(206d) Scale up of Anti-Solvent Crystallization Using in-Line Tools

Anti-solvent crystallization can be used as an alternative to cooling or evaporation for the separation and purification of solid product in the pharmaceutical industry. Addition of anti-solvent reduces the solubility of a solute in solution thereby facilitating the generation of supersaturation, the driving force for nucleation and growth. Some of the problems associated with anti-solvent crystallization are that it is poorly understood, can produce fine irregularly shaped crystals and is prone to the formation of polymorphs, solvates and hydrates.

The scale up of antisolvent crystallization is particularly difficult due to the importance of mixing in the system. An increase in the intensity of mixing can maintain a better crystal suspension, reduce the degree of settling and improve the impurity profile. However, the same increase may entrain air from the headspace, increase the level of attrition and cause shear damage to the crystals. Mixing intensity can also affect the growth and nucleation kinetics of the system. For anti-solvent crystallization additional challenges associated with mixing include the need to blend anti-solvent and solution phases sufficiently rapidly, so as to ensure a homogenous level of supersaturation throughout the vessel, and the change in volume over the course of a batch.

To characterize anti-solvent crystallization the use of in-line tools is vital. Focussed Beam Reflectance Measurement (FBRM) can be used to monitor key properties relating to the size and number of crystals in a crystallizer. Particle Vision and Measurement (PVM) is an in-line video camera that provides high quality pictures of the crystal suspension. Attenuated Total Reflectance Fourier Transform Infra Red Spectroscopy (ATR-FTIR) can be used to measure the concentration of both solute and anti-solvent in solution. These in-line tools eliminate the need to sample. A sample may not be representative and there is also the possibility that crystallization may continue between sampling and analysis. Additionally, many off-line methods need the sample to be altered in some way (i.e. dilution, sonication) and this may impact on effective analysis. All of these methods provide in-line real time data that can be combined to provide an extremely useful picture of what is happening inside a crystallizer at any point in the batch. With this information it is possible to assess the impact of many process variables and chose suitable operating parameters to provide the most suitable crystal product.

This work studies the unseeded semi-batch anti-solvent crystallization of benzoic acid from ethanol-water mixtures using water as an anti-solvent. Previous work in this group has established the feasibility of using FBRM and ATR-FTIR to measure solubility and the metastable zone width. The extent to which the anti-solvent addition and agitation conditions influence nucleation behaviour has also been indicated.

For the current work presented here these in-line tools are used to examine system behaviour at a variety of scales ranging from less thank 500 mL to up to 70 L. The relationship between both agitation and addition regimes (including the role of feed point location) and the resultant crystal size distribution is examined. The potential for a reliable scale up procedure is assessed, based on the results obtained.