Rheological, morphological and structural properties of ... - UBO

Rheological, morphological and structural properties of PEÕPAÕnanoclay ternary blends: Effect of clay weight fraction Jacques Huitric,a) Julien Ville, Pascal Médéric, Michel Moan, and Thierry Aubry LIMATB Equipe Rhéologie, Université Européenne de Bretagne (UBO), 6 Avenue Victor Le Gorgeu, CS 93837 29238 Brest Cedex 3, France (Received 12 December 2008; final revision received 19 May 2009兲

Synopsis The effect of an organically modified layered silicate on the rheological, morphological, and structural properties of immiscible polyethylene/polyamide 共PE/PA兲 blends was investigated. The blends have been prepared for PA weight fractions ranging from 10 to 90% and at clay weight fractions from 1 to 6%. Scanning electron microscopy and transmission electron microscopy have been used to study the morphology and the structure of the blends. The dispersed phase size was shown to decrease with increasing clay content up to 2% and tends to stabilize at higher fractions. For PE matrix blends, clay particles were shown to be essentially located at the interface of the two polymers, forming an interphase whose thickness grows with clay fraction. For PA matrix blends with 2% of clay, the interphase thickness is stabilized at 11 nm; further clay addition leads to dispersion of clay within PA. Oscillatory and steady shear measurements have shown that PE matrix ternary blends behaved like polymer blends and underlined the contribution of an interphase at high clay fractions. For sufficiently filled PA matrix blends, a yield behavior was observed. The behavior of PA matrix ternary blends, dominated by the organoclay dispersed in PA, is similar to that of nanocomposites. © 2009 The Society of Rheology. 关DOI: 10.1122/1.3153551兴

I. INTRODUCTION Mixing a polymer with another one is an attractive way to develop new materials. The final properties of the blend are strongly influenced by the droplet size and droplet size distribution resulting from the balance between the effects of break-up and coalescence of droplets induced by the flow during processing. A compatibilizer is usually added in order to promote a fine dispersion, to stabilize the morphology by suppressing coalescence, and to enhance interfacial adhesion 关Paul and Newman 共1976兲; Xanthos 共1988兲; Utracki 共1991兲兴. The compatibilizing molecules are either diblock copolymers, which are miscible with both phases 关Utracki 共1991兲兴, or block or graft copolymers formed by reaction at the interface of the two phases during mixing 关Moan et al. 共2000兲兴. The copolymer, which is active at the surface of the droplets, lowers the interfacial tension 关Paul and Newman 共1976兲; Wu 共1987兲兴 and hence, facilitates break-up but, above all, inhibits coalescence significantly 关Macosko et al. 共1996兲; Lepers and Favis 共1999兲; Hu et al. 共2000兲; Wang et al. 共2004兲; Van Hemelrijck et al. 共2004, 2005兲; Huitric et al. 共2007兲兴. a兲

Author to whom correspondence should be addressed; electronic mail: [email protected]

© 2009 by The Society of Rheology, Inc. J. Rheol. 53共5兲, 1101-1119 September/October 共2009兲

0148-6055/2009/53共5兲/1101/19/$27.00

1101

1102

HUITRIC et al.

The main result of these studies is that the size reduction of the dispersed phase is essentially due to the suppression of the coalescence. This effect is the result of a steric repulsion between the block copolymers present at the interface and the Marangoni effect induced by the local concentration gradient of the compatibilizer at the interface. In the last decade, particular attention was paid to polymer/layered silicate nanocomposites. Recent developments in synthesis, processing, and characterization of these nanomaterials were thoroughly reviewed by Sinha Ray and Okamoto 共2003兲. Now, it is well acknowledged that the addition of low volume fractions of anisotropic organoclay nanoparticles with a large aspect ratio in thermoplastic matrix leads to the improvement of flame retardancy 关Gilman 共1999兲; Zanetti et al. 共2001兲; Tang et al. 共2003兲兴, barrier properties 关Massersmith and Giannelis 共1995兲; Ke and Yongping 共2005兲; Krook et al. 共2004兲兴, and mechanical properties 关Alexandre and Dubois 共2000兲; Reichert et al. 共1998兲; Fornes and Paul 共2004兲兴, so highlighting the potential commercial interest of these nanostructured materials. A polyamide 共PA兲 matrix filled with organically modified montmorillonite 共OMMT兲 nanoparticles is certainly the most commonly studied nanocomposite among all polymer/layered silicate systems 关Reichert et al. 共1998兲; Usuki et al. 共1993兲; Kojima et al. 共1994兲; Fornes and Paul 共2004兲; Aubry et al. 共2005兲; Médéric et al. 共2006兲兴. More recently, in the early 2000s, ternary blends composed of nanoclays dispersed in polymer blends have attracted particular attention because of possible synergetic effects between the nanoparticle network and the blend morphology 关Steinmann et al. 共2002兲兴. The study of such ternary blends requires the use of many investigation techniques. Morphological and structural characterization techniques, such as transmission electron microscopy 共TEM兲 and scanning electron microscopy 共SEM兲, are combined with either mechanical characterization techniques in the solid state 关Chen et al. 共2005兲; Mehrabzadeh and Kamal 共2002兲; Sinha Ray et al. 共2006兲兴 or, less frequently, with rheological experiments in the melt state 关Gahleiter et al. 共2006兲; Hong et al. 共2007兲兴. Moreover, to characterize the structural state of organoclay in the polymer blends, x-ray diffraction experiments are commonly used. Most published studies on ternary blends composed of nanoclay particles dispersed in two immiscible polymer blends have focused on the compatibilizing and reinforcing effects of organoclay presenting selective affinity toward one polymer of the blend. Wang et al. 共2001兲 reported that montmorillonite particles organically modified with octadecylammonium had a good affinity toward polar polymers and acted as a compatibilizer in a blend composed of 90 % wt PA-6 and 10 % wt polypropylene 共PP兲, reducing the dispersed phase size and enhancing the tensile modulus and strength significantly. Gelfer et al. 共2003兲 showed that in polystyrene 共PS兲 and poly共methyl methacrylate兲 blends, OMMT acted as inert fillers and that the reduction in PS domain size was due to the compatibilizing function of the excessive surfactant present in the organoclay and to the increase in the matrix viscosity. Mehrabzadeh and Kamal et al. 共2002兲 showed that adding 5 % wt of organoclay in PA-6/polyethylene 共PE兲 blends increased the interfacial adhesion between the phases and suppressed coalescence and agglomeration of PA nodules. Khatua et al. 共2004兲 reported that, in the case of a Nylon-6 共N6兲/poly共ethylene-ran-propylene兲 rubber 共EPR兲 blends, a sharp decrease in the EPR dispersed phase size was observed with only 1 % wt clay followed by a slow and gradual decrease in the average number diameter of EPR nodules for larger clay fractions. In this work, the TEM observations have shown that the clay was not present at the interface between the two polymers and the mechanical characterization has proved that the addition of organoclay did not enhance interfacial adhesion. So the authors concluded that exfoliated clay particles in the N6 matrix prevented the EPR dispersed phase from coalescence and that the nodule size reduction could not be explained by a decrease in the interfacial tension nor by a steric repulsion mediated by clay at the interface. Clearly, they

