Effect of selective localization of carbon nanotubes in PA6 dispersed phase of PP/PA6 blends on the morphology evolution with time, part 1: droplet deformation under simple shear flows.
Multicomponent polymer systems are mostly thermodynamically immiscible. Hence, blending of such materials typically forms heterogeneous structures such as sea-island or co-continuous morphologies. It is a well-known fact that the final properties of such hybrid systems are dependent on the morphological characteristics of the blend. Consequently, a considerable number of studies attempt to predict and control such a crucial factor during the design and process of immiscible polymer blends [1-7]. Indeed, the desired morphology can only be achieved provided that a thorough understanding of the involved phenomena is available. For instance, the unfavorable interfacial tension between the phases must be balanced against the external forces in order to reach a stable morphology. A classic method to improve the interfacial adhesion and enhance the mechanical performance of immiscible polymer blends is the use of compatibilizing agents [1, 8, 9]. Macaubas et al.  used a linear triblock copolymer (styrene-ethylene-butylene-styrene, SEBS) to compatibilize an immiscible polypropylene (PP)/polystyrene (PS) blend. They reported that the PS droplets encapsulated by SEBS are more resistant to shear deformations. With the same purpose, nanoparticles have been incorporated for many years. It is only recently that their role in controlling the morphology and behavior of immiscible polymer blends has become the motive for orientated theoretical and experimental investigations [11-16]. The available literature on the subject is mainly focused on the impacts of nanoparticles on the final morphology as well as the mechanical, rheological, and electrical properties of the blends [17-19]. As an example, Zou et al.  discussed that the morphology of poly(p-phenylene sulfide)/polyamide 66 (60/40 w/w) was changed by the addition of a small amount of acid-treated carbon nanotubes (CNTs). According to their results, the morphology was changed from sea-island to co-continuous structure. They found that at higher CNT contents it changed back to sea-island with an increased domain size. Such effects of nanoparticles on the morphology formation are shown to be largely dependent on the distribution of nanoparticles within the blends [21, 22]. An increased domain size is expected for the localization in the dispersed phase. On the contrary, for the localization of nanoparticles at the interface or in the matrix, a decreased domain size and a more stable morphology is usually the case. Recently. Baudouin et al.  incorporated CNTs to form a dense barrier at the interface to prevent the coalescence of the dispersed phase. They reported a stable morphology in which the distribution of droplet size was relatively narrow. Recently, it is reported that the spatial distribution and localization of solid particles in polymer blends are controlled by the thermodynamics and processing conditions as well as the aspect ratio of the particle itself .
The effect of solid inclusions on the structural evolutions during flow and after the cessation of the flow is an interesting subject. However, it has not received enough attention yet. The complexity of the phenomena involved during droplet deformation and relaxation could be the main reason for this shortage [25-41]. The main purpose of this two-part work is to investigate these topics. By utilizing computational simulations based on the lattice Boltzmann technique, Usta et al.  demonstrated that the deformation of a sheared droplet could be significantly reduced at intermediate capillary numbers upon the addition of spherical nanoparticles. This reduction was mainly ascribed to the resistance of the suspended particles to the shearing motions imposed by the flow. A similar approach was adopted by Skartlien et al.  to model the evolution of morphology of an emulsion in startup of shear flows in the presence of a surfactant. They attributed a transient viscoelastic contribution to the interfacial stresses under constant strain rates. Kong et al.  found that with an increase in silica nanoparticles content, the critical break aspect ratio of filled-polyamide 6 (PA6) fibrillar droplets during relaxation in PS matrix was improved. Despite all these attempts, the lack of certain experimental data as well as the necessity to nourish the fundamental concepts of the subject could not be denied.
CNTs are distinguished fillers because of their intrinsic characteristics such as semiflexibility, high aspect ratio, etc. [45-47], In recent years, it has become a common practice to incorporate the superior mechanical, thermal, and electrical properties of CNTs via dispersing them in polymers [48-54]. A fair amount of data is also reported particularly on the significance of CNTs localization in multiphase systems [14, 15, 19], Indeed, one possible scenario is the preferred localization of CNTs in the dispersed domains.
In this study, we attempt to identify the mechanisms governing the shear-induced deformation and relaxation of the CNT-containing PA6 droplets in the PP matrix. The first part of this work deals with the morphology evolutions under simple shear flows. PA6 droplets filled with various amounts of CNT are traced under simple shear flows utilizing online optical microscopy. In this way, the morphology evolution was followed during processing. Detailed theoretical and experimental analyses are provided on the dispersion quality of CNTs in PA6, morphology of blends with various CNT contents, CNT localization in blends, and the rheology of blends. Such analyses enable us to draw conclusions on the governing mechanisms of morphology evolution and droplet deformation with time. Moreover, the droplet deformation data are compared with the Maffettone-Minale (MM) model , Based on the results, the contributions of the elastic and interfacial forces to the droplet deformation are discussed. In the second part of this work, the shape relaxation of the deformed CNT-filled droplets is studied after the cessation of flow. The effects of the shear-rate of the pre-shear flow on the relaxation kinetics are investigated. Also, consequences of CNT inclusions in PA6 on the final morphology are revisited. The results of this work help pave the way to design industrially significant advanced materials such as in situ fibrillated full-polymer composites, PP/PA6 blends being one of them .
The deformation of an isolated droplet in shear flows has been the subject of many studies [25-34]. These efforts provide a valuable insight into the response of emulsions and immiscible polymer blends to complex flow fields. However, they are simplified by neglecting certain aspects of real systems such as droplet-droplet interactions . The pioneering works of Taylor [25, 26] provoked many to consider the subject of droplet deformation in shear flows as a promising area of the science of emulsions. Taylor investigated an isolated Newtonian droplet embedded in a Newtonian matrix experiencing a simple steady shear flow in the region of small deformations. He found that the deformation of such a droplet is described by two dimensionless parameters, namely the viscosity ratio, [lambda], and the capillary number, Ca. He defined the viscosity ratio between the two phases as
[lambda] = [[eta].sub.d]/[[eta].sub.m] (1)
where [[eta].sub.d] and [[eta].sub.m] are the zero-shear viscosities of the dispersed phase and the matrix phase, respectively. The second parameter, i.e. the capillary number is
Ca = [[eta].sub.m] [?] [r.sub.0]/[alpha] (2)
where [??] is the applied shear-rate, [r.sub.0] is the un-deformed droplet radius, and [alpha] is the interfacial tension between the two phases. The capillary number could be described as the ratio between the viscous stresses (trying to deform the droplet) to the interfacial stresses (trying to keep the shape of the droplet unchanged). Taylor predicted that the deformation parameter, D, is related to the viscosity ratio and the capillary number in the following form:
D = [a - b/a + b] = Ca [19[lambda] + 16/16[lambda] + 16] (3)
In this equation, a and b are lengths of the major and minor axes of the deformed droplet, respectively. On the basis of Taylor's equation, the deformation parameter is a linear function of Ca, and only weakly dependent on 2. Moreover, the orientation angle, i.e. the angle between the major axis of the droplet and the flow direction, is assumed constant and equal to 45[degrees]. Today, a number of analytical solutions are available which not only relate the deformation parameter to higher orders of Ca, but also define a correlation between the orientation angle and Taylor's dimensionless parameters [27-31, 37].
