Introduction
The blending of polymeric melts with inorganic clays yield polymer nanocomposites which exhibit not only a dramatic increase in tensile strength and heat resistance but also a drastic decrease in gas permeability compared with those of a pure polymer matrix (Kojima et al. 1993;
Yano et al. 1993; Vaia et al. 1995; Messersmith and Giannelis 1995; Krishnamoorti et al. 1996). These unique properties render the nanocomposites as ideal materials for products that range from high barrier packaging for food and electronics to strong, heat- resistant automotive components. From the structural point of view, two types of polymer-clay nanocompos- ites are possible. Intercalated nanocomposites are formed when one or a few molecular layers of the polymer are inserted into the clay gallery with ®xed
interlayer spacings. Exfoliated nanocomposites are formed when the silicate nanolayers are individually dispersed in the polymer matrix, the average distance between the segregated layers being dependent on the clay loading. The separation between the exfoliated nanolayers may be uniform (regular) or variable (disor- dered). Exfoliated nanocomposites show greater phase homogeneity than intercalated nanocomposites. This structural dierence is one of the primary reasons that the exfoliated state is so eective in improving the performance properties of clay composite materials.
Owing to their unique phase morphology and improved interfacial properties, the nanocomposites exhibit im- proved performance properties compared to conven- tional composites. From the processing and application points of view, the mechanical and rheological proper- ties of these nanocomposites are of vital importance. It is Yong Taik Lim
O. Ok Park
Phase morphology and rheological behavior
of polymer/layered silicate nanocomposites
Received: 20 March 2000 Accepted: 11 September 2000
Y. T. LimáO. O. Park (&)
Center for Advanced Functional Polymers Department of Chemical Engineering Korea Advanced Institute of Science and Technology, 373-1, Kusong-dong Yusong, Taejon, Korea
e-mail: [email protected]
Abstract Rheological behavior of polymer/layered silicate nanocom- posites are strongly dependent not only upon their microstructure but also upon the interfacial character- istics. Dierent phase morphology (intercalated or exfoliated) of poly- mer/clay is obtained according to interfacial characteristics between polymer chains and clay. In interca- lated structure, the presence of ran- domly oriented anisotropic stacks of silicate layers is responsible for the enhancement of both moduli. The PS/clay nanocomposites exhibit a slight enhancement at low frequency because of its simple intercalated structure and little interaction. On the other hand, the PS-co-ma/clay nanocomposites have a similar in-
tercalated structure but exhibit a distinct plateau-like behavior at low frequency since the PS-co-ma has a strong attractive interaction with the silicate layers. Finally, PE-g-ma/clay nanocomposites display an exfoliat- ed structure, which exhibit both a distinct plateau-like behavior at low frequency and enhanced moduli at high frequency. Percolation struc- ture as well as large interfacial area between polymer chains and clay are responsible for the rheological behavior.
Key words Nanocompositeá Layered silicateáIntercalated structureá Exfoliated structureá Non-terminal behavior
also important to relate their mechanical and rheological properties to the nature and microstructure of the nanocomposites. Furthermore, these nanocomposites appear to be an ideal system to probe the dynamics and statics of con®ned polymers.
Recently, some studies have reported on the structure and dynamics of the layered silicate based polymer nanocomposite (Krishnamoorti et al. 1996; Krishnamo- orti and Giannelis 1997). The rheology of two end- tethered exfoliated hybrid systems prepared by in-situ polymerization (poly(e-caprolactone)/montmorillonite, nylon-6/montmorillonite) wherein the polymer chains are end-tethered to the silicate surface via cationic surfactants, were also studied (Krishnamoorti and Giannelis 1997). The reports revealed that the power- law dependence of both moduli in the terminal zone is dierent from that observed in homopolymers. And they suggested several hypotheses to explain the non-terminal rheological behavior of their systems. They also discussed the similarity to the non-terminal low-frequency response shown in ordered block copolymers and smectic liquid- crystalline small molecule (Rosedale and Bates 1990;
Kawasaki and Onuki 1990; Chen et al. 1997; Larson et al. 1993; Kannan and Korn®eld 1994; Tepe et al.
1997). But their system is limited only to the exfoliated structure wherein the silicate surface and polymer chains are end-tethered with strong interaction. The rheological behavior of polymer/clay nanocomposite materials is determined not only by the parent components but also by the composite phase morphology and interfacial properties. Until now, few studies have reported the systematic relationship of phase morphology and interfacial properties with the rheological behavior of polymer/layered silicate nanocomposites.
