Upload
nguyenkhanh
View
218
Download
1
Embed Size (px)
Citation preview
LLNL-CONF-578735
Analytical Approaches TowardsUnderstanding Structure-PropertyRelationships in End-Linked Model PDMSNetworks
J. P. Lewicki , R. L. F. Albo, C. T. Alviso, M.Ashmore, S. J. Harley, J. A. Finnie, R. S. Maxwell
September 4, 2012
MODEST 2012Prauge, Czech RepublicSeptember 2, 2012 through September 7, 2012
Disclaimer
This document was prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore National Security, LLC, nor any of their employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States government or Lawrence Livermore National Security, LLC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or Lawrence Livermore National Security, LLC, and shall not be used for advertising or product endorsement purposes.
LLNL-PRES-XXXXXXThis work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344. Lawrence Livermore National Security, LLC
James P. Lewicki, Rebecca L. F. Albo, Cynthia T. Alviso, Jasmine Finnie, Stephen J. Harley, Michael Ashmore and
Robert S. Maxwell
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx2
Δαβɣ-H2OLOAD
Bulk
Netw
ork structure
Chem
ical identity
Structural complexity over a range of size-scales
Complex engineering silicones
Aging and degradation mechanisms
operate over a broad range of
size scales
Macro
Micro
Molecular
Suffer aging and degradation
Aging and degradation over a range of size-scales
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx3
Solid state NMR
Therm. Ana.
Mech. Ana. BDS sorption
studies
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx4
Pyrolysis-Gas Chromatography/Mass Spectrometry
Interrogative thermal analysis of siloxane network architecture
• Sample• He carrier
gas
Probe head
• Δ• Analytes
evolvedFurnace
• Analyteseparation
• MS detection
GCMS
Small sample size (0.05mg):
• Non-diffusion limited regime
Versatile probe designs:
• ‘low & slow’ out-gassing • thermal degradation
Cryo-GC/MS:
• High sensitivity• Efficient separation
• broad mass detection
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx5
What Information can Py-GCMS provide?Pyrolysis-evolved gas analysis (Py-EGA)
(20 ⁰C/min) with GC column in a non-eluting mode, Carbon NT/PDMS composites
250 500 750 10000
1000000
2000000
3000000
4000000
5000000
6000000
7000000
8000000
9000000
MS
D R
espo
nse
Temperature /oC
CA 0_2 (0% SWNT) CA 1_2 (6% SWNT) CA 2_2 (13% SW NT) CA 10_2 (63%% SW NT)
Py-EGA gives us ‘bulk’ degradation profiles:
• Onset & peak degradation temp.
• ‘Yield’ of volatile degradation products
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx6
5 10 15 20 25 30 350
2000000
4000000
6000000
8000000
10000000
12000000
14000000
16000000M
SD
det
ecto
r res
pons
e
Retention time (min)
Unfilled
(1000 ⁰C/min) with GC column in an eluting mode
Ballistic PY yields chemical information
• Speciation of degradation products
• yields of individual species
• Informs mechanistic models
What Information can Py-GCMS provide?Ballistic pyrolytic analysis
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx7
Siloxanes thermally degrade via well established chemistries to form a series of cyclic oligomers
We’ve shown that network structure can influence the stability of a PDMS network in the bulk*.
?
1) Lewicki JP, Mayer, BP, Alviso CT, and Maxwell, RS.” Journal of Inorganic and Organometallic Polymers and Materials (2012) 22:636–645).
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx8
OSi
O
SiO
Si
1 Lewicki JP, Mayer, BP, Alviso CT, and Maxwell, RS.” Journal of Inorganic and Organometallic Polymers and Materials (2012) 22:636–645).2) Lewicki JP, Liggat JJ and Patel M. Polymer Degradation and Stability 2009, 94, 1548-1557.
Using Quantitative GC-Pyrolysis we have shown that the D3 cyclic remains the principle product of degradation, irrespective of network
structure1,2
Studied Architectures:
• Mono-modal• Bi-Modal• Free-chain end systems• Above and below
entanglement• (liquid precursors)
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx9
• The thermodynamic product of thermal degradation in PDMS is not sensitive to the subtleties of network architecture.
