Search

The introduction of metal nanocrystals (NCs) has been confirmed to improve electrical uniformity of oxide-based resistive random access memory (RRAM) devices significantly; however, the current reports do not systematically elucidate the relationship between the size/distribution of NCs and the electrical uniformity of RRAM devices. In this paper, we focused on the impact of metal NCs size and areal density on the resistive switching (RS) performances of oxide RRAM by atomic layer deposition (ALD) based on the experimental results and theoretical calculation. The dependence of ALD cycles of 50–130 during Pt or CoPtx NCs growth on the RS parameters of Al2O3 or HfO2 memory units has been evaluated systematically. The RRAM embedded Pt or CoPtx NCs shows the trends: with increasing ALD cycles, the forming voltage, set/reset voltage, the resistance in off and on state, and ${R}_{ \mathrm{\scriptscriptstyle OFF}}/{R}_{ \mathrm{\scriptscriptstyle ON}}$ ratio entirely first decrease, then flatten, and increase later with a minimum value at about 100 cycles. Although all metal NCs with various sizes enhance the electric field strength compared to at the planar region, only metal NCs with proper NCs size and areal density (9 nm/6– $10\times 10^{\textsf {11}}$ /cm2 in this paper) can effectively produce stronger localized electric field at the tip of metal NCs, leading to optimal RS behavior. [ABSTRACT FROM AUTHOR]

Resistive switching memory (RRAM) features many optimal properties for future memory applications that make RRAM a strong candidate for storage-class memory and embedded nonvolatile memory. This paper addresses the cycling-induced degradation of RRAM devices based on a HfO2 switching layer. We show that the cycling degradation results in the decrease of several RRAM parameters, such as the resistance of the low-resistance state, the set voltage V\mathrm{ set} , the reset voltage V\mathrm{ reset} , and others. The degradation with cycling is further attributed to enhanced ion mobility due to defect generation within the active filament area in the RRAM device. A distributed-energy model is developed to simulate the degradation kinetics and support our physical interpretation. This paper provides an efficient methodology to predict device degradation after any arbitrary number of cycles and allows for wear leveling in memory array. [ABSTRACT FROM AUTHOR]

In spite of the apparent simplicity of the system under study, compact modeling of complementary resistive switching (CRS) devices, i.e., two antiserially connected memristive structures, is by no means straightforward. This requires a deep understanding of the voltage drops occurring across the circuit elements and correct treatment of the so-called snapback (SB) effect typical of the breakdown process of a thin dielectric film in an MIS or a MIM structure. The SB effect fundamentally consists in a sudden reduction of the device resistance caused by the formation of a filamentary path spanning the insulator. Beyond this point, the filament widens keeping the voltage drop across its extremes constant. The consequence of disregarding this effect is clearly distinguishable in the simulated curves: progressive openings of the on-state windows for both positive and negative voltages instead of abrupt ones. In this paper, a modification is introduced in the original memdiode model for memristive devices in order to account for this remarkable feature. Ta2O5-based CRS devices are considered to demonstrate the suitability of our approach. [ABSTRACT FROM AUTHOR]

In this paper, we investigate the relationship between size, shape, and the performance of the voltage-controlled magnetic anisotropy (VCMA)-based magnetic tunnel junctions (MTJs) suitable for gigabit scale MRAMs with 10 years of retention time. A Fokker–Planck simulation framework is developed to model the magnetization dynamics in the presence of thermal noise. Here, we numerically show that the optimization of the MTJ geometry can significantly improve the performance of the VCMA-based MTJs. Using an elliptical MTJ with aspect ratio (AR) of 3 reduces the required supply voltage and energy consumption by 80.7% and 92%, respectively, compared to a circular MTJ when the minimum feature size is 50 nm. The design requirements on the VCMA coefficient, ${\xi }$ , are also reduced by 67%. However, the influence of AR is observed to diminish with reduction in the minimum feature size of the MTJ. For instance, when the minimum feature size is 20 nm, the required supply voltage and energy consumption reduce only by 60.29% and 65.30%, respectively. We, then, perform a comprehensive scaling analysis on the VCMA-based MTJs by varying size, geometry, and ${\xi }$ of the MTJs. The predicted scaling trends are then compared with those of the STT MTJs. [ABSTRACT FROM AUTHOR]

