Citation
High Impact Strength Polymers Having Novel Nano-Structures Produced via Reactive Extrusion

Material Information

Title:
High Impact Strength Polymers Having Novel Nano-Structures Produced via Reactive Extrusion
Creator:
TORTORELLA, NATHAN FRASER ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alloys ( jstor )
Copolymers ( jstor )
Crystallinity ( jstor )
Elastomers ( jstor )
Impact strength ( jstor )
Melting ( jstor )
Molecular weight ( jstor )
Monomers ( jstor )
Polymers ( jstor )
Styrenes ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Nathan Fraser Tortorella. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
4/17/2006
Resource Identifier:
77078655 ( OCLC )

Downloads

This item has the following downloads:

tortorella_n ( .pdf )

tortorella_n_Page_043.txt

tortorella_n_Page_037.txt

tortorella_n_Page_172.txt

tortorella_n_Page_073.txt

tortorella_n_Page_063.txt

tortorella_n_Page_204.txt

tortorella_n_Page_296.txt

tortorella_n_Page_222.txt

tortorella_n_Page_233.txt

tortorella_n_Page_192.txt

tortorella_n_Page_138.txt

tortorella_n_Page_250.txt

tortorella_n_Page_061.txt

tortorella_n_Page_097.txt

tortorella_n_Page_223.txt

tortorella_n_Page_048.txt

tortorella_n_Page_220.txt

tortorella_n_Page_122.txt

tortorella_n_Page_184.txt

tortorella_n_Page_304.txt

tortorella_n_Page_201.txt

tortorella_n_Page_098.txt

tortorella_n_Page_135.txt

tortorella_n_Page_106.txt

tortorella_n_Page_088.txt

tortorella_n_Page_199.txt

tortorella_n_Page_036.txt

tortorella_n_Page_008.txt

tortorella_n_Page_198.txt

tortorella_n_Page_130.txt

tortorella_n_Page_276.txt

tortorella_n_Page_185.txt

tortorella_n_Page_143.txt

tortorella_n_Page_302.txt

tortorella_n_Page_300.txt

tortorella_n_Page_081.txt

tortorella_n_Page_116.txt

tortorella_n_Page_131.txt

tortorella_n_Page_001.txt

tortorella_n_Page_039.txt

tortorella_n_Page_045.txt

tortorella_n_Page_224.txt

tortorella_n_Page_207.txt

tortorella_n_Page_148.txt

tortorella_n_Page_092.txt

tortorella_n_Page_057.txt

tortorella_n_Page_227.txt

tortorella_n_Page_023.txt

tortorella_n_Page_225.txt

tortorella_n_Page_163.txt

tortorella_n_Page_265.txt

tortorella_n_Page_269.txt

tortorella_n_Page_030.txt

tortorella_n_Page_307.txt

tortorella_n_Page_258.txt

tortorella_n_Page_268.txt

tortorella_n_Page_195.txt

tortorella_n_Page_140.txt

tortorella_n_Page_018.txt

tortorella_n_Page_205.txt

tortorella_n_Page_096.txt

tortorella_n_Page_011.txt

tortorella_n_Page_293.txt

tortorella_n_Page_022.txt

tortorella_n_Page_256.txt

tortorella_n_Page_004.txt

tortorella_n_Page_089.txt

tortorella_n_Page_238.txt

tortorella_n_Page_243.txt

tortorella_n_Page_108.txt

tortorella_n_Page_200.txt

tortorella_n_Page_129.txt

tortorella_n_Page_027.txt

tortorella_n_Page_082.txt

tortorella_n_Page_133.txt

tortorella_n_Page_056.txt

tortorella_n_Page_009.txt

tortorella_n_Page_309.txt

tortorella_n_Page_112.txt

tortorella_n_Page_303.txt

tortorella_n_Page_053.txt

tortorella_n_Page_100.txt

tortorella_n_Page_124.txt

tortorella_n_Page_230.txt

tortorella_n_Page_253.txt

tortorella_n_Page_016.txt

tortorella_n_Page_275.txt

tortorella_n_Page_050.txt

tortorella_n_Page_214.txt

tortorella_n_Page_067.txt

tortorella_n_Page_170.txt

tortorella_n_Page_044.txt

tortorella_n_Page_221.txt

tortorella_n_Page_160.txt

tortorella_n_Page_283.txt

tortorella_n_Page_226.txt

tortorella_n_Page_277.txt

tortorella_n_Page_274.txt

tortorella_n_Page_118.txt

tortorella_n_Page_064.txt

tortorella_n_Page_028.txt

tortorella_n_Page_125.txt

tortorella_n_Page_142.txt

tortorella_n_Page_245.txt

tortorella_n_Page_014.txt

tortorella_n_Page_145.txt

tortorella_n_Page_046.txt

tortorella_n_Page_180.txt

tortorella_n_Page_291.txt

tortorella_n_Page_015.txt

tortorella_n_Page_267.txt

tortorella_n_Page_193.txt

tortorella_n_Page_161.txt

tortorella_n_Page_290.txt

tortorella_n_Page_025.txt

tortorella_n_Page_074.txt

tortorella_n_Page_072.txt

tortorella_n_Page_285.txt

tortorella_n_Page_120.txt

tortorella_n_Page_155.txt

tortorella_n_Page_013.txt

tortorella_n_Page_101.txt

tortorella_n_Page_289.txt

tortorella_n_Page_176.txt

tortorella_n_Page_228.txt

tortorella_n_Page_244.txt

tortorella_n_Page_280.txt

tortorella_n_Page_035.txt

tortorella_n_Page_076.txt

tortorella_n_Page_114.txt

tortorella_n_Page_126.txt

tortorella_n_Page_059.txt

tortorella_n_Page_007.txt

tortorella_n_Page_093.txt

tortorella_n_Page_052.txt

tortorella_n_Page_213.txt

tortorella_n_Page_121.txt

tortorella_n_Page_252.txt

tortorella_n_Page_040.txt

tortorella_n_Page_177.txt

tortorella_n_Page_212.txt

tortorella_n_Page_147.txt

tortorella_n_Page_083.txt

tortorella_n_Page_051.txt

tortorella_n_Page_152.txt

tortorella_n_Page_058.txt

tortorella_n_Page_168.txt

tortorella_n_Page_196.txt

tortorella_n_Page_217.txt

tortorella_n_Page_069.txt

tortorella_n_Page_141.txt

tortorella_n_Page_149.txt

tortorella_n_Page_211.txt

tortorella_n_Page_086.txt

tortorella_n_Page_175.txt

tortorella_n_Page_020.txt

tortorella_n_Page_247.txt

tortorella_n_Page_209.txt

tortorella_n_Page_099.txt

tortorella_n_Page_105.txt

tortorella_n_Page_298.txt

tortorella_n_Page_062.txt

tortorella_n_Page_134.txt

tortorella_n_Page_287.txt

tortorella_n_Page_216.txt

tortorella_n_Page_017.txt

tortorella_n_Page_294.txt

tortorella_n_Page_038.txt

tortorella_n_Page_070.txt

tortorella_n_Page_288.txt

tortorella_n_Page_259.txt

tortorella_n_Page_033.txt

tortorella_n_Page_117.txt

tortorella_n_Page_187.txt

tortorella_n_Page_239.txt

tortorella_n_Page_254.txt

tortorella_n_Page_110.txt

tortorella_n_Page_078.txt

tortorella_n_Page_257.txt

tortorella_n_Page_179.txt

tortorella_n_Page_231.txt

tortorella_n_Page_075.txt

tortorella_n_Page_150.txt

tortorella_n_Page_144.txt

tortorella_n_Page_154.txt

tortorella_n_Page_264.txt

tortorella_n_Page_249.txt

tortorella_n_Page_190.txt

tortorella_n_Page_178.txt

tortorella_n_Page_297.txt

tortorella_n_Page_094.txt

tortorella_n_Page_003.txt

tortorella_n_Page_006.txt

tortorella_n_Page_010.txt

tortorella_n_Page_234.txt

tortorella_n_Page_188.txt

tortorella_n_Page_281.txt

tortorella_n_Page_041.txt

tortorella_n_Page_261.txt

tortorella_n_Page_260.txt

tortorella_n_Page_246.txt

tortorella_n_Page_235.txt

tortorella_n_Page_167.txt

tortorella_n_Page_242.txt

tortorella_n_Page_166.txt

tortorella_n_Page_026.txt

tortorella_n_Page_219.txt

tortorella_n_Page_115.txt

tortorella_n_Page_111.txt

tortorella_n_Page_049.txt

tortorella_n_Page_301.txt

tortorella_n_Page_103.txt

tortorella_n_Page_306.txt

tortorella_n_Page_087.txt

tortorella_n_Page_165.txt

tortorella_n_Page_128.txt

tortorella_n_Page_194.txt

tortorella_n_Page_236.txt

tortorella_n_Page_071.txt

tortorella_n_Page_136.txt

tortorella_n_Page_054.txt

tortorella_n_Page_240.txt

tortorella_n_Page_079.txt

tortorella_n_Page_090.txt

tortorella_n_Page_012.txt

tortorella_n_Page_270.txt

tortorella_n_Page_158.txt

tortorella_n_Page_032.txt

tortorella_n_Page_091.txt

tortorella_n_Page_255.txt

tortorella_n_Page_159.txt

tortorella_n_Page_202.txt

tortorella_n_Page_189.txt

tortorella_n_Page_113.txt

tortorella_n_Page_241.txt

tortorella_n_Page_162.txt

tortorella_n_Page_206.txt

tortorella_n_Page_208.txt

tortorella_n_Page_066.txt

tortorella_n_Page_305.txt

tortorella_n_Page_060.txt

tortorella_n_Page_308.txt

tortorella_n_Page_102.txt

tortorella_n_Page_237.txt

tortorella_n_Page_077.txt

tortorella_n_Page_157.txt

tortorella_n_Page_286.txt

tortorella_n_Page_055.txt

tortorella_n_Page_210.txt

tortorella_n_Page_272.txt

tortorella_n_Page_080.txt

tortorella_n_Page_248.txt

tortorella_n_Page_109.txt

tortorella_n_Page_278.txt

tortorella_n_Page_191.txt

tortorella_n_Page_271.txt

tortorella_n_Page_183.txt

tortorella_n_Page_186.txt

tortorella_n_Page_218.txt

tortorella_n_Page_156.txt

tortorella_n_Page_068.txt

tortorella_n_Page_024.txt

tortorella_n_Page_153.txt

tortorella_n_Page_107.txt

tortorella_n_Page_284.txt

tortorella_n_Page_123.txt

tortorella_n_Page_119.txt

tortorella_n_Page_095.txt

tortorella_n_Page_139.txt

tortorella_n_Page_085.txt

tortorella_n_Page_232.txt

tortorella_n_Page_203.txt

tortorella_n_Page_164.txt

tortorella_n_Page_042.txt

tortorella_n_Page_002.txt

tortorella_n_Page_029.txt

tortorella_n_Page_279.txt

tortorella_n_Page_047.txt

tortorella_n_Page_266.txt

tortorella_n_Page_299.txt

tortorella_n_Page_251.txt

tortorella_n_Page_174.txt

tortorella_n_Page_229.txt

tortorella_n_Page_034.txt

tortorella_n_Page_005.txt

tortorella_n_Page_146.txt

tortorella_n_Page_132.txt

tortorella_n_Page_031.txt

tortorella_n_Page_273.txt

tortorella_n_Page_181.txt

tortorella_n_Page_065.txt

tortorella_n_Page_104.txt

tortorella_n_Page_310.txt

tortorella_n_Page_282.txt

tortorella_n_Page_263.txt

tortorella_n_Page_295.txt

tortorella_n_Page_197.txt

tortorella_n_Page_084.txt

tortorella_n_Page_171.txt

tortorella_n_Page_292.txt

tortorella_n_Page_173.txt

tortorella_n_Page_127.txt

tortorella_n_Page_215.txt

tortorella_n_Page_137.txt

tortorella_n_pdf.txt

tortorella_n_Page_021.txt

tortorella_n_Page_169.txt

tortorella_n_Page_019.txt

tortorella_n_Page_182.txt

tortorella_n_Page_151.txt

tortorella_n_Page_262.txt


Full Text












HIGH IMPACT STRENGTH POLYMERS HAVING NOVEL NANO-STRUCTURES
PRODUCED VIA REACTIVE EXTRUSION















By

NATHAN FRASER TORTORELLA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Nathan Fraser Tortorella
































This dissertation is dedicated to my wife, Michelle, and my daughter, Katelynn.















ACKNOWLEDGMENTS

I would like to thank my advisor and mentor, Professor Charles L. Beatty. He

fostered my creativity and desire to succeed over the last several years and the KISS

principle is forever instilled in me. My committee members have been an invaluable

resource to me, namely Drs. Jack Mecholsky, Abbas Zaman, Hassan El-Shall, and Ken

Wagener. Their guidance on both a personal and professional level helped me succeed as

a graduate student. Funding for this research was graciously provided by Proctor and

Gamble. Chuck Yeazell and colleagues at P&G were extremely helpful in furthering the

development of this new material.

My wife Michelle has been unwavering in support, love, and understanding and our

beautiful daughter Katelynn is an inspiration. Stevan and Donna, my parents, and Harry

and Carol, Michelle's parents, deserve to be recognized for their encouragement and love

through thick and thin. I would like to thank my sisters, Stevany, Emily, and Danica, for

being there when I needed a quick laugh or to relax.

Many people have helped me throughout the years to eventually complete the

dissertation. In the Beatty research group, Ajit Bhaskar, Dr. Kunal Shah, and Xiosong

Huang were outstanding friends and colleagues. Thanks go to Mike Tollon for the SEM

pictures, Dr. Valentine Craciun for XRD, Dr. Laurie Gower and her students for optical

microscopy, James Leonard in the Wagener research group for GPC, and Phil McCartney

at Virginia Tech for TEM photos. Dr. Tony Brennan, Dr. Clay Bohn, Dr. Leslie Wilson,

and Michelle Carman were extremely helpful in training and support on several









instruments in the polymer characterization lab. Lastly, I would like to acknowledge the

staff in the Materials Science and Engineering Department for their help and dedication

over the last eight years.

















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES .............................................. .. ......................

LIST O F FIG U RE S ......................................................... ......... .. ............. xiii

ABSTRACT .............. .................. .......... .............. xxi

CHAPTER

1 GENERAL INTRODUCTION ...................................................... .....................

1.1 Introduction .................................................................................................. ....... 1
1.2 C chapter Sum m aries........... .............................................................. ...............

2 BACKGROUND AND LITERATURE REVIEW ................................................ 6

2 .1 P oly m ers of Interest .................................................................. .. .............. .
2.1.1 Isotactic Polypropylene ........................................ .......................... 6
2.1.2 Ethylene-1-Octene Copolym er.......................................... .. ............... 7
2 .2 P ro c e ssin g ................................ ............................................ 9
2.2.1 Reactive Twin Screw Extrusion ................................................................
2.2.2 Morphology Development in an Extruder Dispersive and Dissipative
M ixing ........................................... .. ............. .. ........ ......... 11
2.2.3 Melting and Droplet Breakup Mechanisms..........................................13
2.2.4 Morphology Development Dependency on Rate of Reaction..................16
2.2.5 Viscosity and Reaction Effects on Morphology Development.................. 18
2.3 Toughening of Polymers............. ............................ 19
2.3.1 Origins of Polymer Toughening............... .............................................. 19
2.3.2 Elastomer/Rubber Toughened Blends................................ ... ..................20
2 .4 F ree R adical R action s .............................................................. .....................23
2.4.1 Initiator D ecom position.......................................................................... 23
2.4.2 What Happens After Peroxide Decomposition? ............ ...............26
2.4.3 Polym erization...... .......... .... ........................ ...... ...... .............. 26
2.4.4 C opolym erization ............................................... ............................ 28
2.4.5 M ultifunctional M onom er ........................................ ....... ............... 30
2.4.6 H ydrogen A bstraction ........................................ .......................... 31









2.4.7 Polyethylene Crosslinking Reactions ............... ....... ................... 32
2.4.8 D egradation of Polypropylene .................. ............ ................................ 33
2.5 Previous Efforts to Reduce/Prevent Degradation of Polypropylene ....................35
2.5.1 Free Radical Grafting of Polymers..................................... 35
2.5.2 Solid State G rafting ........................................................ ............... 36
2.5.3 Reactions in the Polym er M elt ....................................... ............... 37
2.5.4 Fundamentals of Free Radical Grafting ...............................................39
2.5.5 Melt Grafting Monomers onto Polyolefins .............................................41
2.6 Melt Grafting and In-situ Compatibilization ............. ..... ..................43
2 .7 C o n c lu sio n s ..................................................................................................... 4 7

3 PHYSICAL BLENDING OF AN IMPACT MODIFIER WITH POLYPROPYLENE48

3 .1 In tro d u ctio n ..................................................................................................... 4 8
3 .2 E x p erim en ta l ................................................................................................4 8
3.2.1 M materials ......... ....... ............................................................ ... .. ............ 48
3 .2 .2 M eth o d s ...............................................................4 8
3 .2 .2 .1 P ro c e ssin g ................................................................................... 4 8
3.2.2.2 M mechanical properties .......................................... 50
3.2.2.3 M orphology ........... ... .... .. ..... .................. .. ............. 51
3.2.2.4 Thermal analysis and theology ................. ................. ....52
3.3 R results and D discussion ................... .....................53
3.3.1 Mechanical and Rheological Properties ............................................ 53
3.3.2 M orphology ................................................................... 59
3.3.3 V iscoelasticity ................ .................. ........................................62
3.3.4 C rystallinity ............................................ .. .....68
3.4 C onclu sions....................... .................................. ......72

4 TOUGHENED POLYPROPYLENE BASED ALLOYS .......................................74

4 .1 In tro d u ctio n .............. ........ .... ......................................................................7 4
4 .2 E xperim mental ........................... ..............................................74
4 .2 .1 M materials ......... ..... .................................................... ................ 7 4
4.2.2 M methods ..................... .................. ... ...... .......... ...... 76
4.2.2.1 Processing............................................. ............... 76
4.2.2.2 M mechanical properties ............................................... ............77
4.2.2.3 Morphological characterization................ ..... ............78
4.2.2.4 Chemical composition and molecular structure .............................. 81
4.2.2.5 Thermal analysis and theology ..................................... ..... 82
4.3 Results and Discussion .............................................................. ......... 84
4.3.1 Notched Izod Impact Analysis .................................. ..................86
4.3.2 Stress-Strain B ehavior........................................... .................... ... 91
4.3.3 Grafting onto Polyolefins ............. ............................. ............... 97
4.3.4 M orphology ........................................... ......... ..... ...............101
4.3.5 Possible Crosslinking of the System ............................. .................. 111
4.3.6 Rheological Properties....................................................112
4.3.7 Crystallinity and Crystallization ............. .......................... .................116









4 .3.8 V iscoelasticity ...................... ................ .................... ........138
4 .4 C o n c lu sio n s ................................................................................................... 14 5

5 THE EFFECT OF PROCESSING CONDITIONS ON ALLOY PROPERTIES ....147

5 .1 In tro d u ctio n ................................................................................................... 14 7
5.2 Experim mental ................................................... ............ ......... 147
5.2.1 M materials ................................ ........................... ...... 147
5.2.2 M methods ................................................................. ... ........ 149
5.2 .2 .1 P processing ............ ... ............................................... ............ 149
5.2.2.2 M mechanical properties ...................................................... 150
5.2.2.3 Chem ical com position......................................... ............... 151
5.2.2.4 R heology ........................................ ............................... 152
5.3 R results and D discussion .................... ......... .. ... .......... .............................. 152
5.3.1 Effect of Processing Tem perature .................................... ............... 152
5.3.2 Effect of Screw Speed ........................ .............. ................................ 155
5.3.3 Effect of Initiator Concentration .......... .... ...... ...... ............ 158
5.3.4 Effect of Multifunctional Monomer Concentration ..............................160
5.3.5 Effect of Styrene Concentration with DEGDA as Multifunctional
M o n o m er .................................. ..... ................ ............. ........ ...............16 3
5.3.6 Effect of Styrene Concentration with TMPTA as Multifunctional
M on om er....................................................................................... ..... 16 5
5 .4 C on clu sion s................................................. ................ 16 7

6 CONTROLLING ALLOY PERFORMANCE BY VARYING ELASTOMER
PROPERTIES .......... ..... .. ............................... .... ...... 168

6.1 Introduction ................. ...... .....................168
6.2 Experim mental ........... .. ..................................... .............. ... 168
6.2.1 Material s ................... ....... ..............168
6.2.2 Methods ....................................................................... ......... ..................170
6 .2 .2 .1 P processing ............ ... ............................................... ..... ....... 170
6.2.2.2 M echanical properties ......................................... ............... .... 171
6.2.2.3 Chemical composition and molecular structure..........................172
6.2.2.4 Thermal analysis and rheology ............................................... 174
6.3 R results and D discussion ............... ......... .................................................. 174
6.3.1 Effect of Elastomer Density on Alloy Performance.................................174
6.3.1.1 Physical blends ................................................ .......... ............ .. 175
6.3.1.2 A lloys .............. ..... ........ .... .............. ........ ...... 183
6.3.2 Effect of Elastomer Molecular Weight on Alloy Performance..............190
6.3.2.1 P physical blends ......................... ...................... .. .. .... .......... 193
6.3.2.2 A lloys ................ ............................ ............. .................... 200
6.3.3 Supercritical CO2 as a Possible Route to Improve Grafting Efficiency ...210
6.4 Conclusions..................................................................... ... ....... ..... 211









7 REACTIVE EXTRUSION OF HIGH DENSITY POLYETHYLENE ...................213

7 .1 In tro d u ctio n ................................................................................................... 2 1 3
7.2 Experim mental ........... .. ..................................... .............. ... 213
7.2.1 M materials .................................................................................... ............... 2 13
7 .2 .2 M eth o d s .............................................................................................. 2 1 5
7.2 .2 .1 P processing ............ ... ....... .. .............. .......... ...... ............2 15
7.2.2.2 M mechanical properties ....................................... ............... 216
7.2.2.3 D esign of experim ents................................................................ 217
7 .3 R results and D iscu ssion ............................................................ .....................2 18
7.3.1 M echanical Properties ................................................... ............... 218
7.3.2 Design of Experiments Results ......... .................................. 221
7.3.2 .1 Im pact strength ......... .......................................................... .....22 1
7.3.2.2 E plastic m odulus ........................................ ......................... 225
7.3.2.3 Yield strength ............................ ............ ..... .............. .228
7.3.2.4 Elongation at break.. ....................... ........................................ 231
7 .4 C o n c lu sio n s................................................................................................... 2 3 4

8 CONCLUSIONS AND FUTURE WORK............................ ......... ............ 236

8.1 Sum m ary and Conclusions ......................... ..........................................236
8.2 Future W ork ........................................ ...................... 240

APPENDIX

A CALIBRATION CURVE FOR ABSOLUTE STYRENE CONCENTRATIONS IN
R EA C TIV E B LEN D S ........................................... .................. ............... 244

B STRESS-STRAIN GRAPHS AND STATISTICS..................................................246

C F T IR G R A P H S .................... .. .. ...... .... .... .. .............................. .......... .. .. ...2 56

D IM A G E A N A L Y SIS ........................................................................ ..................262

E ALIASED TERMS FROM CHAPTER 7 ..................................... ...............265

LIST OF REFEREN CES ................................... ............................. ............... 266

BIOGRAPH ICAL SKETCH ...................................................... 288
















LIST OF TABLES


Table page

2-1 The mechanisms of main reactions in the grafting process.................................23

2-2 Characteristics of Lupersol 101......................................... ........................... 24

2-3 Half life of Lupersol 101 based on temperature.....................................................25

2-4 Typical free radical kinetic values in solution. .............................. ......... ...... .27

2-5 Copolymerization constants for styrene and ethyl acrylate monomers....................29

2-6 Structural and physical information about the monomers of interest. ...................29

3-1 ENGAGE product data table ........................................................... 49

3-2 DSC consecutive heating/cooling cycles ...................................... ............... 53

3-3 Image analysis averages taken from etched SEM images of PP:8407 physical
blends. At least three images were analyzed from Figures 3(a) thru (e), with data
com piled in A ppendix D ............................................... .............................. 61

3-4 DSC endothermic data comparing pure PP and 8407 to physical blends of PP and
8407. Standard deviation for pure PP is from an average of four samples ............70

3-5 DSC exothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10 B. An
average of 3 runs were performed on 90:10_B ..................................................72

4-1 ENGAGE product data table ........................................................... 75

4-2 Structures of reactive materials of interest .............................. .... ............75

4-3 PLM consecutive heating/cooling cycles...... .......... ...............81

4-4 D SC consecutive heating/cooling cycles ..................................... .................83

4-5 Identification of formulations.... ... .............................................. .. ............... 86

4-6 Image analysis of etched SEM surfaces of several blends and alloys. This
information is gathered from Appendix D, Figure 4-16, and Figure 3-5. Ligament
thickness is calculated using the number average diameter..............................102









4-7 Melt behavior of PP, PP-8407 physical blends, and PP-8407 alloys ..................... 115

4-8 DSC endothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10_B. An
average of three runs were performed on 90:10_B ......................................127

4-9 DSC exothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10_B.....131

5-1 ENGAGE product data table............................ ....... ....................148

5-2 Structures of reactive m materials of interest............................................................ 148

5-3 Three different temperature profiles for the extrusion of 80:20_C......................152

5-4 Stress-strain behavior of 80:20_C as a function of temperature. .........................154

5-5 Screw speed relationship to residence time................................. ...... ............ ...156

5-6 Stress-strain behavior of 80:20_C as a function of screw speed .........................158

5-7 Stress-strain behavior of PP:8407 alloys at a ratio of 90:10 as a function of initiator
concentration. .......................................................................160

5-8 Stress-strain behavior of PP:8407 alloys at a ratio of 90:10 as a function of
multifunctional monomer concentration. ............. .......................... ............... 163

5-9 Stress-strain performance of PP:8407 alloys at a ratio of 90:10 as a function of
styrene concentration................... ..... .......... ......... ....... 165

6-1 ENGAGE product data table............................ ....... ....................169

6-2 Structures of reactive m materials of interest............................................................ 170

6-3 DSC consecutive heating/cooling cycles .................................... ............... 174

6-4 Identification of Formulations in relation to Figure 6-2 .....................................175

6-5 Stress-strain data of 8407_0, 8401_0, and 8402 0. ........................ ............177

6-6 DSC endothermic data of 8407 0, 8401 0, and 84020 ......................................179

6-7 DSC exothermic data of 8407 0, 8401 0, and 8402_0 ......................................181

6-8 Stress-strain data of PP:ENGAGE alloys as a function of elastomer density........186

6-9 DSC endothermic data of 8407_A, 8407 B, 8402 A, and 8402_B ......................188

6-10 DSC exothermic data of 8407 A, 8407 B, 8402 A, and 8402B. .......................188

6-11 DSC endothermic data of 002 0, 8407 0, 8200 0, and 8842 0 .........................197









6-12 DSC exothermic data of 002_0, 8407_0, 8200_0, and 88420 .............................199

6-13 Stress-strain behavior of 002 B, 8407 B, 8200 B, and 8842 B. .......................205

6-14 DSC endothermic data of 8407_A, 8407 B, 8842 A, and 8842_B ......................208

6-15 DSC exothermic data of 8407 A, 8407 B, and 8842 A, and 8842B ..................209

7-1 EN G A GE product data table .......................................... ...... .....................214

7-2 Structures of reactive materials of interest..................................... ..................214

7-3 Factors of interest, coded variables, and levels for the impact modification of
H D P E .............................................................................2 17

7-4 Basic fractional factorial design using coded variables from Table 7-3..............218

7-5 Impact strength, elastic modulus, yield stress, and elongation at break results from
the fractional factorial design created for the modification of HDPE....................220

7-6 Analysis of variance table (partial sum of squares) for impact strength..............223

7-7 Analysis of variance table (partial sum of squares) for elastic modulus..............226

7-8 Analysis of variance table (partial sum of squares) for yield stress.......................230

7-9 Analysis of variance table (partial sum of squares) for elongation at break..........232

B-l Actual stress-strain values with standard deviations from figures in both Chapter 3
and C chapter 4. ........................................................................246

B-2 Stress-strain properties of PP:ENGAGE blends and alloys as a function of
elastom er m elt flow index. ............................................. ............................. 253

E-1 Aliased terms from the fractional factorial design given in Chapter 7 ................265
















LIST OF FIGURES


Figure page

2-1 The phase inversion mechanism proposed by Shih. ..............................................14

2-2 Morphology development of binary polymer blends proposed by Scott and
M acosko. ............................................................................15

2-3 Chem ical structure of Lupersol 101 .............................................. ............... 24

2-4 Schematic of H abstraction from PP and subsequent P-chain scission ..................34

3-1 Schematic drawing of the reactive twin screw extruder and a common temperature
profile. ..............................................................................50

3-2 Effect elastomer concentration on room temperature notched impact strength and
melt flow index ................................... ..... ......... 54

3-3 Stress-strain behavior of PP:8407 physical blends ............................................... 56

3-4 Stress-strain properties of PP-8407 blends as a function of elastomer content. ......57

3-5 SEM images of etched, cryo-fractured surfaces of PP:8407 physical blends as a
function of elastomer concentration. (a) Virgin PP at 2,000X, (b) 95:5_0 at
10,000X, (c) 90:10_0 at 10,000X, (d) 80:20_0 at 10,000X, and (e) 70:30_0 at
2,500X. The bar markers for (a) and (e) = 10tm, and for (b), (c), and (d) = l1 m. 60

3-6 Storage modulus vs. temperature for PP:8407 Blends...........................................63

3-7 Tan6 vs. temperature for PP:8407 blends from -1200C to 1200C. Insert represents
Tan6 vs. temperature from -600C to 600C. ................................... ............... 65

3-8 Loss modulus vs. temperature for PP:8407 physical blends...............................68

3-9 DSC melting endotherm of both pure PP and 8407 as well as blends of the two
polym ers. .............................................................................69

3-10 Percent crystallinity of the physical blends and pure PP as a function of PP
concentration ...................................................... ................. 70

3-11 DSC cooling exotherm of both pure PP and 8407 as well as blends of the two
polym ers. .............................................................................7 1









4-1 Schematic drawing of the reactive twin screw extruder and a common temperature
profile. .............................................................................. 77

4-2 FTIR image of PP:polyisoprene reactive blend. ............ ................................... 85

4-3 Representation of sample reference code........................... ..................85

4-4 Room temperature notched Izod impact strength of PP blends and alloys.
Reference material is virgin PP at 0.99 ft-lbs/in. ............. ........................ ......... 87

4-5 Izod Impact test specimens post-fracture. Left- Pure PP, Middle 90:10_0, Right
90 :10 B ............................................................................88

4-6 SEM image showing the tip of an arrested crack from a room temperature notched
Izod impact test of 90:10_0. Insert is a magnified image of crack tip ...................89

4-7 SEM image showing the tip of an arrested crack from a room temperature notched
Izod impact test of 90: 10 B .......................... .............. ............................90

4-8 SEM image of 80:20_C (left) and Pure PP (right) fractured at room temperature
without etching. Each image is located at the center of the impact bar, with
magnification = 2,000X and marker bar = 10m..................................................91

4-9 Stress-strain properties of 95:5 alloys compared to the 95:5 physical blend...........92

4-10 Stress-strain properties of 90:10 alloys as compared to the 90:10 physical blend...93

4-11 Stress-strain properties of 80:20 alloys as compared to the 80:20 physical blend...93

4-12 Stress-strain properties of 70:30 alloys as compared to the 70:30 physical blend...94

4-13 FTIR image of a typical PP:8407 alloy containing styrene and multifunctional
a c ry la te ........................................................................... 9 8

4-14 Styrene grafting efficiency at various 8407 concentrations both with and without
m ultifunctional m onom er ............................................... .............................. 99

4-15 Molecular weight averages for pure 8407 and three grafted 8407 materials.........101

4-16 SEM images of(a) 95:5_B at 10,000X, (b) 90:10_B at 10,000X, (c) 80:20_B at
10,000X, and (d) 70:30_B at 5,000X. Each bar marker = 1m ..........................103

4-17 Matrix ligament thickness of PP blends and alloys as a function of volume % of
84 07 ...............................................................................10 5

4-18 Room temperature notched Izod impact strength as a function of matrix ligament
thickness for blends and alloys at various volume % 8407.................................105









4-19 TEM bright field images of (a) virgin PP and (b) 90:10_0 stained with Ru04 at a
magnification of 63,500X. Bar marker on insert = 100nm.............................. 107

4-20 TEM image of 90:10_A at 63,500X. All marker bars = 100nm. .......................... 108

4-21 TEM image of 90:10B at 63,500X. All marker bars = 100nm. ........................109

4-22 TEM images of 80:20_A (left) and 80:20_B (right) at 63,500X. ..........................110

4-23 TGA graphical comparison of a physical blend (90:10_0) and alloy (90:10 B)...112

4-24 Melt flow index of physical blends and alloys as a function of 8407 content....... 114

4-25 XRD pattern of pure PP, 90:10_0, 90:10_A, and 90:10_B............... ................ 118

4-26 Alignment of lamellae within spherulites of a-PP (left) and P-PP (right) from
reference 278. ........................................................................12 1

4-27 Etched lamellar morphology of isotactic polypropylene of well defined spherulite
at 15kX magnification (a), bar marker = 2am and periphery of a spherulite at 20kX
magnification (b), bar marker = 2 m ......................................... ............... 123

4-28 Etched lamellar morphology of 90:10_0 at 10.5kX magnification, with the bar
m arker = 5[tm ........................................................................ 124

4-29 Etched lamellar morphology of 90:10_0 at 20kX magnification, with the bar
m arker = 2 am ........................................................................ 125

4-30 DSC heating traces of a physical blend (90:10_0) and alloys (90:10_A and
90:10 B)...................................................................... ........... 126

4-31 DSC heating traces of a physical blend (80:20_0) and alloys (80:20_A, 80:20_B,
and 80:20_C without styrene). ............................................................................ 129

4-32 DSC cooling traces of a physical blend (90:10_0) and alloys (90:10_A and
90:10 B)..................................... .................. .............. ........... 130

4-33 Polarized optical images of virgin PP cooled from the melt. (a) = 115.5C at 50X,
(b) = 113.50C at 50X, (c) = 1100C at 50X, and (d) = 250C at 20X magnification.133

4-34 Polarized optical images of 90:10_0 cooled from the melt. (a) = 1180C at 50X, (b)
= 114.5C at 50X, (c) = 105C at 50X, and (d) = 25C at 20X magnification. A
bubble (artifact) appears in the lower left hand of (b) and (c) and along the edges of
(d) ........................................................ ................................ 134

4-35 Polarized optical images of 90:10_A cooled from the melt. (a) = 1280C at 50X, (b)
= 1240C at 50X, (c) = 1220C at 50X, and (d) = 250C at 20X magnification.........135









4-36 Polarized optical images of 90:10_B cooled from the melt. (a) = 1350C at 50X, (b)
= 1330C at 50X, (c) = 1300C at 50X, and (d) = 250C at 20X magnification.........137

4-37 Storage modulus (E') and Tan6 comparison of 95:5_0 and 95:5 B as a function of
tem perature ............................................................ ... .... ......... 139

4-38 Storage modulus (E') and Tan6 comparison of 90:10_0, 90:10_A, and 90:10_B as a
function of tem perature. ...................... ................... ................. ............... 140

4-39 Tan6 comparison of 90:10_0, 90:10_A, and 90:10_B between -500C and 600C to
show a magnified graph of p8407 and pp ............................. ..... ..................... 141

4-40 Storage modulus (E') and Tan6 comparison of 80:20_0 and 80:20_B as a function
of tem p eratu re .................................................................... 14 3

4-41 Storage modulus (E') and Tan6 comparison of 70:30_0 and 70:30_B as a function
of tem perature. ......................................................................144

5-1 Schematic drawing of the reactive twin screw extruder and a common temperature
p ro file .......................................................................... 14 9

5-2 Effect of extruder barrel temperature on room temperature impact strength, melt
flow index, and grafting efficiency of 80:20_C. ............. ..................................... 153

5-3 Impact strength, MFI, and grafting efficiency of 80:20_C as a function of screw
speed .............. .......... ......... ........................................... 157

5-4 Impact strength, MFI, and grafting efficiency of 90:10_B as a function of initiator
concentration. .......................................................................159

5-5 Notched impact strength, melt flow index, and grafting efficiency of PP:8407
alloys at a ratio of 90:10 as a function of multifunctional monomer concentration. 161

5-6 Notched impact strength, melt flow index, and grafting efficiency of PP:8407
alloys at a ratio of 90:10 as a function of styrene monomer concentration ..........164

5-7 Notched impact strength, melt flow index, and grafting efficiency of PP:8407
alloys at a ratio of 80:20 as a function of styrene concentration............................166

6-1 Schematic drawing of the reactive twin screw extruder and a common temperature
p ro file .......................................................................... 17 1

6-2 Representation of sample reference code........................ ... ..... .......... 175

6-3 Impact strength, melt flow index, and melting temperature of physical blends as a
function of the density of the copolymer. ................................... ............... 176









6-4 Digital images of room temperature fractured Izod impact bars as a function of
elastomer density in the physical blends. From left to right: 8407_0, 8401_0,
8402 0. .............................................................................. 177

6-5 Stress-strain performance of PP/elastomer physical blends as a function of
elastom er density........ ... ...... ................................ .. ... ... .... .. ...... .... 178

6-6 DSC melting endotherm of 8407_0, 8401_0, and 8402_0................................179

6-7 DSC cooling exotherm of PP/elastomer physical blends as a function of elastomer
density. Insert is for the temperature range or 20 to 100C. ...............................180

6-8 Dynamic mechanical analysis of 8407_0, 8401_0, and 8407_0. Insert is a
magnified graph from -600C to 600C of the 0 relaxation of PE........................... 182

6-9 Room temperature impact strength, melt flow index, and grafting efficiency of
8407 B, 8401 B and 8402 B ............................... ... ................................. 183

6-10 Room temperature impact strength, melt flow index, and grafting efficiency of
8407 C 8401 C and 8402 C ........................................ .......................... 185

6-11 DSC melting endotherms of 8407_A, 8407_B, 8402_A, and 8402_B ranging from
60C to 1800C. Insert is a magnified graph of the a melting peak of PP.............. 187

6-12 DSC cooling exotherms of PP/elastomer physical blends as a function of elastomer
density. Insert is for the temperature range of 20 to 1000C ...............................189

6-13 Dynamic Mechanical Analysis comparison of (a) 8402_0 and 8402 B, (b) 8401_0
and 8401_B, and (c) 8407_0 and 8407_B ........................................................... 190

6-14 Molecular weight averages of the polyolefin elastomers of interest, ranked
according to their m elt flow index. ............................................. ............... 192

6-15 Polydispersity (molecular weight distribution) of the polyolefin elastomers of
interest. ............................................................................. 193

6-16 MFI and room temperature notched Izod impact strength of 8842_0 (highest
molecular weight), 8200_0, 8407_0, and 002_0 (lowest molecular weight). .......194

6-17 Stress-strain performance of physical blends as a function of MFI of the copolymer. 196

6-18 DSC melting endotherms of 002_0, 8407_0, 8200_0 and 8842_0......................198

6-19 DSC crystallization exotherms of 002_0, 8407_0, 8200_0, and 8842_0...............199

6-20 Impact Strength and melt flow index of 8842_A, 8200_A, 8407_A, and 002_A..201

6-21 Izod impact strength and melt flow index of 8842_C, 8200_C, 8407_C, and 002_C.202









6-22 Izod impact strength and melt flow index of 8842_B, 8200_B, 8407_B, and 002_B.203

6-23 Dynamic mechanical behavior (E' and Tan6) of various alloys and blends. (a)
002_0 vs. 002 B, (b) 8407_0 vs. 8407 B, (c) 8200_0 vs. 8200_B, and (d) 8842_0
v s. 8 84 2 B .......................................................................... 2 0 6

6-24 DSC melting endotherm of 8407_A, 8407_B, 8842_A, and 8842_B....................208

6-25 DSC crystallization exotherm of 8407_A, 8407_B, 8842_A, and 8842_B..........209

6-26 Effect of supercritical carbon dioxide at 1500 psi on the grafting efficiency of
8842_B Injection w as in zone 3............ .................. ..................... ............... .211

7-1 Schematic drawing of the reactive twin screw extruder and a common temperature
p ro file .......................................................................... 2 15

7-2 Half normal plot showing, in coded variables, the most significant effects on
impact strength. A = initiator concentration and C = concentration of 8842........222

7-3 Normal plot of residuals (a) and outliers (b) show the diagnostic results of the
m odel for im pact strength. ............................................ ............................. 224

7-4 Cube graph of the effect of % initiator, 8842 content, and % styrene on impact
strength at constant screw speed, % DEGDA, and temperature ..........................225

7-5 Half normal plot showing, in coded variables, the most significant effects on elastic
m odulu s. ...........................................................................226

7-6 Normal plot of residuals (a) and outliers (b) show the diagnostic results of the
m odel for elastic m odulus. ............................................ ............................. 227

7-7 Cube graph of the effect of % initiator, 8842 content, and % styrene on elastic
modulus at constant screw speed, % DEGDA, and temperature .........................228

7-8 Half normal plot showing, in coded variables, the most significant effects. .........229

7-9 Normal plot of residuals (a) and outliers (b) show the diagnostic results of the
m odel for yield stress. ...................... .. .... .............. ........................... 230

7-10 Cube graph of the effect of % initiator, 8842 content, and % styrene on yield stress
at constant screw speed, % DEGDA, and temperature ................ .....................231

7-11 Half normal plot showing, in coded variables, the most significant effects. .........232

7-12 Normal plot of residuals (a) and outliers (b) show the diagnostic results of the
m odel for elastic m odulus. ............................................ ............................. 233

7-13 Cube graph of the effect of % initiator, 8842 content, and % styrene on elongation
at break at constant % styrene, % DEGDA, and temperature.............................234


xviii









8-1 Schematic drawing of the likely free radical initiated processes during the reactive
extrusion of PP, 8407, initiator, styrene, and multifunctional monomer. ..............237

8-2 Interpretation of the effect of in-situ grafted polymeric chains at the PP-elastomer
interface (blue circles = elastomer domains, black lines = grafted polymers). The
physical blend (left) has no grafting and the alloy (right) has a high degree of
g raftin g ......................................................................... 2 3 9

B-l Stress-strain graph comparison of 95:5 (PP:8407) physical blend and alloys.......247

B-2 Stress-strain graph comparison of 90:10 (PP:8407) physical blend and alloys. ....247

B-3 Stress-strain graph comparison of 80:20 (PP:8407) physical blend and alloys.....248

B-4 Stress-strain graph comparison of 70:30 (PP:8407) physical blend and alloys.....248

B-5 Stress-strain graph comparison of the effect of extruder barrel temperature in
relation to 5-2 and Table 5-4. 1= low temperature, 2=middle temperature, 3=high
tem p eratu re................................................... ......... .................................2 4 9

B-6 Stress-strain graph comparison of the effect of extruder screw speed in relation to
T able 5-6. ..........................................................................249

B-7 Stress-strain graph comparison of the effect of initiator concentration in relation to
T able 5-7 ............................................................................250

B-8 Stress-strain graph comparison of the effect of DEGDA concentration in relation to
T ab le 5 -8 ............................................................................ 2 5 0

B-9 Stress-strain graph comparison of the effect of styrene concentration in relation to
T able 5-9. ................................................................ .. ..... ........ 251

B-10 Stress-strain graph comparison of the effect of elastomer density in relation to
Figure 6-5 and Table 6-5 ...................................................................... 251

B-11 Stress-strain graph comparison of the effect of elastomer density in relation to
T able 6-8. .......................................................................... 252

B-12 Stress-strain graph comparison of the effect of elastomer density in relation to
T able 6-8. .......................................................................... 252

B-13 Stress-strain graph comparison of the effect of elastomer density in relation to
T able 6-8. .......................................................................... 253

B-14 Stress-strain graph of physical blends of ENGAGE elastomers with PP as a
function of elastomer melt flow index ...................................... ............... 254

B-15 Stress-strain graph comparison of the effect of elastomer MFI..........................254









B-16 Stress-strain graph comparison of the effect of elastomer MFI ...........................255

C-1 FTIR graph comparison of 95:5_0, 95:5_A, 95:5 B, and 95:5_C.........................256

C-2 FTIR graph comparison of 90:10_0, 90:10_A, 90:10_B, and 90:10_C.................257

C-3 FTIR graph comparison of 80:20_0, 80:20_A, 80:20_B, and 80:20_C............257

C-4 FTIR graph comparison of 70:30_0, 70:30_A, 70:30_B, and 70:30_C.................258

C-5 FTIR graphs from Chapter 5 (Figure 5-2) of alloys processed at varying
temperatures. 1 = low barrel temperature, 2 = middle barrel temperature, 3 = high
barrel tem perature. ........................... ................... ...... .............258

C-6 FTIR graphs from Chapter 5 (Figure 5-3) of alloys processed at varying screw
sp eed s. .......................................................................... 2 5 9

C-7 FTIR graphs from Chapter 5 (Figure 5-4) of alloys processed at varying
concentration of initiator................................ ......... ......... ............................259

C-8 FTIR graphs from Chapter 5 (Figure 5-5) of alloys processed at varying
concentration of multifunctional monomer............................ .............260

C-9 FTIR graphs from Chapter 5 (Figure 5-6) of alloys processed at varying
concentration of styrene, with DEGDA as multifunctional monomer.................260

C-10 FTIR graphs from Chapter 5 (Figure 5-7) of alloys processed at varying
concentration of styrene, with TMPTA as multifunctional monomer .................261

D-1 Histogram of average particle diameters for (a) 70:30_0 and (b) 70:30_B ..........262

D-2 Histograms of particle roundness for (a) 70:30_0 and (b) 70:30_B ......................262

D-3 Histogram of average particle diameters for (a) 80:20_0 and (b) 80:20B ...........263

D-4 Histograms of particle roundness for (a) 80:20_0 and (b) 80:20_B ......................263

D-5 Histogram of average particle diameters for (a) 90:10_0 and (b) 90:10B ...........263

D-6 Histograms of particle roundness for (a) 90:10_0 and (b) 90:10B ..................264

D-7 Histogram of average particle diameters for (a) 95:5_0 and (b) 95:5_B .............264

D-8 Histograms of particle roundness for (a) 95:5_0 and (b) 95:5_B .........................264















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

HIGH IMPACT STRENGTH POLYMERS HAVING NOVEL NANO-STRUCTURES
PRODUCED VIA REACTIVE EXTRUSION

By

Nathan Fraser Tortorella

December 2005

Chair: Charles L. Beatty
Major Department: Materials Science and Engineering

A major focus of scientists and engineers over the last century has been to increase

the impact strength and therefore reduce the brittleness of materials. By altering and

adding energy absorption mechanisms, brittle failure can be averted. Isotactic

polypropylene (PP) is the focus of this dissertation because it is an extremely low cost,

high volume, versatile plastic but behaves in a brittle manner at or below room

temperature or in a notched state. Early work on impact modification of polypropylene

focused on blending energy-absorbing low density elastomers and rubbers. These binary

blends all had a common problem an increase in impact strength was paralleled by a

significant decrease in both elastic modulus and yield stress.

Reactive extrusion processing has allowed the in-situ compatibilization of isotactic

polypropylene and metallocene-catalyzed ethylene-octene copolymers (EOCs). This

process involves combining both the comonomer and vector fluid approaches to grafting

polyolefins. Styrene monomer and a multifunctional acrylate monomer undergo









peroxide-induced copolymerization and grafting in the presence of both PP and EOC.

This results in a phase separated alloy with an impact strength over 13 times that of pure

polypropylene and double that of the physical blend. There is also a significant

improvement in stress-strain performance when comparing the alloys to physical blend

counterparts.

Many researchers have categorized the necessary components to toughening

polypropylene as pertaining to the amorphous phase. The alloys described in this

dissertation meet the criteria put forth by these researchers, namely low density,

crystallinity, and modulus of the elastomer phase, sub-micron particle diameter, close

inter-particle distance, and a high degree of entanglements of both the PP matrix phase

and EOC minor phase. But many people neglect to study the crystalline state of impact

modified PP in conjunction with the amorphous phase. This work shows that the typical

10-100 tm diameter spherulitic structures found in pure PP are not present in the alloys.

In fact, the spherulites are less than a micron in diameter, are uniformly distributed

throughout the sample, and crystallize at much higher temperatures. SEM images, when

coupled with DSC and XRD, reveal the presence of a high number of small lamellar

crystals composed of a unique highly dense cross-hatched structure. Thus, impact

strength and stiffness can be simultaneously improved by controlling the size and cross-

hatch density of the lamellar crystals and applying phase transformation toughening

concepts.














CHAPTER 1
GENERAL INTRODUCTION

1.1 Introduction

The field of materials science and engineering has emerged as a premier discipline

which encompasses the development, synthesis, and processing of new materials.

Calister describes materials science as investigating the relationships that exist between

the structures and properties of materials, whereas materials engineering is based upon

designing or engineering the structure of a material to produce a predetermined set of

properties [1]. The structure-processing-properties relationship holds true for the creation

of all advanced materials, with polymers, metals, ceramics, and electronic materials as

the core classes. Polymers are the focus of the dissertation, which are essentially organic

macromolecules chemically based on carbon, hydrogen, and other nonmetallic elements.

Polymer blending is an economic process to create a material with a balance of

properties that would otherwise not be possible [2-12]. If a material can be generated

that will lower the cost while maintaining or improving performance of a particular

product then the manufacturer must use it to remain competitive. Many engineering

resins may lack chemical resistance, impact toughness, flame retardency, high

temperature performance, or weatherability, which can be solved via blending with other

engineering or commodity plastics. The development of a new blend or compound from

existing materials is generally more rapid than that of an entirely new polymer.

Polymer blends can be characterized by their phase behavior as being either

miscible or immiscible immisciblee blends having multiple amorphous phases) [10, 12-









14]. A blend's mechanical, thermal, and theological properties, amongst others, depend

strongly on its state of miscibility. Nearly all polymer pairs are immiscible, forming two-

phased systems in which the interface is a source of weakness. Polymer-polymer

immiscibility is a direct consequence of the high molecular weights of the polymer

molecules. Blending two polymers is not thermodynamically favorable because the

mixing of a relatively low number of molecules leads to a positive enthalpy of mixing

and low entropy of mixing [8, 15-18].

By simply blending two immiscible polymers, the resulting material has improved

properties but usually at the expense of another property. For example, adding rubbery or

elastomeric particles to a polypropylene (PP) matrix increases impact strength but

sacrifices yield strength and elastic modulus [19-24]. Previous routes to toughen PP have

their drawbacks, such as the migration phenomenon with plasticizers and costly in-

reactor copolymerization of ethylene and propylene. In order to overcome these

problems, many researchers have found that by controlling the size and distribution of the

rubber particles, cohesive strength of the elastomer, the degree of physical entanglements

in the system, and interfacial adhesion between the rubber and the matrix, all properties

are vastly improved over the original blend.

Physically blending an elastomer with a brittle semicrystalline polymer is not

sufficient for overall property improvement, so a reactive extrusion process

(compatibilization) is often applied to create novel polymeric alloys. The term alloy has

been defined by the Polymer Technology Dictionary as a composition, or blend, which is

based on two or more polymers, the properties of which are significantly better than

would be expected from a simple blend [25]. The system is typically phase separated









with a certain degree of chemical bonding or grafting between phases. Compatibilizers

are sometimes used to control the adhesion aspect of these blends, which therefore results

in finer dispersed phase morphology, uniform distribution of domains, better

processability, enhanced mechanical properties, and increased thermal stability. They

effectively act as high molecular weight surfactants by locating at the interface between

the immiscible polymers [12, 14, 26-40].

Compatibilizers are traditionally thought of as block or graft copolymers which

contain functional groups that may or may not react with the polymers present in the

blend. Three drawbacks of these pre-made copolymers are that they are expensive,

unstable upon annealing, and can cause a substantial increase in viscosity during

processing [41] The effectiveness of these materials are also diffusion dependent and

may not entirely wet the interface of the dispersed domain [31, 42, 43].

In-situ compatibilization is a process in which a fine dispersion of a minor phase

can be generated quickly by reactive extrusion. This process involves using liquid

reactants, such as functional monomers, to locate at the interface between immiscible

polymers and subsequently polymerize. The purpose is to graft the monomers onto both

polymers in a binary blend and create what is believed to be a network-like bridge

between the phases. The ultimate product has enhanced mechanical properties, better

processability, and a unique morphology.

This dissertation focuses on the in-situ compatibilization of two phases the

isotactic polypropylene matrix and an ethylene-1-octene copolymer (otherwise known as

linear low density polyethylene) minor phase. PP is known to behave in a brittle manner

at or below room temperature and in a notched state, thus limiting its use in blow molded









bottles, for example. The elastomeric copolymer, which has a glass transition

temperature well below room temperature, will act as the impact modifier for

polypropylene and contains a fully saturated backbone which limits environmental

degradation. But addition of elastomer alone compromises both stress-strain

performance and processability of the blend. The in-situ compatibilization technique

established by previous researchers has been applied to this system, and drastic

improvements in all macro-scale properties have been achieved.

In order to understand this dissertation, one must be versed in processing of

polymers, polymer blends and toughening of polymers, solid state and melt free radical

grafting of polymers, and free radical polymerization. This is a very complex process,

not only because of the number of components and variables, but because of the

dependency on both the rate of reaction and rate of morphology development. The

following six chapters describe various aspects of this unique polymeric alloy.

1.2 Chapter Summaries

Chapter 2 is a background/review of much of the research that has explored the

complex issues involved in reactive extrusion. The description includes the polymers of

interest (i.e., isotactic polypropylene and ethylene-1-octene copolymers), the type of

processing equipment involved, how the morphology of polymer blends develop in an

extruder, toughening brittle semicrystalline thermoplastics, many characteristics of free

radical polymerization and free radical induced grafting of polyolefins, and how all of

these aspects tie in to create novel in-situ compatibilized polymers.

A fundamental study of the physical blends of a certain grade of copolymer with

polypropylene is undertaken in Chapter 3. Little research has been conducted on using

low molecular weight ethylene-octene copolymers to toughen polypropylene, but this









chapter proves that these new copolymers are effective PP modifiers up to a certain

concentration. Many concepts and characterization techniques are introduced and

explained in detail so as to aid in analysis of results gathered in following chapters.

A complete picture of the mechanisms of toughening in several alloyed systems is

described in Chapter 4. This includes defining the effect of elastomer content on

mechanical, chemical, theological, and morphological properties of alloys. A

comparison is made between physical blends and alloys containing various levels of

liquid reactants. The crystalline as well as amorphous states are described in detail, with

both contributing to the overall performance of the alloys.

In Chapter 5, the actual processing characteristics of the alloys are elaborated upon.

Extruder screw speed and barrel temperature are directly tied into alloy performance.

The effects of varying reactant concentrations are also systematically studied.

Chapter 6 delves deeper into the effect of elastomer molar mass and crystallinity on

the behavior of both the physical blends and alloys. It explains many aspects of alloy

behavior that would not have otherwise been possible.

For Chapter 7, a design of experiments has been conducted to see the effect of this

in-situ compatibilization technique on high density polyethylene's impact strength and

stress strain behavior. High density polyethylene is a high volume commodity plastic and

modification to a high impact polymer is the goal. This chapter is followed by a

concluding chapter with future research possibilities along with appendices and

references.














CHAPTER 2
BACKGROUND AND LITERATURE REVIEW

2.1 Polymers of Interest

2.1.1 Isotactic Polypropylene

Isotactic polypropylene (PP) combines low price with attractive performance; e.g.,

heat distortion temperature > 1000C, strength, stiffness, corrosion resistance, and

versatility in applications, ranging from automotive moldings to films and textile fibers

[44-47]. PP is a semi-crystalline thermoplastic so its properties are strongly dependent on

molecular weight and defect distributions, which in turn affect both rheology and

crystallinity. Most end-use properties of PP homopolymers, such as stiffness, hardness,

and high temperature mechanical properties, are positively influenced by their overall

crystallinity, whereas impact strength and elongation are negatively influenced.

A major factor in the profitability of PP is the availability of low-cost propylene

monomer. There are two main sources of the monomer: co-production with ethylene or

separation from gasoline cracker steams in a petroleum refinery. When polymerizing PP,

three requirements are always present: its chain must be linear (monomer always adds to

the chain end), regiospecific (monomer is always added in head-to-tail manner), and

stereospecific (monomer always adds in the same stereo arrangement, or same side of

chain) [46]. In 1954 Giulio Natta polymerized propylene by means of a modified Ziegler

catalyst and obtained a blend of isotactic and atactic polypropylene [456b, 45c]. For his

pioneering invention, he and Karl Zeigler received the Nobel Prize for chemistry in 1963.

The present day polymerization process medium can be either liquid or gaseous









propylene or an inert hydrocarbon such as hexane. The process can be either bulk or gas

phase or a combination of both. Most polyolefin manufacturing processes presently

utilize heterogeneous Ziegler-Natta catalysts. Because these catalysts have more than

one type of active site, they produce PP with a broad molecular weight distribution

(MWD) and non-uniform stereoregularity.

Isotactic polypropylene is such a versatile plastic that its applications are

innumerable [45]. Some markets include fibers, carpet and upholstery, films, medical

devices, automotive (under the hood, exterior, interior), containers, construction,

nonwoven fabrics, appliances, and transportation. The resistance of PP to chemicals is

well documented and one of the principal reasons automobile batteries are made of PP.

Having a HDT above 1000C allows use with hot aqueous liquids, including steam-

sterilized medical goods. The processability of PP also makes it attractive. These

methods include extrusion, oriented and melt blown fibers, biaxially oriented film, blown

film, sheet, thermoforming, profiles and pipe, wire and cable coating, injection molding,

extrusion and injection blow molding, and compression molding.

2.1.2 Ethylene-1-Octene Copolymer

Dupont Dow Elastomers produce and license novel ethylene-1-octene copolymers

with the tradename ENGAGE via INSITE technology, which allows extraordinary

control over polymer structure, properties, and rheology [48-52]. They use a relatively

new single-site metallocene catalyst to polymerize a wide variety of bulky monomers,

including linear a-olefins. In contrast to Zeigler Natta (Z-N) catalysts, they yield

polymers which incorporate higher levels of the a-olefin to achieve lower polymer

density or crystallinity, and a uniform comonomer distribution with a polydispersity of

about two [45a, 53, 54]. The microstructural uniformity from metallocene catalysts









allows greater dimensional stability, higher impact resistance, greater toughness at low

temperatures, and higher resistance to environmental stress cracking [49]. The ability to

incorporate higher levels of comonomer has allowed densities of the copolymers to reach

0.87 g/cm3, previously unattainable by Z-N catalysts. The development of new

metallocene catalyst generations has bridged the gap between rubber and thermoplastic

technology [44].

These materials exhibit an enormous span of theological, mechanical, and thermal

properties [48, 55, 56]. They have excellent low temperature properties, clarity and crack

resistance. Their superior UV, ozone and weather resistance are primary advantages over

other impact modifiers such as EPDM (ethylene-propylene-diene monomer), EPR

(ethylene-propylene rubber), and SBS (styrene-butadiene-styrene).

In a comparison of elastomers, an EOC had a melting temperature 100C higher than

EPDM of similar crystallinity and molecular weight [57]. A more homogeneous

distribution of crystal morphology is apparent for the copolymers, with the more defect-

ridden EPDM providing less mechanical integrity. The maximum strength and

extensibility of the ethylene-octene copolymer are greater than EPDM even though the

EOC is lower in molecular weight.

An extensive study of several ethylene-a-olefin copolymers was conducted by

Bensason et al. [52] who classified these novel materials into four types: Type 1

copolymers are those with densities less than 0.89 g/cm3 and show a low degree of

crystallinity, low melting temperature, and the absence of cooling rate effects. Spherulites

are nonexistent and the granular, nonlamellar morphology suggest that the crystalline

regions should be described as fringed micelles. Type 2 copolymers range in density









from 0.91 0.9 g/cm3 and form poorly developed, unbanded spherulites containing both

bundled and lamellar crystals. Type 3 materials (0.93 0.91 g/cm3) form smaller

spherulites with thinner lamellae than HDPE homopolymer. Although the branches

restrict crystallization to an extent, the ethylene sequences are long enough to crystallize

in the lamellae. The fourth type has a density of 0.93 g/cm3 or greater and exhibits

lamellar morphology with well-developed spherulites. Lamellar thickness is strongly

related to the kinetics of crystallization because of the lack of long chain branching.

Over the last decade, much effort has been put forth to understand the crystalline

morphology and crystallization processes of these copolymers. A common consensus is

that as the concentration or length of comonomer increases, crystallinity decreases [28,

48-52, 58-64]. The introduction of more comonomeric units hinders the chain regularity

necessary for crystallization to take place [56, 57]. There is a distortion of the crystalline

lattices with an increase of 1-octene content but even at very low density (0.882 g/cm3),

certain amounts of lamellar crystal is still present [60]. The melting enthalpy is reduced

with increasing 1-octene content in the copolymer. Melting temperature was shown by

DSC to be inversely proportional to comonomer content [51]. Reorganization of polymer

chains occurs at room temperature for copolymers having a comonomer content higher

than 2.1 mol% of 1-octene [61] but this seems less likely for very high comonomer

contents because during annealing the branches would have to be drawn through crystals

[65].

2.2 Processing

2.2.1 Reactive Twin Screw Extrusion

Polymer processing in a twin screw extruder has been developed since the 1930s

and 40s, with several varieties offering numerous advantages over the other [2-4, 31, 66].









Twin screw extrusion (TSE) has been shown to be a versatile, cost effective method to

produce a uniform, optimized polymer based product. An ideal compounder will have a

uniform shear and elongational stress field, flexible control over uniform temperature,

pressure and residence time, compatibility for homogenization of liquids with large

differences in theological properties, efficient homogenization before degradation, and

flexibility for change in mixing parameters. TSE's are useful because of the ease of

feeding materials, excellent dispersive and distributive mixing, temperature control,

control over residence time distribution, reaction under pressure, continuous processing,

unreacted monomer and byproduct removal, post-reaction modification, and viscous melt

discharge. Most of the mixing is achieved with kneading paddles.

The main geometrical features that distinguish twin screw extruders are the sense of

rotation and the degree of intermeshing. Twin screw extruders whose screws rotate in the

same direction are co-rotating. The intermeshing twin screw extruder is self-wiping in

nature and helps to minimize the very long residence time tail frequently found with

extruders. They give a relatively uniform shear rate distribution and because the feed rate

is independent of screw speed, high screw speeds are possible (500 rpm) with

correspondingly high throughput rates. With this high speed, small sized equipment can

achieve high melting and mixing capacities. Two drawbacks are the cost to purchase and

maintain and a metered feeding device is needed in starve feeding mode. APV- Baker

Perkins is the manufacturer of our extruder, with some unique features being a clam shell

barrel, greater free volume, and barrel valves.

The first developments in the use of extruders as reactors were made about 60 years

ago and melt phase modification of polymers has been done for over 35 years. Reactive









extrusion (REX) refers to the deliberate use of chemical reactions during continuous

extrusion of polymers and/or polymerizable monomers [4, 31, 67-69]. Reactions have

been performed on molten polymers, on liquefied monomers, or on polymers dissolved,

suspended in, or plasticized by solvent. The types of chemical reactions that have been

performed by reactive extrusion include bulk polymerization, graft reaction, interchain

copolymer formation, coupling/crosslinking reactions, controlled degradation, and

functionalization/functional group modification. The attainment of proper mixing is

undoubtedly the single most important consideration when specifying or designing an

extruder-reactor. Chemical reaction is a molecular event, so proper mixing in reactive

extrusion means mixing at the molecular level and maximizing the interface between

dispersed phases and the matrix [4, 31, 70-72].

Polyolefins have proven to be preferred substrates for reactive extrusion

experiments largely due to their ready availability, low cost, and commercial

applications. The advantages of synthesizing graft copolymers by reactive extrusion as

opposed to alternating technologies include little or no use of solvents, simple product

isolation, short reaction times, continuous process, and relatively low infrastructure costs

[68]. The ability of an extruder to handle materials having high viscosities without any

solvents results in a dramatic raw material cost reduction, no solvent recovery equipment,

an ready-to-use products [73]. Some potential disadvantages or difficulties are the need

to achieve intimate mixing of reactants and substrates, high reaction temperatures

necessary to form a polymer melt, and polymer degradation or crosslinking.

2.2.2 Morphology Development in an Extruder Dispersive and Dissipative Mixing

The performance of extruded materials is determined, amongst others, by the final

morphology and dispersion [74, 75]. For blends and alloys, the morphology depends on









the composition, theological and physical characteristics of the components, relative

compatibility, and the nature and intensity of the mixing. When purely compounding two

plastics, they go through dispersive as well as distributive mixing stages [3, 31, 32, 76,

77]. Dispersive mixing is the breaking up of clumps or aggregates of solids into the

ultimate particulate size, or of immiscible polymers into the desired domain size. It is

dependent upon shear and elongational stress and is achieved by shearing the particulate

matter under high stress usually by kneading disks. In distributive mixing, spatial

uniformity of all components throughout the mixture is desired. This is best achieved by

frequent reorientation of flow elements under strain, including dividing, stretching,

distorting, and/or reorienting the flow. Mixing performance is known to decrease with

increasing viscosity [72].

The most significant evolution of morphology occurs in the initial melting zone of

the extruder. During the initial stages of blending, the elastic behavior is most important

but the viscous and interfacial behavior of the components in a system is undoubtedly

important in the later stages of mixing. The maximum shear stress, accompanied with

frictional and extensional forces, is usually generated at the melting zone of the extruder

and imparts a high degree of mixing [13, 31, 32, 60, 78, 79]. The melting mechanism

arises from the dissipation of the energy created by interparticle friction, rather than by

friction against the barrel wall or by heat transfer through the barrel wall [3, 75]. The rate

of melting controls the rate of reaction and morphology development [31]. Each

polymeric component changes into very small particles as droplets within a very short

time and distance (0.1-10 seconds and a few millimeters). A thermoplastic is dispersed in









the rubber phase with the plastic pellet size reduced from 3 mm to 5-20 am, then to

approx 1 [tm with eventual coalescence [31, 78, 79].

2.2.3 Melting and Droplet Breakup Mechanisms

As the solid begins to melt, the feed mixture may go through its most viscous stage,

that of a highly filled slurry or paste of unmelted solids in just-melted resin [3, 32, 75, 80-

82]. Often the minor phase softens first and will coat particles of the major phase, which

will delay its melting [13]. Transformation of a solid pellet involves three steps:

melting/plastification of the pellet, deformation/stretching of the molten polymer, and

formation of fine particles which may be subject to coalescence [31].

Several theories have been formulated which describe the process of melting and

morphology development [32, 75, 78, 81, 83]. The first to model this behavior was Shih

[78, 79] with the phase inversion mechanism. He found that polymer blends go through a

number of sequential physical changes before being combined into a cohesive mixture to

minimize free energy. For semicrystalline polymer/rubber mixtures, plastic pieces are

initially torn from the pellet surfaces and form a mixture with drawn out layers of rubber.

A lower melting or softening polymer dispersed into a higher melting polymer of major

phase volume follows a 4-stage inversion mechanism as shown in Figure 2-1: A. The

rubber forms a continuous phase closely packed pellets are suspended in it; B. Plastic

pellets break up layer by layer as the pellet surfaces begin to soften, shear, and pull off

from the unmelted solid core and are dispersed in the rubber phase; C. In the region of

maximum torque (0.7 am particles), an abrupt phase inversion occurs due to coalescence










Stlg. Continuous DWfpord
_Ph"* Phaf

A RUBBER PLASTIC PELLETS
(UnmeUetd)



SOFTENED PLASTIC
BI RUBBER PARTICLES &
UNMELTED PELLETS



PARTICLES


c PHASE INVER;ION


D MOLTEN RUBBER
PLASTIC PARTICLES

Figure 2-1: The phase inversion mechanism proposed by Shih.

of molten plastic particles and finely divided rubber droplets are formed in a continuous

plastic matrix; D. The final stage is a viscoelastic fluid matrix with finely divided rubber

droplets suspended in it followed by a continued decrease in torque.

Although the onset of phase inversion is abrupt, the completion of the phase

inversion during phase C is not instantaneous. A small amount of the high melting

polymer remains trapped in the rubber phase at the end of the mixing cycle. During the

phase inversion, the mixture morphology changes from a continuous rubber phase with a

very high concentration of high melting particles (80%) to a continuous molten plastic

phase with a smaller amount of dispersed rubber particles (20%). The overall viscosity is

expected to drop significantly, simply from the change in dispersed phase concentration.

Sundararaj [32] and Scott and Macosko [83] proposed a droplet breakup and

coalescence theory for morphology development (Figure 2-2). An initial mechanism of

droplet breakup involves the formation of sheets or ribbons of the dispersed phase in the









matrix, which are drawn out of a large mass of the dispersed phase. The pellet breakup is

primarily controlled by the rate of deformation and subsequent relaxation of the pellet

phase. As the relaxation time of the pellet decreases (more elastic behavior), it becomes

more difficult to create a sheet. Owing to the effects of flow and interfacial tension, these

sheets are unstable and holes begin to form in them. A high stress level followed by a

lower stress level is required to achieve efficient mixing [84]. In the high stress level, the

dispersed phase is stretched and extended into shapes, which undergo instabilities and

break up upon entering the low stress level.








@l o












2pW-10i0m
Figure 2-2: Morphology development of binary polymer blends proposed by Scott and
Macosko.

As the sheet grows, the holes are filled with the matrix phase, which surrounds the

sheet on either side. When the holes in the sheet or ribbon attain a sufficient size and

concentration, a fragile lace structure is formed, which begins to break apart into

irregularly shaped pieces of a wide distribution in size. These pieces are approximately









the diameter of the particles generated in the blend at long mixing times. The irregular

pieces continue to break down until all of the particles become nearly spherical.

This proposed mechanism results in the generation of very small particles at very

short timescales. As a piece of dispersed phase undergoes this deformation mechanism,

many very small particles may be generated very quickly. As the mixing time proceeds, a

greater proportion of dispersed phase is cycled through this mechanism.

2.2.4 Morphology Development Dependency on Rate of Reaction

Typically, small minor phase drops less than 1 micron are desired. If the dispersed

phase is dilute, this is relatively easy but at dispersed phase concentrations greater than

1%, collisions between drops occur and domain size increases due to coalescence [13, 32,

42, 78, 79]. Also, the particle size distribution broadens at higher concentrations.

Coalescence of the dispersed domains has been shown to be dramatically reduced (up to

30%) by bonding at the interface [32, 42, 85, 86]. Reaction increases the effective

interfacial tension and both delays and intensifies the phase inversion process. The

morphology development of the reactive system parallels its non-reactive counterpart, but

the final number average particle size is two orders of magnitude smaller. Steric

stabilization of the dispersed phase is more important than interfacial tension decrease

[32, 87-89]. When the blend is reactive, very high concentrations of the major phase can

exist as the dispersed phase for longer periods of time before phase inversion occurs. If

the particles are not monodisperse and are not perfect spheres, then much higher

concentrations (>74%) of the dispersed phase are possible [32].

One can imagine that when the major phase envelops the minor phase, small

particles of the major phase generated during melting will be trapped inside the minor









phase. If the system is reactive, then the occluded domains will be stabilized and will not

be able to coalesce with the matrix [32].

During reactive extrusion, a time scale will exist for both the morphology

development and reaction. Polymeric alloys all have one thing in common the rate of

melting dictates when the reaction will begin [13, 31, 32, 42]. For these systems, reaction

occurs almost immediately after the melting of the polymers. There is a fine balance

between the time for morphology generation and reaction time, so the ratio must be

manipulated to obtain the best processing conditions and optimum morphologies. When

compounding all components together, the benefits of co-melting are realized by good

dispersion of the materials and a greater probability of locating reactive ingredients at the

interfaces to facilitate grafting/bonding [27, 90]. If liquid reactants are added

downstream after melting has occurred, dispersion is difficult due to the highly viscous

molten polymer which will lead to polymerization rather than grafting onto the polymer.

