Site of hydrate nucleation

Long and Sloan (1996) performed a series of measurements to investigate the site of nucleation for natural gas and carbon dioxide hydrate initiation in a sapphire tube. 

A hydrate nucleating agent (precipitated amorphous silica) and a quiescent surface inhibitor (sodium dodecyl sulfate) were used in an attempt to initiate hydrates in the bulk phase. While the induction time (for detectable hydrate formation) was not predictable, in every case hydrate was initiated at a surface—usually at the vapor-water interface, but infrequently along the sides of the sapphire tube in the gas phase, and at the metal end-plate below the liquid phase. 

A number of other researchers have also confirmed that nucleation and subsequent growth typically occurs at the water-hydrocarbon interface for methane hydrate (Huo et al., 2001 0stergaard et al., 2001 Taylor, 2006) and carbon dioxide hydrate (Kimuro et al., 1993 Fujioka et al., 1994 Hirai et al., 1995 Mori, 1998). 

Molecular dynamics (MD) simulation studies also indicate that the initial formation of methane hydrate occurs preferentially near the water-methane interface where there is a significant concentration gradient (Moon et al., 2003). 

With such low concentrations of components available to form critical nuclei, hydrate formation seems unlikely in the bulk phases. However, at an interface where higher concentrations exist through adsorption (particularly at the vapor-liquid interface where both phases appear in abundance) cluster growth to a supercritical size is a more likely event. High mixing rates may cause interfacial gas + liquid + crystal structures to be dispersed within the liquid, giving the appearance of bulk nucleation from a surface effect. 

---