Resistive random access memory (ReRAM), a new emerging nonvolatile memory technology based on changes in electrical resistivity of a dielectric film, offers promising advantages such as scalability, fast switching, and low operation voltage. However, for ReRAM to become a successful technology, it is necessary to accurately control the stochastic nature of the conductive nanoscale filaments (CNFs) that governs the resistive switching (RS) behavior of the device and limits its long-term stability and reliability. In this paper, we developed a highly scalable nanostructured/textured electrode that is composed of an array of Al nanotips based on an anodic aluminum oxide template. The nanotips improve the RS characteristics by intensifying the electric field at the apex of each nanotip which is demonstrated using numerical simulations. The localized electric field induces the repetitive nucleation/ formation/rupture of the CNFs in a more controlled fashion compared to a flat Al electrode. As a result, the nanotip sample exhibits uniform and reduced forming/reset voltages as low as 4.70 ± 0.98V/1.00 ± 0.19 V, stable endurance, and long-term retention. As a result, we were able to achieve ultralow-power and error-free operation of 100 cells covering a large area, significantly demonstrating improved uniformity and reliability compared to devices made using flat Al electrodes. This universal bottom-up strategy of self-organized nanostructured-electrodes provides a pathway toward large-scale, highly reliable, and RS memory devices. [ABSTRACT FROM AUTHOR]

This paper presents a comparison of switching performances between the in-situ oxide, gallium nitride (GaN) interlayer FET (OG-FET) and the conventional GaN metal–oxide–semiconductor FET (MOSFET), and explores the influence of the channel electron mobility on the device switching performance. GaN OG-FET is a novel structure with a pristine GaN layer grown between the gate oxide and the p-GaN enhancing the channel mobility up to 185 cm2/ $\text {V}\cdot \text {s}$ , which is over $3{\times}$ larger than that of the typical reported value (50 cm2/ $\text {V}\cdot \text {s}$) in GaN MOSFET. Owing to the high channel electron mobility, the GaN OG-FET showed a switching loss 30% lower than that of the conventional GaN trench MOSFET. Our results indicate that GaN OG-FET has the potential to attain greater efficiency, particularly at higher frequencies, showing a possible patch toward megahertz range conversions. [ABSTRACT FROM AUTHOR]

When developing silicon carbide (SiC) devices, the reliability of the internal body diode is an important issue. Stacking faults (SFs) are expanded from basal plane defects by using a body diode. Although it is well known that ON-voltage degradation occurs due to SFs, their effects on dynamic reliability have not been studied well. Furthermore, including the effects of SFs, there are not many reports about the dynamic reliability of SiC-MOSFETs. In this paper, a double-pulse switching test using 3.3-kV SiC-MOSFETs was carried out to clarify the failure mechanism of the switching operation of SiC-MOSFETs and effects of SFs on the safe operation area (SOA). Before the switching test, current stress was applied to the body diode of the devices under test (DUTs) to expand the SFs. The circuit configuration was half-bridge type, and a double-pulse gate signal was applied to the lower arm DUT. The switching voltage was 1.8 kV, and the switching current increased at about 8-A steps to failure. As a result, reverse recovery SOA (RRSOA) reliability decreased depending on the amount of SFs in the SiC-MOSFET. Because RRSOA failure was caused by avalanche due to the hole concentration during reverse recovery and the SFs raised local current density, reverse bias SOA (RBSOA) hardly decreased even if SFs containing SiC-MOSFETs were used. This is because RBSOA failure was caused by degradation of the gate isolation layer due to overheating and the temperature coefficient of the SFs electric resistance indicated negative. [ABSTRACT FROM AUTHOR]

In this paper, we demonstrate the linear conductance-change characteristics of a conductive-bridging RAM (CBRAM) to be employed as an artificial synapse device in neuromorphic systems. The CBRAM with a bilayer electrolyte structure (${\mathrm {Cu/Cu}}_{{2}-{x}}\text{S}$ / $\mathrm {WO}_{{3}-{x}}$ /W) exhibits analog switching behavior during the depression process due to the well-controlled dissolution of the conductive filament. To analyze the origin of this motion, we investigate the effective voltage applied to $\mathrm {Cu}_{{2}-{x}}\text{S}$ and $\mathrm {WO}_{{3}-{x}}$. Our findings reveal that $\mathrm {Cu}_{{2}-{x}}\text{S}$ , acting as a voltage divider, helps in suppressing the large voltage drop in $\mathrm {WO}_{{3}-{x}}$ , where the formation/dissolution of filament occurs. Furthermore, due to the diode-like characteristics of $\mathrm {Cu}_{{2}-{x}}\text{S}$ and the division of voltage drop between $\mathrm {WO}_{{3}-{x}}$ and $\mathrm {Cu}_{{2}-{x}}\text{S}$ , an optimum programming energy is applied to $\mathrm {WO}_{{3}-{x}}$ during the depression process. This leads to linear conductance-change characteristics under identical pulses. However, abrupt conductance-change characteristics are observed during the potentiation process. Thus, we use only the device characteristics of the depression part for the neuromorphic system. An excellent classification accuracy is achieved due to the linear conductance-change characteristics and optimized pulse conditions. [ABSTRACT FROM AUTHOR]