The limitation of Taylor's studies to Newtonian systems inspired many to investigate the effects of viscoelastic components on the deformation of droplets . Such studies often consider the viscoelasticity of one component while the other component is Newtonian. This approach simplifies the interpretation. Nevertheless, theoretical difficulties such as defining proper dimensionless groups and probable dependence of the results on the chosen constitutive equation still remain. A general approach is proposed by Greco [30, 31] for non-Newtonian fluids in slow, steady-state flows. He suggested using a dimensionless number to weight the elastic stresses with respect to the interfacial stresses. It was stated that this number could be considered as the non-Newtonian counterpart of the capillary number. Yet, the main drawback of his analysis is that the predictions are restricted to steady-state flows.
The concept of ellipsoidal models was first developed by Maffettone and Minale , They proposed a phenomenological model to predict the shape of the droplet by introducing a symmetric, positive, second-order tensor S. The eigenvalues of S represent the square semiaxes of the ellipsoidal droplet. The model describes the evolution of S as
dS/dt - [omega] x S + S x [omega] = -[f.sub.1]/[tau] [S - g(S)I] + [f.sub.2] ([??] x S + S x [??]) (4)
in which [tau] = [alpha]/[mu][r.sub.0] is a characteristic time, [alpha] is the interfacial tension, [mu] is the matrix viscosity, [r.sub.0] is the un-deformed droplet radius, and I is the unit second-order tensor, [??] and [omega] are the deformation rate and vorticity tensors defined as [??] = ([nabla]V + [nabla][V.sup.T])/2 and [omega] = ([nabla]V - [nabla][V.sup.T])/2, respectively. Here, [nabla]V is the velocity gradient and [nabla] [V.sup.T] is its transpose. The function g(S) = 3[III.sub.s]/[II.sub.s] is introduced in order to preserve the volume of the droplet, in which [II.sub.s] is defined as [II.sub.s] = [[(I : S).sup.2] - (I : [S.sup.2])]/2, and [III.sub.s] is the third invariant of S. The constants [f.sub.1] and [f.sub.2] are functions of the viscosity ratio and as suggested by Maffettone and Minale , are [f.sub.1] = 40([lambda] + 1)/(2[lambda] + 3)(19[lambda] + 16) and [f.sub.2] = 5/(2[lambda] + 3). This model is considered to be the foundation of all ellipsoidal equations of droplet shape.
The subject of droplet deformation is a crucial area of emulsions science because of its importance in polymer processing. The commercial need for advanced materials has led researchers to add a third component such as nanoparticles to immiscible polymer blends. Besides, development of new materials such as microfibrillar reinforced composites (MFCs) introduces new prospects of the subject of droplet deformation in shear and extensional flow fields [6, 7, 16]. Such complications challenge the present state of the knowledge of the droplet deformation theory.
The blends under study are composed of two immiscible polymers. PA6 is Akulon F223-D with a density of 1.13 g x [cm.sup.-3] supplied from DSM, Netherlands. PP is Jampilen HP525J with a density of 0.9 gx [cm.sup.-3] and a melt flow index (MFI) of 3 g x [(10 min).sup.-1] at 230[degrees]C (under a weight of 2.16 kg) obtained from Jam Polypropylene, Iran.
CNTs are Nanocyl[TM] NC7000 with a carbon purity of 90%, produced via catalytic carbon vapor deposition process, supplied from Nanocyl S.A., Belgium. CNTs are multiwalled with an average diameter of 9.5 nm and an average length of 1.5 [micro]m according to the supplier. The surface area is 250-300 [m.sup.2] x [g.sup.-1].
In this study two sets of samples were prepared: PA6/CNT nanocomposites and PP/PA6-CNT blends. All samples were prepared by melt compounding technique in a 60 [cm.sup.3] internal mixer (Braebender Plasticorder W50, Germany) equipped with a Banbury type rotor design. Prior to any processing, all materials were dried for 48 h at 80[degrees]C under vacuum.
PA6/CNT Nanocomposites. Nanocomposites of PA6 and CNT were prepared in the internal mixer at 240[degrees]C with a rotational speed of 100 rpm for 20 min. In order to have the same thermomechanical history in all samples, neat PA6 was processed at the same conditions and used as reference. This mixing protocol was utilized in order to reach a reasonable dispersion state after completion of the mixing. The applied mixing torque of the internal mixer also reached a steady-state value in all the samples before the completion of the process.
The CNT contents of the nanocomposites were chosen carefully so that we could cover all concentration regions of suspensions, i.e. dilute, semidilute, concentrated, and nematic. Based on suspensions theory , the transitions from dilute to semidilute for rod-like particles with a diameter of d and a length of L occurs at a volume fraction of [[phi].sub.0] [approximately equal to] 30([pi][d.sup.2]/4[L.sup.2]), from semidilute to concentrated at [[phi].sub.1] [approximately equal to] [pi]d/4L, and from concentrated to nematic based on Onsager's theory  at [[phi].sub.2] [approximately equal to] 3.3d/L. For the PA6/CNT nanocomposites studied here, these values correspond to concentrations of 0.14657, 0.7722, and 3.2 wt%, respectively. Thus, by mixing 0.1, 0.2, 0.5, 1, 2, and 5 wt% of CNTs in PA6, one can cover all concentration regions.
PP/PA6-CNT Blends. The PA6/CNT nanocomposites from PA6/CNT Nanocomposites Section were incorporated into PP (PP/PA6-CNT: 80/20 wt/wt) as the dispersed phase so that we could study the effect of CNT inclusions on the droplet dynamics. The components were blended in the internal mixer at 240[degrees]C with a rotational speed of 50 rpm for 5 min. The nomenclatures of the samples are summarized in Table 1.
Field Emission Scanning Electron Microscopy. The morphology of the blends are determined by performing Field Emission Scanning Electron Microscopy (FE-SEM) on cryofractured surfaces of the compression molded samples (pressed at 240[degrees]C) using a S-4160 (Hitachi, Japan) operating at 25 kV. The samples were gold sputtered to avoid charging.
Transmission Electron Microscopy. The CNT distribution in the blends has been examined using an EM 208S (Philips, Netherlands) Transmission Electron Microscopy (TEM) with an accelerator voltage of 100 kV on ultrathin sections of approximately 70 nm thickness.