Recently, we investigated the rheological evidence for the microstructure of intercalated polystyrene/layered silicate nanocomposites (Lim and Park 2000). The rheological measurement method monitored the micro- structural change which causes non-terminal behavior.
However, there have been no reports which relate the rheological behavior of polymer/layered silicate nano- composites to their phase morphology (intercalated or exfoliated) and to the interfacial characteristics between the polymer chains and layered silicates. In this study, polystyrene (PS)/clay, polystyrene-co-maleic anhydride (PS-co-ma)/clay and polyethylene-graft-maleic anhy- dride (PE-g-ma)/clay nanocomposites, which have dif- ferent interfacial properties and phase morphology, are fabricated. PS-co-ma/clay nanocomposites are prepared to observe the eect of the interface on the phase morphology and rheological behavior compared with those of PS/clay nanocomposites. In addition, PE-g-ma is used to fabricate exfoliated nanocomposites with clay.
Pure PE can produce only an immiscible state between the polymer and clay since there are no interacting moieties in the PE backbone. Furthermore, we also ®nd
that the ®nal morphology of PS-co-ma/clay and PE-g- ma/clay nanocomposites are very dierent although the maleic anhydride groups both in PS-co-ma/clay and PE- g-ma/clay composite systems act as attractive moieties with the polar group of silicate surface and stick to the silicate surface. The rheological behavior of polymer/
clay nanocomposites is also dependent on their micro- structure (Intercalatedor Exfoliated).
In the following section, we begin with the phase morphology of three dierent polymer/clay systems.
In addition, three dierent types of rheological behavior of polymer/layered silicate nanocomposites are identi-
®ed according to their microstructure and interfacial properties.
Experimental
Materials
Three polymer resins with dierent characteristics were used to fabricate the nanocomposites with clay: polystyrene (PS), polysty- rene-co-maleic anhydride (PS-co-ma) and polyethylene-graft- maleic anhydride (PE-g-ma). The weight-average molecular weight (Mw) of PS, which was obtained from LG Chem. Co. Ltd., is 346,000 and the polydispersity index is 2.8 (characterized by Mw/ Mn). The weight-average molecular weight (Mw) of PS-co-ma (random copolymer), which was obtained from Aldrich Chem. Co.
Ltd., is 224,000. The content of maleic anhydride group is 7 wt%.
The weight-average molecular weight (Mw) of PE-g-ma, which was also obtained from Aldrich Chem. Co. Ltd., is 212,000. The content of maleic anhydride in PE-g-ma is 0.8 wt%. Commercial- ized organophilic clays (Cloisite-6A, hereafter M6A) produced by Southern Clay Products were used. Prinstine Na+-Montmorillo- nite was changed to organophilic by cation exchange reaction with dimethyl dihydrogenated tallow, ammonium ion and the cationic exchange capacity (CEC) is 140 meq/100 g.
Preparation of the nanocomposites
Three dierent series of PS/clay, PS-co-ma/clay and PE-g-ma/clay nanocomposites were fabricated in a Brabender batch mixer by introducing the premixed PS/clay, PS-co-ma/clay and PE-g-ma/
clay (simply stirred polymer pellet with organoclay) into the mixer heated to 210°C. Mixing was continued for 10 min at a 45 rpm mixing speed. The weight percent of clay is 3 wt%, 5 wt%, and 10 wt% in each series of composites.
Rheometry
Three dierent series of composites prepared in a Brabender mixer were compression-molded (cylinder form: 25 mm diameter, 2 mm height) at 210°C to make samples for measuring the rheological behavior. The rheological properties of each nanocomposite series were measured by ARES (Advanced Rheometric Expansion System) in an oscillatory mode with a parallel plate geometry using 25 mm diameter plates at 220°C. All measurements were performed with a 2 K FRTN1 transducer with a lower resolution limit of 0.02 g cm. Typical sample thickness ranged from 0.5 mm to 1.3 mm. Some measurements were also performed with thinner samples (thickness: 0.30.4 mm) to ensure whether the surface eects resulted from the possible surface-induced ordering of the layered structure is negligible. The data for the thinner samples
were in good agreement with those from the thicker samples: this indicates that there was little or no in¯uence of the surface on the microstructure.
