high temperature hybrid elastomers - drexel university · high temperature hybrid elastomers a...
TRANSCRIPT
High Temperature Hybrid Elastomers
A Thesis
Submitted to the Faculty
of
Drexel University
by
Kerry Anthony Drake
in partial fulfillment of the
requirements for the degree
of
Doctor of Philosophy
January 2011
© Copyright 2011
Kerry Drake. All Rights Reserved
iii
Dedications
Dedicated to my wife Suzie whose love, support and encouragement helped me
throughout the entire journey, my children Ben and Logan, my parents Anthony and
Sophia Drake, my uncle Richard Domalavage, and to the memory of my late
grandmother Isabel Domalavage.
iv
Acknowledgements
The completion of my doctoral studies could not have taken place without the
support of many people.
It is difficult to express the depth of gratitude I have for my advisor, Dr. Yen Wei,
whose breadth of knowledge and enthusiasm in science and chemistry is unparalleled.
From my first polymer chemistry classes with Dr. Wei up through the completion of this
body of work, Dr. Wei’s tutelage has been invaluable. Dr. Wei has been a role model for
scientific accomplishment in the face of adversity, and a much-appreciated mentor for
both technical management and professional growth.
I would also like to give a special acknowledgement to Dr. Anthony Addison,
who has helped me in innumerable ways over my long association with the Drexel
University Department of Chemistry. Dr. Addison’s encouragement to enter the part-time
Ph.D. program and advice on many topics, personal and professional, has helped me
greatly throughout the past several years.
The Management team of Greene Tweed deserves special merit for encouraging
my personal development through the pursuit of a doctoral degree, along with the
financial support to complete my studies: Michael Delfiner, John Jorgensen, Henry
Steuber, Dr. Brock Alexander, George Rawa, Mike Brewer, Glenn Doell, and Michael
Glessner have all been supportive of this endeavor. Dr. Brock Alexander in particular
has been extremely supportive and has been an excellent mentor for my growth as an
industrial scientist.
v
I am deeply indebted to Dr. Shux Li, who taught me many invaluable lessons in
synthetic chemistry and polymer analysis, and also for the many stimulating discussions
on potential synthetic paths to accomplish the particular objectives of my research
project.
Dr. Neil Mukherjee also deserves special recognition for his many contributions
to my research project, including assistance with the development and fine-tuning of
synthetic procedures found in the literature, assistance with data collection (especially
some of the overnight NMR experiments) and enthusiasm for the generation of new
knowledge.
I would also like to thank the Drexel Chemistry Department and its support staff
for the facilities, equipment, chemicals, and for running and maintaining an exceptional
chemistry department, and all the fellow Drexel Graduate students who I’ve had the
privilege of working with these past few years: Dr. Khalid Mirza, for his expert
knowledge of organic synthesis and spectroscopy, Dr. David Berke–Schliessel and
Sudipto Das for their assistance during my studies, and Dr. Tom Measy and Jon Soffer
from Dr. Stenner’s group for assistance with Raman analysis.
In conclusion, I would also like to acknowledge Dr. Sally Solomon, Dr. Jean-
Claude Bradley, Dr. Anthony Addison, Dr. Joe Foley, and Dr. Frank Ji for serving on
both my Oral Defense Committee and my Ph.D. Committee, Dr. Giuseppe Palmese, who
served on my Oral Defense Committee, and Dr. Susan Jansen-Varnum of Temple
University who is serving on my Ph.D. Committee.
vi
Dr. Frank Ji warrants special recognition for chairing my Ph.D. Committee and
his assistance on preparation of papers for publication and this thesis while Dr. Wei was
on sabbatical.
vii
Table of Contents
List of Tables .....................................................................................................................xv
List of Figures ................................................................................................................. xvii
List of Schemes .............................................................................................................. xxiv
List of Symbols .............................................................................................................. xxvi
List of Abbreviations .......................................................................................................xxx
Abstract ......................................................................................................................... xxxii
Chapter 1: Overview of the structure-property relationships of elastomeric
materials..............................................................................................................................1
1.1. Organization of this dissertation ...................................................................................2
1.2. Historical background of rubber and elastomer technology .........................................2
1.3. Structure property relationships of elastomers .............................................................4
1.3.1. Flexible linkages ..................................................................................................4
1.3.2. Viscoelasticity ......................................................................................................9
1.3.2.1. Mechanical properties definitions ...............................................................9
1.3.2.2. Temperature dependence of properties of elastomers ..............................18
1.3.2.3. Free volume concept, and the dependence of free volume on
temperature ...........................................................................................................20
1.3.2.4. Time dependence of polymer properties (Deborah Number) ...................22
1.3.2.5. Polymer self-diffusion, and its effect on elastomeric properties ..............23
1.3.2.6. Effects of physical entanglements on mechanical properties of
polymers .................................................................................................................24
viii
1.4. Cross-linking of polymers for elastomeric properties ................................................30
1.5. Crystallinity of a polymer, and the effect on elasticity ...............................................31
1.6. Thermal stabilities of polymers ..................................................................................33
1.7. Thermal stabilities of cross-link sites .........................................................................38
1.7.1. Diethynyl cross-linking ......................................................................................39
1.7.2. Ethynyl cross-linking mechanisms. ...................................................................39
1.8 Survey of commercially available high temperature elastomers .................................41
1.8.1 Fluoropolymer based elastomers ........................................................................42
1.8.1.1. Perfluoroelastomers (FFKM) ....................................................................44
1.8.1.2. Fluoroelastomers (FKM) ..........................................................................45
1.9. Olefinic elastomers .....................................................................................................45
1.10. Silicone elastomers ...................................................................................................46
1.10.1. Phenoxysilanes .................................................................................................48
1.11. Analytical tools used for characterization of polymers in this work ........................49
1.12. References .................................................................................................................58
Chapter 2: Synthesis of polymers for subsequent endcapping reactions ...................66
2.1. Introduction .................................................................................................................66
2.2. Experimental section ...................................................................................................73
2.2.1. Reagents .............................................................................................................73
2.2.2. Synthesis ............................................................................................................73
ix
2.2.2.1. Diphenylsilane-hydroquinone copolymer melt polymerization (Scheme
2.6) .........................................................................................................................73
2.2.2.2. Diphenylsilane-hydroquinone solution polymerization (Scheme 2.7) .....75
2.2.2.3. Diphenylsilane-biphenol copolymer melt polymerization (Scheme 2.8) .76
2.2.2.4.Diphenylsilane-hydroquinone copolymer polymerization in CH2Cl2
(Scheme 2.9) ..........................................................................................................78
2.2.2.5. Diphenyldichlorosilane-biphenol copolymer polymerization in CH2Cl2
(Scheme 2.10) ........................................................................................................79
2.2.2.6.Diphenylsilane-biphenol copolymer polymerization in toluene
(Scheme 2.11) ........................................................................................................81
2.2.2.7. Diphenylsilane-hydroquinone copolymer polymerization in mixed
THF/toluene system (Scheme 2.12). .....................................................................83
2.2.2.8. Dichlorodiphenylsilane-biphenol polymerization in THF
(Scheme 2.13). .......................................................................................................86
2.3. Characterization ..........................................................................................................87
2.4. Results and Discussion. ..............................................................................................88
2.4.1. Solubility in methanol, acetone, and its relationship with molecular weight
of aryloxysilanes. .........................................................................................................88
x
2.4.2.Diphenylsilane-hydroquinone polymer (melt polymerization) ..........................90
2.4.2.1. Spectroscopy .............................................................................................90
2.4.2.2. Molecular weight ......................................................................................93
2.4.2.3. Thermal analysis .......................................................................................94
2.4.3. Diphenylsilane-biphenol polymer (melt polymerization) ................................103
2.4.3.1. Molecular weight ....................................................................................103
2.4.3.2. Thermal analysis .....................................................................................105
2.5. Conclusions for hydrosilane condensation reactions ................................................106
2.6. Dichlorosilane diol condensations ............................................................................106
2.6.1. Dichlorosilane- hydroquinone condensations ..................................................106
2.6.1.1. Molecular weights ...................................................................................106
2.6.1.2. Thermal analysis of dichlorodiphenylsilane-hydroquinone polymers. ...109
2.6.2. Dichlorosilane- biphenol condensations ..........................................................111
2.6.2.1. Molecular weights ..................................................................................111
2.6.2.1.1. Estimated molecular weights vs. theoretical molecular weights
from the Carothers Equation .........................................................................112
2.6.2.1.2. Molecular weight summary. ..........................................................113
xi
2.6.2.2. Thermal analysis of dichlorodiphenylsilane-biphenol polymers. ...........114
2.6.2.3. Spectroscopic analysis ............................................................................117
2.7. Conclusions ...............................................................................................................121
2.8. References .................................................................................................................125
Chapter 3: Phenylethynyl and phenol end capping studies of
polybiphenyloxydiphenylsilanes for cross-linking and enhanced thermal
stability. ..........................................................................................................................128
3.1. Introduction ...............................................................................................................128
3.2. Experimental Section ................................................................................................133
3.2.1. Materials ..........................................................................................................133
3.3. Polymer synthesis .....................................................................................................131
3.4. General procedure for endcapping ............................................................................134
3.4.1. Endcapping with lithiumphenylacetylide .......................................................135
3.4.2. Endcapping with (4[(4-fluorophenylethynyl)]phenol) ....................................135
3.4.2. Endcapping with phenol. .................................................................................136
3.5. Polymer purification .................................................................................................137
3.6. Characterization ........................................................................................................138
3.7. Results and discussion ..............................................................................................139
xii
3.7.1. Molecular weight summary .............................................................................139
3.8. Uncapped polymers ..................................................................................................141
3.9. Characterization of lithiumphenylacetylide endcapped polymer .............................148
3.9.1. Molecular weight .............................................................................................148
3.9.2. Spectroscopic characterization .........................................................................148
3.9.3. Thermal analysis ..............................................................................................152
3.10. Characterization of (4[(4-fluorophenylethynyl)]phenol) endcapped polymer. ......154
3.10.1. Molecular weight ...........................................................................................154
3.10.2. Spectroscopic characterization .......................................................................155
3.10.3. Thermal analysis ............................................................................................159
3.11. Comparison of physical properties of endcapped aryloxysilanes with
dimethylsiloxane and FFKM. ..........................................................................................161
3.12. Conclusions. ............................................................................................................162
3.13. References. ..............................................................................................................163
Chapter 4: Novel Diacetylphenoxysilane polymers ...................................................167
4.1. Introduction ...............................................................................................................167
4.2. Experimental Section ................................................................................................168
4.2.1. Materials ..........................................................................................................168
4.2.2. Polymer Synthesis. ...........................................................................................169
4.3. Characterization ........................................................................................................170
4.4. Results and Discussion .............................................................................................171
xiii
4.4.1. Molecular weight .............................................................................................171
4.4.2. Spectroscopic characterization .........................................................................172
4.4.3. Thermal analysis ..............................................................................................179
4.5. Elastomeric properties of BDDP:DCDPS polymer ..................................................195
4.6. Summary ...................................................................................................................196
4.7. References .................................................................................................................197
Chapter 5: General conclusions and suggested future research ...............................200
5.1. Summary ...................................................................................................................200
5.2. Future studies. ..........................................................................................................201
5. 2.1. Endcapped materials .......................................................................................201
5. 2.2. Diacetylphenoxysilane polymer......................................................................201
5.2.2.1. Elastomeric material ...............................................................................202
5.2.2.2. Thermoset polymer .................................................................................203
5.2.2.3. Preceramic polymer ................................................................................203
5.2.2.4. Solvent cast coating ................................................................................203
5. 3. Conclusions ..............................................................................................................203
Appendix A: PhD Candidacy Research Proposal Electrochemically Assisted
Sol Gel Deposition of Lanthanum and Cerium Oxides for Enhanced
Corrosion Resistance of Solid Oxide Fuel Cell Components .....................................205
Vita ..................................................................................................................................239
xiv
List of Tables
Table 1.1. Comparison of Tgs of selected linear polymers with flexible backbone
linkages, and predominantly methylene secondary linkages ...............................................6
Table 1.2. Comparative table, tensile properties of selected material types ......................15
Table 1.3. Typical hardness values for elastomers when tested
at ambient conditions ........................................................................................................18
Table 1.4. Comparison of generic classes of polymeric materials, and their respective
fulfillments of property requirements to allow classification as elastomers. .....................29
Table 1.5. Bond dissociation energies of general classes of polymeric bonds ..................36
Table 1.6. Summary of the most common high temperature cross-linkable
end-groups..........................................................................................................................38
Table 1.7. Maximum operating temperatures of selected elastomers, adapted from
Threadingham et al. ...........................................................................................................42
Table 1.8. Service temperatures and fluorine contents of various fluorinated
polymers ............................................................................................................................44
Table 2.1. Calculated weighted average molecular weight (Mw), number average
molecular weight (Mn,) and polydispersity Index (Mw/Mn),PDI ....................................89
Table 2.2. Molecular weight summary of several diphenylsilane-diol
melt condensation polymers. ............................................................................................94
Table 2.3. Calculated molecular weight changes for diphenylsilane
polymer after heating to 300 °C for 30 minutes. ...............................................................99
xv
Table 2.4. Summary of molecular weights, dichlorosilane-hydroquinone polymers
prepared with different reaction solvents and different monomer ratios .........................108
Table 2.5. Summary of molecular weights, dichlorosilane-biphenol polymers prepared
with different reaction solvents and different monomer ratios ........................................112
Table 2.6. Summary of Tg, TGA 5% weight loss and residue content in nitrogen
and air..............................................................................................................................123
Table 3.1. Polymer molecular weight summary (relative to polystyrene standards).......141
Table 4.1. Summary of thermal properties of uncured, partially cured, and fully cured
diethynyl polymer specimens. .........................................................................................193
xvi
List of Figures
Figure 1.1. Common flexible chemical linkages for elastomers. Clockwise from upper
left: methylene, ether, thioether, phophazene, siloxane, fluoromethylene ..........................5
Figure 1.2. Tensile test specimen (left), tensile test in progress(right), showing the change
in gauge length/strain from 2 inches to >2 inches as test progresses (increase in distance
between horizontal marks in middle to right figures) ........................................................11
Figure 1.3. Representative stress/strain curve, showing modulus (OP tangent line),
Proportional limit (P), Elastic Limit (E), Yield (Y), ultimate strength(U),
and strain at break (R) .....................................................................................................12
Figure 1.4. Stress-strain curves for perfectly elastic material (a) and
perfectly viscous material(b). .............................................................................................13
Figure 1.5. Stress/strain curves of representative material types .......................................15
Figure 1.6. Tensile Stress/strain curves for styrene- butadiene block
thermoplastic elastomer ....................................................................................................19
Figure 1.7. Representation of reptation of a polymer chain P, moving
through fixed obstacles , O ...............................................................................................20
Figure 1.8. Specific volume of a typical amorphous polymer
vs. temperature. ..................................................................................................................21
Figure 1.9. Rectangular Torsion DMA test of PEEK, showing shear modulus
as a function of temperature. ..............................................................................................26
Figure 1.10. Deformation under load vs. temperature for a typical cross-linked
elastomer. ...........................................................................................................................27
xvii
Figure 1.11. Trends in shear modulus relative to cross-link density (mass between
cross-links). ........................................................................................................................31
Figure 1.12. Trend in shear modulus relative to % crystallinity (physical cross-links). ...32
Figure 1.13. Schematic of bond energy, bond length, and energy required to cleave a
bond....................................................................................................................................34
Figure 1.14. Initial cure mechanism of phenylethynyl cross-link sites ............................41
Figure 1.15. Possible cure products of thermally cured phenylethynyl polyimides ..........42
Figure 1.16. FKM, FFKM chemical structures of starting monomers. ............................44
Figure 1.17. Monomer and general structures for the most common olefinic rubbers. ....47
Figure 1.18. General siloxane structure, with most common side group chemical
structures. ...........................................................................................................................48
Figure 1.19. Carborane structure from Peters et al ............................................................49
Figure 1.20. Schematic of a GPC column..........................................................................51
Figure 1.21. Representative DSC scan ..............................................................................51
Figure 1.22. Schematic of thermogravimetric instrument .................................................54
Figure 1.23. Schematic of parallel plate rheometer ..........................................................55
Figure 1.24. Representative parallel plate data, showing the change in storage and loss
modulus as a function of time, and the gel point (cross-over point)..................................56
Figure 1.25. Schematic of NMR instrument ......................................................................57
Figure 2.1 Pendant and main chain ethynyl groups. ..........................................................73
Figure 2.2.Overlay of molecular weight distributions of insoluble polymer fractions. .....90
Figure 2.3. FTIR Spectrum of Polymer 2.1.3 ....................................................................92
xviii
Figure 2.4. FTIR of diphenylsilane ....................................................................................93
Figure 2.5. FTIR of hydroquinone .....................................................................................93
Figure 2.6. TGA thermogram of polymer 2.6-2 (hydroquinone/diphenylsilane) in air,
10 ºC per minute heating rate. ............................................................................................96
Figure 2.7. Parallel plate rheogram, of sample 2.6-2. Plot of storage modulus, G’, loss
modulus G’’, and tan δ. ......................................................................................................97
Figure 2.8. Parallel plate rheogram of sample 2.6-.3. Plot of complex viscosity η*
during isothermal test conducted at 300 °C. ......................................................................98
Figure 2.9. Overlay of DSC scans of sample 2.6-2, as synthesized vs. after rheology
test (30 minutes, 300 °C).H= half-height for Tg calculation. (20 °C/min heating rate,
nitrogen atmosphere). ......................................................................................................101
Figure 2.10.TGA plot of sample 2.6-2 Y1 Axis=Weight %, Y2 Axis = temperature. ....102
Figure 2.11. Overlay of TGA thermograms of sample 2.6-2, showing enhancement in
thermal stability of isothermally heated sample. ............................................................103
Figure 2.12. GPC chromatogram of biphenol-diphenylhydrosilane. ...............................105
Figure 2.13. DSC thermogram of biphenol-diphenylsilane melt polymerized sample
(2nd
heat) at 20°C/min in nitrogen atmosphere. ..............................................................106
Figure 2.14. TGA of polymer 2.12B (DCDPS, hydroquinone) in air ( 10°C/min) .........110
Figure 2.15. DSC of polymer 2.12B (DCDPS, hydroquinone)
xix
in nitrogen ( 20 °C/min) ..................................................................................................112
Figure 2.16. TGA in nitrogen of 2.11-2 Dichlorodiphenyl silane-biphenol
(10 °C/min). .....................................................................................................................116
Figure 2.17. TGA in air of 2.11-2 (Dichlorodiphenyl silane-biphenol)
in air (10°C/min heating rate). .........................................................................................117
Figure 2.18. DSC scan of 2.11-2 ( dichlorodiphenylsilane- biphenol),
20 °C/min, nitrogen atmosphere .....................................................................................118
Figure 2.19. FTIR of polymer 2.11, with FTIR spectra of the starting reagents
overlaid for comparison ...................................................................................................119
Figure 2.20. 13
C NMR of a representative DCDPS biphenol polymer,
20,000 Mn(CDCl3). ..........................................................................................................120
Figure 2.21. 13
C NMR of dichlorodiphenylsilane (starting reagent) ...............................121
Figure 2.22. Predicted 13
C NMR of biphenol ( starting reagent) .....................................122
Figure 3.1. Polydiphenylsiloxane ...................................................................................129
Figure 3.2. Repeat unit of polyaryloxydiphenylsilane. ...................................................130
Figure 3.3. Aryloxydiphenylsilane polymer with chlorosilane endgroups available for
endcapping reactions (general structure of samples 1a, 1b, 2b, 3b, 4b). .........................132
Figure 3.4. First heat DSC comparison of samples with Mn of 7,000 (1a)
and 22,000 (1b) and phenol endcapped polymer, Mn=22,000 (4a). ................................143
xx
Figure 3.5. First heat DSC comparison of samples with Mn of 7,000 (1a)
and 22,000 (1b) and phenol endcapped polymer, Mn=22,000 (4a),
enlarged to show the exothermic peak at 275 °C. ............................................................144
Figure 3.6. 2nd
Heats of unendcapped polymers, with phenol capped
polymer for comparison. ..................................................................................................145
Figure 3.7. Rheogram of complex viscosity vs. cure time at 300 °C for samples 1a, 1b,
and 4a ...............................................................................................................................147
Figure 3.8. Overlay of capped and lithiumphenylacetylide capped polymer
(full scale). ......................................................................................................................150
Figure 3.9. IR spectra of lithiumphenylacetylide capped and uncapped polymer,
showing the characteristic ethynyl peak at 2,159cm-1
. ...................................................151
Figure 3.10. 13
CNMR spectrum of lithiumphenylacetylide capped and uncapped
polymer ............................................................................................................................152
Figure 3.11. 13
C NMR spectra of lithiumphenylacetylide capped and uncapped
polymer, showing ethynyl peaks at 89 and 109 ppm in the capped sample ...................153
Figure 3.12. DSC comparison of first and second heats of 2a, lithiumphenylacetylide
capped polymer, showing the expected ethynyl cure peak at 324 °C. .............................154
Figure 3.13. Comparison of viscosity change on heating of lithiumphenylacetylide
capped vs. uncapped polymer ..........................................................................................155
Figure 3.14. Raman spectrum of fluorophenylphenol capped polymer, with spectra of
fluorophenylphenol reagent and uncapped polymer overlaid for comparison
xxi
(full scale) .......................................................................................................................157
Figure 3.15. Raman spectrum of fluorophenylphenol capped polymer, with spectra of
4,4’fluorophenylphenol reagent and uncapped polymer overlaid for comparison
(zoom). .............................................................................................................................158
Figure 3.16. 13
C NMR spectrum of fluorophenylethynylphenol capped polymer
with an uncapped polymer spectrum overlaid for comparison. .......................................159
Figure 3.17. 13
C NMR spectrum of fluorophenylethynylphenol capped polymer with
an uncapped polymer spectrum overlaid for comparison, enlarged to show ethynyl
carbons in the capped polymer. .......................................................................................160
Figure 3.18. DSC thermogram of polymer 3c 4[(4-fluorophenylethynyl)] capped
polymer, enlarged to show the exothermic peak at 355 °C .............................................161
Figure 3.19. Rheogram comparing the changes in complex viscosity on heating of
4[(4-fluorophenylethynyl)] capped polymer, 3a, relative to a comparable
molecular weight uncapped polymer 1a .........................................................................162
Figure 3.20. Shear modulus comparison of polydimetyhylsiloxane, FFKM, and capped
aryloxysilanes when tested above their respective Tgs ....................................................162
.
Figure 4.1. GPC chromatogram of BDDP:DCDPS polymer. Mw=4,600; Mn=2,000;
Polydispersity= 2.3 ..........................................................................................................172
Figure 4.2. FTIR of BDDP, showing the weak ethynyl stretch absorption at
2150cm-1
..........................................................................................................................173
Figure 4.3. Raman spectra of BDDP and BDDP:DCDPS polymer, with THF blank
overlaid for comparison. ..................................................................................................174
Figure 4.4. 13
C NMR of BDDP monomer, showing the acetylinic carbon signals
xxii
at 81.8 and 72.4ppm ........................................................................................................175
Figure 4.5. FTIR of solid polymer, showing the Si-O-C6H5 stretch
at 904 cm-1
and the acetylinic stretches at 2212 and 2150cm-1
. .......................................176
Figure 4.6. 13
C NMR of DCDPS BDDP polymer, in deuterated DMSO. .......................177
Figure 4.7. Comparison of 13
C NMR of the BDDP reagent and
the polymer, showing the additional aromatic bands in the polymer sample. ................178
Figure 4.8. MDSC thermogram of BDDP monomer, heat flow signal shown. ...............179
Figure 4.9. DSC scans of DCDPS:BDDP polymer at different heating rates, showing the
exothermic peak shift due to changes in heating rates .....................................................181
Figure 4.10. Standard DSC of BDDP:DCDPS polymer, with identified
transitions. .......................................................................................................................183
Figure 4.11. MDSC thermogram showing separate reversing and nonreversing
heat flow curves. ..............................................................................................................184
Figure 4.12. MDSC plot of 1st heat, reversing and non-reversing heat capacities vs.
temperature for BDDP:DCDPS polymer .........................................................................185
Figure 4.13. MDSC plot of 2nd heat, reversing and non-reversing
heat capacities vs. temperature for BDDP:DCDPS polymer. .........................................186
Figure 4.14. TGA of DCDPS-BDDP polymer in nitrogen. .............................................187
Figure 4.15. TGA of DCDPS-BDDP Polymer in air .......................................................188
Figure 4.16. Temperature sweep of DCDPS BDDP polymer, in a nitrogen atmosphere.
Crossover point was 242°C. .............................................................................................189
Figure 4.17. Plot of G’ and its derivative as a function of temperature. .........................191
Figure 4.18. Isothermal parallel plate cure experiment, DCDPS:BDDP polymer. .........192
xxiii
Figure 4.19. DSC curves of partially cured DCDPS:BDDP polymer. ..........................193
xxiv
List of Schemes
Scheme 1.1. Natural rubber monomer and polymer (cis1,4 polyisoprene). ........................2
Scheme 1.2. Backbone cleavage ( top) vs. side group cleavage (bottom) .........................35
Scheme 2.1. Uncatalyzed nucleophilic substitution at silicon atom. .................................68
Scheme 2.2. Mechanism of catalyzed nucleophilic substitution of hydridosilane ............68
Scheme 2.3. General reaction scheme, aminosilane – diol condensation
polymerization, in the melt state (no solvents used). ........................................................70
Scheme 2.4. General reaction scheme, catalyzed condensation of dihydridiosilane
with diols ............................................................................................................................71
Scheme 2.5. General reaction scheme, catalyzed condensation of dihydridiosilane
with diols. ..........................................................................................................................72
Scheme 2.6. diphenylsilane condensation melt polymerization, with hydroquinone
comonomer ........................................................................................................................74
Scheme 2.7. Diphenylsilane solution based condensation polymerization with
hydroquinone as the comonomer. ......................................................................................76
Scheme 2.8. Diphenylsilane condensation melt polymerization, with biphenol
comonomer ........................................................................................................................78
Scheme 2.9. Hydroquinone-dichlorodiphenylsilane condensation. ...................................79
Scheme 2.10. biphenol-dichlorodiphenylsilane condensation. ..........................................79
Scheme 2.11. Biphenol-dichlorodiphenylsilane condensation in toluene. ........................81
Scheme 2.12. Hydroquinone-dichlorodiphenylsilane condensation
in THF/toluene mixed solvent system. ............................................................................84
Scheme 2.13. Biphenol-dichlorodiphenylsilane condensation in THF .............................87
xxv
Scheme 3.1. Condensation polymerization reaction between dichlorodiphenylsilane
and biphenol. Dichlorosilanes were added in 5% excess to obtain chlorosilane
endgroups ........................................................................................................................132
Scheme 3.2. Lithiumphenylacetylide endcapping reaction .............................................133
Scheme 3.3. Fluorophenyethynylphenol endcapping reaction. .......................................133
Scheme 3.4. Phenol endcapping reaction ........................................................................138
Scheme 3.5. Silanol thermally induced cross-linking reaction with phenyl
substituted siloxane. .........................................................................................................148
Scheme 4.1. Polymerization reaction of 4,4'-buta-1,3-diyne-1,4-diyldiphenol
and dichlorodiphenylsilane. ............................................................................................168
xxvi
List of Symbols
Tg glass transition temperature
Tm melting point
F Helmholtz free energy of the system
∆F change in Helmholtz free energy
U internal energy of a system
T temperature
S entropy
f applied force
σ stress
A area over which force is applied
ε strain
L final dimension of test sample
Lo initial dimension of test sample
E Modulus
Vf free volume of a polymer
Vtot total volume occupied by a polymer
xxvii
Vpol volume of polymer chains (assuming 100% packing efficiency, no free
volume)
αR volumetric ( cubic) expansion coefficient of rubbery state
αG volumetric ( cubic) expansion coefficient of glassy state
De Deborah number
t(inh) inherent relaxation time of the polymer
t(obs) time of observed deformation process
L one dimensional diffusion length
t time
D Fick’s diffusion constant
M molecular weight of a polymer chain
G* complex shear modulus
H Planck’s constant
c speed of light
λ wavelength
ν wavenumber
Mw weight average molecular weight
Mn number average molecular weight
Mn number average molecular weight
γ ratio of functional groups in monomer 1 to the functional groups of
monomer 2
ρ extent of reaction (ρ=1=100% reaction completed),
xxviii
Mrepeat mass of the repeat unit of the polymer
G’ storage shear modulus
G” loss shear modulus
tanδ G’/G”
η* complex viscosity
η0 zero shear viscosity
K proportionality constant of Fox-Flory Power law
Ea activation energy
R Rydberg (Gas) constant
Cp heat capacity
T/1/2∆Cp temperature at half-height of the step change in heat capacity, often
reported as the vitrification point of a thermoset
Rate heating rate of DSC experiment, for use with Ozawa activation energy
calculation
Tpeak peak temperature of exothermic transition measured in a DSC curing
experiment
α extent of reaction, also called degree of cure
∆Htotal total exothermic heat of reaction for a fully cured specimen
xxix
∆Hresidual residual exothermic heat of reaction for partially cured specimen
xxx
List of Abbreviations
FTIR Fourier transform infrared spectroscopy
NMR nuclear magnetic resonance spectroscopy
DSC differential scanning calorimetery
TGA thermogravimetric analysis
BDE bond dissociation energy
PTFE polytetrafluoroethylene
TFE tetrafluoroethylene
MVE fluorinated methyl vinyl ether
HFP fluorinated propylene
VF2 vinylidine fluoride (CH2CF2)
ASTM American Society for Testing and Materials (standards organization)
FFKM ASTM designation for perfluorocarbon elastomers
FKM ASTM designation for fluorocarbon elastomers
EPDM ASTM designation for ethylene-propylene-diene elastomers
SBR ASTM designation for styrene-butadiene rubber (elastomer)
NBR ASTM designation for acrylonitrile-butadiene rubber (elastomer)
xxxi
GPC gel permeation chromatography (also known as SEC)
SEC size exclusion chromatography
THF tetrahydrofuran
DCDPS dichlorodiphenylsilane
DPS diphenylsilane
HQ hydroquinone
TEA triethylamine
PDI Polydispersity index Mw/Mn
PETI phenylethynylpolyimide
BDDP 4,4'-buta-1,3-diyne-1,4-diyldiphenol
xxxii
Abstract:
High Temperature Hybrid Elastomers
Kerry Drake
Advisor: Dr. Yen Wei
Conventional high temperature elastomers are produced by chain polymerization
of olefinic or fluorinated olefinic monomers. Ultimate thermal stabilities are limited by
backbone bond strengths, lower thermal stability of cross-link sites relative to backbone
bonds, and depolymerization or “unzipping” at high temperatures. In order to develop
elastomers with enhanced thermal stability, hybrid thermally cross-linkable polymers that
consisted only of organic-inorganic and aromatic bonds were synthesized and evaluated.
The addition of phenylethynyl or phenylacetylinic functional groups to these polymers
resulted in conversion of the polymers into high temperature elastomers when cross-
linked by thermal curing.
Polyphenyoxydiphenylsilanes were synthesized via several different condensation
reactions. Results of these synthetic reactions, which utilized both hydroquinone and
biphenol as monomers, were systematically evaluated to determine the optimal synthetic
conditions for subsequent endcapping reactions. It was determined that
dichlorodiphenylsilane condensations with biphenol in toluene or THF were best suited
for this work. Use of excess dichlorodiphenylsilane yielded polymers of appropriate
molecular weights with terminal reactive chlorosilane groups that could be utilized for
coupling with phenylethynyl reagents in a subsequent reaction.
xxxiii
Two new synthetic routes were developed to endcap biphenoxysilanes with
ethynyl containing substituents, to yield polymers with cross-linkable end groups.
Endcapping by lithiumphenylacetylide and 4[(4-fluorophenylethynyl)]phenol yielded two
new polymers that could be thermally cross-linked on heating above 300 °C. Successful
endcapping was verified chemically by 13
C NMR, FTIR and Raman analysis. Exothermic
peaks consistent with ethynyl curing reactions were observed in endcapped polymers by
DSC.
A new diacetylinic polymer was prepared through reaction of 4,4'-buta-1,3-diyne-
1,4-diyldiphenol and dichlorodiphenylsilane. This aromatically substituted siloxane
polymer contained thermally cross-linkable diacetylene links in the backbone. FTIR,
Raman, and 13
C NMR analysis confirmed the diethynyl group was present in the
polymer. DSC analysis showed the polymer had a Tg of 130 °C, and a strong exothermic
cure peak at 260 °C. TGA analysis in nitrogen showed a 5% weight loss temperature of
541 °C and a pyrolysis yield of 82% at 800°C. Parallel plate rheological testing
confirmed the polymer cross-linked through monitoring of changes in viscosity during
heating. After curing above 260 °C, the polymer vitrified, with no detectable Tg observed
on subsequent DSC analyses. Curing at 210 °C for 30 minutes in nitrogen resulted in a
partially cross-linked material that exhibited elastomeric properties above Tg. Curing
under these conditions resulted in an estimated 25% degree of cross linking, and an
increase in Tg to 146 °C. The activation energy of thermally initiated curing of the
diacetylene groups was estimated to be 100 kJ/mol from DSC data using the Ozawa
method.
1
Chapter 1: Overview of the structure-property relationships of elastomeric
materials
1.1. Organization of this dissertation
Based on societal needs for higher temperature elastomers, the demonstrated high
thermal stabilities of aryloxysilanes, and the possibility of utilizing new phenylethynyl
cross-linking chemistry to produce elastomeric materials, this research project was
designed to explore the synthesis and characterization of new hybrid elastomers that
contain both inorganic silicon-oxygen and organic carbon-carbon bonds in the polymer
backbone.
Chapter 1 of this dissertation provides the basic concepts and background on high
temperature elastomeric materials and their characterization. Those readers who are well
versed in this area may skip this chapter. Chapter 2 details the synthetic investigations for
producing aryloxysilane polymers. This includes evaluation of several different synthetic
methods and the results, including synthesis with stoichiometric excesses of
dichlorosilanes to yield reactive end groups for subsequent functionalization with
phenylethynl groups. Chapter 3 describes the synthesis and characterization of new
phenylethynyl endcapped aryloxysilanes. In addition, uncapped and phenol capped
polymers were synthesized and characterized as well, for use as baselines for evaluating
the impact of endcapping on thermal and rheological properties. Chapter 4 presents the
synthesis and characterization of main-chain phenylethynyl aryloxysiloxanes. New
polymers were synthesized by reaction of 4,4'-buta-1,3-diyne-1,4-diyldiphenol and
dichlorodiphenylsilane. Chemical, thermal, and rheological characterization of this
polymer system was performed and is detailed in this chapter. Chapter 5 provides an
2
overall conclusion for these studies, along with recommendations for future work to
further develop an understanding of these systems and their potential for use to extend the
range of elastomeric materials in demanding services. Appendix A, an original research
proposal prepared to fulfill Drexel’s PhD candidacy requirements, is also included for
reference.
1.2. Historical background of rubber and elastomer technology
Elastomers have been used for centuries. The first elastomers known were based
on natural rubber. Reports from Christopher Columbus’ 2nd
expedition to the new world
in 1495 include references to the natives playing games using a rubber ball made from
the milk of a tree 1. The tree species, which was later named Hevea brasiliensis, yields a
milk-like substance which contains a high proportion of organic molecules that contain
unsaturated double bonds resulting from polymerization of a diene monomer, i.e.,
isoprene (Scheme 1.1).
CH2
C C
CH2
CH3 H
CH2
C C
CH2
CH3
H
x
Scheme 1.1. Natural rubber monomer and polymer (cis 1,4 polyisoprene).
These double bonds could be thermally activated for cross linking2,3
but the thermal
process is not very efficient, relative to vulcanization processes developed later by
Goodyear. The balls exhibited elastomeric properties (bouncing, etc), but would not hold
their shape well, and would flatten when allowed to sit under stress such as gravity for a
period of time.
3
The material from the rubber tree was later found to be useful as an eraser for
pencil lead, hence the name “rubber”: to rub [on paper for erasing] 4. In the 1750s,
technology was developed to concentrate and purify the milk to produce durable goods,
such as shoes and tubes. However, since cross-linking technology had not yet been
developed, product performance was not optimal. The materials would still exhibit creep
and flow over time, shoes would lose their shape, material would sag over time, etc. It
was not until Mr. Goodyear developed the process using sulfur as a cross-linking agent
termed “vulcanization” in 1839 to controllably cross link natural rubber that commercial
applications of this new class of material were truly enabled1. Note that the term “cross-
linking” often has the same meaning as “curing” in industry.
