Page 1 of 35 Journal of Polymer Science Part A: Polymer Chemistry

Chemical degradation of an uncross-linked pure fluororubber in an

alkaline environment
Fo
rP

Journal: J. Polym Sci. Part A: Polym. Chem.

Manuscript ID: JPOL-A-04-0504.R1

ee

Manuscript Type: Original Article

Date Submitted by the

n/a
Author:
rR

Chemical degradation, FKM, Fluororubber, XPS and ATR-FTIR, TGA

Keywords:
and DTG, Gel fraction (%)
ev
iew

ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901

 Journal of Polymer Science Part A: Polymer Chemistry Page 2 of 35

Chemical degradation of an uncross-linked pure fluororubber in an

alkaline environment

SUSANTA MITRAa, AFSHIN GHANBARI-SIAHKALI *a, PETER KINGSHOTTa,

SØREN HVILSTEDb, KRISTOFFER ALMDALa

a
The Danish Polymer Centre, Risø National Laboratory, Building 124, PO Box 49,
Fo

Frederiksborgvej 399, DK-4000 Roskilde, Denmark.

b
The Danish Polymer Centre, Department of Chemical Engineering,
r

Technical University of Denmark, Building 423, DK-2800 Kgs. Lyngby, Denmark.

Pe
er
Re
vie
w

* Corresponding author:

Afshin Ghanbari-Siahkali (E-mail: afshin.ghanbari-siahkali@risoe.dk)

1
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 3 of 35 Journal of Polymer Science Part A: Polymer Chemistry

ABSTRACT

The chemical degradation of an uncrosslinked pure fluoroelastomer, FKM (Viton A), in

an alkaline environment (10% NaOH, 80 0C) is investigated in this study. Scanning

electron microscopy (SEM) images show that on a microscopic level significant

degradation substantially increase the surface roughness after prolonged exposure (e.g., 12

weeks). The molecular mechanisms of the chemical degradation processes at the surface

are evaluated using X-ray photoelectron spectroscopy (XPS) and attenuated total

reflectance-Fourier transform infrared spectroscopy (ATR-FTIR). The results reveal that

Fo

the early degradation proceeds primarily via dehydrofluorination reactions, creating double

bonds in the rubber backbone. This further accelerates the degradation after longer
r

exposure times. Furthermore, the resulting double bonds undergo nucleophilic attack by
Pe

aqueous NaOH solution to form several oxygenated species. All these species ultimately
er

recombine to form cross-links as evidenced by the increase in gel fraction (%) and surface

hardness (Shore A). The pronounced effect of chemical degradation through reduction in
Re

thermal stability of the pure FKM rubber upon exposure is also evident in

thermogravimetric analysis (TGA) and differential thermogravimetry (DTG).

vie
w

Keywords: Chemical degradation; FKM (Viton A); Fluororubber; XPS and ATR-FTIR;

TGA and DTG; Gel fraction (%); Surface hardness (Shore A)

2
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 4 of 35

INTRODUCTION

Fluororubbers (FKM) are one of the very important classes of synthetic rubber and find use

in many crucial applications viz., in aerospace, automotives, seals, gaskets and hoses.

Fluororubbers are well known for their inherent resistance to fuel/oil and heat by virtue of

the strong nature of their structural chemistry (i.e., C – F bond energy, 485 kJ mol-1)1. To
2-4
date studies of FKM degradation have mainly emphasised thermal and irradiative5-7

induced degradation. Nevertheless, there are a number of studies reported that focus on

the chemical degradation of FKM rubbers in chemical environments8-12 .

Fo

However, no systematic studies of the alkaline induced chemical degradation of pure

FKM rubber have been made, probably because the pure rubber never finds use in any
r Pe

practical applications. It is beneficial to investigate the degradation behaviour of the pure

rubber, in the absence of additives and crosslinking agents, to elucidate whether any
er

improvements to the native structure of the compound can be designed. This should

improve resistance under the conditions of actual application. Our previous study on the
Re

chemical degradation of di-amine cross-linked fluoroelastomer, FKM (Viton A) in an

alkaline environment (10% NaOH, 80°C)13 has led us to re-consider several issues viz.,
vie

whether the nature of the cross-link is mainly responsible for the nucleophilic attack, and
w

thus provides the initiation points of chemical degradation, or whether the rubber backbone

is equally vulnerable under the studied alkaline environment. If the cross-link site alone is

responsible, then the type of cross-linking needs to be reviewed. If the rubber backbone is

equally vulnerable, then the basic architecture of the rubber needs to be modified, in order

to improve resistance to the chemical degradation environment studied here.

The aim of the present study is thus to determine the molecular mechanisms of chemical

degradation of a specific type of pure FKM rubber (i.e. vinylidene fluoride-co-hexafluoro

propylene) with fluorine content of around 66% when exposed to an aqueous alkaline

3
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 5 of 35 Journal of Polymer Science Part A: Polymer Chemistry

environment. The very same type of FKM rubber used in our previous study on cross-

linked system is used in this study. This enables us to compare the results with the same

frame of reference.

