Você está na página 1de 10

Polymer Degradation and Stability 90 (2005) 471e480

www.elsevier.com/locate/polydegstab

Investigation of the hydrothermal stability of cross-linked


liquid silicone rubber (LSR)
Afshin Ghanbari-Siahkali a,*, Susanta Mitra a, Peter Kingshott a,
Kristoer Almdal a, Carsten Bloch b,
Helle Kem Rehmeier b
a

The Danish Polymer Centre, Ris National Laboratory, Building 124, PO Box 49, Frederiksborgvej 399, DK-4000 Roskilde, Denmark
b
Research Department, Grundfos Management A/S, Poul Due Jensens Vej 7, DK-8850 Bjerringbro, Denmark
Received 11 February 2005; received in revised form 7 March 2005; accepted 11 April 2005
Available online 14 June 2005

Abstract
We investigate the hydrothermal stability of cross-linked liquid silicone rubber (LSR) in water at 100  C up to period of two
years. Optical microscopy of cross-sections of the exposed samples reveal that only the outer 100 mm of the surface layer is aected
after two years. However, the surface chemistry of the material after prolonged exposure becomes signicantly modied, as
monitored by X-ray photoelectron spectroscopy (XPS) and attenuated total reectance Fourier transform infrared (ATR-FTIR),
which probes depths of 10 nm and 1 mm, respectively. In addition, changes to the bulk physical properties of the rubber samples,
prior to and after the exposure, were investigated by thermogravimetric analysis (TGA) and dierential scanning calorimetry (DSC).
Micro-hardness analysis showed that surface roughness of the two year exposed sample increased from 60 (IRHD) to 75 (IRHD).
Furthermore, the volume change (%) measurement showed a signicant decrease in the course of exposure at prolonged time. The
results provide the experimental basis for development of LSR materials suitable for numerous technical applications.
2005 Elsevier Ltd. All rights reserved.
Keywords: Chemical degradation; Liquid silicone rubber (LSR); Surface hardness; XPS; ATR-FTIR; TGA; DSC; Physical properties; Volume
change

1. Introduction
The essential physiochemical properties of silicone
rubber materials in terms of excellent weatherability and
thermal stability, oxidation resistance, low-temperature
exibility, and good dielectric properties, as well as low
production costs, make this class of materials very
attractive candidates for use at elevated temperatures in
liquid environments [1,2]. It is reported that the thermal
stability of silicone rubber could be improved even to

* Corresponding author. Tel.: C45 46 77 4795; fax: C45 46 77 4791.


E-mail address: afshin.ghanbari-siahkali@risoe.dk (A. Ghanbari-Siahkali).
0141-3910/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymdegradstab.2005.04.016

temperatures higher than 300  C in air by substitution


of the methyl groups in the main building block of the
rubber backbone (e.g., polydimethylsiloxane, PDMS)
with phenyl groups [3,4]. Nevertheless, these silicone
rubbers have limited hydrothermal stability, which
limits their utilities at service to a maximum temperature
of 120  C in the presence of water, steam, and acidic or
alkaline environments.
The aim of the present study is to examine the
hydrothermal stability of a commercial grade crosslinked LSR sample exposed to deionised water at
100  C for prolonged time (104 weeks), using a combination of surface analysis techniques and mechanical
tests.

472

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480

2. Experimental
2.1. Materials
The cross-linked LSR used in this study is a
two-component commercial grade (Elastosil) rubber
produced by Wacker-Chemie GmbH in Germany. The
representative chemical structures of the LSRs key
constituents used in this study (Scheme 1) consists of
siloxane cross-linkers (component A) and siloxane
(component B) in a mixing ratio of A:B Z 1:1. Silica
was already added as a ller (z25%) in component B.
This rubber was press cured by utilizing a Pt-based
catalyst rst at ambient temperature in 5 min and then
at 170  C followed by post-curing at 200  C for 4 h. The
samples were cut into rectangular slabs of dimensions,
2 mm ! 5 mm ! 100 mm, that were used for aging
studies. PDMS sample with a viscosity of 300 000 cSt
was purchased from ABCR and used as received. The
surface treated (silane) silica was kindly provided by
AVK Gummi A/S e Denmark.
2.2. Aging and characterisation methods
The LSR samples were immersed in deionised water
(70 ml) in closed Pyrex glass bottles (100 ml). The
bottles were then placed in an oven at 100  C and kept
for a long period of time (104 weeks). The aged samples
were tested for physical properties including volume
change (%) according the ISO 1817, 3rd edition