As new synthetic sources of rubbery materials in addition to rubber tree milk were
developed, the term “elastomer” was coined to describe these materials. An elastomer is
defined as “any of a variety of elastic materials whose properties resemble rubber.” This
term was first used circa 1939, and its etymology is based on the combination of the
words elastic+ (poly)mer. Elasto is descriptive of the general elastic properties of the
material, and –mer, is shortened form of polymer, which is the generic class of this
material5.
A more technical definition of elastomer is “a polymer used above its Tg (i.e. its
glass transition temperature)”6. Ambient temperature is commonly used as a cutoff point
for Tg for rubbers, i.e. if the Tg < 23 °C the material fits the classification of a rubber
under normal conditions (i.e. 1 atmosphere pressure and typical room temperature of
~20°C). However, for high temperature service, the main requirement for an elastomer is
actually that the Tg must be below the service temperature, not necessarily ambient
4
temperature. Hendrick was one of the first researchers to realize that this paradigm was a
limiting factor for the development of elastomeric materials. Hendrick’s research group
developed high temperature cross-linked polyketone elastomers with Tgs in the 150 °C
range, while targeting service temperatures in the 250-350 °C range7. These materials
exhibited superior performance to conventional high temperature siloxane elastomers.
They concluded that as long as the service temperature was well above Tg, the material
would perform as an elastomer. Therefore it does not matter if a material has a Tg of 130
°C if it is being used at 300°C, it will still perform as an elastomer at 300 °C. This
concept of defining the properties of a material at its service temperature, rather than at
ambient temperature, is a cornerstone of the development of the elastomeric materials
described in this work. Many of the materials synthesized in our studies are glassy at
room temperature, but exhibit elastomeric properties at high temperatures, thus
performing as elastomers in this temperature regime.
1.3. Structure property relationships of elastomers
1.3.1. Flexible linkages
The properties of elastomers are based on the conformations of the polymer
chains, their potential for extensibility, and also the retractive forces imparted by cross-
linking. Linear polymer chains consist of both rigid and flexible segments. The chains
can move and change conformation by rotation around flexible linkages. Some common
flexible linkages for elastomeric polymers are methylene, ether, thioether,
difluoromethylene, siloxane (silylether), and phosphazene, which are shown in Figure
1.1. A summary of properties for representative polymers that contain the flexible
linkages in Figure 1.1 is given in Table 1.1.
5
CH3
CH2
CH2
CH2
CH3CH3
O
CH2
O
CH3CH3
S
CH2
S
CH3
F3C
CF2
CF2
CF2
CF3
P N
R
Rn
Si
O
Si
O
R
R
R
R n
Figure 1.1. Common flexible chemical linkages for elastomers. Clockwise from upper
left: methylene, ether, thioether, phophazene, siloxane, fluoromethylene.
6
Table 1.1. Comparison of Tgs of selected linear polymers with flexible backbone
linkages, and predominantly methylene secondary linkages8-12
.
Polymer
Crystalline/
amorphous Flexible Link
Other
Links
Tg
(amorphous
region)
Tm
PTFE crystalline -CF2- -CF2 127⁰C
Plus strong
γ transition
at -110ºC in
amorphous
domain
due to CF
segmental
movements.
330⁰C
Polyoxy-
methylene
crystalline -O-(ether) -CH2- -10 to -
80⁰C
Polyethylene
crystalline -CH2- -CH2- -24⁰C
PVDF crystalline -CF2- -CH2- -40⁰C
Polydimethyl-
phosphazene
amorphous
P N
R
R
-- -50⁰C N/A
Polysulfide
Rubber
amorphous -S-
(thioether)
-CH2- -50⁰C N/A
Polymethyl
hydrido-
siloxane
amorphous -O-
(silylether)
-CH3SiH- -138⁰C N/A
When comparing Tgs of materials with these chemical linkages in the backbone
and similar non- bulky substituents, several key findings are notable (Table 1.1):
1) Crystallinity restricts molecular motion and increases Tg of amorphous domains in
a polymer, relative to amorphous materials with the same chemical composition.
7
2) Inorganic links (silylether, phosphazene) are more flexible than carbon-carbon
links, as evidenced by their lower Tgs.
3) Tg Trends CF2>-O- > -CH2->phosphazene >thioether >silylether
4) Silylether compounds possess the lowest Tgs.
These findings can be understood by considering molecular conformations of the
polymers, and the effects flexible links that allow movement of the polymer backbone
have on molecular mobilities13-15
. Preferred conformations of collections of polymer
chains are coiled rather than extended (extended chains=less disorderedness=lower
entropy). When coiled chains are stretched, the number of possible molecular
conformations are reduced, which results in a decrease in entropy of the system. Thus
when the force is removed, entropy increases through the re-coiling of the extended
polymer chains. This means the retractive force for a stretched elastomer is entropically
driven, rather than driven by changes in internal energy. The derivation of this is fairly
straightforward.
If we consider the Helmholtz free energy of a system (analogous to Gibbs free
energy, but for nonisobaric conditions), we see that the free energy of the system
increases (when a force f, is applied and work is done on the system. Therefore the
increase in free energy is equivalent to the Helmholtz free energy of the system. The
Helmholtz free energy can be described by the following equation:
F=U-TS
where F= the free energy of the system, U= internal energy, T=Temperature, and S=
entropy. Changes in free energy can be described by a modified equation:
8
∆F=∆U-T∆S
So f, applied force, is equal to ∆F, the change in Helmholtz free energy. At conditions of
constant temperature and volume the retractive force can be described by the following
equation:
VTVTVT L
ST
L
U
L
Ff
,,,
∂
∂−
∂
∂=
∂
∂=
Where F,U, and L are the Helmholtz free energy, internal energy, and length of the
specimen, respectively.(Note the constant T,V constraints. For a rubber, the Poisson
ratio~ 0.5, which means volume does not change on stretching, there is a change in length
only. This satisfies the constant volume constraint).
For rubbers, the energies of most possible conformations are equal, this means a
change from one conformation to another does not result in a change in internal energy,
i.e. 0
,
=
∂
∂
VTL
U
However, on stretching or deformation the uncoiling of the polymer chains results
in a reduction in the number of available conformations results in a decrease in entropy,
so the total free energy of the system is increased. Experimentally, the percentage of
retractive force due to energetic vs. entropic considerations can be measured through
extension experiments at different temperatures. Entropic contributions to the retractive
forces for several common polymers range from 80% to 90% of the total retractive force
16. So 80 to 90% of a rubber’s retractive force is due to entropy changes.
9
Entanglement of adjacent polymer chains is also thermodynamically favored, as
the entropy of mixing is positive 17
(favors mixing- higher disorder in a mixed system
than a segregated system). Therefore, intertwined chain coiling is a preferred state in an
ensemble of polymer molecules. However, the entanglement is not a static situation, it is
a dynamic system governed by processes that can be described by statistical
thermodynamics.
At any given moment there will be a certain number of physical entanglements of
polymer chains. The entanglements impart properties such as resilience and resistance to
deformation when an outside force is imposed on the material. As will be discussed later,
the timeframe of the applied stress is critical. If force is applied over a short time frame
relative to the timeframe for molecular motion, the polymer will respond elastically. If a
force is applied over a long timescale, the polymeric material will exhibit viscous
deformation under the applied load. The demonstration of both viscous and elastic
properties is a fundamental concept of polymer mechanics, known as viscoelasticity.
1.3.2. Viscoelasticity Viscoelasticity is a term used to describe the behavior of a material which has a
combination of viscous and elastic properties. In order to review the concept of
viscoelasticity and its impact on the molecular design and the resulting mechanical
properties of elastomeric materials, some definitions of material properties and
measurement methods are required.
1.3.2. 1. Mechanical properties definitions
10
Stress is the term used to describe the force per unit area imposed on a material according
to the equation:
A
f=σ
Where σ= stress, f = applied force, and A=area over which force is applied:.18
.
Strain is the change in dimensions of an object when acted upon by an outside force. It is
described the following equation:
Lo
LoL −=ε
Where ε=strain, L= final dimension of test sample, and Lo= initial dimension of test
sample 18
.
Modulus is defined as the slope of the curve when stress and strain are plotted. It is
described by the equation:
ε
σ=E
Where E=Modulus, σ= stress, and ε=strain
A graphical representation of a tensile test is instructive in demonstrating the
changes in part geometry as stress is applied. A tensile test consists of applying a force to
a bar of material and measuring the stretching or displacement caused by that stress. This
is shown in Figure 1.2 below.
11
Figure 1.2 . Tensile test specimen(left), Tensile test in progress(right), showing the
change in gauge length/strain from 2 inches to >2 inches as test progresses (increase in
distance between horizontal marks in middle to right figures)19
.
Plots of stress on a y axis, and strain on an x-axis are commonly used to represent
the response of a material to a given stress, at the specified temperature and strain rates of
the test. A representative stress-strain plot is shown in Figure 1.3.
12
Figure 1.3. Representative stress/strain curve, showing modulus (OP tangent line), Proportional limit (P), Elastic Limit (E), Yield (Y), ultimate strength(U), and strain at
break (R) 20
.
High resistance to deformation (higher force required per unit of deformation) is
seen when testing a material with a high modulus. Lower force required for deformation
is found when testing a material with a low modulus. For an elastic material, when a
force acts upon a fixed object, the deformation induced by the imposed force is
reversible, thus once the force is removed, the deformation will reverse and the object
will return to its original conformation and dimensions. For viscous materials, such as
gelatin or petroleum grease, displacement due to an imposed force is irreversible. Viscous
materials show stress/strain curve similar to those in Figure 1.4. 1.4(a) will show an
13
unloading curve that exactly overlaps the loading curve.1.4(b) will retain its deformed
state after the load is removed, and will not return to its starting dimensions.
Figure 1.4. Stress-strain curves for perfectly elastic material (a) and perfectly viscous
material (b). 20
However, for most materials there is a point where the deformation is no longer 100%
recoverable. This is known as the elastic limit, (E on Figure 1.3).
Once a material is stressed past the elastic limit, the material will exhibit a plastic
region, where the stress/strain relationship is non-linear. In this range the stress level is
large enough to actually free dislocations (crystal defects) in a crystalline material, or
uncurl polymer chains and entanglements in an amorphous polymer. The movement of
the crystal defects and/or unfurling of the chains allows for irreversible plastic movement
of the material.
The point where the slope of the stress strain curve reaches zero is known as the
yield point (Y in Figure 1.3). Once a polymer is above its yield point, the defects and/or
14
polymer chains continue their movement with no additional increase in applied force.
When applied force is maintained on the material after yielding, the material may either
break, or reach a point where additional force is required to continue extension. This is
known as strain or work hardening21
.
A material is classified as elastic when it exhibits completely recoverable
deformation. Examples of elastic materials are metals, plastics and cross-linked rubbers.
Ceramics also exhibit elastic properties22
, however, they are in general very brittle
materials and exhibit only very low strains before yield/break. Metals typically have yield
points of a few percent, plastics can have elastic limits of a ~8 to 10 %, whereas rubbers
can be deformed to several hundred percent strain, and show full recovery to initial
dimensions. Values of Moduli, yield and break strengths, breaking strains and
proportional limits for some common materials are listed in Table 1.2. Representative
stress-strain curves for several different types of materials are shown in Figure 1.5.
15
Figure 1.5. Stress/strain curves of representative material types. Note the differences in
modulus, yield behavior, and strain to break20
.
In a crystalline material above the yield point, dislocations in the crystalline
domains begin to move. In an amorphous polymeric material above the yield point,
polymer chains begin to extend. In a semicrystalline polymeric material, both chain
extension and dislocations occur simultaneously. The degree of each can be estimated by
evaluating the amount of recoverable strain after yield. The amount that is recoverable is
related to the chain extension and the ability of the chains to return to their initial state
and position20
. The unrecoverable portion is related to the irreversible movement of the
extended polymer chains and dislocations. If the goal is to develop an elastomeric
16
material, crystallinity should be avoided, except in the case where the crystallites are used
as physical cross-links.
Table 1.2. lists some representative values of material properties for several
different classes of materials. Note the differences in proportional limit values vs.
ultimate tensile values for many of the materials, and the lack of proportional limit data
for “brittle” materials, such as cast iron and concrete.
Table 1.2. Comparative table, tensile properties of selected material types23
.
17
Resilience, the ability of a material to absorb energy under stress and then return
to its original state when unloaded 20
, is a key property for elastomers. High resilience
equates to high resistance to irreversible deformation. Hardness is used as a metric for
quantifying the resilience of a material. Hardness is measured by indentation of a probe
into a material. The depth of penetration of the probe is related to the modulus of the
material25
, with high modulus materials showing smaller depth of penetration and high
hardness. Lower modulus correlates with lower hardness24,25
. There are exceptions in the
special case of filled materials where the indenter width is of the same magnitude as filler
particles and/or crystallites in semicrystalline polymers26
. In these cases, the indenter may
penetrate between particles and give a lower hardness value than shown by macro-
mechanical properties, such as modulus; however for unfilled materials hardness
measurements and modulus are very interrelated.
For rubbers, a cylindrical indenter with a truncated cone is used, and hardness is
reported in Shore units (higher Shore number=harder material) 25
. Table 1.3 lists
hardnesses of some common rubber types, where Hardness scales O<A<D.
18
Table 1.3. Typical hardness values for elastomers when tested at ambient conditions25
.
1.3.2.2. Temperature dependence of properties of elastomers Significant differences are seen for performance of elastomeric materials at
different temperatures, especially when properties below and above Tg are compared.
Below Tg, polymers are stiff and brittle; above Tg, the polymers are soft and resilient if
they are cross-linked or semicrystalline (Figure 1.6 below). If polymers are amorphous
and uncrosslinked, they will exhibit viscous flow (not shown).
19
Figure 1.6. Tensile stress-strain curves for styrene-butadiene block thermoplastic
elastomer (Tg=-17°C, +88°C)27
.Note the transition from glassy behavior with high
modulus and low strain at break below Tg, (-60 to -30°C) to rubbery behavior with low
modulus and high elongation at 23°C.
The mechanical behavior of polymers is a physical manifestation of
viscoelasticity. The molecular basis of viscoelasticity is the actual movement of the
polymer molecules in the bulk material. Movement is due to both thermal vibrations, and
response as a result of applied stress from an outside stimulus (loading/unloading,
stretching, hydrostatic pressure, etc.).
As with all materials at temperatures above absolute zero, polymer molecules are
in a constant state of motion, sliding around and over one another due to random thermal
fluctuations. One concept used to describe this behavior is reptation, the “snake-like”
movement of large polymer chains.28
29-31
. The polymer chains are represented by
20
analogy to a collection of snakes, constantly writhing and moving. The polymer chain
cannot move sideways, due to the obstacles blocking its path, therefore it can only move
forward, in a movement pattern analogous to that of a snake, writhing between obstacles.
Figure 1.7. Representation of reptation of a polymer chain P, moving through fixed
obstacles, O28
.
In this analogy, the space between the chains is defined as free volume, and would
be equivalent to the volume between obstacles, O, in the Figure 1.7. The effective
volume occupied by the polymers is actually significantly higher than the volume that
would be occupied if the chains were neatly ordered and packed. A review of the
concept of free volume is useful to understand its role in the properties of elastomers.
1.3.2.3. Free volume concept, and the dependence of free volume on temperature
The volume of a polymeric material can be broken down into two components,
total volume, Vtot, and volume occupied by the polymer chains, Vpol. Vtot is always larger
21
than Vpol due to the presence of interstitial vacancies. The volume of the interstitial space
between the polymer molecules is known as the free volume of the polymer, Vf
Vf=Vtot-Vpol
Below Tg, the ratio Vf/Vtot is fixed. Above Tg, due to long range segmental motion, Vf
increases more rapidly than Vpol. This is shown schematically in Figure 1.8. Note the
different volumetric expansion coefficients for rubbery and glassy phases.
Figure 1.8. Specific volume of a typical amorphous polymer vs. temperature. αR and
αG=volumetric ( cubic) expansion coefficients of rubbery and glassy states,
respectively32
.
Above Tg, the specific volume curve slope is proportional to αR which is the
volumetric expansion coefficient of the rubbery phase. Below Tg, where long range
segmental motion is no longer possible, the slope is proportional to αG .
22
As thermal energy is absorbed by the system, the polymer chains rotate and
vibrate faster and thus take up more space. As the polymer is heated above Tg, long range
segmental motion can occur, and free volume increases at a significantly faster rate. If
heat is removed from the system, the chains will move more slowly, and take up less
space (lower free volume). At temperatures below the glass transition temperature, long
range segmental motion is no longer possible, and polymer chains essentially “freeze” in
place with only short range segmental movement possible over the timeframe of most
short term mechanical property measurement experiments (<1day), and the material
exhibits “glassy” properties.
1.3.2.4 Time dependence of polymer properties (Deborah Number)
Over long periods of time, long range segmental motion can still occur. The
equivalency of long term deformation and shorter term higher temperature deformation is
one of the key differentiators of polymers relative to other material classes (metals,
ceramics, etc). The timeframe of the deformation relative to the timeframe for molecular
movement/relaxation is known as the “Deborah Number” and is described by the
following equation33,34
:
)(
)(
inht
obstDe =
Deborah Number, where t(inh)=inherent relaxation time of the polymer, t(obs) =time of
observed deformation process. Note: this term was designated the Deborah number in
reference to a passage in the Bible, where the prophetess Deborah observed “The
mountains flowed before the Lord”.
If De>1, relaxation processes are much faster than the time of the applied stress, and the
material behaves as a liquid; when De<1, relaxation process are slower than the time of applied
stress, and the material will behave like a solid ( no measureable deformation under load)33, 34
.
23
The significance of this parameter is profound; if we take into account the time frame of
our experiment as well as the temperature, we see that long times are equivalent to high
temperatures. This means that properties of viscoelastic materials over long periods of time can
be modeled using the time-temperature-frequency superposition principle, where molecular
movement over long time periods is essentially equivalent to shorter timeframe measurements at
elevated temperatures.
1.3.2.5. Polymer self-diffusion, and its effect on elastomeric properties Another manifestation of molecular movement via reptation is that polymer chains can
actually physically migrate from one area of a polymer matrix to another (self-diffusion). The
speed of this movement is related to the relative free volume of the polymer. Higher free volume
between polymer chains allows for faster movement of all chains and a higher probability for a
single chain to migrate from one area in a polymer matrix to another area.
Polymer molecules, although seeming to be fixed in space, are actually
continually moving, and will over time diffuse from one region to another in a bulk
matrix (self-diffusion), especially when the polymer is above its Tg and higher free
volume is available for movement. The diffusion length at a specified temperature is in
general described by the mass diffusion length equation derived from the Einstein
Smolouchowski relation35
:
L=(2Dt)1/2
Where L= one dimensional diffusion length, t=time, and D=Fick’s Diffusion constant
(cm2/sec).
Diffusion rates in polymers are temperature dependent, again based on free
volume availability and also if the polymer is above its Tg, where long range segmental
24
motion is possible and transport rates are greatly enhanced 36, 37
. Self-diffusion rates in
amorphous polymers are proportional to molecular weights of the polymer chain,
following the relationship shown in the equation below28
:
D α M-2
Where D= diffusion coefficient and M= molecular weight of the polymer chain.
Values of D for polymers above Tg have been reported in the literature and
typically range from 10-12
to 10-6
cm2/sec
38 . Diffusion constants for polymers below Tg
are much lower. Note that higher rates of self diffusion result in lower viscosities and
thus less resistance to irreversible deformation under applied loads. The temperature
dependence of the Diffusion coefficient, D is significant, especially when comparing
values above and below Tg). At T> Tg, D increases by several orders of magnitude16
. This
has the net effect of increasing polymer mobility, and thus decreasing viscosity, and
increasing flow rates under applied force.
1.3.2.6. Effects of physical entanglements on mechanical properties of polymers
The statistical nature of entanglements governs instantaneous properties, along
with the relative free volume of the polymer at the testing temperature. At fixed free
volume (below Tg), long range segmental movement is inhibited, so polymers will have a
greater resistance to deformation due to free-volume restricted self-diffusion and to
entanglements. When stress is applied and then released, entanglements will tend to pull
the chains back into original conformations39
. At higher free volumes (testing
temperatures above Tg), individual chains have more room to move and slide over each
other when stress is applied, resulting in a lower “effective” number of entanglements
25
The physical manifestation of this is a greatly reduced modulus above Tg, which is
desired for elastomers. For uncrosslinked, amorphous polymers, individual
entanglements will have a higher likelihood of sliding over each other due to the
enhanced mobility of the chains above Tg. The result is viscous flow, not elastic
deformation.
This is illustrated by examining the modulus vs. temperature curve generated by
DMA for a semicrystalline polymer, such as PEEK (Figure 1.9). The modulus typically
drops around 80% when the material passes through its Tg, which is due to the ability of
the molecules to slide over each other due to the increase in free volume and the greater
capability for movement. However, the chains are restricted from full freedom of
movement (diffusion) because the crystallites act as physical tie layers, and thus a
significant portion of total mechanical stiffness is retained. The polymers are fixed within
the crystalline domains, and thus can extend but not diffuse away from their initial
positions. Thus the modulus drop-off at Tg is related mainly to the loss of entanglement
movement restrictions in the amorphous domains. In the case of amorphous polymers,
the drop off above Tg is >99%40
. Note that above the melting point, Tm, of the
semicrystalline PEEK material, where physical cross-links are no longer present, the
modulus drops by several orders of magnitude ( Figure 1.9, T>330 °C). This is the point
where viscous flow occurs; the polymer is essentially in a molten state.
26
0 50.0 100.0 150.0 200.0 250.0 300.0 350.0
temperature (°C)
10000
1.000E5
1.000E6
1.000E7
1.000E8
1.000E9
1.000E10
|G*|
(P
a)
Tg=143C
Tm=343C
Figure 1.9. Rectangular Torsion DMA test of PEEK, showing shear modulus as a
function of temperature. Note the large decrease in shear modulus, G*, as the polymer
passes through Tg.
In addition, over time when exposed to an applied external force, the polymer
chains will actually flow so resistance to deformation will decrease over time. The
decrease in resistance to deformation is due to movement of the individual chains under
applied stress to generate a lower energy conformation that will reduce the stress16
. The
mechanism of the stress relaxation is the irreversible movement of the polymer chains,
via self diffusion-reptation.
27
Above Tg, as free volume increases, modulus decreases due to the availability of a higher
percentage of free volume, which makes it easier for individual polymers to self-diffuse
through the bulk matrix. By definition an elastomer is a material used above its Tg 6, so
understanding of the physics of this situation is critical in designing of a material to
function above its Tg without undergoing significant creep and/or actual irreversible
viscous polymer flow (“melting”).
Another way of viewing these transformations and the effect on the material is
through a plot of induced strain vs. temperature (Figure 1.10).
Figure 1.10. Deformation under load vs. Temperature for a typical cross-linked
elastomer. Region 1 =glassy state, Region II = transition state, Region III = rubbery
plateau, Region IV =viscous flow34
.
By examining Figure 1.10, one can see that below Tg (Region I) the material
exhibits limited deformation. In Region II, the material passes through its glass transition,
and higher strains are seen. In Region III, the material exhibits a “rubbery plateau”- this
is the temperature region where the material can function as an elastomer. In Region IV,
28
the material exhibits viscous flow, either due to melting if semicrystalline, or due to
degradation of the polymer.
In summary, in order to obtain a polymer with desired elastomeric properties, one
needs a material with:
1) Low moduli relative to the force required to deform them, say for a sealing
application
2) High resiliency, where deformation is essentially 100% reversible when the
external stress is removed
3) Some method of physically locking polymer molecules into place, either through
crystallite ties, or chemical cross-links
Table 1.4 provides a summary of the various categories of polymers, and the important
properties that dictate if the material might perform acceptably as an elastomer.
29
Table 1.4. Comparison of generic classes of polymeric materials, and their respective
fulfillments of property requirements to allow classification as an elastomer.
Polymer Type Modulus
Below Tg
Modulus
Above Tg
Elasticity
below Tg
Elasticity above
Tg
Amorphous High Low High, until
yield point
(<10%)
None
( viscous flow)
Semicrystalline High medium High, until
yield point
(<10%)
Limited to yield
point at test
temperature
Thermoset High Low to high
depending on
level of cross
linking
High, until
yield point
(<5%),
typically
brittle
materials
Dependent on
level of cross
linking
( low cross-
linking=high
extensibility,
high cross
linking =low
extensibility)
Based on the criteria listed above, the only classes of polymeric materials that can
effectively act as elastomers are semicrystalline polymers or thermoset polymers when
used above their glass transition temperatures.
30
1.4. Cross-linking of polymers for elastomeric properties
In the situation described earlier for rubber articles manufactured prior to 1839,
most rubber products were amorphous in nature and only lightly thermally cross-linked.
Thus these materials had somewhat limited utility, due to undesired flow properties.
When Goodyear developed vulcanization, what he actually did was to develop a method
to chemically tie molecules in a fixed position so that their initial conformation was
frozen. The matrix could then deform under external applied stresses, but return to
original conformation when the external force was removed. Due to the chemical bonds
between adjacent chains, the local environment was now fixed, so stretched, uncoiled
polymers chemically bonded to nearest neighbors would return to their coiled
conformation when stress was released, and since the molecules were tied together via
cross linking, polymer molecules would return to the same conformation with respect to
nearest neighbor molecules. The practical result of this is that the bulk matrix would
return to its original configuration, so a stretched rubber band would return to its original
length, or a compressed rubber ball would return to its original spherical configuration
when the applied external stresses were removed.
The number of cross-links in a polymeric material has a profound effect on the
mechanical properties and the Tg of the material. The degree of cross linking is typically
defined using units of cross-links per gram. The change in Tg, ∆ Tg, is proportional to
cross-links/gram and can be described by the following equation:
∆ Tg = Z*χ’
31
Where∆ Tg = increase in Tg due to an increase in cross-link density, χ’=cross-link
density, units=cross-links/gram, and Z=constant (specific to each polymer system,
representative values 2 to 4 x104)41
.
The degree of cross-linking also affects modulus of a polymeric material at all
temperatures. General trends are shown in Figure 1.11.
Figure 1.11. Trends in shear modulus relative to cross-link density42
Note Mc~1/χ’.
1.5. Crystallinity of a polymer, and the effect on elasticity
The ability of a polymer to crystallize is a function of the structure of the
backbone. In general, if a polymer backbone has a regular repeat structure that allows
packing into a crystalline array. Factors such as tacticity, number of carbon atoms in any
aliphatic repeat units in the backbone, and bulkiness of substituents affect the ability to
form a crystalline structure32
. Crystalline structures are desirable in some cases, but this
comes at the expense of reduced elasticity, since the crystallites act as both physical
32
cross-links and reinforcing agents (increase modulus). This is shown graphically in
Figure 1.12. At high crystallinity levels, thermoplastic elastomers exhibit high Moduli,
and thus may have less utility in applications that require softness and resilience.
Figure 1.12. Trend in shear modulus relative to% crystallinity (physical cross-links). Note
Elastomers are typically >30% crystalline 42
.
In summary, the chemical requirements for a material to structurally perform as
an elastomer include:
1) Flexible links to allow coiling of the molecules in an unstrained state, and
allow for uncoiling under applied stress
33
2) A method for locking molecules in-place to eliminate the possibility of
viscous flow by reducing or eliminating self diffusion. This is accomplished by
cross-linking or by formation of crystalline domains that act both as tie layers and
as reinforcing agents41
3) Control of the cross-link density or % crystallinity to optimize properties
for the specific requirements of the material
4) Dependence on crystallinity as the sole tie mechanism should be avoided,
as material can eventually melt , crystallize over time, or will be too stiff for use
as an elastomer ( may perform more like a thermoplastic)
However, fulfillment of the structural and mechanical requirements alone will not
guarantee an elastomeric material will perform well at high temperatures. Consideration
of the chemical and structural properties that govern thermal stability must now be
considered.
1.6. Thermal stabilities of polymers
Many excellent references are available that review thermal stabilities of polymers 43, 44
.
A typical starting point for discussion in most of these sources is the bond dissociation
energies of the constituent chemical bonds of the component molecules. Bond
dissociation energy is defined as the energy required to homolytically cleave a chemical
bond. Bond enthalpy is defined as the bond dissociation energy at 298K 45
. The potential
energy diagram of a chemical bond relative to its bond length is shown in Figure 1.13.
These values can be measured by several experimental methods: photoionization mass
34
spectrometry, radical kinetics experiments, or gas phase acidity/electron affinity
experiments.
Figure 1.13. Schematic of bond energy, bond length, and energy required to cleave a
bond, from Blanksby et al45
.
According to Dr. Tibor Kelen, author of a monograph on polymer degradation,
and an expert on thermal stability of polymers: “The dissociation energies of the various
bonds in the polymer may determine the course of degradation: the process always begins
with the scission of the weakest available bond43
.” Kelen also cites a correlation with
50% weight loss temperature and bond dissociation energy (higher bond dissociation
energies= higher 50% weight loss temperatures).43
35
Polymer chemical bonds can be separated into two categories: backbone, and side
chain. Cross-linked polymers add a third category: the cross-linking bond.
For ultimate thermal stability, backbone bonds should be as stable as possible. If a
backbone bond is cleaved, the molecular weight of that polymer chain is immediately
decreased by the number of monomers in the smallest fragment. Side chain cleavage is
not necessarily desirable, but cleavage of a side chain bond results in a reduction in
molecular weight of only the weight of the fragment, not multiple repeat units. This is
shown graphically in Scheme 1.2.
SiH O Si O Si O Si OH SiH O Si Si O Si OHO+
SiH O Si O Si O Si OHSiH O Si O Si O Si OH
C
+
Scheme 1.2: backbone cleavage (top) vs. side group cleavage (bottom).
Both side group and backbone bonds can be grouped into several broad
categories: aliphatic (C-C, C-H), aromatic (C-C, C-H) , heteroatom (C-O,C-N) and a
category that can loosely be termed “inorganic” bonds. (P=N, Si-C, Si-O, etc.). If one
36
compares tabulated data on bond dissociation energies, the stabilities of the bonds can be
ranked according to their bond dissociation energies (Table 1.5).
Table 1.5. Bond dissociation energies of general classes of polymeric bonds2, 12, 45-49
.
Bond Bond Dissociation Energies
(Kcal/mol)
Si-C
( aliphatic)
69
P=N 72.5
Si-C
( aromatic)
79
C-O-C
(aliphatic)
80
C-C 83
C-O-C
(aromatic)
87
C-H 89
Si-O 108
C-F 123
C=C 145
In addition, other factors such as the possibility of resonance stabilization and
inductive effects of substituents may also play a significant role in bond dissociation
energies, and can in some cases even change the values43
. For example the bond
37
dissociation energy of a methyl group from toluene (methyl-phenyl bond) is 95 kcal/mol,
while the bond dissociation energy of a methyl group from propane is only 83 kcal/mol43
.
The highest bond dissociation energies are those of aromatic carbon; however this
link is highly rigid, so a polymer composed of mainly aromatic links would not exhibit a
Tg. If the bond dissociation energies in Table 1.4 are evaluated along with the Tgs of the
flexible links shown in Table 1.1 to determine the polymer type with the best properties
for an elastomer, we find the following:
• The silylether bond has both the highest bond dissociation energy, and the
lowest Tg (when linking comparable functionalities)
• The phosphazene linkage has a much lower bond dissociation energy, with
a slightly higher Tg. The result would be an elastomeric material with
lower thermal stability than a phenyl substituted siloxane. This is in fact
what has been reported in the literature for this material11
• The highest thermal stability elastomers available today are
fluoropolymers, but their ultimate thermal stabilities (based on bond
dissociation energies) are not as high as siloxanes, and their Tg s are higher
than silylethers. In addition, the tendency of fluoropolymers to crystallize
requires extensive modification of the polymer to obtain an elastomeric
material
1.7. Thermal stabilities of cross-link sites
38
Cross-link bonds are often the weakest link in high performance materials.
Hergenrother provides an excellent summary of several of the most common cross-link
chemistries used for high temperature polyimides66
(Table 1.6).
Table 1.6. Summary of the most common high temperature cross-linkable end-groups66
.
Nadic, vinyl, and other allylic cross linking agents have lower thermal stability
than the aromatic groups which make up the polymer backbone. Of all the cross-linking
39
chemistries listed in Table 1.6, phenylethynyl moieties to yield the most thermally stable
cross-links.
1.7.1. Diethynyl cross-linking
One variant of this cross-link chemistry, diethynyl cross-linking, has also been
investigated extensively, mainly in the form of diethynylsilyl cross linking groups50-53
.
Diethynylbenzene has also been evaluated as a stand-alone cross-linkable oligomer, again
with relatively good thermal stabilities demonstrated in the cross-linked polymer54,55
.
1.7.2. Ethynyl cross-linking mechanisms Numerous research groups utilize phenylethynyl cross-linking chemistries for
high temperature thermosets56-61
. All authors agree that this cross-link chemistry is very
stable at high temperatures. However, the mechanisms of curing are still not clear.
Several research groups have performed cure studies on model compounds to try
to understand the curing mechanism59, 62-64
. NMR analysis performed in these studies
indicated multiple curing reactions are possible, and multiple complex cure products are
formed. Initial curing appears to occur through addition across the triple bonds to form
polyenes (Figure 1.14).
40
Figure 1.14. Initial cure mechanism of phenylethynyl cross-link sites64
. .
Postcuring at higher temperatures facilitates development of complex conjugated
structures and polyaromatics. One study, which used 13
C labeled ethynyl reagents
determined the identities of several possible structures of cured products via solid state
NMR . These are shown in Figure 1.15.
.
Figure 1.15. Possible cross-linking products of thermally cured phenylethynyl
polyimides63
.
41
In addition, several researchers have observed that the curing of lower molecular
weight oligomers and model compounds appears to progress through different pathways
than higher molecular weight materials, with higher proportions of polyenes in the final
products. The authors attributed this to the higher mobilities of the chains in lower
molecular weight/lower viscosity resins which facilitate addition polymerization, rather
than cyclic formation62, 63, 65
. Whatever the mechanism, there is significant data available
that shows phenylethynyl curing chemistries generate very thermally stable cross-linked
materials. Therefore, the use of these cross-linking moieties is a logical step for the
development of new high temperature hybrid elastomers.
1.8. Survey of commercially available high temperature elastomers
In order to demonstrate the need for a new high temperature elastomeric material,
a survey of high temperature elastomers is useful to show where current performance is
limited with existing materials. Numerous elastomeric materials are commercially
available today, each of which has its own particular set of strengths and weaknesses. For
the purposes of this work, the main property of interest for improvement is the maximum
operating temperature. Table 1.7 shows the maximum operating temperature of several
commercial polymers, where the maximum operating temperature is defined as the
temperature at which aging for 1000 hours still shows a 50% retention in elongation8.
42
Table 1.7. Maximum operating temperatures of selected elastomers, adapted from
Threadingham et al 8, 9
.
Polymer
Maximum
Operating
Temperature,°C
Chemraz FFKM 315
FKM(perfluorocarbon) 250
Methyl vinyl silicone 250
FKM(fluorocarbon) 230
Fluorinated Silicone 210
Phosphazene 190
Ethyl Vinyl Acetate 175
Cross-linkable acrylate 160
HNBR ( butadiene-nitrile) 150
EPDM 150
Isoprene 140
NBR 125
acrylonitrile 125
chloroprene 100
polyurethane 80
Another tests often used to indicate thermal stability of polymers is weight loss
via TGA. Hergenrother cites 5% weight loss as a good indicator of a polymer’s ceiling of
thermal stability66
.
1.8.1. Fluoropolymer based elastomers
PTFE (polytetrafluoroethylene) is a highly crystalline thermoplastic polymer,
which melts at 327 °C9. However, its crystallinity inhibits its use as an elastomer in
homopolymer form. In order take advantage of the high thermal stability of fully
fluorinated polymers in an elastomeric form, copolymers were developed which
combined TFE with other fluorinated monomers that contained additional flexible
linkages and side chains to inhibit crystallization9. Co-monomers that contain functional
groups for cross linking are also used to allow further improvement via formation of a
43
cross-linked network67
. Cross-linking in general results in enhanced thermal stability of
polymers, relative to comparable uncrosslinked polymers68
, so addition of the cross
linking sites in FFKMs should by extension improve ultimate thermal stability as well.
