Study And Application Of Novel Biomaterial Based Memristive Devices

The big data era has propelled the development of various fields but has also brought challenges on how to store and compute massive data.To better address this challenge and overcome the performance bottleneck of von Neumann architecture,a series of post-Moore storage and computing devices are urgently needed.As one of the candidates for high-performance post-Moore’s devices,the memristor has been extensively researched in fields such as neuromorphic computing and data storage due to its inherent advantages.The switching mechanisms of memristors based on various materials are different,resulting in certain differences in their electrical performance.Therefore,exploring the characteristics of different materials to fabricating high-performance memristors is still an important direction of research.Compared with inorganic materials,organic materials such as biomolecules exhibit advantages such as low cost,mechanical flexibility,biocompatibility,and degradability.Proteins,as a type of biomolecule,were often studied for the fabrication of memristor.Peptides,as components of proteins,have simpler and purer structures that are easier to analyze,and they can also be used as templates to reduce metal nanoparticles,giving materials more electrical properties.Currently,research on peptide-based memristive devices is almost nonexistent in China.Therefore,developing high-performance peptide-based memristors and exploring their application prospects are of great significance.Based on this,this paper designed and explored a novel resistive switching biomaterial memristor based on the amyloid peptide chain(hIAPP_20-29).The specific work was divided into the following three parts:Firstly,the possibility of introducing the amyloid polypeptide chain(hIAPP_20-29)material into memristors was discussed.On the one hand,its surface polarity can provide a channel for charge migration,which may meet the requirements of the resistive switching layer for memristor.On the other hand,it can be used as a template for the in-situ reduction of gold nanoparticles to prepare hybrid materials.The introduction of gold nanoparticles can promote the stable growth of conductive filaments,thereby obtaining better device performance.Multiple characterization methods were conducted to analyze the structure of hIAPP_20-29 and gold nanoparticle hybrid materials(Au-hIAPP_20-29 hybrids).The preparation method and process of the device were introduced detailly.Two types of memristor devices,Ag/hIAPP_20-29/Pt and Ag/Au-hIAPP_20-29 hybrids/Pt,were successfully prepared.Secondly,multiple electrical tests and analyses were carried out on the Ag/hIAPP_20-29/Pt and Ag/Au-hIAPP_20-29 hybrids/Pt devices.The results showed that the Ag/hIAPP_20-29/Pt device exhibited volatile unipolar threshold switching behavior.The switching behavior of the Ag/Au-hIAPP_20-29hybrids/Pt device can be regulated by compliance current.Under low current limits,it exhibits volatile unipolar threshold switching behavior,while under high current limits,it exhibits non-volatile bipolar resistive switching behavior.Moreover,compared with the Ag/hIAPP_20-29/Pt device,the Ag/Au-hIAPP_20-29 hybrids/Pt device has higher performance,including non-volatility,high switching ratio,more stable switching voltage,reliable cycling and holding characteristics.Meanwhile,detailed discussion was conducted on the conduction mechanism of the device based on the SCLC model.The resistance switching conduction model based on Ag conductive filaments was proposed.Finally,the application prospects of the prepared biomaterial memristor were investigated.Firstly,the device’s conductivity was altered through different pulse stimulations to simulate its interaction with biological synapses,resulting in the achievement of various synaptic functions,including STP to LTP conversion,EPSC,PPF,STDP,etc.This demonstrated the potential of the prepared device as an artificial synapse.Secondly,the existing simulation models for memristor circuits were investigated,and model units were established using the electrical characteristic curves obtained from device measurements for simulation in software.Various logic unit circuits were also constructed and simulated to explore the potential of the devices in building logic units.