EXPERIMENTAL

Materials

Pure FKM rubber used in our study was a commercial grade rubber (Viton® A) that

consists of vinylidene fluoride (~70%) and hexafluoro propylene (~30%) co-monomers

Fo

having Mooney viscosity of 60 (ML-1+10 @100°C), and kindly provided by AVK Gummi

A/S, Denmark. The representative chemical structure of the FKM rubber used in this study
r

is shown in Scheme 1. The pure rubber was cut into small pieces of around 2 mm
Pe

thicknesses with an approximate length of 10 mm and a width of 5 mm each. All the

er

samples were sonicated in methanol for 2 minutes in order to clean and remove any

contamination present on the surface. Analytical grade NaOH was dissolved in deionised
Re

water to make 10% solution (weight/volume) with pH 14.5 as an exposure media.

vie

Aging and characterisation methods

The pure FKM samples were exposed to a solution of NaOH (10 %) at 800C for various
w

periods of time from 1 to 12 weeks.

The interfacial chemical degradation was elucidated using X-ray photoelectron

spectroscopy (XPS) analysis on all samples using a Sage 100 (SPECS, Berlin, Germany),

which was operated at a pressure less than 10-7 Torr. Analyses were performed using an

unmonochromated Mg K X-ray source operated at a power of 300 W with a take-off angle

of 90°, resulting in a maximum probe depth of around 10 nm. Exposure time was

minimised (e.g., less than 40 minutes) in order to prevent X-ray induced sample damage.

4
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 6 of 35

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 5 – 10%. High-resolution XPS analysis was performed on the carbon 1s (C 1s)

region, and the spectra were deconvoluted using curve fitting techniques that relied on a

least squares minimization routine in the software. The full width at half maximum

(FWHM) for all peaks was constrained to 2.0 eV. The binding energies were corrected by

referencing to the hydrocarbon component (C-C/C-Hx) at 285.0 eV.

The degradation/modification at a slightly deeper level into the surface than XPS was
Fo

probed (e.g., 1 µm) by attenuated total reflectance Fourier transform infrared (ATR-FTIR)

spectroscopy using a Perkin Elmer Instruments (Spectrum One), and run with 32 scans and
r

a resolution of 4 cm-1.
Pe

Surface topography before and after the exposure was monitored using scanning electron
er

spectroscopy, SEM, (Zeiss DSM-960). Prior to the SEM imaging, the samples were coated

with a thin layer of gold for minimising the charging effect and increasing the image
Re

contrast.

The change in the thermal stability was also monitored by thermogravimetric analysis
vie

(TGA) and differential thermogravimetry (DTG) using a TA Instruments (Q 500). The

samples were subjected to a constant heating rate (20 0C min-1) from the ambient
w

temperature to 800 0C, in a nitrogen atmosphere maintaining a constant flow rate of 20 ml

min-1. The temperatures at 50% weight loss were noted. The residue (%) for the samples

was noted at 700 0C from TGA thermograms. The temperature (Tmax), at the maxima of the

peak of the differential thermogravimetry (DTG) curves was also noted along with the

corresponding maximum rate of decomposition (percent change in weight per min. i.e.,

dw/dt) at the corresponding Tmax. All the analyses were done using the TA instruments

Universal Analysis 2000 software version 3.2B.

5
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 7 of 35 Journal of Polymer Science Part A: Polymer Chemistry

The gel fraction (%) study and surface hardness (Shore A) measurements reveal the

outcome, through chain scission and/or cross-linking of the pure rubber as a consequence

of the chemical degradation upon exposure. The gel fraction (%) for the samples was

obtained according to the following equation.

Gel fraction (%) = (W2/W1) X 100

Where, W1 is the initial weight before extraction and W2 is the weight of the insoluble

portion after extraction and drying. The rubber samples with known weight (W1) were

Soxhlet extracted using methyl ethyl ketone (MEK) as a solvent for 16 hours. The
Fo

remaining insoluble portion after extraction was dried under vacuum at 700C for 24 hours.

The dried materials were Soxhlet extracted again using deionised water for 8 hours in order
r

to remove any free unreacted NaOH present in the material. The remaining insoluble
Pe

portion was finally dried under vacuum at 100 0C for 24 hours and weighed (W2).
er

Surface hardness was measured using a Zwick durometer (Shore A) that complies with

DIN 53505/ASTM D676. Hardness measurements on the surfaces were done as per
Re

ASTM D2240-03.

In the course of all the analyses, a control sample (not exposed) of pure FKM rubber was
vie

also characterised under similar conditions for comparison purposes.

w

RESULTS AND DISCUSSIONS

XPS

As per the representative chemical structure of the FKM rubber investigated in this study

(Scheme 1), one could expect the theoretical C/F ratio to be around 0.63. But, considering

the high vinylidene fluoride (~70%) content of this type of random co-polymer, the C/F

ratio theoretically could be around 0.72. Our experimental C/F (0.72) agrees well with this.

