Search

This article is the second of a series of two articles. In the first article of the series, a new K v distribution model and an experimental methodology to measure the K v distribution were introduced. In this second part, the K v distribution is integrated into a lyo-simulation tool, to more accurately predict the variability of the product temperature, primary drying time, total sublimation mass flow and Pirani signal. The K v distribution is also integrated into the graphical design space. The impact of incorporating the R p distribution is briefly discussed. The comparison of the simulation tool with actual product temperature monitoring, Pirani signal or overall sublimation flow shows very good agreement in the case studies presented. Overall, the lyo-simulation incorporating the K v distribution is a very useful tool to support industrial development, i.e. process optimization, scale assessment, technology transfer, and troubleshooting of the lyophilization process., Competing Interests: Declaration of Competing Interests The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: All the authors worked for Sanofi at the time where the research work was done, except one working for Astek - IT&M STATS (contracting company for Sanofi). All the work was Sanofi financed., (Copyright © 2024. Published by Elsevier Inc.)

The freezing of biologics has been widely studied from the physical chemistry point of view, for instance in terms of cryo-concentration, excipient crystallization, pH swing, potential protein denaturation, etc. In contrast, considerations on the processing aspects are very limited. For instance, the impact of freezer temperature, container size, freezer load, and freeze chilling capacity on the freezing rate in the most frequent case of freezing in a bottle have not been reported. In this study, the freezing time of either water or buffer solution was measured in various processing conditions. Experimental trials were conducted using containers ranging from 1 to 20 L in two types of freezers: a normal freezer (-30°C set point) and an ultra freezer (-70°C set point). These trials showed that both the container size and the freezer load influenced the freezing times. The current study demonstrated that the use of the well-established Plank model for freezing, coupled with freezer performance characterization, allows the description of the actual freezing kinetics in a very simple and accurate manner. The kinetics can then be modeled to accurately predict both the actual freezer temperature (possibly above the set point) and the freezing time based on freezer load., (© PDA, Inc. 2021.)