Research On Organic Transistor Neuromorphic Structure Based On Polymer Ferroelectric/electre

There are billions of neurons in the brain,and their interaction patterns generate our consciousness,emotions,and behavior.Among them,the synapse is the most important information channel,that is,the specific connection between two nerve cells,through which signals can complete the interaction between nerve cells.Inspired by biological neuromorphic systems,neuromorphic devices can simultaneously perceive,memorize,and process massive amounts of information through parallel and efficient processes,demonstrating enormous potential in overcoming the Van Neumann bottleneck in neuromorphic computing.Most of the existing p-type organic transistors used ferroelectric mechanism and charge capture mechanism to simulate neural synaptic function.Due to different mechanisms,the transfer characteristic curve of bidirectional scanning in p-type ferroelectric transistors generally exhibits clockwise hysteresis,while the transfer characteristic curve of bidirectional scanning in p-type charge capture transistors generally exhibits counterclockwise hysteresis.In terms of simulating neural synaptic function,on the p-type ferroelectric transistor,the response current after pulse stimulation shows the same trend as that of pulse stimulation response.On the other hand,on p-type charge capture transistors,the response current after pulse stimulation shows a reverse trend compared to the current response to pulse stimulation.In this study,in order to achieve the goal of reaching a higher level of synaptic function,the two mechanisms are combined based on the p-type transistor,which can simulate not only the unique synaptic behavior of the two mechanisms,but also other complex synaptic behavior.This master’s thesis is based on the research results of organic transistors,on the basis of in-depth understanding of ferroelectric polarization effect and charge capture effect,combined with the advantages of process preparation,such as adjusting device performance with the introduction of variables(new characteristics)of materials,taking ferroelectric transistors and transistors combined with ferroelectric effect and charge capture effect as research objects,taking device performance optimization and regulation of multiple plasticity as research perspectives,through adjusting device carrier process a series of organic neural morphology transistors have been successfully constructed through strategies such as device structure,exploration of new mechanisms,introduction of parameterized evaluation of neural morphology performance,and various regulatory signals,aiming to provide the best solution for simulating complex neural systems in living organisms.The main research content and achievements are as follows:1.Conduct research on ferroelectric organic transistors.Firstly,a thin film of organic ferroelectric material(P(VDF-Tr FE-CFE))was prepared,and physical analysis methods were used to analyze the excellent ferroelectric crystallization of the prepared P(VDFTr FE-CFE)thin film(conforming to the rod-shaped structure of ferroelectric crystallization and typical ferroelectric crystallization diffraction peak results).On this basis,a p-type ferroelectric transistor was constructed using P(VDF-Tr FE-CFE)as the gate dielectric layer,and its simulated neural morphology performance was evaluated.Finally,the improvement direction of the device is explored,which is to reduce the influence of ferroelectric polarization fluctuation and improve the quality of semiconductor crystal.2.In order to further tailor the simulated neural morphology performance of ferroelectric transistors to the biological neural performance,different thicknesses of alumina layers were used as passivation layers between the P(VDF-Tr FE-CFE)layer and the semiconductor layer to improve device performance.The improvement of performance parameters such as switch ratio,storage window,and mobility indicates that adjusting the thickness of the alumina layer plays an important role in improving device performance.By using the method of evaluating neural morphological performance parameters,the optimal thickness device is selected.On this basis,a series of explorations on neural morphological properties were completed.Finally,the neural morphology behavior of the device under pulse stimulation with adjustable conductivity characteristics was studied,and the simulation of artificial neural network functions was achieved.3.In order to conduct dynamic neural morphology simulation,electret materials with charge capture function were selected to replace the alumina layer,so that the device has additional electrical properties(having properties opposite to pure ferroelectric transistors).The thickness of the P(VDF-Tr FE-CFE)layer has been adjusted to achieve a small range of voltage operation.On this basis,the new neural morphological characteristics(derived from charge capture)coexist with the old neural morphological characteristics(derived from ferroelectric polarization).Furthermore,a comprehensive exploration was conducted on the neural morphology of the new and old characteristics of the device.Finally,by combining the new neural morphology function with the old neural morphology function,a higher level of neural morphology is generated,making the operating model of the device closer to the actual biological neural operating model,and achieving the simulation of the temperature perception and regulation system in the sympathetic nervous system.