On-Line Monitoring of Roller Compaction Ribbon Density Using NIR Spectroscopy

Solid dosage forms, such as tablets and capsules, have been predominately manufactured through batch processes in pharmaceutical industry for years. The development of continuous manufacturing processes is of high interest to the pharmaceutical industry to reduce the huge costs associated with process scale-up. Roller compaction is a dry granulation technique which is readily applicable to continuous processing and is currently a part of a test bed at Purdue University to demonstrate the feasibility of manufacturing tablets continuously. PAT implementation is a key component to the success of the continuous process since we must now measure critical quality parameters on-line. Near infrared spectroscopy has been selected as the analytical “eye” for monitoring ribbon density since it is an analytical technique which is non-destructive, fast, and measures both chemical and physical information about the sample. Previous research has shown that roller compacted ribbons have a three dimensional density distribution which can affect the particle size distribution of granules after milling. Our specific interest is in understanding and monitoring the density distribution of ribbons produced by roller compaction with near infrared spectroscopy. We will also investigate the distribution of active pharmaceutical ingredients within the ribbons. Therefore, by understanding the effects of formulation, operating conditions, and equipment design on ribbon density distribution we can develop optimized monitoring strategies to help produce a more uniform final product. The active pharmaceutical ingredient, acetaminophen (APAP), and excipients, including microcrystalline cellulose (MCC), magnesium stearate (MgSt) and Cal-O-Sil® (SiO2) were selected as the model system to study the roller compaction process. To understand the effect of formulation on ribbon density distribution acetaminophen concentration, microcrystalline cellulose grades, and lubricant levels were changed. The effect of operating conditions on ribbon density distribution was studied by changing the roll force and the feed screw speed to roll speed ratio. Finally, the effect of equipment design was studied by altering the penetration of the feed screw into the slip region of the rolls. A multi-point, non-contact diffuse reflectance near infrared probe was set-up at the outlet of the rolls to interrogate the entire width of the ribbon. Offline envelop density measurements were used to create and confirm a multivariate density calibration model for each formulation. The multi-point spectra were combined together to create density contour maps. In this work, we show the results of the multi-point experiments to characterize ribbon density distribution and acetaminophen distribution. We will also discuss how the variance of ribbon density changes with the scale of scrutiny. Based on the analysis of density variation, a near infrared monitoring strategy will be outlined. Finally, we will offer some general conclusions about how the experimental results can be used and extended to process control.