Artide
Molecular weight effects on the mechanical properties of novel epoxy thermoplastics
High Peiformance Polymers
•• • C?l
High Performance Polymers 24(3) 161-172 ©The Author(~) 2012 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOl: IO.IIn/0954008311431017 hip.sagepub.com
~SAGE
Christine D. Caroselli, Monoj Pramanik, Brandon C. Achord and James W. Rawlins 1
f
Abstract A series of epoxy thermoplastics (ETPs) with varying molecular weights were synthesized from a difunctional diglycidyl ether of bisphenoi-A (DGEBA)-based epoxy resin and an aromatic secondary diamine. The materials possessed glass transition temperatures varying between 73.61 and 85.36°C. The ETP series was characterized for fracture toughness, flexural, and compression properties. In general, fracture properties increased with increasing molecular weight and yet were consistently shown to decrease when higher molecular weight values were the result of increased branching. Flexural ultimate strength and strain-at-break increased with increasing molecular weight while flexural modulus decreased to a plateau. Compression properties were relatively unaffected by changes in molecular weight over the range of materials synthesized.
:r tt
e :i
1-
If l) ).
y tr ~r
::I ;:r
el
i;e :d :II I.
Keywords epoxy thermoplastics, chain extender, nuclear magnetic resonance, differential scanning calorimetry, gel permeation chromatography, mechanical properties
'
:~;:;s:;;
*~ .~ §
Introduction
,.
·~;t
Epoxy polymers, normally thermosets, are commonly used matrix materials for composites in aerospace, automotive, aircraft, military, and marine applications due to their high thermal stability, high modulus, high chemical/corrosion resistance, and manufacturing ease. 1'2 The ease of processability, the high number of commercially available starting materials and the variability of crosslinkers and )~>mlbin!ed potential results have been well studied in literA large volume of data supporting the tunable and :>net-te:rm performance characteristics are drivers for conuse/utility. The favorable properties are attributed .., . -·. the epoxy thermosets' highly crosslinked network. .·.·. Unfortunately, the inherent rigidity of thermoset epoxies . 'also results in a brittle matrix, which in turn, hinders its utility and limits its application. Many techniques have been explored to improve the toughness of epoxy thermosets, for . example, rubber incorporation, 3- 7 thermoplastic blending,S-10 epoxy blending, 11 - 14 and increasing molecular weight between crosslinks. 6 •8·15- 18 A negative but common ...' ._ .. occurrence is that improvement in toughness is achieved at ·~ the expense of other properties, particularly the glass tran'" sition temperature (Tg) and Young's modulus. 4·5·7·8 •15- 17 -, ·:h,:~.~·
In contrast, high molecular weight thermoplastics are characterized by high impact strength, high fracture resistance, and many possess the ability to deform plastically under high strain, namely ductile failure. 1 While epoxy thermoplastics exhibit better toughness than their thermoset counterparts, their Tg values are lower than desired for use in high performance composites. Polymers that combine the facile handling and processability, and the higher Tg of thermosetting polymers with the toughness and ductile failure modes and high fatigue resistance of a thermoplastic would be valuable to composite scientists and engineers. Very high molecular weight epoxy thermoplastics (ETPs), greater than 10 times the entanglement molecular weight (Me), are traditionally difficult to synthesize for a
School of Polymers and High Performance Materials, The University of Southern Mississippi, Hattiesburg, MS, USA
Corresponding author: James W. Rawlins, School of Polymers and High Performance Materials, The University of Southern Mississippi, 118 College Drive # 5217, Hattiesburg , MS 39406, USA Email:
[email protected]
1'''
-
162
