University of Massachusetts Amherst ScholarWorks@UMass Amherst
Doctoral Dissertations 1896 - February 2014
1-1-1995
The modification of epoxy/amine resin systems with poly(ether imide)/
Cynthia Dawn Athanasiou University of Massachusetts Amherst
Follow this and additional works at: https://scholarworks.umass.edu/dissertations_1
Recommended Citation Athanasiou, Cynthia Dawn, "The modification of epoxy/amine resin systems with poly(ether imide)/" (1995). Doctoral Dissertations 1896 - February 2014. 844. https://scholarworks.umass.edu/dissertations_1/844
This Open Access Dissertation is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Doctoral Dissertations 1896 - February 2014 by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].
THE MODIFICATION OF EPOXY/AMINE RESIN SYSTEMS WITH POLY(ETHER IMIDE)
A Dissertation Presented
by
CYNTHIA DAWN ATHANASIOU
Submitted to the Graduate School of the University of Massachusetts Amherst in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
September 1995
Polymer Science and Engineering THE MODIFICATION OF EPOXY/AMINE RESIN SYSTEMS WITH POLY(ETHER EVflDE)
A Dissertation Presented
by
CYNTHIA DAWN ATHANASIOU
Approved as to style and content by:
Richard J. Farris, Chair
Roger S. Porter, Member
C. Peter Lillya, Member
William J. MacKnight, Department Head Polymer Science and Engineering Department To William Athanasiou ACKNOWLEDGMENTS
It is difficult to know where to begin, since there are so many people who have contributed to this work in so many ways. First I wish to thank my advisor, Richard J.
Farris. During my years at UMass, Amherst he has given me the opportunity to learn many things, including how to deal with setbacks, design my own equipment and speak in public. He has taught me to question everything, even if at first glance it doesn't seem to need questioning. Next I would like to thank my other committee members. Dr. Roger
Porter has given me many suggestions and asked me some tough questions after reading
my work. Dr. C. Peter Lillya has helped me with the chemistry questions which arose and
led me to some of the equipment in his lab when necessary. His student, Erdem Cetim, has
helped me in making some PAEK oligomers and showing me how to run the VPO in his
lab.
I am indebted to all of the members of Dr. Farris' research group, both past and
present. They are a never-ending source of help and inspiration. A few people I need to
mention specifically are: Gregg Bennett for helping me get started in this area of research.
Christian Lietzau for being a great office-mate and asking many tough questions during
our group meetings. Joan Vrtis, Bob Fleming, Verna Lo, Mario Perez and Anthony
Karandinos who all started the same year I did and helped me get through the cumes.
Kapil Sheth who has been another great office-mate. Dave Macon who's made life in the
with I've lab a bit more interesting. And Meredith White, a fellow RPI graduate whom
shared good times and bad, and a few good bike rides to boot. All of the other students and professors in the Polymer Science and Engineering
Department also need to be thanked. It has been a great place to work and learn. The students all help one another out, and one can turn to any of the professors for guidance at any time. Of course, a great debt of gratitude is owed to the secretaries that have worked with me throughout my career at UMass. Eileen Besse has kept track of all of the records, acted as a go-between for me with the graduate school, and arranged the on-campus interviews. JoAnne Purdy has read all of my papers for me and made sure everything was running smoothly with the finances. Carol Moro kept the CUMIRP program running smoothly and helped me with all of my purchase orders while she was with the department. Norm Page and John Domian kept the electrical equipment up and running, and Lou Raboin helped when electron microscopy questions arose.
I want to thank my parents, Charlie and Jean Barnes, for all of their support. My
Dad always encouraged me to try everything and not to let anything stand in my way. He
taught me that I can do anything I set my mind to. I guess he was right. My mother. Dawn
Barnes, could not really understand why I wanted to go to graduate school, but I am sure
she is looking down fi-om Heaven now and smiling. Thank you for giving me my artistic talent and a good head on my shoulders, as well as the strength and stubbornness to
succeed. I also want to thank my brother, sister-in-law and niece, Craig, Linda and
Amanda Barnes for also being supportive and coming to my defense. It meant a lot.
for all of his And most importantly, I want to thank my husband, Will Athanasiou,
a lot more love, support and encouragement throughout this whole undertaking. He had
vi faith in me than I did sometimes. Without him this work would not have been possible.
This dissertation is dedicated to him.
vii ABSTRACT
THE MODIFICATION OF EPOXY/AMINE RESIN SYSTEMS WITH POLY(ETHER IMIDE
SEPTEMBER 1995
CYNTHIA DAWN ATHANASIOU, B.S., RENSSELAER POLYTECHNIC INSTITUTE
M.S., UNIVERSITY OF MASSACHUSETTS AMHERST
Ph.D., UNIVERSITY OF MASSACHUSETTS AMHERST
Directed by: Professor Richard J. Farris
A diglycidylether of bisphenol A, DGEBA, base epoxy resin has been modified with several different types of poly(ether iniide)s with the goal of increasing the toughness
of this high Tg epoxy system. Initially, Ultem® 1000-1000 fi-om GE was blended into
Epon 828/DDS. Phase separated systems resulted with a 2.5-fold increase in the fi-acture energy over the neat epoxy resin (676 jW vs. 265 J/m^). As seen under the SEM, the included phase pulled out of the continuous phase, indicating poor adhesion between the two phases. Next an amine terminated poly(ether imide), ATPEI, of Mn equal to 8600 was reacted into the same epoxy resin system. There was no increase in the fi-acture energy with the addition of the ATPEI except at the 40 weight percent loading level. At
this level, the fi-acture energy was comparable to that of the 40% Ultem®/E828/DDS
viii (700 J/m Phase separation ). was not observed in the TEM above 10 weight percent of the
ATPEI in the E828/DDS.
A mix of the Ultem® and the ATPEI were added to the epoxy resin with favorable results. Phase separation was present. Good adhesion between the phases was evident.
And at 20 weight percent, the mixed PEI modified E828/DDS had a higher fracture energy than either of the other two systems investigated previously in this study (409 J/m^ for the mix vs. 176 jWfor the ATPEI and 353 JW for the Ultem®).
Finally, an amine terminated poly(ether imide) with Mn equal to 25,500 was reacted into the same epoxy resin. The fracture energy at 20 weight percent was higher
than any of the systems studied previously (45 1 J/m^ vs. 409 jW for the mixed PEI modified system). Phase separated morphologies occurred at lower loading levels than
anticipated - co-continuous at 20% and phase inverted at 30%. In all cases the Tg's remained above 200°C.
ix TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS y ABSTRACT ^
LIST OF TABLES xiv
LIST OF FIGURES xvii
CHAPTER
L INTRODUCTION 1
1.1 Introduction 1 1.2 Background 3 1.3 Dissertation Overview 8 1.4 References 11
2. BLENDS OF POLY(ETHER IMIDE) WITH EPOXY RESINS: ULTEM® 1000-1000 MODIFIED EPON 828/DDS 17
2.1 Introduction 17 2.2 Materials 18 2.3 Property Determinations for Neat Ultem® 1000-1000 19
2.3.1 Glass Transition Temperature 20 2.3.2 Thermal Stability 23 2.3.3 Modulus 26 2.3.4 Fracture Energy 27 2.3.5 Coefficients of Thermal Expansion 28 2.3.6 Solubility in Epon 828 30
2.4 Fabrication of Plaques 31
2.4.1 Solventless Processing of Ultem® 1000-1000 in Epon 828/DDS 31 2.4.2 A DSC Study: Is Ultem® 1000-1000 Reacting with Epon 828? 32 2.4.3 Cure Studies 36
43 2.5 Mechanical Properties
43 2.5.1 Modulus 45 2.5.2 Fracture Energy
^° 2.6 Morphology
^° 2.6.1 Phase Separation 2.6.2 Fracture Surface
2.7 Thermal Analysis
X 2.7.1 Thermal Stability 62 2.7.2 Glass Transition Temperature 66
2.7.2.1 Differential Scanning Calorimetry 66 2.7.2.2 Dynamic Mechanical Analysis 68
2.7.4 Thermal Expansion Coefficient 71
2.8 Solubility in Various Solvents 79 2.9 Machinability 84 2.10 Conclusions 89 2.11 References 91
REACTIONS OF POLY(ETHER IMIDE) WITH EPOXY RESINS: EPON 828/DDS MODIFIED WITH AMINE TERMINATED POLY(ETHER IMIDE) 92
3.1 Introduction 92 3.2 Materials 93 3.3 Property Determinations for Neat ATPEI 94
3.3.1 Equivalent Weight Determination 94
3.3.1.1 Calculated from information received from GE 94 3.3.1.2 Determined using Gel Permeation Chromatography (GPC) 95 3.3.1.3 Determined using Vapor Phase Osmometry (VPO) 95
3.3.2 Thermal Stability 97
3.3.3 Glass Transition Temperature 100 3.3.4 Solubility in Epon 828 102
3.4 Determination of Stoichiometrically Optimum Amount of ATPEI to be Added to System ..102
3.4.1 Fabrication of Plaques: Assuming Varied Molecular Weights of ATPEI 103 3.4.2 Thermal Stability 104 3.4.3 Glass Transition Temperature 107
3.4.3.1 Differential Scanning Calorimetry, DSC 107 3.4.3.2 Dynamic Mechanical Analysis, DMA HO
3.4.4 Modulus 112 3.4.5 Fracture Energy 3.4.6 Summary & Conclusions
in Epon 828/DDS 117 3.5 Fabricating Plaques: Amine Terminated Poly(Ether Imide) Reacted
in E828/DDS 1 17 3.5. 1 SolvenUess Processing of 0-40% by Weight ATPEI 119 3.5.2 DSC Study: Is ATPEI Reacting with Epon 828?
122 3.6 Thermal Analysis
3.6.1 Thermal Stability
xi 3.6.2 Glass Transition Temperature 12' 3.6.3 Thermal Expansion CoefiBcients (CTE) 13^
3.7 Mechanical Properties 1^
3.7.1 Modulus 3.7.2 Fracture Energy |35
3.8 Morphology 13g
3.8.1 Phase Separation ^139 3.8.2 Fracture Surface 142
3.9 Solubility in Various Solvents 149 3.10 Machinability I53 3.11 Phase Separation Studies using Optical Microscopy 156 3.12 Summary and Conclusions 159 3.13 References 161
ULTEM®/ATPEI MODIFIED EPON 828/DDS RESIN SYSTEMS: COMBINATION OF HIGHER MN PEI WITH AMINE FUNCTIONALITY 163
4.1 Introduction 163 4.2 Preliminary Study: 20% PEI/E828/DDS, 1/3 Ultem® & 2/3 ATPEI 164
4.2. 1 Plaque Fabrication 164 4.2.2 Phase Separation Determination by TEM 166 4.2.3 Modulus 167 4.2.4 Fracture Energy 168 4.2.5 Thermal Analysis 171
4.2.5.1 Thermal Stability 171 4.2.5.2 Glass Transition Temperature 173 4.2.5.3 Linear Coefficients of Thermal Expansion (CTE) 175
4.2.6 Solubility in Various Solvents 177 4.2.7 Summary 181
4.3 Optimization of Ultem® & ATPEI Mix 182
4.3.1 Fabricating Plaques 182 4.3.2 Thermal Analysis 183
4.3.2.1 Thermal Stability 183 4.3.2.2 Glass Transition Temperature 184 4.3.2.3 Thermal Expansion Coefficient 185 4.3.2.4 Summary and Conclusion 186
1^^ 4.3.3 Mechanical Properties
4.3.3.1 Modulus 1^^ 4.3.3.2 Fracture Energy
xii 4.3.4 Morphology 191
4.4 Machinability 4.5 Summary and Conclusions 4.6 References
5. REACTION OF A HIGHER MOLECULAR WEIGHT AMINE TERMINATED POLY(ETHER IMIDE) WITH EPOXY RESINS: EPON 828/DDS MODIFIED WITH A HIGH MN ATPEI
5.1 Introduction 198 5.2 Property Determination for the Neat HMWA Powder 200
5.2.1 Thermal Stability 200 5.2.2 Glass Transition Temperature 203 5.2.3 Solubility in Epon 828 203
5.3 Fabricating Plaques 205 5.4 Thermal Analysis 206
5.4.1 Thermal Stability 206 5.4.2 Glass Transition Temperature 209 5.4.3 Thermal Expansion Coefficients 210
5.5 Mechanical Properties 212
5.5.1 Modulus 212 5.5.2 Fracture Energy 213
5.6 Morphology 216
5.6.1 Phase Separation 216 5.6.2 Fracture Surface 218
5.7 Machinability 222 5.8 Summary and Conclusions 224 5.9 References 226
227 6. CONCLUSIONS AND FUTURE WORK
6.1 Conclusions 227 6.2 Future Work 234 236 6.3 References
APPENDICES
A THREE POINT BEND TESTING FOR MODULUS DETERMINATION 238 TENSION TESTING...239 B. FRACTURE ENERGY DETERMINATION THROUGH COMPACT
BIBLIOGRAPHY
xiii LIST OF TABLES
Table Page
2.1 Tg's of Ultem® 1000-1000 by various analytical methods 20
2.2 Thermal stability of Ultem® 1000-1000 23
2.3 Mechanical properties for Ultem® 1000-1000 28
2.4 Coefficients of thermal expansion for the Ultem® 1000-1000 30
2.5 Tg's for various curing conditions of 10% Ultem® in Epon 828/DDS 37
2.6 Moduli of Epon 828/DDS epoxy resins modified with Ultem® 1000-1000 43
2.7 Critical stress intensity factors and fracture energies of the Ultem® modified E828/DDS systems 46
2.8 Thermal stability data for Epon 828/DDS resins modified with Ultem® 1000-1000 63
2.9 Glass transition temperatures for Ultem® modified epoxy systems as determined from DSC 68
2. 10 Tg's of Ultem® modified resin systems by both DSC and DMA 69
2. 1 1 Thermal expansion coefficients for a neat Epon 828/DDS epoxy resin system through the thickness and across the width of the sample 71
2. 12 Glassy and rubbery coefficients of thermal expansion for the epoxy resins modified with Ultem® 76
96 3.1 Results fi-om VPO analysis of standard and ATPEI
97 3.2 Thermal stability data for neat ATPEI
molecular 3.3 Results of the TGA analysis on the samples assuming differing weights of the ATPEI
modified resins with 3.4 Summary of Tg's as determined by DSC for the ATPEI varying assumed molecular weights
xiv 3.5 Summary of Tg's as determined by DMA for the ATPEI modified resins with varying assumed molecular weights 1 lo
3.6 Flexural moduli of the three resin systems with varying assumed Mn's 1 12
3.7 Kic & Gic information for 20% ATPEI/E828/DDS resins with differing assumed Mn's 114
3.8 Summary of all of the data presented in section 3.4 116
3.9 Summary of thermal stability data by TGA 127
3.10 DSC Tg data for the ATPEI/E828/DDS resin systems 128
3.11 Linear coefficients of thermal expansion for the ATPEI modified Epon 828/DDS systems 130
3.12 Modulus data for the ATPEI modified Epon 828/DDS systems 134
3.13 Critical stress intensity factors vs. amount of ATPEI in modified epoxy system 135
3.14 Fracture energy data for the ATPEI modified Epon 828/DDS epoxy resin system 137
4. 1 Critical stress intensity factor and fi-acture energy of the 20% mixed PEI/E828/DDS 169
4.2 TGA data for the mixed PEI/E828/DDS system 171
4.3 CTE data for 20% mixed PEI/E828/DDS 175
4.4 TGA data for the various mixed PEI's in E828/DDS 183
4.5 Tg's of the different mixture of PEI's with E828/DDS 184
185 4.6 CTE data for the various mixed PEI/E828/DDS samples
186 4.7 Modulus data for the various mixed PEI/E828/DDS samples
the different mixed 4 8 Critical stress intensity factors and fracture energies for PEI systems
200 5.1 Thermal stability data of HMWA
XV 5.2 TGA data for the HMWA modified E828/DDS systems 209
5.3 Glass transition temperatures of HMWA/E828/DDS systems 210
5.4 Glassy and rubbery CTE's for the HMWA modified E828/DDS 212
5.5 Moduli of the HMWA/E828/DDS systems 213
5.6 Critical stress intensity factor and fi-acture energy data fi-om compact tension test 214
xvi LIST OF FIGURES
Figure Page
2.1 Chemical structures of materials used in this research 19
2.2 DSC trace showing Tg of Ultem® 1000- 1000 21
2.3 DMA trace showing Tg of Ultem® 1000-1000 22
2.4 TGA trace showing the entire temperature range from room temperature to 750°C 24
2.5 TGA trace showing onset of 1% decomposition 25
2.6 Typical force/deflection curve for Ultem® 1000-1000 in three point bending. ... 27
2.7 TMA trace of neat Ultem® 1000-1000 showing CTE's 29
2.8 DSC trace of 35% Ultem®/E828. First heating 33
2.9 DSC trace of 65% Ultem®/E828. First heating 34
2.10 DSC trace of E828/DDS. First heating. Notice the large exothermic peak around 200°C 35
2.1 1 DSC thermogram for sample cured at 170°C for two hours 38
2. 12 DSC thermogram of sample cured for 170°C for 4 hrs 39
2.13 DSC thermogram of sample cured for 170°C for two hours and 190°C for an additional two hours 40
2.14 DSC thermogram of sample cured at 170°C for two hours and 210°C for an additional two hours
2.15 DSC thermogram of sample cured for two hours at 170°C and an additional two hours at 230°C
Epon 828/DDS 44 2. 16 Typical force/deflection curve for Ultem® modified
xvii 2.17 Modulus vs. weight % Ultem® blended into Epon 828/DDS epoxy resin systems
2.18 Critical stress intensity factor vs. weight % Ultem® blended into Epon 828/DDS 4-7
2. Fracture energy 19 vs. weight % Ultem® blended into Epon 828/DDS 47
2.20 micrograph TEM of 10% Ultem® in Epon 828/DDS. Note the phase separated morphology 49
2.21 TEM micrograph of 20% Ultem® in Epon 828/DDS. Note the phase separated morphology 49
2.22 TEM micrograph of 20% Ultem® in Epon 828/DDS at a higher magnification. Note the epoxy rich phase included in the poly(ether imide) rich inclusions 50
2.23 TEM micrograph of 30% Ultem® in Epon 828/DDS. Note the co-continuous morphology 50
2.24 TEM micrograph of 40% Ultem® in Epon 828/DDS. Note the phase inverted morphology 51
2.25 Side view of compact tension test specimen showing line where it was cut for SEM viewing 52
2.26 SEM micrograph of 10% Ultem® in Epon 828/DDS. Magnification = 6000X 53
2.27 SEM micrograph of 10% Ultem® in Epon 828/DDS. Magnification - 10,000X 54
2.28 SEM micrograph of 20% Ultem® in Epon 828/DDS. Magnification = lOOOX 55
2.29 SEM micrograph of 20% Ultem® in Epon 828/DDS. Magnification = 6,000X 56
2.30 SEM micrograph of 30% Ultem® in Epon 828/DDS. Magnification = lOOOX 57
2.31 SEM micrograph of 30% Ultem® in Epon 828/DDS. Magnification = 6000X 58
xviii 2.32 SEM micrograph of 40% Ultem® in Epon 828/DDS. Magnification = lOOOX 59
2.33 SEM micrograph of 40% Ultem® in Epon 828/DDS. Magnification = 10,000X 50
2.34 SEM micrograph of neat Ultem®. Magnification = lOOOX 61
2.35 TGA thermogram of 30% Ultem® in Epon 828/DDS showing 50% weight loss 54
2.36 TGA thermogram of 30%) Ultem® in Epon 828/DDS showing 1% weight loss 65
2.37 DSC thermogram of 20%) Ultem® in Epon 828/DDS showing Tg 67
2.38 DMA trace of 10% Ultem® in Epon 828/DDS showing only one Tg 70
2.39 TMA trace of neat Epon 828/DDS. First and second heatings through the thickness of the sample 72
2.40 TMA trace of neat Epon 828/DDS. Second heating through the thickness of the sample 73
2.41 TMA trace of neat Epon 828/DDS. Third and fourth heatings across the width of the sample 74
2.42 TMA trace of neat Epon 828/DDS. Fourth heating across the width of the sample 75
2.43 TMA trace of 30%) Ultem® in E828/DDS showing first and second heatings. . 77
2.44 TMA trace of 30% Ultem® in E828/DDS showing second heating 78
2.45 Coefficients of thermal expansion vs. weight % Ultem® in Ultem®/E828/DDS 79
2.46 Effect of water on Ultem® modified E828/DDS 81
2.47 Effect of methylene chloride on Ultem® modified E828/DDS 81
2.48 Effect of methyl ethyl ketone on Ultem® modified E828/DDS 82
2.49 Effect of JP5 Jet Fuel on Ultem® modified E828/DDS 82
xix 1
2.50 Effect of deicing fluid on Ultem® modified E828/DDS 83
2.51 Effect of Skydol hydraulic fluid on Ultem® modified E828/DDS 83
2.52 Photo of a drilled sample of neat epoxy 85
2.53 Photo of a drilled sample of neat Ultem® 86
2.54 Photo of a drilled sample of 30% Ultem in E828/DDS 87
2.55 Photo of a drilled sample of 40% Ultem in E828/DDS 88
3.1 Chemical structures of materials used in this research 93
3.2 TGA trace of neat ATPEI showing no onset of 50% weight loss 98
3.3 TGA trace of neat ATPEI showing onset of 1% weight loss 99
3.4 DSC Thermogram of ATPEI 101
3.5 TGA Trace of 20% ATPEI/E828/DDS assuming 8200 Mn of ATPEI showing onset of 50% weight loss 105
3.6 TGA Trace of 20% ATPEI/E828/DDS assuming 8200 Mn of ATPEI showing onset of 1% weight loss 106
3.7 DSC trace of 20% ATPEI/E828/DDS assuming 8200 Mn. Second heat data. . 109
3.8 DMA trace of 20% ATPEI/E828/DDS with assumed Mn of 8200 1 1
3.9 Typical force/deflection curve for the 20% ATPEI/E828/DDS resin system under three point bending 113
3.10 Critical stress intensity factor vs. assumed Mn of ATPEI in 20% ATPEI/E828/DDS 114
3.11 Fracture energy vs. assumed Mn of ATPEI in 20% ATPEI/E828/DDS 115
3.12 DSC Trace of ATPEI/E828 mix 120
3.13 DSC trace of ATPEI/E828 mix, second heating 121
3.14 TGA trace of 10% ATPEI/E828/DDS showing onset 50% weight loss 123
3.15 TGA trace of 10% ATPEI/E828/DDS showing onset of 1% weight loss 124
XX 3.16 TGA trace of 40% ATPEI/E828/DDS showing onset 50% weight loss 125
3.17 TGA trace of 40% ATPEL^828/DDS showing onset of 1% weight loss 126
3.18 Second heat DSC trace of 20% ATPEI/E828/DDS 129
3.19 TMA trace for 30% ATPEI in Epon 828/DDS showing glassy and rubbery ^^^'^ 131
3.20 CTE vs. % ATPEI for ATPEI/E828/DDS 132
3.21 typical A force/deflection curve for a three point bend test on the ATPEI modified resins I33
3.22 Modulus vs. amount of ATPEI or Ultem in the Epon 828/DDS resin system I34
3.23 Kic vs. amount and type of PEI used in modifying the E828/DDS resin system 136
3.24 Fracture energy vs. type and amount of poly(ether imide) incorporated in Epon 828/DDS I37
3.25 TEM micrograph of 10% ATPEI/E828/DDS 140
3.26 TEM micrograph of 20% ATPEI/E828/DDS 140
3.27 TEM micrograph of 30% ATPEI/E828/DDS 141
3.28 TEM micrograph of 40% ATPEI/E828/DDS 141
3.29 SEM micrograph of 10% ATPEI/E828/DDS fracture surface. 6000X magnification 143
3.30 SEM micrograph of 20% ATPEI/E828/DDS fracture surface. lOOOX magnification 144
3.31 SEM micrograph of 20% ATPEI/E828/DDS fracture surface. 6000X magnification 145
3.32 SEM micrograph of 30% ATPEI/E828/DDS fracture surface. 1500X magnification 146
xxi 3.33 SEM micrograph of 40% ATPEI/E828/DDS fracture surface. lOOOX magnification
3.34 SEM micrograph of 40% ATPEI/E828/DDS fracture surface. lOOOOX magnification
3.35 Effect of water on ATPEI modified E828/DDS 150
3.36 Effect of methylene chloride on ATPEI modified E828/DDS 150
3.37 Effect of methyl ethyl ketone on ATPEI modified E828/DDS 151
3.38 Effect of JP5 Jet Fuel on ATPEI modified E828/DDS 151
3.39 Effect of deicing fluid on ATPEI modified E828/DDS 152
3.40 Effect of Skydol hydraulic fluid on ATPEI modified E828/DDS 152
3.41 20% ATPEI/E828/DDS after drilling 154
3.42 40% ATPEI/E828/DDS after drilling 155
3 .43 Typical optical micrograph of a sample of ATPEI/E828/DDS after cure study 157
3.44 Typical optical micrograph of a sample of ATPEI/E828/DDS after cure study 158
4.1 TEM micrograph of 20% mixed PEI modified Epon 828/DDS 166
4.2 Modulus vs. type of PEI modifier at the 20% loading level for the 20% PEI/E828/DDS resin systems 168
4.3 Kic vs. type of modifier used at 20% loading level 169
4.4 Gic vs. type of modifier used at 20% loading level 170
4.5 TGA trace of 20% mixed PEI/E828/DDS showing onset of 1% weight loss.... 172
4.6 TGA trace of 20% mixed PEI/E828/DDS showing onset of 50% weight loss.. 172
4.7 Second heat DSC trace of the 20% mixed PEI/E828/DDS showing Tg 174
4.8 TMA trace of 20% mixed PEI/E828/DDS showing glassy and rubbery CTE's 176
xxii 4.9 Effect of water on the modified E828/DDS systems 178
4.10 Effect of methylene chloride on the modified E828/DDS systems 179
4.1 Effect of methyl 1 ethyl ketone on the modified E828/DDS systems 179
4. Effect of JP5 12 jet fuel on the modified E828/DDS systems 180
4.13 Effect of deicing fluid on the modified E828/DDS systems 180
4. Effect 14 of Slcydol hydraulic fluid on the modified E828/DDS systems 181
4.15 Modulus vs. amount of ATPEI added to mixed systems 187
4.16 Comparison of critical stress intensity factors with changing amounts of ATPEI in Mix 188
4. 17 Comparison of fracture energies with changing amounts of ATPEI in Mix .... 190
4.18 SEM micrograph of the fracture surface of 20% (33% Ultem®, 67% ATPEI)/E828/DDS 192
4.19 SEM micrograph of the fi-acture surface of 20% (25% Ultem®, 75% ATPEI)/E828/DDS 193
4.20 SEM micrograph of the fi-acture surface of 20% (50% Ultem®, 50% ATPEI)/E828/DDS 194
4.21 20% mixed PEI modified E828/DDS after drilling 195
5.1 TGA trace of HMWA showing onset of 1% weight loss 201
5.2 TGA trace of HMWA showing onset of 50% weight loss 202
5.3 Second heating DSC scan showing glass transition temperature 204
5.4 TGA trace of 20% HMWA/E828/DDS showing onset of 1% weight loss 207
5.5 TGA trace of 20% HMWA/E828/DDS showing onset of 50% weight loss 208
5.6 TMA trace of 20%. HMWA/E828/DDS showing the glassy and rubbery CTE's 211
xxiii 5.7 Critical stress intensity factor vs. type of modifier at 20% weight percent loading level 215
5.8 Fracture energy vs. type of modifier at 20% weight percent loading level 215
5.9 TEM micrograph of 20% HMWA/E828/DDS showing a co-continuous morphology 217
5.10 TEM micrograph of 30% HMWA/E828/DDS showing a phase inverted morphology 217
5.1 micrograph 1 SEM of fracture surface of 20% HMWA/E828/DDS magnified lOOOX 219
5.12 SEM micrograph of fracture surface of 20% HMWA/E828/DDS magnified 10,000X 220
5.13 SEM micrograph of fracture surface of 30% HMWA/E828/DDS magnified 5,000X 221
5.14 20% HMWA/E828/DDS after drilling 223
B.l Compact tension sample geometric definitions 239
B.2 Above Tg crack insertion apparatus 240
xxiv CHAPTER 1
INTRODUCTION
1.1 Introduction
Ideally, for a resin to be considered as a high performance material, it should meet certain criteria. These are: 1.) a high glass transition temperature (greater than 135°C),
2.) a large modulus (greater than 2 GPa), 3.) good adhesion to a variety of substrates,
4.) ease of processing, 5.) impervious to a variety of solvents, including water, and 6.) a large fracture energy.
