Publisher’s version / Version de l'éditeur:
Rheologica Acta, 41, 5, pp. 383-482, 2002-08-01
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
https://nrc-publications.canada.ca/eng/copyright
Vous avez des questions?
Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la
première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez
pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the
first page of the publication for their contact information.
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. /
La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version
acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien
DOI ci-dessous.
https://doi.org/10.1007/s003970100211
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Dynamic melt flow of nanocomposites based on poly-ε-caprolactam
Utracki, Leszek A.; Lyngaae-Jorgensen, Jorgen
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site
LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
NRC Publications Record / Notice d'Archives des publications de CNRC:
https://nrc-publications.canada.ca/eng/view/object/?id=f4f94f87-d0d3-4d72-9c44-611087628ee8
https://publications-cnrc.canada.ca/fra/voir/objet/?id=f4f94f87-d0d3-4d72-9c44-611087628ee8
Introduction
Polymeric nanocomposites (NC) are materials that
contain dispersed nanometer-size particles in a single
or multi-component polymeric matrix. The nano-parti-
cles can be lamellar, ®brillar, tubular, shell-like, spher-
ical, etc. To ascertain good dispersion in a hydrophobic
melt, they are usually reacted with organic compounds,
e.g., sodium montmorillonite (Na-MMT) with an onium
salt, forming organoclay.
The majority of commercial NC contain 2±10 wt% of
organoclay. NCs have been prepared with virtually all
commodity and engineering, thermoplastic and thermo-
set resins (Utracki and Kamal 2002). However, from the
long list, the historically ®rst and commercially the most
important are NCs based on polyamides, especially on
poly-e-caprolactam (PA-6). It has been reported (Ube
Industries 2000) that addition of as little as 2 wt% of
organoclay (in relation to PA) increases the density by
less than 1%, tensile and ¯exural modulus by 70 and
130%, respectively, HDT by 70 °C, and reduces oxygen
and moisture permeability by 50%, ¯ammability by
70%, etc. The NCs are being used in the transport,
packaging, and aerospace industries (Dagani 1999).
From the fundamental point of view, the reinforcing
eect of nano-particles is related to the aspect ratio, p
Rheol Acta (2002) 41: 394±407
Ó Springer-Verlag 2002
ORIGINAL CONTRIBUTION
Leszek A. Utracki
Jùrgen Lyngaae-Jùrgensen
Dynamic melt ¯ow of nanocomposites based
on poly-e-caprolactam
Received: 24 May 2001
Accepted: 27 August 2001
L. A. Utracki (&)
National Research Council Canada,
IMI, Boucherville, QC,
Canada J4B 6Y4
e-mail: Leszek.utracki@nrc.ca
J. Lyngaae-Jùrgensen
Dept. Chem. Eng.,
Technical University of Denmark,
Lyngby, Denmark
Abstract The dynamic ¯ow behav-
ior of polyamide-6 (PA-6) and a
nanocomposite (PNC) based on it
was studied. The latter resin con-
tained 2 wt% of organoclay. The
two materials were blended in pro-
portions of 0, 25, 50, 75, and
100 wt% PNC. The dynamic shear
rheological properties of well-dried
specimens were measured under N
2
at T 240 °C, frequency x 0.1±
100 rad/s, and strains c 10 and
40%. At constant T, c, and x the
time sweeps resulted in signi®cant
increases of the shear moduli. The c
and x scans showed a complex
rheological behavior of all clay-con-
taining specimens. At c 10% the
linear viscoelasticity was observed
for all compositions only at
x>1 rad/s, while at c 40% only
for 0 and 25 wt% of PNC. However,
the eect was moderate,
namely decreasing G¢ and G¢¢
(at x 6.28 rad/s; c 50%) by 15
and 7.5%, respectively. For compo-
sitions containing >25 wt% PNC
two types of non-linearity were de-
tected. At x £ x
c
1.4 0.2 rad/s
yield stress provided evidence of
a 3-D structure. At x > x
c
,G¢
and G¢¢ were sensitive to shear
history ± the eect was reversible.
From the frequency scans at x > x
c
the zero-shear relative viscosity vs
concentration plot was constructed.
The initial slope gave the intrinsic
viscosity from which the aspect ratio
of organoclay particles, p 287 9
was calculated, in agreement with
the value calculated from the re-
duced permeability data, p 286.