Microstructure of nanocomposites
The degree of swelling and the interlayer distance of the clay in the nanocomposites were studied by means of wide angle X-ray diraction. The X-ray diraction spectra were collected on a Rigaku Inc. h-h diractometer equipped with an intrinsic germa- nium detector system using Cu Ka (k1.54 AÊ) radiation at a generator voltage 45 kVand current of 40 mA. The scanning rate is 2°/min from 1.2°to 10°. Transmission electron micrographs were taken from 60±100 nm thick, microtomed sections of polymer/clay composites using a JEOL 1200 TEM with 120 kVaccelerating voltage.
Results and discussion
Phase morphology in polymer/layered silicate nanocomposites
Figure 1 shows that dierent ®nal phase morphologies were formed in PS/M6A, PS-co-ma/M6A and PE-g-ma/
M6A nanocomposites. Both PS/M6A and PS-co-ma/
M6A composites have intercalated structure, while the original ordered structure of M6A is delaminated and exfoliated structure is formed in case of the PE-g-ma/
M6A. The X-ray diraction pattern shows that the basal re¯ections from unintercalated M6A, p-(001) and p- (002) are observed at 2h2.75 and 2h5.20, respec- tively. But, the basal peak of the M6A was shifted to higher angle (2h2.98) with heat treatment, which means that the oxidative decomposition of the oligomer
which reside in the gallery of silicate layer and collapse of the silicate layer as was reported in the previous works (Ogawa 1994). The heat treatment was done at T210°C for 10 min in the heating chamber to see the thermal eect of the processing condition on the change of the microstructure of M6A. And, the repeating distance perpendicular to layer corresponds to d0012.94 nm. The basal re¯ection of intercalated M6A in PS/M6A and PS-co-ma/M6A are observed at 2h2.56 and 2h2.62, respectively, and the repeating distance perpendicular to the silicate layer increases to d0013.448 nm and d0013.369 nm, respectively. It means that the distance of silicate layer increases with intercalation of polymer chains. And the highly magni-
®ed TEM images in Fig. 2a, b show the intercalated structure. Recently, the microstructural change of layered silicate with intercalation of PS chains (Vaia et al. 1993, 1995; Vaia and Giannelis 1997) was moni- tored by rheological measurement method (Lim and Park 2000). The ®nal phase morphology of PS-co-ma/
M6A is similar to that of PS/M6A by judging from their X-ray diraction data, even though PS-co-ma has more polar groups which act as more attractive moieties to the silicate surface. However, if we can consider the TEM images shown in Fig. 2, we can clearly see the eect of the maleic anhydride group on the ®nal phase morphol- ogy of the PS-co-ma/M6A nanocomposites in Fig. 2. If we compare Fig. 2c and Fig. 2d, more homogeneous dispersion of M6A is noticed in PS-co-ma/M6A nano- composites. The more eective distribution of silicate particles in the nanocomposites may be due to the strong interaction between the maleic anhydride group of PS- co-ma and silicate surface, which experiences more eective shear forces during the compounding process.
Another interesting aspect of the phase morphology of the nanocomposites, by judging from their X-ray diraction data, is that the ®nal morphology of PS-co- ma/clay and PE-g-ma/clay nanocomposites is very dierent, even though the maleic anhydride groups in PS-co-ma/clay and PE-g-ma/clay composite systems act as attractive moieties with the polar group in silicate surface and stick to it.
To elucidate the dierent morphology in PS-co-ma/
clay and PE-g-ma/clay, the interaction between PS or PE with the clay should be understood. Previous studies (Vaia et al. 1993, 1995; Vaia and Giannelis 1997), which have been done for the fabrication of PS/clay nanocom- posites by direct melt intercalation, showed that the main driving force of PS intercalation to the gallery of clay is the Lewis acid-base interaction between the PS and the silicate surface. On the other hand, there are no interacting moieties in polyole®n such as PE which leads to an immiscible state between the polymers and clay. So, a functional polar group such as maleic anhydride should be introduced to the polymer backbone in order to fabricate the nanocomposite with clay. When the PE-g-
Fig. 1 X-ray diraction patterns of PS/M6A, PS-co-ma/M6A and PE-g-ma/M6A nanocomposites (5 wt% clay loading)
ma chains intercalate into the interlayer of the layered silicate due to the interaction of the maleic anhydride group with the silicate surface, the PE backbone itself does not have any interaction with the silicate surface and it tends to retain its coil-like conformation in the gallery of layered silicate; this results in forcing the adjacent silicate layers to separate and leads to the fully exfoliated structure. On the other hand, the PS chains intercalate into the silicate layer because of the Lewis acid-base interaction with the silicate layer and reside in the gap of the silicate layer with the planar single chain morphol-
ogy, and this results in intercalated morphology. The PS- co-ma chains have a maleic anhydride group which functions as a stronger active site with the silicate layer.
