| Condensation is initiated by nucleation, followed by growth, and finally by aging. Among these three steps, nucleation is the least well understood, hardest to predict, and hardest to measure. Most of the experimental investigation on nucleation has been performed using water, propanol and higher n-alcohols, and alkanes. Nucleation experiments using short chain length alcohols such as methanol and ethanol are less common and more problematic because these molecules associate in the gas phase to form small stable clusters.;The goals in this work are to obtain a detailed view of methanol condensation under the highly supersaturated conditions in a supersonic nozzle, and to measure methanol nucleation rates in this device. Through combined static pressure, spectroscopic temperature, Fourier Transform Infra-Red spectroscopy, and small angle x-ray scattering (SAXS) measurements we obtain a detailed picture of methanol condensation. We observe three distinct stages of methanol condensation in the supersonic nozzle. In the beginning, small methanol clusters form, increase in concentration, and evolve in size. This is followed by a period where the cluster size distribution remains relatively stable while the cluster concentration increases. Finally, as the vapor becomes supersaturated enough, liquid droplets form via nucleation and growth, consuming more monomer and reducing the cluster concentration. At the point where the liquid droplets first observed, up to 30% of the monomer has been depleted to form clusters. The amount of clusters formed is significantly more than that predicted by a simple model that describes the vapor as an equilibrium mixture of monomer, dimer, and tetramer. A simple calculation using an energy balance suggests a significant fraction of the clusters in the distribution are larger than the tetramer. Preliminary SAXS measurements suggest that the cluster distribution is centered around the hexamer. The detailed picture of methanol condensation is also critical to accurately measuring the nucleation rate in the nozzle. The methanol data appear consistent with data from longer chain alcohol measured in the same nozzle. In addition, the methanol data in the nozzle appear reasonably consistent with the methanol data measured in the nucleation pulse chamber. Finally, the measured nucleation rates are within one order of magnitude of the rates predicted by classical nucleation theory (CNT). |