The monomers used for fluoroelastomers are fluorinated ethers (MVE),
fluorinated propylene (HFP), and/or vinylidine fluoride (VF2), and are shown in Figure
1.16.
n
F
F
F
F
F
F
F
O CF3
H
H
F
F
F
F
F
CF3
Tetrafluoroethene (TFE)
fluoromethylvinylether (MVE)
vinylidene fluoride(VF2)
hexafluoropropene (HFP)
Cure Site R R'
m x
FFKM
R=TFE
R'=MVE
FKM
R=TFE
R'=VF2,HFP
Figure 1.16. FKM,FFKM -chemical structures of starting monomers.
Thermal stability and thus upper continuous use temperatures track well with % fluorine
content 69,70
(See Table 1.8).
44
Table 1.8. Service temperatures and fluorine contents of various fluorinated polymers9.
The highest % fluorine elastomeric materials are perfluoroelastomers, designated by the
abbreviation FFKM71
.
1.8.1.1. Perfluoroelastomers (FFKM)
This material consists of copolymers of tetrafluoroethylene (TFE),
perfluropolyethers (MVE), and monomers with active sites for cross-linking9. Even
though the polymer consists mainly of carbon-fluorine bonds, the backbone is still
composed of aliphatic C-C links, so at high temperatures the polymer unzips (rapidly
depolymerizes into monomers), rather than breaking down into larger fragments via
random scission processes. Decomposition via unzipping is the most common
mechanism of thermal decomposition of chain polymerized polyolefins, and results in
lower thermal stabilities of polyolefins than one would predict based solely on bond
dissociation energies67
. This trend is also observed in fluorinated olefinic polymers69
.
45
1.8.1.2. Fluoroelastomers (FKM)
Fluoroelastomers, are the next highest thermal stability commercially available
elastomers. They are copolymers of TFE, VDF and HFP. FKMs also exhibit unzipping
reactions on heating above their ceiling temperature for depolymerization. In addition to
the unzipping behavior, the introduction of C-H bonds in the vinylidene fluoride
monomer (-CH2F2-) essentially introduces “weaker” links in terms of thermal stability,
with bond dissociation energies from Table 1.5 of 89 kcal/mol for C-H bonds as
compared to 123 kcal/mol for C-F bonds. Loss of hydrogen (C-H cleavage) can also
result in formation of a double bond, and thus initiate the unzipping depolymerization
reaction at a significantly lower temperature than a fully fluorinated polymer, due to the
inclusion of this” weak link” 43
.
1.9. Olefinic elastomers
Olefinic elastomers are predominantly composed of alkene monomers that are
polymerized by chain polymerization. The most common olefinic elastomers are natural
rubber (polyisoprene) and synthetic (EPDM, nitrile-butadiene, and styrene-butadiene)8.
The structures of several of the most common monomers and polymers are shown in
Figure 1.17.
In general these materials have good mechanical properties and fairly low glass
transition temperatures, but thermal stabilities are poor due to the preponderance of
aliphatic linkages with associated low bond dissociation energies. These links are also
vulnerable to oxidative chain scission43
. See Table 1.6 for details.
46
H
H
H
H
H
H
H
CH3
ethylene
propylene
isoprene
acrylonitrile
R R' R"
m x
EPDM
R=ethylene
R'=propylene
R" diene
SBR
R=styrene
R'=butadiene
R"=N/A
CH2
C C
CH2
CH3 H
N
H
H
H
y
CH3
CH2
CH
CH
CH2
CH2
CH2
CH
CH
CH2
CH3
n
diene ( generic representation)
H
H
CH3
styrene
NBR
R=acrylonitrile
R'=butadiene
R"=N/A
Figure 1.17. Monomer and general structures for the most common olefinic rubbers.
1.10. Silicone elastomers
Silicone elastomers have very high thermal stabilities, due to their hybrid nature
and the high bond dissociation energies of the silylether bonds. Thermal stabilities of
siloxanes are often limited by the pendant groups, and by the possibility of cyclic
depolymerization, as the siloxane chains tend towards a curled conformation that on
heating results in transetherfication and significant depolymerization via cyclic
formation72
. Common siloxane pendant groups are shown in Figure 1.18.
47
Si
R
R'
O
n
CH3 Methyl
Phenyl
Vinyl
R,R'=
Figure 1.18. General siloxane structure, with common side group chemical structures.
Methyl groups in methyl-substituted siloxanes tend to react with oxygen in air at
temperatures above 220°C12
. Cross-linking occurs by abstraction of H by oxygen,
followed by radical cross-linking43
. However, when stabilized with appropriate
antioxidants, even dimethylsiloxane can survive aging at 300 °C in hot air environments
for significant lengths of time 73
.
Replacement of methyl groups with phenyl groups results in greater thermal
stability12,74
. However, the replacement of all methyl groups with phenyl groups results in
crystallization of the polymer. In the fully substituted polydiphenylsiloxane, the material
actually exhibits liquid crystal transitions at 240°C, a melting point/decomposition at
>500°C, and has such a high degree of crystallinity that the Tg is almost undetectable72
.
Due to the tendency to crystallize, polydiphenylsiloxane is not useable as an elastomer,
even when heated above its Tg of 40°C.
Silylphenylene links (Si-Phenyl-Si) have been investigated as spacer units to
impart greater thermal stability as well as for disruption of molecular symmetry to inhibit
crystallization75-77
.
48
Silylcarborane links have also been investigated for the same reasons78-80
(Figure
1.19). Note in carborane siloxanes the CH bonds in Figure 1.18 are replaced by C –Si
bonds. M-carboranes were the most commonly used carboranes, as this isomer is
produced in highest yield when carboranes are synthesized.
Figure 1.19. Carborane structure from Peters et al79
.
Variants of this material were commercialized by the Olin Corporation in the
1980s78-80
under the trade name “Dexsil”. Several researchers attributed the higher than
expected thermal stability of this material to a resonance effect; the carborane is electron
deficient, and can actually act as an electron-sink, thus increasing the bond dissociation
energy of the silylether links78
.
1.10.1. Phenoxysilanes
Other researchers investigated phenoxysiloxane polymers, where the repeat unit
contained Si-O-phenyl-O links81-83
. These materials were very thermally stable, but cross
linking attempts with vinyl substituents were not successful. The authors attributed this to
49
residual aniline and anilosilane monomers, which may have acted as free radical
scavengers and inhibited radical cross linking of the vinyl groups81
. Up until this writing,
no systematic evaluation of this polymer has been performed with ethynyl cross linking
substituents.
It would be instructive to compare the bond dissociation energies of bonds in this
polymer to determine if the fully aromatic nature of the material produces significant
enhancements in bond dissociation energies due to resonance stabilization, as was seen in
the carborane systems. Unfortunately, phenoxysilane bond dissociation energies were not
readily available in the literature. However, thermal stability of this chemical bond
appears to be exceptionally high, as inferred from other related data. For example, an
evaluation of tetraphenoxysilane as a potential heat transfer fluid showed the material did
not degrade to any appreciable extent at high temperatures, even after heating at 400⁰C
for 100 days in a nitrogen atmosphere.84
. Given this data, it appears that phenoxysilane
polymers, when paired with thermally stable cross-linking functionality, have the
potential to yield an elastomeric system with exceptional thermal stability.
1.11. Analytical tools used for characterization of polymers in this work
In order to communicate the findings of this work, a basic understanding of the
analytical equipment used for data generation is required. In the following paragraphs,
brief descriptions are given of the primary testing procedures used for the body of this
50
work. The goal is not to give an exhaustive description, but rather a high level overview
of the basic principles, equipment, and type of information generated for each type of
test, so the results presented in the following pages can be clearly understood by the
reader.
GPC:
Gel permeation chromatography, also called size exclusion chromatography
(SEC) is a technique that is used to determine the molecular weight distributions of
polymers. It is a chromatographic technique where the polymer is dissolved in a mobile
phase and passed through a column with a stationary phase, for separation of the sample
by molecular weight. THF is the most common solvent used. Stationary phase is usually
a cross-linked polystyrene. The eluted polymer solution passes through a detector, which
is usually a refractive index detector or a UV detector.85
Higher molecular weight
polymer elutes earlier, while low molecular weight fractions elute later, as they spend
more time in the pores of the stationary phase. A graphical representation of the operating
principle of this technique is shown in Figure 1.20.
Figure 1.20. Schematic illustration of a GPC column. Note the larger molecules eluting
first, while the smaller molecules are associated with the pores in the column packing86
.
51
This is a secondary technique, which gives molecular weight distributions based
on hydrodynamic volume of standards, which are often monodisperse polystyrene. Even
though it is a secondary technique, the information gains is still useful in that it can
provide information on molecular weight distributions, and also allow for a quantitative
comparison of different reactions ( i.e. did a change in reaction conditions result in an
increase or decrease in molecular weight?). This technique is used extensively for
polymer analysis, provided the polymer can dissolve in a suitable solvent. This technique
cannot be used for cross-linked polymers, as they will not dissolve.
DSC:
Differential scanning calorimetry is a technique whereby the thermal
characteristics of a sample are measured. This is accomplished by measuring the heat
flow into or out of a sample and comparing that to heat flow into or out of a reference
sample heated under the same conditions. Samples are typically ~10 mg and are
encapsulated in small aluminum pans which may or may not be hermetically sealed. The
reference sample is usually an empty pan. Ideally, the differential signal between the
sample and reference pans will be zero; any differences in heat flow are due only to the
sample. Common transitions are endothermic (melting, evaporation), exothermic (curing
reactions) or step changes (glass transitions). The temperatures of the transitions are
characteristic of the polymers- each polymer has a characteristic Tg range, melting range
(if crystalline), and curing temperature (if cross-linkable). The temperatures of the
transitions are affected by heating rates and atmosphere used for testing, so these details
52
are generally included when DSC data is presented. A typical scan is shown in Figure
1.21.
Tg
Cold Crystallization
Melting
Cure
Decomposition
He
at
Flo
w
Figure 1.21. Representative DSC scan. Y axis=heat flow, X axis= temperature (adapted
from Choudry et al87
).
TGA:
Thermogravimetric analysis is the study of weight changes of a sample when it is
heated, either in an oxidizing atmosphere (typically air) or in an inert atmosphere
(typically nitrogen, argon, or helium). The sample is suspended on a platinum pan, and
placed into a small heating chamber. The platinum plan is suspended from a
microbalance, and the weight of the sample is plotted relative to the temperature of the
chamber. This is shown schematically in Figure 1.22.
53
Figure 1.22. Schematic of thermogravimetric instrument87
.
Temperatures of weigh losses are often characteristic of the gases evolved and/or
thermal changes in the polymer. For polymeric samples, adsorbed solvents will be lost
near their boiling points, encapsulated solvents will be lost when the samples are heated
above Tg, and losses due to decomposition are also measured. 5% weight losses are often
used as a cutoff for the start of thermal decomposition2. 50% weight losses correlate well
with bond dissociation energies of the samples being tested43
. As with DSC, heating rates
and atmosphere can affect the temperature of transitions, so this information is usually
reported when TGA data is presented.
Rheology:
54
Rheology is the study of flow properties of a material. Viscosity is a measure of a
materials resistance to flow. A parallel plate rheometer was used to measure the
viscosities of samples as a function of temperature in an inert atmosphere. A schematic of
a parallel plate rheometer is shown in Figure 1.23. An oscillating torque is applied to a
sample of molten polymer that is held between two plates. Given the sample geometry,
applied torque, and the resulting displacement, as well as the time lag after the force is
initially applied until the displacement occurs, one can calculate numerous physical
parameters for the sample.
Figure 1.23. Schematic illustration of parallel plate rheometer 88
.
Quantities such as apparent viscosity, and storage and loss modulus can be
calculated (storage modulus= elastic response of the polymer, loss modulus=viscous
response of the polymer). Each of these parameters is related to the inherent properties of
the sample. Viscosity at very low shear rates is proportional to the molecular weight of a
55
polymer to the 3.4th
power33
. Storage modulus is a measure of the network structure
present in a sample. Loss modulus is a measure of the viscous nature of the sample.
Changes in molecular weight due to cross linking can be measured by changes in
viscosity, storage modulus, and loss modulus88
. Tracking these variables as time and or
temperature are varied enables the determination of whether or not cross-linking is taking
place, as well as a means to determine the effective cure completion. Above Tg, when the
polymer can flow, the loss modulus is greater than the storage modulus. When a material
cross-links, the storage modulus increases while the loss modulus decreases. When the
storage modulus =loss modulus, this is defined as the gel point, where the sample first
becomes a cross-linked polymer network. An example is shown in Figure 1.24.
Figure 1.24. Representative parallel plate data, showing the change in storage and loss
modulus as a function of time, and the gel point (cross-over point).89
NMR:
56
Nuclear magnetic resonance is used to determine the chemical structure of
organic molecules. Nuclei with odd numbers of protons and neutrons have nuclear spins
that align with strong magnetic field. When these nuclei are exposed to RF radiation the
spins can be made to flip out of alignment with the magnetic field. This is shown
schematically in Figure 1.25. The energy required to flip the spins is related to the
magnetic field strength and the chemical environment of the atoms of interest86
. The
frequencies are scaled to reference samples (chemical shift), which are either
tetramethylsilane (defined as 0 ppm) or solvent signals which have known chemical
shifts, relative to tetramethylsilane. Chemical shifts are plotted as the x-axis in a
spectrum, and usually have 0 on the right side of the scan. The units used are ppm. H
and13
C are the most common atoms analyzed by NMR.
Figure 1.25. Block Diagram of NMR instrument86
.
57
IR:
Infrared absorption spectroscopy is commonly used to identify chemical bonds
that might be present in an unknown. The sample is irradiated with infrared radiation, and
the transmitted radiation is measured at a detector. The wavelengths where significant
absorption occurs give information about the structures of the sample. Spectra are
usually plotted as transmission vs. wavenumber, ν (ν =hc/λ, where h=Planck’s constant,
and λ is the wavelength of the photon.) The presence or absence of bonds in a product
that are different than starting reagents is usually supporting evidence that a reaction had
occurred. For example, if a dried starting reagent had a broad absorption at 3400 cm-1
,
and the dried final product did not, that could be taken as an indication that the OH had
reacted (OH has a strong absorption at 3400 cm-1
).
For the purposes of this work, several bond types are of interest. Ethynyl bonds
absorb at 2100-2200 cm-1
, OH from starting phenolic type materials have absorption
bands at 3400 cm-1
, Si-O-Si bonds absorb at 1100 cm-1
, and Si-O-aromatic bonds absorb
~ 900 cm-1 54, 82
. One requirement for IR absorption is that the molecules must be
asymmetric to absorb infrared radiation by a change in dipole moment of the bond of
interest. If they are symmetric, IR absorption will not occur. This presents a potential
challenge when evaluating ethynyl substituents, as they are symmetrical; absorption
bands are only seen when substituents are sufficiently different to allow absorption, and
usually the transitions are very weak90
.
Raman:
Raman spectroscopy is similar to infrared, in that the information yielded is
chemical in nature, so one can determine chemical bonds present in a sample. The
58
difference is that Raman actually measures inelastic light scattering, not absorption.
Typically a high energy laser is used to irradiate a sample. The sample will scatter the
monomchromatic incident radiation and the incoherent scattered radiation is then
measured at a detector. However, scattering only occurs if the sample has symmetrical
bonds that may be polarized by the incident radiation, so the molecules in the sample can
absorb some of the energy of the incident photon, resulting in inelastic scattering86
.
Raman is thus used primarily for qualitative and quantitative determination of symmetric
chemical bonds. For the reagents used in this work, many of the ethynyl moieties used
were highly symmetric, and thus infrared absorption was very weak. Raman provided
much better qualitative proof of the presence of ethynyl groups in several of the materials
synthesized during the course of this research. Ethynyl bands of interest typically are seen
in the 2,200cm-1
region90
.
1.12. References
(1) Greve, H.; Threadingham, D. In Rubber, 1. Survey; Ullmann's Encyclopedia of
Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA, Darmstadt,
Germany: 2000.
(2) Lewis, R. J., Sr. In Hawley's Condensed Chemical Dictionary; John Wiley & Sons:
Hoboken, NJ, 2009.
(3) Brandrup, J.; Immergut, E. H.; Grulke, E. A.; Abe, A.; Bloch, D. R. Polymer
Handbook (Fourth Edition). John Wiley & Sons: Hoboken, NJ, 2005, 5.
(4) Etymology of rubber.
http://www.etymonline.com/index.php?search=rubber&searchmode=none
(5) Definition of elastomer. http://www.merriam-webster.com/dictionary/elastomer
(6) Kuhn, H.; Medlin, D. In ASM Handbook, Volume 08 - Mechanical Testing and
Evaluation; ASM International: Materials Park, OH , 2000.
59
(7) Hedrick, J. L.; Yang, A. C.; Scott, J. C.; Economy, J. E.; McGrath, J. E., Elastomeric
behaviour of cross-linked poly(aryl ether ketone)s at elevated temperatures. Polymer
1992, 33, 5094-5097.
(8) Threadingham, D.; Obrecht, W.; Lambert, J. P.; Happ, M. In Rubber, 3. Synthetic;
Ullmann's Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co.
KGaA: Darmstadt, Germany: 2000.
(9) Carlson, D. P.; Schmiegel, W. In Fluoropolymers, Organic; Ullmann's Encyclopedia
of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Darmstadt,
Germany: 2000.
(10) Dvornic, P.; Lenz, R. High Temperature Siloxane Elastomers; Huethig & Wepf,
Heidelberg, Germany: 1990, 193-195.
(11) Carraher, C.; Pittman, C. In Inorganic Polymers; Ullmann's Encyclopedia of
Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Darmstadt,
Germany: 2000.
(12) Jones, R. G.; Ando, W.; Chojnowski, J. Silicon-Containing Polymers - The Science
and Technology of Their Synthesis and Applications. Springer - Verlag: Vienna,
Austria: 2000, 186.
(13) Schut, J.; Bolikal, D.; Khan, I. J.; Pesnell, A.; Rege, A.; Rojas, R.; Sheihet, L.;
Murthy, N. S.; Kohn, J.,Glass transition temperature prediction of polymers through
the mass-per-flexible-bond principle. Polymer 2007, (48), 6115-6124.
(14) Schneider, H. A., Polymer class specificity of the glass temperature. Polymer 2005,
46, (7), 2230-2237.
(15) Schneider, H. A., Flexibility and phase transitions of polymers. Journal of Applied
Polymer Science 2003, 88, (16), 1590-1599.
(16) Sperling, L. H. In Introduction to Physical Polymer Science. Fourth Edition; John
Wiley & Sons: Hoboken, NJ 2006.
(17) Wang, S.; Ravindranath, S.; Wang, Y.; Boukany, P., New theoretical considerations
in polymer rheology: Elastic breakdown of chain entanglement network. Journal of
Chemical Physics 2007, 127, (6), 064903.
(18) Hosford, W. In Overview of Tensile Testing; Han, P., Ed.; Tensile Testing; ASM
International: Materials Park, OH, 1993.
(19) Campo, E. A. In Complete Part Design Handbook - For Injection Molding of
Thermoplastics; Hanser Publishers: Munich, Germany, 2006.
60
(20) Mitchell, B. S. Mechanics of Materials, In An Introduction to Materials Engineering
and Science: For Chemical and Materials Engineers, John Wiley & Sons, Inc.,
Hoboken, NJ, 2004.
(21) Hosford, W. In Tensile Testing of Plastics; Turner, S., Ed.; ASM International:
Materials Park, OH , 1993.
(22) Lassettre, M.; Everhart, J. O., Stress-strain relations in ceramic materials. Journal of
the American Ceramic Society 1946, 29, (9), 261-266.
(23) Case, J.; Chilver, L.; Ross, C. T. F. In Strength of Materials and Structures, Fourth
Edition; Elsevier; Burlington, MA, 1999.
(24) Qi, H. J.; Joyce, K.; Boyce, M. C., Durometer hardness and the stress-strain behavior
of elastomeric materials. Rubber Chem. Technol. 2003, 76, (2), 419-435.
(25) ASTM D2240 - 05. Standard Test Method for Rubber Property—Durometer
Hardness.In ASTM Book of Standards Volume 09.01: Rubber, Natural and Synthetic
- General Test Methods; Carbon Black, ASTM International:Conshohocken, PA,
2006, .
(26) Magonov, S. N.; Reneker, D. H., Characterization of polymer surfaces with atomic
force microscopy. Annual Review of Material Science 1997, (27), 175-222.
(27) Adhikari, R.; Godehardt, R.; Huy, T. A.; Michler, G. H. Low temperature tensile
deformation behaviour of styrene/butadiene based thermoplastic elastomer. Kautsch
Gummi Kunstst 2003, 56, (11), 573-577.
(28) de Gennes, P. G., Reptation of a Polymer Chain in the Presence of Fixed Obstacles.
Journal of Chemical Physics 1971, 55, (2), 572-579.
(29) Pethrick, R. A., Molecular motion in polymer systems. Current Opinion in Solid
State and Materials Science 2002, 6, (3), 221-225.
(30) McLeish, T. C. B. Tube theory of entangled polymer dynamics. Advances in Physics
2002, 51, (6), 1379-1527.
(31) Erman, B. Advances in constraint theories of rubber-like elasticity of polymers.
Current Opinion in Solid State and Materials Science. 2010, 14, (2), 35-37.
(32) Sperling, L. H. In Introduction to Physical Polymer Science. Fourth Edition; John
Wiley & Sons: Hoboken, NJ 2006, 383.
(33) Malkin, A. Y. In Rheology Fundamentals; ChemTec Publishing: Toronto, Canada.
1994, 120-121.
61
(34) Bukhina, M. F.; Kurlyand, S. K. In Low-Temperature Behaviour of Elastomers; VSP
Publishing : Amsterdam,Netherlands. 2007, 4-7.
(35) Atkins, P.W.;Physical Chemistry. Third Edition; W.H. Freeman and Company: New
York, NY. 1986, 676.
(36) Vrentas, J. S.; Vrentas, C. M., A new equation relating self-diffusion and mutual
diffusion coefficients in polymer-solvent systems. Macromolecules 1993, 26, (22),
6129-6131.
(37) De Kee, D.; Liu, Q.; Hinestroza, J., Viscoelastic (Non-Fickian) Diffusion. The
Canadian Journal of Chemical Engineering 2005, 83, (6), 913-929.
(38) Finerman, T. M.; Crist, B., Self-diffusion of polydisperse linear polymers. Journal of
Non Crystalline Solids 1991, 131, (2), 620-623.
(39) Lin, D. C.; Douglas, J. F.; Horkay, F., Development of minimal models of the elastic
properties of flexible and stiff polymer networks with permanent and
thermoreversible cross-links Soft Matter 2010, 6, (15), 3548-3561.
(40) Pocket Performance Specs for Thermoplastics;Cannon Communications: Los
Angeles, CA 2003, 628.
(41) Sperling, L. H. In Introduction to Physical Polymer Science. Fourth Edition; John
Wiley & Sons: Hoboken, NJ 2006, 395.
(42) Hertz, D. L. In Mechanics of Elastomers at High Temperatures; Veneruso, A. F.,
Arnold, C. and Simpson, R. S., Eds.; High temperature electronics and
instrumentation seminar proceedings, US Department of Energy Office of Scientific
and Technical Information: Oak Ridge, TN 1980, 227.
(43) Kelen, T. In Polymer Degradation; Van Nostrand Reinhold Company: New York,
NY, 1983,4-6.
(44) Allcock, H. R.; Lampe, F. W. In Contemporary Polymer Chemistry. Second Edition;
Prentice Hall: Englewood Cliffs, NJ 1990 , 236.
(45) Blanksby, S. J. S., Bond Dissociation Energies of Organic Molecules. Accounts of
Chemical Research 2003, 36, (4), 255-263.
(46) Fine, D. H.; Westmore, J. B., Heats of Formation of some alkylthio radicals.
Canadian Journal of Chemistry 1970, (48), 395.
(47) Fortman, G. C.; Captain, B.; Hoff, C. D. Thermodynamic Investigations of the
Staudinger Reaction of Trialkylphosphines with 1-Adamantyl Azide and the
62
Isolation of an Unusual s-cis Phosphazide. Inorganic Chemistry 2009, 48, (5), 1808-
1810.
(48) O'Brien, J. P.; Ferrar, W. T.; Allcock, H. R., Photolysis of Poly(alkoxy-and
aryloxyphosphazenes), [NP(OR)2] n. Macromolecules 1979, 12, (1), 108-113.
(49) Calle, L. M.; Kana'an, A. S., Enthalpies and entropies of sublimation of tetraphenyl
silane and hexaphenyl disilane. The bond dissociation energies of Si-C and Si-Si.
The Journal of Chemical Thermodynamics 1974, 6, (10), 935-942.
(50) Homrighausen, C. L.; Keller, T. M., Synthesis and Characterization of silarylene-
siloxane diacetylene Polymer and its Conversion to Thermosetting Plastic. Polymer
2002, 43, (9), 2619.
(51) Kolel-Veetil, M.; Keller, T.M.,The effects of concentration dilution of cross-linkable
diacetylenes on the plasticity of poly(m-carborane-disiloxane-diacetylene)s. Journal
of Materials Chemistry 2003, 13, (17), 1652-1656.
(52) Kolel-Veetil, M.;Beckham, H,; W Keller, T.M., Dependence of thermal properties
on the copolymer sequence in diacetylene-containing polycarboranylenesiloxanes.
Chemistry of Materials 2004, 16, (16), 3162-3167.
(53) Kolel-Veetil, M.; Keller, T.M., Alternatives to thermal curing in diacetylene-
containing carboranylenesiloxanes. Fire and Polymers IV: Materials and Concepts
for Hazard Prevention 2006, (922), 366-377.
(54) Sastri, S. B.; Keller, T. M.; Jones, K. M.; Armistead, J. P., Studies on Cure
Chemistry of New Acetylenic Resins. Macromolecules 1993, 26, (23), 6171-6174.
(55) Bucca, D.; Keller, T. M., Thermally and oxidatively stable thermosets derived from
preceramic monomers. Journal of Polymer Science Part A-Polymer Chemistry.
1997, 35, (6), 1033-1038.
(56) Connell, J. W.; Smith, J. G.; Criss, J. M., High temperature transfer molding resins:
Laminate properties of PETI-298 and PETT-330. High Performance Polymers
2003, 15, (4), 375-394.
(57) Smith, J. G.; Connell, J. W., Microwave synthesis of phenylethynyl imide
oligomers: Neat resin and composite properties of PETI-298. High Performance
Polymers 2006, 18, (3), 341-354.
(58) Bain, S.; Ozawa, H.; Criss, J. M., Development of a cure/postcure cycle for PETI-
330 laminates fabricated by resin transfer molding. High Performance Polymers
2006, 18, (6), 991-1001.
63
(59) Li, Y.; Murphy, L. A.; Lincoln, J. E.; Morgan, R. J., Phenylethynyl endcapped
fluorinated imide oligomer AFR-PEPA-N: Morphology and processability
characteristics. Macromolecular Materials and Engineering 2007, 292, (1), 78-84.
(60) Chen, J.; Qu, X.; Liu, J.; Yang, H.; Fan, L.; Yang, S., Novel phenylethynyl-
endcapped polyimide oligomers: Synthesis and characterization. Polymer
Engineering and Science 2008, 48, (5), 918-924.
(61) Ghose, S.; Watson, K. A.; Cano, R. J.; Britton, S. M.; Jensen, B. J.; Connell, J. W.;
Herring, H. M.; Lineberry, Q. J., High Temperature VARTM of Phenylethynyl
Terminated Imides. High Performance Polymers 2009, 21, (5), 653-672.
(62) Back, S., Aerospace Organic Matrix Composite Materials Task Order 0005:
Complex Architecture and Analytical Methods −Synthesis and Cure Mechanism
Characterization of Phenylethynyl-Terminated Imide Oligomers. University of
Dayton Research Institute Technical Report UDR-TR-2009-132 2009, 26.
(63) Roberts, C. C.; Apple, T. M.; Wnek, G. E., Curing chemistry of phenylethynyl-
terminated imide oligomers: Synthesis of C-13-labeled oligomers and solid-state
NMR studies. Journal of Polymer Science Part A-Polymer Chemistry 2000, 38,
(19), 3486-3497.
(64) Fang, X. M.; Rogers, D. F.; Scola, D. A.; Stevens, M. P., A study of the thermal cure
of a phenylethynyl-terminated imide model compound and a phenylethynyl-
terminated imide oligomer (PETI-5). Journal of Polymer Science Part A-Polymer
Chemistry 1998, 36, (3), 461-470.
(65) Mecham, S., Synthesis and characterization of phenylethynyl-terminated
poly(arylene ether sulfone)s as thermosetting structural adhesives and composite
matrices. PhD Dissertation,Virginia Polytechnic Institute and State University,
Blacksburg, VA 1997.
(66) Hergenrother, P. M., The Use, Design, Synthesis, and Properties of High
Performance/High Temperature Polymers: An Overview. High Performance
Polymers 2003, 15, (1), 3-45.
(67) Heller, M., Thermal stability and sealing performance of perfluoroelastomer seals as
a function of cross-linking chemistry; Journal of Vacuum Science & Technology. A,
Vacuum, Surfaces, and Films 1999, 17, (4), 2119-2124.
(68) Wilkie, C., TGA/FTIR: an extremely useful technique for studying polymer
degradation. Polymer Degradation and Stability 1999, 66 (3), 301-306.
(69) Giannetti, E. Thermal stability and bond dissociation energy of fluorinated
polymers: A critical evaluation. Journal of Fluorine Chemistry 2005, 126, (4), 623-
630.
64
(70) Ameduri, B.; Boutevin, B., Update on fluoroelastomers: from perfluoroelastomers to
fluorosilicones and fluorophosphazenes. Journal of Fluorine Chemistry 2005, 126,
(2), 221-229.
(71) ASTM D1418-05. Standard Practice for Rubber and Rubber Latices-Nomenclature.
In ASTM Book of Standards Volume 09.01: Rubber, Natural and Synthetic -
General Test Methods; Carbon Black; ASTM International: Conshohocken, PA,
2006.
(72) Meier, D.; Lee, M., Synthesis and properties of diarylsiloxane and
(aryl/methyl)siloxane polymers: 1. Thermal properties. Polymer 1993, 34, (23),
4882-4892.
(73) Moretto, H.; - Schulze, M.; Wagner, G., In Silicones; Ullmann's Encyclopedia of
Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA:Darmstadt, Germany
2000, 24.
(74) Deshpande,G.; Rezac,M. The effect of phenyl content on the degradation of
poly(dimethyl diphenyl) siloxane copolymers . Polymer Degradation and Stability
2001, 74, (2), 363-370.
(75) Dvornic, P. R.; Lenz, R. W., Exactly alternating silarylene-siloxane polymers. 9.
Relationships between polymer structure and glass transition temperature.
Macromolecules 1992, 25, (14), 3769-3778.
(76) Lauter, U.; Kantor, S. W.; Schmidt-Rohr, K.; MacKnight, W. J., Vinyl-Substituted
Silphenylene Siloxane Copolymers: Novel High-Temperature Elastomers.
Macromolecules 1999, 32, (10), 3426-3431.
(77) Hani Rahim; Lenz Robert W. In Synthesis and Properties of Silphenylene-Siloxane
Polymers; Silicon-Based Polymer Science; Ziegler,T.; Fearon,F. Ed; American
Chemical Society: Washington, DC 1989,Volume 224, 741-752.
(78) Peters, E. N., The development of carborane-siloxane polymers. Industrial &
Engineering Chemistry Product Research and Development 1984, 23, (1), 28-32.
(79) Peters,E.N., In Structure and Property Modification of m-Carborane Siloxanes;
Modification of Polymers, Carraher, C.; Tsuda, M., Ed.; American Chemical
Society: Washington, DC 1980, Volume 121, 449-456.
(80) Stewart, D. D.; Peters, E. N.; Beard, C. D.; Dunks, G. B.; Hedaya, E.; Kwiatkowski,
G. T.; Moffitt, R. B.; Bohan, J. J., D2-m-Carborane Siloxanes. 7. Synthesis and
Properties of Ultra-High Molecular Weight Polymer. Macromolecules 1979, 12, (3),
373-377.
65
(81) Steffen, K. ,Poly- dioxyarylen( dioxycycloalky1en)-diphenylsilane II.
Molekulargewichte, Erweichungstemperatuen, Thermostabilitaten. Die Angewandte
Makromolekulare Chemie 1972, (24), 21-33.
(82) Dunnavant, W. R.; Markle, R. A.; ; Sinclair, R. G.; Stickney, P. B. ; Curry, J. E.;
Byrd, J. D., p,p Biphenol Dianilosilane Condensation Polymers. Macromolecules
1968, (1), 249-254.
(83) Curry, J. Method of Producing Alternating Ether Siloxane Polymers. US Patent
3346515, 1967.
(84) Kaufman, H. C., Properties of several arylated silicon compounds. Journal of
Chemical and Engineering Data 1962, 7, (2), 556-8.
(85) Allcock, H. R.; Lampe, F. W. In Contemporary Polymer Chemistry; Prentice Hall:
Englewood Cliffs, NJ, 1990, 396-398.
(86) Cheremisinoff, N. P. Polymer Characterization - Laboratory Techniques and
Analysis. William Andrew Publishing/Noyes: Park Ridge, NJ 1996, 60-66.
(87) Choudhury, N. R.; De, P. P.; Dutta, N. K. In Thermal Analysis of Rubbers and
Rubbery Materials; Smithers Rapra Technology: Billingham, UK 2010, 18-23.
(88) Meynard, K. In Dynamic Testing; Dynamic Mechanical Analysis; CRC Press:
Washington, DC, 1999 .
(89) Mukherjee, I.; Drake, K.; Berke-Schlessel, D.; Lelkes, P. I.; Yeh, J.; Wei, Y., Novel
Thermally Cross-Linkable Poly[(arylenedioxy)(diorganylsilylene)]s Based on
Curcumin: Synthesis and Characterization. Macromolecules 2010, 43, (7), 3277-
3285.
(90) Speight, J. G. Lange's Handbook of Chemistry,16th Edition, McGraw Hill:
Columbus, OH 2005, 3.9-3.11,3.37-3.42.
66
Chapter 2: Synthesis of polymers for subsequent endcapping reactions
2.1. Introduction
In order to synthesize ethynl terminated aryloxysilane polymers via endcapping,
one must first synthesize the polymer to obtain a high enough molecular weight to
facilitate separation and processing. Another constraint is that the molecular weight must
be low enough to allow a significant percentage of labile end groups for reaction in the
subsequent step of endcapping with ethynl containing reagents. Stoichiometric excess of
dichlorosilane monomer was used to yield a polymer with the required molecular weight
and reactive Si-Cl groups on the ends for subsequent functionalization in a later step.
Target molecular weights were selected to yield polymers of approximately 40 repeat
units. 40 repeat units with 2 end groups would result in cross-link densities of 5 molar %
for bifunctional cross linking endgroup moieties. For reference, typical cross-link
densities of high performance elastomeric materials range from 3-7%1. For the
diphenylsilane-hydroquinone polymer with a repeat unit mass of 290, and the
diphenylsilane-biphenol polymer with a repeat unit mass of 366, target number average
molecular weights of 11,000 and 15,000, respectively, were selected.
Several synthetic methods for the preparation of aryloxysilane polymers have
been documented by previous researchers2-4
. All reactions fall under the general category
of nucleophilic substitution at the silicon atom. General mechanisms can be classified
into catalyzed and uncatalyzed nucleophilic substitution. Uncatalyzed reactions involve
nucleophilic attack of the electropositive silicon atom by nucleophiles5, 6
(Scheme 2.1).
67
R
R
Si ClCl
δ+ δ−+ Nu:
R
R
SiNuCl + Cl-
Nu:= OH OH
OH OH
::
::
Scheme 2.1. Uncatalyzed nucleophilic substitution at silicon atom.
Catalyzed nucleophilic substitutions involve the use of Lewis bases to activate the
silicon atom and allow the hydride ion to act as a leaving group. The hydride ion then
coupling with the hydrogen of an alcohol 7-9
. This reaction is shown in Scheme 2.2.
