461286 Process Optimization Using a Distillation Model Coupled with a Solubility Model for Increased Process Yield Following Multiple Distillations and Solvent Swap

Monday, November 14, 2016: 3:40 PM
Union Square 5 & 6 (Hilton San Francisco Union Square)
Christopher J. Morrison, Particle Sciences and Engineering, GlaxoSmithKline, King of Prussia, PA and John Guo, GSK, King of Prussia, PA

Yield variability and extended distillation times lead to the need for process modeling. The combined distillation and solubility model predicts both the acetonitrile concentration and resulting yield loss in the mother liquors. The model was used to reduce yield loss and cycle time while providing additional potential for solvent recovery of acetonitrile. The distillation model was generated using Aspen properties with the system at 0.1 bar absolute pressure and Euler integration for a stepwise mass balance approach while the solubility model was generated using data regression from mother liquors. The solubility data was necessary to be used from mother liquor rather than in a pure system due to increased solubility of the intermediate while in the presence of the byproducts of the desired reaction. The model has predicted the acetonitrile composition to be within 0.03 mass fraction of the experimental data. The yield loss using the mother liquors model with modeled acetonitrile mass fraction resulted in a difference of no more than 1.30% yield over the 0.001 to 0.114 acetonitrile mass fraction range. The model was used to maximize yield and was verified on pilot plant scale.

Extended Abstract: File Not Uploaded
See more of this Session: Pharmaceutical Process Development & Pilot Plants
See more of this Group/Topical: Process Development Division