 

LIBRARY W
Michigan State
‘ University

 

 

J

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

l
7L_l-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

*1??? ~

MSU Is An Affirmative Action/Equal Opportunity institution
chnG-nt

 

 

STRUCTURAL PROPERTIES OF HIGH PERFORMANCE
POLYMER FIBERS AND THEIR EFFECTS ON
FIBER-MATRIX ADIIESION

by

Javad Kalantar

A DISSERTATION

Submitted to
Michigan State University
in partial ﬁdﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemical Engineering

1991

ABSTRACT

STRUCTURAL PROPERTIES OF HIGH PERFORMANCE POLYMER FIBERS
AND THIHR EFFECTS ON FIBER-MATRIX ADHESION

by

Javad Kalantar

Duﬁngthepastthreedeeades,manyimportanttypesofhigh-strengthand
high-modulus polymer reinforcing ﬁbers have been developed. These ﬁbers possess
combinations of stiffness, high strength, high toughness, and low density that rival the
properties of inorganic reinforcing ﬁbers such as glass and carbon ﬁbers. However,
these polymer ﬁbers generally exhibit weak adhesive and interfacial properties. This
study sought to develop a fundamental understanding of structural and chemical
properties of polymer ﬁbers that inﬂuence their adhesive and interfacial behavior. This
knowledge is critical for developing approaches to improve the engineering performance
of these ﬁbers.

Several physical and chemical treatmmts of polymer ﬁbers that produce different
extents of structural and chemical alterations were examined in this study. Polymer
treatments with coupling agents (titanium and zirconium organometallic complexes),
polymer coatings (butadiyne, Parylene-N, Parylenc—C), chemical treatments (ﬂuorination,
sulfonation), plasma treatments (0,, CE, He, C0,, N11,, N20, Ar, 11,0), ion
implantations ('I‘i“,Ar*,N*,He*), and structural modiﬁcations (sol-gel infusion, Friedel-
Craftschaincrosslinldng)wereexamined. Thesmdyconcentratedonpolyaramidﬁbers
(Kevlar. 29, 49, 149, and Technora'), aromatic heterocyclics (p-Phenylene
Benzobisoxazole), and ultrahigh-modulus polyethylene abet: (Spectra. 1000) as well as

model polycarbonate and polyethylene in sheet and bulk forms. The knowledge
developed, however, should be applicable to other high performance polymer ﬁbers.

Examinations of plasma treatments and coupling agents provided insights to the
interfacial limitations of ﬁber-matrix adhesion. Ion implantation, sol-gel, Friedel—Crafts,
and sulfonation treatments were examined because of their ability to affect the inter-ﬁber
cohesive interactions. Results show that the skin-core morphology and/or wetting
properties of high performance polymer ﬁbers can limit their adhesive and interfacial
load transfer properties. However, once these limitations are overcome, ﬁber lateral
cohesive properties become the limiting factor. Therefore, the key to improving the
ﬁber-matrix interfacial load transfer of high performance polymer ﬁbers is both to
improve the wettability, and to increase the ﬁber lateral cohesive strength.

The study also developed a new sample preparation and spectroscopic analysis
method to quantify the composition of the polymer inter-phase at high resolution. This
method signiﬁcantly facilitates atomic and chemical analysis of the polymer inter-phase.
The application of the method to determine sulfur distribution in sulfonated polycarbonate
samples provided experimental data for mass-transfer modeling of the sulfonation surface

Wt.

To my parents and family,
for all the love and support

you’ve given me.

iv.

Acknowledgements

I like to acknowledge and thank those individuals who have greatly inﬂuenced my
intellectual as well as personal development leading to this milestone. My parents,
Hassan and Parvin, who stressed the importance of education and through many
sacriﬁces provided me with unconditional emotional and ﬁnancial support. Professor
Lawrence T. Drzal who gave me his full support during my Master and Ph.D. education.
Professor Drzal gave me the freedom to grow, mature and achieve under his capable
guidance. He taught me excellence by his example.

My personal associations at the Composite Materials and Structures Center (CMSC)
will remain very special to me. My colleagues, Shri Iyer, Tesh Rao, Brent Larson,
Craig Chmielewski, Richard Schalek, Husan Almossawi, Pedro Franco-Hermra, Sanjay
Padaki, Himanshu Asthana, Murty Vyakamam, Matt Kendziorsln', Greg Fisher, Ed
Drown, and many others made CMSC the exceptional place it is. My association with
Richard Schalek was particularly productive both in research and in friendship. Thanks
to you all for your support and, more importantly, for your friendship.

As with most undertakings of this size, this dissertation was shaped by the ideas of
many people. My Ph.D. committee members, Prof. D. S. Grummon, Dr. R]. Morgan,
Prof. M.C. Hawley, and Prof. K. Berglund gave me direction and support. I
particularly thank Prof. D.S. Grummon for his professional guidance as well as his
personal friendship. I was always fascinated by Dr. Morgan’s stimulating ideas and
excellent insight into my area of dissertation. It was always a pleasure to discuss my
work with Prof. Hawley who generously gave me his time. Prof. Berglund taught me
about sol-gels and gave me excellent suggestions. Special thanks to Dr. E. Grulke and
W. Walles for their insights into sulfonation treatments and taking the time to share
them.

My appreciation to the staff of CMSC, Mike Rich and Dr. Kevin Hook for teaching
me a variety of instrumentations. Arlene Klingbiel always assisted me in the many ways
that only a good secretary can with a added bonus of a friendly personality. Thanks to
Brian Rock for his helpful hints and suggestions.

My time at Michigan State University was also highlighted by many people who will
remain my life-long friends. Farshid and Shabnam shared and comforted me during my
highs and lows. Farzad and Setareh amazed me with their hospitality and generosity.
Farshid and Kathy always made me feel welcomed.

My sincere gratitude to the El. du Pont de Nemours & Co. and Dow Chemical Co.
for providing ﬁnancial support for this research.

Finally, I would like to express my appreciation to the staff of Center for Electron
Optics for their help and informative discussions.

V

Table of Contents

List of Tables

us of ﬁgures
Nomenclature

Chapter 1: Introduction

1.1 High Performance Polymer Fibers
1.2 Dissertation Overview

Chapter 2: Emerimental
2.1 Materials

2.2 Interfacial Shear Strength Characterization
2.3 Surface Energy Characterization
2.4 Surface Chemistry Characterization

2.4.1 X-ray Photoelectron Spectroscopy
2.4.2 Auger Electron Spectroscopy

2.5 Microscopic Characterization
Chapter 3: Adhesive Properties of High Performance Polymer Fibers.

3.1 Introduction

3.2 Experimental
3.3 Results and Discussion

3.4 Conclusions

Chapter 4: Coupling Agent Ireatrnents of High Performance Polymer Fibers

4.1 Introduction

4.2 Experimental
4.3 Results and Discussion

4.4 Conclusions

4.3.1 Coupling Agents
4.3.2 Butadiyne Coatings
4.3.3 Parylene Coatings

Chapter 5: Plasma and Corona Treatments of High Performance Polymer Fibers

5.1 Introduction

5.2 Experimental
5.3 Results and Discussion

5.4 Conclusions

5.3.1 PBO Plasma Treatments
5.3.2 Polyethylene Corona and Plasma Treatments

WQQ¥HHQHQ°

sue-
Uh

16
17
19
20
21
27
28
47
43
49
53
56
56

383‘:

33333333

Chapter 6: Sulfonation and Fluorination Ireatmem of Polymers
6.1 Introduction
6.2 Experimental
6.3 Results and Discussion
6.3.1 Aramid Fluorination

6.3.2 Sulfonations of High Performance
Polymer Frbers

6.3.3 Polycarbonate Sulfonation
6.3.4 Polyethylene Sulfonation
6.4 Conclusions
Chapter 7: Ion Implantation of High Performance Polymer Fibers
7.1 Introduction
7.2 Experimental
7.3 Results and Discussion
7.3.1 Aramid Ion Implantation
7.3.2 Polyethylene Ion Implantation
7.4 Conclusions
Chapter 8: Structural Modiﬁcation of High Performance Polymer Fibers
8.1 Introduction
8.1.1 Infusion ofa Second Phase
8.1.2 Friedel-Craﬁs Reactions
8'1'3c33ﬂi'3i‘ynné‘ﬁ'éé" ““m"
8.2 Experimental

8.3 Results and Discussion
8.3.1 Sol-Gel Treatments
8.3.2 Epoxy Infusion

8.3.3ChainC ' ‘ b Freidel-Crafts
_ rosslrnlnng y

8.4 Conclusions

Chapter 9: Conclusions and Recommendations
9.1 Conclusions
9.2 Recommendations

Appendix A: Cpgnbeypressiw Strength Measurements of High Performance Polymer
rs

A.l Compression Measurement Techniques
A.2 Single Fiber Compression (SFC) Techniques
Appendix B: Thermodynamics of Surface Tension

vii

8828

1
100
105

107
113
123
125
126
131
133
133
145
157
159
160
161
163
164

169
171
176
183
185

187
188
189
191
193

194
197
201

Appendix C: A Novel le

sites
F.1 Introduction
F.2 Procedure
Appendix G: Wilhelrrry Program
Appendix H: Experimental Data
Bibliosraphy

viii

Preparation Technique rIonandElectronBeam 204
Analysis of Fiber-Matrix Interphase III? Polymer Composites

Appendix D: Unsteady Diﬂirsion in a Semi-Inﬁnite Slab
Appendix E: Sulfonation Treatments of High Density Polyethylene Gas Tanlcr
Appendix F: Procedures for Ultra-Thin Microtomy of Fiber Reinforced

213
217
220

220
221
226
229
247

Table 1.1
Table 2.1

Table 3.1
Table 3.2
Table 3.3
Table 4.1

Table 4.2
Table 5.1
Table 5.2

Table 6.1
Table 6.2
Table 7.1
Table 7.2

Table 8.1
Table 8.2

Table 8.3
Table 8.4

Table 8.5
Table 8.6

Table 8.7
Table 8.8
Table 9.1

Table A.l
Table E.l

List of Tables

Effectsofdifferent hesonthemo hol andchemi of
highperf 1PM rP 0%? stry

orrnance po ymer ﬁbers.

Surface energy components (dynelcm) of reference liquids used for
erhelmy contact angle measurements.

Tensile properties of high performance polymer ﬁbers.
Interfacial shear strength (18S) of high performance ﬁbers.
Material property data.

lene expenm' ental protocol and cunng' conditions examined for
eac ﬁber treatment.

Tensiles anddiameterofBand7mearylene-Ctreated
Kevlar-49 E—Glass ﬁbers.

Plasma treatment conditions for the ﬁrst set of PEG plasma

treatments. The expected major surface changes are also listed.

Plasma treatment conditions for the second set of PEG plasma
treatments.

XPS atomic composition of ﬂuorinated Kevlar-49 aramid ﬁbers.
XPS atomic composition of sulfonated polycarbonate surfaces.
Ion implantation protocols for polyaramid and polyethylene ﬁbers.

Mechanical Properties of untreated and 400 keV 10” He‘lcm2
implanted Spectra-1000 polyethylene ﬁbers.

Fiber diameter for dried, wet, and solvent exchanged PBO ﬁbers.
TGA weight loss measurements for various treatments of Wet-P130

Material property data for solvent exchanged Wet-P80 ﬁbers.

Atomic concentrations for Dow PBO ﬁbers, measured by XPS
(average of three runs for each treatment). ‘

Material data for Foster-Miller - sol- el treated and
untreatedpﬂ. (F M) g

Wet-P30 sol- el treatments and correspondrng' compressive
strengths. g

Compressive strengths of epoxy impregnations of Wet-P130 ﬁbers.
Material property data for Friedel—Crafts treated Wet-P130 ﬁbers.

Adhesive enhancement mechanisms introduced by various
treatments of the hrgh performance polymer ﬁbers.

Fiber compressive strengths (MPa) from various test methods.

XPS atomic% com sition and AES elemental line-scans
nonldepths or the sulfonated hrgh-densrty polyethylene gas
samp es.

4
15

29
31
31
55

60

132
153

172
172

174

176

218

ListofFigures

Figure 1.1 Speciﬁc tensile properities of reinforcing ﬁbers and conventional 2

bulk matenals.
Figure 1.2 Overview of the dissertation plan 5
Figure 2.1 Schematic of single ﬁber ﬁagmertadon process. 10
Figure2.2 Asamplemoldandtensilejigusedforthesingleﬁbe' 11
fragmertatron test.
Figure 2.3 Droplet test apparatus. 13
Figure3.1 Monomer structure ofsomeli uid stalline l ersusedin 22
high performance polymer ﬁbgrs. cry p0 ym
Figure3.2 Tensilefracturestrainofhi ormance lymerﬁbe'sveses 28
their tensile modulus. gh perf p0

Figure3.3 Experimentalandtheoretical interfacialshearstrength ofvarious 32

Figure 3.4 tical micrographs of ﬁber ments. A Kevlar-49 ' t-ﬁeld, 33
Kevlar-.49 cross-polarized, ( AS-4 bright-ﬁeld, (Brig-4

cross-polanzed.

Figune3.5 InterfacialshcarstrergthofyvashedPBO, Kevlar49andTechnora 34
ﬁbers for three different currng conditions.

Figure 3.6 Interfacial shear strer (188 of 'as received" (sized) and . 35
'w'asdlhtied' (unsrzed) lat-4 and Technora ﬁbers for two cunng
co ons.

Figure 3.7 Polar and disperssive components of surface energy for seveal 36
hé'ghperformance lymerﬁbersand ' uid x . Onl Kevlar-
4 andTechnoraffgersweesized. qu GPO! y

Figure 3.8 SEM micrographs of an untreated Kevlar-49 showin torn ﬁber 39
skininaformofahelicalribbon. g

Figure 3.9 SEM micrographs of a PBO ﬁber skin separation. 40

Figure 3.10 TEM mi hs of a radiall sectioned unheated Kevlar-49 ﬁber. 41
Fibe-mmm showsyboth interfacial separations and ﬁber
cohesive ﬁbrillatrons.

Figure 3.11 TEM rnicrographs of A 'as received“ and 'washed' Technora 42
ﬁbers. .The ”as recei( " ﬁber shows moreggtensive interfacial

ﬁbrillanons than the “washed" ﬁber.

Figure3.12 TEMmi hsofan'asreceived'PBOﬁber,showin athin 43
1a oftheEberslrinisadheringtotheeportysideoftheg

mgphase separations.

Figure 3.13 'cal micrograph of 'as received" PBO ﬁbes, showing preserce 44
o compressrve bands.

Figure 3.14 SEM micrograph of “as received" PBO ﬁbe's. Kink bands are not 45
apparett on the ﬁber surface.

Figure 3.15 TEM micrographs of an "as received“ Spectra-1000 l eth less 46
ﬁber, showing extensive interfacial adhesive failures 9.3.}. A.
ﬁber cohesive failures.

Figure 4.1 Genealreaction schene ofa titanium or zirconium based coupling
agent wrth the hydroxyl groups of a polymer substrate.

Figure 4.2 Polyme'rzah' 'on of a dracety' lere monomer uces a combination
of substituted (A) polyere and (B) polyacegemghuchrres.

Figure 4.3 Various types of parylere polymer available.

Figure 4.4 Interfacial shear; strength of Kevlarg49 aramid and AS-4 carbon
ﬁbers wrth hqurd couplmg agent mrxed 175°C cured epoxy.
Figure 4.5 terfacial shear shength of butadiyne treated Kevlar-49 aramid

In

ﬁbes with 175°C cured epoxy.

Figure 4.6 TEM micrograph of a 0.84 wt% butadiyne treated Kevlar-49 ﬁber.

Figurutﬂ Interfaeialshearsh'er ofheahnerts3and7mearylene;C
coatedKevlar-49, P , AS-4and E—Glass ﬁbersattwocunng
condrhons.

Figure 4.8 Lon itudinal thermal expansion ofthe unheated and Parylene-C
heaéd Kevlar-49 ﬁbers. The lene coatings terd to

longitudinally compress the ﬁber unng the cool down.

Figure4.9 Fracturin rocessof? m lere—CcoatedKevlar-49aramid
andAS cgrbonﬁbers.“ Pary

Figure4.10 'l'EMmi hsofradialsectionsofa3 m lene-Ccoated
Kevlar-49 in the 175°C cured epoxy 5mg”

Figure4.11 TEMmicro hsofradialsectionsofathin lere-Ncoated
Kevlar-49 mmeahnent F). WY

Figure5.l Schematicdragrarn‘ ofthesurfacemodiﬁcationof lymesina
glow discharge reactor. p0

ﬁgure5.2 Percertchan esintersileandinterfacialslwarshergthsoftheﬁrst
set of plasmagh'eated PBO ﬁbers. .

Figure5.3 Percertchan esintensileandinterfacialshearshengthsofthe
second set of plasma heated PBO ﬁbe's.

Figure 5.4 SEM mi hs of A 0.5, 0.75 and C 1.0 min.
0,!CF., 39 watts PHD plasmmhzeahnerts. ( 'D)

Figure5.5 TEMmicrographsofal.Omin. CF, 397um lasma
heated ﬁber (heatmert Y). 0,! 4 p

Figure5.6 Polaranddrspersr’ 'vecom tsofsurfaceerergyforsecondset
of plasma heated PBO ﬁg?“

Figure5.7 Intefacialshearshetgthofunheated,coronaheatedandplasma
heated Specha-lOOO polyethylene ﬁbe's as determmed by the
droplet test.

Figure 5.8 Polar and ' ’ve components of surface erergy for unheated,
corona and plasma heated Specha-lOOO polyethylete ﬁbers.

Figure 5.9 XPS atomic concenhations of unheated, corona heated and plasma
heated Specha-lOOO polyethylene ﬁbes.

Figure5.10 SuperimposedXPSsi sofox on 'on forcoronaand lasma
heated pecha-IOOO ghalay'lethylenzﬁ'rbe‘r:£1 p

388§38

$2

Figure 5.11
Figure 6.1
Figure 6.2
Figure 6.3
Figure 6.4
Figure 6.5
Figure 6.6
Figure 6.7
Figure 6.8

Figure 6.9

Figure 6.10

Figure 6.11
Figure 6.12
Figure 6.13
Figure 6.14
Figure 6.15
Figure 6.16

Figure 6.17

Figure 6.18

EM m1 hs of (A) unheated
lasma heated pecha-lOOO polyeth

Gereal reaction schemes for fluorination and sulfonation of
hydrocarbons.

Ratios of atomic% of hi er, oxygen, and ﬂuorine over carbon,
for the ﬂuonnated Kevlar

Tersile shength and interfacial shear shength (fragmertation test)
of the fluorinated and unheated Kevlar-49 aramid ﬁbers.

'I'EM micrographs of a radially sectioned fluorinated Kevlar-49
aramid ﬁber (heahnert D).

masses“ m (C)

SEM micro hs of A unheated and hcahnent

fluorinated ﬁlm-4 9(ﬁliers. (B) (B)
Relative ratios of tensile and interfacial shear shergth of
sulfonated ﬁbers compared eir unheated values.

TEM micrographs of a radially sectioned, sulfonated Kevlar-49
aramid ﬁber showing extensive ﬁber cohesive failure.

ABS analysis of sulfonated polycarbonates. (BSAiascan alow
magniﬁcation view of an analysrs surface
for a 1-hour gas phase sulfonated sample.

Sulfur penetration depths for the chilled (~ 17°
ambiert (~ 22°C) and warm (~ 37°C) solution sulfonated

Rimmed edby ”X1193 11.3%.?” ’°°‘ °f m

SEM mi crograpjhs of sulfonated and unheated pol carbonate

3 cmElaflnheatedosample, (B)l-hourgas sulfonated
sample. surfaces.(cs -hour solution phase heatedsam sam.ples
XPS atomic co sition of sulfonated Spech'a- -1000 without
preheatmert) )polgoethylene (
XPS atomic composition offl sulfonated Specha- ~1000 (Corona
preheated) spechaﬁ
XPS atomic composition of sulfonated Specha-IOOO (plasma
preheated) specha ﬁbers.

Sulfur atomic96 of sulfonated un-preheated,corona corona,preheated and
plasma preheated Specha-lOOO polyethylere ﬁbers.

Tensile shength and modulus changes of sulfonated ﬁbers relative
totheirunheated values.

Polar and dispersive componeratlrkl of surface erergy fosrllun-
gpecha 1000 oopolyethl ylere 06ml) ‘

Plotma of sulfur atomic% for un-preheated corona ted, and
sulfonatedS 4000 polyety ﬁbers ve'sus
compon etts of s erergy

m...” “a, massif. have «ﬁnesse...

sulfonated, Spoons-1000 pol lyethy

xii

91
94
101

110

112

114
114
115
115
117

119

119

120

Figure 6.19

Figure 6.20
Figure 7.1
Figure 7.2

Figure7.3

Figure 7.4
Figure 7.5
Figure 7.6

Figure 7.7

Figure 7.8
Figure 7.9
Figure 7.10
Figure 7.11
Figure 7.12
Figure 7.13

Figure 7.14

Figure 7.15

Figure 8.1

Figure8.2
Figure8.3

Interfacial shear earshength mcreasc of sulfonated S -1000
lyrlzgylerf ﬁbes compared to unheated value measured by the

DigitizedopticalmicrographofasulfonatedPBOﬁberundcrshear

XPS clemertal composition of 400 keV N * implanted Kevlar-49
aramid fabers relative to unheated ﬁber.

croﬂhs of 400 keV 10" N‘lcm’ irradiated Kevlar-49
ﬁbcr.(A).w ﬁbchinmah'ixE, (B)dctailedviewofthc

1ntcrphasc
TEMmicro of 400 keV 10" N +lcm2 irradiated Kevlar-49
ﬁbtgbhégwho ﬁbervicw,(B)highmagniﬁcationvicwofthc

TEMmrcrogEPhso of 30 keV 10" N“/cm2 irradiated Kevlar-49
ﬁber. (A)w cﬁbervrcw, (B)dcta11cdv1ewofthcmtcrphasc

TEM micrographs of 100 keV 10" Ti‘lcm2 irradiated Kevlar-49
ﬁber. (A) who ﬁber view, (B) ﬁber-mah'ix interphasc.

TEM micro hs. of 400 keV 10" N *lcm irradiated Kevlar-49
ﬁbtgtrpaftcr. uremISS test. (A) wholcﬁbervrcw, (B) skrn-bulk
1n hase

Polaranddispersivccom erts..ofsurfacefreeercrgyforN+
implanted Kevlar-49 ﬁbergonand liquid epoxy.

Interfacial shear shergth of N " implanted Kevlar-49 ﬁbers.
Tersilc shetgth and implantation depth of irradiated aramid ﬁbers.

of 10" Nilcm2 400 keV implanted aramid ﬁber.
E3480% Iaetm'insﬁram,initialcuringkinks (B) 2%shainﬁbcrfrach1res,
5%sha1n mah1x,fracturc (Dsham) released,kmks kmksreappear.

136

138
139
140

hs of 100 keV 10" Ti‘lcm2 irradiated Specha- -1000 148
ﬁber. (A)w wliléic

ﬁbcrvicw, (B) dctailcdvicwofthcinterphasc
micrograpmdw s hof A 75 keV 10“ Ar“/cm2 and 400 keV
Hc°lcm1rrad1ated ( 5pecha -1000 ﬁbers. (B)

4 +
$311an crographs. of 1%)po 75 1(ch0 10l ﬁg [cm2 and (B) 100 keV
new . asap... We gem... Sta-a.

keV 13*Hc cm2 implantedﬁbes

813an crographs of a pulled-out droplet on a 400 keV 10"
He‘lcm imglantedpec Spec ha- 1000 ﬁber. (A) Back side, (B) overall
view, and( )bladcsrdeofthcdroplet

An overall scheme of Frieda-Crafts acylation reaction.
drfuétglhorglacidhalidccancross-linkadjacertPBOchainsat
sev s1

Compressive ofsev hrgh ormance lymer and
«MM by the hi8singlggfﬁber comgorcssion test.

AnopticalﬁnucrographofcompressionkinkbandsinPBOﬁbers.

xiii

149
151

155

156

163

165
165

Figure 8.4
Figure 8.5

Figure 8.6
Figure 8.7

Figure 8.8

Figure 8.9

Figure 8.10

Figure A.l
Figure C.1

Figure C.2
Figure C.3

Figure C.4

FigureC.5

Figure C.6
Figure E.1

Figure F.1
Figure F12

Figure E3

SEM micrographs ofa ﬁbrillated PBO ﬁber.

Superim sed XPS signals of carbon, ox
for the hour water soxhlet cxhacted
after 400°C drying.

SEM micrographs of Foster-Miller sol-gel heated PBO ﬁbers.
MESline-scanofsilicgg‘argerrmﬁber scdofaSEM micrographofa

Foster-Millersolel-g ﬁberimbeddedinancpoxy
matrix.

TEMmi
heatedP ﬁ
ﬁbcrcxterior.

AES ni err line-scan of a DEI‘A inﬁltrated wet-PBO ﬁber,
showing d1g1hzed1magcofﬁbercross-section embedded
epoxy mahix.

TEMmicrographsoonxalland succinlheatedPBO
ﬁhcrssho showrng only a(~)100 ninof rgghonperi'chahon.

Test specimen geometries for the single ﬁber compression tests.

0118
a-ﬁnagd ﬁbers regi

hs of an axially sectioned Foster- Miller sol-gel
. Note the accumulation ofkinkbandsnear the

111811

PERKIN- ELMER fracture stag samba: holders. A sample block
is shown laced at the center of the ldersand another sample
block rs own from its side view.

Major sectioning orientations of a ﬁber.

TraB‘ezoidal block reparation for radial cuts. (A Blade markin s
'on cf ) around the ﬁbegr

 

isolate ede ﬁber location. (B)Mah1

meter is himmed ﬁoff .(C) A shallow hapezoidal block 1s
mmedaroundth eﬁ
Trapezor dalblock reﬁrationforlateralcuts. A)Thcﬁberis .
1nihallycovcred.mah1xishimmedun aﬁberportion 1s

exposed. sed'ﬁberThc. lock 1s formed 1n the region adjacent to the

SEM microgm hs of an axially cut nickel coated carbon ﬁber-
epoxy com Cut was made at 6° angle from the ﬁber long
axrs. (A) rapczoidal block face. (B) Oblong ﬁber cross section.

SEMmicrographofaradiallycutcoppcrcoatedcarbonﬁhcr.

SEM micrographs of sulfonated samples and their elemental line-
scans.

Major sectioning orientations ofa ﬁber.

midal. block preparation for radial cuts. (A) Blade markings
faccisolatc the ﬁber location .(B)Mah'1x around the ﬁber
perimeter is himmed off. (C)A shallow hapezoidal block 1s

himmed around the ﬁber.

idal block on for lateral cuts. (A) The ﬁber 1s
iruhallyco ) mah1xis himmed until a ﬁber
exposed.ﬁ(be)r'1hcoc l ockisformcdinthcregion adjacent

 

xiv

orcand

166
175

178
179

180

184

186

197
210

210
211

211

212

212
219

225

<am~evsfrnvpafffrﬂﬁgpﬂ“79°999??>

hMmMMme

Interface area (mm’)
Fiber cross-sectional area
Mahix cross-sectional area
Molar concenhation of component i
Concenhation of species 1
Initial concenhation of species 1
Fiber diameter (pm)
Diﬁ'usion coefficient
Helmholtz free energy per unit area
Wetting force (mg)
Fiber compressive modulus (GPa)
Mahix compressive modulus (GPa)
Axial ﬁber elastic modulus (GPa)
binding energies of the K shell atomic orbital
binding energies of the L shell atomic orbital
binding energies of the M shell atomic orbital
Kinetic energy ofan clechon emitted by a KLM Auger hansition
Kinetic energy of photoelechon orbital
Binding energy of the atomic orbital
Gibbs free energy
Mahix shear modulus (GPa)
mass-transfer ﬂux of species 1
Reaction equilibrium constant
Fiber critical length (mm)
Theoretical ﬁber critical length (mm)
Number of moles of component i
Presure of the phase (atm)
Rate of production per volume of species 1
Gas constant '
Enhopy of the system (1)
Temperature (‘0)
Volume of the phase (1)

Surface excess of component i = n/A

Surface tension (dynelcm)

Dispersivc component of surface energy of liquid (dynelcm)
Polar component of surface energy of liquid (dynelcm)
Total surface energy of liquid (dynelcm)

Dispersivc component of surface energy of solid (dynelcm)
Polar component of surface energy of solid (dynelcm)
Toml surface energy of solid (dynelcm)

Fiber surface energy (dynelcm)

Mahix surface energy (dynelcm)

Fiber and Mahix strain

Cross-chemical potential of component i = df/c,

Liquid contact angle (degree)

Fiber compressive shength (MPa)

Mahix shess (MPa)

Fiber tensile shength (GPa)

Interfacial shear shength (MPa)

Chemical potential of component i (J I mol)

Fiber Poisson’s ratio

Spectrometer work function.

CHAPTER I

 

Introduction

 

1-1 W

The term ”high performance polymer ﬁbers“ refers to organic ﬁbers that posses high
axial tensile properties comparable to those of the inorganic reinforcing ﬁbers (Dctcresa
1985). These ﬁbers have tensile properties that are at least an order of magnitude greater
than more common textile ﬁbers. High performance ﬁbers possess a unique combination
of stiffness, high shength, high toughness, and low density that makes them an athactivc
alternativetoinorganicreinforcing ﬁberssuchasglassandcarhonﬁhcrs. Zahretal.
(1989) have presented an overview discussion of the unique properties of aramid ﬁber
polymer composites. In general, on a per weight bases, high performance polymer ﬁbers
have a signiﬁcant advantage over other inorganic reinforcing materials. Figure 1.1
shows a plot of speciﬁc tensile shength and modulus of some reinforcing ﬁbers as well
as some conventional materials (Agrawal er al. 1980, Kumar 1989). Tensile properties
of the high performance polymers ﬁbers approach their theoretical maximums, which is

achievedbyahighdcgreeofpolymerchainalignmentandreductionofdefectsinthe

1

2
ﬁber shucture. However, the high performance polymer ﬁbers, generally exhibit weak

adhesive properties. The level of ﬁber-mahix adhesion conhols many properties of
ﬁber-reinforced composites such as hansvcrse shength, shear shength, and ﬂexure. The
weak adhesive properties of high performance polymer ﬁbers signiﬁcantly limits their
shuchrral applications.

Signiﬁcant amounts of research have been devoted to the study of the interfacial
properties of glass and carbon ﬁbers and many surface heahnent techniques and coupling
agents have been developed. For glass and carbon ﬁbers these surface heahnent
techniques can double or hiple the interfacial bond shength (Riggs at al. 1982, Wu 1982,
Bjorkstcn er al. 1952). For the liquid crystalline polymer ﬁbers, many workers have

attempted to obtain similar adhesion improvements with interfacial heahnent methods but

 

 

 

 

they have been generally unsuccessful.
34 PBO
”\ O
E Spectra-1 000
°~ O
E 3 i Technora
8 Kevlar-29 Kevlar-49 . PST
3, O O . Kevlar-149
7)
£2 - .. S-Glcss
; E—Gloss HT graphite
or . . HM graphite
E1 . . Boron .
in”
g .T Glcsts
un 5 en
.33 0 . 5 Stgel, Aluginum 1 . j . .
0 50 50 200 250

1 0 1
Modulus? Density (GPo.cm’/g)

Figure 1.1 - Speciﬁc tensile properties of reinforcing ﬁbers and conventional bulk
materials.

3
Cooke (1987) and Allred (1983) have documented various attempts on the

development of surface heahnent techniques for the Kevlar aramid ﬁbers. These works
assume that the low adhesive properties of the aramid ﬁbers are mainly the result of a
chemically inert ﬁber surface (Penn et al. 1985) and to a lesser degree mechanical
properties of the ﬁbers (Dr-ml 1983). Despite many efforts, promising coupling agents
have not been developed (Penn et a1. 1983) and surface heahnents such as surface
oxidation techniques and plasma heahnents improve adhesion but are usually
accompanied by signiﬁcant loses in ﬁber tensile shength (Wcrtheimcr at at. 1981).
There have been approaches that suggest forming chemically active groups on the aramid
ﬁber surfacecandouble the interfacialhond shength (Allredetal. 1985, Wu etal. 1986)
without tensile shength losses but these results have not been substantiated. A previous
study of ararnid-epoxy adhesion by the Kalantar and Drzal (1990‘) has shown that the
morphologyofmcammidﬁbersandnotthesurfacechenﬁshyofhwﬁberishmihng
their adhesive properties.

This study examines ﬁve approaches to chemical and morphological modiﬁcation of
surfaceandbulkpropcrtiesofthchighperformanccpolymerﬁbers. Table 1.lliststhc
cxammedappmachesandmdrexpwtedeﬂechmdreﬁbamorphologicalandchenucd
properties. These approaches systematically explore shucturc-property relations of the
high performance polymer ﬁbers. Each heahnent technique produces different extent of
ﬁbuchemicalandmorphologicalalteaﬁmuﬂrataﬂowaparﬁcuhraspectofﬁber
adhesivebehaviortoheexamined. Effectsofthesetechniquesonstruchrre-property
rdahmsofhighperformancepolymcrﬁbmamdemﬂedmmdrnspechvechapm.

4

Table 1.1 - Effects of different approaches on the morphology and chemishy of high
performance polymer ﬁbers.

Treahnent ChemistryBulk SMorphologyulk
Surface

Coupling Agents, Fiber Coatings

 

     
    
   
   
   

 

PlasmaandCoronaTreatments

 

Chemical Treahnents

 

Ion Implantation

 

Structural Modiﬁcations

 

 

1212mm
ihcgoalsofthisstudyamzmmveshgatcshucmmlpmpaﬁesofhighperfonnmce

polymer ﬁbers that affect their adhesive behavior; to develop a fundamental
understanding ofthe ﬁber shuctural limitations; to evaluate several novel techniques that
can enhance the ﬁber adhesive performance properties; and to suggest ways to improve
adhesivcproperticsoftlrehighperformancepolymerﬁhers. V
AnexpcrimentalovcrviewofthcdissertationisshowninFigurelJ. Although,thc
maindrcmcofthisdissatahmisdrerdahmsbctwemdrcadhcsivcmdshucmm
properhesofﬂwhighperfomancepolymaﬁbm,mesmdyﬂsomveshgatessevaﬂ
important polymer heahnents that merit their own particular discussion. Hence, the
dissahﬁmisorgmﬂedhbsevualdnptersmachwnhiﬂngadiscussimofaspedﬁc

 

Structure-Property Relations of
High Performance Polymer Fibers

 

 

 

  
     
   
 

 

 

 

Tensile
Interfacial Shear

Temp. dependence
Compression
Surface Energy

Epoxy (four curing

Figurel.2- Overviewoftheexperimentalplan.

 

Treatment
Techniques

 

hemical Treatmen
Fluorination

Sulfonation (gas, solu.)

Ion 1m lantation
"1» T1, Ar", N", He

Second Phase Infusion
(Sol-Gel, Epoxy)

Chemical Cross-linking
(Frieda-Cram)

 

6
type of polymer heahnent. Conclusions at the end of each chapter mainly address the

examined heahnent, but the relevancy of the conclusions to the main theme of the
dissertation is also discussed. The main conclusions of the dissertation are ﬁnally
coalesced in the last chapter to provide a comprehensive overview of the shucturc-
property relations of the high performance polymer ﬁbers.

Chapter2detailstheexperimentaltechniquesusedinthisdissertahon. Some
discussions on background literature for the examined techniques are also presented.
Chapter 3 provides general discussion and observation on the adhesive properties of high
performancepolymcrﬁbcrs. Datapresentedinthischaptcrarcusedasthebaselinedata
throughout the dissertation. Chapter 4 examines effects of coupling agents and ﬁber
coatings on adhesive properties of the high performance polymer ﬁbers. Chapter 5
cmwenhatesmdrcplasmaandcormasurfaceheahnurtsofPBOandSpecha-looo
ﬁbers. Chapter6presentsadiscussionofsulfonahonsurfaceheahncntandabrief
examination of the ﬂuorinated Kevlar-49 ﬁbers. Sulfonation of polymers both in ﬁber
andsheetformsamalsomveshgatedmdcvdopanmdcrstandingofmass-hansfcr
phenomena that may limit the extent of polymer heahnent penehations. In Chapter 7,
eﬂ'easofionimphntahononnwchanicalandchemicalpmpaﬁesoftheammidand
polyethylene ﬁbers-arc investigated. Chapter 8 examines approaches to inﬁlhatc the high
performance polymer ﬁbers and reinforce hansvcrse cohesive properties of the ﬁber.
Chapter 9 presents the conclusions of this dissertation and proposes some
recommendation for further developments.

CHAPTER 2

 

Experimental

 

2-1 MATERIALS
Aramid ﬁbers examined in this study were Kevlar-29°, Kevlar-49°, Kevlar-149°

ﬁbers (13.1. (111 Pont, Wilmington, DE) and Technoraa ﬁbers (Tcijin Limited, Japan).
Polyethylene ﬁbers were the ulha-high-modulus SpechaO-1000 (Allied Signal,
Morristown, NJ). To eliminate possible interference by ﬁber sizing, the ﬁbers were
soxhlet exhacted in absolute ethanol for 24 hours and then dried overnight at 125°C.
PBO (p-Phcnylenc BenzobisOxazole) ﬁbers were provided by Dow Chemical (ref. #
XV-0383-C8700975-008). These PBO ﬁbers contained no sizing and were used “as
received“. Otherexarnincdﬁber'swercAS-4Ocarhonﬁbe1's(Hercules,Magna,U'1'),
and B-glass ﬁbers (Pittsburgh Paint Glass, Pittsburgh, PA).

. The epoxy resin was D.E.R. 331 which is a Diglycidyl Ether of Bisphenol A
(DGEBA) epoxy (Dow Chemical, Midland, MI). Four different curing conditions,
ambient, 75-100°C,'75-125°C and 175°C curing were selected for this study.

8
For the ambient curing condition, the curing agent was DiEthyleneTriAmine (DETA)

(Aldrich, Milwaukee, WI). The D.E.R.331/DETA system contained a 11.0/ 100 mass
ratio of curing agent to epoxy. The mixture was degassed in a vacuum oven for 15
minutes at -29 in.Hg (gauge pressure). DETA is highly reactive and its epoxy mixture
was degassed only at room temperature to avoid gelling. At low curing temperatures this
epoxy system was still too brittle for the critical length testing. Subsequent post—curing
of the DEI‘A systems was required to increase its fracture strain. The post-curing time
and temperatures were determined by the glass transition temperature (1“) of the matrix.
During the post-cure, the oven temperature was maintained below the T, of the matrix.
This was to avoid building up thermal stresses. At each post-curing temperature, initially
theT,wasonlyafewdegreesabovetheoven temperature. Afteracertaintimetheglass
uansiﬁontempuammwasmcreasedanowingtheoventemperammmbemisedmsteps
of 10°C. For the DETA systems, the curing schedule was: 25°C for 48 hours, followed
by 4 hours post-curing at 40°C, 50°C, 60°C, 70°C, and 80°C. Subsequent TMA
examinations showed only ~ 0.1 % post-curing shrinkage for this epoxy system (Kalantar
et al. 1990‘).

For the 75-100°C and 75-125°C curing, the curing agent was m—PhenyleneDiAmine
(mPDA) (Aldrich, Milwaukee, WI). For the D.E.R.331/MPDA system, a 14.5/ 100
mass ratio of MPDA and epoxy were combined. The 75-100°C curing was used for the
droplet test and involved curing for 24 hour at ambient temperature followed by 2 hours
at 75°C and 3 hours at 100°C. The 75-125°C curing was used for fragmentation and

singleﬁbercompressiontests,andconsistedofcuringat75°Cfor2hoursandpost-

curing post-curing at 125°C for 2 hours.

For the 175°C curing condition a mixture of two curing agents MPDA and
DiEthleolueneDiAmine (DETDA) (Ethyl Corp. , Baton Rouge, LA) were combined.
For D.E.R.331IMPDA/DEI'DA system, a 7.25/ 100 mass ratio of MPDA and a
11.75/100 mass ratio of DETDA were mixed. The mixture was degassed in a vacuum
oven at 75°C for 5 minutes at -29 in.Hg to reduce the viscosity of the solution as well

as removing entrapped bubbles. The 175°C system was used for the fragmentation test,

and involved a 3 hour cure at 175°C.

 

Characterization oftheinterfacial adhesionwasdoneby fragmentationanddroplet
techniques. Drzaletal. (1991) havepresentedacomprehensivereview of theseadhesion
tests. ﬁgureZJshowsaschematicofasingleﬁberfragmentationprocess. Axialstress
istransferredtotheﬁberthroughshearattheinterface(A). Theﬁberaxialstressrises
untiltheﬁberfracturesu'engthisreachedm). Continuedapplieationofstresstothe
specimenrendtsindwrepeﬁﬁmofmeﬁagmentaﬁonpmwssunﬁlanﬂwﬁagment
lengthbecomeshorterthanﬂrelengthneedmtransfertheﬁacmresuess(C). This
maximumfragmentationlengthiscalledthecriticallengthag.

Fmﬂreﬁagmulmﬁmtesgasingleﬁberwasembeddedinadogboneslmpedpolymer
matrix. Thedogbmesamplewassubjectedtoatensileloadusingatensiletesﬁngjig
(Figun21)andtbeﬁberﬁagmentaﬁonprocesswasmonitoredunderan0pﬁeal
microscopeuntilthecriticsllengthwasreached. Forthepolymerﬁbersthatexhibita

10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a
A E
(I)
LENGTH
B {3’
— - — E
LENGTH
3
c E
% _ _ — 5

 

 

 

LENGTH

ﬁgureZJ-Schematicofsingleﬁberfragmentationprocess.

 

ll

    
 

mi
N

“wan-m." "v—‘wo :ucwvv's..,.,._

-. ..~, 2": ,r. -
'1.” C m» ,l to O

 

Figure 2.2 - A sample mold and tensile jig used for the single ﬁber fragmentation test.

12
ﬁbrillar fragmentation, the average critieal lengths were obtained by counting the number

offailedregionswithina22mmﬁberlength. This221engthwasmarkedbyaglass
slide over the sample. For the carbon and glass ﬁbers the fragmentation process
produces easily distinguishable ﬁber ends, permitting a direct measurement of each
fragment length. The relation between critieal length (1,) and interfacial shear strength
(1) is easily obtainable by a force balance between the ﬁber surface shear load and its
tensile strength (Kelly er a1. 1965):
sud—2" (5’: ) (2 . 1)

where oiistheﬁbertensilestrengthanddistheﬁberdiameter.

Forthedroplettest,aﬁberisembeddedinadropletofamatlixand tensileload
needed to pull the ﬁber free from the droplet is measured. Plots of each debonding load
versesmeﬁber-dropletmterfadalareapmvidesameasumofmeirmterfacial shear
strength. FigureSJ showstheapparatususedforthedroplettest. Theexperimental
procedure for thedroplet test has been described by Rao etal. (l991°). Droplets of
liquid epoxy are deposited on the ﬁber by lightly passing a syringe of resin over the
ﬁber. The surface tension of the epoxy pulls the liquid into concentric droplets. To
reducelossofcuﬁngagentfommﬂdmplasbydiffusionuhightemperamresmm
er a]. 1991'), the droplets are allowed to gel at ambient conditions for 24 hour before
they are cured at 75°C for 2 hour and post-cure at 100°C for 3 hour. About a dozen
droplets were deposited on each ﬁber sample. Only the droplet with perfect cylindrical
symmetry were used for the measurements. Further discussion of the droplet test have
been presented by Miller et a1. (1987), Gilbert et al. (1990) and Mcalea et al. (1988).

1 LOAD CELL

2 BLADE MICROMETER

3 X-Y TRANSLATION STAGE
4 ACTUATOR

5 BASE

n run-«4‘4 a
.s

. M’ .I ~* A.
- s N J. 2’» “JWWJ’
e .. x
. 3? .
’U

 

ﬁgure 2.3 - Droplet test apparatus.

14

2.3 W

Contact angles of ﬁbers with three liquids, water, ethylene glycol, and methylene
iodide were measured using a Wilhelmy technique. The instrument is similar to the
setup used by Hammer et al. (1980). A single ﬁber is earefully mounted on an
aluminum hook with a cyanoacrylate adhesive. Fiber and hook are then suspended on
the arm of a Cahn microbalance. A small beaker of the liquid is slowly raised to the
ﬁber tip. The force before and after contact with the liquid is recorded by a digital data
acquisition system. The instrument (Waterbury 1991) automatically lowers the ﬁber ﬁve
times at 0.5 mm steps and records the force changes after a few seconds of equilibration

time. The measured force (F) is related to the contact angle through use of the equation:

F 87211:! c036 (2-2)
where 7,} is the total surface free energy of the probing liquid, d is the ﬁber diameter,
and Gisthecontactangle. Themeasured contactangleswiththereferenceliquidsare
then converted to polar and dispersive surface energy components using the method
proposed by Kaelble et a1. (1974):

MI was» .55.),7J15 (2.3,
2H 71.

where 1', 1‘, and 7‘ refer to polar, dispersive, and total surface energies, and subscript

 

L and 8 refer to liquid and solid material. The measured contact angles of equation (2.2)
were converted into the polar and dispersive components using Program WILHEMY

(Appendix G). This program performs a regression ﬁt to equation (2.3) using the surface

15

Table 2.1 - Surface energy components (dynelcm) of reference liquids used for Wilhelmy
contact angle measurements (Hammer et al. 1980).

 

Ethylene Glycol

 

Formamide

 

Methylene iodide

 

 

 

energy components of the probing liquids. Surface energy components of some reference
liquids are listed in the Table 2.1.

2.4 W

Baun (1980) has listed more than eighty different surface analysis techniques to study
adhesion. For polymer-polymer interfaces, the number of these techniques is limited
because of the fragile nature of polymer surfaces. Gillberg (1987) haspresented an
overviewofsomeofthesetechniquesmchasAES,ESCA,andelecnonnncmscopy for
polymer surface analysis. Occhiello et at. (1989) also reviewed spectroscopic techniques
for characterization of polymer composite interfaces. A brief description of several
relevant analytical techniques is presented here.

16

2.4.1 MW
Surface analysis by X-ray Photoelectron Spectroscopy (XPS), also known as Electron

Spectroscopy for Chemical Analysis (ESCA), is accomplished by irradiating a sample
with a monoenergetic x-ray beam and analyzing the electrons emitted. Mg Ka x-rays
(1253.6 eV) or Al Kat x-rays (1486.6 eV) are commonly used. These irradiated photons
causephotoionizationoftheatomsinthesurfaceregionofthesamplethatresultsin
emission of two types of electrons, photoelectrons and Auger electrons. Probabilities of
thcinteractionsoftheseemittedelectrons withmatterfarexceeds thoscoftheirradiated
photons, so while the photons can penetrate the solid sample in orders of l-10 um,
emitted electrons can escape only tens of Angstroms of solid. Therefore, the electrons
usedinXPSanalysisofthesolidsoriginatewithinwnsofAngstromsofthetopsurface
region. The emitted electrons have kinetic energies (En) given by:

Eng-bro -E‘-¢, (2.4)
where by is energy of the x-ray photons, Em is the binding energy of the atomic orbital
from which the electron originates, and 4:, is spectrometer work function (energy needed
for the electron to leave the spectrometer).

