Clickable Image Publications Undergrad Opportunities Research Research Group Research Group Home

Research Projects

Nucleation and Condensation

The nucleation and condensation experiments in my lab are conducted using the continuous flow supersonic nozzle apparatus illustrated below. The expansion rates in our nozzle are comparable to those found in turbomachinery, jet exhausts and steam turbines. In addition to examining pure and completely miscible systems, the design of the vapor production system also allows us to produce multicomponent gas mixtures for systems that exhibit a bulk miscibility gap.This is the device we use in the aerosol-SANS experiments to study the structure of multicomponent nanometer-sized liquid droplets. 

 

Our conventional Laval nozzles converge to the throat and then diverge monotonically in the supersonic region. A condensible gas mixture flowing through the nozzle quickly expands and cools (at ~1K/ms), creating a highly supersaturated state (S=10-100). In the absence of foreign condensation nuclei, spontaneous homogeneous nucleation creates new droplets that grow rapidly, depleting the vapor and quenching further particle formation. The transition is accompanied by latent heat release to the flow, and, thus, condensation is observed as the deviation of a state variable (pressure p, temperature T, or density) from its isentropic expansion value. The point of deviation is called the `onset of condensation’. At the nozzle exit the droplets have radii 10-20nm and their number concentration is about 1011-1012  per cm3
One of the most exciting new developments in the past year has been our ability to extract nucleation rates in supersonic nozzles directly from experimentally determined quantities for the first time. The idea is quite simple. We use Small Angle Neutron Scattering experiments to determine N by fitting the scattering spectrum assuming the droplets are log-normally distributed.  The pressure measurements determine the characteristic time Dt associated with the peak nucleation rate, and the corresponding supersaturation and temperature.  From the number density and characteristic time we can directly estimate the peak nucleation rate as J(S,T) = N/Dt.
The figure to the left summarizes the nucleation rates we have measured using a conventional nozzles. Three different nozzles were used to measure rations as a function of supersaturations. The solid lines are the predictions of  Classical Nucleation Theory modified by a temperature dependent correction factor that was developed by our collaborators Dr. Judith Wölk and Prof. Reinhard Strey at the University of Cologne, Germany.

Finally, we are interested in being able to measure concentrations in the gas phase directly using spectroscopic methods. We need this information in order to determine the overall composition of multicomponent droplets when less than 100% of the incoming material condenses. The neutron scattering length density of the aerosol is a strong function of composition, and uncertainty in composition translates directly into uncertainty in number density. To this end we have purchased a Tunable Diode Laser Absorption Spectrometer and have coupled it to a supersonic nozzle.In the future we hope to be able to determine the rotational temperature of the gas directly and thus place additional constraints on our data.