In 1986 the worldwide market for composites and high performance materials was
20 million pounds'. Approximately half of this was epoxy resins. By the year 2000 this
number is expected to reach 40 million pounds. Epoxy resins constitute such a large percentage of this number because of their excellent adhesion to various substrates, their low shrinkage upon cure, and their ease of processing. Epoxy resins can be B-staged and
exhibit tack and drape. The final cure can then be completed at a later time. Epoxy resin properties can be tailored with respect to adhesion and glass transition temperature through the use of various curing agents to meet the needs of most applications. But due to their high crosslink density, epoxies are inherently brittle. They are also sensitive to a number of solvents, including water.
1 Because of these shortcomings, researchers have looked to other materials to fiU the high performance material gap. Aromatic thermoplastic materials have been used with varying degrees of success'"". They are more ductile and therefore tougher than their crosslinked counterparts. The semi-crystalline aromatic thermoplastics are impervious to a large number of solvents'' Thermoplastics exhibit excellent properties in composites when there is good adhesion between the fiber and the matrix^'^''''^ But they are very difficult
13 • to process with high melting , temperatures and/or high viscosities, sometimes requiring the use of solvents'^ These materials cannot be B-staged and are generally hard as a board at room temperature, causing difficulties in composite fabrication. The finished products can also exhibit time dependent properties, such as creep.
In previous work, it has been seen that thermosets and thermoplastics can be successfully combined to bring out the best properties of each systeml'7-20 At high enough loading levels of reactive thermoplastic in a thermoset matrix, the systems can exhibit toughness and solvent insensitivity properties approaching that of the neat thermoplastic, while maintaining the processability of the neat thermoset. If the
thermoplastic used has a lower glass transition temperature, Tg, than the epoxy, there is a
decrease in the Tg of the modified epoxy system. It is the goal of this research to incorporate the best properties of the thermoplastic into the epoxy resin while maintaining the processability and the high Tg.
2 1.2 Background
As has been stated above, epoxy resins are the most widely used matrix material for advanced composites applications. This is due in part to their excellent adhesion to a variety of substrates and their processability'. A delay in cure can be affected if desired to enhance the processing, i.e., B-staging, filament winding. Epoxy resins also have good mechanical and electrical properties, superior dimensional stability and good resistance to heat and chemical attack^'. Upon curing, no small molecule, such as water, is liberated.
This results in very low shrinkage, and no by-products are formed to create voids in the system or to act as plasticizers*'^\
There are a number of commercially available curing agents for epoxy resins. This allows the tailoring of the cured epoxy properties to meet application specific needs. For
high performance materials, the aromatic amines are generally better than the aliphatic^^.
Longer pot life, higher heat distortion temperatures and better chemical resistance in the cured resin can be attained using aromatic amines. Aromatic amines are less reactive,
allowing the resin to be "B" staged at room temperature but not to cure for months at that temperature. Heat must be added for a complete cure. 4,4'-diaminodiphenylsulfone has
been chosen for use in this project. It is an aromatic diamine with a melting temperature between 170-180°C. Very high heat distortion temperatures are possible. Strength properties are maintained to higher temperatures. This curing agent is generally used in laminated circuit boards for high temperature applications, prepreg, tooling and casting applications.
3 The toughening of thermosets, particularly epoxy/amine resins, is a subject that has received much attention. Thermosets have extremely good processing and adhesion properties, but can be very brittle. Attempts have been made to alleviate this problem by altering the crosslink density, the molecular weight between crosslinks, Mc, and the distribution of Mc through the use of different epoxy prepolymers and blends of
" prepolymers . and by varying the stoichiometry of the reactants^^"^^ While this can alter the thermal behavior of the systems, the glassy elastic tensile and fracture toughness characteristics are not appreciably changed^^.
It was thought that altering the crosslink density of the system would improve the solvent sensitivity of the epoxy/amine systems^^ The absorption of solvent acts as a plasticizer and reduces the bulk properties^^'^''"^^ It has been shown that the solvent sensitivity is not dependent on the crosslink density^^'^"*, but rather on the number of specific interaction sites available for hydrogen bonding.
Some of the first work in the area of toughening epoxy/amine systems has been
through rubber toughening . The first successful attempt has been through the use of a carboxyl terminated butadiene-acrylonitrile elastomer as an additive to the epoxy
" network . The elastomer is initially soluble in commercial epoxy resins, but forms a second rubbery phase upon cure of the material. Other soluble rubbers which have been used successfully to toughen epoxy/amine networks are: modified poly(n-butyl acrylate)'*^,
4 1 42 carboxyl carboxyl and hydroxyl terminated polybutadienes , amine terminated siloxanes , terminated poly(butadiene acrylonitrile) copolymers (CTBN) of varying Mn and acrylonitrile contents''^ and CTBN and solid rubber mixtures'^''*^
4 The fracture toughness of these systems is dependent upon the toughness of the
original epoxy, the particle size and volume fraction of the rubbery particles within the
resin, the interfacial bonding between the particles and the epoxy, and the properties of the
rubbery component'*^ The determining factors in the size of the particle and the interfacial
bonding are the relative compatibility of the modifier to the epoxy and the rate of reaction
between the epoxy and the rubbery phase. There are two assumptions with regard to the
failure mechanism in rubber toughened systems. One is that the failure energy is dissipated
in the plastic deformation in the epoxy matrix initiated by the voiding of the rubber. The
second is that the crack is bridged by the rubbery component, and as the crack propagates,
the strain energy in the fibrils is lost. Because of the large mismatch in the modulus and
glass transition temperature, the bulk properties are reduced. The overall result is an
increase in the toughness with the lower Tg epoxies, but an associated decrease in the
glass transition temperature and modulus There is little toughness increase with the
addition of rubber to high Tg epoxies^^.
Another attempt at toughening epoxies has been through the use of epoxidized vegetable oils^\ These oils were incorporated into a DGEBA epoxy resin and cured with a
diamine curing agent. The amine reacts with the epoxide groups of both the epoxy resin
and the epoxidized vegetable oil, creating good adhesion between the two phases upon
cure. Phase inversion occurs at around 30% vegetable oil content. The results show toughening comparable to other rubber modified systems, and low water absorption. But the Tg and Young's moduli are lower than the neat thermoset material.
5 More recently, researchers have been studying the thermoplastic toughening of
epoxy resins''''°'"-^\ The proper selection of thermoplastic alleviates the large mismatch
of the modulus and the glass transition temperature between the epoxy and the toughening
agent. The bulk properties are not affected as they are with the rubber toughened systems.
One attempt to toughen epoxy resins with thermoplastics has been through the use of
poly(ether sulfone) in triflmctional and tetrafunctional epoxy resins. The system phase
separates upon curing, but only a marginal increase in properties was reported", due to
poor interfacial adhesion between the two phases. More success was reported with the
blending of poly(ether imide)s in biRinctional, triflinctional and tetrafunctional epoxy
resins when the amount of thermoplastic incorporated was sufficient to cause the system
to phase invert to a thermoplastic rich continuous phase system^^'^". These systems were
created by mixing the components with the use of solvents.
The most recent work has been through the use of end functionalized
thermoplastic oligomers that are reacted into the epoxy matrix'^"^". This approach has led
to a dramatic improvement in the adhesion between the two phases, improving the
toughness characteristics. Initially the system is miscible, but phase separates on
curing'^^°'^'^^ When the system is phase inverted, the toughness values can approach that
of the neat thermoplastic materiaP°'^\ Initial work in this area was through the use of phenol terminated poly(aryl ether sulfone/^ leading to the development of amine terminated poly(aryl ether sulfone)^^'^^. This material, reacted into epoxy systems, resulted in toughening of the systems at low loading levels of reactive thermoplastic oligomer.
Amine terminated poly(aryl ether sulfone)s, of number average molecular weights less
6 than 5000, incorporated into commercial Heloxy 69 resorcinol-based diglycidyl ether
resulted in even tougher materials'^'''. In all cases, the highest degree of toughness
occurred when the thermoplastic/epoxy system reached the phase inverted morphology
with epoxy rich inclusions in a thermoplastic rich continuous phase.
G.R. Almen al. et define the criteria for the selection of thermoplastics to be used
as the toughening agents for thermoset materials as the following: "1.) The thermoplastic
must have a high Tg and high modulus; 2.) The polymer must have suitable reactive end
groups to interact with the thermoset resin; 3.) The thermoplastic must have a well defined
molecular weight distribution; 4.) The polymer must be compatible with the thermoset.""
In this way the thermoplastic will have the largest impact on the toughening of the
thermoset.
Prior work in this lab with amine terminated oligomers of poly(aryl ether ketone)s
incorporated into epoxy systems has been very successfiiP". These oligomers were reacted
into the epoxy resins via a solventless mixing and processed with conventional
thermosetting technology. The resulting materials were initially miscible but phase
separated upon curing. All maintained room temperature tack and drape. When sufficient
oligomeric modifier was added to induce phase inversion, the toughness characteristics
approached those of the neat thermoplastic material, 826 J/m vs. 260 J/m for the neat
epoxy resin. The modulus decreased only slightly fi'om that of the neat epoxy resin (2. 1 vs.
2.4 GPa). There was a corresponding decrease in the glass transition temperature due to the lower Tg of the thermoplastic with respect to the neat thermoset (156°C vs. 219°C).
7 Poly(ether imide), PEI, is an amorphous thermoplastic material with high strength, high modulus, good heat resistance, high Tg, good electrical properties and remains stable over a wide temperature range'^ Poly(ether imide) has been solvent blended with various epoxy systems to try to increase their toughness characteristics''''"'"""". It has been shown that adding PEI substantially increases the fracture toughness of the resin system. The mechanism is thought to be the ductile drawing of the PEl". The fracture energy is seen to correlate linearly with the amount of thermoplastic incorporation, but the values are all lower than would be predicted by the rule of mixtures'".
1.3 Dissertation Overview
The objective of this work is to develop a toughened epoxy/amine resin system that can be used in high performance composite matrix applications, while furthering the
understanding of the toughening mechanisms in thermoplastic toughened epoxy resins.
This was accomplished by toughening the epoxy resins with poly(ether imide). Poly(ether imide)s of varying molecular weights and end groups were added to the epoxy resin prior to cure. The mechanical, morphological and thermal properties of the various systems were investigated and structure/property relationships determined. Earlier work in our lab developed thermoplastic toughened epoxies with Tg's around 150°C. A goal of this
research is to increase the Tg to nearly 200°C.
8 Chapter 2 discusses the results of blending an unreactive poly(ether imide) of intermediate molecular weight into a commercial epoxy resin system. The properties of the
neat PEI resin are determined. Plaques - were made up of 0 40% PEI by weight in the epoxy resin and the mechanical properties determined. The morphology of the systems at these different conditions was determined and correlated to the mechanical properties to determine a structure/property relationship. Also investigated were the glass transition temperature, thermal stability, linear coefficients of thermal expansion, solubility in various solvents, and machinability.
Chapter 3 involves the reacting of a lower molecular weight amine terminated poly(ether imide) into the same epoxy network. Again the thermal, mechanical and morphological properties were determined and compared. They were also compared to those observed with the unreactive PEI modified epoxy resins. Since no phase separation was present, a study was undertaken to determine the cause.
In Chapter 4 a mixture of the previous two types of PEI modifier, the unreactive
PEI and the amine terminated PEI, was used as a modifier. All of the same tests were performed as in the previous chapters. The ratio of the reactive to unreactive PEI's used in the modifier was altered to determine the effect of the phase separation vs. the amine
termination and to see if there was an optimum ratio.
Chapter 5 examines the use of a higher molecular weight amine terminated poiy(ether imide) as the modifier for the epoxy/amine resin system. Only two levels of modification were examined due to lack of material, 20 and 30% by weight of the PEI in the epoxy resin. The results obtained with these systems were compared to all of those
9 previously collected for the other modified epoxy systems, and the efifect of the higher molecular weight as well as the amine termination discussed.
Chapter 6 summarizes the major conclusions of this dissertation. Suggestions for future work in this area are also presented.
Finally, several appendices are provided which discuss the more complicated experimental techniques used in most chapters of this dissertation.
10 '
1 .4 References
1. Ronald Bauer, S "Application of Epoxy Resins in Advanced Composites" 34th Int' SAMPE Sympos. ,1889-1 900, (1989) '
2. Steven Peake, Art Maranci, Diana Megna, "High Temperature Polyetherimide Composites", 34th Int. SAMPE Sympos., 1235-1246, (1989)
3. S.D. Wu, J.L. Hedrick, D.K. Mohanty, B.K. Carter, G.L. Wilkes and J.E. McGrath, "Synthesis and Characterization of Poly(Arylene Ether Ketone Sulfone) Copolymers" 31st Int. SAMPE Sympos., 933-944, (1986)
Chen-Chi 4. M. Ma, Jiann-Tsun Hu, Wen-Liang Liu, Hung-Chung Hsia, Biing - Yng Shieh and Rong-Shian Liu, "The Thermal, Rheological and Morphological Properties of Polyphenlyene Sulfide and Polyetherether Ketone Resins and Composites", 3 1st Int.
SAMPE Sympos. , 420-43 3,(1986)
5. D.C. Leach, F.N. Cogswell, E. Nield, "High Temperature Performance of Thermoplastic Aromatic Polymer Composites", 31st Int. SAMPE Sympos. 434-448 (1986)
6. R.M. Turner, "The Effect of the Thermoplastic Matrix on the Performance of Continuous Fibre Reinforced Composites", I9th Int. SAMPE Tech. Conf., 443-453, (1987)
7. A.R. Wedgewood, K.B. Su, and J. A. Narin, "Toughness Properties and Service
Performance of High Temperature Thermoplastics and Their Composites", 19th Int. SAMPE Tech. Conf., 454-463, (1987)
8. T.E. Attwood, P C. Dawson, J.L. Freeman, L.R.J. Hoy, J.B. Rose, and P A. Staniland, "Synthesis and Properties of Polyaryletherketones", Polymer, 22, 1097-1 103, (1981)
9. Rhymer B. Rigby, "Polyetheretherketone Peek", Polym. News, 9, 325-328, (1984)
10. Walter R. Bergenn and R.B. Rigby, "PES and PEEK: Tough Engineering
Thermoplastics", Chem. Eng. Pr., 81, 36-38, (1985)
11. P.M. Hergenrother, S.J. Havens and B.J. Jensen, "New Polyarylene Ethers", ISthlnt. SAMPE Tech. Conf., 454-461, (1986)
12. John Timothy Hartness, "Polyetheretherketone Matrix Composites", 1 4th Nat. SAMPE Tech. Conf, 26-37, (1982)
11 13. J.B. Cattanach, G. Guff and F.N. Cogswell, "The Processing of Thermoplastics Containing High Loadings of Long and Continuous Reinforcing Fibers" J ofPolvm ' ^ 6(1-4), 345-362, (1986) '
14. Daniel Leeser and Brad Banister, "Amorphous Thermoplastic Matrix Composites for New Applications", 2 1st Int. SAMPE Tech. Con/., 507-513, (1989)
15. F.N. Cogswell and D.C. Leach, "Continuous Fibre Reinforced Thermoplastic- A Change in the Rules for Composite Technology", Plastics and Rubber Proc andAppl ' 4(3), 271-276, (1984)
16. D.P. Jones, D.C. Leach and D.R. Moore, "Mechanical Properties of Poly(Ether-Ether- Ketone) for Engineering Applications", Polymer, 26, 1385-1393, (1985)
17. J.A. Cecere, J.L Hedrick. and J.E. McGrath, "Morphology and Properties of Amine Terminated Poly(Arylene Ether Ketone) and Poly(Arylene Ether Sulfone) Modified Epoxy Resin Systems", 3 1stint. SAMPE Sympos., 580-588, (1986)
18. H.G. Recker, "Toughened Thermosets for Damage Tolerant Carbon Fiber Reinforced Composites", 2 1stint. SAMPE Tech. Conf., 283-293, (1989)
19. G.S. Bennett, R.J. Farris and S.A. Thompson, "Amine-Terminated Poly(Aryl Ether Ketone)-Epoxy/Amine Resin Systems as Tough High Performance Materials", Polymer 32(9), 1633-1641,(1991)
20. Greggory S. Bennett, Ph.D. Thesis, University of Massachusetts, (1991)
21. John Gannon, "Epoxy" , Modern Plastics, (Mid-Oct), 129-131, (1991)
22. Rafael J. Perez, "Commercial Epoxy Resin Curing Agents", in: Epoxy Resin
Technology . Ed: Paul F. Bruins, Interscience Publishers, NY, (1968)
23. S.C. Misra, J.A. Manson and L.H. Sperling, "Effect of Cross-Link Density
Distribution on the Engineering Behavior of Epoxies", in: Epoxy Resin Chemistry , ed. Ronald S. Bauer, ACS Symposium Series 114, (1979)
24. V.B. Gupta, C. Brahathee Swaran, "Packing Density of Amine Cross-Linked Epoxy
Networks", J. Appl. Polym. Sci., 41, 2533-2535, (1990)
25. Shinkichi Murakami, Osamu Watanabe, Hiroshi Inoue, Mitsukazu Ochi, Takuya Shiraishi, Masaki Shimbo, "Fracture Behavior of Epoxy Resins with Different Crosslinking Structures", 34th Int. SAMPE Sympos., 2194-2205, (1989)
26. G.R. Palmese and R.L. McCuUough, "Effect of Epoxy-Amine Stoichiometry on Cured
Resin Material Properties", J. Appl. Polym. Sci., 46, 1863-1873, (1992)
12 27. Charles D. Wingard and Charles L. Beatty, "Crosslinking of an Epoxy with a Mixed Amine as a Function of Stoichiometry. II. Final Properties Via Dynamic Mechanical Spectroscopy", J. Appl. Polym. ScL, 41, 2539-2554, (1990)
28. Joseph D. Keenan, James C Seferis. and John T Quinlivan, "Effects of Moisture and Stoichiometry on the Dynamic Mechanical Properties of a High-Performance Structural Epoxy", J. Appl. Polym. Set, 24, 2375-2387, (1979)
29. E. Morel, V. Bellenger, M. Bocquet and J. Verdu, "Structure-Properties Relationships for Densely Cross-Linked Epoxide-Amine Systems Based on Epoxide or Amine Mixtures; Part 3. Elastic Properties", J. Mat. Set, 24, 69-75, (1989)
30. Charles M. Demonet, "Interaction of Moisture with Resin Matrix Composites" 21st SAMPE Tech. Conf. ,231-238,(1989)
31. Roger J. Morgan, James E. O'Neal and Dale L. Fanter, "The Effect of Moisture on the Physical and Mechanical Integrity of Epoxies", J. Mat. ScL, 15, 751-764, (1980)
32. Michael .J. Adamson, "Thermal Expansion and Swelling of Cured Epoxy Resin Used in Graphite/Epoxy Composite Materials", J. Mat. Sci., 15, 1736-1745, (1980)
33. Y. Diamant, G. Marom and L.J. Broutman, "The Effect of Network Structure on Moisture Absorption of Epoxy Resins", J. Appl. Polym. Sci., 26, 3015-3025, (1981)
34. V. Bellenger, J. Verdu and E. Morel, "Structure-Properties Relationships for Densely
Cross-Linked Epoxide-Amine Systems Based on Epoxide or Amine Mixtures; Part 2. Water Absorption and Diflaision", J. Mat. ScL, 24, 63-68, (1989)
35. C. Keith Riew and John K. Gillham (Eds), Rubber Modified Thermoset Resins . Advances in Chemistry Series 208, (1984)
36. J.N. Sultan and F.J. McGarry, "Effect of Rubber Particle Size on Deformation
Mechanisms in Glassy Epoxy", Polm. Eng. and ScL, 13(1), 29-34, (1973)
37. Z.N. Sanjana and J.H. Testa, "Toughened Epoxy Resins for Filament Winding", 30th Nat. SAMPE Sympos., 1221-1230, (1985)
38. A.J. Kinloch, S.J. Shaw, D.A. Tod and D.L. Hunston, "Deformation and Fracture
Behavior of a Rubber-Toughened Epoxy: 1 . Microstructure and Fracture Studies", Polymer, 24, 1341-1354, (1983)
39. A.J. Kinloch, S.J. Shaw and D.L. Hunston, "Deformation and Fracture Behavior of a
Rubber-Toughened Epoxy: 2. Failure Criteria", Polymer, 24, 1355-1363, (1983)
13 40. S.L. Kirschenbaum, S. Gazit and J.P. Bell, "An Alternative Liquid Rubber for Epoxy Resin Toughening: Improving Poly(n-butyl acrylate) Rubber-Epoxy Compatibility by Use ot Acrylomtnle and Acrylic Acid Copolymers and Terpolymers", Rubber Modified Thermoset Resins, Advances in Chemistry Series 208, (1984)
41. K. Dusek, F. Lednicky, S. Lunak, M. Mach and D. Duskova, "The Toughening of Epoxy Resins with Reactive Polybutadienes", Rubber Modified Thermoset Jle.^m^ Advances in Chemistry Series 208, (1984)
42. E M. Yorkgitis, C. Tran, N.S. Eiss, T. Y. Hu, I. Yilgor, G.L. Wilkes and J.E McGrath, "Siloxane Modifiers for Epoxy Resins", Rubber Modified Thermoset Ref;in
43. D. Verchere, H. Sautereau, J.P. Pascault, S.M. Moschiar, C.C. Riccardi and R.J.J. Williams, "Miscibility of Epoxy Monomers with Carboxyl-Terminated Butadiene- Acrylonitrile Random Copolymers", Polymer, 30, 107-1 15, (1989)
44. W.D. Bascom and R.Y. Ting, "The Fracture of an Epoxy Polymer Containing Elastomeric Modifiers", J. Mat. ScL, 16, 2657-2664, (1981)
45. M. Ochi and J.P. Bell, "Rubber-Modified Epoxy Resins Containing High Functionality Acrylic Elastomers", J. Appl. Polym. ^c/., 29, 1381-1391, (1984)
46. A.C. Garg, "Fracture Behavior of Toughened Epoxy Resins", 18th Int. SAMPE Tech. Con/., 682-693, (1986)
47. Sung C. Kim and Hugh R. Brown, "Impact Modified Epoxy Resin with Glassy Second Component", J. Mats. Sci., 22, 2589-2594, (1987)
48. R.S. Raghava, "Criteria to Evaluate Effectiveness of Impact Modifiers Used in Toughening of Brittle Polymers such as Epoxies", 29th Nat. SAMPE Sympos. 1384-1392, , (1984)
49. S.D. Sjoerdsma, "The Tough-Brittle Transition in Rubber Modified Polymers", Polym. Comm., 30, 106-108,(1989)
50. L.C. Chan, J.K. Gillham, A.J. Kinloch, S.J. Shaw, "Rubber Modified Epoxies:
Morphology, Transitions, and Mechanical Properties", Rubber Modified Epoxies . Am. Chem. Soc. Ser., 261, (1984)
51. Isabelle Frischinger and Stoil Dirlikov, "Toughening of Epoxy Resins by Epoxidized Vegetable Oils", Polym. Comm., 32(17), 536-537, (1991)
14 52. G.R. Almen, P. Mackenzie, V. Malhotra, R.K. Maskell, P.T. McGrail and M S Sefton, "Thermoset/Thermoplastic Alloys: An Approach to Tough, High Performance Resin Matrices", 20th Int. SAMPE Tech. Con/., 46-59, (1988)
53. M.S. Sefton, P.T. McGrail, J.A. Peacock, S.P. Wilkinson, R.A. Crick, M. Davies and G. Almen, "Semi-Interpenetrating Polymer Networks as a Route to Toughening of Epoxy Resm Matrix Composites", J 9th Int. SAMPE Tech. Con/., 700-710, (1987)
54. E. Martuscelli, P. Musto, G. Ragosta and G. Scarinzi, "A New Supertough Epoxy Formulation with High Crosslinking Density", Die Angewandte Makromolekulare Chemie, 204, 153-159, (1993)
55. G.R. Almen, R.M. Byrens, P.D. Mackenzie, R.K. Maskell, P.T. McGrail and M.S. Sefton, "977-A Family of New Toughened Epoxy Matrices", 34th Int. SAMPE Svmposy 259-270,(1989) ^
C.B. 56. Bucknall, and I.K. Partridge, "Phase Separation in Crosslinked Resins Containing Polymeric Modifiers", Poly. Eng. andSci., 26(1), 54-62, (1986)
57. Keizo Yamanaka and Takashi Inoue, "Structure Development in Epoxy Resin Modified with Poly(Ether Sulphone)", Polymer, 30, 662-667, (1989)
58. Clive B. Bucknall and Ivana K. Partridge, "Phase Separation in Epoxy Resins Containing Polyethersulphone", Polymer, 24, 639-644, (1983)
59. Clive B. Bucknall and Adrian H. Gilbert, "Toughening Tetrafiinctional Epoxy Resins Using Polyetherimide", Polymer, 30, 213-217, (1989)
60. Ronald S. Bauer, "Toughened High Performance Epoxy Resins: Modification with Thermoplastics. I.", 34th Int. SAMPE Sympos., 899-909, (1989)
61. S.A. Thompson, Ph.D. Thesis, University of Massachusetts, (1987)
62. J.L. Hedrick, M.J. Jurek, I. Yilgor, J.E. McGrath, "Chemical Modification of Matrix Resin Networks with Engineering Thermoplastics", ACS Polym. Preprints, 26(2), 293, (1985)
63. U.S. Patent number 4,656,207, "Epoxy Thermosets Having Improved Toughness", (1987)
64. U.S. Patent number 4,656,208, "Thermosetting Epoxy Resin Compositions and Thermosets Therefrom", (1987)
15 65. James L. Hedrick, I. Yilgor, Garth L. Wilkes, James E. McGrath "Chemical Modification of Matrix Resin Networks with Engineering Thermoplastics" Polym • Bull 13,201-208,(1985) > ^
66. F. Biddlestone, A.A. Goodwin, J.N. Hay and G.A.C. Mouledous, "The Relaxation Spectrum and Physical Aging of Polyetherimide", Polymer, 32(17), 31 19-3125, (1991)
67. A. Murakami, D. Saunders, K. Ooishi and T. Yoshiki, "Fracture Behavior of Thermoplastic Modified Epoxy Resins", J. Adhesion, 39, 227-242, (1992)
68. Douglas J. Hourston, Jaqueline M. Lane and Nigel A. MacBeath, "Toughening of Epoxy Resins with Thermoplastics. II. Tetraflinctional Epoxy Resin - Polyetherimide Blends", Polym. Int., 26, 17-21, (1991)
69. D J. Hourston and J.M. Lane, "The Toughening of Epoxy Resins with Thermoplastics 1. Triftinctional Epoxy Resin - Polyetherimide Blends", Polymer 33(7) 1379-1383 (1992)
70. Masaki Kimoto, "Microstructure and Mechanical Properties of CFRP Consisting of Epoxy-Polyetherimide Blend Matrix", Sen-! Gakkaishi, 49(2), 72-78, (1993)
16 CHAPTER 2
BLENDS OF POLY(ETHER IMIDE) WITH EPOXY RESINS:
ULTEM® 1000-1000 MODIFIED EPON 828/DDS
2. 1 Introduction
In this chapter, Ultem® 1000-1000, an unreactive poly(ether imide) produced by the General Electric Company, was blended into a commercial epoxy resin system. This chapter discusses the systems produced by blending these commercially available polymers. It has been noted in the first chapter that thermoplastic modification of epoxy resins generally results in an increase in the fi-acture energy of the ensuing resin systems
without the decrease in the bulk properties that is associated with rubber toughened systems'"'. The use of Ultem® as a modifier of several different types of epoxy systems
has been investigated previously^'". In previous work toughening did occur, but all of the materials were processed with the aid of solvents, usually methylene chloride. In this
work, all of the processing is accomplished without the use of solvents.