series of reasons, foremost being the fact that side reactions . o11gomers . c . t9-2t sueh as the 1ormation of eye1IC are common, especially in reactions performed in dilute solutions, 19 while etherification occurs primarily at high reaction temperatures.19·22 Solution polymerization poses problems when accurate determination of mechanical properties is necessary in composite materials as the residual solvent(s) often plague the data by acting as a plasticizer and altering bulk properties. Consequently, such polymers have been synthesized only on a narrow and limited basis, often with high degrees of branching and broad polydispersity. Renner and Michaelis advanced a DGEBA-based epoxy oligomer using aliphatic and aromatic difunctional amines with and without solvent to obtain ETPs of number average molecular weights (Mn) in the range of9 500-33 300 g mol- 1.22 Klee and coworkers employed secondary diamines to advance a DGEBA epoxy oligomer in dilute 19 and concentrated20 solutions to yield linear epoxies with Mn values between 8000 and 18 000 g mol- 1. Blanco et al. reacted a DGEBA-based epoxy oligomer with p-toluidine to obtain an ETP with Mn of 19 400 g mol- 1.23 Kawakami and coworkers also synthesized a 26 000 g mol- 1 ETP by advancing a DGEBA-based epoxy oligomer with aniline. 24·25 Fracture toughness of epoxy thermosets has been enhanced by various techniques such as incorporating thermoplastics/toughening agents or varying crosslink density. In general, epoxy thermosets are reported to gain fracture toughness at the expense of modulus and glass transition temperature. 4·5·7 •8 •15 - 17 The approach used in the present study was to develop maximum achievable molecular weight linear epoxy thermoplastics that were fully characterized in terms of Tg, modulus, strength, strain and fracture toughness before subsequent crosslinking. In future, a series of embedded crosslinkers will be employed to crosslink the thermoplastics at varying proportions to synthesize a series of epoxy thermosetting systems. The results reported herein focus on synthesis without the difficulties discussed in brief above. High molecular weight ETPs are expected to follow established trends, namely higher toughness as molecular weight increases, resulting in greater durability during mechanical testing. It was anticipated that higher fracture toughness to correlate with increasing ETP molecular weight and ETPs to be drastically tougher than common epoxy thermosets. Bulk synthesis methods were used to eliminate the possible effects of residual solvent on mechanical properties. Characterization and mechanical properties are reported herein.
Experimental section
Materials Epon® 828, donated by Hexion Specialty Chemicals (Momentive), Inc., is a liquid resin with a reported epoxy equivalent weight (EEW) of 185-192 g/equivalent and
High Performance Polymers 24(3)
density of 1I60 kg m- 3 at 25°C. Prior to use, the EEW of Epon 828 was determined using the procedures defined in ASTM DI652-97 to be 188.72 g/equiv. PolyLink® 4200, supplied by The Hanson Group, is a liquid aromatic secondary diamine with vapor pressure < I mmHg at 20°C. All materials were used as received.
Synthesis
of epoxy thermoplastics
The epoxy oligomer, Epon 828, and amine chain extender, PolyLink 4200, were weighed into a glass jar and pla.ced on a roller mill for bulk homogenization. Chemical structures of monomers and the synthesized ETP are shown in Scheme I. The monomer blend was transferred into silicone molds designed to the dimensions prescribed for each mechanical test. The molds were placed in an inert gas oven under a nitrogen atmosphere and heated as specified in Table 1. Varying molecular weights were achieved in ETPs 1-3 via a stepwise temperature increase of the monomer blend to 120, 150, and 180°C (method A). ETPs 4-8 were synthesized by staging the monomer blend at 100°C, heating to 180°C, and reacting isothermally over different periods of time to achieve multiple degrees of molecular weight advancement (method B). A summary of monomer composition and reaction conditions of the ETPs which have been synthesized with epoxy and amine is shown in Table I.
Analysis · Differential scanning calorimetry (DSC) studies were conducted on a DSC-Q2000 (TA Instruments) to select reaction temperature and determine activation energy. A sample size of ~ 8 mg from a stoichiometric blend of Epon 828 and Polylink 4200 was weighed accurately into aT-zero aluminum hermetic pan and sealed. Samples were heated at four different heating rates, namely, 1, 2, 3, and 4°C min- 1. The activation energy of reaction between Epon 828 and PolyLink 4200 was determined using the . Ki ssmger and 0 zawa method s. 26-29
-In(!!_) = -In (ZR) + !!__ Tg E
RTp
(Kissinger equation)
where is the heating rate, Tp is the peak temperature ofthe exotherm, Z is the pre-exponential factor, R is the universal gas constant (8.314 J mol- 1 K- 1), and E is the activation energy. A plot of -tn(a;rg) ~gainst !IT~ yields a straight line from the slope of which activatiOn energy was calculated. ZE gaR
E RT
log,B =log-(-)-- c -1-(Flynn-Wa/1-0zawa equation)
where g(cx) is the conversion dependence function, c and I are couple tabulated coefficients. Variables c and I are
-
Caroselli et at.
