Micro-Scale-Down a small-volume physical model to assist freeze-thaw process design

When a volume of solution is freezing, which can take several hours or even days, drug substance is under two situations that are represented in Figure 1: within ice crystals (in stress) or in the liquid bulk (without stress).
The amount of time that drug substance stays at each situation depends of its location in the container (e.g. near the walls or at the center) and process parameters (size of the container and heat-transfer properties).

Any scale-down strategy that disregards this duality of stresses within the container’s microscale will fail to provide a reasonable assessment of the process. The Micro-Scale-Down (Figure 1) is a system that was
developed by Smartfreez to replicate the stress-time curve of a manufacturing process using samples of only 5 mL to 10 mL. For example, the green areas in Figure 2 represent the stress-time of 1 mL vials that were frozen to match the curve of the 20 L bottle frozen in laminar airflow.

Smartfreez technology (equipment) enables to control accurately nucleation and heat transfer in small volumes, without interference of agents that can also compromise protein stability (as particles or ultrasounds). This is critical because in small volumes, samples can supercool and freeze abruptly (in a faction of a second), as the container walls can easily have more mass than the sample, compromising completely the matching of the stress-time. If freezing is well controlled, the ice-interface in the micro-scale-down vials is expected to be similar to the corresponding area in the large container because the stress-time is equivalent. Note that stress-time is an equivalent definition to cooling-rate (Tf-Tg)/(stress-time), because (Tf-Tg) is relatively constant.

Overall, the Micro-Scale-Down enables to run small volume testing under freeze-thaw rates that are relevant to a selected manufacturing equipment, or alternatively, select a manufacturing equipment and conditions (among available alternatives) that can deliver a selected freeze-thaw requirement.

Figure 1 – During freezing the stresses have a binary distribution illustrated on the left. When ice is present there are interfaces, low temperature and high concentration; when ice is not present drug substance is at refrigeration conditions near initial composition. The time that drug substance is inside the ice structure until it reaches Tg’ (loss of molecular mobility) is considered the stress-time. The image at the center shows the stress time distributions for 2 L of drug substance frozen in bottles under different regimes (airblast at 5 m/s or low airflow 1 to 2 m/s). The dashed curves show 5 mL samples that were frozen in the Micro-Scale Down, in this case to replicate the stresses of the 2 L bottle frozen under different conditions. On the right is shown a photograph of the equipment.