MODELS OF MATTER TRANSPORT - THEIR VERIFICATION AND USE FOR DILUTE SYSTEMS WITH THE AID OF STATISTICAL-THEORIES

Statistical expressions for the Onsager phenomenological coefficients are now available for several models of matter transport in systems wh ich are dilute in solutes, vacancies and interstitials. They describe transport by unpaired and paired defects, as in the familiar five-freq uency model. The role of atomistic calculations of the parameters of s uch models in validating, simplifying or improving them is illustrated . In general it may be difficult to convincingly determine from experi ment all the parameters of a given model or to distinguish between clo sely related models, such as between those retaining first or first an d second neighbour binding between a solute and a vacancy. Nevertheles s, there is particular utility in knowledge of all the Onsager coeffic ients for models suggested by atomistic calculations and other informa tion, when one interprets phenomena involving coupled fluxes. This is illustrated by the example of coupled fluxes of a solute, vacancies an d interstitials in a dilute alloy under steady irradiation by high ene rgy particles. Here, essential contributions to modelling solute segre gation to grain boundaries can be made; striking qualitative differenc es between closely related models can occur in this case. Finally the limitations of the current statistical results are examined for system s where the solutes and defects carry effective charges. Limitations a re suggested at low temperatures by the existence of a phase transitio n in the restricted primitive model of dilute aqueous electrolytes. Th e range of concentration and temperature within which the current proc edures, stemming from the early work of Teltow and Lidiard, have been properly validated is limited. In the absence of recent theoretical ad vances Monte Carlo simulation is becoming the method of choice in inve stigating such limitations and in studying the transition to higher co ncentrations where there are larger clusters of defects.