Rheological Measurements. The rheological measurements were performed on samples using a rheometric mechanical spectrometer (Paar Physica UDS200, Austria). All measurements were done at 240[degrees]C in nitrogen environment utilizing a parallel plate fixture with the diameter equal to 25 mm and a constant 1 mm gap. In order to minimize the deformation history, all samples with or without CNT were held at the testing temperature for 30 min in nitrogen environment prior to any measurement. The storage modulus of annealed samples was measured against time at a constant strain and oscillation frequency of 0.5% and 1 rad x [s.sup.-1] respectively. The independency of storage modulus against time guaranteed microstructural stability of the samples during the rheological measurements and optical microscopy. Melt-state viscoelastic analyses of the specimens were carried out in linear viscoelastic region. The linear viscoelastic region was determined by monitoring the storage modulus in dynamic strain amplitude sweep experiments. A deformation strain of 1% was detected to be in linear response region for all of the samples. Linear rheological behavior of the specimens was studied using frequency sweep experiments in small strain oscillatory shear deformations.
Optical Microscopy. In situ morphological analysis of microstructural evolutions was performed on a shear-optical system in startup of steady shear flow. The microscope was combined of a phase contrast optical microscope (Leica DMRX, UK) and a double-side heated shearing stage (Linkam CSS450, UK). A CCD camera mounted directly on the microscope allowed to record the morphology evolutions of blends in real time. A controlled gap of 100 [+ or -] 5 [micro]m was set in all tests. The samples were melted and then pressed at 240[degrees]C into thin films with a thickness of 110 [+ or -] 4 [micro]m. Samples were put into the shearing stage at room temperature, then heated up to 240[degrees]C with a rate of 25[degrees]C x [min.sup.-1]. After reaching 240[degrees]C, the material was kept constant for 10 min in order to minimize the deformation history. Then, a steady-state shear flow with a shear-rate of 2 or 4 [s.sup.-1] was applied for 120 s and the morphology was recorded every 1 s.
RESULTS AND DISCUSSION
Morphology of Blends
When dealing with polymer nanocomposites, one has to ensure a fine state of dispersion of the nanoparticles. In order to obtain such information, dynamic rheological measurements have been conducted on the PA6/CNT samples. Figure 1 shows the complex viscosity of the nanocomposites with different CNT contents. At low frequencies, at a critical concentration between 0.5 and 1 wt% of CNT a transition from Newtonian to non-Newtonian response is observed. Previous studies on PA6/CNT nanocomposites have evidenced fine dispersion states of CNT in PA6 with the emergence of non-Newtonian behavior at low frequencies at a comparable CNT concentration , It should be noted that favorable thermodynamic affinity between CNT and PA6 promotes dispersion of nanoparticles in PA6 [14, 15, 19, 23],
Some authors suggest using the value of apparent yield stress as a direct indication of physical interactions between particles [16, 57-59]. The value of the yield stress increases with the dispersion extent. Therefore, one can infer the level of dispersion from this value. The apparent yield stress is determined by fitting the complex viscosity data, [[eta].sup.*] ([omega]), with a Carreau-Yasuda model including an additional term of yield stress, [[tau].sub.0]/[omega],
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)
where [[eta].sub.0] is the zero-shear viscosity, [lambda] is an average relaxation time, n is the flow index, a is the curvature parameter, and [[tau].sub.0] is the apparent yield stress . In the fitting procedure, we followed the method of Lertwimolnum et al. , First, the parameters were determined for the neat PA6 by taking the yield stress to be zero. Afterwards, these parameters were used as starting values for the nanocomposites. To simplify, it was assumed that the flow index n of the nanocomposites vary in a trust-region of 50% of its starting value. This assumption is reasonable since the slopes of the shear-thinning region of the viscosity curves at high frequencies are almost constant (Fig. 1). The results are presented in Table 2. The agreement between the experiments and the fitted curves were excellent. The parameters vary for different CNT contents. The curvature parameter a decreases with the CNT content. The relaxation time [lambda] increases slowly at first and then shows a sudden increase for higher concentrations. The zero-shear viscosity [[eta].sub.0] also increases with the addition of CNTs to the system. Furthermore, the value of the yield stress of nanocomposites increases with the CNT content. It confirms the emergence of CNT percolated networks inside the PA6. Considering the low CNT contents, the apparent yield stress data prove a well-dispersed state of CNTs in PA6 .
FE-SEM was performed on the samples in order to investigate the morphology of the blends. Figure 2 shows the morphology of the blends with different CNT contents. The blends exhibit the typical sea-island morphology for the CNT contents below 2 wt%. At concentrations of 2 and 5 wt%, nodular microstructures are formed because of the difficulties of dispersing the minor phase. It is well-known that during the blending step, the morphology is controlled mainly by two parameters, namely the capillary number, Ca, and the viscosity ratio, [lambda], (see Theory section for full parameter definitions). The presence of CNTs in the dispersed phase increases [lambda] and makes it more difficult to disperse the minor phase. Figure 3 shows FE-SEM micrographs of the PP/PA6-2 and PP/PA6-5 samples. The existence of non-uniform nodules of the PA6 phase in the continuous phase is obvious. Such irregular structures are formed during the mixing process by the intense stresses of the rotor.
Figure 4 displays the FE-SEM images of the surface of unfilled-PA6 droplets and the cavities in the matrix because of the PA6 droplets removal after cryofracture. The existence of such a distinct and smooth interface between the components of the blend points to the low thermodynamic affinity of the two phases [14, 15, 19, 23], By adding CNTs to the samples, one should first determine its effect on such interactions between the phases. It is also necessary to establish the localization of CNTs. Classical thermodynamics is often used to predict the location of fillers in immiscible polymer blends. It is based on the concept that the minimization of the interfacial energy is the driving force for the fillers to locate in a specific phase or at the interface.
