How Freeze Drying Works: Sublimation Without A Liquid Phase

The reason to do it this way rather than by evaporation is structural. Boiling a solution away concentrates the solutes into whatever is left, drives them through a liquid-air interface, and finishes as a film or a crust. Subliming ice out of a frozen matrix leaves the solutes sitting in the space the ice crystals occupied, which is why a lyophilised product is a porous cake of roughly the same volume as the liquid that was filled, rather than a residue at the bottom of the vial.

The process runs in three stages with distinct physics: freezing, primary drying and secondary drying. Freezing decides the ice crystal geometry and therefore the pore structure. Primary drying removes the ice by sublimation, and accounts for the large majority of the water and of the run time. Secondary drying removes water that is bound to the solid rather than frozen, by desorption at higher temperature.

A finished run is measured in hours to days. Primary drying of a 1 mL fill in a 3 mL vial commonly occupies 20 to 40 hours; a 5 mL fill in a larger vial can run several times longer, because the sublimation front has to travel through a thicker cake and the vapour has to escape through it.

Stage one: freezing, which sets the pore structure

Shelves are cooled to between -40 and -50 degrees Celsius, and the solution nucleates and grows ice crystals. Cooling rate controls crystal size: slow cooling gives large crystals and wide channels, fast cooling gives small crystals and narrow ones. Wide channels let vapour out faster, which shortens primary drying; narrow channels give a finer cake with a larger surface area.

Some processes include an annealing hold, where the frozen material is warmed to an intermediate temperature such as -20 degrees Celsius for one to three hours and then re-cooled. The hold allows small crystals to merge into larger ones and lets crystallisable excipients such as mannitol reach their crystalline form before drying begins.

Freezing also concentrates whatever is dissolved. As ice forms, solutes are excluded from the crystal lattice into the remaining unfrozen fraction, which can reach solute concentrations of tens of percent by weight before it solidifies. Buffer salts in that fraction can crystallise selectively and shift its pH by more than one unit.

Stage two: primary drying, and the one number that matters

With the chamber evacuated to roughly 0.05 to 0.3 mbar, heat is supplied through the shelf and ice sublimes from the top of the frozen mass downward. The sublimation front recedes through the vial, leaving dried cake above it. The enthalpy of sublimation of ice is about 51 kJ per mole, close to 2,830 kJ per kilogram, and every bit of that has to be delivered as heat or the front stalls.

Product temperature is not shelf temperature, and the gap between them is the whole control problem. Sublimation is endothermic, so the drying material sits colder than the shelf beneath it, commonly by 5 to 20 degrees Celsius depending on chamber pressure, fill depth and vial heat transfer. Shelf setpoint is what an operator selects; product temperature is what decides the outcome.

The limit on how hard primary drying can be pushed is the collapse temperature, or for a crystallising system the eutectic temperature. Above it the concentrated unfrozen fraction has too little viscosity to hold its shape once the ice around it has gone.

Stage three: secondary drying, for water that is not ice

When the ice is gone, water remains adsorbed to the solid surface and dissolved in the amorphous phase. Removing it is desorption rather than sublimation, and it needs higher temperature rather than lower pressure: shelves are ramped to +20 to +40 degrees Celsius and held for several hours. Residual moisture targets of 1 to 3 percent by mass are common, measured by Karl Fischer coulometric titration on a weighed portion of cake.

At the end of the run, shelves close down onto the vials and seat the stoppers while the chamber is still under reduced pressure, or after a backfill with nitrogen. That is why a lyophilised vial ordinarily holds a vacuum, and why piercing the stopper draws air inward.

The failure mode: product temperature crossing collapse

The specific thing that goes wrong in primary drying is a product temperature excursion above the collapse temperature of the formulation. It happens when shelf temperature is raised to shorten the run, or when chamber pressure drifts up and raises heat transfer to the vial, or simply when a vial at the edge of the shelf receives extra radiant heat from the chamber wall. Edge vials commonly run 1 to 3 degrees Celsius warmer than centre vials for that reason.

It shows up as a cake that has lost its shape: shrunken away from the vial wall, glassy rather than matte, or slumped into a fraction of its original height. The material is chemically the same material, so appearance is the only signal, and it is a process signal rather than a purity measurement.

Where Aurum's records stop

We publish reverse-phase HPLC purity by area percent at 214 nm and a mass consistent with the stated formula. Those are measurements on the material. They say nothing about the drying cycle that produced the cake, and shelf temperature profiles, chamber pressure traces and residual moisture figures for catalogue items are not among the specifications Aurum publishes. We also do not perform or claim sterility or endotoxin testing, which the paragraph above should be read against rather than around.

Cake appearance is the one process signal visible without instruments, and its usefulness is limited to exactly that: a visual record of whether the structure held.

Common questions

Why does freeze drying need a vacuum at all?

Because sublimation of ice is only fast enough to be useful when the pressure of water vapour above the ice is kept far below its equilibrium value. At 6.11 mbar and above, water can exist as a liquid, and the material melts instead of drying.

Is the cake the same volume as the liquid that was filled?

Close to it, in a well-run cycle. That is the point of the method: the solid occupies the space the ice left, minus modest shrinkage. A cake at a fraction of the fill height indicates the structure did not hold.

Why is primary drying so much longer than secondary?

Because primary drying moves the bulk of the water, and it moves it through a growing layer of dried cake that resists vapour flow. Secondary drying moves a few percent by mass and faces no ice front.

Does a longer run mean a drier product?

Not necessarily. Once the ice is gone, additional primary drying time removes nothing. Residual moisture is set by the secondary drying temperature and hold, and confirmed by titration rather than inferred from how long the run lasted.