• Using a statistical Analysis technique we can examine the relationship between network architecture and minor product formation
• We observe clear clustering of degradation profiles as a function of network structure
Quantifiable relationships exist between basic network architectures and the distributions of degradation derived species in PDMS networks.
Lewicki JP, Mayer, BP, Alviso CT, and Maxwell, RS.” Journal of Inorganic and Organometallic Polymers and Materials(2012) 22:636–645).
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx10
Our analysis suggests that at shorter chain lengths/high x-link densities larger cyclicsare favored. However at low x-link densities and long chain lengths smaller cyclicsdominate
A) Short chains high X-link density
B) long chains low X-link density
• The additional constraint of high x-link density networks makes the formation of larger cyclics somewhat more conformationalyfavorable
• In low x-link density systems above entanglement random coil behavior dominates and the majority of a chain does not ‘see’ a x-link site, the main driver is now thermodynamics and small cyclic formation dominates
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx11
• Analytical pyrolysis coupled with PCA can map the degradation profiles of chemically similar engineering silicone elastomers
• Materials group by Network modality and catalyst chemistry
• Neither aging or irradiation significantly alter a materials pyrolysis degradation profile.
Lewicki, JP, Albo RLF, Alviso, CT and Maxwell, RS. “Pyrolysis-gas chromatography/mass spectrometry for the forensic fingerprinting of silicone engineering elastomers” Journal of Analytical and Applied Pyrolysis (accepted)
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx12
• As with the model systems, architectural features such as phenyl free chain ends or high x-link densities ‘skew’ the profile of minor degradation products providing a fingerprint
Lewicki, JP, Albo RLF, Alviso, CT and Maxwell, RS. “Pyrolysis-gas chromatography/mass spectrometry for the forensic fingerprinting of silicone engineering elastomers” Journal of Analytical and Applied Pyrolysis (accepted, in press)
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx13
The synthesis of well defined end-linked model PDMS networks allows us to decouple and investigate the effects of individual architectural motifs on the properties of a elastomer system
X-link density, chain length, modality and level of FCE’s all influence the degradation chemistry of a silicone material
Here we have demonstrated the utility of interrogative thermal analytical analysis methodologies for the study of otherwise intractable materials
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx14
Thank you.
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx15
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx16
Matrix of data
Variables
Sam
ples
(n)
(m)
(X)
Principle component analysis
Small number of new variables or “components” that maintain structure of the data (i.e., the basic information) but that can better capture the data’s variance.
These components give information on:-How samples are (or aren’t) related to one another-The degree to which variables influence the overall structure of the data.
COMPLEX = scale, similarity, etc.Examples of variables (n): Pressure, temperature, absorbance, intensity, etc.Examples of samples (m): Data samples at specific time points, # of samples taken in a study.
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx17
Ex.: Elliptically distributed x-y-z data.
Variable 2Variable 1
Varia
ble
3
Component 2
Component 1
Principle Component 1:- Accounts for the largest variance (spread) in the data.
Principle Component 2:-Orthogonal to PC1, accounts for next largest variance.
Principle Component 3:-Direction forced by # of variables (3)-Not necessary!!! – little variance in last remaining new dimension
Dimension Reduction by discarding unnecessary components…
…unimportant ‘variables’…spectral noise…etc.
PCA can also identify outliers and discriminate between outliers within the model and outside the model.
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx18
Loadings identify the variables that contribute to the grouping and/or separation.
Scores describe the relationship of the samples – how the samples group to (or separate from) each other.
Matrix of data
Variables
Sam
ples
Scores
AAA
AB
B B
CC
CC
D
DD
D
*
13
14 15
16
17 12
18
1920
21
22
Loadings
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx19
-20000 0 20000 40000 60000 80000100000120000140000160000180000
4.5
4.6
4.7
4.8
4.9E
ps'
Time [s]
Model ExpDec1Equation y = A1*exp(-x/t1) + y0
Reduced Chi-Sqr
1.91314E-6
Adj. R-Square 0.99935Value Standard Error
Eps' y0 4.48995 3.10282E-4Eps' A1 0.41781 0.00103Eps' t1 51045.60943 189.84508
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx20
Purity & polydispersity of the starting materials
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
0.0
0.5
1.0
1.5
2.0 68K 8K 10K 32K 54K 136K
WF/
dLog
M
log(mol Wt.)
