Mechanical Deagglomeration

Intimate contacts (up to almost the molecular level) of the nanoparticles with the polymeric matrix are a prerequisite for the design of well-functioning nanocomposites. To satisfy this condition, a thorough homogenization of the components must be obtained. Winkler et al. [125] developed a theory for the deagglomeration of ultrafine particle clusters in fluid systems, using the mechanical forces in a dispersion machinery. They showed that the probability that filler particles are dispersed in a dispersion medium, pT, is the product of two separate, partial probabilities: pt, the probability with which the agglomerates encounter a potentially active dispersion site, and pE, the probability that enough energy per unit volume (energy density) is available to overcome the forces holding the agglomerates together. By applying their theory to dispersion experiments carried out in a bead mill and a high-speed impeller [126], the effective volumes per second (in which a dispersion of the agglomerates could take place) were found to be on the order of 1% of the total volume, and the transfer efficiency of the energy into the agglomerates was approximately 1% as well. According to these findings, it was suggested that the key to success for dispersing particles is to guide their agglomerates in a constrained path through well-defined zones of suitable energy densities. In addition, Winkler and Dulog found that the deagglomeration in most dispersion machinery is governed by the shear stress within the mill base [127]. This means that smearing-type mechanisms are far more important than smashing-type stresses for dispersions.

From a practical point of view, compounding polymers in their molten state with additives is generally achieved by various devices, like two-roll mills, internal mixers, single-screw extruders, and twin-screw extruders [128]. Two-roll mills and internal mixers are not suitable for production on an industrial scale. Single-screw extruders cannot provide sufficient homogenization. Corotating twin-screw extruders are therefore the most popular machines because of their continuous operation and high productivity. Both the screw and the barrel of the extruders are constructed from segments and can be assembled to produce the necessary conditions for efficient homogenization.

Fundamental dispersion studies have identified some useful rules for optimum compounder performance, for example [129-131]:

• Since dispersion is a phase reduction process, it pays to generate a well-mixed feed in the feeder train or feed section of the compounder.

• High stresses applied during melting are important and friction dependent, and care should be taken to reserve lubricating fluids, powders, and low melting solids for post-addition.

• Compatibilizers reduce the interfacial tension and result in dramatically improved dispersions.

• Solid particles in polymer melts break up like a fatigue failure, which accounts for the fact that there must be a minimum residence time established to achieve dispersion in addition to a minimum shear stress. Above the minimum residence time and minimum shear stress, the shear and time values are interchangeable in achieving the same dispersion effect.

• Shear yield stress of particle filled polymer melts increases with decreasing particle size.

When a twin-screw extruder is used, the effects of extrusion processing conditions on the final state of dispersion should be known in advance, so as to control the formation of agglomerates. Although most of the available literature is actually based on the approaches developed for microcomposites, the original principles stated above might also be applicable to nanocomposites. This is due to the facts that (a) the size of the nanoparticle agglomerates also falls in to the micrometer regime, and (b) the initial particle size gives only slight differences in dispersion, as demonstrated by Ess and Hornsby in their paper dealing with CaCO3 particles varying in diameter from 80 nm to 3 pm [132].

Ess and Hornsby discussed the dispersive mixture quality of CaCO3-filled PP and PA 66 composites compounded by an intermeshing corotating twin-screw extruder. For the system under investigation, the zone of polymer fusion at the end of the first stage of the screw profile was identified as having a crucial influence on enhancing the filler dispersion because of the high levels of shear developed in this region. Furthermore, there was some evidence of filler agglomeration due to compactive forces in the solid bed formed before polymer melting. Compounding experiments, using pellets of CaCO3 preconditioned under various pressure, temperature, and moisture conditions, indicated that too high values of these quantities gave a marked reduction in the degree of filler dispersion. These results therefore highlight the importance of controlling these variables (for example, in premix-ing operations) for an optimized filler dispersion.

Recently, Bories et al. made a study of this aspect by using polypropylene (PP) and CaCO3 (0.7 pm) as the experimental materials [133]. They found that reducing the extrusion flow rate or increasing the rotation speed and barrel temperature can yield improved dispersion. Furthermore, the main concerns of their work were focused on twin-screw extruder design and its relation to particle agglomerate dispersion. Four different screw geometries were employed; each one had two separate kneading sections, referred to as the melting and mixing sections (Table 6). To evaluate the conditions that minimize the relative agglomerate fraction, the dispersion efficiency of the four different screw configurations was compared with the use of either hopper feeding or mid-extruder feeding of CaCO3. As shown in Figure 20a, where the screw configurations have a common severe melting section and differ in the mixing section, better dispersions are achieved when PP and CaCO3 are fed together at the hopper. It is possible that a distribution of the mineral can be achieved even during solid conveying before PP melting, leading to low levels of agglomeration. When feeding takes place at the primary hopper, the state of dispersion does not change significantly whether severe or gentle mixing sections are used downstream, indicating that the agglomeration/dispersion process is completed after the melting zone. The conclusions are quite different when feeding of CaCO3

Table 6. Summary of screw configuration severity.

Melting section

Mixing section

Screw configuration

(first mixing section)

(second mixing section)

1

+

-

2

+

+

3

-

+

4

-

-

Note: + and — represent severe and gentle, respectively. Source: Reprinted with permission from [133], M. Bories et al., Int. Polym. Process. 14, 234 (1999). © 1999 Carl Hanser Verlag.

Note: + and — represent severe and gentle, respectively. Source: Reprinted with permission from [133], M. Bories et al., Int. Polym. Process. 14, 234 (1999). © 1999 Carl Hanser Verlag.

ra oc

I I Screw configuration 1 (+/-) V//A Screw configuration 2 (+/+)

Hopper

Run/Screw speed [min 1]

TO TO

I I Screw configuration 3 (-/+) Screw configuration 4 (-/-)

Hopper

Run/Screw speed [min 1]

Figure 20. Relative agglomerate volume fraction obtained with different screw configurations for a 60 wt.% CaCO3/PP composite (barrel temperature 200 °C). Reprinted with permission from [133], M. Bories et al., Int. Polym. Process. 14, 234 (1999). © 1999, Carl Hanser Verlag.

takes place at the mid-extruder. In this case, significantly higher agglomerate volume fractions are observed for the severe mixing zone, which includes a reverse element that is used to fill the mixing zone. This may cause additional agglomeration, since large undistributed mineral pockets can be rapidly compressed into agglomerates in this section.

When the screws have a common gentle melting section and differ in the second mixing zone severity (Fig. 20b), the best results are obtained by mid-extruder feeding. For both hopper and mid-extruder feeding, the best improvements are achieved with configuration 4, which has gentle melting and mixing zones. The use of a severe mixing section is detrimental for the dispersion in both configurations, leading to higher agglomeration levels. The effect of the second mixing zone is also observed, even when feeding of the CaCO3 takes place in the primary hopper. This indicates that the dispersion process is not completed at the end of the melting zone, contrary to the observations with the severe melting section. The PP, being only partially molten in the melting zone, is still highly viscous at the mid-extruder position. This seems to be a good compromise. The polymer has sufficient fluidity to help in the wetting of the mineral while it is still viscous enough to generate high hydrodynamic stresses. Best overall results are thus obtained when the mineral is fed at mid-extruder with the gentlest configuration.

Was this article helpful?

0 0

Post a comment