Scientists Pin Down When Drying Droplets Turn Sticky, Boosting Powder Production

Every year, the food and pharmaceutical industries transform staggering volumes of liquids into powders by spray drying, a process in which a liquid feed is atomized into a hot chamber and dried within seconds. The quality of the resulting powder depends heavily on agglomeration, the deliberate clumping of particles into larger, porous structures that dissolve more easily and flow better. Agglomeration hinges on a deceptively simple question: when does a drying droplet stick to a passing particle, and when does it bounce away? A new study published in Current Research in Food Science by S.W.C. de Leeuw and colleagues at Wageningen University & Research provides one of the most quantitative answers yet, combining high-speed imaging, infrared thermography and predictive modeling to map the stickiness of drying droplets across a range of drying conditions.

The stakes are higher than they might appear. In the nozzle zone of an industrial spray dryer, freshly formed droplets collide with so-called fines, tiny dry particles smaller than about 100 micrometers that are recirculated from the fluidized bed or cyclone back into the chamber. When a wet droplet meets a fine, the particle merges completely into the liquid. As the droplet surface partially dries, collisions instead produce sticking, the desired outcome that builds agglomerates. Dry too far, and the droplet surface becomes glassy and rigid, so particles bounce off and are lost as waste. The researchers set out to determine exactly where that sticky window lies, and whether it can be predicted from first principles rather than empirical rules.

The team used a sessile single droplet drying platform, in which droplets of roughly 400 micrometers were dispensed onto a surface and dried by a controlled air stream. Maltodextrin with a dextrose equivalence of 38 served as the model material, chosen because its diffusivity and sorption behavior are well characterized, making it ideal for physical modeling. Glass beads in the 70 to 110 micrometer range acted as stand-ins for fines particles, an approach validated in earlier work showing that collision outcomes are governed primarily by the state of the drying droplet rather than the identity of the collision partner.

A central concept in the study is the locking point, the moment when a drying droplet stops behaving like an ideal shrinking sphere and begins to deviate from its spherical shape. This deviation signals the onset of skin formation, when evaporation at the surface outpaces the diffusion of solute from the interior, creating a steep concentration gradient and a mechanically distinct outer layer. The researchers found that the locking point shifts linearly with drying conditions: raising the relative humidity of the drying air from 0 to 15 percent at 80 degrees Celsius delayed locking from about 6.6 to 11.8 seconds, while raising the air temperature from 70 to 90 degrees Celsius at zero humidity shortened it from roughly 9.4 to 5.1 seconds. Crucially, the normalized droplet radius at locking remained statistically unchanged across all conditions, hovering near 0.85 of the initial radius, suggesting that a critical surface solute concentration triggers skin formation regardless of how fast the droplet dries.

To probe stickiness directly, the team fired glass beads at droplets at controlled times during drying and classified each collision as merging, sticking or bouncing. The results were strikingly consistent: across relative humidities from 0 to 15 percent, sticking collisions clustered between 0.75 and 1.5 times the locking point time. Increasing humidity delayed the entire sequence of collision regimes but did not fundamentally change its structure. This confirms and generalizes earlier findings from experiments with different maltodextrin formulations and particle shapes, pointing to the state of the droplet skin, not the drying schedule itself, as the dominant factor determining collision outcomes.

The experimental work was paired with a one-dimensional droplet drying model that solves coupled heat and mass transfer equations in a Lagrangian reference frame, discretizing the droplet into 40 concentric shells. Water diffuses between shells while solids remain fixed, and the model tracks the evolving glass transition temperature at the droplet surface using the Couchman-Karasz relation, which captures how moisture depresses the glass transition. The model was validated against two independent measurements: the volume decrease of droplets up to the locking point, captured by automated image analysis of high-speed video, and the droplet temperature over time, measured with an infrared camera through a special AMTIR-1 window at 125 frames per second.

The thermography revealed a familiar but now precisely quantified pattern. Droplets initially hold a constant temperature set by evaporative cooling, which rises with ambient humidity because higher moisture in the air slows evaporation. Measured constant temperatures of 28, 33, 39 and 44 degrees Celsius at 0, 5, 10 and 15 percent relative humidity matched the corresponding wet bulb temperatures. Once the skin forms and evaporation slows, the droplet temperature climbs toward the air temperature. The model reproduced these trends well, with minor deviations attributed to the difficulty of characterizing the thermal contact between droplet and drying surface, a resistance the team captured with a single fitted parameter.

The key predictive insight emerged when the researchers plotted collision outcomes against the ratio of glass transition temperature to actual droplet temperature at the droplet surface, a dimensionless measure of how close the skin is to becoming glassy. Merging dominated below a ratio of about 0.6, when the droplet is still liquid-like. Sticking occurred between roughly 0.55 and 0.9, with the densest concentration of sticking collisions between 0.6 and 0.8. Bouncing took over above 0.8, as the skin approached the glassy state. Some overlap between regimes was expected and observed, since drying droplets develop spatially heterogeneous surfaces with softer and stiffer patches, and buckling after locking creates irregular geometries that influence individual collisions.

When the researchers normalized drying time by the locking point time, the curves for all tested humidities and temperatures nearly collapsed onto a single master curve, with the glass transition ratio reaching about 0.6 at locking. This collapse means the stickiness regime can be predicted for new drying conditions without running collision experiments for each one: model the droplet, find the locking point, and the sticky window follows as 0.75 to 1.5 times that time. The authors propose this as a general methodology for formulations where glass transition behavior is well defined, while noting limitations for materials such as globular proteins, for very high feed solids concentrations, and for the higher air temperatures used industrially, where the rapidly shortening locking time makes visual determination harder.

The practical payoff could be substantial. Better control of the nozzle zone means manufacturers can tune air temperature, humidity and fines recirculation to maximize agglomeration, improving powder solubility and reducing the fraction of product lost as unagglomerated fines. Because the framework rests on fundamental heat and mass transfer rather than formulation-specific calibration, it offers a route from single droplet physics to pilot and industrial scale spray drying. For an industry that dries everything from milk and coffee to enzymes and probiotics, knowing precisely when a droplet will stick, and being able to compute it in advance, turns a long-standing empirical art into a predictive science.

Subject of Research: Predicting the stickiness regime of drying maltodextrin droplets during spray drying using single droplet experiments and heat and mass transfer modeling

Article Title: Evaluating and predicting the stickiness regime of drying droplets across drying conditions

Article References: de Leeuw, S., Bekebrede, J., Schutyser, M., & van der Sman, R. (2026). Evaluating and predicting the stickiness regime of drying droplets across drying conditions. Current Research in Food Science, Article 101588. https://doi.org/10.1016/j.crfs.2026.101588

Image Credits: AI Generated

DOI: 10.1016/j.crfs.2026.101588

Keywords: spray drying, droplet drying, stickiness, agglomeration, glass transition, maltodextrin, locking point, thermography, mass transfer, food powders, collision regimes, relative humidity