5. Experimental evidence for anisotropic surface melting on ice crystal
This theoretical aspect for the habit change of snow crystal gives an elegant explanation for the problem which has alluded many researchers for a long time. In order to examine this model, it is very important to study the ice crystal surfaces by the various methods. The existence of quasi-liquid layers on ice surfaces had been claimed by many researchers, in connection with many interesting phenomena related to the surface structures[19]. Those, however, produce indirect evidence for surface melting, and the exact structures and the dynamic properties of ice surfaces have remained illusive. Recently, studies of ice crystal surfaces by the modern experimental methods such as ellipsometry[26,27,28], X- ray scattering[29,30], Nuclear Magnetic Resonance(NMR)[31,32], and Atomic Force Microscopy (AFM)[33] have been carried out at the various laboratories. As a result, the qlls on ice surfaces have been detected by every experimental methods and their physical properties have been directly investigated.
For example, Fig. 7 displays the qll thickness as a function of temperature, which was obtained by our ellipsometric technique[27]. It should be noted that the transition layers were detected at temperatures above －2 and －4oC for both the {0001} and {1010} faces, respectively, and the layer thickness increased as the temperature approaches the bulk melting point. Furthermore, the index of refraction for the layer was also measured as 1.330, which was very close to that for the bulk water, 1.333, but slightly smaller. So we can conclude that the transition layers observed on the ice crystal surfaces are regarded as quasi-liquid layers originating from surface melting. In addition, we have observed a halo pattern of diffracted X-rays from ice surfaces above －2oC[29]. The interpretation is that the arrangement of water molecules in the transition layer should be disordered. Even though the absolute values of critical temperatures for surface melting obtained in these experiments are completely different from the theoretical argument proposed to explain the habit change of snow crystals, we can claim that these experiments provide direct qualitative evidence for anisotropic surface melting on an ice crystal.