~ EEW lefined ·Link® omatic t20°C.
Epon828
163
[>/to~~OXQo~ OH
+ IN~,~ I~NI ~ ~
PolyLink4200 ~ ~
n
H
tender, pedon 1ctures •wn In to sili,r each ~rt gas ecified ved in monoPs 4-8 md at y over ·ees of nmary of the amme
Bulk, N 2 atm
Epoxy Thermoplastic Scheme I. Synthesis of ETP. Table I. Epoxy and amine molar ratios and reaction conditions. Epoxy: amine molar ratio
1.03: 1.00
were select :nergy. end of ly into swere 3, and ~tween
ng the
on)
:ofthe iversal vation traight y was
rtion) ~and I l I are
1.02: 1.00
ETP
I 2 3 4 5 6 7 8
Reaction conditions
Method
48 h ati20°C 48 h at120°C + 48 hat 150°C 48 h ati20°C + 48 hat 150°C + 48 hat I80°C 12 h at100°C + 3.5 hat 180°C 12 h ati00°C + 5 hat I80°C 12 h ati00°C + 8 hat 180°C 12 h at100°C + 16 hat 180°C 12 h atiOOoc + 24 hat 180°C
Method A
Method B
dependent on the value of the EIRT ratio, in which c and I are 1600 and 0.4880 respectively if the E/RTvalue is in the range of 13-20.29 Activation energy of the reaction betw~n the DGEBA epoxy and secondary. diamine was also calculated from the slope of the line generated in the (og versus 1/TP plot. The Tg of each ETP was determined via DSC analysis by heating a sample at 10°C/min- 1• All DSC analyses were conducted in triplicate, and the average Tg and standard deviation were reported. Proton nuclear magnetic resonance eH-NMR) was used to verify the composition of the monomer blend at time zero and fmal structure of the synthesized ETPs. 1HNMR samples were prepared in deuterated dimethyl sulfoxide (DMSO-d6 ) at a 10% by weight solution. Proton nuclear magnetic resonance spectra were obtained using a Varian mercury NMR spectrometer operating at a frequency of 300.13 MHz with delay times and scan numbers
of 2 s and 128 scans for polymer samples, and 5 s and 64 scans for small molecules respectively. All chemical shifts (indicated as{) ppm) were referenced either automatically by the software or manually using the resonance frequency of the deuterated solvent (DMSO-d6 ). Thermogravimetric analysis (TGA) was performed on a TGA-Q500 (TA Instruments) to quantify weight loss of the individual monomers and the monomer blend as a function of time, mimicking both synthesis methods. Sample preparation for TGA analysis of each synthesis method consisted of blending the two monomers in the desired ratio into a 20 mL scintillation vial at ambient and mixing at 2400 rpm for 3 min in a DAC 150 FVZ speed mixer from FlackTek® Inc. A sample of ~25 mg was transferred into a TGA pan for testing. The ramp rates were set to 10°C min -I and isothermal holds were used to mimic both synthesis methods A and B. Method A was mimicked with an isothermal hold for 300 min at l20°C, a 120 min hold at l80°C, and a final temperature of 400°C. Method B was replicated by heating to I 00°C, holding isothermally for 300 min, heating to l80°C, holding isothermally for l80min, and heating to the final temperature of 400°C. Gel permeation chromatography (GPC) was performed on a Varian PL-GPC 50 Plus from Polymer Laboratories equipped with dual-angle light scattering, differential pressure, and refractive index detectors to determine Mn, weight average molecular weight (Mw), peak average molecular weight (Mp), and the polydispersity index (PDI) of the synthesized ETPs. Low molecular weight samples were prepared at a concentration of 5 mg mL -I in
,..,..,,.,~-:111' :·· . J
.·
i'
I I I
i I
-
164
anhydrous tetrahydrofuran (THF), while high molecular weight samples were prepared at 10 mg mL -I in anhydrous THF. The samples were analyzed at a flow rate of 0.8 mL min- 1 at 40°C through a series of four columns (three Polypore® and one 50A PLGel® columns). Toluene was employed as the flow marker (52.3 min) and the data was evaluated against polystyrene standards. All analysis was conducted in triplicate, and the average values were reported along with the corresponding standard deviation. In addition to molecular weight, GPC was utilized to detect the presence of cyclic species as well as calculate average branching numbers (B0 ). Dual-angle light scattering was used in order to calculate (from known standards) a light-scattering constant (KLs), which contains contributions from both concentration and refractive index of the solvent The branching calculation used a reference slope of radius of gyration (R~ef) versus molecular weight, where the deviation from the experimental radius of gyration (Rg) represents the branching ratio, g (Equation (1)). When the branching ratio deviates to values below unity it is an indication of cyclic oligomer formation, as the radius of gyration would be an over-estimation in relation to the molecular weight The same is true in deviations greater than unity; however, this indicates branch points and an underestimation for a given molecular weight The branching ratio was then input into equation 2, and the ternary number average branching number (Bn) was calculated. 30 Additional quaternary averages or weight averages were calculated but are not reported here.
