Effect of mixing on high density polyethylene/clay nanocomposites mechanical properties and morphology

Polyethylene (PE) is the highest volume commodity polymer. The largest volume applications are flexible and rigid packaging. Key properties for packaging application materials include barrier, mechanical, and thermal properties. Material, energy and capital costs must be minimised, for economic and environmental reasons. For example, too little material may create shrinkage or warpage in the final product, or too long cooling time will decrease productivity. Hence the design and the manufacturing processing must be optimized against constraints.

PE/clay nanocomposites have been studied extensively over the last two decades as nanoclay improves many key packaging properties. Melt compounding is a popular technique to produce nanocomposites because of the ready availability of equipment and its low environmental impact. Many studies investigated PE nanocomposite properties in terms of material variables (such as polymer and filler type, and filler and compatibiliser level). However, only a limited number of studies have investigated the effects of processing conditions. Most of these focused on extruders (with various types of extruder and screw designs) as these are commonly used in PE compounding or applications. Very few have investigated internal mixers, as these are not commonly used in industry. However, they have key advantages in research work on a laboratory scale, as volumes are small, process variables are independently controlled, and the rotors are easily changed.

This research explores the effect of processing conditions (temperature, rotor speed and mixing time) on mechanical properties and morphology of high density PE (HDPE)/clay nanocomposites prepared in a double-rotor internal batch mixer. The nanocomposite formulation was HDPE blow moulding grade (HD5148 MFI 0.83 g/10 mins from Qenos), 2 wt% organoclay (Cloisite 93A from Southern Clay Products), 5 wt% HDPE-g-ma compatibiliser (Polybond® 3009 from Chemtura) and 0.1 wt% processing stabilisers (1:1 Songnox™ 1010 and Songnox™ 1680 from SunAce). Three different rotor designs (Roller, Banbury, and Sigma) were trialled. A three level Box-Behnken experimental design was used to produce samples at various process conditions for each rotor design for the one nanocomposite formulation. Mechanical properties including modulus, strength and impact were measured. Empirical models for secant modulus were developed and used to predict optimum or best process settings for each rotor design. Morphology of samples was studied using Small Angle X-ray Scattering (SAXS), Scanning Electron Microscopy (SEM), and Transmission Electron Microscopy (TEM). Mechanical properties, significant terms in the empirical models, and morphology analyses were compared to identify the dominant dispersion mechanism for each rotor design.

All three rotors improved nanocomposite modulus compared to natural PE, with improvements ranging from 2 to 27%, with an average improvement of 15%. All rotors produced similar ranges of improvement. Results for strength were quite different. Some samples improved significantly with filler addition, while some deteriorated significantly. Results ranged from a decrease of -10 (for the Sigma rotor) to an improvement of 9% (for the Banbury rotor) with an average change of 2%. The Banbury rotor had significantly better results than the Roller or Sigma rotors. Impact strength decreased significantly for all rotors, by an average of 70%. The Roller rotor had the worst results. These results are within published ranges from other studies using both extruders and internal mixers.

There were also significant differences in modulus and strength between best and worst samples for all rotors. This shows that both rotor and process conditions had significant effects on mechanical properties. An optimum set of process conditions to maximize modulus was identified for the Roller rotor (close to the mid-point), but not for the other two rotors. This suggests the optimums for the other two rotors were outside the experimental range of process conditions. Best conditions for the Banbury and Sigma rotors were identified as the boundary conditions. These settings were low, high, low settings for the Banbury, and low, low, high for the Sigma rotor. The effect of variables, and the optimum or best conditions, depended on the rotor type.

The morphology results showed a hybrid composite structure, with some exfoliation, some intercalation, as well as some microstructure (agglomerates). There was a weak correlation between better mechanical properties and morphology with more exfoliation and thinner intercalated particles. The Banbury had the most exfoliated or thin intercalated particles.

There are two main mechanisms for dispersion of filler in melt mixing. Shear forces break the platey filler into thinner particles, and diffusion causes polymer to enter the clay galleries, leading to gallery expansion and delamination. The dependence of optimum process conditions on rotor type suggests there is a different interplay of dispersion mechanisms for each rotor. The data sets from different analyses were compared to draw conclusions about the dispersion mechanisms present or dominant for each rotor.

The modulus was a maximum at high torque for the Roller and Banbury rotors (indicating high shear), and medium torque for the Sigma rotor (indicating medium shear). The statistical analysis of DoE data showed that shear was dominant in the Roller rotor, while both shear and diffusion were present in the Banbury but shear was dominant, and conversely, diffusion was more dominant for the Sigma samples. The Response optimizer analysis showed the optimum or best conditions to achieve the highest modulus were medium shear and diffusion for the Roller, high shear and low diffusion for the Banbury, and low shear and high diffusion for the Sigma roller. While there is some disagreement between the data sets, overall it appears that shear is the dominant mechanisms in the Roller and Banbury rotor samples, and diffusion is dominant in the Sigma rotor samples.

Overall the Banbury rotor is recommended as the best rotor, as it produced similar impact and better modulus and strength than the other rotors. Its best process conditions were at low temperatures and short mixing times. This would be attractive commercially as energy costs would be lower and productivity higher, generating both economic and environmental benefits.

In conclusion, an internal mixer is a useful laboratory scale equipment to evaluate the sensitivity of a nanocomposite to process variables and equipment design and to estimate the maximum properties that can be achieved.