloride). Plastics, Rubber and Composites Processing and Applications. 21: 53-61. Ash, B.J., Schadler, L.S. and Siegel, R.W.(2002). Glass Transition Behaviour of Alumina/Poly(methyl methacrylate) Nanocomposites. Materials Letters. 55: 83-87. 267 Azman Hassan (1996). Impact Modification and Properties of Solid and Foamed PVC-U. Loughborough University of Technology, United Kingdom: Ph.D. Thesis. Azman Hassan and Kannan, S.K.(2001). Effect of Titanate, Zirconate and Silane Coupling Agents in Rice Husk Ash-Filled PVC-U Composites. Journal of the Institute of Materials Malaysia. 2(2): 32-42. Azman Hassan and Sivaneswaran, K. (2002). The Effect of Rice Husk Ash on Processability of ABS Impact Modified PVC-U. 3rd National Symposium on Polymeric Materials. December 30-31. Universiti Teknologi Malaysia, 174179. Bairdo, M., Frisoni, G., Scandola, M. and Licciardello, A.(2002). Surface Chemical Modification of Natural Cellulose Fibers. Journal of Applied Polymer Science. 83:38-45. Beall, F.C. and Eickner, H.W. (1970). Thermal Degradation of Wood Components: A Review of the Literature. USDA Forest Service Research Paper. 1-27. Benjamin, P. (1980). Quality and Quality Control of uPVC Pressure Pipe. Plastics and Rubber: Materials & Applications. 5(4):151-160. Bohme, K. (1975). Processing Aids. Angew. Makromol. Chem. 47: 243. Bower, J.D. (1986).The Function of Lubricants in Processing of Rigid PVC. Journal of Vinyl Technology. 8(4):179-182. Bucknall, C.B. and Smith, R.R.(1965). Stress-Whitening in High-Impact Polystyrene. Polymer. 6:437-446. Butters, G.(1982). Particulate Nature of PVC –Formation, Structure, and Processing. New York : Marcel Dekker. Bystedt, J., Flensjo, M., Svensson, G. and Terselius, B. (1988). The Effect of Shear on Fusion and Mechanical Properties of Rigid PVC Pipes. Journal of Vinyl Technology.10(2):100-102. Calvert, D.J., Haworth, B. and Stephenson, R.C. (1991). The Use of Fracture Mechanics to describe the Impact Strength of PVC Window Profiles. Plastics and Rubber : Materials & Applications. 15(4): 229-242 Casey, W.J. and Okano, K. (1986). An Investigation of Acrylic Processing Aids for Low Molecular Weight PVC. Journal of Vinyl Technology. 8(1): 37-40. 268 Chen, C.H, Mao, C.F and Lo, Y.W (2001). Study of Fusion Percolation Thresholds of Rigid PVC Compounds. Journal of Applied Polymer Science.81: 30223029. Chen, C.H, Wesson R.D, Collier, J.R. and Lo,Y.W. (1995). Studies of Rigid PVC Compounds IV: Fusion Characteristics and Morphology Analyses. Journal of Applied Polymer Science. 58: 1107-1115. Chen, C.H. and Lo. Y.W. (1999). Influences of Chlorinated Polyethylene and Oxidized Polyethylene on the Fusion of Rigid Poly(vinyl chloride) Compounds. Journal of Applied Polymer Science.74: 699-705. Chen, N., Wan, C., Zhang, Y. and Zhang, Y.(2004).Effect of Nano-CaCO3 on Mechanical Properties of PVC and PVC/Blendex Blend. Polymer Testing. 23: 169-174. Chiu, H.T. and Chiu, W.M. (1996). Influence of Mechanical Properties in Carbon Black (CB) Filled Isotactic Polypropylene (iPP) and Propylene-Ethylene Block Copolymer.Journal of Applied Polymer Science. 61: 607-612. Chiu, H.T. and Chiu, H.T.(1998). The Toughening Behaviour in Propylene-Ethylene Block Copolymer Filled with Carbon Black and Styrene-Ethylene ButylenesStyrene Triblock Copolymer. Materials Chemistry and Physics. 56: 108-115. Coaker, A.W. and Wypart, R.W.(1993). Vinyl Chloride Resins: Suspension, Emulsion,Microsuspension and Bulk(Mass). In: Wickson, E.J. Handbook of Polyvinyl Chloride Formulating. New York: John Wiley and Sons Inc: 15-55. Cogswell, F.N., Player, J.M. and Young, R.C. (1984). Influence of Processing Aids on the Extensibility of PVC Melts. Plastics and Rubber: Processing. 