1999-03-01 and micro-hardness (IRHD) according ISO


48, 3rd edition amendment 1, 1999-08-15. The surface
topography before and after the exposure was monitored using a Zeiss LSM 5 laser-scanning microscope
(Carl Zeiss, Inc., Germany) using a Zeiss !10 Epiplan
Neouar lens (NA 0.3), !20 Epiplan Neouar lens
(NA 0.5), or !50 Epiplan Neouar (NA 0.75) objective
lens.
The surface chemistry of the samples was monitored
using XPS and ATR-FTIR. XPS analysis was performed on all samples using a Sage 100 (SPECS, Berlin,
Germany) operated at a pressure of !107 Torr.
Analysis was performed using a unmonochromated
MgKa X-ray source operated at a power of 300 W at
a take-o angle of 90  , resulting in a maximum probe
depth of around 10 nm. Sample exposure was minimised
to prevent X-ray induced damage. Atomic concentrations of each element were calculated by determining the
relevant integral peak intensities using a linear type
background. The systematic error is estimated to be
in the order of 5e10%. The binding energies were
corrected by referencing to the oxygen component of
SieO (O1s Z 532.0 eV) for maintaining the consistency
of all the samples [5]. High resolution XPS analysis on
carbon (C1s), oxygen (O1s) and silicone (Si2p) was also
carried out to monitor the change in the chemical
association of the atoms under probe. Curve tting
techniques were utilized which rely on least-square
minimization routine in the software. Full width at half
maximum (FWHM) for C1s and O1s are constrained to

Scheme 1. The representative chemical structures of the LSRs key constituents.

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480

1.8 eV and 2.0 eV, respectively, to maintain consistency


in curve tting. For Si2p the tting protocol was set on
the doublet separation (2p3/2 and 2p1/2) at 0.61 eV with
an area ratio of 2:1 [5].
ATR-FTIR measurements were performed on
a Perkin Elmer Instruments (Spectrum One) and run
with 32 scans and a resolution of 4 cm1.
Thermal stability of the exposed sample was assessed
by TGA using a TA Instruments (Q 500). The samples
were subjected to a constant heating rate (10  C min1)
from ambient temperature to a nal temperature of
1000  C under N2 environment with a constant ow rate
of 90 ml min1. DSC was conducted on a TA Instruments (Q 1000), over the temperature range of 165  C
to 100  C at a constant heating rate of 10  C min1
under N2 environment with a constant ow rate of
50 ml min1. The weight loss (%) in TGA and phase
transitions in DSC (i.e., glass transition temperature, Tg,
at the inection point of the heat ow vs. temperature,
melting temperature, Tm, at endothermic peak, and
crystallisation temperature, Tc, at the exothermic peak)
in DSC analysis was determined via the TA Instruments
Universal Analysis 2000 software (version 4.0).
In the course of all analysis, a control sample of LSR
(not exposed) along with the PDMS and surface treated
silica (SiO2) were also studied, here, for comparison
purposes.

473

with a pure PDMS sample and a sample of surface


treated silica ller typically used for rubber compounding for comparison purposes. In addition, Table 1 shows
the corresponding atomic ratios (Si/O, O/C, and C/Si)
for these samples.

3. Results
3.1. Microscopy
The initial visual inspections showed that the exposed
sample after 104 weeks has become somewhat rigid and
opaque in contrast to exible and semi-transparent
control sample. Optical microscopy was used to see the
extent of modication of the exposed sample. Fig. 1
shows the optical images of control, exposed (water at
100  C for 104 weeks), at both surface and cross-section.
The results clearly show the dramatic changes over
time compared to the fairly smooth surface of the
control sample. The cross-section image of the exposed
sample clearly shows (Fig. 1c) that some portion of the
rubber materials has been aected. The depth of the
aected region was measured to be around 100 mm (as
indicated by arrows).
3.2. XPS
XPS is a surface sensitive method that probes the
composition to a depth of about 10 nm, thereby allowing for the elucidation of interfacial chemical degradation. Fig. 2 illustrates high-resolution XPS spectra of
LSR rubber samples for the C1s, O1s, and Si2p regions
prior (designated control) and after exposure, along

Fig. 1. Optical images of LSR sample (a) control, (b) exposed, and
(c) cross-section of exposed to water at 100  C for 104 weeks (the
arrows highlight the aected portion due to the exposure in close
proximity to the sample surface, z100 mm).

