二维材料的结构调控及其人工突触器件研究

随着AI浪潮席卷而来，芯片训练大模型对能耗的要求日益增长。以忆阻器为代表的神经仿生芯片，因其与神经突触以及神经元结构相仿，成为构建神经仿生芯片，以解决算力和能耗问题的基石。但目前忆阻器的应用面临着稳定性、制造复杂性、模型和理论的不完善、集成和兼容性问题。

二维材料作为阻变层的忆阻器有着低能耗、高开关速度、实现高密度集成的优势，但存在制备效率低、性能稳定性差等缺点。本研究通过红磷辅助的高能球磨法解理石墨，并辅以稀硝酸的进一步解理，成功构建了磷酸化石墨烯纳米片。利用该材料的良好水散性，设计共溶剂旋涂策略，构建了忆阻器阻变层。并构筑忆阻器，该器件有大于1000次直流循环的电流-电压（I-V）性能、高良品率和在10个器件中出色一致性。其由界面势垒产生的易失性特性，被用以模拟人工神经元，解决了传统碳基阻变层的均匀性和可重复性问题。

进一步地，我们通过离子配位阻变层和电极材料的更换，将其转化为非易失性忆阻器，展现了集感知、存储与计算于一体的能力，其表现出优异的仿生性能，为模拟神经突触提供了独特优势，实现了碳基材料的电导精确调控。

在机理分析方面，本文利用导电原子力显微镜（C-AFM）深入探讨了器件的工作原理。得益于阻变层的稳定性和均匀性，利用原子力显微镜的探针尖端，构建了微观尺度的器件，提供了碳基阻变层的导电模型、纳米尺度的解决方案，同时开发了具有出色抗弯曲性能和寿命的柔性忆阻器。

最后，本研究通过对神经突触数据的仿真学习，建立了人工神经网络（ANN），并成功展现了优越的类脑芯片性能。

With the wave of AI coming, the demand for energy consumption of chips training model is increasing day by day. Neuromimetic chips have emerged as the choice for solving arithmetic and energy problems. Because of the structural similarity of the memristor to the synapse and to the neuron. It becomes an excellent candidate for a neuromimetic chip. However, the application of memristors faces problems of stability, manufacturing complexity, imperfections in models and theories, integration and compatibility at present.

Memristors featuring two-dimensional (2D) materials as resistive layers offer advantages such as low energy consumption, rapid switching, and high-density integration. Nevertheless, they are challenged by inefficient production and unstable performance. In this study, we synthesized phosphorylated graphene nanosheets (phos-GPs) by high-energy ball milling of red phosphorus and graphite, followed by mixing with dilute nitric acid. To leverage the superior water diffusivity of this material, we developed a co-solvent spin-coating method to create the resistive layer. This approach enabled the construction of a memristor that exhibited robust Current-Voltage (I-V) performance over 1000 DC cycles, with high yields and excellent consistency across 10 devices. The volatility of this device supports the emulation of neuronal behavior, addressing issues of uniformity and reproducibility commonly encountered with conventional carbon-based resistive layers.

Moreover, modifying the memristor's properties through ion coordination within the resistive layer and replacing the electrodes transitions the device from volatile to non-volatile. This modification facilitates the integration of sensing, storage, and computation, while demonstrating exceptional biomimetic characteristics suitable for simulating synapses. Additionally, precise modulation of conductivity in carbon-based materials was successfully achieved.

In the principle analysis, we provide insights into the device's working mechanism using a conductive Atomic Force Microscope (C-AFM). Benefiting from the stability and uniformity of the resistive layer, we constructed microscale devices using the AFM probe tip. We also modeled the conductivity of carbon-based resistive change layers, offering nano-scale solutions. Additionally, we developed a flexible memristor characterized by excellent bending resistance and extended lifespan.

Lastly, we established an Artificial Neural Network (ANN) that learned from the simulation of synaptic data, successfully demonstrating the superior performance of the bionic chip.

[

[1] VON NEUMANN J. First draft of a report on the EDVAC [J]. IEEE Annals of the History of Computing, 1993, 15(4): 27-75.

[2] ZENG S, TANG Z, LIU C, et al. Electronics based on two-dimensional materials: Status and outlook [J]. Nano Research, 2021, 14: 1752-67.

[3] SCHNEIDER M L, DONNELLY C A, RUSSEK S E, et al. Ultralow power artificial synapses using nanotextured magnetic Josephson junctions [J]. Science Advances, 2018, 4(1): e1701329.

