Poly(ethylene terephthalate) (PET) maintains a good balance of thermal and mechanical properties compared to other conventional polymers and therefore occupies a privileged position in the plastics industry. Although it is well known that incorporating clay minerals into PET enhances its physical, mechanical, and barrier properties, industrial applications of PET/clay nanocomposites have been limited because clay disperses poorly in the PET matrix.

Smectite clays, such as montmorillonite and hectorite, have been used extensively in polymer/clay nanocomposites because of their high aspect ratio, plate morphology, natural abundance, and low cost.¹ These types of clays are naturally hydrophilic and contain Na⁺, K⁺, and Ca²⁺ ions in free state between layers. One way to improve the compatibility with the polymer matrix is to exchange these ions with organophilic surfactants. After modification with quaternary long-chain ammonium or phosphonium salts, they are called organoclays, which can be dispersed in polymer matrix.²

However, extrusion can cause thermal decomposition of commonly used alkyl ammonium ions in modified clays, and so also degrade the polymer matrix. ^3–5 As a result, this simple and versatile method, which plays a very important role in preparing PET/clay nanocomposites, has significant drawbacks (in addition to being inefficient).

In recent years, a new method called ‘thermokinetic mixing’ has begun to be employed for preparing polymer nanocomposites.^6–9 In a thermokinetic mixer, blades on a high-speed shaft accelerate the particles and impart them high kinetic energy, which is converted to thermal energy when they hit the chamber wall.⁷ This system, which requires no external heating, is specifically designed to handle difficult compounding and dispersion applications by completely heating, mixing, and compounding products within a few minutes. It has the advantage of short processing times and high shear rates.

We investigated the suitability of thermokinetic mixing for preparing PET/organoclay nanocomposites. We prepared nanocomposites by both thermokinetic mixing and melt extruding, and used scanning electron microscopy and x-ray diffraction to examine their dispersion properties.¹⁰ To our knowledge, this is the first time these methods have been compared for this nanocomposite system.

We compared two clay types, one of which (Resadiye, RK) is mined in Turkey and is not available commercially, while the other (Rockwood, RW) is commercially available. We first modified the raw clays using the synthesized quaternary ammonium salt consisting of tertiary amine ethoxylate and benzyl chloride and afterwards prepared PET/organoclay (96/4 w/w) nanocomposites by either a twin screw extruder and or a thermokinetic mixer. We name the resulting clays RKQR (quaternary Resadiye) and RWQR (quaternary Rockwood).

The d-spacing (the distance between clay platelets) observed by x-ray diffraction (XRD) analysis has been used to describe the nanoscale dispersion of the clay platelets in the polymer (see Figure 1). XRD showed there was no expansion in the d-spacing of the Resadiye organoclay (RKQR) independent of the processing type used (14.3Å for the 96/4 w/w PET/RKQR nanocomposite prepared on the extruder and 15.8Å for the nanocomposite processed on the thermokinetic mixer. The spacing for pure RKQR was 18Å). Extruded PET/RWQR (96/4 w/w) nanocomposite showed a large peak centered at ca. 2θ = 2.28^° (39.1Å) and a relatively smaller peak located at ca. 2θ = 5.42^° (16.3Å). In addition, the PET/RWQR (96/4 w/w) nanocomposite prepared on the thermokinetic mixer displayed larger d-spacings (44.4 and 15.1Å). Apart from that, we noticed a peak broadening and an intensity loss in the low scattering angle peak of the nanocomposite prepared on the thermokinetic mixer, indicating a decrease in the degree of coherent layer stacking (i.e., a more disordered system). It can be seen that the thermokinetic mixer produced more expanded and more homogeneous dispersed structures than the extrusion process did.

Scanning electron microscope (SEM) images allowed us to compare particle sizes for the different nanocomposites. RKQR had initial particles of larger than 20μm, while particles of RWQR were larger than 10μm: see Figure 2(a) and (b). Regardless of processing method, the clay particles substantially decreased in size. The extruded PET/RKQR (96/4 w/w) nanocomposite had clay domains that were mostly 3–4μm and partly 15–20μm in size: see Figure 2(c). Preparing the same nanocomposite (PET/RKQR) on the thermokinetic mixer reduced the particle sizes mostly to 0.1–1.5μm and partly to 4–5μm: see Figure 2(d). We saw the same trend in PET/RWQR nanocomposites. Extruded PET/RWQR (96/4 w/w) nanocomposite had clay domains whose sizes change between 3–5μm and also 8–10μm: see Figure 2(e). We observed clay particles 0.2-0.6μm and partly 1–2μm in size for the PET/RWQR nanocomposite prepared using the thermokinetic mixer: see Figure 2(f). Changing from extruder to thermokinetic mixer reduced the clay particle size substantially: processing on the thermokinetic mixer obviously dispersed the clay particles in the PET matrix phase better than the extrusion process. Results from SEM analyses were in excellent agreement with those obtained from XRD data.

In summary, we have shown that larger shear rates induced during thermokinetic mixing result in better dispersion of the clay in the matrix phase. Using the thermokinetic mixer also enables short processing times and so saves energy and time. We are currently working to modify different clay types that have high cation exchange capacities (>100meq/100g) and working on polymer/organoclay nanocomposites for a fully dispersed polymer nanocomposite system with enhanced properties and novel industrial applications.