PE/PA/NANOCLAY TERNARY BLENDS

1103

TABLE I. Characteristics of the pure components. Mw 共g/mol兲

Mn 共g/mol兲

␩ⴱ0

Material

共Pa s兲a

Melting point 共°C兲

PE PA

140 000 37 000

37 000 20 000

10 750 2 000

121 183

a

Measured at 200 ° C.

have not observed any compatibilizing effect. Nevertheless, Hong et al. 共2007兲 showed, using TEM analysis and extensional force experiments, that the organoclay present at the polybutylene terephthalate 共PBT兲/PE interface could act as a compatibilizer by lowering interfacial tension between the two polymers. They measured a reduction in the interfacial tension, from 0.576 to 0.014 mN/m, when adding 1 % wt clay. When organoclay has a good affinity toward the two polymers of the blend, it acts like a coupling agent at the interface and reduces the interfacial tension between the two polymers 关Yoo et al. 共2005兲; Fang et al. 共2007兲兴, thus playing the role of a compatibilizer. As discussed above, different interpretations have been proposed to explain the effect of organoclay in ternary blends. Still, the compatibilizing effect is questionable. Does nanoclay addition induce a reduction in interfacial tension and/or a suppression of coalescence? The aim of the present work is to investigate the effects of a commercial OMMT, referenced as Cloisite® C30B, on the structural, morphological, interfacial, and rheological properties of PE/PA-12 blends, over the whole range of PA fractions, and for clay weight fractions up to 6%, with particular attention to the samples exhibiting a nodular morphology. II. EXPERIMENTAL A. Materials The blends studied in the present work were composed of a linear low-density PE and a PA-12, fully immiscible, supplied by Enichem 共Flexirene® FG 20F兲 and Arkema 共Rilsan® AECHVO兲, respectively. The viscosity ratio of these blends, i.e., the ratio of the dispersed phase viscosity to the matrix viscosity was 0.19 when the dispersed phase is PA, and 5.26 when the dispersed phase is PE. Table I shows the main characteristics of the two components: the melting point, the number, and weight average molecular weight denoted by Tm, M n and M w, respectively. The layered silicate added to these polymer blends was OMMT clay, supplied by Southern Clay Products 共Cloisite® C30B兲. This organoclay is a methyl tallow bis-2-hydroxyethyl ammonium exchanged montmorillonite clay with a modifier concentration of 90 meq per 100 g and has a good affinity toward PA 关Aubry et al. 共2005兲兴, but a very poor affinity toward PE 关Médéric et al. 共2005兲兴. The specific gravity of this organophilic clay is very close to 2. The characteristic dimensions of an individual silicate platelet are width and length ⬃500 nm and thickness ⬃0.7 nm, corresponding to an average aspect ratio of about 350. B. Blending The blends have been prepared at PA weight fractions ␾PA, ranging from 10 to 90%, by mixing simultaneously all the components in a Haake mixer 共Rheocord E U 5兲. The weight fractions of added C30B varied from 1 to 6% relative to PA. The temperature imposed during mixing was 200 ° C chosen to minimize the degradation of the components, especially that of the organic modifier 关Aubry et al. 共2005兲兴. The blending condi-

1104

HUITRIC et al.