'l‘heelectronsleavingthesamplearedetectedbyanelectron spectrometeraccording
totheirkineticenergy. Theanalyzerisoperatedtoacceptelectrons thathaveenergies
within a ﬁxed narrow range, this ﬁxed window is called the "pass energy", hence, the
narrowerthepassenergythehighertheresolutionoftheenergy scan. Scanning for
different energies is accomplished by electrostatically retarding the electrons before they
reachthedetectors. 'I'hisretardationvoltagemaybevariedfromaerouptothephoton

17
energy (energies of the electrons emitted cannot exceed the energy of the ionizing

photons). The probe for XPS is X-ray photons, which are less disruptive to the surface
than the electron beam of ABS. XPS is inherently more sensitive than ABS to the
chemical environment of the elements but examines a larger area of the sample surface.

For a typical XPS investigation where the surface composition is unknown, a wide
scan meyspecu'umofﬂtesurfaceisobtainedﬁrstmidulﬁfyﬂleelementsﬂlatam
present. Once the elemental composition is determined, narrower detailed scans of the
selected peaks are used for a more comprehensive analysis of the chemical composition.

XPS analyses were conducted on a Perkin-Elmer PHI 5400 ESCA system using an
Al Ka toroidal monochromatic source (PHI 10-410). Spectra were collected at a base
pressure of approximately 10’ torr and electron take-off angle of 65° using a position
sensitive detector (PSD) on a 180° hemispherical analyzer set at 44.75 eV pass energy
for the survey scans (0-1000 eV) and 35.75 eV for the narrow scans of the elemental

regions used for composition analyses. Size of the analysis area was 1x3 mm.

241W

The Auger Electron Spectroscopy (AES) technique for the chemical analysis is based
onmesimﬂarpmcessastheXPStechnique,excepthESﬂleanalysissurfaceis
irradiatedwithabeamofelectrons. Theincidentelectronsionizeatomsofthesurface
creatingvacanciesintheirinnerelectronshell. Theionizedatomsrelaxtoalower
enagysmtebyﬁlﬁngmeinnershenvacanciesbyﬂwdecmmﬁomﬂlelowermergy
shells. ‘I‘hisrelaxationprocessreleasecharacteristic “Auger electrons“ aswellasx-ray

18
photons that can be used to identify the excited atoms. Auger transitions are typically

denotedbythreecapitalletters suchasKLL, KLM, LMM, etc. Theletterontheleft
refers to the electron shell in which the initial vacancy occurred; the middle letter refers
totheshell from whichanelectron comesto ﬁlltheinitialvacancy; and theletteron the
right refers to the shell from which the Auger electron is emitted. Therefore, kinetic
energy ofan emitted KLM Auger electron (En) is given by:
am-zK-zL-z,-¢, (2.5)
where E‘, EL, and En are the binding energies of the atomic orbital from which the
electrons originate, and ¢ 3 is the spectrometer work function.
SimilarmtheXPS,theAFStecthueonlyexanunesmedecn’onsﬂmtoﬁginate
within the tens of Angstroms of the top surface region. The principle advantage of the
movermtechniqueistheabilitytofocusandscantheprobing electronbeam, and
obtain information on the spatial distribution of surface elements at high magniﬁcations.
However, to analyze the surface composition in practical time scales requires ABS to
utilize a fast electron spectrometer. Therefore, AES technique is less sensitive to the
chemical environment of the elements than the XPS technique.
AHAESanalysiswerecarﬁedoutusingaPerldn—Elmeer66OScanningAuger
Microprobe. Samples were analyzed at 1000 to 30000x magniﬁcations. ABS beam
conditionsforanalyseswere1.5tolOnAbeamcurrentand3t010kaeamenergy.
For the surfaces of unknown composition, initially a survey spectrum was obtained to
determine the surface composition. The spatial distribution of the interested element was
then monitored by its AES signal peak height. Signal intensities were plotted in line-scan

19
or map fashions.
For some samples, a short (> 50 nm) ion beam sputtering of the analysis surface was
conducted to remove the surface contaminates. However, longer sputtering times were
avoided because a high dose sputtering of polymers would preferentially remove the non-

carbon surface elements and produces a carbonized surface composition (see Chapter 7).

2.5 MW:

Microscopic techniques such as light microscopy, scanning electron microscopy
(SEM), and transmission electron microscopy (TEM), are useful in the study of
interfacial microstructure. Light microscopy requires little sample preparation and is
non-destructive, but its magniﬁcations are low. SEM microscopy can deliver higher
magniﬁcation and resolution than light microscopy but provides only topographical
information. TEM is the most useful microscopic method for interfacial investigations.
Samples as thin as 50-60 nm can be shaved from the interface by ultra-microtoming.
TEM can show the details of the interface up to 500,000x magniﬁcations. A procedure
for ultra thin microtomy of composite material is presented in Appendix F.

Observation of ﬁber-matrix interfacial morphology was obtained by transmission
electron microscopy ('I'EM) using a JEOL CXlOO T'EM. In the TEM rnicrographs
presentedinthisdissertation,directionofthesectioningisshownbyanarrowand
designated magniﬁcations are shown by scale bars. Topography of sample surface were
characterized by scanning electron microscopy (SEM) using a JEOL ISM-T330 SEM.

CHAPTER 3

 

Adhesive Properties of

High Performance Polymer Fibers

 

Thischapterpresents somediscussionsontheadhesiveand structuralpropertiesof
high performance polymer ﬁbers. Tensile and adhesive behavior of the untreated ﬁbers
areexaminedandcomparedtotheirpredictedvalues. Thesediscussionsprovidevaluable
insights 'to mechanisms that control the adhesive properties of high performance polymer
ﬁbers. Thedatapresentedinthischapteralsoserveasthereferenceproperties forother
results of this dissertation.

20

21
3-1 W

To produce high performance polymer ﬁbers a highly ordered extended chain
morphology is required. Flaws and cracks that are detrimental to the ﬁber strength must
also be minimized. The most successful high performance polymer ﬁbers have been
prepared from rigid-rod liquid crystalline polymers. The liquid crystalline polymers with
their highly ordered liquid morphologies are good candidates to initiate the high order
required for the high performance polymer ﬁbers. Indeed, the commercial synthesis of
high-modulus ﬁbers came about with the advent of rigid-chain polymers and ﬁber
spinning from their liquid crystalline solutions. Today, the primary commercial high
performance polymer ﬁbers are made from liquid crystalline polymers. An ultra-high-
modulus polyethylene ﬁber that is spun from a gel solution has also been
commercialized.

A "liquid crystal" is a substance with optical anisotropy like a crystal but with a low
viscosity like a liquid. The liquid crystalline state with one-dimensional order is called
nematic. In nematic solutions, the long axis of molecules are generally parallel, but their
positions may be random. The nematic solutions of polymers usually involve rigid-chain
polymers. The rigid-chain polymers are elongated molecules with ﬂat segments such as
benzene rings and posses high rigidity along their long axis. Monomer structures of
some liquid crystalline polymer ﬁbers are shown in Figure 3.1. These rigid-rod
polymers exhibit extremely high viscosities when melted or tend to decompose before
melting at high temperatures (>250°C). Therefore, to prepare these ﬁbers, organic

solvents or inorganic acids are employed to dissolve them into liquid crystalline solutions

22

before spinning. High performance polymer ﬁbers are generally produced from nematic
liquid crystalline polymer precursors.

There are three main types of liquid crystalline polymers which exhibit the rigid-rod
liquid morphology: aramids (aromatic polyamides) such as p-phenylene terephthalamide
(PPTA); aromatic heterocylic polymers such as p-phenylene benzobisoxazole (PBO) and
p-phenylene benzobisthiazole (PET); and the family of thermotropic aromatic
copolyesters (Sawyer er a1. 1986) such as naphthyl-phenyl copolyesters (NTP). White
(1985) has presented a historical survey of development of liquid crystalline polymers.
Itisinterestingtonotethatsomeofthestrongestnaturalﬁbers suchassilkandcellulose

 

0

II
N— c_<: :>.c
I ll
H o
KEVIAR-p-PhenylmeTerephtlnlAmideMA)

Eng-is» 4&0 @3- ”Q8

TBCHNORA - p-Phenylene DiPhenylEther TaepbthalAmide

N N
- r c \\
°\O:.:o/°©’

PBO-p—PhenyleneBalzobisOxazole

O

 

 

 

Figure 3.1 - Monomer structure of some liquid crystalline polymers used in high
performance polymer ﬁbers.

23
also form liquid crystalline states when dissolved in solvents. There are also many

natural ﬁbers such as those present in coconut husk, pineapple, banana, and bamboo,
which exhibit morphologies similar to the synthetic liquid crystalline polymers (Chand
er a1. 1988).

A process for the formation of aramid (PPTA) ﬁbers has been reported by Morgan
et al. (1989‘). Aramid ﬁbers are produced by the condensation polymerization of
terephthaloyl chloride and p-phenylene diamine. The PPTA is polymerized using a
stoichiometric ratio of the reactants. The HCl formed during polymerization is
neutralized with a NaOH wash. The PPTA polymers are then dissolved in a
concentrated H180, solvent to produce low viscosity PPTA liquid crystalline solution for
the ﬁber fabrication. The solution (~20 wtﬁ PPTA) is extruded at 80°C from spinneret
oriﬁces into ﬁber form by a ”dry-jet wet spinning” process (Blades 1973). The resulting
yarns are neutralized with NaOH and water ”washed” to remove the resulting N580.
salt. Further drying and drawing treatments increase the ﬁber stiffness and strength.

Reviews of aramid ﬁber morphology have been presented by Kalantar et a]. (1990')
and Panar er al. (1983). Aramid ﬁbers consist of cylindrical crystallites about 60 nm in
diameterandabout200nminlength. Theserodsarealignedintheﬁberdirection,
longitudinally connected by macromolecules passing through the rods and radially
connected by hydrogen bonding. The ﬁbers have a different morphology between their
intaior and exterior regions because of the extrusion and coagulation processes of their
fabrication. This morphology ofthe ﬁber is called the “skin-core" morphology and has

beenpostulatedtobethemainmechanismresponsible fortheweakadhesiveproperties

24
of the ﬁbers (Kalantar et al. 1990 ", Greszczuk 1969). Similar skin-core morphologies

for other liquid crystalline polymers such as PBO ﬁbers have been reported (Krause et
al. 1988, Woodward 1988) which are also attributed to the extrusion and coagulation
steps in their production. There are also other solvent ﬁber processing techniques such
as those used for the fabrication of Technora aramid ﬁbers (p-pherlylene/0,4-
diphcnyletherwrephthalamide) thatareexpected toresultinalowerdegreeofskin—core
morphological difference than the ﬁbers spun by a dry-jet wet spinning process.
(Technora ﬁbers are spun from a N-methyl pyrrolidone solution and then into a water
bath, Morgan 1989 " ).

Commercial high performance ultra-high-molecular-weight (UHMW) polyethylene
ﬁbers are manufactured by a gel-spinning process (O’Sullivan 1991). The starting
polyethylenehasamolecularweightintherangeof2to6millionwhichisas muchas
100 times more than the commercial grade high-density polyethylene. A solution of ~5
wt% of the UHMW polyethylene in decahydronaphthalenc solvent is extruded through
spinneretstoformgel-likeﬁbers. Theseﬁbersarethenstretchedtoabout300x their
lengthatatemperamreclosetothepolymer’s meltingpointof ~145°Ctoformtheﬁnal
highperformancepolymerﬁbers. Theﬁnalﬁberspossesgreaterthan95$chain
orientation and upto 85% degree of crystallinity. For a given polymer molecular
weight, the ultimate mechanical properties of gel-spun ﬁbers are controlled by choice of
solvent, polymer concentration, and spinning temperature (Kalb et at. 1980). Other
ﬁbers such as ethylene-vinyl alcohol (EV OH) ﬁbers have also been produced by gel-
spinning procedures (Schellekens et at. 1990).

25
For high-modulus polyethylene ﬁbers, different ﬁber morphologies may result from

the gel-spinning process depending on the drawing conditions of the ﬁber (Hofmann et
al. 1989). Under low drawing, a ”slush-kebab“ morphology which consists of a central
rod or ribbon with overgrowths of folded chains are formed (Billmeyer 1984). Under
high degrees of drawing, the initial shish-kebab morphology transforms into extended
polymer chains (Brady at al. 1989) and a ﬁbrillar microstructure results. Commercial
high performance ultra-high-molecular-weight (UHMW) polyethylene ﬁbers are hot-
drawn under high stress conditions and exhibit the extended chain ﬁbrillar morphology.
Thealmostperfectchainaﬁgnmentofhighpaformancepolymaﬁbasrennmmmdr
hightensileproperties. Tensilepropertiesofthehighperformancepolymerﬁbersare
relatively close to predicted values of their perfect crystal properties (Smith 1990) which
is evidence for their highly ordered morphologies. However, composites made with high
performance polymer ﬁbers generally exhibit low transverse properties, independent of
matrix (Mittleman er al. 1985, Smith et al. 1985, Brady er al. 1990). The level of
ﬁber-matrix interactions controls many properties of ﬁber-reinforced composites such as
transverse strength, shear strength, and flexure. The low transverse properties ofhigh
performance polymer ﬁbers signiﬁcantly limits their structural applications.
Compositematerialscombmetwommomdissimilarmatmialsmyiddacomposite
with properties superior to those of its constituents. For example, reinforcing glass
ﬁbers embedded in a polymer matrix can form strong ﬁber-glass panels. Since
composites, by deﬁnition, are heterophase materials, interphase regions are inherent
featmesoftheirstructure. Thecompositeinterphase maybedeﬁnedasatransition

26
region between constituent materials, across which a gradation of mechanical and

chemical properties occurs. The ”interphase" contains the ”interface”, formed by the
contact of two surfaces plus the region on both sides where the material is different from
the bulk. The structure and composition of the composite interphase controls the extent
of interaction between its constituents and signiﬁcantly affects the behavior of the
compoﬁte (Grezczuk 1969, Chamis 1974, Tsai et al. 1974). In general the optimum
condition of the composite interphase depends on the particular application and its
expected loads. For example, for continuous ﬁber reinforced composites a strong
interphase improves off-axis properties of the composite such as transverse strength and
ﬂexure. For some applications such as fracture toughness, however, a weak interphase
is desirable.

There are two approaches to the investigation of the composite interphase. One
approachdedswidldremterfacialaspecmofbondfOMngandcmwenuawsmdw
physical-chemistry of the interphase. The other approach deals with macrostructural
aspectsand mechanicalanalysis oftheinterphaseregion. Thetwoapproaches mustbe
combined for a complete explanation of the composite interphase. An extended review
ofﬂrecumthteramremdleamnud-epoxyinterphaseanddleadhesionrdated
properties ofthearamid ﬁbers has beenpresented by Kalantaretal. (1990‘).

27
3.2 W

Aramid ﬁbers examined in this study were Kevlar-29, Kevlar-49, Kevlar-149 (E.I.
du Pont, Wilmington, DE) and Technora ﬁbers (Teijin Limited, Japan). The
‘ polyethylene ﬁbers were the ultra-high-modulus Spectra-1000 (Allied Signal, Morristown,
NJ). PBO ﬁbers were provided by Dow Chemical (ref. I XV-0383-C8700975-008). To
eliminatepossibleinterferencebyﬁbersizing, theﬁbersweresoxhletextractedin
absolute ethanol for 24 hours and then dried overnight at ambient conditions.

Four epoxy systems were used: DER331/MPDA/DE1'DA 175 °Cl3hr, DERB3 l/DEI‘A
and ambient cure with the fragmentation test, and DER331/MPDA RT124hr/75 °C-
2hrl100°Cl3hr with the droplet test (described in Chapter 2). Single ﬁber tensile
strengths were measured by ASTM D3379 tensile test. Surface compositions of the
ﬁberswerecharacterizedbePS. Fiber-matrixinterfacial morphologywasexamined
byuanmussimdecuonmimowopya'ﬁhoofulmminmiaomnwdsecdmswtnormal
totheﬁberaxis. Fibersurfaceenergieswerecharacterized byWilhelmy measurements
using water, ethylene glycol and methylene iodide as probing liquids. These
experimental conditions are detailed in Chapter 2.

28

3.3W

Measumdtensileproperﬁesoftheexmninedhighperfonnancepolymerﬁbusam
listedinTable 3.1. Thescdataareusedthroughoutthisdissertationasthebaseline
properties. Figure 3.2 compares the tensile modulus of the high performance polymer
ﬁberswiththeirtensilefracturestrain. Thehighermodulusﬁberstendtoshowlower
fracturestrainsandviseversa. Generally,tensilcfracturesaredefectcontrolled,while
tensilemoduliarecontrolled bythebulkproperties ofthe ﬁbers. However, Figure3.2
ahowsacorrelationcouldexistbetweenthetwoproperties. Thehighmodulioftbese
ﬁbershavebeenachievedbyordaingthepolymerchainsmdwaxialﬁbadhecﬁm.
AsmeextaltofﬁbaoﬁentaﬁmmcreasesdwovaaUﬁbawnsilemodulusmcreases,
butmemovementsofmechainsbecomcmomresuictedanddleﬁbermsnainis

 

 

 

 

180 - 0 P30
1515b '
o
‘g',’ I Kevlar-149
3 140 r
.0
é’
2 120 ”
.2 I Kevlar—49
12100 -
2 _ I Kevlar—29
- Technora
b. 80 _ O
~ . Spectra
60 J I L l a l a J n l 4 l a
1 2 3 4 5 6 7 8

Fiber Fracture Strain (7.)

Flgure3.2- Talsilefracturestrainofhighperformancepolymerﬁbersversestheir
mnsilemodulus.

29

Table 3.1 - Tensile properties of high performance polymer ﬁbers.

 —w W

No. of Tests

 

Diameter (pm)

12.2 :1: 0.9

12.1 :l: 0.5

13.0 :1: 0.5

 

Tensile Modulus (GPa)

94:1:9

108:]:7

14816

 

Tensile Strength (GPa)

3.09 :l: 0.46

3.49 :l: 0.24

2.38 :l: 0.25

 

No. of Tests

 

3.67 :l: 0.40

14

 

3.15 :l: 0.24

22

 

1.72 i 0.23

12

 

12.0 :I: 0.9

18.8 d: 2.1

28.7 :I; 2.7

 

Tensile Modulus (GPa)

84:4

181 j: 17

68:9

 

Tensile Strength (GPa)

3.72 i 0.27

3.42 :l: 0.55

2.65 :1: 0.29

 

Fracture Strain (%)

 

5.08 :t 0.32

 

2.05 :t 0.45

 

6.91 :l: 1.30

 

30
reduced. This trend is evident in the Kevlar ﬁbers that have the same polymer

chemistry. Kevlar 29 and 49 reportedly have a pleated sheet morphology (Dobb et al.
1977). Conversely, Kevlar 149, which exhibits 80-90% of its theoretically predicted
tensile modulus, does not show the pleated morphology which suggests 'stlaightening'
of the pleated sheet Structure and increased orientation (Krause er al. 1989). The
implication of previous observations is that reducing relative mobility of adjacent polymer
chains (e. g. , by cross-linking), Should result in increased tensile modulus but reduced
fracture strain of the ﬁbers (also see section 7.3.2).

Table 3.2 compares the interfacial shear strength (ISS) of the examined high
performance polymer ﬁbers. The “as received“ Kevlar-49 and Technora ﬁbers possess
proprietary sizings that were removed by an ethanol washing process. Both “washed"
and ”as received" ﬁberswereexaminedtoevaluatetheeffectsofﬁbersizingontheﬁber
interfacial properties. For other unsized ﬁbers, “washed" ﬁbers were also examined to
ascertaindrewaslungprocessdoesnotaltertheﬁberadhesivepmperﬁes. Formeliquid
crystalline polymer ﬁbers, three curing conditions were examined to assess the effect of
thermalstresscsontheﬁberinterfacial shearstrength.

ToexaminesomeoftrendsintheISS data, athreedimensional stress model (Whitney
etal. l980)hasbeenexamined. Previously, itwasdemonstratedthatthemodelcould
notreasonablypredicttheactualISSvaluesofaramidﬁbersbecauseofthefailureofits
linear elastic ﬁber fragmentation assumption (Kalantar er al. 1990 " ). However, the
model provides valuable insights to the expected trends for various ﬁbers. A
combinationofexperimentaland theoreticaldataareused topredicttheISS trends. The

31

Table 3.2 - Interfacial shear strength (188) of high performance ﬁbers.

’ Kevlar-149

ISS Fragmentation test (MPa)

 

Ambient Cure

75/125°C cure

16.711302)
16..3;t14(12)

175°C cure

17.3 :1: 1.4 (10)

 

Kevlar-49
As Rec.

17.5 :t 1.3 (27)

17.9 :1; 1.2 (12)
18.2 :I: 1.3 (12)

18.0 :1: 1.0 (25)
17.6 :1: 1.3 (17)

 

Kevlar-29 Washed

As Rec.

19.5 :1: 2.7 (20)
19.9 :1: 1.8 (10)

20.9 :1: 3.2 (7)
20.6 :1: 2.8 (8)

 

Technora Washed

As Rec.

naizcan
33.2 i 2.2 (10)

28.7 :1: 1.9 (7)
36.6 :I: 2.7 (5)

 

PBO Washed

11.5 :1: 1.5 (27)

16.1 :1: 1.4 (14)

17.8 :1: 1.4 (14)

 

E-Glass As Rec.

28.9 :1: 11.4 (219

 

As Rec.

 

29.6 :1: 9.3 (614)

 

41 j: 12 (572)

 

 

ISS Droplet test 75/100°c cure (MPa)

 

As Rec.

Washed: washed with Ethanol

AsRec.: testedasreceived

2.95 :1: 0.17 (18)

 

(the numbers in parentheses represent the number of samples tested)

Table3.3-Materialpropertydata.

Material ensile Modul Tensrle Fracture
(pm) (GPa) (GPa)gth Strain (%)

17..1:|:12

l.08:t0.22

 

AS4 E.(J"(Carborl)

1 8.1 :l: 0.3

5.862

 

Epoxy Ambient Cure
' 75l125°C cure‘
175°C cure

‘Raoetal. 1991'
2nominalliterataure value

 

 

0.0947 :1: 0.0012

 

 

32
theoretical critical length (1“,) is given by:

 

l a - 49 G (3.1)
. . v w ..

where Evis the axial ﬁber elastic modulus, ﬂu is axial ﬁber Poisson’s ratio, and G, is
the matrix elastic modulus. To predict a theoretical ISS value, I,” is inserted into
equation 2.1 using the experimental tensile strength data of Table 3.1.

Figure 3.3 compares the experimental and theoretical ISS values of the "as received'
ﬁbers for the 175°C/3hr curing condition. Material property data for E-glass ﬁbers, AS-
4carbon ﬁbers, andvariousepoxy systemsarelistedinTable3.3. Flgure3.3shows
thatinterfacial shearsu'engthaSS)oftheinorganicﬁbersagreewiththetheore6cal
values, whereas the organic ﬁbers exhibit ISS strengths about half their predicted values.
Thedismepmmybetwealdleoredcalandexpaimamlresulucanbeamibutedmdw

 

1 0.0

L m Experimental _
_ :1 Theoretical

J: or an
o o o
I T

l

J

N
O

Interfacial Shear Strength (MPa)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i i

A54 E-Glass PBO Technora Kevlar Kevlar Kevlar
29 49 149

Bguu3J-Expaimenmlmdthwmdcalmtu'facialshearmulgthofvaﬁmmﬁbers.

 

33
model’s assumptions that the ﬁbers are transversely isotropic and undergo linear elastic

deformations. Figure 3.4 shows the optical micrographs of Kevlar-49 and AS-4 carbon
ﬁber fragments under bright-ﬁeld and cross-polarized lighting. For the carbon ﬁber, the
ﬁber fragments into whole segments, whereas the Kevlar-49 ﬁbers fragment by axial
cracks that result in ﬁbrillated segments. Therefore, for the high performance polymer
ﬁbers, the predicted results of this model could only provide a qualitative measure of
various trends. For example, the model suggests that Technora ﬁbers, with their lower
tensile modulus than Kevlar-49 ﬁbers, are expected to exhibit higher ISS values than

Kevlar-49 for the same epoxy nratrix.

 

 

 

ﬁgure 3.4- Optical micrographs of ﬁber fragments. (A) Kevlar-49 bright-ﬁeld,
(B) Kevlar-49 cross-polarized, (C) AS-4 bright-ﬁeld, (D) AS-4 cross-polarized.
8 100 um

34
ﬁgun3.5exanﬁnestheeffectofcuﬁngdwrmalsuessesmmemmrfacialslear

strength properties of “washed" PBO, Kevlar-49 and Technora ﬁbers. Of the three
curing conditions, the 175°C curing exerts the highest thermal stresses and the Ambient
curing the lowest thermal stresses (Kalantar er al. 1990”). For the Technora ﬁber the
ambient cure data was not obtainable because this epoxy system is more brittle when
cured at ambient conditions compared to the other curing conditions. The ambient cured
Technora samplebrokebeforetheﬁbercritical lengthcouldbemeasured. Figure3.5
showsthatinterfacialshearstrengthoftheKevlar-49 sarnpleswerenotaffectedbythe
curing conditions, whereas Technora and PBO samples Show increased ISS values for the
higherthermalstresses. Theseresults suggestthatthefailuremodeoftheKevlar ﬁbers
areindependentofthethermalstressesoftheirsurroundingmatrix. Itwillbe
demonsnatedlaterthattheadhesionoftheTechnomandPBOﬁbershavesome

U
01

 

U
0

P' ..... :‘t,2;t;: o . Kee.v.l .o.r.;.4.9. ................ . ..................... . ..........

 

N
01
r

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

ooooooooooooooooooooooooooooooooooooooooooooo

N
O
' r

_.b
U)

._a
O

Interfacial Shear Strength (MPa)

01

 

   

   

 

 

 

 

 

 

 

O

 

 

Ambient 75/1 25°C
Cure Cure
Figun3.5-InmrfacialshearsnengthofwashedPBO,Kevhr49andTechnomﬁbas
for three different curing conditions.

35
addiﬁonallimitaﬁonsﬂratmaybeaffectedbythestressenvimnmentofﬂle matrix.

TechnoraandKevlar-49ﬁbersweretheonlyﬁbersthatweresizedbytheir
manufactures to improve their handling and/or adhesive properties. Figure 3.6 shows
the interfacial shear strength (ISS) changes between the "as received“ (sized) and
”washed" (unsized) ﬁbers for two curing conditions. The Kevlar-49 ﬁber shows similar
ISS values for the ”washed” and ”as received” ﬁbers, whereas, theTechnora ﬁbers show
over30% higherISS valuesforthe "as received" ﬁbersthan the ”washed” ﬁbers. These
observations suggests that ﬁber sizing is enhancing Technora-epoxy adhesion but is not
affecting Kevlar-epoxy adhesion.

ﬂgun3.7wmparesmesurfaceulergyofsevaalhighperformancepolymerﬁbas
with the liquid epoxy. The ”as received” Technora ﬁber shows surface energy

components that closely match the epoxy surface energy components. Kaelble (1971) has

 

[:1 Washed

L... ................................... . ...............................
m As Received

.5
C)

   
       
         
       
       
       
       
      
     

   

    
    
       
       
   
       
   
     
   
       
     

       
       
 

  

 

 

 

 

 

 

 

 

 

 

 

     

 

A
C,
O.
2 -
.c 3.0.!
*5. :35: 3:02
C130 - ................................................... éé‘ §§ .n.
D O 9 9 . 6
8 D:0:¢ 1:0:
4! Doe’s Doe’s
U) :eze: >3:
5 rte: 1:0:
0 20 .. ............. . ....................... . ........ . . :0: 1 . :0}:
.C I . I V0.1 ’0’4
m {0% ’0‘ 1.9.4
D 0.4 ’9” 30.4
"' 3.0.1 1.0% 1.0%
.9 ’0’. v.9” t’e’q
0 :0: so.” 3:.”
O D e 1 . 99.4 v e’«
1 O '- ’00.! """" 3.961 ' .000.
t 1.9.4 5.0% 9.9.4
0 t’e’s s’e’r v’e’a
e o e o o e
H D e c b o d v e t
C We” Doe’s D.O.<
_ e O o e o e e
v.0.s . o v.0.1 1.0;
1.9; b as v.0.4 0.0.4
0 ‘7 7. D 0 G
75/125°c 175°C 75/125°c 175-c
Cure Cure Cure Cure
Kevlar-49 Technora

Figure 3.6 - Interfacial shear strength (ISS) of “as received“ (sized) and ”washed“
(unsized) Kevlar-49 and Technora ﬁbers for two curing conditions.

36
suggested that a close match between polar and dispersive components of surface free

energy of the adherents can result in their optimum wetting properties. The “washed“
Technora ﬁber shows over twice the polar component of surface energy of the epoxy
which my result in less than optimum wetting properties. The surface energy similarities
of the ﬁber sizing and epoxy may also result in sizing-epoxy mutual diffusion and
increased ﬁber-matrix coupling. The sizing may also increase the modulus of the
surrounding matrix enhancing the load carrying capacity of the interphase. Conversely,
“washed“ and ”as received” Kevlar ﬁbers have similar surface energy components. Note
that the unsized Kevlar-29 ﬁbers show similar surface energy components for their
'washed" and ”as received” ﬁbers which suggest that the washing process does not affect
the ﬁber surface wetting properties.

Work by Gutowski (1990) has shown that in the absence of chemical bonding,

 

 

 

 

 

 

 

 

 

 

 

 

30
am As Received
‘- l:l Washed
2 20 ,.. ...........................................
0 _
0.
A10... .......... ......., ........ ..
E
{
o o , ..
c t’t v ’e‘
‘6. 1’. '9‘
V i:‘ ’ ‘ :‘1
‘ o .. .......................... . ’9‘ '0‘ .
0).) :9: :0: {t
a :3: :1: if
L- . ’0‘ '0‘ 3.4
a, 20 ,— .............. t . ........ . . . '0‘ ’2‘ s.
0. '3 :0: :4
g {:1 .z. :2;
30 t— ............... H ........... 3: :2: :z:
D C D 4
40
Epoxy Kevlar Kevlar Technora PBO Spectra
29 49 l 000

Figure 3.7 - Polar and dispersive components of surface energy for several high
performance polymer ﬁbers and liquid epoxy. Only Kevlar-49 and Technora ﬁbers were
sized.

37

maximum adhesion between the matrix and ﬁber occurs when the surface energy of the
ﬁber (7,) and matrix (7.) are equal and when (717.) < 1 only incomplete wetting could
occur. PBO and Spectra-1000 ﬁbers show lower total surface free energy than the liquid
epoxy which suggests incomplete wetting with the epoxy. Incorporation of polar
functional groups could potentially enhance their wetting compatibility with the liquid
epoxy (discussed in Chapters 5 and 6).

The weak lateral interactions of the high performance polymer ﬁbers is demonstrated
by their skin separation. Figure 3.8 shows SEM micrographs of a Kevlar-49 aramid
ﬁberthatshowsﬁberskinseparationintheformofaheliealribbon. Figure3.9shows
a similar skin separation for an untreated PBO ﬁber. Technora ﬁbers also exhibit sldn
separation (Takata 1987), although they are expected to have less of the skin-core
morphological differences than other examined liquid crystalline polymer ﬁbers.

TEM micrographs of the liquid crystalline polymer ﬁbers also demonstrate their weak
lateral cohesive properties. Figure 3.10 shows a typieal section of an ”as received"
Kevlar-49 ﬁber. Interfacial separation is parallel to the cutting direction and there are
cohesive ﬁber ﬁbrillation near the ﬁber-matrix parting areas. Similar ﬁber surface
ﬁbrillation are also observed for other aramid ﬁbers. Figure 3.11 shows TEM
micrographs of "as received” and ”washed“ Technora ﬁbers. The ”as received" ﬁber
(Figure 3.11A) shows more ﬁber surface ﬁbrillation than the ”washed” ﬁber (Figure
3.113), suggesting stronger adhesive interactions of the latter. Figure 3.6 also showed
that the "as received” Technora ﬁbers exhibit over 30 % higher interfacial shear strength

values than the "washed' ﬁbers.

38
Figure 3.12 illustrates TEM micrographs of an ”as received“ PBO ﬁber that shows

a thin layer of the ﬁber skin is adhering to the epoxy side of the interphase separation.
This observation suggests that PBO ﬁbers have a thin surface layer that fails during the
application of shear stress. There is additional evidence for the presence of the PBO skin
layer. Figure 3.13 shows an optical micrograph of ”as received” PBO ﬁbers that exhibit
thepresence ofkinkbandsin some ﬁbers. Thekinkbandsareprobablytheresultof
compressive stresses induced during the ﬁber manufacturing process. Figure 3.14 shows
SEMmicrographsofthesamePBOﬁbers,whichdonotshowanykinksontheﬁber
surfacessuggestingthattheldnkbandsareinternal. IntheChapteré, itisshownthat
etching removal of the PBO surface layer exposes its sub-surface kinked line structure.
These observations conﬁrm the presence of a thin surface layer on the PBO ﬁbers.

Spectra-1000 polyethylene ﬁbers also exhibit a combination of interfacial and cohesive
ﬁber-matrix separation similar to the liquid crystalline polymers. Figure 3.15 shows
TEM micrographs of a radially sectioned Spectra-1000 ﬁber. The rnicrographs show
extensive interfacial adhesive failure which suggest weak ﬁber-matrix bonding. A higher
magniﬁcation view (Figure 3.153) shows ﬁber ﬁbrillation near some interfacial parting
areas. Therefore, for the untreated Spectra ﬁbers interfacial shear failure is dominated
by interfacial adhesive failure along with some ﬁber cohesive failure.

 

, “—
Figure 3.8 - SEM micrographs of untreated Kevlar-49 showing torn ﬁber sla’n in a form
ofahelicalribbon. (A)bar =10um (B)bar=5um

 

 

“Em-e 3.9 - SEM rnicrographs of a PBO ﬁber skin separation.
(A)bar=5um (B)bar= lum

 

‘
E

a

“sure 3.10 - TEM micrographs of a radially sectioned untreated Kevlar-49 ﬁber. Fiber-
malrix interphase shows both interfacial separation and ﬁber cohesive ﬁbrillation.
(A)bar=1pm (B)bar=250nm

    

  

. ,..
.. (4,7 . we. . '
;‘ 'Crh>>"r'~'_§ _ u' 7

Baun 3.11 - TEM micrographs of (A) "as received' and (B) 'washcd' Technora ﬁbers.
The 'as received' ﬁber is more ﬁbrillated than the “washed" ﬁber. bar = 1 pm

43

 

Figure3.12-TEMmicrographs of an 'as received' PBO ﬁber, showingathinlayer of
theﬁbersldnisadheringtotheepoxysideoftheinterphaseseparations.
(A)bar=2um (B)bar=100nm

 

 

 

 

Figure 3.13 - Optieal micrograph of 'as received” PBO ﬁbers, showing presence of
compressive kink bands.

45

81.060 IOFn

 

Figure 3.14 - SEM micrograph of 'as received' PBO ﬁbers. Kink bands are not
apparent on the ﬁber surface.

 

1"} , " u i ..
Figure 3.15 - TEM micrographs of an 'as received“ Spectra-1000 polyethylene ﬁber,
showing extensive interfacial adhesive failures and a few ﬁber cohesive failures.
(A)bar=5nm (B)bar=lOOnrn

47

3-4 QQHCLHSIQHS

For the examined curing conditions, the higher thermal stresses at higher curing
temperatures increased interfacial shear strength (ISS) of the Technora and PBO
ﬁbers, whereas, ISS values of the Kevlar-49 ﬁbers were unaffected by the increased
thermal stresses.

PBO and Spectra-1000 ﬁbers exhibit low polar components of ﬁber surface energy,
whereas Technora ﬁbers exhibit excessive polar component. Both eases result in
less than optimum wetting compatibility with liquid epoxies.

PBO ﬁbers exhibit a cohesively weak skin layer that fails within the ﬁber during
PBO-epoxy interfacial separation.

Comparison of experimental and predicted interfacial shear strength (ISS) of the
polymer ﬁbers suggests that increasing lateral interactions of the aligned polymer
chains could signiﬁcantly increase their interfacial load carrying capacity.
Enhancementofhteralchaininteracﬁmsisexpectedbreducemeirrelaﬁve

mobility, thus reducing the ﬁber fracture strain but increasing its tensile modulus.

Results of this chapter suggest that wetting properties and/or skin-core morphology of

the high performance polymer ﬁbers can signiﬁcantly affect their adhesive behavior. The

examined high performance polymer ﬁbers exhibit internal ﬁber ﬁbrillation which

suggest weak lateral interactions between adjacent polymer chains. This observation

suggests that the cohesive ﬁbrillation of the ﬁbers is ultimately responsible for their shear

failure mechanism.

CHAPIFR 4

 

Fiber Coating and Coupling Agent
Treatments of

High Performance Polymer Fibers

 

In this chapter, effects of several types of ﬁber coatings and caupling agents an
adhesive properties of Kevlar-49 aramid ﬁbers are examined. These treatments only
affect the ﬁber-matrix interphase properties, therefore, allowing the effects of the
interphase on ﬁber-matrix adhesion to be examined without perturbation form changes
inﬁber-matrixproperties. Thisapproachshouldprovidevaluableinsightstothe
signiﬁcance of ﬁber-matrix interphase on the adhesive properties of high performance

polymer ﬁbers.

48

49
4.1 W

Coupling agents are materials that are able to strongly interact by physical or
chemieal means between two substrates. Coupling agents can be either directly applied
to the ﬁber or be dissolved in the liquid resin and diffuse to the ﬁber-matrix interface.
Coupling agents are typically applied in minute quantities since their excessive presence
could introduce a weak boundary layer at the ﬁber-matrix interface.

liquid coupling agents based on organometallic complexes have recently claimed
improved bonding between polymer ﬁbers and polymer resins. Gabayson at al. (1988)
have reported increased interfacial bonding and enhanced processing for a variety of
polymer systems with the application of organometallic coupling agents. Sugerman er
al. (1989) have reported up to 20% increase in ﬂexural and compressive strength, and
80% increase in impact strengths of Kevlar-49/Novalak-MNA short ﬁber composites with
the applieation of various titanium and zirconium based coupling agents. Figure 4.1
illustrates the principle of the organometallic coupling agents. The titanium or zirconium
derivedwupﬁngagaitreactswimmefmepmmnmthesubsuamintafaceresmﬁngin
formation of an organic monomolecular layer on the surface of the substrate. The new
organichyercanthaimteractwimdwpolymermaaixandmhmwetheﬁber-mauix
interactions. These coupling agents, however, should be applied only in low
concentration (parts per thousand) to avoid producing weak boundary layers on the
substratesurfaces. Inthisstudy, twotypesoftitaniumorzirconiumbasedcoupling
agents have been examined.

Fiber coatings are applied to ﬁber-matrix interface in much larger quantities than the

50
coupling agents and form macroscopic layers at the ﬁber matrix interface with

composition independent of the substrate; therefore, the mechanical properties of the
coating material is expected to signiﬁcantly affect the ﬁber-matrix interactions. Fiber
coating can enhance ﬁber-matrix interactions by enhancing interfacial adhesion
mechanisms. Forexarnple, inchapter3itwas shownthattheinter'facial shearstrength
reductionsofthesizedTechnoraﬁberismorethanBO% higherthantheunsizedﬁber.
This ISS increase was attributed the improved epoxy compatible surface energy
components of the sized ﬁber. Therefore, applieation of the ﬁber coatings to the
Technora ﬁber can enhance their thermodynamic wetting properties.

Vapor deposited ﬁber coatings are particularly good candidates for enhancing the
adhesive properties of high performance polymer ﬁbers. The mobile vapor molecules
canproducegoodﬁbersurfacewetﬁngmdmaywenpareuatemeﬁbermteﬁormform
mechanical anchors within the ﬁber structure. Reagents such as butadiyne, ethylene, and

OH 0
OH mogéE/e :1} c
on c

M-TlorZr

 

 

 

 

 

Figure4.1 -Generalreaction schemeofatitaniumorzirconiumbasedcouplingagent
with the hydroxyl groups of a polymer substrate.

51
p-xylene can form strong and reactive ﬁlms of coatings on the ﬁber exterior that can then

be bonded with the matrix. In this study, butadiyne and p-xylene gas deposited ﬁber
coatings have been investigated.

Butadiyne (C411,) is a diacetylene monomer that can undergo a thermal polymerization
to form a polymer ﬁlm from its vapor phase (Snow 1985). Butadiyne polymerization
producesahighlyreactivecoatingonthesurfaceoftheﬁberthatcantheninteractwith
the matrix. The reaction scheme of butadiyne polymerization is shown in Figure 4.2.
The deposited polymer structure has been characterized as a complex combination of
polyene and polyacene structures (Snow 1985). Armistead at al. (1987) have examined
interfacial shear strength (ISS) of butadiyne coated AS-4 and HMS-4 earbon ﬁbers using
theISSfragmentationtest. Theyreportedupto50% ISSincreases fortheAS-4 ﬁbers
but no measurable ISS enhancement for the HMS-4 ﬁbers was detected.

Parylenesaretheothergasphaseﬁbercoatingsexaminedinthisstudy. Paryleneis

 

31—0-3 O- OED—:0
:D E
I E
ill—E
4.
3
I
3

(A) (B)

 

 

 

Figure 4.2 - Polymerization of diacetylene monomers produce a combination of
substituted (A) polyene and (B) polyacene structures.

 

 

52
the generic name for members of a p-xylylene thermoplastic polymer series developed

by Union Carbide. There are three type of Parylene available commercially which are
shown in Figure 4.3. Parylene-N is a primary dielectric and is used in electronics
devices. Parylene-C has several useful electrical properties along with very low
permeability to moisture and is used for the coating of electronic circuitry. Parylene-D
has similarelectriealpropertiesasotherparylenes buthasthebestchemicalresistance
of all other parylene grades. The parylene polymers are deposited from the vapor phase
at pressures around 0.1 torr and ambient temperatures. The initial vapor monomer is
highly reactive which results in its simultaneous adsorption and polymerization on the
substrate (Beach 1987). Parylene polymerization produces poly-para-xylylene chains that
are greater than 5000 units long. Applieation of parylene for adhesion promotion has not
beenreported intheliterature. Inthis study, paryleneswereconsideredascoatingsfor
aramid ﬁbers bceause of their chemieal compatibility with the ﬁbers.

 

 

/ ‘
orig—.042} Parylene-N
\
Cl
{CH2 ‘01,} Parylene - C

we}

Figure 4.3 - Various types of parylene polymer available.

 

 

 

 

 

53
4.2 W

Four types of ﬁbers were examined, Kevlar-49 aramid, PBO aromatic heterocyclic
AS-4 carbon, and E-glass ﬁbers. Three curing conditions, ambient, 75°C/2hr-
125°C/2hr, and 175°C/3hr were selected for this study. Interfacial shear strengths were
characterized by the fragmentation tests. Single ﬁber tensile strengths were measured
usingamodiﬁedASTMD3379testwith25mmtestgagelengthandanominal
elongation rate of 0.135 mm/min. Thermal expansion of the parylene coated ﬁbers were
evaluated by a du Pont thermal mechanieal analyzer (TMA model 943) using a 10 mm
ﬁber gauge length. Material and experimental conditions are detailed Chapter 2.

The liquid coupling agents were designated as KRSS (titanium IV tetrakis bis
2-propenolato methyl -l-butanolato adduct 2 moles di—tridecyl hydrogen) and L237
(zirconium IV 2,2-bis propenolato methyl butanolato, tris 4-arnino—besoato—O) (Kenrich
Petrochemieals, Inc., Bayonne, NJ). The recommended amount of coupling agent was
0.2-0.3 weight percent. Both the recommended amounts and a ~ 1.5% concentration of
each coupling agents were examined in this study.

For the butadiyne treatments, three tows of treated and one untreated Kevlar-49 ﬁbers
were supplied by Dr. A.W. Snow at the Naval Research lab. The butadiyne treatments
oftheKevlar-49 ﬁberswereconductedas follow: Atowofﬁberwaswrappedaround
arectangularwinderandinsertedintoasoxhletextractor. Fiberswerewashedwith
chloroform for 3 hour, followed by 1 hour of vacuum drying at room temperature. The
ﬁberswaethurﬁansfenedmambuhrreactmwhichwasheatedm150°c,evacuated
and backﬁeld to 650 torr with butadiyne. The butadiyne deposition was measured by

 

 

 

 

54
percent weight gain. Three deposition quantities 0.50 wt% (2 hour reaction time), 0.84

wt% (4.25 hr), and 1.39 wt% (8 hr) were conducted, which displayed a progressive
development of brown coloration. The treated ﬁber tows showed some ﬁber clustering
especially for the 1.39% treated and to a lesser extend for the 0.84% treated ﬁbers. The
agglomeration made separation of the individual ﬁbers difﬁcult. The interior ﬁbers
sometimes had a lighter color than the exterior ﬁbers, which indicated the non-uniformity
of the coatings.

Parylene coatings of Kevlar-49, PBO, A84 and E-Glass ﬁbers were conducted at
Novatran Corporation (Clear Lake, WI). The parylene deposition process has been
described by the Novatran Publications. Tire process consists of three distinct steps.
The ﬁrst step is vaporization ofthe solid dimer (~250°C and ~1torr). The second
step is the division (pyrolysis) of the dimer (~680°C and ~0.5 torr) to produce the
monomeric diradical, p-xylylene. Finally, the monomer enters the deposition chamber
(~25°c and ~0.l torr) where it simultaneously adsorbs and polymerizes on the
substrates. Both thick and thin coatings of different parylenes have been investigate.
Table 4.1 lists the experimental protocol for parylene treatments of various ﬁbers. The
propﬁetaryadhedmpmmoterA-lﬂwasalmexmﬁnedmmhancemewupﬁngbetween
theparyleneanditssubstrates. ThetreatmentGwasimmersedinliquidepoxy
immediatdyaﬁerﬂleﬁbercoaﬁngstoevaluateeffectofagingonﬂlecoaﬁng. On
arrival, samplerasrinsedwithacetonetoremoveexcessepoxyandafterabrief
vacuumdryingtheﬁberswereeastandcured.

55

Table 4.1 - Parylene experimental protocol and curing conditions examined for each
ﬁber treatment.

3 pm Parylene-C f 175 175 175
Ambient Ambient Ambient

 

7 pm Parylene-C 175 175 175
Ambient Ambient Ambient

 

10 nm Parylene-N ' 175
75/ 125

 

100 nm Parylene-N 175
75/ 125

 

A-l74 on ﬁber + 175
10 nm Parylene-N 75/ 125

 

A-l74 on ﬁber + 175
10 nm Parylene-N + 75/ 125
A-l74 on parylene

 

Sameastutstored
andshippedinepoxy

 

 

 

 

 

56
4.3 W

Effects of various ﬁber coating and coupling agent treatments an adhesive properties

of high performance polymer ﬁbers are presented separately.

4.3.1 W

Both low and high concentration of KRSS and L237 liquid coupling agents were
examined. Figure 4.4 shows the interfacial shear strength (ISS) of the treated and
untreated Kevlar-49 aramid and A84 carbon ﬁbers for the 175°C curing condition. For

the Kevlar-49 ﬁbers their ISS values are unaffected by the coupling agents but for the

m Kevlar—49
[:1 AS—4

_T __

 

o

o
l
I

I

l

 

 

 

 

 

 

m
0
l

 

 

 

 

.p.
O
I

N
O
I

 

Interfacial Shear Strength (MPa)
03
O

 

 

 

 

 

 

 

 

 

 

 

 

. E Q E g
Untreated 0.347. 1 .1 472 0.3272 1 .1 673
KRSS KR55 L237 L237

Figun4.4-hterfacialsheusﬁalgdlofKevhr49ammidandAS4carbonﬁbaswim
liquid coupling agent mixed 175°C cured epoxy.

