S.L. Rough

Effects of liquid phase migration on extrusion of microcrystalline cellulose pastes.

The behaviour of water-based microcrystalline cellulose pastes undergoing ram extrusion has been investigated. Factors affecting the redistribution of water within the extruding paste and the upstream barrel compact, such as the initial water content, extrusion rate and die geometry, have been considered. The rates of dewatering for these given systems were characterised by the gradients of the extrusion pressure-ram displacement profiles. A linear relationship between the pressure-displacement gradient and the inverse square root of the paste velocity was obtained for a given paste and extrusion geometry. At velocities where water migration was significant, the extrudate was found to have a higher water content than that of the paste in the barrel at any given time; both the extrudate and the barrel paste decreased in water content with increasing ram displacement. Spheronisation of extrudate samples has shown that the redistribution of liquid during extrusion is an important factor affecting the quality of the spheres. A paste flow model, incorporating pseudo-plastic and shear deformation terms, was used to predict the change in extrusion pressures caused by liquid phase migration. The model parameters were obtained as functions of water content and gave good agreement with the experimental extrusion profiles.

A model describing liquid phase migration within an extruding microcrystalline cellulose paste

A quantitative, one-dimensional model has been developed to predict liquid phase migration, and hence the redistribution of water, within a microcrystalline cellulose (MCC) paste undergoing ram extrusion. The process features an initial compaction stage, followed by paste convection after the onset of flow. Compaction is treated similarly to the consolidation of soils, with Darcy-type liquid flow behaviour and stresses estimated using the Janssen–Walker analysis. The forces required in the convection stage are calculated using the modified Benbow–Bridgwater model, incorporating water-content-dependent pseudo-plastic and shear deformation terms. The model estimates extrusion pressure–time and liquid content distributions. Factors affecting liquid phase migration, such as the extrusion rate and die geometry, have been considered. The model shows reasonable comparison with experiment and mimics the observed trends. A dimensionless number has been defined, being the ratio of the maximum initial liquid phase velocity relative to the solids phase, to the absolute solids velocity, which suggests a criterion for the occurrence of significant liquid phase migration. The model does not incorporate pore suction pressure effects, which are likely to be important in other paste systems.

Extrusion-spheronisation of highly loaded 5-ASA multiparticulate dosage forms.

The aim of the current work was to develop an extrusion-spheronisation (E-S) route to manufacture pellets with a high loading (≥90wt%) of 5-aminosalicylic acid (5-ASA). Ram extrusion studies, supported by centrifuge testing, were employed to investigate the effect of the chemical (acidity) and physical (particle size and shape) characteristics of 5-ASA on the ability of microcrystalline cellulose (MCC)-based pastes to retain water when subjected to pressure. Liquid phase migration (LPM) within the paste during the extrusion, and hence variation in water content of extrudates and reproducibility of the final E-S product, was generally observed. The extent of LPM was found to be related to both the drug loading and its physical properties, most notably the particle shape (needle-like). A reduction in particle size, combined with a change in the shape of the 5-ASA particles, allowed LPM to be reduced considerably or eliminated. The performance of colloidal grades of MCC (Avicel RC591 and CL611) as alternative extrusion aids to the standard Avicel PH101 was also investigated: these proved to be superior aids for the highly loaded 5-ASA pastes as their greater water retention capacity mitigated LPM. Combining these results yielded a route for manufacturing pellets with 5-ASA loading ≥90wt%.

A comparison of ram extrusion by single-holed and multi-holed dies for extrusion–spheronisation of microcrystalline-based pastes

The use of multi-holed dies as an alternative to single-holed dies for generating extrudates for spheronisation was investigated both in terms of extrusion and spheronisation performance. A model 45wt% microcrystalline cellulose (MCC)/water paste was employed in ram extrusion tests with square-ended dies with 1, 6, 33 and 137 holes, all of diameter 1mm and length 2mm. The extrudates generated using the multi-holed dies yielded pellets with comparable sphericity to those using the single-holed die. Multi-holed dies could also be operated with lower paste flow rates before encountering liquid phase migration (LPM). The characteristic processing velocity for the onset of LPM was determined for each die configuration and supported the hypothesis that LPM was caused by suction effects. A simple model of the flow pattern in a lab-scale Fuji-Paudal frontal screen extruder is presented which yields estimates of velocities and shear rates involved in these devices. The pressure required to extrude the paste through multi-holed dies was compared with the model proposed by Benbow and co-workers. The paste rheology was characterised using the Benbow-Bridgwater approach, employing 1, 2 and 3mm diameter dies of various lengths. The Benbow et al. model under-predicted the observed extrusion pressure, which was attributed to its failure to account for the redundant work contribution in these complex flows.

Tapping characterisation of high shear mixer agglomerates made with ultra-high viscosity binders

Abstract The time-dependent consistency regimes produced during high shear mixing of zeolite powder and an ultra-high viscosity binder (linear alkylbenzene sulphonate (LAS) paste) were quantitatively described by the bulk aerated and tapped densities, which were determined using both hand and automated tapping techniques. The Hausner ratio was calculated, providing information on the inter-granular friction and cohesivity of the granular bulk as a function of mixing time. An agglomeration mechanism is proposed based upon the trend in Hausner ratio, which was confirmed by optical micrographs and granule size distribution data. The suggested mechanism comprises layering of zeolite particles onto an LAS paste core, breakage of powder-coated paste granules, micro-mixing of the granules, granule growth via coalescence, and finally granule consolidation. The bulk tapping data were analysed using the Kawakita equation and a logarithmic compaction approach. Three distinct compaction regions were identified with the latter analysis, the first of which was related to weak agglomerate break up, and the second and third to granule movement involving elastic and plastic granule deformations respectively during rearrangement. A variety of bulk compaction parameters were obtained, and their variation with mixing time is discussed. At least 10 times as many automated taps were required to reach the final tapped density in comparison to the hand-tapping procedure, and the final density was always lower than that obtained via hand-tapping. When the automated tapping data were scaled in terms of total number of taps and analysed, the parameters describing the bulk compressibility showed similar trends to those obtained from the hand-tapping procedure.

