Memristive Switching Research Articles

Mimicking somatic behavior of neurons using the integrate and fire model of 2D SnS memristive switching characteristics

This Letter presents the fabrication and characterization of a 2D SnS memristor, proposing its integrate and fire (I&F) model as a potential hardware implementation of neuronal somatic behavior. The memristor comprises a thin layer of tin (II) sulfide (SnS) sandwiched between copper (Cu) electrodes on a silicon (Si) substrate. This structure exhibits an impressive Roff:Ron ratio of 103 at a read voltage Vrd of 0.25 V with exceptionally low switching Vsw and set Vset voltages of 0.3 and 0.35 V, respectively, with ∼3 order variation between the maximum Rmax and Rmin resistances offered during single voltage sweep cycle. We have explained the memristive behavior using the dual ionic conduction mechanism in the SnS active layer. We extracted the real-time band diagram of SnS using ultraviolet photoelectron spectroscopy, explaining the low Vsw observed. We propose that the emulation of the I&F artificial neuron model exhibited by the fabricated device could serve as a promising application in the field of artificial neuron spiking.

A universal resist-assisted metal transfer method for 2D semiconductor contacts

Abstract With the explosive exploration of two-dimensional (2D) semiconductors for device applications, ensuring effective electrical contacts has become critical for optimizing device performance. Here, we demonstrate a universal resist-assisted metal transfer method for creating nearly free-standing metal electrodes on the SiO2/Si substrate, which can be easily transferred onto 2D semiconductors to form van der Waals (vdW) contacts. In this method, polymethyl methacrylate (PMMA) serves both as an electron resist for electrode patterning and as a sacrificial layer. Contacted with our transferred electrodes, MoS₂ exhibits tunable Schottky barrier heights and a transition from n-type dominated to ambipolar conduction with increasing metal work functions, while InSe shows pronounced ambipolarity. Additionally, using α-In2Se3 as an example, we demonstrate that our transferred electrodes enhance resistance switching in ferroelectric memristors. Finally, the universality of our method is evidenced by the successful transfer of various metals with different adhesion forces and complex patterns.

Scalable fabrication of vertically arranged Bi2Se3 crossbar arrays for memristive device applications

Despite the unique advantages of the memristive switching devices based on two-dimensional (2D) transition metal dichalcogenides, scalable growth technologies of such 2D materials and wafer-level fabrication remain challenging. In this work, we present the gold-assisted large-area physical vapor deposition (PVD) growth of Bi2Se3 features for the scalable fabrication of 2D-material-based crossbar arrays of memristor devices. This work indicates that gold layers, prepatterned by photolithography processes, can catalyze PVD growth of few-layer Bi2Se3 with 100-folds larger crystal grain size in comparison with that grown on bare Si/SiO2 substrates. We also present a fluid-guided growth strategy to improve growth selectivity of Bi2Se3 on Au layers. Through the experimental and computational analyses, we identify two key processing parameters, i.e., the distance between Bi2Se3 powder and the target substrate and the distance between the leading edges of the substrate and the substrate holder with a hollow interior, which plays a critical role in realizing large-scale growth. By optimizing these growth parameters, we have successfully demonstrated cm-scale highly-selective Bi2Se3 growth on crossbar-arrayed structures with an in-lab yield of 86%. The whole process is etch- and plasma-free, substantially minimizing the damage to the crystal structure and also preventing the formation of rough 2D-material surfaces. Furthermore, we also preliminarily demonstrated memristive devices, which exhibit reproducible resistance switching characteristics (over 50 cycles) and a retention time of up to 106 s. This work provides a useful guideline for the scalable fabrication of vertically arranged crossbar arrays of 2D-material-based memristive devices, which is critical to the implementation of such devices for practical neuromorphic applications.

Leveraging Tunability of Localized-Interfacial Memristors for Efficient Handling of Complex Neural Networks.