474

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480


285.2 eV
Pure PDMS
LSR-Control
LSR-Exposed
Surface treated Silica

Normalized Intensity (a.u.)

1,0

Table 1
Atomic ratio from XPS analysis for the studied samples

284.5 - 285.7 eV

0,6

0,4

288

O/C

C/Si

0.79
0.61
1.00
0.93

0.56
1.61
0.50
0.55

2.26
1.04
2.00
1.98

286

284

282

Si

Binding Energy, eV

CH3

532 eV

1,0

Normalized Intensity (a.u.)

Si/O
LSR-control
LSR-exposed
Theoretical LSR
Pure PDMS

CH3
290

Pure PDMS
LSR-Control
LSR-Exposed
Surface treated Silica

0,8

Repeating unit in LSR

0,6
0,4
0,2
0,0
535

534

533

(b)

532

531

530

529

528

Binding Energy, eV
103.5 eV

102.6 eV
102.4 eV

104.2 eV

1,0

Normalized Intensity (a.u.)

Atomic ratio

0,2

(a)

Pure PDMS
LSR-Control
LSR-Exposed
Surface treated Silica

0,8

0,6

0,4

0,2

0,0
108

(c)

Sample

0,8

106

104

102

100

Binding Energy, eV

Fig. 2. XPS high-resolution scans of (a) carbon, (b) oxygen, and


(c) silicone for the studied samples.

After 104 weeks exposure the LSM rubber has


undergone a dramatic chemical change as indicated by
the nearly three-fold increase (up to 1.61) in O/C
compared to the control sample (O/C Z 0.56). On the
other hand, the C/Si ratio is decreased by about half
from 2.26 to 1.04, whereas the Si/O ratio is reduced to
only 23% of the original value (0.79) after exposure
(Table 1). The atomic ratios (Si/O, O/C, and C/Si) for
pure PDMS are very close to those of the unexposed
control sample. The minor dierences could possibly be
due to the presence of additives in the control sample.

However, deviations from the theoretical stoichiometry


(C:Si:O Z 2:1:1) of the control LSR is probably due to
the adventitious hydrocarbon surface contamination.
After exposure the XPS results clearly point towards
the fact that the LSR sample has undergone severe
oxidation with relatively high siliceous containing
species detected at surface.
In order to assess the extent of degradation into the
bulk, XPS analysis was performed on a cross-section
of exposed sample (not shown). An identical elemental
composition to the unexposed bulk suggests that no
chemical changes had occurred in the bulk during
exposure. Thus, these ndings clearly show that the
vulnerability (i.e., chemical degradation) of the rubber
system to the exposure media is only conned to the
surface.
In order to probe changes in the chemical environments surrounding the carbon (C1s), oxygen (O1s) and
silicone (Si2p) atoms after exposure, high-resolution XPS
was performed. Curve tting is undertaken by referencing
the charge correction as SieO peak at 532 eV for all the
samples in order to maintain the consistency and thus
results obtained for dierent samples are comparable on
the same reference scale. Pure PDMS was also analysed
under identical experimental conditions in order to
correlate the results against known binding energies for
the expected chemical shifts for each of the elements. A
sample of surface treated silica (which is a typical
reinforcing ller for rubbers) was also analysed under
the same condition in order to compare the result
particularly for the exposed LSR sample. Normalized
C1s, O1s, and Si2p spectra for all the samples are shown
in Fig. 2(aec). The curve tting results are also
summarised in Table 2. High-resolution C1s, O1s, and
Si2p spectra clearly reveal that the pure PDMS and the
control LSR have almost identical molecular features
(Fig. 2) and this is also, clearly, reected in the data
analysis given in Table 2.

