Atom probe tomography field evaporation characteristics and compositional corrections of ZrB 2

• Published in: Materials characterization Vol. 156
• Main Authors: Engberg, David L. J., Tengdelius, Lina, Högberg, Hans, Thuvander, Mattias, Hultman, Lars
• Format: Journal Article
• Language: English
• Published: 2019
• Subjects: Atom probe tomography (APT); Compositional correction procedures; Elastic recoil detection analysis (ERDA); Field evaporation characteristics of borides; Transition metal diborides; Zirconium diboride (ZrB)
• ISSN: 1044-5803; 1873-4189
• DOI: 10.1016/j.matchar.2019.109871

The microstructure of stoichiometric ZrB2.0 and B over-stoichiometric ZrB2.5 thin films has been studied using atom probe tomography (APT), X-ray diffraction, and transmission electron microscopy. Both films consist of columnar ZrB2 grains with AlB2-type crystal structure. The narrow stoichiometry range of ZrB2 results in the presence of separate disordered B-rich boundaries even in ZrB2.0. At higher average B content, specifically ZrB2.5, the formation of a continuous network around the sides of the ZrB2 columns is promoted. In addition, the APT field evaporation characteristics of ZrB2 and its influence on the measured local composition has been studied and compared to the average composition from elastic recoil detection analysis (ERDA). Differences in the measured average compositions of the two techniques are explained by the APT detector dead-time/space. A new pile-up pairs correction procedure based on co-evaporation correlation data was thus employed here for the APT data and compared with the 10B-method (the B equivalence of the 13C-method), as well as the combination of both methods. In ZrB2.0, all of the applied compositional correction methods were found to reduce the compositional difference when appropriate isotopic abundances were used. In ZrB2.5, the inhomogeneity of the film likely increased the local APT composition to such an extent that even conservative correction procedures overestimated the B content compared to the ERDA reference. The strengths of the pile-up pairs correction compared the 10B and the combined methods are higher precision, due to it being less dependent on the accuracy of estimated isotopic abundances, and that the correction itself is not dependent on careful background correction of the mass spectrum.