NUCLEATION IN CONDENSATION

Nucleation in condensation is defined to occur by local fluctuations within a supersaturated system. These density fluctuations are sufficient for local volume elements to increase from the supersaturated parent state density to the density at the spinodal boundary. If i(o) monomers are required to form a nucleus, then, for delta-i(o) monomers in a fluctuation, spinodal decomposition to (delta - 1)i(o) vapor atoms of parent density and i(o) atoms of a single liquidlike cluster may occur when delta-i(o)upsilon(s) = (delta - 1)i(o)upsilon(p)) + i(o)upsilon(c). Delta is a multiplier determined by conservation of matter, and upsilon(i) is the appropriate specific volume. If the i(o) atoms in the liquid cluster are in unstable equilibrium with the parent vapor, nucleation will have occurred. A capillarity approximation which depends on the density difference across the interface (i.e., a Macleod-Sugden capillarity) and the classical Kelvin equation provide a measure of the size of fluctuation, delta-i(o) monomers, which must occur to yield a nucleus. Fluctuation theory provides a measure for the rate of occurrence of density fluctuations containing delta-i(o) monomers from the parent phase density to the spinode boundary. Spinodal decomposition and completion time theory provides a measure of the fraction of fluctuations which survive long enough to decompose to a droplet within a supersaturated parent phase during an experimental observation. The nucleation rate is determined by application of all three concepts, and this theory is supported by the best experimental data available on nucleation during condensation.