научная статья по теме RHEOLOGY OF SILICA-FILLED POLYSTYRENE: FROM MICROCOMPOSITES TO NANOCOMPOSITES Физика

Текст научной статьи на тему «RHEOLOGY OF SILICA-FILLED POLYSTYRENE: FROM MICROCOMPOSITES TO NANOCOMPOSITES»

ВЫСОКОМОЛЕКУЛЯРНЫЕ СОЕДИНЕНИЯ, Серия А, 2012, том 54, № 6, с. 877-893

КОМПОЗИТЫ

УДК 541.64:535.132

RHEOLOGY OF SILICA-FILLED POLYSTYRENE: FROM MICROCOMPOSITES TO NANOCOMPOSITES1

© 2012 г. G. Havet and A. I. Isayev

Department of Polymer Engineering, The University of Akron, Akron, Ohio 44325-0301, USA e-mail: aisayev@uakron.edu Received November 29, 2011 Revised Manuscript Received January 23, 2012

Abstract — Rheology and viscoelastic behavior of polystyrene (PS)/silica microcomposites and nanocompos-ites were studied. The apparent viscosity, transient shear stress growth after startup shear flow and stress relaxation after cessation of flow at various shear rates, the complex dynamic viscosity, the storage and loss moduli at small and large strain amplitudes and various frequencies were performed. The effect of size, shape and volume concentration of silica was discussed. The maximum volume concentration, corresponding to the concentration at which the relative viscosity of mixtures goes to infinity, with respect to the hydrodynamic contribution of the particles and to polymer-filler interactions was obtained. The difference between the yield stress and residual stress is shown. The domain of equivalence between the apparent viscosity as a function of the shear rate in steady state flow and the complex dynamic viscosity as a function of the strain rate amplitude in highly nonlinear region of oscillatory flow was established. The viscoelastic behavior was interpreted based on the morphology of microcomposites and nanocomposites observed by SEM.

INTRODUCTION

The rheological behavior of polymers reinforced by nanometer scale particles (carbon blacks, CB) has been extensively characterized. Early works by Payne and Whittaker [1] on the oscillatory shear flow and by Vinogradov et al. [2] in the steady shear flow were concerned with elastomer/CB mixtures. Various features of their behavior were attributed to the particle-particle interactions and breakdown of aggregates. The most extensive review on the effect of interactive fillers in elastomeric systems on their dynamic properties was published by Donnet [3]. Four factors contributing mainly to the elastic modulus of these mixtures were defined: matrix properties, hydrodynamic effects, filler-rubber and inter-aggregate interactions. Contributions of these factors to the overall properties are considered to be additive. In particular, the strain dependence of the modulus of mixtures was attributed to inter-aggregate interactions and matrix properties. However, methods to evaluate other contributions were not provided. Also, the focus on dynamic properties alone would not lead to full understanding of this problem.

Rheology of polystyrene (PS) filled with various fillers was extensively studied by White and coworkers [4, 5] at concentrations up to 30% by volume. The principal concern of their studies was mixtures of PS with small particles (<1 ^m) inducing the yield stress. In particular, Tanaka and White [5] evaluated mixtures

1 This article is dedicated to Professor Valery Grigorevich Kuli-chikhin on the occasion of his 70th Birthday.

of PS with CB, titanium oxide and calcium carbonate with and without a surface treatment. The appearance of the yield stress was attributed to particle-particle interactions due to the London—van der Waals and electrostatic forces. The surface treatment was shown to have an important effect on the rheological behavior and yield stress. In particular, the surface treatment strongly affected the particle-particle interactions. Accordingly, it was concluded that the yield value was due to the particle-particle interaction and the tendency of small particles with the large surface area to form a gel-like structure. The problem was theoretically treated by Tanaka and White [6]. Also, by utilizing a Casson model, Suetsugu and White [4] determined the yield stress of four mixtures of 30 vol% of calcium carbonate of different particle sizes in PS. The yield value was found to linearly depend on the reciprocal value of particle diameters.

Mewis and Spaull [7] reviewed rheology of mixtures with colloidal particles. For mixtures of titanium oxide in the non-Newtonian linseed oil, Mewis and de Bleyser [8] assumed the similar dependence of the shear stress and the first normal stress difference on concentration. In the oscillatory shear flow, at low particle concentrations corresponding to the linear dependence of viscosity on the volume fraction, properties of the mixture relative to those of the matrix were found to be independent of frequency and were in agreement with properties obtained in the steady shear flow. At higher particle concentrations, the relative values decrease with increasing frequency. This behav-

ior was attributed to variations of the hydrodynamic interaction between the particles.

