Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

Fabrication and Mechanical Properties of Silica/Perfluoropolymer

Nanocomposites by Direct Melt-Compounding
without Surface Modification on Nano-Silica
Mitsuru Tanahashi1, Yusuke Watanabe1, Jeong-Chang Lee2,
Kunihiko Takeda3 and Toshiharu Fujisawa1
1. Department of Materials, Physics and Energy Engineering, Nagoya University, Nagoya, Japan
2. Technical Center, Du Pont-Mitsui Fluorochemicals Co., Ltd., Shizuoka, Japan
3. Institute of Science and Technology Research, Chubu University, Kasugai, Japan
Abstract: The authors research group has previously developed a novel method for the fabrication of
silica/perfluoropolymer nanocomposites, wherein nano-sized silica particles without surface modification
were dispersed uniformly through breakdown of loosely packed agglomerates of silica nanoparticles with
low fracture strength in a polymer melt during direct melt-compounding. The method consists of two
stages: the first stage involves preparation of the loose silica agglomerate, and the second stage involves
melt-compounding of a completely hydrophobic perfluoropolymer, PFA (poly(tetrafluoroethylene-coperfluoropropylvinylether)), with the loose silica agglomerates. By using this simple method without any
lipophilic treatment of silica surfaces, silica nanoparticles with a primary diameter of 190 nm could be
dispersed uniformly into the PFA matrix. The main purpose of the present study is to evaluate the tensile
properties of silica/PFA nanocomposites fabricated by the above method. In order to elucidate the effects
of the size of the dispersed silica in the PFA matrix on the properties of the composites, silica/PFA composite samples exhibiting the dispersion of larger-sized silica particle-clusters were fabricated as negative
controls exhibiting poor dispersion state of the silica additives. The results obtained under the present experimental conditions showed that the size of the dispersed silica in the PFA matrix exerts a strong influence on the ultimate properties, such as tensile strength and elongation at break, of the composite materials. Furthermore, uniform dispersion of isolated silica nanoparticles was found to improve not only the
Youngs modulus but also the ultimate tensile properties of the composite.
Keywords: Nanocomposite, Perfluoropolymer, Silica Nanoparticle, Colloidal Stability, Tensile Property
1. INTRODUCTION
Fluorocarbon polymers are used as tubing, gaskets and
valves that are extensively used in various advanced engineering applications, such as chemical plants, medical
and semiconductor manufacturing equipments, because
of their smoother surfaces, self-lubrication properties and
excellent resistance to extremes of temperature and
chemical reagents [1-3]. However, since fluorocarbon
polymers, in particular, perfluorocopolymers, have much
weaker intermolecular interactions than ordinary hydrocarbon polymers [1,2,4,5], their mechanical properties
and dimensional stability are becoming critical issues for
applications of these polymers. There have been some
attempts to improve these characteristics of fluorocarbon
polymers by blending the polymers with inorganic fillers
[6,7]. Recent developments in medical and advanced engineering fields require the downsizing of fluorocarbon
products with corresponding high-reliability. Therefore, a
technique that enables uniform dispersion of nanoscale
fillers in fluorocarbon polymers must be developed.
However, it has generally been considered to be difficult
to fabricate fluorocarbon polymer nanocomposites containing inorganic nanofillers for two reasons. The first is
the extremely high cohesive force acting between nanosized materials, and the second is the extremely low
chemical affinity of the completely hydrophobic fluorocarbon polymer with inorganic fillers that have hy-

droxyl-rich hydrophilic surfaces. Thus, fluorocarbon

polymer composites containing inorganic fillers have
been conventionally fabricated only using melt-compounding methods involving additional and complicated
chemical treatments such as intercalation of organic
swelling agents into clay layers [6] and modification of
the silica surface with silane coupling agents [7]. However, thermal decomposition of these chemical agents
contained in the composites likely occurs during the
melt-compounding and molding processes, since these
processes for composites of fluorocarbon polymers, such
as PFA (poly(tetrafluoroethylene-co-perfluoropropylvinylether)), with high melting points must be operated at
elevated temperature. Moreover, these agents in the
composites, in the form of impurities, exert a harmful influence on semiconductor manufacturing clean processes.
Therefore, in order to ensure the purity of fluorocarbon
polymer composites to be as high as possible, surface
modifiers to achieve good dispersion of the nanofillers
should not be used for fabricating the composites.
The authors research group has proposed a simple
method for dispersing silica nanoparticles into a perfluoropolymer by direct melt-compounding without requiring any surface modification of the silica [8]. For the
direct melt-compounding process, in addition to the shear
stress induced in the melt-compounded polymer, the
critical threshold value necessary for breaking down the