Therefore, the PS-co-ma chains in the silicate gallery stick to the silicate layers more strongly than PS chains, and this also results in an intercalated morphology. Sche- matics for the ®nal phase morphology in PS/M6A, PS-co- ma/M6A and PE-g-ma/M6A are described in Fig. 3.
Recent theoretical descriptions (Balazs et al. 1998;
Zhulina et al. 1999) of the phase morphology in polymer/layered silicate nanocomposites are consistent
Fig. 2a±d Transmission Elec- tron Micrograph of PS/M6A and PS-co-ma/M6A nanocom- posites:a, cPS/M6A;b, d PS-co-ma/M6A
with our experimental results. Both numerical and analytical self-consistent theory (SCF) models were proposed to isolate a robust scheme for creating stable dispersion from polymers and clays that are immiscible.
Speci®cally, they showed that it is possible to lead to the formation of ``exfoliated'' hybrids if a small fraction of end-functionalized polymers is added to the melt of chemically identical, non-functionalized chains.
Consider the case where the interaction parameter between the polymer and silicate layers, v is less than zero (or vsurf< 0). Then the polymer and silicate surface would experience an attractive interaction so that the polymer diuses through the energetically favorable gallery. As the polymer contacts with the two con®ning layers, the polymer ``glues'' the two surfaces together as it moves through the interlayer.
This ``fused'' condition could represent the attraction between the polymer and clay sheets would only lead to intercalated, rather than exfoliated structures. Converse- ly, in the case wherev> 0, the polymer can separate the sheets, as the chain tries to retain its coil-like confor- mation and gain entropy. The SCF calculations indicate that for v> 0, polymers and sheets are immiscible.
However, if the polymer contains a fragment that is highly attracted to the silicate surface, then the facile penetration into the gallery can be easily accomplished.
In addition, this polymer must contain a longer fragment that is not attracted to the silicate. The non-
attractive block will attempt to gain entropy by pushing the layered silicate apart. In fact, one may wonder if the eect of molecular weight for each polymer is also important factor to the ®nal phase morphology of polymer/clay nanocomposites. But, as discussed in other research (Vaia et al. 1993; Vaia and Giannelis 1997), the molecular weight aects only the kinetics of the intercalation of polymer chains into the silicate layer, not on the ®nal phase morphology.
Eect of interface on rheological behavior of polymer/layered silicate nanocomposites
The eect of ®ller on the rheological behavior of polymer/®ller composites has been widely studied in conventional composite systems (Aoki 1987; Poslinski et al. 1988; Baradollet et al. 1995; Kyu et al. 1996;
Lakdawala and Salovey 1987; Gandhi and Salovey 1988). And the results showed that the interaction between polymer chains and ®ller strongly aect the rheological behavior of the composites. To understand the eect of the interfacial properties on the rheological behavior of polymer/layered silicate nanocomposite, the rheological behavior of the PS-co-ma/M6A nanocom- posites is compared with that of the PS/M6A nanocom- posites. The dynamic mechanical properties of the PS/
M6A and PS-co-ma/M6A nanocomposites are shown in Figs. 4 and 5, respectively. The rheological behavior of the PS-co-ma/M6A nanocomposites at low frequency is very dierent from that of the PS/clay nanocomposites.
In particular, enhancement of the storage modulus is very large compared to that of the PS/clay nanocom- posites and plateau-like behavior is also exhibited.
In PS/M6A nanocomposites, the PS chains are simply inserted between layers and there is no notable strong interaction between the PS and the silicate surface except for weak Lewis acid-base interaction (Vaia et al.
