example of such molecules are given on figure 1. Alternatively random copolymers, also shown on figure 1, can reinforce the interface by effectively reducing the immiscibility and increasing the interfacial width. Figure 1: Schematic of connecting molecules at interfaces between A and B immiscible polymers. (Dark beads represent monomers of A and light beads represent B monomers). From left, diblock copolymers, triblock copolymers and random copolymers. The latter type of molecule acts by broadening the interface. EXPERIMENTAL METHODS In order to obtain the necessary information to bridge the gap between the molecular scale and the continuum scale, several techniques must be used. At the molecular level, one must be able to work with well-defined molecules, in particular for the block copolymers which are used as connecting chains. The molecular structure of the interface must be characterized before and after fracture if possible. For this purpose partial deuteration of the molecules present at the interface is invaluable and allows the use of ion beam techniques (to measure the amount of deuterium present on each surface after fracture) or neutron reflectivity (to measure interfacial width before fracture). If nitrogen is present, X-ray photoelectron spectroscopy (XPS) can also be quantitative to measure the amount of block copolymer at the interface. At the macroscopic level, suitable fracture mechanics tests are needed to characterize the fracture toughness of the interface Gc. The double cantilever beam test combines ease of sample preparation and measurement with a great flexibility in controlling the degree of mode mixity at the interface during crack propagation[2]. This last point is important because the Gc of interfaces generally depends on the phase angle of loading in a non-trivial way. We found this to be particularly true of interfaces between glassy polymers where very small amounts of shear stresses could significantly modify the measured Gc by causing small crazes to grow in the bulk polymers[2,3]. Finally at the intermediate microscopic level, it is important to use observation tools such a electron or optical microscopy to investigate how is local plasticity near the interface is affected by a modification of the molecular structure at the interface or by a change in the plastic deformation properties of the bulk polymers. CONNECTING CHAINS BETWEEN GLASSY POLYMERS The simplest and most informative case is that of an interface between immiscible A/B glassy polymers reinforced with a variable amount of A-B diblock copolymer. In that case one can safely assume that all the stress transfer capability is due to the presence of the block copolymers. For long copolymer chains, which do not disentangle, the stress transfer capability of the interface σint is given by: σint = fb Σ (1) where fb is the force to break a covalent bond (approximately 2nN) and Σ is the areal density of connector chains present at the interface. As long as σint remains below the crazing stress of both A and B polymers, Gc is low and the interface fails by simple chain fracture. If σint > σcraze, i.e. if Σ > Σ ∗ = σint/fb, a craze precedes the propagating crack and Gc becomes much higher. If shorter copolymer chains are used, and the interface
RkJQdWJsaXNoZXIy MjM0NDE=