Aral and Kalyon [9] investigated viscoelastic properties of polydimethyl siloxanes (PDMS) containing non-colloidal hollow spheres. Their results agreed with those obtained in earlier observations. Utilizing the small amplitude oscillatory flow, where the rheo-logical properties are independent on the strain amplitude, an elastic plateau was observed at low frequencies. A qualitative interpretation of the rheological behavior observed in these experiments was given. An attempt for a quantitative evaluation was also made assuming the particle-particle interaction alone. However, in some cases the behavior appears to be also due to the particle-matrix interaction.

A few studies [10—15] have specifically been directed to investigation of the effect of the polymer-filler interaction. In particular, Han et al. [13, 14] studied the effect of coupling agents on rheology of polyolefin containing calcium carbonate, talc, glass beads and fiber glass. They showed that with an addition of coupling agents a decrease in viscosity along with an increase of elasticity or vice versa took place. This was attributed to a plasticizing effect of short chains of coupling agents. SEM micrographs showed a lower interfacial bonding in the samples containing coupling agents. These additives were acting more like processing aids or nucleating agents than coupling agents.

Schreiber and coworkers [15] interpreted the ease of dispersion of various plasma treated calcium carbonates in polyethylene (PE) and polyvinylchloride (PVC) in terms of acid-base interactions between the filler and polymer. Later, similar studies were carried out on LDPE and chlorinated PE containing surface coated rutile titanium oxide [11]. Also, Bomal and Godard [12] studied rheology of LLDPE—calcium carbonate mixtures treated by various carboxylic acids. They showed a strong dependence of the maximum volume fraction on the matrix-particle interaction. However, the presence of gel-like structure was attributed to the particle-particle interactions.

Kosinski and Caruthers [16, 17] studied the stress growth in the steady shear flow of PDMS of various molecular weights filled with fumed silica. Aranguren et al. [10] studied the influence of the hydrogen bonding on the rheology of mixtures of PDMS of various molecular weights with fumed silica particles subjected to various surface treatments. They stated that within the range of concentrations under investigation (less than 30 phr), the strain amplitude dependence was not due to the aggregate-aggregate interactions, but rather to the aggregate-polymer-aggregate interactions. In addition to the primary entanglements of the adsorbed polymer chains and the secondary entanglements of the non-adsorbed polymer chains, a significant contribution arises due to the direct bridging of polymer chains.

The interaction of PS melt with fumed silica through hydrogen bonding was characterized by Lipa-

tov [18]. However, its influence on rheology was not established. Also, no conclusion about changes in viscoelastic properties upon addition of fillers was made. In some cases, the viscosity of the mixture was increased, while the elastic response was decreased. In other cases, an addition of fillers simply increased viscoelastic properties of the melt.

Zhang and Archer [19] studied rheology of PEO-silica mixtures. The linear viscoelastic data indicated the presence of a solid-like behavior at low frequencies at particle volume fractions as low as 2%. This concentration is significantly lower than the concentration corresponding to the theoretical percolation threshold. Accordingly, filler networking mechanism was proposed wherein nanosized silica particles surrounded by an immobilized PEO shell bridged by larger polymer molecules.

Nanocomposites of PS filled with silica nanoparti-cles were extensively studied by Kumar et al. [20, 21] and Jouault et al. [22]. In the former studies, the effect of melt processing conditions on the dispersion was studied. In the later study, the problem of dispersion was solved by using solution processing with dimethy-lacetamide being a good solvent for PS matrix. Similar to PEO [19], at a silica volume fraction lower than the percolation threshold, a new additional elastic contribution was detected corresponding to a longer terminal time than that of the matrix. Since the observed dispersion was good, the effect was attributed to the polymer-filler interaction and to an additional polymer conformational contribution due to the existence of indirect long-range bridging of the filler by the chains.

Havet and Isayev [23, 24] proposed a model of the rheology of polymer containing interactive fillers based on a double network created by the entangled polymer matrix and the adsorbed polymer. They provided a close set of equations arising from a thermody-namic description of the physics of the problem consistent with the thermodynamics of the chemical interactions. The adsorbed polymer molecules were considered to be present as brushes and bridges. Both networks were described by a nonlinear constitutive equation with the pro

Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.

Показать целиком