Keywords Polymeric nanocompos-
ites á Dynamic melt ¯ow á
Organoclay á Montmorillonite á
Poly-e-caprolactam á Platelets sus-
pensions á Aspect ratio
(ratio of ®ber length or platelet thickness to its
diameter), and to the particle-matrix interactions. Inde-
pendent of the actual dimension, for p ³ 5p
c
the
reinforcing eect is the same as of an in®nitely large
particle, increasing the strength of the composite to that
of the platelet strength. The critical value of the aspect
ratio, p
c
, is given by the ratio of the tensile strength to
the interfacial shear strength ± for most polymeric
composites p
c
@ 100 (Piggott 1980). The available orga-
noclays have p @ 10±2000. Evidently, the importance of
p depends very much on the source (mineral or
synthetic), method of preparation, and that of process-
ing. High p-values are crucial for the NC that is to be
used to reduce permeability, as well as for generation of
anisotropic mechanical performance, but less so for, e.g.,
reduction of ¯ammability (Gilman et al. 2000).
Preparation of organoclays for the use in polymers is
a laborious and exacting process that starts with the
puri®cation of the ore, grinding, ion exchange, then
intercalation and drying (Knudson and Jones 1992;
Clarey et al. 2000). Thus the organoclay price is
relatively high, namely for mineral and synthetic,
respectively, 1.6 and 2.3 US$/kg to be compared with
2.9 US$/kg for PA-6 (all mid-2001 prices).
The technology of NC that is an object of this paper
(abbreviated as PNC) has been described in several
patents and articles from the Toyota Research Corpo-
ration (now Toyota Central R & D Labs.) and Ube
Industries (namely Okada et al. 1988; Deguchi et al.
1992; Okada and Usuki 1995). To produce PNC, ®rst
Na-MMT was intercalated with 12-aminolauric acid in
an aqueous medium, which increased the interlayer
spacing from d
001
0.96 nm (dry Na-MMT) to 1.3 nm.
Next, the organoclay was dispersed in the monomer and
polymerized ± during the polymerization d
001
increased
to about 7 nm. The ®nal exfoliation of clay platelets
takes place during processing.
Because of its small thickness and large p, the
exfoliated clay has the speci®c surface area of about
750±800 m
2
/g. This high value is responsible for the
large eects clay has on the NC performance. It has been
postulated that the optimal nanocomposite structure
consist of disordered, exfoliated clay platelets dispersed
in a polymeric matrix. However, as the aspect ratio
increases, the encompassed volume of the platelet
increases with the cube of its diameter. Thus, platelets
with high aspect ratio are characterized by a low value of
the maximum packing volume fraction, /
m,plate
(3/
2p)/
m,sphere
@ 1/p, above which they are unable to rotate
freely, and locally have to orient parallel to each other
(Utracki 1995). In consequence, as Okada and Usuki
(1995) showed, the interlayer spacing hyperbolically
decreases with increasing clay content (from 22 nm at
5 wt% to 5 nm at 25 wt% of clay).
The ¯ow of PNC is sensitive to the dispersed particle
size, shape, and surface characteristics as well as to the
transient geometrical structure and complex interactions
that involve the matrix polymer, clay, and a compati-
bilizer. Thus, rheology complements the traditional
methods of NC characterization, such as X-ray dirac-
tion (XRD), transmission electron microscopy (TEM),
permeability, or mechanical testing.
Few rheological studies on NCs have been published.
Krishnamoorti et al. (1996) reported on the dynamic
and steady-state ¯ow of polydimethylsiloxane (PDMS)
containing up to 15 wt% layered silicate. Linear increase
of the zero-shear viscosity (g
o
) with silicate loading was
observed, but at high rates of deformation the matrix
viscosity was nearly recovered. Surprisingly, exfoliation
reduced the enhancement of g
o
. The authors also tested
PA-6 with organoclay (PNC). This preliminary work
was expanded in the following publication that consid-
ered behavior of PNC containing 2 and 5 wt% of
organoclay and NC of poly-e-caprolactone with up to
10 wt% of organoclay (Krishnamoorti and Giannelis
1997). Since these systems were obtained by intercala-
tion followed by polymerization that engendered a direct
bonding between MMT surface and macromolecules,
the authors labeled them as ``end-tethered polymer
layered silicate nanocomposites''. The G¢ and G¢¢ moduli
were found to increase with clay loading. Their power-
law dependence in the terminal zone was dierent from
that observed for homopolymers. At low frequencies the
rheological response of samples containing high level of
clay showed almost a solid-like response.
