Analysis of agglomerate dispersion mechanisms of multiwalled carbon nanotubes during melt mixing in polycarbonate

Abstract

Dispersion of primary nanotube agglomerates in polymer melts is one of the difficult tasks when applying melt mixing for nanocomposite preparation. Hence, there is a need for a better understanding of the ongoing processes. Filler agglomerates generally undergo dispersion by rupture and erosion mechanisms, which usually occur simultaneously. To analyse these mechanisms and their corresponding dispersion kinetics 1 wt% multiwalled carbon nanotubes (MWNT) were incorporated into polycarbonate using a microcompounder. Different mixing speeds at constant melt temperature were applied thereby changing the applied stress. The states of MWNT agglomerate dispersion at different mixing times were assessed by quantifying the agglomerate area ratio and particle size distribution using image analysis of optical transmission micrographs. A model is proposed to estimate the fractions of rupture and erosion mechanisms during agglomerate dispersion. At low mixing speeds, the dispersion was found to be governed by both mechanisms, whereas rupture dominance increases with increasing mixing speed. Further, the relationship between electrical resistivity and dispersion was studied indicating a critical behaviour. A dependency on the amount of dispersed nanotubes was found only in a certain range of state of dispersion.

Introduction

In recent years, manufacturing of polymer-multiwalled carbon nanotubes (MWNT) composites has acquired a lot of attention due to the extraordinary properties of MWNT. In this context, the state of MWNT dispersion in a polymer matrix plays a decisive role for the application of polymer–MWNT composites. A high cohesive strength of the “as produced” primary MWNT agglomerates restricts their dispersion into individualized tubes within the polymer matrix during melt compounding. This leads in many cases to the presence of un-dispersed MWNT agglomerates in composites as shown e.g. in [1], [2], [3], [4], [5], [6], [7], [8]. The dispersive mixing operation during melt compounding of polymer-filler systems consists of several stages, namely filler incorporation, wetting and infiltration (by polymer melt), followed by dispersion, distribution, and flocculation (of filler in the polymer melt). During the stage of filler dispersion, the large initial filler agglomerates are reduced in size up to the smallest dispersible unit. In a first approximation, the change in size of large agglomerates into smaller parts can be attributed to two mechanisms, namely rupture (a bulk phenomenon) and erosion (a surface phenomenon). In rupture mechanism, the large agglomerates are broken down into smaller ones in short times; whereas in erosion mechanism, large agglomerates are eroded into smaller ones by removing single or bundles of nanotubes from the agglomerate surface needing comparatively longer time. As the dispersion step is the most difficult one, it is considered to determine the rate at which filler disperses in the polymer melt [9]. Therefore, it is necessary to investigate the melt compounding conditions that can govern such underlying dispersion mechanisms of MWNT agglomerates in the polymer melt.

For fillers like carbon black (CB), silica, calcium carbonate (CaCO₃), etc. such kinds of dispersion mechanisms have been extensively investigated and reported. Kao and Mason [10] and later Powell and Mason [11] studied the process of dispersion of non-cohesive clusters (made of lucite and polystyrene) in silicon oil under pure shear flow. They observed dispersion by erosion mechanism and proposed a model, stating that the rate at which particles leave the surface of the cluster is proportional to the surface area of the cluster. An “onion peeling model” of CB dispersion in rubber which seems to be similar to erosion mechanism was proposed by Shiga and Furuta [12]. They stated that aggregates could be dispersed by either individually or collectively scraping from the surface of agglomerates. For dispersion of agglomerates by “rupture” mechanism, Bolen and Colwell [13] proposed that it occurs when shearing stresses exceed a certain threshold value dependent on the properties of filler. Later, extensive work has been reported on agglomerate dispersion mechanisms by the group of Manas-Zloczower and Feke [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]. One of the findings reported by them [16], [17], [18] reveals that in the course of CB dispersion, rupture mechanism generates a lower number of coarse fragments than erosion. Moreover, they stated that the critical shear stress required for dispersion by erosion process is much lower than the one required for the rupture process and based on an earlier reported work [24] they emphasised that the ratio of applied shear stress and cohesive strength of agglomerates could be decisive in agglomerate dispersion. This ratio was named as fragmentation number (Fa) by Hansen et al. [25]. The same group stated that the mechanism of agglomerate dispersion is dependent on a certain critical value of applied shear stress [26]. If Fa is smaller than Fa_critical, particles undergo dispersion by erosion and if Fa is larger than Fa_critical, particles undergo dispersion by rupture. Further, Potente et al. [27] proposed a mathematical model for the dispersion mechanisms which was used to simulate the dispersion process. A comparison was made with experimental data for the dispersion of CaCO₃ in polypropylene. Although the simulated and experimental results of degree of dispersion along screw length showed some deviations, the tendencies were similar. In a parallel work, Lozano et al. [28] proposed a model for dispersion mechanisms to simulate the particle size distribution of CaCO₃ in polypropylene. The predicted values could be fitted with the experimental values by assuming share of both dispersion mechanisms by trial and error method. Although both models consider dispersion by erosion and rupture mechanism, they do not directly quantify the share of both mechanisms during agglomerate dispersion.

Recently, we reported the influence of melt processing conditions on electrical resistivity and morphology of polycarbonate (PC)–MWNT composites [8]. It was stated that depending on compounding conditions, MWNT agglomerates could be dispersed by either rupture or erosion dominated mechanisms. In the prevailing work, to distinctly differentiate the occurrence of these dispersion mechanisms the dispersion kinetics of MWNT agglomerates was investigated by melt compounding PC with 1 wt% MWNT. Three different mixing speeds were employed, thereby applying three different shear rates, using a small scale DACA microcompounder. The effects on agglomerate size distribution and evolution of morphology due to different mixing conditions are reported. In any melt compounding process, filler agglomerate particles are subjected to dispersion by both rupture and erosion mechanisms as both mechanisms run parallel. A model is proposed to quantify the splitting of dispersion mechanisms during a melt mixing process into rupture and erosion. Furthermore, we report about electrical resistivities of composites produced under different mixing conditions.