Activation energy of hydrogen, oxygen, and nitrogen diffusion in metals

Abstract

Significant progress has recently been made in experimental studies of the diffusion of atoms of light elements in metals. In particular, for a number of metals, the diffusion coefficients of tracer atoms of hydrogen (tritium), oxygen, nitrogen, and boron were measured. Previously, such data were available only for carbon. New experimental data enabled us in this work to formulate the problem of developing a theoretical criterion for the reliability of experimental results on the activation energy Q for the interstitial diffusion mechanism. The necessity of formulating such a problem is primarily caused by the fact that the scatter of experimental data on the diffusion of interstitial impurities in metals is often very wide. For example, experimental values of the activation energy Q of hydrogen diffusion in α -Ti are within the range 24‐62 kJ/mol [1]. In this context, in this work, we analyzed experimental data on the diffusion of interstitial atoms in metals that were obtained by reliable methods. It was found that all such data on the activation energy of hydrogen, oxygen, and nitrogen diffusion are consistent with a calculation model [2, 3] based on the theory of elasticity. The calculation accuracy corresponds to the level of a modern precision experiment (on the order of 1%). The corresponding data for carbon and boron were excluded from the analysis. This is because the adequacy of the elastic model of the potential barrier for carbon was not confirmed by the results of measuring the anisotropy of the diffusion coefficients in α -Ti [4, 5]. Data on the diffusion of tracer boron atoms in metals are extremely meager and did not allow us to analyze relationships for boron. Elastic model of the potential barrier. The currently dominant concept of the nature of Q and procedures for calculating Q was formulated in the midtwentieth century [2, 3]. Within this concept, the activation energy Q for the interstitial diffusion mechanism is determined by the elastic energy of crystal strain when an impurity atom is placed at a saddle point; i.e., the larger the diffusant atom, the higher the activation energy Q . The expression for Q has the form [2]