Why a Freeze-Dried Cake Keeps Its Shape: Collapse and the Glass Transition
A freeze-dried cake is the solid, porous block left in the container after the water has been removed as vapour. Whether it stands up as a neat block or slumps into a shrunken, sticky mass is decided by two temperatures: the collapse temperature and the glass transition temperature of the freeze-concentrated solution. Both are described in the freeze-drying literature as important physicochemical characteristics that guide cycle development [1].
The two temperatures that set the limits
The collapse temperature, written Tc, is the point at which the drying solid can no longer hold its own structure. The glass transition temperature of the freeze-concentrated solution, written Tg', describes the frozen concentrate that surrounds the ice crystals: below it the concentrate behaves as a rigid glass, above it it softens. Together with the crystallisation behaviour of the other ingredients in the formula, these values guide how a freeze-drying cycle is developed [1].
The practical consequence is plain. Supercooling during the freezing of solutions often leads to poor results — long primary drying times, or a collapse in the cake structure [2]. Collapse is not a cosmetic defect alone; it is the structure failing while the water is still leaving.
How these values are measured
The most frequently used methods to determine Tc and Tg' are differential scanning calorimetry, usually shortened to DSC, and freeze-drying microscopy, shortened to FDM [1]. DSC measures the heat a sample takes up or gives off as its temperature changes, which is how a glass transition is detected. FDM freezes a thin film of the solution under a microscope and dries it while the operator watches for the moment the structure gives way.
A 2018 study published in Frontiers in Chemistry evaluated an optical fibre system, or OFS, as an alternative tool for analysing Tc, Tg' and crystallisation events, and as a possible online monitoring tool during freeze-drying [1]. The OFS and FDM results correlated well, and the crystallisation behaviour of mannitol could be monitored by the OFS during freeze/thawing as it can be by DSC [1]. The authors concluded that the OFS enabled analysis of both Tc and crystallisation processes, work usually done by FDM and DSC, and that detecting these events during drying itself makes online monitoring possible [1].
For a brand, the point is that these are measurable properties of a specific formula, not general rules. Two different sugar systems give two different numbers.
Annealing: changing the frozen state before drying starts
A 2022 study in Pharmaceutics set out to show that annealing causes a rearrangement of water molecules between ice crystals, and between the freeze-concentrated amorphous matrix and the crystalline ice phase, in a frozen binary aqueous solution [2]. Ice crystal sizes and the volume fractions of the crystalline and amorphous phases of 10% (w/w) sucrose and trehalose solutions were quantified after annealing, using freeze-drying microscopy with image labelling [2]. Depending on annealing time and temperature, the amorphous phase decreased in volume as vitreous water crystallised — described as glassy state relaxation — while the crystalline phase coarsened, described as Ostwald ripening [2]. The two phenomena occur simultaneously but are distinct processes following different kinetics [2].
What the formula itself contributes
The glass transition is not fixed by the drying method alone; it moves with what is in the mix. In freeze-dried plant-based powders with added inulin, both the glass transition temperature of the anhydrous material and the characteristic Tg' of maximum freeze-concentration increased as more inulin was added [4]. The glass transition lines in that work were developed using the Gordon–Taylor model, and water adsorption isotherms were measured at 25 °C by a static-gravimetric method [4]. Powders based on plant raw materials were noted to have low storage stability because of their sorption and thermal properties, and to cause problems during processing — which is why carrier agents are sought [4].
Sugar content can also work the other way. In freeze-dried agar gels, dried samples with a lower concentration of sugars showed lower shrinkage and higher porosity, as well as higher crispness and a higher glass transition temperature [5]. Freeze drying combined with maltodextrin gave a very low water activity of 0.038 ± 0.000, but the resulting powder was less free-flowing than microwave-dried material and tended to have lower solubility [6].
Water pulls the glass transition down
The glass transition is not a one-off measurement taken at the end of drying. Absorbed water lowers it. In freeze-dried maca powders, the anhydrous glass transition temperature was 79.5–80.2 °C, with negligible difference between samples, but Tg fell as the powders took up water [7]. The depression was more gradual in the freeze-dried powders than in a commercial comparison powder, attributed to a difference in water-insoluble material content [7].
From this the authors defined a critical water activity — the water activity at which Tg falls to 25 °C [7]. Below a water activity of 0.328 there was negligible caking; above it, the degree of caking increased markedly, with large variation between samples [7]. The degree of caking could be described uniformly as a function of water activity divided by critical water activity, which the authors proposed as an empirical way to predict caking [7]. Higher glass transition temperatures have elsewhere been read as indicating better stability at room temperature [3].
What a brand can take from this
Three things follow for anyone deciding how a freeze-dried powder should be made.
- The cycle has to be built around measured values for your formula. Tc and Tg' guide cycle development, and they are determined by DSC and FDM, with optical fibre monitoring now shown to correlate well with FDM [1].
- The number that matters at the end is not only the cake you see on day one. Glass transition falls as the powder picks up water, and caking rises sharply past a critical water activity [7]. Carrier choice shifts those temperatures [4][6], and so does how much sugar the system carries [5].
A cake that holds its shape is the visible result of a frozen structure that stayed below its limits while the water left, and a dry solid kept below its glass transition afterwards.
References
- Horn J, Friess W. Detection of Collapse and Crystallization of Saccharide, Protein, and Mannitol Formulations by Optical Fibers in Lyophilization — Frontiers in chemistry, 2018-01-26.
- Kharatyan T, Gopireddy SR, Ogawa T, Kodama T, Nishimoto N, Osada S, Scherließ R, Urbanetz NA. Quantitative Analysis of Glassy State Relaxation and Ostwald Ripening during Annealing Using Freeze-Drying Microscopy — Pharmaceutics, 2022-05-30.
- Yildiz G, Gao Y, Ding J, Zhu S, Chen G, Feng H. Enhancing physicochemical, bioactive, and nutritional properties of sweet potatoes: Ultrasonic contact drying with slot jet nozzles compared to hot-air drying and freeze drying — Ultrasonics sonochemistry, 2024-12-30.
- Stępień A, Witczak M, Witczak T. The Thermal Characteristics, Sorption Isotherms and State Diagrams of the Freeze-Dried Pumpkin-Inulin Powders — Molecules (Basel, Switzerland), 2022-03-29.
- Jakubczyk E, Kamińska-Dwórznicka A, Ostrowska-Ligęza E. The Effect of Composition, Pre-Treatment on the Mechanical and Acoustic Properties of Apple Gels and Freeze-Dried Materials — Gels (Basel, Switzerland), 2022-02-10.
- Pashazadeh H, Ali Redha A, Johnson JB, Koca I. Valorization of okra waste: Microencapsulation of okra flower polyphenol-rich extract with maltodextrin and gum Arabic by freeze drying, spray drying, and microwave drying — Journal of food science, 2025-03-01.
- Granados AEA, Kawai K. Browning, Starch Gelatinization, Water Sorption, Glass Transition, and Caking Properties of Freeze-dried Maca ( Lepidium meyenii Walpers) Powders — Journal of applied glycoscience, 2020-11-20.