68
R
R
Si HH
δ+ δ−+
Nu:= OH OH
OH OH
::
::
CH3O-
K+
::
R
R
Si HH
OCH3
δ−δ+
O
H
:
:
R
R
Si OH
OH
+
H2
(catalyst)
CH3O-
K+
::
Scheme 2.2. Mechanism of catalyzed nucleophilic substitution of hydridosilane.
Of the various synthetic reactions available to make this polymer, melt
condensation reactions between aminosilanes and diols, substitution at a silicon atom
with amine leaving groups, typically yields the highest molecular weights (Scheme 2.3)
10.
69
+
n
R OHOHO R O Si
R'
SiN NH
R'
H
+ NH2
Melt Polymerization
6 hours
300 −325 °C
N2
Where R=
or
R' =
Scheme 2.3. General reaction scheme, aminosilane – diol condensation polymerization,
in the melt state (no solvents used).
However, the aminosilane-diol reaction has several disadvantages when
considered for the preparation of polymers that contain ethynyl moieties. Polymerization
temperatures of 300 to 325 °C are required, and residual aniline and unreacted anilosilane
monomer are difficult to be completely removed from the system. In some cases 1-2% of
unreacted monomers are unaccounted for after the reactions are completed11
. These
reaction conditions pose two significant issues:
1) Polymerization temperatures of 300 to 325 °C fall in the range of cure
temperatures of many examples of ethynyl-containing polymers12-14
. Due to this
fact, undesired side reactions of the ethynyl groups are more likely to take place
during synthesis using this route than with other reactions that are performed
under milder conditions.
70
2) Dunnavant et al. identified residual aniline and anilosilane monomers as
impurities that likely interfered with cross-linking reactions of vinyl functional
groups, presumably through their free-radical scavenger properties11
. Free radical
scavenging by residual impurities would be a likely issue with ethynyl containing
polymers as well, since they often initially cross-link through radical addition
processes across the triple bond 15
.
Given these two concerns, for purposes of this study and the production of cross-
linkable ethynyl aryloxysilanes, it was decided not to pursue an aminosilane synthetic
route.
Two other general synthetic routes to generate aryloxysilanes are catalyzed
condensation of hydridosilanes with diols (Scheme 2.4) and condensation of
dichlorosilanes with diols (Scheme 2.5).
+
n
R OHOHO R O Si
SiH H + H2
Melt or Solvent
Polymerization
4-6 hours
180 −200 °C for melt
<100°C for solvent
N2
R =
or
Scheme 2.4. General reaction scheme, catalyzed condensation of dihydridiosilane with
diols.
71
+
n
R OHOHO R O Si
SiCl Cl + H Clsolvent polymerization
N2
R =
Scheme 2.5. General reaction scheme, catalyzed condensation of dihydridiosilane with
diols.
Both synthetic routes were explored in detail during the course of the studies in
this thesis. Temperatures, solvents, monomer ratios, and reaction times were
systematically varied to determine the optimal synthetic conditions to fulfill two separate
requirements:
1) Generation of a polymer with a molecular weight in the desired range, with the
proper end group chemistry to allow for facile addition of cross linking moieties
in a subsequent step, via addition directly into the polymerization reactor
(endcapping reactions are the focus of Chapter 3 in this dissertation).
2) Polymerization of a silane with a diol, where one of the monomers would contain
ethynylinic moieties. The ethynyl groups could be present as either pendant
72
groups or as part of the polymer backbone. These two categories, pendant and
main chain, are represented in Figure 2.1. Note that the focus of Chapter 3 is
ethynyl endcapping of aryloxysilanes, and the focus of Chapter 4 is main chain
ethynyl aryloxysilanes.
Ethynyl End Caps Main Chain Ethynyl
Figure 2.1. Pendant and main chain ethynyl groups.
Schemes 2.6 through 2.12 outline the specific reactions performed in this study.
Details of each synthesis are described in the following pages, along with
characterization data for polymers produced from each reaction type. A general summary
follows at the end of the chapter, with conclusions on the optimal reactions for further
study during the endcapping and ethynylic diol synthetic work.
73
2.2. Experimental section
2.2.1. Reagents
Diphenylsilane (DPS), 99%, formula weight 184.31 Gelest Catalog number
SID4559.0; dichlorodiphenyl silane (99%), formula weight 253.2, Gelest Catalog
number SID4510.1; 4,4’–biphenol, 99%, formula weight 186.21; hydroquinone, 99%,
formula weight 110.1; potassium methylate (95%), formula weight=70.13; Toluene, ACS
grade, anhydrous were all used as purchased, unless noted.
2.2.2. Synthesis
2.2.2.1. Diphenylsilane-hydroquinone copolymer melt polymerization (Scheme 2.6)
OHOH + SiH H
n
O O Si
KOCH3 catalyst
∆ ( 175-200 °C)
Scheme 2.6. Diphenylsilane condensation melt polymerization, with hydroquinone
comonomer.
In a typical procedure, 6.8 grams (60 mmol) of hydroquinone solids were added
to a three neck round bottom flask. To this flask, 11 grams (60 mmol) of liquid
74
diphenylsilane were added. A catalytic amount of potassium methylate (0.01g) was
added to the mixture. The reaction vessel was placed in a heat transfer fluid bath
(dibutylphthalate), and purged with nitrogen, at a flow rate of ~ 1 bubble per second in a
bubbler attachment filled with silicone oil. A reflux condenser was placed in line with the
bubbler, in order to capture any evolved diphenylsilane (boiling point 95 to 97 °C at
13mm Hg16
). A thermometer was placed in one neck of the flask, while a collection
vessel was connected to the third neck, in order to capture any material that might
overflow during the course of the reaction. The collection vessel had a gas outlet port for
venting of generated hydrogen gas to a hood vent. In addition a 2nd
thermometer was
placed in the oil bath to monitor temperature of the heat transfer fluid. A magnetic stir bar
was also added to the reactor prior to final assembly. The heat transfer bath was then
slowly heated to reaction temperatures, over approximately 40 minutes to one hour.
In the initial stage of the reaction, the mixture consisted of two phases, a solid
hydroquinone phase suspended in a liquid diphenylsilane phase. Initial bubbling was seen
in the liquid phase at approximately 150 °C. As the temperature approached the melting
point of the hydroquinone (171°C), vigorous bubbling was seen. The temperature then
spiked to >200 °C and a violent reaction took place. Within 5-10 minutes the reaction
appeared to be substantially complete as judged by the absence of significant bubbling,
and yielded a clear viscous molten polymer. As per the procedure outlined by Steffen2,
the reaction was allowed to continue for another two hours at 170 °C under nitrogen. The
reaction then proceeded for an additional two hours while pulling a vacuum to remove
any trapped volatiles and unreacted diphenylsilane from the melt. The viscous clear-gray
75
semi-solid polymer was poured from the reactor into sample vials for further analysis.
Yield was quantitative.
2.2.2.2. Diphenylsilane-hydroquinone solution polymerization (Scheme 2.7)
OH OH + SiH H O O Si
n
+ H2
Toluene
KOCH3 catalyst
∆ ( 120-130 °C)
Scheme 2.7. Diphenylsilane solution based condensation polymerization with hydroquinone as
the comonomer.
Due to the vigorous nature of the melt polymerization and difficulty in
maintaining controlled conditions, solution polymerization was also attempted. The
procedure outlined by Steffen9 was followed, except for the substitution of toluene for
xylene as a solvent (toluene was successfully used as a solvent for aryloxysilanes
produced via condensation of dichlorosilanes with hydroquinone and biphenol). In a
typical procedure, 4.4 grams (40 mmol) of hydroquinone solids were added to a 50ml
two- neck round bottom flask. To this flask, 7.6 grams (40 mmol) of liquid
diphenylsilane was added, along with 30 ml of anhydrous toluene. A catalytic amount of
potassium methylate (5 drops of 5% potassium methylate in methanol solution, ca. 0.2 ml
total) was added to the mixture. The reaction vessel was placed in a heat transfer fluid
76
bath (dibutylphthalate), and purged with nitrogen, at an initial flow rate of ~ 3 bubbles
per second in a bubbler attachment filled with silicone oil. A reflux condenser was placed
in line with the bubbler. A flexible tube was connected to the bubbler outlet, and bubbled
through a water trap in order to observe the rate of bubbling to watch for the start of
higher levels of gas flow, which would indicate hydrogen gas evolution, and thus indicate
that the condensation reaction was occurring. A thermometer was placed in one neck of
the flask. In addition a 2nd
thermometer was placed in the oil bath to monitor temperature
of the heat transfer fluid. A magnetic stir bar was also added to the reactor prior to final
assembly. The heat transfer bath was then slowly heated to reaction temperatures of 110
ºC (oil temperature=130 ºC), the boiling point of toluene, over approximately 30 minutes.
The reaction was allowed to proceed for an additional 5 hours.
For this system, hydroquinone was found to be insoluble in toluene; however, in
another system studied (dichlorosilane-diol condensation) in toluene solvent, the
polymerization still proceeds to build high molecular weight, as oligomers are soluble. It
was concluded that solubility in toluene may not be a limiting factor, so the condensation
reaction of hydridosilane in toluene was attempted. This reaction was repeated several
times, unfortunately in all cases, no apparent build in viscosity was noted, and no
polymer was obtained on precipitation into methanol (reaction products, if any, were
fully soluble in methanol, or resulted in an oily residue which could not be separated
from the solution).
2.2.2.3. Diphenylsilane-biphenol copolymer melt polymerization (Scheme 2.8)
77
OH OH + SiH H O O Si
n
KOCH3 catalyst
∆ ( 175 - 200 °C)
Scheme 2.8. Diphenylsilane condensation melt polymerization, with biphenol
comonomer.
1.86 grams (10 mmol) of 4,4-biphenol solids were added to a three neck round
bottom flask. To this flask, 1.84 grams (10 mmol) of liquid diphenylsilane was added. A
catalytic amount of potassium methylate solution was added to the mixture (5 drops of
5% potassium methylate in methanol solution, ca 0.2 ml total). The reaction vessel was
placed in a heat transfer fluid bath (dibutylphthalate), and purged with nitrogen, at a flow
rate of ~ 1 bubble per second in a bubbler attachment filled with silicone oil. A reflux
condenser was placed in line with the bubbler, in order to capture any evolved
diphenylsilane (boiling point 95 to 97 °C at 13mm Hg), a thermometer was placed in one
neck of the flask, while a collection vessel was connected to the third neck, in order to
capture any material that might overflow during the course of the reaction. The
collection vessel has a gas outlet port for venting of generated hydrogen gas to a hood
vent. In addition a 2nd
thermometer was placed in the oil bath to monitor temperature of
the heat transfer fluid. A magnetic stir bar was also added to the reactor prior to final
assembly. The heat transfer bath was then slowly heated to reaction temperatures, over
approximately 40 minutes to one hour.
78
In the initial stage of the reaction, the mixture consisted of two phases, a solid
white biphenol phase dispersed via stirring throughout the liquid diphenylsilane
supernatant phase. Initially, bubbling was first observed in the liquid phase at
approximately 170 °C. As the reaction temperature rose to 210 to 220 ºC, the solution
became clearer and bubbling increased. Biphenol has a melting point of 245 ºC, so unlike
the hydroquinone silane condensation, the reaction did not proceed violently at the
melting point of the diol, since it did not melt under reaction conditions. After
approximately 30 minutes at 220 ºC, the solution viscosity increased to the point at which
the magnetic stir bar stopped spinning. The solid could not be poured from the reaction
flask, so after cooling, it was dissolved in THF and precipitated in methanol. The
precipitate was filtered to yield a white fluffy precipitate. The precipitate was dried in a
vacuum oven at 70°C for 3 days, to yield 1.42 g of a gummy white material (39% of
theoretical yield). The precipitate was almost fully soluble in acetone, which was an
indication that molecular weight was relatively low.
2.2.2.4. Diphenylsilane-hydroquinone copolymer polymerization in CH2Cl2 (Scheme
2.9).
OH OH + SiCl Cl
n
CH2Cl
2
pyridine
0 −5 °C
N2
O O SiH
Scheme 2.9. Hydroquinone-dichlorodiphenylsilane condensation.
79
In a typical procedure, 2.75 grams (25 mmol) of solid hydroquinone were added
to a 100 ml three neck round bottom flask. To this flask, 25 ml /33grams of CH2Cl2 and
4.2 grams (53mmol) of pyridine were added, along with a magnetic stir bar. A reflux
condenser was added in the central neck of the flask, capped with a nitrogen bubbler for
inert atmosphere purge during the course of the reaction.. A thermometer was placed in
the 2nd
neck, positioned so the tip of the thermometer was immersed in the reactor liquor.
A vented addition funnel was charged with 8-10 ml of CH2Cl2 and 6.33 g (26.5 mmol) of
dichlorodiphenylsilane, and mounted in the third neck of the flask. An ice bath was used
to control the temperature during addition of the dichlorosilane. The contents of the
addition funnel were added dropwise over the course of approximately one hour. The ice
bath was removed, and the reaction was allowed to proceed for an additional three hours.
At the conclusion of the reaction, the solution was very viscous. An additional 20 ml of
CH2Cl2 was added to the reactor to reduce viscosity and allow for easy transfer of the
material to a precipitating bath. The reaction was filtered to remove precipitated
pyridine:HCl salt condensation byproducts, and then precipitated in methanol ( non-
solvent for the polymer). Polymer was dried overnight under vacuum at ambient
temperature. 4.64 grams of white polymer was obtained ( 64% yield). Elevated
temperature drying was not used, since the Tg of this polymer is ~ 60 °C. Heating near Tg
for an extended period of time would result in formation of a gummy solid.
2.2.2.5. Diphenyldichlorosilane-biphenol copolymer polymerization in CH2Cl2
(Scheme 2.10)
80
OH OH + SiCl Cl O O Si
n
CH2Cl
2
pyridine
0 −5 °C
N2
Scheme 2.10. Biphenol-dichlorodiphenylsilane condensation.
In a typical procedure, 4.65 grams (25 mmol) of solid biphenol were added to a
100 ml three neck round bottom flask. To this flask, 25 ml ca. 33grams of CH2Cl2 and
4.15 grams (53 mmol) of pyridine were added, along with a magnetic stir bar. A reflux
condenser was added in the central neck of the flask, capped with a nitrogen bubbler for
inert atmosphere purge during the course of the reaction. A thermometer was placed in
the 2nd
neck, positioned so the tip of the thermometer was immersed in the reactor liquor.
A vented addition funnel was charged with 8-10 ml of CH2Cl2 and 6.33g (26.5mmol) of
dichlorodiphenylsilane, and mounted in the third neck of the flask (Note
dichlorodiphenylsilane was added in 5% excess to yield predominantly chlorosilane end
groups). The reactor was immersed in an ice bath to control the temperature in the 0 to 5
ºC range during addition of the dichlorosilane, as per the procedure outlined by Tamai et
al17
. The contents of the addition funnel were added dropwise over the course of
approximately one hour. After addition was complete, the ice bath was removed, and the
81
reaction was allowed to proceed for an additional three hours at ambient temperature (20
to 23 ºC). As with the hydroquinone system, at the conclusion of the reaction the solution
was highly viscous. Therefore additional CH2Cl2 was added to the reactor to reduce
viscosity and allow easy transfer of the material to a vacuum filter apparatus, as well as to
reduce viscosity for more efficient filtration. The reaction was vacuum filtered through a
medium porosity fritted glass filter to remove precipitated the pyridine:HCl salt
condensation byproduct. The filtrate containing the soluble polymer was then precipitated
in methanol and dried overnight in a vacuum oven at 70 °C. 6.9g of white powdery
polymer was obtained (75% yield).
2.2.2.6. Diphenylsilane-biphenol copolymer polymerization in toluene (Scheme 2.11)
OH OH + SiCl Cl O O Si
n
toluene
triethylamine
65 °C
N2
Scheme 2.11. Biphenol-dichlorodiphenylsilane condensation in toluene.
A representative reaction for this synthetic method is described in the following
procedure: In a sealed 100ml three neck round bottom flask equipped with a magnetic
stir bar, a reflux condenser, and a nitrogen purge, 3.72 (0.020 moles) grams of biphenol
82
and 4.25 grams (0.042 moles) of triethylamine was added to 30 ml of toluene. A
thermometer was used to monitor reaction temperature by inserting into the reaction
liquor through a stopper with a rubber adapter. The solution was heated to 80 °C on a hot
plate with a magnetic stirrer. To this solution, 5.3 grams (0.021moles) of
dichlorodiphenylsilane was added, along with 30 ml of toluene via an addition funnel
with an inert gas pressure equalizer. The DCDPS was added dropwise over
approximately one hour. After addition, the solution was heated under reflux overnight
(13-17 hours). Temperature was maintained at 65-70 °C for the remainder of the reaction.
Biphenol was insoluble in toluene, so at the start of the reaction, the solution consisted of
a suspension of biphenol solids. After addition of a few ml of the dichlorodiphenylsilane
solution, a large amount of globular tan precipitate formed. As the remaining
dichlorodiphenylsilane was added, the tan precipitate gradually broke up and dissolved,
which left only fine particles of TEA: HCl salts suspended in a viscous solution.
Initially, filtration was attempted to remove the amine salts. However, due to the
high viscosity of the solution even when diluted with additional toluene, filtration was
very inefficient, took several hours to complete, and resulted in significant loss of
polymer (verified by analysis of the filter cake, see Table 2.1 for details). Therefore,
precipitation in a non-solvent which could dissolve the amine salts was pursued as an
alternative purification route. The polymer solution was transferred dropwise into
methanol, in a ratio of approximately 10 to 20 parts methanol to 1 part of polymer
solution. The precipitated polymer was separated by filtration, redissolved in THF at a
10% concentration, filtered with a standard gravity filtration to remove any remaining
83
TEA:HCl salts, and reprecipitated in methanol. Samples were typically dried in a
vacuum oven at 70 °C overnight. Yields were usually in the 60 to 70% range.
2.2.2.7. Dichlorodiphenylsilane-hydroquinone copolymer polymerization in mixed
THF/toluene system (Scheme 2.12)
OH OH + SiCl Cl
n
O O Si
1: 7 THF: toluene
triethylamine
(THF removed
after 3 hrs)
0 −5 °C
N2
Scheme 2.12. Hydroquinone-dichlorodiphenylsilane condensation in THF/toluene mixed
solvent system.
Initial experiments with the toluene system yielded polymers with somewhat low
to medium molecular weights, but the fact that the diols were insoluble in the reaction
solvent was a concern. Insolubility of phenoxy compounds in non-polar solvents can
sometimes be a limiting factor in nucleophilic substitution reactions; this is why phase
transfer catalysts are often used to solubilize insoluble phenoxy compounds 18
.
In order to generate high molecular weight polymer via condensation reactions, 1:
1 stoichiometry must be maintained. However, if one of the monomers is not soluble, the
result is a two phase system, where one monomer and the growing polymer chains are
soluble, but the second monomer is insoluble. This is a situation that often results in the
84
formation of cyclic oligomers, and not high molecular weight polymers 19
. Essentially,
when one monomer is soluble and the other is insoluble, the result is equivalent to a high
dilution synthesis; this situation is actually termed “pseudo-high dilution synthesis”, and
is a common technique used to prepare cyclic oligomers. A growing polymer chain in
solution is much more likely to react with itself (end- to end reaction) rather than with the
non-soluble 2nd
oligomer which is in a different phase in the reactor.
In order to investigate if solubility was truly a limiting factor for this system in
terms of molecular weight attainment, a mixed solvent system was developed and
evaluated. Hydroquinone was insoluble in toluene, but soluble in THF. It was found
through solubility experiments that a ratio of 7 ml of THF to 20 ml of toluene would
dissolve 25 mmol of hydroquinone, which was the typical reaction scale of this work.
Once the optimum solvent ratio was determined, reactions were performed in this system.
Two mixed solvent system processes were explored.
1) An evaluation was performed on reactions in the mixed THF: toluene system
where the mixed system was used for the entire reaction.
2) An evaluation was performed where the mixed system was used for only 2-3
hrs (comparable time for monomer disappearance in the toluene system).
After 3 hrs, the THF was removed via distillation. The reaction was then run
for the remaining reaction time at a higher temperature in an essentially pure
toluene system.
In both cases the reaction was initially performed as follows: In a 3 necked 100ml
round bottom flask with a magnetic stirrer and a nitrogen purge, 2.2 grams (20 mmol) of
85
hydroquinone and 4.25 grams (42mmol) of triethylamine were added to 20 ml of toluene
and 7ml of THF). A reflux condenser capped with nitrogen bubbler was fitted in the
central neck of the reactor. An addition funnel with a pressure equalizer was charged with
5.32g (21 mmol) of dichlorodiphenylsilane, 10 ml of toluene and 4ml of THF, and
mounted in a side neck of the reactor. A thermometer was inserted in the 3rd
neck of the
flask, and positioned so the tip was fully immersed in the reaction liquor.
The sample was heated to 60 ºC, and then the dichlorodiphenylsilane was added
dropwise over the course of an hour. The reaction was allowed to proceed for an
additional 2 hours, then the reflux condenser was moved to a side arm, so the top was
slightly below the level of the reactor, to allow for distillation and removal of the THF
from the reactor solution. Once all THF was collected, the condenser was replaced in the
central arm of the reaction flask and the reaction was allowed to continue overnight
(approximately 14 hrs). After stopping of the reaction and cooling, the salts were either
filtered from the reactor liquor and the filtrate precipitated in a non-solvent, or the entire
contents of the reactor were precipitated. Molecular weight data is presented in Table 2.4
to show the effects of different precipitating methods and the effects of THF and the
related initial dissolution of the diol on molecular weights.
2.2.2.8. Dichlorodiphenylsilane-biphenol polymerization in THF (Scheme 2.13).
86
OH OH + SiCl Cl O O SiH
n
THF
triethylamine
60 °C
N2
Scheme 2.13. Biphenol-dichlorodiphenylsilane condensation in THF.
A representative reaction for this synthetic method is described in the following
procedure: In a sealed 100ml three neck round bottom flask equipped with a magnetic stir
bar, a reflux condenser, and a nitrogen purge, 1.86 (0.010 moles) grams of biphenol and
2.13 grams (0.021 moles) of triethylamine was added to 30 ml of THF (biphenol was
fully soluble in the THF). A thermometer was used to monitor reaction temperature by
inserting into the reaction liquor through a stopper with a rubber adapter. The solution
was heated to 60 °C on a hot plate with a magnetic stirrer. To this solution, 30 ml of THF
and 2.65 grams (0.0105 moles) of dichlorodiphenylsilane was added via an addition
funnel with an inert gas pressure equalizer. The DCDPS was added dropwise over
approximately one hour. After addition, the solution was heated under reflux for 2-3 hrs.
Temperature was maintained at 65-70 °C for the remainder of the reaction.
Both anhydrous THF and freshly distilled THF were used in separate reactions in
order to evaluate the need for distillation immediately prior to polymerization. THF was
distilled by placing 100 ml in a distillation column with ca. 1 gram of LiAlH4, and heating
to boiling (~66 °C). The first 20 ml of distilled liquid were discarded; the remaining 80ml
of solvent was used for the synthesis. Once the polymerizations were completed, samples
87
were precipitated in methanol, filtered, and reprecipitated in methanol a 2nd
time, then
dried overnight in a vacuum oven at 70 °C. For the non-distilled THF sample, 3.1 grams
were obtained after drying (84% yield). The distilled THF sample yield was not
calculated, as it was used for an endcapping experiment in Chapter 3. Molecular weight
only was measured on this sample.
2.3. Characterization
Molecular weight
Molecular weights were determined using a Waters 590 LC system (Waters
Corp., Milford, MA) equipped with a Phenogel 1µ 103A GPC column (Phenomenex,
Torrance, CA) and a Waters 410 refractive index detector. Elution rate was maintained at
1 mL/min. THF was used as the mobile phase. Polystyrene standards were used for the
molecular weight calibrations.
Spectroscopy
Infrared spectra were collected on a Perkin Elmer (Perkin-Elmer Co., Norwalk,
CT) Spectrum One spectrometer. Spectra of solid samples were collected using an
attenuated total reflectance (ATR) accessory. Spectra of polymers in solution were
collected via transmission IR by dissolving small amounts of sample in THF and
mounting between NaCl plates. ATR correction was performed on all ATR spectra.
Thermal analysis
Differential scanning calorimetry thermograms were collected using a TA
Instruments Q100 DSC (TA Instruments, New Castle, DE). Samples were analyzed in
hermetic pans under a nitrogen atmosphere. Pinhole vents were introduced in the pans to
88
allow any volatiles generated during heating to escape, and also to remove residual
oxygen which might otherwise be trapped in the pan and participate in undesired
oxidation reactions. Heating rates of 20 °C per minute and cooling rates of 10 °C per
minute were used. Thermogravimetric analyses were performed on a TA Instruments
Q50 TGA. Samples were loaded onto platinum pans and temperature was increased at a
rate of 10 ºC per minute, except where otherwise noted. Rheometric analysis was
performed on a TA Instruments AR2000 Rheometer with 8mm parallel plates.
Continuous oscillation testing was performed using a 1 % strain and a frequency of 1
Hertz. Testing was performed at 300 °C under a nitrogen purge. Approximate shear rate
for these tests was 1x10-5
sec-1
.
2.4. Results and discussion
2.4.1. Solubility in methanol, acetone, and its relationship with molecular weight of
aryloxysilanes
Through our studies, we have found that the depending on the solvent and the
polymer’s molecular weight, solubility of aryloxysilanes can vary from completely
soluble to insoluble. In order to quantify these observations and potentiality utilize
solvent extractions as a method of fractionation, an exploration of solubility in different
solvents was performed. For example, one experiment was performed where a biphenol-
silane polymer of a known molecular weight was dissolved in THF, divided into two
equal portions, and precipitated in 1) methanol and 2) acetone. The precipitated polymers
were then filtered and dried in a vacuum oven at 80°C, a temperature above the boiling
point of THF, acetone, and methanol (66 °C, 55 °C, and 65 °C, respectively). GPC
89
analysis of the insoluble fractions showed significant differences in molecular weight
distributions (Figure 2.2, Table 2).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000
No
rmaliz
ed
Inte
nsity
Molecular Weight
Biphenoxydiphenylsilane Molecular Weight distributions of insoluble fractions
acetone insoluble f raction
methanol insoluble f raction
Figure 2.2. Overlay of molecular weight distributions of insoluble polymer fractions.
Table 2.1. Calculated weighted average molecular weight (Mw), number average
molecular weight (Mn,) and PolyDispersity Index (Mw/Mn, PDI) .
Methanol Insoluble
Fraction
Acetone Insoluble
Fraction
Mw 26,900 65,000
Mn 12,100 50,000
PDI 2.2 1.3
90
Based on this data, it was determined that solubility in acetone may be used as a
qualitative measure of molecular weight. If a polymer was soluble in acetone, polymer
molecular weight was likely lower than 12,000. Therefore, when results of some of the
experiments are reported in the following pages, solubility was used as an initial
screening test for a reaction. If polymers were found to be soluble in acetone it was
concluded the polymers were relatively low molecular weights. If the polymers were
soluble in acetone, it was suggested that the material was likely oligomeric, and no
further workups or analysis were performed.
2.4.2. Diphenylsilane-hydroquinone polymer
2.4.2.1. Spectroscopy
FTIR comparative analysis of the starting reagents(figure 2.4 and 2.5) and the
reaction products (Figure 2.3) showed a much lower free OH band at 3400cm-1
, and the
absence of Si-H bands which should be present at 2100cm-1 20
. Both of these facts are
indications that the starting materials had reacted to form an adduct. In addition a new,
broad band was observed at 907 cm-1
, which was likely a band due to a Si-O –phenyl
bond21
.
91
3500 3000 2500 2000 1500 1000 500
80
90
100
% T
ransm
itta
nce
cm-1
Diphenylsilane-hydroquionone polymer 2.1.3
Figure 2.3. FTIR Spectrum of Polymer 2.1.3, showing the Si-O-Ph absorption band at
910cm-1
.
92
Figure 2.4. FTIR of diphenylsilane22
. Note the strong band at 2100cm-1
, for the Si-H bond
absorption.
Figure 2.5. FTIR of hydroquinone22
. Note the broad band at 3400cm-1
, from the OH
stretching.
93
2.4.2.2. Molecular weight
Molar ratios of the monomers were varied in order to obtain polymers of the
desired Mn for subsequent endcapping. The Carothers equation was used to calculate the
theoretical monomer ratios required for the target Mn23
:
repeatMMn *)21(
)1(
−+
+=
γργ
γ
Where Mn=number average molecular weight, γ=ratio of functional groups in monomer 1
to the functional groups in monomer 2, and ρ=extent of reaction (ρ=1=100% reaction
completed), and Mrepeat=mass of the repeat unit of the polymer.
The results of several melt condensation reactions with various starting monomer
ratios of diphenylsilane to hydroquinone are presented in Table 2.2.
94
Table 2.2. Molecular weight summary of several diphenylsilane-diol melt condensation
polymers.
Sample Mw Mn PDI DPS/HQ Ratio
2..6-1 25,400 13,500 1.9 1
2.6-2 7,000 3,000 2.36 1.13
2.6-3 13,000 5,200 2.50 1.13
Note sample codes are based on polymerization scheme numbers. For example, Sample
2.6= sample prepared from Scheme 2.6.The final characters are used for differentiation of
the specimens, if several reaction results of a specific type are listed.
2.4.2.3. Thermal analysis of melt polymerized hydroquinone-biphenol polymer
TGA analysis of 2.1.3 showed good thermal stability, with a 5 % weight loss in
air of 454ºC. The measured residue after full oxidation of 19.7% compares well with the
theoretical residue value of 20.7% for hydroquinone-diphenylsilane polymer. Theoretical
residue weight is calculated as follows: The aryloxydiphenylsilane polymer repeat unit
weight is 290.4 Daltons. One silicon atom is present per repeat unit, which results in a
weight % silicon of 9.7%. Conversion of silicon to SiO2 on complete oxidation of the
polymer in air yields a theoretical SiO2 weight of 21%.
95
454.3°C 5.0% Loss
Residue:19.7%(2.6mg)
0
20
40
60
80
100
120
We
ight
(%)
0 200 400 600 800Temperature (°C)
Figure 2.6. TGA thermogram of polymer 2.6-2 (hydroquinone/diphenylsilane) in air, 10
ºC per minute heating rate.
Rheological testing of the polymer showed good thermal stability at temperatures
of 300 ºC (Figure 2.2). Both storage and loss modulus showed some increase as a
function of time, but cross linking did not appear to occur, as evidenced by the lack of a
crossover point in G’ and G”.
After heating in the rheometer for 30 minutes at 300 ºC, solubility of the sample
was tested by immersion in a vial of THF. The sample dissolved within a few hours. The
solubility of the polymer after rheological testing was another supporting piece of
96
evidence that indicated cross-linking of this polymer did not occur during prolonged
heating at 300ºC.
0 500.00 1000.0 1500.0 2000.0
time (s)
10.00
100.0
1000
10000
G' (
Pa)
10.00
100.0
1000
10000
G'' (P
a)
0
1.000
2.000
3.000
4.000
5.000
6.000
7.000
8.000
9.000
tan(d
elta)
Figure 2.7. Parallel plate rheogram of Sample 2.6-2. Plot of storage modulus, G’, loss
modulus G’’, and tan δ.
97
0 500.00 1000.0 1500.0 2000.0
time (s)
1.000
10.00
100.0
1000
|n*|
(P
a.s
)
Figure 2.8. Parallel plate rheogram, of sample 2.6-.3 .Plot of complex viscosity η* during
isothermal test conducted at 300°C.
Melt viscosities of linear polymers at very low shear rates follow the Fox-Flory
Power Law, described by the equation below24
:
2.04.3 ±= KMoη
Where η0=zero shear viscosity (typically extrapolated to zero from several progressively
lower shear measurements), M= weight molecular weight of the polymer, and K=
proportionality constant.
98
If one measures the viscosity of a polymer at very low shear rates, it is
proportional to the molecular weight of the polymer to the 3.4 power. If any change in
viscosity is measured in a sample, this equation can be used to estimate molecular weight
changes for polymers during thermal cross-linking. If the starting molecular weight, M, is
a known quantity, and an initial viscosity is measured at a specified temperature at
time=0, the power law equation can be solved for K via the following equation:
4.3
)(
MK oη
=
Zero shear viscosity at a specified time, η0, and the calculated K can then be used to
estimate the molecular weight, M’, after heating for the specified period of time.
The shear rates used for the melt testing reported in this work are very low, 10-5
sec-1
, so viscosities measured during this testing should approximate η0. Based on this
approximation, the complex viscosity of Sample 2.6-2 as a function of time at 300 °C is
shown in Figure 2.8.
Starting viscosity was 4 Pa-sec, final viscosity after heating for 30 minutes was
300 Pa-sec. As reported in Table 2.3, this corresponds to a molecular weight increase of
100%. This molecular weight change is what would be expected from one additional
coupling reaction of the polymer chains (recall that in condensation polymerization,
molecular weights double for each coupling reaction: (monomer+monomer=dimer,
dimer+ dimer=tetramer, etc. 25
).
99
Table 2.3. Calculated molecular weight changes for diphenylsilane polymer after heating
to 300 °C for 30 minutes.
Sample
ηo
(measured by
parallel plate
test)
K MW
2.6-2i ( t=0 min) 4 Pa Sec 1.69e-12
(calculated)
7,000
(GPC)
2.6-2f ( t=30
min) 226 Pa Sec 1.69e-12
14,000
(Calculated)
Comparison of DSC thermograms before and after rheological testing showed the
Tg did not shift significantly (Figure 2.9). For reference, note that an estimate of the
repeatability of Tg values measured by differential scanning calorimetry range from 1.9 to
2.8 °C, as measured on polyurethane, polystyrene and epoxy glass composites26
.
100
Figure 2.9. Overlay of DSC scans of sample 2.6-2, as synthesized vs. after rheology test
(30 minutes, 300°C).H= half-height for Tg calculation. (20 °C/min heating rate, nitrogen
atmosphere).
TGA analysis with an isothermal step of 300 ºC for 30 minutes, corresponding to
the rheological test dwell time, was also performed (Figure 2.10). Minimal weight loss
of 0.2% was observed during the dwell at 300 ºC, which is another indication of good
thermal stability of this polymer system. Note the test atmosphere was switched from
nitrogen to air after the isothermal hold in order to investigate the possibility of changes
in thermal stability after heating, due either to removal of volatiles or possible cross-
66.9°C(H)63.5°C
70.3°C
65.5°C(H) 63.0°C
68.°C
-0.6
-0.5
-0.4
-0.3
-0.2
0 50 100 150 200 250 300Temperature (°C)
Sample 2.6-2i, as synthesized––––––– Sample 2.6-2f, after Rheology Test– – – –
Exo Up
He
at
Flo
w,
W/g
101
linking reactions, and to generate SiO2 residue for semi-quantitative estimates of silicon
content.
Figure 2.10. TGA plot of sample 2.6-2 Y1 Axis=Weight %, Y2 Axis = temperature.
A comparison of the TGA scans in air with and without the isothermal hold under
nitrogen shows some slight differences in weight loss at higher temperatures. There were
also some changes in the weight loss curve above the 5% decomposition point, but no
significant change in the residue weight (Figure 2.11).
472.1C, 95.00%
Residue: 19.86%
(2.529mg)
0.2357% Weight Loss, 300C Isothermal (0.03002mg)
0
200
400
600
800
Tem
pera
ture
(°C
)
0
20
40
60
80
100
120
Weig
ht
(%)
0 20 40 60 80 100 120 Time (min)
102
The 5% weight loss temperature increased about 18 ºC, which can be interpreted
as an increase in thermal stability due to the isothermal hold. This is a possible indication
of some chain extension due to heating, rather than cross linking. Thermal stabilities
typically increase with increasing molecular weight 27
and since the polymer was melt
polymerized and analyzed without further purification, the reaction components,
including catalyst were likely still present in the sample.
Figure 2.11. Overlay of TGA thermograms of sample 2.6-2, showing enhancement in
thermal stability of isothermally heated sample. Isothermal scan was performed up to 330
°C in nitrogen, and then switched to an air atmosphere.