g
=
Rg Rref
(1)
High Performance Polymers 24(3)
1.20
--1°C/min 1.00
~
E?, 3:
...
- - - 2"CCmin - • - · 3"Cimin - · · - 4"Citnln
0.80 0.00
~ 0.40 Ill
0.20 0.00
. 0
Ia!
t
~-20
35
85
135 T~rature
185
235
("C)
Figure I. DSC of Epon 828 and Polylink 4200 blend (stoichiometric) at different heating rates.
Fracture toughness properties were measured on a MTS Insight instrument Samples were prepared with a width of 15.12 (± 0.03) mm, thickness of7.57 (± 0.08) mm, and crack length of 7.15 ( ± 0.32) mm. Fracture bars were notched using a milling machine, and a crack was introduced at the tip of the notch by tabbing a razor blade. The tests were operated in flexural mode using a single-edgenotch bending (SENB) geometric arrangement according to ASTM D 5045-99, with a load cell of2.5 kN (562.02lbf) and crosshead speed of 1.30 mm min -I. The density of each thermoplastic was determined via a water submersion method using the Mettler Toledo density determination kit with Excellence XP/XS analytical balances. Sample sizes were approximately 0.2 g. Measurement was performed in triplicate, and the average values and standard deviation were reported.
g
Results and discussion (2)
Compression properties were measured on a MTS 810 Materials Test System (from MTS System Corporation) with appropriate sample clamps. Compression tests were conducted according to ASTM D 695-02a. Cylindrical samples were prepared with a length of 25.4 (± 0.01) mm and diameter of 12.7 (± 0.01) mm. The tests were conducted with a 111.2 kN (25 000 lbf) load cell at a cross-head speed of 1.27 mm min- 1 without the aid of an extensometer. Modulus of elasticity was calculated from the slope of the initial linear portion of the stress-strain curve. Compression analysis yielded modulus of elasticity, yield stress, and strain-at-yield. Flexural tests were conducted on a MTS Insight with a 10 kN (2248.09 lbf) load cell and cross-head speed of 5 mm min- 1 • Samples with a thickness of 3.2 (± 0.02) mm, width of 12.7 (± 0.03) mm, and support span of 16 times the sample thickness were tested according to ASTM D 790.
Synthesis Dynamic DSC curves of Epon® 828 and PolyLink® 4200 blend (stoichiometric ratio) at four different heating rates are displayed in Figure 1. It was observed that epoxy and secondary diamine amine start to react at 100-125°C and reach maximum rate of reaction at 185-223°C. This confirmed that the initial reaction temperature of Epon 828 with PolyLink 4200 can be set at ,...., l20°C to synthesize the ETPs. Although DSC curves were not complete for rise and fall of the reaction exotherm in the temperature range of study, it does not restrict the application of the Kissinger and Ozawa method to calculate activation energy of reaction, as it requires only the peak temperature of the exotherm where the rate of conversion is a maximum. Typical Kissinger and Ozawa plots are shown in Figures 2 and 3, respectively. Activation energy of DGEBA epoxy (Epon 828) and aromatic secondary diamine (PolyLink 4200) reaction is 62.48 kJ mol- 1 as calculated by the Kissinger method and 62.65 kJ mol- 1 by the Ozawa method. Zvetkov et al. 29 showed that low molecular weight DGEBA-based epoxy and aromatic diamines (primary) have an epoxy-amine
-
Caroselli et a/.