128-132. Collyer, A.A. and Clegg, D.W. (1986). Mechanical Properties of Reinforced Thermoplastics. London : Elsevier Applied Science Publisher. Conley, R. (1972). Infrared Spectroscopy.2nd Edition. Boston ,USA: Allyn and Bacon Publisher. Cora, B., Daumas, B. and Zegres, A. (1999). Impact Strength and Morphology in Poly (vinyl chloride) Window Profiles-Relationship with Gelation Level. Plastics,Rubber and Composites. 28(4): 165-169. Covas, J.A. (1992). Single Extrusion of PVC: Effect on Fusion & Properties. Journal of Vinyl Technology. 32(11): 743-750. 269 Covas, J.A and Gilbert, M. (1992). Single Screw Extrusion of Poly (vinyl Chloride): Effect on Fusion and Properties. Polymer Engineering and Science. 32(11): 743-750. Covas, J.A., Gilbert, M. and Marshall, D.E. (1988).Twin Screw Extrusion of a Rigid PVC Compound-Effect on Fusion and Properties. Plastics and Rubber Processing and Applications. 9: 107-116. Deanin, R.D. and Chuang, W.Z.L. (1986). ABS and CPE Modifiers in Rigid PVC. ANTEC ’86. 1239-1240. Dunkelberger, D.L., Lutz, J. and Szamborski, E.C. (1993). Processing Aids. In : Wickson, E.J. Handbook of Polyvinyl Chloride Formulating. New York : John Wiley and Sons Inc: 639-647. Fabbri, P., Champon, G., Castellano, M., Belgacem, M.N. and Gandini, A.(2004). Rapid Report Reactions of Cellulose and Wood Superficial Hydroxy Groups with Organomettalic Compounds. Polymer International. 53: 7-11. Faulkner, P.G. (1975). The Use of a Temperature Programmable Brabender Mixing Head for the Evaluation of the Processing Characteristics of PVC. Journal Macromolecules Science-Physics. B11(2):251-279. Fried, J.R. (2003). Polymer Science and Technology, 2nd Edition. USA: Prentice Hall Inc. Gardiner, D. Pepperl, G. and Wisse J.D.M. (1997). Natural and Artificial Wethering of White PVC Window Profiles. Plastics, Rubber and Composites Processing and Applications. 26 (2): 59-65. Geil, P.H. (1977). Morphology-Characterization Terminology. Journal Macromolecules Science-Physics.B14 (1):171. Gerlach, D. ( 1998). CPE as Impact Modifier. Kunststoffe. 88: 1215-1216. Gilbert, M. (1985). Fusion of PVC Compounds. Plastics and Rubber International. 10(3):16-19. Gilbert, M. and Vyoda, J.C. (1981).Thermal Analysis Technique for Investigating Gelation of Rigid PVC Compounds. Polymer. 22: 1134-1136. Gilbert. M., Hemsley, D.A., and Miadonye, A. (1983). Assessment of Fusion in PVC Compounds. Plastics and Rubber Processing and Applications. 3: 343-351. Graessley, W.W.(1974). Advanced Polymer Science. In: Terselius and Jansson. (1984).Gelation of PVC-Part 3: Effect on Tensile Properties. 4(4): 291-299. 270 Grohman, M. (1994). A Novel Impact Modifier System for Vinyl Siding and Profile Substrate. Journal of Vinyl Technology. 16(2): 107-111. Gurram, S., Julson, J.L., Muthukuarrapan, K. Stokke, D.D. and Mahapatra, A.K. (2002). Application of Biorenewable Fibers in Composites. ASAE/CSAE North Central Intersectional Conference. Sept. 27-28. University of Saskatchewan, Saskatoon, Canada, 1-20. Han, J.S. and Rowell, J.S. (1997). Chemical Composition of Fibers. In Rowell, R.M. Young, R.A. and Rowel, J.K. Paper and Composites from Agro-Based Resources.New York : Lewis Publishers. Hattori, T., Tanaka, K. and Matsuo, M.(1972). Fusion of Particulate Structure in PVC During Powder Extrusion. Polymer Engineering Science.12: 199. Hawkins, T.(1982). Evaluation of PVC Compound on the Brabender Torque Rheometer. Journal of Vinyl Technology. 4 (3):110-114. Hiltner, A., Siegmann, A. and Baer, E. (1984). Impact Modification of PVC with Chlorinated Polyethylene. ANTEC ’84. 555-557. Hill, C.A.S. and Abdul Khalil, H.P.S.(2000). Effect Fibre Treatments on Mechanical Properties of Coir or Oil Palm Fibre Reinforced Polyester Composites. Journal of Applied Polymer Science.78 (9): 1685-1697. Ichazo, M.N., Albano, C. and Gonzalez, J.(2000).Behaviour of Polyolefin Blends with Acetylated Sisal Fibres. Polymer International. 49: 1409-1416. Ismail, H. and Mega, L. (2001). The Effects of Compatibilizer and Silane Coupling Agent on the Mechanical Properties of White Rice Husk Ash Filled Polypropylene/Natural Rubber Blend. Polymer-Plastics Technology Engineering. 