Corrosion sous contrainte des verres de silice – CORCOSIL

The major goal of CORCOSIL is to shed light on the mechanisms of slow crack propagation in glasses at the nanoscale. In several cases, phenomenological equations such as Wiederhorn's can explain the dependence of crack velocity on stress and on environmental parameters. Yet the stress-corrosion mechanisms that occur at the crack tip are still debated, and their modelling at the nanometer scale is necessary to relate phenomenological parameters to the specific composition and structure of glasses. The teams of the Laboratoire des Colloïdes, Verres et Nanomatériaux (LCVN) of Montpellier and of the Service de Physique et Chimie des Surfaces et des Interfaces (SPCSI) at CEA-Saclay have recently performed, in collaboration, in situ Atomic Force Microscopy (AFM) observations of the stress corrosion of silicate glasses, which have completely changed the classical picture and resulted into an exceptional visibility on the international scene. Indeed, these teams have shown that fracture of silicate glasses proceeds through the nucleation, growth and coalescence of damage cavities ahead of the main crack tip, while classical models predict the successive breaking of atomic bonds right at the crack tip, like in cleavage. These results, at the very centre of discussions on the mechanisms of crack propagation in glass, call for further investigation, in which the unique experimental setup conceived in cooperation by the Montpellier and Saclay groups has a crucial role to play. In this setup, a crack can be generated in a vitreous sample using DCDC (Double Cleavage Drilled Compression) geometry and followed in situ by AFM into a clean nitrogen atmosphere with controlled variation of relative humidity (between ~0% and 80%). The evolution of the crack speed as a function of the stress intensity factor and of the environmental conditions can be tracked in an extremely wide range of velocities (between 10-3 and 10-13 m/s). When the crack velocity is slower than 10-9 m/s, the morphology of the crack tip neighborhood can be studied in real time (in regions down to 50 nm size for lowest velocities). Moreover, AFM phase imaging techniques are used to measure heterogeneities in local mechanical and chemical properties accompanying the advancing crack tip. Instrumental changes based on the purchase of a new generation AFM and the development of a fast digital acquisition system will allow for crucial improvements of the observations. The first goal of CORCOSIL is to understand what damage takes place in the bulk of the glass sample: stress and environmental conditions here are different from those at the free surface, where cavities have been observed up to now. Major assets of the new AFM such as nanometric lateral probe stability or 10 fold increase in the image resolution will be used to get relevant experimental post-mortem topographical observations of the crack surfaces. FRASTA techniques will be used to analyse the mismatch between corresponding opposite surfaces. Statistical analysis will then be performed with the expertise of CEA-Saclay, paying special attention to the development of clear statistical estimators so that cavity remnants can be unambiguously detected and separated from the inherent noise of the experimental setup. In parallel, roughness measurements will be performed to determine the size of the fracture process zone from fracture surface scaling properties. A second part of CORCOSIL project will be devoted to study the critical role of water in the mechanisms of stress corrosion. Combining topographic and phase AFM images, the Montpellier group has recently shown that a region 100 nm long inside the crack cavity is indeed filled with a hydrous condensate in typical ambient conditions. The extent of the observed condensation is significantly larger than what can be predicted by classical capillarity models and has tremendous implications on the physics of crack propagation. CORCOSIL will include extensive studies of the kinetics of growth and evaporation of the water condensate in relation to changes in the stress intensity factor, crack velocity and relative humidity. Modelling of this behaviour requires a deeper investigation of the wetting properties of the new fracture surfaces that will benefit of the expertise of the Laboratoire de Physique de la Matière Condensée et Nanostructures (LPMCN) of Lyon. This group has recently established the nature of the capillary forces acting through nanometric bridges linking wettable surfaces by coupling accurate adhesion measurements with a custom developed Surface Force Apparatus (SFA) and fine theoretical modelling. CORCOSIL intends to develop this approach to study phenomena in the neighbourhood of the crack tip by combining SFA and colloidal probe measurements, and considering the effects of the local stress field, surface roughness and charge distribution.