Corresponding author: M. Tanahashi, mtana@numse.nagoya-u.ac.jp

1238

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

agglomerates of nanofillers are the main factors that control the dispersion state of the fillers. Therefore, in the
authors method, loosely packed agglomerates of silica
nanoparticles with low fracture strength (loose silica agglomerates) were prepared intentionally by destabilizing
an aqueous solution of nano-sized colloidal silica. By
melt-compounding PFA with the prepared loose silica
agglomerate, silica/PFA nanocomposites exhibiting uniform dispersion of nano-sized silica particles without
surface modification could be fabricated through the
breakdown of the agglomerates in the PFA polymer melt.
However, the characteristics of the fabricated silica/PFA
nanocomposites were not discussed in the previous study
[8].
The present study was aimed at elucidating the mechanical properties of the silica/PFA composites filled
with unmodified silica nanoparticles fabricated by the
proposed method. First, using aqueous solutions of
nano-sized colloidal silica as starting materials, the silica
agglomerates with different strength were prepared from
the destabilized colloidal silica dispersion systems via pH
control and salt addition. And then, by melt-compounding
the PFA with these silica agglomerates, silica/PFA composite samples exhibiting different silica dispersion states
were fabricated. The composite samples obtained were
subjected to tensile testing. On the basis of the results
obtained, the effects of the size and content of the dispersed silica additives in the PFA matrix on the Youngs
modulus, tensile strength and elongation at break of the
composites are discussed.

uniformly, fabricated with varying silica volume fractions,

Vf,Silica, from 2.8 to 14.4 % in the authors previous study
[8] were subjected to tensile testing. As negative controls
of the silica dispersion state, composites B and C exhibiting poor dispersion state of silica (see Figs. 3 (b) and
(c)), were fabricated, respectively, using sintered agglomerates of silica particles with dp,Silica = 12 nm with a
high fracture strength and commercially available powders of single-nm-sized fused silica. The procedure for
the fabrication of composite A, including the preparation
of loose silica agglomerate with dp,Silica = 190 nm, was
described in the previous paper [8]. In this paper, the
procedure for the fabrication of composites B and C will
be described in the following sections 2.2 and 2.3.
2.2. Materials for fabrication of silica/PFA composites
B and C
The same PFA powders (Teflon PFA 350J, Du
Pont-Mitsui Fluorochemicals Co., Ltd.) used as the matrix for composite A were used as the matrix polymer for
composites B and C as well. Commercially available
powder of fused silica with dp,Silica of around 7 nm
(AEROSIL 300, Nippon Aerosil Co., Ltd.) was used as
the additive for the fabrication of composite B. For the
fabrication of composite C, a commercially available
aqueous colloidal solution of spherical SiO2 with dp,Silica =
12 nm (SNOWTEX-30, Nissan Chemical Industries,
Ltd.) was used as the starting material for silica agglomerate preparation.
2.3. Fabrication and characterization of silica/PFA
composites B and C
The silica agglomerates used as additives for the fabrication of composite C were prepared from the destabilized colloidal silica dispersion system with dp,Silica = 12
nm under the same conditions of pH, with a value of 4,
and KBr concentration (volume ratio of KBr to SiO2
(KBr/SiO2) = 70/30) as that used for the preparation of
loose silica agglomerates with dp,Silica = 190 nm [8]. In
order to ensure an increase in the agglomerate strength,
the silica agglomerates obtained with dp,Silica = 12 nm
were sintered at 873 K for 2 h in air. The packing arrangement of the primary silica particles and pore size
distribution in the sintered agglomerates were characterized, respectively, by a scanning electron microscope operated at an accelerating voltage of 10-20 kV (SEM; Hitachi S-800, Hitachi, Ltd.) and the volumetric method
using nitrogen adsorption (TriStar3000, Micromeritics
Instrument Corp.). In addition, a microcompression test
(MCT-W500, Shimadzu Co.) was used to evaluate the
fracture strength of the agglomerates.