The first section investigates the thermal, mechanical and morphological properties
of the neat Ultem® to determine its potential for toughening the epoxy resin. Next, a study was done to ensure that the Ultem® was not reacting with the epoxy resin, but was actually being blended into the system. Amounts equal to 10 - 40% by weight of the
17 Ultem® were blended into the epoxy/amine resin system and the thermal, mechanical and morphological properties of the resulting resins determined. The solubility of the modified resins was determined in a variety of solvents, and a mention of the change in machinability with the addition of Ultem® is included. Finally, an analysis is made of the effect of the Ultem® resin on the epoxy/amine system, and structure/property relationships discussed.
2.2 Materials
Ultem® 1000-1000 was obtained fi-om the General Electric Company and blended into the commercially available diglycidyl ether of bisphenol A, DGEBA, based epoxy resin, Epon 828. The system was cured with a stoichiometric amount of 4,4'- diaminodiphenylsulfone (DDS). The Ultem® has a Mn of 17,000. The Epon 828 has a Mn of 380 with an epoxide equivalent weight of 188 g/mole. The DDS has an amine equivalent weight of 62 g/mole. This corresponds to a functionality of four, one DDS can
react with four epoxides. The chemical structures are shown in Figure 2.1.
18 PoIy(ether imide)
Figure 2. 1 Chemical structures of materials used in this research.
2.3 Property Determinations for Neat Ultem® 1000-1000
Before any modification could be efifected, it was necessary to determine the
thermal and mechanical properties of the neat Ultem® resin.
19 2.3.1 Glass Transition Temperature
The glass transition temperature, Tg, was determined for the neat Ultem® 1000-
1000 resin and compared to that reported by the General Electric Company. The Tg was
examined by Differential Scanning Calorimetry (DSC) using a DuPont Instruments
DSC 2910 and the data analyzed using a DuPont Instruments Thermal Analyst 2000. The
Tg was taken as the inflection point of the endothermic transition from the third heat data.
The Tg was also determined using Dynamic Mechanical Analysis (DMA). A DuPont
Instruments DMA 983 was used for the testing and the data analyzed using the DuPont
Instruments Thermal Analyst 2000. The Tg was determined as the peak in tan 6. Copies of the thermograms are in Figures 2.2 & 2.3. The results are reported below in Table 2.1.
Table 2.1 Tg's of Ultem® 1000-1000 by various analytical methods
Tg (°C) - as reported Tg PC) - DSC Tg m - DMA 217 217 211
From this chart it can be seen that the values are all within 5% of each other and the Tg of
the material is taken as the reported value of 2 1 7°C for the remainder of this research.
20 21 H
o o I— o o @ §
o H00
2
22 2.3.2 Thermal Stability
The thermal stability of the Ultem® 1000-1000 was measured to ensure no
degradation would occur during the processing and curing of the modified epoxy resins.
The onset of and 1% 50% weight loss was ascertained using Thermal Gravimetric
Analysis (TGA) under nitrogen with a DuPont Instruments TGA 2950. The samples were heated at a rate of 20°C/min. to 750°C, and the onset of 1% weight loss determined. The
TGA traces are shown in Figures 2.4 & 2.5 and the results listed in Table 2.2. The weight loss did not reach 50% at temperatures below 750°C.
Table 2.2 Thermal stability of Ultem® 1000-1000
Onset 1% Wei2ht Loss rC) Onset 50% Weieht Loss (°C) 469 N/A
As can be seen fi-om this data, the Ultem® 1000-1000 is stable over a wide
temperature range and will not degrade during the processing at 190°C, or the curing at a
final temperature of 230°C. Because of the great thermal stability of the Ultem®, it is
23 24 25 envisioned that a higher degree of thermal stability can be imparted to the system with the addition of Ultem®.
2.3.3 Modulus
The flexural modulus of the Ultem® 1000-1000 was determined as outlined in
Appendix A. The material was used in the as received form of a bar with dimensions of
x 63mm 12.5mm x 3mm. The result is given below, and a representative force/deflection curve shown in Figure 2.6.
Flexural Modulus (GPa): 2.7
As can be seen from the curve, there is no breakage of the bar at 5% deflection and
the modulus is relatively high. For isotropic materials the flexural modulus is equal to the
elastic modulus. Because the Ultem® is an isotropic material, this value of the flexural
modulus, 2.7 GPa, will be used in the calculations of the fracture energy where the value
of the elastic modulus is needed.
26 Force/Deflection Curve for Ultem 1000-1000
300 -r
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 Deflection [m]
Figure 2.6 Typical force/deflection curve for Ultem® 1000-1000 in three point bending.
2.3.4 Fracture Energy
The fracture energy is used as the measure of the toughness of the materials
throughout this dissertation. The fracture energy is determined from a miniature compact tension test as outlined in Appendix B. The samples were loaded at a rate of 0.05cm/min.
until the crack propagated. The critical load for crack propagation was noted and used to
calculate the critical stress intensity factor, Kic, and fracture energy, Gic. The results are presented in Table 2.3. Table 2.3 Mechanical properties for Ultem® 1000-1000
Kir mWm'"] Gir Wn^^ 3.1 3530
These are numbers large as will be evidenced when compared to the critical stress intensity factor and fracture energy of the neat epoxy resin in the next sections.
2.3.5 Coefficients of Thermal Expansion
The linear coefficients of thermal expansion, CTE's, were determined on the neat
Ultem® 1000-1000 bar using a DuPont Instruments TMA2940 Thermal Mechanical
Analyzer (TMA). The samples were heated from room temperature to 275°C at 5°C/min.
and the glassy and rubbery CTE's taken from the second heat data. The glassy CTE is determined from the data collected between 100 and 150°C, and the rubbery CTE's
measured between 237 and 250°C. The TMA scan is shown in Figure 2.7 and the results shown in Table 2.4. These numbers are compared to those of the neat epoxy resin and the modified resins in section 2.6.4.
28 o cu
29 Table 2.4 Coefficients of thermal expansion for the Ultem® 1000-1000
Glassy CTE (^m/m°r) Rubbery CTE (uni/m°0 25 499
2.3.6 Solubility in Epon 828
Finally, it was necessary to ascertain the solubility of the Ultem® 1000-1000 pellets in the Epon 828 epoxy resin. 21.6 g of the epoxy resin was heated in a 50ml beaker to 190°C, the processing temperature used for sample preparation. 2.07 g of the Ultem® pellets were added, an amount equal to nearly 10% of the weight of the epoxy resin. The combination was allowed to stir at \90°C for two hours. The resultant mixture in the
beaker was a clear brown liquid. This indicated that the Ultem® pellets had dissolved in
the epoxy resin. At this point it was decided that the Ultem® 1000-1000 pellets would be a suitable material for the attempted toughening of the Epon 828/DDS resin system without the aid of solyents.
30 2.4 Fabrication of Plagues
2.4.1 Solventless Processing nf TTIt^m® iqoq.iooo Fpnn ROR/pps
All of the plaques in this dissertation were formulated without the use of solvents.
Solvents are becoming an ever-increasing problem in industry. The regulations for use and
disposal of the solvents are getting tougher every day. Because of this, and the health risks
associated with many solvents, it was deemed necessary to process these materials without the use of solvents. To this end, the plaques processed in this chapter were all prepared in the following manner:
1 Epon 828 was heated in a 50ml . beaker in an oil bath to a temperature of 190°C while
stirring.
2. Ultem® pellets were added in amounts from 10 - 40% by weight and the mixture
allowed to stir and heat until it went clear. This took approximately two hours.
3 The temperature oil . of the bath and the resin mixture was reduced to 1 70°C.
4. 4,4'-diaminodiphenylsulfone was added for a 1 : 1 stoichiometry of reactive hydrogens
on the amine to epoxide groups. This mixture was allowed to stir until it went clear,
about 5 minutes.
5. The beaker containing the modified epoxy was placed in a vacuum oven at 170°C and
allowed to outgas until no more bubbles were present, approximately 5-10 minutes
depending on the viscosity of the resin mixture.
31 The 6. outgassed resin mixture was removed from the oven and qu.ckly poured into a preheated sheet mold. The filled mold was then replaced m the oven.
7. The cure schedule proceeded as follows:
170°C for two hours with no vacuum
230°C for two hours under vacuum to prevent epoxy fi-om degrading in air.
Oven turned off and the mold assembly allowed to come to room
temperature very slowly under vacuum to minimize residual stresses.
The plaques were removed from the molds and labeled. They were then machined into the
sample sizes necessary for mechanical and thermal analysis. The samples were stored at
room temperature and ambient conditions.
2.4.2 A DSC Study; Is I Jltem® 1000-1 OOP Reacting with Epon 828?
Mixtures of 35% and 65% by weight of Ultem® 1000-1000 in Epon 828 were
prepared and tested in a DuPont Instruments DSC2910 Differential Scanning Calorimeter
(DSC) to look for evidence of a reaction occurring between the resins, i.e., a large
exothermic or endothermic peak upon heating. This was compared to a system that is
known to react, Epon 828 with DDS. The traces follow in Figures 2.8 - 2. 10. As can be
seen from the traces of the mixtures of Ultem® 1000-1000 in Epon 828, Figures 2.8 &
2.9, there are no features to the curve below 250°C except the Tg around 200°C. This
32 33 34 (6/M) MOij ^eaH
35 indicates that there is no reaction occurring between the Ultem® and the Epon 828.
Figure 2. 10 shows Epon 828 and DDS, which are known to react and do so starting
around 150°C. In comparing Figures 2.8 and 2.9 with 2. 10, it becomes clear that there
does not appear to be any reaction talcing place between the Epon 828 and the Ultem®
pellets under the curing conditions.
2.4.3 Cure Studie.s
A cure study was performed on a resin system comprised of 10% Ultem® by weight in Epon 828/DDS to see if different curing conditions could produce cured resins with different glass transition temperatures. One of the goals in this work was to make a resin that can be used at the temperatures demanded of a high performance composite matrix material, so a high Tg is essential. All mixtures were processed under the same conditions outlined in section 2.3.1 steps 1-6. The following cure schedules were followed for the resins being tested:
170°Cfor two hours
170°C for four hours
170°C for two hours, 190°C for two hours
170°C for two hours, 210°C for two hours
170°C for two hours, 230°C for two hours.
36 Differential scanning calorimetry was performed on the cured resin systems and the
Tg's determined from the second heat data. The thermograms are shown in Figures 2.1 1 -
2.15 and the results are listed in Table 2.5.
Table 2.5 Tg's for various curing conditions of 10% Ultem® in Epon 828/DDS
Cure Conditions Glass Transition Temperature (°C) 170°C, 2 hrs. 192
170°C,4hrs. 195
170°C, 190°C 192
170°C,210°C 196
170°C,230°C 194
It can be seen that there is no trend for the Tg of these materials with curing temperature. Because of this, and to keep continuity with other researchers who have
worked toughening epoxies in this lab, it was decided that the resins would be cured at
170°C for two hours then 230°C for two hours under vacuum. This will be followed by a
slow cool to room temperature under vacuum as presented in section 2.3 . 1.
37 (B/M) MOtd :^BaH
38 39 40 41 42 2.5 Mechanical Properties
The modulus and the fracture energy were measured to determine the applicability of these materials to high performance composite matrix materials.
2.5.1 Modulus
The moduli were again measured following the procedure outlined in Appendix A.
A typical force/deflection curve is shown on the next page in Figure 2.16 and the results for the Ultem® 1000-1000 modified Epon 828/DDS epoxy/amine resin systems are summarized in Table 2.6.
Table 2.6 Moduli of Epon 828/DDS epoxy resins modified with Ultem® 1000-1000
% Uitem® in E828/DDS Modulus (GPa)
0 2.3
10 2.4
20 2.5
30 2.4
40 2.9
100 2.7
43 A plot of the data summarized in Table 2.6 is shown in Figure 2. 17. The modulus
for both the neat epoxy and the neat thermoplastic are high, so it is no surprise that the
moduli of the modified epoxies are also high. The moduli of all of the modified epoxies
remain above 2 GPa.
Force/Deflection Curve for 40% UItem/E828/DDS ^under Three Point Bending
250 -r
200
150
100
50
\ H 1 \ \- 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005 Deflection [ml
Figure 2. 16 Typical force/deflection curve for Ultem® modified Epon 828/DDS
44 Modulus vs. Ultem Incorporation 3 T
2.8 -
& 2.6 + s
^ 2.4
2.2 -
10 15 20 25 30 35 40 Weight % Ultem
Figure 2. 17 Modulus vs. weight % Ultem® blended into Epon 828/DDS epoxy resm systems.
2.5.2 Fracture Energy
The critical stress intensity factor and the fracture energies of the modified epoxies
are determined in the same manner as the neat Ultem® resin following the procedure in
Appendix B. The results of all of the Ultem® modified resins are shown below in Table
2.7. The results for the neat epoxy and the neat Ultem® 1 000- 1 000 are included for comparison.
45 % Ultem® in E828mDS KiciMNAnfl GicG,.fW(J/mn 0 .78 265 10 .95 376 20 .94 353 30 1.06 468 40 1.40 676 100 3.09 3530
Plots of the data summarized above are shown in Figures 2. 18 & 2. 19. It can be
seen that the fracture energy does increase with the addition of Uhem® to the epoxy resin
system. The toughening does not become significant until the material contains 30% by
weight of the thermoplastic. There is roughly a 2x increase in the fracture energy at the
30% loading level and a 2.5x increase with at the 40% level. Thus the goal of toughening the epoxy resin with poly(ether imide) has been realized with the blending of the Ultem® into the Epon 828/DDS.
46 Critical Stress Intensity Factor vs. Ultem Incorporation
2.50
2.00
1 50 Ca
1.00
0.50 --
0.00
10 20 30 40 Weight % Ultem
Figure 2. 18 Critical stress intensity factor vs. weight % Ultem® blended into Epon 828/DDS
Fracture Energy vs. Ultem Incorporation
700
600
500 - B P5.
400 --
o 300
r? 200 -
100
0 10 20 30 40 Weight % Ultem
Figure 2. 19 Fracture energy vs. weight % Ultem® blended into Epon 828/DDS
47 2.6 Morphology
The morphology of the modified resin systems was studied to determine their
phase separation behavior.
2.6.1 Phase Separation
A JEOL 100-CX Scanning-Transmission Electron Microscope was used in the
transmission electron microscopy mode to study the phase separated morphology of the Epon 828/DDS systems modified with Ultem® 1000-1000. The TEM was used in its high
resolution mode and photomicrographs taken at various magnifications. The samples were
ultramicrotomed at room temperature using a diamond knife, collected onto copper grids
and stained with osmium tetroxide for 3-4 hours before viewing. The resuUs are shown in
Figures 2.20 - 2.24.
From these micrographs, several things become evident. No matter what the morphology, there is a sharp interphase between the included phase and the continuous
phase. There is no smearing of the phase boundaries. This is true even of the epoxy rich inclusions within the thermoplastic rich included phase in the epoxy rich matrix seen at
This 20%. particular morphology seems to suggest that complete phase separation is not
taking place, but that it is arrested at some intermediate level.
48 Figure 2.20 micrograph TEM of 10% Ultem® in Epon 828/DDS. Note the phase separated morphology. The dark areas are the poly(ether imide).
1 |jm
Figure 2.21 TEM micrograph of 20% Ultem® in Epon 828/DDS. Note the phase separated morphology. The dark areas are the poly(ether imide)
49 Figure 2.22 TEM micrograph of 20% Ultem® in Epon 828/DDS at a higher magnification. The dark areas are the poIy(ether imide). Note the epoxy rich phase included in the poly(ether imide) rich inclusions.
Figure 2.23 TEM micrograph of 30% Ultem® in Epon 828/DDS. Note the co-continuous morphology. The dark areas are the poly(ether imide). There are areas of
thermoplastic rich included phase in an epoxy rich continuous phase and areas of epoxy rich inclusions in a thermoplastic rich continuous phase.
50 TEM micrograph of 40% Ultem® in Epon 828/DDS. Note the phase inverted morphology. The dark areas are the poly(ether imide). There is now a thermoplastic rich continuous phase with epoxy rich inclusions.
At the loading 10% level, Figure 2.20, there is an epoxy rich continuous phase with thermoplastic Ultem® rich inclusions. The size of the included phase is on the order of Ifim. At 20% incorporation, Figures 2.21 & 2.22, the morphology remains the same, but the included particles are now larger, on the order of 2|am. At 30%, Figure 2.23, the
co-continuous morphology becomes evident, with areas of thermoplastic rich inclusions in epoxy rich continuous phase, and areas of epoxy rich inclusions in a thermoplastic rich
continuous phase. At 40%, Figure 2.24, a phase inverted morphology exists. In other work done on thermoplastic toughening of epoxy resins, the best results were realized
with either a co-continuous or phase inverted morphology. The same is true of this work.
51 The modified epoxy becomes significantly tougher at the 30% loading level where the co- continuous morphology exists, and a total of 2.5k tougher than the neat epoxy with a
phase inverted morphology.
2.6.2 Fracture Snrfarp
The fracture surfaces of the compact tension specimens were studied to determine
which type of failure occurred, ductile or brittle, and qualitatively describe the interfacial
adhesion between the matrix and the included phase. Samples were prepared from the
broken halves of the compact tension specimens tested previously. The fracture surface
was removed from the rest of the sample as depicted in Figure 2.25.
Cut
Figure 2.25 Side view of compact tension test specimen showing line where it was cut for SEM viewing
52 The sample was held in a vice and a coping saw used to cut the fracture surface
from the bulk of the sample. This fracture surface was then mounted using either carbon
paint or silver paint, depending upon availability, to a stub made specifically for the JEOL 100-CX Scanning Transmission Electron Microscope. The mounted samples were sputter
coated with a thin layer of gold and viewed with a JEOL 100-CX STEM in the SEM
mode at 20KeV. The photomicrographs shown below in - Figures 2.26 2.34 were taken at
the magnifications indicated.
Figure 2.26 SEM micrograph of 10% Ultem® in Epon 828/DDS. Magnification = 6000X. Notice surface irregularities.
53 i
Figure 2.27 SEM micrograph of 10% Ultem® in Epon 828/DDS. Magnification = 10,000X. Notice the bumps and dimples caused by the thermoplastic inclusions.
54 Figure 2.28 SEM micrograph of 20% Ultem® in Epon 828/DDS. Magnification = lOOOX. Notice surface irregularities.
55 Figure 2.29 SEM micrograph of 20% Ultem® in Epon 828/DDS. Magnification = 6,000X. Notice the irregular surface morphology.
56 Figure 2.30 SEM micrograph of 30% Ultem® in Epon 828/DDS. Magnification = lOOOX.
This is the point at which there is a co-continuous morphology via TEM.
57 Figure 2.31 SEM micrograph of 30% Ultem® in Epon 828/DDS. Magnification = 6000X. This is the point at which there is a co-continuous morphology via TEM.
58 Figure 2.32 SEM micrograph of 40% Ultem® in Epon 828/DDS. Magnification = lOOOX. Notice the non-smooth surface showing a ductile failure. This is the point at which there is a phase inverted morphology via TEM.
59 Figure 2.33 SEM micrograph of 40% Ultem® in Epon 828/DDS. Magnification =
10,000X. Notice there is ductile failure occurring around brittle particles. Some spheres are clearly visible on the surface while other appear to have
been broken in half This is the point at which there is a phase inverted morphology via TEM.
60 Figure 2.34 SEM micrograph of neat Ultem®, Magnification - lOOOX. Notice the gross ductile failure of the fracture surface.
61 From these micrographs it can be seen that the phase separation was not complete.
A neat epoxy SEM micrograph would show a featureless surface as smooth as glass,
indicating a brittle fracture. The PEI was making the epoxy matrix more ductile. Discrete
particles were not generally evident in the SEM micrographs with the formulations
including <30% Ultem® in the epoxy. The 40% Ultem® in epoxy seemed to show a
ductile matrix (Ultem® rich) with brittle spherical particles (epoxy rich). Some of these
particles were clearly evident on the surface indicating the particles have pulled out from the matrix denoting poor interfacial adhesion.
2.7 Thermal Analysis
After fabricating plaques by blending 10-40 % Ultem® by weight into Epon
828/DDS, thermal measurements were performed to determine the modified resins'
thermal stability, glass transition temperature and thermal expansion coefficients.
2.7.1 Thermal Stability
To determine the resins' appHcability to high temperature situations a thermal stability measurement was made with the DuPont Instruments TGA2950 Thermal
Gravimetric Analyzer, TGA. The samples were heated from room temperature to 600°C at a rate of 20°C per minute and the temperature at which 1% and 50% weight loss
62 occurred noted. The results are shown in Table 2.8 and a typical TGA trace shown i
Figures 2.36 & 2.37.
