Reactive diffusion in a prototype system: nickel - aluminum II: The ordered Cu3Au rule and the sequence of phase formation, nucleation
Abstract
In most metallic compounds (and also in others) it is anticipated that the majority atoms will diffuse faster than the minority atoms. This effects has been named the ordered Cu3Au rule after the well known compound. Some extensions are possible. Because the majority atoms are mostly bound to like atoms, they should display bonding and diffusion properties similar to those displayed in the elemental state. Thus nickel atoms in AlNi3 should, and do, diffuse at about the same rate as in pure nickel. Furthermore, when comparing two compounds such as AlNi3 and Al3Ni, one would anticipate that aluminum atoms would diffuse much more readily in Al3Ni than nickel atoms in AlNi3. Without entering into the largely fruitless debate about the identity of the first phase formed, these considerations lead to the expectation that Al3Ni should form and grow before AlNi3. This is indeed what is generally observed. A reversal of the usual order is probably due to impurity effects. In the nickel-aluminum system the absence of some phases, here Al3Ni5, may safely be attributed to a high nucleation barrier. © 1992.