Temperature dependence of coercivity for isolated Ni nanowires unraveled by high-sensitivity micromagnetometry

TL;DR

This work addresses how the magnetic coercivity of Ni nanowires differs when studied as isolated wires versus in dense arrays, highlighting the roles of dipolar and magnetoelastic interactions. Using a highly sensitive micromechanical torsional oscillator, the authors measure the hysteresis of a small group of released Ni NWs over 5–200 K, extracting magnetization from tiny resonance-frequency shifts. They find that isolated wires exhibit a nearly temperature-independent coercivity around 0.1 T and squared hysteresis with high remanence, contrasting with the more gradual coercivity behavior in arrays and indicating non-coherent, Bloch-domain-wall–mediated reversal in isolation. The results provide direct insight into intrinsic NW magnetism and underscore how interwire and template interactions distort coercivity in arrays, with important implications for designing NW-based magnetic devices.

Abstract

Magnetic nanowires are critical components in fields such as data storage and spintronics, where precise control of their magnetic properties is essential for device optimization. In particular, the behavior of isolated nanowires is often different from that of an ensemble, offering an opportunity to explore the role that dipolar and magnetoelastic interactions play in the latter system. Unfortunately, the comparison between a collection of nanowires and single ones is often poorly characterized, as measuring individual nanowires with weak magnetic signals is a challenging task. In this work, we employ a highly-sensitive micromechanical torsional oscillator to measure the magnetic response of few individual Ni nanowires with 72 +/- 5 nm average diameter, fabricated by electrodeposition in anodic aluminum oxide templates as an array and subsequently released from this membrane. When comparing the magnetic properties as a function of temperature between single nanowires and the array, we show that coercivity values of individual nanowires are at least twice as large as for the array in the range 5 K - 200 K. Also, we characterize the differences in the hysteresis loops, which are more squared for isolated nanowires, with a high magnetic remanence close to 80 % of the saturation value. Our results highlight the crucial role of dipolar and mechanical interactions in modifying the magnetic behavior of nanowires arrays, providing valuable insights for the design and application of nanowires-based magnetic devices.

Figures (9)

• Figure 1: (a) SEM image of the silicon micromechanical oscillator showing the NWs stuck to its plate. This structure is attached to two springs which are in turn anchored to the substrate, allowing torsional oscillations forced by an alternating voltage signal connected to the electrodes. When a magnetic field $\mathbf{H}$ is applied, the resonance frequency of the device changes, allowing us to infer the magnetic moment of the sample. (b) SEM image of a typical Ni NW.
• Figure 2: SEM image showing a bundle made up of NWs of approximately $l=9$$\mu$m in length.
• Figure 3: Resonance curves obtained when $\mathbf{H}$ or $\mathbf{M}$ (or both) are zero, and therefore, there is no additional magnetic torque (cases I and II), and when these fields are parallel (case III) or antiparallel (case IV), which results in the appearance of a restoring or antirestoring magnetic torque, respectively.
• Figure 4: Sketches corresponding to cases I, II, III, and IV. No additional magnetic torque arises when (a) $\mathbf{H}=0$ or (b) $\mathbf{M}=0$ . However, when both fields are nonzero, (c) a restoring torque appears if $\mathbf{H}$ and $\mathbf{M}$ are parallel, while (d) this torque is antirestoring if $\mathbf{H}$ and $\mathbf{M}$ are antiparallel.
• Figure 5: (a) Changes in the resonant frequency $\Delta \nu$ as a function of the external field $\mu_0 H$ for different temperatures, as indicated. (b) Corresponding magnetic hysteresis loops calculated using Eqs. (\ref{['rootMplus']}) and (\ref{['constZ']}). In both figures, the arrows indicate the direction in which $\Delta \nu$ and $\mu_0 M$ evolve as the external field changes.