Leidenfrost temperature on an extremely smooth surface

Most existing heat transfer models on transition boiling employ the hypothesis on the mechanism limiting the so-called liquid-solid contact fraction that the activation of preexisting nuclei upon liquid-solid contact to be intermittent and localized and that conventional activation criteria can be used to predict the activation of preexisting nuclei in the wetted area upon contact. Supposing this hypothesis to be valid, the behavior of liquid-solid contact on an extremely smooth surface such as a single-crystal surface is expected to be very different from that on a metal surface of standard roughness. In this report, to examine the hypothesis, the Leidenfrost temperature is measured on single-crystal and metal plates. The maximum surface roughness of the former is 0.03 mu m, and that of the latter is 1.25 mu m. Results of the experiment show that the Leidenfrost temperatures on these two surfaces do not differ from each other as long as the surfaces are the same in wettability and thermal conductivity (or thermal diffusivity). Visual observation of the transient and localized wetted area on the single-crystal surface shows that numerous tiny dry spots are formed in the wetted area even at initial surface superheats lower than the minimum superheat to activate the preexisting nucleus of 0.03 mu m radius. These results imply that the hypothesis employed in most of the existing heat transfer models on transition boiling must be reexamined.