reaction activation energy of ~50-55 kJ mol- 1• The data reported herein is consistent with single reactions per nitrogen using a secondary versus primary amine, which is most often cited in the literature. Figures4 and 5 show the 1H-NMR spectra for the monomer blend at time zero and a synthesized ETP (ETP 3) respectively. As the PolyLink 4200 reacted with Epon 828, the intensity of terminal protons of the -CHz- on the epoxy ring (protons marked a in Figures 4 and 5) decreased as expected. Overlap occurred between the diminishing peak of the proton directly attached to the nitrogen of the PolyLink 4200 monomer (proton I in Figure 4) with the proton on the secondary hydroxyl, resulting from the epoxy ring opening reaction via nucleophilic attack of the secondary amine (proton r in Figure 5). Peaks m and n, pertaining to the aromatic protons of the amine shown in Figure 4, shifted upon reacting with the epoxy to peaks m* and n* in Figure 5. The peak at 2.48 ppm pertained to the solvent, DMSO-d6 . Moisture in DMSO-d6 appeared at approximately 3.33 ppm in both spectra . Thermogravimetric analysis of individual monomers and the monomer blend (epoxy: amine= 1.03) was conducted mimicking synthesis methods A and B to determine the degree of mass loss or volatilization that occurred during bulk epoxy thermoplastic synthesis in an open and elevated temperature environment. Figure 6 shows the TGA curves of individual monomers as well as the
UAO.-------------------, EIR= 7515.89 E=62.48kJ!mol 12.00
'· 11.20
R'1 10.80
5
of straight llne = 0.9977
+-T""""r-r~--,-~-.--,......,......,.-,-~~~.--..~-.-....,..-J
0.00205
0.002
0.0021
0.00215
0.0022
ur. :hio-
MTS lth of , and were intro. The edgerding 2lbf)
Figu~e 2. equation.
-ln(P/"r~)
against 1/TP plot from the Kissinger
0.70 -0.4880EIR=-3677.58 E= 62.65kJ!mol
0.60 0.50
a
0.40
'Of. 0.30
.E
0.20 0.10 0.00
R 2 of stralgbt llne = 0.9982
-0.10 0.002
0.00205
0.0021
via a
0.0022
1/l'p
~nsity
l ballsureralues
0.00215
165
Figure 3. log(,B) vs //TP plot from the Flynn--Waii-Ozawa equation.
f k
'4200 ~ rates ~y and C and s con·n 828 ize the seand nge of
e,m d
n
0
a
~erand
ion, as where ~er and :tively. td aroion is od and ~t al.29 epoxy -amine
a
c
c
'-'
f
...___, I
6.5
~L~l
~
I
6
5~5
b,j
~
I
4.5
Figure 4. 1H-NMR of the monomer blend at time zero.
I
4 1H (ppm)
I
3.5
I
3
1 I
2.5
I
2
I
1.5
h
'""'''- II', ,,
-
166
High Performance Polymers 24(3)
0
~
p
/~l
a
OH I'
h h
k
e,n* p,c
d
m*
o,p
r ,l
JG
\ 6.5
7
6
5.5
4.5
5
L
4
3.5
3
2.5
1.5
2
1H (ppm) I
,I
Figure 5. 1H-NMR of synthesized ETP 3.
)
120
I
100 &
'\
I
25.98%1 27.80% 80
.. --
..
·~:: 1.:.:-; -~/'.
~{{
::i:
ilii 'I' I
i',d:'
i'I''~ :j
--
~
I
~
-~
·a;
~
40
"
\
3lt \
74.02%
98.5 I
\
'._ __ --- ......
200
250
300
75
\
350
100 Tem 125 15PCJ 175 perature
---
'
--------
0
150
0.9372%
I ------
\97.550 \ \
100
.......
