Study On The Resistive Switching Behaviors And Neural Morphological Functions Of The Oxide-based Memristors

As the fourth passive electronic device besides resistance,capacitance and inductance,memristor has been developed tremendously in non-volatile memory,intelligent logic operation and computer image processing since it was proposed by Leon in 1971.Compared with the active devices,such as MOSFET transistors,BJTs,etc.,memristors not only have the merits of small size,simple structure,fast response speed,and low power consumption,but also possess the unique storage and calculation functions.More importantly,the traditional von Neumann computer architecture can be broken by applying the integration of memory and calculation memristor.In summary,the memristor has become one of the indispensable elements in the next generation of computer systems.Recently,the application of memristors is mainly implemented in Resistive Random Access Memory(RRAM)and artificial synapses.At this stage,with low capacity and low storage density in RRAM,and low synaptic expansion functionality in artificial synapses,the practical application of memristors have been severely limited.In order to solve the above technical problems,researchers have proposed a series of effective solutions.For example,in the resistance state transition operation of RRAM,adjusting the set voltage applied to the device or compliance current to achieve multi-resistance in a single device state transition.And when operating electronic synapses,third-party parameters are introduced to realize the functions of high-level biological synapses such as vision,hearing,and smell.In order to solve the above-mentioned bottleneck more effectively,this research prepared high-performance multi-value RRAM and advanced optoelectronic artificial synapse devices based on oxide materials.Based on the researchers'existing technical means and improved methods,the resistive effect and neuromorphic behavior of oxide-based memristors have been further expanded and improved,and the physical mechanism that causes its resistive behavior has been studied in detail.The specific contents are as follows:Chapter 1:Firstly,it offers a briefly introduction about the structure,resistive switching materials and resistive switching mechanism.Then we introduce the application method and research significance of the memristor in detail.Chapter 2:Starting from the process of preparing memristive devices,several methods for preparing devices are systematically depicted,including some typical chemical methods and physical methods.By comparison,the advantages and disadvantages of the preparation methods are described.Finally,the test equipment and test process of the memristor characteristic are introduced.Chapter 3:A facile chemical solution method was applied to prepare non-toxic and harmless zinc oxide quantum dots(ZnO QDs)in the ethanol phase,and spin-coated them on the polymer PMMA as a trap trap layer to construct PMMA/ZnO QDs/PMMA multilayers-based resistive random access memory(RRAM).By adjusting the concentration of ZnO QDs,the effect of trap density on the ON/OFF ratio of the device was studied.In addition,a physical model that promotes the formation of conductive filaments by a local electric field is proposed.Moreover,the physical mechanism of the resistive switching behaviors is explained in detail.Most importantly,by adjusting the compliance current imposed on the device,the multi-resistance state transition in a single device is realized,which solves the requirements for large-scale,high-density storage in the era of big data.Chapter 4:A non-stoichiometric titanium oxynitride(TiN_xO_2-x)with abundant oxygen defects was selected as the resistive active layer,and combined with the two-dimensional material Mo S₂ to construct a new type of photoelectron synaptic device based on type-II heterojunction.By applying electrical and optical signals,the migration behavior of oxygen vacancies in the resistive active layer is regulated,thereby simulating the various functions and behaviors of human synapses.More importantly,benefiting from the sensitivity of the device to light signals,we fabricated the device into an array to simulate the visual behavior of the human body,and successfully realized the functions of the human visual system,including the perception and memory of different light intensity signals from the outside world.This device provides a new solution for realizing novel optoelectronic neural computing and building artificial visual neural network.Chapter 5:The work of this thesis is summarized,and the future development trend and advanced application fields of memristors are prospected.