tions were the same for all blends: temperature 200 ° C, residence time 12 min, and blade rotational speed 32 rpm, corresponding to a shear rate ⬃17 s−1, estimated from the model by Bousmina et al. 共1999兲. Then, samples were pelletized and processed by compression molding at 200 ° C to get 2 mm thick sheets. Owing to the hygroscopic nature of PA, all blends were dried for 4 h at 85 ° C in a vacuum oven before experiments. C. Morphological characterization SEM was used in order to investigate the dispersed phase morphology: nodular, fibrillar, or cocontinuous morphology. SEM observations were made from cryofractured samples, whose surface was vacuum-metallized and examined with a Hitachi S-3200N scanning electron microscope, with an accelerating voltage of 15 kV. In the case of nodular morphology, the average number diameter Dn and average volume diameter Dv were obtained from the SEM micrographs by measuring at least 350 particles with SigmaScan® PRO 5.0 image analysis software. The particle size polydispersity index I was defined as the ratio I = Dv / Dn. The localization, dispersion, and exfoliation degree of organoclay particles were determined by TEM. Ultrathin sections were prepared at −100 ° C with an ultracryomicrotome 共Reichert & Jung兲 using a diamond knife. Imaging was performed on a JEOL JEM 1230 at 80 kV. D. Rheological measurements The rheological characterization was performed using oscillatory and steady shear tests, with a Rheometrics Dynamic Analyzer 共ROA II兲 rheometer, equipped with a parallel plate geometry: diameter of 25 mm and gap of 2 mm. All experiments were carried out at a temperature of 200 ° C, under a continuous purge of dry nitrogen, in order to avoid sample degradation and absorption of moisture. The oscillatory shear measurements were performed at a fixed strain of 0.04, which is smaller than the limit of the linear viscoelastic regime for all systems studied. The material stability at 200 ° C, of all systems studied in this work, was systematically investigated by performing time sweep experiments. No significant variation in the rheometrical data was observed and experiments were shown to be reproducible within ⫾ 5%. III. RESULTS AND DISCUSSION A. Morphology In a blend, the morphology of the dispersed phase results from of a delicate balance between shear forces, which tend to deform the droplets, and interfacial tension forces, which tend to resist to the deformation, so that the particle shape and size result from the continuous competition between break-up and coalescence of the dispersed phase during mixing 关Grace 共1981兲兴. Moreover, the elasticity of the two components may play an important role 关Starita 共1972兲兴. As the aim of the present work is to show the influence of the nanoclay addition on the morphology and rheology of PE/PA blends, it is necessary to first characterize the morphology of the blends without clay over the whole range of PA weight fractions. Morphological observations of these blends are reported in Table II. More precisely, the blend morphology depends on PA content, leading to the determination of three different regions. A nodular morphology is observed for PA weight fractions up to 30% 关Fig. 1共a兲, region 1兴. Region 2 is characterized by an ill-defined morphology for which, nodular, elongated droplets, and fibrillar morphology coexist, as illustrated in Fig. 1共b兲.


1105

TABLE II. Morphological description of the blends.

␾ PA 10, 20, 30 40, 50, 60 70, 80, 90

Morphology description Nodular 共region 1兲 Fibrillar or very stretched 共region 2兲 Nodular 共region 3兲

Beyond a PA12 weight fraction of 70%, PE nodules are observed 关Fig. 1共c兲, region 3兴. In the two nodular regions, it must be noticed that the dispersed phase size grows with the fraction of the dispersed phase. The existence of such three regions is well documented and can be attributed to an enhancement of coalescence, as the dispersed phase fraction increases 关Favis and Chalifoux 共1988兲兴. In the rest of this study, we focused exclusively on blends having a nodular morphology, i.e., within regions 1 and 3, where morphology is easier to characterize qualitatively and quantitatively than in region 2. Figure 2共a兲 shows the number average diameter of the dispersed phase Dn as a function of PA weight fraction for PE/PA blends and ternary blends containing different amounts of C30B. In both regions 1 and 3, Dn increases with the dispersed phase weight fraction for PE/PA blends. At the same minor phase weight fraction, Dn is larger for PE nodules in a PA matrix than for PA nodules in a PE matrix. This result is expected since PE viscosity is higher than PA viscosity. Moreover, PE and PA elastic properties were

FIG. 1. 共a兲 SEM micrograph of 90/10 PE/PA blend. 共b兲 SEM micrograph of 50/50 PE/PA blend. 共c兲 SEM micrograph of 10/90 PE/PA blend.

1106

HUITRIC et al.

FIG. 2. 共a兲 Number average diameter as a function of PA weight fraction, for different C30B contents: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, and 共䊊兲 4%.共b兲 SEM micrograph of 70/30 PE/PA with 2% C30B.

measured in previous studies, respectively, by Huitric et al. 共1998兲 and Médéric et al. 共2006兲 and it was found that PE elasticity is 20 times higher than PA elasticity. So PE droplets are more difficult to deform and to break than PA droplets, as suggested by Starita’s analysis 关Starita 共1972兲兴. Figure 2共a兲 also shows that whatever the amount of C30B added, at a fixed PA weight fraction, a strong decrease of Dn is observed in the two regions. However, the effect of adding C30B is not the same in both regions. Indeed, in region 1 for 70/30 PE/PA blend, the addition of 2% clay induces a change in morphology from nodules to very elongated PA droplets, close to fibrillar morphology, as illustrated in Fig. 2共b兲, favored by the possible presence of anisotropic clay platelets in the PA dispersed phase. A similar observation was made by Mehrabzadeh and Kamal 共2002兲兴 in the case of an 80/20 high density polyethylene 共HPDE兲/PA-6 blend, for which adding 5% of clay induced both a


1107

FIG. 3. Number average diameter as a function of C30B weight fraction, for 共䉱兲 80/20 and 共쎲兲 20/80 PE/PA blends.

lower size of the dispersed phase and a change of droplet shape, from spherical to lamellar type. Whereas in region 3, for a 30/70 PE/PA blend, the droplets keep a nodular shape at all C30B fractions. The diminution of dispersed phase size when adding C30B is also clearly illustrated in Fig. 3: Dn decreases with increasing C30B content for a 20/80 PE/PA blend and an 80/20 PE/PA blend. Let us recall that the amount of C30B added is relative to the PA content in the blend, meaning that for 2% added C30B, the 20/80 PE/PA blend contains 1.6% organoclay, whereas the 80/20 PE/PA blend only contains only 0.4% clay. Figure 3 shows that the decrease in the dispersed phase size is very marked, up to a nanoclay content close to 2% for the two blends, especially for the PA matrix blend. Beyond 2% added C30B, the size of the dispersed phase exhibits a quasi-plateau or at least a relatively smooth decrease in the droplet size. An important qualitative observation is that adding C30B to 20/80 and 80/20 PE/PA blends keeps the nature of the morphology unchanged, that is, nodular dispersion of the minor phase. The polydispersity index values are summarized in Table III. For PE/PA blends, we can observe that the PA droplet polydispersity index is higher than that of PE droplets, which are nearly monodispersed. Moreover, Table III shows that the addition of nanoclay leads to a narrowing of the size distribution in the case of PA nodules, whereas it keeps the size distribution nearly unchanged for PE nodules. Figure 4 shows TEM micrographs of 80/20 PE/PA blends with 1%, 2%, and 4% added C30B. The PE phase is easily identifiable due to the presence of crystallites, which make it slightly noisy on the micrographs. We can observe that clay is present at the polymer interface and tends to create an interphase. For the blend containing 1% C30B 关Fig. 4共a兲兴, the interphase is discontinuous with a non-homogeneous clay layer whose average thickness is close to 7 nm. For 2% added C30B 关Fig. 4共b兲兴, the interphase becomes quasicontinuous and its thickness has a mean value of 14 nm. It is necessary to reach a value of 4% of organoclay 关Fig. 4共c兲兴 to observe a continuous 21 nm thick interphase combined with the presence of very few C30B entities in the PA dispersed phase. A similar result was obtained by Hong et al. 共2006兲 in the case of a 90/10 PE/PBT blend for which it was

