A Two-Step Dry Etching Model for Non-Uniform Etching Profile in Gate-All-Around Field-Effect Transistor Manufacturing.

The Gate-All-Around Field-Effect Transistor (GAAFET) is proposed as a successor to Fin Field-Effect Transistor (FinFET) technology to increase channel length and improve the device performance. The GAAFET features a complex multilayer structure, which complicates the manufacturing process. One of the most critical steps in GAAFET fabrication is the selective lateral etching of the SiGe layers, essential for forming the inner-spacer. Industry commonly encounters a non-uniform etching profile during this step. In this paper, a continuous two-step dry etching model is proposed to investigate the mechanism behind the formation of the non-uniform profiles. The model consists of four modules: anisotropic etching simulation, Ge atom diffusion simulation, Si/SiGe etch selectivity calculation and SiGe selective etching simulation. By calibrating and verifying this model with experimental data, the edge rounding and gradient etching rates along the sidewall surface are successfully simulated. Based on further examination of the influence of chamber pressure on the profile using this model, the inner-spacer shape is improved experimentally by appropriately reducing the chamber pressure. This work aims to provide valuable insights for etching process recipes in advanced GAAFETs manufacturing.

Electrochemical-reaction-induced synaptic plasticity in MoOx-based solid state electrochemical cells.

Solid state electrochemical cells with synaptic functions have important applications in building smart-terminal networks. Here, the essential synaptic functions including potentiation and depression of synaptic weight, transition from short- to long-term plasticity, spike-rate-dependent plasticity, and spike-timing-dependent plasticity behavior were successfully realized in an Ag/MoOx/fluorine-doped tin oxide (FTO) cell with continual resistance switching. The synaptic plasticity underlying these functions was controlled by tuning the excitatory post-synaptic current (EPSC) decay, which is determined by the applied voltage pulse number, width, frequency, and intervals between the pre- and post-spikes. The physical mechanism of the artificial synapse operation is attributed to the interfacial electrochemical reaction processes of the MoOx films with the adsorbed water, where protons generated by water decomposition under an electric field diffused into the MoOx films and intercalated into the lattice, leading to the short- and long-term retention of cell resistance, respectively. These results indicate the possibility of achieving advanced artificial synapses with solid state electrochemical cells and will contribute to the development of smart-terminal networking systems.

Moisture effects on the electrochemical reaction and resistance switching at Ag/molybdenum oxide interfaces.

An important potential application of solid state electrochemical reactions is in redox-based resistive switching memory devices. Based on the fundamental switching mechanisms, the memory has been classified into two modes, electrochemical metallization memory (ECM) and valence change memory (VCM). In this work, we have investigated a solid state electrochemical cell with a simple Ag/MoO3-x/fluorine-doped tin oxide (FTO) sandwich structure, which shows a normal ECM switching mode after an electroforming process. While in the lower voltage sweep range, the switching behavior changes to VCM-like mode with the opposite switching polarity to the ECM mode. By current-voltage measurements under different ambient atmospheres and X-ray photoemission spectroscopy analysis, electrochemical anodic passivation of the Ag electrode and valence change of molybdenum ions during resistance switching have been demonstrated. The crucial role of moisture adsorption in the switching mode transition has been clarified based on the Pourbaix diagram for the Ag-H2O system for the first time. These results provide a fundamental insight into the resistance switching mechanism model in solid state electrochemical cells.

MEMS Oscillators-Network-Based Ising Machine with Grouping Method.

Combinatorial optimization (CO) has a broad range of applications in various fields, including operations research, computer science, and artificial intelligence. However, many of these problems are classified as nondeterministic polynomial-time (NP)-complete or NP-hard problems, which are known for their computational complexity and cannot be solved in polynomial time on traditional digital computers. To address this challenge, continuous-time Ising machine solvers have been developed, utilizing different physical principles to map CO problems to ground state finding. However, most Ising machine prototypes operate at speeds comparable to digital hardware and rely on binarizing node states, resulting in increased system complexity and further limiting operating speed. To tackle these issues, a novel device-algorithm co-design method is proposed for fast sub-optimal solution finding with low hardware complexity. On the device side, a piezoelectric lithium niobate (LiNbO3) microelectromechanical system (MEMS) oscillator network-based Ising machine without second-harmonic injection locking (SHIL) is devised to solve Max-cut and graph coloring problems. The LiNbO3 oscillator operates at speeds greater than 9 GHz, making it one of the fastest oscillatory Ising machines. System-wise, an innovative grouping method is used that achieves a performance guarantee of 0.878 for Max-cut and 0.658 for graph coloring problems, which is comparable to Ising machines that utilize binarization.

