Does a surface spin-flop occur in antiferromagnetically coupled multilayers? Magnetic states and reorientation transitions in antiferromagnetic superlattices

TL;DR

This study models magnetic states in antiferromagnetically coupled multilayers, showing that finite systems do not exhibit surface spin-flop transitions but instead undergo various volume transitions with metastable states, explaining experimental observations.

Contribution

It demonstrates that finite multilayer systems lack surface spin-flop transitions and instead exhibit volume transitions, clarifying previous experimental interpretations.

Findings

Finite multilayers do not have surface spin-flop transitions.

Observed jumps are first-order volume transitions, not surface effects.

Multiple metastable states cause hysteresis in magnetization.

Abstract

Equilibrium spin configurations and their stability limits have been calculated for models of magnetic superlattices with a finite number of thin ferromagnetic layers coupled antiferromagnetically through (non-magnetic) spacers as Fe/Cr and Co/Ru multilayers. Depending on values of applied magnetic field and unaxial anisotropy, the system assumes collinear(antiferromagnetic, ferromagnetic, various "ferrimagnetic") phases, or spatially inhomogeneous (symmetric spin-flop phase and asymmetric, canted and twisted, phases)via series of field induced continuous and discontinuous transitions. Contrary to semi-infinite systems a surface phase transition, so-called "surface spin-flop", does not occur in the models with a finite number of layers. It is shown that "discrete jumps" observed in some Fe/Cr superlattices and interpreted as "surface spin-flop" transition are first-order "volume" transitions between different canted phases. Depending on the system several of these collinear and canted phases can exist as metastable states in broad ranges of the magnetic fields, which may cause severe hysteresis effects. The results explain magnetization processes in recent experiments on antiferromagnetic Fe/Cr superlattices.