Phase Change Memory Research Articles

A Survey on Optical Phase-Change Memory: The Promise and Challenges

Silicon photonics (SiPh) technology has facilitated the deployment of integrated photonics across different application domains, from ultra-fast communication in Datacom applications to energy-efficient optical computation in emerging hardware accelerators for machine learning. More recently, the integration of SiPh and phase change materials has created a unique opportunity to realize adaptable, reconfigurable, and programmable photonic platforms. In particular, the nonvolatile programmability in phase change materials has made them a promising candidate for implementing photonic memory cells and architectures. Accordingly, photonic memory systems and even in-memory photonic computing paradigms are on the rise, especially given their potential for improving data access in electronic and photonic processors. However, there are still many challenges in the design and fabrication of phase-change photonic integrated circuits, which need to be addressed. This article presents a comprehensive survey on the recent advances and challenges for the integration of phase change materials with contemporary photonic devices while focusing on the photonic memory application. In particular, we explore phase-change photonic memory from the material level to the architecture level by presenting an overview of different material-level characteristics of phase change materials with their optical, electrical, and thermal properties as well as their integration into SiPh devices and photonic memory architectures and their application for in-memory photonic computing. We also present a comparison with electronic memory and discuss open research challenges that must be addressed to further advance phase-change photonic memory towards successful integration into emerging computing systems.

Hybrid CMOS-PCM Ternary Logic for Digital Circuit Applications

This paper presents hybrid design of a Ternary Inverter (TI) circuit by integrating an Ovonic-Threshold-Switching (OTS) based Phase-Change-Memory (PCM) cell between Complementary-Metal-Oxide-Semiconductor (CMOS) transistors. Volatile OTS behaviour of the PCM structure helps to generate the ternary state. Reliability of the proposed TI design has been demonstrated through the Monte-Carlo simulations in Cadence-Spectre simulator incorporating both Process and Mismatch variability at all corners. The proposed inverter cell consumes 10.44 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mu$</tex-math></inline-formula> W power and 1.47 ns delay on UMC-180 nm technology node. Moreover to demonstrate the scaling potential, we have simulated the proposed inverter at various technology nodes i.e 180/90/65/28 nm. At advanced feature sizes, power and delay are improved by significant amount. For benchmarking, we have compared the proposed TI design with various state-of-the-art hybrid structures. Furthermore, to demonstrate potential applications, we designed and simulated ternary universal logic gates i.e NAND/NOR. Benefits of the proposed TI in circuit applications are shown through a Ternary Multiplier (TMUL) arithmetic circuit implementation. Finally, a Ternary-Static-Random-Access-Memory (TSRAM) is proposed with two back-to-back connected CMOS-PCM TIs with three bit storing capability.

Design of all-phase-change-memory spiking neural network enabled by Ge-Ga-Sb compound

The implementation of artificial spiking neural network (SNN) usually takes advantage of multiple heterogeneous circuits to mimic either neurons which generate spiking pulses, or synapses which store the weights of event correlations. Here, we design a homogeneous device using Ge-Ga-Sb (GGS) as a phase-change-memory (PCM) material which can do both jobs. The GGS compound shows high stability when used in data storage, such as high working temperature (281°C) and high 10-years data retention temperature (230°C), as well as low resistance drift. Interestingly, when the as-fabricated GGS device is set by iterative narrow-width electric pulses, it first experiences an abrupt resistance drop by two orders of magnitude, followed by a continuous resistance decrease. This unique abrupt-to-progressive transition can be used to mimic both neuronal and synaptic functions, mechanistically enabled by the formation of conductive channels and the continuous growth with the phase separation of crystalline areas. To this end, we propose an all-PCM SNN, which is emulated to have high accuracy (90%) in the standard pattern recognition.