475

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480


Table 2
High-resolution XPS data analysis for the studied samples
Sample

Region

Peak position (eV)

Possible structure

FWHM (eV)

Pure PDMS

C1s
O1s
Si2p3/2

284.8 (100%)
532.0 (100%)
102.4 (66.7%)

CeSi
SieO
SieO

2.0
2.0
1.7

LSR-control

C1s
O1s
Si2p3/2

284.7 (88.0%)
532.0 (90.8%)
102.6 (66.7%)

CeSi
SieO
SieO

2.1
2.2
2.0

LSR-exposed

C1s

284.5
285.9
532.0
533.0
103.5

(85.0%)
(15.0%)
(75.8%)
(24.2%)
(66.7%)

CeSi
CeO
SieO
CeO
SieO

2.0
2.2
2.1
2.3
2.7

285.2
286.8
288.7
532.0
533.9
104.2

(69.2%)
(23.5%)
(15.0%)
(94.1%)
(5.9%)
(66.7%)

CeC
CeO/CeN
OaCeO/OaCeN
SieO/CaO
CeO
SieO

2.5
2.1
1.7
2.6
2.9
2.1

O1s
Si2p3/2
Surface treated silica

C1s

O1s
Si2p3/2
a

Values in parenthesis represent the respective area percentages under the peak.

After exposure, there is a clear, yet, small change in


the C1s spectra where another new CeO species at
285.9 eV emerges as a consequence of degradation
(Table 2). This observation is once again consistent
with O1s spectra of the exposed sample, where a new
peak evolves at 533.0 eV due to CeO.
In order to obtain more detailed information, the Si2p
curve tting was undertaken. With a tting protocol of
doublets (2p3/2 and 2p1/2) area ratio of 2:1 at a separation
of 0.61 eV [5], the curve tting yielded results as shown in
Table 2. Both, Table 2 and Fig. 2(c) show that there is
a very signicant broadening of the Si2p peak due to
degradation upon prolonged exposure. The almost
identical Si2p spectra for both control LSR and PDMS
control samples indicate that they have similar molecular
structures. One could argue that the presence of silica
ller (SiO2) in the processed control LSR sample might
lead to some contribution to the Si2p spectrum,
however, it appears that the particles are either covered
by LSR, or probably buried too deep (O10 nm) inside
the material to be detected by XPS. Upon exposure
a three-fold increase in the O/C ratio, two-fold decrease
in C/Si, and 23% decrease in Si/O ratio, in combination
with the signicant change in the high-resolution study
at Si2p, reveal that surface erosion of the rubber makes
the surface more enriched in SiO2. The Si2p spectrum for
surface treated silica supports this postulation. The LSR
degradation occurs from the siloxane linkages (SieO)
undergoing attack by water, leading to chain scission.
The shifting of the Si2p spectrum of the LSR upon
exposure from 102.6 eV to 103.5 eV, which is getting
closer to Si2p of the surface treated silica (Fig. 2(c)),
further consolidates this postulation [6]. Broadening of
the Si2p peaks for the exposed sample is most likely due

to a contribution from both the fragmented rubber


chains and SiO2, which cannot be resolved further.
The decreased intensity of the C1s spectra for the
exposed LSR in Fig. 2(a) is also attributed to the
inhomogeneous surface enriched in SiO2 as a manifestation of chain scission of LSR upon hydrothermal
degradation and thus signal is attenuated. For the
surface treated silica very low signal intensity of C1s
spectra is plausibly due to the presence of very thin layer
of organic chemicals on the surface, which has been used
for surface treatment of the silica ller. During curve
tting the contribution from nitrogen was included
(Table 2) considering its presence (w2.4%). Furthermore, our assumption on charge correction by referencing to SieO at 532 eV is justied by Fig. 2(b) and
Table 2 as the major contribution of O1s signals are
arising from SieO for all the samples.
3.3. ATR-FTIR
ATR-FTIR analysis was performed to study the
chemical functionality of the changes that occurred to
the LSR after exposure to the water at 100  C for 104
weeks. The spectra for the control and exposed LSR
samples at various depths are shown in Fig. 3. Table 3
summaries the characteristic FTIR bands and assignments for the control LSR [7].
Most of the characteristic FTIR bands for the LSR
decreased sharply in the intensity after exposure. The
only band, which remains intact, is centred at 1081 cm1
and is assigned to n(SieOeSi) vibration (Fig. 3(d), I),
however, the band at 2905 cm1 assigned to symmetric
n(CH) vibrations in the methyl group (CH3) is almost
non-existence (Fig. 3(d), II). On the other hand, there is

