Search

A quantitative evaluation of the intrinsic reliability limits of floating-gate (FG) memories in the decananometer range due to the reduction of collective phenomena and to the dominance of single-electron stochastic behaviors is presented. A model that predicts the intrinsic dispersions of the memory retention time and programming window is proposed, and the experimental results obtained on ultrasound memory devices are used to validate this model.

A study was conducted on aggressive technology scaling down to achieve increasingly higher very large scale integration densities and performance. It reports the results of an extensive examination of the data retention of 0.4-micron EPROM devices under a broad range of baking temperatures. Data retention dependence on temperature is described by a new 'T' model which considerably differs from the typically assumed Arrhenius law.

Abstract: Phase change memory can provide a remarkable artificial synapse for neuromorphic systems, as it features excellent reliability and can be used as an analog memory. However, this approach is complicated by the fact that crystallization and amorphization differ radically: crystallization can be realized in a very gradual manner, very similarly to synaptic potentiation, while the amorphization process tends to be abrupt, unlike synaptic depression. Addressing this non‐biorealism of amorphization requires system‐level solutions that have considerable energy cost or limit the generality of the approach. This work demonstrates experimentally that an adaptation of the memory structure associated with an initialization electrical pulse followed by a sequence of identical fast pulses can overcome this challenge. A single device can then naturally implement gradual long‐term potentiation and depression, much like synapses in biology. This study evidences through statistical measurements the reproducibility of the approach, discusses its physical origin, as well as the importance of the device architecture and of the initial electrical pulse. Through the use of system‐level simulation, it is shown that this device is especially adapted to a neuroscience‐inspired learning. These results highlight how nanodevices can be suitable for bioinspired applications while retaining the qualities of industrial technology. [ABSTRACT FROM AUTHOR]

In this paper, we present an alternative approach to perform spike sorting of complex brain signals based on spiking neural networks (SNN). The proposed architecture is suitable for hardware implementation by using resistive random access memory (RRAM) technology for the implementation of synapses whose low latency (<1μs) enables real-time spike sorting. This offers promising advantages to conventional spike sorting techniques for brain-computer interfaces (BCI) and neural prosthesis applications. Moreover, the ultra-low power consumption of the RRAM synapses of the spiking neural network (nW range) may enable the design of autonomous implantable devices for rehabilitation purposes. We demonstrate an original methodology to use Oxide based RRAM (OxRAM) as easy to program and low energy (<75 pJ) synapses. Synaptic weights are modulated through the application of an online learning strategy inspired by biological Spike Timing Dependent Plasticity. Real spiking data have been recorded both intra- and extracellularly from an in-vitro preparation of the Crayfish sensory-motor system and used for validation of the proposed OxRAM based SNN. This artificial SNN is able to identify, learn, recognize and distinguish between different spike shapes in the input signal with a recognition rate about 90% without any supervision. [ABSTRACT FROM AUTHOR]

We investigate in detail the effects of metal electrodes on the switching performance and conductive filament (CF) stability of HfO2-based RRAM. The current–voltage characteristics of the devices exhibit different electrode-dependent RESET profiles which we attempt to clarify. With the insight from the experimental data, we employ first-principles calculations to have a better microscopic understanding of the devices. We study the charge injection, formation of Frenkel pairs, and diffusion of oxygen defects (oxygen vacancies Vo and oxygen interstitials Oi) that are important in the CF creation and stability during the device operation. Since the presence of Ti in RRAM has been associated with the creation of substoichiometric TiOy region at the Ti/HfO2 interface, we also explore different Ti and Hf suboxides to understand the possible composition of that interface. Our calculations suggest that the composition of the interface would be Ti2O/Hf2O3 from thermodynamic perspective. By combining the experimental and calculations results, we show that the concentration of oxygen interstitial (Oi) ions in the oxide after CF formation is larger for RRAM devices with inert electrodes (like Pt) compared with O reactive electrodes (like Ti) which results in degraded device performance. The lower Oi concentration in HfO2 layer with Ti electrodes results in improved CF thermal stability and device variability. [ABSTRACT FROM AUTHOR]

