Wednesday, March 14, 2012

Process Wednesday: Oiling out

Credit: Practical Process Research and Development
Neal Anderson
From our mentor-by-literature Neal Anderson comes more thoughts on crystallization and the formation of oils (ugh!):
Gradual cooling without seeding leads to one nucleation event and the formation of relatively small crystals (Path B). Slowing applying the crystallization pressure (path C) can produce large crystals, depending on the size and number of seeds. If the crystallization pressure is too great (path A), supersaturation occurs, and molecules may be forced out of solution too quickly as small crystalline solids, non-crystalline precipitates, or oils... 
The formation of oil droplets may be verified by light microscopy examination. Oils are difficulty to displace from a filter cake by washing. Oils tend to foul filters, slowing filtration rates, and they occlude impurities in the isolated product. Conditions that produce oiling and precipitations should be avoided. Seed crystals are added to solutions of compounds adjusted to the metastable conditions in order to prompt good crystal growth. 
Anderson seems to support using seed crystals to encourage crystallizations to take path C. He later mentions using hold times (e.g. hold temperature of crystallization at 70°C for 1 hour) to encourage crystal growth.

There's a lot to learn about crystallization... 


  1. Crystallisation is an art not a science. We moved into brand new labs once, nothing would crystallize. After around 12 months the compounds started to crystallise and would not stop after that! I had to re-write several procedures to cope with the new situation(s).

  2. Sounds familiar - you hear the urban myths of crystallisations that have run on plant for years, then one day they make a dodgy batch of a different form, and that's all they can produce for ever more.

    It's surprising how wide the MSZW can be with some compounds - equipment like Avantium's Crystal 16 is useful for finding the boundaries. I'd be interested in alternative kit/methods used in other labs.

  3. I disagree. Crystallization is very much a science. A complex system of conditions can be controlled to give reproducible results. Finding the boundaries (other polymorphs) is risk-based. How many conditions should be screened? There is always the chance a new form will crop up, but due diligence, using Xtl16 and other screens/methods can reduce the chances of a new polymorph greatly. And its super important to get a good solubility curve at the beginning of the process development.

    1. If you don't have access to a superawesome crystallization machine, what do you use?

    2. seed crystals or cigarette smoke!