476

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480


0,60
0,55

1015

(I)

0,50

794

1081

0,45

Absorbance

0,40
1260

0,35
0,30

865

0,25

1413

701

0,20

(a)

0,15

(b)

0,10

(c)

0,05

(d)

0,00

(e)

2000

1500

1000

Wavenumbers (cm-1)

0,050
0,045

2964

(II)

Absorbance

0,040
0,035
0,030

2905
3421

2516

0,025

(a)
0,020

(b)
0,015

(c)

0,010

(d)
(e)

0,005
4000

3500

3000

2500

Wavenumbers (cm-1)
Fig. 3. ATR-FTIR for the LSR sample control (a) surface, (b) bulk, and exposed to water at 100  C for 104 weeks (c) bulk, (d) surface, and
(e) surface treated silica for regions (I) 2000e650 cm1 and (II) 4000e2000 cm1.

a tendency of growing a broad band in the region of


3000e3500 cm1 possibly due to formation of hydroxyl
groups (OH), which form in the course of degradation
(Fig. 3(d), II). Nevertheless, one cannot exclude the fact
that the broad band could also be due to residual
moisture from aging.
Table 3
Characteristic FTIR band assignments for the control LSR sample
No.

Wavenumbers (cm1)

Assignment

1
2
3
4
5
6
7
8

3500e3200
2962e2960
1440e1410
1270e1255
1100e1000
870e850
840e780
700

OH
CH in CH3
CH
SieCH3
SieOeSi
Si (CH3)3
Si (CH3)2
Si (CH3)3

The main bands presented for the exposed sample


are, primarily, very similar to a silica surface environment. This becomes clearer when one observes that the
FTIR absorbance bands for surface treated silica
(Fig. 3(e)), used as ller, in the rubber industry, are
identical compared to that of 104 weeks exposed LSR
sample at the surface (Fig. 3(d)). Interestingly, the FTIR
spectrum of the cross-section (i.e., bulk) of the exposed
sample (Fig. 3(c)) shows no evidence of degradation,
since it is very similar to spectra of the control sample,
both at the surface (Fig. 3(a)) and the bulk (Fig. 3(b)).
The observation made in XPS regarding formation of
the CeO species at 285.9 eV (Table 2), which is likely to
be as a consequence of degradation process. However,
this species is obscured in ATR-FTIR spectra since
the absorption peak for n(CeO) vibration at around
1100 cm1 is largely overlapped by the broad and
dominated n(SieOeSi) vibration band. As far as the

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480

FTIR spectrum of the surface treated silica is concerned


(Fig. 3(e)), apart from the band at 1081 cm1 due to
n(SieOeSi) vibration, the presence of a small band
centred at around 794 cm1 could be due to an organic
agent (CH), which is utilized in the treatment of the
silica as ller. This band, however, is not obvious at
higher wavenumbers (e.g., around 3000 cm1) probably
due to the broad band originated from hydroxyl groups
(OH) possibly as a result of residual moisture from
aging process or, as mentioned earlier, hydrolysis
product (Scheme 2).
The ATR-FTIR observations are supportive of the
XPS results in demonstrating that signicant changes in
the rubber backbone occur after 104 weeks of exposure.

3.4. Physical properties


3.4.1. TGA
In order to reect the true picture of TGA of the
exposed sample, we tried to use that portion of the
sample for analysis identied by other methods (e.g.,
Optical microscopy, XPS, and ATR-FTIR) as degraded
portion by cutting only the top layer (!100 mm) of
the material. Fig. 4 shows the TGA thermograms
along with DTG (dierential thermogravimetry) for
the exposed LSR rubber sample at prolonged exposure
time. The weight loss (%) is similar for both control and
exposed samples when the temperature is increased from
ambient temperature to 200  C. The real dierences

CH3

CH3

Si

Si

CH3

CH3

Repeating unit in the LSR

CH3

Si
CH3

OH

H2O

Attack on
Si-O-Si

+ HO

Subsitution of
methyl with OH

CH3

OH

Si

Si

CH3

CH3

Chain scission of the LSR backbone

Minor reaction product

CH3

CH3

Si

Si
CH

CH3

CH2
Unreacted vinyl unit in the LSR after crosslinking

Attack on
Vinyl bond

CH3

H2O

CH3

CH3

Si

477

Si
CH

OH

CH3
Minor reaction product

Scheme 2. Proposed hydrothermal degradation reaction mechanisms for the LSR sample.