This paper reports on the suppression of single-transistor latch, an extreme condition of impact ionization (I.I.), in n-type double-gate germanium (Ge) junctionless (JL) transistors. It is shown that JL transistors can be latched to the on-state due to an increase in the I.I. generated power per unit volume with an increase in drain bias, film thickness, oxide thickness, and doping. Latch, being detrimental to switching, can be effectively suppressed by applying an appropriate negative (positive) back-gate bias for n-type (p-type) JL MOSFET. The independent gate operation with appropriate back bias allows the device to overcome latching and regain switching action with a subthreshold swing (${S}$ -swing) <10 mV/decade at room temperature. The work showcases the limit imposed on back bias due to an increase in off-state current due to band-to-band tunneling along with the utility of sidewall spacer to effectively suppress the same. We also show the tunability of hysteresis using back-gate bias. The work highlights a systematic methodology to eliminate single-transistor latch while preserving sub-60 mV/decade switching in Ge JL MOSFETs. [ABSTRACT FROM AUTHOR]

Bipolar redox-based resistive random-access memory cells are intensively studied for new storage class memory and beyond von Neumann computing applications. However, the considerable variability of the resistance values in ON and OFF state as well as of the SET voltage remains challenging. In this paper, we discuss the physical origin of the significant reduction in the switching variability of HfO2-based devices achieved by the insertion of a thin TiOx layer between the HfO2 layer and the oxygen exchange metal layer. Typically, HfO2 single layer cells exhibit an abrupt SET process, which is difficult to control. In contrast, self-compliance effects in the HfO2/TiOx bilayer devices lead to an increased stability of SET voltages and OFF-state resistances. The SET process is gradual and the RESET becomes abrupt for higher switching currents. Comparison of the experimental data with simulation results achieved from a physics-based compact model for the full description of the switching behavior of the single layer and bilayer devices clearly reveal three major effects. The TiOx layer affects the temperature distribution during switching (by modifying the heat dissipation), forms an additional series resistance and changes the current conduction mechanism in the OFF state of the bilayer device compared to the single layer device. [ABSTRACT FROM AUTHOR]

Performance of the crossbar memory array highly depends on the selector characteristics. In this paper, rigorous transient analyses are performed for a large-size crossbar memory array using novel NbO2-based selectors with a threshold switching behavior. To enable accurate and efficient array-level simulation, an electrostatic discharge-based compact model is employed to effectively describe the ${I}$ – ${V}$ characteristics of the selector. Multiple key design parameters of the selector are investigated, such as the threshold voltage, leakage current, and intrinsic switching speed. A sensitivity analysis is performed to evaluate the impact of hypothetical improvements in various selector parameters. In addition, the impacts of resistances of interconnect and memory element on the array-level access delay and energy dissipation are quantified. The results show that reducing the threshold voltage of selectors provides the most significant performance improvement, where up to 80% of the energy-delay product saving is observed if the threshold voltage is reduced by 50%. [ABSTRACT FROM AUTHOR]

A comprehensive study on the dielectric modulated (DM) drift region for power devices is presented in this paper. The performance of this drift region structure is theoretically analyzed and compared with other two structures: conventional and superjunction. In this paper, a study is focused on switching performance. The C – V relationships during switching are analytically derived, and the depletion widths predicted by the proposed analytical model agree well with simulation in a wide reverse bias region. The switching energy loss is evaluated based on the analytical model and its accuracy is verified by 2-D simulations. An effective depletion charge QD is defined as E\mathrm {loss} / breakdown voltage (BV), from which a figure of merit (FOM) of R\mathrm{\scriptscriptstyle ON,\mathrm {sp}}{}^\ast QD is proposed to compare the performances of these different drift region structures, considering both the conduction and switching losses. The discussions in part I on Si and silicon carbide device structures are extended in part II to include FOMs. It is found that the DM drift region structures show better R\mathrm{\scriptscriptstyle ON,\mathrm {sp}} -BV tradeoffs compared with the conventional structures, but slighter inferior FOMs. However, DM structures are still attractive for applications, where switching loss is less critical than the conduction loss, such as low frequency or soft switching. [ABSTRACT FROM PUBLISHER]

