This Crystal Doesn’t Melt Like Ice: Physicists Capture a Strange New Phase

Why It Matters

The discovery reshapes fundamental understanding of phase transitions in 2D materials, opening pathways for engineered melting behavior in nanotechnologies and advanced electronics.

Key Takeaways

• First direct observation of hexatic phase in covalent 2D crystal.
• AI‑driven analysis tracked atomic movements during melting.
• Hexatic‑to‑liquid transition found to be abrupt, contrary to theory.
• Findings expand understanding of 2D phase transitions for nanotech.

Pulse Analysis

Melting in bulk solids is usually a sudden, first‑order transition, but theory has long predicted that atomically thin layers can melt via an intermediate hexatic phase. First described in the 1970s, the hexatic state retains orientational order while losing translational symmetry, a hybrid that only appears when a material is confined to two dimensions. Experimental confirmation has been limited to colloidal spheres and simulations, leaving a gap in our knowledge of how real covalent crystals behave under extreme thinness. This behavior also influences thermal stability in emerging 2D technologies.

The University of Vienna team captured the elusive phase by sandwiching a single‑layer silver iodide crystal between graphene sheets and heating it inside a scanning transmission electron microscope equipped with a Protochips Fusion stage. Temperatures were raised past 1100 °C, allowing the researchers to record thousands of atomic‑resolution frames. A custom neural‑network, trained on simulated datasets, identified each atom’s position and bond angles, revealing a narrow 25 °C window where the crystal entered the hexatic state before liquefying. Electron diffraction corroborated the intermediate ordering. The graphene encapsulation prevented oxidation, preserving crystal integrity throughout the heating cycle.

The observation overturns the conventional Kosterlitz‑Thouless‑Halperin‑Nelson‑Young scenario, which predicts continuous transitions on both sides of the hexatic phase. Instead, the researchers saw a continuous solid‑to‑hexatic shift followed by an abrupt hexatic‑to‑liquid jump, suggesting new physics in covalent two‑dimensional systems. Beyond fundamental science, the ability to engineer and monitor such intermediate states could inform the design of flexible electronics, sensors, and nanoscale devices where controlled melting or phase‑change is advantageous. Moreover, the successful integration of AI‑driven image analysis showcases a powerful workflow for future atomic‑scale investigations. These insights may accelerate the development of phase‑change memory and reconfigurable nanomaterials.