 

 

 

 

 

 

57
A84 ﬁbers there are about 12% reduction in their ISS values for the L137 coupling

agents. The reductions for the AS-4 carbon ﬁber occurs despite the abundant presence
of hydroxyl groups on these ﬁbers (Hook er al. 1990).

The enhanced mechanieal properties of polymer composites reported by Gabayson at
al. (1988) and Sugerman at al. (1989) may be due to improved processing introduced by
the organometallic coupling agents. The single ﬁber test represent an ideal case of ﬁber-
matrix wetting, whereas, in the multiﬁlament composites, tow impregnation of the ﬁber
bundles are important consideration. A liquid coupling agent could improve the ﬁber
tow impregnation by enhancing ﬁber wetting or the resin ﬂow properties, thus helping
to produce defect-free composites. Improvements of interfacially controlled composite
properties by producing defect-free parts, however, is only a processing enhancement
which is different from an intrinsic improvement of the ﬁber-matrix interfacial shear
strength.

433mm
Figure 4.5 shows the interfacial shear strength (ISS) of the butadiyne treated

Kevlar-49 aramid ﬁbers for the 175°C curing condition. No signiﬁeant changes in the
ISS results is observed. Fiber tensile strengths were also unaffected by the treatments.
Figure 4.6 illustrates a TEM micrograph of a 0.84% butadiyne treated ﬁber that exhibits
increased interfacial ﬁbrillation. Therefore, the butadiyne coating has increased the
aramid-matrix adhesion as evidenced by extensive interfacial ﬁbrillation, however, the

hnpmvedadhesimhasnotovercometheﬁbersurfacesuucmmﬁnutaﬁonsandthe

5 8
Kevlar-epoxy interfacial shear strength is unaffected. Nonetheless, the butadiyne treated

aramid surface is at least as strong as the untreated surface, suggesting that butadiyne is
not producing a new weak surface layer. The vapor deposition and fast reaction of
butadiyne should enhance wetting properties of non-wetting surfaces. Snow (1981 ' ,
1981 ‘ , 1984) has demonstrated that butadiyne may be deposited onto polyethylene or
Teﬂon substrates. Therefore, butadiyne treatment may be a useful coating for surfaces

that exhibit poor wetting compatibilities with epoxy resins.

 

N
O

 

 

 

 

 

 

 

 

—L
O)
sitar

 

—I
m N
7' Its

Interfacial Shear Strength (MPa)
.rs

 

 

 

 

 

 

 

 

 

 

 

O

Untreated 0.50 0.84 1 .39
Butadiyne wt7.

Figure 4.5 - Interfacial shear strength of butadiyne treated Kevlar-49 aramid ﬁbers with
175°C cured epoxy.

59

 

Figure 4.6 - TEM micrograph of a 0.84 wt% butadiyne treated Kevlar-49 ﬁber.

4.3.3 Balsam
Effects of thick parylene coatings (treatments A and B) on tensile properties of

Kevlar-49 and E-Glass ﬁbers were examined. Because thickness of these coatings are

comparable to the ﬁber diameters, the contribution of the coatings was evaluated using:

P

a . T
AI+—£A¢ (‘01)
31

where a... is the ﬁber tensile strength corrected for the coating contribution, P is the load
applied to the sample, A, is the ﬁber cross-sectional area, A, is the coating cross-
sectional area, E, is the coating tensile modulus (2.76 GPa for Parylene-C), and E, is the
ﬁber tensile modulus (Tables 3.1 and 3.2). Equation 4.1 is derived in Appendix A.
Table4.21iststhetensilestrength and diameterof3and7mearylene—C treatﬁbers.
ThecoatedKevlar-49 thecoatedﬁbersdonotshow signiﬁeantchangesintheirtensile
strength, however, fortheE-Glass ﬁberstherearemeasurableincreasesinthewnsile
strengthofthecoatedﬁbers. TensilestrengthincreasesoftheE-Glassﬁberscanbe

Table 4.2 - Tensile strength and diameter of 3 and 7 pm Parylene—C treated Kevlar-49
and E-Glass ﬁbers.

“_ Tm mg“ (“I")
Kevlar 49 12.7 i 0.5 2.81 :1; 0.46
3 pm Parylene-C 19.2 :1; 0.7 1.23 i 0.16

7 pm Parylene-C 27.9 :1; 1.0 0.609 1; 0.088

 

3 pm Parylene-C 0,. 2.73 i 0.36
7 um Parylene-C a... 2.70 :1; 0.39
E-Glass 17.1 :1: 1.2 1.08 :1; 0.22
7 um Parylene-C 33.7 :1; 2.2 0.382 :1: 0.084
7 pm Parylene—C a... 1.33 i 0.35

 

 

 

61
attributed to the deposition process and hydrophobic nature of the parylene coating.

Michalske er al. (1987) has shown that water ean weaken a glass by chemieally attacking
theglassmoleculesatacracktip. Asurfacecoatingthateanblocktheopeningofthe
cracks and restrict the passage of water molecules, should increase the strength of the
glass. Parylene vacuum deposition allows for water removal and its small monomers
permitcrackpenetration. Parylenehasalsoalowmoistureandgaspermeability that
should block moisture from reaching the crack tips. The ﬁber tensile tests have been
performed at 25 mm gage lengths, greater tensile strength increases are expected for the
shorter gage length samples because of statistical reduction of the number of large defects
in the shorter gage length samples.

To evaluate effects oftherrnal stresses on the ﬁber-matrix interfacial shear strength,
both the ambient and 175°C cured epoxies were examined. Figure 4.7 shows the
interfacialshearstrengthsoftheKevlar49,PBO,AS-4andE-Glassﬁbersforthe
treatmentsAandB. Foralltheﬁberstheparylenecoatingsreducetheinterfacialshear
strength. The ISS reductions can be attributed to the effects of the parylene’s low
modulus. Equation 2.1 does not explicitly include effects of thermal stresses and the
matrix modulus, but more complex models such as those proposed by Whitney at al.
(1980) (see equation 3.1) or Cox et al. (1952) demonstrate that lowering the modulus of
ﬁber—matrix interphase modulus would reduce efﬁciency of interfacial load transfer. The
ISS reductions of the parylene coated ﬁbers are much more pronounced for the inorganic
B-glass and A84 carbon ﬁbers than the organic Kevlar-49 and PBO polymer ﬁbers.

Theseobservaﬁomsuggestﬂutthemwrfacialpmperﬁesofdworganicﬁbemamnm

62
much superior to the mechanieal properties of the parylene layers.

Figure 4.7 also shows that for the parylene coated ﬁbers there are ISS increases at
the higher temperature curing condition. Untreated Kevlar-49 ﬁbers do not exhibit
temperamretrends (seeFigure3.4),buttheparylenecoatedﬁbers showISSincreases
for the higher curing temperature. This temperature trend of the coated Kevlar-49 ﬁbers
may be attributed to the thermal expansion effects of the thick parylene layer. Figure
4.8 illustrates the longitudinal thermal expansion of untreated and Parylene-C coated
Kevlar-49 ﬁbers. The effects of the coating is to introduce additional longitudinal
compressive stress on the embedded ﬁber during the cool down from oven temperatures
to ambient conditions. This longitudinal compression in turn results in increased
interfacial radial compressive stress due to Poisson’s ratio effects (Kalantar er al. 1990
‘). Theshearloadeanyingcapacitymparylenemterphasemaybehwreasedbydle

 

‘ [:1 Untreated
- 3pm Parylene—C
_ m 7pm Parylene-C

P
b
b

-I
O
O

m
0

.p

O
I
}-—-i

N
O
I

T

Interfacial Shear Strength (MPa)
0)
o

 

 

 

 

 

 

 

 

0 Hrs Hﬁg ﬁﬁa HE 3

Ambient 175‘C Ambient 1 75‘C Ambient 175’C 1 75‘C
Kevlar-49 PBO AS-4 E-Glass
Figure4.7-Interfacialshearstrengthoftreatments3and7umParylene-Ccoatingsfor
Kevlar-49, PBO, A84 and E-Glass ﬁbers at two curing conditions.

 

63
increased excess radial stress, resulting in the increased ISS values.

Figure 4.9 illushates the fracture process of 7 am ParyleneC coated Kevlar-49
aramid and AS-4 carbon ﬁbers. The ﬁbers have been embedded in the 175°C cured
epoxy matrices. On load application, for the Kevlar-49 sample, the parylene coating
fractures before the ﬁber (A), whereas, for the A84 ﬁbers, ﬁber fracture occurs before
the parylene fracture (D).- This observation suggest that the parylene coating is more
brittle than the Kevlar-49 ﬁber but more ductile than the AS-4 ﬁber. For both ﬁbers,
their fractures results in displacements and conieal fractures of the parylene coating (B),
(C). (E). and (F)-

 

0.20

0.15 -

Dimension Change (96)
i

 

 

 

 

 

llllTlll
20406080100120140160180200

Temperature ('0)

Figure 4.8 - Longitudinal thermal expansion of the untreated and Parylene—C heated
Kevlar-49 ﬁbers. The parylene coatings tend to longitudinally compress the ﬁber during
the cool down.

 

 

64
Figure 4.10 shows the TEM micrographs of a 3 pm parylene-C coated Kevlar-49

ﬁber sectioned radially. The micrographs show extensive ﬁbrillation of the coating both
on the ﬁber and epoxy sides. The parylene ﬁbrillation suggest that parylene itself has
an ordered morphology with even weaker lateral cohesive properties than those of the
Kevlar-49 ﬁbers. Therefore, cohesive failures of the parylene layer is limiting the load
hansfer between ﬁber and mahix.

Thin Parylene-N coatings of Kevlar-49 ﬁbers resulted in similar observations as the
thick coating. Treahnents C, D, E, F, and G resulted in an average ISS value of 9.7
MPa (compared to 18 MPa for unheated ﬁbers), which is lower than the average values
obtained for the thick parylene-C coatings (12.4 MPa). This observation correlates with
the lower modulus of parylene-N (2.41 GPa) than parylene-C (2.76 GPa).

Figure 4.11 shows TEM micrographs of a Kevlar-49 ﬁber with coating F. This
coating has the A—174 adhesion promoter and shows only limited interfacial separation.
The adhesion promoter has altered the epoxy morphology to form what appear to be two
concenhic bands around the ﬁber perimeter. Each band is about 50 to 80 nm thick. The
similarities in interfacial shear shength of thin Parylene-N coatings with and without the
adhesion promoter suggest that the parylene layer is itself the weak boundary layer.
Therefore, interfacial shear shength of Parylene-N heated aramid ﬁbers are still limited

by the ﬁbrillation of the parylene layer.

 

 

r-l

 

 

 

 

 

a“-.. .r 1 as m-.- . ,
’“ "w (""1“ 1'“: 2“ * 4 1 1 1
“LIMKWI 1 1 x . TI

F soum

 

Figun4.9-Fractuﬁngprocessof7mearylale-CcoatedKevlar49aramidmdAS-4
carbonﬁbers

 

"W

4.10 - TEM micrographs of radial sections of a 3 um Parylene-C coated Kevlar-
49 ﬁber in the 175°C cured epoxy system. (A) bar = 1 pm (B)bar = 1pm

 

Figure 4.11 - TEM micrographs of radial sections of a thin Parylene-N coated Kevlar-49
ﬁber (heahnentF). (A)bar-= lum (B)bar= 100nm

4.4 commas

Examined organometallic coupling agents marginally reduced the interfacial shear
shength (ISS) of AS4 carbon ﬁbers but did not affect Kevlar-49 ISS values.
Butadiyne heated Kevlar-49 ﬁbers exhibit improved ﬁber-mahix bonding, however,
the ISS of Kevlar-49 ﬁbers are unaffected suggesting that ﬁber cohesive failure is
still limiting their ﬁber-mahix adhesion.

Parylene coatings could enhance tensile properties of E-glass ﬁbers by providing a
moisture banier but did not affect tensile properties of the polymer ﬁbers.
Interfacial shear shength of the parylene coated ﬁbers were drastieally reduced due
to the low modulus and tensile shength of the parylene polymer. The parylene
coating also demonshatcd that the interfacial shear shength of the organic ﬁbers are
not much shonger than to the mechanieal properties of the parylene layers.
Tiwshearloadearryingeapacitytheparylaieinterphaseeanbehwreasedby

compressive shesses that are induced by the sample curing process.

The similarity of the ISS values of unheated and parylene coated polymer ﬁbers

suggest that the high performance polymer ﬁbers possess mechanically weak adhesive

properties. Since, surface chemishy modiﬁcation ean not enhance the bulk mechanical

properties of a polymer ﬁber, therefore, coupling agents and/or ﬁber coatings can only

be effective if the wetting or chemishy of the ﬁber-mahix interface is limiting their

adhesive interactions.

69
Parylene results also suggest that the thermal shesses of the curing process can affect

shear load carrying capacity of the ﬁber-mahix interphase. This observation may
explain the ISS temperature trends that were observe for the unheated Technora and PBO
ﬁbers but were absent for the Kevlar-49 ﬁbers (see Chapter 3). Interfacial bonding of
the Kevlar-49 ﬁbers are merely limited by the internal ﬁber ﬁbrillation, whereas,
Technora and PBO ﬁbers exhibit additional wetting and weak surface layer limitations.
The surface property limitations of the Technora and PBO ﬁbers may be affected by the

shess environment of the mahix resulting in their observed 188 temperature heads.

CHAPTER 5

 

Plasma and Corona Treatments of

High Performance Polymer Fibers

 

Inthischapter, the'effects ofcoronaandplasma heahnentsonmechaniealand
chemical properties of the PBO and polyethylene ﬁbers have been investigated. These
heahnents can alter chemishy and morphology of polymer surfaces without modifying
their bulk properties; allowing examination of surface limited adhesion mechanisms and

how these limitations inﬂuence adhesive properties of high performance polymer ﬁbers.

70

71
5.1 W

A plasma is an excited gas region where positive and negative space charges are
created. To generate a plasma the gas must be excited by some power input such as AC,
DC, elechomagnetic radiation, or nuclear reactions. There are three categories of
plasma: hot, mixed, and cold plasma. A hot plasma such as solar corona has high
equilibrium temperatures and must be maintained by nuclear reactions or laser
excitations. Acold plasmasuchasthatinaneonlamphasitsbulkgasinambient
temperatures but its free elechons may have high kinetic energies that result in their
highly reactive chemical environments. A mixed plasma is between the hot and cold
plasma. In this report plasma heahnents refer to cold plasma phenomena exclusively.

A plasma may contain atoms, molecules, ions, ﬁee radicals, free elechons, and
metastablespecies. Theexcitedspeciespresentinaplasmaeaninteractwiththetreating
subshatetoproduceavarietyofeffectssuchassurfacelayerremoval (etchingand
cleaning), chemical modiﬁeation, cross-linking, and polymerimtion (Kinloch 1987). In
polymer materials the low molecular weight polymers and contaminants tend to
accumulateonthesurfacetominimizethesurfacefreeenergy (seeAppendix B). These
lowmobcuhrwdghtspwiescreammterfadﬂweakbomdaryhyershmtmdehimennl
toadhesion. Theexcited species oftheplasmacanremovethese surfacespecies from
thesubshate. Thesurfacedegradationcausedbyplasmaheatmenteanreducethe
molecular weight of the surface polymer or contaminants which then vaporize into the
reactionchamber. Thesurfacedegradationcanalsoetchthesurfaceandincreaseits

contactsurfacearea. Plasmamaycontainchemicallyrcactiveexcitedspeciessuchas

 

 

72
free radieals and ions that chemieally interact with the subshate to produce reactive

surface sites and increase surface free energy. Without active oxygen and nihogen
species the surface groups may interact to cross-link the surface molecules (Cross-linking
by Activated Species of INert Gas, 'CASING"). If the plasma contains polymerinble
gas monomers, then the deposition of the polymer phase onto the subshate surface is
possible.

A corona discharge is the flow of elechicity from a high voltage conductor through
ionized air between the conductor and an insulator, and is usually exhibited as a faint
glow adjacent to the surface of the conductor. Normally air is a nonconductor, however,
it contains small number of ions produced, for example, by the cosmic rays. A charged
conductor draws the oppositely charged ions to neuhalize itself. When the drawn ions
receive a high acceleration, their collision with other air molecules produces more ions
and the surrounding air becomes conductive. Corona discharge is the result of an
elechical breakdown in the surrounding gas. Corona discharge ean inhoduce
oxygen-containing groups onto the subshate surface through reactions with ozone, water,
oxygen, nitrogen, and different free radicals (Briggs at al. 1983). The extent of each
reaction depends on the composition of the reaction gas. The corona discharge reactions
can degrade and clean the subshate surface (Kim at al. 1971, Carley er al. 1978), form
sites for hydrogen bonding (Owens 1975, Blythe at al. 1978, Lanauze at al. 1990),
increase the surface free energy (Carley at al. 1978, Baszkin at al. 1978), and cause
surface cross-linking (Kim et al. 1971).

Aglowdischargeplasmaissimilartocoronadischargebutitisproducedataless

 

 

 

73

than ahnospheric pressure. Glow discharge can be established by AC, DC, or
elechomagnetic power inputs, but radio-frequency (RF) plasma are the most common
method because of their efﬁciency in sustaining the plasma. Plasma generated by RF
excitementcanbecontainedintheRFﬁeldmrimaryplasma) orearried outsidebygas
ﬂowand diffusion(secondary plasma). Dependingon thereactorconﬁgurationprimary
or secondary plasma may cover the working volume (Rose er al. 1986). In general,
conholling parameters for a glow discharge plasma are: process gas, power dissipation,
excitation frequency, gas ﬂow rate, and plasma chamber geomehy. A schematic diagram
ofaglowdischargereactionchamberisshowninFigureSJ.

 

 

 

was i

we; "w i

Figure5.1-Schemaﬁcdiagramofthesurfacemodiﬁcaﬁonofpolymersinaglow
dischargereactor.

 

 

 

 

 

 

 

 

74
The most commonly used gases in a glow discharge are oxygen, nihogen, air, argon,

helium, nihous oxide, ammonia, water, and tehaﬂuoromethane (CF4). Each gas or
mixture has a distinct plasma characteristic. The efﬁciency of the chemieal processes
alsodependonthegaspressureandenergyinput. Decreasing thegaspressureand/or
increasing the RF power increases the mean free path, degree of excitation, and
concenhation of the active species that are important in aggressive processes such as
cleaning and degradation. In general, low pressure and high RF power provides fast but
more surface sensitive reactions.

Rose et al. (1986) and Liston (1989) have reported on the effects of plasma
heahnents on polymers. In general, plasma heahnents of polymers produces four major
effects: cleaning (removal of unbounded contaminations), surface activation (wetting or
non-wetting), chain cross-linking and etching (removal of the polymer subshate). For
any set of process condition and gas chemishy, all effects are present to different extent.
Typically, an oxygen and/or nitrogen containing plasma can produce surface cleaning,
highersurfaceenergiesandreactivepolargroups. Theefﬁciencyofchainscissionis
vastlyenhancedbymixingCEintoaprocessgas. AgasplasmaofCF.andO,isa
particularly aggressiveplasmaformostpolymersandisusedfortheetching heahnents
ofpolymers. Inertgasplasmaseanbeusedtocross-linkthepolymersurfaces. Inan
inertgasplasma, thepolymerchainsthatarebrokenbytheactivatedspeciesofthe
plasmaarewithoutchemicaﬂyacﬁvemdicalsinhlegasphasetoreactwith,sothe
polymersitesrcactwitheachotherandcross-linkthepolymerchains. Inertgasescen
alsoenhancethesurfacewettingbecausetheycreatesites forpolarfunctionalgroupsto

 

 

 

75
attach to when exposed to oxygen or nihogen in the air (Nakayama er al. 1991). Polar

groups could be inhoduced to the polymer subshates when the plasma gas contains
nihogen, ammonia, nihous oxide, oxygen, or water (Hansen et al. 1965, Holmes er al.
1990). A pure CF, gas plasma, however, lowers the surface energy of the polymer
subshate and makes the polymer non-wettable because the polymer chains become similar
to ﬂuorocarbon chains.

Plasmareactors canalsobeused toformpolymermaterials fromamonomerplasma
gas. Plasma polymerintion has several advantages over the conventional polymerization
techniques including thin ﬁlm formation, perfect conformance to the subshate contours,
one-step monomer synthesis and polymerintion, and cleaning and/or conditioning of the
subshate surface before deposition. Plasma polymerization processes have good subshate
surfacepenehationandcoveragethatpromotesgrafting thedepositedpolymerﬁlmbthe
subshate. The polymer grafting resulting ﬁom plasma polymerization is not notably
affected bythenatureofthesubshateandprovideslittlealteration ofthebulkproperties
of the subshate (Yasuda 1985). Polymer ﬁlms produced by plasma polymerintion can
beconsidaedascoupﬁngagalhfmﬁba-mahixadhesimbecauseofdleirabiﬁtym
graftwiththesubshateﬁberandthencovalentlyreactwith theresinmahix. Plasma
polymerizationhastheabilityofdepositingaﬁlmofvirtuallyany monomerthathasa
vapor phase (Bell 1983). In a study by Inagaki er al. (1982) several polymer subshates
waeﬁrstexposedtoargonplasmatocleandlesubshateanddlulamonomergas
(himethylsilydimethlamine or hexamethyldisiloxane) was injected to form the polymer
ﬁlms. They have reported improved adhesion between hexamethyldisilazane heated

76
ﬁbers with epoxy resins.

Occhiello er al. (1991) have examined the adhesive properties of oxygen plasma
heated polypropylene (PP) sheets. The locus of failure for PP-epoxy joints was found
to be adhesive for the unheated samples and cohesive for the plasma heated samples.
For the plasma heated PP, the PP-epoxy joints failed within the PP bulk, but close to the
modiﬁed PP interphase which suggests the cohesive shength of the PP had beconne the
limiting factor in the PP-epoxy adhesion. The PP-epoxy joints also showed increases in
the shear shength which was athibuted to the inhoduction of polar functional groups to
the plasma heated PP surfaces. Similarly, for plasma heated polyethylene ﬁbers
Ladzesky at al. (1983) have reported a shift from an adhesive failure for the unheated
sample to a cohesive failure for the heated samples.

Therearereportsontheimpmvementofaramidadhesivepropertiesbyglow
discharge heahnents, however, these reports show that ﬁber-matrix adhesion
improvements are accompanied by ﬁber sherngth deteriorations. Wertheimer er al.
(1981) have reported 10% to 85% increases in the peel shength of the aramid-epoxy
composites after exposure to a microwave plasma, but along with a 35 % reduction in
ﬁber tensile sherngth. Allred er al. (1985) have reported a glow discharge heatment for
ararnnid ﬁbers that improves their adhesive properties without ﬁber shength deterioration.
UsingaRFplasmaindlepresenceofammoniagasAlhedhasreponedatwo-fold
increase in the interlaminar peel shength of heated Kevlar-49/epoxy composites arnd
faﬂummodechangesﬁominterfacefaﬂuretomixMresofﬁberandnnhixfaﬂure. 6th
er al. (1990) have also report doubling of interfacial shear shength betweern various

77
plasmaheatedKevlar-49aramidﬁbersandanepoxymahixasdeterminedbyadroplet

test. Conversely, using a droplet test method, Kupper er a1. (1991) have reported 10-
20% increase in the ISS values of various plasma heated Kevlar-49 ﬁber, but 10-20%
ISS reduction for the plasma heated Technora ﬁbers (note that the droplet test technique
typieally show about 20% error). In general, the large increase in the interfacial slnear
shengthoftheplasmaheatedaramidﬁbersreportedbyAllrederaI. (l985)andGaur
er al. (1990) have not been widely validated by other researchers and plasma heahnents

of arannid ﬁbers merit closer examination.

5-2 EXEERIMENIAL

The polymer ﬁbers examined in this study were p-Phenylene BenzobisOxazole (PBO)
aromatic heterocych supplied by Dow Chemical (Midland, MI) (ref. 4' XV-0383-
C8700975-008) and ulha-high-modulus Specha— 1000 polyethylene ﬁbers supplied by
Allied Signal (Morristown, NJ). Botln ﬁbers were unsized and were used 'as received“.

Plasma heahnernts of the PBO ﬁbers were conducted by Plasma Science, Irnc.
(Belmont, CA). Plasma heated PBO ﬁbers were sealed in nihogen-purged bags and
wereshippedovernight. Plasmaandcorornaheated Specha-lOOO ﬁberswere provide by
the Allied Signal, heated with their proprietary heahnernt condition.

The epoxy systems used were the DER331/MPDAIDETA 175°C/3hr (fragmentation
test), and DER331/MPDA RT/24hr/75 °Cl2hrl 100°C/3hr (droplet test). Single ﬁber
tensileshengthsweremeasuredbyASTMD3379tensiletest. Theseexperimental
conditions are detailed in Chapter 2.

 

 

78
5.3 W
Results of PBO and polyethylene plasma heahnents are discussed separately but
conclusions for both ﬁbers are combined to develop a comprehensive understanding of
the effects of plasma and corona heahnents on the surface properties of high performance

polymer ﬁbers.

53.1 We

InChapter3,itwasdemonshatedthatadhesivepropertiesofthePBO ﬁbersare
limited by a cohesively weak surface layer that fails within the ﬁber during interfacial
separation. Plasma heahnent could improve the ﬁber adhesive properties by variety of
mechanisms such as etching the ﬁber skin and/or cross-linking the surface polymers to
shengthen the ﬁber surface properties. PBO ﬁbers also exhibited lower polar surface
energy titan liquid epoxy which suggest incomplete wetting with liquid epoxy. Plasma
heahnents could inhoduce polar functional groups to enhance the PBO-epoxy wetting
compatibility.

In this study, plasma heahnents of PBO ﬁbers with variety of plasma gases have been
examined to determine which heahnernt is the most effective for enhancing the adhesive
propertiesofthePBOﬁber. Table5.1liststheheatmentprotocolfortheﬁrstsetof
plasma heahnernts. A variety of different plasma gases and conditions were examined.
Table 5.1 also lists the possible polymer surface alteration meclulnisms that are expected
to dominate for each heahnent condition. Figure 5.2 shows percent changes in tensile
mldmterfadalsheushungmsofthephsmheamdﬁberswithmspectwdleunheawd

 

 

79

Table 5.1 — Plasma heahnent condition for the ﬁrst set of PBO plasma heahnernts. The
expected major surface changes are also listed.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A O, 50% 300 3.00 Cleaning
B O, 25% CF, 25 % 300 3.00 Etching
C He 50% 300 3.00 Cross-linking
D CO, 50% 299 3.00 Polar-sites
E NH, 50% 299 3.00 Polar-sites
F NH, 50% 445 3.00 Polar-sites
G N20 50% 299 3.00 Polar-sites
II N20 50% 455 3.00 Polar-sites
I Ar 50% 300 3.00 Cross-linking
112/N, 50% 299 3.00 Cleaning
K Ar 50% 299 3.00 Cross-linking
+ 100% CO, 0 3.00 + Polar-sites
L Ar 50% 299 3.00 Crow-linking
+ 100% O, 0 3.00 + Polar-sites
M H20 100% 299 3.00 Clearning

 

80
ﬁber. In general, tensile shengths are not affected by the plasma heahnents except for

the OJCF, (heahnent B) which slnow about 28% tensile shengtln reduction. The OJCF.
plasma is a highly corrosive heahnent that results in etching removal of the polymer
subshates. Figure 5.2 also shows that some heahnents have increased the interfacial
shear shength (ISS) of the PBO ﬁbers. For example, heahnent D (C0,) exhibits we
40% ISS increase. The ONE and CO, plasma heahnents slnowed the greatest inﬂuence
on the PBO properties and wee chosen for further examination.

For the second set of PBO plasma heahnents, O,ICF, and CO, heahnents were
examinedwithvarious inputenegyandexposuretimes. Table5.21iststheexperimenta1
protocol for the second set of PBO plasma heahnents. Figure 5.3 shows percent

 

50 - - Tensile Strength
- 1:] Interfacial Shear Strength
40 r i-

OF L
0, 11:1..[1_IU__
O

ABCDEFGHIJKLM
Treatment

Figure5.2-Percentchangesintensileandinterfacialshearshengthsoftheﬁrstsetof
plasmaheatedPBOﬁbers.

 

 

 

 

 

 

 

 

 

 

2 Change from Control Value
I

l
N
O

 

 

 

 

I
U
0

81

Table 5.2 - Plasma heahnent condition for the second set of PBO plasma heahnents.

Gas Power Time
(Fiow %) ' (watts) (min.)
N 301 0.50

 

 

 

 

 

 

 

 

 

 

 

ca2 50%

o ca, 50% 301 0.75
r ca, 50% 301 1.00
Q ca, 50% 479 0.50
R (:0, 50% 479 0.75
s ca2 50% 479 1.00
r 02 25% CF, 2595 301 0.50
U o, 25% or, 25% 301 0.75
v o, 2595 CF, 25% 301 1.00
w o, 25% CF, 25% 395 0.50
x 02 259': or, 25% 397 0.75
Y o, 25% CF, 25% 395 1.00

 

 

 

82
changesmtensilemdmmrfacialshearmmgmsofmephsmaheawdﬁbeswimrespect

to the unheated ﬁber. All plasma heahnents show interfacial shear shength increases,
but the OJCF, heahnents generally show greater increases than CO, heahnents. In
particular heahnent W (Oz/CF“ 395 watts, 0.5 min) shows about 56% ISS increase.
Figure 5.3 also shows that except for heahnent Y (Oz/CF“ 395 watts, 1.0 min), the
examined plasma conditions do not signiﬁcantly affect the PBO tensile shength. These
results illushate the viability of plasma heahnents for improving interfacial shear shength
of the PBO ﬁbers. The mechanisms of adhesion improvements for the plasma heated
PBO ﬁbers are examined next.

Asdiscussed previouslyinChapter 3, virginPBOexhibitinternalldnkbandsthatare

 

60
- - Tensile Strength ﬂ

: [:1 Interfacial Shear Strength

45
T m

30

' T j I T V

15

 

 

 

 

 

 

 

 

 

 

 

I
—h
0'1
v I v v

7. Change from Control Value

I
U
0

I

 

 

 

l
rs
U"

NOPQRSTUVWXY
Treatment

FiguRSJ-Peemtchangesmtelsileandmwrfacialdlearshengthsofmesecmdse
ofplasmaheatedPBOﬁbels.

83
not apparent on the ﬁber surface. Figure 5.4 illushates the SEM micrographs of 0.5,

0.75 and 1.0 minutes O,/CF4, 397 watts PBO plasma heahnents (W, X, Y). These
rnicrographs show the development of PBO surface etching by the plasma heahnent, as
evidenced by the gradual appearance of the ﬁber internal kinks. Initially, the O,ICF,
plasma heahnents increase the PBO-epoxy interfacial shear shength, but as the surface
etching proceeds, eventually the ﬁber tensile properties begin to deteriorate. Treahnent
Y exhibit about 40% tensile shength reduction.

TEM micrographs of heahnent W are shown in Figure 5.5. The rnicrographs still
show internal ﬁbrillation of the PBO ﬁber but the external ﬁber surface layer that was
observed for the unheated ﬁbers (Figure 3.12) is now absent. Plasma heahnent results
suggest that the etching removal of the weak surface layer of the PBO ﬁbers earn
sigrniﬁcantly (~50%) enhance their interfacial shear shengths with an epoxy resin;
howeve, once the ﬁber skin weak boundary layer is removed, the cohesive ﬁbrillation
of the ﬁber limits the interfacial load hansfer of the PBO ﬁbes. Therefore, for the PBO
ﬁbers plasma etching must be optimized to remove the weak ﬁber surface layer without
exceeding and deteriorating the ﬁber tensile properties. ‘

SEMexaminationofthecozplasmaheated ﬁbersdidnotexhibitdiscemable surface
morphological changes like those observed with the OJCF, plasma heahnents.
Therefore, the ISS increases of the CO; plasma heated ﬁbers (heahnents No8) should be
mainly tlne result of sonne chemical modiﬁcation of the PBO ﬁber surfaces. Figure 5.6
compares the surface energy ofseveal plasma heated ﬁbers with theliquid epoxy. Note

that the heahnent P (C02) exhibits a much improved epoxy compatible surface energy

 

 

ﬁgure 5.4 - SEM micrographs of (A) 0.5, (B) 0.75 and (C,D) 1.0 min. 0,/CF., 397
watts PBO plasma heatments.

 

Figure 5.5 - TEM nnicrographs of a 1.0 min. 02/CF., 397 watts PBO plasma heated
ﬁbe (heatmentY). (A)bar =2um (B)bar =250 nm

 

86
components compared to the unheated PBO ﬁbers. The 0,ICF4 plasma heated ﬁber

(heahnents U, V, and W) also exhibit increased polar component of surface energy
which enhances their wetting properties with the liquid epoxy. However, the surface
energy components of the O,/CF4 are less than optimum, and better wetting properties
are expected if the surface energy components could better match the epoxy components
(Gutowski 1990). Inhoduction of ﬁber-matrix covalent chemieal bonding for the CO,

plasma heated ﬁbers is another possible mechanism of adhesion improvement.

a.
O

 

U
o
l
J .
j

N
O

p ...................................................................

Polar
8
LI

M

O

 

(dyne/cm)

_a
O

 

 

 

 

 

 

--------------------------------------------------------

Dispersive
N
O

(A
O

-------------------------------------------------------

 

 

 

'l/I/I/I/I/I/Il

 

 

 

 

 

Epoxy Untreated P U V W

Figure 5.6 - Polar and dispersive components of surface energy for second set of plasma
heated PBO ﬁbers.

 

In Chapter 3, the low adhesive propeties of the Spectra-1000 polyethylene ﬁbers
were attributed to their poor surface energy compatibility with the liquid epoxy. In this
section, effects of plasma and corona heahnents on adhesive properties of Specha-1000
M are examined.

Figure 5.7 compares the interfacial shear shength (ISS) of the plasma and corona
treated Spectra-1000 ﬁbers as determined by the droplet test. Both heahnents produce
about 300% increases in the ISS values compared to the unheated ﬁbers. Figure 5.8
compares the surface energy components of the ﬁbes with a liquid epoxy. There is now
a polar component of surface energy for the plasma heated ﬁbers which is quite
signiﬁeant when compared with the unheated ﬁbers. However, the surface energy of the
cororna heated ﬁbe's only exhibit an increase in the dispersive component of their surface

 

_a
-F O) m 0
I 1

Interfacial Shear Strength (MPa)
N

 

 

 

 

 

O

Untreated Corona Plasma
Treated Treated
Figun5.7-mtefacialshearshengdnofunheated,cotheawdandphsmaheawd
Specha-IOOO polyethylene ﬁbe's as determined by the droplet test.

88
energy. Figure 5.9 shows the atomic concenhation of the examined ﬁbers as determine

by XPS. The oxygen contents of the corona and plasma heated ﬁbers are about 4 to 5
times higher than the unheated ﬁbers which suggests presence of new chemical
functiornalities on the heated ﬁbers. These new chemieal fnnnctionalities may provide
sites for covalent chemical bonding between ﬁber and mahix molecules.

Figure 5.10 shows superimposed XPS narrow scans of oxygen signal for the plasma
and corona heated ﬁbers. Note that the plasma heated ﬁbers show additional presence
of oxygen signal in the 532-534 eV region. For the plasma heated ﬁbers, their increased
polar component of surface energy may be due to this additional type of oxygen. Other
researchers have also reported the addition of new polar functional groups to the surface
of plasma heated Specha—lOOO ﬁbers (Mona et al. 1990, Nguyen er al. 1988). The
increased numbe of polar functional groups on the plasma heated ﬁbers increases their

 

 

(dyne/cm)

20

Dispersive

 

 

30

 

 

 

 

 

 

 

 

40
Epoxy Untreated Corona Plasma

Treated Treated

Figure 5.8 - Polar and dispersive components of surface energy for unheated, corona
heated and plasma heated Specha-lOOO polyethylene ﬁbers.

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

120
A _ Oxygen
N [:1 Carbon
V100 ..................... . ........................................................
c ﬂ __
.9 l- -———-
45 80 ............ ., ..................................................
L.
44 r-
8
O 50 y. ..............................................................
c
o l
O 40 . .................................................................
.9
E
O 20 _ ......................................................................
.o—I
“ N N
Untreated Corona Plasma
Treated Treated
Figure 5.9 - XPS atomic concenhations of unheated, corona heated and plasma heated
Specha-IOOO polyethylene ﬁbers.
1 : n s . . : 1f . t e .
v 1» -
, .. " :1
1|- lb 3:
. T «n- :3
lb 1b i
I " 0 L
0 Pm 0 3g
‘ 0 ' i' 2-
g s 0 1n g
‘ " 1» T
.. Corona 0
’ 1- 1p
lb 4-
2 tr 1-
.2‘ .
l .I
0 c t : 4 + s 3 +4 : ¢ ; :
ﬂ 9! a u H
mm d

Figure 5.10 - Superimposed XPS signals of oxygen region for corona and plasma heated
Specha-lOOO polyethylene ﬁbers.

2.
‘ s
._
~
‘

i
.2
2
a
a.
h
.a
g
a.
p
‘
‘.
I

'Pjifﬂnr'

 

a.

 

 

90
polar component of surface energy, thus, enhancing the ﬁber thermodynamic wetting

with a liquid epoxy.

The previous observations indicate that despite the chemical and surface energy
diffeences between the corona and plasma heated polyetlnylene ﬁbers, their ISS increases
are similar, suggesting an adhesion improvement mechanism that is only moderately
dependent on the chemieal modiﬁeation of the ﬁber surfaces.

Etching of the polyethylene surfaces by oxygen plasma and corona heahnents that
could inhoduce mechanieal interlocking adhesion mechanisms that can signiﬁcantly
enhance ﬁber-matrix mechanical interactions (Nguyen et al. 1988, Choe er al. 1990).
Figure 5.11 compares SEM nnicrographs of an unheated Specha—lOOO ﬁber with plasma
and corona heated ﬁbers. The unheated ﬁber has a relatively smooth surface with some
surface crazing that is attributed to their manufacturing process (Postema et al. 1987).
Conversely, botln corona and plasnna heated ﬁbers exhibit rough and pitted surface
topography. Theplasmatreatedﬁbersshow moresurfaceetchingtlnanthecoronaheated
ﬁbers. Ladizesky et al. (1982) have reported that plasma heahnents of ulha-high-
modulus polyethylene ﬁbers produces an etched surface, into which resin ean penetrate
to produce mechanical interlocking between ﬁbe and resin. Unheated polyethylene
ﬁbes exhibited smooth surfaces and the locus of ﬁber-mahix failure is interfacial,
whereas, the plasma heated ﬁbers showed ruptures within the ﬁlaments during ﬁbe-
nnatrixseparatibn. Therefore, theISSincreasesofthecoronaandplasmaheated
Spectra-1000 ﬁbe's ean be mainly attributed to the mechanical intelocking mechanism

inhoduced by these heatments.

 

 

Wmhmmp

{(30)}:an .. "ll ”W3 “3' 3'9 It
"dj" “('30 u \‘g .‘,“‘,d ‘lﬂﬂy’;

I?“ {)3me ”(all 10“.] cu.
...emw , 3.9 flank m‘nwt, '

t MM '
1’,“ '1) ”s:’\"‘. ,’/’.s”. .‘. _ a vLVI¢'

.I‘Cv“".w.s P ‘ if}: "‘t‘ film "I ’ ’ 1’

MHH‘ mn-

. (W ‘0': " w-;! u'-.9
lam. ft"; linen-.1393 W

Etc '38:! mfvljl“ 4‘" ' 7" ' ‘U '0‘"

Figure 5.11 - SEM micrographs of (A) unheated, (B) corona heated, and (C) plasma
heated Specha-IOOO polyethylene ﬁbers.

         
 

  
 
         

5-4 W

0 Plasma heahnents of PBO ﬁbe's suggest that the etching removal of the weak
surface layer oftlne ﬁbers ean signiﬁcantly (~50%) enhance their interfacial shear
shengths with an epoxy resin; howeve, once the ﬁber skin weak boundary layer
is removed, the cohesive ﬁbrillation of the ﬁber limits the interfacial load hansfer
of the PBO ﬁbers. Plasma heahnents can also signiﬁcantly enhance the ﬁber
wetting compatibility with liquid epoxy resins.

0 Plasma and corona heahnents of polyethylene ﬁbers ean increase that inte'facial
shear shengths by about 300% . The polyethylene-epoxy adhesion improvements
are mainly due to mechanical interlocking mechanisms inhoduced by micropitting
oftlneﬁbersurfaces. Wettingandchemiealbondingpropetiestheﬁber-mahix

interphasemayalsobeenhancedbytheplasmaandcoronaheatments.

Reulhofmisshldydemmshatematphsmaundcomsurfaceheahnenttechnique
can enhance wetting propeties of polyme ﬁbe, produce surface roughness, remove
weak surface layers and inhoduce functional groups for covalent chemical bonding.
Therefore, for high performance polyme M such as PBO, Specha-lOOO, and
Technora which exhibit adhesive surface limitations, plasma and corona surface
heahnnents are viable approaches to oveconne their respective surface limitations.
However, once these surface limitations are overcome, the ﬁbe lateal cohesive

' propeties become the limiting factor.

CHAPTER 6

 

Sulfonation and Fluorination

Treatments of Polymers

 

This chapte presents a discussion of ﬂuorination and sulfonation polymer surface
heahnents. Through these chemical heahnents, effects of ﬁber-mahix inte'phase
chemishymadhedvepropehesofhighpefomancepolymerﬁbesareexamined.
Examination of polymer sulfonation also helped to develop valuable insights on unass-
hansfe phenomena that may limit the extent of ﬁber shuctural modiﬁcation (discussed
inChapter8). Thesndfonationporﬁonofthisstudyhasbeenconductedincollaboration
with Dr. Y. Muraoka (Muraoka er al. 1991“", Kalantttr er a1. 1991' ).

93

6-1 WON

Chemical modiﬁcation of polymer surfaces by ﬂuorination and/or sulfonation
heahnents have reported to enhance adhesive properties of the polymer subshate. Dixon
er al. (1976‘,l977‘) have reported improved adhesion and water hansport properties for
the ﬂuorinated synthetic ﬁbe's such as polyesters, polyoleﬁns, and polyacrylonihiles.
Walles (1989) has also reported improved adhesion, wettability, abrasiorn resistance,
vapor barrie W of polymers through sulfonation surface heahnents.

Figure 6.1 shows the geneal reaction schemes for the fluorination and sulfonation
of hydrocarbons. Generally, ﬂuorination of the organic polymes proceed to form a
ﬂuorinated carboxylatcd (acid ﬂuorides, -FC=O) on the polymer subshate; tlnese

 

Fluorinaﬁon —(|3H + F2 —->-(|3F + HF
I l
Sulfonation — CH + $0 —> - 0803H

l 3 I

| | _
Neutralization ”0303” + "Ha —>—cso NH"

I I 3 4

 

 

 

Figure 6.1 - Geneal reaction schemes for fluorination and sulfonation of hydrocarbons.

95
carboxylate groups are only the ultimate reaction products, and are formed nnainly after

the polyme is exposed to oxygen (Dixon er al. 1976 ", 1977 |'). The fluorination
reactions also produce the highly corrosive HF species as a byproduct ofthe ﬂuorination
reaction. Conversely, Sulfonation of polyme's do not produce corrosive byproducts like
the fluorination heahnents. Furthermore, sulfonated polymers can be neuhalized with
any numbe of cations to produce a variety of barrier properties (Walles 1989). Muraoka
er al. (1991') have presented a discussion on the reaction schemes of polycarbonate
snnlfonation.

Fluorination of polyme's is carried out by exposing the polymer surface to a
ﬂuorinating mixture comprising from 0.1 to 20 vol% elemental ﬂuorine gas and the
renaindeacarriegassuchasargonornihogen. Thelevelofoxygeninthe
ﬂuorinatingmixmreiskepttoaminimumsincehighlevelsofoxygencouldbe
dehimental to the heahnent (Dixon er al. 1977’).

Sulfonation of polymes can be carried out with sulfonating agents such as sulfur
hioxide or its adduct with various organic compounds, sulfuric acid, pyrosulfuric acid
(oleum), chlorosulfonic acid. Gas phase sulfonation of automobile gas tanks is a
common indushial application of polyme sulfonation heahnent and is conducted by
exposing the polyethylene tank to ~10 vol% SO, in N, carrier gas for about 10 minutes,
and then neuhalizing with NH, gas.

Othetypesofchennicalheahnentstoimproveadhesivepropetiesofhigh
performance polyme ﬁbe's have also been reported. Mercx er al. (1990) have reported
up to 70% ﬁbe pull-out shength increases for the oxalylchloride surface heahnent of

96
Twaron arannid ﬁbers. Wu et al. (1986) have reported on incorporation of amine

functional groups on the Kevlar-49 aramid ﬁbers by bronnination reactions followed by
ammonolysis, nihation, and reduction. They report doubling of the T-peel shength and
up to 50% interlarnirnar shear shength increases for the heated Kevlar-49 composite
specimen. The improved ﬁber-mahix of these ﬁber heahnents were accompanied by a
shift of locus of failure from adhesive ﬁbe failures to cohesive ﬁber ﬁbrillation.

In this study, fluorination of Kevlar-49 ﬁbers and sulfonation of Kevlar-49, Technora,
and PBO ﬁbe's are investigated to assess the effects of ﬁber-mahix intephase chennishy
on adhesive properties of high performance polymer. Sulfonation of polyethylene and
polycarbonate sheets have also been examined to develop an understanding of mass-

hansfer phenomena that may be limiting the extent of polymer heahnent penehations.

6-2 W

Polymer ﬁbe's examined in this study wee Kevlar-49 (3.1. du Pont, Wilmington,
DE) and Technora (Teijin Limited, Japan) aramid ﬁbes, Specha-1000 ulha-high.
modulus polyethylene ﬁbers (Allied Signal, Morristown, NJ), PBO (p-Phenylene
Benzobis Oxazole) aromatic heterocyclic ﬁbes (Dow Chennical, Midland, MI). The
polycarbonate material was LEXAN 8050 sheets (thiclmess 20 mils, non-UV stabilized)
supplied by Geneal Elechic Co. (Pittsﬁeld, MA).

Three epoxy systems wee used in this study,: DERS31IMPDA/DEI'DA 175°C/3hr,
DERB3l/MPDA 75 °C/2h/125°C12hr (fragmentation test) and DER331IMPDA
RT/24hr/75°C/2hr/100°C/3hr (droplet test). Single ﬁbe tensile shengtlns wee measured

97
by ASTM D3379 tensile test. These and other experimental procedures are detailed in

Charmit 2-

The ﬂuorination heahnents of Kevlar-49 ﬁbes were done by Air Products and
Chemicals Inc. (Emmaus, PA). Four proprietary gaseous surface heahrnents designated
as 9629-80-A, 9629-80-13, 9629-80-C, and 9629-80-D (called heahnents A, B, C, and
D respectively) were provided. These heatments showed a progressive development of
a green coloration. Fibers wee slnipped via express mail in nihogen purged, heat sealed
bags, and wee cast in epoxy immediately after their arrival.

Thesulfurhioxideused forthesulfonationheahrnentswas suppliedinthestablesolid
form (Aldrich Chemical, WI). The Freon solvent used for solution heahnents was 1,1,2-
hichloro—l,2,2-hiﬂuoro ethane (Eashnan Kodak, NY). A solution of SO, in the Freon
solvent was used for the sulfonation medium. The SO,/Freon solution was prepared by
pouring Freon solvent into a one liter glass vessel precharged with solid SO, solid
polymers. The solution was leﬂ for a week at ambient tempeature to equilibrate S0,
and Freon solubility. Glass-wool was inserted into the bottle to prevent the polymer
sample from coming into contact with solid S0, material at the bottom ofthe ﬂask. The
sulfonationwascarried outbyplacing mateialsintotheSOJFreonﬂask. Theﬂaskwas
shaken by hand occasionally.

