Kinetic energy density and agglomerate abrasion rate during blending of agglomerates into powders

https://doi.org/10.1016/j.ejps.2011.11.012Get rights and content

Abstract

Problems related to the blending of a cohesive powder with a free flowing bulk powder are frequently encountered in the pharmaceutical industry. The cohesive powder often forms lumps or agglomerates which are not dispersed during the mixing process and are therefore detrimental to blend uniformity. Achieving sufficient blend uniformity requires that the blending conditions are able to break up agglomerates, which is often an abrasion process. This study was based on the assumption that the abrasion rate of agglomerates determines the required blending time.

It is shown that the kinetic energy density of the moving powder bed is a relevant parameter which correlates with the abrasion rate of agglomerates. However, aspects related to the strength of agglomerates should also be considered. For this reason the Stokes abrasion number (StAbr) has been defined. This parameter describes the ratio between the kinetic energy density of the moving powder bed and the work of fracture of the agglomerate.

The StAbr number is shown to predict the abrasion potential of agglomerates in the dry-mixing process. It appeared possible to include effects of filler particle size and impeller rotational rate into this concept. A clear relationship between abrasion rate of agglomerates and the value of StAbr was demonstrated.

Introduction

Mixing of powders is an important topic in industry (Muzzio et al., 2002), while it receives significant attention in academia as well (Donald and Roseman, 1962, Das Gupta et al., 1991, Sudah et al., 2002a, Sudah et al., 2002b). One of the challenges is that there are considerable differences in powder behavior. Too often it is assumed that an entire powder bed can be considered as either cohesive or non-cohesive. In practice however, a blend mostly consists of powders with different properties. In this respect, a cohesive powder that needs to be blended into a free-flowing bulk powder is an often encountered system. A typical example is a small amount of micronised drug that is blended in non-cohesive diluents (Nase et al., 2001, McCarthy, 2003, Li and McCarthy, 2003).

Cohesive powders tend to form agglomerates (Kuwagi and Horio, 2002). These need to be broken and dispersed to achieve blend uniformity. Ideally, the blending process is capable of breaking up the agglomerates without additional screening or application of shear-intensifying tools such as a chopper. In such situations the rate of agglomerate removal will be the factor that determines the required blending time. A (quantitative) understanding of the parameters that determine the rate of agglomerate removal is essential for example to enable process scale-up exercises and the implementation of process analytical technology (PAT) strategies.

Previous papers showed that the shear forces in a dry powder bed in a blender are very low: test particles with known yield strength did not deform at all (Tardos et al., 2004, Willemsz et al., 2010). It appeared that brittle test particles with known mechanical properties (the “brittle calibrated test particles”) reduced in size via abrasion. This abrasion process is typically characterized by a high frequency of impacts by filler particles on the surfaces of the agglomerates.

Mechanical properties of the agglomerates and certain process related parameters like particle size of the filler and the rotational rate of the impeller were found to affect the size reduction rate of the agglomerates (Willemsz et al., 2010). The velocity of the powder is an important parameter and it is often assumed that particle velocity in the blender is the same as, or at least proportional with the impeller tip speed when the blender is a convective blender (Iveson et al., 2001).

However, the actual particle velocity depends on many more parameters, such as filler particle size, the cohesiveness of the powder and relative fill volume of the mixer (Muguruma et al., 2000, Russell et al., 2003, Willemsz et al., 2011). Therefore direct measurement of (filler) particle velocity can be considered to provide more relevant data regarding the mixing process.

The aim of this paper was to investigate the effects of true particle velocity on agglomerate abrasion rate and establish the relevance of this parameter for blending processes. This knowledge can be applied in future process analytical technology exercises.

Section snippets

Materials

The materials used were microcrystalline cellulose (Avicel PH-101, FMC, Philadelphia, USA) and α-lactose monohydrate with different particle sizes (Pharmatose® 100, and 450 M from DMV Fonterra Excipients, Goch, Germany, with bulk densities of 750 and 470 kg/m3, respectively).

Manufacturing of model agglomerates (brittle calibrated test particles, bCTPs)

The model agglomerates or spherical brittle calibrated test particles (bCTPs) were prepared as described before, Willemsz et al. (2010). A selection of these bCTPs was used in the blending tests; the other part was used for

Blending intensity and abrasion rate of the test particles

Brittle calibrated test particles with different porosities were added to α-lactose monohydrate and blending tests were performed at different conditions. After a certain time interval, the bCTPs were collected and the mass reduction was monitored. The bCTPs were added to the powder again and mixing was continued. In that way the bCTP mass reduction over time was determined. Fig. 2 gives an example of the mass reduction of bCTPs over time. The mass reduction follows apparent first order

Conclusion

The abrasion of agglomerates during dry-mixing with two different filler particles has been investigated. This study reveals that the kinetic energy density of the moving powder bed (Wb) is a relevant parameter to explain abrasion rates of agglomerates. However, aspects related to the strength of agglomerates could not be included directly in this approach. For this reason, the Stokes abrasion number (StAbr) has been defined as the ratio between the kinetic energy density of the moving powder

Acknowledgment

This study was performed within the framework of Top Institute Pharma project number D6-203.

The authors would like to thank Bianca van der Werff for performing the statistical analyses and Uwe Thissen for his helpful comments on the statistical analyses.

Cited by (6)

View full text