Primary And Secondary Drying Remove Two Different Kinds Of Water

Water in a frozen solution is not all in one state. Cooling separates most of it into ice crystals, but a fraction stays behind in a concentrated, glassy matrix with the solutes, unfrozen and immobilised rather than crystalline. Those two populations of water leave the vial by different physical mechanisms, and a freeze drying cycle addresses them in sequence.

Primary drying is sublimation. With chamber pressure held below the vapour pressure of ice at the product temperature, ice converts straight to vapour and the vapour travels to a condenser held far colder than the product. Secondary drying is desorption: no ice remains, and the water still present is being coaxed off the surface of the dried solid by raising shelf temperature once the risk of melting has passed.

The order is not negotiable and the transition between them is the sensitive moment in the whole cycle. Primary drying must be complete, in every vial, before shelf temperature rises for secondary drying. A vial that still holds ice when the shelves warm has its product temperature pushed above the point where the concentrated matrix can support its own structure, and the cake loses its shape.

The three phases in order

Freezing is not a drying phase, but it determines how the drying phases behave. The ice crystal size set during freezing becomes the pore network of the dried cake, and that network is the path every subsequent water molecule has to travel. Fast freezing gives small crystals, fine pores, high resistance to vapour flow and a slow primary drying. Slow or annealed freezing gives larger crystals and a more open, faster-drying structure.

Why primary drying stays cold

During sublimation the product sits below the shelf, thermally, because the phase change absorbs heat. Sublimation of water requires roughly 2,830 kJ per kilogram, and that energy comes from the shelf through the vial base. The result is a self-regulating temperature difference: a typical product temperature during primary drying runs about 5 °C to 20 °C below the shelf setting, with the gap widening as the sublimation rate rises.

The ceiling on product temperature is a material property of the frozen solution. For a formulation that stays amorphous, the relevant figure is the glass transition temperature of the maximally freeze-concentrated solute, written Tg prime, commonly in the range of −45 °C to −25 °C for sugar and polyol-based solutions. For a crystallising formulation it is the eutectic melting temperature. Primary drying is run with product temperature below that figure, usually with a margin of 2 °C to 5 °C.

Higher shelf temperature shortens primary drying, and the entire economics of a cycle pushes toward the highest shelf temperature that keeps product temperature under the limit. That is why the endpoint of primary drying is measured rather than assumed.

How the endpoint is detected

The usual signal is comparative pressure measurement. A Pirani gauge reads pressure by thermal conductivity and is calibrated for nitrogen, so it over-reads in the water-vapour-rich atmosphere of active sublimation. A capacitance manometer reads pressure by diaphragm deflection and is gas-independent. While ice is subliming the two gauges disagree; as sublimation ends and the chamber atmosphere loses its water vapour, the Pirani reading falls toward the manometer reading and the curves converge.

• Comparative pressure measurement: convergence of Pirani and capacitance manometer readings marks the end of sublimation across the batch.
• Pressure rise test: the chamber is isolated briefly and the rate of pressure increase is recorded. A small rise indicates little remaining sublimation.
• Product thermocouples: probes in individual vials show product temperature rising toward shelf temperature once ice is gone. They report on the vials they are in, and probed vials dry faster than unprobed ones.
• Tunable diode laser absorption spectroscopy: measures water vapour concentration and flow in the duct to the condenser, giving a batch-level sublimation rate in real time.

Every one of these methods reports on the batch as a whole or on a handful of instrumented vials, which is the structural weakness. Vials at the edge of a shelf receive radiant heat from the chamber walls and dry faster than vials in the middle of a tray, a difference often described as edge effect and observed as a spread of several hours in primary drying completion across a single shelf.

The failure mode: warming the shelves too early

The characteristic failure is premature transition to secondary drying while the slowest vials still hold ice. Shelf temperature climbs, the remaining ice in those vials is no longer absorbing heat fast enough to hold product temperature down, product temperature crosses Tg prime, and the concentrated matrix softens. A softened matrix cannot support the pore structure that sublimation has already carved out of it, so the structure gives way.

It shows up as a visibly shrunken, glassy or partly melted-looking cake in a minority of vials, usually those in the centre of a tray rather than at the edges, while the rest of the batch looks normal. The collapsed vials also dry less completely, because the open pore network that secondary drying depends on has closed, so they tend to carry higher residual moisture as well as looking wrong.

What the published record says about a cycle

Cycle data are not among the specifications Aurum publishes. Shelf temperature profiles, chamber pressure traces, product thermocouple records and endpoint determinations are generated by the freeze-drying operation that produced the cake, and none of them appear in the published record for a lot. Nothing on a listing should be read as a statement about how a particular cycle was run.

Residual moisture, the one figure that would let a downstream party infer anything about how secondary drying went, is also not among the published specifications. What is published is purity by reverse-phase HPLC and identity by mass spectrometry, and a chromatogram is silent on drying. That leaves cake appearance as the only cycle-related observation available at the point a vial is received, and appearance is a coarse instrument.

Common questions

Which phase takes the most time?

Primary drying, almost always, and by a wide margin. It removes the bulk of the water at the coldest temperature the cycle uses, which is the slow combination.

Can the two phases overlap?

Across a batch they unavoidably do, because vials do not finish sublimation simultaneously. Within a single vial they do not: desorption of bound water only becomes the dominant process once that vial's ice is gone.

Why is chamber pressure kept low during secondary drying too?

Low pressure maintains the concentration gradient that drives water vapour off the solid and toward the condenser. Some cycles raise pressure slightly during secondary drying to aid heat transfer to the vials, which is a deliberate trade rather than an oversight.

Does a longer secondary drying phase always give a drier cake?

Up to a point, then the curve flattens. Desorption slows asymptotically, so additional hours buy progressively less water removal, and some amorphous solids are physically less stable when dried very aggressively.

Is the condenser part of the drying, or just plumbing?

It is part of the drying. The condenser is the cold sink that keeps water vapour moving out of the chamber, typically held at −50 °C to −80 °C. If its capacity is exceeded, chamber pressure rises and sublimation slows regardless of the shelf setting.