Ni tracer diffusion in the B2-compound NiAl: influence of temperature and composition Original Research Article

The effect of composition and temperature on Ni bulk self-diffusion is investigated for nine different single crystalline NiAl alloys with well-defined compositions between 46.8 and 56.6 at.% Ni in the temperature range from 1050 to 1630K. The diffusion penetration profiles of Ni in NiAl were determined by applying two different techniques of profile detection. Radiotracer experiments have been carried out using the 63Ni tracer, a serial sectioning technique, and sensitive liquid scintillation counting for the high temperature measurements, while at lower temperatures the diffusion profiles were analyzed by secondary ion mass spectrometry (SIMS) using the highly enriched stable isotope 64Ni. In contrast to the literature data on Ni self-diffusion in NiAl alloys by , the present measurements show an unexpected concentration dependence of the Ni diffusion coefficients D with nearly constant diffusivities for stoichiometric and Al-rich alloys and increasing D values with increasing Ni content on the Ni-rich side of the NiAl composition range. The effective diffusion activation enthalpy Q is equal to about 3.0±0.07 eV for the Al-rich, stoichiometric, and slightly Ni-rich NiAl alloys, while for the compositions with larger Ni content a decrease of Q was observed with increasing Ni content, for example, Q=2.39 eV for the Ni56.6Al43.4 alloy. The present experimental results imply that mainly the same diffusion mechanism operates on both sides of stoichiometry in NiAl. This mechanism is identified with the triple defect mechanism. Its contribution is compositionally independent. The activation energy of Q=3.18 eV was calculated for the triple defect mechanism using empirical EAM potentials in agreement with the experimental data. The decrease of Q at large Ni concentrations on the Ni-rich side is explained by an additional contribution of the anti-structure bridge mechanism with the activation energy of Q=1.73 eV.