40(4): 463-478. Katz, H.S and Brandmaier, H. (1982). Concise Fundamentals Fibre-Reinforced Composites. In Milewski, J.V and Katz, H.S. (Editors). Handbook of Reinforcements For Plastics. New Jersey:Van Nostrand Reinhold Company Inc. Kerner, E.H. (1969). In : Maiti and Subbarao (1991). Mechanical Properties of Isotactic Polypropylene/Wood Flour Composites. International Journal of Polymeric Materials.15: 1-16. 271 Klaric, I., Vrandecic, N.S. and Roje, U. (2000). Effect of Poly(vinyl chloride)/Chlorinated Polyethylene Blend Composition on Thermal Stability. Journal of Applied Polymer Science. 79(1):166-172. Kokta, B.V.,Maldas, D., Daneault, C. and Beland, P.(1990). Composites of Poly(vinyl chloride)-Wood Fibers I. Effect of Isocynate as a Bonding Agent. PolymerPlastic Technology and Engineering. 29(1 and 2): 87-118. Kosior, E.I. and Stachurski, Z.H.(1986). Processing Aids: The Effect of Poly(methyl methacrylate) on the Fusion of Rigid Poly(vinyl chloride). British Polymer Journal. 18(2): 94-104. Kovacic, T., Klaric, I., Nardelli, A. and Baric, B. (1993). Thermal Degradation of Poly(vinyl chloride)/ Poly(a-methylstyrene-acrylonitrile) Blends. Polymer Degradation and Stability. 40: 91-96. Krzewki, R.J. and Collins E.A. (1981). Rheology of PVC Compounds.I. Effects of Processing Variables on Fusion. J. Macromol. Sci.-Phys. B20 (4): 443-464. Kunori, T. and Geil, P.H.(1980). Morphology-Property Realtionships in Polycarbonate-Based Blends. 1. Modulus. J. Macromol. Sci. Phys. B18(1): 93134. Lacatus, E.E, and Rogers, C.E. (1986). The Effect of Fusion and Physical Aging on the Toughness of Poly (vinyl Chloride).ANTEC 1986. 1271-1275. Leong. Y.W., Abu Bakar, M.B., Mohd Ishak, Z.A. and Ariffin,A.(2003). Mechanical and Morphological Study of Talc/Calcium Carbonate Filled Polypropylene Hybrid Composites Weathered in Tropical Climates. Jurnal Teknologi. 39(A): 23-34. Liang, J.Z. and Li, R.K.Y. (2000). Rubber Toughening in Polypropylene : A review. Journal of Applied Polymer Science.77(2):409-417. Lu, J., Wei, G.X., Sue, H.J. and Chu, J.(2000). Toughening Mechanism in Commercial Thermoplastic Polyolefin Blends. Journal of Applied Polymer Science.76: 311-319. Lutz, J.T. (1993). Impact Modifiers for PVC. Journal of Vinyl Technology. 15(2): 8298. Mahmood Husein Datoo (1992). Mechanics of Fibrous Composite. London : Elsevier Applied Science. 272 Maiti, S.N. and Subbarao, R.(1991). Mechanical Properties of Isotactic Polypropylene/Wood Flour Composites. Journal of Polymeric Materials.15: 1-16. Maiti, S.N. and Lopez, B.H.(1992). Tensile Properties of Polypropylene/Koalin Composites. Journal of Applied Polymer Science. 44: 353-360. Maldas, D. and Kokta, B.V. (1990). Effect of Extreme Conditions on the Mechanical Properties of Wood Fiber-Polystyrene Composites.II.Sawdust as a Reinforcing Filler. Polymer-Plastics Technology and Engineering. 29(1 and 2): 119-165. Maldas, D. and Kokta, B.V. (1990). Influence of Phthalic Anhydride as a Coupling Agent on the Mechanical Behavior of Wood Fiber-Polystyrene Composites. Journal of Applied Polymer Science. 41: 185-194. Maldas, D. and Kokta, B.V. (1991). Influence of Maleic Anhydride as a Coupling Agent on the Performance of Wood Fiber-Polystyrene Composites. Polymer Engineering and Science. 31(18): 1351-1357. Marcilla, A. and Beltran, M. (1995). Thermogravitmetric Kinetic Study of Poly(vinyl chloride) Pyrolisis. Polymer Degradation and Stability. 48: 219-229. Marinda, R., Yang, J., Roy, C. and Vasile, C. (1999). Vacuum Pyrolisis of PVC I. Kinetic Study. Polymer Degradation and Stability. 66(1): 127-144. Marshall, G.P. (1982). Design for Toughness in Polymers 1-Fracture Mechanics. Plastics and Rubber Processing and Applications. 2: 169-182. Marshall, D.E., Higgs, R.P. and Obande, O.P. (1983). The Effect of Extrusion Conditions on the Fusion, Structure and Properties of Rigid PVC. Plastics and Rubber Processing and Applications. 