1993). And the eect of the neutral wall is a weak slowdown of the chain relaxations and the segment mobilities. In contrast, 7 wt% maleic anhydride groups in PS-co-ma are attractive to the surface of the silicate layer and the interaction between the polymer chains and the silicate layer is very strong. Thus, the distinct plateau-like behavior of the PS-co-ma/M6A shown in Fig. 5 is due to the strong interaction between the maleic anhydride group of PS-co-ma and polar groups on the clay surface, such as oxygen and hydroxyl groups.
In addition, the more uniform dispersion of clay particles are formed in PS-co-ma/M6A nanocomposites as shown in Fig. 2. As the micro-sized stacked clay particles are delaminated into more smaller particles, the contacting area with polymer chains increases. So the interaction between polymer chains and clay surface is more larger in PS-co-ma/M6A nanocomposites than in PS/M6A nanocomposites at the same silicate loading,
Fig. 3 Proposed mechanism of nanocomposite formation for PS, PS-co-ma, and PE-g-ma with M6A
partly due to maleic anhydride groups and partly due to the large contacting interface area. Furthermore, the clay-clay interactions between clay particles are also very important when the clay particle with large aspect ratio are dispersed in polymer matrix. As shown in Fig. 5, liquid-like behavior of PS-co-ma gradually changes to the pseudo solid-like behavior with silicate loading in excess of 7 wt%. These features suggest that presence of some mesoscopic arrangement of anisotropic silicate layers is also responsible for the rheological behavior of PS-co-ma/M6A and will be also discussed in the section of rheology of shear-aligned nanocomposites.
Phase morphology and rheology
PS/M6A, and PS-co-ma/M6A nanocomposites have an intercalated structure while the PE-g-ma/M6A nano-
composites have an exfoliated structure as suggested by Fig. 1. To understand the relationship between phase morphology and rheological behavior in the polymer/
clay nanocomposites, the dynamic mechanical proper- ties of the PE-g-ma/M6A nanocomposites are measured.
The rheological behavior of the PE-g-ma/M6A nano- composites is shown in Fig. 6. The storage and loss moduli increase and the frequency dependence of both moduli decreases with clay loading compared to those of the matrix polymer. The enhancement of both moduli in PE-g-ma/M6A is larger than those of PS/M6A and PS- co-ma/M6A nanocomposites at all frequency regions.
Although strong interaction between maleic anhydride with clay also exist in PE-g-ma/M6A nanocomposites as in PS-co-ma/M6A nanocomposites, remarkable en- hancement of both moduli are displayed even at low silicate loadings such as 3 wt% and 5 wt%. The
Fig. 4a, b Dynamic mechanical properties of PS/M6A nanocompos- ite at 220°C:astorage modulus;bloss modulus (PS/M6A-03, PS/
M6A-05, and PS/M6A-10 means that 3 wt%, 5 wt%, and 10 wt% of M6A are mixed with PS, respectively)
Fig. 5a, b Dynamic mechanical properties of PS-co-ma/M6A nano- composite at 220°C:astorage modulus;bloss modulus (PS-co-ma/
M6A-03, PS-co-ma/M6A-05, and PS-co-ma/M6A-10 means that 3 wt%, 5 wt%, and 10 wt% of M6A are mixed with PS-co-ma, respectively)
dierent phase morphology of polymer/layered silicate nanocomposites may be responsible for the dierent rheological behavior of the nanocomposite series.
In intercalated nanocomposites, the locally ordered layer structures of alternating polymer chains and silicate layers are dispersed in the polymer matrix with macroscopically disordered state. But their layer struc- ture is similar to the unintercalated layered structure with the only dierence being an expansion of the gallery height due to intercalated polymer chains. So, the rheological behavior of intercalated nanocomposites are similar to that of conventional composites, except that the contacting interface between polymer and ®ller is larger in intercalated nanocomposites.