Semantic memory-based dynamic neural network using memristive ternary CIM and CAM for 2D and 3D vision.

The brain is dynamic, associative, and efficient. It reconfigures by associating the inputs with past experiences, with fused memory and processing. In contrast, AI models are static, unable to associate inputs with past experiences, and run on digital computers with physically separated memory and processing. We propose a hardware-software co-design, a semantic memory-based dynamic neural network using a memristor. The network associates incoming data with the past experience stored as semantic vectors. The network and the semantic memory are physically implemented on noise-robust ternary memristor-based computing-in-memory (CIM) and content-addressable memory (CAM) circuits, respectively. We validate our co-designs, using a 40-nm memristor macro, on ResNet and PointNet++ for classifying images and three-dimensional points from the MNIST and ModelNet datasets, which achieves not only accuracy on par with software but also a 48.1 and 15.9% reduction in computational budget. Moreover, it delivers a 77.6 and 93.3% reduction in energy consumption.

Spontaneous Threshold Lowering Neuron using Second-Order Diffusive Memristor for Self-Adaptive Spatial Attention.

Intrinsic plasticity of neurons, such as spontaneous threshold lowering (STL) to modulate neuronal excitability, is key to spatial attention of biological neural systems. In-memory computing with emerging memristors is expected to solve the memory bottleneck of the von Neumann architecture commonly used in conventional digital computers and is deemed a promising solution to this bioinspired computing paradigm. Nonetheless, conventional memristors are incapable of implementing the STL plasticity of neurons due to their first-order dynamics. Here, a second-order memristor is experimentally demonstrated using yttria-stabilized zirconia with Ag doping (YSZ:Ag) that exhibits STL functionality. The physical origin of the second-order dynamics, i.e., the size evolution of Ag nanoclusters, is uncovered through transmission electron microscopy (TEM), which is leveraged to model the STL neuron. STL-based spatial attention in a spiking convolutional neural network (SCNN) is demonstrated, improving the accuracy of a multiobject detection task from 70% (20%) to 90% (80%) for the object within (outside) the area receiving attention. This second-order memristor with intrinsic STL dynamics paves the way for future machine intelligence, enabling high-efficiency, compact footprint, and hardware-encoded plasticity.

A low-power vertical dual-gate neurotransistor with short-term memory for high energy-efficient neuromorphic computing.

Neuromorphic computing aims to emulate the computing processes of the brain by replicating the functions of biological neural networks using electronic counterparts. One promising approach is dendritic computing, which takes inspiration from the multi-dendritic branch structure of neurons to enhance the processing capability of artificial neural networks. While there has been a recent surge of interest in implementing dendritic computing using emerging devices, achieving artificial dendrites with throughputs and energy efficiency comparable to those of the human brain has proven challenging. In this study, we report on the development of a compact and low-power neurotransistor based on a vertical dual-gate electrolyte-gated transistor (EGT) with short-term memory characteristics, a 30 nm channel length, a record-low read power of ~3.16 fW and a biology-comparable read energy of ~30 fJ. Leveraging this neurotransistor, we demonstrate dendrite integration as well as digital and analog dendritic computing for coincidence detection. We also showcase the potential of neurotransistors in realizing advanced brain-like functions by developing a hardware neural network and demonstrating bio-inspired sound localization. Our results suggest that the neurotransistor-based approach may pave the way for next-generation neuromorphic computing with energy efficiency on par with those of the brain.

Interface-type tunable oxygen ion dynamics for physical reservoir computing.

Reservoir computing can more efficiently be used to solve time-dependent tasks than conventional feedforward network owing to various advantages, such as easy training and low hardware overhead. Physical reservoirs that contain intrinsic nonlinear dynamic processes could serve as next-generation dynamic computing systems. High-efficiency reservoir systems require nonlinear and dynamic responses to distinguish time-series input data. Herein, an interface-type dynamic transistor gated by an Hf0.5Zr0.5O2 (HZO) film was introduced to perform reservoir computing. The channel conductance of Mott material La0.67Sr0.33MnO3 (LSMO) can effectively be modulated by taking advantage of the unique coupled property of the polarization process and oxygen migration in hafnium-based ferroelectrics. The large positive value of the oxygen vacancy formation energy and negative value of the oxygen affinity energy resulted in the spontaneous migration of accumulated oxygen ions in the HZO films to the channel, leading to the dynamic relaxation process. The modulation of the channel conductance was found to be closely related to the current state, identified as the origin of the nonlinear response. In the time series recognition and prediction tasks, the proposed reservoir system showed an extremely low decision-making error. This work provides a promising pathway for exploiting dynamic ion systems for high-performance neural network devices.

