Glassy Shields Keep Freeze-Dried Probiotics Alive, Study Finds

Freeze-dried probiotics live or die by the microscopic architecture of the powder that surrounds them. A new study from researchers at Konkuk University, Kangwon National University, and collaborators in South Korea shows that whether the protective matrix around the cells is glassy or crystalline, and how quickly it succumbs to water-induced molecular mobility, can make the difference between losing half a log of viable bacteria and losing nearly two and a half. The work, published in Food Science and Biotechnology, offers a physically grounded explanation for why some probiotic powders survive months on a shelf while others quietly collapse.

The team, led by Sejun Park and corresponding author Mi-Jung Choi, used Leuconostoc mesenteroides, a lactic acid bacterium common in fermented foods, embedded in matrices built from whey protein isolate combined with either lactose or inulin. Whey protein isolate provides a protein scaffold, but the carbohydrate component determines much of the matrix’s thermal and hygroscopic behavior. The researchers deliberately manipulated how the samples were frozen before freeze-drying, because the freezing step sets the physical state of the solid matrix that will ultimately shelter the cells.

Two freezing protocols were compared. In the first, samples were plunged directly to −100°C in a single step. In the second, a two-step protocol first held the samples at −5°C, a temperature that encourages ice nucleation and slow crystal growth, before finishing at −100°C. In some experiments the nucleation step was seeded to control when ice crystals formed. X-ray diffraction and Fourier-transform infrared spectroscopy then revealed what each treatment had wrought: unseeded one-step freezing produced predominantly amorphous matrices, disordered solids in which molecules are frozen in place like a snapshot of the liquid state, whereas seeded two-step freezing produced matrices with clear crystalline features and a higher degree of molecular organization.

That distinction matters because amorphous and crystalline solids behave very differently toward water. Differential scanning calorimetry measured the glass transition temperature, the threshold at which an amorphous solid softens from a rigid glass into a rubbery, mobile state. Water sorption experiments tracked how much moisture each matrix absorbed at different relative humidities, and Gordon–Taylor analysis quantified how strongly that absorbed water depressed the glass transition temperature. Water is a notorious plasticizer: even small amounts slip between macromolecules, lowering the transition temperature and unlocking molecular motion that was previously locked out by the glass.

The carbohydrate identity proved decisive here. Lactose-based matrices were far more sensitive to water-induced glass transition depression, sagging toward rubbery behavior at comparatively modest moisture gains. Inulin-based matrices, by contrast, retained their glass transition temperature much more effectively and showed lower plasticization. Inulin, a longer-chain fructose polymer, simply holds its glassy character better in the presence of water, a property that has been exploited in other dried food and pharmaceutical systems but is now quantified here in the specific context of probiotic preservation with whey protein isolate.

The consequences for the bacteria were stark. Immediately after freeze-drying, cells housed in glassy matrices had lost only 0.50 to 0.70 log units of viability, while cells in crystalline matrices lost 2.48 to 2.65 log units, a difference of roughly two orders of magnitude in surviving colony-forming units. The crystallization process itself appears to be the culprit: as the matrix organizes into crystals, it expels and concentrates solutes, generates mechanical stresses, and strips away the intimate molecular contact that sugars and proteins need to stabilize bacterial membranes and proteins during drying. A crystal lattice, however orderly, is a poor bodyguard.

Storage told a parallel story. Over the storage period, glassy matrices continued to protect their cargo, with viability losses of only 0.41 to 0.77 log units, whereas crystalline matrices shed another 1.71 to 1.87 log units. The mechanism linking these numbers is molecular mobility. When absorbed water depresses the glass transition temperature toward or below the storage temperature, the matrix enters a rubbery regime in which molecules can diffuse, rearrange, and collide. In that regime, damaging reactions accelerate, and the physical barriers that immobilize and shield bacterial cells break down. The Williams–Landel–Ferry equation, a standard model for how relaxation times accelerate as temperature rises above the glass transition, connected the measured transition depressions directly to the observed inactivation rates.

What makes the study technically notable is that it treats the freezing protocol, the carbohydrate chemistry, and the storage environment as one coupled system rather than isolated variables. The degree of crystallinity imprinted during freezing determines the initial physical state; the carbohydrate type determines how fast that state degrades under humidity; and the WLF framework ties the resulting molecular mobility to bacterial death. This chain of causation gives formulators a set of levers: freeze in a way that preserves the amorphous glass, choose carbohydrates such as inulin that resist water plasticization, and package to keep moisture out, and the same probiotic strain can be expected to survive dramatically longer.

The findings arrive at a moment when probiotic powders, functional foods, and microbiome therapies are a fast-growing market, and shelf stability remains a persistent bottleneck. Live bacteria are fragile cargo, and the industry has long relied on empirical trial and error to find protective formulations. By showing that the glassy state, quantified through calorimetry, sorption isotherms, and spectroscopy, is a predictive criterion for both drying survival and storage survival, the Konkuk team adds rigor to what has often been craft knowledge. The work also echoes and extends earlier observations that dried bacterial cells survive in proportion to how deeply their surroundings are locked below the glass transition.

There are, of course, practical caveats. The study examined one organism and one protein-carbohydrate platform, and real products face fluctuating temperatures and humidities that no single isotherm can capture. Crystalline matrices, for all their poor performance here, may offer advantages in other contexts, such as controlled release or reduced hygroscopicity. But the central message is hard to escape: for freeze-dried probiotics, the enemy is not cold or vacuum but mobility, and mobility is water’s gift. Keep the matrix glassy, keep it dry, and the bacteria keep their count. The research was supported by the Basic Research Program of the National Research Foundation of Korea, funded by the Ministry of Education.

Subject of Research: Stability of freeze-dried probiotics in whey protein isolate-carbohydrate matrices as governed by matrix physical state and water-induced molecular mobility

Article Title: Matrix physical state and water-induced molecular mobility influence the stability of freeze-dried probiotics in whey protein isolate-carbohydrate matrices

Article References: Park, S., Kim, J., Song, H., Kim, S.-H., Jo, Y.-J., & Choi, M.-J. (2026). Matrix physical state and water-induced molecular mobility influence the stability of freeze-dried probiotics in whey protein isolate-carbohydrate matrices. Food Science and Biotechnology. https://doi.org/10.1007/s10068-026-02315-5

Image Credits: AI Generated

DOI: 10.1007/s10068-026-02315-5

Keywords: freeze-drying, probiotics, glass transition temperature, molecular mobility, whey protein isolate, inulin, lactose, Leuconostoc mesenteroides, water plasticization, X-ray diffraction, differential scanning calorimetry, WLF model