Residual Moisture: The Water a Freeze-Dried Cake Keeps After Drying
No freeze-dried cake is completely dry. The question for a brand is not whether water remains but how much, how evenly it is spread across a batch, and what that water does over the months the product sits on a shelf.
The numbers published studies report
Residual moisture levels vary widely with what is being dried. A study of disaccharide formulations reported amorphous cakes with a residual water content of 0.8% w/w after lyophilisation [1]. A study of a freeze-dried reference material found that a phosphate buffer gave well-formed cakes but a high residual moisture content of 2–4% w/w [5]. Freeze-dried emulsions built on sodium alginate and whey protein isolate, developed for cosmetic and dermatological use, came out between 2.3% and 10.9% [6]. Freeze-dried bigel materials made from a hydrogel and an oleogel phase ran higher still, from 5% to 20% [7].
Those four figures come from four different material systems, so they are not a ranking of good and bad. They show the range a formulator is working inside, and that the formulation decides most of it.
Formulation changes the cake, and the cake decides the water
The clearest demonstration of this comes from the reference-material study. Adding sodium chloride — a formulant often viewed as unfavourable for freeze-drying — did not impair the freeze-dried appearance or the process, and cut residual moisture considerably, to below 1% w/w [5]. The authors postulated that the salt produced a more open cake structure, and supported this with thermal analysis and scanning electron microscopy [5].
That is the mechanism worth holding onto: water leaves through the pores. A more open structure gives the vapour an easier path out. The bigel study points the same way from the other side — residual moisture there increased with ethylcellulose content and decreased with whey protein isolate and oleogel content [7]. The same materials also varied in porosity, from 35% to 58%, and in density, from 100 to 200 mg/mL [7]. In the alginate and whey protein emulsions, porosity ran from 59% to 95% and density from 116 to 308 mg/mL, and the physicochemical properties depended on the concentration of whey protein isolate, the aqueous-to-oily phase ratio, and the type and amount of cryoprotectant, oil and beeswax added [6]. Glycerin, propylene glycol, sorbitol, mannitol and trehalose were the cryoprotectants tested [6].
So the choices made before anything is frozen — which polymers, which sugars or polyols, how much oil phase, how the emulsion is homogenised [6] — are already setting the moisture the finished cake will carry.
What drives water out during secondary drying
Secondary drying is where the bound water is removed, and its rate can be measured. One study used near-infrared spectroscopy inline to build and validate a model predicting residual moisture content in a 10% sucrose formulation during a continuous spin freeze-drying process, then used that model to test which process settings mattered [4].
The ranking was clear. Product temperature, controlled by a feedback mechanism on infrared heaters, had the highest positive impact on the drying rate during secondary drying [4]. A higher cooling rate during the freezing step significantly increased the desorption rate as well [4]. A higher filling volume had a smaller negative effect [4]. Chamber pressure, between 10 and 30 Pa, had no significant effect at all [4].
Two things follow. First, what happens at the freezing step still shows up in the moisture level at the end — freezing and drying are not independent decisions. Second, filling more material into a container slows drying down, so a change in container or fill is a change in the drying cycle.
Heating method also affects how evenly moisture is distributed. A microwave-assisted vacuum drying study measured heating uniformity through residual moisture analysis across a batch and found a normalised standard deviation of 12.7% in the microwave-assisted cycle against 15.3% in the conventional cycle [2]. At 18 GHz and 60 W, the method gave nearly three times the speed-up in primary drying of sucrose relative to a conventional cycle for typical laboratory batches [2]. The conventionally used 2.45 GHz frequency was shown to be far from optimum [2].
Too much water collapses the cake
The disaccharide study tested this directly. Cakes that started at 0.8% w/w residual water were deliberately given extra moisture by standing the containers half open in a 25 °C / 60% relative humidity chamber, bringing them to between 7.5% and 10% w/w [1]. That addition caused rapid collapse of the cake for every disaccharide except trehalose [1]. By contrast, storage of the properly dried cakes at 40 °C / 75% RH did not cause physical collapse [1].
Heat alone did not wreck the structure; water did. And a half-opened container in a humid room was enough to take the moisture content up by roughly an order of magnitude [1]. Packaging and the moment of closing are part of the drying process, not an afterthought.
Measuring it, and keeping it the same at scale
Residual moisture content is one of the standard characterisations reported for freeze-dried materials, alongside porosity, density, mechanical properties and swelling [3][6][7]. It can also be followed in real time: the near-infrared model predicted moisture inline during drying [4], and process analytical tools have been used to monitor freeze-drying at both pilot and production scale, including using the vacuum valve open/close frequency to determine the endpoint of primary drying at production scale [8].
Scale-up is the usual worry. In that pilot-versus-production comparison, the duration of primary drying and the residual moisture content were found to be independent of freeze-dryer scale under two different protocols, which the authors took as evidence of functional equivalence between the two machines [8].
What a brand can take from this
Ask for a residual moisture figure, and ask what method produced it. Materials stored dry and swollen with water directly before use are being proposed precisely because that form travels and stores more easily [3].
References
- Quaak SG, Haanen JB, Beijnen JH, Nuijen B. Naked plasmid DNA formulation: effect of different disaccharides on stability after lyophilisation — AAPS PharmSciTech, 2010-03-04.
- Abdelraheem A, Tukra R, Kazarin P, Sinanis MD, Topp EM, Alexeenko A, Peroulis D. Statistical electromagnetics for industrial pharmaceutical lyophilization — PNAS nexus, 2022-05-16.
- Prus-Walendziak W, Kozlowska J. Lyophilized Emulsions in the Form of 3D Porous Matrices as a Novel Material for Topical Application — Materials (Basel, Switzerland), 2021-02-17.
- Leys L, Nuytten G, Lammens J, Van Bockstal PJ, Corver J, Vervaet C, De Beer T. A NIR-Based Study of Desorption Kinetics during Continuous Spin Freeze-Drying — Pharmaceutics, 2021-12-16.
- Matejtschuk P, Bird C, Ezeajughi E, MacLellan-Gibson K, Wadhwa M. Impact of Formulation Choices on the Freeze-Drying of an Interleukin-6 Reference Material — Frontiers in molecular biosciences, 2022-07-04.
- Walendziak W, Douglas TEL, Kozlowska J. Design, Optimization, and Characterization of Freeze-Dried Emulsions Based on Sodium Alginate and Whey Protein Isolate Intended for Cosmetic and Dermatological Applications — ACS omega, 2025-06-03.
- Walendziak W, Douglas TEL, Kozlowska J. Physicochemical Properties of Freeze-Dried Bigel-Based Materials Composed of Sodium Alginate/Whey Protein Isolate Hydrogel and Ethylcellulose/Sunflower Oil Oleogel — Biomacromolecules, 2025-03-25.
- Ten Have R, Reubsaet K, van Herpen P, Kersten G, Amorij JP. Demonstrating Functional Equivalence of Pilot and Production Scale Freeze-Drying of BCG — PloS one, 2016-03-16.