Processing of Polymer Nanocomposites

1.2.1

Materials

This chapter focuses on polypropylene (PP) based nanocomposites, but the main results can be generalizd to other matrices. The compatibilizer is a polypropylene grafted with maleic anhydride (PP-g-MA). For a NC content of 5 wt%, a systematic study of the influence of the PP-g-MA amount on the microstructure was made [6]. Exfoliation was shown to be maximal at a 2 : 1 PP-g-MA–clay weight ratio. Above this ratio, a decrease in the mechanical properties of the nanocomposites was observed, due to the low viscosity of PP-g-MA with respect to the PP. However, this optimal ratio is not universal but depends also on the MA content [7].

The choice of an organoclay adapted to the chosen matrix is also an important issue. Many organomodified montmorillonites (OMMTs) are marketed, which differ depending on the type and amount (cation exchange capacity, CEC) of the surfactant. We recently compared three different OMMTs, supplied by two different companies [8]. Dellite© 67G (D67G) (from Laviosa Chimica Mineraria, Italy) and Cloisite© 20 (C20) (from BYK Additives, Germany) have the same organic modifier (dimethyl dihydrogenated tallow ammonium), while Cloisite© 30B (C30B) has another modifier (methyl, tallow, bis-2-hydroxyethyl ammonium). C30B is supposed to have a particular affinity with polar matrices, such as polyamide, while C20 and D67G are recommended for non-polar polymers, such as polyolefins [9]. After mixing under the same conditions and with the same composition, it was observed that C30B always resulted in microcomposites, without exfoliation, while C20 and D67G had a similar behavior and resulted in a mixed intercalated–exfoliated structure. In the following, these two OMMTs will be used to prepare the nanocomposites.

1.2.2

Processing

Although twin-screw extrusion remains the main compounding process, it is often useful to use an internal mixer to more easily characterize the influence of processing conditions. Indeed, in such a system, the rotor speed and the mixing time can be controlled independently. The torque is measured continuously during mixing, which easily leads to an accurate determination of the specific mechanical energy (SME) provided to the nanocomposite:

where C(t) is the torque at time t, N is the rotation speed (in rad/s), tmix is the mixing time, and m is the mass of the sample.

The sequence of introduction of the OMMT (before or after the melting of the matrix, in one or more steps) can also be easily tested on the internal mixer.

On a twin-screw extruder, the first step is to define the screw profile. It usually includes a melting section, followed by a few mixing sections, generally composed of blocks of kneading discs with different configurations (disc thickness, number of tips, staggering angle), depending on the type of mixing we expect to promote (dispersive or distributive) [10]. Compared to the internal mixer, only an overall SME can be measured on the extruder, including solid conveying and melting steps. Therefore, the real SME provided to the nanocomposite is more difficult to determine and can only be accessed from numerical modeling.

In the twin-screw experiments, the samples can be collected at the die exit, but also all along the screws after performing dead-stop experiments: once the steady-state is reached, the screw rotation and the feeding are suddenly stopped, and the barrel is quickly cooled and opened to give access to the screws. Figure 1.1 shows an example of this type of experiment.

1.2.3

Characterization

The size of the entities present in the nanocomposites can vary from a few tenths of microns for large agglomerates to 1–10 μm for aggregates, and a tenth of nanometers for tactoids [12]. Therefore, to quantify the dispersion state, it is crucial to consider these different characteristic scales and to use appropriate techniques.

Scanning electron microscopy (SEM) is used to observe the large aggregates (Figure 1.2(a)) and to quantify the so-called area ratio, that is, the ratio of the area of the aggregates, Aclay, to the area of the observed sample, Atotal.

In contrast, transmission electron microscopy (TEM) provides a detailed observation at the nanoscale, but over a very small area (Figure 1.2(b)). Generally, it does not make it possible to quantify the dispersion but permits us to check the local morphology and to confirm a possible exfoliation of the clay.

X-ray diffraction (XRD) is generally used to measure the interlamellar distance between platelets in the tactoids, and thus to quantify the intercalation of the OMMT by polymer chains. The absence of a diffraction peak may indicate an exfoliated structure, but it may also result from an insufficient amount of diffracting entities or an inappropriate orientation of the tactoids with respect to the incident beam.

In fact, one of the most sensitive techniques to quantify the dispersion state of nanocomposites is the rheometry in small amplitude oscillatory shear. Indeed, as shown in Figure 1.3, partial exfoliation of the platelets leads to a large increase of the complex viscosity at low frequency, explained by the interactions between the platelets and the development of a percolated network [13–16]. It has been shown that this behavior can be perfectly described by a Carreau–Yasuda law with a yield stress [5, 17]:

where ω is the angular frequency, σ0 is the apparent yield stress, η0 is the zero shear viscosity, λ is a characteristic time, a is the Yasuda parameter, and n is the dimensionless power law index. In fact, σ0 is a plateau modulus, but it is called yield stress by analogy with the behavior that could be observed in continuous shear. It was shown that, for a fixed composition, the level of exfoliation was directly related to the value of σ0 [5]. In the following, the apparent yield stress will be systematically used to quantify exfoliation.