Three heahnent temperatures nominally called warm, ambient, and chilled wee
examined. Fordnechilledsulfonaﬁon,tlneﬂaskwasplacedinafreezermaintairnedat
-17 j; 2°C. Ambient temperatures wee maintained at 22 :l: 3°C. For the warm

sulfonation, the ﬂask was place in a convection oven maintained at 37 :1: 3°C.

98
Detemination of the SO, concenhation in the solution was carried out by exhacting SO,

from the Freon solvent into water and then tihating with a NaOH solution. The
equilibrium SO, solution was determined to be 0.012 N for the chilled solution, 0.037
N for the ambient temperature solution, and 0.052 N for the warm solution. For the
polycarbonate sulfonation all three heahnent tempeatures wee exannined, but for the
polyethylene sulfonation only the ambient heatnnent tempeature was examined.

For the polyetlnylene samples, color changes from an irnitial bright white to brown
wee observed as the sulfonation reaction proceeded. For polycarbonate and arannid
samples no color changes wee observed during the sulfonation.

To assess the effects of neuhalization process on the polycarbonate sulfonation,
samples wee neuhalized either with l N aqueous NFLOH solution for 30 minutes, or
lNAgNO, for 1 hour, orleftunneuhalizedbyrinsingonlywithFreonorwater. For
the unneuhalized polycarbonate samples, a thin brown ﬁlm was obseved to form when
sampleswereexposedtoarnbientair. Thisbrownliquidﬁlmisprobablydueto
envhmmentalmoMnthatreachwithdwexcessSQontheheatedsurfaces. The
liquidﬁlmwaslesspresentonthechilledsulfonated samplesandmoreonthewarm
sulfonated samples.

Forthesulfonationheahnentoftheﬁbers, ﬁbeweeﬁrstdriedat125°Cfor4hours
toremovetlneirabsorbed water. Theﬁbesweetheninseted in0.0003Nsolutionof
SO,inFreon for30hours. Fibers weedriedatambientconditions for l hourarndthen
cast in the 75°C/2hr/l25°Cl2hr epoxy system. .

SulfonahonpenehationofheatedsampleweedeeminedbyanAEStechnique.

99
DetaﬂsofﬂnesampleprepamﬁontechniquefmﬂneAESanalysisareprwentedin
AppendixC(Kalantaretal. 1990‘). Insummary,thesampleis initiallycoatedwitha
thick gold laye (~50 nm) andthenthe analysissurfaceisslnaved witlnadiamond krnife
to prepare a smooth and clean surface surrounded by gold boundaries. The analysis
surface is recoated with a thin gold layer (~ 2 nm) which helps to avoid sample charging
butisthinenoughtoallowtheemittedAugerelechonsfromthematerialbelowthegold
layertobedetected.

All ABS analyses wee carried out using a Perkin-Elmer PHI 660 Scanning Auger
Microprobe. Samples were analyzedat 10000to 30000x magniﬁcations. AES beam
cornditionsforanalysiswere 1.5to3nAbeamcurrentand3tolOkaeamenergy. The
unfurAESsigrnalmtensitywasmomwredueachpointmmesampleﬁomhwsigml
peakheightandwasplottedinline-scanfaslniorn. Ashort(>50nm)ionbeamsputtering
oftheheahdmrfaceswasconductedbrenovednemrfacecontaminathutlmge
sputte'ing times wee avoided because high dose sputtering of polymers would
prefeentially remove the non-carbon (sulfur and oxygen bee) surface elements and
wouldproduceacarbonizedsurfacecomposition(seeChapte7).

The surface compositions of the samples wee examined by XPS. Sample surface
texturebeforeandafteheatrnents wereexaminedby SEM,operatingat15.0kaeam

energy and 5000X magrniﬁcation.

1w
6.3 W

The results of the ﬂuorination and sulfonation heahnents are discussed separately to

provide an exclusive discussion for each type of chemical heahnent.

6.3.1 Aramiililuorination
Table 6.1 shows XPS atomic composition of the ﬂuorinated Kevlar-49 ﬁbe's. Note

that the unheated Kevlar-49 ﬁbers exhibit an oxidized surface composition as evidenced
by their large oxygen content compared to their monomer stoichiomehy. Treahnents A,
C, arnd D have resulted in similar levels of ﬂuorination (~ 22 atomic%), while,
heahnent B has resulted in higher ﬂuorine content (~32%) and lower oxygen content
(~12%) than the other heahnents. The atomic% ratios of oxygen/carbon,
nihogen/carbon, and ﬂuorine/carbon are plotted in the Figure 6.2. Treahnents A, C,
andDhaveresultedinO/Cratioshighethantheunheatedﬁberswhichsuggests

Table 6.1 - XPS atomic composition of ﬂuorinated Kevlar-49 aramid ﬁbers.

Theoretical 78 ll 11 '

Unheated 73 20 7.4 -

Treatment A 52 20 5.0 23
Treahnent B 51 12 4.7 32
Treahnent C 53 20 5.2 22
Treahnent D 51 19 4.5 25

 

 

 

 

 

~4% Mnmwﬁrudnwm

101
formation of acid ﬂuorides on the heated ﬁbers. Treahnent B, however, shows O/C

raﬁolowethanunheatedﬁbeswhichmaybeathibutcdtoremovalofﬂneoﬁginal
oxidized laye of the aramid subshate by the ﬂuorination heahnent.
Tensﬂeshengthsofﬂwﬂuoﬁnatedﬁbesalsosuggesthmtechingalterahmofﬂle
aramidsurfacesbytheheahnentB. Figure6.3comparestensileshengthandinterfacial
shear shength (fragmentation test) of the ﬂuorinated Kevlar-49 ﬁbes with the unheated
ﬁber. TensileshengthofheahnentBﬁbersisabouth% lowerthanotheﬁbeswhich
conﬁrrrns substantial ﬁber chain scission has occurred by this heahnent. The interfacial
shear shengths ofthe ﬂuorinated ﬁbers, howeve, is unaffected by the heahnents.
Figure 6.4 shows TEM nnicrographs of ﬁber-mahix interface for the heahnent D.
The micrograph shows improved ﬁbe-nnahix adhesion but the interfacial ﬁbrillation have

beenshiftedtowardtheﬁberinterior. Otherﬂuorinationheahnentsshowsimilar

 

[:1 N/Cratlo
0-5 * 0/c ratio
m F/C ratio

.99
.0...‘
6.00000094

1
d
..
.0
O

0
..°_
0
O

O
O
O

O

9
4a
I
9:.

O

O
9
6

O
O

O
O
O

0......
:0
9.:

o

O
0

Ratio
0

at:

e 0’.

O

0.9.0‘0 O O O O'
to?
.0 O
O O

O
O

O
0..
e
e

O

O
0

O
O

O
O
O

s
o
e
e
e
o
e

O

. O

.0

O

O
O
0

6’0...
9 O
O...

O

.99.
O.
0.6...

O

OOO‘OOO'OO
O

' O
.9.’
OOI’O.O...O.O.O...O.O O 0.6 0..
O

O
O
O

/////////////////1
//////////1

O O
9.6

O
O

O
.0

 

 

           

//////////////////A

   

l/I/I/I/I/I/I/I/I/A

     

O O O Q
.0.0.0.0.0

H H

Untreated A C D

Figure 6.2 - Ratios of atomic% of nihogen, oxygen, and ﬂuorine ove carbon, for the
ﬂuorinated Kevlar-49 arannid ﬁbe's.

O

0.0

 

[I]

102
improvedadhesiontotheheatedaramidﬁbesurfaces. Thelackofincreaseinthe

aramid-epoxy intefacial shear shength despite the improved ﬁber-mahix adhesion
suggeststlnattheadhesivepropeties ofthearamidﬁbersarestilllimitedbythecohesive
failures ofthe ﬁber rather than the extent of ﬁber-mahix interfacial bonding.

SEM rnicrographs of unheated and heahnent B ﬁbe are shown in Figure 6.5. The
heatedﬁbeshowssmmdnersurfacetexmreaigure65mhnntheunhcatedﬁbes
suggesting morphological modiﬁcation of the ﬁbe surface by the heahnent. The other
ﬂuorinated ﬁbe's do not show any discernable surface topography difference form the
unheated ﬁbers.

Therefore, ﬂuorination of Kevlar-49 aramid ﬁbers can increase the level of ﬁbe-
mahixadhesion, butthearamid-epoxyinterfacialshearshengthisnotaffectedbythis

 

 

 

 

 

 

 

 

enhanced ﬁber-mahix adhesion.

3.5 ~ 30
3-0 ‘ - 25?
A 3
a? 2.5 - 5
8 . r 20 gr
s q. 0
~ 2.0 - A b
on em (I)
E, ~ % ﬁx? V % 15 3
at .5 . 2
.2 . tn
o ”“5
. l’ U
+9 _ .g
. r 5 B
0.5 r O Tensnle Strength 5

r O Interfacial Shear Strength
0.0 0
Untreated A C D B

Figure6.3- Tensileshelghnarndintefacialslnearshengdn(fragmentaﬁonm)ofthe
ﬂuorinated and unheated Kevlar-49 aramid ﬁbes.

103

 

it.
t

Figure 6.4 - TEM micrographs of a radially sectioned ﬂuorinated Kevlar-49 aramid ﬁber
(heahnent D). (A) bar = 500 nm (B) bar = 100 nm

 

Figure 6.5 - SEM nnicrographs of (A) unheated and (B) heahnent B ﬂuorinated
Kevlar-49 ﬁbers. bar = 1 am

105

63.2 W

Figum6.6showmemhhvemhosoftensileshengthmdinterfacialshearshengm
(fragmentation test) of Kevlar-49, Technora, arnd PBO ﬁbers. All of the sulfonated ﬁbers
show some tensile shength reduction. Sulfonated Technora and PBO ﬁbers show slight
interfacial shear shengtln increases, whereas, Kevlar-49 ﬁbers ISS are unchanged.
Figure 6.7 shows TEM micrographs of radially sectioned, sulfonated Kevlar-49 ﬁber.
Themicrographs showtlnatasurfacelayeroftlneﬁberisadheringtotheepoxyand
cohesive separation is taking place within the ﬁber. Similarly, other sulfonated high
performance polymer ﬁbe's show that the sulfonated ﬁber exterior adhering to the mahix
andthelocusoffailureshiftingtowardtheﬁbeinterior. InChapter3and5, itwas

 

'an

[:I Tensile Strength
- s\\\\‘ Interfacial Shear Strength

 

 

 

 

 

Treated/Untreated Ratio

 

 

 

 

 

 

 

 

 

 

 

 

 

.0
o

 

Kevlar—49 Technora PB

Figun6.6-Rdaﬁvemhosoftensileshmgthmndmtefacialshershengmofmﬂmned
ﬁbe'scompared totlneirunheated values.

106

 

Figure 6.7 - TEM nnicrographs of a radially sectioned, sulfonated Kevlar-49 aramid
ﬁber showing extensive ﬁber cohesive failure.

107
demonshatcd that the adhesive properties of high performance polymer ﬁbes are

ultimately linnitcd by the ﬁber cohesive ﬁbrillation, however, epoxy adhesion of
Technora and PBO ﬁbers are furtlner linnited by other surface linnitations. Both Technora
and PBO ﬁbers have less that optimum surface energy compatibility with liquid epoxy.
Furthermore, PBO ﬁbers have a weak surface boundary layer that separates witlnin the
ﬁber during intefacial shear failure. The ISS increases of sulfonated Technora and PBO
ﬁbers are probably due to reduction of their respective surface linnitations which enhance
the ﬁber-mahix load hansfer efﬁciency.

TEM analyses of sulfonated high performance ﬁbers also suggested that sulfur
penchation is limited to only a few tlnousand angshoms into the ﬁber shucture.
Sulfonation of polycarbonate sheets is examined next to help to develop an undestanding

of these diffusion-limited polymer heahnents.

6.3.3 WW
Table 6.2 shows the XPS atomic composition of the sulfonated polycarbonate

samples. Some samples exhibit hace amounts of iron which are probably deposited
during the manufacturing of the sheets. The amount of sulfur on the polycarbonate
samples decreases after water rinse, AgNO,, or NH.OH neuhalization steps. In
particular, the NILOH neuhalization reduces the sulfur content to hace amounts. The
sulfur content reduction during the NH.OH neuhalization may be due to dissolution of
the sulfonated species (NI-LOH is a solvent for the polycarbonate), or the basic attack of
the sulfonated functionalities.

108
AB analysis of polycarbonate samples for both solution and gas phase sulfonation

heahnents were conducted. Figure 6.8A shows a low magniﬁcation view of an analysis
surface. Thesurfaceiscutthroughtlnetlnicknessoftlneheatcd sheetandthesulfonated
regions are at the left and right edges of the analysis surface. Figure 6.8B shows the
line-scan spechum for a l-hour gas phase heated and NFLOH neuhalized polycarbonate
sample. The line-scan is supe'imposed on the SEM image of the sample. The thicknness
ofthe sulfonated region appears to be about 1.5 microns which is sinnilar to the values
obtained for the l-hour solution phase sulfonation. The SEM image slnows a bright
shucturally modiﬁed region about 1 micron tlnick which is tlninner than the sulfonated

Table 6.2 - XPS atomic composition of polycarbonate surfaces.

 

Unheated
As Received
Freon washed

 

Freon rinsed
Solution rxn‘

 

Water rinsed
Solution rxn‘

AgNO, rinsed ‘
Solution rxn2

NFLOH rinsed
Solution rxn3
Gaseous rxn‘

 

 

 

 

 

 

 

 

‘ combination of 20, 45 hours, and 3, 7 days data.

3 20 hours solution sulfonnation followed by 2 hours in 1N AgNO, solution, and
1 hour water rinse.

’combinationofthree30minutesandthree2hoursdata.

‘combinationofthree lOminutesandthree l hourdata.

@QQ’ rim: '

W

 

Figure 6.8 - AFS analysis of sulfonated polycarbonate. (A) a low magrniﬁcation view
of an analysis surface. (B) line-scan for a 1-hour gas phase sulfonated sample.

110
region indicated by the line-scan. This bright region was observed for all other NH.OH

neuhaliaed samples; theFreonandwaterrinsedsamplesdidnotshow such shucturally

Figure 6.9 shows sulfur penehation depths for solution phase sulfornated
polycarbonate. The ﬁgure covers tinne spans from 10 minutes to a week of heahnent
timesforthethreesulfonationtempeahlresexamined. Penehationdepthsareplotted
versus square root of reaction time to assess concenhation dependency of sulfonation
penehation diffusion coefﬁcient (see Appendix D). The chilled (~17°C) sulfornated

samplesshowacmstantslopefmdneexaminedmﬂfonaﬁonhmeswhichmggeﬂsa

 

 

 

 

9

A

158 -

3

£7 i

36 -

O

.35 g

24-

g .

333’

1.2 _ I Warm

3

t f, 0 Ambient

51v 1 Chilled

O l l a r l a 1 n4 n l l 1 l a a J l l l a l L

o 5 no 15 20 25

Treatment Time (hour°'5)
Figure6.9-Sulfurpenehationdepthsforthechilled(~ -17°C),ambient(~22°C)and
warm ( ~ 37°C) solution phase sulfonated polycarbonate, plotted veses square root of
time (detemined by ABS line-scans).

111

diffusion coefﬁcient that is independent of the extent of concentration variation (see
equation D.8). For the ambient (22°C) and warm (37°C) sulfonation, however, the
proﬁles show that after about 3.5 pm of sulfonation penetration the diffusion coefﬁcient
becomes highly concentration dependent. These observations suggest that for the ambient
and warm sulfonation of polycarbonate, the sulfur penetration into a polycarbonate ﬁlm
islimitedbytheformationofabarricrlayerofsulfonatedmaterialthatreducethe
diffusion coefﬁcient and thus penetration depth of the sulfonation treatment.

Figure 6.10 shows SEM micrographs of untreated and sulfonated polycarbonate
surfaces. The untreated polycarbonate sample is essentially flat and featureless (Figure
6.10A). The l-hour gas phase treated and NEOH neutralized sample exhibits an etched
surface with many pitted and ﬂaked regions (Figure 6.1013). The 2-hour solution treated
and NI-LOH neutralized polycarbonate appears to have pock marked futures and many
specks of 200 nm particles (Figure 6.100. These SEM micrographs suggests the
soludonﬁeaﬁnmtscanﬂushmteﬁalsfrommesurfaceproducingamomsmoom

topography than the gas phase sulfonation treatments.

112

 

hs of sulfonated and untreated polycarbonate surfaces. (A)

rmcrograp

Figure 6.10 - SEM

Untreated sample, (B) l-hour gas phase sulfonated sample. (C) 2-hour solution phase

treated samples.

113
63.4 Eolyethxlenammnation
In Chapter 5, plasma and corona treatments of Spectra-1000 ultra-high-molecular-

weight polyethylene ﬁbers showed about 350% increase in the interfacial shear shength
(ISS) of the heated ﬁbers. ISS increases of the plasma and corona heated Specha-lOOO
ﬁbers were mainly athibuted to a combination of the surface etching that could inhoduce
mechanical interlocking adhesion mechanisms and improved wetting properties of the
ﬁbers with the epoxy resin. In this section, ambient sulfonation heahnents of unheated,
corona, and plasma heated Specha-lOOO polyethylene ﬁbers are examined to assess the
effects further chemical alterations of these ﬁbers.

Figures 6.11, 6.12, and 6.13 show XPS atomic compositions of the sulfonated
Specha- 1000 polyethylene ﬁbers. The sulfur and oxygen contents increases are due to
SO, addition and nihogen is the result of the NI-LOH neutralization step. Note that
ammicwmposiﬁmofplasmaprehemedﬁbersreachesihplateaumrermandwwrma
preheatedﬁberswhichinmmplateaussoonerthantheunpreheated ﬁbers. Thishend
is noticed more clearly in the plot of sulfur atomic% shown in Figure 6.14. The initial
(5 min. sulfonation) increase in sulfur contents of plasma-pretreated, corona-preheated,
and unpretreated sulfonated ﬁbers are 5.10, 2.81, and 1.15% respectively. All
sulfonated ﬁbersreacha ~8% sulfurcontentafterabout l hourofsulfonation.

The sulfonation rate dependency of the polyethylene ﬁbers on the surface
pretreatments suggests that the etching or conditioning of the polyethylene surface makes
them more accessible for the sulfonation reactions to occur. Spectra-1000 ﬁbers are
made from ultra-high-molecular-weight polyethylene and are highly crystalline. It is

114

 

 

 

 

 

 

 

 

 

1 00
0 Carbon
CI Oxygen
A Sulfur
80 ................................................... , ....... <>. ..... N iirbgne.“ . .
N 60 .. . . . ........... . ............................................................ -
.2
E V _9
< 40 _. ............................................... H ..................... ;EI ...
20 ............................. . ....... . ..................................... ..
—$ 3 '
O a 1 a . J a a 1 a r r a a r r a 1
O 30 60 90 120 150 1 80

Treatment Time (min.)

Figure 6.11 - XPS atomic composition of sulfonated Spectra-1000 (without preheatrnent)
polyethylene ﬁbers.

 

 

 

 

 

 

 

 

 

 

1 OO
0 Carbon
D Oxygen
........................................................ A.....$H'f9r.....
<> Nitrogen
N ...................................................................... _
.2
E o a
< .......................................................................... ..
{J
+ 9 ‘
O 4 J a J l a a 1 a a l a a L a a l
O 30 60 90 1 20 1 50 180

Treatment Time (min.)

ﬁgure 6.12 - XPS atomic composition of sulfonated Specha-IOOO (Corona preheated)
spectra ﬁbers.

115

 

 

 

 

 

 

 

 

 

 

1OO ’
0 Carbon
Cl Oxygen i
A Sulfur
80 ......................................................... O. ..... N ;t.r.o.g..e.n. . .
.t
N 60 ...................................................................... ..
.2
< 40 ...................... {3 .................. . ........ .,
*EI
20 ....................................................................... .1
O a a a a l a a l a # l a a l a a l
O 30 60 90 1 20 1 50 1 80

Treatment Time (min.)

Figure 6.13 - XPS atomic composition of sulfonated Specha-lOOO (Plasma preheated)
polyethylene ﬁbers.

 

 

 

 

  
 

 

 

 

1 O

N

.2

E .................................................................. .

.9

<

L

3 .......................................................................

('15) Corona Pretreoted

Untreated <

0 30 60 90 1 20 150 1 80

Treatment Time (min.)

Figure 6.14 - Sulfur atomic% of sulfonated un-preheated, corona preheated, and plasma
preheated Spectra-1000 polyethylene ﬁbers.

116
demonshatcd (next) that lower molecular weight polyethylene exhibits much higher

sulfonation penehation rate than the higher molecular weight polymer. Therefore, any
mechanism such as plasma etching that can reduce the molecular weight or crystallinity
of the polyethylene polymer should increase their sulfonation rate.

ABS examinations of sulfonated polyethylene ﬁbers suggest that the molecular weight
or crystallinity of the polyethylene polymer affects their sulfonation rate. APS linescan
of sulfonation penetration depth for the unpreheated Spectra-1000 ﬁbers measures only
0.9 :l: 0.2 after 1-hour, and 1.2 d; 0.2 pm after 3-hour sulfonation. However,
sulfonation of high density polyethylene produced over 15 pm of sulfur penehation after
only 10 to 15 minutes of 80, exposure (see Appendix E). Similar observations have
been reported by Holden er al. (1985). Their study of gas barrier properties of highly
oriented polyethylene ﬁlms has shown that the solubility of gases are proportional to the
amorphous volume fraction of the ﬁlms. They showed that increasing the crystallinity
of the ﬁlms signiﬁcantly reduces the diffusion coefﬁcient of the gases. This effect is
particularly marked for the larger gas molecules.

The previous observations on molecular-weight effects on polyethylene sulfonation
penehation also tend to negate the possibility of presence of a low molecular weight layer
on the ﬁber surfaces. If a low molecular-weight layer is present on the unpreheated
ﬁbers then they should exhibit faster surface sulfonation than the ”cleaned” plasma and
corona preheated samples; this is not observed experimentally. In addition, the
polyethylene ﬁber had undergone soxhlet extractions in ethanol for 24 hours and ﬁbers
were sulfonated in a Freon solvent. Both ethanol and Freon could remove such weakly

117
bonded low molecular weight surface polymers.

Figure 6.15 shows tensile property changes of sulfonated Spectra-1000 polyethylene
ﬁbers compared to their initial preheated values. The sulfonated ﬁbers show much
greater tensile shength reductions than modulus reductions. Introduction of new surface
defects by the sulfonation heahnents can signiﬁcantly affect tensile strength of the ﬁbers
since tensile shengths are defect conholled. Whereas, tensile modulus is determined by
bulk properties, and its reduction can be accounted by the reduced effective cross—
sectional area of the ﬁber as the result of the sulfonation (~l um sulfur penehation).

Figure 6.16 compares the surface energy of the sulfonated Spectra-1000 ﬁbers with
epoxy resin. For all ﬁbers, the polar components of surface energy increases as the
sulfonation heahnent times are increased. Again the plasma preheated ﬁber shows a

fasternteofpohrcomponmtincreasesmanmecommpreheatedandunprehcawd

 

 

 

 
 

 

 

 

  

 

 

 

1 O .....................................................................
\n ' Tensile Modulus
3..
9 _ . . ...................................................................
a:
a ’
306 .. .................... = ...................................................
e Tensile Strength
a u
g 0.4 _. .......................................................................
(I)
C
.S.’
0 2 . . . .9 ..... Unprstrsstss! ...................................................
' E1 Corona pretreated
A Plasma pretreated
0.0 L L I L 4 l A L l a a l 1 J 1 a L 1
0 30 60 90 1 20 150 180

Sulfonation Time (min.)

Figure 6.15 - Tensile shength and modulus changes of sulfonated ﬁbers relative to their
unheated values.

118
ﬁbers.

Figure 6.17 plots ﬁber surface sulfur concernhation versus the polar component of
surfaceenergy, showing increasing polar surfaceenergyasSO, functional groups become
incorporatedintotheheatedsurfaces. Notethattheplasmapreheatedﬁbersexhibit
higher polar surface energy component at any given sulfur concenhation than the un-
preheatedorthecoronapreheated ﬁbers. InChapterSitwasdemonshatedthatthe
plasma heated polyethylene ﬁbers have a polar component of surface energy that is
absent in the unheated and corona heated ﬁber (Figure 5.8). Evidently, the initial polar
functionalities of the plasma heated ﬁbers are retained after the sulfonation treatrnernt
resulting in their increased polar componernt of surface ernergy.

Figure 6.18 shows SEM micrographs of epoxy droplets on unheated, plasma
preheated/5 minutes sulfonated, and plasma preheated/3 hours sulfornated, polyethylene
ﬁbers. Themeniscusoftheepoxyondneseﬁbersshowareducingcontactangleas
sulfonation time increases indicating in improved ﬁber-mahix wetting.

Figure 6.19 shows the percent increase in the interfacial shear sherngth (ISS) oftlne
sulfonate Specha-IOOO polyethylerne ﬁbers relative to their preheated and unpreheated
samples. Sulfonated ﬁbers exhibit up to 450% ISS increases which are much higher than
those obtained by corona or plasma heatments alone (~300%). In Chapter 7, it is
demonshatcd that ion implantation of Specha-lOOO polyethylerne ﬁbers can cross-link the
ﬁbersurfaceshucmreandincreaseitsISSvaluesbyabout350% beforetlnelocusof
shear failure shifts completely into the ﬁber interior. Therefore, the ISS increases

beyond 350% arelikelyduetomechanisrrnsotherthan ﬁberproperty modiﬁcation. For

 

119

 

 

 

  
 
 
 

 

 

  

 

 

 

 

 

  
   

 

 

     

 

 

 

 

 

 

 

C: Un-pretreoted

 

 

 

50

 

 

 

AEo\oc.€v x905 ooot:m1..o.oa

.n
one”one“.nonsuoueuonenﬁenoHeneu. neuoueuonenouowu. 0 . m
W\\ 8
\\\\\\\\\5 \\\\\\\\\t 1 m my A 4 9
:M V. AA A
was???Honououououououououou nonououououonouou. O 1W... n 8
\\\\\\\\\\\\\r V\\\\\\\. 2 m A D
1
m o D r
1
) . r 7
w .m W. O
( S A r 6
....... e f
.0....euo.ouo.ouo.9no.euou9nou menonﬂououonoHou. 0
\\\\\\\\a S\\\\\\\t m m. B m D L
T
m mm A .4 5
smouonononouonouououom ﬁ&&&&&&&&. .1.... J
\\\S ‘\\\\\\\\\\e O m m
d d A 4| .0 d r 4
m m o e m
0 a H e Mus o
e m U .V d e e
h t Housman. Honououonououououononn. S .m r r n
m m . \\\\\\\\\\\\h 5 o .m... .m r 3
P P e m m H or. D
O O «m t
m m .m m o o
r S . toteteetetotototetetetor. D. n m
POU % DDDDDDDDDDDD m p — m S l 2
n o .m
v .m .m U C P .
“ m w. n 1
x. . O D A
— n p 6
0 O O 0 O 0 0 l
m 3 2 .l 0 1 2 3 4 60 . p . p L . p \u‘ 0
¢to\o .53 O O O O 0
BBQ m>_m..mam_o m m 4 3 2

Sulfur Atomic7.

Figure 6.17 - Plot of sulfur atomic% for un-preheated, corona preheated, and plasma
preheated sulfonated Specha-1000 polyethylene ﬁbers versus their polar components of

surfaceenergy.

Figure 6.18 - SEM micrographs of epo y droplets on (A) unheated, (B) plasma
pretreated/5 minutes sulfonated, and (C) plasma preheated/3 hours sulfonated, Specha-
1000 polyethylene ﬁbers.

 

121
the > 30 minutes sulfonated ﬁbers the sulfur composition is about 8 atomic% . These

highly sulfonated surfaces could alter the chemishy of the ﬁber-mahix interphase
resulting in a brittle and increased modulus ﬁber-matrix interphase. Rao et al. (1991 '
) have shown that increasing modulus of a ﬁber-mahix interphase increases the efﬁciency
of their interfacial load hansfer, hence increases ﬁber-mahix ISS value. The
introduction of a brittle (increased modulus) interphase by the sulfonation heahnent is
eviderntinthefracturingprocessofthesulfonated ﬁbers. Figure6.20showsashear
failed sulfonated PBO ﬁber. The sample exhibits radial mahix cracks around the heated
ﬁbers which were absent in the unheated ﬁber samples. Other sulfonated high
performance polymer ﬁbers exhibit similar mahix cracks. These observations suggest
that the sulfonation heatments of the Specha-lOOO ﬁbers can enhance their ISS by
producing an increased modulus ﬁber-matrix interphase, thus, increasing the interfacial
load hansfer efﬁciency.

122

 

U"
0
O

.p.
O
O

 

300

 

 

Interfacial Shear Strength Increase (7.)

 

 

200 P O Unpretreated ‘
I Corona pretreated ‘
A Plasma pretreated
100 l r a l r 1 1 r r l r L 4 a r I
O 30 6O 90 120 150 180

Sulfonation Treatment Time (min.)

Figure 6.19 - Interfacial shear sherngth increase of sulfonated Spectra-1000 polyethylene
ﬁbers compared to unheated value (measured by the droplet test).

 

 

 

 

 

Figun620-DigiﬁzedopﬁcalmicmgmphofandfmatedPBOﬁbuundershearsheu.

123
6-4 CONCLUSIONS

0 BothﬂuorinaﬁonandsndfonaﬁonheahnentsofKevlar-49ararnidﬁberscordd
increase the ﬁber-mahix adhesion as evidenced by the shift in the locus of ﬁber-
matrix separations from the interface into the ﬁber interior, however, aramid-epoxy
interfacial shear shength (ISS) is not affected by these heatments. Conversely,
sulfonated Technora and PBO ﬁbers exhibit increased 188 values.

0 Sulfonation heahnents of polycarbonate ﬁlms demonstrated that the diffusion of
sulfur species into the polycarbonate ﬁlm is limited by a formation of a barrier layer
of sulfonated materials that reduce the pernehation depth of sulfornationn heatrrnents.

O Sulfonation heatments of Spectra-1000 polyethylene ﬁbers show that the
preheaunantofﬁberswimaphsmamcormaheahnantcouldnotablyinaeasedndr
sulfonation rate. Surface polarity of Spectra-1000 ﬁbers was also increased as the
result of 80, surface incorporation. Sulfonated Specha-lOOO ﬁbers showed ISS
increases higher than those obtained by corona and plasma heatments alone. The
ISSincreasesofthesulfonatedﬁbersareattributedtoenhancementof
polyethylene-epoxy wetting compatibility and inhoduction of an increased modulus
ﬁber-matrix interphase.

124
Resulhofthissnndyshowdmtchemicalsurfaceheamnenhofhighperfonmnce

polynner ﬁbers can inhoduce functional groups which can improve chemical and
mechanical properties of the ﬁber surfaces. Therefore, interfacial shear shengtln of ﬁbers
such as Technora, PBO and Specha-lOOO that exhibit surface limited adhesive properties
can be enhanced by these chemical heatments. Chennical heatments can also alter
chemicalandmechmﬁcalproperﬁesoftheﬁba-mahixmtaphasemhwreasehslmd
carrying capacity, thus increasing the interfacial shear shength oftlne high performance

polymer ﬁbers.

CHAPTER 7

 

Ion Implantation of High Performance

Palymer Fibers

 

In this chapter, effects of ion implantation on mechanical and chemical properties of
the aramid and polyethylerne ﬁbers are investigated. Unlike the heahnents discussed in
the previous chapters, ion implantation is a nonequilibrium technique that permits
modiﬁcation of polymer chemishy and morphology without dependency on conditions
such as diffusivity or reaction rate. Therefore, by simply conholling the ion beam
parameters, theexterntanddepthofpolymermodiﬁcatiorncanbepreselected. Through
theionimplantationapproach, eﬁectsofboththeﬁber—mahixinterphaseandﬁberbulk
properﬁesmadhesivepmperﬁesofhighperfomancepolymerﬁbasueexamhwd.

This study of ion implantation of polymers has been a collaborative research among
several investigators and portions of these results have been published elsewhere: Ozzello
end. 1989, Kalantar etal. 1989, Kalantar eral. 1990‘, and Grummon etal. 1991.

125

126
7.1 W

Ion irnplantatiorn is a process by which a beam of high-velocity ions are injected into
thenearsurfaceregionofasolid. Typically, tlneionbeamisgeneratedbyanarc
chamber that is placed inside a high voltage electric ﬁeld. The are chamber generates
a positive ion plasma which is tlnen exhacted by the elechic ﬁeld into a positive ion
beam. Thebeamofposidveionsthusformedaredirectedtoamass-separaﬁngmgne
tlnatselects onlyoneion specie. Theresultingisotopicallypureionbeamisthenfocused
andacceleatedusingothearmysofmagnehanddechodesandacquhesteu-to-
hundreds of kilovolts in kinetic energy before striking the target. The whole ion
implantation process is conducted inside a vacuum chamber.

Ion beam implantation technology has been employed by elechonic industries
extensively, particularly for the doping of the semiconductors. Other applications of ion
beam implantation include enhancing fatigue and corrosion resistance, producing novel
mate-ials, and precision machining. Implantation to enhance polymer-polymer adhesion,
however, has not been extensively reported in the literature.

Anincidentionlosesihinidalldneﬁcenegythmughinteacﬁonswithnudeand
elechonsofthetargetmate-ials. Theenergylossoccursthroughelasticscatte'ingbythe
target atomic nuclei or by excitation of its elechons. The excited arnd recoiled target
atoms and electrons transfer and dissipate their energy by inte'actions witln other target
materials. Theincidention eventually distributes its initialkineticenergy inacascade
ofammcrecoﬂsanddechonexcimﬁommatslowanduopthehncidentionwimmme

targetorrecoilitofftlnetarget.

127
There are two types of ion-target interactions, atomic processes and elechonic

processes. Theatomicprocessesoftheionbeamareresponsibleformostofthe
shuctural changes of the target material, while, the elechonic processes mainly affect the
chemicalstateoftlnetargetmaterials. Theenergyexpendedinbothtypesofion-target
interactions can lead to signiﬁcant physical and chemical changes in target material arnd
thespaﬁaldishibutionofeachinteacdondee'nunestheextentofdnepmpety
modiﬁcation. For most materials, the combinations of mass-energy interrelations makes
the effects of different atomic and elechonic interactions hard to distinguish.

During the ion-target interactions the ions can become deﬂected from their original
direction of motion resulting in different depths of implantation. Some of the target
shuctural properties that can complicate the modeling of the ion implantation distribution
arethecrystallinityofthetargetatomsandthetargetphysical surface. Incrystallirne
targetsthepropertiesofthetargetare shornglydirectional, whichcanresultinacomplex
phenomenon of correlated collisions called “channeling". The target physical surface
also complicates the ion-target interactions because of the possibility of ion deﬂection.
Geneally, thecaseofamorphoustargeswheemostoftheionspenehatethetargetare
the simplest cases for analytical solutions of ion implantation distribution (Brice 1975,
Ziegler 1985). Othe factors affecting the ion implantation distribution are the ion
species, ion energy, ion ﬂux (dose), and temperature. The spatial distribution of
iorn-target inteactions is generally called "range dishibution'. Thee are several reviews
on the evaluation of range distributions of ion implantation (Kumakhov er al. 1981,
Mayer et al. 1970, Wilson er al. 1973, Carter er al. 1968).

128
Ion beam implantation can alter the chemishy, surface energy, arnd morphology of

a polymer surface which could potentially improve its adhesive properties. Ion-target
elechonic interactions can produce sites for covalent chemical bonding between the ﬁber
and matrix molecules or cross-link the ﬁber surface polymes and hinder the development
of a cohesive weak boundary layer. Elechonic interactions can also increase the surface
free elegy of ﬁber and enhance its wetting prope‘ties. Ion-target atomic inteactions
can improve the ﬁber adhesive properties by etching the ﬁber surface to eitlner increase
its surface area or remove the ﬁber surface weak boundary layer; however, exeessive
atomic interactions may weaken the ﬁber surface. Work by Dresselhaus er al. has shown
that the ion implantation tends to cross-link the polymes with unbranched chains, while,
polymeswith complex sidechainstendtodegrade. Theirworkalsoslnowsthation
implantation tends to more selectively remove non-carbon atoms from the polymes
similartoapyrolysisprocesses. Ingeneal, toimprovepolynneradlnesiontlnemain
thmstofionimphntaﬁonshouldbewmaximizedwdecheucinteracﬁonsdlatcan
impmvethewetﬁngandchemicalbondingwimmmeﬁbesurfaceandmminimizedw
atomic inteactions that may degrade the polyme in the implanted region.
Puglisi(l989)hasreviewedsomeoftheeffectsoftheionbeamonthepolymer
subshate. For polymer materials, ion doses equal or greater than 10" ions/cm2 produce
amorphouscarbonizedmateialswithonlyparﬁalmemoryofmeoﬁginalpolyme
composition. At these moderate to high doses the nuclear inteactions produce random
atomic displacements, while, elechonic inteactions drive to maintain chemical stability.
At doses lower than 10" ions/cm2 botln nuclear and elechonic inteactions can produce

129
scission (bond-breaking) and aggregation (bond-forming) reactions. The scission
reactions give products such as H,, C211,, etc., and aggregation reactions produce
products with higher molecular weight than the target polymer. The extent of scission
andaggregationreactionsdependsonthechenishyofthetarget. Forexample, for
polymethylmetlnacrylate the scission reactions dominate, but for polystyrene the
aggregation reactions prevail (Puglisi 1989).

Previous work on polymer surface modiﬁcation by ion implantation has been shown
to increase the adhesion of polyethylene (PE) to metallic titanium ﬁlms (Bodo er al.
1986) and to increase UHMW-PFJepoxy interfacial shear strength (Ozzello at al. 1989).
Licciardello er al. (1987) have reported that ion implantation of polystyrene produces
cross-linked shuctures that are chemically different from that obtained by electron
bombardnnents. Adem er al. (1988) have shown that oxygen implantation into
polyvinyledene fluoride results in hydrogen and fluorine rearrangement on the polyme'ic
chain. Ishitani et al. (1989) have reported on the applications of SIMS, FITR, Raman
and ESR to study oxygen implanted polyethylene. They conclude that ion implantation
of the polymer inhoduces complicated chemical reactions which ultimately result in the
geneation ofamorphouscarbon. Yoshidaeral. (1987) haveexaminedtlnestructureand
morphology of ion-implanted polyimide ﬁlms and reports that carbonimtion of the
implanted polyme laye' as a result of the polymer scission and loss of oxygen arnd
nihogen atoms. Bertrand er al. (1987) have examined helium implantation of
polyethylene terephalate polymes which show bornd breaking and surface damage as the
result of implantation. Suzuki et al. (1988) have reported that ion implantation brakes

130
up the silicone rubber polymer to form new radicals which can signiﬁcantly enhance the

polymer wettability. Similarly, Torrisi er al. (1988) have demonshatcd the degradation
of the polytetraﬂuoroethylene polymer as the result of helium ion implantation. Ion
implantation of polymer subshates is a new area of research which is actively urnder
investigation, however, applications of ion implantation to enhance adhesive properties
of polymes has not been investigated extensively. In this study, the effects of ion
implantation on adhesive properties of Kevlar-49 arannid and Spectra-1000 polyethylene

ﬁbersareexamined.

131
7.2 WAT.
Two epoxy systems we'e used in this study, the DER331/MPDA/DETDA 175°C/3hr

(fragmentation test) and the DER331/MPDA RT/24hr/75 °C/2hr/100°C/3hr (droplet test)
systems. Single ﬁber tensile measurements were done by ASTM D3379 tensile test.
Fiber surface elemental compositions we'e characterized by XPS. Fiber-matrix interfacial
morphologieswereexaminedbyTEM. Theseexperimentalproceduresaredetailedin
Chapter 2-

Both Kevlar-49 aramid and Specha-IOOO were soxhlet exhacted with ethanol before
they were sent for the implantation heatnnents. The ion implantation were dorne by Spire
Corporation (Bedford, MA) according to the protocol slnown in Table 7.1. Ion beam
currents were between 0.2 to 5 uA. Fibes were implanted by laying individual ﬁbers
ona75 mm squarealuminumframewhich held themincontactwithachilledaluminum
hcatsink. IaterXPSstudiesshowednoevidenceofre-sputteedaluminumonanyof
the irradiated samples. For each implantation condition, several hundred lengths of
ﬁbe's and several strips offabric (10x75 mm’) m irradiated. After irradiation, the
ﬁbes wee left on their frames and sealed in nitrogen-purged bags to prevent absorption
of atnnosphe'ic moisture. Most of the implanted ﬁbers were embedded in epoxy within
four days after irradiation.

132

Table 7.1. Ion implantation protocols for polyaramid and polyethylene ﬁbes.

Energy Dose Implant Depth
(keV) (ions/cm?) (nm :1; 10% var.)

Kevlar-49 Polyararnid

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Spectra-1000 Polyethylene

 

 

 

 

 

 

133
7.3 W

Results of arannid and polyethylene ion implantation are discussed separately to
provide an exclusive discussion for each ﬁber type. Conclusions for botln ﬁbers are then

combined to develop a comprehensive understanding of the effects of ion implantation

on adhesive properties of high performance polymer ﬁbers.

7.3.1 Minimum

Figure 7.1 shows the XPS elemental composition of 400 keV N * implanted ﬁbe's
(in fabric form) for various doses. For tlnese implantation, 10" N ” lcm2 dose exhibit
about 6% increase in carbon, 15% reduction in nihogen, and 22% reduction in oxygen.

Lower dose implantation, however, show only marginal compositiorn changes. In Figure

10 r 1 1 rT]
K7
V
3
c 0 t... .....................................
O
.C
U
4.:
C A
Q)
§_10 ...........................................................................
Q
C
.9
.‘1
m
8_20 .............................................................................
E A Carbon
0 0 Oxygen
U .
Cl Nitrogen
_30 l r r r r r l r l r l l r r l
1012 2 3 4 s 2 3 4 5 1014

Figure 7.1 - XPS elemental composition changes of 400 keV N“ implanted Kevlar-49

 

 

 

 

 

 

.1013 .
Implantation Dose (tons/cm’)

aramid ﬁbe's relative to unheated ﬁber.

 

134
7.1,carbmizaﬁmoftheaMdmhfaceappeamwtakephceatdoseshighethan10”

ions/cm”, however, the aramid carbonization dose can increase for lighter and lower
elergyions (Velkatesanetal. 1987). Theinitial slightincreasesinnihogenandcarbon,
and reduction in oxygen relative to unheated ﬁbes may be attributed to sputteing oftlne
Kevlar-49 oxidized surface layer (see Table 6.1).
Eﬂ‘eesofﬁbesurfaceearbonizaﬁonuealsoapparentindleﬁbemorphologyas
observed by TEM. Generally, ﬁbers implanted with doses at or above 10“ ions/cm’
showconhasteffectsinTEMmieographswhichappearasadarkeningintheﬁbe
external region, corresponding roughlytotlneion projected range. Figure7.2 showsthis
darkened region fora 10“ INF/cm2 400 keV ﬁber embedded in an epoxy mahix. The
darkenedzoneonthisﬁberappearsonlyontheleﬁsideduetoline-of-sightshieldingby
anadjacentﬁber. Notethattlnethicknessofthecarbonizedregionisaboutlumas
predictedbytheTRM-90program(6uimaraeseral. 1989,Biersacketal. l980)shown
in Table 7.1. The implanted region shows good adhesion around the ﬁbe-matrix
intefaceexceptovemeregionwheetheﬁbewasnotimplantedmppeﬁbepeimee
in Figure 7.2A). This observation is evidence for the improvenent of aramid-epoxy
adlnesion by the ion implantation heahnent. TheXPS results forthe10"N*/cm’400
ktheahnent(Figun7.l)conﬁmdmtdnedarkregionammdﬂneﬁbepeimeeisdue
totlneﬁbercarbonization.
Conversely,atdosesatorbelow10”ions/cm’thisnearsurfacedarbningisabsent
forthe400keVN+implantedaramidﬁbes Figure7.3showsTEMrnicrographsof
a 10” I‘VIcm2 400 keV implanted ﬁber. The section show excellent ﬁber-mahix

”1‘4: 1. v "
ﬁfty ' ' .
I

O . __

 

Figure 7.2 - TEM micrographs of 400 keV 10" N’lcrrl2 irradiated Kevlar-49 ﬁber.
(A) whole ﬁber K in mahix E, (B) detailed view of the interphase.
(A)bar = lam (B)bar =200nm

 

Figure 7.3 - TEM rnicrographs of 400 keV 10" N‘ch2 irradiated Kevlar-49 ﬁber.
(A) whole ﬁber view, (B) high magniﬁcation view of the interphase. ’
(A)bar=1pm (B)bar=100nm

137
adhesion but without detectable skin darkening. This observation is consistent with the
XPS results which showed a surface composition close to the unheated ﬁbes (Figure
7.1). Figures 7.2 and 7.3 exhibit increased ﬁber-matrix adhesion compared to the
untreated (Figure 3.9) ﬁbers. Similar increased ﬁber-mahix adhesion is observed for
all other ion implanted aramid ﬁbers (discussed later).

Although, thearamidcarbonintionappearstotakeplaceatdosesatorabove 10“
ions/cm’,thecarbonizaﬁondosealsodependsmﬂneimplantaﬁonenegyenddw
implanted ion mass. Generally, the polymer carbonizing dose increases for lighter and
lower energy ions (Venkatesan er al. 1987). Figure 7.4 shows tlne TEM micrographs
of a 10“ N“/cm2 30 keV implanted ﬁber. The ﬁber does not show the near surface
darkened region observed in the 10" N‘lcm2 400 lseV implanted ﬁber (Figure 7.2).
Figure 7.4 also shows good overall aramid-epoxy adhesion, but on closer inspection of
the intephase, extensive interfacial ﬁbrillation is observed (Figure 7.43). This
interfacial ﬁbrillation is within the ﬁber which suggests cohesive ﬁber failure.

Figure 7.5 shows TEM micrographs of the 10ls T‘i‘*lcm2 100 keV implanted ﬁber.
This combinationofenergyanddoseiswellabovethecarbonizingdoseofthearamid
ﬁbesandthecarbeuzedsldnlayeiscleadyvisibleamunddnepeimeeofdwﬁbe-
matrixinterface(~150nm thick). Thecarbonizedﬁbersurfacehasadheredwelltotlne
matrixandtlnelocusoffailureisbetweentheirnplantedskinandtlneitsbulkinterior.
Thepresenceofbubbleshappedatdneﬁbe—mahixintefacewnﬁnmdmdwﬁbesldn
adherestothematrixandtheinterfacialfailurearewithintlneﬁber. Similarly,Figure
7.6 shows the TEM micrograph of a 10“ N"!cm2 400 keV implanted ﬁber, nnade after

G

 

Figure 7.4 - TEM micrographs of 30 keV 10“ N‘lcm2 irradiated Kevlar-49 ﬁber.
(A) whole ﬁber view, (B) detailed view of the interphase.
(A)bar = lam (B)bar = 100nm

 

Figure 7.5 - TEM micrographs of 100 keV 10” Ti’lcm2 irradiated Kevlar-49 ﬁber.
(A) whole ﬁber view, (B) ﬁber-matrix interphase.
(A)bar = lam (B)bar =500nm

a

"3
r.
r“, . _.' V
~j:up;.-ae’~d '-“"’"

 

Figure 7.6 - TEM micrographs of 400 keV 10“ N +Icm2 irradiated Kevlar-49 ﬁber after
failure in ISS test. (A) whole ﬁber view, (B) skin-bulk intephase.
(A)bar = lum (B)bar =200nm

141
failure in the interfacial shear shength test. Again, the implanted ﬁber exterior adheres

welltothematrixandthelocusoffailurehasshiftedtoward tlneﬁberinterior.

