425191 Prediction of Lubrication-Based Tensile Strength Reduction Using a Compartment Modeling Approach

Thursday, November 12, 2015: 1:30 PM
Ballroom B (Salt Palace Convention Center)
William R. Ketterhagen, Drug Product Design, Process Modeling & Engineering Technology, Pfizer Worldwide Research and Development, Groton, CT, Matthew P. Mullarney, Drug Product Design, Pfizer Inc., Groton, CT, John Kresevic, Drug Product Design, Pfizer, Inc., Groton, CT and Daniel O. Blackwood, Drug Product Design, Pfizer Worldwide Research and Development, Groton, CT

In the production of pharmaceutical tablets and capsules, a lubricant such as magnesium stearate is frequently added to the powder blend or granulation to reduce friction between the powder and the tablet press or encapsulator components.  While the presence of the lubricant is generally necessary to improve manufacturability of the dosage form, if the lubricated blend is exposed to excessive shear strain during processing and handling prior to tableting or encapsulation, adverse effects on quality attributes of the final dosage form may be observed.  These effects can include an increase in wetting contact angle, a slowdown in disintegration and/or dissolution, and a reduction in tensile strength. 

In this work, the extent of lubrication in a powder feeding process is studied experimentally and with a compartment model approach.  Since a placebo blend is used to study this phenomenon, the tensile strength is the quality attribute used to quantify the extent of lubrication.  Here, a compartment modeling approach is proposed to model the powder flow pattern and predict the lubrication-based tensile strength reduction that may occur during the production of a large batch.  Parameter estimation is conducted through the use of a separate experiment utilizing an input step change in powder feed from undyed to dyed material.  Model predictions are compared with experimental results as well as predictions from a separate discrete element method (DEM) modeling approach.  Finally, additional predictions from the compartment model are made to assess the impact of various processing modifications.

Extended Abstract: File Not Uploaded