Table 1 shows the atomic ratios (C/F, O/F and O/C) for the pure FKM rubber after different

6
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 8 of 35

exposure times to the alkaline environment that was studied. After one week, the pure

FKM rubber has undergone a dramatic chemical change as indicated by the ten-fold

increase in O/C ratio (0.32) compared to that of the control sample (0.03). The O/C ratio

remains almost the same from 1 week to 5 weeks of exposure. From 7 weeks of exposure

and onwards, O/C continues to increase again as an indication of significant progress in the

degradation on the surface of the pure FKM rubber, and it finally reaches to 0.94 after 12

weeks of exposure. As a consequence, the C/F and O/F ratios increase dramatically with

the exposure time as shown in Table 1. This is probably due to the elimination of hydrogen
Fo

fluoride (HF) from the pure rubber backbone through a dehydrofluorination reaction

induced by the harsh alkaline environment. The elimination of HF leaves a double bond
r

for further nucleophilic attack by the aqueous alkaline medium and this lead to several
Pe

consequent reactions. All these reactions seem to be competing with each other and
er

continue upon prolong exposure.

In order to assess the extent of degradation into the bulk, XPS analysis was also performed
Re

on the interface of the control and 12 weeks exposed samples after sectioning to a depth of

approximately 1mm from the surface. Interestingly, the C/F value for the bulk (0.69)
vie

agrees well with the postulated theoretical stoichiometry (C/F – 0.72) of the FKM rubber

and thus indicates the identical composition of surface and bulk for the control pure FKM
w

rubber. The O/C ratio from the bulk of the 12 weeks exposed sample (0.20) is found to be

4 times higher than that for the bulk of the control sample (0.05). The O/F ratio from the

bulk of the 12 weeks exposed sample (0.98) is significantly higher than that of the bulk of

the control sample (0.04). The C/F ratio of the 12 weeks exposed sample from the bulk

(4.89) is also much higher than the bulk of the control sample (0.69). This demonstrates

that the chemical degradation of the pure FKM rubber is not only confined to the surface

but also reaches to the bulk upon prolonged exposure (e.g. 12 weeks).

7
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 9 of 35 Journal of Polymer Science Part A: Polymer Chemistry

High resolution XPS analysis was performed in order to obtain more detailed information

about the chemical changes occurring to the pure FKM rubber samples. Curve fitting was

used to deconvolute both the control and exposed samples for the C 1s region. Curve

fitting allows assignment of changes to the chemical environment of the carbon atom. Fig.

1(a) and (b) show the C1s spectra of the control and the 12 weeks exposed FKM rubber

samples, respectively. In addition Fig. 1(c) shows the normalised superimposed C1s

spectra for all samples before and after exposure. The results of the curve fitting are

summarised in Table 2. The C 1s spectrum for the control sample in Fig. 1(a) can be fitted
Fo

to 5 components corresponding to the C-C/C-H (285.0 eV), C-CF2 (286.9 eV), CF (289.3

eV), CF2 (291.4 eV) and CF3 (293.9 eV). These assignments are in accordance with those
r

found in the Beamson and Briggs database14. We do not include any secondary shifts in
Pe

the fitting since we are using lower resolution non-monochromatic Mg K X-rays. The C-F
er

environment is found to be quite stronger than C-C/C-H component as expected for the

pure FKM rubber control sample14. After exposure to the NaOH solution, there is a clear
Re

change in the chemical functionality, as shown in Fig. 1(c) and Table 2. The most

significant changes are the gradual loss of the C-F component associated with
vie

dehydrofluorination. Simultaneously the degree of oxygen functionality increases as a

w

function of exposure time, and the spectral deconvolution of the 12 weeks exposed sample

results in species that are almost solely assigned to oxygen functionality. From 7 weeks of

exposure, peaks at 286.9 eV, 288.3 eV and at 289.6 eV can independently be assigned to C-

OH/C-O-C, C=O and O=C-OH species respectively14. This is supported by the changes in

the atomic ratios (Table 1). It is worthy to note that the side chain, CF3 units, within the

backbone remained intact even after 12 weeks of exposure. However, the CF3 signal

present in the 12-weeks exposed sample (Fig. 1(b)) is attenuated compared to the control

sample (Fig. 1(a)). This is plausibly due to the covering of the top layers on surface by the

8
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 10 of 35

species generated by chemical degradation. This observation can be explained by the fact

that no -hydrogen is present in this unit and hence it cannot participate in

dehydrofluorination reactions.