The flow in an extruder promotes the reaction in two ways: First, it either breaks or

deforms the suspended droplets and increases the interfacial area available for the

reaction. Second, the flow convectively increases the mass transport to supply fresh

reactants to the reaction [71]. The improved reaction rate by back and forth flow can be

attributed to more efficient production of new interfacial area [72].

Many studies have shown that the majority of reaction occurs in the melting zone.

For monomer grafting onto polypropylene, the conversion can be up to 70% of the final

value upon melting [91]. The level of styrene grafting onto a polyolefin can be 85% after

the materials pass through a kneading block zone [92]. Another study has shown that

free radical grafting of a monomer onto PE or PP has already gone to completion just









after the kneading block [90]. Machado et al. found that regardless of the type of

polyolefin, the majority of grafting reaction occurred in the melting zone of the extruder

[93]. They also found that higher grafting yields exist for lower viscosity polymers,

possibly due to the melting phenomenon. Grafting may actually occur below the melting

temperature of the polymer because of significant levels of peroxide decomposition [94,

95].

Diffusion of reactants is not the most important issue for an alloyed system because

if the reaction rate is fast, the size of the dispersed phase will be very small [42]. A good

compatibilizing chemistry has to be fast enough compared with the rate of interfacial area

generation so that once the interface is created, it is stabilized quickly by a layer of

copolymer so as to minimize coalescence [31]. The shearing or extensional flow

experienced by the high molecular weight polymers creates a very large interfacial area,

thus reducing the need for long range diffusion and facilitating reaction at the interface.

2.2.5 Viscosity and Reaction Effects on Morphology Development

When two immiscible polymers are blended, one phase is mechanically dispersed

inside the other. The size and shape of the dispersed phase depend on several processing

parameters including rheology, interfacial properties, and composition [31, 32, 96].

Viscosity ratio and surface tension between the major and minor phases play

important roles in determining droplet size [31, 39, 69, 75, 82]. With the viscoelastic

fluids generally encountered in polymer processing, the lower the elastic nature of

dispersed phase, the lower the matrix stresses required to break up and stabilize them.

Taylor was the first to theoretically describe droplet breakup and formation while

suspended in another liquid medium [97, 98]. When the rate of distortion of the fluid or

the radius of the drop is great enough, the drops tend to break up. For a very small









viscosity ratio (TrDrop/TlMatrix), the drop remains coherent in spite of the fact that it gets very

long and narrow. The act of bursting is always an elongation to a threadlike form

followed by degeneration into drops which are of the order of 1/100th of the size of the

original drop. A low viscosity minor phase will break up into small droplets early in the

extrusion process [32, 76, 96] but beyond a minimum viscosity coalescence is favored

[39].

2.3 Toughening of Polymers

2.3.1 Origins of Polymer Toughening

Polymer toughness, or the property of resisting fracture by absorbing and

dissipating energy, is a highly sought after characteristic of a material or product. It

depends on many parameters including temperature, pressure, deformation rate, shape of

specimen, and type of load, aside from material properties like molecular weight,

polydispersity, chain packing, chain entanglements, crystallinity, heterogeneity, etc.

Brittle fracture occurs at high strain rates, low temperatures, and in thick sections because

each restricts the extent of the yield zone. Plastic deformation itself is a complex

phenomenon and involves both crystalline and amorphous phases [99]. Energy is

absorbed within the sample by viscoelastic deformation of the polymer chains, and

finally by the creation of new surface areas [100].

Fracture resistance in rubber toughened polymers is generally attributed to three

major mechanisms that absorb or dissipate energy as cracks advance through polymers

chains: rubber cavitation [29, 99, 101-105], matrix crazing [28, 106, 107], and/or shear

yielding [5, 8, 12, 101, 103, 105, 106, 108, 109] with chain breakage accompanying the

failure of polymers. Impact resistance has also been correlated with the presence of a

secondary transition at least 500C below the testing temperature [100, 105, 107]. The









speed of the impact test effectively raises the temperature of this secondary transition by

500C (time/temperature equivalent); therefore, secondary transitions occurring near the

test temperature at low frequency are effectively shifted into the glassy region at testing

frequency. Below its Tg, a rubber particle will not cavitate and will therefore behave in a

brittle manner. Secondary transitions must be associated with motion of the polymer

backbone (glass transition temperature), not pendant side chain groups, for them to be

effective at improving impact resistance.

When a toughened polymer blend is subjected to a uniaxial stress, the localized

stress experienced by the matrix material in the vicinity of a rubber particle will be

magnified by the local stress concentration factor. The initial cavitation of a rubber

particle relieves the triaxial stress existing at a crack tip and enhances localized yielding

in the matrix, thus avoiding a brittle catastrophic failure of this material. If the applied

stress is increased further, the crack tip may be bridged by the stretching rubber particle,

provided there is sufficient adhesion between the rubber and the matrix and that the

particle can stretch sufficiently rapidly in terms of the speed of the crack advance. For

this reason, it is desirable that the rubber should have as low a Tg as possible, while

allowing the rubber to fibrillate and maintain a degree of structural integrity in response

to impact loading [12]. Conventional wisdom states that for toughening polymers, the

rubber droplets must be at least as large as the cracks they are trying to stop, putting the

minimum size at several hundred angstroms to 300-500 nm [100].

2.3.2 Elastomer/Rubber Toughened Blends

A fundamental understanding of how rubber particles affect the creep response of

polymers is described by the Erying theory, which states that energy barriers at the

molecular level control the macroscopic rates of flow [110]. This is a fundamental









property which depends on rubber composition, volume fraction of dispersed phase,

rubber particle size and distribution, and rubber-matrix interfacial interactions [2, 3, 8, 9,

22, 28, 69, 107, 111-115].

Toughness is the greatest at an optimum rubber particle size for polymers that

dissipate fracture energy mainly by matrix crazing (PS and PMMA) [70, 116]. But for

semicrystalline polymers, close interparticle distance is just as important as small particle

size and high interfacial adhesion [5, 21, 69, 113, 116-118]. A critical interparticle

distance (matrix ligament) exists below which a material behaves in a ductile manner and

above which behaves in a brittle manner [69, 103, 113, 119-124]. Ligament thickness is

an inherent property of the polymer and is independent of rubber volume fraction or

particle size. At large separation distances, the stress field in the matrix is simply a

superposition of those around isolated particles, and the polymer blend will remain

brittle. However, when the particle surfaces are sufficiently close, the stress field is no

longer simply additive, and the fields around the particles will interact. This will result in

enhanced matrix yielding, and a transition to tough behavior [119]. Even if rubber

particles are chemically bound to the matrix, a polymer blend will still be brittle if the

interparticle distance or particle size is greater than a critical value.

Bartczak [125] and Muratoglu et al. [126] expanded upon the matrix ligament

theory and revealed that a layer of anisotropic crystalline material having a lower plastic

resistance than the bulk surrounds the dispersed particles. Van Dommelen et al. have

actually modeled this behavior and show that this unique layer is a highly efficient

method for toughening semicrystalline polymers by altering matrix craze formation,

reducing principal stresses, and inducing extensive matrix shearing [115].









In general, the critical particle size for toughening decreases with increasing

ductility of the matrix polymer. For a Nylon 6,6-rubber blend, it was found that if rubber

particles are large, a greater amount of rubber is needed to achieve toughening and vice

versa [119]. Chou et al. showed that rubber particles narrowly distributed around 0.5 to 1

micron in size are effective for toughening PP [103]. A large number of small rubber

particles are preferred when toughening PP because they are more efficient at promoting

shear yielding and crazing throughout the matrix [22, 69, 127]. Also, a bimodal

distribution of rubber gives a good balance of toughness and stiffness in PP [23].

If rubber is well bonded to the matrix, stresses can become redistributed between

phases subsequent to yielding, so that the plastic zone is just as capable of resisting crack

extension as the homopolymer. The effect of poor matrix-rubber bonding is a weaker

plastic zone, thus canceling out the benefits to be derived from a reduction in yield stress.

Good rubber-matrix adhesion and small particle size are necessary criterion for screening

rubbers to toughen PP and increase impact strength [19, 22, 110, 128]. Blends with very

small particle size have a relatively high cavitation stress, which results in a high yield

stress of the blend [104].

A final contribution to toughening semicrystalline polymers is from stress-induced

phase transformations [129]. If changes in crystallinity take place during deformation,

energy will be absorbed by melting or released by crystallization. A comparison can be

made of the softening effects due to a rise in temperature to those associated with an

isothermal reduction in crystallinity. Adiabatic heating of polymers fractured at high

deformation rates is common [21, 130, 131] and influences the melting/recrystallization

process which in turn affects plastic deformation [21, 129-135].









In summary, for optimum impact strength improvement in PP/elastomer blends, the

following conditions must be satisfied [136]:

1. Elastomer particles are finely and uniformly distributed in the PP matrix,

2. The modulus of the elastomer is much less than that of the PP,

3. The crystallinity of the elastomer is low,

4. A certain degree of interfacial adhesion is present between the elastomer particles
and the PP matrix,

5. The cohesive strength of the elastomer is large,

6. A certain degree of entanglement of high MW polymer chains is present in the PP
matrix.

2.4 Free Radical Reactions

The free radical grafting process is very complex and not completely understood.

The following table lists the most likely reactions to be encountered while grafting

monomers onto polyolefin substrates.

Table 2-1: The mechanisms of main reactions in the grafting process.
Initiator Decomposition R'OOR' 2R'O
Hydrogen Abstraction R'O + P R'OH + P
P-Chain Scission (PP phase) P- P1 + P2'
Crosslinking (PE phase) P. + P- P-P
Graft Initiation P- + M PM-
Graft Propagation PMn- + M PMn+1"
Homopolymerization R'O- + M R'OH + Mn'
Termination by Recoupling PMn+1r + Mn- PMm


2.4.1 Initiator Decomposition

Previous research has shown that a peroxide initiator is necessary to create enough

free radicals to graft monomers onto polyolefin substrates which may be due to the

presence of stabilizers in the polymer resin [4, 7, 90, 94, 137-143]. Initiator









decomposition is known to be the rate limiting step for hydrogen abstraction [92].

Dialkyl peroxides are the initiators of choice because they are amongst the most stable of

all commercially available organic peroxides and the free radicals generated from

decomposition have a variety of uses [144]. 2,5-dimethyl-2,5-di-(t-butylperoxy) hexane

(tradename Lupersol 101) is often used to degrade polypropylene, crosslink various types

of polyethylenes, and graft monomers onto these polyolefins because of its efficiency

[73, 92, 94, 147, 148]. Figure 2-3 gives the chemical structure while Table 2-2 lists

characteristics of the peroxide.

CH3 CH3

(CH3)3C-00-C-CH2-CH2-C-OO-C(CH3)3

CH3 CH3

Figure 2-3: Chemical structure of Lupersol 101

Table 2-2: Characteristics of Lupersol 101 [145, 146].
Molar Mass (g/mol) 290.44
Peroxide content (%) 91-93
Oxygen content (%) 10.03-10.25
Physical form Liquid
Melting point (OC) 8
Boiling Point (C) 249
Specific gravity (cm3 at OC) 0.865 @ 25C
Viscosity (mPa.s) @ 200C 6.52
Typical Decomposition products in inert media: methane, ethane, ethylene, acetone, t-
butyl alcohol. Lupersol 101 is considered a suitable food additive.

The half life of a peroxide is very important, defined as the time it takes for one

half of a given quantity of peroxide in dilute solution to decompose at a given

temperature. The melt free radical grafting rate is dictated by peroxide efficiency, or half

life [138, 146]. One has to keep in mind that the half life of Lupersol 101 in molten

LDPE is reported to be 2-3 times longer than in organic solvents [4, 90, 94]. A similar









result is found when peroxide decomposition is in the solid (glassy) state of a polymer

[149]. The low reactivity of polymers compared to model compounds is attributed to

several factors, including the lower concentration of tertiary C-H reaction sites, coiled

conformations, and high viscosity [150, 138].

Conventional wisdom shows that half life time should be about 5 times that of the

residence time of the polymers [68, 92]. If the half life is too long, the initiator may not

be completely utilized. A short half life may increase the crosslinking of radical-radical

combination or grafting yield may be limited by the rate of monomer diffusion to the site

of reaction, especially a heterogeneous melt. Other work has shown that no grafting

occurs after consumption of the initiator (as estimated by half life) [151]. Half life is

calculated as Ti/2 = 0.693/kd, where kd = Ae-ERT, Ea is the activation energy, R is the

universal gas constant, T = temperature (K), and A is an integration constant [146].

Table 2-3 is a manufacturer's generated list of estimated half lives for Lupersol 101 at

various temperatures.

Table 2-3: Half life of Lupersol 101 based on temperature
Temp (C) 165 190 210 230
Half life (seconds) 280 (4.7 minutes) 28 5.4 1.2

The mechanism of free radical production from Lupersol 101 involves a primary

and secondary reaction, with decomposition of each group independent of eachother [68].

Upon addition of heat, this peroxide decomposes homolytically into 4 alkoxy radicals at a

bond dissociation energy of about 36 kcal/mol [145, 152, 153]:

CH3 CH3
(CH33C-0 0O-C-C H2-CH2-C-O -O-C(CH33
CH3 CH3









A secondary reaction can occur in which the tertiary alkoxy radicals can undergo further

fragmentation (e.g. P-scission) to form ketones and alkyl radicals:

CH3 0
CH3-C- O* CH3-C-CH3 + CH3*
P-chain scission
CH3
t-butoxy radical acetone + methyl radical

The extent of reaction at a given position is proportional to the amount of peroxide

initiator that has decomposed at that position. In regions of high temperature, the

peroxide reacts more rapidly, inducing a concentration gradient that drives additional

peroxide to diffuse into the hot region and react. The faster the initiator is consumed, the

less time it has to diffuse to other portions of the channel and react [154].

2.4.2 What Happens After Peroxide Decomposition?

A radical produced by either primary or secondary reactions can abstract a

hydrogen (from the polymer backbone) or add to a double bond (vinyl monomer). The

secondary reaction is strongly temperature dependent, more so than abstraction or

addition [68, 155]. So with an increase in temperature, there is an increase in the number

of methyl radicals, which attack bonds in a much more random nature than t-butoxy

radicals and usually add to vinyl monomers rather than abstract hydrogen atoms from the

polymer [156]. Methyl radicals add to styrene some three orders of magnitude faster than

the model compound 2,2,4 trimethylpentane [68].

2.4.3 Polymerization

The feasibility of radical polymerization of a monomer depends primarily on the

polarity and size of the substituents on the double bond and the tendency to chain transfer

[153, 157-160]. The reactivity of the monomer is influenced by two factors:

1. The stability of the monomer toward addition of a free radical









2. The stability of the monomer radical thus formed (more important).

Monomers with less resonance stabilization represent much higher energy states.

The order of reactivity of the radicals is the reverse of that for monomers; i.e., styrene

monomer is more likely than methyl acrylate to consume a radical but addition of styrene

monomer to a polymerizing chain is 100 times slower than methyl acrylate [153, 161]. A

larger propagation rate constant will lead to higher molecular weight. Resonance

stabilization depresses the activity of the radical. Low activation energies are indicative

of a greater decrease in energy from reactants to products [157, 162, 163]. Table 2-4

gives typical activation energies for several aspects of free radical polymerization in

solution [146].

Propagation is bimolecular and its rate is independent of chain length. The initial

chain formed rapidly produces a high molecular weight polymer [146]. A monomer's

ceiling temperature is defined as the temperature above which monomer cannot be

converted into long chain polymer. A relatively low ceiling temperature is suggested for

grafting single monomer units onto polymers or to limit homopolymerization. For

styrene and acrylate-based monomers whose ceiling temperatures are well above

polypropylene extrusion processing temperatures, depropagation or "unzipping" of the

chain is not a concern.

Table 2-4: Typical free radical kinetic values in solution.
Type of Reaction Activation Energy (kcal/mol)
Initiator Decomposition 30-50
Initiation 5-7
Propagation 4-10
Chain Transfer 10-20
Termination 0-6
Less stable 7r-bond more stable G-bond: -12 to -23.9kcal/mol: exothermic









During free radical polymerization [164, 165] and graft copolymerization [166,

167] bimolecular termination is still energetically favorable but the macroradicals are

large with slow diffusion, so termination will be almost exclusively diffusion controlled.

The rate constant of termination for styrene is inversely proportional to the viscosity of

the medium over a thousand fold range of viscosity.

Chain transfer may occur, defined as the termination of one macroradical to

produce another macroradical which serves as a branch point. The new free radical

produced may or may not be initiate another polymer chain formation, depending on its

activity. Styrene is more to likely add monomer than undergo chain transfer to either

polypropylene or polystyrene [162a]

Styrene monomer (Table 2-6) is an aromatic hydrocarbon which, under normal

conditions, is a clear, colorless, flammable liquid. It is a versatile material, the

derivatives of which are styrene-based polymers. Styrene is one of the few vinyl

monomers that undergo rapid thermal polymerization [168, 169].

2.4.4 Copolymerization

When studying the copolymerization of two monomers, their reactivity ratios result

from a combination of steric, resonance, and polar effects [153, 159, 160].

* Steric effects: bulky substituents decrease reactivity in radical polymerizations.

* Resonance: If a radical can be stabilized by resonance, it is more likely to form
(monomer is more reactive). However, resonance stabilization of the radical also
makes the radical less reactive towards propagation.

* Polar effects: A monomer with an electron-withdrawing substituents is more likely
to react (cross-propagate) with a monomer having an electron-donating substituent
than it is to self-react. So, ethyl acrylate will want to add styrene monomer units
over other ethyl acrylate units.









Two monomers can either undergo self-polymerization or copolymerization. The

terms rl and r2 are reactivity ratios, and define the relative tendencies to self-propagate

or cross-propagate. Ifrl > 1, then monomer 1 tends to self-propagate, whereas ifrl <1,

copolymerization is preferred. In Q-e schemes, Q is a measure of monomer reactivity

(resonance stabilization), and e relates to monomer polarity. Q values increase with

increasing resonance stabilization, while e values become less negative as groups

attached to the double bond become more electron attracting. Table 2-5 gives Q and e

values [162c] and reactivity ratios [162b] for styrene and ethyl acrylate, while Table 2-6

gives basic information on the three monomers of interest.

Table 2-5: Copolymerization constants for styrene and ethyl acrylate monomers.
Monomer (r) Q value e value reactivity ratio
Styrene (r2) 1.00 -0.8 0.699
Ethyl acrylate (ri) 0.41 0.55 0.139

Because rl *r2 deviates from unity and since e is much different for the monomers,

alternating copolymerization will exist [161]. The reactivity ratio of styrene does not

change with temperature but ethyl acrylate can be strongly affected. The higher the

temperature, the more ethyl acrylate units will be incorporated into the copolymer [170].

The initiator has no significant effect on the reactivity ratio values. For two monomers

far apart on the polarity series, the rate of copolymerization is often much higher than for

either monomer alone [160].

Table 2-6: Structural and physical information about the monomers of interest.
Monomer Styrene Diethyleneglycol diacrylate Trimethylolpropane
(DEGDA) Triacrylate (TMPTA)
Structure H o o o
C=C II 11---CB=CB
CC- U


-Ib-C--CH=CH,










Table 2-6 Continued
Functiona 1 2 3
lity
Molar 104 214 296
Mass
(g/mol)
Viscosity 0.76 12 106
@ 25C
(cps)

2.4.5 Multifunctional Monomer

The random nature of free radical polymerization processes likely results in the

assemblage of units in an irregularly patterned structure. Nonlinear polymers are

obtained from monomers at least some of which posses a functionality exceeding two.

Gelation can be avoided in nonlinear polymerization by limiting the extent of reaction or

using small proportions of reactants [161].

In a study which compared homo- and co-polymerizing multifunctional monomers,

Diffusion of the monomer played a significant role in bond conversion. As the

composition of the crosslinker is increased, the maximum double bond conversion

decreases due to diffusional limitations to the polymerization process [163, 171]. The

acrylate copolymers are more homogeneous than methacrylate copolymers [163]. When

comparing di-functional vs. tri-functional monomers, the concentration of crosslinking

double bonds is 43% higher for tri- due to higher functionality per monomer [172]. For

neat monomers, the maximum polymerization rate for di-functional is twice that of tri-

functional.

A survey of multifunctional acrylates was conducted to assess the effect of

monomer rank (number of atoms between acrylate functionalities), size of functional

group, and number of functional groups [171]. As the rank increased, the conversion of









double bonds increased from 74.5% to 84.2%. This was explained by the increased

mobility in the double bonds in the latter part of the reaction. Activation energy and rate

of polymerization decreases as the number of alkyl groups increases [173]. The opposite

trend was observed for increasing bulkiness of pendant functional group and increasing

functionality of the monomers. The increase in the number of double bonds of the

monomer did not necessarily increase the possible crosslinking concentration. Acrylates

reacted faster and had higher enthalpy of polymerization than methacrylates. The

monomers with smaller pendant groups and lower functionality exhibited higher rate

constants during polymerization due to increased diffusivity of the monomer and pendant

double bonds in the reacting gel. The reactive pendant group could introduce hindrance,

reduce the reactivity of the double bonds and reduce the conversion.

2.4.6 Hydrogen Abstraction

The specificity of hydrogen abstraction depends on an array of steric, polar, and

stereoelectronic factors, including bond dissociation energies and kinetic effects [68, 140,

150, 152, 174-178]. The lower the bond dissociation energy, the more stable the radical

and therefore the more reactive. Alkyl radical stability increases in the order primary (1)

< secondary (2) < tertiary (3) < allyl z benzyl. Barriers for tertiary, secondary, and

primary radical formation are 10, 11.5, and 12.2 kcal/mol, respectively, which indicates

an order of magnitude difference in kinetics between tertiary and primary formation [155,

156]. For isotactic polypropylene at 1300C, the relative reactivity of CH:CH2:CH3

groups with benzoyl peroxide as initiator is 50:10:1 (reactivities per H atom) [68].

Another study showed that the PP macroradical actually forms a mixture of primary,

secondary, and tertiary free radicals (1:14:18) [179]. Abstraction from the tertiary C-H

position for a PP model accounts for two-thirds of the total product distribution [150].









In modeling polyethylene, secondary radicals were seen but there was no evidence

of signals from primary radicals [156]. As previously stated, methyl radical formation

begins to dominate at high temperatures. At 2000C, CH:CH3 hydrogen abstraction for

methyl radicals is 8:1 and CH2:CH3 is 3:1. At 2000C, CH:CH3 hydrogen abstraction for

t-butoxy radicals is 6:1, and CH2:CH3 is 5:1 [68, 180]. It is possible that, if secondary or

primary radicals are produced, they may be transformed into the more stable tertiary

radicals by subsequent inter- or intra-molecular abstraction reactions [68].

When comparing polymers, bulky side groups can also reduce the reactivity of the

substrate [151, 181-183]. For grafting branched polyethylenes, tertiary-hydrogen atoms

are three to four times more reactive than secondary hydrogen atoms [155]. The

introduction of a branch leads to the replacement of 4 secondary hydrogen atoms by one

tertiary and three relatively unreactive primary H atoms.

For an ethylene-octene copolymer with 10% octene, the site of grafting is 12:1 for

secondary:tertiary [150]. Therefore, a greater number of secondary groups results in

predominant grafting at these points offsetting the greater reactivity of the tertiary C-H

reaction sites. As the octene content of the copolymer changes to 5%, the ratio of

secondary to tertiary sites changes to 22:1.

2.4.7 Polyethylene Crosslinking Reactions

Polyethylene will crosslink in the presence of free radicals because of the long

lived macroradical present after hydrogen abstraction. Cross-termination of PE has a

high rate constant [155]. ENGAGE polyolefin elastomers have been found to form an

insoluble gel only above 0.3 wt% initiator [184]. It is now known that the highest

molecular weight fractions of polyethylene will be consumed first and a higher amount of

crosslinking agent shifts the distribution towards lower molecular weights [149, 185].









The low MW fraction is believed to act as grafted pendant chains in the presence of

longer polymer chains. The network formed within the low MW fraction consists mainly

of chemical crosslinks, whereas high MW material comprises both physical

entanglements and chemical crosslinks. Trapped entanglements generate the major part

of the crosslinking points at low peroxide concentrations, especially in HDPE. At high

number average molecular weight values, dense networks are easily created with only a

small mount of chemical crosslinks, as the probability of entanglement formation is very

high.

The presence of polyfunctional monomers greatly increases the efficiency of

polyethylene crosslinking [186]. The degree of crosslinking levels off rapidly at 3%

monomer, but monofunctional monomers such as styrene, methyl methacrylate, and vinyl

acetate were observed to have no effect on PE crosslinking.

Yang et al. desired to reduce crosslinking when trying to graft glycidyl

methacrylate onto polyethylene and used several inhibitors and monomers to do so

[187]. They found that the chain transfer agent/inhibitor p-benzoquinone gave an

acceptable grafting degree with minimal crosslinking. Styrene monomer, on the other

hand, gave an unusually high gel content along with grafting degree.

2.4.8 Degradation of Polypropylene

Chain scission is the most energetically and kinetically favorable process after

hydrogen abstraction from the 0 position on the PP backbone. Disproportionation,

another term for p-chain scission, results in a saturated product and an unsaturated

molecule, depicted in Figure 2-4 [137, 153, 188]. This degradation discolors the plastic,

reduces crystallinity, compromises mechanical strength, and lowers the viscosity and

melt strength [189, 190]. But some authors have shown that peroxide addition can









narrow the molecular weight distribution and eliminate the high molecular weight tail of

polypropylene to facilitate processing [73, 189, 191-194].

H C CH3 CH3I CH3 H CHH3 CH3I CHHl CH3
.C-C--C-- -C-C--~ ,l a-C ---C-C-C- -C-C-C"-
F Hi

+ (CH3)3C-OH
(CH3)3C-O*



H CH, H CH3H CHHs CH3 CH 9 H31 CH H
C -C. C=C- c -C-C-C' -c c-c-c -C
SH H H HH H H
Figure 2-4: Schematic of H abstraction from PP and subsequent P-chain scission.

The degree of chain scission is found to be proportional to the initiator level. Each

reaction occurs randomly, leading to random chain scission. Under typical extrusion

processing temperatures and low levels of peroxide, the backbone radical center is highly

unstable and undergoes P-chain scission almost immediately after H abstraction. The

activation energy for P-chain scission has been reported to be between 29.6-32 kcal/mol

but the relative rate is a function of temperature [137, 179]. Bimolecular termination,

which is highly diffusion controlled, may be negligible because of the low concentration

and short lifetime of the backbone radical [152, 154, 195]. The scission occurs randomly

along the chains and higher MW chains have a greater number of bonds, so longer chains

will experience P-scission preferentially [152, 154]. The number average molecular

weight is inversely proportional to the degree of chain scission, which in turn is linear

with respect to the amount of initiator [194].









2.5 Previous Efforts to Reduce/Prevent Degradation of Polypropylene

2.5.1 Free Radical Grafting of Polymers

Numerous processes of grafting monomers onto polymer backbones exist, with free

radical melt, solution, and solid state grafting common practice [7, 68, 90. 91, 138, 148,

155, 177, 179, 190, 196-199]. Gamma [200] and UV [201, 202] radiation as well as

ozonation to produce hydroperoxide functionalities on PP [203, 204] are also interesting

grafting methods. The goal of each process is to achieve a high grafting yield and a low

incidence of side reactions, which require that the radical sites on the backbone are

efficiently transformed to graft sites [68, 205].

This dissertation revolves around free radical grafting of polyolefins via peroxide

decomposition. All reactions are done primarily in the melt state rather than in solution

for several reasons. In the past, solvents would dissolve the polymer and high graft

yields ensued, but solvents had to be distilled for separation, and the grafted copolymer

had to be dried of solvent before use. Not to mention the high solvent loadings needed

which makes the process expensive and poses environmental issues [7, 206].

The idea of grafting monomers onto polyolefin backbones is not new and several

patents have been issued over past several decades. The original purpose was to impart

polarity, resistance to degradation, and improved adhesion [199, 207-211]. The

monomer(s) and initiator mixture can be either mixed with polymer before melting or

added to the melt above 1200C. Grafting is done continuously in a reactor that provides

intimate contact between the components (i.e. extrusion) with devolatization prior to

exiting the die. Typicaly, the reactive components are a peroxide or azo initiator (at 0.1-

0.7 wt%) to create the free radicals and a monomer (at 0.5-5 wt%) with a functionality









that can further react. A low molecular weight product with a sharp increase in flow rate

is common.

2.5.2 Solid State Grafting

There may be some solid state (surface) grafting for a reactive extrusion process

when all materials are added at once into the extruder [94, 95, 143, 212]. For grafting

below the surface, initiator decomposition rate may be of secondary importance to

monomer diffusion. But during extrusion, monomer and initiator may become trapped

into small pockets of molten polymer which are then broken as the melt is sheared.

Diffusion of liquid reactants into the amorphous region (not crystalline [213]) may occur.

The coalescent state of the monomer droplet facilitates an increased rate of

polymerization but also homopolymer formation [172, 178]. The grafting yield of PP as

a solid particle is much lower compared to in the melt even though the amount of P-chain

scission is three orders of magnitude less. For grafting at 1200C (solid state), melt flow

index is over 100 times greater than at that 2200C (molten state) [7, 180].

High melt strength polypropylene has been created by absorbing styrene and

butadiene and subsequently polymerizing to produce a branched alloy with enhanced

strain hardening, melt strength, and drawability [179]. This branched architecture can

also be created by irradiation of PP with a difunctional monomer [214]. Higher

functionality monomers were not as effective at crosslinking PP and caused the formation

of an undesirable gel. Shorter chain monomers are better than longer chain monomers

for improving melt strength and an acrylate-based functionality was better than

methacrylate at the same molecular weight because the reactivity of acrylate monomer is

higher than methacrylate [215].









2.5.3 Reactions in the Polymer Melt

In order to minimize side reactions, the radicals formed on the polyolefin backbone

must be trapped as soon as they are formed [68]. Some monomers are more effective

than others, which may be due to relative solubility of the monomers in polyolefin melt

or inherent reactivity of monomer. By increasing grafting yields, degradation and side

reactions such as homopolymerization will decrease. This strategy involves choosing a

monomer combination such that the primary monomer has a high reactivity towards free

radicals and can effectively trap radicals on the polyolefin backbone and the secondary

monomer facilitates this grafting process by creating branch sites or speed up the

polymerization process of the primary monomer. The higher grafting yields are

attributed to:

* Longer chain grafts rates of copolymerization of electron donor acceptor pairs are
greater than for homopolymerization.

* More grafting sites more efficient trapping of radical sites on polymer backbone.

A straightforward method to crosslink PP was to simply add an excess amount of

peroxide initiator to induce addition reactions of fragmented chains [216-219].

Generation of free radicals on PP leads to degradation because of the low stability of

macroradicals on the tertiary carbon. Crosslinking efficiency is low due to the

fragmentation of a large portion of macroradicals. Decrease of crosslinking efficiency

occurs with increasing temperature because the rate of fragmentation increases (a higher

energy process) compared to recombination. The rate of fragmentation depends primarily

on temperature while the rate of recombination depends on both initiator concentration

and the rate of its decay. The activation energy of recombination is close to zero while

the energy of fragmentation is greater than 29 kcal/mol.









Another, more effective method to crosslink PP is to use a polyfunctional

monomer. The idea is based on two assumptions: the double bonds of the monomer will

react with tertiary PP macroradicals to suppress fragmentation and the double bonds of

the monomer will increase the number of crosslinks between PP chains [68, 114, 180,

190]. Formation of stable radicals after addition of PP radicals is the most important role

of the monomer double bonds and is a very effective free radical trap. The contribution

of a few long polymer chains can drastically affect the elongational viscosity by

increasing the degree of chain entanglement in PP and creating chains 2-3 times bigger

than can be formed by intermolecular combination [215].

Ludwig and Moore found that peroxide initiated grafting of a hexafunctional

coupling agent can reduce chain scission of PP dramatically [220]. The high tail end of

the molecular weight distribution increased and was attributed to the formation of PP

crosslinks through the coupling agent. Notched Izod impact strength and tensile strength

improved with the coupling agent. The decrease in the degradation of PP is due to the

primary radicals reacting preferentially with multifunctional monomer [190].

A branched PP was created by the use of a polyfunctional acrylate monomer

(trimethylolpropane triacrylate), a peroxide initiator, and an iniferter compound [221].

The iniferter acts as a free radical initiator, chain transfer agent, and chain terminator and

was found to facilitate the long chain branching from TMPTA. The iniferter also

prevented gel formation and homopolymerization of the TMPTA, both of which are

deleterious to the ultimate properties of the grafted PP. Several authors have tried to

crosslink isotactic polypropylene in order to improve the melt extensibility and









mechanical properties but also to reduce chain scission while grafting functional

monomers.

Normally, thermal degradation of PP during its processing can be avoided by

adding thermal stabilizers (antioxidants). However, this kind of stabilizing mechanism

cannot be applied to the grafting process because the stabilizer eliminates the free

radicals, which initiate grafting. The major purpose of crosslinking is to both enhance the

mechanical properties of grafted PP and increase its melt viscosity, which may facilitate

its dispersion into other phases during melt blending [7, 222].

Monomers alone do not necessarily reduce degradation of polypropylene, but some

studies have shown that styrene grafting can cause cross-termination and increase

branching of the copolymer [92, 94]. Styrene-GMA copolymerization and grafting can

create long chain branching of PP macromolecules [94]. The resulting structure of

grafted PS materials is more likely a highly branched, entangled network, as opposed to a

crosslinked network. The entangled chain ends behave as temporary junctions. Although

theological testing revealed a gel point, soxhlet extraction with xylene yielded no

insoluble material.

2.5.4 Fundamentals of Free Radical Grafting

The proposed chemical mechanisms for free radical grafting reactions onto

polymers include initiation, propagation, transfer, and termination [94]. These are the

same reactions for free radical polymerization [154, 158, 159]. Once the initiator

decomposes and primary radicals are generated, they must diffuse away from eachother.

A concern when dealing with a highly viscous system is that two primary free radicals

recombine to yield an inactive species, known as the cage effect. If it breaks free of the

cage, it may abstract a hydrogen atom from the polymer backbone, creating potential free









radical grafting sites. Most of the functional monomers used for graft modification are

capable of homopolymerizing, which generally enhances the ability of the monomer to

graft onto the polymer [138]. Termination of the free radicals is either by kinetically

identical combination or disproportionation, which is negligible for primary radicals.

Chain transfer is not a likely source for grafting sites onto the polymer backbone [141].

In fact, the rate of propagation in melt grafting is over 35 times that of chain transfer [92].

The activation energy for grafting is slightly higher than initiator decomposition, which is

the rate determining step of the reaction [223].

Few studies have probed free radical reactions under the conditions likely to be

encountered in melt phase polymer reactions. These involve relatively high

temperatures, relatively high pressures, and media of relatively high viscosity [68, 138,

181]. As homopolymer is produced, there is some phase separation; the monomer that is

trapped in the homopolymer phase cannot graft onto the polymer, resulting in a lower

degree of grafting and grafting efficiency than the model system. Homopolymerization

does compete successfully with the grafting reactions in model systems [111].

The kinetics and mechanism of grafting onto a saturated polymer (like

polybutadiene) is the same as homopolymerization in solution regardless of whether the

polybutadiene is present [179]. Both free and grafted chains are in the same environment

and presumed to grow at equivalent rates [139-142, 224, 225]. Styrene

homopolymerization readily competes with grafting but if benzyl acrylate is used in place

of styrene, grafting is severely reduced. This is attributed to the fact that the benzyl

acrylate monomer is much less reactive than styrene and the rather inactive polybutadiene

macroradical cannot compete for benzyl acrylate monomer [140]. The grafting of styrene









and homopolymerization is 10 times that of a monomer which tends to graft as single

units because of styrene's high ceiling temperature [92].

When grafting polypropylene, the ratio of tertiary to secondary to primary grafting

sites in neat solution is 5.6:1.5:1, respectively, and 3.9:0.3:1 in benzene, respectively

[150]. This study suggests that the position of t-butoxy radical initiated grafting in

LLDPE and PP is most likely different, with LLDPE grafting mainly at the secondary

sites.

2.5.5 Melt Grafting Monomers onto Polyolefins

Gaylord and Mishra have found that free radical based functionalization of

polypropylene results in random attachment of the functional group in conjunction with

degradation of PP by 0-chain scission [226]. In fact, some authors have found that when

grafting monomers onto PP, the majority of the grafts are formed after chain scission [68,

94, 180].

Doney and Salsman were successful in patenting a reactive extrusion process for

creating block copolymers of isotactic polypropylene in the presence of an alkenically

unsaturated polar monomer and a peroxide initiator [227]. Polypropylene is first

preferentially degraded via P-chain scission by peroxide-derived free radicals. This is

followed by addition of a peroxide/monomer mixture to the degraded chains which

therefore link up PP chains to form block and graft copolymers with tailored

hydrophilicity.

When grafting acrylate functional monomers onto polyolefins, their ceiling

temperatures are typically above 4000C and homopolymerization will occur alongside

grafting. Methacrylate monomers, on the other hand have ceiling temperatures

approximately 2000C and tend to graft as very short chain branches with limited









homopolymerization. The initiator derived primary radicals have a relatively low

reactivity towards the acrylated monomer and preferentially abstract hydrogen atoms

from the hydrocarbon substrate. Addition of the monomer to the resulting polyolefin

radical has a rate constant four orders of magnitude faster than homopolymerization. On

the other hand, styrene undergoes more rapid addition to hydrocarbon radicals [156].

Over the years, several routes have been developed to improve the grafting

efficiency of functional monomers. One early process used toluene as an interfacial

agent [207]. The graft yield increased for two reasons: improved solubility of the

monomer in the solvent and more surface area by swelling the surface of the polymer.

The main type of PP grafting proposed at elevated temperatures is P-scission followed by

graft initiation on one of the chain fragments [94]. The use of a multifunctional monomer

greatly reduced the extent of degradation but did not seem to improve the grafting yield

of GMA [7].

Hu et al. were the first to systematically study chemical methods for improving the

free radical grafting yield onto PP while minimizing degradation [191]. They proposed

three routes, all based on increasing the reactivity of the monomer and/or rate of reaction.

The comonomer concept was derived, which is fundamentally dependent upon the

reactivity ratio of free radical copolymerization. This means that for a monomer to be

grafted onto a macromolecular chain, adding comonomer will be beneficial for improving

the monomer grafting yield only if this comonomer reacts with a macroradical more

rapidly than the grafting monomer and the resulting macroradical is capable of

copolymerizing with the grafting monomer.









A consensus from several authors is that grafting yields are higher when styrene

comonomer is used, so styrene must preferentially graft onto the polymer macroradical

before chain scission occurs [7, 40, 92, 94, 206, 228]. Styrene then forms a stable styryl

macroradical which is long lived and invites copolymerization with polar monomers.

The intrinsic free radical grafting rate is so much accelerated by the presence of styrene

that the overall grafting kinetics becomes controlled primarily by PP melting. Another

reason for the high grafting yields may be due to the fact that polyolefins such as HDPE,

LLDPE, and PP are soluble in refluxing styrene [138]. A third effect of styrene is that is

forms a charge transfer complex with the glycidyl methacrylate, which is more reactive

than GMA alone.

2.6 Melt Grafting and In-situ Compatibilization

One of the first methods to bond polypropylene and polyethylene was to melt blend

them in an extruder in the presence of a peroxide initiator. It was expected that the chain

decay of PP and the chain buildup of PE should be balanced in PP-PE blends where

coupling of PP and PE macroradicals can lead to graft copolymer chains PE-g-PP. Short

PP oligomers were expected to be tied to PE chains so as to prevent crosslinking [70,

229-234]. Contrary to popular belief, the polymers reacted fairly independent of each

other. A similar effect was seen for PE-PP copolymers [91, 235, 236].

This interchain coupling is favored only in one phase blends where the compounds

are in intimate molecular contact. In two phase blends, the components can react only at

the interface between phase domains so grafting is hindered. Although limited bonding

exists between the phases, the resulting alloy does have higher toughness and melt

strength with low gel content [232, 237]. High melt strength is characteristic of materials

with long side branches (LDPE) or high molecular weight (PS) [238].









In order to reduce or prevent the degradation seen in the peroxide/PP/PE blends, a

multifunctional monomer was added to the mix. A recent patent focused on using a small

percentage of multifunctional coagent in addition to peroxide for the compatibilization of

PP and an ethylene copolymer. A very high melt strength is observed and degradation of

PP is essentially eliminated without the formation of substantial gel [239]. Yoon et al.

reported a tri-functional acrylate monomer (TMPTA) is effective at minimizing PP chain

scission by bonding to PP macroradicals before chain scission and stabilizing them via

resonance [222]. By crosslinking a blend of PP and LDPE, the interface becomes

obscure (indicating some bonding between phases), with a 6X increase in impact strength

over pure PP [233]. But it has also been shown that the multifunctional coagent only

raises the crosslinking efficiency in PP, not PE [240].

A process called dynamic vulcanization can be used to crosslink rubber in a PP

matrix in-situ to improve bonding and stress transfer between phases. A themoplastic

vulcanizate (TPV) is thus formed, showing improved impact properties, solvent

resistance, and long term elastic recovery over the physical blend, but an insoluble gel is

typically formed [136, 241, 242]. Dispersion with these alloys is typically sub-micron

and dispersed phases can be as small as 30 nm in size [243].

A nano-structured polypropylene/polyamide 6 (PP/PA6) alloy has been created in-

situ by anionically polymerizing e-caprolactam in the presence of functionalized PP [244-

246]. The resulting material had PP as the continuous matrix and PA6 homopolymer

dispersed as very small domains. The reaction is very fast and by proper control of the

kinetics of both reactions, it is not difficult to control the particle size of the dispersed

domains, leading to enhanced properties.









Flaris and Baker used a unique approach to compatibilize PE and PS in an extruder

[247]. A vector fluid approach was utilized, which is simply using a carrier fluid to bring

a reactive ingredient to the interface of the immiscible polymers. A vector fluid may be a

low molecular weight material which can stay long enough in the blend to carry the

reactive material (a peroxide in this case) to the interface and is then fugitive at some

later point. For example, it may be drawn off at a vent port close to the end of a twin

screw extruder barrel. Using a viscous vector fluid led to little grafting and large

dispersed domains in a compatibilized PE-PS blend. This was attributed to low

dispersability of the reactive ingredients and a reduction in peroxide activity.

During in-situ compatibilization, continuously new interfaces can be generated

with proper mixing equipment. An appropriate vector fluid must be used so that the

coupling reaction is restricted to the interface of the blend components. It must dissolve

the reactive ingredient much more easily than the polymer melt so that little reactive

ingredient can diffuse into the polymer phases. If the reactive ingredient contains both a

monomer and initiator, both grafting and homopolymerization may occur. But this is

likely to happen only at the surface of the minor phase polymer particles, which is

surrounded with the vector fluid. In this case, the graft copolymer formed should locate

right at the interface and possibly entangle with the other blend components at the

polymer-polymer interface [143].

Grafting level and interfacial adhesion are known to be higher for vector fluids of

low viscosity. Styrene monomer is used as a reactive vector fluid and grafts onto

polyethylene macroradicals. Bridging may occur between PE and PS due to the

miscibility imparted by the grafts. The highest graft levels are found with low MW fluids









that are not volatile, partly miscible with the PS phase, but yet remain at the interface of

PE and PS. Lagardere and Baker used the vector fluid approach to compatibilize a blend

of LDPE and PS in an extruder [248]. A peroxide initiator with styrene monomer was

shown to be a very effective system because bonds were created at the interface, thus

reducing interfacial tension, altering the morphology, and improving overall mechanical

properties. Using this approach, PS domains were approximately 50 nm in diameter

which means that the immiscible units of PS and LDPE are interconnected [249]. The in-

situ polymerization process yielded no crosslinked material and a broader molecular

weight distribution than LDPE alone. A similar result was found by Teh and Rudin

[250].

Styrene is not the only monomer used to compatibilize two polymers by in-situ

reactive processing. A patent describes a composition in which an acrylate monomer,

initiator, and a diacrylate are adsorbed into one or more polyolefins and reactively

processed to create what they call a thermoplastic elastomer [251]. The bonding between

polymers and interactions between functional groups enhances properties. In-situ graft

copolymers can lead to dispersed domain sizes on the order of 50 nm [86]. HDPE and PP

have been reactively processed with a dialkyl peroxide and n-butyl methacrylate to

suppress unwanted PP side reactions and improve compatibility [148].

Tang's dissertation focused on reactively extruding polyolefins with high molecular

weight ethylene-a-olefin copolymers [11]. The alloy was created by adding a peroxide

initiator, styrene monomer, and an epoxide-functional methacrylate to PP and

ENGAGE pellets as they entered the extruder so as to graft both polymers in-situ.

Impact strength jumped by over 8X from the pure polymer, the morphology contained









finer and better dispersed domains of elastomer, and stress strain behavior improved

compared to physical blends.

2.7 Conclusions

The process of reactive extrusion is a complex one, to say the least. Once a suitable

system is chosen (i.e. free radical polymerization mechanism using high molar mass

polymers in a reactive twin screw extruder), many variables have to be defined and

optimized in order to get the best possible performance of the alloy. The following

chapters will attempt to tie in many aspects of reactive extrusion to the in-situ grafting of

two polyolefins for the creation of high impact alloys.














CHAPTER 3
PHYSICAL BLENDING OF AN IMPACT MODIFIER WITH POLYPROPYLENE

3.1 Introduction

A fundamental study of the physical blends of a low molecular weight grade of

ethylene-1-octene copolymer (EOC) with isotactic polypropylene (PP) is undertaken.

Little research has been conducted using low viscosity ethylene-octene copolymers to

toughen polypropylene, but this chapter proves that these new copolymers are effective

PP modifiers up to a certain concentration. Many concepts and characterization

techniques are introduced and explained in detail so as to aid in analysis of results both in

this chapter and all following chapters.

3.2 Experimental

3.2.1 Materials

Table 3-1 gives a list of pertinent ethylene-1-octene copolymers produced by

Dupont Dow elastomers under the tradename ENGAGE, but the EOC grade of interest

is 8407 [48]. Isotactic polypropylene homopolymer was supplied by Equistar Chemical

(grade PP 31S07A) and is contact translucent. All polymers were received in pellet form.

3.2.2 Methods

3.2.2.1 Processing

All polymers were dried in an air circulating oven at 40C for 24 hours prior to

compounding. Before processing, the resins were premixed by hand for about 10

minutes. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin









Table 3-1: ENGAGE product data table.
ENGAGE Grade (decreasing 8842 8407 8200 8401 8402
comonomer content)
Comonomer Content wt% 13C 45 40 38 31 22
NMR/FTIR
Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902
Melt Index, dg/min ASTM D- 1.0 30 5.0 30 30
1238, 190C, 2.16kg
Mooney Viscosity ASTM D- 26 < 5 8 < 5 < 5
1646 ML 1 + 4 at 121C
Durometer Hardness, Shore A 50 72 75 85 94
ASTM D-2240
DSC melting Peak, oC Rate: 33 60 60 78 98
10C/min
Glass Transition Temp, oC DSC -61 -57 -56 -51 -44
inflection point
Flexural Modulus, MPA ASTM 3.5 12.1 12.1 25.8 69.9
D-790, 2% Secant
Ultimate Tensile Strength, MPa 2.1 3.3 6.9 6.4 12.9
ASTM D-638, 508mm/min
Ultimate Elongation, % ASTM 975 >1000 >1000 950 790
D-638, 508mm/min

screw extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D

ratio of 39. The temperature of the extruder was regulated by electrical resistance and

water circulation in the barrels. The screw speed, unless otherwise noted was 150 rpm.

The dried, pre-mixed resins were then introduced into the extruder from the hopper

of the extruder at 60 g/min through a screw driven dry material feeder from Accu Rate,

Inc. Devolatilization was carried out by a vacuum pump, VPS-10A, Brooks Equipment

Company. This was placed near the die and created a vacuum of about 15 in Hg. The

extruder was always starved to feed. Figure 3-1 is a schematic of the extruder, with a

typical temperature profile. After compounding, the resultant strands which exit the die

are quenched in a water bath, pelletized, and dried in a vacuum oven at 1000C, 28 in Hg

for 24 hours.










Zenith Pump
Power


eeder
SControl Panel

Reactive Species Power 340 V
vacuum| |
Pelletizer Water Bat
Water Bath



ater in water in

Water out
Water out

Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed)
195C 205C 210C 210C 200C 190C 180C 165C
Figure 3-1: Schematic drawing of the reactive twin screw extruder and a common
temperature profile.

3.2.2.2 Mechanical properties

In order to measure the strength of the materials at very high testing rates, a

notched Izod impact test was performed according to ASTM D256 standards. The pellets

were placed in a mold with 6 slots, each measuring 0.5x0.5x2.5 in3. The mold is put in a

Carver press (Fred S. Carver, Inc.) at 2000C and after the material is melted, pressed up

to 5000 psi. After waiting for 5-10 minutes, the pressure is slowly increased up to 10,000

psi. After another 5 minutes, the heat is turned off and the sample is let to cool down to

room temperature at about 1.50C/min. The bars were then taken out and notched (0.1 in

deep, 0.01 in radius) with a Testing Machines, Inc. (TMI) notching machine. Before

testing, they were conditioned at room temperature for 24 hours and a 30 ft-lb hammer

was used with test method A on a TMI Izod impact tester. At least 5 bars were broken

and impact strength is recorded regardless of full or partial break.

For stress-strain measurements, dried pellets were placed in a mold measuring 15

cm2 x 1 mm thick. The mold was put into the Carver press at 2000C and after the









material melts, pressed up to 5000 psi. After a 5-10 minute wait, the sample was slowly

pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath.

Specimens were tested according to ASTM D638 standards. Type V specimens were

punched out of the compression molded sheet with a die, measuring 1 + 0.15 mm

thickness, 2.95 mm gauge width and 9.5 mm gauge length. Five samples were tested

after conditioning at room temperature for 48 hours. The machine used to test the

samples was an MTS Model 1120 Instron, using a 1000 lb load cell at a test speed of 12.7

mm/min.

A Seiko DMS220 interfaced with a Seiko Rheostation model SDM/5600H was

used to test dynamic mechanical specimens. Testing was conducted from -120C to

1500C at a heating rate of 5C/min in dry nitrogen atmosphere maintained at an

approximate flow rate of 100 mL min-. Rectangular samples (20xl0xlmm3) were cut

from the compression molded sheet and tested in bending mode at a frequency of 1Hz.

3.2.2.3 Morphology

Scanning Electron Microscopy (SEM) was performed on a JEOL 6335F Field

Emission SEM. The microscope was kept under vacuum at 1x105 Pa, with an

accelerating voltage of 5 kV and working distance of 13.4 mm with secondary electron

detection at various magnifications. For better phase contrast, etching was done to

remove amorphous, elastomeric material. Samples were etched by the following

procedure: A notched impact bar was immersed in liquid nitrogen for 10 minutes and

immediately fractured using a TMI Izod impact tester. A section of the cryofractured

surface (2 cm thick) was then immersed in xylene (purchased from Fisher Scientific) at

600C for one hour. The sample was removed and dried under vacuum at 400C for 12

hours. Sample mounting was on an aluminum stub with conductive carbon paint from









Ted Pella, Inc. The sample was then coated with carbon then vacuum dried at room

temperature for one hour prior to examination in the microscope.

Image analysis was done using ImagePro software. Three images per sample were

recorded and domain size and distribution were quantified. The domains in all images

had to be manually outlined for the software to recognize them as discreet phases. For

the majority of samples, over 1000 particles were considered for diameter and roundness

measurements. The number average (Dn) and weight average (Dw) diameter were

determined using the following equations:

- n= NiDi (3.1)
S;Ni


D S ND i (3.2)
ENCD
where Ni represents the number of particles with diameter Di.

3.2.2.4 Thermal analysis and rheology

Differential scanning calorimetry (DSC) was used to study the different

thermodynamic transitions present in the blends. DSC was performed on a Seiko SII

DSC 220C-SSC/5200, Seiko Instruments, equipped with a Seiko Rheostation model

SDM/5600H and calibrated with indium and tin standards. Samples (approx. 7 mg in

weight) were sealed in crimped aluminum pans, with the reference being 99.99% pure

alumina. Purging of the sample was done with dry nitrogen at a flow rate of 100 ml/min.

Each sample experienced two heating and cooling cycles (shown in Table 3-2) with the

first to erase prior thermal history. The second cycle is reported in all graphs. The %

crystallinity is found by first integrating the heat flow curve to a flat baseline then

dividing by the heat of fusion of a perfect PP crystal (207 J/g).









Melt Flow Index (MFI) testing was done according to ASTM D1238 (230C and

2.16 kg weight) on a Tinius Olsen model MP 933 Extrusion Plastometer. For materials

Table 3-2: DSC consecutive heating/cooling cycles
Step Start End Temp Heating/Cooling Rate Hold Time Sampling
Temp (OC) (oC/min) (min) (s)
(C)
1 -70 200 20 3 3
2 200 -80 20 5 3
3 -80 200 10 5 1
4 200 -80 10 3 1

with a flow rate of 0.5-3.5 g/10 min, the weight of the samples was approx. 3 g, whereas

materials with flow rates of 3.5-300 g/10 min, the sample weight was approx. 6 g. All

materials were dried under vacuum then conditioned at room temperature before testing.

3.3 Results and Discussion

3.3.1 Mechanical and Rheological Properties

Impact strength (IS) is the ability to resist a high loading rate (approx. 3.6 m/s) and

is one of the most important properties for plastics part designers to consider because it

sets up the worst possible condition for plastics [252, 253]. It is a critical measure of

service life, product safety, and liability. The standard Izod notch ideally functions as an

artificial crack because of its sharpness and acts to concentrate the applied stress,

minimize plastic deformation, and direct the fracture to the part of the specimen behind

the notch. Fracture of the Izod specimen is dominated by bending-moment-induced

tensile stress, but extensive plastic deformation is possible.

As can be seen from Figure 3-2, impact strength increases, although not

monotonically, with increasing elastomer content and reaches a peak at 20 wt% 8407.

This increase is to be expected because the elastomer phase has a much lower glass

transition temperature (Tg) than the matrix material and thus promotes various energy







54


absorbing mechanisms such as cavitation, shear yielding, and crazing. Also, in

semicrystalline polymers, elastomers or rubbers act as nucleating agents and thereby

reduce spherulite diameter [103, 241, 254-258]. This reduction in spherulite size is

known to improve impact strength due to an increase in interfacial thickness, better inter-

spherulitic chain mobility, and reduction of spherulitic defects [254, 255, 259, 260].


8 -

7 I Impact Strength
_" I I MFI

6
Eo
'-0

c X




0






0 5 10 15 20 25 30

Concentration of 8407 in Blend
Figure 3-2: Effect elastomer concentration on room temperature notched impact strength
and melt flow index.

Above 20 wt% 8407, a drastic decrease in impact strength is observed. At this

level of elastomer, the interface between phases becomes of great importance. These two

phases are only partially miscible at low elastomer concentrations, so gross phase

separation may lead to brittle behavior because of poor stress transfer between phases

[254, 261, 262]. At high strain, property deterioration takes place as incompatibility

leads to cracks and failure at inter-phase boundaries. Also, at high elastomer loadings,

intra-spherulitic regions of PP become very diffuse and do not have the load bearing









capacity of a fully developed spherulite [257, 261, 263, 264]. It should be noted that the

impact strengths of samples fractured at liquid nitrogen temperatures (for morphological

analysis) were about 0.4 ft-lbs/in, regardless of elastomer concentration.

Melt flow index is a measure of the uniformity of the flow rate of a polymer and is

not a fundamental property. It is a relative test which allows one to make inferences on

molecular weight and viscosity. MFI is proportional to molecular weight in the region of

entanglement and gives a good indication of zero shear viscosity.

From figure 3-2, one can see that the melt flow index increases in a somewhat

exponential fashion with increasing elastomer content. The blend with 30 wt% 8407 has

the highest MFI, which is expected because the MFI of this elastomer is over 35 times

that of PP. As melt flow index increases, impact strength is known to decrease [265].

This is only true for samples containing greater than 20% elastomer.

In stress-strain experiments of semicrystalline polyolefins, deformation first occurs

in the amorphous phase followed by activation of crystallographic mechanisms [105,

266-268]. For the initial slope of the force-length curve (2-3% strain is usually reversible

and is known as the elastic modulus), the deformation of the disordered interlamellar

regions (loose chain folds, tie molecules, cilia, chain entanglements, as well as

completely unincorporated molecules) are involved, and the lamellar structure remains

intact [269, 270]. The lamellae present in the sample also behave effectively as crosslink

junctions and provide resistance to deformation [271], but some authors dispute this

mechanism [272]. The initial elastic part is followed quickly by a viscoelastic part, in

which the stress gradually increases to reach a maximum at the yield point [273].







56


Figure 3-3 shows typical stress-strain responses of the physical blends while Figure

3-4 represents tensile property trends as a function of elastomer content. As can be seen

from these figures, elastic modulus is a roughly additive function of blend composition

and gradually declines with elastomer content [274, 275, 262]. The low crystallinity and

relatively low average molar mass of the elastomer means that in the amorphous phase of

PP, more chain ends are present (reduction in entanglement density), chain mobility is

enhanced, total % crystallinity is reduced, as well as chain stiffness. It should be noted

that standard deviations for stress-strain data are given in Appendix B.


50 --
-5 Pure PP
................... (95:5) PP:8407
-----. (90:10) PP:8407
.........- (80:20) PP:8407
40- -- (70:30) PP:8407



CO 30

20
20 .... .-- --



10 -



0 \
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Strain
Figure 3-3: Stress-strain behavior of PP:8407 physical blends

When a polymer reaches its yield stress (also termed proportional limit), energy

barriers are overcome along with dilation and long range diffusion [269]. At this point,

plastic flow localizes in a neck and the stress decreases towards a plateau value (cold

drawing) [273]. The dominating mechanisms of yielding is partial or local melting,







57


stretching of amorphous chains, and subsequent recrystallization, with competing

mechanisms controlled by dislocation motion [105, 269, 272, 276]. Yield stress is

proportional to lamellar thickness, unlike the elastic modulus.


80 Yield Stress 2000 r 7000
cU` I I Stress at Break
70 Elongation at break 6
S-o-- Elastic Modulus -1800
E, -- Energy to Break
SE 60 m 5000 m
S_ e\- 1600 c
-c 50 4000<

S40 1400 3000
I I Of I \ \ -
S30 2000
C)\- 1200 3
vC C 3
S020 1000
U- 1000
10 -

S 0 II, II II II II 80oo
100 to0 95 to 5 90 to 10 80 to 20 70 to 30
Ratio of PP:ENG8407
Figure 3-4: Stress-strain properties of PP-8407 blends as a function of elastomer content.

Yield stress, like elastic modulus, appears to be a monotonic function of blend

composition and decreases almost linearly with an increase in elastomer concentration.

This is directly related to the linear decrease in total % crystallinity of the blend with

increasing 8407 content (explained in section 3.3.4). Pure PP shows a characteristically

sharp yield point whereas addition of elastomer broadens this peak and increases the

strain at yield. The reduction in yield stress with elastomer may be due to the

dependence of yield stress on lamellar thickness (the greater elastomer content, the

smaller the lamellar thickness). Also the greater mobility of the amorphous region

reduces the activation energy needed for crystallographic slip, dislocation motion, and

chain disentanglement.









The stress drop after yield (cold drawing) is associated with a decrease in unit cell

volume and cross-sectional area which is a direct result of shearing and fragmentation of

less perfect crystals [266, 277]. Necking (a form of shear yielding) accompanies this

stress drop, where a very extensive reorganization of the polymer is taking place.

Spherulites are broken up and the polymer becomes oriented in the direction of the

stretch [100, 273]. The number of chain folds decreases, and the number of tie molecules

between the new fibrils is increased.

Deviations in post yield behavior can be seen in Figure 3-3 for the various blends.

Pure PP exhibits the largest drop in stress but this occurs over a relatively small strain.

The high crystallinity of the pure polymer and limited thickness of amorphous layer lends

it to a greater degree of rearrangement followed by orientation. The effect of elastomer

on cold drawing is obvious at 20% 8407, where the stress drop is relatively small and its

slope is much less than pure PP. Diffuse spherulites are likely, with elastomeric material

existing in both inter- and intra-spherulitic regions. The presence of elastomer appears to

enhance this spherulite breakup process. At 20% 8407 impact strength is highest and this

may be related to the high strain reached at the end of cold drawing. The area under the

stress-strain curve before strain hardening occurs may be related to energy absorption at

high strain rates.

A process known as strain hardening occurs post-yield and is indicated by a

positive slope in the stress-strain curve. This is a result of chain unfolding and

orientation involving the so-called fibrillar transformation which leads to a continuous

increase in crystallinity. Mechanical work reduces the thermodynamic barrier between

the metastable and stable crystals and helps chains find their way to more stable potential









energy wells during deformation. The more defective crystals are destroyed during

drawing and rebuilt into more perfect crystals with a narrower size distribution [132, 161,

266, 269, 277-279]. The slope therefore decreases with increasing elastomer

concentration, indicating PP spherulites have an increasing degree of disorder and

decreasing load-bearing capability. Strain hardening doesn't occur at the 30 wt% 8407

level, so the ordering and alignment of crystallizable chains are severely hindered by the

large domains of 8407, which are highly mobile, primarily non-crystalline, and

immiscible with PP.

Elongation and energy at break for elastomer modified PP are poor at high

elastomer concentrations because the high interfacial energy between phases dominates

failure [261, 262]. The ability to neck and draw is very defect or morphology sensitive,

so at 30% 8407, the amorphous content may act as a defect by promoting spherulite

breakup and inhibiting recrystallization.

3.3.2 Morphology

Mechanical performance of the physical blend is highly dependent upon its

morphology. For SEM imaging, etching of the elastomer phase was required because the

domains could only be seen at 30% concentration. From Figures 3-5(a) thru (e), the

etching procedure produces dark pits where elastomeric material once was. Another

obvious trend is shown increasing elastomer concentration results in an increase in

domain size and decrease in overall matrix-domain interface. In turn, the ligament

thickness or distance between elastomer particles [119] is increased. These ideas will be

elaborated upon in the proceeding paragraphs.





















(a) (b)


(c) (d)


(e)
Figure 3-5: SEM images of etched, cryo-fractured surfaces of PP:8407 physical blends as
a function of elastomer concentration. (a) Virgin PP at 2,000X, (b) 95:5_0 at
10,000X, (c) 90:10_0 at 10,000X, (d) 80:20_0 at 10,000X, and (e) 70:30_0 at
2,500X. The bar markers for (a) and (e) = 10,tm, and for (b), (c), and (d) =
l1 m.









Table 3-3: Image analysis averages taken from etched SEM images of PP:8407 physical
blends. At least three images were analyzed from Figures 3(a) thru (e), with
data compiled in Appendix D.
Sample Average Roundness for Ligament
ID Dn (nm) Dw (nm) particles >1* Thickness (nm)
95:5 0 77 104 1.08 166
90:10 0 113 167 1.1 194
80:20 0 224 778 1.2 305
70:30 0 269 2630 1.2 321
*Roundness is a measure of how closely the particle's shape matches that of a
perfect circle. A value of 1 = a perfect circle.

Table 3-3 (generated from Appendix D) represents image analysis of the SEM

pictures and reveals that a relatively narrow particle size distribution exists at low

elastomer concentrations. The number average particle diameter (Dn) for the 95:5 blend

is approx. 77 nm while at 90:10, Dn increases to 113 nm. The viscosity of the minor

phase plays a huge role in determining these values. The matrix viscosity is much higher

than the elastomer viscosity, so coalescence is suppressed at low 8407 concentrations.

As the morphology of the blend develops in the extruder and the minor phase is

elongated, dispersed, and distributed, the diffusion process of coalescence is kinetically

much slower than breakup. So, for the short residence times experienced in the extruder,

the domains remain submicron [17, 75]. Also, the particles remain relatively spherical at

these low concentrations.

With increasing elastomer concentration, the anisotropy and average domain size of

the particles increase. These results are typical of what previous researchers have found

regarding PP-elastomer blends [18, 75, 128, 254, 274, 275, 280]. There also seems to be

less of a monomodal distribution of elastomer domains at higher concentrations. The

smaller interfacial area of the large domains, as well as the large distance between

domains, can be directly correlated with mechanical properties. At such high loadings of









8407, the stress field around the elastomer domains is modified and interactions with the

matrix are hindered so the blend reacts in a brittle manner to an applied load.

The large size of the dispersed phase at high concentrations is due to coalescence,

which is the recombination of particles known to take place during the mixing process

arising from forced collisions. Coalescence (collisions) of the minor phase is the rate

limiting step during morphology development.

3.3.3 Viscoelasticity

Dynamic mechanical behavior is studied at the molecular level and structural

factors affect it, including molecular weight, crosslinking, crystallinity, lamellar

thickness, and interfacial interaction between phases [57, 81, 235, 281, 282]. Polymers

are examples of viscoelastic materials, which have characteristics of both viscous liquids

(only dissipate energy) and elastic solids (only store mechanical energy) [94, 235].

Deformation in solid polymers is dominated by relaxation processes, which are sensitive

to morphology and crystallinity [81]. A necessary condition for a highly plastic

deformation is the possibility of motions of kinetic elements on a time scale similar to the

deformation rate. Dynamic mechanical analysis (DMA) is a valuable tool in the

characterization of viscoelastic behavior, which is the mechanical behavior dependence

upon time and temperature. It is able to separate the viscous (loss modulus or E") and

elastic response (storage modulus or E') of the material and relates the two by Tan6 (E"/

E').

The first quantity to be studied is the storage modulus (Figure 3-6), which is an

indication of the stiffness of the polymer and may be considered inversely proportional to

impact strength in physical blends [283]. The stiffness typically increases with

increasing density (crystallinity) or entanglement density (from high molar mass) [284].









The behavior in Figure 3-6 is similar to plasticizing a brittle glassy polymer with a low

molecular weight compound [285].

There are typically five regions of viscoelastic behavior for semicrystalline

polymers [100]. The low temperature (glassy) region in Figure 3-6 occurs at

temperatures less than about -450C. Long range motion is frozen in at these

temperatures, so the stiffness of the blend is dependent upon the free volume and


5.5e+9 -
5.0e+9
4.5e+9
4.0e+9
3.5e+9
n 3.0e+9
Eu 2.5e+9
2.0e+9
1.5e+9
1.0e+9
5.1e+8
5; ne+ -


-100 -80 -60 -40 -20 0 20 40 60 80 100 120
Temperature (oc)
Figure 3-6: Storage modulus vs. temperature for PP:8407 Blends

mobility of side and end groups as well as secondary forces between chains. The long

side chains of 8407 renders the blend more pliable due to the increase in free volume.

The glass transition (Tg) region is accompanied by a drop in modulus of about 2-e9 over a

range of about 300C, starting at about -450C for the elastomeric phase and 100C for the

matrix phase. This is the onset of long range, coordinated molecular motion.

A rubbery plateau region after Tg begins to take shape for the elastomer phase at

high concentrations. Long range rubber elasticity is present in this short temperature


SPure PP
........... (95:5) PP:8407
-_----- (90:10) PP:8407
S.. (80:20) PP:8407
-\ (70:30) PP:8407


N \

'\ \ ^



\\









range and is likely due to the entanglements associated within the amorphous phase. The

rubbery flow region follows at even higher temperatures in the range of about 10-30C

for 8407 and 40-700C for PP phase. In this region, the polymer is marked by both rubber

elasticity (short time scale) and flow properties (long time scale). The response is

dependent upon physical entanglements and with the more mobile elastomeric phase

present, chains are able to move in a more coordinated motion leading to flow. The

liquid flow region occurs at the highest temperatures and represents chain reputation out of

entanglements and flow as individual molecules. Segments are free to move from one

lattice site to the other, and the hard polymer becomes soft and rubbery. Keep in mind

that this behavior is strictly in the amorphous (non-crystalline) phase of the material.

Isotactic polypropylene and ethylene-octene copolymers have three primary

viscoelastic relaxations [81, 110, 235, 281, 282, 284, 286-288]: the y relaxation around

-600C for PP and -1200C for 8407 is attributed to local motion of side groups, end

groups, and short main chain -CH2- links; P relaxation located between 0C and 300C for

PP and -500C to -100C for 8407, represents large scale (non-crystalline amorphous) chain

motion due to an increase in amorphous volume and is much broader than in wholly

amorphous polymers; a relaxation localized between 400C and 900C for PP and 600C to

800C for 8407 exists only in the presence of the crystalline phase and originates from

diffusion of crystallographic defects, interlamellar slip, and motions of the interfacial

regions containing tie molecules, folds, loops, etc.

It is obvious from Figure 3-7 that upon addition of elastomer, two P relaxations are

present the lowest temperature peak representing 8407 and the peak located about

200C, representing PP. When determining the relative solubility between polymers,









separate Tan6 peaks will exist for phase separated blends [8, 9, 81, 289]. The magnitude

of each 0 peak is characteristic of the relative concentrations of the two components,

regardless of a dispersed or co-continuous morphology. Broadened peaks and shifts

towards an intermediate temperature signify mutual solubility of polymers. Studies have

shown that PP is largely immiscible with polyethylene and polyethylene copolymers

[290-298] but at low copolymer concentrations, limited partial miscibility is known to


PurePP
0.6 -..(95:5) PP:8407
0.6 ------- (90:10) PP:8407
-. (80:20) PP:8407
-- (70:30) PP:8407
0.5
/\ /
0.4 0.12 -
/ i
../ / /
S0.3 -0.08- PP /
0 P8407
0.2 /.04



0.1 -60 -40 -20 0 20 40 6/ /.


0.0 -
-100 -80 -60 -40 -20 0 20 40 60 80 100 120
Temperature (C)
Figure 3-7: Tan6 vs. temperature for PP:8407 blends from -1200C to 1200C. Insert
represents Tan6 vs. temperature from -600C to 600C.

exist [18, 299, 300]. The limited miscibility is obvious from the shift of the maximum

38407 peak toward higher temperatures with increasing copolymer content. The shift to

higher temperatures indicates that the apparent activation energy of the 18407 peak slightly

increases with concentration in the blends [283, 301]. There may also be a restriction in









8407 mobility due to the trapping of chains in the spherulitic structure of PP. The

intensity of both p8407 and 3pp peaks are known to increase with more elastomer [302] or

decrease with increasing degree of crystallinity [81]. PP typically does not show major

changes in peak position with crystallinity. The 3pp maximum becomes more of a

shoulder at decreasing concentrations and overlaps with 08407. The PP peak position is

also stationary for all values of elastomer. These findings are similar to that of Xiao et al.

[258].

Only at low elastomer concentrations can viscoelasticity of the blend be correlated

with impact strength. The time scale involved is comparable to the order of magnitude as

the relaxation time of viscoelastic relaxations (milliseconds). In elastomer toughened

blends, toughness shows some correlation with the area of the Tan6 due to the primary or

secondary transition of rubbery component. For an impact test carried out at room

temperature, a material will be ductile if it contains one or more prominent subambient

relaxation. [109, 281, 283, 303, 304].

Above 400C, the slope of the Tan6 curve increases for all samples but a trend is

seen in that the slope upturn is more drastic with increasing elastomer concentration.

This should not be surprising because the enhanced mobility afforded by the elastomer

enables long range motion and relaxation of chains at lower temperatures. A broad

peak/shoulder at about 800C is seen for Pure PP as well as 95:5 and 90:10 blends. This

represents the a relaxation of PP and dramatically increases in intensity with high

elastomer concentration. The mobility of defects in the crystalline phase as well as on the

surface of the crystallites is enhanced by the presence of highly mobile amorphous

elastomeric chains. There is also a slight shift in the a relaxation to lower temperatures









with increasing 8407 concentration, indicating a lower activation energy needed for chain

motion. Two distinct high intensity peaks are seen at about 75C and 1000C for the

70:30 blend and are indicative of the diffuse spherulitic regions of PP, thin lamellae, high

defect concentration, and possibly even the onset of crystal melting [283, 305]. Large

deformations may be expected to involve the motion or deformation of crystallites or

aggregates of segments. So, the high temperature peaks may be associated with the yield

stress of the blends at room temperature [306]. The lower activation energies associated

with the 70:30 blend can be related to the neck formation during a stress-strain

experiment, where unstable deformation and brittle failure occur [105].