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480


0.8

100

0.6

80

0.4

60

0.2

40

0.0
800

0.2

(a)

-105.4C(I)

0.0

Heat Flow (W/g)

120

Deriv. Weight ( /C)

Weight ( )

478

(b)

-104.1C(I)

(c)

-0.2
-126.0C(I)

(d)

-0.4

-0.6
0

200

400

600

Temperature (C)
Fig. 4. TGA/DTG for the LSR (dd) control and (- - - -) exposed to
water at 100  C for 104 weeks.

between the two samples become more apparent at


temperatures above 300  C. At 400  C most polymeric
materials start to decompose which is reected by weight
loss (%). For control sample the weight loss (%) is
almost twice as high as the exposed sample. The
dierence could mainly be related to the fact that the
LSR, after long period of exposure to water at 100  C,
has a siliceous-like surface that inhibits breakdown of the
bulk at this temperature. Furthermore, the calculated
residue (%) at the nal temperature (e.g., 1000  C) is
found to be higher for exposed sample compared to the
control one. The probable reason might be that since
the exposed sample is left with higher level of silica
relative to the control sample, which cannot be driven
o from the system at higher temperature (1000  C). In
addition, the dierence between DTG curves, which
represent the rate of weight loss (%) as a function of
temperature, for control and exposed sample (e.g., the
area under the curve reduced in the exposed sample)
can easily be noticed (Fig 4).
In order to assess the extent of degradation into the
bulk, TGA analysis was performed on a cross-section
(bulk) of exposed sample (not shown). As expected,
the results correlate with that of control sample since the
degradation was limited only to 100 mm top layer of the
LSR sample.
3.4.2. DSC
The selection of the degraded portion of exposed
sample for DSC examination was followed as described
for TGA analysis. Fig. 5 illustrates the DSC transition
signals for LSR samples along with the PDMS and
surface treated silica studied in this work. The melting
temperature, Tm, for the control (45.0  C) and
exposed (46.0  C) samples of LSR are not signicantly
dierent. However, the Tm curve of the exposed LSR
becomes somewhat broader than that of the control
sample. In general, an ill-dened feature of Tm is
attributed to the presence low molecular mass com-

-0.8
-165
Exo Up

-115

-65

-15

35

85

Temperature (C)

Fig. 5. DSC transition curves on heating cycle of (a) surface treated


silica, the LSR (b) control, and (c) exposed to water at 100  C for 104
weeks, and (d) PDMS.

pounds, which mostly associated with polymeric materials [9]. Although this value can also depends on the
crystallisation temperature, sample pre-treatment, and
the DCS analysis conditions (e.g., heating rate) [8].
The DSC of pure PDMS sample, on the other hand,
exhibits, two resolved Tm peaks (40.50  C and 47.1  C)
along with one Tc (84.7  C) and a well-dened Tg
(126.0  C).
The Tg of LSR samples compared to PDMS, as
expected, were less well dened and could be identied
at higher temperature for control (105.4  C) and
exposed (104.1  C). Although, the dierence in Tg
for the control and exposed sample is not that
signicant, however, one could assume the increase in
Tg of exposed LSR is a direct consequence of surface
modication (e.g., become more siliceous) that ultimately
leads to less chain mobility of the rubber. The poor
denition of this point relative to that of PDMS is
most likely due to the reduction of chain mobility upon
cross-linking of the rubber [8]. It is also worth
mentioning that Tc was absent in LSR samples, as
expected, since the presence of ller and cross-linking
suppresses crystallisation. In addition, the DSC for silica
ller sample did not show any of these endothermic or
exothermic transition curves (Fig. 5(a)). Our DSC results
agree with a similar observation reported for some
implant silicone breast samples studied by an extensive
DSC investigation [8].
3.4.3. Micro hardness and volume change (%)
Fig. 6 shows the micro-hardness and volume change
(%) as a function of exposure time. The surface
hardness increases linearly as a function of exposure
time. The micro-hardness measurement showed an increase in the hardness of exposed sample due to exposure in comparison to the control sample. The increase
in micro-hardness could possibly be due to degradation