\
\
20
-----
0.2275% "ir_, ___
57.90°
\
\
50
~.901%
\
I:
-20
I
...
\
l: C)
G.7887%
·;;;
',
60
p~t-
~.5 s=
I
Blend (r =1.031- Method A Blend (r =1.03J • Method B
f:___
100.5
I
400
Temperature (•C)
450
20l••.ll
PolyLink 4200 Epon828 Blend (r=1.03) ·Method A Blend (r=1.03) ·Method B
500
550
600
650 700 Universal V4
Figure 6. TGA Polylink 4200, Epon 828, and monomer blends mimicking synthesis methods A and B for volatility of raw materials over time and conversion.
~
-
Caroselli et a/.
167
Table 2. GPC results of synthesized ETPs. ETP
Mn (g mol- 1)
I
II 110 15800 18120 7440 10950 14281 15950 16920
2 3 4 5 6 7 8
± ± ± ± ± ± ± ±
170 290 830 100 390 280 740 200
Mw (g mol- 1) 19060 35540 55460 11750 18680 25470 32270 38820
Mp (g mol- 1)
± 330 ± 1440 ± 4850 ± 160 ± 710 ± 500 ± 1260 ± 620
17050 22600 25 110 11720 16820 21 180 23540 24900
± 230 ± 380 ± 900 ± ISO ± 430 ± 370 ± 630 ± 380
Maximum branching
POl 1.72 2.25 3.06 1.58 1.70 1.78 2.02 2.29
± ± ± ± ± ± ± ±
0.01 0.05 0.15 0.00 0.01 0.01 0.02 0.03
0 0 3.2 0 0 0 0 >0
1.4 1.3 1.2
> g :c 0
~
,a:..
.e .!:1 0
z
1.1
0.9 0.8 0.7 0.6 0.5 04 0.3 0.2 0.1 1000
10000
100000 Extrapolated MN (Daltons)
1e6
2.5e6
Figure 7. GPC chromatogram overlay of method A ETPs in increasing order of molecular weight (ETP 1-3).
over
monomer blend. PolyLink 4200 lost its mass nearly 26 and 74% during isothermal holds at 120 and 180°C, respectively, while Epon 828 did not lose any weight at l20°C but did lose 28% during the isothermal hold at 180°C. The TGA curves also showed that the isothermal hold at l20°C resulted in 1.90% weight loss of the monomer blend and was attributed to amine volatilization as noticed from the PolyLink 4200 TGA curve. After acquiring the data above and to reduce amine mass loss from volatilization, a modified synthesis method was evaluated and the initial reaction temperature stage was reduced to l00°C. Data confirmed that < 1% by mass of the monomer blend was lost during the isothermal hold at 100°C. ··.. ;':)j$:;Was .observed from DSC analysis that the epoxy~e reaction between Epon 828 and PolyLink 4200 commences at "'ll0°C. Accordingly, the bulk reaction between Epon 828 and PolyLink 4200 (1.03 : 1.00 molar ratio) was initiated at l20°C. This blend was allowed to react further at 150 and 180°C to achieve higher molecular weights (Method A). A mass loss of 1.9% was noticed via thermogravimetric analysis (mimicking method A) of an Epon 828-PolyLink: 4200 blend (1.03 : 1.00 molar ratio) during cure and attributed to amine volatilization. Additionally, it was confirmed that the longer reaction period (48 h) at 180oc resulted in branching in the linear thermoplastic (discussed later). To reduce the mass loss and shorten the reaction period, Method B was developed (initial
temperature l00°C for 12 hand further heated at 180°C for varying periods of3.5 to 24 h). Table 2 summarizes the GPC results of each ETP. For example, GPC chromatograms of ETPs 1-3 are displayed in Figure 7. The range of Mn, Mw, Mp, and PDI values for the synthesized ETPs were in the range of 7440--18 120 g mol- 1, 11 750--55 460 g mol- 1, 11 720--25 110 g mol- 1, and 1.58-3.06, respectively. It was observed that molecular weight increased with increasing temperature from l20°C/ 48 h to 150°C/48 h to 180°C/48 h as expected, and the authors attribute this consistent rate to less diffusion limits as the system is well above the final material glass transition temperature. ETP 3 possessed Mw and PDI values much higher than others, with Mw greater than 50 000 g mol- 1 and PDI > 3. This particular sample was the only one to experience prolonged exposure (48 h) to the highest reaction temperature of 180°C. The bulk conditions, Tg, and reaction mechanism each add to a broadened molecular weight distribution when conditions are focused on driving the highest attainable molecular weight. The results from the light-scattering data are given in Table 2. Branching was detected in the ETP with high Mw and large PDI (ETP 3), with a maximum ternary number average branching number of 3.2. Deviation from the reference radius of gyration was not evident in any other molecular weights other than the Mn range shown in Figure 8. ETP 8 showed slight scattering in the high
-
168
High Performance Polymers 24(3)
3.4
2.7
.....E
2.0
z"
"'
.5
.c u
:;
1.4
.n
0.7
0.0
ETP 1,2,4-7
I
-0.7 244771.4
268430.6
292089.8
363067.4
339408.2
3157 49.0
Mol WI
:.-1
Figure 8. GPC light-scattering curves indicate the presence of branch points in ETP 3. The remaining ETPs shown lie directly along the x-axis (negligible branching or cyclization).