We study in detail the impact of alloying HfO2 with Al (Hf1–xAl2xO2+x) on the oxide-based resistive random access memory (RRAM) (OxRRAM) thermal stability through material characterization, electrical measurements, and atomistic simulation. Indeed, migration of oxygen atoms inside the dielectric is at the heart of OxRRAM operations. Hence, we performed comprehensive diffusion barrier calculations in HfO2, Hf1–xAl2xO2+x, and Hf1–xTixO2 relative to the oxygen vacancy ( Vo ) movement involved in low-resistance state ( R\mathrm{\scriptscriptstyle ON} ) thermal stability. Calculations are performed at the best level using ab initio techniques. This paper provides an insight on the improved R\mathrm{\scriptscriptstyle ON} stability of our Hf1–xAl2xO2+x-based RRAM devices and predicts the degraded retention of Hf1–xTixO2-based RRAM measured in the literature. Our theoretical calculations link the origin of R\mathrm{\scriptscriptstyle ON} retention failure to the lateral diffusion of oxygen vacancies at the constriction/tip of the conductive filament in HfO2-based RRAM. [ABSTRACT FROM PUBLISHER]

In this paper, we investigate in depth Forming, SET, and Retention of conductive-bridge random-access memory (CBRAM). A kinetic Monte Carlo model of the CBRAM has been developed considering ionic hopping and chemical reaction dynamics. Based on inputs from ab initio calculations and the physical properties of the materials, the model offers the simulation of both the Forming/SET and the Data Retention operations. It aims to create a bond between the physics at atomic level and the device behavior. From the model and experimental results obtained on decananometric devices, we propose an understanding of the physical mechanisms involved in the CBRAM operations. Using the consistent Forming/SET and Data Retention model, we obtained good agreement with the experimental data. Finally, the impact of each layer of the CBRAM on the Forming/SET behavior is decorrelated, allowing an optimization of the performance. [ABSTRACT FROM AUTHOR]

Through ab initio calculations, we propose that the conductive filaments in Pt/HfO2/Pt resistive random access memories are due to HfOx suboxides, possibly tetragonal, where x\leq 1.5 . The electroforming process is initiated by a continuous supply of oxygen Frenkel defect pairs through an electrochemical process. The accumulation of oxygen vacancies leads to metallic suboxide phases, which remain conductive even as ultranarrow 1- {\rm nm}^{2} filaments embedded in an insulating {\rm HfO}_{2}$ matrix. Our experiments further show that the filaments remain as major leakage paths even in the OFF-state. Moreover, thermal heating may increase the OFF-state resistance, implying that there are oxygen interstitials left in the oxide layer, which may recombine with the oxygen vacancies in the filaments at high temperature. [ABSTRACT FROM PUBLISHER]

In this paper, we examine the problem of the drift of the low-resistance state (LRS) in phase change memories based on C or N doped and undoped Ge-rich Ge2Sb2Te5 . A novel procedure, named R–SET technique, is proposed to boost the SET speed of these innovative phase change materials by overcoming the decrease of crystallization speed caused by Ge enrichment. The R–SET technique allows, at the same time, an optimized SET programming of the memory cell and the reduction of the LRS drift with respect to standard SET procedures. A circuit that generates the desired R–SET pulse based on a time reference scheme is proposed and discussed. [ABSTRACT FROM PUBLISHER]

In this paper, we show performance and reliability improvement of Ag-GeS2-based conductive bridge RAM (CBRAM) devices by addition of a 2-nm-thick HfO2 layer between the electrolyte and the W bottom electrode. Our optimized dual-layer electrolyte stack (2-nm HfO2\--30\-nm~GeS2) leads to a resistance ratio (ROFF/RON) higher than 10^6 and projected 10 years read disturb immunity at 0.04 V. The improved memory resistance ratio is explained by means of physical modeling. Using compact modeling and circuit level simulations, we show that our optimized CBRAM device, integrated in a 1T-2R architecture, fits well with the aggressive requirements of field programmable gate array-type reconfigurable applications. Nonvolatility, back-end-of-line compatibility, and 1.3-nA leakage current during continuous reverse read operation at 1 V are strong benefits demonstrated on our device for such applications. [ABSTRACT FROM AUTHOR]

In this paper, we show that performances of Ge2Sb2Te5-based phase-change memory (PCM) cells can be improved by the insertion of a thin HfO2 interfacial layer between the phase-change material and the tungsten plug. Significant reduction of the RESET and SET currents and of the energy required to switch the PCM cells are demonstrated. In addition, compared with pure GST reference cells, the presence of the HfO2 layer leads to an enhanced endurance (>10^8~cycles) and an improved data-retention (ten years at 172^\circC). Using cross-sectional TEM-Energy Dispersive X-Ray (EDX) analyses and technology computer-aided design simulations, the decreases of the programming currents and energy are explained through the reduction of the PCM cell active area, due to the creation of conductive paths in the HfO2 layer during the initial forming procedure. Electrical measurements and modeling of the forming procedure indicate that the size of the conductive paths is controlled by the maximum current flowing through the PCM cells during the current overshoot of the forming procedure. [ABSTRACT FROM AUTHOR]

