Primary Drying: How Shelf Temperature and Chamber Pressure Set the Pace
Two settings do most of the work: the temperature of the shelf, and the pressure inside the chamber. Together they set the pace of sublimation. Push them too hard and the structure of the product is lost. Hold them too gently and the cycle runs for days longer than it needs to.
What the two settings actually control
The shelf delivers heat into the frozen material. How much of that heat actually arrives is described by a heat transfer coefficient, usually written Kv — a number that depends on the vial and on the chamber pressure. Chamber pressure changes it: in one study the effect of pressure on Kv was assessed by running tests at 8, 13 and 26 Pa [4].
Working against the heat is the layer of already-dried material sitting on top of the ice. Vapour has to travel out through it. This is why the ideal setpoint is not a single number. One modelling study notes that the optimal setpoint varies with time, because product resistance varies continuously [1].
Those two coefficients — heat in, vapour out — are the inputs to the models used to plan a cycle. Both were estimated for a crystalline solution (5% mannitol) and an amorphous one (5% sucrose) in 6R vials, and the resulting process maps compared against conventional laboratory methods [4].
The limit: collapse temperature
There is a ceiling on how fast you can drive the process. At the sublimation interface — the moving boundary between dry material above and ice below — the product has a temperature, Ti. If Ti rises above the collapse temperature of the formulation, Tc, the structure is lost [2].
This matters most for amorphous formulations. A collapsed cake is a shrunken, glassy or sunken mass instead of an even, porous block. The risk of collapse is what any sensible optimisation is trying to minimise while still keeping process time economical [3].
There is a second, less obvious limit. Push sublimation hard enough and the vapour flow through the chamber reaches choked flow conditions, at which point pressure control is lost [2]. So the usable operating window is bounded on both sides: product structure at one edge, the machine's ability to hold pressure at the other.
Design space: the usable window, drawn out
The region between those limits is called the design space — the set of shelf temperature and chamber pressure combinations that will dry the product without breaking it. A primary drying calculator performing one-dimensional heat and mass transfer analysis in a vial has been extended to generate a design space across a range of chamber pressures and shelf temperatures, and to pick the best setpoints from it [1]. The same calculator predicted drying time with an average deviation of 3% from experiments [1].
Drawing that space requires measuring Kv and Rp, which has traditionally meant gravimetric tests — weighing vials to see how much ice left — and thermocouples pushed into the product. Newer tools do it with less material. An infrared-based process analytical method allows Kv to be estimated without weighing, and a micro freeze-dryer allows gravimetric tests with a very small number of vials [4]. The coefficients and the resulting design spaces obtained this way were in very good agreement with those from a laboratory-scale unit [4]. The practical gain is saving time and formulation material when evaluating Rp [4].
What optimisation is worth
Standard practice is to run the dryer to a fixed protocol: each phase programmed at fixed time points, with the critical process parameters inside each phase held constant or interpolated in a straight line between two setpoints. This has been shown to be time consuming and inefficient [2].
Replacing fixed setpoints with profiles that change over time produces large savings. An optimiser that supplies varying chamber pressure and shelf temperature as functions of time achieved 62% faster primary drying for 5% mannitol and 50% faster for 5% sucrose, compared with typical cycle conditions [1].
A different approach keeps watch while the cycle runs. In a real-time control strategy, a design space is recomputed at each iteration to select an optimal set of process parameters, with the model's output uncertainty taken into account [2]. The method was tested by deliberately disturbing the chamber pressure and the shelf fluid system, and was confirmed capable of handling moderate disturbances [2]. The end of primary drying was predicted using both uncertainty analysis and a comparative pressure measurement technique, and the predictions agreed with the experimental data [2].
A third route optimises the shelf temperature in two stages, to maximise sublimation without risking collapse [3]. That approach builds in high-resolution variability data for the heat transfer coefficient, vial dimensions and dried layer resistance, and feeds them into an uncertainty analysis that estimates the risk of the protocol failing [3]. It was verified on two formulations, one allowing aggressive drying and one requiring conservative drying [3]. The outcome was protocols both faster and more robust than the classical approach [3].
What a brand can take from this
Primary drying is not one setting applied for a fixed number of hours. It is a window, and where the window sits depends on the formulation. A crystalline material such as mannitol and an amorphous one such as sucrose behave differently enough that they were studied as separate cases [4], and one produced a 62% time saving under optimisation where the other produced 50% [1].
References
- Shivkumar G, Kazarin PS, Strongrich AD, Alexeenko AA. LyoPRONTO: an Open-Source Lyophilization Process Optimization Tool — AAPS PharmSciTech, 2019-10-31.
- Vanbillemont B, Nicolaï N, Leys L, De Beer T. Model-Based Optimisation and Control Strategy for the Primary Drying Phase of a Lyophilisation Process — Pharmaceutics, 2020-02-20.
- Vanbillemont B, Greiner AL, Ehrl V, Menzen T, Friess W, Hawe A. A model-based optimization strategy to achieve fast and robust freeze-drying cycles — International journal of pharmaceutics: X, 2023-04-10.
- Harguindeguy M, Fissore D. Micro Freeze-Dryer and Infrared-Based PAT: Novel Tools for Primary Drying Design Space Determination of Freeze-Drying Processes — Pharmaceutical research, 2021-03-08.