No abstract

The adjustable conductance of a two-terminal memristor in a crossbar array can facilitate vector-matrix multiplication in one step, making the memristor a promising synapse for efficiently implementing neuromorphic computing. To achieve controllable and gradual switching of multi-level conductance, important for neuromorphic computing, a theoretical design of a superlattice-like (SLL) structure switching layer for the multi-level memristor is proposed and validated, refining the growth of conductive filaments (CFs) and preventing CFs from the abrupt formation and rupture. Ti/(HfO x /AlO y ) SLL /TiN memristors are shown with transmission electron microscopy , X-ray photoelectron spectroscopy , and ab initio calculation findings corroborate the SLL structure of HfO x /AlO y film. The optimized SLL memristor achieves outstanding conductance modulation performance with linearly synaptic weight update (nonlinear factor 𝜶 = 1.06), and the convolutional neural network based on the SLL memristive synapse improves the handwritten digit recognition accuracy to 94.95%. Meanwhile, this improved synaptic device has a fast operating speed (30 ns), a long data retention time (≥ 10 4 s at 85 °C), scalability, and CMOS process compatibility. Finally, its physical nature is explored and the CF evolution process is characterized using nudged elastic band calculations and the conduction mechanism fitting. In this work, as an example the HfO x /AlO y SLL memristor provides a design viewpoint and optimization strategy for neuromorphic computing.

No abstract

No abstract

In this letter, a method to realize all-optical helicity-dependent magnetic switching (AO-HDS) using a first-order azimuthally polarized vortex (FAPV) beam is demonstrated. Numerical calculations of the focal fields of FAPV beams reveal that left-handed and right-handed circular polarizations are generated due to the interaction between the polarization singularity and the helical wave front. Its feasibility for AO-HDS is experimentally demonstrated in Gd27Fe63.87Co9.13 under low numerical aperture (NA) conditions and within a narrow fluence window. It is numerically predicted that under high NA conditions, the lateral size of magnetic bits recorded by FAPV beams can be nearly 30% smaller than that obtained by circularly polarized beams, which opens a promising route to realize ultrafast and ultrahigh-density magnetic recording.

We report an investigation of temperature dependent spin Hall magnetoresistance (SMR) and anisotropic magnetoresistance (AMR) in Ta/CoFe2O4 nanostructures. The AMR of the Ta/CoFe2O4 nanostructure starts to appear at 50 K and its magnitude enhances dramatically with the decrease in temperature due to the suppressed spin-flip scattering. However, the SMR shows a complex temperature dependence correlated with the thickness of Ta layers. It increases monotonically and slightly with the decrease in temperature in thicker (7 nm) Ta. Moreover, Ta/CoFe2O4 nanostructures with thinner (3 nm) Ta exhibit a significant peak of SMR at about 75 K, probably owing to a good matching between the Ta layer thickness and its spin diffusion length. The fundamental distinct temperature dependences reveal different transport mechanisms of the two magnetoresistance effects. Our results will contribute to the further understanding and optimization of the magnetoresistance effects in spinel magnetic insulator heterojunctions.

Influence of high temperature on corrosion behavior of 304 stainless steel in chloride solutions AIP Advances 6, 115301 (2016) With the advance of HAMR technology, the storage potential of L1 0 -FePt nanoparticles with practical considerations are of great significance. We present an L1 0 -FePt nanoparticle model based on atomistic spin model with Langevin thermodynamics to simulate the magnetic behaviors of L1 0 -FePt nanoparticles at Curie temperature and room temperature to explore their practical design margins. Given specific composites, the Curie temperatures of L1 0 -FePt nanoparticles decrease with their volume sizes decrease starting from 8 nm, meaning no more laser power increment needed for smaller L1 0 -FePt nanoparticles. However, L1 0 -FePt nanoparticles get unstable more easily while their volume sizes decrease at room temperature within 10 years. Above all, a reasonable size of L1 0 -FePt nanoparticles for stable information retaining should not be less than 8 nm at a certain aspect ratio. © 2016 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license

SmCo 5 is an emerging perpendicular magnetic material for super-high density magnetic recording, due to its large magnetic anisotropy energy. In this paper, the magnetic moments of SmCo 5−x Cu x have been studied using first principles calculation based on density-functional theory (DFT). Calculations are performed using the pseudopotential plane wave DFT code Vienna ab initio simulation package (VASP) with the projector augmented wave (PAW) method. The local density approximation LDA+U method is used for the calculation of the exchange correlation energy of Sm. The calculation results show that the average Co magnetic moment of SmCo 5−x Cu x decreases with the increase of Cu doping concentrations, and the influence of the Cu doping on the spin state of Co is greater than that of Sm. The magnetic anisotropy energy of SmCo 5 is analyzed. The electronic density of states and the differential in spin densities of atoms show that the spatial distribution of 4f electron and the 4f -3d coupling are the controlling factors of the magnetic anisotropy energy of SmCo 5 .