472.1°C 95.0%
454.3°C 95.0%
0
20
40
60
80
100
120
We
igh
t (%
)
0 200 400 600 800Temperature (°C)
Sample 2.6-2, with isothermal––––––– Sample 2.6-2, no isothermal– – – –
103
2.4.3. Diphenylsilane-biphenol polymer (melt polymerization)
2.4.3.1. Molecular weight
GPC analysis of one representative sample (Figure 2.12) showed the material had
a weight average molecular weight of ~1,500, which is only about 8 repeat units (the
mass per repeat unit of this polymer is 366 Daltons). In addition, significant peaks were
seen at points on the chromatogram that corresponded to molecular masses of 741 and
320. Even though GPC is a secondary method of molecular weight determination, it is
likely that these peaks correspond to 4 repeat units, 2 repeat units and 1 repeat unit of the
polymer. In cases such as this where dimers, trimers, etc are seen as separate distinct
peaks, this type of distribution is a clear indication that the polymerization reaction did
not proceed to build high molecular weight. It is interesting to note that even though high
molecular weight was not obtained, the viscosity of this melt polymerized sample was
such that a magnetic stir bar ceased spinning. This brings up yet another limitation of
melt polymerization, the need for higher torque mechanical stirrers, rather than the more
commonly used magnetic stirrers.
104
Figure 2.12. GPC chromatogram of biphenol-diphenylhydrosilane.
In addition, the melt polymerized sample was found to be completely soluble in
acetone. This was a further indication of low molecular weight.
2.4.3.2. Thermal analysis
DSC analysis of the polymer showed a very low Tg of 76 °C (Figure 2.13), rather
than a Tg in the expected 120 °C range. A small endotherm was also seen at 191 °C. It is
6 5 4 3 2 1
-2
0
2
4
6
8
10
12
Inte
nsity (
Arb
. units)
Log Molecular Weight
2.45 (MW=320)
3.17 (MW=1,500)
2.87 (MW= 741)
105
likely this endotherm was either a melting point or a boiling point. This temperature did
not correspond to the melting or boiling point of either starting material. For reference,
diphenylsilane has a predicted boiling point of 217 °C, and biphenol has a melting point
of 245 °C.
75.7°C(H)72.9°C
78.6°C191.4°C
-0.8
-0.6
-0.4
-0.2
0.0
0.2
Heat
Flo
w (
W/g
)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
Figure 2.13. DSC thermogram of biphenol-diphenylsilane melt polymerized sample (2nd
heat).20°C/min, nitrogen atmosphere.
Lower Tg s such as this are typically a sign of a low molecular weight. Lower
molecular weight polymers have higher proportions of end groups, which have higher
106
mobility than backbone segments (due to less entanglement). Therefore the lower
molecular weight polymer will have usually have a lower Tg than a high molecular
weight polymer 28
.
2.5. Conclusions for hydrosilane condensation reactions
This synthesis yielded molecular weights close to targets for subsequent
endcapping for the hydroquinone system. However, the overly-vigorous nature of the
melt polymerization, the failure to polymerize in solution reactions made the system
impractical for large scale production. The biphenol system did not yield useable
molecular weights in either melt or solution systems.
In addition, the dearth of potential monomers for either endcapping reagents or
for main chain ethynl synthesis disqualified this system from further studies. Based on
these factors, this synthetic avenue was not vigorously pursued for further exploration in
this work.
2.6. Dichlorosilane diol condensations
The focus of this part of the research was to understand differences between the
various solvent systems to determine an optimum solvent system for endcapping and for
main-chain ethynyl synthesis. The main goal was to prepare the required molecular
weight. Therefore, the results in this section will be discussed together, rather than system
by system.
2.6.1. Dichlorosilane-hydroquinone condensations
2.6.1.1. Molecular weights
Table 2.4 presents a summary of the molecular weights of all solvent systems
explored for the condensation reaction of dichlorodiphenylsilane with hydroquinone.
107
Comparisons of samples with the same stoichiometric ratios and different synthetic
solvent/base systems gives insight into relative efficiencies of the systems ( e.g..
comparison of 2.12T-1, 2.12T-2 , 2.12MF 2.12MS). Comparison of samples made with
the same solvent systems and different stoichiometric ratios gives insight into the
sensitivity of the system to stoichiometric imbalances (e.g. 2.9-1 and 2.9-2).
108
Table 2.4. Summary of Molecular Weights, Dichlorosilane-hydroquinone polymers
prepared with different reaction solvents and different monomer ratios.
Polymer Mw Mn PDI
Reaction
Solvent
DCPDS/HQ
Ratio Base
Temperature
,
C
Biphenol <200 <200 -- -- -- -- --
DCDPS <200 <200 -- -- -- -- --
2.9-1
19,200
8,100 2.37 CH2Cl2 1.00 pyridine 0
2.9-2
8,900
3,700 2.41 CH2Cl2 1.05 pyridine 0
2.12H-1
8,600
4,400 1.95 THF 1.05 TEA 68
2.12-H2
9,500
4,900 1.94 THF 1.05 TEA 68
2.12-H3
990
621 1.59 THF 1.10 TEA 68
2.12T-1
35,400
15,300 2.31 Toluene 1.00 TEA 78-80
2.12T-2
14,700
6,000 2.45 Toluene 1.05 TEA 78-80
2.12MF
( Mixed-
full rxn)
26,700
11,600 2.30
Toluene
+THF full
rxn 1.00 TEA 78-80
2.12MS
( Mixed,
start of
rxn)
16,900
7,400 2.28
Toluene
+THF-full
toluene at
end 1.00 TEA 78-80
Note samples are labeled by Schemes used for the synthetic reactions. Example: 2.12 is a
reaction which followed Scheme 2.12. Other characters are used to represent replicate
runs(-1, -2) and solvents (H=THF, T=toluene, MF= mixed for full reaction, MS=mixed
for start of reaction only.
Based on this data, several conclusions can be drawn.
1. With 1:1 stoichiometry, toluene systems yielded the highest molecular weights.
109
2. For a 1.05:1 ratio (excess dichlorodiphenylsilane), theoretical Mn is 11,300 based on
the Carothers equation.
300,114.290*)1952.02952.01(
)952.01(*
)21(
)1(=
××−+
+=
−+
+= repeatMMn
γργ
γ
Toluene gave the highest Mn. However, theoretical values were not attained for 1.05
stoichiometries.
3. The mixed THF: toluene system did not appear to help in terms of increase in
molecular weights over a full toluene system.
4. Increasing the stoichiometric excess of dichlorodiphenylsilane to 1.1:1 severely
dropped the molecular weight (Theoretical value was 7,000).
2.6.1.2. Thermal analysis of dichlorodiphenylsilane-hydroquinone polymers
TGA and DSC analysis of representative polymers of this type showed properties
that were slightly lower than the analogous diphenylhydrosilane-hydroquinone polymers.
5% weight loss in air was ~ 20 °C lower, and Tg was ~ 4 °C lower, but this is within the
statistical error of measurement by DSC26
.
110
437.3°C 95.0%
Residue:20.0%(3.0mg)
0
20
40
60
80
100
120W
eig
ht (%
)
0 200 400 600 800Temperature (°C)
Figure 2.14. TGA of polymer 2.12B (DCDPS, hydroquinone) in air (10°C/min)
111
61.4°C(H)59.6°C
63.3°C
-0.06
-0.04
-0.02
0.00
0.02
0.04
0.06
0.08H
ea
t F
low
(W
/g)
-50 0 50 100 150 200 250 300Temperature (°C)
Exo Up
Figure 2.15. DSC of polymer 2.12 B (DCDPS, hydroquinone) in nitrogen (20 °C/min)
2.6.2. Dichlorosilane-biphenol condensations
2.6.2.1. Molecular weights
Comparison of the molecular weights of samples prepared from DCDPS and biphenol
in various solvent systems are summarized in Table 2.5. Again, through review of
the number average molecular weights (Mn values) of the different synthetic systems,
one can determine the optimum system for hitting the target molecular weights, and
also evaluate the sensitivity of the system to stoichiometric imbalances.
112
Table 2.5. Summary of Molecular Weights, dichlorosilane-biphenol polymers prepared
with different reaction solvents and different monomer ratios
Polymer Mw Mn PDI
Reaction
Solvent
DCDPS
Biphenol
Ratio Base Temp.,C
Biphenol <200 <200 -- -- -- -- --
DCDPS <200 <200 -- -- -- -- --
2.9-1
18,500
7,400 2.50 CH2Cl2 1.00 pyridine 0
2.9-2
16,900
7,300 2.32 CH2Cl2 1.05 pyridine 0
2.11-1
40,800
19,900 2.05 Toluene 1.00 TEA 78-80
2.11-2
51,400
20,300 2.53 Toluene 1.05 TEA 78-80
2.11-3
5,400
2,600 2.08 Toluene 1.10 TEA 78-80
2.11P
Tan ppt
start of rxn
1,950
608 3.21 Toluene 1.15 TEA 78-80
2.13D
23,300
12,300 1.89
THF
(distilled) 1.05 TEA 68
2.13U 12,700 6,800 1.9
THF
(undistilled) 1.05 TEA 68
Note samples are labeled by the synthetic used for their synthesis. For example Samples
2.9-1 and 2.9-2 are made using conditions specified in Scheme 2.9. The last characters
are used either to designate replicates, or special sample conditions (P=precipitate,
D=distilled, U=undistilled).
2.6.2.1.1. Estimated molecular weights vs. theoretical molecular weights from the
Carothers equation
Theoretical number average molecular weight based on a 1.05:1 monomer ratio is
15,000. Samples prepared using this ratio synthesized in toluene (double precipitated)
113
had higher values than theoretical (Higher than predicted Mn values were also measured
for several samples discussed in Chapter 3). Since GPC is a secondary technique that
calculates molecular weights based on hydrodynamic volumes, not absolute molecular
weights, this may be an indication that the biphenol system may deviate from the
polystyrene calibration standards in terms of relative hydrodynamic volumes per
equivalent molecular weights.
Biphenol is a bulky, rigid molecule. Its inclusion in the DCDPS polymer increases
the Tg by ~ 60°C, due to its rigidity, relative to hydroquinone links. This could
potentially impart more rigid-rod character to the polymer. Increased rigidity could result
in an increase its hydrodynamic volume. This would lead to an overestimate of molecular
weight. For rigid rod systems, it has been shown that GPC values can actually
substantially overestimate molecular weights, by as much as a factor of 229, 30
. One other
possibility is that the dual precipitations used in the workups of these polymers actually
resulted in some fractionation of the polymer. i.e. removal of some of the lower
molecular weight fractions through the 2nd
precipitation step. In either case, for the
purposes of this work the absolute values of molecular weight are of secondary
importance; the relative values were sufficient for evaluating the effectiveness of the
different synthetic systems.
2.6.2.1.2. Molecular weight summary
For the biphenol polymer, toluene systems yielded the highest molecular weights
followed by THF systems, which also yielded high molecular weights. The CH2Cl2
system did not appear to be effective for biphenol. Based on this data, toluene and THF
were both acceptable for end capping systems or for ethynyl monomer systems, to be
114
explored in Chapters 3 and 4 of this work. Also, in terms of stoichiometric ratios, again
1.05 appeared to be an optimal value both for subsequent end capping with the targeted
number of end groups, and also from a purification perspective. Ratios of 1.10 and 1.15
yielded lower molecular weights, and may be acceptable for work to produce oligomers,
but not polymers that can be easily obtained by precipitation into methanol (based on
solubility in methanol of polymers with GPC Mn values <10,000).
2.6.2.2. Thermal analysis of dichlorodiphenylsilane-biphenol polymers
Thermal characterization of these polymers was also performed in order to verify
the high thermal stabilities reported in the literature, and to verify silicon content through
TGA in an air atmosphere. The TGA (Figures 2.16, 2.17) indicate relative thermal
stabilities of the polymers. Higher temperatures before significant mass loss indicate
higher thermal stability. When a sample is fully oxidized at the end of a TGA run (at the
highest temperature of the run), the residual weight can be used to estimate the silicon
content in the polymer. The DSC scan (Figure 2.18) shows the Tg of the material. Table
2.6 summarizes the thermal analysis data for these polymers.
115
515.8°C 95.0%
Residue:16.5%(1.2mg)
0
20
40
60
80
100
120W
eig
ht (%
)
0 200 400 600 800Temperature (°C)
Figure 2.16. TGA in nitrogen of 2.11-2 Dichlorodiphenylsilane-biphenol (10 °C/min).
Note atmosphere was switched to air after 330° C in order to oxidize material and obtain
residue for Si quantification.
116
501.4°C 95.0%
Residue:15.4%(0.4mg)
0
20
40
60
80
100W
eig
ht (%
)
0 200 400 600 800Temperature (°C)
Figure 2.17. TGA in air of 2.11-2 (Dichlorodiphenylsilane-biphenol) in air (10 °C per
min heating rate).
117
137.0°C(H)134.6°C
139.4°C
-4
-2
0
2
4H
ea
t F
low
(W
/g)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
Figure 2.18. DSC scan of 2.11-2 ( dichlorodiphenylsilane- biphenol), 20°C/min, nitrogen
atmosphere.
2.6.2.3. Spectroscopic analysis
FTIR analysis of the polymer when compared to the starting reagents (Figure
2.19) showed several key differences that confirmed new bonds had formed in the
polymerization reaction. The band at 3,400 cm-1
for biphenol monomer, corresponding to
OH stretching, was greatly reduced in the polymer sample. In addition, the absorbance at
900cm-1
that corresponded to an Si-O-aromatic stretching was not seen in either starting
reagent. Both observations were indications that the terminal OH groups in biphenol
reacted with the dichlorodiphenylsilane to form the Si-O-aromatic linkage.
118
4000 3500 3000 2500 2000 1500 1000
DCDPS Reagent
( Nujol)
Biphenol Reagent
cm-1
DCDPS-Biphenol Polymer
Figure 2.19. FTIR of polymer 2.11, with FTIR spectra of the starting reagents overlaid
for comparison.
13C NMR analysis of a representative biphenol polymer showed a spectrum with
peaks that were characteristic of an adduct of dichlorodiphenylsilane (Figure 2.20). The
peak at 153 ppm is characteristic of an ipso aromatic carbon bonded to an oxysilane (C-
O-Si bond). Literature values for related phenoxysilane model systems show chemical
shifts of 155-154 ppm for the ipso carbon31
. This carbon appears to have a slight upfield
shift from the biphenol ipso carbon, which has a predicted chemical shift of ~ 158 ppm32
.
The peak positions are different from the 13
C NMR spectra of the starting materials,
which are shown in Figures 2.21 and 2.22).
119
160 150 140 130 120 110 100
0.00E+000
1.00E+009
2.00E+009
3.00E+009
4.00E+009
5.00E+009
6.00E+009
Arb
. U
nits
ppm
Figure 2.20. 13
C NMR of a representative DCDPS biphenol polymer, 20,000 Mn
(CDCl3).
120
Figure 2.21.
13C NMR of dichlorodiphenylsilane (starting reagent)
22.
121
Figure 2.22. 13
C NMR of biphenol (starting reagent)32
.
2.7. Conclusions
Table 2.6 summarizes the relevant thermal data for the various polymers
synthesized and characterized during the course of this work. Comparison of samples
collected under the same atmosphere allows us to determine which material has the
highest thermal stability. From this data, we can conclude that diphenylsilane
hydroquinone polymers (represented by Sample 2.6-C) have better thermal stability than
dichlorodiphenylsilane polymers (represented by Sample 2.12-C), as evidenced by the
higher 5% weight loss values.
We can also conclude that dichlorodiphenylsilane biphenol polymers (Sample
2.11) have better thermal stabilities than the hydroquinone polymers. Comparison of the
Tg values shows that Tgs are fairly close for polymers with the same repeat units,
122
indicating that at these molecular weights the identities of the endgroups (Si-H vs. Si-
OH) does not have a significant effect on the Tg. One can also conclude that the identity
of the aromatic diol included in the repeat unit has a significant effect on the Tg. The
biphenol polymer has a Tg which is 75-80 °C higher than the hydroquinone unit.
Evaluation of the residue levels of all materials show relatively good agreement
with theoretical values.
Table 2.6. Summary of Tg, TGA 5% weight loss and residue content in nitrogen and air.
Sample Description Atmosphere
Tg in
nitrogen
(20°C/min)
5%
Weight
Loss
Residue
Content
(%)
Theoretical
Residue
(%)
2.6-2
DPS-HQ
Polymer Air
65°C 454.3 19.7 20.7
2.6-2
DPS-HQ
Polymer Nitrogen
65°C 472.1 19.9 20.7
2.12-C
DCDPS-HQ
polymer Air
61°C 437.3 20.0 20.7
2.11-2
DCPS-
biphenol
polymer air
137°C 501.4 15.4 16.5
2.11-2
DCPS-
biphenol
polymer Nitrogen
137°C 515.8 16.7 16.5
Samples are designated by the synthetic scheme used to prepare the polymer.
Criteria for selection of the optimum polymer system for further work in chapters 3
and 4 are the following:
1) High thermal stability
123
2) Acceptable molecular weight, for target end group concentration and/or facile
synthesis of ethynyl containing monomers, and
3) Chemistry that is compatible with ethynyl substituents and ethynyl containing
monomers
Based on these criteria, the following systems were disqualified from further study:
Aminosilane-diol: was not explored due to high required reaction temperatures (in
ethynyl curing temperature range), and due to cleanup issues reported in the literature,
with monomers and condensation byproducts potentially interfering with radical cross-
linking reactions (this could interfere with ethynyl curing as well).
Hydrosilane-diol (hydroquinone and biphenol): disqualified due to difficulty
controlling the melt polymerization, inability to duplicate solution based synthesis, and
limited availability of ethynyl monomers and end capping reagents.
Dichlorosilane-hydroquinone: disqualified due to lower thermal stabilities, and some
issues with oxidation of polymers (many polymers had grayish-black colors, which was
attributed to some possible oxidation of hydroqinone).
Dichlorosilane-hydroquinone in CH2Cl2/pyridine: this system yielded low molecular
weights when compared to THF and toluene systems, with triethylamine as the acid
scavenger.
The system selected for further development was the dichlorodiphenylsilane-
biphenol system in toluene. Dichlorodiphenylsilane-biphenol condensation in THF was
also acceptable, albeit appeared to yield slightly lower molecular weights. THF was still
chosen as the solvent for several reactions in Chapters 3 and 4 due to its superior
124
solubilization of diols ( relative to toluene), and for use with the lithiumphenylacetylide
endcapping reagent, as this reagent can only be obtained as a 1.0M solution in THF.
125
2.8. References
(1) Klingender, R. C., Ed.; In Handbook of Specialty Elastomers; CRC Press: Boca
Raton, FL 2008,102.
(2) Steffen, V.K.-D, Poly-dioxyarylen (dioxycycloalky1en)-diphenylsilane I. Synthese
durch Polykondensation von Dihydrido-diphenylsilanen und anderen, teilweise
neuen bifunktionellen Diphenylsilanen mit Diphenolen bzw. Cycloalkylendiolen.
Die Angewandte Makromolekulare Chernie 1972 24, (313), 1-20.
(3) Liaw, D. L., Synthesis and Characterization of Novel Polyaryloxydiphenylsilane
Derived From 2,2'- Dimethyl-biphenyl-4,4'-Diol. Journal of Polymer Science, Part
A: Polymer Chemistry 1999, 37, (24), 4591-4595.
(4) Dunnavant, W. R.; Markle, R. A.; ; Sinclair, R. G.; Stickney, P. B. ; Curry, J. E.;
Byrd, J. D., p,p Biphenol Dianilosilane Condensation Polymers. Macromolecules
1968, 1, (3), 249-254.
(5) Sommer, L., Stereochemistry of Asymmetric Silicon. V. Coupling Reactions with
Organometallic Reagents and Displacements of Chloride and Fluoride Leaving
Groups. Journal of the American Chemical Society 1967, 89, (4), 862-868.
(6) Sommer, L. ,Stereochemistry of Asymmetric Silicon. IV. The Sn2-Si
Stereochemistry Rule for Good Leaving Groups. Journal of the American Chemical
Society 1967, 89, (4), 857-861.
(7) Denmark, S. E.; Beutner, G. L. Lewis base catalysis in organic synthesis.
Angewandte Chemie, International Edition 2008, 47, (9), 1560-1638.
(8) Eaborn, C.; Jenkins, I. D., Mechanism of the base-catalyzed alcoholysis of
triorganosilanes. Journal of Organometallic Chemistry. 1974, 69, (2), 185-92.
(9) Steffen, V.K.-D. Preparation of High Molecular Weight Thermoplastic
Polyorganosilicic acid esters. US Patent 4026827, 1977.
(10) Dvornic, P. R., Degradative side reactions in the syntheses of exactly alternating
silarylene-siloxane polymers. Polymer Bulletin 1992, 28, (3), 339-344.
(11) Dunnavant, W. R.; Markle, R. A.; Sinclair, R. G.; Stickney, P. B. Second Annual
Summary Report on Process Development and Pilot-Plant Production of Silane
Polymers and Diols. Batelle Memorial Institute: Columbus, OH 1967, 77-88.
126
(12) Sastri, S. B.; Keller, T. M.; Jones, K. M.; Armistead, J. P., Studies on Cure
Chemistry of New Acetylenic Resins. Macromolecules 1993, 26, (23), 6171-6174.
(13) Wang, F.; Xu, J.; Zhang, J.; Huang, F.; Shen, Y.; Du, L., Synthesis and thermal cure
of diphenyl ethers terminated with acetylene and phenylacetylene. Polymer
International 2006, 55, (9), 1063-1068.
(14) White, D.; Levy, G., 13C Nuclear Magnetic Resonance Spectra of m-
Diethynylbenzene Polymers and Related Polyacetylenes. Macromolecules 1972, 5,
(4), 526- 531.
(15) Back, S., Aerospace Organic Matrix Composite Materials Task Order 0005:
Complex Architecture and Analytical Methods −Synthesis and Cure Mechanism
Characterization of Phenylethynyl-Terminated Imide Oligomers. University of
Dayton Research Institute Technical Report UDR-TR-2009-132 2009, 26.
(16) diphenylsilane, catalog SID4559.0. www.gelest.com.
(17) Tamai, S.; Tanaka, C.; Abe, T.; Kuroki, T.; Ishikawa, J., Preparation and Properties
of Optically Clear Poly[(arylene dioxy) (diphenylsilanes)], and poly[(alkylene)
(dioxydiphenylsilane)]. High Performance Polymers 2003, 15, (3), 361.
(18) Brunelle, D. ; Singleton, D. A. N-alkyl-4-(N',N'-dialkylamino)pyridinium salts:
thermally stable phase transfer catalysts for nucleophilic aromatic displacement.
Tetrahedron Letters 1984, 25, (32), 3383.
(19) Brunelle, D., Cyclic oligomer chemistry. Journal of Polymer Science. Part A,
Polymer Chemistry 2008, 46, (4), 1151-1164.
(20) Kniseley, R. N.; Fassel, V. A.; Conrad, E. E., Observations on the silicon-hydrogen
vibrational bands in alkyl and aryl substituted silanes. Spectrochimica Acta 1959,
(15), 651-655.
(21) Curry, J. E.; Byrd, J. D., Silane polymers of diols. Journal of Applied Polymer
Science 1965, 9, (1), 295-311.
(22) National Institute of Advanced Industrial Science and Technology Spectral Database
for Organic Compounds (SDBS). http://riodb01.ibase.aist.go.jp/sdbs/ (accessed
12/01, 2010).
(23) Allcock, H. R.; Lampe, F. W. In Contemporary Polymer Chemistry. Second Edition;
Prentice Hall: Englewood Cliffs, NJ 1990, 275.
(24) Dealy, J. M.; Larson, R. G. In Structure and Rheology of Molten Polymers - From
Structure to Flow Behavior and Back Again. Hanser Publishers: Cincinnati, OH
2006, 132-135.
127
(25) Allcock, H. R.; Lampe, F. W. In Contemporary Polymer Chemistry. Second Edition;
Prentice Hall: Englewood Cliffs, NJ 1990, 258.
(26) Standard Test Method for Transition Temperatures and Enthalpies of Fusion and
Crystallization of Polymers by Differential Scanning Calorimetry; ASTM Volume
08.02 Plastics (II): D3222 - D5083; ASTM International: Conshohocken, PA, 2003.
(27) Kelen, T., In Polymer Degradation. Van Nostrand Reinhold Company: New York,
NY, 1983,2-6.
(28) Sperling, L. H. In Introduction to Physical Polymer Science. Fourth Edition; John
Wiley & Sons: Hoboken, NJ 2006, 845.
(29) Moroni, M.; Le Moigne, J.; Luzzati, S., Rigid rod conjugated polymers for nonlinear
optics: 1. Characterization and linear optical properties of poly(aryleneethynylene)
derivatives. Macromolecules 1994, 27, (2), 562-571.
(30) Vanhee, S.; Rulkens, R.; Lehmann, U.; Rosenauer, C.; Schulze, M.; Kohler, W.;
Wegner, G., Synthesis and Characterization of Rigid Rod Poly(p-phenylenes).
Macromolecules 1996, 29, (15), 5136-5142.
(31) Hudrlik, P. F.; Minus, D. K., NMR spectra of phenoxysilanes with various silyl
groups. Journal of Organometallic Chemistry 1996, 521, (1-2), 157-162.
(32) ACD Labs Predicted 13
C NMR structure of 4,4' biphenol. www.scifinder.org
(accessed 12/01, 2010).
128
Chapter 3: Phenylethynyl and phenol end capping studies of
polybiphenyloxydiphenylsilanes for cross-linking and enhanced thermal stability.
3.1. Introduction
Thermally stable polymers are critical for many applications. Numerous industries
are pushing the limits of current organic backbone polymers 1-3
. One possible solution is
to use hybrid materials, with organic-inorganic bonds in the polymer backbone, for
enhanced thermal stability 4-8.
Siloxanes are one such family of materials, however until
now their ultimate thermal stabilities have been limited by degradative depolymerization
which occurs at lower temperatures than anticipated based on Si-O bond strengths 9.
Aromatically substituted polysiloxanes have been studied for decades10-11
. The
repeat unit of polydiphenylsiloxane, one of the major polymers in this category, is shown
in Figure 3.1:
n
Si O
Figure 3.1. Polydiphenylsiloxane.
129
Polydiphenylsiloxane has very good short term thermal stability. It is a polymer
liquid crystal with a glass transition of 265 °C and a melting point of 540 °C.
Thermogravimetric analysis (TGA) has shown a 10% weight loss temperature of 511 °C.
However processability, achievement of high molecular weights, and long term thermal
stability have been continuing issues with this material13
.
Depolymerization via cyclization is a likely thermal degradation mechanism6,12
.
In order to inhibit this reaction, several research groups have evaluated the use of
aromatic spacers (biphenoxy units) in the backbone8, 14
, yielding polymers with the repeat
structure shown in Figure 3.2.:
O O Si
n
Figure 3.2. Repeat unit of polyaryloxydiphenylsilane.
Thermal stabilities of polyaryloxydiphenylsilanes are also quite good. TGA
analysis showed a 10% weight loss of 565 °C15
. However, the uncrosslinked polymer
130
has a softening point of ~150 °C 11
, which limits its ultimate utility as a high
temperature polymer.
Several reaction pathways are available for synthesis of this class of polymers16
.
Highest molecular weights are obtained via reaction of anilinosilanes with biphenols11
,
in which aniline is produced as a byproduct of the reaction. Attempts at cross-linking of
this polymer via inclusion of vinyl moieties yielded materials of limited utility, which the
authors attributed as likely due to residual aniline impurities from the synthesis (aniline is
a possible free radical scavenger) 17
.
In order to fully exploit the potential high temperature properties of this polymer
system, cross-linking is required. Cross-linking enables the polymer to remain solid
above its Tg, and provides enhanced chemical resistance. Ethynyl groups have been
successfully used for thermal cross-linking of a family of high temperature polyimides,
PETI (phenylethynyl terminated polyimides)18
, so their use with aryloxysilanes was a
logical extension of this cross-linking technology. The focus of the research performed in
this chapter was to explore the feasibility of endcapping of this polymer system with
phenylethynyl groups, and to develop methods to verify endcapping was successful.
Based on the evaluation of various synthetic routes for preparation of this polymer
performed in Chapter 2, the following reactions were selected for synthesis of endcapped
polymers:
131
OH OH + SiCl Cl O O Si
n
THF or toluene,
TEA
60 -80 °C
N2
Scheme 3.1. Condensation polymerization reaction between dichlorodiphenylsilane and
biphenol. Dichlorosilanes were added in 5% excess to obtain chlorosilane endgroups.
This DCDPS diol condensation reaction typically generates somewhat lower
molecular weights than other synthetic routes, including the aforementioned
anilinosilane-diol route11
, but had the advantage of utilizing commercially available
reagents and also the flexibility gained from the use of relatively labile chlorosilane
functional group for subsequent endcapping reactions (Figure 3.3). In addition, the lower
molecular weights obtained by this route would yield more reactive end-groups, and thus
in principle more ethynyl groups per mole in the final capped polymers.
O SiO ClOSi OCl O SiOSi
n
Figure 3.3. Aryloxydiphenylsilane polymer with chlorosilane end-groups available for
endcapping reactions (general structure of samples 1a, 1b, 2b, 3b, 4b).
132
Endcapping reactions that were evaluated are listed in Schemes 3.2 and 3.3. Both
these reactions were selected for evaluation because the chemistries of endcapping are
similar to the polymerization reaction. Both involve nucleophilic displacement of the Cl
from the chlorosilane end groups, and both endcapping reactions are feasible in the
solvent systems used in this study (THF and toluene).
OSiCl O Si Cl
Endcap
Li
OSi O Si
Polymer 2a
Scheme 3.2. Lithiumphenylacetylide endcapping reaction.
133
OSiCl O Si Cl
FOH
Endcap
OSi O SiOF O F
Polymer 3a
Scheme 3.3. Fluorophenyethynylphenol endcapping reaction.
3.2. Experimental Section
3.2.1. Materials
Dichlorodiphenylsilane (DCDPS), 99% purity, was purchased from Gelest
(catalog number SID4510.1); 4,4’–biphenol, 99.9% purity, was purchased from TCI
(catalog number B0464); triethylamine(TEA), 99%, (catalog number 15791) was
purchased from Acros; 4[(4-fluorophenylethynyl)]phenol, 99% purity, was purchased
from Wako (catalog number 320-90143) ; lithium phenylacetylide 1 M solution in THF,
was purchased from Sigma Aldrich (catalog number 340677) ; phenol, 99% purity, was
purchased from Sigma Aldrich (catalog number P3653); toluene (99.9% anhydrous) was
purchased from Sigma Aldrich ( catalog number 244511); anhydrous tetrahydrofuran
(catalog number. 181500010) was purchased from Acros ; methanol was purchased from
Pharmco-Aaper. Unstabilized THF was freshly distilled in the presence of LiAlH4 and
used as the reaction solvent immediately after distillation.
134
3.3. Polymer Synthesis
Polymers were synthesized by polycondensation of DCDPS with biphenol in
toluene or in THF with an inert atmosphere purge under reflux at 65-70 °C following the
procedure outlined in Chapter 2, Section 2.2.2.8. TEA was used as an acid scavenger to
remove the condensation byproduct HCl from the system.
A typical synthesis in toluene was performed as follows: In a sealed 100ml three
neck round bottom flask equipped with a magnetic stir bar, a reflux condenser, and a
nitrogen purge, 3.72 (0.020 moles) grams of biphenol and 4.25 grams (0.042 moles) of
triethylamine was added to 30 ml of toluene. A thermometer was used to monitor
reaction temperature by inserting it into the reaction liquor through a stopper with a
rubber adapter. The solution was heated to 65 to 70 °C on a hot plate with a magnetic
stirrer. To this solution, 5.3 grams (0.021 moles) of DCDPS along with 30 ml of toluene
was added via an addition funnel with an inert gas pressure equalizer. The DCDPS was
added dropwise over approximately one hour. After addition, the solution was heated
under reflux overnight (13-17 hours). Temperature was maintained at 65-70 °C for the
remainder of the reaction. Polymers were purified via multiple precipitations in methanol,
as detailed in Section 3.5. Synthesis in freshly distilled THF was performed following the
same procedure, except that reaction time was 1-2 hours. Typical yields were 65 to 70%
of theoretical after the multiple precipitations.
Both the THF and toluene solvent systems were found to yield roughly equivalent
polymer molecular weights with longer polymerization times employed for the toluene
135
system (1-2 hours for THF vs. 12-24 hours for toluene). Since the polymerization
reaction is nucleophilic in nature, the use of a non-polar aprotic solvent such as toluene
would likely result in slower reactions than those obtained in a polar aprotic solvent,
such as THF19
, which is the likely reason why toluene reactions took longer to reach
equivalent molecular weights.
3.4. General procedure for endcapping
Prior to performing an endcapping reaction, an aliquot of the reaction liquor was
removed as a control (unendcapped sample) and allowed to continue to polymerize in a
separate flask. Immediately thereafter, endcapping agents were added dropwise to the
remaining polymer reaction liquor through an addition funnel. Endcapping reagents were
added in approximately 4X excess, based on nominal number average molecular weight
(Mn) of 12,000. The endcapping reactions were allowed to proceed for a time period
equal to initial polymerization at 65 to 70 °C (1 additional hour for THF solutions, 12
additional hours for toluene reactions).
3.4.1. Endcapping with lithiumphenylacetylide
Lithiumphenylacetylide was added after 1 hr of polymerization in THF or 12 hrs
in toluene. A portion of the solution was separated prior to addition and allowed to
continue to react and build molecular weight (uncapped sample). Both THF and toluene
endcapping reactions were allowed to proceed for an equivalent time period after the
addition of endcaps at the same reaction temperature (65 to 70°C). In order to try to drive
the endcapping reaction to the product side by removal of the condensation product,
toluene reactions were also performed. LiCl, the condensation product of the endcapping
136
reaction, is highly soluble in THF, but insoluble in toluene 20
. Due to the lower polarity of
toluene and slower reaction kinetics, toluene polymerizations were allowed to proceed
overnight (approximately 12-14 hours), THF is a better solvent for nucleophilic reactions,
so reactions were allowed to proceed approximately 1 hour to obtain target molecular
weights. Total polymer yield for a 20 mmol reaction, after precipitation in 0.1M HCl,
redissolving in THF, filtration and reprecipitation in methanol was 4.7 grams total. 4.3
grams of capped material and 0.4 grams of uncapped material were obtained (67%
theoretical).
3.4.2. Endcapping with (4[(4-fluorophenylethynyl)]phenol)
4[(4-fluorophenylethynyl)]phenol was added after 1 hr of polymerization in THF
or 12 hrs in toluene. In a typical procedure, 1.06 grams (0.005 moles) of 4[(4-
fluorophenylethynyl)]phenol and 10ml of THF was transferred into an addition funnel
with an inert glass pressure equalizer. The solution was then added dropwise to the
polymerization reaction over 1 hour at a reaction temperature of 65-70 °C. After the
addition was completed, the reaction was allowed to proceed for an additional hour at 65-
70 °C. A portion of the solution was separated prior to addition and allowed to continue
to react and build molecular weight (uncapped control sample). At the completion of the
reactions, the polymers were precipitated in 0.1M HCl, redissolved in THF, and then
reprecipitated in methanol. Yields after the dual precipitations for a 0.010 mole scale
endcapping reaction were 2.1 grams of capped polymer, 0.5 grams of uncapped polymer,
2.6 grams total (71% of theoretical yield).
137
3.4.3. Endcapping with phenol
Phenol was added after 17 hrs of polymerization in toluene. In a typical
procedure, 0.47 grams (0.005 moles) of phenol and 10ml of toluene was transferred into
an addition funnel equipped with an inert gas pressure equalizer. The solution was then
added dropwise to the polymerization reaction over 1 hour at a reaction temperature of
65-70 °C. The reaction was allowed to proceed for an additional 12 hours at 65-70 °C. A
portion of the solution was separated prior to addition and allowed to continue to react
and build molecular weight (uncapped sample).Yields for a 0.010 mole scale endcapping
reaction were 2.2 grams of capped polymer, 0.3 grams of uncapped polymer, 2.5 grams
total (69 % theoretical yield).