When the single layers of clay are dispersed in the polymer matrix, more polymer chains come into contact with the silicate surface and ®ller-®ller interactions
between clay particles are also very important, even at low silicate loading. When the layered silicate with 10 stacking layers (300 nm´ 500 nm´ 1 nm), for instance, are exfoliated to a single layer, the contacting interface increase is about 10 times. The eective surface area increases more rapidly as the content of silicate increas- es. To see the eective domain structure of the exfoliated nanostructure, we consider the layered silicate as a disk type. Each disk is considered to be circular with radiusR and thickness 2a. We denote the average number density of the disks by n. The average volume fraction of the disks is thus
/2paR2n 1
The disk radius to thickness aspect ratio is denoted by a R
2a 2
and is assumed to be much greater than unity. The concentration of disks are delineated by the value of the dimensionless parameternR3. (Note that this can also be written nR3/a/p). A relatively concentrated region can be obtained with a low content of ®ller correspond- ing to the situation nR31, but /1. It turns out that both inequalities are easily satis®ed in composites with high aspect ratio particles such as clay, so that the relatively concentrated regime spans a rather broad range of concentration. As an example, a composite with only 5 vol% particles (/0.05) and particle aspect ratio a1000 is characterized bynR3» 16, which is well into the relatively concentrated regime. So, the ®ller-®ller interaction between clay particles (even at low silicate loading) as well as the large contacting interface with polymer chains are responsible for the dierent rhe- ological behavior of exfoliated PE-g-ma/M6A nano- composites, compared with those of intercalated nanocomposites.
To compare the eects of interaction and phase morphology between polymer chains and clay on the rheological behavior of polymer/clay nanocomposites more thoroughly, the relative increase (G¢Com/G¢Homo) of storage modulus in each nanocomposite series were plotted against clay loading in Fig. 7. The enhancement of storage modulus at low frequency region (x 0.12 rad/s) was large in PE-g-ma/M6A and PS- co-ma/M6A nanocomposites. In PS-co-ma/M6A nano- composites, strong interaction between polymer and clay is responsible for the rheological behavior, while clay- clay interactions as well as large interfacial area between polymer chains and clay is responsible for the rheolog- ical behavior of PE-g-ma/M6A. And the PE-g-ma/M6A nanocomposite, which has exfoliated nanostructure, showed large enhancement of the storage modulus both at low and high frequency (x119 rad/s). Especially, the enhancement of storage modulus was largest in PE- g-ma/M6A nanocomposites, even at low clay loading.
Fig. 6a, b Dynamic mechanical properties of PE-g-ma/M6A nano- composite at 220°C:astorage modulus;bloss modulus (PE-g-ma/
M6A-03, PE-g-ma/M6A-05, and PE-g-ma/M6A-10 means that 3 wt%, 5 wt%, and 10 wt% of M6A are mixed with PE-g-ma, respectively)
Rheology of shear-aligned nanocomposites
The eects of phase morphology and interfacial prop- erties on the rheological behavior of polymer/layered silicate nanocomposites are also demonstrated by the small amplitude oscillatory frequency test after large amplitude oscillatory shear (LAOS). Application of LAOS leads to a shear-aligned sample. These measure- ments were carried out with 10 wt% ®ller samples and during-shear moduli showed a decrease with continual shearing and ®nally reached a plateau value. The modulus decreased monotonically and the stress signal remains sinusoidal. The plateau value was reached after about t1800 s for PS/M6A and PS-co-ma/M6A nanocomposites and t7200 s for PE-g-ma/M6A nano- composites, respectively. Figure 8 shows the storage moduli of PS/M6A, PS-co-ma/M6A and PE-g-ma/M6A nanocomposites after LAOS (c120%, x1 rad/s).
Very dierent rheological behavior is shown in each
Fig. 7a, b Relative increase of storage modulus of PS, PS-co-ma, PE- g-ma nanocomposites with M6A at low and high frequency region:
alow frequency (x0.12 rad/s);bhigh frequency (x119 rad/s)
Fig. 8a±c Storage modulus of polymer/clay nanocomposite after LAOS (Large Amplitude Oscillatory Shear). (LAOS was conducted c120%,x1 rad/s):aPS/M6A-10;bPS-co-ma/M6A-10;cPE-g- ma/M6A-10
nanocomposite series. The G¢ of the shear-aligned PS/
M6A nanocomposite is lower than that of the initially unaligned sample and close to that of the matrix polymer. A large decrease of G¢is also observed in the PS-co-ma/M6A nanocomposites, while the magnitude of G¢ at low frequency is larger than that of PS/M6A, which re¯ects the strong interaction between the PS-co- ma and silicate layer. That LAOS can signi®cantly alter the small-strain linear viscoelastic response, and indi- cates that there is some mesoscopic arrangement of the silicate layers, which is organized by the application of large amplitude shear. The large decrease of G¢, after LAOS is conducted, suggests that randomly oriented silicate layers and some particle network structure are formed in non-sheared composites, which is observed in PS-co-ma/M6A nanocomposites (in excess of 7 wt%
clay loading), by judging from the rheological behavior of nanocomposites. But the particle network structure in PS-co-ma/M6A nanocomposites, is easily aligned by LAOS, as observed in shear-aligned composites.