A scalable (<130 nm) resistive switching memristor that features both filamentary and interfacial switching aimed at neuromorphic computing is developed in this study. The typically perceived noise or volatility was effectively harnessed as a controlled mechanism for interfacial switching. The multilayer structure for the proposed memristor enhances switching stability by curbing ionic overmigration and mitigating leakage paths. Furthermore, the memristors showcased their reliability by demonstrating more than 15 M cycles in the filamentary mode and 1 M pulses in the interfacial mode. Additionally, retention tests at 85 °C for 104 s confirmed the stability across different states, affirming its reliability as a nonvolatile CMOS-compatible element. While many studies validate performance solely on the MNIST data set, this work also evaluates more complex data sets, demonstrating the robustness of the demonstrated memristor in supervised learning. Specifically, supervised learning simulations on MNIST and fashion MNIST data sets indicated a high learning rate with <4% deviations from numerical training, while offline inference trained on CIFAR-10 and CIFAR-100 data sets revealed <2.5% and <7% deviations caused by programing error accumulation, even with increased memristor counts for these highly complex data sets. Unsupervised learning via spike-timing-dependent plasticity further highlights the potential of the developed memristor in bridging artificial and biological paradigms, offering a significant advance toward efficient and biologically inspired computing architectures.

Laser assembly of CeO2 nanobrushes and their resistive switching performance

Memristors in a metal/oxide/metal configuration emerge as an advanced non-volatile information storage technology due to their high operating speed and low power consumption. Understanding the role of the microstructure of the active oxide layer in its resistive switching (RS) performance is scientifically important for revealing the conduction mechanism of oxides, and technically critical for the development of high-performance memristors. Here, self-assembly of vertically aligned CeO2 nanobrushes, a typical RS material, is demonstrated in pulsed laser epitaxy. The growth map of CeO2 is fully explored, and optimized growth conditions for CeO2 nanobrushes are established. Microstructure and crystallinity characterizations reveal the single-crystalline epitaxy nature of CeO2 nanobrushes. A nanobrush memristor exhibits a threshold switching feature with an On/Off ratio of 50, as 16 times high as a thin-film memristor, and excellent endurance of up to 500 cycles and retention time of up to 104 seconds. Theoretical fitting of the I-V curves of the thin-film and nanobrush memristors show interfacial switching in the thin-film memristor and filamentary switching in the nanobrush memristor. The distinct RS mechanisms between thin films and nanobrushes suggest the fundamental role of the structural design of active oxide layers in the RS performance of oxide-based memristors.

A multilevel resistive switching memristor based on flexible organic–inorganic hybrid film with recognition function

Abstract Brain-inspired neuromorphic computing systems fueled the emergence of memristor-based artificial synapses, however, conventional silicon-based devices restricted their usage in the wearable field because of their difficulty in bending. To tackle the above challenge, a vertically structured flexible memristor with aluminum-based hydroquinone organic–inorganic hybrid film and Al2O3 as the functional layer, ITO and Pt as the bottom and top electrodes, and PET as the substrate has been developed utilizing molecular/atomic layer deposition to achieve a tradeoff between the resistive transition properties and the flexibility of memristors. The obtained devices combine stable resistive switching behavior and flexibility, showing high switching ratio of 103, better retention (up to 105 s) and endurance properties (up to 104 cycles), and robustness at radius of curvature of 4.5 mm after 104 bending cycles. Furthermore, the presence of multilevel resistive states in these devices ensures that the memristor can emulate synaptic properties such as paired-pulse facilitation, transition from short-term plasticity to long-term plasticity, long-term potentiation and depression, and spike-time-dependent plasticity. The resistive switching mechanism and the role of the bending state on the electrical performance of the device are explored. The fully connected artificial neural network based on the memristor can achieve a recognition accuracy of 90.2% for handwritten digits after training and learning. Flexible memristor will bring feasible advances to the integration of neuromorphic computing and wearable functionality.

