Stability-filtered lattice matching for the discovery of supercooling nucleants in gallium and its low-melting alloys

Gallium and its low-melting alloys possess a high latent heat of fusion and melt near room temperature, yet they supercool by several tens of kelvin before solidification, rendering freezing unreliable and limiting their application as phase change materials. Heterogeneous nucleants are commonly selected by lattice matching to the solidifying phase. However, lattice matching alone is insufficient for gallium because its strong oxophilicity causes the reduction of candidate oxides before they can act as effective nucleants. We combine a classical nucleation model with a thermodynamic stability criterion accounting for lattice matching and chemical stability in the melt. Using the literature value of the gallium interfacial energy (55 mJ m −2 ), the model reproduces the measured homogeneous supercooling of pure gallium (68 K versus 67.8 K) and predicts the observed nucleant ranking. The stability criterion shows that the best reported oxide, TeO₂, is unstable in liquid gallium, reducing to GaTe and β-Ga₂O₃, both poor crystallographic templates. Screening oxides, carbides, and nitrides identifies HfN and ScN as the strongest templates, while first-principles calculations of the work of adhesion show stronger interfacial bonding than the ZrN benchmark. Yet these strongest templates do not meet the remaining requirements: an effective nucleant must also be wetted by the melt and have a density close to that of liquid gallium to remain dispersed. HfN and ScN satisfy neither condition. Thus none of the screened material satisfies all four requirements simultaneously. Nucleant selection requires optimisation of lattice matching, chemical stability, wetting, and density, with a transient cell model defining the residual supercooling below the thermal margin between the melting point and the cold plate.