However, rheological behavior of shear-aligned PE- g-ma/M6A nanocomposites is very dierent from those of PS/M6A and PS-co-ma/M6A nanocomposites. A decrease of G¢is exhibited only at high frequency: this also re¯ects that the silicate layers are partially oriented.
The decrease of G¢in shear-aligned composite is not so large compared with that of PS/M6A and PS-co-ma/
M6A. The interaction between maleic anhydride group in PE-g-ma and clay surface is partly responsible for the enhanced elastic eect of shear-aligned PE-g-ma/M6A composites, as in PS-co-ma/M6A composites. But the content of maleic anhydride group in PE-g-ma is small compared with that in PS-co-ma. So the facts suggest that some particle network structure still remains in the exfoliated PE-g-ma/M6A nanocomposites after LAOS is conducted.
In fact, the anisotropic particles, which have a very large aspect ratio, are easily percolated when they are dispersed in polymer matrix. Some theoretical and experimental results (Boissonade et al. 1983; Balberg et al. 1984; Philipse 1996) for the percolation of anisotropic particles over a wide range of aspect ratios are reported. And, the principal conclusion is that the particles should have a high aspect ratio to obtain
percolation structure at low volume fraction. When the single layers are dispersed (exfoliated), three-dimension- al percolated network structures are formed at lower silicate loading due to the anisotropy associated with each layer. And the network structure is more complex when the content of clay increases. So the randomly oriented silicate sheets and percolated particle network structure in exfoliated structure (especially in high silicate loading) cannot be completely destroyed by LAOS (suggested by Fig. 8c), although clay particles are partially oriented. And the peculiar particle network structure is responsible for the distinctly enhanced elastic eect of PE-g-ma/M6A nanocomposites, in addition to large interfacial area between polymer and clay.
Conclusions
The relationship between rheological behavior and the microstructure of polymer/clay nanocomposites which have dierent interfacial characteristics is discussed.
Dierent phase morphology (intercalated or exfoliated) of polymer/clay is obtained according to interfacial characteristics between polymer chains and clay. Both storage and loss moduli of the polymer/clay nanocom- posites increased, and the frequency dependence de- creased with silicate loading. In intercalated structure, the presence of randomly oriented anisotropic stacks of silicate layers is responsible for the enhancement of both moduli, which is suggested by LAOS experiments. In particular, large enhancement of elastic eect at low frequency is also shown in intercalated nanocomposites where strong interaction between polymer chains and clay exist. Exfoliated structures show not only plateau- like behavior at low frequency but also enhanced moduli at high frequency. Percolation structure, which re¯ects strong clay-clay interactions, as well as large interfacial area between polymer chains and clay are mainly responsible for the rheological behavior.
Acknowledgement The authors are grateful for the ®nancial supports of the Center for Advanced Functional Polymers (CAFPoly) appointed by the Korea Science and Engineering Foundation. This work is also partially supported by the Brain Korea 21 Project of Ministry of Education(MOE) of Korea.
References
Aoki Y (1987) Dynamic viscoelastic prop- erties of ABS polymers in the molten state. 5. Eect of grafting degree.
