(4dn) Developing the Next Generation of Tools for the Study of Nucleation and Metastability in Liquids

Nucleation - the initiation of many types of phase transitions ? is an important phenomenon, but one difficult to study experimentally. At the molecular scale, nucleation is a rare event that can involve tens or hundreds of molecules and we do not have yet good experimental and computational tools to investigate it. While the outcome of nucleation (creation of a new phase) is easily observable macroscopically, macroscopic investigations of nucleation are difficult because nucleation is often a stochastic phenomenon. A nucleation experiment must be repeated many times to acquire good statistical data. Microfluidic technology has characteristics that make it an attractive choice as a platform for the study of nucleation in liquids: it handles small amounts of liquids that can be kept very clean (creating reproducible experimental conditions), and the typical duration of experiments is short (allowing a high repetition rate). Several groups have already used microfluidics to investigate how crystals nucleate from solutions. We had chosen to look at a different class of problems ? the nucleation of solids from supercooled liquids ? and we have developed an instrument for the study of ice nucleation in supercooled water. This apparatus has the highest data throughput and state-of-the-art accuracy among instruments designed for the study of ice nucleation. We have used it to measure with the rate of homogenous nucleation in pure water, to study the effect of external electric fields on homogenous ice nucleation, and to measure the temperature dependence of heterogeneous ice nucleation due to ice-nucleating particles immersed in water. The apparatus can investigate the freezing of hydrocarbon liquids as well; we have supercooled and frozen drops of alcohols and alkanes. In all these measurements we were able to achieve well-controlled experimental conditions such as temperatures and cooling rates, high measurement accuracy, and better statistics than those achieved previously.