2. EXPERIMENTAL DETAILS
2.1. Conditions for control of the silica dispersion state
in the PFA matrix
Since the main purpose of this study was to elucidate
the relationship between the silica dispersion state in the
PFA composite and the mechanical properties of the
composite, the authors made attempts to control the silica
dispersion state in PFA matrix from a good state (i.e.
homogeneous distribution of isolated nano-sized silica
particles) to a poor one (i.e. uneven distribution of unbroken agglomerates as micron-sized silica particle-clusters). The experimental conditions used for the
fabrication of the silica/PFA composites exhibiting different dispersion states of silica are summarized in Table
1. In the present study, as shown in this table, three types
of nano-silica were used as additives in the PFA-matrix
composite. As examples of silica/PFA nanocomposites
exhibiting good dispersion state of nano-silica (see Fig.
3 (a)), composites A, A-2 and A-3, wherein silica particles with a diameter, dp,Silica, of 190 nm were dispersed

Table 1. Conditions for the fabrication of silica/PFA composites by the direct melt-compounding method.
Composite
sample No. Volume fraction, Form in mixing stage with PFA
V f,Silica / %
A

2.8

A-2

7.7

A-3

14.4

Silica additive
Diameter of primary
particle, d p,Silica / nm

Starting material

Preparation conditions

pH 4 & (KBr/SiO2) = 70/30

Non-sintered loose agglomerate

190

Colloidal silica aqueous solution (MP-2040)

Fused silica (AEROSIL 300)

2.8

Powder

~7

2.8

873 K-sintered agglomerate

12

Colloidal silica aqueous solution (SNOWTEX-30) pH 4 & (KBr/SiO2) = 70/30

1239

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

2.4. Tensile testing of fabricated silica/PFA composites

A-C
The tensile properties of the composite samples, such
as Youngs modulus, tensile strength and elongation at
break, were measured at a controlled relative humidity of
around 50 % at room temperature, according to JIS K
7127 using the TENSILON (UTM-1T or SS-207-EP,
Toyo Baldwin) instrument with a cross-head speed of 50
mmmin1.
3. RESULTS AND DISCUSSION
3.1. Pore structure and fracture strength of prepared
silica agglomerates
Figure 1 shows SEM micrographs of a selected area on
the surface of (a, b) non-sintered and (c) 873 K-sintered
agglomerates of silica particles with dp,silica = 12 nm prepared under various conditions. For the non-sintered agglomerates, the structure of the sample prepared from the
destabilized colloidal silica dispersion system via pH
control and KBr addition (Fig. 1 (a)) was more porous
than that of a blank sample of the silica agglomerate prepared from the stable colloidal solution without pH control or KBr addition (Fig. 1 (b)). A similar relationship
between the pore structure of the prepared agglomerate
and stability of the colloidal silica solution (the starting
material) has also been obtained in the case of the
non-sintered silica agglomerates with dp,silica = 190 nm [8].
By controlling the stability of the dispersion system in a
colloidal silica aqueous solution via pH control and KBr
addition, the packing arrangement of the primary silica
particles in the agglomerates prepared could be varied
from a dense structure to a porous structure. A compari-

1240

Fracture Strength,

f,Silica / MPa

The fabrication of both

(a)
(b)
(c)
composites B and C follows
the procedures described in
the previous paper [8]. Composite C was fabricated by
melt-compounding the PFA
with the 873 K-sintered silica
agglomerates in an intensive
batch mixer (Labo Plastomill,
50 nm
50 nm
50 nm
KF70 model, Toyo Seiki Seisaku-sho, Ltd.). On the other
Fig. 1. SEM micrographs of selected areas of the surfaces of prepared agglomerates
hand, fused nano-silica in
of
silica particles with dp,Silica = 12 nm: (a) Non-sintered dense agglomerate prepowder form was added into
pared
from colloidal silica solution without pH control or KBr addition (pH 10 and
the same mixer and mixed
(KBr/SiO
2) = 0/100), (b) non-sintered and (c) 873- K sintered porous agglomerates
with the melt-compounded
prepared
from
destabilized colloidal silica solution with pH control and KBr addiPFA phase to obtain composite
tion
(pH
4
and
(KBr/SiO
2) = 70/30).
B. In the melt-compounding
processes,
appropriate
amounts of the sintered silica agglomerates or the fused
2
10
silica powders were mixed with PFA to control Vf,silica in
each composite to be 2.8 %. Fracture surfaces were prepared by immersing the fabricated composite samples in
1
10
liquid nitrogen and causing them to fracture; the surfaces
were examined using SEM to characterize the dispersion
state of the silica additives in the PFA matrix. The results
0
dp,Silica
10
dp
obtained from the SEM observations (SEM micrographs)
,Silica = 12
nm
= 7 nm
were analyzed using image-processing software (Cos-1
mos32, Library, Inc.) to calculate the equivalent circular
10
diameter (ESD) of the silica phase identified on the fracdp
,Silica = 19
0 nm
ture surfaces of each composite.
-2