3: 353-358. Marshall, B.I.(1969). Lubrication of Rigid PVC. British Plastics. August: 70-78. Mathew, W.R. (1993). Colorants. In Wickson,E.J. Handbook of Polyvinyl Chloride Formulating. New York: John Wiley and Sons Inc: 449-473. Mathur, K.K and Vanderheiden, D.B. (1982). Precipitated Calcium Carbonate as Ultraviolet Stabilizers and Impact Modifiers in Poly(vinyl chloride) Siding and Profiles. Proceeding of International Symposium on Polymer Additives. 371-338. Mathur, K.K. and Vanderheiden, D.B.(1986). Fillers: Types, Properties and Performance. In: Kresta, J.E. Polymer Science and Technology. New York: Pelnum Press. 503-544 273 Matuana, L.M., Park, C.B. and Balatinecz, J.J.(1997). Processing and Cell Morphology Relationships for Microcellular Foamed PVC/Wood-Fiber Composites. Polymer Engineering and Science. 37(7): 1137-1147. Matuana, L.M., Balatinecz, J.J. and Park, C.B.(1998). Effect of Surface Properties on the Adhesion between PVC and Wood Veneer Laminates. Polymer Engineering and Science. 38(5): 765-773. Matuana, L.M. Kamdem, D.P. and Jun Zhang (2001). Photoaging and Stabilization of Rigid PVC/Wood Fibers. Journal of Applied Polymer Science. 80(11): 19431950. Manikandan Nair, K.C., Diwan, S.M. and Sabu Thomas. (1996). Tensile Properties of Short Sisal Fiber Reinforced Polystyrene Composites. Journal of Applied Polymer Science. 60: 1483-1497. Marinda, R., Yang, J., Roy, C. and Vasile, C. (1999). Vacuum Pyrolisis of PVC 1: Kinetic Study. Polymer Degradation and Stability. 66(1): 127-144. McCrum, Buckley, C.P. and Bucknall, C.B. (1997). Reinforced Polymers in Principles of Polymer Engineering.2nd . London: Edition Oxford Science Publisher. McGill, W.J. and Wittstock, T. (1983). Microstructure and Impact Strength of PVC Blends with EVAgPVC. Plastics and Rubber Processing and Applications. 3(1): 77-83. McGill, W.J., and Wittstock, T. (1987). The Effect of Processing Temperature on the Impact Strength of PVC Blended with a Particulate Acrylic Modifier.Plastics and Rubber: Processing and Applications. 7: 185-188. McNeill, I.C., Memetea, L. and Cole, W.J. (1995).A Study of the Products of PVC Thermal Degradation.Polymer Degradation and Stability. 49: 181-191. Mengeloglu, F.(2001). Rigid Poly(vinyl chloride)/Wood-Flour Composites and Their Foams. Michigan Technological University: Ph.D. Thesis. Menges, G. and Berndsten, N.(1976). Processing Aids. In: Wickson, E.J. Handbook of Polyvinyl Chloride Formulating. New York: John Wiley and Sons Inc: 639647. Mertz, E.H., Cleaver, G.C. and Bear, M.(1956). Studies on Heterogenous Polymeric Systems. Journal of Polymer Science. 22: 325-341. Miller, E. (1996). Introduction to Plastics and Composites: Mechanical Properties and Engineering Applications. New York:Marcel Dekker,Inc. 274 Mohd Ishak, Z.A., Aminullah, A., Ismail, H. and Rozman, H.D.(1998). Effect of Silane Based Coupling Agents and Acrylic Acid Based Compatibilizer on Mechanical Properties of Oil Palm Empty Fruit Bunch Filled High Density Polyethylene Composites. Journal Applied Polymer Science. 68: 2189-2203. Mohd Ishak, Z.A., Ishiaku, U.S. and Kocsis, J. K. (2000). Hygrothermal Aging and Fracture Behaviour of Short-Glass-Fibre-Reinforced Rubber–Toughened Poly(butylenes terephthalate) Composites. Composite Science and Technology. 60: 803-815. Montaudo, G. and Puglisi, C. (1991).Evolution of Aromatics in the Thermal Degradation of Poly(vinyl chloride) : A Mechanistic Study. Polymer Degradation and Stability. 33: 229-262. Morten-Jones, D.H.(1993). Polymer Processing. London : Chapmman & Hall. Murphy, J. (2001). Additives for Plastics Handbook. 2nd Edition. UK:Elsevier Science Ltd. Myers, G.E., Chahyadi, I.S., Coberly, C.A. and Ermer, D.S.(1991). Wood Flour/ Polypropylene Composites: Influence of Maleated Polypropylene and Process and Composition Variables on Mechanical Properties. International Journal of Polymeric Materials. 