ATR-FTIR

The ATR-FTIR spectra for the control and exposed pure FKM rubber samples at various

exposure times are shown in Fig. 2(a) and (b) for the regions 1800 – 800 cm-1 and 3700 –

2700 cm-1 respectively. The spectra for all the exposed and control samples are presented
Fo

in the Fig. 2(a) and (b). It must be emphasised that the spectra are displayed on an absolute

absorbance scale. In Fig. 2(a), the strongest band at 1195 cm-1 is assigned to (CF2)
r Pe

vibrations, the band at 1397 cm-1 is ascribed to (CF) and the other band at around 884 cm-
1
reveals (CF3), for the control sample. These are the main characteristic bands for pure
er

5,8,10,13,15,16
FKM rubber . The band at 1195 cm-1 decreases gradually with exposure time

and disappears almost completely after 12 weeks of exposure.

Re

In agreement with the XPS results the ATR-FTIR band at 884 cm-1 associated with the
vie

(CF3) also remains unchanged through the exposure. On the other hand, the band centred

at around 1439 cm-1 emerges noticeably and simultaneously as a function of exposure time.
w

This band is initially present as a very weak shoulder in the control sample. The band at

1439 cm-1 can be assigned to scissoring vibration from the methylene groups (-CH2-). This

indicates the increase in dominance of methylene environment over the fluorine

environment in the rubber backbone with increase in exposure time as a manifestation of

the degradation process. This fact is well supported by the appearance of two new bands

2925 cm-1 and 2854 cm-1 (Fig. 2(b)) upon exposure, which can be assigned to assym (CH2)

and sym (CH2) respectively. As a result of the chemical degradation, several new bands

emerge due to the formation of oxygen containing species. These are mainly at 3340 cm-1

9
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 11 of 35 Journal of Polymer Science Part A: Polymer Chemistry

(due to hydroxyl groups) (Fig. 3 (b)), at 1733 cm-1 (due to carbonyl groups, C=O), at 1575

cm-1 (due to carboxylate ions, COO-) and at 1003 cm-1 (due to ester or ether (C-O-C)

groups) (Fig. 2(a)) 5,8,10,13,15,16.

In summary, the ATR-FTIR observations demonstrate the significant changes in the rubber

backbone upon exposure to the alkaline conditions and support our postulations on the

molecular mechanism of degradation as determined by XPS.

SEM
Fo

The SEM image for 12 weeks exposed FKM rubber sample clearly shows severe surface

damage through the significant development of surface roughness and irregularities when
r

compared to the control sample (Fig. 3). The apparently smooth surface of the control
Pe

sample has completely turned to a very rough and irregular surface with the sporadic
er

presence of small holes after 12 weeks of exposure. This indicates the changes in the

surface due to the chemical degradation of the pure FKM rubber as a result of the exposure
Re

media that was studied. The damaged surface confirms the changes in chemical

functionality owing to the chemical degradation as observed in our surface analyses with
vie

XPS and ATR-FTIR.

w

TGA and DTG

Fig. 4 shows the TGA and DTG thermograms for the control pure FKM rubber and 12

weeks exposed samples. The results are summarised in Table 3. Clearly the thermal

stability of the sample weakens upon exposure during this alkaline exposure condition.

TGA shows a weight loss for the 12 weeks exposed sample starting at 50 0C in comparison

to the control sample. The temperature at 50% weight loss for the 12 weeks exposed

sample is found to be 12 0C lower than that of control sample. Whereas, the % residue at

10
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 12 of 35

700 0C for 12 weeks exposed sample is found to be approximately 39% in comparison to

that of approximately 3% for the control sample. As a consequence, in DTG, the peak

maxima temperature (Tmax) is shifted towards the lower end and is approximately 24 0C

lower than that of control sample. On the other hand, the rate of decomposition i.e., dw/dt,

(%weight loss min-1) is decreased by 45% between the control and 12 weeks exposed

samples. These two conflicting findings indicate the presence of two primary processes in

the degradation of pure FKM rubber upon exposure to 10% NaOH solution at 80 0C. The

dehydrofluorination leaves double bonds on the rubber backbone. These undergo

Fo

subsequent nucleophilic attack by the aqueous alkaline medium, resulting in chain scission.

The formation of these relatively lower molecular weight species accounts for the early
r

degradation as manifested by lower temperature at 50% weight loss in TGA, and the
Pe

shifting of Tmax towards the lower end in DTG. On the other hand, the recombination of
er

these degradative species in the form of cross-linking results in significantly lower dw/dt,

in DTG. The probable and plausible reason for very high %residue at 700 0C might be due
Re

to some unreacted salts and/or salt formation via favourable chemical reaction of

carboxylic groups (i.e., generated as a result of the pure FKM degradation) with the NaOH
vie

solution. Such salts cannot be driven off easily from the system even at higher temperature

(700 0C).
w

Gel fraction (%)

The gel fraction (%) for the FKM rubber as a function of exposure time is shown in Fig. 5.