The loss modulus (E") is a measure of the energy absorbed due to relaxation and is

useful in clarifying the mechanisms of internal motions. From Figure 3-8, the height of

each peak represents the relative mobility of the polymer chains. The peak position (at a

given temperature) depends on the chemical structure, flexibility of the molecular chain,

steric hindrance, and bulkiness of side groups. The high temperature side of the y peak of

8407 is seen below -800C and its intensity characteristically increases with increasing

elastomer concentration.

The Tg (0) peak of PP is found at about 200C and with increasing elastomer

concentration it decreases in temperature and intensity. The energy barrier for chain

motion therefore decreases with increasing elastomer concentration and this peak

becomes more of a shoulder simply due to the fact that the PP concentration in the blend

is decreasing. The high temperature peak at about 700C decreases in intensity due to the

lower PP concentration but sharpens possibly due to increased mobility and free volume

from the elastomer.










2e+8

PurePP
2e+8 / \ ................... (95:5) PP:8407
-/ (90:10) PP:8407
2e+8 ..-..-- (80:20) PP:8407
\- (70:30) PP:8407



Sle+8 A


1e+8 -
\\..
8e+7 -

6e+7 -

4e+7 -

2e+7
-100 -80 -60 -40 -20 0 20 40 60 80 100
Temperature (C)
Figure 3-8: Loss modulus vs. temperature for PP:8407 physical blends

It should be noted that the Tg of a polymer is commonly recorded as either the

onset or peak temperature on either E" or Tan6 graphs. The values reported in this

section are peak temperatures. Differential scanning calorimetry (DSC) is another

method to measure Tg and has shown that the Tg of 8407 is about -500C and PP about -

3C. This is concurrent with the onset temperatures for the 0 relaxations in both E" and

Tan6 graphs.

3.3.4 Crystallinity

The thermal transitions and crystalline character of the blends have been

characterized by DSC. This method measures heat flow into (endotherm) and out of

(exotherm) the sample in relation to a reference at varying temperatures. For a first order

thermodynamic transition like the melting temperature, there is a discontinuity in specific









volume vs. temperature and the DSC thermogram will show an endothermic peak

representing the melting of crystalline material in the sample.

From Figure 3-9 and Table 3-4, the a-crystalline phase of isotactic polypropylene

(monoclinic bravais lattice) can be identified as a peak at approximately 1650C. The

elastomer has a broad melting endotherm from about 300C to 800C, signifying that a

small degree of fringed micelle material is present [56, 64]. An accurate measure of 8407

% crystallinity is not available due to the broad melting range of the copolymer.


-4000

-5000

-6000

$ -7000

S -8000 -
w Pure PP
0................... (95:5) PP:8407
S-9000 ------ (90:10) PP:8407
--- (80:20) PP:8407
-10000 (70:30) PP:8407 7/i
----- Pure 8407

-11000

-12000 ...
40 55 70 85 100 115 130 145 160 175

Temperature (C)
Figure 3-9: DSC melting endotherm of both pure PP and 8407 as well as blends of the
two polymers.

The overall % crystallinity of the blends decreases with increasing elastomer

content (Figure 3-10) because of the hindrance effect of the melt for the arrangement of

the crystallizable chains of PP [128, 190, 264, 307-309]. There is only a slight decrease

in the % crystallinity of the PP phase with increasing elastomer concentration, which










indicates that chain alignment and order is not significantly hindered by large pockets of

immiscible material. The heat of melting per unit of PP in the blends is independent of

elastomer content, signifying that no co-crystallization is taking place [103, 263, 310].


--- Total % Crystallinity
--- PP % Crystallinity


65 70 75 80 85 90
PP Concentration in Blend


95 100 105


Figure 3-10: Percent crystallinity of the physical blends and pure PP as a function of PP
concentration.

Table 3-4: DSC endothermic data comparing pure PP and 8407 to physical blends of PP
and 8407. Standard deviation for pure PP is from an average of four samples.
Ratio of a Phase Melting Enthalpy of
PP:8407 Peak (C) Tm Onset (C) Melting (J/g) % Crystallinity
100:0 165.9 + 0.4 154.5 + 0.5 93.9 + 0.7 45.4 + 0.4
95:5 165.5 154.8 87 42
90:10 165.2 152.9 82.3 39.8
80:20 164.7 153.9 72.6 36.6
70:30 164.4 152.5 62.5 30.2
0:100 64.5--

Addition of elastomer is also known to decrease the onset of melting and peak

melting temperatures which may be due to the localization of elastomer in the intra-

spherulitic regions and disturbance of spherulite regularity [128, 255, 258, 264, 294, 308

310]. Also, smaller spherulties have a lower heat capacity, so the melting range of the

polymer blend should shift to lower temperatures [255]. When PP crystallizes in the










presence of amorphous, low molar mass, or immiscible material, the second component

can either be incorporated into the growing spherulite as occlusions (no effect on growth

rate) or rejected into the interspherulitic regions as deformed domains (considerable

depression in growth rate) [100, 236, 255, 257, 260, 263, 297, 307, 311-313].

Similar to several other references, elastomer commonly acts as a nucleating agent

for the PP and HDPE, therefore increasing crystallization temperature (T,) and

decreasing the average spherulite diameter [17, 23, 24, 103, 241, 255-258, 294]. This is

exemplified in Figure 3-11 and Table 3-5. The decrease in spherulite size is an indication

of increasing rate of nucleation, likely due to enhanced mobility of PP segments [100,

310]. This reduction also reduces the inhomogeneity of the sample and thus leads to

increased impact strength and elongation to break [259]. There may be local defects

within the spherulite at high elastomer concentrations which lead to weak spots or holes,

thereby reducing impact energy.


16000
Pure PP
14000 -(95:5) PP:8407
----- (90:10) PP:8407
12000 -- (80:20) PP:8407
-200 (70:30) PP:8407
S Pure 8407
10000 -i
E
8000 -

u 6000 -

"I .
4000 \

2000 --

0 10 20 30 40 50 60 70 80 90 100 110 120 130
Temperature (C)
Figure 3-11: DSC cooling exotherm of both pure PP and 8407 as well as blends of the
two polymers.









Table 3-5: DSC exothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10 B.
An average of 3 runs were performed on 90:10_B.
Ratio of Peak Temperature of Enthalpy of
PP:8407 Crystallization (C) Crystallization (J/g)
100:0 109.6 + 0.9 -90.7 + 0.6
95:5 114.9 -85.8
90:10 115.2 -82.7
80:20 116.1 -72.1
70:30 116.1 -60.6
0:100 43.6 -34.1


3.4 Conclusions

By simply melt blending a low molar mass grade of ethylene-octene copolymer, a

modest jump in impact strength can be seen up to 20 wt% elastomer, above which the

material acts in a brittle manner. This is directly related to the morphology of the blends,

where elastomer domains increase in size and change in shape from spherical to

ellipsoidal. The coalescence of the particles reduces their overall surface area so as to

limit interfacial interaction with PP. Melt flow index increases with increasing elastomer

content because the elastomer MFI is over 30 times that of the matrix phase. This low

molar mass elastomer affects both the crystalline and amorphous phases which in turn

deleteriously affect stress-strain performance.

Viscoelastic analysis confirms the thought that PP and 8407 are largely immiscible

phases. The low temperature peak of 8407 signifies chain mobility at low temperatures,

which is a good indication of impact strength performance. The sheer speed of the

impact test renders pure PP brittle but the blends ductile because elastomer chains are

able to relax and diffuse in response to the high stress before breaking. The crystallinity

of the blend is reduced with addition of 8407, but the % crystallinity of PP remains the

same regardless of the amount of elastomer. Also, the crystallization temperature






73


increases with increasing elastomer content because of the heterogeneous nucleation

effect and enhanced molecular mobility.














CHAPTER 4
TOUGHENED POLYPROPYLENE BASED ALLOYS

4.1 Introduction

The focus of this dissertation is based on the results obtained in this chapter, which

elaborates on the fundamental differences between PP alloys and physical blends. It will

be shown that modification on the nano-scale is necessary to achieve performance

enhancements previously unattainable. Many different toughening mechanisms are

thought to be involved in improving the impact strength of PP via alloying with an

elastomer. This chapter will tie in experimental results to many references and will

shown that a novel material has indeed been created. When modifying a polymer via

reactive blending, the resulting mechanical properties depend on a foundation of

controlled morphology, rheology, crystallinity, and chemical structure. A thorough

explanation of the alloy's impact strength and stress strain properties will be followed by

morphological, chemical, theological, and crystallographic characterization.

4.2 Experimental

4.2.1 Materials

Table 4-1 gives a list of pertinent ethylene-1-octene copolymers (EOCs) produced

by Dupont Dow elastomers under the tradename ENGAGE, but the grade of interest is

8407 [48]. Isotactic polypropylene homopolymer was supplied by Equistar Chemical

(grade PP 31S07A) and is contact translucent. All polymers were received in pellet form.

Polyisoprene (NATSYN) was donated by Goodyear. The peroxide and monomers used

in this study were reagent grade chemicals (structures are shown in Table 4-2). The









Table 4-1: ENGAGE product data table.
ENGAGE Grade (decreasing 8842 8407 8200 8401 8402
comonomer content)
Comonomer Content wt% 13C 45 40 38 31 22
NMR/FTIR
Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902
Melt Index, dg/min ASTM D- 1.0 30 5.0 30 30
1238, 190C, 2.16kg
Mooney Viscosity ASTM D- 26 < 5 8 < 5 < 5
1646 ML 1 + 4 at 121C
Durometer Hardness, Shore A 50 72 75 85 94
ASTM D-2240
DSC melting Peak, oC Rate: 33 60 60 78 98
10C/min
Glass Transition Temp, oC DSC -61 -57 -56 -51 -44
inflection point
Flexural Modulus, MPA ASTM 3.5 12.1 12.1 25.8 69.9
D-790, 2% Secant
Ultimate Tensile Strength, MPa 2.1 3.3 6.9 6.4 12.9
ASTM D-638, 508mm/min
Ultimate Elongation, % ASTM 975 >1000 >1000 950 790
D-638, 508mm/min

monomers were purified by passing through an activated alumina column before use.

Styrene monomer, inhibited by 10-15 ppm t-butyl catechol, was purchased from Fisher.

The initiator, 2,5dimethyl-2,5-di-(t-butylperoxy) hexane, was purchased from Atofina

under the tradename Lupersol 101. Diethyleneglycol diacrylate (DEGDA), inhibited by

80 ppm Hq and 120 ppm MEHQ, and trimethylolpropane triacrylate (TMPTA), inhibited

by 125 ppm HQ and 175 ppm MEHQ, were graciously donated by Sartomer, an Atofina

company. Irganox B215 was purchased from Ciba specialty chemicals.

Table 4-2: Structures of reactive materials of interest
Name Lupersol 101 DEGDA
Structure CHI CH,
\o o
(CH3)3C-O-C-CCH-CH2-0-0-C(CH3)3 (- H-O- -CH -O-2 -H
CH3 CH3 "










Table 4-2 Continued
Name Styrene TMPTA

Structure ?
H .1o--C--
CH.:=CH-R-O -rC-CHI--CH,

0


4.2.2 Methods

4.2.2.1 Processing

All polymers were dried in an air circulating oven at 400C for 24 hours prior to

compounding. Before processing, the resins were premixed by hand for about 10

minutes. Monomer/initiator mixtures were magnetically stirred for 10 minutes before

processing and a choice amount of the mixture was added to the dry polymer pellets

before processing.

The blending was carried out in a 34 mm non-intermeshing, co-rotating twin screw

extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D ratio of

39. The temperature of the extruder was regulated by electrical resistance and water

circulation in the barrels. The dried, pre-mixed resins were then introduced into the

extruder from the hopper of the extruder at 60 g/min through a screw driven dry material

feeder (Accu Rate, Inc). A Zenith pump controlled the rate of monomer/initiator solution

addition into the extruder. The screw speed, unless otherwise noted was 150 rpm.

Devolatilization was carried out by a vacuum pump, VPS-10A, Brooks Equipment

Company. This was placed near the die and created a pressure of about 15 in Hg. The

extruder was always starved to feed. Figure 4-1 is a schematic of the extruder along with

a typical temperature profile. After compounding, the resulting strands which exit the die










are quenched in a water bath, pelletized, and dried in a vacuum oven at 1000C, 28 in Hg

for 24 hours.

Zenith Pump
Power


eeder
Control Panel

Reactive Species Power 340 V
vacuum| |
Pelletizer Water Bath



ater in water in

Water out
Water out

Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed)
195C 205C 210C 210C 200C 190C 180C 165C
Figure 4-1: Schematic drawing of the reactive twin screw extruder and a common
temperature profile.

4.2.2.2 Mechanical properties

In order to measure the strength of the materials at very high testing rates, a

notched Izod impact test was performed according to ASTM D256 standards. The pellets

were placed in a mold with 6 slots, each measuring 0.5x0.5x2.5in3. The mold is put in a

Carver press (Fred S. Carver, Inc.) at 2000C and after the material melts, pressed up to

5000 psi. After waiting for 5-10 minutes, the pressure was slowly increased up to 10,000

psi. Five minutes later, the heat was turned off and the sample was let to cool down to

room temperature at about 1.50C/min. The bars were then taken out and notched with a

Testing Machines, Inc. (TMI) notching machine. Before testing, they were conditioned

at room temperature for 24 hours and a 30 ft-lb hammer was used with test method A on

a TMI Izod impact tester. At least 5 bars were broken and impact strength is recorded

regardless of full or partial break.









For stress-strain measurements, dried pellets were placed in a mold measuring 15

cm2 x 1 mm thick. The mold was put into the Carver press at 2000C and after the

material melts, pressed up to 5000 psi. After a 5-10 minute wait, the sample was slowly

pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath.

Specimens were tested according to ASTM D638 standards. Type V specimens were

punched out of the compression molded sheet with a die, measuring 1 mm thickness, 2.95

mm gauge width and 9.5 mm gauge length. Five samples were tested after conditioning

at room temperature for 48 hours. The machine used to test the samples was an MTS

Model 1120 Instron, using a 1000 lb load cell at a test speed of 12.7 mm/min

A Seiko DMS220 interfaced with a Seiko Rheostation model SDM/5600H was

used to test dynamic mechanical specimens. Testing was conducted from -1200C to

1500C at a heating rate of 50C/min in a dry nitrogen atmosphere maintained at an

approximate flow rate of 100 mL min-. Rectangular samples (20xl0xlmm3) were cut

from the compression molded sheet and tested in bending mode at a frequency of 1Hz.

4.2.2.3 Morphological characterization

Scanning electron microscopy (SEM) was performed on a JEOL 6335F Field

Emission SEM. The microscope was kept under vacuum at 1x105 Pa, with an

accelerating voltage of 5 kV and secondary electron detection at various magnifications.

Two different etching procedures were performed for better phase contrast one using

fractured impact bars and the other using compression molded films. The etching was

done to remove amorphous (or elastomeric) material. Impact bar samples were etched by

the following procedure: A notched impact bar was immersed in liquid nitrogen for 10

minutes and immediately fractured using a TMI impact tester. A section of the

cryofractured surface (2 cm thick) was immersed in xylene (purchased from Fisher




Full Text

PAGE 1

HIGH IMPACT STRENGTH POLYMERS HAVING NOVEL NANO-STRUCTURES PRODUCED VIA REACTIVE EXTRUSION By NATHAN FRASER TORTORELLA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Nathan Fraser Tortorella

PAGE 3

This dissertation is dedicated to my wife, Michelle, and my daughter, Katelynn.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my advisor and ment or, Professor Charles L. Beatty. He fostered my creativity and desire to succeed over the last severa l years and the KISS principle is forever instilled in me. My committee members have been an invaluable resource to me, namely Drs. Jack Mechol sky, Abbas Zaman, Hassan El-Shall, and Ken Wagener. Their guidance on both a personal and professional level helped me succeed as a graduate student. Funding for this res earch was graciously provided by Proctor and Gamble. Chuck Yeazell and colleagues at P&G were extremely helpful in furthering the development of this new material. My wife Michelle has been unwavering in support, love, and understanding and our beautiful daughter Katelynn is an inspiration. Stevan and Donna, my parents, and Harry and Carol, Michelles parents, deserve to be recognized for their encouragement and love through thick and thin. I would like to thank my sisters, Stevany, Emily, and Danica, for being there when I needed a quick laugh or to relax. Many people have helped me throughout th e years to eventually complete the dissertation. In the Beatty research gr oup, Ajit Bhaskar, Dr. Kunal Shah, and Xiosong Huang were outstanding friends and colleagues. Thanks go to Mike Tollon for the SEM pictures, Dr. Valentine Craciun for XRD, Dr. Laurie Gower and her students for optical microscopy, James Leonard in the Wagener re search group for GPC, and Phil McCartney at Virginia Tech for TEM photos. Dr. T ony Brennan, Dr. Clay Bohn, Dr. Leslie Wilson, and Michelle Carman were extremely he lpful in training and support on several

PAGE 5

v instruments in the polymer characterization lab. Lastly, I would like to acknowledge the staff in the Materials Science and Engineeri ng Department for their help and dedication over the last eight years.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...............................................................................................................x LIST OF FIGURES.........................................................................................................xiii ABSTRACT.....................................................................................................................xxi CHAPTER 1 GENERAL INTRODUCTION....................................................................................1 1.1 Introduction.............................................................................................................1 1.2 Chapter Summaries.................................................................................................4 2 BACKGROUND AND LI TERATURE REVIEW......................................................6 2.1 Polymers of Interest................................................................................................6 2.1.1 Isotactic Polypropylene................................................................................6 2.1.2 Ethylene-1-Octene Copolymer.....................................................................7 2.2 Processing...............................................................................................................9 2.2.1 Reactive Twin Screw Extrusion...................................................................9 2.2.2 Morphology Development in an Extr uder Dispersive and Dissipative Mixing..............................................................................................................11 2.2.3 Melting and Droplet Breakup Mechanisms................................................13 2.2.4 Morphology Development Dependency on Rate of Reaction....................16 2.2.5 Viscosity and Reaction Eff ects on Morphology Development..................18 2.3 Toughening of Polymers.......................................................................................19 2.3.1 Origins of Polymer Toughening.................................................................19 2.3.2 Elastomer/Rubber Toughened Blends........................................................20 2.4 Free Radical Reactions.........................................................................................23 2.4.1 Initiator Decomposition..............................................................................23 2.4.2 What Happens After Pe roxide Decomposition?.........................................26 2.4.3 Polymerization............................................................................................26 2.4.4 Copolymerization.......................................................................................28 2.4.5 Multifunctional Monomer..........................................................................30 2.4.6 Hydrogen Abstraction................................................................................31

PAGE 7

vii 2.4.7 Polyethylene Crosslinking Reactions.........................................................32 2.4.8 Degradation of Polypropylene....................................................................33 2.5 Previous Efforts to Reduce/Prev ent Degradation of Polypropylene....................35 2.5.1 Free Radical Grafting of Polymers.............................................................35 2.5.2 Solid State Grafting....................................................................................36 2.5.3 Reactions in the Polymer Melt...................................................................37 2.5.4 Fundamentals of Free Radical Grafting.....................................................39 2.5.5 Melt Grafting Monomers onto Polyolefins................................................41 2.6 Melt Grafting and In-situ Compatibilization........................................................43 2.7 Conclusions...........................................................................................................47 3 PHYSICAL BLENDING OF AN IMPACT MODIFIER WITH POLYPROPYLENE48 3.1 Introduction...........................................................................................................48 3.2 Experimental.........................................................................................................48 3.2.1 Materials.....................................................................................................48 3.2.2 Methods......................................................................................................48 3.2.2.1 Processing.........................................................................................48 3.2.2.2 Mechanical properties......................................................................50 3.2.2.3 Morphology......................................................................................51 3.2.2.4 Thermal analysis and rheology........................................................52 3.3 Results and Discussion.........................................................................................53 3.3.1 Mechanical and Rheological Properties.....................................................53 3.3.2 Morphology................................................................................................59 3.3.3 Viscoelasticity............................................................................................62 3.3.4 Crystallinity................................................................................................68 3.4 Conclusions...........................................................................................................72 4 TOUGHENED POLYPROPYLENE BASED ALLOYS..........................................74 4.1 Introduction...........................................................................................................74 4.2 Experimental.........................................................................................................74 4.2.1 Materials.....................................................................................................74 4.2.2 Methods......................................................................................................76 4.2.2.1 Processing.........................................................................................76 4.2.2.2 Mechanical properties......................................................................77 4.2.2.3 Morphological characterization........................................................78 4.2.2.4 Chemical composition a nd molecular structure...............................81 4.2.2.5 Thermal analysis and rheology........................................................82 4.3 Results and Discussion.........................................................................................84 4.3.1 Notched Izod Impact Analysis...................................................................86 4.3.2 Stress-Strain Behavior................................................................................91 4.3.3 Grafting onto Polyolefins...........................................................................97 4.3.4 Morphology..............................................................................................101 4.3.5 Possible Crosslinking of the System........................................................111 4.3.6 Rheological Properties..............................................................................112 4.3.7 Crystallinity and Crystallization...............................................................116

PAGE 8

viii 4.3.8 Viscoelasticity..........................................................................................138 4.4 Conclusions.........................................................................................................145 5 THE EFFECT OF PROCESSING C ONDITIONS ON ALLOY PROPERTIES....147 5.1 Introduction.........................................................................................................147 5.2 Experimental.......................................................................................................147 5.2.1 Materials...................................................................................................147 5.2.2 Methods....................................................................................................149 5.2.2.1 Processing.......................................................................................149 5.2.2.2 Mechanical properties....................................................................150 5.2.2.3 Chemical composition....................................................................151 5.2.2.4 Rheology........................................................................................152 5.3 Results and Discussion.......................................................................................152 5.3.1 Effect of Processing Temperature............................................................152 5.3.2 Effect of Screw Speed..............................................................................155 5.3.3 Effect of Initiator Concentration..............................................................158 5.3.4 Effect of Multifunctional Monomer Concentration.................................160 5.3.5 Effect of Styrene Concentratio n with DEGDA as Multifunctional Monomer........................................................................................................163 5.3.6 Effect of Styrene Concentratio n with TMPTA as Multifunctional Monomer........................................................................................................165 5.4 Conclusions.........................................................................................................167 6 CONTROLLING ALLOY PERFORMANCE BY VARYING ELASTOMER PROPERTIES...........................................................................................................168 6.1 Introduction.........................................................................................................168 6.2 Experimental.......................................................................................................168 6.2.1 Materials...................................................................................................168 6.2.2 Methods....................................................................................................170 6.2.2.1 Processing.......................................................................................170 6.2.2.2 Mechanical properties....................................................................171 6.2.2.3 Chemical composition a nd molecular structure.............................172 6.2.2.4 Thermal analysis and rheology......................................................174 6.3 Results and Discussion.......................................................................................174 6.3.1 Effect of Elastomer Density on Alloy Performance.................................174 6.3.1.1 Physical blends...............................................................................175 6.3.1.2 Alloys.............................................................................................183 6.3.2 Effect of Elastomer Molecula r Weight on Alloy Performance................190 6.3.2.1 Physical blends...............................................................................193 6.3.2.2 Alloys.............................................................................................200 6.3.3 Supercritical CO2 as a Possible Route to Improve Grafting Efficiency...210 6.4 Conclusions.........................................................................................................211

PAGE 9

ix 7 REACTIVE EXTRUSION OF HI GH DENSITY POLYETHYLENE...................213 7.1 Introduction.........................................................................................................213 7.2 Experimental.......................................................................................................213 7.2.1 Materials...................................................................................................213 7.2.2 Methods....................................................................................................215 7.2.2.1 Processing.......................................................................................215 7.2.2.2 Mechanical properties....................................................................216 7.2.2.3 Design of experiments....................................................................217 7.3 Results and Discussion.......................................................................................218 7.3.1 Mechanical Properties..............................................................................218 7.3.2 Design of Experiments Results................................................................221 7.3.2.1 Impact strength...............................................................................221 7.3.2.2 Elastic modulus..............................................................................225 7.3.2.3 Yield strength.................................................................................228 7.3.2.4 Elongation at break.........................................................................231 7.4 Conclusions.........................................................................................................234 8 CONCLUSIONS AND FUTURE WORK...............................................................236 8.1 Summary and Conclusions..........................................................................236 8.2 Future Work.................................................................................................240 APPENDIX A CALIBRATION CURVE FO R ABSOLUTE STYRENE CO NCENTRATIONS IN REACTIVE BLENDS..............................................................................................244 B STRESS-STRAIN GRAPHS AND STATISTICS...................................................246 C FTIR GRAPHS.........................................................................................................256 D IMAGE ANALYSIS................................................................................................262 E ALIASED TERMS FROM CHAPTER 7................................................................265 LIST OF REFERENCES.................................................................................................266 BIOGRAPHICAL SKETCH...........................................................................................288

PAGE 10

x LIST OF TABLES Table page 2-1 The mechanisms of main reac tions in the grafting process......................................23 2-2 Characteristics of Lupersol 101................................................................................24 2-3 Half life of Lupersol 101 based on temperature.......................................................25 2-4 Typical free radical ki netic values in solution.........................................................27 2-5 Copolymerization constants for st yrene and ethyl acrylate monomers....................29 2-6 Structural and physical informati on about the monomers of interest......................29 3-1 ENGAGE product data table.................................................................................49 3-2 DSC consecutive heating/cooling cycles.................................................................53 3-3 Image analysis averages taken from etched SEM images of PP:8407 physical blends. At least three images were analyzed from Figures 3(a) thru (e), with data compiled in Appendix D..........................................................................................61 3-4 DSC endothermic data comparing pure PP and 8407 to physical blends of PP and 8407. Standard deviation for pure PP is from an average of four samples.............70 3-5 DSC exothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10_B. An average of 3 runs were performed on 90:10_B........................................................72 4-1 ENGAGE product data table.................................................................................75 4-2 Structures of reactive materials of interest..............................................................75 4-3 PLM consecutive heating/cooling cycles................................................................81 4-4 DSC consecutive heating/cooling cycles................................................................83 4-5 Identification of formulations..................................................................................86 4-6 Image analysis of etched SEM surfaces of several blends and alloys. This information is gathered from Appendix D, Figure 4-16, and Figure 3-5. Ligament thickness is calculated using th e number average diameter...................................102

PAGE 11

xi 4-7 Melt behavior of PP, PP-8407 physical blends, and PP-8407 alloys.....................115 4-8 DSC endothermic data comparing pure PP to 90:10_0, 90:10_A, and 90:10_B. An average of three runs were performed on 90:10_B................................................127 4-9 DSC exothermic data compari ng pure PP to 90:10_0, 90:10_A, and 90:10_B.....131 5-1 ENGAGE product data table...............................................................................148 5-2 Structures of reactive materials of interest.............................................................148 5-3 Three different temperature prof iles for the extrusion of 80:20_C........................152 5-4 Stress-strain beha vior of 80:20_C as a f unction of temperature............................154 5-5 Screw speed relationship to residence time............................................................156 5-6 Stress-strain beha vior of 80:20_C as a f unction of screw speed............................158 5-7 Stress-strain behavior of PP:8407 alloys at a ratio of 90:10 as a function of initiator concentration..........................................................................................................160 5-8 Stress-strain behavior of PP:8407 alloys at a ratio of 90:10 as a function of multifunctional monomer concentration................................................................163 5-9 Stress-strain performance of PP:8407 all oys at a ratio of 90:10 as a function of styrene concentration..............................................................................................165 6-1 ENGAGE product data table...............................................................................169 6-2 Structures of reactive materials of interest.............................................................170 6-3 DSC consecutive heating/cooling cycles...............................................................174 6-4 Identification of Formulati ons in relation to Figure 6-2........................................175 6-5 Stress-strain data of 8407_0, 8401_0, and 8402_0................................................177 6-6 DSC endothermic data of 8407_0, 8401_0, and 8402_0.......................................179 6-7 DSC exothermic data of 8407_0, 8401_0, and 8402_0.........................................181 6-8 Stress-strain data of PP:ENGAGE all oys as a function of elastomer density........186 6-9 DSC endothermic data of 8407_A, 8407_B, 8402_A, and 8402_B......................188 6-10 DSC exothermic data of 8407_A, 8407_B, 8402_A, and 8402_B........................188 6-11 DSC endothermic data of 002_0, 8407_0, 8200_0, and 8842_0...........................197

PAGE 12

xii 6-12 DSC exothermic data of 002_0, 8407_0, 8200_0, and 8842_0.............................199 6-13 Stress-strain behavior of 002_B, 8407_B, 8200_B, and 8842_B..........................205 6-14 DSC endothermic data of 8407_A, 8407_B, 8842_A, and 8842_B......................208 6-15 DSC exothermic data of 8407_A, 8407_B, and 8842_A, and 8842_B..................209 7-1 ENGAGE product data table...............................................................................214 7-2 Structures of reactive materials of interest.............................................................214 7-3 Factors of interest, coded variables, and levels for the impact modification of HDPE.....................................................................................................................217 7-4 Basic fractional factorial design us ing coded variables from Table 7-3................218 7-5 Impact strength, elastic modulus, yield stress, and elongation at break results from the fractional factorial design crea ted for the modification of HDPE....................220 7-6 Analysis of variance table (partial sum of squares) for impact strength................223 7-7 Analysis of variance table (partial sum of squares) for elastic modulus................226 7-8 Analysis of variance table (parti al sum of squares) for yield stress.......................230 7-9 Analysis of variance table (partial sum of squares) for elongation at break..........232 B-1 Actual stress-strain values with standa rd deviations from figures in both Chapter 3 and Chapter 4....................................................................... ..................................246 B-2 Stress-strain propertie s of PP:ENGAGE blends and alloys as a function of elastomer melt flow index......................................................................................253 E-1 Aliased terms from the fractional factorial design given in Chapter 7..................265

PAGE 13

xiii LIST OF FIGURES Figure page 2-1 The phase inversion mechanism proposed by Shih.................................................14 2-2 Morphology development of binary polymer blends proposed by Scott and Macosko...................................................................................................................15 2-3 Chemical structure of Lupersol 101.........................................................................24 2-4 Schematic of H abstrac tion from PP and subsequent -chain scission....................34 3-1 Schematic drawing of the reactive twin screw extruder and a common temperature profile.......................................................................................................................50 3-2 Effect elastomer concentration on room temperature notched impact strength and melt flow index.........................................................................................................54 3-3 Stress-strain behavior of PP:8407 physical blends..................................................56 3-4 Stress-strain properties of PP-8407 blends as a function of elastomer content.......57 3-5 SEM images of etched, cryo-fracture d surfaces of PP:8407 physical blends as a function of elastomer con centration. (a) Virgin PP at 2,000X, (b) 95:5_0 at 10,000X, (c) 90:10_0 at 10,000X, (d) 80:20_0 at 10,000X, and (e) 70:30_0 at 2,500X. The bar markers for (a) and (e) = 10 m, and for (b), (c), and (d) = 1 m.60 3-6 Storage modulus vs. te mperature for PP:8407 Blends.............................................63 3-7 Tan vs. temperature for PP:8407 blends from -120C to 120C. Insert represents Tan vs. temperature from -60C to 60C...............................................................65 3-8 Loss modulus vs. temperature for PP:8407 physical blends...................................68 3-9 DSC melting endotherm of both pure PP and 8407 as well as blends of the two polymers...................................................................................................................69 3-10 Percent crystallinity of the physical blends and pure PP as a function of PP concentration............................................................................................................70 3-11 DSC cooling exotherm of both pure PP and 8407 as well as blends of the two polymers...................................................................................................................71

PAGE 14

xiv 4-1 Schematic drawing of the reactive twin screw extruder and a common temperature profile.......................................................................................................................77 4-2 FTIR image of PP:polyisoprene reactive blend.......................................................85 4-3 Representation of sample reference code................................................................85 4-4 Room temperature notched Izod impact strength of PP blends and alloys. Reference material is virg in PP at 0.99 ft-lbs/in......................................................87 4-5 Izod Impact test specimens post-fracture Left Pure PP, Middle 90:10_0, Right 90:10_B.................................................................................................................88 4-6 SEM image showing the tip of an arrest ed crack from a room temperature notched Izod impact test of 90:10_0. Insert is a magnified image of crack tip....................89 4-7 SEM image showing the tip of an arrest ed crack from a room temperature notched Izod impact test of 90:10_B.....................................................................................90 4-8 SEM image of 80:20_C (left) and Pure PP (right) fractured at room temperature without etching. Each image is located at the center of the impact bar, with magnification = 2,000X and marker bar = 10 m.....................................................91 4-9 Stress-strain properties of 95:5 alloys compared to the 95:5 physical blend...........92 4-10 Stress-strain properties of 90:10 alloys as compared to the 90:10 physical blend...93 4-11 Stress-strain properties of 80:20 alloys as compared to the 80:20 physical blend...93 4-12 Stress-strain properties of 70:30 alloys as compared to the 70:30 physical blend...94 4-13 FTIR image of a typical PP:8407 all oy containing styrene and multifunctional acrylate.....................................................................................................................98 4-14 Styrene grafting efficiency at various 8407 concentrations both with and without multifunctional monomer.........................................................................................99 4-15 Molecular weight averages for pure 8407 and three grafte d 8407 materials.........101 4-16 SEM images of (a) 95:5_B at 10,000X (b) 90:10_B at 10,000X, (c) 80:20_B at 10,000X, and (d) 70:30_B at 5,000X. Each bar marker = 1 m............................103 4-17 Matrix ligament thickness of PP blends and alloys as a function of volume % of 8407........................................................................................................................105 4-18 Room temperature notched Izod impact st rength as a function of matrix ligament thickness for blends and alloys at various volume % 8407....................................105

PAGE 15

xv 4-19 TEM bright field images of (a) vi rgin PP and (b) 90:10_0 stained with RuO4 at a magnification of 63,500X. Bar marker on insert = 100nm...................................107 4-20 TEM image of 90:10_A at 63,500X All marker bars = 100nm...........................108 4-21 TEM image of 90:10_B at 63,500X All marker bars = 100nm...........................109 4-22 TEM images of 80:20_A (left) and 80:20_B (right) at 63,500X...........................110 4-23 TGA graphical comparison of a physical blend (90:10_0) and alloy (90:10_B)...112 4-24 Melt flow index of physical blends a nd alloys as a function of 8407 content.......114 4-25 XRD pattern of pure PP, 90:10_0, 90:10_A, and 90:10_B....................................118 4-26 Alignment of lamellae within spherulites of -PP (left) and -PP (right) from reference 278..........................................................................................................121 4-27 Etched lamellar morphology of isotactic polypropylene of well defined spherulite at 15kX magnification (a), bar marker = 2 m and periphery of a spherulite at 20kX magnification (b), bar marker = 2 m.....................................................................123 4-28 Etched lamellar morphology of 90:10_0 at 10.5kX magnification, with the bar marker = 5 m.........................................................................................................124 4-29 Etched lamellar morphology of 90:10_0 at 20kX magnification, with the bar marker = 2 m.........................................................................................................125 4-30 DSC heating traces of a physical blend (90:10_0) and al loys (90:10_A and 90:10_B).................................................................................................................126 4-31 DSC heating traces of a physical bl end (80:20_0) and alloys (80:20_A, 80:20_B, and 80:20_C without styrene)................................................................................129 4-32 DSC cooling traces of a physical blend (90:10_0) and al loys (90:10_A and 90:10_B).................................................................................................................130 4-33 Polarized optical images of virgin PP cooled from the melt. (a) = 115.5C at 50X, (b) = 113.5C at 50X, (c) = 110C at 50X, and (d) = 25C at 20X magnification.133 4-34 Polarized optical images of 90:10_0 cooled from the melt. (a) = 118C at 50X, (b) = 114.5C at 50X, (c) = 105C at 50X, and (d) = 25C at 20X magnification. A bubble (artifact) appears in the lower left hand of (b) and (c) and along the edges of (d)...........................................................................................................................134 4-35 Polarized optical images of 90:10_A cool ed from the melt. (a) = 128C at 50X, (b) = 124C at 50X, (c) = 122C at 50X, a nd (d) = 25C at 20X magnification.........135

PAGE 16

xvi 4-36 Polarized optical images of 90:10_B cool ed from the melt. (a) = 135C at 50X, (b) = 133C at 50X, (c) = 130C at 50X, a nd (d) = 25C at 20X magnification.........137 4-37 Storage modulus (E') and Tan comparison of 95:5_0 and 95:5_B as a function of temperature.............................................................................................................139 4-38 Storage modulus (E') and Tan comparison of 90:10_0, 90:10_A, and 90:10_B as a function of temperature..........................................................................................140 4-39 Tan comparison of 90:10_0, 90:10_A, and 90:10_B between -50C and 60C to show a magnified graph of 8407 and PP................................................................141 4-40 Storage modulus (E') and Tan comparison of 80:20_0 and 80:20_B as a function of temperature........................................................................................................143 4-41 Storage modulus (E') and Tan comparison of 70:30_0 and 70:30_B as a function of temperature........................................................................................................144 5-1 Schematic drawing of the reactive twin screw extruder and a common temperature profile.....................................................................................................................149 5-2 Effect of extruder barre l temperature on room temperature impact strength, melt flow index, and grafting efficiency of 80:20_C.....................................................153 5-3 Impact strength, MFI, and grafting effi ciency of 80:20_C as a function of screw speed.......................................................................................................................157 5-4 Impact strength, MFI, and grafting effici ency of 90:10_B as a function of initiator concentration..........................................................................................................159 5-5 Notched impact strength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 90:10 as a function of multifunctional monomer concentration.161 5-6 Notched impact strength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 90: 10 as a function of styren e monomer concentration............164 5-7 Notched impact strength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 80:20 as a f unction of styrene concentration............................166 6-1 Schematic drawing of the reactive twin screw extruder and a common temperature profile.....................................................................................................................171 6-2 Representation of sample reference code...............................................................175 6-3 Impact strength, melt flow index, and me lting temperature of physical blends as a function of the density of the copolymer...............................................................176

PAGE 17

xvii 6-4 Digital images of room temperature fract ured Izod impact bars as a function of elastomer density in the physical bl ends. From left to right: 8407_0, 8401_0, 8402_0....................................................................................................................177 6-5 Stress-strain performance of PP/elas tomer physical blends as a function of elastomer density....................................................................................................178 6-6 DSC melting endotherm of 8407_0, 8401_0, and 8402_0.....................................179 6-7 DSC cooling exotherm of PP/elastomer phys ical blends as a function of elastomer density. Insert is for the te mperature range or 20 to 100C..................................180 6-8 Dynamic mechanical analysis of 8407_0, 8401_0, and 8407_0. Insert is a magnified graph from -60C to 60C of the relaxation of PE.............................182 6-9 Room temperature impact strength, melt flow index, and grafting efficiency of 8407_B, 8401_B, and 8402_B...............................................................................183 6-10 Room temperature impact strength, melt flow index, and grafting efficiency of 8407_C, 8401_C, and 8402_C...............................................................................185 6-11 DSC melting endotherms of 8407_A 8407_B, 8402_A, and 8402_B ranging from 60C to 180C. Insert is a magnified graph of the melting peak of PP..............187 6-12 DSC cooling exotherms of PP/elastomer phys ical blends as a function of elastomer density. Insert is for the te mperature range of 20 to 100C..................................189 6-13 Dynamic Mechanical Analysis comparison of (a) 8402_0 and 8402_B, (b) 8401_0 and 8401_B, and (c) 8407_0 and 8407_B..............................................................190 6-14 Molecular weight averages of the pol yolefin elastomers of interest, ranked according to their melt flow index.........................................................................192 6-15 Polydispersity (molecular weight distri bution) of the polyolefin elastomers of interest....................................................................................................................193 6-16 MFI and room temperature notched Izod impact strength of 8842_0 (highest molecular weight), 8200_0, 8407_0, and 002_0 (lowest molecular weight).........194 6-17 Stress-strain performance of physical ble nds as a function of MFI of the copolymer.196 6-18 DSC melting endotherms of 002_0, 8407_0, 8200_0 and 8842_0........................198 6-19 DSC crystallization exotherms of 002_0, 8407_0, 8200_0, and 8842_0...............199 6-20 Impact Strength and melt flow index of 8842_A, 8200_A, 8407_A, and 002_A..201 6-21 Izod impact strength and melt fl ow index of 8842_C, 8200_C, 8407_C, and 002_C.202

PAGE 18

xviii 6-22 Izod impact strength and melt fl ow index of 8842_B, 8200_B, 8407_B, and 002_B.203 6-23 Dynamic mechanical behavior (E' and Tan ) of various alloys and blends. (a) 002_0 vs. 002_B, (b) 8407_0 vs. 8407_B, (c) 8200_0 vs. 8200_B, and (d) 8842_0 vs. 8842_B..............................................................................................................206 6-24 DSC melting endotherm of 8407_A, 8407_B, 8842_A, and 8842_B....................208 6-25 DSC crystallization exotherm of 8407_A, 8407_B, 8842_A, and 8842_B...........209 6-26 Effect of supercritical carbon dioxide at 1500 psi on the grafting efficiency of 8842_B. Injection was in zone 3...........................................................................211 7-1 Schematic drawing of the reactive twin screw extruder and a common temperature profile.....................................................................................................................215 7-2 Half normal plot showing, in coded va riables, the most si gnificant effects on impact strength. A = initiator concen tration and C = concentration of 8842........222 7-3 Normal plot of residuals (a) and outliers (b) show th e diagnostic results of the model for impact strength......................................................................................224 7-4 Cube graph of the effect of % initia tor, 8842 content, and % styrene on impact strength at constant screw sp eed, % DEGDA, and temperature............................225 7-5 Half normal plot showing, in coded variab les, the most significa nt effects on elastic modulus..................................................................................................................226 7-6 Normal plot of residuals (a) and outliers (b) show th e diagnostic results of the model for elastic modulus......................................................................................227 7-7 Cube graph of the effect of % initia tor, 8842 content, and % styrene on elastic modulus at constant screw spee d, % DEGDA, and temperature...........................228 7-8 Half normal plot showing, in coded va riables, the most si gnificant effects..........229 7-9 Normal plot of residuals (a) and outliers (b) show th e diagnostic results of the model for yield stress.............................................................................................230 7-10 Cube graph of the effect of % initiato r, 8842 content, and % styrene on yield stress at constant screw speed, % DEGDA, and temperature..........................................231 7-11 Half normal plot showing, in coded va riables, the most significant effects..........232 7-12 Normal plot of residuals (a) and outliers (b) show th e diagnostic results of the model for elastic modulus......................................................................................233 7-13 Cube graph of the effect of % initia tor, 8842 content, and % styrene on elongation at break at constant % styren e, % DEGDA, and temperature................................234

PAGE 19

xix 8-1 Schematic drawing of the likely free radi cal initiated processes during the reactive extrusion of PP, 8407, initiator, styr ene, and multifunctional monomer...............237 8-2 Interpretation of the effect of in-situ grafted polymeric chains at the PP-elastomer interface (blue circles = elastomer domains, black lines = grafted polymers). The physical blend (left) has no grafting and the alloy (right) ha s a high degree of grafting...................................................................................................................239 B-1 Stress-strain graph comparison of 95: 5 (PP:8407) physical blend and alloys.......247 B-2 Stress-strain graph comparison of 90: 10 (PP:8407) physical blend and alloys.....247 B-3 Stress-strain graph comparison of 80: 20 (PP:8407) physical blend and alloys.....248 B-4 Stress-strain graph comparison of 70: 30 (PP:8407) physical blend and alloys.....248 B-5 Stress-strain graph comparison of the effect of extruder ba rrel temperature in relation to 5-2 and Table 5-4. 1= low te mperature, 2=middle temperature, 3=high temperature.............................................................................................................249 B-6 Stress-strain graph comparison of the eff ect of extruder screw speed in relation to Table 5-6................................................................................................................249 B-7 Stress-strain graph comparison of the effect of initiator concentration in relation to Table 5-7................................................................................................................250 B-8 Stress-strain graph comparison of the e ffect of DEGDA concentr ation in relation to Table 5-8........................................................................................ ........................250 B-9 Stress-strain graph comparison of the eff ect of styrene concentration in relation to Table 5-9........................................................................................ ........................251 B-10 Stress-strain graph comparison of the e ffect of elastomer density in relation to Figure 6-5 and Table 6-5........................................................................................251 B-11 Stress-strain graph comparison of the e ffect of elastomer density in relation to Table 6-8................................................................................................................252 B-12 Stress-strain graph comparison of the e ffect of elastomer density in relation to Table 6-8................................................................................................................252 B-13 Stress-strain graph comparison of the e ffect of elastomer density in relation to Table 6-8................................................................................................................253 B-14 Stress-strain graph of physical blen ds of ENGAGE elastomers with PP as a function of elastomer melt flow index.................................................................. 254 B-15 Stress-strain graph comparison of the eff ect of elastomer MFI.............................254

PAGE 20

xx B-16 Stress-strain graph comparison of the eff ect of elastomer MFI............................. .255 C-1 FTIR graph comparison of 95:5_0, 95:5_A, 95:5_B, and 95:5_C......................... 256 C-2 FTIR graph comparison of 90:10_0, 90:10_A, 90:10_B, and 90:10_C.................257 C-3 FTIR graph comparison of 80:20_0, 80:20_A, 80:20_B, and 80:20_C.................257 C-4 FTIR graph comparison of 70:30_0, 70:30_A, 70:30_B, and 70:30_C.................258 C-5 FTIR graphs from Chapter 5 (Figure 5-2) of alloys processed at varying temperatures. 1 = low barrel temperature, 2 = middle barrel temperature, 3 = high barrel temperature.................................................................................................. 258 C-6 FTIR graphs from Chapter 5 (Figure 53) of alloys processed at varying screw speeds.....................................................................................................................259 C-7 FTIR graphs from Chapter 5 (Figure 5-4) of alloys processed at varying concentration of initiator........................................................................................259 C-8 FTIR graphs from Chapter 5 (Figure 5-5) of alloys processed at varying concentration of multifunctional monomer............................................................ 260 C-9 FTIR graphs from Chapter 5 (Figure 5-6) of alloys processed at varying concentration of styrene, with DEGDA as multifunctional monomer................... 260 C-10 FTIR graphs from Chapter 5 (Figure 5-7) of alloys processed at varying concentration of styrene, with TMPTA as multifunctional monomer................... 261 D-1 Histogram of average particle di ameters for (a) 70:30_0 and (b) 70:30_B........... 262 D-2 Histograms of particle roundness for (a) 70:30_0 and (b) 70:30_B...................... 262 D-3 Histogram of average particle di ameters for (a) 80:20_0 and (b) 80:20_B........... 263 D-4 Histograms of particle roundness for (a) 80:20_0 and (b) 80:20_B...................... 263 D-5 Histogram of average particle di ameters for (a) 90:10_0 and (b) 90:10_B........... 263 D-6 Histograms of particle roundness for (a) 90:10_0 and (b) 90:10_B...................... 264 D-7 Histogram of average particle di ameters for (a) 95:5_0 and (b) 95:5_B............... 264 D-8 Histograms of particle roundne ss for (a) 95:5_0 and (b) 95:5_B.......................... 264

PAGE 21

xxi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HIGH IMPACT STRENGTH POLYMERS HAVING NOVEL NANO-STRUCTURES PRODUCED VIA REACTIVE EXTRUSION By Nathan Fraser Tortorella December 2005 Chair: Charles L. Beatty Major Department: Materials Science and Engineering A major focus of scientists and engineers ove r the last century has been to increase the impact strength and therefore reduce the br ittleness of materials. By altering and adding energy absorption mechanisms, brittle failure can be averted. Isotactic polypropylene (PP) is the focus of this disser tation because it is an extremely low cost, high volume, versatile plastic but behaves in a brittle manner at or below room temperature or in a notched state. Early work on impact modification of polypropylene focused on blending energy-absorbing low densit y elastomers and rubbers. These binary blends all had a common problem an increa se in impact strength was paralleled by a significant decrease in both el astic modulus and yield stress. Reactive extrusion processing has allowed the in-situ compatibilization of isotactic polypropylene and metallocene-catalyzed ethyle ne-octene copolymers (EOCs). This process involves combining both the comonome r and vector fluid approaches to grafting polyolefins. Styrene monomer and a multifunctional acrylate monomer undergo

PAGE 22

xxii peroxide-induced copolymerization and grafti ng in the presence of both PP and EOC. This results in a phase separa ted alloy with an impact stre ngth over 13 times that of pure polypropylene and double that of the physical blend. There is also a significant improvement in stress-strain performance when comparing the alloys to physical blend counterparts. Many researchers have categorized the necessary components to toughening polypropylene as pertaining to the amorphous phase. The alloys described in this dissertation meet the criteria put forth by these researchers, namely low density, crystallinity, and modulus of the elastomer phase, sub-micr on particle diameter, close inter-particle distance, and a high degree of entanglements of both the PP matrix phase and EOC minor phase. But many people neglect to study the cr ystalline state of impact modified PP in conjunction with the amorphous phase. This work shows that the typical 10-100 m diameter spheruliti c structures found in pure PP ar e not present in the alloys. In fact, the spherulites are less than a micron in diameter are uniformly distributed throughout the sample, and crystallize at much higher temperatures. SEM images, when coupled with DSC and XRD, reveal the pres ence of a high number of small lamellar crystals composed of a unique highly dens e cross-hatched structure. Thus, impact strength and stiffness can be simultaneously improved by controlli ng the size and crosshatch density of the lamella r crystals and applying pha se transformation toughening concepts.

PAGE 23

1 CHAPTER 1 GENERAL INTRODUCTION 1.1 Introduction The field of materials science and engin eering has emerged as a premier discipline which encompasses the development, synthe sis, and processing of new materials. Calister describes materials science as inves tigating the relationships that exist between the structures and properties of materials, whereas materials engineering is based upon designing or engineering the structure of a material to produce a predetermined set of properties [1]. The structure-processing-prope rties relationship holds true for the creation of all advanced materials, with polymers, metals, ceramics, and electronic materials as the core classes. Polymers are the focus of the dissertation, which are essentially organic macromolecules chemically based on carbon, hydrogen, and other nonmetallic elements. Polymer blending is an economic process to create a material with a balance of properties that would otherwise not be possible [2-12]. If a material can be generated that will lower the cost while maintaining or improving performance of a particular product then the manufacturer must use it to remain competitive. Many engineering resins may lack chemical resistance, impact toughness, flame retardency, high temperature performance, or weatherability, which can be solved via blending with other engineering or commodity plastics. The de velopment of a new blend or compound from existing materials is generally more rapid than that of an entirely new polymer. Polymer blends can be characterized by their phase behavior as being either miscible or immiscible (immiscible blends having multiple amorphous phases) [10, 12-

PAGE 24

2 14]. A blends mechanical, thermal, and rh eological properties, amongst others, depend strongly on its state of miscibil ity. Nearly all polymer pair s are immiscible, forming twophased systems in which the interface is a source of weakness. Polymer-polymer immiscibility is a direct consequence of the high molecular weights of the polymer molecules. Blending two polymers is not thermodynamically favorable because the mixing of a relatively low number of molecu les leads to a positive enthalpy of mixing and low entropy of mixing [8, 15-18]. By simply blending two immiscible polymer s, the resulting material has improved properties but usually at the expense of anot her property. For example, adding rubbery or elastomeric particles to a polypropylene (PP) matrix increases impact strength but sacrifices yield strength and el astic modulus [19-24]. Previous routes to toughen PP have their drawbacks, such as the migration phenomenon with plastici zers and costly inreactor copolymerization of ethylene and propylene. In order to overcome these problems, many researchers have found that by controlling the size a nd distribution of the rubber particles, cohesive strength of the el astomer, the degree of physical entanglements in the system, and interfacial adhesion between the rubber and the matrix, all properties are vastly improved over the original blend. Physically blending an elastomer with a brittle semicrystalline polymer is not sufficient for overall property improvem ent, so a reactive extrusion process (compatibilization) is often applied to create novel polymeric alloys. The term alloy has been defined by the Polymer Technology Dictionary as a composition, or blend, which is based on two or more polymers, the propert ies of which are significantly better than would be expected from a simple blend [25] The system is typically phase separated

PAGE 25

3 with a certain degree of chemical bonding or grafting between phases. Compatibilizers are sometimes used to control the adhesion asp ect of these blends, which therefore results in finer dispersed phase morphology, uniform distribution of domains, better processability, enhanced mechanical propertie s, and increased thermal stability. They effectively act as high molecular weight surf actants by locating at the interface between the immiscible polymers [12, 14, 26-40]. Compatibilizers are traditionally thought of as block or graft copolymers which contain functional groups that may or may not react with the polymers present in the blend. Three drawbacks of these pre-made copolymers are that they are expensive, unstable upon annealing, and can cause a substantial increase in viscosity during processing [41] The effectiveness of these materials are also diffusion dependent and may not entirely wet the interface of the dispersed domain [31, 42, 43]. In-situ compatibilization is a process in whic h a fine dispersion of a minor phase can be generated quickly by reactive extrusion. This pr ocess involves using liquid reactants, such as functiona l monomers, to locate at the interface between immiscible polymers and subsequently polymerize. The purpose is to graft the monomers onto both polymers in a binary blend and create what is believed to be a network-like bridge between the phases. The ultimate product has enhanced mechanical properties, better processability, and a unique morphology. This dissertation focuses on the in-situ compatibilization of two phases the isotactic polypropylene matrix and an ethylen e-1-octene copolymer (otherwise known as linear low density polyethylene) minor phase. PP is known to behave in a brittle manner at or below room temperature and in a notched state, thus limiting its use in blow molded

PAGE 26

4 bottles, for example. The elastomeric copolymer, which has a glass transition temperature well below room temperature, will act as the impact modifier for polypropylene and contains a fully satura ted backbone which limits environmental degradation. But addition of elasto mer alone compromises both stress-strain performance and processability of the blend. The in-situ compatibilization technique established by previous researchers has b een applied to this system, and drastic improvements in all macro-scale properties have been achieved. In order to understand this dissertation, one must be versed in processing of polymers, polymer blends and toughening of polymers, solid state and melt free radical grafting of polymers, and free radical polymeriz ation. This is a very complex process, not only because of the number of compone nts and variables, but because of the dependency on both the rate of reaction a nd rate of morphology development. The following six chapters describe various as pects of this unique polymeric alloy. 1.2 Chapter Summaries Chapter 2 is a background/review of much of the research that has explored the complex issues involved in reactive extrusi on. The description includes the polymers of interest (i.e., isotactic pol ypropylene and ethylene-1-octene copolymers), the type of processing equipment involved, how the mor phology of polymer blends develop in an extruder, toughening brittle se micrystalline thermoplastics, many characteristics of free radical polymerization and free radical induced grafting of polyolefins, and how all of these aspects tie in to create novel in-situ compatibilized polymers. A fundamental study of the physical blends of a certain grade of copolymer with polypropylene is undertaken in Chapter 3. Li ttle research has been conducted on using low molecular weight ethylen e-octene copolymers to tough en polypropylene, but this

PAGE 27

5 chapter proves that these new copolymers are effective PP modifiers up to a certain concentration. Many concepts and characte rization techniques are introduced and explained in detail so as to aid in analysis of results gath ered in following chapters. A complete picture of the mechanisms of toughening in several alloyed systems is described in Chapter 4. This includes de fining the effect of elastomer content on mechanical, chemical, rheological, and morphological properties of alloys. A comparison is made between physical blends and alloys containing various levels of liquid reactants. The crystal line as well as amorphous states are described in detail, with both contributing to the overall performance of the alloys. In Chapter 5, the actual processing characte ristics of the alloys are elaborated upon. Extruder screw speed and barrel temperature ar e directly tied into alloy performance. The effects of varying reactant concentrati ons are also systematically studied. Chapter 6 delves deeper into the effect of elastomer molar mass and crystallinity on the behavior of both the physical blends and alloys. It explains many aspects of alloy behavior that would not have otherwise been possible. For Chapter 7, a design of experiments has b een conducted to see the effect of this in-situ compatibilization technique on high density polyethylen es impact strength and stress strain behavior. High density polyethylene is a high volume commodity plastic and modification to a high impact polymer is th e goal. This chapter is followed by a concluding chapter with future research possibilities along with appendices and references.

PAGE 28

6 CHAPTER 2 BACKGROUND AND LITERATURE REVIEW 2.1 Polymers of Interest 2.1.1 Isotactic Polypropylene Isotactic polypropylene (PP) combines low pr ice with attractive performance; e.g., heat distortion temperature > 100C, streng th, stiffness, corrosion resistance, and versatility in applications, ranging from automotive moldings to films and textile fibers [44-47]. PP is a semi-crystal line thermoplastic so its prop erties are strongly dependent on molecular weight and defect distributions which in turn aff ect both rheology and crystallinity. Most end-use properties of PP homopol ymers, such as stiffness, hardness, and high temperature mechanical properties, are positively influenced by their overall crystallinity, whereas impact strength and elongation are negatively influenced. A major factor in the profitability of PP is the availability of low-cost propylene monomer. There are two main sources of th e monomer: co-producti on with ethylene or separation from gasoline cracker steams in a petroleum refinery. When polymerizing PP, three requirements are always present: its chai n must be linear (monomer always adds to the chain end), regiospecific (monomer is always added in head-to-tail manner), and stereospecific (monomer always adds in the sa me stereo arrangement, or same side of chain) [46]. In 1954 Giulio Natta polymeri zed propylene by means of a modified Ziegler catalyst and obtained a blend of isotactic and atactic polypr opylene [456b, 45c]. For his pioneering invention, he and Karl Zeigler rece ived the Nobel Prize for chemistry in 1963. The present day polymerization process me dium can be either liquid or gaseous

PAGE 29

7 propylene or an inert hydrocarbon such as hexa ne. The process can be either bulk or gas phase or a combination of both. Most polyolefin manufact uring processes presently utilize heterogeneous ZieglerNatta catalysts. Because these catalysts have more than one type of active site, they produce PP w ith a broad molecular weight distribution (MWD) and non-uniform stereoregularity. Isotactic polypropylene is such a versatil e plastic that its applications are innumerable [45]. Some markets include fi bers, carpet and upholst ery, films, medical devices, automotive (under the hood, exteri or, interior), containers, construction, nonwoven fabrics, appliances, and transportatio n. The resistance of PP to chemicals is well documented and one of the principal reasons automobile batteries are made of PP. Having a HDT above 100C allows use with hot aqueous liquids, including steamsterilized medical goods. The processability of PP also makes it attractive. These methods include extrusion, oriented and melt bl own fibers, biaxially oriented film, blown film, sheet, thermoforming, profiles and pipe, wire and cable coating, injection molding, extrusion and injection blow molding, and compression molding. 2.1.2 Ethylene-1-Octene Copolymer Dupont Dow Elastomers produce and license novel ethylene-1 -octene copolymers with the tradename ENGAGE via INSITE technology, which allows extraordinary control over polymer structure, properties, and rheology [48-52]. They use a relatively new single-site metallocene cat alyst to polymerize a wide variety of bulky monomers, including linear -olefins. In contrast to Zeigle r Natta (Z-N) catalysts, they yield polymers which incorporate higher levels of the -olefin to achieve lower polymer density or crystallinity, and a uniform comono mer distribution with a polydispersity of about two [45a, 53, 54]. The microstructura l uniformity from metallocene catalysts

PAGE 30

8 allows greater dimensional stability, higher impact resistance, gr eater toughness at low temperatures, and higher resistance to environm ental stress cracking [49]. The ability to incorporate higher levels of comonomer has allowed densities of the copolymers to reach 0.87 g/cm3, previously unattainable by ZN catalysts. The development of new metallocene catalyst generations has bridged the gap between rubber and thermoplastic technology [44]. These materials exhibit an enormous span of rheological, mechanical, and thermal properties [48, 55, 56]. They have excellent lo w temperature propertie s, clarity and crack resistance. Their superior UV, ozone and w eather resistance are primary advantages over other impact modifiers such as EPDM (ethylene-propylene-diene monomer), EPR (ethylene-propylene rubber), and SBS (styrene-butadiene-styrene). In a comparison of elastomers, an EOC had a melting temperature 10C higher than EPDM of similar crystallinity and molecu lar weight [57]. A more homogeneous distribution of crystal morphol ogy is apparent for the copolymers, with the more defectridden EPDM providing less mechanical integrity. The maximum strength and extensibility of the ethyleneoctene copolymer are greate r than EPDM even though the EOC is lower in molecular weight. An extensive study of several ethylene-olefin copolymers was conducted by Bensason et al. [52] who clas sified these novel materials in to four types: Type 1 copolymers are those with densities less than 0.89 g/cm3 and show a low degree of crystallinity, low melting temperature, and the absence of cooling rate effects. Spherulites are nonexistent and the granular, nonlamella r morphology suggest that the crystalline regions should be described as fringed mice lles. Type 2 copolymers range in density

PAGE 31

9 from 0.91 0.9 g/cm3 and form poorly developed, unba nded spherulites containing both bundled and lamellar crystals. Type 3 materials (0.93 0.91 g/cm3) form smaller spherulites with thinner lamellae than HDPE homopolymer. Although the branches restrict crystallization to an extent, the et hylene sequences are long enough to crystallize in the lamellae. The fourth type has a density of 0.93 g/cm3 or greater and exhibits lamellar morphology with well-developed sphe rulites. Lamellar thickness is strongly related to the kinetics of crystallization because of the lack of long chain branching. Over the last decade, much effort has been put forth to understand the crystalline morphology and crystallization processes of th ese copolymers. A common consensus is that as the concentration or length of com onomer increases, crysta llinity decreases [28, 48-52, 58-64]. The introduction of more como nomeric units hinders the chain regularity necessary for crystallization to take place [56, 57]. There is a distortion of the crystalline lattices with an increase of 1-octene cont ent but even at very low density (0.882 g/cm3), certain amounts of lamellar crystal is still pr esent [60]. The melting enthalpy is reduced with increasing 1-octene cont ent in the copolymer. Mel ting temperature was shown by DSC to be inversely proportio nal to comonomer content [51] Reorganization of polymer chains occurs at room temperature for copolymers having a com onomer content higher than 2.1 mol% of 1-octene [61] but this seems less likely for very high comonomer contents because during annealing the branches would have to be drawn through crystals [65]. 2.2 Processing 2.2.1 Reactive Twin Screw Extrusion Polymer processing in a twin screw extr uder has been developed since the 1930s and 40s, with several varieties offering nume rous advantages over the other [2-4, 31, 66].

PAGE 32

10 Twin screw extrusion (TSE) has been shown to be a versatile, cost effective method to produce a uniform, optimized polymer based product. An ideal compounder will have a uniform shear and elongational stress field, fl exible control over uniform temperature, pressure and residence time, compatibility for homogenization of liquids with large differences in rheological pr operties, efficient homogeniza tion before degradation, and flexibility for change in mixing parameters. TSEs are useful because of the ease of feeding materials, excellent dispersive a nd distributive mixing, temperature control, control over residence time distribution, reac tion under pressure, cont inuous processing, unreacted monomer and byproduct removal, pos t-reaction modification, and viscous melt discharge. Most of the mixing is achieved with kneading paddles. The main geometrical features that distingu ish twin screw extruders are the sense of rotation and the degree of intermeshing. Twin screw extruders whose sc rews rotate in the same direction are co-rotating. The intermes hing twin screw extruder is self-wiping in nature and helps to minimize the very long residence time tail frequently found with extruders. They give a relatively uniform shear rate distribution and because the feed rate is independent of screw speed, high sc rew speeds are possible (500 rpm) with correspondingly high throughput rates. With this high speed, small sized equipment can achieve high melting and mixing capacities. Two drawbacks are the cost to purchase and maintain and a metered feeding device is need ed in starve feeding mode. APVBaker Perkins is the manufacturer of our extruder, with some unique features being a clam shell barrel, greater free volume, and barrel valves. The first developments in the use of extrude rs as reactors were made about 60 years ago and melt phase modification of polymers has been done for over 35 years. Reactive

PAGE 33

11 extrusion (REX) refers to the deliberate use of chemical reactions during continuous extrusion of polymers and/or polymerizable monomers [4, 31, 67-69]. Reactions have been performed on molten polymers, on lique fied monomers, or on polymers dissolved, suspended in, or plasticized by solvent. The types of chemical reactions that have been performed by reactive extrusion include bul k polymerization, graft reaction, interchain copolymer formation, coupling/crosslinking reactions, controlled degradation, and functionalization/functional group modificati on. The attainment of proper mixing is undoubtedly the single most important consid eration when specify ing or designing an extruder-reactor. Chemical reaction is a mo lecular event, so proper mixing in reactive extrusion means mixing at the molecular le vel and maximizing the interface between dispersed phases and th e matrix [4, 31, 70-72]. Polyolefins have proven to be prefer red substrates for reactive extrusion experiments largely due to their ready availability, low cost, and commercial applications. The advantages of synthesizi ng graft copolymers by reactive extrusion as opposed to alternating technologies include li ttle or no use of solvents, simple product isolation, short reaction times, continuous pro cess, and relatively low infrastructure costs [68]. The ability of an ex truder to handle materials havi ng high viscosities without any solvents results in a dramatic raw material cost reduction, no solvent recovery equipment, an ready-to-use products [73]. Some potential disadvantages or diffi culties are the need to achieve intimate mixing of reactants a nd substrates, high reaction temperatures necessary to form a polymer melt, and polymer degradation or crosslinking. 2.2.2 Morphology Development in an Extrud er Dispersive and Dissipative Mixing The performance of extruded materials is determined, amongst others, by the final morphology and dispersion [74, 75]. For ble nds and alloys, the morphology depends on

PAGE 34

12 the composition, rheological a nd physical character istics of the components, relative compatibility, and the nature and intensity of the mixing. When purely compounding two plastics, they go through disp ersive as well as distributive mixing stages [3, 31, 32, 76, 77]. Dispersive mixing is the breaking up of clumps or aggregates of solids into the ultimate particulate size, or of immiscible polymers into the desired domain size. It is dependent upon shear and elongational stress and is achieved by shearing the particulate matter under high stress usually by kneading di sks. In distributive mixing, spatial uniformity of all components throughout the mi xture is desired. This is best achieved by frequent reorientation of flow elements under strain, including dividing, stretching, distorting, and/or reorienting the flow. Mixing performance is known to decrease with increasing viscosity [72]. The most significant evolution of morphology occurs in the init ial melting zone of the extruder. During the initial stages of ble nding, the elastic behavior is most important but the viscous and interfacial behavior of the components in a system is undoubtedly important in the later stages of mixing. The maximum shear stress, accompanied with frictional and extensional forces, is usually generated at the melting zone of the extruder and imparts a high degree of mixing [ 13, 31, 32, 60, 78, 79]. The melting mechanism arises from the dissipation of the energy crea ted by interparticle friction, rather than by friction against the barrel wall or by heat transfer through the barrel wall [3, 75]. The rate of melting controls the rate of reaction and morphology development [31]. Each polymeric component changes into very small particles as droplets within a very short time and distance (0.1-10 seconds and a few millimeters). A thermoplastic is dispersed in

PAGE 35

13 the rubber phase with th e plastic pellet size re duced from 3 mm to 5-20 m, then to approx 1 m with eventual coalescence [31, 78, 79]. 2.2.3 Melting and Droplet Breakup Mechanisms As the solid begins to melt, the feed mi xture may go through its most viscous stage, that of a highly filled slurry or paste of unmelted solids in just-melted resin [3, 32, 75, 8082]. Often the minor phase softens first and wi ll coat particles of the major phase, which will delay its melting [13]. Transformation of a solid pellet i nvolves three steps: melting/plastification of the pellet, deform ation/stretching of the molten polymer, and formation of fine particles which ma y be subject to coalescence [31]. Several theories have been formulated which describe the process of melting and morphology development [32, 75, 78, 81, 83]. The fi rst to model this behavior was Shih [78, 79] with the phase inversion mechanism. He found that polymer blends go through a number of sequential physical changes before being combined into a cohesive mixture to minimize free energy. For semicrystalline po lymer/rubber mixtures, plastic pieces are initially torn from the pellet surfaces and form a mixture with drawn out layers of rubber. A lower melting or softening polymer disper sed into a higher melting polymer of major phase volume follows a 4-stage inversion mech anism as shown in Figure 2-1: A. The rubber forms a continuous phase closely packed pellets are suspended in it; B. Plastic pellets break up layer by layer as the pellet surfaces begin to soften, shear, and pull off from the unmelted solid core and are disperse d in the rubber phase; C. In the region of maximum torque (0.7 m particles), an abrupt phase inversion occurs due to coalescence

PAGE 36

14 Figure 2-1: The phase inversion mechanism proposed by Shih. of molten plastic partic les and finely divided rubber dr oplets are formed in a continuous plastic matrix; D. The final stage is a viscoela stic fluid matrix with finely divided rubber droplets suspended in it followed by a continued decrease in torque. Although the onset of phase inversion is abrupt, the completion of the phase inversion during phase C is not instantaneous. A small amount of the high melting polymer remains trapped in the rubber phase at the end of the mixing cycle. During the phase inversion, the mixture morphology change s from a continuous rubber phase with a very high concentration of high melting par ticles (80%) to a continuous molten plastic phase with a smaller amount of dispersed rubber particles (20%). Th e overall viscosity is expected to drop significantly, simply from the change in dispersed phase concentration. Sundararaj [32] and Scott and Macosko [83] proposed a droplet breakup and coalescence theory for morphology development (Figure 2-2). An initial mechanism of droplet breakup involves the formation of sheet s or ribbons of the dispersed phase in the

PAGE 37

15 matrix, which are drawn out of a large mass of the dispersed phase. The pellet breakup is primarily controlled by the rate of deforma tion and subsequent rela xation of the pellet phase. As the relaxation time of the pellet de creases (more elastic behavior), it becomes more difficult to create a sheet. Owing to the effects of flow and interfacial tension, these sheets are unstable and holes begin to form in them. A high stress level followed by a lower stress level is required to achieve effici ent mixing [84]. In the high stress level, the dispersed phase is stretched and extended into shapes, which undergo instabilities and break up upon entering the low stress level. Figure 2-2: Morphology devel opment of binary polymer blends proposed by Scott and Macosko. As the sheet grows, the holes are filled with the matrix phase, which surrounds the sheet on either side. When the holes in the sheet or ribbon attain a sufficient size and concentration, a fragile lace structure is fo rmed, which begins to break apart into irregularly shaped pieces of a wide distribut ion in size. These pieces are approximately

PAGE 38

16 the diameter of the particle s generated in the blend at long mixing times. The irregular pieces continue to break down until all of the particles become nearly spherical. This proposed mechanism results in the gene ration of very small particles at very short timescales. As a piece of dispersed phase undergoes this deformation mechanism, many very small particles may be generated ve ry quickly. As the mixing time proceeds, a greater proportion of dispersed phase is cycled through this mechanism. 2.2.4 Morphology Development Dependency on Rate of Reaction Typically, small minor phase drops less than 1 micron are desired. If the dispersed phase is dilute, this is relatively easy but at dispersed phase concentrations greater than 1%, collisions between drops occur and domain size increases due to coalescence [13, 32, 42, 78, 79]. Also, the particle size distribu tion broadens at higher concentrations. Coalescence of the dispersed domains has been shown to be dramatically reduced (up to 30%) by bonding at the interface [32, 42, 85, 86] Reaction increas es the effective interfacial tension and both delays and intensifies the pha se inversion process. The morphology development of the reactive system parallels its non-reac tive counterpart, but the final number average particle size is two orders of magnitude smaller. Steric stabilization of the dispersed phase is more important than interfacial tension decrease [32, 87-89]. When the blend is reactive, very high concentrat ions of the major phase can exist as the dispersed phase for longer periods of time before phase inversion occurs. If the particles are not monodisperse and are not perfect spheres, then much higher concentrations (>74%) of the di spersed phase are possible [32]. One can imagine that when the major phase envelops the minor phase, small particles of the major phase generated duri ng melting will be trapped inside the minor

PAGE 39

17 phase. If the system is reactive, then the occluded domains will be stabilized and will not be able to coalesce with the matrix [32]. During reactive extrusion, a time scal e will exist for both the morphology development and reaction. Polymeric alloys al l have one thing in co mmon the rate of melting dictates when the reaction will begin [13, 31, 32, 42]. For these systems, reaction occurs almost immediately after the melting of the polymers. There is a fine balance between the time for morphology generation a nd reaction time, so the ratio must be manipulated to obtain the best processing conditions and optimum morphologies. When compounding all components together, the be nefits of co-melting are realized by good dispersion of the materials and a greater probab ility of locating reactive ingredients at the interfaces to facilitate grafting/bonding [27, 90]. If liquid reactants are added downstream after melting has occurred, dispersion is difficult due to the highly viscous molten polymer which will lead to polymerizati on rather than grafting onto the polymer. The flow in an extruder promotes the reac tion in two ways: First, it either breaks or deforms the suspended droplets and increases the interfacial ar ea available for the reaction. Second, the flow convectively incr eases the mass transport to supply fresh reactants to the reaction [71]. The improved reaction rate by back and forth flow can be attributed to more efficient productio n of new interfacial area [72]. Many studies have shown that the majority of reaction occurs in the melting zone. For monomer grafting onto polypropylene, the conversion can be up to 70% of the final value upon melting [91]. The level of styrene grafting onto a polyolefin can be 85% after the materials pass through a kneading block zone [92]. Another study has shown that free radical grafting of a m onomer onto PE or PP has alre ady gone to completion just

PAGE 40

18 after the kneading block [90]. Machado et al. found that regardless of the type of polyolefin, the majority of gr afting reaction occurred in the melting zone of the extruder [93]. They also found that higher grafting yields exist fo r lower viscosity polymers, possibly due to the melting phenomenon. Graf ting may actually occur below the melting temperature of the polymer because of signi ficant levels of peroxide decomposition [94, 95]. Diffusion of reactants is not the most impor tant issue for an alloyed system because if the reaction rate is fast, the size of the di spersed phase will be ve ry small [42]. A good compatibilizing chemistry has to be fast enough compared with the rate of interfacial area generation so that once the in terface is created, it is stabil ized quickly by a layer of copolymer so as to minimize coalescence [ 31]. The shearing or extensional flow experienced by the high molecular weight polyme rs creates a very large interfacial area, thus reducing the need for long range diffusion and facilitating reacti on at the interface. 2.2.5 Viscosity and Reaction Effects on Morphology Development When two immiscible polymers are blende d, one phase is mechanically dispersed inside the other. The size and shape of the dispersed phase depend on several processing parameters including rheology, interfacial properties, and composition [31, 32, 96]. Viscosity ratio and surface tension be tween the major and minor phases play important roles in determining droplet si ze [31, 39, 69, 75, 82]. With the viscoelastic fluids generally encountered in polymer pr ocessing, the lower the elastic nature of dispersed phase, the lower the matrix stress es required to break up and stabilize them. Taylor was the first to theoretically de scribe droplet breakup and formation while suspended in another liquid medi um [97, 98]. When the rate of distortion of the fluid or the radius of the drop is great enough, the drops tend to break up. For a very small

PAGE 41

19 viscosity ratio ( Drop/ Matrix), the drop remains coherent in sp ite of the fact that it gets very long and narrow. The act of bursting is al ways an elongation to a threadlike form followed by degeneration into drops which are of the order of 1/100th of the size of the original drop. A low viscosity minor phase wi ll break up into small droplets early in the extrusion process [32, 76, 96] but beyond a mi nimum viscosity coalescence is favored [39]. 2.3 Toughening of Polymers 2.3.1 Origins of Polymer Toughening Polymer toughness, or the property of resisting fracture by absorbing and dissipating energy, is a highly sought after charac teristic of a material or product. It depends on many parameters including temperatur e, pressure, deformation rate, shape of specimen, and type of load, aside from ma terial properties like molecular weight, polydispersity, chain packing, chain entanglem ents, crystallinity, heterogeneity, etc. Brittle fracture occurs at high strain rates, low temperatures, and in thick sections because each restricts the extent of the yield zone. Plastic deformation itself is a complex phenomenon and involves both crystalline and amorphous phases [99]. Energy is absorbed within the sample by viscoelastic deformation of the polymer chains, and finally by the creation of new surface areas [100]. Fracture resistance in rubber toughened polym ers is generally attributed to three major mechanisms that absorb or dissipate energy as cracks advance through polymers chains: rubber cavitation [ 29, 99, 101-105], matrix crazi ng [28, 106, 107], and/or shear yielding [5, 8, 12, 101, 103, 105, 106, 108, 109] with chain breakage accompanying the failure of polymers. Impact resistance has also been correlated with the presence of a secondary transition at least 50C below the testing temperature [100, 105, 107]. The

PAGE 42

20 speed of the impact test eff ectively raises the temperature of this secondary transition by 50C (time/temperature equivalent); therefore, secondary transitions occurring near the test temperature at low frequency are effectiv ely shifted into the gl assy region at testing frequency. Below its Tg, a rubber particle will not cavitate and will therefore behave in a brittle manner. Secondary transitions must be associated with motion of the polymer backbone (glass transition temperature), not pendant side chain groups, for them to be effective at improving impact resistance. When a toughened polymer blend is subjec ted to a uniaxial stress, the localized stress experienced by the matrix material in the vicinity of a r ubber particle will be magnified by the local stress concentration factor. The initial cavitation of a rubber particle relieves the triaxial stress existing at a crack tip and enhances localized yielding in the matrix, thus avoiding a brittle catastrophi c failure of this mate rial. If the applied stress is increased further, the crack tip ma y be bridged by the stretching rubber particle, provided there is sufficient adhesion between the rubber and the matrix and that the particle can stretch sufficientl y rapidly in terms of the speed of the crack advance. For this reason, it is desirable that the rubber should have as low a Tg as possible, while allowing the rubber to fibrillate and maintain a degree of structural integrity in response to impact loading [12]. C onventional wisdom states that for toughening polymers, the rubber droplets must be at least as large as the cracks they are trying to stop, putting the minimum size at several hundred angstroms to 300-500 nm [100]. 2.3.2 Elastomer/Rubber Toughened Blends A fundamental understanding of how rubber pa rticles affect the creep response of polymers is described by the Erying theory, which states that en ergy barriers at the molecular level control the m acroscopic rates of flow [110] This is a fundamental

PAGE 43

21 property which depends on rubber compositi on, volume fraction of dispersed phase, rubber particle size and distri bution, and rubber-matrix interfaci al interactions [2, 3, 8, 9, 22, 28, 69, 107, 111-115]. Toughness is the greatest at an optimum rubber particle size for polymers that dissipate fracture energy mainly by matr ix crazing (PS and PMMA) [70, 116]. But for semicrystalline polymers, close interparticle dist ance is just as important as small particle size and high interfacial adhesion [5, 21, 69, 113, 116-118]. A critical interparticle distance (matrix ligament) exists below which a material behaves in a ductile manner and above which behaves in a brittle manner [69, 103, 113, 119-124]. Ligament thickness is an inherent property of the polymer and is independent of rubber volume fraction or particle size. At large separation distances, the stress field in the matrix is simply a superposition of those around isolated part icles, and the polymer blend will remain brittle. However, when the particle surfaces are sufficiently close, the stress field is no longer simply additive, and the fields around the particles will interact. This will result in enhanced matrix yielding, and a transition to tough behavior [119]. Even if rubber particles are chemically bound to the matrix, a polymer blend will stil l be brittle if the interparticle distance or particle size is greater than a critical value. Bartczak [125] and Muratoglu et al. [126] expanded upon the matrix ligament theory and revealed that a la yer of anisotropic crystalline material having a lower plastic resistance than the bulk surr ounds the dispersed particles. Van Dommelen et al. have actually modeled this behavior and show that this unique layer is a highly efficient method for toughening semicrystalline polym ers by altering matrix craze formation, reducing principal stresses, and indu cing extensive matrix shearing [115].