15: 21-44. Nabi Saheb. D. and Jog. J.P. (1999). Natural Fibre Polymer Composites: A review. Advances in Polymer Technology. 18(4): 351-363 Nakamura,Y., Fukuoka,Y., and Iida, T. (1998) . Tensile Test of Poly(vinyl chloride) Filled with Ground Calcium Carbonate Particles.Journal of Applied Polymer Science. 70: 311-316. Naqvi, M.K. and Sen, A.R. (1991). Thermal Stabilty and Thermal Characteristics of PVC/cis-PBR Blends. Polymer Degradation and Stability. 33: 367-375. Newman, S. and Strella, S.(1965).Stress-strain Behaviour of Rubber-Reinforced Glassy Polymers. Journal of Applied Polymer Science.9: 2297-2309. Nguyen, T., Zavarin, E.and Barrall, E.M. (1981). Thermal Analysis of Lignocellulosic Materials. Part 1. Journal Macromolecular Science- Rev. Macromolecular Chemistry. C20(1): 1-65. Nielsen, L.E. (1966). Simple Theory of Stress-Strain Properties of Filled Polymers. Journal of Applied Polymer Science. 10(1): 97-103. Nielsen, L.E. (1974). Mechanical Properties of Polymers and Composites, Vol 2. New York: Marcel-Dekker. 275 Nicolais, L. and Narkis, M.(1971). Polymer Engineering Science. 11:194. In : Gupta and Punawar (1984). Tensile Yield Behaviour of PP and SEBS Blends. Journal of Applied Polymer Science. 29: 3513-3531. Okieimen, F.E. and Eromonsele, O.C. (2000). Thermal Stabilization of PVC with Khaya Seed Oil: Thermogravimetic Studies.Journal of Applied Polymer Science. 77(7): 1432-1438. Owen, E.D.(1989). Degradation and Stabilisation of PVC. New York:Elsevier Applied Science Publisher. Parey,J. and Menges, G. (1981).PVC- Gelation Model. Journal of Vinyl Technology. 3: 152-156. Pavia, D.L., Lampman, G.M. and Kriz, G.S.(1979). Introduction to Spectroscopy: A Guide for Students of Organic Chemistry. Philadelphia: Saunders College Publishing. Pavlinec, J., Lazar, M. and Csomorova, K. (1997). Thermal Degradation of Multilayer Methacrylate-Acrylate Perticle-Bead Polymer Powders and Melts. Polymer Degradation and Stability. 55: 65-71. Pedersen, T.C. (1984). Process and Material Considerations in the Industrial Application of Lubricants in Rigid PVC Extrusion. Journal of Vinyl Technology. 6 (3):104-109. Pedro, B.M.A, Monteiro, E.E.C. and Dweck, J. (2001). PVC and Agalmatolite Composite Characterization. Polymer Testing. 20: 269-273. Penn, W.S. (1971). PVC Technology. London : Applied Science Publishers Limited. Peng, L.C.(2002). Influence of Process Variables on Totally Chlorine Free Dissolving Pulp From Oil Palm Lignocellulosic as Evaluated by Response Surface Methodology. Universiti Sains Malaysia. Ph.D. Thesis. Peters, B. and Bruch, C. (2001). A Flexible and Stable Numerical Method for Simulating the Thermal Decomposition of Wood Particles. Chemosphere. 42: 481-490. Petrich, R.P.(1973). Impact Reinforcement of Poly(vinyl chloride). Polymer Engineering and Science.13(4): 248-254. Pfening, J.L.G. and Dunkelberger, D.L. ( 1986). Effect of Acrylic Processing Aids on PVC Die Swell. J. of Vinyl Technology. 8 (3): 126-131. 276 Portingell, G.C. (1982). Processing Properties. In Butters, G. Particulate Nature of PVC –Formation, Structure and Processing. New York : Marcel Dekker. 135227 Povolo, F., Schwartz, G. and Hermida, E.B.(1996). Temperature and Strain Rate Dependence of the Tensile Yield Stress of PVC. Journal of Applied Polymer Science. 61:109-117. Porter, R.S. and Johnson, J.(1966). In : Terselius, B. and Jansson, J.F. (1984).Gelation of PVC, Part 3: Effect on Tensile Properties. Plastics and Rubber Processing and Applications. 4(4): 291-299. Prasad, S.V., Pavithran, C. and Rohatgi, P.K.(1983).Alkali Treatment of Coir Fibres for Coir-Polyester Composites. Journal of Materials Science.18: 1443-1454. Raghi, S.E, Zahran, R.R. and Gebril, B.E.(2000). Effect of Weathering on some Properties of Poly(vinyl chloride)/Lignin Blends. Materials Letters.46: 332342. Raj, R.G., Kokta, B.V. Groleau, G. and Daneault, C. (1989). Use of Wood Floor Fiber as a Filler in Polyethylene: Studies on Mechanical Properties.Plastics and Rubber Processing and Applications. 