The gel fraction shows a significant increase from the control (0.2%) to the 12 weeks

exposed sample (34.5%). This observation indicates that “cross linking” through

recombination of oxygenated scission species is the dominant feature upon prolonged

exposure. The increase in gel fraction is dramatic even after one week of exposure and

11
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 13 of 35 Journal of Polymer Science Part A: Polymer Chemistry

thereafter it increases gradually with increase in exposure time. The cross-linking on the

surface after 1-week exposure probably results in a protective layer, and thus only small

changes are observed in XPS, ATR-FTIR from 2 to 5 weeks of exposure. However, upon

longer exposure (e.g., 7 weeks) the protective barrier breaks and further degradation

proceeds.

Surface hardness (Shore A)

Although the surface hardness measurements are generally used for the cured rubber
Fo

surface, the same technique is included in this study primarily because the hardness of the

pure FKM rubber is relatively high, and thus a reliable result is attainable. Moreover, the
r

hardness measurement is indication of the change in the material properties particularly in a

Pe

degradative environment17. Fig. 6 shows the increase in hardness with exposure time for
er

pure FKM rubber samples. Interestingly, the increase in surface hardness is found to be

quite dramatic after one week of exposure, and thereafter the increase levels off with an
Re

increase in exposure time. This correlates well with the Gel fraction (%) results. The

increase in hardness at the 12 weeks exposed sample is as high as 12 units compared to the
vie

control sample, which is quite substantial. The very high increases in surface hardness

upon exposure can be attributed to the contribution from several plausible factors. Cross-
w

linking through recombination of degradative species definitely contributes substantially to

the overall increase in the surface hardness. Transformation of a relatively flexible rubber

backbone into more rigid backbone owing to the presence of different oxygen

functionalities and subsequent salt formation on the surface as a result of degradation can

also contribute significantly to the overall increase in the surface hardness.

12
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 14 of 35

REACTION MECHANISMS FOR DEGRADATION PROCESS

We have found that the degradation of pure FKM rubber in an alkaline environment

proceeds through backbone cleavages or chain scission via a process of

dehydrofluorination followed by a nucleophilic attack on the resulting double bonds by the

aqueous alkaline medium.

Later on, the degradative species containing several oxygen functionalities recombine with

each other and this result in cross-linking. These mechanisms are presented in Scheme 1

and Scheme 2, respectively, and are fully supported by XPS and ATR-FTIR results
Fo

showing changes in chemical functionality on the FKM rubber surface.

The high vinylidene fluoride (VDF) content of the FKM rubber (Viton®A) provides a
r

source of -hydrogen. This participates actively in the dehydrofluorination reactions on

Pe

heating in the presence of alkali and proceeds through an autocatalytic reaction11. Once a
er

double bond (C=C) is generated in the backbone due to dehydrofluorination, the

elimination of HF from the adjacent site is facilitated. The double bond formed on the
Re

backbone is susceptible to nucleophilic attack by aqueous NaOH solution at 80 0C. This

leads to the formation of hydroxyl groups, which can undergo further oxidation to form
vie

carbonyl, and later on to carboxylic groups through the conventional reaction mechanism.
w

With exposure, the scission products recombine to cause cross-linking. It is highly

plausible that chain-scission or backbone cleavage compete with the cross-linking in the

degradation process studied here.

The entire plausible mechanism can be depicted as follows:

1) Backbone cleavage due to dehydrofluorination (Scheme 1). The evidence comes

from the depletion of fluorine and inclusion of oxygen as shown in the XPS study

as well as the presence of cleaved species in the form of methylene groups and

13
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 15 of 35 Journal of Polymer Science Part A: Polymer Chemistry

carboxylic derivative as shown in the high resolution XPS and ATR-FTIR studies.

These results are also well supported by TGA and DTG.

2) Recombination of the degradative species containing several oxygen functionalities

(through reactions of carboxyl and hydroxyl and self reactions between hydroxyl

groups) to form cross-links (Scheme 2). The evidence for this comes from the Gel

fraction (%) study, the DTG study, and the increase in surface hardness.

In fact, all these reactions may take place simultaneously, and ultimately lead to the

generation of significant irregularities and roughness on the surface (Fig. 3).

Fo

Interestingly, our previous study on the same FKM rubber cross-linked by DIAK® 4 [4,4'-

methylenebis(cyclohexylamine) carbamate] under the same exposure environment has led

r

to significant crack formation on the surface after 12 weeks of exposure13. Cross-links

Pe

offer resistance to the rubber chains undergoing any change in the conformation due to the
er

change in chemical functionalities on the surface. These surface conformational changes

result from the chemical degradation and thus lead to initiation of cracks. In absence of
Re

any cross-links, the rubber chains are free to undergo changes in the conformation, and thus

generate significant surface irregularities and roughness upon prolong exposure, as we have
vie

observed in this study. However, the chain scission through dehydrofluorination followed

by nucleophilic attack proceeds in the same manner both in the di-amine cross-linked and
w

in the uncross-linked FKM rubber in the same chemical degradation environment.