1108

HUITRIC et al.

TABLE III. Weight fraction of C30B located at the interphase, and in PA phase 共nodules or matrix兲 and specific particle density for 共a兲 80/20 and 共b兲 20/80, PE/PA blends.

C30B%

Dn共␮m兲

e 共nm兲

nclay

␾int

0 1 2 4

3.08 2.24 1.43 1.33

7 14 21

3 6 8

1 2 ⬃4

C30B%

Dn共␮m兲

e 共nm兲

nclay

␾int

0 1 2 4 5 6

3.76 2.77 1.50 1.34 1.30 1.22

9 14 13 12 11

4 6 5 5 5

0.30 0.84 0.85 0.82 0.82

共a兲

共b兲

␾mat

PA

0.70 1.16 3.15 4.18 5.18

␾nod

PA

I

A few platelets

1.17 1.05 1.02 1.03

Dsp

I

20 10 9 5 4

1.01 1.02 1.01 1.01 1.03 1.01

FIG. 4. 共a兲 TEM micrograph of 80/20 PE/PA blend with 1% C30B. 共b兲 TEM micrograph of 80/20 PE/PA blend with 2% C30B. 共c兲 TEM micrograph of 80/20 PE/PA blend with 4% C30B.


1109

FIG. 5. 共a兲 TEM micrograph of 20/80 PE/PA blend with 1% C30B. TEM micrograph of 20/80 PE/PA blend with 2% C30B. 共c兲 TEM micrograph of 20/80 PE/PA blend with 4% C30B. 共d兲 TEM micrograph of 20/80 PE/PA blend with 5% C30B. 共e兲 TEM micrograph of 20/80 PE/PA blend with 6% C30B.

necessary to reach an organoclay mass fraction of 5% to observe the presence of clay entities in PBT nodules. Finally, the morphology of 80/20 PE/PA blends filled with nanoclay particles looks quite similar to that of 80/20 PE/PA blends, except that the nodules are covered with a continuous shell composed of stacked clay particles as shown in Fig. 6共a兲. Figure 5 shows the TEM micrographs for 20/80 PE/PA blends with C30B up to 6%. Once 1% organoclay is added to the blend, a continuous interphase with a relatively

1110

HUITRIC et al.

FIG. 6. 共a兲 Schematic description of the interphase between PE matrix and PA nodule. 共b兲 Schematic description of the interphase between PA matrix and PE nodule.

regular thickness combined with the presence of few exfoliated clay layers and stacks in the PA matrix is observed, without any clay particles in the PE nodules. Moreover, for further organoclay addition we observe an increase in the interphase thickness, which stabilizes at ⬃2% C30B, while the amount of clay present within the PA matrix increases with C30B weight fraction. So, in this case, the morphology of 20/80 PE/PA blends filled with C30B is characterized by PE nodules covered with a shell composed of organoclay particles intercalated with PA chains dispersed in a PA/clay nanocomposite matrix. The structure is schematically presented in Fig. 6共b兲. From the measured average nodule diameter Dn, we have calculated the average nodule area and the number of individual silicate layers necessary to cover the whole nodule surface without overlapping. Besides, the average thickness e of the interphase between the two polymers can be estimated from TEM observations, taking into account


1111

the possible variation of thickness of this interphase. In a previous study 关Médéric et al. 共2006兲兴, for intercalated nanocomposites composed of the same PA and C30B used in the present work and using wide-angle x-ray diffraction, we measured an inter-reticular distance dir of 3 nm. So, knowing the thickness of a clay particle, that is, 0.7 nm, we can determine the number of stacked silicate layers nclay present in the interphase and the amount of clay dispersed in the PA nodules 共region 1兲 or within the PA matrix 共region 3兲. Experimental values of Dn and e, and calculated values of nclay and C30B weight fraction either at the interphase, named ␾int, or in PA nodules denoted as ␾nod PA 共region 1兲, or in PA matrix denoted as ␾mat PA, are reported in Table III. These data clearly show that for blends with a PE matrix, nearly all C30B particles are localized at the interface matrixnodules and that the intercalated clay interphase thickness grows with C30B content. On the other hand, for blends with a PA matrix, only a fraction of added Cloisite® is present at the interphase. Moreover, the measurements show that once 2% C30B is added, the interphase thickness stabilizes at ⬃13 nm, corresponding to a stack of ⬃4 PA intercalated clay layers. Anyway for both systems, the results show that a continuous interphase forms prior to the presence of C30B in the PA phase; even though, in the case of PA nodules, only very few organoclay entities were observed at 4% total clay fraction. Starting from the above observations and results, we suggest the following mechanism of size reduction. In the case of blends with a PE matrix, for which C30B is localized only at the interphase, we suggest that the size reduction could result from coalescence inhibition due to the significant rigidity of the intercalated PA/clay nanocomposite layer covering PA nodules. Indeed, in flow-driven coalescence encountered during melt blending, collisions between nodules occur and interphases act as solid-like barriers which repulse each other. Moreover, the Marangoni forces, which could exist when the repartition of the organoclay at the interface is not homogeneous, are expected to be negligible because of the high aspect ratio of C30B particles, inducing a very low mobility of the organoclay. At last we should point out that a lowering of the interfacial tension, due to the presence of the clay at the interface, cannot be excluded. For blends with PE nodules, where C30B particles are present in a continuous interphase and possibly in the PA matrix, the dispersed phase size reduction mechanism is expected to be the same, but even more reinforced by the enhancement of break-up mechanism and coalescence barrier effect, due to the dispersion of clay in the PA phase. Indeed, the presence of C30B in the PA matrix gives rise both to an increase in the matrix viscosity which favors the break-up of PE nodules and inhibition of coalescence due to the anisotropic silicate layers 关Sinha Ray et al. 共2006兲兴. The dispersion of the organoclay in the PA matrix can be estimated by the specific particle density. Indeed, from the TEM micrographs, the determination of the specific particle density Dsp, that is, the average number particle per ␮m2 divided by the clay mass fraction 关Fornes et al. 共2001兲兴 allows us to estimate the degree of exfoliation of the organoclay in the PA matrix. Table III共b兲 shows the exfoliation degree decrease with increasing clay mass fraction in the PA matrix and that Dsp decreases from 20 to 4 when the mass fraction of C30B in the matrix increases from 0.7% to 5.2%. A similar result was already reported by Médéric et al. 共2006兲 in the case of a PA/C30B nanocomposite. B. Rheology 1. Oscillatory shear rheology