InChapter3, itwasdemonstratedthatfortheunheatedKevlar-49aramidﬁbestlne
ﬁbe-matrix failure mode is interfacial with some ﬁber surface cohesive ﬁbrillations
(Figure 3.9). Similar failure modes are also observed for ﬁbers implanted below the
carbonizing dose as shown in Figures 7.3 and 7.4. However, for carbonized
implantation (210“ ions/cm’) the ﬁber-matrix failure mode is more cohesive. Figures
7.5, 7.6 and other nnicrographs of carbonized aramid ﬁbers show the locus of failure to
be between the carbonized implanted skin and the ﬁber bulk interior. The difference
between the failure modes of the carbonized and the lower dose implantation may be due
to mechanical weakening of the carbonized ﬁber region. Degradation of mechanical
properties of the implanted aramids are more severe for the higher dose implantation
(discussed later). Thecarbonized skinisaweakboundarylayerthatshiﬁsthelocusof
failureintotlneﬁberinteior. Thecarbonizedﬁberskinmayalsohaveincreased
microscopic porosity that can promote its epoxy adhesion through mechanical
interlocking mecharnisms. Although, SEM obse'vations of the carbonized aramid ﬁbers
actually showed somewhat smoother surface topography than unheated ﬁbers (similar to
Figure6.5), themorphologicalchanges maybeatascale smallerthanthatdiscernable
at 20 kx magniﬁcation.

Adequate wetting of the ﬁber surface by liquid resin is a prerequisite for good
adhesion. Figure'7.7comparesthesurfaceenergyofsevealN+ implantedaramidﬁbe's
withtlneepoxyresin. TheirradiatedKevlarﬁbesshowabouthalfthepolarenergy

142
component of the untreated ﬁber along with marginal changes in their dispersive energy
compornents. Gutowski (1990) has shown that in the absence of chemical bonding,
maximumadhesionbetween themahixandﬁberoccurswhenthesurfaceenergyofthe
ﬁber (1,) and mahix (7.) are equal. Gutowski proposes theoretical relations that suggest
for (7.17.) > 1 complete wetting occurs but the shength of ﬁber-nnahix adhesion
decreases as the ratio increases. For the surface energy ratios (1.11.) < 1 incomplete
wetﬁngoccummndﬁbe—mahixshengﬂnmpidlydmpswzeoasdnemﬁoapproaches
0.7, hence the ion implanted Kevlar-49 ﬁbes are expected to have incomplete wetting
with the epoxy resin. Theefore, the enhanced ﬁber-matrix adhesion of the ion implanted

ﬁbers must be due to physical mechanisms such as mechanical inte'locking and/or

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20
b
2
o ..........................................................
CL 10
A
E
0
>0
C
>.
t
V
10 g ......................................................
G)
.2 ﬂ
0) I
B 20 ﬂ ........................................................
o. I
.a g _
he
0 3° g __ ....... Lg ...... A ................ ._. ....... . ................
I l_l
A
‘0 1 1 4L 1 l l P P
> > it it is is
U an at x x x x
2 4‘ 4‘ o o o o
a, o o o o o o
o n n .- <r ‘t' ¢
x" 3 v to v N n at
8' Oi '— v- v- v- '- v-
1* to U U U U laJ
U K P '- e— v- e— '—

Figure 7.7 - Polar arnd dispersive compornents of surface free energy for N+ implanted
Kevlar-49 ﬁbe's arnd liquid epoxy.

143
mechanical interpelhation, or introduction of active chemical sites that can form covalent
bonds with the matrix molecules.

Figure 7.8 shows the interfacial shear shength of N+ implanted aramid ﬁbe's for
various doses. There are no signiﬁcant increases in the aramid-epoxy interfacial shear
shengtln despite the apparent improved interfacial adhesion observed for all ion implanted
aramidﬁbers. Thereisaboutal6% interfacialshearshengthreductionforthe
carbonized aramid ﬁbers, that, along with the previous TEM observations, suggest
creation ofaweakinterphasebetween thecarbonized ﬁberskinandtheﬁberinterior.

The ﬁbers examined in Figure 7.8, all were impregnated immediately after their
removal from their sealed bags. For some samples, ﬁbes were kept exposed for one

montln in the atrnospheic condition. This environmental exposure caused no detectable
ISS change.

 

N
O

eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

t?-
r!
.2...
as
._._.
l—I—l
l-Bﬂ
3...
._f._.
l—H
l—-I-l

O
l

eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

I Corbonized

lnterfocial Shear Strength (MPa)
0|
I

 

 

 

E14 30keV
1E13100l<eV
1E14100keV
‘lElZ ‘OOIteV
5512 400keV
1E13 400l<eV
1E14 400keV
2E14 400keV

Control

p

Figure 7.8 - Interfacial shear shengtln of N * implanted Kevlar-49 ﬁbes.

144

Aramiddegradation by ionimplantationisalsoexhibitedinthetensilepropertiesof
the ﬁber. Figure 7.9 shows the tensile shength of the implanted arannid ﬁber along with
their expected implantation depths (Table 7.1). The implanted aramid ﬁbers that were
carbonized, were distinguishable by their brownish coloration compared to the bright
yellow color of unheated and the low dose implanted ﬁbers. The qualitatively identiﬁed
carbonized ﬁbers are shown by the solid symbols. In general, ion implantation produces
tensile strength losses that increase with the implantation depth (energy) and dose. All
ﬁbers implanted at 400 keV (~l to 1.5 um implantation depth) show a loss of tensile
shength, however, the carbonized ﬁbes exhibit the lowest tensile strengths. This
suggest that thescission reactions dominate the chemishy of ion-target inteactions in

polyaramids, as they reportedly do in the case of ion irradiated polymethylmethacrylate

 

    
  

 

 

 

3.5 3.5
A28 - ................... E ............. { WP§ ....................... .238
cf .2
o l i E i E
V V
5 2.1 n. ..................... . ..................................................... r-l 2.15
0" O.
C Q)
2 i O
(‘5 Cl Stoichiometric g
0 1.4 ........... . ...... . ........................................................ .l 1.4.;
.-: I Carbonlzed O
(D ‘ . ‘0"
5 O lmplantotlon Depth 3
l- ‘5.
0.7 .. .......................................................................... .. 0.7g
0.0 % 0.0
>
E 3 i i s 3 E E E 3 g 3 3
s: a a ﬁ § .3. .3. .3, § § ’4 § §
v c en 2 to n 1 N or n 2 «e v
'9 a: a: a: g is a: a a a an ‘3 .5 an
‘é:“.’:+:‘2:‘21’?+:‘1
8 z z z 2 l: 2 z z z 2 1'? z 2
Figure 7.9 - Tensile shengtln and implantation depth ofirradiated aramid ﬁbe's.

145
(Vankatesan er al. 1987). The tensile shength reductions in the low energy implantation
are less evident because the low penehation implantation only modify a small volume of
ﬁber m.

The combirnation of enhanced ﬁber-mahix adhesion and reduced mechanical
propertiesoftheionimplantedaramidﬁbersisalsoexhibitedintlnefracturingprocess
of their impregnated samples. Figure 7.10 shows the optical nnicrographs of fracturing
process of a 10“ Iii/cm2 400 keV sample impregrnated in an epoxy system. The ion
implantedarmnidﬁbepmducesmahixfmcmresdlatdonotoccmindneunheated'
ararnnidﬁbers. Thistypeofmahixfrachnreindicatesthepresenceofabrittleﬁber-mahix
interphase (see Figure 6.20). For the carbornized implantation, the brittle skin layer
probablyﬁacmresendcausescmcksdutpmpagateinwdwmahix,nmedwless,dw
ﬁbe-mahixadhesionhasmbeshongbecauseeacksamnotdeﬂeeedalongtheﬁbe-

matrix interface.

7.3.2 WW
Grummon er al. (1991“) have reported on the effects of implantation dose on

polyethylene composition. Ther XPS results show that at doses below 10" ions/cm’,
increasesinoxygelanddecreasesincarbmconteltrdaﬁvemthemheatedﬁbes
occurs. Atdosesabove10'5ionslcm’, however,theﬁbercarbonizationbeginsto
predominate (Ol, Ct). Like the aramid ﬁbes, the carbornization oftlne polyethylene
ﬁbers were accompanied by visible darbning of the ﬁbe' exteior. Implanted
polyethylene also exhibited surface morphological changes such as nnelting of the ﬁbrillar

E
3.
O
O
‘-

 

 

 

 

 

 

Flgure 7.10 - Fracturing of 10“ N“lcm2 400 keV implanted aramid ﬁber. (A) 0% strain,
initial curing kinks, (B) 2% strain, ﬁber fractures, (C) 5% strain, matrix fractures,
(D) strain released, kinks reappear.

147
tendrils that are present on the untreated ﬁbe's. These surface modiﬁcation wee

attributed to the thermal effects of the implantation process (T. = ~ 150°C).

In Chapter 3, it was shown that the slnear failure of untreated polyethylene ﬁbers
were primarily interfacial, with a few loeations of ﬁber cohesive failure. TEM
micrographs of untreated polyethylene ﬁbers demonstrated extensive ﬁber-epoxy
interfacial separation with the sporadic presence of ﬁber surface ﬁbrillation (Figure
3.14). TEM micrographs of the ion implanted polyethylene ﬁbers show a marked
difference in their interfacial morphology and shear failure behavior. Figure 7.11 shows
a 10" Ti”/cm2 100 keV polyethylene ﬁber imbedded in an epoxy system. This ﬁber has
beenimplantedabovetheearbonizingdoseandtheearbonized skinappmrsasadark
band around ﬁber perimeter. A marked improvement in the interfacial bonding is
apparent in the ion implanted ﬁber.

Other polyethylene implantation also exhibit improved ﬁber-matrix adhesion,
however, the implanted skin appears to adhere to the epoxy and the locus of failure shifts
to within the ﬁber structure. Figure 7.12 shows the TEM nnicrographs of a 10“
Ar’lcm’ 75 keV and a 10u He“/cm2 400 keV polyethylene ﬁbers. For botln of these
implantation the implanted region is discernable around the ﬁber perimeter with its
smootln texture, reduced folding, and relatively dark coloration. The smooth texture and
reduced folding appearance of the implanted region may be the result of reduced
toughness of the implanted region. The stretching of the ﬁbe during the sectioning
process produces the ﬁber folding observed in the TEM rnicrographs. Ion implantation

of polyethylene ﬁbers below the carbonizing dose tends to cross-link the ﬁber which

148

 

Figure 7.11 - TEM nnicrographs of 100 keV 10" 'l‘i’lcm2 irradiated Spectra-1000 ﬁber.
(A) whole ﬁber view, (B) detailed view of the ﬁber-matrix interphase.
(A)bar sSum (B)bar =100nm

 

Figure 7.12 - TEM micrographs of (A) 75 keV 10" Ar‘ch2 and (B) 400 keV 10”
I-Ie“/cm2 irradiated Spectra-1000 ﬁbers.

150
reduces the ﬁber toughness, thus, producing a smoother texture and fewer folding.

TheappeamnceofdneimplantedregionsintheTEMmicmgmphswasnotobserved
for the aramid ﬁbers that were irradiation below the carbonizing condition, which may
be due to the greater thermal sensitivity of the polyethylene ﬁbers (T. =- 150°C) than
the aramid ﬁbers ('1‘, = 350°C, T4,... = 500°C).

Figure 7.13 shows the TEM micrographs of a 10“ Ar‘lcm’ 75 keV and a no15
Ti+lcm2 400 keV irradiated polyethylene ﬁbers. These rnicrographs show that the ﬁber-
matrix separation is taking place between the implanted skin and the ﬁber bulk interior.
Figure 7.13 and other micrographs of ion implanted polyethylene ﬁbers show that the
ion implantation treatment shifts the locus of polyethylene-epoxy separation from an
adhesive separation for the untreated ﬁbers to a cohesive separation between the
implanted region and the ﬁber interior for the treated ﬁbers.

Surface energy measurements by Grumman er al. (1991‘) have suggested that the
improved adhesion ean be attributed to the enhanced wetting properties ofthe implanted
ﬁbe's. The polar component of surface energy wee increased sharply (ZOO-300%) to
values close to the epoxy component. The dispersive components, however, wee not
signiﬁearntly altered but remained close to the epoxy component. Therefore, the ion
implanted polyethylene ﬁbers showed much improved surface energy compatibility with
the epoxy which can signiﬁcantly enhance their thermodynamic wetting properties.

The interfacial shear strength (ISS) of the ion implanted polyethylene ﬁbers
characterizedbythedroplettechniquehasbeenreportedbyOmlloetal. (1989).
Generally, ﬁbe's implanted near or above the earbonizing dose show about 250% to

It!"

”1’; k.
new can?
t f , f
n

 

ﬁgure 7.13 - TEM micrographs of (A) 75 keV 10“ Ar"/cm2 and (B) 100 keV 10“
‘I‘i‘lcm' irradiated Spectra-1000 ﬁber.

152
300% 188 increases relative to the untreated ﬁbers. These ISS results wee not affected

bytheclnemicalnatureoftheimplantedionssuggestingthatonlythedensityof
irradiationenergycontrolstheextent of adhesion modiﬁcation. T'heimprovedISSresults
along with the TEM observations suggest that polyethylene ion implantation can
signiﬁcantly increase the ﬁber-matrix adhesion, but then the locus of failure shifts to
withintheﬁbestrucnrreandbetweentheimplantedand‘unueatedregions.

Grummonetal. (1991') havealsoidentiﬁed seveal polar functionalgroupsonthe
implanted polyethylene ﬁbe's using an ATR-FI'IR technique. Covalent chemical bonding
between these functional groups and the epoxy may be possible, however, their
contribution to the ove'all adhesion level is not clear. Hook :3 al. (1990) have
deenunedthatﬂeconuibuﬁonofmewvalentchenucalbmdingmshearbmdsuengm
betweenAS4carbonﬁberandanepoxy-aminepolymeislessthan5%. T'hepresence
ofthenewpohrgroupsondwimphntedpolyehyleneﬁbescanexplainmemeesein
thepohrwmpewntofdwsurfaceenegyandcensequenﬂyenhancedthemodymmic
wetting withtheepoxy resin, butthecontribution of covalentchemicalbondingbetween
epoxypolymeanddnesepolargrouptodneimpmvedﬁbe—maﬂixadhedonrequire
furtherevaluation.

Effects of ion implantation on the mechanical properties of the polyethylene ﬁber can
beobservedintheﬁbetensileprope'ties. Table7.200mparesthemechanical
properties of the untreated and 400 keV 10" He‘lcm’ implanted Spectra—1000 ﬁbe.
Datashowsareducﬁoninthetoughnesofuneimphnwdﬁbeoensilesuengthand
fracturestrain)butthetensilemodulusisincreased. Similartensilepropetieeof

Table 7.2 - Mechanical Properties of untreated and 400 keV 10l3 I-Ie°/cm2 implanted

Spectra-1000 polyethylene ﬁbers.

Implantation Depth (pm :1; 10%
var.)

1.6

 

No. of Tests

12

12

 

Diameter (pm)

28.7 :1: 2.7

28.7 :I: 2.5

 

Tensile Modulus (GPa)

67.8 :I: 9.2

82.5 :I: 3.0

 

Tensile Strength (GPa)

2.65 :1: 0.29

1.77 :1: 0.26

 

Fracture Strain (76)

6.91 :1: 1.30

3.42 :l: 0.42

 

ISS droplet test (MPa)

2.95 :t 0.17

10.3 :1: 0.5

 

 

 

implanted polyethylene ﬁbers with implanted ranges less than 0.15 um (Ozrello er al.
1989) did not show signiﬁcant tensile property changes probably because only a marginal
volume of the ﬁber structure was modiﬁed. For polyethylene, aggregation reactions are
known to dominate the low dose implantation (Puglisi 1989) that can result in cross-
linking of the polyethylene structure. The increased inteactions among adjacent polyme
chainsinthecross-ﬁnkedpolymetendtoreducednerrelaﬁvemobilitywhich reduces
toughnessbutincreases modulusoftlnestructure. Tlnereduced toughnessandincreased
modulus of the implanted polyethylene ﬁber shown in the Table 7.2, suggests that the
implantation prosess produces a highe modulus but nnore brittle polyethylene structnnre
for irradiations below the carbonization condition.

154
Otlne researchers have also reported modulus increases for the cross-linked

polyethylene ﬁbes. Chapiro (1988) has examined the radiation chemistry of
polyethylene and has proposed a cross-linking chenical mechanism. He reports that
cross-linking increases the modulus and hardness ofthe polyethylene and imparts a non-
melting behavior to the material. Postena et al. (1988) have reported that
chlorosulfrmation of ultra-high strength polyethylene ﬁbers decreases the ﬁbe tensile
strength but increases the ﬁber modulus by more than 50%.

The brittle nature of the ion implanted polyethylene is also evident in their fracturing
process. Figure 7.14 shows the optical micrographs of shear fractured ion implanted
polyethylene ﬁbe's that were imbedded in an epoxy matrix. In Figure 7.14A the 100
keV 10" Ti‘lcm2 implanted ﬁber shows thin surface fractures along the ﬁbe radius
whicharelesstlnanhalfaﬁberdiameterapart. InFignu'e7.14Bthe400keV10”
I-Ie”/cm2 implanted ﬁber shows thick fracture lines that are angled arnd about two ﬁber
diametesapart. Theseangledfracnrrelinesarealongthecompressivehelicalkinklines
oftheﬁberthatareproducedduringtheﬁbemanufacnnringprocess. Thethe'mal
suesesmducedduﬁngmeﬁbeimbeddingprowssmmeepoxymauixreNBingreate
displacementsalongthesekinklinesthantherestoftheﬁbe, therefore, tlneimplanted
ﬁberskinismoreapttobreakalongthesekinklines. Thecloselypackedfracturesof
thecarborﬁzedﬁbeMgure7.l4A)aremorefrequentthantheﬁbeldnkline, which
suggests that carbonizing implantation are producing a brittle structure. The 400 keV
10" IIe‘lcm2 implanted polyethylene ﬁber shows a moderately highe ISS increase
(~360%, Table 7.2) than the increases reported by Ozzello et al. (1989) for the

155
carbonizing implantation conditions (250 to 300%). For the implanted ﬁbers, mechanical

coupling between polyethylene bulk and its implanted surface may be weaker for the
brittle carbonized surfaces than the cross-linked non-carbonized surface.

The shear strength limitations of the implanted polyethylene ﬁbers wee clearly
demonstrated during tlneir droplet-pullout ISS test. Figure 7.15 shows SEM nnicrographs
of a separated draplet on a 400 kev 10” £18ch2 implanted Spectra-1000 ﬁbe. In
Figure7.153thebladesideofthedropletshowsasection oftheﬁbethathasbeen
fractured within the ﬁbe, while, on the back side of the ﬁber the separated skin has
folded (Figure 7.15C). In Figure 7.15, the ﬁber fractures are not symmetrical around
the ﬁber probably because the implanted ﬁbe was not uniformly irradiated. Non-
uniformity of the ﬁbe implantation is also evident in the ﬁbe-droplet meniscus, where
the fractured ﬁbe side shows a epoxy wetted surface while othe sides appear unwetted.
TheeSEMmicmgmphsdemonsnatethatmelocusofshearfaﬂumhasbeensluﬁed from
dneﬁbe-mauixintefacemwiﬂninﬂneﬁbewucMeandbeweendneimplantedmd

untreated regions.

 

 

 

 

 

 

 
        

“I

-‘w

253

 

 

 

Figure 7.14 - Digitized optical micrograph of sheared ion implanted Spectra-1000 ﬁber.
(A) 100 keV 10" Ti‘*/cm2 implanted, arnd (B) 400 keV 10‘3 He‘lcm’ implanted ﬁber.

 

Figure 7.15 - SEM nnicrographs of a pulled-out droplet on a 400 keV 10“ Helen2
implanted Spectra-1000 ﬁbe. (A) Back side, (3) overall view, and (C) blade side
of the droplet.

157
7.4 CQHCLHSIQNS

0 For botln arannid and polyethylene ﬁbes, surface carbonization tales place at ion
dosesnearorabovetlne 10“ionslcm’. Belowthecarbonizingdou, aggregation
reacﬁonstendtodonunatedchemicalmodiﬁcaﬁonofpolymerswithaliphaﬁc
backbonednaimsuchupolyednyleneneulﬁnginaeoss-ﬁnkedﬁbesﬂucmwdnat
has reduced toughness but highe modulus propeties. For polymes with non-
carbonelementsintheirbackbonesuchasaramids, thescission reactionstendto
domirnatethechennical modiﬁcationresultinginreducedmecharnicalpropetiesof
the implanted region.

0 For the polyethylene ﬁber ion implantation produced more epoxy compatible
urrfaceenegycomponentsdnatcouldsigniﬁcanﬂyenhancetherthemodymmic
wetting prope'ties. Convesely, ion implantation reduced the epoxy wetting
compatibility of the arannid ﬁbe's. Nonetheless, for botln ﬁbes improved ﬁbe-
matrix adhesion wee obseved.

e Ion implanted Spectra-1000 polyethyleneﬁbesexhibit250to 360% increases in tlne
ﬁber-matrix interfacial shear strengths (ISS), however, Kevlar-49 aramid ﬁbes did
not exhibit any signiﬁcant (ISS) increases. For both ﬁbe, improved interfacial
adhesimchangeddwlocusofshearfaﬂureﬁomﬁbe-manixinmhasetowithin
theﬁbesnucnrreandbeweendneimplantedregionandtheunneatedbulk.

158
Result of this study demonstrate that increasing the inteactions between ﬁbe polyme

chains increases the ﬁber tensile modulus and intefacial shear strength, but decreases the
ﬁbefracturestrainandtensilestrength. However,ifthelatealchaininteactionsare
not advanced tlnroughout the ﬁber structure, then the locus of shear failure shifts to an
untreated region of the ﬁber structure. Therefore, the key to improving the ﬁber-matrix
intefacial load transfe of high peformance polyme ﬁbes is to introduce polyme chain

inteaction throughout the ﬁber structure.

CHAPTER 8

 

Structural Modiﬁcation of

High Performance Polymer Fibers

 

Polymeueannenmdiscussedmthepmviouschapteshavebeenablemincreaseﬂw
intefacialloadnansfecapacityofmehighpefommncepolymeﬁbeswdnepoint
wheetheﬁbelateralcohesivestrengthbecomesthelimitingfactor. Although,
interfacial shear strength (ISS) increases obtained by some of tlnese treatment techniques
are signiﬁcant relative to their untreated values, these ISS improvements are still far
lowetlnantheir theoretically predicted valuesﬂ'igure3.3). Forthehighpeformance
polyme ﬁbes, previous results indicatednathigheinte‘facialshearstrengthsmaybe
obtainableifﬂwﬁbeuansvesecohesivepropeﬁecouldbeenhmcedmmughoutﬂw
ﬁbestructnnre. Inthischapte,approachestostrucnnralmodiﬁcationofhigh

performance polyme ﬁbes are investigated.

159

160
8.1115139121311th

AsdiscussedintheChaptersBand7,thehighlyalignedstr-uctureofhigh
peformance polyme ﬁbers results in their excellent tensile propeties, howeve, since
ornlyweaksecondaryforcesconnectthechainsradially,theﬁbersexhibitweakshearand
compressive properties. Therefore, increasing cross-chain interactiorns is the key to
increasing the shear propeties ofthe high performance polymer ﬁbers. In this study,
approaches to structnrral modiﬁcation of high peformance polymer ﬁbers are
investigated.

Mechanical evaluation of structurally modiﬁed polymer ﬁbes require special
consideations. An experimental and theoretical study of axial compressive behavior of
high peformance polyme ﬁbes by Deteesa (1985) has suggested that the longitudinal
shearmodulusofﬁbesareequaltotlneircompressivestrengthwhen ﬁbe'sfailbya
coope'ativechainbucklingmode. Thisconceptisimportantforthemechanical
evaluation of the structurally modiﬁed polyme ﬁbers. Using ISS tests to evaluate
Wymodiﬁedﬁbesmyproduceeronwuﬂylowresmmbecauseofmepresence
ofaweaksurfacelayersproducedbythetreatment. Obsevationofthecompressive
propertiesoftheﬁbesisanindirectbutmorereliableapproachtoevaluationofshear
propetychangesofthehighpeformancepolymeﬁbers. InAppendixA,itisshown
that the single ﬁbe compressive tests are insensitive to the extent of ﬁber-matrix bornd
snengthaslongasﬁbeandmauixundegothesamestraincondidons,(i.e. no
debonding occurs before the ﬁber compressive failure, see Equation A.3). Furthermore,
Deteeea (1985) has shown that the longitudinal shear modulus of high peformance

161
polyme ﬁbes are equal their compressive strength, therefore, changes in shear moduli

of treated ﬁbers should be reﬂected directly in their compressive strength measurements.
In this study, single ﬁbe compressive test wee adopted to evaluate the effects of ﬁber
structural modiﬁcation. 1

Two approaches to reinforce lateral chain interactions of polyme ﬁbers wee
examined in this study: infusion of SiOz and epoxy into interﬁbrillar space of liquid
crystalline polymers to provide lateral support arnd to mechanically couple the ﬁbrils, arnd

Friedel-Crafts acylation reactions to covalently bond adjacent polyme chains.

8.1.llnﬁrsien_af_a.8ecand_2hase
Theuseofasecondphaserationaltoinﬁltrateandreinforcetheﬁbrillar

microstructure of PZT ﬁlms have been patented by Kovar et al. (1989). Howeve, the
patentappliestoﬁlmsanddoesnotclaimapplicationtopolymerﬁbes. Inthisstudy,
two types of infusion agents have been investigated: SiO, by sol-gel reactions and epoxy
resins dissolved in a solvent.

Sol-gel reactions geneally refe to chemical systems where a colloidal suspension of
snnallparticles (sol)1inktogetlneformingasolidmass(gel). Metalalkoxidesarea
common type of sol reagents. Metal alkoxides have the general chemical formula
M(OR), whee M is the metal, R is an alkyl group (C11,, (‘41,, etc.), and x is the
valencestateofthemetalatom. Thesemetalalkoxidescanbehydrolyzedtoformtheir
corresponding hydroxide. One of the alkoxides used in this study is tetra-ethyl-
orthosilicate (TEOS). The oveall hydrolysis reactiorn of TEOS can be represented as

162

follows:
Si(OC,H,), + nH20 - Si(OI-I),(OC,I-I,) + nC2H50H
Si(OI-I),(OC,H,) - SiO, + (n-1)H,O + C2H50H

This hydrolysis reaction can be acid or base catalyzed. The ﬁnal morphological state
oftbesilicagelisdependentonthepHofthesol solution (LaCourse 1988andJones
1989). A high pH (<9) tend to form non-inteacting spheical particles. At pH 5-8 the
particles coalesce into microgel regions. A low pH (>3) tends to form linear molecules
that are occasionally cross-linked. The othe silicon alkoxide examined is tetra-methyl-
orthosilicate (TMOS) that undergoes similar hydrolysis as TEOS. The wate solubility
ofTMOSismuchgreatedunTEOSwhichisanimportantconsideaﬁm(discussed
late). Silica sol-gel reactions are easy to process and have a high tensile modulus plus
high compressive mecharnical propeties. In this study, sol-gel reactions of silicon
alkoxides to form silica gels (SiOz) wee considered to mechanically couple the ﬁbrillar
structure of liquid crystalline polymes.

Epoxyresins weetheotheinfusionagentthatwasinvestigated. Unlikethebrittle
and high modulus sno, gels, epoxy systems can provide a ductile and low modulus
matrixaroundtlneﬁbrils. Thisductilitycouldbeacriticalfactorsincetlnehigh
performancepolyme ﬁbes undegoasigniﬁcantamountofprocessing stresses thatmay
destroy the reinforcing network of a brittle infusion agent.

163
8.1.2 BMW

The otlne approach to increase the inteaction between adjacent polymer chains is
Friedel-Crafts acylation reactions to covalently bond the benzene rings in the PBO
polyme chains. Friedel-Crafts reactions are types of chemical reactions in which an
alkyl (RC-) or acyl (RCO-) group is substituted into a benzene ring by reaction of an
alkyl or acyl halide (Streitwiese et al. 1981). In this study, Friedel-Crafts reactions of
difunctional acid chlorides have been examined to bridge the benzene rings on adjacent
PBO chains. Figure 8.1 shows an overall scheme of the Friedel-Crafts acylations.
Using a difunctional acid halide, it slnould be possible to cross-link adjacent PBO polyme

chains at various sites. Figure 8.1 also shows some possible sites for cross-linking the

 

f? f?
.+RCX —> R +HX
X-F,C|,Br,l
Friedel-CraftAcylationReaction

 

/ \ f \
\f/ \f/

PossiblesitesforPBOchainstocrosslink

 

 

 

Figure 8.1 - An oveall scheme of Friedel-Crafts acylation reaction. A difunctional
acid halide can cross-link adjacent PBO chains at seveal site.

164

PBO chains, of course, othe combinations of intramolecular and intemolecular cross-
linking are also possible.

Applications of the Friedel-Crafts reactions to high performance polymer ﬁbers have
been reported by Mecx et al. (1990), whee, oxalylchloride (Cl-OC-CO-Cl) was used
tosurfacetreatTwaron arannid ﬁbes. Theirproposed heahnents assunnedtlnatornly
nitrogen acylation occurs rathe than the benzene ring reactions. This work suggests that

for the arannid polymes there are more sites for the chennicel cross-linking by Friedel-

Crafts reactions than there are for the PBO polyme.

 

liquid crystalline polyme ﬁbers exhibit low compressive strength and limited
compressive elasticity compared to inorganic reinforcing ﬁbes such as carbon and glass
ﬁber. Composites made from liquid crystalline reinforcing ﬁbes have a tensile to
compressive strengtln ratio of 5:1, wheeas, carborn and glass ﬁber reinforced composites
have about the same tensile and compressive strengths. Figure 8.2 compares some
compressive strength values determined from single ﬁbe measurements (see Appendix
A). Thecompressive strengthsofpolymeﬁbesweemeasuredbyobserving theonset
ofkinkband formations. Figure 8.3 shows an optical micrograph ofcompressive kink
bands for PBO ﬁbers. The low compressional propeties of the high performance
polyme ﬁbers result from their rigid-rod morphology (Deteesa 1985, Dobb etal. 1981).
These ﬁbes are composed of ﬁbrillar eystallites with poor lateal interactions between
adjacentﬁbrils. Figure8.4slnowsSEMmicrographsofaﬁacturedPBO ﬁbethat

165

 

\l

‘ 82222 High Performance Polymer Fibers 2‘:
’ C: Inorganic Fibers

._

PBT Technora Kevlar Kevlar E- Glass AS— 4
29 49 Carbon

0)

U'1
I

M U
l l

Compressive Strength (GPa)
—‘ A

 

 

 

 

O

 

Figure 8.2 - Compressive strength of seveal high performance polyme and inorganic
ﬁbes determined by the single ﬁber compression test.

 

   
 

 

 

\\I.\A\o.w
V‘ \ _

 

 

 

 

Figure 8.3 - An optical micrograph of compression kink bands in PBO ﬁbers.

166

,. .. _..--—~
.. .p¢_'° '

.. a... r

B W ..
.ﬁ-vu . l'.-" . V ' " i I i

 

Figure 8.4 - SEM micrographs of a ﬁbrillated PBO ﬁber.
(A)bar=10um (B)bar=2um

167
clearly demonstrates the ﬁbrillar morphology of the ﬁber. On compression the absence

of strong lateral inteactions results in cooperative buckling of the ﬁbrils that are
exhibited as formation of kink bands at oblique angles to the ﬁber axis (Figure 8.3).
Examining the local bending of the PBO ﬁbers by high resolution electron microscopy,
Martin et al. (1991) have determined that the formation of the kink bands involves
bending and/or covalent bond breaking of the rigid-rod polymer. They also observe that
kinksareinitiatedattheﬁbesurfaceandthengrow towardstheﬁbeinteior. This
observationsuggests surfacehardeningtechniquestlnatcanmakethekinkshardeto
initiate may be useful for improving the compressive strength of polyme ﬁbers. Previous
investigations suggest that to increase compressive propeties of the rigid-rod polyme
ﬁbesthelateralinteactionsbetweentheﬁbrils mustbeincrease. Howeve, Martins:
al. (1991) warn that such microstructural modiﬁcations will be successful only ifthey
do not introduce additional sites for kink irnitiations.

For polyethylene ﬁbes, Attenburrow er al. (1979) have demonstrated that the
compression ldnkbandformationoftheﬁbesaretheresultofthesheardeformaﬁonof
their lamellar crystals. Kazuyo et al. (1975) have reported that during kink band
formation oftlne polyetlnylene ﬁbes, the ﬁbe axis rotates from the original axis by 70-
75° which explains the formation of the helical kink bands.

Deteesa (l985)has investigated tlnecompressive belnavioroftlneararnid ﬁbes. On
axial compression the arannid ﬁbe forms regularly spaced helical kink bands at 50° to
60° withrespecttotheﬁbelongaxis. Thecompressionkinksoccuratabouto.5%
bending strainbutthey disappearonremovaloftlneloadwithoutany apparent affecton

168
the ﬁbe tensile properties. Only afte approaching 3% compressive strains do the ﬁbes

showa10%lossintensilestrength. Deteresa(l985)hasproposedamodelfor
compressive failure by elastic nnicrobuckling of the extended-chain polymers. The model
shows, and the expeimental results conﬁrm, that the stress required to buckle tlnese
ﬁbesisequaltothe minimumlongitudinalshear modulusoftheﬁber(nottlneshear
strengtln). Deteresa also concludes from his model that the compressive strength of
extended-chin polyme ﬁbers is about one-third of their torsion moduli.

Cohen (1986) has investigated the structural elements of PBT ﬁbes and ﬁlms. He
proposestlnattlnebasic structuralfeaturesresponsible forthennechanicalpropetiesare
seduﬁngdnemiﬁalcoaguhﬁmpmssmﬂnethanthehteheatneannentmddrying
processes. Cohen reports that the PET ﬁbe's consist of an inteoonnected network of
oriented microﬁbrils with 'Y-shaped' junctions between the microﬁbrils. TEM
rnicrographs ofthebuckled ﬁbes showthatthecompressive failureoftheﬁbeistlne
result of the buckling of the individual microﬁbrils. Cohen suggests that the dimension
of the microﬁbrils signiﬁcantly influences the compressive strength of the ﬁbers and the
post-treatments may perfect the chain packing but do not alter the ﬁbrillar morphology.
Themeasuredcompressive strengthofaramidﬁbesisabouttwicePBOorPBTﬁbes
(Figure 8.2). PBO and PET ﬁbes lack the intramolecular hydrogen bornding inteactions
thatarepresentinthearamid ﬁbes(Figure3.1). Inaddition,thepresenceof
'Y-shaped' microﬁbril junctions has not been reported for the aramid ﬁbes. Since the
'Y-shape' junctions can be the locus of compressive failure, the presence of these
junctions may be anotlne limitation of the PBO arnd PBT morphology.

169
32W

The polyme ﬁbe examined in this study was p-Phenylene BenzobisOxazole (PBO)
aromatic heteocyclic supplied by Dow Chemical (Midland, MI). Through an special
agreement with Dow Chemical, the PBO ﬁbes wee supplied in two conditions: ﬁbes
tlnatweejustcoagulatedinwateandweestillwateswollen(referedtoasWet-PBO
munisrepon),andﬁbesdnathadundegoneapmpnetarydryingprowssaﬁedner
coagulation (refered to as AR—PBO). The Wet-PBO ﬁbe's wee examined because their
swenedstrumwasconsideedmoreconducivemtheinﬁnsionofuneexamhned
treatmentsthanthehigh crystallinestructureoftheAR-PBO ﬁbes.

The epoxy matrix was the DER331/MPDA 75°C/2hr/125°C/2hr system used for the
measurements of single ﬁbe compressive test (Appendix A) and interfacial shear strength
byfragmentationtest(describedindetailsintheChapte2). Singleﬁbetensile
strengthsweemeasuredbyASTMD3379tensiletest. Fiberdiameteswee
characterized with the aid of a video calipe. Wate content of the Wet-PBO afte
various treatments wee measured by a DuPont9511hemogravimetric analyze (TGA),
ramped at 10°Clmin. from ambient temperatnrre to 400°C. Fiber-matrix internal
morphologywascharacterizedbyTEMmicroscopy. Elementalcharacte'izationsofthe
treatmentsweeconductedbyAFSandXPStechniques.

SolventexchangesoftheWet-PBOﬁbesweedonebyattachinga~5inchlongwet
ﬁbertowonastainlesssteelrodusinganalunninumﬁrnewire. Thesarnplerodwastlnen
placedina6inchlongtesttubewithatwist-offcapandthetubewasﬁlledwiththe

solvent of inteest. For mutually immiscible solvents such as water and Freon, a third

170
transitory solvent such as acetone with mutual solubility in botln water and Freon was

used to exchange between the two immiscible solvents.
For the sol-gel reactions, two types of silicone alkoxides wee examined:
TMOS (tetra-methyl-orthosilicate) Si(OCH,).
T'EOS (tetra-ethyl-orthosilicate) Si(OC,H,),

Because TEOS is immiscible in water and TMOS is only slowly miscible in wate,
for some treatments methyl or ethyl alcohols wee used as a co-solvent. The various sol-
gel reagents and treatment conditions are described in the results section.

For the epoxy treatments, the water in the Wet-PBO ﬁbers wee ﬁrst exchanged with
acetone, and tlnen ﬁbe tows wee immersed in a dilute solution of DGEBA epoxy
(DER331), DDS (DiaminoDiphenyl Sulfone), DEI‘A (DiEtyleneTriAmine), or H31 (a
proprietary Dow Curing Agent), dissolved in acetone. The various heahnent conditions
arealsodescribedintheresults section.

For the FriedeLCrafts reactions, tlnree diflunctional acid chlorides wee used:

oxalyl chloride (Cl-OC-CO-Cl)
succinyl chloride (Cl-OC-CHz-CHz-CO-Cl)
adipyl chloride (Cl-OC-CHz—CHz-Cﬂz-CHz-CO-Cl)

The Friedel-Crafts reaction wee carried out by exchanging wate in the Wet-PBO
ﬁbes with dichlorometlnane (CHzClz) and then adding acid chlorides into the solvent.
Initially, SnCl, was used as a catalyst, but the reactions wee vigorous even without the
catalystasevidencedbyahighrateofHClgasbubblingﬁomtheﬁbes. Thetreatrnent
reagentsconsisted of ~ 1.5 grarnacid chloridemixedin ~40gramsofdichloromethane.

171
8.3 W

Wet-PBO ﬁbers wee chosen for the study of structural modiﬁcations because tlneir
open and wate swollen structure was considered more readily accessible by various
inﬁltrationproeesses. Aseiesofﬁbediametermeasurements ofPBOﬁbeswee
conductedtodeterminehow much swellingispresentinthewetﬁbesarndhowthe
swelling changes when wate is exchanged with another solvent. Fibe diamete's were
measuredwhiletlneﬁbesweebeingsoakedinthesamemediumastheirﬁnalsolvent,
to prevent ﬁbers from drying during the examirnation.

Table 8.1 lists the PBO diameters for wet, dried, and various solvent exchanged
ﬁbes. Ingeneal, ﬁbesthatweenotdriedremainedswollenevenwhenwatewas
exchanged with othe solvents. Howeve, once the ﬁbes wee dried, the ﬁbers could
not be reswollen by the solvents exchange. These ﬁbe diamete measurements show that
inswollenstatePBOﬁbesareabout80% largethanthedriedstate. Theabilityto
exchange wate with otlne solvent without reducing the ﬁbe volume is critical for
conducting treatments that are wate inhibited. Themogravimetric (TGA) measurenents
oftheWet-PBOﬁbersalsoconﬁrmtheprevious results. TGAweightlossmeasurements
forvarioustreatrnnents ofWet-PBOarelistedintheTable 8.2. TheeeTGA results
suggest that the wet or solvent exchanged Wet-PBO ﬁbe have about 35 wt% of wate
contentbeforetheyaredried. Howeve,oncetheﬁbesaredriedtheycannotbe
reswollenbysoakinginwateagain. BothdiameteandTGAmeasurementssuggestthat
the highe tempeature drying conditions (400°C) tend to reduce the ﬁbe free-volume
more than the ambient drying.

172

Table 8.1- Fibe diamete for dried, wet, and solvent exchanged PBO ﬁbers.

 

 

 

Water swollen 26. 9 d: 4.1
2 weeks in Ethanol Ethanol 26.0 :1: 4.6
1 day Acetone, 1 day Freon, 1 day Acetone, wate 26.3 :1; 4.6
1 day water

1 day Ethanol, 1 day Acetone, 5 days Freon Freon 27.9 :1; 4.7

 

1 day liquid N; Frozen, 1 day ambient dried

   

21.5 :1: 4.3

 

  

    

 

 

 

 
     

 

  

 

 

 

    

 

 

 

 

 

 

‘ Aveage of 30 measurements.

1 day ambient dried, 1 day liquid N, Frozen, water 21.5 :1: 4.3
1 day ambient dried

1 day ambient dried, 2 day Acetone Acetone 20.6 d: 3.6
2 weeks ambient dried, 24 hour wate soaked wate 19.3 i 3.6
1 day ambient dried dry 18.8 :1: 3.8
1 day Etlnanol, 1 day Acetone, 1 day Freon, dry 21.7 :1: 2.9

L 5 hour ambient dried

5 days ambient dried, 400°C dried water 19.2 :1: 3.5
5 days ambient dried, 440°C dried dry 18.4 :1: 3.1

 

 

 

Table 8.2 - TGA weight loss measurements for various treatnnents of Wet-PBO ﬁbers.

w... ..

Wet-PBO

36.6 :1: 8.3

 

Wet-PBO -. EtOH -9 Water

34.7 :i: 4.6

 

Wet-PBO 10 min liquid N, frozen, 24 hrs in wate

'nsias

 

Wet-PBO 24 hrs ambient dried, 48 hrs in water

9.7 :1: 1.1

 

Wet-PBO 4W°C dried, 1 week in water

 

 

3.64 :1: 2.9

173
For the Wet-PBO ﬁbes, exchanging wate with otlne solvents appears to affect

tensilepropetiesoftlneﬁnaldried ﬁbe. Table8.31iststhetensilepropertiesofthe
solvent exchanged Wet-PBO ﬁbes. In general, drying ﬁbes at high tempe'ature
increased ﬁbe tensile modulus which suggest a highe degree of chain orientation.
Exchanging wate with acetone or alcohol tend to reduce modulus and tensile strength
of the ﬁbe's which may be due to partial solubility of the ﬁber polymer in these solvents,
decreasing molecular orientation. These obsevations suggest that the choice of solvents
used in a treatment not only affects the chemistry of the treatrnnent but it could also
inﬂuence the morphology the ﬁbers.

Table 8.4 lists XPS elemental composition of the AR-PBO, Wet-PBO, 12-hour wate

soxhlet extracted Wet-PBO, 12-hour metlnanol extracted Wet-PBO, and 12-hour water

Table 8.3 - Mateial prope'ty data for solvent exchanged Wet-PBO ﬁbe's.

Wet-PBO-eRT dried 3.28 :1: 0.51 110 :t 11 21.8 :1; 2.6

Wet-PBO - RT dried 3.55 d; 0.87 146 j; 14 20.8 :I: 2.6
RT-D8IIH350°C-O8IIMRT

 

We-PBO -0 Acetone-0 RT dried 2.62 :1; 0.31 83 :l: 18 22.0 :1: 3.0
We-PBO-O MeOH - RT dried 2.78 1: 0.26 77 :l; 8 22.3 :1: 2.4

Wet-PBO- Acetone 2.86 :l; 0.46 115 j; 7 21.7 :1; 1.3
RT~8Iu-2W°C~8h~RT

 

Wet-PBO -. EtOH 2.09 :l: 0.60 78 :1; 9 20.8 1; 2.5
-0 CO, (liq. -0 gas) -o RT dried

 

 

 

 

 

RT-Room‘l‘enperanlre

174
soxlnlet extracted Wet-PBO that was dried at 400°C for 1 hour in a nitrogen environment.

The All-PBO appears to have a slightly oxidized surface, but all of the Wet-PBO ﬁbers
before heat treatment exhibit about twice the oxygen content of heat treated ﬁbe's.
Figure 8.5 shows superimposed narrow scans of carbon, oxygen, and nitrogen for the
12-hour wate soxhlet extracted Wet-PBO ﬁbers before and after the 400°C drying.
Carbon and nitrogen signals show similar compositions, although, thee are some
resolution difference between signals of the two treatnnents. The oxygen signal for the
Wet-PBO ﬁbers, however, shows the presence of another type of oxygen in the 531-532
eV region. The lower binding energy oxygen signals typieally correspond to more
electronegative oxygens; in this ease most likely tlnose of earboxylic or ester oxygens in
the ketone position. The non-stoichiometric amount of oxygen on the Wet-PBO ﬁbers
maybeduetopresenceacidic functionalitiesthatareleﬁfromtheﬁberspinningprowss
andhavenotbeenneutralizedbythecoagulationprocess.

Table 8.4 - Atomic concentrations for Dow PBO ﬁbes, measured by XPS (average of
three runs for each treatment).

AR-PBO 79.0 :1; 0.4 13.2 :1; 0.5 7.78 :1: 0.37
Wet-PBO 69.7 :1: 0.4 23.0 1: 0.8 7.29 :l: 0.37

We-PBO, 12 hrs soxhlet extracted in water 70.0 :1; 0.9 23.3 i; 0.8 6.75 t 0.13
Wet-PBO, 12 hrs soxhlet extracted in MeOH 68.2 1; 0.5 25.4 :I: 1.0 6.38 :l; 0.50

Vie-PBO, 12 hrs soxhlet extracted in water 77.5 i 0.4 13.8 :I; 0.9 8.63 :1: 0.47
1bourO4N°Cin70cclmim N,

 

 

 

 

175

 

   

VVVVV'V'VV‘
AAAAAA

A4 A A A A A

.t
V V V V V V V V V V V V V V V V 1 V V V

N(E)/E, shf, smo7

enuoauonooi

 
 

' V V U V

IAJAAAIAAAA‘A‘AJAAJ
I

 

 

 

V V V

o s- to us a. an on ~s . o 8
A l A A A A A A A A A A A A A A A A A
v v v v v vﬁ I V V I V V

v

 

 

 

 

 

 

.1
”L
1
81'
P
h .1»
qt
3 ,.
1.
$7 T
‘-
5v
1»
‘1»
o
'4-
z ..
20
l.
‘1
D
. A A A A A J A A A l A A A A A A A
w v r v v v v r r v v wﬁ 1 v v v r ‘1
a I “ . .4 a a at . II ‘

BINDING ENERGY, eV

Figure 8.5 - Supeimposed XPS signals of carbon, oxygen, and nitrogen regions for the
12 hour water soxhlet extracted Wet-PBO ﬁbes before and after 400°C drying.