Si O Si
CH3
CH3
O OH
CH3
CH3
Si
CH3
CH3
OHSi[O(CH2)2CH3]4+
SnII[O2C(C2H5)CH(CH2)3CH3]
n
+CH2
CH2
CH3OH
x4
Hydroxy-terminated PDMS Crosslinked PDMS
2H2O R-Sn(II)[OH]x + OH
O
CH3
CH3
Tin(II)ethylhexanoate
Si-Si-O-PDMS-O-SiO
OSi-O-PDMS-O-Si-
OSi-O-PDMS-O-Si-
-Si-O-PDMS-O-SiO
Active catalyst
Ethylhexanoic acid
Tetrapropylorthosilicate (TPOS) crosslinker
Unwanted side reactions
Complex cure kinetics
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx21
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.00.00000
0.00002
0.00004
0.00006
0.00008
0.00010
0.00012
0.00014
0.00016
0.00018
0.00020
0.00022
0.00024
Ave
. XLD
stoic eqv. on silane
0.0
0.2
0.4
0.6
0.8
1.0
Sol.
frac
tion
2 3 4 5 6 7 8-0.000005
0.000000
0.000005
0.000010
0.000015
0.000020
0.000025
0.000030
0.000035
0.000040
0.000045
0.000050
Ave
. XLD
stoic eqv. on silane
0.0
0.2
0.4
0.6
0.8
1.0
Sol.
frac
tion
0 1 2 3 4 5 6 7 8
0.000000
0.000002
0.000004
0.000006
0.000008
0.000010
0.000012
0.000014
Ave
. XLD
Stoic eqv. on silane
0.0
0.2
0.4
0.6
0.8
1.0So
l fra
ctio
n
Assessment of crosslink density with the application of F-H network theory.
~6000 g mol-1 ~28,000 g mol-1
~115,000 g mol-1
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx22
4.6 4.7 4.8 4.9 5.0
0
10000
20000
30000
8:17.5:1
6.5:16:1
5.5:15:1
ppm
1:1
5.0 5.5 6.0 6.5 7.0 7.50
5000
10000
2:1
ppm
1:1
1H MAS – Can directly observe the excesses of functional groups in partially gelled systems
Decreasing vinyl signal
Increasing levels of Silanol FCE’s
Magic Sandwich Echo allows protons associated with regions of differing
mobility to be assessed
Under-cure Vinyl
excess
Fully cured
network
Under-cure silane
excess
Processed data yields distributions of protons with distinct relaxation times
A homogenous network structure is approached
=> distinct pops. converge
The progress of a structure towards a homogenous, ordered network can therefore be followed using convenient, non-invasive and easy to implement solid state NMR
methodologies
Maus et. al Macro. Chem. Phys. 207, 13, 2006
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx23
Water sorption studies yield insight into the dynamics
of siloxanenetworks
• In contrast to commercial and end-use formulations, the uptake behavior of the optimized model systems is linear and single component.
• Residual functionality and void volume in Sylgard-184 and the LLNL end-use formulation (M97) lead to significant Langmuir and pooling sorption modes – absent in the model system.
• In the single component Henrys law behavior of the model network, we are observing the fundamental sorption behavior of the fully dense network
Void occupation
Interaction withFG’s
Simple networkoccupation
LLNL’s M97
Sylgard 184
Monomodal Model Network
Lawrence Livermore National Laboratory LLNL-PRES-xxxxxx24
0.01 0.1 1 10 100 1000 10000
0.0
0.2
0.4
0.6
0.8
1.0 68K Mono 8 & 68K Bimodal 68K 20% FCE's
T2 /ms
Solid state NMR can be tasked as a powerful tool for the analysis of otherwise intractable polymeric network materials
Basic T2 measurements have been shown to be
sensitive network changes in the solid state
Multiple quantum NMR yield NMR derived distributions of MW through <DQ> - the residual
dipolar coupling constant. MQ NMR is also sensitive to chain entanglements and Mc
Mayer, Lewicki ,Weisgraber et al. MACROMOLECULES 44 8106 2011