Scientists uncover how ice forms inside clouds

Water can remain liquid far below its familiar freezing point, but in Earth’s atmosphere it rarely gets the chance to cool that far. Tiny particles of mineral dust provide surfaces where water molecules can organize into ice, triggering the formation of cloud ice, snow and, ultimately, precipitation. Now, researchers at Bielefeld University and the University of Vienna, working with colleagues from the University of Helsinki, have identified at molecular resolution why one common mineral is exceptionally effective at starting this process. Their findings show that microcline can promote ice formation on a widespread, stable surface rather than relying on rare cracks, steps or other defects.

The discovery addresses a long-standing question in atmospheric science. Pure water does not normally freeze until it reaches approximately minus 38 degrees Celsius, a condition known as homogeneous freezing. In clouds, however, water droplets often freeze at warmer temperatures because airborne particles act as ice nucleators. Feldspars, a major group of rock-forming minerals present in atmospheric dust, are among the most important of these particles. Microcline, a potassium aluminum silicate, has repeatedly emerged as one of the most powerful ice-nucleating minerals, even though its chemical composition differs only slightly from that of related feldspars. Until now, scientists lacked a detailed explanation for its unusual activity.

The new study reveals that microcline’s ordinary (001) cleavage plane is sufficient to initiate ice formation. This surface is the most common and thermodynamically stable face of the mineral, meaning that it is energetically favored and widely exposed when microcline breaks apart. Earlier models had placed greater emphasis on unusual surface features, such as atomic steps, cracks and damaged regions, where water molecules might be expected to become trapped and organized. The researchers’ observations instead indicate that a broad, stable mineral surface can itself provide the molecular template required for ice nucleation.

At the heart of the process are aluminol groups—chemical units containing aluminum, oxygen and hydrogen—arranged across the microcline surface. These groups interact strongly with water molecules through hydrogen bonding. As water accumulates, the aluminol groups help hold the molecules in specific positions and orientations, reducing the molecular disorder that normally makes crystallization difficult. The surface effectively stabilizes the earliest ice-like clusters, allowing them to grow rather than dissolve back into disorganized liquid water. According to the study, microcline’s stable surface contains roughly twice as many of these groups as the corresponding surface of the closely related mineral sanidine.

The researchers also discovered that ice does not grow on microcline in the most straightforward orientation. The crystal lattice of ice is not simply aligned with one of the common, low-index surfaces of hexagonal ice. Instead, the ice adopts a less common, higher-index plane that matches the structure of the mineral at a fixed angle. This type of alignment is known as epitaxial growth. In epitaxy, one crystal guides the organization of another by providing a repeating arrangement of atomic or molecular binding sites. The finding suggests that uncommon faces of ice may be more important in atmospheric freezing than previously recognized.

Direct evidence for the mechanism came from high-resolution atomic force microscopy performed under ultra-high-vacuum conditions. The researchers cooled carefully prepared mineral samples and used the microscope to map the surface while observing the emergence of nanoscale ice clusters. Atomic force microscopy detects the forces between a sharp probe and the sample, allowing scientists to reconstruct surface structures with extraordinary precision. In this case, the technique made it possible to identify where the first clusters formed and how they developed on the microcline surface. Computer simulations independently reproduced the observed arrangement and supported the interpretation of the experimental images.

The comparison with sanidine provided an important control. Although sanidine has the same overall chemical composition as microcline, its surface structure differs because the atoms are arranged differently within the crystal. On sanidine, ice formation occurred mainly at step edges, consistent with conventional theories of heterogeneous nucleation. The contrast indicates that chemical composition alone cannot explain the behavior of feldspar minerals. The precise geometry and density of reactive surface groups matter just as much. Microcline’s ability to provide a regularly spaced network of hydrogen-bonding sites appears to allow ice to form across a stable terrace, while sanidine depends more heavily on localized defects.

The implications extend well beyond mineral physics. The temperature at which cloud droplets freeze influences the balance between liquid water and ice in clouds, their lifetime, their brightness and the way they reflect incoming sunlight. Ice-containing clouds can also alter the pathways by which precipitation develops, affecting when and where rain or snow reaches the ground. Climate models must therefore estimate how efficiently airborne particles initiate ice under changing atmospheric conditions. If common, stable mineral surfaces are more active than previously assumed, models based mainly on rare defects could underestimate the contribution of dust to cloud formation and climate feedbacks.

The study also demonstrates the value of examining atmospheric processes at the scale of individual molecules. A dust particle may be only a few micrometers across, yet its capacity to trigger ice depends on arrangements of atoms separated by less than a billionth of a meter. By connecting nanoscale surface chemistry with cloud-scale behavior, the researchers offer a more precise framework for understanding how geological materials influence the atmosphere. Their results suggest that the search for efficient ice nucleators should not focus exclusively on damaged or exotic surfaces. Common crystal faces, when decorated with the right chemical groups, may be powerful atmospheric actors in their own right.

The work, led by Dr Florian Schneider and Professor Dr Angelika Kühnle at Bielefeld University in collaboration with scientists at the University of Vienna and the University of Helsinki, was published in Nature Communications. The authors describe the direct visualization of nanoscale ice clusters as a methodological breakthrough and argue that their findings close a key knowledge gap concerning microcline’s exceptional performance. As mineral dust continues to circulate through the atmosphere, the molecular architecture of surfaces such as microcline may help determine whether supercooled cloud water remains liquid or suddenly becomes ice—an apparently small decision with consequences for weather, precipitation and the global climate.

News Publication Date: 25-Aug-2026

Web References: https://www.uni-bielefeld.de/fakultaeten/chemie/ag/pc1-kuehnle/ ; https://www.uni-bielefeld.de/forschung/profil/fokusbereiche/anbauen/ ; https://www.youtube.com/watch?v=iXU5epU4nmc

References: Nature Communications, DOI: 10.1038/s41467-026-76548-7