Transport properties of realistic phosphorene nanoribbons (PNRs) with edge defects are studied by using statistical atomistic quantum transport simulations. Regarding the impact of nanoribbon width downscaling and increasing of edge defect percentage, we show that PNRs exhibit qualitatively similar behavior of the ON- and OFF-state conductance, and of the ON–OFF conductance ratio as graphene nanoribbons (GNRs). However, we demonstrate that PNRs are superior to GNRs in terms of the absolute values of the conductance parameters, and that PNRs are much more immune to edge defects than their graphene counterparts. This paper identifies PNRs as a more promising channel material than GNRs for the extremely scaled postsilicon transistor technology. [ABSTRACT FROM PUBLISHER]

The efficiency of the human brain in performing classification tasks has attracted considerable research interest in brain-inspired neuromorphic computing. The spiking neuromorphic architectures attempt to mimic the computations performed in the brain through a dense interconnection of the neurons and synaptic weights. A leaky-integrate-fire (LIF) spiking model is widely used to emulate the dynamics of the biological neurons. In this paper, we propose a spin-based LIF spiking neuron using the magnetoelectric (ME) switching of ferromagnets. The voltage across the ME oxide exhibits a typical leaky-integrate behavior, which in turn switches an underlying ferromagnet. Due to the effect of thermal noise, the ferromagnet exhibits probabilistic switching dynamics, which is reminiscent of the stochasticity exhibited by biological neurons. The energy efficiency of the ME switching mechanism coupled with the intrinsic nonvolatility of ferromagnets results in lower energy consumption, when compared with a CMOS LIF neuron. A device to system-level simulation framework has been developed to investigate the feasibility of the proposed LIF neuron for a hand-written digit recognition application. [ABSTRACT FROM AUTHOR]

The effect of voltage on the anisotropy of a magnetic tunnel junction (MTJ) is of substantial interest for low-power nonvolatile memory applications. In this paper, we develop a device-to-bit-cell level simulation framework for voltage-assisted switching of the MTJs, which can satisfactorily reproduce the published experimental data. Our simulation framework is based on a coupled Landau–Lifshitz–Gilbert–Slonczewski equation and nonequilibrium greens function formalism. Using this simulation framework, we investigate the effect of scaling the oxide thickness of an MTJ to enhance the electric field effect. Although it seems attractive for an isolated device, yet, in a bit-cell configuration, reducing the oxide thickness leads to an increase in the supply voltage. In addition, we demonstrate that the unipolar characteristic of the voltage-assisted switching leads to write failures depending on the initial state stored in the MTJ. We, therefore, suggest a read before write scheme exhibiting approximately $2\times $ improvement in the write energy consumption at a bit-cell level compared with a standard STT-MRAM for iso-oxide thickness. However, the latency overhead associated with voltage-assisted MTJs results in a $1.55\times $ degradation of the write speed. Moreover, our layout analysis indicates that the bit-cell area reduction is constrained by the metal pitch, in spite of smaller access transistors. [ABSTRACT FROM AUTHOR]

This paper presents a comprehensive simulation analysis of the dynamic operation of T-RAM cells. The analysis addresses the evolution of cell electrostatics and of carrier concentrations and flows in the device when the gate voltage undergoes a fast low-to-high switch, as required by both read and write operations. The results clarify, first of all, the basic physics making holes in the p-base the physical element allowing information storage in the device. From this starting point, the working principles exploited for DRAM operation of the memory cell are then explained. [ABSTRACT FROM AUTHOR]

Gallium nitride (GaN) devices are gathering momentum, with a number of recent market introductions for a wide range of applications such as point-of-load converters, OFF-line switching power supplies, battery chargers, and motor drives. This paper studies the basic characteristics of a 600 V cascode GaN switch, such as voltage distribution during the turn-ON and turn-OFF transition. The switching loss mechanism of the cascode GaN switch is analyzed in detail, including the impact of the package parasitic inductance in both hard- and soft-switching modes. A soft-switching 5 MHz boost converter is developed and shows the advantages and the potential of the cascode GaN. [ABSTRACT FROM AUTHOR]

