Morphology and magnetism of multifunctional nanostructured $\gamma$-Fe$_2$O$_3$ films: Simulation and experiments

TL;DR

This paper presents a multistage simulation approach combining Cell Dynamic System and Monte Carlo methods to accurately model the magnetic properties of nanostructured gamma-Fe2O3 films, validated by experiments.

Contribution

The novel multistage simulation method accurately predicts magnetic behavior of nanocomposites by incorporating realistic 3D structures and particle distributions.

Findings

Simulation closely matches experimental T_MAX (54 K vs. 53 K)

Predicted remanence and coercivity are higher than experimental values

Approach effectively models morphology and magnetic properties of nanostructured films

Abstract

This paper introduces a new approach for simulating magnetic properties of nanocomposites comprising magnetic particles embedded in a non-magnetic matrix, taking into account the 3D structure of the system in which particles' positions correctly mimic real samples. The proposed approach develops a multistage simulation procedure in which the size and distribution of particles within the hosting matrix is firstly attained by means of the Cell Dynamic System (CDS) model. The 3D structure provided by the CDS step is further employed in a Monte Carlo (MC) simulation of zero-field-cooled/field-cooled (ZFC/FC) and magnetic hysteresis loops ($M \times H$ curves) for the system. Simulations are aimed to draw a realistic picture of the as-produced ultra-thin films comprising maghemite nanoparticles dispersed in polyaniline. Comparison (ZFC/FC and $M \times H$ curves) between experiments and simulations regarding the maximum of the ZFC curve ($T_{\scriptsize MAX}$), remanence ($M_R/M_s$) and coercivity ($H_C$) revealed the great accuracy of the multistage approach proposed here while providing information about the system's morphology and magnetic properties. For a typical sample the value we found experimentally for $T_{\scriptsize MAX}$ (54 K) was very close to the value provided by the simulation (53 K). For the parameters depending on the nanoparticle clustering the experimental values were consistently lower ($M_R/M_s$ = 0.32 and $H_C$ = 210 Oe) than the values we found in the simulation ($M_R/M_s$ = 0.53 and $H_C$ = 274 Oe). Indeed, the approach introduced here is very promising for the design of real magnetic nanocomposite samples with optimized features.