Low power flexible monolayer MoS2 integrated circuits.

Monolayer molybdenum disulfide (ML-MoS2) is an emergent two-dimensional (2D) semiconductor holding potential for flexible integrated circuits (ICs). The most important demands for the application of such ML-MoS2 ICs are low power consumption and high performance. However, these are currently challenging to satisfy due to limitations in the material quality and device fabrication technology. In this work, we develop an ultra-thin high-κ dielectric/metal gate fabrication technique for the realization of thin film transistors based on high-quality wafer scale ML-MoS2 on both rigid and flexible substrates. The rigid devices can be operated in the deep-subthreshold regime with low power consumption and show negligible hysteresis, sharp subthreshold slope, high current density, and ultra-low leakage currents. Moreover, we realize fully functional large-scale flexible ICs operating at voltages below 1 V. Our process could represent a key step towards using energy-efficient flexible ML-MoS2 ICs in portable, wearable, and implantable electronics.

Long-Term Accuracy Enhancement of Binary Neural Networks Based on Optimized Three-Dimensional Memristor Array.

In embedded neuromorphic Internet of Things (IoT) systems, it is critical to improve the efficiency of neural network (NN) edge devices in inferring a pretrained NN. Meanwhile, in the paradigm of edge computing, device integration, data retention characteristics and power consumption are particularly important. In this paper, the self-selected device (SSD), which is the base cell for building the densest three-dimensional (3D) architecture, is used to store non-volatile weights in binary neural networks (BNN) for embedded NN applications. Considering that the prevailing issues in written data retention on the device can affect the energy efficiency of the system's operation, the data loss mechanism of the self-selected cell is elucidated. On this basis, we introduce an optimized method to retain oxygen ions and prevent their diffusion toward the switching layer by introducing a titanium interfacial layer. By using this optimization, the recombination probability of Vo and oxygen ions is reduced, effectively improving the retention characteristics of the device. The optimization effect is verified using a simulation after mapping the BNN weights to the 3D VRRAM array constructed by the SSD before and after optimization. The simulation results showed that the long-term recognition accuracy (greater than 105 s) of the pre-trained BNN was improved by 24% and that the energy consumption of the system during training can be reduced 25,000-fold while ensuring the same accuracy. This work provides high storage density and a non-volatile solution to meet the low power consumption and miniaturization requirements of embedded neuromorphic applications.

Oscillator-Network-Based Ising Machine.

With the slowdown of Moore's law, many emerging electronic devices and computing architectures have been proposed to sustain the performance advancement of computing. Among them, the Ising machine is a non-von-Neumann solver that has received wide attention in recent years. It is capable of solving intractable combinatorial optimization (CO) problems, which are difficult to be solve using conventional digital computers. In fact, many CO problems can be mapped to finding the corresponding ground states of Ising model. At present, Ising machine prototypes based on different physical principles, such as emerging memristive oscillators, have been demonstrated, among which the Ising Hamiltonian solver based on the coupled oscillator network simultaneously holds the advantages of room-temperature operation, compact footprint, low power consumption, and fast speed to solution. This paper comprehensively surveys the recent developments in this important field, including the types of oscillators, the implementation principle of the Ising model, and the solver's performance. Finally, methods to further improve the performance have also been suggested.

Local resistance switching at grain and grain boundary surfaces of polycrystalline tungsten oxide films.

Resistance switching behavior has been investigated in as-prepared and oxygen-annealed polycrystalline tungsten oxide films using conductive atomic force microscopy. The oxygen-annealed film appeared more insulative than the as-prepared films. The local current distributions demonstrated the lower conductivity at the grain boundaries than at the grains in the oxygen-annealed films. Reversible resistance switching behavior only occurred at the grain surface region of the oxygen-annealed films and the resistance switching process was described by the local valence change of tungsten ions induced by electrochemical migration of protons or oxygen vacancies. This different resistance switching behavior between the grain and grain boundary surface was attributed to the different oxygen vacancy density caused by the post-annealing process. The present results would be especially meaningful for the fabrication of nanoscale resistive nonvolatile memory devices.

Hybrid memristor-CMOS neurons for in-situ learning in fully hardware memristive spiking neural networks.