176
8.3.1 We
The sol-gel approach attempts to fornn a three-dimensional gel network inside the
ﬁber structure that can reinforce the ﬁbrillar morphology of the liquid crystalline polymer
and improve its sheer and compressive properties.

The ﬁrst set of sol-gel treated PBO ﬁbers were provided by Foster-Miller, Inc.
(Waltham, MA). A Wet-PBO ﬁber tow was inﬁltrated with a silica sol-gel reagent while
mounted on a spring-loaded C-shaped metal ﬁxture, and then dried to 350°C. Another
tow ofuntreated PBO ﬁbers undewentthe samedryingproceduretoprovidethecontrol
mate'ial. The target sol-gel concentration was 5 wt% in the ﬁber.

Table 8.5 compares the tensile, compressive, and interfacial shear strength of the
'as received“ (DOW PBO XV-0383-C8700975-008) and Foster-Miller control and sol-gel
treatedPBO ﬁbers. Thecompressive strengthdataaredeterminedbyEquatlonAJ

Table 8.5 - Mateial propety data for Foster-Miller (F-M) sol-gel treated and untreated
PBO ﬁbe's.

Property Dow F-M F-M
i As Received control Sol-Gel

Tensile Strength (GPa)

3.42 :l: 0.55

2.97 :1: 0.55

2.51 :1: 0.29

 

Tensile Modulus (GPa)

181 j: 17

197117

135 :1: 15

 

Fracture Strain (%)

2.05 :1: 0.45

1.63 :1; 0.26

2.29 :1: 0.39

 

Diameter (pm)

18.8 d: 2.1

16.9 :I: 2.3

19.8 :1: 2.4

 

Compressive Strength‘ (MPa)

368199

457 j; 107

294th

 

Interfacial Shear Strength’ (MPa)

 

17.8 :1: 1.4

 

15.0 :I: 3.0

 

 

‘ from single ﬁber compressive measurements using uniform cross-section specimen.

3 for 175°C cured epoxy matrix.

177
using uniform-Cross-Section specimen. The sol-gel ﬁbers do not show any compressive

shength improvement over the control ﬁbers, however, there is about 15 % reduction in
their tensile strength. Interfacial shear strength of the sol-gel treated ﬁbers is also about
half the untreated values.

The interfacial shear strength reduction of the sol-gel treated ﬁbers can be attributed
totlnepresenceofSiozweakboundarylayeronthesurfaceofthetreatedﬁbers. Figure
8.6 shows SEM micrographs of the sol-gel treated ﬁbers. The rnicrographs show two
typesofsilica gel topography, geliseitlnerintheform oflargeand thickislandsorthin
nonuniform coatings. With a nominal 5 wt% gel concentration and its accumulation on
theﬁberextelior, theinternalgelconcentrationisexpectedtobelow. EDXandAPS
examinationsofthesol—geltreatedﬁberswe'eunabletodetectthepresenceofsilicon
morethanamicronbelowthesurfaceoftheﬁber. Figure8.7showstheAESline—scan
of Si atoms superimposed on a SEM image of a sol-gel treated PBO-epoxy intephase.
Onlyabout1pmofSiispresentneartheﬁber-matrixinterphaseandthereisno
signiﬁcant SiO, penetration into the ﬁber structure. TEM micrographs of a sol-gel
treated PBO ﬁbers also demonstrate SiOz accumulation on the ﬁber exterior. Figure 8.8
showsanaxially sectioned Sol-GeltreatedPBOﬁber, showingthegelcoating (dark
band) around the ﬁbe perimeter. These observations suggest that SiO, gel did not
penetrateintothebulkofthePBO structure. Furthermore, presenceoftlneSiqul-face
layeronthesol-geltreatedﬁberscreatesanewweakboundarylayebetweentheﬁber
and matrix which reduces the ﬁber-matrix interfacial shear strength. Compressive
strengtlns ofthe sol-gel treated ﬁbers wee unaffected by the sol-gel treatments since the

E

178

 

Figure 8.6 - SEM micrographs of Foster-Mine sol-gel treated PBO ﬁbers.
(A)bar=50um (B)bar=10um (C)bar=2um (B)bar=-Hum

179
$0; gel network did not penetrate into the ﬁber bulk.

The previous results demonstrate that evaluation of ﬁber structural modiﬁcations from
interfacial shear strength measurements can lead to inaccurate conclusions because of the
aberration that may be introduced by the ﬁber-matrix interphase. For example, a
structurally modiﬁed polymer ﬁber with increased shear properties may exhibit reduced
interfacial shear strength if a mechanically weak layer accumulates on its exterior.
Conversely, compressive tests are insensitive to the extent of ﬁber-matrix bond strength
as long as no debonding occurs before compressive failure (see Appendix A).
Consequently, compression tests are emphasized for the remainder of this study.

Other sol-gel treatments of the Wet-PBO ﬁbers with various silica sol reagents were

.3
l' 't ':I

 

7/08/8? 15.8kV 29.8kX 1

ﬁgure 8.7 - AES line-scan of silicon superimposed on a SEM micrograph of a Foste-
Miller sol-gel treated PBO ﬁber imbedded in an epoxy matrix. (0 cross-sectional View).

 

      

ﬁgure 8.8 - TEM micrographs of an ' y sectioned Foster-Miller sol-gel treated PBO
ﬁber. Note the accumulation of kink bands near the ﬁber exterior.
(A)bar=5;lm (B)bar=lum

181

also examined. Table 8.6 summarizes the other examined sol-gel treatments and their
compressive strength results. None of these sol-gel treatments produced any signiﬁcant
compressive strength increase. Their SEM, TEM and ABS observations produced
similar results to the Foster-Miller sol-gel treated ﬁbers, suggesting that the gel network
is not penetrating into the ﬁber structure for any of these examined conditions.

TheinabilityoftheSiozgeltoinfuseintotheﬁbercoremaybeduetoadiffusion
limited mass-transfer phenomena. In Chapter 6, it was demonstrated that sulfornation
penetration into a polycarbonate ﬁlm is limited by a formation of a barrier layer of
sulfonated materials. A similar phenomena could be occurring with the sol-gel
treatments of the PBO ﬁbers. A fast conversion of sol species into an immobile gel
could form a barrier to further sol diffusion into the ﬁber interior. Furthermore, the
acidic surface of tlw PBO ﬁbers (Figure 8.5) may catalyze the sol-gel reactions and
accelerate the formation of the gel barrier or fornn long and linear gel molecules which

would have difﬁculty diffusing into the ﬁber structure.

182

Table 8.6 - Wet-PBO sol-gel treatments and corresponding compressive strengths.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Condition E, (GPa) ed (MPa)
As Rweived (heat treated) 181 a; 17 315 1 7—8'
Wet-PBO 24 hour RT dried 110 :i: 11 144 :t 31
1 day RT dried, RT-08hr-t350°C-8hr-eRTM' 146 :t 14 507 :t 102
1 day RT dried, RT-e8hrv350°C-t8}nr-ORT50g T 146' 443 i 74
l W =
1 hr sonicated in TMOS, RT-tlhr-v150°C-Ihr-eRTNI' 84' 186 j: 78
== =
TMOS/H10/HZSO‘ (25ml/25ml/O.3ml) for 6hr 84' 133 i 78
(removed before gelation) RT-OO. 51119125 °C—-0. 75HTM‘
mos/H,0/H,so. (30ml/30ml/0.3ml) for 6hr 84' 324 :t 50
(removed before gelation) RT»3hH350°C-8hr~>RTM'

TMOS/HZOIHZSO. (35ml/35m1/0.3ml) for 6hr 84' 281 :t 56
(removed before gelation) RT-s3hH350°C-8hr-RT 50g T

=— —_ =
80hr in TMOS, 24hr tensioned in water, 24hr RT dried 84' 174 :t 27
RT~3hH350°C¢8hr+RT M'

E _
TMOS/HZOIEtOI-l (30ml/30ml/15ml) for 3 day 74 :t 4 187 j; 66
(removed aﬁer gelation), 1 day RT dried
RT-tanr-350°C-e8hr-RT 50g T
TMOS/Hp/EtOH (30m1/30ml/30ml) for 2 day 84' 169 :l: 42
(removed after gelation), 1 day RT dried under tension
RT—tﬂnn350’C—WWRT50g T
TMOS/Hp/EtOI-i (30ml/30ml/30ml) for 2 day 84' 207 j; 59
(removed after gelation), 1 day RT dried under tension
RT-uanr-e350°C-.8hr~eRTM'

TMOS/H20/EtOl-i (30ml/30ml/30ml) for 2 day 84‘ 190 i 59
(removed aﬁer gelation), 1 day RT dried tension-free
RTWIH350°GD8IIHRTNT

m
TMOS/EtOH (30ml/30rnl) for 3 day, 84' 70 :t 10
24 hr in 100% humidity @ 80°C, 24hrs RT dried
TMOS/EtOH/26020 (20m1/20ml/20ml) for 2 day, 2 day in 84' 80 j: 15
water, 1 day RT dried
TMOS/EtOH/ZGO2O (20ml/20ml/20rnl) for 2 day, 2 day in 124 :1; 8 298 :t 98
water, 1 day RT dried, RTM'C—MTM‘

m =
TMOS/EtOH (equirnolar) 3 day, 12hr water, 12hr RT dried 146' 287 :1: 74
RT-e8hr-350°C-012hr-RT NI‘

TMOS/EtOH (equimolar + lcc Acetic Acid) for 3 day, 146' 312 :t 156
12hr water, 12hr RT dried, RT-e8hr-e350°C-.12hr-RTM
TEOS/EtOH (equimolar) 2 day, 12hr water, 12hr RT dried 146' 232 :l: 93
RT-nﬂhrdSO'C-thr-eRTNI'

TEOS/BOH (equimolar + lcc Formic Acid) for 2 day, 146' 343 :t 159

12hr water, 12hr RT dried, RT-e8hr-O350°C-12hr-RTM'

 

 

 

 

 

‘Assunnedmodulus RT-RoomTempeature

NT-NoTension T-Tensioned

183

8.3.2 W

The objective of the epoxy treatments is to produce a cross-linked epoxy network
within the ﬁber structure. Table 8.7 lists the various attempts to impregnate the Wet-
PBO ﬁbers with epoxy resin and the resulting compressive strengtlns. In this approach,
ﬁrst the water content of the wet-PBO ﬁbers was exchanged with acetone since acetone
isasolvent forbotlntheepoxyand thecuringagent. Curingagent moleculesare smaller
than the epoxy monomer molecules, hence, ﬁbers were ﬁrst soaked in the curing agent
solution to enhance the possibility of their complete inﬁltration. The acetone exchanged
ﬁbers wee inserted in a solution of curing agent (DETA, DDS, or H31) in acetone, and

were allowed to equilibrate. Next, the ﬁbers were inserted in an epoxy (DER331) and

Table 8.7 - Compressive strengths of epoxy impregrnated Wet-PBO ﬁbers.

,_——

As Received (heat treated) 181 j; 17 315 j; 78
Wet-PBO 24 hour RT dried 110 :1: 11 144 :l: 31
Wet-PBO 48 hour Acetone washed, RT dried 83 :1: 18 99 :1: 34

Wet-PBO 13 hour MeOH washed, RT dried 77 j; 8 109 j; 40

 
 
  
  
       
   
   
   
       
    
 

         
       
   

AR-PBO DEI‘AIAcetone (1/1) 19 hour
- DER33l/Acetone (1/6) 5 day, Acetone washed, ar Dried

Wet-PBO DDS/Acetone (1/4) 12 hour

- DER33l/Acetone (1/3) 2.5 day, Acetone waslned,
RT Dried 77‘ 91 j: 35
RT»Ihr-180°C-1hr+180°C-¢12hr-.RT 425g 7‘ 122 j; 7 119 :1: 31
RT-tlhr-220°C-t1hr+220°C-t12hr-RT 1503 T 122' 139 :1: 63
RT~1hH220°C-Ihr-220°C-’12hr-RT NT 122‘ 176 :1: 71
RT-t8hr-350°C-12hr-RTNT 146‘ 309 :1: 97

Wet-PBO Acetone washed
100% H31 2.5 hour, RT Dried
100% H31 4 days, RT Dried

 

 

 

 

 

NT-NoTension T-Tenioned RT-RoomTempeature 'Assumadmoduhns
DETA I DiBtyleneTriAmine DDS I DiaminoDiplnenyl Sulfone
H31-PropriehryDowCuringAgent DER331-ProprietaryDowDGEBAspoxy

184
acetone solution to allow the inﬁltration and reaction of the epoxy with the curing agent

inside the ﬁber. Finally, ﬁbers were rinsed with acetone to remove excess epoxy or
curing agent, and were thermally cured to complete the reaction of the inﬁltrated epoxy
system. Figure 8.9 shows an ABS nihogen linescan of BETA/Acetone soaked wet-PBO
ﬁber. In this line-scan, contribution of the nitrogen from the ﬁber and the matrix
moleculeshasbeensubtracted fromtheAESbackgroundandtheremainingsignalisdue
to BETA distribution. This nitrogen line-scan conﬁrms that DEl‘A has completely
inﬁltrated the ﬁber.

None of the attempts given in Table 8.7 resulted in signiﬁcant compressive strength
increasesorﬁbemoduluschanges, which suggeststhattlneepoxynetworkisnot
penetrating into the ﬁber interior. The lack of epoxy inﬁltration is probably due to its
large monomer size. Apparently, even though swollen wet-PBO ﬁbes have 80% large
volume than the dried ﬁbers, their free-volume is not freely accessed by large molecules
such as the epoxy.

 

 

 

 

 

 

Figure8.9-AFSnitrogen line-scan ofaDETA inﬁltrated wet-PBO ﬁbe, showingthe
digitizedimageofaﬁbercross-sectionembeddedinanepoxymatrix. bar =25pm

185

83.3me
Table8.8comparesthetensilepropertiesofthePBOﬁbestreated withtlnethree

typesofFriedel-Craftstreatmentreagents. Allofthetreatedﬁbe showreducedtensile
strengthandfracturestrain, however, theﬁbertensilemoduli wee unaffected. Since
thetensilestrengthoftheﬁbersaredefectcontrolled,while,theﬁbetensilemodulus
isbulkcontrolled,thedatasuggesttlnatthetreatmentshavenotpenetratedintotheﬁbe
structure and have ornly affected ﬁber surface morphology. Figure8.10 shows TEM
micrographs of the oxalyl and succinyl treated PBO ﬁbes. These TEM micrographs
showthepresenceofadarkbandaroundtheﬁbepeimetethatisaboutlOOnmwide.
'I‘hisdarkbandispresumablytheboundarylaye of tlne reaction penetration. 'Theefore,
WobservaﬁemwggestthatmeFﬁedd-Cmﬂsreacdeuhavenmwoceededinwdne
ﬁbebulkanditspenetrationhasbeenlimitedtotheﬁberexteior. Thediffusion
ﬁmitaﬁonofdneFﬁedd-Cmﬁsheannenmissimﬂarmthemdfmadmdiffusimﬁmited
mass—transfe problems that have was discussed previously; the treated surface forms a
diffusion barrier that block the access of mobile reactive species into the ﬁbe inte'ior.

Table as - Mateial property data for Friedel—Crafts treated Wet-PBO ﬁbes.

Treatment Tensile Tensile Diameter Fracture
Strength Modulus (pm) Strain
. i (GPa) f (GPa) (5) .

Wet-PBO .. RT dried 3.28 :1; 0.51 110 j; 11 21.8 :1; 2.6
Oxalyl Chloride reacted 1.78 3; 0.57 112 j; 9 19.8 i 4.4 1.87 :1: 0.57
Succinyl Chloride reacted 2.11 :1: 0.40 108 :1: 15 22.5 :1: 3.7
Adipyl Chloride reacted 1.24 i 0.32 101 :l: 16 20.8 :1: 3.1

     
         
        

 

 

 

 

 

        

 

Figure 8.10 - TEM nnicrographs of (A) oxalyl and (B) succinyl treated PBO ﬁbers
showing only a ~ 100 nm of reaction penetration.

187
8.4 W

0 Examination of water swollen PBO ﬁbers show about 80% large ﬁber volume for
theswollen ﬁberscomparedtodriedﬁbers. ’IheswellingofthePBOﬁberremained
unchanged through different solvent exchanges, however, once the PBO ﬁbers wee
dried their structures could not be reswollen.

O The Wet PBO ﬁbers show an excess presence of carboxylic oxygen which was
removed by the 400°C drying condition.

0 Despite the swollen structure of the PBO ﬁbe, the results of sol-gel, epoxy and
Friedel-Crafts treatments showed that the reactive species were not penetrating into
the ﬁbe interior. Formation of a diffusion barrie by the reacted species is

postulatedtobethereasonforthelimitedtreatmentpenetration.

Results of this study suggest that the structure of the swollen polymer ﬁber has eitical
inﬂuence on the mechanical propeties of the dried ﬁber. The free-volume of the swollen
wet—PBO ﬁbes is not easily accessible to large molecules making the infusion of a
secondreinforcingphaseadifﬁcultapproachtoexecute. 'Ihemecharnicalpropetiesof
tlnedriedPBOﬁbewereaffectedbythesolventusedduringtheswollenphase,
mggesdngﬂnatdwpolymerchainreoﬁenmdmmaybepossiblemﬂwswouenphase.
These observations suggest that modiﬁcations of the ﬁbe morphology in the swollen
phaseisﬂnekeytostrucmmmodiﬁcadonofdnehighpeformancepolymeﬁbes.

CHAPTER 9

 

Conclusions & Recommendations

 

The goals of this dissetation wee: to investigate structural propeties of high
performaMe polymer ﬁbers that affect their adhesive behavior; to develop a fundamental
undestanding of the ﬁber structural limitations; to evaluate seveal novel techniques that
can enhance the ﬁber adhesive peformance propeties; and to suggest ways to improve
adhesive propeties of the high peformance polyme ﬁbes.

During this study, seveal important polyme treatments wee investigated and their
particularresultshavebeendiscussedintheircorrespondingchaptes. Thischapte
presents themainconclusionsofthisdissetationonthestrucmmpmpeﬁesofhigh
pefornnance polynne ﬁbes and their effects on the ﬁbe-matrix adhesion. Sonne
recommendations for further studies on adhesive properties of high peformance polymer
ﬁbers are presented.

188

189

9.1 W

Resultsofthisstudy suggestthattheadhesivepropertiesofthehighperformance
polymer ﬁbes may be limited by the ﬁbe surface morphology and/or wetting propeties.
Polymer treatment techniques examined in this study demonstrated their ability to
overcome the surface limitations of the polyme ﬁbes. Table 9.1 lists the particular
effectiveness of each polyme treatment on ﬁber-matrix adhesion enhancement
mechanisms. In general, ﬁbe-matrix wetting compatibility could be improved by all of
the examine polyme treatments. Surface treatments which etch the polyme substrate
to remove a weak surface laye on the ﬁbe (eg. PBO) and/or introduce sites for
mecharnical interlocking with the resin, wee vey effective for enhancing the ﬁber-matrix
mechanical interactions. The polyme heahnents also introducte new active sites for

chemical bonding between ﬁber and matrix molecules.

Table 9.1 - Adhesion enhancement mechanisms introduced by various treatments of the
high performance polymer ﬁbers.

 

Fibe-Matrix Adhesion Enhancement Mechanism

 

Weak skin Active surface Mechanical
removal functionalities interlocking

Coupling Agents &
Polyme Coatings

 

Plasma Treatments ‘ + +

 

Chemical Treatments

 

Ion Implantations

 

 

 

190
The examined polyme treatments could increase the intefacial load transfe capacity

of the high perforrnnance polymer ﬁbers to the point whee intenal ﬁber ﬁbrillations
would occur, indicating that the ﬁbe lateal cohesive strength has become the limiting
factor. Although, interfacial shear strength (ISS) increases obtained by some of these
treatment techniques are signiﬁcant relative to the untreated values, tlnese ISS
improvements are still far lowe than values measured for inorganic reinforcing ﬁbes.

The results of polyme surface treatments suggest that surface morphology and/or
wetting properties of high performance polyme ﬁbers can limit their adhesive propeties,
howeve, once these limitations are overcome ﬁbe lateal cohesive strength becomes the
limiting factor. Therefore, the key to improving the ﬁber-matrix adhesion is to enhance
both adhesive and cohesive properties of high performance polymer ﬁbe's.

Structural modiﬁcations of tlne high performance polynne ﬁbes to reinforce lateral
cohesivesﬁengthofthepolymeﬁbesweeauenptedusingapproachetoinfusea
secondary reinforcing phase (sol-gel, epoxy) or to chemically cross-link adjacent polyme
ﬁbrils (Friedel-Crafts reactions). These attempt wee unsuccessful since the treatments
werenotpenetrating intotheﬁbebulkstructure. Intheseattempts, tlnepenetrationof
thereactingphasewasfoundtobehinderedbyforrnationofabarrielayeofreacted
nnaterials that blocked diffusion of the reaction front.

191
9.2mm
Several other workers (Dctcresa 1985, Dobb et al. 1981, Cohen 1986) have

investigated the morphology-property relations and compressional behavior of high
performance polyme ﬁbers, howeve, the proeess of the compressive failure is still not
well understood. For example, it is not clear how a compressive failure initiates and
propagates. An undestanding of the compressive failure mechanism is important and
should be pursued furthe.

The complication introduced by formation of a diffusion barrier during structural
treatments ofthepolymeﬁbessuggestsdnatpost—ﬁeaﬁnentofdneseﬁbesisadifﬁellt
approach totheﬁberstructulal modiﬁcation. Manipulationofﬁbeinternaland surface
morphologyduﬁngimmanufacmﬁngshwdbeexploredasmenextstepmmodifymg
morphological properties of high performance polymer ﬁbers. Infusion of a second
phase intothepolymedopebeforespinningandcoagulation mayprovidethedesired
mechanical reinforcement of the ﬁbe ﬁbrillar structure. Silicon alkoxides would be
suitable candidates for the addition into the polymer dope because of their intrinsically
highshearandcompressivemodulusaswdlastherabiﬁtymwidnstandinghigh
prowssingtempe'atures. Aftertheﬁbespinninganditsintroductionintothewater
coagulationbatln,thesiliconalkoxidescanquicklyreactwitlnwatetoformthegel
network. The presence of the solvent acid should catalyzed the sol-gel reaction and form
long silicapolymeschainstlnatcantiebetweentheadjacentﬁbrils.

Anotheareaofmtereumatshouldbeplusuedisdnemodeﬁngofvaﬁouspolymer
surface heahnents. Currently, thee are numerous polyme surface treatment techniques

192
available; howeve, modeling of tlnese treatment processes has been hindeed by the lack

of information on their treated interphase composition. The new sample preparation
technique for electron beam analysis of polymes that was developed during this
dissertation allows for data collection on interphase composition and distribution of the
polyme treatments. Improved control of polymer treatments can be achieved by
modeling and undestanding their treatment processes.

Developing an arnalytical undestanding of the diffusion-limited polyme surface
treatments should provide valuable insights to approaches for controlling the penetration
depths of these treatments. For the structural modiﬁcation techniques examined in this

dissetation, this knowledge is critical to achieve deep inﬁltration of treatments.

APPENDICES

APPEADDI A

 

Compressive Strength Measurements of

High Performance Polymer Fibers

 

Compressive properties of reinforcing ﬁbers are difﬁcult properties to measure. Test
methoddependencyandvaﬁousfaﬂurecﬁteionhavepmducedhrgescadeinthe
reported values oftlne reinforcing ﬁbers compressive propeties. In this appendix, some
oftlnetechniquesformeasuring singleﬁbercompressivestrengtlnsarereviewedand
various ﬁbe compressive failure criteia are discussed. Particular attention is given to
the compressive properties ofhigh performance polynne ﬁbe's. A new variation ofan
embedded single ﬁber compression test is also introduced.

193

194 AppendixA

AJW

Thee have been many techniques developed for the compression testing of ﬁbe
reinforced composites, howeve, presently thee appears to be no univesally accepted
standard. Theinconsistenciesofcompressiontestingaretlneresultofmauixand
interfacial dependencies of compressive properties arnd variabilities of compressive failure
modes. Someaspectsofﬂnesemauixandinterfacialdependencieambﬁeﬂydiscussed
anddnensingleﬁbetechniquesanddneirﬁbefailurecriteriaarereviewed.

Compressive failures of ﬁber reinforced composites are typically the result of the
mieobucklirng oftheﬁbes (Agrawaleal. 1980). Most inorganicﬁberssuchasglass
orcarbonﬁbeshavemuchhighecompressivestrengththanpolymemauices. During
the compression of an inorganic ﬁbe and polyme matrix composite, the Poisson’s ratio
diﬁ‘eencebeweenﬁbeandmauixcanmuoducemsvesesuesesattheﬁbe-mauix
mtefaceMrendmmmauixyiddingand/eﬁbe-mauixmtefacialdebondingbefom
the ﬁbe microbuckling occurs. For these composites, a strong interface and/or high
matrixmoduluscanhelptodelaytheonsetoftheﬁbermicrobuclding. Therefore,
matrix and interfacial conditions could critically affect the onset of ﬁbe compressive
failures.

Mauixandmtefadalpmpedecanalsohnﬂuenceﬂneulﬁmatetensilepropeﬁebut
toalesseextendthantlnecompressivepropeties. T‘ypically,intheﬁbereinforced
compoﬁteswidnpolymemauices,dwﬁbeismorebﬁnledmnthemanixandtensile
failuresareinitiatedbytheﬁberbreakageatadefectorweakpoint. Onceaeackis
Mdateditcanthenpropagatethmughdnemahixmpmduceanmheﬁbefaﬂunejoin

195 Appendix A
otlne cracks and eventnnally cause ultimate failure of the composite. Therefore, matrix

and interfacial conditions that affect the crack prorogation can also affect the ultimate
composite tensile properties. Madhukar et al. (in press) have investigated the effects of
surface heahnents on tensile and compressive propeties of carbon ﬁber-epoxy
composites. Using the same carbon ﬁber with different surface heahnents, their work
showed that the ultirrnate compressive properties are more matrix sensitive than the tensile
propeties for these composites.

Many workers have compared the various compression test methods for composite
mateials. Schoeppne et al. (1990) have published a comprehensive review of the
compression test methods for polyme matrix composites, concluding, that no simple and
reproducible compression test technique is yet available. Chou et al. (1980) have
examined the test conditions that complicate compressional properties measurements.
Working with glass-epoxy composites, they showed that the method of specimen
gripping, ﬁbevolume fraction, and ﬁbealignmentallhaveapronouncedeffecton the
measured compressive propeties. Using three test ﬁxtures with different specinnen
loadings, Bergeal. (l989)haveshownthatdirectendloadingcanresultinpremature
faﬂurebecauseofendeuslungmndspﬁtﬁngofdnetenspecimenandmcommended
shearloading suchasthoseusedinthelITRIorCelaneseﬁxtures. Finally,becauseof
memongmanixdependencyofﬂnecompresivepmpeﬁesdwvaﬁaﬁmindnespednwn
fabrication technique can introduce sigrniﬁcant scatte in the composite compressive
values.

Problems encountered in compression testing of the ﬁbe reinforced composites are

196 AppendixA
evenmoreseveeinthecaseoforganicreinforcingﬁbessuchasaramidsand

polyethylene ﬁbers. Rueda er al. (1990) have shown that the standard compression tests
are inadequate to measure the compressive propeties of the Kevlar-Epoxy composites.
The low level of the ﬁbe-nnatrix adhesion aggravates the specimen loading problems
(end crushing and splitting). Chou et al. (1980) also demonstrated that tlne critical Eule
buckling load is signiﬁcantly reduced if transvese failure occur during testing. Another
difﬁculty is the lack of a clear failure criteria for organic ﬁber compression testing. As
shown in the Chapter 8, the organic ﬁbe have low inherent compressive propeties and
exhibit compressive failure by formation of the compressive kink bands (Figures 3.11,
3.12, 5.4, and 8.2). Howeve, during composite compression testing, only gross failures
oftlnetestspecimenisrecorded, andtheirndividualldnkbandformationsoftheﬁbes
are not monitored. The aggravated problems with compression testing oftlne organic
ﬁbecompositessuggestsﬂnatthesingleﬁbetechniquesmaybebeﬂesuited fortlne
unambiguous compressional evaluation of these ﬁbes.

Singleﬁbetechniquesdimhateorsimpﬁfythemaﬁixdependencymndaﬂowthe
in-situ ﬁbe failure process to be closely monitored. The quantiﬁes of ﬁber needed in
dnesingleﬁbetechniquearefarlessdnanindnecompositetesttechniques, makingthem
attractive at the early stages of ﬁber development. Although, among various single ﬁber
techniques,theeissigruﬁcantscaueinreporteddanmanypardcuhrﬁbe,the
interpretation of this data is simpler than for composites.

197 Appendix A
A.25' l E] C . [SE31 1 .
Inthistestasingleﬁbeisembedded alongthecenterofaO.75 inchlong uniform

cross-section or curved neck epoxy coupon. Figure A.l illustrates the three specimen
geometriestlnatcanbeused. Thedogbonesamplegeometryistlnenewvariationoftlne
SFCteststhatisintroducedinthisreport. TheseSFCtestsareconductedbyloadinga
specimen at its ends and compressing it slowly until a ﬁbe compressive failure is
detected in the gauge section. The ﬁber compressive failure process is monitored with
theaid ofa transmitted light microscopeat ~200x magniﬁcation. Forthecurve-neck

specimentheﬁmtwmpressivefaﬂumoccummthespecimenneckregimwhemsmss

 

 

 

0.75 >

 

'/ z \6
=>/12t~-.-

Dogbone Uniform Goes-Section Curve-Neck

 

 

 

 

 

 

 

 

 

FigureAJ-Testspecimengeomeuiesfordnesingleﬁbecompressionwsts.

198 Appendix A
is concentrated, facilitating detection of the ﬁrst failure. For the otlne two test

geometries the ﬁrst ﬁber failure could occur anywhee along their 0.75 inch long gauge
length making the detection of the ﬁrst failure somewhat more difﬁcult. The dogbone
geometry, howeve, has a major advantage ove the other two sample geometries in case
of specimen alignment. Because a dogbone specimen could be gripped at its ends, with
the grips loose the sample could be gently pulled to align the sample, then grips are
fastened tightly and tire sample is tested for compression. The ease of sample alignment
and the rigid gripping of the dogbone samples is also a major safety advantage especially
for the measurements of high compressive strengtln inorganic ﬁbes.
Theﬁbecompressive strength (ad)canbe related totheload (P)attheobservation

of tlne ﬁrst compressive event:

P -= 0.11,. +01A1- e(E,A_ +E,A,)

(n.3,)

where 0.,IMatrixstress eIFiberandMatrixstrain
A. I Matrixcross-sectiornalarea A,IFibecross-sectiomlarea
E.IMatlixcompressivemoduli EIﬁnercompressivemoduli

Equation A.l assumes that ﬁber and matrix undego the same strain deformation (e).

Equation A.1 can be rearrange to obtain:

__Ii_
Art—£34. (A.2)
5!

”or

In general, the ﬁber Cross-sectional area (A,) at most contributes by less than 0.2%

to above equation and its contribution can be ignored:
P a,

""1742

(A.3)

199 Appendix A
Equation A.3 applies to any SFC specimen geometry as long as the ﬁber and matrix

undergo the same strain deformations (i.e. no debonding occurs before the ﬁbe
compressive failure). The equation becomes invalid when debonding occurs (similarities
of the organic ﬁbe and matrix Poisson’s ratio restrain such large gap formations). The
curve-neck specimen is more sensitive to the effects of ﬁbe-matlix debonding that the
uniform cross-section test geometries. The load concentrations in the neck region of the
curve-neck specimen causes tine debonding to always occur in the neck region, wheeas,
with the uniform cross-section specimen even if the debonding occurs in some regiorns,
there may be bonded regions that could be used to monitor the ﬁber compressive failure
events.

All Study by Bazhenov er al. (1989) have shown that prior to the ﬁber failure the
tensile and compressive modulus of organic ﬁbers are similar, tlneefore, Equation A.3
isevaluatedusingﬁberandmatrixtensilemoduli. TableA.1 comparestheresults of
the SFC tests with otlne literature values for some high peformance polyme ﬁbes.
The SFC test specimen were cast witln the DERSBl/MPDA epoxy systen cured 2 hours
at 75°C followed by 2 hours at 125°C (see Chapter 2).

FortheSFCteststhedogbonesamplegeometrytend toprovidehighervaluestlnan
the curve-neck sample. For the high peformance polyme ﬁbers, detection of the ﬁrst
ldnkbandismorelaboriousinthelonggaugelengthofthedogbone samplesthaninthe
short neck region of the curve-wk samples. Therefore, the vey ﬁrst kink band
formation could be missed when examining the dogbone samples leading to slightly
highe results than those obtained by the curve-neck samples.

200 Appendix A
The SFC and Cantilever-Bending tests show similar compressive strength values for

Kevlar-49, PBO, and PET ﬁbers. Recoil, Loop, and composite tests, howeve, only
show trend similarities. Comparing arannid and coppe wire compressive yield,
Bazhenov er al. (1989) have suggested that the epoxy matrix constrain the kink formation
of organic ﬁbe failure, thus, the compressive strength of the embedded ﬁbers should be
greatetlnananisolatedﬁbe. Thehighematlixcontentthehighethisconstraining
effect, therefore, single ﬁbe composites (such as SFC and Cantilever Bending) should
yield highe compressive strength values than multiﬁlament composites or isolated ﬁber
tests (Recoil and Loop).

Table A.1 - Fibe compressive strengths (MPa) from various test methods.

 

682 :1: 152

 

545:1:118 626176

 

565 :1: 57 614 :1: 53
315 :1: 78 369 :1: 154
276 :1: 117’

 

 

 

 

 

 

 

 

 

 

' Allen (1987) 3 Detereea (1985) ’ Kurmr (1989) ‘ Tahta (1987) ’ Dml (1986)

APPENDIX B

 

Thermodynamics of Surface Tension

 

Newly formed liquid surfaces and intefaces can assume thermodyrnarrnic equilibrium
quickly, howeve, the same is not true for solid surfaces. Therefore, the changes
introduced by surface treatments on solids can result in tlnemodynannically metastable
surfaceproperties. Thereisevidenceofchangingadhesivepropertiesonagingoftreated
polyme surfaees. For cororna treated polyethylene ﬁlms Carley er al. (1978) has shown
tlnedecayoftlnepolarcomponentofsurfacefreeenegywith timethatmaybecaused
by reaction of surface oxidized species with airbornne contaminates (Kinloch 1987).
Corona treated poly(ethylene teephthalate) exhibits similar changing adlnesive propeties
that are attributed to redistribution of the surface polar groups to form internal hydrogen
bonding (Kinloch 1987). The dynamic surface propeties of treated solid substrates are
theresultofdnednemodynanﬁctendencyofdnesurfacemachieveitssmteof

equilibrium.

201

202 Appendix B
Defay et al. (1966) have presented a discussion of non-equilibrium surface tension

thatprovideinsightintotlnetlnernnodynarrnicsofsurfaceaging. Whentwophases’and
"areseparatedbyaninterfacebagenealchangeofGibbsfreeenergyofthesystemis

given by:

dc - -SdT+V’dp’+V” dp” wall-+2 u,’ drn,’ +2 u,” an,” +2: ufdrn,’ (a. 1)
i i 8

whee G I Gibbsfreeenegy S I Entropy ofthe system
TITempeature VIVolumeofthephase
pIPressureofthephase AIInterfacearea

n, I Numbe of moles of component i 7 I Surface tension
u, I Chemical potential of compornent i
Atconstanttempeamremndpressuretheconuibuﬁonofﬂneintefacemdeibbs
freeenegyofthesystem (surfacefreeenergy)is:

dG‘ - 7dr! + 227%”): (8.2)

Thecondiﬁonforspontaneouschangeatconstanttempeamnandpressumisthe
reduction of Gibbs free energy ((16 z 0), theefore, equation (3.2) shows that the
mtefacehasammmlwndmcymdimimmiuammrfacetensim,eaccumuhtelow
freeenegyspecies.
Defay also deives a genealization of the classical Gibbs surface tension equation that

is valid for non-equilibrium systems:

d7 - -s*trr - 2mm: + E e,’ dc,’ + 2e,” dc,” (3.3)
i l i

203 Appendix B
whee I“ I Surface excess of componenti = ni/A

C. I molar concentration of component i
e I cross-chemical potential of componenti = df/q
f I Helmholtzfreeenergypeunitarea
Function 5; represents the inﬂuence of concentrations of species 1' on eithe side of the
interface on the surface free energy. At equilibrium the cross-chemical potential terms
become zero and equation (3.3) converges to the classienl Gibbs surface tension

equation, thus, at constant temperature and equilibrium:

67 ' {313614 (3.4)
I

If only surfactant D accumulates at the interface, tlnen equation (13.4) becomes:

:11 - -r,,dup (a. 5)
Assuming dilute surfactant concentration Le.

dc
duo - RlencD - mail (8.6)

D

whee R I Gas constant. Then equation (3.5) becomes:

47 .- r,
32$ RTE; (3.7)

Equation (3.7) shows the reduction of surface tension when tlne surfactant accumulate
at the inteface (Atkins 1978).

APPENDDI C

 

A Novel Sample Preparation Technique for
Ion and Electron Beam Analysis of the
Fiber-Matrix Interphase

in Polymer Composites

 

Applications of ion and electron beam analysis to non-conductive and polynneic
material surfaces can be hindered by sample charging problems. Inteaction of the probe
beamwiththesarnpleinsurfacesensitivetechniquessuchasAugeElecuon
Spectroscopy (ABS), Ion Scattering Spectroscopy (ISS), and Secondary Ion Mass
Spectroscopy (SIMS) can produce charged surfaces that interfee with the analysis
process (Werne er al. 1976). A novel sannple preparation technique for ion and electron
beam analysis of the ﬁbe-matrix interphase in polymeic composite materials has been

204

205 Appendix C
developed. The proposed technique can also be adapted for analysis of othe
nonconductive materials.

Sample preparation by fracture and polishing techniques (Gabriel 1985) have seveal
signiﬁcant linnitations in composite material applications. Use of typical fracture
approaches for preparation of ﬁbrous composite material samples can result in ﬁbe
pullout and rough surface topographies that promote surface charging. Polymers are also
sensitive to surface polishing techniques and different constituents within the composite
may be left with various topographies because of diffeences in abrasion rates. The
sample preparation technique described in this report employs a diamond knife to prepare
a smooth surface that assists in preventing surface charging, with the furthe advantage
of producing highly clcan surfaces and ﬁne control for positioning of the interested area.

Our sample preparation technique is a modiﬁcation of the standard Transmission
Electron Microscopy (TEM) microtoming technique (Klomparens er al. 1986, Dawes
1971, Sawye er al. 1987). To examinne the ﬁber-matrix intephase, the ﬁber must be
embeddedinapolyme matrix. Low ﬁbervolumefractioncompositesorasingle ﬁbe
embeddedinamatrixcouponaretheappropriatesamples. Ingeneal, theemustbe
enough matrixaroundtheﬁbertoallowrazorbladetrimmingofasampleblock. With
high ﬁbe volume fraction composites (12¢. many ﬁbes pe given cross section)
preparation of the sample block is difﬁcult.

The sample block must be trimmed ﬁrst to ﬁt the instrument sample holde. The
dimensions ofa block depends on the speciﬁc sample holde used; a rectangular block
of 2x2 x8 mm is typical. Figure C.l shows an unmounted sample block and PERKIN

206 Appendix C
ELMER sample holders (fracture type) with a sample block mounted in tlneir cente. To

preparethesamplefortheholde,thesamplecanbescoredwitharazorbladeandthen
tappedtofracturethematrix. Alltherazorbladesusedinthecuttingprocessesarenew
bladesandtheiredgesareclcaned withmethanolsoalredcottonswabstoremove
contaminants. Once the sample block is formed it is cleaned in methanol and then
handledwith tweeters only.

Figure C.2 shows the major sectioning orientations possible for a ﬁbe. Radial cuts
provide circular cross-sections of the ﬁber-matrix interphase but the analysis area is
relatively small. Axialandlatealcutsprovideagreateanalysisareabutaremore
difﬁculttoprepare. Theaxialcutsaregenerallypreferredovethelateralcutsbecause
menialcuummadeparallelmdneﬁber-mauixmwrfaceﬂutWanysmeaﬁng
ofphaseboundaries.

Secﬁeﬁngorientaﬁonofdneﬁbemustbedecidedonbeforeprepaﬁngdnesample
block because it can affect the trimming procedure. In radial cuts the block face always
containstheﬁbeendanduimmingcanstanamundtheﬁbeendmgureCJA). In
axialandlatealcutstheﬁbemaybeinitiallycoveredbypolymematnixsothematrix
mustberemovedbeforereaehingtheﬁbermgureCAA). Radialsectioningis
describedﬁrst. Axialandlatealcutsrequireadditionalstepsthataredescribedlate.

For all sectioning orientations, the top of the sample block must be hand trimmed into
auapezoidofdimensionsofabouto.25 mm (ﬁgumCJQtoavoiddamagingthehnife
edgeduringdiamondknifesectioning. Totnimthetrapeeoidalbloektlnesampleisplaced

inaspecimengripmndeasecﬁmﬁngnﬁeoscopeanddnetopfaceofthesampleis

207 Appendix C
viewed at 10x to 50X magniﬁcation.

Figure C.3 shows the process of trapezoidal block preparation for the radial cuts.
First, thelocation oftlneﬁberisisolated by fourlarge razorblade marks about 1 mm
apart (Figure C.3A). With the ﬁber at the center of the blade marks the epoxy outside
the marks is trimmed at 45(de angles (Figure C.3B). Next, the matrix around the ﬁbe
is gradually trimmed to a shallow depth (~0.5 mm) until two parallel edges and two
angled edges about 0.25 mm apart around the ﬁber form the trapezoidal block (Figure
C.3C). Razor blade cuts are always made at 45° and away from the trapezoidal block
since otherwise a crack can innitiate that fractures the small trapezoidal block at its base.

Figure C.4 shows the process of trapezoidal block formation for the lateal cuts. The
ﬁbeisirnitiallycoveed bythematnixa‘igureCAA). Thematnixovetheﬁbeis
removed at a shallow angle (~ 10°) until a ﬁber portion is exposed (Figure CAB). The
regionofthinmatrixcoverageadjacenttotheexposedﬁbeendistheregionfor
trimming the trapezoidal block. This trapezoidal block is prepared in a similar procedure
astheradialcuts. Thethinmatrixovetheﬁbercanberenovedduringtlnediamond
sectioning or partially left for sputte deptln proﬁling through the ﬁbe-matrix intephase.
Thesameprocedureisapplied fortlneaxialcuts.

Before the ﬁnal trimming of the trapezoidal block face with a diamond knife, the
sannpleblockisremoved fromtlnespecimengripandisplaced (faceup)inagoldplasma
coate. The block is covered with a tlnick gold coating (~ 100 nm). This coating is
important in alleviating charge buildup on the analysis surface. The gold coated sample
blockisnowready fortheﬁnalsurfacetrimmingbyadiamondlnnifeandisplacedback

208 Appendix C
in the microtoming setup.

Sectioning of high peformance ﬁbes requires a diamond krnife. A solution of
deionized wate with a low concentration of acetone (~1 droplet of acetone pe 20 nnl
ofwate) isusedtoﬁlltlnediamondlmifeboat. Acetonereducesthecontactangleofthe
watetoproducebetterwetting ofthekrnifeedge. Thetrapcroidal shapeoftlnesample
blockisselectedtospecifythediamond knifecutting directionwhichisfromthewide
base edge to the opposing parallel edge. The trapezoidal block is oriented so its wide
base edge contacts the knife edge ﬁrst. The knife is advanced manually toward the
tapezoidalblockfaceuntilreﬂectionoftheblockfaceoffthewatermeniscusis
observed. The krnife is then slowly advanced toward the block at 1 pm steps until cutting
begins. Several 1 pm sections are cut to remove artifacts from the razor blade trimming.
Thethiclmesssettingisadjustedtocutsevealultra-thinsections(50to100nm)to
prepare a smooth and clean ﬁnal surface.

At the completion of this stage a small, ﬂat, and clean surface for analysis having
gold coated boundaries is obtained. Every analysis area is at most 0.12 m away from
theconductiveboundaries. Formostmaterialsthisisenoughtoeliminatetlnecharging
problems, howeve, iftlnechargingisstillpresent,aﬁnegoldcoating(<1nm)canbe
applied onto the surface.

Two examples of sample preparations are illustrated. All sample preparations wee
carried out on a Reichet-Jung ULTRACUT E nnicrotonne and setup for analysis on a
PERKIN ELMER PHI 660 Scanning Auge Multiprobe. ﬁgure C.5 shows SEM
micrographsofanaxially cutnickelcoatedcarbonﬁbeembeddedinanepoxy matrix.

209 AppendixC
Thecutwasmadeat6°anglefromtheﬁberlongaxisproducinganoblongﬁbe

cross-section. Figure C.5A shows the trapezoidal block face. The block face still has
some of its initial gold coating covering its right half because of the angled cutting
direction. Figure C.SB shows the oblong ﬁber cross-section. For this sample, line-scan
or nnapping analysis can be carried out at the ﬁbe—matrix intephase, or, the adjacent
matrixovertheﬁberontheadvancing sideofthesectioningdirection (rightside) could
be the site for sputter depth proﬁle analysis. Figure C.6 slnows a SEM micrograph of
a copper coated carbon ﬁber radial cross-section.

210 Appendix C

   

Figure C.l - PMIN-ELMER fracture stage sample holders. A sample block is shown
placed at the center of the holders and another sample block isshown from its side view.

/

/v<—

 

 

 

/ /

L“ V“ VL’

Radial Cut Axial Cut Lateral Cut

 

 

 

 

 

 

 

 

Figure C.2 - Major sectioning orientations of a ﬁber.

211 Appendix C

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

blade marks 0.25 mm
I x ‘7
[I
./ _/
(A) (B) (C)

Figure C.3 - Trapezoidal block preparan'on for radial cuts. (A) Blade markings on the
face isolate the ﬁbe location. (B) Matrix around the ﬁbe perimeter is trimmed off. (C)
A shallow trapezoidal block is trimmed around the ﬁbe.

] Areaofaectlonlng

 

 

 

 

/’9§7
/ J' J

(A) (B) (C)

Figure C.4 - Trapezoidal block preparation for lateal cuts. (A) The ﬁbe is initially
coveed. (B)'I'hematrixistrimmeduntilaﬁberportionisexposed.(C)Theblockis
forrnedintlneregionadjacenttotlneexposedﬁbeend.

 

 

 

 

 

 

 

 

 

212 Appendix C

 

Figure C.5 - SEM micrographs of an axially cut nickel coated carbon ﬁber-epoxy
composite. Cut was nnade at 6° angle from the ﬁber long axis. (A) Trapezoidal block
face. (B) Oblong ﬁber cross section.

n————+
10.0w BJZJkX 5.0a

 

FigureC.6-SEMmicrographofaradiallycutcoppecoatedcarbonﬁber.

APPENDDI D

 

Unsteady Diffusion in a Semi-Inﬁnite Slab

 

To help to understand the mass-transfer phenomena affecting the penetration depth
of sulfur into polycarbonate, tlne general concept of unsteady diffusion in a semi-inﬁnite
slab is reviewed here. This discussion is a modiﬁcation and expansion of the topic as
presented by Cussler (1984).