Формирование периодических двухфазных структур на поверхности аморфных пленок Ge-=SUB=-2-=/SUB=-Sb-=SUB=-2-=/SUB=-Te-=SUB=-5-=/SUB=- при воздействии ультракоротких лазерных импульсов различной длительности и частоты следования

Phase change memory materials due to their high susceptibility to low-intensity light fields are extremely attractive for active microphotonics and integrated optics devices. As a result of fast phase switching, these materials change the refractive index in a wide spectral range, which has found application in information storage systems. In this work, we studied the formation of two-phase periodic structures consisting of alternating lines of amorphous and crystalline phases on the surface of thin-film Ge2Sb2Te5 phase-change memory materials exposed to ultrashort laser pulses. Periodic structures were formed at a wavelength of 1030 nm at different durations and repetition rates of light pulses. It has been established that the ordering of two-phase structures obtained at a constant energy fluence remains practically unchanged with an increase in the repetition rate from 10 kHz to 1 MHz, but a change in the pulse duration from 180 fs to 10 ps leads to a violation of the periodic structure due to the formation of extended continuously crystallized regions.

Recent advances in doped Ge2Sb2Te5 thin film based phase change memories

Nowadays, chalcogenide-based phase change materials (PCMs) are proving themselves superior in the category of memory devices due to a combination of their unique set of properties.

Ultrahigh overall-performance phase-change memory by yttrium dragging

Benefiting from the dragging effect of yttrium, an ultrahigh overall-performance phase-change memory is reported, including low resistance drift, high data retention, low power consumption, fast operating speed, and good cycling endurance.

Surface effects on the crystallization kinetics of amorphous antimony.

Elemental antimony (Sb) is regarded as a promising candidate to improve the programming consistency and cycling endurance of phase-change memory and neuro-inspired computing devices. Although bulk amorphous Sb crystallizes spontaneously, the stability of the amorphous form can be greatly increased by reducing the thickness of thin films down to several nanometers, either with or without capping layers. Computational and experimental studies have explained the depressed crystallization kinetics caused by capping and interfacial confinement; however, it is unclear why amorphous Sb thin films remain stable even in the absence of capping layers. In this work, we carry out thorough ab initio molecular dynamics (AIMD) simulations to investigate the effects of free surfaces on the crystallization kinetics of amorphous Sb. We reveal a stark contrast in the crystallization behavior between bulk and surface models at 450 K, which stems from deviations from the bulk structural features in the regions approaching the surfaces. The presence of free surfaces intrinsically tends to create a sub-nanometer region where crystallization is suppressed, which impedes the incubation process and thus constrains the nucleation in two dimensions, stabilizing the amorphous phase in thin-film Sb-based memory devices.

Phase-change control of anomalous Hall effect in ferromagnetic MnBi thin films

We have experimentally demonstrated a phase change control of Hall effects in ferromagnetic MnBi thin films at room temperature. Two distinct ferromagnetic phases, i.e., a low-temperature phase and a quenched high-temperature phase, were obtained by slow-cooling or quenching the sputtered MnBi films. The measurement of Hall effects showed that the anomalous Hall resistivity has different coercivity and remanence between the two phases. By optimizing the Mn/Bi ratio, it is even possible to repeatedly switch the sign of Hall resistivity, which could be utilized as a concept of phase-change memory based on ferromagnetic transport.

Architecting Optically Controlled Phase Change Memory

Phase Change Memory (PCM) is an attractive candidate for main memory, as it offers non-volatility and zero leakage power while providing higher cell densities, longer data retention time, and higher capacity scaling compared to DRAM. In PCM, data is stored in the crystalline or amorphous state of the phase change material. The typical electrically controlled PCM (EPCM), however, suffers from longer write latency and higher write energy compared to DRAM and limited multi-level cell (MLC) capacities. These challenges limit the performance of data-intensive applications running on computing systems with EPCMs. Recently, researchers demonstrated optically controlled PCM (OPCM) cells with support for 5 bits / cell in contrast to 2 bits / cell in EPCM. These OPCM cells can be accessed directly with optical signals that are multiplexed in high-bandwidth-density silicon-photonic links. The higher MLC capacity in OPCM and the direct cell access using optical signals enable an increased read/write throughput and lower energy per access than EPCM. However, due to the direct cell access using optical signals, OPCM systems cannot be designed using conventional memory architecture. We need a complete redesign of the memory architecture that is tailored to the properties of OPCM technology. This article presents the design of a unified network and main memory system called COSMOS that combines OPCM and silicon-photonic links to achieve high memory throughput. COSMOS is composed of a hierarchical multi-banked OPCM array with novel read and write access protocols. COSMOS uses an Electrical-Optical-Electrical (E-O-E) control unit to map standard DRAM read/write commands (sent in electrical domain) from the memory controller on to optical signals that access the OPCM cells. Our evaluation of a 2.5D-integrated system containing a processor and COSMOS demonstrates 2.14 × average speedup across graph and HPC workloads compared to an EPCM system. COSMOS consumes 3.8× lower read energy-per-bit and 5.97× lower write energy-per-bit compared to EPCM. COSMOS is the first non-volatile memory that provides comparable performance and energy consumption as DDR5 in addition to increased bit density, higher area efficiency, and improved scalability.

