The resulting high-field estimates were therefore more consistent with the intrinsic energy barrier of individual nanoparticles, although the fitted value remained an effective quantity that could include core, shell, shape and surface contributions. Effective magnetic anisotropy controls the barrier to thermally activated reversal of a nanoparticle’s magnetic moment. For an isolated particle with uniaxial anisotropy, the barrier is K_effV, where K_eff is the effective anisotropy constant and V is the magnetic-core volume.

The Néel relaxation time follows an Arrhenius-like relation, τ_N = τ_0 exp[K_effV/(k_BT)], with τ_0 typically between 10^-9 and 10^-13 seconds depending on material properties and damping. Moments fluctuate rapidly and produce a superparamagnetic response when τ_N is much shorter than the measurement time. They remain blocked and exhibit hysteresis when τ_N is much longer.

At the blocking temperature T_B, the two are comparable; a conventional analysis assumes τ_0 = 10^-10 seconds and a 100-second DC-magnetometry measurement time, yielding K_eff = 25k_B〈T_B〉/〈V〉. This conversion depends on assumptions about anisotropy symmetry, relaxation mechanism, attempt time and particle-volume distribution that real ensembles may not satisfy. The review emphasizes that nanoparticle ensembles combine several effects that can bias a single low-field analysis.

Size and shape distributions produce a range of barriers rather than one sharp value. Surface and interface contributions become increasingly prominent as particles become smaller, while dipolar interactions can either raise or lower the apparent barrier. Strongly correlated aggregates may also violate the equilibrium Boltzmann–Gibbs description used by conventional interaction corrections.

Magnetite adds another complication because its magnetocrystalline anisotropy is cubic, whereas many blocking-temperature formulas assume a uniaxial particle with only two equivalent easy directions. The Stoner–Wohlfarth model assigns a geometrical coercivity factor of about 0.479 to randomly oriented uniaxial particles, compared with about 0.32 for cubic systems with K_1 greater than zero and about 0.19 for cubic systems with K_1 less than zero. Treating magnetite as uniaxial can therefore underestimate its intrinsic anisotropy.

Morphology also affects conversion from size to volume: approximating particles as cubic rather than spherical increased the inferred volume by a factor of about 1.9 in the review’s comparison. The first approach extracted a blocking-temperature distribution from the temperature derivative of the zero-field-cooled and field-cooled magnetization difference. This derivative identifies the rate at which particles cross their blocking barriers, but the review notes that the measured distribution is proportional to T f(T_B), not simply f(T_B), because larger particles carry greater magnetic moments.

A correction accounting for this weighting reduced the influence of large particles and produced nearly identical characteristic blocking temperatures of about 30 K for bare and coated magnetite samples with similar magnetic cores. By contrast, the conventional distribution returned about 51 K for the coated sample and 68 K for the bare, more aggregated material. These results were not contradictory measurements of an individual barrier; they reflected different weighting of a polydisperse, interacting ensemble.

The review judged the corrected distribution more accurate for polydisperse or moderately and strongly interacting samples, while warning that even that result should be checked with complementary methods. More difficult cases exposed the limitations of the Néel analysis. A bimodal sample containing large cubic particles and elongated bars had no conventional zero-field-cooled maximum, an almost temperature-independent field-cooled curve and no clear intersection between the curves.

The larger bars remained blocked at room temperature because of their size and shape anisotropy. Cobalt-doped samples also strained the uniaxial formula because of their cubic symmetry. A more strongly doped and larger particle produced a lower apparent anisotropy than a smaller, less doped one because the increase in volume outweighed the increase in blocking temperature and reduced the relative contribution of surface anisotropy.

Deviations from nominal stoichiometry in co-precipitated material could also affect these results. Coercive-field thermal decay provided a second route to K_eff. Kneller’s law, H_C(T) = H_C(0)[1 − (T/T_B)^γ], is often fitted with γ = 1/2 for aligned uniaxial particles, 2/3 for randomly oriented uniaxial particles or approximately 3/4 for cubic systems.

Its monodisperse idealization compresses a continuous distribution of barriers into a single effective value. Applied to the nanoparticle samples, it reproduced only the broad trend of thermal decay and underestimated anisotropy in soft magnetites relative to the high-field estimates. In strongly anisotropic cobalt-doped material, it instead produced implausibly high values, including about 81.5 ± 0.5 × 10^4 J/m³ for CIONP9.1, because collective rigidity and frozen surface spins were interpreted as a larger single-particle barrier.

A generalized model incorporating a log-normal distribution of blocking temperatures fitted most samples better, but could not obtain physically meaningful parameters for the mixed nanorod-and-sphere ensemble. Such generalized fits still left the extracted anisotropy entangled with dipolar interactions. Remanent-magnetization decay supplied a third low-field check.