[4] PEREDA A E. Electrical synapses and their functional interactions with chemical synapses [J]. Nature Reviews Neuroscience, 2014, 15(4): 250-63.

[5] STRUKOV D B, SNIDER G S, STEWART D R, et al. The missing memristor found [J]. Nature, 2008, 453(7191): 80-3.

[6] 马可. 基于晶体管人工突触的光电调控的研究 [D]; 南京邮电大学.

[7] BAI B, SHU H, WANG X, et al. Towards silicon photonic neural networks for artificial intelligence [J]. Science China Information Sciences, 2020, 63: 1-14.

[8] ZHUGE X, WANG J, ZHUGE F. Photonic synapses for ultrahigh‐speed neuromorphic computing [J]. Physica Status Solidi (RRL)–Rapid Research Letters, 2019, 13(9): 1900082.

[9] WANG G, WANG R, KONG W, et al. Simulation of retinal ganglion cell response using fast independent component analysis [J]. Cognitive Neurodynamics, 2018, 12: 615-24.

[10] PARK H L, LEE Y, KIM N, et al. Flexible neuromorphic electronics for computing, soft robotics, and neuroprosthetics [J]. Advanced Materials, 2020, 32(15): 1903558.

[11] LE V-Q, DO T-H, RETAMAL J R D, et al. Van der waals heteroepitaxial AZO/NiO/AZO/muscovite (ANA/muscovite) transparent flexible memristor [J]. Nano Energy, 2019, 56: 322-9.

[12] CHEN F-F, ZHU Y-J, CHEN F, et al. Fire alarm wallpaper based on fire-resistant hydroxyapatite nanowire inorganic paper and graphene oxide thermosensitive sensor [J]. ACS Nano, 2018, 12(4): 3159-71.

[13] SHANG J, LIU G, YANG H, et al. Thermally stable transparent resistive random access memory based on all‐oxide heterostructures [J]. Advanced Functional Materials, 2014, 24(15): 2171-9.

[14] YE L, GAO Z, FU J, et al. Overview of memristor-based neural network design and applications [J]. Frontiers in Physics, 2022, 10: 839243.

[15] CAO Z, SUN B, ZHOU G, et al. Memristor-based neural networks: a bridge from device to artificial intelligence [J]. Nanoscale Horizons, 2023, 8(6): 716-745.

[16] XU W, WANG J, YAN X. Advances in memristor-based neural networks [J]. Frontiers in Nanotechnology, 2021, 3: 645995.

[17] YAO P, WU H, GAO B, et al. Fully hardware-implemented memristor convolutional neural network [J]. Nature, 2020, 577(7792): 641-6.

[18] WANG Z, JOSHI S, SAVEL’EV S, et al. Fully memristive neural networks for pattern classification with unsupervised learning [J]. Nature Electronics, 2018, 1(2): 137-45.

[19] PARK S-O, JEONG H, PARK J, et al. Experimental demonstration of highly reliable dynamic memristor for artificial neuron and neuromorphic computing [J]. Nature Communications, 2022, 13(1): 2888.

[20] KUMAR S, WANG X, STRACHAN J P, et al. Dynamical memristors for higher-complexity neuromorphic computing [J]. Nature Reviews Materials, 2022, 7(7): 575-91.

[21] ZIDAN M A, STRACHAN J P, LU W D. The future of electronics based on memristive systems [J]. Nature Electronics, 2018, 1(1): 22-9.

[22] CHOI S, JANG S, MOON J-H, et al. A self-rectifying TaOy/nanoporous TaOx memristor synaptic array for learning and energy-efficient neuromorphic systems [J]. NPG Asia Materials, 2018, 10(12): 1097-106.

[23] KRISHNAPRASAD A, DEV D, SHAWKAT M S, et al. Graphene/MoS2/SiOx memristive synapses for linear weight update [J]. 2D Materials and Applications, 2023, 7(1): 22.

[24] ZHANG Y, MAO G-Q, ZHAO X, et al. Evolution of the conductive filament system in HfO2-based memristors observed by direct atomic-scale imaging [J]. Nature Communications, 2021, 12(1): 7232.

[25] STATHOPOULOS S, KHIAT A, TRAPATSELI M, et al. Multibit memory operation of metal-oxide bi-layer memristors [J]. Scientific Reports, 2017, 7(1): 17532.

[26] WEDIG A, LUEBBEN M, CHO D-Y, et al. Nanoscale cation motion in TaOx, HfOx and TiO x memristive systems [J]. Nature Nanotechnology, 2016, 11(1): 67-74.

