Spin Canting in Exchange Coupled Bi-Magnetic Nanoparticles: Interfacial Effects and Hard / Soft Layer Ordering

Abstract

We investigate the spatial distribution of spin orientation in magnetic nanoparticles consisting of hard and soft magnetic layers. The nanoparticles are synthesized in a core / shell spherical morphology where the magnetically hard, high anisotropy layer is CoFe 2 O 4 (CFO) while the lower anisotropy material is Fe 3 O 4 (FO). The nanoparticles have a mean diameter of ∼9.2 -9.6 nm and are synthesized as two variants: a conventional hard / soft core / shell structure with a CFO core / FO shell (CFO@FO) and the inverted structure FO core / CFO shell (FO@CFO). High resolution electron microscopy confirms the coherent spinel structure across the core / shell boundary in both variants while magnetometry indicates the nanoparticles are superparamagnetic at 300 K and develop a considerable anisotropy at reduced temperatures. Low temperature M vs. H loops suggest a multi-step reversal process. Temperature dependent small angle neutron scattering (SANS) with full polarization analysis reveals a strong perpendicular plane alignment of the spins near zero field, indicative of spin canting, but the perpendicular alignment quickly disappears upon application of a weak field and little spin ordering parallel to the field until the coercive field is reached. Above the coercive field of the sample, spins orient predominantly along the field direction. At both zero field and near saturation, the parallel magnetic SANS peak coincides with the structural peak, indicating the magnetization is uniform throughout the nanoparticle volume, while near the coercive field the parallel scattering peak shifts to higher momentum transfer (Q), suggesting that the coherent scattering volume is smaller and likely originates in the softer Fe 3 O 4 portion of the nanoparticle.

Macrospin model of an assembly of magnetically coupled core-shell nanoparticles

“…The width d is chosen to be equal to the unit cell length of the associated material. For the core spins, the easy axis anisotropy is set along z direction whereas for the surface atoms the anisotropy axis direction is randomly selected for each spin site in agreement with experimental evidence 9 . Bulk J ij values are collected in Table I for CFO and FO nanoparticles 13 .…”

“…Experimental measurements have been performed on dense assemblies of CFO@FO and FO@CFO core-shell nanoparticles 9 . Under these experimental conditions, the dipolar interaction can affect the magnetic response, especially for NPs composed of CFO and FO due to their considerable magnetic moments.…”

“…In blue is shown the experimental hysteresis loop taken from Ref. [9] for comparison with the prediction of the theoretical model.…”

“…Moreover, as CFO is in a hard magnetic phase while FO is in a soft magnetic phase at normal conditions, the variants of these compounds can possess interesting exchanged coupled related properties. Thus, two variants of core/shell NP assemblies have already been synthesized and magnetically characterized 9 . More specifically, a conventional assembly of NPs is experimentally synthesised by adding CFO to the core of the NP and capping the core with a shell of FO (CFO@FO).…”

“…Zero-field cooling (ZFC) and field cooling (FC) curves for the assembly nanoparticles CFO@FO (left panel) and FO@CFO (right panel). The predictions of the theoretical model (circle symbol) are compared with the experimental measurements taken from Ref [9]. (solid line).…”