11: 215-221. Raj, R.G., Kokta, B.V. Groleau, G. and Daneault, C. (1990). The Influence of Coupling Agents on Mechanical Properties of Composites Containing Cellulosic Fillers. Polymer-Plastics Technology and Engineering. 29(4):339353. Raj. R.G., Kokta B.V., Maldas, D. and Daneault, C.(1989).Use of Wood Fibers in Thermoplastics.VII. The Effects of Coupling Agents in Polyethylene-Wood Fiber Composites. Journal of Applied Polymer Science. 37: 1089-1103. Robinovic, I.S. (1983). The Acrylic Weatherable Impact Modifier for PVC, Journal of Vinyl Technology. 5(4): 179-182. Robinovitch, E.B., Lacatus, E. and Summers, J.W. (1984). The Lubrication Mechanism of Calcium Stearate/Paraffin Wax Systems in PVC Compounds. Journal of Vinyl Technology. 6(3): 98-103. Rowell, R.M. (1997). Chemical Modification of Agro –Resources for Property Enhancement. In Rowell, R.M., Young, R.A. and Rowell, J.K. Paper and Composites from Agro-Based Resources. London : Lewis Publishers. 351-375. Rozman, H.D., Lai C.Y., Ismail, H. and Mohd Ishak, Z.A. (2000).The Effect of Coupling Agents on the Mechanical and Physical Properties of Oil Palm 277 Empty Fruit Bunch-Polypropylene Composites. Polymer International. 49: 1273-1278. Rozman, H.D., Peng, G.B. and Mohd Ishak, Z.A. (1998). The Effect of Compounding Techniques on the Mechanical Properties of Oil Palm Fruit BunchPolypropylene Composites. Journal of Applied Polymer Science. 70: 26472655. Rozman, H.D., Tan, K.W., Kumar, R.N. and Abu Bakar A. (2001). The Effect of Hexamethhylene Diisocyanate Modified ALCELL Lignin as a Coupling Agent on the Flexural Properties of Oil Palm Empty Fruit BunchPolypropylene Composites.Polymer International. 50: 561-567. Rozman, H.D., Tay, G.S., Kumar, R.N., Abusamah, A., Ismail, H and Mohd Ishak, Z.A.(2001). The Effect of Oil Extraction of the Oil Palm Empty Fruit Bunch on the Mechanical Properties of Polypropylene-Oil Palm Empty Fruit BunchGlass Fibre Hybrid Composites. Polymer-Plastics Technology and Engineering. 40 (2): 103-115. Saad, A.L.G., Sayed, W.M., Ahmed, M.G.M. and Hassan, A.M.(1999). Preparation and Properties of Some Filled Poly (vinyl Chloride) Compositions. Journal of Applied Polymer Science.73: 2657-2670. Sameni, J.K., Ahmad, S.H. and Zakaria, S.(2003). Performance of Rubberwood FiberThermoplastics Natural Rubber Composites. Polymer-Plastics Technology and Engineering. 42(1): 139-152. Sanadi, A.R., Young, R.A., Clemons, C. and Rowell, R.M.(1994). Recycled Newspaper Fibres as Reinforcing Fillers in Thermoplastics: Part I-Analysis of Tensile and Impact Properties in Polypropylene. Journal Reinforced Plastics and Composites.13:54-67. Sapieha, S., Pupo, J.F. and Schreiber, H.P. (1989). Themal Degradation of CelluloseContaining Composites During Processing. Journal of Applied Polymer Science. 37: 233-240. Sarani, Z., Hazira, H., July Andrianny, M. and Mohamad, D. (2001). Chemical Modification on Lignocellulosic Polymeric Oil Palm Empty Fruit Bunch for Advanced Material. Advances in Polymer Technology. 20 (4): 289-295. Sarmoria, C. and Enrique, V. (2004). Model for a Scission-Crosslinking Process with both H and Y Crosslinks. Polymer.44: 5661-5669. 278 Sarvetnick, H.A.(1969). Polyvinyl Chloride. London: Reinhold Plastics Applications Series. Schlumpf, H.P.(1993). Fillers and Reinforcement. In : Wee, Y.J. (2001). Effect of Calcium Carbonate Fillers on Mechanical Properties of Impact-Modified PVC-U. Universiti Teknologi Malaysia : Thesis MSc. Schramm, G. (1965). Measuring The Fusion Rate of Rigid PVC Dry Blend. International Plastic Engineering. 1-7. Schwartz, W.T. (1987). The Effect of Fillers on the Impact Properties of Oxytuf High Impact PVC. Journal of Vinyl Technology. 9 (2): 46-52. Shackelford, J.F. (1996). Introduction to Materials Science for Engineers. 