GCNBs were prepared by chemical vapor deposition at Tokai Carbon Co. Ltd. The detailed preparation procedure has been reported previously [18]. The structure of GCNBs was studied by X-ray diffraction (XRD) (Rigaku, Rint2500), Raman spectroscopy (JovinYvon, T-64 000), and TEM (Hitachi-9000).For the fabrication of GCNB electrodes, each GCNB sample was mixed with a solution of poly(vinylidene difluoride)/N-methylpyrrolidinone (PVdF/NMP) (KF # 1120, Kureha) to make a slurry of a suitable viscosity. The weight ratio of GCNBs to PVdF was adjusted to 9:1. Then, the slurry was spread onto a copper foil thinly and evenly to fabricate the electrodes. The electrode was allowed to stand in a draft overnight to evaporate most of the NMP solvent, and was then vacuum dried at 80°C for 1 day. The electrode thickness was ca. 100 lm. For electrochemical measurements, 1 mol dm -3 LiClO 4 dissolved in PC and 1 mol dm -3 LiClO 4 dissolved in EC:DEC (1:1 by volume) were used as electrolytes. The former and the latter electrolytes are referred to as PC-and EC-based electrolytes, respectively. CV measurements were performed in a three-electrode cell using a HSV-100 (Hokuto Denko) instrument. Alternating current (AC) impedance measurements were also conducted with a three-electrode cell using a Solartron SI 1255 impedance analyzer coupled with a SI 1480 multi-channel electrochemical interface over a frequency range from 100 kHz to 10 mHz with an AC oscillation of 10 mV. Lithium metal was used as the counter and reference electrodes, and the GCNB electrode served as the working electrode. Unless otherwise stated, potentials were referenced to lithium metal.

We experimentally investigate spin-orbit torques and spin pumping in NiFe/Pt bilayers with direct and interrupted interfaces. The damping-like and field-like torques are simultaneously measured with spin-torque ferromagnetic resonance tuned by a dc bias current, whereas spin pumping is measured electrically through the inverse spin Hall effect using a microwave cavity. Insertion of an atomically thin Cu dusting layer at the interface reduces the damping-like torque, field-like torque, and spin pumping by nearly the same factor of ≈1.4. This finding confirms that the observed spin-orbit torques predominantly arise from diffusive transport of spin current generated by the spin Hall effect. We also find that spin-current scattering at the NiFe/Pt interface contributes to additional enhancement in magnetization damping that is distinct from spin pumping.

We measured spin--transport in nonferromagnetic (NM) metallic multilayers from the contribution to damping due to spin pumping from a ferromagnetic Co90Fe10 thin film. The multilayer stack consisted of NM1/NM2/Co90Fe10(2 nm)/NM2/NM3 with varying NM materials and thicknesses. Using conventional theory for one--dimensional diffusive spin transport in metals, we show that the effective damping due to spin pumping can be strongly affected by the spin transport properties of each NM in the multilayer, which permits the use of damping measurements to accurately determine the spin transport properties of the various NM layers in the full five--layer stack. We find that due to its high electrical resistivity, amorphous Ta is a poor spin conductor, in spite of a short spin-diffusion length of 1.0 nm, and that Pt is an excellent spin conductor by virtue of its low electrical resistivity and a spin diffusion length of only 0.5 nm. Spin Hall effect measurements may have underestimated the spin Hall angle in Pt by assuming a much longer spin diffusion length.The accurate measurement and control of damping in nanomagnets, are vital for future device applications, such as magnetic random access memory (MRAM)[1], spin--torque MRAM [2], spin--torque nano--oscillators [3], racetrack memory [4], and energy--assisted magnetic recording in hard--disk drives [5,6]. Most of these applications require structures with nanoscale ferromagnets in ohmic contact with nonferromagnetic (NM) metals. A significant source of damping in such metallic multilayer (ML) structures is spin--pumping [7,8,9,10,11], whereby time--varying magnetization generates a pure spin--current into the contacting NM. Indeed,

We demonstrate a strong dependence of the effective damping on the nanomagnet size and the particular spin-wave mode that can be explained by the theory of intralayer transverse-spin pumping. The effective Landau-Lifshitz damping is measured optically in individual, isolated nanomagnets as small as 100 nm. The measurements are accomplished by use of a novel heterodyne magneto-optical microwave microscope with unprecedented sensitivity. Experimental data reveal multiple standing spin-wave modes that we identify by use of micromagnetic modeling as having either localized or delocalized character, described generically as end and center modes. The damping parameter of the two modes depends on both the size of the nanomagnet as well as the particular spin-wave mode that is excited, with values that are enhanced by as much as 40% relative to that measured for an extended film. Contrary to expectations based on the ad hoc consideration of lithography-induced edge damage, the damping for the end mode decreases as the size of the nanomagnet decreases. The data agree with the theory for damping caused by the flow of intralayer transverse spin currents driven by the magnetization curvature. These results have serious implications for the performance of nanoscale spintronic devices such as spin-torque-transfer magnetic random access memory.

We demonstrate that the spectroscopic g−factor can be determined with high precision and accuracy by broadband ferromagnetic resonance measurements and applying an asymptotic analysis to the data. Spectroscopic data used to determine the g−factor is always obtained over a finite range of frequencies, which can result in significant errors in the fitted values of the spectroscopic g−factor. We show that by applying an asymptotic analysis to broadband datasets, precise values of the intrinsic g−factor can be determined with errors well below 1 %, even when the exact form of the Kittel equation (which describes the relationship between the frequency and resonance field) is unknown. We demonstrate this methodology with measured data obtained for sputtered Ni 80 Fe 20 ("Permalloy") thin films of varied thicknesses, where we determine the bulk g−factor value to be 2.109 ± 0.003. Such an approach is further validated by application to simulated data that includes both noise and an anisotropy that is not included in the Kittel equation that was used in the analysis. Finally, we show a correlation of thickness and interface structure to the magnitude of the asymptotic behavior, which provide insight into additional mechanisms that may induce deviations from the Kittel equation.2