Furthermore, cross-links undergo hydrolysis in the alkaline environment, and this leads to

de-crosslinking along with chain scission and thus a decrease in cross-link density is

observed as the predominant feature in the di-amine cross-linked sample13. The chemical

degradation of the uncross-linked FKM rubber under alkaline conditions primarily

proceeds via cross-linking through recombination of the scission products. This is

manifested by the increase in gel fraction (%) as observed in this study.

14
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 16 of 35

CONCLUSIONS

The conclusion of the present work can be summarised as the follows:

1) A pure FKM rubber with high VDF content undergoes chemical degradation upon

exposure to 10% NaOH solution at 80 0C to a significant extent. The extent of

degradation is found to be extensive after 12 weeks of exposure. The degradation

proceeds via dehydrofluorination followed by the nucleophilic attack onto the

resulting double bonds. This subsequently leads to the chain scission/backbone

cleavage. Simultaneously, the resulting scission products recombine to form cross-

Fo

links.

2) The severe surface degradation proceeds into the bulk upon prolonged exposure and
r

is found to be strong enough to affect the thermal stability of the material.

Pe

3) The plausible reaction mechanism proposed here indicates that VDF content as well
er

as presence of -hydrogen with respect to >CF- or -CF2- of pure FKM rubbers are

the two main determining factors in this chemical degradation environment. The
Re

extent of chemical degradation is extensive even in this un-stressed or un-strained

condition.
vie
w

15
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 17 of 35 Journal of Polymer Science Part A: Polymer Chemistry

The authors wish to acknowledge financial support from MONEPOL (Danish Centre for

the Study of polymer Degradation and Stability) a Danish Research Agency funded centre

contract (J.nr. 2000-603/4001-51). The authors would also like to acknowledge the co-

operation extended by AVK Gummi A/S, Denmark, and the Research Department,

Grundfos Management A/S, Denmark. The authors would like to thank Dr. Keld West at

the Danish Polymer Centre, Risø National Laboratory, Denmark, for his very generous help

with the SEM analyses. The authors would also like to convey their sincere thanks to Dr.

Charles M. Hansen, FORCE Technology A/S, Denmark, for his generous comments and
Fo

suggestions on this manuscript.

REFERENCES
r

1. Améduri, B.; Boutevin, B.; Kostov, G. Prog.Polym.Sci. 2001, 26,105-187.

Pe

2. Thomas, D. K. J. Appl Polym Sci 1964; 8: 1415-1427.

er

3. Kader, A. M.; Bhowmick, A. K. Polym Degrad Stab 2003, 79, 283-295.

4. Banik, I.; Bhowmick, A. K.; Raghavan, S. V.; Majali, A. B.; Tikku, V. K. Polym
Re

Degrad Stab 1999; 63(3): 413-421.

5. Cuccuru, A.; Sodi, F. Thermochim Acta 1975, 12, 281-286.

vie

6. Boullier, I.; Esnouf, S.; Le Mዊel, A. J. Polym Sci Part B Polym Phys 2003, 41,1509-
w

1517.

7. Ito, M. Radiat Phys Chem 1996, 47, 607-610.

8. Sugama, T. Mater Lett 2001, 50, 66-72.

9. Sugama, T., Sullivan, B. J. Mater Sci. Letters 2001, 20, 1737-1740.

10. Harwood, H. J. J. Test Eval 1983, 11, 289-298.

11. Smith, G. C.; Park, D.; Titchener, K. J.; Davies, R. E.; West, R. H. Appl Surf Sci

1995, 90, 357-371.

12. Wang, S.; Legare, J. M. J. Fluorine Chem 2003, 122,113-119.

16
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 18 of 35

13. Mitra, S.; Ghanbari-Siahkali, A.; Kingshott, P.; Almdal, K.; Rehmeier, H. K.;

Christensen, A. G. Polym Degrad Stab 2004, 83, 195-206.

14. Beamson, G.; Briggs, D. In ‘High Resolution XPS of Organic Polymers. The Scienta

ESCA300 Database.’ John Wiley and Sons, Chichester, 1992.

15. Theodore, A. N.; Carter III, R.O. J. Appl Polym Sci 1993, 49, 1071-1080.

16. Church, J. F.; Evans, D. J. J. Appl Polym Sci 1995, 57, 1585-1594.

17. Dinzburg, B. N., Gillen, K. T. Lubrica Eng 2002, 58(2), 21-25.

Fo
r Pe
er
Re
vie
w

17
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 19 of 35 Journal of Polymer Science Part A: Polymer Chemistry

FIGURE CAPTIONS

Fig. 1. High-resolution XPS Study: C 1s spectra of pure FKM rubber (a) control (b) 12

weeks exposed to NaOH (10 %) at 80 0C, (c) Normalised overlaying of C1s spectra for all

samples before and after exposure.