4th Edition. London: Prentice Hall Inc. Siegmann, A. and Hiltner, A.(1984). Impact Modification of Poly(vinyl chloride) with Chlorinated Polyethylene I. Blend Morphology. Polymer Engineering and Science. 24(11): 869-876. Sivaneswaran, K.K. (2002). Effect of Rice Husk Fillers on Mechanical Properties of ABS Impact Modified PVC-U. Universiti Teknologi Malaysia: Thesis MSc. Smith, W.F. (1996). Principle of Materials Science and Engineering.3rd Edition. New York: McGraw-Hill, Inc. Sombatsompop, N., Chaochanchaikul, K., Phromchirasuk, C. and Thongsang, S.(2003). Effect of Wood Sawdust Content on Rheological and Structural Changes, and Thermo-Mechanical Properties of PVC/Sawdust Composites. Polymer International. 52: 1847-1855. Sombatsompop, N. and Chaochanchaikul, K. (2004). Effect of Moisture Content on Mechanical Properties, Thermal and Structural Stability and Extrudate Texture of Poly(vinyl chloride)/Wood Sawdust Composites. Polymer International. 53: 1210-1218. Sombatsompop, N. and Phromchirasuk, C.(2004). Effects of Acrylic-Based Processing Aids on Processibility, Rheology, Thermal and Structural Stability, and Mechanical Properties of PVC/Wood-Sawdust Composites. Journal of Applied Polymer Science. 492: 782-790. Soni, P.L., Geil, P.H., and Collin E.A. (1981). Microdomain Structure in Plasticized PVC. J. Macromol. Sci.-Phys. B20 (4): 479-482. 279 Sreekala, M.S., Thomas, S. and Neelakantan, N.R.(1996/1997). Utilization of Short Oil Palm Empty Fruit Bunch Fiber(OPEFB) as a Reinforcement in PhenolFormaldehyde Resins: Studies on Mechanical Properties. Journal of Polymer Engineering. 16: 265-294. Sreekala, M.S. and Thomas, M.S.(2003). Effect of Fibre Surface Modification on Water-Absorption Characteristics of Oil Palm Fibres. Composites Science and Technology. 63: 861-869. Stark, N.M. and Matuana, L.M.(2004). Surface Chemistry Changes of Weathered HDPE/wood-flour Composites Studied by XPS and FTIR Spectroscopy. Polymer Degradation and Stability. 86(1):1-9. Stevenson, J.C.(1995). Impact Modifiers: Providing a Boost to Impact Performance. Journal of Vinyl and Additive Technology. 1(1): 41-45. Stokke, D.D.(2002). Fibre/Plastics Composite Materials: An Open House Biocomposite Group, Centre for Crops Utilization Research. January 23. Iowa State University, Ames, Iowa. Streitwieser, A. and Heathcook, C. (1981). Introduction to Organic Chemistry, 2nd Edition. New York:Macmillan Pub. Co., Inc. Strong, A.B.(2000). Plastics-Materials and Processing. 2nd Edition. London: Prentice Hall. Summers, J.W. (1981). The Nature of PVC Crystallinity-the Microdomain Structure. Journal. of Vinyl Technology.3(2):107-110. Summers, J.W.(1982). Poly(vinyl chloride) Processing Morphology. ANTEC 1982:501-503. Takase, S. and Shiraishi, N.(1989). Studies Composites from Wood and Polypropylene.II. Journal of Applied Polymer Science.37: 64-659. Teh., J.W., Rudin, A., Cooper, A.A. and Batisite J.L.H. (1989). Estimation of the Degree of Fusion of Rigid PVC from Thermal Analysis Measurements.” Makrmol. Chem., Macromol. Symp. 29: 123-132. Terselius, B. and Jansson, J.F. (1984). Gelation of PVC, Part 3: Effect on Tensile Properties. Plastics and Rubber Processing and Applications. 4: 291-299. Terselius, B. and Jansson, J.F. (1985). Gelation of PVC. Part 5 : Effect on Structure and Physical Properties. Plastics and Rubber Processing and Applications. 5:193-201. 280 Terselius, B. and Jansson, J.F. and Bystedt, J. (1981).Gelation of Rigid PVC-Pipes. J. Macromol. Sci.-Phys. B20: 403-414. Terselius, B. and Jansson, J.F.(1984). Gelation of PVC, Part 2 : Effect on Internal Pressure Resistance. Plastics and Rubber Processing and Applications. 4: 285-290. Terselius, B., Jansson, J.F. and Bystedt, J. (1985). Fracture Toughness of Rubber Modified PMMA.Plastics and Rubber Processing and Applications. 5: 1-7. Thiebaud, S., Borredon, M.E., Baziard, G. and Senocq,F. (1997). Properties of Wood Esterified by Fatty-Acid Chlorides. Bioresources Technology. 