Computer modelling the freeze drying process

Published on: 

Pharmaceutical Technology Europe

Despite numerous benefits, computer modelling has mainly remained in the domain of academic research.

The first model of the freeze‑drying process was developed in the ‘70s, but, despite numerous benefits, computer modelling has mainly remained in the domain of academic research. A few research teams have proposed commercial user‑friendly modelling software, but these require the experimental determination of various parameters (specific to both the product and the equipment), which makes using them in the industrial field difficult.

Large pharmaceutical companies with high expertise in the process have shown a growing interest in the modelling approach because it offers numerous benefits, including facilitating process design and scale‑up. The most important operational advantage, however, is the ability to remove empirism in the design of freeze drying cycles, which sharply reduces the experimental effort because a design cycle that considers quality and economic objectives can be quickly designed. In addition, the model can provide better insight into a poorly instrumented process and allows information to be obtained that is difficult to access by measurement, such as product temperature or moisture content.

There are two main challenges associated with computer modelling. The first is to develop easy and rapid experimental tests to determine the parameters of heat and mass transfer of the models. Some approaches using the methodology of the pressure rise test have been recently developed to facilitate the determination of these complex parameters. The second challenge is to integrate the heterogeneity of drying conditions into the model — the freezing step, the equipment and the vials contribute to the whole heterogeneity. The result is that each vial within the same batch has its own drying behaviour and final quality properties. Combining modelling and statistical approaches are a promising means to develop spatially distributed models of product quality.

It is quite difficult to speak of innovation — what we observe is the application of already‑known concepts to the freeze drying process. However, three recent applications can be mentioned: the development of software sensors for in‑line control of the process; the use of computational fluid dynamics to model vapour flow in freeze drying; and the modelling of other product conditioning, such as spherical particles of few microns in the case of spray freeze drying. Spray freeze drying is a relatively new method to produce micronised powders with enhanced dissolution properties for purposes such as microencapsulation, pulmonary delivery or intra‑epidermal needle‑free injection.


In the future, the FDA’s Quality by Design approach will probably help to increase the use of computer modelling. From an academic point of view, the modelling of the freeze drying process will move from the macroscopic scale to the microscopic scale to predict the degradation degree of the active ingredient. But this is a big challenge that will take more than 10 years to come to fruition!

Based on a contribution by Stéphanie Passot, Assistant Professor at the Paris Institute of Technology for Life, Food and Environmental Sciences (AgroParisTech, France).