For species (1), concentration (C,) and flux 0,) as a function of position (z) and tirrne
(t)canberelatedbyamassbalanceonathinlayeofareaAandthicknessAz,

(accumulation in volume M2) = (rate of diffusion into the laye at z)

- (rate of diffusion into the laye at z+Az)
+ (amount produced by reaction in AM)
in mathematical terms, this is,
{Emits-re) - rnj,|z - ALL,“ + rlAAz (13.1)

whee r, is the rate of production per volume of species (1). By rearranging and using

213

214 Appendix D
the deﬁnition of a derivative equation D.1 becomes,

SCI 8 _ Sj, D 2
TE T2 + r1 ( )
The Fick’s law of diffusion states,
dC
'jn ' D721 (003)

where D is the diffusion coefﬁcient. Combining equation 11.3 with equation D.2,

6c, :Dazc, + an ac, +1.
‘5? 572 231; i (”-4)

For the case whee diffusion coefﬁcient is independent of position, we get,

incl 62c,
TE ‘ D 62’ i r‘ (”'5’

 

When the reacting solute is present in two forms, the mobile form (1) and the
irnmobilereactedforrn (2),andthereactionisfasterthandiffusion, then,aﬁrstorde
reaction can be represented as,

c2 = xc, (0.6)
whee C, is the concentration of the reacted and immobile solute, and K is the

equilibrium constant of reaction. A similar mass balance on species (2) gives,

'6'? - -r, (0.7)

Adding equations D.5 and D.7 in combination with equation 13.6 results,

set D 62c1
":17 3175—27 (”-3)

215 Appendix D
The boundary conditions of this equation are,

t= 0, C, =0 forallz (D.9)
t>0, C1=Cli atz=0
Cj= atz=00

where C“ is the solute concentration outside the slab. The differential equation D.8 has
been solved using the metlnod of ”combination of variables.‘ A new variable I is
deﬁned,

t = — z
0.10
[4.1.1: ( )

 

 

1+K

The diffeential equation D.8 is then rewritten as,

35'; era-5;! -o (0.11)
The boundary conditions D.9 become,
2 =0, {=0 C‘ =C“ (D.12)
z=oo,ort=0 {=09 C,=0
Integration of equation D.ll results,
%% . ae’rz (0.13)

whee a is an integration constant. A second integration with the use of boundary
conditions D.13 gives,

cu ' Cr

- er! I = art _.
Cit

 

 

 

(0.14)

 

216 Appendix D
whee, , 2
equation D.15 shows that for any ﬁxed concentration of solute the position of that
concentration is proportional to the square root of time. For example, for the position
inside the slab where C, is only 1% of the Cu (i.e. erf 1' = 0.01), the value of {can be

read from an 'eror function” table, thus,

 

 

 

 

§=~2.33=; z
(0.16)
4 D t
1+1:
- D (0.17)
a - ~20 4 t
z [ 33! 1+Kin/—

Equation D.l‘7 shows that when the diffusion coefﬁcient is concentration independent
tlnen the penetration depth of the diffusing phase is linearly proportional to tine square
root of time.

For the case where diffusion coefﬁcient is concentration dependent, equation D.4 can

berewritten as,

ea, 3 Daze, + an re, ta, + r
‘6'? 7;: nan-2'1? ' (1)-18)

so, 620, an 6c, 2
a» 8 __ '0-
TE 0622 Tc,(Tz) +r, (0.19)

a model for concentration dependency of diffusion coefﬁcient can be assunned (e. g.
WLF model) and equation D.l9 can then be solved numerically by ﬁtting expeimerntal
data such those presented in Figure 6.8.

APPENDDI E

 

Sulfonation Treatments of High Density

Polyethylene Gas Tanks

 

Three samples of blow molded high-density polyethylene gas tanks that contained
activated carbon, wee sulfonated with ~12 vol% SO,/N, gas phase treatments for 10
to 15 minutes. The samples were subsequently neutralization with tlnree diffeent catiorns,
chromium (sample #2), calcium (sample #33), and coppe (sample #46). Sample titration
showed 1200 pg of SO, pe square inch for all samples. Table 13.1 lists the results of
XPS atomic concentration and ABS elemental depth penetrations for the three examined
samples. Figure E.l shows the elemental line scans supeimposed on the SEM
rrnicrographs of the treated samples.

217

218 Appendix B

Table E.1 - XPS atomic% composition and AES elemental line-scans penetration depths
for the sulfonated high-density polyethylene gas tanks samples.

 

 

 

Chromium Calcium Copper

XPS Atomic% °

C 38.4 1: 4.0 50.0 d: 0.4 28.4 i 1.2

O 48.8 i 3.1 38.5 :l: 0.4 11.1 :1: 1.4

S 1.69 :1: 0.07 7.42 :l: 0.02 25.8 :I: 0.4

N - 1.14 i 0.05 -

Cr 9.94 :i: 1.22 - -

Ca 1.20 :1: 0.23 2.89 :1: 0.03 -

Cu - - 34.7 1; 0.7

AFE line-scan

S 15.5 :I: 3.9 (4) 15.2 1; 3.0 (6) 15.3 i 0.9 (5)

Cr 3.9 :1; 0.9 (6)

Ca 7.0 d: 0.6 (6)

Cu 2.5 :t 0.2 (4)

 

 

 

’ aveage of two runs.
Mamber in parenthesis represents the number of runs averaged.

Appendix E

7.21an 4.ak>< 1am
Sample #33, Ca treated [

)
I
I
I

F—-=—————-%l
. 7.8kV 4.0kX 10.0Mm‘
- Sample #48

Figure 13.1 - SEM micrographs of sulfonated samples and their elennental line-scans.

 

APPEADDI F

 

Procedures for Ultra-Thin Microtomy of

Fiber Reinforced Composites

 

F-IWQN

AnessenﬁalpanofuansmissionelectronmicroscopyCI'EM)isﬂneulua-ﬂnin
nnicrotoming ofsarnples. TheTEMsamples mustbethinenoughtotransmitsufﬁcient
electronstoformanirnage. Thesamplesmustalsobestableundetlneelectronbeam
andinahighvacuum. TEMimagesarefornrnedbycombinationofelastic,inelastic,and
absorptioninteactionsoftheelectronbeamwitlntlnesample. lrncreasingthesannple
thiclmess ineeases-tlnebeamabsorpﬁonandreduwstheimageresoluﬁonandsample
stability. ForalOOkVelecuonbeunthepmcticalspecimenthicknessislinutedtoIOO
nm.

Klomparens er al. (1986) have described the microtomy techniques for the biological
materialsandMalisetaI. (1990) havereviewedtlneultramicrotomy techniquesfor

220

221 Appendix F
metallic and ceannic nnaterials. Howeve, sectioning of high performance reinforcing

ﬁbes requires additional considerations than otlne materials. The epoxy resins used in
composite are more brittle than tlnose used in biological applications but more ductile than
ceamic materials. In this report a procedure for ultra-thin microtomy of ﬁbe-matrix
interface is described.

film

To produce ultra-tlnin sections of reinforcing ﬁbes, they must be ﬁrst embedded in
a polyme matrix. Low ﬁber volume fraction composites or single ﬁbes embedded in
amatrixcouponaretheappropriatesamples. Ingeneal,tlneremustbeenoughmatrix
around the ﬁbe to allow formation of a sectioning block. \Vrth high ﬁbe volume
fraction composites trimming the sectioning block is difﬁcult.

Topreparethesectioningblock, thesampleisplacedinasampleholder. The
dimensions of a sample depends on the speciﬁc sample holde used; a rectangular block
of3x10mmistypical. Topreparetlnesampleforitsholder,tlnesamplecanbecutby
arazorblade,normallypositionedandthentappedtofrachnretheparts. Alltherazor
bladesusedinthecutﬁngpmcessesarenewbladeandtheredgeamcleanedwith
ethanol soaked cotton swabs to remove contanninant.

Orientationoftlnesectioning mustbedecidedonbeforepreparingthesample block
because it can affect the trimming procedure. Figure C.1 shows various sectioning
orientationsofaﬁbe. Inradialcutstheblockfacealwayscontainstheﬁbeendand
trimmingcanstartaroundtheﬁbeend. Inaxialandlateralcutstheﬁbemaybe

222 Appendix F
initially coveed by polyme matrix and the matrix must be trimmed before reaching the

ﬁbe. Radialsectioningisdescribedﬁrst. Axialandlateralcutsrequireadditionalsteps
thatwillbedescribedlate.

Thesectioningblockmustbehand trimmedintoatrapezoidofdimensionsnolornge
than 0.25 mm before diamond lmife is used. Thedimensions and quality of the trapezoid
blockdeterminestlnequalityoftheﬁnalsections. Totrimtlnetrapezcidblock,the
sarnpleholdeisplacedundeasectioningnnicroscopeandthetopfaceofthesampleis
viewed at 10x to 50x magniﬁcation. The top sample face may have to be removed
before trimming the trapezoid block. This is to remove the artifacts of the blade cuts or
toapproachtlneﬁberfortlnelateralandaxialcuts.

FigunCJshowsdneprocessofsecﬁoningblockpreparaﬁmfordneradialcum.
First,tlnelocationoftheﬁbeisisolated byfourlargeblademarksaboutlmmapart
(FigureCJA). Mdntheﬁberatthecenteoftheblademarks,theepoxyoutsidethe
marksaretrimmedat45°angles(FigureC.2B). Next,thematrixaroundtheﬁbeis
graduanynimmedwiﬂnashanowdepthunﬁltwopamlldedgeandtwoanglededge
about0.25mmapartaroundtheﬁbeareobtained(FigureC.2C). Itisimportantto
havethetwoparalleledgesasparallelaspossibletoobtaingoodribbonformationduring
tlnemicrotoming. Cutsarealwaysmadeat45°andawayfromthetapezoidblockface
sinceoﬂnewiseacrackcaniniﬁatednatﬁachesoffthesmaﬂﬁapeaoidbloek. Theﬁnal
trapezoid block should be less than 0.25 mm tlnick since tlnicke blocks are more
susceptible to vibration than the shallowe blocks.

Foraxialandlatealcutstheﬁbeisinitiallycoveredbythematrix(FigureCJA).

223 AppandixF
Tlnematrixovetlneﬁberisremovedatashallowangle(<10°)tillaﬁbeportionis

exposed(Figure C.3B). Theregion ofthinmatrixcoveageovetheﬁbeistheregion
for trimming the trapezoid sectioning block. The trapezoid sectioning block is prepared
inasimilarprocedureastheradialcuts. Thethinnnatrixovetheﬁbeisremoved
duringthediamondsectioning.

Ultra-thin nnicrotoming of high peformance ﬁbers requires a diamond knife.
Sectiorningoftlnetrapezoid blockfaceshouldstartfromadulllmifeedgeandoncethe
trapezoid block face is ready for ultra-thin sectioning the face is sectioned with a sharp
lmife edge. The mounting angle of the knife varies depending on the krnife edge angle
set by manufacture (~55°) and type of sample being sectioned. A solution of
de-ionizcdwatewithlowconcentrationofacetone ~1dropletofacetonepe200cof
water)isusedto‘ﬁllthediamondknifeboat. Acetonereducestlnecontactangleofthe
watetoproducebettewettingoftheknifeedge. Theknifeboatisﬁlledusinga
syringe. Initially,theboatinﬁlledtobrinkofoverﬂowandthenexcesswaterisdrawn
offtoforrnaconcave ﬂuid surfacebehindthediamond edge. Thisprocedureinsures
good wetting of the knife edge. The lmife is advanced manually toward the block face
untilreﬂectionof‘tlneblockfaceoffthewatemeniscusisobseved. museum
slowly advanced toward the block at 1 am steps until cutting begins. Seveal 1 pm
seeﬁonsamcutbremovednehimmhguﬁfacmmprepamaclcanunoothfacefeune
ultra-thin sectioning. Thelmifeistlnenwithdrawnandshiftedtoasharplmifeedge.
Again the lmifeis slowly advanced until tlnin sections of gold orpurple colorare cnnt (0.2
to 1 am). Motorized sample advancingistumedonandacuttingspeedofOAO mm/sec

224 Appendix F
is selected. The tlniclmess setting is adjusted to obtain silve-gold colored sections (50
to 100 nm).

In a successful ultra-tlnin sectioning, the sections remain attached and form ribbons
that ﬂoat on the water. To manipulate ribbons, an eyelash applicator is prepared. An
eyelash applicator is simply an eyelash mounted on a wooden applicator using a drop of
nail-polish. With the eyelash applicator, the ribbons are assembled for the pick up by
a TEM grid.

TocollectTEMsections, ﬁrsttlnelmifeisrecededandwaterisaddedtotlneboat.
The ribbons are tlnen arranged away from the diamond edge. The grids used are 200 or
300meshﬁnewirecoppergridstlnathavesnnallendtabsforeasypickup. Agridis
picked upbyatweezerandpassed oveaﬂametoburn offits hydrocarbon contaminates '
and increase its wetting. Under the microscope oftlne nnicrotome a grid is submeged
inwaterandapproachestheassembledribbonsfromundeneaththesurface. Theribbons
arepickedupandtlnegridisdrainedonaﬁlterpape. Thegridsarestoredina

dessicator for late sample staining and carbon coating.

0

/\__27“‘"'

/

 

 

 

/ /

v v
Radial Cut Meal Cut /Lateral Cut

Figure F.l - Major sectioning orientations of a ﬁbe.

 

 

 

 

 

 

 

 

225 Appendix F

/ Area of sectioning
w il— a

/ / /

(A) (B) (C)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure F12 - Trapezoidal block preparation for radial cuts. (A) Blade rrnarkings on the
faceisolatetheﬁbelocation. (B)Matrixaroundtheﬁbepeimeteistrimmedoff. (C)
A shallow trapezoidal block is trimmed around the ﬁber.

 

I .9":

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(A)

Figure F.3 - Trapezoidal block preparation for lateal cuts. (A) The ﬁbe is irnitially
coveed. (B)Thematrixistrimmeduntilaﬁbeportionisexposed. (C)Tlneblockis
formedintheregionadjacenttotheexposedﬁbeend.

APPENDDI G

 

Wilhelmy Program

 

c.
c.
C.
c.

Linear ragroaaion program to determine aurtaca anargy componenta
from contact angle maaaurcmanta

integer I,J,R,N,DONB,NUH(10)

roal SLOP,YINT,SP,SD,A,B,C,TSB,STDTSB,STDSL,8TDYI
rcal SUHX,SONY,SUHXX,SUMYY,SUHXY,SUHXH

real uaaux,xx111,xxrz,xx122,sz,anc(zoon,rtaoon,x(2oon
raal HBAN(10),STD(10)

charactar*99 nararntno),uana(1on,narnour
charactar*1 QUE

DON! I 0

print *, ' ENTER THE OUTPUT FILE RAM! 8 '

read (*,'(an') naraour

OPBR (unitIl,filo-DATAOUT,atatuaI'unknown')
rewind l

DONE I 0

do NEIL! (DON! .lt. l)

x I 0

print *, 'now many liquids? '

road *, I

do 100 JIl, I

print *, 'w - Watar'

print 9, 'l - Ethylene Glycol'
print *, 'r - Formamida'

print *, 'u - Methylene Iodida'
print *, 'B - Buxadacana'

print *

print *, 'Choao liquid (w,a,r,u,n): '
20.6 (*.'(A)') 903
I! (one .oq. 'w') than

AI 72.8

B- 21.8

CI 51.0

ﬂaunts) I 'ﬂatcr'
3L8!!! (90! .oq. 'a') than

n- 48.3

BI 29.3

CI 19.0

NAHB(J) I 'lthylana Glycol'
ILSBI! (QUE .oq. '2') then

R- 58.3

B- 32.3

CI 26.0

226

50

200
250
100

300

350

400
410

227 Aqnxuuﬁx13

RAMB(J) I 'rormamida'
3L8!!! (QUE .cg. 'h') than
AI 27.6
BI 27.6
CI 0
NAME”) I 'Baxadacana'
8L8!!! (90! .ag. 'm') than
AI 50.8
B- 48.4
c. 20‘
RAH3(J) I 'Hathylana Iodida'
ELSE
GOTO 1
andif
writ. (*,50) naun(an
format ('lntar ',al7,' data til-r ')
road (*,'(A)') DATAIN(J)
print *
OPBN(unitI2,filo-DATAIN(J),atatuaI'unknoun')
rewind 2
DO 200 iIl, 1000
READ (2,*,andI250) ANG(i)
K I K‘O’l
Y(X) I (A*(l+OOS(ANG(i)*0.0l74533)))/(2*((B)**O.5))
x(x) I (C/B)**0.5
NUH(J) I i-l
CALL STAT (ANG,NUH(J),HBAN(J),8TD(J))

suux - o
suuxx - 0
son! - o
sour! - o
coax! - o
souxu - 0
do 300 3-1, x
suux - suux + x(J)
suaxx - suuxx + (X(J)*X(J))
sour - sun! + 2(a)
sour! - suns! + (Y(J)*!(J))
suuxr - suuxr + (X(J)*!(J))
uaaax - SUHX/x
do 350 3-1, x
suuxu - suuxu + (x(J) - uaauxntea
xxrrn - suuxx / (x-suuxu)
xxra - -suux / (xtsuuxu)
xxraz - 1/suuxu
rrar - (xxrnresuur) + (xxrztsuuxr)
anon . (xxrzasuu!) + (xxrzztsuux!)
S2 - (sour! - ((rrurtsuuxn + (stop-suux!)))/(x~2)
sworn - (xxrnntszntto.s
srnsn - (xx122t32)-*o.5
sr - zesnortsrnsn
an - zerrurtsrnrr
rsa - (rrurtezn + (snorttzn
srnrsa - so + as

no 400 jIl, I
write (t,snon narn1u(sn,uaaa(J),uaan¢a).srntan,uuu(an
write (1,410) nararu(J),naua(J),naau(an,srn(an,nuu(J)
format (alO,a16,' contact angles ',£6.2,' +- ',£4.2,' (',i2,')')
print *
write (l.*)

420

430

440

450

460

10

20

228

writa p.420) unmsrnnrr

writa (1,420) YINT,8TDYI

format ('Sqrt Diap. Comp. I ',f6.3,' +- ',f5.3)
writa (*,430) SLOP,STDSL

writa (1,430) SLOP,8TDSL

format ('Sqrt Polar Comp. I ',f6.3,' +- ',f5.3)
print *

writa (1,*)

writa (*,440) (YINT)**2,SD

writa (1,440) (YINT)**2,SD

format ('Diap. Comp. mN/m I ',f6.2,' +- ',f5.2)
writa (*,450) (snop)**2,sr

urita (1,450) (SLOP)**2,8P

format ('Polar Comp. mN/m I ',f6.2,' +- ',f5.2)
print *

urita (l,*)

writa 0,460) ISLSTDTSI

urita (1,460) rsa,srnrsa

Aqnxmuﬁxts

format ('Total aurfaca anargy mN/m I ',f6.2,' +- ',f4.2)

print *
writa (1,*)

print *, 'DONB 7 YIN : '

tOId (*p'thl') 003

I! (90! .aq. 'Y'.or. qua .aq. 'y') than
dona I 1

INDIP

IND DO

END

SUBROUTINI STAT (AVB,K,MEAN,8TD)
RIAL VIR,AVI(200),HEAN,8TD
INTEGER J,R

sou - o
no no J-1.x
sun - sun + ava(J)
aaan . sou/x
van - o
no 20 J-1,x
van - van + (ava(J)-aaaN)**z
are - (van/(x-n))*-o.s
ano

APPENDDK H

 

Experimental Data

 

rluorintad Ravlar-49

I Of Braaka
0ntraatad-175 25 18.8000 1
ap-a 12 18.5833 1
ap-b 14 21.7143 1
ap-c 12 18.8333 1
ap-d 12 19.6667 1
Diamatar
0ntraatad.dia 30 12.6853
a.dia 16 12.0938
b.dia 16 12.7963
c.dia 15 12.3787
d.dia 16 12.2425
Tanaila Strangth
k49-javid.dia 30 2.8770
a.gpa 16 2.9935
b.gpa 14 2.6078
c.gpa 14 3.0221
d.gpa 16 3.1175
188
21L! RAH! I or BRRARS
xun75a0.nar 18.8 t 1.1 (25)
AP-B 21.7 t 1.8 (14)
AP-c 18.8 t 2.5 (12)
AP-D 19.7 t 1.4 (12)
Compraaaiva Strangth
UNTREATID 107 752 1 168
AP-A 10 742 t 132
AP-B 10 713 t 167
an-c 10 767 t 104
AP-D 15 862 t 162

(t-taat)

1.0301
1.6765 6.1.- 35 1 - 0.476
1.7723 3.1.- 37 1 - 6.400
2.5166 3.1.- 35 t - 0.057
1.3707 6.1.- 35 t - 2.093
1 0.5970
1 0.5599 6.1.- 44 e - 3.269
1 0.4273 6.1.- 44 1 - 0.657
1 0.6442 3.1.- 43 t - 1.533
1 0.6239 6.1.- 44 t - 2.359
1 0.3167

1 0.2637 6.2.- 44 1 - 1.255

1 0.3765 6.1.- 42 t - 2.473

1 0.3533 3.1.- 42 t - 1.364

1 0.3531 a.s.- 44 t - 2.357

L.(luu laid r (HP-1

1.17 97.5 15.62 1 0.90

1.13 93.7 15.27 1 1.33

1.01 34.4 16.49 1 1.35

1.17 97.3 16.00 1 2.14

1.12 93.2 17.07 1 1.19
a.r.- 120 t - 2.336

229

230 ntppendbtll
coupling aganta
384 L. (WI) 14/6 1’ (UPI)
175aa4md 266.3 t 74.8 (572) 34.6 t 9.7 84.73 1 23.81
KR-A84-L.Lc 266.1 a 79.8 (202) 34.6 a 10.4 84.79 a 25.43
la-aa4-h 304.9 a 96.7 (189) 39.6 1 12.6 73.99 a 23.47
la-aa4-1 302.7 1 96.6 (200) 39.3 t 12.5 74.53 a 23.77
Ravlar-49 # or nnnans 1., (nan) 1../d 1 (MPa)
ku175md 18.8 t 1.1 (25) 1.17 97.5 16.97 t 0.98
kr551 17.7 t 0.8 ( 7) 1.24 103.5 15.99 1 0.68
kr55h 19.0 t 1.1 ( 8) 1.16 96.5 17.15 1 0.97
12371 19.0 t 1.3 ( 8) 1.16 96.5 17.15 1 1.18
la37h 17. 1 1.4 ( 8) 1.23 102.6 16.14 t 1.22
r I 10.509
1! r > r (tabla) than thara ia anough data for a difararnca
Untraatad ribara
e or anaaxs 1.. (an) r../d r (IlPa)
TlCHNORA-AR 36.3 t 2.2 (15) 0.607 47.7 32.28 1 1.95
TlCBNORA-waahad 28.3 a 2.5 (15) 0.776 60.7 25.22 1 2.25
RZ9-175HD-AR 24.0 t 3.3 ( 8) 0.917 72.2 19.95 1 2.74
x29-175HD-Waahad 24.4 1 3.7 ( 7) 0.901 70.9 20.31 t 3.11
R49-175HD-AR 18.8 t 1.1 (25) 1.170 92.1 15.63 1 0.90
PBO-AR 10.7 1 0.8 (14) 2.053 102.7 16.56 1 1.28
PBOIWaahad 10.8 t 1.6 ( 9) 2.041 102.1 16.66 1 2.42

Dogbona SFC taata
Ravlar-49 ar
Ravlar-29 ar
PBO-ar
Tachnora-ar
Tachnora-waahad

643.200 1 95.053 (
502.000 1 61.059 (
401.000 1 31.646 (
610.625 1 51.264 (
613.125 1 57.275 (

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)

Sacond Hodulua (GPa)
Fractura Strain (t)

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)
Practura Strain (t)

Diamatar (um)

Tanaila Strangth (GPa)
Modulua (GPa)

Sacond Modulua (GPa)
Fractura Strain (t)

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)
Fractura Strain (t)

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)
Fractura Strain (5)

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)
Fractura Strain (t)

Spactra-1000 untraatad

Diamatar (um)

Tanaila Strangth (GPa)
Hodulua (GPa)
Fractura Strain (t)

Spactra-1000, 400 kaV 10

Diamatar (um)

Tanaila Strangth (GPa)
Aodulua (GPa)
tractura Strain (t)

231

Tanaila Propartiaa

Al ROG. Ravlar-l49

 

13.01 t
2.318 2
145.6 t
108.5 2
1.678 t

0.524 (11)
0.322 (10)
5.423 (10)
5.137 ( 9)
0.307 (10)

Aa Rac. Tachnora

 

13.23 t
3.201 2
66.87 t
4.880 t

0.912 (11)
0.246 (11)
3.090 (11)
0.260 (11)

Combinad Ravlar-149

 

12.97 t
2.376 2
147.9 2
108.6 2
1.720 t

0.459
0.252
5.607
6.203
0.229

(21)
(20)
(2°)
(19)
(20)

Aa Rac. PBO

 

19.40 1 1.322 (15)
2.347 1 0.461 (15)
165.9 1 11.729 (15)

1.787 1 0.294 (15)
P80 crit. point driad

 

20.75
2.088
78.16
2.439

t 2.491
t 0.599
t 9.252
t 0.292

(10)
(10)
(10)
(10)
w-PBO Adipoyl (PCA3.1)
a 3.09 (10)

1.24 1 0.32 ( 8)

100.3 1 16.2 ( 3)
1.30 1 0.23 ( 3)

 

20.77

Pull aat

 

29.5
2.56
64.5
7.62

1 3.3 (12)
1 0.43 (11)
1 11.9 (14)
1 2.00 (11)

" Ita‘lcan2 ion implantad
rull aat

 

28.7
1.66
77.0
3.94

1 2.5 (14)
1 0.30 (12)
1 12.5 (11)
1 0.73 (13)

nappendhrll

Waahad Ravlar-149

 

12.94 t 0.400
2.434 t 0.150
150.3 2 4.963
108.7 1 7.279
1.762 t 0.110

(10)
(1°)
(1°)
(1°)
(1°)

Waahad Tachnora

 

12.92 1 0.734 (11)
3.195 1 0.211 (11)
70.00 1 2.455 (11)
4.990 1 0.320 (10)

Combinad Tachnora

 

13.08 t
3.198 2
68.44 t

0.323 (22)
0.224 (22)
3.159 (22)

4.932 t 0.288 (21)

“ht-PBO RT driad

 

20.05 t
2.949 2
105.6 2
3.245 t

3.453 (10)
0.309 ( 3)
5.459 (10)
0.791 ( 3)

abnao Sulfonatad (331)

 

25.17
1.492
64.38
2.327

1 3.265 (10)
1 0.357 (10)
1 5.479 (10)
1 0.433 (10)

Corractad

 

23.7 1 2.7 (11)
2.65 1 0.29 (10)
67.3 1 9.22 (12)
6.91 1 1.30 ( 9)

Corractad

 

27.8 t 1.3
1.77 t 0 2
82.5 t 3.0
3.42 t 0 4

 

 

 

 

 

232 nAppendthI
Ion Implantad ribara

Kavlar-49 Diamatar (um) Tan. Str. (GPa) Comp. Str. (MPa)
Untraatad 12.59 1 0.61 (37) 2.88 1 0.31 (37) 752 1 168 (107)
30 nt 10“ 2.91 1 0.34 746 1 164
30 8+ 4x10“ 2.32 1 0.30 743 1 140
30 at 10” 2.93 1 0.31 706 1 123
75 Art 10” 2.74 1 0.20 690 1 96
100 11* 10“ . 2.77 1 0.36 745 1 161
100 nt 2310“ 12.22 1 0.31 (10) 2.77 1 0.36 (10) 320 1 34 (10)
100 nt 10“ 12.75 1 0.66 (11) 2.59 1 0.30 (10) 777 1 107 (10)
390 8* 2x10" 12.33 1 0.39 (12) 2.67 1 0.27 (10) 723 1 61 (9)
400 n+ 5x10“ 12.45 1 0.26 (12) 2.72 1 0.25 (11) 750 1 70 (10)
400 nt 2x10” 12.52 1 0.32 (14) 2.46 1 0.17 (11) 735 1 104 (10)
400 8+ 10“ 12.73 1 0.73 (10) 2.23 1 0.13 (10) 743 1 33 (9)
390 8* 2x10“ 12.46 1 0.29 (19) 2.27 1 0.13 (20) 376 +- 66 (3)
390 36* 10” 12.41 1 0.46 (11) 2.50 1 0.32 (11) 713 +- 133 (3)
ribar I or BREAKS 1., (nan) I../d 7 (MPa)
Untraatad 13.3 1 1.1 (25) 1.170 92.9 15.49 1 0.39
30 nt 10“ 13.6 1 2.1 ( 7) 1.135 94.1 15.30 1 1.71
30 nt 10"old 17.9 1 1.1 (13) 1.227 97.5 14.7 71 0.92
100 nt 10” 19.0 1 2.2 (11) 1.153 92.0 15.06 1 1.77
100 3* 10"old 17.2 1 0.9 (10) 1.279 101.6 13.63 1 0.73
100 3* 10“ 17.5 1 1.4 (11) 1.260 100.1 12.94 1 1.07
100 8+ 10“old 17.0 1 0.9 (10) 1.294 102.3 12.60 1 0.70
100 rt’ 10"old 16.0 1 1.2 (10) 1.375 109.2 11.36 1 0.92 noon
400 at 10'2 13.3 1 1.5 (10) 1.170 92.9 14.36 1 1.13
400 3* 5x10" 22.4 1 0.3 (10) 0.932 73.0 17.43 1 0.66
390 36* 10” 21.4 1 1.1 (11) 1.030 31.3 15.23 1 0.30
400 nt 10” 22.4 1 2.1 (20) 0.932 73.0 15.77 1 1.49
400 at 10“ 19.5 1 2.3 (13) 1.130 39.3 12.70 1 1.32
400 nt 10“ 19.9 1 2.3 (23) 1.107 37.9 12.91 1 1.50 “"-
400 11* 10"old 21.0 1 1.9 (10) 1.048 83.2 13.70 1 1.23
400 I" 10“old 21.0 1 1.4 ( 8) 1.048 83.2 13.70 1 0.92 MPDA

233

Appendix H

Ion Implantad Ravlar-49 XPS DATA (t-taat)

ion/kco 19.810 1 .414 (13)

ion/R120 19.083 1 .370 ( 3) d.f.I 14 t I 2.783

ion/ken 7.397 1 .257 (13)

ion/k12n 7.573 1 .145 ( 3) d.f.I 14 t I 1.126

icn/kcc 72.785 1 .440 (12)

ion/kl2c 73.343 1 .515 ( 3) d.f.I 13 t I 1.912

ion/kco 19.810 1 .414 (13)

ion/k512o 20.290 1 .529 ( 4) d.f.I 15 t I 1.911

ion/ken 7.397 1 .257 (13)

ion/k512n 7.320 1 .156 ( 4) d.f.I 15 t I .559

ion/kcc 72.785 1 .440 (12)

ion/k512c 72.390 1 .589 ( 4) d.f.I 14 t I 1.437

ion/kco 19.810 1 .414 (13)

ion/R130 18.515 1 .114 ( 4) d.f.I 15 t I 6.064

ion/ken 7.397 1 .257 (13)

ion/k13n 7.610 1 .181 ( 4) d.f.I 15 t I 1.526

ion/kcc 72.785 1 .440 (12) .

ion/k13c 73.872 1 .259 ( 4) d.f.I 14 t I 4.615

ion/kco 19.810 1 .414 (13)

ion/R140 16.207 1 .080 ( 3) d.f.I 14 t Il4.644

ion/kcn 7.397 1 .257 (13)

ion/k14n 6.423 1 .156 ( 3) d.f.I 14 t I 6.189

ion/kcc 72.785 1 .440 (12)

ion/k14c 77.370 1 .079 ( 3) d.f.I 13 t I 17.491
If t > t(tab1a) than thara ia anough data for a difararnca

vat-PBO Diamatara (Vidao calipara)

 

 

rila Condition taat madium (um)

pbo-l vat watar 26.9 1 4.1 (30)
pbo-2 ona day RT driad dry 18.8 1 3.8 (30)
pbo-3 2 waaka in neon BtOB 26.0 1 4.6 (30)
pbo-4 1d neon, 1d Act, 1d Bra, 5hr RT dry dry 21.7 1 2.9 (30)
pbo-S 2 aka RT driad, 24hra watar aoak watar 19.3 1 3.6 (30)
pbo-6 ld Acat, 1d Ira, 1d Acat, 1d watar watar 26.3 1 4.6 (30)
pbo-7 LR2 froaan, 1d RT driad uatar 21.5 1 4.3 (30)
pbo-8 1d RT dry, LR2 froaan, 1d RT dry watar 20.3 1 3.6 (30)
pbo-9 ld RT driad, 2d Acatona Acatona 20.6 1 3.6 (30)
pbo-lo 1d neon, 1d Acatona, 5d Praon Praon 27.9 1 4.7 (30)
pbo-l1 5d RT driad, 4000 driad uatar 19.2 1 3.5 (30)
pbo-12 5d RT driad,440¢ draid dry 18.4 1 3.1 (30)
t-taata

pbo-2 18.8 1 3.8 (30)

pbo-4 21.7 1 2.9 (30) d.f.I 58 t I 3.330

pbo—4 21.7 1 2.9 (30)

pbo-7 21.5 1 4.3 (30) d.f.I 58 t I 0.195

pbo-2 18.8 1 3.8 (30)

pbo-9 20.6 1 3.6 (30) d.f.I 58 t I 1.932

pbo-ll 19.2 1 3.5 (30)

pbo-12 18.4 1 3.1 (30) d.f.I 58 t I 0.927

pbo-l 26.9 1 4.1 (30)

pbo-lo 27.9 1 4.7 (30) d.f.I 58 t I 0.823

 

PBO-10
PBO-11
PBO-12
PBO-13

ribar

PBO-RI RIC.

PBO-1
PBO-2
PBO-3
PBO-4
PBO-5
PBO-6
PBO-7
PBO-8

PBO-10
PBO-11
PBO-12
PBO-13

Plan-a Traatad PBO (Sat Ona)

10-22-88
10-19-88
10-17-88
10-17-88
10-24-89
10-17-89

234

 

# of braaka

10.6 1 0.5 (7)
12.6 1 1.3 (5)
10.3 1 0.4 (5)
16.2 1 1.5 (6)
14.2 1 1.5 (5)
12.0 1 1.4 (6)

Diamatar (Ricrona)

 

 

Tanaila Strangth (GPa)

 

 

19.46 1 2.52 (15) 3.37
20.87 1 1.93 (17) 3.37
17.18 1 2.43 (14) 2.44
18.88 1 1.91 (13) 3.22
20.78 1 2.83 (13) 3.34
20.18 1 2.29 (14) 3.38
19.36 1 1.83 (15) 3.41
20.16 1 2.00 (14) 3.41
19.85 1 2.08 (15) 3.15
20.31 1 1.77 (14) 3.34
19.54 1 1.85 (15) 3.53
20.07 1 2.02 (15) 3.38
19.38 1 1.79 (15) 3.30
19.45 1 1.78 (15) 3.35
Ccmpraaaiva Str. (MPa)
397 1 98 (27)

392 1 121 (10)

359 1 114 (11)

366 1 115 (12)

344 1 80 (12)

321 1 78 (12)

379 1 95 (14)

409 1 127 (15)

397 1 107 (30)

347 1 76 (14)

374 1 100 (16)

341 1 105 (14)

374 1 123 (15)

331 1 99 (15)

0.50
0.51
0.41
0.49
0.49
0.47
0.58
0.46
0.30
0.29
0.45
0.58
0.35
0.43

“HRH’WH‘H’MH‘H'H’H'HH'

(13)
(13)
(14)
(13)
(13)
(13)
(14)
(11)
(13)
(14)
(15)
(15)
(14)
(14)

Traatmant

nona
02 50‘

O2 25 CF4 25‘

Ba 50‘
CO2 50‘
R83 50‘
RB3 50‘
R20 50‘
R20 50‘
Ar 50‘
82/82 50‘

Ar 50‘, 002 100‘
Ar 50‘, O2 100.

H20 1008

PBO-CI
PBO-1
?BO-2
PBO-3
PBO-4
PBO-5
PBO-6
PBO-7
PBO-8
PBO-9
PBO-11
PBO-12
PBO-13

ICC .

ICC .

Plan-n 2103104 280 (But Two,

 

I of Break-
10.7 1 0.8
15.3 1 1.0
12.9 1 1.6
13.0 1 1.6
12.1 1 0.9
12.4 1 1.9
12.4 1 2.0
14.8 1 1.8
15.3 1 1.3
12.8 1 1.4
13.8 1 2.1
16.8 1 2.2
19.3 1 2.2

(14)
(9)
(10)
(1°)
(9)
(9)
(3)
(3)
(8)
(3)
(1°)
(1‘)
(1°)

dianntor (um)

 

 

19.95 1 1.95 (39)
20.50 1 1.83 (7)
21.66 1 1.45 (7)
22.15 1 2.59 (6)
19.49 1 1.27 (7)
20.67 1 1.47 (7)
20.35 1 1.12 (10)
16.60 1 1.04 (6)
19.90 1 0.73 (6)
19.57 1 0.97 (8)
20.20 1 2.16 (6)
16.10 1 2.24 (6)
18.34 1 1.54 (10)
t or BREAKS

10.7 1 0.8 (14)
15.3 1 1.0 ( 9)
12.9 1 1.6 (10)
13.0 1 1.6 (10)
12.1 1 0.9 ( 9)
12.4 1 1.9 ( 9)
12.4 1 2.0 ( 6)
14.6 1 1.8 ( 6)
15.3 1 1.3 ( 6)
12.6 1 1.4 ( 6)
13.8 1 2.1 (10)
16.7 1 2.1 (14)
19.3 1 2.2 (10)

 

 

235 JAppcndthI
1-30-89)

Trontnont
cournon
00, 50 1.00 am 301 wane
00, 50 0.75 an 301 wan-6
00, 50 0.50 am 301 name
00, 50 1.00 an: 479 WATTS
00, 50 0.75 am 479 wane
00, 50 0.50 am! 476 mum's
0, 25 0!. 25 1.00 am 301 wars
0, 25 Ct. 25 0.75 am 301 WATTS
0, 25 0!. 25 0.50 14117 301 wane
0, 25 Ct. 25 0.75 an: 397 WATTS
0, 25 Ct. 25 0.50 an: 395 wan-s
0, 25 or. 25 1.00 am 395 mus
Ton. 5:1. (02.) Comp. 511. (“91)
3.19 1 0.48 (29) 397 1 96 (27)
3.44 1 0.56 (7) 356 1 107 (7)
3.41 1 0.47 (7) 367 1 96 (5)
3.29 1 0.43 (6) 362 1 108 (5)
3.45 1 0.51 (7) 392 1 112 (6)
3.56 1 0.41 (7) 405 1 74 (5)
3.25 1 0.60 (9) 451 1 57 (5)
3.46 1 0.49 (7) 501 1 106 (5)
3.22 1 0.39 (7) 391 1 60 (5)
3.69 1 0.49 (6) 359 1 75 (5)
3.31 1 0.46 (7) 404 1 106 (9)
3.49 1 0.52 (6) 367 1 62 (5)
2.02 1 0.55 (9) 645 1 159 (12)
1.. RID f (1121)
2.05 114.1 16.556 1 1.276
1.43 79.7 23.697 1 1.545
1.71 94.7 19.936 1 2.465
1.69 94.0 20.091 1 2.416
1.62 100.9 16.717 1 1.434
1.77 96.2 19.232 1 3.004
1.76 96.6 19.125 1 3.064
1.49 62.9 22.795 1 2.832
1.44 60.1 23.566 1 1.961
1.73 95.9 19.705 1 2.146
1.59 66.6 21.327 1 3.323
1.32 73.1 25.631 1 3.232
1.14 63.3 17.721 1 2.032

 

 

236

among 011:1

k29aro lthylono Glycol contact anqlo:
k29arn Hothylono Iodido contact anqlo:
k29arw Water contact anglo:
Sqrt 0139. Comp. I 5.264 1 .089
8qrt Polar Comp. I 4.102 1 .090
010p. Comp. (dyno/cn) I 27.71 1
Polar Coup. (dyna/cu) I 16.83 1
Total surface onorgy (dyno/cn) I

.94
.74
44.54 1

k29wo lthylono Glycol contact anqlo:
k29wu Hothylono Iodido contact anqlo:
k29ww Wator contact anqlo:
8qrt Dicp. Comp. I 5.434 1 .056
8qrt Polar Comp. I 4.014 1 .062
Dilp. Comp. (dyno/cn) I 29.53 1
Polar Comp. (dyna/cn) I 16.11 1
Total surface anorqy (dyno/cm) I

.61
.50
45.64 1

Eazlnr_12_hl_312:122§

k49-ar-o lthylono Glycol contact anglo:
k49-ar-m Hothylono Iodido contact anqlo:
k49-ar-w Wntor contact anglo:
Sqrt 013p. Comp. I 5.231 1 .212

Sqrt Polar Comp. I 3.632 1 .212

0119. Camp. (dyno/cn) I 27.37 1 2.21
Polar Comp. (dyno/cn) I 13.19 1 1.54
Total ourfaco onorgy (dyno/cn) I 40.56 1

k49-w-o lthylono Glycol contact anqlo:
k49-w-1 rornanido contact anglo:
k49-w-n Hothylono Iodido contact anqlo:
k49-w-w Wator contact anqlo:
Sqrt Dilp. Comp. I 5.498 1 .219

Sqrt Polar Comp. I 4.022 1 .194

Dilp. Comp. (dyno/cn) I 30.23 1 2.41
Polar Comp. (dynelcm) I 16.18 1 1.56 .
Total surfaco onorgy (dyno/cn) I 46.40 1

2&9.Bl.3£££1¥1§

pbo-ar-o lthylono Glycol contact anglo:
pbo-ar-n Mothylono Iodido contact anqlo:
pbo-nr-w Water contact anglo:

8qrt 010p. Comp. I 5.779 1 .094

8qrt Polar Chap. I 2.458 1 .083

010p. Comp. (dyno/cn) I 33.39 1 1.09
Polar Comp. (dynelcm) I 6.04 1 .41
Total ourflco onorgy (dyno/cn) I 39.43 1
2&9.ﬂllhlﬂ

pbo-w-o lthylono Glycol contact anglo:

pbo-HIV. Wator

8qrt Dinp. Comp. I
8qrt Polar Coup. I
Diop. Coup. (dyno/cn) I
Polar Coup. (dyno/cn) I
Total curtaco onorgy (dyno/cn) I

contact anqlo:
5.206 1 .470
2.964 1 .368
27.10 1 4.89
8.78 1 2.18
35.88 1

27.22
42.32
60.52

1.67

27.70
40.49
59.94

1.11

57.24
31.89
62.70

3.75

30.13
24.97
36.32
59.93

3.97

45.43
40.05
76.73

1.50

”0104‘”

”I.”

6.30
12.7
3.57

5.52
5.77
3.47

6.86
6.45
3.24

5.18
4.62
5.44
4.48

4.78
7.43
3.59

(30)
(25)
(25)

”WNW

“A-“
vvvv

(2°)
(2°)
(35)

45.16 1 6.15 ( 5)
75.54 1 3.66 ( 7)

7.07

.Appendhtll

237

tech-ar-e Ethylene Glycol contact angle:
tech-er-m Methylene Iodide contact angle:
tech-nr-w Water contact angle:
8qrt Diep. Comp. I 5.784 1 .079
8qrt Polar Coup. I 3.271 1 .081
Dilp. Comp. (dynelcm) I 33.45 1
Polar Comp. (dynelcm) I 10.70 1
Total eurfnce energy (dynelcm) I

.91
.53

tech-w-e 8thylene Glycol contact angle:
tech-u-n Methylene Iodide contact angle:
tech-w-w Water contect engle:
Sqrt Diep. Comp. I 5.364 1 .147

8qrt Polar Comp. I 4.905 1 .146

Diep. Comp. (dyne/cn) I 28.77 1 1.58
Polar Comp. (dynelcm) I 24.06 1 1.43
Total eurface energy (dynelcm) I

W

kn10014e Ethylene Glycol contact angle:

kn10014m Methylene Iodide contact angle:

kn10014w Water contact angle:

8qrt Dicp. Comp. I 5.321 1 .177

Sqrt Polar Coup. I 3.351 1 .168

Diep. Comp. mM/n I 28.31 1 1.89

Polar Coup. nN/n I 11.23 1 1.12

Total eurtace energy uN/u I 39.54 1 3.01
- + l4

kn3014e lthylene Glycol contact angle:

kn3014n Methylene Iodide contact angle:

kn3014w Hater contact angle:

Sgrt Diep. Coup. I 5.519 1 .107

8qrt Polar Coup. I 2.955 1 .092

Diep. Comp. (dynelcm) I 30.46 1 1.18

Polar Comp. (dynelcm) I 8.73 1 .54

Total eurface energy (dynelcm) I

- + 13.

kn3015e Ethylene Glycol contact angle:
kn3015m Methylene Iodide contact angle:
kn3015w water contact angle:
8qrt Diep. Coup. I 5.692 1 .068
8qrt Polar Coup. I 2.913 1 .070
Diop. Comp. (dyne/ce) I 32.39 1
Polar Comp. (dynelcm) I 8.49 1
Total curtace energy (dyne/ce) I

.77
.41

+ 12

W

kn40012e Ethylene Glycol contact angle:
kn40012m Methylene Iodide contact angle:
kn40012w weter contact angle:
Sqrt Diep. Coup. I 5.548 1 .061
8grt Polar Comp. I 2.968 1 .070
Diep. Comp. (dyne/cn) I 30.77 1
Polar comp. (dynelcm) I 8.81 1
Total eurface energy (dynelcm) I

.68
.42

38.38
32.18
65.76

44.15 1 1.44

24.48
26.60
45.56

52.83 1 3.01

49.01
34.70
67.41

41.76
43.62
73.02

39.19 1 1.73

42.19
38.88
71.48

40.88 1 1.18

36.44
45.50
73.20

39.58 1 1.10

”It”

“I?”

”It":

”I.”

