Void and bubble lattice formation in molybdenum: A mechanism based on two-dimensional self-interstitial diffusion

The recent experimental conclusion by Jacques and Robrock that self-interstitial atoms (SIA) in molybdenum migrate at low temperatures by two-dimensional diffusion on [011] planes has led to an exploration of the spatial effects that might arise if this property was extended to temperatures where voids are found under irradiation. Qualitative analysis supported by rate theory shows that the introduction of two-dimensional interstitial migration allows concentration variations of SIA from plane to plane which, in turn, can lead to the small natural spatial variations in void density and radius parameters being amplified with positive feedback. The practical result of such an effect would be the planar ordering of voids on [011] planes. Although the analysis examines effects on only one set of the six [011] planes, extension of planar ordering onto all six sets will, it is argued, lead to a bcc void lattice. The circumstantial evidence to support the model seems fairly strong. For example the tendency for planar ordering is most efficient for high void densities and low dislocation densities while the long range ordering effects on the close packed planes, prior to perfect lattice formation, is a natural consequence of the model. Furthermore a ratio of void lattice parameter to void radius of the correct experimental order can be obtained. The paper discusses these points in detail together with an extension of the model to cover bubble lattice formation in molybdenum. Other metals are also considered; while there is no difficulty for bcc metals, application of the model to fcc and hcp lattices is more speculative. (orig.)