Effect of solids formulation on the manufacture of high shear mixer agglomerates

Abstract The strength and morphology of granules generated by high shear agglomeration of powder with a highly viscous surfactant paste were investigated as a function of mixing time and powder particle size. The formulations featured a fine powder (zeolite or calcium carbonate) and linear alkylbenzene sulfonate (LAS) in a ratio of 1.7 : 1 by volume. During agglomeration, the solid material progressed through a number of consistency regimes up to the formation of a smooth dough. Differences in calcium carbonate (calcite) particle size had a significant effect on the agglomeration behavior, in that an increase in the mean diameter reduced the time required to form the dough state. Micrographic visualization, coupled with bulk density and Hausner ratio measurements, suggested that all the systems studied followed the same agglomeration mechanisms up to a certain point, albeit at different rates. The LAS-calcite mixtures appeared to be more compressible than zeolite mixes with similar particle size, although the overall trends in the compaction parameters were similar. None of the LAS-calcite mixtures attained an intermediate ‘crumble-agglomerate’ regime observed with the zeolite, which is attributed to the extensive hydrate-forming properties of the latter.

A regime map for stages in high shear mixer agglomeration using ultra-high viscosity binders

Abstract The porosities and saturation levels of high shear mixer agglomerates, containing zeolite powder bound with a highly viscous surfactant paste [linear alkylbenzene sulphonate (LAS)], were determined as a function of mixing time by performing a series of simple liquid immersion experiments. The results validated, to a certain extent, the use of the secondary break point ( P 2 ) obtained from a logarithmic tapping analysis of the bulk material as a measure of the individual granule plastic yield during tapping. The bulk material tapping parameter P 2 was also used to represent the dynamic yield stress ( Y g ) of the granules. This enabled the determination of the Stokes deformation number (St def = p g U b c /2Y g ) for the granules, which was employed on an agglomeration regime map as described by Iveson et al. Powder Technol. 117, 83-97 (2001). It was found that the variation in granule morphology with time was reflected accurately by the parameter trajectory on the regime map. The results were in accord with the previously reported agglomeration mechanism for this system.

In situ measurements of porosities and permeabilities of alumina pastes

Abstract The compaction behaviour of various water-based pastes prepared from mixtures of different sized α-alumina powders was investigated by determining the paste porosity as a function of the applied total stress, and the results confirmed a logarithmic stress–porosity relationship for a given paste. The corresponding paste permeabilities were measured using a test cell in which the local solid stresses on the matrix and the liquid pore pressures within the paste were independently controlled. The data were interpreted using a Darcys law approach commonly used in soil mechanics, and the generated permeability coefficients were shown to vary exponentially with the porosity of a specified paste. The experimental permeabilities were compared with those predicted from a Carman–Kozeny analysis, and the results suggest that entrapped air (unsaturation), as well as particle size, shape, size distribution and residual surface charging, are important factors that influence the permeability of a consolidated paste.

Chapter 3 Extrusion—spheronisation

Publisher Summary Extrusion–spheronisation (E–S) is used to manufacture spherical or cylindrical pellets by extruding a semi-solid wet powder mass through a single die, a series of dies, or a screen featuring many holes, and then breaking up and rounding the extrudate on a rotating friction plate. E–S is also known as extrusion–marumerisation, where a “Marumeriser” is the name for the spheroniser originating from the Japanese word for “pellet,” which appeared on the original Fuji–Paudal patent for their device. When optimized, the process yields a dense and quite spherical product with good integrity. These properties can be tailored via coating, which is amenable to the spherical shape of the granules. E–S is used widely in the manufacture of controlled release pharmaceuticals, polymers, detergents, fertilizers, and herbicides (particularly in water-dispersible granule form). The spheronisation step is often omitted when cylindrical pellets are the desired product form, as in fertilizer and herbicide manufacture where the extrudates can be broken up in a fluidized bed drier to manufacture products with the required size. This chapter focuses on the use of E–S in pharmaceutical granulation and related applications, where the material extruded is a wet powder mass and extrudate diameters are in the range of 500–1500 μm. In practice, the largest particle sizes achievable are ∼5 mm. The free-flowing granules obtained by E–S are used in tabletting or in filling capsules for oral dosage forms. This particular process is favored by the pharmaceutical industry because pellets formed with a small amount of active ingredient show a slower release profile compared to those made by other techniques.

A novel lab-scale screen extruder for studying extrusion-spheronisation.

A novel apparatus, the laboratory roller screen extruder (termed the LRS), was developed to replicate key aspects of the geometry and shear strain rates generated near the screen of industrial screen extruders. The configuration of the LRS is reported alongside a commissioning study employing a cohesive 45 wt% water/microcrystalline cellulose paste. The key operating parameters which controlled the extrudate mass flowrate, force on the screen and roller torque were (i) the size of the gap between the top of the roller blade and the screen, and (ii) the roller rotational speed. The data suggest that the apparent shear rate, based on the blade-screen clearance, provides a quantitative criterion for scale-up. The amount of screen flex showed good agreement with a simple bending deformation model. Spheronisation of the extrudates gave pellets with a narrow size distribution and acceptable sphericity which would be acceptable for capsule filling. Optimisation of the pellet shape was not performed. The results indicate that the LRS can be used to assess formulations for industrial screen extrusion-spheronisation.