Table 2.8 Thermal stability data for Epon 828/DDS resins modified with Ultem® 1000- 1000
Amount of Ultem® in Temperature at Temperature at Epon 828/DDS 1% Weight Lo.ss r^C) 50% Weight Loss (^C)
0 282 412 10 198 415 20 196 420
30 292 443 40 272 452 100 469 N/A
The rate of weight loss with temperature is much greater for the 10 and 20% incorporations of Ultem® in the epoxy, accounting for the lower temperatures for 1% weight loss. As the amount of Ultem® added becomes sufficient to cause an Ultem®
continuous phase to occur, the temperature at which a 1% weight loss is noted jumps by
nearly 100°C. These differences tend to even out by the time a weight loss of 50% is
achieved. As the percent of Ultem® added to the system is increased, the temperature at
which a weight loss of 50% occurs is also increased. The neat Ultem® never sees a 50%
63 64 o cu
65 weight loss at the temperatures used in this experiment (up to 600°C). Adding Ultem®
1000-1000 to the commercial epoxy resin system Epon 828/DDS increases its thermal
stability as indicated by TGA measurements.
2.7.2 Glass Transition Temperature
2.7.2. Dif 1 ferential Scanning Caiorimetry
The Tg was first determined by DSC. Approximately 10 mg of the sample material was weighed out into a DSC pan and sealed hermetically. An empty reference pan was placed on the back probe and the sample pan on the front. The temperature was ramped from room temperature to at 275X a heating rate of 5°C per minute and the heat flow vs. temperature plotted with endotherms shown in the downward direction. Samples of 0 -
40% Ultem® in Epon 828/DDS as well as the neat Ultem® 1000-1000 were analyzed. A typical trace appears in Figure 2.38 and the results summarized below in Table 2.9. The
Tg's were determined from the second heating run.
66
Table 2.9 Gte^ran^sition temperatures for Ultem® modified epoxy systems as determined
Amount of Ultem® in EX2X/nns ^o^^
^ 206 207 204 30 202 40
100 217
Several things can be noted from this table. First only one Tg has been identified for each case. Since there is a phase separated morphology present in these samples, two Tg's wei expected. It can also be seen that the Tg's of these materials do not change appreciably and remain above 200°C. This was as expected since the Tg's of the neat materials are over 200°C.
2.7.2.2 Dynamic Mechanical Analysis
Because only one Tg was in evidence with the DSC measurements, it was decided that the samples should be examined by DMA, a more sensitive technique for Tg measurements. It was anticipated that two Tg's would be seen using this technique. The
sample used was a bar with the dimensions 12mm x 67mm x 3 mm. It was run on a
68 DuPont Instruments DMA983 Dynamic Mechanical Analyzer in the vibration mode with a
gage length of 35mm and a rate of 1 Hz. The samples were heated from room temperature
to 275°C at 1°C per minute. A typical DMA trace for the samples of 0 - 40%
Ultem®/E828/DDS is shown in Figures 2.39 and the results summarized in Table 2.10
Table 2. Tg's of 10 Ultem® modified resin systems by both DSC and DMA
Vo Ultem® in E828/DDS Tg: DMA rC) Tg: DSC (°C) 0 215 206 10 218 207 20 211 204 100 211 217
As can be seen from the results and the trace, only one Tg was present in this case
as well. The Tg's correspond to those determined using DSC to within a few percent. It is suggested that only one Tg can be seen because the Tg's of the neat Epon 828/DDS and the neat Ultem® 1000-1000 are similar, approximately 210°C for the Epon 828/DDS and
217°C for the Ultem®.
69 70 2.7.4 Thermal Expansion Coefficient
Finally, the linear coefficients of thermal expansion (CTE) were determined for
each of the resin systems. First a study was done to determine whether or not the CTE's
were the same in all directions as should be the case for isotropic materials. The
measurements were made across the thickness of half of a compact tension test sample of
neat Epon 828/DDS. After two runs were performed through the thickness of the sample,
two additional runs were performed on the same sample across its width. The results are
shown in Table 2. 1 1 and the traces shown in Figures 2.40 - 2.43.
Table 2. 1 Thermal 1 expansion coefficients for a neat Epon 828/DDS epoxy resin system through the thickness and across the width of the sample
Glassy CTE (mn/m°C) Rubbery CTE (mii/m°C) Through Thickness: 65 163
Across Width: 69 168
These numbers are within a few percent of each other, within experimental error. The samples exhibit isotropic behavior and the thermal expansion coefficients will be measured through the thickness of each of the samples for ease.
71 72 73 74 o o o cu
75 After this experiment was performed, the remaining samples of Epon 828/DDS modified with Ultem® ,000-1000 were analyzed. A representative trace is shown in
Figures 2.44 2.45 and the & results are summarized in Table 2.12 and in Figure 2.46.
Table 2. Glassy 12 and rubbery coefficients of thermal expansion for the epoxy resins modified with Ultem®
Ultem® % in K828/nDS Glassy CTK rmm/mor) Ruhhcry CTE (mm/mor)
^ 65 163 10 67 169 20 68 174 30 64 168 40 70 170 100 25 499
As can be seen from the data, the CTE's for both the glassy and rubbery states of all of the modified samples mimic that of the neat epoxy. This would seem to suggest that even though the sample is co-continuous at 30% and phase inverted at 40%, situations when there is a continuous phase of the thermoplastic Ultem®, the networked structure of the epoxy matrix continues to dominate the thermal expansion properties.
76 77 78 Coefficients of Thermal Expansion vs. Ultem Incorporation
ISO
160 f
140
120 -
100
80
60
40
20
0 4 -1 1 1-
10 15 20 25 30 35 40 Weight % Ultem
Glassy CTE Rubbery CTE
Figure 2.45 Coelficients of thermal expansion vs. weight % Ultem® ini Ultem®/E828/DDS.
2.8 Solubility in Various Solvents
Since these materials could conceivably be used in the aerospace industry, it is
important that they be able to withstand exposure to a variety of different solvents. The
specification used here is from Boeing for high performance composite matrix materials
with aerospace applications. The modified epoxies are subjected to six different solvents:
distilled water, methylene chloride, methyl ethyl ketone, JP5 jet fuel, deicing fluid and
79 Skydol hydraulic flu,d. The samples used are the broken halves of the compact tension test specimens. This gives samples that are all fairly equal in size and shape.
The samples are initially placed in a one dram vial and dried in a vacuum oven
overnight. They are then weighed and this number used as the initial dry weight. The vials are filled with the appropriate solvent and allowed to stand at room temperature. At specified times, the samples are removed from the vials, dried quickly by blotting with a
KimWipe and the weight measured using a Sartorious balance. After weighing, the sampl'e is quickly replaced in the vial. The weights are recorded over a period of time (~ 2 weeks in duration). At the end of the study the samples are removed fi-om the solvents and dried in a vacuum oven for a period of two weeks. The final dry weight is then measured to determine whether any of the samples have lost weight or retained some solvent and gained weight. Charts showing the effects of the various solvents on the different samples are presented in Figures 2.47 - 2.52.
80 I
Effect of Water^n^Jlte^^io^iii;^28/DDs"s^^^^^
3.50
3.00 --
2.50
2.00 --
1.50
1.00
0.50 0 0.00 — 200 400 600 800 1000 0.50 1200 1400 1600 1800
1.00 -- o 0% U D 10%U A 20%U 30%U 1.50 --
Time (hrs.)
Figure 2.46 Effect of water on Ultem® modified E828/DDS. The final set of points are the final dry weights.
Effect of Methylene Chloride on Ultem Modified E828/DDS Systems |
70 -r
60
50 + a 5 40 --
-a 5 30
20
A a 10 o 0% U 10%U A 20%U * 30%U
0 ^
200 400 600 800 1000 1200 1400 1600 1800
Time (hrs.)
Figure 2.47 Effect of methylene chloride on Ultem® modified E828/DDS. The final set of points are the final dry weights.
81 Time (hrs.)
Figure 2.48 Effect of methyl ethyl ketone on Ultem® modified E828/DDS. The final set of points are the final dry weights.
Effect of JP5 Jet Fuel on Ultcm Modifcd E828/DDS Systems [
1.00
0.80 -
0.60 -
0.40 ^ ft A
0.20
0.00 «K:N!S * 1^ I 1 1 — M 1 200 400 600 800 1000 1200 1400 -0.20 - 1600 1800
-0.40
-0.60
-0.80 0%U D 10%U a20%U * 30%U
-1.00 -i-
-1.20
Time (hrs.)
Figure 2.49 Effect of JP5 Jet Fuel on Ultem® modified E828/DDS. The final set of points are the final dry weights.
82 Effect of Deicing Flt^T^nljit^nrM^difi^ E828/DDSS^^;;;n7|
2.00 -r
1.50 -
1.00
g s 0.50 •a
0.00 200 600 800 1000 1200 1400 1600 1800 -0.50 --
-1.00 -
0% U A -1.50 -- 10%U 20%U « 30%U I
Time (hrs.)
Figure 2.50 Effect of deicing fluid on Ultem® modified E828/DDS. The final set of point! are the final dry weights.
Effect of Hydraulic Fluid on Ultcm Modified E828/DDS System
0.60
0.40 4-M
0.20
0.00 ^ (I 200 400 600 800 1000 § -0.20 - 1200 1400 1600 1800
S -0.40 -
-0.60 4>
-0.80
-1.00
-1.20
-1.40 0% U D 10%U A 20%U » 30%U
-1.60
Time (hrs.)
Figure 2.51 Effect of Skydol hydraulic fluid on Ultem® modified E828/DDS. The final set
of points are the final dry weights.
83 All of the solvents swelled the modified epoxy systems. The methylene chloride
swelled the system with 30% Ultem® the most (see Figure 2.48). The CH.Cb and the
MEK could not be removed completely fi-om the swelled systems by drying in a vacuum
oven, so the measure of weight lost to degradation could not be ascertained. The systems
exposed to water, jet fuel, deicing fluid or hydraulic fluid showed weight losses due to
extractables in the system. The weight loss was lessened with the addition of the Ultem®
in all cases except when the systems were exposed to water. The jet fuel, deicing fluid and
hydraulic fluid had especially deleterious effects on the modified resins. The addition of
Ultem® to the systems lessened the effects fi-om the hydraulic fluid and deicing fluid, but caused more degradation to occur at the 30% loading level with the jet fiiel. Overall the addition of Ultem® is beneficial when the systems were exposed to water, MEK, deicing fluid and hydraulic fluid, but detrimental with exposure to methylene chloride and jet fuel.
2.9 Machinabilitv
As can be seen in Figures 2.53 - 2.56, there is an improvement in the machinability of the epoxy resin systems when modified with the Ultem® resin. Machining the neat
epoxy resin results in a powdery residue (see Figure 2.53). The epoxy is brittle and subject
to breaking, chipping and cracking unless extreme care is taken during the machining process. Machining the neat Ultem® results in neat spirals of material and no residue (see
84 Figure 2.54). Machining the Ultem® modified systems result m less powdering, with neat
spirals of material and less likelihood of chipping or breaking (see Figures 2.55 & 2.56). In
short, the machinability is greatly improved with the addition of Ultem® to the
epoxy/amine resin system.
^.^^\ V 1)|)S
Figure 2.52 Photo of a drilled sample of neat epoxy. Notice the powdery residue.
85 /
Figure 2.53 Photo of a drilled sample of neat Ultem®. Notice the neat spirals of material and no powdery residue.
86 I
Figure 2.54 Photo of a drilled sample of 30% Ultem in E828/DDS. Notice the neat spirals of material and very little powdery residue.
87 lilllllMII^^
Figure 2.55 Photo of a drilled sample of 40% Ultem in E828/DDS. Notice the neat spirals of material and very little powdery residue.
88 2.10 Concliisinns
The data presented in this chapter shows that the unreactive poly(ether imide), Uhem® 1000-1000 produced by GE, can be used to toughen the epoxy resin system Epon
828/4,4'-diaminodiphenylsulfone. The Ultem® has been blended into the E828/DDS
without the use of solvents. The result is a material with toughness increased 2.5X, high
Tg's (>200°C), good modulus (>2.5 GPa), and a phase separated morphology. Initially
the system starts with PEI rich inclusions on the order of lum in an epoxy rich continuous
phase. At intermediate loading levels, approximately 30% by weight of the Ultem®, the
system becomes co-continuous with both epoxy rich inclusions in an Ultem® rich
continuous phase and Ultem® rich inclusions in an epoxy rich continuous phase. At 40%
of the Ultem® in the epoxy, the system phase inverts with epoxy rich inclusions in an
Ultem® rich continuous phase. The solubility data shows that the resistance to solvents such as hydraulic fluid and deicing fluid is enhanced, but resistance to water, methylene chloride and jet fuel is decreased. And finally, the machinability of the epoxy resin systems is improved with the addition of Ultem®.
The SEM micrographs indicate that the phase separation is not complete and that the interfacial adhesion between the inclusions and the matrix is not as solid as it could be.
This leads us to the next section where an amine terminated poly(ether imide) is added to the same epoxy matrix. The goal is to achieve better interfacial adhesion between the particles and the matrix, resulting in an increase in the toughness. The amine groups on the
89 PEI will be allowed to react with the epoxide groups in the epoxy to create chemical bonds between the two phases^ It is anticipated that this will increase the adhesion between the two phases.
90 2. 1 1 Referenr.pR
hG.R. Almen P. Mackenzie, V. Malhotra, R.K. Maskell, P.T. McGrail and M S Sefton. Thermoset/Thermoplastic Alloys: An Approach to Tough, High PerformanceResin^^^^^ance Resm Matrices", 20th Int. SAMPE Tech. Conf, 46-59, (1988)
2. M^S. Sefton, P.T. McGrail, J.A. Peacock, S.P. Wilkinson, R.A. Crick, M. Davies and ^^"^•-^"t^'-pe^etrating Polymer Networks as a Route D Tf' • to Toughening of Epoxv Resin Matrix Composites", J9th Int. SAMPE Tech. Conf., 700-710, (1987)
3. E. Martuscelli, P. Musto, G. Ragosta and G. Scarinzi, "A New Supertough Epoxy Formulation with High Crosslinking Density", Die Angewandte Makromolekulare Chemie, 204, 153-159, (1993)
4. G.R. Almen, R.M. Byrens, P.D. Mackenzie, R.K. Maskell, P.T. McGrail and M S Sefton, "977-A Family of New Toughened Epoxy Matrices", 34th Int. SAMPE Svmpos 259-270,(1989) ^ '
5. Keizo Yamanaka and Takashi Inoue, "Structure Development in Epoxy Resin Modified with Poly(Ether Sulphone)", Polymer, 30, 662-667, (1989)
6. Masaki Kimoto, "Micro-structure and Mechanical Properties of CFRP Consisting of Epoxy-Polyetherimide Blend Matrix", Sen-I Gakkaishi, 49(2), 72-78, (1993)
7. Clive B. Bucknall and Adrian H. Gilbert, "Toughening Tetraftinctional Epoxy Resins Using Polyetherimide", Polymer, 30, 213-217, (1989)
8. A. Murakami, D. Saunders, K. Ooishi, T. Yoshiki, M. Saitoo, O. Watanabe, M. Takezawa, "Fracture Behavior of Thermoplastic Modified Epoxy Resins" J. Adhesion 39, 227-242, (1992)
9. Ronald S. Bauer, "Toughened High Performance Epoxy Resins: Modification with Thermoplastics. I.", 34th Int. SAMPE Sympos., 899-909, (1989)
10. D.J. Hourston, JM. Lane, "The Toughening of Epoxy Resins with Thermoplastics: 1. Trifiinctional Epoxy Resin-Polyetherimide Blends", Polymer, 33(7), 1379-1383, (1992)
11. D.J. Hourston, JM. Lane, N.A. MacBeath, "Toughening of Epoxy Resins with Thermoplastics. IL Tetraftinctional Epoxy Resin-Polyetherimide Blends", Polymer International, 26, 17-21, (1991)
91 CHAPTER 3
REACTIONS OF POLY(ETHER IMIDE) WITH EPOXY RESINS: EPON 828/DDS MODIFIED WITH AMINE TERMINATED POLY(ETHER MDE)
3.1 Introduction
As seen in Chapter the unreactive 2, poly(ether imide), Ultem, toughened the Epon
828/DDS when it was blended into the epoxy resin system. Other research has shown that
reactive thermoplastics toughen epoxy resin systems better than the unreactive
thermoplastics'-'^ It is hypothesized that a reactive poly(ether imide), PEI, will toughen
the epoxy system to a greater extent than blending the Ultem did. The reactive group will
be able to react with the epoxide groups of the epoxy to create chemical bonds thus
increasing the adhesion between the included particles and the matrix material in the phase
separated systems.
An amine-terminated poly(ether imide), ATPEI, was received from the General
Electric Company and used as received. First some properties of the ATPEI were
measured. Once the properties of the neat ATPEI were determined, the ATPEI was
reacted into the same epoxy system as above, Epon 828/DDS, and the thermal, mechanical
and morphological properties studied. The results have been analyzed and compared to
those of the unreactive PEI of Chapter 2.
92 3.2 Materials
Amine terminated poly(ether imide), ATPEI, was obtained from the General
Electric Company and reacted mto the commercially available diglycidyl ether of bisphenol
A, DGEBA, based epoxy resin, Epon 828. The system was cured with an amount of 4,4'-
diaminodiphenylsulfone (DDS) to a stoichiometric number of equivalents. The Epon 828
has a Mn of 380 with an epoxide equivalent weight of 188 g/mole. The DDS has an amine
equivalent weight of 62 g/mole. The equivalent weight of the ATPEI will be determined in
this chapter. The chemical structures are shown in Figure 3.1.
DDS
Poly(ether imide)
Figure 3.1 Chemical structures of materials used in this research.
93 3.3 Property Determinations for Npat AJPFT
In this section the equivalent weight is determined, the Tg and thermal stability
measured and the solubility of ATPEI in Epon 828 studied.
3.3.1 Equivalent Weight Determination
A 1 : stoichiometry of amine to epoxide 1 groups was maintained throughout this
dissertation. The amine groups came from either the ATPEI, the DDS or a combination of
the two. In order to determine the amount of curing agent necessary to make up the 1:1
stoichiometry, the equivalent weight of the ATPEI had to be determined. This was
accomplished in 3 ways and the results compared: first by calculations from the
information received from GE, second by using GPC, and third by using vapor phase
osmometry. These procedures are outlined below.
3.3.1.1 Calculated from information received from GE
ATPEI with an estimated DP of 10 was received from the General Electric
Company, Schenectady, NY. No Mn data was available for this material, but was available
for an ATPEI sample with DP = 29. For this material the Mn is equal to 25,500. Using this data the Mn was worked out to be:
94 25,500 _ Mn 29 ~ 10
25,500(10) 29
8793 = Mn
This Mn was then divided by two to get the equivalent weight of the amine in the ATPEI
Equivalent weight: 4396 g/eq.
3.3. 1.2 Determined usin g Gel Permeation Chromatography rOPC)
The ATPEI received from GE was analyzed using GPC to determine the number average molecular weight, Mn. The Mn was found to be 8620. Dividing this by two give an equivalent weight of 43 lOg/eq.
3.3. 1.3 Determined using Vapor Phase Osmometry (VPO)
As a final check, the Mn of the ATPEI sample was also determined using VPO.
First the standards were run, then the sample. The results follow in Table 3.1.
95 3. Results from 1 VPO analysis of standard and ATPEI
Standard ATPEI Concentration Reading Concentration Reading 0.025 0.573 0.025 0.185 0.039 1.122 0.041 0.379 0.063 2.142 0.060 0.720 slope = 41.42 slope = 15.37
From this data, and knowing the Mn of the standard to be 3336, the Mn of the ATPEI was calculated:
slope^,, * Mn^^, = slope,, * Mn,,,^
M._.^lm^ = 8991 15.37
This divided by two gives an equivalent weight of 4495.5. This number corresponds well with the numbers determined from the other techniques. The Mn of the ATPEI is nearly half that of the Ultem at 17,000. The Mn of the ATPEI was taken as 8600 for a starting point and tests conducted with Mn's assumed slightly higher than 8600 and slightly lower than 8600 to determine which assumption gives the best properties in the cured resin (See section 3.4).
96 3.3.2 ThcnnMl Shjhility
The Ultem was stable to very high te.Tiperaturcs and it was desired to see whether
the ATPEI was also stable to these high temperatures The ATI>r:i must not degrade at
the processing or curing temperatures needed for fabricating the plaques of material.
Thermal gravimetric analysis was perlbrmed on the neat ATPIt! to determine the
temperatures at which 1% and 50% weight losses occur The sample was heated in a
DuPont Instruments TGA2950 at a rate of 20"C/min. from room temperature to 500°C.
The scan is shown in Figures 3.2 & 3. The results are listed in Table 3.2.
Table 3.2 Thermal stability data for neat A TPEI
Oiise( 1% Weight l>oss Oiiscf 50% Weight I.OSS 423 "C N/A
97 cu oc oo am. a
in
in
o
tl
a0 6 o V-
o
o
H
1 I I CD 10 C\J o 0) CT.
98 99 3.3.3 Glass Transition Temperature
To determine the Tg, a sample was examined on a DuPont Instruments DSC2910
Differential Scanning Calorimeter. The sample was heated at a rate of 5°C/min. and the Tg
reported from the second heating run. The scan is shown in Figure 3.4 and the result
shown below.
Tg(ATPEI): 195°C
This number is lower than that for the Ultem (217°C). This is not unexpected since the Mn of the ATPEI is almost half that of the Ultem, and it also contains reactive end groups.
Because of the lower Tg it should be possible to process these materials at lower temperatures than the Ultem modified systems.
100 101 3.3.4 Soluh ilitv in Rpon R9R
Since all of the processing in this project is accomplished without the use of
solvents, it is important that the ATPEI be soluble in the epoxy resin it is modifying. To
determine this, the Epon 828 was heated to 170X and the ATPEI added in an amount
equivalent to 20% by weight. This mixture was allowed to heat and stir. The mixture was observed until it went clear. It did so after 30 minutes at 170°C. Thus it was deemed soluble in Epon 828.
3.4 Determination of Stoichiometricallv Optimum Amount of ATPEI to be Added to System
It was estimated that the Mn of the ATPEI received fi-om GE was 8600. To try to get the best possible properties from the modified systems, plaques were made of
E828/DDS modified with ATPEI assuming different ATPEI Mn's: one slightly higher than
8600, one slightly lower than 8600, and 8600. The plaques were then tested for Tg,
thermal stability, modulus and fracture energy. The techniques used, as well as the results obtained, are discussed in this section.
102 3.4.1 Fabrication of Plaq ues- AssnminR Varied Molecular Weights of ATPF.T
Three different plaques were fabricated for this portion of the project. All plaques
were made with 20 weight percent ATPEI in Epon 828/DDS. Three different Mn's of the
ATPEI were assumed - 8200, 8600, and 9000. The plaques were fabricated using these
numbers as the basis for the equivalent weight and were created as follows:
1. A specified amount of Epon 828 was heated to 170°C in a 50 ml beaker over an oU
bath while stirring with a mechanical stirrer.
2. An amount of ATPEI equal to 20% of the weight of the Epon 828 was added to the
beaker and allowed to heat and stir until the mixture went clear (approximately
30 mins.).
The 3. number of equivalents of ATPEI added to the system was determined using the
equation below:
{Wt. ATPEI [g]}/{ (Assumed Mn/2)[g/eq]} = # of Equivalents
4. 4,4'-diaminodiphenylsulfone was added to the system to make up the difference in
amine equivalents for a 1:1 stoichiometry of amine to epoxide groups in the final
system. This was allowed to mix and heat until the mixture went clear (approximately
7 mins.). The mixture 5. was placed into a vacuum oven preheated to 170°C and aUowed to
outgas until no further bubbles were produced (about 10 mins.).
The mixture 6. was poured into a preheated sheet mold and cured following the schedule
outlined below:
A. 2 hours at 1 70°C without vacuum
B. 2 hours at 230°C with vacuum
C. Heat turned off and the sample allowed to come slowly to room
temperature overnight under vacuum.
The cure schedule was kept the same as the Ultem modified samples for consistency.
3.4.2 Thermal Stability
To determine which resin system would have the highest thermal stability, a measurement was made for all three samples using thermal gravimetric analysis, TGA, as in section 3.3.2. The samples were heated at a rate of 20°C/min. from room temperature to 500°C under nitrogen. A representative trace can be seen in Figures 3.5 & 3 .6 and the resuhs shown in Table 3.3.
104 105 o o o cu
106 Table 3.3 Results of the TGA analysis on the samples assumingo differing molecularm weights of the ATPEI &
Assumed Mn of ATPFT Onset l"/o W.iahf T n.. (or) Onset SQQ/n W.ipht T (qq 8200 362 436 8600 306 463 9000 363 447
There seems to be no trend in the data shown here. The 8600 resin has a lower
onset of 1% weight loss, but a higher onset of 50% weight loss than the other two resins.
Comparing this data to that of the Ultem modified resins, it can be seen that the onset of
1% weight loss in the ATPEI modified resins occurs at higher temperatures than that for
the Ultem modified resins. This is also true for the onset of 50% weight loss. The ATPEI
modified resins are stable to higher temperatures than the unreactive PEI (Ultem) modified
resins.
3.4.3 Glass Transition Temperature
3.4.3.1 Differential Scanning Calorimetrv. DSC
The Tg's of the three materials were determined and compared. This was done using dififerential scanning calorimetry, DSC. 10 mg samples were weighed out and
107 hermetically sealed in DSC pans. The sample pan was placed on the front probe and the empty reference pan was placed on the rear probe. This forces the exothermic reactions to appear in the positive direction on the traces. The samples were heated from room temperature to 275°C at a heating rate of 5°C/min. under nitrogen. The Tg's from the second heating are reported. A typical scan is DSC shown in Figure 3.7 and the results are summarized below in Table 3.4.