It is possible to find the equilibrium position of fillers according to Young's equation. A quantitative estimation of the so-called wetting coefficient, [[omega].sub.a], can help determine the localization. This is only possible provided that the surface free energy of the components and their temperature dependencies are known :
[[omega].sub.a] = [[gamma].sub.Particle-Pol B] - [[gamma].sub.Particle-Pol A]/[[gamma].sub.Pol A-Pol B] (6)
where [[gamma].sub.Particle-Pol A], [[gamma].sub.Particle-Pol B], and [[gamma].sub.Pol A-Pol B] are the interfacial energies between fillers and polymer A, between fillers and polymer B, and between the two polymers, respectively. If the wetting coefficient is higher than 1, the particles will be located in polymer A; if it is lower than -1, the particles will prefer to distribute in polymer B; and finally, if it is between -1 and 1, the particles will be located at the interface between the two polymers. The interfacial energy can be determined from surface energies of the components. Based on the type of the surfaces, mainly two approaches, i.e. the harmonic-mean equation (typically valid between low-energy materials) and the geometric-mean equation (typically valid between a low-energy material and a high-energy material) are used. The harmonic-mean equation is :
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)
and the geometric-mean equation is 
[[gamma].sub.12] = [[gamma].sub.1] + [[gamma].sub.2] - 2([square root of ([[gamma].sup.d.sub.1] [[gamma].sup.d.sub.2]]) = [square root of ([[gamma].sup.p.sub.1] [[gamma].sup.p.sub.2]])]) (8)
where[gamma].sub.1] and [gamma].sub.2] are the surface tensions of components 1 and 2; [gamma].sup.d.sub.1] and [gamma].sup.d.sub.2] are the dispersive parts of the surface tensions of components 1 and 2; and [gamma].sup.p.sub.1] and [gamma].sup.p.sub.2] are the polar parts of the surface tensions of components 1 and 2. The respective values of the surface energies for PA6 and CNTs are available in the literature at 240[degrees]C . For PP, such data are taken from  at 200[degrees]C and using the temperature dependency relationship of its surface tension (-d[gamma]/dT = 0.058 mN [m.sup.-1] [K.sup.-1] ) were evaluated at 240[degrees]C. The extrapolated values of the surface energies for all components at 240[degrees]C are given in Table 3. Assuming PP as polymer A and PA6 as polymer B in Eq. 6, by incorporating the harmonic-mean and the geometric-mean equations, the wetting coefficients at 240[degrees]C were -1.44 and -2.25, respectively. This means that the CNTs should selectively distribute in the dispersed phase, i.e. PA6.
It has been shown that beside the thermodynamic driving force, the kinetic parameters can also play a significant role in the nanoparticles transfer from one phase to the other [14, 22]. This can determine the final localization of the particles in polymer blends. If there were any CNTs in PP, the stronger interfacial interaction between PA6 and CNTs could promote the migration of CNTs from PP to PA6 or the interface. The relatively lower viscosity of PA6 (with a zero-shear viscosity of 355 Pa.s at 240[degrees]C) in comparison with PP (with a zero-shear viscosity of 703 Pa.s at 240[degrees]C) also induces CNTs to migrate to PA6 with the aid of shear stresses during the melt blending process. In order to make sure that the CNTs remain in the PA6 phase while testing the droplet deformation, it is crucial to inquire into the effects of the flow fields on the final localization of nanoparticles in the samples. The general mechanisms of particle migration inside a molten heterogeneous medium imply that at first the particles should approach the interface [61, 62]. Such mobility could be explained by three mechanisms: Brownian motions, motions induced by shear stresses, and motions of particles trapped in the zone between two droplets during a droplet-droplet collision. It is safe to assume that the last mechanism is not dominant in the current system since according to our optical imaging, almost no collisions were observed during the entire analysis procedure. A measure of the relative importance of the Brownian and the shear-induced motions is quantitatively given by the nondimensional Peclet number, Pe. If the relaxation of the individual nanotubes is the primary mechanism for the relaxation process, then the rotational Peclet number in a shear flow field of [??] shear-rate, is given as Pe = [??]/[D.sub.rot] in which [D.sub.rot] is the rotational diffusivity of the cylindrical nanotubes. For cylinders of length L and diameter d, the rotational diffusivity is expressed as [D.sub.rot] = 3[k.sub.B] T[ln(L/d) -0.446]/[pi][[eta].sub.0] [L.sup.3], where [[eta].sub.0] is the zero-shear viscosity of the polymer and [k.sub.B] T is the Boltzmann factor. In this study, if the CNTs are located in the lower viscosity material (PA6) and the shear-rate is 0.01 [s.sup.-1] 1 [less than or equal to] [??], the Peclet number is 1 [much less than] Pe. Consequently, one can verify that such Brownian motions are not the dominant mode of movement towards the interface in the probable particles migration. For CNT localization in the PP phase, the corresponding Peclet number would be still well above unity. As a result, the probable migration may occur in the convective transport regime where it is expected that shear stresses control the mobility of the nanoparticles. Therefore, rheological measurements were carried out on the samples after melt blending. Such experiments investigate the effects of strong stress fields of the internal mixer on the migration of CNTs to the matrix (Rheological Characterization Section).
As a final remark on the migration of CNTs, one should notice that a comprehensive picture of the particle migration phenomenon should also consider the surface effects. If a CNT particle migrates from PA6 phase to PP phase, the adsorbed macromolecules of PA6 must desorb to be replaced by macromolecules of PP . Based on the available data in literature [14, 15, 19, 23], the interfacial energies between CNTs and the PP phase are calculated to be 9.084 and 6.572 mJ x [m.sup.-2] according to the harmonic-mean and geometric-mean equations, respectively. The corresponding values of the interfacial energies between CNTs and the PA6 phase are 4.425 and 2.360 mJ x [m.sup.-2], and between the two polymers are 3.247 and 1.870 mJ x [m.sup.-2], in the same respect. Obviously, the thermodynamic affinity of CNT particles towards PP chains is less than their affinity to PA6 chains. Hence, desorption of PA6 macromolecules followed by adsorption of PP macromolecules during a migration process is not an energetically favored phenomenon. Moreover, it is interesting to note that the surface tension between the CNTs and the PP phase is even higher than that between the two immiscible polymers. As a result, the addition of CNTs to the dispersed phase is expected to induce no improvement on the interfacial strength of the two polymer phases. Figure 5a shows the surface of a filled-PA6 droplet in PP/PA6-0.5 sample. A cavity in the matrix because of the pulling out of a droplet after cryofracture is also shown in Fig. 5b for the same sample. As it is obvious, the interface between the two phases is quite distinct and smooth. Such finely fractured interfaces were also observed for the blends without CNT inclusions (see Fig. 3). Apparently, the nanotubes did not strengthen the interface in the samples as it was anticipated.
Figure 6 demonstrates TEM micrographs of the PP/PA6-0.5 sample. It further proves the selective localization of CNTs in PA6 phase. Nanoparticles form a network within the droplets which controls the response of the minor phase to the applied stresses. Such internal structures change the behavior of the minor phase because of the altered capillary number, viscosity and elasticity ratios , It should be noted that some CNTs would of course locate in the matrix because of the intensive mixing stresses applied. However, in Rheological Characterization Section, the dynamic rheological measurements of the blends show this amount to be negligible.
In order to determine the effect of shear flow on the final distribution of CNTs in the blends, linear viscoelastic behavior of the samples were monitored in frequency sweep experiments. Figure 7 demonstrates the variation of storage modulus, G', and complex viscosity, [[eta].sup.*], of the blends with frequency. The melt-compounded blends were taken from the internal mixer and kept at 240[degrees]C under nitrogen atmosphere for 30 min before testing. Such testing protocol investigates the final state of the distribution in the samples after intense mixing flow fields. It also ensures the distribution stability during the study by verifying the migration probability. One can clearly observe that for different CNT loadings below 5 wt%, the results are almost the same as if there is no change in the CNT concentration. Considering the thermodynamic affinity of CNTs to the PA6 phase, for these concentrations it can be stated that the majority of CNT particles are selectively located in the dispersed phase rather than the matrix . For the PP/PA6-5 sample, the appearance of a plateau in the storage modulus at low frequencies (Fig. 7a) implies the formation of a percolated structure with a yield stress within the continuous phase. Such a structure is because of the increase in the continuity of the dispersed phase (with high solid-like elasticity) in the matrix (see Fig. 3) . As a result, a Theologically percolated structure forms comprised of highly-elastic PA6 regions which claims the melt-state behavior in the low frequencies region. This structure could be deduced from either the apparent plateau in storage modulus (Fig. 7a) or the non-Newtonian trend in complex viscosity at long relaxation times (Fig. 7b) [50-54].