OSiCl O Si Cl
Endcap
OH
O OSi O Si O
Polymer 4a
138
Scheme 3.4. Phenol endcapping reaction
3.5. Polymer purification
Reaction solutions of capped polymers and uncapped controls prepared in THF
were precipitated dropwise into a cold (refrigerated, 7-10 °C) 0.1M aqueous HCl solution
to neutralize any unreacted TEA and remove the soluble TEA:HCl salt21
. Toluene forms
an emulsion in water based systems, so initial precipitations for the toluene systems were
performed utilizing methanol as the non-solvent. The insoluble polymers were then
filtered, dried and redissolved in THF. The THF solutions were filtered to remove any
residual insoluble material, mainly TEA:HCl salts, and then precipitated for a 2nd time in
methanol. The dual precipitation was performed in order to remove unreacted monomers
and excess capping reagents from the final polymer prior to analysis, as both biphenol
and 4[(4-fluorophenylethynyl)]phenol (99%) are highly soluble in methanol. As reviewed
in the preceding section, yields for endcapped polymers that were purified using this
procedure typically were in the 65-70% of theoretical range.
3.6. Characterization
Molecular weight
Molecular weights were determined by gel permeation chromatography using a
Waters 590 LC system (Waters Corp., Milford, MA) equipped with a Phenogel 1µ 103A
GPC column (Phenomenex, Torrance, CA) and a Waters 410 refractive index detector.
Elution rate was maintained at 1 mL/min. THF was used as the mobile phase. Polystyrene
standards were used for the molecular weight calibrations.
139
Spectroscopy
Infrared spectra were collected on a Perkin Elmer (Perkin-Elmer Co., Norwalk,
CT) Spectrum One spectrometer. Spectra of polymers in solution were collected via
transmission IR by dissolving small amounts of sample in THF and mounting between
NaCl plates.
Raman spectra were collected on a Renishaw (Renishaw Plc., Gloucestershire,
UK) RM1000 Raman microscope. Samples were dissolved in THF and mounted between
two glass slides. A laser with a 513nm wavelength was used as the excitation source.
13C NMR spectra were collected using a Varian Unity Inova 300MHz NMR
(Varian Inc., Palo Alto, CA). Samples were dissolved in CDCl3. Scans were collected in
proton decoupled mode. Chemical shifts were calibrated using the CDCl3 solvent peak as
an internal reference.
Thermal analysis
Differential scanning calorimetry thermograms were collected using a TA
Instruments Q100 DSC (TA Instruments, New Castle, DE). Samples were analyzed in
hermetic pans under a nitrogen atmosphere. Pinhole vents were introduced in the pans to
allow any volatiles generated during heating to escape, and also to remove residual
oxygen which might otherwise be trapped in the pan and participate in undesired
oxidation reactions. Pinholes were also introduced into the reference pans, for
consistency. Heating rates of 20 °C per minute and cooling rates of 10 °C per minute
were used. Rheometric analysis was performed on a TA Instruments AR2000 Rheometer
with 8mm parallel plates. Continuous oscillation testing was performed using a 1 % strain
140
and a frequency of 1 Hertz. Testing was performed at 300 °C under a nitrogen purge.
Approximate shear rate for these tests was 1x10-5
sec-1
.
3.7. Results and discussion
3.7.1. Molecular weight summary
Condensation in THF yielded polymers with Mn in the 12,000 to 23,000 range,
relative to the polystyrene standards. This molecular weight range corresponded to
roughly 32 to 62 repeat units, this would yield an approximate a molar concentration of
6% to 3% end groups per mole of polymer (higher molecular weight =lower molar
percentage of end groups). Uncapped control samples had somewhat higher molecular
weights than the capped samples, due to the fact that the reaction was allowed to proceed
for an equivalent time as the endcapped reaction. Capped and uncapped materials
synthesized from the same starting polymerization reactions were compared, even though
they exhibited differences in molecular weights, in order to reduce the possibility of
measuring differences in thermal stabilities due solely to batch to batch variations rather
than the change in thermal properties due to endcapping. This also gave further insight
into the effect of molecular weight on thermal stability.
High molecular weight uncapped samples were obtained by allowing the reactions
to proceed for longer reaction times. It was found that 4 hours in THF, and 34 hours in
toluene were sufficient to allow molecular weights to build to the 20,000 range.
A low molecular weight (ca. 7,000 Mn) uncapped polymer reference sample was
also prepared by using unstabilized THF which was not freshly distilled and only running
the reaction for 15 minutes rather than 2 hours after addition of DCDPS was completed.
141
An Mn of 7,000 corresponds to 19 repeat units. Assuming 100% conversion, this would
correspond to about 10 molar % end groups. All the molecular weight data is summarized
in Table 3.1.
Table 3.1. Polymer molecular weight summary (relative to polystyrene standards).
Sample
Designation Mn Mw
Polydispersity
Index
(Mw/Mn) Description
1a 6,800 12,700 1.9
Low Mn uncapped
polymer
1b 23,400 43,700 1.9
High Mn uncapped
polymer
2a
7,700 16,200 2.1
Lithiumphenylacetyli
de capped polymer
2b 12,000 31,500 2.6
uncapped polymer
(lithiumphenylacetylid
e control sample)
3a 6,500 12,400 1.9
4[(4-
fluorophenylethynyl)]p
henol capped polymer
3b 12,300 23,300 1.9
Uncapped polymer
(4[(4-
fluorophenylethynyl)]p
henol control sample)
4a 22,400 42,000 1.9
Phenol capped
polymer
4b 22,000 40,900 1.9
Uncapped polymer,
phenol control sample
142
3.8. Uncapped polymers
Analysis of uncapped polymers as controls showed cross-linking occurred in
uncapped polymers. In order to characterize this competing cross-linking reaction,
thermal analysis was performed on several reference samples with different molecular
weights as well as a phenol capped control.
DSC scans were collected on two different Mn polymers and compared to the
thermogram of a phenol capped polymer (Figures 3.4 and 3.5):
-3
-2
-1
0
1
He
at
Flo
w (
W/g
)
50 100 150 200 250 300 350Temperature (°C)
Polymer 1a, 6,800 Mn––––––– Polymer 1b. 23,400 Mn– – – – Polymer 4a, Mn 22,400 Phenol capped––––– ·
Exo Up
Figure 3.4. First heat DSC comparison of samples with Mn of 7,000 (1a) and 22,000 (1b)
and phenol endcapped polymer, Mn=22,000 (4a).
143
-4.5
-3.5H
ea
t F
low
(W
/g)
200 250 300Temperature (°C)Exo Up
6,800Mn (11molar % endgroups)
23,000 Mn ( 3 molar % endgroups)
23,000 Mn phenol capped
(>>3% endgroups)
Figure 3.5. First heat DSC comparison of samples with Mn of 7,000 (1a) and 22,000 (1b)
and phenol endcapped polymer, Mn=22,000 (4a), enlarged to show the exothermic peak
at 275 °C.
An exotherm at 275 °C was seen in both uncapped samples 1a and 1b (Figure
3.5), but was more pronounced in the lower molecular weight sample 1a. This exotherm
was not observed in the phenol endcapped higher Mn polymer, 4a. In addition, the
exotherm was not seen in the 2nd heats of either uncapped samples 1a or 1b (Figure 3.6).
The fact that the exotherm peak area appeared to correlate with theoretical end group
concentrations, coupled with the absence of a peak in the 2nd
heats appeared to indicate
that an irreversible end-group related exothermic reaction was taking place on heating
above 275 °C.
144
144.4°C(H)
126.9°C(H)
117.3°C(H)
-4
-2
0
2
4H
ea
t F
low
(W
/g)
50 100 150 200 250 300 350Temperature (°C)
Polymer 1a. 6,800 Mn––––––– Polymer 4a, Mn 22,400 Phenol cap– – – – Polymer 1b. 23,400 Mn––––– ·
Exo Up
Figure 3.6. 2nd
Heats of unendcapped polymers, with phenol capped polymer for
comparison.
The Tg of heat-treated polymers, as measured from the 2nd
heat DSC scans,
showed an inverse relationship with Mn (Figure 3.6). Since Tg generally increases with an
increase in cross-link density22
, this is another possible indication of cross-linking. The
magnitude of the Tg shift indicated a higher cross-link density correlation with higher end
group content (lower Mn=more endgroups). The phenol capped sample, which should
have the lowest percentage of reactive endgroups due both to molecular weight and
capping with an inactive functional group, exhibited the lowest Tg after thermal cycling,
as expected. Note that 1a had a lower Tg than 4a and 1b on first heating (Figure 3.4).
145
The uncapped samples were insoluble in THF after heating to 400 °C in the DSC
test. This was further confirmation that cross-linking reactions had occurred. The phenol
endcapped polymer dissolved as expected in the absence of any significant cross-linking.
The combination of the exotherm at 275 °C seen in the first heat, the Tg increase which
tracked with relative endgroup content, and the insolubility of uncapped polymers after
heating was strong evidence that uncapped samples cross-linked via a reactive endgroup
mechanism, whereas the phenol endcapped polymer did not react to any appreciable
extent.
Samples 1a (6,800 Mn) and 1b (22,000Mn) were tested via parallel plate
rheometry at 300 °C under a nitrogen atmosphere. A sample of 4a (22,000 Mn) phenol
endcapped polymer was also tested for comparison. The testing temperature of 300 °C
was selected to avoid cross-linking of ethynyl endcaps, which typically occurs at
temperatures of 320 °C to 350 °C23
, and focus solely on reactive end group cross-linking,
i.e. the chlorosilane ends of the polymer. Qualitative comparison of viscosity changes
on heating over time is a common method of monitoring the curing of thermosets24
.
Application of this model for the study of cross-linking of end groups is a valid extension
of this methodology. Higher viscosity changes should correlate with relative
concentrations of reactive endgroups, and thus provide a robust probe of endcapping
effectiveness.
146
0 5.0 10.0 15.0 20.0 25.0 30.0time (min)
0
10000
20000
30000
40000
50000
60000
70000
80000
|n*|
(P
a.s
)
1a (Mn=6,800)
4a (Phenol Cap)
1b (Mn=22,000)
Figure 3.7. Rheogram of complex viscosity vs. cure time at 300 °C for samples 1a, 1b,
and 4a.
The rate of change of the complex viscosity of the lower Mn sample 1a was much
higher than that of the high Mn sample 1b. Again, this would be expected if more reactive
end groups were present (more reactive end groups = faster rate of reaction). The final
complex viscosity of 1a was 4.7 times higher than that 1b after heating for 30 minutes.
Note that 4b, the higher molecular weight phenol endcapped sample, had an order of
magnitude lower viscosity change than the other samples. This was a strong indication
that the phenol endcapping was successful.
147
Though not understood yet, it has been reported by several researchers that silanol
groups react with phenyl substituted silane in a cross-linking reaction at high
temperatures via Scheme 3.5 13
:
SiR
R
R
OH + SiO OR R
∆
+
Si
R
RR
O
SiO OR R
Scheme 3.5. Possible silanol thermally induced cross-linking reaction with phenyl
substituted siloxane.
This reaction, though not experimentally confirmed, is consistent with the cross-
linking behavior observed in the uncapped polymers. Curing via this mechanism is not
desired, as it is relatively uncontrollable, results in the production of benzene, an
undesired volatile cure byproduct (boiling point ~80 °C). This reaction also occurs at
lower temperatures than ethynyl capped polymer cross-linking reactions, and thus could
interfere with ethynyl reactions due to the marked viscosity increase shown in Figure
3.7..
It was also observed after testing was completed that all uncapped samples
exhibited a rubbery consistency at temperatures above Tg (>150 °C). The samples
stretched when removing them from the parallel plate test fixture. The stretched polymers
148
would then contract once stress was released if T > Tg. This was a classical “rubbery”
response, typical of a lightly cross-linked elastomer subjected to an applied stress at a
temperature above Tg 25
. Once the materials cooled below Tg, they exhibited behavior
more typical of glassy materials. The samples were stiff and brittle, with very limited
deformation before fracturing or bending irreversibly. The phenol capped polymer had
characteristics more typical of a molten thermoplastic heated above its Tg, with
irreversible deformation and viscous melt stretching seen when stress was applied to the
sample. The divergent behaviors exhibited by the uncapped polymers and the phenol
capped polymer showed that phenol capping inhibited thermal cross-linking. It was also
an indication of the limitations of a non-cross-linked aryloxysilane, which would have
limited functional utility above Tg due to its viscous, molten state. These findings further
validated the exploration of controlled cross-linking to enhance the utility of this polymer
system.
3.9. Characterization of lithiumphenylacetylide endcapped polymer
3.9.1. Molecular weight
GPC weight analysis showed that polymerization continued in the uncapped
solution, while molecular weight building ceased in the capped solution.
3.9.2. Spectroscopic characterization
FTIR analysis of the capped polymer showed the presence of a weak sharp
ethynyl band at 2,159cm-1
(Figures 3.8 and 3.9). This peak was not observed in the FTIR
spectrum of the uncapped polymer. 13
C NMR analysis showed the presence of ethynyl
peaks at 108 and 89 ppm (Figures 3.10 and 3.11). The Si C Cpeak has a chemical
149
shift of 108-109 ppm, while a C C ipso to an aromatic carbon has a chemical shift of
88-89 ppm. These peaks are characteristic of aromatic ethynyl silanes26-28
. The presence
an ethynyl silane peak at 109 ppm in particular was strong evidence that the silicon
ethynyl adduct had formed. Neither of these peaks was observed in the 13
C spectrum of
the uncapped polymer.
4500 4000 3500 3000 2500 2000 1500 1000 500
74.4
83.7
93.0
102.3
Uncapped Polymer
wavenumber, cm-1
4500 4000 3500 3000 2500 2000 1500 1000 500
74.4
83.7
93.0
102.3
% T
ran
sm
itta
nce
Lithiumphenylacetylide Capped Polymer
Figure 3.8. Overlay of capped and lithiumphenylacetylide capped polymers (full scale).
150
2500 2000
93.6
96.0
98.4
100.8
Uncapped Polymer
wavenumber, cm-1
2500 2000
93.6
96.0
98.4
100.8
% T
ransm
itta
nce
Lithiumphenylacetylide Capped Polymer
2159cm-1
Figure 3.9. IR spectra of lithiumphenylacetylide capped and uncapped polymers,
showing the characteristic ethynyl peak at 2,159 cm-1
.
151
200 150 100 50 0
0.00E+000
8.90E+008
1.78E+009
2.67E+009
Arb
. u
nits
Chemical Shift (ppm)
Lithiumphenylacetylide Capped Polymer
Uncapped Polymer
200 150 100 50 0
0.00E+000
8.90E+008
1.78E+009
2.67E+009
Arb
Un
its
Figure 3.10. 13
C NMR spectrum of lithiumphenylacetylide capped and uncapped polymer
(full scale).
152
110 100 90
-1.00E+008
0.00E+000
Arb
. u
nits
Chemical Shift (ppm)
Lithiumphenylacetylide Capped Polymer
Uncapped Polymer
110 100 90
-1.00E+008
0.00E+000A
rb U
nits
Figure 3.11. 13
C NMR spectra of lithiumphenylacetylide capped and uncapped polymer,
showing ethynyl peaks at 89 and 109 ppm in the capped sample.
3.9.3. Thermal Analysis
An exothermic peak was present at 321 °C in the first heat, but was not observed
in the 2nd
heat (Figure 3.12). Ethynl containing compounds typically exhibit exothermic
peaks during heating at temperatures of 320-350 °C29-30
. The presence of this peak in 3a
is further confirming evidence of the presence of ethynyl groups in the sample.
153
Figure 3.12. DSC comparison of first and second heats of 2a, lithiumphenylacetylide
capped polymer, showing the expected ethynyl cure peak at 324 °C.
Rheological stability was determined at 300 °C under an inert nitrogen
atmosphere. 300 °C was chosen as the test temperature so as to avoid curing the polymer
via ethynyl cross-linking reactions. Thus thermal stability at this temperature should
mainly be a function of reactive endgroup concentration: lower reactive endgroup
concentration should yield a more stable melt.
324°
First heat
Second heat
-
-
-
-
-
-H
eat
Flo
w
(W/g
)
250 270 290 310 330 350 370 390Temperature Exo Up
154
0 500.00 1000.0 1500. 2000.
time (s)
0
10000
20000
30000
40000
50000
60000
70000
80000|n
*| (
Pa.s
)
(Pa
.s)
Lithiumphenylacetylide
capped 2a ( Mn=7,700)
Uncapped Polymer 1a
(Mn=6,800)
Figure 3.13. Comparison of viscosity change on heating of lithiumphenylacetylide
capped vs. uncapped polymer.
The lithiumphenylacetylide capped polymer 2a showed a much slower change in
viscosity over time relative to an uncapped polymer with comparable molecular weight,
1a (Figure 3.13). This was another indication that endcapping had occurred, as the
phenylethynyl cap should be more thermally stable at 300 °C than the reactive endgroup
of an uncapped polymer.
3.10. Characterization of (4[(4-fluorophenylethynyl)]phenol) endcapped polymer
3.10.1. Molecular weight
Molecular weight was found to stop building at Mn of 6,500 in the capped system,
whereas the remaining uncapped portion of the reaction continued to build molecular
weight. The uncapped system resulted in an Mn roughly double that of the 4[(4-
155
fluorophenylethynyl)] phenol capped material. These results indicated that 4[(4-
fluorophenylethynyl)]phenol acted as a chain stopper to inhibit continued chain growth,
and did in fact endcap the polymer as expected.
3.10.2. Spectroscopic Characterization
Since the ethynyl band in the 4[(4-fluorophenylethynyl)] moiety is symmetric and
a very weak absorber in conventional IR spectroscopy, Raman spectroscopy was used to
verify the presence of the endcap in our polymer system. The Raman spectrum of the
4[(4-fluorophenylethynyl)]phenol endcapping reagent showed a strong absorption band at
2225cm-1
, which corresponds to a stretching mode for ethynyl groups bonded to
aromatic rings31
. This peak was also observed at reduced intensity in the
fluorophenylethynyl capped polymer, but was not observed in an uncapped polymer
sample (Figures 3.14 .and 3.15).
156
3520 2640 1760 8800
3300
6600
99003520 2640 1760 880
2400
4800
7200
96003520 2640 1760 880
0
3300
6600
9900
Arb
. u
nits
Uncapped Polymer
Fluorophenylethynylphenol Reagent
Fluorophenylethynylphenol Capped Polymer
Figure 3.14. Raman spectrum of fluorophenylphenol capped polymer, with spectra of
fluorophenylphenol reagent and uncapped polymer overlaid for comparison (full scale).
157
2366 2275 2184 2093
1080
1440
1800
21602366 2275 2184 2093
2400
4800
7200
96002366 2275 2184 2093
1080
1440
1800
2160
Arb
. u
nits
Uncapped Polymer
Fluorophenylethynylphenol Reagent
Fluorophenylethynylphenol Capped Polymer
2225cm-1
Figure 3.15. Raman spectrum of fluorophenylphenol capped polymer, with spectra of
4,4’fluorophenylphenol reagent and uncapped polymer overlaid for comparison (zoom).
Due to the short-range symmetry of the 4[(4-fluorophenylethynyl)]phenol
molecule, C C ethynyl 13
C NMR spectra exhibit only a slight peak splitting, depending
on substituents on the attached benzene rings32
. A pure 4[(4-fluorophenylethynyl)]
phenol reagent sample showed ethynyl peaks at 88.8, and 88.3 ppm. Peaks in the 4[(4-
fluorophenylethynyl)]phenol capped polymer were observed at 88.7 and 88.4 ppm, but
were not present in the uncapped polymer (Figure 4.16 and 4.17). Since 4[(4-
fluorophenylethynyl)]phenol is highly soluble in methanol, the precipitating solvent for
the endcapped polymer, it is unlikely the starting reagent was present merely as
158
component in a physical mixture. All the spectroscopic results suggest that the
endcapping has been successful.
200 150 100 50 0
0.00E+000
1.40E+007
2.80E+007
4.20E+007
Chemical Shift (ppm)
Arb
. u
nits
200 150 100 50 0
-1.00E+009
-5.00E+008
0.00E+000
5.00E+008
1.00E+009
Arb
Units
4,4' Fluorophenylethynylphenol Capped Polymer
Polymer No Capping
Figure 3.16. 13
C NMR spectrum of fluorophenylethynylphenol capped polymer with an
uncapped polymer spectrum overlaid for comparison.
159
90 85 80
-500000
0
Chemical Shift (ppm)
Arb
. un
its
90 85 80
0.00E+000A
rb U
nits
4,4' Fluorophenylethynylphenol Capped Polymer
Polymer No Capping
Figure 3.17. 13
C NMR spectrum of fluorophenylethynylphenol capped polymer with an
uncapped polymer spectrum overlaid for comparison, enlarged to show ethynyl carbons
in the capped polymer.
3.10.3. Thermal Analysis
An exothermic peak was clearly seen in the first heat of this sample, but was
absent in the 2nd
heat (Figure 3.18). The presence of this peak in the 320 to 350 ºC range
is consistent with curing of an ethynyl group. After heating in the DSC test, this sample
did not dissolve in THF.
160
355°C
-2.5
-2.0
-1.5H
ea
t F
low
(W
/g)
250 270 290 310 330 350 370 390Temperature (°C)
1st Heat2nd Heat
Exo Up
Figure 3.18. DSC thermogram of polymer 3c 4[(4-fluorophenylethynyl)] capped
polymer, enlarged to show the exothermic peak at 355 °C
Rheological characterization showed a clear enhancement in stability during
prolonged heating at 300 °C for a 4[(4-fluorophenylethynyl)] capped polymer, relative to
a polymer with similar molecular weight that was not endcapped (Figure 3.19). This is
further supporting evidence that endcapping was successful, as the fluorophenylethynyl
end group will not cross-link to any large degree at these temperatures, and thus acts as
inert capping agent. Heating above 330 °C should initiate cross-linking of the ethynyl
groups, as evidenced by the strong exotherm in Figure 3.18.
161
0 500.00 1000.0 1500.0 2000.0time (s)
0
10000
20000
30000
40000
50000
60000
70000
80000
|n*|
(Pa.s
)
)1a Uncapped polymer (Mn=6,800)
3a Fluorophenylethynl-phenol capped polymer (Mn=6,500)
Figure 3.19. Rheogram comparing the changes in complex viscosity on heating of 4[(4-
fluorophenylethynyl)] capped polymer, 3a, relative to a comparable molecular weight
uncapped polymer 1a.
3.11. Comparison of physical properties of endcapped aryloxysilanes with
dimethylsiloxane and FFKM
Once the capped aryloxysilanes are cured, the shear modulus can be directly
compared to those of other elastomers reported in the literature. A comparison of the
shear moduli of these materials are shown in Figure 3.20. Note that the moduli of these
materials are within the same range as those of FFKM and dimethylsiloxane, thus
indicating that the capped aryloxysilanes should be elastomeric in nature.
162
100,000
16,000 300
12,000
264,000
0
50,000
100,000
150,000
200,000
250,000
300,000
Shear Modulus, Pa
Figure 3.20 Shear modulus comparison of polydimetyhylsiloxane33
, FFKM34
, and capped
aryloxysilanes when tested above their respective Tgs.
Note that FFKM modulus was calculated from the elastic modulus using the relationship
)1(2 ν+=
EG
Where E= tensile modulus, ν=Poisson ratio (0.5 for rubbers), and G= shear modulus35
.
3.12. Conclusions
Successful endcapping of aryloxysilanes with phenylethynyl functional groups
was achieved utilizing two different reagents: lithium phenylacetylide and 4[(4-
fluorophenylethynyl)]phenol. Endcapping was verified by several spectroscopic and
163
thermal methods. Phenol endcapping was also successfully performed, as verified by
thermal analyses. Shear moduli of capped materials were of the same magnitude as
dimethylsiloxane and FFKM elastomers, thus demonstrating in principle that the
materials should have mechanical properties comparable to commercial elastomers.
Thermal characterization of samples of uncapped aryloxysilane polymers showed
trends in cure exotherms, viscosity changes over prolonged heating, and solubility
changes after heating that were indicative of reactive end group chemistry. Thermally
cured polymers exhibited rubber-like behavior, with pliability observed when the
polymers were heated above their glass transition temperatures. These results verified the
need for controlled endcapping for enhanced thermal stability in phenoxysilane polymers
obtained from chlorosilane monomers.
Results of this work also demonstrated that rheological testing can be used as an
effective screening tool to determine if an endcapping reaction was successful by
evaluating viscosity changes at temperatures above the curing temperature of the reactive
end groups of the unmodified polymer, but below the temperature of curing of the
thermally cross-linkable endcapping moiety.
3.13. References
(1) Pater, R; Curto,P., Advanced Materials for Space Applications. Acta Astronautica
2007, (61), 1121.
(2) Mangalgiri, P.D. Polymer-matrix composites for high-temperature applications.
Defence Science Journal 2005, 55, (2), 175-193.
164
(3) Murari, A; Barzon,A. Ultra high vacuum properties of some engineering polymers.
IEEE Transactions on Dielectrics and Electrical Insulation 2004, 11, (4) 613-619.
(4) Allcock, H. R.; Lampe, F. W. In Contemporary Polymer Chemistry. Second Edition;
Prentice Hall: Englewood Cliffs, NJ 1990. 236.
(5) Carraher, C.; Pittman, C. In Inorganic Polymers; Ullmann's Encyclopedia of
Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Darmstadt,
Germany: 2000.
(6) Jones, R. G.; Ando, W.; Chojnowski, J. Silicon-Containing Polymers - The Science
and Technology of Their Synthesis and Applications. Springer - Verlag: Vienna,
Austria: 2000, 201.
(7) Haas, K.H.; Wolter,H., Synthesis, properties and applications of inorganic–organic
copolymers (ORMOCERs). Current Opinion in Solid State and Materials Science
1999, (4), 571-580.
(8) MacKenzie, J., Structures, Properties and Potential Applications of Ormosils. Journal
of Sol-Gel Science and Technology, 1998, 13, (3) 371-377.
(9) Simoes, J. ; Beauchamp, J., Transition metal-hydrogen and metal-carbon bond
strengths: the keys to catalysis, Chemical Reviews. 1990, (90), 629-688,
(10) Dunnavant, W. R.; Markle, R. A.; ; Sinclair, R. G.; Stickney, P. B. ; Curry, J. E.;
Byrd, J. D., p,p Biphenol Dianilosilane Condensation Polymers. Macromolecules
1968, (1), 249-254.
(11) Curry, J., Method of Producing Alternating Ether Siloxane Polymers.U.S. Patent
3346515, 1967.
(12) Meier, D.; Lee, M., Synthesis and properties of diarylsiloxane and
(aryl/methyl)siloxane polymers: 1. Thermal properties. Polymer 1993, 34, (23),
4882-4892.
(13) Peters, E. N., The development of carborane-siloxane polymers. Industrial &
Engineering Chemistry Product Research and Development 1984, 23, (1), 28-32.
(14) Steffen, K. ,Poly- dioxyarylen(dioxycycloalky1en)-diphenylsilane II.
Molekulargewichte, Erweichungstemperatuen, Thermostabilitaten. Die Angewandte
Makromolekulare Chemie 1972, 24, (335), 21-33.
165
(15) Steffen, V.K.-D, Poly-dioxyarylen(dioxycycloalky1en)-diphenylsilane I. Synthese
durch Polykondensation von Dihydrido-diphenylsilanen und anderen, teilweise
neuen bifunktionellen Diphenylsilanen mit Diphenolen bzw. Cycloalkylendiolen.
Die Angewandte Makromolekulare Chernie 1972, 24, (335), 1-20.
(16) Dvornic, P. R., Degradative side reactions in the syntheses of exactly alternating
silarylene-siloxane polymers. Polymer Bulletin 1992, 28, (3), 339-344.
(17) Dunnavant, W. R.; Markle, R. A.; Sinclair, R. G.; Stickney, P. B. Second Annual
Summary Report on Process Development and Pilot-Plant Production of Silane
Polymers and Diols. Batelle Memorial Institute: Columbus, OH 1967. 277.
(18) Scola, D., ed. Polyimide Resins. In ASM Handbook: Volume 21 Composites , ed. S.
Donaldson, S.; Miracle,D ., Editors. ASM International : Material Park, OH 2001,
105-119.
(19) Brown, W.: Foote, C., Organic Chemistry. Harcourt Brace & Co: Orlando,
FL1998,335.
(20) Dhara, M.G.; D. Baskaran, D.; Sivaram,S.Effect of LiClO4 and LiCl additives on
the kinetics of anionic polymerization of methyl methacrylate in toluene-
tetrahydrofuran mixed solvent. Macromolecular Chemistry and Physics, 2003, 204,
(12),1567-1575.
(21) Homrighausen, C. L.; Keller, T. M., Synthesis and Characterization of silarylene-
siloxane diacetylene Polymer and its Conversion to Thermosetting Plastic. Polymer
2002, (4), 2619-2623.
(22) Moita, I. and S. Akiyama, Molecular Design of Network Polymers. In
Macromolecular Design of Polymeric Materials, H. K., T. Kitayama, and O. Vogl,
Ed.. Marcel Dekker: New York, NY 1997. 394-396.
(23) Bucca, D.; Keller, T. M., Thermally and oxidatively stable thermosets derived from
preceramic monomers. Journal of Polymer Science Part A-Polymer Chemistry.
1997, 35, (6), 1033-1038.
(24) Crompton, T.R., Monitoring Resin Cure. In Polymer Reference Book, S. Shawbury,
Shropshire, Ed.., Rapra Technology Limited:Billingham, UK 2004, 385.
(25) Nielsen, L.E., Mechanical behavior of some lightly crosslinked rubbers. Journal of
Applied Polymer Science, 1964, 8, (1), 511-520.
166
(26) Ichitani, M.; Yorezaya, K.; Okada, K.; Sugimoto, T.,Silyl-carborane hybridized
diethynylbenzene-silylene polymers Polymer Journal, 1999, 31, (11), 908-912.
(27) Palitzsch, W.; Nitsche, S.; Seichter, W.; Weber,E.; Roewer, G.,New aryl ethynylene
substituted silicon-centered molecules. Silicon Chemistry, 2003 ,(2), 33-44.
(28) Ohshita, J.; Taketsugu, R.; Nakahara, Y.; Kuna,A.,Convenient synthesis of
alkoxyhalosilanes from hydrosilanes. Journal of Organometallic Chemistry, 2004,
689,(20), 3258-3264.
(29) Sundar, R.A.; Keller, T.,Linear diacetylene polymers containing bis(dimethylsilyl)
phenyl and/or bis(tetramethyldisiloxane)carborane residues: Their synthesis,
characterization and thermal and oxidative properties. Journal of Polymer Science
Part A-Polymer Chemistry, 1997. 35,(12),2387-2394.
(30) Hergenrother, P. M., The Use, Design, Synthesis, and Properties of High
Performance/High Temperature Polymers: An Overview. High Performance
Polymers 2003, 15, (1), 3-45
(31) Mohammed, A.; Minaev, B.; Agren, H.; Lindgren, M.; Norman,P., Classification of
Raman active modes of platinum(II) acetylides: A combined experimental and
theoretical study. Chemical Physics Letters, 2009, 481, (4-6), 209-213.
(32) Suzuki, T.; Mita, I . ,Synthesis and Properties of Silethynylene siloxane Alternating
Copolymers. European Polymer Journal, 1992, 28,(11), 1373-1376.
(33) Enami, H,; Hamada, Y.; Nakamura, A.; Saiki, T. Silicone gel composition and
silicone gel for use in sealing and filling of electrical and electronic parts. US Patent
6001943, 1999.
(34) Technoflon PFR 04HTS Perfluoroelastomer Datasheet. http://www.ides.com
(35) Mitchell, B. S. Mechanics of Materials, In An Introduction to Materials
Engineering and Science: For Chemical and Materials Engineers, John Wiley &
Sons, Inc., Hoboken, NJ, 2004.
167
Chapter 4: Novel Diacetylphenoxysilane Polymers
4.1. Introduction
Ethynyl groups are commonly used as cross-linking sites in high performance
polymers. Typically they are introduced as end groups on oligomers1-5
. Several
researchers have attempted to introduce diethynyl/diacetylene groups into the backbone
itself for subsequent cross-linking through thermal or other means. The most commonly
used methods of polymerization involve either in-situ generation of diacetylene groups
by conversion of hexachlorobutadiene through reaction with n-butyl lithium 6, 7
, or
coupling of pendant acetylene groups or acetylene end groups with catalysts8.
In situ reactions require complex chemical reactions to take place immediately
prior to the polymerization reaction in the same reaction vessel. This poses two issues:
small changes in theoretical yields could have significant impact on molecular weights
through stoichiometric imbalances, and residual n-butyl lithium, a very strong base, could
potentially result in actual breaking of the polymer backbone9.
Post polymerization couplings require use of organometallic catalysts, and also
have the disadvantage of slower reaction kinetics due to higher viscosity of polymer
solutions relative to solutions of monomeric reactants. In addition, most systems couple
the acetylenes directly to silicon atoms which are usually methyl substituted. Methyl
substituted silicones are known to be less thermally stable than phenyl substituted
silicones10
.
In order to produce a cross-linkable aryloxysilane with high thermal stability with
ethynyl functional groups in the polymer backbone, a system was developed which
utilized diacetylenes in monomeric form, namely 4,4'-buta-1,3-diyne-1,4-diyldiphenol.
168
This monomer was reacted with dichlorodiphenylsilane in the presence of triethylamine
to form a fully phenyl substituted polymer with a diethynyl functionality in the backbone.
Several synthetic routes are possible for the coupling of a diphenylsilane with a diol11-13
.
The synthetic reaction chosen to make this polymer is shown in Scheme 4.1. The THF
condensation solution reaction was chosen because it is performed under the mildest
conditions (65-70 °C), the aromatic diol and TEA are fully soluble in the reaction
solution, and the condensation byproduct, HCl, precipitates out as a salt to push
equilibrium to the product side. In addition, no catalysts were required, which minimized
the possibility of unintended addition reactions of the diacetylene linkages, and also
eliminated the need for removal of catalysts after polymerization.
+
SiCl ClOH OH O O Si
n
THF,
TEA
60 - 70 °C
N2
Scheme 4.1. Polymerization reaction of 4,4'-buta-1,3-diyne-1,4-diyldiphenol and
dichlorodiphenylsilane.
4.2. Experimental Section
4.2.1 Materials
Dichlorodiphenylsilane (DCDPS), 99% purity, was purchased from Gelest
(catalog number SID4510.1); 4,4'-buta-1,3-diyne-1,4-diyldiphenol (BDDP), 99% purity,
was purchased from PolySciences (catalog number 13832-0); triethylamine (TEA), 99%,
(catalog number 15791) was purchased from Acros; anhydrous tetrahydrofuran (catalog
169
number 401757) was purchased from Sigma Aldrich; methanol was purchased from
Sigma Aldrich.
4.2.2. Polymer Synthesis
Polymers were synthesized by polycondensation of DCDPS with 4,4'-buta-1,3-
diyne-1,4-diyldiphenol in anhydrous, unstabilized THF with an inert atmosphere purge
under reflux at 65-70 °C, as per conditions outlined in Chapter 2, Section 2.2.2.8. TEA
was used as an acid scavenger to remove the condensation byproduct HCl from the
system.
As a typical synthesis, in a sealed 200ml three neck round bottom flask equipped
with a magnetic stir bar, a reflux condenser, and a nitrogen purge, 4.68 grams
(0.020moles) of 4,4'-buta-1,3-diyne-1,4-diyldiphenol and 4.25 grams (0.042 moles) of
triethylamine was added to 30 ml of THF. The THF was transferred from a sealed
reagent bottle to the reactor via 100 ml syringe. The 4,4'-buta-1,3-diyne-1,4-diyldiphenol
and TEA were fully soluble in THF. The solution was heated to 70 °C in an oil bath on a
magnetic stirrer. The heating bath was heated with a heating mantle equipped with a
thermocouple controller, which allowed precise control of the heating bath temperature.
To this solution, 5.3 grams (0.021 moles) of DCDPS along with 30 ml of anhydrous THF
was added via an addition funnel with an inert gas pressure equalizer. The DCDPS was
added dropwise over approximately half an hour. Formation of white solids, presumably
TEA:HCl salts, were noted soon after addition of DCDPS. This was an indication that the
reaction was proceeding as planned, as TEA:HCl salt is the condensation byproduct of
this reaction. After DCDPS addition was completed, the solution was heated under reflux
for an additional 2.5 hours. Temperature was maintained at 65-70 °C for the remainder of
170
the reaction. The reaction solution was precipitated in methanol, filtered, and dried in a
vacuum oven at 100 °C overnight. Yields after precipitation, filtration and drying were in
the 90% range.