Table 3.4 Summary of Tg's as determined by DSC for the ATPEI modified resins with varying assumed molecular weights
Assumed Mn of ATPEI Je (°C) 8200 209 8600 206 9000 205
These numbers are comparable with those of the Ultem modified resins and to each other. Again only one Tg is visible. These numbers will be checked with DMA.
108 o o o o I I I I fB/M) MOId qE9H
109 3.4 3.2 Dynamic Mechanical Analysis, DMA
As a check, the Tg's were also determined using DMA. The samples consisted of bars with a gage length of 40 mm, a width of 12 mm and a thickness of 3 mm. They were analyzed at a frequency of 1 Hz from room temperature to 275°C at TC/min. A representative trace is shown in Figure 3.8 and the results outlined in Table 3.5.
Table 3.5 Summary of Tg's as determined by DMA for the ATPEI modified resins with varying assumed molecular weights
Assumed Mn of ATPEI jg (qq) 8200 204
8600 197 9000 200
These numbers are about 5°C lower than those obtained with DSC, but in the same range.
Only one Tg is noted.
110 Ill 3.4.4 Modulus
Next the mechanical properties of the modified resins were determined. The
ultimate goal is to toughen these resin systems. The material with the best mechanical
properties will be used for the rest of this study since the thermal properties are so similar.
The moduli of these materials were determined fi-om three point bend testing as
described in Appendix A. The samples were loaded in three point bending at a rate of 0.05
cm/min. until a 5% strain was reached. The slope of the force/deflection curve was used to
calculate the bending modulus. A typical force/deflection curve is shown in Figure 3.9 and
the results summarized in Table 3 .6.
Table 3.6 Flexural moduli of the three resin systems with varying assumed Mn's
Assumed Mn of ATPEI E (GPa)
8200 2.7
8600 2.8
9000 2.7
The modulus of the 20% ATPEI/E828/DDS resin system with the assumed Mn of the
ATPEI equal to 8600 is the highest at 2.8 GPa.
112 Force/Deflection Curve for 20% ATPEI/E828/DDS under Three Point Bending
0.001 0.002 0.003 0.004 0.005 0.006 Deflection [m]
Figure 3.9 Typical force/deflection curve for the 20% ATPEI/E828/DDS resin system under three point bending.
3.4.5 Fracture Energy
The most important factor to determine in this section is the fracture energy. The
fracture energy is the gauge of the toughness of these materials. The fracture energy is
determined by performing a compact tension test on the samples. This procedure is
outlined in Appendix B. The samples were loaded at a rate of 0.05 cm/min. on an Instron model TTBM Universal Testing Instrument and the load to propagate the precrack noted.
This number was used to calculate the critical stress intensity factor. The Kic and the
113 modulus were used to calculate the fracture energy. These nutnbers are hsted in Table 3.7.
The data is shown graphicaUy in Figures 3.10-3.11.
Table 3 .7 Kk & Gic information for 20% ATPEI/E828/DDS resins Mn's with differing assumed
Assumed Mn ATPET Kir (MWm'") 8200 .83 255 8600 .95 322 9000 .79 231
Critical Stress Intensity Factor vs. Assumed Mn of ATPEI for 20% ATPEI/E828/DDS
S
8200 8600 9000 Assumed Mn
Figure 3.10 Critical stress intensity factor vs. assumed Mn of ATPEI in 20% ATPEI/E828/DDS
114 l^racture Energy vs. Assumed Mn~ofI^^ m 20% ATPEI/E828/DDS
8600 9000 Assumed Mn
Figure 3.11 Fracture energy vs. assumed Mn of ATPEI in 20% ATPEI/E828/DDS
It can be seen that the K.c of the 8600 Mn resin is larger than the other two. The Qc is
also slightly higher. Because of this, 8600 will be used as the Mn of the ATPEI received from GE for the remainder of this research
3.4.6 Summarv & ConclnsinnQ
Below is a table summarizing the data obtained in this section. It compares all the properties of the three different resin formulations. AU are 20% ATPEI in Epon 828/DDS, but the assumed Mn's for the equivalent weight calculations were as follows: 8200, 8600,
115 and 9000, The numbers were based on .ha. de.e™„eC .hrough GPC analysis, as weU as
ATPE, gave .he bes. properties. Th.s is .he number .ha. wU, be used .hroughou. .Ms work.
Table 3.8 Summaiy of all of the data presented in section 3.4
As.sunied 1% 50% Te DSC Tg-DMA Mn Weight Weieht Kic ATPEI Loss (°C) Loss f°C) (°C) (°C) (GPa) (MN/m^'^) (J/m^) 8200 362 436 209 204 2.7 .83 255 8600 306 463 206 197 2.8 .95 322 9000 363 447 206 201 2.7 .79 231
After comparing the above data, it was decided that the material possessing the
best overall properties was that formulated under the assumption that the Mn of the
ATPEI is 8600. 8600 will be used as the molecular weight of the ATPEI for all future work in this dissertation.
116 inEpon 828/nns;
The processing of these plaques is very similar to that for the Ultem modified resin
systems except that the initial processing temperature is lower. It seems either the amine
functionality or the lower molecular weight of the poly(ether imide) or a combination of
these two factors allows the ATPEI to dissolve into the Epon 828 at lower temperatures
and within shorter periods of time than was necessary for blending the Ultem into the E828. The process used is as follows:
1 Epon 828 was heated in a 50 ml . beaker in an oil bath to a temperature of 1 70°C with
stirring.
2. ATPEI powder was added in varying amounts fi-om 10 - 40% by weight and allowed
to stir and heat until the mixture was clear. This took approximately one-half hour.
3. The equivalent weight of the ATPEI added was calculated by taking the weight (in
grams) and multiplying it by I/(4300g/eq) to get the # of equivalents added.
4. 4,4'-diaminodiphenylsulfone was added to make up a 1 : 1 stoichiometry of the amine
to epoxide groups. The mixture was again allowed to stir until it went clear, about 5
minutes.
117 5. The modified epoxy material was transferred from the beaker to a Nalgene evaporating dsh. TWs was placed in a vacuum oven at 170°C and the resin was
aUowed to outgas until no more bubbles were present, approximately 5-10 minutes
depending on the viscosity of the resin mixture.
6. The outgassed resin mixture was removed from the oven and quickly poured into a
preheated sheet mold, which was then replaced in the oven.
7. The cure schedule proceeded as follows:
A. 1 70°C for two hours with no vacuum.
B 230°C for two hours under vacuum to prevent the system from degrading
m air.
C Oven turned off and plaque allowed to come to room temperature very
slowly under vacuum to minimize residual stresses.
At this point the plaques were removed from the molds, machined into the sample sizes
necessary for testing, and labeled for future use. The samples were stored at room
temperature at ambient conditions.
One very noticeable difference in the processing of the ATPEI/E828/DDS vs.
Ultem/E828/DDS is that the viscosity of the ATPEI modified resins were lower for the
comparable weight percents added to the epoxy resin. For example, at 30% the Ultem
modified system was like taffy at 170°C while the ATPEI modified system was closer to the consistency of honey. This allowed for faster outgassing of the ATPEI modified
118 these materials commercially processable.
^-^•^^SCStudxJsiamEeacM^^
To determine if the ATPEI is reacting with the Epon 828 epoxy resin or simply
being blended like the Ultem modified epoxy, a DSC study was done. The mixture
consisted of 79 equivalents Epon 828 to 1 equivalem ATPEI. This mixture was heated at a rate of 5°C/min. from room temperature to 300°C The trace is in Figure 3. 12. It can be seen that a reaction of some kind is taking place upon heating during the first heating run.
The second heating in Figure 3.13 shows essentially a flat line. No endotherms or
exothenms are visible in this trace. This shows that a reaction is occurring between the
Epon 828 and the ATPEI.
119 120 121 l^OtonaLAnalysis
This section analyzes the thermal propen.es of the plaques of material .ade fro. 0-40% by weight of ATPEI in Epon 828/DDS.
3.6.1 Thermal Stability
It was seen in section 3 3 that the thermal stability of the 20% ATPEI modified
E828/DDS was higher than that of the Ultem modified epoxy system. In this section the onset of and 1% 50% weight loss will be determined and compared. All of the samples consisted of approximately 20 mg of the material being analyzed. The sample was placed in the TGA pan and heated at a rate of 20°C/min. from room temperature to 500»C under nitrogen. The traces were recorded and the onset of 1% and 50% weight loss determined.
Representative traces are shown in Figures 3.14 - 3.17 and the results summarized in
Table 3.9.
122 —
CO
o o o o
oc o oo (DO. 3 o Q c o(/I o > 2P
I
O OCO o
'53
01 cn O
o o ao £ o I
o o
00
W o p-( O
o I o o O
< I o s H *53 f I o o o T I OJ o o o o (0
123 124 125 o o o
126 Table 3.9 Summary of thermal stability data by TGA
ATPEf % Qnseti%_Weisht Loss fOQ Onset SQo/o Wei.hM 0 282 412 10 430 494 20 436 496 30 443 499 40 360 447 100
* There is no 50% weight loss before 600°C
Several things can be seen from this data. First, the onset of 1% weight loss is much higher than that for the Ultem modified epoxy resins. It is also much higher than the
neat Epon 828/DDS resin. Also, the onset of 50% weight loss is not much higher than that
for the onset of 1% weight loss. As one can see from the traces, there is essentially a flat
line from the begimiing of the heating to the point where there is significant drop in the
weight. The exception is the 40% ATPEI in E828/DDS.
3.6.2 Glass Transition Temperature
The Tg's of these materials were determined using Differential Scanning
Calorimetry. The samples were run in A DuPont Instruments DSC2910 at a heating rate of 10°C/min. from room temperature to 275°C under nitrogen. The Tg's were taken as the
127 inflection point on the second heating curve. The results are Hsted below in Table 3.10 and a representative DSC trace shown in Figure 3.18.
Table 3 . 10 DSC Tg data for the ATPEI/E828/DDS resin systems.
% ATPET Ts (OC) 0 207 10 208 20 214 30 209
40 191
100 195
It can be seen that the Tg's are all high, around 200°C. The number for the 40%
ATPEI/E828/DDS system is slightly lower than the others.
128 129 3 .6.3 Thermal FjcgansjonCoefficie^^
The linear coefficients of thermal expansion were detennined for the ATPEI
modified epoxy resin systems. The samples studied were one-half of a miniature compact tension test specimen that had been tested and broken. The CTE was measured through
the thickness of the sample usmg a standard expansion probe on a DuPont Instruments TMA2940 Thermal Mechanical Analyzer. The samples were heated from room
temperature to 275°C at a rate of 5X/min. under nitrogen. The glassy and rubbery
coefficients of thermal expansion were determined for each of the samples. A trace is
shown in Figures 3. and 19 the glassy and rubbery CTE's reported in Table 3. 1 1. A graph
of these results is presented in Figure 3.20.
Table 3.11 Linear coefficients of thermal expansion for the ATPEI modified Eoon 828/DDS systems
ATPEI in E828/nDS % Glassy CTE (nm/ruQC^ Rubbery CTE (am/n^QC^
0 65 163
10 74 160
20 67 170
30 65 171
40 59 179
130 131 Coemcien,, »f THc™a, Expansion v. Weiii^^^TTFiT in ATPEI Modified E828/DDS
200
180
160 ^140
120 °a 4 100
80 --
60
40 -
20
0
40 Weight % ATPEI • Glassy CTE's Rubbery CTE's
Figure 3.20 CTE vs. % ATPEI for ATPEVE828/DDS
It can be seen that the CTE's do not change significantly with increasing amounts
of ATPEI in the epoxy system. They remain nearly the same as the neat epoxy resin
system, indicating that the epoxy network is dominating the thermal expansion properties
at all loading levels.
3.7 Mechanical Prop ertie.s
This section analyzes the modulus, critical stress intensity factor and the fi-acture energy of each of the samples.
132 3.7.1 ModnliiQ
The moduli of these ATPEI modified Epon 828/DDS resin systems were
determined using a three point bend test as described in Appendix A. Bars of material
12 mm wide and 3 mm thick were placed across a fixture with a span of 42 mm. The
samples were pressed with a third dowel at a rate of 0.05 cm/min to an outer dement
strain of 5«/.. The force deflection curve was analyzed and the modulus detemuned from
the slope of the force/deflection curve. A typical force/deflection curve for these materials
is shown in Figure 3.21 and the results are shown m Table 3.12. A plot of the data is
shown in Figure 3 .22.
Force/Deflection Curve for 20% ATPEI/E828/DDS under Three Point Bending
g 100 -I-
0.001 0.002 0.003 0.004 0.005 0.006
Deflection [mj
Figure 3.21 A typical force/deflection curve for a three point bend test on the ATPEI modified resins.
133 Table 3.12 Modulus data Ld forlor methe AlFhlATPFT modifiedr..^^ ^^ ^ Epon 828/DDS systems
% ATPET in FX-)«/nT^c^ E (GPa) 0 2.3 10 2.5 20 2.7 30 2.7 40 2.8
Modulus vs. PEI Inco rporation
3 T
2.8 -
2.6 -
2.4
2.2 -
—t- 10 20 30 40 Weight % PEI
• Ultem a ATPEI
Figure 3.22 Modulus vs. amount of ATPEI or Ultem in the Epon 828/DDS resin system,
As can be seen from the data, the moduli of the ATPEI modified epoxy resins are all above 2 GPa. For the 10 - 30% resins, the moduli are all above those of the
134 corresponding Ultem modified resins. The 40% resin has a higher modulus than the 40% Ultem modified Epon 828/DDS. The moduli of the ATPEI modified resins increased
slightly with low incorporations of the ATPEI, but remained fairly constant after that. The Ultem modified resins remained somewhat constant in the begimiing, showing a marked
increase only for the 40% incorporation of the Ultem.
3.7.2 Fracture Energy
The toughness of the modified networks was determined using compact tension testing. The samples were studied as outlined in Appendix B. The samples were run at a rate of 0.05 cm/min. on the screw driven Instron model TTBM tensile testing machine.
From this test, the critical load for crack propagation was determined and used to calculate the critical stress intensity factor. The data is presented in Table 3. 13 and in
Figure 3 .23.
Table 3.13 Critical stress intensity factors vs. amount of ATPEI in modified epoxy system
% ATPEI in E828/DDS Kir (MN/m^^)
0 .78
10 .72
20 .69
30 .72
40 1.4
135 2.50
2.00
^fla 1.50
0.50
0.00 ' h- H 10 20 30 40 Weight % Poly(ether imide)
o ATPEI • Ultem |
Figure 3.23 vs. Kk amount and type of PEI used in modifying the E828/DDS resm system
From these numbers and the modulus of the sample, the fracture energies were calculated. The fracture energy values are shown below in Table 3 .14 and the chart of Gic vs. amount and type of poly(ether imide) used to modify the Epon 828/DDS system presented in Figure 3.24.
It can be seen that the Kic of the ATPEI/E828/DDS systems is lower than that of the Ultem/E828/DDS systems, with the exception of the 40% level of PEI incorporation.
In this case the two Kic's are nearly identical. It was expected that the Kic's of the ATPEI modified epoxy systems would be higher than those of the Ultem modified systems. The amine fiinctionality can react with the epoxide groups forming a covalent bond between
136 Table 3 .14 Fmctoe energy data for the ATPEI modified Epon 828/DDS epoxy resin
% ATPEI in FX?x/nng
0 265 10 207
20 176
30 192 40 700
Fracture Energy vs. PEI Incorporation
700
600
500 -
- 'a' 400
- J^300 1
200 -
100 -
0 0 10 20 30 40 Weight % PEI
ATPEI • Ultem
Figure 3.24 Fracture energy vs. type and amount of poly(ether imide) incorporated in Epon 828/DDS.
137 toughness. This was not the case. The G,c's of the ATPEI modified epoxy systems
decreases until the 40% inoon^oration ,eve,, wh.,e the Ulte. modified systems have G^ic's that increase substantially at the 30 and 40% loading levels. Even in the case of 40»/.
2.2X increase in the fracture energy whh the ATPEI and a 2.5X increase in the fracture
energy of the Ultem modified systems at the highest loading levels. A look at the morphology helps to explain this.
3.8 Morphology
Since it has been seen that phase separation is necessaiy in the epoxy systems for toughening to occur, it was essential to determine the morphology of these systems. To do this, the samples were observed with both TEM and SEM. The techniques and results follow.
138 3.8.1 Phase Separation
In order to determine the phase separation behavior in these systems, small samples were removed from the plaques and studied using transmission electron
microscopy. The technique used is the same as that for the Ultem modified systems. The
samples were ultramicrotomed at room temperature using a diamond knife, collected onto
copper grids and stained with osmium tetroxide prior to observation in the TEM. The samples were observed using a JEOL 100 CX Scamnng-Transmission Electron
Microscope in the TEM mode at 100 KeV. The specimens were viewed and photos taken of representative areas on the grid. Photomicrographs follow in Figures 3.25 - 3.28.
It can be seen that there is phase separation in the 10% ATPEI/E828/DDS samples on the order of 0. ^im. No phase 1 separation is visible in the 20, 30 or 40%
ATPEI/E828/DDS samples. It has been shown for other thermoplastic modified epoxy systems that there is a critical size of the included phase necessary for toughening the resin system. This critical size is generally somewhere between 0.5 - 10 ^m in diameter. The size of the particles for the 10% case is smaller than that, and there seems to be incomplete phase separation in the other cases. This could account for the lack of toughening in the
10-30% samples, but does not explain the sharp increase seen in the 40% case.
139 Figure 3.25 TEM micrograph of 10% ATPEI/E828/DDS. Notice the phase separation on the order of 0. 1 |im.
Figure 3.26 TEM micrograph of 20% ATPEI/E828/DDS. Note the lack of sharp phase separation.
140 Figure 3.27 TEM micrograph of 30% ATPEI/E828/DDS. Note the lack of sharp phase separation. ^ ^
Figure 3.28 TEM micrograph of 40% ATPEI/E828/DDS. Note the lack of sharp phase separation.
141 3.8.2 Fractiirp S^Mrfpirp
As in Chapter U was des.red 2. to see the way these modified epoxies fractured
The fracture surfaces of the compact tension test specimens were viewed using an SEM. The fracture surface was removed from the rest of the sample with the use of a coping
saw (see Figure 2.25). and mounted on an SEM stub using either carbon or silver pamt. These samples were then sputter coated with a thh, layer of gold to prevent charging while
in the electron beam. The samples were then viewed with a JEOL 100-CX STEM in the
scamnng electron microscopy mode at 20 KeV. The whole surface of each sample was viewed and photos taken of representative fracture areas at different magnifications. The
photomicrographs are shown below in Figures 3.29 - 3.34.
From these micrographs it can be seen that there is a different mechanism at work
here. There are no particulate inclusions evident in most of these samples. There is also a
more ductile fracture occurring in most if not all of these samples as evidenced by the
striations and irregular surfaces. It seems that the ATPEI is not phase separating from the
epoxy as the Ultem did, but is rather becoming incorporated into the matrix. It seems at
40% incorporation of the ATPEI into the epoxy, there is enough thermoplastic in the
system to increase the toughness based solely on the amount of thermoplastic added and the ductility it imparts to the epoxy.
142 Figure 3.29 SEM micrograph of 10% ATPEI/E828/DDS fracture surface. 6000X
magnification. Although there is debris on the surface, no particulate inclusions are evident.
143 Figure 3.30 SEM micrograph of 20% ATPEI/E828/DDS fracture surface. lOOOX magnification. Notice the striations on the surface indicative of a more ductile fracture.
144 Figure 3.31 SEM micrograph of 20% ATPEI/E828/DDS fracture surface. 6000X magnification. Again there is debris on the surface, but no particulate inclusions are evident.
145 !.32 SEM micrograph of 30% ATPEI/E828/DDS fracture surface. 1500X magnification. The striations are no longer evident, but the surface is not smooth either. This does not show either a clearly ductile or clearly brittle fracture.
146 I
Figure 3.33 SEM micrograph of 40% ATPEI/E828/DDS fracture surface. lOOOX magnification. The surface is similar to the 30% ATPEI/E828/DDS sample viewed previously
147 Figure 3.34 SEM micrograph of 40% ATPEI/E828/DDS fracture surface. lOOOOX magnification. The surface irregularity indicates a clearly ductile fi-acture. Again there is surface debris but possibly also some included particulate matter.
148 3.9 Solubility in VaHpus Solvent«;
Since it is conceivable that these materials could be used m the aerospace field, it is
important that they be able to withstand a vanety of different solvents. The specification
used here is from Boeing for high performance composite matrix materials with aerospace
applications. The modified epoxies are subject to six different solvents: distilled water,
methylene chloride, methyl ethyl ketone, JP5 jet fuel, deicing fluid and Skydol hydraulic
fluid. The samples are the broken halves of the compact tension test specimens. This gives
samples that are all fairiy equal in size and shape.
The samples are initially placed in a one dram vial and dried in a vacuum oven
overnight. They are then weighed and this number used as the initial dry weight. Next, the
vials are filled with the appropriate solvent and allowed to stand at room temperature. At
specified times, a sample is removed from the vial, dried quickly by blotting with a
KimWipe and the weight measured using a Sartorious balance. The sample is then quickly replaced in the vial and the weights recorded over a period of time (approximately eight weeks in duration). At the end of the study the samples are removed from the solvents and dried in a vacuum oven for a period of two weeks. The final dry weight is then measured to determine whether any of the samples have lost weight or possibly retained some of the solvent and gained weight. The charts showing the effects of the various solvents on the different samples are presented below in Figures 3.35 - 3.40.
149 —
Effect ofWateronATPEI Modified E828/DDS 3.50 J
3.00 --
2.50
-1.00 - Time (hrs.) .1.50 --
Figure 3 35 Effect of water on ATPEI modified E828/DDS. Tlie last set of points are the rinal dry weights.
Effect of Methylene Chloride on ATPEI Modified E828/DDS
60.00 -r
50.00 -
^ 40.00 - O A 5 30.00 A I >o 20.00
10.00 AO 0% A a 10%A A 20%A o 30%A
0.00 i 1 — 1 0 200 400 600 800 1000 1200 1400 1600 1800
Time (hrs.)
Figure Effect 3.36 of methylene chloride on ATPEI modified E828/DDS. The last set of points are the final dry weights.
150 I
14 -r
12 -
a> 10 4-
(J 8 +
*a» 6
0% A B 10%A A 20%A o 30%A
0 200 — 400 600 800 1000 1200 1400 1600 1800 Time (hrs.)
Figure 3.37 Effect of methyl ethyl ketone on ATPEI modified E828/DDS. The last set of points are the final dry weights.
Effect of JP5 Jet Fuel on ATPEI Modified E828/DDS
1.00 -r
0.50 -
0.00 M A fA-
400 600 800 1000 1200 1400 1600 1800
5 -0.50 Time (hrs.)
.1.00
-1.50 -- 0% A o 10%A A 20%A o 30%A
2.00 -L
Figure 3.38 Effect of JP5 Jet Fuel on ATPEI modified E828/DDS. The last set of points are the final dry weights.
151 EffcctofPeicing Fluid on ATPEI Modified E828/DDS
2.00 -r-
1.50 --
1.00 --
S 0.50 u.A Ja 0.00 .SP 200 600 800 ^ -0.50 1000 1200 i 400 1600 1800
Time (hrs.) -1.00
-1.50
-2.00 -L ^-°!^A^JO%Aa 20%A o 30%A
Figure 3.39 EfTect of deicing fluid on ATPEI modified E828/DDS The last set of points are the final dry weights.
Effect of Skydol Hydraulic Fluid on ATPEI Modified E828/DDS
0.20 -r
0.00 4^ H 200 ^ 400 600 800 1000 •0.20 4 1200 1400 1600 1800
4> •0.40 Time (hrs.)
U -0.60
•S -0.80 +
^ -1.00 4-
•1.20
1.40 0%A n 10%A A 20%A o 30%A 1.60 ^
Figure 3.40 Effect of Skydol hydraulic fluid on ATPEI modified E828/DDS. The last set of points are the final dry weights.
152 couM no. be removed co.p,e.e,y fr„™ .He swe„ed systems d^., , , __ so the measure of weight los, .0 degradaUon could not be ascertained. The systems exposed water, ,0 jet m, deicing fluid or hydraulic fluid showed weight losses due to extractables in the system. The weight loss was lessened with the addition of the ATPEI in all cases except when the systems were exposed ,0 water or jet ft,el. There was little
difference with the addition of the ATPEI to the ES28/DDS systems that were subjected to distilled water. The ATPEI had some beneficial effects to the systems subjected .0 hydraulic fluid, deicing fluid and MEK, but was a detriment to the systems exposed to
CH2CI2 and jet ftiel.
3.10 Machinability
As can be seen in Figures 3.41 & 3.42 below, there is an improvement in the machinability of the epoxy resin systems when modified with the ATPEI. Machining the neat epoxy resin results in a powdery residue. The epoxy is brittle and subject to breaking, chipping and cracking unless extreme care is taken during the machining process.
Machining the ATPEI modified systems result in neat spirals of material with less likelihood of chipping or breaking. Debris is still evident on and around the samples,
153 though. Overall, the machinability is improved with the addition of ATPEI to the
epoxy/amine resin system, but less so than when Ultem® is added.
'''' A ri>i;i/{.;s2s//)IXS
Figure 3.41 20% ATPEL^828/DDS after drilling. Note the neat spirals of material and the excess debris created by the machining.
154 .()', .\ ri>i:i/i-:s2s/DDS
?ure 3.42 40% ATPEI/E828/DDS after drilling. Note the neat spirals of material and the excess debris created by the machining.