PAGE 44

22 In general, the critical particle size for toughening decreases with increasing ductility of the matrix polymer. For a Nyl on 6,6-rubber blend, it was found that if rubber particles are large, a greater amount of rubber is needed to achieve toughening and vice versa [119]. Chou et al. showed that rubbe r particles narrowly distributed around 0.5 to 1 micron in size are effective for toughening PP [103]. A large number of small rubber particles are preferred when toughening PP beca use they are more efficient at promoting shear yielding and crazing throughout the matrix [22, 69, 127]. Also, a bimodal distribution of rubber gives a good balan ce of toughness and stiffness in PP [23]. If rubber is well bonded to the matrix, stre sses can become redistributed between phases subsequent to yielding, so that the plastic zone is just as capable of resisting crack extension as the homopolymer. The effect of poor matrix-rubber bonding is a weaker plastic zone, thus canceling out the benefits to be derived from a reduction in yield stress. Good rubber-matrix adhesion and small particle size are necessary cr iterion for screening rubbers to toughen PP and increase impact st rength [19, 22, 110, 128]. Blends with very small particle size have a relatively high cavit ation stress, which re sults in a high yield stress of the blend [104]. A final contribution to toughening semicrys talline polymers is from stress-induced phase transformations [129]. If changes in crystallinity take place during deformation, energy will be absorbed by me lting or released by crystall ization. A comparison can be made of the softening effects due to a rise in temperature to those associated with an isothermal reduction in crystall inity. Adiabatic heating of polymers fractured at high deformation rates is common [21, 130, 131] a nd influences the melting/recrystallization process which in turn affects plastic deformation [21, 129-135].

PAGE 45

23 In summary, for optimum impact strength improvement in PP/elastomer blends, the following conditions must be satisfied [136]: 1. Elastomer particles are finely and un iformly distributed in the PP matrix, 2. The modulus of the elastomer is much less than that of the PP, 3. The crystallinity of the elastomer is low, 4. A certain degree of interfaci al adhesion is present betw een the elastomer particles and the PP matrix, 5. The cohesive strength of the elastomer is large, 6. A certain degree of entanglem ent of high MW polymer chains is present in the PP matrix. 2.4 Free Radical Reactions The free radical grafting process is very complex and not completely understood. The following table lists the most likely re actions to be encountered while grafting monomers onto polyolefin substrates. Table 2-1: The mechanisms of main reactions in the grafting process. Initiator Decomposition R'OOR' 2R'O Hydrogen Abstraction R'O + P R'OH + P -Chain Scission (PP phase) P P1 + P2 Crosslinking (PE phase) P + P P-P Graft Initiation P + M PM Graft Propagation PMn + M PMn+1 Homopolymerization R'O + M R'OH + Mn Termination by Recoupling PMn+1 + Mn PMm 2.4.1 Initiator Decomposition Previous research has shown that a peroxi de initiator is necessary to create enough free radicals to graft monomers onto polyolef in substrates which may be due to the presence of stabilizers in the polym er resin [4, 7, 90, 94, 137-143]. Initiator

PAGE 46

24 decomposition is known to be the rate lim iting step for hydrogen abstraction [92]. Dialkyl peroxides are the initia tors of choice because they ar e amongst the most stable of all commercially available organic peroxi des and the free radicals generated from decomposition have a variety of uses [144]. 2,5-dimethyl-2,5-di-(tbutylperoxy) hexane (tradename Lupersol 101) is often used to degrade polypropylene, crosslink various types of polyethylenes, and graft monomers onto th ese polyolefins because of its efficiency [73, 92, 94, 147, 148]. Figure 2-3 gives the chem ical structure while Table 2-2 lists characteristics of the peroxide. Figure 2-3: Chemical st ructure of Lupersol 101 Table 2-2: Characteristics of Lupersol 101 [145, 146]. Molar Mass (g/mol) 290.44 Peroxide content (%) 91-93 Oxygen content (%) 10.03-10.25 Physical form Liquid Melting point (C) 8 Boiling Point (C) 249 Specific gravity (cm3 at C) 0.865 @ 25C Viscosity (mPa.s) @ 20C 6.52 Typical Decomposition products in inert media: methane, ethane, ethylene, acetone, tbutyl alcohol. Lupersol 101 is c onsidered a suitable food additive. The half life of a peroxide is very impor tant, defined as the time it takes for one half of a given quantity of peroxide in dilute solution to decompose at a given temperature. The melt free radical grafting rate is dictated by peroxide efficiency, or half life [138, 146]. One has to keep in mind that the half life of L upersol 101 in molten LDPE is reported to be 2-3 times longer than in organic solvents [4, 90, 94]. A similar

PAGE 47

25 result is found when peroxide decomposition is in the solid (glassy ) state of a polymer [149]. The low reactivity of polymers compar ed to model compounds is attributed to several factors, including the lower concentration of tertiary C-H reaction sites, coiled conformations, and high viscosity [150, 138]. Conventional wisdom shows that half life time should be about 5 times that of the residence time of the polymers [ 68, 92]. If the half life is too long, the initiator may not be completely utilized. A short half life ma y increase the crosslinking of radical-radical combination or grafting yield may be limited by the rate of monomer diffusion to the site of reaction, especially a heterogeneous melt. Other work has shown that no grafting occurs after consumption of the initiator (as estimated by half life) [151]. Half life is calculated as 1/2 = 0.693/kd, where kd = Ae-Ea/RT, Ea is the activation energy, R is the universal gas constant, T = temperature (K), and A is an integration constant [146]. Table 2-3 is a manufacturers generated list of estimated half lives for Lupersol 101 at various temperatures. Table 2-3: Half life of Lupersol 101 based on temperature Temp (C) 165 190 210 230 Half life (seconds) 280 (4.7 minutes) 28 5.4 1.2 The mechanism of free radical producti on from Lupersol 101 involves a primary and secondary reaction, with decomposition of each group independent of eachother [68]. Upon addition of heat, this peroxide decompos es homolytically into 4 alkoxy radicals at a bond dissociation energy of a bout 36 kcal/mol [145, 152, 153]:

PAGE 48

26 A secondary reaction can occur in which the tertiary alkoxy radicals can undergo further fragmentation (e.g. -scission) to form ket ones and alkyl radicals: The extent of reaction at a given position is proportional to the amount of peroxide initiator that has decomposed at that pos ition. In regions of high temperature, the peroxide reacts more rapidly, inducing a concentration grad ient that drives additional peroxide to diffuse into the hot region and react. The faster the initiator is consumed, the less time it has to diffuse to other por tions of the channe l and react [154]. 2.4.2 What Happens After Peroxide Decomposition? A radical produced by either primary or secondary reactions can abstract a hydrogen (from the polymer backbone) or add to a double bond (vinyl monomer). The secondary reaction is strongly temperature dependent, more so than abstraction or addition [68, 155]. So with an increase in temp erature, there is an increase in the number of methyl radicals, which a ttack bonds in a much more random nature than t-butoxy radicals and usually add to vinyl monomers ra ther than abstract hydrogen atoms from the polymer [156]. Methyl radicals add to styren e some three orders of magnitude faster than the model compound 2,2,4 trimethylpentane [68]. 2.4.3 Polymerization The feasibility of radical polymerizati on of a monomer depends primarily on the polarity and size of the substituents on the double bond and the tendency to chain transfer [153, 157-160]. The reactivity of the mono mer is influenced by two factors: 1. The stability of the monomer toward addition of a free radical

PAGE 49

27 2. The stability of the monomer radical thus formed (more important). Monomers with less resonance stabilization represent much higher energy states. The order of reactivity of the radicals is th e reverse of that for monomers; i.e., styrene monomer is more likely than methyl acrylate to consume a radical but addition of styrene monomer to a polymerizing chain is 100 times slower than methyl acrylate [153, 161]. A larger propagation rate constant will lead to higher molecular weight. Resonance stabilization depresses the activity of the radi cal. Low activation en ergies are indicative of a greater decrease in en ergy from reactants to produc ts [157, 162, 163]. Table 2-4 gives typical activation energies for several aspects of free radical polymerization in solution [146]. Propagation is bimolecular and its rate is independent of chain length. The initial chain formed rapidly produces a high mol ecular weight polymer [146]. A monomers ceiling temperature is defined as the te mperature above which monomer cannot be converted into long chain polymer. A relative ly low ceiling temperat ure is suggested for grafting single monomer units onto polyme rs or to limit homopolymerization. For styrene and acrylate-based monomers whose ceiling temperatures are well above polypropylene extrusion processing temperatur es, depropagation or unzipping of the chain is not a concern. Table 2-4: Typical free radica l kinetic values in solution. Type of Reaction Activation Energy (kcal/mol) Initiator Decomposition 30-50 Initiation 5-7 Propagation 4-10 Chain Transfer 10-20 Termination 0-6 Less stable -bond more stable -bond: -12 to -23.9kcal/mol: exothermic

PAGE 50

28 During free radical polymerization [164, 165] and graft copol ymerization [166, 167] bimolecular termination is still energe tically favorable but the macroradicals are large with slow diffusion, so termination w ill be almost exclusively diffusion controlled. The rate constant of terminat ion for styrene is inversely pr oportional to the viscosity of the medium over a thousand fold range of viscosity. Chain transfer may occur, defined as th e termination of one macroradical to produce another macroradical which serves as a branch point. The new free radical produced may or may not be initiate anothe r polymer chain formation, depending on its activity. Styrene is more to likely add monomer than underg o chain transfer to either polypropylene or polystyrene [162a] Styrene monomer (Table 2-6) is an aromatic hydrocarbon which, under normal conditions, is a clear, colorless, flammable liquid. It is a versatile material, the derivatives of which are styr ene-based polymers. Styren e is one of the few vinyl monomers that undergo rapid th ermal polymerization [168, 169]. 2.4.4 Copolymerization When studying the copolymerizat ion of two monomers, their reactivity ratios result from a combination of steric, res onance, and polar effects [153, 159, 160]. Steric effects: bulky substituents decreas e reactivity in radical polymerizations. Resonance: If a radical can be stabilized by resonance, it is more likely to form (monomer is more reactive). However, re sonance stabilization of the radical also makes the radical less re active towards propagation. Polar effects: A monomer with an electronwithdrawing substituents is more likely to react (cross-propagate) with a monomer having an electron-donating substituent than it is to self-react. So, ethyl acry late will want to add styrene monomer units over other ethyl acrylate units.

PAGE 51

29 Two monomers can either undergo self-pol ymerization or copolymerization. The terms r1 and r2 are reactivity ratios, and define the relativ e tendencies to self-propagate or cross-propagate. If r1 > 1, then monomer 1 tends to self-propaga te, whereas if r1 <1, copolymerization is preferred. In Q-e sche mes, Q is a measure of monomer reactivity (resonance stabilization), and e relates to monomer polarity. Q values increase with increasing resonance stabilization, while e values become less negative as groups attached to the double bond become more elec tron attracting. Tabl e 2-5 gives Q and e values [162c] and reactivity ratios [162b] for styrene and ethyl acrylate, while Table 2-6 gives basic information on the three monomers of interest. Table 2-5: Copolymerization constants fo r styrene and ethyl acrylate monomers. Monomer (r) Q value e value reactivity ratio Styrene (r2) 1.00 -0.8 0.699 Ethyl acrylate (r1) 0.41 0.55 0.139 Because r1*r2 deviates from unity and sin ce e is much different for the monomers, alternating copolymerization will exist [161] The reactivity ratio of styrene does not change with temperature but ethyl acrylate can be strongly affected. The higher the temperature, the more ethyl acrylate units will be incorporated into the copolymer [170]. The initiator has no significant effect on the reactivity ratio values. For two monomers far apart on the polarity series, the rate of copolymerization is often much higher than for either monomer alone [160]. Table 2-6: Structural and physical inform ation about the monomers of interest. Monomer Styrene Diethyleneglycol diacrylate (DEGDA) Trimethylolpropane Triacrylate (TMPTA) Structure

PAGE 52

30 Table 2-6 Continued Functiona lity 1 2 3 Molar Mass (g/mol) 104 214 296 Viscosity @ 25C (cps) 0.76 12 106 2.4.5 Multifunctional Monomer The random nature of free radical polyme rization processes like ly results in the assemblage of units in an irregularly pa tterned structure. Nonlinear polymers are obtained from monomers at least some of which posses a functi onality exceeding two. Gelation can be avoided in nonlinear polymeri zation by limiting the extent of reaction or using small proportions of reactants [161]. In a study which compared homoand co -polymerizing multifunctional monomers, Diffusion of the monomer played a signi ficant role in bond conversion. As the composition of the crosslinker is incr eased, the maximum double bond conversion decreases due to diffusional limitations to the polymerization pr ocess [163, 171]. The acrylate copolymers are more homogeneous than methacrylate copolymers [163]. When comparing di-functional vs. tri-functional mono mers, the concentrati on of crosslinking double bonds is 43% higher for tridue to hi gher functionality per monomer [172]. For neat monomers, the maximum polymerization ra te for di-functional is twice that of trifunctional. A survey of multifunctional acrylates wa s conducted to assess the effect of monomer rank (number of atoms between acr ylate functionalities) size of functional group, and number of functional groups [171]. As the rank increased, the conversion of

PAGE 53

31 double bonds increased from 74.5% to 84.2%. This was explained by the increased mobility in the double bonds in the latter part of the reaction. Activ ation energy and rate of polymerization decreases as the number of alkyl groups increases [173]. The opposite trend was observed for increasing bulkiness of pendant functional group and increasing functionality of the monomers. The in crease in the number of double bonds of the monomer did not necessarily increase the possi ble crosslinking concentration. Acrylates reacted faster and had higher enthalpy of polymerization than methacrylates. The monomers with smaller pendant groups and lower functionality exhibited higher rate constants during polymerization due to increas ed diffusivity of the monomer and pendant double bonds in the reacting gel. The reactiv e pendant group could introduce hindrance, reduce the reactivity of the double bonds and reduce the conversion. 2.4.6 Hydrogen Abstraction The specificity of hydrogen abstraction de pends on an array of steric, polar, and stereoelectronic factors, including bond dissoci ation energies and ki netic effects [68, 140, 150, 152, 174-178]. The lower the bond dissociati on energy, the more stable the radical and therefore the more reactive. Alkyl radical stability increases in the order primary (1) < secondary (2) < tertiary (3) < allyl benzyl. Barriers for tertiary, secondary, and primary radical formation are 10, 11.5, and 12.2 kcal/mol, respectively, which indicates an order of magnitude differen ce in kinetics between tertia ry and primary formation [155, 156]. For isotactic polypropylene at 130 C, the relative reactivity of CH:CH2:CH3 groups with benzoyl peroxide as initiator is 50 :10:1 (reactivities pe r H atom) [68]. Another study showed that the PP macrorad ical actually forms a mixture of primary, secondary, and tertiary free radicals (1:14:18) [179]. Abstraction from the tertiary C-H position for a PP model accounts for two-thirds of the total product distribution [150].

PAGE 54

32 In modeling polyethylene, secondary radica ls were seen but there was no evidence of signals from primary radicals [156]. As previously stat ed, methyl radical formation begins to dominate at high temperatures. At 200C, CH:CH3 hydrogen abstraction for methyl radicals is 8:1 and CH2:CH3 is 3:1. At 200C, CH:CH3 hydrogen abstraction for t-butoxy radicals is 6:1, and CH2:CH3 is 5:1 [68, 180]. It is possi ble that, if secondary or primary radicals are produced, they may be tr ansformed into the more stable tertiary radicals by subsequent interor intramolecular abstraction reactions [68]. When comparing polymers, bulky side groups can also reduce the reactivity of the substrate [151, 181-183]. For grafting branch ed polyethylenes, tertiary-hydrogen atoms are three to four times mo re reactive than secondary hydrogen atoms [155]. The introduction of a branch leads to the repl acement of 4 secondary hydrogen atoms by one tertiary and three relatively unreactive primary H atoms. For an ethylene-octene copolymer with 10% octene, the site of grafting is 12:1 for secondary:tertiary [150]. Ther efore, a greater number of secondary groups results in predominant grafting at these points offsetting the greater reactivity of the tertiary C-H reaction sites. As the octe ne content of the copolymer changes to 5%, the ratio of secondary to tertiary sites changes to 22:1. 2.4.7 Polyethylene Crosslinking Reactions Polyethylene will crosslink in the presence of free radicals because of the long lived macroradical present af ter hydrogen abstraction. Cros s-termination of PE has a high rate constant [155]. ENGAGE polyolefin elastomers have been found to form an insoluble gel only above 0.3 wt% initiator [ 184]. It is now known that the highest molecular weight fractions of polyethylene wi ll be consumed first and a higher amount of crosslinking agent sh ifts the distribution towards lo wer molecular weights [149, 185].

PAGE 55

33 The low MW fraction is believed to act as grafted pendant chains in the presence of longer polymer chains. The network formed w ithin the low MW fraction consists mainly of chemical crosslinks, whereas high MW material comprises both physical entanglements and chemical crosslinks. Tra pped entanglements generate the major part of the crosslinking points at low peroxide c oncentrations, especially in HDPE. At high number average molecular weight values, dens e networks are easily created with only a small mount of chemical crosslinks, as the pr obability of entanglement formation is very high. The presence of polyfunctional monomers greatly increases the efficiency of polyethylene crosslinking [186]. The degree of crosslinking levels off rapidly at 3% monomer, but monofunctional monomers such as styrene, methyl methacrylate, and vinyl acetate were observed to have no effect on PE crosslinking. Yang et al. desired to reduce crosslin king when trying to graft glycidyl methacrylate onto polyethylene and used se veral inhibitors and monomers to do so [187]. They found that the chain transfer agent/inhibitor p-benzoquinone gave an acceptable grafting degree with minimal crosslinking. Styrene monomer, on the other hand, gave an unusually high gel cont ent along with grafting degree. 2.4.8 Degradation of Polypropylene Chain scission is the most energetically and kinetically favorable process after hydrogen abstraction from the position on the PP backbone. Disproportionation, another term for -chain scission, results in a satu rated product and an unsaturated molecule, depicted in Figure 2-4 [137, 153, 188]. This degradation di scolors the plastic, reduces crystallinity, compromises mechanical strength, and lowers the viscosity and melt strength [189, 190]. But some authors have shown that peroxide addition can

PAGE 56

34 narrow the molecular weight distribution and el iminate the high molecular weight tail of polypropylene to facilitate processing [73, 189, 191-194]. Figure 2-4: Schematic of H abst raction from PP and subsequent -chain scission. The degree of chain scission is found to be proportional to the initiator level. Each reaction occurs randomly, leading to random chain scission. Under typical extrusion processing temperatures and low levels of pe roxide, the backbone radical center is highly unstable and undergoes -chain scission almost immediat ely after H abstraction. The activation energy for -chain scission has been reporte d to be between 29.6-32 kcal/mol but the relative rate is a function of temp erature [137, 179]. Bimolecular termination, which is highly diffusion controlled, may be negligible because of the low concentration and short lifetime of the backbone radical [152, 154, 195]. The scission occurs randomly along the chains and higher MW chains have a greater number of bonds, so longer chains will experience -scission preferential ly [152, 154]. The number average molecular weight is inversely proportiona l to the degree of chain scis sion, which in turn is linear with respect to the am ount of initiator [194].

PAGE 57

35 2.5 Previous Efforts to Reduce/Prev ent Degradation of Polypropylene 2.5.1 Free Radical Grafting of Polymers Numerous processes of grafting monomers onto polymer backbones exist, with free radical melt, solution, and solid state grafting common practice [7, 68, 90. 91, 138, 148, 155, 177, 179, 190, 196-199]. Gamma [200] and UV [201, 202] radiation as well as ozonation to produce hydroperoxi de functionalities on PP [20 3, 204] are also interesting grafting methods. The goal of each process is to achieve a high grafting yield and a low incidence of side re actions, which require that the radical site s on the backbone are efficiently transformed to graft sites [68, 205]. This dissertation revolves around free radi cal grafting of polyolefins via peroxide decomposition. All reactions are done primarily in the melt state rath er than in solution for several reasons. In the past, solvents would dissolv e the polymer and high graft yields ensued, but solvents had to be dist illed for separation, and the grafted copolymer had to be dried of solvent before use. Not to mention the high so lvent loadings needed which makes the process expensive and pos es environmental i ssues [7, 206]. The idea of grafting monomers onto polyol efin backbones is not new and several patents have been issued over past several decades. The original purpose was to impart polarity, resistance to degradation, a nd improved adhesion [199, 207-211]. The monomer(s) and initiator mixtur e can be either mixed with polymer before melting or added to the melt above 120C. Grafting is done continuously in a reactor that provides intimate contact between the components (i.e extrusion) with de volatization prior to exiting the die. Typicaly, th e reactive components ar e a peroxide or az o initiator (at 0.10.7 wt%) to create the free radicals and a m onomer (at 0.5-5 wt%) with a functionality

PAGE 58

36 that can further react. A low molecular weight product with a sharp increase in flow rate is common. 2.5.2 Solid State Grafting There may be some solid state (surface) grafting for a reactive extrusion process when all materials are added at once into the extruder [94, 95, 143, 212]. For grafting below the surface, initiator decomposition ra te may be of secondary importance to monomer diffusion. But during extrusion, m onomer and initiator may become trapped into small pockets of molten polymer which are then broken as the melt is sheared. Diffusion of liquid reactants into the amorphous region (not crystalline [213]) may occur. The coalescent state of the monomer dropl et facilitates an increased rate of polymerization but also homopolymer formati on [172, 178]. The graftin g yield of PP as a solid particle is much lower compared to in the melt even though the amount of -chain scission is three orders of magnitude less. For grafting at 120C (solid state), melt flow index is over 100 times greater than at that 220C (molten state) [7, 180]. High melt strength polypropylene has been created by absorbing styrene and butadiene and subsequently polymerizing to produce a branched alloy with enhanced strain hardening, melt strength, and drawability [179]. This branched architecture can also be created by irradiation of PP w ith a difunctional monomer [214]. Higher functionality monomers were not as effective at crosslinki ng PP and caused the formation of an undesirable gel. Shorter chain monome rs are better than longer chain monomers for improving melt strength and an acryla te-based functionality was better than methacrylate at the same molecular weight beca use the reactivity of acrylate monomer is higher than methacrylate [215].

PAGE 59

37 2.5.3 Reactions in the Polymer Melt In order to minimize side reactions, the radicals formed on the polyolefin backbone must be trapped as soon as they are formed [68]. Some monomers are more effective than others, which may be due to relative so lubility of the monome rs in polyolefin melt or inherent reactivity of mono mer. By increasing grafting yields, degradation and side reactions such as homopolymerization will de crease. This strategy involves choosing a monomer combination such that the primar y monomer has a high reactivity towards free radicals and can effectively trap radicals on the polyolefin backbone and the secondary monomer facilitates this grafting process by creating branch sites or speed up the polymerization process of the primary m onomer. The higher grafting yields are attributed to: Longer chain grafts rates of copolymeri zation of electron donor acceptor pairs are greater than for hom opolymerization. More grafting sites more efficient tra pping of radical sites on polymer backbone. A straightforward method to crosslink PP was to simply add an excess amount of peroxide initiator to induce addition react ions of fragmented chains [216-219]. Generation of free radicals on PP leads to de gradation because of the low stability of macroradicals on the tertiary carbon. Cro sslinking efficiency is low due to the fragmentation of a large portion of macroradicals. Decrease of crosslinking efficiency occurs with increasing temperature because the rate of fragmentation increases (a higher energy process) compared to recombination. Th e rate of fragmenta tion depends primarily on temperature while the rate of recombina tion depends on both ini tiator concentration and the rate of its decay. The activation en ergy of recombination is close to zero while the energy of fragmentation is greater than 29 kcal/mol.

PAGE 60

38 Another, more effective method to cr osslink PP is to use a polyfunctional monomer. The idea is based on two assumptions: the double bonds of the monomer will react with tertiary PP macroradicals to suppress fragmentation and the double bonds of the monomer will increase the number of crosslinks between PP chains [68, 114, 180, 190]. Formation of stable radicals after addi tion of PP radicals is the most important role of the monomer double bonds and is a very eff ective free radical tra p. The contribution of a few long polymer chains can drasti cally affect the elong ational viscosity by increasing the degree of chain entanglement in PP and creating chains 2-3 times bigger than can be formed by intermolecular combination [215]. Ludwig and Moore found that peroxide initiated grafting of a hexafunctional coupling agent can reduce chain scission of PP dramatically [220]. The high tail end of the molecular weight distribution increased a nd was attributed to the formation of PP crosslinks through the coupling agent. Notche d Izod impact strength and tensile strength improved with the coupling agent. The decrea se in the degradation of PP is due to the primary radicals reacting preferentially with multifunctional monomer [190]. A branched PP was created by the use of a polyfunctional acrylate monomer (trimethylolpropane triacrylate) a peroxide initiator, and an iniferter compound [221]. The iniferter acts as a free radi cal initiator, chain transfer ag ent, and chain terminator and was found to facilitate the long chain bran ching from TMPTA. The iniferter also prevented gel formation and homopolymeri zation of the TMPTA, both of which are deleterious to the ultimate prope rties of the grafted PP. Se veral authors have tried to crosslink isotactic polypropylene in orde r to improve the melt extensibility and

PAGE 61

39 mechanical properties but also to redu ce chain scission while grafting functional monomers. Normally, thermal degradation of PP dur ing its processing can be avoided by adding thermal stabilizers (antioxidants). Howe ver, this kind of stabilizing mechanism cannot be applied to the graf ting process because the stabilizer eliminates the free radicals, which initiate grafting. The major pur pose of crosslinking is to both enhance the mechanical properties of grafted PP and increase its melt viscosity, which may facilitate its dispersion into other pha ses during melt blending [7, 222]. Monomers alone do not necessarily reduce degradation of polypropylene, but some studies have shown that st yrene grafting can cause cross-termination and increase branching of the copolymer [92, 94]. St yrene-GMA copolymerization and grafting can create long chain branching of PP macromol ecules [94]. The resu lting structure of grafted PS materials is more likely a highly branched, entangled network, as opposed to a crosslinked network. The entangled chain e nds behave as tempor ary junctions. Although rheological testing revealed a gel point, soxhlet extraction with xylene yielded no insoluble material. 2.5.4 Fundamentals of F ree Radical Grafting The proposed chemical mechanisms for free radical grafti ng reactions onto polymers include initiation, propagation, transf er, and termination [94]. These are the same reactions for free radical polymer ization [154, 158, 159]. Once the initiator decomposes and primary radicals are generate d, they must diffuse away from eachother. A concern when dealing with a highly viscous system is that two primary free radicals recombine to yield an inactive species, known as the cage effect If it breaks free of the cage, it may abstract a hydrogen atom from the polymer backbone, creating potential free

PAGE 62

40 radical grafting sites. Most of the functi onal monomers used for graft modification are capable of homopolymerizing, which generally enhances the ability of the monomer to graft onto the polymer [138]. Termination of the free radicals is either by kinetically identical combination or disp roportionation, which is negligible for primary radicals. Chain transfer is not a likely source for gr afting sites onto the pol ymer backbone [141]. In fact, the rate of propagation in melt grafting is over 35 times th at of chain transfer [92]. The activation energy for grafting is slightly higher than in itiator decomposition, which is the rate determining step of the reaction [223]. Few studies have probed free radical reactions under th e conditions likely to be encountered in melt phase polymer reactions. These involve relatively high temperatures, relatively high pressures, and media of relatively high viscosity [68, 138, 181]. As homopolymer is produced, there is so me phase separation; the monomer that is trapped in the homopolymer phase cannot graft onto the polymer, resulting in a lower degree of grafting and grafting efficiency th an the model system. Homopolymerization does compete successfully with the graf ting reactions in model systems [111]. The kinetics and mechanism of graf ting onto a saturated polymer (like polybutadiene) is the same as homopolymerizat ion in solution regard less of whether the polybutadiene is present [179]. Both free and grafted chains are in the same environment and presumed to grow at equivale nt rates [139-142, 224, 225]. Styrene homopolymerization readily competes with grafti ng but if benzyl acryla te is used in place of styrene, grafting is severely reduced. This is attributed to the fact that the benzyl acrylate monomer is much less reactive than styrene and the rather inactive polybutadiene macroradical cannot compete for benzyl acryla te monomer [140]. The grafting of styrene

PAGE 63

41 and homopolymerization is 10 times that of a monomer which tends to graft as single units because of styrenes high ceiling temperature [92]. When grafting polypropylene, th e ratio of tertiary to s econdary to primary grafting sites in neat solution is 5.6:1.5:1, respec tively, and 3.9:0.3:1 in benzene, respectively [150]. This study suggests that the positi on of t-butoxy radical initiated grafting in LLDPE and PP is most likely different, with LLDPE grafting mainly at the secondary sites. 2.5.5 Melt Grafting Monomers onto Polyolefins Gaylord and Mishra have found that free radical ba sed functionalization of polypropylene results in random attachment of the functional group in conjunction with degradation of PP by -chain scission [226]. In fact, some authors have found that when grafting monomers onto PP, the majority of the grafts are formed after chain scission [68, 94, 180]. Doney and Salsman were successful in pa tenting a reactive ex trusion process for creating block copolymers of is otactic polypropylene in the presence of an alkenically unsaturated polar monomer a nd a peroxide initiator [227] Polypropylene is first preferentially degraded via -chain scission by peroxide-deriv ed free radicals. This is followed by addition of a peroxide/monomer mixture to the degraded chains which therefore link up PP chains to form bloc k and graft copolymers with tailored hydrophilicity. When grafting acrylate functional mono mers onto polyolefin s, their ceiling temperatures are typically above 400C a nd homopolymerization will occur alongside grafting. Methacrylate monomers, on the other hand have ceiling temperatures approximately 200C and tend to graft as ve ry short chain branches with limited

PAGE 64

42 homopolymerization. The initiator derived primary radicals have a relatively low reactivity towards the acrylated monomer a nd preferentially abst ract hydrogen atoms from the hydrocarbon substrate. Addition of the monomer to the resulting polyolefin radical has a rate constant f our orders of magnitude faster than homopolymerization. On the other hand, styrene undergoes more rapi d addition to hydro carbon radicals [156]. Over the years, several routes have been developed to improve the grafting efficiency of functional monomers. One early process used toluene as an interfacial agent [207]. The graft yield increased for two reasons: improved solubility of the monomer in the solvent and more surface ar ea by swelling the surf ace of the polymer. The main type of PP grafting proposed at elevated temperatures is -scission followed by graft initiation on one of the chain fragments [94]. The use of a multifunctional monomer greatly reduced the exte nt of degradation but did not s eem to improve the grafting yield of GMA [7]. Hu et al. were the first to systematica lly study chemical methods for improving the free radical grafting yield onto PP while mini mizing degradation [191]. They proposed three routes, all based on incr easing the reactivity of the monomer and/or rate of reaction. The comonomer concept was derived, which is fundamentally dependent upon the reactivity ratio of free radical copolymerization. This mean s that for a monomer to be grafted onto a macromolecular chain, adding comonomer will be beneficial for improving the monomer grafting yield only if this co monomer reacts with a macroradical more rapidly than the grafting monomer and the resulting macroradical is capable of copolymerizing with the grafting monomer.

PAGE 65

43 A consensus from several authors is that grafting yields are higher when styrene comonomer is used, so styrene must preferen tially graft onto the polymer macroradical before chain scission occurs [7, 40, 92, 94, 206, 228] Styrene then forms a stable styryl macroradical which is long lived and invite s copolymerization with polar monomers. The intrinsic free radical grafti ng rate is so much accelerated by the presence of styrene that the overall grafting kine tics becomes controlled prim arily by PP melting. Another reason for the high grafting yields may be due to the fact that polyolefins such as HDPE, LLDPE, and PP are soluble in refl uxing styrene [138]. A third eff ect of styrene is that is forms a charge transfer complex with the gl ycidyl methacrylate, wh ich is more reactive than GMA alone. 2.6 Melt Grafting and In-situ Compatibilization One of the first methods to bond polypropyl ene and polyethylene was to melt blend them in an extruder in the presence of a peroxide initiator. It was expected that the chain decay of PP and the chain buildup of PE should be balanced in PP-PE blends where coupling of PP and PE macroradic als can lead to graft copolym er chains PE-g-PP. Short PP oligomers were expected to be tied to PE chains so as to pr event crosslinking [70, 229-234]. Contrary to popular belief, the pol ymers reacted fairly independent of each other. A similar effect was seen for PE-PP copolymers [91, 235, 236]. This interchain coupling is favored only in one phase blends where the compounds are in intimate molecular contact. In two phase blends, the compone nts can react only at the interface between phase domains so grafting is hindered. Although limited bonding exists between the phases, the resulti ng alloy does have higher toughness and melt strength with low gel content [232, 237]. High melt strength is characteristic of materials with long side branches (LDPE) or high molecular weight (PS) [238].

PAGE 66

44 In order to reduce or prevent the degradat ion seen in the peroxide/PP/PE blends, a multifunctional monomer was added to the mix. A recent patent focused on using a small percentage of multifunctional coagent in addition to peroxide for the compatibilization of PP and an ethylene copolymer. A very high me lt strength is observe d and degradation of PP is essentially eliminated without the form ation of substantial ge l [239]. Yoon et al. reported a tri-functional acrylate monomer (TMP TA) is effective at minimizing PP chain scission by bonding to PP macroradicals before chain scission and stabilizing them via resonance [222]. By crosslinking a ble nd of PP and LDPE, the interface becomes obscure (indicating some bonding between phases) with a 6X increase in impact strength over pure PP [233]. But it has also been shown that the multifunctional coagent only raises the crosslinking effici ency in PP, not PE [240]. A process called dynamic vulcanization can be used to crosslink rubber in a PP matrix in-situ to improve bonding and stress transf er between phases. A themoplastic vulcanizate (TPV) is thus formed, showi ng improved impact properties, solvent resistance, and long term elastic recovery over the physical bl end, but an insoluble gel is typically formed [136, 241, 242]. Dispersion w ith these alloys is typically sub-micron and dispersed phases can be as small as 30 nm in size [243]. A nano-structured polypropylene/polyamide 6 (PP/PA6) alloy has been created insitu by anionically polymerizing -caprolactam in the presen ce of functionalized PP [244246]. The resulting material had PP as the continuous matrix and PA6 homopolymer dispersed as very small domains. The reacti on is very fast and by proper control of the kinetics of both reactions, it is not difficult to control the particle size of the dispersed domains, leading to enhanced properties.

PAGE 67

45 Flaris and Baker used a unique approach to compatibilize PE and PS in an extruder [247]. A vector fluid approach was utilized, which is simply using a carrier fluid to bring a reactive ingredient to the in terface of the immiscible polyme rs. A vector fluid may be a low molecular weight material which can stay long enough in the blend to carry the reactive material (a peroxide in this case) to the interface and is then fugitive at some later point. For example, it may be drawn off at a vent port close to the end of a twin screw extruder barrel. Usi ng a viscous vector fluid led to little grafting and large dispersed domains in a compatibilized PE -PS blend. This was attributed to low dispersability of the reactiv e ingredients and a reduc tion in peroxide activity. During in-situ compatibilization, continuously new interfaces can be generated with proper mixing equipment. An appropriate vector fluid must be used so that the coupling reaction is restricted to the interface of the blend components. It must dissolve the reactive ingredient much more easily th an the polymer melt so that little reactive ingredient can diffuse into the polymer phases. If the reactive ingr edient contains both a monomer and initiator, both grafting and homopolymerization may occur. But this is likely to happen only at the surface of the minor phase polymer particles, which is surrounded with the vector fluid. In this case, the graft c opolymer formed should locate right at the interface and possibly entangle with the other blend components at the polymer-polymer interface [143]. Grafting level and interfacial adhesion are know n to be higher for vector fluids of low viscosity. Styrene monomer is used as a reactive vector fl uid and grafts onto polyethylene macroradicals. Bridging ma y occur between PE and PS due to the miscibility imparted by the grafts. The highest graft levels are found with low MW fluids

PAGE 68

46 that are not volatile, partly mi scible with the PS phase, but yet remain at the interface of PE and PS. Lagardere and Baker used the ve ctor fluid approach to compatibilize a blend of LDPE and PS in an extruder [248]. A pe roxide initiator with styrene monomer was shown to be a very effective system because bonds were created at the interface, thus reducing interfacial tension, altering the morphology, and improving overall mechanical properties. Using this approach, PS domain s were approximately 50 nm in diameter which means that the immiscible units of PS and LDPE are interconnected [249]. The insitu polymerization process yielded no crossl inked material and a broader molecular weight distribution than LD PE alone. A similar result was found by Teh and Rudin [250]. Styrene is not the only monomer us ed to compatibilize two polymers by in-situ reactive processing. A patent describes a co mposition in which an acrylate monomer, initiator, and a diacrylate are adsorbed into one or more polyolefins and reactively processed to create what th ey call a thermoplastic elasto mer [251]. The bonding between polymers and interactions between func tional groups enhances properties. In-situ graft copolymers can lead to dispersed domain sizes on the order of 50 nm [86]. HDPE and PP have been reactively processe d with a dialkyl peroxide and n-butyl methacrylate to suppress unwanted PP side reactions and improve compatibility [148]. Tangs dissertation focused on reactively extruding polyolefins with high molecular weight ethylene-olefin copolymers [11]. The all oy was created by adding a peroxide initiator, styrene monomer, and an ep oxide-functional methacrylate to PP and ENGAGE pellets as they entered the ex truder so as to graft both polymers in-situ Impact strength jumped by over 8X from the pure polymer, the morphology contained

PAGE 69

47 finer and better dispersed do mains of elastomer, and stress strain behavior improved compared to physical blends. 2.7 Conclusions The process of reactive extrusion is a complex one, to say the least. Once a suitable system is chosen (i.e. free radical poly merization mechanism using high molar mass polymers in a reactive twin screw extruder), many variables have to be defined and optimized in order to get the best possibl e performance of the alloy. The following chapters will attempt to tie in many aspects of reactive extrusion to the in-situ grafting of two polyolefins for the creation of high impact alloys.

PAGE 70

48 CHAPTER 3 PHYSICAL BLENDING OF AN IMPACT MODIFIER WITH POLYPROPYLENE 3.1 Introduction A fundamental study of the physical blends of a low molecular weight grade of ethylene-1-octene copolymer (E OC) with isotactic polypropyle ne (PP) is undertaken. Little research has been conducted using lo w viscosity ethylene-octene copolymers to toughen polypropylene, but this chapter proves that these new copolymers are effective PP modifiers up to a certain concentration. Many con cepts and characterization techniques are introduced and explained in detail so as to aid in analysis of results both in this chapter and all following chapters. 3.2 Experimental 3.2.1 Materials Table 3-1 gives a list of pertinent ethylene-1-octene copolymers produced by Dupont Dow elastomers under the tradename E NGAGE, but the EOC grade of interest is 8407 [48]. Isotactic pol ypropylene homopolymer was supplied by Equistar Chemical (grade PP 31S07A) and is contact translucent. All polymers were received in pellet form. 3.2.2 Methods 3.2.2.1 Processing All polymers were dried in an air circ ulating oven at 40C for 24 hours prior to compounding. Before processing, the resi ns were premixed by hand for about 10 minutes. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin

PAGE 71

49 Table 3-1: ENGAGE product data table. ENGAGE Grade (decreasing comonomer content) 8842 8407 8200 8401 8402 Comonomer Content wt% 13C NMR/FTIR 45 40 38 31 22 Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902 Melt Index, dg/min ASTM D1238, 190oC, 2.16kg 1.0 30 5.0 30 30 Mooney Viscosity ASTM D1646 ML 1 + 4 at 121oC 26 < 5 8 < 5 < 5 Durometer Hardness, Shore A ASTM D-2240 50 72 75 85 94 DSC melting Peak, oC Rate: 10oC/min 33 60 60 78 98 Glass Transition Temp, oC DSC inflection point -61 -57 -56 -51 -44 Flexural Modulus, MPA ASTM D-790, 2% Secant 3.5 12.1 12.1 25.8 69.9 Ultimate Tensile Strength, MPa ASTM D-638, 508mm/min 2.1 3.3 6.9 6.4 12.9 Ultimate Elongation, % ASTM D-638, 508mm/min 975 >1000 >1000 950 790 screw extruder, APV Chemical Machinery (n ow B&P Process Systems) with an L/D ratio of 39. The temperature of the extruder was regulated by electr ical resistance and water circulation in the barrels. The scre w speed, unless otherwise noted was 150 rpm. The dried, pre-mixed resins were then in troduced into the ex truder from the hopper of the extruder at 60 g/min through a screw dr iven dry material feeder from Accu Rate, Inc. Devolatilization was carried out by a vacuum pump, VPS-10A, Brooks Equipment Company. This was placed near the die and created a vacuum of about 15 in Hg. The extruder was always starved to feed. Figure 3-1 is a schema tic of the extruder, with a typical temperature profile. After compoundi ng, the resultant strands which exit the die are quenched in a water bath, pelletized, and dried in a vacuum oven at 100C, 28 in Hg for 24 hours.

PAGE 72

50 Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed) 195C 205C 210C 210C 200C 190C 180C 165C Figure 3-1: Schematic drawing of the r eactive twin screw extruder and a common temperature profile. 3.2.2.2 Mechanical properties In order to measure the strength of the materials at very hi gh testing rates, a notched Izod impact test was performed according to ASTM D256 standards. The pellets were placed in a mold with 6 slots, each measuring 0.5x0.5x2.5 in3. The mold is put in a Carver press (Fred S. Carver, Inc.) at 200C and after the material is melted, pressed up to 5000 psi. After waiting for 5-10 minutes, th e pressure is slowly increased up to 10,000 psi. After another 5 minutes, the heat is turn ed off and the sample is let to cool down to room temperature at about 1.5C/min. The bars were then taken out and notched (0.1 in deep, 0.01 in radius) with a Testing Machines Inc. (TMI) notching machine. Before testing, they were conditioned at room te mperature for 24 hours and a 30 ft-lb hammer was used with test method A on a TMI Izod imp act tester. At least 5 bars were broken and impact strength is recorded regardless of full or partial break. For stress-strain measurements, dried pelle ts were placed in a mold measuring 15 cm2 x 1 mm thick. The mold was put into the Carver press at 200C and after the

PAGE 73

51 material melts, pressed up to 5000 psi. Afte r a 5-10 minute wait, the sample was slowly pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath. Specimens were tested according to ASTM D638 standards. Type V specimens were punched out of the compression molded sheet with a die, measuring 1 0.15 mm thickness, 2.95 mm gauge widt h and 9.5 mm gauge length. Five samples were tested after conditioning at room temperature for 48 hours. The machine used to test the samples was an MTS Model 1120 Instron, using a 1000 lb load cell at a test speed of 12.7 mm/min. A Seiko DMS220 interfaced with a Seiko Rheostation model SDM/5600H was used to test dynamic mechanical specimens. Testing was conducted from -120C to 150C at a heating rate of 5C/min in dr y nitrogen atmosphere maintained at an approximate flow rate of 100 mL min-1. Rectangular samples (20x10x1mm3) were cut from the compression molded sheet and tested in bending mode at a frequency of 1Hz. 3.2.2.3 Morphology Scanning Electron Microscopy (SEM) wa s performed on a JEOL 6335F Field Emission SEM. The microscope was kept under vacuum at 1x10-5 Pa, with an accelerating voltage of 5 kV and working distance of 13.4 mm with secondary electron detection at various magnificat ions. For better phase contrast, etching was done to remove amorphous, elastomeric material. Samples were etched by the following procedure: A notched impact bar was imme rsed in liquid nitrogen for 10 minutes and immediately fractured using a TMI Izod impact tester. A section of the cryofractured surface (2 cm thick) was then immersed in xy lene (purchased from Fisher Scientific) at 60C for one hour. The sample was remove d and dried under vacuum at 40C for 12 hours. Sample mounting was on an aluminum stub with conductiv e carbon paint from

PAGE 74

52 Ted Pella, Inc. The sample was then coat ed with carbon then vacuum dried at room temperature for one hour prior to examination in the microscope. Image analysis was done using ImagePro software. Three images per sample were recorded and domain size and distribution we re quantified. The domains in all images had to be manually outlined for the software to recognize them as discreet phases. For the majority of samples, over 1000 particles were considered for diameter and roundness measurements. The number average (Dn) and weight average (Dw) diameter were determined using the following equations: where Ni represents the number of particles with diameter Di. 3.2.2.4 Thermal analysis and rheology Differential scanning calorimetry (DSC ) was used to study the different thermodynamic transitions present in the bl ends. DSC was performed on a Seiko SII DSC 220C-SSC/5200, Seiko Instruments, equipped with a Seiko Rheostation model SDM/5600H and calibrated with indium and tin standards. Samples (approx. 7 mg in weight) were sealed in crimped aluminum pans, with the reference being 99.99% pure alumina. Purging of the sample was done with dry nitrogen at a flow rate of 100 ml/min. Each sample experienced two heating and coo ling cycles (shown in Table 3-2) with the first to erase prior thermal history. The sec ond cycle is reported in all graphs. The % crystallinity is found by first in tegrating the heat flow curve to a flat baseline then dividing by the heat of fusion of a perfect PP crystal (207 J/g). (3.1) (3.2)

PAGE 75

53 Melt Flow Index (MFI) testing was done according to ASTM D1238 (230C and 2.16 kg weight) on a Tinius Olsen model MP 933 Extrusion Plastometer. For materials Table 3-2: DSC consecutive heating/cooling cycles Step Start Temp (C) End Temp (C) Heating/Cooling Rate (C/min) Hold Time (min) Sampling (s) 1 -70 200 20 3 3 2 200 -80 20 5 3 3 -80 200 10 5 1 4 200 -80 10 3 1 with a flow rate of 0.5-3.5 g/10 min, the wei ght of the samples was approx. 3 g, whereas materials with flow rates of 3.5-300 g/10 mi n, the sample weight was approx. 6 g. All materials were dried under vacuum then condi tioned at room temperature before testing. 3.3 Results and Discussion 3.3.1 Mechanical and Rheological Properties Impact strength (IS) is the ability to re sist a high loading ra te (approx. 3.6 m/s) and is one of the most important properties for pl astics part designers to consider because it sets up the worst possible condition for plastics [252, 253]. It is a critical measure of service life, product safety, and liability. The standard Izod notch ideally functions as an artificial crack because of its sharpness and acts to concentrate the applied stress, minimize plastic deformation, and direct the fr acture to the part of the specimen behind the notch. Fracture of the Izod specimen is dominated by bending-moment-induced tensile stress, but extensive pl astic deformation is possible. As can be seen from Figure 3-2, imp act strength increases, although not monotonically, with increasing elastomer content and reaches a peak at 20 wt% 8407. This increase is to be expected because the elastomer phase has a much lower glass transition temperature (Tg) than the matrix material and thus promotes various energy

PAGE 76

54 absorbing mechanisms such as cavitati on, shear yielding, and crazing. Also, in semicrystalline polymers, elastomers or r ubbers act as nucleating agents and thereby reduce spherulite diameter [103, 241, 254-258]. This reduction in spherulite size is known to improve impact strength due to an incr ease in interfacial thickness, better interspherulitic chain mobility, and reduction of spherulitic defects [254, 255, 259, 260]. Concentration of 8407 in Blend 051015202530 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 Impact Strength MFI Figure 3-2: Effect elastomer concentration on room temperature notched impact strength and melt flow index. Above 20 wt% 8407, a drastic decrease in imp act strength is observed. At this level of elastomer, the interface between phase s becomes of great importance. These two phases are only partially miscible at low elastomer concentrations, so gross phase separation may lead to brittle behavior beca use of poor stress tr ansfer between phases [254, 261, 262]. At high strain, property de terioration takes place as incompatibility leads to cracks and failure at inter-phase boundaries. Also, at high elastomer loadings, intra-spherulitic regions of PP become very diffuse and do not have the load bearing

PAGE 77

55 capacity of a fully developed spherulite [ 257, 261, 263, 264]. It should be noted that the impact strengths of samples fractured at li quid nitrogen temperatur es (for morphological analysis) were about 0.4 ft-lbs/in, rega rdless of elastomer concentration. Melt flow index is a measure of the uniform ity of the flow rate of a polymer and is not a fundamental property. It is a relative test which allows one to make inferences on molecular weight and viscosity. MFI is propor tional to molecular weight in the region of entanglement and gives a good indi cation of zero shear viscosity. From figure 3-2, one can see that the me lt flow index increases in a somewhat exponential fashion with increasing elastomer content. The blend with 30 wt% 8407 has the highest MFI, which is expected because the MFI of this elastomer is over 35 times that of PP. As melt flow index increases, impact strength is known to decrease [265]. This is only true for samples cont aining greater than 20% elastomer. In stress-strain experiments of semicrystall ine polyolefins, deformation first occurs in the amorphous phase followed by activati on of crystallographic mechanisms [105, 266-268]. For the initial slope of the force-leng th curve (2-3% strain is usually reversible and is known as the elastic modulus), the de formation of the disordered interlamellar regions (loose chain folds, tie molecules, cilia, chain entanglements, as well as completely unincorporated molecules) are involved, and the lamellar structure remains intact [269, 270]. The lamellae present in the sa mple also behave effectively as crosslink junctions and provide resistance to deforma tion [271], but some authors dispute this mechanism [272]. The initial elastic part is followed quickly by a viscoelastic part, in which the stress gradually increases to reach a maximum at the yield point [273].

PAGE 78

56 Figure 3-3 shows typical stre ss-strain responses of the physical blends while Figure 3-4 represents tensile property trends as a function of elastomer content. As can be seen from these figures, elastic modulus is a roughly additive function of blend composition and gradually declines with elastomer cont ent [274, 275, 262]. The lo w crystallinity and relatively low average molar mass of the elas tomer means that in the amorphous phase of PP, more chain ends are present (reduction in entanglement density), chain mobility is enhanced, total % crystallinity is reduced, as well as chain sti ffness. It should be noted that standard deviations for stress-s train data are given in Appendix B. Strain 0.00.51.01.52.02.53.0 Stress (MPa) 0 10 20 30 40 50 Pure PP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Figure 3-3: Stress-s train behavior of PP:8407 physical blends When a polymer reaches its yield stress (also termed proportional limit), energy barriers are overcome along with dilation and lo ng range diffusion [269]. At this point, plastic flow localizes in a neck and the st ress decreases towards a plateau value (cold drawing) [273]. The dominating mechanisms of yielding is partial or local melting,

PAGE 79

57 stretching of amorphous chains, and subse quent recrystallization, with competing mechanisms controlled by dislocation motion [105, 269, 272, 276]. Yield stress is proportional to lamellar thickness, unlike the elastic modulus. Ratio of PP:ENG8407 100 to 095 to 590 to 1080 to 2070 to 30 Yield Stress (MPa), Break Stress (MPa), Elongation at break (mm) 0 10 20 30 40 50 60 70 80 Elastic Modulus (MPa) 800 1000 1200 1400 1600 1800 2000 Energy to Break (N*mm) 0 1000 2000 3000 4000 5000 6000 7000 Yield Stress Stress at Break Elongation at break Elastic Modulus Energy to Break Figure 3-4: Stress-strai n properties of PP-8407 blends as a function of elastomer content. Yield stress, like elastic modulus, appear s to be a monotonic function of blend composition and decreases almost linearly with an increase in elastomer concentration. This is directly related to the linear decrea se in total % crystallinity of the blend with increasing 8407 content (explained in section 3.3.4). Pure PP shows a characteristically sharp yield point whereas addi tion of elastomer broadens this peak and increases the strain at yield. The reduction in yield stress with elastomer may be due to the dependence of yield stress on lamellar thickness (the greater elastomer content, the smaller the lamellar thickness). Also th e greater mobility of the amorphous region reduces the activation energy needed for cr ystallographic slip, di slocation motion, and chain disentanglement.

PAGE 80

58 The stress drop after yield (cold drawing) is associated with a decrease in unit cell volume and cross-sectional area which is a dire ct result of shearing and fragmentation of less perfect crystals [266, 277]. Necking (a form of sh ear yielding) accompanies this stress drop, where a very extensive reorgani zation of the polymer is taking place. Spherulites are broken up a nd the polymer becomes oriented in the direction of the stretch [100, 273]. The number of chain folds decreases, and the number of tie molecules between the new fibrils is increased. Deviations in post yield beha vior can be seen in Figure 3-3 for the various blends. Pure PP exhibits the largest dr op in stress but this occurs over a relatively small strain. The high crystallinity of the pure polymer a nd limited thickness of amorphous layer lends it to a greater degree of rearrangement follo wed by orientation. The effect of elastomer on cold drawing is obvious at 20% 8407, where th e stress drop is relatively small and its slope is much less than pure PP. Diffuse sphe rulites are likely, with elastomeric material existing in both interand intra-spherulitic re gions. The presence of elastomer appears to enhance this spherulite breakup process. At 20% 8407 impact strength is highest and this may be related to the high strain reached at the end of cold draw ing. The area under the stress-strain curve before strain hardening o ccurs may be related to energy absorption at high strain rates. A process known as strain hardening o ccurs post-yield and is indicated by a positive slope in the stress-strain curve. This is a result of chain unfolding and orientation involving the so-called fibrillar transformation which le ads to a continuous increase in crystallinity. Mechanical wo rk reduces the thermodynamic barrier between the metastable and stable crystals and helps chains find their way to more stable potential

PAGE 81

59 energy wells during deformation. The more defective crystals are destroyed during drawing and rebuilt into more perfect crystals with a na rrower size distribution [132, 161, 266, 269, 277-279]. The slope therefore decreases with increasing elastomer concentration, indicating PP spherulites have an increasing degree of disorder and decreasing load-bearing capability. Strain hardening doesnt occur at the 30 wt% 8407 level, so the ordering and alignment of crysta llizable chains are severely hindered by the large domains of 8407, which are highly m obile, primarily noncrystalline, and immiscible with PP. Elongation and energy at break for elas tomer modified PP are poor at high elastomer concentrations because the high interfacial energy between phases dominates failure [261, 262]. The ability to neck and dr aw is very defect or morphology sensitive, so at 30% 8407, the amorphous content may act as a defect by promoting spherulite breakup and inhibiting recrystallization. 3.3.2 Morphology Mechanical performance of the physi cal blend is highly dependent upon its morphology. For SEM imaging, etching of the elastomer phase was required because the domains could only be seen at 30% concentra tion. From Figures 35(a) thru (e), the etching procedure produces dark pits where elastomeric material once was. Another obvious trend is shown increasing elastomer concentration results in an increase in domain size and decrease in overall matrix -domain interface. In turn, the ligament thickness or distance between elas tomer particles [119] is incr eased. These ideas will be elaborated upon in the proceeding paragraphs.

PAGE 82

60 (a) (b) (c) (d) (e) Figure 3-5: SEM images of et ched, cryo-fractured surfaces of PP:8407 physical blends as a function of elastomer concentration. (a) Virgin PP at 2,000X, (b) 95:5_0 at 10,000X, (c) 90:10_0 at 10,000X, (d) 80:20_0 at 10,000X, and (e) 70:30_0 at 2,500X. The bar markers for (a) and (e) = 10 m, and for (b), (c), and (d) = 1 m.

PAGE 83

61 Table 3-3: Image analysis averages taken from etched SEM images of PP:8407 physical blends. At least three images were anal yzed from Figures 3(a) thru (e), with data compiled in Appendix D. Sample ID Dn (nm) Dw (nm) Average Roundness for particles >1* Ligament Thickness (nm) 95:5_0 77 104 1.08 166 90:10_0 113 167 1.1 194 80:20_0 224 778 1.2 305 70:30_0 269 2630 1.2 321 *Roundness is a measure of how closely the particle's shape matches that of a perfect circle. A value of 1 = a perfect circle. Table 3-3 (generated from Appendix D) represents im age analysis of the SEM pictures and reveals that a relatively narrow particle size distribution exists at low elastomer concentrations. The number average particle diameter (Dn) for the 95:5 blend is approx. 77 nm while at 90:10, Dn increases to 113 nm. Th e viscosity of the minor phase plays a huge role in determining these va lues. The matrix viscosity is much higher than the elastomer viscosit y, so coalescence is suppresse d at low 8407 concentrations. As the morphology of the blend develops in the extruder and the minor phase is elongated, dispersed, and distributed, the diffusi on process of coales cence is kinetically much slower than breakup. So, for the short residence times experien ced in the extruder, the domains remain submicron [17, 75]. Also, the particles remain relatively spherical at these low concentrations. With increasing elastomer concentration, th e anisotropy and average domain size of the particles increase. These results are typi cal of what previous researchers have found regarding PP-elastomer blends [18, 75, 128, 254, 274, 275, 280]. There also seems to be less of a monomodal distributi on of elastomer domains at higher concentrations. The smaller interfacial area of the large domain s, as well as the large distance between domains, can be directly correla ted with mechanical properties. At such high loadings of

PAGE 84

62 8407, the stress field around the elastomer domains is modified and interactions with the matrix are hindered so the blend reacts in a brittle manner to an applied load. The large size of the dispersed phase at hi gh concentrations is due to coalescence, which is the recombination of particles known to take place during the mixing process arising from forced collisions. Coalescence (collisions) of the minor phase is the rate limiting step during morphology development. 3.3.3 Viscoelasticity Dynamic mechanical behavior is studied at the molecular level and structural factors affect it, includin g molecular weight, crossli nking, crystallinity, lamellar thickness, and interfacial interaction between phases [57, 81, 235, 281, 282]. Polymers are examples of viscoelastic materials, whic h have characteristics of both viscous liquids (only dissipate energy) and elastic solids (only store me chanical energy) [94, 235]. Deformation in solid polymers is dominated by relaxation processes, which are sensitive to morphology and crystallinity [81]. A necessary condition for a highly plastic deformation is the possibility of motions of kinetic elements on a time scale similar to the deformation rate. Dynamic mechanical anal ysis (DMA) is a valuable tool in the characterization of viscoelas tic behavior, which is the m echanical behavior dependence upon time and temperature. It is able to sepa rate the viscous (loss modulus or E'') and elastic response (storage modulus or E') of the material and relates the two by Tan (E''/ E'). The first quantity to be studied is the st orage modulus (Figure 3-6), which is an indication of the stiffness of the polymer and may be considered inversely proportional to impact strength in physical blends [283]. The stiffness typically increases with increasing density (crystallinity) or entanglem ent density (from high molar mass) [284].

PAGE 85

63 The behavior in Figure 3-6 is similar to plas ticizing a brittle glassy polymer with a low molecular weight compound [285]. There are typically five regions of visc oelastic behavior for semicrystalline polymers [100]. The low temperature (gla ssy) region in Figure 3-6 occurs at temperatures less than about -45C. Long range motion is frozen in at these temperatures, so the stiffness of the blend is dependent upon the free volume and Temperature ( ) -100-80-60-40-20020406080100120 E' (MPa) 5.0e+6 5.1e+8 1.0e+9 1.5e+9 2.0e+9 2.5e+9 3.0e+9 3.5e+9 4.0e+9 4.5e+9 5.0e+9 5.5e+9 Pure PP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 C Figure 3-6: Storage modulus vs temperature for PP:8407 Blends mobility of side and end groups as well as secondary forces between chains. The long side chains of 8407 renders the blend more plia ble due to the increase in free volume. The glass transition (Tg) region is accompanied by a dr op in modulus of about 2e9 over a range of about 30C, starting at about -45 C for the elastomeric phase and 10C for the matrix phase. This is the onset of long range, coordinated molecular motion. A rubbery plateau region after Tg begins to take shape for the elastomer phase at high concentrations. Long range rubber elasti city is present in this short temperature

PAGE 86

64 range and is likely due to the entanglements associated within the amorphous phase. The rubbery flow region follows at even higher te mperatures in the ra nge of about 10-30C for 8407 and 40-70C for PP phase. In this region, the polymer is marked by both rubber elasticity (short time scale) and flow pr operties (long time scale). The response is dependent upon physical entanglements and w ith the more mobile elastomeric phase present, chains are able to move in a mo re coordinated motion leading to flow. The liquid flow region occurs at the highest temper atures and represents chain reptation out of entanglements and flow as individual molecule s. Segments are free to move from one lattice site to the other, a nd the hard polymer becomes soft and rubbery. Keep in mind that this behavior is strictly in the amor phous (non-crystalline) phase of the material. Isotactic polypropylene and ethylene-octe ne copolymers have three primary viscoelastic relaxations [81, 110, 235, 281, 282, 284, 286-288]: the relaxation around -60C for PP and -120C for 8407 is attributed to local motion of side groups, end groups, and short main chain -CH2links; relaxation located between 0C and 30C for PP and -50C to -10C for 8407, represents la rge scale (non-crystalline amorphous) chain motion due to an increase in amorphous volum e and is much broader than in wholly amorphous polymers; relaxation localized between 40C and 90C for PP and 60C to 80C for 8407 exists only in the presence of the crystalline phase and originates from diffusion of crystallographic de fects, interlamellar slip, and motions of the interfacial regions containing tie molecules, folds, loops, etc. It is obvious from Figure 3-7 that upon addition of elastomer, two relaxations are present the lowest temperature peak re presenting 8407 and the peak located about 20C, representing PP. When determining th e relative solubility between polymers,

PAGE 87

65 separate Tan peaks will exist for phase separated blends [8, 9, 81, 289]. The magnitude of each peak is characteristic of the relativ e concentrations of the two components, regardless of a dispersed or co-continuous morphology. Broadened peaks and shifts towards an intermediate temperature signify mu tual solubility of polymers. Studies have shown that PP is largely immiscible with polyethylene and polyet hylene copolymers [290-298] but at low copolymer concentratio ns, limited partial miscibility is known to Temperature ( ) -100-80-60-40-20020406080100120 Tan Delta 0.0 0.1 0.2 0.3 0.4 0.5 0.6 C -60-40-200204060 0.04 0.08 0.12 PurePP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Figure 3-7: Tan vs. temperature for PP:8407 blends from -120C to 120C. Insert represents Tan vs. temperature from -60C to 60C. exist [18, 299, 300]. The limited miscibility is obvious from the shift of the maximum 8407 peak toward higher temperatures with in creasing copolymer content. The shift to higher temperatures indicates that th e apparent activation energy of the 8407 peak slightly increases with concentration in the blends [283, 301]. There ma y also be a restriction in PP 8407

PAGE 88

66 8407 mobility due to the trapping of chains in the spherulitic structure of PP. The intensity of both 8407 and PP peaks are known to increase with more elastomer [302] or decrease with increasing degree of crystallin ity [81]. PP typically does not show major changes in peak position with crystallinity. The PP maximum becomes more of a shoulder at decreasing concen trations and overlaps with 8407. The PP peak position is also stationary for all values of elastomer. Th ese findings are similar to that of Xiao et al. [258]. Only at low elastomer concentrations can viscoelasticity of the blend be correlated with impact strength. The time scale involved is comparable to the order of magnitude as the relaxation time of viscoelastic relaxa tions (milliseconds). In elastomer toughened blends, toughness shows some correlation with the area of the Tan due to the primary or secondary transition of rubbery component. For an impact test carried out at room temperature, a material will be ductile if it contains one or more prominent subambient relaxation. [109, 281, 283, 303, 304]. Above 40C, the slope of the Tan curve increases for all samples but a trend is seen in that the slope upturn is more dras tic with increasing elastomer concentration. This should not be surprising because the enhanced mobility afforded by the elastomer enables long range motion and relaxation of chains at lower temp eratures. A broad peak/shoulder at about 80C is seen for Pure PP as well as 95:5 and 90:10 blends. This represents the relaxation of PP and dramatically increases in intensity with high elastomer concentration. The mobility of def ects in the crystalline phase as well as on the surface of the crystallites is enhanced by the presence of highly mobile amorphous elastomeric chains. There is also a slight shift in the relaxation to lower temperatures

PAGE 89

67 with increasing 8407 concentra tion, indicating a lower activa tion energy needed for chain motion. Two distinct high intensity peaks are seen at about 75C and 100C for the 70:30 blend and are indicative of the diffuse spherulitic region s of PP, thin lamellae, high defect concentration, and possibly even the onset of crystal melting [283, 305]. Large deformations may be expected to involve th e motion or deformation of crystallites or aggregates of segments. So, the high temperat ure peaks may be associ ated with the yield stress of the blends at room temperature [ 306]. The lower activation energies associated with the 70:30 blend can be related to the neck formation during a stress-strain experiment, where unstable deformati on and brittle failu re occur [105]. The loss modulus (E'') is a measure of the energy absorbed due to relaxation and is useful in clarifying the mechan isms of internal motions. From Figure 3-8, the height of each peak represents the relative mobility of the polymer chains. The peak position (at a given temperature) depends on the chemical structure, flexibility of the molecular chain, steric hindrance, and bulkiness of side groups. The high temperature side of the peak of 8407 is seen below -80C and its intensity char acteristically increa ses with increasing elastomer concentration. The Tg ( ) peak of PP is found at about 20C and with increasing elastomer concentration it decreases in temperature a nd intensity. The energy barrier for chain motion therefore decreases with increasing elastomer concentration and this peak becomes more of a shoulder simply due to the fact that the PP concentration in the blend is decreasing. The high temperature peak at about 70C decreases in intensity due to the lower PP concentration but sharpens possibly due to increased mobility and free volume from the elastomer.

