What is the effect of agitation on crystallization?

The common crystallization kettle stirring (Stirred tank crystallizer) is the axial paddle type as shown in Fig. a. Here the authors designed an anchor-like scraped-wall stirring paddle (scraped surface crystallizer) and selected three different types of APIs to study the effect of stirring on crystallization.

A comprehensive look at crystallization techniques, this one was a real eye-opener!
The first was ibuprofen (IBP), which was crystallized by adding a droplet of water to a 25 °C solution of IBP in DMSO. When crystallizing with ST axial paddle stirring at 60 rpm, the stirring was not good and by 25 minutes there was an aqueous phase in the upper layer of DMSO, resulting in a high yield when sampled and analyzed at 30 minutes. Continuing to agitate and sample, crystallization using ST axial flow stirring at 200 rpm and SS wall scraper anchor stirring at 60 rpm resulted in similar yields at 30 min and 60 min.

How to apply efficient continuous flow production process in biopharmaceutical? With case sharing!
The hard crust on the surface is related to the nature of the crystallization API itself as well as the smoothness and hardness of the surface of the stirring paddle and the crystallization kettle. With the stainless steel ST paddle type at 200 rpm (Fig. c), there was more severe surface crusting, but with the SS anchor type at 60 rpm using PTFE, there was less surface crusting.

Diphenhydramine Hydrochloride (DPH) is a lamellar crystal that is precipitated by cooling starting at 60 °C to 5 °C and then holding.SS Anchor 60 rpm crystallization kinetics are similar to those of the ST 200 rpm, and both precipitate faster than the ST Paddle 60 rpm.

Differences and connections between small and scaled-up experiments and pilot production

There were more lumpy solids in the crystals obtained using ST paddle stirring (Fig. b, 60 rpm; c, 200 rpm) (Fig. a).

 

Discharge was easier with both the SS Anchor 60rpm and the ST Paddle 200rpm with no residual chunky solids, but the ST 60rpm had more surface crusting, and chunky residue on discharge.

For Fluoxetine Hydrochloride (FLU) in needle form, also cooled and decrystallized, the SS Anchor 60rpm decrystallized significantly faster than the ST Paddle.

Using these three crystallization conditions, long, thin, needle-like FLU crystals were obtained, with the SS-anchor type yielding longer crystals (Fig. a) than the others (Fig. a, ST, 60 rpm; Fig. c, ST, 200 rpm). (Dry Sharing) Industrial Crystallization Process Control Points and Application Cases

No discharge problems were encountered using the SS anchored 60 rpm (Fig. a), but crystals obtained at different speeds using the ST were poorly mobile and could not be poured out directly, but only after making up the solvent (Fig. b).

Comprehensive Comparison:
1. ST Paddle Stirring 200 rpm promotes crystal breakage, agglomeration, and secondary nucleation, so the crystal size distribution is wider.
2. ST paddle stirring 60 rpm, DPH crystals grow to a certain size and start to precipitate and agglomerate to the bottom of the kettle. Because of poor agitation, crystal growth is affected by mass transfer, resulting in slower crystal precipitation.
3. SS anchor stirring 60 rpm, because the gap between the anchor and the kettle wall is smaller, the solid suspension is better, and the CFD of fluid simulation shows that the average shear of the system is smaller. Less crystal breakage, such as the crystallization of FLU, than in other stirring systems with finely broken crystals. Low rotational speed and low shear, so less crystal-crystal and crystal-kettle wall interaction, so less agglomeration.