In order to determine the extent of the linear viscoelastic regime, strain sweep experiments were realized at a fixed frequency of 1 Hz. This test was carried out for all the blends with a nodular morphology. The so-called critical strain ␥c, defining the limit of

1112

HUITRIC et al.

FIG. 7. Critical strain as a function of C30B weight fraction for 共䉱兲 80/20 and 共쎲兲 20/80 blends.

the linear viscoelastic regime, is plotted in Fig. 7 for 20/80 and 80/20 PE/PA blends. In this figure, for the blend with the PE matrix, the curve of ␥c versus C30B weight fraction exhibits two regions. At low C30B weight fraction, a significant decrease in the critical strain is observed, which is inversely proportional to the C30B fraction. Above 2% C30B, the critical strain seems to be relatively independent of the clay fraction, even though we have only two data points. It must be reminded that 2% C30B is the fraction needed to cover the nodules with a quasi-continuous interphase. For 20/80 PE/PA blends, Fig. 7 also displays two well separate regions. The first one is quite similar to that described for the blends with the PE matrix. In the second region, that is, above 2% C30B, ␥c goes on decreasing with increasing the clay, still with a weaker dependence on the clay fraction. TEM micrographs have shown that 2% C30B was necessary to obtain an interphase with a regular thickness and that further clay addition led to an increase in C30B content in the PA matrix, which could be responsible for the decrease of ␥c above 2% solid fraction 关Aubry et al. 共2005兲兴. The effect of clay addition on the frequency dependence of the complex viscosity is shown in Figs. 8 and 9, for the 80/20 and 20/80 PE/PA blends, respectively. The behavior of the blends filled with clay differs from that observed for the blends without clay. Indeed, a higher viscosity is obtained at the lowest frequencies and the viscosity level increases with increasing the C30B fraction 共Fig. 8兲. Nevertheless, in the case of blends with the PE matrix, it was possible to determine a Newtonian complex viscosity for all C30B weight fractions. The complex viscosity curve, at 4 % wt of clay, clearly displays two plateaus, as illustrated in Fig. 8, which could be due to the relaxation of the viscoelastic interphase. On the other hand, for blends with the PA matrix, the viscosity curves look similar to those for blends with the PE matrix for weak C30B fractions. At 2% of organoclay, the existence of two plateau regions in the complex viscosity curve is also observed 共Fig. 9兲. The low frequency plateau could still be attributed to the viscoelastic contribution of the interphase since the clay fraction in the matrix 关Table III共b兲兴 is below the percolation threshold, close to 3 % wt 关Aubry et al. 共2005兲兴. But, above 2% clay, a major difference between the two systems is evidenced: it was not possible to determine a Newtonian complex viscosity. Moreover, for blends with 5% and 6% C30B, the slope of the complex viscosity curve, close to ⫺1 at low frequencies, is indicative of


1113

FIG. 8. 80/20 PE/PA blend complex viscosity as a function of frequency for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, and 共䊊兲 4%.

a yield behavior which is related to the large presence of clay entities in the PA matrix, as shown by TEM micrographs 共Fig. 5兲. Similar results were obtained for PA-C30B nanocomposite by Aubry et al. 共2005兲. The storage moduli G⬘ as a function of frequency in the linear viscoelastic regime are shown in Figs. 10 and 11, for the 80/20 and 20/80 PE/PA blends, respectively. As in the case of viscosity curves, results are qualitatively the same for the two systems, for low C30B weight fractions, but differ completely above 2% clay content. In the case of the PE matrix, Fig. 10 displays a behavior similar to that of immiscible polymer blends, with a nodular morphology for all clay fractions added. At high frequencies, there is no significant effect of C30B content. All G⬘ curves are close to that of the PE matrix, most likely due to the absence of organoclay in the PE matrix. More interestingly, a shoulder in the G⬘ curves clearly appears in the presence as well as in the absence of added

FIG. 9. 20/80 PE/PA blend complex viscosity as a function of frequency for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, 共䊊兲 4%, 共䉱兲 5%, and 共䉭兲 6%.

1114

HUITRIC et al.

