Optimized Operation of Batch Crystallization

On-line Monitoring and Optimization of Batch Cooling Crystallization targeting an optimal Particle Size Distribution (PSD)

Batch Crystallization is a widely used production step for high-value specialty chemicals. Among other things, the quality of the products depends mainly on crystal purity, crystal size, product filterability and dry solid flow properties [1]. The correspondent operation goals can generally be expressed in terms of the particle size distribution (PSD), i.e. a PSD can be determined that fits best the operation objectives.
With a sufficiently accurate model, one can optimize a priori the cooling profile in order to produce optimal product properties [1]. However, model uncertainties and disturbances always existent, this strategy does not lead necessarily to the desired results. Furthermore, the a priori optimization assumes certain initial conditions. In industrial practice it is not possible to reproduce exactly these initial conditions.
In this project, we want to overcome these difficulties by monitoring the process state and optimizing on-line the cooling profile. For this purpose, the Focused Beam Reflectance Measurements (FBRM) provides particle size information. A method previously developed in our group allows for the calculation of the PSD from FBRM measurements [2, 3, 4]. In addition, concentration and temperature are monitored [1].
The measurement information detects instantly deviations from the previously predicted system behavior. Its use in the on-line optimization (closed-loop approach) of the cooling profile warrants that the system follows the optimal trajectory and results in the desired PSD.


Study how to influence particle size and shape in batch crystallization

In batch crystallization some of the properties of the produced crystalline material like its particle size distribution (PSD), particle shape, fines content, etc. greatly influence down-stream processing steps like filtering and drying, thus process operation can substantially be improved by controlling these crystal properties. This requires a thorough understanding of the process, in the form of both theoretical knowledge, as well as online information on important process parameters like PSD, particle shape, temperature and composition.
For this purpose, a population balance model is developed in order to accurately describe crystallization kinetics. Experimental evidence shows that particularly for organic material different crystal faces have different growth kinetics leading to typical crystal shapes (very often needle-like). Therefore the model considers two characteristic crystal lengths to characterize the particle population.
Moreover the properties of the crystal population are related to downstream processing characteristics and establishing this relationship would enable one to identify the optimal crystal properties with regard to process time or other process objectives.
Furthermore, information on the PSD and particle shape can be extracted from FBRM (Focussed Beam Reflectance Measurement) measurements and PVM (Process Video Microscope) images. The combined use of such advanced on-line monitoring tools and a kinetic model significantly enhances the possibilities for an effective tuning of product properties such as size and shape. This approach will be applied to both batch cooling crystallization and anti-solvent crystallization.



[1] Worlitschek J., M. Mazzotti: Model-based Optimisation of PSD in Batch Cooling Crystallisation of Paracetamol, Crystal Growth and Design, in press (2004)

[2] Ruf A., J. Worlitschek, M. Mazzotti: Modeling and experimental analysis of PSD measurements through FBRM, Particle & Particle System Characterization, 17, pp. 168-180 (2000)

[3] Worlitschek J., T. Hocker, M. Mazzotti: Restoration of PSD from Chord Length Distribution data. I. Method of Constrained Least Squares Minimization, submitted to Particle & Particle System Characterization (2003)

[4] Worlitschek J., T. Hocker, M. Mazzotti: Restoration of PSD from Chord Length Distribution data. II. Method of Projections onto Convex Sets, submitted to Particle & Particle System Characterization (2003)


Wichtiger Hinweis:
Diese Website wird in älteren Versionen von Netscape ohne graphische Elemente dargestellt. Die Funktionalität der Website ist aber trotzdem gewährleistet. Wenn Sie diese Website regelmässig benutzen, empfehlen wir Ihnen, auf Ihrem Computer einen aktuellen Browser zu installieren. Weitere Informationen finden Sie auf
folgender Seite.

Important Note:
The content in this site is accessible to any browser or Internet device, however, some graphics will display correctly only in the newer versions of Netscape. To get the most out of our site we suggest you upgrade to a newer browser.
More information

© 2014 ETH Zurich | Imprint | Disclaimer | 31 October 2005