However, the size and direction of the current induced effective field seems to vary depending on the system and the underlying mechanism of such field generation is not well understood. For example, the effective field in Pt|Co|AlOx is reported to be ~3000 to ~10000 Oe for a current density of 10 8 A/cm 2 , pointing perpendicular to both the film normal and the current flow direction (defined as a transverse field hereafter) 4,13 . More recently, in the same system, signs of current induced effective field directed along the current flow, i.e. Here we show a systematic study of the current induced effective field in Ta|CoFeB|MgO.We use a low current excitation technique to quantitatively evaluate the size and direction of the effective field. We find that the size and even the sign of the transverse and longitudinal effective fields vary as the Ta layer thickness is changed, suggesting competing contributions from two distinct sources. We find that the transverse effective field is larger than the longitudinal field, by nearly a factor of three, when the Ta layer thickness is large. In contrast, the relative size of the two components shows somewhat an oscillatory dependence on the Ta thickness for films with thin Ta.Films are sputtered on a highly resistive silicon substrate coated with 100 nm thick thermally oxidized SiO 2 . We use a linear shutter during the sputtering to vary the thickness of one layer in each substrate. Two film stacks are prepared here: Ta wedge: Si-sub|d Ta Ta|1 Co 20 We set the thickness of each layer, d Ta and t CoFeB , to vary from ~0 to ~2 nm. Throughout this paper, the nominal thickness is used for the Ta wedge. For the CoFeB wedge film, correction of the thickness was required due to technical reasons, and thus we use our resistivity results to calibrate the thickness. All films are annealed at 300 °C for one hour ex-situ after the film deposition. Photo-lithography and Ar ion etching are used to pattern Hall bars from the film and a lift off process is used to form the contact electrodes (10 Ta|100 Au). Prior to the deposition of the contact electrodes, we etch the Ta capping layer and nearly half of the MgO layer to avoid large contact resistance. Although etching of the MgO layer significantly influences the magnetic anisotropy of the CoFeB layer under the etched region 24 , here we assume that this has little effect on the evaluation of the current induced effective fields since we limit the applied field smaller than the magnetization switching field. 4Schematic illustration of the experimental set up and definition of the coordinate system are shown in Fig 1(a). The width and length of typical wires measured are 10 m and 60 m, respectively. We measure wires with different width, ranging from 5 m to 30 m, and find little dependence on the width for most of the parameters shown here. Positive current is defined as current flowing along the +y direction in Fig 1(a). Current is fed into the wire and the Hall voltage is measured in all experiments. Using the Extraordinary Hal...

Recent advances in the understanding of spin orbital effects in ultrathin magnetic heterostructures have opened new paradigms to control magnetic moments electrically. The Dzyaloshinskii-Moriya interaction (DMI) is said to play a key role in forming a Néel-type domain wall that can be driven by the spin Hall torque. Here we show that the strength and sign of the DMI can be changed by modifying the adjacent heavy-metal underlayer (X) in perpendicularly magnetized X/CoFeB/MgO heterostructures. The sense of rotation of a domain wall spiral is reversed when the underlayer is changed from Hf, Ta to W and the strength of DMI varies as the filling of 5d orbitals, or the electronegativity, of the heavy-metal layer changes. The DMI can even be tuned by adding nitrogen to the underlayer, thus allowing interface engineering of the magnetic texture in ultrathin magnetic heterostructures.

Conversion of charge current into pure spin current and vice versa in non-magnetic semiconductors or metals, which are called the direct and inverse spin Hall effects (SHEs), provide a new functionality of materials for future spin-electronic architectures. Thus, the realization of a large SHE in a device with a simple and practical geometry is a crucial issue for its applications. Here, we present a multi-terminal device with a Au Hall cross and an FePt perpendicular spin injector to detect giant direct and inverse SHEs at room temperature. Perpendicularly magnetized FePt injects or detects perpendicularly polarized spin current without magnetic field, enabling the unambiguous identification of SHEs. The unprecedentedly large spin Hall resistance of up to 2.9 mOmega is attributed to the large spin Hall angle in Au through the skew scattering mechanism and the highly efficient spin injection due to the well-matched spin resistances of the chosen materials.

Spin Hall magnetoresistance (SMR) is studied in metallic bilayers that consist of a heavy metal (HM) layer and a ferromagnetic metal (FM) layer. We find a nearly tenfold increase of SMR in W/CoFeB compared to previously studied HM/ferromagnetic insulator systems. The SMR increases with decreasing temperature despite the negligible change in the W layer resistivity. A model is developed to account for the absorption of the longitudinal spin current to the FM layer, one of the key characteristics of a metallic ferromagnet. We find that the model not only quantitatively describes the HM layer thickness dependence of SMR, allowing accurate estimation of the spin Hall angle and the spin diffusion length of the HM layer, but also can account for the temperature dependence of SMR by assuming a temperature dependent spin polarization of the FM layer. These results illustrate the unique role a metallic ferromagnetic layer plays in defining spin transmission across the HM/FM interface.

No abstract