FIG. 10. 80/20 PE/PA blend elastic modulus as a function of frequency for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, and 共䊊兲 4%.

organoclay, which might be attributed to the form relaxation of the nodules. The characteristic time ␭d of this relaxation mechanism is very slightly shifted toward higher values in comparison with the blend without clay. The analytical expression of ␭d derived from Palierne’s model 关Palierne 共1990兲兴 and proposed by Graebling et al. 共1994兲 can be used to determine an apparent interfacial tension ␣. The results show that the value of ␣, close to 10 mN m−1 for PE/PA blends, drops to 4.5 mN m−1 for 1% added C30B and tends to a value of 3 mN m−1 for 2% and 4% clay. This suggests that the size reduction in the dispersed phase could be due, at least partly, to the decrease in the apparent interfacial tension. However, this suggestion must be cautiously examined since the simplified version of Palierne’s model used was developed for “uncompatibilized” blends, and the nature of an apparent interfacial tension is somewhat questionable in the presence of a viscoelastic interphase.

FIG. 11. 20/80 PE/PA blend elastic modulus as a function of frequency for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, 共䊊兲 4%, 共䉱兲 5%, and 共䉭兲 6%.


1115

FIG. 12. 80/20 PE/PA blend flow curve for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, and 共䊊兲 4%.

For blends with a PA matrix, the storage modulus versus frequency curves 共Fig. 11兲 show that for blends with clay weight fractions lower than 2%, the behavior of the ternary blends is similar to that observed for blends with a PE matrix. In particular, a shoulder in the G⬘ curve is also observed. Whereas, for blends with a C30B fraction higher than 4%, G⬘ exhibits clearly a solid-like behavior. At low frequency, the elastic modulus is nearly independent of frequency. This behavior was observed for many layered silicate nanocomposites and was attributed to the formation of a percolated network 关Krishnamoorti and Yurekli 共2001兲; Médéric et al. 共2006兲兴. Blend with 4% clay displays an intermediate behavior characterized by a weak frequency dependence at low frequencies, still with no tendency to plateau. Moreover, contrary to blends with the PE matrix, an increase in storage modulus G⬘ with C30B content is observed at high frequencies. A similar behavior was observed by Lim and Park 共2001兲 for polymer/layered silicate nanocomposites and by Aubry et al. 共2005兲 for PA-12/C30B nanocomposites. This behavior can be explained by the presence of organoclay in the PA matrix for C30B contents higher than 1%, as revealed by TEM micrographs. Interestingly, when the clay weight fraction of clay present in the matrix is higher than 2%; the shoulder attributed to the form relaxation process of the nodules is hidden by the solid-like behavior of the PA/C30B nanocomposite matrix. 2. Steady shear rheology

Flow curves, at low shear rates, are displayed in Figs. 12 and 13 for 80/20 and 20/80 PE/PA blends with different clay contents. All blends with a PE matrix 共Fig. 12兲 exhibit a shear-thinning behavior very similar to that of the 80/20 PE/PA blend without added clay. This moderate shear-thinning behavior, characterized by a power-law index n = 0.85, is typical of immiscible polymer blends with a nodular morphology at relatively low shear rates that are below 1 s−1 关Huitric et al. 共1998兲兴. So we can consider that in the case of blends with a PE matrix, the shear viscosity is governed by the PA nodules whatever the clay content. However, the magnitude of this shear-thinning is independent of the amount of clay added. Therefore, we can conclude that at these low deformation rates, the shear viscosity is not influenced by the characteristics of the interphase 共thickness, number of stacked clay layers兲 discussed above. For blends with a PA matrix, Fig.

1116

HUITRIC et al.

FIG. 13. 20/80 PE/PA blend flow curve for C30B weight fractions: 共䊏兲 0%, 共䊐兲 1%, 共쎲兲 2%, 共䊊兲 4%, 共䉱兲 5%, and 共䉭兲 6%.

13 shows for the lowest clay content that is 1%, a similar weak shear-thinning, as observed for the blends with a PE matrix and, again, we can assume that the shear viscosity is mainly governed by the nodular PE dispersed phase. For the highest clay contents that are for 5% and 6%, an apparent yield stress is observed, which can be related to the nanocomposite structure of the matrix when a sufficient amount of clay is added, as revealed by the linear viscoelastic measurements. This type of behavior has been already reported for PA/C30B nanocomposites 关Aubry et al. 共2005兲兴. These authors have shown that a transition between shear-thinning behavior and yield behavior occurs at a C30B content higher than 3% and that this yield behavior is due to a mesoscopic structure composed of networked domains of correlated clay particles. In this study, we have shown by TEM analysis that the clay fraction present in the matrix is close to 4% when the total amount of clay added to the blend is 5%. This result suggests that the shear viscosity of these 20/80 PE/PA blends is fully governed by the PA/clay nanocomposite matrix and that the apparent yield stress is related to the existence of networked clay domains within the matrix. The 20/80 PE/PA blends with 2% and 4% added clay exhibit an intermediate behavior 共Fig. 13兲, in agreement with the linear viscoelastic behavior previously discussed. Indeed, these blends exhibit a shear-thinning behavior which is all the more marked as the clay content increases, with n = 0.7 and n = 0.6, for 2% and 4% clay, respectively. Such a behavior highlights the increasing contribution to the shear viscosity of the PA/clay nanocomposite matrix containing an increasing amount of clay, or alternatively a decreasing contribution of PE nodules. IV. CONCLUSION We have shown that the addition of an organoclay in PE/PA blends leads to a more refined morphology. Indeed, it was shown by SEM micrographs that the dispersed phase size decreases when the weight fraction of the added C30B increases. TEM micrographs have shown that the organoclay was previously localized at the interface of the two polymers and second was present in the PA phase for sufficient high weight fraction of added C30B. Moreover, we have shown that the addition of organoclay leads to a lower interfacial tension.


