Designing carbon conductive filament memristor devices for memory and electronic synapse applications

Zhou, Zhenyu; Zhao, Jianhui; Chen, Andy Paul; Pei, Yifei; Xiao, Zuoao; Gong, Wenyu; Chen, Jingsheng; Fu, Guangsheng

doi:10.1039/c9mh01684h

Cited by 70 publications

(56 citation statements)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yan et al have recently reported that electron hopping between sulfur and metal vacancies is the reason for ultra-fast memristive behavior in two-dimensional metal chalcogenide films. 16 , 23 , 52 , 53 The charge hopping is voltage-dependent, which leads to voltage-dependent PPF as observed in Fig. 3 c. We believe that the same is the underlying mechanism behind observing the PPF/PPD in MoS 2 QD systems as well, and has been discussed in detail later on in this work.…”

Section: Resultsmentioning

confidence: 53%

See 1 more Smart Citation

Programmable electronic synapse and nonvolatile resistive switches using MoS2 quantum dots

Thomas

Resmi

Ganguly

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Brain-inspired computation that mimics the coordinated functioning of neural networks through multitudes of synaptic connections is deemed to be the future of computation to overcome the classical von Neumann bottleneck. The future artificial intelligence circuits require scalable electronic synapse (e-synapses) with very high bit densities and operational speeds. in this respect, nanostructures of two-dimensional materials serve the purpose and offer the scalability of the devices in lateral and vertical dimensions. in this work, we report the nonvolatile bipolar resistive switching and neuromorphic behavior of molybdenum disulfide (MoS 2) quantum dots (QD) synthesized using liquid-phase exfoliation method. The ReRAM devices exhibit good resistive switching with an On-Off ratio of 10 4 , with excellent endurance and data retention at a smaller read voltage as compared to the existing MoS 2 based memory devices. Besides, we have demonstrated the e-synapse based on MoS 2 QD. Similar to our biological synapse, paired pulse facilitation / Depression of short-term memory has been observed in these MoS 2 QD based e-synapse devices. this work suggests that MoS 2 QD has potential applications in ultra-high-density storage as well as artificial intelligence circuitry in a costeffective way.

show abstract

Section: Resultsmentioning

confidence: 53%

Section: Resultsmentioning

confidence: 90%

Programmable electronic synapse and nonvolatile resistive switches using MoS2 quantum dots

Thomas

Resmi

Ganguly

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Combined with the I-V curve, we found that the conductance or resistance of the device can be regulated by electrical signals, similar to the change of synaptic weight. This proves that our device has the potential to simulate the synaptic function, learning, and memory behaviors [27][28][29][30][31]. The synapse consists of presynaptic membrane, synaptic space, and postsynaptic membrane.…”

Section: Resultsmentioning

confidence: 63%

Artificial nociceptor based on TiO2 nanosheet memristor

et al. 2021

Self Cite

View full text Add to dashboard Cite

With the development of technology, the learning and memory functions of artificial memristor synapses are necessary for realizing artificial neural networks and neural neuromorphic computing. Owing to their high scalability performance, nanosheet materials have been widely employed in cellular-level learning, but the behaviors of nociceptor based on nanosheet materials have rarely been studied. Here, we present a memristor with an Al/TiO 2 /Pt structure. After electroforming, the memristor device showed a gradual conductance regulation and could simulate synaptic functions such as the potentiation and depression of synaptic weights. We also designed a new scheme that verifies the pain sensitization, desensitization, allodynia, and hyperalgesia behaviors of real nociceptors in the fabricated memristor. Memristors with these behaviors can significantly improve the quality of intelligent electronic devices. Data fitting showed that the high resistance and low resistance states were consistent with the hopping conduction mechanism. This work promises the application of TiO 2-based devices in next-generation neuromorphological systems.

show abstract

“…Because of the gradually change in current–voltage characteristics at both bias polarities of the Gd x O y memristors with H 2 plasma surface modified CSA graphene BEs, as shown in Figure 4a, a superior adoptability of the devices in artificial synapse is expected. [ 41,63,64 ] Consequently, the basic biomimetic characteristics of potentiation, depression, and spike‐timing‐dependent plasticity (STDP) for the Gd x O y memristors with CSA graphene BEs were measured and are presented in Figure . Figure 6a,b shows the potentiation and depression processes of the Gd x O y memristors with a pristine and a 10 min H 2 plasma surface modified CSA graphene BEs, respectively.…”

Section: Resultsmentioning

confidence: 99%

“…[ 39 ] It has been reported that if MLG with defects or carbon quantum dots were placed between the electrode and RS layer of memristors, the carbon atoms can drift into the RS layer to form carbon conductive filaments. [ 40,41 ] Nevertheless, these studies did not investigate the plasma treatment on graphene electrodes and discuss its influences on RS behaviors by redox reaction mechanism and biomimetic properties as well. As we know, chemical‐vapor‐deposited (CVD) graphene has been extensively used in a variety of fields, but it is too expensive to be applied in semiconductor manufacturing.…”

Section: Introductionmentioning

confidence: 99%

Compacted Self‐Assembly Graphene with Hydrogen Plasma Surface Modification for Robust Artificial Electronic Synapses of Gadolinium Oxide Memristors

Chan

et al. 2020

Adv Materials Inter

View full text Add to dashboard Cite

increasing the requirements of smart data computing and storage. However, for the conventional von Neumann computer architecture, owing to the separation between memory and computation functions, the data calculation rules and transmission speed between the central processing unit and memory segments limit the overall efficiency and computation time and dissipate a high processing power, which cannot satisfy the scenario of real-time applications. [1] By contrast, the human brain is a complex organ composed of ≈100 billion neurons and 60 trillion synapses with an average consumed power of ≈20 W and superior learning and cognitive abilities. [2] Therefore, an entirely new architecture concept of artificial intelligence (AI) has to be proposed to solve the limitations of latency and the issues of power consumption. For this purpose, deep learning is the core of AI based on the artificial neural networks, which is designed to mimic how the brain works. The neural networks consist of synapses, neurons, pulsed neural networks, perceptual systems, and so on. To simulate synapses, some nonvolatile memory devices have been proposed such as floating-gate (FG) flash memory, phase-change memory (PCM), memristor in resistive random-access memory (RRAM), and ferroic tunnel junctions. [3-6] Since the scaling of

show abstract

Designing carbon conductive filament memristor devices for memory and electronic synapse applications

Abstract: Utilizing the instability of the edge atoms of graphene defects, carbon conductive filaments were formed under the regulation of the electric field and the synaptic function was achieved.

Cited by 70 publications

References 68 publications

Programmable electronic synapse and nonvolatile resistive switches using MoS2 quantum dots

Programmable electronic synapse and nonvolatile resistive switches using MoS2 quantum dots

Artificial nociceptor based on TiO2 nanosheet memristor

Compacted Self‐Assembly Graphene with Hydrogen Plasma Surface Modification for Robust Artificial Electronic Synapses of Gadolinium Oxide Memristors

Contact Info

Product

Resources

About