10

Agglomerate sample

d p,Silica

Non-sintered
Non-sintered

~ 7 nm

873 K-sintered
Non-sintered

12 nm
190 nm

-3

10

0.2

0.3

0.4

Experimental results Rumpf's eq. (Eq. (2))

0.5

0.6

0.7

Porosity, Silica
Fig. 2. Relationship between the fracture strength of prepared silica agglomerates and their porosity. The experimental results represented by the closed squares with the
superscript asterisk (*) were obtained from the authors
previous study [8].
son of Fig. 1 (b) and 1 (c) show that by sintering at 873 K,
the primary silica particles forming the agglomerate coalesce and the center-to-center distance of the particles
shrinks, resulting in an increase in the interparticle contact area, which is frequently referred to as sinter neck
formation and growth.
The effects of the porosity of the silica agglomerate
with dp,Silica = 12 nm on the fracture strength are shown in
Fig. 2 together with the results for the agglomerates with
dp,Silica = 190 nm obtained in the previous study [8]. The
value of f,Silica was determined as the agglomerate tensile
strength from the following equation [9]:

f,Silica =

2.8 F0,Silica

d a,Silica

(1)

Here, F0,Silica is the load required for the breakdown of the

silica agglomerate as measured by the microcompression
test, and da,Silica is the mean diameter of the test pieces of

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

silica agglomerates, which corresponds to the distance

between the points of loading. The porosity of the agglomerates, Silica, was calculated using the results of the
pore volume of the silica agglomerates measured by the
volumetric method using nitrogen adsorption. In this figure, the theoretical tensile strength of a packed agglomerate of spherical particles derived on the basis of the binding force acting between two particles forming the agglomerate (Rumpfs equation) [10]:

f,Silica =

AH,Silica

2
8 Silica 24 a d p, Silica
9 1 Silica

(2)

resulting in the formation of three-dimensional networks

of silica particles. Deviations of the experimental results
from the theoretical strength in Fig. 2 might be caused by
the formation of chemical bonds, such as the
above-mentioned bridging via K+, partly between the
primary silica particles in the prepared silica agglomerate.
Moreover, fracture of the silica agglomerate with
non-uniform pore structure (i.e. local existence of some
dense packing regions with relatively higher porosity)
during the microcompression test is a typical local phenomenon that obeys the weakest link theory, although the
porosity of the prepared silica was determined as average
data to determine the pore distribution. The difference
between the control modes of f,Silica and Silica may also
cause a deviation from the theoretical strength of agglomerate derived on the assumption of porous structure
with mono-pore size.
In the present study, since fused silica used for the fabrication of composite B is in a powder form, the values of
f,Silica and Silica for this sample were not evaluated by
the microcompression test and the volumetric method
using nitrogen adsorption. It is expected from Fig. 2 that
the stress required for the dispersion of the cluster of
fused silica powders (AEROSIL 300) to their primary
particles with dp,Silica 7 nm is a little less than twice as
high as the strength of the prepared silica agglomerate
with dp,Silica = 12 nm.

was also plotted as a function of Silica and dp,Silica. In Eq.

