From macro fracture energy to micro bond breaking mechanisms – Shorter is tougher

We show a novel fracture behavior and properties of brittle materials not previously explored. These were made possible when merging macro to micro in fracture. Our findings are based on macro-scale fracture cleavage experiments of dynamic cracks propagating in brittle single-crystal silicon specimens, focusing on the crack's energy-speed relationships, and the fine details of the crack front. From the micro-scale, we developed an atomistic model for bond-breaking mechanisms. These mechanisms are in the form of low-energy kink advance (migration) and high-energy kink formation (nucleation) along the crack front. The energy release rate (ERR) at crack initiation, G₀, and its derivative, dG₀/da≡Θ, in particular, play a major role in the novel behavior and properties. G₀ and Θ are influenced by the precrack length, a₀. The macro-scale experiments, the micro-scale atomistic model, and the gradient of the ERR, Θ, enabled the conclusions portrayed in this work. We identified that the cleavage energy is not a constant but bounded by the Griffith Barrier as the lower bound while the upper bound (apparently the lattice-trapping barrier) may reach 3 times the lower bound. We further suggest that a brittle material can be envisaged as comprising a pseudo-R-Curve behavior mechanism typical of metallic materials. A new and essential fracture mechanism was identified, which we term ‘quasi-propagation’, a transitional mechanism between initiation and propagation. During this mechanism, the sequence of the bond-breaking mechanisms is varying, causing an increase in the macroscale cleavage energy. The evaluated cleavage energy shows another novel behavior, that of ‘shorter is tougher’, namely, shorter cracks require higher energy to propagate, and, therefore, specimens with shorter cracks are stronger than that predicted by Griffith's theory.