Particle production by supercritical antisolvent processing techniques

Abstract

This thesis discusses particle production by supercritical antisolvent processing (SAS) techniques by looking the fundamentals and applications of the method with some case studies. The final aim of this work is however to consider the SAS particle production process feasibility. In the process studied the solid is dissolved in a conventional solvent and the solution is sprayed continuously through a nozzle into the subcritical or supercritical fluid. The dispersion of solution in the fluid leads to an expansion of the droplets and at the same time an extraction of the liquid into the fluid occurs. The solvent power of the conventional solvent decreases dramatically and supersaturation leads to the precipitation of particles.

A static variable volume view cell (VVV-cell) is a useful and fast way to find out the appropriate combinations of the solvent and the gaseous antisolvent for a given solid. Often pharmaceutical materials are expensive and not available in large amounts, which makes it impossible to make phase separation studies with laboratory or production scale SAS equipment. However in VVV-cell experiments it is possible to use small amounts of materials and fast examine the influence of recrystallization temperature, pressure and concentration. But it is not possible to conclude by using only VVV-cell experiments, what kind of particles (size distribution, crystal habit and morphology) there will be produced in SAS process, because of the different formation dynamics and residence times.

The present study showed that in the supercritical state the variables, such as density of CO2 and temperature, have a greater effect on the particle size than the model of droplets predicts. The liquid side mass transfer seems to control the studied polymer material particle size. In poly(L-lactic acid) particle formation with dichloromethane solvent and CO2 antisolvent by SAS technique it is advantageous to use low temperature and high pressure, in which conditions the mass transfer effect and volumetric expansion of droplets to produce high supersaturation will be favourable. In SAS process it is not possible to influence the initial droplet size by varying process variables (temperature, pressure and flow rate) in a typical operating range with a similar nozzle. Therefore the mass transfer coefficient of the liquid phase should be maximized to produce a high supersaturation fast when a small particle size is needed.

Supercritical fluid technology is considered to be an innovative and promising way to design particles. In this thesis the applicability of two special supercritical precipitation techniques was studied. In the first case the results show, that it is possible to produce completely amorphous particles by spraying a methanol solution of sodium cromoglycate into supercritical carbon dioxide. The most significant parameter affecting the crystallinity was the residual methanol concentration in the particles. In the second case the results show, that cholesterol can be selectively extracted / crystallized at high purity, from a one phase mixture which contains lipids and cholesterol dissolved in pressured CO2. In this simple cholesterol production process no liquid solvents are needed. The egg yolk phospholipid was the most suitable raw material for the tested method. Because the concentration of cholesterol in the fluid stream is low, this method is not economically viable in industrial scale.

The work clearly demonstrated that to recrystallize fine particles with SAS techniques in an industrial scale the price of the products must be high. These products would be chemical intermediates, biological and pharmaceutical compounds. On the other hand, if the compound properties are waxy or soft or the products are thermally unstable compounds, it may be feasible to use antisolvent formation techniques also for less expensive products, if the method makes the production possible. The manufacturing cost of SAS process is capital intensive. An estimated manufacturing cost for a new GMP plant is around 50-300 Euro/kg product without a feedstock price. This is for a 4000 to 8000 kg/year production rate and 5-10 wt% feed concentration of the starting material in an organic solvent. An effective way to decrease the manufacturing cost is to increase the raw material concentration in solvent. It is favourable to design a process for production rate over 2000-3000 kg/year and to use over 5 wt% feed concentration. Below those values the manufacturing cost increases dramatically.