Electrical control of magnetic tunnel junctions (MTJs) provides opportunities to introduce MTJs into high-performance applications requiring low power consumption. The magnetic state of an MTJ can be electrically controlled through: 1) the spin transfer torque (STT) effect; 2) the voltage controlled magnetic anisotropy (VCMA) effect; and 3) the fusion of STT and VCMA. Several compact models have been published for MTJs. All of these models consider an MTJ whose magnetic state is controlled through the STT effect. In this paper, a model of an MTJ comprising a free layer, an analysis layer, and a spin polarizing layer is described. The MTJ compact model, adaptive compact MTJ (ACM) model, includes the effects of asymmetry on the MTJ behavior, and models a device controlled through the STT, VCMA, or a fused STT–VCMA mechanism. The ACM model includes the dynamics of the junction temperature. The proposed model can be adapted to experimental configurations including in-plane MTJ (IMTJ), IMTJ with a perpendicular-to-the-plane polarizer, perpendicular-to-the-plane MTJ (PMTJ), and PMTJ with an additional easy axis. The ACM model is validated with published experimental data, showing reasonably accurate results with an average error of less than 6%. [ABSTRACT FROM AUTHOR]

This paper reports the design, fabrication, characterization, and scaling analysis of novel buckling-based nanoelectromechanical relays that use, for the first time, piezoelectric actuation. The generated stress from anchored piezoelectric films is employed to buckle a clamped-clamped beam to connect suspended source and drain terminals, while the residual stress in the beam is used to control the actuation voltage. This demonstration is the first of its kind to exploit residual stress to deliver a highly scalable switching mechanism that exhibits low actuation voltage ( \(\sim 1.8\) V), very fast switching ( \(\sim 80\) ns), and uniquely achieves an equivalent electric body bias via mechanical methods. Finite-element analysis and analytical simulations show a linear dependence of the switching voltage on the residual stress, while the voltage versus stress tuning slope is reduced linearly with scaling the piezoelectric film thickness (−16.6 mV/MPa for 10-nm thick aluminum nitride film). A scaling analysis shows that millivolt switching is possible for aggressively miniaturized relays, which highlights the great impact this technology could have on demonstrating ultralow power digital circuits and memories. [ABSTRACT FROM PUBLISHER]

Resistive switching memory (RRAM) relies on the voltage-driven formation/disruption of a conductive filament (CF) across a thin insulating layer. Due to the 1-D structure of the CF and discrete nature of defects, the set and reset states of the memory device generally display statistical variability from cycle to cycle. For projecting cell downscaling and designing improved programming operations, the variability as a function of the operation parameters, such as the maximum current in the set process and maximum voltage in the reset process, need to be evaluated and understood. This paper addresses set/reset variability, presenting statistical data for HfOx-based RRAM and introducing a physics-based Monte Carlo model for switching statistics. The model can predict the distribution of the set state as a function of the compliance (maximum) current during set and distribution of the reset state as a function of the stop (maximum) voltage during reset. Numerical modeling results are finally presented to provide additional insight into discrete fluctuation events. [ABSTRACT FROM AUTHOR]

The resistive random access memory (RRAM) crossbar array has been extensively studied as one of the most promising candidates for future high-density nonvolatile memory technology. However, some problems caused by circuit and device interaction, such as sneak leakage paths, result in limited array size and large power consumption, which degrade the array performance significantly. Thus, the analysis on circuit and device interaction issue is imperative. In this paper, a simulation method is developed to investigate the critical issues correlated with the interaction between devices and the circuit. The simulations show that a large off/on ratio of resistance states of RRAM is beneficial for large readout margin (i.e., array size). The existence of the selector connected in series with an RRAM device can eliminate the need for high Ron resistance, which is critical for the array consisted of only RRAM cells. The readout margin is more sensitive to the variation of Ron and is determined by the nonlinearity of the I–V characteristics of RRAM, whereas the nonlinear characteristics of the selector device are beneficial for a larger readout margin. An optimal design scheme for turn-on voltage and conductance of the selector is proposed based on the simulation. [ABSTRACT FROM PUBLISHER]