I read a few things on spintronics during my university studies, and I found the field promising and interesting. This spintronics paper works with antiferromagnets and attempts to streamline control of antiferrmagnetic domains. Electric manipulation of these materials tend to create large amounts of heat waste. Thermoplasmonics uses the collective electron oscillations in metallic structures to funnel nanoscale heating mechanisms into useful applications.
An AFM thin film of NiO has the Hamiltonian
(Eq. 1) with n the Neel vector characterizing the magnetization difference of the two sublattices. The first term is the magnetoelastic energy and the second the intrinsic magnetic anisotropy. uij is the strain vector component and ui the displacement vector along i-th direction. The out-of-plane anisotropy field H|| imposes a strong easy-plane anisotropy in NiO, and n is confined to the film plane, while the in-plane anisotropy H⊥ determines the 4-fold anisotropy in the plane. The total energy variance under θ has four local minima at either Neel vector component in the xy-plane equals ±1. Assuming a nonzero ume, one pair of states remains stable. Given ume > H⊥/Hme only the ny = ±1 are stable, conversely, if ume < -H⊥/Hme only the nx = ±1 are stable. The critical value for ume is dependent on material, for NiO, it's about 10-5. The underlying mechanics are due to the hot regions expanding and creating an additional stress field enhancing the local deformation in neighboring areas, and the lattice displacement reaching a maximum at intermediate temperature regions. For majority oriented temperature gradient, the strain component defined as the derivative of the displacement dominates and determines the sign of ume.
When a laser illuminates metallic nanowires, its free electrons are collectively excited, which forms surface plasmons, which dissipate the electromagnetic energy into heat and raising the material temperature. The heat diffuses from the hot nanowire into the environment, creating a spatial temperature gradient and thermal strain in the magnetic substrate. This is simulated using the COMSOL Multiphysics package on a gold nanowire patterned on an NiO insulator. In this simulation, the anisotropic strain distribution is highly tunable through the nanowire's dimensions, the polarization of the laser selectively exciting distinct plasmon modes. In general, a metallic nanowire supports two characterisitc plasmon modes:
- longitudinal mode associated with electron oscillations along the long wire axis
- transverse mode corresponding to oscillations along its short axis
The excitation efficiency depends strongly on the polarization of the incident wave. They have very different spatial distributions of temperature. The induced strain field can be harnessed to manipulate the magnetic domain orientation. An absorption peak at around λ = 1.1 μm can be measured with distinct local resonance peak in absorption and temperature rise appears corresponding to the excitation of the longitudinal plasmon mode. The thermal-induced strain effect is maximized at this critical wavelength. The analytical calculation of the resonance position is challenging because the wavelength of the incident wave is comparable with the length of the nanowire. For s-polarized illumination, at λ = 0.55 μm is consistent with the expected transverse plasmon response of metallic nanowires with comparable dimensions reported in literature. The responses for the polarization provides a filter.
When an s-wave with frequency near the longitudinal plasmon resonance illuminates the structure, the electrons in the vertical arms of the square are primarily excited. The resulting strain field in the frame is dominated by heating in the two vertical arms is typically negative inside the square. When a p-wave near the longitudinal resonance frequency is incident, excitation in the vertical arms is suppressed, and instead the horizontal arms will be excited and the thermal induced strain field reverses its sign. The Neel vector inside the square switches stable configurations as well. Reversibility is achieved by switching back to s-polarized illumination after p-polarized at small wavelengths. A polarization-selective plasmon excitation provides a robust and fully reversible means for controlling the Neel-vector orientation in the nanoframe (p.4). Immediately after optical excitation, hot electrons are generated on an ultrafast timescale, followed by e--e- scattering, which redistributes their energy and elevates the temperature of the material. Eventually, the thermal energy is transferred to the underlying AFM layer, giving rise to inhomogeneous thermal strains. The Neel vector rapidly reorients to follow the direction of the strain field. The switching energy via optical plasmons is on the otder of 0.001μJ, which is six orders of magnitude smaller than current-driven techniques (p. 5).