[27] SANGWAN V K, JARIWALA D, KIM I S, et al. Gate-tunable memristive phenomena mediated by grain boundaries in single-layer MoS2 [J]. Nature Nanotechnology, 2015, 10(5): 403-6.

[28] BESSONOV A A, KIRIKOVA M N, PETUKHOV D I, et al. Layered memristive and memcapacitive switches for printable electronics [J]. Nature Materials, 2015, 14(2): 199-204.

[29] QIAN K, TAY R Y, NGUYEN V C, et al. Hexagonal boron nitride thin film for flexible resistive memory applications [J]. Advanced Functional Materials, 2016, 26(13): 2176-84.

[30] FANG Y, ZHAI S, CHU L, et al. Advances in halide perovskite memristor from lead-based to lead-free materials [J]. ACS Applied Materials & Interfaces, 2021, 13(15): 17141-57.

[31] YANG J-Q, WANG R, WANG Z-P, et al. Leaky integrate-and-fire neurons based on perovskite memristor for spiking neural networks [J]. Nano Energy, 2020, 74: 104828.

[32] JOHN R A, SHAH N, VISHWANATH S K, et al. Halide perovskite memristors as flexible and reconfigurable physical unclonable functions [J]. Nature Communications, 2021, 12(1): 3681.

[33] ZHANG T, WANG L, DING W, et al. Rationally designing high-performance versatile organic memristors through molecule-mediated ion movements [J]. Advanced Materials, 2023, 35(40): 2302863.

[34] ZHAO Z, EL-KHOULY M E, CHE Q, et al. Redox-active azulene-based 2D conjugated covalent organic framework for organic memristors [J]. Angewandte Chemie International Edition, 2023, 62(7): 202217249.

[35] SUN B, NGAI J H, ZHOU G, et al. Voltage-controlled conversion from CDS to MDS in an azobenzene-based organic memristor for information storage and logic operations [J]. ACS Applied Materials & Interfaces, 2022, 14(36): 41304-15.

[36] HUH W, LEE D, LEE C H. Memristors based on 2D materials as an artificial synapse for neuromorphic electronics [J]. Advanced Materials, 2020, 32(51): 2002092.

[37] ZHANG X, ZHAO X, SHAN X, et al. Humidity effect on resistive switching characteristics of the CH3NH3PbI3 memristor [J]. ACS Applied Materials & Interfaces, 2021, 13(24): 28555-63.

[38] PRUDNIKOV N V, LAPKIN D A, EMELYANOV A V, et al. Associative STDP-like learning of neuromorphic circuits based on polyaniline memristive microdevices [J]. Journal of Physics D: Applied Physics, 2020, 53(41): 414001.

[39] LUO X, MING J, GAO J, et al. Low-power flexible organic memristor based on PEDOT: PSS/pentacene heterojunction for artificial synapse [J]. Frontiers in Neuroscience, 2022, 16: 1016026.

[40] YUAN L, LIU S, CHEN W, et al. Organic memory and memristors: from mechanisms, materials to devices [J]. Advanced Electronic Materials, 2021, 7(11): 2100432.

[41] SANGWAN V K, HERSAM M C. Neuromorphic nanoelectronic materials [J]. Nature Nanotechnology, 2020, 15(7): 517-28.

[42] GOI E, ZHANG Q, CHEN X, et al. Perspective on photonic memristive neuromorphic computing [J]. PhotoniX, 2020, 1: 1-26.

[43] YOUNGBLOOD N, RíOS OCAMPO C A, PERNICE W H, et al. Integrated optical memristors [J]. Nature Photonics, 2023, 17(7): 561-72.

[44] SUNG S H, KIM T J, SHIN H, et al. Simultaneous emulation of synaptic and intrinsic plasticity using a memristive synapse [J]. Nature Communications, 2022, 13(1): 2811.

[45] DU N, ZHAO X, DI VENTRA M, et al. Synaptic plasticity in memristive artificial synapses and their robustness against noisy inputs [J]. Frontiers in Neuroscience, 2021, 15: 660894.

[46] GAO S, LIU G, YANG H, et al. An oxide schottky junction artificial optoelectronic synapse [J]. ACS Nano, 2019, 13(2): 2634-42.

[47] KUMAR M, ABBAS S, KIM J. All-oxide-based highly transparent photonic synapse for neuromorphic computing [J]. ACS Applied Materials & Interfaces, 2018, 10(40): 34370-6.

