The breakage of needle-shaped crystals under pressure filtration

• Main Author: Shier, Andrew Philip
• Other Authors: Muller, F. L.
• Published: University of Leeds 2017
• Subjects: 660
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.714302

Pressure filtration is one of the most common solid-liquid separation techniques in use in the fine chemical and pharmaceutical industry. Typically pressure filtrations are conducted within the same unit operation as agitated drying, thus the opportunity to study the individual effects of filtration are often negated. Recent findings have shown pressure filtration to be a major cause of particle breakage, which has consequences for product quality, bioavailability and downstream operations like formulation. A new method for the investigation of the effect of hydrodynamic stress on static crystals has been developed by continuously recirculating fluid through a bed of freshly crystallised particles. The method is compared to traditional batch pressure filtration tests with good correlation between results. Levels of breakage of needle-like glutamic acid crystals are found to increase with rises in pressure and fluid flow rate. Particle size and shape data obtained from optical microscopy is critically analysed in order to determine the type and extent of crystal breakage. Automated methods have also been developed to filter particle data in order to provide a more representative measurement. Small-scale centrifugal testing of particle beds has been conducted with an analytical photo-centrifuge. Needle-like crystal beds are found to consolidate to greater extents than more rounded particles, and also show evidence of inelastic behaviour. The more isometric particles conversely display elastic behaviour in response to the application of stress. Modelling of particle data has simulated algorithmic breakage events to predict the breakage observed under pressure filtration, with good agreement between modelled and experimental particle size distributions. The structure of particle beds has also been modelled by generation of simulated needle-beds from particle length data.