During the last half century, the tremendous development of computers based on von Neumann architecture has led to the revolution of the information technology. However, von Neumann computers are outperformed by the mammal brain in numerous data‐processing applications such as pattern recognition and data mining. Neuromorphic engineering aims to mimic brain‐like behavior through the implementation of artificial neural networks based on the combination of a large number of artificial neurons massively interconnected by an even larger number of artificial synapses. In order to effectively implement artificial neural networks directly in hardware, it is mandatory to develop artificial neurons and synapses. A promising advance has been made in recent years with the introduction of the components called memristors that might implement synaptic functions. In contrast, the advances in artificial neurons have consisted in the implementation of silicon‐based circuits. However, so far, a single‐component artificial neuron that will bring an improvement comparable to what memristors have brought to synapses is still missing. Here, a simple two‐terminal device is introduced, which can implement the basic functions leaky integrate and fire of spiking neurons. Remarkably, it has been found that it is realized by the behavior of strongly correlated narrow‐gap Mott insulators subject to electric pulsing.

No abstract

Progress of silicon-based technology is nearing its physical limit, as the minimum feature size of components is reaching a mere 10 nm. The resistive switching behavior of transition metal oxides and the associated memristor device is emerging as a competitive technology for next-generation electronics. Significant progress has already been made in the past decade, and devices are beginning to hit the market; however, this progress has mainly been the result of empirical trial and error. Hence, gaining theoretical insight is of the essence. In the present work, we report the striking result of a connection between the resistive switching and shock-wave formation, a classic topic of nonlinear dynamics. We argue that the profile of oxygen vacancies that migrate during the commutation forms a shock wave that propagates through a highly resistive region of the device. We validate the scenario by means of model simulations and experiments in a manganese-oxide-based memristor device, and we extend our theory to the case of binary oxides. The shock-wave scenario brings unprecedented physical insight and enables us to rationalize the process of oxygen-vacancy-driven resistive change with direct implications for a key technological aspect-the commutation speed. DOI: 10.1103/PhysRevX.6.011028 Subject Areas: Condensed Matter Physics, Materials Science, Nonlinear DynamicsThe information age we live in is made possible by a physical underlayer of electronic hardware, which originates in condensed-matter physics research. Despite the great progress made in recent decades, the demand for faster and power-efficient devices continues to grow. Thus, there is urgent need to identify novel materials and physical mechanisms for future electronic-device applications. In this context, transition metal oxides (TMOs) are attracting much attention for nonvolatile memory applications [1]. In particular, TMO are associated with the phenomenon of resistive switching (RS) [2] and the memristor device [3], which is emerging as a competitive technology for nextgeneration electronics [1,[4][5][6][7][8][9][10]. The RS effect is a large, rapid, nonvolatile, reversible change of the resistance, which may be used to encode logic information. In the simplest case, one may associate high and low resistance values to binary states, but multibit memory cells are also possible [11,12].Typical systems where RS is observed are two-terminal capacitor-like devices, where the dielectric might be a TMO and the electrodes are ordinary metals. The phenomenon occurs in a strikingly large variety of systemsranging from simple binary compounds, such as NiO, TiO 2 , ZnO, Ta 2 O 5 , HfO 2 , and CuO, to more complex perovskite structures, such as superconducting cuprates and colossal magnetoresistive manganites [2,4,6,9,13].From a conceptual point of view, the main challenges for a nonvolatile memory are as follows: (i) to change its resistance within nano seconds (required for modern electronics applications), (ii) to be able to retain the state for years (i.e., n...

Bipolar resistive switching in low cost n-Si/La2/3Ca1/3MnO3/M (M = Ti + Cu) devices was investigated. For low SET compliance currents (CC), an interfacial-related resistive switching mechanism, associated to the migration of oxygen vacancies close to the manganite/metal interface, is operative. Simulations using the voltage enhanced oxygen vacancies drift model validate our experimental results. When further increasing the CC, we have observed the onset of a second, filamentary, resistive switching regime with a concomitant collapse of the ON/OFF ratio. We finally demonstrate that it is possible to delay the onset of the filamentary regime by controlling the film thickness.