4.3. Characterization
Molecular weight
Molecular weights were determined via gel permeation chromatography. A
Waters LC 1515 (Waters Corp., Milford, MA) equipped with a Phenogel 106A GPC
column (Phenomenex, Torrance, CA) and a Waters 2414 refractive index detector.
Elution rate was maintained at 1 mL/min. THF was used as the mobile phase. Polystyrene
standards were used for the molecular weight calibrations.
Spectroscopy
Infrared spectra were collected on a Perkin Elmer (Perkin-Elmer Co., Norwalk,
CT) Spectrum One spectrometer. Spectra of solid samples were collected using an
Attenuated Total Reflectance accessory.
Raman spectra were collected on a Renishaw RM1000 Raman microscope
(Renishaw Plc., Gloucestershire, UK). Samples were dissolved in THF and mounted
between two glass slides. A laser with a 513nm wavelength was used as the excitation
source.
13C NMR spectra were collected using a Varian Unity Inova 300MHz NMR
(Varian Inc., Palo Alto, CA). Samples were dissolved in deuterated DMSO. Scans were
collected in proton decoupled mode. Chemical shifts were calibrated using the deuterated
DMSO solvent peaks as an internal reference.
Thermal analysis
171
Differential scanning calorimetry thermograms were collected using a TA
Instruments Q100 DSC (TA Instruments, New Castle, DE) with Modulated DSC
capability. Samples were analyzed in hermetic pans under a nitrogen atmosphere. Pinhole
vents were introduced in the pans to allow any volatiles generated during heating to
escape, and also to remove residual oxygen which might otherwise be trapped in the pan
and participate in undesired oxidation reactions. Pinholes were also introduced into the
reference pans, for consistency. Heating rates of 5 °C per minute were used. Modulated
thermograms were collected in modulated mode with heat-only modulation.
Thermogravimetric analysis was performed on a TA Instruments Q50 TGA. Heating
rates of 5 °C per minute were used. Samples were analyzed in nitrogen and air
atmospheres.
Rheometric analysis was performed on a TA Instruments AR2000 Rheometer
with 8mm parallel plates. Continuous oscillation testing was performed using a 1 % strain
and a frequency of 1 Hertz. Testing was performed at temperatures from 120 °C to 300
°C under a nitrogen purge. Heating rates of 5 °C per minute were used. Samples were
preconditioned at the starting temperatures for one minute prior to starting the tests.
Approximate shear rate for these tests was 1x10-5
sec-1
.
4.4. Results and discussion
4.4.1 Molecular weight
Molecular weight of a representative polymer was in the 4,000 range (Figure 1).
The GPC chromatogram showed a symmetrical distribution, with no large shoulders due
to unreacted monomers. This indicated that the polymerization reaction proceeded as
172
planned, and the precipitation in methanol removed unreacted monomers. Molecular
weights were somewhat lower than those obtained in our laboratory under similar
conditions with DCDPS biphenol polymerization 14
. This may be an indication of either
slower kinetics with BDDP or possibly some stoichiometric imbalance due to impurities
in the monomer or a side reaction.
Figure 4.1. GPC chromatogram of BDDP:DCDPS polymer. Mw=4,600; Mn=2,000;
Polydispersity= 2.3.
4.4.2. Spectroscopic Characterization
FTIR spectrum of the BDDP monomer showed a weak ethynyl band at 2150 cm-1
(Figure 4.2). This is consistent with acetylinic in plane stretching. For example IR of a
related compound 1,4 diphenylbutadiyne shows an acetylinic stretching at 2152 cm-1
15
173
4000 3500 3000 2500 2000 1500 1000
75
80
85
90
95
100
Tra
nsm
issio
n
A (1/CM)
2300 2250 2200 2150 2100 2050 200095.0
95.5
96.0
96.5
97.0
97.5
98.0
Figure 4.2. FTIR of BDDP, showing the weak ethynyl stretching absorption at 2150 cm-1
.
.
Since BDDP is a highly symmetrical molecule that shows only weak ethynyl IR
absorptions, Raman analysis was also performed. Solution state Raman showed a strong
absorption band at 2215 cm-1
(Figure 4.3). This again was consistent with an ethynyl
stretching mode. Reported literature values of the related compound 1, 4-
diphenylbutadiyne, showed a Raman active ethynyl stretching band at 2214 cm-1
16
.
174
Figure 4.3. Raman spectra of BDDP and BDDP:DCDPS polymer, with THF blank
overlaid for comparison.
13
C NMR of BDDP monomer, showed carbon signals at 81.8 and 72.4 ppm.
These chemical shifts correspond well with the ethynylic carbons of 1,4
diphenylbutadiyne. Analysis performed in deuterated acetone by White et al. concluded
that peaks with chemical shifts of 81.7 ppm could be attributed to acetylinic carbons
bonded to the phenyl ring. A chemical shift of 74 ppm was attributed to the internal
acetylinic carbons17
.
2500 2450 2400 2350 2300 2250 2200 2150 2100 2050 2000
450
900
1350
1800
-18000
0
18000
36000
-18000
0
18000
36000
Abs.
cm-1
B
BDDP Reagent
Abs.
THF(Blank)
BDDP:DCDPS Polymer
Abs.
175
200 180 160 140 120 100 80 60 40 20 0
-5.00E+007
0.00E+000
5.00E+007
1.00E+008
1.50E+008
2.00E+008
2.50E+008
B
90 88 86 84 82 80 78 76 74 72 70
0.00E+000
5.00E+007
Figure 4.4. 13
C NMR of BDDP monomer, showing the acetylinic carbon signals at 81.8
and 72.4ppm.
FTIR analysis of the solid polymer showed a strong band at 900cm-1
, which is
consistent with Si-O-C6Hx stretching18
. This band was not observed in the IR spectrum of
the monomer. The formation of an aromatic siloxane bond through the condensation
reaction is the desired result of the synthesis. Therefore, the presence of this band is
confirmation that the condensation reaction had in fact taken place, as there are no Si-O-
C6Hx bonds in either of the reactants.
This spectrum also showed two weak ethynyl bands at 2215 cm-1
and 2150 cm-1
.
The presence of two weak bands in the polymer, whereas there was only one IR active
band in the reagent, may be an indication that the conformation of the phenyl groups
176
attached to the diacetylinic group are staggered rather than coplanar. Analysis of the IR
spectra of solutions of the model compound 1,4-diphenylbutadiyne showed two IR bands
in solution which the authors attributed to rotational isomers that were solvent
stabilized15
. In this case, since we are evaluating a solid polymer system, the attachment
of the diphenylsiloxane linkages to the diphenylbutadiyne may induce a staggered
arrangement in the solid phase.
Raman analysis of the polymer showed an acetylinic peak at 2215 cm-1
, which
was further supporting evidence of the presence of ethynylic functional groups in the
polymer.
4000 3500 3000 2500 2000 1500 1000
40
45
50
55
60
65
70
75
80
85
90
95
100
cm-1
Tra
nsm
itta
nce
2300 2250 2200 2150 2100 2050 200096
97
98
177
Figure 4.5. FTIR of solid polymer, showing the Si-O-C6H5 stretch at 904 cm-1
and the
acetylinic stretches at 2212 and 2150 cm-1
.
150 100 50 0
0.00E+000
2.00E+008
4.00E+008
Arb
. U
nits
ppm
84 82 80 78 76 74 72 70
Figure 4.6. 13
C NMR of DCDPS BDDP polymer, in deuterated DMSO. The inset graph
shows the acetylinic peaks.
13C NMR analysis of the polymer showed acetylinic carbon peaks at 81.8 and
72.3 ppm (Figure 4.6). Comparisons of the NMR spectra of the BDDP monomer and the
polymer (Figure 4.7) show additional bands in the aromatic region at 126 and 121.1 ppm
in the polymer that are not observed in the monomer. These peaks correspond to aromatic
carbons of phenyl groups attached to silyl ether19
.
178
160 140 120 100 80
0.00E+000
2.00E+007
4.00E+007
ppm
Arb
. U
nits
160 140 120 100 80
0.00E+000
2.00E+007
4.00E+007
DBBP:DCDPS Polymer
DBBP Reagent
Figure 4.7. Comparison of 13
C NMR of the BDDP reagent and the polymer, showing the
additional aromatic bands in the polymer sample.
Raman analysis of the polymer showed the acetylinic peak present at 2215 cm-1
, the same
position where the acetylinic peaks were seen in the DBBP starting monomer (Figure
4.3).
4.4.3. Thermal Analysis
An MDSC scan of the ethynyl monomer from 50 °C to 400 °C showed a very
strong exothermic peak centered at 210 °C (Figure 4.8).
179
210.2°C
199.6°C463.0J/g
-1
0
1
2
3
4H
ea
t F
low
(W
/g)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
Figure 4.8. MDSC thermogram of BDDP monomer, heat flow signal shown.
Since the formula weight of BDDP is 234.2 Daltons, the enthalpy of 463 J/g is
equivalent to 108 kJ/mol. This agrees fairly well with values reported in the literature for
enthalpies of 120 kJ/mol for thermally induced polymerization enthalpies of related non-
contiguous diethynyl monomers20,21
.
In addition, by measuring the exothermic peak temperature at different heating
rates (Figure 4.9), the activation energy can be estimated by using the following equation
180
first developed by Ozawa15
, based on the Arrhenius relationship and relative reaction
rates of thermosetting polymers.
)/1(567.0 peakT
RateREa
∆
∆×
−=
Where Ea=Activation energy, R= gas constant, and the second term is the slope from a
plot of log heating rate vs. 1/ Tpeak(K).
DSC scans of the neat reagent were performed at several heating rates ( Figure
4.9). When plotted as per Ozawa method the slope of the line of log heating rate vs. 1/
Tpeak(K) can be used to calculate an activation energy of 122 kJ/mol. These preliminary
results agree well with activation energies reported in the literature of 125 to 166 kJ/mol
reported for related non-contiguous phenylethynylbenzene compounds with multiple
ethynyl substituents22
.
181
273.3°C256.6°C
297.2°C
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
He
at F
low
(W
/g)
200 250 300 350 400Temperature (°C)
BDDP Polymer, 10C per min––––––– BDDP Polymer, 5C per min– – – – BDDP Polymer, 20C per min––––– ·
Exo Up
Figure 4.9. DSC scans of DCDPS:BDDP polymer at different heating rates, showing the
exothermic peak shift due to changes in heating rates.
Calculation of activation energies for the polymer using the Ozawa method
yielded estimated activation energy of 100 kJ/mol. These preliminary results indicate the
activation energy of the polymer may be slightly lower than the pure monomer. The
lower activation energy may be due to long range inductive effects of the siloxane bonds
on the acetylinic carbons, which may enhance reactivity. This effect has been seen in
other acetylinic systems, where substituents on phenyl rings attached to ethynyl groups
can reduce activation energies of the ethynyl polymerization4,21,22
. Future testing may
182
confirm the differences, if any, between activation energies of the monomer and the
polymer.
A comparison of the DSC heat flow signals of the BDDP monomer and the
polymer (Figures 4.8 and 4.10, respectively) show several interesting differences. The
exothermic peak for cross-linking occurs at a higher temperature in the polymer relative
to the BDDP monomer. In addition, the monomer shows a sharp peak with a maximum at
210 °C, while the polymer exotherm peak maximum is observed at 260 °C, with a
shoulder on the higher temperature side centered at 277 °C.
The exothermic enthalpies of neat BDDP and the BDDP:DCDPS polymer show a
close correspondence to what one would predict based on the theoretical content of
ethynyl monomer. For an alternating BDDP:DCDPS polymer, the repeat unit weight is
414.5 Daltons. The formula weight of the diacetylinic component of this alternating
copolymer is 232.2 Daltons. The weight percent of acetylinic component in the polymer
is 56% (232.2⁄ 414.5=56%). The ratio of the enthalpies of pure monomer to polymer is
53% (244 kJ/mol ⁄ 463 kJ/mol). This 53% enthalpy ratio compares well to the 56%
theoretical weight % ratio (6% relative error).
183
129.7°C(H)123.9°C
135.5°C182.6°C
256.9°C
277.0°C
257.8°C
228.2°C244.1J/g
-0.4
-0.3
-0.2
-0.1
0.0
0.1H
ea
t F
low
(W
/g)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
.
Figure 4.10. Standard DSC of BDDP:DCDPS polymer, with identified transitions.
When reversing and non-reversing signals are deconvoluted and plotted
separately, the Tg is much easier to observe (Figure 4.11). Also, the shift in the baseline
in the reversing heat flow signal is a clear indication of a shift in heat capacity due to
vitrification. Lower heat flow into the polymer indicates that the heat capacity of the
sample has been reduced. Reduction in heat capacity is commonly seen after polymers
are cross-linked, due to the reduced mobility of the chains in a cross-linked network,
relative to uncrosslinked linear polymers23
. A plot of reversing and non reversing heat
capacity , Cp (Figure 4.12) shows the reversing heat capacity is has been reduced by 40%,
184
from 2.5 J/g°C to 1.4 J/g°C. The temperature at half-height of the step change in heat
capacity T/1/2∆Cp corresponds to the vitrification point of the polymer24
.The T/1/2∆Cp for
this polymer was measured at 275 °C from the reversing Cp curve of Figure 4.12.
131.7°C(H)126.0°C
137.3°C-0.3
-0.2
-0.1
0.0
0.1
0.2
[ ] N
on
rev H
ea
t F
low
(W
/g)
– –
– –
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
Re
v H
ea
t F
low
(W
/g)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
Figure 4.11. MDSC thermogram showing separate reversing and nonreversing heat flow
curves. Note the prominence of the Tg in Figure 4.11 relative to Figure 4.10.
185
272.6°C(H)
265.2°C
280.0°C
239.3°C(H)
229.1°C
249.5°C
0.9
1.0
1.1
1.2
1.3
1.4
[ ] R
ev C
p (
J/(
g·°
C))
– –
– –
-2
-1
0
1
2
3N
on
rev C
p (
J/(
g·°
C))
0 50 100 150 200 250 300 350 400Temperature (°C)
Figure 4.12. MDSC plot of 1st heat, reversing and non-reversing heat capacities vs.
temperature for BDDP:DCDPS polymer. Note the rise in Reversing Cp signal between
100 and 150 °C, due to the sample passing through its glass transition prior to onset of
curing.
A plot of the 2nd
heat of the polymer (Figure 4.13) shows the cross-linked material
has vitrified and no longer exhibits a glass transition. In addition the heat capacity
showed a relatively smooth increase in Cp, with no major step changes. A small step was
observed at 275 °C, with a subsequent offset in the reversing Cp baseline. This was likely
due to some additional curing taking place during the 2nd
heat. The cured sample showed
186
no observable swell in THF after heating, which was another strong indication that
extensive cross-linking had taken place.
0
1
2
3
4
[ ] N
on
rev C
p (
J/(
g·°
C))
––
––
–
0.6
0.7
0.8
0.9
1.0
1.1
Re
v C
p (
J/(
g·°
C))
0 50 100 150 200 250 300 350 400Temperature (°C)
Figure 4.13. MDSC plot of 2nd heat, reversing and non-reversing heat capacities vs.
temperature for BDDP:DCDPS polymer.
TGA analysis of the polymer in nitrogen (Figure 4.14) showed it had excellent
thermal stability. The 5% weight loss temperature was 540 °C, which compares favorably
to other high temperature polymers25
. Weight loss through completion of curing was
<1%, , of which 0.4% was likely due to residual solvents, based on the weight loss
immediately after the sample temperature exceeded Tg.
187
This polymer also had a very high residue content of 82% at 800°C. This is
comparable to char yields of several preceramic polymers26
, where high residue content
correlates to better ceramic yields in the fully converted polymer.
540.5°C 95.0%
Residue:82.2%(6.9mg)
1.0%(0.1mg)
80
85
90
95
100
105
We
igh
t (%
)
0 200 400 600 800Temperature (°C)
Figure 4.14. TGA of DCDPS-BDDP polymer in nitrogen.
TGA analysis of the polymer in air (Figure 4.15) showed a slightly lower 5%
weight loss temperature, and a residue content of 15.7%. This value was very close to the
theoretical residue content for the polymer of DCDPS-BDDP, which is 15.9% (based on
conversion of Si to SiO2.).
188
521.5°C 95.0%
Residue:15.7%(1.9mg)
0
20
40
60
80
100
We
igh
t (%
)
0 200 400 600 800Temperature (°C)
Figure 4.15. TGA of DCDPS-BDDP Polymer in air.
Parallel plate rheology of the polymer showed behavior characteristic of a
thermally cross-linkable material. At 200 °C, approximately 70 °C above the Tg, the
polymer showed a higher loss modulus, G” than its storage modulus, G’. This is
typically seen when an amorphous polymer is tested above its Tg; the polymer exhibits
189
viscous characteristics, and shows irreversible deformation (flow) when subjected to
shear. As the polymer was heated and began to cross-link, G’ increased. The point where
G’=G”, or “crossover point” is commonly defined as the gel point, when the polymer has
converted from a collection of individual strands into a cross-linked network27
. The
crossover point for this system was observed at 242 °C (Figure 4.16).
190.0 200.0 210.0 220.0 230.0 240.0 250.0 260.0 270.0 280.0
temperature (°C)
1000
10000
1.000E5
1.000E6
1.000E7
1.000E8
G' (
Pa
)
1000
10000
1.000E5
1.000E6
1.000E7
1.000E8
G'' (P
a)
0
0.5000
1.000
1.500
2.000
2.500
3.000
3.500
tan
(de
lta
)
Figure 4.16. Temperature sweep of DCDPS BDDP polymer, in a nitrogen atmosphere.
Crossover point was 242 °C.
The viscosity of this material increased above the measurement capabilities of the
instrument at temperatures above the gel point, due to the extensive thermally initiated
cross-linking.
190
In order to investigate the properties of this material while still in an elastomeric
form rather than a fully cross-linked network, additional experiments were performed to
generate partially-cured material. Heating to temperatures above 240 °C resulted in a
heavily cross-linked material, so temperature selection to allow slower curing was
necessary.
Storage modulus of a material in a cross linking experiment is a parameter that
allows one to evaluate the extent of network formation. At the start of the experiment,
storage modulus is low, at the end of the experiment, storage modulus is much higher. In
addition the change in slope of the storage modulus as a function of temperature can also
be used to judge the onset of a curing reaction. This can be evaluated graphically by
plotting the derivative of G’ as a function of temperature (Figure 4.17).
191
200.0 202.5 205.0 207.5 210.0 212.5 215.0 217.5 220.0 222.5 225.0
temperature (°C)
1000
10000
1.000E5
G' (
Pa)
-0.050
0.075
d (lo
g (G
')) / d(te
mp
era
ture
) (Ce
ntra
l Diffe
rence)
Figure 4.17. Plot of log G’ and its first derivative d(logG’)/dt as a function of
temperature.
Like most amorphous polymers tested above their Tg, viscosity of the
DCDPS:BDDP polymer decreases as temperature is increased, until it reaches a
minimum at 205°C. As the temperature is increased and thermally induced cross-linking
begins to occur, G’ begins to increase. The first derivative, d logG’/dt also changes as
cross linking starts, with the steepest increase seen from 205 to 210 °C followed by a
very steady rate of increase from 209 to 225 °C. Based on this data, 210 °C was selected
as the target temperature to perform a cure cycle. The rate of viscosity increase in this
192
temperature range was slow, which indicated that the rate of cure was also slow at this
temperature. This indicated that partial curing should be possible.
Samples were heated isothermally at 210 °C for 30 minutes. G’, G”, and tan delta
were tracked as a function of time in order to investigate the thermosetting properties of
this material (Figure 4.18). The crossover point was measured at 4.8 minutes. In addition
when the sample was removed from the rheometer after completion of the test while still
at a temperature above its Tg, it exhibited a very rubbery consistency. The material
stretched when removed from the parallel plate fixture. Within a few minutes of cooling,
the material solidified and converted to a stiff, glassy material.
0 5.0 10.0 15.0 20.0 25.0 30.0
time (min)
1000
10000
1.000E5
1.000E6
G' (
Pa)
1000
10000
1.000E5
1.000E6
G'' (P
a)
0
0.2500
0.5000
0.7500
1.000
1.250
1.500
1.750
tan(d
elta
)
193
Figure 4.18. Isothermal parallel plate cure experiment, DCDPS:BDDP polymer.
In order to evaluate the extent of cure which took place during the isothermal test,
a DSC was performed on a sample after the rheology test. It was found that the Tg had
increased, but was still measureable. In addition, a reduced exotherm was also observed.
292.2°C
247.1°C180.4J/g
146.2°C(H)136.0°C
156.6°C
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
Hea
t F
low
(W
/g)
0 50 100 150 200 250 300 350 400Temperature (°C)
Exo Up
Figure 4.19. DSC curve of partially cured DCDPS:BDDP polymer.
194
The combination of the increase in Tg, the reduction in exothermic reaction
energy, and the observation of rubbery properties in the polymer at temperatures above
Tg confirmed that the polymer has been partially cured to generate an elastomeric
material.
The degree of cure can be estimated by comparison of the exothermic heats of
reaction of the sample being evaluated with a model sample that has been fully cured
using the equation below28
:
total
residualtotal
H
HH
∆
∆−∆=α
Where α= extent of reaction, also called degree of cure,∆Htotal= total exothermic heat of
reaction to produce a fully cured specimen, and ∆Hresidual= residual exothermic heat of
reaction for a partially cured specimen.
The thermal properties of uncured polymer, partially cured polymer, and fully
cured diethynyl polymer specimens are summarized in Table 4.1.
195
Table 4.1. Summary of thermal properties of uncured, partially cured, and fully cured
diethynyl polymer specimens.
4.5. Elastomeric properties of BDDP:DCDPS polymer
A comparison of the shear moduli of the BDDP:DCDPS polymers cured at 210
°C and 270 °C is instructive (Figure 4.20). Partial curing of the BDDP polymer results in
a shear modulus that is comparable to those of commercial polydimethylsiloxane29
and
FFKM30
elastomers. BDDP polymers cured at 270°C have much higher shear moduli,
due to the heavy cross-linking and exhibit properties more typical of cured thermoset
polymers. One can infer from this preliminary testing that the partially cured material
should have elastomeric properties when tested above its Tg.
Sample Tg(°C) Exotherm Area (J/g)
Normalized
Exotherm area Degree of
Cure
Uncured
polymer
129.7 244.2 100% 0%
Partially cured
polymer
146.2 180.4 74% 26%
Fully cured
polymer
No Tg (vitrified) 0 0% 100%
196
100,000
16,000 12,000
72,000
0
20,000
40,000
60,000
80,000
100,000
120,000
Shear Modulus, Pa7,700,000
Figure 4.20. Compsrison of shear moduli of BDDP polymers with those of
representative polydimethylsiloxane and FFKM polymers.
4.6. Summary
A new diacetylinic polymer was prepared through reaction of 4,4'-buta-1,3-diyne-
1,4-diyldiphenol and dichlorodiphenylsilane. Molecular weight analysis confirmed a
polymer had formed. FTIR, Raman, and 13
C NMR confirmed acetylinic functional
groups were present in the polymer. Thermal analysis was performed to characterize
thermosetting properties of this polymer. TGA showed the polymer had excellent thermal
stability and a high char yield. DSC analysis showed the polymer exhibited strong
exothermic transitions, with subsequent shifts in heat capacity and no detectable glass
197
transition after heating, features consistent with a thermal cross-linking. Rheological
testing showed significant viscosity increased on heating. Analysis of the data allowed
selection of a cure temperature that yielded a partially cross-linked material which
exhibited elastomeric characteristics, namely a low shear modulus and demonstration of
reversible stretching under deformation.
4.7. References
(1) Connell, J. W.; Smith, J. G.; Criss, J. M., High temperature transfer molding resins:
Laminate properties of PETI-298 and PETI-330. High Performance Polymers
2003, 15, (4), 375-394.
(2) Liu, H. B.; Simone, C. D.; Katiyar, P. S.; Scola, D. A., Adhesive properties of low-
viscosity phenylethynyl (PE) endcapped co-oligomides and co-oligomide blends.
International Journal of Adhesion and Adhesives 2005, 25, (3) 219-226.
(3) Kejian, W.; Ogasawara, T.; Ishida, Y., Optimization of liquid molding cycle for a
phenylethynyl terminated polyimide composite. Journal of Reinforced Plastics and
Composites 2006, 25, (4), 361-377.
(4) Sasaki, T.; Yokota, R. Synthesis and properties of an addition-type imide oligomer
having pendent phenylethynyl groups: Investigation of curing behavior. High
Performance Polymers 2006, 18, (2) 199-211.
(5) Chen, J.; Zuo, H.; Fan, L.; Yang, S. Synthesis and Properties of Novel Meltable
Fluorinated Aromatic Oligoimides Endcapped with 4-Phenylethynylphthalic
Anhydride. High Performance Polymers 2009, 21, (2) 187-204.
(6) Kolel-Veetil, M.;Beckham, H,; W Keller, T.M., Dependence of thermal properties on
the copolymer sequence in diacetylene-containing polycarboranylenesiloxanes.
Chemistry of Materials 2004, 16, (16), 3162-3167. .
(7) Homrighausen, C. L.; Keller, T. M., Synthesis and Characterization of silarylene-
siloxane diacetylene Polymer and its Conversion to Thermosetting Plastic. Polymer
2002, 43, (9), 2619.
(8) Boileau, S Bouteiller, L Kowalewska,,A. Polysiloxanes containing cross-linkable
diacetylene units in the main chain. Polymer International 2004, 53, (2), 191-197.
198
(9) Grzelka, A.; Chojnowski, J.; Cypryk, M.; Fortuniak, W.; Hupfield, P. C.; Taylor, R.
G. Polycondensation and disproportionation of an oligosiloxanol in the presence of
a superbase. Journal of Organometallic Chemistry 2002, 660, (1) 14-26.
(10) Dvornic, P.; Lenz, R. High Temperature Siloxane Elastomers; Huethig & Wepf,
Heidelberg, Germany: 1990, 325.
(11) Dunnavant, W. R.; Markle, R. A.; ; Sinclair, R. G.; Stickney, P. B. ; Curry, J. E.;
Byrd, J. D., p,p Biphenol Dianilosilane Condensation Polymers. Macromolecules
1968, (1), 249-254.
(12) Steffen, V.K.-D, Poly-dioxyarylen(dioxycycloalky1en)-diphenylsilane I. Synthese
durch Polykondensation von Dihydrido-diphenylsilanen und anderen, teilweise
neuen bifunktionellen Diphenylsilanen mit Diphenolen bzw. Cycloalkylendiolen.
Die Angewandte Makromolekulare Chernie 1972 24, (313), 1-20.
(13) Liaw, D. L., Synthesis and Characterization of Novel Polyaryloxydiphenylsilane
Derived From 2,2'- Dimethyl-biphenyl-4,4'-Diol. Journal of Polymer Science, Part
A: Polymer Chemistry 1999, 37, (24), 4591-4595.
(14) Drake, K.; Mukherjee, I.; Mirza, K.; Ji,F.;, Wei, Y. Phenylethynyl and phenol end
capping studies of polybiphenyloxydiphenylsilanes for cross-linking and enhanced
thermal stability. Manuscript to be submitted for publication.
(15) Nyquist, R. A.; Putzig, C. L,. Infrared and Raman study of 1,4-diphenylbutadiyne in
the solid state and in solution. Vibrational Spectroscopy 1992, 4,(1), 35-38.
(16) Zimmermann, B.; Baranović, G. Two-dimensional infrared correlation spectroscopic
study on thermal polymerization of diphenylbutadiyne. Vibrational Spectroscopy
2006, 41, (1) 126-135.
(17) White, D.; Levy, G., 13C Nuclear Magnetic Resonance Spectra of m-
Diethynylbenzene Polymers and Related Polyacetylenes. Macromolecules 1972,
5,(4), 526- 531.
(18) Curry, J. E.; Byrd, J. D., Silane polymers of diols. Journal of Applied Polymer
Science 1965, 9, (1), 295-311.
(19) Gädda, T. M.; Weber, W. P., Polydiphenylsiloxane polydimethylsiloxane
polydiphenylsiloxane triblock copolymers. Journal of Polymer Science Part A:
Polymer Chemistry 2006, 44, (11) 3629-3639.
(20) Back, S., Aerospace Organic Matrix Composite Materials Task Order 0005:
Complex Architecture and Analytical Methods −Synthesis and Cure Mechanism
Characterization of Phenylethynyl-Terminated Imide Oligomers. University of
Dayton Research Institute Technical Report UDR-TR-2009-132 2009, 26.
199
(21) Wang, F.; Xu, J.; Zhang, J.; Huang, F.; Shen, Y.; Du, L., Synthesis and thermal cure
of diphenyl ethers terminated with acetylene and phenylacetylene. Polymer
International 2006, 55, (9), 1063-1068.
(22) Sastri, S. B.; Keller, T. M.; Jones, K. M.; Armistead, J. P., Studies on Cure
Chemistry of New Acetylenic Resins. Macromolecules 1993, 26, (23), 6171-6174.
(23) Swier, S.; Van Assche, G.; Van Mele, B., Reaction kinetics modeling and thermal
properties of epoxy-amines as measured by modulated-temperature DSC. I. Linear
step-growth polymerization of DGEBA plus aniline. Journal of Applied Polymer
Science 2004, 91, (5) 2798-2813.
(24) Van Assche, G.; Van Hemelrijck, A.; Rahier, H.; Van Mele, B. Modulated
temperature differential scanning calorimetry: Cure, vitrification, and
devitrification of thermosetting systems. Thermochimica Acta 1997, 304-305, 317-
334.
(25) Hergenrother, P. M., The Use, Design, Synthesis, and Properties of High
Performance/High Temperature Polymers: An Overview. High Performance
Polymers 2003, 15, (1), 3-45.
(26) Riedel, R., Silicon-based polymer-derived ceramics: Synthesis properties and
applications - A review. Nihon Seramikkusu Kyōkai gakujutsu ronbunshi 2006,
114, (1330) 425-444.
(27) Mukherjee, I.; Drake, K.; Berke-Schlessel, D.; Lelkes, P. I.; Yeh, J.; Wei, Y. Novel
Thermally Cross-Linkable Poly[(arylenedioxy)(diorganylsilylene)]s Based on
Curcumin: Synthesis and Characterization. Macromolecules 2010, 43, (7) 3277-
3285.
(28) McHugh, J.; Fideu, P.; Herrmann, A.; Stark, W., Determination and review of
specific heat capacity measurements during isothermal cure of an epoxy using TM-
DSC and standard DSC techniques. Polym. Test. 2010, 29, (6) 759-765.
(29) Enami, H,; Hamada, Y.; Nakamura, A.; Saiki, T. Silicone gel composition and
silicone gel for use in sealing and filling of electrical and electronic parts. US Patent
6001943, 1999.
(30) Technoflon PFR 04HTS Perfluoroelastomer Datasheet. http://www.ides.com
200
Chapter 5: General conclusions and future research
5.1. Summary
Chapter 1 provided a review on the structural requirements for elastomers.
Structural requirements are related to the flexible links in the polymer backbone, along
with cross linking to yield a material that is useable above its Tg. Chapter 1 also reviewed
basic concepts of thermal stability of various polymeric linkages, and provided an
overview of current state of the art elastomers including FFKMS, FKMS, siloxanes, and
olefinic elastomers. Based on available thermal data, the conclusion was drawn that
hybrid organic-inorganic ethynyl cross-linked materials should in principle yield higher
thermal stability materials than those currently commercially available.
Chapter 2 presented synthetic methods which would yield aryloxysilane materials.
Many reactions were reviewed, and both hydroqinone and biphenol were evaluated as co-
monomers for this type of polymer system. Based on data generated during this research,
a dichlorodiphenylsilane–biphenol system prepared in toluene and THF gave the best
results.
Chapter 3 detailed the results of endcapping studies of dichlorodiphenylsilane-
biphenol polymers. New endcapped polymers were prepared utilizing
lithiumphenylacetylide and 4,4” fluorophenylethynylphenol as endcapping reagents.
Successful endcapping with was confirmed through chemical, thermal and rheological
analysis.
Chapter 4 detailed the synthesis and thermal characterization of a novel diethynyl
siloxane material, a copolymer of 4,4'-buta-1,3-diyne-1,4-diyldiphenol and
201
dichlorodiphenylsilane. Chemical, thermal and rheological testing confirmed the
monomer had reacted to form a polymer, which demonstrated the polymerization reaction
was successful. This polymer could be thermally crosslinked, and based on selection of
the cure temperatures and times, the material exhibited properties that ranged from
elastomeric to vitreous thermoset with no observed Tg. Based on the knowledge
generated during the course of these studies, and the novel materials made, there are
several avenues of research that could be explored in future work
5.2. Future studies
5.2.1. Endcapped materials
A logical next step for evaluation of endcapped material would be the synthesis of
the lithiumphenylacetylide and 4,4” fluorophenylethynlphenol end-capped polymers in
sufficient quantities for thermal curing and cross-linking. This would allow for
preliminary physical properties testing, which would provide further details as to the
utilities of these systems as high temperature cross-linked rubbers.
5.2.2. Diacetylphenoxysilane polymers
Further work could be performed to explore several different types of applications
of this polymer system. This material showed potential utility as an elastomeric material
when partially cured, a thermoset material when fully cured, and its high ceramic yield of
82% shows it could also be used as a preceramic polymer. In addition, the solubility of
this polymer, and the inclusion of ethynylinic moieties that could be thermally or UV
cured should allow for its use as a solvent cast coating.
202
5.2.2.1. Elastomeric material
The demonstration of elastomeric properties in a partially cured diethynyl
polymer shows that the material might be useful as an elastomer. Substitution of a portion
of the diethynyldiol with biphenol would allow tuning of the cross linking density for
specific applications. Variation of biphenol:diethynyldiol ratios would allow one to
synthesize materials that would range from lightly cross-linked (high biphenol content) to
heavily cross-linked (high diethynyl diol content). Future studies may also be performed
to evaluate relative reactivities of the diethynyl diol vs. biphenol via variation of feed
ratios and evaluation of actual monomer ratios in the polymer. This would allow for
better control of the monomer ratios incorporated into the polymer for obtaining targeted
cross-link densities for a particular application. Evaluation of mechanical properties of
these materials would allow future researchers to easily control mechanical properties via
simple variation of monomer stoichiometry, rather than utilizing other cure site
monomers for cross-linking.
5.2.2.2. Thermoset polymer
Given the high thermal stability of this polymer, the fact that its cure temperature
is 210° C to 260 °C, and the vitrification which yields a glassy material with no observed
Tg, its properties as a thermoset resin could be comparable or superior to current high
performance thermosets. Studies for evaluation of the physical and chemical properties of
this material when fully cured should show if the laboratory scale thermal properties
203
yield performance advantages over current state of the art high temperature thermosets
(epoxy, bismaleimide or polyimides).
5.2.2.3. Preceramic polymer
Synthesis of the polymer as detailed in Chapter 4, with 100% of the diol monomer
as the diethynyl species, resulted in a material with a pyrolysis yield of 82%. The high
ceramic yield is competitive with those of commercially available preceramic polymers.
Studies to evaluate this polymer as a preceramic material would be instructive. The
polymer could be mixed with other fillers such as ceramic fibers or whiskers, carbon
fibers, etc., and cured in a nitrogen atmosphere to form a ceramic material. The
properties of these composites might prove useful where current ceramic composite
performance might be limited by poor ceramic yields.
5.2.2.4. Solvent cast coating
The solubility of this polymer allows for its use as a solvent cast coating. The
polymer could be dissolved in a suitable solvent, such as THF, and spin cast as a coating
or used as a dip coating. Inclusion of the ethynylic moieties could allow for curing via
UV or thermal treatments. The high thermal stability should make it competitive to other
solvent cast thermoset coatings (epoxies, polyimides)
5.3. Conclusions
The new materials prepared in this research have potential utility for numerous
high performance applications. Further evaluation of these materials to produce
204
elastomers, thermosets, preceramics, and coatings should yield a rich area for future
materials research. Further synthetic studies, process development, and actual properties
generation are all potential areas of focus for future studies.