479

Change in physical property

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480


Table 4
TGA data analysis for the LSR samples

18
16
14
12
10
8
6
4
2
0
-2
-4
-6
-8
-10
-12
-14
-16
-18

Micro hardness (IRHD)


Volume change (

15

30

45

60

75

90

105

Exposure time, week(s)


Fig. 6. Change in physical properties of the LSR sample including
micro-hardness (IRHD) and volume change (%) as a function of
exposure time to water at 100  C.

(e.g., the chain scission), which in turn results in surface


erosion and ultimately surface becomes SiO2 enriched.
The increase in the surface hardness was already
reected in the optical micrographs (Fig. 1(b)), and
conrmed by these measurements. In addition, this
result could be rationalised by the chemical changes/
modications in the chemistry of the LSR sample at the
surface level already observed by XPS and FTIR data
for the exposed material. These observations pointed
towards more siliceous (OeSieO) like surface (Figs. 2(c)
and 3) with a sharp increase in the amount of oxygen
(Table 1).
The volume change (%) study, on the other hand,
showed a decrease upon exposure by 17% compared to
the control sample. Previously, it has been reported [10]
that silicone rubber materials are very likely to lose (i.e.,
leach out) some low molecular species once they are
exposed to high temperature (e.g., 170e210  C). In
addition, this report showed [10] that silicone rubber
after thermal degradation processes gave rise to weight
loss via formation of cyclic oligomeric species as a result
of random chain scission along the backbone. This
process could possibly be accelerated under hydrothermal degradation (i.e., in the presence of water) even
at 100  C, where water is an eective leachant for
oligomeric species. This nding could also account for
our observation in the decrease of the volume change
(%) of exposed LSR samples compared to that of
control LSR sample. Cleavage in the main backbone
of the LSR upon hydrothermal exposure results in
relatively shorter rubber chains, which erodes away (i.e.,
transport into the aqueous solution, e.g., water) upon
prolonged exposure. Furthermore, the volume change
(%) decreases linearly with respect to the exposure time
(Fig. 6), and the TGA results provide supporting
evidence in this respect as a lesser reduction in the
weight loss (%) is observed for the exposed LSR sample
relative to control one (Fig. 4 and Table 4).

Sample

Weight loss (%) at various temperature (  C)


25e200

200e400

400e800

Residue

LSR-control
LSR-exposed

0.6
0.4

2.0
1.1

55.9
50.5

41.5
48.0

3.4.4. Reaction mechanisms for hydrothermal


degradation
From a bond dissociation energy point of view, the
SieC bond is more favourable to break (327 kJ mol1)
than SieO bond (452 kJ mol1). Nevertheless, the latter
bonds are more likely to break during degradation;
suggesting that depolymerisation in the case of silicone
rubber is mainly governed by the molecular structure
and kinetic considerations and not by the bond
dissociation energies [12,13]. The plausible degradation
mechanisms of LSR are outlined in Scheme 2. Based on
our nding in both XPS and FTIR and reported studies
in literature [9e12] the hydrothermal degradation could
be rationalised via two dierent modes. The most
plausible is the chain scission via hydrolytic attacks on
SieOeSi bonds upon prolonged exposure. One of the
less likely possibilities is substitution of the methyl side
groups with formation of hydroxyl groups (OH) since
this mode of degradation would be more realistic if the
reaction had taken place in the presence of acid or basic
media [12]. The other minor possibility of degradation
pathway could be the attack on unreacted vinyl groups,
used as cross-link sites, resulting in the formation of
hydroxyl groups (OH). In all cases these reactions would
result in chemical alteration of the LSR backbone.
Thus, the major pathway for the degradation on
the basis of relative bond dissociation energy is not
appropriate for our study, as the homolytic cleavage of
the bonds is not found. In our case, which could be
dened as a plausible heterolytic scission of bonds, the
solvation energy plays more important role in deciding
the major routes for degradation than the bond
dissociation energy.

