Tuning field amplitude to minimise heat-loss variability in magnetic hyperthermia

In this work, we theoretically investigate how shape-induced anisotropy dispersion and magnetic field amplitude jointly control both the magnitude and heterogeneity of heating in magnetite nanoparticle assemblies under AC magnetic fields. Using real time Landau-Lifshitz-Gilbert simulations with thermal fluctuations, and a macrospin model that includes both the intrinsic cubic magnetocrystalline anisotropy and a shape-induced uniaxial contribution, we analyze shape-polydisperse systems under clinically and technologically relevant field conditions. We show that for relatively large particles, around 25 to 30 nm, the relative dispersion of local (single-particle) losses exhibits a well-defined minimum at moderate field amplitudes (between 4 to 12 mT), hence identifying an optimal operating regime that minimizes heating heterogeneity while maintaining substantial power dissipation. The position of this critical field depends mainly on particle size and excitation frequency, and only weakly on shape dispersion, offering practical guidelines for improving heating uniformity in realistic MFH systems.

💡 Research Summary

Magnetic fluid hyperthermia (MFH) relies on the heat generated by magnetic nanoparticles (MNPs) when subjected to an alternating magnetic field. While maximizing the average specific loss power (SLP) is a common goal, the variability of heat production among individual particles can lead to local over‑ or under‑heating, compromising therapeutic efficacy and safety. This paper investigates how dispersion in particle shape‑induced anisotropy, together with the amplitude of the driving field, controls both the magnitude and heterogeneity of heating in magnetite (Fe₃O₄) nanoparticle ensembles.

The authors adopt a macro‑spin description in which each particle is represented by a single magnetic moment. The total magnetic energy includes (i) the intrinsic cubic magnetocrystalline anisotropy of magnetite (K_c = ‑1.1 × 10⁴ J m⁻³), (ii) a uniaxial anisotropy K_u arising from small deviations from sphericity, modeled as prolate ellipsoids with aspect ratio r = c/a, and (iii) the Zeeman interaction with a sinusoidal field H(t) = H_max sin(2πft). The uniaxial constant is calculated from demagnetizing factors (K_u = μ₀²(N_a‑N_c)M_s²/2). Shape dispersion is introduced by assuming a Gaussian distribution of r with mean ⟨r⟩ = 1.1 and standard deviations σ_r = 0, 0.1, 0.2.

Dynamic magnetization trajectories are obtained by integrating the stochastic Landau‑Lifshitz‑Gilbert (LLG) equation using the OOMMF micromagnetic package, including a thermal field at T = 300 K and a Gilbert damping α = 0.1. Simulations are performed for N = 1000 non‑interacting particles, with particle diameters D = 15, 20, 25, 30 nm (volume kept constant for each D), frequencies f = 100 kHz and 1 MHz, and field amplitudes H_max ranging from 1 mT to 40 mT. For each set of parameters, 30 independent realizations are generated to ensure statistical robustness. Hysteresis loops are recorded, and the SLP is calculated from the loop area (SLP = f ρ ∫ M(H) dH). The dispersion of local heating is quantified by the standard deviation σ_SLP of the particle‑wise SLP values.

Key findings:

1. Shape‑monodisperse systems – Increasing the aspect ratio from r = 1.0 (perfect sphere) to r = 1.2 progressively widens the hysteresis loops and raises SLP/f. At r ≈ 1.1 the uniaxial shape anisotropy becomes comparable to the cubic term, lowering the effective energy barrier and allowing measurable heating already at low fields (~4 mT at 100 kHz). At r ≈ 1.2 the barrier is much larger, so appreciable losses appear only above ~10 mT. Frequency elevation to 1 MHz amplifies dynamic losses, especially for the most elongated particles.

2. Shape‑polydisperse ensembles – When a Gaussian distribution of r is applied (σ_r > 0), the ensemble SLP follows the behavior of the more elongated sub‑populations at higher fields, while at low fields the contribution is dominated by particles near r = 1.1. The standard deviation σ_SLP grows sharply as H_max approaches the saturation region, reflecting the increasing dominance of high‑anisotropy particles and thus larger heating heterogeneity.

3. Optimal field amplitude – Plotting σ_SLP normalized by the mean SLP versus H_max reveals a pronounced minimum in the intermediate field range of 4–12 mT for particle diameters 25–30 nm. This “optimal” field (H_crit) minimizes heating variability while still delivering substantial power. The position of H_crit shifts modestly with particle size and frequency (higher f moves the minimum to slightly larger H_max) but is largely insensitive to the degree of shape dispersion σ_r.

The practical implication is that, within clinically acceptable H·f products (≤5 × 10⁸ A m⁻¹ s⁻¹), one can select a moderate field amplitude (4–12 mT) to achieve a uniform heating profile across a realistic, shape‑polydisperse nanoparticle batch. This strategy does not require eliminating shape dispersion; controlling the mean aspect ratio (⟨r⟩≈1.1) suffices to place the ensemble near the optimal regime.

The study thus provides a quantitative framework for balancing average heating efficiency against spatial uniformity in MFH, guiding both nanoparticle synthesis (targeting modest elongations) and device operation (selecting appropriate field amplitudes and frequencies). Future work should incorporate inter‑particle dipolar interactions, Brownian rotation, and the influence of the surrounding medium to extend the model toward in‑vivo conditions.