1117

We have proposed a mechanism of size reduction. In the case of the PA nodules, the C30B is localized at the interphase without modification of the bulk properties of the polymers; so we have suggested that the size reduction could result from the coalescence inhibition due to steric repulsions resulting from the compression of the intercalated PA chains of the interphase. On the other hand, for blends with PE nodules, it has been observed, in addition to the organoclay present at the interphase, that the C30B entities are in the the PA matrix. So, the dispersed phase size reduction mechanism is reinforced by the dispersion of clay in PA phase, which gives rise to an increase in the matrix viscosity which then favors the break-up of the PE nodules, and to coalescence inhibition. Rheological measurements have shown that in the case of the PE matrix, whatever the C30B weight fraction, a behavior similar to that of immiscible polymer blends in the terminal zone. At high C30B content, the contribution of the interphase was evidenced. For blends with a PA matrix with clay weight fractions lower than 2%, the behavior and the interpretation for the ternary blends are similar to those observed for PE matrix blends. Whereas for a weight fraction of C30B higher than 4%, the linear viscoelastic and steady shear properties show a solid-like behavior similar to that of sufficiently filled nanocomposites. Moreover, it was interesting to note that morphological analysis and rheological measurements are in good agreement. ACKNOWLEDGMENTS The authors are most grateful to Dr. Benoît Brulé 共Arkema, France兲 for fruitful discussion and to Dr. Dufaure Nicolas 共Arkema, France兲 for his precious help concerning TEM experiments. Finally, financial support from “Région Bretagne” is gratefully acknowledged.

References Alexandre, M., and P. Dubois, “Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials,” Mater. Sci. Eng. R. 28, 1–63 共2000兲. Aubry, T., T. Razafinimaro, and P. Médéric, “Rheological investigation of the melt state elastic and yield properties of a polyamide-12 layered silicate nanocomposite,” J. Rheol. 49, 425–440 共2005兲. Bousmina, M., A. Ait-Kadi, and J. B. Faisant, “Determination of shear rate and viscosity from batch mixer data,” J. Rheol. 43, 415–434 共1999兲. Chen, G. X., H. S. Kim, E. S. Kim, and J. S. Yoon, “Compatibilization-like effect of reactive organoclay on the poly共L-lactide兲/poly共butylene succinate兲 blends,” Polymer 46, 11829–11836 共2005兲. Fang, Z., Y. Xu, and L. Tong, “Effect of clay on the morphology of binary blends of polyamide 6 with high density polyethylene and HDPE-graft-acrylic acid,” Polym. Eng. Sci. 47, 551–559 共2007兲. Favis, B. D., and J. P. Chalifoux, “Influence of composition on the morphology of polypropylene/polycarbonate blends,” Polymer 29, 1761–1767 共1988兲. Fornes, T. D., and D. R. Paul, “Structure and properties of nanocomposites based on nylon-11 and -12 compared with those based on nylon-6,” Macromolecules 37, 7698–7709 共2004兲. Fornes, T. D., P. J. Yoon, H. Keskkula, and D. R. Paul, “Nylon 6 nanocomposites: The effect of matrix molecular weight,” Polymer 42, 09929–09940 共2001兲. Gahleitner, M., B. Kretzschmar, G. V. Vliet, J. Devaux, D. Pospiech, K. Bernreitner, and E. Ingolic, “Rheology/ morphology interactions in polypropylene/polyamide-6 nanocomposites,” Rheol. Acta 45, 322–330 共2006兲. Gelfer, M. Y., H. H. Song, L. Liu, B. S. Hsiao, B. Chu, M. Rafailovich, M. Si, and V. Zaitsev, “Effects of organoclays on morphology and thermal and rheological properties of polystyrene and poly共methyl methacrylate兲 blends,” J. Polym. Sci., Part B: Polym. Phys. 41, 44–54 共2003兲. Gilman, J. W., “Flammability and thermal stability studies of polymer-layered silicate 共clay兲 nanocomposites,”

1118

HUITRIC et al.