To further evaluate the influence of CNTs in the blends, the complex viscosity data were fitted with the modified Carreau-Yasuda model. The details of the fitting procedure are the same as in Fig. 1. The apparent yield stress values reported in Table 4 show that physical interactions are almost absent in the blends with CNT contents lower than 5 wt%. It therefore suggests that the presence of a few CNTs located in the matrix can be ignored for blends with CNT contents below 5 wt%. Considering the previous results, for these samples one can assume that each PA6 nodule contains a percolated network of well-dispersed CNTs, but these networks are isolated from each other. However, for the sample with the highest CNT content, these structures grow to form a percolated structure in the entire material and dominate the rheological response. The emergence of an apparent yield stress is a sign of this dominance .
Therefore, it should be noted that there is the possibility that some CNTs are present in the matrix. However, their presence is not influencing enough to be observed in the rheological tests since the concentration is not passing the rheological percolation threshold. Nevertheless, minor CNT inclusions in the matrix can be ignored since no significant alteration of the rheology of the blends is observed upon CNT addition to the dispersed phase. Moreover, no apparent yield stress is emerged for blends with CNT contents below 5 wt%. Thus, from all thermodynamic statements as well as rheological investigations, it is clear that the majority of CNTs selectively choose to distribute and remain in PA6 for the entire testing duration.
Droplet Deformation in Steady Shear Flows
The optical-shear apparatus was employed to capture the effect of shear flow fields on droplet deformation in the blends. It was noted before that the accumulation of CNTs in the dispersed phase results in the gradual emergence of solid-like elastic structures within PA6 nodules. These structures could not be deformed easily, especially at high CNT contents. Figure 8 shows the optical images of PP/PA6-2 sample before startup of flow and after 60 s of startup of flow. One can clearly observe the stretched skin of PA6 in alignment with the flow direction, around the strong solid-iike internal structure of the dispersed phase. Such a skeleton-skin-like structure is too strong to yield under applied shear-rates in this study and thus preserves the shape of the nodules. Therefore, in order to measure the evolution of droplet deformation, blends with CNT contents up to 0.5 wt% in the dispersed phase were tested, i.e. dilute and semidilute suspensions. Other samples with higher concentrations, i.e. concentrated and nematic suspensions, were assumed un-deformable under the specific shear-rates applied here.
Figures 9-12 show the optical micrographs of the evolution of droplet deformation with time for the samples under study at the shear-rate of 2 [s.sup.-1]. As observed, droplets deform at first and then reach a steady value. With increasing CNT loading, it appears that the droplet deformation rate and the final steady value gradually decrease. An image analysis software was utilized to quantify the deformation data of the optical images. For each sample, the respective dimensions of five droplets with an average equilibrium diameter of 30 [+ or -] [micro]m were measured. This droplet diameter assures that the deformation occurs under non-confined conditions [34, 36]. It was noticed that the droplets would not undergo breakup or coalescence processes. One should bear in mind that the optical images provided here are taken normal to the vorticity plane and thus some droplets may seem to interact with each other. By tracking the droplets under flow, it was found that the droplets collision and coalescence processes are not dominant in the window of our interest. For breakup, the situation is a bit different and will be pointed out in the following.
In order to quantitatively describe the deformation parameter of the droplets, Taylor's deformation parameter [25, 26] was calculated against time in the vorticity plane. The results are given in Fig. 13. It is obvious that an increase in CNT content in the dispersed phase results in the reduced deformability of the droplets. In the sample without any nano inclusion, the deformation parameter passes through a maximum and then reaches the steady slate. The occurrence of such a behavior could be ascribed to droplet breakups during the flow. To avoid any breakups during the measurements careful attention was paid while selecting the droplets. However, it appears that some minor breakups at the tips of the droplets may have occurred on time spots outside of imaging. By adding CNTs to the droplets, such undesired breakups are put to rest to a great extent and therefore the maximum in the diagram disappears. In accordance with previous discussions, the increased viscosity and solid-like elasticity of the dispersed phase should be the reasons behind the decreased breakups.
With increasing CNT content, the droplet deformation is slowed down. Moreover, the steady deformation value decreases. The former case could be because of the slowed-down dynamics of the PA6 macromolecules in the presence of nanoparticles . The linear rheological behavior of the corresponding nanocomposites of PA6 and CNT confirms the reduced chains mobility, see Fig. 1. The overall result of both energetic and topologic interactions between the chains and the particles hinders the mobility of the macromolecules. Hence, the bulk would respond more slowly to the applied flow field [63-65]. The latter case should be because of the inefficiency of the hydrodynamic viscous forces to deform the filled-droplets.
The deformation is a balance between the viscous forces trying to deform the droplet and the shape-conserving forces. In his classic theory, Taylor [25, 26] assumed the interfacial tension between the phases as the shape-conserving mechanism of droplets. He defined the nondimensional capillary number as the ratio of the viscous forces to the interfacial forces. In a filled-droplet there must be other governing forces in action too. In fact, the formation of a solid-like elastic structure within the droplets makes it harder for the viscous forces to deform the bulk of the dispersed phase. It is rational to consider this internal solid-like elasticity as a conflicting force along with the interfacial tension against the droplet deformation. With increasing the CNT loading and the gradual development of the percolated network, this solid-like structure dominates the behavior of the PA6 phase. It goes on until the complete freezing of the deformation of the dispersed phase against the external flow field, as it was shown for PP/PA6-2 sample at the applied shear-rate of 2 [s.sup.-1] (see Fig. 8). From the data, it appears that for the concentrations above 0.5 wt% at the applied shear-rate of 2 [s.sup.-1], the droplet deformation analysis seems meaningless. According to previous tests, it is clear that the percolated network of CNTs and polymer chains formed in the droplets is responsible for the response of the dispersed phase to the flow field . This mechanism not only describes the experimental data here, but also is in accordance with previous simulation studies. Some instances include the introduction of elastic stresses by adding polymer chains to the Newtonian droplets , the resistance of the suspended particles in the droplets against flow , and addition of a viscoelastic contribution to the interfacial stresses in surfactant-stabilized emulsions .