205
Appendix A: Ph.D. Candidacy Research Proposal
Electrochemically Assisted Sol Gel Deposition of Lanthanum and Cerium Oxides for
Enhanced Corrosion Resistance of Solid Oxide Fuel Cell Components
A research proposal by Kerry Drake
Department of Chemistry
Drexel University
April 23, 2009
Committee Members:
Dr. J.C. Bradley (Chair)
Dr. A. Addison
Dr. J. Foley
Dr. F. Ji
Dr. G. Palmese
Dr. S. Solomon
206
Abstract:
The goal of this research proposal is to develop a pulsed electrochemically
controlled sol gel deposition process for cerium and lanthanum oxides. This process will
yield uniform dense coatings which can be used to enhance corrosion resistance of
metallic solid oxide fuel cell components.
Significant research has been done over the last decade on fuel cells as the next
generation of power sources 1-4
. Solid oxide fuel cells (SOFCs) are one variant of this
technology with promise for both industrial scale power generation, as well as smaller
scale applications in commercial transportation or in the home. However, material
limitations have had a significant impact on commercialization of this technology.
Specifically, the use of metallic components is limited due to corrosion issues. This can
be mitigated by deposition of rare earth oxide coatings (lanthanides, etc.) for enhanced
corrosion resistance.
Sol gel is a convenient technique for deposition of metal oxides at relatively low
temperatures (500 oC to 800
oC). However, thickness limitations due to application
method, potential cracking from release of condensation products from thick films during
sintering, and precise control of coating thickness and deposition is often an issue.
Electrochemical techniques can be used to accelerate and control both the
hydrolysis step as well as condensation step in development of a sol gel film. In
nonaqueous solutions, acid or base catalysts can be generated at the cathode or anode
respectively through controlled addition of water as a reagent. By controlling current and
207
thus controlling the concentration of acid or base catalysts, the rate of deposition at the
surface of the electrode can be modified to produce a more consolidated film, and thus
allow more condensation products to escape prior to sintering.
Pulsed deposition, where potentials or currents are cycled on/off, may also be
used to further control the rate by even tighter control of the quantities of generated
catalyst as well as allowing time for the charged layer to dissipate between pulses. This
would allow charged species (monomer ions or catalyst) to more readily come in contact
with the growing surface, thus enhancing coating deposition condensation rates. In
addition, the ability to reverse cycle to neutralize catalyst (switch from acidic conditions
to basic conditions via reversing potentials) would allow for very tight control of
deposition conditions, which should yield a significant improvement over current
methods.
A.1. Solid Oxide Fuel Cell Overview:
Fuel cells generate energy through electrochemical reactions of fuel and oxygen,
rather than the production of heat through combustion and its subsequent use to perform
work (by turning a turbine generator, for example). Solid Oxide Fuel Cells (SOFCs)
have the highest system efficiencies (~85%, including use of high temperature steam for
power cogeneration), and operate at the highest temperatures.
Fuels for SOFCs include hydrogen gas, methane or other hydrocarbons.
Hydrocarbons are actually used as precursors for CO and hydrogen fuel, since they react
at high temperatures with steam to form hydrogen and carbon monoxide). Figure A.1
shows a basic schematic for a SOFC.
208
Figure A.1. Schematic of solid oxide fuel cell5.
Overall cell voltages are in the 0.9 to 1 volt range 6. Cells must be stacked in series in
order to obtain higher voltages.
The main advantages of solid oxide fuel cells over other fuel cell variants are as
follows (due mainly to the high operating temperatures, relative to other types):
• H2 and CO fuels can be directly formed from hydrocarbons during cell operation.
• No precious metal catalysts are required for the electrochemical reactions.
209
• High operating efficiencies and total system efficiencies. High temperature steam,
a byproduct of the SOFC reaction can be used for cogeneration of power (to turn a
generator) or for formation of fuel 7.
Unfortunately, the high operating temperature (850 oC to 1000
oC) is also one of
the main disadvantages of SOFCs. These temperatures necessitate the use of ceramics for
many of the components, including the cell interconnects.
Typical cell configurations are either tubular or planar. Examples of each are
given below (Figures A.2 and A.3). Tubular SOFCS components are approximately 1.5
meters in length by 2.2 cm in diameter, and are closed at one end.
A.1.1 Solid Oxide Fuel Cell Components
The most common electrolyte for SOFCs is yttria stabilized zirconia (YSZ). 8%
molar yttrium is the optimum concentration for enhanced O-2
conductivity for cell
operation (Y+3
ion substitution for Zr+4
results in oxide vacancies in the lattice which are
available for O-2
conduction). 10 % Y is the concentration which yields the most stable
zirconia phase, but results in lower O-2
conductivity than 8% YSZ. The electrolyte is
placed in the cell by either electrochemical vapor deposition or ceramic tape casting 8.
Anodes are usually porous cermets (ceramic/metal composites). Pores are
intentionally produced in order to allow ionic movement of fuel, with inclusion of
sufficient metallic phase content to allow electronic conduction. 8
210
Figure A. 2 Tubular Solid Oxide Fuel Cell7
211
Figure A. 3 Planar Solid Oxide Fuel Cell7
The cathode is also a porous ceramic perovskite, such as lanthanum manganite 7.
Porous cermets and ceramics are often produced via hot isostatic pressing, a common
method used to form ceramic parts 9. Machining to final dimensions is usually required.
Cells are stacked via interconnects, which are also called bipolar plates in planar fuel cell
designs. Interconnects have several key requirements:
• Good electrical conductivity, so there is minimal loss of generated power through
resistive processes.
212
• Corrosion resistance, so the resistivity does not change over time via buildup of
non-conductive corrosion products (oxides, etc.).
• The interconnects must be able to withstand both oxidative reactions and
reductive reactions, since they are in direct contact with the cathode of one cell, and the
anode of another cell.
In planar cells, interconnects also contain channels for fuel and oxygen to enter
the cell electrolytes for reaction. Often, the channels are cut at different angles on each
side of the bipolar plate (0o on front side, 90
o on the back side) See Figure A.3 for
details.
Commonalities of most designs include relatively large interconnecting
components (0.5 to 1.5 meters long) with complex geometries. Interconnects are exposed
to both oxidizing and reducing atmospheres at high temperatures, and thus require
excellent corrosion resistance and chemical stability. These requirements are currently
met to some extent with ceramic materials, such as LaCrO3 at operating temperatures of
850oC to 1000
oC
10. Unfortunately, ceramic materials are relatively difficult and
expensive to fabricate in large shapes 11
, and can usually only be machined by grinding.
Interconnects are often the most expensive component in a SOFC. Traditional
metallic corrosion resistant alloys are much easier to fabricate into complex shapes via
standard machining methods (milling, etc.). Unfortunately, the high temperatures and
aggressive environments in SOFCs have in many cases exceeded the limits of these
alloys. If corrosion resistance can be improved in a cost effective manner, metallic
components may be substituted for ceramics.
213
Recently, there has been a trend towards lower operating temperatures (500 oC to
800 oC), in part to allow the use of ferritic alloys (alloys that contain 10-30% Cr).
However, corrosion resistance and the associated increase in resistivity due to formation
of scale is an ongoing issue. 8
One promising method of enhancing corrosion resistance of ferritic stainless steel
alloys is by coating with a thin layer of an oxide of a reactive element such as Ce, La, Zr,
or Y.
A.2. Rare Earth Coatings for SOFC Components to Enhance Corrosion Resistance:
Seo et al 12
evaluated the corrosion resistance of Fe–22Cr–0.5Mn ferritic stainless
steel alloys which contained small quantities (0.05 to 0.06 wt %) of rare earth elements
Y, Ce, or La dispersed throughout the specimen (not as a coating). Unfortunately, no
undoped alloy was evaluated as a baseline, so general improvement gained by reactive
element doping was not evaluated.
The authors rated the corrosion protection order Y>La>Ce. However distribution
of the rare earth elements appears to play a critical role in the corrosion inhibition, and
the elements studied. Y appeared to disperse more evenly throughout the alloy, while La
and Ce concentrated mainly at grain boundaries. In principle, application of the rare
earths/reactive elements as a coating, to concentrate them on the surface of the parts,
should provide further enhancement of corrosion protection.
Fontana et al13
studied both conductivity and corrosion protection of La, Nd, and
Y oxides coatings deposited on ferritic steels via chemical vapor deposition (CVD).
Coating thicknesses were estimated to be in the 100-200nm range, based on the process
214
used for coatings. La and Nd oxide showed the lowest resitivitities after aging in air at
800oC. La and Nd coated specimens also showed the lowest weight gains. Measurements
of the oxide scales that formed on the rare earth coated surfaces were roughly ½ the
thickness of oxides that formed on uncoated surfaces.
Zhua et al14
coated type 444 stainless steel with lanthanum oxide via reactive
sputtering (sample was oxidized under low pressure to form Cr2O3, then PVD sputter
coated with La2O3, then annealed to form a LaCrO3 phase) and sol gel deposition.
Oxidation tests of the coated substrates showed a slow, even weight gain, whereas the
untreated substrate showed weight loss, likely due to spallation.
Sol gel coatings had the advantage of no observed formation of Kirkendall voids
in the La2O3/Cr2O3 interface. These voids were found in the reactive sputtered coatings,
and could have a significant detrimental effect on the coating strength. (Kirkendall voids
are crystal defects/vacancies which allow inter material diffusion into a boundary layer
from an alloy. In the system above, this could result in voids that allow diffusion of
oxygen or chromium to the boundary layer, and thus result in corrosion buildup at the
interface, rather than on the surface of the coating)15
.
Thermal spray techniques have also been used to produce coatings on ferritic
SOFC interconnects16
. Unfortunately, this technique is a “line of sight” deposition
process which results in relatively porous coatings. Other techniques such as e-beam
deposition, vacuum plasma deposition, and magnetron sputtering have also been used to
prepare coatings on SOFC components17
. However, high equipment costs, low deposition
215
rates and also some line of sight process restrictions have limited the use of these
techniques for commercial or “near commercial” SOFC component manufacturing.
These and other studies have shown that coating of ferritic alloys with reactive,
rare earth elements has significant beneficial effects in terms of reduction of corrosion
growth and maintenance of low resistivity. Unfortunately, most methods of application
are either highly directional (thermal spray) or very expensive, slow processes (PVD). In
addition, conductivity of the interconnect must be maintained, so precise control of
coating thickness with minimal porosity and crystal lattice voids is critical.
Due to the difficulties in controlled, repeatable application of coatings to these
types of components, the need for an alternative method of coating preparation is clear.
The use of sol gel technology for coatings is relatively well developed for silica based
systems; however, much less work has been done on rare earth systems. The addition of
electrochemically assisted/controlled sol gel deposition has been demonstrated for
production of improved coating properties, relative to “standard” dip or spin coated sol
gel.
A.3. Sol Gel:
Sol Gel is a technique whereby cross linked inorganic networks can be formed
from solutions. The term “sol” generally refers to a colloidal solution, “gel” refers to a
3D interconnected network. Metal alkoxides are typically used as starting reagents, with
TEOS (tetraethoxysilicate) being probably the most well known system studied in the
literature.18
216
The sol gel process for production of inorganic networks proceeds through 2 steps,
hydrolysis, and condensation.
Hydrolysis can be acid catalyzed (Scheme A.1) or base catalyzed (Scheme A.2)
M
OR
OR
OR
O
R
H+
M
OROR
ORO
+
H
R
O
H
H
M
OROR
ORO
+
H
H
O
H
H
+ROH
M
OR
OR
OR
O
H
+ H3
O+
Scheme A. 1 Acid catalyzed hydrolysis, sol gel.
Scheme A. 2 Base catalyzed hydrolysis, sol gel.
Condensation is also catalyzed by both acids and bases (Schemes A. 3 and A.4,
respectively).
217
Scheme A.3 Acid catalyzed condensation, sol gel
Scheme A.4 Base catalyzed condensation, sol gel
Hydrolysis and condensation can occur either stepwise (full hydrolysis followed
by condensation) or simultaneously. The dominant mechanism depends on the system,
catalyst, and other reaction conditions. For non silicate alkoxides, hydrolysis is often very
fast. This necessitates the use of non-aqueous solvents and the use of water as a reagent,
rather than the solvent.19
In addition, chelating agents are often used in to suppress some hydrolysis and
condensation paths, in order to moderate reaction rates. The complexing ligands appear to
218
block one or more of the bonding sites from reacting with other molecules, which results
in better control of film deposition 19
.
A.4. Electrochemically Assisted Sol Gel:
Sol gel technology is commonly used to produce coatings 18, 19
. However,
production of coatings with controlled thickness and full coverage is often difficult to
obtain. Electrochemical generation of acid or base catalyst directly on the surface to be
coated is one way of enhancing coating quality. For example, electrochemically assisted
sol gel deposition of silicates has been shown to enhance corrosion protection 20
.
Work by Hu et al showed that electrochemical deposition of coatings yielded
better coating consistency which resulted in better corrosion resistance than dip coating
with the same solutions.
219
FigureA.4 a,b,c-vinyltrimethoxysilane deposits obtained at several potentials. a) no
applied voltage, b) deposition at -0.8V, c) deposition at-1.2V (SCE) 20
. Note the uneven
coverage in a) and the large voids in c) most likely due to evolution of H2 gas which
generated significant porosity.
A.4.1. Electrochemically assisted sol gel deposition at cathodic and anodic potentials
Evaluation of both acid and base catalyzed sol gel reactions is proposed for this
research proposal, based on seminal work by Shacham et al21
. This group studied the
effects of positive and negative potentials on the deposition of zirconia thin films through
a sol gel process. Zirconium tetra propoxide/propanol solutions were used, along with
0.1M LiClO4 (ostensibly added as a supporting electrolyte).Small amounts of water were
added as a limiting reagent.
Film deposits were obtained at both positive and negative potentials. Significant
parameter interrelationships identified were as follows:
• Added water was necessary for generation of the catalysts via oxidation or
reduction, even though reactions of the alcohol could theoretically also generate acid
(H+) via the reaction:
Me2CHOH�Me2C=O + 2H+ +2e
- Eo~+1.8 (SHE)
220
or base (CH3O-) via the reaction:
2Me2CHOH +2e- �Me2CHO- + H2 (gas) Eo~-0.7(SHE)
• Applied potentials significantly affected coating thicknesses. The
Thickness/Voltage slope was steeper for anodic deposition than for cathodic deposition.
Thicker coatings were obtained for anodic deposition than for cathodic deposition
(1200nm vs. ~400nm).
• Applied anodic potentials also affected initial current densities, but currents
dropped as a function of time and converged for voltages -1V through -1.5V after ~ 15
minutes. The authors suspect this is due to diffusion limitations of catalyst (H+ or OH-)
through the film as it thickens.
• Applied potentials were limited from -1.6V to +2.5V vs. Ag/AgBr electrode due
to generation of large quantities of gas (H2 at negative potentials, O2 at positive
potentials).
• Stirring decreased film thickness, which the authors interpreted as an indication
mass transport of monomer to the film surface was not a limiting factor in deposition. The
thinning of films through stirring appears to substantiate the requirement for a local
concentration gradient of the acid or base catalyst.
There results were encouraging, and will be used as a model for the experiments
in this proposal. However, a review of these conclusions indicates that pulsed deposition
should yield significant improvements over the DC deposition method.
A. 4.2. Pulse Plating
221
Pulse plating as the name implies is a technique where potentials are cycled on/off
or positive/negative (Reverse Pulse). Square wave pulses, such as shown in Figure A.5
are the most common waveforms used in pulse plating.
Figure A. 5. Square wave pulse profile for pulse plating experiments22
.
Where Ip= pulse current, Ia=Average current density. Off time can consist of a period
with zero current (designated “Pulsed Deposition”) or with a reversed current (designated
“Reversed Pulse deposition”).22
222
In standard DC experiments, a charged layer quickly forms near the active
electrode, where the concentrations of ions of interest vary inversely with distance from
the surface of the electrode. This is known as the Nernst diffusion layer 23
.
The maximum theoretical current, also called the limiting current, is described by
the following equation:
δ
CaDnFAti =)(
Where D=Fick’s diffusion coefficient (10-5
cm2/sec in aqueous solutions), F=Faraday’s
constant, A=Electrode area, Ca=bulk solution concentration, and n=number of electrons
in Ox/Redox reaction, δ=the thickness of the diffusion layer.
For stationary electrodes, the thickness of the Nernst layer increases with the square root
of time, Dtπδ = , therefore the current as a function of time follows this relationship,
known as the Cottrell equation23
:
Dt
CaDnFAti
π=)( ,
The practical implication is that Faradaic currents decrease over time, due to both
depletion of active ions in the Nernst layer as well as an increase in charged layer
thickness. This means that deposition rates will also decrease as a function of time in
standard DC experiments due to charging.
Stirring helps enhance transport of ions to the electrode surface to a certain extent
and reduces the thickness of the Nernst layer, but diffusion through the Nernst layer is
223
still a limiting factor in deposition at the electrode surface. Also, stirring when trying to
deposit a sol gel coating would be detrimental, as the convection likely would remove the
fragile outer layer of deposited material
Pulse plating on the other hand allows a replenishment of the species of interest
during the off duty segment of the cycle through diffusion, rather than through
convection. This effectively results in a dual Nernst layer, a pulsed layer δp close to the
electrode surface and a stationary second layer δs (Figure A.6).
Figure A.6 Concentration profiles of pulse plating experiment, with DC concentration
profile overlaid for comparison.
DC concentration gradient
superimposed for reference
224
Note the differences in concentration at the electrode surface between DC and
pulsed experiments, and also the higher effective distance from the cathode through
which an ion must diffuse to get to the surface of the electrode.
When potential is cycled off in a pulsed experiment (“off” time in Figure A.5),
unreacted species of interest can then diffuse towards the surface of the electrode through
the δp. In addition solvent molecules, condensation byproducts, or counter ions that built
up during the “on” pulse can also diffuse away from the electrode when potentials are off,
at least through the δp layer.
The utility of pulsed deposition over DC methods for sol gel deposition should be
significant, especially in light of the difficulties seen in DC electrochemically assisted sol
gel discussed above.
A.5.1. Proposed Research
The goal of this project is to produce and evaluate coatings of lanthanum and
cerium alkoxides via pulsed electrochemically assisted sol-gel deposition. Lanthanum
ions are predominantly in the +3 oxidation state, while cerium is usually in the +4
oxidation state, so this would allow investigation of tervalent vs. tetravalent metal
alkoxides and their comparable amenability to sol-gel type deposition processes.
Pulsed deposition should yield the following advantages over other sol gel techniques:
• Much better control of catalysts concentration and location than standard sol gel
methods. Pulsing can generate controlled amounts of catalyst, reverse pulsing can
neutralize the catalysts.
225
• Rare earth oxides are sensitive to acids, so generation of small controlled amounts
of catalyst should in theory allow the condensation reactions to proceed without
unwanted acidic degradation of growing oxide layers.
• Most metal hydroxides are relatively insoluble in water or ethanol. Pulsed
deposition should in principle generate smaller quantities of fully converted metal
hydroxides. In addition, the generated hydroxides should form very close to the electrode
surface, where deposition is desired, rather than in the bulk solution, as seen in
conventional sol gel methods
• Since the condensation reaction will be limited to the region near the electrode
surface, lower concentrations of the metal alkoxides should be required than necessary
under DC conditions.
• Low effective currents can be used to minimize generation of unwanted gaseous
byproducts (large quantities of H2 or O2), which can be detrimental to film quality20
.
• Improved diffusion characteristics of electrochemically prepared catalysts, as well
as for removal of condensation products. This should result in production of thicker,
more dense deposits.
A.5.2. Chemistry
Sol gel deposition of rare earths is feasible via several routes. Zhu et al10
performed La Cr codeposition on a ferritic stainless steel via sol gel method using
lanthanum nitrate, with citric acid chelating agents. Key issues identified by the authors
226
that still needed to be addressed with the sol gel process were control of coating thickness
and minimization of porosity.
Pohl et al 25
prepared La 0.5Sr 0.5CoO3 coatings via sol gel methods from mixed
alkoxides. Precursors used were La, Sr, and Co methoxyethanol alkoxides, in
toluene:methoxyethanol solutions.Gels were formed by deposition of solutions on
aluminum and subsequent drying. Heating to 800oC yielded a perovskite structure. The
main experimental difficulty found in this work was preparation of a suitable soluble Co
alkoxide precursor, not the lanthanum alkoxide component.
Khalil et al. utilized cerium isopropoxide in isopropanol to prepare mesoporous
CeO2. In this case, ultrasonic dispersion was used to enhance solubility of the cerium
isopropoxide26
. Note: In the case of electrochemically assisted sol gel deposition, low
solubility should be less of an issue, as the reaction will occur only in the vicinity of the
electrode. Precipitation of insoluble hydroxides in the bulk solution (after significant
alkoxide hydrolysis) should be much less of an issue.
Based on these results, and available research on electrochemically assisted sol
gel, cerium and lanthanum isopropoxide systems are the most likely to be successfully
deposited, and thus will be the center of focus for this research proposal.
Both alkoxides are soluble in non-aqueous solvents:
• Cerium isopropoxide is soluble in THF and pyridine 26
.
• Lanthanum isopropoxide is soluble in hot propanol 27
.
227
Also, both isopropoxides are available commercially28
, or can be synthesized 27,29
. Other
alkoxides or acetates might also be tried if the isopropoxide work is not successful.
A. 5.3. Experimental plan
The first experiments performed would be determination of the solubilities of
each metal alkoxide in their preferred solvents. Temperature effects on solubility would
also be evaluated.
The next experiments would be performed to determine of conductivity of the
solutions, with supporting electrolyte, if necessary. LiClO4 should be inert within the
electrochemical window of these experiments, but this should be verified by cyclic
voltammetry (+ and – voltage limits to be determined). If any interfering reactions occur,
a suitable alternative will be identified (ammonium perchlorate, etc.).
Once the solution concentrations of metal alkoxides and supporting electrolytes
are finalized, determination of the electrochemical window of entire system via cyclic
voltammetry would be the next experimental task. Testing of anhydrous solutions should
show potentials at which the solvents break down, or if any precipitation occurs i.e.
La(OH)3, Ce(OH)4 . Since the reduction potentials of Ce+3
and La+3
are -2.33 and -2.37
respectively30
, reduction of these ions to metallic form should not occur. (Note: Ce+4
reduction to Ce+3
has an Eo of+1.77 V, so cathodic processes may result in some
conversion to Ce+3
). The reduction potential of Lithium is -3.2V, so its reduction should
not occur in the electrochemical range of this research proposal.
After the electrochemical windows of the solvent, supporting electrolyte, and
alkoxides have been determined, the effects of added water concentration on both
228
reaction potentials and deposition rates will be evaluated. Cyclic voltammetry, both
stirred and unstirred, will give insight into the reduction potential of water in these non-
aqueous systems.
Once proper potentials and limiting currents are determined for production of H+
or OH –under DC conditions, experiments will be performed to determine double layer
charging and discharging times. From that information, pulse on and off times will be
selected.
Chronoamperometry will be used to calculate the diffusion rate of alkoxide
electrolyte (under anhydrous conditions) and water to working electrodes. This
information will be used to help determine optimum duty cycle in pulse plating
experiments.
• time for current to reach maximum will relate to “on” pulse length.
• time for current to reach zero once potential is switched off will be related to time
required for species of interest to diffuse to electron surface after initial depletion-this
will help in determination of the time required on the off duty pulse cycle to replenish the
Nernst layer.
Figure A.7 Square wave applied current pulse, with actual currents overlaid for
comparison 22.
T(off)
T(on)
229
Both pulsed deposition and reverse pulsed deposition will be performed and
compared, in terms of quality of coatings produced. Potentials determined from the cyclic
voltammetry experiments for oxidation and reduction of water in these systems will be
used for the reversed pulse experiments. Times for pulses will be calculated based on the
criteria of Ibl et al .24
• Pulse Time On >time for current to reach 99% of limiting current (See Figure
A.7-T(on)).
• Pulse Time Off >discharge time=time for current to change from Limiting current
to 1% Limiting current when potential is cycled off (See Figure A.7-T(off)).
• For reverse pulse deposition, Time of reverse cycle >time for current to reach
99% of reverse I limit. This should allow enough time to generate catalyst in the reverse
pulse to neutralize that formed during the on pulse.
Chronoamperometry will again be used to determine charging and discharging
times. Chronoamperometry will also be performed at different times during the
experiment to measure changes in diffusion characteristics with time (likely due to
coating build-up on electrode screening full potential of electrode from bulk solution).
These experiments would be performed at both anodic and cathodic potentials. In
particular, evaluation of feasibility of acidic catalysis (anodic deposition) would need to
230
be performed. There is some possibility of oxidation reaction of stainless steels. This
would have detrimental effects on coating quality due to corrosion of the substrate.
Cathodic reaction likely would not have these issues due to high pH at the electrode
surface. The main issue with cathodic deposition would be uncontrolled generation of H2
gas from reduction of water, which would result in voids in the coating.
Coating depositions will first be attempted on inert electrodes to determine
feasibility of reactions and ability to deposit a coating. Both cathodic and anodic
depositions will be attempted.
Optimum firing temperature will be determined through a combination of
thermogravimetric analysis (TGA), X-ray diffraction (XRD) phase analysis, and
morphology via microscopy.
Infrared and Raman spectroscopy should also be useful techniques to analyze the
coatings for relative intensity levels of M-OH vs. M-O-M bonds as functions of
deposition time, applied potential, and also heat treatment. La2O3 has Raman shifts at
104, 191, 301 and 410 cm-1
31
, while CeO2 has a single Raman shift at 465cm-1 32
.
Residual M(OH)x bonds can be tracked by monitoring the OH stretch at ~3400cm-133
via
IR. IR and Raman bands of LaCrO3, a possible reaction product after high temperature
sintering, would likely be slightly different (MOM’ bonds should be IR active vs. Raman
active MOM bonds) and may need to be determined experimentally as part of this
research proposal.
Coating of a representative grade of ferritic steel will be the next step. Type 444
will be used, as this is a typical steel for low temperature SOFC applications14
.
231
SEM/EDX will be performed on dried samples to verify La / Ce deposition and
look at coating quality (smoothness, cracks, voids, etc.). Cross sectional analysis on
mounted samples will be performed to determine coating thickness. If coating is too thin
for SEM detection, TEM will be performed.
The optimal temperature/time cycle determined for inert electrode deposits should
be used as the starting point. Raman and IR spectroscopy will be used to evaluate the
coatings to help select the best cycle parameters.
High temperature corrosion tests will be performed to determine protective effect
of sol gel coating. Testing would be similar to that used by Seo et al. 12
. Samples would
be coated and then oxidized in an air environment at 700oC-800
oC. Weight changes,
surface morphology changes, and conductivity changes would be determined and
compared to untreated surfaces to determine advantages of coating. A reference sample
of LaCrO3 would also be tested under the same conditions as a benchmark of current
technology.
General electrochemical corrosion testing with QCM will also be performed. This
will allow determination of weight changes due to oxidation layer build-up in an aqueous
environment. Changes in a coated specimen will be compared to an uncoated specimen.
Polarization curves will be run (Tafel Plots, as per Figure A.8, example data) to
determine corrosion currents, icorr. These currents will be compared between uncoated
(reference) samples and coated samples for qualitative evidence of surface passivation
(corrosion protection).
232
Pitting experiments will also be performed in 1M NaCl in deionized water.
Potential will be cycled as per Figure A.9. Evaluation of Epit and Eprot will be performed
on coated samples as well as control samples.
Evaluation of the surfaces after corrosion experiments for pitting, buildup or other
physical changes based on corrosion will be performed. Electron microscopy would be
preferred, as long as specimens can be observed without coating and without the need for
high vacuum. This should be possible with an Environmental SEM.
Figure A.8 Tafel Curve 34
.
233
FigureA. 9. Cyclic polarization curve35
.
A.6. Equipment Requirements
A.6.1. Electrochemical Equipment
In order to perform the experiments, electrochemical equipment capable of
precise control of applied potentials with pulse capability will be needed. Experiments
should be run in a multiple compartment cell to minimize reaction of working electrode
surface with reaction byproducts from the counter electrode.
234
Inert atmosphere purge is also recommended, to maintain moisture control and to
eliminate possibility of interfering oxygen reactions (reduction of oxygen, etc.). Heating
capability, for enhanced solubility of salts if needed would also be desired.
Working electrodes will include noble elements (Pt, Au) as well as representative
ferritic alloys that are targeted for SOFC use. A quartz crystal microbalance (QCM)
electrode setup would also be recommended, to monitor weight changes on the working
electrode over time.
A.6.2. Characterization equipment
• SEM, with low pressure capability (ESEM) to evaluate gels. An SEM with a
heated stage to allow dynamic observation of film changes on heating would be
recommended. Depending on coating thicknesses, mounting and cross-sectional analysis
would also give important information.
• TGA, DSC for thermal analysis to determine decomposition characteristics of the
gels. TGA/MS would be useful as well for determination of the volatile species that are
evolved at different temperatures. Tracking the identities of condensation products as
they relate to different experimental parameters could potentially give useful information.
• Glancing angle x-ray diffraction to allow determination of temperatures that
produce crystalline oxides.
• FTIR and Raman equipment capable of measurement of the active bands
mentioned above.
A6.3. Oven for heat treatment of coatings
235
An oven with air and inert purge is required. The oven must be capable of
temperatures up to 1000oC (750
oC to 800
oC minimum).
A7. Conclusions
Solid oxide fuel cells are a promising technology for future power generation.
However, the technology is currently limited by material capabilities. The ability to use
standard metallic components would be a significant advancement; however corrosion
resistance under operating conditions is a major limiting factor, particularly for
interconnects. Rare earth oxide coatings are one method for enhancing corrosion
resistance of ferritic alloys for SOFCs. Sol gel deposition of rare earth oxides is feasible,
but coating control is often an issue.
Based on the best available literature and review of electrochemical principles,
pulsed electrochemically assisted sol gel should yield significant improvements in
coating quality over standard sol gel deposition and other coatings technologies (plasma
spray, etc.). If successful, this research could yield an enabling technology for SOFCs,
with potential for use for corrosion protection in other applications as well.
References
(1) Stambouli, A. B.; Traversa, E. Renewable & Sustainable Energy Reviews 2002, 6,
297-306.
(2) Cowey, K.; Green, K. J.; Mepsted, G. O.; Reeve, R. Current Opinion in Solid State &
Materials Science 2004, 8, 367-371.
(3) de Bruijn, F. Green Chem. 2005, 7, 132-150.
236
(4) Wee, J. Renewable & Sustainable Energy Reviews 2007, 11, 1720-1738.
(5) Fuel Cells. Ullmann's Encyclopedia of Industrial Chemistry; 2004.
(6) Sasaki, K.; Hori, Y.; Kikuchi, R.; Eguchi, K.; Ueno, A.; Takeuchi, H.; Aizawa, M.;
Tsujimoto, K.; Tajiri, H.; Nishikawa, H.; Uchida, Y. J. Electrochem. Soc. 2002, 149,
A227-A233.
(7) Fuel Cells. Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley and
Sons: 2001; Vol. 12, pp 199-228.
(8) Brett, D.; Atkinson, A.; Brandon, N.; Skinnerd, S. Chemical Society Reviews 2008,
37, 1568-1578.
(9) Singhal, S.; Kendall, K. High Temperature Solid Oxide Fuel Cells: Fundamentals,
Design, and Applications ; Elsevier: 2003.
(10) Zhu, W.; Deevi, S. Materials Research Bulletin 2003, 38, 957-972.
(11) Barsoum, M. Fundamentals of Ceramics; Institute of Physics: 2003.
(12) Seo, H.; Jin, G.; Jun, J. Journal of Power Sources 2008, 178, 1-8.
(13) Fontana, S.; Amendolab, R.; Chevalier, S.; Piccardo, P.; Caboche, G.; Viviani, M.;
Molins, R.; Sennour, M. Journal of Power Sources 2007, 172, 652-662.
(14) Zhua, J.; Zhanga, Y.; Basua, A.; Lua, Z.; Paranthamanb, M.; Leec, D.; Payzantc, E.
Surface and Coatings Technology 2004, 177-178, 65-72.
(15) Nakajima, H. JOM 1997, 49, 15-19.
(16) Hui, R.; Wang, Z.; Kesler, O.; Rose, L.; Jankovic, J.; Yick, S.; Maric, R.; Ghosha, D.
Journal of Power Sources 2007, 170, 308-323.
(17) Pederson, L.; Singh, P.; Zhou, X. Vacuum 2006, 80, 1066-1083.
(18) Wright, J.; Sommerdijk, N. Sol-Gel Materials; CRC Press: 2001.
237
(19) Sol Gel Technology. Kirk-Othmer Encyclopedia of Chemical Technology, John
Wiley and Sons: 2007.
(20) Hu, J.; Liu, L.; Zhang, J.; Cao, C. Progress in Organic Coatings 2007, 58, 265-271.
(21) Shacham, R.; Mandler, D.; Avnir, D. Chem. Eur.J. 2004, 10, 1936-1943.
(22) Chandrasekar, M.; Pushpavanam, M. Electrochimica Acta 2008, 3, 3313-3322.
(23) Analytical Voltammetry and Polarography. Ullman's Encyclopedia of Industrial
Chemistry; 2004.
(24) Ibl, N. Surface Technology 1980, 10, 81-104.
(25) Pohl, A.; Westin, G. journal of The American Ceramic Society 2005, 88, 2099-2105.
(26) Khalil, K.; Elkabee, L.; Murphy, B. Microporous and Mesoporous Materials 2005,
78, 83-89.
(27) Brown, N.; Mazdiyasni, K. Inorganic Chemistry 1970, 9, 2783-2786.
(28) Metal-Organics for Materials, Polymers & Synthesis;Gelest; 2009.
(29) Gavrilenko, V.; Chekulaeva, L.; Savitskaya, I.; Garbuzova, I. Russian Chemistry
Bulletin 1992, 42, 1957-1959.
(30) Electrochemical Series; Vanýsek, P., Ed.; CRC Handbook of Chemistry and
Physics; CRC Press, 2003.
(31) Wang, S.; Wang, W.; Qian, Y. Thin Solid Films 2000, 372, 50-53.
(32) McBride, J.; Hass, K.; Poindexter, D.; Weber, W. Journal of Applied Physics 1994,
76, 2435-2441.
(33) Cutrufello, M.; Ferino, I.; Monaci, R.; E. Rombi, E.; Colo, G.; Nav, J. Phys. Chem.
Chem. Phys. 2001, 3, 2928-2934.
(34) Cheng, Y.; Zheng, Y. F. Surface and Coatings Technology 2006, 200, 4543-4548.
238
(35) Jones, D. Principles and Prevention of Corrosion; Prentice-Hall: Upper Saddle
River, NJ 07458, 1996.
239
Vita
Kerry Anthony Drake, Born in Shenandoah, Pennsylvania, US citizen.
Education
BS Chemistry, Bucknell University.
MS Chemistry, University of Michigan (Ann Arbor).
Kerry resumed his studies at Drexel in 2006, as a Post- Masters’ PhD Student, and joined
Dr. Yen Wei’s research group in the Department of Chemistry.
Professional Experience
1991-1992: Graduate Teaching Assistant, University of Michigan, Ann Arbor, MI.
1992-2001: Senior Research Chemist, Arkema, King of Prussia, PA: specialization in x-
ray fluorescence, x-ray diffraction, electron microscopy and x-ray photoelectron
spectroscopy.
2001-2003: Thermoplastics Laboratory Manager at Greene Tweed, Kulpsville, PA:
specialization in thermoplastic chemical and physical characterization.
2003-2011: Scientist, Polymeric Materials Development at Greene Tweed, Kulpsville,
PA: Lead scientist responsible for all corporate R&D activities related to new
thermoplastic materials development.
Selected Publications
• Martyak, N. M.; Drake, K., Modulus of Elasticity of Electroless Nickel Coatings
Determined from X-Ray Diffraction Studies. Galvanoteknik 2000, 91, 11, 3062-3070.
• Martyak, N. M.; Drake, K., Peak Profile Analysis of Electroless Nickel Coatings.
Journal of Alloys and Compounds 2000, 312, (1-2) 30-40.
• Mukherjee, I.; Drake, K.; Berke-Schlessel, D.; Lelkes, P. I.; Yeh, J.; Wei, Y., Novel
Thermally Cross-Linkable Poly[(arylenedioxy)(diorganylsilylene)]s Based on
Curcumin: Synthesis and Characterization. Macromolecules 2010, 43, (7), 3277-
3285.
• Co-inventor on two provisional patent applications at Drexel and Greene Tweed
which have not been made public at the time of this writing.