Organic Phase-Change Memory Transistor Based on an Organic Semiconductor with Reversible Molecular Conformation Transition.

Phase-change semiconductor is one of the best candidates for designing nonvolatile memory, but it has never been realized in organic semiconductors until now. Here, a phase-changeable and high-mobility organic semiconductor (3,6-DATT) is first synthesized. Benefiting from the introduction of electrostatic hydrogen bond (S···H), the molecular conformation of 3,6-DATT crystals can be reversibly modulated by the electric field and ultraviolet irradiation. Through experimental and theoretical verification, the tiny difference in molecular conformation leads to crystalline polymorphisms and dramatically distinct charge transport properties, based on which a high-performance organic phase-change memory transistor (OPCMT) is constructed. The OPCMT exhibits a quick programming/erasing rate (about 3 s), long retention time (more than 2h), and large memory window (i.e., large threshold voltage shift over 30V). This work presents a new molecule design concept for organic semiconductors with reversible molecular conformation transition and opens a novel avenue for memory devices and other functional applications.

High-density analog image storage in an analog-valued non-volatile memory array

Data stored in the cloud or on mobile devices reside in physical memory systems with finite sizes. Today, huge amounts of analog data, e.g. images and videos, are first digitalized and then compression algorithms (e.g. the JPEG standard) are employed to minimize the amount of physical storage required. Emerging non-volatile-memory technologies (e.g. phase change memory (PCM), resistive RAM (RRAM)) provide the possibility to store the analog information in a compressed format directly into analog memory systems. Here, we demonstrate with hardware experiments an image storage and compression scheme (joint source-channel coding) with analog-valued PCM and RRAM arrays. This scheme stores information in a distributed fashion and shows resilience to the PCM and RRAM device technology non-idealities, including defective cells, device variability, resistance drift, and relaxation.

Physics-Based Modeling Strategies of Phase-Change Random Access Memory

Phase-change random access memory (PCRAM) has been regarded as one of the most promising candidates for future “universal” memory due to its mature fabrication techniques and profitable commercialized prospect. To further improve its physical performance and exploit more application areas, it is imperative to develop a physically realistic model that can closely mimic the electrical, thermal, and phase-change mechanisms of PCRAM. Although plenty of phase-change memory models have been proposed during the last two decades, they were constructed from various physical principles and, thus, targeted for different practical applications. As a result, a comprehensive review to elucidate the state-of-the-art modeling strategies of PCRAMs based on their phase-change mechanisms and application targets remains mysterious. To address this issue, we here reviewed the reported modeling approaches of PCRAMs mainly in terms of the finite-element method, the physics-based analytical model, and compact modeling techniques. For each technique, its adopted strategies to simulate the electrothermal and phase-change processes of phase-change materials at the device level were elaborated, respectively, and their respective merits and drawbacks were also given. The application fields, for which these models suit, were eventually discussed.