[48] SHRIVASTAVA S, KEONG L B, PRATIK S, et al. Fully photon controlled synaptic memristor for neuro‐inspired computing [J]. Advanced Electronic Materials, 2023, 9(3): 2201093.

[49] DANG B, LIU K, WU X, et al. One‐phototransistor–one‐memristor array with high‐linearity light‐tunable weight for optic neuromorphic computing [J]. Advanced Materials, 2023, 35(37): 2204844.

[50] HUANG W, HANG P, WANG Y, et al. Zero-power optoelectronic synaptic devices [J]. Nano Energy, 2020, 73: 104790.

[51] ZHOU L, ZHANG S-R, YANG J-Q, et al. A UV damage-sensing nociceptive device for bionic applications [J]. Nanoscale, 2020, 12(3): 1484-94.

[52] CAO F, HU Z, YAN T, et al. A Dual‐Functional Perovskite‐Based photodetector and memristor for visual memory [J]. Advanced Materials, 2023, 35(44): 2304550.

[53] TAN H, NI Z, PENG W, et al. Broadband optoelectronic synaptic devices based on silicon nanocrystals for neuromorphic computing [J]. Nano Energy, 2018, 52: 422-30.

[54] WANG Y, YIN L, HUANG S, et al. Silicon-nanomembrane-based broadband synaptic phototransistors for neuromorphic vision [J]. Nano Letters, 2023, 23(18): 8460-7.

[55] NOVOSELOV K S, GEIM A K, MOROZOV S V, et al. Electric field effect in atomically thin carbon films [J]. Science, 2004, 306(5696): 666-9.

[56] CAO Y, FATEMI V, DEMIR A, et al. Correlated insulator behaviour at half-filling in magic-angle graphene superlattices [J]. Nature, 2018, 556(7699): 80-4.

[57] CAO Y, FATEMI V, FANG S, et al. Unconventional superconductivity in magic-angle graphene superlattices [J]. Nature, 2018, 556(7699): 43-50.

[58] URI A, GROVER S, CAO Y, et al. Mapping the twist-angle disorder and Landau levels in magic-angle graphene [J]. Nature, 2020, 581(7806): 47-52.

[59] CAO Y, RODAN-LEGRAIN D, RUBIES-BIGORDA O, et al. Tunable correlated states and spin-polarized phases in twisted bilayer–bilayer graphene [J]. Nature, 2020, 583(7815): 215-20.

[60] LEE G-H, YU Y-J, CUI X, et al. Flexible and transparent MoS2 field-effect transistors on hexagonal boron nitride-graphene heterostructures [J]. ACS Nano, 2013, 7(9): 7931-6.

[61] YIN Z, LI H, LI H, et al. Single-layer MoS2 phototransistors [J]. ACS Nano, 2012, 6(1): 74-80.

[62] XUE F, HE X, RETAMAL J R D, et al. Gate-tunable and multidirection-switchable memristive phenomena in a van der waals ferroelectric [J]. Advanced Materials, 2019, 31(29): 1901300.

[63] SAMY O, ZENG S, BIROWOSUTO M D, et al. A review on MoS2 properties, synthesis, sensing applications and challenges [J]. Crystals, 2021, 11(4): 355.

[64] SHEN Y, ZHENG W, ZHU K, et al. Variability and yield in h‐BN‐based memristive circuits: the role of each type of defect [J]. Advanced Materials, 2021, 33(41): 2103656.

[65] PAN C, JI Y, XIAO N, et al. Coexistence of grain-boundaries-assisted bipolar and threshold resistive switching in multilayer hexagonal boron nitride [J]. Advanced Functional Materials, 2017, 27(10): 1604811.

[66] CHEN S, MAHMOODI M R, SHI Y, et al. Wafer-scale integration of two-dimensional materials in high-density memristive crossbar arrays for artificial neural networks [J]. Nature Electronics, 2020, 3(10): 638-45.

[67] MAO J Y, WU S, DING G, et al. A van der waals integrated damage-free memristor based on layered 2D hexagonal boron nitride [J]. Small, 2022, 18(12): 2106253.

[68] WANG M, CAI S, PAN C, et al. Robust memristors based on layered two-dimensional materials [J]. Nature Electronics, 2018, 1(2): 130-6.

[69] ZHU X, LI D, LIANG X, et al. Ionic modulation and ionic coupling effects in MoS2 devices for neuromorphic computing [J]. Nature Materials, 2019, 18(2): 141-8.

