In the brain, learning is achieved through the ability of synapses to reconfigure the strength by which they connect neurons (synaptic plasticity). In promising solid-state synapses called memristors, conductance can be finely tuned by voltage pulses and set to evolve according to a biological learning rule called spike-timing-dependent plasticity (STDP). Future neuromorphic architectures will comprise billions of such nanosynapses, which require a clear understanding of the physical mechanisms responsible for plasticity. Here we report on synapses based on ferroelectric tunnel junctions and show that STDP can be harnessed from inhomogeneous polarization switching. Through combined scanning probe imaging, electrical transport and atomic-scale molecular dynamics, we demonstrate that conductance variations can be modelled by the nucleation-dominated reversal of domains. Based on this physical model, our simulations show that arrays of ferroelectric nanosynapses can autonomously learn to recognize patterns in a predictable way, opening the path towards unsupervised learning in spiking neural networks.

ABINIT is a package whose main program allows one to find the total energy, charge density, electronic structure and many other properties of systems made of electrons and nuclei, (molecules and periodic solids) within Density Functional Theory (DFT), Many-Body Perturbation Theory (GW approximation and Bethe-Salpeter equation) and Dynmical Mean Field Theory (DMFT). ABINIT also allows to optimize the geometry according to the DFT forces and stresses, to perform molecular dynamics simulations using these forces, and to generate dynamical matrices, Born effective charges and dielectric tensors. The present paper aims to describe the new capabilities of ABINIT that have been developed since 2009. It covers both physical and technical developments inside the ABINIT code, as well as developments provided within the ABINIT package. The developments are described with relevant references, input variables, tests and tutorials.

The great interest in fuel cells inspires a substantial amount of research on nonprecious metal catalysts as alternatives to Pt-based oxygen reduction reaction (ORR) electrocatalysts. In this work, bimodal template-based synthesis strategies are proposed for the scalable preparation of hierarchically porous M-N-C (M = Fe or Co) single-atom electrocatalysts featured with active and robust MN 2 active moieties. Multiscale tuning of M-N-C catalysts regarding increasing the number of active sites and boosting the intrinsic activity of each active site is realized simultaneously at a singleatom scale. In addition to the antipoisoning power and high affinity for O 2 , the optimized Fe-N-C catalysts with FeN 2 active site presents a superior electrocatalytic activity for ORR with a half-wave potential of 0.927 V (vs reversible hydrogen electrode (RHE)) in an alkaline medium, which is 49 and 55 mV higher than those of the Co-N-C counterpart and commercial Pt/C, respectively. Density functional theory calculations reveal that the FeN 2 site is more active than the CoN 2 site for ORR due to the lower energy barriers of the intermediates and products involved. The present work may help rational design of more robust ORR electrocatalysts at the atomic level, realizing the significant advances in electrochemical conversion and storage devices. Single-Atom ElectrocatalystsThe ORCID identification number(s) for the author(s) of this article can be found under https://doi.

Atomically thin two-dimensional (2D) materials have received considerable research interest due to their extraordinary properties and promising applications. Here we predict the monolayered indium triphosphide (InP) as a new semiconducting 2D material with a range of favorable functional properties by means of ab initio calculations. The 2D InP crystal shows high stability and promise of experimental synthesis. It possesses an indirect band gap of 1.14 eV and a high electron mobility of 1919 cm V s, which can be strongly manipulated with applied strain. Remarkably, the InP monolayer suggests tunable magnetism and half-metallicity under hole doping or defect engineering, which is attributed to the novel Mexican-hat-like bands and van Hove singularities in its electronic structure. A semiconductor-metal transition is also revealed by doping 2D InP with electrons. Furthermore, monolayered InP exhibits extraordinary optical absorption with significant excitonic effects in the entire range of the visible light spectrum. All these desired properties render 2D InP a promising candidate for future applications in a wide variety of technologies, in particular for electronic, spintronic, and photovoltaic devices.