It is to be expected that the rheological behavior of
the end-tethered systems will somehow dier from that
in others where such direct bonding is missing. Thus,
Homann et al. (2000a) showed that tethering has a
dramatic eect on the degree of dispersion and rheology.
Two types of polystyrene-based nanocomposites were
prepared. In the ®rst the clay was intercalated with
phenyl ethyl amine, while in the second with amine-
terminated PS. In the ®rst system clay was intercalated,
in the second it was exfoliated. A plot of G¢ vs reduced
frequency (xa
T
) showed that neat PS and PS containing
intercalated clay nearly superposed one on top of the
other, whereas the exfoliated nanocomposite showed a
large increase of G¢ (especially at low x ), indicating a
network formation.
In the following publication (Homann et al. 2000b)
melt ¯ow behavior of NC based on polyamide-12 was
studied. Two types of clay were used: synthetic ¯uoro-
mica and mineral MMT. The clay was intercalated using
either protonated amino dodecanoic acid (ADA) or
water, then dispersed in ADA, which in turn was
polymerized. The use of ADA-intercalated clay resulted
in formation of end-tethered structure, with exfoliated
silicate layers chemically bonded to the matrix. By
contrast, the use of water as a swelling agent resulted in
nanocomposites comprising exfoliated silicate layers
well dispersed in the polymer matrix, but without
395
attached polymer chains. A stress-controlled rheometer
was used in the dynamic mode with parallel plate
geometry at frequency x 1±25 rad/s. The data were
collected within the linear viscoelastic region. The
rheological behavior of the NCs diered from that of
neat PA-12 matrix. The presence of a super-structure
was deduced from the low frequency behavior. For
PNCs with polymer molecules not bonded to the clay
surface the ¯ow mainly depended on the matrix with
only a minor in¯uence of clay (at least up to 4 wt%
loading). Tethering enhanced G¢ and G¢¢ by one and by
one-half decade, respectively.
Schmidt et al. (2000) studied shear orientation of
polymer-clay solutions during Couette ¯ow, probed by
birefringence and SANS. Thus, 3 wt% of synthetic
hectorite (platelets: d 30 nm and h @ 1 nm) was dis-
persed in aqueous solution of 2 wt% polyethylene glycol
(PEG, M
w
10
3
kg/mol) at pH 10 and an NaCl
concentration of 10
)3
mol/l. To account for the SANS
and birefringence results, the authors postulated that the
polymer chains are adsorbed onto the clay particles. The
birefringence indicated a mechanical coupling between
clay platelets and polymer; at low rates of shear,
_
c <
_
c
critical
' 30s
1
its value was dominated by clay
platelets, but at high by polymer chains stretched in the
¯ow direction. SANS data indicated that at
_
c >
_
c
critical
the ¯ow is strong enough to induce orientation. The clay
platelets (within aggregates, having diameter d 32±
233 nm) were oriented in the ¯ow direction with the
surface normal in the neutral (not radial) direction.
Linear viscoelastic ¯ow of NCs based on polystyrene-
polyisoprene diblock copolymer was studied by Ren
et al. (2000). MMT was intercalated with dimethyl-
dioctadecyl-ammonium. At >6.7 wt% of organoclay
the low frequency data showed a pseudo-solid-like
behavior, similar to that observed for exfoliated end-
tethered nanocomposites. The behavior was attributed
to the presence of anisotropic stacks of clay platelets,
each stack randomly oriented vis-a
Á
-vis another, and
forming a percolated 3-D network that was incapable of
fully relaxing. The large-amplitude oscillatory shear was
able to orient these structures and increase their liquid-
like character.
Solomon et al. (2001) reported on the linear and
nonlinear rheology of polypropylene (PP) based NCs
with intercalated MMT (interlayer spacing 2.9 nm) The
Na-MMT was ®rst intercalated with several ammonium
salts (interlayer spacing increased from 1.1 to 2.1 nm).