155 No phase separation was observed under ,he TEM, Phase separation seems to be a requirement for toughenrng eposes with po^ether imide)^ An attempt was made to force the phase separation of the ATPEI modified systems by changing the processing or curing condition, temperatures or times. This was done with the aid of an Olympus BH-2 Optical Microscope fitted with a Linlcham TMS9000 Hot Stage.
The samples used for this test were prepared in the usual manner except that small batches of 20»/. ATPEI/E828/DDS were processed at temperatures of either 150»C,
1 70"C or 1 90T. These three samples were B-staged and held in a freezer until ready to be
studied, effectively stopping the curing process. A small piece of the B-staged, frozen
material was cut from the sample, placed between glass slips on the hot stage and quickly
heated up to the curing temperature of either 1 yOT, I WC, 2 1 OT or aSCC The
samples were observed under the optical microscope for a period of 30 minutes to
detenmine whether or not phase separation was occurring. A ramp cure was also tested.
This sample was initially very quickly brought to the temperature it was processed at and
held there for 5 minutes. The temperature of the sample was then quickly dropped to 80»C and held for hours, 2 increased to 130»C for 1 hour, 170°C for two hours, and finally
230''C for two hours. These different curing cycles were tested with each of the three processing conditions to no avail. Several representative photos are shown in Figures 3.41
- 42, but the only features visible are air bubbles. Thus it was determined that the curing
156 Figure 3.43 Typical optical micrograph of a sample of ATPEI/E828/DDS after cure study,
157 gure 3.44 Typical optical micrograph of a sample of ATPEI/E828/DDS after cure study.
158 and processing condmons had no effec. on the presence or absence of phase separation with this system.
3.12 Summary and Cnnrlnsions
ATPEI can be processed into the Epon 828/DDS resin system without the use of solvents. The ATPEI reacts into the Epon 828 resm at a lower temperature and more
quickly than the Ultem blends mto the epoxy. The thermal properties of the cured resms
are comparable to those of the Ultem modified resm systems, but the mechanical and
morphological properties are much different. There is no evidence of phase separation
any of the ATPEI modified resins except the 10% loading level. This had very small
inclusions, below the size necessary for toughening the resins. The fi-acture surfaces showed evidence of some ductility in the fi-acture, indicating that the ATPEI was
incorporated into the matrix making it more ductile than a neat epoxy resin matrix would
be.
The moduli of the ATPEI modified resins were all above 2 GPa and even increased
to levels above the Ultem modified resins in - some cases (10 30%). The resin system
loaded with 40% of the ATPEI had a modulus comparable to that of the 30% ATPEI
modified system (approximately 2.7 GPa), but lower than that of the 40% Ultem modified resm (2.9 GPa). The Kic values were lower than in the case of the Ultem modification in all cases except for the loading 40% level. In this case the two systems were about equal
159 with values on A MN/m^^^. More of a difference was evident with the Qe values. The values of the 10 - 20y. samples were substantially different with the gap widening with an increasing amount of poly(ether imide) modifier added (see Rgure 3.24). At the 40%
loading levels the system modified with the ATPEI was nearly the same as that modified with the Ultem (700 vs. 678 J/m\ respectively).
Studies were run to determine if the cause of the phase separation or lack of phase
separation was due to processing or curing conditions. Samples were processed at varymg
temperatures then the cures completed using different cure schedules. None of this made
any difTerence in the phase separation behavior of the resin system. The resins remained homogeneous as viewed through optical microscopy.
In conclusion, reacting the 8600 Mn ATPEI into the Epon 828/DDS resin system
did not result in the amount of toughening anticipated. It was hypothesized that the amine
functionality on the ATPEI would result in covalent bonding between the epoxy and the poly(ether imide) causing an increase in the fi-acture energy. Although the bonding may have taken place, phase separation did not. The phase separation was not a phenomenon of the processing or curing conditions, so it is speculated that the lack of phase separation was caused by the relatively low molecular weight. In the next section, Ultem and ATPEI are combined and used as the modifying poly(ether imide) in an attempt to get phase separation fi-om the Ultem and adhesion between the phases from the amine functionality of the amine terminated poly(ether imide).
160 2J3_References
Resin Systems". 3 Is, In, SAMPE ^""^ 58^58^ 1 986^
C«S.S^^ carbon P*er Reinforced
LfoneSXineS^^^^^^^^^^ ^o^A^I E.he. P^*™''nce 32(9), 1633-1641, (1991) Materials", Po/ymer,
4. Greggory Bennett, Ph.D. Thesis, University of Massachusetts, (1991)
"T^el'TerheLS^^^^^ Matrices", 20tH In, IaMPE Z. Co^Zt'. 0988)'
6. M.S. Sefton, P T. McGrail, J.A. Peacock, S.P Wilkinson, R.A Crick M Davies and "''"8 Po'^'"- Networks as a Route to TSn-Sf Epoxy tTZResin Matnx Composites", imin, SAMPE Tech. Con/., 700-710, (1987)
7. E. Martuscelli P. Musto, G. Ragosta and G. Scarinzi, "A New^«tirSupertough Eooxv srsri si'fs sr *
R K. Maskell, P.T. McGrail SeS.n' .^7T; p ^-.^'rl?' and M.S. Toughened Epoxy Matrices", 34,h In, SAMPE Sympos., 259-270, (igsl)
9. C.B. Bucknall and I.K. Partridge, "Phase Separation in Crosslinked Resins Containing Polymeric Modifiers", Poly. Eng. anrf &/. , 26( 1 ), 54-62, ( 1 986)
161 10. Keizo Yamanaka and Takashi Inn.iP "Qtm . , Modified With PoWEther Su^,t:^:P:X^%ZX ^rS'^"''"^
ILXtL^r.^^:^^^^^^^ ModifiCon
14. S.A. Thompson. 7%.m, University of Massachusetts. (1987)
162 CHAPTER 4
ULTEM®/ATPEI MODIFIED EPON 828/DDS RESIN SYSTEMS: COMBINATION OF HIGHER MN PEI WITH AMINE FUNCTIONALITY
4. 1 Introdur-tinn
In previous work toughening thermosets with thermoplastics, researchers have found correlations between the morphology and the toughness''^. In Chapters 2 & 3, this
has also been observed. When phase separation was present in the Ultem® modified epoxy resins, the toughness of the cured resin was increased. When the epoxy was
modified with amine terminated poly(ether imide) (ATPEI), no phase separation was
present and no toughening occurred below a 40% loading level of the ATPEI.
Other researchers have found correlations between the toughness of the
thermoplastic modified thermoset resins and adhesion between the phases in phase
separated systems''^. In this chapter the goal is to utilize blends of the two types of
poly(ether imide)s previously studied to obtain both phase separation and good adhesion
between the two phases when used to modify the same epoxyresin system. The Ultem® will provide the phase separating capabilities and the ATPEI will provide the reactive end groups necessary for good adhesion between the phases. It is hypothesized that the combined poly(ether imide)s will show an increase in toughness not realized in the previous sections.
163 feas,.,Uy of ^.„« .He .wo ..pes of PHPs Su.se,ue„.,, o.He. co.bi„a..o„s we. .es.ea to try to optimize .he mix.ure.
4.2. 1 Plaque Fahrinatinn
In this preliminao, study, a plaque was fabrica.ed consis.ing of a .o.ai of 20 weigl,. percent poly(ether i™de) in Epon 828/DDS. Of tl>.s 20»/o. ,/3 of the weigh, was Ul.en,® and 2/3 of .he weigh, was ATPEI, .he same ma.erials described in Chap.ers 2 & 3. This
mix.ure was processed in the same way as the Ultem® modified sys.ems wi.h .he Uhem® and the ATPEI being added at the same time^ The procedure is outlined below:
1. Epon 828 was heated in a 50 ml beaker in an oil bath to a temperature of 190°C while
Stirring.
2. Ultem® pelle.s and ATPEI powder were added in .he appropria.e amoums and
allowed .0 stir and heat until the mixture was clear. This took approximately .wo
hours.
3. The number of equivalents of amine added to the resin via the ATPEI was determined
using the equation below.
164 {Wt, ATPEI (g)j/{4300 = g/eqi # of Equivalents
4. The temperature of the oil bath and the resin njxture was dropped to , TOT.
5. 4.4--dra„.nodipheny.su,fo„e was added to ™ake the ditference for I a : I stoichio.etr, of the amine to epoxide equivalents. This was allowed to mi. until „ went clear, about 5 minutes.
6. The resin mixture was transferred to a Nalgene evaporating dish. The evaporating dish containing the modified epoxy was placed in a vacuum oven at 170°C and allowed to
outgas umil no more bubbles were present, approximately 5-10 minutes, depending
on the viscosity of the resin mixture.
7. The outgassed resin mixture was removed from the oven and quickly poured into a
preheated sheet mold. The filled mold was then replaced in the oven.
The 8. cure schedule proceeded as follows:
A. 1 70°C for two hours with no vacuum
B. 230°C for two hours under vacuum to prevent system from degrading in
air.
C. The oven was turned off and the molded assembly was allowed to cool to
room temperature very slowly under vacuum to minimize residual stresses.
The plaque was removed from the mold, labeled and machined into the sizes necessary for the thermal, mechanical and morphological tests.
165 The goal of this research is to get a toughened system^ Phase separation seems to
be a requirement for this, especially at the lower loading levels of the poly(ether imide) (see Chapters 2 & 3). To determine the mon,hology. the sample was viewed under
transmission electron microscopy. The samples were prepared as in the previous sections.
The sample was ultramicrotomed at room temperature with a diamond knife and the thin
sections collected onto copper grids. The specimens were stained with osmium tetroxide
for 4 hours prior to viewing. They were then observed with a JEOL 100-CX STEM in the
transmission electron microscopy mode at 100 KeV. Photos were taken of representative
areas on the sample. The resuhs are shown below in Figure 4.1.
Figure 4. 1 TEM micrograph of 20% mixed PEI modified Epon 828/DDS
166 There is phase separation present in this mixed PEI modified Epon 828/DDS resi,
separation of this magnitude exists, it is anticipated that toughening will occur.
4.2.3 Modulus
The mechanical properties were studied to determine if there is any toughening of
the system occurring. The modulus was determined as detailed in Appendix A A sample
with a 42mm gage length, 12mm width and 3 mm thickness was loaded at 0.05cm/min to
a final outer strain of 5% on an Instron tensile testing machine. The modulus was
determined fi-om the slope of the force/deflection curve and the result shown below. A graphical comparison of this number with those for the 20% loading levels of the other
PEI modifiers is shown in Figure 4.2.
E(GPa): 2.7
As can be seen in Figure 4.2, this number is comparable to those for the other modified systems and will be used in the calculation of the fi-acture energy.
167 Modulus vs. Type of PEI Modifier at 20% PEI/E828/DDS
Neat Ultem ATPEI Mix Type of PEI Modifier
Figure 4.2 Modulus vs^type of PEI modifier at th. 20% loading level for the 20% fhuhiilJi/DDS resin systems
4.2.4 Fracture Energy
After the modulus was determined to be comparable to those of the systems
modified with the other types of poly(ether imide), the critical stress intensity factor and the fi-acture energy were determined. These two quantities were determined in the same manner as the previous samples. The compact tension sample was loaded at a rate of
0.05cm/min until the sample broke. The critical load to cause crack propagation was noted and the critical stress intensity factor assessed as described in Appendix B. A graphical comparison of the Kic's at the 20% loading level for the different types of modifiers used
168 thus far shown in . Figure 4.3. The fracture energy was calculated fro. the critical stress
the Qc's vs. type of modifier used at the 20% loading level is shown in Figure 4.4.
Critical Stress Intensity Factor vs. Type of PEI Modifier at 20% PEI/E828/DDS
S
u )2
Neat Ultem ATPEI Mix Type of PEI Modifier
Figure 4.3 vs. type Kic of modifier used at 20% loading level.
Table 4.1 Critical stress intensity factor and fi-acture energy of the 20% mixed PEI/E828/DDS
Grr (J/m^) 1.05 409
169 ure Energy vs. Type of PEI Modifier at 20% PEI/E828/DDS
Type of PEI Modifier
Figure 4.4 Gic vs. type of modifier used at 20% loading level.
It can be seen from Figure 4.4 that there is an increase in the critical stress intensity
factor over the K,c's at the 20% loading level of the other types of poly(ether imide) modifiers ("Neat" refers to the neat Epon 828/DDS epoxy resin system). The real difference is apparent in the fracture energies where the modulus is taken into account, see
Figure 4.5. There is a dramatic increase in the fracture energy of the mixed PEI modified system over any of the other systems studied previously at 20 weight percent. This demonstrates that the higher molecular weight is necessary to get phase separation, but
170 mo. .oughe.„g occu. when a reacfve end group is added fo. adhesion he.ween .he .wo phases.
4.2.5 Thermal An^jysH
For completion, the thermal properties of the mtxed PEI modified system were determined.
4.2.5.1 Thermal Stability
The thermal stability was analyzed using a DuPont Instruments TGA2950 Thermal
Gravimetric Analyzer, TGA. The sample was heated from room temperature to 600X at 20T/min. under nitrogen. The onset of and \% 50% weight loss were then determined.
The traces are shown in Figures 4.5 - 6 and the results shown in Table 4.2.
Table 4.2 TGA data for the mixed PEI/E828/DDS system
Onset 1% Weight Loss (°r) Onset 50% Weight Loss (°n 228 435
171 TGA Trace for 20% Mixed PEI Modified E828/DDS 1/3 Ultem, 2/3 ATPEI 100.5 -r
100
1 .037% Wt Loss @ 228^C (0.2078mg)
a»
99 -
98.5
100 200 300 400 Temperature (<>C)
Figure 4.5 TGA trace of 20% mixed PEI/E828/DDS showin g onset of 1% weight loss.
TGA Trace for 20% Mixed PEI Modified E828/DDS 1/3 Ultem, 2/3 ATPEI 120 -r
100 --
49.90% Wt 80 - Loss@ 435°C (9.993 mg)
I 40
20
100 200 300 400 500 600 Temperature (°C)
Figure 4.6 TGA trace of mixed 20% PEI/E828/DDS showing onset of 50% weight loss.
172 These numbers most closely resemble those of the Ultem® modified systems where the average temperature at which 1% weight loss occurs is 240T. and the
temperature a, which 50«/o weigh, loss occurs is 433T^ The weigh, loss temperatures for the ATPEI modified sys.ems are much higher than these listed here, nearly 420T for VA and 485T for 50%. The thermal stability of this material seems to be mim.ck.ng the Ultem® modified Epon 828/DDS systems.
4.2.5.2 Glass Transition Tpmp^r^^ture
The Tg of the material was determined through the use of a DuPont Instruments DSC2910 Differential Scamiing Calorimeter, DSC. The sample was ramped at a rate of lOX from room temperature to 300°C under nitrogen. The Tg was determined from the
second heating of the same sample. The DSC trace is shown in Figure 4.7 and the result shown below.
Tg(°C): 212
Still only one Tg was observed. This value coincides with all of the others seen so far in this study.
173 174 The coefficient of .he™, expansion was de.er„.„ed using a DuPon. Ins.^ts TMA2940 T.e™a, Mechanical Analyzer, TMA. The san>p,e was heated fron, .oon, temperature to 275»C at 5«C/mi„ under ™trogen and the CTE's detennined fro™ the
4.3.
Table 4.3 CTE data for 20% mixed PEI/E828/DDS
Glassy CTE(,.m/mor) Rubbery CTF ..»./^or^^
168
These numbers coincide with those determined earlier for both the ATPEI/E828/DDS and the Ultem®/E828/DDS systems.
175 o o o
176 4.2.6 Soluhility in Vc^rj^^^^^nMrnt-
The solubility of this sample was measured in the same way as those in Chapters 2
The specification & 3. used was from Boeing for high performance composite matrix
materials with aerospace applications. The modified epoxies were subjected to six
different solvents: distilled water, methylene chloride, methyl ethyl ketone, JP5 jet fuel, de-
icing fluid and Skydol hydraulic fluid. The samples were the broken halves of the compact
tension test specimens. This gave samples that were all fairly equal in size and shape.
The samples were initially placed in a one dram vial and dried in a vacuum oven
overnight. They were then weighed and this number used as the initial diy weight. The
vials were then filled with the appropriate solvent and allowed to stand at room
temperature. At specified times, the samples were removed fi-om the vials, dried quickly by
blotting with a KimWipe and the weight measured using a Sartorious balance. The sample
was then quickly replaced in the vial. The weights were recorded over a six week period
of time. At the end of the study the samples were removed fi-om the solvents and dried in a
vacuum oven for a period of two weeks. The final dry weight was then measured to determine whether any of the samples have lost weight or possibly retained some of the solvent and gained weight. Charts showing the effects of the various solvents on all of the different samples are presented below in Figures 4.9 - 4. 14. The last set of points in all charts is that of the final dried weights.
The mixed system seems to fall somewhere in the middle with all of the different
solvents. When subjected to water, jet fuel, de-icing fluid or hydraulic fluid, the modified
177 see ™.Ha,s we. „ .^..^ ^^^^^ ^^^^^^^ ^^^^
greater than the initial drv weiaht t« *u do. we,gh.. In these case, some of the solvent has been trapped w.th,n the system and could not be removed using a vacuum for two weeks.
Effect of Water on Modified E828/DDS Systems
3.50
3.00
2.50
- 4> 2.00
1.50 ~
1.00 - a
0.50 -j
30%A . mix 0.00
200 400 -0.50 " 600 800 ,000 ,200 1400 ,600 ,800
-I.OO Time(hrs) *
-1.50 ^
Figure 4.9 Effect of water on the modified E828/DDS systems.
178 —
Effect of Methylene Chloride on Modified E828/DDS Systems 70
60 --
50 --
5 40
* 0% u 0 10%U A 20%U 30%U 10 X 10%A • 20%A + 30%A < mix 1
0 1 1 h- 1 \ 1 200 400 600 800 1000 1200 1400 1600 1800 Time (hrs.)
Figure 4. 10 Effect of methylene chloride on the modified E828/DDS systems.
Effect of Methyl Ethyl Ketone on Modified E828/DDS Systems
14 -r
12 • 0% U 10%U A 20%U 30%U
10 - « 10%A • 20%A + 30%A mix
S 8
S 6
2 -
0 xhv-# -H 1 h- —I 1 1 0 200 400 600 800 1000 1200 1400 1600 1800
Time (hrs.)
Figure 4. 1 1 Effect of methyl ethyl ketone on the modified E828/DDS systems.
179 — i
1.00 -r
0.50
0.00 L,_ -I
400 600 ,00 — ^.'^^o^^ 1800 -0.50 Time (hrs.)
-1.00 -
• 0% u 1.50 10%U A 20%U 30%U * 10%A • 20%A + 30%A . mix
2.00
Figure 4. 12 Effect of JP5 jet fiiel on the modified E828/DDS systems.