Spiking neural network, inspired by the human brain, consisting of spiking neurons and plastic synapses, is a promising solution for highly efficient data processing in neuromorphic computing. Recently, memristor-based neurons and synapses are becoming intriguing candidates to build spiking neural networks in hardware, owing to the close resemblance between their device dynamics and the biological counterparts. However, the functionalities of memristor-based neurons are currently very limited, and a hardware demonstration of fully memristor-based spiking neural networks supporting in-situ learning is very challenging. Here, a hybrid spiking neuron combining a memristor with simple digital circuits is designed and implemented in hardware to enhance neuron functions. The hybrid neuron with memristive dynamics not only realizes the basic leaky integrate-and-fire neuron function but also enables the in-situ tuning of the connected synaptic weights. Finally, a fully hardware spiking neural network with the hybrid neurons and memristive synapses is experimentally demonstrated for the first time, and in-situ Hebbian learning is achieved with this network. This work opens up a way towards the implementation of spiking neurons, supporting in-situ learning for future neuromorphic computing systems.

Nonvolatile Memory and Artificial Synapse Based on the Cu/P(VDF-TrFE)/Ni Organic Memtranstor.

We demonstrate a flexible nonvolatile multilevel memory and artificial synaptic devices based on the Cu/P(VDF-TrFE)/Ni memtranstor which exhibits pronounced nonlinear magnetoelectric effects at room temperature. The states of the magnetoelectric voltage coefficient αE of the memtranstor are used to encode binary information. By applying selective electric-field pulses, the states of αE can be switched repeatedly among 2n states (n = 1, 2, 3) in a zero dc bias magnetic field. In addition, the magnetoelectric coefficient is used to act as synaptic weight, and the induced magnetoelectric voltage VME is regarded as postsynaptic potentials (excitatory or inhibitory). The artificial synaptic devices based on the Cu/P(VDF-TrFE)/Ni memtranstor display the long-term potentiation (depression) and spiking-time-dependent plasticity behaviors. The advantages of a simple structure, flexibility, multilevel, and self-biasing make the Cu/P(VDF-TrFE)/Ni organic memtranstor a promising candidate for applications in nonvolatile memory as well as artificial synaptic devices with low energy consumption.

Oxide-Based Electrolyte-Gated Transistors for Spatiotemporal Information Processing.

Spiking neural networks (SNNs) sharing large similarity with biological nervous systems are promising to process spatiotemporal information and can provide highly time- and energy-efficient computational paradigms for the Internet-of-Things and edge computing. Nonvolatile electrolyte-gated transistors (EGTs) provide prominent analog switching performance, the most critical feature of synaptic element, and have been recently demonstrated as a promising synaptic device. However, high performance, large-scale EGT arrays, and EGT application for spatiotemporal information processing in an SNN are yet to be demonstrated. Here, an oxide-based EGT employing amorphous Nb2 O5 and Lix SiO2 is introduced as the channel and electrolyte gate materials, respectively, and integrated into a 32 × 32 EGT array. The engineered EGTs show a quasi-linear update, good endurance (106 ) and retention, a high switching speed of 100 ns, ultralow readout conductance (<100 nS), and ultralow areal switching energy density (20 fJ µm-2 ). The prominent analog switching performance is leveraged for hardware implementation of an SNN with the capability of spatiotemporal information processing, where spike sequences with different timings are able to be efficiently learned and recognized by the EGT array. Finally, this EGT-based spatiotemporal information processing is deployed to detect moving orientation in a tactile sensing system. These results provide an insight into oxide-based EGT devices for energy-efficient neuromorphic computing to support edge application.

Artificial Synapse Based on van der Waals Heterostructures with Tunable Synaptic Functions for Neuromorphic Computing.

Two-dimensional (2D) materials and van der Waals heterostructures have attracted tremendous attention because of their appealing electronic, mechanical, and optoelectronic properties, which offer the possibility to extend the range of functionalities for diverse potential applications. Here, we fabricate a novel multiterminal device with dual-gate based on 2D material van der Waals heterostructures. Such a multiterminal device exhibited excellent nonvolatile multilevel resistance switching performance controlled by the source-drain voltage and back-gate voltage. Based on these features, heterosynaptic plasticity, in which the synaptic weight can be tuned by another modulatory interneuron, has been mimicked. A tunable analogue weight update (both on/off ratio and update nonlinearity) of synapse with high speed (50 ns) and low energy (∼7.3 fJ) programming has been achieved. These results demonstrate the great potential of the artificial synapse based on van der Waals heterostructures for neuromorphic computing.

