How glass breaks -- Damage explains the difference between surface and fracture energies in amorphous silica

The difference between free surface energy and fracture toughness in amorphous silica is studied via multi-scale simulations. We combine the homogenization of a molecular dynamics fracture model with a phase-field approach to track and quantify the various energy contributions. We clearly separate free surface energy localized as potential energy on the surface and damage diffusion over a 16-23 A range around the crack path. The plastic contribution is negligible. These findings, which clarify brittle fracture mechanisms in amorphous materials, align with toughness measurements in silica.

💡 Research Summary

**
The paper presents a comprehensive multiscale investigation of the long‑standing discrepancy between the free‑surface energy (γ) and the fracture energy (or fracture toughness, Gc) of amorphous silica glass. By coupling atomistic molecular dynamics (MD) simulations with a continuum phase‑field fracture framework, the authors quantify two distinct contributions to the total energy required for crack propagation: (1) the energy needed to create new free surfaces, and (2) the energy dissipated in a diffuse damage zone that extends roughly 16–23 Å around the crack path.

Atomistic Modelling
Silica samples of two sizes (100 Å cubic and 400 × 300 × 100 Å rectangular) are generated using the BKS potential with Wolf truncation for Coulomb interactions. After high‑temperature equilibration (3000 K) the systems are quenched at 10 K/ps to 10⁻⁵ K, yielding realistic glass structures. A pre‑existing rounded incision (100 Å length, 10 Å radius) is introduced, and a mode‑I K‑field displacement is applied quasi‑statically, increasing KI up to 2 MPa√m. The displacement increments are kept below 0.01 Å to ensure linear‑elastic response.

Coarse‑Graining and Damage Definition
Atomic positions, forces, and energies are coarse‑grained using a Gaussian kernel (width w) to obtain continuous fields: displacement u, logarithmic strain ε, Cauchy stress σ, potential energy density ψpot, and mass density ρ. Damage d is defined as the reduction of the elastic energy density relative to the undamaged state: ψel = (1‑d)² ψ+0 + ψ‑0, where ψ+0 and ψ‑0 are the tensile and compressive parts of the strain‑energy density. Compression is assumed damage‑free, so d ranges from 0 (pristine) to 1 (fully broken).

Free‑Surface Energy Measurement
Separate MD calculations cut the glass at various positions, replace periodic boundaries with free surfaces, and minimize the total energy. The surface energy is obtained from 2γ = ΔΨpot / (Lx Lz). This provides a direct atomistic estimate of the energy required to break Si‑O bonds and relax the newly created surfaces.

Diffuse Damage Zone
Analysis of the coarse‑grained elastic energy field ψε reveals a substantial reduction extending far beyond the immediate crack faces. By fitting the damage distribution to a Gaussian profile, the authors extract a characteristic damage width l of about 20 Å. Structural analysis links this diffuse zone to changes in the silicate ring network and variations in silicon coordination number, rather than to conventional plastic flow. The plastic contribution is found to be negligible.

Phase‑Field Calibration via FEMU
The phase‑field fracture model requires two material parameters: the critical fracture energy gc and an internal length scale lc (which regularizes the crack). To determine these, the authors employ a Finite‑Element Update (FEMU) scheme. They compute a damage field dMD from the atomistic coarse‑graining and compare it with a damage field dFEM obtained from a finite‑element phase‑field simulation using the same boundary conditions. By minimizing the norm of (dMD − dFEM) over 20 equally spaced slices along the thickness, they fit gc and lc. The resulting gc is approximately five times γ, matching experimental measurements of silica toughness, while lc coincides with the measured damage width.

Key Findings and Implications

1. The fracture energy of silica exceeds the free‑surface energy by a factor of ~5, which the authors attribute to a diffuse damage zone rather than plastic dissipation.
2. The damage zone is characterized by structural rearrangements (ring breaking, coordination changes) over a nanometric length scale, providing a physical basis for the internal length parameter used in phase‑field models.
3. Plastic deformation is essentially absent, confirming the brittle nature of silica while revealing a more complex fracture process than simple bond rupture.
4. The FEMU methodology demonstrates that a single snapshot of atomistic deformation can yield macroscopic fracture properties (crack length, gc, lc) without explicit boundary‑condition information.

Broader Impact
By clearly separating surface creation from diffuse damage, the study resolves a decades‑old puzzle in glass fracture mechanics and validates the use of phase‑field models for amorphous materials. The multiscale workflow—MD → coarse‑graining → damage quantification → phase‑field calibration—offers a template for investigating other brittle, non‑crystalline solids such as oxide glasses, chalcogenide glasses, or even certain polymeric glasses where plasticity is minimal. Moreover, the identification of a nanometric internal length scale linked to specific structural motifs opens avenues for designing tougher glasses through compositional or processing strategies that suppress or redistribute the damage zone.