PAGE 90

68 -100-80-60-40-20020406080100 E'' (MPa) 2e+7 4e+7 6e+7 8e+7 1e+8 1e+8 1e+8 2e+8 2e+8 2e+8 Temperature ( )C PurePP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Figure 3-8: Loss modulus vs. temp erature for PP:8407 physical blends It should be noted that the Tg of a polymer is commonly recorded as either the onset or peak temperature on either E'' or Tan graphs. The values reported in this section are peak temperatures. Differential scanning calorimetry (DSC) is another method to measure Tg and has shown that the Tg of 8407 is about -50C and PP about 3C. This is concurrent with the onset temperatures for the relaxations in both E'' and Tan graphs. 3.3.4 Crystallinity The thermal transitions and crystalline character of the blends have been characterized by DSC. This method measures heat flow into (e ndotherm) and out of (exotherm) the sample in relation to a referen ce at varying temperatures. For a first order thermodynamic transition like the melting temperat ure, there is a discontinuity in specific

PAGE 91

69 volume vs. temperature and the DSC therm ogram will show an endothermic peak representing the melting of crystalline material in the sample. From Figure 3-9 and Table 3-4, the -crystalline phase of isotactic polypropylene (monoclinic bravais lattice) can be identified as a peak at approximately 165C. The elastomer has a broad melting endotherm from about 30C to 80C, signifying that a small degree of fringed micelle material is present [56, 64]. An accurate measure of 8407 % crystallinity is not ava ilable due to the broad melt ing range of the copolymer. 40557085100115130145160175 Endotherm ( W) -12000 -11000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 Pure PP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Pure 8407 Temperature ( C) Figure 3-9: DSC melting endothe rm of both pure PP and 8407 as well as blends of the two polymers. The overall % crystallinity of the ble nds decreases with increasing elastomer content (Figure 3-10) because of the hindrance effect of th e melt for the arrangement of the crystallizable chai ns of PP [128, 190, 264, 307-309]. There is only a slight decrease in the % crystallinity of the PP phase with increasing elastomer concentration, which

PAGE 92

70 indicates that chain alignment and order is not significantly hindere d by large pockets of immiscible material. The heat of melting pe r unit of PP in the ble nds is independent of elastomer content, signifying that no co-c rystallization is taking place [103, 263, 310]. 28 30 32 34 36 38 40 42 44 46 4865707580859095100105 PP Concentration in Blend% Crystallinit y Total % Crystallinity PP % Crystallinity Figure 3-10: Percent crystallinity of the physical blends and pure PP as a function of PP concentration. Table 3-4: DSC endothermic data compari ng pure PP and 8407 to physical blends of PP and 8407. Standard deviation for pure PP is from an average of four samples. Ratio of PP:8407 Phase Melting Peak (C) Tm Onset (C) Enthalpy of Melting (J/g) % Crystallinity 100:0 165.9 0.4 154.5 0.5 93.9 0.7 45.4 0.4 95:5 165.5 154.8 87 42 90:10 165.2 152.9 82.3 39.8 80:20 164.7 153.9 72.6 36.6 70:30 164.4 152.5 62.5 30.2 0:100 64.5 Addition of elastomer is also known to decrease the onset of melting and peak melting temperatures which may be due to th e localization of elas tomer in the intraspherulitic regions and disturbance of spherulite regularity [128, 255, 258, 264, 294, 308 310]. Also, smaller spherulties have a lowe r heat capacity, so the melting range of the polymer blend should shift to lower temperat ures [255]. When PP crystallizes in the

PAGE 93

71 presence of amorphous, low molar mass, or immiscible material, the second component can either be incorporated in to the growing spherulite as oc clusions (no effect on growth rate) or rejected into the interspherulitic regions as deformed domains (considerable depression in growth rate) [100, 236, 255, 257, 260, 263, 297, 307, 311-313]. Similar to several other references, elas tomer commonly acts as a nucleating agent for the PP and HDPE, therefore increas ing crystallization temperature (Tc) and decreasing the average spherulite diam eter [17, 23, 24, 103, 241, 255-258, 294]. This is exemplified in Figure 3-11 and Table 3-5. The d ecrease in spherulite size is an indication of increasing rate of nucleation, likely due to enhanced mobility of PP segments [100, 310]. This reduction also reduces the inhomoge neity of the sample and thus leads to increased impact strength and elongation to break [259]. There may be local defects within the spherulite at high elastomer concentr ations which lead to weak spots or holes, thereby reducing impact energy. Temperature ( C) 010203040506070809010011012013 0 Exotherm ( W) 2000 4000 6000 8000 10000 12000 14000 16000 Pure PP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Pure 8407 Pure PP (95:5) PP:8407 (90:10) PP:8407 (80:20) PP:8407 (70:30) PP:8407 Pure 8407 Figure 3-11: DSC cooling exotherm of both pure PP and 8407 as well as blends of the two polymers.

PAGE 94

72 Table 3-5: DSC exothermic data compar ing pure PP to 90:10_0, 90:10_A, and 90:10_B. An average of 3 runs were performed on 90:10_B. Ratio of PP:8407 Peak Temperature of Crystallization (C) Enthalpy of Crystallization (J/g) 100:0 109.6 0.9 -90.7 0.6 95:5 114.9 -85.8 90:10 115.2 -82.7 80:20 116.1 -72.1 70:30 116.1 -60.6 0:100 43.6 -34.1 3.4 Conclusions By simply melt blending a low molar mass grade of ethylene-octene copolymer, a modest jump in impact strength can be seen up to 20 wt% elastomer, above which the material acts in a brittle manne r. This is directly related to the morphology of the blends, where elastomer domains increase in size and change in shape from spherical to ellipsoidal. The coalescence of the particles reduces their overall surface area so as to limit interfacial interaction with PP. Melt fl ow index increases with increasing elastomer content because the elastomer MFI is over 30 ti mes that of the matrix phase. This low molar mass elastomer affects both the crysta lline and amorphous phases which in turn deleteriously affect stress-strain performance. Viscoelastic analysis confirms the thought that PP and 8407 are largely immiscible phases. The low temperature peak of 8407 signi fies chain mobility at low temperatures, which is a good indication of impact streng th performance. The sheer speed of the impact test renders pure PP brittle but the bl ends ductile because elastomer chains are able to relax and diffuse in response to the high stress before breaking. The crystallinity of the blend is reduced with addition of 8407, but the % crystallinity of PP remains the same regardless of the amount of elastome r. Also, the crystallization temperature

PAGE 95

73 increases with increasing elastomer conten t because of the hete rogeneous nucleation effect and enhanced molecular mobility.

PAGE 96

74 CHAPTER 4 TOUGHENED POLYPROPYLENE BASED ALLOYS 4.1 Introduction The focus of this dissertation is based on th e results obtained in this chapter, which elaborates on the fundamental differences between PP alloys and physical blends. It will be shown that modification on the nano-scal e is necessary to achieve performance enhancements previously unattainable. Many different toughening mechanisms are thought to be involved in improving the impact strength of PP via alloying with an elastomer. This chapter will tie in experi mental results to many references and will shown that a novel material has indeed been created. When modifying a polymer via reactive blending, the resulting mechanical properties depend on a foundation of controlled morphology, rheology, crystallinit y, and chemical structure. A thorough explanation of the alloys im pact strength and stress strain properties will be followed by morphological, chemical, rheological, and crystallographi c characterization. 4.2 Experimental 4.2.1 Materials Table 4-1 gives a list of pertinent ethy lene-1-octene copolymers (EOCs) produced by Dupont Dow elastomers under the tradename ENGAGE, but the grade of interest is 8407 [48]. Isotactic polypr opylene homopolymer was supplied by Equistar Chemical (grade PP 31S07A) and is contact translucent. All polymers were received in pellet form. Polyisoprene (NATSYN) was donated by Goodyear The peroxide and monomers used in this study were reagent grade chemicals (structures are shown in Table 4-2). The

PAGE 97

75 Table 4-1: ENGAGE product data table. ENGAGE Grade (decreasing comonomer content) 8842 8407 8200 8401 8402 Comonomer Content wt% 13C NMR/FTIR 45 40 38 31 22 Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902 Melt Index, dg/min ASTM D1238, 190oC, 2.16kg 1.0 30 5.0 30 30 Mooney Viscosity ASTM D1646 ML 1 + 4 at 121oC 26 < 5 8 < 5 < 5 Durometer Hardness, Shore A ASTM D-2240 50 72 75 85 94 DSC melting Peak, oC Rate: 10oC/min 33 60 60 78 98 Glass Transition Temp, oC DSC inflection point -61 -57 -56 -51 -44 Flexural Modulus, MPA ASTM D-790, 2% Secant 3.5 12.1 12.1 25.8 69.9 Ultimate Tensile Strength, MPa ASTM D-638, 508mm/min 2.1 3.3 6.9 6.4 12.9 Ultimate Elongation, % ASTM D-638, 508mm/min 975 >1000 >1000 950 790 monomers were purified by passing through an activated alumina column before use. Styrene monomer, inhibited by 10-15 ppm t-but yl catechol, was purchased from Fisher. The initiator, 2,5dimethyl-2,5-di-(t-butylper oxy) hexane, was purchased from Atofina under the tradename Lupersol 101. Diethylenegl ycol diacrylate (D EGDA), inhibited by 80 ppm Hq and 120 ppm MEHQ, and trimethylol propane triacrylate (TMPTA), inhibited by 125 ppm HQ and 175 ppm MEHQ were graciously donated by Sartomer, an Atofina company. Irganox B215 was purchased from Ciba specialty chemicals. Table 4-2: Structures of reactive materials of interest Name Lupersol 101 DEGDA Structure

PAGE 98

76 Table 4-2 Continued Name Styrene TMPTA Structure 4.2.2 Methods 4.2.2.1 Processing All polymers were dried in an air circ ulating oven at 40C for 24 hours prior to compounding. Before processing, the resi ns were premixed by hand for about 10 minutes. Monomer/initiator mixtures were magnetically stirred for 10 minutes before processing and a choice amount of the mixtur e was added to the dry polymer pellets before processing. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin screw extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D ratio of 39. The temperature of the extruder was re gulated by electrical resistance and water circulation in the barrels. The dried, premixed resins were then introduced into the extruder from the hopper of the extruder at 60 g/min through a screw driven dry material feeder (Accu Rate, Inc). A Zenith pump cont rolled the rate of monomer/initiator solution addition into the extruder. The screw spee d, unless otherwise not ed was 150 rpm. Devolatilization was carried out by a v acuum pump, VPS-10A, Brooks Equipment Company. This was placed near the die and created a pressure of about 15 in Hg. The extruder was always starved to feed. Figure 4-1 is a schema tic of the extruder along with a typical temperature profile. After compoundi ng, the resulting strands which exit the die

PAGE 99

77 are quenched in a water bath, pelletized, and dried in a vacuum oven at 100C, 28 in Hg for 24 hours. Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed) 195C 205C 210C 210C 200C 190C 180C 165C Figure 4-1: Schematic drawing of the r eactive twin screw extruder and a common temperature profile. 4.2.2.2 Mechanical properties In order to measure the strength of the materials at very hi gh testing rates, a notched Izod impact test was performed according to ASTM D256 standards. The pellets were placed in a mold with 6 slots, each measuring 0.5x0.5x2.5in3. The mold is put in a Carver press (Fred S. Carver, Inc.) at 200C and after the material melts, pressed up to 5000 psi. After waiting for 5-10 minutes, th e pressure was slowly increased up to 10,000 psi. Five minutes later, the heat was turned off and the sample was let to cool down to room temperature at about 1.5C/min. The bars were then taken out and notched with a Testing Machines, Inc. (TMI) notching machin e. Before testing, they were conditioned at room temperature for 24 hours and a 30 ft -lb hammer was used with test method A on a TMI Izod impact tester. At least 5 bars were broken and impact strength is recorded regardless of full or partial break.

PAGE 100

78 For stress-strain measurements, dried pelle ts were placed in a mold measuring 15 cm2 x 1 mm thick. The mold was put into th e Carver press at 200C and after the material melts, pressed up to 5000 psi. Afte r a 5-10 minute wait, the sample was slowly pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath. Specimens were tested according to ASTM D638 standards. Type V specimens were punched out of the compression molded sheet with a die, measuring 1 mm thickness, 2.95 mm gauge width and 9.5 mm ga uge length. Five samples were tested after conditioning at room temperature for 48 hours. The mach ine used to test the samples was an MTS Model 1120 Instron, using a 1000 lb load cel l at a test speed of 12.7 mm/min A Seiko DMS220 interfaced with a Seiko Rheostation model SDM/5600H was used to test dynamic mechanical specimens. Testing was conducted from -120C to 150C at a heating rate of 5C/min in a dr y nitrogen atmosphere maintained at an approximate flow rate of 100 mL min-1. Rectangular samples (20x10x1mm3) were cut from the compression molded sheet and tested in bending mode at a frequency of 1Hz. 4.2.2.3 Morphological characterization Scanning electron microscopy (SEM) was performed on a JEOL 6335F Field Emission SEM. The microscope was kept under vacuum at 1x10-5 Pa, with an accelerating voltage of 5 kV a nd secondary electron detection at various magnifications. Two different etching procedures were perf ormed for better phase contrast one using fractured impact bars and th e other using compression molded films. The etching was done to remove amorphous (or elastomeric) ma terial. Impact bar samples were etched by the following procedure: A notched impact bar was immersed in liquid nitrogen for 10 minutes and immediately fractured using a TMI impact tester. A section of the cryofractured surface (2 cm thick) was imme rsed in xylene (purchased from Fisher

PAGE 101

79 Scientific) at 60C for one hour. The sample was removed and dried under vacuum at 40C for 12 hours. Compression molded thin films were etched so as to view the lamellar morphology of the polymers. They were produced by hea ting the Carver press up to 200C, pressing samples to desired thickness (about 10 m), and cooling at a rate of about 30C/min. The films were then etched according to the pr ocedure originally re ported by Olley et al. [314] and duplicated by various other aut hors [315-318]. The following materials and their concentrations were purchase from Sigm a Aldrich and used as received: Potassium permanganate (1.3 wt%), phosphoric acid (32.9 wt%), and sulfuric acid (65.8 wt%). Potassium permanganate crystals were sl owly added to a phosphoric/sulfuric acid solution in a conical flask and stirred until they all dissolved, producing a dark green color. Dissolving the crystals took over an hour. The sample s were then immersed in the permanganic acid solution for five hours. Th e next consecutive steps include: washing with a mixture of 2:7 sulfuric acid:distille d water which has been cooled to near its freezing point, decanting the solution, washing w ith hydrogen peroxide (35% solution in water, purchased from Acros Organics) for two minutes, washing with distilled water several times over two minutes, and washing with acetone (Fisher Scientific) for two minutes. Sample mountings for SEM were on alum inum stubs with conductive carbon paint from Ted Pella, Inc. The sample was then co ated with carbon (for impact bar specimens) or a thin layer of gold-palladium (for thin fi lms) then vacuum dried at room temperature for one hour prior to examination in the microscope.

PAGE 102

80 Image analysis was done using ImagePro so ftware. Domain size and distribution were quantified, but the domains had to be manually outlined for the software to recognize them as discreet phases. For the majority of images, over 1000 particles were considered for diameter and roundness m easurements. The number average (Dn) and weight average (Dw) diameter were determined using the following equations: where Ni represents the number of particles with diameter Di. For transmission electron microscopy (TEM ), Samples were embedded in a room temperature cure epoxy, using a flat silicon mold, and let cu re overnight. These were trimmed with a razor blade to expose as small an area as possible and then cryoultramicrotomed on a Reichert UltraCut at -110C using a 35 diamond knife. Sections were approximately 70-100 nm thick. Ethanol was used in the knife boat to float the sections and the sections were picked up usi ng 600 mesh copper grids. These grids were stained with a 0.5% aqueous solution of ru thenium tetroxide, pur chased from Electron Microscopy Sciences, for 30 minutes. After letting the grids outgas for about 24 hours, the sections were imaged with a Philips 420T TEM at 100 kV. In order to assess the crystalline structure of the polymer samples, wide angle x-ray diffraction (WAXD) was used. This was done on a Phillips MRD Xpert high resolution XRD using Cu k radiation at a wavelength of 1.54056n m. The scan rate was 6/min over a 2 range from 5 to 50. (4.1) (4.2)

PAGE 103

81 A polarized light microscope (PLM), Ol ympus model BX60 with CCD camera and quarter wave plate attachments, was used to study the crystal line morphology of the materials as they cooled from the melt. A hot stage (Linkam TMS 91) allowed for crystallization studies on ~5 m thick compression molded samples. Images were captured using Scion software. Table 4-3 give s the heating and cooling cycles for each specimen. The first cycle was performed to erase previous cooling history and the second cycle is reported in the results and discussion section. Table 4-3: PLM consecutiv e heating/cooling cycles Step Start Temp (C) End Temp (C) Heating/Cooling Rate (C/min) Hold Time (min) 1 30 200 10 5 2 200 30 10 5 3 30 200 10 5 4 200 30 10 For stress whitening qualification of impact bars, digital images were taken postfracture using a Kodak EasyShare CX7300. 4.2.2.4 Chemical composition and molecular structure Fourier Transform Infrared Spectrosc opy (FTIR) was performed on a Nicolet 20SXB Spectrometer. 256 Scans were taken from 3500 to 500 cm-1 wavenumbers with a resolution of 4. Measurements were done in transmission mode on thin films (~2-3 m). Films were produced by melting the polymer in a Carver press at 180C and 10,000 psi for 2 minutes then quenching in a water bath at room temperature. For quantification of the grafted styrene, a calibrati on curve had to be established as explained in Appendix A. Areas under the 700 cm-1 and 1376 cm-1 peaks were compared and related to absolute styrene amounts to get the grafting efficiency (GE). This is defined by the following equation:

PAGE 104

82 (4.3) Amount of Monomer Gr afted to Polymer Backbone GE = X 100% Original Amount of Mo nomer Pumped into Extruder For FTIR, all reactively extruded materials must be purified before quantification of grafting yield. This first involved dissolv ing the pellets of crude graft copolymer in hot xylene at a concentration of ca 4% (wt/vol ). The hot solution was precipitated into ten volumes of acetone (a known non-solvent for the LLDPE, HDPE and PP, and a solvent for styrene monomer and homopolymer based on solubility parameters [162]). The unreacted monomers, styrene, a nd DEGDA or TMPTA homopolymers and copolymers remained soluble in acetone a nd were separated out from the grafted polyolefins. The precipitated graft modified alloy was filtered, washed, and then vacuum dried at 70C for 24 hours. FTIR showed that the GE level remained unaltered upon further rounds of purification. Therefore, one purification step was sufficient for removal of all the residual impurities. Gel permeation chromatography (GPC) was performed on a Waters GPCV 2000 calibrated using crosslinked pol ystyrene standards for relative molar mass determination. The set temperature was 40C with THF as the solvent at a flow rate of 1.0mL/min. Samples (about 9 mg) were dissolved in 9.5 ml (8.44 g) of HPLC grade THF purchased from Acros Organics and passed through 0.45 m filters before analysis. 4.2.2.5 Thermal analysis and rheology Differential scanning calorimetry (DSC ) was used to study the different thermodynamic transitions present in the bl ends. DSC was performed on a Seiko SII DSC 220C-SSC/5200, Seiko Instruments, equipped with a Seiko Rheostation model SDM/5600H and calibrated with indium and tin standards. Samples (approx. 7 mg in weight) were sealed in crimped aluminum pans, with the reference being 99.99% pure

PAGE 105

83 alumina. Purging of the sample was done with dry nitrogen at a flow rate of 100 mL/min. Each sample experienced two heati ng and cooling cycles (shown in Table 3-2) with the first to erase prior thermal history. The second cycle is repo rted in all graphs. The % crystallinity is found by fi rst integrating the heat flow curve to a flat baseline then dividing by the heat of fusion of a perfect PP crystal (207 J/g). Table 4-4: DSC consecutiv e heating/cooling cycles Step Start Temp (C) End Temp (C) Heating/Cooling Rate (C/min) Hold Time (min) Sampling (s) 1 -70 200 20 3 3 2 200 -80 20 5 3 3 -80 200 10 5 1 4 200 -80 10 3 1 Thermal Gravimetric Analysis (TGA) was performed on a select few samples using a SEIKO TG/DTA 320. This measures weight loss as a function of temperature with a typical heating run from 25C to 1000C at 10C/min. To determine the gel content of the alloys, ASTM D2765 test method A was followed and repeated three times for the materials of interest. Samples weighing 0.300 0.015 g were placed in a 120-mesh stainless steel cloth pouch and immersed into a round bottom flask through a reflux condenser. The 500 ml flask was filled with 350 g of xylene and 1% antioxidant (Irganox B215 from Ciba Specialty Chemicals) was added to it. Xylene was boiled at 140C and refluxi ng was done for 12 hours. After extraction, the samples were immediately placed in a v acuum oven preheated to 150C and dried at 28 in Hg for 24 hours. The difference in wei ght between the sample + cage before and after extraction is reported. Melt Flow Index (MFI) testing was done according to ASTM D1238 (230C and 2.16 kg weight) on a Tinius Olsen model MP 933 Extrusion Plastometer. For materials

PAGE 106

84 with an flow rate of 0.5-3.5 g/10 min, the we ight of the samples was approx. 3 g, whereas materials with flow rates of 3.5-300 g/10 mi n, the sample weight was approx. 6 g. All materials were dried under vacuum then condi tioned at room temperature before testing. The extensional flow properties of the melt were investigated using a Gttfert Rheotens tester operated at a constant acceleration of the ro llers (120 mm/s). A polymer melt strand is extruded vertically downward s from a capillary die and drawn by rotating rollers, with the stretched polymer melt unde rgoing uniaxial extension. The force at which the polymer melt breaks is designated the melt strength. 4.3 Results and Discussion ENGAGE polyolefin elastomers have an advantage over polyisoprene (PI), polybutadiene (PB), or elastomers contai ning unsaturated bonds because ENGAGE possesses no main chain unsaturation, theref ore oxidation and disc oloring are limited. An argument may be made that PI may be bett er suited for this reactive extrusion process than ethylene-olefin copolymers because the peroxide initiated grafting is expected to be facilitated by the presence of double bonds in the polymer backbone. A typical formulation of 0.3 wt% initiator, 6 wt% styren e, and 0.8% DEGDA, with a PP:PI ratio of 80:20 was reactively extruded a nd analyzed by FTIR to detect whether all PI double bonds had reacted. Figure 4-2 s hows both a full scan from 500 cm-1 to 3300 cm-1 and an insert of the peaks of interest. The pres ence of styrene, 8407, and polypropylene caused overlapping of some peaks, but some characte ristic peaks have been identified. The carbonyl stretch from DEGDA represents the peak at 1740 cm-1 and the 1600 cm-1 peak is due to the C=C aromatic stretch from the phenyl pendant group of styrene. The peak at 1660 cm-1 is most significant because it represents an alkenyl C=C stretc h, characteristic of the polyisoprene main chain unsaturati on [319, 320]. This signifies that the free

PAGE 107

85 radical grafting process does not consume ev ery double bond in the backbone of PI and therefore unsaturated ru bbers are not suitable alternatives to fully saturated elastomers. Figure 4-2: FTIR image of PP:polyisoprene reactive blend. The elastomer of interest is 8407, havi ng a combination of low viscosity and crystallinity compared to many commercially available ethylene-octene copolymers. A systematic study has been undertaken which en compasses the effect of reactive materials (initiator, styrene, and multifunctional monome r) in alloys with varying amounts of 8407. To simplify referencing samples, a code has been established for each formulation (Figure 4-3 and Table 4-5). 95:5_A Ratio of PP:8407 Designated Formulation from Table 4-5 Figure 4-3: Repr esentation of sample reference code The weight % is defined as the percentage of material added in relation to the total weight of all ingredients. For example, a PP:8407 ratio of 95:5, 0.3 wt% Initiator and 6

PAGE 108

86 wt% styrene means that out of a total of 100 grams, Initiator = 0.3 g, Styrene = 6 g, PP = 89 g, 8407 = 4.7 g. Table 4-5: Identification of formulations Alloy ID wt% Initiator wt% Styrene Multifunctional Monomer (0.8 wt%) 0 (Physical Blend) 0 0 A 0.3 6 B 0.3 6 DEGDA C 0.3 6 TMPTA 4.3.1 Notched Izod Impact Analysis A quick comparative measure of th e toughness of thermoplastics at high deformation rates is the Izod impact test. Figure 4-4 shows that by physical blending 8407 with PP, impact strength is marginally improved. The same holds true when a peroxide initiator and styren e monomer are added to the mi x. It should be noted that impact bars of pure PP and 95:5_A completely broke, but all others were only partial breaks. Addition of 0.8 wt% multifunctional monomer (either dior tri-functional) improves the room temperature notched Izod imp act strength dramatically. The greatest jump in impact strength is seen at a PP:8407 ratio of 80:20, having over a 13X improvement over pure PP. This result is indicative of a modified material with enhanced energy absorption and dissipative mechanisms. The high functionality of DEGDA and TMPTA has several effects on th e alloys. They incorporate a greater degree of entanglements and intercrystalline tie molecules, better efficient trapping of radicals on the polymer backbones, homogeneit y on the micron scale, and they limit the degree of PP degradation. The high deformation rate leads to adiabati c heating in the sample (upwards of 7080C at the surface of a rubber toughened polyp ropylene [116]) which means an increase

PAGE 109

87 in free volume in the amorphous phase and an increase in Tg [129, 242, 321]. More free volume means greater chain mobility, a high degree of grafting, and higher impact strength. Another possibility is the idea of crystalline phase transformation toughening, where melting (energy absorption) and recrystall ization (dissipation of energy) take place as a result of the high temperatures [129, 132, 134]. The phase transformation may be associated with the retractive (elastic) behavi or of the impact bars. Defective or small crystals are destroyed and r ecrystallized into an orient ed thermodynamically stable Sample ID 95:5_90:10_80:20_70:30_ Notched Izod Impact Strength (ft-lbs/in) 0 2 4 6 8 10 12 14 0 A B C Figure 4-4: Room temperature notched Izod im pact strength of PP blends and alloys. Reference material is vi rgin PP at 0.99 ft-lbs/in. phase [287, 322-327]. In other words, a type of strain hardening is experienced. Another phenomenon could be that the elastomer phase experiences a large extension of its chains along with a rise in temperatur e. By applying rubber elasti city theory, a decrease in entropy is opposed by a more energetically favorable ra ndom coil conformation [100,

PAGE 110

88 161]. So during notched impact tests, imp act energy is absorbed primarily during the bending of the specimens (i.e. under tensile deformation) [252, 253]. The tri-functional acrylate gives the highest impact strength at high concentrations of elastomer. TMPTA has a higher grafting efficiency than DEGDA (from its greater functionality), which will be shown in the ne xt section. A more network-like structure may be formed with TMPTA present, therefore improving bonding between phases (improved stress transfer) and increasing th e retractive force of the alloy. At low elastomer concentrations, grafting, crosslinki ng, and entanglements may be so high with TMPTA that the impact strength is not as significantly affected. Figure 4-5: Izod Impact test specimens postfracture. Left Pure PP, Middle 90:10_0, Right 90:10_B. Figure 4-5 shows that stress whitening occu rs when 10% elastomer is added to PP, regardless of whether it is an alloy or physical blend. This is commonly seen in rubber toughened polymer and is attributed to the di lation and cavitation of rubber domains as well as shear yielding and crazing [103, 328]. Pure PP was one of a few materials that broke completely in this test. For the partia lly broken samples, the crack traveled further for the physical blend than reactive blend, wh ich may be due to the increased elastic restorative force caused by the multifunctional monomer.

PAGE 111

89 A corollary to Figure 4-5 would be to use SE M to probe the actual crack tips of the physical and reactive blends. From Figures 4-6 and 4-7, there is a clear difference between the cracks tips for the partially broken samples. An important feature is the shape of the crack tip itse lf and morphology of immediat ely surrounding material. For the physical blend (Fig. 4-6), the crack tip is blunted and curved. There is little Figure 4-6: SEM image showing the tip of an arrested crack from a room temperature notched Izod impact test of 90:10_0. Insert is a magnified image of crack tip. stretching and drawing of material in to the crack front, which spans over 10 m. Elongated voids are present at the crack surf ace and are oriented approximately parallel to the fracture surface, a feature also found by Wei and Sue [329]. For the alloyed polypropylene (Fig. 4-7), the crack reta ins its ellipsoida l shape (approx. 2 m in diameter) and there appears to be a high degree of yielding and melting of material

PAGE 112

90 completely surrounding the crack tip. The uniform shape of the tip may be related to the homogeneous distribution of sub-micron elasto mer particles and thin lamellar crystals. As a crack propagates through the sample, it encounters a large number of extremely small energy absorbing particles which do not blunt the tip as typical large rubber particles would. Figure 4-7: SEM image showing the tip of an arrested crack from a room temperature notched Izod impact test of 90:10_B. The morphology of room temperature fract ured surfaces of pure PP and an alloy are also remarkably different. Figure 4-8 is a comparison between fracture surfaces of brittle PP (a) and tough 80:20_C (b), imaged at the center of the impact bars. The most striking feature is the coarse morphology in the toughened material compared to the relatively flat surface of the brittle material. A flat surface is indicat ive of a fast, brittle fracture, whereas an extremely rough surface indicates energy dissipation and absorption leading to a tough response to the impact test As previously noted, adiabatic heating takes place at the surface of the sample [21, 130, 131, 133]. The rise in temperature

PAGE 113

91 results in a void-free relaxation layer, which is due to a high degree of plastic deformation. The existence of this relaxa tion layer indicates an increase in impact strength and slows the propagation of a crack. Figure 4-8: SEM image of 80:20_C (left) and Pure PP (right) fr actured at room temperature without etching. Each image is located at the center of the impact bar, with magnification = 2,000X and marker bar = 10 m. It should be noted that the impact strengt hs of samples fractured at liquid nitrogen temperatures (for morphological analysis) were around to 0.4 ft-lbs/in, regardless of elastomer concentration. 4.3.2 Stress-Strain Behavior Stress-strain measurements have been conducted on all of the previous impact strength samples. These low deformation rate tests give a different perspective as to how the material responds to a tens ile load. Figures 4-9 thru 412 represent the stress-strain behavior of these materials, with Appe ndix B showing both st atistical data and representative plots of each sample. The low strain portion of the stress stra in curve is dominated by the amorphous phase response to the applied load. When comparing each of these graphs, the elastic modulus is highest for the samples c ontaining multifunctional monomer at all concentrations of elastomer. This is intuitive because of the greater degree of

PAGE 114

92 entanglements and intercrystalline tie molecules afforded by the multifunctional monomer. At a constant percent crystallin ity, the smaller the crystallite size (which generally implies more tie molecules), the higher the modulus [330] Section 4.3.7 will show that unusually small lamellar crystals exist for the alloys. The amorphous component of semicrystalline polymers, alt hough macroscopically isotropic, can consist of microor submicron domains with consid erable degree of order [105]. There is an amorphous layer (fold surface) which increases in thickness because defects locate in this area [331]. Grafted material cannot crystall ize and thus may resi de at the surface of lamellar crystals rather than be incorporated into them. Typicall y, elastic modulus and yield strength increase upon gr afting of monomer [184, 222, 230, 250]. Yield stress remains approximately the sa me between the physical blends and alloys containing only styrene a nd initiator, but is slightly higher for samples containing Sample ID 0 (physical Blend)ABC Break Stress (MPa) and Elongation at Break (mm) 0 10 20 30 40 50 60 70 Elastic Modulus (MPa) 1680 1700 1720 1740 1760 1780 1800 1820 1840 1860 1880 Energy to Break (N*mm) 4000 4200 4400 4600 4800 5000 5200 5400 5600 5800 Yield Stress (MPa) 32 33 34 35 Stress at Break Elongation at Break Elastic Modulus Energy to Break Yield Stress Figure 4-9: Stress-strai n properties of 95:5 alloys compar ed to the 95:5 physical blend.

PAGE 115

93 0 (physical Blend)ABC 0 10 20 30 40 50 60 70 1400 1450 1500 1550 1600 1650 1700 1750 3600 3800 4000 4200 4400 4600 4800 Yield Stress (MPa) 28 29 30 31 32 33 34 Sample IDBreak Stress (MPa) and Elongation at Break (mm)Elastic Modulus (MPa) Energy to Break (N*mm) Stress at Break Elongation at Break Elastic Modulus Energy to Break Yield Stress Figure 4-10: Stress-strain prope rties of 90:10 alloys as comp ared to the 90:10 physical blend. 0 (physical Blend)ABC 0 10 20 30 40 50 60 1100 1150 1200 1250 1300 1350 1400 2600 2800 3000 3200 3400 3600 3800 Yield Stress (MPa) 23 24 25 26 27 Sample IDBreak Stress (MPa) and Elongation at Break (mm)Elastic Modulus (MPa) Energy to Break (N*mm) Stress at Break Elongation at Break Elastic Modulus Energy to Break Yield Stress Figure 4-11: Stress-strain prope rties of 80:20 alloys as comp ared to the 80:20 physical blend.

PAGE 116

94 0 (physical Blend)ABC 0 5 10 15 20 25 900 920 940 960 980 1000 1020 1040 1060 1080 1100 1120 100 300 500 700 900 1100 1300 Yield Stress (MPa) 16 17 18 19 20 21 22 Sample IDBreak Stress (MPa) and Elongation at Break (mm)Elastic Modulus (MPa) Energy to Break (N*mm) Stress at Break Elongation at Break Elastic Modulus Energy to Break Yield Stress Figure 4-12: Stress-strain prope rties of 70:30 alloys as comp ared to the 70:30 physical blend. multifunctional monomer. The higher yield stress may signify a greater energy barrier to molecular motion, lamellar unfolding, and orienta tion. This may also be connected to the unique cross-hatch structure of the alloys. The density of tie molecules affects the deformation mode of lamellar blocks (s lippage, breakup, defolding) and thus the macromolecular rearrangements in the crack tip [134]. From Figure 4-9, energy to break, elongati on at break, and stress at break are all greater for the physical blend than of the high impact strength alloys, a trend not seen at higher concentrations of elastomer. This may be due to the fact that multifunctional monomer cannot completely prevent degrada tion of PP, so small non-crystalline PP chains will fail beyond a certain extension. Al so, slight crosslinking may exist in the elastomer phase, creating a phase separated gel which is known to decrease these tensile properties. Elastic modulus and yield st ress are highest for the DEGDA-containing samples. This is somewhat surprising give n the higher functionality of TMPTA. This

PAGE 117

95 result may be due to the greater overall number of DEGDA molecules (about 1.4:1 molar ratio) and therefore great er number of branch points homogeneously distributed throughout the material. Also, the lower visc osity of DEGDA allows better diffusion in the highly viscous PP melt. At the 90:10 level, the DEGDA containing sample shows the best overall stressstrain performance (Figure 410). In the physical blend, the higher elastomer content leads to a greater degree of phase sepa ration, larger elastomeric domains, altered spherulitic morphology, and greater amorphous inte rfacial thickness. At higher levels of 8407, the elastic modulus increases in magnitude to a greater extent with multifunctional monomer present. This may be an indicati on that the PE phase is imposing a greater influence on the free radical grafting pro cess and a high degree of branches or entanglements may be formed due to excess free radicals. The increase in elastomer concentration has a much more drastic effect on elastic modulus for 70:30_C than 70:30_B. As can be seen in Figure 4-12, the elastic modulus of 70:30_C actually surpasses that of 70:30_B. One may vi sualize that the 8407 phase melts first, acting as a cage by trapping liqui d reactants. Once the initiator decomposes, the 8407 phase is affected to a greater degree than PP, and because of the high functionality of the TMPTA molecule, cros slink density will be higher. At low 8407 concentrations the grafting reac tion happens so fast that the less viscous, less sterically hindered molecule (DEGDA) compensates for the speed. For high elastomer concentrations, radicals are longer lived in the 8407 phase, therefore increasing the probability of being trapped by TMPTA vinyl groups.

PAGE 118

96 The strain hardening for 70:30_0 and 70:30_A are poor (similar to impact strength results), which indicates the ease in which crystalline lamellae are destroyed and recrystallized. The temperatur e rise experienced from adia batic heating, low interfacial energy, and large elastomer domains may all contribute to the poor performance. 70:30_B has a low strain at break and ener gy to break likely due to the greater entanglement or crosslink dens ity of the sample. The sa mple containing DEGDA must limit both the degradation of PP and crosslinking of PE while facilitating a highly branched, diffuse macromolecular structure with crystallizable PP chains in close proximity to one another. A higher tie chai n density indicates an increasing ability to strain harden [132, 332-334]. And a semblan ce of order before strain hardening may facilitate a stress-induced pha se transformation of crystall ine phase stabilization [132, 134]. A material with a less dense crysta l structure and thicker amorphous layer of connecting chains typical of phase of PP, has a higher i nherent ductility and overall macroscopic toughness than the phase [278, 287, 332, 333, 335-340]. Smaller blocks of crystallites translate easier than larger ones and therefore orient preferentially in the direction of deformation lead ing to greater toughness [260] Also, elongation at break and hence energy to break has already been prov en to increase with increasing degree of grafting [7, 250]. For all stress-strain plots, reactive ble nds not containing multifunctional monomer show poor properties in relation to all other materials, including the physical blends. PP is known to undergo -chain scission during melt free ra dical grafting, so isotactic high molecular weight chains that may otherwise be able to crystallize are chopped up and rendered defective or amorphous material. Thus, their contribution to stress strain

PAGE 119

97 performance is a negative one. These materi als have lower percent crystallinity and higher MFI, as will be shown in the follo wing sections, so the energy barriers for molecular motion are reduced. The degradation and lower molecular weight are analogous to adding an excess amount of pe roxide to the system, which reduces the overall crystallinity [148, 220, 233, 341] as well as melting temperature [222, 233]. 4.3.3 Grafting onto Polyolefins FTIR is a very versatile technique which determines the presence of certain types of bonds and functional groups in an orga nic molecule [319, 320]. Molecules are excited to a higher energy state when they absorb infrared (IR) radiation, corresponding to stretching and bending frequencies of covalent bonds and unique to every compound. A brief explanation of the peaks present fo r a typical toughened PP alloy is needed to understand what groups are present and how they respond to IR radiation. Figure 4-13 is a typical FTIR image containing PP, 8407, in itiator, styrene, and multifunctional monomer. The relevant peaks are characte rized according to the strength and type of deformation of the bond and location with regards to frequency. The purpose of using FTIR is to both qua ntify the amount of styrene grafted onto the polyolefin backbone and verify the exis tence of grafted multifunctional monomer. Grafting efficiency, as descri bed in the experimental sect ion and Appendix A, is shown in Figure 4-14. This figure is direct eviden ce that PP is more likely to be grafted over 8407 because GE decreases almost linearly with increasing 8407 content. Chapter 2 explained this decrease as a result of the higher bond dissoci ation energy of primary and secondary hydrogen atoms in 8407 compared to the abundant tertiary hydrogens in PP. The long side groups of the copolymer may also hinder approach of primary free radicals to its backbone.

PAGE 120

98 Figure 4-13: FTIR image of a typical PP:8407 alloy cont aining styrene and multifunctional acrylate. At 3085cm-1, aromatic ring C-H stretch; from 2970-2721cm-1 there are C-H stretching vi brations from methyl (CH3), methylene (CH2) and methyne (CH) groups; At 1738cm-1, C=O stretch from acrylate containing monomers; 1660cm-1 peak is a C=C stretch from styrene; 1460cm-1 is from CH2 scissoring, and both CH2 and CH3 bending; 1376cm-1 gives CH2 wagging, CH3 bending, and C-H deformation of CH3; from 1340-1050cm-1, C-H bending, CH2 twisting and wagging, C-C skelet al stretching vibrations, CCH3 stretching take place; peaks at 997cm-1 and 974cm-1 are from C-H out of plane bending; peaks in the region of 953 790cm-1 are from C-H bending, CH2 rocking, CH3 rocking, C-C stretching, C-CH3 Stretching; 766cm-1 peak from an aromatic C-H out of plane bend; the 720cm-1 peak from CH2 methylene rocking; and 700cm-1 peak from an aromatic C-H out of plane bend.

PAGE 121

99 Sample ID 95:5_90:10_80:20_70:30_ Grafting Efficiency 0 20 40 60 80 100 A B C Figure 4-14: Styrene grafting efficiency at various 8407 concentrations both with and without multifunctional monomer. The amount of styrene grafted onto the pol yolefins is noticeably increased with multifunctional monomer present. The chemistry behind this phenomenon implies that the reaction kinetics are much faster with styrene plus comonomer as opposed to styrene alone. This reduces the extent of PP chain scission and traps radicals before they can cause unwanted side reactions. Also, the fact that the multifunctional monomers have more than one site to start a growing ch ain or bond to anothe r one creates more opportunity to increase the total amount of gr aft copolymer. The grafting of styrene onto 8407 or PP rather than acrylate monomers is en ergetically favored, but it is still unclear whether grafting occurs before or after -chain scission and whether multifunctional monomer links degraded PP chains back together Besides its inherent stability to radical

PAGE 122

100 attack, styrene monomer improves the dissolution of multifunctional monomer, initiator, and the polymers, which is known to improve grafting efficiency [7, 40, 92, 94, 138, 206, 228]. Studies have shown that grafting does take place in the copolymer phase likely because of its lower Tm and lower viscosity than PP [240]. The tri-functional acrylate has a higher GE than the di-functiona l acrylate at all c oncentrations because of the extra free radical-trapping pendant group. Grafting and branching ma y originate from a single monomer unit or include intermolecular bonding of homopolymer or copolymer networks. It should be noted that the ratio of isotactic polypropylene segments can be quantified using FTIR by comparing the area under two peaks: 998 cm-1 (isotactic helical conformation) and 975 cm-1 (internal standard) [342, 343]. The grafted materials do not show any appreciable change in isotacticity. Gel permeation chromatography (GPC) is a universal technique for determining polymer molar mass averages and distribution of chains within the given sample. Figure 4-15 reveals that grafting does occur on 8407 when reactively extruded absent PP. TMPTA is shown to have a much grater infl uence on the higher molecular weight chains than DEGDA, which is due to the higher functio nality of TMPTA. This is evidence that the stress-strain behavior and grafting effici ency is more favorable with TMPTA present at high 8407 concentrations. Teh and Rudin have shown that when grafting polyethylene, molecular weight increases w ith increasing monomer concentration [250]. There is also a more pronounced increase in molecular weight of the longer polymer chains (i.e. Mz and Mz+1). Grafting is shown to occur on this copolymer, but Chapter 6,

PAGE 123

101 section 6.3.1.2 will give conclusive evidence th at the monomers do graft onto 8407 in the presence of PP. Molar Mass Averages MnMwMzMz+1 Molar Mass (g/mol) 3.5e+4 7.0e+4 1.1e+5 1.4e+5 1.8e+5 2.1e+5 2.5e+5 2.8e+5 3.2e+5 Pure 8407 0.3% Initiator, 6% Styrene, 0.8% DEGDA 0.15% Initiator, 3% Styrene, 0.8% DEGDA 0.15% Initiator, 3% Styrene, 0.8% TMPTA Figure 4-15: Molecular weight averages fo r pure 8407 and three grafted 8407 materials. A broadening of the copolymers molecu lar weight distribu tion is due to the random, uncontrolled nature of the free radi cal grafting process. The polydispersity index (or molecular weight distribution) in creases from 1.85 for pure 8407 to 1.97 for the grafted 8407 at 0.3% in itiator, 6% styrene, and 0.8% DEGDA. The PDI of 8407 at lower initiator and styrene conc entrations containing DE GDA and TMPTA are 1.91 and 1.96, respectively. 4.3.4 Morphology Scanning electron microscopy has been uti lized to study the elastomer domain size, shape and distribution in both the physical blends and alloys Figures 4-16 shows SEM pictures of alloys with varying elastomer concentration. In a similar fashion to the

PAGE 124

102 physical blends, the more elastomer in the alloy, the larger the etched domains. The number of domains present for the alloys is much less than the corresponding physical blends, which is an indication of covalent bonding of the elasto mer to the matrix. The diameter of particles in the alloys are significantly smaller and more elongated than the physical blend counter parts, with averages in Table 4-6 and particle size distributions in Appendix D. The reason for both of these effects is likely due to the stabilizing influence of the reactive materials [32, 87-89]. The number average particle diameter is smaller for all alloys when comp ared to the physical blends. This trend is even more obvious for weight average diamet er, where larger part icles are much more influential. Table 4-6: Image analysis of etched SEM su rfaces of several blends and alloys. This information is gathered from Appe ndix D, Figure 4-16, and Figure 3-5. Ligament thickness is calculated using the number average diameter. Sample ID Dn (nm) Dw (nm) Average Roundness for particles >1* Ligament Thickness (nm) Impact Strength (ft-lbs/in) 95:5_0 77 104 1.08 166 1.8 0.3 95:5_B 55 76 1.27 119 5.7 0.8 90:10_0 113 167 1.11 194 5.0 0.3 90:10_B 93 164 1.37 159 9.8 0.8 80:20_0 224 778 1.15 305 6.8 0.8 80:20_B 127 337 1.34 173 12.4 0.4 70:30_0 269 2630 1.16 321 2.9 0.1 70:30_B 162 860 1.35 193 8.4 0.5 *Roundness is a measure of how closel y the particle's shape matches that of a perfect circle. A value of 1 = a perfect circle. The images in Figure 4-16 are composed of non-circular particles, which is unique when compared to the physical blends in the previous chapter. As explained in section 2.4.4, Sundararaj, Macosko, and Scott deve loped their own theory of morphology development when melt blending polymers [32, 83] At an intermediate step of droplet breakup, a fragile lace structure forms which is composed of spheres of the matrix phase

PAGE 125

103 surrounded by a thin sheet of minor phase. Th e sheet then begins to break apart into irregularly shaped pieces of a wide distribut ion in size. With addition of initiator, styrene, and multifunctional monomer, these pi eces are very small (<100 nm in diameter) and are stabilized as non-spherical shapes. Without stabilization, as shown in Chapter 3, the irregular pieces continue to break down until all of the partic les become nearly spherical which is accompanied by coalescence. (a) (b) (c) (d) Figure 4-16: SEM images of (a) 95:5_ B at 10,000X, (b) 90:10_B at 10,000X, (c) 80:20_B at 10,000X, and (d) 70:30_B at 5,000X. Each bar marker = 1 m. Ligament thickness (surface-to-surface di stance between domains) was introduced in Chapter 2 as a key component to toughening brittle semicrystalline polymers. This is

PAGE 126

104 based on volume percent of rubber concentra tion, which is found by using the following equation: Volume fraction of rubber phase ( 8407) = w8407 / [(wPP*( 8407/ PP)) + w8407] (4.4) where w8407 = weight fraction of 8407 phase, wPP = weight fraction of PP phase, 8407 = density of 8407 (0.87 g/cm3), and PP = density of PP (assumed to be 0.91 g/cm3). The matrix ligament thickness (T) is found from Wu [120]: T = d[( /6 )1/3 1] (4.5) where d = rubber par ticle diameter and is the rubber volume fraction. The matrix ligament theory was founded on research that lim ited the dispersed particle size to about 300 nm. From SEM and TEM (explained in future paragraphs), the domain size of these blends and alloys are much smaller, so this theory may not be directly applicable. Many authors have s hown that ligament thickness should decrease with increasing rubber concentrations. They we re able to control the degree of dispersion and coalescence by using high viscosity rubbers (i.e., viscosity ratio is close to one) [119121, 125, 274]. From Table 4-6 and Figure 4-17, the opposite trend is witnessed. The viscosity of 8407 is much less than polypropyl ene, so breakup, diffusion, and coalescence of 8407 domains throughout the melt is easier than higher viscosity elastomers. As the percentage of 8407 increases, the size of the dispersed phase increases. From data in Chapter 3, this is due to diffu sion and coalescence, but grafting is able to prevent coalescence by stabiliz ing the dispersed phase. One aspect of the matrix ligament theory is the fact that below a critical interparticle distance, the material behaves in a tough manner (high impact strength) and ab ove which it behaves in a brittle manner. Figure 4-17 shows that the ligament thickness of the alloys is lower than the physical

PAGE 127

105 blends at all concentrations of elastomer. This translates into higher impact strength, as exemplified in Figure 4-18. 100 150 200 250 300 350 51015202530Volume % of 8407Ligament thickness (nm) physical blends alloys Figure 4-17: Matrix ligament thickness of PP blends and alloys as a function of volume % of 8407. 0 2 4 6 8 10 12 14 100150200250300350 Matrix Ligament Thickness (nm)Notched Izod Impact Strength (ft-lbs/in) physical blends alloys Figure 4-18: Room temperature notched Izod im pact strength as a function of matrix ligament thickness for blends and alloys at various volume % 8407. The alloys and blends have about the same matrix ligament thickness at 175 200 nm, but the impact strength of the alloy is much higher. This is a testament to the need for high interfacial bonding between the ma trix and dispersed phase. For physical

PAGE 128

106 blends, there appears to be a minimum and maximum ligament thickness. At high concentrations of elastomer, coalescence is common which leads to smaller surface area and weaker interfacial interactions. At low elastomer concentrations, the matrix ligament is smallest but a decrease in impact strength is observed, contrary to what the theory predicts. At such a small number average part icle diameter, the particles may not be able to effectively promote energy absorption mechanisms like dilation, cavitation, crazing, and shear yielding. Typically, the rubber phase should be approximately the size of the crack it is trying to stop [100]. Transmission electron microscopy (TEM) is ut ilized because it gives more detailed structural information than SEM. Contrast in bright field TEM depends on the relative electron transparency of the phases present in the material. Ruthenium tetroxide (RuO4) has been used as a stain to impart contrast enhancement in polymers because of its high electron density [344-349]. It locates in amorphous, low density regions of the polymer, thus scattering electrons more efficiently a nd creating a dark area in the TEM image. This is a valuable staining techniqu e for rubber toughened plastics or polymers containing aromatic unsaturation because the rubber/unsaturated materials are preferentially stained. As can be seen in Figure 4-19(a), pure PP shows a blurry picture, which may be because of the limited diffusion of the staining agent in a highly crystalline matrix whose Tg is close to room temperature. For a blend containing 10 wt% 8407, the staining agent preferentially locates in the elastomer phase, which is distributed in a range from about 10 nm to 150 nm diameter domains (Figure 4-19(b )). The particles are dispersed at such

PAGE 129

107 (a) (b) Figure 4-19: TEM bright field images of (a) virgin PP and (b) 90:10_0 stained with RuO4 at a magnification of 63,500X. Ba r marker on insert = 100nm. a small scale presumably because the viscosity of the PP is much larger than 8407. At low elastomer concentrations the higher viscosity matrix does not allow diffusion and coalescence of the el astomer domains so a fine mo rphology is observed. Also, the presence of these domains solidifies the f act that 8407 and PP are immiscible polymers, but at a relatively small scale. Addition of styrene and initiator to th e physical blend results in a unique multiphase morphology, as can be seen in Figure 4-20. There are vari ous levels of darkness in the image the darkest domains being attribut ed to styrene and the lighter, gray domains are 8407. Small dark domains (10 nm or le ss) are distributed throughout the image but some are locally clustered in the elastomer pha se. There are also a few 100 nm solid dark domains present in the image. These larger domains may be styrene material that has phase separated (possibly as homopolymerized material). For the elastomeric domains that contain several sub-10 nm styrene-base d domains, one might conclude that the

PAGE 130

108 Figure 4-20: TEM image of 90:10_A at 63,500X. All marker bars = 100nm. styrene has grafted to the elastomer phase. But it is quite possible from the phase inversion mechanism [78, 79] that some PP pha se may be trapped inside the elastomer phase along with grafted styr ene. Grafts may locate at interlamellar or lateral intercrystalline regions, forming microdomain s which differ structurally from larger domains of PS homopolymer [350]. By adding 0.8 wt% of multifunctional monomer to the styrene/initiator system, the morphology is altered (Figure 4-21). The ove rall area of stained material increases compared to Figure 4-20, a qua lification of the difference in grafting efficiency of the two systems. Similar to Figure 4-20, rela tively large elastomer domains containing a cluster of small dark domains about 10 nm in diameter are present (insert 1, Fig. 4-21).

PAGE 131

109 However, this domain is connected to a larger more highly stained domain with a cluster of particles approx. 50 nm in diameter. Because grafting (and dark staining of styrene) is Figure 4-21: TEM image of 90:10_B at 63,500X. All marker bars = 100nm. thought to occur primarily in the PP phase, these highly stained domains may contain polypropylene chains. With the high degree of bonding/grafting that is taking place between the phases, the phase inversion mechanism of tr apping PP inside 8407 domains is highly likely [78, 79, 89]. Insert 2 in Figure 4-21 shows many dispersed stained domains ranging from 10 -100 nm in size, but unlike Figure 4-20, these domains are within close proximity to eachother. Th e distribution of extremely small domains (presumably elastomeric) has a very noticea ble effect on macro-scale properties like impact strength. The morphology for insert 3 is very interesting because the stained material is a web-like structure encapsulated in an elastomeric domain. Keep in mind 1 2 3 1 2 3

PAGE 132

110 that these TEM images are of non-purified material, so the morphology may look much different when only grafted chains are present. A change in morphology is observed when there is an increase in elastomer concentration (Figure 4-22). As expected, more elastomer results in larger gray domains, with one domain elongated to approximately 1 m in length. There is also an increase in the size of dark stained domains for 80:20_A, which may be an indication of agglomeration or because of the cage effect leading to longer grafted chains. For 90:10_A, the stained material located in the elastomer phase (<10 nm in diameter) were dispersed relatively uniformly on the interior, but at the 80:20 level, these domains are Figure 4-22: TEM images of 80:20_A (left) and 80:20_B (right) at 63,500X. primarily found at the outer edges of the elasto mer and are larger in size. With higher elastomer concentration, the monomers are likel y partitioned in that phase, but because of

PAGE 133

111 the low grafting efficiency and slow kinetics are phase separating from the low viscosity melt as the blend proceeds dow n the barrel of the extruder. 80:20_B shows a more highly stained im age than 80:20_A, which is directly related to its superior styrene grafting efficien cy. The staining agent will partition in the amorphous regions of the polymer and since po lystyrene is non-crys talline, a darker image is to be expected. The number of li ght gray domains (purel y elastomer phase) in the high impact sample is much less th an in the samples without multifunctional monomer. This could mean that the elasto mer phase is distributed throughout the sample as styrene-grafted material. Like 90:10_B, clus ters of highly stained domains exist in the sub-100 nm range. 4.3.5 Possible Crosslinking of the System In order to verify whether the alloy cont ains an insoluble gel, which will hinder processability and compromise mechanical properties, both soxhlet extraction according to ASTM D2765 and thermogravimetric anal ysis (TGA) were performed. Dynamic mechanical analysis, in section 4.3.9, will also indicate any crosslinking in the alloy. The extraction technique was r un on 3 different samples: 95:5_B, 90:10_B, and 80:20_C but did not reveal any insoluble gel. Gel is know n to absent when grafting polyolefins, even for materials susceptible to crosslinki ng such as HDPE and LLDPE [7, 40, 94, 138, 196]. TGA is a popular method to quantify mate rial degradation as a function of temperature [351]. Figure 4-23 shows an unexpe cted response of the alloy it begins to degrade at a lower temperature than the physical blend. The oppos ite trend would be expected for a crosslinked system because the mobility of the chains is restricted and there are a greater number of bonds to break. The alloy degrades in this manner possibly due to residual initiator, mono mer, or reaction byproducts not vented out of the extruder

PAGE 134

112 vacuum port. Another reason for the lower degradation temperature may be due to an increase in the number of tie molecules, wh ich have a lot of st ored energy therefore making them more suscep tible to scission [352]. Temperature ( ) 200400600800 % Loss -100 -80 -60 -40 -20 0 90:10_0 90:10_B C Figure 4-23: TGA graphical comparison of a physical blend (90:10_0) and alloy (90:10_B). 4.3.6 Rheological Properties Melt strength (MS) and melt flow index (MFI) arguably are two of the most important material parameters when the thermoplastic processing technique involves stretching flow (i.e. blow mold ing, film extrusion, and thermoforming). The relevance of MFI has been described in the previous chapte r, but MS needs further explanation. The MS of a polymer is a measure of its resistan ce to extensional deformation. This is a comparative measurement like MFI and is useful in providing information on the drawability of differe nt polymers [353-355]. Isotactic polypropylene has a characteristically sharp melting point and poor melt strength [356], therefore limiting its use in these processes. Poor melt strength is

PAGE 135

113 indicative of parison collapse in blow molding, cell-rupture in foaming, married fibers in melt spinning, excessive sag in thermoformi ng, melt drool in profile extrusion, and low extrusion coating speeds [357, 358]. Several ways to improve melt strength include increase molecular weight, broaden mol ecular weight distribution, incorporate comonomer to reduce crystallinity, add nucleating agents, additives to form high viscosity interpenetrating netw orks, incorporation of short or long chain branches (most effective), grafting, or cro sslinking. High melt strength poly propylene resins have been created, commonly via solid phase grafting, wh ich drastically improve its rheological properties [179, 353]. Solid phase grafting has the lowest incidence of PP chain scission of free radical grafting methods. The melt strength (Table 4-7) and MFI (Figure 4-24) data indicate long chain branched molecules present in the PP a lloy, not necessarily a crosslinked gel. Solidification temperature of 95:5_B is higher than 95:5_0, which is to be expected because DSC results indicate an increase in crystallization temp erature. The upper service temperatures are not very different probably because the Tm of the PP phase does not change significantly. The melt strength of a ll alloys is much grea ter than the physical blends, with MFI decreasing in a parallel fashion [353, 354]. Long chain molecules are present in the system, thereby increasing the degree of entanglement and therefore having a higher resistance to extensional deformati on. A linear relationship exists between melt strength and zero-shear viscos ity which is dependent upon the average molecular weight of the polymer. The alloy has the ability to deform longer di stances and at faster rates than the physical blend, as revealed by its melt exte nsibility and melt toughness.

PAGE 136

114 From previous section, grafting efficiency is the lowest for alloys not containing multifunctional monomer. And because MFI is the highest for these same alloys (Figure 4-24), degradation ( -chain scission) of PP is most likely taking precedence. Without multifunctional monomer, branching is not detectable by MFI measurements. Molecular weight increases with increa sing monomer concentration along with a broadening of MWD and increase in the hi gh MW tail [250]. Some authors have found that at high monomer concentrations, MS a nd MFI decrease while viscosity and graft yield increase [40, 92, 215, 228]. The d ecrease in MS may be related to homopolymerized material. Sample ID 95:5_90:10_80:20_70:30_ Melt Flow Index (g/10 min) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 0 A B C Figure 4-24: Melt flow index of physical blends and alloys as a function of 8407 content.

PAGE 137

115Table 4-7: Melt Behavior of PP, PP -8407 physical Blends, and PP-8407 alloys. Sample ID Solidification Temperature, ST (oC) Upper Service Temperature, UST (oC) Melt Strength, MS (cN) Melt Extensibility, ME (mm/s) Melt Toughness, MS ME (cN*mm/s) 95:5_0 113 155 17.8 60.7 1080 95:5_B 129 156 56.1 86.5 4853 90:10_0 157 23.9 42.6 1018 90:10_B 155 43.1 93.5 4030 80:20_0 157.5 16.8 40.4 679 80:20_B 157.1 50.8 83.0 4216 Pure PP 107 163

PAGE 138

116 The free radicals formed on the backbone of PP are effectively trapped by the styrene-multifunctional monomer branched syst em, therefore eliminating degradation. materials with a low MFI value tend to have longer molecular chains and higher average molecular weights which lead to molecular entanglements in the polymer melt [359]. The MFI of polypropylene is mainly dependent upon the average MW of the polymer. At a given MFI, a polymer with a higher melt strength has a broa der MWD [360], with the long chain polymers (Mz) more important to MFI [73] For a long chain molecule, only short segments of the mo lecular chain can move at ti me, taking to Eryings theory into consideration [110]. As melt flow index increases, impact strength has been shown to decrease [265]. When comparing a di to tri-functional monomer, the higher functionality monomer results in a lower MFI at the 70:30 level. The higher functionality means higher density of crosslink junc tions and expectedly lower MFI. 4.3.7 Crystallinity and Crystallization Isotactic polypropylenes mechanical pr operties are highly dependent on its crystalline state, which is a function of processing c onditions, molecular weight, molecular weight distribution, and chain conf iguration and conformation. PP is a unique semicrystalline polymer because of its four possible crystalline phases [100, 267, 287, 327, 361, 333, 361-364]. Commercial grades of PP essentially crys tallize into the (monoclinic) modification with sporadic occurrence of the (trigonal or pseudohexagonal) phase at higher te mperatures [279, 287, 333, 365, 366]. The (triclinic) phase is only observed in low MW or stereoblock fractions crystallized at high pressures above 200 MPa. The so-called smec tic phase is created by quenching PP and is composed of a mixture of monoclinic and pseudohexagonal structures [334, 367].

PAGE 139

117 Wide angle x-ray diffraction is a precise technique used to determine the crystal structures present in polymers as well as the re lative size and orientat ion of the crystalline lamellae. Both long range and short range orde r must exist for the pe aks to be present in Figure 4-25, so an amorphous halo exists unde rneath them because of the presence of non-crystalline material. The peak positions in the figure represent specific planes which diffract the x-rays and the peak heights at these positions indicate the relative number of crystalline planes diffracting. The alpha pha se has a monoclinic structure with unit cell dimensions: a=6.65 b=20.96 c=6.5 and = 99 [364]. This crystal structure contains four chains/unit cell, unit cell volume of 905 3, and density of 0.936 g/cm3) [162, 338, 339, 364, 368]. Of all the possibl e crystalline phases of PP, the phase is the most thermodynamically stable [267] with a melting enthalpy of 207 J/g [162] and melting temperature of about 165C. For faster quenching, this phase is created due to its higher growth rate. The Bragg reflections at 14.1, 16.8, 18.5, 21.1, 22, 25.5, 28.5, and 42.5 correspond to the following indexed planes of the monoclinic crystal: (110), (040), (130), (111), (131) and (041), (060), (220), a nd (-113), respectively [162, 267, 287, 369]. A strong (110) reflection represents lamellae grow th in a radial direction (perpendicular to the chain axis) along with (130) at a 45 azi muth [339, 369-371]. The (010) planes, with the lowest density of methyl groups, are mo st likely to be involved in epitaxial interactions of tangential la mellae growth [369, 372-374]. Fo ld surfaces are parallel to (101) [375] and the (-113) plane is the best measure of the packing and orientation of chain axes [369]. The most intense (110) diffraction of polyethylene overlaps with the (111) reflections of PP [314].

PAGE 140

118 The alpha phase peaks are labeled in Figure 4-25 and the difference in peak heights could be due to variation in the mean lamellar crystal size or their distribution, deformation at the spherulite boundaries, or any long range order induced in the structure. At low angles (large d-spacing) 90:10_B ha s the largest number of diffracting planes which may signify a large number of smaller lamellar crystals. The (110) peak intensity is high for pure PP, which is to be expected because with long chains and larger, more Figure 4-25: XRD pattern of pure PP, 90:10_0, 90:10_A, and 90:10_B. uniform spherulites comes a greater degree of radial growth out from the spherulite center. This peak is low for 90:10_0 and 90:10_A because impurities or diluents may prevent or inhibit growth in the radi al direction [128, 255, 258, 264, 203, 307, 310]. The fact that 90:10_B has a peak intensity simila r to pure PP reveals that the total number of planes diffracting in the radial direction are high even though this sample has just as

PAGE 141

119 many if not more impurities as 90:10_0 and 90:10_A. The relatively high peak intensity of the (040) plane in 90:10_B may help clarify things. Structurally in PP, the tangential lamella e epitaxially grow on the lateral (010) crystallographic plane of the radial lamellae by a satisfactory interdigitation of the methyl groups of facing planes at a substantial angle (80 or 100) [373, 374, 376]. Only open textures may lead to well developed tangent ial lamellae, which may be prevented by the high degree of branching in 90:10_B. But ep itaxial crystallization of helical polymers like PP is dictated primarily by interchain and in ter-helical distances [373]. So if chains are in close proximity to eachot her, crystallization may be facilitated. The high intensity of the (110) and (040) peaks, therefore, represent a high density of both radial and tangential lamellae. The multifunctional monomer is promoting a large number of very small, highly cross-hatched lamellar structures. Differential scanning calorimetry will show that 90:10_A has a low % crystallinity which may be due to and alteration of the is otacticity of the chains With decreasing isotacticity, the capability to retain such a regular (010) plane is reduced due to the configurational defects caused by random appearance of methyl groups along the chain direction. As a consequence, the epitaxial nucleation process is increasingly hampered and the cross-hatching density is thus reduced. The packing and chain orientation peak in tensity (at 42.5) is smaller for alloys (90:10_B and 90:10_A) than 90:10_0 or pure PP, which is consistent with poor crystallographic packing in the chain di rection and defective crystal surfaces. Organization of such crystals resembles the smectic packing of conformationally distorted chains [364]. Gr afting is known to increase d-spacing and create more

PAGE 142

120 imperfections of the crystal lite surface [331, 350]. A crysta lline lattice can accommodate various kinds of imperfections in the fo rm of short branches [331, 362]. When comparing pure PP to 90:10_B, the d-spacing of the alloy is 0.01 nm lower for the (110) planes and 0.012 nm greater for the (040) plane. The width of the (040) and (130) peaks increase upon modification with initiator, styrene, and DEGDA. This indicates a wide r distribution of crystal size, a size reduction of these crystals, and/or a lo wer degree of order in the cr ystalline phase [267, 279]. At 14, there is a shift in peak position to highe r values for 90:10_A and B, indicating tighter packing in unit cell direction perpendicular to the chain direction [ 369]. Tighter packing of chains in the crystal unit cell of PP can be caused by higher isotacticity and/or higher crystallization temperature. This idea is el aborated upon later on in this section within the discussion about crystallization. The only sample that shows a distinct peak at 16 is 90:10_A, which represents the crystalline phase of PP. All other samp les show more of a broad shoulder than a distinct peak. There are two pr imary peaks assigned to this phase, namely (300) at 16.1 and (301) at 21.1 [162, 279, 287]. Like the (110) plane for the alpha phase, (300) correspond to the crystal growing planes and to the largest lateral dimensions (perpendicular to chain axis) of lamellar crystallites [339]. The crystalline unit cell is composed of 9 chains, with parallel-stacked lamellae tending to pack into bundles [162, 337, 340, 364, 377]. Figure 4-26 compares the relative size and orientation of beta crystals compared to alpha crystals. The dimensions of the beta crystallites are approximate ly three times larger than the corresponding dimension of the alpha crystallites, havi ng a unit cell volume of 3150 3, dimensions of a=10.98

PAGE 143

121 b=10.98 c=6.47 and =90, and density of 0.921g/cm3 [162, 267, 338, 339, 363, 368]. The metastable higher energy trigon al phase has a melting enthalpy of 166 J/g [162, 363, 378] and melting temperature of about 152C. It grows up to 70% faster than the phase in a temperature window extending from 105C up to 141C but is nucleated less profusely than [255, 287, 322, 325, 333, 338, 364, 370, 379-382]. The phase of PP can be created by speci fic nucleating agents (most common), temperature gradients, orientation by shear, orientation by phase separation, copolymerization, or from addition of elastomer [267, 325, 333, 339, 363, 383-385]. But authors have found that use of multifunctiona l monomers during the crosslinking of PP results in the formation of this unique pha se [114, 128, 350]. They attribute it to the presence of multiple molecular structures/architectures composed of branched and/or intermolecularly crosslinked macromolecules. The phase is not created simply by grafting polystyrene onto PP [386]. Figure 4-26: Alignment of lame llae within spherulites of -PP (left) and -PP (right) from reference 278. The smectic phase may be present in this material, which is stable at temperatures below 70C, shows order in the chain axis direction, but the late ral packing of the PP helices are not as well formed as in the alpha monoclinic crysta lline phase [334, 367]. The quenched material is composed of a mixture of monoclinic and pseudohexagonal

PAGE 144

122 structures and has reflections at 14.8 representing inte r-chain packing and 21.2 for inter-chain registration. X-ray diffraction introduces many intere sting possibilities of how the lamellar morphology is structured in these material s. A permanganic acid etching technique pioneered by Olley et al. [387] has shown to be able to reveal lamellar crystals in both isotactic polypropylene and polyethylene [314318]. Figure 4-27 is an etched SEM image of pure polypropylene crystallized noni sothermally from the melt. In Figure 427(a), arrow #1 points to the center of a spherulite which is about 4 m in diameter, while arrow #2 locates many radial lamellae at the fr inge of the spherulite. Figure 4-27(b) shows the periphery of a spherulite where many radial lamellae are visible. These lamellae are less than 100 nm th ick and can extend out over 2 m from the center. The acid etch procedure for these samples is not optimized and the experiment was nonisothermal, so large, well-defined spherul ites were not visible. But throughout this image, holes may be mistaken for amorphous non-cross-hatched material. The cross hatches of pure polypropylene are relatively thick lamellae and will not be etched by the acid treatment. The next image (Figure 4-28) is an etched film of 90:10_0, the physical blend of PP and 8407. Large dark pits about 0.5-2 m in diameter are dispersed in this sample, which are most likely remnants of elastomeric particle s. In this image, both radial (arrow #1) and tangential (arrow #2) lamellae can be seen within a non-circular spherulite in the center of the image.