(2), AH,Silica is the Hamaker constant of the silica particles
and a is the distance between the surfaces of the two particles in the agglomerate. The value of AH for fused silica
in vacuum or air has been measured and calculated by
several research groups [11-13]. There is a slight scatter,
but most of the reported values are on the order of the
magnitude of 1020 Joules. Thus, their average, 5.5
1020 J, was used as an approximate value of AH,Silica for
silica nanoparticles forming the agglomerate in this paper.
For the value of a, the adhesion separation distance that is
often taken to be around 0.4 nm (single-ngstrm-odered
constant) for particles in intimate but chemically unbonded contact with a surface [13-17] was adopted.
The value of f,Silica for the sintered silica agglomerate
with dp,Silica = 12 nm was higher than that for the
3.2. Dispersion states of silica in fabricated silica/PFA
composites
non-sintered one in the case of nearly the same porosity.
This is caused by the difference in the agglomerate strucFigure 3 (a) shows a typical example of SEM microture shown in Figs. 1 (b) and (c). The sinter neck formed
graphs of the fracture surfaces of the silica/PFA composite A fabricated by melt-compounding PFA with 2.8
by strong siloxane (SiOSi) bonds during the sintering
of the agglomerate leads to an increase in the agglomerate
vol.% non-sintered loose silica agglomerates with dp,Silica
strength. Figure 2 also shows that the agglomerate
= 190 nm described in the previous paper [8]. The silica
agglomerate used for the fabrication of composite A has a
strength decreases with decreasing diameter of the primary silica particles and increasing porosity which supporosity, Silica, of around 0.45 and the lowest strength
ports the theoretical agglomerate strength as expressed by
among the other silica agglomerates prepared in the preEq. (2). However, the experimental results for dp,Silica = 12
sent study, as shown in Fig. 2. The micrograph in Fig. 3
and 190 nm are not in
agreement with the theo(a)
(b)
(c)
retical agglomerate strength.
All the experimental values
of the strength determined
from the microcompression
test are higher than the corresponding theoretical values. The causes of these deviations of the experimental
200 nm
5 m
10 m
results from the theoretical
strengths were considered,
Composite A
Composite B
Composite C
as below. In the previous
Ref. 8
[Present work]
[Present work]
study [8], the authors concluded that the destabilizing
Fig. 3. SEM micrographs of selected areas of fracture surfaces of the silica/PFA comprocess of the dispersion
posites (a) A [8], (b) B, and (c) C fabricated by melt-compounding PFA with 2.8 vol.%
system in the colloidal silica
of various types of silica additives. Loose silica agglomerates with dp,Silica = 190 nm
solution via pH control and
and 873 K-sintered agglomerates with dp,Silica = 12 nm were used for the fabrication of
KBr addition leads to bridg+
composites
A and C, respectively. Composite B was fabricated by direct
ing via potassium ions (K )
melt-compounding
of PFA with silica powders with dp,Silica 7 nm. The bright spots
at a limited number of surenclosed by the circles in (a) are examples of isolated primary silica particles on the
face sites on the silica
fracture surface of composite A. The bright regions enclosed by squares in (b) and (c)
nanoparticles in the destabilized colloidal silica solution, are examples of agglomerates that were not broken down and remained as silica particle clusters in composites B and C.

1241

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

Stress, c / MPa

40

Silica/PFA composite A (
pure PFA (
)
Silica/PFA composite B (
)

30

Silica/PFA composite C
20 (
)

10

at Room Temp.
-1
(Cross-Head Speed: 50 mm.min )
0

100

200

Strain,

300

400

500

c / %

Fig. 4. Typical stress-strain curves of pure PFA and silica/PFA composites A, B and C with Vf,Silica = 2.8 % fabricated under various conditions. The plus (+) symbol
shown at the end of each curve represents the breaking
point of the test piece.
Diameter of Primary Silica Particles, dp,Silica = 190 nm

600 (a)
550
500
450
400
(b)

450 (c)
400
350
300
50
0
0.1

Data for pure PFA

Data for pure PFA

50
40
30
20
10

Data for pure PFA

10

Ultimate Tensile Strength,

u,c / MPa

3.3. Tensile properties of silica/PFA composites

Typical stress (c)-strain (c) curves of the fabricated
silica/PFA composites A, B, and C containing the same
amounts of silica (Vf,Silica = 2.8 %) are shown in Fig. 4.
The values of (a) the Youngs modulus, Ec, (b) the ultimate tensile strength, u,c, and (c) the elongation at break,
b,c, for these three composites exhibiting different dispersion states of silica additives are shown in Fig. 5 as a
function of the mean size of the dispersed silica additives
in each composite, dSilica in PFA. In the present study, the
value of dSilica in PFA in the case of composite A was regarded as dp,Silica (= 190 nm), since this composite clearly
exhibited homogeneous distribution of isolated primary
silica particles, as shown in Fig. 3 (a). In the case of

composites B and C, dSilica in PFA was evaluated using the

average ESD of the silica additive in each composite. In
both Figs. 4 and 5, the results for pure PFA prepared by
melt-compounding the virgin PFA powders without silica
addition were also plotted to compare with the results of
the silica/PFA composites.
The stress-strain curves in Fig. 4 display yielding behavior at c in the vicinity of 10 % and plastic deformation within the range of c from each yielding point to its
corresponding breaking point. From Figs. 4 and 5, the
Youngs modulus of all composites A, B and C consis-

