Lowering the substrate onto the liquid -- instead of dropping the liquid -- assures even spreading. When the liquid is molten metal and the solid surface is hot, the standard equation for describing how liquid drops wet solid surfaces fares poorly.

A 200-year-old scientific mystery has been solved, and the solution holds major significance for one of the most basic and ubiquitous of all industrial processes, the bonding of liquid metals to the surfaces of other materials.

Researchers at the Lawrence Berkeley National Laboratory (Berkeley Lab) have discovered why the standard equation for describing how liquid drops wet solid surfaces fares poorly when the liquid is molten metal and the solid surface becomes hot. For their achievement, the research team has received a $50,000 award from the U.S. Department of Energy's Office of Science.

"The answer to the mystery is that in the interaction between liquid metals and solid surfaces at high temperatures during the process of wetting, ridges are formed on the solid surface that can greatly slow or even stop the spread of the liquid," says Antoni Tomsia, a materials scientist with Berkeley Lab's Materials Sciences Division (MSD) who, along with MSD colleagues Eduardo Saiz and Rowland Cannon, specializes in the study of interfaces between metals and ceramics.

The three researchers believe that the ridges are formed in response to capillary forces and evolve through diffusion and solution-precipitation processes. Over the past couple of years, working with a unique high-vacuum wetting furnace equipped with a camera that can record images at the rate of 2,000 frames per second, the Berkeley Lab scientists have been able to make movies that revealed the formation of ridges on the surfaces of ceramic materials. Ridges formed under certain conditions when the surfaces were wetted with drops of various molten metals, including nickel, copper, gold and aluminum.

Under unusual wetting conditions, ridges from a few nanometers to several hundred nanometers in height can form. At high temperatures the ridges may exceed atomic dimensions and limit the speed at which the liquid spreads.

"We have found that in cases where unusual wetting behavior occurs, small ridges ranging in height from a few to several hundred nanometers are formed," says Saiz. "The ridges form in response to the vertical component of force from surface tension. At lower temperatures these are very small elastic distortions, but at higher temperatures, where diffusion or solution-precipitation can occur, the ridges can readily exceed atomic dimensions and can limit the spreading velocity of the liquid."

The process by which liquids "wet" or spread across the surface of a solid is especially important when the liquid is a metal, because it plays a key role in such routine but crucial tasks as soldering, brazing, coating, and composite processing. Consider, for example, where the computing industry would be if metal connecting pins could not be brazed to microprocessing chips. Or, what the lighting industry would do if metal contacts could not be joined to glass bulbs.

"Wetting matters a lot in a great many manufacturing processes," says Tomsia, "but its importance is overlooked until there's a problem."

For the wetting of solid surfaces with drops of liquid metals, the problem has been that the standard means of predicting how the wetting will proceed, the Young-Dupré equation, is not accurate.

"Based on conventional fluid mechanics, you would expect a drop of liquid metal to spread over a ceramic surface and achieve the equilibrium contact angle given by the Young-Dupré equation within a fraction of a second," says Cannon. "However, in real life the wetting process can take minutes, hours, or even days."

This delayed response is called "hysteresis," and though scientists have been well aware of its occurrence in high-temperature wetting, until the experiments at Berkeley Lab no one had been able to explain it.

The Young-Dupré equation is named for Thomas Young, who in 1805 described the shape of a drop of liquid in contact with a rigid surface, based on the angle of contact at the triple junction where liquid, solid, and vapor meet, and for Lewis Dupré, who in 1855 wrote the equation that relates adhesion and surface energies to equilibrium contact angles. Their equation is used to predict the degree to which a liquid will either spread out across the surface of a rigid solid or bead up.

These predictions work well when the liquid is water, such as characterizing the pattern of raindrops on windshields, but not for liquid metals when the temperature is within 20 to 25 percent of the substrate's melting point, a frequent occurrence in the manufacturing of microprocessing chips.

"The higher the temperature, the greater the size of the ridges that are formed and the slower the velocity of the liquid's spreading," says Cannon. "If the spreading velocity slows as the liquid approaches the final equlibrium contact angle, it raises the probability that ridges will be formed. However, guided by our own theory, which predicted the ridges, we now know when equilibrium contact angles actually satisfy the Young-Dupré equation and can be used to deduce fundamental thermodynamic surface energies."

Berkeley Lab scientists (from left) Rowland Cannon, Eduardo Saiz, and Antoni Tomsia used a unique camera-equipped wetting furnace to uncover a 200-year-old secret about drops of molten metals on solid surfaces.

Tomsia, Saiz, and Cannon conducted their wetting experiments in a vacuum (at 10^–6 torr) or in purified gas at temperatures up to about 1,600 degrees Celsius. Substrates of high purity polycrystalline alumina were diamond-polished to make them extremely smooth, then brought into contact with molten liquid droplets. The triple junction line where the liquid, substrate surface, and vapor from the droplet area come into contact were studied, and then the droplets were removed from the substrate by acid etching. The researchers imaged the topography of the wetted region using atomic force microscopy -- and discovered the ridges predicted by their theory.

"The Young-Dupré equation assumed that solid surface substrates were rigid and insoluble," says Tomsia. "We now know that if you put liquid metal on a rigid substrate it is a bit like putting hot oil on water. You will get perturbations on the surface."

The Berkeley Lab researchers are now developing a new model for describing the dynamics of wetting under conditions where the Young-Dupré equation fails. They are also looking into ways to manipulate those conditions so that wetting can be better controlled.

"Liquid metals don't wet ceramic surfaces the way we would like them to for doing such things as manufacturing microprocessing chips," says Tomsia, "so we're looking for ways to trick the metals into wetting these surfaces faster and more efficiently."

This research was funded by the Office of Basic Energy Sciences, within the Office of Science at the U.S. Department of Energy.

Additional information

• More about the behavior of liquid metals on ceramic surfaces
• Other research by Tomsia and Saiz
• More about the Office of Basic Energy Sciences