Methodology, Data Interpretation and Practical Transfer of Freeze-Dry Microscopy

For this thesis, methodology, data interpretation and practical transferability of results measured by freeze-dry microscopy were subject to detailed investigations. Previously published research work and findings were included in the discussion and in some cases were newly interpreted. Based on a theoretical background, experimental results led to a classification of collapse behavior in three collapse phases. For better theoretical and practical understanding and the explanation of recent and previous findings, a model of factors influencing freeze-dry microscopy experiments was developed in this thesis. During freeze-dry microscopy experiments the limiting parameter of the overall balance has to be the temperature or better the heating rate used. Pressure setting should be adequate, but not limiting. Besides temperature and pressure settings, the extent of supercooling, the physical properties of solved substances and their total solid content are primary factors of collapse events. To prove the introduced model, to investigate the importance of single parameters, and to find an optimized freeze-dry microscopy methodology, detailed studies were carried out including various experimental settings. For sucrose, trehalose, glucose, 2-(hydroxypropyl)-s-cyclodextrin, and polyvinylpyrrolidone 10 and 40 kDa, a dependence of collapse temperatures on total solid content was found. Despite different extents and shapes of the plotted curves for the various excipients, three trends became obvious: (1) Within a given excipient class the collapse temperature increases with a higher molecular size or chain length (for polymers) and the difference between onset of collapse and full collapse decreases. (2) For all excipients an exponential correlation was investigated with low collapse temperatures for low concentrations and higher collapse temperatures for intermediate concentrations (10 to 15% w/w). (3) Hygroscopic substances show a decrease of collapse temperature values for high (about 20 to 30% w/w) total solid contents (depending on the excipient). Furthermore, it was found that bovine serum albumin and human serum albumin are interchangeable from freeze-dry microscopy measurements based on their structural similarity. In contrast to glass transition temperature Tg´, their collapse temperatures were clearly and relatively easily detectable by freeze-dry microscopy. Studies on binary mixtures of protein with a disaccharide revealed a high dependence of collapse temperatures (onset and full collapse) on the composition of the mixture. Differences in freeze-drying and collapse behavior were strong as well. For selected excipients and binary mixtures of protein and disaccharide the measured collapse temperatures were compared to glass transition temperatures of the maximally freeze-concentrated solutions. Due to the dependence of the collapse temperature on total solid content of the measured substance(s), differences between collapse temperatures and glass transition temperatures varied strongly for excipient solutions as well as protein/sugar mixtures. By freeze-dry microscopy measurements in combination with lab-scale freeze-drying runs it could be shown that obtained collapse temperature results are in general transferable to freeze-drying cycles with regard to, e.g., the difference detected for collapse temperature between different concentrations of the same excipient. Scanning electron microscopy in combination with BET specific surface area measurements clearly indicated severe damages in cakes for which primary drying was performed at higher product temperatures, even it was performed below the temperature of the onset of collapse.