Macromolecules 20:2208±2213 Balazs AC, Singh C, Zhulina E (1998)
Modeling the interactions between polymers and clay surfaces through self-consistent ®eld theory. Macromol- ecules 31:8370±8381
Balberg I, Binenbaum N, Wagner N (1984) Percolations thresholds in the three- dimensional sticks system. Phys Rev Lett 52:1465±1468
Baradollet P, Bousmina M, Muller R (1995) Relationship between structure and rheological properties in the melt of polymers containing spherical inclu- sions. Poly Adv Tech 6:301±308
Boissonade J, Barreau F, Carmona F (1983) The percolation of ®bres with random orientations: a Monte Carlo sturdy. J Phys A: Math Gen 16:2777±
Chen ZR, Issaian AM, Korn®eld JA, Smith2787 SD, Grothaus JT, Satkowsk MM (1997) Dynamics of shear-induced alignment of a lamellar diblock: a
rheo-optical, electron microscopy, and X-ray scattering study. Macromole- cules 30:7096±7114
Gandhi K, Salovey R (1988) Dynamic mechanical behavior of polymers con- taining carbon black. Poly Eng Sci 28:877±887
Kannan RM, Korn®eld JA (1994) Stress- optical manifestations of molecular and microstructural dynamics in com- plex polymer melts. J Rheol 38:1127±
Kawasaki K, Onuki A (1990) Dynamics1150 and rheology of diblock copolymers quenched into micro-phase separated states. Phys Rev A 42:3664±3666 Kojima Y, Uzuki A, Kawasumi M, Okada
A, Kurauchi T, Kamigaito O (1993) Synthesis of nylon-6/clay hybrid by montmorillonite intercalated with e- caprolactam. J Polym Sci Part A:
Polym Chem 31:983±986
Krishnamoorti R, Vaia RA, Giannelis EP (1996) Structure and dynamics of poly- mer-layered silicate nanocomposites.
Chem Mater 8:1728±1734
Krishnamoorti R, Giannelis EP (1997) Rheology of end-tethered polymer lay- ered silicate nanocomposites. Macro- molecules 30:4097±4102
Kyu T, Zhu GC, Zhou ZL, Tajuddin Y, Qutubuddin S (1996) Novel ®lled poly- mer composites prepared from in- situ polymerization via a colloidal ap- proach. ii. Blends of kaolin/nylon-6 in-situ composites with conventional
nylon-6 and nylon-66. J Polym Sci Part B: Polym Phys 34:1769±1775
Lakdawala K, Salovey R (1987) Rheology of copolymers containing carbon black. Poly Eng Sci 27:1043±1049 Larson RG, Winey KI, Patel SS, Watanabe
H, Bruinsma R (1993) The rheology of layered liquids-lamellar block copoly- mers and smectic liquid crystals. Rheol Acta 32:245±253
Lim YT, Park OO (2000) Rheological evidence for microstructure of interca- lated polymer/layered silicate nano- composites. Macro Rapid Comm 21:231±235
Messersmith PB, Giannelis EP (1995) Synthesis and barrier properties of poly(e-caprolactone)/layered silicate nanocomposites. J Polym Sci Part A:
Polym Chem 33:1047±1057
Ogawa M (1994) Formation of novel oriented transparent ®lms of layered silica-surfactant nanocomposites. J Am Chem Soc 116:7941±7942
Philipse AP (1996) The random contact equation and its implications for col- loidal rods in packing, suspensions, and anisotropic powders. Langmuir 12:1127±1133
Poslinski AJ, Ryan ME, Gupta RK, Sesh- adri SG, Frechette FJ (1988) Rheolog- ical behavior of ®lled polymeric systems. I. Yield stress and shear- thinning eects. J Rheol 32:703±735 Rosedale JH, Bates FS (1990) Rheology of
ordered and disordered symmetric
poly(ethylenepropylene)-poly(ethyleth- ylene) diblock copolymers. Macromol- ecules 23:2329±2338
Tepe TD, Hajduk A, Hillmyer MA, Wei- mann PA, Tirrell M (1997) In¯uence of shear on a lamellar triblock copolymer near order-disorder transition. J Rheol 41:1147±1171
Vaia RA, Ishii H, Giannelis EP (1993) Synthesis and properties of two-dimen- sional nanostructures by direct melt intercalation of polymer melts in lay- ered silicates. Chem Mater 5:1694±1696 Vaia RA, Jandt KD, Kramer EJ, Giannelis EP (1995) Kinetics of polymer melt in- tercalation. Macromolecules 28:8080±
Vaia RA, Giannelis EP (1997) Polymer8085 melt intercalation in organically-mod- i®ed layered silicates: model predictions and experiment. Macromolecules 30:8000±8009
Yano K, Uzuki A, Okada A, Kurauchi T, Kamigaito O (1993) Synthesis and properties of polyimide-clay hybrid.
J Polym Sci Part A: Polym Chem 31:2493±2498
Zhulina E, Singh C, Balazs AC (1999) Attraction between surfaces in a poly- mer melt containing telechelic chains:
guidelines for controlling the surface separation in intercalated polymer-clay composites. Langmuir 15:3935±3943