Search

We consider the effect of a partially contaminated interface on the steady thermocapillary flow developed in a two-dimensional slot of finite extent. The contamination is due to the presence of an insoluble surfactant which is carried away by the flow and forms a region of stagnant surface. This problem, first studied in the classical theoretical paper by Carpenter and Homsy (1985, J. Fluid Mech. 155, 429), is revisited thanks to new experimental data. We show that there is a qualitative agreement between above theory and our experiments: two different regions simultaneously coexist on the surface, one of which is free from surfactant and subject to vigorous Marangoni flow, while the other is stagnant and subject to creeping flow with the surface velocity smaller about two orders of magnitude. We found, however, significant disagreement between theory predictions for the extent of a stagnant surface region and newly obtained experimental data. In this paper, we provide an explanation for this discrepancy demonstrating that the surface temperature distribution is far from suggested earlier. Another effect, not previously taken into account, is a possible phase transition experienced by the surfactant. We obtain a correct analytic solution for the position of the edge of the stagnation zone and compare it with the experimental data.