4. Conclusions
The hydrothermal stability of a cross-linked liquid
silicone rubber has been monitored for a prolonged
period. The conclusions can be summarised as follows:
1. The extent of hydrothermal degradation leading to
hydrolysis and oxidation at prolong time is limited
to only 100 mm of top surface layer. Below this level
the LSR remains intact and resembles the control
(unexposed) sample.
2. The plausible reaction mechanisms involve attack on
SieOeSi bonds, methyl side group in the side chains

480

A. Ghanbari-Siahkali et al. / Polymer Degradation and Stability 90 (2005) 471e480

(eSieCH3) within the repeating units of the LSR as


well as unreacted vinyl side groups (eSieCHaCH2).
These degradation pathways could then lead to
formation of hydroxyl groups as intermediate step
and upon further degradation (i.e., oxidation) into
carbonyl groups. These reactions would ultimately
result in oxidation and breakdown of the LSR (i.e.,
hydrolysis).
3. In terms of physical properties, the exposed LSR
sample exhibits a signicant reduction in volume as
a consequence of leaching the low molecular weight
species (oligomers) into the exposure medium
(water). On the other hand, the micro surface
hardness increases due to the chemical modication
of the surface (i.e., changing to silica rich surfaces)
as well as formation of degradation products (e.g.,
formation of oxygenated species including OH
groups) on the surface (100 mm).

[2]
[3]

[4]

[5]

[6]
[7]

[8]

[9]

Acknowledgement

[10]

The authors wish to acknowledge the technicians


at Research department, Grundfos Management A/S,
Denmark for carrying the necessary tests.

[11]

References
[1] Joo J, Lee CY. High frequency electromagnetic interference
shielding response of mixtures and multilayer lms based on

[12]

[13]

conducting polymers. Journal of Applied physics 2000;88(1):


513e8.
Clarson SJ, Semlyen JA, editors. Siloxane polymers. Englewood
Clis, NJ: PrenticeeHall; 1993.
Kendrick TC, Parbhoo B, White JW. In: Patai S, Rappoport Z,
editors. The chemistry of organic silicon compounds. New York:
John Wiley & Sons; 1989.
Grassie N, Macfarlane IG, Francey KF. The Thermal degradation of polysiloxanes e II poly(methylphenylsiloxane). European
Polymer Journal 1979;15:415e22.
Beamson G, Briggs D. High resolution XPS of organic polymers.
The Scienta ESCA 300 Database. Chichester: John Wiley and
Sons; 1992.
Koerner G, Schulze M, Weis J. Silicones chemistry and
technology. Essen: Vulkan-Verlag; 1991.
Kang DW, Yeo HG, Lee KS. Preparation and characterisation of
liquid silicone rubber nanocomposite containing ultrane magnesium ferrite powder. Journal of Inorganic and Organometallic
Polymers 2004;14(1):73e84.
Birkefeld AB, Eckert H, Peiderer B. A study of aging silicone
breast implants using 29Si, 1H relaxation and DSC measurements.
Biomaterials 2004;25(18):4405e13.
Hemminger WF, Cammenga HK. Methoden der thermischen
analysis. Berlin/FRG: Heidelberg/Spring; 1989.
Kole S, Srivastava SK, Tripathy DK, Bhowmick AK. Accelerated
hydrothermal weathering of silicone rubber, EPDM, and their
blends. Journal of Applied Polymer Science 1994;54(9):1329e37.
Kim SH, Cherney EA, Hackam R, Rutherford KG. Chemicalchanges at the surface of RTV silicone rubber coatings on insulators
during dry-band arching. IEEE Transaction on Dielectric and
Electrical Insulation 1994;1(1):106e23.
Camino G, Lomakin SM, Lazzari M. Polydimethylsiloxane
thermal degradation e part 1. Kinetic aspects. Polymer 2001;42
(6):2395e402.
Korshak VV. The chemical structure and thermal characterisation
of polymers. Jerusalem: Keter Press; 1971.

Você também pode gostar