An international collaboration led by the NIST Center for Nanoscale Science and Technology has used scanning electron microscopy with polarization analysis (SEMPA) to acquire images of the magnetic structure inside patterned nickel-palladium (NiPd) thin film nanostructures, revealing peculiar magnetization textures that can affect the behavior of these ferromagnetic alloys in experimental applications.
The NIST work further confirms the Brian Ahern theory that nanomagnetism plays an important part in low energy assisted nuclear reactions. The nickel and palladium electrodes used in many cold fusion experiments would have this nanomagnetic behavior.
NiPd alloys are used for studying how ferromagnets affect nearby superconductors. They are also good electron spin injectors and spin analyzers being applied to the development of carbon nanotube-based electronics using electron spin (“spintronics”). The magnetic orientation of nano-patterned NiPd thin film contacts was expected to be simple, controlled primarily by the shape of the patterned film and the applied magnetic field. The SEMPA measurements, along with magnetic force microscopy and spin-polarized photoemission electron microscopy, revealed a surprisingly complex spatial structure of the magnetization. In some devices, the magnetization was even perpendicular to the expected direction. The researchers found that complexity arises from stress-induced anisotropies caused by mismatches in both the lattice structures and the thermal expansion coefficients between the NiPd films and the underlying substrates. Although this stress-induced magnetic structure may be a problem for some applications, the researchers believe it can be used as a new route to control the orientation of the magnetization in nano-patterned electrodes.
We have observed peculiar magnetization textures in Ni80Pd20 nanostructures using three different imaging techniques: magnetic force microscopy, photoemission electron microscopy under polarized X-ray absorption, and scanning electron microscopy with polarization analysis. The appearances of diamond-like domains with strong lateral charges and of weak stripe structures bring into evidence the presence of both a transverse and a perpendicular anisotropy in these nanostrips. This anisotropy is seen to reinforce as temperature decreases, as testified by a simplified domain structure at 150 K. A thermal stress relaxation model is proposed to account for these observations. Elastic calculations coupled to micromagnetic simulations support qualitatively this model.
The magnetization distribution of NiPd nanostrips has been imaged using complementary techniques (MFM, XMCD-PEEM and SEMPA), revealing a transverse orientation of the magnetization and the appearance of weak stripes at low temperature or large thickness. The direct observation of a largely transverse magnetization differs from the conclusions previously drawn from AMR measurements only. It however corresponds well with the effect of field orientation on the switching of NiPd electrodes observed in magneto-transport measurements on carbon nanotubes, as reported by Refs. 11 and 14. From these observations, it appears that in order to account for the observed textures, non-negligible out-of-plane and transverse anisotropies have to be present.
Considering the evolution with temperature (increase of both anisotropy constants as temperature decreases), a thermal stress mechanism has been considered as the origin of this surprising magnetization texture, via magnetostriction. Note that the same mechanism was invoked for explaining the spin reorientation transition observed in Ni1−xPdx alloys grown on Cu3Au(100). Performing elastic, magneto-elastic and micromagnetic simulations, all qualitative features of the experiments could be reproduced, however with a disagreement regarding the relative magnitudes of the transverse and out-of-plane anisotropies.
We conclude that in addition, another effect may be present, such as an interfacial strain due to metalsubstrate mismatch, a structural ordering of the alloy in the growth direction (i.e., the film normal), or a plastic strain relaxation. The latter effect may also explain the observed difference in magnetic properties between the infinite film and the nanostructures, at the same thickness. Indeed, weak stripes were seen to appear at lower thickness in the nanostructures, and the value of the perpendicular anisotropy measured by ferromagnetic resonance on infinite films was smaller than what was deduced for nanostructures of the same thickness. This shows also that the evaluation of strain in nanostructures is difficult, and that magnetic patterns in nanostructures made out of magnetostrictive materials should be imaged.
Even though only one composition has been considered in this study, the discussion is general and should apply to other Pd concentrations. For lower nickel concentrations, the thermal strain is anticipated to increase, as well as the Young’s modulus and magnetostriction constant (initially at least), resulting in a fairly constant induced anisotropy. On the other hand, the alloy magnetization will decrease, down to zero, so that the role of the anisotropy induced by the thermal strain will be more and more important as the nickel content decreases. As a result, the easy axis will switch to the direction perpendicular to the plane, at a temperature that depends on composition. This corresponds well to the observations at a Ni atomic concentration of 10 %.
Finally, similar phenomena should occur for nanostructures made of other materials with a large magnetostriction and a small saturation magnetization.