Multidimensional Nucleation from Vapor Mixtures and Polymer Solutions

Long thought to be rare in the atmosphere because particle concentrations are high enough to scavenge vapors and keep vapor supersaturations too low, homogeneous nucleation has recently been shown to nearly everywhere that measurements have been made using state of the art techniques. Ubiquitous atmospheric nucleation is attributed to binary nucleation of sulfuric acid and water vapors, or ternary nucleation when ammonia is also present. Binary nucleation is generally described in terms of passage over a saddle point in the cluster free energy surface that describes the distribution of clusters in a hypothetical constrained-equilibrium with the supersaturated vapor mixture. It will be shown that the constrained-equilibrium assumption is an unnecessary artifice, and that eliminating that assumption resolves other problems with classical binary nucleation theory. Multidimensional nucleation that leads to saddle point crossing in nucleation also occurs when anisotropic crystals are formed as we shall show with experimental observations of nucleation of polymer crystals from solution in levitated small (nanogram) droplets. When the thermodynamic activity of the solvent vapor surrounding the levitated particle is changed, two phase transitions are observed: deliquescence leading to complete dissolution of soluble materials at high vapor activity; and efflorescence in which the solute crystallizes from supersaturated solution at significantly lower vapor activity. Aqueous salt solutions exhibit the expected stochastic homogeneous nucleation process, but the solvent activities at which poly(ethylene oxide) solutions both deliquesces and effloresces vary depending upon the detailed history of the solution phase. Neither foreign seeds nor homogeneous nucleation can explain this behavior, but a model of the surface free energy of the cluster suggests an explanation.