PAGE 145

123 (a) (b) Figure 4-27: Etched lamellar morphology of isotactic polypropylene of a well defined spherulite at 15kX magnifi cation (a), bar marker = 2 m and periphery of a spherulite at 20kX magnifi cation (b), bar marker = 2 m. 1 2

PAGE 146

124 Figure 4-28: Etched lamellar morphology of 90:10_0 at 10.5kX magnification, with the bar marker = 5 m. X-ray diffraction has introduced the possi bility of 90:10_B having a dense crosshatched structure, and SEM reveals its morphology in Figure 4-29. A highly dense network of crystalline lamellae is in fact present throughout th e image. It should be noted that this type of structure was comm only seen throughout the sample (about 5 cm2). The network (or cross-hatched) stru cture is a combination of radi al and tangential lamellae. A defined spherulitic structure wa s not seen, but arrow #1 is beli eved to be radial lamellae and arrow #2 tangential lamellae. The most inte resting aspect of this material is pointed out by arrow # 3, which is a unique combina tion of interwoven ta ngential and radial lamellae. Some lamellae measure to be below 100 nm in length and about 10-20 nm in width. This aspect ratio is much smaller than what is observed in the previous two images. The presence of etched pits, like 90:10_0, represent elas tomeric domains of 1 2

PAGE 147

125 8407. Notice that these are much smaller in di ameter than in the phys ical blend ( a trend observed in section 4.3.4). Figure 4-29: Etched lamellar morphology of 90:10_0 at 20kX magnification, with the bar marker = 2 m. Differential scanning calorimetry is a versatile tool which allows precise measurement of the melting and subsequent recr ystallization of semicrystalline polymers. The heating run described in the experimental section should eliminate all traces of melt orientation [388]. Nonisothermal experiment s were performed instead of isothermal measurements in order to simulate the nonisothermal conditions of processing. From Figure 4-30 and Table 4-10, Tm of the PP phase at about 165C does not decrease with addition of elastomer, but it does decrease upon addition of liquid reactants, similar to the findings of Borsig et al. [343]. When graf ting polyolefins, their melting temperature and % crystallinity are known to decrease compared to the pu re polymer [7, 148, 184, 215]. The decrease in Tm along with decrease in crystallinity is seen for 90:10_A. The point in 123

PAGE 148

126 Temperature ( ) 120130140150160170180 Endotherm ( W) -11000 -10000 -9000 -8000 -7000 -6000 -5000 90:10_0 90:10_A 90:10_B C Figure 4-30: DSC heating traces of a physic al blend (90:10_0) and alloys (90:10_A and 90:10_B). which melting comes to completion occurs th an 90:10_0 (173C vs. 175C, respectively). This means that larger, more perfect crystals composed of longer crystallizable PP chains exist in the physical blend. The disappearan ce of the last traces of crystallinity in 90:10_B means that the crystals are either re latively more imperfect or there is an increase in entropy of fusion [389]. The % crystallinity is estimated by taki ng the area of the melting endotherm and dividing by the melting enthalpy of a perfect cr ystal [390]. A value of 207 J/g was taken to be the standard melting enthalpy, with the alloy containing multifunctional monomer having a higher % crystallinity than both 90:10_0 and 90:10_A. The physical blend has a low crystallinity because the amorphous elasto mer acts as a diluen t. 90:10_A, on the other hand, likely contains degraded PP chai ns which could once crystallize but are now chopped into smaller non-crystallizable segmen ts. The high crystall inity of 90:10_B will

PAGE 149

127 be explained later in this sect ion. Remember that initiator de rived free radicals attack the high molecular weight chains first, and these ar e the same chains that contribute most to the high level of crystallinity in pure PP. Graf ted chains will not crystallize because they are strictly amorphous materials. Table 4-8: DSC endothermic data compar ing pure PP to 90:10_0, 90:10_A, and 90:10_B. An average of three runs were performed on 90:10_B. Sample ID Phase Melting Peak (C) Tm Onset (C) Enthalpy of Melting (J/g) % Crystallinity Pure PP 165.9 0.4 154.5 0.5 93.9 0.7 45.4 0.4 90:10_0 165.2 0.5 152.9 0.9 82.3 1 39.8 0.4 90:10_A 164.3 0.2 149.5 1.1 81.7 0.8 39.5 0.5 90:10_B 164.8 0.4 143.9 1.1 86 1.8 41.5 0.9 The onset of melting temperature (Tm onset) is derived from the intercept of the tangent to the steepest gradient of the leadi ng edge of the fusion curve with the baseline of the thermal curve. By taking a close l ook at 90:10_B, melting begins at a much lower temperature than the physical blend. The me lting point of lamellar crystals reflects the energy stored in the fold surfaces and he nce their stability [362, 369]. Regions of intermediate order should melt at lower te mperatures than highly ordered internal regions. The melting transition broadens for PP due to degradation [148] or creation of larger molecules [215]. The onset of melti ng in a LLDPE/PP/peroxide system decreased by 9C with 0.25% peroxide and the endotherm was consistently bimodal, attributed to the existence of two different crystal forms [222]. The melting point should also be depressed also if the crystallite is very sm all (i.e. smaller crystallizable polymer chain segments), owing to higher surface free energy and lower heat capacity [255, 279, 327, 362, 369, 389, 391-397]. Diluents, copolymerized units, grafts/branches/crosslinks, and end groups should have an equivalent eff ect on depressing the melting point at low

PAGE 150

128 concentrations [161, 185, 260, 331, 389, 393-395, 398]. Small amounts of crosslinks or branches could limit the maximum lamellar thickness because of a reduction in long range mobility, the presence of pre-existing ti e molecules, and/or th e production of stress along the crystallizing chain so as to oppose the incorporation of additional members of the same chain [389, 393-395]. One may argue that there may be a reduction in melting entropy of the crystals in the pres ence of crosslinks, resulting in Tm elevation [393], but this trend is not seen. Another set of experiments focused on the melting of these materials is shown in Figure 4-31. The PP:8407 ratio is 80:20 and th ese samples show many similarities to the 90:10 blend and alloys. 80:20_A shows th e characteristic melting peak of the phase of PP at 149.4C, but without styrene present in the system, this peak is not seen. The sample labeled 80:20_C without styrene cont ains 0.8wt% TMPTA and 0.3% initiator, but not styrene monomer. This material has th e lowest % crystallinity and lowest peak melting temperature, which is indicative of a highly de graded PP phase. This is reinforced by the fact that the melt flow inde x of this material is extremely high (Chapter 5, Figure 5-7). The physical blend (80:20_0) shows an in termediate % crystallinity with a well defined melting endotherm, high Tm onset, and high peak melting temperature. This is to be expected because the high molecular we ight chains of PP are not altered by the grafting process. The high impact strengt h alloy (80:20_B) shows the typical broad melting endotherm (a high number of small, defective lamellar crystals), high % crystallinity, but unlik e 90:10_B, a high peak Tm. Briefly, this difference is the result of

PAGE 151

129 enhanced crystal motion at the 90:10 level a nd reduced motion at the 80:20 level. The viscoelasticity of the alloys (section 4 .3.8) will help explain this phenomenon. Temperature ( ) 130135140145150155160165170175 Endotherm ( W) -10000 -9000 -8000 -7000 -6000 -5000 80:20_0 80:20_A 80:20_C without styrene 80:20_B C Figure 4-31: DSC heating traces of a phys ical blend (80:20_0) and alloys (80:20_A, 80:20_B, and 80:20_C without styrene). Crystallization of long macr omolecular chains involves th e transfer of consecutive sequential chain units from the amorphous to the crystalline phase [100, 161, 313, 362, 392, 393, 399]. It is complicated by the requirement that many consecutive units of each participating chain must enter systematically in the same crystallite via chain sliding and diffusion. Crystallization will be restricted to very long crystallites slightly below Tm but the critical chain length for crystallization decreases with temperature. The minimum sequence length for crystallization of PP ha s been reported to range from 8 to 11 monomer units [400, 401].

PAGE 152

130 From Figure 4-32 and Table 4-9, the crystall ization peak temperature is highest for 90:10_B, followed by 90:10_A, and finally 90:10 _0. Nucleation and growth rate are known to increase with addition of peroxide [220]. Also, Tc increases for grafted polymers because of the nucleating aff ect [7, 148, 179, 221, 245]. Branching in polyolefins has a nucleating effect by in creasing crystallization temperature and crystallization rate [238, 357, 402]. Grafting monomers onto PP increases its crystallization temperature, so they in effect act as nucleating agents. Nucleating agents increase the crystallizati on rate and temperature, gi ve a more uniform morphology, improve mechanical properties and optical properties (from reduc tion of spherulite dimensions), and shorten process cycle time [338, 361]. They also retard the relaxation of chains from the polymer melt, producing more tie molecules which improve molecular entanglement between crystal grains and in crease boundary strength between spherulites [260]. 859095100105110115120125130135140145150155 Exotherm ( W) 2000 4000 6000 8000 10000 12000 14000 90:10_0 90:10_A 90:10_B Temperature ( )C Figure 4-32: DSC cooling traces of a physic al blend (90:10_0) and alloys (90:10_A and 90:10_B).

PAGE 153

131 A higher crystallization temperature implie s a faster crystallization rate, which infers an increase in the st ereoregularity of the chain [378, 401]. Although PP may have the same overall isotacticity (from FTIR), the non-isotactic units may be distributed differently and hence lead to a variation in cr ystallizability. This may be the reason that the percent crystallinity in 90:10_B and 80:20_B are higher than th e physical blends at these levels of elastomer. The most importa nt thermal parameter for crystallization is the supercooling ( T defined as Tm-Tc), which shows how far the polymer crystallizes from its melt-crystal equilibrium st ate [378]. With increasing Tc (smaller supercooling), the fraction of the stable phase increases [375, 398]. The thickness of the lamellar crystallites in the form is inversely proportional to the supercooling below 191C. The onset of nucleation follows the same trend as Tc for the set of samples, which means that there may be a lower number of non-isotactic de fects associated with the small crystals in 90:10_B [401]. The introduction of a small number of cro sslinks is known to considerably increase the nucleation density of the polymer [402]. There may also be a unique interface of the graft copolymer [128]. Beta nucleated samples crystallize with a much lower supercooling than the alpha phase, which might also explain its higher crystallinity [378]. Table 4-9: DSC exothermic data compar ing pure PP to 90:10_0, 90:10_A, and 90:10_B. Sample ID Peak Temperature of Crystallization (C) Enthalpy of Crystallization (J/g) Pure PP 109.6 0.9 -90.7 0.6 90:10_0 115.3 0.4 -83.1 2.1 90:10_A 124 1.4 -80.7 0.7 90:10_B 130.8 0.65 -84.4 1.4

PAGE 154

132 Perturbations in the mobility of chains duri ng crystallization hinder the formation of the most compact alpha phase [327]. It should be noted that the crystallization and melting behavior of blends and alloys at the 95:5 and 70:30 levels fo llow the same trends as those at the 90:10 and 80:20 levels. Polarized light microscopy was utiliz ed to elaborate upon the unique and unexpected crystalline features of these materials. This tool, when used with a heating stage and first order wave plate, gives detailed in-situ morphological information on the crystalline phases of the materials. Figures 4-33 thru 4-36 are collections of still photographs describing the non-isothermal crystallization of pure polypropylene, 90:10_0, and two reactive blends (90:10_A and 90:10_B). Because this is the end of a second heating/cooling cycle, many of the phase spherulites originally present in pure PP has re-crystallized into the thermodynamically more favorable phase. The relatively slow cooling rate (10C/min) allows this transformation to take place [287, 322-327]. Contrast in polarized light microscopy is dependent upon a change in refractive index and is positive when the refractive index of light polarized parallel to the long axis of the object is larger than th at of light polarized perpendicu lar to the long axis [370]. The birefringence of a spherulite is positive wh en the refractive index for light polarized parallel to the radius of the s pherulite is larger than for ta ngentially polarized light (alpha phase of PP) and negative when oriented tangentially (beta phase of PP). The phase spherulite is characterized by a central region composed of an interwoven array of short lamellae, stacked edge-on, and the spherulite periphery of long, radi ally-oriente d lamellae with short lateral branches or iented approximately perpendicu lar to the radial lamellae, therefore building a rigid 3-D crosshatched network with weakly positive birefringence

PAGE 155

133 [337, 370, 371, 374, 375, 388, 403]. The phase spherulties s how a radial lamellar morphology (banded structure) which is not as rigid as the monoc linic cross-hatched morphology and shows a highly negative birefringence [287, 314, 325, 333, 359, 366, 370, 403]. Figure 4-33: Polarized optical images of virgin PP cooled from the melt. (a) = 115.5C at 50X, (b) = 113.5C at 50X, (c) = 110C at 50X, and (d) = 25C at 20X magnification. The alpha phase typically shows the charac teristic maltese cross extinction pattern but may appear blurry as fibrils radiate from the center of the spherulite [396]. This is a mixed spherulite and consists of crosshatched lamellae in which subsidiary lamellae grow tangentially to the primary radiating lamellae. In Figure 4-33, nucle ation of spherulites commences at random times, therefore creating a distribution of spheru lites that range in

PAGE 156

134 size from about 10 m to about 100 m. Notice that the spheru lites, which are composed of twisting lamellae radiating out from a nucleating center, have a mottled maltese cross pattern of positively birefringent material. If the color using a first order wave plate is uniformly blue or yellow, lamellae are arra nged in parallel in one direction and run through domain from end to end [398]. A blue color indicates positive birefringence in the direction of the chain axis [370, 396]. Moving on to the physical blend (Figure 434), crystallization appears to mimic that of pure PP, except for the dark circular domains dispersed throughout the spherulite. These domains are probably the phase separated elastomer particles, incorporated into the Figure 4-34: Polarized optical images of 90: 10_0 cooled from the melt. (a) = 118C at 50X, (b) = 114.5C at 50X, (c) = 105C at 50X, and (d) = 25C at 20X magnification. A bubble (artifact) appears in the lower left hand of (b) and (c) and along the edges of (d).

PAGE 157

135 growing spherulite as occlusions (n o effect on growth rate) [100, 236, 255, 257, 260, 263, 297, 307, 311-313]. The spherulites remain roughly spherical until they impinge on eachother and range in size from about 10 m to about 100 m. The crystallization behavior of grafted PP has already been proven to be much different than the pure polymer. And from Figure 4-35, there appears to be no semblance of the typical spheruli te seen in the previous two exam ples. A distribution of oval and round spherulites nucleate at the beginning stag es of crystallization, with some showing completely positive birefringence and some completely negative. This means that the orientation of the crystallite is either comp letely in-phase with the polarized light or Figure 4-35: Polarized optical images of 90: 10_A cooled from the melt. (a) = 128C at 50X, (b) = 124C at 50X, (c) = 122 C at 50X, and (d) = 25C at 20X magnification.

PAGE 158

136 completely out-of-phase. Also, elliptical cr ystals with uniform optical anisotropy are characteristic of liquid crystalline polymers having rigid molecular chains with local ordering. These morphologies rarely occur in the crystallization of flexible polymers like PP [396]. A network-like color scheme of blue or yellow regions is found for 4-35(c), which could mean a disruption of crystalline or der in the sample. The last image in the Figure (4-35(d)) shows a much fi ner distribution of birefringent material as compared to Figures 4-33(d) and 4-34(d). When monomers are grafted onto polymers, the polymer is known to crystallize at higher temperatures and result in sm aller spherulite dimensions [230, 238, 357, 402]. The DSC and XRD results for 90:10_A show the presence of phase polypropylene, which is characterist ically more highly birefringent than phase. This difference cannot be discerned from the images in Figure 4-35. The sample with the highest Tc and presumably highest crystallization rate, 90:10_B, reveals the most inte resting morphology of all the samples (Figure 4-36). The nuclei are very small and distributed unifo rmly throughout the sample and appear at approximately the same time. The ellipsoidal and spherical shapes common in the previous three figures are not present in Figure 4-36. A network-like appearance of birefringent material is present, likely associated with the branching afforded by multifunctional monomer. The final imag e in the group shows a macroscopically uniform distribution of extremely small, pr esumably crystalline structures below 1 m in diameter. The extensive branching present in this all oy appears to have a significant effect on the entire crystallization process and is e xplained as follows: Cl assical nucleation and growth is known be preceded by an induc tion period, where parallel orientation and

PAGE 159

137 segregation of polymer segments induces a spinodal decomposition type microphase separation [110, 327, 389, 397, 398, 400, 404, 405]. Density fluctuations of about Figure 4-36: Polarized optical images of 90: 10_B cooled from the melt. (a) = 135C at 50X, (b) = 133C at 50X, (c) = 130 C at 50X, and (d) = 25C at 20X magnification. 20 nm periods in the amorphous phase are a ssociated with the i nduction period, but long range ordering and local orientation with domain sizes about 300-1000 nm takes place before the formation of crystal nuclei [404-406] The early stages of crystallization of polymers can be ascribed to a physical gela tion-type process [392, 406, 407]. After the crystal network is set up, the amorphous defects can conti nue locally to crystallize with very little or no long distance diffusion in a secondary crystallizati on process [61-63]. A small number of crosslinks is known to form a sample-spanning network and produce a

PAGE 160

138 considerable increase in nucleation dens ity [390, 393-395, 407]. Als o, crystallinity is easier with a stretched confor mation over some distance [408]. 4.3.8 Viscoelasticity Dynamic mechanical testing can reveal a lot about a polymers response to temperature, time, and load. The transitions allow one to determine the relative compatibility of the existing components as well as degree of crossl inking and crystalline phase characteristics. Storage modulus is an indication of th e stiffness of the polymer and may be considered inversely proportional to impact strength in physical blends [283]. The stiffness typically increases with increasing density (crystallinity) or entanglement density (from high molar mass) [284 ]. Figure 4-37 plots E' and Tan vs. temperature of 95:5_0 and 95:5_B. The alloys response to th e oscillating load deviates from the physical blend in several instances. The stor age modulus of 95:5_B is equal to or lower than the physical blend at all temperatures. This may be a result of a greater free volume from grafted and ungrafted styrene/DEGDA molecules, the presence of smaller PP chains, and a nanometer length scale distribution of elastome r which changes the intrinsic relaxation time distribution of the polymers. Al so, the entire range of molecular chains in the amorphous phase of the physical blend may fr ozen in to form a fully glassy state at low temperatures [302, 409]. Between about 0C and 40C (onset of long range molecular motion of PP) there is a smaller slope for the E for 95:5_B than 95:5_0 which is an indication of crosslinking or decrease in crystallinity [410]. A gel is not present and % crys tallinity actually increases, so this is likely due to the branched arch itecture or an increase in molar mass from the

PAGE 161

139 reaction [81]. For Tan graph comparisons, the alloys intensity is lower than the physical blend up until 40C. This may be due to the simple fact that the overall Temperature ( ) -80-60-40-20020406080100120 E' (MPa) 5.0e+7 5.5e+8 1.1e+9 1.6e+9 2.1e+9 2.6e+9 3.1e+9 3.6e+9 4.1e+9 4.6e+9 Tan Delta 0.05 0.10 0.15 0.20 0.25 0.30 95:5_0 95:5_B C Figure 4-37: Storage modulus (E') and Tan comparison of 95:5_0 and 95:5_B as a function of temperature concentrations of PP and 8407 are reduced because of the presence of the reactive materials. Another explanation is becau se molecular motion is hindered by a highly entangled network of chains or a greater nucleation density of crystals [301]. The higher intensity of 95:5_0 at Tg (~20C) indicates greater amorphous content of the material. DSC measurements have shown that the % crystallinity of the alloy is actually higher than the physical ble nd so the intensity of Tg would be lower. Also, a decrease in intensity reveals an increase in molar mass [81]. Above 40C, the molecular motion in the amorphous phase is more pronounced for the alloy. The smaller lamellar crystals in th e alloy may allow for easier crystal glide and amorphous phase rearrangement. Authors have attributed this increase in shoulder/peak

PAGE 162

140 height at about 80C to the existence of the crystalline phase of PP, which is known to have a wider, diffuse amorphous and crystalline region than the phase [260, 287, 301, 337, 340]. The decrease in intensity afte r this shoulder may be due to lamellar reorganization (annealing) into a more thermodynamically stable phase [132]. There was no evidence of polystyrenes Tg in this or any Tan graph possibly because the concentration was too low a nd became masked by other relaxations. At the highest temperature of the test, the intensity of the alloy is higher than th e physical blend, also lending to the notion that crys tallographic defect motion is probable, surface defects are more mobile, and long range amorphous motion and crystal glide are present. At the 90:10 level of PP:8407 (Figures 438 and 4-39), three different samples are compared: 90:10_0, 90:10_A, and 90:10_B. For the storage modulus graphs, the high impact sample is again lower than the physic al blend at all temperatures and shows a Temperature ( ) -80-60-40-20020406080100120 E' (MPa) 5.0e+7 5.5e+8 1.1e+9 1.6e+9 2.1e+9 2.6e+9 3.1e+9 3.6e+9 4.1e+9 4.6e+9 5.1e+9 5.6e+9 Tan Delta 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 90:10_0 90:10_A 90:10_B C Figure 4-38: Storage modulus (E') and Tan comparison of 90:10_0, 90:10_A, and 90:10_B as a function of temperature.

PAGE 163

141 similar decrease in slope at about 20C. But one subtle difference between 90:10_B and 90:10_0 is at the high temperature region of th e graph, where E' shows a greater negative deviation from the physical blend. This may be due to softening of the material and may be a consequence of a change in the relaxa tion behavior brought about by nano-elastomer domains or the enhanced mobility of smaller lamellar crystals. Tg of both the PP and 8407 phases are lower than the physical blend, a trend explained in the 95:5 comparison. The alloy not containing the multifunctional monomer (90:10_A) has a much lower storage modulus at low temperatures than either 90:10_0 or 90:10_B. This may be the result of an excess homopolymerized (not graf ted) material which may exist as oligomers and have little chain mobility restrictions. But at high temperatur es, 90:10_A mimics the physical blend which may signify that while the amorphous phase is severely affected by this modification, the lamellar crystals and its corresponding defects are not. -40-2002040 0.04 0.06 0.08 0.10 90:10_0 90:10_A 90:10_B Temperature ( )CTan Delta Figure 4-39: Tan comparison of 90:10_0, 90:10_A, and 90:10_B between -50C and 60C to show a magnified graph of 8407 and PP.

PAGE 164

142 When studying the Tan peaks of the materials in Figures 4-38 and 4-39, the elastomer phase appears to be affected to a greater extent in 90:10_B than 90:10_A. This means that in 90:10_B, the reactive materials are able to localize and react in a more efficient fashion than 90:10_A, leading to a branched architect ure and reduction in molecular mobility. At the Tg of polypropylene, 90:10_A has a higher molecular mobility than the other two samples. This is to be expected from melt flow index results, where it was shown that by reactively extr uding without multifunctional monomer, the high molecular weight tail of PP is attacked a nd degradation ensues [233]. An interesting comparison at about 80C can be seen ther e is a more prominent peak/shoulder for 90:10_A but above this temperature the inte nsity is less than both 90:10_B and 90:10_0. X-ray diffraction and DSC studies have already shown that the crystalline phase of PP is more pronounced in 90:10_A. DSC also c onfirmed that the overa ll % crystallinity is lower for this sample. So, one may conclude that this peak/shoul der does represent the unique crystal phase of PP but the material crystallizes predominantly in the alpha phase. The alpha phase may not be distributed as small lamella r crystals, as in 90:10_B. 90:10_B does have a shoulder at about 80C, similar to the shoulder found in XRD for the or smectic phase. At high temperatures, the Tan graph drastically increases for 90:10_B compared to 90:10_0. This behavior is indicative of interand intra-crystalline mobility, and is a direct result of the large di stribution of small lamellar crystals with high surface energy [132]. DSC results have show n that the peak melting temperature of 90:10_B is lower than 90:10_0, reinforcing the idea of greater crysta l mobility in the alloy.

PAGE 165

143 For higher concentrations of elastomer (Figures 4-40 and 4-41), the storage modulus results are similar. In both cases, th e high impact alloys are not as stiff as their physical blend counterparts at lo w temperatures (until the Tg of PP is surpassed). The distribution of nano-elastomer domains throughout the alloy ma y be affecting its overall relaxation behavior. In othe r words, there are likely ma ny soft domains with highly mobile polymer chains composing them, but th ey may be anchored at a few points so as to stabilize them. The slope of the storag e modulus for each alloy is smaller than the physical blends, and MFI and melt strength data reinforce the idea that a high molar mass branched structure is the cause. When pro cessing at these high elastomer levels, the reactive materials may have a tendency to crosslink and branch while trapped in the elastomer phase. Above the rubbery flow regi on (40-70C), E' is higher for the alloys Temperature ( ) -80-60-40-20020406080100120 E' (MPa) 5.0e+7 5.5e+8 1.1e+9 1.6e+9 2.1e+9 2.6e+9 3.1e+9 3.6e+9 4.1e+9 4.6e+9 Tan Delta 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 80:20_0 80:20_B -60-3003060 0.02 0.04 0.06 0.08 0.10 C Figure 4-40: Storage modulus (E') and Tan comparison of 80:20_0 and 80:20_B as a function of temperature

PAGE 166

144 because of chemical and physical links in th e amorphous phase and at the surface of the crystalline phase. Internal friction (Tan ) measurements of these samples show that the Tg of 8407 decreases in intensity but also shifts to a slightly higher temperat ure. A higher energy barrier for molecular motion is present (possi bly due to branching or crosslinking) and the phases are more compatible due to grafting agents [8, 9, 258, 289, 321]. The Tg of PP decreases in intensity upon modi fication, a trend seen in all samples and attributed to a decrease in chain mobility or concentration of PP. At hi gh temperatures, Figures 4-40 and 4-41 show different behavior. 80:20_0 e xperiences a drastic increase in intensity above 60C which indicates that the spheru lite morphology may be disrupted and the amorphous regions surrounding the lamellar crys tals greatly influence interand intracrystalline mobility. For 80:20_B a single peak exists at 80C, which may represent the Temperature ( ) -80-60-40-20020406080100120 E' (MPa) 5.0e+7 5.5e+8 1.1e+9 1.6e+9 2.1e+9 2.6e+9 3.1e+9 3.6e+9 4.1e+9 Tan Delta 0.1 0.2 0.3 0.4 0.5 70:30_0 70:30_B C Figure 4-41: Storage modulus (E') and Tan comparison of 70:30_0 and 70:30_B as a function of temperature.

PAGE 167

145 phase of PP. Like the 70:30 alloy, the slop e of the curve increases continuously due to the slippage and diffusion of amorphous regions around stable lamellar crystals. A close look at 70:30_0 reveals two p eaks at about 60C and 100C. The peak at 60C may represent the melti ng of any 8407 crystalline lamellae. At 100C, there may be pre-melting of the PP crys tals and very high rate of slippage and mobility of the crystals. Very diffuse and unstable spherulites are known to exist in PPelastomer blends at high elastomer con centrations [257, 261, 263, 264]. For 70:30_B, these two peaks disappear and a single p eak forms at 80C, not unlike 80:20_B. The creation of this one peak shows that the r eactive constituents lead to compatibilization between phases. The decrease in intensity of the alloys comp ared to the physical blends at high temperature reveals a restriction in PP chain mobility and a high entanglement density. DSC comparisons of 80:20_0 and 80:20_B have shown that Tm of 80:20_B is higher than 80:20_0. 4.4 Conclusions Many types of reactions are present in th is complex reactive extrusion process. Understanding free radical polymerization mechanisms is extremely important and key to controlling the microstructure of the alloys and the resulting macr o-scale properties. Initiator decomposition is the rate determini ng step, but once the given energy barrier is overcome, several reactions happen at once. These include monomer polymerization, hydrogen abstraction from polymers, graf ting of monomers onto polymers, and degradation reactions. The end result, which is reinforced by all of the data presented in this chapter, is an interconnected, branch ed collection of macromolecules with unique

PAGE 168

146 morphology, excellent mechanical properties a nd melt extensibility, and characteristic crystalline morphology. Impact strength has been shown to be dependent upon a high degree of grafting between the phases present, di stribution of extremely small elastomer domains, a highly entangled amorphous phase, small crystalline st ructures, and a low glass transition impact modifier. Literature has pr imarily concentrated on the modification of either the amorphous phase or crystalline ph ase, but this work was able to tie in both concepts for the creation of a novel super high im pact semicrystalline polymer. The crystalline state of the alloys appear s to be unique, with a finely dispersed, highly cross-hatched spherulitic structure. DS C has shown that relatively smaller crystals are indeed present which have the ability to slide and diffuse more efficiently than pure PP lamellar crystals. A combination of the af orementioned characterist ics give the alloys the unique ability to both absorb energy by typical mechanisms but also to resist deformation by phase transformation t oughening and elastic retractive forces.

PAGE 169

147 CHAPTER 5 THE EFFECT OF PROCESSING CONDITIONS ON ALLOY PROPERTIES 5.1 Introduction This chapter delves deeper into how mech anical properties and grafting efficiency change with varying processing parameters a nd reactive ingredient concentrations. The effect of screw speed, temperature, initi ator concentration, multifunctional monomer concentration, and styrene concentration are studied. Properties such as notched Izod impact strength, elastic modulus, yield strengt h, melt flow index, a nd grafting efficiency can be tailored and optimized by varying these parameters. 5.2 Experimental 5.2.1 Materials Table 5-1 gives a list of pertinent ethylene-1-octene copolymers produced by Dupont Dow elastomers under the trade name ENGAGE, but the EOC grade of interest is 8407 [48]. Isotactic pol ypropylene homopolymer was supplied by Equistar Chemical (grade PP 31S07A) and is contact translucent. All polymers were received in pellet form. The peroxide and monomers used in this st udy were reagent grade chemicals (structures are shown in Table 5-2). The monomers we re purified by passing through an activated alumina column before use. Styrene mono mer, inhibited by 10-15 ppm t-butyl catechol, was purchased from Fisher. The initiator 2,5dimethyl-2,5-di-(t-bu tylperoxy) hexane, was purchased from Atofina under the trad e name Lupersol 101. Diethyleneglycol diacrylate (DEGDA), inhibited by 80 ppm Hq and 120 ppm MEHQ, and trimethylolpropane triacrylate (TMPTA), inhibited by 125 ppm HQ and 175 ppm MEHQ,

PAGE 170

148 were graciously donated by Sartomer, an Atofin a company. Table 5-1 lists structures of the reactive materials. Table 5-1: ENGAGE product data table. ENGAGE Grade (decreasing comonomer content) 8842 8407 8200 8401 8402 Comonomer Content wt% 13C NMR/FTIR 45 40 38 31 22 Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902 Melt Index, dg/min ASTM D1238, 190oC, 2.16kg 1.0 30 5.0 30 30 Mooney Viscosity ASTM D1646 ML 1 + 4 at 121oC 26 < 5 8 < 5 < 5 Durometer Hardness, Shore A ASTM D-2240 50 72 75 85 94 DSC melting Peak, oC Rate: 10oC/min 33 60 60 78 98 Glass Transition Temp, oC DSC inflection point -61 -57 -56 -51 -44 Flexural Modulus, MPA ASTM D-790, 2% Secant 3.5 12.1 12.1 25.8 69.9 Ultimate Tensile Strength, MPa ASTM D-638, 508mm/min 2.1 3.3 6.9 6.4 12.9 Ultimate Elongation, % ASTM D-638, 508mm/min 975 >1000 >1000 950 790 Table 5-2: Structures of reactive materials of interest Name Lupersol 101 DEGDA Structure Name Styrene TMPTA Structure

PAGE 171

149 5.2.2 Methods 5.2.2.1 Processing All polymers were dried in an air circ ulating oven at 40C for 24 hours prior to compounding. Before processing, the resi ns were premixed by hand for about 10 minutes. Monomer/initiator mixtures were magnetically stirred for 10 minutes before processing and a choice amount of the mixtur e was added to the dry polymer pellets before processing. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin screw extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D ratio of 39. Figure 5-1 is a schematic of the extruder along with a typical te mperature profile. Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed) 195C 205C 210C 210C 200C 190C 180C 165C Figure 5-1: Schematic drawing of the r eactive twin screw extruder and a common temperature profile. The temperature of the extruder was regulat ed by electrical resistance and water circulation in the barrels. The dried, premixed resins were then introduced into the extruder at 60 g/min through a screw dry material feeder, Accu Rate, In c. A Zenith pump controlled the rate of monomer/initiator so lution addition into the extruder. The screw

PAGE 172

150 speed, unless otherwise noted was 150 rpm. Devolatili zation was carried out by a vacuum pump, VPS-10A, Brooks Equipment Co mpany. This was placed near the die and created a pressure of about 15 in Hg. The extruder was always starved to feed. After compounding, the resulting strands which exit the die are quenched in a water bath, pelletized, and dried in a vacuum oven at 100C, 28 in Hg for 24 hours. 5.2.2.2 Mechanical properties In order to measure the strength of the mate rial at very high test ing rates, a notched Izod impact test was performed according to ASTM D256 standards. The pellets were placed in a mold with 6 sl ots, each measuring 0.5x0.5x2.5in3. The mold was put in a Carver press (Fred S. Carver, Inc.) at 200C a nd after the material melts, the pressure was raised to 5000 psi. After waiting for 5-10 minutes, the pressure was slowly increased up to 10,000 psi. After another 5 minutes, the heat was turned off and the sample was let to cool down to room temperature at about 1.5 C/min. The bars were then taken out and notched with a Testing Machines, Inc. (TM I) notching machine. Before testing, they were conditioned at room temperature for at least 24 hours. A 30 ft-lb hammer was used with test method A on a TMI Izod impact tester. At least 5 bars were broken and impact strength is recorded as an average regardless of full or partial break. For stress-strain measurements, dried pe llets were placed in a mold measuring 15x15 cm2 x 1 mm thick. The mold was put into the Carver press at 200C and after the material melts, pressed up to 5000 psi. Afte r a 5-10 minute wait, the sample was slowly pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath. Specimens were tested according to ASTM D638 standards. Type V specimens were punched out of the compression molded sheet with a die, measuring 1 mm thick, 2.95 mm gauge width and 9.5 mm ga uge length. Five samples were tested after conditioning

PAGE 173

151 at room temperature for 48 hours. The mach ine used to test the samples was an MTS Model 1120 Instron, using a 1000 lb load cel l at a test speed of 12.7 mm/min 5.2.2.3 Chemical composition Fourier Transform Infrared Spectrosc opy (FTIR) was performed on a Nicolet 20SXB Spectrometer. 256 Scans were take n from 3500 to 500 cm-1 wavenumbers with a resolution of 4. Measurements were done in transmission mode on thin films (~2-3 microns). Films were produced by melting th e polymer in a Carver press at 180C and 10,000 psi for 2 minutes then quenching in a water bath at room temperature. For quantification of the grafted styrene, a calibration curve had to be established as explained in Appendix A. Areas under the 700cm-1 and 1376 cm-1 peaks are compared and related to absolute styrene amounts to ge t the grafting efficiency (GE). This is defined by the following equation: Amount of Monomer Grafted to Polymer Backbone GE = X 100% Original Amount of Monomer Pumped into Extruder All reactively extruded materials must be purified before quantifying the GE. This first involved dissolving th e pellets of crude graft copolymer in hot xylene at a concentration of ca 4% (wt/vol). The hot so lution was precipitated into ten volumes of acetone (a known non-solvent for the LLDPE, HDPE and PP, and a solvent for styrene monomer and homopolymer based on solubil ity parameters [162]). The unreacted monomers, styrene and DEGD A or TMPTA homopolymers and copolymers remained soluble in acetone and were separated out from the grafted polyolefins. The precipitated graft modified alloy was filtered, washed, and then vacuum dried at 70C for 24 hours. FTIR showed that the GE level remained unaltered upon further rounds of purification. Therefore, one purification step was sufficien t for removal of all the residual impurities. (5.1)

PAGE 174

152 5.2.2.4 Rheology Melt Flow Index (MFI) testing was done according to ASTM D1238 (230C and 2.16 kg weight) on a Tinius Olsen model MP 933 Extrusion Plastometer. For materials with an flow rate of 0.5-3.5 g/10 min, the we ight of the samples was approx. 3 g, whereas materials with flow rates of 3.5-300 g/10 mi n, the sample weight was approx. 6 g. All materials were dried under vacuum then condi tioned at room temperature before testing, with an average of two runs reported. 5.3 Results and Discussion 5.3.1 Effect of Processing Temperature Controlling the extruder barrel temperatur e allows one to control or influence several parameters, specifically the poin t in which reactions will start and reach completion, rate of plastification/melting of pol ymers, rate of initiator decomposition and hence rate or reaction, length of polys tyrene grafts and homopolymer, and degradation/crosslinking of polymers. These all have implications on the final mechanical properties of the alloy. Table 5-3 gives the three di fferent temperature profiles of interest, with the highest temperature experienced in Zone Table 5-3: Three different temperature profiles for the extrusion of 80:20_C. Zone 8 (Die) Zone 7Zone 6Zone 5Zone 4Zone 3 Zone 2 Zone 1 (Feed) Sample 1 180C 185 190 190 185 180 175 165 Sample 2 195C 205 210 210 200 190 180 165 Sample 3 215C 225 230 230 220 210 190 165 At low temperatures, the initiator dissoc iation will occur at a later point along the extruder because of a longer half life. Therefore, attack onto molten PP by primary radicals is more likely because the concentr ation of initiator derived radicals will be higher at a later point along the barrel. In itiator and monomer may be caged in the

PAGE 175

153 highly viscous melt as concentrated pockets which likely preferentially attack the major phase polypropylene. So, the benefits of polymers co-melting and maximum monomer conversion may not be realized. C Temperature ( ) 190210230 Grafting Efficiency 10 20 30 40 50 60 70 Notched Izod Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 2 4 6 8 10 12 14 Grafting Efficiency Impact Strength MFI Figure 5-2: Effect of extrude r barrel temperature on room te mperature impact strength, melt flow index, and grafting efficiency of 80:20_C. The sensitivity of alloy perf ormance on barrel temperature is seen in Figure 5-2. Grafting efficiency is lowest for the low te mperature sample, peaks at an intermediate temperature, and decreases at the highest te mperature. It typically increases with temperature because the activation energy for graft initiation is less than homopolymerization [7, 249] but another author found that at 220C, efficiency drops off because of premature reaction of monomer and initiator before proper mixing in the polymer melt [166]. At the highest temperatur e, initiator radicals become less selective as far as hydrogen abstraction is concerned. There will be a large concentration of

PAGE 176

154 radicals, but initiator decomposition occurs at such a fast rate that the grafting reaction cannot keep up with all other competing reac tions. It should also be noted that no grafting will occur after all initiator decomposes [152]. The low impact strength for the low temper ature sample is due to the degradation of PP and reduction of entanglements. But the reason for the low impact strength at the high temperature could be due to the fact that homopolymerization dominates over polyolefin grafting. Excess homopolymerized ma terial can deleteriously affect impact strength [68]. Also, increasing temperature leads to a lower viscosity of both polymer melts, which promotes coalescence of the dispersed phase [411]. Chapters 3 and 4 prove that coalescence of the disp ersed phase is not conducive to high impact strengths. Melt flow index is also graphed in Fi gure 5-2, which is highest for the low temperature samples. This is due to a high concentration of primary free radicals generated in the polypro pylene melt, leading to chain scission. MFI is low for the high temperature sample because consumption of ra dicals by monomers is fast. There may be a highly entangled network of homopolymeri zed material and slightly crosslinked ethylene copolymer, with a relativel y lower degree of PP degradation. Stress-Strain behavior (Table 5-4) parallels the respons e of impact strength at various barrel temperatures. Elastic modulus is a measur e of the stiffness of the amorphous phase of the alloy and is hi ghly dependent upon the chain entanglement density as well as distribution of crystallites which act as pseudo-cros slink junctions. At Table 5-4: Stress-strain behavior of 80:20_C as a function of temperature. Max Extruder Temp (C) Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 190 1220 39 24.1 0.8 50 10 32.9 3.2 3112 844 210 1356 110 25.3 2 56 10 35.9 4.6 3488 1208 230 1203 62 23.8 1.8 44.5 4 32.4 2 2970 745

PAGE 177

155 low extrusion temperatures, de gradation is high and chain en tanglement density is lowest from MFI data. For the highest temper ature, there may be a high degree of homopolymerized materials, which may be enta ngled, but does not have the load bearing capacity of the PP phase which has both a hi ghly entangled amorphous phase and crystal blocks reinforcing it. Yield stress is highl y dependent upon the crystalline state and if mobility at and around the surface of lamellar crysta ls is easier, yield stress will decrease. Yield stress is low for both 190C and 230C samples which coincidentally have low grafting efficiencies resulting in phase se parated homopolymer. Elongation at break, stress at break, and energy to break are a ll dependent upon the breakup of crystals and subsequent recrystallization. Defects and high interfacial te nsion decrease the magnitude of each of these properties [261, 262]. So, a high GE at 210C reduces chain degradation (defects) and improves interfaci al adhesion between phases. 5.3.2 Effect of Screw Speed The size of the dispersed phase is known to decrease with increasing shear rate but at very high shear rates, the particles actua lly increase in size [41]. Polymer melts are shear-thinning materials, so w ith increasing shear rate, the matrix viscosity can decrease sharply. At higher shear rates, the droplets have hi gher approach velocities and thus the coalescence probability can increase [41, 341]. Besides it effect on domain size, screw sp eed also dictates the residence time (length of time polymer exists in the extruder), degree of mixing, and point at which the polymers become fully molten. The overall grafting yield is dictated by the local residence time in the plastification zone and not by the overall time in the entire extruder [90]. Table 5-5 lists the actua l residence times at varying screw speeds for the reactive twin screw extruder. Two heat sources are responsible for plastif ication of polymers:

PAGE 178

156 barrel heating by conduction and viscous di ssipation by the rotation of the screws (interparticle friction). Th e higher the screw speed, the greater amount of heat from viscous dissipation, but the lower amount of heat transferred to the polymers by conduction due to reduced residence time. Pl astification is a physical process, while grafting is a chemical one. Figure 5-3 indi cates that at varying screw speeds, MFI, impact strength, and grafting efficien cy are interrelated for each alloy. Table 5-5 Screw speed rela tionship to residence time Screw Speed (rpm) 75 100 150 200 250 Residence Time (min:sec) 2:38 2:12 1:46 1:35 1:28 An increase in rotation rate results in a decrease of monomer conversion because of a decrease in mean residence time as well as a decrease in the number of fully filled chambers and overall volume of extruder barr el [7, 77, 90]. Also, the reaction will start at a later point in the extruder. Figure 56 shows that there is a reduction in grafting efficiency of PP as screw speed increases. Grafting will start at a later point along the extruder so homopolymerization may dominate because monomer is not able to diffuse within the PP phase. Also, the liquids ma y prematurely react within the copolymer phase, which is known to decrease the grafting efficiency of PP. Immiscible mixtures always have multiple melt phases regardless of the intensity of mixing [10]. Melt flow index decreases with increasing screw rpm or residence time, which is confirmed by another author [72] This may also be related to the grafting or crosslinking of the PE phase, while also reducing the degr adation of PP. MFI continually decreases with increasing screw speed because of be tter dispersion of elastomer and therefore greater surface area to bond the phases at the interface. There is an increasing rate of plastification and a low residence time, so PP degradation is limited.

PAGE 179

157 Screw Speed (rpm) 75100150200250 Notched Izod Impact Strength (ft-lbs/in) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Melt Flow Index (g/10min) 0.6 0.9 1.2 1.5 1.8 2.1 2.4 2.7 3.0 Grafting Efficiency 20 30 40 50 60 70 80 90 100 110 120 130 Impact Strength MFI Grafting Efficiency Figure 5-3: Impact strength, MFI, and graf ting efficiency of 80:20_C as a function of screw speed. Longer residence times lead to larger di mensions of the dispersed phase [411], which may have a negative effect of impact strength, as shown in Figure 5-3. Increasing shear rate enhances reaction rate but levels off after a certain screw speed [71], and 150 rpm appears to be the critical point where both impact strength a nd grafting efficiency level off. At higher screw speeds, particles collide less [74, 148, 245], leading to a higher impact strength. But the slight decrease in impact strength and increase in grafting efficiency at 250 rpm may be due to the fact that polymer melts are shear thinning. So with increasing shear rate, the matrix viscosit y can decrease sharply leading to increasing particle size [32]. Impact st rength increases with screw sp eed but plateaus at high screw speeds along with elongation at break and elastic modulus [148]. From stress strain data in Table 5-6, the optimal screw speed appears to be at 150 rpm. At low screw speeds, degradation of PP is present which means the high molecular weight, crystallizable chains are affected. So, the breakup of lamellar crystals is easier

PAGE 180

158 but the recrystallization and a lignment of chains is severely restricted. Cha and White also found that elongation at break increases with increasing screw speed [147]. Elastic modulus increases with increases with screw sp eed but tails off similar to the behavior of impact strength at the highest speed. This may be due to coalescence of the low viscosity elastomer phase. Table 5-6: Stress-strain behavior of 80:20_C as a function of screw speed. Screw Speed (rpm) Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 75 1309 52 22.3 0.7 30.5 13 26.2 3 1708 822 100 1320 60 24.3 1.7 35 14 28.5 2.1 1981 999 150 1356 110 25.3 2 56 10 35.9 4.6 3488 1208 200 1310 80 24.2 1.1 50 10 33.9 1.8 3292 1123 250 1234 58 24.5 0.3 46 4 33.2 1.9 3089 658 5.3.3 Effect of Initiator Concentration A basic concept of free radica l grafting is that a peroxide initiator must be present so as to abstract hydrogen atoms from the pol yolefin to start the grafting process. But side reactions like -chain scission (PP) and crossli nking (PE and PE copolymers) take effect with excess peroxide. Figure 5-4 shows that grafting efficiency increases continuously as initiator concentration increa sed. By adding more peroxide, the number of primary free radicals increases, th us improving the probability for hydrogen abstraction from the PP or 8407 backbone. Gr afting degree has been shown to increase with initiator le vel for a variety of polyolefins, including LLDPE [95, 196], HDPE [7, 249], EPR [412], and PP [40, 94, 138, 205]. M onomer conversion also increases [7, 92]. Although the grafting efficiency does increase the monomers grafted may actually bond to chains that have already undergone chain scission. Melt flow index continually increases with increasing peroxide concentration. This indicates that the PP phase is preferentially degraded (as opposed to 8407 crosslinking).

PAGE 181

159 This same result has also been found by a variety of authors [ 7, 40, 73, 228]. Initiator concentration and its efficiency were the mo st important variable affecting MW, MWD, Wt % Initiator 0.150.300.450.60 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 10 11 Grafting Efficiency 20 30 40 50 60 70 80 90 100 110 120 Impact Strength MFI Grafting Efficiency Figure 5-4: Impact strength, MFI, and graf ting efficiency of 90:10_B as a function of initiator concentration. and the rheological properties of the product [4]. An incr ease in peroxide efficiency decreases all MW averages, with Mz most sensitive [70, 73, 148]. Izod impact strength increases from 0.15% to 0.3% initiator, but decreases in a linear fashion at increasing concentrations. This could be a direct result of the degradation of the PP chains [7, 73, 148, 233]. Also, the PE phase may be crosslinking to a greater extent, so its mobility and energy absorption capabilities are diminished. There is a drastic increase in grafting efficiency at 0.3% initiator and a peak in impact strength, which are likely interrelated. This mu st be the point in which grafting and

PAGE 182

160 homopolymerization are balanced to give the highest grafting efficiency with the lowest amount of degradation. Tensile properties, such as yield stress, elastic modulus, elongation at break, stress at break, and energy to break, are given in Tabl e 5-7. Similar to impact strength, elastic modulus and break stress reach a peak at 0.3% peroxide and decrease at higher concentrations. Yield stress remains about th e same for all samples, although the lowest initiator concentration has th e lowest yield stress [341]. Elongation, degree of strain hardening, and energy to break decrease continuously with increasing peroxide, concurrent with other re sults [148, 184, 186, 220, 233, 413]. In one case, elongation at break and hence energy to break was not found to be reproducible [222]. Table 5-7: Stress-strain beha vior of PP:8407 alloys at a ra tio of 90:10 as a function of initiator concentration. wt% Initiator Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 0.15 1502 74 30.5 1.7 60 8 40.6 3.3 5164 1091 0.3 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 0.45 1657 44 30.9 0.6 54 8 39.3 3 4002 898 0.6 1617 103 31 1.8 35 11 29 3.4 1944 892 The reason for the low stress-strain propertie s at high initiator c oncentrations could be due to the fact that th e % crystallinity is known to decrease [148, 220, 233, 341] as well as melting temperature [222, 233]. Also, the onset of melting of a PP/LLDPE/peroxide alloy decreases by 9C with 0.25% peroxide [222]. So, the size and distribution of the lamellar crystals are affected. 5.3.4 Effect of Multifunctio nal Monomer Concentration The multifunctional monomer serves several purposes in this reactive extrusion process and is an integral part in improvi ng mechanical properties. Figure 5-5 reveals that melt flow index, as expected, decreases with increasing DEGDA content. This

PAGE 183

161 multifunctional monomer has four effective branching sites, which act to reduce degradation by linking up degraded PP chai ns, enhance crosslinking of the 8407 phase, and consume free radicals. Previous graf ting studies show that MFI decreases and viscosity increases with increasing monomer c oncentration, attributed to reduced chain Wt % DEGDA 0.00.20.40.81.1 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 10 11 Grafting Efficiency 20 30 40 50 60 70 80 90 100 Impact Strength MFI Grafting Efficiency Figure 5-5: Notched impact st rength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 90:10 as a function of multifunctional monomer concentration. degradation [40, 92, 228]. Molecular wei ght increases with increasing monomer concentration as well as broa dening of molecular weight di stribution and increase in the high MW tail [250]. High monomer concentrat ions reduce the melt st rength of the alloy [215] and could have implica tions on elongation at break. Impact strength increases concurrently up to a point (0.8 wt%) then dramatically decreases at the highest concentration, wh ere homopolymerization begins to dominate over grafting. Both homopolymer and grafte d monomer positively affect impact strength

PAGE 184

162 [148]. A monomer coagent is known to impr ove impact strength for PP-LLDPE alloys [230, 250]. At the highest DEGDA concentration, a more densely branched material is present along with the possibi lity of phase separated homopol ymer. Impact strength is reduced because the ability to absorb and di ssipate stress is low and there may be weak interfaces with reduce stress tr ansfer capability. A high degr ee of crosslinks respond in a brittle manner to high deformation rate test s. At high concentration, homopolymer or copolymer can phase separate [68], which dele teriously affects all mechanical properties. Grafting efficiency also drops at the sa me point as impact strength. Increasing monomer concentration increas es amount of material graf ted but grafting efficiency levels off and in some cases decreases at high monomer concentrations [7, 40, 92, 94, 95, 180, 187, 190, 205, 228]. This indicates that DE GDA is trapping radicals and creating a high degree of branched hom opolymer material. Although st yrene is still grafting onto the polymer phases, DEGDA is causing styren e to preferentially copolymerize over grafting. Addition of only 0.2 wt% DEGD A shows the greatest styrene grafting efficiency, which means that at higher concentrations, th e materials being grafted may not necessarily be as single monomer units or chains but as branched macromolecules. At 0% DEGDA, grafting is low because the ra te of reaction is slow er and chain scission more likely [153, 161]. Also, styrene would prefer to add more monomer than chain transfer to PP or PS [162a]. Table 5-8 shows that the stress-strain properties in crease up to 0.8 wt% DEGDA, then decrease just as impact strength and gr afting efficiency decrea se. Typically, elastic modulus and yield strength increase with increasing monomer concentration [184, 222, 230, 250]. Elongation at break does slightly increase with increasing monomer

PAGE 185

163 concentration, a trend also show n in literature [7, 259]. The crosslinked sample gave the lowest energy to break and elongation at break, but a lower modulus than the 0.8 wt% DEGDA sample. The low strain portion of a stress strain curve is dependent upon the amorphous nature of the material, tie molecule s, short range defects, etc. With the possibility of a gel present in the vicinity of crystalline phase and Table 5-8: Stress-strain beha vior of PP:8407 alloys at a ra tio of 90:10 as a function of multifunctional monomer concentration. wt% DEGDA Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 0 1501 74 28.3 0.9 57 4 37 1.1 4205 736 0.2 1458 59 29.6 1.7 56 6 38.3 3.9 4192 649 0.4 1526 14 29.3 0.9 55 11 38.5 3 3961 1300 0.8 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 1.2 1572 52 29.6 1 39 8 34.1 3.7 3064 1200 glassy tie molecules, there may be phase se parated, long-chained bran ched or crosslinked material that prevents tie molecules from b ecoming taut upon applica tion of stress. Also, there may be a reduction in % crystallinity at such high monomer concentrations [7, 148, 184, 215]. The melting transition is seen to br oaden for PP due to degradation [148] or creation of larger molecules [215]. 5.3.5 Effect of Styrene Concentration wi th DEGDA as Multifunctional Monomer Styrene monomer has proven to be an inte gral part in the toughness improvement of these alloys. Figure 5-6 shows the effect of styrene on the room temperature impact strength, grafting efficiency, and MFI of all oys with a PP:8407 ratio of 90:10. For impact strength at various styrene concentrations, a p eak is found at 6% styr ene, an intermediate concentration. As in the previous sec tion, a combination of homopolymer and graft copolymer will improve impact strength [148, 230, 250]. But with more monomer, longer grafts may be possible and the formati on of large glassy polystyrene domains may phase separate and decrease impact strength.

PAGE 186

164 MFI decreases with increasing monomer concentration, presumably because the probability that a PP macroradical will reac t with styrene monomer is enhanced. One author believes that crosslinking of PP during free radical grafting of styrene occurs by termination through combination of two PP-styr ene radicals [95]. With more monomer, a greater concentration of highly en tangled chains is more likely. Wt % Styrene 369 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 10 11 Grafting Efficiency 40 50 60 70 80 90 100 Impact Strength MFI Grafting Efficiency Figure 5-6: Notched impact st rength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 90: 10 as a function of styren e monomer concentration The grafting efficiency increases with incr easing styrene content, which is to be expected because as hydrogen atoms are bei ng abstracted from the polymer backbones, the larger amount of styrene av ailable leads to greater probab ility of styrene grafting onto the polymer rather than chain scission or cr osslinking. Also, longer grafts may also be present at this high st yrene concentration. The stress-strain performance of the materials as a function of styrene concentration is given in Tabl e 5-9. At low styrene conten t, break stress and energy to break are significantly less than high concentr ation samples. This may be due to the

PAGE 187

165 degradation of PP and disruption of crystalline phases. For a high strain rate test like Izod impact, the material acts in a brittle ma nner, but for low strain rates, the polymer chains are given time to disentangle and st retch. For this reason, the stress-strain properties are better for the highest m onomer concentration over the lowest. Table 5-9: Stress-str ain performance of PP:8407 alloys at a ratio of 90:10 as a function of styrene concentration. wt% Styrene Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 3 1612 83 29.8 1.1 50 13 37 4.7 3339 1103 6 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 9 1613 37 29.6 1.4 57 10 41.8 4.6 4955 1071 5.3.6 Effect of Styrene Concentration wi th TMPTA as Multifunctional Monomer In another study at higher ENGAGE cont ent (80:20 ratio of PP:8407 vs. 90:10 in the previous section) and using TMPTA rather than DEGDA as the multifunctional monomer, similar trends are observed at increasing styrene concentration. The data in Figure 5-7 shows that Impact strength and gr afting efficiency follow similar paths. At low styrene contents, degradation of the pol ymer chains dominate because the ratio of initiator-derived free radicals to monomer is great. Addition of only 2 wt% styrene to the system results in a 10-fold decrease in MFI a nd a 4-fold increase in impact strength. This is due to the ability of styrene to both act as a vector fluid and locat e the peroxide at the interface but also to polymerize and graft onto the polyolefin backbones. At the highest concentration of styrene (10 wt%), grafting efficiency falls off most likely due to homopolymerization. For higher 8407 concentrations, there is less PP for styrene to graft onto, which may be the reason why GE drops off in Figure 5-7 and not in Figure 5-6. Hu et al. gives a very good explanation as to why there would be a difference in grafting yield [205]. In PP, each polypropyl ene moiety contains one tertiary hydrogen

PAGE 188

166 atom, two secondary hydrogen atoms, and three primary hydrogen atoms, with concentrations of 0.238, 0.475, and 0.713 mpr (mol es per hundred grams of resin), Wt % Styrene 0246810 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 5 10 15 20 25 30 35 40 Grafting Efficiency 30 35 40 45 50 55 60 65 70 75 80 Impact Strength MFI Grafting Efficiency Figure 5-7: Notched impact st rength, melt flow index, and grafting efficiency of PP:8407 alloys at a ratio of 80:20 as a f unction of styrene concentration. respectively. This means that if a mono mer (molar mass = 196 g/mol) has a grafting yield of 10 phr (0.0497 mpr) it represents only one-fifth of the molar number of the tertiary hydrogen atoms. Because the number of consumed tertiary hydrogen atoms is small, every additional amount of the grafti ng solution leads to a proportional increase in the monomers grafting yield as seen in Fi gure 5-6. Addition of 8407 (or a reduction in polypropylene tertiary hydrogen atoms) will limit the available sites fo r grafting and thus promote homopolymerization while reducing grafting efficiency.

PAGE 189

167 5.4 Conclusions This chapter elaborates on the idea th at when creating and developing materials with complex architectures and many component s, processing can play an important role in defining its macroscopic properties. A hi gh screw speed has been shown to give the best mechanical properties, which is rela ted to better mixing a nd grafting within the alloys. If the barrel temperature of the ex truder is too low, polypropylene will degrade preferentially, but if too hi gh, homopolymerization rather than grafting of monomers will dominate. There is an optimum styrene, initiator, and multifunctional monomer concentration for these alloys, but these results reveal that a high grafting efficiency does not necessarily translate into great mechanical properties. Reactive extrusion of PP and 8407 involves degradation and cr osslinking reactions, which if not restricted, diminish the mechanical integrity of th e alloy. So, a high grafting effi ciency may come at the price of a largely degraded PP phase.

PAGE 190

168 CHAPTER 6 CONTROLLING ALLOY PERFORMANCE BY VARYING ELASTOMER PROPERTIES 6.1 Introduction Very few studies have been conducte d on elastomer toughened polypropylene blends with varying elastomer crystallinity or molecular weight. Regarding reactive blends of PP-elastomer blends, there have been no systematic studies that focus on alloy performance as a function of elastomer molecu lar weight or crystal linity. A comparative study of several alloys with varying grad es of ENGAGE polyolefin elastomers was done in order to assess the robustness and furt her explain the fundament als of the reactive extrusion process introduced in Chapter 4. Th e elastomers were categorized according to their density (% crystallinity) and melt flow i ndex (viscosity or molecular weight). At approximately the same density, low crystall inity elastomers with varying molecular weight are focused on. A high molecular wei ght translates into more entanglements, higher viscosity, and difficult y dispersing, which all should have some effect on the properties of the alloys. The low molecular weight (high MFI) elastomers were chosen because they are easily dispersed in the highly viscous PP matrix and should not compromise processability. 6.2 Experimental 6.2.1 Materials Table 6-1 gives a list of pertinent ethylene-1-octene copolymers produced by Dupont Dow elastomers under the trade name ENGAGE, but the EOC grade of interest

PAGE 191

169 is 8407 [48]. Isotactic pol ypropylene homopolymer was supplied by Equistar Chemical (grade PP 31S07A) and is contact translucent. All polymers were received in pellet form. Table 6-1: ENGAGE product data table. Engage Grade (decreasing comonomer content) 8842 8407 8200 8401 8402 Comonomer Content wt% 13C NMR/FTIR 45 40 38 31 22 Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902 Melt Index, dg/min ASTM D1238 190oC, 2.16 kg 1.0 30 5.0 30 30 Mooney Viscosity ASTM D1646 ML 1 + 4 at 121oC 26 < 5 8 < 5 < 5 Durometer Hardness, Shore A ASTM D-2240 50 72 75 85 94 DSC melting Peak, oC Rate: 10oC/min 33 60 60 78 98 Glass Transition Temp, oC DSC inflection point -61 -57 -56 -51 -44 Flexural Modulus, MPA ASTM D-790, 2% Secant 3.5 12.1 12.1 25.8 69.9 Ultimate Tensile Strength, MPa ASTM D-638, 508 mm/min 2.1 3.3 6.9 6.4 12.9 Ultimate Elongation, % ASTM D-638, 508 mm/min 975 >1000 >1000 950 790 The peroxide and monomers used in this st udy were reagent grade chemicals (structures are shown in Table 6-2). The monomers we re purified by passing through an activated alumina column before use. Styrene mono mer, inhibited by 10-15 ppm t-butyl catechol, was purchased from Fisher. The initiator 2,5dimethyl-2,5-di-(t-bu tylperoxy) hexane, was purchased from Atofina under the trad e name Lupersol 101. Diethyleneglycol diacrylate (DEGDA), inhibited by 80 ppm Hq and 120 ppm MEHQ was graciously donated by Sartomer, an Atofina company. Ta ble 4-3 lists structur es of the reactive materials.

PAGE 192

170 Table 6-2: Structures of reactive materials of interest Name Lupersol 101 TMPTA Structure Name Styrene DEGDA Structure 6.2.2 Methods 6.2.2.1 Processing All polymers were dried in an air circ ulating oven at 40C for 24 hours prior to compounding. Before processing, the resi ns were premixed by hand for about 10 minutes. Monomer/initiator mixtures were magnetically stirred for 10 minutes before processing and a choice amount of the mixtur e was added to the dry polymer pellets before processing. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin screw extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D ratio of 39. The temperature of the extruder was re gulated by electrical resistance and water circulation in the barrels. The dried, premixed resins were then introduced into the extruder at 60 g/min through a screw dry material feeder, Accu Rate, In c. A Zenith pump controlled the rate of monomer/initiator so lution addition into the extruder. The screw speed, unless otherwise noted was 150 rpm. Devolatili zation was carried out by a vacuum pump, VPS-10A, Brooks Equipment Co mpany. This was placed near the die and created a pressure of a bout 15 in Hg. The extruder was always starved to feed.

PAGE 193

171 Figure 6-1 is a schematic of the extruder, w ith a typical temperature profile. After compounding, the resulting strands which exit the die are quenched in a water bath, pelletized, and dried in a vacuum oven at 100C and 28 in Hg for 24 hours. Supercritical CO2-assisted processing was carried out in the twin screw extruder with an Isco 260D syringe pump connected at zone 5. Industrial grade carbon dioxide was supplied by Praxair. Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed) 195C 205C 210C 210C 200C 190C 180C 165C Figure 6-1: Schematic drawing of the r eactive twin screw extruder and a common temperature profile. 6.2.2.2 Mechanical properties In order to measure the strength of the mate rial at very high test ing rates, a notched Izod impact test was performed according to ASTM D256 standards. The pellets were placed in a mold with 6 sl ots, each measuring 0.5x0.5x2.5in3. The mold was put in a Carver press (Fred S. Carver, Inc.) at 200C and after the material melts, pressed up to 5000 psi. After the material melts, the pre ssure was slowly increased up to 10,000 psi. After another 5 minutes, the heat is turned off and the sample is let to cool down to room temperature at about 1.5C/min. The bars were then taken out and notched with a

PAGE 194

172 Testing Machines, Inc. (TMI) notching machin e. Before testing, they were conditioned at room temperature for 24 hours and a 30 ft -lb hammer was used with test method A on a TMI Izod impact tester. At least 5 bars were broken and impact strength is recorded as an average regardless of full or partial break. A Kodak EasyShare CX7300 was used to take digital images of fractured impact bars. For stress-strain measurements, dried pelle ts were placed in a mold measuring 15x15 cm2x1 mm thick. The mold is put into the Carver press at 200C and pressed up to 5000 psi after the material melts. After a 5-10 minute wait, the sample was slowly pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath. Specimens were tested according to ASTM D638 standards. Type V specimens were punched out of the compression molded sheet with a die, measuring 1 mm thickness, 2.95 mm gauge width and 9.5 mm ga uge length. Five samples were tested after conditioning at room temperature for 48 hours. The mach ine used to test the samples was an MTS Model 1120 Instron, using a 1000 lb load cel l at a test speed of 12.7 mm/min A Seiko DMS220 interfaced with a Seiko Rheostation model SDM/5600H was used to test dynamic mechanical specimens. Testing was conducted from -120 to 150C at a heating rate of 5C/min in dry nitrogen atmosphere maintained at an approximate flow rate of 100 mL min-1. Rectangular samples (20x10x1mm3) were cut from the compression molded sheet and tested in bending mode at a frequency of 1Hz. 6.2.2.3 Chemical composition and molecular structure Fourier Transform Infrared Spectrosc opy (FTIR) was performed on a Nicolet 20SXB Spectrometer. 256 Scans were taken from 3500 to 500 cm-1 wavenumbers with a resolution of 4. Measurements were done in transmission mode on thin films (~2-3

PAGE 195

173 microns). Films were produced by melting th e polymer in a Carver press at 180C and 10,000 psi for 2 minutes then quenching in a water bath at room temperature. For quantification of the grafted styrene, a calibration curve had to be established as explained in Appendix A. Area under the 700cm-1 and 1376cm-1 peaks are compared and related to absolute st yrene amounts to get the grafting effi ciency (GE). This is defined by the following equation: Amount of Monomer Gr afted to Polymer Backbone GE = X 100% Original Amount of Mo nomer Pumped into Extruder All reactively extruded materials must be purified first before quantification using FTIR. This first involved dissolving the pell ets of crude graft copolymer in hot xylene at a concentration of ca 4% (wt/vol). The hot solution was precipitated into ten volumes of acetone (a known non-solvent for the LLDPE, HDPE and PP, and a solvent for styrene monomer and homopolymer based on solubil ity parameters [162]). The unreacted monomers, styrene and DEGD A or TMPTA homopolymers and copolymers remained soluble in acetone and were separated out from the grafted polyolefins. The precipitated graft modified alloy was filtered, washed, and then vacuum dried at 70C for 24 hours. FTIR showed that the GE level remained unaltered upon further rounds of purification. Therefore, one purification step was sufficien t for removal of all the residual impurities. Gel permeation chromatography (GPC) was performed on a Waters GPCV 2000 calibrated using crosslinked polystyrene standards for relative molecular weight determination. The set temperature was 40C fo r THF as the solvent at a flowrate of 1.0 mL/min. Samples (about 9 mg) were dissolv ed in 9.5 ml (8.44 g) of HPLC grade THF purchased from Acros Organi cs and passed through 0.45 m filters. (6.1)

PAGE 196

174 6.2.2.4 Thermal analysis and rheology Differential scanning calorimetry (DSC ) was used to study the different thermodynamic transitions present in the bl ends. DSC was performed on a Seiko SII DSC 220C-SSC/5200, Seiko Instruments, equipped with a Seiko Rheostation model SDM/5600H and calibrated with indium and tin standards. Samples (approx. 7 mg in weight) were sealed in crimped aluminum pans, with the reference being 99.99% pure alumina. Purging of the sample was done with dry nitrogen at a flow rate of 100 ml/min. Each sample experienced two heating and coo ling cycles (shown in Table 6-3) with the first to erase prior thermal history. The second cycle is reported in all graphs. Table 6-3: DSC consecutiv e heating/cooling cycles Step Start Temp (C) End Temp (C) Heating/Cooling Rate (C/min) Hold Time (min) Sampling (s) 1 -70 200 20 3 3 2 200 -80 20 5 3 3 -80 200 10 5 1 4 200 -80 10 3 1 Melt Flow Index (MFI) testing was done according to ASTM D1238 (230C and 2.16 kg weight) on a Tinius Olsen model MP 933 Extrusion Plastometer. For materials with an flow rate of 0.5-3.5 g/10 min, the we ight of the sample was approx. 3 g, whereas materials with flow rates of 3.5-300 g/10 mi n, the sample weight was approx. 6 g. All materials were dried under vacuum then condi tioned at room temperature before testing. 6.3 Results and Discussion 6.3.1 Effect of Elastomer Dens ity on Alloy Performance Three different ethylene-octene co polymers (8407, 8401, and 8402) are ranked according to their % crystallinity. All other material properties and processing parameters are held constant unless otherwis e noted. Also, the ratio of PP to EOC was kept constant at 90:10. Ta ble 6-4 and Figure 6-2 give a key code and designated

PAGE 197

175 formulations to ease in the identification of each sample. The first number given is the grade of EOC, the second number/letter signi fies a formulation whic h can be referenced in Table 6-4. 8407_0 Grade of ENGAGE Designated Formulation Figure 6-2: Representation of sample reference code The weight % is defined as the percentage of material added in relation to the total weight of all ingredients. For example, a PP:8407 ratio of 90:10, 0.3 wt% Initiator and 6 wt% styrene means that out of a total of 100 gr ams, Initiator = 0.3 g, Styrene = 6 g, PP = 84.3 g, 8407 = 9.4 g. Table 6-4: Identification of Formul ations in relation to Figure 6-2 Alloy ID wt% Initiator wt% Styrene wt% DEGDA 0 (Physical Blend) 0 0 0 A 0.3 6 0 B 0.3 6 0.8 C 0.15 3 0.8 6.3.1.1 Physical blends From Figure 6-3, blend impact strength scales linearly with both copolymer density and copolymer melting temperature, both of wh ich are interrelated. Melting temperature of the copolymers is known to be inversely proportional to comonome r content [51]. The highly crystalline elastomer behaves in a brittle manner at high impact because the crystalline domains are not able to absorb and dissipate energy efficiently. Low temperature transitions cease to exist for thes e blends because chain mobility is hindered and the number of copolymer side groups is diminished. Low temperature transitions/relaxations contribu te to energy dissipation at high impact [100, 105, 107], so

PAGE 198

176 it is no surprise that impact strength decreases with decreasing number of octene units in the copolymer backbone. A dir ect indication that a polymer absorbs energy upon impact is through stress whitening. The whiter the impact bar, the great er amount of cavitation and dilation, and hence the gr eater impact strength [103, 328]. From Figure 6-6, the blend with the highly crystalline elastome r, 8402_0, completely breaks and shows only a slight amount of stress white ning at the surface of the broken bars. As the amorphous Density of Copolymer (g/cc) 0.8700.8850.902 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Melting Temperature of Copolymer ( C) 55 60 65 70 75 80 85 90 95 100 Impact Strength MFI Tm of Copolymer Figure 6-3: Impact strength, melt flow index, and melting temperature of physical blends as a function of the density of the copolymer. content of the elastomer increases, the imp act strength and degree of stress whitening follow suit. This is a visual confirmati on that elastomer density can have a profound effect on blend performance and morphology, ev en at 10 wt% elastomer. The melt flow index remains approximately the same for all bl ends and this is to be expected because the elastomers each have the same melt flow index as reported by the manufacturer. Stress-strain behavior of et hylene-octene copolymers links crystallinity to stress response [51, 52, 57, 59, 414]. The uniform elastomeric response of these copolymers is

PAGE 199

177 accounted for by a physical network built on fringed micellar crystal junctions and perhaps entanglements. Because the small fri nged micellar crystals are within their broad melting range at ambient temperature, they melt and reform during deformation. By physically blending the polymers, ther e should be an additive relationship between the mechanical performance and densit y of the elastomer [ 262, 274, 275]. Stress strain behavior is shown in Figure 6-4 and ta bulated in Table 6-5. Elastic modulus and Figure 6-4: Digital images of room temperatur e fractured Izod impact bars as a function of elastomer density in the physical blends. From left to right: 8407_0, 8401_0, 8402_0. yield stress are highest for the high density blend (8402_0), whereas elongation at break is lowest for the low crystalline elastomer bl end (8407_0). This is to be expected because the elastomer phase, which resides primarily in the interspherulitic regions of PP, will experience the applied load before the crystalline phase will. A higher elastomer Table 6-5: Stress-strain data of 8407_0, 8401_0, and 8402_0. Effect of Copolymer Density Blend ID Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 8407_0 1450 59 28 1.6 58 12 38.7 4.8 4038 825 8401_0 1459 40 29.6 1.1 68 10 42 4.4 5983 1059 8402_0 1596 89 29.8 1.4 64 5 39.2 3 4978 474

PAGE 200

178 crystallinity leads to a quicke r, stiffer response to the appl ied load, with crystals acting like filler particles. Yield stress is lowest for 8407_0 likely due to its diffuse, fringed micelle crystalline structure which has limited load bearing capability [52, 59]. The intermediate density copolymer blend, 8401_0, exhibits the highest elongation at break, energy to break, and stress at break, which may be due to greater interfacial interaction with PP or optimal combination of chain mobility and lamellar crystal distribution. Density of Copolymer (g/cc) 0.870.8850.902 Yield Stress (MPa), Elongation at Break (mm), Stress at Break (MPa) 20 30 40 50 60 70 Elastic Modulus (MPa) 1400 1450 1500 1550 1600 1650 Energy To Break (N/mm) 4000 4200 4400 4600 4800 5000 Yield Stress Elongation at Break Stress at Break Elastic Modulus Energy to Break Figure 6-5: Stress-str ain performance of PP/elastomer phys ical blends as a function of elastomer density. Because so many physical properties of the blend are dependent upon the crystallinity of the PP, DSC experiments were performed. From Figure 6-6, the general shape of each PP melting peak does not change with elastomer density (i.e. onset temperatures are approximately the same from Table 6-6). There are two distinct differences between the samples, though. Fi rstly, the peak at approximately 95C in 8402_0 can be attributed to the melting of PE crystals [48, 49, 51, 58]. This peak

PAGE 201

179 decreases in temperature and area as the de nsity (% crystallinity) of the elastomer decreases. Secondly, a small dip at about 150C is seen for each sample but decreases in intensity with decreasing elastomer crystallin ity. This melting peak is likely associated with the phase of PP. This crystal phase is known to be nucleated by many routes, including addition of elastomers [267, 325, 333, 339, 363, 383-385]. One reason for its creation is the disturbance of crystallization of PP into the form. So, 8402 appears to have the greatest effect on the crystallization process of PP, but this will be confirmed in Figure 6-7 and Table 6-7. Table 6-6: DSC endothermic data of 8407_0, 8401_0, and 8402_0 Sample ID PP phase melting peak temperature (C) Tm Onset (C) Melting Enthalpy (J/g) % Crystallinity 8407_0 165.2 152.9 82.3 39.8 8401_0 164.7 153.0 78.9 38.1 8402_0 164.9 152.6 75.1 36.3 Temperature ( ) 6080100120140160180 Endotherm ( W) -11000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 8407_0 8401_0 8402_0 C Figure 6-6: DSC melting endot herm of 8407_0, 8401_0, and 8402_0.

