Can solid surface energy be a predictor of ice nucleation ability?

Surface energy is one of the most important properties of materials that dictate many behaviors such as the spreading of water and ice nucleation. It is commonly thought that low surface energy leads to hydrophobicity and delays ice formation. Conventional studies mainly focus on rough surfaces where surface energy is amplified/reduced accordingly and it remains elusive to clearly decouple the effect of surface energy on ice nucleation. Here we experimentally construct the atomically smooth surfaces of halogen chemistry by varying silane self-assembled monolayers with exactly the same aliphatic chain but varied end groups of F, Cl, Br and I. By analyzing ice nucleation rate and free energy barrier based on the measured freezing temperature/delay time and the classical nucleation theory, we demonstrate that the ice nucleation efficiencies of various halogen chemistry surfaces increase in the order of Br, F, I and Cl, which is inconsistent with the surface energy sequence. The results indicate that solid surface energy is not a secure predictor of ice nucleation ability. By further analyzing the surface potentials of different halogen chemistry surfaces, we find that surface polarity plays a key role in ice nucleation maybe through impacting the orientation of interfacial water.