Project 2–5: Nanomechanics of Metal Plasticity in Thin Films and Cylinders

Principal Investigator: Wei Cai (Stanford University)

Anatomy of a Dislocation

Metallic elements and alloys exist as crystalline materials: orderly three-dimensional arrays of atoms that form small particles called grains. Metallic grains adhere to one another to form the macroscale structures we recognize as paper clips, coins, gears, and sheet metal. Thin metal films and microscopic metal devices are only a few grains across, and they behave very differently from their larger counterparts. Metals differ from other crystalline materials such as ceramics or gemstones in their ability to deform— bend, flatten, or be drawn into wires—without breaking. Planes of metal atoms glide past each other like layers of ball bearings and settle into new posi- tions that maintain the overall integrity of the crystal structure.

This process is not perfect, however. Each time a piece of metal bends, planes of atoms slip past each other. In the process, some of these planes become misaligned to form structures called dislocations. Dislocations can travel through the structure, and when enough of them accumulate, a crack forms. In thin metal films, the distance from the interior to the surface is small, so the interaction of dislocations with the surface becomes a driving factor. Disloca- tions that reach the surface of the metal disappear, or “exit,” but can leave behind various types of structural imperfections, depending on the metal’s crystal lattice structure.

Two common structures for metallic crystals are body-centered cubic (BCC) and face-centered cubic (FCC). Metals with a BCC structure, including tungsten, chromium, and molybdenum, tend to be strong and moderately ductile. Soft, ductile metals such as aluminum, copper, nickel, and platinum assume an FCC configuration. (Brittle metals such as zinc, titanium, and magnesium take on a third structure, not addressed under the current AHPCRC project, known as hexagonal close packing.)

Metal micropillars are convenient stand-ins for macroscale test bars when the compression tests are performed at a very small scale. “Forests” of micropillars can be grown and cut from a thin film on a substrate using focused ion beams. Micro-tensile tests are much more challenging because one has to machine “grips” on top of the micro-pillar so that a holder can grab the pillar and pull on it.

Using Molecular Dynamics and Dislocation Dynamics simulations, Wei Cai’s group has discovered a counter- intuitive behavior of dislocations in BCC pillars. When stress is applied to BCC pillars, dislocations do not simply exit the pillar (as they do for FCC pillars). Instead, the dislocations self-replicate and send other dislocation moving in the opposite direction before the first one exits the surface. This means that a single dislocation nucleation event in a BCC pillar can produce a larger amount of plastic deformation than for a corresponding FCC pillar.

It is very difficult to verify these simulation results directly from in situ experiments, but some research groups have started to design experiments that could ultimately test this prediction.

Reference

Surface Controlled Dislocation Multiplication in Metal Micro-Pillars. C.R. Weinberger and W. Cai, Proceedings of the National Academy of Sciences, 105, 14304–14307, 2008. doi:10.1073/pnas.0806118105

Source: AHPCRC Bulletin, Vol. 1 No. 4 (2009)