Abstract
Ice formation is a complex cooperative process typically catalyzed by the presence of an interface. While it has long been proposed that a surface must have a lattice parameter matching bulk ice to be an effective nucleator, this study challenges that assumption. By investigating a series of pseudomorphic tin-alloy surfaces (Sn on Pt, Pd, Rh, Cu, and Ni), researchers found that only the SnPt surface—which has a lattice parameter 7% larger than bulk ice—forms a stable wetting layer. This suggests that matching the oxygen-oxygen (O–O) spacing of the first adsorbed layer is more critical than matching the 3D bulk ice lattice.
The Challenge of Lattice Matching
For decades, the "lattice match" theory has dominated ice nucleation research, suggesting that materials mimicking the hexagonal lattice of bulk ice (Ih) are the most effective seeds. However, isolating the effect of lattice spacing is difficult because changing the substrate usually alters chemical and electronic properties simultaneously. This study employs a series of substitutional alloy surfaces to isolate the geometric effect of lattice parameter on whether water forms a smooth film (wetting) or beads up into droplets (clustering).
Experimental and Computational Methods
The researchers created $(\sqrt{3} \times \sqrt{3})R30^\circ$ Sn templates on five different metal crystals: Pt, Pd, Rh, Cu, and Ni. This provided a range of lattice constants from 2.49 Å (Ni) to 2.78 Å (Pt).
Characterization was performed using Temperature Programmed Desorption (TPD) to distinguish between wetting and clustering, while Low Energy Electron Diffraction (LEED) and Helium Atom Scattering observed structural order. Density Functional Theory (DFT) calculations modeled the binding energies and molecular orientations.
Key Insights: The "H-Down" Flattened Layer
A critical discovery was the orientation of the first water layer. Unlike the traditional "half-ice" bilayer found in bulk crystals, water on these metal surfaces forms a flattened structure.
In this configuration, uncoordinated hydrogen atoms point toward the metal surface ("H-down"), allowing the metal's electrons to screen the molecular dipole. This creates a stable bond that favors a specific O–O distance of approximately 2.72 Å. Surprisingly, surfaces like SnPd and SnRh, which closely match bulk ice spacing, failed to wet. Only SnPt, which matches the O–O distance required for this flattened network, supported a stable 2D film.
Results and Implications
- Wetting Specificity: Only SnPt(111) exhibited a stable 2D wetting layer. All other alloys resulted in immediate 3D ice clustering.
- The Pinning Paradox: A surface that perfectly templates the first layer of water (like SnPt) may actually be a poor ice nucleator. The strong, rigid binding "pins" the water into a geometry that cannot easily transition into the 3D bulk ice structure.
- Binding Energy: DFT showed that the binding energy of water on SnPt is approximately 0.600 eV, which is very close to the cohesive energy of bulk ice, explaining its unique stability among the series.
Conclusion
Lattice matching to bulk ice is an insufficient criterion for predicting ice nucleation. Instead, the stability of the initial water layer depends on matching the specific O–O distance of the adsorbed phase and the binding energy of the surface. Future research into ice nucleators should prioritize surface mobility and binding energy over simple geometric templates.
Summary updated