3.78
5.46
2.43

5.60
5.79
8.58

2.53
10.4
4.31

3.36
7.96
2.27

6.06
8.41
2.87

(25)
(25)
(25)

(40)
(25)
(37)

(25)
(2°)
(45)

(25)
(3°)
(25)

(23)
(5°)
(24)

AppendixH

238
W192
kn40013e Ethylene Glycol contact angle: 45.77 1
kn40013m Methylene Iodide contact angle: 37.26 1
kn40013w Water contact angle: 70.78 1
qut Disp. Comp. I 5.759 1 .096
Sqrt Polar Comp. I 2.860 1 .109
Disp. Comp. (dynelcm) I 33.17 1 1.11
Polar Comp. (dynelcm) I 8.18 1 .63
Total surface energy (dyne/cn) I 41.34 1 1.73
W911
kn40014e Ethylene Glycol contact angle: 53.29 1
kn40014n Methylene Iodide contact angle: 47.35 1
kn40014w Water contact angle: 78.52 1
8grt Disp. Coup. I 5.306 1 .121
8grt Polar Coup. I 2.563 1 .118
Disp. Comp. (dynelcm) I 28.16 1 1.29
Polar Coup. (dynelcm) I 6.57 1 .60
Total surface energy (dynelcm) I 34.72 1 1.89
W
kti-e Ethylene Glycol contect angle: 54.01 1
kti-m Methylene Iodide contect angle: 49.84 1
kti-w Weter contact angle: 71.41 1
Bgrt Disp. Comp. I 4.794 1 .138
Sqrt Polar Comp. I 3.434 1 .129
Disp. Comp. (dynelcm) I 22.99 1 1.32
Polar Comp. (dynelcm) I 11.79 1 .88
Totel surface energy (dyne/cm) I 34.78 1 2.21
W
khe39013-1 Ethylene Glycol contact engle: 38.94
khe39013-1 Water contact angle: 62.95

sqrt DLsp. Comp. I 4.103 1 .355

8grt Polar Coup. I 4.731 1 .300

Disp. Comp. (dynelcm) I 16.83 1 2.91

Polar Comp. (dynelcm) I 22.38 1 2.84

Total surface energy (dyne/ce) I 39.21 1 5.74

+ 15

5.95 (15)
7.06 (30)
3.76 (15)

6.36 (39)
15.5 (35)
5.69 (40)

7.56 (40)
3.40 (25)
5.77 (40)

Appendix H

1 6.76 ( 5)

1 2.54 ( 4)

W

kar-e Ethylene Glycol contact angle: 40.32 1 5.53 ( 5)
her-m Methylene Iodide contact angle: 47.74 1 1.44 ( 4)
kar-w Water contnct angle: 68.20 1 3.14 ( 5)

8grt D1sp. Comp. I 5.187 1 .171
8grt Polar Coup. I 3.542 1 .165

Disp. Comp. (dynelcm) I 26.90 1 1.78
Polar Coup. (dynelcm) I 12.54 1 1.17
Total surface energy (dynelcm) I 39.45 1 2.95

239

AppendixH

Regreuion Plots for Iilhelly Data

 

 

 

 

 

 

 

 

 

 

 

 

Kevlar-49 AR
14
1' = 13.21 1.5
12_ 7" = 27.41 2.2
77 =40.613.8 i
10-
3 ..
6 I I
I l l
0 0.5 1 1.5
Kevlar Ar+ 75keV IBIS 01d
14
7" = 11.3 1 .6
12_ 7" = 29.7 1 1.0
77 = 41.01 15
10—
3 ._
5-
I l I '
0 0.5 1 1.5
Kevlar N+ 400keV 11114
14
7" I 6.61 .6
12‘ 74 228.2113
77 I 34.7 1 1.9
10—
8 ..
5-
1 l l
0 0.5 1 1.5

14

Kevlar-49 Washed

 

7P =16.211.6
7" = 30212.4
77 =46.41 4.0

     
 

 

 

 

I l I
0.5 1 1.5

Kevlar Ti+ lecV lElS

 

1P 1 11.81 .9
7" 123.01 1.3
17:34.8122

 

 

 

P
X= ‘41?

1’: 1704-0056)

24F

240

 

 

 

 

 

 

 

 

 

 

 

 

 

Kevlar N+ lOOkeV 1614
14
7P = 11.21 1.1
12_ 7‘ =28.311.9
77 = 39.5 1 3.0
10—
8-
6-
l l
0 0.5 1 1.5
Kevlar N+ 400keV 1613 old
14
7' I 8.21 .6
12_ 1‘ . 33211.1
7’ a 41.31 1.7
10—
8..
5-
l l
0 0.5 1 1.5
Kevlar N+ 30keV 1615 old
14
1’ I 8.51 .4
12_ 7‘ 132.41 .8
«(7 =40.91 12
10-
8-
6..
l l l
0 05 1 1.5

AppendixH

 

 

 

 

 

 

 

 

Kevlar N-I- 400keV 1E12 01d
14
7" I 8.81 .4
12_ 7‘ I 30.81 .7
77 = 39.61 1.1
Y 10-
8-
5-
I I I
0 0.5 l 1.5
Kevlar N+ 30keV 1514 old
l4
7' I 8.71 .5
12_ 7" I30.S1 1.2
77 = 39211.7
Y 10-
3-
5-
I I
0 0.5 l 1.5
X

Y= 1704-0056)

21F

241

 

 

Appendix H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
  

 

 

 

Kevlar-29 AR Kevlar-29 Washed
14 14
yr 1 15,3 1 .7 7' I 16.11 .5
12_ 14:27.719 12_ 7"=29.51.6
)7 1445117 )7 =45.611.1
10- Y 10-
8 -I 8 —
6- 6-
l 1 l l I l
0 0.5 __ 1 1.5 0 0.5 l 1.5
Technora AR Technora Washed
14 14
7? =10.7 1 05 7? = 24.11 1.4
4 g d =
12_ 77 33.5 1 0.9 12_ 77 28.8 1 1.6
y =44211.4 1 =52.613.0
10- -
Y 10
8 1 I 8-
1
6- 6-
I I I I I I'
0 0.5 1 1.5 o 0.5 1 1.5
PBO AR x
14
7" I 6.04 1 0.41
12_ y‘=33.411.1 F 17(1+0056)
77 I 39.41 1.5 - 243,7
10—
- p
8 x = .77
6 ‘Y
I I I
0 0.5 1 1.5

 

242 Appendix H
Tvo Po ints Ii 1he lay Data
Fiber Water 61.01)). 1'; 1'; 7}
Com. Ang. (°) C0111. Aug. (°) (1me dync/em dynelcm
DER33] Epoxy 36.7 103 47.5

K29 As Rec. 60.5 1 3.6 (25) 27.2 1 6.3 (30) 23.3 1 1.4 19.8 1 1.1 43.1 1 2.5
K29 Washed 59.9 1 3.5 (25) 27.7 1 5.5 (25) 22.3 1 1.4 ”.8 1 1.1 43.1 1 2.5
1:491:61. 62.713.2(25) 57216.9(25) 34110.73 38.7120 42.2127
K49 Washed 60.7 1 4.4 (60) 30.2 1 5.8 (38) 21.4 1 1.5 ”.7 1 1.1 42.2 1 2.7
11+ 100er 1515 74115.7 (25) 55.1145 (30) 16211.6 1421 1.8 30.41 3.4
A1" 75KeV 1515 68.0 1 3.4 (25) 40.01 5.3 (25) 22.1 1 1.6 15.4 1 1.1 37.6 1 2.7
N“ 30KeV 1515 71.5 1 2.3 (25) 42.2 1 3.4 (25) 24.7 1 1.1 12.0 1 0.6 36.7 1 1.8
N‘ 30KeV 1514 73.0 1 4.3 (45) 41.8 1 2.5 (25) 27.6 1 2.1 9.91 1 0.98 37.5 1 3.1
11+ 100KeV1514 75.1168 (30) 47.4136 (30) 24.1123 10211.5 343142
N‘ 400KeV 1514 76.5 1 6.3 (42) 46.8 1 4.4 (22) 26.9 1 3.5 8.34 1 1.5 35.2 1 4.9
1513 1 day old 62.8 1 4.3 (25) 48.1 1 7.2 (24) 9.29 1 1.4 29.5 1 2.0 38.8 1 3.3
1513 2 days old 68.3 1 3.3 (15) 49.0 1 3.7 (15) 13.9 1 1.5 ”.6 1 1.5 34.5 1 3.1
1513 3 days old 65.8 1 5.5 (15) 44.3 1 5.0 (10) 15.4 1 3.2 21.3 1 2.9 36.7 1 6.1
5512 1 day old 63.3 1 4.1 (24) 36.2 1 5.3 (20) 19.7 1 1.8 ”.1 1 1.5 39.8 1 3.3
5512 2 days old 67.9 1 6.4 (13) 37.0 1 5.6 (15) 25.1 1 3.6 13.9 1 2.2 39.0 1 5.8
5512 3 days old 79.5 1 4.7 (15) 44.3 1 5.5 (15) 34.7 1 3.6 4.74 1 1.08 39.4 1 4.6
2512 day I 61.3 1 7.2 (25) 31.11 6.9 (29) 21.5 1 2.5 ”.3 1 2.1 41.7 1 4.6
2512 day 2 52.9 1 6.7 (19) 29.5 1 5.6 (21) 13.8 1 2.1 32.7 1 2.7 46.6 1 4.8
2512 I mouth 73.2 1 2.9 (2.4) 36.4 1 6.1 (23) 33.5 1 1.9 7.74 1 0.74 42.3 1 2.6
1513 day I 62.9 1 1.8 (17) 37.6 1 4.8 (22) 18.1 1 1.0 21.5 1 0.9 39.61 1.9
[513 day 2 64.6 1 4.2 (”) 37.5 1 6.4 (25) 20.0 1 1.8 19.0 1 1.5 39.0 1 3.3
2514dayl 71.313.2(25) 39214.8 (23) 27611.8 10.9103 38512.6
2514 day 2 70.7 1 2.4 (25) 49.1 1 5.0 (21) 16.3 1 1.2 17.1 1 1.0 33.4 1 2.2
2514 1 month 82.8 1 1.4 (25) 57.3 1 4.4 (24) 22.5 1 1.1 6.92 1 0.49 29.5 1 1.6
25141»: wash 76.014.1(29) 51917.6 (30) 19611.9 11.8112 31413.1
Aerrod.B 60.1 16.6 (54) 32.61 5.8(25) 19.0126 22812.1 41814.8
Par-N 10011»: 86.5 1 4.3 (34) 52.8 1 6.2 (25) 34.7 1 2.8 2.41 1 0.57 37.1 1 3.3
Technora as rec. 65.8 1 2.4 (25) 38.4 1 3.8 (25) ”.9 1 1.1 17.7 1 0.8 38.5 1 1.8
Technora washed 45.6 1 6.2 (25) 24.5 1 3.7 (25) 10.8 1 1.3 42.3 1 2.1 53.0 1 3.5
PBO as rec. 76.7 1 3.6 (35) 45.4 1 4.8 (”) 28.7 1 2.2 7.63 1 0.88 36.3 1 3.1
PBO mashed 76.0 1 4.1 (35) 45.3 1 5.5 (24) 27.6 1 2.3 8.40 1 0.98 36.0 1 33
P80” (set two) 68.3 1 3.7 (45) 29.6 1 3.0 C”) 33.0 1 2.3 10.4 1 0.9 43.3 1 3.2
P8044 (set two) 53.5 1 6.2 (30) 31.9 1 5.8 (24) 12.9 1 1.8 33.2 1 2.3 46.1 1 4.1
P8047 (set two) 51.8 1 4.0 (25) 26.7 1 6.6 (22) 14.2 1 1.3 33.4 1 1.7 47.6 1 3.0
P5048 (set two) 53.9 1 9.1 (28) 29.9 1 3.6 (30) 14.9 1 2.4 30.7 1 2.8 45.6 1 5.2
P804‘12 (set two) 58.8 1 5.3 (25) 35.7 1 5.4 (24) 15.2 1 1.7 26.9 1 1.9 42.0 1 3.6

243

Appendix]!

Droplet Raoultl Rogroaaion Plot:

Untreated S 1000, nil-Wed

 

 

50
40..

304

10-

 

 

0

 

 

50
Slope = 0.01358 1 0.00077
40 ' 11.2 .. 0.933
Load 30 q
(luv) 20 _
10 .1
I
0 '1 I ﬂ 1 1
0 150 300 450 600 750
Untreated 81000, 10 min sulfonated
50
Slope = 0.04041 1 0.00391
40 “ 112 = 0.906
Load 30 '-
0“) 20 _
.1!
10 -
I
0 l l 1 1
0 150 300 450 600 750
Untreated S1000, 1 hr mlfonatod
50
Slope = 0.05602 1 0.00388
40 " 112 . 0.959
Load 30 -
(luv) 20 _
10 ..
0

 

 

 

 

 

 

 

1111
0 150300450600750

Droplet Length (uni)

 

Untreated 81000, 5 min sulfonated

 

Slope = 0.03544 1 0111195
112 . 0.957

20_ I

 

 

1 l 1 1
0 150 300 450 6(1) 750
Untreated 81000, 30 min sulfonated

 

 

Slope = 0.05471 1 0.00288
R2 . 0.989

 

 

 

 

1 1 l 1
0 150 300 450 600 750
Untreated 81W. 2 hr sulfonated

 

 

Slope = 0.06045 1 0.01143
R2 . 0.903

  

 

 

 

1 l l l
0 150 300 450 600 750
Droplet Length 0.1m)

 

Untreated 81000, 3 111' sulfonated Comna treated 81000, un-preu'eawd

244

 

Appendix H

 

    

 

 

 

 

50 50
Slope=0.0663710.00250 I Slope=0.0403310.00186‘
40" R2=0.974 ' 40‘ [1220.961
_ 30 ._

Load 30 I J
(HIV) 20 _ 20 _ a I .

10.. 10-

0 0

 

 

1 1 1 1
0 150 33X) 4%) 600 750 0 150 300 450 600 750
Coronau'e'ated81000,5minsulfonated Corona treated SlW,3hrSu1fonaled

 

 

    
  

 
 
  

50 Slope=0.0513510.00180 50 Slope=0.0624110.00337
40" 112:0.977 40" 112:0.974 I
1.661130“ 30'
(luv) 20 _ 20 _
10— 10-

 

 

 

 

 

 

1 1 1 1
150 300 450 600 750
Plasma treated 811110. 5 min sulfonated

1 1 1 1
O 150 300 450 600 750 0
Plasmaueated 81000.nn-preu'eated

 

 

 

 

 

 

 

 

50 50
Slope =- 0.04091 1 0.00255 Slope = 0.05721 1 0.00355
40" 112-0.931 40" 112:0.953
Load 30 - . . 30 -
I
(mV) 20 _ I . . 20 _
10 - I .- 10 -
I
0 1 1 1 1 0 1 1 1 1
0 150 300 450 600 750 0 150 300 450 600 750
Droplet Length (um) Droplet Length (pm)

245 Appendix H

Plasma treated $1000, 2 hr sulfonated Plasma treated 81000, 3 hr sulfonated

 

 

    
 

    
 

 

 

 

 

 

 

 

 

 

50 50
Slope=0.0636710.00288 Slope=0.0604910.00183
40' R2=0.972 40" 11220981
Load 30 - 30 - '
(EN) 20_ 20_
10— 10—
0 ~ I I I I 0 I I I I
01503004506007500150300450600750
Droplet Length (um) Droplet Length (pm)
Untreated Spawn-1000 400keV 1513He+implanted Spectra-10m
20 S1ope=0.0135810.00077 20.. Slope=0.0473210.00”5
R2 = 0.933
15 .-
Load
(mV) 10 -
5 ._
I
0 "1 1 1 I

 

 

 

 

 

 

M ”0.95.8011

246 lappendbtll

Sulfonated Pc sulfur Penetration Depthe (Ala neeulte)

 

 

 

 

    
  

 

 

Time (hr) Chilled (um) Ambient (um) Warn (um)
0.167 - 0.7 1 0.2 [6] 1.4 1 0.3 [6]
0.50 - 1.1 1 0.2 [6] -
10° - 1e5 t 002 [5] '-
2.0 - 2e2 t Del [6] -
3.0 - 2.8 t 0.2 [6] -
7.0 - 3e6 * 00‘ [6) -
10.0 0.9 1 0.2 [6] — -
168.0 3.7 1 0.3 [5] 5.6 1 0.2 [6] 7.6 1 0.9 [6]
BLIP-C 914/31 H.331 BILMImMQRm
FILE: 1117111.: Richest". month-e1. Emu-sled. 11111131320111
”.5 me= 1.1” k €18. m: ..715 k cl: "31..” “ﬂ.“
1. 1 4 . t 4. . 4 : 1 1 . 1
l
3
. 7.53“?“th
7
‘ I
5
‘
l
3
2
l
O - . : : .
... 2.. 1.. 3.. ... I... 12..

men. elm

BIBLIOGRAPHY

BIBLIOGRAPHY

 

Adem, E.H., 8.]. Bean, C.M. Demanet, A. LeMoel, and LP. Duraud, ”XPS as a Tool
for Investigation of Polymers Irradiated by Energetic Ions, " Nuclear Instruments and
Methods in Physics Research, B32, 182 (1988).

Agrawal, DD. and LJ. Broutman, Analysis and Performance of ﬁber Reirgﬁmced
Polymers, John Wiley and Sons, In. N.Y., 48 (1980).

Allen, S.R. , ”Tensile Recoil Measurement of Compressive Strength for Polymeric High
Performance Fibers," J. Mater. Sci, 22, 853 (1987).

Allred, R.E. , Surface Chemical Modiﬁcation of Polyaramid Filaments with Amine
Plasmas, Ph.D. dissertation, Massachusetts Institute of Technology, (1983).

Allred, R.E., E.W. Merrill and D.K. Roylanee, Molecular Characterization of
Composite Interfaces, H. Ishida and G. Kumar, Eds, Plenum Press, New York,
333 (1985).

Armistead, J.P., A.W. Snow, W.D. Bascom, ”Butadiyne Vapor Deposition
Polymerization on Carbon Fibers,‘ 19th International SAMPE Technical Conference,
October, 644 (1987).

Atkins, P.W., Physical Chemistry, W.H.Freeman & Company, San Francisco, 850
(1978).

Attenburrow, 6.13. and D.C. Bassett, “Compliance and Failure Modes of Oriented Chain
Extended Polyethylene“ J. Material Science, 14, 2679 (1979).

Baszkin, A. and L. Ter-Minassin—Saraga, “Effect of Surface Polarity on Self-Adhesion
of Polymers,‘ Polymer, 19, 10833 (1978).

Baun, W.L., Appl. Surface Sci, 4, 291 (1980).

Bazhenov, L., V.V. Kozey, and A.A. Berlin, “Compressive Fracture of Organic Fiber
Reinforced Plastics,” J. Mater. Sci, 24, 4509 (1989).

247

248

Beach, W.F., "A Model for the Vapor Deposition Polymerimtion of p-Xylylene,"
Macromolecules, 11, 72 (987).

Bell, A.T., “The First Annual International Conference of Plasma Chemistry and
Technology, Ed. H.V. Boenig, Technomic Publishing Co. Inc., Iancaster, PA, 29
(1983).

Bellamy, L.T., The Infrared Spectra of Complex Molecules, John Wiley & Sons, New
York (1958).

Berg, J .S. and D.F. Adams, ”An Evaluation of Composite Materials Compression Test
Methods,” J. of Composites Technology & Research, 41 (1989).

Bertrand, P., Y. DePuydt, J.M. Beulmn, P. Lutgen, and G. Feyder, ”Modiﬁcation of
Polymer (PET) surface Reactivity by Low Energy Ion Bombardment, " Nuclear
Instruments and Methods in Physics Research, Bl9/20, 887 (1987).

Biersack, LP. and L. G. Haggmark, Nucl. Instr. Meth., 174, 257 (1980).

Billmeyer, F.W., Textbook of Polymer Science, John Wiley & Sons, New York, 289
(1984).

Bjorksten, 1. and LL Yaeger, “Vinyl Silane Size for Glass Fabrics", Mod. Plast., 29,
124 (1952).

Blades, H., "Dry-Jet Wet Spinning Process,“ U.S. Patent 3,767,756 (1973).

Blythe, 11.11., D. Briggs, 0.11. Kendall, D.G. Rance and v.1.1. Zichy, “Surface
Modiﬁcation of Polyethylene by Electrical Discharge Treatment and the Mechanism
of Autoadhesion,‘ Polymer, 19, 1273 (1978).

Bodo, P. and LE. Sundgren, J. Appl Phys., 60, 1161 (1986).

Bowden, P.B. and R.J. Young, ”Deformation Mechanisms in Crystalline Polymers," J.
Material Science, 9, 2034 (1974).

Brady, J.M. and 12.1.. Thomas, 'The Deformation of Oriented High Density
Polyethylene, Part 2: The Formation of Long Crystalline Fibrils at Elevated
Temperatures,‘ J. Material Science, 24, 3319 (1989).

Brady, J.M. and R.S. Porter, ”Composites Reinforced with Fibers of a Thermotropic
Liquid Crystal Copolyester,‘ J. Thermoplastic Composite Materials, 3, 252 (1990).

249

Brice, D.K. , ”Ion Implantation Range and Energy Deposition Distributions, ” Plenum
Data Company, New York, (1975).

Briggs, D., C.R. Kendall, A.R. Blythe and A.B. Wootton, ”Electrical Discharge
Treatment of Polypropylene Film,” Polymer, 24, 27 (1983).

Cameron, G. and B.R. Main, Polym. Degrad. Stab., ll, 9 (1985).

Carley, LP. and P.T. Kitze, ”Corona-Discharge Treatment of Polyethylene Films. 1.
Experimental Work and Physical Effects,” Polymer Eng. Sci., 18, 326 (1978).

Carter, G. and LS. Colligon, Ion Bombardment quolids, Ch. 5, American Elsvier
Publishing Company Inc. , New York, (1968).

Cazeneuve, C., LE. Castle and LP. Watts, J. Mater. Sci., 25, 1902 (1990).

Chand, N., R.K. Tiwary, and P.I(. Rohatgi, ”Resource Structure Properties of Natural
Cellulosic Fibers - an Annotated Bibliography,” J. Material Science, 23, 381 (1988).

Chamis, C.C., ”Mechanics of Load Transfer at the Interface,” Inted'aces in Polymer
Matrix Composites, E. P. Pluddemann, ed. Academic Press, New York, 31 (1974).

Chapiro, A., ”Chemical Modiﬁcations in Irradiated Polymers,” Nuclear Instruments and
Methods in Physics Research, B32, 111 (1988).

Chou, T. and A. Kelly, ”The Effects of Transverse Shear on the Longitudinal
Compressive Strength of Fiber Composites,” J. of Material Science, 15, 327 (1980).

Choc, C.R. and J. Jang, ”Adhesion Promotion of Ultra High Modulus Polyethylene
Fiber/Epoxy Composite Interface,” in Controlled Interphase in Composite Materials,
H. lshida, Ed., Proceedings of the Third International Conference on Composite
Interfaces, Elsevier, NY, 97 (1990).

Cohen, Y., ”Structure Formation in Solutions of Rigid Polymers Undergoing a Phase
Transition,” Ph.D. dissertation, University of Massachusetts, (1986).

Cooke, T.F., ”High Performance Fiber Composites with Special Emphasis on the
Interface”, J. Poly. 5113., 7, 197 (1987).

Cox, H.L., ”The Elasticity and Strength of Paper and Other Fiberious Materials,” J.
Appl. Phys. B., 3, 72 (1952).

Cussler, E.L., Diﬁirsion, Cambridge University Press, New York, NY (1984).

250

Dawes, C.J., Biological Techniques in Electron Microscopy, Barnes and Noble Inc.,
New York, (1971).

Defay, R., I. Prigogine, A. Bellemans and D.H. Everett, ”Surface Tension and
Adsorption,” Ch. XIX, John Wiley and Sons Inc., New York, (1966).

DeI.ong, J.D., K.J. Hook, M.J. Rich, J. Kalantar and LT. Drzal, ”Spectroscopic
Characterization of Fiber-Matrix Interphase, ” in Controlled Interphase in Composite
Materials, H. Ishida, Ed., Proceedings of the Third International Corﬁrence on
Composite Interfaces, Elsevier, NY, 87 (1990).

Dersselhaus, M.S., B. Wasserman, and GE Wnek, ”Ion Implantation of Polymas,”
28.” A.J. Kinloch, Adhesion and Adhesives, Chapman And Hall, New York, 129

(1987).

Deteresa, S.J., S.R. Allen, R.J. Farris and R.S. Porter, ”Compressive and Torsional
Behavior of Kevlar-49 Fibres,” J. Mater Sci., 19, 57 (1984).

Deteresa, S.J., ”The Axial Compressive Strength of High Performance Polymer Fibers,”
Ph.D. dissertation, University of Massachusetts, (1985).

Dixon, D.D., L.J. Hayes, Sulfa-Fluorination of Synthetic Resins, U.S. Patent 39405”,
(1976').

Dixon, D.D., L.J. Hayes, Fluorination of Polyester: and Polyamide Fibers, U.S. Patent
3988491, (1976’).

Dixon, D. D, W. M. Smith, Fluorinated Polyester Tire Reinforcement Materials, U. S.
Patent 4009304, (1977').

Dixon, D.D., L.J. Hayes, Fluorination of Polyolﬁn and Polyacrylonitrile Fibers, U.S.
Patent 4020223, (1977').

Dobb, M.G., DJ. Johnson and B.P. Seville, ”Supramolecular Structure of a High-
Modulus Polyaromatic Fiber (Kevlar 49),” J. Polymer Science, Polymer Physics 5d. ,
15, 2201 (1977).

Dobb, M.G., D.J. Johnson and B.P. Saville, ”Compressional Behavior of Kevlar 49
Fibers,” Polymer, 22(7), 960 (1981).

Drml, L.T., ”Interfacial Behavior of Aramid and Graphite Fibers in an Epoxy Matrix”,
15th National SAMPE Tech. Cont, Cincinnati OH, (1983).

251

Drzal, L.T., ”The Interfacial and Compressive Properties of Polybenmthiazole Fibers,”
AFWAL-TR-86—4003 (1986).

Drnl, L.T., M.J. Rich, and S. Subrarnoney, ”Fiber-Matrix Interface and Its Effects On
Composite Properties,” Proceedings of the Third Annual Conference on Advanced
Composites, Detroit, MI, 305 (1987).

Drzal, LT. and P.J. Herrera-Franco, ”Composite Fiber-Matrix Bond Tests, ” Engineered
Materials Handbook, 3, 391 (1991).

Gabayson, S.M., G. Sugerman, and SJ. Monte, ”Role of Coupling Agents in Aerospace
Composite Failure,” in Proceedings of the SPE Annual Technical meeting, New York
(1989).

Gabriel, B.L., SEM: A User’s Manual For Materials Science, American Society for
Metals, Metals Park, (1985).

Gaur, U and T. Dividson, ”Interfacial Effects of Plasma Treatment of Fiber Pull-out,”
Materials Research Society Symposium Proceedings 1 70, Pittsburgh, PA, 315 (1990).

Gilbert, A.H., B. Goldstein and G. Marom, ”A Liquid Droplet Measurement Technique
as a Means of Assessing the Interlaminar Shear Strength of Fiber-Reinforced
Composites,” Composites, 21, 408 (1990).

Gillberg, G., ”Polymer Surface Characterization: An Overview,” J. Aﬂresion, 21, 129
(1987). .

Greszczuk, L..,B ”TheoreticalStudiesoftheMechanismoftheFiber-Manixlnterface
in Composites,” Interface 1n Composites, ASTM, STP 452, Am. Soc. Testing and
Materials, Phil. PA, 42 (1969).

Grumman, D.S., R. Schalek, A. Ozzello, and LT. Drzal, ”Modiﬁcation of Fiber-Matrix
Adhesion in Polyethylene Reinforced Composites by Energetic Ion Irradiation,” in
Proceedings of the Sixth Annual ASM/ESD Advanced Composites Conference, Detroit,
MI, 155 (1990).

Grumman, D.S., R. Schalek, A. Ozzello, J. Kalantar and L.T. Drzal, ”High Energy Ion
Implantation of Polymeric Fibers for Modiﬁcation of Reinforcement-Matrix
Adhesion,” Nuclear Instruments and Methods in Physics Research (1991 in press).

Guimaraes, R.B., L. Amaral, M. Behar, and F.C. Zawislak, ”Anomalous Depth Proﬁles
of Light Ions and Nobel Gases Implanted Into Polymers,” Nuclear Instr-torrents and
Methods in Physics Research, B39, 800 (1989).

252

Gutowski, W. , ”Effect of Fiber-Matrix Adhesion on Mechanical Properties of
Composites”, Controlled Interphase in Composite Materials, H. Ishida Ed., Elsevier
Science Publishing Co., New York, NY, 505 (1990).

Hammer, GE. and L.T. Drzal, ”Graphite Fiber Surface Analysis by X-ray Photoelectron

Spectroscopy and Polar/Dispersive Free Energy Analysis, ” Applications of Surface
Science 4, North-Holland Publishing Company, 340, (1980).

Hansen, R.H., J .V. Pascale, T.D. Benedictis, and P.M. Rentzepis, ”Effects of Atomic
Oxygen on Polymers,” J. Polymer Science. A, 3, 2205 (1965).

Holden, P.S., G.A.J. Orchard, and LM. Ward, ”A Study of the Gas Barrier Properties
of Highly Oriented Polyethylene, ” J. Polymer Science. Polymer Physics Edition, 23,
709 (1985).

Holmes, S. and P. Schwatrz, ”Ammination of Ultra-high Strength Polyethylene using
Ammonia Plasma,” Composite Science and Technology, 38, l (1990).

Hook, K.J., R.K. Agrawal and L.T. Drzal, ”Effects of Microwave Processing on Fiber-
Matrix Adhesion. II. Enhanced Chemical Bonding of Epoxy to Carbon Fibers,” J.
Adhesion, 32, 157 (1990).

Ihata, J., J. Polym. Sci., Part A, Polym. arena, 26, 167 (1988).

Inagaki, N., A. Kishi and K. Katsuura, Int. J. Adhesion Adhesives, 2, 233 (1982).

Ishitani, A., K. Shoda, H. Ishida, T. Watanabe, and K. Yoshida, ”Characterization of
Oxygen-Implanted Polyethylene by Various Analytical Techniques, ” Nuclear
Instruments and Methods in Physics Research, B39, 783 (1989).

Jones, R.W., Fundamental Principles of Sol-Gel Technology, The Institute of Metals,
Brookﬁeld, VT (1989)

Kaelble, D.H., Physical Chemistry of Arﬁresion, Wiley-Interscience, New York, (1971).

Kalantar, J., D.S. Grummon and L.T. Drzal, ”Effects of Ti“, Ar+ and N+ Ion
Implantation on Aramid Fiber Adhesive Properties,” in Interfaces in Composites,
C.G. Pantano and E.J.H. Chen, Eds., Materials Research Society symposium
Proceedings 170, Pittsburgh, PA, 315 (1989).

Kalantar, J. and L.T. Drzal, ”The Bonding Mechanism of Aramid Fibers to Epoxy
Matrices, Part I: A Review of the Literature,” J. Mater. Sci., 25, 4186 (1990‘).

253

Kalantar, J. and L.T. Drzal, ”The Bonding Mechanism of Aramid Fibers to Epoxy
Matrices, Part II: An Experimental Investigation,” J. Mater. Sci., 25, 4194 (1990').

Kalantar, J ., L.T. Drzal and K.J. Hook, ”A Novel Sample Preparation Technique for
Ion and Electron Beam Analysis of the Fiber-Matrix Interphase in Polymeric
Composites,” J. Vac. Sci. Technol. A., 8, 3878 (1990‘).

Kalantar, J., L.T. Drzal and D.S. Grummon, ”Structural Properties of Aramid Fibers
and Their Inﬂuence on Fiber Adhesion, ” in Controlled Interphase in Cornposite
Materials, H. Ishida, Ed., Proceedings of the Third International Conference on
Composite Interfaces, Elsevier, NY, 685 (1990”).

Kalantar, Y. Muraoka, and L.T. Drzal, ”Effects of Temperature on Atomic Distribution
of Sulfonated Species in Polymers,” (manuscript in preparation, 1991').

Kalb, B. and A.J. Pennings, ”Maximum Strength and Drawing Mechanism of Hot Drawn
High Modulus Molecular Weight Polyethylene,” J. Material Science, 15, 2584
(1980).

Kazuyo, S., K. Imada, and M. Takayamgi, ”Formation of Kink Bands by Compression
of the Extruded of Solid Linear Polyethylene,” J. Polymer Science, 13, 73 (1975).

Kelly, A. and W.R. Tyson, ”Tensile Properties of Fiber-Reinforced Metals:
Copper/Tungsten and Copper/Molybdenum,” J. Mech. Phys. Solids, 13, 329 (1965).

Kim, C.Y., J. Evans, and D.A.I. Goring, ”Corona-Induced Autohesion of Polyethylene,”
J. Appl. Polymer Sci., 15, 1365 (1971).

Klomparens, K.L., S.L. Flegler, and G.R. Hooper, Procedures ﬂrr Transmission and
Scanning Electron Microscopy for Biological and Medical Science, LADD Research
Industries, Burlington, VT. (1935)-

Kovar, R. and RR. Wallis, ”Interpenetrated Polymer Films,” U.S. Patent 4,845,150
(1989).

Krause, S.J., T.B. Haddock, D.L. Veaie, P.G. Lenhert, W.F. Hwang, G.E. Price, T.E.
Helminiak, J.F. O’Brien, and W.W. Adams, ”Morphology and Properties of
Rigid-Rod Poly(p-Phenylene Benzobisoxaaole) (PBO) and stiff-chain

poly(2,5(6)-Benzoxazole) (ABPBO) Fibers, ” Polymer, 29, 1354 (1988).

Krause, S.J., D.L. Vezie, and W.W. Adams, ”Straightening of Pleated Sheet Structure
in Fibers of Poly(p-Phenylene TerephthalamideyKevlar 149,” Polymer
Communications, 30, 10 (1989).

254

Kumakhov, M.A. and ER Konlarov, ”Energy Loss and Ion Ranges in Solids,” Gordon
and Breach, Science Publishers Inc., New York, (1981).

Kumar, S., ”Compressive Strength of High Performance Fibers,” SAMPE Journal, 26,
51 (1990).

Kupper, K. and P. Schwartz, ”Modiﬁcation of the Fiber-Matrix Interface of p-Aramid
Fibers Using Gas Plasma,” J. Adhesion Sci. Technol., 5, 165 (1991).

LaCourse, W.C., ”Continuous Filament Fibers by the Sol-Gel Process,” in Sol-Gel
Technology for Thin Films, Fibers, Preforms, Electronics and Specialty Shapes, L.C.
Kelin Edt., Noyes Publications, Park Ridge, NJ, 184 (988). -

Ladizesky, N.H. and I.M. Ward, ”A Study of the Adhesion of Drawn Polyethylene
Fiber/Polymeric Resin Systems,” J. Material Science, 18, 533 (1983).

Iadizesky, N.H. and I.M. Ward, ”The Adhesion Behavior of High Modulus
Polyethylene Fibers Following Plasma and Chemical Treatments,” J. Material
Science, 24, 3763 (1989).

Lanauze, LA. and D.L. Myers, ”Ink Adhesion on Corona-Treated Polyethylene Studied
by Chemical Derivatization of Surface Functional Groups, ” J. Applied Polymer
Science, 40, 595 (1990).

Licciardello, A., O. Puglusi, L. Calcagno, and G. Foti, ”XPS Study of Polystyrene Gel
Induced by Ion and Electron Bombardmmt,” Nuclear Instruments and Methods in
Physics Research, Bl9/20, 903 (1987).

Liston, E.M., ”Plasma Treatment for Improved Bonding: A review,” J. Adhesion, 30,
199 (1989).

Madhukar, MS and L.T. Drzal, ”Fiber-Matrix Adhesion and its Effects on Composite
Mechanical Properties. 111. Longitudinal (0°) Compressive Properties of
GraphitelEpoxy Composites, ” (1n press)

Malis, T.F. and D. Steele, ”Ultramicrotomy for Materials Science,” in Workshop on
Specimen Preparation for TEM of Materials 11, R. Anderson, Ed., Materials
Research Society Symposium Proceedings 199, Pittsburgh, PA, 315 (1990).

Mark, V., Ger. Offen. 2,458,968 (1975).

Martin, D.C. and EL. Thomas, ”Micromechanisms of Kinking in Rigid-Rod Polymer
Fibers,” J. Material Science, (in press, 1991).

255

Mayer, W.J., L. Eriksson and J .A. Davies, ”Ion Implantation in Semiconductors,” Ch.
2., Academic Press, New York, (1970).

Mcalea, K. and G. Besio, ”Adhesion Between Polybutylene Terephthalate and E—Glass
Measured with a Microdebond Technique,” Polymer Composites, 9, 285 (1988).

Mercx, F.P.M. and P.J. Lemstra, ”Surface Modiﬁcation of Aramid Fibres,” Polymer
Communications, 31, 252 (1990).

Michalske, T.A. and B.C. Bunker, ”The Fracturing of Glass,” Scientb‘icAmerican, 257,
122 (1987).

Miller, B., P. Muri and L. Rebenfeld, ”A Microbond Method for Determination of the
Shear Strength of a Fiber/Resin Interface,” Comp. Sci. Tech., 28, 17 (1987).

Mittelman, H. , I. Roman, and G. Marom, ”The Morphology of Shear Fracture of Kevlar
Fiber-Reinforced Epoxy Composites, ” J. Material Science Letters, 4, 845 (1985).

Mittlefehldt, ER. and LA. Gardella, 'Jr., Appl. Spectrosc., 48, 1172 (1989).
Montoneri, E., J. Polym. Sci., Part A, Polym. Chem, 27, 3043 (1989).

Montoneri, E., M. Gleria, G. Ricca and G.C. Pappalardo, J. Macromol. Sci., arena,
A26, 645 (1989).

Morgan, R.J., C.O. Pruneda, and W.J.Steele, ”The Relationship Between the Physical
Structure and the Microscopic Deformation and Failure Processes of Poly(p-Phenylene
Terephthalamide) Fibers,” J. Polym. Sci. Part B, 21, 1757, (1983).

Morgan, R.J. and RE Allred, Reﬁrence Book for Composite Tech., S.M. Lee Edt.,
Technomic, Lancaster, PA, 143 (1989').

Morgan, R.J., Private Communication, (1989”).

Mona,.M., E. Occhiello, L. Gila and F. Garbassi, ”Surface Dynamics vs. Adhesion in
Oxygen Plasma Treated Polyoleﬁns,” J. Adhesion, 33, 77 (1990).

Muraoka, Y., J. Kalantar, L.T. Drzal and K.J. Hook, ”Analytical Techniques Used to
Quantify Chemistry and Atomic Distribution of Sulfonated Species in Polymers,” ‘
(manuscript in preparation, 1991').

Muraoka, Y., J. Kalantar, andL..T Drzal, ”SulfonationofUltraHighMolecularWeight
Polyethylene Fibers, Part I: Chemical Investigation,” (manuscript in preparation,
1991”).

256

Muraoka, Y., J. Kalantar, and L.T. Drzal, ”Sulfonation of Ultra High Molecular Weight
Polyethylene Fibers, Part II: Mechanical Investigation, ” (manuscript in preparation,
1991‘).

Nakayama, Y., F. Soeda, and A. Ishitani, ”Surface Analysis of Plasma Treated Poly
(ethylene terephthalate) Films,” Polymer Engineering and Science, 31, 812 (1991).

Nguyen, H.X., G. Riahi, G. Wood and A. Poursartip, ”Optimization of Polyethylene
Fiber Reinforced Composites Using A Plasma Surface Treatment, ” Allied Signal
Technical Publication, (1988).

Occhiello, E., F. Garbassi, M. Mom and L. Nicolais, ”Spectroscopic Characterization
of Interfaces in Polymer Composites,” Campos. Sci. and Technol., 36, 133 (1989).

Occhiello, E., M. Morra, G. Morini, F. Garbassi and D. Johnson, ”On Oxygen Plasma-
Trcated Polypropylene Interface with Air, Water, and Epoxy Resins. II. Epoxy
Resins,” J. Applied Polymer Science, 42, 2045 (1991).

Owens, D.K. , ”Mechanism of Corona-Induced Self-Adhesion of Polyethylene Film, ” J.
Appl. Polymer Sci., 19, 265 (1975).

Ozzello, A., D.S. Grumman, J. Kalantar, L.T. Drzal, I.H. Loh and R.A. Moody,
”Interfacial Shear Strength of Ion Beam Modiﬁed UHMW-PE Fibers in Epoxy Matrix
Composites,” in Interfaces Between Polymers, Metals and Ceramics, B.M. Dekoven,
A.J. Gellman and R. Rosenberg, Eds., Materials Research Society Symposium
Proceedings 153, Pittsburgh, PA, 217 (1989).

Panar, M., P. Avakian, R.C. Blume, K.H. Gardner, T.D. Gierke, and H.H. Yang,
”Morphology of Poly(p-Phenylene Terephthalamide) Fibers,” J. Polymer Science:
Polymer Physics Ed., 21, 1955 (1983).

Penn, L.S., F.A. Bystery and H. Marchionni, ”Relation of Interfacial Adhesion in
Kevlar/Epoxy Systems to Surface Characterization and Performance”, Poly.
Composites, 4(1), 27 (1983).

Penn, L.S., F. Bystry, W. Krap and S. Lee, ”Aramid/Epoxy vs. Graphite/Epoxy: Origin
of the Difference in Strength at the Interface,” Polym. Sci. Technol., 27, 93 (1985).

Perry, R.H. and D. Green, Perry's Chemical Engineers' Handbook, sixth ed.,
McGRAW-HILL, New York, 349 (1984).

P08161113, A.R. and A.J. Pennings, ”Chlorosulfonation of Ultra-High Strength
Polyethylene Fibers,” in Ultra High Modulus Polymers: Approaches to Design and
Development, Marcel Dekker Inc., New Yourk, 431 (1988).

 

257

Postema, A.R., W. Hoogsteen, and A.J. Pennings, ”Grazing in Inna-high Molecular
Weight Polyethylene Gel-Fibers,” Polymer Communications, 28, 148 (1987).

Puglisi, 0., Materials Science and Engineering, B2, 167 (1989).

Rao, V. andL.T. Drzal, ”TheDependenceofInterfacial ShearStrengthonMatrixand
Interphase Properties,” Polymer Composites, 12, 48 (1991”).

Rao, V., P. Herrera-Franco, A.D. Ozzello and L.T. Drzal, ”A Direct Comparison of
theFragmentationTestandtheMicrobondPull-outTest forDeterminingthe
Interfacial Shear Strength,” J. Adhesion, 34, 65 (1991”).

Riggs, D.M., R.J. Shuford and R.W. Lewis, ”Graphite Fibers and composites”,
Hardback of Composites, G. Lubin, ed., Van Nostrand Reinhold Co., New York,
196 (1982).

Rose, P.W. and S.L. Kaplan, Plastic Finishing and Decoration, Ch 4, Ed. Don Sata,
Van Nostrand Reinhold Company, (1986).

Rosenberg, R., Materials Research Society Proceedings 153, Pittsburgh, PA, 217 (1989).

Rueda, G., F.L. Matthews and E.W. Godwin, ”An Experimental Comparison of
Standard Test Methods for the Determination of Tensile, Compressive and Flexural
Properties of Kevlar Fibre/Epoxy Laminates,” J. Reinforced Plastics and Composites,
9, 182 (1990).

Sawyer, L.C. and M. Jaffe, ”The Structure of Thermotropic Copolyesters,” J. Materials
Science, 21, 1897 (1986).

Sawyer, L..C and D..T Grubb, in PolymerMicroscopy, Chap. 4, Chapman and Hall,
New York, (1987).

Schellekens, R. and H. Ketels, ”Preparation and Characterization of Solution(gel)-Spun
Ethylene-Vinyl Alcohol Fibers,” Polymer Camrrmnications, 31, 212 (1990).

Schoepnner, G.A. andR.L. Sierakowski, ”A review of Compression Test Methods for
OrganicMatrix Composites,” J. Comp. Tech. & Res., 12, 3 (1990).

Smith, K.J., ”The Breaking Strength of Perfect Polymer Fibers,” Polymer Engineering
and Science, 30, 437 (1990).

Smith, P.A., D,G, Gilbert and A. Poursartip, ”Matrix Cracking of Composites Inside
a Scanning Bectron Microscope,” J. Materials Science Letters, 4, 1361 (1985).

 

258
Snow, A.W., Nature, 292, 40 (1981”).
Snow, A.W., Carbon, 19, 467 (1981').

Snow, A.W., New Monomers and Polymers, B.M. Culbertson and C.U. Pittman J., eds.,
Plenum, New York, 399 (1984).

Snow, A.W., ”Thermal Characterization of Polybutadiyne,” J. Macromol. Sci. Chem,
A2200), 1429 (1985).

Streitwieser, A. and C.H. Heathcock, Introduction to Organic aremistry, Macmillan
Publishing Co., New York, NY (1981).

Sugerman, G., S.J. Monte, S.M. Gabayson, and WE. Chitwood, ”Enhanced Bonding
of Fiber Reinforcements to Thermoset Resins, ” in Proceedings ofthe SAMPE Annual
Technical Conference, Paper 2C, Minneapolis (1988).

Suzuki, Y. and M. Kusakabe, ”Surface Modiﬁcation of Silicone Rubber by Ion
Implantation,” Nuclear Instruments and Methods in Physics Research, B33, 1”
(1988).

Takata,T., ”InterfaceProblemsBetweenAramidFibersandMatrixRegionsin
Fiber-Reinforced Plastics (ArFRP),” translation from Japanese, Sen-I-Gabkasishi,
44(2). 67 (987)-

Tamikawa, M., H. Kaji and K. Ueda, Kobunshi Kagaku, 24, 385 (1967).

Tirpak, G., J. Silibia, J. Appl. Polym. Sci., 17, 643 (1973).

Tsai, SW. and HT. Hahn, ”Role of the Interface on the Strength of Composite
Materials,” Atﬂresion and Absorption of Polymers, L.H. Lee, ed., Plenum Press,
New York, 217 (1974).

Torrisi, L., L. Calcagno, and A.M. Foti, ”MeV Helium Ion Beam Etching of Poly-
TetraFluoroEthylene,” Nuclear Instrumenn and Methods in Physics Reseanoh, B32,
142 (1988).

Ueno, J., M. Watanabe and Y. Tanaka, Jpn. Patent 24,998 (1965).

Venkatesan, T., L. Calcagno, B.S. Elman and G. Foti, Ion Beam Modiﬁcation of
Insulators, Elsevier, New York, 301 (1987)

Wallet, W.R., U.S. Patent 3,779,840 (1973).

259
Walles, W.E., Amer. Gran. Soc. Meeting, Dallas, TX (1989).

Waterbury, M.C., ”The Inﬂuence of Processing, Chemistry, and Interphase
MicrosuucmreonﬂleAdhesionofCamonFiberstoThennosetandThermophsﬁe
Matrices,” Ph.D. dissertation, Michigan State University, (1991).

Werner, H.W. and A.E. Morgan, J. Appl. Phys., 47, 1232 (1976).

Wertheimer, M.R. and H.P. Schreiber, ”Surface Property Modiﬁcation of Aromatic
Polyamides by Microwave Plasmas”, J. Appl. Poly. Sci., 26, 2087 (1981).

White, J .L. , ”Historical Survey of Polymer Liquid Crystals,” J. Appl. Polym. Sci. , Appl.
Polym. W. 41, 3 (1985).

Whitney, J.M. and L.T. Drzal, ”Three Dimensional Stress Distribution ArounD an
Isolated Fiber Fragment,” in Toughened Composites, ASTM STP 937, 179 (1980).

Wilson, R.G. and G.R. Brewer, Ion Beams, John Wiley & Sons, New York, (1973).
Woodward, E., Atlas of Polymer Morphology, Hanser Publishers, NY, 266 (1988).
Wu, 8., Polymer Interface and adhesion, Marcel Dekker, Inc. New York (1982).

Wu, Y. and G.C. Tesoro, ”Chemical Modiﬁcation of Kevlar Fibers Surface and of
Model Diamides,” J. Appl. Poly. Sci., 31, 1041 (1986).

Yasuda, H., Plasma Polymerization, Academic Press Inc., Orlando, (1985).
Yates, J.B., D.J. Smith and LR. Campbell, Eur. Patent 213,466 (1987).

Yoshida, K. and M. Iwaki, ”Structure and Morphology of Ion-Implanted Polyirnide
Films,” Nuclear Instruments and Methods in Physics Research, B191”, 878 (1987).

Yue, J., A.J. Epstein and A.G. MaeDiarmid, Amer. Chem. Soc. Meeting, Atlanta, GA
(1991).

Zahr, G.E. and P.G. Riewald, ”Composite Systems Containing Aramid Fibers: An
Overview,” 44011AnnualCorﬁrence, Composite Institute, TheSocietyofthePlastics
Industry, 2E-1 (1989).

Ziegler, J.F., J.P. Biersack, U. Littmark, The Stopping and Range of Ions in Solids,
Pergamon Press, New York, (1985).