Effect of Dcicing Fluid on Modified E828/DDS Systems
2.00
1.50 1
1.00
g 0.50 * d 4!* 0.00 M 1 1 — \ \ 1 \ \ -SP 1 'S 200 400 600 800 1000 1200 1400 1600 1800 ^ -0.50 Time (hrs.) S -1.50 0% u 10%U A 20%U * 30%U • K 10%A • 20%A + 30%A ... mix -2,00 i Figure 4. 13 EfTect of deicing fluid on the modified E828/DDS systems. 180 Effect of Hydraulic Fluid on ModifiedE82^;^ Systems 0.50 0.00 200 400 800 ,000 ,200 ,400 1600 1800 5 -0.50 I Tiine(hrs.) ^ -1.00 1.30 • 0% U 10%U * 20%U « 30%U « 10%A • 20%A + 30%A . mix 2.00 Figure 4. Effect of 14 Skydol hydraulic fluid on the modified E828/DDS systems. 4.2.7 Summary This mixed poly(ether imide) system has shown an increase in the critical stress intensity factor and fracture energy over those of the other PEI modified systems at 20 weight percent. The thermal properties and the solubility are comparable to the other systems. Phase separation does occur and is on the order of 0.5|im. The mechanical properties are enhanced with the addition of this mixed modifier. Next, the system will be optimized by testing different ratios of the ATPEI to Ultem® in the poly(ether imide) mix. 181 ^^^-MiaizatioiLoOJ^^ Once it was detennined .hat the mixed Ultem® and ATPEI would wo* as a toughen^ng agent for the Epon S28/DDS epoxy resin system, other ratios of mixtures were created to optimize the properties of the mix of the Ultem® to the ATPEI. 4.3.1 Fabricating PlagnpQ The following plaques were fabricated following the same procedure as in section 4.2: 1. 20% (50% Ultem®, 50% ATPEI) in Epon 828/DDS 2. 20% Ultem®, (33% 67% ATPEI) in Epon 828/DDS 3. 20% (25% Ultem®, 75% ATPEI) in Epon 828/DDS The numbers in the parenthesis indicate how much of the 20% is made up of each component. For example, 20% (25% Ultem®, 75% ATPEI) in Epon 828/DDS means that a mixture consisting of 25% Ultem® by weight and 75% ATPEI by weight was made. This mixture was added to the Epon 828 resin in an amount equal to 20% by weight. The resins were formulated following the procedure outlined in section 4.2.1. The plaques were then labeled and machined into the sizes necessary for the various tests. 182 For the remainder of this section, it will be assumed that aU samples are comprised of 20»/. of the mixed PEIs, and only the ratio of the Ultem® to the ATPEI wiU be listed to identify the different samples. 4.3.2 Thermal Analysis 4.3.2.1 Thermal Stability The onset of 1% and 50% weight loss were determined using TGA. The samples were heated from room temperature to 600X at 20°C/min. under nitrogen. A representative TGA scan can be seen in Figures 4.5 & 4.6, and the results are summarized below in Table 4.4. These values also coincide with those of the Ultem® modified Epon 828/DDS epoxy resins. No significant changes occur by altering the ratio of Ultem® to ATPEI. Table 4.4 TGA data for the various mixed PEI's in E828/DDS Mixture Composition Onset 1% Weight Loss Onset 50% Weight Loss 25% Ultem®, 75% ATPEI 235 433 33% Ultem®, 67% ATPEI 228 435 50% Ultem®, 50% ATPEI 270 432 183 2 ure The glass transition temperature was determined in the same way as in section 4.2 5.2. The samples were heated at a rate of 10»C/min. from room temperature to 275«C under nitrogen in hermetically sealed pans. The Tg's were determined from the second heat runs. A representative trace is shown in Figure 4.7 and the results are outlined in Table 4.5 Again, the Tg's of the neat materials are all so close that it is not unexpected that the Tg's of these modified materials are all similar. There is no difference between these values and those for the Ultem® or ATPEI modified E828/DDS resin systems. Table 4.5 Tg's of the different mixture of PEFs with E828/DDS Composition of Mixfiirp ^o^^ 25% Ultem®, 75% ATPEI 195 3 3 Ultem®, 67% ATPEI % 2 1 50% Ultem®, 50% ATPEI 202 184 4.3.2.3 Thejinal^xmMmCoefficiem The thermal expansion coefficients were deterrrined fron, the brolcen halves of the conrpac tension specimens. The samples were mounted m .he TMA and examined using the normal expansion probe. The samples were heated from room temperature to 275»C a 5T/mi„ under mtrogen. The CTE's were derived from second heating data. A representative trace is shown in Figure 4.8 and the results summarized in Table 4.6. The data are similar to those of all of the other PEI modified Epon 828/DDS systems Table 4 6 CTE data for the various mixed PEI/E828/DDS samples Composition of Mixture Glassv rnm/mor) CTE Rubbery CTF r.,n./..on 25% Ultem®, 75% ATPEI 64 172 33% Ultem®, 67% ATPEI 66 168 50% Ultem®, 50% ATPEI 65 167 185 4.3 .2.4 SummatxandConclusion There was no dramatic change in any of the thermal properties with the addition of more or less ATPEI in the mi«ure. This was not unexpected since the thermal data for all of the materials have been fairly similar. 4.3.3 Mechanical ProperfipQ Again, toughening of the epoxy resins is an important goal to this research. To determine whether or not changing the mixture composition has an effect on the mechanical properties, the modulus, critical stress intensity factor and the fracture energy were determined and compared to those of the preliminary mixture of the PEFs. 4.3.3.1 Modulus First the modulus was determined using a three point bending apparatus as described in Appendix A. The beams tested had a gage length of 42mm, a width of 12mm and a thickness of 3mm. The samples were loaded in three point bending to an outer strain of 5%, then unloaded. The slope of the force/deflection curve was used to calculate the modulus as described in Appendix A and the results are summarized in Table 4.7. 186 Table 4.7 Modulus data for the various mixed PEI/E828/DDS samples ComposiHon of MiYf..r^ E (GPa) 25% Ultem®, 75% ATPEI 2.4 33% Ultem®, 67% ATPEI 2.7 50% Ultem®, 50% ATPEI 2.5 A graphical depiction of these results is shown m Figure 4.15. The modulus for the 33% either higher or lower amounts of ATPEI dropped the modulus. Modulus vs. Amount of ATPEI for 20% PEI/E828/DDS 50 67 % ATPEI in ATPEI/Ultem Mix Figure 4. 15 Modulus vs. amount of ATPEI added to mixed systems. 187 4.3.3.2 Frartiir^ FngrQT The modulus of the in.t,al composition tested was the Wghest. In this section the compact tendon specimens were loaded at a rate of 0.05 cm/min until the samples broke The critical load to cause crack propagation was noted and the cnt.cal stress intensity factor calculated as shown in Appendix B. These results are shown in Figure 4. 16. Critical Stress Intensity Factor vs. Amount of ATPEI in 20% PEI/E828/DDS 50 67 Amount of ATPEI in Ultem/ATPEI Mix Figure 4. 16 Comparison of critical stress intensity factors with changing amounts of ATPEI in Mix Using the modulus and the critical stress intensity factor, the fracture energy determined as shown in Appendix B. A comparison of the fracture energies of the 188 1 summarized below in Table 4.8. Although the K,e values do not differ, once the moduli are taken into account the values Qc become different. The sample of the 20% (250/. Ultem®, 75% ATPEI) / E828 / DDS has the highest fracture energy of any of the mixed PEI modified Epon 828/DDS epoxy systems. These systems are tougher than either the 20% Ultem®/E828/DDS or the 20% ATPEI/E828/DDS systems examined in Chapters 2 & 3. Table 4.8 Critical stress intensity factors and fracture energies for the different mixed PEI systems Composition of Mixture K,r (MWm'") Qic U/m') 25% Ultem®, 75% ATPEI 1 .04 45 33% Ultem®, 67% ATPEI 1 .04 401 50% Ultem®, 50% ATPEI 1 .03 424 189 Fracture Energy vs. A mount of ATPEI in 20o/o PEI/E828/DDS % ATPEI in ATPEIAJItem Mix Figure 4 . 17 Comparison of fracture energies with changing amounts of ATPEI in Mix 190 ^.MNlorphoiogy These systems should also be phase separated. The fracture surfaces of the miniature compact te„s.o„ samples were observed with scaling electron microscopy The SEM photos of the fracture surfaces are shown in Figures 4,8- 4.20. The samples were prepared as in previous sections. The fracture surface of a compact tension test specm,en was removed using a coping saw. then mourned on a SEM stub using either carbon or silver paint. The sample was sputter coated with a thin layer of gold to encourage conduction of electrons. It is apparent from all three of the photos that the systems are phase separating. The matrix is still becoming more ductile with the incorporation of the poly(ether imide)s, implying that the systems are not completely phase separated and there is still some poly(ether imide) within the epoxy matrix. There is evidence of good adhesion between the two phases. While there are raised bumps on the surface, no dimples are evident. This leads to the conclusion that the inclusions were pulled apart rather than pulled out of the matrix material. This is especially evident in Figure 4.20. Not only are the raised bumps evident, there are craters around each bump signifying a stretching and pulling apart of the inclusions, rather than a pulling out of them. 191 Figure 4. SEM 18 micrograph of the fracture surface of 20% (33% Ultem® 67% ATPEI)/E828/DDS 192 Figure 4. SEM 19 micrograph of the fracture surface of 20% (25% Ultem® 75% ATPEI)/E828/DDS 193 Figure 4.20 SEM micrograph of the fracture surface of 20% (50% Ultem® 50% ATPEI)/E828/DDS 194 4.4 Machinahility As can be seen in Figure 4.21 below, there is an improvement in the machinabUity of the epoxy resin systems when modified with the Ultem®/ATPEI mix, Machimng the neat epoxy resin results in a powdery residue (see Figure 2.53). The epoxy is brittle and subject to breaking, chipping and cracking unless extreme care is taken during the machining process. The mixed PEI modified systems resuh in less powdering, with neat spirals of material upon machining, and less likelihood of chipping or breaking. Figure 4.21 20% mixed PEI modified E828/DDS after drilluig. Note the spirals of material and a reduced amount of powdery debris over the neat E828/DDS. 195 I. was seen tha. the ther„.> properties of the th.ee different forntulations of the ™ixed PE./ES2S/DDS resin systents did not vary wide,. Their properties were on average consisted of Tg's around 200»C. onsets of 50% decomposition around 434T. glassy CTE of 65^t^Vm«C and mbbery CTE's around 170,m/mT. The modulus of the 20% mixed system with 33% Ultem® and 67% ATPEl was slightly higher than the rest of the samples (2.7 vs. 2.5 GPa), but the fracture energy of the system containing 25% Ultem® and 75% ATPEI was higher than the others at 45 1 W vs. 424 J/m^ for the next closest. All three systems showed a phase separated morphology and good adhesion between the two phases. The inclusions were stretched and pulled apart rather than out of the matrix. In conclusion, the Epon 828/DDS epoxy resin system can be successfuUy toughened with a combination of the unreactive poly(ether imide) Ultem® and the reactive amine-terminated poly(ether imide). The thermal properties are comparable to those of the Ultem® modified systems, but the fracture energies are much higher. 451 J/m' vs. 353 at the 20% loading level by weight of each The systems are phase separated with inclusions on the order of 0.5nm. And aU was accomplished without the use of any solvents. 196 4,6 Referenrfig 2. H.G. Recker, V. Altstadt, W. Eberle T Folda D Gertli w w u Greggory 3. S. Bennett, Ph.D. Thesis, University of Massachusetts, (1991) P^^^' ?19^)'^^'' ' ^' 34(23), 4832-4836, 5. R.A. Pearson, A.F. Yee, Polym. Mater. Sci. Eng., 63, 311, (1990) 6. G.R. Almen, P. Mackenzie, V. Malhotra, R.K. Maskell P T McGrail M 9 ^^ftnn^"""^ 20th ' Int. SAMPE Tech. Con/., 46-59, (1988) J.L. Hendrick, 7. I. Yilgor, G.L. Wilkes, J.E. McGrath, Polym. Bull., 13, 201, (1985) 197 CHAPTER 5 REACTION OF A mOHER MOLECULAR WEIGHT AMINE TERMINATED POLY(ETHER IMIDE) WITH EPOXY RESINS: EPON 828/DDS MODIFIED WITH A HIGH MN ATPEI SJJntroduction G.R. AJmen et al. have determined certain properties that a thermoplastic must possess in order to have the most impact in toughening a thermoset^. Several of these ar that the thermoplastic should have a high glass transition temperature, Tg, and modulus, reactive end groups, and be compatible with the thermoset. Many researchers have also observed that phase separation is a necessity for thermoplastic toughening of thermosets^-^. Bemiett^ has concluded that sharp boundaries between the included and continuous phases are also important for optimum toughening. The results that have been seen thus far in this study are: 1 Ultem® modified . E828/DDS results in toughened, phase separated systems with sharp phase boundaries (676 J/m^ vs. 265 J/m^ for the neat Epon 828/DDS). (Chapter 2) 2. When the E828/DDS is modified with an amine terminated poly(ether imide), ATPEI, of 8600 number average molecular weight, no phase 198 separation is present and the effect on toughening is less than that of the Ultem/E828/DDS systems with incorporations up to 30o/o by weight (192 J/m^ for the 30% ATPEI/E828/DDS vs. 468 JW for the 30O/O Ultem(e)/E828/DDS). At 40% there is little difference between the two systems (676 jW for the Ultem®/E828/DDS and 700 for the ATPEI/E828/DDS). (Chapter 3) Combining 3. the two materials, Ultem® and ATPEI, in varying amounts and using this as the modifier results in systems that are phase separated and tougher than either of the other two systems (451 jW for the mixed PEI modified system vs. 176 J/m' for the ATPEI modified epoxy or 353 J/m^ for the Ultem® modified epoxy all at 20 weight percent incorporations). (Chapter 4) In situation 1, phase separation with sharp phase boundaries are present, but no reactive end groups are present to react with the epoxide groups of the Epon 828. In situation 2, there are reactive end groups but no phase separation. This results in diminished properties. With situation the 3, phase separation is present with sharp phase boundaries, and reactive end groups. It seems the molecular weight is playing a role in the phase separation behavior, and the reactive end groups are enhancing the adhesion between the two phases in the PEI modified epoxy systems. This chapter investigates combining high molecular weight with reactive end groups in one material which will be used to modify the epoxy resins. This 199 will be accomplished by using a h,gi,er Mn ATPEI. The molecular weigh, of ,he high ..olecular weight ATPEI (HMWA) is 25,500 and it .s endcapped with amine groups. 5.2 Property P-etenninaimikHieNgMj^ The properties of the neat HMWA will be determined in this section to characterize the material before it is put to use modifying the epoxy resin 5.2. 1 Thermal Stability The thermal stability of the HMWA powder was analyzed using a DuPont Instruments TGA2950 thermal gravimetric analyzer, TGA.. The sample was heated from room temperature to 600X at 20°C/min. under nitrogen. The onset of 1% and 50% degradation were determined. The traces are shown in Figures 5.1 & 5.2 and the results are listed in Table 5.1. Table 5.1 Thermal stability data of HMWA Onset 1% Weieht Loss (°n Onset 50% Weight Loss (°C) 454 N/A 200 o o o T-OO ! inD- Q >in CO o o I o m u 0) CO £- 3 O (D -113 a "53 E O tD O H CXJ r-H O CCO o o o CO o O o T o H t-H 01 B in o) 01 a; 201 202 The onset of X% weight loss occurs at a higher temperature than the ATPEI (4230C), but lower than the Ultem® (469°C). SQo/o No weight loss was observed over the temperature range studied, as was the case with the ATPEI and the Ultem®. AU of the PEI's tested are stable to high temperatures. 5.2.2 Glass Transition Temperature The Tg was determined using a DuPom Instruments DSC2910 Differential Scamiing Calorimeter, DSC. The sample was heated at a rate of lOT/min. from room temperature to 275°C and the Tg determined from the second heat data. The DSC trace is shown in Figure 5.3 and the result shown below. The Tg of this material is the same as the Ultem (217°C) and higher than the ATPEI (195°C). TgrC): 216 5.2.3 Solubility in Epon 828 1.47 of the g HMWA was added to 5.9 g of the Epon 828 at a temperature of 170°C and allowed to stir and heat to determine if the mixture was compatible. It took nearly an hour for this to occur, but the mixture of the Epon 828 and HMWA did become clear, indicating the two materials were miscible and the HMWA was soluble in the Epon 828. 203 iJJabricatingPlagues Plaques of this material were fabricated in the same mamier as the other PEI modified epoxies in this dissertation. 1 Epon was 828 heated in a 50 ml beaker in . an oil bath to a temperature of 1 70°C while Stirring. 2. HMWA powder was added to make up either 2Q% or 30o/o by weight and allowed to stir and heat until the mixture was clear. This took approximately one hour. The number 3. of equivalents of amine added to the resin via the HMWA was determined from the equation below. (Wt. ATPEI (g)}/{ 12750 g/eq) = # of Equivalents 4. 4,4'-diaminodiphenylsulfone was added to make up a 1 : 1 stoichiometry of the amine to epoxide groups. This was allowed to mix until it became clear, about 5 minutes. 5. The modified epoxy was poured into an evaporating dish which was placed in a vacuum oven at 170°C and allowed to outgas until no more bubbles were present, approximately 10 minutes. 6. The outgassed resin mixture was removed from the oven and quickly poured into a preheated sheet mold. The filled mold was then replaced in the oven. 7. The cure proceeded as follows: A. 1 70°C for two hours with no vacuum 205 B. 230X for two hours under vacuum to prevent system from degrading in air. C. Oven turned off and the molding apparatus allowed to cool to room temperature very slowly under vacuum .0 minimize residual stresses. This procedure produced cured plaques of material that were an opaque tan/brown in appearance. These plaques were labeled and subsequently machined into samples for thermal and mechanical analysis. 5.4 Thermal Analysis The plaques of HMWA/E828/DDS were examined and the various thermal properties determined. 5.4.1 Thermal Stability The thermal stability of the material was determined using the TGA. The sampL were heated from room temperature to 600°C at 20°C/min under nitrogen. The traces < shown in Figures 5.4-5 and the results shown below. 206 207 o o o cu oc oo Q 1 I O to o o dp o -CI s H in 208 Table 5.2 TGA data for the HMWA modified E828/DDS systems OnseUAffiaghtLoss li^i^ QnMiffiiWeightLoss 20 220 30 453 These numbers are comparable to the Ultem modified systems (~200T/420«C). The ATPEI modified systems showed much higher thermal stability (~430°C/485T) 5.4.2 Glass Transition Temperature The glass transition temperature of the HMWA/E828/DDS materials were determined using differential scanning calorimetry. The samples were heated from room temperature to 350°C at lOT/min. under nitrogen. The Tg's were taken from the second heat data. The results Usted below in Table 5.3. 209 Table 5.3 Glass transition temperatures of HMWA/E828/DDS systems %HMWAJnE828™s ,o^^ 20 201 30 208 These numbers are comparable with the other PEI modified E828/DDS systems. 5.4.3 Thermal Expan.sinn Coefficients The coefficients of thermal expansion were determined using a DuPont Instruments TMA2940 Thermal Mechanical Analyzer. The sample were heated at a rate of 5°C/min from room temperature to 275°C under nitrogen. The coefficients of thermal expansion were measured from the second heating through the thickness of the samples using the macro expansion probe. A representative trace is shown in Figure 5.6 and the results outlined in Table 5.4. 210 211 Table 5.4 Glassy and rubbeiy CTE's for the HMWA modified E828/DDS 01..^^ ^,.,,0,^^ 67 172 These numbers are also in the range of all of the other samples for both the rubbery and glassy coefficients of thermal expansion. 5.5 Mechanical Prop ertie«; As in all the previous cases, one very important goal is to toughen the epoxy resin with the modifying material. The toughness of the HMWA/E828/DDS systems will be examined in this section. 5.5.1 Modulus The modulus was determined following the procedure outlined in Appendix A, with the exception that only one sample of the 30% HMWA/E828/DDS was examined due to the amount of material available. The bar sample with a 42 mm gage length, 12 mm 212 The results are shown in Table 5.5 Table 5 .5 ModuH of the HMWA/E828/DDS systems % HMWA in FR9R/nps EiGPa): 20 2.5 30 2.3 The results are comparable to those that were determined for the other types of PEI modifiers. 5.5.2 Fracture Energy The critical stress intensity factor and the fi-acture energy were determined as outlined in Appendix B. The compact tension specimens were loaded at a rate of 0.05 cm/min. until the sample failed. The critical load to crack propagation was noted and the critical stress intensity factor calculated fi-om this as shown in Appendix B. Using this critical stress intensity factor and the modulus determined in the previous section, the 213 fracture energy was calculated. The results are shown below in Table 5.6 and the graphical types of modifiers all at 20 weight percent are shown in Figures 5.7 - 5.8 Table 5.6 Critical stress intensity factor and fracture energy data from compact tension 20 1.07 458 30 1.03 461 The Kic of the 20% can be seen to be sUghtly higher than that for the mixed PEI modified system (1 .07 vs. 1 .05 MN/m"\ and significantly higher than either the Ultem® (0.94 MNW^) or the ATPEI (0.69 MN/m'") modified systems. The Kic of the 30% PiMWA/E828/DDS is comparable to the Ultem® (1.06 MN/m'^') and again much higher than the ATPEI modified system (0.72 MN/m'''). The Gic's are comparable to the 30% Ultem®/E828/DDS (468 JW) and much higher than the 30% ATPEI/E828/DDS J/m (192 ). The morphology of this system will be investigated next to explain these results. 214 Critical Stress IntensitTFaiti,^^^ at 20% PEI/E828/DDS 1.2 a 0.6 u fa? 0.4 0.2 Neat Ultem ATPEI Mix HMWA TypeofPEI Modifier Figure 5.7 Critical stress intensity factor vs. type of modifier at 20% weight ^percent loading level Fracture Energy vs. Type of PEI Modifier at 20% PEI/E828/DDS Neat Ultem ATPEI Mix HMWA Type of PEI Modifier Figure 5.8 Fracture energy vs. type of modifier at 20% weight percent loading level 215 5.6 Morphnlngy 5.6.1 Phase Separation The phase separated morphology was examined through the use of transmission electron microscopy. The samples were ultramicrotomed at room temperature with a diamond knife and the sections collected onto copper grids. The samples were then stained for four hours in osmium tetroxide prior to viewing. The grids were observed with a JEOL 100-CX STEM in the TEM mode at 100 KeV. All sections were viewed and photos taken of a representative area. The micrograph of the HMWA^828/DDS samples are shown in Figures 5.9 and 5.10. As can be seen from Figure 5.9, the system is exhibiting a co-continuous phase morphology at the 20% level, with areas of thermoplastic rich inclusions m an epoxy rich matrix and areas of epoxy rich inclusions in the thermoplastic rich matrix. Having this type of morphology at this low a loading level is incongruous with the rest of this study. This morphology is not usually seen until at least 30% incorporation of the poly(ether imide). 216 Figure 5.9 TEM micrograph of 20o/o HMWA/E828/DDS showing a co-continuous morphology. The dark areas are the poly(ether imide). Figure 5.10 TEM micrograph of 30% HMWA/E828/DDS showing a phase inverted morphology. There are now epoxy rich inclusions in a PEI rich continuous phase. The dark areas are the poly(ether imide). 217 At 30»/. incorporation of HMWA. the systems sliowed the phase inverted morphology comparable to that of the 40% Ul,em®/E828/DDS (see Figure 5. 10). These findings are most likely due to the larger molecular weight of th.s po,y(ether imide). The Ultem®, with a Mn of .7,000, phase separated. The ATPEI with a Mn of 8600 did not. The HMWA with a Mn of 25,500 phase separated and showed a mon^hology usually demonstrated only at higher loading levels. 5.6.2 Fractnrp. S,urf!^r^ The fracture surface of the miniature compact tension test samples were also observed to determine the type of fracture that occurred. The samples were prepared in the same way as in previous chapters. The fracture surface was removed from the sample with a coping saw and mounted on an SEM stub using carbon paint. The sample was gold sputter coated to reduce charging. The sample was then viewed with the JEOL 100-CX STEM in SEM mode at 20 KeV. The entire sample was viewed and photos taken of representative areas on the surface. The micrographs can be seen in Figures 5.11-5.13. It can be seen from the micrographs in Figures 5. 11 & 5. 12 that the fracture has occurred in a ductile manner. There is much deforming of the fracture surface. But Figure 5. 12 also shows that there are also some inclusions in the smoother, less ductile areas. This coincides with the TEM micrograph showing the co-continuous morphology. It is to be expected that some areas, where the HMWA is the continuous phase, would fracture in 218 a more ductile and son,e areas, those in wMch _ the epoxy is the con..nuous phase. again good adhesion between the phases can be seen^ There are buntps, hut no cavities. And there is evidence of craters around the bun,ps or particles indicating that the materia, was stretched and broken rather than pulled out. Figure 5.11 SEM micrograph of fracture surface of 20% HMWA/E828/DDS magnified lOOOX. 219 SEM . 12 micrograph of fracture surface of 20% HMWA/E828/DDS magnifiedia^iuiicu 10,000X. 220 Figure 5.13 SEM micrograph of fracture surface of 30% HMWAy^828/DDS magnified 221 A. 30% HMWA/H828/DDS, chere ,s again evidence of a ductile fa„ure. Th,s is true of .he entire sample This would be expected with the phase inverted morphology seen in the TEM micrographs previously. 5.7 Machinahility As can be seen in Figure 5, 14 below, there is an improvement in the machinability of the epoxy resin systems when modilicd with the high molecular weight amine terminated poly(elher imide). HMWA, Machining the neat epoxy resin results in a powdery residue (see Figure 2.53). The epoxy is brittle and subject to breaking, chipping and cracking unless extreme care is taken during the machining process. The HMWA modified systems result in very little powdering, with neat spirals of material upon machining, and less likelihood of chipping or breaking 222 Figure 5.14 20% HMWA/E828/DDS after drilling. Note the neat spirals of material with very little powdery debris on the surface. 223 5.8.Surnmary^nd_Co^^ The above data has shown a material that has a modulus and thermal properties comparable to those of the Ultem modified systems and the Ultem/ATPEI mixed modified systems. But the fracture energy is much greater at the 20% level of incorporation, 458 J/m^ vs. 409 or 353 It is comparable to the fracture energy of the 30o/o Ultem/E828/DDS system where the co-continuous morphology exists, 468 JW. The 20o/o HMWA/E828/DDS system also exhibits a co-contmuous morphology, not usually found at this low a loading level of the poly(ether imide). The SEM data backs this up and also shows that good adhesion between the phases is present. Significant improvement in the fracture energy with the addition of 30% HMWA to the Epon 828/DDS over that of the 20% HMWA/E828/DDS is not realized. At 30% loading level, the fracture energy is similar to that of the 20% loading level, even though a phase inverted morphology is present with inclusions of an epoxy rich phase in a continuous poly(ether imide) rich phase. Some researchers have found that the highest toughening occurs with a co- continuous morphology^'', while others have found the best toughness with a phase inverted morphology^ The Kic values seem to support the former for this poly(ether imide) modified system. Hendrick' et. al. determined that a maximum in toughness was associated with a strong interface with a polysulfone modified DGEBA/DDS type epoxy resin system. In this chapter, a high molecular weight amine terminated poly(ether imide) was used as the modifier with the intent of getting a phase separated morphology and 224 When .he various PEI. are compared a. s™,ar loading ievels, the HMWA toughens the epoxy resin system most effectively. From all of the systems studied ,n this dissertation, this material seems to have the best potential for toughening the Epon 828/DDS epoxy networks^ The HMWA dissolves properties than the non-reactive Ul.em. And the co-continuous morphology occurs at a lower loading level as well. 225 5.9 ReferenrpR H.G. 2. Recker, V. Altstadt, W. Eberle T Folda D Gprth w tr i Cho, J.W. Hwang, K. Cho, J.H. An, C.E. 09^) Park, Polymer, ZAi:!