The NCs were prepared in an internal mixer by melt
compounding with 1.3±6.2% intercalated-MMT and
maleic anhydride grafted-PP (MA-PP) (3 parts per 1
part of organoclay). Addition of clay signi®cant
increased G¢ and G¢¢. To study the viscoelastic non-
linearity the steady-state shear ¯ow reversal was
conducted. First, a PNC specimen was sheared for
300 s at the rate of shear
_
c 0:1 s
1
recording the value
of the shear stress (r
12
), then the ¯ow was stopped for a
time, t
rest
, and the specimen was re-sheared at the same
rate of shear, but in the opposite direction. Character-
istically the magnitude of the stress overshoot (r
max
)
increased with t
rest
. The master curve was constructed
plotting r
max
=r
1
1=cvst
rest
(where r
1
is the value
of r
12
at long time and c is MMT loading). The authors
concluded that an anisometric structure is responsible
for the non-linear ¯ow behavior.
Galgali et al. (2001) investigated creep behavior of
melt intercalated NC of PP with organoclay and MA-
PP. The latter was added at a ratio of either 0:1 or 1:1 in
respect to organoclay. The NCs were characterized by
TEM and wide-angle X-ray diraction (WAXD) at
T 200 °C. The creep compliance at shear stresses,
r
12
10 and 50 Pa was signi®cantly lower for NCs
containing MA-PP, and the eect increased with
annealing. TEM and WAXD showed the presence of
clay aggregates dispersed within the polymer matrix and
a small amount of exfoliated platelets. The zero shear
viscosity (g
o
) strongly depended on clay content. The g
o
of compatibilized NCs containing ³3 wt% clay was at
least three orders of magnitude higher than that of
matrix resin and the uncompatibilized NC. Importantly,
the large increase of g
o
was not accompanied by any
increase in the ¯ow activation energy compared to the
matrix polymer. The compatibilized NC also showed an
apparent yield stress. This solid-like response of molten
NC apparently originates not from macromolecules
entrapped between clay platelets, but rather from the
frictional interactions of clay aggregates. The addition of
MA-PP did not signi®cantly aect WAXD spectra, but
had a signi®cant in¯uence on rheology. Its presence
increased the probability of 3-D formation (the solid-
like rheological response) to the level similar to that
observed for the end-tethered chains. The responsible
mechanism was not identi®ed ± it may be that the MA-
PP enhanced exfoliation of clay platelets, but it could
also form a third phase inside NC that formed a
percolating 3-D network. Even in face of the WAXD
results, the authors clearly prefer the ®rst possibility ±
they postulated that clay platelets exfoliated upon
addition of MA-PP to form bridges between the
aggregates forming a 3-D percolating network. The
creep was found sensitive to the microstructural changes
occurring in the nanocomposite.
Recently Okamoto et al. (2001) studied the exten-
sional ¯ow behavior of polypropylene (PP)/clay NC
containing 0.2 wt % MA-PP and 4 wt% MMT interca-
lated with stearyl ammonium ion via melt extrusion at
200 °C. TEM showed ®ne dispersion of the silicate stacks
ca 150 nm long and about 5 nm thick, while according to
XRD the interlayer spacing was only d
001
@ 2.79 nm.
For the study the rotating clamps, Meissner-type elon-
gational RME rheometer was used at 150 °C at Hencky
strain rates
_
e 0:001 1:0s
1
. The linear viscoelastic
396
envelope of the stress growth function in elongation was
about tenfold higher than that computed from shear
¯ow. Furthermore, a strong strain hardening (SH) was
observed. The former type of behavior is expected from
the multiphase systems with yield stress; however the
second is contrary to expectation ± strain softening has
been reported for composites (Takahashi 1996). Further-
more, unlike single-phase polymer melts, the low defor-
mation rate (
_
c
_
e 0:001 s
1
) stress growth functions of
NC (in shear and elongation) increase with test time,
t £ 300 s, not showing a tendency of reaching a steady
state. The authors concluded that ¯ow-induced internal
structure is dierent in shear than in elongational. TEM
observations indicated that in specimen elongated at
_
e 1:0s
1
the MMT platelets aligned perpendicularly to
the stretching direction, but at
_
e 0:001 s
1
the platelet
orientation was more random, with many plates aligned
in the ¯ow direction. In the later case, according to the
authors a house of cards-like structure was created by
¯occulation, similar to that observed in low intensity
shear ¯ow.
