Increasing process understanding of wet granulation by spectroscopic methods and dimension reduction tools

Jørgensen, A.C., 2004. Increasing process understanding of wet granulation by spectroscopic methods and dimension reduction tools Dissertationes Biocentri Viikki Universitatis Helsingiensis, 20/2004, 62 pp. ISBN 952-10-1955-7 ISBN 952-10-1956-5 (pdf) ISSN 1239-9469 Wet granulation is a common unit operation applied in the pharmaceutical industry. It is a complex process where several interrelated phenomena take place simultaneously. Moreover, it exposes the processed materials to harsh conditions which may alter the solid-state of these. Thus, in order to increase process understanding of wet granulation, methods providing real-time information from the process would be valuable. The aim of the present study was to investigate the use of spectroscopic methods, near-infrared (NIR) and Raman spectroscopy, in elucidating phenomena taking place during wet granulation. More specifically, a processing-induced transformation, hydrate formation, which takes place during wet granulation, was studied. In addition, the use of near-infrared spectroscopy in the process monitoring of high-shear wet granulation was studied by comparing it to impeller torque measurements, which is an established process monitoring method. The measurements were performed offor at-line. Moreover, the difficulty to grasp the large data amounts produced by different process monitoring methods was addressed by combining the data and visualising it with projection methods. Two different approaches were investigated, principal components analysis and self-organizing maps, which are linear and non-linear methods, respectively. It was possible to follow the processing-induced transformation by both spectroscopic methods. Common excipients did not disable the measurements, but altered the rate of transformation. NIR reflected also macroscopic changes taking place during the high-shear granulation process, such as the increase in size and consolidation of the agglomerates. The combination of process data enabled the study of the state of the process in a way which none of the individual process measurements allowed. Both projection methods were able to solve the task of visualising the state of the process. Hence, the use of all of the available process data in a multidisciplinary way, allowed by the projection methods, may contribute to the creation of better process understanding. Table of contents Table of contents _______________________________________________________ i Preface ______________________________________________________________ iii List of abbreviations and acronyms _______________________________________ iv List of original publications ______________________________________________v