PHARMACEUTICAL INSTRUMENTATION DEVELOPMENT FOR IN-PROCESS MONITORING OF FREEZE-DRYING CYCLES

Background: Product temperature during the primary drying stage is a critical process parameter in the freeze drying of biopharmaceutical drug products and for a successful optimum drying process, ice interface temperature should stay just below the critical temperature of the glass transition – little below collapse temperature. The ice interface temperature is determined by manometric temperature measurement (MTM) and tuneable diode laser absorption spectroscopy (TDLAS) but only provides batch average temperatures whereas the glass transition/collapse temperature is determined using offline lab techniques employing sample conditions that do not mimic the true freeze drying process conditions. Aim: The aim of this project is two-fold: The first is to establish whether it's possible to determine the vapour pressure within the chamber in the space above the vial using a THz and laser absorption approach from which ice interface temperature can be determined. The second is to establish the use of impedance spectroscopy for the in-vial determination of the phase behaviour of surrogate formulations of mannitol and sucrose, with a view to determining the Tg’ and crystallization characteristics that are relevant to the effectiveness of the lyophilization cycle. Method: A Multiplex PAT freeze dryer system (including a silicon oil cooled shelf) was designed and built with the water vapour sensors (PATs): THz-TDS, laserIR system and chilled mirror hygrometer (Optidew) installed for vapour measurement, CFD modelling and TVIS were used to study vapour distribution in chamber and thermal behaviours of mannitol respectively. Results: THz-TDS measurements of VP in chamber identifies 3 important absorption frequencies including 1.153 THz, 2.242 THz and 3.135 THz where water vapour absorption resonance demonstrates high values of SNR. It suggests an uncertainty of 25 µBar with a limit of detection (3α) of 75 µBar cm-1 and the uncertainty dependent temperature for measured and extrapolated data at -40 °C are +1.6/-1.9 °C and +0.5/-0.5 °C respectively. LaserIR demonstrated sensitivity to WVP in the chamber at 26 oC but was unable to measure down to concentration relevance to freeze-drying (e.g., below 1750 mBar) due to absorption by other gas molecules in addition to noise and turbulence in the chamber while chilled mirror hygrometer measurement showed up to 92% agreement with theoretical data and this was confirmed by the CFD simulation. TVIS technology demonstrated ability to measure Tgʹ and crystallization events of freeze-drying sample in the vial. Conclusions: Although the attempts to develop the novel approaches to vapour measurement in freeze-drying chamber failed to deliver a satisfactory outcome, Optidew has been identified as a new alternative for localised WVP measurement in freeze-drying chamber. The most significant finding of this study was the successful application of TVIS technology to identify mannitol glass transition (Tgʹ), crystallization and polymorph transition within original container