One can easily deduce from the above that the balance between shear and resisting forces could occur at larger droplet deformations if the applied shear flows were more powerful. In order to investigate this idea, PP/PA6-0.2 and PP/PA6-0.5 samples were studied under a shear-rate of 4 [s.sup.-1], see Figs. 14 and 15. Figure 16 compares the deformation parameter data of these figures. Clearly, the deformation rate and the steady value have both increased with increasing the applied shearing forces. Besides, the breakup process could be intensified under stronger shearing, which is likely the reason of the emerging maximum point at higher shear-rates. Therefore, one can infer that the hydrodynamic forces are in control and dictate the deformation response of the filled-droplets.
In order to investigate the governing elastic forces, the ellipsoidal model of Malfetone and Minale (MM model)  is incorporated to predict the deformation parameter against time. The MM model is often applicable for the droplets which always keep ellipsoidal shape in Bow conditions. Nevertheless, it is quite capable to predict the transient deformability in immiscible polymer blends with a good precision . The MM model incorporates the capillary number to weight the relative effects of the shape distorting viscous forces to the shape reserving interfacial forces. It was shown that with the addition of CNTs to the droplets, one should expect the solid-like elastic forces to couple with the interfacial tension against the viscous forces. Here, an attempt was made to modify the capillary number in order to account for such effects. The proposed modified capillary number, [Ca.sub.modified], would be
[Ca.sub.modified] = [[eta].sub.m] [??]/[alpha] / [r.sub.0] + E (9)
where E is the contribution of the solid-like elastic forces to the total resistance against the deformation. The other parameters were described before. This modification simply suggests that for the definition of the capillary number, one should account for all the resisting mechanisms along with the interfacial contribution ([alpha]/[r.sub.0]). Another modification of the capillary number was addressed by Frijters et al.  where it was suggested to account for the distortion of the linear shear gradient caused by the presence of the droplets.
The model was utilized to predict the droplet deformation. By incorporating the viscosity ratio of each composition, [lambda], and the modified capillary number, [Ca.sub.modified], the resisting forces ([alpha]/[r.sub.0] + E) were estimated. The viscosity ratio is usually taken as the ratio of the zero-shear viscosities of the two phases. However, the zero-shear viscosity cannot be easily detected for some nanocomposites. Moreover, the viscosity of nanocomposites is a strong function of the shear-rate. Therefore, to overcome such problems we performed startup of steady shear flow experiments at a constant shear-rate of 0.1 [s.sup.-1] (data not shown here). This shear-rate is small enough to be close to the zero-shear condition while not so small to be experimentally problematic to measure the shear stress. For the PA6/CNT samples, the shear stress with time showed a maximum and then reached a steady-state. This maximum is often described as a yielding characteristic of the sample. At the steady-state the sample is in equilibrium under the flow. We have taken the viscosity value at this steady-state for all samples and used it to determine the viscosity ratios in our calculations. Figure 17 demonstrates a comparison between the model with [Ca.sub.modified] and the experimental data. While the MM model cannot capture the slow kinetics of droplet deformation at the early stages, it is quite successful to predict the overall trend of the deformation. Table 5 reports the estimated resisting forces against deformation for each sample. With an increase in the CNT content or the applied shear-rate, the value of such resistances increases. With increasing the shear-rate, the response of a viscoelastic fluid will become more elastic. Thus, it is not far from expectations that the solid-like elasticity induced by the nanoparticles couples to a more extent with the elasticity of the chains as they form a percolated network together . It is interesting to note that with the addition of only 0.1 wt% of CNT to PA6, such resisting forces develop very fast and decrease the deformability and breakup of the droplets.
One should notice the magnitude of the estimated resisting forces given in Table 5. A comparison between the interfacial forces and the solid-like elastic forces simply suggests that the key resistance stems from the solid-like elasticity of the internal structures. Assuming an average equilibrium diameter of 30 [micro]m for the droplets and a surface energy of 3.247 mJ x [m.sup.-2] between the polymers at 240[degrees]C, the average interfacial force is estimated to be 216 N x [m.sup.-2]. This value is even smaller than the calculated resisting forces of PP/PA6 sample and constitutes only about 34% of the resistance. Such an interesting point shows the significance of the entangled networks of long macromolecules even before the addition of CNTs. It is in accordance with the general idea that the viscoelasticity of the droplet is against the deformability . With the addition of 0.5 wt% CNTs, the contribution of the interfacial forces decreases down to less than 1% at shear-rates of 2 [s.sup.-1]. Such an increase in the elastic resistance of the droplets is because of the multiple consequences of the embedded CNTs in the bulk of the dispersed phase. Some instances include the formation of solid-like structures , emergence of particle-particle and/or particle-polymer as well as secondary polymer-polymer interactions [64-68], reduced chain mobility , and increased steric hindrance against diffusion . Hence, it is clear that the elasticity of the droplets instead of the interfacial contribution is the dominant mechanism for such a remarkable decline in the deformability of the droplets.
In previous studies, some researchers have shown that a viscoelastic matrix could intensify the deformation and breakup of the droplets  while others point to the contrary . However, there is a general consensus on that the viscoelastic droplets tend to deform less and reduce (or in some cases stop) the breakup process [16, 36, 42]. Usta et al.  studied the deformation of droplets filled with circular nanoparticles or long polymer chains with computational modeling techniques. They reported that nanoparticles reduced droplet deformation. They found that nanoparticles increase viscous shear stresses because of the resistance of the suspended particles to the shearing motion imposed by flow. Nanoparticles also alter the interfacial wetting forces coupled with the Brownian diffusion. Besides, by incorporating polymer chains in the droplets, deformation reduced even more efficiently than the case with the nanoparticles. Such an effect was ascribed to the introduction of elastic stresses that arise from the interactions between bonded monomers. Taking into account the unique physical characteristics of the CNTs such as high aspect ratio as well as semiflexibility of the individual particles [45, 46], it is rational to expect that they behave like polymer chains to a great extent . The ideas presented so far back up the earlier remarks on the mechanism of reduced deformability of droplets in presence of CNTs. To sum up, a schematic representation of the deformation of unfilled- and filled-droplets is shown in Fig. 18.
The present paper focuses on the effects of incorporating CNTs in the PA6 droplets and deals with morphology evolution and droplet deformation under simple shear flows. CNT-based nanocomposites of PA6 were prepared with different solid contents to cover all concentration regions according to classic suspensions theory. Linear rheological measurements proved the increased complex viscosity of the nanocomposites compared with pure PA6. It was shown to be because of the formation of percolated networks of CNTs and macromolecules. Such nanocomposites were added to the PP matrix to produce immiscible blends. FE-SEM was utilized in order to study the morphology of the blends. Typical sea-island morphology was observed for the samples with CNT contents lower than 2 wt%. Irregular nodules were found for higher CNT contents. Based on the classical thermodynamics framework and the linear rheology measurements of the blends, it was noted that the majority of CNTs would locate in the PA6 phase for CNT concentrations up to 2 wt%. A detailed analysis suggested that particle migration was not energetically favored in the blends.