Elongation at Break, Young's Modulus,

Ec / MPa
b,c / %

(a) clearly shows that the isolated primary silica particles

with dp,Silica = 190 nm are dispersed uniformly in the PFA
matrix. This implies that the prepared loose silica agglomerates was sufficiently weak for them to be broken
down to the primary nanoparticles in the PFA melt by the
shear stress induced under the melt-compounding conditions of the previous study [8]. The dispersion states of
silica particles in composites A-2 and A-3 have already
been described in the previous paper [8] for varying volume fraction of the non-sintered loose silica agglomerates
with dp,Silica = 190 nm mixed with PFA. It was found that
both composites A-2 with Vf,Silica = 7.7 % and A-3 with
Vf,Silica = 14.4 %, as well as composite A with Vf,Silica =
2.8 %, exhibit a homogeneous distribution of isolated
primary silica particles in the PFA matrix.
Figures 3 (b) and (c), respectively, show the dispersion
states of silica in the negative control samples of composites B and C fabricated by melt-compounding under the
same conditions as used for the fabrication of composite
A. These types of silica additives, such as sintered agglomerates and single-nm-sized powders, used for the
fabrication of composites B and C are so dense that the
additives could not be broken down and they remained as
micron-sized
large
particle-clusters
in
the
melt-compounded PFA. For composites B and C exhibiting poor dispersion of silica, the sizes of the silica particle-clusters remaining in the PFA matrix were examined
as follows. The values of the average and the standard
deviation of the ESD of the silica particle-clusters in
composite B were, respectively, calculated to be 2.0 m
and 0.9 m by analyzing the silica phase appearing as
bright regions in the SEM micrographs of the composite
surfaces (ex. Fig. 3 (b)). The same analysis of the SEM
micrographs of the surfaces of composite C (ex. Fig. 3
(c)) was used to determine the average and the standard
deviation to be 41.1 m and 18.8 m, respectively. The
results obtained from the above analysis showed that the
dispersion state of silica additives in composite C became
poorer than that in composite B. This implies that the
fracture strength of the sintered agglomerate of silica particles with dp,Silica = 12 nm might be higher than that of
the cluster of fused silica powders with dp,Silica 7 nm.
On the basis of the results obtained in the authors previous [8] and current studies, it is clarified that the dispersion state of silica nanoparticles in the PFA matrix by
the direct melt-compounding method is significantly affected by the fracture strength of the particle agglomerates added into the mixer.

100

Mean Size of Dispersed Silica Additives,

dSilica in PFA / m

Fig. 5. Effects of the size of the dispersed silica additives

in the silica/PFA composite with Vf,Silica = 2.8 % on its (a)
Youngs modulus, (b) ultimate tensile strength and (c)
elongation at break at room temperature. The tensile
properties of composites A, B and C are represented by
closed, half-closed and open circles, respectively. The
properties of pure PFA are represented by the hatched
regions.

1242

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

900
800
700
600
500
400
300
100
0

(a)

Data for pure PFA

500

(b)

Data for pure PFA

400
300
100
0

10

12

14

0
16

Elongation at Break,
b,c / %

Young's Modulus,
Ec / MPa

tently show a value higher than that of pure PFA, independent of dSilica in PFA. These improvements in the
modulus are caused by the inclusion of hard silica. On the
other hand, increases in u,c and b,c of the silica/PFA
composites fabricated in the authors previous study [8]
with decreasing size of dispersed silica additives in the
PFA matrix are observed, in contrast to that observed for
Ec. In particular, composite A exhibiting a homogeneous
distribution of isolated primary silica particles with dp,Silica
= 190 nm in the PFA matrix yielded a larger elongation at
break than in the case of pure PFA.
The ultimate tensile properties, such as elongation at
break, of polymer materials have a tendency to decrease,
by the inclusion of micron-sized metal oxide additives
although the elastic modulus increases [18-20]. However,
composite A fabricated by the authors research group [8],
wherein isolated silica nanoparticles are dispersed uniformly, was found to possess both a larger Youngs
modulus as well as a larger elongation at break than pure
PFA.
The values of Ec and b,c for the silica/PFA nanocomposite A are plotted as a function of the silica volume
fraction in Fig. 6 (a) and (b), respectively. This figure
shows the results obtained from the stress-strain curves of
composites A, A-2 and A-3 exhibiting similar good
nano-dispersion of silica. The Youngs modulus increased
with increasing Vf,Silica. The value of b,c was found to initially increase with increasing Vf,Silica and showed a value
of around 400 % at Vf,Silica = 2.8 %. Although further addition of silica decreased the value of b,c, this value could
be kept at a high level close to the elongation at break of
pure PFA even in the case of 14.4 vol.%. A similar tendency of b,c as a function of the silica content has been
observed in polyvinyl chloride nanocomposites containing surface-modified silica nanofillers [21]. Nitta and
coworkers [22,23] have reported that one of main factors
controlling the large deformation characteristics of matrix