The physical mechanisms that regulate carrier transport in polycrystalline chalcogenides, such as \Ge2\Sb2\Te5 (GST), are still debated. Recently, self-induced Joule-heating (SJH) effect has been claimed to be the key factor in explaining the nonlinearity of the I–V characteristics of polycrystalline GST-based phase-change memory (PCM). In this paper, we carefully investigate the SJH occurring in the GST material by analyzing the I–V characteristics of PCM cells at low voltages, i.e., in the memory-cell readout region. To accomplish the study, we use ad hoc fabricated PCM devices allowing an easier evaluation of SJH occurring in the chalcogenide layer. A novel procedure to test the SJH effect is also proposed. A comparison between numerical simulations and compact modeling is discussed as well. Our paper shows that the SJH effect is not sufficient to reproduce the experimental I–V nonlinearity, claiming for new experiments and theoretical investigations. Therefore, this paper can be considered a step forward toward the comprehension of the transport properties of polycrystalline GST, which is a key aspect for robust modeling of PCM devices. [ABSTRACT FROM PUBLISHER]

Based on the material analysis of the SiN layers presented in part I of this paper, we develop accurate atomistic and electrical models for the silicon nitride (SiN)-based nonvolatile memory devices, taking into account the candidate SiN defects responsible for the memory effect. Our analysis points out the role of the hydrogen atoms and Si dangling bonds in the trapping properties of SiN films with different stoichiometries. The atomistic models provide a comprehensive picture describing the energy level and the occupation number of the different defects in the SiN. The electrical model coupled with the atomistic results, for the first time, demonstrates the ability to describe the program/erase curves of charge-trap memory cells with SiN storage layers with diversified composition. Good agreement between simulations and experimental results coming from the material analysis and the electrical characterization of thin (type-B device) and thick (type-A device) tunnel oxide memory cells is shown. [ABSTRACT FROM PUBLISHER]

In this paper, an extensive investigation of hybrid molecular/Si field-effect memories is presented, where redox ferrocene (Fc) molecules play the role of the memory charge storage nodes. Engineering of the organic linkers between Fc and Si is achieved by grafting Fc with different linker lengths. The study shows a clear correlation between results from atomistic computational density functional theory, electrochemical measurements (cyclic voltammetry) and electrical data obtained by a detailed study on capacitors and pseudo-MOS devices. Physical–chemical analyses (atomic force microscopy, high-resolution transmission electron microscopy, and X-ray photoelectron spectroscopy), corroborate the quality of molecular layers on devices. [ABSTRACT FROM PUBLISHER]

This letter studies the intrinsic variability in oxide-based resistive RAM technology, highlighting the presence of a short range ( $\approx 40$ ) correlation of resistances among cycles (for both low resistance state and high resistance state). Experimental results demonstrate the existence of a resistance correlation, and an analytical model is proposed in order to explain the findings. The presence of the correlation seems to indicate that the conductive filament, which is believed to be the basis of resistive RAM behavior, keeps for a limited number of cycling operations a memory of its morphology. The extension of this correlation depends on the programming conditions. [ABSTRACT FROM PUBLISHER]

Proposes an experimental and theoretical investigation of memory-cell structures employing discrete-trap type storage nodes, using either natural nitride traps or semiconductor nano-crystals. Background on the use of discrete-trap storage nodes in memory devices; Methodology; Results and discussion.

Abstract: In this work, cells based on TiN/TaOx/TiN metal–insulator–metal structure, fabricated in a 65nm complementary metal oxide semiconductor technology, are investigated in order to propose a good stack candidate for multiple time programmable (MTP) applications. Firstly, the oxide material thickness and stoichiometry are optimized to minimize the forming voltage and the RESET current. Programming energies of the devices under different current compliance during the SET operation and with different cell size are studied putting in evidence the filamentary mechanism. Very high speed and low voltage for the ON (SET) operation are demonstrated. High thermal stability up to 125°C is demonstrated in the SET/RESET states of the device (with an ROFF/RON ~100). Promising data retention exceeding 24h at 250°C is also shown. Finally, due to the stability of this device, the high potential of TaOx-based resistive random access memory for MTP applications is demonstrated. [Copyright &y& Elsevier]