The remanence ratio M_r/M_s depends on anisotropy symmetry and particle orientation: the Stoner–Wohlfarth model predicts 0.5 for randomly oriented uniaxial particles, while cubic systems with multiple easy axes can reach approximately 0.831 for K_1 greater than zero and 0.866 for K_1 less than zero. Dipolar interactions commonly suppress remanence in uniaxial systems, while exchange interactions can enhance cooperative alignment. Fits of M_r(T), combined with a temperature-dependent saturation magnetization that includes Bloch-law and surface-spin terms, generally gave lower blocking temperatures that closely matched the corrected T f(T_B) analysis.

Bare and coated samples produced nearly indistinguishable fitted mean blocking temperatures despite differing interactions, suggesting that remanence decay primarily reflected the magnetic cores’ barriers. The imposed log-normal distribution may nevertheless underrepresent a minority of large or strongly interacting particles, while an analysis of the bimodal sample remained outside the model’s capabilities. The modified Actis–Zélis model instead analyzed how magnetization approached saturation.

Its statistical-mechanical treatment combines the Zeeman energy, which aligns magnetic moments with an applied field, with the anisotropy energy. When the applied field greatly exceeds the coercive field, the Zeeman energy largely overwhelms interparticle dipolar interactions, allowing K_eff to be estimated for each isotherm and its temperature dependence to be followed directly. For the 6.4-nm magnetite sample IONP6.4, K_eff decreased from 7.9 ± 0.4 × 10^4 J/m³ at 4 K to 4.85 ± 0.01 × 10^4 J/m³ at 300 K. More crystalline samples displayed larger low-temperature values, including 19.0 ± 1.0 × 10^4 J/m³ at 6 K for IONP8 and 14.5 ± 0.5 × 10^4 J/m³ at 2 K for IONP7.9.

Cobalt doping produced magnetic hardening because Co²⁺ ions preferentially occupy octahedral sites and contribute strong spin-orbit coupling. The most highly doped sample, CIONP9.1, reached 29.7 ± 0.7 × 10^4 J/m³ at 2 K and decreased to 8.9 ± 0.1 × 10^4 J/m³ at room temperature. Moderately doped samples produced values near 15.8–18.8 × 10^4 J/m³.

Low-field estimates were substantially higher in some of these materials: the generalized coercive-field model gave about 50 ± 10 × 10^4 J/m³ for CIONP8.1, compared with 15.8 ± 0.5 × 10^4 J/m³ from high-field fitting at 2 K. Some samples developed a maximum in K_eff below 150 K. The review associated this feature with surface-spin dynamics rather than incomplete magnetization because fits began at fields of at least 1.0 T, while the coercive fields were about 0.6–0.7 T. Fits to the Callen–Callen thermal scaling law often failed or produced non-integer exponents, consistent with surface canting, broken translational symmetry, shape anisotropy and interactions preventing nanoscale ferrites from behaving as ideal bulk ferromagnets. The comparison does not render low-field magnetometry useless. Zero-field-cooled and field-cooled measurements, coercive-field analysis and remanence decay remain useful for examining ensemble thermomagnetic stability, including behavior relevant to magnetic hyperthermia.

Their derived constants, however, cannot automatically be interpreted as the anisotropy of an isolated particle. The review’s most reliable option for extracting a more nearly intrinsic energy barrier was the modified high-field asymptotic analysis, especially for polydisperse, interacting or strongly surface-affected systems. Even that estimate is not a pure bulk constant: it can include magnetocrystalline, shape, core-shell and surface effects and remains dependent on temperature, field window, fitting assumptions and sample characterization.

The practical implication is not that one magnetic measurement is universally superior, but that researchers should match the model to the sample’s symmetry, interaction regime and morphology, and should use multiple methods when the intended quantity is individual-particle anisotropy rather than collective magnetic behavior.

Subject of Research: Comparison of DC-magnetometry methods for estimating effective magnetic anisotropy in nanoparticle ensembles

Article Title: Review: Determining effective magnetic anisotropy in magnetic nanoparticles systems using DC-magnetometry methods

Article References: de Menezes, J. R., Gomes, L. G., Fabris, F., de Almeida, A. A., Tancredi, P., Barbosa, M., da Silva, G. S., & Muraca, D. (2026). Review: Determining effective magnetic anisotropy in magnetic nanoparticles systems using DC-magnetometry methods. Journal of Materials Science, 61(41), 31675-31706. https://doi.org/10.1007/s10853-026-13671-2

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13671-2

Keywords: magnetic nanoparticles, magnetic anisotropy, DC magnetometry, Actis–Zélis model, magnetite, cobalt ferrite, dipolar interactions, superparamagnetism, coercive field, remanent magnetization, Review, Determining