Paper from 2022. Wikipedia has some relevant context[1], with more details here[2].

[1]: https://en.wikipedia.org/wiki/Ritonavir#Polymorphism_and_tem...

[2]: https://www.cmbe.engr.uga.edu/bche4520/Other/Ch9/Bauer%20200...

Asianometry has recently covered Rionavir here:

https://m.youtube.com/watch?v=_xPhxtuA_Qc

Polymorphs are actually super interesting scientific challenges in pharmaceutical manufacturing.

I remember one drug was about to be launched, they had settled on polymorph A, but then during trial manufacturing runs polymorph B started to showed up. They couldn't use polymorph B because it messed up the absorption of the drug.

So they tried a ton of different methods to see if they could force the drug to polymorph A, but to no avail since it was the more thermodynamically stable version.

They thought they were ok because another plant was successfully still making A. However, after a few months polymorph B started to pop up, likely due to cross-contamination of some sample or even some microscopic type B crystals falling off a notebook page. Once B showed up, the entire plant was useless for making type A.

In the end, they were able to identify yet another polymorph, type C, which could be reproducibly created even from type B. The resubmitted their manufacturing protocol to the FDA and it was approved. It delayed launch by over a year.

I just finished re-reading Cat's Cradle by Kurt Vonnegut. Ice-nine is probably the reason non-chemists (with a literary bent anyway) know anything about alternate forms of crystallization.