A highly CMOS compatible hafnia-based ferroelectric diode.

Memory devices with high speed and high density are highly desired to address the 'memory wall' issue. Here we demonstrated a highly scalable, three-dimensional stackable ferroelectric diode, with its rectifying polarity modulated by the polarization reversal of Hf0.5Zr0.5O2 films. By visualizing the hafnium/zirconium lattice order and oxygen lattice order with atomic-resolution spherical aberration-corrected STEM, we revealed the correlation between the spontaneous polarization of Hf0.5Zr0.5O2 film and the displacement of oxygen atom, thus unambiguously identified the non-centrosymmetric Pca21 orthorhombic phase in Hf0.5Zr0.5O2 film. We further implemented this ferroelectric diode in an 8 layers 3D array. Operation speed as high as 20 ns and robust endurance of more than 109 were demonstrated. The built-in nonlinearity of more than 100 guarantees its self-selective property that eliminates the need for external selectors to suppress the leakage current in large array. This work opens up new opportunities for future memory hierarchy evolution.

Mimicking Synaptic Plasticity and Neural Network Using Memtranstors.

Artificial synaptic devices that mimic the functions of biological synapses have drawn enormous interest because of their potential in developing brain-inspired computing. Current studies are focusing on memristive devices in which the change of the conductance state is used to emulate synaptic behaviors. Here, a new type of artificial synaptic devices based on the memtranstor is demonstrated, which is a fundamental circuit memelement in addition to the memristor, memcapacitor, and meminductor. The state of transtance (presented by the magnetoelectric voltage) in memtranstors acting as the synaptic weight can be tuned continuously with a large number of nonvolatile levels by engineering the applied voltage pulses. Synaptic behaviors including the long-term potentiation, long-term depression, and spiking-time-dependent plasticity are implemented in memtranstors made of Ni/0.7Pb(Mg1/3 Nb2/3 )O3 -0.3PbTiO3 /Ni multiferroic heterostructures. Simulations reveal the capability of pattern learning in a memtranstor network. The work elucidates the promise of memtranstors as artificial synaptic devices with low energy consumption.

Spin-induced multiferroicity in the binary perovskite manganite Mn2O3.

The ABO3 perovskite oxides exhibit a wide range of interesting physical phenomena remaining in the focus of extensive scientific investigations and various industrial applications. In order to form a perovskite structure, the cations occupying the A and B positions in the lattice, as a rule, should be different. Nevertheless, the unique binary perovskite manganite Mn2O3 containing the same element in both A and B positions can be synthesized under high-pressure high-temperature conditions. Here, we show that this material exhibits magnetically driven ferroelectricity and a pronounced magnetoelectric effect at low temperatures. Neutron powder diffraction revealed two intricate antiferromagnetic structures below 100 K, driven by a strong interplay between spin, charge, and orbital degrees of freedom. The peculiar multiferroicity in the Mn2O3 perovskite is ascribed to a combined effect involving several mechanisms. Our work demonstrates the potential of binary perovskite oxides for creating materials with highly promising electric and magnetic properties.

Giant magnetoelectric effects achieved by tuning spin cone symmetry in Y-type hexaferrites.

Multiferroics materials, which exhibit coupled magnetic and ferroelectric properties, have attracted tremendous research interest because of their potential in constructing next-generation multifunctional devices. The application of single-phase multiferroics is currently limited by their usually small magnetoelectric effects. Here, we report the realization of giant magnetoelectric effects in a Y-type hexaferrite Ba0.4Sr1.6Mg2Fe12O22 single crystal, which exhibits record-breaking direct and converse magnetoelectric coefficients and a large electric-field-reversed magnetization. We have uncovered the origin of the giant magnetoelectric effects by a systematic study in the Ba2-x Sr x Mg2Fe12O22 family with magnetization, ferroelectricity and neutron diffraction measurements. With the transverse spin cone symmetry restricted to be two-fold, the one-step sharp magnetization reversal is realized and giant magnetoelectric coefficients are achieved. Our study reveals that tuning magnetic symmetry is an effective route to enhance the magnetoelectric effects also in multiferroic hexaferrites.Control of the electrical properties of materials by means of magnetic fields or vice versa may facilitate next-generation spintronic devices, but is still limited by their intrinsically weak magnetoelectric effect. Here, the authors report the existence of an enhanced magnetoelectric effect in a Y-type hexaferrite, and reveal its underlining mechanism.