59: 103-107. Titow,W.V. (1991). PVC Plastics: Properties, Processing and Applications. New Yorks: Elsevier Applied Science Publisher. Torikai, A. and Hirose, H. (1999). Accelerated Photodegradation of Poly(vinyl chloride). Polymer Degradation and Stability.63: 441-445. Truss, R.W.(1985). Understanding Brittle Failure of uPVC (Unplasticised Polyvinyl Chloride) pipe. Pure and Applied Chemistry. 57(7): 993-1000. Uitenham, L.C. and Geil, P.H. (1981). Processing, Morphology and Properties of PVC. Macromolecules Science, Physic. B20(4): 593-622. Varma, A.J., Deshpande, S.V. and Kondapalli, P. (1999). A Comparative Study of the Thermal Behaviour of PVC: A Series of Synthesized Chlorinated Polyethylene and HDPE. Polymer Degradation and Stability. 63(1): 1-3. Ward, M. (1971).The Yield Behaviour of Polymers. Journal of Materials Science. 6. 1397-1417. Wan, C., Qiao, X., Zhang, Y. and Zhang, Y.(2003). Effect of Epoxy Resin on Morphology and Physical Properties of PVC/Organophilic Montmorillonite Nanocomposites. Journal of Applied Polymer Science. 89: 2184-2191. Wan, C., Qiao, X., Zhang, Y. and Zhang, Y.(2003). Effect of Different Clay Treatment on Morphology and Mechanical Properties of PVC-Clay Nanocomposites. Polymer Testing. 22: 453-461. Wee, Y.J. (2001). Effect of Calcium Carbonate Fillers on Mechanical Properties of Impact-Modified PVC-U. Universiti Teknologi Malaysia : Thesis MSc. Wenguang, M. and La Mantia, F.P. (1996). Processing and Mechanical Properties of Recycled PVC and Homopolymer Blends with Virgin PVC. Journal of Applied Polymer Polymer Science. 59(5): 759-767 281 Wiebking, H.(1986). The Use of Mineral Fillers in PVC: A Review of Principles. Pennsylvania:Pfizer Inc. Easton.155-163. Wimolmala, E., Wootthikanokkhan, J. and Sombatsompop, N. (2001). Effects of Composition and Temperature on Extrudate Chracteristics, Morphology, and Tensile Properties of Acrylic Rubber-Blended PVC. Journal of Applied Polymer Science. 80: 2523-2534. Whelan, T. and Goff, J.(1989). Unplasticised Poly(vinyl Chloride (uPVC; RPVC; PVC-U). British Plastics and Rubber. April 1989: 51-56. Woodhams, R.T., Thomas, G. and Rodgers, D.K.(1984). Wood Fibres as Reinforcing Fillers for Polyolefins. Polymer Engineering and Science. 24. 1166-1171. Wu, J., Yu, D., Chan, C.M., Kim,J. and Mai, Y.W.(2000). Effect of Fibre Pretreatment Condition on the Interfacial Strength and Mechanical Properties of Wood Fibre /PP Composites. Journal of Applied Polymer Science.76: 10001010. Wypych, J.(1985). Polyvinyl Chloride Degradation. New York:Elsevier Science Publishers. Xanthos, M. (1983). Processing Conditions and Coupling Agent Effects in PP/Wood Flour Composites. Plastics and Rubber Processing and Applications. 3: 223228. Xie, X.L, Li, B.G., Pan, Z.R., Li, R.K.Y. and Tjong, S.C.(2001). Effect of Talc/MMA in situ Polymerization on Mechanical Properties o PVC-Matrix Composites. Journal of Applied Polymer Science. 80: 2105-2112. Yagoubi, N., Baillet, A., Legendre, B., Raaron, A. and Ferrier, D.(1994). BetaRadiation Effects on PVC Materials: Methodology for Studying Chemical Modifications.Journal of Applied Polymer Science. 54: 1043-1048. Yarahmadi, N., Jakubowicz, I. and Gevert, T. (2001). Effects of Repeated Extrusion on the Properties and Durability of Rigid PVC Scrap. Polymer Degradation and Stability. 73: 93-99. Yap, M.G.S., Que, Y.T., Chia, L.H.L., and Chan, H.S.O.(1991). Thermal Properties of Tropical Wood-Polymer Composites. Journal of Applied Polymer Science.43: 2057-2065. Zaini, M.J., Ismail, Z., Fuad, M.Y.A. and Mustafah, J.(1994). Application of Oil Palm Wood Flour as Fillers in Polypropylene. Polymer Journal. 26(5): 637-642. 282 Zaini, M.J., Fuad, M.Y.A., Ismail, Z., Mansor, M.S., and Mustafah, J. (1996). The Effect of Filler Content and Size on the Mechanical Properties of Polypropylene/Oil Palm Wood Flour Composites. Polymer International. 40(1): 51-55.