Fig. 2. (a) ATR-FTIR spectra of pure FKM rubber samples before and after exposure to

NaOH (10 %) at 80 0C at different periods of time for the region: 1800 – 800 cm-1, (i)

Control, (ii) 1 week, (iii) 2 weeks, (iv) 5 weeks, (v) 7 weeks, and (vi) 12 weeks
Fo

Fig. 2. (b) ATR-FTIR spectra of pure FKM rubber samples before and after exposure to
r

NaOH (10 %) at 80 0C for the region: 3700 – 2700 cm-1.

Pe
er

Fig. 3. SEM images of pure FKM rubber samples for control and 12-weeks sample after

exposed to NaOH (10 %) at 80 0C.

Re

Fig. 4. TGA/DTG thermograms under nitrogen atmosphere of pure FKM rubber samples
vie

for both control (----) and 12 weeks ( ) exposed samples after exposure to NaOH (10 %)

at 80 0C.
w

Fig. 5. Gel fraction (%) for pure FKM rubber as a function of exposure time after exposed

to NaOH (10 %) at 80 0C.

Fig. 6. Increase in surface hardness (Shore A) for pure FKM rubber samples as a function

of exposure time after exposed to NaOH (10 %) at 80 0C. [Change in hardness = (surface

hardness of the exposed sample – surface hardness of the control sample)].

18
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 20 of 35

TABLE CAPTIONS

Table 1 Ratios of atomic percentages of C, F and O of the surface and bulk of pure FKM

rubber samples before and after exposure to NaOH (10%) at 80 0C as obtained from XPS.

Table 2 Summary of the high resolution XPS results from the curve fitting of the C1s

spectra for the FKM rubber samples before and after exposure to NaOH (10%) at 80 0C for

different period of time.

Fo

Table 3 Summary of the TGA and DTG results of the pure FKM rubber for the control
r

sample and the sample after 12 weeks of exposure to NaOH (10%) at 80 0C.
Pe
er
Re
vie
w

19
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 21 of 35 Journal of Polymer Science Part A: Polymer Chemistry

SCHEME CAPTION

Scheme 1. The dehydrofluorination reactions of pure FKM rubber in alkaline media

followed by nucleophilic attack lead to chain scission/backbone cleavage.

Scheme 2. Recombination of oxygenated scission species result in cross-linking upon

prolonged exposure.
Fo
r Pe
er
Re
vie
w

20
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 22 of 35

Fo
r Pe

Fig. 1(a)
er
Re
vie
w

1
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 23 of 35 Journal of Polymer Science Part A: Polymer Chemistry

Fo
r Pe
er

Fig. 1(b)
Re
vie
w

2
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 24 of 35

CF2 CF2-CH2 C-C/C-H

Normalised Intensity (Arbitrary Unit)

1,0
Control
1-week
0,8 CF 2-weeks
5-weeks
7-weeks
CF3 12-weeks
0,6

0,4
Fo

0,2
r

0,0
Pe

294 292 290 288 286 284 282

Binding Energy, eV
er

Fig 1(c)
Re
vie
w

3
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 25 of 35 Journal of Polymer Science Part A: Polymer Chemistry

1439
1003
1397
884

1575
1195

(vi)
Fo

1733
Absorbance (arbitrary value)

r Pe
er

(v)
Re

(iv)
vie

(iii)
w

(ii)

(i)

1800 1600 1400 1200 1000 800

Wave numbers (cm-1)
Fig. 2(a)

4
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 26 of 35

3340

2925
Absorbance (arbitrary value)

2854

12-weeks

Control
Fo
r Pe

3600 3400 3200 3000 2800

er

-1
Wave numbers (cm )

Fig. 2(b)
Re
vie
w

5
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 27 of 35 Journal of Polymer Science Part A: Polymer Chemistry

Fo
r Pe
er

Control 12 weeks
Re

Fig. 3.
vie
w

6
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 28 of 35

120 80

100

Derivative Weight (%/min)

60

80
Weight (%)

40
60

20
Fo

40

0
20
r Pe

0 -20
0 200 400 600 800
Temperature (0C)
er
Re

Fig. 4
vie
w

7
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 29 of 35 Journal of Polymer Science Part A: Polymer Chemistry

36

Gel Fraction (%) 32

28
24
20
16
12
8
4
Fo

0
0 2 4 6 8 10 12
Exposure time, week(s)
r Pe

Fig. 5
er
Re
vie
w

8
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 30 of 35

Increase in Hardness (Shore A)

13
12
11
10
9
8
7
6
5
4
3
2
1
Fo

0
0 2 4 6 8 10 12
Exposure time, week(s)
r Pe
er
Re

Fig. 6
vie
w

9
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 31 of 35 Journal of Polymer Science Part A: Polymer Chemistry

