Results: Xxxxxxx Xxxx Breaking. During the atomistic simulation using the GAP/SW embedding method, the reversed contribution from temperature to the crack propagation velocity was found for the ther- mally activated and catastrophic loading regions. The GAP potential in the simulation has no chemical accuracy confirmed (see section 4.2.3) and thus the question is still open as to the bond breaking process. In the case of no experimental observation available to confirm about the findings, a theoretical model is thus designed to provide a different perspective to the underlying mechanism, which motivated the work to be described below. In this simplified atomic model with LJ interaction potential as depicted in Fig.3.6, we ignore the site correlation of bond breaking near the crack tip in the crack propagation directions, which is appropriate according to the result in GAP/SW. Pulling forces exerted on the edge atoms are used to atomically simulate the stress fields near the crack tip and separation of the bonding structure is carried out at a relative velocity of v . Due to the general atomistic features, this model is not restricted to Silicon and is designated to address the description for bond breaking in a group of brittle fracture system. As one of the significant factors affecting the bond breaking, the stiffness of the back bonds compared to the central bonds was also explored with the stiffness ratio defined as the proportion of the bond energy between the back bonds and the central bond. We adopted uniform bonding which corresponds to the stiffness ratio of 1.0 as shown in panel (a) of Fig.3.6 and triple back bonding with stiffness of ∼ 4/3, as in panel (b), in analogy to the tetrahedral bonding type in Silicon. Figure 3.6: A schematic plot showing the bond-breaking model with different back bond strength ratio with respect to the central bonding as illustrated in (a) and (b), respectively. The edge atoms in both case are considered to be moving with constant velocities.
