Quantitative prediction of the fracture toughness of amorphous carbon from atomic-scale simulations

Fracture is the ultimate source of failure amorphous carbon (a-C) films, however it challenging to measure fracture properties a-C from nano-indentation tests and results reported experiments are not consistent. Here, we use atomic-scale simulations make quantitative mechanistic predictions on a-C. Systematic large-scale K-field controlled crack propagation performed for samples with densities $\rho=2.5, \, 3.0 \text{ } 3.5~\text{g/cm}^{3}$ created by liquid quenches a range quench rates $\dot{T}_q = 10 - 1000~\text{K/ps}$. The show that propagates nucleation, growth, coalescence voids. Distances $ \approx 1\, \text{nm}$ between nucleated voids result in brittle-like toughness. We growth criterion proposed Drugan, Rice \& Sham estimate steady-state toughness based our short crack-length simulations. values $2.4-6.0\,\text{MPa}\sqrt{\text{m}}$ initiation $3-10\,\text{MPa}\sqrt{\text{m}}$ within experimentally range. These findings demonstrate can provide quantitatively predictive even materials ductile mechanism.