Table1

Sample C/F O/F O/C

Control 0.72 (0.002)* 0.02 (0.003) 0.03 (0.005)

1 week 3.77 (0.181) 1.22 (0.011) 0.32 (0.015)

2 weeks 3.46 (0.164) 1.04 (0.016) 0.30 (0.014)

5 weeks 6.37 (0.242) 1.87 (0.025) 0.29 (0.012)

7 weeks 5.29 (0.227) 3.21 (0.043) 0.61 (0.030)

12 weeks 7.53 (0.336) 7.05 (0.055) 0.94 (0.045)

Fo

__________________________________________________________________

Sample from the bulk

r

Control 0.69 (0.001) 0.04 (0.006) 0.05 (0.002)

Pe

12 weeks 4.89 (0.203) 0.98 (0.008) 0.20 (0.007)

__________________________________________________________________
er

*
Figures in the parenthesis represent the respective standard deviation values
Re
vie
w

10
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 32 of 35

Table 2

E
CF
3
A C D B
CH CF CF CF
2 2 2
n m
FKM Rubber (Viton A)
**
C-C /C-H A
Fo C1s (eV)

Pure FKM - Viton A

Observed in Samples
14
285.00 286.77
Possible
rP B
289.82
Possible
C
291.23
Possible
D
291.77
Possible
E
293.86
Possible
*
Week(s) exposed
0 285.0 286.9
Structure
A 289.3 ee Structure
B 291.4
Structure
C
Structure
293.9
Structure
E

1 285.0 286.8 A/C-OH 288.7 rR

B/ C=O/ 290.7 C 291.9 D 293.6 E
HO-C=O
ev
2

5
285.0

285.0
286.4 A/C-OH

286.7 A/ C-OH
288.2

288.7
B/ C=O

C=O/
290.0

290.4
iew
B/C/
HO-C=O

B/C/
291.6

292.1
D

D
293.4

293.9
E

E
/C-O-C HO-C=O HO-C=O

7 285.0 286.9 C-OH 288.3/ C=O/ 291.7 D 293.9 E

/C-O-C 289.6 HO-C=O

12 285.0 286.2 C-OH 287.5/ C=O/ 293.6 E

/C-O-C 289.4 HO-C=O
*: Control sample (i.e.. not exposed)
**: The referenced hydrocarbon component

11
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 33 of 35 Journal of Polymer Science Part A: Polymer Chemistry

Table 3 Analyses of TGA and DTG

Temperature at Peak maximum Rate of Residue at

Sample 50% weight loss temperature decomposition 700°C
(°C) Tmax (°C) dw/dt at Tmax (%)
(from TGA) (from DTG) (% /min) (from TGA)

Fo (from DTG)

Pure FKM rubber - Control

rP
495.0 499.5 61.3 2.5

Pure FKM rubber-12 weeks exposed 483.0

ee 476.0 33.7 38.7

rR
ev
iew

12
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
 Journal of Polymer Science Part A: Polymer Chemistry Page 34 of 35

CF3 CF3

- 2HF
CH2 CF2 CF2 CF CF2 CH2 CF CF2

R
Representative repeating units of FKM rubber (Viton A)

CF3 CF3

CH CF CF2 CF CF2 CH C CF2

- -
OH
OH

CF3 CF3
-HF
CH2 CF CF2 CF CF2 CH2 C CF2
Fo

OH OH

CF3 CF3
r

CH C CF2 CF CF2 CH2 C CF2

Pe

O H OH

CF3 CF3
er

CH2 C CF2 CF CF2 CH2 C CF2

OH
Re

CF3
OH - OH CF3
vie

CH2 C CF2 CF CF2 CH2 C CF2

O OH
-
w

O- Na
+ CF3 CF3

CH3 + C CF2 CF CF2 CH2 C CF2

O OH

- +
O Na CF3 CF3

CH2 CF CF2 CH2 C CF2

C + HCF2

O OH

Scheme 1.

1
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901
Page 35 of 35 Journal of Polymer Science Part A: Polymer Chemistry

O- Na
+ CF3 CF3

C CF2 CF CF2 CH2 C CF2

O OH
+ - +
O Na

CH2 C

O CF3
CF3
+
HCF2 CF CF2 CH2 C CF2

OH
Fo

O CF3 CF3

C CF2 CF CF2 CH2 C CF2

r

CF3 O O
Pe

CH2 C
HCF2 CF CF2 CH2 C CF3
er

CF2 O
Re

CF3 CF3
O- Na
+ CF3 CF3
vie

CF CF2 CH2 C CF2

C CF2 CF CF2 CH2 C CF2 + HCF2

OH OH
O
w

O- Na
+ CF3 CF3

C CF2 CF CF2 CH2 C CF2

O CF3 O

HCF2 CF CF2 CH2 C CF2

CF3

Scheme 2.

2
ScholarOne, 375 Greenbrier Drive, Charlottesville, VA, 22901