We consider the phenomenon of electric Mott transition (EMT), which is an electric induced insulator to metal transition. Experimentally, it is observed that depending on the magnitude of the electric excitation the final state may show a short lived or a long lived resistance change. We extend a previous model for the EMT to include the effect of local structural distortions through an elastic energy term. We find that by strong electric pulsing the induced metastable phase may become further stabilized by the electro-elastic effect. We present a systematic study of the model by numerical simulations and compare the results to new experiments in Mott insulators of the AM4Q8 family. Our work significantly extends the scope of our recently introduced leaky-integrateand-fire Mott-neuron [P. Stoliar Adv Mat 2017] to bring new insight on the physical mechanism of its relaxation. This is a key feature for future neuromorphic circuit implementations. arXiv:1711.05206v1 [cond-mat.mtrl-sci]

Vanadium dioxide (VO2) has attracted much attention owing to its metal–insulator transition near room temperature and the ability to induce volatile resistive switching, a key feature for developing novel hardware for neuromorphic computing. Despite this interest, the mechanisms for nonvolatile switching functioning as synapse in this oxide remain not understood. In this work, we use in situ transmission electron microscopy, electrical transport measurements, and numerical simulations on Au/VO2/Ge vertical devices to study the electroforming process. We have observed the formation of V5O9 conductive filaments with a pronounced metal–insulator transition and that vacancy diffusion can erase the filament, allowing for the system to “forget.” Thus, both volatile and nonvolatile switching can be achieved in VO2, useful to emulate neuronal and synaptic behaviors, respectively. Our systematic operando study of the filament provides a more comprehensive understanding of resistive switching, key in the development of resistive switching-based neuromorphic computing.

The use of mean-field models to describe the activity of large neuronal populations has become a very powerful tool for large-scale or whole brain simulations. However, the calculation of brain signals from mean-field models, such as the electric and magnetic fields, is still under development. Thus, the emergence of new methods for an accurate and efficient calculation of such brain signals is currently of great relevance. In this paper we propose a novel method to calculate the local field potentials (LFP) and magnetic fields from mean-field models. The calculation of LFP is done via a kernel method based on unitary LFP's (the LFP generated by a single axon) that was recently introduced for spiking-networks simulations and that we adapt here for mean-field models. The calculation of the magnetic field is based on current-dipole and volume-conductor models, where the secondary currents (due to the conducting extracellular medium) are estimated using the LFP calculated via the kernel method and the effects of medium-inhomogeneities are incorporated. We provide an example of the application of our method for the calculation of LFP and MEG under slow-waves of neuronal activity generated by a mean-field model of a network of Adaptive-Exponential Integrate-and-Fire (AdEx) neurons. We validate our method via comparison with results obtained from the corresponding spiking neuronal networks. Finally we provide an example of our method for whole brain simulations performed with The Virtual Brain (TVB), a recently developed tool for large scale simulations of the brain. Our method provides an efficient way of calculating electric and magnetic fields from mean-field models. This method exhibits a great potential for its application in large-scale or whole-brain simulations, where calculations via detailed biological models are not feasible.

Functional magnetic resonance imaging relies on the coupling between neuronal and vascular activity, but the mechanisms behind this coupling are still under discussion. Recent experimental evidence suggests that calcium signaling may play a significant role in neurovascular coupling. However, it is still controversial where this calcium signal is located (in neurons or elsewhere), how it operates and how relevant is its role. In this paper we introduce a biologically plausible model of the neurovascular coupling and we show that calcium signaling in astrocytes can explain main aspects of the dynamics of the coupling. We find that calcium signaling can explain so-far unrelated features such as the linear and non-linear regimes, the negative vascular response (undershoot) and the emergence of a (calcium-driven) Hemodynamic Response Function. These features are reproduced here for the first time by a single model of the detailed neuronal-astrocyte-vascular pathway. Furthermore, we analyze how information is coded and transmitted from the neuronal to the vascular system and we predict that frequency modulation of astrocytic calcium dynamics plays a key role in this process. Finally, our work provides a framework to link neuronal activity to the BOLD signal, and vice-versa, where neuronal activity can be inferred from the BOLD signal. This opens new ways to link known alterations of astrocytic calcium signaling in neurodegenerative diseases (e.g. Alzheimer’s and Parkinson’s diseases) with detectable changes in the neurovascular coupling.