Table 3. fg values of ETPs. ETP
3
2
Tg ("C) 76.36
±
1.19 78.55
±
0.88 85.36
±
0.11
73.61
±
molecular weight region, but the resulting branching numbers were sporadic and are not reported here. It is assumed that branching (attributed to etherification) began occurring here when the ETP was maintained at 180° C for 24 h. Also, the branching ratio, g, from each Rg and reference R~ef was never below unity, indicating cyclic oligomer formation was negligible.
I,I
5
4 0.25 77.22
±
0.20 79.15
±
0.43 78.46
86 84
I
I
! )
:i
:I
r r
'I
i.·
.,
·jr
.•
:\j,
'il
' il..£ ,,'.:/·
!
i
0.89 78.85
+Method A
±
0.86
•
•MetbodB
- 82 tJ
'
f-'lla:ving M 0 of II II 0 g mol- 1 and fracture toughness )'~#fi,Q~,~~~a m 112, possesses sufficient strength to endure .. ·· ··~•· ~no(~~gi,and produce fracture toughness specimens. This indicated that the epoxy thermoplastic (Epon 828 & PolyLink4200)molecular weight threshold for toughness must be "'11 000 g mol- 1 for sample preparation without premature failure. Figures 12 and 13 reveal the flexural strength, modulus, and strain-at-break for these ETPs. The results ranged from 9.65-93.25 MPa, 2.26--4.83 GPa, and 0.29-2.86%, respectively. Ultimate strength and strain-at-break consistently increased with increasing molecular weight while the modulus decreased and plateaued as molecular weight increased. The trends correlate well with the improved fracture properties correlating with higher linear molecular
7000
9000
11000
13000
15000
17000
19000
.I.VJ,.(glmol)
Figure 13. Flexural properties (strain-at-break) dependence on ETP molecular weight for synthesis methods A and B.
weight. When a given polymer exhibits higher toughness, the material is also expected to exhibit lower stiffuess and greater elongation prior to breaking. ETP 2 also exhibited the highest flexural strain-at-break, which correlates with it also being the toughest ETP as seen in fracture testing. ETPs I and 5 were synthesized using different conditions, yet resulted in similar molecular weights and flexural properties. ETPs 2 and 7, also of comparable molecular weights achieved through alternate methods, resulted in similar modulus values. ETPs 3, 7, and 8 surpassed the strain limit set for the instrument (based upon specimen dimensions) and did not break during the testing. Compression properties (yield strength, modulus, and strain-at-yield) of the synthesized ETPs ranged between 54.74-81.60 MPa, 1.50-2.47 GPa, and 4.44-6.01%, which are displayed in Figures 14 and 15, and were generally unaffected by increases in ETP molecular weight. ETP density seemed to be unaffected by molecular weight, as shown in Table 4, samples of the same structural composition resulted in similar volumes/densities regardless of varying molecular weight influences. Albeit, the molecular weight range synthesized was not as broad as desired, the compression properties were similar regardless of resulting
.·.;,
-
170
8.00
10.00 100
I I
~80
~
+Yi