[70] RANGANATHAN K, FIEGENBAUM‐RAZ M, ISMACH A. Large‐Scale and robust multifunctional vertically aligned MoS2 photo‐memristors [J]. Advanced Functional Materials, 2020, 30(51): 2005718.

[71] HAO C, WEN F, XIANG J, et al. Liquid‐exfoliated black phosphorous nanosheet thin films for flexible resistive random access memory applications [J]. Advanced Functional Materials, 2016, 26(12): 2016-24.

[72] KUMAR D, LI H, DAS U K, et al. Flexible Solution‐processable black-phosphorus-based optoelectronic memristive synapses for neuromorphic computing and artificial visual perception applications [J]. Advanced Materials, 2023, 35(28): 2300446.

[73] MA H, FANG H, XIE X, et al. Optoelectronic synapses based on MXene/Violet Phosphorus van der waals heterojunctions for visual-olfactory crossmodal perception [J]. Nano-Micro Letters, 2024, 16(1): 1-15.

[74] ZENG J, FENG G, WU G, et al. Multisensory ferroelectric semiconductor synapse for neuromorphic computing [J]. Advanced Functional Materials, 2024: 2313010. 

[75] YOON C, OH G, KIM S, et al. Implementation of threshold-and memory-switching memristors based on electrochemical metallization in an identical ferroelectric electrolyte [J]. NPG Asia Materials, 2023, 15(1): 33.

[76] YANG Y, GAO P, GABA S, et al. Observation of conducting filament growth in nanoscale resistive memories [J]. Nature Communications, 2012, 3(1): 732.

[77] YANG R, HUANG H M, HONG Q H, et al. Synaptic suppression triplet‐STDP learning rule realized in second-order memristors [J]. Advanced Functional Materials, 2018, 28(5): 1704455.

[78] RAO J, FAN Z, HONG L, et al. An electroforming-free, analog interface-type memristor based on a SrFeOx epitaxial heterojunction for neuromorphic computing [J]. Materials Today Physics, 2021, 18: 100392.

[79] CHA J-H, YANG S Y, OH J, et al. Conductive-bridging random-access memories for emerging neuromorphic computing [J]. Nanoscale, 2020, 12(27): 14339-68.

[80] REN Y, SUN R, CHEN S H Y, et al. Exploring Phase‐Change memory: from material systems to device physics [J]. Physica Status Solidi (RRL)–Rapid Research Letters, 2021, 15(3): 2000394.

[81] LIU Z-C, WANG L. Applications of phase change materials in electrical regime from conventional storage memory to novel neuromorphic computing [J]. IEEE Access, 2020, 8: 76471-99.

[82] ZHOU H, SORKIN V, CHEN S, et al. Design‐Dependent switching mechanisms of schottky‐barrier‐modulated memristors based on 2D semiconductor [J]. Advanced Electronic Materials, 2023, 9(6): 2201252.

[83] LECUN Y, BENGIO Y, HINTON G. Deep learning [J]. Nature, 2015, 521(7553): 436-44.

[84] ZHANG A, ZHOU H, LI X, et al. Fast and robust learning in spiking feed-forward neural networks based on intrinsic plasticity mechanism [J]. Neurocomputing, 2019, 365: 102-12.

[85] KANG T S, BANERJEE A. Learning feedforward and recurrent deterministic spiking neuron network feedback controllers [J]. arXiv preprint arXiv:170802603, 2017.

[86] ZUO W, ZHU Q, FU Y, et al. Volatile threshold switching memristor: An emerging enabler in the AIoT era [J]. Journal of Semiconductors, 2023, 44(5): 053102.

[87] PALMER L M, SHAI A S, REEVE J E, et al. NMDA spikes enhance action potential generation during sensory input [J]. Nature neuroscience, 2014, 17(3): 383-90.

[88] UJFALUSSY B B, MAKARA J K, LENGYEL M, et al. Global and multiplexed dendritic computations under in vivo-like conditions [J]. Neuron, 2018, 100(3): 579-92.

[89] YANG J J, STRUKKOV D B, STTEWART D R, et al. Memristive devices for computing [J]. Nature nanotechnology, 2013, 8(1): 13-24.

[90] YUAN R, TIW P J, CAI L, et al. A neuromorphic physiological signal processing system based on VO2 memristor for next-generation human-machine interface [J]. Nature Communications, 2023, 14(1): 3695.

[91] HUANG H M, YANG R, TAN Z H, et al. Quasi‐hodgkin‐huxley neurons with leaky integrate‐and‐fire functions physically realized with memristive devices [J]. Advanced Materials, 2019, 31(3): 1803849.