Tailoring interface mixing and magnetic properties in (111) Permalloy/Pt multilayers

Movaffaq Kateb,Jon Tomas Gudmundsson,Snorri Ingvarsson
DOI: https://doi.org/10.48550/arXiv.2010.06104
2020-10-13
Materials Science
Abstract:We present deposition and characterization of multilayers consisting of 20 repetitions of 15 $\r{A}$ thick permalloy Ni$_{80}$Fe$_{20}$ at. \% (Py) and 5 $\r{A}$ Pt. The samples were prepared by two different sputter deposition methods, namely dc magnetron sputtering (dcMS) and high power impulse magnetron sputtering (HiPIMS), that represent low and moderate ionized flux fraction of the film forming material, respectively, for deposition of the Py layers. The effect of substrate roughness, working gas pressure and sputter power on the in-plane uniaxial magnetic anisotropy of the films are studied. The multilayers were characterized by X-ray reflectivity and diffraction, and by magneto-optical Kerr effect (MOKE). It is shown that HiPIMS deposition produces multilayers with unique surface roughness regardless of the substrate surface roughness. Multilayers prepared by both dcMS and HiPIMS deposition present a strong (111) texture normal to the film plane. The results show that utilizing HiPIMS for deposition of the Py layer leads to a minimum interface mixing between individual layers compared to dcMS deposition performed at sputter power. This is associated with the smooth surface of Py deposited by HiPIMS. However, this sharp interface results in higher coercivity and an opening in the hard axis hysteresis loops while multilayers with intermixing present well defined in-plane uniaxial anisotropy i.e. a linear hard axis. Comparison with Py/Cu and Py/CuPt multilayers, prepared under identical conditions using HiPIMS, suggests that poor in-plane uniaxial anisotropy is obtained in the Py/Pt case, caused by the inverse magnetostriction arising from the large lattice mismatch between Py and Pt. The Py/Pt multilayers that exhibit interface mixing have a more relaxed interface and thus presents negligible inverse magnetostriction and have better defined anisotropy.
What problem does this paper attempt to address?