In summary, the rheological studies in shear and
elongation demonstrated that even at low clay loading
NC ¯ow is complicated: (1) at low deformation rates by
a solid-like yield stress behavior, caused by a 3-D
structure; (2) at high strains the clay platelets get
oriented, which causes the shear viscosity to decrease
nearly to that of the matrix. The later eect is
particularly evident in the steady-state ¯ow. Both eects
are stronger for the end-tethered than for free platelets
systems, especially at higher clay loading.
In the studies on polyamide-based NCs the unstable
nature of the polycondensation-type polymeric matrix
must be of concern. The tests must clearly separate
eects caused by the changes in the matrix from these
brought about by the presence of nano-®ller. The
present article summarizes results of the ®rst stage of
the PNC rheological studies. Thus, the ®rst goal is to
study the stability of the PA, PNC, and their mixture.
The second goal is through systematic and quantitative
studies of selected rheological functions to elucidate
mutual in¯uence of the microstructure and ¯ow. Fur-
thermore, the limiting conditions for onset of complex
behavior are of interest.
Experimental
Materials
Polyamide-6 (PA) and nanocomposite (PNC) that was based on it
were obtained from Ube Intl. (commercial grades PA1015B and
PA1015C2, respectively). Characteristics of these resins are pub-
lished (Ube Industries 2000). The PA has the same molecular
weight as the PNC matrix resin. The latter was produced by the
reactor exfoliation method with 2 wt% of organo-clay (equivalent
to 1.6 wt% or 0.64 vol.% of MMT) (Okada et al. 1988; Deguchi
et al. 1992; Okada and Usuki 1995).
The rheological properties of the following seven samples were
measured (see Table 1): PA and PNC as received and ®ve their
mixtures compounded in a Werner-P¯eiderer twin-screw extruder
(TSE) and re-pelletized: 1. PA, 2. PA with 25 wt% of PNC, 3. PA
with 50 wt% of PNC, 4. PA with 75 wt% of PNC, and 5. PNC.
Note that the two commercial resins were tested using ``as
received'' and re-extruded PA and PNC specimens.
Drying
Before the rheological tests the PA and PA-based materials have to
be dried to a standard level. Thus, pelletized samples were dried
under vacuum at 80 °C for up to a week. The complex shear
modulus at two frequencies vs drying time is shown in Fig. 1.
Calculations show that after 48 h drying the rheological signal is
0.4% below the ultimate level for t®1. Since a similar conclusion
was reached using the weight loss method (Bureau 2001), these
rheological data do represent removal of volatiles, not a chemical
modi®cation of PA-6.
Rheological tests
After drying (48 h at 80 °C under vacuum) the pellets were directly
loaded into a controlled strain rate rheometer (ARES from
Rheometric Scienti®c). Parallel plates of 25 mm diameter were
used for the dynamic tests at T 240 °C. For the specimens
temperature stabilization, 3 min soak time was allowed. The time,
strain, and frequency sweeps were carried out under a blanket of
dry nitrogen. The data were reproducible within 2%. To con®rm
some results, tests were also carried out using the Rheometrics
Mechanical Spectrometer, RMS.
Results
Time sweeps
Time sweeps were carried out for 1 h at frequency
x 6.28 rad/s and strains c 10 and 40%. All seven
Table 1 Polynomial ®t parameters for the time sweeps. The ®rst ®ve samples were re-extruded or compounded, the last two ``as received''
PNC wt% G¢
1
G¢
2
G¢
3
r
G¢
G¢¢
1
G¢¢
2
G¢¢
3
r
G¢¢
0 (PA-6) 39.388 0.040822 0.0000 0.99996 1926.9 0.55258 ±4.5154e-05 0.99983
25 82.285 0.061125 0.0000 0.99993 2194.0 0.48794 ±3.1387e-05 0.99977
50 152.88 0.079951 ±2.6699e-06 0.99987 2560.4 0.50421 ±3.4721e-05 0.99970
75 242.65 0.093212 ±3.8080e-06 0.99978 2898.0 0.44210 ±2.7627e-05 0.99939
100 (PNC) 321.27 0.095999 ±4.2305e-06 0.99971 3133.3 0.37305 ±2.2062e-05 0.99859
0 (PA-6) 35.839 0.047436 0.0000 0.99995 1918.2 0.65015 ±4.9298e-05 0.99958
100 (PNC) 559.74 0.17737 ±1.0141e-05 0.99962 3773.2 0.54600 ±2.8488e-05 0.99888
397