Volume Fraction of Silica,

Vf,Silica / %
Fig. 6. Effect of volume fraction of silica in silica/PFA
nanocomposites on the (a) Youngs modulus and (b)
elongation at break at room temperature. The tensile
properties of composites A, A-2 and A-3 exhibiting homogeneous distribution of isolated primary silica particles with dp,Silica = 190 nm in the PFA matrix are represented by closed circles. The properties of pure PFA are
represented by the hatched regions.

polymers, such as fracture behavior, might be interactions

between filler and matrix phases. However, it was unexpectedly found from Figs. 7 (c) and 8 (b) that the elongation at break of the silica/PFA nanocomposites fabricated
by the proposed method without any surface modification
of the nano-silica could be prevented from deteriorating,
even in the case of the extremely low affinity of silica
with the PFA matrix.
Although, the exact reason for this unexpected fracture
behavior cannot be discussed only on the basis of the
present results, it is possible that the size of the dispersed
silica in the polymer matrix might be too small to behave
as a point of stress concentration. According to fracture
mechanics of materials, the main factor that controls the
mechanical properties, such as elongation at break and
tensile strength, of materials is generally regarded to be
the existence of stress concentration points, which lead to
crack initiation. If dissimilar hard materials with micron-ordered or larger sizes are mixed into a polymer in
the glassy-state, the sites of the hard materials in the
polymer subjected to various external loadings act as
stress concentration points, resulting in the rupture of the
polymer. Therefore, it has been reported that the elongation at break for various polymer-based composites containing micron-sized silica fillers is lower than that for the
corresponding virgin polymer without silica addition
[18-20]. On the other hand, different behavior is observed
in the case when polymer-based nanocomposites containing nano-sized silica fillers, such as silica/PFA nanocomposites A, A-2 and A-3 fabricated in the authors
study, are subjected to various external loadings. Keddie
et al [24]. and Forrest et al [25]. have investigated the relationship between the thickness of the polystyrene (PS)
film and its glass-transition temperature, Tg,PS. In their
previous studies, it was found that Tg,PS decreased with
decreasing PS film thickness when the thickness was below 100 nm. In particular, a glass transition of the PS film
was observed at a temperature just slightly above room
temperature in the study of Forrest et al [25]. The reduction in Tg,PS values was suggested to be a result of the existence of a liquid-like layer, wherein the relaxation of
constraints to molecular motion of PS polymer chains is
occurs at the surface of the glassy film. From the above
findings and suggestion obtained in the studies by Keddie
et al [24]. and Forrest et al [25]., it is possible that polymeric materials with nano-ordered dimensions can be regarded to behave as a viscous liquid even at room temperature, far below the glass-transition temperature of the
corresponding polymer bulk with micron-ordered or larger size.
Thus, the size of the primary silica particles dispersed
in the silica/PFA nanocomposites A, A-2 and A-3 (190
nm) might be close to a dimension level that makes it
possible to enhance the mobility of the PFA polymer
chains. It is expected that since the isolated silica
nanoparticles are surrounded by just a liquid-like PFA
layer in the silica/PFA nanocomposites A, A-2 and A-3
fabricated in the authors previous study [8], a stress
concentration did not occur at the sites of these nanoparticles in the PFA matrix under tensile loading.
4. CONCLUSIONS
The mechanical properties of silica/PFA nanocompo-

1243

Proceedings of International Symposium on EcoTopia Science 2007, ISETS07 (2007)