High-speed Ta2O5-based threshold switching memristor for LIF neurons

Due to their high similarity to biological ion channels, low power consumption, small footprint, and the fact that they do not require reset circuits, threshold switching memristors have been intensively studied for simulating neurons in neuromorphic chips. Switching speed is one of the key challenges which limit the application of threshold switching memristors in chips. In this study, Ta2O5 threshold switching memristors with high switching speeds were prepared by doping with silver. The results show that 14 wt. % Ag doped Ta2O5 threshold switching memristors exhibit excellent bi-directional threshold switching performance, featuring fast switching speeds (&amp;lt;20 ns, &amp;lt;18 ns), low leakage currents (&amp;lt;10 pA), and high switching ratio (&amp;gt;107). According to the field nucleation theory, the rapid switching speed can be attributed to the low nucleation energy (0.26 eV) of silver within the Ta2O5 matrix, which is achieved by incorporating 14 wt. % Ag during the doping process. Based on Pspice, a LIF (leaky integrate-and-fire) neuron based on the silver nanoparticles doped Ta2O5 threshold switching memristors is built, and its firing function has been simulated. The results show that the LIF neuron with a short switching time is able to excite pulse spiking with high frequencies. These results demonstrated that the silver nanoparticles doped Ta2O5-based threshold switching memristors hold significant potential for constructing high-speed artificial neural networks.

Controlling the digital-to-analog switching in HfO2-based memristors via modulating the oxide thickness

The HfO2-based memristors have been acknowledged as a highly potential candidate for nonvolatile memory and neuromorphic computing applications. However, the analog resistive switching behavior of the HfO2-based memristors still needs improvement. In this work, Cu/HfO2/Pd devices with different oxide layer thicknesses were prepared by the magnetron sputtering process. The resistance switching characteristics related to oxide layer thickness were explored, revealing significant differences in the forming, SET, and RESET characteristics depending on oxide thickness. The conduction mechanism of the Cu/HfO2/Pd memristors is mainly attributed to the space charge limited current (SCLC) model. Thinner devices tend to show a more abrupt SET behavior and larger SET voltage, while thicker devices exhibit more gradual SET behavior and a lower SET voltage. In addition, as the thickness of the oxide layer increases, the Cu/HfO2/Pd memristors transition from digital resistive switching to analog resistive switching. The formation and rupture of oxygen vacancy filaments are predominant in thicker device, and the shape of the conduction filaments has been proposed as responsible for the abrupt and gradual changes in resistive switching. This work demonstrates that the thickness of the oxide layer plays an important role in engineering digital logic and neuromorphic behavior in HfO2-based memristors.

Investigation of the TaOx unipolar switching memory on high efficiency computing

Memristor has shown great potential application for data storage, neuromorphic computing chip, and memory logic, but its bi-directional operation from setting in positive bias voltage region to resetting in a negative bias voltage region would introduce a complex control circuit. Herein, a memristor with the structure of Au/TaOx/F-doped SnO2 presents unipolar switching behavior, enabling the memristor to operate the set and reset processing in a single voltage direction. By comparison with bipolar switching memristor, using this unipolar switching memristor to operate the logic expression of the letter “D” based on American standard code for information interchange (ASCII) can reduce power consumption by 55.56% while switching the logic state from “0” to “1” and then back to “0” can save up to 48% of power consumption. This work provides a feasible method for the low power consumption computing as well as a road to simplify circuit design.

Versatile NbOx‐Based Volatile Memristor for Artificial Intelligent Applications

AbstractTo achieve cost‐effectiveness, researchers are exploring various memristors for their adaptation in neuromorphic computing. Recent studies have focused on developing versatile functioning singular memristors, such as those involved in on‐receptor computing, which integrates sensory functions into current neuromorphic computing paradigms. Additionally, adaptations like reservoir computing are being investigated for computing systems. In this study, a memristor composed of a stack of Ti/NbOx/Pt layers is fabricated to explore multifunctional behaviors within a single memristor. By applying bias toward the top Ti electrode, gradual current changes with volatile features are demonstrated, revealing an ion‐migration‐based nonfilamentary switching memristor. Leveraging this volatile functionality, an artificial nociceptor is first implemented, demonstrating key functions of biological nociceptors including thresholding, relaxation, no‐adaptation, and sensitization. Subsequently, synapse emulation akin to the biological brain is achieved through easy conductance potentiation and depression with diverse synapse functions, enabling the memristor to mimic learning activities with spike firing. Lastly, computational applications are explored by adapting edge computing and multi‐bit reservoir computing, expanding the memristor's applications across diverse fields with versatile behaviors.

