Atomic origin of hysteresis during cyclic loading of Si due to bond rearrangements at the crack surfaces

Abstract

The atomistic origin of fatigue failure in micron-sized silicon devices is not fully understood. Two series of density-functional theory calculations on cubic diamond Si explore the effect of surface bond formation on crack healing in systems which exhibit strong surface reconstruction. Both series introduce a separation between Si(100) layers (i.e., the crack) and allow the ions to relax to their minimum-energy configuration. The initial surface ionic positions are either bulk terminated or 2 x 1 reconstructed. A plot of the energy versus the introduced separation reveals that once the surfaces reconstruct, the crack is no longer able to return to the equilibrium configuration. Rather, the healed crack interface contains defects which places the flawed energy minimum at a finite strain of 3% and an increased energy of 1.13 Jm2 relative to the equilibrium configuration. The irreversible plastic deformation supports the mechanism proposed by Kahn et al. [Science 298 1215 (2002)] that invokes mechanically induced subcritical cracking to explain the delayed onset of failure.

A Nanoscale Mechanism of Fatigue in Ionic Solids

“…Although the results shown here apply specifically to α-Al 2 O 3 , it is anticipated that similar atomic-scale fatigue mechanisms would influence a variety of other ionic solids and films. Such atomic-scale fatigue may impact the durability of ceramics and semiconductors in general, since the localized bonding in both materials families requires very specific lattice structures in order to maintain cohesive strength. Lastly, although the effects of water are not considered here, the dramatic impact of water adsorption on oxide surface energies , indicates it could play a crucial role in modifying local bonding at defect sites where crack propagation occurs.…”

First Principles and Multi-Scale Modeling of the Roles of Impurities and Dopants on Thermal Barrier Coating Failure

“…We continue to suggest how to improve them [1,2], as well thinking of spin-offs for both catalysis [3] and microelectronics [4,8]. The design principles we have proposed may help increase the stability and hence the lifetime of TBC's in the future.…”

“…During the last two years, we made significant advances both in developing new methods for describing condensed matter [6,7] and in uncovering the atomic level reasons for failure of thermal barrier coatings and of related materials [12], including fatigue [8,10] and stress buildup due to phase transformations [11] that may lead to spall. We continue to suggest how to improve them [1,2], as well thinking of spin-offs for both catalysis [3] and microelectronics [4,8].…”

“…In work related both to microelectronics and to TBCs, we explored the effect of cyclic tensile loading on the behavior of silicon [8] and alumina [10]. We explored the behavior of both in the threshold region of crack formation, i.e., where the crystal must choose between healing a defect through a uniform elastic strain of the material or creating two free surfaces.…”