Ice nucleation on BaF2(111).

The mechanism of heterogeneous ice nucleation on inorganic substrates is not well understood despite work on AgI and other materials over the past 50 years. We have selected BaF(2) as a model substrate for study since its (111) surface makes a near perfect match with the lattice of the basal face of I(h) ice and would appear to be an ideal nucleating agent. Two series of experiments were undertaken. In one, nucleation of thin film water formed from deposition of vapor on BaF(2)(111) faces was explored with the finding that supercooling to -30 degrees C was required before freezing occurred. In the other series, nucleation of liquid water on submerged BaF(2) crystals was studied. Here supercooling to -15 degrees C was needed before ice formed. The reason why BaF(2) is such a poor nucleating agent contains clues to realistic mechanisms of heterogeneous nucleation. Our explanation of these results follows the model of Fletcher [J. Chem. Phys. 29, 572 (1958)] who showed that heterogeneous ice nucleating ability depends on how well ice wets a substrate. In this view, a smooth BaF(2)(111) face is poor at nucleation because ice only partially wets its surface. In an extension of Fletcher's model, our calculations, consistent with the experimental results demonstrate that the pitting of a submerged BaF(2) crystal dramatically improves its ice nucleating ability.