Das Buch |
Uniaxially compacted tablets are one of the most preferred drug dosage forms. A deeper understanding of the structure-property relationships can put some light on the final performance of the entire tableting operation. Chipping and abrasion in finishing operations, such as drum coating, are undesirable phenomena that might convey to dust formation and the rejection of an entire batch. The extent of each individual failure mechanism is dependent on formulation parameters (elasto-plastic mechanical properties of tablet components and compaction parameters) as well as geometry, stress field, and inertia within subsequent process steps. To provide a comprehensive analysis of the role of single particle properties and manufacturing conditions on the mechanical stability of a tablet, and to establish a predictive numerical tool of the overall process performance by the Discrete Element Method (DEM), a four phases study is hereby presented. First, for a selected number of pharmaceutical excipients, a direct determination of particle properties (e.g. size distribution, true density, module of elasticity or surface energy) is performed. This provides a clear characterisation of the shape, morphology and densification mechanisms on the particle scale and permits the calibration of DEM input parameters.
Second, the effects of compaction stress as well as the inherent anisotropy is analysed. Powder densification through uniaxial compaction is governed by a number of simultaneous processes, taking place on a reduced time as result of the stress gradients within the packing, as well as the frictional and adhesive forces between the powder and the die walls. Because of that, a density and stiffness anisotropy is developed across the axial and radial directions. In order to quantify structural differences, mechanical strength, specific surface area and friability resistance have been extracted for a range of compaction conditions. Micro-indentation has been applied to assess and quantify the variation of the module of elasticity throughout
the surface of tablets.
Third, a multilevel DEM modelling approach accounting for the micro-level (single particle, intra-tablet interactions) and the bulk level (inter-tablet collisions) is depicted. The elasto-plastic cohesive hysteretic Edinburgh Contact Model (ECM) is employed to determine the extent of particle rearrangement and the degree of anisotropy for uniaxial compaction by a DEM simulation.
Therefore, a calibration strategy for all required collisional, frictional or cohesion input parameters is presented. Afterwards, the numerical compaction for a range of compaction pressures and different particle dispersities (mono- and polydisperse) has been conducted. The anisotropic behaviour of compaction has been analysed by computing averaged quantities such as the deviatoric
term and the maximum difference of characteristic roots of the fabric tensor.
Fourth, on the bulk level a dual approach by means of DEM and Population Balance Modelling (PBM) is suggested. DEM provides the description of kinetics and allows the extraction of collisional energy distributions. In this case, the primary shape of the tablets is represented by a multi-sphere approach, a rigid assembly of clumped spheres to resemble non-spherical shape. Then, after solving the PBM equations, the mass loss
of tablets according to their mechanical resistance, as well as the fragment size distribution progeny can be computed. The novelty of the presented approach is the bidirectional coupling: tablet and fragment distribution masses are dynamically updated by the particle replacement method according to the stress history of the system.
|