Appl. Clay Sci. 15, 31–49 共1999兲. Grace, H. P., “Dispersion phenomena in high viscosity immiscible fluid systems and application of static mixers as dispersion devices in such systems,” Chem. Eng. Commun. 14, 225–277 共1982兲. Graebling, D., A. Benkira, Y. Gallot, and R. Muller, “Dynamic viscoelastic behaviour of polymer blends in the melt—Experimental results for PDMS/POE-DO, PS/PMMA, and PS/PEMA,” Eur. Polym. J. 30, 301–308 共1994兲. Hong, J. S., J. K. Kim, Y. H. Ahn, S. J. Lee, and C. Kim, “Interfacial tension reduction in PBT/PE/clay nanocomposite,” Rheol. Acta 46, 469–478 共2007兲. Hong, J. S., H. Namkung, K. H. Ahn, S. J. Lee, and C. Kim, “The role of organically modified layered silicate in the breakup and coalescence of droplets in PBT/PE blends,” Polymer 47, 3967–3975 共2006兲. Hu, Y. T., D. J. Pine, and L. G. Leal, “Drop deformation, breakup and coalescence with compatibilizer,” Phys. Fluids 12, 484–489 共2000兲. Huitric, J., P. Médéric, M. Moan, and J. Jarrin, “Influence of composition on rheological properties of polyethylene/polyamide blends,” Polymer 39, 4849–4856 共1998兲. Huitric, J., M. Moan, P. J. Carreau, and N. Dufaure, “Effect of reactive compatibilization on droplet coalescence in shear flow,” J. Non-Newtonian Fluid Mech. 145, 139–149 共2007兲. Ke, Z., and B. Yongping, “Improve the gas barrier property of PET film with montmorillonite by in situ interlayer polymerization,” Mater. Lett. 59, 3348–3351 共2005兲. Khatua, B. B., J. L. Dong, Y. K. Hwang, and K. K. Jin, “Effect of organoclay platelets on morphology of nylon-6 and poly共ethylene-ran-propylene兲 rubber blends,” Macromolecules 37, 2454–2459 共2004兲. Kojima, Y., A. Usuki, M. Kawasumi, A. Okada, T. Kurauchi, O. Kamigaito, and K. Kaji, “Fine structure of nylon-6-clay hybrid,” J. Polym. Sci., Part B: Polym. Phys. 32, 625–630 共1994兲. Krishnamoorti, R., and K. Yurekli, “Rheology of polymer layered silicate nanocomposites,” Curr. Opin. Colloid Interface Sci. 6, 464–470 共2001兲. Krook, M., A. C. Albertsson, U. W. Gedde, and M. S. Hedenqvist, “Barrier and mechanical properties of montmorillonite/polyesteramide nanocomposite,” Polym. Eng. Sci. 42, 1238–1246 共2002兲. Lepers, J. C., and B. D. Favis, “Interfacial tension reduction and coalescence suppression in compatibilized blends,” AIChE J. 45, 887–895 共1999兲. Lim, Y. T., and O. O. Park, “Phase morphology and rheological behavior of polymer/layered silicate nanocomposites,” Rheol. Acta 40, 220–229 共2001兲. Macosko, C. W., P. Guegan, A. K. Khandpur, A. Nakayama, P. Marechal, and T. Inoue, “Compatibilizers for melt blending: Premade block copolymers,” Macromolecules 29, 5590–5598 共1996兲. Messersmith, P. B., and E. P. Giannelis, “Synthesis and barrier properties of poly共␧-caprolactone兲-layered silicate nanocomposites,” J. Polym. Sci. Part Polym. Chem. 33, 1047–1057 共1995兲. Médéric, P., T. Razafinimaro, T. Aubry, M. Moan, and M. H. Klopffer, “Rheological and structural investigation of layered silicate nanocomposites based on polyamide or polyethylene: Influence of processing conditions and volume fraction effects,” Macromol. Symp. 221, 75–84 共2005兲. Médéric, P., T. Razafinimaro, and T. Aubry, “Influence of melt-blending conditions on structural, rheological and interfacial properties of polyamide-12 layererd silicate nanocomposites,” Polym. Eng. Sci. 46, 986–994 共2006兲. Mehrabzadeh, M., and M. R. Kamal, “Polymer-clay nanocomposites based on blends of polyamide-6 and polyethylene,” Can. J. Chem. 80, 1083–1092 共2002兲. Moan, M., J. Huitric, P. Médéric, and J. Jarrin, “Rheological properties and reactive compatibilization of immiscible polymer blends,” J. Rheol. 44, 1227–1245 共2000兲. Palierne, J. F., “Linear rheology of viscoelastic emulsions with interfacial tension,” Rheol. Acta 29, 204–214 共1990兲. Paul, D. R., and S. Newman, Polymer Blends 共Academic, New York, 1976兲, Vols. 1 and 2. Reichert, P., J. Kressler, R. Thoann, R. Mülhaupt, and G. Stöppelmann, “Nanocomposites based on a synthetic layer silicate and polyamide-12,” Acta Polym. 49, 116–123 共1998兲. Sinha Ray, S., and M. Okamoto, “Polymer/layered silicate nanocomposites: A review from preparation to processing,” Prog. Polym. Sci. 28, 1539–1641 共2003兲. Ray, S., M. Bousmina, and A. Maazouz, “Morphology and properties of organoclay modified polycarbonate/


1119

poly共methyl methacrylate兲 blend,” Polym. Eng. Sci. 46, 1121–1129 共2006兲. Starita, J. M., “Microstructure of melt blended polymer systems,” Trans. Soc. Rheol. 16, 339–347 共1972兲. Steinmann, S., W. Gronski, and C. Friedrich, “Influence of selective filling on rheological properties and phase inversion of two-phase polymer blends,” Polymer 43, 4467–4477 共2002兲. Tang, Y., Y. Hu, S. Wang, Z. Gui, and Z. Chen, “Preparation of poly共propylene兲/clay layered nanocomposites by melt intercalation from pristine montmorillonite 共MMT兲,” Polym. Adv. Technol. 14, 733–757 共2003兲. Usuki, A., Y. Kojima, M. Kawasumi, A. Okada, Y. Fukushima, T. Kurauchi, and O. Kamigaito, “Synthesis of nylon-6-clay hybrid,” J. Mater. Res. 8, 1179–1183 共1993兲. Utracki, L. A., Two Phases Polymer Systems 共Hanser Verlag, Munich, 1991兲. Van Hemelrijck, E., P. Van Puyvelde, S. Velankar, C. W. Macosko, and P. Moldenaers, “Interfacial elasticity and coalescence suppression in compatibilized polymer blends,” J. Rheol. 48, 143–159 共2004兲. Van Hemelrijck, E., P. Van Puyvelde, C. W. Macosko, and P. Moldenaers, “The effect of block copolymer architecture on the coalescence and interfacial elasticity in compatibilized polymer blends,” J. Rheol. 49, 783–798 共2005兲. Wang, H., C. Zeng, M. Elkovitch, L. J. Lee, and K. W. Koelling, “Processing and properties of polymeric nano-composites,” Polym. Eng. Sci. 41, 2036–2046 共2001兲. Wang, S., Y. Hu, L. Song, J. Liu, Z. Chen, and W. Fan, “Study on the dynamic self-organization of montmorillonite in two phases,” J. Appl. Polym. Sci. 91, 1457–1462 共2004兲. Wu, S., “Formation of dispersed phase in incompatible polymer blends: Interfacial and rheological effects,” Polym. Eng. Sci. 27, 335–343 共1987兲. Xanthos, M., “Interfacial agents for multiphase polymer systems: Recent advances,” Polym. Eng. Sci. 28, 1392–1400 共1988兲. Yoo, Y., C. Park, S. D. Lee, K. Y. Choi, D. S. Kim, and J. H. Lee, “Influence of addition of organoclays on morphologies in nylon 6/LLDPE blends,” Macromol. Chem. Phys. 206, 878–884 共2005兲. Zanetti, M., G. Camino, R. Thomann, and R. Mulhaupt, “Synthesis and thermal behaviour of layered silicateEVA nanocomposites,” Polymer 42, 4501–4507 共2001兲.