PAGE 202

180 The % crystallinity noticeably decreases with increasing elastomer density. This may be attributed to the id ea that lamellar crystals in 8401 and 8402 inhibit the secondary crystallization of PP. Secondary crystalliz ation is defined as further crystallization behind the crystallization front (i.e. within the sphe rulite) [362]. The p eak temperature of these alloys is highest for 8407_0, which may be due to the secondary crystallization process of PP being less restrict ed by the highly amorphous 8407. Temperature ( ) 020406080100120 Exotherm (uW) 2000 4000 6000 8000 10000 12000 14000 16000 8407_0 8401_0 8402_0 C 20406080100 1800 2000 2200 2400 2600 Figure 6-7: DSC cooling exot herm of PP/elastomer physical blends as a function of elastomer density. Insert is for th e temperature range or 20 to 100C. The physical blends can also be characterized by their cr ystallization behavior from the melt. Figure 6-7 and Table 6-7 show that the peak crystallization temperature (of PP) is highest for 8402_0, which also has the lowe st crystallization enthalpy. From this result, the crystal phase of 8402 may disrupt the primary or secondary crystallization processes of PP. Towards the end of prim ary crystallization, chain reptation and diffusion are the limiting factors. 8402 ma y be in its induction period before re-

PAGE 203

181 crystallization and towards the end of PP primar y crystallization. Th e insert in Figure 6-6 is a magnified image of crysta llization from 100C down to 20C. A distinct observation is the crystallization of poly ethylene in 8402_0. Long linear chains are able to organize and align in a parallel fashion for crystallizati on into lamellar structures. The PE phase of 8402_0 has a higher Tc, greater peak intensity, and a well defined peak, all attributed to the lower branch content and hence greater percent crystallinity. Table 6-7: DSC exothermic data of 8407_0, 8401_0, and 8402_0 Sample ID Peak temperature of crystallization (C) Enthalpy of crystallization (J/g) 8407_0 115.3 -83.1 8401_0 114.5 -80 8402_0 116.1 -74.5 The viscoelastic behavior of these phys ical blends has been characterized and typical results are seen in Fi gure 6-8. Dynamic mechanical testing has shown that the relaxation of polyethylene, which is related to the crystalline phase, does not exist for the low density copolymers [51, 415]. This is due to increased translational mobility of chain stems from crystal thickness and th e low degree of surface order. The -relaxation, which is attributed to glass transition temperature (Tg) of constrained, non-crystalline segments, increases in intensity with increasing com onomer content (more branching), leading to lower density. In the temperat ure range of -60C to 0C, an increase in molar volume of PE due to branching lead s to the creation of the relaxation. Higher comonomer content in the copolymer increases Tan peak intensity and shifts it to lower temperatures [415, 416]. This is related to the increased chain mobility and the decrease in activation energy for bond rotation and chain movement. The next peak at about 20C represents the glass transition temperature of PP. Both 8401_0 a nd 8407_0 are at lower peak intensities than 8402_0. A decrease is an indication of an incr ease in % crystallinity of the blend. DSC

PAGE 204

182 experiments (Figure 6-5, Table 6-5) come to the same conclu sion that crystalline content in 8402_0 is lower than the other two blends At temperatures above 40C, 8407 shows the highest slope and intensity up to 125C, followed by 8401_0 and finally 8402_0 (with Temperature ( ) -80-4004080120 Tan Delta 0.1 0.2 0.3 0.4 8407_0 8401_0 8402_0 -60-40-200204060 C Figure 6-8: Dynamic mechanic al analysis of 8407_0, 8401_0, and 8407_0. Insert is a magnified graph from -60C to 60C of the relaxation of PE. the smallest slope and lowest intensity). The peak (or shoulder) of PP represents lamellar crystal slippage and defect motion in and around them [45, 55, 197]. This is known to occur at about 80C whic h is 20C higher than the Tm of 8407, 2C higher than the Tm of 8401, and 18C below the Tm of 8402. The diffuse lamellar crystals in 8407 melt early, thereby increasing the intercrystallin e mobility of PP. It should also be noted that at the same molecular weight, 8407 will ha ve a smaller radius of gyration the either 8401 or 8402 because of a higher branch content and hence a lower number of entanglements in the amorphous phase. The gr eater branch content leads to enhanced mobility due to greater free volume.

PAGE 205

183 6.3.1.2 Alloys Upon modification of the physical blends with in itiator, styrene, and multifunctional monomer, mechanical property enhancement is drastic. But this section aims to make relative comparisons between the alloys based on the density of the elastomeric component. Figure 6-9 is an exam ple of how a slight in crease in branch Sample ID 8407_B8401_B8402_B Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 10 11 Grafting Efficiency 30 40 50 60 70 80 90 Impact Strength MFI Grafting Efficiency Figure 6-9: Room temperature impact strength, melt flow index, a nd grafting efficiency of 8407_B, 8401_B, and 8402_B. content can affect mechanical properties and gr afting efficiency of the alloys. From the data gathered in the previous section on phys ical blends, impact st rength is highest for 8407_0, followed by 8401_0 and 8402_0. The same trend follows for alloyed systems, with impact strength decreasing in th e following order: 8407_B > 8401_B > 8402_B. The most striking aspect of this graph is a comparison of grafting efficiency. 8401_B has a GE that is 28% lower than 8407_B, while 8402_B is 35% lower than 8407_B. This result has several important implications. It is direct evidence that styrene monomer grafts onto the elastomer phase. It also proves that the monomer/initiator mixture

PAGE 206

184 partitions in the elastomer phase as well as at the interface of the EOC and PP. The partitioning is likely due to the lower melting temperature of the EOC compared to PP [240]). Free radical grafting reactions happe n fast (> 70% completion after the melting zone [90-93]), so the majori ty of the monomers may react before all 8402 pellets are molten. Researchers have shown that monome rs will reside in the amorphous regions of a solid and not the crystalline portions [212]. With a more crystalline polymer, monomer may polymerize in the extruder before it is ab le to diffuse into the polymer and graft onto it. The higher the Tm of copolymer, the lower the plasti fication rate [90] and the longer the length of the solids conveying zone and melting zones needed for grafting. Another aspect of this graph is the melt fl ow index. MFI appears to increase with increasing density of elastomer, so more ch ain degradation is asso ciated with using a higher Tm elastomer phase. The minor phase will so ften first and coat PP particles, which then delay the melting of PP [13]. 8402 will soften at a later point along the extruder when compared to 8407, so PP will melt quicke r and thus degrade in the presence of a higher concentration of primary radicals. Th is is proof that in order to get a higher grafting efficiency and higher impact strength the elastomer must have some degree of solubility in the monomer mixture. This may promote long chain branching, prevent homopolymerization due to c ontact with the extruder barr el wall, and reduce chain scission of PP. At lower concentrations of styrene and initiator, the differences in grafting efficiency, impact strength, and melt flow inde x are not as contrasti ng as in Figure 6-9. From Figure 6-10, the impact strength does in crease in a linear fashion with decreasing elastomer density, which is to be expected. Melt flow index is only slightly higher for

PAGE 207

185 8402_C than 8401_C or 8407_C. This is similar to the trend in Figure 6-9. The grafting efficiency and impact strength of these sa mples are much lower than in Figure 6-9 because of the lower styrene and initiator con centrations in these samples. The standard deviation of grafting efficiency in Fig. 6-10 is too large to pinpoint a trend; therefore GE Sample ID 8407_C8401_C8402_C Notched Izod Impact Strength and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 Grafting Efficiency 15 20 25 30 35 40 45 50 55 Impact Strength MFI Grafting Efficiency Figure 6-10: Room temperature impact strength, melt flow index, a nd grafting efficiency of 8407_C, 8401_C, and 8402_C does not appear to change much with elastome r density. At these low levels of styrene and initiator concentratio n, homopolymerization may be favored over grafting. Stress-strain performance is exemplified in Table 6-8 with typical graphs in Appendix B. For the alloys that contain only styrene and initiator as the reactive ingredients, tensile properties follow the same trend as the physical blends the higher the density of elastomer, the better the prope rties. The PP chains degrade to a certain extent with initiator and styren e regardless of which elastomer is present, which results in relatively poor mechanical properties.

PAGE 208

186 Upon addition of multifunctiona l monomer to the alloys all properties improve except for elongation at break. The reason fo r the lower elongation at break may be because of a high degree of grafting in the PP pha se. This property is very sensitive to defects (i.e. grafted chains) which may dist urb the linearity of the polymer chains. 8407_B and 8407_C have higher elastic modulus than 8401_B and 8401_C, respectively. This may be due to the fact that the grafting efficiency is much higher for 8407 alloys, so a more highly branched structure exists in amorphous intercry stalline regions. Table 6-8: Stress-strain data of PP:ENGAGE alloys as a function of elastomer density. Effect of Copolymer Density Blend ID Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 8407_A 1501 74 28.3 0.9 57 4 37 1.1 4205 736 8401_A 1557 67 30.1 1.3 62 9 40 1.4 4812 1092 8402_A 1606 34 31.4 1.2 53 7 37.6 2.5 4025 771 8407_B 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 8401_B 1680 51 31.1 0.7 50 11 41.6 1.7 5457 689 8402_B 1744 43 34.5 0.9 33 12 33.7 2.6 3491 693 8407_C 1589 58 31.4 1.3 55 9 40 4 4167 986 8401_C 1509 98 29 1 61 8 39.8 3 4833 946 8402_C 1639 57 32.7 1.5 42 10 32.3 3.3 3282 839 The yield stress is highly dependent upon the material s lamellar thickness, so Yield increases with increasing elastomer density for the A and B series. One interesting note is that alloys containing 8402 ha ve consistently lower elongati on at break, stress at break, and energy to break. These properties are all dependent upon crystal breakup and the ability to strain harden, or recrystallize into a more ordered struct ure. Regarding the PP phase, % crystallinity of alloys containing 8402 are lower than 8407 alloys from Table 67, so the ability of PP to recrystallize ma y be hindered by the mere presence of PE crystals. The melting behavior of four different alloys is shown in Figure 6-11. Alloys with high and low density both w ith and without DEGDA are shown and actual data is given

PAGE 209

187 in Table 6-9. Those containing 8402 show a ch aracteristic melting peak of PE crystals at about 95C. The position and area of this peak does not significantly change with addition of multifunctional monomer, but comp ared to 8402_0, the peak is shifted to lower by about 15C. So, the linear, crystalli zable chains in 8402 are slightly affected by Temperature ( ) 6080100120140160180 Endotherm ( W) -11000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 8407_A 8407_B 8402_A 8402_B C Figure 6-11: DSC melting endotherms of 8407_A, 8407_B, 8402_A, and 8402_B ranging from 60C to 180C. Insert is a magnified graph of the melting peak of PP. the reactive extrusion process, resulting in less ordered, thinner lamellae [389, 393-395, 398]. The melting peak of PP is slightly higher for 8402_B than 8402_A, possibly due to reduced degradation of the longer PP chains. Like in 8402_0, 8402_A has a second melting peak at about 150C attributed to the phase of PP [162, 363, 378]. The crystalline nature of 8402 ma y inhibit the formation of the alpha phase and so the less stable phase forms. Comparing 8407_A and 8402_A, the phase is more prominent for 8407_A. This may be due to the higher gr afting efficiency of 8407, leading to more PS-DEGDA chains to disrupt the crystallization of the phase.

PAGE 210

188 Table 6-9: DSC endothermic da ta of 8407_A, 8407_B, 8402_A, and 8402_B. Sample ID PP phase melting peak temperature (C) Tm Onset (C) Melting Enthalpy (J/g) % Crystallinity 8407_A 164.3 149.5 81.7 39.5 8407_B 164.8 143.9 86.0 41.5 8402_A 164.0 148.7 77.2 37.3 8402_B 164.7 145.7 81.6 39.4 The onset of melting is higher for 8407_A than 8402_A, so the PP phase is degraded more so with 8402 present. 8407 softens early and may encapsulate the monomer/initiator mixture, which promotes grafting onto the copol ymer phase, reduces homopolymerization, and cont rols hydrogen abstraction fr om PP by primary radicals. There are likely a high number of small lame llar crystals in 8407_A whereas for 8402_A the crystallization of PP is inhibited by the crystal nature of the elastomer. Neither 8407_B nor 8402_B show the phase peak of PP, likely due to the reduced degradation of PP and improved crystallization temperat ure (Figure 6-12). Both 8407_B and 8402_B show an increase in the % crystallinity compared to 8407_ A and 8402_A, respectively. Upon studying the crystallization behavior of these alloys (Figure 6-12 and Table 6-10), addition of DEGDA increases the cr ystallization temperature and enthalpy of crystallization of the alloys regardless of ENGAGE crysta llinity. This phenomenon was explained in Chapter 4 and is attributed to an increase in the nucleation density from branching [238, 357, 402]. The peak represen ting the crystallizati on of PE does not Table 6-10: DSC exothermic da ta of 8407_A, 8407_B, 8402_A, and 8402_B. Sample ID Peak temperature of crystallization (C) Enthalpy of crystallization (J/g) 8407_A 124 -80.7 8407_B 130.8 -84.4 8402_A 124.5 -75.7 8402_B 131.4 -80.3

PAGE 211

189 change upon addition of multifunctional monomer. This may be an indication that there is no crosslinking in the elastomer, wh ich would naturally redu ce the % crystallinity and shift the peak to higher temperatures. Temperature ( ) 020406080100120140160180 Exotherm ( W) 2000 4000 6000 8000 10000 12000 14000 8407_A 8407_B 8402_A 8402_B C Figure 6-12: DSC cooling exotherms of PP/el astomer physical blends as a function of elastomer density. Insert is for th e temperature range of 20 to 100C. The viscoelastic behavior of both alloys containing multifunctional monomer and physical blends are compared in Figure 6-13. In each graph, the sample containing DEGDA results in a lower peak intensity for the Tg of polypropylene at about 20C. This means that the % crystallinity is higher, a re sult verified by DSC measurements. Also, a peak becomes more prominent around 80C with addition of DEGDA. The formation of this relaxation is symbolic of the triclinic phase in PP, which has greater toughness and impact strength than the phase [260, 287, 301, 337, 340]. Another intere sting note is that at high temperatures (> 100C), the intensity of the Tan graph is affected to greater

PAGE 212

190 Temperature ( ) -80-4004080120 Tan Delta 0.03 0.06 0.09 0.12 0.15 0.18 8402_0 8402_B C Temperature ( ) -4004080120 Tan Delta 0.05 0.10 0.15 0.20 0.25 8401_0 8401_B C (a) (b) -80-4004080120 0.1 0.2 0.3 0.4 8407_0 8407_B Tan DeltaTemperature ( )C (c) Figure 6-13: Dynamic Mechanical Analys is comparison of (a) 8402_0 and 8402_B, (b) 8401_0 and 8401_B, and (c) 8407_0 and 8407_B. extent with decreasing density of elastomer. This may be related to the grating efficiency of each alloy. At higher elastomer densitie s, the overall grafting efficiency is lower resulting in less molecular mobility at high temperatures. So, 8407 may be dispersed and distributed at a much smaller scale and to a much greater extent than its high density counterparts and so facilitates inter-crystalline motion. 6.3.2 Effect of Elastomer Molecula r Weight on Alloy Performance Four different ethylene-octene copoly mers (002, 8407, 8200, and 8842) are ranked according to their melt flow index and molecu lar weight. All othe r material properties and processing parameters are held constant unless otherwise noted. Also, the ratio of PP

PAGE 213

191 to EOC was kept constant at 90:10. Table 6-4 and Figur e 6-2 give a key code and designated formulations to ease in the identi fication of each sample. The first number is the grade of copolymer, the second number/le tter signifies a formulation which can be referenced in Table 6-4. Figure 6-14 shows the molecular weight aver ages for each of the elastomer grades according to their melt flow index. Polyme rs are mixtures of molecules of various molecular weights and molecular sizes, so molecular weight averages are used to characterize them. With polystyrene standard s, these molecular weights are relative in relation to the hydrodynamic vol ume of the standards. Decreasing melt flow index means increasing average molecula r weight and polydispersity (Mw/Mn). The number average molecular weight (Mn) is simply the total weight of all the polymer molecules in a sample, divided by th e total number of polymer molecules in a sample. This is very sensitive to changes in the weight-fractions of low molecular weight species, which is why its magnitude is less than all other averages. Large molecules have a relatively low effect on this average, so it decreases only slightly with increasing MFI. Higher moments of molecular weight, (Mw or weight average, Mz, and Mz+1) rely more heavily on the number of high molecu lar weight chains. A large Mw often means high tensile strength, Mz and Mz+1 relates to the stiffness and l ong fatigue life of polymers, and Mn is often associated with improving flow properties for processing and low temperature flexibility. At an MFI of 1 or 5, the molecular weight of Mz+1 is much higher than at 30 or 500 g/10min. An interesting not e is that at an MFI of 5, Mz+1 is relatively much higher

PAGE 214

192 Melt Flow Index of Copolymer (g/ 10min) 1530500 Molar Mass (g/mol) 2.0e+4 1.0e+5 1.8e+5 2.6e+5 3.4e+5 4.2e+5 5.0e+5 5.8e+5 6.6e+5 7.4e+5 M n M w M z M z+1 Figure 6-14: Molecular weight averages of the polyolefin elastomers of interest, ranked according to their melt flow index. than Mz, which may favor phase separation of chai ns from entropy considerations. This has implications on mechanical properties whic h will be discussed later in the chapter. The critical entanglement molecular weight (Mc) for polyethylene is 4000 g/mol [185, 417], which is directly relate d to viscoelasticity and chai n mobility and dictates the polymers viscosity and ultimate mechanical pr operties. Each grade of copolymer has a Mn above this value, which means that enough p hysical entanglements are present to give some resistance to deformation. The closer the polydispersity approaches a value of one, the narrower is the molecular weight distribution. The low MF I polymers (Figure 6-15) have a relatively high PDI, which means that the distribu tion is skewed towards larger chains.

PAGE 215

193 ENGAGE Grade 002840782008842 Polydispersity Index 1.80 1.84 1.88 1.92 1.96 2.00 2.04 Figure 6-15: Polydispersity (molecular weight distribution) of the polyolefin elastomers of interest. 6.3.2.1 Physical blends The notched Izod impact strength and stre ss-strain behavior ar e highly dependent upon intercrystalline thickness and molecular wei ght of the constituents [270, 409, 410]. From Figure 6-16, impact strength is highest for th e high molecular weight 8842_0 sample, followed by 8407_0, 8200_0, and 002_0, respectively. The highest impact strength corresponds to the lowest elastomer MFI, so there is a highly entangled network of mobile chains within the amorphous regions in PP which are able to relax, diffuse, and respond at the speed of the impact test. 8200_0 has a lower impact strength than 8407_0, which can be traced back to the molecula r weight and density of 8200. 8200 has a slightly lower comonomer content than 8407, and a disproportionate amount of high molecular weight chains (Mz+1). This means that 8200 is more dense than 8407, the high molecular weight chains in 8200 give stiffne ss to the blend, and there may also be a greater degree of phase separati on within the blend. The lowe st impact strength is seen for the blend containing the lowest molecular we ight elastomer. This is likely due to the

PAGE 216

194 fact that 002 has a high number of chain e nds along with a relativ ely low bulk cohesive strength and so cannot transfer stress effectively. There may also be coalescence of 002 because of its extremely low viscosity, which results in a low degree of interfacial surface area and adhesion [12, 69]. Another plausible theory is that at high deformation rates, rubber elasticity theory can be applied to th e elastomer phase [15, 161]. So, by stretching out the chains, a greater retr active force is needed for further deformation and the temperature rise experienced in an impact test promotes a conformational change from extended to random coil. Copolymer MFI 1530500 Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/min) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Impact Strength MFI Figure 6-16: MFI and room temperature notch ed Izod impact strength of 8842_0 (highest molecular weight), 8200_0, 8407_0, and 002_0 (lowest molecular weight). As expected, the melt flow index of the blends rise as the MFI of the elastomer phase increases. Notice that a proportionate ly large jump is seen from 8407_0 to 002_0, indicative of the 17X jump in elastomer MFI. This can be related to the high number of low molecular weight chains in 002_0. 8407 is essentially made up of chains that are

PAGE 217

195 double in length than that of 002_0, which give s it the cohesive stre ngth and elongational viscosity needed to prevent coal escence and resist deformation. The stress-strain deformation of the elastomers at low strains is related to the sliplink theory of rubber elasticity, which states th at entanglements are visualized as slipping links that can slide along the chain contour between crosslinks [58]. At intermediate and high strains, crosslinks are considered to be entangled knots that tighten as strain increases. By decreasing molecular weight, th ere is an increase in the number of chain ends, which reduces the number of effective network chains. At high MW, deformation is homogeneous a nd is governed primarily by the strainhardening process for stress-strain behavior. A ductile deformation requires that there be an adequate number of seque nces of disordered chain uni ts connecting crystallites. Furthermore, the number of units involved must be large so that each connector is deformable and can sustain large deformations So in order to avoid embrittlement, the interlamellar region must possess chains that display rubber-like behavior. They cannot be straight chains in planar zigzag confor mation. The effective number of deformable sequences will be tempered by chain entanglements and interlinking. These factors are sensitive to MW, % crystallin ity, and interlamellar thickness. For low MW materials, there are not enough disordered sequences c onnecting the crystal lite to transmit the tensile force. And with a high crystallinity le vel, the thickness of the interlamellar region is small. Hence, the disordered connecti ng units would not be ab le to sustain large deformations [269]. A few trends are obvious from Figure 6-17 regarding tensile deformation of the physical blends. All stress-strain properties (yield strength, elongati on at break, stress at

PAGE 218

196 break, elastic modulus, and energy to break) d ecrease in a somewhat linear fashion with decreasing molecular weight to a point. If elastomer particles are trapped within a growing PP spherulite, it will change its gr owth and final morphology, which in-turn will affect these properties. A low elastomer MW will result in a reduction in the number and degree of entanglements in the amorphous phase (tie molecule population), thereby decreasing boundary strength. So the elastic response to an applied load is reduced (decrease in elastic modulus) and the mobility given to lamellar slippage, rotation, and Melt Flow Index of Copolymer 1530500 Yield Stress (MPa), Elongation at Break (mm), Stress at Break (MPa) 20 30 40 50 60 70 80 Elastic Modulus (MPa) 1440 1460 1480 1500 1520 1540 1560 1580 1600 E nergy t o B rea k (N/ mm ) 3500 4000 4500 5000 5500 6000 6500 Yield Stress Elongation at Break Stress at Break Elastic Modulus Energy to Break Figure 6-17: Stress-strain performance of phys ical blends as a function of MFI of the copolymer. orientation is increased due to the greater free volume afforded by more chain ends (decrease in yield stress). Th e short chain nature of 8407 re duces the extent of strain hardening, so stress at break, elongation at br eak, and energy to break all decrease. With increasing MW, the plateau region beyond yiel d decreases, and the slope of strain hardening becomes steeper [269]. However, these trends are reversed when 002_0 and 8407_0 are compared. In fact, the elastic m odulus of 002_0 is comparable to that of

PAGE 219

197 8200_0. This brings up an interesting point at what molecular weight or melt flow index will the elastomer phase be completely rejected from a growing PP spherulite? The spherulitic morphology of these blends need to be studied in detail to further understand the distribution and perfection of the PP spherul ites. Also the extent of physical aging (densification of non-crystalline material below Tg) may give an indication of why this occurs. It should be noted that standard devi ations and typical stre ss strain graphs are given in Appendix B. The melting behavior of the physical ble nds is shown in Figure 6-18 and Table 611. The results indicate that as the molecula r weight of the elastomer increases, the onset of melting temperature and en thalpy of fusion also increa se. The reason that 8842_0 has a higher % crystallinity than 002_0 may be traced to the blends crystallization behavior and will be explained over the next few para graphs. The peak at approximately 150C is related to the melting of the phase of PP and is known to be nucleated by elastomers. Table 6-11: DSC endothermic da ta of 002_0, 8407_0, 8200_0, and 8842_0 Sample ID PP phase melting peak temperature (C) Tm Onset (C) Melting Enthalpy (J/g) % Crystallinity (from Refs. 170, 391) 002_0 165 152.5 82.3 39.8 8407_0 165.2 152.9 82.3 39.8 8200_0 165 152.6 85.2 41.2 8842_0 164.5 153 88.9 43 The early stage of crystallizat ion may be based on Florys theory of crystallization of copolymers [161, 392]. He visualizes crystals linked rather quickly by amorphous defects and set up a metastable, global network of interlinke d crystals. This type of gelation is reached at 5% of crystallization. After the cr ystal network is set up, the amorphous defects can continue locally to crys tallize with very little or no long distance diffusion in a secondary crystallization pro cess [61-63]. Crystallization occurs by a

PAGE 220

198 straightening out of short coil sequences without a long rang e diffusion process [100]. A larger average molecular weight of elastomer leads to a higher crystallization temperature and enthalpy of crystallization. Long se quences of a given polymer repeat unit are known to crystallize at a higher temperatures than shorter sequences [271]. But a recent study found that a higher MFI elastomer reduces the spherulite growth rate, showing that the diffusion process dominates ove r crystallization rate [290]. Temperature ( ) 130140150160170 Endotherm ( W) -12000 -11000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 002_0 8407_0 8200_0 8842_0 C Figure 6-18: DSC melting endotherms of 002_0, 8407_0, 8200_0 and 8842_0. Figure 6-19 and Table 6-12 show the crystall ization behavior of these blends. The long copolymer chains and high degree of entanglements in 8842 prevent long range diffusion of polypropylene chains (melt memo ry effects) and therefore facilitate crystallization [100, 399]. When polypropylen e segments are in close proximity to each other, they are more likely to nucleate. A higher crystallization temperature signifies a higher packing density, faster crystallization ki netics, and thus greater % crystallinity for 8842_0 than the lower molecular weight copolym ers. Also, the lower molecular weight

PAGE 221

199 Table 6-12: DSC exothermic data of 002_0, 8407_0, 8200_0, and 8842_0 Sample ID Peak temperature of crystallization (C) Enthalpy of crystallization (J/g) 002_0 114 -78.9 8407_0 115.3 -83.1 8200_0 116.8 -83.5 8842_0 117.7 -84.3 Temperature ( ) 90100110120130140 Exotherm ( W) 2000 4000 6000 8000 10000 12000 14000 16000 002_0 8407_0 8200_0 8842_0 C Figure 6-19: DSC crystallization ex otherms of 002_0, 8407_0, 8200_0, and 8842_0. material is likely rejected into the interspherulitic regions of PP as deformed domains, resulting in a considerable depression in growth rate [100, 236, 255, 257, 260, 263, 297, 307, 311-313]. The density of 8842 is less than all the other copolymers, so this may not inhibit the crystallization processes of PP. All of these factors point to the fact that a high molecular weight elastomer result in thicker, more perfect PP crystalline lamellae with high melting temperature. This, along with a high degree of entanglements is what gives 8842_0 a balance of both impact strength and stress-strain properties.

PAGE 222

200 6.3.2.2 Alloys Reactively extruding the physical blends wh ile changing elastomer MW allows one to analyze several aspects of the process, namely: Whether the liquid reactants become tr apped in the elastomer phase (higher viscosity = higher MW = lower MFI) and th us limited in dispersion within the PP phase, Whether an increase in the entangleme nt density means a greater degree of grafting, Whether the impact strength can be optimized, And whether chain scission of PP can be reduced. From Figure 6-20, addition of styrene and ini tiator to the blends results in a slight decrease in impact strength for all sample s but 8200_A, when compared to their physical blends. The decrease in impact strength is likely due to the degradation of the PP phase and the increase may be due to a change in the percent crystallinity of 8200 (grafting has been shown to decrease the % crystallinity of polyolefins [7 148, 184, 215]). Degrading PP chains is evidenced by higher MFI values. It appears that 002_A has the highest MFI and is affected the most by the modification with styrene and initiat or. 002_0 may not be trapping liquid reactants as the higher molecula r weight elastomers, so PP is experiencing a higher concentration of primary free ra dicals. During morphology development, 002 may not coat PP pellets or be stretched out like a higher MW copolymer because of its low number of entanglements and relatively low softening temperature [13]. The sheet formation during morphology development is cruc ial to proper dispersion [32, 83] and after phase inversion takes place, the low vi scosity of 002 likely does not trap any of the PP phase via grafting or crosslinking.

PAGE 223

201 Sample ID 8842_A8200_A8407_A002_A Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Impact Strength MFI Figure 6-20: Impact Strength and me lt flow index of 8842_A, 8200_A, 8407_A, and 002_A. Addition of DEGDA and reduction in styrene and initiator concentrations results in a significant increase in impact strength for all alloys but 002_C from Figure 6-21. Once again, 002 likely does not participate in the gr afting process because of its low number of physical entanglements, low viscosity, high num ber of secondary and tertiary hydrogen atoms, and tendency to coalesce from such a drastic difference in viscosity between 002 and PP [420]. The melt flow index of 002_C is 66% lower than 002_A which is significantly greater than all other alloy comb inations. The PP phase is preferentially degraded in 002_0, so addition of a radical trapping agent and reduction in the overall

PAGE 224

202 Sample ID 8842_C8200_C8407_C002_C Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 Grafting Efficiency 10 15 20 25 30 35 40 45 50 Impact Strength MFI Grafting Efficiency Figure 6-21: Izod impact strength and melt flow index of 8842_C, 8200_C, 8407_C, and 002_C. concentration of free radicals s hould give this result. An in teresting trend is the steady increase in grafting efficiency with reduc tion in elastomer molecular weight from 8842_C to 8407_C. Since styrene grafting prim arily occurs on PP rather than the elastomer phase, 8407 is most efficient at dispersing the liquid reactants throughout the system. This grade of ENGAGE appears to have the optimal comb ination of molecular weight, viscosity, and melting or softening te mperature at this styrene concentration. 8842_C has a lower grafting efficiency possi bly due to the cage effect that long, entangled chains may have on the free radical polymerization proce ss. At the lowest elastomer molecular weight, grafting efficiency drops to a very low value. This may be because monomer is lost to homopolymerizat ion by touching extrude r barrel wall (no encapsulation of liquids) and gr oss phase separation from PP.

PAGE 225

203 Alloys having the highest impact strengt hs are seen in Figure 6-22. These materials have the highest concentration of initiator, styrene, and multifunctional monomer which reinforces that fact that th e right combination of initiator and monomers will lead to outstanding alloy performance. A few features in this graph must be noted and elaborated upon. The grafting efficiency remains roughly constant when elastomer MFI stays between 1 and 30 g/10min. The initia tor concentration is likely so high that there are enough free radicals facilitate high grafting yi elds and the high styrene concentration, acting as a vector fluid, encompasses all phases and particles as the morphology evolves in the melting portion of the extruder. The overall degree of Sample ID 8842_B8200_B8407_B002_B Notched Impact Strength (ft-lbs/in) and Melt Flow Index (g/10min) 0 1 2 3 4 5 6 7 8 9 10 11 Grafting Efficiency 30 40 50 60 70 80 90 Impact Strength MFI Grafting Efficiency Figure 6-22: Izod impact strength and melt flow index of 8842_B, 8200_B, 8407_B, and 002_B. entanglements is sufficient even at the 8407 level to improve dispersion of the liquid reactants. The fact that 8407 and 8842 have si milar grafting efficien cies but a different

PAGE 226

204 molecular weight leads one to believe th at at such high monomer and initiator concentrations, the higher moments of molecu lar weight may not pl ay a significant role. In the presence of initiator, the highest molecular weight fractions of polyethylene will be consumed first and a higher amount of crosslinking agent shifts the distribution towards lower molecular weights [149, 185]. The low MW fraction is believed to act as grafted pendant chains in th e presence of longer polymer chains. The network formed within the low MW fraction consists mainly of chemical crosslinks, whereas high MW material comprises both physical entanglemen ts and chemical crosslinks. Trapped entanglements generate the major part of th e styrene grafting poin ts at low peroxide concentrations. At high number average mo lecular weight values, dense networks are easily created with only a small mount of ch emical crosslinks and branches, as the probability of entanglement formation is very high. Impact strength is highest for 8407_B, whic h may be due to the high dispersability of the 8407 phase compared to its high molecular weight counterparts. The smaller elastomeric domains allow for more efficient energy absorption mechanisms. Melt flow index, to no surprise, steadily increases with increasing elastomer MFI. MFI is roughly additive for the alloys, sim ilar to physical blend results. An interesting trend is shown in Table 613 regarding the elastic modulus and yield stress of the high impact specimens. The lo wer the molecular weight of the elastomer, the higher the modulus of the alloy. Th e modulus is directly dependent upon the amorphous phase of the alloy, which is likely highly branched from addition of multifunctional monomer. At both high and low deformation rates, 002_B responds in a brittle manner. The PP phase is affected more drastically with the low molecular weight

PAGE 227

205 elastomer present, having a highly bran ched amorphous phase and many reinforced cross-hatched lamellar crystals. For grafti ng of high molecular weight polyethylenes (like 8842), grafting reactions are known to result in little, if any, crosslinked material. So, for higher viscosity elasto mers, dispersion and distributi on may be improved so as to Table 6-13: Stress-strain beha vior of 002_B, 8407_B, 8200_B, and 8842_B. Effect of Copolymer Molecular weight Blend ID Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 002_B 1782 109 31.8 1.8 46 10 37.6 4 4048 881 8407_B 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 8200_B 1598 54 30.2 0.6 62 7 43.2 1.5 5699 485 8842_B 1545 40 29.3 1.1 56 6 42.7 2.5 4945 710 increase chain mobility in the PP phase. Elonga tion at break, stress at break, and energy to break are all lowest for 002_B a relatively brittle alloy. Dynamic mechanical behavior has already been shown to be a vital technique in establishing relaxation behavior of polymer chai ns in both blends and alloys. It gives a wealth of knowledge regarding miscibility, crys tallinity and/or crosslinking, degradation, branching, and intercrystalline mobility. Figure 6-23(a) thru (d) compares physical blends and alloys for all grades of elastomer. The storage modulus is seen to be higher for 002_B than 002_0 but this trend is opposite fo r all other samples. From stress-strain data, the elastic modulus and yield stress are also extraordinarily high for 002_B. In this alloy, PP may retain its spherulite structure and with a slight degr ee of crosslinks, the cross-hatched structure may be reinforced. At temperatures above 20C, the rubbery plateau regions of 8407_B and 8200_B decrease in a steeper manner when compared to their physical blends, but for 8842_B the graph pl ateaus to a slightly greater extent. The length of the rubbery plateau region is known to increase with molecular weight [81], so

PAGE 228

206 Temperature ( ) -80-60-40-20020406080100120 E' (MPa) 5e+7 6e+8 1e+9 2e+9 2e+9 3e+9 3e+9 4e+9 4e+9 5e+9 5e+9 6e+9 Tan Delta 0.05 0.10 0.15 0.20 0.25 0.30 0.35 002_0 002_B C Temperature (oC) -80-60-40-20020406080100120 E' (MPa) 5e+7 6e+8 1e+9 2e+9 2e+9 3e+9 3e+9 4e+9 4e+9 5e+9 5e+9 6e+9 Tan Delta 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 8407_0 8407_B (a) (b) Temperature ( ) -100-80-60-40-20020406080100120 E' (MPa) 5e+7 6e+8 1e+9 2e+9 2e+9 3e+9 3e+9 4e+9 4e+9 5e+9 5e+9 6e+9 Tan Delta 0.1 0.2 0.3 0.4 0.5 8842_0 8842_C C Tem p erature ( ) -100-80-60-40-20020406080100120 E' (MPa) 5e+7 6e+8 1e+9 2e+9 2e+9 3e+9 3e+9 4e+9 4e+9 Tan Delta 0.05 0.10 0.15 0.20 0.25 8200_0 8200_B C (c) (d) Figure 6-23: Dynamic mechanic al behavior (E' and Tan ) of various alloys and blends. (a) 002_0 vs. 002_B, (b) 8407_0 vs. 8407_B, (c) 8200_0 vs. 8200_B, and (d) 8842_0 vs. 8842_B. the high degree of physical entanglements in 8842 likely promotes rubber-like behavior of the alloy. For Tan comparisons, the Tg of the elastomer phase is slightly lower for all alloys than blends except in the case of 8200_0. This material has a slightly higher % crystallinity than all other elastomers, so the crystalline phase may be disturbed by the grafting of monomer. The intensity of polypropylenes Tg increases with increasing elastomer MW. This is surprising because one may expect that elastomers with more chain ends and more mobility would have th e greatest effect on the intensity of PPs Tg. But these low molecular weight species may be rejected more easily from the interand intra-spherulitic regions in PP. The Tg of 8842_B is highest probably because the long

PAGE 229

207 polymer chains of 8842 are grafted onto PP over a larger volumetric space. In 8407_B there may be smaller domains but a greater num ber of domains. At higher temperatures (approaching and eclipsing the relaxation), 8200_B and 8407_B have greater intensities than 8842_B and 002_B when compared to their physical blends. This poses an interesting question about the morphology of these alloys it appears that at extreme values of elastomer molecular weight, domain size may be too large. So, the lamellar crystals of PP do not slide past one anothe r readily because they are not surrounded by elastomer particles. Also notice the drasti c decrease in intens ity from 8842_0 to 8842_B at high temperatures. A propor tionately high number of entanglements and crosslinks are present which reduce intercrystalline mobility so the elastic nature of the material increases. The relaxation behavior, m echanical properties, and rheological properties are known, but the crystalline state of these alloys has to be defined. Figure 6-24 and Table 6-14 show that the onset of melting is lower and % crystallinity of PP higher for alloys containing DEGDA than those w ithout. This is true regardless of the molecular weight of elastomer and it is significant because it re inforces the idea that the method of radical attack, polymerization, branchi ng, grafting, and PP chain scission are the same as long as the elastomer viscosity remains within a certain range. As stated in the previous sections and chapters, the lamellar cr ystals in 8407_B and 8842_B ar e smaller and greater in number than 8407_0 and 8842_0, respectively. A lower onset of melting temperature is linked to the greater surface energy of the sma ller crystals and the higher % crystallinity is associated with a branched structure c ontaining short crystallizable PP sequences which are in close proximity to eachot her [255, 279, 327, 362, 369, 389, 391-397].

PAGE 230

208 Temperature ( ) 120130140150160170180 Endotherm ( W) -11000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 8407_A 8407_B 8842_A 8842_B C Figure 6-24: DSC melting endothe rm of 8407_A, 8407_B, 8842_A, and 8842_B. The phase of PP at 150C is more pronounced for 8407_A than 8842_A, which may be due to the lower degree of PP chain scission experienced by 8842_A. The presence of this phase has been attributed to degraded PP chain segments and a high degree of homopolymerized monomer that hinder the formation of the traditional phase. The increase in % crystallinity from 8407_A to 8407_B is more drastic than 8842_A to 8842_B. This may be due to the hi gh MFI of 8407_A (more degraded chains) which are then linked up by multifunctional monomer (8407_B). 8842 may trap some of the initiator and slightly cro sslink or extensively branch. Table 6-14: DSC endothermic da ta of 8407_A, 8407_B, 8842_A, and 8842_B. Sample ID PP phase melting peak temperature (C) Tm Onset (C) Melting Enthalpy (J/g) % Crystallinity 8407_A 164.3 149.5 81.7 39.5 8407_B 164.8 143.9 86 41.5 8842_A 163.5 149.7 80.8 39 8842_B 164.9 144.6 85.4 41.3

PAGE 231

209 The crystallization behavior of the alloys is shown in Figure 6-25 and Table 6-15, both including 8407_A, 8407_B, 8842_A, and 8842_B. With addition of multifunctional monomer, Tc and enthalpy of crystalliz ation both increase, regardless of elastomer MFI. So, dispersion of liquid reactant s is adequate, grafting is wi despread, and a network-like morphology is likely present. The increase in Tc from 8407_A to 8407_B is slightly higher than from 8842_A to 8842_B. This may be due to the greater dispersability of 8407 and therefore better distribution of grafts and branch points. Temperature ( ) 100110120130140150 Exotherm ( W) 2000 4000 6000 8000 10000 12000 14000 16000 8407_A 8407_B 8842_A 8842_B C Figure 6-25: DSC crystallization ex otherm of 8407_A, 8407_B, 8842_A, and 8842_B. Table 6-15: DSC exothermic data of 8407_A, 8407_B, and 8842_A, and 8842_B Sample ID Peak temperature of crystallization (C) Enthalpy of crystallization (J/g) 8407_A 124 -80.7 8407_B 130.8 -84.4 8842_A 124.1 -78.6 8842_B 130.3 -83.6

PAGE 232

210 6.3.3 Supercritical CO2 as a Possible Route to Improve Grafting Efficiency A new route to improve the processability of thermoplastics and their blends is to utilize supercritical carbon dioxide (scCO2) to plasticize the system and reduce interfacial tension between phases [421-426]. Many authors have also found scCO2 useful in free radical polymerization of mono mers as well as grafting of polyolefins [425-433]. This high pressure additive has many other uses including extraction and infusion of low molecular weight materials, foaming agent, plasticizer, etc. [425, 426, 434-436]. Carbon dioxide, above its cr itical point of 1037 psi and 31.7C, has the diffusivity of a gas and viscosity similar to a liquid [425, 426]. These properties, along with its safe storage and handling and low cost make CO2 an interesting candidate to improve the processing of polymers. scCO2 has been shown to enhance free radical reactions in the polymerization of vinyl monomers but also to improve the mixing of immiscible polymers. Some authors have found that so lid state and melt free radical grafting of monomers onto polypropylene is facilitate d by the increased diffusivity and high pressures from scCO2. The purpose of the following experiment was to identify whether this additive can improve the grafting efficien cy of reactive extrusion, which correlates positively with mechanical properties. Figure 6-26 reveals that when scCO2 is injected into zone 5 of the extruder, grafting efficiency decreases in a linear fashion. Th is result is not necessarily surprising because scCO2 is a solvent for low molecular weight orga nic liquids, such as styrene. Also, the temperature of the injected fluid is much lower than the normal reaction conditions, which could adversely affect efficiency b ecause the peroxide decomposition rate and polymerization rate is slowed. Many of th e previous studies used batch mixers and experiments that lasted hours to study the graf ting efficiency, but the residence time is so

PAGE 233

211 short in the extruder that this may prevent scCO2 from promoting grafting. Several suggestions for improving the grafting yield using scCO2 are given in Chapter 8 under future recommendations. 30 40 50 60 70 80 012345678scCO2 Feed Rate (mL/min)Grafting Efficienc y Figure 6-26: Effect of supe rcritical carbon dioxid e at 1500 psi on the grafting efficiency of 8842_B. Injection was in zone 3. 6.4 Conclusions This chapter exemplifies the idea that material selection of the elastomer is crucial to balancing blend and alloy mechanical prope rties while maintaining processability. The crystalline content and chain le ngth of the elastomer plays a large role in dictating alloy and blend mechanical properties. Some obvious trends are seen in blends and alloys: as elastomer density increases, impact strengt h decreases but elastic modulus and yield strength increase. The grafting efficiency is drastically lower for 8402_B than 8407_B, which can be traced back to fact that 8402 ha s a higher % crystallinity and Tm so it melts at a later point along the extr uder barrel. This leaves mo re time for homopolymerization to take place. If one were to maximize both impact strength and el astic modulus of PP, a combination of 8402 and 8407 may give the desired result.

PAGE 234

212 Elastomer MFI (molecular weight) also has a profound effect on mechanical properties. A high MFI is pref erential for an optimal balance of properties because below a molecular weight limit, the blend/alloy be haves in a brittle manner. This can be directly related to the higher degree of physical entanglements, higher viscosity, and lower number of chain ends in the low MFI elastomer. The processability of 8842_B may be compromised as evidenced by its low melt flow index value of 0.4 g/10min, which is why a lower molecular weight el astomer may be preferred (i.e. 8407).

PAGE 235

213 CHAPTER 7 REACTIVE EXTRUSION OF HIGH DENSITY POLYETHYLENE 7.1 Introduction Impact modification of polypropylene was very successful, so the reactive extrusion process was applied to another commodity plastic high density polyethylene (HDPE). Polyethylene is one of the larges t volume polymers produced in the last decade [437]. It is expected to have an average a nnual growth rate of 5.2% in terms of demandproduction during the decade spanning 2000-2010. The objective of this chapter is to design a set of experiments so as to identif y the most crucial asp ects of modifying HDPE and to improve its impact strength while balancing stress-strai n properties. HDPE normally crosslinks in the pr esence of peroxide, but the reactive extrusion process described in previous chapte rs has shown that an insol uble gel does not form for PP alloys. 7.2 Experimental 7.2.1 Materials Table 7-1 gives a list of ethylene-1-octene copolymers and manufacturers data. The impact modifier of interest wa s ENGAGE 8842 [48]. HDPE was supplied by ExxonMobil as the grade Paxon AM 55-003. A ll polymers were received and used in pellet form. Table 7-2 lists structures of the reactive materials. The peroxide and monomers used in this study were reagent grad e chemicals. The monomers were purified by passing through an activated alumina column before use. Styrene monomer, inhibited by 10-15 ppm t-butyl catechol, wa s purchased from Fisher.

PAGE 236

214 Table 7-1: ENGAGE product data table. Engage Grade (decreasing comonomer content) 8842 8407 8200 8401 8402 Comonomer Content wt% 13C NMR/FTIR 45 40 38 31 22 Density, g/cm3 ASTM D-792 0.857 0.870 0.87 0.885 0.902 Melt Index, dg/min ASTM D1238 190oC, 2.16 kg 1.0 30 5.0 30 30 Mooney Viscosity ASTM D1646 ML 1 + 4 at 121oC 26 < 5 8 < 5 < 5 Durometer Hardness, Shore A ASTM D-2240 50 72 75 85 94 DSC melting Peak, oC Rate: 10oC/min 33 60 60 78 98 Glass Transition Temp, oC DSC inflection point -61 -57 -56 -51 -44 Flexural Modulus, MPA ASTM D-790, 2% Secant 3.5 12.1 12.1 25.8 69.9 Ultimate Tensile Strength, MPa ASTM D-638, 508 mm/min 2.1 3.3 6.9 6.4 12.9 Ultimate Elongation, % ASTM D-638, 508 mm/min 975 >1000 >1000 950 790 Table 7-2: Structures of reactive materials of interest Name Lupersol 101 DEGDA Structure Name Styrene Structure The initiator, 2,5dimethyl-2,5 -di-(t-butylperoxy) hexane, was purchased from Atofina under the trade name Lupersol 101. Diethyl eneglycol diacrylate (DEGDA), inhibited by 80ppm Hq and 120 ppm MEHQ, was gracious ly donated by Sartomer, an Atofina company.

PAGE 237

215 7.2.2 Methods 7.2.2.1 Processing All polymers were dried in an air circ ulating oven at 40C for 24 hours prior to compounding. Before processing, the resi ns were premixed by hand for about 10 minutes. The monomers were passed trough activated alumina columns to remove residual inhibitor. Monomer/initiator mixtures are magnetically stirred for 10 minutes before processing and a choice amount of th e mixture was added to the dry polymer pellets before processing. The blending was carried out in a 34 mm non-intermeshing, co-rotating twin screw extruder, APV Chemical Machinery (now B&P Process Systems) with an L/D ratio of 39. The temperature of the extruder was re gulated by electrical resistance and water circulation in the barrels. The dried, premixed resins were then introduced into the Die Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Zone 7 (Feed) 195C 205C 210C 210C 200C 190C 180C 165C Figure 7-1: Schematic drawing of the r eactive twin screw extruder and a common temperature profile. extruder at 60 g/min through a screw dry material feeder, Accu Rate, In c. A Zenith pump controlled the rate of monomer/initiator so lution addition into the extruder. The screw

PAGE 238

216 speed, unless otherwise noted was 150 rpm. Devolatili zation was carried out by a vacuum pump, VPS-10A, Brooks Equipment Co mpany. This was placed near the die and created a pressure of a bout 15 in Hg. The extruder was always starved to feed. Figure 7-1 is a schematic of the extruder, w ith a typical temperature profile. After compounding, the resulting strands which exit the die are quenched in a water bath, pelletized, and dried in a vacuum oven at 100C, 28 in Hg for 24 hours. 7.2.2.2 Mechanical properties In order to measure the strength of the mate rial at very high test ing rates, a notched Izod impact test was performed according to ASTM D256 standards. The pellets were placed in a mold with 6 slot s, each measuring 0.5x0.5x2.5 in3. The mold was put in a Carver press (Fred S. Carver, Inc.) at 200C and after the material melts, pressed up to 5000 psi. After waiting for 5-10 minutes, th e pressure was slowly increased up to 10,000 psi. After another 5 minutes, the heat was turned off and the sample is let to cool down to room temperature at about 1.5C/min. Th e bars were then take n out and notched with a Testing Machines, Inc. (TMI) notching machine. Before te sting, they were conditioned at room temperature for 24 hours. A 30 ft-lb hammer was used with test method A on a TMI Izod impact tester. At least 5 bars were broken and impact strength is recorded as an average regardless of full or partial break. For stress-strain measurements, dried pe llets were placed in a mold measuring 15x15 cm2x1 mm thick. The mold is put into the Carver press at 200C and after the material melts, pressed up to 5000 psi. Afte r a 5-10 minute wait, the sample was slowly pressed to 10,000 psi. Five minutes later, the sample was quenched in a water bath. Specimens were tested according to ASTM D6 38 standards. Type V specimens were punched out of the compression molded sheet with a die, measuring 1 mm thick, 2.95

PAGE 239

217 mm gauge width and 9.5 mm gauge length. Five samples were tested after conditioning at room temperature for 48 hours. The mach ine used to test the samples was an MTS Model 1120 Instron, using a 1000 lb load cell at a test speed of 12.7 mm/min. 7.2.2.3 Design of experiments In a characterization experiment, we are usually interested in determining which process variables affect the response [438, 439] Experiments done in Chapters 3 thru 6 have shown screw speed, temperature, and concentrations of additives (initiator, multifunctional monomer, styrene, and elastomer) have some effect on the mechanical properties of PP. These same process variable s will be used to characterize the response of HDPE. Table 7-3 lists th ese factors and their correspond ing codes and levels. The next logical step is to optimize create a response surface using a factorial or central composite design. A fractional factorial desi gn allows an experimenter to screen out factors that do not significantly affect the response variables by limiting the total Table 7-3: Factors of interest coded variables, and levels for the impact modification of HDPE Variable Name -1 Level +1 Level A % Initiator 0.3 1 B % Styrene 4 8 C PE:ENG8842 95:5 85:15 D Screw Speed 100 200 E % DEGDA 0.5 1 F Temperature 190 210 number of experiments. The design of intere st uses 6 factors at two levels each, for a total of 16 experiments. Tables 7-3 and 7-4 describe the experiment al design of interest, including factors and levels. DesignExpert 6.0.11 from Stat-Ease, Inc. was used to generate the design and analyze the data.

PAGE 240

218 Table 7-4: Basic fractional factorial desi gn using coded variables from Table 7-3. Fractional Factorial Design 26-2 with 16 Runs Total Generated Variables = Temperature and % DEGDA Basic Design Run A B C D E = ABC F = BCD 1 2 + + 3 + + + 4 + + + 5 + + + 6 + + + 7 + + 8 + + + + 9 + + 10 + + + + 11 + + + 12 + + + 13 + + + 14 + + + 15 + + + + 16 + + + + + + 7.3 Results and Discussion 7.3.1 Mechanical Properties Table 7-5 includes all materi als, levels, and resultant impact strength and stressstrain properties. The highest modulus, yield stress, and elongation at break are seen for pure HDPE, as expected, and the pure polymer al so has the lowest impact strength. The good tensile performance is due to the high % crystallinity typically found in HDPE. The modulus and yield stress are not as high as those found for pure PP in Chapter 3, even though PP typically has lower % cr ystallinity than HDPE. HDPE is composed of radially oriented lamellar crystals which take shape in the spherulite, whereas PP is composed of cross-hatched spherulitic stru ctures which are much stiffer [337, 370, 371, 375, 388]. The impact strength of pure HDPE is 2.2 ft-lbs /in, which is over two times greater than pure PP at 0.99 ft-lbs/in. This could be due to the less stiff crystalline state of HDPE, but

PAGE 241

219 a more plausible explanation is that HDPE has low temperature relaxations (Tg is reported to be around -130C) in which chains have the mobility to respond to high rate impact tests [100, 105, 107, 109, 304]. The Tg of PP has been reported to be about 10C. HDPE typically shows drastic increase in impact strength alongside a decrease in modulus with addition of a rubbery phase [125] This trend is seen when blending HDPE with 8842 at both the 85:15 and 95:5 levels. The ENGAGE grade 8842 was chosen over 8407 because in a physical blend with HDPE it improved HDPEs impact strength to a greater extent. This may be attributed to better mixing of 8842 with HDPE and the existence of entangled high molar mass chains Also, the density of 8842 is less than 8407. There are many notable features in Table 73. Sample 16 has an impact strength of 16 ft-lbs/in, the highest value of all experiments in the Table and for all PP alloys. But its stress strain performance is poor compared to pure HDPE. The HDPE was originally thought to crosslink and form an entangled network with 8842, th erefore reaping the benefits of high impact strength from the 8842 phase and high tensile strength from the HDPE phase via stress transfer mechanism. Initiator level does have to be above a certain level before any appreciable crossli nking can occur [7, 184, 187] and addition of monomer reduces the tendency to crosslink. Al so, the grafting efficiency is believed to be lower for the HDPE-8842 system than PP8407 system from bond dissociation energy considerations and the fact that HDPE is highly crystalline so it will have to be completely molten before any grafting will take place. For run 16, all variables are at

PAGE 242

220Table 7-5: Impact strength, el astic modulus, yield stress, and elongation at br eak results from the fractional factorial design created for the modification of HDPE. Run wt% Initiator wt% StyreneENG8842 Screw Speed (rpm) wt% DEGDA Max Temp ( C) Impact Strength (ft-lbs/in) Std. Dev. Elastic Modulus (MPa)Std. Dev. Stress at Yield (MPa)Std. Dev. Elongation at break (mm)Std. Dev. 10.3451000.51903.00.25127910124.31.2528 214510011908.30.809086621.10.8255 30.38510012101.50.25152911326.60.3318 41851000.52103.10.2011339823.20.6173 50.341510012107.60.308093319.70.86110 614151000.521014.80.406344718.60.53314 70.38151000.51907.00.308646820.00.6498 81815100119014.70.906933518.20.3295 90.3452000.52102.80.2011997624.51.54715 1014520012106.30.3510337822.71.1312 110.38520011902.60.10130312025.40.95011 121852000.51904.40.4010004822.40.8315 130.341520011907.70.3083212020.01.3607 1414152000.519015.00.606462417.02.0266 150.38152000.52107.70.408457119.80.56312 161815200121016.00.606904118.21.0483 Pure PE---100-1902.20.10 1571 105 29.2 1.0 56 14 100-1903.30.1013568025.50.78911 Physical Blend PE:8842 = (95:5)

PAGE 243

221 their high level, so a high amount of elasto mer in conjunction with a high concentration of free radicals and monomer is beneficial to impact strength This leads one to believe that the grafting of monomer is highest for th is alloy. Also, notice that yield stress is highest for alloys at the low concentration of 8842 and initiator, so the crystal phase is not disrupted at these low levels. Between all alloys in this table, run 3 has the highest modulus and the lowest impact strength. Th e initiator and 8842 levels are low but all other levels are high so the gr afting degree may be very low for this sample. In turn, the crystalline phase is not as dr astically affected and so it retains its mechanical integrity. 7.3.2 Design of Experiments Results The disadvantage of using a fractional fact orial design is that each main-effect and interaction contrast will be confounded with one or more other main-effect and interaction contrasts and so ca nnot be estimated separately. Two factorial contrasts that are confounded are referred to as being alia sed. The aliasing problem comes to the surface when a significant main effect is iden tified, but is not clear whether the observed effect is due to the main effect or to it s interaction with other factors or to the combination of both. Because of the aliasing problem, fr actional factorial designs are most often used as screening experiments. Appendix E lists the factorial effects aliases of interest. 7.3.2.1 Impact strength When analyzing data from unreplicated factorial designs, occasionally real highorder interactions occur. A normal probability plot is thus used to estimate the effects. Figure 7-2 is a half normal plot of the significant effects on im pact strength. The effects that are negligible are normally distributed with a mean of ze ro and a variance of 2 and will tend to fall along a straight line in this plot. Signifi cant effects will have nonzero

PAGE 244

222 means and will not lie along the straight line. A diagnostic check indicates that the only significant effects are in itiator, HDPE:8842 ratio, and the interaction between them. Figure 7-2: Half normal plot showing, in c oded variables, the most significant effects on impact strength. A = initiator concen tration and C = concentration of 8842. Analysis of variance is derived from a partitioning of total variability into its component parts and is exemplified in Table 7-6. Sum of squares is a measure of the overall variability of the data. If the sum of squares is divided by the appropriate number of degrees of freedom, the sample variance would result. The num ber of degrees of freedom of a sum of squares is equal to the number of independent elements in that sum of squares. Each mean square is simply its sum of squares divi ded by its degrees of freedom. F is essentially a te st statistic for the null hypothesi s and the F ratio is the mean square of the treatment of interest divide d by the mean square of the error within treatments. At an = 0.05, the F value is accepted, so the null hypothesis is rejected and both initiator and 8842 concentra tion significantly affect imp act strength. The model F-

PAGE 245

223 value of 76.6 implies the model is significant. There is only a 0.01% chance that a model F-value this large could occur due to noise. Values of "Prob > F" less than 0.05 indicate model terms are significant. In this case A, C, AC are significant model terms. Values greater than 0.10 indicate the m odel terms are not significant. Table 7-6: Analysis of variance table (par tial sum of squares) for impact strength. Source Sum of Squares Degrees of Freedom Mean Square F Value Prob > F Model 348.8 3 116.3 76.6 < 0.0001 significant A (% initiator) 113.96 1 114 75.1 < 0.0001 C (% 8842) 213.89 1 214 141 < 0.0001 AC 20.93 1 21 13.8 0.003 Residual 18.20 12 1.5 Cor Total 366.98 15 Std. Dev. = 1.23 R-Squared = 0.95 Mean = 7.7 Adj. R-Squared = 0.94 C. V. = 16.1 Pred. R-Squared = 0.92 PRESS = 32.4 Adeq. Precision = 20.5 The "Pred R-Squared" of 0.92 is in reasona ble agreement with th e "Adj R-Squared" of 0.94. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable, so the ratio of 20.5 indicates an adequate signal. This model can be used to navigate the design space. The final equation in terms of coded factors: Impact Strength = +7.66 + (2.67 A) + (3.66 C) + (1.14 A C) (7.1) The final equation in te rms of actual factors: Impact Strength = -0.36 + (1.09 % Initia tor) + (0.31 Amt ENG8842) + (0.65 % Initiator Amt ENG8842) (7.2) One more method to check for model adequ acy is in Figure Figure 7-7(a), which is a normal probability plot of studentized residuals. Because of the linearity of this figure, the model is said to have a normal distri bution about the mean. Figure 7-7(b) shows what, if any, outliers (i.e. influential values ) are present. An outlier will have an

PAGE 246

224 unusually high positive or negative value and are eas y to detect from this plot. Outliers may be due to the error va riables not being normally distributed, having different variances, or an incorrect speci fication of model. Run 2 is the only outlier but it also has a higher standard deviation than all other runs This material has a high concentration of initiator and DEGDA but low levels of temper ature, screw speed, and styrene. These each have typically opposite effects on grafting ef ficiency of PP from Chapter 5. A high (a) (b) Figure 7-3: Normal plot of residuals (a) and outliers (b) s how the diagnostic results of the model for impact strength. initiator means more primary free radical s and hence higher grafting, high DEGDA content traps generated free ra dicals and creates more of a branched structure, low styrene means less homopolymerized material, low temperature leads to lower grafting efficiency, and lower screw speed has a te ndency towards higher grafting efficiency. Figure 7-4 is a cube graph of impact stre ngth at various levels for each of these parameters. For high impact strength, initia tor and 8842 concentration must be at their high levels, but all other va riables will not have the same influential effects.

PAGE 247

225 DESIGN-EXPERT Plot Impact Strength X = A: % Initiator Y = C: Amt ENG8842 Z = B: % Styrene Actual Factors D: Screw Speed = 150.00 E: % DEGDA = 0.75 F: Temperature = 200.00Cube GraphImpact Strength A: % InitiatorC: Amt ENG8842B: % Styrene A-A+ CC+ BB+ 2.475 2.475 7.5 7.5 5.525 5.525 15.125 15.125 Figure 7-4: Cube graph of th e effect of % initiator, 8842 content, and % styrene on impact strength at constant scre w speed, % DEGDA, and temperature. 7.3.2.2 Elastic modulus The half normal % probability plot in Figur e 7-5 shows that only two variables (% initiator and 8842 content) significantly affect the response of elastic modulus. Analysis of variance in Table 7-7 gives a lot of info rmation regarding the adequacy of the model and the ability to predict future outcomes. The model F-value of 56.1 implies the model is significant. There is only a 0.01% chan ce that a model F-value this large could occur due to noise. Values of "Prob > F" less than 0.05 indicate model terms are significant. In this case A, C, are significant model terms. Values greater than 0.10 indicate the model terms are not significant. The "Pred R-Squared" of 0.843 is in reasonable agreement with the "Adj R-Squared" of 0.88. "Adeq Precisi on" measures the signal to noise ratio. A

PAGE 248

226 ratio greater than 4 is desirable. The ratio of 16.7 indicates an adequate signal. This model can be used to navigate the design space. Figure 7-5: Half normal plot showing, in c oded variables, the most significant effects on elastic modulus. Table 7-7: Analysis of variance table (par tial sum of squares) for elastic modulus. Source Sum of Squares Degrees of Freedom Mean Square F Value Prob > F Model 9.4E+5 2 4.71E+5 56.1 < 0.0001 significant A 2.3E+5 1 2.31E+5 27.5 0.0002 C 7.1E+5 1 7.1E+5 84.6 < 0.0001 Residual 1.1E+5 13 8392.9 Cor Total 1.05E+6 15 Std. Dev. = 91.6 R-Squared = 0.90 Mean = 962.3 Adj. R-Squared = 0.88 C. V. = 9.5 Pred. R-Squared = 0.84 PRESS = 1.6E+005 Adeq. Precision = 16.68 The final equation in te rms of coded factors: Elastic Modulus = + 962 (120 A) (211 C) (7.3) The final equation in te rms of actual factors:

PAGE 249

227 Elastic Modulus = + 1607 (343 % In itiator) (42 Amt ENG8842) (7.4) (a) (b) Figure 7-6: Normal plot of residuals (a) and outliers (b) s how the diagnostic results of the model for elastic modulus. Because of the linearity of this figure, the model is said to have a normal distribution about the mean. Fi gure 7-6(b) shows what, if a ny, outliers (i.e. influential values) are present. Run 3 is the only outli er, which has an extrao rdinarily high value. The material has a low initiator concentrati on and was extruded at a low temperature, but contains high monomer concentr ations at the higher temper ature. Elastic modulus is essentially the resistance of the non-crystalline phase of the plastic to deformation at low strains. A high initiator c oncentration may promote cross linking of both PE and 8842, which should positively effect el astic modulus. The cube graph in Figure 7-7 shows the opposite trend. There may be pockets of in soluble gel which nega tively effect stressstrain properties as a separate dispersed pha se. More studies have to be conducted

PAGE 250

228 DESIGN-EXPERT Plot Elastic Modulus X = A: % Initiator Y = C: Amt ENG8842 Z = B: % Styrene Actual Factors D: Screw Speed = 150.00 E: % DEGDA = 0.75 F: Temperature = 200.00Cube GraphElastic Modulus A: % InitiatorC: Amt ENG8842B: % Styrene A-A+ CC+ BB+ 1293.19 1293.19 871.813 871.813 1052.81 1052.81 631.438 631.438 Figure 7-7: Cube graph of the effect of % in itiator, 8842 content, a nd % styrene on elastic modulus at constant screw speed % DEGDA, and temperature. regarding gel content an d grafting efficiency. Lower initiator concentrations presumably lead to lower levels of grafting in HDPE, but at high monomer concentrations, there is likely a high percentrage of highly entangl ed, homopolymerized ma terial within the amorphous phase. Further studies also need to be performed on the crystalline state of the alloys because lamellar crys tals will essentially act as fi ller particles in the amorphous phase and thus stiffen the polymer. At high in itiator concentraitons, the crystalline state may be disrupted. 7.3.2.3 Yield strength The half normal plot in Figure 7-8 indicat es that, like elastic modulus, % initiator and 8842 concentration have the gr eatest effect on yield stress.

PAGE 251

229 Figure 7-8: Half normal plot showing, in c oded variables, the most significant effects. From the analysis of variance (Table 78), the model F-value of 92.3 implies the model is significant. There is only a 0.01% chance that a model F-value this large could occur due to noise. In this case A nad C are significant model terms. The "Pred RSquared" of 0.90 is in reasonable agreement with the "Adj R-Squared" of 0.92. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. The ratio of 21 indicates an adequa te signal. This model can be used to navigate the design space. The final equation in terms of coded factors is: Yield Strength = + 21.4 (1.2 A) (2.4 C) (7.5) The final equation in terms of actual factors is: Yield Strength = + 28.4 (3.4 % In itiator) (0.5 Amt ENG8842) (7.6) Figure 7-9(a) is linear with respect to studentized residuals, so the results from this design are normal about the mean. There are no outliers from Fi gure 7-9(b) and the

PAGE 252

230 Table 7-8: Analysis of variance table (p artial sum of squares) for yield stress. Source Sum of Squares Degrees of Freedom Mean Square F Value Prob > F Model 115.9 2 58 92.3 < 0.0001 significant A 22.3 1 22.3 35.5 < 0.0001 C 93.6 1 93.6 149 < 0.0001 Residual 8.2 13 0.6 Cor Total 124.1 15 Std. Dev. = 0.79 R-Squared = 0.93 Mean = 21.4 Adj. R-Squared = 0.93 C. V. = 3.7 Pred. R-Squared = 0.90 PRESS = 12.4 Adeq. Precision = 20.98 distribution of is random about zero. The cube graph in Figure 7-10 indicates that 8842 and initiator level should be lo w in order to maximize yield st ress. The crystalline state may be adversely affected at high initia tor concentrations a nd insoluble gel may deleteriously affect yield stress but more studies have to be conducted on this issue. (a) (b) Figure 7-9: Normal plot of residuals (a) and outliers (b) s how the diagnostic results of the model for yield stress.

PAGE 253

231 DESIGN-EXPERT Plot Yield Strength X = A: % Initiator Y = C: Amt ENG8842 Z = B: % Styrene Actual Factors D: Screw Speed = 150.00 E: % DEGDA = 0.75 F: Temperature = 200.00Cube GraphYield Strength A: % InitiatorC: Amt ENG8842B: % Styrene A-A+ CC+ BB+ 24.9563 24.9563 20.1188 20.1188 22.5938 22.5938 17.7563 17.7563 Figure 7-10: Cube graph of the effect of % initiator, 8842 content, and % styrene on yield stress at constant screw sp eed, % DEGDA, and temperature. 7.3.2.4 Elongation at break Elongation (the ability to neck and draw ) is more sensitive to defects and morphology than the other three mechanical prop erties studies thus far [271, 272]. This is evidenced by the greater number of significan t effects from Figure 7-11. The initiator concentration has the greatest effect, fo llowed by 8842 concentration, screw speed, and the interaction between scre w speed and % styrene. The ANOVA table (Table 7-8) gives direct analysis of these main effects. The model F-value of 27.5 implies the model is si gnificant. There is only a 0.01% chance that a model F-value this large could occur due to noise. Values of "Prob > F" less than 0.05 indicate model terms are significant. In th is case A, C, D, and BD are significant

PAGE 254

232 Figure 7-11: Half normal plot showing, in c oded variables, the most significant effects. model terms. Values greater than 0.10 indicate the model terms are not significant. The "Pred R-Squared" of 0.81 is in reasonabl e agreement with the "Adj R-Squared" of 0.88. "Adeq Precision" measures the si gnal to noise ratio. A ratio greater Table 7-9: Analysis of varian ce table (partial sum of squa res) for elongation at break. Source Sum of Squares Degrees of Freedom Mean Square F Value Prob > F Model 2872.8 4 718.2 27.5 < 0.0001 significant A 1870.6 1 1870.6 71.5 < 0.0001 C 451.6 1 451.6 17.3 0.002 D 72.7 1 72.7 2.8 0.124 BD 333.1 1 333.1 12.7 0.004 Residual 287.7 11 26.2 Cor Total 3160.4 15 Std. Dev. = 5.1 R-Squared = 0.9090 Mean = 40.8 Adj. R-Squared = 0.88 C. V. = 12.5 Pred. R-Squared = 0.81 PRESS = 608.7 Adeq. Precision = 17.1 than 4 is desirable. The ratio of 17.1 indica tes an adequate signal. This model can be used to navigate the design space. The fi nal equation in terms of coded factors is:

PAGE 255

233 Elongation = + 40.8 (10.8 A) + (5.3 C) (4.3 D) + (3.4 B D) (7.7) The normal probability plot in Figure 7-12 is linear, so the results are normal about the mean. There are no outliers from Figure 7-12(b) and the distri bution is random about zero. The cube graph in Figure 7-13 s hows that a high amount of 8842 along with a lower screw speed and initiator concentration will give the greatest elongation at break. This means that the amorphous 8842 may have a higher molar mass than HDPE or crosslinking may not be signifi cant. A low initiator concentration means lower degree of grafting and side reactions like crosslinking. But a lo wer screw speed means longer residence time and less intensive mixing. From data gathered in Chapter 5 regarding PP, lower initiator concen tration gives a high el ongation at break but low screw speed results in low elongation. (a) (b) Figure 7-12: Normal plot of residuals (a) and outliers (b) show the diagnostic results of the model for elastic modulus.

PAGE 256

234 DESIGN-EXPERT Plot Elongation X = A: % Initiator Y = C: Amt ENG8842 Z = D: Screw Speed Actual Factors B: % Styrene = 4.00 E: % DEGDA = 0.76 F: Temperature = 193.78Cube GraphElongation A: % InitiatorC: Amt ENG8842D: Screw Speed A-A+ CC+ DD+ 47.19 45.44 57.81 56.06 25.56 23.81 36.19 34.44 Figure 7-13: Cube graph of the effect of % initiator, 8842 content, and % styrene on elongation at break at co nstant % styrene, % DE GDA, and temperature. 7.4 Conclusions The reactive extrusion of HDPE to increase its impact strength has been successful, but its stress-strain performance suffered which is typical of rubber toughened thermoplastics. The grafting of the polymers was thought to increase the resistance to a tensile load, which it does in the Izod impact test. But at slow deformation rates, homopolymerized and non-polymerized monomer may affect the respons e of the alloy to a greater extent. A design of experiments proved to be i nvaluable in deciding what variables significantly affected responses and the leve ls at which properties are maximized. The wt% initiator and amount of 8842 proved to be most important for overall property control. The initiator dict ates when and to what exte nt the free radical grafting,

PAGE 257

235 polymerization, and crosslinking events will ta ke place. The content of 8842 has such a drastic effect on properties due to its non-crystalline state and low Tg characteristics. Also, 8842 is thought to be phase separated from HDPE, so its interfacial interactions will play a huge role in macroscopic properties.

PAGE 258

236 CHAPTER 8 CONCLUSIONS AND FUTURE WORK 8.1 Summary and Conclusions This work successfully described how to improve isotactic polypropylenes (PP) impact strength and why this occurs without compromising stress-strain performance. A novel ethylene-1-octene copolymer (EOC) was used as the impact modifier in the blends and alloys due to its outstanding low temper ature toughness, limited de fect distribution, excellent processability, and a fully satura ted backbone. A low molecular weight grade has been shown to be an excellent candidate for modifying PP because it can be easily dispersed and distributed within the matrix. By simply blending the impact modifier wi th PP, a modest jump in impact strength is seen but the all too comm on decrease in tensile strength modulus, and elongation at break are experienced. So, a reactiv e extrusion process was employed to: Stabilize the phase separated domains within the matrix phase, Reduce average domain size, and Create an efficient stress transfer mechanism between phases. This process used free radical chemistry to polymerize monomers using a so-called in-situ compatibilization techni que as a one-step synthetic method to produce a polypropylene alloy. The combination of a peroxide initiator, styrene, and a multifunctional acrylate have proven to give a high grafting efficiency, low melt flow index, outstanding melt strength, and extr emely high impact strength along with improved stress-strain behavior over the physical blend.

