Effect Of Small Size, Stress Localization And Stress Gradient On The Strength Of Silicon

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

This award supports an investigation of the failure behavior of silicon at small scales. Most micro-nano mechanical systems use silicon beams as their structural components. These beams are typically subjected to bending during operation. But silicon is brittle at room temperature. This limits the design space of silicon devices. Bending, however, localizes high stresses near the surface of the beams close to the anchors. In addition, the stresses decrease from the surface towards the middle plane of the beam, giving rise to stress gradients. The effects of small size, stress localization and stress gradient on the failure mechanisms of silicon remain elusive to date. Recent evidence suggests that silicon at small scale can be ductile (i.e., not brittle) at very high yield stresses. If so, then small size, stress localization and stress gradients together may offer the virtues of both ductility and high strength to silicon. Such failure resistance would present a yet untapped paradigm to the design space of silicon devices. A detailed understanding of the failure and deformation mechanisms of silicon at small scale under bending would be transformative for both semiconductor physics and industry, and will be a fundamental advance for the field of mechanics. The goal of this project is to explore the mechanics and mechanisms of deformation and failure in small silicon samples under bending by combining theory and experiments. The working hypothesis of the project is that dislocation is the primary mechanism of deformation in silicon under bending at small scale. Small samples are dislocation free. Small size and stress localization in bending offer high flaw tolerance against fracture. This is due to the low probability of flaw incidence in the small stressed region. Bending results in dislocation nucleation from the surface before any flaw induced fracture. These dislocations enter the bulk yielding the silicon. But the yield stress increases with decreasing size due to stress gradient. This hypothesis will be tested by undertaking three tasks: (1) mechanistic modeling and molecular dynamics simulations of silicon samples under bending, (2) bending experiments on micro-nano fabricated single crystal silicon samples with various sizes and at different temperatures, and (3) in situ bending experiments in transmission electron microscopes (TEM) to reveal the mechanisms of deformation (in collaboration with Max Planck Institute at Dusseldorf, Germany). A novel micro mechanical stage will be developed for tasks 2 and 3. The research will be integrated with education and outreach activities involving K-12 to graduate students.