sites were investigated by a simple melt-compounding

method without any complicated lipophilic treatments of
the silica surfaces. The following results were obtained
under the present experimental conditions:
(1) With regard to the fabrication method for the silica/PFA nanocomposites, the fracture strength of an
agglomerate of silica nanoparticles was found to determine the dispersion state of silica in the PFA matrix. This indicated that a reduction in the agglomerate strength of silica before the melt-compounding
stage is one of the key strategies for obtaining uniform dispersion of the silica nanoparticles in this
method.
(2) The size of the dispersed silica in the PFA matrix exerts a strong influence on the ultimate tensile properties, such as tensile strength and elongation at break,
of the composite materials.
(3) In the case of dispersion of micron-sized silica in the
PFA matrix, the ultimate tensile properties of the
composites were lower than those of the pure PFA
although the Youngs modulus became higher than
that of the pure PFA.
(4) In the case of uniform dispersion of nano-sized silica
in the PFA matrix, in addition to the modulus, the ultimate properties were successfully improved compared to the pure PFA. This implies that the sites of
the isolated silica nanoparticles distributed homogeneously in the PFA matrix did not act as points of
stress concentration during various external loadings.
(5) The method proposed by the authors research group
was found to be a simple fabrication process for high
performance silica/PFA nanocomposites exhibiting
excellent mechanical properties.
REFERENCES
1. T. Satokawa, Y. Kometani, A. Yamada and S. Koizumi, Fussojushi, Nikkan Kogyo Shimbun, Ltd., Tokyo (1969), (in
Japanese).
2. T. Satokawa, Kinousei Ganfusso Koubunshi, Nikkan Kogyo
Shimbun, Ltd., Tokyo (1982), (in Japanese).
3. Modern Fluoropolymers: High Performance Polymers for
Diverse Applications, J. Scheirs (ed.), Wiley, Chichester,
(1997).
4. M. P. Krafft and J. G. Riess, Biochimie, 80 (1998), pp.
489-514.
5. G. P. Koo, Fluoropolymers, L. A. Wall (ed.), Wiley-Interscience, New York, (1972), pp. 507-543.

6. M. W. Ellsworth, JP Patent 2001-523278, (2001) (in Japanese).

7. Y.-C. Chen, H.-C. Lin and Y.-D. Lee, J. Polym. Res., 10
(2003), pp. 247-258.
8. M. Tanahashi, M. Hirose, Y. Watanabe, J.-C. Lee and K.
Takeda, J. Nanosci. Nanotechnol. 7 (2007), pp. 2433-2442.
9. Y. Hiramatsu, Y. Oka, and H. Kiyama, J. Min. Metall. Inst.
Jpn., 81 (1965), pp. 1024-1030.
10. H. Rumpf, Agglomeration, W. A. Knepper (ed.), Interscience,
New York (1962), pp. 379-418.
11. B. V. Derjaguin, Y. I. Rabinovich and N. V. Churaev, Nature,
272 (1978), pp. 313-318.
12. D. B. Hough and L. R. White, Adv. Colloid Interface Sci., 3
(1980), pp. 3-41.
13. J. N. Israelachvili, Intermolecular and Surface Forces, 2nd
ed., Academic Press, London, (1992).
14. R. A. Bowling, Particles on Surfaces 1: Detection, Adhesion,
and Removal, K. L. Mittal (ed.), Plenum Press, New York
(1988), pp. 129-142.
15. F. Arai, D. Ando, T. Fukuda, Y. Nonoda and T. Oota, Proc. of
the 1995 IEEE/RSJ Int. Conf. on Intelligent Robots and
Systems (IROS95), Vol. 2, IEEE Computer Society Press,
Los Alamitos, (1995), pp. 236-241.
16. K. Higashitani, Engineering System for Fine Particles, Fundamental Technology I, H. Yanagida (supervised), Fuji
Technosystem, Tokyo, (2001), pp. 398-402 (in Japanese).
17. K. Okuyama and K. Higashitani, Powder Technology
Handbook, 3rd ed., H. Masuda, K. Higashitani and H. Yoshida (eds.), CRC Press, Boca Raton, (2006), pp. 183-197.
18. A. Morikawa, Y. Iyoku, M. Kakimoto and Y. Imai, J. Mater.
Chem., 2 (1992), pp. 679-689.
19. A. Morikawa, Y. Iyoku, M. Kakimoto and Y. Imai, Polym. J.,
24 (1992), pp. 107-113.
20. Y. Chen and J. O. Iroh, Chem. Mater., 11 (1999), pp.
1218-1222.
21. K. M. Asif, M. I. Sarwar and Z. Ahmad, Mat. Res. Soc.
Symp. Proc., 576 (1999), pp. 351-356.
22. B. Liu, K. Asuka, M. Terano and K. Nitta, Current
Achievements on Heterogeneous Olefin Polymerization
Catalysts, M. Terano (ed.), Sankeisha, Nagoya, (2004), pp.
119-126.
23. K. Nitta, Seikei-Kakou, 18 (2006), pp. 114-117, (in
Japanese).
24. J. L. Keddie, R. A. L. Jones and R. A. Cory, Europhys. Lett.,
27 (1994), pp. 59-64.
25. J. A. Forrest, K. Dalnoki-Veress, J. R. Stevens and J. R.
Dutcher, Phys. Rev. Lett., 77 (1996), pp. 2002-2005.

1244