^^ 4832-4836, R.A. 4. Pearson, A.F. Yee, Polym. Mater. Sci. Eng., 63, 311, (1990) Greggory 5. S. Bennett, Ph.D. Thesis, University of Massachusetts, (1991) ^ ^ S^ft^"' P T. McGrail, S.P. WUkinson Soc.W Adv.L^r?'Mater. Proc.f Eng. 33rd Internal Symp., 979-989, (1988) J.L. Hendrick, 7. I. Yilgor, G.L. Wilkes, J.E. McGrath, Polym. Bull, 13, 201, (1985) 226 CHAPTER 6 CONCLUSIONS AND FUTURE WORK 61 ConclnsinnQ Throughout this work the hypothesis that high molecular weight is needed for phase separation and a reactive end group on the thennoplastic is necessaty for better adhesion between the phases for thermoplastic toughened thermoset systems has been tested. Initially it was unclear that the molecular weight was playing a role in the phase separation, but this has been shown to be true. And the best fracture energy data has come from the systems where the phase separation is present and there is good adhesion between the two phases. Much work has been seen in the literature on thermoplastic toughening of thermosets or more particularly, epoxies'"^'. But in most of these studies, the thermoplastic has been blended into the epoxy resin through the use of various solvents. More success has been reported when a thermoplastic with reactive end groups is reacted into the epoxy matrix''"^'. Work by Bennett^' has shown that it is possible to incorporate thermoplastics into epoxy systems without the use of any solvents by reacting oligomers of polyaryletherketone into commercial epoxy resins. The epoxy acts as the solvent for the thermoplastic. Using this work as a guide, thermoplastics with higher molecular weights 227 (up to 25,500) and higher Tg values (195-220OC) have been reacted into the commercial epoxy resin system, Epon 828/DDS. These Wgher molecular weight and higher Tg poly(ether imide)s have provided new challenges. And because of this work, there IS now a better understanding of the mechanisms of the toughening epoxy resm wUh these types of thermoplastics, and the structure - property relationships involved therein. In Chapter 2, the unreactive, commercial poly(ether imide), PEI, Ultem® 1000- 1000 was blended into a diglycidylether of bisphenol A, DGEBA, based epoxy resin. This was the commercial Epon 828 with diaminodiphenylsulfone used as the curing agent. The neat epoxy resin system has a modulus of 2.3 GPa, a critical stress intensity factor, , as measured by compact tension testing of 0.78 MN/m"\ a fracture energy, Gic, of 265 ]/m\ and a glass transition temperature, Tg, around 205°C. The values for the neat Ultem® 1000-1000 are generally much higher than those of the neat epoxy resin. The modulus is 2.7 GPa, the critical stress intensity factor, K,c, is 3.09 MN/m"\ and the fracture energy, Gic, is 3530 JW. The Tg is very similar at 217°C. The Ultem® was blended into the epoxy in amounts ranging from 10 to 40% by weight. The greatest effects on the critical stress intensity factor and the fracture energy occurred with the incorporation 40% level. The Kic value nearly doubled (1.4 vs. 3/2 .78 MN/m and the fracture ) energy increased by 2.5 times (676 jW vs. 265 JW). While some researchers have reported a Unear increase in the critical stress intensity factor with increasing amounts of PEI addition, regardless of the morphology^ that was not seen here. The Kic's increased to .95 MN/m^'^ at the 10 and 20% incorporations and increased to 1.06 MN/m^^^ and 1.40 MN/m^^^ for the 30 and 40% loading levels respectively. The 228 modulus showed an increase from 2.3 GPa for the neat epoxy resin ,o 2 9 GPa at 40% incorporation of Ultem®. The increase in the critical stress intensity factors coiresponds to a change in the phase separated morphology. At 10 and 10% loadmg levels of the Ultem®, the systems exhibit phase separation with PEI rich inclusions in an epoxy rich continuous phase. There is evidence that the phase separation is not complete (see Figure 2.22). At the 30% loading level, the systems becomes co-continuous with areas of PEI rich inclusions in an epoxy rich continuous phase and areas of epoxy rich inclusions in a PEI rich continuous phase. The system phase inverts by 40% to give epoxy rich inclusions in a PEI rich continuous phase. In this instance, the best toughness properties occur with the phase inverted morphology. This concurs with the findings of Cho et al. with PEI modification of a TGDDM based epoxy resin^ and with Hourston and Lane'' for PEI modification of trifunctional epoxy systems. However, Hourston, Lane and MacBeath^' found no trend in the toughness with increasing amounts of Ultem®, or with the changing morphology. There was not a large variation in the thermal properties of these systems with the addition of the Ultem®. The Tg's stayed fairiy constant around 205 - 210°C. The onset of weight 50% loss occurred at slightly higher temperatures with an increasing amount of Ultem, fi-om 412°C for the neat epoxy to 452°C for the 40% loading level of Ultem®. The onset of 1% weight loss did vary. With small amounts of Ultem addition, less than 30%, there was a decrease in the 1% weight loss temperature fi-om 282°C down to 196°C. At 30 and 40% loading levels, the temperature at which the onset of 1% weight loss occurred came back up to that of the neat epoxy. The glassy and rubbery coefficients of thermal 229 expansion remained steady a. 67 and ,69 ,™/m"C respectively, regardless of the morphology of the system. The machinablilty of the samples became easier as more Ultem® was added to the blend. Initially, with the neat epoxy, the samples powdered and cracked upon machining unless extreme care was taken. With the addition of the Ultem®, the samples no longer powdered, but neat spirals of material were produced. Also, much less cracking and chipping was observed with the Ultem® modified resins. In Chapter 3, an amine terminated poIy(ether imide), ATPEI, of number average molecular weight 8600 was reacted into the same epoxy resin system. The anticipated effects of higher toughness values using a reactive thermoplastic were not realized. The modulus did continue to increase with the addition of the ATPEI, from 2.3 to 2.8 GPa. But the critical stress intensity factor, Kic, and the fracture energy, Gic, both decreased with the addition of ATPEI up to 30%. Kic decreased from 0.78 to 0.69 MN/m'^, while Gic decreased from 265 to 176 jW. An anomaly occurred with the reaction of 40% by weight of the ATPEI into the Epon 828. The Kic and the dc increased dramatically to 1.4 MN/m^^ and 700 jW respectively, numbers similar to those found at 40% Ultem® blended into the epoxy. The morphology of all of the epoxy samples reacted with ATPEI were studied. At 10% ATPEI in E828/DDS phase separation was present on a small scale. Thermoplastic inclusions on the order of 0. 1 ^im were visible in an epoxy rich matrix. For the other conditions, no phase separation was present. Other researchers have found phase 230 separation necessao^ for tougherung of thermoplastic modified "."." epoxy resins' This was true of this system as well, at least up to 30% ATPEI reacted into the epoxy. I. seems the 0. mm inclusion size is below that necessary for toughemng. At 40% there is enough of the more ductile poly(ether imide) in the system to enhance the toughness even without phase separation. Because it was believed that phase separation is necessaiy for the toughening of the thermosets, curing and processing conditions were varied to try to cause phase separation. This strategy did not work. Phase separation could not be forced by changing the processing or curing conditions. The Tg of the ATPEI modified materials remained around 210°C. But the onset of 1% and 50% weight loss occurred at higher temperatures than the Ultem® modified systems in Chapter 2. The onset of 1% weight loss was seen at temperature near 430°C, while the onset of 50% weight loss occurred around 495°C. The glassy and rubbery coefficients of thermal expansion were nearly identical to those of the Ultem® modified systems at all levels of ATPEI incorporation up to 30%. At 40%, the glassy CTE dropped to 59 |im/m°C and the rubbery CTE increased to 179 ^im/m°C. The machinability of these samples was also enhanced, but not to the extent of the Ultem® modified epoxy systems. In Chapter 4, Ultem® and ATPEI were combined and used as the modifier of the epoxy systems. Several different ratios of the Ultem® to the ATPEI were examined to determine the optimum ratio for toughening the epoxy systems. All three systems were similar in their properties, with the 25% Ultem® / 75% ATPEI system having a slightly 231 higher fracture energy than the other two systems (45 1 vs. 424 or 409 J/m^ These numbers are all hrgher than either the Ultem® modified or ATPEI modified epoxy resin systems at 20% by weigh, incorporation of the PEI (353 JW and 176 ,/n.^ respectively). These Ultem®/ATPEI modified systems all showed phase separated mon^hologies with poly(ether imide) rich included phases in epoxy rich continuous phases. The inclusions were on the order of 0.5,m. SEM photos of the fracture surfaces also shows that there is good adhesion between the phases. The panicles are stretched and broken rather than being pulled out of the matrix. And finally, in Chapter 5, a high molecular weight amine terminated poly(ether imide) was reacted mto the Epon 828/DDS epoxy resin system in amounts equal to 20 & 30% by weight. The moduli of the two cured materials was similar, 2.5 GPa for the 20% modification and 2.3 GPa for the 30% loading level. The fracture energies were also quite close at nearly 460 Vm\ At 20%, this number is higher than that of any of the other systems studied in this research. For 30%, this number is very comparable to that of the 30% Ultem® or 30% Ultem®/ATPEI mix. And this does constitute nearly a 2-fold increase in the fi-acture energy over the neat epoxy resin. The morphologies of these systems were unexpected. At 20% loading level, a co- continuous morphology exists. This has not generally been noted in systems with loading levels less than 30%^^''^''^ At 30% incorporation of the HMWA, the system is phase inverted with epoxy rich inclusions in a PEI rich continuous phase. Some researchers have found that optimum toughening occurs with the co-continuous morphology^^'^*' , while 232 others have seen the Wghest increases in toughness with the phase inverted mon.hCogy'. In this instance, the morphology does not seem to make a difference in the fracture energy. This work has shown that phase separation is a necessity for increasing the toughness with poly(ether imide) modified epoxy resms at loading levels up to 30o/o. The best numbers are obtained whh the systems that are not only phase separated, but also in which there is good adhesion between the two phases. At the 10% loading level the high molecular weight poly(ether imide) toughened the systems better than the other types of PEI modifiers. This corresponds to the co-continuous morphology. Linear increases in K.c with the addition of PEI were not seen, as was reported by Bucknall et. al.«. But neither can it be stated that the best toughening corresponds to a given morphology in all of the situations studied here. The highest increases in fracture energy were realized with the 40% incorporation of either the unreactive Ultem® or the lower molecular weight amine terminated poly(ether imide), irrespective of morphology and phase separation. More work must be done with the other systems, especially the higher molecular weight amine terminated poly(ether imide), to determine if this is true in all cases. In conclusion, the thermoplastic poly(ether imide) can be used as a toughening agent for the DGEBA based epoxy resin system cured with DDS. Increases of over 2.5 times were realized with some of the PEI's studied. There is no corresponding decrease in the modulus or glass transition temperature as would be seen with rubber modification. These modified systems have higher glass transition temperatures than those studied in this 233 lab in the pas. (2,0«C vs. ,60»C)- All of this work has been done without the use of solvents. The limiting factor in the processing of these materials will be the viscosity of the heated resins. And finally, a better understandmg of the structure/property relationships in these systems has been gained. 6.2 Future WnrV Although much has been accomplished through this work, there are still many directions in which this work can proceed. A few of them are outlined here. As was stated in the last section, a more thorough study of the HMWA resin over a more complete range of loading levels would give a better understanding of the structure/property relationships with poly(ether imide) toughened DGEBA type epoxy resins. With the lower molecular weight materials, there does seem to be a dependence of the fracture energy on the morphology of the material, up to intermediate loading levels. At higher loading levels the (40%), fracture energies for the phase inverted system and the system that did not phase separate were identical. Does the higher molecular weight amine terminated poly(ether imide) modified DGEBA type epoxy resin also show no dependence on phase separation at higher loading levels? The properties of these materials have all been studied at room temperature, but these modified epoxy systems are designed to be able to have upper use temperatures around 200°C. It would be very interesting to see how these materials will behave at these 234 higher temperatures. It is proposed that research be done on the mechanical properties at higher temperatures. Conduct three pomt bend and compact tension tests on an Instron with an environmental chamber where the temperature can be controlled and determine the high temperature modulus and fracture energy. And finally, a very important test that should be done is putting these materials into actual composites to see how well the properties translate from the neat resin systems to the composite matrix material. It is feasible that the morphology will be affected with the constraints imposed by the reinforcing material. Also the stresses in that confined environment will be different. Previous studies have been done on the wetting characteristics of an epoxy modified with a tertiary butyl PEEK oligomer with a carbon fiber''. It was seen that the epoxy phase preferentially wet the carbon fiber. If this is also true with the poly(ether imide) modified epoxies, it is expected that the final morphology of the system will be different form that seen here. This may also affect the mechanical properties of the composite. 235 6.3 Referenrp! 1_G.R. Almen P. Mackenzie, V. Malhotra, R.K. Maskell P T A Thermoset/Thermoplastic Alloys: An Approach to Tough High Matrices", 20th Int SAMPE Tech. Conf A6-59, (1988) 2. M^S. Sefton, P.T. McGrail, J.A. Peacock, S.P. Wilkinson, R.A R.M. Byrens, P.D. Mackenzie, R.K. Maskell, P.T. McGrail and M S ^Family of New Toughened Epoxy Matrices", S4th Int. SAMPE Sympos., 6. Keizo Yamanaka and Takashi Inoue, "Structure Development in Epoxy Resin Modified with Poly(Ether Sulphone)", Polymer, 30, 662-667, (1989) 7. Clive B. Bucknall and Ivana K. Partridge, "Phase Separation in Epoxy Resins Containing Polyethersulphone", Polymer, 24, 639-644, (1983) 8. Clive B. Bucknall and Adrian H. Gilbert, "Toughening Tetraflinctional Epoxy Resins Using Polyetherimide", Polymer, 30, 213-217, (1989) 9. J.B. Cho, J.W. Hwang, K. Cho, J.H. An and C.E. Park, "Efifects of Morphology on Toughening of Tetraflinctional Epoxy Resins with Poly(Ether Imide)" Polymer 34(23) 4832-4836,(1993) 10. Ronald S. Bauer, "Toughened High Performance Epoxy Resins: Modification with Thermoplastics. I.", 34th Int. SAMPE Sympos. , 899-909, (1989) 11. Douglas J Hourston, Jacqueline M. Lane, Nigel A. MacBeath, "Toughening of Epoxy Resins with Thermoplastics. II. Tetraflinctional Epoxy Resin-Polyetherimide Blends", Polymer International, 26, 17-21, (1991) 12. D.L. Hourston, J.M. Lane, "Toughening of Epoxy Resins with Thermoplastics: 1. Trifiinctional Epoxy Resin-Polyetherimide Blends", Polymer, 33(7), 1379-1383, (1992) 236 -d Properties of A™„e Te™tterPdy(C1 aL^^^^^^^^^ Jf?" Resin Syste.s"51^/„rI:5:pS«St09LT^ """"'^ ^^"^ ^ Performance Materials", Polymer, 32(9), 1633-1641, (1991) 16. Greggoo' S. Bennett, Ph.D. Thesis, University of Massachusetts, (1991) S.A. 17. Thompson, Ph.D. Thesis, University of Massachusetts, (1987) ^ Heckmam,, P H "esch T tebrr Tt' ° * Itteman, "^"^^ T'^':!'- Con/., (1989) 283-293, 19. R.A. Pearson, A.F. Yee, Pofym. Mater. Sci. Eng., 63, 31 1, (1990) 20. G^R. Almen, R.K. Maskell, V. Malhotra, M.S. Sefton, P.T. McGrail, S.P. WUkinson, ioc. Adv. Mater Proc. Eng. 33rd Internat. Symp., 979-989, (1988) 237 . APPENDIX A THREE POINT BEND TESTING FOR MODULUS DETERMINATION The flexural modulus was detemuned using a three point bend test following ASTM D790M. The material tested was in the form of a bar with dimensions of 63mm I2.5mm X 3mm. These tests were run on an Insfron Model TTBM screw-driven tensile testing machine with a SM500 500 pound load cell. The three point bend apparatus consisted oftwo flat supports with a span of 42mm, and the central unit, a 4.8mm diameter steel rod which applied the force to the beam. The bar was stressed until the outer elements of the bar reached 5% strain, and the modulus determined from equation A.1 Al Where E = flexural modulus (Pa) L = beam span (m) M ^ slope of the force/deflection curve (N/m) B = beam width (m) D ^ beam depth (m) The signal was collected into LabTech Notebook using a PC for data analysis. Three samples were tested for each condition and the average result reported. 238 APPENDIX B FRACTURB ENERGY DETERMIK.TON TROUGH COMPACT TENSION lESTING The fracture energy is determined from a miniature compact tension test. Tl^s test follows the ASTM standard E399. but the size of the sample tested is reduced to ,2.5 mm X 12 mm X mm. This sample 3 size has been tested against the larger standard samples and the resuhs are comparable. A schematic of the test specimen geometry is shown below. Figure B. 1 Compact tension sample geometric definitions. The samples are cut from the cast plaque of material and holes drilled using an aluminum guide. The crack is inserted using an above Tg crack insertion technique'. The drilled samples are placed in the fixture shown below in Figure B.2 equipped with a new 239 single-edged razor blade. The apparatus is heated to 50 - 90°C above the Tg of the material, and the razor blade slid across the top of the samples to create a crack that is approximately 3mm deep. The samples were then replaced in the oven and allowed to cool slowly to room temperature to minimize the effects of residual stresses in the samples. This technique provides sharp and consistent precracks. The samples are then placed in cleavis brackets mounted in the Instron tensile testing machine with an SM-100 100 pound load cell. The sample is pulled at a rate of 240 0.05cn.n.n. and the critical load at failure measured. Fro. tMs load the critical stress intensity factor, K,, is determined using the equations below. ^ ^PcF{alW) where Kic - critical stress intensity factor [N/m^^^] Pc = crack propagation load [N] F(aAV) = function of geometric parameters as shown above [dimensionless] a - crack length as shown in Figure B. 1 [m] W = sample dimension as shown in Figure B. 1 [m] B = sample thickness [m] The fracture energy can then be calculated from the flexural modulus and the critical stress intensity factor as shown in the equation below. E where, Gic = fracture energy [J/m^] Kic = critical stress intensity factor [N/m^^^] E = flexural modulus [Pa] 241 References 1 S. Thompson, RJ. Farris, "A High . Quality A Pre-crack Insertion Technique for Resistance Determination m Thermoset Materials", SAMPEJ, 24(1), 47, (1988) 242 BIBLIOGRAPHY ' ^' M^Grail, ' P T '^sTftl'^M sX'rp'' r ' and ^'"^^^ ^P^^ ^atnces", 34th 2^2^^^^^^^^^^^ Int. UpE Sefton, M.S., Thermoset/Thermoplastic Alloys- An Annroach to Ton ah w' u Performance Resm Matrices", 20th Int. r^/Tco^^^^^^^^ '''' ' S^ft^"' S., ^ ' M McGrail P T Wilkinson S P Soc. Adv. Mater. ' Proc. Eng. 33rd ' ' Internat Symp., 979-989, (1988) ^" ' Rose, ' ' J.B. and Staniland SynthesisWh ' ^r'""'-^-^ P A and Properties of Polyaryletherketones", Polymer, 11, 1097-1 107(1981) Bascom, W.D. and Ting, R. Y., "The Fracture of an Epoxy Polymer Containing Elastomenc Modifiers", J. Mat. Sci., 16, 2657-2664, (1981) "^PP'^^^^i^" «f Epo^y Resins in Advanced Composites", 34th SAMPE Sympos.,m9-\900,(m9) Int Bauer, Ronald S., "Toughened High Performance Epoxy Resins: Modification with Ihermoplastics. I ", 34th Int. SAMPE Sympos., 899-909, (1989) Bellenger, v., Verdu, J., and Morel, E., "Structure-Properties Relationships for Densely Cross-Linked Epoxide-Amine Systems Based on Epoxide or Amine Mixtures- Part 2 Water Absorption and Diffusion", J. Mat. Sci. , 24, 63 -68, ( 1 989) Bennett, G.S., Farris, R.J. and Thompson, S.A., "Amine-Terminated Poly(Aryl Ether Ketone)-Epoxy/Amine Resin Systems as Tough High Performance Materials" Polymer ' 32(9), 1633-1641, (1991) ' ^ Bennett, Greggory S., Ph.D. Thesis, University of Massachusetts, (1991) Bergenn, Waher R. and Rigby, R.B., "PES and PEEK: Tough Engineering Thermoplastics", Chem. Eng. Pr., 81, 36-38, (1985) Biddlestone, F., Goodwin, A.A., Hay, J.N. and Mouledous, G.A.C., "The Relaxation Spectrum and Physical Ageing of Polyetherimide", Polymer, 32(17), 31 19-3125, (1991) 243 Bucknall, Clive B., and Partridge, Ivana K., "Phase Separation in Epoxy Resins Containing Polyethersulphone",Po/v7wer, 24, 639-644, (1983) Cattanach, J.a, Guff, G. and Cogswell, F.N., "The Processing of Thermoplastics "^""^ ^^^"'^^^^"^ ''''''' ^"^6^^^^^^^^^^^ ' ^f'^^y-- Cecere, J A Hedrick, J.L. and McGrath, J.E., "Morphology and Properties of Amine Terminated Poly(Ai7lene Ether Ketone) and Poly(Arylene Ether Sulfone) Modified Epoxy Resm Systems", 3Jst Int. SAMPE Sympos., 580-588, (1986) Chan, L.C., Gillham, J.K., Kinloch, A.J., Shaw, S.J., "Rubber Modified Epoxies" Morphology, Transitions, and Mechanical Properties", Rubber Modified EnnviP. Am ' Chem. Soc. Ser. ,261,(1984) ' Cho, J.B., Hwang, J.W., Cho, K., An, J.H. and Park, C.E., "Effects of Morphology on Toughemng of Tetrafimctional Epoxy Resins with Poly(Ether Imide)" Polvmer 34^23^h 4832-4836, (1993) y Cogswell, F.N. and Leach, DC, "Continuous Fibre Reinforced Thermoplastic: A Change in the Rules for Composite Technology", Plastics and Rubber Proc andAppl 4(3) 271-276,(1984) Demonet, Charles M., "Interaction of Moisture with Resin Matrix Composites" 21st Tech. SAMPE Con/., 23 1 -23 8, ( 1 989) Diamant, Y., Marom, G. and Broutman, L.J., "The Effect of Network Structure on Moisture Absorption of Epoxy Resins", J. Appl. Polym. Sci., 26, 3015-3025, (1981) Dusek, K., Lednicky, F., Lunak, S., Mach, M. and Duskova, D., "The Toughemng of Epoxy Resins with Reactive Polybutadienes", Rubber Modified Thermoset Resins . Advances in Chemistry Series 208, (1984) Frischinger, Isabelle and Dirlikov, Stoil, "Toughening of Epoxy Resins by Epoxidized Vegetable Oils", Polym. Comm., 32(17), 536-537, (1991) Gannon, John, "Epoxy", Modern Plastics, (Mid-Oct), 129-131, (1991) 244 ^^"f Density of Amine Cross-Linked "^nlltiNetworks J. Epoxy , TZ,Appl. Polym.P f Sci., 41, 2533-2535, (1990) ' ' "C^^-^^l ' Modification of Matrix "Rlt'N.^'Re.n Networks ' with Engineering Thermoplastics", ACS Polym. Preprints, 26(2) Ss, Hedrick James L Yilgor, I., Wilkes, Garth L., McGrath, James E., "Chemical "^tlt^St^' " -ermoplastics" 1,. Hourston, D.I and Lane, J.M., "The Toughening of Epoxy Resins with Thermoplastics- Intunctional 1 Epoxy Resm - Polyetherimide Blends", Polymer, 33(7), 1379-1383, (1992) Hourston, Douglas J., Lane, Jaqueline M. and MacBeath, Nigel A "Toughening of Epoxy Resms with Thermoplastics. IL Tetrafunctional Epoxy Resin - Polyetherimide Blends", Polym. Int., 26, 17-21, (1991) Jones, D.P., Leach, D.C. and Moore, D R., "Mechanical Properties of Poly(Ether-Ether- Ketone) for Engineering Applications", Polymer, 26, 1385-1393, (1985) Keenan, Joseph D., Seferis,James C. and Quinlivan, John T, "Effects of Moisture and Stoichiometry on the Dynamic Mechanical Properties of a High-Performance Structural Epoxy", J. Appl Polym. Sci., 24, 2375-2387, (1979) Kim, Sung C. and Brown, Hugh R., "Impact Modified Epoxy Resin with Glassy Second Component", J. Mats. Sci., 22, 2589-2594, (1987) Kimoto, Masaki, "Microstructure and Mechanical Properties of CFRP Consisting of Epoxy-Polyetherimide Blend Matrix", Sen-I Gakkaishi, 49(2), 72-78, (1993) Kinloch, A.J., Shaw, S.J., Tod, D.A. and Hunston, D.L., "Deformation and Fracture Behavior of a Rubber-Toughened Epoxy: 1. Microstructure and Fracture Studies", Polymer, 24, 1341-1354, (1983) 245 Ma Chen-Chi M Hu Jiann-Tsun. Liu, Wen-Liang. Hsia. Hung-Chung, Shieh Biing - Yng and L,u, Rong-Sh.an, "Tl,e Thermal, Rheological and Morphological P opmles of PobpherJyene Sulfide and Polyetherether Ketone Resins and clposL\ i7S SAMPESympos., 420-433, (me) Martuscelli, E., Musto, P., Ragosta, G. and Scarinzi, G., "A New Supertough Epoxy Formulation with High Crosslinking Density", Die Angewandte Makromolekulare Chemie, 204, 153-159, (1993) Misra S.C., Manson, J. A. and Sperling, L.H., "Eflfect of Cross-Link Density Distribution on the Engineering Behavior of Epoxies", in: Epoxv Resin rhe.mi.try ed Ronald S Bauer, ACS Symposium Series 114, (1979) Morel, E., Bellenger, V., Bocquet, M. and Verdu, J., "Structure-Properties Relationships for Densely Cross-Linked Epoxide-Amine Systems Based on Epoxide or Amine Mixtures; Part 3. Elastic Properties", J. Mat. ScL, 24, 69-75, (1989) Morgan, Roger J., O'Neal, James E. and Fanter, Dale L., "The Effect of Moisture on the Physical and Mechanical Integrity of Epoxies", J. Mat. Sci., 15, 751-764, (1980) Murakami, A., Saunders, D., Ooishi, K., Yoshiki, T., Saitoo, M., Watanabe, O., Takezawa, M., "Fracture Behavior of Thermoplastic Modified Epoxy Resins"', J. Adhesion, 39, 227-242, (1992) Ochi, M. and Bell, J.P., "Rubber-Modified Epoxy Resins Containing High Functionality Acrylic Elastomers", J. Appl. Polym. Palmese, G.R. and McCuUough, R.L., "Effect of Epoxy-Amine Stoichiometry on Cured Resin Material Properties", J. Appl. Polym. Sci., 46, 1863-1873, (1992) 246 Peake, Steven, Maranci, Art Meena Diana "Hir.v, t . ^ . Composites", SAMpI^^.^^^^^^^ Pearson, R.A., Yee, A.F., Pofym. Mater. Sci. Eng., 63, 311, (1990) «f I-P-t ^ToZ^^^^^ f Modifiers Used in Recker, H.G. "Toughened Thermosets for Damage Tolerant Carbon Fiber ^eimorcedReinforced Composites", 2Jst Int. SAMPE Tech. Conf., 283-293, (1989) ^^^^ ' ^ ' ^ ' I«-an, P., ""Ttsch H\tl>e?T \ ^ > (19895 ' 283^293, Riew, C. Keith and Gillham, John K. (Eds), Rubber Modified Then.n..t i^.e^.o " ' Advances m Chemistry Series 208, (1984) Rigby, Rhymer B., "Polyetheretherketone Peek", Polym. News, 9, 325-328, (1984) Sanjana Z.N. and Testa, J.H., "Toughened Epoxy Resins for Filament Winding" 30th Nat. SAMPE Sympos. ,1221-1230,(1985) Sefton, M.S., McGrail, P.T., Peacock, J.A., Wilkinson, S.P., Crick, R.A., Davies M and Almen, G., "Semi-Interpenetrating Polymer Networks as a Route to Toughening of Epoxy Resin Matrix Composites", 19th Int. SAMPE Tech. Conf., 700-710, (1987) Shinkichi Murakami, Osamu Watanabe, Hiroshi Inoue, Toa Nenryo Kogyo K.K., Mitsukazu Ochi, Takuya Shiraishi, Masaki Shimbo, "Fracture Behavior of Epoxy Resins with Dififerent Crosslinking Structures", 34th Int. SAMPE Sympos. , 2194-2205, (1989) Sjoerdsma, S.D., "The Tough-Brittle Transition in Rubber Modified Polymers" Polym Comm. 1 06- , 30, 108,(1 989) Sultan, J.N. and McGarry, F.J., "Effect of Rubber Particle Size on Deformation Mechanisms in Glassy Epoxy", Polm. Eng. and Sci., 13(1), 29-34, (1973) Thompson, S.A., Ph.D. Thesis, University of Massachusetts, (1987) Thompson, S.A, Farris, R.J., "A High Quality Pre-crack Insertion Technique for Fracture Resistance Determination in Thermoset Materials", SAMPE J, 24(1), 47, (1988) 247 Aco^lonitnle Random Copolymers", Polymer, 30, 107-115™ 989) UimoseLKesins, Advances in Chemistry Series 208,(1 984) Wedgewood, A.R Su K B. and Narin, J.A., "Toughness Properties and Service Temperature Thermoplastics and Their Composites" J9th Int II!ZTtVI^'^^ ' SAMPE Tech. Conf., 454-463, (1987) Wingard, Charles D. and Beatty, Charies L., "Crosslinking of an Epoxy with a Mixed Amine as a Function of Stoichiometry. IT Final Properties Via Dynamic Mechanical Spectroscopy J. Appl. , Polym. Set, 41, 2539-2554, (1990) S.D., Wu Hedrick, J.L., Mohanty, D.K., Carter, B.K., Wilkes, G.L. and McGrath J E Synthesis and Characterization of Poly(Arylene Ether Ketone Sulfone) Copolvmers"' 3Jst Int. SAMPE Sympos., 933-944, (19S6) 248