Font Size: a A A

Study On The Characteristics And Neurosynaptic Biomimetic Of Zirconia/Two-dimensional Chalcogenide Resistance Random Access Memory

Posted on:2021-04-21Degree:MasterType:Thesis
Country:ChinaCandidate:C Y QinFull Text:PDF
GTID:2381330623476428Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
Memristor refers to a device that implements a controllable change in resistance by applying voltage and current signals,and has storage and calculation functions.In the memory application direction,an important application of memristors is resistance random access memory?RRAM?.RRAM has the characteristics of non-volatile,fast speed,high integration,simple structure and other advantages,and has the potential to start the next memory era.However,in practical applications,the RRAM has a problem of insufficient reliability,that is,the voltage dispersion during set and reset is large,and the distribution is not uniform.Therefore,seeking new materials and new structures to improve device reliability is our main direction at present.In addition,in the direction of neural bionics,how to realistically simulate synaptic performance needs to be improved.In order to improve the performance of the above RRAM and research needs in the field of electronic artificial synapses,the main research contents of this subject are as follows:Firstly,a two-dimensional?2D?tungsten disulfide?WS2?nanosheet layer was inserted under a traditional RRAM dielectric layer zirconia?ZrO2?to construct a RRAM having an Ag/ZrO2/WS2/Pt structure.Under the DC scanning voltage of the test instrument,the Ag/ZrO2/WS2/Pt device exhibits the reversible behavior of a bipolar resistive switching?RS?.Compared with two single-layer resistive switching devices?Ag/WS2/Pt and Ag/ZrO2/Pt?,the Ag/ZrO2/WS2/Pt devices have a significantly reduced switching voltage?0.16 V/-0.06 V?,a more concentrated voltage distribution range?on:3.75%;off:13.33%?,and a greatly decreasing dispersion.Meanwhile,the endurance?109?,retention time?4×104 s?,switching speed?on:10ns;off:18 ns?and other aspects are relatively optimized.In addition,this topic explores the RS mechanism of the double-layer ZrO2/WS2 device.It is believed that the reason for the significant improvement in performance is that the fracture/recovery of the conductive filament is limited to the double-layer interface region.Secondly,in the research direction of improving the device function by combining 2D WS2 with ZrO2,we have selected 2D Group IV sulfide GeS for research,which has a larger forbidden area band width and smaller effective mass than WS2,and a RRAM with Ag/ZrO2/GeS/Pt structure was prepared.This device exhibits reversible bipolar resistance switching behavior.Compared with single-layer Ag/ZrO2/Pt and Ag/GeS/Pt RRAM,this double-layer RRAM has the effect of improving electrical performance,including a significantly reduced on/off threshold voltage and a significantly concentrated voltage distribution.Meanwhile,the data retention capability and the switching speed of the double-layer RRAM device are relatively improved.In addition,comparing the Ag/ZrO2/GeS/Pt device here with the Ag/ZrO2/WS2/Pt device,the threshold voltage is further reduced?on:0.10 V?,and the thermal stability is better?normal operation at 180oC?.Thirdly,the synaptic plasticity of Ag/ZrO2/WS2/Pt and Ag/ZrO2/GeS/Pt memristor devices were tested and simulated.These two new memristor devices can successfully perform two-way progressive regulation of resistance under the action of pulse voltage,respectively.Then a variety of synaptic functions were simulated,such as spiking-timing-dependent plasticity?STDP?and paired-pulse facilitation?PPF?.Finally,we take the Ag/ZrO2/WS2/Pt memristive device as an example,use its bidirectional progressive regulation data,and combine it with Python to simulate handwritten digital recognition system.In the end,the device realized analog image recognition of 0-9 ten handwritten digits.This research work demonstrates that the new memristive devices we have good synaptic plasticity and can be used as artificial synaptic electronic bionic devices to further promote the development of synaptic bionic devices.
Keywords/Search Tags:Memristor, RRAM, WS2 nanosheet, GeS nanosheet, Neural biomimetic
PDF Full Text Request
Related items