Growth and Kerr magnetometry of Mn2Au on a gold-capped Nb(001) substrate

Abstract

We report on the epitaxial growth of antiferromagnetic Mn2Au on a Nb(001) substrate capped with a pseudomorphic layer of gold. We observe a layer-by-layer growth by means of medium-energy electron diffraction and confirm stoichiometry and surface structure by Auger electron spectroscopy and low-energy electron diffraction. Evaporation of 15 ML of ferromagnetic Fe on 12--17 ML of Mn2Au results in an exchange-coupled bilayer system with an exchange-bias shift that can be set by field-cooling from 400 K. Areas with and without exchange bias, with domain sizes in the range of tens of μm, are identified by Kerr microscopy. Postannealing the sample at or above 450 K after Mn2Au layer growth decreases the amount of areas where Fe magnetically couples to Mn2Au. We conclude that exchange coupling to an interfacial Fe layer depends on the interface termination of Mn2Au. Our findings provide insight into the growth process of Mn2Au and the coupling to an Fe layer. Our results point out the importance of growth, interface quality and termination on the magnetic properties of a Mn2Au/Fe bilayer which may help to improve material properties for spintronic applications.

Figures (7)

• Figure 1: LEED images of surface-oxidized Nb (a), after flash annealing of Au (b) and after evaporation of 15 ML of $\mathrm{Mn_2Au}$ (c) at electron energies of 140--148 eV. Additional LEED spots in (a) indicate an oxygen superstructure while (b) and (c) show clear (001)-spots corresponding to a flat surface with long-range crystallographic order.
• Figure 2: (a): MEED oscillations measured in-situ during deposition. Black dotted lines indicate start and end of deposition. Blue dashes indicate a complete ML, with the second and third ML marked at likely positions interpolated from the other peaks. (b): Auger electron spectrum of the clean flash-annealed Nb(001)/NbAu substrate and after deposition of 15 ML of $\mathrm{Mn_2Au}$.
• Figure 3: Temperature dependence of hysteresis loops of Nb(001)/NbAu/15 ML $\mathrm{Mn_2Au}$/15 ML Fe before (a) and after (b) FC, together with the original hysteresis loop measured at RT for reference. (c) Comparison of hysteresis loops between 15 ML and 20 ML of Fe.
• Figure 4: (a): Temperature-dependent magnetization loops of a 450 K postannealed sample after FC. (b): Hysteresis loops after an additional second FC process in reversed field, compared to the initial hysteresis loop at 300 K after the first FC process. (c-e): Hysteresis loops for different postannealing temperatures (c) 350 K, (d) 400 K (FC field applied in negative direction, 17 ML of $\mathrm{Mn_2Au}$) and (e) 475 K. (f): Hysteresis loops after FC for a sample with only 12 ML of $\mathrm{Mn_2Au}$, postannealed at 350 K.
• Figure 5: (a) Plot of the temperature dependence of the coercive field (top) and EB shift (bottom), taken after a FC process, for different postannealing temperatures ranging from 300 K to 475 K. The black dashed horizontal line at 12.5 mT indicates the coercivity of the non-shifted part. (b) Ratio between the Kerr signal amplitude of the exchange-bias-shifted part of the hysteresis loop and the total amplitude of the saturated hysteresis loops for samples with 15--17 ML of $\mathrm{Mn_2Au}$.