Invited: Mimicking Biological Synapse Activity with LDH-Based Memristor

Memristors, mimicking biological synapse response, are believed to have a great potential for use as an adaptive memory unit in future neuromorphic computing systems. Here we report on fabrication of two-dimensional memristor comprised of layered double hydroxide (LDH) material, covered with graphene layer, serving as a terminal for connecting to electrodes. We demonstrated nonvolatile resistive switching of LDH memristor. Device performance is studied upon replacing graphene semimetal with alternative 2D materials. Comprehensive characterization of the memristor materials was performed with correlated Raman, Electron, and Scanning Probe Microscopy methods.ACKNOWLEDGMENTThis work was supported by the US Department of Defense (Contract #W911QY2220006), 2DCC-MIP NSF cooperative agreement DMR-2039351. Work was performed at JSNN, a member of the Southeastern Nanotechnology Infrastructure Corridor (SENIC) and National Nanotechnology Coordinated Infrastructure (NNCI), which is supported by the NSF (ECCS-1542174)

A Temperature Sensory Leaky Integrate-and-Fire Artificial Neuron Based on Chitosan/PNIPAM Bilayer Volatile Complementary Resistive Switching Memristor.

The presence of neurons is crucial in neuromorphic computing systems as they play a vital role in modulating the strength of synapses through the release of either excitatory or inhibitory stimuli. Hence, the development of sensory neurons plays a pivotal role in broadening the scope of brain-inspired neural computing. The present study introduces an artificial sensory neuron, which is constructed using a temperature-sensitive volatile complementary resistance switch memristor based on the functional layer of the chitosan/PNIPAM bilayer. The resistive switching behavior arises from the formation and ionization of oxygen vacancy filaments, whereby the threshold voltage and low resistive resistance of the device exhibit a temperature-dependent increase within the range of 290-410 K. A functional replication of a neuron with leaky integration and firing has been successfully developed, effectively simulating essential biological functions such as firing triggered by threshold, refractory period implementation, and modulation of spiking frequency. The artificial sensory neuron exhibits characteristics similar to those of leaky integrated firing neurons that receive temperature inputs. It has the potential to control the output frequency and amplitude under varying temperature conditions, making it suitable for temperature-sensing applications. This study presents a potential hardware implementation for developing efficient artificial intelligence systems that can support temperature detections.

Resistive switching of IGZO-based memristors with electrode modulation and multi-value storage application

Resistive switching of IGZO-based memristors with electrode modulation and multi-value storage application

A zinc oxide-based threshold switching memristor for simulating synaptic plasticity and artificial nociceptor

A zinc oxide-based threshold switching memristor for simulating synaptic plasticity and artificial nociceptor

Cyclopentadithiophene-based conjugated polymer artificial synapses for dual information encryption

High-performance polymer memristors with analog-type switching behaviors and low power consumption, which is a perfect candidate for brainoid intelligence, have triggered the research campaign of the novel in-memory computing paradigm. A novel poly[cyclopentadithiophene- alt-thiophene] with triazole and ferrocene moieties in the sidechains (PCTF) is synthesized by copper-catalyzed azide-alkyne Click chemistry reaction of poly[4,4-bis(6-azidohexyl)-4H- cyclopenta(1,2-b: 5,4-b′) dithiophene-alt-thiophene], which is prepared by the reaction of poly[thiophene-alt-4,4- bis(6-bromohexyl)- 4H-cyclopenta(1,2-b:5,4-b′)dithiophene](PTT-Br) with NaN3, with ethynylferrocene. The PTT-Br film sandwiched between the Al and ITO electrodes shows a typical history-dependent memristive switching performance at a small sweep voltage range of ± 1 V, with 6 distinguishable conductance states. The device current can be modulated continuously when being subjected to consecutive positive and negative voltage sweeps. In contrast to PTT-Br, PCTF exhibits excellent nonvolatile memristive performance, with 16 distinguishable conductance states due to the combined action of charge transfer between the triazole and ferrocene moieties, and redox activity from both the ferrocene and polymer backbone. As an important component related to human memory, the short-term synaptic plasticity to long-term potentiation/plasticity transition is, for the first time, successfully used to realize dual information encryption at a high information security level. Only by inputting a given number of voltage pulses and at a specific time can classified information be correctly read out.

Capacitive and Inductive Characteristics of Volatile Perovskite Resistive Switching Devices with Analog Memory.

