The origin of interfacial turbulence, spontaneous agitation of the interface between two unequilibrated liquids, has been explained in terms of classical flow, diffusion, and surface processes. The essence of the explanation is the long-known though much neglected Marangoni effect, wherein movement in an interface is caused by longitudinal variations of interfacial tension. It is proposed that interfacial turbulence is a manifestation of hydrodynamic instability, which is touched off by ever present, small, random fluctuations about the interface.
A simplified mathematical model has been analyzed in order to detail the mechanism of the “interfacial engine” which supplies the mechanical energy of interfacial turbulence. In its present form the analysis incorporates several drastic simplifications, though ways of removing some of these have been suggested. The groundwork has been laid for the more elaborate analyses that are needed for a decisive test of the theory.
The analysis shows how some systems may be stable with solute transfer in one direction yet unstable with transfer in the opposite direction, a striking result. It also suggests that interfacial turbulence is usually promoted by (1) solute transfer out of the phase of higher viscosity, (2) solute transfer out of the phase in which its diffusivity is lower, (3) large differences in kinematic viscosity and solute diffusivity between the two phases, (4) steep concentration gradients near the interface, (5) interfacial tension highly sensitive to solute concentration, (6) low viscosities and diffusivities in both phases, (7) absence of surface-active agents, and (8) interfaces of large extent.
That some of these effects have been observed in the laboratory lends credence to the theory.