The Brittle-Ductile Transition of Silicon

At the preceding workshop 1981 on Corsika one of the authors analysed the early measurements of St. John (1) on the b/d transition of pre-cleaved Si single crystals. The analysis was based on the assumption that dislocations can be easily produced at the crack-tip in an otherwise perfect crystal. The condition for ductile behaviour was that the dislocation could blunt the crack fast enough in comparison with the built-up of the stress intensity at a constant crack opening rate δ so that the critical KIC would never be reached. The theory of the b/d transition temperature TC made use of the linear high stress-dislocation velocity behavior that is well documented in t he literature but contained two adjustible parameters: One is the distance Δ between dislocations emitted in sequence from the crack tip (the width of the “dislocation-free zone”). This is related to the “friction stress” τ*, to the level of which the stress ahead of the crack is reduced by the shielding of the dislocations emitted (τ* = Gb/ Δ, G = shear modulus). The second parameter is KIC at TC, taken to be several times the low T value KCO = 0.9 MPa √m according to the experiments (1,2). The rate-of-blunting -and shielding criterion for the b/d transition rather than that of dislocation formation at the crack tip is in contradiction to the classical Rice-Thomson theory (3) which calculates a very high activation energy of dislocation creation.