No abstract

There has recently been an increased interest in predicting the tensile strength of binary tablets from the properties of the individual components. In this paper, measurements are reported for tensile strength of tablets compressed from single-component and binary powder mixtures of lactose with microcrystalline cellulose (MCC), and lactose with two types of silicified microcrystalline cellulose (SMCC and SMCC-HD), which are different in compressibility. Measurements show the tensile strength increases with the relative density for single powders, and both with the relative density and the mass fraction of cellulose in the mixtures. It was also observed, for binary mixtures compacted at 50 and 150 MPa, that there was a slight variation in porosity with the mass fraction of celluloses. The predictions of the tensile strength of binary tablets from the characteristics of the single-components was analysed with the extended Ryshkewitch-Duckworth model by assuming both linear and power law mixing rules for the determination of the parameters "tensile strength at zero porosity and bonding capacity constant". As consequence, four models were analysed and compared with measurements using criteria based on the standard deviation from the mean values. Results showed a good prediction using a linear mixing rule combined with the power law. However, as the predictions of these models depend on the powders and the porosity range for the characterization of single-components, none of them can be systematically considered as being the best to predict binary behaviour from data for individual powders.

Homogeneity of properties over the width of strips produced by roll compaction of microcrystalline cellulose powder (MCC) has been examined by light transmission through the compact, by measurements of the porosity distributions and by three-dimensional finite element modeling. Light transmission through compacts revealed periodic heterogeneity in the form of alternate dark and light zones. The period seems to be connected to the geometry of the screw and independent on the feed screw velocity which was varied with the roll speed with a constant ratio. Measurements of the porosity of samples cut from the compacted strip show heterogeneity of the density over the width of strips with a higher density in the centre of the strip and a lower density on the sides. These two techniques clearly showed the heterogeneous behavior across the width of the compacted strip of MCC. However, the light zones (respectively dark zones) did not correspond to the lower porosity zones (respectively higher porosity zones). Three-dimensional finite element modeling (FEM) of roll compaction of powders was conducted with two inlet feed conditions: constant feed pressure and constant feed velocity. Results of the simulations using the constant feed pressure show a uniform maximum principal stress and density across the width of the strip. When a constant inlet feed velocity is assumed the maximum principal stress over the strip width was higher at the centre of the strip and decreases to the sides. This profile also corresponds to the density profile over the width of the strip. In this case, the predicted results present a similar tendency to that found by mercury intrusion porosimetery and are in agreement with the measured bulk density of strips produced with different roll speeds.

International audiencePermeating air is known to have a negative impact on the roller compaction process, because the feed is destabilized by the flow of escaping gas, causing pressure to build-up and potentially damage compacts at release. Airflow during powder roller compaction and its effect on the rolling process are investigated in the rolling direction (1D), using an extension of the Johanson model for the solid. Fluid transport obeys Darcy's law, with permeability being a function of both material density and particle size, through the Kozeny-Carman relationship. In this modeling, the effect of the air pressure on the solid is neglected in the compaction zone. Assuming air at atmospheric pressure at the feeding angle and ignoring airflow through the gap, predictions of air pressure as a function of the rolling angle for bentonite material powder are presented and discussed. Results suggest the existence of two different stability zones within the operating conditions, where industrial systems could function without being affected by airflow effects. The model highlights the importance of the permeability/rotation speed ratio, which governs the proportion of air trapped in the compacts to the portion evacuated through the feed. We also investigate the effect of particle fragmentation during the rolling process. Finally, we provide guidelines for the scale-up of roller presses subjected to air flow issues, through a study of the effect of the system dimensions and rotation speed on the elimination of air. In spite of the lack of available experimental data, this model allows for a better understanding of how air escapes by diffusing through the material during the rolling process, and opens interesting perspectives for the mitigation of its effect on the process

Dry granulation by roll compaction is a continuum manufacturing process to produce granules with improved flowability which can further be easily used in tableting process. However, the granules are non-homogeneous in density and have non-spherical shapes which impact their densification behaviour during die-compaction. The aim of this study was to investigate both the densification mechanism and the failure strength of granules of microcrystalline cellulose (MCC) and mannitol using Cooper-Eaton and Adams models. For both materials, the Cooper-Eaton approach led to the quantification of fractional volume compaction by particle rearrangement and by plastic deformation respectively to explain the difference in densification behaviour of raw material and granules. Moreover, the model showed its ability to capture the effect of granule density and granule sizes and to differentiate the densification mechanisms of MCC as a plastic material and mannitol as a brittle material. The Adams model was used to compute the failure strength of single granule from in-die compression data. The obtained results of the granules were in the range [0.6-1.43 MPa]. However, regarding the effect of granule density, the model showed mixed results indicating that the model is not representative of the studied granules which are not spherical and have a relatively wide range of sizes, nevertheless, the model was derived for near spherical particles with a narrow size distribution.