PAGE 259

237 There are many types of reactions presen t in this complex reactive extrusion process. A key to controlling the microstruc ture of the alloys and the resulting macroscale properties is to understand free radica l grafting mechanisms. Figure 4-38 allows one to visualize the reactions likely occurring during the reac tive extrusion process. As can be seen, initiator decomposition is th e rate determining step, but once that Figure 8-1: Schematic drawing of the likely free radical initiated processes during the reactive extrusion of PP, 8407, initia tor, styrene, and multifunctional monomer. energy barrier is overcome, several reactions happen at once. These include monomer polymerization, hydrogen abst raction from polymers, gr afting of monomers onto polymers, and degradation of polymers. The end result is likely a diffusely

PAGE 260

238 interconnected, branched collection of macromolecules with unique morphology, improved melt properties, and exce llent mechanical properties. Many interesting amorphous and crystalline fe atures have been created through this reactive extrusion process. Using TEM for example, elastomer domains in the physical blends are distributed bimodally as spherica l domains that range in size from 50 to 150 nm. For high impact samples, domains range in size from 10 nm (d istributed throughout the sample) to oblong 300 nm domains having a variety of stained material dispersed throughout. SEM has shown that the physic al blends etched domains are roughly spherical, but the alloys ha ve irregularly shaped domains which are much smaller in diameter and closer in proximity to eachother. During an impact test, the extension of the amorphous phase and a rise temperature likely results in both a phase transformation and a highly elastic retractive force, thereby limiting chain breakage. Figure 8-2 shows an interpretation of how grafted material locates in and around dispersed elastomer domains in a polypropyl ene-based alloy. For the high impact strength alloys, a high grafting efficiency coupled with sma ll elastomer domains leads to a high degree of physical entanglements, covale nt bonding, and an alte red stress field at the interface of the pa rticles. Also, grafting is tho ught to occur throughout these small domains. The mobility of the amorphous phase of PP is restricted in the presence of these branching components but dispersion of the elastomer n such a small scale may change the relaxation time distribution of PP. For the physical blends, large particles exist with a greater interparticle distance. This results in a low particle-matrix surface area and the ability to abso rb energy through typical mechanisms like shear yielding, crazing, and cavitation is reduced.

PAGE 261

239 Figure 8-2: Interpretation of the effect of in-situ grafted polymeric chains at the PPelastomer interface (blue circles = elas tomer domains, black lines = grafted polymers). The physical blend (left) ha s no grafting and the alloy (right) has a high degree of grafting. The crystalline state of the alloys is unique in that many small crystals with a high cross-hatch density are distri buted throughout the sample. Th e high surface area leads to a higher energy state of the crystal, so the process of melt and r ecrystallization during deformation is facilitated. This enhan ces mechanical properties by absorbing and dissipating energy effi ciently. Also, the phase of PP is present in the alloys, which is known to have more energy ab sorption capability than the phase due to its larger lattice structure and lower melting enthalpy. Pola rized light microscopy coupled with DSC proved invaluable in the study of the crysta llization of these materials. The high branching and interconnected network of the alloys drastically increases the crystallization temperature and decreases th e spherulite size compared to its physical blend counterparts. The unique molecular architecture appears to promote the local ordering process known as the induction pe riod before the well known nucleation and growth phenomenon. Processing variables have a significant eff ect on overall properties of the alloys. Initiator concentration, which is the rate de termining step in the grafting process, has a drastic effect on overall properties of the al loy. A high grafting efficiency does not

PAGE 262

240 necessarily translate into hi gh impact strength because of the degradation of PP through chain scission. Screw speed, barrel temperat ure, and monomer concentrations all play important roles in the performance of the all oy. Grafted material positively influences the overall mechanical properties of PP, but it is still not known whether this is due to a reduction in the amount of homopolymerized ma terial, reduction in degradation of PP, and/or improvement of interfacial adhesion. By controlling the properties of the imp act modifier, both the blend and alloys macro-scale performance can be dictated. A high density elastomer will give a high elastic modulus and yield strength, but poor imp act strength. It will melt at a later point along the extruder barrel which results in a lo w grafting efficiency for the alloy. An elastomer with a very high MFI (low MW) w ill not give good impact properties because of its lack of molecular entanglements and te ndency to coalesce. On the other hand, there is a sharp transition of elastomer molecular weight where all mechanical, rheological, and chemical properties drastically improve. High density polyethylene, when modified using the reactive extrusion process with a high MW elastomer, gives an extremely high impact strength but at the expense of a low modulus and yield stress. A high degr ee of homopolymerized material as well as disturbance of the crystalline state of HDPE may be the cause for the imbalance of properties. A design of expe riments approach allowed objec tive results to be obtained regarding which parameters affected imp act strength, yield stress, modulus, and elongation at break. 8.2 Future Work Reactive extrusion is such a diverse, vers atile system that innumerable possibilities exist for the modification of polymers. One unique study that should be done is to use

PAGE 263

241 fluorescent or deuterated mono mers during the in-situ reactive extrusion process. One can then monitor and map out, either by c onfocal fluorescent microscopy, SANS, or NMR, the length of grafts, the placement of grafts, and the exact composition of grafts. The morphology of the alloys was so interest ing that further studies must be done to clarify exactly what the deeply stained domains represent. Positron annihilation will give the free volume of the system, which is sure to change upon modification of polymers by grafting monomers. TEM coinciding with electron diffraction will elucidate the exact lamellar structure(s) present. Light scat tering is another impo rtant technique when studying crystalline morphology. By anneali ng the high impact samples and comparing DSC and XRD results to pure PP, one should ge t a better idea of the limiting effect of branching on chain folding. Isothermal crystallization experiments will show the crystallization kinetics of the system and verify the belief that multifunctional monomer drastically increases crystallization rate. Also, in-situ small angle neutron scattering will allow one to explore the induc tion period before nucleation and growth, which should be unique to this alloy. Studies should be conducted on the deform ation and fracture mechanisms of this system. In-situ deformation experiments using TEM is vital to understand the structure of these alloys will the stained regions cavitate, craze, or delaminate from the surrounding material? Stained TEM sections of the region surrounding the arrested crack tip in fractured impact bars may be very informative. Distinct ductile-to-brittle transitions should be found, even though it is already known that at high monomer and elastomer concentrations the materials cha nges its behavior from tough and ductile to brittle. The crack su rfaces should be studied in more detail, as well as the notch

PAGE 264

242 sensitivity. AFM or nanoindentation is valu able in polymer characterization because it should be able to pinpoint the modulus and vi scoelasticity of the established nanometersized domains. Also, in-situ XRD/tensile deformation e xperiments should be done to determine exactly what phase transf ormations occur during deformation. The processing conditions can further be varied to include an array of monomers with diverse viscosities and functionalities. The type and half life of initiator can be varied, powder instead of PP granules may be used, pellets could be imbibed with monomer mixtures before pro cessing, and design should be cr eated with a more complete set of experiments using supercritical carbon dioxide (scCO2) as a processing/grafting aid. Supercritical CO2 should be added at different ports along the extruder, it should be used with a variety of peroxides having lowe r half lives at 200C, extruder screw speed should be lowered and barrel temperatur e should be increased when using scCO2. Adding high strength fillers to polymers is of great interest due to their ability to enhance mechanical, thermal, and rheological properties. Further studies should include the use of layered silicate cl ays (possibly montmorillonite), exfoliated graphite, carbon nanotubes, silica nanoparticles, and tungsten ca rbide whiskers, to name a few. A unique study may encompass surface functionalization of the particles to facilitate bonding between phases. Combining rigid fill ers with thermoplastics is done through inexpensive, continuous processi ng routes, but a problem facing materials scientists is to efficiently disperse and distribute all t ypes of particles having different surface chemistries. Alignment of high aspect ratio fillers may also be a desirable feature while melt processing to leverage their isotropic stru ctures. Application of supercritical carbon dioxide is one method to do this. By functi onalizing the surface of the nanoparticles with

PAGE 265

243 scCO2-philic functional groups, the shear for ces in an extruder along with scCO2 will facilitate the needed dispersion and distribution. scCO2 will also allow production of nano-elas tomer domains in thermosets which will improve toughening to a greater extent than micron-sized elastomers. The RESS process (Rapid Expansion from a Supercritical Solution) has been modified recently to include Rapid Expansion of a supercritical solution into a solvent (RESolv). This shows promise in being able to dissolve polymers in scCO2 and upon release of pressure shoot small polymeric particles into the matrix of choice.

PAGE 266

244 APPENDIX A CALIBRATION CURVE FO R ABSOLUTE STYRENE CO NCENTRATIONS IN REACTIVE BLENDS For quantification of the grafted styrene, a calibration curve (Figure A-2) had to be established. This was done by extruding pol ypropylene/polystyrene blends of varying PS concentrations. By taking the ratio of the areas under the 700 cm-1 and 1376 cm-1 peaks and relating this to the amount of styrene a dded, absolute amounts of styrene present can be easily calculated. A ratio of areas is taken from Figure A-1 so that thickness variations will not affect the results. Figure A-1: FTIR calibration gr aphs of iPP/PS physical blen ds with varying PS content; A = 0%, B = 2%, C = 4%, D = 6%, E = 8%, F = 10%. A B C D E F

PAGE 267

245 y = 0.235x R2 = 0.9905 0 5 10 15 20 25 020406080100PP:PS RatioAbsorbance Ratio_1376 cm^-1 pea k area/702 cm^-1 peak area Figure A-2: Calibration curve for abso lute styrene content determination

PAGE 268

246 APPENDIX B STRESS-STRAIN GRAPHS AND STATISTICS This Appendix contains both gr aphs and statistical data fo r stress-strain behavior in Chapters 3, 4, 5, and 6. Table B-1: Actual stress-strai n values with standa rd deviations for Figures in Chapter 3 and Chapter 4. Sample ID Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) Pure PP 1833 57 36.8 1.7 74 2 49.4 1.4 6000 274 95:5_0 1707 84 32.6 0.3 70 8 42.8 2.5 5683 943 95:5_A 1740 36 33.4 1.1 49 8 38.3 1 4254 680 95:5_B 1858 105 34.7 1.6 54 7 43.8 5.2 5149 1298 95:5_C 1761 60 33.1 0.9 53 9 41.3 4.2 4800 1392 90:10_0 1450 59 28.0 1.9 58 12 38.7 4.8 4038 825 90:10_A 1502 74 28.3 0.9 57 4 37 1.1 3840 736 90:10_B 1695 110 30.9 1.5 60 4 42.6 1.5 4294 155 90:10_C 1621 47 29.9 1 59 4 41.8 1.7 4458 723 80:20_0 1125 66 23.4 1.1 39 8 24.5 3.5 3065 1094 80:20_A 1275 94 24.6 1.8 35 14 29 4.8 2677 896 80:20_B 1347 77 24.8 0.5 54 10 35.8 4.3 3635 1269 80:20_C 1356 110 25.3 2 56 10 35.9 4.6 3488 1208 70:30_0 928 91 17.1 1.1 6.1 3 9.8 1.1 179 117 70:30_A 996 37 18.1 2.6 7 3 15 1.9 237 79 70:30_B 1054 33 18.5 1 19 7 19.8 4.3 1308 765 70:30_C 1109 53 20.3 1.6 11 5 20.3 1.5 768 200

PAGE 269

247 Figure B-1: Stress-str ain graph comparison of 95:5 (PP: 8407) physical blend and alloys. Figure B-2: Stress-str ain graph comparison of 90:10 (PP: 8407) physical blend and alloys.

PAGE 270

248 Figure B-3: Stress-str ain graph comparison of 80:20 (PP: 8407) physical blend and alloys. Figure B-4: Stress-str ain graph comparison of 70:30 (PP: 8407) physical blend and alloys.

PAGE 271

249 Figure B-5: Stress-strai n graph comparison of the effect of extruder barrel temperature in relation to Figure 5-2 and Table 5-4. 1= low temperature, 2=middle temperature, 3=high temperature. Figure B-6: Stress-str ain graph comparison of the effect of extruder screw speed in relation to Table 5-6.

PAGE 272

250 Figure B-7: Stress-strain graph comparison of the effect of initiator concentration in relation to Table 5-7. Figure B-8: Stress-str ain graph comparison of the eff ect of DEGDA concentration in relation to Table 5-8.

PAGE 273

251 Figure B-9: Stress-str ain graph comparison of the effect of styrene concentration in relation to Table 5-9. Figure B-10: Stress-strain graph comparison of the effect of elastomer density in relation to Table 6-5.

PAGE 274

252 Figure B-11: Stress-strain graph comparison of the effect of elastomer density. Figure B-12: Stress-strain graph comparison of the effect of elastomer density.

PAGE 275

253 Figure B-13: Stress-strain graph comparison of the effect of elastomer density. Table B-2: Stress-strain properties of PP:E NGAGE blends and all oys as a function of elastomer melt flow index. Blend ID (MFI in g/10min) Elastic Modulus (MPa) Yield (MPa) Elongation at Break (mm) Break (MPa) Energy to Break (N*mm) 002_0 (500) 1387 72 28.3 1.6 60 9 38.7 4.1 4636 956 8407_0 (30) 1450 59 28.5 1.9 58 12 38.7 4.8 4038 825 8200_0 (5) 1489 49 28.8 1.3 74 4 45 0.6 6543 196 8842_0 (1) 1591 54 30.5 0.6 76 8 49.5 3.3 6814 1426

PAGE 276

254 Figure B-14: Stress-strain graph of physical blends of ENGAGE elas tomers with PP as a function of elastomer melt flow index. Figure B-15: Stress-strain graph comparison of the effect of elastomer MFI.

PAGE 277

255 Figure B-16: Stress-strain graph comparison of the effect of elastomer MFI. Figure B-17: Stress-strain graph comparison of the effect of elastomer MFI.

PAGE 278

256 APPENDIX C FTIR GRAPHS This Appendix lists representative graphs used to find grafting efficiency for many of the alloys found in Chapters 4 and 5. Figure C-1: FTIR graph comparison of 95:5_0, 95:5_A, 95:5_B, and 95:5_C.

PAGE 279

257 Figure C-2: FTIR graph comparison of 90:10_0, 90:10_A, 90:10_B, and 90:10_C. Figure C-3: FTIR graph comparison of 80:20_0, 80:20_A, 80:20_B, and 80:20_C.

PAGE 280

258 Figure C-4: FTIR graph comparison of 70:30_0, 70:30_A, 70:30_B, and 70:30_C. Figure C-5: FTIR graphs from Chapter 5 (Fig ure 5-2) of alloys processed at varying temperatures. 1 = low barrel temperatur e, 2 = middle barrel temperature, 3 = high barrel temperature.

PAGE 281

259 Figure C-6: FTIR graphs from Chapter 5 (Fig ure 5-3) of alloys processed at varying screw speeds. Figure C-7: FTIR graphs from Chapter 5 (Fig ure 5-4) of alloys processed at varying concentration of initiator.

PAGE 282

260 Figure C-8: FTIR graphs from Chapter 5 (Fig ure 5-5) of alloys processed at varying concentration of multifunctional monomer. Figure C-9: FTIR graphs from Chapter 5 (Fig ure 5-6) of alloys processed at varying concentration of styrene, with DEGDA as multifunctional monomer.

PAGE 283

261 Figure C-10: FTIR graphs from Chapter 5 (Fi gure 5-7) of alloys processed at varying concentration of styrene, with TMPTA as multifunctional monomer.

PAGE 284

262 APPENDIX D IMAGE ANALYSIS This Appendix is a compilation of histogr ams representing image analysis of the SEM pictures from Chapters 3 and 4. It in cludes side-by-side comparisons of the mean diameter and roundness of physical blends vs. alloys at varying ratios of PP to 8407. The y axiz is log scale for easier interpretation of results. Thes e histograms are gathered from at least three different pictures take n from cryo-fractured, etched surfaces. (a) (b) Figure D-1: Histogram of av erage particle diameters for (a) 70:30_0 and (b) 70:30_B. (a) (b) Figure D-2: Histograms of particle roundness for (a) 70:30_0 and (b) 70:30_B

PAGE 285

263 (a) (b) Figure D-3: Histogram of av erage particle diameters for (a) 80:20_0 and (b) 80:20_B (a) (b) Figure D-4: Histograms of particle roundness for (a) 80:20_0 and (b) 80:20_B (a) (b) Figure D-5: Histogram of av erage particle diameters for (a) 90:10_0 and (b) 90:10_B

PAGE 286

264 (a) (b) Figure D-6: Histograms of particle roundness for (a) 90:10_0 and (b) 90:10_B (a) (b) Figure D-7: Histogram of av erage particle diameters for (a) 95:5_0 and (b) 95:5_B (a) (b) Figure D-8: Histograms of particle roundness for (a) 95:5_0 and (b) 95:5_B

PAGE 287

265 APPENDIX E ALIASED TERMS FROM CHAPTER 7 Table E-1: Aliased terms from the fracti onal factorial design given in Chapter 7 [Est. Terms] Aliased Terms [Intercept] = Intercep t 1.75 ACE 1.75 CDF [A] = A 1.75 CE 2.75 BCE + DEF [B] = B + 0.75 ACE + 0.75 CDF [C] = C + 2.33 BC + 1.33 ABE + 1.33 BDF [D] = D 1.75 CF + AEF 2.75 BCF [E] = E + 2.33 BE + 1.33 ABC + ADF [F] = F 1.75 CD + ADE 2.75 BCD [AB] = AB + 0.75 CE + 1.75 BCE [AC] = AC + 1.33 BE + 2.33 ABC [AD] = AD + EF 1.75 ACF 1.75 CDE [AE] = AE + 1.33 BC + DF + 2.33 ABE + 2.33 BDF [AF] = AF + DE 1.75 ACD 1.75 CEF [BD] = BD + 0.75 CF + 1.75 BCF [BF] = BF + 0.75 CD + 1.75 BCD [ABD] = ABD + 0.75 ACF + BEF + 0.75 CDE [ABF] = ABF + 0.75 ACD + BDE + 0.75 CEF

PAGE 288

266 LIST OF REFERENCES 1. W. D., Calister Jr., Materials Science and Engi neering: An Introduction, John Wiley and Sons, New York, (1997). 2. C. Rauwndaal, Plastics Formulating and Compounding, 15-26 (November/December, 1995). 3. D. B. Todd, Plastics Compounding, Hanser/Gardner Inc ., Cincinnati, OH (1998). 4. M. Xanthos, Reactive Extrusion Pr inciples and Practice, Hanser/Gardner Inc., Cincinnati, OH (1992). 5. R. A. Pearson, ACS Symp. Ser. 759, Toughening of Plastics, American Chemical Society, Washington, DC, Chapters 1-10 (2000). 6. W. P. Benoit II, SPE ANTEC, 3, 4298 (1995). 7. L. Yao, The Functional Monomers Grafted Poly olefins and Their Applications in the Compatibilization of Polymer Blends, Ph.D. Dissertation, Un iversity of Florida (1996). 8. D. R. Paul and C. B. Bucknall (Eds.), Polymer Blends, Volumes 1 and 2, John Wiley & Sons, Inc., New York (2000). 9. A. A. Collyer (Ed.), Rubber Toughened Engineering Plastics, Chapman & Hall, London (1994). 10. J. W. Barlow and D. R. Paul, Polym. Eng. Sci., 21, 985 (1981). 11. H. Tang, Novel Polyolefin Elastomer-Base d Blends and Their Applications, Ph.D. Dissertation, Universi ty of Florida (2000). 12. I. S. Miles and S. Rostami, Multicomponent Polymer Systems, Longman Scientific & Technical, Essex, UK (1992). 13. C. W. Macosko, Macromol. Symp., 149, 171 (2000). 14. A. Ajji and L. A. Utracki, Polym. Eng. Sci., 36, 1574 (1996).

PAGE 289

267 15. L. H. Sperling, (Ed.), Recent Advances in Polyme r Blends, Grafts, and Blocks, Plenum, New York (1974). 16. C. B. Bucknall, Toughened Plastics, Applied Science Publishers, London (1977). 17. K. J. Ganzeveld and L.P.B.M. Janssen, Polym. Eng. Sci., 7, 457 (1992). 18. M. Kontopoulou, W. Wang, T. G. Gopakumar, and C. Cheung, Polymer, 44, 7495 (2003). 19. F. C. Stehling, T. Huff, C. S. Speed, and G. Wissler, J. Appl. Polym. Sci., 26, 2693 (1981). 20. C. Lotti, C. A. Correa, and S. V. Canevarolo, Mater. Res., 3, 37 (2000). 21. A. van der Wal, R. Nijhof, and R. J. Gaymans, Polymer, 40, 6031 (1999). 22. B. Z. Jang, D. R. Uhlmann, and J. B. Vander Sande, Polym. Eng. Sci., 25, 643 (1985). 23. Y. Liu, X. Zhang, J. Gao, F. Huang, B. Tan, G. Wei, and J. Qiao, Polymer, 45, 275 (2004). 24. F. O. M. S. Abreu, M. M. C. Forte, and S. A. Liberman, J. Appl. Polym. Sci., 95, 254 (2005). 25. T. Whelan, Polymer Technology Dictionary, Chapman & Hall London, UK (1994). 26. H. F. Guo, S. Packirisamy, N. V. Gvozdic, and D. J. Meier, Polymer, 38, 785 (1997). 27. R. Potluri, C. G. Gogos, M. R. Libera, and S. S. Dagli, SPE ANTEC, 3, 3172 (1995). 28. N. C. Liu, and W. E. Baker, Polym. Eng. Sci., 32, 1695 (1992). 29. N. C. Liu, and W. E. Baker, Polymer, 5, 988 (1994). 30. J.-R. H. Schauder, U. S. Patent 6,303,688 (2001). 31. W. E. Baker, C. E. Scott, and G.-H. Hu, Reactive Polymer Blending, Hanser Gardner, Cincinatti, OH, Chapters 6 and 7 (2001). 32. U. Sundararaj, Morphology Development During Polymer Blending, Ph.D. Dissertation, University of Minnesota (1994). 33. D. Gersappe, D. Irvine, A. C. Balazs, Y. Liu, J. Sokolov, M. Rafailovich, S. Schwarz, and D. G. Peiffer, Science, 265, 1072 (1994).

PAGE 290

268 34. B. Z. Jang, D. R. Uhlmann, and J. B. van der Sande, Rubb. Chem. Techn., 57, 291 (1984). 35. S. Thomas, and R. E. Prudhomme, Polymer, 33, 4260 (1992). 36. D. W. Park, and R. J. Roe, Macromolecules, 24, 5324 (1991). 37. S. Datta, and D. J. Lohse, Macromolecules, 26, 2064 (1993). 38. S. Endo, K. Min, J. L. White, and T. Kyu, Polym. Eng. Sci., 26, 45 (1986). 39. P. Pieroni, D. Ercoli, G. Goizueta, and N. Capiati, Elast. Plast., 34, 131 (2002). 40. L. F. Chen, B. Wong, and W. E. Baker, Polym. Eng. Sci., 36, 1594 (1996). 41. M. S. Lee, T. P. Lodge, and C. W. Macosko, J. Polym. Sci. B: Polym. Phys., 35, 2835 (1997). 42. H. K. Jeon, and J. K. Kim, Macromolecules, 33, 8200 (2000). 43. G. H. Fredrickson and L. Leibler, Macromolecules, 29, 2674 (1996). 44. J. Suhm, J. Heinmann, C. Worner, P. Muller, F. Stricker, J. Kr essler, J. Okuda, and R. Mulhaupt, Macromol. Symp., 129, 1 (1998). 45. J. Karger-Kocsis, Polypropylene: An A-Z Reference, Kluwer Academic Publishers, Dordrecht, The Netherlands (1999). (a) R. Mulhaupt, Metallocene catalysts and tailor-made polyolefins, pp. 455-475. (b) R. Mulhaupt, Ziegl er-Natta catalysis and propylene polymerization, pp. 896-920. (c) H. Ledwinka, and W. Neibl, Industrial polymerization processes, pp. 314-319, (d) E. Andreassen, Infrared and Raman spectroscopy of polypropylen e, pp. 324-328, (e) A. Galeski, Nucleation, pp. 545-553. 46. E. P. Moore, Jr., Polypropylene Handbook, Hanser, Munich (1996). 47. P. Walter, D. Mader, P. Reichert, and R. Mulhaupt, J.M.S.-Pure Appl. Chem., A36, 1613 (1999). 48. DuPont Dow Elastomers, ENGAGE Pr oduct Information, September 2001, http://www.dupontdow.com/Products/Engage/engage.asp Last accessed June 4th, 2004. 49. J. Scheirs, and W. Kaminsky, Metallocene-based Polyolefins, Vol. 2, John Wiley & Sons, New York (2000). 50. F. J. Timmers, U. S. Patent 5,703,187, (1997). 51. J. Koivumaki, and J. Seppala, Polymer, 34, 1958 (1993).

PAGE 291

269 52. A. Bensason, J. Minick, A. Moet, S. Chum, A. Hiltner, and E. Baer, J. Appl. Polym. Sci., 34, 1301 (1996). 53. W. Kaminsky, Macromol. Chem. Phys., 197, 3907 (1996). 54. M. Arnold, J. Knorr, F. Koller, and S. Bornemann, J.M.S. Pure Appl. Chem., A36, 1655 (1999). 55. S. Lai and G. W. Knight, SPE ANTEC, 1, 1188 (1993). 56. L. T. Kale, T. A. Plumley, R. M. Patel, and P. Jain, SPE ANTEC, 2, 2249 (1993). 57. D. E. Turek and B. G. Landes, SPE ANTEC, 2, 2270 (1995). 58. S. Bensason, V. Stepanov, S. Chum, A. Hiltner, and E. Baer, Macromolecules, 30, 2436 (1997). 59. P. Thirion and T. Weil, Polymer, 25, 609 (1984). 60. M. L. Cerrada, R. Benavente, and E. Perez, J. Mater. Res., 16, 1103 (2001). 61. M. Peeters, B. Goderis, H. Reynaers, and V. Mathot, J. Polym. Sci. B: Polym. Phys., 37, 83 (1999). 62. B. Goderis, H. Reynaers, and M. H. Koch, Macromolecules, 35, 5840 (2002). 63. C. G. Vonk, J. Polym. Sci. C: Polym. Lett., 24, 305 (1986). 64. R. Androsch, Polymer, 40, 2805 (1999). 65. R. Androsch, and W. Bernhard, Macromolecules, 33, 9076 (2000). 66. C. I. Chung, Extrusion of Polymers Theory and Practice, Hanser, Munich (2000). 67. W. Michaeli, A. Grefensten, and U. Berghaus, Polym. Eng. Sci., 35, 1485 (1995). 68. G. Moad, Prog. Polym. Sci., 24, 81 (1999) and all references therein. 69. S. Wu, Polym. Eng. Sci., 27, 335 (1987). 70. P Cheung, D. Suwanda, and S. T. Balke, Polym. Eng. Sci., 30, 1063 (1990). 71. I.-K. Yang and J. D. Lin, Polym. Eng. Sci., 42, 753 (2002). 72. G. Shearer and C. Tzoganakis, Polym. Eng. Sci., 39, 1584 (1999). 73. C. Tzoganakis, Peroxide Degradation of Polyprop ylene During Reactive Extrusion, Ph.D. Thesis, McMaster University, (1988).

PAGE 292

270 74. A. V. Machado, J. A. Covas, and M. van Duin, J. Appl. Polym. Sci., 71, (1999), 135-141. 75. A. P. Plochocki, S. S. Dagli, and R. D. Andrews, Polym. Eng. Sci., 30, 741 (1990). 76. W. Huneault, M. F. Champagne, L. E. Daigneault, and M. M. Dumoulin, SPE ANTEC, 2, 2020 (1995). 77. K. J. Ganzeveld and L. P. B. M. Janssen, Polym. Eng. Sci., 32, 457 (1992). 78. C.-K. Shih, Polym. Eng. Sci, 35, 1688 (1995). 79. C.-K. Shih, D. G. Tynan, and D. A. Denelsbeck, Polym. Eng. Sci., 31, 1670 (1991). 80. S. Tyagi, and A. K. Ghosh, Polym. Eng. Sci., 42, 1309 (2002). 81. C. Jordan, J. Y. Cavaille, and J. Perez, J. Polym. Sci. B: Polym. Phys., 27, 2361 (1989) and all references therein. 82. M. A. Huneault, M. F. Champagne, and A. Luciani, Polym. Eng. Sci., 36, 1694 (1996). 83. C. E. Scott and C. W. Macosko, Polymer, 36, 461 (1995). 84. U. Sundararaj and C. W. Macosko, Polym. Eng. Sci., 35, 100 (1995). 85. M. H. Wilhelm and M. I. Felisberti, J. Appl. Polym. Sci., 85, 847 (2002). 86. H. K. Jeon and J. K. Kim, Macromolecules, 31, 9273 (1998). 87. R. G. Cox, J. Fluid Mech., 37, 601 (1969). 88. D. Barthes-Biesel, D. and A. Acrivos, J. Fluid Mech., 61, 1 (1973). 89. B. D. Favis, C. Lavallee, and A. Derdouri, A., J. Mater. Sci., 27, 4211-4218 (1992). 90. H. Cartier and G. H. Hu, Polym. Eng. Sci., 38, 177 (1999). 91. A. V. Machado, J. A. Covas, and M. Van Duin, Polymer, 42, 3649 (2001). 92. J. Cha and J. L. White, Polym. Eng. Sci., 41, 1238 (2001). 93. A. V. Machado, M. van Duin, and J. A. Covas, J. Polym. Sci. A: Polym. Chem., 38, 3919 (2000). 94. B. Wong, Free Radical Melt Grafting of Gl ycidyl Metheacrylate and Styrene onto Polypropylene: A Study of the Rh eological and Chemical Properties, Masters Thesis, Queens University (1995).

PAGE 293

271 95. Z. Song, and W. E. Baker, Die Angew Makromol. Chem., 181, 1, (1990). 96. H. F. Guo, S. Packirisamy, N. V. Gvozdic, and D. J. Meier, Polymer, 38, 785 (1997). 97. G. I. Taylor, Proc. R. Soc. London., A138, 41 (1932). 98. G. I. Taylor, Proc. R. Soc. London., A146, 501 (1934). 99. A. Galeski and Z. Bartczak, Macromol. Symp., 194, 47 (2003). 100. L. H. Sperling, Introduction to Physical Polymer Science, 2nd Ed., John Wiley & Sons (1992). 101. X. Chen and Y. Mai, Polym. Eng. Sci., 10, 1763 (1998). 102. C. B. Bucknall, A. Karpodinis, and X. C. Zhang, J. Mater. Sci., 29, 3377 (1994). 103. C. J. Chou, K. Vijayan, D. Kirby, A. Hiltner, and E. Baer, J. Mater. Sci., 23, 2521 (1988). 104. K. Dijkstra, A. Van der wal, and R. J. Gaymans, J. Mater. Sci., 29, 3489 (1994). 105. A. Galeski, Prog. Polym. Sci., 28, 1643 (2001) and all re ferences therein. 106. H. R. Brown, Macromolecules, 24, 2752 (1991). 107. J. T. Lutz, Jr. and R. F. Grossman, Polymer Modifiers and Additives, Chapters 8 and 11, Marcel Dekker, New York (2001). 108. K. A. Mazich, M. A. Samus, C. A. Smith, and G. Rossi, Macromolecules, 24, 2766 (1991). 109. S. H. Jafari and A. K. Gupta, J. Appl. Polym. Sci., 78, 962 (2000). 110. C. B. Bucknall and C. J. Page, J. Mater. Sci., 17, 808 (1982). 111. D. C. Clark, W. E. Baker, K. E. Russell, and R. A. Whitney, J. Polym. Sci. A: Polym. Chem., 38, 2456 (2000). 112. M. E. Fowler, H. Keskkula, and D. R. Paul, Polymer, 28, 1703 (1987). 113. F. Stricker, Y. Thomann, and R. Mulhaupt, J. Appl. Polym. Sci., 68, 1891 (1998). 114. X. Wang, C. Tzoganakis, and G. L. Rempel, J. Appl. Polym. Sci., 61, 1395 (1996). 115. J. A. W. van Dommelen, W. A. M. Brekelmans, and F. P. T. Baaijens, Mech. Mater., 35, 845 (2003).

PAGE 294

272 116. A. van der Wal, A. J. J. Verheul, and R. J. Gaymans, Polymer, 40, 6057 (1999). 117. M. Ishikawa, M. Sugimoto, and T. Inoune, J. Appl. Polym. Sci., 62, 1495 (1996). 118. M. Zhang, Y. Liu, X. Zhang, J. Gao, F. Huang, Z. Song, G. Wei, and J. Qiao, Polymer, 43, 5133 (2002). 119. S. Wu, Polymer, 26, 1855 (1985). 120. S. Wu, J. Appl. Polym. Sci., 35, 549 (1988). 121. S. Wu, J. Polym. Sci. B.: Polym. Phys., 21, 699 (1983). 122. M. H. Ha, B. K. Kim, and E. Y. Kim, J. Appl. Polym. Sci., 93, 179 (2004). 123. F. C. Stehling, T. Huff, C. S. Speed, and G. Wissler, J. Appl. Polym. Sci., 26, 2693 (1981). 124. W. Grellmann, S. Seidler, K. Jung, and I. Kotter, J. Appl. Polym. Sci., 79, 2317 (2001). 125. Z. Bartczak, J. Macromol. Sci., B41, 1205 (2002). 126. O. K. Muratoglu, A. S. Argon, and R. E. Cohen, Polymer, 36, 921 (1995). 127. B. Z. Jang, D. R. Uhlmann, and J. B. Vander Sande, J. Appl. Polym Sci., 30, 2485 (1985). 128. W. Wang, Q. Wu, and B. Qu, Polym. Eng. Sci., 43, 1798 (2003). 129. S. Hillmansen and R. N. Haward, Polymer, 42, 9301 (2001). 130. A. van der Wal, and R. J. Gaymans, Polymer, 40, 6045 (1999). 131. I. Brough, R. N. Haward, G. Healey, and A. Wood, Polymer, 45, 3115 (2004). 132. R. Segula, E. Staniek, B. Escaig, and B. Fillon, J. Appl. Polym. Sci., 71, 1873 (1999). 133. A. C. Steenbrink, H. Janik, and R. J. Gaymans, J. Mater. Sci., 32, 5505 (1997). 134. J. Karger-Kocsis, Polym. Eng. Sci., 36, 203 (1996). 135. J. T. Dickenson, L. C. Jensen, S. C. Langford, and R. P. Dion, J. Polym. Sci B: Polym. Phys., 32, 779 (1994). 136. T. Inoue, J. Appl. Polym. Sci., 54, 723 (1994).

PAGE 295

273 137. E. Borsig, A. Fiedlerova, and M. Lazar, J. Macromol. Sci.Chem., A16, 513 (1981). 138. E. Niki and Y. Kamiya, J. Org. Chem., 38, 1403 (1973). 139. N.-J. Huang and D. C. Sundberg, J. Polym. Sci. A: Polym. Chem., 33, 2533 (1995). 140. N.-J. Huang and D. C. Sundberg, J. Polym. Sci. A: Polym. Chem., 33, 2551 (1995). 141. N.-J. Huang and D. C. Sundberg, J. Polym. Sci. A: Polym. Chem., 33, 2571 (1995). 142. N.-J. Huang and D. C. Sundberg, J. Polym. Sci. A: Polym. Chem., 33, 2587 (1995). 143. X. Zhang and W. E. Baker, J. Appl. Polym. Sci., 67, 427 (1998). 144. P. A. Callais and R. T. Kazmierczak, SPE ANTEC, 1, 1368 (1989). 145. Atofina Chemicals, Inc., Organic Peroxi des Product Bulletin Dialkyl Peroxides, March 2001, http://www.atofinachemicals.com Last accessed January 6th, 2004. 146. C. E. Carraher, Jr., Polymer Chemistry, 5th Ed., Marcel Dekker New York (2000). 147. P. A. Callais and R. T. Kazmierczak, SPE ANTEC, 1, 1921 (1990). 148. J. Cha and J. L. White, Polym. Eng. Sci., 41, 1238 (2001). 149. M. Lazar, A. Kleinova, A., Fiedlerova, I. Janigova, and E. Borsig, J. Polym. Sci.: A: Polym. Chem., 42, 675 (2004). 150. P. Dokolas, M. Looney, S. Musgrave, S. Poon, and D. H. Solomon, Polymer, 41, 3137 (2000). 151. G. H. Hu, Y. J. Sun, and M. Lambla, Polym. Eng. Sci., 36, 676 (1996). 152. W. Zhou and S. Zhu, Ind. Eng. Chem. Res., 36, 1130 (1997). 153. J. Fossey, D. Lefort, and J. Sorba, (Eds.), Free Radicals in Organic Chemistry, John Wiley & Sons. New York (1995). 154. D. W. Mead, J. Appl. Polym. Sci., 57, 151, (1995). 155. K. E. Russell, Prog. Poylm. Sci., 27, 1007 (2002) and all references therein. 156. S. Camara, B. C. Gilbert, R. J. Meier, M. van Duin, and A. C. Whitwood, Org. Biomol. Chem., 1, 1181 (2003). 157. M. P. Stevens, Polymer Chemistry: An Introduction, 3rd Ed., Oxford, New York (1999).

PAGE 296

274 158. W. A. Pryor, Free Radicals, McGraw-Hill, New York (1966). 159. F. M. Lewis, C. Walling, W. Cummings E. R. Briggs, and F. R. Mayo, J. Am. Chem. Soc., 70, 1515 (1948). 160. F. M. Lewis, C. Walling, W. Cummings E. R. Briggs, and F. R. Mayo, J. Am. Chem. Soc., 70, 1529 (1948). 161. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, Inc., New York (1953). 162. J. Bandrup and E. H. Immergut (Eds.), Polymer Handbook, 3rd Ed., WileyInterscience, New York, (1989); (a) K. C. Be rger and G. Brandrup, pp. 81II, (b) R. Z. Greenley, pp. 153, (c) R. Z. Greenley, pp.267. 163. A. Kannurpatti, J. W. Anseth, and C. N. Bowman, Polymer, 39, 2507 (1998). 164. S. W. Benson and A. M. North, J. Am. Chem. Soc., 84, 935 (1962). 165. G. T. Russell, D. H. Napper, and R. G. Gilbert, Macromolecules, 21, 2133 (1988). 166. K. E. Oliphant, K. E., Russell, and W. E. Baker, Polymer, 36, 1597 (1995). 167. C. Kiparissides amd D. Achilias, J. Appl. Polym. Sci., 35, 1303 (1988). 168. A. Husain and A. E. Hamielec, J. Appl. Polym. Sci., 22, 1207 (1978). 169. G. Moad, E. Rizzardo, and D. H. Solomon, Macromolecules, 15, 909 (1982). 170. N. Sahloul and A. Penlidis, Adv. Polym. Tech., 23, 186 (2004). 171. E. J. Dietz and N. A. Peppas, Polymer, 38, 3767 (1997). 172. J. S. Young, A. Kannuroatti, and C. N. Bowman, Macromol. Chem. Phys., 199, 1043 (1998). 173. A. Thakur, K. Banthia, and B. R. Maiti, J. Appl. Polym. Sci., 58, 959 (1995). 174. Y. R. Luo, Handbook of Bond Dissociation Energies in Organic Compounds, CRC Press, Boca Raton, Fl (2003). 175. S. Fliszar, Atoms, Chemical Bonds, and Bond Dissociation Energies, SpringerVerlag, New York (1994). 176. A. A. Zavitsas, J. Chem. Education, (78) 417 (2001). 177. M. Ratzsch, H. Bucka, A. Hesse, and M. Arnold, J.M.S. Pure Appl. Chem., A33, 913 (1996).

PAGE 297

275 178. G. G. Cameron and M. Y. Qureshi, J. Polym. Sci. A: Polym. Chem., 18, 2143 (1980). 179. M. Ratzsch, M. Arnold, E. Bors ig, H. Bucka, and N. Reichelt, Prog. Polym. Sci., 27, 1195 (2002) and all references therein. 180. J. B. Wong Shing, W. E. Baker, K. E. Russell, and R. A. Whitney, J. Polym. Sci A: Polym. Chem., 32, 1691 (1994). 181. Q. Yu and S. Zhu, Polymer, 40, 2961 (1999) 182. M. Saule, S. Navarre, O. Babot, W. Ma slow, L. Vartommen, and B. Maillard, Macromolecules, 36, 7469 (2003). 183. P. Dokolas, Q. G. Guanghua, and D. H. Solomon, J. Appl. Polym. Sci., 83, 898 (2002). 184. K. Siriinha and D. Meksawat, J. Appl. Polym. Sci., 93, 1179 (2004). 185. L. H. U. Andersson, B. Gu stafsson, and T. Hjertberg, Polymer, 45, 2577 (2004). 186. G. Odian and B. S. Bernstein, J. Polym. Sci. A: Polym. Chem., 2, 2835 (1964). 187. J. Yang, Z. Yao, D. Shi, H. Huang, Y. Wang, and J. Yin, J. Appl. Polym. Sci., 79, 535 (2001). 188. V. R. Landi and E. K. Easterbrook, Polym. Eng. Sci., 18, 1135 (1978). 189. C. Tzoganakis, J. Vlachopoulos, and A. E. Hamielec, Polym. Eng. Sci., 28, 170 (1988). 190. G. H. Hu, J. J. Flat, and M. Lambla, SPE ANTEC, 1, 2775 (1994). 191. D. Suwanda, R. Lew, and S. T. Balke, J. Appl. Polym. Sci., 35, 1019 (1988). 192. D. Suwanda, R. Lew, and S. T. Balke, J. Appl. Polym. Sci., 35, 1033 (1988). 193. F. Berzin and B. Vergnes, Polym. Eng. Sci., 40, 344 (2000). 194. V. J. Triacca, P. E. Gloor, S. Zhu, A. N. Hrymak, and A. E. Hamielec, Polym. Eng. Sci., 33, 445 (1993). 195. P. V. Krishna and R. H. Boyd, Macromolecules, 26, 679 (1993). 196. C. Rosales, R. Perera, M. Ichazo, J. Gon zales, H. Rojas, A. Sanchez, and A. D. Barrios, J. Appl. Polym. Sci., 70, 161 (1998). 197. H. G. Fritz, J.M.S. Pure Appl. Chem., A36, 1683 (1999).

PAGE 298

276 198. W. Heinen, S. W. Erkens, M. van Duin, and J. Lugtenburg, J. Polym. Sci. A: Polym. Chem., 37, 4368 (1999). 199. R. S. Auda, W. M. Davis, D. Y.-L. Chung, L. W. Flatley, Jr., B. G. Jones, D. A. White, and H. G. Woudboer, U. S. Patent 5,651,927 (1997). 200. S. K. Singh, S. P. Tambe, A. B. Samui, and D. Kumar, J. Appl. Polym. Sci., 93, 2802 (2004). 201. Y. Qing and X. Wenying, Polym. Eng. Sci., 34, 446 (1994). 202. J. Martinez, R. Benavides, C. Guerrero, and B. E. Reyes, Polym. Degrad. Stabil., 86, 129 (2004). 203. J. Yamauchi, A. Yamaoka, K. Ikemoto, and T. Matsui, J. Appl. Polym. Sci., 43, 1197 (1991). 204. B. Boutevin, J. J. Robin, N. Torres, and J. Casteil, Macromol. Chem. Phys., 203, 245 (2002). 205. G.-H. Hu, H. Li, L.-F. Feng, and L. A. Pessan, J. Appl. Polym. Sci., 88, 1799 (2003). 206. R. Rengarajan, M. Vicic, and S. Lee, Polymer, 30, 834 (1989). 207. L. Wild, J. A. Hinnenkamp, B.-H. Chang, B.-H., and W. W. Kiang, U. S. Patent 5,344,888 (1994). 208. S. W. Coe, U. S. Patent 6,426,389 (2002). 209. M. S. Rodriguez, P. N. Tutunjian, C. L. Edwards, P. K. Casey, C. E. Ash, C. C. Hwo, and P. K. Wong, U. S. Patent 6,448,333 (2002). 210. L. Wild, J. A. Hinnenkamp, B. H. Cha ng, and W. W. Kiang, U. S. Patent 2,344,888, (1994). 211. K. Nakashima, R. Muroi, M. Imai, and K. Sato, U. S. Patent 3,981,958 (1976). 212. F. Picchioni, J. G. P. Goossens M. van Duin, and P. Magusin, J. Appl. Polym. Sci., 89, 3279 (2003). 213. S. A. Michaels and H. J. Bixler, J. Polym. Sci., 50, 393-409 (1961). 214. F. Yoshii, K. Makuuchi, S. Kikukawa, T. Tanaka, J. Saitoh, and K. Koyama, J. Appl. Polym. Sci., 60, 617 (1996). 215. X. Yunshu, F. Yoshii, and K. Makuuchi, J. Macromol. Chem., A32, 1803 (1995).

PAGE 299

277 216. M. Lazar, L. Hrckova, E. Borsig, A. Marcincin, and N. Reichelt, J. Appl. Polym. Sci., 78, 886 (2000). 217. I. Chodak, M. Lazar, and M. Capla, J. Polym. Sci. A: Polym. Chem., 29, 581 (1991). 218. I. Chodak and E. Zimanyova, Eur. Polym. J., 20, 81 (1984). 219. E. Borsig, L. Hrchova, A. Fiedlerova, and M. Lazar, J.M.S.-Pure Appl. Chem., A35, 1313 (1998). 220. K. N. Ludwig and R. B. Moore, SPE ANTEC, 1, 1844 (1999). 221. D. Graebling, Macromolecules, 35, 4602 (2002). 222. L. K. Yoon, C. H. Choi, and B. K. Kim, J. Appl. Polym. Sci., 56, 239 (1995). 223. R. Greco, G. Maglio, and P. V. Musto, J. Appl. Polym. Sci., 33, 2513 (1987). 224. A. Brydon, G. M. Burnett, and G. G. Cameron, J. Polym. Sci., 11, 3255 (1973). 225. A. Brydon, G. M. Burnett, and G. G. Cameron, J. Polym. Sci., 12, 1011 (1974). 226. N. G. Gaylord and M. K. Mishra, J. Polym. Sci. C: Polym. Lett., 21, 23 (1983). 227. G. Doney and K. Salsman, U. S. Patent 6,855,771 (2005). 228. Y. Li, X. M. Xie, and B. H. Guo, Polymer, 42, 3419 (2001). 229. D. Braun, S. Richter, G. P. Hellmann, and M. Ratzch, J. Appl. Polym. Sci., 68, 2019 (1998). 230. I. H. Do, L. K. Yoon, B. K. Kim, and H. M. Jeong, Eur. Polym. J., 32, 1387 (1996). 231. J. C. Randall, Jr. and T. Huff U. S. Patent 5,066,723 (1991). 232. B. I. Chaudhary, T. H. Ho, T. Oswald, a nd J. DellaMea, U. S. Patent Application 20020151609 (2002). 233. W. Y. Chiu and S. J. Fang, J. Appl. Polym. Sci., 30, 1473 (1985). 234. J. W. Teh, A. Rudin, and J. C. Keung, Adv. Polym. Techn., 13, 1 (1994). 235. T. Murayama, Dynamic Mechanical Analysis of Polymeric Material, Materials Science monographs; 1, Elsevier, The Netherlands (1978). 236. C. D. Lee, SPE ANTEC, 1, 2035 (2005).

PAGE 300

278 237. H. G. Heck, D. R. Parikh, L. B. Weaver, J. R. Bethea, L. A. Meiske, and M. F. Martin, U. S. Patent 6,506,842 (2003). 238. D. Graebling, M. Lambla, and H. Wautier, J. Appl. Polym. Sci., 66, 809 (1997). 239. K. L. Walton, U. S. Patent 6,548,600 (2003). 240. E. Borsig, A. Fiedlerova, L. Rychla, M. Lazar, M Ratzsch, and G. Haudel, J. Appl. Polym. Sci., 37, 467 (1989). 241. T. Inoue and T. Suzuki, J. Appl. Polym. Sci., 56, 1113 (1995). 242. T. Inoue and T. Suzuki, J. Appl. Polym. Sci., 59, 1443 (1996). 243. J. Yun, R. Patel, and D. C. Worley, SPE ANTEC, 3, 4172 (2004). 244. J. Teng, J. U. Otaigbe, and E. Taylor, SPE ANTEC, 1, 1362 (2003). 245. J. Teng, J. U. Otaigbe, and E. P. Taylor, Polym. Eng. Sci., 4, 648 (2004). 246. G.-H. Hu, H. Carteir, and C. Plummer, Macromolecules, 32, 4713 (1999). 247. V. Flaris and W. E. Baker, SPE ANTEC, 3164 (1995). 248. P. Lagardere and W. E. Baker, SPE ANTEC, 2, 2433 (1998). 249. B. S. Kim and S. C. Kim, J. Appl. Polym. Sci., 69, 1307 (1998). 250. J. W. Teh and A. Rudin, Polym. Eng. Sci., 22, 1678 (1992). 251. T. Vestberg, V. Lonnberg, T. Vainio, K. Hanhi, and H. Jukarainen, U. S. Patent 5,981,665 (1999). 252. A. van der Wal and R. J. Gaymans, Polymer, 40, 6067 (1999). 253. S. Havriliak, C. A. Cruz, and S. E. Slavin, Polym. Eng. Sci., 36, 2327 (1996). 254. S. Jose, A. S. Aprem, B. Francis, M. C. Chandy, P. Werner, V. Alstaedt, and S. Thomas, Eur. Polym. J., 40, 2105 (2004). 255. J. Karger-Kocsis, A. Kallo, A. Szafner, G. Bodor, and Z. Senyei, Polymer, 20, 37 (1979). 256. P. Phinyocheep, F. H. Axtell, and T. Laosee, J. Appl. Polym. Sci., 86, 148 (2002). 257. E. Martuscelli, C. Silvestre, and G. Abate, Polymer, 23, 229 (1984). 258. H.-W. Xiao, S.-Q. Huang, T. Jiang, and S.-Y. Cheng, J. Appl. Polym. Sci., 83, 315 (2002).

PAGE 301

279 259. A. Galeski and E. Piorkowska, J. Polym. Sci. B: Polym. Phys., 21, 1313 (1983). 260. M. M. Winram, D. Grubb, and A. Keller, J. Mater. Sci., 13, 791 (1978) 261. X.-Q. Zhou and J. N. Hay, Polymer, 34, 4710 (1993). 262. R. E. Robertson and D. R. Paul, J. Polym. Sci., 17, 2579 (1973). 263. Z. Bartczak, A. Galeski, and M. Pracella, Polymer, 27, 537 (1986). 264. S. George, K. T. Varughese, and S. Thomas, Polymer, 41, 5485 (2000). 265. M. Gahleitner, J. Wolfschwenger, C. B achner, K. Bernreitner, and W. Neibl, J. Appl. Polym. Sci., 61, 649 (1996). 266. D. S. Li, H. Garmestani, R. G. Alamo, and S. R. Kalidindi, Polymer, 44, 5355 (2003). 267. G. Machado, E. L. G. Denardin, E. J. Kina st, M. C. Goncalves, M. A. de Luca, S. R. Teixeira, and D. Samios, Eur. Polym. J., 41, 129 (2005). 268. P. J. Flory and D. Y. Yoon, Nature, 272, 226 (1978). 269. M. A. Kennedy, A. J. Peacock, and L. Mandelkern, Macromolecules, 27, 5297 (1994). 270. V. Petraccone, I. C. Sanchez, and R. S. Stein, J. Polym. Sci. B: Polym. Phys., 13, 1991 (1975). 271. L. E. Nielsen and F. E. Stockton, J. Polym. Sci. A: Polym. Chem., 1, 1995 (1963). 272. B. A. G. Schrauwen, R. P. Janssen, L. E. Govaert, and H. E. H. Meijer, Macromolecules, 37, 6069 (2004). 273. P. T. S. Dijkstra, D. J. Van Dijk, and J. Huetink, Polym. Eng. Sci., 42, 152 (2002). 274. K. Premphet and W. Paecharoenchai, J. Appl. Polym. Sci., 82, 2140 (2001). 275. K. Premphet and W. Paecharoenchai, J. Appl. Polym. Sci., 85, 2412 (2002). 276. D. W. van der Meer, B. Pukanszky, B., and G. J. Vancso, J. Macomol. Sci., B41, 1105 (2002). 277. R. Seguela and F. Rietsch, Polymer, 27, 532 (1986). 278. T. A. Huy, R. Adhikari, T. Lupke S. Henning, and G. H. Michler, J. Polym. Sci. B: Polym. Phys., 42, 4478 (2004).

PAGE 302

280 279. C. Passingham, P. J. Hendra, M. E. A. Cudby, V. Zichy, and M. Weller, Eur. Polym. J., 26, 631 (1990). 280. S. Thomas and G. Groeninckx, Polymer, 40, 5799 (1999). 281. C. Grein, K. Bernreitner, and M. Gahleitner, J. Appl. Polym. Sci., 93, 1854 (2004). 282. C. Jourdan, C., Cavaille, and J. Perez, J. Polym. Sci. B, Polym. Phys., 27, 2361 (1989). 283. M. L. Cerrada, O. Prieto, J. M. Perena, R. Benavente, and E. Perez, Macromol. Symp., 198, 91 (2003). 284. Y. P. Khanna, E. A. Turi, T. J. Taylor, V. V. Vickroy, and R. F. Abbott, Macromolecules, 18, 1302 (1985). 285. J. Stepek and H. Daoust, Additives for Plastics, Springer-Verlag, New York (1983). 286. N. McCrum, J. Polym. Sci. B: Polym. Phys., 2, 495 (1964). 287. P. Jacoby, B. H. Bersted, W. J. Kissel, and C. E. Smith, J. Polym. Sci. B: Polym. Phys. 24, 461 (1986). 288. R. Popli, M. Glotin, and L. Mandelkern, J. Polym. Sci. B: Polym. Phys., 22, 407 (1984) and all references therin. 289. Z. Liang and H. L. Williams, J. Appl. Polym. Sci., 44, 699 (1992). 290. J. Li, R. A. Shanks, and Y. Long, Polymer, 42, 1941 (2001). 291. J. Li, R. A. Shanks, R. H. Olley, and G. R. Greenway, Polymer, 42, 7685 (2001). 292. M. Yamaguchi, H. Miyata, and K.-H. Nitta, J. Appl. Polym. Sci., 62, 87 (1996). 293. J. J. Rajasekaran, J. G. Curro, and J. D. Honeycutt, Macromolecules, 28, 6843 (1995). 294. J. Gao, D. Wang, M. Yu, and Z. Yao, J. Appl. Polym. Sci., 93, 1203 (2004). 295. A. L. N. D. Silva, M. C. G. Rocha, F. M. B. Coutinho, R. E. S. Bretas, and C. Scuracchio, J. Appl. Polym. Sci., 79, 1634 (2001). 296. M. Yamaguchi, H. Miyata, and K. H. Nitta, J. Polym. Sci. B: Polym. Phys., 35, 953 (1997). 297. W. Wenig and K. Meyer, Coll. Polym. Sci., 258, 1009 (1980). 298. Z. Bartczak and A. Galeski, Polymer, 27, 544 (1986).

PAGE 303

281 299. N. Kukaleva, F. Cser, M. Jollands, and E. Kosior, J. Appl. Polym. Sci., 77, 1591 (2000). 300. T. McNally, P. McShane, G. M. Nally, W. R. Murphy, M. Cook, and A. Miller, Polymer, 43, 3785 (2002). 301. A. Romankiewicz, J. Jurga, and T. Sterzynski, Macromol. Symp., 202, 281 (2003). 302. S. George, N. R. Neelakantan, K. T. Varughese, and S. Thomas, J. Polym. Sci. B: Polym. Phys., 35, 2309 (1997). 303. J. Heijboer, J. Polym. Sci. C: Polym. Lett., 16, 3755 (1968). 304. L. Woo, S. Westphal, and M. T. K. King, Polym. Eng. Sci., 34, 420 (1994). 305. W. G. Oakes and D. W. Robinson, J. Polym. Sci., 14, 505 (1954). 306. J. A. Roetling, J. Polym. Sci., 7, 303 (1966). 307. F. Avalos, M. A. Lopez-Machado, and M. Arroyo, Polymer, 25, 5681 (1996). 308. W. Wenig and M Asresahegn, Polym. Eng. Sci., 33, 877 (1993). 309. A. K. Gupta, S. K. Rana, and B. L. Deopura, J. Appl. Polym. Sci., 44, 719 (1992). 310. V. Choudhary, H. S. Varma, and I. K. Varma, Polymer, 32, 2534 (1991). 311. S. Laihonen, U. W. Gedde, P.-E. Werner, M. Westdahlo, P. Jaaskelainen, and J. Martinez-Salazar, Polymer, 38, 371 (1997). 312. S. Hoshino, E. Meinecke, J. Powers, R. S. Stein, and S. Newman, J. Polym. Sci. A:: Polym. Chem., 3, 3041 (1965). 313. R. Mehta and T. Kyu, J. Polym. Sci. B.: Polym. Phys., 42, 2892 (2004). 314. R. H. Olley, D. C. Bassett, Polymer, 23, 1707 (1982). 315. K. L. Naylor and P. J. Phillips, J. Polym. Sci. B: Polym. Phys., 21, 2011 (1983). 316. R. H. Olley and D. C. Bassett, Polymer, 30, 399 (1989). 317. M. Aboulfaraj, B. Ulrich, A. Dahoun, and C. GSell, Polymer, 34, 4817 (1993). 318. L. C. Sawyer and D. T. Grubb, Polymer Microscopy, 2nd Ed., Chapman & Hall, London, Chapter 4 (1996). 319. G. Socrates, Infrared and Raman Characteristic Group Frequencies, 3rd Ed., John Wiley & Sons, London (2001).

PAGE 304

282 320. J. Coates, Interpretation of Infrared Spectra, A Practical Approach, in Encyclopedia of Analytical Chemistry, R. A. Meyers (Ed.), John Wiley & Sons Ltd., Chichester, pp. 10815-10837 (2000). 321. B. D. Favis, C. Lavallee, and A. Derdouri, J. Mater. Sci., 27, 4211 (1992). 322. B. Fillon, A. Thierry, J. C. Wittmann, and B. Lotz, J. Polym. Sci. B: Polym. Phys., 31, 1407 (1993). 323. Q. Yuan, W. Jiang, and L. An, Coll. Polym. Sci., 282, 1236 (2004). 324. S. Vleeshouwers, Polymer, 38, 3213 (1997). 325. A. J. Lovinger, J. O. Chua, and C. G. Gryte, J. Polym. Sci. B: Polym. Phys., 15, 641 (1977). 326. K. Cho, D. N. Saheb, H. Tang, B.-I. Kang, J. Kim, and S.-S. Lee, Polymer, 44, 4053 (2003). 327. J. Garbarczyk, D. Paukszta, and S. Borysiak, J. Macrcomol. Sci., B41, 1267 (2002). 328. T. Pressly, N. Jivraj, C. Bos nyak, G. Young, and V. Tharp, SPE ANTEC, 1, 2020 (2004). 329. G.-X. Wei and H.-J. Sue, Polym. Eng. Sci., 40, 1979 (2000). 330. L. E. Nielsen and F. D. Stockton, J. Polym. Sci. A: Polym. Chem., 1, 1995 (1963). 331. F. J. B. Calleja, J. C. G. Ortega, and J. Martinez de Salazar, Polymer, 19, 1094 (1978). 332. T. Labour, G. Vigier, R. Seguela, C. Gauthier, G. Orange, and Y. Bomal, J. Polym. Sci. B: Polym. Phys., 40, 31 (2002). 333. J. Varga, J. Macromol. Sci., B41, 1121 (2002) and all references therein 334. Y. Cohen, and R. F. Saraf, Polymer, 42, 5865 (2001). 335. C. Grein, C. J. G. Plummer, H.-H. Kausch, Y. Germain, and P. Beguelin, Polymer, 43, 3279 (2002). 336. M. Raab, J. Kotek, J. Baldrian, and W. Grellmann, J. Appl. Polym. Sci., 69, 2255 (1998). 337. T. Labour, C. Gauthier, R. Segula, G. Vigier, Y. Bomal, and G. Orange, Polymer, 42, 7127 (2001). 338. T. Sterzynski, P. Calo, M. Lambla, and M. Thomas, Polym. Eng. Sci., 37, 1917 (1997).

PAGE 305

283 339. J. Kotek, M. Raab, J. Baldrian, and W. Grellmann, J. Appl. Polym. Sci., 85, 1174 (2002). 340. T. Labour, L. Ferry, C. Gauthier, P. Hajji, and G. Vigier, J. Appl. Polym. Sci., 74, 195 (1999). 341. K. J. Kim, Y. S. Ok, and B. K. Kim, Eur. Polym. J., 28, 1487 (1992). 342. R. A. Ozzetti, A. P. D. O. Filho, U. Schuchardt, and D. Mandelli, J. Appl. Polym. Sci., 85, 734 (2002). 343. A. Kotschkina and M. Grell, J. Polym. Sci. Part. C, 16, 3731 (1968). 344. D. Montezinos, B. G. Wells, and J. L. Burns, J. Appl. Polym. Sci., 23, 421 (1985). 345. S. Hosoda, and K. Kojima, Polym. Comm., 30, 83 (1989). 346. J. M. Kim, M. Lee, S. Kim, and S. Y. Kang, SPE ANTEC, 1, 1931 (2003). 347. J. X. Li and W. L. Cheung, J. Appl. Polym. Sci., 72, 1529 (1999). 348. G. M. Brown and J. H. Butler, Polymer, 38, 3937 (1997). 349. H. G. Haubruge, A. M. Jonas, and R. Legras, Polymer, 44, 3229 (2003). 350. A. Smit and A. Bezjak, Polymer, 22, 590 (1981). 351. Y. Kodera and B. J. McCoy, Energy & Fuels, 16, 119 (2002). 352. D. Dudic, D. Kostoski, V. Djokovic, and Z. Stojanovic, Polym. Deg. Stab., 67, 233 (2000). 353. H. C. Lau, S. N. Bhattacharya, and G. J. Field, Polym. Eng. Sci., 38, 1915 (1998). 354. S. Muke, I. Ivanov, N. Kao, and S. N. Bhattacharya, Polym. Int., 50, 515 (2001). 355. H. M. Laun and H. Schuch, J. Rheol. 33, 119 (1989). 356. J. J. Morad, SPE ANTEC, 1, 783 (1995). 357. A. Dharia and R. Folland, SPE ANTEC, 1, 2156 (2005). 358. M. Ratzsch, J.M.S. Pure Appl. Chem., A36, 1759 (1999). 359. C. He, S. Costeux, P. Wood-Adams, and J. M. Dealy, Polymer, 44, 7181 (2003). 360. P. Micic and S. N. Bhattacharya, Polym. Eng. Sci., 40, 1571 (2000). 361. M. Obadal, R. Cermak, and K. Stoklasa, SPE ANTEC, 1, 2151 (2005).

PAGE 306

284 362. J. M. Schultz, Polymer Crystallization, Oxford University Press, New York (2001). 363. S. V. Meille, D. R. Ferro, S. Bruckner, A. Lovinger, and F. J. Padden, Macromolecules, 27, 2615 (1994). 364. B. Lotz, J. C. Wittmann, and A. J. Lovinger, Polymer, 37, 4979 (1996). 365. J. Varga, J. Macromol. Sci., B41, 1121 (2002) and all references therein 366. F. J. Padden and H. D. Keith, J. Appl. Phys., 30, 1479 (1959). 367. A. Zambelli, P. Ammendola, A. Grassi, P. Longo, and A. Proto, Macromolecules, 19, 2703 (1986). 368. B. Wunderlich, Macromolecular Physics, 3, Academic Press, New York (1980). 369. Y. Jin, M. Rogunova, A. Hiltner, E. Baer, R. Nowacki, A. Galeski, and E. Piorkowska, J. Polym. Sci. B: Polym. Phys., 42, 3380 (2004). 370. F. L. Binsbergen and B. G. M. de Lange, Polymer, 9, 23 (1968). 371. A. J. Lovinger, J. Polym. Sci. B: Polym. Phys., 21, 97 (1983). 372. B. Lotz, and J. C. Wittmann, J. Polym. Sci. B: Polym. Phys., 24, 1541 (1986). 373. C. Mathieu, A. Thierry, J. C. Wittmann, and B. Lotz, J. Polym. Sci. B: Polym. Phys., 40, 2504 (2002). 374. J. J. Janimak, S. Z. D. Cheng, P. A. Giusti, and E. T. Hsieh, Macromolecules, 24, 2253 (1991). 375. V. Petraccone, B. Pirozzi, and S. V. Meille, Polymer, 27, 1665 (1986). 376. P. S. Dai, P. Cebe, M. Capel, R. G. Alamo, and L. Mandelkern, Macromolecules, 36, 4042 (2003). 377. J. X. Li and W. L. Cheung, Polymer, 39, 6935 (1998). 378. J. Varga, I. Mudra, and G. W. Ehrenstein, J. Appl. Polym. Sci., 74, 2357 (1999). 379. I. A. Al-Raheil, A. M. Qudah, and M. Al-Share, J. Appl. Polym. Sci., 67, 1259 (1998). 380. A. Turner-Jones, J. M. Aizlewood, and D. R. Beckett, Makromol. Chem., 75, 134 (1964). 381. J. Varga, Angew. Makromol. Chem., 104, 79 (1982).

PAGE 307

285 382. B. Fillon, B. Lotz, J. Thierry, and J. C. Wittmann, J. Polym. Sci. B: Poym. Phys., 31, 1395 (1993). 383. J. Karger-Kocsis, A. Kallo, A. Szafner, and G. Bodor, Polymer, 20, 37 (1979). 384. C. S. Ha and S. C. Kim, J. Appl. Polym. Sci., 35, 2211 (1988). 385. J. Garbarczyk and D. Paukszta, Coll. Polym. Sci., 263, 985 (1985). 386. L. DOrazio, R. Guarino, C. Mancarella, E. Martuscelli, and G. Cecchin, J. Appl. Polym. Sci., 75, 553 (2000). 387. R. H. Olley, A. M. Hodge, and D. C. Bassett, J. Polym. Sci. B: Polym. Phys., 17, 627 (1979). 388. F. Rybnikar, J. Appl. Polym. Sci., 27, 1479 (1982). 389. R. M. Gohil and P. J. Phillips, Polymer, 27, 1687 (1986). 390. H. F. Mark, N. Bikales, C. G. Overberger, and G. Merges, Encyclopedia of Polymer Science and Engineering, 2nd Ed., John Wiley & Sons Inc., New York, 482 (1986). 391. P. J. Flory, Trans. Faraday Soc., 51,848 (1955). 392. M. J. Richardson, P. J. Flory, and J. B. Jackson, Polymer., 4, 221 (1963). 393. Y. H. Kao and P. J. Phillips, Polymer, 27, 1669 (1986). 394. P. J. Phillips and Y. H. Kao, Polymer, 27, 1679 (1986). 395. R. M. Gohil and P. J. Phillips, Polymer, 27, 1699 (1986). 396. G. Teramoto, T. Oda, H. Saito, H. Sano, and Y. Fujita, J. Polym. Sci. B: Polym. Phys., 42, 2738 (2004). 397. M. Iijima and G. Strobl, Macromolecules, 33, 5204 (2000). 398. F. L. Binsbergen, J. Polym. Sci. B: Polym. Phys., 11, 117 (1973). 399. S. Yamazaki, M. Hikosaka, A. Toda, I. Wataoka, and F. Gu, Polymer, 43, 6585 (2002). 400. X. Zhu, D. Yan, and Y. Fang, J. Phys. Chem. B., 105, 12461 (2001). 401. D. R. Burfield, P. S. T. Loi, Y. Doi, and J. Mejzik, J. Appl. Polym. Sci., 41, 1095 (1990). 402. X. Wang, C. Tzoganakis, and G. L. Rempel, J. Appl. Polym. Sci., 61, 1395 (1996).

PAGE 308

286 403. D. A. Hemsley (Ed.), Applied Polymer Light Microscopy, Elsevier, New York (1989). 404. Z.-G. Wang, B. S. Hsiao, E. B. Sirota, P. Agarwal, and S. Srinivas, Macromolecules, 33, 978 (2000). 405. M. Imai, K. Mori, T. Mizukami, K. Kaji, and T. Kanaya, Polymer, 33, 4451 (1992). 406. N. V. Pogodina, S. K. Siddiquee, J. W. van Egmond, and H. H. Winter, Macromolecules, 32, 1167 (1999). 407. N. V. Pogodina and H. H. Winter, Macromolecules, 31, 8164 (1998). 408. F. L. Binsbergen, Polymer, 11, 253 (1970). 409. C. E. Locke and D. R. Paul, J. Appl. Polym. Sci., 17, 2597 (1973). 410. J. A. Sauer, L. J. Merrill, and A. E. Woodward, J. Appl. Polym. Sci., 58, 19 (1962). 411. S. Filipe, M. T. Cidade, M. Wilhelm, and J. M. Maia, Polymer, 45, 2367 (2004). 412. G.-H. Hu and H. Cartier, J. Appl. Polym. Sci., 71, 125 (1999). 413. K. Sirsinha and D. Meksawat, J. Appl. Polym. Sci., 93, 1179 (2004). 414. O. Darras and R. Seguela, J. Polym. Sci.: B, Polym. Phys., 31, 759 (1993). 415. P. Starck, Eur. Polym. J., 3, 339 (1997). 416. X. Guan and P. J. Phillips, SPE ANTEC, 2, 2076 (2004). 417. U. W. Gedde, Polymer Physics, Chapman and Hall, London, pp. 106 (1995). 418. J. M. Schultz, Polym. Eng. Sci., 24, 770 (1982). 419. A. Lustiger, C. N. Marzinsky, and R. R. Mueller, J. Polym. Sci. B: Polym. Phys., 36, 2047 (1998). 420. A. Smedberg, T. Hjertberg, and B. Gustafsson, Polymer, 3395 (2003). 421. M. D. Elkovitch, L. J. Lee, and D. L. Tomasko, Polym. Eng. Sci., 41, 2108 (2001). 422. V. Khan, C. Kwag, C. W. Manke, and E. Gulari, SPE ANTEC, 1, 1418 (1998). 423. A. Xue and C. Tzoganakis, SPE ANTEC, 1, 2276 (2003). 424. H. Li, L. J. Lee, and D. L. Tomasko, Ind. Eng. Chem. Res., 43, 509-514 (2004).

PAGE 309

287 425. Y.-P. Sun, Supercritical Fluid Technology in Materials Science and Engineering, Marcel Dekker, Inc., New York (2002). 426. A. I. Cooper, J. Mater. Chem., 10, 207 (2000). 427. J. J. Watkins and T. J. McCarthy, Macromolecules, 27, 4845 (1994). 428. J. A. Hyatt, J. Org. Chem., 49, 5097 (1984). 429. S. Lee, K. Soonjong, and A. Fouad, U. S. Patent 5,663,237 (1997). 430. B. M. Dorscht and C. Tzoganakis, J. Appl. Polym. Sci., 87, 1116 (2003). 431. Z. Dong, Z. Liu, B. Han, J. He, T. Jiang, and G. Yang, J. Mater. Chem., 12, 3565 (2002). 432. S. Beuermann, M. Buback, C. Isemer, I. Lacik, and A. Wahl, Macromolecules, 35, 3866 (2002). 433. Q. Dong and Y. Liu, J. Appl. Polym. Sci., 90, 853 (2003). 434. A. I. Cooper, Adv. Mater., 15, 1049 (2003). 435. S. Sicardi, L. Manna, and M. Banchero, J. Supercrit. Fluids, 17, 187 (2000). 436. S.-J. Park and S.-D. Yeo, Sep. Sci. Tech., 34, 101 (1999). 437. Univation Technologies, News and Information PE Markets, January 2003, http://www.univation.com. Last accessed July 13th, 2005. 438. D. C. Montgomery, Design and Analysis of Experiments, John Wiley & Sons, Inc., New York (2001). 439. A. Dean and D. Voss, Design and Analysis of Experiments, Springer-Verlag, New York (1999).

PAGE 310

288 BIOGRAPHICAL SKETCH Nathan Fraser Tortorella was born on January 9, 1979, in Attleboro, MA. He spent most of his childhood in Massac husetts but attended high sch ool in Kinnelon, NJ. His family moved to Florida during the summer of 1997, the same year he would start school at the University of Florida. He declared materials science and engineering as his major not long after entering the University and has flourished ever since. A Bachelor of Science degree in May 2002, Master of Science degree in May 2005, and Ph.D. in December 2005 within the department soon followed. The author enjoys Gator Athletics, fishi ng, mountain biking and hiking. He also likes to play an occasional pickup game of ba sketball. He married his wife, Michelle, on April 24, 2004, and they had their daught er, Katelynn Victoria, on June 2, 2005.