Modeling the Gradual RESET of Phase Change Memory With Confined Geometry

This work develops a model of the temperature and resistance for the RESET of confined phase change memory (PCM). The confined PCM, unlike the mushroom type, has heating inside the cell and higher resistance sensitivity to active region size. The ideal temperature profile is derived from a steady-state heat conduction formulation based on a quasi-3D approach. A thermal subcircuit is then proposed to implement the heating module, considering the thermoelectric effect (TEE) and temperature dependence of phase change materials, to obtain the active region size. The resistance of the irregular active region is then calculated using an integral method. Calibrated TCAD simulations and experimental data are used to verify the model with desired scalability and accuracy.

Xbar-Partitioning: A Practical Way for Parasitics and Noise Tolerance in Analog IMC Circuits

Conventional in-memory computing (IMC) architectures consist of analog memristive crossbars to accelerate matrix-vector multiplication (MVM), and digital functional units to realize nonlinear vector (NLV) operations in deep neural networks (DNNs). These designs, however, require energy-hungry signal conversion units which can dissipate more than 95% of the total power of the system. Fully-analog IMC circuits remove the need for signal converters by realizing both MVM and NLV operations in the analog domain leading to significant energy savings. However, they are more susceptible to errors caused by interconnect parasitic and noise. Here, we propose Xbar-partitioning, a practical approach to divide large IMC arrays into multiple partitions to alleviate the impacts of noise and parasitics while keeping the computation in the analog domain. The SPICE circuit simulation results for the deployment of a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$400\times 120\times 84\times 10$ </tex-math></inline-formula> DNN model on various fully-analog IMC architectures with five different 2-terminal and 3-terminal resistive devices, six different bitcell layouts, and four different partitioning schemes show that the highest accuracy of 98.08% can be obtained for a design using phase-change memory (PCM) devices and 1T-1R bitcell with 13, 4, and 3 horizontal partitions, and 4, 3, and 1 vertical partition for the first, second, and third layers of the DNN, respectively. Finally, we provide a signal-to-noise ratio (SNR) analysis which shows that the IMC architectures can be made more noise-tolerant by using smaller-sized bitcells, higher partitions in the crossbar, and device technologies with higher <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$R_{off}/R_{on}$ </tex-math></inline-formula> ratio.

Nitrogen tuned crystal structure, optical band gap, localized states on amorphous and crystalline GeTe films

The nitrogen (N) doping effect on crystal structure, optical band gap, and localized states for amorphous/crystalline GeTe films was studied. The crystallization temperature and sheet resistance were improved by N doping. The local bonding structure of amorphous and crystalline N x ( GeTe ) 1 - x was elucidated by Raman spectra. The changes in optical band gap and localized states density for the films were evaluated by transmittance spectra. The band gap broadening after N doping is crucial to reducing the threshold current for phase change memory.

Dielectric functions evolution and electronic bandgap manipulation by silicon doping for Sb2Te3 phase change films: Temperature dependent spectroscopic ellipsometry study

The dielectric function evolution and electronic bandgap manipulation by silicon (Si) doping for Sb2Te3 phase change films have been investigated by temperature dependent spectroscopic ellipsometry measurements. During the phase change from amorphous to rhombohedral structures, the significant contrast of electronic band structure for pure Sb2Te3 and Si-doped Sb2Te3 (SST) films as functions of temperature (210–620 K) and Si concentration (0%–12%) has been systematically studied by analyzing the dielectric functions, Tauc gap energy, and partial spectral weight integral. The distinct differences can be mainly attributed to the increment of structure order degree, originated from the change of local bonding arrangement. Based upon the evolutions of Tauc gap energy and partial spectral weight integral with increasing temperature for all four samples, it can be concluded that Si doping can inhibit the crystallization of amorphous films and accelerate the phase change process by serving as nanoscale heaters, which is helpful in improving the thermal stability of amorphous films. The elevated crystallization temperature and phase change rate by Si doping contribute to the dependability and endurance for SST-based phase change memory. The present data provide an important direction on the physical mechanism investigation of Si doping Sb2Te3 by optical techniques.