Shear-optical microscope was employed to study the deformation of unfilled- and filled-droplets. The results clearly indicated that with the addition of CNTs to the PA6 droplets, solid-like elastic structures begin to develop within the droplets. Such structures dominate the response of the droplets to the imposed flows. A gradual reduction in Taylor's deformation parameter was reported with increasing CNT content. It was shown that with the formation of such structures an elastic force develops within the droplets which tends to keep the shape unchanged. TEM also confirmed such structures inside the droplets. The transient ellipsoidal MM model was used to predict the experimental data. It was coupled with a modified capillary number which accounts for both interfacial and elastic resistances. The evaluated elastic resisting stresses increased with increasing CNT concentration. Indeed, deformability was discussed to be meaningless for concentrations above 0.5 wt%. A skeleton-skin-like structure was proposed for the deformation characteristics of such highly filled-droplets. Also, it was concluded that the increased viscosity ratio as well as the solid-like elasticity of the internal structures of the droplets hindered the droplet breakups.
In the second part of this work, the relaxation kinetics of unfilled- and filled-droplets is studied with time after cessation of shear flows.
[1.] J. Duvall, C. Sellitti, V. Topolkaraev, A. Hiltner, E. Baer, and C. Myers, Polymer, 35, 3948 (1994).
[2.] Y. Seo, B. Kim, and K. Ung Kim, Polymer, 40, 4483 (1999).
[3.] J. Piglowski, I. Gancarz, M. Wlazlak, and H.-W. Kammer, Polymer, 41, 6813 (2000).
[4.] D. Shi, Z. Ke, J. Yang, Y. Gao, J. Wu, and J. Yin, Macromolecules, 35, 8005 (2002).
[5.] A. Sacchi, L. Di Landro, M. Pegoraro, and F. Severini, Eur. Polym. J., 40, 1705 (2004).
[6.] X. Li, M. Chen, Y. Huang, and G. Cong, Polym. Eng. Sci., 39, 881 (1999).
[7.] J. Teng, J.U. Otaigbe, and E.P. Taylor, Polym. Eng. Sci., 44, 648 (2004).
[8.] L.A. Pinheiro, C.S. Bittencourt, and S.V. Canevarolo, Polym. Eng. Sci., 50, 826 (2010).
[9.] G. Jannerfeldt, L. Boogh, and J.-A.E. Manson, Polym. Eng. Sci., 41, 293 (2001).
[10.] P. Macaubas, N.R. Demarquette, and J.M. Dealy, Rheol. Acta, 44, 295 (2005).
[11.] S.-N. Li, B. Li, Z.-M. Li, Q. Fu, and K.-Z. Shen, Polymer, 47, 4497 (2006).
[12.] L. Elias, F. Fenouillot, J.C. Majeste, and P. Cassagnau, Polymer, 48, 6029 (2007).
[13.] L. Elias, F. Fenouillot, J.-C. Majeste, G. Martin, and P. Cassagnau, J. Polym. Sci., Part B: Polym. Phys., 46, 1976 (2008).
[14.] A.-C. Baudouin, C. Badly, and J. Devaux, Polym. Degrad. Stab., 95, 389 (2010).
[15.] A.-C. Baudouin. J. Devaux, and C. Badly, Polymer, 51, 1341 (2010).
[16.] D. Chomat, J. Soulestin, M.-F. Lacrampe, M. Sclavons, and P. Krawczak, J. Appl. Polym. Sci., 132, (2015).
[17.] L. Zhang, C. Wan, and Y. Zhang, Compos. Sci. Technol., 69, 2212 (2009).
[18.] S. Bose, C. Ozdilek, J. Leys, J.W. Seo, M. Wubbenhorst, J. Vermant, and P. Moldenaers, ACS Appl. Mater. Interfaces, 2, 800 (2010).
[19.] A. Cayla, C. Campagne, M. Rochery, and E. Devaux, Synth. Met., 161, 1034 (2011).
[20.] H. Zou, K. Wang, Q. Zhang, and Q. Fu, Polymer, 47, 7821 (2006).
[21.] L. Liu, Y. Wang, Y. Li, J. Wu, Z. Zhou, and C. Jiang, Polymer, 50, 3072 (2009).
[22.] L. Liu, H. Wu, Y. Wang, J. Wu, Y. Peng, F. Xiang, and J. Zhang, J. Polym. Sci.. Part B: Polym. Phys., 48, 1882 (2010).
[23.] A.-C. Baudouin, D. Auhl, F. Tao, J. Devaux, and C. Bailly, Polymer, 52, 149 (2011).
[24.] A. Goldel, A. Marmur, G.R. Kasaliwal, P. Potschke, and G. Heinrich, Macromolecules, 44, 6094 (2011).
[25.] G.I. Taylor, Proc. R. Soc. A, 138, 41 (1932).
[26.] G.I. Taylor, Proc. R. Soc. A, 146, 501 (1934).
[27.] P.L. Maffettone and M. Minale, J. Nonnewton. Fluid Merit., 78, 227 (1998).
[28.] S. Guido and F. Greco, Rheol. Acta, 40, 176 (2001).
[29.] D. Megias-Alguacil, K. Feigl, M. Dressier, P. Fischer, and E.J. Windhab, J. Nonnewton. Fluid Merit., 126, 153 (2005).
[30.] F. Greco, Phys. Fluids, 14, 946 (2002).
[31.] F. Greco, J. Nonnewton. Fluid Merit., 107, 111 (2002).
[32.] H.S. Lee and M.M. Denn, J. Nonnewton. Fluid Mech., 93, 315 (2000).
[33.] P. Tanpaiboonkul, W. Lerdwijitjarud, A. Sirivat, and R.G. Larson, Polymer, 48, 3822 (2007).
[34.] P. van Puyvelde, A. Vananroye, R. Cardinaels, and P. Moldenaers, Polymer, 49, 5363 (2008).
[35.] M. Gabriele, R. Pasquino, and N. Grizzuti, Macromol. Mater. Eng., 296, 263 (2011).
[36.] S. Guido, Curr. Opin. Colloid Interface Sci., 16, 61 (2011).
[37.] J.M. Rallison, J. Fluid Mech., 98, 625 (1980).
[38.] S. Frijters, F. Gunther, and J. Harting, Soft Matter, 8, 6542 (2012).
[39.] L.A. Utracki and Z.H. Shi, Polym. Eng. Sci., 32, 1824 (1992).
[40.] A. Bigdeli, H. Nazockdast, A. Rashidi, and M.E. Yazdanshenas, J. Polym. Res., 19 (2012).
[41.] B. Lin, U. Sundararaj, and P. Guegan, Polym. Eng. Sci., 46, 691 (2006).
[42.] O.B. Usta, D. Perchak. A. Clarke, J.M. Yeomans, and A.C. Balazs, J. Client. Phys., 130 (2009).
