Application of Online Near Infrared for Process Understanding of Spray-Drying Solution Preparation

• Published in: Journal of pharmaceutical sciences 2019-03, Vol.108 (3), p.1203-1210
• Main Authors: Lee, Yung-Chi, Zhou, George, Ikeda, Craig, Chouzouri, Georgia, Howell, Logan
• Format: Article
• Language: English
• Subjects: amorphous solid dispersion; Chemistry, Pharmaceutical - instrumentation; Chemistry, Pharmaceutical - methods; Chemistry, Pharmaceutical - standards; Chromatography, High Pressure Liquid - instrumentation; Chromatography, High Pressure Liquid - methods; Desiccation - instrumentation; Desiccation - methods; dissolution; Drug Compounding - instrumentation; Drug Compounding - methods; Drug Compounding - standards; Feasibility Studies; NIRs; PAT; Quality Control; Solubility; Solvents - chemistry; Spectroscopy, Fourier Transform Infrared - instrumentation; Spectroscopy, Fourier Transform Infrared - methods; spray drying; Water - chemistry
• ISSN: 0022-3549; 1520-6017
• DOI: 10.1016/j.xphs.2018.10.022

Solution preparation is the first unit operation of the manufacturing process for spray-dried solid dispersions. Visual inspection and offline high-performance liquid chromatography analysis are routinely used to assess the solution preparation end point as well as the final solution composition. However, the accuracy and appropriateness of these approaches are challenged by the scale of production and solvent evaporation during sample handling. Thus an appropriate online process analytical tool is needed to improve process and quality control for the solution preparation process. The objective of this report is to develop near infrared (NIR) models for real-time monitoring of the spray solution preparation process. These models were built and refined via 2 different experiments designs with different production scale. The potency of spray-dried intermediate was analyzed by high-performance liquid chromatography and used to verify the quantitative model. The results indicated that the quantitative NIR models can be used to predict the active pharmaceutical ingredient concentration of the final spray solution accurately with a standard error of prediction of 2.4 wt%. Based on this investigation, online NIR was deemed to be a suitable analytical tool on process and quality control for spray solution preparation.