With the increasing demands and complexity of the neuromorphic computing schemes utilizing highly efficient analog resistive switching devices, understanding the apparent capacitive and inductive effects in device operation is of paramount importance. Here, we present a systematic array of characterization methods that unravel two distinct voltage-dependent regimes demonstrating the complex interplay between the dynamic capacitive and inductive effects in volatile perovskite-based memristors: (1) a low voltage capacitance-dominant and (2) an inductance-dominant regime evidenced by the highly correlated hysteresis type with nonzero crossing, the impedance responses, and the transient current characteristics. These dynamic capacitance- and inductance-dominant regimes provide fundamental insight into the resistive switching of memristors governing the synaptic depression and potentiation functions, respectively. More importantly, the pulse width-dependent and long-term transient current measurements further demonstrate a dynamic transition from a fast capacitive to a slow inductive response, allowing for the tailored stimulus programming of memristor devices to mimic synaptic functionality.

Memristor Effect in Layered Film Structures

Equivalent electrical circuits of multilayer film structures with memristor switching of resistance at interlayer boundaries and at the boundaries of crystal grains in each layer are proposed. Numerical modeling of the current-voltage characteristics of such structures has shown that their loop-shaped form, typical of memristors, is transformed into a linear ohmic dependence of the total current on the magnitude of the applied external voltage as both the number of layers and the number of grains in each layer increase. A certain combination of the number of layers and grains in a layer has been established, at which the maximum total current flowing through the structure and the ratio of resistances in the “off” and “on” states reach the highest values.

Compliance-free, analog RRAM devices based on SnOx

Brain-inspired resistive random-access memory (RRAM) technology is anticipated to outperform conventional flash memory technology due to its performance, high aerial density, low power consumption, and cost. For RRAM devices, metal oxides are exceedingly investigated as resistive switching (RS) materials. Among different oxides, tin oxide (SnOx) received minimal attention, although it possesses excellent electronic properties. Herein, we demonstrate compliance-free, analog resistive switching behavior with several stable states in Ti/Pt/SnOx/Pt RRAM devices. The compliance-free nature might be due to the high internal resistance of SnOx films. The resistance of the films was modulated by varying Ar/O2 ratio during the sputtering process. The I–V characteristics revealed a well-expressed high resistance state (HRS) and low resistance states (LRS) with bipolar memristive switching mechanism. By varying the pulse amplitude and width, different resistance states have been achieved, indicating the analog switching characteristics of the device. Furthermore, the devices show excellent retention for eleven states over 1000 s with an endurance of > 100 cycles.

Low-power perovskite-based threshold switching memristor for artificial nociceptor

Organic-inorganic halide perovskites (OHPs) with good ion diffusion paths are effective functional materials for simulating biological neurons. In this work, we utilize the Ag diffusion threshold switching phenomenon in FAPbI3 films to investigate the nociceptor function. The threshold switching device exhibits a low threshold voltage (0.075 V) and a large characteristic switching ratio of 107 in volatile resistance switching, along with a minimum switching slope (SS< 2.6 mV/dec), and the device achieves low power consumption of 150 pW. Importantly, the nociceptor functions of threshold firing and sensitization have been successfully simulated based on the volatile threshold switching characteristics, which demonstrates the potential in practical neuromorphic applications.

Memristive switching of nanofluidic diodes by ionic concentration gradients

We demonstrate that nanofluidic diodes exhibit a memristive behavior that can be controlled not only by the amplitude and frequency of the driving oscillatory signal but also by changing the relative orientations of the externally imposed ionic concentration difference with respect to the pore charge density gradient. Because ionic concentration differences, memristive properties, and nanofluidic pores are of broad interest to electrochemical technology and membrane bioelectrochemistry, this study can be a useful contribution to the field. In our case, the memristor is a polymer membrane with current rectifying conical nanopores. The electrical interaction between the mobile ions in the aqueous solution and the pore surface charges leads to rich physico-chemical phenomena, including memory effects and the neuromorphic-like potentiation of the membrane conductance following voltage pulses (spikes). The multipore membrane can be used in the design of electrochemical circuits for signal conversion and information processing in iontronic hybrid devices.