The origin of surface instabilities in fracture of brittle crystals or When cracks meet atoms

David Wang Auditorium, 3rd floor Dalia Maydan Bldg.
Liron Ben-Bashat Bergman, PhD candidate

Liron Ben-Bashat Bergman, PhD candidate
Department of Materials Science and Engineering
Technion, Haifa 3200003, Israel
Fracture surface instabilities and their origin in brittle crystals were impossible to detect so far due to the lack of high resolution fracture experiments. Over the last decade, we have developed such capabilities that enabled us to study the origin of surface instabilities down to the atomistic scale of the events.
Surface instabilities in the form of micron scale ridges in brittle single crystal silicon were investigated in our lab by cleaving macro scale silicon specimens along the {111} low energy cleavage plane under three-point bending (3PB). The specimens contained boron dopants with two distinct concentrations (1015 and 1019 atoms/cm-3). The fracture surface ridges were generated at crack speed below 1100 m/sec and their density has shown to increase significantly with increasing boron concentration. The origin of the above ridges was experimentally revealed by in situ fracture experiments of miniaturized bending specimens under ultra-high vacuum (UHV) and their surface analysis using scanning tunneling microscope (STM). These show that when a crack propagates at a speed of nearly 1200 m/sec it collides with individual boron atoms along the crack front, generating atomic height jogs which grow (similar to dislocations jogs) by over 3 orders of magnitude, and terminate as micron scale ridges. We then generalized this phenomenon by additional fracture cleavage experiments of silicon with oxygen as interstitials, and silicon and germanium doped with phosphorus and gallium, all revealed that surface instabilities are related to atomic lattice strain and material independent. Findings were supported by multi scaled quantum mechanical molecular mechanics calculations. Surprisingly, same collision but when cracks were propagated on the {110} low energy cleavage plane yielded local, undeveloped jog.
An energy balance describing crack stability was developed demonstrating the relationship between fracture surface undulations, lattice strain, crack speed and material anisotropy. The requirements for finding the critical speed dependent atomic strain are presented determining the origin of fracture surface instabilities at the lowest scale.
The fundamental query is whether fracture surface in ideal, defects free, crystal contains no surface instabilities, meaning, it is stable. We will answer this query intuitively, as an ideal dopant free crystals are not available.

Supervisor: Prof. Dov Sherman