[43.] R. Skartlien, E. Solium, A. Akselsen, and P. Meakin, Rheol. Acta, 51, 649 (2012).
[44.] M. Kong, Y. Huang, G. Chen, Q. Yang, and G. Li, Polymer, 52, 5231 (2011).
[45.] M. Moniruzzaman and K.I. Winey, Macromolecules, 39, 5194 (2006).
[46.] P.-C. Ma, N.A. Siddiqui, G. Marom, and J.-K. Kim, Composites Part A, 41, 1345 (2010).
[47.] B.P. Grady, J. Polym. Sci., Part B: Polym. Phys., 50, 591 (2012).
[48.] Q. Zhang, F. Fang, X. Zhao, Y. Li, M. Zhu, and D. Chen, J. Phys. Client. B, 112, 12606 (2008).
[49.] L. Moreira, R. Fulchiron, G. Seytre, P. Dubois, and P. Cassagnau, Macromolecules, 43, 1467 (2010).
[50.] P. Potschke, M. Abdel-Goad, I. Alig, S. Dudkin, and D. Lellinger, Polymer, 45, 8863 (2004).
[51.] U.A. Handge and P. Potschke, Rheol. Acta, 46, 889 (2007).
[52.] T. Chatterjee and R. Krishnamoorti, Macromolecules, 41, 5333 (2008).
[53.] Y.Y. Huang, S.V. Ahir, and E.M. Terentjev, Phys. Rev. B: Condens. Matter Mater. Pliys., 73 (2006).
[54.] F. Du, R.C. Scogna, W. Zhou, S. Brand, J.E. Fischer, and K.I. Winey, Macromolecules, 37, 9048 (2004).
[55.] R.G. Larson, The Structure and Rheology of Complex Fluids, Oxford University Press, New York (1999).
[56.] M. Wang, W. Wang, T. Liu, and W.-D. Zhang, Compos. Sci. Technol., 68, 2498 (2008).
[57.] M.J. Kayatin and V.A. Davis, Macromolecules, 42, 6624 (2009).
[58.] F. Berzin, B. Vergnes, and L. Delamare, J. Appl. Polym. Sci., 80, 1243 (2001).
[59.] W. Lertwimolnun and B. Vergnes, Polym. Eng. Sci., 46, 314 (2006).
[60.] S. Wu, Polymer Interface and Adhesion, Marcel Dekker Inc, New York, Basel (1982).
[61.] F. Fenouillot, P. Cassagnau, and J.-C. Majeste, Polymer, 50, 1333 (2009).
[62.] L. Elias, F. Fenouillot, J.C. Majeste, P. Alcouffe, and P. Cassagnau, Polymer, 49, 4378 (2008).
[63.] G.N. Toepperwein, N.C. Karayiannis, R.A. Riggleman, M. Kroger, and J.J. Pablo, Macromolecules, 44, 1034 (2011).
[64.] J.B. Hooper and K.S. Schweizer, Macromolecules, 39, 5133 (2006).
[65.] B.J. Anderson and C.F. Zukoski, Macromolecules, 41, 9326 (2008).
[66.] B. Anderson and C. Zukoski, Macromolecules, 42, 8370 (2009).
[67.] B.J. Anderson and C.F. Zukoski, Langmuir, 26, 8709 (2010).
[68.] L.M. Hall and K.S. Schweizer, Macromolecules, 44, 3149 (2011).
[69.] M. Mu, R.J. Composto, N. Clarke, and K.I. Winey, Macromolecules, 42, 8365 (2009).
A. Gooneie, (1) H. Nazockdast, (2) F. Shahsavan (2)
(1) Chair of Polymer Processing, Department of Polymer Engineering and Science, Montanuniversitat Leoben, Otto Glockel-Strasse 2, 8700-Leoben, Austria
(2) Department of Polymer Engineering and Color Technology, Amirkabir University of Technology, 424 Hafez Ave Tehran 15875-4413, Iran
Correspondence to: Ali Gooneie; e-mail: email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
TABLE 1. Formulation and nomenclature of the samples. Name PA6 wt% PP wt% CNT wt% PA6/CNT nanocomposites PA6 100 -- -- PA6-0.1 99.9 -- 0.1 PA6-0.2 99.8 -- 0.2 PA6-0.5 99.5 -- 0.5 PA6-1 99 -- 1 PA6-2 98 -- 2 PA6-5 95 -- 5 PP/PA6-CNT blends PP/PA6 20 80 -- PP/PA6-0.1 19.98 80 0.02 PP/PA6-0.2 19.96 80 0.04 PP/PA6-0.5 19.9 80 0.1 PP/PA6-1 19.8 80 0.2 PP/PA6-2 19.6 80 0.4 PP/PA6-5 19 80 1 TABLE 2. Fitted parameters of the Carreau-Yasuda model in PA6/CNT samples. [[tau].sub.0] [[eta].sub.0] [lambda] Sample (Pa) (Paxs) (s) a n PA6 0 436 9.9043 e-4 0.6686 0.0701 PA6-0.1 10 439 5.6592 e-4 0.5264 0.0743 PA6-0.2 15 659 5.5218 e-4 0.4400 0.0684 PA6-0.5 148 1278 3.1372 e-4 0.2863 0.0702 PA6-1 2416 7070 1.9851 e-4 0.1627 0.0363 PA6-2 14607 157034 0.0072 0.1194 0.0547 PA6-5 126722 1216091 0.8260 0.1783 0.0433 TABLE 3. Values of the surface energies for components at 240[degrees]C. Dispersive Polar surface Total surface surface energy, energy, energy mJ x [[gamma].sup.d] [[gamma].sup.p] Material [m.sup.-2] mJ x [m.sup.-2] mJ x [m.sup.-2] PA6 26.4 23.2 3.2 PP 17.1 16.7 0.4 CNT 27.8 17.6 10.2 TABLE 4. Apparent yield stress, [[tau].sub.0], in PP/PA6-CNT samples. Sample [[tau].sub.0] (Pa) PP/PA6 0 PP/PA6-0.1 0 PP/PA6-0.2 2 PP/PA6-0.5 16 PP/PA6-1 20 PP/PA6-2 41 PP/PA6-5 944 TABLE 5. Values of the estimated resisting forces. Sample ([alpha]/[r.sub.0]) + E (N x [m.sup.-2]) [??]=2[s.sup.-1] PP/PA6 628.5 PP/PA6-0.1 6064.7 PP/PA6-0.2 6927.4 PP/PA6-0.5 24,026.8 [??]=4[s.sup.-1] PP/PA6-0.2 16,960.0 PP/PA6-0.5 46,037.6
|Printer friendly Cite/link Email Feedback|
|Author:||Gooneie, A.; Nazockdast, H.; Shahsavan, F.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 2015|
|Previous Article:||Processing of composite polystyrene foam with a honeycomb structure.|
|Next Article:||Multilayer coextrusion of rubber compounds.|