In complex mechanical systems, lubricants reduce friction and protect moving parts against wear. Ultra-thin film lubricants are prevalent in today's technological world in all areas where tolerances are tight. In order for lubricants to work properly, their viscosity must be suitable for the particular application. For example, high viscosity may be required in automobile engines whereas low viscosity is needed in biomedical applications. Numerous authors have demonstrated that the glass transition temperature can be altered by interactions with specific surfaces when the lubricant film is thin. Since the viscosity is an exponential function of temperature, we therefore postulated that these interactions may also affect the viscosity of thin films. To our knowledge, no commercial viscometer exists that can provide the very high shear required to measure films thinner than one micron. Hence, we proposed a new method for measuring viscosity of films in-situ near a surface. This method is based on the theory of Brochard for a liquid/liquid bi-layer.

When two liquid films are in contact, the energy of the bi-layer is minimized when the layer with the lowest surface tension is exposed. Hence, if the surface tension of the upper layer is higher than the lower one it will dewet by forming holes. . . .Hence, one can measure the viscosity of a lubricant on a surface by covering it with a more viscous, high surface energy polymer, and observing the dewetting velocity.

We tested this concept by constructing bi-layers where a low viscosity . . .film is spun cast on a silicon wafer and covered by a high viscosity polystyrene (PS) probe film. We first determined that the velocity is mainly a function of the PMMA viscosity and hence the concept worked.

From the previous work of Li we knew that the diffusion coefficient, D, of PMMA near a silicon interface decreased by two orders of magnitude with film thickness due to pinning of the polymer chains by surface interactions. Since the Stokes-Einstein equation allows one to predict . . ., we measured the viscosity of the PMMA films as a function of film thickness. We found that the viscosity increases dramatically as the films become thinner. In fact, the viscosity of the PMMA at 20nm became higher than the PS probe layer.

In order to obtain a microscopic view of the holes, we examined the bi-layers with scanning force microscopy. We first found that the depth of the dewetting holes was larger than the film thickness. Melt fracture occurs when bulk polymers are extruded and the center of the liquid flows at a rapid rate while the edges are adsorbed to the pipe walls. We therefore postulated that the dewetting was occurring in a plane below the polymer interfaces due to melt fracture. We hypothesized that the analogy of bulk melt fracture to thin films occurs when the polymer at the interface between the two films moves with the velocity of the opening hole, while the other end of the film is strongly adsorbed to the substrate. We proved this hypothesis by screening the silicon surface with a very thin layer of Polyvinyl Pyridine (PVP). Dewetting then occurred only at the interface and melt fracture was greatly decreased.

This research was supported by the Simons Foundation.

* Shira Billet and Dora Sosnowik shared top team prize of $100,000 in Siemens Westinghouse Science and Technology Competition. Congratulations!