Coupling a phase field model with an electro-thermal solver to simulate PCM intermediate resistance states for neuromorphic computing

A key attribute for Phase-Change Memories to be used for neuromorphic computing is the possibility to create intermediate resistance states by the progressive crystallization during the successive application of set pulses. To complete an experimental approach, we studied progressive crystallization by simulation. Insight on the real mechanisms of progressive crystallization is obtained by coupling Joule heating, diffusion of heat and a phase change model relying on the Phase Field Method.

Simultaneously introducing tetrahedral and pseudo-octahedron by single-element doping enables faster and more stable phase change memory

Phase change memory (PCM) is one of the most promising candidates for the next generation non-volatile memory, but how to balance the contradiction between thermal stability and speed remains the bottleneck problem that limit its application as a universal memory. Here, we innovatively introducing the tetrahedral cluster and the pseudo-octahedron composed of fourfold abab rings at the same time into Sb2Te3 amorphous matrix by elemental Cu doping. The increased Cu tetrahedral clusters bring about enhanced amorphous stability in Sb2Te3 and raise its crystallization temperature from 153.3 ℃ to 203.7 ℃. Besides, the reduced number of the lone-pair electrons, the stronger bond strength and the greater Peierls-like distortions also play crucial roles. Simultaneously, crystallization process simulation results show the fourfold abab ring structure, as a crystalline precursor, accelerates the crystallization speed remarkably. Experimentally, 14 ns SET speed is achieved in Cu0.2(Sb2Te3)0.8 based PCM device, which is four times faster than in pure Sb2Te3 device. Therefore, by a simple single-element doping process, both amorphous stability and crystallization speed are improved at the same time.

Spectroscopic study of nitrogen incorporation in Ge, Sb, and Te elemental systems: A step toward the understanding of nitrogen effect in phase-change materials

Nitrogen doping in chalcogenide materials represents a promising way for the improvement of material properties. Indeed, N doping in GeSbTe phase-change alloys have demonstrated to greatly enhance thermal stability of their amorphous phase, necessary to ensure the data retention of the final phase-change memory device. Although it is suggested that the N doping in such alloys leads to the preferential formation of Ge-N bonds, further questions concerning the bonding, in particular, Sb-N and Te-N, and the structural arrangement remain unclear. In this paper, we present a study of as-deposited elemental Ge, Sb, and Te systems and their nitrides (i.e., GeN, SbN, and TeN alloys), using a large range of N content from 0 up to about 50 at. %. The as-deposited alloys are investigated by Fourier transform infrared and Raman spectroscopy. We identify the active vibrational modes associated with the formation of Ge-N, Sb-N, and Te-N bonds, highlighting the impact of N incorporation on the structure of these elemental systems. We further qualitatively compare the GeN, SbN, and TeN experimental spectra with the “ab initio” simulations of the related ideal nitride structures. Finally, the analysis of elemental nitride layers is extended to N-doped GeSbTe alloys, providing deeper understanding of nitrogen bonding in such ternary systems, employed in memory technology.

Joule heating induced non-melting phase transition and multi-level conductance in MoTe2 based phase change memory

Phase change memory (PCM) is considered as a leading candidate for next generation data storage as well as emerging computing device, but the advancement has been hampered by high switching energy due to the melting process and amorphous relaxation induced large resistance drift. Polymorphic crystal-crystal transition without amorphization in metal dichalcogenides (TMDs) could be employed to solve these issues. Yet, the mechanism is still controversy. A melting-free PCM made of two dimensional (2D) MoTe2, which exhibits unipolar resistive switching (RS) and multi-level states with substantially reduced resistance drift via joule heating, is reported in this work. The device is first prepared based on the temperature dependence of Raman spectrum and electrical transport investigations on MoTe2 films. Significantly improved device performances on energy efficiency, switching speed, and memory window are further achieved by electrode size scaling down, indicating the key role of localized heating. Then, device scale transmission electron microscopy images reveal that the resistive switching stems from the transition between semiconducting 2H phase and metallic 1T′ phase. An entropy induced Te vacancies model is proposed to explain the reversible phase change mechanism in the MoTe2 based device. This study paves the way for further development of PCM based on atomically thin 2D TMDs, aiming for high density storage-class memory and high-precision neuromorphic computing.